Compare commits

...

895 Commits

Author SHA1 Message Date
Johann
0c0a05046d Release v1.6.1 Long Tailed Duck
Change-Id: If27447472417c7ed34238295427ddb9da0561725
2017-01-12 12:27:27 -08:00
Johann Koenig
cabc29ba24 Merge "Add mips dspr2 partial idct tests" 2017-01-09 19:49:02 +00:00
Johann Koenig
8a7847c2c9 Merge "Fix mips dspr2 idct32x32 functions for large coefficient input" 2017-01-09 19:47:47 +00:00
Johann Koenig
bf168b24f5 Merge "Fix mips dspr2 idct16x16 functions for large coefficient input" 2017-01-09 19:47:00 +00:00
Johann Koenig
08d0a7fd0f Merge "Fix mips dspr2 idct8x8 functions for large coefficient input" 2017-01-09 19:46:18 +00:00
Johann Koenig
ab20869221 Merge "Fix mips dspr2 idct4x4 functions for large coefficient input" 2017-01-09 19:45:54 +00:00
Johann Koenig
7b18202e74 Merge "Add mips dspr2 vp9 intrapred tests" 2017-01-09 19:39:13 +00:00
Johann Koenig
9af97fb630 Merge "postproc: vpx_mbpost_proc_across_ip_neon" 2017-01-09 18:17:26 +00:00
Marco Paniconi
ebe0b57c91 Merge "vp9: 1 pass cbr mode: increase threshold for gf_cbr_boost_pct usage." 2017-01-09 17:23:12 +00:00
Kaustubh Raste
6377f9d966 Add mips dspr2 partial idct tests
Change-Id: Idf4003ea6f9a2a42a9f26e156bee73697acb7a37
2017-01-09 17:30:16 +05:30
Kaustubh Raste
50dd3eb62c Fix mips dspr2 idct32x32 functions for large coefficient input
Change-Id: If9da7099f226a27a09cc9e2899eb66a1158909d2
2017-01-09 17:21:09 +05:30
Kaustubh Raste
c06991fce6 Fix mips dspr2 idct16x16 functions for large coefficient input
Change-Id: I9be3d3d040837f658c6314606e28db8c31092a1a
2017-01-09 16:35:28 +05:30
Kaustubh Raste
24d804f79c Fix mips dspr2 idct8x8 functions for large coefficient input
Change-Id: If011dd923bbe976589735d5aa1c3167dda1a3b61
2017-01-09 16:22:19 +05:30
Kaustubh Raste
afd2d797eb Fix mips dspr2 idct4x4 functions for large coefficient input
Change-Id: I06730eec80ca81e0b7436d26232465b79f447e89
2017-01-09 15:28:30 +05:30
Kaustubh Raste
c6ccd1e939 Add mips dspr2 vp9 intrapred tests
Change-Id: I6be8c59ee220af0597bc2d7213f2779ac2e88db9
2017-01-09 14:11:57 +05:30
Hui Su
c7e2bd6298 Merge "Add support for VP9 level targeting" 2017-01-07 00:55:41 +00:00
Johann
4dca923454 postproc: vpx_mbpost_proc_across_ip_neon
The speedup is pretty poor. I would be concerned except the SSE2 is
worse:
Existing SSE2 improvement: 22%
New neon improvement: 35%

BUG=webm:1320

Change-Id: Ied598a261134aa6cbe69f96f58589d2bae17bf62
2017-01-06 16:39:17 -08:00
Marco
f1909d26f8 vp9: 1 pass cbr mode: increase threshold for gf_cbr_boost_pct usage.
Increase the boost threshold below which GOLDEN update will use same
rate correction factor as INTER_NORMAL.

Improves performance when gf_cbr_boost_pct is set (between 0 and 100)
in CBR mode.

Change-Id: I9f54cc18664786a100b13a416b7137ae03bd0cab
2017-01-06 15:37:10 -08:00
Jerome Jiang
316071d79c Merge "vp9: Enable more aggresive short circuit for speed 8." 2017-01-06 22:38:40 +00:00
Marco Paniconi
b632626ec0 Merge "vp9: Add some controls to sample encoder: vpx_temporal_svc_encoder" 2017-01-06 22:34:49 +00:00
Jerome Jiang
b87ebd7af8 Merge "vp9: Compute source sad for every superblock when partition copy is on." 2017-01-06 21:57:27 +00:00
Marco
bf5cdbdf9d vp9: Add some controls to sample encoder: vpx_temporal_svc_encoder
Add the gf boost and frame_parallel controls.
Set as default to off.

Change-Id: Id85fcb16a4fae97f51c09e9ebadb5cdcd510c2f5
2017-01-06 11:34:04 -08:00
Jerome Jiang
267e73446c vp9: Enable more aggresive short circuit for speed 8.
Set short_circuit_low_temp_var to 3 for speed 8 for all res.
No strong visual difference on all clips.

Change-Id: Ia6d9a314291ab1c14d5421bbdd769974083aeb2a
2017-01-06 10:23:34 -08:00
hui su
337ad83e58 Add support for VP9 level targeting
Constraints on encoder config:
-target_bandwidth is no larger than 80% of level bitrate limit
-target_bandwidth * (1 + max_over_shoot_pct) is no larger than
88% of level bitrate limit
-min_gf_interval is no smaller than level limit
-tile_columns is no larger than level limit

Constraints on rate control:
-current frame size plus previous three frames' size is no larger
than the CPB level limit
-current frame size is no larger than 50%/40%/20% of the CPB
level limit if it's a key/alt-ref/other frame.

Change-Id: I84d1a2d6d6e3c82bfd533b3309ce999cfaba2c8b
2017-01-06 10:07:31 -08:00
Jerome Jiang
afc8c4836f vp9: Compute source sad for every superblock when partition copy is on.
The source sad could be used to copy the partition without going into
choose_partitioning function to speed up vp9 encoding. Computing source
sad takes little time. Speed test on Android and Linux shows little
encoding time gain (less than 1.4%).

Turned off for now since partition copy is turned off.

Change-Id: I61c9d5b8f22329760cb29a4ee30a7f9c232ce8d3
2017-01-06 17:59:02 +00:00
Linfeng Zhang
2d12a52ff0 Merge "Add high bitdepth 8x8 idct NEON intrinsics" 2017-01-06 16:47:23 +00:00
Linfeng Zhang
90f889a56d Merge "Clean DC only idct NEON intrinsics" 2017-01-06 01:16:19 +00:00
Jerome Jiang
72746c079d vp9: Set short circuit to level 3 for VGA for speed 8.
vp9: Set short circuit to level 3 for VGA for speed 8. Also change the
threshold_32x32 to 5/8*thresholds[1] to improve quality regression
caused to VGA clips.

Change-Id: Ia1590e91e7cb22be78d5b85013387bb1be4272e3
2017-01-04 11:28:31 -08:00
Marco Paniconi
1ca1515dd3 Merge "vp9: 1 pass cbr: allow noise estimation down to 360p." 2017-01-04 17:24:08 +00:00
Marco
768b1f7281 vp9: 1 pass cbr: allow noise estimation down to 360p.
Also adjust some thresholds for noise level setting.

Change-Id: I7e03d7057ef2061c9447728deb9c6aff5d3da4b7
2017-01-03 16:26:22 -08:00
Marco
63a8257fb7 vp9: SVC unittests: fix to use y4m source.
Comment out check on buffer underrun, as it currently fails
on some of the svc tests.

Also cast the update of bits_in_buffer_model_, as this can
go negative now due to the buffer underrun.
This fixes the issue in #1352.

BUG=webm:1350
BUG=webm:1352

Change-Id: Ibd4ef23921daf09e5c15b000aca904aa4573599c
2017-01-03 15:29:04 -08:00
Yunqing Wang
99c573f018 Merge "Fix for out of range motion vector bug in joint motion search" 2017-01-03 17:46:15 +00:00
Ranjit Kumar Tulabandu
b67e1f701f Fix for out of range motion vector bug in joint motion search
Clamped the initial mv in vp9_refining_search_8p_c.

BUG=webm:1354

Change-Id: I47d302b350937e3e6e52e95c983b5fb0b4c64fba
2017-01-03 09:12:32 -08:00
Yunqing Wang
ecdb6a00c2 Merge "Make sub-pixel mv search's return value consistent with the return type" 2016-12-29 19:16:01 +00:00
Yunqing Wang
c96a8dcb5b Merge "Bug fix to avoid random crashes during ARNR filtering" 2016-12-29 17:24:24 +00:00
Gabriel Marin
e6b9609fc0 Merge "Remove superfluous conditional on 'shortcut'" 2016-12-29 06:03:43 +00:00
Linfeng Zhang
911bb980b1 Clean DC only idct NEON intrinsics
BUG=webm:1301

Change-Id: Iffc83854218460b3f687f3774e71d45b552382a5
2016-12-28 13:51:44 -08:00
Linfeng Zhang
9b187954df Add high bitdepth 8x8 idct NEON intrinsics
BUG=webm:1301

Change-Id: I56e3bc3aab9214e2debac93796389a7194991084
2016-12-27 16:28:53 -08:00
Yunqing Wang
1d12559b09 Make sub-pixel mv search's return value consistent with the return type
For out-of-range cases, returned UINT_MAX instead of INT_MAX in the
sub-pixel mv search to be consistent with the "uint32_t" return type.

Change-Id: I8e206d771228c13d89bafbbe9f14722c8ecc6a7a
2016-12-27 12:08:38 -08:00
Ranjit Kumar Tulabandu
7cf13826b7 Bug fix to avoid random crashes during ARNR filtering
The function 'vp9_find_best_sub_pixel_tree_pruned_more' is modified
to return INT_MAX for handling invalid MV cases from UINT32_MAX.

yunqingwang:
patch 3: rebased on top of the tree.
patch 4: The return type of vp9_find_best_sub_pixel_tree* was changed
to uint32_t to fix ubsan warnings. Changing UINT_MAX back to INT_MAX
was not quite right. Patch 4 modified vp9_temporal_filter.c to accept
uint32_t.
(Note: Inconsistency exists in vp9_find_best_sub_pixel_tree*, which
will be fixed in a separate CL.)

Change-Id: Ib1a79dc2aa41ea6335c21669c76883cdbb7e0535
2016-12-27 11:20:08 -08:00
Linfeng Zhang
3c47a0dc6f Merge "Clean idct 8x8 neon functions" 2016-12-27 17:59:28 +00:00
James Zern
78a24171a6 Revert "vp9: SVC unittests: fix to use y4m source."
This reverts commit f0b491a524.

This change results in unsigned integer overflows (as reported by
-fsanitize=integer) in datarate_test.cc,
for many of --gtest_filter=VP9/DatarateOnePassCbrSvc.OnePassCbrSvc*:
unsigned integer overflow: 167198 - 185560 cannot be represented in type
'unsigned long'

As the encoder didn't change, but the input with the change to
(correctly) use Y4mVideoSource, this revert is merely masking the issue.

BUG=webm:1352

Change-Id: Iecd9a6c83b3fca67c566732a5c92d36193cc2060
2016-12-23 14:18:18 -08:00
Marco Paniconi
36e767c147 Merge "vp9: SVC unittests: fix to use y4m source." 2016-12-22 17:26:42 +00:00
James Zern
90ceaba3e4 libs.mk/stress.sh,curl: set --retry to 1
provide some resilience for transient errors

Change-Id: I8db3d4eb5ef3cccc235a8c4c0052199c0ce23a27
2016-12-22 08:29:15 -05:00
Marco
f0b491a524 vp9: SVC unittests: fix to use y4m source.
Comment out check on buffer underrun, as it currently fails
on some of the svc tests.

BUG=webm:1350

Change-Id: I73c88b800cdcc06bd2f900f7b7e2a5fd08248065
2016-12-21 22:59:35 -08:00
Linfeng Zhang
6d5a3fe583 Clean idct 8x8 neon functions
BUG=webm:1301

Change-Id: I05f47dca1fddc155c8396e627cfccf6449677307
2016-12-21 14:24:17 -08:00
Marco
e7c453b613 vp9: 1 pass vbr: Skip find_predictors in pickmode when source is altref.
When source frame is altref, we only do zero-mv mode, so we can skip
the find_predictors(). No change in compression.
Small speed gain, ~1%.

Only affects 1 pass vbr with lookhead altref, for ytlive with
the macro flag USE_ALTREF_FOR_ONE_PASS on.

Change-Id: I9318c5da8521f017bf54919cd652438b3a6313d1
2016-12-21 12:12:55 -08:00
Marco Paniconi
b5770a2007 Merge "vp9; Fix to unitest for high noise." 2016-12-21 19:38:00 +00:00
Marco
9ba77ed45b vp9; Fix to unitest for high noise.
Source if y4m, and fix comment.

Change-Id: I1eb84977d42dd0f9009c276b56b3fdb03949bfc2
2016-12-21 10:22:34 -08:00
Marco Paniconi
9ba45fa510 Merge "vp9: Add datarate test for denoiser, for high noise case." 2016-12-21 03:56:13 +00:00
Marco
3fcd595dfb vp9: Add datarate test for denoiser, for high noise case.
Also breakout the denoiser tests, as the denoiser only
runs for real-time speed >=5.

Change-Id: I921b785860c35e9d1ebfad0833673a98490186c2
2016-12-20 16:48:25 -08:00
Jerome Jiang
f27276f44f Merge "vp9: Add feature to copy partition from the last frame." 2016-12-20 21:46:44 +00:00
Gabriel Marin
fce163cd54 Remove superfluous conditional on 'shortcut'
Remove superfluous test. Produces a small improvement in instruction scheduling.
Measured a 1% to 1.5% reduction in execution time for routine vp9_optimize_b
with different compilers.

No change in behavior.

TEST=Verified that encoded files match bit for bit, with and without this
change.
BUG=b/33678225

Change-Id: I2bf248d4c25fc0256147d7a8766ff9108ae9cba3
2016-12-20 12:20:21 -08:00
Kaustubh Raste
8a152a55f7 Merge "Add mips msa vp9 intrapred tests" 2016-12-20 02:27:08 +00:00
Jerome Jiang
1d5ca84df6 vp9: Add feature to copy partition from the last frame.
Add feature to copy partition from the last frame.
The copy is only done under certain conditions that SAD is below threshold.
Feature is currently disabled, until threshold is tuned.
Feature will be initially used for Speed 8 (ARM).

Under extreme case of always copying partition for speed 8:
Encode time is reduced by 5.4% on rtc_derf and 7.8% on rtc.
Overall PSNR reduced by 2.1 on rtc_derf and 0.968 on rtc.

Change-Id: I1bcab515af3088e4d60675758f72613c2d3dc7a5
2016-12-19 16:24:03 -08:00
Gabriel Marin
85aead1790 Merge "Simplify address arithmetic in vp9_optimize_b" 2016-12-19 23:25:39 +00:00
James Zern
80474bf65e Merge "vpx_idct32x32_1024_add_neon: quiet uninitialized warning" 2016-12-19 22:39:01 +00:00
Marco Paniconi
c1f5194842 Merge "vp9 denoiser: Fix the logic for re-evaluating zeromv after denoising." 2016-12-19 21:15:37 +00:00
Gabriel Marin
0549f5aae9 Simplify address arithmetic in vp9_optimize_b
Simplify address arithmetic on token_costs to reduce the number of generated
instructions that are used for address arithmetic inside routine
vp9_optimize_b. It also helps improve instruction scheduling depending on
compiler and optimization level.

Measured a 9.3% reduction in retired instructions and 5.3% reduction in
execution time for this routine with GCC v4.8.4 and optimization flags -O3,
and a reduction of up to 11.6% in execution time with other compilers.

No change in behavior.

TEST=Verified that encoded files match bit for bit, with and without this
change.
BUG=b/33678225

Change-Id: I6098650fb5cd2aa04e014fe6e68ca20761f3a21f
2016-12-19 13:10:04 -08:00
James Zern
a68b36c752 vpx_idct32x32_1024_add_neon: quiet uninitialized warning
relocate the assignment to 'in' outside of the for loop. this quiets a
spurious warning in visual studio builds since:
86e340c enable vpx_idct32x32_1024_add_neon in hbd builds

+ give the variable a more descriptive name

BUG=webm:1294

Change-Id: I5c3da5c7939621477e0fc0ad3a1b2a3045c5bffd
2016-12-19 12:49:44 -08:00
Marco
6e8dbc76ad vp9: With denoising on, only estimate noise level for higher resolns.
Allow it for resolns above 640x360 for now.

Change-Id: I087d0d8173f96b316164fdd4a499110ce2e7a233
2016-12-19 10:05:54 -08:00
Marco
61b569b461 vp9 denoiser: Fix the logic for re-evaluating zeromv after denoising.
Correctly set interp_filter to SWITCHABLE for INTRA mode.
Also reduce threshold on noise level for re-evaluating zeromv.

Change-Id: Id32c01e193209fb380aa07204f0be3babf29f70a
2016-12-19 09:30:16 -08:00
Linfeng Zhang
7e23f895ca Merge "Clean hbd idct 4x4 neon functions and other" 2016-12-19 17:09:26 +00:00
Kaustubh Raste
1f3e079a35 Add mips msa vp9 intrapred tests
Change-Id: I49b91464a87cad8692f4b1477e45e5f567b4fe87
2016-12-19 17:32:38 +05:30
Johann Koenig
9b63cb057a Merge "post proc test: add padding for sse2 tests" 2016-12-17 01:12:34 +00:00
Marco Paniconi
d1eca240fb Merge "vp9: Change condition to enable recheck_zeromv_after_denoising." 2016-12-16 23:53:33 +00:00
Marco
4260a7f2b3 vp9: Change condition to enable recheck_zeromv_after_denoising.
For when denoising enabled: change condition to enable
the recheck_zeromv_after_denoising for only very high noise level.
This is causing an issue, so enabling it for very high noise
to effectively shut it off.

Change-Id: Ic40d6025f3f398338cedd270d17c0ccd9a3daa84
2016-12-16 15:00:21 -08:00
Johann
5993b808f0 post proc test: add padding for sse2 tests
Avoid valgrind warnings for reading out of bounds when the width is not
divisible by 16.

Change-Id: I5670d7cfbbce00874b98cfb7472f99c7936c2c47
2016-12-16 14:06:06 -08:00
Johann
4781a67737 postproc test: disable new down and across test
The new test is causing valgrind failures:
[ RUN      ] SSE2/VpxPostProcDownAndAcrossMbRowTest.CheckCvsAssembly/0
==28923== Invalid read of size 16
28923==    at 0x724016: ??? (deblock_sse2.asm:146)

Disable during investigation. The test is new but the code is not.

Change-Id: I5521e5fd48a595e3798b833bf7e3cc97b81c1975
2016-12-16 12:19:00 -08:00
Jim Bankoski
318a1ff5ec vp8 : use threading mutex's for tsan only.
To avoid decode performance hit of 2% when running on hyperthreaded
cores.

This patch only uses the mutex's when we are running tsan.

This is safe because 32 bit operations like read and store are atomic
on all the platforms we care about. Tsan warns about race situations,
but in this case either situation ( read occurs before write or write
before read) the worst case is that we go around one extra time in the
loop.  So the ordering doesn't really matter.

That said a few other things have been tried :

for instance as per here:
webrtc/base/atomicops.h#52

In this patch they use:
__atomic_load_n(i, __ATOMIC_ACQUIRE);
__atomic_store_n(i, value, __ATOMIC_RELEASE);

This code works on gcc, clang ( replacing protected write and read), and
avoids tsan errors. Incurring no penalty in performance.  In C11 its
replaced by straight atomic operands.

However there is no equivalent in the visual studio's we support as
int32 on all windows platforms is already atomic.  To avoid tsan like
warnings on windows we'd need to use interlocked exchange and the
end result doesn't gain us any thing.

Change-Id: I2066e3c7f42641ebb23d53feb1f16f23f85bcf59
2016-12-16 08:50:55 -08:00
Marco Paniconi
2b1ec65b5d Merge "vp9: Fix to usage of flag USE_ALTREF_FOR_ONE_PASS" 2016-12-15 19:48:16 +00:00
Johann
41b0888a84 postproc: neon down and across macroblock filter
Implement vpx_post_proc_down_and_across_mb_row in NEON.
Runs about 6-7x faster than C.

BUG=webm:1320

Change-Id: Ic5c7d3552a88cfcf999ec5bf2bd46fee460642c2
2016-12-14 15:11:28 -08:00
Marco
5de798f2b2 vp9: Fix to usage of flag USE_ALTREF_FOR_ONE_PASS
The flag USE_ALTREF_FOR_ONE_PASS allows for alt-ref lookahead
in 1 pass vbr (from https://chromium-review.googlesource.com/#/c/365498).
This change is to make sure this macro flag only has effect if
the config flag cpi->oxcf.enable_auto_altef is also on.

No change in ytlive encoding, as USE_ALTREF_FOR_ONE_PASS is not
yet enabled.

Change-Id: I1a69681e4a15c5244581a3dab4587fca08f02e0f
2016-12-14 15:07:38 -08:00
Linfeng Zhang
c8f25fa5c0 Clean hbd idct 4x4 neon functions and other
BUG=webm:1301

Change-Id: I387b7eae716a7df15c691dc6f368b07602df7342
2016-12-14 11:38:28 -08:00
Yaowu Xu
27e1bacdb3 Change order of operation to avoid ubsan warnings
This commit change an order of operation to avoid left shifts of
negative numbers.

Change-Id: I607c7eb91658c7a5ef397fc1504721d1b10e3dd6
2016-12-14 09:37:14 -08:00
Linfeng Zhang
3dd20456ab Merge "Update idct test code to test 8-bit & high bitdepth simultaneously" 2016-12-14 17:05:34 +00:00
Linfeng Zhang
201dcefafe Update idct test code to test 8-bit & high bitdepth simultaneously
Change-Id: Icc0eb9c0ddf2a13ec832877a089450972134e8ec
2016-12-13 17:25:04 -08:00
James Bankoski
3486abd54a Merge "Reapply 'Amend and improve VP8 multithreading implementation'" 2016-12-14 01:21:50 +00:00
James Zern
86e340c76e enable vpx_idct32x32_1024_add_neon in hbd builds
BUG=webm:1294

Change-Id: Ibdda54e6d1303b0f73bc7bc71417e4041d7618de
2016-12-12 19:28:35 -08:00
Jim Bankoski
85a541a421 Reapply 'Amend and improve VP8 multithreading implementation'
Reapply this patch:
ff0107f Amend and improve VP8 multithreading implementation

Amended the patch to add a unit test, and fix an asan error.

BUG=webm:851

Change-Id: I6572c03256169c64e80248bf5a5e99f59a2fc93c
2016-12-13 02:11:34 +00:00
Linfeng Zhang
5d4aa325a6 Cosmetics by unifying dest_stride to stride in idct
Change-Id: Ie9336a808a3c3592bb4fd5d4ad3839028bfcafba
2016-12-12 15:13:22 -08:00
James Bankoski
282f3b3d78 Merge "vp8: adds multithread testing." 2016-12-10 00:01:32 +00:00
Marco Paniconi
817488be47 Merge "vp9: Fix to crash in svc code." 2016-12-09 23:47:02 +00:00
Jim Bankoski
121e161115 vp8: adds multithread testing.
The test is disabled because of TSAN errors until we resolve
BUG=webm:851

Change-Id: I0b21c8d815bc1ea365da024b1e2ee5e1fc5715c2
2016-12-09 15:05:59 -08:00
Johann
2c24f7178d Move load_and_transpose to transpose_neon.h
Allows for use outside the idcts without pulling in idct_neon.h

Change-Id: I4a94c1af3dac3e1b5bc8296ec9eab0ddcc8cfecf
2016-12-09 12:54:55 -08:00
Marco
076d4bd91a vp9: Fix to crash in svc code.
use_base_mv assumes 2x2 scaling, so fix is to shutoff
this feature unless spatial scale factors are 2.

Added svc unittest for 2 spatial layers with 5x5 scaling,
which generates the issue without this fix.

Also fix some settings in svc unittest:
let the speed setting vary (from 5 to 8), and enable static threshold.

BUG=webm:1344

Change-Id: Idfd0a6c633c21b49a0479601506302cfe974e30e
2016-12-09 08:57:09 -08:00
James Zern
7ba9d31e3f Merge "idct16x16_add_neon: fix arm visual studio builds" 2016-12-09 03:19:16 +00:00
Marco
cd6f742980 vp8 multi_res_encoder: Ajust some settings in sample encoder.
Set #threads to default 1 for all streams, change bit allocaton
for 3 temporal layers, and enable denoiser on middle resolution layer.

Change-Id: I4a57adbfdb2c319002b8f3cf359613842dc00d75
2016-12-08 15:27:16 -08:00
James Zern
6defef4ab2 idct16x16_add_neon: fix arm visual studio builds
after:
2d3d95f enable vpx_idct16x16_256_add_neon in hbd builds

reorder INCLUDEs and fix indent of IF/ENDIFs

remove vpx_config.asm to avoid multiple symbol definitions in windows
builds and shift idct_neon.asm.S to the top to allow use of
CONFIG_VP9_HIGHBITDEPTH in the export list.

Change-Id: I0dacfbae62a6ec8fe4a26940c1a52da2dfad2029
2016-12-08 15:17:57 -08:00
Yunqing Wang
880adc3355 Merge "Remove an unused first pass statistic" 2016-12-08 22:46:44 +00:00
Yunqing Wang
394020383d Remove an unused first pass statistic
One of the first pass stats "new_mv_count" is no longer used in VP9,
and is removed. This also makes it easy to implement a multi-threaded
first pass. This change doesn't affect the coding performance, which
has been verified by borg tests.

Change-Id: I4c7c7bf9465fda838eb230814ef0c631c068c903
2016-12-07 15:32:25 -08:00
Marco Paniconi
e4c6f8fde7 Merge "vp9: Fix some TODOs in svc code." 2016-12-07 22:06:01 +00:00
Linfeng Zhang
385599b553 Merge "Update TEST_P(PartialIDctTest, RunQuantCheck)" 2016-12-07 21:05:05 +00:00
Linfeng Zhang
174528de1e Merge "Update idct NEON optimization to not use narrowing saturating shift" 2016-12-07 21:03:21 +00:00
Marco
5778a7c9cb vp9: Fix some TODOs in svc code.
Change-Id: Ie9f441245987ade9dab38af69adf4dd1fb38ca3f
2016-12-07 13:02:48 -08:00
James Zern
f16a0a1aa4 Merge "enable vpx_idct16x16_256_add_neon in hbd builds" 2016-12-07 20:26:44 +00:00
Linfeng Zhang
834feffe08 Update TEST_P(PartialIDctTest, RunQuantCheck)
1. Use correct projections when copying real dct/quant outputs.
2. Remove local random number generator and combine loops.
3. Quantization with minimum allowed step sizes instead of maximum.
   This may generate larger inputs.

Change-Id: I154afc26230c894d564671cff4b8fd5485b69598
2016-12-07 11:34:00 -08:00
Marco Paniconi
17c403d0ab Merge "vp9: Adjust the weight factor for segment rate cost for aq-mode=3." 2016-12-07 19:31:13 +00:00
Linfeng Zhang
018a2adcb1 Update idct NEON optimization to not use narrowing saturating shift
Change-Id: Iae517017217dbacd638d40fcfeeb0f4bba7b8b8b
2016-12-07 10:25:09 -08:00
James Zern
2d3d95f7ac enable vpx_idct16x16_256_add_neon in hbd builds
BUG=webm:1294

Change-Id: Ib421c150b0d29dee0a81390a612bf01a4a28cff1
2016-12-06 18:32:21 -08:00
James Zern
228c9940ea Merge changes Ibad079f2,I7858a0a1
* changes:
  enable vpx_idct16x16_10_add_neon in hbd builds
  idct16x16,NEON: rm output_stride from pass1 fns
2016-12-07 01:40:28 +00:00
James Zern
8befcd0089 enable vpx_idct16x16_10_add_neon in hbd builds
BUG=webm:1294

Change-Id: Ibad079f25e673d4f5181961896a8a8333a51e825
2016-12-06 16:09:19 -08:00
James Zern
af9d7aa9fb idct16x16,NEON: rm output_stride from pass1 fns
vpx_idct16x16_256_add_neon_pass1, vpx_idct16x16_10_add_neon:
this was a constant 8 in all cases meaning the results are stored
contiguously, this allows the number of stores to be reduced.

Change-Id: I7858a0a15a284883ef45c13dfd97c308df9ea09e
2016-12-06 15:13:33 -08:00
Linfeng Zhang
cb339d628f Refine 8-bit 8x8 idct NEON intrinsics
Change-Id: I4ec4ad1928ec2ed87f596f52f097bc52065278dd
2016-12-05 17:50:14 -08:00
Marco
360ac89885 vp9: Adjust the weight factor for segment rate cost for aq-mode=3.
Use the segment weight factor based on the target (cr->percent_refresh)
if it less than the current estimate (avergae of past usage and target).
Small improvement at low bitrates.

Change-Id: Iba8fd909e203f94458901366d3a991f7ea854d49
2016-12-05 12:42:56 -08:00
Linfeng Zhang
a8eee97b43 Check in vpx_lpf_vertical_4_dual_neon() assembly
This replaces its C version.

Change-Id: Ie39e9324305fdc0fff610ced608a037e44a85a1a
2016-12-02 15:54:30 -08:00
James Zern
a7fa1314da Merge changes I4afc130e,Iaa64d23f
* changes:
  Add high bitdepth 4x4 idct NEON intrinsics
  Update idct x86 intrinsics to not use saturated add and sub
2016-12-02 04:01:28 +00:00
Linfeng Zhang
17a8cf5cc3 Add high bitdepth 4x4 idct NEON intrinsics
Change-Id: I4afc130effa05b8be2e9f982967216b1beb2ce4b
2016-11-30 13:07:13 -08:00
Linfeng Zhang
264f6e70ec Update idct x86 intrinsics to not use saturated add and sub
Change-Id: Iaa64d23fdb45ca1f235b0ea57e614516e548eca4
2016-11-29 17:06:08 -08:00
James Zern
c6641782c3 idct16x16,NEON,cosmetics: normalize fn signatures
+ remove unused parameters from vpx_idct16x16_10_add_neon_pass2

Change-Id: Ie5912a4abdd308fab589380bca054a2e7234a2c4
2016-11-28 16:46:01 -08:00
James Zern
12566c3d0f Merge changes Ide6d3994,I164cfcbe
* changes:
  enable vpx_idct32x32_135_add_neon in hbd builds
  idct_neon: rename load_tran_low_to_s16 -> ...s16q
2016-11-29 00:12:45 +00:00
James Zern
33ddc645ce Merge "build/make/Android.mk: correct rtcd template var refs" 2016-11-28 23:39:37 +00:00
James Bankoski
68991d7f87 Merge "svc_test: fix two warnings" 2016-11-28 22:27:26 +00:00
Jim Bankoski
27b5cc31e6 svc_test: fix two warnings
Use of possibly uninitialized variable and missing test initializer.

Change-Id: I2192c81c39ef4239cc11a309850c0ee8781ef17e
2016-11-28 12:53:39 -08:00
Jerome Jiang
f68cf8ba19 Cosmetic changes to variable names in deblocker tests.
Change kExpectedOutput to expected_output in function parameters in
the deblocker test.

Change-Id: I5baf8d1285ac47922950887406c7aa519ddc512a
2016-11-28 10:08:12 -08:00
James Zern
a58e0b2a74 build/make/Android.mk: correct rtcd template var refs
the expansion of findstring and rtcd_dep_template_CONFIG_ASM_ABIS needs
to be deferred until the block is parsed as makefile syntax rather than
eval time where rtcd_dep_template_CONFIG_ASM_ABIS will be unset. this
ensures vpx_config.asm is properly created.

Change-Id: I7c38c6c082da78397936467482789dd468adc316
2016-11-24 17:55:16 -08:00
James Zern
120234fa17 Merge changes I6b4cd56e,I88f91b92
* changes:
  Android.mk,armv7: fix idct_neon.asm.S creation
  build/make/Android.mk: set/use qexec appropriately
2016-11-24 07:22:04 +00:00
James Zern
21a1abd8e3 enable vpx_idct32x32_135_add_neon in hbd builds
BUG=webm:1294

Change-Id: Ide6d3994fe01c4320c9d143e6d059b49568048e4
2016-11-23 19:59:43 -08:00
James Zern
568d4b1d63 idct_neon: rename load_tran_low_to_s16 -> ...s16q
BUG=webm:1294

Change-Id: I164cfcbe9bc4511d1d04af9206cf351a0ec2957b
2016-11-23 19:57:48 -08:00
James Zern
d757d7e998 Merge changes Icc4ead05,Ib019964b,I3b5fd3b3,Ieedadee2
* changes:
  Update vpx_idct4x4_16_add_neon() to pass SingleExtremeCoeff test
  Refine 8-bit 4x4 idct NEON intrinsics
  Add idct speed test.
  Update partial_idct_test.cc to support high bitdepth
2016-11-24 03:31:25 +00:00
Jerome Jiang
f63eb66ecd Merge "Change C/MSA post proc to match SSE2." 2016-11-24 01:56:34 +00:00
Jerome Jiang
95eb505660 Merge "Cover more filter levels in unit tests for post proc." 2016-11-24 01:56:22 +00:00
James Zern
2c598d0858 Android.mk,armv7: fix idct_neon.asm.S creation
force this to be created before any other .S files. this change
additionally removes the file from the source list as it doesn't need to
be compiled on its own.

Change-Id: I6b4cd56ef6059d08f75f06fb749cddf76e0e165e
2016-11-23 16:49:19 -08:00
James Zern
1136db0db0 build/make/Android.mk: set/use qexec appropriately
commands are echo'd when V=1; libs.mk depends on this variable as well

Change-Id: I88f91b9260f16686cfccdf6bd3f29d246521b62e
2016-11-23 16:46:50 -08:00
Marco
d793950ec8 vp9: Adjust cyclic refresh parameters for low bitrates.
Increase the motion threshold and qp-delta for segment#2 boost.
This can increase the frame-drop at low bitrates, but generally
better spatial quality.

Only affects real-time mode with aq-mode=3, at very low bitrates.

Change-Id: I5ccb784667f70d0c27d369806b93b1f93d5605d1
2016-11-23 12:14:28 -08:00
James Zern
af290bfe3b Merge "use storage.googleapis for testdata download" 2016-11-23 19:27:20 +00:00
Jerome Jiang
97ec6291ee Change C/MSA post proc to match SSE2.
BUG=webm:1321

Change-Id: I719023375dc48cf7d8ed72188853f0f1ccc4ad7f
2016-11-23 10:42:11 -08:00
Jerome Jiang
755fb3d4ec Cover more filter levels in unit tests for post proc.
For some filter level, the C/MSA doesn't match SSE2. Part of unit tests
are disabled. They will be re-enabled when C/MSA funcs are fixed.
BUG=webm:1321

Change-Id: Ib16b98b5eecb15d2252aa4ea267b782ee2b27533
2016-11-23 10:31:41 -08:00
Marco Paniconi
8b2cbaefcf Merge "vp9: Use more aggressive skip when short_circuit_low_temp_var = 1." 2016-11-23 18:15:58 +00:00
James Zern
22f7aca097 use storage.googleapis for testdata download
replace downloads.webmproject.org with the canonical
storage.googleapis.com/... form. this appears less likely to fail when
dealing with multiple concurrent connections.

Change-Id: I0dcbd04df9e4057fa851f458b3ef7e3589f1f2f1
2016-11-22 23:03:12 -08:00
James Zern
d7f1d60c51 Merge "avoid redefining WIN32_LEAN_AND_MEAN" 2016-11-23 00:43:22 +00:00
James Zern
198a046d3e Merge "vp9,read_inter_block_mode_info: quiet msan warning" 2016-11-23 00:42:24 +00:00
James Zern
cb22359d02 vp9,read_inter_block_mode_info: quiet msan warning
best_sub8x8[1] won't be used meaningfully when is_compound is false, but
may trigger an msan warning as the value is copied around and later
clamped.

BUG=667044

Change-Id: Icc24c3b72cdb550bebea44d4aaa4ff8bf3fbab56
2016-11-22 15:32:00 -08:00
Linfeng Zhang
05e2b5a59f Merge "Add 32x32 d45 and 8x8, 16x16, 32x32 d135 NEON intra prediction" 2016-11-22 23:20:53 +00:00
James Zern
446d1ee624 avoid redefining WIN32_LEAN_AND_MEAN
fixes redef errors when the macro is supplied elsewhere, e.g., the
command line

Change-Id: Ic15726817a43e30595d50562ef1f077060c193cf
2016-11-22 15:15:53 -08:00
Marco
b6597745f9 vp9: Use more aggressive skip when short_circuit_low_temp_var = 1.
Use the same feature as https://chromium-review.googlesource.com/#/c/411327/,
but allow it to be used for speed  = 6 and 7, where
short_circuit_low_temp_var = 1.

Speed up of ~2-3% for speed 7, with little/no loss in compression.

Change-Id: I263a0f261ad9929034392d68f0153dc6376fdb5f
2016-11-22 14:54:28 -08:00
Jerome Jiang
0966757874 Cosmetic changes to post proc unit tests.
Remove unnecessary "virtual" before some functions. Change *_btm_* in
variable names to *_bottom_*.

Change-Id: Ifd4ce667537617f451cdfed47dd8c48817fd983b
2016-11-22 22:28:17 +00:00
James Zern
7d2690e658 Merge "build/make/Android.mk: use -fPIC w/ENABLE_SHARED=1" 2016-11-22 20:14:12 +00:00
Linfeng Zhang
6cc76ec73f Update vpx_idct4x4_16_add_neon() to pass SingleExtremeCoeff test
Change-Id: Icc4ead05506797d12bf134e8790443676fef5c10
2016-11-22 11:35:05 -08:00
James Bankoski
4bb01229cd Merge "vp9-tests : split VpxEncoderThreadTest into two tests." 2016-11-22 19:34:16 +00:00
Linfeng Zhang
974e81d184 Refine 8-bit 4x4 idct NEON intrinsics
Change-Id: Ib019964bfcbce7aec57d8c3583127f9354d3c11f
2016-11-22 11:26:03 -08:00
Linfeng Zhang
45876b4550 Add idct speed test.
Change-Id: I3b5fd3b36cac1fb3a93e27fd8fd0781c91d412ce
2016-11-22 11:19:24 -08:00
Linfeng Zhang
d479c9653e Update partial_idct_test.cc to support high bitdepth
BUG=webm:1301

Change-Id: Ieedadee221ce539e39bf806c41331f749f891a3c
2016-11-22 11:11:58 -08:00
Jim Bankoski
719f39f44e vp9-tests : split VpxEncoderThreadTest into two tests.
VpxEncoderThreadTest was taking a very long time for some runs and
timing out a lot.   This is an attempt to split the test into runs
that can be run nightly ( speeds 2 through 9) and runs that can
be run weekly ( speeds 0-1 ).

Change-Id: Iee6f61a561006d3a30381dd3b52b9a4dce07a70c
2016-11-22 07:31:04 -08:00
Kaustubh Raste
ecc5998bcf Fix mips dspr2 build warning
Change-Id: Ia8fb3ed124f01384e7896e309c9ff22c05b40719
2016-11-22 17:49:17 +05:30
Yaowu Xu
0ffbb36ddc Add validation of frame_parallel_decoding_mode
This is a boolean value that is written into bitstream, any value other
than 0 or 1 could have led to unexpected behavior. This commit fix the
issue by adding validation of the value to make sure it is boolean.

BUG=webm:1339

Change-Id: I2d3e69e8dbefcab9a0db9cb39a91a40ce531c5a1
2016-11-21 10:53:25 -08:00
Jingning Han
f473e892f7 Merge "Enable asymptotic closed-loop encoding decision" 2016-11-19 04:12:55 +00:00
Kaustubh Raste
a38e9f412d Merge "Fix SingleLargeCoeff idct test" 2016-11-19 03:37:29 +00:00
James Zern
3d55311062 vpx_temporal_svc_encoder.sh: fix comment (// -> #)
Change-Id: Ib13152a9ff523b1c29e8519e4f7ed01ad9874525
2016-11-18 19:11:55 -08:00
James Zern
7317ce8bd4 build/make/Android.mk: use -fPIC w/ENABLE_SHARED=1
fixes reloc errors like:
R_X86_64_PC32
vpx_dsp/x86/deblock_sse2.o:
requires dynamic R_X86_64_PC32 reloc against 'vpx_rv' which may overflow
at runtime

Change-Id: I218fc0e7c8258197f890d395f335e5a4fe82dccb
2016-11-18 18:54:34 -08:00
James Zern
cbeae53e76 Merge "Clean horizontal intra prediction NEON optimization" 2016-11-19 01:29:37 +00:00
James Zern
7adeccb33d Merge "partial_idct_test: s/SingleLargeCoef/SingleExtremeCoeff/" 2016-11-18 20:02:44 +00:00
Jerome Jiang
23f1bfbd85 Merge "Change *_xmm to *_sse2 in deblocker assembly functions." 2016-11-18 00:23:45 +00:00
Jerome Jiang
de5fd00ec5 Change *_xmm to *_sse2 in deblocker assembly functions.
Some cosmetic changes because xmm is an anachronism.

Change-Id: I436a5b78a3c52776c20d6640939311f2a84a9bc7
2016-11-17 23:38:04 +00:00
James Zern
f6921412d4 partial_idct_test: s/SingleLargeCoef/SingleExtremeCoeff/
tests with 'Large' in the name are reserved for slow running tests which
may not be run on all platforms

Change-Id: I2a7d6dd46b29b50469893e46433844132fb727c2
2016-11-17 12:28:57 -08:00
Marco Paniconi
485a49d0b8 Merge "vpx_temporal_svc_encoder.sh: Run all tests for 1-4 threads for vp8/vp9." 2016-11-17 20:22:32 +00:00
Kaustubh Raste
c56e5dd620 Fix SingleLargeCoeff idct test
Updated idct code to handle single large coefficient (-32768)

Change-Id: Ia13ab1ab434a9a1b9954a5914088977a88841cc7
2016-11-17 11:41:07 +00:00
Jerome Jiang
5d48663e04 Merge "Change C and msa to match results from sse2." 2016-11-17 05:16:27 +00:00
Jerome Jiang
cb1b1b8fef Change C and msa to match results from sse2.
Re-enable the tests to check CvsAssembly.
BUG=webm:1321

Change-Id: Id7f7d74b06c469fb6c8f5d04e91359e9cd9097a6
2016-11-16 17:05:26 -08:00
Marco
2ef2243804 vpx_temporal_svc_encoder.sh: Run all tests for 1-4 threads for vp8/vp9.
Change-Id: I079ee87cb32e36a1486c377c0037945b4bb89626
2016-11-16 14:11:25 -08:00
Jim Bankoski
f667cc7a4e stress.sh: Runs multiple libvpx encodes and decodes in parallel
This runs multiple encodes and decodes of vp8 and vp9 in parallel,
with so many threads that problems with synchronization can show up.

Change-Id: I2b297e7f43d1e741323c7ad9f50a3931ae609f16
2016-11-16 06:59:26 -08:00
James Zern
12fe34516e Merge "build/make/Android.mk: fix cpufeatures import" 2016-11-15 23:41:48 +00:00
James Zern
f09c687ea9 Merge changes I3950c883,I2b679b04
* changes:
  partial_idct_test: use <limits> for int16_min/max
  vpx_timer.h,x86.h: define NOMINMAX for windows.h
2016-11-15 23:41:18 +00:00
Linfeng Zhang
011fdec1e6 Merge "Add high bitdepth intra prediction NEON optimization (mode tm)" 2016-11-15 23:30:48 +00:00
Jerome Jiang
4ddae8f524 Merge "vp9: Speed 8: More aggresive golden skip for low res." 2016-11-15 22:50:58 +00:00
Linfeng Zhang
85c1ee434d Add high bitdepth intra prediction NEON optimization (mode tm)
BUG=webm:1316

Change-Id: Ib014de06836ac12726f4a2c9f0833ec4eb4d233b
2016-11-15 14:19:46 -08:00
Jerome Jiang
360217a233 vp9: Speed 8: More aggresive golden skip for low res.
Add a new, more aggresive short circuit: short_circuit_low_temp_var = 3 to skip
golden of any mode when variance is lower than threshold for low res.
This change only affects speed = 8, low resolution.

Metrics for avgPSNR/SSIM on rtc_derf (low resolution) show loss of
0.27/0.31%.
On Nexus 6, the encoding time is reduced by ~2.3% on average across all
low-res clips.

Visually little change on rtc_derf clips.

Change-Id: Ia8f7366fc2d49181a96733a380b4dbd7390246ec
2016-11-15 13:56:27 -08:00
James Zern
2218a4c292 partial_idct_test: use <limits> for int16_min/max
this removes the need for __STDC_LIMIT_MACROS which is defined in
vpx_integer.h, but may be preceded by earlier includes of stdint.h;
fixes build with the r13 ndk

Change-Id: I3950c8837cf90d5584a20ce370ae370581c2182c
2016-11-15 12:18:38 -08:00
James Zern
0412193bb9 vpx_timer.h,x86.h: define NOMINMAX for windows.h
avoids the definition of min/max macros in headers that may appear in
c++ unit tests. the codebase uses VPXMIN/MAX for this purpose in any
case

Change-Id: I2b679b045d64fb34fd8780f704e3caf10a758d82
2016-11-15 12:18:38 -08:00
James Zern
f938ab5e6a build/make/Android.mk: fix cpufeatures import
use 'android/cpufeatures' rather than 'cpufeatures'; this matches the
documentation, fixes compilation with r12b/r13 and still works with
older ndks.

Change-Id: I2f34233c164e6d4d46428f8905d5502cea4288a2
2016-11-14 13:25:51 -08:00
Jerome Jiang
eff68a3a4d vp9: Speed 8: Turn off 4x4avg for low-res non-key frames.
Changes only affects speed = 8 for low resolutions.

Metrics for avgPSNR/SSIM on rtc_derf (low resolutions) show loss of
0.5/0.6%.
On Nexus 6, the encoding time is reduced by ~5.9% on average across all
low-res clips.
Visually little/no change on rtc_derf clips.

Change-Id: I68dd50e558d72dcc1af8317d224bfae5e3bd872d
2016-11-14 11:17:14 -08:00
Jingning Han
44f8ee7258 Enable asymptotic closed-loop encoding decision
This commit enables asymptotic closed-loop encoding decision for
the key frame and alternate reference frame. It follows the regular
rate control scheme, but leaves out additional iteration on the
updated frame level probability model. It is enabled for speed 0.

The compression performance is improved:

lowres 0.2%
midres 0.35%
hdres  0.4%

Change-Id: I905ffa057c9a1ef2e90ef87c9723a6cf7dbe67cb
2016-11-14 09:22:55 -08:00
Linfeng Zhang
a3128ad33a Add high bitdepth intra prediction NEON optimization (h and v)
BUG=webm:1316

Change-Id: I47eeac698a98a31d1af5f72441052302e9fa4f46
2016-11-12 12:00:19 -08:00
Jerome Jiang
186dc40e8e Merge "Add unit tests for post proc." 2016-11-12 04:38:27 +00:00
Jerome Jiang
b282048fe4 Add unit tests for post proc.
Some tests are disabled since C and msa don't match sse2.
BUG=webm:1321

Change-Id: I61f303348e5292844a822612f100dbe006489e3e
2016-11-11 15:17:53 -08:00
Marco Paniconi
b6f6169348 Merge "vp9: Adjust thresholds for limiting cyclic refresh for noisy content." 2016-11-11 17:11:19 +00:00
James Zern
ba016b710a Merge "*ppflags.h: remove unused *_DEBUG_* enum values" 2016-11-10 20:53:22 +00:00
James Zern
80f6b243a7 Merge changes I339088b2,Iaade219e,If142afb1,I4257c4b3
* changes:
  fdct8x8_test: add vpx_idct8x8_64_add_neon in hbd
  fdct4x4_test: add vpx_idct4x4_16_add_neon in hbd
  partial_idct_test,NEON: add missing idct variants
  enable vpx_idct32x32_34_add_neon in hbd builds
2016-11-10 05:02:39 +00:00
James Zern
a1c40a2c1a fdct8x8_test: add vpx_idct8x8_64_add_neon in hbd
this was enabled in:
3ae2597 idct,NEON: add a tran_low_t->s16 load adapter

+ enable it for all NEON configs, both intrisincs and assembly versions
exist

BUG=webm:1294

Change-Id: I339088b2a398200f95658d040034fb9b2a7c8ce0
2016-11-09 20:04:27 -08:00
Linfeng Zhang
40ab0424d4 Add high bitdepth intra prediction NEON optimization (mode d45 and d135)
BUG=webm:1316

Change-Id: I6a330874348df04df24a6d9efdc06f567e04bf8e
2016-11-09 12:04:04 -08:00
James Zern
4807f1584c *ppflags.h: remove unused *_DEBUG_* enum values
usage of the vp8 versions was removed in:
3f72509 vp8: remove VP8_SET_DBG* control support

vp9 had the usage stripped even earlier.

Change-Id: I978142eb6492552cd29c9c6feb1e89acfc5f7b84
2016-11-08 21:09:16 -08:00
James Zern
cfbb599335 fdct4x4_test: add vpx_idct4x4_16_add_neon in hbd
this was enabled in:
3ae2597 idct,NEON: add a tran_low_t->s16 load adapter

+ enable it for all NEON configs, both intrisincs and assembly versions
exist

BUG=webm:1294

Change-Id: Iaade219e9d1de7b69423670d3ea6271b0965e068
2016-11-08 18:29:40 -08:00
James Zern
c344dee463 partial_idct_test,NEON: add missing idct variants
idct4x4 and idct8x8 were universally enabled for high-bitdepth builds
in:
3ae2597 idct,NEON: add a tran_low_t->s16 load adapter

BUG=webm:1294

Change-Id: If142afb169c48728cc4b222e7c41aa4a63f95f0f
2016-11-08 18:29:35 -08:00
James Zern
738c8f23c6 enable vpx_idct32x32_34_add_neon in hbd builds
replace load_and_transpose_s16_8x8() in idct32_6_neon() with a separate
load_tran_low_to_s16() and transpose_s16_8x8(). the combined function is
used in idct32_8_neon() where the input is the correctly sized output
from the earlier stage.

BUG=webm:1294

Change-Id: I4257c4b3a421b2cf5d13651f966eee0680ef98a9
2016-11-08 17:03:36 -08:00
Marco
18794d8ddc vp9: Adjust thresholds for limiting cyclic refresh for noisy content.
For noisy content, be more aggressive in skippping some blocks for
delta-qp to reduce noise pulsing artifact. Also treat frame boundary
case when dimension is not multiple of superblock size/64.

Only affects non-screen content case, and when source noise
is measured to be high (at least level kMedium).

Change-Id: Ib13a2a20ed1ce37ff3c44d95c3ef2635fd695222
2016-11-08 15:50:46 -08:00
Johann
f5141ea45f Refine vp8_refining_search_sadx4 targeting
This uses the same sdx4df pointers as vp8_diamond_search_sadx4 and
should therefore target the same optimizations.

See e4ddf9db6a

Change-Id: Ic298e9b25c34bbe6b7a0799509355b0addb56675
2016-11-08 15:22:44 -08:00
Johann
50b40f114c Optimize idct32x32_135_add for NEON
BUG=webm:1295

Change-Id: I7f80ef4d29813fcb401fc6075babf19e3c195462
2016-11-08 22:06:07 +00:00
Linfeng Zhang
64a5a8fd6f Merge "Add high bitdepth intra prediction NEON optimization (mode dc)" 2016-11-08 16:53:42 +00:00
James Zern
3fdfbcb73d Merge "partial_idct_test: set MinSupportedCoeff for NEON" 2016-11-08 01:05:16 +00:00
Johann Koenig
5c64c01c7c Merge "ads2gas: remove RN stanza" 2016-11-08 00:44:17 +00:00
Johann
271de2c9fb ads2gas: remove RN stanza
The matching on ads2gas_apple.pl is too liberal and catches
CONFIG_EXTERNAL_BUILD and CONFIG_INTERNAL_STATS because they have RN in
the names.

The RN renaming feature is not used in any existing assembly files. It
was used in some armv6 files but they were removed.

Change-Id: Ib65abf1947d3e89f0d1584e2a5de399d24008f95
2016-11-07 16:21:16 -08:00
James Zern
40bcb96abd partial_idct_test: set MinSupportedCoeff for NEON
vpx_idct4x4_16_add_neon fails with INT16_MIN, +1 is all right

BUG=webm:1335

Change-Id: I25830c8ab0782822fc3c9db6cc669c2e65f2700e
2016-11-07 15:47:09 -08:00
Linfeng Zhang
d545c19afa Rename vpx_highbd_idct8x8_10{*}() to vpx_highbd_idct8x8_12{*}()
Also update its trigger threshold from 10 to 12.

Change-Id: Ib8dddd87a5a22a12ca66e7084d342fbb027b0a2f
2016-11-07 09:07:55 -08:00
Linfeng Zhang
a9874961f0 Merge "Replace highbd_dct_const_round_shift with dct_const_round_shift" 2016-11-07 16:55:01 +00:00
Johann Koenig
ac495218fb Merge "idct test: use coeff consistently" 2016-11-06 00:13:05 +00:00
Johann Koenig
a139ecd0c9 Merge "partial_idct_test: Add large coefficient test" 2016-11-06 00:13:00 +00:00
Johann Koenig
5d0a271ded Merge "Update vp9_fdct8x8_quant_ssse3 for highbitdepth" 2016-11-06 00:12:13 +00:00
James Zern
6e179dacd0 Merge "vp9-svc: Add unittest for svc-decoding." 2016-11-05 02:47:11 +00:00
Johann
e851160642 idct test: use coeff consistently
Change-Id: I913a13066993a3315a0ff8310b3cad1572d4cdd7
2016-11-04 18:41:59 -07:00
Johann
9ad3e14015 partial_idct_test: Add large coefficient test
Two functions do not pass this test:
vpx_idct8x8_64_add_ssse3
vpx_idct8x8_12_add_ssse3

The test has been modified to avoid triggering an issue with those
functions but they still must be investigated.

BUG=webm:1332

Change-Id: I52569a81e8e6e0b33c4a4d060d0b69c3fc4f578e
2016-11-04 18:37:58 -07:00
Marco
eefc7d1412 vp9-svc: Add unittest for svc-decoding.
To test the VP9_DECODE_SVC_SPATIAL_LAYER decoder control
introduced in 86b0042.

Change-Id: I3d164a41d7bbab14c0aee80fd890870704a18f6e
2016-11-05 01:29:51 +00:00
Johann
e10c95dc83 Update vp9_fdct8x8_quant_ssse3 for highbitdepth
Borrow transition functions from fdct.h nee vpx_quantize_b_sse2

BUG=webm:1304

Change-Id: I9c88c3eec3ff8bb461411d98c26c3c236ea28ef1
2016-11-05 01:23:07 +00:00
Linfeng Zhang
04c3bf3c85 Replace highbd_dct_const_round_shift with dct_const_round_shift
They are identical.

Change-Id: I1ccaf03c81c3cbf88e82d77ffeb8204f5b063c61
2016-11-04 16:15:02 -07:00
Linfeng Zhang
32326c2f13 Merge "Cosmetics of inv_txfm.c" 2016-11-04 22:40:03 +00:00
Johann Koenig
900ec31bea Merge "Extract high bit depth helper functions" 2016-11-04 21:03:17 +00:00
Linfeng Zhang
b68d8107cb Cosmetics of inv_txfm.c
Unify code of 8-bit and high bitdepth.

Change-Id: I3fe441577af0249030ca3a1ef769eb9030711434
2016-11-04 13:24:41 -07:00
Johann
cf35ffc025 Extract high bit depth helper functions
These can be used in the vp9 fdct as well.

Change-Id: I4f3875e0cba1b8cad209c3a0581e121deba7675e
2016-11-04 18:13:51 +00:00
James Zern
232221b83a Merge "configure: disable tools for armv7-win32-vs1[24]" 2016-11-04 17:48:24 +00:00
Martin Storsjo
34c35b6fb6 Add a missing END directive in idct_neon.asm
This fixes building with MS armasm.

Change-Id: I2629eeed859b775ca667a65ba109f8d1bf7b0e03
2016-11-04 12:21:18 +02:00
Martin Storsjo
c559cc6191 Fix producing vcxproj files with *.S arm assembly files
These cases were leftover in
1ddb4c0362.

Change-Id: Ie058fb6c78580e60205c47a1d314bd66e794cde4
2016-11-04 12:21:13 +02:00
James Zern
90a135854c configure: disable tools for armv7-win32-vs1[24]
this shares the same prohibition as the examples

Change-Id: I17d65e4f26847af8cbb1d1a3c4a114ed021a8b9f
2016-11-03 22:54:35 -07:00
Marco Paniconi
cca774c7df Merge "vp9: Non-rd pickmode: fix logic in reference masking." 2016-11-03 23:12:05 +00:00
Marco
86b0042f44 vp9-svc: Add decoder control to decode up to x spatial layers.
Change-Id: I85536473b8722424785c84c5b5520960b4e5744a
2016-11-03 11:18:00 -07:00
Marco
da9f762e24 vp9: Non-rd pickmode: fix logic in reference masking.
Add condition that usable_ref_frame > LAST.
This is to avoid potentially skipping all last-nonzero mv modes,
if golden is used as a reference but skipped completely for the
current block.

This has no effect currenty, as we always consider testing golden
mode for each block.

Change-Id: I3182cf44664081935a90ed43aa7b32e710e60e22
2016-11-03 10:32:57 -07:00
Debargha Mukherjee
f93305aa07 Merge "Speed-up recode loop for extreme bitrate diffs" 2016-11-03 17:04:17 +00:00
Jerome Jiang
cb5a2ac920 Merge "pp_filter_test.cc,cosmetics:adjust name convention" 2016-11-03 04:31:35 +00:00
Jerome Jiang
3e961c09be pp_filter_test.cc,cosmetics:adjust name convention
Change-Id: I81b6fc9b83f0febbb12975aef92768bbd273fd61
2016-11-02 13:50:00 -07:00
Linfeng Zhang
1338c71dfb Clean horizontal intra prediction NEON optimization
Change-Id: I1ef0a5b2655cbc7e1cc2a4a1a72e0eed9aa41f05
2016-11-02 11:43:45 -07:00
Linfeng Zhang
1868582e7d Add 32x32 d45 and 8x8, 16x16, 32x32 d135 NEON intra prediction
Change-Id: I852616794244490123eb615ac750da50265f0fa5
2016-11-02 11:40:37 -07:00
Johann Koenig
5ac7a59a05 Merge "arm idct: move to-be-shared code to header" 2016-11-02 18:09:45 +00:00
Paul Wilkins
295cd3b493 Merge "Fixed bug in formatting of debug stats." 2016-11-02 17:10:07 +00:00
paulwilkins
de76d2e315 Fixed bug in formatting of debug stats.
Fixed formatting bug introduced by the fix to BUG=webm:1322
( Iedc4477aef1746aa0a4f84d88a1156296fd3ba87)

Change-Id: I715ee446c0e8584967ab87ba4e355759dd394187
2016-11-02 09:38:18 +00:00
James Zern
1961a92a94 vp9,tile_worker_hook: correctly set jmp target
vp9_init_macroblockd() resets the error_info to cm's global copy; this
needs to be set to the thread-level target to avoid jumping to the
incorrect stack, resulting in hang or crash.
broken since:
1f4a6c8 vp9/tile_worker_hook: add multiple tile decoding
includes v1.5.0, v1.6.0

BUG=629481

Change-Id: Icbf1696b25ba8c479e845fbf227b3c3ca73542f5
2016-11-01 18:45:50 -07:00
Linfeng Zhang
3b74066b10 Add high bitdepth intra prediction NEON optimization (mode dc)
BUG=webm:1316

Change-Id: I984d6004ea2445e86f213fb6fa4d794a9955af8f
2016-11-01 17:07:36 -07:00
Johann
bf8ab194ee arm idct: move to-be-shared code to header
Change-Id: I67458cd358b4dc4434bbdbfcdd571769561b619e
2016-11-01 15:43:56 -07:00
James Zern
1b275ab898 Merge "idct32x32_1_add_neon: clear a couple conv warnings" 2016-11-01 22:34:59 +00:00
James Zern
9de91855ef Merge changes I08af3a54,If5959a25,I6763e62e
* changes:
  build/make/Android.mk: s/armv8/arm64/
  build/make/Android.mk: fix armeabi-v7a build
  use .S suffix rather than .s for NEON asm
2016-11-01 21:43:13 +00:00
Linfeng Zhang
05ee241493 Add high bitdepth intra prediction optimization speed test
BUG=webm:1316

Change-Id: I99feec867d5b8ea06b43cdd3fcd7c90238f5efdb
2016-11-01 13:57:01 -07:00
Linfeng Zhang
0c88014592 Merge "Refine 8-bit intra prediction NEON optimization (mode tm)" 2016-11-01 19:38:07 +00:00
Linfeng Zhang
cc5f49767a Refine 8-bit intra prediction NEON optimization (mode tm)
Change-Id: I98b9577ec51367df5e5d564bedf7c3ea0606de4c
2016-11-01 09:45:16 -07:00
Paul Wilkins
84dcfced5b Merge "Change to KF boost calculation." 2016-11-01 09:29:30 +00:00
James Zern
7625c803b3 idct32x32_1_add_neon: clear a couple conv warnings
int16_t -> uint8_t

Change-Id: I3c5e0985bc3584dce289c35b5973de24cdc73b76
2016-10-31 18:56:34 -07:00
James Zern
2e076ffe50 build/make/Android.mk: s/armv8/arm64/
the configure target is arm64-android-gcc which generates .mk files of
the same form

Change-Id: I08af3a54ef203b1496d185a0f8c8fe702881a173
2016-10-31 18:35:23 -07:00
James Zern
ae32318170 build/make/Android.mk: fix armeabi-v7a build
vpx_config.asm and idct_neon.asm.S are required since:
3ae2597 idct,NEON: add a tran_low_t->s16 load adapter

Change-Id: If5959a25edb370dd7dcdca71c96e9a5aad0840ce
2016-10-31 18:34:16 -07:00
James Zern
1ddb4c0362 use .S suffix rather than .s for NEON asm
for compatibility with other build systems

Change-Id: I6763e62e3126850ad4f8ad29e388b8dad0bbc4c3
2016-10-31 16:39:05 -07:00
Marco Paniconi
7cf0c000cf Merge "vp9-svc: Fix some stats in vp9_spatial_svc_encoder." 2016-10-31 23:16:36 +00:00
Marco
41ad80f69d vp9-svc: Fix some stats in vp9_spatial_svc_encoder.
Correction to rate control stats output under -rcstat.

Change-Id: I46fa5d2a66ed657121ee3d685608a148bb9a7bb3
2016-10-31 15:24:13 -07:00
James Zern
410d947c5f Merge "idct,NEON: add a tran_low_t->s16 load adapter" 2016-10-31 21:59:12 +00:00
Peter Boström
11b099ea46 Merge "Add vp9_spatial_svc_encoder to .gitignore." 2016-10-31 20:01:11 +00:00
Linfeng Zhang
cde5d5db13 Merge "Refine 8-bit intra prediction NEON optimization (mode h and v)" 2016-10-31 19:57:23 +00:00
Peter Boström
39bcb49909 Add vp9_spatial_svc_encoder to .gitignore.
Change-Id: I3c90d657cca533264dd62bb7749c53a862d0352a
2016-10-31 14:55:56 -04:00
Marco Paniconi
702b3e1ee5 Merge "vp9-svc: Add checks to layer bitrates in vp9_spatial_svc_encoder." 2016-10-31 18:23:14 +00:00
James Zern
3ae25974fd idct,NEON: add a tran_low_t->s16 load adapter
enable idct4x4* and idct8x8* which are compatible for 8-bit decodes in
high-bitdepth mode. the adapter narrows 32-bit input to 16, whether the
expansion can be avoided at all in this case remains a TODO. roughly
matches sse2.

BUG=webm:1294

Change-Id: I3ea94e5a2070dfd509b5de0c555aab4e1f4da036
2016-10-31 11:21:16 -07:00
Linfeng Zhang
a347118f3c Refine 8-bit intra prediction NEON optimization (mode h and v)
Change-Id: I45e1454c3a85e081bfa14386e0248f57e2a91854
2016-10-31 10:33:44 -07:00
Marco
e1cdb50298 vp9-svc: Add checks to layer bitrates in vp9_spatial_svc_encoder.
Add some checks to the layer bitrates passed in through the command line.

Change-Id: I16f270035a6034d63313fe3019aa90dca9a3eefb
2016-10-31 10:07:24 -07:00
James Bankoski
fb9fef83c7 Merge "vpxdec.c : don't double count corrupted frames" 2016-10-31 13:58:15 +00:00
Jim Bankoski
30f3017697 vpxdec.c : don't double count corrupted frames
A past patch made it so that every frame that had a decode error
caused a corrupted frame to be counted.  Unfortunately it was possible
to get both a decode error and a corrupt frame for the same frame
and thus double count an error. This code makes that impossible.

Change-Id: Iea973727422a3bf093ffda72fa358a285736048b
2016-10-31 06:09:58 -07:00
James Zern
086aab7e13 tiny_ssim: fix visual studio build
s/inttypes.h/vpx_integer.h/
clear a uint64_t -> double conversion warning

Change-Id: I58d108b083787a754152eb79ef6df61c2c5f95b1
2016-10-29 13:04:07 -07:00
Peter Boström
ae206924a6 Merge "Add temporal-layer support to tiny_ssim." 2016-10-29 01:25:01 +00:00
Johann Koenig
5724e8c4c7 Merge "partial_idct_test: add _add_ test" 2016-10-29 00:34:16 +00:00
Peter Boström
fd4efde489 Add temporal-layer support to tiny_ssim.
Permits skipping 0, 1/2 or 3/4 of the frames, corresponding to
temporal layers 2, 1 and 0 of a 3-temporal-layer encoding. 1/2
corresponds to TL0 in a 2-layer encoding.

Change-Id: I7f6d131f63707e5262fc67d111bfb3a751ede90d
2016-10-28 14:56:05 -04:00
Marco Paniconi
7042137e60 Merge "vp9: Updates to SVC sample encoder." 2016-10-28 18:32:41 +00:00
Marco
a8fdb3926e vp9: Updates to SVC sample encoder.
Allow for passing in the layer bitrates at command line.
Fix to allow passing in bitrate for each spatial-temporal layer.

Change to some default values for 1 pass cbr mode:
spatial scale and qp-max/min.

Small fixes to some build warnings.

Change-Id: I3f9a776262712480a6570bb863a835b2fc49935a
2016-10-28 10:59:58 -07:00
Yaowu Xu
9205f54744 Merge "Add tools/tiny_ssim for generating SSIM/PSNR." 2016-10-28 17:21:48 +00:00
Peter Boström
7c75cae74a Add tools/tiny_ssim for generating SSIM/PSNR.
Change-Id: Icc3e5aaa6636ffe17dc9da5f7a80afaccbde509a
2016-10-28 12:39:49 -04:00
Linfeng Zhang
4d305dab34 Merge "Refine 8-bit intra prediction NEON optimization (mode d45 and d135)" 2016-10-28 15:58:01 +00:00
Paul Wilkins
715c65914b Change to KF boost calculation.
This  change is a step in a larger change to the way boost and interval are
determined for ARF and Key frames.

This patch contains some pluming for the general case but focuses on the
key frame boost calculation. This now relies more heavily on the rate at
which the error score increases between the primary and secondary reference
frame. This seems to be less fragile when dealing with different frame sizes.
For example larger image formats tend in the first pass to see a higher
% of intra coded blocks and the use of this number in calculating the frame
decay factor was leading to much lower boost numbers for 4K, for example,
than the same clip coded at 2K.

This change does give overall gains but they are MUCH larger for the 4K Netflix
set. For the 4K Netflix set the average gain is around 3% with some clips > 20%
whereas for the same set at 2K the average gain is 0.5-1%.

In general for small image formats the boost is most often reduced a little whereas
4K clips the boost is increased. There are some -ve cases such as Akiyo at 352x288
where the reduced boost hurts the metrics, especially for SSIM, even while
the set as a whole improves. This is most notable at very low Q and may be the
subject of a future patch.

Some common code for KF and ARF was separated in this patch for the purposes of
tuning but may later be re-merged if appropriate.

Change-Id: Iaa15ac5a58d2be89181100d95cef6a8dc4b12d0d
2016-10-28 15:35:59 +01:00
Debargha Mukherjee
4f7a59c802 Merge "Force recode if framesize exceeds max allowed size" 2016-10-28 04:21:44 +00:00
Linfeng Zhang
4ae9f5c092 Refine 8-bit intra prediction NEON optimization (mode d45 and d135)
dst += stride behaving better with gcc/clang.
Unroll loops.

Change-Id: I83f85df2bc9f17c6159542f57680b509395db2b1
2016-10-27 14:24:50 -07:00
Debargha Mukherjee
1cd987d922 Speed-up recode loop for extreme bitrate diffs
Adjusts the q adjustement step depending on how far the
projected and target rates differ.

Change-Id: I498d03523ca233a270512ca3972c372daa4ca2a8
2016-10-27 11:08:44 -07:00
Debargha Mukherjee
54e03017b6 Force recode if framesize exceeds max allowed size
Fixes a case where recode is not triggered based on the value
of maxq passed into the recode loop test function.

BUG=b/32375284

Change-Id: I15ad985d0525c68e0443cfaf842440d2754b2266
2016-10-27 09:52:51 -07:00
Johann Koenig
4555c50ecd Merge "partial_idct_test: consolidate block size" 2016-10-27 15:41:44 +00:00
Paul Wilkins
aadfde4687 Merge "Changes to KF boost calculation." 2016-10-27 10:21:23 +00:00
Paul Wilkins
02deeea447 Merge "Removal of a couple of two pass adjustments." 2016-10-27 10:21:06 +00:00
Johann
7994dba6c0 partial_idct_test: add _add_ test
The result of the transform is added to the destination buffers. In the
existing tests the destination buffer is always empty so that portion of
the code was never exercised.

Change-Id: I1858c4fed2274f1b9faf834d2ba4186a4510492a
2016-10-26 21:35:49 -07:00
Johann
ed2c240538 partial_idct_test: consolidate block size
Use *input_block_ for sizeof() calculation like the other test

Change-Id: I1e4bd227131662056405af78c5052ad6ef769e9f
2016-10-26 21:35:03 -07:00
Johann
08e0da30ca Refactor partial idct test
Switch to using correctly sized inputs and outputs. This simplifies
adding tests with varying strides.

Change-Id: I716a0d8173dcf6a86d56656ac9d3101b7ec27642
2016-10-26 12:28:18 -07:00
Paul Wilkins
de859676dd Changes to KF boost calculation.
Remove double counting of decay. Limit maximum KF boost.

Change-Id: I0fb2344d0f78b5e95bb899dfad12b0ca84034b2c
2016-10-26 17:53:29 +01:00
paulwilkins
ccd6a8e2fa Removal of a couple of two pass adjustments.
Removed a couple of adjustments that no longer move the needle
much but complicate the process of tuning.

Change-Id: Ie320f5cf155e6aac14a4757ea9ada2cd59f27590
2016-10-26 17:52:37 +01:00
Linfeng Zhang
9c0680bd43 Merge "Refine 8-bit intra prediction NEON optimization (mode dc)" 2016-10-26 16:51:44 +00:00
Johann
9720b58aac Optimize idct32x32_34_add for NEON
Approximately 3 times faster than the 1024 version which was used
previously.

BUG=webm:1295

Change-Id: Id15fb3d096029ec38ef01c53e5f6eb08254347c9
2016-10-25 15:43:58 -07:00
James Zern
98ffc49204 Merge "Update vp9_intrapred_test.cc to support 8-bit" 2016-10-25 21:59:29 +00:00
Yunqing Wang
4b8b1bae52 Merge "Modify the encoder multi-thread unit test" 2016-10-25 20:54:26 +00:00
Yunqing Wang
c327b3f0b0 Modify the encoder multi-thread unit test
Modified the encoder multi-thread test so that it included cpu-used=0 and
frame-parallel=0.

frame_parallel_decoding_mode is 1 by default, which disables probability
updating and gives lower encoding quality. Current VP9 multi-threading
encoder and decoder support probability updating. To test this part, we
should turn on it in the unit test, namely, setting frame-parallel to 0.

Change-Id: Ia1f86e01f0de628f50d819ae31509de3e1b6c755
2016-10-25 11:35:01 -07:00
James Bankoski
f53d3363ac Merge "vpxdec: return fail if frame fails to decode." 2016-10-25 18:34:07 +00:00
Yunqing Wang
c192def8f3 Change 2 motion search counts to be tile data
This patch modified the motion search counts used in:
https://chromium-review.googlesource.com/#/c/305640/

These 2 counts were originally added as thread data, and used to
make decisions in motion search. The tile encoding order can be
inconsistent while using different number of threads, which can
cause bitstream mismatch. Here moved them to tile data to solve
the issue.

BUG=webm:1322

Change-Id: Iedc4477aef1746aa0a4f84d88a1156296fd3ba87
2016-10-25 10:12:41 -07:00
Linfeng Zhang
ce88b8f5c5 Refine 8-bit intra prediction NEON optimization (mode dc)
dst += stride behaving better with gcc/clang
Expanding inline function dc_SIZExSIZE() save intructions for
vpx_dc_predictor_SIZExSIZE_neon().

Change-Id: Id0ccbd58b6a31df539141fd33bdf28633339150d
2016-10-24 13:18:51 -07:00
Linfeng Zhang
d1c74c149b Update vp9_intrapred_test.cc to support 8-bit
BUG=webm:1316

Change-Id: Ic9309bbeeef52e9d07fb4a4c95c12efa813cbf8c
2016-10-24 13:13:55 -07:00
Jim Bankoski
7ef094c02f vpxdec: return fail if frame fails to decode.
A failure to decode is most likely equivalent to a corrupt
frame for the purpose of returning a failure.

Change-Id: Ie53db2b8130b40b725841f5f7a299d63aa56913d
2016-10-24 12:05:59 -07:00
James Zern
2e6a1976a0 Merge "remove idct32x32*_add_neon.asm" 2016-10-22 02:29:56 +00:00
James Zern
5d91752a98 Merge "vpx_highbd_convolve_copy_neon: use multi reg loads" 2016-10-22 02:28:15 +00:00
Vignesh Venkatasubramanian
9a032fa262 Merge "vp9_bitstream: Encode tiles in parallel" 2016-10-22 02:23:06 +00:00
Vignesh Venkatasubramanian
5deffa1175 vp9_bitstream: Encode tiles in parallel
Re-use the tile worker threads to pack the bitstream in parallel
on a per-tile basis.  Restricting this to real-time only for now
(further testing is needed to ensure this does not make 2-pass
worse in any case).

BUG=webm:1309

Change-Id: I8a80da7c5089b837d0df79a5c49d5e3022dfc8ec
2016-10-21 17:35:03 -07:00
Marco
ee1b3f34c0 vp9: Nonrd variance partition: increase threshold for using 4x4 avg.
In variance partition low resolutions may use varianace based on
4x4 average for better partitioning.
Increase the threshold for doing this at speed = 8.

Improves speed by ~5%, with little loss, < 1%, on RTC_derf set.

Change-Id: Ib5ec420832ccff887a06cb5e1d2c73199b093941
2016-10-21 11:51:06 -07:00
James Zern
9dbb3ad396 remove idct32x32*_add_neon.asm
the intrinsics are neutral to ~20% faster on cros/android
devices when using gcc-4.9/clang-3.8.1 and gcc-4.9/clang-3.8.x from the
r13 ndk. neutral results typically came with gcc-4.9 while larger
positive gains were achieved with clang 3.8.x.

BUG=webm:1303

Change-Id: I4d31f9c017944681b881493525d4573a7a5b1e16
2016-10-20 19:47:14 -07:00
Marco
a7d116aa67 vp9: Speed=8 real-time: Keep the bias_golden feature on.
Small/no change in metrics on RTC set, speed increase by 2-3%.

Change-Id: Iee997bd7433e8e508216e9267b1c31c5a9aa5121
2016-10-20 17:03:51 -07:00
Marco Paniconi
32e63efcfb Merge "vp8: Apply gf target-size boost only when refresh_golden_frame = 1." 2016-10-20 22:38:55 +00:00
Vignesh Venkatasubramanian
83ca63582a Add vp9cx_set_ref to .gitignore
Get rid of the 'git status' clutter when building with examples.

Change-Id: I20b715ddfc6c8ccb4993de7ebb2b4ad6df9ea437
2016-10-20 12:07:30 -07:00
Marco
9fdae93858 vp8: Apply gf target-size boost only when refresh_golden_frame = 1.
Change only affects 1 pass cbr, error resilience off.

Change-Id: I68b896b09d722995a71c44331233e97bd862bcfc
2016-10-20 11:32:29 -07:00
James Zern
995a967f19 Merge "third_party: roll libwebm snapshot" 2016-10-19 22:34:25 +00:00
Marco
9624964832 vp8: Adjust threshold to set the gf_noboost flag.
Change only affects 1 pass cbr, with error_resilient off.

Change-Id: Ibf254d8772fa2a8f188c9932d37b2f42362d8003
2016-10-19 12:55:37 -07:00
Marco
ff38b8dfae vp8: Add control for gf boost for 1 pass cbr.
Control already exists for vp9, adding it to vp8.
Usage is only when error_resilient is off.
Added a datarate unittest for non-zero boost.

Change-Id: I4296055ebe2f4f048e8210f344531f6486ac9e35
2016-10-19 09:43:53 -07:00
James Zern
7f31bfeddb Revert "vp9_bitstream: Encode tiles in parallel"
This reverts commit 9e8efa5b18.

this change causes ubsan warnings, failures in
vpxenc_vp9_webm_rt_multithread_tiled

BUG=webm:1309

Change-Id: I020c7be985c771bfff4b3de1afe51cc8edb980da
2016-10-18 22:47:48 -07:00
James Zern
68833c7f85 third_party: roll libwebm snapshot
git log --no-merges 32d5ac4..9732ae9
9732ae9 EbmlElementSize: quiet uint64->int32 conv warning
da04eba SetProjectionPrivate: quiet uint64->size_t conv warning
6db32d5 mkvparser,Projection::Parse: fix int->bool conv
3bb0dfa cosmetics: fix a couple lint warnings
0e179d6 update .clang-format
fc5f88d Fix temp files being left on system.
c04a134 Add support for overriding PixelWidth and PixelHeight.
c0160e0 Add support to explicitly set segment duration.
02bc809 Add support to estimate file duration.
c97e3e7 Add support to output sub-sample encryption information.
26f4344 MakeUID: quiet unused param warning in Android builds
d6af52a Change check to fix compile error.
1720020 webm_parser: Add Mesh value for ProjectionType
78f2c5a webm_parser: Use ./ prefix for includes
da62f65 webm_parser: Remove webm/ prefix from public includes
e15e8f2 webm_parser: Update README build instructions
5023f2b mkvmuxer: Fix Colour::Valid()
cf16204 mkvmuxer_tests: Actually test cue points in the cue point test.
93e9fb3 Validate Colour element values.
8036925 mkvparser_tests: Add Projection element test.
f52d38c mkvparser_tests: Add Colour element test.
826436a mkvparser: minor SeekHead::Entry clean up.
24fb44a mkvmuxer_tests: Add Projection element test.
1e0a8ea mkvmuxer_tests: Add Colour element test.
0278616 mkvmuxer: Colour accessors/mutators.
2346f8f Add mkvparser wrapper functions.
54d6b6b webm_info: Add Projection element support.
65fee06 mkvmuxer_sample: Add support for Projection element.
9a3f2b5 mkvparser_sample: Add support for Projection element.
41e814a mkvparser: Add Projection element support.
483a0ff mkvmuxer: Add Projection element support.
676a713 Add support for the Projection element
725f362 mkvmuxer: Fix memory leak when Colour is set multiple times.
fa182de mkvparser_sample: Add output of audio track codec private size.
8f521f2 mkvparser_tests: Add invalid BlockGroup test.
39137d7 Remove docs saying binary elements default to 0
80685d3 Do not skip over unknown elements at the root level
c147504 Fix legacy Makefile.
58711e8 mkvparser_sample: Fix version info string.
837746f mkvparser_tests: Add invalid block test.
207cd80 Disambiguate sample sources and targets.
a112d71 mkvparser_tests: Refactor invalid file loading code.
5dea33e Disambiguate test source and target names.
125049e parser_tests: Add another truncated chapter string test.
1de8d4c parser_tests: Add truncated chapter string test.
ff8c2b6 parser_tests: Move cue validation to test_util.
4b0690f parser_tests: Add invalid lacing test.
9828e39 mkvmuxer: Set default doc type version to 4.
5495a59 webm_parser: Reference more files in CMakeLists.txt.
0c0ecd0 vpxpes_parser: Add start code emulation prevention support.
639a4bc webm2pes: Remove debug printfs().
9a51102 webm2pes: fflush() in the correct conversion function.
dc7f155 webm2pes: Track total bytes written.
d518128 webm_parser: Enable usage of werror.
e1fe762 webm2pes: Add test for mux/demux of large input.
1b24a79 vpxpes_parser: Read and store PTS when present.
6cf0a0f vpxpes_parser: Store frame payloads.
25d2602 webm_parser: Convert style to match the rest of libwebm
24be76d webm2pes: Replace VpxFrame with VideoFrame.
b451c3b Add a basic video frame storage class.
05c90eb libwebm_util: Clarify error text in superframe parser.
e6415af webm2pes: Make WritePesPacket() a public method.
8f840dd webm2pes: Move frame read out of PES packet write method.
448af97 webm2pes: Restore frame fragmentation support.
f8bb714 cmake: Integrate new parsing API and tests.
cb8ce0b Add a new incremental parsing API
900d322 vpxpes_parser/webm2pes: BCMV and PTS fixes.
4b73545 webm2pes: Add start code emulation prevention.
82903f3 Add column tiles and frame parallel to webm_info
5d91edf style_clean_up: Remove unnecessary parentheses
a95aa4b vp9_level_stats: correct total_uncompressed_bits_ calculation
f46566f mkvreader: Fix shorten-64-to-32 warning in 32 bit builds.
76630ca mkvwriter: Fix shorten-64-to-32 warning in 32 bit builds.
a8ffbd4 webm2pes: Fix format specifier warnings.
faf89d4 Add MaxLumaSampleRate grace percent to stats.
d31e6c9 Fix profile 2 in vp9_header_parser.
bd3ab3a Add flag to estimate last frame's duration to stats.
c182ed9 Fix lint issue in hdr_util.h
cc62ecd Add test for Cluster memory leak
196708a Change MaxLumaSampleRate to be based on frame resolution.
cbd676b mkvmuxer: Fix leak when a Cluster isn't finalized
9a235e0 mkvmuxer: Set doctype to matroska when muxing non-WebM codecs.
47f2843 Add parsing support for new features in CodecPrivate.
e3c9576 Add VP9 level output to webm_info.
5cf549f cmake: Log compiler flag at check time.
bbaaf2d Add class to gather VP9 level stats.
8bb68c2 Add file to parse data from VP9 frames.
296429a Add support to parse VP9 profile.
df3412f Add support for setting VP9 profile and level to sample_muxer.
87832d4 mkvmuxer: Fix Segment::Finalize in kLive mode
6df3e56 mkvmuxerutil.hpp: Add using directives for overloaded size utils.
ec47928 mkvmuxerutil: Revert to using mkvmuxertypes.
a1dc4f2 Fix parsing of VP9 level.
4e3d037 Add support to output Colour elements to webm_info.
d3656fd muxer_tests: ignore iwyu re gtest-message.h
e76dd5e Fix file name in mkvmuxertypes shim.
1be5889 Add temporary include shims at old file locations.
039df94 Add TEST_TMPDIR environment variable

Change-Id: I84bc1401b0aad71ad6727b687f1bede9953a7a08
2016-10-18 18:11:36 -07:00
James Zern
a60dd5c83a Merge "Fix warnings reported by -Wshadow: Part1: vpx_dsp directory" 2016-10-18 22:09:29 +00:00
James Zern
53d8ff6f14 Merge "Revert "third_party: Roll libwebm snapshot."" 2016-10-18 20:06:48 +00:00
Kaustubh Raste
8ff5af773a Merge "Optimize sad_64width_x4d_msa function" 2016-10-18 07:46:02 +00:00
James Zern
171e2ccf99 Revert "third_party: Roll libwebm snapshot."
This reverts commit 808a560be6.

causes build warnings under visual studio

Change-Id: I2e49a75d72469f316e8b01929b783e6f727f756c
2016-10-17 23:24:47 -07:00
Kaustubh Raste
b7310e2aff Optimize sad_64width_x4d_msa function
Reduced HADD_UH_U32 macro calls

Change-Id: Ie089b9a443de516646b46e8f72156aa826ca8cfa
2016-10-18 04:05:33 +00:00
Urvang Joshi
e084e05484 Fix warnings reported by -Wshadow: Part1: vpx_dsp directory
While we are at it:
- Rename some variables to more meaningful names
- Reuse some common consts from a header instead of redefining them.

Change-Id: I75c4248cb75aa54c52111686f139b096dc119328
(cherry picked from aomedia 09eea21)
2016-10-17 19:25:19 -07:00
James Zern
68cd3052ca vpx_highbd_convolve_copy_neon: use multi reg loads
for copy16/32/64

BUG=webm:1299

Change-Id: I5080d736bde7e487c80ef3d7024dda1e96a57eaf
2016-10-17 17:15:03 -07:00
Marco Paniconi
f6980ca68e Merge "vp9: Non-rd variance partition: add condition for 64x64 split." 2016-10-18 00:03:17 +00:00
Linfeng Zhang
b0cc8d5cc6 Merge "add vpx high bitdepth convolve8 NEON intrinsics optimization" 2016-10-17 23:57:14 +00:00
Linfeng Zhang
9c8981c666 add vpx high bitdepth convolve8 NEON intrinsics optimization
BUG=webm:1299

Change-Id: I236bfa0441e357b6ff05add8269a2cfb543924d1
2016-10-17 15:23:54 -07:00
Marco
55a2b67368 vp9: Non-rd variance partition: add condition for 64x64 split.
Add stronger condition for splitting 64x64, for low noise content.
This reduces dragging artifact near moving head.

Little/no change in metrics on RTC set.

Change-Id: I39b38cfd20f2ece53ff49c2aaf76ba9f82761be1
2016-10-17 12:54:27 -07:00
Frank Galligan
808a560be6 third_party: Roll libwebm snapshot.
fc5f88d Fix temp files being left on system.
c04a134 Add support for overriding PixelWidth and PixelHeight.
c0160e0 Add support to explicitly set segment duration.
02bc809 Add support to estimate file duration.
c97e3e7 Add support to output sub-sample encryption information.
26f4344 MakeUID: quiet unused param warning in Android builds
d6af52a Change check to fix compile error.
1720020 webm_parser: Add Mesh value for ProjectionType
78f2c5a webm_parser: Use ./ prefix for includes
da62f65 webm_parser: Remove webm/ prefix from public includes
e15e8f2 webm_parser: Update README build instructions
5023f2b mkvmuxer: Fix Colour::Valid()
cf16204 mkvmuxer_tests: Actually test cue points in the cue point test.
93e9fb3 Validate Colour element values.
8036925 mkvparser_tests: Add Projection element test.
f52d38c mkvparser_tests: Add Colour element test.
826436a mkvparser: minor SeekHead::Entry clean up.
24fb44a mkvmuxer_tests: Add Projection element test.
1e0a8ea mkvmuxer_tests: Add Colour element test.
0278616 mkvmuxer: Colour accessors/mutators.
2346f8f Add mkvparser wrapper functions.
54d6b6b webm_info: Add Projection element support.
65fee06 mkvmuxer_sample: Add support for Projection element.
9a3f2b5 mkvparser_sample: Add support for Projection element.
41e814a mkvparser: Add Projection element support.
483a0ff mkvmuxer: Add Projection element support.
676a713 Add support for the Projection element
725f362 mkvmuxer: Fix memory leak when Colour is set multiple times.
fa182de mkvparser_sample: Add output of audio track codec private size.
8f521f2 mkvparser_tests: Add invalid BlockGroup test.
39137d7 Remove docs saying binary elements default to 0
c147504 Fix legacy Makefile.
80685d3 Do not skip over unknown elements at the root level
58711e8 mkvparser_sample: Fix version info string.
837746f mkvparser_tests: Add invalid block test.
207cd80 Disambiguate sample sources and targets.
a112d71 mkvparser_tests: Refactor invalid file loading code.
5dea33e Disambiguate test source and target names.
125049e parser_tests: Add another truncated chapter string test.
1de8d4c parser_tests: Add truncated chapter string test.
ff8c2b6 parser_tests: Move cue validation to test_util.
4b0690f parser_tests: Add invalid lacing test.
9828e39 mkvmuxer: Set default doc type version to 4.
5495a59 webm_parser: Reference more files in CMakeLists.txt.
0c0ecd0 vpxpes_parser: Add start code emulation prevention support.
639a4bc webm2pes: Remove debug printfs().
9a51102 webm2pes: fflush() in the correct conversion function.
dc7f155 webm2pes: Track total bytes written.
d518128 webm_parser: Enable usage of werror.
e1fe762 webm2pes: Add test for mux/demux of large input.
1b24a79 vpxpes_parser: Read and store PTS when present.
6cf0a0f vpxpes_parser: Store frame payloads.
25d2602 webm_parser: Convert style to match the rest of libwebm
24be76d webm2pes: Replace VpxFrame with VideoFrame.
b451c3b Add a basic video frame storage class.
05c90eb libwebm_util: Clarify error text in superframe parser.
e6415af webm2pes: Make WritePesPacket() a public method.
8f840dd webm2pes: Move frame read out of PES packet write method.
448af97 webm2pes: Restore frame fragmentation support.
f8bb714 cmake: Integrate new parsing API and tests.
cb8ce0b Add a new incremental parsing API
900d322 vpxpes_parser/webm2pes: BCMV and PTS fixes.
4b73545 webm2pes: Add start code emulation prevention.
82903f3 Add column tiles and frame parallel to webm_info
5d91edf style_clean_up: Remove unnecessary parentheses
a95aa4b vp9_level_stats: correct total_uncompressed_bits_ calculation
f46566f mkvreader: Fix shorten-64-to-32 warning in 32 bit builds.
76630ca mkvwriter: Fix shorten-64-to-32 warning in 32 bit builds.
a8ffbd4 webm2pes: Fix format specifier warnings.
faf89d4 Add MaxLumaSampleRate grace percent to stats.
d31e6c9 Fix profile 2 in vp9_header_parser.
bd3ab3a Add flag to estimate last frame's duration to stats.
c182ed9 Fix lint issue in hdr_util.h
cc62ecd Add test for Cluster memory leak
196708a Change MaxLumaSampleRate to be based on frame resolution.
cbd676b mkvmuxer: Fix leak when a Cluster isn't finalized
47f2843 Add parsing support for new features in CodecPrivate.
9a235e0 mkvmuxer: Set doctype to matroska when muxing non-WebM codecs.
e3c9576 Add VP9 level output to webm_info.
bbaaf2d Add class to gather VP9 level stats.
5cf549f cmake: Log compiler flag at check time.
8bb68c2 Add file to parse data from VP9 frames.
df3412f Add support for setting VP9 profile and level to sample_muxer.
296429a Add support to parse VP9 profile.
87832d4 mkvmuxer: Fix Segment::Finalize in kLive mode
6df3e56 mkvmuxerutil.hpp: Add using directives for overloaded size utils.
ec47928 mkvmuxerutil: Revert to using mkvmuxertypes.
4e3d037 Add support to output Colour elements to webm_info.
a1dc4f2 Fix parsing of VP9 level.
039df94 Add TEST_TMPDIR environment variable
d3656fd muxer_tests: ignore iwyu re gtest-message.h
e76dd5e Fix file name in mkvmuxertypes shim.
1be5889 Add temporary include shims at old file locations.

Change-Id: I6a1026814560be80d604a5ecb9b66406a1186dd9
2016-10-17 12:45:05 -07:00
Vignesh Venkatasubramanian
9e8efa5b18 vp9_bitstream: Encode tiles in parallel
Re-use the tile worker threads to pack the bitstream in parallel
on a per-tile basis.  Restricting this to real-time only for now
(further testing is needed to ensure this does not make 2-pass
worse in any case).

BUG=webm:1309

Change-Id: Ia2c982da56697756e12f02643f589189b3271d98
2016-10-17 10:42:03 -07:00
Jerome Jiang
4c3d539baa Merge "VP8: Add realtime speed to datarate_test.cc" 2016-10-15 06:01:41 +00:00
Jerome Jiang
acd21e053a VP8: Add realtime speed to datarate_test.cc
Change-Id: Ia56f0e8dfba20143be3e69666d9184dd3ca5b563
2016-10-14 17:09:27 -07:00
Linfeng Zhang
6c309c1f59 Merge "add vpx_highbd_convolve_{copy,avg}_neon()" 2016-10-14 23:04:59 +00:00
James Bankoski
e49a02b113 Merge "Drop empty frames." 2016-10-14 16:38:56 +00:00
Jim Bankoski
3e21d703ce Drop empty frames.
Change-Id: I2d45a6eb3aaca97eb61e8e7ef9e5114221091244
2016-10-14 06:28:14 -07:00
Linfeng Zhang
f910d14a1a add vpx_highbd_convolve_{copy,avg}_neon()
BUG=webm:1299

Change-Id: Ib87ac466ada63251eb06ae2abd1e13e61e0d1538
2016-10-13 15:21:14 -07:00
Marco
f5b8b473db vp8: Adjust thresholds in VP8/DatarateTestLarge tests.
Fix unit_tests_ubsan failure VP8/DatarateTestLarge.BasicBufferModel.
Failure was triggered by commit: df66f8e8.

Change-Id: I2c49e5cc24094b15063161bab27b09ec7e6f2045
2016-10-13 09:28:40 -07:00
James Zern
1909270f65 Merge "cosmetics,*loopfilter_neon.c: s/tranpose/transpose/" 2016-10-13 07:12:51 +00:00
Vignesh Venkatasubramanian
3e3475321c Merge "vp9_bitstream: Parameterize interp_filter_selected" 2016-10-13 06:33:31 +00:00
Vignesh Venkatasubramanian
769292017b vp9_bitstream: Parameterize interp_filter_selected
Facilitates encoding tiles in parallel.

BUG=webm:1309

Change-Id: I37aa336d47babffc8352188dc767eebdb8a99474
2016-10-12 20:22:03 -07:00
Kaustubh Raste
9e75c01353 Merge "Optimize vpx_mbpost_proc_across_ip_msa function" 2016-10-13 02:12:33 +00:00
Kaustubh Raste
99adf8b22e Merge "Optimize vpx_get4x4sse_cs_msa function" 2016-10-13 02:12:00 +00:00
James Zern
fd270437f0 cosmetics,*loopfilter_neon.c: s/tranpose/transpose/
Change-Id: I267d6a9d715ddb6110f0881c2e820c37fc673fe1
2016-10-12 16:12:56 -07:00
Vignesh Venkatasubramanian
04a6010742 Merge "vp9_bitstream: Parameterize max_mv_magnitude" 2016-10-12 21:52:42 +00:00
Vignesh Venkatasubramanian
d03d1c8cd3 vp9_bitstream: Parameterize max_mv_magnitude
Facilitates encoding tiles in parallel.

BUG=webm:1309

Change-Id: I614a5a492c30b6773c30e7294cd6a6f456e02ab4
2016-10-12 12:50:17 -07:00
Linfeng Zhang
b894d95b32 Merge "[vpx highbd lpf NEON 6/6] vertical 16" 2016-10-12 19:31:39 +00:00
Linfeng Zhang
f664d3f6d5 Merge "[vpx highbd lpf NEON 5/6] horizontal 16" 2016-10-12 19:31:25 +00:00
Linfeng Zhang
3b06acd4e2 Merge "[vpx highbd lpf NEON 4/6] vertical 8" 2016-10-12 19:31:06 +00:00
Linfeng Zhang
01454ec485 [vpx highbd lpf NEON 6/6] vertical 16
BUG=webm:1300

Change-Id: I29d0b482d66f05e278325ddebcf108fbf0b6e222
2016-10-11 22:59:19 -07:00
Linfeng Zhang
27479775c4 [vpx highbd lpf NEON 5/6] horizontal 16
BUG=webm:1300

Change-Id: I21da32d6cfb8a1a6f58bc9756d17f48f13a59a12
2016-10-11 22:59:19 -07:00
Linfeng Zhang
251cbfbec8 [vpx highbd lpf NEON 4/6] vertical 8
BUG=webm:1300

Change-Id: If06b12bc081bab60059b100414dd7018f83ac62d
2016-10-11 22:59:19 -07:00
Kaustubh Raste
56b1be1889 Merge "Optimize vp8 loopfilter msa functions" 2016-10-12 05:44:11 +00:00
James Zern
356f95b423 Merge "[vpx highbd lpf NEON 3/6] horizontal 8" 2016-10-12 05:35:48 +00:00
Linfeng Zhang
96c7206ede [vpx highbd lpf NEON 3/6] horizontal 8
BUG=webm:1300

Change-Id: Ica2379e294be60b7f80fcfcec110dca4c3b59d81
2016-10-12 00:48:31 +00:00
Marco
065ba0c486 vp8: Adjust threshold on VP8/DatarateTestLarge.DenoiserOffOn.
Fix unit_tests_ubsan failure for VP8/DatarateTestLarge.DenoiserOffOn.
Failure was triggered by commit: df66f8e8.

Change-Id: I7cc5bd309e85950cfc5755e01d0eb942d9ca6984
2016-10-11 16:18:14 -07:00
Marco
57c6bf291e 1 pass vbr: Allow for lookahead alt-ref in real-time mode.
For 1 pass vbr real-time mode:
Allow for the usage of alt-ref frame when non-zero lag-in-frames is used.
Use non-filtered alt-ref, and select usage based on fast scene/content
analysis/detection within the lag of frames.

Positive gains on ytlive set: overall avgPSNR ~3-4%.
Several clips are up between 5-14%, a few clips are neutral/small change.

Current speed decrease is about ~5-10%.

Use the flag USE_ALTREF_FOR_ONE_PASS to enable this feature
(off by default for now).

Change-Id: I802d2bf3d44f9cf01f6d15c76be9c90192314769
2016-10-11 10:13:17 -07:00
Marco
cdbd89197e vp9: 1 pass vbr: some adjustments to gf interval.
Put limit on gf interval based on lag, and allow
for the adjustment on next gf group also on key frame.

Small/neutral change on ytlive metrics.

Change only affects 1 pass vbr real-time mode.

Change-Id: I339c8f4398848698b6e10fe9482c52ca661b94a5
2016-10-11 08:34:12 -07:00
Marco Paniconi
294a734a5f Merge "vp8: Change default gf behavior for 1 pass cbr." 2016-10-10 23:06:31 +00:00
Linfeng Zhang
57e4cbc632 Merge "[vpx highbd lpf NEON 2/6] vertical 4" 2016-10-10 16:57:55 +00:00
Linfeng Zhang
19046d9963 Merge "[vpx highbd lpf NEON 1/6] horizontal 4" 2016-10-10 16:56:23 +00:00
Kaustubh Raste
3da752fe00 Optimize vpx_mbpost_proc_across_ip_msa function
Removed HADD_SW_S32 calculation

Change-Id: I7384dc881451d197404d09beb7c27b222e1d6875
2016-10-10 18:03:28 +05:30
Kaustubh Raste
d05104b488 Optimize vpx_get4x4sse_cs_msa function
Reuse CALC_MSE_B macro

Change-Id: I39f0a92ac2dbb5fa8628df1a5d556cfdc42a3648
2016-10-10 16:31:57 +05:30
Kaustubh Raste
8b5eddf709 Merge "Optimize vp9 loopfilter msa functions" 2016-10-08 05:05:16 +00:00
Kaustubh Raste
3c2f7eb339 Optimize vp9 loopfilter msa functions
Updated code to process in 8bit as saturation/clipping takes care of
overflow
Removed unused macro

Change-Id: I113df60286fb28b216df800d95b2d3695ef71440
2016-10-07 19:26:26 -07:00
Marco
df66f8e830 vp8: Change default gf behavior for 1 pass cbr.
In 1 pass CBR, with error_resilience off, allow for
special logic to change the default gf behaviour.
In this CL: boost is turned off and the gf period
is set to a multiple of cyclic refresh period.

Change only affect 1 pass CBR mode, i.e, when the flag
gf_update_onepass_cbr is set.

Including the previous change (3ec8e11: to allow cyclic refresh
for error_resilience off), comparing metrics on RTC set for
error_resilience off vs on: avgPSNR/SSIM up by ~6%.

Change-Id: Id5b3fb62a4f04de5a805bd1b418f2b349574e0bc
2016-10-07 11:13:06 -07:00
Vignesh Venkatasubramanian
e83e828998 Merge "write_modes: add MACROBLOCKD as a parameter" 2016-10-07 18:09:09 +00:00
Vignesh Venkatasubramanian
ed50e7710c write_modes: add MACROBLOCKD as a parameter
This will enable bit stream packing of each tile column in
parallel.

BUG=webm:1309

Change-Id: Ie349d8cc5825326218ffda893a50730b2e68ed34
2016-10-07 10:25:02 -07:00
Kaustubh Raste
06a6b28d75 Optimize vp8 loopfilter msa functions
Updated code to process in 8bit as saturation/clipping takes care of overflow

Change-Id: I35fb2c0e702fd91309cc391c5a7745a3b619a64c
2016-10-07 15:48:31 +05:30
James Zern
5e4d2548cf Merge "Fix build failure in libvpx_example_test-multi-target." 2016-10-07 01:53:40 +00:00
Linfeng Zhang
49aa9b1f12 [vpx highbd lpf NEON 2/6] vertical 4
BUG=webm:1300

Change-Id: Ia33a9f2d6c7e2e6b3497ad6f1a09439a85b33983
2016-10-06 14:22:26 -07:00
Linfeng Zhang
7aa27bd62f [vpx highbd lpf NEON 1/6] horizontal 4
BUG=webm:1300

Change-Id: Idf441806e6bf397ff5ecd8776146b3f781f50c40
2016-10-06 14:03:04 -07:00
James Zern
ac00db7948 Merge changes from topic '8bit-hbd-idct'
* changes:
  vpx_dsp/idct*_neon.asm: simplify immediate loads
  enable idct*_1_add_neon in high-bitdepth builds
2016-10-06 19:37:19 +00:00
Marco
c7072ae2f4 Fix build failure in libvpx_example_test-multi-target.
Due to change in command line to sample encoder from:
7eff8f3 Update to vpx_temporal_svc_encoder command line.

This caused the tests in vpx_temporal_svc_encoder.sh to fail.

Change-Id: Ic667da81955ad117d04610af21877fed1d4f188f
2016-10-06 12:22:32 -07:00
Alex Converse
fd918cf9a3 Merge "Remove vpx_realloc()" 2016-10-06 18:42:05 +00:00
Kaustubh Raste
f875267ad0 Merge "Modify vp8 idct msa functions store method" 2016-10-06 02:25:42 +00:00
James Zern
1e1caad165 vpx_dsp/idct*_neon.asm: simplify immediate loads
mov supports 0-65535

Change-Id: I019de0d784836d7bd60e6b36f2cdeefb541cb3fd
2016-10-05 14:28:32 -07:00
James Zern
a6be7ba1aa enable idct*_1_add_neon in high-bitdepth builds
these are compatible as they only load one element of the input so the
larger size of tran_low_t makes no difference in little endian builds.
note the asm is incompatible with big-endian, but there are other points of
failure there so currently it's considered unsupported.

BUG=webm:1294

Change-Id: Icd2665a0699bccae92d1bea43a95b0a83fb17028
2016-10-05 11:14:25 -07:00
Marco Paniconi
efb56ec3ff Revert "Revert "vp8/encoder/onyx_if.c: apply clang-format""
This reverts commit a7456144ce.

Change-Id: I400987fb26a09e9b9ea42c91f48ea12f7bc37356
2016-10-05 17:59:55 +00:00
Alex Converse
3063c37600 Remove vpx_realloc()
It only handles the realloc constraint (preserving low elements) by
serendipity, and we don't actually rely on that behavior anyway.
Meanwhile the calls may do extra copying that gets immediately clobbered
by the callers.

Change-Id: I8dfa89e4a81084b084889c27bd272fdf85184e8d
2016-10-05 10:57:56 -07:00
Marco Paniconi
a7456144ce Revert "vp8/encoder/onyx_if.c: apply clang-format"
This reverts commit 891a87dccd.

Change-Id: I067b3b6a3cfb5bc760166999948b8087d4c5cb80
2016-10-05 15:45:48 +00:00
Kaustubh Raste
68f6f6c4cc Modify vp8 idct msa functions store method
vp8_short_inv_walsh4x4_msa - Optimized to process in short vector type
Updated below functions to store exact number of bytes in output rather than complete vector
idct4x4_addblk_msa
idct4x4_addconst_msa
dequant_idct4x4_addblk_msa
dequant_idct4x4_addblk_2x_msa
dequant_idct_addconst_2x_msa

Change-Id: Ic1b3752e2421dc7d70a082dcdaab9d140d7e5d9c
2016-10-05 10:12:12 +05:30
clang-format
891a87dccd vp8/encoder/onyx_if.c: apply clang-format
after:
955b3b6 vp8: Allow for cyclic refresh even if error_resilience it off.

Change-Id: Iba189b18c84be8f5140754280c6801cfc387cfcd
2016-10-04 21:12:06 -07:00
Marco
955b3b66bd vp8: Allow for cyclic refresh even if error_resilience it off.
cyclic_refresh was tied to error_resilience mode.
Allow it to be on also for 1 pass CBR mode even if
error_resilience is off.

Other option to use new control for this, but prefer to avoid
that for now.

Change-Id: I3625b292ee059a890e31338b514e211bf0ab5c3e
2016-10-04 14:19:49 -07:00
Sarah Parker
8978704970 Merge "Remove rate deviation metric from vp8" 2016-10-04 18:56:14 +00:00
Sarah Parker
d556d435f3 Remove rate deviation metric from vp8
BUG=b/31780679

Change-Id: I2b2a43b154eeacb4f51a11f6362cc535cfe318da
2016-10-04 11:20:55 -07:00
Johann Koenig
3db06394e7 Merge "Connect partial IDCT tests" 2016-10-04 18:01:19 +00:00
Johann
24c0146403 Connect partial IDCT tests
Change-Id: Ie8d5d9123f5a9d39db4ec9c74f77ee979ae4e685
2016-10-04 10:31:01 -07:00
Angie Chiang
5d635365bb Merge "Move highbd txfm input range check from 2d iht transform to 1d idct/iadst" 2016-10-04 16:57:37 +00:00
Kaustubh Raste
0a92dd7319 Merge "Fix vpx_plane_add_noise_msa functionality bit-mismatch" 2016-10-04 06:35:47 +00:00
Angie Chiang
5b073c695b Move highbd txfm input range check from 2d iht transform to 1d idct/iadst
This change will make the highbd txfm input range check more comprehensive

The 25-bit highbd input range is composed by
12 signal input bits + 7 bits for 2D forward transform amplification + 5 bits for
1D inverse transform amplification + 1 bit for contingency in rounding and quantizing

BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1286
BUG=https://bugs.chromium.org/p/chromium/issues/detail?id=651625

Change-Id: I04c0796edd7653f8d463fba5dc418132986131e7
2016-10-03 17:21:08 -07:00
James Zern
577221bc87 Merge "invalid_file_test: quiet unused const warning" 2016-10-03 22:51:06 +00:00
James Zern
fb020805f9 Merge "Fix warning when building with GCC 5." 2016-10-03 22:42:52 +00:00
James Zern
c6bc7499d9 Merge "cosmetics,*_neon.c: rm redundant return from void fns" 2016-10-03 22:40:42 +00:00
Kaustubh Raste
6922fc8230 Fix vpx_plane_add_noise_msa functionality bit-mismatch
Change-Id: I04961afb592ae6a67fdcfd8c9066e920dd4b30e7
2016-10-03 18:15:59 +00:00
Marco
7eff8f3b1d Update to vpx_temporal_svc_encoder command line.
Set the  #threads at command line.

Change-Id: Id0daa2393880c3da2d903c11a793072d3008b34b
2016-10-03 09:49:15 -07:00
James Zern
50b9c467da Merge "vpx_convolve8_neon,load/store*: correct param type" 2016-10-01 23:52:14 +00:00
Geza Lore
0dc12b4a1c Fix warning when building with GCC 5.
These caused the following warning with GCC 5:
     warning: logical not is only applied to the left hand side of
     comparison [-Wlogical-not-parentheses]
     assert(!is_compound == (cm->reference_mode == SINGLE_REFERENCE));

Change-Id: If296aabb2311ceb7d903b395c1549ef81c2cbf9b
(cherry picked from commit c6cf7a6111)
2016-10-01 12:23:15 -07:00
James Zern
fca2196a2e invalid_file_test: quiet unused const warning
with --disable-vp9

Change-Id: I81bd603b02ee5d1b45a50aa9e7534f9da498b0e0
2016-10-01 11:49:02 -07:00
James Zern
c449983c56 vpx_convolve8_neon,load/store*: correct param type
stride/pitch in convolve is expressed with a ptrdiff_t

Change-Id: Ia5a6732dc509f06ccf7035386fa8ae721b4b1a71
2016-10-01 11:03:29 -07:00
Martin Storsjo
9255328f27 Remove a stray END declaration in loopfilter_4_neon.asm
Change-Id: Ic8c359a5677f9c663787aac74f530e886163bc69
2016-10-01 14:12:42 +03:00
James Zern
3c00132181 Merge "vp8,frame_buffers: remove unused use_frame_threads" 2016-10-01 01:35:55 +00:00
Linfeng Zhang
da14d23e44 Merge "Refactor vpx lpf NEON files (step 2/2)" 2016-10-01 00:07:51 +00:00
Linfeng Zhang
edbca72a53 Merge "Refactor vpx lpf NEON files (step 1/2)" 2016-10-01 00:07:31 +00:00
Marco Paniconi
0a9f56f146 Merge "vp9: On change_config() only call update_frame_size if needed." 2016-09-30 21:43:33 +00:00
Marco Paniconi
5e908aff34 Merge "vp9 real-time mode: Change loopfilter speed feature at speed 8." 2016-09-30 21:42:05 +00:00
James Zern
db80c23fd4 cosmetics,*_neon.c: rm redundant return from void fns
+ a couple of 'break's after a return

Change-Id: Ia21f12ebcef98244feb923c17b689fc8115da015
2016-09-30 13:09:57 -07:00
James Zern
b6277a47c7 Merge changes from topic '8bit-hbd-idct'
* changes:
  *idct*_neon.c: add missing rtcd include
  idct,msa/neon: exclude idct files from hbd build
  *rtcd_defs.pl: remove empty specialize calls
2016-09-30 19:36:08 +00:00
James Zern
1396d12103 *idct*_neon.c: add missing rtcd include
+ correct declarations as necessary

BUG=webm:1294

Change-Id: I719602df9a56e79188a78e7f8b31257c6d3cc11d
2016-09-30 11:41:26 -07:00
James Zern
b51c4df93a idct,msa/neon: exclude idct files from hbd build
these functions are incompatible currently and unreferenced in rtcd,
exclude them from the build.

BUG=webm:1294

Change-Id: I7790c195a91e1b142f56c04d2a5e305d9133b896
2016-09-30 11:32:47 -07:00
Linfeng Zhang
ca2fe7a8c7 Refactor vpx lpf NEON files (step 2/2)
Change-Id: I0744407cd3361ff752bd7f6e654b70ab6b41a58f
2016-09-30 09:56:28 -07:00
Linfeng Zhang
4779f5308d Refactor vpx lpf NEON files (step 1/2)
Change-Id: I4016d096d46ca691f3b17199b259b7231e983cfb
2016-09-30 09:48:54 -07:00
Linfeng Zhang
8c744fd978 Merge "Unify loopfilter function names" 2016-09-30 15:58:08 +00:00
Linfeng Zhang
c435b7fbdd Merge "Refine vpx convolve8 NEON intrinsics optimization" 2016-09-30 15:56:31 +00:00
Linfeng Zhang
bde905cba1 Merge "Refine vpx_convolve_copy_neon() and vpx_convolve_avg_neon()" 2016-09-30 15:54:02 +00:00
James Zern
ed62d27c71 *rtcd_defs.pl: remove empty specialize calls
add_proto adds a 'c' specialization

Change-Id: I0ed0c2240d45264b0e0056ce7c8f63f4a00780bc
2016-09-29 20:38:26 -07:00
James Zern
f38616e1a2 vp8,frame_buffers: remove unused use_frame_threads
this was never fully implemented

Change-Id: I4640cf84c40ea2cc9c6c12acf116d39df4b04578
2016-09-29 20:24:15 -07:00
James Zern
39ff0de810 Merge "configure: test for -Wshorten-64-to-32 in non hbd builds" 2016-09-30 03:01:56 +00:00
Johann Koenig
cb4aa6d589 Merge changes I158f631a,I0555f639
* changes:
  vp8: remove mmx functions
  Rename _xmm functions to _sse2
2016-09-30 01:47:41 +00:00
Yunqing Wang
9afe2cf599 Merge "Fix an issue in vp9_first_pass for non-mulitple of 16 resolutions" 2016-09-30 00:49:06 +00:00
Linfeng Zhang
7f1f35183a Unify loopfilter function names
Rename vpx_lpf_horizontal_edge_8() to vpx_lpf_horizontal_16().
Rename vpx_lpf_horizontal_edge_16() to vpx_lpf_horizontal_16_dual().

Change-Id: I798ca8fbbd657d06d3db2bfb0fb3321168f49e52
2016-09-29 16:25:42 -07:00
Linfeng Zhang
85a9e48d25 Refine vpx_convolve_copy_neon() and vpx_convolve_avg_neon()
BUG=webm:1290

Change-Id: Ia27e58521eba5a4852b50381c56746fa5767f6d6
2016-09-29 16:19:39 -07:00
Deepa K G
2745f94deb Fix an issue in vp9_first_pass for non-mulitple of 16 resolutions
This patch sets the 16x16 src_diff to zero and ensures correct calculation
of this_error for block sizes smaller than 16x16.

Change-Id: I7b7c02d267433c9f22c8ac9b8d5df2f499175172
2016-09-29 16:19:23 -07:00
Johann Koenig
ad55b1d270 Merge changes Ia3e9122f,Id33eb6c8,I956bd8ce
* changes:
  Remove vp8_clear_system_state
  vpx_dsp: clean up rtcd
  vp8: clean up rtcd
2016-09-29 23:16:45 +00:00
James Zern
7b9c86167e Merge "vp9_detokenize,decode_coefs: fix signed int overflow" 2016-09-29 22:36:13 +00:00
Johann
721354fe7f vp8: remove mmx functions
When they have sse2 equivalents.

Change-Id: I158f631a3bcecba57b36093ac10114b1904767a7
2016-09-29 15:25:27 -07:00
Johann
2663b092ae Rename _xmm functions to _sse2
Avoid the extra level of indirection/confusion.

Change-Id: I0555f639d67835df9fb7dac0c75085e9954805f1
2016-09-29 15:23:11 -07:00
Johann
1364cb58b4 Remove vp8_clear_system_state
Use vpx_clear_system_state instead.

Change-Id: Ia3e9122f69a2c690ddd7c7bc54f92ccb9ec18b3e
2016-09-29 13:22:49 -07:00
Marco
e765435293 vp9: On change_config() only call update_frame_size if needed.
change_config() may be called often in real-time application,
to update bitrate/framerate or qp-max/min.
No need to do update_frame_size() unless frame size has changed.

Change-Id: I23a51deade1e03adc91c468f9ffde3235298770c
2016-09-29 13:03:26 -07:00
Marco
d017548be6 vp9 real-time mode: Change loopfilter speed feature at speed 8.
For real-time mode at speed 8: turn off MINIMAL_LF at speed 8,
for non-screen content mode.

Visually better, avgPSNR/SSIM on rtc set go up by ~4-5%.
Speed decrease of about ~3%.

Change-Id: I8eb69330f02e0ceece1507d43cfc8a049a1d8291
2016-09-29 12:59:01 -07:00
Linfeng Zhang
b3cb065ee4 Refine vpx convolve8 NEON intrinsics optimization
BUG=webm:1290

Change-Id: I5d7fce62270f9d76ef9ce98b3d188ad11fb21873
2016-09-29 12:48:59 -07:00
James Zern
691ef20272 Merge changes I11786887,Ia91823ad
* changes:
  vpx_dsp/get_prob: relocate den == 0 test
  vpx_dsp/get_prob: make clip_prob branchless
2016-09-29 19:11:35 +00:00
Johann
7b5a348088 vpx_dsp: clean up rtcd
Remove avx2+ssse3 specialization. Disabling ssse3 now automatically
disables avx2.

Change-Id: Id33eb6c85d1c4ee57128ebe45c995eb15cfcc765
2016-09-29 12:10:07 -07:00
Johann
c7f9d0719d vp8: clean up rtcd
Remove lines which specify the same name for a function.

Change-Id: I956bd8ce2b81a2a8feab5621d28bd2499c2b4c2d
2016-09-29 12:10:01 -07:00
James Zern
450d89034b vp9_detokenize,decode_coefs: fix signed int overflow
when decoding an invalid bitstream with --enable-vp9-highbitdepth

BUG=webm:1297

Change-Id: I401d87033b4293f2ca595bc51678aad9951ecf15
2016-09-28 22:42:03 -07:00
James Zern
93c823e24b vpx_dsp/get_prob: relocate den == 0 test
to get_binary_prob(). the only other caller mode_mv_merge_probs() does
its own test on 0.

BUG=chromium:639712

Change-Id: I1178688706baeca2883f7aadbc254abb219a44ce
2016-09-28 17:42:49 -07:00
James Zern
e094e151de Merge "vp9: fix compilation for g++ 6.2.x" 2016-09-28 23:36:23 +00:00
Johann Koenig
bb27be0dfe Merge "Hook up vp8_diamond_search_sad_sse3" 2016-09-28 20:54:25 +00:00
James Zern
63f7e131fe Merge "vpxdec: avoid memory leaks under most conditions" 2016-09-28 19:35:16 +00:00
James Zern
7481edb33f vpx_dsp/get_prob: make clip_prob branchless
+ inline the function directly as there was only one consumer
(get_prob())

this is an attempt to reduce the amount of branches to workaround an amd
bug. this change is mildly faster or neutral across x86-64, arm.

http://support.amd.com/TechDocs/44739_12h_Rev_Gd.pdf
665 Integer Divide Instruction May Cause Unpredictable Behavior

BUG=chromium:639712

Suggested-by: Pascal Massimino <pascal.massimino@gmail.com>
Change-Id: Ia91823aded79aab469dd68095d44300e8df04ed2
2016-09-28 11:51:46 -07:00
Tristan Matthews
32c375447c vp9: fix compilation for g++ 6.2.x
Inline function called from test/dct16x16_test.cc wouldn't build due to:
  invalid operands of types ‘__gnu_cxx::__enable_if<true, double>::__type
  {aka double}’ and ‘int’ to binary ‘operator>>’
  return (abs(ref->row) >> 3) < COMPANDED_MVREF_THRESH &&

this converts the test to abs() < COMPANDED_MVREF_THRESH << 3 which
hides the promotion issue.

Regression from commit de993a847f

BUG=webm:1291

Change-Id: I73b5943d07d5b61b709d299114216a2371a8fd62
2016-09-27 23:17:31 -07:00
Linfeng Zhang
240726ac85 Merge "Clean convolve_test.cc" 2016-09-28 00:20:28 +00:00
Debargha Mukherjee
74f038e6f8 Merge "Fix for compile error with range checking" 2016-09-28 00:05:37 +00:00
James Zern
06abc1ecd9 configure: test for -Wshorten-64-to-32 in non hbd builds
provides msvc-like warnings for implicit conversions from 64-bit to
32-bit types

--enable-vp9-highbitdepth still requires some work

this also skips CXXFLAGS for now as some work would be needed to cleanup
third_party/*.cc or split it from test/*.cc where it comes to flags.

Change-Id: Ic9a095b73286eba5ed39bfc27ff69593748cbbf4
2016-09-27 16:52:21 -07:00
Johann
3a57ce4478 Cast strto[u]l down
Change-Id: I945b2f8754cf484a08e5ba511cfd2d4a44181b08
2016-09-27 15:37:10 -07:00
Johann
e4ddf9db6a Hook up vp8_diamond_search_sad_sse3
The original commit never set any 'specialize' line:
61311e6103

It appears the sadx4 version of function uses sdx4df calls to speed up
the search. There are no sse3 versions of the sdx4df functions, but
there are sse2 and msa versions.

There is a neon version of vpx_sad16x16x4d but not any of the smaller
versions. Perhaps if they existed this function could be expanded to use
them.

Change-Id: I936d7d6b1a3ff6dcd5a4d2322272708c47cdec13
2016-09-27 15:31:49 -07:00
James Zern
e61d82bd4f vpxdec: avoid memory leaks under most conditions
avoids false positives when fuzzing with ASan+LSan.

Change-Id: I0d23b530ae80e5692b6951fe6e3690ea44159a5a
2016-09-27 14:29:18 -07:00
Johann Koenig
348cff040a Merge changes from topic 'wextra'
* changes:
  Expand -Wextra to more of the library
  mips: clean up wextra warnings
  Add compiler flag -Wsign-compare
  Add compiler warning flag -Wextra and fix related warnings.
2016-09-27 21:13:50 +00:00
Linfeng Zhang
81ff7a065f Clean convolve_test.cc
Combine test MatchesReferenceSubpixelFilter and
MatchesReferenceAveragingSubpixelFilter.

Change-Id: I75f96befbbb118cdc6b8c6001b4cdda8d88fbbd3
2016-09-27 13:36:31 -07:00
Johann
c3a135b5b8 Expand -Wextra to more of the library
Suppress warnings in third_party/.

vp8 -Wclobbered issue is tracked here:
BUG=webm:1246

BUG=webm:1069

Change-Id: I9b94bf546d7b690c26a59ae67967facdce8ec45b
2016-09-27 13:19:27 -07:00
Johann
02fa245d15 mips: clean up wextra warnings
Remove unused zbin variable:
warning: unused parameter ‘zbin’

Use int for loop variables to avoid unsigned conversion:
warning: comparison between signed and unsigned integer expressions

Change-Id: Icea74b870c0ee68a8bf687e796a69392af25a8ad
2016-09-27 13:19:18 -07:00
Urvang Joshi
097b31c7f0 Add compiler flag -Wsign-compare
Also, fix the warnings generated by this flag.

(cherry picked from commit ebeb1155d4fa6d28e2f40c92265245f8df097fcb)

From AOM. Don't actually add -Wsign-compare. It will be covered by
-Wextra.

Switch to vpx_integer.h from df9c9d6d4c43f02c58d4e776c53323788e013cbc

BUG=webm:1069

Change-Id: I1dc6e61caa5d56af4a55b6692ab620bb3144652a
2016-09-27 12:39:36 -07:00
Urvang Joshi
0aa3e2564f Add compiler warning flag -Wextra and fix related warnings.
Note: some of these warnings are enabled by a combination of -Wunused
(added earlier) and -Wextra.

Cherry-picked from AOM 4790a69faaec8f03d65f64ff070f6ab4307dbb16

Expands use of (void)x; on unused variables. AOM only supports one codec
in codec_factory.h

Does not include changes to HandleDecodeResult. AOM removed
invalid_file_test.cc which does use the video parameter.

Does not enable -Wextra yet. There are more issues to fix.

BUG=webm:1069

Change-Id: I322a1366bd4fd6c0dec9e758c2d5e88e003b1cbf
2016-09-27 12:05:01 -07:00
Paul Wilkins
b3ebea5e8a Merge "Limit max arf boost and scale motion breakout for image size." 2016-09-27 14:08:29 +00:00
Peter de Rivaz
8db503063f Fix for compile error with range checking
Current version does not build with options:
  --enable-vp9-highbitdepth --enable-coefficient-range-checking

Change-Id: Ic3285f1a3e0d6be88da7f2cd8fa5a631368dd03b
2016-09-27 09:28:44 +01:00
Marco Paniconi
70240a77b8 Merge "vp9: Reduce frame loopfilter-level for 1 pass cbr." 2016-09-26 22:05:44 +00:00
Johann Koenig
b165451ad5 Merge "Un-Revert "Restore vp8_sixtap_predict4x4_neon"" 2016-09-26 19:11:00 +00:00
Johann Koenig
37798711aa Merge "Use shifted value for sinpi8sqrt2" 2016-09-26 19:10:57 +00:00
Marco
d9fc28c0a1 vp9: Reduce frame loopfilter-level for 1 pass cbr.
Reduce the filt_guess for 1 pass cbr on inter-frames.
This reduces visual artifact seen in rtc clip (jimred.vga),
and improves metrics on rtc set.

Metrics on rtc set for cbr mode overall positive, most clips are up:
Speed 7 rtc: avgPSNR/SSIM up by: ~2.6/3.9%
Speed 8 rtc: avgPSNR/SSIM up by: ~1.3/2.5%

Change-Id: Ia4eccea1c19d65b583516df28823cd756c49464f
2016-09-26 10:12:43 -07:00
Linfeng Zhang
b46243d7ff Merge "Refactor lpf (size 4 and 8) NEON intrinsics optimization" 2016-09-26 16:11:12 +00:00
paulwilkins
0421d8e318 Limit max arf boost and scale motion breakout for image size.
Added a cap on the maximum boost for an arf based on interval length.
Fixed bug where by the image size was not accounted for in determining
two of the motion breakout thresholds.

Overall small gains of 0.2-0.4% psnr but on large image format clips with
slow zooms the gain may be as much as 20% or more (e.g. in_to_tree
at 1080P)

Change-Id: Id0a47391203026742daa9c97afac5705fd8c4dfb
2016-09-26 15:38:29 +01:00
Scott LaVarnway
60624aa53a Merge "VP9: token decoder expansion" 2016-09-26 12:06:50 +00:00
James Zern
f8c056a895 Merge "vp9_idct: delete dead TODOs" 2016-09-24 01:47:00 +00:00
Johann
ab0e7a237a Use shifted value for sinpi8sqrt2
The value 35468 changes sign when stored in int16_t:
implicit conversion from 'int' to 'int16_t' (aka 'short')
changes value from 35468 to -30068

This negation requires adding back the original value to compensate.
Shifting the value keeps the value positive and saves a post-vqdmulh
shift.

This technique is used in webp and idct_dequant_full_2x_neon

BUG=b/28027557

Change-Id: I0c5ce09bea170fe08061856c2af6f841a557e0c3
2016-09-23 17:04:18 -07:00
Johann
1d14e42df7 Un-Revert "Restore vp8_sixtap_predict4x4_neon"
This restores d9dce2f48e

Switched to using signed shift-and-narrow. Instead of saturating
negative results to 0, it was saturating them to 255.

BUG=webm:817
BUG=webm:1273

Change-Id: I571095336aa4182e3288b17924fcaaece42b0a49
2016-09-23 14:58:57 -07:00
Scott LaVarnway
87b689f97a VP9: token decoder expansion
This version is based on Change 267683, but does not
use the macros.

Change-Id: I0619fa618decf8bdeef250584d75d70318b5d9a7
2016-09-23 06:24:20 -07:00
Scott LaVarnway
ada850786c Merge "VP9: pass TileWorkerData instead of MACROBLOCKD and vpx_reader." 2016-09-23 11:59:16 +00:00
James Zern
deadda3dea Merge "vpx_idct32x32_34_add_sse2: rm unneeded transposes" 2016-09-23 02:49:26 +00:00
James Zern
a914ffad97 Merge "variance_neon: sync variance*() w/c,sse2" 2016-09-23 02:18:49 +00:00
Scott LaVarnway
7a34f85955 VP9: pass TileWorkerData instead of MACROBLOCKD and vpx_reader.
Change-Id: I869ef0f113c022143b531c44aefa0f1bb267052d
2016-09-22 13:18:36 -07:00
James Zern
fdd1186f97 vpx_idct32x32_34_add_sse2: rm unneeded transposes
this change is neutral to mildly positive across various x86-64
platforms

Change-Id: I28fb5ae598fc1317b7a42c9a846ac5d57d104784
2016-09-21 19:49:25 -07:00
Angie Chiang
99ef84c65a Merge "Detect invalid highbd iht input" 2016-09-22 01:06:38 +00:00
James Zern
e372bfd5ac variance_neon: sync variance*() w/c,sse2
removes some unnecessary casts and adds a few explicit uint32 ones for
larger sizes to quiet -Wshorten-64-to-32 warnings

Change-Id: I63c5fce8e62c426d5cf5c10a66a113c119a43518
2016-09-21 18:04:45 -07:00
James Zern
fcf281b6a1 Merge "vp8: remove VP8_SET_DBG* control support" 2016-09-22 00:43:35 +00:00
Angie Chiang
80338b91d3 Detect invalid highbd iht input
Do nothing in vp9_highbd_iht#x#_##_add_c when input magnitude is beyond
20 bits. Note that, sign bit is not included here.

In the 20 bits, we use 12 bits for input signal, 7 bits for forward
transform amplification, and 1 bit for contingency in rounding and
quantizing

BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1286

Change-Id: I332c6f68df4614fc2e7d2dc4c5bb0d0cff8a245c
2016-09-21 17:15:19 -07:00
Johann
2bed8b6acd Keep vp8 sixtap read within bounds
When filtering it needs 6 pixels: 2 prior to the source, the source, and
3 after the source.

When filtering 16 wide, that means 21. To accomplish this the SSE2 reads
[-2] to [5], [6] to [13], and [14] to [21], a total of 24 bytes (reading
in groups of 8 is easy)

The filter then shifts this last set to the top half of the register and
uses 'or' to combine it with the previous set.

Valgrind detected an issue reading pixels [19], [20] and [21]:
Address 0x7f581c2 is 434 bytes inside a block of size 441 alloc'd

Note: we only need pixels [16], [17], and [18] as context for [15].

To fix this, it now reads 8 bytes starting at [11], which re-loads [11]
through [13], but stops at [18] and does not over-read any values.

This is shifted by 5 and 'or'd with xmm1. Although the lower bits are
not cleared, they overlap directly with [11] through [13], so 'or'
produces the correct results.

Change-Id: I0c89c03afa660fc9b0108ac055d7bd403e493320
2016-09-21 16:17:07 -07:00
Johann
35ebc1cddf predict_test: align dst buffer to 16
On 32 bit machines 'new' does not always appear to allocate sufficiently
aligned buffers, causing intermittent test failures.

Change-Id: I0db4fc73782012e4eef71dc0fb540e74fdbfcebe
2016-09-21 13:35:47 -07:00
James Zern
3f72509587 vp8: remove VP8_SET_DBG* control support
the --enable-postproc-visualizer configure option remains as a no-op as
do the control names and values for compatibility
+ remove the corresponding debug flags from vpxdec: --pp-*

Change-Id: I4a001cd9962b59560d7d6bda6272d4ff32b8d37c
2016-09-20 20:19:36 -07:00
James Zern
cec6433e41 vp9_idct: delete dead TODOs
Change-Id: Icdd5494f557d83026dc078bce37997a76aa288fb
2016-09-20 19:46:27 -07:00
James Zern
b6e686b1ea Merge changes from topic 'Wshorten'
* changes:
  vp8: convert some uses of unsigned long to size_t
  vp8/encoder: quiet some -Wshorten-64-to-32 warnings
2016-09-20 23:17:20 +00:00
James Zern
c31d02615d Merge "variance_avx2: sync variance functions with c-code" 2016-09-20 22:33:39 +00:00
James Zern
2351a73531 Merge "examples: quiet -Wshorten-64-to-32 warnings" 2016-09-20 22:32:58 +00:00
James Zern
feb4313c5f Merge "vp9_rtcd: remove non-existent highbd convolve fns" 2016-09-20 22:07:09 +00:00
Johann Koenig
8478f97105 Merge "Enable ssse3 bilinear tests" 2016-09-20 21:46:50 +00:00
Johann Koenig
18fd69ee91 Merge "Add vp8_bilinear_filter test" 2016-09-20 20:30:48 +00:00
Alex Converse
0d2687ef87 Merge "Code class0 using vpx_read() / vpx_write()." 2016-09-20 19:19:29 +00:00
James Zern
5841929fde vp9_rtcd: remove non-existent highbd convolve fns
these were moved to vpx_dsp

Change-Id: I307b07ae05e2333277d4b7011cba36dcf8409959
2016-09-19 20:01:23 -07:00
James Zern
08b8b6bb8f examples: quiet -Wshorten-64-to-32 warnings
all around usage of strtol/strtoul

Change-Id: If907c89f107a068987aa71ddd93cee9a7389e4cd
2016-09-19 19:02:49 -07:00
James Zern
8281da74b9 vp8: convert some uses of unsigned long to size_t
similar to changes that were done in vp9 for encoded frame size
reporting. has the side-effect of quieting a -Wshorten-64-to-32 warning.

Change-Id: I89f74cb617fc29334ee351dc8dfaa3b8cfd4e5af
2016-09-19 18:35:59 -07:00
James Zern
0ce98b423b vp8/encoder: quiet some -Wshorten-64-to-32 warnings
this code is similar to other existing uses and/or vp9

Change-Id: I56e646931379759d9f7332ea6d746060007c75ee
2016-09-19 18:35:59 -07:00
Linfeng Zhang
761e5ec2f6 Refactor lpf (size 4 and 8) NEON intrinsics optimization
Also check in 8x8 8-bit transpose NEON intrinsics optimization
transpose_u8_8x8()

Change-Id: I32d321cf97ea21eab158ac4896990fc9a51681c4
2016-09-19 16:41:37 -07:00
James Zern
6acd061aad variance_avx2: sync variance functions with c-code
add missing int64 -> uint32 cast; quiets -Wshorten-64-to-32 warnings

Change-Id: I4850b36e18dc8b399108342be4bfe0b684aefb78
2016-09-19 16:19:29 -07:00
Johann Koenig
0695843a21 Merge "Remove -fno-strict-aliasing flag" 2016-09-19 22:49:23 +00:00
Johann
fad70a358b Remove -fno-strict-aliasing flag
The referenced bug was fixed by saving neon registers. That this had any
effect was coincidental.

Both chromium and Android build with clang and neither uses this flag.

Change-Id: I470247d6fd9226fc207b42a187105581a94badc3
2016-09-19 12:16:03 -07:00
Nathan E. Egge
de7f5ce9e5 Code class0 using vpx_read() / vpx_write().
The vp9_mv_class0_tree is a balanced tree with two leafs and can
simply be coded as a boolean with probability class0[0].

Change-Id: If294dac825a5f945371092c74aa8e3f84cd962b6
(cherry picked from commit be8a8ab62ebdd111c6f2e9a33b15630570671eba)
2016-09-19 10:50:39 -07:00
Alex Converse
01e2902521 Zero the whole rd_counts struct rather than the each member
Change-Id: I495aa9cec2b2b8f1ae69bdab8b3feeca76358472
2016-09-19 10:04:47 -07:00
James Zern
aa0eb67bf7 loopfilter_mb_neon: remove unused load_8x8()
quiets a -Wunused-function warning for arm targets

Change-Id: I293a7e3d3d7d61d6af2fbedad5e8c25126c418b6
2016-09-17 11:00:31 -07:00
Linfeng Zhang
5d73639d8f Merge "Refactor lpf (size 16) NEON intrinsics optimization" 2016-09-17 00:33:30 +00:00
James Zern
112eb54c1b Merge "vpx_codec_control: return incapable for unmatched control" 2016-09-16 17:30:44 +00:00
Linfeng Zhang
8107368000 Refactor lpf (size 16) NEON intrinsics optimization
Extract shared code so later lpf size 4 and 8 functions can reuse.

Change-Id: Ibb43ef1fd8651bd2e32fcc4c56cf6fa7ca237401
2016-09-16 09:12:13 -07:00
James Zern
33aef48f29 vpx_subpixel_8t_intrin_avx2: tolerate unversioned clang
assume __clang_major__==0 has the latest version of
_mm256_broadcastsi128_si256. fixes builds with custom clang toolchains.

BUG=b/30970831

Change-Id: I90becd56278e4716bd46e2ba9d910af977e8dfa6
2016-09-16 07:14:17 +00:00
James Zern
7a9e476072 Merge changes from topic 'clang-format'
* changes:
  apply clang-format
  .clang-format: update to 3.8.1
2016-09-16 07:11:33 +00:00
Johann
e813c2b416 Enable ssse3 bilinear tests
The code only has issues when xoffset == 0 and yoffset == 0 which
represents a simple copy. Presumably this case does not need to be
handled because the issue has existed since 2010.

BUG=webm:1287

Change-Id: Ic47e2653f3b729e99b40e53d8d2d8d1501edaaa9
2016-09-15 23:16:26 -07:00
Johann
caf9a7841e Add vp8_bilinear_filter test
Build out the sixtap_predict test because the filters are
interchangeable. Add verbose failures and border checking.

Change-Id: I962f50041750dca6f8d0cd35a943424cf82ddcb1
2016-09-15 23:16:19 -07:00
James Zern
6ae58fd55e Merge "Revert "Restore vp8_sixtap_predict4x4_neon"" 2016-09-16 06:13:42 +00:00
Johann Koenig
7795e99296 Revert "Restore vp8_sixtap_predict4x4_neon"
This reverts commit d9dce2f48e.

Appears to be failing the SixtapPredict tests in some configurations and possibly test vectors as well.

Change-Id: Ica6aa83ebac47d0a76e451846e7da67b1c17a7d7
2016-09-16 06:12:49 +00:00
Johann Koenig
fdbe249991 Merge "Restore vp8_bilinear_predict4x4_neon" 2016-09-16 05:33:50 +00:00
Johann Koenig
102eae06e9 Merge "zero structures completely" 2016-09-16 04:41:22 +00:00
Johann
43743b1d3e Restore vp8_bilinear_predict4x4_neon
This function was removed when clang started introducing alignment hints
which caused the 32 bit vld1_lane_u32/vst1_lane_u32 to fail:
https://llvm.org/bugs/show_bug.cgi?id=24421

The load has been rendered safe with an implementation ~indiscernible
performance-wise that uses _u8 and over-reads just a touch.

It is still ~5x faster than C in the unaligned case and doing both
filters.

BUG=webm:892
BUG=webm:1273

Change-Id: Icf7167189391b46202f47233bb585c24c42bcc36
2016-09-15 21:16:11 -07:00
Johann Koenig
7bc0733c27 Merge "Restore vp8_sixtap_predict4x4_neon" 2016-09-16 04:12:08 +00:00
Johann
d5054504a7 zero structures completely
Use vp[89]_zero when possible.

Expand the {} set when neither is available or nearby.

Change-Id: Ifc1f46f60100916cd798bf7be3a10f09321c99bd
2016-09-16 03:54:11 +00:00
Johann
1d2aaf58dd vp8 postproc: expand CONFIG_POSTPROC guard
postproc.c is overloaded and used for both postproc and internal stats.
If only --enable-internal-stats is specified there are issues with
non-existent struct members and unused functions.

Change-Id: I82367f1ffce659c3918c9f964dbce94a716fbb89
2016-09-16 03:52:19 +00:00
Johann
f2be831885 altref test: comment out 'pass'
All the other test which do not use 'pass' (which appears to be almost
all of them) do this.

Cleans -Wextra/-Wunused-parameter:
unused parameter ‘pass’

Change-Id: I1ff3acf3f3d1e831f94dcb00ea36337afe0aefe0
2016-09-15 17:45:47 -07:00
Johann Koenig
c53aacf408 Merge "vp9 frame parallel test: Initialize cfg differently" 2016-09-15 23:46:56 +00:00
Marco
4c1a9fb8db vp9: Small code cleanup.
Remove the experiment LIMIT_QP_ONEPASS_VBR_LAG, as its
not currently used and no plan to use in near future.

Change-Id: Ib069f8d7225195be04b765d0ab477510dfba6a3b
2016-09-15 15:17:17 -07:00
clang-format
5f6d143b41 apply clang-format
Change-Id: I501597b7c1e0f0c7ae2aea3ee8073f0a641b3487
2016-09-15 15:07:53 -07:00
James Zern
30b1abd6e6 .clang-format: update to 3.8.1
based on --style=Google with the following differences:
3a4
> # Generated with clang-format 3.8.1
13c14
< AllowShortCaseLabelsOnASingleLine: false
---
> AllowShortCaseLabelsOnASingleLine: true
41c42
< ConstructorInitializerAllOnOneLineOrOnePerLine: true
---
> ConstructorInitializerAllOnOneLineOrOnePerLine: false
44,45c45,46
< Cpp11BracedListStyle: true
< DerivePointerAlignment: true
---
> Cpp11BracedListStyle: false
> DerivePointerAlignment: false
73c74
< PointerAlignment: Left
---
> PointerAlignment: Right
75c76
< SortIncludes:    true
---
> SortIncludes:    false

SortIncludes will like be enabled in a future commit

Change-Id: I5c404f44081b65354e7f526411c91fbbe31ac5af
2016-09-15 15:05:52 -07:00
Johann
d9dce2f48e Restore vp8_sixtap_predict4x4_neon
This function was removed when clang started introducing alignment hints
which caused the 32 bit vld1_lane_u32/vst1_lane_u32 to fail:
https://llvm.org/bugs/show_bug.cgi?id=24421

The load has been rendered safe with an implementation ~indiscernible
performance-wise that uses _u8 and over-reads just a touch.

The store, when unaligned, has a version that is ~25% slower but safe
when xoffset = 0 (second pass filter only). When the first pass filter
(or both) are in play, the new version is almost identical in speed.

Worst case performance (both filters, unaligned stores) is roughly 3-4x
faster than C.

BUG=webm:817
BUG=webm:1273

Change-Id: I1e490e94453e0872151fe0dafb05557463f6247d
2016-09-15 14:56:47 -07:00
Johann
284cb5314e vp9 frame parallel test: Initialize cfg differently
Use the canonical 'vpx_codec_dec_cfg_t()' as opposed to 'vp9_zero()'
which just hammered everything to 0.

Change-Id: Id820efef700ad92a625797f8fd58e465b15eeca4
2016-09-15 12:19:25 -07:00
Johann Koenig
ee01b78ddd Merge "Documentation for building unit tests for Android" 2016-09-15 19:17:14 +00:00
Johann
a3400f4376 Documentation for building unit tests for Android
BUG=webm:1258

Change-Id: Iea142f7b0df0e047720e8c5362464932de57d564
2016-09-15 19:16:14 +00:00
James Zern
4282d29355 Merge "cosmetics,vp8: join some lines, fix table format" 2016-09-14 00:41:51 +00:00
Johann
4c6819d0fc vp8 decoder: cast decoding_thread_count to int
For some reason allocated_decoding_thread_count is signed, but decoding_thread_count is not.

Cleans -Wextra/-Wsign-compare:
comparison between signed and unsigned integer expressions

Change-Id: Id0ada78100acff27c1c4ed7493c563d13c55cdcd
2016-09-13 14:51:14 -07:00
Johann
75fe2d4409 vp9 frame parallel test: Initialize cfg to 0
Use vp9_zero() to set every element.

Cleans -Wextra/-Wmissing-field-initializers:
missing initializer for member ‘vpx_codec_dec_cfg::w’
missing initializer for member ‘vpx_codec_dec_cfg::h’

Change-Id: I5b41ce7d55a912e29b1d4c3e840cea80e8510fbe
2016-09-13 14:51:14 -07:00
Johann
db32581650 vp9cx_set_ref.c: remove unused 'cfg' parameter
Cleans -Wextra/-Wunused-parameter warning:
warning: unused parameter ‘cfg’

Change-Id: I84eae57a50306cb66c625bb648b0a330678818db
2016-09-13 14:51:06 -07:00
Johann
bce23ab36b webmenc: remove unused 'fps' parameter
Cleans -Wextra/-Wunused-parameter warning:
warning: unused parameter ‘fps’

Change-Id: Ia5f9338f11ae8d0708a87c6d4e7d7e924fc3b19b
2016-09-13 14:25:40 -07:00
James Zern
6eca31be5f vpx_codec_control: return incapable for unmatched control
VPX_CODEC_INCAPABLE rather than the more generic VPX_CODEC_ERROR

Change-Id: Id1ed7fb23a2910192713c6b2389c0b7320201f52
2016-09-09 17:40:10 -07:00
James Zern
a22a455899 cosmetics,vp8: join some lines, fix table format
Change-Id: Idcf3b68f0e59bd74c9d332bbd4a7c1484ddb691a
2016-09-09 16:39:34 -07:00
Marco
421f376568 vp8: Set the skin model to mode 1.
This change was reverted before due to a hangouts encode-time
regression investigation. But since then this change has been
cleared of causing any noticeable regression.

This mode reduces some false detection, and uses the
same model as in vp9.

Change-Id: I9c82a748c5f601d0aca9f61ee218abfbd58c62bd
2016-09-09 09:09:43 -07:00
James Zern
66241b9579 Merge "vp8: Remove TSAN warning around end of encode." 2016-09-09 03:08:18 +00:00
Alexander Potapenko
948a1f51d0 vp8: Remove TSAN warning around end of encode.
Tsan warns when run in one pass and there is a recode
loop.

Change-Id: Ice2ecb2270f09ebd49efbd49c0e4f77d32e23c0f
2016-09-08 14:36:32 +02:00
James Zern
4b0e78bfda Merge "vpx_dsp: added vpx_highbd_idct32x32_1_add_sse2()" 2016-09-08 01:05:18 +00:00
James Zern
bcbc4761fa vpx_mem.c: remove unnecessary inline
these aren't overly speed critical, best to leave it to the compiler.

Change-Id: I231c14abee5b845d7b8e8454832f2feb22c6ce45
2016-09-07 12:49:21 -07:00
Scott LaVarnway
309125b1e7 vpx_dsp: added vpx_highbd_idct32x32_1_add_sse2()
Change-Id: I140d93aebadb0eaf6220881e61a0451450081227
2016-09-07 05:58:29 -07:00
Sarah Parker
c892521b1d Fix missing write to opsnr in internal stats
Change-Id: I21c8ad0b5ed7f8d843cae45c18f5727bceb8f859
2016-09-03 12:15:32 -07:00
James Zern
4a25b59bbd Merge "invalid_file_test: quiet -Wunused-const-variable warnings" 2016-09-03 01:14:55 +00:00
James Zern
e6f0c26268 invalid_file_test: quiet -Wunused-const-variable warnings
present when --disable-vp8(-decoder) or --disable-vp9(-decoder) was used

Change-Id: I31ebb7a55c6f1af3c744982f56b78e80116cc845
2016-09-01 19:54:34 -07:00
James Zern
3d253b0c71 vp8_cx_iface: quiet -Wshorten-64-to-32 warning
set_reference_and_update(): use the correct type for flags,
vpx_enc_frame_flags_t

Change-Id: I257da784537ff18686f6db8665f99af6ea6a86ba
2016-09-01 19:54:00 -07:00
James Zern
d6d3d4ba31 get_cpu_count: quiet -Wshorten-64-to-32 warnings
sysconf returns a long; cast (unsigned) dwNumberOfProcessors to int for
good measure

Change-Id: I1f181d7bd9a060c0898db41f66a5065394afdc4e
2016-09-01 19:54:00 -07:00
Johann Koenig
4d1540f8ce Merge changes from topic 'Wundef'
* changes:
  Enable -Wundef by default
  Define VP8_TEMPORAL_ALT_REF to !CONFIG_REALTIME_ONLY
  Remove CONFIG_DEBUG guards from assert()
  Remove unused function vpx_de_mblock
  Fix -Wundef warning for OUTPUT_FPF
  Fix -Wundef warning for __SANITIZE_ADDRESS__
2016-09-02 01:39:18 +00:00
Yaowu Xu
594e53514b Merge "Fix formatting in internal stats for vp8 and vp9" 2016-09-01 23:55:23 +00:00
Yaowu Xu
454139ae13 Merge "Casts to remove some warnings." 2016-09-01 23:37:04 +00:00
Debargha Mukherjee
a6bc3dfb0f Merge "Refactor uv tx size with lookup arrays" 2016-09-01 16:46:32 +00:00
Paul Wilkins
009116cb6f Merge "Modified resize loop constraints." 2016-09-01 15:59:30 +00:00
paulwilkins
3e9e77008c Casts to remove some warnings.
Added casts to remove warnings:
BUG=webm:1274

In regards to the safety of these casts they are of two types:-

- Normalized bits per (16x16) MB stored in a 32 bit int (This is safe as bits
per MB even with << 9 normalization cant overflow 32 bits. Even raw 12
bits hdr source even would only be  29 bits :- (4+4+12+9) and the encoder
imposes much stricter limits than this on max bit rate.

- Cast as part of variance calculations.  There is an internal cast up to 64 bit
for the Sum X Sum calculation, but after normalization dividing by the number
of points the result will always be <= the SSE value.

Change-Id: I4e700236ed83d6b2b1955e92e84c3b1978b9eaa0
2016-09-01 16:10:12 +01:00
Johann
4d1c117f5b Enable -Wundef by default
BUG=webm:1069

Change-Id: I43728f9fd007542718a55d5fdcbc63a8d2f86682
2016-08-31 23:01:57 -07:00
Johann
1139f0dbc2 Define VP8_TEMPORAL_ALT_REF to !CONFIG_REALTIME_ONLY
Previously VP8_TEMPORAL_ALT_REF was only defined for non-realtime-only
builds. However, its value was checked with #if, not #ifdef.

Fixes -Wundef warnings.

BUG=webm:1069

Change-Id: If78d8731298f3f0d3662ffa25f973e7adaf67152
2016-08-31 23:01:57 -07:00
Johann
18b6691105 Remove CONFIG_DEBUG guards from assert()
When 'NDEBUG' is set, assert() generates no code.

Change-Id: Icf61cfc1a8f6e5f0770b3626d8c73ae968df1108
2016-08-31 23:01:57 -07:00
Johann
24f534ac90 Remove unused function vpx_de_mblock
vpx_config.h was not included so CONFIG_POSTPROC was never defined.

Change-Id: I777de499823afa286734549a8e7f4a93e7ad97f3
2016-08-31 23:01:45 -07:00
Johann
7b3c2e3269 Fix -Wundef warning for OUTPUT_FPF
BUG=webm:1069

Change-Id: I3d13d07cf0934e6e262c8033bd77d7197d03ce21
2016-08-31 22:59:59 -07:00
Johann
42ccd79b27 Fix -Wundef warning for __SANITIZE_ADDRESS__
BUG=webm:1069

Change-Id: Iad8811939a910a8f31cf5788220712a255ddf36a
2016-08-31 22:59:53 -07:00
Linfeng Zhang
113f9721d1 Merge "Rename test/lpf_8_test.cc to test/lpf_test.cc" 2016-08-31 22:46:07 +00:00
Linfeng Zhang
5399613889 Rename test/lpf_8_test.cc to test/lpf_test.cc
It actually tests all sizes lpf functions.

Change-Id: Ie31798f90165e6e0c13cbac0e0ab9648ab568bce
2016-08-31 15:16:48 -07:00
Linfeng Zhang
bee7d837ab Update NEON transpose functions.
Unify coding style.

Change-Id: I5826f40c02c882df7353391e0c9dd6cef6bd4b97
2016-08-31 14:58:40 -07:00
Debargha Mukherjee
e6446b4b60 Refactor uv tx size with lookup arrays
Change-Id: Ife6a3d301c5faaba89d16d188d638631083511f7
2016-08-31 13:15:38 -07:00
Linfeng Zhang
3dfba04dec Merge "Update vpx_lpf_vertical_16_dual_neon() intrinsics" 2016-08-31 19:41:25 +00:00
paulwilkins
6fc07a217d Modified resize loop constraints.
Using a tighter resize constraint on undershoot seems to help
results (especially SSIM) as significant undershoot on a frame
seems to have more of a damaging impact than overshoot.

This patch has been tuned so that in local testing using the
derf set it is encode speed neutral for speed  setting 2.

Average quality result for speed 2 (psnr,ssim) were  as follows:-

 lowres  0.039,  0.453
 midres  0.249, 0.853
 hdres  0.159, 0.659
 NetFlix -0.241, 0.360

Change-Id: Ie8d3a0d7d6f7ea89d9965d1821be17f8bda85062
2016-08-31 12:45:49 +01:00
Jim Bankoski
66b2266a22 libyuv: update to de944ed8c74909ea6fbd743a22efe1e55e851b83
Fixes windows build issue:
==> tests::VS10_x64 is broken
         LINK : warning C4742: 'kYvuI601Constants' has different alignment in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': 32 and 2 [.build-x86_64-win64-vs10\vpxdec.vcxproj]
         LINK : warning C4744: 'kYvuI601Constants' has different type in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': '__declspec(align(32)) struct (224 bytes)' and 'struct (224 bytes)' [.build-x86_64-win64-vs10\vpxdec.vcxproj]
         LINK : warning C4742: 'kYuvI601Constants' has different alignment in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': 32 and 2 [.build-x86_64-win64-vs10\vpxdec.vcxproj]
         LINK : warning C4744: 'kYuvI601Constants' has different type in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': '__declspec(align(32)) struct (224 bytes)' and 'struct (224 bytes)' [.build-x86_64-win64-vs10\vpxdec.vcxproj]
         LINK : warning C4742: 'kYvuI601Constants' has different alignment in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': 32 and 2 [.build-x86_64-win64-vs10\vpxenc.vcxproj]
         LINK : warning C4744: 'kYvuI601Constants' has different type in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': '__declspec(align(32)) struct (224 bytes)' and 'struct (224 bytes)' [.build-x86_64-win64-vs10\vpxenc.vcxproj]
         LINK : warning C4742: 'kYuvI601Constants' has different alignment in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': 32 and 2 [.build-x86_64-win64-vs10\vpxenc.vcxproj]
         LINK : warning C4744: 'kYuvI601Constants' has different type in 'third_party\libyuv\source\row_common.cc' and 'third_party\libyuv\source\planar_functions.cc': '__declspec(align(32)) struct (224 bytes)' and 'struct (224 bytes)' [.build-x86_64-win64-vs10\vpxenc.vcxproj]
         LINK : error C2220: warning treated as error - no 'executable' file generated [.build-x86_64-win64-vs10\vpxdec.vcxproj]
         LINK : error C2220: warning treated as error - no 'executable' file generated [.build-x86_64-win64-vs10\vpxenc.vcxproj]

Change-Id: Ic3c4fff9209f5a52ff8f8ff321548d49ba09ec06
2016-08-30 14:24:35 -07:00
Linfeng Zhang
f7cbfed682 Update vpx_lpf_vertical_16_dual_neon() intrinsics
Process 16 samples together.

Change-Id: If6ee8e3377aa2786417f2fc411ba7d87ea8b6799
2016-08-30 11:17:33 -07:00
Paul Wilkins
129814fcb4 Merge "Adjust coefficient optimization and tx_domain rd speed features." 2016-08-30 16:54:40 +00:00
Linfeng Zhang
3a3169be59 Merge "Update vpx_lpf_horizontal_edge_16_neon() intrinsics" 2016-08-29 21:37:07 +00:00
Marco Paniconi
e66cd132f0 Merge "vp8: Move loopfilter synchronization to end of encode_frame call." 2016-08-29 05:52:40 +00:00
Linfeng Zhang
4916515511 Update vpx_lpf_horizontal_edge_16_neon() intrinsics
Process 16 samples together.

Change-Id: I9cfbe04c9d25d8b89f63f48f519e812746db754d
2016-08-27 14:47:48 -07:00
James Zern
3a98508775 Merge "vpx_mem,align_addr: use ~ to create mask" 2016-08-27 21:27:45 +00:00
James Zern
19d881290d vpx_mem,align_addr: use ~ to create mask
removes the need for an intermediate cast to int, which was missing in
the call added in:
69c5ba1 vpx_mem: Refactor code

quiets a visual studio warning:
C4146: unary minus operator applied to unsigned type, result still
unsigned

Change-Id: I76c4003416759c6c76b78f74de7c0d2ba5071216
2016-08-27 11:39:18 -07:00
James Zern
2917737879 vp9_alt_ref_aq_set_nsegments: harmonize fn signature
Change-Id: I5f232664652a8dc3a71e43b8b1fa05ddb4a84ecc
2016-08-27 11:16:03 -07:00
Yury Gitman
507d272265 Move vp9_alt_ref_aq_private.h to vp9_alt_ref_aq.c
+ add a temporary dummy element to ALT_REF_AQ to avoid a warning about
an empty struct

Change-Id: Ib6e5c39ff62ad96eb4e3686d4882228a42b3843f
2016-08-27 10:53:41 -07:00
James Zern
a19b9b6185 Merge changes Ia81004d6,I74b80fb6,I38fcb62b,I2da9cd5d
* changes:
  vpx_mem: add basic size check
  vpx_mem: normalize function names
  vpx_realloc correction.
  vpx_mem: Refactor code
2016-08-26 23:52:04 +00:00
James Zern
ed11abbc36 Merge changes I353da4a2,I423f2153
* changes:
  vp8_decoder_create_threads: check sem/pthread returns
  vp8_create_decoder_instances: add missing setjmp
2016-08-26 23:48:08 +00:00
Johann Koenig
a70861c435 Merge "Remove halfpix specialization" 2016-08-26 21:28:01 +00:00
James Zern
58a497dc29 Merge "add_noise,vpx_setup_noise: correct 'char_dist' type" 2016-08-26 18:47:39 +00:00
James Bankoski
fcc4f3fa21 Merge "libyuv: update to c244a3e9" 2016-08-26 18:06:06 +00:00
Jingning Han
dd2a475e43 Merge "Fix VS build warnings in vp9_alt_ref_aq files" 2016-08-26 17:19:12 +00:00
Paul Wilkins
badd32d914 Merge "Add ALLOW_RECODE_FIRST speed mode." 2016-08-26 15:46:45 +00:00
Jingning Han
84fccfe475 Fix VS build warnings in vp9_alt_ref_aq files
Change-Id: I5b19ec00a1eb8b148026f665d217c12eb50b614a
2016-08-26 08:43:36 -07:00
paulwilkins
dc42f343ae Add ALLOW_RECODE_FIRST speed mode.
This patch is to address concerns that changes to allow
recodes on the first frame in each ARF group do not give a
good enough speed quality trade off for speed 2. Though the
average impact  on encode speed is 1-2%, for some hard clips
it is > 5% rise.  For speed 1 this is less an issue and for Speed 0
the previous patch actually  improves speed.

Change-Id: Ie1bcefdbfdf846d3f4428590173f621465dffe3a
2016-08-26 11:43:47 +01:00
James Zern
a91fe33c6d Merge "vp8: fix decoder crash with invalid leading keyframes" 2016-08-26 07:01:42 +00:00
Sarah Parker
37e83789f1 Fix formatting in internal stats for vp8 and vp9
This corrects a formatting error introduced in:
I1e9d548ce445d29002f0c59ebfd3957a6f15e702
where spaces were used as delimiters instead of tabs.

The corresponding fix for vp10 is in
Ica3d625d6672b3c47e0e208b45eede29b9004030.

Change-Id: Ibc4eb8fd82e6b926ba259a679dc98557cadba9b1
2016-08-25 17:46:18 -07:00
Marco
b6a5f6f740 vp8: Move loopfilter synchronization to end of encode_frame call.
Allow loopfilter to continue until encode_frame is completed.

Change-Id: I7bbccc3d409e263aab6a6ff24588d8b2a964a96e
2016-08-25 12:37:30 -07:00
Yury Gitman
292d221fed Create interface for the ALT_REF_AQ class
Current commit is just an API template  for the rest of the code, and
I will add inner logic later.

Altref  frames  generate a  lot  of  bitrate  and  at the  same  time
other  frames  refer to  them  a  lot, so  it  makes  sense to  apply
special  compensation-based adaptive  quantization scheme  for altref
frames. E.g.,  for blocks  that are  good predictors  for the  future
apply rate-control  chosen quantizer  while for bad  predictors apply
worse one.

Change-Id: Iba3f8ec349470673b7249f6a125f6859336a47c8
2016-08-25 10:55:14 -07:00
Yury Gitman
c018032579 Merge "Add --alt-ref-aq=<int> option" 2016-08-25 17:49:41 +00:00
paulwilkins
635ae8bdc1 Adjust coefficient optimization and tx_domain rd speed features.
Previously Tx domain rd was used in all cases above speed 0.
Coefficient optimization was only enabled for best and speed 0.

This patch selectively sets these features at other speed settings
based on block complexity.

For the Netflix and HD sets in particular the quality gains are
large compared to the speed hit. At speed 1 the average psnr
gain in the NF set  is > 2.5% with one clip coming in at 18%
and some points almost 30%.  Average gains for the lower
resolution test sets are around 1%.

The gains are biggest at low Q so some further optimization
may be possible.

Change-Id: I340376c7b2a78e5389a34b7ebdc41072808d0576
2016-08-25 15:36:16 +01:00
Jim Bankoski
6d7a9f3e9c libyuv: update to c244a3e9
Fixes color issue when scaling without breaking mingw.

BUG=https://bugs.chromium.org/p/libyuv/issues/detail?id=605
BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1252

Change-Id: I09437d93fd65964ad57113274d8c819f3eaf2e57
2016-08-25 06:39:38 -07:00
James Zern
3ddff4503a add_noise,vpx_setup_noise: correct 'char_dist' type
fixes SSE2/AddNoiseTest.CheckCvsAssembly/0 with -funsigned-char.
visibly broken since:
0dc69c7 postproc : fix function parameters for noise functions.
where the types diverged (char vs. int8)
but likely the return changed in:
2ca24b0 postproc - move filling of noise buffer to vpx_dsp.
when multiple implementations were merged.

Change-Id: I176ca1f170217f05ba7872b0c4de63e41949e999
2016-08-24 21:46:26 -07:00
Marco Paniconi
ce634bbf4d Merge "Add datarate tests for encoder multi-threads (vp8 and vp9)." 2016-08-25 03:13:36 +00:00
James Zern
4699aca87f vpx_mem: add basic size check
set a max allocable size to prevent overflows in 32-bit and extremely
large allocation attempts in 64-bit. this could be amended to allow size
or num parameters to be 64-bits with the correct size being used at each
call site.

BUG=webm:819

Change-Id: Ia81004d6c4279680714c4488b4f6cf287ab396a5
2016-08-24 19:22:57 -07:00
James Zern
963291217f vpx_mem: normalize function names
use lower case + '_' rather than capital followed by camel case

Change-Id: I74b80fb660d281228e25edc8b6509455ffe2920e
2016-08-24 19:22:56 -07:00
Urvang Joshi
28c6207bcd vpx_realloc correction.
vpx_realloc was allocating 1 byte more than needed every time.
Fixed this, and took this opportunity to do a small refactoring.

Change-Id: I38fcb62b698894acbbab43466c1decd12f906789
(cherry picked from aom: 2a876b4 aom_realloc correction.)
2016-08-24 19:22:52 -07:00
Urvang Joshi
69c5ba1910 vpx_mem: Refactor code
Change-Id: I2da9cd5da48ae97e770bccfd1233bcc70b484688
(cherry picked from aom: 83c95f5 aom_mem: Refactor code)
2016-08-24 19:22:41 -07:00
Marco
dde8004716 Add datarate tests for encoder multi-threads (vp8 and vp9).
Change-Id: I7f9b23026aaee309095cc3f4724125ae319875af
2016-08-24 16:25:36 -07:00
Yury Gitman
d7c20079a6 Add --alt-ref-aq=<int> option
In the future this option will activate adaptive quantization special
for altref frames. Encoder will  create the adaptive quantization map
on the basis of lookahead buffers similarity which is the estimate of
the future motion compensation performance.

Change-Id: Ia0088b3babb0f9a4899c79d8d819947ba5a03df2
2016-08-24 15:49:25 -07:00
Jacky Chen
5260a6675e Merge "vp9: Refactor set_low_temp_var_flag." 2016-08-24 22:02:53 +00:00
James Zern
a6efe6d437 vp8_decoder_create_threads: check sem/pthread returns
Change-Id: I353da4a2f988ca51d48d0ca91236e8cc0bb48ff5
2016-08-23 19:19:57 -07:00
James Zern
13338a481f vp8_create_decoder_instances: add missing setjmp
vp8_decoder_create_threads() has allocations that expect one is set.

Change-Id: I423f2153a2969c88d48ba45cc9ead4a01443ce65
2016-08-23 18:29:42 -07:00
Johann
d393885af1 Remove halfpix specialization
This function only exists as a shortcut to subpixel variance with
predefined offsets. xoffset = 4 for horizontal, yoffset = 4 for vertical
and both for "hv"

Removing this allows the existing optimizations for the variance
functions to be called. Instead of having only sse2 optimizations, this
gives sse2, ssse3, msa and neon.

BUG=webm:1273

Change-Id: Ieb407b423b91b87d33c4263c6a1ad5e673b0efd6
2016-08-23 17:05:39 -07:00
James Zern
0f42d1fa85 vp8: fix decoder crash with invalid leading keyframes
decoding the same invalid keyframe twice would result in a crash as the
second time through the decoder would be assumed to have been
initialized as there was no resolution change. in this case the
resolution was itself invalid (0x6), but vp8_peek_si() was only failing
in the case of 0x0.
invalid-vp80-00-comprehensive-018.ivf.2kf_0x6.ivf tests this case by
duplicating the first keyframe and additionally adds a valid one to
ensure decoding can resume without error.

BUG=b/30593765

Change-Id: If0859035908b7870d67a7f3f646b5a080252eb6d
2016-08-23 16:27:52 -07:00
Yury Gitman
c325fb748a Correct CHECK_MEM_ERROR macro (release builds)
The previous macro doesn't work with &cpi->common as a first argument

Change-Id: Iddf7a1f5d56d7abafd9b2b8707aa611d349e7a68
2016-08-23 22:46:04 +00:00
jackychen
8d4c0ec1f1 vp9: Refactor set_low_temp_var_flag.
No need to pass in force_split, since we should use sb_type in the
condition.

Change-Id: Ide27243ef46e017bbb98d676347fc566a6c828f7
2016-08-23 15:11:40 -07:00
Yunqing Wang
f6c5410cd4 Merge "Disable split mode in 4k video encoding" 2016-08-23 15:35:33 +00:00
Yunqing Wang
ef98f49cb0 Disable split mode in 4k video encoding
Disabled the split mode while encoding 4k video to speed
up the encoder.

Borg test result on 4k set:
Overall PSNR: +0.029%; SSIM: +0.009%.
Average encoder speedup at speed 2 is 2.5%.

Change-Id: I1519c658f07c3ac838affbe5aff0ed9b94f3f8f4
2016-08-22 19:46:44 -07:00
Yury Gitman
bf7a02a4cf Correct CHECK_MEM_ERROR macro
The previous macro doesn't work with &cpi->common as a first argument

Change-Id: Ic3f5c49a94cf8b17de6569811b957c963341bb58
2016-08-22 14:25:57 -07:00
Marco Paniconi
f5bd76f5c1 Merge "Revert "vp8: Move loopfilter synchronization to end of encode_frame call."" 2016-08-22 15:46:57 +00:00
Marco Paniconi
de075a95e0 Revert "vp8: Move loopfilter synchronization to end of encode_frame call."
This reverts commit c2fe9acced.

This change break linux browser test in chromium:
https://build.chromium.org/p/chromium.webrtc/builders/Linux%20Tester

Change-Id: I226782fad480c17a99ec6c785ad93cf4ab88f0ae
2016-08-22 15:46:20 +00:00
Yunqing Wang
37169c0bd4 Merge "Adjust speed features for 4k video encoding" 2016-08-19 23:11:05 +00:00
Yunqing Wang
fe488cceff Adjust speed features for 4k video encoding
Adjusted speed 2 features to speed up 4k video encoding.
BDBR results from borg test:
PSNR: +0.313%; SSIM: +0.268%.
Average speedup: 8.5%

Change-Id: I1e2695a01fb3f3817c1df4480e184c2aed8f2eba
2016-08-19 09:30:32 -07:00
James Zern
149d082377 vp9_pickmode: quiet float conversion warnings
Change-Id: I591e4f958955b3f2edb2f95a83c54cd83c8ef075
2016-08-19 01:28:01 -07:00
James Zern
8b4c31584e vp9_alloc_context_buffers: clear cm->mi* on failure
this fixes a crash in vp9_dec_setup_mi() via
vp9_init_context_buffers() should decoding continue and the decoder
resyncs on a smaller frame

BUG=b/30593752

Change-Id: I9ce8d94abe89bcd058697e8bd8599690e61bd380
2016-08-19 00:18:11 -07:00
Jacky Chen
52db2b1690 Merge "vp9 svc: SVC encoder speed up." 2016-08-18 21:21:29 +00:00
Johann Koenig
33dedd0628 Merge "Remove '-chromium' flag from ads2gas_apple.pl" 2016-08-18 19:54:55 +00:00
JackyChen
8be7e572a7 vp9 svc: SVC encoder speed up.
Bias towards base_mv and skip 1/4 pixel motion search when using base mv.
2~3% speed up for 2 spatial layers, 3~5% speed up for 3 spatial layers.
PSNR loss:
(2 layers) 0.07dB for gips_stationary, 0.04dB for gips_motion;
(3 layers) 0.07dB for gips_stationary, 0.06dB for gips_motion.

Change-Id: I773acbda080c301cabe8cd259f842bcc5b8bc999
2016-08-18 11:25:45 -07:00
Marco Paniconi
1c07abca18 Merge "vp9 non-rd pickmode: Add limit on newmv-last and golden bias." 2016-08-18 18:03:48 +00:00
Marco Paniconi
37a39ac138 Merge "vp8: Move loopfilter synchronization to end of encode_frame call." 2016-08-18 02:46:31 +00:00
Marco
7eb7d6b227 vp9 non-rd pickmode: Add limit on newmv-last and golden bias.
Add option, for newmv-last, to limit the rd-threshold update for early exit,
under a source varianace condition.
This can improve visual quality in low texture moving areas,
like forehead/faces.

Also add bias against golden to improve the speed/fps,
will little/negligible loss in quality.

Only affects CBR mode, non-svc, non-screen-content.

Change-Id: I3a5229eee860c71499a6fd464c450b167b07534d
2016-08-17 14:33:44 -07:00
Johann
1b982cc64f Remove '-chromium' flag from ads2gas_apple.pl
The flag was added because Apple clang and Chromium clang disagreed
for certain versions of instructions.

qsubaddx, qaddsubx, ldrneb and ldrneh were used in armv6 assembly
which was removed in d55724fae9

vqshrun was used in some neon assembly but superseded by
dcbfacbb98

.include was used for obj_int_extract/asm_offsets and removed in
6eec73a747

Change-Id: I32f4c9b536d0318482101c0b8e91e42b8f545f18
2016-08-17 14:05:16 -07:00
paulwilkins
af3b0de732 Add casting to fix warning.
Frame bits can safely be stored int but group bits
(kf or arf) use 64bit.

Change-Id: I0800f2a28070f8749110a95721c116fc56987885
2016-08-17 11:18:07 +01:00
paulwilkins
ab7cd6d068 Add {} to try and keep Jenkins happy.
Change-Id: If1ca3cf83e058317c9751d7da6caa7cd75eb6845
2016-08-17 11:17:36 +01:00
Marco
c2fe9acced vp8: Move loopfilter synchronization to end of encode_frame call.
Change-Id: I5bdfea7f51df1f1fa5d9c1597e96988acce6c2f2
2016-08-16 11:22:23 -07:00
Linfeng Zhang
f9efbad392 NEON asm of vpx_lpf_{horizontal,vertical}_8_dual_neon()
Also expose the NEON intrinsics version.

BUG=webm:1261, webm:1266.

Change-Id: I8c4ae658467dcf66ebf7a75982b2ef712dbb4535
2016-08-16 08:50:57 -07:00
paulwilkins
5d881770e5 Change default recode rule for good speed 0 and best.
Changes the default recode rule for Speed 0 and best quality
from ALLOW_RECODE to ALLOW_RECODE_KFARFGF.

Tested on the NF, hdres, midres and lowres test sets, this setting
when combined with patch I40cb559... now performs "as well" in
metrics terms (in fact it came out a tiny amount better overall)
but encode time is 9.6%  faster (measured as the average
from 27 mid rate local encodes on clips in the derf/lowres set.

Change-Id: I8c781c0cdfa3a9929cd9406d15582fce47d6ae3b
2016-08-15 10:52:54 +01:00
paulwilkins
de3b769524 Change to recode rules.
Allow recodes for the first inter frame in each arf group
even when the recode rule is set to ALLOW_RECODE_KFARFGF.

Small gains of 0.05%.

Change-Id: I40cb559d36a2bf0ebf5cf758c3f92e452b480577
2016-08-15 10:52:02 +01:00
Paul Wilkins
fe4dd4f43f Merge "Modified ARF group allocation." 2016-08-15 09:42:30 +00:00
Yunqing Wang
fafec95702 Merge "Fix another motion vector out of range bug" 2016-08-12 23:52:14 +00:00
James Zern
dfcefe06fa Merge "variance_impl_avx2: restore table layout" 2016-08-12 23:02:27 +00:00
James Zern
bd7cfb46fb variance_impl_avx2: restore table layout
disable clang-format for bilinear_filters_avx2

restores the row layout prior to:
099bd7f vpx_dsp: apply clang-format
but keeps the justification used by clang-format

Change-Id: Icf1733a37edb807e74c26b23a93963c03bd08fd7
2016-08-12 11:52:53 -07:00
Linfeng Zhang
f09b5a3328 NEON intrinsics for 4 loopfilter functions
New NEON intrinsics functions:
vpx_lpf_horizontal_edge_8_neon()
vpx_lpf_horizontal_edge_16_neon()
vpx_lpf_vertical_16_neon()
vpx_lpf_vertical_16_dual_neon()

BUG=webm:1262, webm:1263, webm:1264, webm:1265.

Change-Id: I7a2aff2a358b22277429329adec606e08efbc8cb
2016-08-12 09:58:17 -07:00
Yunqing Wang
a413dbe594 Fix another motion vector out of range bug
This patch fixed a motion vector out of range bug:
vpxenc: ../libvpx/vp9/encoder/vp9_mcomp.c:69:
 mv_cost: Assertion `mv->col >= -((1 << (11 + 1 + 2)) - 1) &&
 mv->col < ((1 << (11 + 1 + 2)) - 1)' failed.

For blocks that returned without having full-pixel search, the original
MV limits were not restored, which caused the failure. Moved the set
MV limit function down to fix the bug.

Change-Id: Id7d798fc7214e95c6e4846c588f0233fcf1a4223
2016-08-12 09:27:58 -07:00
Marco
f1e12c1bf3 vp8: Fix denoiser setting in multi-res sample encoder.
Change-Id: I9222f3b252e5ed883659f1a14cd705944ee9da07
2016-08-10 16:22:08 -07:00
paulwilkins
656f4a88cf Modified ARF group allocation.
Small average gains in the range 0.05 - 0.1

Change-Id: I30e85c04be615cc84726427c5057388b20a6ff60
2016-08-10 14:22:01 -07:00
Aleksey Vasenev
343b6b09a1 Align thread entry point stack
_beginthreadex does not align the stack on 16-byte boundary as expected
by gcc.

On x86 targets, the force_align_arg_pointer attribute may be applied to
individual function definitions, generating an alternate prologue and
epilogue that realigns the run-time stack if necessary. This supports
mixing legacy codes that run with a 4-byte aligned stack with modern
codes that keep a 16-byte stack for SSE compatibility.
https://gcc.gnu.org/onlinedocs/gcc/x86-Function-Attributes.html

Change-Id: Ie4e4ab32948c238fa87054d5664189972ca6708e
Signed-off-by: Aleksey Vasenev <margtu-fivt@ya.ru>
2016-08-10 11:57:34 -07:00
James Zern
4916a87bfc Merge changes I1d3edbdb,I8b49fd05
* changes:
  tests: use scoped_ptr for local video source vars
  y4m_test: init members in the constructor
2016-08-10 00:05:58 +00:00
Alex Converse
941fe20336 Merge "Refactor mv limits." 2016-08-09 17:12:50 +00:00
James Zern
475e9d26e0 tests: use scoped_ptr for local video source vars
prevents leak warnings on ASSERT*() failures

Change-Id: I1d3edbdbb18dbbe3b17691971348a8121cf09afa
2016-08-08 14:43:14 -07:00
Yury Gitman
c37d012ada Merge "Add cpi parameter for forcing segmentation update" 2016-08-08 21:29:42 +00:00
James Zern
9e9722bc79 y4m_test: init members in the constructor
prevents use of an uninitialized value in the deconstructor should the
test fail before tmpfile_ is set.

Change-Id: I8b49fd05f0d05e055fdf653bd46983d30f466a68
2016-08-08 14:27:34 -07:00
Yury Gitman
7a730d5901 Add cpi parameter for forcing segmentation update
Change-Id: I1b0bcb1ffe7604117bfaa0b9989d0e25ff04d28c
2016-08-08 13:20:42 -07:00
James Zern
cfd92dab18 Merge changes from topic 'clang-tidy'
* changes:
  *_perf_test.cc: correct DoDecode signature
  test: apply clang-tidy google-readability-braces-around-statements
2016-08-08 20:12:42 +00:00
Alex Converse
6554333b59 Refactor mv limits.
Change-Id: Ifebdc9ef37850508eb4b8e572fd0f6026ab04987
2016-08-08 11:54:00 -07:00
Yunqing Wang
6a8d4631a8 Merge "Fix a motion vector out of range bug" 2016-08-08 17:59:50 +00:00
James Zern
2c17d54681 *_perf_test.cc: correct DoDecode signature
+ delete unused kMaxPsnr from decode_perf_test.cc

Change-Id: Id93347631e7870491069a8b7c5bb1f6b2828425f
2016-08-05 20:21:02 -07:00
clang-format
9c9d92ae3a test: apply clang-tidy google-readability-braces-around-statements
applied against a x86_64 configure with and without
--enable-vp9-highbitdepth

clang-tidy-3.7.1 \
  -checks='-*,google-readability-braces-around-statements' \
  -header-filter='.*' -fix
+ clang-format afterward

Change-Id: Ia2993ec64cf1eb3505d3bfb39068d9e44cfbce8d
2016-08-05 20:02:28 -07:00
Linfeng Zhang
2d1e63d0c5 Remove duplicates in Loop8Test6Param and Loop8Test9Param
Extract the duplicated data generation code in OperationCheck() of
Loop8Test6Param and Loop8Test9Param, and put in function InitInput().

Change-Id: Ied39ba4ee86b50501cc5d10ebf54f5333c4708f0
2016-08-05 19:51:01 -07:00
James Zern
c12f2f3187 Merge "remove tools/vpx-style.sh" 2016-08-06 01:23:13 +00:00
James Zern
19d2e73dea Merge changes Ice037acb,I806af11b,I344a7dd0,Ib7cb87fa
* changes:
  vp9: normalize vpx_enc_frame_flags_t usage
  args.c: add some explicit casts
  webmdec: quiet -Wshorten-64-to-32 warning
  test/decode_test_driver: rm unused deadline member
2016-08-06 01:20:52 +00:00
Linfeng Zhang
ba42ce64b7 Fix a bug in test/lpf_8_test.cc
This bug is introduced in 36608af524,
where buffer tmp_s is not fully initialized.

Change-Id: I125b966cf054a82bc63c72647cdd463f434eda17
2016-08-05 17:52:10 -07:00
Yunqing Wang
2fb826c4d5 Fix a motion vector out of range bug
This patch fixed a motion vector(MV) out of range bug, which was caused
by not restoring the original values of the MV min/max thresholds after
the sub8x8 full pixel motion search. It occurred rarely and only was seen
while encoding a 4k clip for 200 frames.

BUG=webm:1271

Change-Id: Ibc4e0de80846f297431923cef8a0c80fe8dcc6a5
2016-08-05 15:23:05 -07:00
James Zern
7104833085 vp9: normalize vpx_enc_frame_flags_t usage
quiets -Wshorten-64-to-32 warnings

Change-Id: Ice037acb675d1d81bfedf2dfcfa91a8a29a19dfd
2016-08-04 23:37:49 -07:00
James Zern
d772d55704 args.c: add some explicit casts
values are range checked before returning; quiets -Wshorten-64-to-32
warnings

Change-Id: I806af11b2aaf6760c7ab234a2fe2fdf40e7bece7
2016-08-04 23:37:49 -07:00
James Zern
c79665d0ad webmdec: quiet -Wshorten-64-to-32 warning
track->GetNumber() will fit in an int in well-behaved files

Change-Id: I344a7dd05d04daf3df2d67358ea69f8014a03a5b
2016-08-04 23:37:49 -07:00
James Zern
1b1e40c0b2 test/decode_test_driver: rm unused deadline member
has the side-effect of removing some lint and -Wshorten-64-to-32
warnings

Change-Id: Ib7cb87fa65cd65534096921f243d15288e97256d
2016-08-04 23:36:53 -07:00
James Zern
958ae5af9c remove tools/vpx-style.sh
update ftfy.sh to use clang-format

Change-Id: I8ac740c5b3842beed2b8878fbe506f381f4c57e4
2016-08-04 20:17:09 -07:00
Johann Koenig
57f49db81f Merge changes I6ef79702,Id332c641,I354b5d22,I84438013
* changes:
  Use common transpose for vpx_idct32x32_1024_add_neon
  Use common transpose for vpx_idct8x8_[12|64]_add_neon
  Use common transpose for vp9_iht8x8_add_neon
  Use common transpose for vpx_idct16x16_[10|256]_add_neon
2016-08-04 22:30:47 +00:00
Johann Koenig
17720b60bb Merge "Remove armv6 target" 2016-08-04 22:21:13 +00:00
James Zern
7f7c888c14 Merge "correct break placement" 2016-08-04 22:19:30 +00:00
Johann
0325b95938 Use common transpose for vpx_idct32x32_1024_add_neon
Change-Id: I6ef7970206d588761ebe80005aecd35365ec50ff
2016-08-04 20:13:18 +00:00
Johann
f4e4ce7549 Use common transpose for vpx_idct8x8_[12|64]_add_neon
Change-Id: Id332c641f05336ef9a45e17493ff149fd0a168f0
2016-08-04 20:13:12 +00:00
Johann
7103b5307d Use common transpose for vp9_iht8x8_add_neon
Change-Id: I354b5d22130d76b0eceda0748db1f871f58fa372
2016-08-04 20:13:03 +00:00
Johann
8619203ddc Use common transpose for vpx_idct16x16_[10|256]_add_neon
Change-Id: I84438013f483e82084d33ba9a63c33273d35fcaa
2016-08-04 20:12:53 +00:00
Johann Koenig
b757d89ff9 Merge "Extract neon transpose for re-use" 2016-08-04 20:12:38 +00:00
James Zern
4db9bd324d Merge "vp9_ratectrl.c: apply clang-format" 2016-08-04 20:01:46 +00:00
James Zern
70a7885a65 correct break placement
these should be placed within {}s when present

Change-Id: Ia775fac5373603e77360398f19b07958fb43f476
2016-08-04 13:00:14 -07:00
Johann Koenig
caac87b05b Merge "Don't expand to Q register for 4x4 intrapred" 2016-08-04 19:55:50 +00:00
Johann
d55724fae9 Remove armv6 target
Change-Id: I1fa81cc9cabf362a185fc3a53f1e58de533a41e5
2016-08-04 12:55:06 -07:00
Johann Koenig
476e8fc855 Merge "Pad 'Left' when building under ASan" 2016-08-04 19:27:45 +00:00
Linfeng Zhang
36608af524 Merge "Update Loop8Test{6,9}Param to test filter8() in mb_lpf_vertical_edge_w()" 2016-08-04 19:21:22 +00:00
Johann
377cfa31f0 Extract neon transpose for re-use
Change-Id: I5e1c7f4c80d1c6f7fd582ac468c6eaaa3603a06c
2016-08-04 19:04:25 +00:00
James Zern
374f0ff4a0 Merge changes from topic 'clang-format'
* changes:
  README: add a note about clang-format
  README: update target list
  README: fix typo
2016-08-04 19:03:03 +00:00
clang-format
3a4002b94d vp9_ratectrl.c: apply clang-format
after:
ff0a87c vp9 1pass vbr: Adjustment to gf interval.

Change-Id: I1296e53e601bf0c2b562e3a34082ac45c294a5f1
2016-08-04 11:57:00 -07:00
Johann
df69c751a7 Don't expand to Q register for 4x4 intrapred
The code was expanding to Q registers so that vqrshn could be used, for
vector quad round shift and narrow. If 4 values are added together,
there is a shift by 2. If 8 values, a shift by 3. Since this accounts
for any possibility of overflow, we can skip the narrowing shift.

This allows keeping the values in D registers and casting the 16 bit
value to 8 bits.

Change-Id: I8d9cfa07176271f492c116ffa6a7b351af0b8751
2016-08-04 18:51:46 +00:00
Linfeng Zhang
bbf4c91f79 Update Loop8Test{6,9}Param to test filter8() in mb_lpf_vertical_edge_w()
One branch of filter8() in mb_lpf_vertical_edge_w() was not tested.

Change-Id: I194202d771d9acd6b4e5e600ee2bae89986b49f3
2016-08-04 11:33:14 -07:00
Marco Paniconi
9fdeeaf411 Merge "vp9 1pass vbr: Adjustment to gf interval." 2016-08-04 17:50:55 +00:00
Yaowu Xu
7a79fa1362 Fix msvc compiler warnings
MSVC 2013 complained about using 32 shift where 64 bit shift should be
used.

Change-Id: I7a2b165d1a92d3c0a91dd4511b27aba7709b5e55
2016-08-03 18:33:06 -07:00
James Zern
b51d127c82 Merge "Resolve -Wshorten-64-to-32 warnings in prob.h." 2016-08-04 00:38:08 +00:00
James Zern
15f29ef092 README: add a note about clang-format
Change-Id: I835401e3befffcbc68e7d2bdd2fd556a19948e91
2016-08-03 17:34:03 -07:00
James Zern
77f5c3d2e8 README: update target list
Change-Id: I80293720a5f12bc2449ceaadbb2ad0f924141552
2016-08-03 17:30:45 -07:00
James Zern
5ea8712b82 README: fix typo
Change-Id: I2c3ecc62b1fd1e600b3d70b623c8b11e1e8e4d13
2016-08-03 17:30:45 -07:00
James Zern
068281751c Merge "test: apply clang-format" 2016-08-04 00:27:59 +00:00
James Zern
a412c004e4 Merge "vp9/decoder,vp9/*.[hc]: apply clang-format" 2016-08-04 00:22:59 +00:00
Johann
a7a8e07a44 Pad 'Left' when building under ASan
The neon intrinsics are not able to load just the 4 values that are
used. In vpx_dsp/arm/intrapred_neon.c:dc_4x4 it loads 8 values for both
the 'above' and 'left' computations, but only uses the sum of the first
4 values.

BUG=webm:1268

Change-Id: I937113d7e3a21e25bebde3593de0446bf6b0115a
2016-08-03 16:38:51 -07:00
Marco
ff0a87ce38 vp9 1pass vbr: Adjustment to gf interval.
Increase the minimum distance.
Reduces the overshoot somewhat on some clips,
small gain in avgPSNR (~0.1%) on ytlive set.

Change-Id: Id5ddde20c2907dbdb536e79542eff775019c142b
2016-08-03 15:36:27 -07:00
clang-format
08131055e4 vp9/decoder,vp9/*.[hc]: apply clang-format
Change-Id: Ic38ea06c7b2fb3e8e94a4c0910e82672a1acaea7
2016-08-03 14:29:31 -07:00
Yaowu Xu
85e111b3ba Merge "vp9 svc: Fix a valgrind error." 2016-08-03 20:53:05 +00:00
clang-format
8ff40f8bec vp9/common: apply clang-format
Change-Id: Ie0f150fdcfcbf7c4db52d3a08bc8238ed1c72e3b
2016-08-02 18:27:07 -07:00
clang-format
e0cc52db3f vp9/encoder: apply clang-format
Change-Id: I45d9fb4013f50766b24363a86365e8063e8954c2
2016-08-02 16:47:11 -07:00
JackyChen
f7032713af vp9 svc: Fix a valgrind error.
This error was introduced by the patch:
8ce67d7 vp9 svc: Enable different speed setting for each spatial layer.
To use svc, svc_param should be cleared to 0 at the beginning.

Change-Id: I222f03ddae8a50e84b4690b78263abb742fae91e
2016-08-02 16:16:22 -07:00
Alex Converse
d089ac4dda Resolve -Wshorten-64-to-32 warnings in prob.h.
Change-Id: I1244ee908d81467f0fc8a8fce979fc8077a325b4
2016-08-02 15:40:23 -07:00
Alex Converse
3a04c9c9c4 Merge "Resolve -Wshorten-64-to-32 in variance." 2016-08-02 22:26:55 +00:00
Yaowu Xu
039f9e08f0 change HBD pixel value from uint8_t to uint16_t
This fixes a regression in 10/12 bit encoding results.

Change-Id: I438877352a41aae0a864a8d9979afe4aa2061d81
2016-08-02 11:01:39 -07:00
Yaowu Xu
dc5618f3bb Add pointer conversion for HBD buffers
This fixes a crash in HBD build.

Change-Id: I7f688f50227323e69bba65df0d56f4360f01771b
2016-08-01 15:56:43 -07:00
Alex Converse
004eebed31 Merge "Unfork 8-bit in HBD path in vp9_model_rd_from_var_lapndz callers." 2016-08-01 16:42:39 +00:00
Alex Converse
2c3807b89f Merge "Cache optimizations in optimize_b()." 2016-08-01 16:30:05 +00:00
Alex Converse
e446ffda45 Cache optimizations in optimize_b().
Move best index into the token state. Shrink it down to one byte. This
is more cache friendly (access are group together) and uses less total
memory.

Results in 4% fewer cycles in optimize_b().

Change-Id: I75db484fb3dc82f59928d54b659d79c80ee40452
2016-07-29 12:06:49 -07:00
Johann Koenig
d4ab234869 Merge "replace by VSTM/VLDM to reduce one of VST1/VLD1" 2016-07-29 14:25:10 +00:00
Min Chen
407c2e2974 replace by VSTM/VLDM to reduce one of VST1/VLD1
Change-Id: I596567570580babb1a52925541d1fd1045c352f5
2016-07-28 23:01:38 +00:00
JackyChen
6fbb4c3061 vp8: Switch skin model to mode 0 to save some cycle.
This change will speed up vp8 encoder by 1.5% ~ 2% on linux. No
much speed change on Mac.

Change-Id: Id957f19ddd89805baa2af84c5027d52d9a48553f
2016-07-28 13:32:50 -07:00
Jacky Chen
462a7c9f0a Merge "vp9 svc: Enable different speed setting for each spatial layer." 2016-07-28 20:21:30 +00:00
Alex Converse
c0241664aa Resolve -Wshorten-64-to-32 in variance.
The subtrahend is small enough to fit into uint32_t.

Change-Id: Ic4d7128aaa665eaf6b25d562610ba8942c46137f
2016-07-28 10:16:31 -07:00
Alex Converse
4508eb3123 Merge "Fix 64 to 32 narrowing warning." 2016-07-28 16:36:46 +00:00
clang-format
956af1d478 vpx_dsp/x86/quantize_sse2.c: apply clang-format
post:
e429080 .clang-format: disable DerivePointerAlignment

Change-Id: I21a0546668edb2b09660e216d4875a1d2ad24d53
2016-07-27 21:41:18 -07:00
James Zern
6b374abc86 Merge "vp9 denoiser: Derefencing pointer should be after null check." 2016-07-28 00:43:19 +00:00
Alex Converse
335cf67d8b Fix 64 to 32 narrowing warning.
- Solves potential integer overflow on 12-bit
- Fixes Visual Studio build

Change-Id: I26dd660451bbab23040e4123920d59e82585795c
2016-07-27 12:40:23 -07:00
James Zern
341919d038 Merge "vpx_scale: apply clang-format" 2016-07-27 01:59:21 +00:00
clang-format
33e40cb5db test: apply clang-format
Change-Id: I0d9ab85855eb723f653a7bb09b3d0d31dd6cfd2f
2016-07-27 01:58:52 +00:00
JackyChen
47cc64cdf8 vp9 denoiser: Derefencing pointer should be after null check.
BUG=webm:1267

Change-Id: I899fc9e8d784c6eefcbe27945c619845adb7b6f0
2016-07-26 17:31:17 -07:00
James Zern
e4290800b2 .clang-format: disable DerivePointerAlignment
everything outside of third_party should follow 'PointerAlignment:
right' i.e., associate the '*' with the variable

+ add a note about the clang-format that generated this file

Change-Id: I13e3f4f5fb6e22a8fa7fc3d06879c995b7c41a39
2016-07-26 16:46:54 -07:00
clang-format
f4be884466 vpx_scale: apply clang-format
Change-Id: Ia07ba57756f75911d3d06318e1f9b1982e1ca8c5
2016-07-26 15:57:41 -07:00
James Zern
fbf256da41 Merge "vpx_ports: apply clang-format" 2016-07-26 22:54:31 +00:00
Alex Converse
34201e50c1 Unfork 8-bit in HBD path in vp9_model_rd_from_var_lapndz callers.
BUG=b/29583530

Change-Id: Ia88a75f9572e08f228559ab84b8a77efb5aff0af
2016-07-26 21:57:58 +00:00
James Zern
1a3c4f91f6 Merge "vpx_mem: apply clang-format" 2016-07-26 21:19:17 +00:00
James Zern
9f9a8d2aaa Merge "vpx_util: apply clang-format" 2016-07-26 21:18:25 +00:00
Alex Converse
1c85230344 Merge "Only consider visible 4x4s in pixel domain error." 2016-07-26 19:39:54 +00:00
James Zern
7987686397 Merge "register_state_check: simplify Check() methods" 2016-07-26 18:49:18 +00:00
clang-format
6565c17f24 vpx_util: apply clang-format
Change-Id: Ie7eab608e2906b9a2b3533db95292ebc430ad377
2016-07-25 22:33:21 -07:00
James Zern
f8c27d164c register_state_check: simplify Check() methods
- make Check() void as the EXPECT's are sufficient to document failure

cumulatively this has the effect of avoiding reporting incorrect Check()
failures due to earlier test failures.

Change-Id: I2cf775449f18c90c1506b8eadd7067adbc3ea046
2016-07-25 15:14:02 -07:00
jackychen
8ce67d714a vp9 svc: Enable different speed setting for each spatial layer.
This change only affects 1 pass cbr svc mode.

Change-Id: If0da87bb200f7e7762755340c40c8157cc7a16ca
2016-07-25 15:11:43 -07:00
Alex Converse
d6c5ef4557 Only consider visible 4x4s in pixel domain error.
BDRATE change
derf144: -0.327
lowres: -0.048
midres: -0.125
hdres: -0.238

Change-Id: I789aba9870b5c2952373a7dd4fc8ed45590c3c54
2016-07-25 21:44:06 +00:00
clang-format
580f14b68b vpx_ports: apply clang-format
Change-Id: Ice343335a40238fd21490bce0ce2972bdcb87055
2016-07-25 14:29:06 -07:00
clang-format
d7a3b781d3 vpx_mem: apply clang-format
Change-Id: I0440686fc03f1ee02bd0168c91e671a0a2d0056a
2016-07-25 14:17:59 -07:00
clang-format
099bd7f07e vpx_dsp: apply clang-format
Change-Id: I3ea3e77364879928bd916f2b0a7838073ade5975
2016-07-25 14:14:19 -07:00
James Zern
82070ae939 Merge "configure: test for -Wfloat-conversion" 2016-07-25 19:47:38 +00:00
Johann Koenig
5c0f5cdda8 Merge "Fix compilation error under Clang 4.0." 2016-07-25 19:44:05 +00:00
Ivan Krasin
91369fd9b7 Fix compilation error under Clang 4.0.
The LLVM trunk has reached 4.0 and now __clang_major__ is not enough
to distinguish between old XCode Clang and the new 'real' Clang.
Using __apple_build_version__ allows to make this distinction.

BUG=chromium:631144

Change-Id: I0b6e46fddfe4f409c7b7e558bda34872e60ee2d9
2016-07-25 19:18:49 +00:00
James Zern
889ed5b158 configure: test for -Wfloat-conversion
supported by clang, gcc-4.9+

Change-Id: I893766de7307fef9a8b68c0cfae137c9d3b0dbe8
2016-07-25 19:05:15 +00:00
James Zern
7aa0c748b3 Merge "vp9: fix frame-level threaded decode shutdown" 2016-07-25 19:00:37 +00:00
Alex Converse
511bf49b7e Merge "Minor skip segment simplification." 2016-07-25 17:50:43 +00:00
Scott LaVarnway
ad5fea03e6 Merge "VP9: get_pred_context_switchable_interp() -- encoder side" 2016-07-25 11:58:24 +00:00
James Zern
54b2071bf4 vp8/decodeframe: fix signed/unsigned comparison
quiets a visual studio warning

Change-Id: Ic7725616bc2cb837e6f79294d4fcff36b67af834
2016-07-23 11:41:52 -07:00
James Zern
f368f86df6 vp9: fix frame-level threaded decode shutdown
Shutdown all threads before reclaiming any memory. The frame-level
parallel decoder may access data from another worker.

BUG=webm:1259

Change-Id: I26856ebd1f77cc4a4545331baa19bbf3e01c4ea4
2016-07-23 10:59:15 -07:00
clang-format
c42d54c3a3 vp8/postproc.c: disable clang-format for RGB_TO_YUV
Change-Id: Id2a936301ec1e3d5648b4f8adbf4e6625002589d
2016-07-23 10:55:44 -07:00
James Zern
18e53642b7 Merge "vpx/: apply clang-format" 2016-07-23 01:27:14 +00:00
James Zern
5cf8eda308 Merge changes I0089e884,Icb0ecb9e
* changes:
  vp8/postproc: fix implicit float conversion
  blockiness_test: fix implicit float conversion
2016-07-23 01:18:32 +00:00
James Zern
256a4af4d1 Merge "resize_test: fix implicit float->int conversion" 2016-07-23 01:17:49 +00:00
James Bankoski
7381e3330f Merge "vp8:fix threading issues" 2016-07-23 00:51:37 +00:00
Jim Bankoski
0fff2fb34c vp8:fix threading issues
1 - stops de allocating before threads are closed.
2 - limits threads to mb_rows when mb_rows < partitions

BUG=webm:851

Change-Id: I7ead53e80cc0f8c2e4c1c53506eff8431de2a37e
2016-07-23 00:50:55 +00:00
James Zern
b2542417cd vp8/postproc: fix implicit float conversion
float->int as reported by -Wfloat-conversion

Change-Id: I0089e8847b218c47526bcfbb0fffd9aad7c5adb3
2016-07-22 16:01:52 -07:00
Yury Gitman
3d3f51262c Add VPX_SWAP macro
Change-Id: I60e233eddef238ad918183392794084673f27d2d
2016-07-22 15:41:25 -07:00
James Zern
5e2791b54d blockiness_test: fix implicit float conversion
float->int as reported by -Wfloat-conversion

Change-Id: Icb0ecb9e2d54edb95813d9f2de34cb6c27b63cbd
2016-07-22 15:35:42 -07:00
Alex Converse
9a62ecbd35 Minor skip segment simplification.
Change-Id: I34863fce1abe94f9539e9a5a6149ae1efb6501bd
2016-07-22 15:31:18 -07:00
Marco Paniconi
53db633349 Merge "vp9 1pass-vbr: Adjust gf setting for nonzero-lag case." 2016-07-22 21:27:05 +00:00
James Zern
325bdddc38 resize_test: fix implicit float->int conversion
Change-Id: I1efc16fa158740a06da719a1ea90c6dd6a182bb4
2016-07-22 13:11:07 -07:00
Marco
c06a4b9df2 vp9 1pass-vbr: Adjust gf setting for nonzero-lag case.
Change-Id: I230c586c6d5ae56ee9a6d37b7d9452351bb4bd80
2016-07-22 11:48:09 -07:00
Paul Wilkins
830fa866a5 Merge "Sample points to reduce encode overhead." 2016-07-22 09:27:34 +00:00
Paul Wilkins
063e4a2914 Merge "Noise energy Experiment in first pass." 2016-07-22 09:27:19 +00:00
clang-format
e3e9fee419 vpx/: apply clang-format
Change-Id: I95922a64568bf289863c1564212b6be5beec36df
2016-07-21 20:49:07 -07:00
Yunqing Wang
4b073bc39a Add back header in threading.h
Added back the header needed in threading.h

Change-Id: I2ce66ad4fe58004997623f6c3f3b8dd11640aa98
2016-07-21 17:26:05 -07:00
Yunqing Wang
930773a1ed Merge "Revert "Amend and improve VP8 multithreading implementation"" 2016-07-21 21:32:55 +00:00
Yunqing Wang
87c6c5224d Revert "Amend and improve VP8 multithreading implementation"
Reverted the patch because of possible performance issue.

Change-Id: I49944f827ccd38ed194c9f8d9cb9036fa9bf79e1
2016-07-21 12:28:25 -07:00
Scott LaVarnway
c969b2b02b VP9: get_pred_context_switchable_interp() -- encoder side
Change-Id: I7217c90d5cf38c51b76759a2dc4f10070f3a40ac
2016-07-21 11:47:51 -07:00
Alex Converse
18c7f46c12 MinArfFreqTest: Don't leak video on failure.
Change-Id: I250379f0ac8d4929c9032e7343290e2980fc2e77
2016-07-21 11:40:51 -07:00
Alex Converse
92e91bd3a1 Make test encoder test driver less likely to leak on failure.
Individual tests still need to be updated.

Change-Id: Ic433d0f742e13560b136f136b72b2a9973970d78
2016-07-21 11:39:47 -07:00
James Zern
16e069b8bb Merge changes from topic 'clang-tidy'
* changes:
  vp8/onyx_if.c: rework #if's to avoid dangling else's
  vp8/bitstream.c: rework #if to avoid dangling else
2016-07-21 03:05:20 +00:00
Johann
a6cc74b987 Merge remote-tracking branch 'origin/khakicampbell' 2016-07-20 18:49:04 -07:00
jackychen
71f9cbcfc8 vp9: Fix the clang warning of unsigned int type.
Change-Id: I6308db16bd626fa5943925471e9171f567669350
2016-07-20 15:58:35 -07:00
James Zern
b19f8b1607 vp8/onyx_if.c: rework #if's to avoid dangling else's
Change-Id: Ieda8958a3da1000424fcff91a1315d0049612202
2016-07-20 12:36:09 -07:00
James Zern
77a31eb3c5 vp8/bitstream.c: rework #if to avoid dangling else
Change-Id: I9178ae75876f3df3fa3271314db39830552b9549
2016-07-20 12:36:09 -07:00
James Zern
7bb35db872 Merge changes from topic 'clang-tidy'
* changes:
  vp8/{bitstream,rdopt},y4minput: correct break placement
  y4minput.c: correct empty loop formatting
  vp8: simplify a few #if's
  vp8: remove extra semicolons
2016-07-20 19:34:59 +00:00
Yaowu Xu
690fcd793b Change to call vp9_post_proc_frame()
This commit changes the call in vp9 encoder from vp9_deblock() to
vp9_post_proc_frame() to ensure the data structures used in the call
are properly allocated. This fixes an encoder crash when configured
with --enable-internal-stats.

Change-Id: I2393b336c0f566665336df4f1ba91c405eb56764
2016-07-20 11:01:49 -07:00
James Zern
fd85664ae6 vp8/{bitstream,rdopt},y4minput: correct break placement
these should be placed within {}s when present

Change-Id: If00e9766fa8cb039cc070467f353a468f99460fb
2016-07-19 20:51:25 -07:00
James Zern
1b048e966a y4minput.c: correct empty loop formatting
prefer {}s over ';'

Change-Id: I563fc82717e1deb4f42a40e03dca318c6adaa0c1
2016-07-19 20:46:39 -07:00
James Zern
b5164f55a0 vp8: simplify a few #if's
bitstream.c: asserts are disabled when CONFIG_DEBUG is unset
vp8_dx_iface.c: split |s into 2 statements across #if bounds

Change-Id: I307d1e969134db5c9c0edd7690589b6b29116cbd
2016-07-19 20:45:28 -07:00
James Zern
96797e43b4 vp8: remove extra semicolons
Change-Id: I84e1a293ee033865f82c244e8aaaadfb2fb27e63
2016-07-19 20:44:14 -07:00
clang-format
033dab9ca0 top-level: apply clang-format
Change-Id: Ibd5395bf8956a80f7c0df4d539c7a42c927a1fc7
2016-07-19 14:34:19 -07:00
James Zern
6e336f6e5f Merge "vp8: apply clang-tidy google-readability-braces-around-statements" 2016-07-19 21:24:30 +00:00
clang-tidy
7f3e07f1c8 vp8: apply clang-tidy google-readability-braces-around-statements
applied against an x86_64 configure

clang-tidy-3.7.1 \
  -checks='-*,google-readability-braces-around-statements' \
  -header-filter='.*' -fix
+ clang-format afterward

Change-Id: I6694edeaee89b58b8b3082187e6756561136b459
2016-07-19 12:38:03 -07:00
James Zern
5b55bcc564 Merge "examples: apply clang-format" 2016-07-19 19:21:08 +00:00
James Zern
e3f7991f99 Merge changes Ia6004c08,I1954f9d6
* changes:
  cosmetics: Add a few explanatory comments
  cosmetics: Correct grammar/spelling in comments
2016-07-19 19:12:23 +00:00
Yury Gitman
e4ac882007 cosmetics: Add a few explanatory comments
Change-Id: Ia6004c08e6f5fd269a1bbd4df51ce9b76345150d
2016-07-19 10:39:00 -07:00
Marco Paniconi
e90d4f0a03 Merge "vp9: Allow usage of lookahead for real-time, 1 pass vbr." 2016-07-19 17:17:16 +00:00
Johann Koenig
451211cb01 Merge "Change 'git cl upload' default to --no-squash" 2016-07-19 16:37:42 +00:00
James Zern
ea3d324f13 Merge changes I18982dbf,I15c8976c
* changes:
  build/make/Makefile: add a 'test_*' default target
  build/make/Makefile: remove default suffix rules
2016-07-19 06:09:36 +00:00
Pascal Massimino
7e4740156f Merge "take II: variance_test partial clean-up" 2016-07-19 03:52:55 +00:00
clang-format
ef45540927 examples: apply clang-format
Change-Id: Icc3bbb07c99a31a70030baec7e51b881902a7b5e
2016-07-18 19:04:56 -07:00
James Bankoski
c69cc4ce1f Merge "configure: turn on all unused warnings by default" 2016-07-19 00:57:46 +00:00
James Zern
25085a6ac2 build/make/Makefile: add a 'test_*' default target
allows 'make test_libvpx', etc. some reworking of the makefiles would be
needed to avoid hard coding targets here.

Change-Id: I18982dbf691e7d36ab8bcf5934bab9340687b061
2016-07-18 16:30:58 -07:00
James Zern
23d0f73838 build/make/Makefile: remove default suffix rules
Change-Id: I15c8976c6478bf75ec617398f49461b310ab7569
2016-07-18 16:30:40 -07:00
skal
7d72ebaa5c take II: variance_test partial clean-up
remove some (but not all yet!) tuple mis-use, and revamp the code a lot.
Factorize some common chores into MainTestClass.

Change-Id: Id37b7330eebe80d19b9d12a454f24ff9be6b1116
2016-07-18 16:18:26 -07:00
Marco
05fe0f20a6 vp9: Allow usage of lookahead for real-time, 1 pass vbr.
Allow usage of lookahead for VBR in real-time mode, for 1 pass vbr.

Current usage is for fast checking of future scene cuts/changes,
and adjusting rate control (gf interval and active_worst/target size).

Added unittests (datarate) for 1 pass vbr mode, with non-zero lag.

Added an experimental option to limit QP based on lookahead.

Overall positive gain in metrics on ytlive set:
avgPNSR/SSIM up on average ~1-3%; several clips up by 5, 7%.

Change-Id: I960d57dfc89de121c4824b9a9bf88d2814e74b56
2016-07-18 15:20:17 -07:00
Johann
1afbd88e81 Change 'git cl upload' default to --no-squash
Chromium changed the upstream default to --squash but this conflicts
with libvpx historical defaults.

Change-Id: I80f2f2b48e2ba08e02184b50e6d5f8f5e76fec24
2016-07-18 14:15:24 -07:00
Yury Gitman
bdfdd7d993 cosmetics: Correct grammar/spelling in comments
Change-Id: I1954f9d6e33abff9081fe7a5cf59d5497768e0df
2016-07-18 12:49:00 -07:00
Jim Bankoski
3e04114f3d prepend ++ instead of post in for loops.
Applied the following regex  :
search for: (for.*\(.*;.*;) ([a-zA-Z_]*)\+\+\)
replace with: \1 ++\2)

This misses some for loops:
ie : for (mb_col = 0; mb_col < oci->mb_cols; mb_col++, mi++)

Change-Id: Icf5f6fb93cced0992e0bb71d2241780f7fb1f0a8
2016-07-18 06:54:50 -07:00
James Zern
106a8a1536 Merge "Revert "variance_test partial clean-up"" 2016-07-16 22:01:19 +00:00
James Zern
3d791194f8 vpx_plane_add_noise_c: normalize int types
quiets signed/unsigned mismatch warning

Change-Id: Iaabd7dfff110ba26056258457541f5635d2e85e6
2016-07-16 11:56:55 -07:00
James Zern
090aa88b5a Revert "variance_test partial clean-up"
This reverts commit f993ed5c86.

build warnings under msvc, segfaults with high-bitdepth enabled.

Change-Id: I67502651107830bcadb6ef56d8f2709cccbfdf2b
2016-07-16 11:55:07 -07:00
Pascal Massimino
f44db1487d Merge "sad_test: add some const to methods" 2016-07-16 04:50:45 +00:00
Pascal Massimino
8d681b36c7 Merge "vp9_error_block_test: simplify fn wrapper generation" 2016-07-16 04:41:46 +00:00
Pascal Massimino
5319e83843 sad_test: add some const to methods
Change-Id: I6f2481509b0aa94338ed6185f80c4a6b65532280
2016-07-15 21:07:00 -07:00
Pascal Massimino
22c36dd464 Merge "remove tuple from 'sad_test.cc'" 2016-07-16 03:56:10 +00:00
James Zern
037a50ed36 vp9_error_block_test: simplify fn wrapper generation
Change-Id: I1f1d396b9456e52e863c4c75f23c3d17420668b4
2016-07-15 20:27:45 -07:00
skal
3fc29ae3ee remove tuple from 'sad_test.cc'
+ general clean-up

Change-Id: Ib9dca3d1a3b7f0c1bedef2a26c9ff5ae1c289e8a
2016-07-15 19:52:56 -07:00
clang-format
81a6739533 vp8: apply clang-format
Change-Id: I7605b6678014a5426ceb45c27b54885e0c4e06ed
2016-07-15 19:28:44 -07:00
James Zern
65daa41378 add .clang-format, based on Google style
derived from clang-format 3.7.1; same as used in libaom

Change-Id: I8ea915a41d1f2ea3b0d4e4dab9ebc808e9116f11
2016-07-15 19:26:24 -07:00
James Bankoski
ce6678fdc9 Merge "addnoise : clear out static size for generated noise" 2016-07-16 01:48:07 +00:00
Jim Bankoski
cb957c302a addnoise : clear out static size for generated noise
Change-Id: I5d4343f2da9cd4b01dd37be7a048d159fec109d1
2016-07-15 15:52:45 -07:00
Jim Bankoski
0dfede2e79 configure: turn on all unused warnings by default
Change-Id: I7f6cb446cd3ac57ac39835cf065d9501a66acd5b
2016-07-15 15:26:20 -07:00
Jim Bankoski
da1bda0fb2 vp9_postproc.c : unused variable if not vp9_highbitdepth.
Change-Id: Ib89b128f23767934c40b5add3fcf9dbd875e82f9
2016-07-15 15:04:57 -07:00
James Bankoski
302e425453 Merge "postproc : fix function parameters for noise functions." 2016-07-15 17:33:53 +00:00
Jim Bankoski
0dc69c70f7 postproc : fix function parameters for noise functions.
Change-Id: I582b6307f28bfc987dcf8910379a52c6f679173c
2016-07-15 08:27:34 -07:00
James Zern
60ada7edb4 Merge "variance_test partial clean-up" 2016-07-15 07:31:30 +00:00
skal
f993ed5c86 variance_test partial clean-up
remove some (but not all yet!) tuple mis-use, and revamp the code a lot.
Factorize some common chores into MainTestClass.

Change-Id: Ia14f3924140e8545e4f10d0504475681baae8336
2016-07-14 22:06:44 +00:00
James Zern
6d2b79e3a2 Merge "vp9_intrapred_test: follow-up cleanup" 2016-07-14 19:48:52 +00:00
James Zern
a07bb84215 gtest-all.cc: quiet an unused variable warning
under windows / mingw builds

Change-Id: I93f9a5df77cea0c28d4afb272abcde5a9732e355
2016-07-13 18:36:17 -07:00
skal
3386ca7496 vp9_intrapred_test: follow-up cleanup
address few comments from ce050afaf3

Change-Id: I5d8fc9dab35c4ee5ec3671134c4eef4ec241e309
2016-07-14 00:40:55 +00:00
James Bankoski
7eec1f31b5 Merge "postproc: noise style fixes." 2016-07-13 22:04:47 +00:00
Hui Su
8dd3bef7ef Merge "Revert "Eliminate isolated and small tail coefficients:"" 2016-07-13 21:30:12 +00:00
Pascal Massimino
8c7751e1c2 Merge "clean-up vp9_intrapred_test" 2016-07-13 21:26:15 +00:00
Yaowu Xu
d6197b621d Merge "Fix encoder crashes for odd size input" 2016-07-13 20:05:09 +00:00
Jim Bankoski
e736691a6d postproc: noise style fixes.
Change-Id: Ifdcb36b8e77b65faeeb10644256e175acb32275d
2016-07-13 12:39:01 -07:00
skal
ce050afaf3 clean-up vp9_intrapred_test
remove tuple and overkill VP9IntraPredBase class.

Change-Id: I85b85bdd33d7fe417895e75f77db219f713dfea3
2016-07-13 18:16:32 +00:00
hui su
248f6ad771 Revert "Eliminate isolated and small tail coefficients:"
This reverts commit ff19cdafdb.

Change-Id: I81f68870ca27a1ff683ee22090530b6997815fb2
2016-07-13 11:14:44 -07:00
Jingning Han
fed14a3e94 Merge "Disable trellis optimization when lossless is on" 2016-07-13 16:01:01 +00:00
James Bankoski
e93f2fdb83 Merge "postproc - move filling of noise buffer to vpx_dsp." 2016-07-13 15:31:17 +00:00
Jim Bankoski
2ca24b0075 postproc - move filling of noise buffer to vpx_dsp.
Change-Id: I63ba35dc0ae9286c9812367a531e01d79a4c1635
2016-07-13 07:35:25 -07:00
Jim Bankoski
b24373fec2 deblock: missing const on extern const.
Change-Id: I0df08f7c431daf939e266f008bf5158b0c97358b
2016-07-13 07:27:29 -07:00
Jim Bankoski
6f424a768e vp9_postproc.c missing extern.
BUG=webm:1256

Change-Id: I5271e71bc53cce033fb906040643dcdd5ccb2381
2016-07-12 17:47:49 -07:00
James Zern
6a3ff0b617 Merge changes from topic 'webp-thread-update'
* changes:
  vpx_thread: use CreateThread for windows phone
  vpx_thread: use WaitForSingleObjectEx if available
  vpx_thread: use InitializeCriticalSectionEx if available
  vpx_thread: use native windows cond var if available
  vpx_thread.[hc]: update webp source reference
2016-07-13 00:08:05 +00:00
Yaowu Xu
98431cde07 Fix encoder crashes for odd size input
Change-Id: Id5c30c419282369cc8c3280d9a70b34a859a71d8
2016-07-12 11:11:26 -07:00
Jacky Chen
19c157afe2 Merge "vp9 svc: Reuse scaled_temp in two stage downscaling." 2016-07-12 17:59:09 +00:00
JackyChen
110a2ddc9b vp9 svc: Reuse scaled_temp in two stage downscaling.
This change eliminates redundant computation in the two stage
downscaling, which saves ~1% encoding time in 3-layer svc encoding.

Change-Id: Ib4b218811b68499a740af1f9b7b5a5445e28d671
2016-07-12 10:09:55 -07:00
Jingning Han
efccbc9fb5 Disable trellis optimization when lossless is on
Disable trellis coefficient optimization when the lossless mode
is turned on.

Change-Id: I9001bf626e86dc3c8c32331ede04fd39036e5f7c
2016-07-12 09:00:16 -07:00
Jim Bankoski
88e6951465 deblock filter : moved from vp8 code branch
The deblocking filters used in vp8 have been moved to vpx_dsp for
use by both vp8 and vp9.

Change-Id: I5209d76edafc894b550f751fc76d3aa6799b392d
2016-07-12 05:53:00 -07:00
James Zern
45ed7effed Merge "remove *debugmodes.c from the default build" 2016-07-11 23:49:29 +00:00
Scott LaVarnway
2e93fcf893 Merge "vp9_rd_pick_intra_mode_sb(): set interp_filter to" 2016-07-11 22:31:06 +00:00
paulwilkins
3a986eac57 Sample points to reduce encode overhead.
Only noise filter sampled points in first pass to reduce
any first pass speed overhead.

Change-Id: Ic80d4400e59146d1c3332336c4350faf28ff8b17
2016-07-11 11:45:52 +01:00
Scott LaVarnway
ed7786869a vp9_rd_pick_intra_mode_sb(): set interp_filter to
SWITCHABLE_FILTERS.  This is a partial fix for the build
issues with Change 357240.

Change-Id: I4e507c196175bae729a4f1397878ec8776b0146c
2016-07-09 09:47:34 -07:00
Yaowu Xu
5adb43b8be Fix non-highbitdepth coding path for HBD build
Change-Id: I38eb42b8d051924a7cd1ccc3421a4057cf6e170f
2016-07-08 11:26:34 -07:00
Marco Paniconi
20946cdd3b Merge "vp9: Adjustment of gfu_boost and af_ratio for 1 pass vbr." 2016-07-08 16:26:06 +00:00
Yaowu Xu
dc008cc17d Merge "Enable HBD support in real time encoding path" 2016-07-07 22:32:48 +00:00
Marco
cc431ad50a vp9: Adjustment of gfu_boost and af_ratio for 1 pass vbr.
Modify the gfu_boost and af_ratio setting based on the
average frame motion level.

Change only affects 1 pass vbr.

Metrics overall positive on ytlive set.
On average up by ~1%, several clips up by 2-4%.

Change-Id: Ic18c49eb2df74cb4986b63cdb11be36d86ab5e8d
2016-07-07 15:18:14 -07:00
Marco Paniconi
a75965fa94 Merge "vp9: Adjustment to mv bias for non-rd pickmode." 2016-07-07 21:07:37 +00:00
Jingning Han
2f28f9072e Enable coeff optimization for intra modes
This further improves the coding performance by
lowres 0.3%
midres 0.5%
hdres  0.6%

Change-Id: I6a03b6da210b9cbc261474bad4a103e0ba021c68
2016-07-07 12:25:41 -07:00
Jingning Han
44354ee7bf Use precise context to estimate coeff rate cost
Use the precise context to estimate the zero token cost in trellis
optimization process. This improves the speed 0 coding performance
by 0.15% for lowres and 0.1% for midres. It improves the speed 1
coding performance by 0.2% for midres and hdres.

Change-Id: I59c7c08702fc79dc4f8534b64ca594da909e2c91
2016-07-07 12:25:33 -07:00
Jingning Han
62aa642d71 Enable uniform quantization with trellis optimization in speed 0
This commit allows the inter prediction residual to use uniform
quantization followed by trellis coefficient optimization in
speed 0. It improves the coding performance by

lowres 0.79%
midres 1.07%
hdres  1.44%

Change-Id: I46ef8cfe042a4ccc7a0055515012cd6cbf5c9619
2016-07-07 12:25:33 -07:00
Jingning Han
541eb78994 Refactor coeff_cost() function
Move the operations that update the context buffers outside this
function. The coeff_cost() takes all input as const value and returns
the coefficient cost.

This makes preparation for the next coefficient optimization CLs.

Change-Id: I850eec6e5470b91ea84646ff26b9231b09f70a0c
2016-07-07 18:09:39 +00:00
Jingning Han
7c1fdf02cd Merge "Support measure distortion in the pixel domain" 2016-07-07 18:09:20 +00:00
Marco
f451b404ea vp9: Adjustment to mv bias for non-rd pickmode.
Replace the existing mv bias with a bias only for
NEWMV, and based on the motion vector difference of
its top/left neighbors.

For cbr non-screen-content mode.

Change-Id: I8a8cf56347cfa23e9ffd8ead69eec8746c8f9e09
2016-07-07 10:33:06 -07:00
Yunqing Wang
9976ff8c79 Merge "Fix Visual Studio build warning" 2016-07-07 16:48:49 +00:00
Yunqing Wang
9ef37860cd Fix Visual Studio build warning
Fixed signed/unsigned mismatch warning.

Change-Id: I1634d0634de752f4b8baa8059e8f3e2891fa53b6
2016-07-07 08:43:57 -07:00
paulwilkins
2580e7d63e Noise energy Experiment in first pass.
Use a measure of noise energy to adjust Q estimate and
arf filter strength.

Gains 0.3-0.5% on Lowres and |Netflix sets.
Hdres and Midres neutral.

Change-Id: Ic0de552e7b6763e70eeeaa3651619831b423e151
2016-07-07 14:50:21 +01:00
Paul Wilkins
f037cf80c9 Merge "Add experimental spatial de-noise filter on key frames." 2016-07-07 13:30:07 +00:00
Jingning Han
e357b9efe0 Support measure distortion in the pixel domain
Use pixel domain distortion metric in speed 0. This improves the
compression performance by 0.3% for both low and high resolution
test sets.

Change-Id: I5b5b7115960de73f0b5e5d0c69db305e490e6f1d
2016-07-06 18:25:17 -07:00
Yaowu Xu
884c2ddc48 Enable HBD support in real time encoding path
BUG=webm:1223

Change-Id: If83a613784e3b2a33c9c93f9ad0ba39dd4d23056
2016-07-06 14:18:37 -07:00
Debargha Mukherjee
adbad6092f Merge "Remove decode asserts from better-hw-compatibility" 2016-07-06 20:55:29 +00:00
Jacky Chen
aa6108382e Merge "vp9: Choose the scheme for modeling rd for 32x32 based on skin color." 2016-07-06 20:03:55 +00:00
Debargha Mukherjee
4b6e4e1813 Remove decode asserts from better-hw-compatibility
Safer to have the decoder operate normally and have
better-hw-compatibility only implement encoding changes.
Fixes some test failures.

Change-Id: I0dd70d002e4e893992f0cd59774b9363e6f7fe76
2016-07-06 12:26:38 -07:00
Yunqing Wang
a921444fdb Merge "Modify the name of vp9cx_set_ref example" 2016-07-06 18:55:15 +00:00
JackyChen
2678aefc48 vp9: Choose the scheme for modeling rd for 32x32 based on skin color.
For real time CBR mode, use model_rd_for_sb_y for 32x32 if the sb is
a skin sb to avoid visual regression on the slowly moving face.

Refer to the cl: https://chromium-review.googlesource.com/#/c/356020/

Change-Id: I42c36666b2b474ce5ee274239d52ae8ab400fd46
2016-07-06 11:12:03 -07:00
Min Ye
ff19cdafdb Eliminate isolated and small tail coefficients:
Improve hdres PSNR by 0.696%
Improve midres PSNR by 0.313%
Improve lowres PSNR by 0.142%

Change-Id: Icabde78aa9689f539f6a03ec09f712c20758796c
2016-07-06 11:08:23 -07:00
Yunqing Wang
fba5c354ad Modify the name of vp9cx_set_ref example
Modified the name of vp9cx_set_ref example so that the test script
ran correctly.

Change-Id: I0ab2de66220b0a88b7af7ea1633a088ab78dd9ff
2016-07-06 10:05:51 -07:00
Jingning Han
51aad61c8c Merge "Remove txfrm_block_to_raster_xy() from vp9 encoder" 2016-07-06 16:00:18 +00:00
Yunqing Wang
825bb86044 Merge "Make set_reference control API work in VP9" 2016-07-06 00:08:52 +00:00
Jingning Han
14011f037d Remove txfrm_block_to_raster_xy() from vp9 encoder
The transform block row and column positions are always available
outside the callees. There is no need to re-compute these values
again. This approach has been used by the decoder. This commit
removes txfrm_block_to_raster_xy() function.

Change-Id: I5b90f91a0d8b7c35cfa7d171da9edf8202630108
2016-07-04 18:41:47 -07:00
James Zern
5afa3b9150 Merge "improve vpx_filter_block1d* based on replace paddsw+psrlw to pmulhrsw" 2016-07-02 03:08:33 +00:00
James Zern
3197172405 Merge "Update vpx subpixel 1d filter ssse3 asm" 2016-07-02 03:08:17 +00:00
James Zern
3007081a87 vpx_thread: use CreateThread for windows phone
BUG=b/29583578

original webp change:

commit d2afe974f9d751de144ef09d31255aea13b442c0
Author: James Zern <jzern@google.com>
Date:   Mon Nov 23 20:41:26 2015 -0800

    thread: use CreateThread for windows phone

    _beginthreadex is unavailable for winrt/uwp

    Change-Id: Ie7412a568278ac67f0047f1764e2521193d74d4d

100644 blob 93f7622797f05f6acc1126e8296c481d276e4047 src/utils/thread.c
100644 blob 840831185502d42a3246e4b7ff870121c8064791 src/utils/thread.h

Change-Id: Iade8fff6367b45534986c77ebe61abeb45bce0f8
2016-07-01 19:36:58 -07:00
James Zern
7954e67bb8 vpx_thread: use WaitForSingleObjectEx if available
BUG=b/29583578

original webp change:

commit 0fd0e12bfe83f16ce4f1c038b251ccbc13c62ac2
Author: James Zern <jzern@google.com>
Date:   Mon Nov 23 20:40:26 2015 -0800

    thread: use WaitForSingleObjectEx if available

    Windows XP and up

    Change-Id: Ie1a46a82722b8624437c8aba0aa4566a4b0b3f57

100644 blob d58f74e5523dbc985fc531cf5f0833f1e9157cf0 src/utils/thread.c
100644 blob 840831185502d42a3246e4b7ff870121c8064791 src/utils/thread.h

Change-Id: If165c38b378c6e0c55e17a1b071efd3ec3e7dcdd
2016-07-01 19:36:58 -07:00
James Zern
a48d42a804 vpx_thread: use InitializeCriticalSectionEx if available
BUG=b/29583578

original webp change:

commit 63fadc9ffacc77d4617526a50c696d21d558a70b
Author: James Zern <jzern@google.com>
Date:   Mon Nov 23 20:38:46 2015 -0800

    thread: use InitializeCriticalSectionEx if available

    Windows Vista / Server 2008 and up

    Change-Id: I32c5b4e5384d614c5a821ef511293ff014c67966

100644 blob f84207d89b3a6bb98bfe8f3fa55cad72dfd061ff src/utils/thread.c
100644 blob 840831185502d42a3246e4b7ff870121c8064791 src/utils/thread.h

Change-Id: I9ce49b3a86857267e504cd8ceab503b7b441d614
2016-07-01 19:36:58 -07:00
James Zern
8cc525e82b vpx_thread: use native windows cond var if available
BUG=b/29583578

original webp change:

commit 110ad5835ecd66995d0e7f66dca1b90dea595f5a
Author: James Zern <jzern@google.com>
Date:   Mon Nov 23 19:49:58 2015 -0800

    thread: use native windows cond var if available

    Vista / Server 2008 and up. no speed difference observed.

    Change-Id: Ice19704777cb679b290dc107a751a0f36dd0c0a9

100644 blob 4fc372b7bc6980a9ed3618c8cce5b67ed7b0f412 src/utils/thread.c
100644 blob 840831185502d42a3246e4b7ff870121c8064791 src/utils/thread.h

Change-Id: Iede7ae8a7184e4b17a4050b33956918fc84e15b5
2016-07-01 19:36:58 -07:00
James Zern
d4d6c58e37 vpx_thread.[hc]: update webp source reference
+ drop the blob hash, the updated reference will be updated in the
commit message

BUG=b/29583578

Change-Id: Ifabbe52a2f07ac29e1881f5c8a62d7f3eb3c2c04
2016-07-01 19:36:58 -07:00
James Zern
d2b40894be remove *debugmodes.c from the default build
these are debug-only modules that can be added in manually when needed.
leave a reference in vp8_common.mk / vp9_common.mk for easy addition.

quiets -Wmissing-prototypes warning

BUG=b/29584271

Change-Id: Ifc8637d877edfbd562b34dc5c540428bba7951fc
2016-07-01 19:10:04 -07:00
Yunqing Wang
0a075cb39c Make set_reference control API work in VP9
Moved the API patch from NextGenv2. An example was included.
To try it, for example, run the following command:
$ examples/vpx_cx_set_ref vp9 352 288 in.yuv out.ivf 4 30

Change-Id: I4cf8f23b86d7ebd85ffd2630dcfbd799c0b88101
2016-07-01 17:58:02 -07:00
James Zern
3ef9c0ba03 vp8/common/reconintra4x4.c: add missing include
quiets -Wmissing-prototypes warning

BUG=b/29584271

Change-Id: I806e3475ebee579dce0073dd1784a7c2899e7de0
2016-07-01 16:20:42 -07:00
James Bankoski
f5a15f270a Merge "Revert "libyuv: update to 2f101fdb"" 2016-07-01 19:14:53 +00:00
James Bankoski
c5372cf077 Revert "libyuv: update to 2f101fdb"
Compile failures on linux platform.  

BUG=webm:1253

This reverts commit aa81375d73.

Change-Id: Ibab2c4827bc21518dc03c6e9716b5015cff56fc7
2016-07-01 19:14:28 +00:00
Johann Koenig
e616012d69 Merge changes I59a11921,I296a0b81,I397d7753
* changes:
  configure: remove x86inc.asm distinction
  test: remove x86inc.asm distinction
  vpx_dsp: remove x86inc.asm distinction
2016-07-01 18:13:41 +00:00
James Zern
fbbd3f0d8d Merge "convolve_test: fix byte offsets in hbd build" 2016-07-01 01:54:30 +00:00
Jacky Chen
ee78c541a4 Merge "vp9 postproc: Bug fix and code clean." 2016-06-30 21:59:44 +00:00
James Bankoski
892ebd9760 Merge "libyuv: update to 2f101fdb" 2016-06-30 19:11:30 +00:00
Johann
571f00cb95 configure: remove x86inc.asm distinction
BUG=b:29583530

Change-Id: I59a1192142e89a6a36b906f65a491a734e603617
2016-06-30 11:14:14 -07:00
Johann
0266e70c52 test: remove x86inc.asm distinction
BUG=b:29583530

Change-Id: I296a0b81755e3086bc0a40cb126d0200ff03c095
2016-06-30 11:14:10 -07:00
Johann Koenig
3c41d7358c Merge "vp9: remove x86inc.asm distinction" 2016-06-30 17:42:17 +00:00
Johann Koenig
89771f2c2c Merge "Require x86inc.asm" 2016-06-30 17:41:33 +00:00
Paul Wilkins
1d3f1983b2 Merge "Fix error in get_ul_intra_threshold() for 10/12 bit." 2016-06-30 16:26:14 +00:00
Paul Wilkins
f7c2d2a3de Merge "Fix error in get_smooth_intra_threshold() for 10/12 bit." 2016-06-30 16:25:55 +00:00
Jim Bankoski
aa81375d73 libyuv: update to 2f101fdb
Fixes color issue when scaling without breaking mingw.

BUG=https://bugs.chromium.org/p/libyuv/issues/detail?id=605
BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1252

Change-Id: Ifba747feb0c6a08f2b353b820a24c6c145d440ad
2016-06-30 13:25:39 +00:00
paulwilkins
e25d6252a4 Fix error in get_ul_intra_threshold() for 10/12 bit.
The scaling of the threshold for 10 and 12 bit here appears
to be in the wrong direction. For 10 and 12 bit we expect sse
values to be higher and hence the threshold used should be
scaled up not down.

Change-Id: I2678116652b539aef48100e0f22873edd4f5a786
2016-06-30 13:38:57 +01:00
paulwilkins
f9a3d08f1b Fix error in get_smooth_intra_threshold() for 10/12 bit.
This function seems to scale the threshold for testing an
SSE value in the wrong direction for 10 and 12 bit inputs.

Also for a true SSE the scalings should probably be << 4 and 8

Change-Id: Iba8047b3f70d04aa46d9688a824f3d49c1c58e90
2016-06-30 13:34:11 +01:00
Jacky Chen
e85607410e Merge "vp9: Change the scheme for modeling rd for 32x32 on newmv_last mode." 2016-06-30 05:59:46 +00:00
James Zern
f5a6079141 convolve_test: fix byte offsets in hbd build
CONVERT_TO_BYTEPTR(x) was corrected in:
003a9d2 Port metric computation changes from nextgenv2
to use the more common (x) within the expansion. offsets should occur
after converting the pointer to the desired type.

+ factorized some common expressions

Change-Id: I171c3faaa5606d098e984baa9aa74bb36042f57f
2016-06-29 20:39:07 -07:00
Johann
1b833d63d9 vpx_dsp: remove x86inc.asm distinction
BUG=b:29583530

Change-Id: I397d77536b0d3cee0a92cdfe8b76bc4e434d0720
2016-06-29 18:55:58 -07:00
Johann
fe96dbda15 vp9: remove x86inc.asm distinction
BUG=b:29583530

Change-Id: I952da3fc0d4716dec897be0d2e9806af6612722b
2016-06-29 18:55:28 -07:00
Johann
d11c97e8e2 Require x86inc.asm
Force enable x86inc.asm when building for x86. Previously there were
compatibility issues so a flag was added to simplify disabling this
code.

The known issues have been resolved and x86inc.asm is the preferred
abstraction layer (over x86_abi_support.asm).

BUG=b:29583530

Change-Id: Ib935e97b37ffb22d7af72ba0f04564ae6280f1fd
2016-06-29 18:55:12 -07:00
James Zern
0462263765 configure: restore vs_version variable
inadvertently lost in the final patchset of:
078dff7 configure: remove old visual studio support (<2010)

this prevents an empty CONFIG_VS_VERSION and avoids make failure

Change-Id: I529d52eca59329e2715309efd63d80f0e1fed462
2016-06-29 16:57:28 -07:00
JackyChen
5fc2d6cb9f vp9: Change the scheme for modeling rd for 32x32 on newmv_last mode.
For real time CBR mode, use model_rd_for_sb_y for 32x32 if the mode is
newmv last, which is less aggressive in skipping transform and
quantization, to avoid quality regression in some conditions.

Change-Id: Ifa30be587f2a8a4a7f182a172de6ce277c0f8556
2016-06-29 16:28:15 -07:00
James Bankoski
c8f6ed77b9 Merge "Revert "libyuv: update to b8ddb5a2"" 2016-06-29 23:20:25 +00:00
James Bankoski
291033032e Revert "libyuv: update to b8ddb5a2"
This reverts commit b8f83282f8.

Update was to wrong version and still has: 

BUG=webm:1252

Change-Id: I80f3a7c0581ab5e2dd1a84f7840e51d7c362afac
2016-06-29 23:09:10 +00:00
James Zern
3a6a81fc9a Merge changes I9433d858,Iafd05637,If08ce6ca
* changes:
  tests: remove redundant round() definition
  remove visual studio < 2010 workarounds
  configure: remove old visual studio support (<2010)
2016-06-29 23:07:16 +00:00
Yaowu Xu
b458f42966 Merge "Remove effectless initialization" 2016-06-29 22:51:14 +00:00
James Zern
0a64929f19 tests: remove redundant round() definition
use vpx_ports/msvc.h for compatibility

BUG=b/29583530

Change-Id: I9433d8586cd0b790e7f4d697304298feafe801f1
2016-06-29 14:57:47 -07:00
Yaowu Xu
c02a4beed8 Merge "Prevent negative variance" 2016-06-29 20:53:37 +00:00
Linfeng Zhang
6b350766bd Update vpx subpixel 1d filter ssse3 asm
Speed test shows the new vertical filters have degradation on Celeron
Chromebook. Added "X86_SUBPIX_VFILTER_PREFER_SLOW_CELERON" to control
the vertical filters activated code. Now just simply active the code
without degradation on Celeron. Later there should be 2 set of vertical
filters ssse3 functions, and let jump table to choose based on CPU type.

Change-Id: Iba2f1f2fe059a9d142c396d03a6b8d2d3b981e87
2016-06-29 13:48:41 -07:00
Yaowu Xu
63a37d16f3 Prevent negative variance
Due to rounding, hbd variance may become negative. This commit put in
check and clamp of negative values to 0.

Change-Id: I610d9c8aa2d4eebe7bc5f2c5624a9e3cadad4c94
2016-06-29 11:08:17 -07:00
James Bankoski
f14c323b4c Merge "libyuv: update to b8ddb5a2" 2016-06-29 17:58:40 +00:00
Jim Bankoski
b8f83282f8 libyuv: update to b8ddb5a2
Fixes color issue when scaling without breaking mingw.

BUG=https://bugs.chromium.org/p/libyuv/issues/detail?id=605
BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1252

Change-Id: I3920c5664def7ae7a23f60fb160d26d23bc86a27
2016-06-29 17:53:14 +00:00
paulwilkins
be013eb396 Add experimental spatial de-noise filter on key frames.
For forced key frames in particular this helps to make them
blend better with the surrounding frames where noise tends
to be suppressed by a combination of quantization and alt
ref filtering.

Currently disabled by default under and IFDEF flag pending
wider testing.

Change-Id: I971b5cc2b2a4b9e1f11fe06c67ef073f01b25056
2016-06-29 17:25:41 +01:00
Scott LaVarnway
74bb78df82 Merge "VP9: handle_inter_mode()... Use interp_filter" 2016-06-29 11:41:52 +00:00
James Zern
c125f4a594 remove visual studio < 2010 workarounds
BUG=b/29583530

Change-Id: Iafd05637eb65f4da54a9c857e79204a77646858a
2016-06-28 20:58:49 -07:00
James Zern
078dff72ca configure: remove old visual studio support (<2010)
BUG=b/29583530

Change-Id: If08ce6ca352f377ac4db6b9b1909b507bba6d872
2016-06-28 20:40:22 -07:00
jackychen
6b4463dc1f vp9 postproc: Bug fix and code clean.
Bug fix: The crash is caused by not allocating buffer for prev_mip in
postproc_state and prev_mip in postproc_state is only used for MFQE,
ohter postproc modules, deblocking and etc., should not use it.

BUG=webm:1251

Change-Id: I3120d2f50603b4a2d400e92d583960a513953a28
2016-06-28 16:13:44 -07:00
Scott LaVarnway
feb7e9a372 VP9: handle_inter_mode()... Use interp_filter
only if above/left is inter.

Change-Id: I0cc1f926425c021c84536df8271e9ee5f3f87caf
2016-06-28 14:09:59 -07:00
Jacky Chen
d004c64013 Merge "vp9: Increase thr_var for 32x32 blocks in var-based partitioning." 2016-06-28 20:54:06 +00:00
Jacky Chen
4736e5f9d1 Merge "vp9: Move chroma sensitivity check out from choose_partitioning." 2016-06-28 20:53:23 +00:00
Yaowu Xu
43ae6c1e22 Remove effectless initialization
Change-Id: Iec117841a7ecf6f99d2b718057d8646e221c5c64
2016-06-28 12:28:45 -07:00
James Zern
0afe5e405d Merge "*.asm: normalize label format" 2016-06-28 19:22:10 +00:00
jackychen
91038e0eb6 vp9: Move chroma sensitivity check out from choose_partitioning.
Change-Id: Ie78185a30cac4d1841be3708bd23e6505d3733b6
2016-06-28 09:58:51 -07:00
Yaowu Xu
b2d690187e Merge "psnr.c: use int64_t for sum of differences" 2016-06-28 16:55:44 +00:00
Yaowu Xu
d34b49d7b9 psnr.c: use int64_t for sum of differences
Since the values can be negative.

Change-Id: Idda69e9fb47bb34696aeb20170341a0191c5d85e
2016-06-28 09:53:11 -07:00
Parag Salasakar
10b4753179 Merge "mips added p6600 cpu support" 2016-06-28 08:45:01 +00:00
James Zern
f51f67602e *.asm: normalize label format
add a trailing ':', though it's optional with the tools we support, it's
more common to use it to mark a label. this also quiets the
orphan-labels warning with nasm/yasm.

BUG=b/29583530

Change-Id: I46e95255e12026dd542d9838e2dd3fbddf7b56e2
2016-06-27 19:46:57 -07:00
James Bankoski
32ac7cabdf Merge "Revert "libyuv: update to 1b3e4aee47"" 2016-06-27 22:59:11 +00:00
James Bankoski
7f2628152a Revert "libyuv: update to 1b3e4aee47"
This reverts commit 0c6caf187c.

BUG=https://bugs.chromium.org/p/webm/issues/detail?id=1252

Fails mingw_64 builds.

Change-Id: I83e7204bf1be48b499dc32b2597693b95ec49d06
2016-06-27 22:29:52 +00:00
jackychen
8cbd4f8701 vp9: Increase thr_var for 32x32 blocks in var-based partitioning.
For real-time mode, increase variance threshold for 32x32 blocks in
var-based partitioning for resolution >= 720p, so that it is more
likely to stay at 32x32 for high resolution which accelerates the
encoding speed with little/no PSNR drop.

PSNR effect on different speed settings:
speed 8 rtc: 0.02 overall PSNR drop, 0.285% SSIM drop
speed 7 rtc: 0.196% overall PSNR increase, 0.066% SSIM increase
speed 5 rtc_derf: no effect.

Speed up:
gips_motion_WHD, 1mbps: 2.5% faster on speed 7, 2.6% faster on speed8
gips_stat_WHD, 1mbps: 4.6% faster on speed 7, 5.6% faster on speed8

Change-Id: Ie7c33c4d2dd7d09294917e031357fc5476c3a4bb
2016-06-27 14:44:27 -07:00
James Bankoski
71aacf39c7 Merge "libyuv: update to 1b3e4aee47" 2016-06-27 19:32:50 +00:00
Yaowu Xu
7676defca9 Merge "Port metric computation changes from nextgenv2" 2016-06-27 19:18:00 +00:00
Min Chen
b2fb48cfcf improve vpx_filter_block1d* based on replace paddsw+psrlw to pmulhrsw
Change-Id: I14c0c2e54d0b0584df88e9a3f0a256ec096bea6e
2016-06-27 17:50:45 +00:00
Parag Salasakar
7c184c6e1f mips added p6600 cpu support
Removed -funroll-loops

Change-Id: I6684bcac62902c10f945a6dcc4ed803203fcd829
2016-06-27 13:02:55 +05:30
Yaowu Xu
b9ec759bc2 Fix ubsan warnings: vp9/encoder/vp9_pickmode.c
This commit fixes a number of integer out of range issue in HBD build.

BUG=webm:1219

Change-Id: Ib4192dc74a500e1b86c37a399114c7f6d4ed5185
2016-06-27 05:53:46 +00:00
James Zern
913081ab02 Merge "s/UINT32_MAX/UINT_MAX/" 2016-06-25 21:09:55 +00:00
James Zern
ca88d22f39 s/UINT32_MAX/UINT_MAX/
provides better toolchain compatibility

Change-Id: I8561a6de668a68ff54fe3886a4ee6300f0ae9c04
2016-06-25 12:15:51 -07:00
Jim Bankoski
0c6caf187c libyuv: update to 1b3e4aee47
Color issue when scaling.   https://codereview.chromium.org/2084533006/

Change-Id: I84d74346f754c02a5b770b87b6e0b6885d03bb20
2016-06-24 21:57:48 +00:00
Yaowu Xu
003a9d20ad Port metric computation changes from nextgenv2
Change-Id: I4aceffcdf7af59ffeb51984f0345c3a4c7e76a9f
2016-06-24 13:52:50 -07:00
841 changed files with 118208 additions and 127810 deletions

91
.clang-format Normal file
View File

@@ -0,0 +1,91 @@
---
Language: Cpp
# BasedOnStyle: Google
# Generated with clang-format 3.8.1
AccessModifierOffset: -1
AlignAfterOpenBracket: Align
AlignConsecutiveAssignments: false
AlignConsecutiveDeclarations: false
AlignEscapedNewlinesLeft: true
AlignOperands: true
AlignTrailingComments: true
AllowAllParametersOfDeclarationOnNextLine: true
AllowShortBlocksOnASingleLine: false
AllowShortCaseLabelsOnASingleLine: true
AllowShortFunctionsOnASingleLine: All
AllowShortIfStatementsOnASingleLine: true
AllowShortLoopsOnASingleLine: true
AlwaysBreakAfterDefinitionReturnType: None
AlwaysBreakAfterReturnType: None
AlwaysBreakBeforeMultilineStrings: true
AlwaysBreakTemplateDeclarations: true
BinPackArguments: true
BinPackParameters: true
BraceWrapping:
AfterClass: false
AfterControlStatement: false
AfterEnum: false
AfterFunction: false
AfterNamespace: false
AfterObjCDeclaration: false
AfterStruct: false
AfterUnion: false
BeforeCatch: false
BeforeElse: false
IndentBraces: false
BreakBeforeBinaryOperators: None
BreakBeforeBraces: Attach
BreakBeforeTernaryOperators: true
BreakConstructorInitializersBeforeComma: false
ColumnLimit: 80
CommentPragmas: '^ IWYU pragma:'
ConstructorInitializerAllOnOneLineOrOnePerLine: false
ConstructorInitializerIndentWidth: 4
ContinuationIndentWidth: 4
Cpp11BracedListStyle: false
DerivePointerAlignment: false
DisableFormat: false
ExperimentalAutoDetectBinPacking: false
ForEachMacros: [ foreach, Q_FOREACH, BOOST_FOREACH ]
IncludeCategories:
- Regex: '^<.*\.h>'
Priority: 1
- Regex: '^<.*'
Priority: 2
- Regex: '.*'
Priority: 3
IndentCaseLabels: true
IndentWidth: 2
IndentWrappedFunctionNames: false
KeepEmptyLinesAtTheStartOfBlocks: false
MacroBlockBegin: ''
MacroBlockEnd: ''
MaxEmptyLinesToKeep: 1
NamespaceIndentation: None
ObjCBlockIndentWidth: 2
ObjCSpaceAfterProperty: false
ObjCSpaceBeforeProtocolList: false
PenaltyBreakBeforeFirstCallParameter: 1
PenaltyBreakComment: 300
PenaltyBreakFirstLessLess: 120
PenaltyBreakString: 1000
PenaltyExcessCharacter: 1000000
PenaltyReturnTypeOnItsOwnLine: 200
PointerAlignment: Right
ReflowComments: true
SortIncludes: false
SpaceAfterCStyleCast: false
SpaceBeforeAssignmentOperators: true
SpaceBeforeParens: ControlStatements
SpaceInEmptyParentheses: false
SpacesBeforeTrailingComments: 2
SpacesInAngles: false
SpacesInContainerLiterals: true
SpacesInCStyleCastParentheses: false
SpacesInParentheses: false
SpacesInSquareBrackets: false
Standard: Auto
TabWidth: 8
UseTab: Never
...

7
.gitignore vendored
View File

@@ -37,9 +37,9 @@
/examples/twopass_encoder
/examples/vp8_multi_resolution_encoder
/examples/vp8cx_set_ref
/examples/vp9cx_set_ref
/examples/vp9_lossless_encoder
/examples/vp9_spatial_scalable_encoder
/examples/vpx_temporal_scalable_patterns
/examples/vp9_spatial_svc_encoder
/examples/vpx_temporal_svc_encoder
/ivfdec
/ivfdec.dox
@@ -50,6 +50,9 @@
/samples.dox
/test_intra_pred_speed
/test_libvpx
/tools.dox
/tools/*.dox
/tools/tiny_ssim
/vp8_api1_migration.dox
/vp[89x]_rtcd.h
/vpx.pc

15
AUTHORS
View File

@@ -7,6 +7,8 @@ Adam Xu <adam@xuyaowu.com>
Adrian Grange <agrange@google.com>
Aex Converse <aconverse@google.com>
Ahmad Sharif <asharif@google.com>
Aleksey Vasenev <margtu-fivt@ya.ru>
Alexander Potapenko <glider@google.com>
Alexander Voronov <avoronov@graphics.cs.msu.ru>
Alexis Ballier <aballier@gentoo.org>
Alok Ahuja <waveletcoeff@gmail.com>
@@ -27,6 +29,7 @@ Christian Duvivier <cduvivier@google.com>
Daniele Castagna <dcastagna@chromium.org>
Daniel Kang <ddkang@google.com>
Deb Mukherjee <debargha@google.com>
Deepa K G <deepa.kg@ittiam.com>
Dim Temp <dimtemp0@gmail.com>
Dmitry Kovalev <dkovalev@google.com>
Dragan Mrdjan <dmrdjan@mips.com>
@@ -37,6 +40,7 @@ Fabio Pedretti <fabio.ped@libero.it>
Frank Galligan <fgalligan@google.com>
Fredrik Söderquist <fs@opera.com>
Fritz Koenig <frkoenig@google.com>
Gabriel Marin <gmx@chromium.org>
Gaute Strokkenes <gaute.strokkenes@broadcom.com>
Geza Lore <gezalore@gmail.com>
Ghislain MARY <ghislainmary2@gmail.com>
@@ -48,6 +52,7 @@ Hangyu Kuang <hkuang@google.com>
Hanno Böck <hanno@hboeck.de>
Henrik Lundin <hlundin@google.com>
Hui Su <huisu@google.com>
Ivan Krasin <krasin@chromium.org>
Ivan Maltz <ivanmaltz@google.com>
Jacek Caban <cjacek@gmail.com>
Jacky Chen <jackychen@google.com>
@@ -61,6 +66,7 @@ Jean-Yves Avenard <jyavenard@mozilla.com>
Jeff Faust <jfaust@google.com>
Jeff Muizelaar <jmuizelaar@mozilla.com>
Jeff Petkau <jpet@chromium.org>
Jerome Jiang <jianj@google.com>
Jia Jia <jia.jia@linaro.org>
Jian Zhou <zhoujian@google.com>
Jim Bankoski <jimbankoski@google.com>
@@ -75,6 +81,7 @@ Joshua Litt <joshualitt@google.com>
Julia Robson <juliamrobson@gmail.com>
Justin Clift <justin@salasaga.org>
Justin Lebar <justin.lebar@gmail.com>
Kaustubh Raste <kaustubh.raste@imgtec.com>
KO Myung-Hun <komh@chollian.net>
Lawrence Velázquez <larryv@macports.org>
Linfeng Zhang <linfengz@google.com>
@@ -91,8 +98,11 @@ Michael Kohler <michaelkohler@live.com>
Mike Frysinger <vapier@chromium.org>
Mike Hommey <mhommey@mozilla.com>
Mikhal Shemer <mikhal@google.com>
Min Chen <chenm003@gmail.com>
Minghai Shang <minghai@google.com>
Min Ye <yeemmi@google.com>
Morton Jonuschat <yabawock@gmail.com>
Nathan E. Egge <negge@mozilla.com>
Nico Weber <thakis@chromium.org>
Parag Salasakar <img.mips1@gmail.com>
Pascal Massimino <pascal.massimino@gmail.com>
@@ -101,16 +111,19 @@ Paul Wilkins <paulwilkins@google.com>
Pavol Rusnak <stick@gk2.sk>
Paweł Hajdan <phajdan@google.com>
Pengchong Jin <pengchong@google.com>
Peter Boström <pbos@google.com>
Peter de Rivaz <peter.derivaz@gmail.com>
Philip Jägenstedt <philipj@opera.com>
Priit Laes <plaes@plaes.org>
Rafael Ávila de Espíndola <rafael.espindola@gmail.com>
Rafaël Carré <funman@videolan.org>
Ralph Giles <giles@xiph.org>
Ranjit Kumar Tulabandu <ranjit.tulabandu@ittiam.com>
Rob Bradford <rob@linux.intel.com>
Ronald S. Bultje <rsbultje@gmail.com>
Rui Ueyama <ruiu@google.com>
Sami Pietilä <samipietila@google.com>
Sarah Parker <sarahparker@google.com>
Sasi Inguva <isasi@google.com>
Scott Graham <scottmg@chromium.org>
Scott LaVarnway <slavarnway@google.com>
@@ -130,6 +143,8 @@ Thijs Vermeir <thijsvermeir@gmail.com>
Tim Kopp <tkopp@google.com>
Timothy B. Terriberry <tterribe@xiph.org>
Tom Finegan <tomfinegan@google.com>
Tristan Matthews <le.businessman@gmail.com>
Urvang Joshi <urvang@google.com>
Vignesh Venkatasubramanian <vigneshv@google.com>
Yaowu Xu <yaowu@google.com>
Yi Luo <luoyi@google.com>

View File

@@ -1,3 +1,19 @@
2017-01-09 v1.6.1 "Long Tailed Duck"
This release improves upon the VP9 encoder and speeds up the encoding and
decoding processes.
- Upgrading:
This release is ABI compatible with 1.6.0.
- Enhancements:
Faster VP9 encoding and decoding.
High bit depth builds now provide similar speed for 8 bit encode and decode
for x86 targets. Other platforms and higher bit depth improvements are in
progress.
- Bug Fixes:
A variety of fuzzing issues.
2016-07-20 v1.6.0 "Khaki Campbell Duck"
This release improves upon the VP9 encoder and speeds up the encoding and
decoding processes.

32
README
View File

@@ -1,4 +1,4 @@
README - 20 July 2016
README - 9 January 2017
Welcome to the WebM VP8/VP9 Codec SDK!
@@ -47,10 +47,9 @@ COMPILING THE APPLICATIONS/LIBRARIES:
--help output of the configure script. As of this writing, the list of
available targets is:
armv6-linux-rvct
armv6-linux-gcc
armv6-none-rvct
arm64-android-gcc
arm64-darwin-gcc
arm64-linux-gcc
armv7-android-gcc
armv7-darwin-gcc
armv7-linux-rvct
@@ -60,6 +59,7 @@ COMPILING THE APPLICATIONS/LIBRARIES:
armv7-win32-vs12
armv7-win32-vs14
armv7s-darwin-gcc
armv8-linux-gcc
mips32-linux-gcc
mips64-linux-gcc
sparc-solaris-gcc
@@ -73,15 +73,13 @@ COMPILING THE APPLICATIONS/LIBRARIES:
x86-darwin12-gcc
x86-darwin13-gcc
x86-darwin14-gcc
x86-darwin15-gcc
x86-iphonesimulator-gcc
x86-linux-gcc
x86-linux-icc
x86-os2-gcc
x86-solaris-gcc
x86-win32-gcc
x86-win32-vs7
x86-win32-vs8
x86-win32-vs9
x86-win32-vs10
x86-win32-vs11
x86-win32-vs12
@@ -93,13 +91,12 @@ COMPILING THE APPLICATIONS/LIBRARIES:
x86_64-darwin12-gcc
x86_64-darwin13-gcc
x86_64-darwin14-gcc
x86_64-darwin15-gcc
x86_64-iphonesimulator-gcc
x86_64-linux-gcc
x86_64-linux-icc
x86_64-solaris-gcc
x86_64-win64-gcc
x86_64-win64-vs8
x86_64-win64-vs9
x86_64-win64-vs10
x86_64-win64-vs11
x86_64-win64-vs12
@@ -132,7 +129,22 @@ VP8/VP9 TEST VECTORS:
$ ./configure --enable-unit-tests
$ LIBVPX_TEST_DATA_PATH=../libvpx-test-data make testdata
CODE STYLE:
The coding style used by this project is enforced with clang-format using the
configuration contained in the .clang-format file in the root of the
repository.
Before pushing changes for review you can format your code with:
# Apply clang-format to modified .c, .h and .cc files
$ clang-format -i --style=file \
$(git diff --name-only --diff-filter=ACMR '*.[hc]' '*.cc')
Check the .clang-format file for the version used to generate it if there is
any difference between your local formatting and the review system.
See also: http://clang.llvm.org/docs/ClangFormat.html
SUPPORT
This library is an open source project supported by its community. Please
please email webm-discuss@webmproject.org for help.
email webm-discuss@webmproject.org for help.

125
args.c
View File

@@ -8,12 +8,12 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include <stdlib.h>
#include <string.h>
#include <limits.h>
#include "args.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/msvc.h"
#if defined(__GNUC__) && __GNUC__
@@ -22,42 +22,36 @@ extern void die(const char *fmt, ...) __attribute__((noreturn));
extern void die(const char *fmt, ...);
#endif
struct arg arg_init(char **argv) {
struct arg a;
a.argv = argv;
a.argv = argv;
a.argv_step = 1;
a.name = NULL;
a.val = NULL;
a.def = NULL;
a.name = NULL;
a.val = NULL;
a.def = NULL;
return a;
}
int arg_match(struct arg *arg_, const struct arg_def *def, char **argv) {
struct arg arg;
if (!argv[0] || argv[0][0] != '-')
return 0;
if (!argv[0] || argv[0][0] != '-') return 0;
arg = arg_init(argv);
if (def->short_name
&& strlen(arg.argv[0]) == strlen(def->short_name) + 1
&& !strcmp(arg.argv[0] + 1, def->short_name)) {
if (def->short_name && strlen(arg.argv[0]) == strlen(def->short_name) + 1 &&
!strcmp(arg.argv[0] + 1, def->short_name)) {
arg.name = arg.argv[0] + 1;
arg.val = def->has_val ? arg.argv[1] : NULL;
arg.argv_step = def->has_val ? 2 : 1;
} else if (def->long_name) {
const size_t name_len = strlen(def->long_name);
if (strlen(arg.argv[0]) >= name_len + 2
&& arg.argv[0][1] == '-'
&& !strncmp(arg.argv[0] + 2, def->long_name, name_len)
&& (arg.argv[0][name_len + 2] == '='
|| arg.argv[0][name_len + 2] == '\0')) {
if (strlen(arg.argv[0]) >= name_len + 2 && arg.argv[0][1] == '-' &&
!strncmp(arg.argv[0] + 2, def->long_name, name_len) &&
(arg.argv[0][name_len + 2] == '=' ||
arg.argv[0][name_len + 2] == '\0')) {
arg.name = arg.argv[0] + 2;
arg.val = arg.name[name_len] == '=' ? arg.name + name_len + 1 : NULL;
arg.argv_step = 1;
@@ -70,8 +64,7 @@ int arg_match(struct arg *arg_, const struct arg_def *def, char **argv) {
if (arg.name && arg.val && !def->has_val)
die("Error: option %s requires no argument.\n", arg.name);
if (arg.name
&& (arg.val || !def->has_val)) {
if (arg.name && (arg.val || !def->has_val)) {
arg.def = def;
*arg_ = arg;
return 1;
@@ -80,15 +73,12 @@ int arg_match(struct arg *arg_, const struct arg_def *def, char **argv) {
return 0;
}
const char *arg_next(struct arg *arg) {
if (arg->argv[0])
arg->argv += arg->argv_step;
if (arg->argv[0]) arg->argv += arg->argv_step;
return *arg->argv;
}
char **argv_dup(int argc, const char **argv) {
char **new_argv = malloc((argc + 1) * sizeof(*argv));
@@ -97,9 +87,8 @@ char **argv_dup(int argc, const char **argv) {
return new_argv;
}
void arg_show_usage(FILE *fp, const struct arg_def *const *defs) {
char option_text[40] = {0};
char option_text[40] = { 0 };
for (; *defs; defs++) {
const struct arg_def *def = *defs;
@@ -109,15 +98,12 @@ void arg_show_usage(FILE *fp, const struct arg_def *const *defs) {
if (def->short_name && def->long_name) {
char *comma = def->has_val ? "," : ", ";
snprintf(option_text, 37, "-%s%s%s --%s%6s",
def->short_name, short_val, comma,
def->long_name, long_val);
snprintf(option_text, 37, "-%s%s%s --%s%6s", def->short_name, short_val,
comma, def->long_name, long_val);
} else if (def->short_name)
snprintf(option_text, 37, "-%s%s",
def->short_name, short_val);
snprintf(option_text, 37, "-%s%s", def->short_name, short_val);
else if (def->long_name)
snprintf(option_text, 37, " --%s%s",
def->long_name, long_val);
snprintf(option_text, 37, " --%s%s", def->long_name, long_val);
fprintf(fp, " %-37s\t%s\n", option_text, def->desc);
@@ -127,110 +113,103 @@ void arg_show_usage(FILE *fp, const struct arg_def *const *defs) {
fprintf(fp, " %-37s\t ", "");
for (listptr = def->enums; listptr->name; listptr++)
fprintf(fp, "%s%s", listptr->name,
listptr[1].name ? ", " : "\n");
fprintf(fp, "%s%s", listptr->name, listptr[1].name ? ", " : "\n");
}
}
}
unsigned int arg_parse_uint(const struct arg *arg) {
long int rawval;
char *endptr;
uint32_t rawval;
char *endptr;
rawval = strtol(arg->val, &endptr, 10);
rawval = (uint32_t)strtoul(arg->val, &endptr, 10);
if (arg->val[0] != '\0' && endptr[0] == '\0') {
if (rawval >= 0 && rawval <= UINT_MAX)
return rawval;
if (rawval <= UINT_MAX) return rawval;
die("Option %s: Value %ld out of range for unsigned int\n",
arg->name, rawval);
die("Option %s: Value %ld out of range for unsigned int\n", arg->name,
rawval);
}
die("Option %s: Invalid character '%c'\n", arg->name, *endptr);
return 0;
}
int arg_parse_int(const struct arg *arg) {
long int rawval;
char *endptr;
int32_t rawval;
char *endptr;
rawval = strtol(arg->val, &endptr, 10);
rawval = (int32_t)strtol(arg->val, &endptr, 10);
if (arg->val[0] != '\0' && endptr[0] == '\0') {
if (rawval >= INT_MIN && rawval <= INT_MAX)
return rawval;
if (rawval >= INT_MIN && rawval <= INT_MAX) return (int)rawval;
die("Option %s: Value %ld out of range for signed int\n",
arg->name, rawval);
die("Option %s: Value %ld out of range for signed int\n", arg->name,
rawval);
}
die("Option %s: Invalid character '%c'\n", arg->name, *endptr);
return 0;
}
struct vpx_rational {
int num; /**< fraction numerator */
int den; /**< fraction denominator */
};
struct vpx_rational arg_parse_rational(const struct arg *arg) {
long int rawval;
char *endptr;
struct vpx_rational rat;
long int rawval;
char *endptr;
struct vpx_rational rat;
/* parse numerator */
rawval = strtol(arg->val, &endptr, 10);
if (arg->val[0] != '\0' && endptr[0] == '/') {
if (rawval >= INT_MIN && rawval <= INT_MAX)
rat.num = rawval;
else die("Option %s: Value %ld out of range for signed int\n",
arg->name, rawval);
} else die("Option %s: Expected / at '%c'\n", arg->name, *endptr);
rat.num = (int)rawval;
else
die("Option %s: Value %ld out of range for signed int\n", arg->name,
rawval);
} else
die("Option %s: Expected / at '%c'\n", arg->name, *endptr);
/* parse denominator */
rawval = strtol(endptr + 1, &endptr, 10);
if (arg->val[0] != '\0' && endptr[0] == '\0') {
if (rawval >= INT_MIN && rawval <= INT_MAX)
rat.den = rawval;
else die("Option %s: Value %ld out of range for signed int\n",
arg->name, rawval);
} else die("Option %s: Invalid character '%c'\n", arg->name, *endptr);
rat.den = (int)rawval;
else
die("Option %s: Value %ld out of range for signed int\n", arg->name,
rawval);
} else
die("Option %s: Invalid character '%c'\n", arg->name, *endptr);
return rat;
}
int arg_parse_enum(const struct arg *arg) {
const struct arg_enum_list *listptr;
long int rawval;
char *endptr;
long int rawval;
char *endptr;
/* First see if the value can be parsed as a raw value */
rawval = strtol(arg->val, &endptr, 10);
if (arg->val[0] != '\0' && endptr[0] == '\0') {
/* Got a raw value, make sure it's valid */
for (listptr = arg->def->enums; listptr->name; listptr++)
if (listptr->val == rawval)
return rawval;
if (listptr->val == rawval) return (int)rawval;
}
/* Next see if it can be parsed as a string */
for (listptr = arg->def->enums; listptr->name; listptr++)
if (!strcmp(arg->val, listptr->name))
return listptr->val;
if (!strcmp(arg->val, listptr->name)) return listptr->val;
die("Option %s: Invalid value '%s'\n", arg->name, arg->val);
return 0;
}
int arg_parse_enum_or_int(const struct arg *arg) {
if (arg->def->enums)
return arg_parse_enum(arg);
if (arg->def->enums) return arg_parse_enum(arg);
return arg_parse_int(arg);
}

27
args.h
View File

@@ -8,7 +8,6 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#ifndef ARGS_H_
#define ARGS_H_
#include <stdio.h>
@@ -18,29 +17,33 @@ extern "C" {
#endif
struct arg {
char **argv;
const char *name;
const char *val;
unsigned int argv_step;
const struct arg_def *def;
char **argv;
const char *name;
const char *val;
unsigned int argv_step;
const struct arg_def *def;
};
struct arg_enum_list {
const char *name;
int val;
int val;
};
#define ARG_ENUM_LIST_END {0}
#define ARG_ENUM_LIST_END \
{ 0 }
typedef struct arg_def {
const char *short_name;
const char *long_name;
int has_val;
int has_val;
const char *desc;
const struct arg_enum_list *enums;
} arg_def_t;
#define ARG_DEF(s,l,v,d) {s,l,v,d, NULL}
#define ARG_DEF_ENUM(s,l,v,d,e) {s,l,v,d,e}
#define ARG_DEF_LIST_END {0}
#define ARG_DEF(s, l, v, d) \
{ s, l, v, d, NULL }
#define ARG_DEF_ENUM(s, l, v, d, e) \
{ s, l, v, d, e }
#define ARG_DEF_LIST_END \
{ 0 }
struct arg arg_init(char **argv);
int arg_match(struct arg *arg_, const struct arg_def *def, char **argv);

View File

@@ -29,11 +29,6 @@
# include $(CLEAR_VARS)
# include jni/libvpx/build/make/Android.mk
#
# There are currently two TARGET_ARCH_ABI targets for ARM.
# armeabi and armeabi-v7a. armeabi-v7a is selected by creating an
# Application.mk in the jni directory that contains:
# APP_ABI := armeabi-v7a
#
# By default libvpx will detect at runtime the existance of NEON extension.
# For this we import the 'cpufeatures' module from the NDK sources.
# libvpx can also be configured without this runtime detection method.
@@ -42,28 +37,44 @@
# --disable-neon-asm
# will remove any NEON dependency.
# To change to building armeabi, run ./libvpx/configure again, but with
# --target=armv6-android-gcc and modify the Application.mk file to
# set APP_ABI := armeabi
#
# Running ndk-build will build libvpx and include it in your project.
#
# Alternatively, building the examples and unit tests can be accomplished in the
# following way:
#
# Create a standalone toolchain from the NDK:
# https://developer.android.com/ndk/guides/standalone_toolchain.html
#
# For example - to test on arm64 devices with clang:
# $NDK/build/tools/make_standalone_toolchain.py \
# --arch arm64 --install-dir=/tmp/my-android-toolchain
# export PATH=/tmp/my-android-toolchain/bin:$PATH
# CROSS=aarch64-linux-android- CC=clang CXX=clang++ /path/to/libvpx/configure \
# --target=arm64-android-gcc
#
# Push the resulting binaries to a device and run them:
# adb push test_libvpx /data/tmp/test_libvpx
# adb shell /data/tmp/test_libvpx --gtest_filter=\*Sixtap\*
#
# Make sure to push the test data as well and set LIBVPX_TEST_DATA
CONFIG_DIR := $(LOCAL_PATH)/
LIBVPX_PATH := $(LOCAL_PATH)/libvpx
ASM_CNV_PATH_LOCAL := $(TARGET_ARCH_ABI)/ads2gas
ASM_CNV_PATH := $(LOCAL_PATH)/$(ASM_CNV_PATH_LOCAL)
ifneq ($(V),1)
qexec := @
endif
# Use the makefiles generated by upstream configure to determine which files to
# build. Also set any architecture-specific flags.
ifeq ($(TARGET_ARCH_ABI),armeabi-v7a)
include $(CONFIG_DIR)libs-armv7-android-gcc.mk
LOCAL_ARM_MODE := arm
else ifeq ($(TARGET_ARCH_ABI),armeabi)
include $(CONFIG_DIR)libs-armv6-android-gcc.mk
LOCAL_ARM_MODE := arm
else ifeq ($(TARGET_ARCH_ABI),arm64-v8a)
include $(CONFIG_DIR)libs-armv8-android-gcc.mk
include $(CONFIG_DIR)libs-arm64-android-gcc.mk
LOCAL_ARM_MODE := arm
else ifeq ($(TARGET_ARCH_ABI),x86)
include $(CONFIG_DIR)libs-x86-android-gcc.mk
@@ -93,10 +104,10 @@ LOCAL_CFLAGS := -O3
# like x86inc.asm and x86_abi_support.asm
LOCAL_ASMFLAGS := -I$(LIBVPX_PATH)
.PRECIOUS: %.asm.s
$(ASM_CNV_PATH)/libvpx/%.asm.s: $(LIBVPX_PATH)/%.asm
@mkdir -p $(dir $@)
@$(CONFIG_DIR)$(ASM_CONVERSION) <$< > $@
.PRECIOUS: %.asm.S
$(ASM_CNV_PATH)/libvpx/%.asm.S: $(LIBVPX_PATH)/%.asm
$(qexec)mkdir -p $(dir $@)
$(qexec)$(CONFIG_DIR)$(ASM_CONVERSION) <$< > $@
# For building *_rtcd.h, which have rules in libs.mk
TGT_ISA:=$(word 1, $(subst -, ,$(TOOLCHAIN)))
@@ -124,7 +135,7 @@ endif
# Pull out assembly files, splitting NEON from the rest. This is
# done to specify that the NEON assembly files use NEON assembler flags.
# x86 assembly matches %.asm, arm matches %.asm.s
# x86 assembly matches %.asm, arm matches %.asm.S
# x86:
@@ -132,31 +143,44 @@ CODEC_SRCS_ASM_X86 = $(filter %.asm, $(CODEC_SRCS_UNIQUE))
LOCAL_SRC_FILES += $(foreach file, $(CODEC_SRCS_ASM_X86), libvpx/$(file))
# arm:
CODEC_SRCS_ASM_ARM_ALL = $(filter %.asm.s, $(CODEC_SRCS_UNIQUE))
CODEC_SRCS_ASM_ARM_ALL = $(filter %.asm.S, $(CODEC_SRCS_UNIQUE))
CODEC_SRCS_ASM_ARM = $(foreach v, \
$(CODEC_SRCS_ASM_ARM_ALL), \
$(if $(findstring neon,$(v)),,$(v)))
CODEC_SRCS_ASM_ADS2GAS = $(patsubst %.s, \
$(ASM_CNV_PATH_LOCAL)/libvpx/%.s, \
CODEC_SRCS_ASM_ADS2GAS = $(patsubst %.S, \
$(ASM_CNV_PATH_LOCAL)/libvpx/%.S, \
$(CODEC_SRCS_ASM_ARM))
LOCAL_SRC_FILES += $(CODEC_SRCS_ASM_ADS2GAS)
ifeq ($(TARGET_ARCH_ABI),armeabi-v7a)
ASM_INCLUDES := vpx_dsp/arm/idct_neon.asm.S
CODEC_SRCS_ASM_NEON = $(foreach v, \
$(CODEC_SRCS_ASM_ARM_ALL),\
$(if $(findstring neon,$(v)),$(v),))
CODEC_SRCS_ASM_NEON_ADS2GAS = $(patsubst %.s, \
$(ASM_CNV_PATH_LOCAL)/libvpx/%.s, \
CODEC_SRCS_ASM_NEON := $(filter-out $(addprefix %, $(ASM_INCLUDES)), \
$(CODEC_SRCS_ASM_NEON))
CODEC_SRCS_ASM_NEON_ADS2GAS = $(patsubst %.S, \
$(ASM_CNV_PATH_LOCAL)/libvpx/%.S, \
$(CODEC_SRCS_ASM_NEON))
LOCAL_SRC_FILES += $(patsubst %.s, \
%.s.neon, \
LOCAL_SRC_FILES += $(patsubst %.S, \
%.S.neon, \
$(CODEC_SRCS_ASM_NEON_ADS2GAS))
NEON_ASM_TARGETS = $(patsubst %.S, \
$(ASM_CNV_PATH)/libvpx/%.S, \
$(CODEC_SRCS_ASM_NEON))
# add a dependency to the full path to the ads2gas output to ensure the
# includes are converted first.
ifneq ($(strip $(NEON_ASM_TARGETS)),)
$(NEON_ASM_TARGETS): $(addprefix $(ASM_CNV_PATH)/libvpx/, $(ASM_INCLUDES))
endif
endif
LOCAL_CFLAGS += \
-DHAVE_CONFIG_H=vpx_config.h \
-I$(LIBVPX_PATH) \
-I$(ASM_CNV_PATH)
-I$(ASM_CNV_PATH) \
-I$(ASM_CNV_PATH)/libvpx
LOCAL_MODULE := libvpx
@@ -177,7 +201,8 @@ endif
$$(rtcd_dep_template_SRCS): vpx_scale_rtcd.h
$$(rtcd_dep_template_SRCS): vpx_dsp_rtcd.h
ifneq ($(findstring $(TARGET_ARCH_ABI),x86 x86_64),)
rtcd_dep_template_CONFIG_ASM_ABIS := x86 x86_64 armeabi-v7a
ifneq ($$(findstring $(TARGET_ARCH_ABI),$$(rtcd_dep_template_CONFIG_ASM_ABIS)),)
$$(rtcd_dep_template_SRCS): vpx_config.asm
endif
endef
@@ -187,16 +212,17 @@ $(eval $(call rtcd_dep_template))
.PHONY: clean
clean:
@echo "Clean: ads2gas files [$(TARGET_ARCH_ABI)]"
@$(RM) $(CODEC_SRCS_ASM_ADS2GAS) $(CODEC_SRCS_ASM_NEON_ADS2GAS)
@$(RM) -r $(ASM_CNV_PATH)
@$(RM) $(CLEAN-OBJS)
$(qexec)$(RM) $(CODEC_SRCS_ASM_ADS2GAS) $(CODEC_SRCS_ASM_NEON_ADS2GAS)
$(qexec)$(RM) -r $(ASM_CNV_PATH)
$(qexec)$(RM) $(CLEAN-OBJS)
ifeq ($(ENABLE_SHARED),1)
LOCAL_CFLAGS += -fPIC
include $(BUILD_SHARED_LIBRARY)
else
include $(BUILD_STATIC_LIBRARY)
endif
ifeq ($(CONFIG_RUNTIME_CPU_DETECT),yes)
$(call import-module,cpufeatures)
$(call import-module,android/cpufeatures)
endif

View File

@@ -26,7 +26,7 @@ test-no-data-check:: .DEFAULT
testdata:: .DEFAULT
utiltest: .DEFAULT
exampletest-no-data-check utiltest-no-data-check: .DEFAULT
test_%: .DEFAULT ;
# Note: md5sum is not installed on OS X, but openssl is. Openssl may not be
# installed on cygwin, so we need to autodetect here.
@@ -90,7 +90,7 @@ all:
.PHONY: clean
clean::
rm -f $(OBJS-yes) $(OBJS-yes:.o=.d) $(OBJS-yes:.asm.s.o=.asm.s)
rm -f $(OBJS-yes) $(OBJS-yes:.o=.d) $(OBJS-yes:.asm.S.o=.asm.S)
rm -f $(CLEAN-OBJS)
.PHONY: clean
@@ -180,13 +180,13 @@ $(BUILD_PFX)%.asm.o: %.asm
$(qexec)$(if $(CONFIG_DEPENDENCY_TRACKING),,mkdir -p $(dir $@))
$(qexec)$(AS) $(ASFLAGS) -o $@ $<
$(BUILD_PFX)%.s.d: %.s
$(BUILD_PFX)%.S.d: %.S
$(if $(quiet),@echo " [DEP] $@")
$(qexec)mkdir -p $(dir $@)
$(qexec)$(SRC_PATH_BARE)/build/make/gen_asm_deps.sh \
--build-pfx=$(BUILD_PFX) --depfile=$@ $(ASFLAGS) $< > $@
$(BUILD_PFX)%.s.o: %.s
$(BUILD_PFX)%.S.o: %.S
$(if $(quiet),@echo " [AS] $@")
$(qexec)$(if $(CONFIG_DEPENDENCY_TRACKING),,mkdir -p $(dir $@))
$(qexec)$(AS) $(ASFLAGS) -o $@ $<
@@ -198,8 +198,8 @@ $(BUILD_PFX)%.c.S: %.c
$(qexec)$(if $(CONFIG_DEPENDENCY_TRACKING),,mkdir -p $(dir $@))
$(qexec)$(CC) -S $(CFLAGS) -o $@ $<
.PRECIOUS: %.asm.s
$(BUILD_PFX)%.asm.s: %.asm
.PRECIOUS: %.asm.S
$(BUILD_PFX)%.asm.S: %.asm
$(if $(quiet),@echo " [ASM CONVERSION] $@")
$(qexec)mkdir -p $(dir $@)
$(qexec)$(ASM_CONVERSION) <$< >$@
@@ -418,7 +418,6 @@ ifneq ($(call enabled,DIST-SRCS),)
DIST-SRCS-yes += build/make/gen_asm_deps.sh
DIST-SRCS-yes += build/make/Makefile
DIST-SRCS-$(CONFIG_MSVS) += build/make/gen_msvs_def.sh
DIST-SRCS-$(CONFIG_MSVS) += build/make/gen_msvs_proj.sh
DIST-SRCS-$(CONFIG_MSVS) += build/make/gen_msvs_sln.sh
DIST-SRCS-$(CONFIG_MSVS) += build/make/gen_msvs_vcxproj.sh
DIST-SRCS-$(CONFIG_MSVS) += build/make/msvs_common.sh
@@ -449,3 +448,5 @@ all: $(BUILD_TARGETS)
install:: $(INSTALL_TARGETS)
dist: $(INSTALL_TARGETS)
test::
.SUFFIXES: # Delete default suffix rules

View File

@@ -138,14 +138,6 @@ while (<STDIN>)
s/DCD(.*)/.long $1/;
s/DCB(.*)/.byte $1/;
# RN to .req
if (s/RN\s+([Rr]\d+|lr)/.req $1/)
{
print;
print "$comment_sub$comment\n" if defined $comment;
next;
}
# Make function visible to linker, and make additional symbol with
# prepended underscore
s/EXPORT\s+\|([\$\w]*)\|/.global $1 \n\t.type $1, function/;

View File

@@ -18,12 +18,6 @@
# Usage: cat inputfile | perl ads2gas_apple.pl > outputfile
#
my $chromium = 0;
foreach my $arg (@ARGV) {
$chromium = 1 if ($arg eq "-chromium");
}
print "@ This file was created from a .asm file\n";
print "@ using the ads2gas_apple.pl script.\n\n";
print "\t.set WIDE_REFERENCE, 0\n";
@@ -126,18 +120,6 @@ while (<STDIN>)
s/DCD(.*)/.long $1/;
s/DCB(.*)/.byte $1/;
# Build a hash of all the register - alias pairs.
if (s/(.*)RN(.*)/$1 .req $2/g)
{
$register_aliases{trim($1)} = trim($2);
next;
}
while (($key, $value) = each(%register_aliases))
{
s/\b$key\b/$value/g;
}
# Make function visible to linker, and make additional symbol with
# prepended underscore
s/EXPORT\s+\|([\$\w]*)\|/.globl _$1\n\t.globl $1/;
@@ -218,18 +200,5 @@ while (<STDIN>)
s/\bMEND\b/.endm/; # No need to tell it where to stop assembling
next if /^\s*END\s*$/;
# Clang used by Chromium differs slightly from clang in XCode in what it
# will accept in the assembly.
if ($chromium) {
s/qsubaddx/qsax/i;
s/qaddsubx/qasx/i;
s/ldrneb/ldrbne/i;
s/ldrneh/ldrhne/i;
s/(vqshrun\.s16 .*, \#)0$/${1}8/i;
# http://llvm.org/bugs/show_bug.cgi?id=16022
s/\.include/#include/;
}
print;
}

View File

@@ -635,7 +635,7 @@ setup_gnu_toolchain() {
AS=${AS:-${CROSS}as}
STRIP=${STRIP:-${CROSS}strip}
NM=${NM:-${CROSS}nm}
AS_SFX=.s
AS_SFX=.S
EXE_SFX=
}
@@ -680,9 +680,6 @@ process_common_toolchain() {
aarch64*)
tgt_isa=arm64
;;
armv6*)
tgt_isa=armv6
;;
armv7*-hardfloat* | armv7*-gnueabihf | arm-*-gnueabihf)
tgt_isa=armv7
float_abi=hard
@@ -883,36 +880,6 @@ process_common_toolchain() {
if disabled neon && enabled neon_asm; then
die "Disabling neon while keeping neon-asm is not supported"
fi
case ${toolchain} in
# Apple iOS SDKs no longer support armv6 as of the version 9
# release (coincides with release of Xcode 7). Only enable media
# when using earlier SDK releases.
*-darwin*)
if [ "$(show_darwin_sdk_major_version iphoneos)" -lt 9 ]; then
soft_enable media
else
soft_disable media
RTCD_OPTIONS="${RTCD_OPTIONS}--disable-media "
fi
;;
*)
soft_enable media
;;
esac
;;
armv6)
case ${toolchain} in
*-darwin*)
if [ "$(show_darwin_sdk_major_version iphoneos)" -lt 9 ]; then
soft_enable media
else
die "Your iOS SDK does not support armv6."
fi
;;
*)
soft_enable media
;;
esac
;;
esac
@@ -959,7 +926,7 @@ EOF
;;
vs*)
asm_conversion_cmd="${source_path}/build/make/ads2armasm_ms.pl"
AS_SFX=.s
AS_SFX=.S
msvs_arch_dir=arm-msvs
disable_feature multithread
disable_feature unit_tests
@@ -969,6 +936,7 @@ EOF
# only "AppContainerApplication" which requires an AppxManifest.
# Therefore disable the examples, just build the library.
disable_feature examples
disable_feature tools
fi
;;
rvct)
@@ -1011,47 +979,50 @@ EOF
;;
android*)
if [ -z "${sdk_path}" ]; then
die "Must specify --sdk-path for Android builds."
fi
if [ -n "${sdk_path}" ]; then
SDK_PATH=${sdk_path}
COMPILER_LOCATION=`find "${SDK_PATH}" \
-name "arm-linux-androideabi-gcc*" -print -quit`
TOOLCHAIN_PATH=${COMPILER_LOCATION%/*}/arm-linux-androideabi-
CC=${TOOLCHAIN_PATH}gcc
CXX=${TOOLCHAIN_PATH}g++
AR=${TOOLCHAIN_PATH}ar
LD=${TOOLCHAIN_PATH}gcc
AS=${TOOLCHAIN_PATH}as
STRIP=${TOOLCHAIN_PATH}strip
NM=${TOOLCHAIN_PATH}nm
SDK_PATH=${sdk_path}
COMPILER_LOCATION=`find "${SDK_PATH}" \
-name "arm-linux-androideabi-gcc*" -print -quit`
TOOLCHAIN_PATH=${COMPILER_LOCATION%/*}/arm-linux-androideabi-
CC=${TOOLCHAIN_PATH}gcc
CXX=${TOOLCHAIN_PATH}g++
AR=${TOOLCHAIN_PATH}ar
LD=${TOOLCHAIN_PATH}gcc
AS=${TOOLCHAIN_PATH}as
STRIP=${TOOLCHAIN_PATH}strip
NM=${TOOLCHAIN_PATH}nm
if [ -z "${alt_libc}" ]; then
alt_libc=`find "${SDK_PATH}" -name arch-arm -print | \
awk '{n = split($0,a,"/"); \
if [ -z "${alt_libc}" ]; then
alt_libc=`find "${SDK_PATH}" -name arch-arm -print | \
awk '{n = split($0,a,"/"); \
split(a[n-1],b,"-"); \
print $0 " " b[2]}' | \
sort -g -k 2 | \
awk '{ print $1 }' | tail -1`
fi
fi
if [ -d "${alt_libc}" ]; then
add_cflags "--sysroot=${alt_libc}"
add_ldflags "--sysroot=${alt_libc}"
fi
if [ -d "${alt_libc}" ]; then
add_cflags "--sysroot=${alt_libc}"
add_ldflags "--sysroot=${alt_libc}"
fi
# linker flag that routes around a CPU bug in some
# Cortex-A8 implementations (NDK Dev Guide)
add_ldflags "-Wl,--fix-cortex-a8"
# linker flag that routes around a CPU bug in some
# Cortex-A8 implementations (NDK Dev Guide)
add_ldflags "-Wl,--fix-cortex-a8"
enable_feature pic
soft_enable realtime_only
if [ ${tgt_isa} = "armv7" ]; then
soft_enable runtime_cpu_detect
fi
if enabled runtime_cpu_detect; then
add_cflags "-I${SDK_PATH}/sources/android/cpufeatures"
enable_feature pic
soft_enable realtime_only
if [ ${tgt_isa} = "armv7" ]; then
soft_enable runtime_cpu_detect
fi
if enabled runtime_cpu_detect; then
add_cflags "-I${SDK_PATH}/sources/android/cpufeatures"
fi
else
echo "Assuming standalone build with NDK toolchain."
echo "See build/make/Android.mk for details."
check_add_ldflags -static
soft_enable unit_tests
fi
;;
@@ -1064,7 +1035,7 @@ EOF
STRIP="$(${XCRUN_FIND} strip)"
NM="$(${XCRUN_FIND} nm)"
RANLIB="$(${XCRUN_FIND} ranlib)"
AS_SFX=.s
AS_SFX=.S
LD="${CXX:-$(${XCRUN_FIND} ld)}"
# ASFLAGS is written here instead of using check_add_asflags
@@ -1153,13 +1124,13 @@ EOF
if [ -n "${tune_cpu}" ]; then
case ${tune_cpu} in
p5600)
check_add_cflags -mips32r5 -funroll-loops -mload-store-pairs
check_add_cflags -mips32r5 -mload-store-pairs
check_add_cflags -msched-weight -mhard-float -mfp64
check_add_asflags -mips32r5 -mhard-float -mfp64
check_add_ldflags -mfp64
;;
i6400)
check_add_cflags -mips64r6 -mabi=64 -funroll-loops -msched-weight
i6400|p6600)
check_add_cflags -mips64r6 -mabi=64 -msched-weight
check_add_cflags -mload-store-pairs -mhard-float -mfp64
check_add_asflags -mips64r6 -mabi=64 -mhard-float -mfp64
check_add_ldflags -mips64r6 -mabi=64 -mfp64
@@ -1396,10 +1367,6 @@ EOF
fi
fi
if [ "${tgt_isa}" = "x86_64" ] || [ "${tgt_isa}" = "x86" ]; then
soft_enable use_x86inc
fi
# Position Independent Code (PIC) support, for building relocatable
# shared objects
enabled gcc && enabled pic && check_add_cflags -fPIC
@@ -1429,6 +1396,7 @@ EOF
*-win*-vs*)
;;
*-android-gcc)
# bionic includes basic pthread functionality, obviating -lpthread.
;;
*)
check_header pthread.h && add_extralibs -lpthread

View File

@@ -1,490 +0,0 @@
#!/bin/bash
##
## Copyright (c) 2010 The WebM project authors. All Rights Reserved.
##
## Use of this source code is governed by a BSD-style license
## that can be found in the LICENSE file in the root of the source
## tree. An additional intellectual property rights grant can be found
## in the file PATENTS. All contributing project authors may
## be found in the AUTHORS file in the root of the source tree.
##
self=$0
self_basename=${self##*/}
self_dirname=$(dirname "$0")
. "$self_dirname/msvs_common.sh"|| exit 127
show_help() {
cat <<EOF
Usage: ${self_basename} --name=projname [options] file1 [file2 ...]
This script generates a Visual Studio project file from a list of source
code files.
Options:
--help Print this message
--exe Generate a project for building an Application
--lib Generate a project for creating a static library
--dll Generate a project for creating a dll
--static-crt Use the static C runtime (/MT)
--target=isa-os-cc Target specifier (required)
--out=filename Write output to a file [stdout]
--name=project_name Name of the project (required)
--proj-guid=GUID GUID to use for the project
--module-def=filename File containing export definitions (for DLLs)
--ver=version Version (7,8,9) of visual studio to generate for
--src-path-bare=dir Path to root of source tree
-Ipath/to/include Additional include directories
-DFLAG[=value] Preprocessor macros to define
-Lpath/to/lib Additional library search paths
-llibname Library to link against
EOF
exit 1
}
generate_filter() {
local var=$1
local name=$2
local pats=$3
local file_list_sz
local i
local f
local saveIFS="$IFS"
local pack
echo "generating filter '$name' from ${#file_list[@]} files" >&2
IFS=*
open_tag Filter \
Name=$name \
Filter=$pats \
UniqueIdentifier=`generate_uuid` \
file_list_sz=${#file_list[@]}
for i in ${!file_list[@]}; do
f=${file_list[i]}
for pat in ${pats//;/$IFS}; do
if [ "${f##*.}" == "$pat" ]; then
unset file_list[i]
objf=$(echo ${f%.*}.obj \
| sed -e "s,$src_path_bare,," \
-e 's/^[\./]\+//g' -e 's,[:/ ],_,g')
open_tag File RelativePath="$f"
if [ "$pat" == "asm" ] && $asm_use_custom_step; then
# Avoid object file name collisions, i.e. vpx_config.c and
# vpx_config.asm produce the same object file without
# this additional suffix.
objf=${objf%.obj}_asm.obj
for plat in "${platforms[@]}"; do
for cfg in Debug Release; do
open_tag FileConfiguration \
Name="${cfg}|${plat}" \
tag Tool \
Name="VCCustomBuildTool" \
Description="Assembling \$(InputFileName)" \
CommandLine="$(eval echo \$asm_${cfg}_cmdline) -o \$(IntDir)\\$objf" \
Outputs="\$(IntDir)\\$objf" \
close_tag FileConfiguration
done
done
fi
if [ "$pat" == "c" ] || \
[ "$pat" == "cc" ] || [ "$pat" == "cpp" ]; then
for plat in "${platforms[@]}"; do
for cfg in Debug Release; do
open_tag FileConfiguration \
Name="${cfg}|${plat}" \
tag Tool \
Name="VCCLCompilerTool" \
ObjectFile="\$(IntDir)\\$objf" \
close_tag FileConfiguration
done
done
fi
close_tag File
break
fi
done
done
close_tag Filter
IFS="$saveIFS"
}
# Process command line
unset target
for opt in "$@"; do
optval="${opt#*=}"
case "$opt" in
--help|-h) show_help
;;
--target=*) target="${optval}"
;;
--out=*) outfile="$optval"
;;
--name=*) name="${optval}"
;;
--proj-guid=*) guid="${optval}"
;;
--module-def=*) link_opts="${link_opts} ModuleDefinitionFile=${optval}"
;;
--exe) proj_kind="exe"
;;
--dll) proj_kind="dll"
;;
--lib) proj_kind="lib"
;;
--src-path-bare=*)
src_path_bare=$(fix_path "$optval")
src_path_bare=${src_path_bare%/}
;;
--static-crt) use_static_runtime=true
;;
--ver=*)
vs_ver="$optval"
case "$optval" in
[789])
;;
*) die Unrecognized Visual Studio Version in $opt
;;
esac
;;
-I*)
opt=${opt##-I}
opt=$(fix_path "$opt")
opt="${opt%/}"
incs="${incs}${incs:+;}&quot;${opt}&quot;"
yasmincs="${yasmincs} -I&quot;${opt}&quot;"
;;
-D*) defines="${defines}${defines:+;}${opt##-D}"
;;
-L*) # fudge . to $(OutDir)
if [ "${opt##-L}" == "." ]; then
libdirs="${libdirs}${libdirs:+;}&quot;\$(OutDir)&quot;"
else
# Also try directories for this platform/configuration
opt=${opt##-L}
opt=$(fix_path "$opt")
libdirs="${libdirs}${libdirs:+;}&quot;${opt}&quot;"
libdirs="${libdirs}${libdirs:+;}&quot;${opt}/\$(PlatformName)/\$(ConfigurationName)&quot;"
libdirs="${libdirs}${libdirs:+;}&quot;${opt}/\$(PlatformName)&quot;"
fi
;;
-l*) libs="${libs}${libs:+ }${opt##-l}.lib"
;;
-*) die_unknown $opt
;;
*)
# The paths in file_list are fixed outside of the loop.
file_list[${#file_list[@]}]="$opt"
case "$opt" in
*.asm) uses_asm=true
;;
esac
;;
esac
done
# Make one call to fix_path for file_list to improve performance.
fix_file_list file_list
outfile=${outfile:-/dev/stdout}
guid=${guid:-`generate_uuid`}
asm_use_custom_step=false
uses_asm=${uses_asm:-false}
case "${vs_ver:-8}" in
7) vs_ver_id="7.10"
asm_use_custom_step=$uses_asm
warn_64bit='Detect64BitPortabilityProblems=true'
;;
8) vs_ver_id="8.00"
asm_use_custom_step=$uses_asm
warn_64bit='Detect64BitPortabilityProblems=true'
;;
9) vs_ver_id="9.00"
asm_use_custom_step=$uses_asm
warn_64bit='Detect64BitPortabilityProblems=false'
;;
esac
[ -n "$name" ] || die "Project name (--name) must be specified!"
[ -n "$target" ] || die "Target (--target) must be specified!"
if ${use_static_runtime:-false}; then
release_runtime=0
debug_runtime=1
lib_sfx=mt
else
release_runtime=2
debug_runtime=3
lib_sfx=md
fi
# Calculate debug lib names: If a lib ends in ${lib_sfx}.lib, then rename
# it to ${lib_sfx}d.lib. This precludes linking to release libs from a
# debug exe, so this may need to be refactored later.
for lib in ${libs}; do
if [ "$lib" != "${lib%${lib_sfx}.lib}" ]; then
lib=${lib%.lib}d.lib
fi
debug_libs="${debug_libs}${debug_libs:+ }${lib}"
done
# List Keyword for this target
case "$target" in
x86*) keyword="ManagedCProj"
;;
*) die "Unsupported target $target!"
esac
# List of all platforms supported for this target
case "$target" in
x86_64*)
platforms[0]="x64"
asm_Debug_cmdline="yasm -Xvc -g cv8 -f win64 ${yasmincs} &quot;\$(InputPath)&quot;"
asm_Release_cmdline="yasm -Xvc -f win64 ${yasmincs} &quot;\$(InputPath)&quot;"
;;
x86*)
platforms[0]="Win32"
asm_Debug_cmdline="yasm -Xvc -g cv8 -f win32 ${yasmincs} &quot;\$(InputPath)&quot;"
asm_Release_cmdline="yasm -Xvc -f win32 ${yasmincs} &quot;\$(InputPath)&quot;"
;;
*) die "Unsupported target $target!"
;;
esac
generate_vcproj() {
case "$proj_kind" in
exe) vs_ConfigurationType=1
;;
dll) vs_ConfigurationType=2
;;
*) vs_ConfigurationType=4
;;
esac
echo "<?xml version=\"1.0\" encoding=\"Windows-1252\"?>"
open_tag VisualStudioProject \
ProjectType="Visual C++" \
Version="${vs_ver_id}" \
Name="${name}" \
ProjectGUID="{${guid}}" \
RootNamespace="${name}" \
Keyword="${keyword}" \
open_tag Platforms
for plat in "${platforms[@]}"; do
tag Platform Name="$plat"
done
close_tag Platforms
open_tag Configurations
for plat in "${platforms[@]}"; do
plat_no_ws=`echo $plat | sed 's/[^A-Za-z0-9_]/_/g'`
open_tag Configuration \
Name="Debug|$plat" \
OutputDirectory="\$(SolutionDir)$plat_no_ws/\$(ConfigurationName)" \
IntermediateDirectory="$plat_no_ws/\$(ConfigurationName)/${name}" \
ConfigurationType="$vs_ConfigurationType" \
CharacterSet="1" \
case "$target" in
x86*)
case "$name" in
vpx)
tag Tool \
Name="VCCLCompilerTool" \
Optimization="0" \
AdditionalIncludeDirectories="$incs" \
PreprocessorDefinitions="WIN32;_DEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
RuntimeLibrary="$debug_runtime" \
UsePrecompiledHeader="0" \
WarningLevel="3" \
DebugInformationFormat="2" \
$warn_64bit \
$uses_asm && tag Tool Name="YASM" IncludePaths="$incs" Debug="true"
;;
*)
tag Tool \
Name="VCCLCompilerTool" \
Optimization="0" \
AdditionalIncludeDirectories="$incs" \
PreprocessorDefinitions="WIN32;_DEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
RuntimeLibrary="$debug_runtime" \
UsePrecompiledHeader="0" \
WarningLevel="3" \
DebugInformationFormat="2" \
$warn_64bit \
$uses_asm && tag Tool Name="YASM" IncludePaths="$incs" Debug="true"
;;
esac
;;
esac
case "$proj_kind" in
exe)
case "$target" in
x86*)
case "$name" in
*)
tag Tool \
Name="VCLinkerTool" \
AdditionalDependencies="$debug_libs \$(NoInherit)" \
AdditionalLibraryDirectories="$libdirs" \
GenerateDebugInformation="true" \
ProgramDatabaseFile="\$(OutDir)/${name}.pdb" \
;;
esac
;;
esac
;;
lib)
case "$target" in
x86*)
tag Tool \
Name="VCLibrarianTool" \
OutputFile="\$(OutDir)/${name}${lib_sfx}d.lib" \
;;
esac
;;
dll)
tag Tool \
Name="VCLinkerTool" \
AdditionalDependencies="\$(NoInherit)" \
LinkIncremental="2" \
GenerateDebugInformation="true" \
AssemblyDebug="1" \
TargetMachine="1" \
$link_opts \
;;
esac
close_tag Configuration
open_tag Configuration \
Name="Release|$plat" \
OutputDirectory="\$(SolutionDir)$plat_no_ws/\$(ConfigurationName)" \
IntermediateDirectory="$plat_no_ws/\$(ConfigurationName)/${name}" \
ConfigurationType="$vs_ConfigurationType" \
CharacterSet="1" \
WholeProgramOptimization="0" \
case "$target" in
x86*)
case "$name" in
vpx)
tag Tool \
Name="VCCLCompilerTool" \
Optimization="2" \
FavorSizeorSpeed="1" \
AdditionalIncludeDirectories="$incs" \
PreprocessorDefinitions="WIN32;NDEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
RuntimeLibrary="$release_runtime" \
UsePrecompiledHeader="0" \
WarningLevel="3" \
DebugInformationFormat="0" \
$warn_64bit \
$uses_asm && tag Tool Name="YASM" IncludePaths="$incs"
;;
*)
tag Tool \
Name="VCCLCompilerTool" \
AdditionalIncludeDirectories="$incs" \
Optimization="2" \
FavorSizeorSpeed="1" \
PreprocessorDefinitions="WIN32;NDEBUG;_CRT_SECURE_NO_WARNINGS;_CRT_SECURE_NO_DEPRECATE;$defines" \
RuntimeLibrary="$release_runtime" \
UsePrecompiledHeader="0" \
WarningLevel="3" \
DebugInformationFormat="0" \
$warn_64bit \
$uses_asm && tag Tool Name="YASM" IncludePaths="$incs"
;;
esac
;;
esac
case "$proj_kind" in
exe)
case "$target" in
x86*)
case "$name" in
*)
tag Tool \
Name="VCLinkerTool" \
AdditionalDependencies="$libs \$(NoInherit)" \
AdditionalLibraryDirectories="$libdirs" \
;;
esac
;;
esac
;;
lib)
case "$target" in
x86*)
tag Tool \
Name="VCLibrarianTool" \
OutputFile="\$(OutDir)/${name}${lib_sfx}.lib" \
;;
esac
;;
dll) # note differences to debug version: LinkIncremental, AssemblyDebug
tag Tool \
Name="VCLinkerTool" \
AdditionalDependencies="\$(NoInherit)" \
LinkIncremental="1" \
GenerateDebugInformation="true" \
TargetMachine="1" \
$link_opts \
;;
esac
close_tag Configuration
done
close_tag Configurations
open_tag Files
generate_filter srcs "Source Files" "c;cc;cpp;def;odl;idl;hpj;bat;asm;asmx"
generate_filter hdrs "Header Files" "h;hm;inl;inc;xsd"
generate_filter resrcs "Resource Files" "rc;ico;cur;bmp;dlg;rc2;rct;bin;rgs;gif;jpg;jpeg;jpe;resx;tiff;tif;png;wav"
generate_filter resrcs "Build Files" "mk"
close_tag Files
tag Globals
close_tag VisualStudioProject
# This must be done from within the {} subshell
echo "Ignored files list (${#file_list[@]} items) is:" >&2
for f in "${file_list[@]}"; do
echo " $f" >&2
done
}
generate_vcproj |
sed -e '/"/s;\([^ "]\)/;\1\\;g' > ${outfile}
exit
<!--
TODO: Add any files not captured by filters.
<File
RelativePath=".\ReadMe.txt"
>
</File>
-->

View File

@@ -55,16 +55,11 @@ indent_pop() {
parse_project() {
local file=$1
if [ "$sfx" = "vcproj" ]; then
local name=`grep Name "$file" | awk 'BEGIN {FS="\""}{if (NR==1) print $2}'`
local guid=`grep ProjectGUID "$file" | awk 'BEGIN {FS="\""}{if (NR==1) print $2}'`
else
local name=`grep RootNamespace "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
local guid=`grep ProjectGuid "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
fi
local name=`grep RootNamespace "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
local guid=`grep ProjectGuid "$file" | sed 's,.*<.*>\(.*\)</.*>.*,\1,'`
# save the project GUID to a varaible, normalizing to the basename of the
# vcproj file without the extension
# vcxproj file without the extension
local var
var=${file##*/}
var=${var%%.${sfx}}
@@ -72,13 +67,8 @@ parse_project() {
eval "${var}_name=$name"
eval "${var}_guid=$guid"
if [ "$sfx" = "vcproj" ]; then
cur_config_list=`grep -A1 '<Configuration' $file |
grep Name | cut -d\" -f2`
else
cur_config_list=`grep -B1 'Label="Configuration"' $file |
grep Condition | cut -d\' -f4`
fi
cur_config_list=`grep -B1 'Label="Configuration"' $file |
grep Condition | cut -d\' -f4`
new_config_list=$(for i in $config_list $cur_config_list; do
echo $i
done | sort | uniq)
@@ -103,25 +93,6 @@ process_project() {
eval "${var}_guid=$guid"
echo "Project(\"{8BC9CEB8-8B4A-11D0-8D11-00A0C91BC942}\") = \"$name\", \"$file\", \"$guid\""
indent_push
eval "local deps=\"\${${var}_deps}\""
if [ -n "$deps" ] && [ "$sfx" = "vcproj" ]; then
echo "${indent}ProjectSection(ProjectDependencies) = postProject"
indent_push
for dep in $deps; do
eval "local dep_guid=\${${dep}_guid}"
[ -z "${dep_guid}" ] && die "Unknown GUID for $dep (dependency of $var)"
echo "${indent}$dep_guid = $dep_guid"
done
indent_pop
echo "${indent}EndProjectSection"
fi
indent_pop
echo "EndProject"
}
@@ -191,11 +162,7 @@ process_makefile() {
IFS=$'\r'$'\n'
local TAB=$'\t'
cat <<EOF
ifeq (\$(CONFIG_VS_VERSION),7)
MSBUILD_TOOL := devenv.com
else
MSBUILD_TOOL := msbuild.exe
endif
found_devenv := \$(shell which \$(MSBUILD_TOOL) >/dev/null 2>&1 && echo yes)
.nodevenv.once:
${TAB}@echo " * \$(MSBUILD_TOOL) not found in path."
@@ -204,7 +171,7 @@ ${TAB}@echo " * You will have to build all configurations manually using the"
${TAB}@echo " * Visual Studio IDE. To allow make to build them automatically,"
${TAB}@echo " * add the Common7/IDE directory of your Visual Studio"
${TAB}@echo " * installation to your path, eg:"
${TAB}@echo " * C:\Program Files\Microsoft Visual Studio 8\Common7\IDE"
${TAB}@echo " * C:\Program Files\Microsoft Visual Studio 10.0\Common7\IDE"
${TAB}@echo " * "
${TAB}@touch \$@
CLEAN-OBJS += \$(if \$(found_devenv),,.nodevenv.once)
@@ -221,16 +188,9 @@ clean::
${TAB}rm -rf "$platform"/"$config"
.PHONY: $nows_sln_config
ifneq (\$(found_devenv),)
ifeq (\$(CONFIG_VS_VERSION),7)
$nows_sln_config: $outfile
${TAB}\$(MSBUILD_TOOL) $outfile -build "$config"
else
$nows_sln_config: $outfile
${TAB}\$(MSBUILD_TOOL) $outfile -m -t:Build \\
${TAB}${TAB}-p:Configuration="$config" -p:Platform="$platform"
endif
else
$nows_sln_config: $outfile .nodevenv.once
${TAB}@echo " * Skipping build of $sln_config (\$(MSBUILD_TOOL) not in path)."
@@ -255,23 +215,12 @@ for opt in "$@"; do
;;
--ver=*) vs_ver="$optval"
case $optval in
[789]|10|11|12|14)
10|11|12|14)
;;
*) die Unrecognized Visual Studio Version in $opt
;;
esac
;;
--ver=*) vs_ver="$optval"
case $optval in
7) sln_vers="8.00"
sln_vers_str="Visual Studio .NET 2003"
;;
[89])
;;
*) die "Unrecognized Visual Studio Version '$optval' in $opt"
;;
esac
;;
--target=*) target="${optval}"
;;
-*) die_unknown $opt
@@ -281,16 +230,7 @@ for opt in "$@"; do
done
outfile=${outfile:-/dev/stdout}
mkoutfile=${mkoutfile:-/dev/stdout}
case "${vs_ver:-8}" in
7) sln_vers="8.00"
sln_vers_str="Visual Studio .NET 2003"
;;
8) sln_vers="9.00"
sln_vers_str="Visual Studio 2005"
;;
9) sln_vers="10.00"
sln_vers_str="Visual Studio 2008"
;;
case "${vs_ver:-10}" in
10) sln_vers="11.00"
sln_vers_str="Visual Studio 2010"
;;
@@ -304,14 +244,7 @@ case "${vs_ver:-8}" in
sln_vers_str="Visual Studio 2015"
;;
esac
case "${vs_ver:-8}" in
[789])
sfx=vcproj
;;
10|11|12|14)
sfx=vcxproj
;;
esac
sfx=vcxproj
for f in "${file_list[@]}"; do
parse_project $f

View File

@@ -82,7 +82,7 @@ generate_filter() {
| sed -e "s,$src_path_bare,," \
-e 's/^[\./]\+//g' -e 's,[:/ ],_,g')
if ([ "$pat" == "asm" ] || [ "$pat" == "s" ]) && $asm_use_custom_step; then
if ([ "$pat" == "asm" ] || [ "$pat" == "s" ] || [ "$pat" == "S" ]) && $asm_use_custom_step; then
# Avoid object file name collisions, i.e. vpx_config.c and
# vpx_config.asm produce the same object file without
# this additional suffix.
@@ -203,7 +203,7 @@ for opt in "$@"; do
# The paths in file_list are fixed outside of the loop.
file_list[${#file_list[@]}]="$opt"
case "$opt" in
*.asm|*.s) uses_asm=true
*.asm|*.[Ss]) uses_asm=true
;;
esac
;;
@@ -452,7 +452,7 @@ generate_vcxproj() {
done
open_tag ItemGroup
generate_filter "Source Files" "c;cc;cpp;def;odl;idl;hpj;bat;asm;asmx;s"
generate_filter "Source Files" "c;cc;cpp;def;odl;idl;hpj;bat;asm;asmx;s;S"
close_tag ItemGroup
open_tag ItemGroup
generate_filter "Header Files" "h;hm;inl;inc;xsd"

View File

@@ -384,13 +384,8 @@ if ($opts{arch} eq 'x86') {
}
close CONFIG_FILE;
mips;
} elsif ($opts{arch} eq 'armv6') {
@ALL_ARCHS = filter(qw/media/);
arm;
} elsif ($opts{arch} =~ /armv7\w?/) {
@ALL_ARCHS = filter(qw/media neon_asm neon/);
@REQUIRES = filter(keys %required ? keys %required : qw/media/);
&require(@REQUIRES);
@ALL_ARCHS = filter(qw/neon_asm neon/);
arm;
} elsif ($opts{arch} eq 'armv8' || $opts{arch} eq 'arm64' ) {
@ALL_ARCHS = filter(qw/neon/);

View File

@@ -2,3 +2,4 @@
GERRIT_HOST: chromium-review.googlesource.com
GERRIT_PORT: 29418
CODE_REVIEW_SERVER: chromium-review.googlesource.com
GERRIT_SQUASH_UPLOADS: False

60
configure vendored
View File

@@ -22,6 +22,7 @@ show_help(){
Advanced options:
${toggle_libs} libraries
${toggle_examples} examples
${toggle_tools} tools
${toggle_docs} documentation
${toggle_unit_tests} unit tests
${toggle_decode_perf_tests} build decoder perf tests with unit tests
@@ -97,11 +98,9 @@ EOF
# all_platforms is a list of all supported target platforms. Maintain
# alphabetically by architecture, generic-gnu last.
all_platforms="${all_platforms} arm64-android-gcc"
all_platforms="${all_platforms} arm64-darwin-gcc"
all_platforms="${all_platforms} arm64-linux-gcc"
all_platforms="${all_platforms} armv6-linux-rvct"
all_platforms="${all_platforms} armv6-linux-gcc"
all_platforms="${all_platforms} armv6-none-rvct"
all_platforms="${all_platforms} armv7-android-gcc" #neon Cortex-A8
all_platforms="${all_platforms} armv7-darwin-gcc" #neon Cortex-A8
all_platforms="${all_platforms} armv7-linux-rvct" #neon Cortex-A8
@@ -132,9 +131,6 @@ all_platforms="${all_platforms} x86-linux-icc"
all_platforms="${all_platforms} x86-os2-gcc"
all_platforms="${all_platforms} x86-solaris-gcc"
all_platforms="${all_platforms} x86-win32-gcc"
all_platforms="${all_platforms} x86-win32-vs7"
all_platforms="${all_platforms} x86-win32-vs8"
all_platforms="${all_platforms} x86-win32-vs9"
all_platforms="${all_platforms} x86-win32-vs10"
all_platforms="${all_platforms} x86-win32-vs11"
all_platforms="${all_platforms} x86-win32-vs12"
@@ -152,8 +148,6 @@ all_platforms="${all_platforms} x86_64-linux-gcc"
all_platforms="${all_platforms} x86_64-linux-icc"
all_platforms="${all_platforms} x86_64-solaris-gcc"
all_platforms="${all_platforms} x86_64-win64-gcc"
all_platforms="${all_platforms} x86_64-win64-vs8"
all_platforms="${all_platforms} x86_64-win64-vs9"
all_platforms="${all_platforms} x86_64-win64-vs10"
all_platforms="${all_platforms} x86_64-win64-vs11"
all_platforms="${all_platforms} x86_64-win64-vs12"
@@ -162,7 +156,7 @@ all_platforms="${all_platforms} generic-gnu"
# all_targets is a list of all targets that can be configured
# note that these should be in dependency order for now.
all_targets="libs examples docs"
all_targets="libs examples tools docs"
# all targets available are enabled, by default.
for t in ${all_targets}; do
@@ -241,8 +235,6 @@ ARCH_EXT_LIST_X86="
avx2
"
ARCH_EXT_LIST="
edsp
media
neon
neon_asm
@@ -272,7 +264,6 @@ CONFIG_LIST="
install_bins
install_libs
install_srcs
use_x86inc
debug
gprof
gcov
@@ -334,7 +325,6 @@ CMDLINE_SELECT="
gprof
gcov
pic
use_x86inc
optimizations
ccache
runtime_cpu_detect
@@ -342,6 +332,7 @@ CMDLINE_SELECT="
libs
examples
tools
docs
libc
as
@@ -487,7 +478,7 @@ EOF
#
# Write makefiles for all enabled targets
#
for tgt in libs examples docs solution; do
for tgt in libs examples tools docs solution; do
tgt_fn="$tgt-$toolchain.mk"
if enabled $tgt; then
@@ -578,26 +569,29 @@ process_toolchain() {
check_add_cflags -Wall
check_add_cflags -Wdeclaration-after-statement
check_add_cflags -Wdisabled-optimization
check_add_cflags -Wfloat-conversion
check_add_cflags -Wpointer-arith
check_add_cflags -Wtype-limits
check_add_cflags -Wcast-qual
check_add_cflags -Wvla
check_add_cflags -Wimplicit-function-declaration
check_add_cflags -Wuninitialized
check_add_cflags -Wunused-variable
case ${CC} in
*clang*)
# libvpx and/or clang have issues with aliasing:
# https://code.google.com/p/webm/issues/detail?id=603
# work around them until they are fixed
check_add_cflags -fno-strict-aliasing
;;
*) check_add_cflags -Wunused-but-set-variable ;;
esac
check_add_cflags -Wunused
# -Wextra has some tricky cases. Rather than fix them all now, get the
# flag for as many files as possible and fix the remaining issues
# piecemeal.
# https://bugs.chromium.org/p/webm/issues/detail?id=1069
check_add_cflags -Wextra
# check_add_cflags also adds to cxxflags. gtest does not do well with
# -Wundef so add it explicitly to CFLAGS only.
check_cflags -Wundef && add_cflags_only -Wundef
if enabled mips || [ -z "${INLINE}" ]; then
enabled extra_warnings || check_add_cflags -Wno-unused-function
else
check_add_cflags -Wunused-function
fi
if ! enabled vp9_highbitdepth; then
# Avoid this warning for third_party C++ sources. Some reorganization
# would be needed to apply this only to test/*.cc.
check_cflags -Wshorten-64-to-32 && add_cflags_only -Wshorten-64-to-32
fi
fi
@@ -646,17 +640,9 @@ process_toolchain() {
vs*) enable_feature msvs
enable_feature solution
vs_version=${tgt_cc##vs}
case $vs_version in
[789])
VCPROJ_SFX=vcproj
gen_vcproj_cmd=${source_path}/build/make/gen_msvs_proj.sh
;;
10|11|12|14)
VCPROJ_SFX=vcxproj
gen_vcproj_cmd=${source_path}/build/make/gen_msvs_vcxproj.sh
enabled werror && gen_vcproj_cmd="${gen_vcproj_cmd} --enable-werror"
;;
esac
VCPROJ_SFX=vcxproj
gen_vcproj_cmd=${source_path}/build/make/gen_msvs_vcxproj.sh
enabled werror && gen_vcproj_cmd="${gen_vcproj_cmd} --enable-werror"
all_targets="${all_targets} solution"
INLINE="__forceinline"
;;

View File

@@ -76,6 +76,7 @@ vpxdec.SRCS += tools_common.c tools_common.h
vpxdec.SRCS += y4menc.c y4menc.h
ifeq ($(CONFIG_LIBYUV),yes)
vpxdec.SRCS += $(LIBYUV_SRCS)
$(BUILD_PFX)third_party/libyuv/%.cc.o: CXXFLAGS += -Wno-unused-parameter
endif
ifeq ($(CONFIG_WEBM_IO),yes)
vpxdec.SRCS += $(LIBWEBM_COMMON_SRCS)
@@ -215,6 +216,17 @@ vp8cx_set_ref.SRCS += vpx_ports/msvc.h
vp8cx_set_ref.GUID = C5E31F7F-96F6-48BD-BD3E-10EBF6E8057A
vp8cx_set_ref.DESCRIPTION = VP8 set encoder reference frame
ifeq ($(CONFIG_VP9_ENCODER),yes)
ifeq ($(CONFIG_DECODERS),yes)
EXAMPLES-yes += vp9cx_set_ref.c
vp9cx_set_ref.SRCS += ivfenc.h ivfenc.c
vp9cx_set_ref.SRCS += tools_common.h tools_common.c
vp9cx_set_ref.SRCS += video_common.h
vp9cx_set_ref.SRCS += video_writer.h video_writer.c
vp9cx_set_ref.GUID = 65D7F14A-2EE6-4293-B958-AB5107A03B55
vp9cx_set_ref.DESCRIPTION = VP9 set encoder reference frame
endif
endif
ifeq ($(CONFIG_MULTI_RES_ENCODING),yes)
ifeq ($(CONFIG_LIBYUV),yes)

View File

@@ -65,8 +65,7 @@ static void get_image_md5(const vpx_image_t *img, unsigned char digest[16]) {
static void print_md5(FILE *stream, unsigned char digest[16]) {
int i;
for (i = 0; i < 16; ++i)
fprintf(stream, "%02x", digest[i]);
for (i = 0; i < 16; ++i) fprintf(stream, "%02x", digest[i]);
}
static const char *exec_name;
@@ -86,12 +85,10 @@ int main(int argc, char **argv) {
exec_name = argv[0];
if (argc != 3)
die("Invalid number of arguments.");
if (argc != 3) die("Invalid number of arguments.");
reader = vpx_video_reader_open(argv[1]);
if (!reader)
die("Failed to open %s for reading.", argv[1]);
if (!reader) die("Failed to open %s for reading.", argv[1]);
if (!(outfile = fopen(argv[2], "wb")))
die("Failed to open %s for writing.", argv[2]);
@@ -99,8 +96,7 @@ int main(int argc, char **argv) {
info = vpx_video_reader_get_info(reader);
decoder = get_vpx_decoder_by_fourcc(info->codec_fourcc);
if (!decoder)
die("Unknown input codec.");
if (!decoder) die("Unknown input codec.");
printf("Using %s\n", vpx_codec_iface_name(decoder->codec_interface()));
@@ -111,8 +107,8 @@ int main(int argc, char **argv) {
vpx_codec_iter_t iter = NULL;
vpx_image_t *img = NULL;
size_t frame_size = 0;
const unsigned char *frame = vpx_video_reader_get_frame(reader,
&frame_size);
const unsigned char *frame =
vpx_video_reader_get_frame(reader, &frame_size);
if (vpx_codec_decode(&codec, frame, (unsigned int)frame_size, NULL, 0))
die_codec(&codec, "Failed to decode frame");
@@ -121,14 +117,13 @@ int main(int argc, char **argv) {
get_image_md5(img, digest);
print_md5(outfile, digest);
fprintf(outfile, " img-%dx%d-%04d.i420\n",
img->d_w, img->d_h, ++frame_cnt);
fprintf(outfile, " img-%dx%d-%04d.i420\n", img->d_w, img->d_h,
++frame_cnt);
}
}
printf("Processed %d frames.\n", frame_cnt);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_reader_close(reader);

View File

@@ -84,18 +84,16 @@ int main(int argc, char **argv) {
exec_name = argv[0];
if (argc != 4)
die("Invalid number of arguments.");
if (argc != 4) die("Invalid number of arguments.");
reader = vpx_video_reader_open(argv[1]);
if (!reader)
die("Failed to open %s for reading.", argv[1]);
if (!reader) die("Failed to open %s for reading.", argv[1]);
if (!(outfile = fopen(argv[2], "wb")))
die("Failed to open %s for writing.", argv[2]);
n = strtol(argv[3], &nptr, 0);
m = strtol(nptr + 1, NULL, 0);
n = (int)strtol(argv[3], &nptr, 0);
m = (int)strtol(nptr + 1, NULL, 0);
is_range = (*nptr == '-');
if (!n || !m || (*nptr != '-' && *nptr != '/'))
die("Couldn't parse pattern %s.\n", argv[3]);
@@ -103,8 +101,7 @@ int main(int argc, char **argv) {
info = vpx_video_reader_get_info(reader);
decoder = get_vpx_decoder_by_fourcc(info->codec_fourcc);
if (!decoder)
die("Unknown input codec.");
if (!decoder) die("Unknown input codec.");
printf("Using %s\n", vpx_codec_iface_name(decoder->codec_interface()));
@@ -116,8 +113,8 @@ int main(int argc, char **argv) {
vpx_image_t *img = NULL;
size_t frame_size = 0;
int skip;
const unsigned char *frame = vpx_video_reader_get_frame(reader,
&frame_size);
const unsigned char *frame =
vpx_video_reader_get_frame(reader, &frame_size);
if (vpx_codec_decode(&codec, frame, (unsigned int)frame_size, NULL, 0))
die_codec(&codec, "Failed to decode frame.");
@@ -139,8 +136,7 @@ int main(int argc, char **argv) {
}
printf("Processed %d frames.\n", frame_cnt);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
printf("Play: ffplay -f rawvideo -pix_fmt yuv420p -s %dx%d %s\n",
info->frame_width, info->frame_height, argv[2]);

View File

@@ -68,12 +68,10 @@ int main(int argc, char **argv) {
exec_name = argv[0];
if (argc != 3)
die("Invalid number of arguments.");
if (argc != 3) die("Invalid number of arguments.");
reader = vpx_video_reader_open(argv[1]);
if (!reader)
die("Failed to open %s for reading.", argv[1]);
if (!reader) die("Failed to open %s for reading.", argv[1]);
if (!(outfile = fopen(argv[2], "wb")))
die("Failed to open %s for writing", argv[2]);
@@ -81,8 +79,7 @@ int main(int argc, char **argv) {
info = vpx_video_reader_get_info(reader);
decoder = get_vpx_decoder_by_fourcc(info->codec_fourcc);
if (!decoder)
die("Unknown input codec.");
if (!decoder) die("Unknown input codec.");
printf("Using %s\n", vpx_codec_iface_name(decoder->codec_interface()));
@@ -91,26 +88,25 @@ int main(int argc, char **argv) {
if (res == VPX_CODEC_INCAPABLE)
die_codec(&codec, "Postproc not supported by this decoder.");
if (res)
die_codec(&codec, "Failed to initialize decoder.");
if (res) die_codec(&codec, "Failed to initialize decoder.");
while (vpx_video_reader_read_frame(reader)) {
vpx_codec_iter_t iter = NULL;
vpx_image_t *img = NULL;
size_t frame_size = 0;
const unsigned char *frame = vpx_video_reader_get_frame(reader,
&frame_size);
const unsigned char *frame =
vpx_video_reader_get_frame(reader, &frame_size);
++frame_cnt;
if (frame_cnt % 30 == 1) {
vp8_postproc_cfg_t pp = {0, 0, 0};
vp8_postproc_cfg_t pp = { 0, 0, 0 };
if (vpx_codec_control(&codec, VP8_SET_POSTPROC, &pp))
die_codec(&codec, "Failed to turn off postproc.");
if (vpx_codec_control(&codec, VP8_SET_POSTPROC, &pp))
die_codec(&codec, "Failed to turn off postproc.");
} else if (frame_cnt % 30 == 16) {
vp8_postproc_cfg_t pp = {VP8_DEBLOCK | VP8_DEMACROBLOCK | VP8_MFQE,
4, 0};
vp8_postproc_cfg_t pp = { VP8_DEBLOCK | VP8_DEMACROBLOCK | VP8_MFQE, 4,
0 };
if (vpx_codec_control(&codec, VP8_SET_POSTPROC, &pp))
die_codec(&codec, "Failed to turn on postproc.");
};
@@ -125,8 +121,7 @@ int main(int argc, char **argv) {
}
printf("Processed %d frames.\n", frame_cnt);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec");
printf("Play: ffplay -f rawvideo -pix_fmt yuv420p -s %dx%d %s\n",
info->frame_width, info->frame_height, argv[2]);

View File

@@ -34,10 +34,8 @@ void usage_exit(void) {
static int parse_dim(char *v, int *width, int *height) {
char *x = strchr(v, 'x');
if (x == NULL)
x = strchr(v, 'X');
if (x == NULL)
return 0;
if (x == NULL) x = strchr(v, 'X');
if (x == NULL) return 0;
*width = atoi(v);
*height = atoi(&x[1]);
if (*width <= 0 || *height <= 0)
@@ -93,30 +91,25 @@ int main(int argc, char *argv[]) {
else
frames = INT_MAX;
printf("Input size: %dx%d\n",
width, height);
printf("Target size: %dx%d, Frames: ",
target_width, target_height);
printf("Input size: %dx%d\n", width, height);
printf("Target size: %dx%d, Frames: ", target_width, target_height);
if (frames == INT_MAX)
printf("All\n");
else
printf("%d\n", frames);
inbuf = (uint8_t*)malloc(width * height * 3 / 2);
outbuf = (uint8_t*)malloc(target_width * target_height * 3 / 2);
inbuf = (uint8_t *)malloc(width * height * 3 / 2);
outbuf = (uint8_t *)malloc(target_width * target_height * 3 / 2);
inbuf_u = inbuf + width * height;
inbuf_v = inbuf_u + width * height / 4;
outbuf_u = outbuf + target_width * target_height;
outbuf_v = outbuf_u + target_width * target_height / 4;
f = 0;
while (f < frames) {
if (fread(inbuf, width * height * 3 / 2, 1, fpin) != 1)
break;
vp9_resize_frame420(inbuf, width, inbuf_u, inbuf_v, width / 2,
height, width,
outbuf, target_width, outbuf_u, outbuf_v,
target_width / 2,
target_height, target_width);
if (fread(inbuf, width * height * 3 / 2, 1, fpin) != 1) break;
vp9_resize_frame420(inbuf, width, inbuf_u, inbuf_v, width / 2, height,
width, outbuf, target_width, outbuf_u, outbuf_v,
target_width / 2, target_height, target_width);
fwrite(outbuf, target_width * target_height * 3 / 2, 1, fpout);
f++;
}

View File

@@ -8,7 +8,6 @@
* be found in the AUTHORS file in the root of the source tree.
*/
// VP8 Set Active and ROI Maps
// ===========================
//
@@ -86,8 +85,7 @@ static void set_roi_map(const vpx_codec_enc_cfg_t *cfg,
roi.static_threshold[3] = 0;
roi.roi_map = (uint8_t *)malloc(roi.rows * roi.cols);
for (i = 0; i < roi.rows * roi.cols; ++i)
roi.roi_map[i] = i % 4;
for (i = 0; i < roi.rows * roi.cols; ++i) roi.roi_map[i] = i % 4;
if (vpx_codec_control(codec, VP8E_SET_ROI_MAP, &roi))
die_codec(codec, "Failed to set ROI map");
@@ -98,14 +96,13 @@ static void set_roi_map(const vpx_codec_enc_cfg_t *cfg,
static void set_active_map(const vpx_codec_enc_cfg_t *cfg,
vpx_codec_ctx_t *codec) {
unsigned int i;
vpx_active_map_t map = {0, 0, 0};
vpx_active_map_t map = { 0, 0, 0 };
map.rows = (cfg->g_h + 15) / 16;
map.cols = (cfg->g_w + 15) / 16;
map.active_map = (uint8_t *)malloc(map.rows * map.cols);
for (i = 0; i < map.rows * map.cols; ++i)
map.active_map[i] = i % 2;
for (i = 0; i < map.rows * map.cols; ++i) map.active_map[i] = i % 2;
if (vpx_codec_control(codec, VP8E_SET_ACTIVEMAP, &map))
die_codec(codec, "Failed to set active map");
@@ -115,7 +112,7 @@ static void set_active_map(const vpx_codec_enc_cfg_t *cfg,
static void unset_active_map(const vpx_codec_enc_cfg_t *cfg,
vpx_codec_ctx_t *codec) {
vpx_active_map_t map = {0, 0, 0};
vpx_active_map_t map = { 0, 0, 0 };
map.rows = (cfg->g_h + 15) / 16;
map.cols = (cfg->g_w + 15) / 16;
@@ -125,25 +122,21 @@ static void unset_active_map(const vpx_codec_enc_cfg_t *cfg,
die_codec(codec, "Failed to set active map");
}
static int encode_frame(vpx_codec_ctx_t *codec,
vpx_image_t *img,
int frame_index,
VpxVideoWriter *writer) {
static int encode_frame(vpx_codec_ctx_t *codec, vpx_image_t *img,
int frame_index, VpxVideoWriter *writer) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(codec, img, frame_index, 1, 0,
VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK)
die_codec(codec, "Failed to encode frame");
const vpx_codec_err_t res =
vpx_codec_encode(codec, img, frame_index, 1, 0, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK) die_codec(codec, "Failed to encode frame");
while ((pkt = vpx_codec_get_cx_data(codec, &iter)) != NULL) {
got_pkts = 1;
if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!vpx_video_writer_write_frame(writer,
pkt->data.frame.buf,
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts)) {
die_codec(codec, "Failed to write compressed frame");
@@ -167,12 +160,11 @@ int main(int argc, char **argv) {
VpxVideoInfo info;
VpxVideoWriter *writer = NULL;
const VpxInterface *encoder = NULL;
const int fps = 2; // TODO(dkovalev) add command line argument
const int fps = 2; // TODO(dkovalev) add command line argument
const double bits_per_pixel_per_frame = 0.067;
exec_name = argv[0];
if (argc != 6)
die("Invalid number of arguments");
if (argc != 6) die("Invalid number of arguments");
memset(&info, 0, sizeof(info));
@@ -182,40 +174,36 @@ int main(int argc, char **argv) {
}
assert(encoder != NULL);
info.codec_fourcc = encoder->fourcc;
info.frame_width = strtol(argv[2], NULL, 0);
info.frame_height = strtol(argv[3], NULL, 0);
info.frame_width = (int)strtol(argv[2], NULL, 0);
info.frame_height = (int)strtol(argv[3], NULL, 0);
info.time_base.numerator = 1;
info.time_base.denominator = fps;
if (info.frame_width <= 0 ||
info.frame_height <= 0 ||
(info.frame_width % 2) != 0 ||
(info.frame_height % 2) != 0) {
if (info.frame_width <= 0 || info.frame_height <= 0 ||
(info.frame_width % 2) != 0 || (info.frame_height % 2) != 0) {
die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
}
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
info.frame_height, 1)) {
info.frame_height, 1)) {
die("Failed to allocate image.");
}
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res)
die_codec(&codec, "Failed to get default codec config.");
if (res) die_codec(&codec, "Failed to get default codec config.");
cfg.g_w = info.frame_width;
cfg.g_h = info.frame_height;
cfg.g_timebase.num = info.time_base.numerator;
cfg.g_timebase.den = info.time_base.denominator;
cfg.rc_target_bitrate = (unsigned int)(bits_per_pixel_per_frame * cfg.g_w *
cfg.g_h * fps / 1000);
cfg.rc_target_bitrate =
(unsigned int)(bits_per_pixel_per_frame * cfg.g_w * cfg.g_h * fps / 1000);
cfg.g_lag_in_frames = 0;
writer = vpx_video_writer_open(argv[5], kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing.", argv[5]);
if (!writer) die("Failed to open %s for writing.", argv[5]);
if (!(infile = fopen(argv[4], "rb")))
die("Failed to open %s for reading.", argv[4]);
@@ -239,15 +227,15 @@ int main(int argc, char **argv) {
}
// Flush encoder.
while (encode_frame(&codec, NULL, -1, writer)) {}
while (encode_frame(&codec, NULL, -1, writer)) {
}
printf("\n");
fclose(infile);
printf("Processed %d frames.\n", frame_count);
vpx_img_free(&raw);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_writer_close(writer);

View File

@@ -8,7 +8,6 @@
* be found in the AUTHORS file in the root of the source tree.
*/
// Simple Decoder
// ==============
//
@@ -103,12 +102,10 @@ int main(int argc, char **argv) {
exec_name = argv[0];
if (argc != 3)
die("Invalid number of arguments.");
if (argc != 3) die("Invalid number of arguments.");
reader = vpx_video_reader_open(argv[1]);
if (!reader)
die("Failed to open %s for reading.", argv[1]);
if (!reader) die("Failed to open %s for reading.", argv[1]);
if (!(outfile = fopen(argv[2], "wb")))
die("Failed to open %s for writing.", argv[2]);
@@ -116,8 +113,7 @@ int main(int argc, char **argv) {
info = vpx_video_reader_get_info(reader);
decoder = get_vpx_decoder_by_fourcc(info->codec_fourcc);
if (!decoder)
die("Unknown input codec.");
if (!decoder) die("Unknown input codec.");
printf("Using %s\n", vpx_codec_iface_name(decoder->codec_interface()));
@@ -128,8 +124,8 @@ int main(int argc, char **argv) {
vpx_codec_iter_t iter = NULL;
vpx_image_t *img = NULL;
size_t frame_size = 0;
const unsigned char *frame = vpx_video_reader_get_frame(reader,
&frame_size);
const unsigned char *frame =
vpx_video_reader_get_frame(reader, &frame_size);
if (vpx_codec_decode(&codec, frame, (unsigned int)frame_size, NULL, 0))
die_codec(&codec, "Failed to decode frame.");
@@ -140,8 +136,7 @@ int main(int argc, char **argv) {
}
printf("Processed %d frames.\n", frame_cnt);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec");
printf("Play: ffplay -f rawvideo -pix_fmt yuv420p -s %dx%d %s\n",
info->frame_width, info->frame_height, argv[2]);

View File

@@ -109,32 +109,27 @@ static const char *exec_name;
void usage_exit(void) {
fprintf(stderr,
"Usage: %s <codec> <width> <height> <infile> <outfile> "
"<keyframe-interval> <error-resilient> <frames to encode>\n"
"See comments in simple_encoder.c for more information.\n",
"<keyframe-interval> <error-resilient> <frames to encode>\n"
"See comments in simple_encoder.c for more information.\n",
exec_name);
exit(EXIT_FAILURE);
}
static int encode_frame(vpx_codec_ctx_t *codec,
vpx_image_t *img,
int frame_index,
int flags,
VpxVideoWriter *writer) {
static int encode_frame(vpx_codec_ctx_t *codec, vpx_image_t *img,
int frame_index, int flags, VpxVideoWriter *writer) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(codec, img, frame_index, 1,
flags, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK)
die_codec(codec, "Failed to encode frame");
const vpx_codec_err_t res =
vpx_codec_encode(codec, img, frame_index, 1, flags, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK) die_codec(codec, "Failed to encode frame");
while ((pkt = vpx_codec_get_cx_data(codec, &iter)) != NULL) {
got_pkts = 1;
if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!vpx_video_writer_write_frame(writer,
pkt->data.frame.buf,
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts)) {
die_codec(codec, "Failed to write compressed frame");
@@ -155,7 +150,7 @@ int main(int argc, char **argv) {
int frame_count = 0;
vpx_image_t raw;
vpx_codec_err_t res;
VpxVideoInfo info = {0};
VpxVideoInfo info = { 0, 0, 0, { 0, 0 } };
VpxVideoWriter *writer = NULL;
const VpxInterface *encoder = NULL;
const int fps = 30;
@@ -172,8 +167,7 @@ int main(int argc, char **argv) {
exec_name = argv[0];
if (argc != 9)
die("Invalid number of arguments");
if (argc != 9) die("Invalid number of arguments");
codec_arg = argv[1];
width_arg = argv[2];
@@ -181,50 +175,44 @@ int main(int argc, char **argv) {
infile_arg = argv[4];
outfile_arg = argv[5];
keyframe_interval_arg = argv[6];
max_frames = strtol(argv[8], NULL, 0);
max_frames = (int)strtol(argv[8], NULL, 0);
encoder = get_vpx_encoder_by_name(codec_arg);
if (!encoder)
die("Unsupported codec.");
if (!encoder) die("Unsupported codec.");
info.codec_fourcc = encoder->fourcc;
info.frame_width = strtol(width_arg, NULL, 0);
info.frame_height = strtol(height_arg, NULL, 0);
info.frame_width = (int)strtol(width_arg, NULL, 0);
info.frame_height = (int)strtol(height_arg, NULL, 0);
info.time_base.numerator = 1;
info.time_base.denominator = fps;
if (info.frame_width <= 0 ||
info.frame_height <= 0 ||
(info.frame_width % 2) != 0 ||
(info.frame_height % 2) != 0) {
if (info.frame_width <= 0 || info.frame_height <= 0 ||
(info.frame_width % 2) != 0 || (info.frame_height % 2) != 0) {
die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
}
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
info.frame_height, 1)) {
info.frame_height, 1)) {
die("Failed to allocate image.");
}
keyframe_interval = strtol(keyframe_interval_arg, NULL, 0);
if (keyframe_interval < 0)
die("Invalid keyframe interval value.");
keyframe_interval = (int)strtol(keyframe_interval_arg, NULL, 0);
if (keyframe_interval < 0) die("Invalid keyframe interval value.");
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res)
die_codec(&codec, "Failed to get default codec config.");
if (res) die_codec(&codec, "Failed to get default codec config.");
cfg.g_w = info.frame_width;
cfg.g_h = info.frame_height;
cfg.g_timebase.num = info.time_base.numerator;
cfg.g_timebase.den = info.time_base.denominator;
cfg.rc_target_bitrate = bitrate;
cfg.g_error_resilient = strtol(argv[7], NULL, 0);
cfg.g_error_resilient = (vpx_codec_er_flags_t)strtoul(argv[7], NULL, 0);
writer = vpx_video_writer_open(outfile_arg, kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing.", outfile_arg);
if (!writer) die("Failed to open %s for writing.", outfile_arg);
if (!(infile = fopen(infile_arg, "rb")))
die("Failed to open %s for reading.", infile_arg);
@@ -239,20 +227,19 @@ int main(int argc, char **argv) {
flags |= VPX_EFLAG_FORCE_KF;
encode_frame(&codec, &raw, frame_count++, flags, writer);
frames_encoded++;
if (max_frames > 0 && frames_encoded >= max_frames)
break;
if (max_frames > 0 && frames_encoded >= max_frames) break;
}
// Flush encoder.
while (encode_frame(&codec, NULL, -1, 0, writer)) {};
while (encode_frame(&codec, NULL, -1, 0, writer)) {
}
printf("\n");
fclose(infile);
printf("Processed %d frames.\n", frame_count);
vpx_img_free(&raw);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_writer_close(writer);

View File

@@ -61,25 +61,21 @@ static const char *exec_name;
void usage_exit(void) {
fprintf(stderr,
"Usage: %s <codec> <width> <height> <infile> <outfile> "
"<frame limit>\n",
"<frame limit>\n",
exec_name);
exit(EXIT_FAILURE);
}
static int get_frame_stats(vpx_codec_ctx_t *ctx,
const vpx_image_t *img,
vpx_codec_pts_t pts,
unsigned int duration,
vpx_enc_frame_flags_t flags,
unsigned int deadline,
static int get_frame_stats(vpx_codec_ctx_t *ctx, const vpx_image_t *img,
vpx_codec_pts_t pts, unsigned int duration,
vpx_enc_frame_flags_t flags, unsigned int deadline,
vpx_fixed_buf_t *stats) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(ctx, img, pts, duration, flags,
deadline);
if (res != VPX_CODEC_OK)
die_codec(ctx, "Failed to get frame stats.");
const vpx_codec_err_t res =
vpx_codec_encode(ctx, img, pts, duration, flags, deadline);
if (res != VPX_CODEC_OK) die_codec(ctx, "Failed to get frame stats.");
while ((pkt = vpx_codec_get_cx_data(ctx, &iter)) != NULL) {
got_pkts = 1;
@@ -96,20 +92,16 @@ static int get_frame_stats(vpx_codec_ctx_t *ctx,
return got_pkts;
}
static int encode_frame(vpx_codec_ctx_t *ctx,
const vpx_image_t *img,
vpx_codec_pts_t pts,
unsigned int duration,
vpx_enc_frame_flags_t flags,
unsigned int deadline,
static int encode_frame(vpx_codec_ctx_t *ctx, const vpx_image_t *img,
vpx_codec_pts_t pts, unsigned int duration,
vpx_enc_frame_flags_t flags, unsigned int deadline,
VpxVideoWriter *writer) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(ctx, img, pts, duration, flags,
deadline);
if (res != VPX_CODEC_OK)
die_codec(ctx, "Failed to encode frame.");
const vpx_codec_err_t res =
vpx_codec_encode(ctx, img, pts, duration, flags, deadline);
if (res != VPX_CODEC_OK) die_codec(ctx, "Failed to encode frame.");
while ((pkt = vpx_codec_get_cx_data(ctx, &iter)) != NULL) {
got_pkts = 1;
@@ -117,8 +109,8 @@ static int encode_frame(vpx_codec_ctx_t *ctx,
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts))
pkt->data.frame.sz,
pkt->data.frame.pts))
die_codec(ctx, "Failed to write compressed frame.");
printf(keyframe ? "K" : ".");
fflush(stdout);
@@ -128,14 +120,12 @@ static int encode_frame(vpx_codec_ctx_t *ctx,
return got_pkts;
}
static vpx_fixed_buf_t pass0(vpx_image_t *raw,
FILE *infile,
static vpx_fixed_buf_t pass0(vpx_image_t *raw, FILE *infile,
const VpxInterface *encoder,
const vpx_codec_enc_cfg_t *cfg,
int max_frames) {
const vpx_codec_enc_cfg_t *cfg, int max_frames) {
vpx_codec_ctx_t codec;
int frame_count = 0;
vpx_fixed_buf_t stats = {NULL, 0};
vpx_fixed_buf_t stats = { NULL, 0 };
if (vpx_codec_enc_init(&codec, encoder->codec_interface(), cfg, 0))
die_codec(&codec, "Failed to initialize encoder");
@@ -145,40 +135,33 @@ static vpx_fixed_buf_t pass0(vpx_image_t *raw,
++frame_count;
get_frame_stats(&codec, raw, frame_count, 1, 0, VPX_DL_GOOD_QUALITY,
&stats);
if (max_frames > 0 && frame_count >= max_frames)
break;
if (max_frames > 0 && frame_count >= max_frames) break;
}
// Flush encoder.
while (get_frame_stats(&codec, NULL, frame_count, 1, 0,
VPX_DL_GOOD_QUALITY, &stats)) {}
while (get_frame_stats(&codec, NULL, frame_count, 1, 0, VPX_DL_GOOD_QUALITY,
&stats)) {
}
printf("Pass 0 complete. Processed %d frames.\n", frame_count);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
return stats;
}
static void pass1(vpx_image_t *raw,
FILE *infile,
const char *outfile_name,
const VpxInterface *encoder,
const vpx_codec_enc_cfg_t *cfg,
static void pass1(vpx_image_t *raw, FILE *infile, const char *outfile_name,
const VpxInterface *encoder, const vpx_codec_enc_cfg_t *cfg,
int max_frames) {
VpxVideoInfo info = {
encoder->fourcc,
cfg->g_w,
cfg->g_h,
{cfg->g_timebase.num, cfg->g_timebase.den}
};
VpxVideoInfo info = { encoder->fourcc,
cfg->g_w,
cfg->g_h,
{ cfg->g_timebase.num, cfg->g_timebase.den } };
VpxVideoWriter *writer = NULL;
vpx_codec_ctx_t codec;
int frame_count = 0;
writer = vpx_video_writer_open(outfile_name, kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing", outfile_name);
if (!writer) die("Failed to open %s for writing", outfile_name);
if (vpx_codec_enc_init(&codec, encoder->codec_interface(), cfg, 0))
die_codec(&codec, "Failed to initialize encoder");
@@ -188,17 +171,16 @@ static void pass1(vpx_image_t *raw,
++frame_count;
encode_frame(&codec, raw, frame_count, 1, 0, VPX_DL_GOOD_QUALITY, writer);
if (max_frames > 0 && frame_count >= max_frames)
break;
if (max_frames > 0 && frame_count >= max_frames) break;
}
// Flush encoder.
while (encode_frame(&codec, NULL, -1, 1, 0, VPX_DL_GOOD_QUALITY, writer)) {}
while (encode_frame(&codec, NULL, -1, 1, 0, VPX_DL_GOOD_QUALITY, writer)) {
}
printf("\n");
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_writer_close(writer);
@@ -215,8 +197,8 @@ int main(int argc, char **argv) {
vpx_fixed_buf_t stats;
const VpxInterface *encoder = NULL;
const int fps = 30; // TODO(dkovalev) add command line argument
const int bitrate = 200; // kbit/s TODO(dkovalev) add command line argument
const int fps = 30; // TODO(dkovalev) add command line argument
const int bitrate = 200; // kbit/s TODO(dkovalev) add command line argument
const char *const codec_arg = argv[1];
const char *const width_arg = argv[2];
const char *const height_arg = argv[3];
@@ -225,19 +207,17 @@ int main(int argc, char **argv) {
int max_frames = 0;
exec_name = argv[0];
if (argc != 7)
die("Invalid number of arguments.");
if (argc != 7) die("Invalid number of arguments.");
max_frames = strtol(argv[6], NULL, 0);
max_frames = (int)strtol(argv[6], NULL, 0);
encoder = get_vpx_encoder_by_name(codec_arg);
if (!encoder)
die("Unsupported codec.");
if (!encoder) die("Unsupported codec.");
w = strtol(width_arg, NULL, 0);
h = strtol(height_arg, NULL, 0);
w = (int)strtol(width_arg, NULL, 0);
h = (int)strtol(height_arg, NULL, 0);
if (w <= 0 || h <= 0 || (w % 2) != 0 || (h % 2) != 0)
if (w <= 0 || h <= 0 || (w % 2) != 0 || (h % 2) != 0)
die("Invalid frame size: %dx%d", w, h);
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, w, h, 1))
@@ -247,8 +227,7 @@ int main(int argc, char **argv) {
// Configuration
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res)
die_codec(&codec, "Failed to get default codec config.");
if (res) die_codec(&codec, "Failed to get default codec config.");
cfg.g_w = w;
cfg.g_h = h;

File diff suppressed because it is too large Load Diff

View File

@@ -8,7 +8,6 @@
* be found in the AUTHORS file in the root of the source tree.
*/
// VP8 Set Reference Frame
// =======================
//
@@ -52,6 +51,7 @@
#include "vpx/vp8cx.h"
#include "vpx/vpx_encoder.h"
#include "vp8/common/common.h"
#include "../tools_common.h"
#include "../video_writer.h"
@@ -64,25 +64,21 @@ void usage_exit(void) {
exit(EXIT_FAILURE);
}
static int encode_frame(vpx_codec_ctx_t *codec,
vpx_image_t *img,
int frame_index,
VpxVideoWriter *writer) {
static int encode_frame(vpx_codec_ctx_t *codec, vpx_image_t *img,
int frame_index, VpxVideoWriter *writer) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(codec, img, frame_index, 1, 0,
VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK)
die_codec(codec, "Failed to encode frame");
const vpx_codec_err_t res =
vpx_codec_encode(codec, img, frame_index, 1, 0, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK) die_codec(codec, "Failed to encode frame");
while ((pkt = vpx_codec_get_cx_data(codec, &iter)) != NULL) {
got_pkts = 1;
if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!vpx_video_writer_write_frame(writer,
pkt->data.frame.buf,
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts)) {
die_codec(codec, "Failed to write compressed frame");
@@ -98,55 +94,53 @@ static int encode_frame(vpx_codec_ctx_t *codec,
int main(int argc, char **argv) {
FILE *infile = NULL;
vpx_codec_ctx_t codec = {0};
vpx_codec_enc_cfg_t cfg = {0};
vpx_codec_ctx_t codec;
vpx_codec_enc_cfg_t cfg;
int frame_count = 0;
vpx_image_t raw;
vpx_codec_err_t res;
VpxVideoInfo info = {0};
VpxVideoInfo info;
VpxVideoWriter *writer = NULL;
const VpxInterface *encoder = NULL;
int update_frame_num = 0;
const int fps = 30; // TODO(dkovalev) add command line argument
const int bitrate = 200; // kbit/s TODO(dkovalev) add command line argument
const int fps = 30; // TODO(dkovalev) add command line argument
const int bitrate = 200; // kbit/s TODO(dkovalev) add command line argument
vp8_zero(codec);
vp8_zero(cfg);
vp8_zero(info);
exec_name = argv[0];
if (argc != 6)
die("Invalid number of arguments");
if (argc != 6) die("Invalid number of arguments");
// TODO(dkovalev): add vp9 support and rename the file accordingly
encoder = get_vpx_encoder_by_name("vp8");
if (!encoder)
die("Unsupported codec.");
if (!encoder) die("Unsupported codec.");
update_frame_num = atoi(argv[5]);
if (!update_frame_num)
die("Couldn't parse frame number '%s'\n", argv[5]);
if (!update_frame_num) die("Couldn't parse frame number '%s'\n", argv[5]);
info.codec_fourcc = encoder->fourcc;
info.frame_width = strtol(argv[1], NULL, 0);
info.frame_height = strtol(argv[2], NULL, 0);
info.frame_width = (int)strtol(argv[1], NULL, 0);
info.frame_height = (int)strtol(argv[2], NULL, 0);
info.time_base.numerator = 1;
info.time_base.denominator = fps;
if (info.frame_width <= 0 ||
info.frame_height <= 0 ||
(info.frame_width % 2) != 0 ||
(info.frame_height % 2) != 0) {
if (info.frame_width <= 0 || info.frame_height <= 0 ||
(info.frame_width % 2) != 0 || (info.frame_height % 2) != 0) {
die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
}
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
info.frame_height, 1)) {
info.frame_height, 1)) {
die("Failed to allocate image.");
}
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res)
die_codec(&codec, "Failed to get default codec config.");
if (res) die_codec(&codec, "Failed to get default codec config.");
cfg.g_w = info.frame_width;
cfg.g_h = info.frame_height;
@@ -155,8 +149,7 @@ int main(int argc, char **argv) {
cfg.rc_target_bitrate = bitrate;
writer = vpx_video_writer_open(argv[4], kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing.", argv[4]);
if (!writer) die("Failed to open %s for writing.", argv[4]);
if (!(infile = fopen(argv[3], "rb")))
die("Failed to open %s for reading.", argv[3]);
@@ -178,15 +171,15 @@ int main(int argc, char **argv) {
}
// Flush encoder.
while (encode_frame(&codec, NULL, -1, writer)) {}
while (encode_frame(&codec, NULL, -1, writer)) {
}
printf("\n");
fclose(infile);
printf("Processed %d frames.\n", frame_count);
vpx_img_free(&raw);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_writer_close(writer);

View File

@@ -14,6 +14,7 @@
#include "vpx/vpx_encoder.h"
#include "vpx/vp8cx.h"
#include "vp9/common/vp9_common.h"
#include "../tools_common.h"
#include "../video_writer.h"
@@ -21,32 +22,28 @@
static const char *exec_name;
void usage_exit(void) {
fprintf(stderr, "vp9_lossless_encoder: Example demonstrating VP9 lossless "
"encoding feature. Supports raw input only.\n");
fprintf(stderr,
"vp9_lossless_encoder: Example demonstrating VP9 lossless "
"encoding feature. Supports raw input only.\n");
fprintf(stderr, "Usage: %s <width> <height> <infile> <outfile>\n", exec_name);
exit(EXIT_FAILURE);
}
static int encode_frame(vpx_codec_ctx_t *codec,
vpx_image_t *img,
int frame_index,
int flags,
VpxVideoWriter *writer) {
static int encode_frame(vpx_codec_ctx_t *codec, vpx_image_t *img,
int frame_index, int flags, VpxVideoWriter *writer) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
const vpx_codec_err_t res = vpx_codec_encode(codec, img, frame_index, 1,
flags, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK)
die_codec(codec, "Failed to encode frame");
const vpx_codec_err_t res =
vpx_codec_encode(codec, img, frame_index, 1, flags, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK) die_codec(codec, "Failed to encode frame");
while ((pkt = vpx_codec_get_cx_data(codec, &iter)) != NULL) {
got_pkts = 1;
if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!vpx_video_writer_write_frame(writer,
pkt->data.frame.buf,
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts)) {
die_codec(codec, "Failed to write compressed frame");
@@ -66,43 +63,40 @@ int main(int argc, char **argv) {
int frame_count = 0;
vpx_image_t raw;
vpx_codec_err_t res;
VpxVideoInfo info = {0};
VpxVideoInfo info;
VpxVideoWriter *writer = NULL;
const VpxInterface *encoder = NULL;
const int fps = 30;
vp9_zero(info);
exec_name = argv[0];
if (argc < 5)
die("Invalid number of arguments");
if (argc < 5) die("Invalid number of arguments");
encoder = get_vpx_encoder_by_name("vp9");
if (!encoder)
die("Unsupported codec.");
if (!encoder) die("Unsupported codec.");
info.codec_fourcc = encoder->fourcc;
info.frame_width = strtol(argv[1], NULL, 0);
info.frame_height = strtol(argv[2], NULL, 0);
info.frame_width = (int)strtol(argv[1], NULL, 0);
info.frame_height = (int)strtol(argv[2], NULL, 0);
info.time_base.numerator = 1;
info.time_base.denominator = fps;
if (info.frame_width <= 0 ||
info.frame_height <= 0 ||
(info.frame_width % 2) != 0 ||
(info.frame_height % 2) != 0) {
if (info.frame_width <= 0 || info.frame_height <= 0 ||
(info.frame_width % 2) != 0 || (info.frame_height % 2) != 0) {
die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
}
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
info.frame_height, 1)) {
info.frame_height, 1)) {
die("Failed to allocate image.");
}
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res)
die_codec(&codec, "Failed to get default codec config.");
if (res) die_codec(&codec, "Failed to get default codec config.");
cfg.g_w = info.frame_width;
cfg.g_h = info.frame_height;
@@ -110,8 +104,7 @@ int main(int argc, char **argv) {
cfg.g_timebase.den = info.time_base.denominator;
writer = vpx_video_writer_open(argv[4], kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing.", argv[4]);
if (!writer) die("Failed to open %s for writing.", argv[4]);
if (!(infile = fopen(argv[3], "rb")))
die("Failed to open %s for reading.", argv[3]);
@@ -128,15 +121,15 @@ int main(int argc, char **argv) {
}
// Flush encoder.
while (encode_frame(&codec, NULL, -1, 0, writer)) {}
while (encode_frame(&codec, NULL, -1, 0, writer)) {
}
printf("\n");
fclose(infile);
printf("Processed %d frames.\n", frame_count);
vpx_img_free(&raw);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec.");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec.");
vpx_video_writer_close(writer);

View File

@@ -20,7 +20,6 @@
#include <string.h>
#include <time.h>
#include "../args.h"
#include "../tools_common.h"
#include "../video_writer.h"
@@ -54,8 +53,9 @@ static const arg_def_t spatial_layers_arg =
static const arg_def_t temporal_layers_arg =
ARG_DEF("tl", "temporal-layers", 1, "number of temporal SVC layers");
static const arg_def_t temporal_layering_mode_arg =
ARG_DEF("tlm", "temporal-layering-mode", 1, "temporal layering scheme."
"VP9E_TEMPORAL_LAYERING_MODE");
ARG_DEF("tlm", "temporal-layering-mode", 1,
"temporal layering scheme."
"VP9E_TEMPORAL_LAYERING_MODE");
static const arg_def_t kf_dist_arg =
ARG_DEF("k", "kf-dist", 1, "number of frames between keyframes");
static const arg_def_t scale_factors_arg =
@@ -75,46 +75,59 @@ static const arg_def_t min_bitrate_arg =
static const arg_def_t max_bitrate_arg =
ARG_DEF(NULL, "max-bitrate", 1, "Maximum bitrate");
static const arg_def_t lag_in_frame_arg =
ARG_DEF(NULL, "lag-in-frames", 1, "Number of frame to input before "
"generating any outputs");
ARG_DEF(NULL, "lag-in-frames", 1,
"Number of frame to input before "
"generating any outputs");
static const arg_def_t rc_end_usage_arg =
ARG_DEF(NULL, "rc-end-usage", 1, "0 - 3: VBR, CBR, CQ, Q");
static const arg_def_t speed_arg =
ARG_DEF("sp", "speed", 1, "speed configuration");
static const arg_def_t aqmode_arg =
ARG_DEF("aq", "aqmode", 1, "aq-mode off/on");
static const arg_def_t bitrates_arg =
ARG_DEF("bl", "bitrates", 1, "bitrates[sl * num_tl + tl]");
#if CONFIG_VP9_HIGHBITDEPTH
static const struct arg_enum_list bitdepth_enum[] = {
{"8", VPX_BITS_8},
{"10", VPX_BITS_10},
{"12", VPX_BITS_12},
{NULL, 0}
{ "8", VPX_BITS_8 }, { "10", VPX_BITS_10 }, { "12", VPX_BITS_12 }, { NULL, 0 }
};
static const arg_def_t bitdepth_arg =
ARG_DEF_ENUM("d", "bit-depth", 1, "Bit depth for codec 8, 10 or 12. ",
bitdepth_enum);
static const arg_def_t bitdepth_arg = ARG_DEF_ENUM(
"d", "bit-depth", 1, "Bit depth for codec 8, 10 or 12. ", bitdepth_enum);
#endif // CONFIG_VP9_HIGHBITDEPTH
static const arg_def_t *svc_args[] = {
&frames_arg, &width_arg, &height_arg,
&timebase_arg, &bitrate_arg, &skip_frames_arg, &spatial_layers_arg,
&kf_dist_arg, &scale_factors_arg, &passes_arg, &pass_arg,
&fpf_name_arg, &min_q_arg, &max_q_arg, &min_bitrate_arg,
&max_bitrate_arg, &temporal_layers_arg, &temporal_layering_mode_arg,
&lag_in_frame_arg, &threads_arg, &aqmode_arg,
static const arg_def_t *svc_args[] = { &frames_arg,
&width_arg,
&height_arg,
&timebase_arg,
&bitrate_arg,
&skip_frames_arg,
&spatial_layers_arg,
&kf_dist_arg,
&scale_factors_arg,
&passes_arg,
&pass_arg,
&fpf_name_arg,
&min_q_arg,
&max_q_arg,
&min_bitrate_arg,
&max_bitrate_arg,
&temporal_layers_arg,
&temporal_layering_mode_arg,
&lag_in_frame_arg,
&threads_arg,
&aqmode_arg,
#if OUTPUT_RC_STATS
&output_rc_stats_arg,
&output_rc_stats_arg,
#endif
#if CONFIG_VP9_HIGHBITDEPTH
&bitdepth_arg,
&bitdepth_arg,
#endif
&speed_arg,
&rc_end_usage_arg, NULL
};
&speed_arg,
&rc_end_usage_arg,
&bitrates_arg,
NULL };
static const uint32_t default_frames_to_skip = 0;
static const uint32_t default_frames_to_code = 60 * 60;
@@ -128,7 +141,7 @@ static const uint32_t default_temporal_layers = 1;
static const uint32_t default_kf_dist = 100;
static const uint32_t default_temporal_layering_mode = 0;
static const uint32_t default_output_rc_stats = 0;
static const int32_t default_speed = -1; // -1 means use library default.
static const int32_t default_speed = -1; // -1 means use library default.
static const uint32_t default_threads = 0; // zero means use library default.
typedef struct {
@@ -155,7 +168,7 @@ void usage_exit(void) {
static void parse_command_line(int argc, const char **argv_,
AppInput *app_input, SvcContext *svc_ctx,
vpx_codec_enc_cfg_t *enc_cfg) {
struct arg arg = {0};
struct arg arg = { 0 };
char **argv = NULL;
char **argi = NULL;
char **argj = NULL;
@@ -165,7 +178,7 @@ static void parse_command_line(int argc, const char **argv_,
const char *fpf_file_name = NULL;
unsigned int min_bitrate = 0;
unsigned int max_bitrate = 0;
char string_options[1024] = {0};
char string_options[1024] = { 0 };
// initialize SvcContext with parameters that will be passed to vpx_svc_init
svc_ctx->log_level = SVC_LOG_DEBUG;
@@ -229,8 +242,8 @@ static void parse_command_line(int argc, const char **argv_,
} else if (arg_match(&arg, &threads_arg, argi)) {
svc_ctx->threads = arg_parse_uint(&arg);
} else if (arg_match(&arg, &temporal_layering_mode_arg, argi)) {
svc_ctx->temporal_layering_mode =
enc_cfg->temporal_layering_mode = arg_parse_int(&arg);
svc_ctx->temporal_layering_mode = enc_cfg->temporal_layering_mode =
arg_parse_int(&arg);
if (svc_ctx->temporal_layering_mode) {
enc_cfg->g_error_resilient = 1;
}
@@ -240,6 +253,9 @@ static void parse_command_line(int argc, const char **argv_,
} else if (arg_match(&arg, &scale_factors_arg, argi)) {
snprintf(string_options, sizeof(string_options), "%s scale-factors=%s",
string_options, arg.val);
} else if (arg_match(&arg, &bitrates_arg, argi)) {
snprintf(string_options, sizeof(string_options), "%s bitrates=%s",
string_options, arg.val);
} else if (arg_match(&arg, &passes_arg, argi)) {
passes = arg_parse_uint(&arg);
if (passes < 1 || passes > 2) {
@@ -278,7 +294,7 @@ static void parse_command_line(int argc, const char **argv_,
enc_cfg->g_input_bit_depth = 10;
enc_cfg->g_profile = 2;
break;
case VPX_BITS_12:
case VPX_BITS_12:
enc_cfg->g_input_bit_depth = 12;
enc_cfg->g_profile = 2;
break;
@@ -360,9 +376,8 @@ static void parse_command_line(int argc, const char **argv_,
"num: %d, den: %d, bitrate: %d,\n"
"gop size: %d\n",
vpx_codec_iface_name(vpx_codec_vp9_cx()), app_input->frames_to_code,
app_input->frames_to_skip,
svc_ctx->spatial_layers, enc_cfg->g_w, enc_cfg->g_h,
enc_cfg->g_timebase.num, enc_cfg->g_timebase.den,
app_input->frames_to_skip, svc_ctx->spatial_layers, enc_cfg->g_w,
enc_cfg->g_h, enc_cfg->g_timebase.num, enc_cfg->g_timebase.den,
enc_cfg->rc_target_bitrate, enc_cfg->kf_max_dist);
}
@@ -399,7 +414,7 @@ struct RateControlStats {
// Note: these rate control stats assume only 1 key frame in the
// sequence (i.e., first frame only).
static void set_rate_control_stats(struct RateControlStats *rc,
vpx_codec_enc_cfg_t *cfg) {
vpx_codec_enc_cfg_t *cfg) {
unsigned int sl, tl;
// Set the layer (cumulative) framerate and the target layer (non-cumulative)
// per-frame-bandwidth, for the rate control encoding stats below.
@@ -408,22 +423,18 @@ static void set_rate_control_stats(struct RateControlStats *rc,
for (sl = 0; sl < cfg->ss_number_layers; ++sl) {
for (tl = 0; tl < cfg->ts_number_layers; ++tl) {
const int layer = sl * cfg->ts_number_layers + tl;
const int tlayer0 = sl * cfg->ts_number_layers;
if (cfg->ts_number_layers == 1)
rc->layer_framerate[layer] = framerate;
else
rc->layer_framerate[layer] =
framerate / cfg->ts_rate_decimator[tl];
rc->layer_framerate[layer] = framerate / cfg->ts_rate_decimator[tl];
if (tl > 0) {
rc->layer_pfb[layer] = 1000.0 *
(cfg->layer_target_bitrate[layer] -
cfg->layer_target_bitrate[layer - 1]) /
(rc->layer_framerate[layer] -
rc->layer_framerate[layer - 1]);
rc->layer_pfb[layer] =
1000.0 * (cfg->layer_target_bitrate[layer] -
cfg->layer_target_bitrate[layer - 1]) /
(rc->layer_framerate[layer] - rc->layer_framerate[layer - 1]);
} else {
rc->layer_pfb[tlayer0] = 1000.0 *
cfg->layer_target_bitrate[tlayer0] /
rc->layer_framerate[tlayer0];
rc->layer_pfb[layer] = 1000.0 * cfg->layer_target_bitrate[layer] /
rc->layer_framerate[layer];
}
rc->layer_input_frames[layer] = 0;
rc->layer_enc_frames[layer] = 0;
@@ -443,35 +454,38 @@ static void printout_rate_control_summary(struct RateControlStats *rc,
vpx_codec_enc_cfg_t *cfg,
int frame_cnt) {
unsigned int sl, tl;
int tot_num_frames = 0;
double perc_fluctuation = 0.0;
int tot_num_frames = 0;
printf("Total number of processed frames: %d\n\n", frame_cnt - 1);
printf("Rate control layer stats for sl%d tl%d layer(s):\n\n",
cfg->ss_number_layers, cfg->ts_number_layers);
cfg->ss_number_layers, cfg->ts_number_layers);
for (sl = 0; sl < cfg->ss_number_layers; ++sl) {
tot_num_frames = 0;
for (tl = 0; tl < cfg->ts_number_layers; ++tl) {
const int layer = sl * cfg->ts_number_layers + tl;
const int num_dropped = (tl > 0) ?
(rc->layer_input_frames[layer] - rc->layer_enc_frames[layer]) :
(rc->layer_input_frames[layer] - rc->layer_enc_frames[layer] - 1);
if (!sl)
tot_num_frames += rc->layer_input_frames[layer];
const int num_dropped =
(tl > 0)
? (rc->layer_input_frames[layer] - rc->layer_enc_frames[layer])
: (rc->layer_input_frames[layer] - rc->layer_enc_frames[layer] -
1);
tot_num_frames += rc->layer_input_frames[layer];
rc->layer_encoding_bitrate[layer] = 0.001 * rc->layer_framerate[layer] *
rc->layer_encoding_bitrate[layer] / tot_num_frames;
rc->layer_avg_frame_size[layer] = rc->layer_avg_frame_size[layer] /
rc->layer_enc_frames[layer];
rc->layer_avg_rate_mismatch[layer] =
100.0 * rc->layer_avg_rate_mismatch[layer] /
rc->layer_enc_frames[layer];
rc->layer_encoding_bitrate[layer] /
tot_num_frames;
rc->layer_avg_frame_size[layer] =
rc->layer_avg_frame_size[layer] / rc->layer_enc_frames[layer];
rc->layer_avg_rate_mismatch[layer] = 100.0 *
rc->layer_avg_rate_mismatch[layer] /
rc->layer_enc_frames[layer];
printf("For layer#: sl%d tl%d \n", sl, tl);
printf("Bitrate (target vs actual): %d %f.0 kbps\n",
cfg->layer_target_bitrate[layer],
rc->layer_encoding_bitrate[layer]);
printf("Average frame size (target vs actual): %f %f bits\n",
rc->layer_pfb[layer], rc->layer_avg_frame_size[layer]);
printf("Average rate_mismatch: %f\n",
rc->layer_avg_rate_mismatch[layer]);
printf("Number of input frames, encoded (non-key) frames, "
printf("Average rate_mismatch: %f\n", rc->layer_avg_rate_mismatch[layer]);
printf(
"Number of input frames, encoded (non-key) frames, "
"and percent dropped frames: %d %d %f.0 \n",
rc->layer_input_frames[layer], rc->layer_enc_frames[layer],
100.0 * num_dropped / rc->layer_input_frames[layer]);
@@ -483,19 +497,18 @@ static void printout_rate_control_summary(struct RateControlStats *rc,
rc->variance_st_encoding_bitrate / rc->window_count -
(rc->avg_st_encoding_bitrate * rc->avg_st_encoding_bitrate);
perc_fluctuation = 100.0 * sqrt(rc->variance_st_encoding_bitrate) /
rc->avg_st_encoding_bitrate;
rc->avg_st_encoding_bitrate;
printf("Short-time stats, for window of %d frames: \n", rc->window_size);
printf("Average, rms-variance, and percent-fluct: %f %f %f \n",
rc->avg_st_encoding_bitrate,
sqrt(rc->variance_st_encoding_bitrate),
rc->avg_st_encoding_bitrate, sqrt(rc->variance_st_encoding_bitrate),
perc_fluctuation);
if (frame_cnt != tot_num_frames)
die("Error: Number of input frames not equal to output encoded frames != "
"%d tot_num_frames = %d\n", frame_cnt, tot_num_frames);
"%d tot_num_frames = %d\n",
frame_cnt, tot_num_frames);
}
vpx_codec_err_t parse_superframe_index(const uint8_t *data,
size_t data_sz,
vpx_codec_err_t parse_superframe_index(const uint8_t *data, size_t data_sz,
uint32_t sizes[8], int *count) {
// A chunk ending with a byte matching 0xc0 is an invalid chunk unless
// it is a super frame index. If the last byte of real video compression
@@ -508,7 +521,6 @@ vpx_codec_err_t parse_superframe_index(const uint8_t *data,
marker = *(data + data_sz - 1);
*count = 0;
if ((marker & 0xe0) == 0xc0) {
const uint32_t frames = (marker & 0x7) + 1;
const uint32_t mag = ((marker >> 3) & 0x3) + 1;
@@ -516,8 +528,7 @@ vpx_codec_err_t parse_superframe_index(const uint8_t *data,
// This chunk is marked as having a superframe index but doesn't have
// enough data for it, thus it's an invalid superframe index.
if (data_sz < index_sz)
return VPX_CODEC_CORRUPT_FRAME;
if (data_sz < index_sz) return VPX_CODEC_CORRUPT_FRAME;
{
const uint8_t marker2 = *(data + data_sz - index_sz);
@@ -525,8 +536,7 @@ vpx_codec_err_t parse_superframe_index(const uint8_t *data,
// This chunk is marked as having a superframe index but doesn't have
// the matching marker byte at the front of the index therefore it's an
// invalid chunk.
if (marker != marker2)
return VPX_CODEC_CORRUPT_FRAME;
if (marker != marker2) return VPX_CODEC_CORRUPT_FRAME;
}
{
@@ -537,8 +547,7 @@ vpx_codec_err_t parse_superframe_index(const uint8_t *data,
for (i = 0; i < frames; ++i) {
uint32_t this_sz = 0;
for (j = 0; j < mag; ++j)
this_sz |= (*x++) << (j * 8);
for (j = 0; j < mag; ++j) this_sz |= (*x++) << (j * 8);
sizes[i] = this_sz;
}
*count = frames;
@@ -558,32 +567,27 @@ void set_frame_flags_bypass_mode(int sl, int tl, int num_spatial_layers,
for (sl = 0; sl < num_spatial_layers; ++sl) {
if (!tl) {
if (!sl) {
ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
ref_frame_config->frame_flags[sl] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
} else {
if (is_key_frame) {
ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_LAST |
VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
ref_frame_config->frame_flags[sl] =
VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
} else {
ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
ref_frame_config->frame_flags[sl] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
}
}
} else if (tl == 1) {
if (!sl) {
ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF;
ref_frame_config->frame_flags[sl] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF;
} else {
ref_frame_config->frame_flags[sl] = VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF;
ref_frame_config->frame_flags[sl] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
}
}
if (tl == 0) {
@@ -602,9 +606,9 @@ void set_frame_flags_bypass_mode(int sl, int tl, int num_spatial_layers,
}
int main(int argc, const char **argv) {
AppInput app_input = {0};
AppInput app_input = { 0 };
VpxVideoWriter *writer = NULL;
VpxVideoInfo info = {0};
VpxVideoInfo info = { 0 };
vpx_codec_ctx_t codec;
vpx_codec_enc_cfg_t enc_cfg;
SvcContext svc_ctx;
@@ -618,14 +622,14 @@ int main(int argc, const char **argv) {
int end_of_stream = 0;
int frames_received = 0;
#if OUTPUT_RC_STATS
VpxVideoWriter *outfile[VPX_TS_MAX_LAYERS] = {NULL};
VpxVideoWriter *outfile[VPX_TS_MAX_LAYERS] = { NULL };
struct RateControlStats rc;
vpx_svc_layer_id_t layer_id;
vpx_svc_ref_frame_config_t ref_frame_config;
int sl, tl;
unsigned int sl, tl;
double sum_bitrate = 0.0;
double sum_bitrate2 = 0.0;
double framerate = 30.0;
double framerate = 30.0;
#endif
struct vpx_usec_timer timer;
int64_t cx_time = 0;
@@ -634,10 +638,10 @@ int main(int argc, const char **argv) {
exec_name = argv[0];
parse_command_line(argc, argv, &app_input, &svc_ctx, &enc_cfg);
// Allocate image buffer
// Allocate image buffer
#if CONFIG_VP9_HIGHBITDEPTH
if (!vpx_img_alloc(&raw, enc_cfg.g_input_bit_depth == 8 ?
VPX_IMG_FMT_I420 : VPX_IMG_FMT_I42016,
if (!vpx_img_alloc(&raw, enc_cfg.g_input_bit_depth == 8 ? VPX_IMG_FMT_I420
: VPX_IMG_FMT_I42016,
enc_cfg.g_w, enc_cfg.g_h, 32)) {
die("Failed to allocate image %dx%d\n", enc_cfg.g_w, enc_cfg.g_h);
}
@@ -668,14 +672,14 @@ int main(int argc, const char **argv) {
if (!(app_input.passes == 2 && app_input.pass == 1)) {
// We don't save the bitstream for the 1st pass on two pass rate control
writer = vpx_video_writer_open(app_input.output_filename, kContainerIVF,
&info);
writer =
vpx_video_writer_open(app_input.output_filename, kContainerIVF, &info);
if (!writer)
die("Failed to open %s for writing\n", app_input.output_filename);
}
#if OUTPUT_RC_STATS
// For now, just write temporal layer streams.
// TODO(wonkap): do spatial by re-writing superframe.
// TODO(marpan): do spatial by re-writing superframe.
if (svc_ctx.output_rc_stat) {
for (tl = 0; tl < enc_cfg.ts_number_layers; ++tl) {
char file_name[PATH_MAX];
@@ -683,15 +687,13 @@ int main(int argc, const char **argv) {
snprintf(file_name, sizeof(file_name), "%s_t%d.ivf",
app_input.output_filename, tl);
outfile[tl] = vpx_video_writer_open(file_name, kContainerIVF, &info);
if (!outfile[tl])
die("Failed to open %s for writing", file_name);
if (!outfile[tl]) die("Failed to open %s for writing", file_name);
}
}
#endif
// skip initial frames
for (i = 0; i < app_input.frames_to_skip; ++i)
vpx_img_read(&raw, infile);
for (i = 0; i < app_input.frames_to_skip; ++i) vpx_img_read(&raw, infile);
if (svc_ctx.speed != -1)
vpx_codec_control(&codec, VP8E_SET_CPUUSED, svc_ctx.speed);
@@ -699,7 +701,8 @@ int main(int argc, const char **argv) {
vpx_codec_control(&codec, VP9E_SET_TILE_COLUMNS, (svc_ctx.threads >> 1));
if (svc_ctx.speed >= 5 && svc_ctx.aqmode == 1)
vpx_codec_control(&codec, VP9E_SET_AQ_MODE, 3);
if (svc_ctx.speed >= 5)
vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
// Encode frames
while (!end_of_stream) {
@@ -729,23 +732,22 @@ int main(int argc, const char **argv) {
// over all the spatial layers for the current superframe.
vpx_codec_control(&codec, VP9E_SET_SVC_LAYER_ID, &layer_id);
set_frame_flags_bypass_mode(sl, layer_id.temporal_layer_id,
svc_ctx.spatial_layers,
frame_cnt == 0,
svc_ctx.spatial_layers, frame_cnt == 0,
&ref_frame_config);
vpx_codec_control(&codec, VP9E_SET_SVC_REF_FRAME_CONFIG,
&ref_frame_config);
// Keep track of input frames, to account for frame drops in rate control
// stats/metrics.
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
for (sl = 0; sl < (unsigned int)enc_cfg.ss_number_layers; ++sl) {
++rc.layer_input_frames[sl * enc_cfg.ts_number_layers +
layer_id.temporal_layer_id];
}
}
vpx_usec_timer_start(&timer);
res = vpx_svc_encode(&svc_ctx, &codec, (end_of_stream ? NULL : &raw),
pts, frame_duration, svc_ctx.speed >= 5 ?
VPX_DL_REALTIME : VPX_DL_GOOD_QUALITY);
res = vpx_svc_encode(
&svc_ctx, &codec, (end_of_stream ? NULL : &raw), pts, frame_duration,
svc_ctx.speed >= 5 ? VPX_DL_REALTIME : VPX_DL_GOOD_QUALITY);
vpx_usec_timer_mark(&timer);
cx_time += vpx_usec_timer_elapsed(&timer);
@@ -764,12 +766,11 @@ int main(int argc, const char **argv) {
uint32_t sizes[8];
int count = 0;
#endif
vpx_video_writer_write_frame(writer,
cx_pkt->data.frame.buf,
vpx_video_writer_write_frame(writer, cx_pkt->data.frame.buf,
cx_pkt->data.frame.sz,
cx_pkt->data.frame.pts);
#if OUTPUT_RC_STATS
// TODO(marpan/wonkap): Put this (to line728) in separate function.
// TODO(marpan): Put this (to line728) in separate function.
if (svc_ctx.output_rc_stat) {
vpx_codec_control(&codec, VP9E_GET_SVC_LAYER_ID, &layer_id);
parse_superframe_index(cx_pkt->data.frame.buf,
@@ -782,26 +783,25 @@ int main(int argc, const char **argv) {
VP9E_TEMPORAL_LAYERING_MODE_BYPASS) {
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
++rc.layer_input_frames[sl * enc_cfg.ts_number_layers +
layer_id.temporal_layer_id];
layer_id.temporal_layer_id];
}
}
for (tl = layer_id.temporal_layer_id;
tl < enc_cfg.ts_number_layers; ++tl) {
vpx_video_writer_write_frame(outfile[tl],
cx_pkt->data.frame.buf,
cx_pkt->data.frame.sz,
cx_pkt->data.frame.pts);
tl < enc_cfg.ts_number_layers; ++tl) {
vpx_video_writer_write_frame(
outfile[tl], cx_pkt->data.frame.buf, cx_pkt->data.frame.sz,
cx_pkt->data.frame.pts);
}
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
for (tl = layer_id.temporal_layer_id;
tl < enc_cfg.ts_number_layers; ++tl) {
tl < enc_cfg.ts_number_layers; ++tl) {
const int layer = sl * enc_cfg.ts_number_layers + tl;
++rc.layer_tot_enc_frames[layer];
rc.layer_encoding_bitrate[layer] += 8.0 * sizes[sl];
// Keep count of rate control stats per layer, for non-key
// frames.
if (tl == layer_id.temporal_layer_id &&
if (tl == (unsigned int)layer_id.temporal_layer_id &&
!(cx_pkt->data.frame.flags & VPX_FRAME_IS_KEY)) {
rc.layer_avg_frame_size[layer] += 8.0 * sizes[sl];
rc.layer_avg_rate_mismatch[layer] +=
@@ -815,7 +815,7 @@ int main(int argc, const char **argv) {
// Update for short-time encoding bitrate states, for moving
// window of size rc->window, shifted by rc->window / 2.
// Ignore first window segment, due to key frame.
if (frame_cnt > rc.window_size) {
if (frame_cnt > (unsigned int)rc.window_size) {
tl = layer_id.temporal_layer_id;
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
sum_bitrate += 0.001 * 8.0 * sizes[sl] * framerate;
@@ -831,43 +831,42 @@ int main(int argc, const char **argv) {
}
// Second shifted window.
if (frame_cnt > rc.window_size + rc.window_size / 2) {
tl = layer_id.temporal_layer_id;
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
sum_bitrate2 += 0.001 * 8.0 * sizes[sl] * framerate;
}
if (frame_cnt >
(unsigned int)(rc.window_size + rc.window_size / 2)) {
tl = layer_id.temporal_layer_id;
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
sum_bitrate2 += 0.001 * 8.0 * sizes[sl] * framerate;
}
if (frame_cnt > 2 * rc.window_size &&
frame_cnt % rc.window_size == 0) {
rc.window_count += 1;
rc.avg_st_encoding_bitrate += sum_bitrate2 / rc.window_size;
rc.variance_st_encoding_bitrate +=
(sum_bitrate2 / rc.window_size) *
(sum_bitrate2 / rc.window_size);
sum_bitrate2 = 0.0;
}
if (frame_cnt > (unsigned int)(2 * rc.window_size) &&
frame_cnt % rc.window_size == 0) {
rc.window_count += 1;
rc.avg_st_encoding_bitrate += sum_bitrate2 / rc.window_size;
rc.variance_st_encoding_bitrate +=
(sum_bitrate2 / rc.window_size) *
(sum_bitrate2 / rc.window_size);
sum_bitrate2 = 0.0;
}
}
}
#endif
}
/*
printf("SVC frame: %d, kf: %d, size: %d, pts: %d\n", frames_received,
!!(cx_pkt->data.frame.flags & VPX_FRAME_IS_KEY),
(int)cx_pkt->data.frame.sz, (int)cx_pkt->data.frame.pts);
*/
if (enc_cfg.ss_number_layers == 1 && enc_cfg.ts_number_layers == 1)
si->bytes_sum[0] += (int)cx_pkt->data.frame.sz;
++frames_received;
break;
}
case VPX_CODEC_STATS_PKT: {
stats_write(&app_input.rc_stats,
cx_pkt->data.twopass_stats.buf,
stats_write(&app_input.rc_stats, cx_pkt->data.twopass_stats.buf,
cx_pkt->data.twopass_stats.sz);
break;
}
default: {
break;
}
default: { break; }
}
}
@@ -880,8 +879,8 @@ int main(int argc, const char **argv) {
// Compensate for the extra frame count for the bypass mode.
if (svc_ctx.temporal_layering_mode == VP9E_TEMPORAL_LAYERING_MODE_BYPASS) {
for (sl = 0; sl < enc_cfg.ss_number_layers; ++sl) {
const int layer = sl * enc_cfg.ts_number_layers +
layer_id.temporal_layer_id;
const int layer =
sl * enc_cfg.ts_number_layers + layer_id.temporal_layer_id;
--rc.layer_input_frames[layer];
}
}
@@ -895,8 +894,7 @@ int main(int argc, const char **argv) {
}
#endif
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec");
if (app_input.passes == 2)
stats_close(&app_input.rc_stats, 1);
if (app_input.passes == 2) stats_close(&app_input.rc_stats, 1);
if (writer) {
vpx_video_writer_close(writer);
}
@@ -908,8 +906,7 @@ int main(int argc, const char **argv) {
}
#endif
printf("Frame cnt and encoding time/FPS stats for encoding: %d %f %f \n",
frame_cnt,
1000 * (float)cx_time / (double)(frame_cnt * 1000000),
frame_cnt, 1000 * (float)cx_time / (double)(frame_cnt * 1000000),
1000000 * (double)frame_cnt / (double)cx_time);
vpx_img_free(&raw);
// display average size, psnr

442
examples/vp9cx_set_ref.c Normal file
View File

@@ -0,0 +1,442 @@
/*
* Copyright (c) 2016 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
// VP9 Set Reference Frame
// ============================
//
// This is an example demonstrating how to overwrite the VP9 encoder's
// internal reference frame. In the sample we set the last frame to the
// current frame. This technique could be used to bounce between two cameras.
//
// The decoder would also have to set the reference frame to the same value
// on the same frame, or the video will become corrupt. The 'test_decode'
// variable is set to 1 in this example that tests if the encoder and decoder
// results are matching.
//
// Usage
// -----
// This example encodes a raw video. And the last argument passed in specifies
// the frame number to update the reference frame on. For example, run
// examples/vp9cx_set_ref 352 288 in.yuv out.ivf 4 30
// The parameter is parsed as follows:
//
//
// Extra Variables
// ---------------
// This example maintains the frame number passed on the command line
// in the `update_frame_num` variable.
//
//
// Configuration
// -------------
//
// The reference frame is updated on the frame specified on the command
// line.
//
// Observing The Effects
// ---------------------
// The encoder and decoder results should be matching when the same reference
// frame setting operation is done in both encoder and decoder. Otherwise,
// the encoder/decoder mismatch would be seen.
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include "vpx/vp8cx.h"
#include "vpx/vpx_decoder.h"
#include "vpx/vpx_encoder.h"
#include "vp9/common/vp9_common.h"
#include "./tools_common.h"
#include "./video_writer.h"
static const char *exec_name;
void usage_exit() {
fprintf(stderr,
"Usage: %s <width> <height> <infile> <outfile> "
"<frame> <limit(optional)>\n",
exec_name);
exit(EXIT_FAILURE);
}
static int compare_img(const vpx_image_t *const img1,
const vpx_image_t *const img2) {
uint32_t l_w = img1->d_w;
uint32_t c_w = (img1->d_w + img1->x_chroma_shift) >> img1->x_chroma_shift;
const uint32_t c_h =
(img1->d_h + img1->y_chroma_shift) >> img1->y_chroma_shift;
uint32_t i;
int match = 1;
match &= (img1->fmt == img2->fmt);
match &= (img1->d_w == img2->d_w);
match &= (img1->d_h == img2->d_h);
for (i = 0; i < img1->d_h; ++i)
match &= (memcmp(img1->planes[VPX_PLANE_Y] + i * img1->stride[VPX_PLANE_Y],
img2->planes[VPX_PLANE_Y] + i * img2->stride[VPX_PLANE_Y],
l_w) == 0);
for (i = 0; i < c_h; ++i)
match &= (memcmp(img1->planes[VPX_PLANE_U] + i * img1->stride[VPX_PLANE_U],
img2->planes[VPX_PLANE_U] + i * img2->stride[VPX_PLANE_U],
c_w) == 0);
for (i = 0; i < c_h; ++i)
match &= (memcmp(img1->planes[VPX_PLANE_V] + i * img1->stride[VPX_PLANE_V],
img2->planes[VPX_PLANE_V] + i * img2->stride[VPX_PLANE_V],
c_w) == 0);
return match;
}
#define mmin(a, b) ((a) < (b) ? (a) : (b))
static void find_mismatch(const vpx_image_t *const img1,
const vpx_image_t *const img2, int yloc[4],
int uloc[4], int vloc[4]) {
const uint32_t bsize = 64;
const uint32_t bsizey = bsize >> img1->y_chroma_shift;
const uint32_t bsizex = bsize >> img1->x_chroma_shift;
const uint32_t c_w =
(img1->d_w + img1->x_chroma_shift) >> img1->x_chroma_shift;
const uint32_t c_h =
(img1->d_h + img1->y_chroma_shift) >> img1->y_chroma_shift;
int match = 1;
uint32_t i, j;
yloc[0] = yloc[1] = yloc[2] = yloc[3] = -1;
for (i = 0, match = 1; match && i < img1->d_h; i += bsize) {
for (j = 0; match && j < img1->d_w; j += bsize) {
int k, l;
const int si = mmin(i + bsize, img1->d_h) - i;
const int sj = mmin(j + bsize, img1->d_w) - j;
for (k = 0; match && k < si; ++k) {
for (l = 0; match && l < sj; ++l) {
if (*(img1->planes[VPX_PLANE_Y] +
(i + k) * img1->stride[VPX_PLANE_Y] + j + l) !=
*(img2->planes[VPX_PLANE_Y] +
(i + k) * img2->stride[VPX_PLANE_Y] + j + l)) {
yloc[0] = i + k;
yloc[1] = j + l;
yloc[2] = *(img1->planes[VPX_PLANE_Y] +
(i + k) * img1->stride[VPX_PLANE_Y] + j + l);
yloc[3] = *(img2->planes[VPX_PLANE_Y] +
(i + k) * img2->stride[VPX_PLANE_Y] + j + l);
match = 0;
break;
}
}
}
}
}
uloc[0] = uloc[1] = uloc[2] = uloc[3] = -1;
for (i = 0, match = 1; match && i < c_h; i += bsizey) {
for (j = 0; match && j < c_w; j += bsizex) {
int k, l;
const int si = mmin(i + bsizey, c_h - i);
const int sj = mmin(j + bsizex, c_w - j);
for (k = 0; match && k < si; ++k) {
for (l = 0; match && l < sj; ++l) {
if (*(img1->planes[VPX_PLANE_U] +
(i + k) * img1->stride[VPX_PLANE_U] + j + l) !=
*(img2->planes[VPX_PLANE_U] +
(i + k) * img2->stride[VPX_PLANE_U] + j + l)) {
uloc[0] = i + k;
uloc[1] = j + l;
uloc[2] = *(img1->planes[VPX_PLANE_U] +
(i + k) * img1->stride[VPX_PLANE_U] + j + l);
uloc[3] = *(img2->planes[VPX_PLANE_U] +
(i + k) * img2->stride[VPX_PLANE_U] + j + l);
match = 0;
break;
}
}
}
}
}
vloc[0] = vloc[1] = vloc[2] = vloc[3] = -1;
for (i = 0, match = 1; match && i < c_h; i += bsizey) {
for (j = 0; match && j < c_w; j += bsizex) {
int k, l;
const int si = mmin(i + bsizey, c_h - i);
const int sj = mmin(j + bsizex, c_w - j);
for (k = 0; match && k < si; ++k) {
for (l = 0; match && l < sj; ++l) {
if (*(img1->planes[VPX_PLANE_V] +
(i + k) * img1->stride[VPX_PLANE_V] + j + l) !=
*(img2->planes[VPX_PLANE_V] +
(i + k) * img2->stride[VPX_PLANE_V] + j + l)) {
vloc[0] = i + k;
vloc[1] = j + l;
vloc[2] = *(img1->planes[VPX_PLANE_V] +
(i + k) * img1->stride[VPX_PLANE_V] + j + l);
vloc[3] = *(img2->planes[VPX_PLANE_V] +
(i + k) * img2->stride[VPX_PLANE_V] + j + l);
match = 0;
break;
}
}
}
}
}
}
static void testing_decode(vpx_codec_ctx_t *encoder, vpx_codec_ctx_t *decoder,
unsigned int frame_out, int *mismatch_seen) {
vpx_image_t enc_img, dec_img;
struct vp9_ref_frame ref_enc, ref_dec;
if (*mismatch_seen) return;
ref_enc.idx = 0;
ref_dec.idx = 0;
if (vpx_codec_control(encoder, VP9_GET_REFERENCE, &ref_enc))
die_codec(encoder, "Failed to get encoder reference frame");
enc_img = ref_enc.img;
if (vpx_codec_control(decoder, VP9_GET_REFERENCE, &ref_dec))
die_codec(decoder, "Failed to get decoder reference frame");
dec_img = ref_dec.img;
if (!compare_img(&enc_img, &dec_img)) {
int y[4], u[4], v[4];
*mismatch_seen = 1;
find_mismatch(&enc_img, &dec_img, y, u, v);
printf(
"Encode/decode mismatch on frame %d at"
" Y[%d, %d] {%d/%d},"
" U[%d, %d] {%d/%d},"
" V[%d, %d] {%d/%d}",
frame_out, y[0], y[1], y[2], y[3], u[0], u[1], u[2], u[3], v[0], v[1],
v[2], v[3]);
}
vpx_img_free(&enc_img);
vpx_img_free(&dec_img);
}
static int encode_frame(vpx_codec_ctx_t *ecodec, vpx_image_t *img,
unsigned int frame_in, VpxVideoWriter *writer,
int test_decode, vpx_codec_ctx_t *dcodec,
unsigned int *frame_out, int *mismatch_seen) {
int got_pkts = 0;
vpx_codec_iter_t iter = NULL;
const vpx_codec_cx_pkt_t *pkt = NULL;
int got_data;
const vpx_codec_err_t res =
vpx_codec_encode(ecodec, img, frame_in, 1, 0, VPX_DL_GOOD_QUALITY);
if (res != VPX_CODEC_OK) die_codec(ecodec, "Failed to encode frame");
got_data = 0;
while ((pkt = vpx_codec_get_cx_data(ecodec, &iter)) != NULL) {
got_pkts = 1;
if (pkt->kind == VPX_CODEC_CX_FRAME_PKT) {
const int keyframe = (pkt->data.frame.flags & VPX_FRAME_IS_KEY) != 0;
if (!(pkt->data.frame.flags & VPX_FRAME_IS_FRAGMENT)) {
*frame_out += 1;
}
if (!vpx_video_writer_write_frame(writer, pkt->data.frame.buf,
pkt->data.frame.sz,
pkt->data.frame.pts)) {
die_codec(ecodec, "Failed to write compressed frame");
}
printf(keyframe ? "K" : ".");
fflush(stdout);
got_data = 1;
// Decode 1 frame.
if (test_decode) {
if (vpx_codec_decode(dcodec, pkt->data.frame.buf,
(unsigned int)pkt->data.frame.sz, NULL, 0))
die_codec(dcodec, "Failed to decode frame.");
}
}
}
// Mismatch checking
if (got_data && test_decode) {
testing_decode(ecodec, dcodec, *frame_out, mismatch_seen);
}
return got_pkts;
}
int main(int argc, char **argv) {
FILE *infile = NULL;
// Encoder
vpx_codec_ctx_t ecodec;
vpx_codec_enc_cfg_t cfg;
unsigned int frame_in = 0;
vpx_image_t raw;
vpx_codec_err_t res;
VpxVideoInfo info;
VpxVideoWriter *writer = NULL;
const VpxInterface *encoder = NULL;
// Test encoder/decoder mismatch.
int test_decode = 1;
// Decoder
vpx_codec_ctx_t dcodec;
unsigned int frame_out = 0;
// The frame number to set reference frame on
unsigned int update_frame_num = 0;
int mismatch_seen = 0;
const int fps = 30;
const int bitrate = 500;
const char *width_arg = NULL;
const char *height_arg = NULL;
const char *infile_arg = NULL;
const char *outfile_arg = NULL;
const char *update_frame_num_arg = NULL;
unsigned int limit = 0;
vp9_zero(ecodec);
vp9_zero(cfg);
vp9_zero(info);
exec_name = argv[0];
if (argc < 6) die("Invalid number of arguments");
width_arg = argv[1];
height_arg = argv[2];
infile_arg = argv[3];
outfile_arg = argv[4];
update_frame_num_arg = argv[5];
encoder = get_vpx_encoder_by_name("vp9");
if (!encoder) die("Unsupported codec.");
update_frame_num = (unsigned int)strtoul(update_frame_num_arg, NULL, 0);
// In VP9, the reference buffers (cm->buffer_pool->frame_bufs[i].buf) are
// allocated while calling vpx_codec_encode(), thus, setting reference for
// 1st frame isn't supported.
if (update_frame_num <= 1) {
die("Couldn't parse frame number '%s'\n", update_frame_num_arg);
}
if (argc > 6) {
limit = (unsigned int)strtoul(argv[6], NULL, 0);
if (update_frame_num > limit)
die("Update frame number couldn't larger than limit\n");
}
info.codec_fourcc = encoder->fourcc;
info.frame_width = (int)strtol(width_arg, NULL, 0);
info.frame_height = (int)strtol(height_arg, NULL, 0);
info.time_base.numerator = 1;
info.time_base.denominator = fps;
if (info.frame_width <= 0 || info.frame_height <= 0 ||
(info.frame_width % 2) != 0 || (info.frame_height % 2) != 0) {
die("Invalid frame size: %dx%d", info.frame_width, info.frame_height);
}
if (!vpx_img_alloc(&raw, VPX_IMG_FMT_I420, info.frame_width,
info.frame_height, 1)) {
die("Failed to allocate image.");
}
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
res = vpx_codec_enc_config_default(encoder->codec_interface(), &cfg, 0);
if (res) die_codec(&ecodec, "Failed to get default codec config.");
cfg.g_w = info.frame_width;
cfg.g_h = info.frame_height;
cfg.g_timebase.num = info.time_base.numerator;
cfg.g_timebase.den = info.time_base.denominator;
cfg.rc_target_bitrate = bitrate;
cfg.g_lag_in_frames = 3;
writer = vpx_video_writer_open(outfile_arg, kContainerIVF, &info);
if (!writer) die("Failed to open %s for writing.", outfile_arg);
if (!(infile = fopen(infile_arg, "rb")))
die("Failed to open %s for reading.", infile_arg);
if (vpx_codec_enc_init(&ecodec, encoder->codec_interface(), &cfg, 0))
die_codec(&ecodec, "Failed to initialize encoder");
// Disable alt_ref.
if (vpx_codec_control(&ecodec, VP8E_SET_ENABLEAUTOALTREF, 0))
die_codec(&ecodec, "Failed to set enable auto alt ref");
if (test_decode) {
const VpxInterface *decoder = get_vpx_decoder_by_name("vp9");
if (vpx_codec_dec_init(&dcodec, decoder->codec_interface(), NULL, 0))
die_codec(&dcodec, "Failed to initialize decoder.");
}
// Encode frames.
while (vpx_img_read(&raw, infile)) {
if (limit && frame_in >= limit) break;
if (update_frame_num > 1 && frame_out + 1 == update_frame_num) {
vpx_ref_frame_t ref;
ref.frame_type = VP8_LAST_FRAME;
ref.img = raw;
// Set reference frame in encoder.
if (vpx_codec_control(&ecodec, VP8_SET_REFERENCE, &ref))
die_codec(&ecodec, "Failed to set reference frame");
printf(" <SET_REF>");
// If set_reference in decoder is commented out, the enc/dec mismatch
// would be seen.
if (test_decode) {
if (vpx_codec_control(&dcodec, VP8_SET_REFERENCE, &ref))
die_codec(&dcodec, "Failed to set reference frame");
}
}
encode_frame(&ecodec, &raw, frame_in, writer, test_decode, &dcodec,
&frame_out, &mismatch_seen);
frame_in++;
if (mismatch_seen) break;
}
// Flush encoder.
if (!mismatch_seen)
while (encode_frame(&ecodec, NULL, frame_in, writer, test_decode, &dcodec,
&frame_out, &mismatch_seen)) {
}
printf("\n");
fclose(infile);
printf("Processed %d frames.\n", frame_out);
if (test_decode) {
if (!mismatch_seen)
printf("Encoder/decoder results are matching.\n");
else
printf("Encoder/decoder results are NOT matching.\n");
}
if (test_decode)
if (vpx_codec_destroy(&dcodec))
die_codec(&dcodec, "Failed to destroy decoder");
vpx_img_free(&raw);
if (vpx_codec_destroy(&ecodec))
die_codec(&ecodec, "Failed to destroy encoder.");
vpx_video_writer_close(writer);
return EXIT_SUCCESS;
}

View File

@@ -28,9 +28,7 @@
static const char *exec_name;
void usage_exit(void) {
exit(EXIT_FAILURE);
}
void usage_exit(void) { exit(EXIT_FAILURE); }
// Denoiser states, for temporal denoising.
enum denoiserState {
@@ -41,7 +39,7 @@ enum denoiserState {
kDenoiserOnAdaptive
};
static int mode_to_num_layers[13] = {1, 2, 2, 3, 3, 3, 3, 5, 2, 3, 3, 3, 3};
static int mode_to_num_layers[13] = { 1, 2, 2, 3, 3, 3, 3, 5, 2, 3, 3, 3, 3 };
// For rate control encoding stats.
struct RateControlMetrics {
@@ -86,14 +84,14 @@ static void set_rate_control_metrics(struct RateControlMetrics *rc,
// per-frame-bandwidth, for the rate control encoding stats below.
const double framerate = cfg->g_timebase.den / cfg->g_timebase.num;
rc->layer_framerate[0] = framerate / cfg->ts_rate_decimator[0];
rc->layer_pfb[0] = 1000.0 * rc->layer_target_bitrate[0] /
rc->layer_framerate[0];
rc->layer_pfb[0] =
1000.0 * rc->layer_target_bitrate[0] / rc->layer_framerate[0];
for (i = 0; i < cfg->ts_number_layers; ++i) {
if (i > 0) {
rc->layer_framerate[i] = framerate / cfg->ts_rate_decimator[i];
rc->layer_pfb[i] = 1000.0 *
(rc->layer_target_bitrate[i] - rc->layer_target_bitrate[i - 1]) /
(rc->layer_framerate[i] - rc->layer_framerate[i - 1]);
rc->layer_pfb[i] = 1000.0 * (rc->layer_target_bitrate[i] -
rc->layer_target_bitrate[i - 1]) /
(rc->layer_framerate[i] - rc->layer_framerate[i - 1]);
}
rc->layer_input_frames[i] = 0;
rc->layer_enc_frames[i] = 0;
@@ -114,29 +112,31 @@ static void printout_rate_control_summary(struct RateControlMetrics *rc,
unsigned int i = 0;
int tot_num_frames = 0;
double perc_fluctuation = 0.0;
printf("Total number of processed frames: %d\n\n", frame_cnt -1);
printf("Total number of processed frames: %d\n\n", frame_cnt - 1);
printf("Rate control layer stats for %d layer(s):\n\n",
cfg->ts_number_layers);
cfg->ts_number_layers);
for (i = 0; i < cfg->ts_number_layers; ++i) {
const int num_dropped = (i > 0) ?
(rc->layer_input_frames[i] - rc->layer_enc_frames[i]) :
(rc->layer_input_frames[i] - rc->layer_enc_frames[i] - 1);
const int num_dropped =
(i > 0) ? (rc->layer_input_frames[i] - rc->layer_enc_frames[i])
: (rc->layer_input_frames[i] - rc->layer_enc_frames[i] - 1);
tot_num_frames += rc->layer_input_frames[i];
rc->layer_encoding_bitrate[i] = 0.001 * rc->layer_framerate[i] *
rc->layer_encoding_bitrate[i] / tot_num_frames;
rc->layer_avg_frame_size[i] = rc->layer_avg_frame_size[i] /
rc->layer_enc_frames[i];
rc->layer_avg_rate_mismatch[i] = 100.0 * rc->layer_avg_rate_mismatch[i] /
rc->layer_enc_frames[i];
rc->layer_encoding_bitrate[i] /
tot_num_frames;
rc->layer_avg_frame_size[i] =
rc->layer_avg_frame_size[i] / rc->layer_enc_frames[i];
rc->layer_avg_rate_mismatch[i] =
100.0 * rc->layer_avg_rate_mismatch[i] / rc->layer_enc_frames[i];
printf("For layer#: %d \n", i);
printf("Bitrate (target vs actual): %d %f \n", rc->layer_target_bitrate[i],
rc->layer_encoding_bitrate[i]);
printf("Average frame size (target vs actual): %f %f \n", rc->layer_pfb[i],
rc->layer_avg_frame_size[i]);
printf("Average rate_mismatch: %f \n", rc->layer_avg_rate_mismatch[i]);
printf("Number of input frames, encoded (non-key) frames, "
"and perc dropped frames: %d %d %f \n", rc->layer_input_frames[i],
rc->layer_enc_frames[i],
printf(
"Number of input frames, encoded (non-key) frames, "
"and perc dropped frames: %d %d %f \n",
rc->layer_input_frames[i], rc->layer_enc_frames[i],
100.0 * num_dropped / rc->layer_input_frames[i]);
printf("\n");
}
@@ -145,11 +145,10 @@ static void printout_rate_control_summary(struct RateControlMetrics *rc,
rc->variance_st_encoding_bitrate / rc->window_count -
(rc->avg_st_encoding_bitrate * rc->avg_st_encoding_bitrate);
perc_fluctuation = 100.0 * sqrt(rc->variance_st_encoding_bitrate) /
rc->avg_st_encoding_bitrate;
printf("Short-time stats, for window of %d frames: \n",rc->window_size);
rc->avg_st_encoding_bitrate;
printf("Short-time stats, for window of %d frames: \n", rc->window_size);
printf("Average, rms-variance, and percent-fluct: %f %f %f \n",
rc->avg_st_encoding_bitrate,
sqrt(rc->variance_st_encoding_bitrate),
rc->avg_st_encoding_bitrate, sqrt(rc->variance_st_encoding_bitrate),
perc_fluctuation);
if ((frame_cnt - 1) != tot_num_frames)
die("Error: Number of input frames not equal to output! \n");
@@ -167,20 +166,20 @@ static void set_temporal_layer_pattern(int layering_mode,
switch (layering_mode) {
case 0: {
// 1-layer.
int ids[1] = {0};
int ids[1] = { 0 };
cfg->ts_periodicity = 1;
*flag_periodicity = 1;
cfg->ts_number_layers = 1;
cfg->ts_rate_decimator[0] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// Update L only.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[0] =
VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
break;
}
case 1: {
// 2-layers, 2-frame period.
int ids[2] = {0, 1};
int ids[2] = { 0, 1 };
cfg->ts_periodicity = 2;
*flag_periodicity = 2;
cfg->ts_number_layers = 2;
@@ -189,22 +188,24 @@ static void set_temporal_layer_pattern(int layering_mode,
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
#if 1
// 0=L, 1=GF, Intra-layer prediction enabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF;
layer_flags[1] = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_REF_ARF;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF;
layer_flags[1] =
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_REF_ARF;
#else
// 0=L, 1=GF, Intra-layer prediction disabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF;
// 0=L, 1=GF, Intra-layer prediction disabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF;
layer_flags[1] = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_REF_LAST;
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_REF_LAST;
#endif
break;
}
case 2: {
// 2-layers, 3-frame period.
int ids[3] = {0, 1, 1};
int ids[3] = { 0, 1, 1 };
cfg->ts_periodicity = 3;
*flag_periodicity = 3;
cfg->ts_number_layers = 2;
@@ -212,16 +213,17 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[1] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, Intra-layer prediction enabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] =
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] = layer_flags[2] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST;
break;
}
case 3: {
// 3-layers, 6-frame period.
int ids[6] = {0, 2, 2, 1, 2, 2};
int ids[6] = { 0, 2, 2, 1, 2, 2 };
cfg->ts_periodicity = 6;
*flag_periodicity = 6;
cfg->ts_number_layers = 3;
@@ -230,19 +232,18 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[2] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF, Intra-layer prediction enabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[3] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST;
layer_flags[1] =
layer_flags[2] =
layer_flags[4] =
layer_flags[5] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[3] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[1] = layer_flags[2] = layer_flags[4] = layer_flags[5] =
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_LAST;
break;
}
case 4: {
// 3-layers, 4-frame period.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 4;
cfg->ts_number_layers = 3;
@@ -251,39 +252,41 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[2] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF, Intra-layer prediction disabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[1] =
layer_flags[3] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[1] = layer_flags[3] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
break;
}
case 5: {
// 3-layers, 4-frame period.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 4;
cfg->ts_number_layers = 3;
cfg->ts_number_layers = 3;
cfg->ts_rate_decimator[0] = 4;
cfg->ts_rate_decimator[1] = 2;
cfg->ts_rate_decimator[2] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF, Intra-layer prediction enabled in layer 1, disabled
// in layer 2.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] = layer_flags[3] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] =
layer_flags[3] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
break;
}
case 6: {
// 3-layers, 4-frame period.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 4;
cfg->ts_number_layers = 3;
@@ -292,18 +295,19 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[2] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF, Intra-layer prediction enabled.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] =
layer_flags[3] = VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] = layer_flags[3] =
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
break;
}
case 7: {
// NOTE: Probably of academic interest only.
// 5-layers, 16-frame period.
int ids[16] = {0, 4, 3, 4, 2, 4, 3, 4, 1, 4, 3, 4, 2, 4, 3, 4};
int ids[16] = { 0, 4, 3, 4, 2, 4, 3, 4, 1, 4, 3, 4, 2, 4, 3, 4 };
cfg->ts_periodicity = 16;
*flag_periodicity = 16;
cfg->ts_number_layers = 5;
@@ -313,28 +317,21 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[3] = 2;
cfg->ts_rate_decimator[4] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
layer_flags[0] = VPX_EFLAG_FORCE_KF;
layer_flags[1] =
layer_flags[3] =
layer_flags[5] =
layer_flags[7] =
layer_flags[9] =
layer_flags[11] =
layer_flags[13] =
layer_flags[15] = VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] =
layer_flags[6] =
layer_flags[10] =
layer_flags[14] = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_GF;
layer_flags[4] =
layer_flags[12] = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[8] = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_REF_GF;
layer_flags[0] = VPX_EFLAG_FORCE_KF;
layer_flags[1] = layer_flags[3] = layer_flags[5] = layer_flags[7] =
layer_flags[9] = layer_flags[11] = layer_flags[13] = layer_flags[15] =
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] = layer_flags[6] = layer_flags[10] = layer_flags[14] =
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_GF;
layer_flags[4] = layer_flags[12] =
VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[8] = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_REF_GF;
break;
}
case 8: {
// 2-layers, with sync point at first frame of layer 1.
int ids[2] = {0, 1};
int ids[2] = { 0, 1 };
cfg->ts_periodicity = 2;
*flag_periodicity = 8;
cfg->ts_number_layers = 2;
@@ -346,17 +343,17 @@ static void set_temporal_layer_pattern(int layering_mode,
// key frame. Sync point every 8 frames.
// Layer 0: predict from L and ARF, update L and G.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[0] =
VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_ARF;
// Layer 1: sync point: predict from L and ARF, and update G.
layer_flags[1] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
// Layer 0, predict from L and ARF, update L.
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
// Layer 1: predict from L, G and ARF, and update G.
layer_flags[3] = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ENTROPY;
VP8_EFLAG_NO_UPD_ENTROPY;
// Layer 0.
layer_flags[4] = layer_flags[2];
// Layer 1.
@@ -365,11 +362,11 @@ static void set_temporal_layer_pattern(int layering_mode,
layer_flags[6] = layer_flags[4];
// Layer 1.
layer_flags[7] = layer_flags[5];
break;
break;
}
case 9: {
// 3-layers: Sync points for layer 1 and 2 every 8 frames.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 8;
cfg->ts_number_layers = 3;
@@ -378,20 +375,21 @@ static void set_temporal_layer_pattern(int layering_mode,
cfg->ts_rate_decimator[2] = 1;
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
layer_flags[1] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[3] = layer_flags[5] =
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[3] =
layer_flags[5] = VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
layer_flags[4] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[6] = VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[6] =
VP8_EFLAG_NO_REF_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ARF;
layer_flags[7] = VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_ENTROPY;
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_ENTROPY;
break;
}
case 10: {
@@ -399,7 +397,7 @@ static void set_temporal_layer_pattern(int layering_mode,
// and is only updated on key frame.
// Sync points for layer 1 and 2 every 8 frames.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 8;
cfg->ts_number_layers = 3;
@@ -409,21 +407,21 @@ static void set_temporal_layer_pattern(int layering_mode,
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF.
// Layer 0: predict from L and ARF; update L and G.
layer_flags[0] = VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_REF_GF;
layer_flags[0] =
VPX_EFLAG_FORCE_KF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF;
// Layer 2: sync point: predict from L and ARF; update none.
layer_flags[1] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ENTROPY;
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ENTROPY;
// Layer 1: sync point: predict from L and ARF; update G.
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST;
layer_flags[2] =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
// Layer 2: predict from L, G, ARF; update none.
layer_flags[3] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ENTROPY;
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ENTROPY;
// Layer 0: predict from L and ARF; update L.
layer_flags[4] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_REF_GF;
layer_flags[4] =
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF;
// Layer 2: predict from L, G, ARF; update none.
layer_flags[5] = layer_flags[3];
// Layer 1: predict from L, G, ARF; update G.
@@ -438,7 +436,7 @@ static void set_temporal_layer_pattern(int layering_mode,
// This was added to compare with vp9_spatial_svc_encoder.
// 3-layers, 4-frame period.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 4;
cfg->ts_number_layers = 3;
@@ -448,20 +446,20 @@ static void set_temporal_layer_pattern(int layering_mode,
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF, Intra-layer prediction disabled.
layer_flags[0] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
layer_flags[2] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[1] = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
layer_flags[3] = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_REF_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF;
break;
}
case 12:
default: {
// 3-layers structure as in case 10, but no sync/refresh points for
// layer 1 and 2.
int ids[4] = {0, 2, 1, 2};
int ids[4] = { 0, 2, 1, 2 };
cfg->ts_periodicity = 4;
*flag_periodicity = 8;
cfg->ts_number_layers = 3;
@@ -471,15 +469,15 @@ static void set_temporal_layer_pattern(int layering_mode,
memcpy(cfg->ts_layer_id, ids, sizeof(ids));
// 0=L, 1=GF, 2=ARF.
// Layer 0: predict from L and ARF; update L.
layer_flags[0] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_REF_GF;
layer_flags[0] =
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF;
layer_flags[4] = layer_flags[0];
// Layer 1: predict from L, G, ARF; update G.
layer_flags[2] = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
layer_flags[6] = layer_flags[2];
// Layer 2: predict from L, G, ARF; update none.
layer_flags[1] = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ENTROPY;
VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_ENTROPY;
layer_flags[3] = layer_flags[1];
layer_flags[5] = layer_flags[1];
layer_flags[7] = layer_flags[1];
@@ -489,7 +487,7 @@ static void set_temporal_layer_pattern(int layering_mode,
}
int main(int argc, char **argv) {
VpxVideoWriter *outfile[VPX_TS_MAX_LAYERS] = {NULL};
VpxVideoWriter *outfile[VPX_TS_MAX_LAYERS] = { NULL };
vpx_codec_ctx_t codec;
vpx_codec_enc_cfg_t cfg;
int frame_cnt = 0;
@@ -502,21 +500,21 @@ int main(int argc, char **argv) {
int got_data;
int flags = 0;
unsigned int i;
int pts = 0; // PTS starts at 0.
int pts = 0; // PTS starts at 0.
int frame_duration = 1; // 1 timebase tick per frame.
int layering_mode = 0;
int layer_flags[VPX_TS_MAX_PERIODICITY] = {0};
int layer_flags[VPX_TS_MAX_PERIODICITY] = { 0 };
int flag_periodicity = 1;
#if VPX_ENCODER_ABI_VERSION > (4 + VPX_CODEC_ABI_VERSION)
vpx_svc_layer_id_t layer_id = {0, 0};
vpx_svc_layer_id_t layer_id = { 0, 0 };
#else
vpx_svc_layer_id_t layer_id = {0};
vpx_svc_layer_id_t layer_id = { 0 };
#endif
const VpxInterface *encoder = NULL;
FILE *infile = NULL;
struct RateControlMetrics rc;
int64_t cx_time = 0;
const int min_args_base = 11;
const int min_args_base = 12;
#if CONFIG_VP9_HIGHBITDEPTH
vpx_bit_depth_t bit_depth = VPX_BITS_8;
int input_bit_depth = 8;
@@ -526,37 +524,38 @@ int main(int argc, char **argv) {
#endif // CONFIG_VP9_HIGHBITDEPTH
double sum_bitrate = 0.0;
double sum_bitrate2 = 0.0;
double framerate = 30.0;
double framerate = 30.0;
exec_name = argv[0];
// Check usage and arguments.
if (argc < min_args) {
#if CONFIG_VP9_HIGHBITDEPTH
die("Usage: %s <infile> <outfile> <codec_type(vp8/vp9)> <width> <height> "
"<rate_num> <rate_den> <speed> <frame_drop_threshold> <mode> "
"<Rate_0> ... <Rate_nlayers-1> <bit-depth> \n", argv[0]);
"<rate_num> <rate_den> <speed> <frame_drop_threshold> <threads> <mode> "
"<Rate_0> ... <Rate_nlayers-1> <bit-depth> \n",
argv[0]);
#else
die("Usage: %s <infile> <outfile> <codec_type(vp8/vp9)> <width> <height> "
"<rate_num> <rate_den> <speed> <frame_drop_threshold> <mode> "
"<Rate_0> ... <Rate_nlayers-1> \n", argv[0]);
"<rate_num> <rate_den> <speed> <frame_drop_threshold> <threads> <mode> "
"<Rate_0> ... <Rate_nlayers-1> \n",
argv[0]);
#endif // CONFIG_VP9_HIGHBITDEPTH
}
encoder = get_vpx_encoder_by_name(argv[3]);
if (!encoder)
die("Unsupported codec.");
if (!encoder) die("Unsupported codec.");
printf("Using %s\n", vpx_codec_iface_name(encoder->codec_interface()));
width = strtol(argv[4], NULL, 0);
height = strtol(argv[5], NULL, 0);
width = (unsigned int)strtoul(argv[4], NULL, 0);
height = (unsigned int)strtoul(argv[5], NULL, 0);
if (width < 16 || width % 2 || height < 16 || height % 2) {
die("Invalid resolution: %d x %d", width, height);
}
layering_mode = strtol(argv[10], NULL, 0);
layering_mode = (int)strtol(argv[11], NULL, 0);
if (layering_mode < 0 || layering_mode > 13) {
die("Invalid layering mode (0..12) %s", argv[10]);
die("Invalid layering mode (0..12) %s", argv[11]);
}
if (argc != min_args + mode_to_num_layers[layering_mode]) {
@@ -564,7 +563,7 @@ int main(int argc, char **argv) {
}
#if CONFIG_VP9_HIGHBITDEPTH
switch (strtol(argv[argc-1], NULL, 0)) {
switch (strtol(argv[argc - 1], NULL, 0)) {
case 8:
bit_depth = VPX_BITS_8;
input_bit_depth = 8;
@@ -577,13 +576,11 @@ int main(int argc, char **argv) {
bit_depth = VPX_BITS_12;
input_bit_depth = 12;
break;
default:
die("Invalid bit depth (8, 10, 12) %s", argv[argc-1]);
default: die("Invalid bit depth (8, 10, 12) %s", argv[argc - 1]);
}
if (!vpx_img_alloc(&raw,
bit_depth == VPX_BITS_8 ? VPX_IMG_FMT_I420 :
VPX_IMG_FMT_I42016,
width, height, 32)) {
if (!vpx_img_alloc(
&raw, bit_depth == VPX_BITS_8 ? VPX_IMG_FMT_I420 : VPX_IMG_FMT_I42016,
width, height, 32)) {
die("Failed to allocate image", width, height);
}
#else
@@ -612,31 +609,29 @@ int main(int argc, char **argv) {
#endif // CONFIG_VP9_HIGHBITDEPTH
// Timebase format e.g. 30fps: numerator=1, demoninator = 30.
cfg.g_timebase.num = strtol(argv[6], NULL, 0);
cfg.g_timebase.den = strtol(argv[7], NULL, 0);
cfg.g_timebase.num = (int)strtol(argv[6], NULL, 0);
cfg.g_timebase.den = (int)strtol(argv[7], NULL, 0);
speed = strtol(argv[8], NULL, 0);
speed = (int)strtol(argv[8], NULL, 0);
if (speed < 0) {
die("Invalid speed setting: must be positive");
}
for (i = min_args_base;
(int)i < min_args_base + mode_to_num_layers[layering_mode];
++i) {
rc.layer_target_bitrate[i - 11] = strtol(argv[i], NULL, 0);
(int)i < min_args_base + mode_to_num_layers[layering_mode]; ++i) {
rc.layer_target_bitrate[i - 12] = (int)strtol(argv[i], NULL, 0);
if (strncmp(encoder->name, "vp8", 3) == 0)
cfg.ts_target_bitrate[i - 11] = rc.layer_target_bitrate[i - 11];
cfg.ts_target_bitrate[i - 12] = rc.layer_target_bitrate[i - 12];
else if (strncmp(encoder->name, "vp9", 3) == 0)
cfg.layer_target_bitrate[i - 11] = rc.layer_target_bitrate[i - 11];
cfg.layer_target_bitrate[i - 12] = rc.layer_target_bitrate[i - 12];
}
// Real time parameters.
cfg.rc_dropframe_thresh = strtol(argv[9], NULL, 0);
cfg.rc_dropframe_thresh = (unsigned int)strtoul(argv[9], NULL, 0);
cfg.rc_end_usage = VPX_CBR;
cfg.rc_min_quantizer = 2;
cfg.rc_max_quantizer = 56;
if (strncmp(encoder->name, "vp9", 3) == 0)
cfg.rc_max_quantizer = 52;
if (strncmp(encoder->name, "vp9", 3) == 0) cfg.rc_max_quantizer = 52;
cfg.rc_undershoot_pct = 50;
cfg.rc_overshoot_pct = 50;
cfg.rc_buf_initial_sz = 500;
@@ -647,11 +642,11 @@ int main(int argc, char **argv) {
cfg.rc_resize_allowed = 0;
// Use 1 thread as default.
cfg.g_threads = 1;
cfg.g_threads = (unsigned int)strtoul(argv[10], NULL, 0);
// Enable error resilient mode.
cfg.g_error_resilient = 1;
cfg.g_lag_in_frames = 0;
cfg.g_lag_in_frames = 0;
cfg.kf_mode = VPX_KF_AUTO;
// Disable automatic keyframe placement.
@@ -659,9 +654,7 @@ int main(int argc, char **argv) {
cfg.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
set_temporal_layer_pattern(layering_mode,
&cfg,
layer_flags,
set_temporal_layer_pattern(layering_mode, &cfg, layer_flags,
&flag_periodicity);
set_rate_control_metrics(&rc, &cfg);
@@ -688,15 +681,14 @@ int main(int argc, char **argv) {
snprintf(file_name, sizeof(file_name), "%s_%d.ivf", argv[2], i);
outfile[i] = vpx_video_writer_open(file_name, kContainerIVF, &info);
if (!outfile[i])
die("Failed to open %s for writing", file_name);
if (!outfile[i]) die("Failed to open %s for writing", file_name);
assert(outfile[i] != NULL);
}
// No spatial layers in this encoder.
cfg.ss_number_layers = 1;
// Initialize codec.
// Initialize codec.
#if CONFIG_VP9_HIGHBITDEPTH
if (vpx_codec_enc_init(
&codec, encoder->codec_interface(), &cfg,
@@ -710,16 +702,20 @@ int main(int argc, char **argv) {
vpx_codec_control(&codec, VP8E_SET_CPUUSED, -speed);
vpx_codec_control(&codec, VP8E_SET_NOISE_SENSITIVITY, kDenoiserOff);
vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
vpx_codec_control(&codec, VP8E_SET_GF_CBR_BOOST_PCT, 0);
} else if (strncmp(encoder->name, "vp9", 3) == 0) {
vpx_svc_extra_cfg_t svc_params;
memset(&svc_params, 0, sizeof(svc_params));
vpx_codec_control(&codec, VP8E_SET_CPUUSED, speed);
vpx_codec_control(&codec, VP9E_SET_AQ_MODE, 3);
vpx_codec_control(&codec, VP9E_SET_GF_CBR_BOOST_PCT, 0);
vpx_codec_control(&codec, VP9E_SET_FRAME_PARALLEL_DECODING, 0);
vpx_codec_control(&codec, VP9E_SET_FRAME_PERIODIC_BOOST, 0);
vpx_codec_control(&codec, VP9E_SET_NOISE_SENSITIVITY, kDenoiserOff);
vpx_codec_control(&codec, VP8E_SET_STATIC_THRESHOLD, 1);
vpx_codec_control(&codec, VP9E_SET_TUNE_CONTENT, 0);
vpx_codec_control(&codec, VP9E_SET_TILE_COLUMNS, (cfg.g_threads >> 1));
if (vpx_codec_control(&codec, VP9E_SET_SVC, layering_mode > 0 ? 1: 0))
if (vpx_codec_control(&codec, VP9E_SET_SVC, layering_mode > 0 ? 1 : 0))
die_codec(&codec, "Failed to set SVC");
for (i = 0; i < cfg.ts_number_layers; ++i) {
svc_params.max_quantizers[i] = cfg.rc_max_quantizer;
@@ -760,14 +756,12 @@ int main(int argc, char **argv) {
layer_id.temporal_layer_id);
}
flags = layer_flags[frame_cnt % flag_periodicity];
if (layering_mode == 0)
flags = 0;
if (layering_mode == 0) flags = 0;
frame_avail = vpx_img_read(&raw, infile);
if (frame_avail)
++rc.layer_input_frames[layer_id.temporal_layer_id];
if (frame_avail) ++rc.layer_input_frames[layer_id.temporal_layer_id];
vpx_usec_timer_start(&timer);
if (vpx_codec_encode(&codec, frame_avail? &raw : NULL, pts, 1, flags,
VPX_DL_REALTIME)) {
if (vpx_codec_encode(&codec, frame_avail ? &raw : NULL, pts, 1, flags,
VPX_DL_REALTIME)) {
die_codec(&codec, "Failed to encode frame");
}
vpx_usec_timer_mark(&timer);
@@ -777,12 +771,12 @@ int main(int argc, char **argv) {
layer_flags[0] &= ~VPX_EFLAG_FORCE_KF;
}
got_data = 0;
while ( (pkt = vpx_codec_get_cx_data(&codec, &iter)) ) {
while ((pkt = vpx_codec_get_cx_data(&codec, &iter))) {
got_data = 1;
switch (pkt->kind) {
case VPX_CODEC_CX_FRAME_PKT:
for (i = cfg.ts_layer_id[frame_cnt % cfg.ts_periodicity];
i < cfg.ts_number_layers; ++i) {
i < cfg.ts_number_layers; ++i) {
vpx_video_writer_write_frame(outfile[i], pkt->data.frame.buf,
pkt->data.frame.sz, pts);
++rc.layer_tot_enc_frames[i];
@@ -825,8 +819,7 @@ int main(int argc, char **argv) {
}
}
break;
default:
break;
default: break;
}
}
++frame_cnt;
@@ -836,16 +829,13 @@ int main(int argc, char **argv) {
printout_rate_control_summary(&rc, &cfg, frame_cnt);
printf("\n");
printf("Frame cnt and encoding time/FPS stats for encoding: %d %f %f \n",
frame_cnt,
1000 * (float)cx_time / (double)(frame_cnt * 1000000),
1000000 * (double)frame_cnt / (double)cx_time);
frame_cnt, 1000 * (float)cx_time / (double)(frame_cnt * 1000000),
1000000 * (double)frame_cnt / (double)cx_time);
if (vpx_codec_destroy(&codec))
die_codec(&codec, "Failed to destroy codec");
if (vpx_codec_destroy(&codec)) die_codec(&codec, "Failed to destroy codec");
// Try to rewrite the output file headers with the actual frame count.
for (i = 0; i < cfg.ts_number_layers; ++i)
vpx_video_writer_close(outfile[i]);
for (i = 0; i < cfg.ts_number_layers; ++i) vpx_video_writer_close(outfile[i]);
vpx_img_free(&raw);
return EXIT_SUCCESS;

View File

@@ -46,7 +46,8 @@ int file_is_ivf(struct VpxInputContext *input_ctx) {
is_ivf = 1;
if (mem_get_le16(raw_hdr + 4) != 0) {
fprintf(stderr, "Error: Unrecognized IVF version! This file may not"
fprintf(stderr,
"Error: Unrecognized IVF version! This file may not"
" decode properly.");
}
@@ -69,14 +70,13 @@ int file_is_ivf(struct VpxInputContext *input_ctx) {
return is_ivf;
}
int ivf_read_frame(FILE *infile, uint8_t **buffer,
size_t *bytes_read, size_t *buffer_size) {
char raw_header[IVF_FRAME_HDR_SZ] = {0};
int ivf_read_frame(FILE *infile, uint8_t **buffer, size_t *bytes_read,
size_t *buffer_size) {
char raw_header[IVF_FRAME_HDR_SZ] = { 0 };
size_t frame_size = 0;
if (fread(raw_header, IVF_FRAME_HDR_SZ, 1, infile) != 1) {
if (!feof(infile))
warn("Failed to read frame size\n");
if (!feof(infile)) warn("Failed to read frame size\n");
} else {
frame_size = mem_get_le32(raw_header);

View File

@@ -18,11 +18,11 @@ extern "C" {
int file_is_ivf(struct VpxInputContext *input);
int ivf_read_frame(FILE *infile, uint8_t **buffer,
size_t *bytes_read, size_t *buffer_size);
int ivf_read_frame(FILE *infile, uint8_t **buffer, size_t *bytes_read,
size_t *buffer_size);
#ifdef __cplusplus
} /* extern "C" */
} /* extern "C" */
#endif
#endif // IVFDEC_H_

View File

@@ -13,10 +13,8 @@
#include "vpx/vpx_encoder.h"
#include "vpx_ports/mem_ops.h"
void ivf_write_file_header(FILE *outfile,
const struct vpx_codec_enc_cfg *cfg,
unsigned int fourcc,
int frame_cnt) {
void ivf_write_file_header(FILE *outfile, const struct vpx_codec_enc_cfg *cfg,
unsigned int fourcc, int frame_cnt) {
char header[32];
header[0] = 'D';

View File

@@ -19,17 +19,15 @@ struct vpx_codec_cx_pkt;
extern "C" {
#endif
void ivf_write_file_header(FILE *outfile,
const struct vpx_codec_enc_cfg *cfg,
uint32_t fourcc,
int frame_cnt);
void ivf_write_file_header(FILE *outfile, const struct vpx_codec_enc_cfg *cfg,
uint32_t fourcc, int frame_cnt);
void ivf_write_frame_header(FILE *outfile, int64_t pts, size_t frame_size);
void ivf_write_frame_size(FILE *outfile, size_t frame_size);
#ifdef __cplusplus
} /* extern "C" */
} /* extern "C" */
#endif
#endif // IVFENC_H_

19
libs.mk
View File

@@ -12,7 +12,7 @@
# ARM assembly files are written in RVCT-style. We use some make magic to
# filter those files to allow GCC compilation
ifeq ($(ARCH_ARM),yes)
ASM:=$(if $(filter yes,$(CONFIG_GCC)$(CONFIG_MSVS)),.asm.s,.asm)
ASM:=$(if $(filter yes,$(CONFIG_GCC)$(CONFIG_MSVS)),.asm.S,.asm)
else
ASM:=.asm
endif
@@ -106,9 +106,6 @@ ifeq ($(CONFIG_VP9_DECODER),yes)
CODEC_DOC_SECTIONS += vp9 vp9_decoder
endif
VP9_PREFIX=vp9/
$(BUILD_PFX)$(VP9_PREFIX)%.c.o: CFLAGS += -Wextra
ifeq ($(CONFIG_ENCODERS),yes)
CODEC_DOC_SECTIONS += encoder
endif
@@ -116,6 +113,12 @@ ifeq ($(CONFIG_DECODERS),yes)
CODEC_DOC_SECTIONS += decoder
endif
# Suppress -Wextra warnings in third party code.
$(BUILD_PFX)third_party/googletest/%.cc.o: CXXFLAGS += -Wno-missing-field-initializers
# Suppress -Wextra warnings in first party code pending investigation.
# https://bugs.chromium.org/p/webm/issues/detail?id=1069
$(BUILD_PFX)vp8/encoder/onyx_if.c.o: CFLAGS += -Wno-unknown-warning-option -Wno-clobbered
$(BUILD_PFX)vp8/decoder/onyxd_if.c.o: CFLAGS += -Wno-unknown-warning-option -Wno-clobbered
ifeq ($(CONFIG_MSVS),yes)
CODEC_LIB=$(if $(CONFIG_STATIC_MSVCRT),vpxmt,vpxmd)
@@ -230,7 +233,7 @@ LIBS-$(if yes,$(CONFIG_STATIC)) += $(BUILD_PFX)libvpx.a $(BUILD_PFX)libvpx_g.a
$(BUILD_PFX)libvpx_g.a: $(LIBVPX_OBJS)
SO_VERSION_MAJOR := 4
SO_VERSION_MINOR := 0
SO_VERSION_MINOR := 1
SO_VERSION_PATCH := 0
ifeq ($(filter darwin%,$(TGT_OS)),$(TGT_OS))
LIBVPX_SO := libvpx.$(SO_VERSION_MAJOR).dylib
@@ -363,7 +366,7 @@ endif
#
# Add assembler dependencies for configuration.
#
$(filter %.s.o,$(OBJS-yes)): $(BUILD_PFX)vpx_config.asm
$(filter %.S.o,$(OBJS-yes)): $(BUILD_PFX)vpx_config.asm
$(filter %$(ASM).o,$(OBJS-yes)): $(BUILD_PFX)vpx_config.asm
@@ -388,7 +391,7 @@ LIBVPX_TEST_SRCS=$(addprefix test/,$(call enabled,LIBVPX_TEST_SRCS))
LIBVPX_TEST_BIN=./test_libvpx$(EXE_SFX)
LIBVPX_TEST_DATA=$(addprefix $(LIBVPX_TEST_DATA_PATH)/,\
$(call enabled,LIBVPX_TEST_DATA))
libvpx_test_data_url=http://downloads.webmproject.org/test_data/libvpx/$(1)
libvpx_test_data_url=https://storage.googleapis.com/downloads.webmproject.org/test_data/libvpx/$(1)
TEST_INTRA_PRED_SPEED_BIN=./test_intra_pred_speed$(EXE_SFX)
TEST_INTRA_PRED_SPEED_SRCS=$(addprefix test/,$(call enabled,TEST_INTRA_PRED_SPEED_SRCS))
@@ -402,7 +405,7 @@ CLEAN-OBJS += libvpx_test_srcs.txt
$(LIBVPX_TEST_DATA): $(SRC_PATH_BARE)/test/test-data.sha1
@echo " [DOWNLOAD] $@"
$(qexec)trap 'rm -f $@' INT TERM &&\
curl -L -o $@ $(call libvpx_test_data_url,$(@F))
curl --retry 1 -L -o $@ $(call libvpx_test_data_url,$(@F))
testdata:: $(LIBVPX_TEST_DATA)
$(qexec)[ -x "$$(which sha1sum)" ] && sha1sum=sha1sum;\

View File

@@ -20,19 +20,17 @@
* Still in the public domain.
*/
#include <string.h> /* for memcpy() */
#include <string.h> /* for memcpy() */
#include "md5_utils.h"
static void
byteSwap(UWORD32 *buf, unsigned words) {
static void byteSwap(UWORD32 *buf, unsigned words) {
md5byte *p;
/* Only swap bytes for big endian machines */
int i = 1;
if (*(char *)&i == 1)
return;
if (*(char *)&i == 1) return;
p = (md5byte *)buf;
@@ -47,8 +45,7 @@ byteSwap(UWORD32 *buf, unsigned words) {
* Start MD5 accumulation. Set bit count to 0 and buffer to mysterious
* initialization constants.
*/
void
MD5Init(struct MD5Context *ctx) {
void MD5Init(struct MD5Context *ctx) {
ctx->buf[0] = 0x67452301;
ctx->buf[1] = 0xefcdab89;
ctx->buf[2] = 0x98badcfe;
@@ -62,8 +59,7 @@ MD5Init(struct MD5Context *ctx) {
* Update context to reflect the concatenation of another buffer full
* of bytes.
*/
void
MD5Update(struct MD5Context *ctx, md5byte const *buf, unsigned len) {
void MD5Update(struct MD5Context *ctx, md5byte const *buf, unsigned len) {
UWORD32 t;
/* Update byte count */
@@ -71,9 +67,9 @@ MD5Update(struct MD5Context *ctx, md5byte const *buf, unsigned len) {
t = ctx->bytes[0];
if ((ctx->bytes[0] = t + len) < t)
ctx->bytes[1]++; /* Carry from low to high */
ctx->bytes[1]++; /* Carry from low to high */
t = 64 - (t & 0x3f); /* Space available in ctx->in (at least 1) */
t = 64 - (t & 0x3f); /* Space available in ctx->in (at least 1) */
if (t > len) {
memcpy((md5byte *)ctx->in + 64 - t, buf, len);
@@ -104,8 +100,7 @@ MD5Update(struct MD5Context *ctx, md5byte const *buf, unsigned len) {
* Final wrapup - pad to 64-byte boundary with the bit pattern
* 1 0* (64-bit count of bits processed, MSB-first)
*/
void
MD5Final(md5byte digest[16], struct MD5Context *ctx) {
void MD5Final(md5byte digest[16], struct MD5Context *ctx) {
int count = ctx->bytes[0] & 0x3f; /* Number of bytes in ctx->in */
md5byte *p = (md5byte *)ctx->in + count;
@@ -115,7 +110,7 @@ MD5Final(md5byte digest[16], struct MD5Context *ctx) {
/* Bytes of padding needed to make 56 bytes (-8..55) */
count = 56 - 1 - count;
if (count < 0) { /* Padding forces an extra block */
if (count < 0) { /* Padding forces an extra block */
memset(p, 0, count + 8);
byteSwap(ctx->in, 16);
MD5Transform(ctx->buf, ctx->in);
@@ -147,8 +142,8 @@ MD5Final(md5byte digest[16], struct MD5Context *ctx) {
#define F4(x, y, z) (y ^ (x | ~z))
/* This is the central step in the MD5 algorithm. */
#define MD5STEP(f,w,x,y,z,in,s) \
(w += f(x,y,z) + in, w = (w<<s | w>>(32-s)) + x)
#define MD5STEP(f, w, x, y, z, in, s) \
(w += f(x, y, z) + in, w = (w << s | w >> (32 - s)) + x)
#if defined(__clang__) && defined(__has_attribute)
#if __has_attribute(no_sanitize)
@@ -166,8 +161,8 @@ MD5Final(md5byte digest[16], struct MD5Context *ctx) {
* reflect the addition of 16 longwords of new data. MD5Update blocks
* the data and converts bytes into longwords for this routine.
*/
VPX_NO_UNSIGNED_OVERFLOW_CHECK void
MD5Transform(UWORD32 buf[4], UWORD32 const in[16]) {
VPX_NO_UNSIGNED_OVERFLOW_CHECK void MD5Transform(UWORD32 buf[4],
UWORD32 const in[16]) {
register UWORD32 a, b, c, d;
a = buf[0];

View File

@@ -45,8 +45,7 @@ struct rate_hist *init_rate_histogram(const vpx_codec_enc_cfg_t *cfg,
hist->samples = cfg->rc_buf_sz * 5 / 4 * fps->num / fps->den / 1000;
// prevent division by zero
if (hist->samples == 0)
hist->samples = 1;
if (hist->samples == 0) hist->samples = 1;
hist->frames = 0;
hist->total = 0;
@@ -78,18 +77,16 @@ void update_rate_histogram(struct rate_hist *hist,
int64_t avg_bitrate = 0;
int64_t sum_sz = 0;
const int64_t now = pkt->data.frame.pts * 1000 *
(uint64_t)cfg->g_timebase.num /
(uint64_t)cfg->g_timebase.den;
(uint64_t)cfg->g_timebase.num /
(uint64_t)cfg->g_timebase.den;
int idx = hist->frames++ % hist->samples;
hist->pts[idx] = now;
hist->sz[idx] = (int)pkt->data.frame.sz;
if (now < cfg->rc_buf_initial_sz)
return;
if (now < cfg->rc_buf_initial_sz) return;
if (!cfg->rc_target_bitrate)
return;
if (!cfg->rc_target_bitrate) return;
then = now;
@@ -98,20 +95,16 @@ void update_rate_histogram(struct rate_hist *hist,
const int i_idx = (i - 1) % hist->samples;
then = hist->pts[i_idx];
if (now - then > cfg->rc_buf_sz)
break;
if (now - then > cfg->rc_buf_sz) break;
sum_sz += hist->sz[i_idx];
}
if (now == then)
return;
if (now == then) return;
avg_bitrate = sum_sz * 8 * 1000 / (now - then);
idx = (int)(avg_bitrate * (RATE_BINS / 2) / (cfg->rc_target_bitrate * 1000));
if (idx < 0)
idx = 0;
if (idx > RATE_BINS - 1)
idx = RATE_BINS - 1;
if (idx < 0) idx = 0;
if (idx > RATE_BINS - 1) idx = RATE_BINS - 1;
if (hist->bucket[idx].low > avg_bitrate)
hist->bucket[idx].low = (int)avg_bitrate;
if (hist->bucket[idx].high < avg_bitrate)
@@ -120,8 +113,8 @@ void update_rate_histogram(struct rate_hist *hist,
hist->total++;
}
static int merge_hist_buckets(struct hist_bucket *bucket,
int max_buckets, int *num_buckets) {
static int merge_hist_buckets(struct hist_bucket *bucket, int max_buckets,
int *num_buckets) {
int small_bucket = 0, merge_bucket = INT_MAX, big_bucket = 0;
int buckets = *num_buckets;
int i;
@@ -129,10 +122,8 @@ static int merge_hist_buckets(struct hist_bucket *bucket,
/* Find the extrema for this list of buckets */
big_bucket = small_bucket = 0;
for (i = 0; i < buckets; i++) {
if (bucket[i].count < bucket[small_bucket].count)
small_bucket = i;
if (bucket[i].count > bucket[big_bucket].count)
big_bucket = i;
if (bucket[i].count < bucket[small_bucket].count) small_bucket = i;
if (bucket[i].count > bucket[big_bucket].count) big_bucket = i;
}
/* If we have too many buckets, merge the smallest with an adjacent
@@ -174,13 +165,10 @@ static int merge_hist_buckets(struct hist_bucket *bucket,
*/
big_bucket = small_bucket = 0;
for (i = 0; i < buckets; i++) {
if (i > merge_bucket)
bucket[i] = bucket[i + 1];
if (i > merge_bucket) bucket[i] = bucket[i + 1];
if (bucket[i].count < bucket[small_bucket].count)
small_bucket = i;
if (bucket[i].count > bucket[big_bucket].count)
big_bucket = i;
if (bucket[i].count < bucket[small_bucket].count) small_bucket = i;
if (bucket[i].count > bucket[big_bucket].count) big_bucket = i;
}
}
@@ -188,8 +176,8 @@ static int merge_hist_buckets(struct hist_bucket *bucket,
return bucket[big_bucket].count;
}
static void show_histogram(const struct hist_bucket *bucket,
int buckets, int total, int scale) {
static void show_histogram(const struct hist_bucket *bucket, int buckets,
int total, int scale) {
const char *pat1, *pat2;
int i;
@@ -232,8 +220,7 @@ static void show_histogram(const struct hist_bucket *bucket,
pct = (float)(100.0 * bucket[i].count / total);
len = HIST_BAR_MAX * bucket[i].count / scale;
if (len < 1)
len = 1;
if (len < 1) len = 1;
assert(len <= HIST_BAR_MAX);
if (bucket[i].low == bucket[i].high)
@@ -241,8 +228,7 @@ static void show_histogram(const struct hist_bucket *bucket,
else
fprintf(stderr, pat2, bucket[i].low, bucket[i].high);
for (j = 0; j < HIST_BAR_MAX; j++)
fprintf(stderr, j < len ? "=" : " ");
for (j = 0; j < HIST_BAR_MAX; j++) fprintf(stderr, j < len ? "=" : " ");
fprintf(stderr, "\t%5d (%6.2f%%)\n", bucket[i].count, pct);
}
}
@@ -268,14 +254,13 @@ void show_q_histogram(const int counts[64], int max_buckets) {
show_histogram(bucket, buckets, total, scale);
}
void show_rate_histogram(struct rate_hist *hist,
const vpx_codec_enc_cfg_t *cfg, int max_buckets) {
void show_rate_histogram(struct rate_hist *hist, const vpx_codec_enc_cfg_t *cfg,
int max_buckets) {
int i, scale;
int buckets = 0;
for (i = 0; i < RATE_BINS; i++) {
if (hist->bucket[i].low == INT_MAX)
continue;
if (hist->bucket[i].low == INT_MAX) continue;
hist->bucket[buckets++] = hist->bucket[i];
}

View File

@@ -23,9 +23,7 @@ class ACMRandom {
explicit ACMRandom(int seed) : random_(seed) {}
void Reset(int seed) {
random_.Reseed(seed);
}
void Reset(int seed) { random_.Reseed(seed); }
uint16_t Rand16(void) {
const uint32_t value =
random_.Generate(testing::internal::Random::kMaxRange);
@@ -52,17 +50,11 @@ class ACMRandom {
return r < 128 ? r << 4 : r >> 4;
}
int PseudoUniform(int range) {
return random_.Generate(range);
}
int PseudoUniform(int range) { return random_.Generate(range); }
int operator()(int n) {
return PseudoUniform(n);
}
int operator()(int n) { return PseudoUniform(n); }
static int DeterministicSeed(void) {
return 0xbaba;
}
static int DeterministicSeed(void) { return 0xbaba; }
private:
testing::internal::Random random_;

View File

@@ -17,8 +17,8 @@
namespace {
// Check if any pixel in a 16x16 macroblock varies between frames.
int CheckMb(const vpx_image_t &current, const vpx_image_t &previous,
int mb_r, int mb_c) {
int CheckMb(const vpx_image_t &current, const vpx_image_t &previous, int mb_r,
int mb_c) {
for (int plane = 0; plane < 3; plane++) {
int r = 16 * mb_r;
int c0 = 16 * mb_c;
@@ -37,8 +37,9 @@ int CheckMb(const vpx_image_t &current, const vpx_image_t &previous,
for (; r < r_top; ++r) {
for (int c = c0; c < c_top; ++c) {
if (current.planes[plane][current.stride[plane] * r + c] !=
previous.planes[plane][previous.stride[plane] * r + c])
previous.planes[plane][previous.stride[plane] * r + c]) {
return 1;
}
}
}
}

View File

@@ -39,6 +39,7 @@ class ActiveMapTest
encoder->Control(VP8E_SET_CPUUSED, cpu_used_);
} else if (video->frame() == 3) {
vpx_active_map_t map = vpx_active_map_t();
/* clang-format off */
uint8_t active_map[9 * 13] = {
1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0,
1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0,
@@ -50,6 +51,7 @@ class ActiveMapTest
0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 1, 1,
1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0,
};
/* clang-format on */
map.cols = (kWidth + 15) / 16;
map.rows = (kHeight + 15) / 16;
ASSERT_EQ(map.cols, 13u);
@@ -77,8 +79,8 @@ TEST_P(ActiveMapTest, Test) {
cfg_.rc_end_usage = VPX_CBR;
cfg_.kf_max_dist = 90000;
::libvpx_test::I420VideoSource video("hantro_odd.yuv", kWidth, kHeight, 30,
1, 0, 20);
::libvpx_test::I420VideoSource video("hantro_odd.yuv", kWidth, kHeight, 30, 1,
0, 20);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}

View File

@@ -13,22 +13,21 @@
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_dsp_rtcd.h"
#include "vpx/vpx_integer.h"
#include "vpx_dsp/postproc.h"
#include "vpx_mem/vpx_mem.h"
namespace {
// TODO(jimbankoski): make width and height integers not unsigned.
typedef void (*AddNoiseFunc)(unsigned char *start, char *noise,
char blackclamp[16], char whiteclamp[16],
char bothclamp[16], unsigned int width,
unsigned int height, int pitch);
static const int kNoiseSize = 3072;
class AddNoiseTest
: public ::testing::TestWithParam<AddNoiseFunc> {
// TODO(jimbankoski): make width and height integers not unsigned.
typedef void (*AddNoiseFunc)(uint8_t *start, const int8_t *noise,
int blackclamp, int whiteclamp, int width,
int height, int pitch);
class AddNoiseTest : public ::testing::TestWithParam<AddNoiseFunc> {
public:
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
virtual ~AddNoiseTest() {}
};
@@ -40,71 +39,19 @@ double stddev6(char a, char b, char c, char d, char e, char f) {
return sqrt(v);
}
// TODO(jimbankoski): The following 2 functions are duplicated in each codec.
// For now the vp9 one has been copied into the test as is. We should normalize
// these in vpx_dsp and not have 3 copies of these unless there is different
// noise we add for each codec.
double gaussian(double sigma, double mu, double x) {
return 1 / (sigma * sqrt(2.0 * 3.14159265)) *
(exp(-(x - mu) * (x - mu) / (2 * sigma * sigma)));
}
int setup_noise(int size_noise, char *noise) {
char char_dist[300];
const int ai = 4;
const int qi = 24;
const double sigma = ai + .5 + .6 * (63 - qi) / 63.0;
/* set up a lookup table of 256 entries that matches
* a gaussian distribution with sigma determined by q.
*/
int next = 0;
for (int i = -32; i < 32; i++) {
int a_i = (int) (0.5 + 256 * gaussian(sigma, 0, i));
if (a_i) {
for (int j = 0; j < a_i; j++) {
char_dist[next + j] = (char)(i);
}
next = next + a_i;
}
}
for (; next < 256; next++)
char_dist[next] = 0;
for (int i = 0; i < size_noise; i++) {
noise[i] = char_dist[rand() & 0xff]; // NOLINT
}
// Returns the most negative value in distribution.
return char_dist[0];
}
TEST_P(AddNoiseTest, CheckNoiseAdded) {
DECLARE_ALIGNED(16, char, blackclamp[16]);
DECLARE_ALIGNED(16, char, whiteclamp[16]);
DECLARE_ALIGNED(16, char, bothclamp[16]);
const int width = 64;
const int width = 64;
const int height = 64;
const int image_size = width * height;
char noise[3072];
int8_t noise[kNoiseSize];
const int clamp = vpx_setup_noise(4.4, noise, kNoiseSize);
uint8_t *const s =
reinterpret_cast<uint8_t *>(vpx_calloc(image_size, sizeof(*s)));
ASSERT_TRUE(s != NULL);
memset(s, 99, image_size * sizeof(*s));
const int clamp = setup_noise(3072, noise);
for (int i = 0; i < 16; i++) {
blackclamp[i] = -clamp;
whiteclamp[i] = -clamp;
bothclamp[i] = -2 * clamp;
}
uint8_t *const s = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
memset(s, 99, image_size);
ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
bothclamp, width, height, width));
ASM_REGISTER_STATE_CHECK(
GetParam()(s, noise, clamp, clamp, width, height, width));
// Check to make sure we don't end up having either the same or no added
// noise either vertically or horizontally.
@@ -122,60 +69,52 @@ TEST_P(AddNoiseTest, CheckNoiseAdded) {
// Initialize pixels in the image to 255 and check for roll over.
memset(s, 255, image_size);
ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
bothclamp, width, height, width));
ASM_REGISTER_STATE_CHECK(
GetParam()(s, noise, clamp, clamp, width, height, width));
// Check to make sure don't roll over.
for (int i = 0; i < image_size; ++i) {
EXPECT_GT((int)s[i], 10) << "i = " << i;
EXPECT_GT(static_cast<int>(s[i]), clamp) << "i = " << i;
}
// Initialize pixels in the image to 0 and check for roll under.
memset(s, 0, image_size);
ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
bothclamp, width, height, width));
ASM_REGISTER_STATE_CHECK(
GetParam()(s, noise, clamp, clamp, width, height, width));
// Check to make sure don't roll under.
for (int i = 0; i < image_size; ++i) {
EXPECT_LT((int)s[i], 245) << "i = " << i;
EXPECT_LT(static_cast<int>(s[i]), 255 - clamp) << "i = " << i;
}
vpx_free(s);
}
TEST_P(AddNoiseTest, CheckCvsAssembly) {
DECLARE_ALIGNED(16, char, blackclamp[16]);
DECLARE_ALIGNED(16, char, whiteclamp[16]);
DECLARE_ALIGNED(16, char, bothclamp[16]);
const int width = 64;
const int width = 64;
const int height = 64;
const int image_size = width * height;
char noise[3072];
const int clamp = setup_noise(3072, noise);
for (int i = 0; i < 16; i++) {
blackclamp[i] = -clamp;
whiteclamp[i] = -clamp;
bothclamp[i] = -2 * clamp;
}
int8_t noise[kNoiseSize];
const int clamp = vpx_setup_noise(4.4, noise, kNoiseSize);
uint8_t *const s = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
uint8_t *const d = reinterpret_cast<uint8_t *>(vpx_calloc(image_size, 1));
ASSERT_TRUE(s != NULL);
ASSERT_TRUE(d != NULL);
memset(s, 99, image_size);
memset(d, 99, image_size);
srand(0);
ASM_REGISTER_STATE_CHECK(GetParam()(s, noise, blackclamp, whiteclamp,
bothclamp, width, height, width));
ASM_REGISTER_STATE_CHECK(
GetParam()(s, noise, clamp, clamp, width, height, width));
srand(0);
ASM_REGISTER_STATE_CHECK(vpx_plane_add_noise_c(d, noise, blackclamp,
whiteclamp, bothclamp,
width, height, width));
ASM_REGISTER_STATE_CHECK(
vpx_plane_add_noise_c(d, noise, clamp, clamp, width, height, width));
for (int i = 0; i < image_size; ++i) {
EXPECT_EQ((int)s[i], (int)d[i]) << "i = " << i;
EXPECT_EQ(static_cast<int>(s[i]), static_cast<int>(d[i])) << "i = " << i;
}
vpx_free(d);

View File

@@ -0,0 +1,157 @@
/*
* Copyright (c) 2012 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "test/codec_factory.h"
#include "test/encode_test_driver.h"
#include "test/i420_video_source.h"
#include "test/util.h"
namespace {
class AltRefAqSegmentTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
protected:
AltRefAqSegmentTest() : EncoderTest(GET_PARAM(0)) {}
virtual ~AltRefAqSegmentTest() {}
virtual void SetUp() {
InitializeConfig();
SetMode(GET_PARAM(1));
set_cpu_used_ = GET_PARAM(2);
aq_mode_ = 0;
alt_ref_aq_mode_ = 0;
}
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
if (video->frame() == 1) {
encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
encoder->Control(VP9E_SET_ALT_REF_AQ, alt_ref_aq_mode_);
encoder->Control(VP9E_SET_AQ_MODE, aq_mode_);
encoder->Control(VP8E_SET_MAX_INTRA_BITRATE_PCT, 100);
}
}
int set_cpu_used_;
int aq_mode_;
int alt_ref_aq_mode_;
};
// Validate that this ALT_REF_AQ/AQ segmentation mode
// (ALT_REF_AQ=0, AQ=0/no_aq)
// encodes and decodes without a mismatch.
TEST_P(AltRefAqSegmentTest, TestNoMisMatchAltRefAQ0) {
cfg_.rc_min_quantizer = 8;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_VBR;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_target_bitrate = 300;
aq_mode_ = 0;
alt_ref_aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
// Validate that this ALT_REF_AQ/AQ segmentation mode
// (ALT_REF_AQ=0, AQ=1/variance_aq)
// encodes and decodes without a mismatch.
TEST_P(AltRefAqSegmentTest, TestNoMisMatchAltRefAQ1) {
cfg_.rc_min_quantizer = 8;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_VBR;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_target_bitrate = 300;
aq_mode_ = 1;
alt_ref_aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
// Validate that this ALT_REF_AQ/AQ segmentation mode
// (ALT_REF_AQ=0, AQ=2/complexity_aq)
// encodes and decodes without a mismatch.
TEST_P(AltRefAqSegmentTest, TestNoMisMatchAltRefAQ2) {
cfg_.rc_min_quantizer = 8;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_VBR;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_target_bitrate = 300;
aq_mode_ = 2;
alt_ref_aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
// Validate that this ALT_REF_AQ/AQ segmentation mode
// (ALT_REF_AQ=0, AQ=3/cyclicrefresh_aq)
// encodes and decodes without a mismatch.
TEST_P(AltRefAqSegmentTest, TestNoMisMatchAltRefAQ3) {
cfg_.rc_min_quantizer = 8;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_VBR;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_target_bitrate = 300;
aq_mode_ = 3;
alt_ref_aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
// Validate that this ALT_REF_AQ/AQ segmentation mode
// (ALT_REF_AQ=0, AQ=4/equator360_aq)
// encodes and decodes without a mismatch.
TEST_P(AltRefAqSegmentTest, TestNoMisMatchAltRefAQ4) {
cfg_.rc_min_quantizer = 8;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_VBR;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_target_bitrate = 300;
aq_mode_ = 4;
alt_ref_aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
VP9_INSTANTIATE_TEST_CASE(AltRefAqSegmentTest,
::testing::Values(::libvpx_test::kOnePassGood,
::libvpx_test::kTwoPassGood),
::testing::Range(2, 5));
} // namespace

View File

@@ -21,7 +21,7 @@ const int kLookAheadMin = 5;
const int kLookAheadMax = 26;
class AltRefTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<int> {
public ::libvpx_test::CodecTestWithParam<int> {
protected:
AltRefTest() : EncoderTest(GET_PARAM(0)), altref_count_(0) {}
virtual ~AltRefTest() {}
@@ -31,9 +31,7 @@ class AltRefTest : public ::libvpx_test::EncoderTest,
SetMode(libvpx_test::kTwoPassGood);
}
virtual void BeginPassHook(unsigned int pass) {
altref_count_ = 0;
}
virtual void BeginPassHook(unsigned int /*pass*/) { altref_count_ = 0; }
virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
libvpx_test::Encoder *encoder) {
@@ -75,11 +73,8 @@ class AltRefForcedKeyTestLarge
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
protected:
AltRefForcedKeyTestLarge()
: EncoderTest(GET_PARAM(0)),
encoding_mode_(GET_PARAM(1)),
cpu_used_(GET_PARAM(2)),
forced_kf_frame_num_(1),
frame_num_(0) {}
: EncoderTest(GET_PARAM(0)), encoding_mode_(GET_PARAM(1)),
cpu_used_(GET_PARAM(2)), forced_kf_frame_num_(1), frame_num_(0) {}
virtual ~AltRefForcedKeyTestLarge() {}
virtual void SetUp() {
@@ -94,8 +89,8 @@ class AltRefForcedKeyTestLarge
if (video->frame() == 0) {
encoder->Control(VP8E_SET_CPUUSED, cpu_used_);
encoder->Control(VP8E_SET_ENABLEAUTOALTREF, 1);
// override test default for tile columns if necessary.
#if CONFIG_VP9_ENCODER
// override test default for tile columns if necessary.
if (GET_PARAM(0) == &libvpx_test::kVP9) {
encoder->Control(VP9E_SET_TILE_COLUMNS, 6);
}
@@ -147,13 +142,11 @@ TEST_P(AltRefForcedKeyTestLarge, ForcedFrameIsKey) {
}
}
VP8_INSTANTIATE_TEST_CASE(
AltRefForcedKeyTestLarge,
::testing::Values(::libvpx_test::kOnePassGood),
::testing::Range(0, 9));
VP8_INSTANTIATE_TEST_CASE(AltRefForcedKeyTestLarge,
::testing::Values(::libvpx_test::kOnePassGood),
::testing::Range(0, 9));
VP9_INSTANTIATE_TEST_CASE(
AltRefForcedKeyTestLarge,
::testing::Values(::libvpx_test::kOnePassGood),
::testing::Range(0, 9));
VP9_INSTANTIATE_TEST_CASE(AltRefForcedKeyTestLarge,
::testing::Values(::libvpx_test::kOnePassGood),
::testing::Range(0, 9));
} // namespace

View File

@@ -57,7 +57,7 @@ TEST_P(AqSegmentTest, TestNoMisMatchAQ1) {
aq_mode_ = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
@@ -77,7 +77,7 @@ TEST_P(AqSegmentTest, TestNoMisMatchAQ2) {
aq_mode_ = 2;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
@@ -97,7 +97,7 @@ TEST_P(AqSegmentTest, TestNoMisMatchAQ3) {
aq_mode_ = 3;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 100);
30, 1, 0, 100);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}

View File

@@ -31,7 +31,7 @@ class AverageTestBase : public ::testing::Test {
AverageTestBase(int width, int height) : width_(width), height_(height) {}
static void SetUpTestCase() {
source_data_ = reinterpret_cast<uint8_t*>(
source_data_ = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBlockSize));
}
@@ -40,9 +40,7 @@ class AverageTestBase : public ::testing::Test {
source_data_ = NULL;
}
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
// Handle blocks up to 4 blocks 64x64 with stride up to 128
@@ -55,65 +53,65 @@ class AverageTestBase : public ::testing::Test {
}
// Sum Pixels
unsigned int ReferenceAverage8x8(const uint8_t* source, int pitch) {
static unsigned int ReferenceAverage8x8(const uint8_t *source, int pitch) {
unsigned int average = 0;
for (int h = 0; h < 8; ++h)
for (int w = 0; w < 8; ++w)
average += source[h * pitch + w];
for (int h = 0; h < 8; ++h) {
for (int w = 0; w < 8; ++w) average += source[h * pitch + w];
}
return ((average + 32) >> 6);
}
unsigned int ReferenceAverage4x4(const uint8_t* source, int pitch) {
static unsigned int ReferenceAverage4x4(const uint8_t *source, int pitch) {
unsigned int average = 0;
for (int h = 0; h < 4; ++h)
for (int w = 0; w < 4; ++w)
average += source[h * pitch + w];
for (int h = 0; h < 4; ++h) {
for (int w = 0; w < 4; ++w) average += source[h * pitch + w];
}
return ((average + 8) >> 4);
}
void FillConstant(uint8_t fill_constant) {
for (int i = 0; i < width_ * height_; ++i) {
source_data_[i] = fill_constant;
source_data_[i] = fill_constant;
}
}
void FillRandom() {
for (int i = 0; i < width_ * height_; ++i) {
source_data_[i] = rnd_.Rand8();
source_data_[i] = rnd_.Rand8();
}
}
int width_, height_;
static uint8_t* source_data_;
static uint8_t *source_data_;
int source_stride_;
ACMRandom rnd_;
};
typedef unsigned int (*AverageFunction)(const uint8_t* s, int pitch);
typedef unsigned int (*AverageFunction)(const uint8_t *s, int pitch);
typedef std::tr1::tuple<int, int, int, int, AverageFunction> AvgFunc;
class AverageTest
: public AverageTestBase,
public ::testing::WithParamInterface<AvgFunc>{
class AverageTest : public AverageTestBase,
public ::testing::WithParamInterface<AvgFunc> {
public:
AverageTest() : AverageTestBase(GET_PARAM(0), GET_PARAM(1)) {}
protected:
void CheckAverages() {
const int block_size = GET_PARAM(3);
unsigned int expected = 0;
if (GET_PARAM(3) == 8) {
expected = ReferenceAverage8x8(source_data_+ GET_PARAM(2),
source_stride_);
} else if (GET_PARAM(3) == 4) {
expected = ReferenceAverage4x4(source_data_+ GET_PARAM(2),
source_stride_);
if (block_size == 8) {
expected =
ReferenceAverage8x8(source_data_ + GET_PARAM(2), source_stride_);
} else if (block_size == 4) {
expected =
ReferenceAverage4x4(source_data_ + GET_PARAM(2), source_stride_);
}
ASM_REGISTER_STATE_CHECK(GET_PARAM(4)(source_data_+ GET_PARAM(2),
source_stride_));
unsigned int actual = GET_PARAM(4)(source_data_+ GET_PARAM(2),
source_stride_);
ASM_REGISTER_STATE_CHECK(
GET_PARAM(4)(source_data_ + GET_PARAM(2), source_stride_));
unsigned int actual =
GET_PARAM(4)(source_data_ + GET_PARAM(2), source_stride_);
EXPECT_EQ(expected, actual);
}
@@ -124,23 +122,20 @@ typedef void (*IntProRowFunc)(int16_t hbuf[16], uint8_t const *ref,
typedef std::tr1::tuple<int, IntProRowFunc, IntProRowFunc> IntProRowParam;
class IntProRowTest
: public AverageTestBase,
public ::testing::WithParamInterface<IntProRowParam> {
class IntProRowTest : public AverageTestBase,
public ::testing::WithParamInterface<IntProRowParam> {
public:
IntProRowTest()
: AverageTestBase(16, GET_PARAM(0)),
hbuf_asm_(NULL),
hbuf_c_(NULL) {
: AverageTestBase(16, GET_PARAM(0)), hbuf_asm_(NULL), hbuf_c_(NULL) {
asm_func_ = GET_PARAM(1);
c_func_ = GET_PARAM(2);
}
protected:
virtual void SetUp() {
hbuf_asm_ = reinterpret_cast<int16_t*>(
hbuf_asm_ = reinterpret_cast<int16_t *>(
vpx_memalign(kDataAlignment, sizeof(*hbuf_asm_) * 16));
hbuf_c_ = reinterpret_cast<int16_t*>(
hbuf_c_ = reinterpret_cast<int16_t *>(
vpx_memalign(kDataAlignment, sizeof(*hbuf_c_) * 16));
}
@@ -169,9 +164,8 @@ typedef int16_t (*IntProColFunc)(uint8_t const *ref, const int width);
typedef std::tr1::tuple<int, IntProColFunc, IntProColFunc> IntProColParam;
class IntProColTest
: public AverageTestBase,
public ::testing::WithParamInterface<IntProColParam> {
class IntProColTest : public AverageTestBase,
public ::testing::WithParamInterface<IntProColParam> {
public:
IntProColTest() : AverageTestBase(GET_PARAM(0), 1), sum_asm_(0), sum_c_(0) {
asm_func_ = GET_PARAM(1);
@@ -195,15 +189,14 @@ class IntProColTest
typedef int (*SatdFunc)(const int16_t *coeffs, int length);
typedef std::tr1::tuple<int, SatdFunc> SatdTestParam;
class SatdTest
: public ::testing::Test,
public ::testing::WithParamInterface<SatdTestParam> {
class SatdTest : public ::testing::Test,
public ::testing::WithParamInterface<SatdTestParam> {
protected:
virtual void SetUp() {
satd_size_ = GET_PARAM(0);
satd_func_ = GET_PARAM(1);
rnd_.Reset(ACMRandom::DeterministicSeed());
src_ = reinterpret_cast<int16_t*>(
src_ = reinterpret_cast<int16_t *>(
vpx_memalign(16, sizeof(*src_) * satd_size_));
ASSERT_TRUE(src_ != NULL);
}
@@ -235,7 +228,7 @@ class SatdTest
ACMRandom rnd_;
};
uint8_t* AverageTestBase::source_data_ = NULL;
uint8_t *AverageTestBase::source_data_ = NULL;
TEST_P(AverageTest, MinValue) {
FillConstant(0);
@@ -286,7 +279,6 @@ TEST_P(IntProColTest, Random) {
RunComparison();
}
TEST_P(SatdTest, MinValue) {
const int kMin = -32640;
const int expected = -kMin * satd_size_;
@@ -320,92 +312,86 @@ using std::tr1::make_tuple;
INSTANTIATE_TEST_CASE_P(
C, AverageTest,
::testing::Values(
make_tuple(16, 16, 1, 8, &vpx_avg_8x8_c),
make_tuple(16, 16, 1, 4, &vpx_avg_4x4_c)));
::testing::Values(make_tuple(16, 16, 1, 8, &vpx_avg_8x8_c),
make_tuple(16, 16, 1, 4, &vpx_avg_4x4_c)));
INSTANTIATE_TEST_CASE_P(
C, SatdTest,
::testing::Values(
make_tuple(16, &vpx_satd_c),
make_tuple(64, &vpx_satd_c),
make_tuple(256, &vpx_satd_c),
make_tuple(1024, &vpx_satd_c)));
INSTANTIATE_TEST_CASE_P(C, SatdTest,
::testing::Values(make_tuple(16, &vpx_satd_c),
make_tuple(64, &vpx_satd_c),
make_tuple(256, &vpx_satd_c),
make_tuple(1024, &vpx_satd_c)));
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(
SSE2, AverageTest,
::testing::Values(
make_tuple(16, 16, 0, 8, &vpx_avg_8x8_sse2),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_sse2),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_sse2),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_sse2),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_sse2),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_sse2)));
::testing::Values(make_tuple(16, 16, 0, 8, &vpx_avg_8x8_sse2),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_sse2),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_sse2),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_sse2),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_sse2),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_sse2)));
INSTANTIATE_TEST_CASE_P(
SSE2, IntProRowTest, ::testing::Values(
make_tuple(16, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
make_tuple(32, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
make_tuple(64, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c)));
SSE2, IntProRowTest,
::testing::Values(make_tuple(16, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
make_tuple(32, &vpx_int_pro_row_sse2, &vpx_int_pro_row_c),
make_tuple(64, &vpx_int_pro_row_sse2,
&vpx_int_pro_row_c)));
INSTANTIATE_TEST_CASE_P(
SSE2, IntProColTest, ::testing::Values(
make_tuple(16, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
make_tuple(32, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
make_tuple(64, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c)));
SSE2, IntProColTest,
::testing::Values(make_tuple(16, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
make_tuple(32, &vpx_int_pro_col_sse2, &vpx_int_pro_col_c),
make_tuple(64, &vpx_int_pro_col_sse2,
&vpx_int_pro_col_c)));
INSTANTIATE_TEST_CASE_P(
SSE2, SatdTest,
::testing::Values(
make_tuple(16, &vpx_satd_sse2),
make_tuple(64, &vpx_satd_sse2),
make_tuple(256, &vpx_satd_sse2),
make_tuple(1024, &vpx_satd_sse2)));
INSTANTIATE_TEST_CASE_P(SSE2, SatdTest,
::testing::Values(make_tuple(16, &vpx_satd_sse2),
make_tuple(64, &vpx_satd_sse2),
make_tuple(256, &vpx_satd_sse2),
make_tuple(1024, &vpx_satd_sse2)));
#endif
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(
NEON, AverageTest,
::testing::Values(
make_tuple(16, 16, 0, 8, &vpx_avg_8x8_neon),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_neon),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_neon),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_neon),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_neon),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_neon)));
::testing::Values(make_tuple(16, 16, 0, 8, &vpx_avg_8x8_neon),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_neon),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_neon),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_neon),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_neon),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_neon)));
INSTANTIATE_TEST_CASE_P(
NEON, IntProRowTest, ::testing::Values(
make_tuple(16, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
make_tuple(32, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
make_tuple(64, &vpx_int_pro_row_neon, &vpx_int_pro_row_c)));
NEON, IntProRowTest,
::testing::Values(make_tuple(16, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
make_tuple(32, &vpx_int_pro_row_neon, &vpx_int_pro_row_c),
make_tuple(64, &vpx_int_pro_row_neon,
&vpx_int_pro_row_c)));
INSTANTIATE_TEST_CASE_P(
NEON, IntProColTest, ::testing::Values(
make_tuple(16, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
make_tuple(32, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
make_tuple(64, &vpx_int_pro_col_neon, &vpx_int_pro_col_c)));
NEON, IntProColTest,
::testing::Values(make_tuple(16, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
make_tuple(32, &vpx_int_pro_col_neon, &vpx_int_pro_col_c),
make_tuple(64, &vpx_int_pro_col_neon,
&vpx_int_pro_col_c)));
INSTANTIATE_TEST_CASE_P(
NEON, SatdTest,
::testing::Values(
make_tuple(16, &vpx_satd_neon),
make_tuple(64, &vpx_satd_neon),
make_tuple(256, &vpx_satd_neon),
make_tuple(1024, &vpx_satd_neon)));
INSTANTIATE_TEST_CASE_P(NEON, SatdTest,
::testing::Values(make_tuple(16, &vpx_satd_neon),
make_tuple(64, &vpx_satd_neon),
make_tuple(256, &vpx_satd_neon),
make_tuple(1024, &vpx_satd_neon)));
#endif
#if HAVE_MSA
INSTANTIATE_TEST_CASE_P(
MSA, AverageTest,
::testing::Values(
make_tuple(16, 16, 0, 8, &vpx_avg_8x8_msa),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_msa),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_msa),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_msa),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_msa),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_msa)));
::testing::Values(make_tuple(16, 16, 0, 8, &vpx_avg_8x8_msa),
make_tuple(16, 16, 5, 8, &vpx_avg_8x8_msa),
make_tuple(32, 32, 15, 8, &vpx_avg_8x8_msa),
make_tuple(16, 16, 0, 4, &vpx_avg_4x4_msa),
make_tuple(16, 16, 5, 4, &vpx_avg_4x4_msa),
make_tuple(32, 32, 15, 4, &vpx_avg_4x4_msa)));
#endif
} // namespace

View File

@@ -26,11 +26,9 @@
#include "vpx_mem/vpx_mem.h"
extern "C"
double vp9_get_blockiness(const unsigned char *img1, int img1_pitch,
const unsigned char *img2, int img2_pitch,
int width, int height);
extern "C" double vp9_get_blockiness(const unsigned char *img1, int img1_pitch,
const unsigned char *img2, int img2_pitch,
int width, int height);
using libvpx_test::ACMRandom;
@@ -40,9 +38,9 @@ class BlockinessTestBase : public ::testing::Test {
BlockinessTestBase(int width, int height) : width_(width), height_(height) {}
static void SetUpTestCase() {
source_data_ = reinterpret_cast<uint8_t*>(
source_data_ = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
reference_data_ = reinterpret_cast<uint8_t*>(
reference_data_ = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
}
@@ -53,14 +51,12 @@ class BlockinessTestBase : public ::testing::Test {
reference_data_ = NULL;
}
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
// Handle frames up to 640x480
static const int kDataAlignment = 16;
static const int kDataBufferSize = 640*480;
static const int kDataBufferSize = 640 * 480;
virtual void SetUp() {
source_stride_ = (width_ + 31) & ~31;
@@ -68,8 +64,8 @@ class BlockinessTestBase : public ::testing::Test {
rnd_.Reset(ACMRandom::DeterministicSeed());
}
void FillConstant(uint8_t *data, int stride, uint8_t fill_constant,
int width, int height) {
void FillConstant(uint8_t *data, int stride, uint8_t fill_constant, int width,
int height) {
for (int h = 0; h < height; ++h) {
for (int w = 0; w < width; ++w) {
data[h * stride + w] = fill_constant;
@@ -104,10 +100,11 @@ class BlockinessTestBase : public ::testing::Test {
void FillCheckerboard(uint8_t *data, int stride) {
for (int h = 0; h < height_; h += 4) {
for (int w = 0; w < width_; w += 4) {
if (((h/4) ^ (w/4)) & 1)
if (((h / 4) ^ (w / 4)) & 1) {
FillConstant(data + h * stride + w, stride, 255, 4, 4);
else
} else {
FillConstant(data + h * stride + w, stride, 0, 4, 4);
}
}
}
}
@@ -135,9 +132,9 @@ class BlockinessTestBase : public ::testing::Test {
}
}
int width_, height_;
static uint8_t* source_data_;
static uint8_t *source_data_;
int source_stride_;
static uint8_t* reference_data_;
static uint8_t *reference_data_;
int reference_stride_;
ACMRandom rnd_;
@@ -152,32 +149,32 @@ class BlockinessVP9Test
BlockinessVP9Test() : BlockinessTestBase(GET_PARAM(0), GET_PARAM(1)) {}
protected:
int CheckBlockiness() {
return vp9_get_blockiness(source_data_, source_stride_,
reference_data_, reference_stride_,
width_, height_);
double GetBlockiness() const {
return vp9_get_blockiness(source_data_, source_stride_, reference_data_,
reference_stride_, width_, height_);
}
};
#endif // CONFIG_VP9_ENCODER
uint8_t* BlockinessTestBase::source_data_ = NULL;
uint8_t* BlockinessTestBase::reference_data_ = NULL;
uint8_t *BlockinessTestBase::source_data_ = NULL;
uint8_t *BlockinessTestBase::reference_data_ = NULL;
#if CONFIG_VP9_ENCODER
TEST_P(BlockinessVP9Test, SourceBlockierThanReference) {
// Source is blockier than reference.
FillRandomBlocky(source_data_, source_stride_);
FillConstant(reference_data_, reference_stride_, 128);
int super_blocky = CheckBlockiness();
const double super_blocky = GetBlockiness();
EXPECT_EQ(0, super_blocky) << "Blocky source should produce 0 blockiness.";
EXPECT_DOUBLE_EQ(0.0, super_blocky)
<< "Blocky source should produce 0 blockiness.";
}
TEST_P(BlockinessVP9Test, ReferenceBlockierThanSource) {
// Source is blockier than reference.
FillConstant(source_data_, source_stride_, 128);
FillRandomBlocky(reference_data_, reference_stride_);
int super_blocky = CheckBlockiness();
const double super_blocky = GetBlockiness();
EXPECT_GT(super_blocky, 0.0)
<< "Blocky reference should score high for blockiness.";
@@ -187,10 +184,10 @@ TEST_P(BlockinessVP9Test, BlurringDecreasesBlockiness) {
// Source is blockier than reference.
FillConstant(source_data_, source_stride_, 128);
FillRandomBlocky(reference_data_, reference_stride_);
int super_blocky = CheckBlockiness();
const double super_blocky = GetBlockiness();
Blur(reference_data_, reference_stride_, 4);
int less_blocky = CheckBlockiness();
const double less_blocky = GetBlockiness();
EXPECT_GT(super_blocky, less_blocky)
<< "A straight blur should decrease blockiness.";
@@ -201,17 +198,16 @@ TEST_P(BlockinessVP9Test, WorstCaseBlockiness) {
FillConstant(source_data_, source_stride_, 128);
FillCheckerboard(reference_data_, reference_stride_);
int super_blocky = CheckBlockiness();
const double super_blocky = GetBlockiness();
Blur(reference_data_, reference_stride_, 4);
int less_blocky = CheckBlockiness();
const double less_blocky = GetBlockiness();
EXPECT_GT(super_blocky, less_blocky)
<< "A straight blur should decrease blockiness.";
}
#endif // CONFIG_VP9_ENCODER
using std::tr1::make_tuple;
//------------------------------------------------------------------------------
@@ -219,9 +215,7 @@ using std::tr1::make_tuple;
#if CONFIG_VP9_ENCODER
const BlockinessParam c_vp9_tests[] = {
make_tuple(320, 240),
make_tuple(318, 242),
make_tuple(318, 238),
make_tuple(320, 240), make_tuple(318, 242), make_tuple(318, 238),
};
INSTANTIATE_TEST_CASE_P(C, BlockinessVP9Test, ::testing::ValuesIn(c_vp9_tests));
#endif

View File

@@ -17,8 +17,9 @@
namespace {
class BordersTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class BordersTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
BordersTest() : EncoderTest(GET_PARAM(0)) {}
virtual ~BordersTest() {}
@@ -78,6 +79,6 @@ TEST_P(BordersTest, TestLowBitrate) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
VP9_INSTANTIATE_TEST_CASE(BordersTest, ::testing::Values(
::libvpx_test::kTwoPassGood));
VP9_INSTANTIATE_TEST_CASE(BordersTest,
::testing::Values(::libvpx_test::kTwoPassGood));
} // namespace

View File

@@ -36,29 +36,26 @@ struct ByteAlignmentTestParam {
};
const ByteAlignmentTestParam kBaTestParams[] = {
{kLegacyByteAlignment, VPX_CODEC_OK, true},
{32, VPX_CODEC_OK, true},
{64, VPX_CODEC_OK, true},
{128, VPX_CODEC_OK, true},
{256, VPX_CODEC_OK, true},
{512, VPX_CODEC_OK, true},
{1024, VPX_CODEC_OK, true},
{1, VPX_CODEC_INVALID_PARAM, false},
{-2, VPX_CODEC_INVALID_PARAM, false},
{4, VPX_CODEC_INVALID_PARAM, false},
{16, VPX_CODEC_INVALID_PARAM, false},
{255, VPX_CODEC_INVALID_PARAM, false},
{2048, VPX_CODEC_INVALID_PARAM, false},
{ kLegacyByteAlignment, VPX_CODEC_OK, true },
{ 32, VPX_CODEC_OK, true },
{ 64, VPX_CODEC_OK, true },
{ 128, VPX_CODEC_OK, true },
{ 256, VPX_CODEC_OK, true },
{ 512, VPX_CODEC_OK, true },
{ 1024, VPX_CODEC_OK, true },
{ 1, VPX_CODEC_INVALID_PARAM, false },
{ -2, VPX_CODEC_INVALID_PARAM, false },
{ 4, VPX_CODEC_INVALID_PARAM, false },
{ 16, VPX_CODEC_INVALID_PARAM, false },
{ 255, VPX_CODEC_INVALID_PARAM, false },
{ 2048, VPX_CODEC_INVALID_PARAM, false },
};
// Class for testing byte alignment of reference buffers.
class ByteAlignmentTest
: public ::testing::TestWithParam<ByteAlignmentTestParam> {
protected:
ByteAlignmentTest()
: video_(NULL),
decoder_(NULL),
md5_file_(NULL) {}
ByteAlignmentTest() : video_(NULL), decoder_(NULL), md5_file_(NULL) {}
virtual void SetUp() {
video_ = new libvpx_test::WebMVideoSource(kVP9TestFile);
@@ -74,8 +71,7 @@ class ByteAlignmentTest
}
virtual void TearDown() {
if (md5_file_ != NULL)
fclose(md5_file_);
if (md5_file_ != NULL) fclose(md5_file_);
delete decoder_;
delete video_;
@@ -89,8 +85,7 @@ class ByteAlignmentTest
const vpx_codec_err_t res =
decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
CheckDecodedFrames(byte_alignment_to_check);
if (res == VPX_CODEC_OK)
video_->Next();
if (res == VPX_CODEC_OK) video_->Next();
return res;
}
@@ -98,8 +93,7 @@ class ByteAlignmentTest
for (; video_->cxdata() != NULL; video_->Next()) {
const vpx_codec_err_t res =
decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
if (res != VPX_CODEC_OK)
return res;
if (res != VPX_CODEC_OK) return res;
CheckDecodedFrames(byte_alignment_to_check);
}
return VPX_CODEC_OK;
@@ -135,7 +129,7 @@ class ByteAlignmentTest
void OpenMd5File(const std::string &md5_file_name_) {
md5_file_ = libvpx_test::OpenTestDataFile(md5_file_name_);
ASSERT_TRUE(md5_file_ != NULL) << "MD5 file open failed. Filename: "
<< md5_file_name_;
<< md5_file_name_;
}
void CheckMd5(const vpx_image_t &img) {
@@ -163,8 +157,8 @@ class ByteAlignmentTest
TEST_F(ByteAlignmentTest, SwitchByteAlignment) {
const int num_elements = 14;
const int byte_alignments[] = { 0, 32, 64, 128, 256, 512, 1024,
0, 1024, 32, 512, 64, 256, 128 };
const int byte_alignments[] = { 0, 32, 64, 128, 256, 512, 1024,
0, 1024, 32, 512, 64, 256, 128 };
for (int i = 0; i < num_elements; ++i) {
SetByteAlignment(byte_alignments[i], VPX_CODEC_OK);

View File

@@ -12,7 +12,7 @@
#include "./vpx_config.h"
#if ARCH_X86 || ARCH_X86_64
# include "vpx_ports/x86.h"
#include "vpx_ports/x86.h"
#endif
namespace libvpx_test {

View File

@@ -32,15 +32,12 @@ class CodecFactory {
virtual ~CodecFactory() {}
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
unsigned long deadline) const = 0;
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg) const = 0;
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags,
unsigned long deadline) // NOLINT(runtime/int)
const = 0;
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags) const = 0;
virtual Encoder* CreateEncoder(vpx_codec_enc_cfg_t cfg,
virtual Encoder *CreateEncoder(vpx_codec_enc_cfg_t cfg,
unsigned long deadline,
const unsigned long init_flags,
TwopassStatsStore *stats) const = 0;
@@ -53,19 +50,25 @@ class CodecFactory {
* to avoid having to include a pointer to the CodecFactory in every test
* definition.
*/
template<class T1>
class CodecTestWithParam : public ::testing::TestWithParam<
std::tr1::tuple< const libvpx_test::CodecFactory*, T1 > > {
};
template <class T1>
class CodecTestWithParam
: public ::testing::TestWithParam<
std::tr1::tuple<const libvpx_test::CodecFactory *, T1> > {};
template<class T1, class T2>
class CodecTestWith2Params : public ::testing::TestWithParam<
std::tr1::tuple< const libvpx_test::CodecFactory*, T1, T2 > > {
};
template <class T1, class T2>
class CodecTestWith2Params
: public ::testing::TestWithParam<
std::tr1::tuple<const libvpx_test::CodecFactory *, T1, T2> > {};
template<class T1, class T2, class T3>
class CodecTestWith3Params : public ::testing::TestWithParam<
std::tr1::tuple< const libvpx_test::CodecFactory*, T1, T2, T3 > > {
template <class T1, class T2, class T3>
class CodecTestWith3Params
: public ::testing::TestWithParam<
std::tr1::tuple<const libvpx_test::CodecFactory *, T1, T2, T3> > {};
template <class T1, class T2, class T3, class T4>
class CodecTestWith4Params
: public ::testing::TestWithParam<
std::tr1::tuple<const libvpx_test::CodecFactory *, T1, T2, T3, T4> > {
};
/*
@@ -74,15 +77,13 @@ class CodecTestWith3Params : public ::testing::TestWithParam<
#if CONFIG_VP8
class VP8Decoder : public Decoder {
public:
VP8Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
: Decoder(cfg, deadline) {}
explicit VP8Decoder(vpx_codec_dec_cfg_t cfg) : Decoder(cfg) {}
VP8Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
unsigned long deadline) // NOLINT
: Decoder(cfg, flag, deadline) {}
VP8Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag)
: Decoder(cfg, flag) {}
protected:
virtual vpx_codec_iface_t* CodecInterface() const {
virtual vpx_codec_iface_t *CodecInterface() const {
#if CONFIG_VP8_DECODER
return &vpx_codec_vp8_dx_algo;
#else
@@ -98,7 +99,7 @@ class VP8Encoder : public Encoder {
: Encoder(cfg, deadline, init_flags, stats) {}
protected:
virtual vpx_codec_iface_t* CodecInterface() const {
virtual vpx_codec_iface_t *CodecInterface() const {
#if CONFIG_VP8_ENCODER
return &vpx_codec_vp8_cx_algo;
#else
@@ -111,28 +112,32 @@ class VP8CodecFactory : public CodecFactory {
public:
VP8CodecFactory() : CodecFactory() {}
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
unsigned long deadline) const {
return CreateDecoder(cfg, 0, deadline);
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg) const {
return CreateDecoder(cfg, 0);
}
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags,
unsigned long deadline) const { // NOLINT
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags) const {
#if CONFIG_VP8_DECODER
return new VP8Decoder(cfg, flags, deadline);
return new VP8Decoder(cfg, flags);
#else
(void)cfg;
(void)flags;
return NULL;
#endif
}
virtual Encoder* CreateEncoder(vpx_codec_enc_cfg_t cfg,
virtual Encoder *CreateEncoder(vpx_codec_enc_cfg_t cfg,
unsigned long deadline,
const unsigned long init_flags,
TwopassStatsStore *stats) const {
#if CONFIG_VP8_ENCODER
return new VP8Encoder(cfg, deadline, init_flags, stats);
#else
(void)cfg;
(void)deadline;
(void)init_flags;
(void)stats;
return NULL;
#endif
}
@@ -142,6 +147,8 @@ class VP8CodecFactory : public CodecFactory {
#if CONFIG_VP8_ENCODER
return vpx_codec_enc_config_default(&vpx_codec_vp8_cx_algo, cfg, usage);
#else
(void)cfg;
(void)usage;
return VPX_CODEC_INCAPABLE;
#endif
}
@@ -149,32 +156,30 @@ class VP8CodecFactory : public CodecFactory {
const libvpx_test::VP8CodecFactory kVP8;
#define VP8_INSTANTIATE_TEST_CASE(test, ...)\
INSTANTIATE_TEST_CASE_P(VP8, test, \
::testing::Combine( \
::testing::Values(static_cast<const libvpx_test::CodecFactory*>( \
&libvpx_test::kVP8)), \
#define VP8_INSTANTIATE_TEST_CASE(test, ...) \
INSTANTIATE_TEST_CASE_P( \
VP8, test, \
::testing::Combine( \
::testing::Values(static_cast<const libvpx_test::CodecFactory *>( \
&libvpx_test::kVP8)), \
__VA_ARGS__))
#else
#define VP8_INSTANTIATE_TEST_CASE(test, ...)
#endif // CONFIG_VP8
/*
* VP9 Codec Definitions
*/
#if CONFIG_VP9
class VP9Decoder : public Decoder {
public:
VP9Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
: Decoder(cfg, deadline) {}
explicit VP9Decoder(vpx_codec_dec_cfg_t cfg) : Decoder(cfg) {}
VP9Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
unsigned long deadline) // NOLINT
: Decoder(cfg, flag, deadline) {}
VP9Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag)
: Decoder(cfg, flag) {}
protected:
virtual vpx_codec_iface_t* CodecInterface() const {
virtual vpx_codec_iface_t *CodecInterface() const {
#if CONFIG_VP9_DECODER
return &vpx_codec_vp9_dx_algo;
#else
@@ -190,7 +195,7 @@ class VP9Encoder : public Encoder {
: Encoder(cfg, deadline, init_flags, stats) {}
protected:
virtual vpx_codec_iface_t* CodecInterface() const {
virtual vpx_codec_iface_t *CodecInterface() const {
#if CONFIG_VP9_ENCODER
return &vpx_codec_vp9_cx_algo;
#else
@@ -203,28 +208,32 @@ class VP9CodecFactory : public CodecFactory {
public:
VP9CodecFactory() : CodecFactory() {}
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
unsigned long deadline) const {
return CreateDecoder(cfg, 0, deadline);
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg) const {
return CreateDecoder(cfg, 0);
}
virtual Decoder* CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags,
unsigned long deadline) const { // NOLINT
virtual Decoder *CreateDecoder(vpx_codec_dec_cfg_t cfg,
const vpx_codec_flags_t flags) const {
#if CONFIG_VP9_DECODER
return new VP9Decoder(cfg, flags, deadline);
return new VP9Decoder(cfg, flags);
#else
(void)cfg;
(void)flags;
return NULL;
#endif
}
virtual Encoder* CreateEncoder(vpx_codec_enc_cfg_t cfg,
virtual Encoder *CreateEncoder(vpx_codec_enc_cfg_t cfg,
unsigned long deadline,
const unsigned long init_flags,
TwopassStatsStore *stats) const {
#if CONFIG_VP9_ENCODER
return new VP9Encoder(cfg, deadline, init_flags, stats);
#else
(void)cfg;
(void)deadline;
(void)init_flags;
(void)stats;
return NULL;
#endif
}
@@ -234,6 +243,8 @@ class VP9CodecFactory : public CodecFactory {
#if CONFIG_VP9_ENCODER
return vpx_codec_enc_config_default(&vpx_codec_vp9_cx_algo, cfg, usage);
#else
(void)cfg;
(void)usage;
return VPX_CODEC_INCAPABLE;
#endif
}
@@ -241,11 +252,12 @@ class VP9CodecFactory : public CodecFactory {
const libvpx_test::VP9CodecFactory kVP9;
#define VP9_INSTANTIATE_TEST_CASE(test, ...)\
INSTANTIATE_TEST_CASE_P(VP9, test, \
::testing::Combine( \
::testing::Values(static_cast<const libvpx_test::CodecFactory*>( \
&libvpx_test::kVP9)), \
#define VP9_INSTANTIATE_TEST_CASE(test, ...) \
INSTANTIATE_TEST_CASE_P( \
VP9, test, \
::testing::Combine( \
::testing::Values(static_cast<const libvpx_test::CodecFactory *>( \
&libvpx_test::kVP9)), \
__VA_ARGS__))
#else
#define VP9_INSTANTIATE_TEST_CASE(test, ...)

View File

@@ -15,11 +15,13 @@
namespace {
class ConfigTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class ConfigTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
ConfigTest() : EncoderTest(GET_PARAM(0)),
frame_count_in_(0), frame_count_out_(0), frame_count_max_(0) {}
ConfigTest()
: EncoderTest(GET_PARAM(0)), frame_count_in_(0), frame_count_out_(0),
frame_count_max_(0) {}
virtual ~ConfigTest() {}
virtual void SetUp() {
@@ -32,12 +34,12 @@ class ConfigTest : public ::libvpx_test::EncoderTest,
frame_count_out_ = 0;
}
virtual void PreEncodeFrameHook(libvpx_test::VideoSource* /*video*/) {
virtual void PreEncodeFrameHook(libvpx_test::VideoSource * /*video*/) {
++frame_count_in_;
abort_ |= (frame_count_in_ >= frame_count_max_);
}
virtual void FramePktHook(const vpx_codec_cx_pkt_t* /*pkt*/) {
virtual void FramePktHook(const vpx_codec_cx_pkt_t * /*pkt*/) {
++frame_count_out_;
}

View File

@@ -26,12 +26,10 @@
#include "vpx_dsp/ssim.h"
#include "vpx_mem/vpx_mem.h"
extern "C"
double vpx_get_ssim_metrics(uint8_t *img1, int img1_pitch,
uint8_t *img2, int img2_pitch,
int width, int height,
Ssimv *sv2, Metrics *m,
int do_inconsistency);
extern "C" double vpx_get_ssim_metrics(uint8_t *img1, int img1_pitch,
uint8_t *img2, int img2_pitch, int width,
int height, Ssimv *sv2, Metrics *m,
int do_inconsistency);
using libvpx_test::ACMRandom;
@@ -41,20 +39,18 @@ class ConsistencyTestBase : public ::testing::Test {
ConsistencyTestBase(int width, int height) : width_(width), height_(height) {}
static void SetUpTestCase() {
source_data_[0] = reinterpret_cast<uint8_t*>(
source_data_[0] = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
reference_data_[0] = reinterpret_cast<uint8_t*>(
reference_data_[0] = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
source_data_[1] = reinterpret_cast<uint8_t*>(
source_data_[1] = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
reference_data_[1] = reinterpret_cast<uint8_t*>(
reference_data_[1] = reinterpret_cast<uint8_t *>(
vpx_memalign(kDataAlignment, kDataBufferSize));
ssim_array_ = new Ssimv[kDataBufferSize / 16];
}
static void ClearSsim() {
memset(ssim_array_, 0, kDataBufferSize / 16);
}
static void ClearSsim() { memset(ssim_array_, 0, kDataBufferSize / 16); }
static void TearDownTestCase() {
vpx_free(source_data_[0]);
source_data_[0] = NULL;
@@ -68,14 +64,12 @@ class ConsistencyTestBase : public ::testing::Test {
delete[] ssim_array_;
}
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
// Handle frames up to 640x480
static const int kDataAlignment = 16;
static const int kDataBufferSize = 640*480;
static const int kDataBufferSize = 640 * 480;
virtual void SetUp() {
source_stride_ = (width_ + 31) & ~31;
@@ -122,9 +116,9 @@ class ConsistencyTestBase : public ::testing::Test {
}
}
int width_, height_;
static uint8_t* source_data_[2];
static uint8_t *source_data_[2];
int source_stride_;
static uint8_t* reference_data_[2];
static uint8_t *reference_data_[2];
int reference_stride_;
static Ssimv *ssim_array_;
Metrics metrics_;
@@ -142,18 +136,17 @@ class ConsistencyVP9Test
protected:
double CheckConsistency(int frame) {
EXPECT_LT(frame, 2)<< "Frame to check has to be less than 2.";
return
vpx_get_ssim_metrics(source_data_[frame], source_stride_,
reference_data_[frame], reference_stride_,
width_, height_, ssim_array_, &metrics_, 1);
EXPECT_LT(frame, 2) << "Frame to check has to be less than 2.";
return vpx_get_ssim_metrics(source_data_[frame], source_stride_,
reference_data_[frame], reference_stride_,
width_, height_, ssim_array_, &metrics_, 1);
}
};
#endif // CONFIG_VP9_ENCODER
uint8_t* ConsistencyTestBase::source_data_[2] = {NULL, NULL};
uint8_t* ConsistencyTestBase::reference_data_[2] = {NULL, NULL};
Ssimv* ConsistencyTestBase::ssim_array_ = NULL;
uint8_t *ConsistencyTestBase::source_data_[2] = { NULL, NULL };
uint8_t *ConsistencyTestBase::reference_data_[2] = { NULL, NULL };
Ssimv *ConsistencyTestBase::ssim_array_ = NULL;
#if CONFIG_VP9_ENCODER
TEST_P(ConsistencyVP9Test, ConsistencyIsZero) {
@@ -205,7 +198,6 @@ TEST_P(ConsistencyVP9Test, ConsistencyIsZero) {
}
#endif // CONFIG_VP9_ENCODER
using std::tr1::make_tuple;
//------------------------------------------------------------------------------
@@ -213,9 +205,7 @@ using std::tr1::make_tuple;
#if CONFIG_VP9_ENCODER
const ConsistencyParam c_vp9_tests[] = {
make_tuple(320, 240),
make_tuple(318, 242),
make_tuple(318, 238),
make_tuple(320, 240), make_tuple(318, 242), make_tuple(318, 238),
};
INSTANTIATE_TEST_CASE_P(C, ConsistencyVP9Test,
::testing::ValuesIn(c_vp9_tests));

File diff suppressed because it is too large Load Diff

View File

@@ -23,10 +23,8 @@ class CpuSpeedTest
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
protected:
CpuSpeedTest()
: EncoderTest(GET_PARAM(0)),
encoding_mode_(GET_PARAM(1)),
set_cpu_used_(GET_PARAM(2)),
min_psnr_(kMaxPSNR),
: EncoderTest(GET_PARAM(0)), encoding_mode_(GET_PARAM(1)),
set_cpu_used_(GET_PARAM(2)), min_psnr_(kMaxPSNR),
tune_content_(VP9E_CONTENT_DEFAULT) {}
virtual ~CpuSpeedTest() {}
@@ -42,9 +40,7 @@ class CpuSpeedTest
}
}
virtual void BeginPassHook(unsigned int /*pass*/) {
min_psnr_ = kMaxPSNR;
}
virtual void BeginPassHook(unsigned int /*pass*/) { min_psnr_ = kMaxPSNR; }
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
@@ -61,8 +57,7 @@ class CpuSpeedTest
}
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
if (pkt->data.psnr.psnr[0] < min_psnr_)
min_psnr_ = pkt->data.psnr.psnr[0];
if (pkt->data.psnr.psnr[0] < min_psnr_) min_psnr_ = pkt->data.psnr.psnr[0];
}
::libvpx_test::TestMode encoding_mode_;
@@ -153,9 +148,9 @@ TEST_P(CpuSpeedTest, TestLowBitrate) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
VP9_INSTANTIATE_TEST_CASE(
CpuSpeedTest,
::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kOnePassGood,
::libvpx_test::kRealTime),
::testing::Range(0, 9));
VP9_INSTANTIATE_TEST_CASE(CpuSpeedTest,
::testing::Values(::libvpx_test::kTwoPassGood,
::libvpx_test::kOnePassGood,
::libvpx_test::kRealTime),
::testing::Range(0, 9));
} // namespace

View File

@@ -24,14 +24,12 @@ const int kCQLevelStep = 8;
const unsigned int kCQTargetBitrate = 2000;
class CQTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<int> {
public ::libvpx_test::CodecTestWithParam<int> {
public:
// maps the cqlevel to the bitrate produced.
typedef std::map<int, uint32_t> BitrateMap;
static void SetUpTestCase() {
bitrates_.clear();
}
static void SetUpTestCase() { bitrates_.clear(); }
static void TearDownTestCase() {
ASSERT_TRUE(!HasFailure())
@@ -128,7 +126,6 @@ TEST_P(CQTest, LinearPSNRIsHigherForCQLevel) {
EXPECT_GE(cq_psnr_lin, vbr_psnr_lin);
}
VP8_INSTANTIATE_TEST_CASE(CQTest,
::testing::Range(kCQLevelMin, kCQLevelMax,
kCQLevelStep));
VP8_INSTANTIATE_TEST_CASE(CQTest, ::testing::Range(kCQLevelMin, kCQLevelMax,
kCQLevelStep));
} // namespace

View File

@@ -1,6 +1,6 @@
#!/bin/sh
##
## Copyright (c) 2014 The WebM project authors. All Rights Reserved.
## Copyright (c) 2016 The WebM project authors. All Rights Reserved.
##
## Use of this source code is governed by a BSD-style license
## that can be found in the LICENSE file in the root of the source
@@ -8,30 +8,27 @@
## in the file PATENTS. All contributing project authors may
## be found in the AUTHORS file in the root of the source tree.
##
## This file tests the libvpx vp8cx_set_ref example. To add new tests to this
## This file tests the libvpx cx_set_ref example. To add new tests to this
## file, do the following:
## 1. Write a shell function (this is your test).
## 2. Add the function to vp8cx_set_ref_tests (on a new line).
## 2. Add the function to cx_set_ref_tests (on a new line).
##
. $(dirname $0)/tools_common.sh
# Environment check: $YUV_RAW_INPUT is required.
vp8cx_set_ref_verify_environment() {
cx_set_ref_verify_environment() {
if [ ! -e "${YUV_RAW_INPUT}" ]; then
echo "Libvpx test data must exist in LIBVPX_TEST_DATA_PATH."
return 1
fi
}
# Runs vp8cx_set_ref and updates the reference frame before encoding frame 90.
# $1 is the codec name, which vp8cx_set_ref does not support at present: It's
# currently used only to name the output file.
# TODO(tomfinegan): Pass the codec param once the example is updated to support
# VP9.
# Runs cx_set_ref and updates the reference frame before encoding frame 90.
# $1 is the codec name.
vpx_set_ref() {
local encoder="${LIBVPX_BIN_PATH}/vp8cx_set_ref${VPX_TEST_EXE_SUFFIX}"
local codec="$1"
local output_file="${VPX_TEST_OUTPUT_DIR}/vp8cx_set_ref_${codec}.ivf"
local encoder="${LIBVPX_BIN_PATH}/${codec}cx_set_ref${VPX_TEST_EXE_SUFFIX}"
local output_file="${VPX_TEST_OUTPUT_DIR}/${codec}cx_set_ref_${codec}.ivf"
local ref_frame_num=90
if [ ! -x "${encoder}" ]; then
@@ -46,12 +43,18 @@ vpx_set_ref() {
[ -e "${output_file}" ] || return 1
}
vp8cx_set_ref_vp8() {
cx_set_ref_vp8() {
if [ "$(vp8_encode_available)" = "yes" ]; then
vpx_set_ref vp8 || return 1
fi
}
vp8cx_set_ref_tests="vp8cx_set_ref_vp8"
cx_set_ref_vp9() {
if [ "$(vp9_encode_available)" = "yes" ]; then
vpx_set_ref vp9 || return 1
fi
}
run_tests vp8cx_set_ref_verify_environment "${vp8cx_set_ref_tests}"
cx_set_ref_tests="cx_set_ref_vp8 cx_set_ref_vp9"
run_tests cx_set_ref_verify_environment "${cx_set_ref_tests}"

View File

@@ -18,8 +18,9 @@
namespace {
class DatarateTestLarge : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class DatarateTestLarge
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
public:
DatarateTestLarge() : EncoderTest(GET_PARAM(0)) {}
@@ -29,6 +30,7 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
virtual void SetUp() {
InitializeConfig();
SetMode(GET_PARAM(1));
set_cpu_used_ = GET_PARAM(2);
ResetModel();
}
@@ -41,12 +43,16 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
duration_ = 0.0;
denoiser_offon_test_ = 0;
denoiser_offon_period_ = -1;
gf_boost_ = 0;
}
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
if (video->frame() == 0)
if (video->frame() == 0) {
encoder->Control(VP8E_SET_NOISE_SENSITIVITY, denoiser_on_);
encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
encoder->Control(VP8E_SET_GF_CBR_BOOST_PCT, gf_boost_);
}
if (denoiser_offon_test_) {
ASSERT_GT(denoiser_offon_period_, 0)
@@ -71,8 +77,7 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
// http://code.google.com/p/webm/issues/detail?id=496 is fixed.
// For now the codec assumes buffer starts at starting buffer rate
// plus one frame's time.
if (last_pts_ == 0)
duration = 1;
if (last_pts_ == 0) duration = 1;
// Add to the buffer the bits we'd expect from a constant bitrate server.
bits_in_buffer_model_ += static_cast<int64_t>(
@@ -83,11 +88,11 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
* empty - and then stop showing frames until we've got enough bits to
* show one. As noted in comment below (issue 495), this does not currently
* apply to key frames. For now exclude key frames in condition below. */
const bool key_frame = (pkt->data.frame.flags & VPX_FRAME_IS_KEY)
? true: false;
const bool key_frame =
(pkt->data.frame.flags & VPX_FRAME_IS_KEY) ? true : false;
if (!key_frame) {
ASSERT_GE(bits_in_buffer_model_, 0) << "Buffer Underrun at frame "
<< pkt->data.frame.pts;
<< pkt->data.frame.pts;
}
const int64_t frame_size_in_bits = pkt->data.frame.sz * 8;
@@ -99,8 +104,7 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
bits_total_ += frame_size_in_bits;
// If first drop not set and we have a drop set it to this time.
if (!first_drop_ && duration > 1)
first_drop_ = last_pts_ + 1;
if (!first_drop_ && duration > 1) first_drop_ = last_pts_ + 1;
// Update the most recent pts.
last_pts_ = pkt->data.frame.pts;
@@ -119,8 +123,8 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
duration_ = (last_pts_ + 1) * timebase_;
// Effective file datarate includes the time spent prebuffering.
effective_datarate_ = (bits_total_ - bits_in_last_frame_) / 1000.0
/ (cfg_.rc_buf_initial_sz / 1000.0 + duration_);
effective_datarate_ = (bits_total_ - bits_in_last_frame_) / 1000.0 /
(cfg_.rc_buf_initial_sz / 1000.0 + duration_);
file_datarate_ = file_size_in_kb / duration_;
}
@@ -139,6 +143,8 @@ class DatarateTestLarge : public ::libvpx_test::EncoderTest,
int denoiser_on_;
int denoiser_offon_test_;
int denoiser_offon_period_;
int set_cpu_used_;
int gf_boost_;
};
#if CONFIG_TEMPORAL_DENOISING
@@ -156,9 +162,6 @@ TEST_P(DatarateTestLarge, DenoiserLevels) {
// For the temporal denoiser (#if CONFIG_TEMPORAL_DENOISING) the level j
// refers to the 4 denoiser modes: denoiserYonly, denoiserOnYUV,
// denoiserOnAggressive, and denoiserOnAdaptive.
// For the spatial denoiser (if !CONFIG_TEMPORAL_DENOISING), the level j
// refers to the blur thresholds: 20, 40, 60 80.
// The j = 0 case (denoiser off) is covered in the tests below.
denoiser_on_ = j;
cfg_.rc_target_bitrate = 300;
ResetModel();
@@ -166,7 +169,7 @@ TEST_P(DatarateTestLarge, DenoiserLevels) {
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.3)
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
}
@@ -190,7 +193,7 @@ TEST_P(DatarateTestLarge, DenoiserOffOn) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.3)
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
#endif // CONFIG_TEMPORAL_DENOISING
@@ -221,8 +224,7 @@ TEST_P(DatarateTestLarge, BasicBufferModel) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.3)
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
}
@@ -256,8 +258,202 @@ TEST_P(DatarateTestLarge, ChangingDropFrameThresh) {
}
}
class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
// Disabled for tsan, see:
// https://bugs.chromium.org/p/webm/issues/detail?id=1049
#if defined(__has_feature)
#if __has_feature(thread_sanitizer)
#define BUILDING_WITH_TSAN
#endif
#endif
#ifndef BUILDING_WITH_TSAN
TEST_P(DatarateTestLarge, DropFramesMultiThreads) {
denoiser_on_ = 0;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 30;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
cfg_.g_threads = 2;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 140);
cfg_.rc_target_bitrate = 200;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
#endif // !BUILDING_WITH_TSAN
class DatarateTestRealTime : public DatarateTestLarge {
public:
virtual ~DatarateTestRealTime() {}
};
#if CONFIG_TEMPORAL_DENOISING
// Check basic datarate targeting, for a single bitrate, but loop over the
// various denoiser settings.
TEST_P(DatarateTestRealTime, DenoiserLevels) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 1;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 140);
for (int j = 1; j < 5; ++j) {
// Run over the denoiser levels.
// For the temporal denoiser (#if CONFIG_TEMPORAL_DENOISING) the level j
// refers to the 4 denoiser modes: denoiserYonly, denoiserOnYUV,
// denoiserOnAggressive, and denoiserOnAdaptive.
denoiser_on_ = j;
cfg_.rc_target_bitrate = 300;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
}
// Check basic datarate targeting, for a single bitrate, when denoiser is off
// and on.
TEST_P(DatarateTestRealTime, DenoiserOffOn) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 1;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 299);
cfg_.rc_target_bitrate = 300;
ResetModel();
// The denoiser is off by default.
denoiser_on_ = 0;
// Set the offon test flag.
denoiser_offon_test_ = 1;
denoiser_offon_period_ = 100;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
#endif // CONFIG_TEMPORAL_DENOISING
TEST_P(DatarateTestRealTime, BasicBufferModel) {
denoiser_on_ = 0;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 1;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
// 2 pass cbr datarate control has a bug hidden by the small # of
// frames selected in this encode. The problem is that even if the buffer is
// negative we produce a keyframe on a cutscene, ignoring datarate
// constraints
// TODO(jimbankoski): Fix when issue
// http://bugs.chromium.org/p/webm/issues/detail?id=495 is addressed.
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 140);
// There is an issue for low bitrates in real-time mode, where the
// effective_datarate slightly overshoots the target bitrate.
// This is same the issue as noted above (#495).
// TODO(jimbankoski/marpan): Update test to run for lower bitrates (< 100),
// when the issue is resolved.
for (int i = 100; i <= 700; i += 200) {
cfg_.rc_target_bitrate = i;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
}
TEST_P(DatarateTestRealTime, ChangingDropFrameThresh) {
denoiser_on_ = 0;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_max_quantizer = 36;
cfg_.rc_end_usage = VPX_CBR;
cfg_.rc_target_bitrate = 200;
cfg_.kf_mode = VPX_KF_DISABLED;
const int frame_count = 40;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, frame_count);
// Check that the first dropped frame gets earlier and earlier
// as the drop frame threshold is increased.
const int kDropFrameThreshTestStep = 30;
vpx_codec_pts_t last_drop = frame_count;
for (int i = 1; i < 91; i += kDropFrameThreshTestStep) {
cfg_.rc_dropframe_thresh = i;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_LE(first_drop_, last_drop)
<< " The first dropped frame for drop_thresh " << i
<< " > first dropped frame for drop_thresh "
<< i - kDropFrameThreshTestStep;
last_drop = first_drop_;
}
}
// Disabled for tsan, see:
// https://bugs.chromium.org/p/webm/issues/detail?id=1049
#ifndef BUILDING_WITH_TSAN
TEST_P(DatarateTestRealTime, DropFramesMultiThreads) {
denoiser_on_ = 0;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 30;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
// Encode using multiple threads.
cfg_.g_threads = 2;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 140);
cfg_.rc_target_bitrate = 200;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
#endif
TEST_P(DatarateTestRealTime, GFBoost) {
denoiser_on_ = 0;
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_dropframe_thresh = 0;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
cfg_.g_error_resilient = 0;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 300);
cfg_.rc_target_bitrate = 300;
ResetModel();
// Apply a gf boost.
gf_boost_ = 50;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, effective_datarate_ * 0.95)
<< " The datarate for the file exceeds the target!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.4)
<< " The datarate for the file missed the target!";
}
class DatarateTestVP9Large
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
public:
DatarateTestVP9Large() : EncoderTest(GET_PARAM(0)) {}
@@ -307,8 +503,8 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
if (num_temp_layers == 2) {
if (frame_num % 2 == 0) {
// Layer 0: predict from L and ARF, update L.
frame_flags = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
frame_flags =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
} else {
// Layer 1: predict from L, G and ARF, and update G.
frame_flags = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
@@ -317,15 +513,15 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
} else if (num_temp_layers == 3) {
if (frame_num % 4 == 0) {
// Layer 0: predict from L and ARF; update L.
frame_flags = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_REF_GF;
frame_flags =
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_REF_GF;
} else if ((frame_num - 2) % 4 == 0) {
// Layer 1: predict from L, G, ARF; update G.
frame_flags = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
} else if ((frame_num - 1) % 2 == 0) {
} else if ((frame_num - 1) % 2 == 0) {
// Layer 2: predict from L, G, ARF; update none.
frame_flags = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST;
frame_flags =
VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
}
}
return frame_flags;
@@ -353,8 +549,7 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
if (video->frame() == 0)
encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
if (video->frame() == 0) encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
if (denoiser_offon_test_) {
ASSERT_GT(denoiser_offon_period_, 0)
@@ -374,8 +569,8 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
vpx_svc_layer_id_t layer_id;
layer_id.spatial_layer_id = 0;
frame_flags_ = SetFrameFlags(video->frame(), cfg_.ts_number_layers);
layer_id.temporal_layer_id = SetLayerId(video->frame(),
cfg_.ts_number_layers);
layer_id.temporal_layer_id =
SetLayerId(video->frame(), cfg_.ts_number_layers);
encoder->Control(VP9E_SET_SVC_LAYER_ID, &layer_id);
}
const vpx_rational_t tb = video->timebase();
@@ -383,15 +578,13 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
duration_ = 0;
}
virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
// Time since last timestamp = duration.
vpx_codec_pts_t duration = pkt->data.frame.pts - last_pts_;
if (duration > 1) {
// If first drop not set and we have a drop set it to this time.
if (!first_drop_)
first_drop_ = last_pts_ + 1;
if (!first_drop_) first_drop_ = last_pts_ + 1;
// Update the number of frame drops.
num_drops_ += static_cast<int>(duration - 1);
// Update counter for total number of frames (#frames input to encoder).
@@ -407,7 +600,7 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
// Buffer should not go negative.
ASSERT_GE(bits_in_buffer_model_, 0) << "Buffer Underrun at frame "
<< pkt->data.frame.pts;
<< pkt->data.frame.pts;
const size_t frame_size_in_bits = pkt->data.frame.sz * 8;
@@ -425,7 +618,7 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
virtual void EndPassHook(void) {
for (int layer = 0; layer < static_cast<int>(cfg_.ts_number_layers);
++layer) {
++layer) {
duration_ = (last_pts_ + 1) * timebase_;
if (bits_total_[layer]) {
// Effective file datarate:
@@ -450,8 +643,8 @@ class DatarateTestVP9Large : public ::libvpx_test::EncoderTest,
int denoiser_offon_period_;
};
// Check basic rate targeting for VBR mode.
TEST_P(DatarateTestVP9Large, BasicRateTargetingVBR) {
// Check basic rate targeting for VBR mode with 0 lag.
TEST_P(DatarateTestVP9Large, BasicRateTargetingVBRLagZero) {
cfg_.rc_min_quantizer = 0;
cfg_.rc_max_quantizer = 63;
cfg_.g_error_resilient = 0;
@@ -471,7 +664,34 @@ TEST_P(DatarateTestVP9Large, BasicRateTargetingVBR) {
}
}
// Check basic rate targeting for CBR,
// Check basic rate targeting for VBR mode with non-zero lag.
TEST_P(DatarateTestVP9Large, BasicRateTargetingVBRLagNonZero) {
cfg_.rc_min_quantizer = 0;
cfg_.rc_max_quantizer = 63;
cfg_.g_error_resilient = 0;
cfg_.rc_end_usage = VPX_VBR;
// For non-zero lag, rate control will work (be within bounds) for
// real-time mode.
if (deadline_ == VPX_DL_REALTIME) {
cfg_.g_lag_in_frames = 15;
} else {
cfg_.g_lag_in_frames = 0;
}
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 300);
for (int i = 400; i <= 800; i += 400) {
cfg_.rc_target_bitrate = i;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.75)
<< " The datarate for the file is lower than target by too much!";
ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.25)
<< " The datarate for the file is greater than target by too much!";
}
}
// Check basic rate targeting for CBR mode.
TEST_P(DatarateTestVP9Large, BasicRateTargeting) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
@@ -495,6 +715,30 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting) {
}
}
// Check basic rate targeting for CBR mode, with 2 threads and dropped frames.
TEST_P(DatarateTestVP9Large, BasicRateTargetingDropFramesMultiThreads) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_dropframe_thresh = 30;
cfg_.rc_min_quantizer = 0;
cfg_.rc_max_quantizer = 63;
cfg_.rc_end_usage = VPX_CBR;
cfg_.g_lag_in_frames = 0;
// Encode using multiple threads.
cfg_.g_threads = 2;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 140);
cfg_.rc_target_bitrate = 200;
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.85)
<< " The datarate for the file is lower than target by too much!";
ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.15)
<< " The datarate for the file is greater than target by too much!";
}
// Check basic rate targeting for CBR.
TEST_P(DatarateTestVP9Large, BasicRateTargeting444) {
::libvpx_test::Y4mVideoSource video("rush_hour_444.y4m", 0, 140);
@@ -520,7 +764,7 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting444) {
ASSERT_LE(static_cast<double>(cfg_.rc_target_bitrate),
effective_datarate_[0] * 1.15)
<< " The datarate for the file missed the target!"
<< cfg_.rc_target_bitrate << " "<< effective_datarate_;
<< cfg_.rc_target_bitrate << " " << effective_datarate_;
}
}
@@ -590,8 +834,7 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting2TemporalLayers) {
cfg_.temporal_layering_mode = VP9E_TEMPORAL_LAYERING_MODE_BYPASS;
if (deadline_ == VPX_DL_REALTIME)
cfg_.g_error_resilient = 1;
if (deadline_ == VPX_DL_REALTIME) cfg_.g_error_resilient = 1;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 200);
@@ -605,10 +848,12 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting2TemporalLayers) {
for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.85)
<< " The datarate for the file is lower than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.15)
<< " The datarate for the file is greater than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
}
}
}
@@ -648,12 +893,14 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayers) {
// Adjust the thresholds to be tighter than .75.
ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.75)
<< " The datarate for the file is lower than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
// TODO(yaowu): Work out more stable rc control strategy and
// Adjust the thresholds to be tighter than 1.25.
ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.25)
<< " The datarate for the file is greater than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
}
}
}
@@ -693,10 +940,12 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayersFrameDropping) {
for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
ASSERT_GE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 0.85)
<< " The datarate for the file is lower than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
ASSERT_LE(effective_datarate_[j], cfg_.layer_target_bitrate[j] * 1.15)
<< " The datarate for the file is greater than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
// Expect some frame drops in this test: for this 200 frames test,
// expect at least 10% and not more than 60% drops.
ASSERT_GE(num_drops_, 20);
@@ -705,8 +954,13 @@ TEST_P(DatarateTestVP9Large, BasicRateTargeting3TemporalLayersFrameDropping) {
}
#if CONFIG_VP9_TEMPORAL_DENOISING
class DatarateTestVP9LargeDenoiser : public DatarateTestVP9Large {
public:
virtual ~DatarateTestVP9LargeDenoiser() {}
};
// Check basic datarate targeting, for a single bitrate, when denoiser is on.
TEST_P(DatarateTestVP9Large, DenoiserLevels) {
TEST_P(DatarateTestVP9LargeDenoiser, LowNoise) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
@@ -733,9 +987,37 @@ TEST_P(DatarateTestVP9Large, DenoiserLevels) {
<< " The datarate for the file is greater than target by too much!";
}
// Check basic datarate targeting, for a single bitrate, when denoiser is on,
// for clip with high noise level.
TEST_P(DatarateTestVP9LargeDenoiser, HighNoise) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_dropframe_thresh = 1;
cfg_.rc_min_quantizer = 2;
cfg_.rc_max_quantizer = 56;
cfg_.rc_end_usage = VPX_CBR;
cfg_.g_lag_in_frames = 0;
::libvpx_test::Y4mVideoSource video("noisy_clip_640_360.y4m", 0, 200);
// For the temporal denoiser (#if CONFIG_VP9_TEMPORAL_DENOISING),
// there is only one denoiser mode: denoiserYonly(which is 1),
// but may add more modes in the future.
cfg_.rc_target_bitrate = 1000;
ResetModel();
// Turn on the denoiser.
denoiser_on_ = 1;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(effective_datarate_[0], cfg_.rc_target_bitrate * 0.85)
<< " The datarate for the file is lower than target by too much!";
ASSERT_LE(effective_datarate_[0], cfg_.rc_target_bitrate * 1.15)
<< " The datarate for the file is greater than target by too much!";
}
// Check basic datarate targeting, for a single bitrate, when denoiser is off
// and on.
TEST_P(DatarateTestVP9Large, DenoiserOffOn) {
TEST_P(DatarateTestVP9LargeDenoiser, DenoiserOffOn) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
@@ -766,11 +1048,15 @@ TEST_P(DatarateTestVP9Large, DenoiserOffOn) {
}
#endif // CONFIG_VP9_TEMPORAL_DENOISING
class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
class DatarateOnePassCbrSvc
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
public:
DatarateOnePassCbrSvc() : EncoderTest(GET_PARAM(0)) {}
DatarateOnePassCbrSvc() : EncoderTest(GET_PARAM(0)) {
memset(&svc_params_, 0, sizeof(svc_params_));
}
virtual ~DatarateOnePassCbrSvc() {}
protected:
virtual void SetUp() {
InitializeConfig();
@@ -788,8 +1074,7 @@ class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
mismatch_psnr_ = 0.0;
mismatch_nframes_ = 0;
}
virtual void BeginPassHook(unsigned int /*pass*/) {
}
virtual void BeginPassHook(unsigned int /*pass*/) {}
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
if (video->frame() == 0) {
@@ -798,12 +1083,17 @@ class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
svc_params_.max_quantizers[i] = 63;
svc_params_.min_quantizers[i] = 0;
}
svc_params_.speed_per_layer[0] = 5;
for (i = 1; i < VPX_SS_MAX_LAYERS; ++i) {
svc_params_.speed_per_layer[i] = speed_setting_;
}
encoder->Control(VP9E_SET_SVC, 1);
encoder->Control(VP9E_SET_SVC_PARAMETERS, &svc_params_);
encoder->Control(VP8E_SET_CPUUSED, speed_setting_);
encoder->Control(VP9E_SET_TILE_COLUMNS, 0);
encoder->Control(VP8E_SET_MAX_INTRA_BITRATE_PCT, 300);
encoder->Control(VP9E_SET_TILE_COLUMNS, (cfg_.g_threads >> 1));
encoder->Control(VP8E_SET_STATIC_THRESHOLD, 1);
}
const vpx_rational_t tb = video->timebase();
timebase_ = static_cast<double>(tb.num) / tb.den;
@@ -811,21 +1101,21 @@ class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
}
virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
vpx_codec_pts_t duration = pkt->data.frame.pts - last_pts_;
if (last_pts_ == 0)
duration = 1;
if (last_pts_ == 0) duration = 1;
bits_in_buffer_model_ += static_cast<int64_t>(
duration * timebase_ * cfg_.rc_target_bitrate * 1000);
const bool key_frame = (pkt->data.frame.flags & VPX_FRAME_IS_KEY)
? true: false;
const bool key_frame =
(pkt->data.frame.flags & VPX_FRAME_IS_KEY) ? true : false;
if (!key_frame) {
ASSERT_GE(bits_in_buffer_model_, 0) << "Buffer Underrun at frame "
<< pkt->data.frame.pts;
// TODO(marpan): This check currently fails for some of the SVC tests,
// re-enable when issue (webm:1350) is resolved.
// ASSERT_GE(bits_in_buffer_model_, 0) << "Buffer Underrun at frame "
// << pkt->data.frame.pts;
}
const size_t frame_size_in_bits = pkt->data.frame.sz * 8;
bits_in_buffer_model_ -= frame_size_in_bits;
bits_in_buffer_model_ -= static_cast<int64_t>(frame_size_in_bits);
bits_total_ += frame_size_in_bits;
if (!first_drop_ && duration > 1)
first_drop_ = last_pts_ + 1;
if (!first_drop_ && duration > 1) first_drop_ = last_pts_ + 1;
last_pts_ = pkt->data.frame.pts;
bits_in_last_frame_ = frame_size_in_bits;
++frame_number_;
@@ -838,16 +1128,13 @@ class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
}
}
virtual void MismatchHook(const vpx_image_t *img1,
const vpx_image_t *img2) {
virtual void MismatchHook(const vpx_image_t *img1, const vpx_image_t *img2) {
double mismatch_psnr = compute_psnr(img1, img2);
mismatch_psnr_ += mismatch_psnr;
++mismatch_nframes_;
}
unsigned int GetMismatchFrames() {
return mismatch_nframes_;
}
unsigned int GetMismatchFrames() { return mismatch_nframes_; }
vpx_codec_pts_t last_pts_;
int64_t bits_in_buffer_model_;
@@ -864,37 +1151,31 @@ class DatarateOnePassCbrSvc : public ::libvpx_test::EncoderTest,
int mismatch_nframes_;
};
static void assign_layer_bitrates(vpx_codec_enc_cfg_t *const enc_cfg,
const vpx_svc_extra_cfg_t *svc_params,
int spatial_layers,
int temporal_layers,
int temporal_layering_mode) {
const vpx_svc_extra_cfg_t *svc_params,
int spatial_layers, int temporal_layers,
int temporal_layering_mode) {
int sl, spatial_layer_target;
float total = 0;
float alloc_ratio[VPX_MAX_LAYERS] = {0};
float alloc_ratio[VPX_MAX_LAYERS] = { 0 };
for (sl = 0; sl < spatial_layers; ++sl) {
if (svc_params->scaling_factor_den[sl] > 0) {
alloc_ratio[sl] = (float)(svc_params->scaling_factor_num[sl] *
1.0 / svc_params->scaling_factor_den[sl]);
alloc_ratio[sl] = (float)(svc_params->scaling_factor_num[sl] * 1.0 /
svc_params->scaling_factor_den[sl]);
total += alloc_ratio[sl];
}
}
for (sl = 0; sl < spatial_layers; ++sl) {
enc_cfg->ss_target_bitrate[sl] = spatial_layer_target =
(unsigned int)(enc_cfg->rc_target_bitrate *
alloc_ratio[sl] / total);
(unsigned int)(enc_cfg->rc_target_bitrate * alloc_ratio[sl] / total);
const int index = sl * temporal_layers;
if (temporal_layering_mode == 3) {
enc_cfg->layer_target_bitrate[index] =
spatial_layer_target >> 1;
enc_cfg->layer_target_bitrate[index] = spatial_layer_target >> 1;
enc_cfg->layer_target_bitrate[index + 1] =
(spatial_layer_target >> 1) + (spatial_layer_target >> 2);
enc_cfg->layer_target_bitrate[index + 2] =
spatial_layer_target;
enc_cfg->layer_target_bitrate[index + 2] = spatial_layer_target;
} else if (temporal_layering_mode == 2) {
enc_cfg->layer_target_bitrate[index] =
spatial_layer_target * 2 / 3;
enc_cfg->layer_target_bitrate[index + 1] =
spatial_layer_target;
enc_cfg->layer_target_bitrate[index] = spatial_layer_target * 2 / 3;
enc_cfg->layer_target_bitrate[index + 1] = spatial_layer_target;
}
}
}
@@ -925,16 +1206,16 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayers) {
cfg_.kf_max_dist = 9999;
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 200);
// TODO(wonkap/marpan): Check that effective_datarate for each layer hits the
// TODO(marpan): Check that effective_datarate for each layer hits the
// layer target_bitrate.
for (int i = 200; i <= 800; i += 200) {
cfg_.rc_target_bitrate = i;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
@@ -973,10 +1254,10 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayersSmallKf) {
cfg_.kf_max_dist = j;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
@@ -1007,15 +1288,14 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayers4threads) {
svc_params_.scaling_factor_den[1] = 288;
cfg_.rc_dropframe_thresh = 10;
cfg_.kf_max_dist = 9999;
::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
30, 1, 0, 300);
::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 0, 300);
cfg_.rc_target_bitrate = 800;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.15)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
@@ -1047,15 +1327,14 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayers) {
svc_params_.scaling_factor_den[2] = 288;
cfg_.rc_dropframe_thresh = 10;
cfg_.kf_max_dist = 9999;
::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
30, 1, 0, 300);
::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 0, 300);
cfg_.rc_target_bitrate = 800;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.22)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
@@ -1086,8 +1365,7 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayersSmallKf) {
svc_params_.scaling_factor_num[2] = 288;
svc_params_.scaling_factor_den[2] = 288;
cfg_.rc_dropframe_thresh = 10;
::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
30, 1, 0, 300);
::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 0, 300);
cfg_.rc_target_bitrate = 800;
// For this 3 temporal layer case, pattern repeats every 4 frames, so choose
// 4 key neighboring key frame periods (so key frame will land on 0-2-1-2).
@@ -1095,10 +1373,10 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayersSmallKf) {
cfg_.kf_max_dist = j;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.30)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
@@ -1131,25 +1409,67 @@ TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc3SpatialLayers4threads) {
svc_params_.scaling_factor_den[2] = 288;
cfg_.rc_dropframe_thresh = 10;
cfg_.kf_max_dist = 9999;
::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720,
30, 1, 0, 300);
::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 0, 300);
cfg_.rc_target_bitrate = 800;
ResetModel();
assign_layer_bitrates(&cfg_, &svc_params_, cfg_.ss_number_layers,
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
cfg_.ts_number_layers, cfg_.temporal_layering_mode);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_GE(cfg_.rc_target_bitrate, file_datarate_ * 0.85)
<< " The datarate for the file exceeds the target by too much!";
<< " The datarate for the file exceeds the target by too much!";
ASSERT_LE(cfg_.rc_target_bitrate, file_datarate_ * 1.22)
<< " The datarate for the file is lower than the target by too much!";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
}
VP8_INSTANTIATE_TEST_CASE(DatarateTestLarge, ALL_TEST_MODES);
// Run SVC encoder for 1 temporal layer, 2 spatial layers, with spatial
// downscale 5x5.
TEST_P(DatarateOnePassCbrSvc, OnePassCbrSvc2SpatialLayers5x5MultipleRuns) {
cfg_.rc_buf_initial_sz = 500;
cfg_.rc_buf_optimal_sz = 500;
cfg_.rc_buf_sz = 1000;
cfg_.rc_min_quantizer = 0;
cfg_.rc_max_quantizer = 63;
cfg_.rc_end_usage = VPX_CBR;
cfg_.g_lag_in_frames = 0;
cfg_.ss_number_layers = 2;
cfg_.ts_number_layers = 1;
cfg_.ts_rate_decimator[0] = 1;
cfg_.g_error_resilient = 1;
cfg_.g_threads = 3;
cfg_.temporal_layering_mode = 0;
svc_params_.scaling_factor_num[0] = 256;
svc_params_.scaling_factor_den[0] = 1280;
svc_params_.scaling_factor_num[1] = 1280;
svc_params_.scaling_factor_den[1] = 1280;
cfg_.rc_dropframe_thresh = 0;
cfg_.kf_max_dist = 999999;
cfg_.kf_min_dist = 0;
cfg_.ss_target_bitrate[0] = 300;
cfg_.ss_target_bitrate[1] = 1400;
cfg_.layer_target_bitrate[0] = 300;
cfg_.layer_target_bitrate[1] = 1400;
cfg_.rc_target_bitrate = 1700;
::libvpx_test::Y4mVideoSource video("niklas_1280_720_30.y4m", 0, 300);
ResetModel();
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
}
VP8_INSTANTIATE_TEST_CASE(DatarateTestLarge, ALL_TEST_MODES,
::testing::Values(0));
VP8_INSTANTIATE_TEST_CASE(DatarateTestRealTime,
::testing::Values(::libvpx_test::kRealTime),
::testing::Values(-6, -12));
VP9_INSTANTIATE_TEST_CASE(DatarateTestVP9Large,
::testing::Values(::libvpx_test::kOnePassGood,
::libvpx_test::kRealTime),
::testing::Range(2, 9));
#if CONFIG_VP9_TEMPORAL_DENOISING
VP9_INSTANTIATE_TEST_CASE(DatarateTestVP9LargeDenoiser,
::testing::Values(::libvpx_test::kRealTime),
::testing::Range(5, 9));
#endif
VP9_INSTANTIATE_TEST_CASE(DatarateOnePassCbrSvc,
::testing::Values(::libvpx_test::kRealTime),
::testing::Range(5, 9));

View File

@@ -25,20 +25,12 @@
#include "vpx/vpx_codec.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/mem.h"
#include "vpx_ports/msvc.h" // for round()
using libvpx_test::ACMRandom;
namespace {
#ifdef _MSC_VER
static int round(double x) {
if (x < 0)
return static_cast<int>(ceil(x - 0.5));
else
return static_cast<int>(floor(x + 0.5));
}
#endif
const int kNumCoeffs = 256;
const double C1 = 0.995184726672197;
const double C2 = 0.98078528040323;
@@ -62,16 +54,16 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
double temp1, temp2;
// step 1
step[ 0] = input[0] + input[15];
step[ 1] = input[1] + input[14];
step[ 2] = input[2] + input[13];
step[ 3] = input[3] + input[12];
step[ 4] = input[4] + input[11];
step[ 5] = input[5] + input[10];
step[ 6] = input[6] + input[ 9];
step[ 7] = input[7] + input[ 8];
step[ 8] = input[7] - input[ 8];
step[ 9] = input[6] - input[ 9];
step[0] = input[0] + input[15];
step[1] = input[1] + input[14];
step[2] = input[2] + input[13];
step[3] = input[3] + input[12];
step[4] = input[4] + input[11];
step[5] = input[5] + input[10];
step[6] = input[6] + input[9];
step[7] = input[7] + input[8];
step[8] = input[7] - input[8];
step[9] = input[6] - input[9];
step[10] = input[5] - input[10];
step[11] = input[4] - input[11];
step[12] = input[3] - input[12];
@@ -89,13 +81,13 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
output[6] = step[1] - step[6];
output[7] = step[0] - step[7];
temp1 = step[ 8] * C7;
temp1 = step[8] * C7;
temp2 = step[15] * C9;
output[ 8] = temp1 + temp2;
output[8] = temp1 + temp2;
temp1 = step[ 9] * C11;
temp1 = step[9] * C11;
temp2 = step[14] * C5;
output[ 9] = temp1 - temp2;
output[9] = temp1 - temp2;
temp1 = step[10] * C3;
temp2 = step[13] * C13;
@@ -113,40 +105,40 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
temp2 = step[13] * C3;
output[13] = temp2 - temp1;
temp1 = step[ 9] * C5;
temp1 = step[9] * C5;
temp2 = step[14] * C11;
output[14] = temp2 + temp1;
temp1 = step[ 8] * C9;
temp1 = step[8] * C9;
temp2 = step[15] * C7;
output[15] = temp2 - temp1;
// step 3
step[ 0] = output[0] + output[3];
step[ 1] = output[1] + output[2];
step[ 2] = output[1] - output[2];
step[ 3] = output[0] - output[3];
step[0] = output[0] + output[3];
step[1] = output[1] + output[2];
step[2] = output[1] - output[2];
step[3] = output[0] - output[3];
temp1 = output[4] * C14;
temp2 = output[7] * C2;
step[ 4] = temp1 + temp2;
step[4] = temp1 + temp2;
temp1 = output[5] * C10;
temp2 = output[6] * C6;
step[ 5] = temp1 + temp2;
step[5] = temp1 + temp2;
temp1 = output[5] * C6;
temp2 = output[6] * C10;
step[ 6] = temp2 - temp1;
step[6] = temp2 - temp1;
temp1 = output[4] * C2;
temp2 = output[7] * C14;
step[ 7] = temp2 - temp1;
step[7] = temp2 - temp1;
step[ 8] = output[ 8] + output[11];
step[ 9] = output[ 9] + output[10];
step[10] = output[ 9] - output[10];
step[11] = output[ 8] - output[11];
step[8] = output[8] + output[11];
step[9] = output[9] + output[10];
step[10] = output[9] - output[10];
step[11] = output[8] - output[11];
step[12] = output[12] + output[15];
step[13] = output[13] + output[14];
@@ -154,25 +146,25 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
step[15] = output[12] - output[15];
// step 4
output[ 0] = (step[ 0] + step[ 1]);
output[ 8] = (step[ 0] - step[ 1]);
output[0] = (step[0] + step[1]);
output[8] = (step[0] - step[1]);
temp1 = step[2] * C12;
temp2 = step[3] * C4;
temp1 = temp1 + temp2;
output[ 4] = 2*(temp1 * C8);
output[4] = 2 * (temp1 * C8);
temp1 = step[2] * C4;
temp2 = step[3] * C12;
temp1 = temp2 - temp1;
output[12] = 2 * (temp1 * C8);
output[ 2] = 2 * ((step[4] + step[ 5]) * C8);
output[14] = 2 * ((step[7] - step[ 6]) * C8);
output[2] = 2 * ((step[4] + step[5]) * C8);
output[14] = 2 * ((step[7] - step[6]) * C8);
temp1 = step[4] - step[5];
temp2 = step[6] + step[7];
output[ 6] = (temp1 + temp2);
output[6] = (temp1 + temp2);
output[10] = (temp1 - temp2);
intermediate[8] = step[8] + step[14];
@@ -188,18 +180,18 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
temp1 = temp2 + temp1;
output[13] = 2 * (temp1 * C8);
output[ 9] = 2 * ((step[10] + step[11]) * C8);
output[9] = 2 * ((step[10] + step[11]) * C8);
intermediate[11] = step[10] - step[11];
intermediate[12] = step[12] + step[13];
intermediate[13] = step[12] - step[13];
intermediate[14] = step[ 8] - step[14];
intermediate[15] = step[ 9] - step[15];
intermediate[14] = step[8] - step[14];
intermediate[15] = step[9] - step[15];
output[15] = (intermediate[11] + intermediate[12]);
output[ 1] = -(intermediate[11] - intermediate[12]);
output[1] = -(intermediate[11] - intermediate[12]);
output[ 7] = 2 * (intermediate[13] * C8);
output[7] = 2 * (intermediate[13] * C8);
temp1 = intermediate[14] * C12;
temp2 = intermediate[15] * C4;
@@ -209,28 +201,24 @@ void butterfly_16x16_dct_1d(double input[16], double output[16]) {
temp1 = intermediate[14] * C4;
temp2 = intermediate[15] * C12;
temp1 = temp2 + temp1;
output[ 5] = 2 * (temp1 * C8);
output[5] = 2 * (temp1 * C8);
}
void reference_16x16_dct_2d(int16_t input[256], double output[256]) {
// First transform columns
for (int i = 0; i < 16; ++i) {
double temp_in[16], temp_out[16];
for (int j = 0; j < 16; ++j)
temp_in[j] = input[j * 16 + i];
for (int j = 0; j < 16; ++j) temp_in[j] = input[j * 16 + i];
butterfly_16x16_dct_1d(temp_in, temp_out);
for (int j = 0; j < 16; ++j)
output[j * 16 + i] = temp_out[j];
for (int j = 0; j < 16; ++j) output[j * 16 + i] = temp_out[j];
}
// Then transform rows
for (int i = 0; i < 16; ++i) {
double temp_in[16], temp_out[16];
for (int j = 0; j < 16; ++j)
temp_in[j] = output[j + i * 16];
for (int j = 0; j < 16; ++j) temp_in[j] = output[j + i * 16];
butterfly_16x16_dct_1d(temp_in, temp_out);
// Scale by some magic number
for (int j = 0; j < 16; ++j)
output[j + i * 16] = temp_out[j]/2;
for (int j = 0; j < 16; ++j) output[j + i * 16] = temp_out[j] / 2;
}
}
@@ -256,8 +244,7 @@ void idct16x16_ref(const tran_low_t *in, uint8_t *dest, int stride,
vpx_idct16x16_256_add_c(in, dest, stride);
}
void fht16x16_ref(const int16_t *in, tran_low_t *out, int stride,
int tx_type) {
void fht16x16_ref(const int16_t *in, tran_low_t *out, int stride, int tx_type) {
vp9_fht16x16_c(in, out, stride, tx_type);
}
@@ -359,11 +346,10 @@ class Trans16x16TestBase {
}
}
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(test_input_block,
test_temp_block, pitch_));
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(test_input_block, test_temp_block, pitch_));
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(test_temp_block, dst, pitch_));
ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(
@@ -374,18 +360,17 @@ class Trans16x16TestBase {
for (int j = 0; j < kNumCoeffs; ++j) {
#if CONFIG_VP9_HIGHBITDEPTH
const int32_t diff =
bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
bit_depth_ == VPX_BITS_8 ? dst[j] - src[j] : dst16[j] - src16[j];
#else
const int32_t diff = dst[j] - src[j];
#endif
const uint32_t error = diff * diff;
if (max_error < error)
max_error = error;
if (max_error < error) max_error = error;
total_error += error;
}
}
EXPECT_GE(1u << 2 * (bit_depth_ - 8), max_error)
EXPECT_GE(1u << 2 * (bit_depth_ - 8), max_error)
<< "Error: 16x16 FHT/IHT has an individual round trip error > 1";
EXPECT_GE(count_test_block << 2 * (bit_depth_ - 8), total_error)
@@ -401,8 +386,9 @@ class Trans16x16TestBase {
for (int i = 0; i < count_test_block; ++i) {
// Initialize a test block with input range [-mask_, mask_].
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
}
fwd_txfm_ref(input_block, output_ref_block, pitch_, tx_type_);
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_block, output_block, pitch_));
@@ -426,16 +412,14 @@ class Trans16x16TestBase {
input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
}
if (i == 0) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = mask_;
} else if (i == 1) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = -mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = -mask_;
}
fwd_txfm_ref(input_extreme_block, output_ref_block, pitch_, tx_type_);
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_extreme_block,
output_block, pitch_));
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(input_extreme_block, output_block, pitch_));
// The minimum quant value is 4.
for (int j = 0; j < kNumCoeffs; ++j) {
@@ -464,12 +448,12 @@ class Trans16x16TestBase {
for (int j = 0; j < kNumCoeffs; ++j) {
input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
}
if (i == 0)
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = mask_;
if (i == 1)
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = -mask_;
if (i == 0) {
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = mask_;
}
if (i == 1) {
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = -mask_;
}
fwd_txfm_ref(input_extreme_block, output_ref_block, pitch_, tx_type_);
@@ -483,8 +467,9 @@ class Trans16x16TestBase {
// quantization with maximum allowed step sizes
output_ref_block[0] = (output_ref_block[0] / dc_thred) * dc_thred;
for (int j = 1; j < kNumCoeffs; ++j)
for (int j = 1; j < kNumCoeffs; ++j) {
output_ref_block[j] = (output_ref_block[j] / ac_thred) * ac_thred;
}
if (bit_depth_ == VPX_BITS_8) {
inv_txfm_ref(output_ref_block, ref, pitch_, tx_type_);
ASM_REGISTER_STATE_CHECK(RunInvTxfm(output_ref_block, dst, pitch_));
@@ -492,17 +477,15 @@ class Trans16x16TestBase {
} else {
inv_txfm_ref(output_ref_block, CONVERT_TO_BYTEPTR(ref16), pitch_,
tx_type_);
ASM_REGISTER_STATE_CHECK(RunInvTxfm(output_ref_block,
CONVERT_TO_BYTEPTR(dst16), pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(output_ref_block, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif
}
if (bit_depth_ == VPX_BITS_8) {
for (int j = 0; j < kNumCoeffs; ++j)
EXPECT_EQ(ref[j], dst[j]);
for (int j = 0; j < kNumCoeffs; ++j) EXPECT_EQ(ref[j], dst[j]);
#if CONFIG_VP9_HIGHBITDEPTH
} else {
for (int j = 0; j < kNumCoeffs; ++j)
EXPECT_EQ(ref16[j], dst16[j]);
for (int j = 0; j < kNumCoeffs; ++j) EXPECT_EQ(ref16[j], dst16[j]);
#endif
}
}
@@ -539,15 +522,16 @@ class Trans16x16TestBase {
}
reference_16x16_dct_2d(in, out_r);
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
coeff[j] = static_cast<tran_low_t>(round(out_r[j]));
}
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, 16));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
16));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16), 16));
#endif // CONFIG_VP9_HIGHBITDEPTH
}
@@ -559,9 +543,8 @@ class Trans16x16TestBase {
const uint32_t diff = dst[j] - src[j];
#endif // CONFIG_VP9_HIGHBITDEPTH
const uint32_t error = diff * diff;
EXPECT_GE(1u, error)
<< "Error: 16x16 IDCT has error " << error
<< " at index " << j;
EXPECT_GE(1u, error) << "Error: 16x16 IDCT has error " << error
<< " at index " << j;
}
}
}
@@ -603,8 +586,8 @@ class Trans16x16TestBase {
} else {
#if CONFIG_VP9_HIGHBITDEPTH
ref_txfm(coeff, CONVERT_TO_BYTEPTR(ref16), pitch_);
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif // CONFIG_VP9_HIGHBITDEPTH
}
@@ -616,9 +599,8 @@ class Trans16x16TestBase {
const uint32_t diff = dst[j] - ref[j];
#endif // CONFIG_VP9_HIGHBITDEPTH
const uint32_t error = diff * diff;
EXPECT_EQ(0u, error)
<< "Error: 16x16 IDCT Comparison has error " << error
<< " at index " << j;
EXPECT_EQ(0u, error) << "Error: 16x16 IDCT Comparison has error "
<< error << " at index " << j;
}
}
}
@@ -631,32 +613,25 @@ class Trans16x16TestBase {
IhtFunc inv_txfm_ref;
};
class Trans16x16DCT
: public Trans16x16TestBase,
public ::testing::TestWithParam<Dct16x16Param> {
class Trans16x16DCT : public Trans16x16TestBase,
public ::testing::TestWithParam<Dct16x16Param> {
public:
virtual ~Trans16x16DCT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
tx_type_ = GET_PARAM(2);
bit_depth_ = GET_PARAM(3);
pitch_ = 16;
pitch_ = 16;
fwd_txfm_ref = fdct16x16_ref;
inv_txfm_ref = idct16x16_ref;
mask_ = (1 << bit_depth_) - 1;
#if CONFIG_VP9_HIGHBITDEPTH
switch (bit_depth_) {
case VPX_BITS_10:
inv_txfm_ref = idct16x16_10_ref;
break;
case VPX_BITS_12:
inv_txfm_ref = idct16x16_12_ref;
break;
default:
inv_txfm_ref = idct16x16_ref;
break;
case VPX_BITS_10: inv_txfm_ref = idct16x16_10_ref; break;
case VPX_BITS_12: inv_txfm_ref = idct16x16_12_ref; break;
default: inv_txfm_ref = idct16x16_ref; break;
}
#else
inv_txfm_ref = idct16x16_ref;
@@ -676,17 +651,11 @@ class Trans16x16DCT
IdctFunc inv_txfm_;
};
TEST_P(Trans16x16DCT, AccuracyCheck) {
RunAccuracyCheck();
}
TEST_P(Trans16x16DCT, AccuracyCheck) { RunAccuracyCheck(); }
TEST_P(Trans16x16DCT, CoeffCheck) {
RunCoeffCheck();
}
TEST_P(Trans16x16DCT, CoeffCheck) { RunCoeffCheck(); }
TEST_P(Trans16x16DCT, MemCheck) {
RunMemCheck();
}
TEST_P(Trans16x16DCT, MemCheck) { RunMemCheck(); }
TEST_P(Trans16x16DCT, QuantCheck) {
// Use maximally allowed quantization step sizes for DC and AC
@@ -694,36 +663,27 @@ TEST_P(Trans16x16DCT, QuantCheck) {
RunQuantCheck(1336, 1828);
}
TEST_P(Trans16x16DCT, InvAccuracyCheck) {
RunInvAccuracyCheck();
}
TEST_P(Trans16x16DCT, InvAccuracyCheck) { RunInvAccuracyCheck(); }
class Trans16x16HT
: public Trans16x16TestBase,
public ::testing::TestWithParam<Ht16x16Param> {
class Trans16x16HT : public Trans16x16TestBase,
public ::testing::TestWithParam<Ht16x16Param> {
public:
virtual ~Trans16x16HT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
tx_type_ = GET_PARAM(2);
bit_depth_ = GET_PARAM(3);
pitch_ = 16;
pitch_ = 16;
fwd_txfm_ref = fht16x16_ref;
inv_txfm_ref = iht16x16_ref;
mask_ = (1 << bit_depth_) - 1;
#if CONFIG_VP9_HIGHBITDEPTH
switch (bit_depth_) {
case VPX_BITS_10:
inv_txfm_ref = iht16x16_10;
break;
case VPX_BITS_12:
inv_txfm_ref = iht16x16_12;
break;
default:
inv_txfm_ref = iht16x16_ref;
break;
case VPX_BITS_10: inv_txfm_ref = iht16x16_10; break;
case VPX_BITS_12: inv_txfm_ref = iht16x16_12; break;
default: inv_txfm_ref = iht16x16_ref; break;
}
#else
inv_txfm_ref = iht16x16_ref;
@@ -743,17 +703,11 @@ class Trans16x16HT
IhtFunc inv_txfm_;
};
TEST_P(Trans16x16HT, AccuracyCheck) {
RunAccuracyCheck();
}
TEST_P(Trans16x16HT, AccuracyCheck) { RunAccuracyCheck(); }
TEST_P(Trans16x16HT, CoeffCheck) {
RunCoeffCheck();
}
TEST_P(Trans16x16HT, CoeffCheck) { RunCoeffCheck(); }
TEST_P(Trans16x16HT, MemCheck) {
RunMemCheck();
}
TEST_P(Trans16x16HT, MemCheck) { RunMemCheck(); }
TEST_P(Trans16x16HT, QuantCheck) {
// The encoder skips any non-DC intra prediction modes,
@@ -761,9 +715,8 @@ TEST_P(Trans16x16HT, QuantCheck) {
RunQuantCheck(429, 729);
}
class InvTrans16x16DCT
: public Trans16x16TestBase,
public ::testing::TestWithParam<Idct16x16Param> {
class InvTrans16x16DCT : public Trans16x16TestBase,
public ::testing::TestWithParam<Idct16x16Param> {
public:
virtual ~InvTrans16x16DCT() {}
@@ -774,7 +727,7 @@ class InvTrans16x16DCT
bit_depth_ = GET_PARAM(3);
pitch_ = 16;
mask_ = (1 << bit_depth_) - 1;
}
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
@@ -792,9 +745,8 @@ TEST_P(InvTrans16x16DCT, CompareReference) {
CompareInvReference(ref_txfm_, thresh_);
}
class PartialTrans16x16Test
: public ::testing::TestWithParam<
std::tr1::tuple<FdctFunc, vpx_bit_depth_t> > {
class PartialTrans16x16Test : public ::testing::TestWithParam<
std::tr1::tuple<FdctFunc, vpx_bit_depth_t> > {
public:
virtual ~PartialTrans16x16Test() {}
virtual void SetUp() {
@@ -863,10 +815,10 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vpx_highbd_fdct16x16_c, &idct16x16_12, 0, VPX_BITS_12),
make_tuple(&vpx_fdct16x16_c, &vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
#else
INSTANTIATE_TEST_CASE_P(
C, Trans16x16DCT,
::testing::Values(
make_tuple(&vpx_fdct16x16_c, &vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(C, Trans16x16DCT,
::testing::Values(make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
0, VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if CONFIG_VP9_HIGHBITDEPTH
@@ -903,31 +855,28 @@ INSTANTIATE_TEST_CASE_P(C, PartialTrans16x16Test,
VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
NEON, Trans16x16DCT,
::testing::Values(
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_neon, 0, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct16x16_c, &vpx_idct16x16_256_add_neon,
0, VPX_BITS_8)));
#endif
#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, Trans16x16DCT,
::testing::Values(
make_tuple(&vpx_fdct16x16_sse2,
&vpx_idct16x16_256_add_sse2, 0, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct16x16_sse2,
&vpx_idct16x16_256_add_sse2, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
SSE2, Trans16x16HT,
::testing::Values(
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 0,
VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 1,
VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 2,
VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2, 3,
VPX_BITS_8)));
::testing::Values(make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2,
0, VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2,
1, VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2,
2, VPX_BITS_8),
make_tuple(&vp9_fht16x16_sse2, &vp9_iht16x16_256_add_sse2,
3, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans16x16Test,
::testing::Values(make_tuple(&vpx_fdct16x16_1_sse2,
VPX_BITS_8)));
@@ -937,16 +886,14 @@ INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans16x16Test,
INSTANTIATE_TEST_CASE_P(
SSE2, Trans16x16DCT,
::testing::Values(
make_tuple(&vpx_highbd_fdct16x16_sse2,
&idct16x16_10, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct16x16_c,
&idct16x16_256_add_10_sse2, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct16x16_sse2,
&idct16x16_12, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct16x16_c,
&idct16x16_256_add_12_sse2, 0, VPX_BITS_12),
make_tuple(&vpx_fdct16x16_sse2,
&vpx_idct16x16_256_add_c, 0, VPX_BITS_8)));
make_tuple(&vpx_highbd_fdct16x16_sse2, &idct16x16_10, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct16x16_c, &idct16x16_256_add_10_sse2, 0,
VPX_BITS_10),
make_tuple(&vpx_highbd_fdct16x16_sse2, &idct16x16_12, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct16x16_c, &idct16x16_256_add_12_sse2, 0,
VPX_BITS_12),
make_tuple(&vpx_fdct16x16_sse2, &vpx_idct16x16_256_add_c, 0,
VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
SSE2, Trans16x16HT,
::testing::Values(
@@ -959,26 +906,24 @@ INSTANTIATE_TEST_CASE_P(
// that to test both branches.
INSTANTIATE_TEST_CASE_P(
SSE2, InvTrans16x16DCT,
::testing::Values(
make_tuple(&idct16x16_10_add_10_c,
&idct16x16_10_add_10_sse2, 3167, VPX_BITS_10),
make_tuple(&idct16x16_10,
&idct16x16_256_add_10_sse2, 3167, VPX_BITS_10),
make_tuple(&idct16x16_10_add_12_c,
&idct16x16_10_add_12_sse2, 3167, VPX_BITS_12),
make_tuple(&idct16x16_12,
&idct16x16_256_add_12_sse2, 3167, VPX_BITS_12)));
::testing::Values(make_tuple(&idct16x16_10_add_10_c,
&idct16x16_10_add_10_sse2, 3167, VPX_BITS_10),
make_tuple(&idct16x16_10, &idct16x16_256_add_10_sse2,
3167, VPX_BITS_10),
make_tuple(&idct16x16_10_add_12_c,
&idct16x16_10_add_12_sse2, 3167, VPX_BITS_12),
make_tuple(&idct16x16_12, &idct16x16_256_add_12_sse2,
3167, VPX_BITS_12)));
INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans16x16Test,
::testing::Values(make_tuple(&vpx_fdct16x16_1_sse2,
VPX_BITS_8)));
#endif // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
MSA, Trans16x16DCT,
::testing::Values(
make_tuple(&vpx_fdct16x16_msa,
&vpx_idct16x16_256_add_msa, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(MSA, Trans16x16DCT,
::testing::Values(make_tuple(&vpx_fdct16x16_msa,
&vpx_idct16x16_256_add_msa,
0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
MSA, Trans16x16HT,
::testing::Values(

View File

@@ -25,18 +25,11 @@
#include "vpx/vpx_codec.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/mem.h"
#include "vpx_ports/msvc.h" // for round()
using libvpx_test::ACMRandom;
namespace {
#ifdef _MSC_VER
static int round(double x) {
if (x < 0)
return static_cast<int>(ceil(x - 0.5));
else
return static_cast<int>(floor(x + 0.5));
}
#endif
const int kNumCoeffs = 1024;
const double kPi = 3.141592653589793238462643383279502884;
@@ -44,10 +37,10 @@ void reference_32x32_dct_1d(const double in[32], double out[32]) {
const double kInvSqrt2 = 0.707106781186547524400844362104;
for (int k = 0; k < 32; k++) {
out[k] = 0.0;
for (int n = 0; n < 32; n++)
for (int n = 0; n < 32; n++) {
out[k] += in[n] * cos(kPi * (2 * n + 1) * k / 64.0);
if (k == 0)
out[k] = out[k] * kInvSqrt2;
}
if (k == 0) out[k] = out[k] * kInvSqrt2;
}
}
@@ -56,21 +49,17 @@ void reference_32x32_dct_2d(const int16_t input[kNumCoeffs],
// First transform columns
for (int i = 0; i < 32; ++i) {
double temp_in[32], temp_out[32];
for (int j = 0; j < 32; ++j)
temp_in[j] = input[j*32 + i];
for (int j = 0; j < 32; ++j) temp_in[j] = input[j * 32 + i];
reference_32x32_dct_1d(temp_in, temp_out);
for (int j = 0; j < 32; ++j)
output[j * 32 + i] = temp_out[j];
for (int j = 0; j < 32; ++j) output[j * 32 + i] = temp_out[j];
}
// Then transform rows
for (int i = 0; i < 32; ++i) {
double temp_in[32], temp_out[32];
for (int j = 0; j < 32; ++j)
temp_in[j] = output[j + i*32];
for (int j = 0; j < 32; ++j) temp_in[j] = output[j + i * 32];
reference_32x32_dct_1d(temp_in, temp_out);
// Scale by some magic number
for (int j = 0; j < 32; ++j)
output[j + i * 32] = temp_out[j] / 4;
for (int j = 0; j < 32; ++j) output[j + i * 32] = temp_out[j] / 4;
}
}
@@ -96,8 +85,8 @@ class Trans32x32Test : public ::testing::TestWithParam<Trans32x32Param> {
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
version_ = GET_PARAM(2); // 0: high precision forward transform
// 1: low precision version for rd loop
version_ = GET_PARAM(2); // 0: high precision forward transform
// 1: low precision version for rd loop
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
}
@@ -147,8 +136,8 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
ASM_REGISTER_STATE_CHECK(inv_txfm_(test_temp_block, dst, 32));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(inv_txfm_(test_temp_block,
CONVERT_TO_BYTEPTR(dst16), 32));
ASM_REGISTER_STATE_CHECK(
inv_txfm_(test_temp_block, CONVERT_TO_BYTEPTR(dst16), 32));
#endif
}
@@ -160,8 +149,7 @@ TEST_P(Trans32x32Test, AccuracyCheck) {
const int32_t diff = dst[j] - src[j];
#endif
const uint32_t error = diff * diff;
if (max_error < error)
max_error = error;
if (max_error < error) max_error = error;
total_error += error;
}
}
@@ -187,8 +175,9 @@ TEST_P(Trans32x32Test, CoeffCheck) {
DECLARE_ALIGNED(16, tran_low_t, output_block[kNumCoeffs]);
for (int i = 0; i < count_test_block; ++i) {
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
}
const int stride = 32;
vpx_fdct32x32_c(input_block, output_ref_block, stride);
@@ -220,11 +209,9 @@ TEST_P(Trans32x32Test, MemCheck) {
input_extreme_block[j] = rnd.Rand8() & 1 ? mask_ : -mask_;
}
if (i == 0) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = mask_;
} else if (i == 1) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = -mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = -mask_;
}
const int stride = 32;
@@ -281,8 +268,9 @@ TEST_P(Trans32x32Test, InverseAccuracy) {
}
reference_32x32_dct_2d(in, out_r);
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
coeff[j] = static_cast<tran_low_t>(round(out_r[j]));
}
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(inv_txfm_(coeff, dst, 32));
#if CONFIG_VP9_HIGHBITDEPTH
@@ -298,9 +286,8 @@ TEST_P(Trans32x32Test, InverseAccuracy) {
const int diff = dst[j] - src[j];
#endif
const int error = diff * diff;
EXPECT_GE(1, error)
<< "Error: 32x32 IDCT has error " << error
<< " at index " << j;
EXPECT_GE(1, error) << "Error: 32x32 IDCT has error " << error
<< " at index " << j;
}
}
}
@@ -372,18 +359,13 @@ using std::tr1::make_tuple;
INSTANTIATE_TEST_CASE_P(
C, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_highbd_fdct32x32_c,
&idct32x32_10, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct32x32_rd_c,
&idct32x32_10, 1, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct32x32_c,
&idct32x32_12, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct32x32_rd_c,
&idct32x32_12, 1, VPX_BITS_12),
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c,
&vpx_idct32x32_1024_add_c, 1, VPX_BITS_8)));
make_tuple(&vpx_highbd_fdct32x32_c, &idct32x32_10, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct32x32_rd_c, &idct32x32_10, 1, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct32x32_c, &idct32x32_12, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct32x32_rd_c, &idct32x32_12, 1, VPX_BITS_12),
make_tuple(&vpx_fdct32x32_c, &vpx_idct32x32_1024_add_c, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c, &vpx_idct32x32_1024_add_c, 1,
VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
C, PartialTrans32x32Test,
::testing::Values(make_tuple(&vpx_highbd_fdct32x32_1_c, VPX_BITS_8),
@@ -392,11 +374,10 @@ INSTANTIATE_TEST_CASE_P(
#else
INSTANTIATE_TEST_CASE_P(
C, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c,
&vpx_idct32x32_1024_add_c, 1, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct32x32_c, &vpx_idct32x32_1024_add_c, 0,
VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c, &vpx_idct32x32_1024_add_c,
1, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(C, PartialTrans32x32Test,
::testing::Values(make_tuple(&vpx_fdct32x32_1_c,
VPX_BITS_8)));
@@ -405,21 +386,19 @@ INSTANTIATE_TEST_CASE_P(C, PartialTrans32x32Test,
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
NEON, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_neon, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c,
&vpx_idct32x32_1024_add_neon, 1, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct32x32_c, &vpx_idct32x32_1024_add_neon,
0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_c,
&vpx_idct32x32_1024_add_neon, 1, VPX_BITS_8)));
#endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_fdct32x32_sse2,
&vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_sse2,
&vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct32x32_sse2,
&vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_sse2,
&vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans32x32Test,
::testing::Values(make_tuple(&vpx_fdct32x32_1_sse2,
VPX_BITS_8)));
@@ -447,21 +426,19 @@ INSTANTIATE_TEST_CASE_P(SSE2, PartialTrans32x32Test,
#if HAVE_AVX2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
AVX2, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_fdct32x32_avx2,
&vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_avx2,
&vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct32x32_avx2,
&vpx_idct32x32_1024_add_sse2, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_avx2,
&vpx_idct32x32_1024_add_sse2, 1, VPX_BITS_8)));
#endif // HAVE_AVX2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
MSA, Trans32x32Test,
::testing::Values(
make_tuple(&vpx_fdct32x32_msa,
&vpx_idct32x32_1024_add_msa, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_msa,
&vpx_idct32x32_1024_add_msa, 1, VPX_BITS_8)));
::testing::Values(make_tuple(&vpx_fdct32x32_msa,
&vpx_idct32x32_1024_add_msa, 0, VPX_BITS_8),
make_tuple(&vpx_fdct32x32_rd_msa,
&vpx_idct32x32_1024_add_msa, 1, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(MSA, PartialTrans32x32Test,
::testing::Values(make_tuple(&vpx_fdct32x32_1_msa,
VPX_BITS_8)));

View File

@@ -28,7 +28,7 @@ TEST(DecodeAPI, InvalidParams) {
&vpx_codec_vp9_dx_algo,
#endif
};
uint8_t buf[1] = {0};
uint8_t buf[1] = { 0 };
vpx_codec_ctx_t dec;
EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_dec_init(NULL, NULL, NULL, 0));
@@ -51,8 +51,7 @@ TEST(DecodeAPI, InvalidParams) {
vpx_codec_decode(&dec, buf, NELEMENTS(buf), NULL, 0));
EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
vpx_codec_decode(&dec, NULL, NELEMENTS(buf), NULL, 0));
EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
vpx_codec_decode(&dec, buf, 0, NULL, 0));
EXPECT_EQ(VPX_CODEC_INVALID_PARAM, vpx_codec_decode(&dec, buf, 0, NULL, 0));
EXPECT_EQ(VPX_CODEC_OK, vpx_codec_destroy(&dec));
}
@@ -77,12 +76,9 @@ TEST(DecodeAPI, OptionalParams) {
// Test VP9 codec controls after a decode error to ensure the code doesn't
// misbehave.
void TestVp9Controls(vpx_codec_ctx_t *dec) {
static const int kControls[] = {
VP8D_GET_LAST_REF_UPDATES,
VP8D_GET_FRAME_CORRUPTED,
VP9D_GET_DISPLAY_SIZE,
VP9D_GET_FRAME_SIZE
};
static const int kControls[] = { VP8D_GET_LAST_REF_UPDATES,
VP8D_GET_FRAME_CORRUPTED,
VP9D_GET_DISPLAY_SIZE, VP9D_GET_FRAME_SIZE };
int val[2];
for (int i = 0; i < NELEMENTS(kControls); ++i) {
@@ -91,9 +87,7 @@ void TestVp9Controls(vpx_codec_ctx_t *dec) {
case VP8D_GET_FRAME_CORRUPTED:
EXPECT_EQ(VPX_CODEC_ERROR, res) << kControls[i];
break;
default:
EXPECT_EQ(VPX_CODEC_OK, res) << kControls[i];
break;
default: EXPECT_EQ(VPX_CODEC_OK, res) << kControls[i]; break;
}
EXPECT_EQ(VPX_CODEC_INVALID_PARAM,
vpx_codec_control_(dec, kControls[i], NULL));
@@ -150,10 +144,9 @@ TEST(DecodeAPI, Vp9PeekSI) {
// size 10, this should return VPX_CODEC_UNSUP_BITSTREAM and after that it
// should return VPX_CODEC_CORRUPT_FRAME.
const uint8_t data[32] = {
0x85, 0xa4, 0xc1, 0xa1, 0x38, 0x81, 0xa3, 0x49,
0x83, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0x85, 0xa4, 0xc1, 0xa1, 0x38, 0x81, 0xa3, 0x49, 0x83, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
};
for (uint32_t data_sz = 1; data_sz <= 32; ++data_sz) {
@@ -162,9 +155,9 @@ TEST(DecodeAPI, Vp9PeekSI) {
if (data_sz >= 8) {
vpx_codec_ctx_t dec;
EXPECT_EQ(VPX_CODEC_OK, vpx_codec_dec_init(&dec, codec, NULL, 0));
EXPECT_EQ((data_sz < 10) ?
VPX_CODEC_UNSUP_BITSTREAM : VPX_CODEC_CORRUPT_FRAME,
vpx_codec_decode(&dec, data, data_sz, NULL, 0));
EXPECT_EQ(
(data_sz < 10) ? VPX_CODEC_UNSUP_BITSTREAM : VPX_CODEC_CORRUPT_FRAME,
vpx_codec_decode(&dec, data, data_sz, NULL, 0));
vpx_codec_iter_t iter = NULL;
EXPECT_EQ(NULL, vpx_codec_get_frame(&dec, &iter));
EXPECT_EQ(VPX_CODEC_OK, vpx_codec_destroy(&dec));

View File

@@ -28,7 +28,6 @@ namespace {
#define VIDEO_NAME 0
#define THREADS 1
const int kMaxPsnr = 100;
const double kUsecsInSec = 1000000.0;
const char kNewEncodeOutputFile[] = "new_encode.ivf";
@@ -70,8 +69,7 @@ const DecodePerfParam kVP9DecodePerfVectors[] = {
power/temp/min max frame decode times/etc
*/
class DecodePerfTest : public ::testing::TestWithParam<DecodePerfParam> {
};
class DecodePerfTest : public ::testing::TestWithParam<DecodePerfParam> {};
TEST_P(DecodePerfTest, PerfTest) {
const char *const video_name = GET_PARAM(VIDEO_NAME);
@@ -92,8 +90,7 @@ TEST_P(DecodePerfTest, PerfTest) {
}
vpx_usec_timer_mark(&t);
const double elapsed_secs = double(vpx_usec_timer_elapsed(&t))
/ kUsecsInSec;
const double elapsed_secs = double(vpx_usec_timer_elapsed(&t)) / kUsecsInSec;
const unsigned frames = video.frame_number();
const double fps = double(frames) / elapsed_secs;
@@ -111,17 +108,13 @@ TEST_P(DecodePerfTest, PerfTest) {
INSTANTIATE_TEST_CASE_P(VP9, DecodePerfTest,
::testing::ValuesIn(kVP9DecodePerfVectors));
class VP9NewEncodeDecodePerfTest :
public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class VP9NewEncodeDecodePerfTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
VP9NewEncodeDecodePerfTest()
: EncoderTest(GET_PARAM(0)),
encoding_mode_(GET_PARAM(1)),
speed_(0),
outfile_(0),
out_frames_(0) {
}
: EncoderTest(GET_PARAM(0)), encoding_mode_(GET_PARAM(1)), speed_(0),
outfile_(0), out_frames_(0) {}
virtual ~VP9NewEncodeDecodePerfTest() {}
@@ -160,8 +153,9 @@ class VP9NewEncodeDecodePerfTest :
virtual void EndPassHook() {
if (outfile_ != NULL) {
if (!fseek(outfile_, 0, SEEK_SET))
if (!fseek(outfile_, 0, SEEK_SET)) {
ivf_write_file_header(outfile_, &cfg_, VP9_FOURCC, out_frames_);
}
fclose(outfile_);
outfile_ = NULL;
}
@@ -171,8 +165,9 @@ class VP9NewEncodeDecodePerfTest :
++out_frames_;
// Write initial file header if first frame.
if (pkt->data.frame.pts == 0)
if (pkt->data.frame.pts == 0) {
ivf_write_file_header(outfile_, &cfg_, VP9_FOURCC, out_frames_);
}
// Write frame header and data.
ivf_write_frame_header(outfile_, out_frames_, pkt->data.frame.sz);
@@ -180,11 +175,9 @@ class VP9NewEncodeDecodePerfTest :
pkt->data.frame.sz);
}
virtual bool DoDecode() { return false; }
virtual bool DoDecode() const { return false; }
void set_speed(unsigned int speed) {
speed_ = speed;
}
void set_speed(unsigned int speed) { speed_ = speed; }
private:
libvpx_test::TestMode encoding_mode_;
@@ -196,10 +189,7 @@ class VP9NewEncodeDecodePerfTest :
struct EncodePerfTestVideo {
EncodePerfTestVideo(const char *name_, uint32_t width_, uint32_t height_,
uint32_t bitrate_, int frames_)
: name(name_),
width(width_),
height(height_),
bitrate(bitrate_),
: name(name_), width(width_), height(height_), bitrate(bitrate_),
frames(frames_) {}
const char *name;
uint32_t width;
@@ -225,10 +215,8 @@ TEST_P(VP9NewEncodeDecodePerfTest, PerfTest) {
const char *video_name = kVP9EncodePerfTestVectors[i].name;
libvpx_test::I420VideoSource video(
video_name,
kVP9EncodePerfTestVectors[i].width,
kVP9EncodePerfTestVectors[i].height,
timebase.den, timebase.num, 0,
video_name, kVP9EncodePerfTestVectors[i].width,
kVP9EncodePerfTestVectors[i].height, timebase.den, timebase.num, 0,
kVP9EncodePerfTestVectors[i].frames);
set_speed(2);
@@ -268,6 +256,6 @@ TEST_P(VP9NewEncodeDecodePerfTest, PerfTest) {
printf("}\n");
}
VP9_INSTANTIATE_TEST_CASE(
VP9NewEncodeDecodePerfTest, ::testing::Values(::libvpx_test::kTwoPassGood));
VP9_INSTANTIATE_TEST_CASE(VP9NewEncodeDecodePerfTest,
::testing::Values(::libvpx_test::kTwoPassGood));
} // namespace

124
test/decode_svc_test.cc Normal file
View File

@@ -0,0 +1,124 @@
/*
* Copyright (c) 2016 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <string>
#include "test/codec_factory.h"
#include "test/decode_test_driver.h"
#include "test/ivf_video_source.h"
#include "test/test_vectors.h"
#include "test/util.h"
namespace {
const unsigned int kNumFrames = 19;
class DecodeSvcTest : public ::libvpx_test::DecoderTest,
public ::libvpx_test::CodecTestWithParam<const char *> {
protected:
DecodeSvcTest() : DecoderTest(GET_PARAM(::libvpx_test::kCodecFactoryParam)) {}
virtual ~DecodeSvcTest() {}
virtual void PreDecodeFrameHook(
const libvpx_test::CompressedVideoSource &video,
libvpx_test::Decoder *decoder) {
if (video.frame_number() == 0)
decoder->Control(VP9_DECODE_SVC_SPATIAL_LAYER, spatial_layer_);
}
virtual void DecompressedFrameHook(const vpx_image_t &img,
const unsigned int frame_number) {
ASSERT_EQ(img.d_w, width_);
ASSERT_EQ(img.d_h, height_);
total_frames_ = frame_number;
}
int spatial_layer_;
unsigned int width_;
unsigned int height_;
unsigned int total_frames_;
};
// SVC test vector is 1280x720, with 3 spatial layers, and 20 frames.
// Decode the SVC test vector, which has 3 spatial layers, and decode up to
// spatial layer 0. Verify the resolution of each decoded frame and the total
// number of frames decoded. This results in 1/4x1/4 resolution (320x180).
TEST_P(DecodeSvcTest, DecodeSvcTestUpToSpatialLayer0) {
const std::string filename = GET_PARAM(1);
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
video.reset(new libvpx_test::IVFVideoSource(filename));
ASSERT_TRUE(video.get() != NULL);
video->Init();
total_frames_ = 0;
spatial_layer_ = 0;
width_ = 320;
height_ = 180;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
ASSERT_EQ(total_frames_, kNumFrames);
}
// Decode the SVC test vector, which has 3 spatial layers, and decode up to
// spatial layer 1. Verify the resolution of each decoded frame and the total
// number of frames decoded. This results in 1/2x1/2 resolution (640x360).
TEST_P(DecodeSvcTest, DecodeSvcTestUpToSpatialLayer1) {
const std::string filename = GET_PARAM(1);
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
video.reset(new libvpx_test::IVFVideoSource(filename));
ASSERT_TRUE(video.get() != NULL);
video->Init();
total_frames_ = 0;
spatial_layer_ = 1;
width_ = 640;
height_ = 360;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
ASSERT_EQ(total_frames_, kNumFrames);
}
// Decode the SVC test vector, which has 3 spatial layers, and decode up to
// spatial layer 2. Verify the resolution of each decoded frame and the total
// number of frames decoded. This results in the full resolution (1280x720).
TEST_P(DecodeSvcTest, DecodeSvcTestUpToSpatialLayer2) {
const std::string filename = GET_PARAM(1);
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
video.reset(new libvpx_test::IVFVideoSource(filename));
ASSERT_TRUE(video.get() != NULL);
video->Init();
total_frames_ = 0;
spatial_layer_ = 2;
width_ = 1280;
height_ = 720;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
ASSERT_EQ(total_frames_, kNumFrames);
}
// Decode the SVC test vector, which has 3 spatial layers, and decode up to
// spatial layer 10. Verify the resolution of each decoded frame and the total
// number of frames decoded. This is beyond the number of spatial layers, so
// the decoding should result in the full resolution (1280x720).
TEST_P(DecodeSvcTest, DecodeSvcTestUpToSpatialLayer10) {
const std::string filename = GET_PARAM(1);
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
video.reset(new libvpx_test::IVFVideoSource(filename));
ASSERT_TRUE(video.get() != NULL);
video->Init();
total_frames_ = 0;
spatial_layer_ = 10;
width_ = 1280;
height_ = 720;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
ASSERT_EQ(total_frames_, kNumFrames);
}
VP9_INSTANTIATE_TEST_CASE(
DecodeSvcTest, ::testing::ValuesIn(libvpx_test::kVP9TestVectorsSvc,
libvpx_test::kVP9TestVectorsSvc +
libvpx_test::kNumVP9TestVectorsSvc));
} // namespace

View File

@@ -21,9 +21,8 @@ const char kVP8Name[] = "WebM Project VP8";
vpx_codec_err_t Decoder::PeekStream(const uint8_t *cxdata, size_t size,
vpx_codec_stream_info_t *stream_info) {
return vpx_codec_peek_stream_info(CodecInterface(),
cxdata, static_cast<unsigned int>(size),
stream_info);
return vpx_codec_peek_stream_info(
CodecInterface(), cxdata, static_cast<unsigned int>(size), stream_info);
}
vpx_codec_err_t Decoder::DecodeFrame(const uint8_t *cxdata, size_t size) {
@@ -35,9 +34,8 @@ vpx_codec_err_t Decoder::DecodeFrame(const uint8_t *cxdata, size_t size,
vpx_codec_err_t res_dec;
InitOnce();
API_REGISTER_STATE_CHECK(
res_dec = vpx_codec_decode(&decoder_,
cxdata, static_cast<unsigned int>(size),
user_priv, 0));
res_dec = vpx_codec_decode(
&decoder_, cxdata, static_cast<unsigned int>(size), user_priv, 0));
return res_dec;
}
@@ -67,7 +65,7 @@ void DecoderTest::HandlePeekResult(Decoder *const decoder,
void DecoderTest::RunLoop(CompressedVideoSource *video,
const vpx_codec_dec_cfg_t &dec_cfg) {
Decoder* const decoder = codec_->CreateDecoder(dec_cfg, flags_, 0);
Decoder *const decoder = codec_->CreateDecoder(dec_cfg, flags_);
ASSERT_TRUE(decoder != NULL);
bool end_of_file = false;
@@ -80,16 +78,14 @@ void DecoderTest::RunLoop(CompressedVideoSource *video,
stream_info.sz = sizeof(stream_info);
if (video->cxdata() != NULL) {
const vpx_codec_err_t res_peek = decoder->PeekStream(video->cxdata(),
video->frame_size(),
&stream_info);
const vpx_codec_err_t res_peek = decoder->PeekStream(
video->cxdata(), video->frame_size(), &stream_info);
HandlePeekResult(decoder, video, res_peek);
ASSERT_FALSE(::testing::Test::HasFailure());
vpx_codec_err_t res_dec = decoder->DecodeFrame(video->cxdata(),
video->frame_size());
if (!HandleDecodeResult(res_dec, *video, decoder))
break;
vpx_codec_err_t res_dec =
decoder->DecodeFrame(video->cxdata(), video->frame_size());
if (!HandleDecodeResult(res_dec, *video, decoder)) break;
} else {
// Signal end of the file to the decoder.
const vpx_codec_err_t res_dec = decoder->DecodeFrame(NULL, 0);
@@ -101,8 +97,9 @@ void DecoderTest::RunLoop(CompressedVideoSource *video,
const vpx_image_t *img = NULL;
// Get decompressed data
while ((img = dec_iter.Next()))
while ((img = dec_iter.Next())) {
DecompressedFrameHook(*img, video->frame_number());
}
}
delete decoder;
}
@@ -116,8 +113,6 @@ void DecoderTest::set_cfg(const vpx_codec_dec_cfg_t &dec_cfg) {
memcpy(&cfg_, &dec_cfg, sizeof(cfg_));
}
void DecoderTest::set_flags(const vpx_codec_flags_t flags) {
flags_ = flags;
}
void DecoderTest::set_flags(const vpx_codec_flags_t flags) { flags_ = flags; }
} // namespace libvpx_test

View File

@@ -26,13 +26,11 @@ class DxDataIterator {
explicit DxDataIterator(vpx_codec_ctx_t *decoder)
: decoder_(decoder), iter_(NULL) {}
const vpx_image_t *Next() {
return vpx_codec_get_frame(decoder_, &iter_);
}
const vpx_image_t *Next() { return vpx_codec_get_frame(decoder_, &iter_); }
private:
vpx_codec_ctx_t *decoder_;
vpx_codec_iter_t iter_;
vpx_codec_ctx_t *decoder_;
vpx_codec_iter_t iter_;
};
// Provides a simplified interface to manage one video decoding.
@@ -40,20 +38,17 @@ class DxDataIterator {
// as more tests are added.
class Decoder {
public:
Decoder(vpx_codec_dec_cfg_t cfg, unsigned long deadline)
: cfg_(cfg), flags_(0), deadline_(deadline), init_done_(false) {
explicit Decoder(vpx_codec_dec_cfg_t cfg)
: cfg_(cfg), flags_(0), init_done_(false) {
memset(&decoder_, 0, sizeof(decoder_));
}
Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag,
unsigned long deadline) // NOLINT
: cfg_(cfg), flags_(flag), deadline_(deadline), init_done_(false) {
Decoder(vpx_codec_dec_cfg_t cfg, const vpx_codec_flags_t flag)
: cfg_(cfg), flags_(flag), init_done_(false) {
memset(&decoder_, 0, sizeof(decoder_));
}
virtual ~Decoder() {
vpx_codec_destroy(&decoder_);
}
virtual ~Decoder() { vpx_codec_destroy(&decoder_); }
vpx_codec_err_t PeekStream(const uint8_t *cxdata, size_t size,
vpx_codec_stream_info_t *stream_info);
@@ -63,17 +58,9 @@ class Decoder {
vpx_codec_err_t DecodeFrame(const uint8_t *cxdata, size_t size,
void *user_priv);
DxDataIterator GetDxData() {
return DxDataIterator(&decoder_);
}
DxDataIterator GetDxData() { return DxDataIterator(&decoder_); }
void set_deadline(unsigned long deadline) {
deadline_ = deadline;
}
void Control(int ctrl_id, int arg) {
Control(ctrl_id, arg, VPX_CODEC_OK);
}
void Control(int ctrl_id, int arg) { Control(ctrl_id, arg, VPX_CODEC_OK); }
void Control(int ctrl_id, const void *arg) {
InitOnce();
@@ -87,7 +74,7 @@ class Decoder {
ASSERT_EQ(expected_value, res) << DecodeError();
}
const char* DecodeError() {
const char *DecodeError() {
const char *detail = vpx_codec_error_detail(&decoder_);
return detail ? detail : vpx_codec_error(&decoder_);
}
@@ -97,38 +84,34 @@ class Decoder {
vpx_get_frame_buffer_cb_fn_t cb_get,
vpx_release_frame_buffer_cb_fn_t cb_release, void *user_priv) {
InitOnce();
return vpx_codec_set_frame_buffer_functions(
&decoder_, cb_get, cb_release, user_priv);
return vpx_codec_set_frame_buffer_functions(&decoder_, cb_get, cb_release,
user_priv);
}
const char* GetDecoderName() const {
const char *GetDecoderName() const {
return vpx_codec_iface_name(CodecInterface());
}
bool IsVP8() const;
vpx_codec_ctx_t * GetDecoder() {
return &decoder_;
}
vpx_codec_ctx_t *GetDecoder() { return &decoder_; }
protected:
virtual vpx_codec_iface_t* CodecInterface() const = 0;
virtual vpx_codec_iface_t *CodecInterface() const = 0;
void InitOnce() {
if (!init_done_) {
const vpx_codec_err_t res = vpx_codec_dec_init(&decoder_,
CodecInterface(),
&cfg_, flags_);
const vpx_codec_err_t res =
vpx_codec_dec_init(&decoder_, CodecInterface(), &cfg_, flags_);
ASSERT_EQ(VPX_CODEC_OK, res) << DecodeError();
init_done_ = true;
}
}
vpx_codec_ctx_t decoder_;
vpx_codec_ctx_t decoder_;
vpx_codec_dec_cfg_t cfg_;
vpx_codec_flags_t flags_;
unsigned int deadline_;
bool init_done_;
vpx_codec_flags_t flags_;
bool init_done_;
};
// Common test functionality for all Decoder tests.
@@ -143,37 +126,35 @@ class DecoderTest {
virtual void set_flags(const vpx_codec_flags_t flags);
// Hook to be called before decompressing every frame.
virtual void PreDecodeFrameHook(const CompressedVideoSource& /*video*/,
Decoder* /*decoder*/) {}
virtual void PreDecodeFrameHook(const CompressedVideoSource & /*video*/,
Decoder * /*decoder*/) {}
// Hook to be called to handle decode result. Return true to continue.
virtual bool HandleDecodeResult(const vpx_codec_err_t res_dec,
const CompressedVideoSource& /*video*/,
const CompressedVideoSource & /*video*/,
Decoder *decoder) {
EXPECT_EQ(VPX_CODEC_OK, res_dec) << decoder->DecodeError();
return VPX_CODEC_OK == res_dec;
}
// Hook to be called on every decompressed frame.
virtual void DecompressedFrameHook(const vpx_image_t& /*img*/,
virtual void DecompressedFrameHook(const vpx_image_t & /*img*/,
const unsigned int /*frame_number*/) {}
// Hook to be called on peek result
virtual void HandlePeekResult(Decoder* const decoder,
virtual void HandlePeekResult(Decoder *const decoder,
CompressedVideoSource *video,
const vpx_codec_err_t res_peek);
protected:
explicit DecoderTest(const CodecFactory *codec)
: codec_(codec),
cfg_(),
flags_(0) {}
: codec_(codec), cfg_(), flags_(0) {}
virtual ~DecoderTest() {}
const CodecFactory *codec_;
vpx_codec_dec_cfg_t cfg_;
vpx_codec_flags_t flags_;
vpx_codec_flags_t flags_;
};
} // namespace libvpx_test

View File

@@ -27,7 +27,7 @@ TEST(EncodeAPI, InvalidParams) {
&vpx_codec_vp9_cx_algo,
#endif
};
uint8_t buf[1] = {0};
uint8_t buf[1] = { 0 };
vpx_image_t img;
vpx_codec_ctx_t enc;
vpx_codec_enc_cfg_t cfg;

View File

@@ -26,10 +26,7 @@ const double kUsecsInSec = 1000000.0;
struct EncodePerfTestVideo {
EncodePerfTestVideo(const char *name_, uint32_t width_, uint32_t height_,
uint32_t bitrate_, int frames_)
: name(name_),
width(width_),
height(height_),
bitrate(bitrate_),
: name(name_), width(width_), height(height_), bitrate(bitrate_),
frames(frames_) {}
const char *name;
uint32_t width;
@@ -45,8 +42,8 @@ const EncodePerfTestVideo kVP9EncodePerfTestVectors[] = {
EncodePerfTestVideo("macmarcostationary_640_480_30.yuv", 640, 480, 200, 718),
EncodePerfTestVideo("niklas_640_480_30.yuv", 640, 480, 200, 471),
EncodePerfTestVideo("tacomanarrows_640_480_30.yuv", 640, 480, 200, 300),
EncodePerfTestVideo("tacomasmallcameramovement_640_480_30.yuv",
640, 480, 200, 300),
EncodePerfTestVideo("tacomasmallcameramovement_640_480_30.yuv", 640, 480, 200,
300),
EncodePerfTestVideo("thaloundeskmtg_640_480_30.yuv", 640, 480, 200, 300),
EncodePerfTestVideo("niklas_1280_720_30.yuv", 1280, 720, 600, 470),
};
@@ -61,12 +58,8 @@ class VP9EncodePerfTest
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
VP9EncodePerfTest()
: EncoderTest(GET_PARAM(0)),
min_psnr_(kMaxPsnr),
nframes_(0),
encoding_mode_(GET_PARAM(1)),
speed_(0),
threads_(1) {}
: EncoderTest(GET_PARAM(0)), min_psnr_(kMaxPsnr), nframes_(0),
encoding_mode_(GET_PARAM(1)), speed_(0), threads_(1) {}
virtual ~VP9EncodePerfTest() {}
@@ -107,24 +100,18 @@ class VP9EncodePerfTest
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
if (pkt->data.psnr.psnr[0] < min_psnr_) {
min_psnr_= pkt->data.psnr.psnr[0];
min_psnr_ = pkt->data.psnr.psnr[0];
}
}
// for performance reasons don't decode
virtual bool DoDecode() { return 0; }
virtual bool DoDecode() const { return false; }
double min_psnr() const {
return min_psnr_;
}
double min_psnr() const { return min_psnr_; }
void set_speed(unsigned int speed) {
speed_ = speed;
}
void set_speed(unsigned int speed) { speed_ = speed; }
void set_threads(unsigned int threads) {
threads_ = threads;
}
void set_threads(unsigned int threads) { threads_ = threads; }
private:
double min_psnr_;
@@ -139,11 +126,12 @@ TEST_P(VP9EncodePerfTest, PerfTest) {
for (size_t j = 0; j < NELEMENTS(kEncodePerfTestSpeeds); ++j) {
for (size_t k = 0; k < NELEMENTS(kEncodePerfTestThreads); ++k) {
if (kVP9EncodePerfTestVectors[i].width < 512 &&
kEncodePerfTestThreads[k] > 1)
kEncodePerfTestThreads[k] > 1) {
continue;
else if (kVP9EncodePerfTestVectors[i].width < 1024 &&
kEncodePerfTestThreads[k] > 2)
} else if (kVP9EncodePerfTestVectors[i].width < 1024 &&
kEncodePerfTestThreads[k] > 2) {
continue;
}
set_threads(kEncodePerfTestThreads[k]);
SetUp();
@@ -157,10 +145,8 @@ TEST_P(VP9EncodePerfTest, PerfTest) {
const unsigned frames = kVP9EncodePerfTestVectors[i].frames;
const char *video_name = kVP9EncodePerfTestVectors[i].name;
libvpx_test::I420VideoSource video(
video_name,
kVP9EncodePerfTestVectors[i].width,
kVP9EncodePerfTestVectors[i].height,
timebase.den, timebase.num, 0,
video_name, kVP9EncodePerfTestVectors[i].width,
kVP9EncodePerfTestVectors[i].height, timebase.den, timebase.num, 0,
kVP9EncodePerfTestVectors[i].frames);
set_speed(kEncodePerfTestSpeeds[j]);
@@ -197,6 +183,6 @@ TEST_P(VP9EncodePerfTest, PerfTest) {
}
}
VP9_INSTANTIATE_TEST_CASE(
VP9EncodePerfTest, ::testing::Values(::libvpx_test::kRealTime));
VP9_INSTANTIATE_TEST_CASE(VP9EncodePerfTest,
::testing::Values(::libvpx_test::kRealTime));
} // namespace

View File

@@ -30,8 +30,7 @@ void Encoder::InitEncoder(VideoSource *video) {
cfg_.g_timebase = video->timebase();
cfg_.rc_twopass_stats_in = stats_->buf();
res = vpx_codec_enc_init(&encoder_, CodecInterface(), &cfg_,
init_flags_);
res = vpx_codec_enc_init(&encoder_, CodecInterface(), &cfg_, init_flags_);
ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
#if CONFIG_VP9_ENCODER
@@ -53,17 +52,17 @@ void Encoder::InitEncoder(VideoSource *video) {
}
void Encoder::EncodeFrame(VideoSource *video, const unsigned long frame_flags) {
if (video->img())
if (video->img()) {
EncodeFrameInternal(*video, frame_flags);
else
} else {
Flush();
}
// Handle twopass stats
CxDataIterator iter = GetCxData();
while (const vpx_codec_cx_pkt_t *pkt = iter.Next()) {
if (pkt->kind != VPX_CODEC_STATS_PKT)
continue;
if (pkt->kind != VPX_CODEC_STATS_PKT) continue;
stats_->Append(*pkt);
}
@@ -83,15 +82,15 @@ void Encoder::EncodeFrameInternal(const VideoSource &video,
}
// Encode the frame
API_REGISTER_STATE_CHECK(
res = vpx_codec_encode(&encoder_, img, video.pts(), video.duration(),
frame_flags, deadline_));
API_REGISTER_STATE_CHECK(res = vpx_codec_encode(&encoder_, img, video.pts(),
video.duration(), frame_flags,
deadline_));
ASSERT_EQ(VPX_CODEC_OK, res) << EncoderError();
}
void Encoder::Flush() {
const vpx_codec_err_t res = vpx_codec_encode(&encoder_, NULL, 0, 0, 0,
deadline_);
const vpx_codec_err_t res =
vpx_codec_encode(&encoder_, NULL, 0, 0, 0, deadline_);
if (!encoder_.priv)
ASSERT_EQ(VPX_CODEC_ERROR, res) << EncoderError();
else
@@ -106,60 +105,57 @@ void EncoderTest::InitializeConfig() {
void EncoderTest::SetMode(TestMode mode) {
switch (mode) {
case kRealTime:
deadline_ = VPX_DL_REALTIME;
break;
case kRealTime: deadline_ = VPX_DL_REALTIME; break;
case kOnePassGood:
case kTwoPassGood:
deadline_ = VPX_DL_GOOD_QUALITY;
break;
case kTwoPassGood: deadline_ = VPX_DL_GOOD_QUALITY; break;
case kOnePassBest:
case kTwoPassBest:
deadline_ = VPX_DL_BEST_QUALITY;
break;
case kTwoPassBest: deadline_ = VPX_DL_BEST_QUALITY; break;
default:
ASSERT_TRUE(false) << "Unexpected mode " << mode;
default: ASSERT_TRUE(false) << "Unexpected mode " << mode;
}
if (mode == kTwoPassGood || mode == kTwoPassBest)
if (mode == kTwoPassGood || mode == kTwoPassBest) {
passes_ = 2;
else
} else {
passes_ = 1;
}
}
// The function should return "true" most of the time, therefore no early
// break-out is implemented within the match checking process.
static bool compare_img(const vpx_image_t *img1,
const vpx_image_t *img2) {
bool match = (img1->fmt == img2->fmt) &&
(img1->cs == img2->cs) &&
(img1->d_w == img2->d_w) &&
(img1->d_h == img2->d_h);
static bool compare_img(const vpx_image_t *img1, const vpx_image_t *img2) {
bool match = (img1->fmt == img2->fmt) && (img1->cs == img2->cs) &&
(img1->d_w == img2->d_w) && (img1->d_h == img2->d_h);
const unsigned int width_y = img1->d_w;
const unsigned int width_y = img1->d_w;
const unsigned int height_y = img1->d_h;
unsigned int i;
for (i = 0; i < height_y; ++i)
for (i = 0; i < height_y; ++i) {
match = (memcmp(img1->planes[VPX_PLANE_Y] + i * img1->stride[VPX_PLANE_Y],
img2->planes[VPX_PLANE_Y] + i * img2->stride[VPX_PLANE_Y],
width_y) == 0) && match;
const unsigned int width_uv = (img1->d_w + 1) >> 1;
width_y) == 0) &&
match;
}
const unsigned int width_uv = (img1->d_w + 1) >> 1;
const unsigned int height_uv = (img1->d_h + 1) >> 1;
for (i = 0; i < height_uv; ++i)
for (i = 0; i < height_uv; ++i) {
match = (memcmp(img1->planes[VPX_PLANE_U] + i * img1->stride[VPX_PLANE_U],
img2->planes[VPX_PLANE_U] + i * img2->stride[VPX_PLANE_U],
width_uv) == 0) && match;
for (i = 0; i < height_uv; ++i)
width_uv) == 0) &&
match;
}
for (i = 0; i < height_uv; ++i) {
match = (memcmp(img1->planes[VPX_PLANE_V] + i * img1->stride[VPX_PLANE_V],
img2->planes[VPX_PLANE_V] + i * img2->stride[VPX_PLANE_V],
width_uv) == 0) && match;
width_uv) == 0) &&
match;
}
return match;
}
void EncoderTest::MismatchHook(const vpx_image_t* /*img1*/,
const vpx_image_t* /*img2*/) {
void EncoderTest::MismatchHook(const vpx_image_t * /*img1*/,
const vpx_image_t * /*img2*/) {
ASSERT_TRUE(0) << "Encode/Decode mismatch found";
}
@@ -172,34 +168,37 @@ void EncoderTest::RunLoop(VideoSource *video) {
for (unsigned int pass = 0; pass < passes_; pass++) {
last_pts_ = 0;
if (passes_ == 1)
if (passes_ == 1) {
cfg_.g_pass = VPX_RC_ONE_PASS;
else if (pass == 0)
} else if (pass == 0) {
cfg_.g_pass = VPX_RC_FIRST_PASS;
else
} else {
cfg_.g_pass = VPX_RC_LAST_PASS;
}
BeginPassHook(pass);
Encoder* const encoder = codec_->CreateEncoder(cfg_, deadline_, init_flags_,
&stats_);
ASSERT_TRUE(encoder != NULL);
testing::internal::scoped_ptr<Encoder> encoder(
codec_->CreateEncoder(cfg_, deadline_, init_flags_, &stats_));
ASSERT_TRUE(encoder.get() != NULL);
video->Begin();
ASSERT_NO_FATAL_FAILURE(video->Begin());
encoder->InitEncoder(video);
ASSERT_FALSE(::testing::Test::HasFatalFailure());
unsigned long dec_init_flags = 0; // NOLINT
// Use fragment decoder if encoder outputs partitions.
// NOTE: fragment decoder and partition encoder are only supported by VP8.
if (init_flags_ & VPX_CODEC_USE_OUTPUT_PARTITION)
if (init_flags_ & VPX_CODEC_USE_OUTPUT_PARTITION) {
dec_init_flags |= VPX_CODEC_USE_INPUT_FRAGMENTS;
Decoder* const decoder = codec_->CreateDecoder(dec_cfg, dec_init_flags, 0);
}
testing::internal::scoped_ptr<Decoder> decoder(
codec_->CreateDecoder(dec_cfg, dec_init_flags));
bool again;
for (again = true; again; video->Next()) {
again = (video->img() != NULL);
PreEncodeFrameHook(video);
PreEncodeFrameHook(video, encoder);
PreEncodeFrameHook(video, encoder.get());
encoder->EncodeFrame(video, frame_flags_);
CxDataIterator iter = encoder->GetCxData();
@@ -212,12 +211,11 @@ void EncoderTest::RunLoop(VideoSource *video) {
switch (pkt->kind) {
case VPX_CODEC_CX_FRAME_PKT:
has_cxdata = true;
if (decoder && DoDecode()) {
if (decoder.get() != NULL && DoDecode()) {
vpx_codec_err_t res_dec = decoder->DecodeFrame(
(const uint8_t*)pkt->data.frame.buf, pkt->data.frame.sz);
(const uint8_t *)pkt->data.frame.buf, pkt->data.frame.sz);
if (!HandleDecodeResult(res_dec, *video, decoder))
break;
if (!HandleDecodeResult(res_dec, *video, decoder.get())) break;
has_dxdata = true;
}
@@ -226,20 +224,16 @@ void EncoderTest::RunLoop(VideoSource *video) {
FramePktHook(pkt);
break;
case VPX_CODEC_PSNR_PKT:
PSNRPktHook(pkt);
break;
case VPX_CODEC_PSNR_PKT: PSNRPktHook(pkt); break;
default:
break;
default: break;
}
}
// Flush the decoder when there are no more fragments.
if ((init_flags_ & VPX_CODEC_USE_OUTPUT_PARTITION) && has_dxdata) {
const vpx_codec_err_t res_dec = decoder->DecodeFrame(NULL, 0);
if (!HandleDecodeResult(res_dec, *video, decoder))
break;
if (!HandleDecodeResult(res_dec, *video, decoder.get())) break;
}
if (has_dxdata && has_cxdata) {
@@ -252,21 +246,14 @@ void EncoderTest::RunLoop(VideoSource *video) {
MismatchHook(img_enc, img_dec);
}
}
if (img_dec)
DecompressedFrameHook(*img_dec, video->pts());
if (img_dec) DecompressedFrameHook(*img_dec, video->pts());
}
if (!Continue())
break;
if (!Continue()) break;
}
EndPassHook();
if (decoder)
delete decoder;
delete encoder;
if (!Continue())
break;
if (!Continue()) break;
}
}

View File

@@ -33,19 +33,17 @@ enum TestMode {
kTwoPassGood,
kTwoPassBest
};
#define ALL_TEST_MODES ::testing::Values(::libvpx_test::kRealTime, \
::libvpx_test::kOnePassGood, \
::libvpx_test::kOnePassBest, \
::libvpx_test::kTwoPassGood, \
::libvpx_test::kTwoPassBest)
#define ALL_TEST_MODES \
::testing::Values(::libvpx_test::kRealTime, ::libvpx_test::kOnePassGood, \
::libvpx_test::kOnePassBest, ::libvpx_test::kTwoPassGood, \
::libvpx_test::kTwoPassBest)
#define ONE_PASS_TEST_MODES ::testing::Values(::libvpx_test::kRealTime, \
::libvpx_test::kOnePassGood, \
::libvpx_test::kOnePassBest)
#define TWO_PASS_TEST_MODES ::testing::Values(::libvpx_test::kTwoPassGood, \
::libvpx_test::kTwoPassBest)
#define ONE_PASS_TEST_MODES \
::testing::Values(::libvpx_test::kRealTime, ::libvpx_test::kOnePassGood, \
::libvpx_test::kOnePassBest)
#define TWO_PASS_TEST_MODES \
::testing::Values(::libvpx_test::kTwoPassGood, ::libvpx_test::kTwoPassBest)
// Provides an object to handle the libvpx get_cx_data() iteration pattern
class CxDataIterator {
@@ -58,8 +56,8 @@ class CxDataIterator {
}
private:
vpx_codec_ctx_t *encoder_;
vpx_codec_iter_t iter_;
vpx_codec_ctx_t *encoder_;
vpx_codec_iter_t iter_;
};
// Implements an in-memory store for libvpx twopass statistics
@@ -75,15 +73,12 @@ class TwopassStatsStore {
return buf;
}
void Reset() {
buffer_.clear();
}
void Reset() { buffer_.clear(); }
protected:
std::string buffer_;
std::string buffer_;
};
// Provides a simplified interface to manage one video encoding pass, given
// a configuration and video source.
//
@@ -97,13 +92,9 @@ class Encoder {
memset(&encoder_, 0, sizeof(encoder_));
}
virtual ~Encoder() {
vpx_codec_destroy(&encoder_);
}
virtual ~Encoder() { vpx_codec_destroy(&encoder_); }
CxDataIterator GetCxData() {
return CxDataIterator(&encoder_);
}
CxDataIterator GetCxData() { return CxDataIterator(&encoder_); }
void InitEncoder(VideoSource *video);
@@ -115,9 +106,7 @@ class Encoder {
void EncodeFrame(VideoSource *video, const unsigned long frame_flags);
// Convenience wrapper for EncodeFrame()
void EncodeFrame(VideoSource *video) {
EncodeFrame(video, 0);
}
void EncodeFrame(VideoSource *video) { EncodeFrame(video, 0); }
void Control(int ctrl_id, int arg) {
const vpx_codec_err_t res = vpx_codec_control_(&encoder_, ctrl_id, arg);
@@ -156,12 +145,10 @@ class Encoder {
cfg_ = *cfg;
}
void set_deadline(unsigned long deadline) {
deadline_ = deadline;
}
void set_deadline(unsigned long deadline) { deadline_ = deadline; }
protected:
virtual vpx_codec_iface_t* CodecInterface() const = 0;
virtual vpx_codec_iface_t *CodecInterface() const = 0;
const char *EncoderError() {
const char *detail = vpx_codec_error_detail(&encoder_);
@@ -175,11 +162,11 @@ class Encoder {
// Flush the encoder on EOS
void Flush();
vpx_codec_ctx_t encoder_;
vpx_codec_enc_cfg_t cfg_;
unsigned long deadline_;
unsigned long init_flags_;
TwopassStatsStore *stats_;
vpx_codec_ctx_t encoder_;
vpx_codec_enc_cfg_t cfg_;
unsigned long deadline_;
unsigned long init_flags_;
TwopassStatsStore *stats_;
};
// Common test functionality for all Encoder tests.
@@ -221,36 +208,35 @@ class EncoderTest {
virtual void EndPassHook() {}
// Hook to be called before encoding a frame.
virtual void PreEncodeFrameHook(VideoSource* /*video*/) {}
virtual void PreEncodeFrameHook(VideoSource* /*video*/,
Encoder* /*encoder*/) {}
virtual void PreEncodeFrameHook(VideoSource * /*video*/) {}
virtual void PreEncodeFrameHook(VideoSource * /*video*/,
Encoder * /*encoder*/) {}
// Hook to be called on every compressed data packet.
virtual void FramePktHook(const vpx_codec_cx_pkt_t* /*pkt*/) {}
virtual void FramePktHook(const vpx_codec_cx_pkt_t * /*pkt*/) {}
// Hook to be called on every PSNR packet.
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t* /*pkt*/) {}
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t * /*pkt*/) {}
// Hook to determine whether the encode loop should continue.
virtual bool Continue() const {
return !(::testing::Test::HasFatalFailure() || abort_);
}
const CodecFactory *codec_;
const CodecFactory *codec_;
// Hook to determine whether to decode frame after encoding
virtual bool DoDecode() const { return 1; }
// Hook to handle encode/decode mismatch
virtual void MismatchHook(const vpx_image_t *img1,
const vpx_image_t *img2);
virtual void MismatchHook(const vpx_image_t *img1, const vpx_image_t *img2);
// Hook to be called on every decompressed frame.
virtual void DecompressedFrameHook(const vpx_image_t& /*img*/,
virtual void DecompressedFrameHook(const vpx_image_t & /*img*/,
vpx_codec_pts_t /*pts*/) {}
// Hook to be called to handle decode result. Return true to continue.
virtual bool HandleDecodeResult(const vpx_codec_err_t res_dec,
const VideoSource& /*video*/,
const VideoSource & /*video*/,
Decoder *decoder) {
EXPECT_EQ(VPX_CODEC_OK, res_dec) << decoder->DecodeError();
return VPX_CODEC_OK == res_dec;
@@ -262,15 +248,15 @@ class EncoderTest {
return pkt;
}
bool abort_;
vpx_codec_enc_cfg_t cfg_;
vpx_codec_dec_cfg_t dec_cfg_;
unsigned int passes_;
unsigned long deadline_;
TwopassStatsStore stats_;
unsigned long init_flags_;
unsigned long frame_flags_;
vpx_codec_pts_t last_pts_;
bool abort_;
vpx_codec_enc_cfg_t cfg_;
vpx_codec_dec_cfg_t dec_cfg_;
unsigned int passes_;
unsigned long deadline_;
TwopassStatsStore stats_;
unsigned long init_flags_;
unsigned long frame_flags_;
vpx_codec_pts_t last_pts_;
};
} // namespace libvpx_test

View File

@@ -19,16 +19,13 @@ namespace {
const int kMaxErrorFrames = 12;
const int kMaxDroppableFrames = 12;
class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, bool> {
class ErrorResilienceTestLarge
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, bool> {
protected:
ErrorResilienceTestLarge()
: EncoderTest(GET_PARAM(0)),
svc_support_(GET_PARAM(2)),
psnr_(0.0),
nframes_(0),
mismatch_psnr_(0.0),
mismatch_nframes_(0),
: EncoderTest(GET_PARAM(0)), svc_support_(GET_PARAM(2)), psnr_(0.0),
nframes_(0), mismatch_psnr_(0.0), mismatch_nframes_(0),
encoding_mode_(GET_PARAM(1)) {
Reset();
}
@@ -66,81 +63,69 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
// LAST is updated on base/layer 0, GOLDEN updated on layer 1.
// Non-zero pattern_switch parameter means pattern will switch to
// not using LAST for frame_num >= pattern_switch.
int SetFrameFlags(int frame_num,
int num_temp_layers,
int pattern_switch) {
int SetFrameFlags(int frame_num, int num_temp_layers, int pattern_switch) {
int frame_flags = 0;
if (num_temp_layers == 2) {
if (frame_num % 2 == 0) {
if (frame_num < pattern_switch || pattern_switch == 0) {
// Layer 0: predict from LAST and ARF, update LAST.
frame_flags = VP8_EFLAG_NO_REF_GF |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
} else {
// Layer 0: predict from GF and ARF, update GF.
frame_flags = VP8_EFLAG_NO_REF_LAST |
VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
}
if (frame_num % 2 == 0) {
if (frame_num < pattern_switch || pattern_switch == 0) {
// Layer 0: predict from LAST and ARF, update LAST.
frame_flags =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
} else {
if (frame_num < pattern_switch || pattern_switch == 0) {
// Layer 1: predict from L, GF, and ARF, update GF.
frame_flags = VP8_EFLAG_NO_UPD_ARF |
VP8_EFLAG_NO_UPD_LAST;
} else {
// Layer 1: predict from GF and ARF, update GF.
frame_flags = VP8_EFLAG_NO_REF_LAST |
VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
}
// Layer 0: predict from GF and ARF, update GF.
frame_flags = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
}
} else {
if (frame_num < pattern_switch || pattern_switch == 0) {
// Layer 1: predict from L, GF, and ARF, update GF.
frame_flags = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST;
} else {
// Layer 1: predict from GF and ARF, update GF.
frame_flags = VP8_EFLAG_NO_REF_LAST | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_ARF;
}
}
}
return frame_flags;
}
virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
::libvpx_test::Encoder * /*encoder*/) {
frame_flags_ &= ~(VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF);
virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video) {
frame_flags_ &=
~(VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF);
// For temporal layer case.
if (cfg_.ts_number_layers > 1) {
frame_flags_ = SetFrameFlags(video->frame(),
cfg_.ts_number_layers,
pattern_switch_);
frame_flags_ =
SetFrameFlags(video->frame(), cfg_.ts_number_layers, pattern_switch_);
for (unsigned int i = 0; i < droppable_nframes_; ++i) {
if (droppable_frames_[i] == video->frame()) {
std::cout << "Encoding droppable frame: "
<< droppable_frames_[i] << "\n";
std::cout << "Encoding droppable frame: " << droppable_frames_[i]
<< "\n";
}
}
} else {
if (droppable_nframes_ > 0 &&
(cfg_.g_pass == VPX_RC_LAST_PASS || cfg_.g_pass == VPX_RC_ONE_PASS)) {
for (unsigned int i = 0; i < droppable_nframes_; ++i) {
if (droppable_frames_[i] == video->frame()) {
std::cout << "Encoding droppable frame: "
<< droppable_frames_[i] << "\n";
frame_flags_ |= (VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF);
return;
}
}
}
if (droppable_nframes_ > 0 &&
(cfg_.g_pass == VPX_RC_LAST_PASS || cfg_.g_pass == VPX_RC_ONE_PASS)) {
for (unsigned int i = 0; i < droppable_nframes_; ++i) {
if (droppable_frames_[i] == video->frame()) {
std::cout << "Encoding droppable frame: " << droppable_frames_[i]
<< "\n";
frame_flags_ |= (VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF);
return;
}
}
}
}
}
double GetAveragePsnr() const {
if (nframes_)
return psnr_ / nframes_;
if (nframes_) return psnr_ / nframes_;
return 0.0;
}
double GetAverageMismatchPsnr() const {
if (mismatch_nframes_)
return mismatch_psnr_ / mismatch_nframes_;
if (mismatch_nframes_) return mismatch_psnr_ / mismatch_nframes_;
return 0.0;
}
@@ -158,8 +143,7 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
return 1;
}
virtual void MismatchHook(const vpx_image_t *img1,
const vpx_image_t *img2) {
virtual void MismatchHook(const vpx_image_t *img1, const vpx_image_t *img2) {
double mismatch_psnr = compute_psnr(img1, img2);
mismatch_psnr_ += mismatch_psnr;
++mismatch_nframes_;
@@ -167,32 +151,32 @@ class ErrorResilienceTestLarge : public ::libvpx_test::EncoderTest,
}
void SetErrorFrames(int num, unsigned int *list) {
if (num > kMaxErrorFrames)
if (num > kMaxErrorFrames) {
num = kMaxErrorFrames;
else if (num < 0)
} else if (num < 0) {
num = 0;
}
error_nframes_ = num;
for (unsigned int i = 0; i < error_nframes_; ++i)
for (unsigned int i = 0; i < error_nframes_; ++i) {
error_frames_[i] = list[i];
}
}
void SetDroppableFrames(int num, unsigned int *list) {
if (num > kMaxDroppableFrames)
if (num > kMaxDroppableFrames) {
num = kMaxDroppableFrames;
else if (num < 0)
} else if (num < 0) {
num = 0;
}
droppable_nframes_ = num;
for (unsigned int i = 0; i < droppable_nframes_; ++i)
for (unsigned int i = 0; i < droppable_nframes_; ++i) {
droppable_frames_[i] = list[i];
}
}
unsigned int GetMismatchFrames() {
return mismatch_nframes_;
}
unsigned int GetMismatchFrames() { return mismatch_nframes_; }
void SetPatternSwitch(int frame_switch) {
pattern_switch_ = frame_switch;
}
void SetPatternSwitch(int frame_switch) { pattern_switch_ = frame_switch; }
bool svc_support_;
@@ -265,15 +249,14 @@ TEST_P(ErrorResilienceTestLarge, DropFramesWithoutRecovery) {
// In addition to isolated loss/drop, add a long consecutive series
// (of size 9) of dropped frames.
unsigned int num_droppable_frames = 11;
unsigned int droppable_frame_list[] = {5, 16, 22, 23, 24, 25, 26, 27, 28,
29, 30};
unsigned int droppable_frame_list[] = { 5, 16, 22, 23, 24, 25,
26, 27, 28, 29, 30 };
SetDroppableFrames(num_droppable_frames, droppable_frame_list);
SetErrorFrames(num_droppable_frames, droppable_frame_list);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
// Test that no mismatches have been found
std::cout << " Mismatch frames: "
<< GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int) 0);
std::cout << " Mismatch frames: " << GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int)0);
// Reset previously set of error/droppable frames.
Reset();
@@ -306,8 +289,7 @@ TEST_P(ErrorResilienceTestLarge, DropFramesWithoutRecovery) {
// layer, so successful decoding is expected.
TEST_P(ErrorResilienceTestLarge, 2LayersDropEnhancement) {
// This test doesn't run if SVC is not supported.
if (!svc_support_)
return;
if (!svc_support_) return;
const vpx_rational timebase = { 33333333, 1000000000 };
cfg_.g_timebase = timebase;
@@ -337,14 +319,13 @@ TEST_P(ErrorResilienceTestLarge, 2LayersDropEnhancement) {
// The odd frames are the enhancement layer for 2 layer pattern, so set
// those frames as droppable. Drop the last 7 frames.
unsigned int num_droppable_frames = 7;
unsigned int droppable_frame_list[] = {27, 29, 31, 33, 35, 37, 39};
unsigned int droppable_frame_list[] = { 27, 29, 31, 33, 35, 37, 39 };
SetDroppableFrames(num_droppable_frames, droppable_frame_list);
SetErrorFrames(num_droppable_frames, droppable_frame_list);
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
// Test that no mismatches have been found
std::cout << " Mismatch frames: "
<< GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int) 0);
std::cout << " Mismatch frames: " << GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int)0);
// Reset previously set of error/droppable frames.
Reset();
@@ -355,8 +336,7 @@ TEST_P(ErrorResilienceTestLarge, 2LayersDropEnhancement) {
// sequence, the LAST ref is not used anymore.
TEST_P(ErrorResilienceTestLarge, 2LayersNoRefLast) {
// This test doesn't run if SVC is not supported.
if (!svc_support_)
return;
if (!svc_support_) return;
const vpx_rational timebase = { 33333333, 1000000000 };
cfg_.g_timebase = timebase;
@@ -385,20 +365,19 @@ TEST_P(ErrorResilienceTestLarge, 2LayersNoRefLast) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
// Test that no mismatches have been found
std::cout << " Mismatch frames: "
<< GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int) 0);
std::cout << " Mismatch frames: " << GetMismatchFrames() << "\n";
EXPECT_EQ(GetMismatchFrames(), (unsigned int)0);
// Reset previously set of error/droppable frames.
Reset();
}
class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class ErrorResilienceTestLargeCodecControls
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
ErrorResilienceTestLargeCodecControls()
: EncoderTest(GET_PARAM(0)),
encoding_mode_(GET_PARAM(1)) {
: EncoderTest(GET_PARAM(0)), encoding_mode_(GET_PARAM(1)) {
Reset();
}
@@ -437,8 +416,8 @@ class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
if (num_temp_layers == 2) {
if (frame_num % 2 == 0) {
// Layer 0: predict from L and ARF, update L.
frame_flags = VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF |
VP8_EFLAG_NO_UPD_ARF;
frame_flags =
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_ARF;
} else {
// Layer 1: predict from L, G and ARF, and update G.
frame_flags = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
@@ -451,9 +430,9 @@ class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
VP8_EFLAG_NO_REF_GF | VP8_EFLAG_NO_REF_ARF;
} else if ((frame_num - 2) % 4 == 0) {
// Layer 1: predict from L, G, update G.
frame_flags = VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST |
VP8_EFLAG_NO_REF_ARF;
} else if ((frame_num - 1) % 2 == 0) {
frame_flags =
VP8_EFLAG_NO_UPD_ARF | VP8_EFLAG_NO_UPD_LAST | VP8_EFLAG_NO_REF_ARF;
} else if ((frame_num - 1) % 2 == 0) {
// Layer 2: predict from L, G, ARF; update ARG.
frame_flags = VP8_EFLAG_NO_UPD_GF | VP8_EFLAG_NO_UPD_LAST;
}
@@ -467,7 +446,7 @@ class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
if (frame_num % 2 == 0) {
layer_id = 0;
} else {
layer_id = 1;
layer_id = 1;
}
} else if (num_temp_layers == 3) {
if (frame_num % 4 == 0) {
@@ -484,16 +463,16 @@ class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
virtual void PreEncodeFrameHook(libvpx_test::VideoSource *video,
libvpx_test::Encoder *encoder) {
if (cfg_.ts_number_layers > 1) {
int layer_id = SetLayerId(video->frame(), cfg_.ts_number_layers);
int frame_flags = SetFrameFlags(video->frame(), cfg_.ts_number_layers);
if (video->frame() > 0) {
encoder->Control(VP8E_SET_TEMPORAL_LAYER_ID, layer_id);
encoder->Control(VP8E_SET_FRAME_FLAGS, frame_flags);
}
const vpx_rational_t tb = video->timebase();
timebase_ = static_cast<double>(tb.num) / tb.den;
duration_ = 0;
return;
int layer_id = SetLayerId(video->frame(), cfg_.ts_number_layers);
int frame_flags = SetFrameFlags(video->frame(), cfg_.ts_number_layers);
if (video->frame() > 0) {
encoder->Control(VP8E_SET_TEMPORAL_LAYER_ID, layer_id);
encoder->Control(VP8E_SET_FRAME_FLAGS, frame_flags);
}
const vpx_rational_t tb = video->timebase();
timebase_ = static_cast<double>(tb.num) / tb.den;
duration_ = 0;
return;
}
}
@@ -519,26 +498,28 @@ class ErrorResilienceTestLargeCodecControls : public ::libvpx_test::EncoderTest,
virtual void EndPassHook(void) {
duration_ = (last_pts_ + 1) * timebase_;
if (cfg_.ts_number_layers > 1) {
if (cfg_.ts_number_layers > 1) {
for (int layer = 0; layer < static_cast<int>(cfg_.ts_number_layers);
++layer) {
++layer) {
if (bits_total_[layer]) {
// Effective file datarate:
effective_datarate_[layer] = (bits_total_[layer] / 1000.0) / duration_;
effective_datarate_[layer] =
(bits_total_[layer] / 1000.0) / duration_;
}
}
}
}
double effective_datarate_[3];
private:
libvpx_test::TestMode encoding_mode_;
vpx_codec_pts_t last_pts_;
double timebase_;
int64_t bits_total_[3];
double duration_;
int tot_frame_number_;
};
private:
libvpx_test::TestMode encoding_mode_;
vpx_codec_pts_t last_pts_;
double timebase_;
int64_t bits_total_[3];
double duration_;
int tot_frame_number_;
};
// Check two codec controls used for:
// (1) for setting temporal layer id, and (2) for settings encoder flags.
@@ -582,10 +563,12 @@ TEST_P(ErrorResilienceTestLargeCodecControls, CodecControl3TemporalLayers) {
for (int j = 0; j < static_cast<int>(cfg_.ts_number_layers); ++j) {
ASSERT_GE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 0.75)
<< " The datarate for the file is lower than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
ASSERT_LE(effective_datarate_[j], cfg_.ts_target_bitrate[j] * 1.25)
<< " The datarate for the file is greater than target by too much, "
"for layer: " << j;
"for layer: "
<< j;
}
}
}

View File

@@ -15,7 +15,7 @@
example_tests=$(ls $(dirname $0)/*.sh)
# List of script names to exclude.
exclude_list="examples tools_common"
exclude_list="examples stress tools_common"
# Filter out the scripts in $exclude_list.
for word in ${exclude_list}; do

View File

@@ -34,21 +34,18 @@ struct ExternalFrameBuffer {
// Class to manipulate a list of external frame buffers.
class ExternalFrameBufferList {
public:
ExternalFrameBufferList()
: num_buffers_(0),
ext_fb_list_(NULL) {}
ExternalFrameBufferList() : num_buffers_(0), ext_fb_list_(NULL) {}
virtual ~ExternalFrameBufferList() {
for (int i = 0; i < num_buffers_; ++i) {
delete [] ext_fb_list_[i].data;
delete[] ext_fb_list_[i].data;
}
delete [] ext_fb_list_;
delete[] ext_fb_list_;
}
// Creates the list to hold the external buffers. Returns true on success.
bool CreateBufferList(int num_buffers) {
if (num_buffers < 0)
return false;
if (num_buffers < 0) return false;
num_buffers_ = num_buffers;
ext_fb_list_ = new ExternalFrameBuffer[num_buffers_];
@@ -64,11 +61,10 @@ class ExternalFrameBufferList {
int GetFreeFrameBuffer(size_t min_size, vpx_codec_frame_buffer_t *fb) {
EXPECT_TRUE(fb != NULL);
const int idx = FindFreeBufferIndex();
if (idx == num_buffers_)
return -1;
if (idx == num_buffers_) return -1;
if (ext_fb_list_[idx].size < min_size) {
delete [] ext_fb_list_[idx].data;
delete[] ext_fb_list_[idx].data;
ext_fb_list_[idx].data = new uint8_t[min_size];
memset(ext_fb_list_[idx].data, 0, min_size);
ext_fb_list_[idx].size = min_size;
@@ -83,11 +79,10 @@ class ExternalFrameBufferList {
int GetZeroFrameBuffer(size_t min_size, vpx_codec_frame_buffer_t *fb) {
EXPECT_TRUE(fb != NULL);
const int idx = FindFreeBufferIndex();
if (idx == num_buffers_)
return -1;
if (idx == num_buffers_) return -1;
if (ext_fb_list_[idx].size < min_size) {
delete [] ext_fb_list_[idx].data;
delete[] ext_fb_list_[idx].data;
ext_fb_list_[idx].data = NULL;
ext_fb_list_[idx].size = min_size;
}
@@ -104,7 +99,7 @@ class ExternalFrameBufferList {
return -1;
}
ExternalFrameBuffer *const ext_fb =
reinterpret_cast<ExternalFrameBuffer*>(fb->priv);
reinterpret_cast<ExternalFrameBuffer *>(fb->priv);
if (ext_fb == NULL) {
EXPECT_TRUE(ext_fb != NULL);
return -1;
@@ -119,7 +114,7 @@ class ExternalFrameBufferList {
void CheckXImageFrameBuffer(const vpx_image_t *img) {
if (img->fb_priv != NULL) {
const struct ExternalFrameBuffer *const ext_fb =
reinterpret_cast<ExternalFrameBuffer*>(img->fb_priv);
reinterpret_cast<ExternalFrameBuffer *>(img->fb_priv);
ASSERT_TRUE(img->planes[0] >= ext_fb->data &&
img->planes[0] < (ext_fb->data + ext_fb->size));
@@ -133,8 +128,7 @@ class ExternalFrameBufferList {
int i;
// Find a free frame buffer.
for (i = 0; i < num_buffers_; ++i) {
if (!ext_fb_list_[i].in_use)
break;
if (!ext_fb_list_[i].in_use) break;
}
return i;
}
@@ -161,16 +155,15 @@ class ExternalFrameBufferList {
int get_vp9_frame_buffer(void *user_priv, size_t min_size,
vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferList *const fb_list =
reinterpret_cast<ExternalFrameBufferList*>(user_priv);
reinterpret_cast<ExternalFrameBufferList *>(user_priv);
return fb_list->GetFreeFrameBuffer(min_size, fb);
}
// Callback used by libvpx to tell the application that |fb| is not needed
// anymore.
int release_vp9_frame_buffer(void *user_priv,
vpx_codec_frame_buffer_t *fb) {
int release_vp9_frame_buffer(void *user_priv, vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferList *const fb_list =
reinterpret_cast<ExternalFrameBufferList*>(user_priv);
reinterpret_cast<ExternalFrameBufferList *>(user_priv);
return fb_list->ReturnFrameBuffer(fb);
}
@@ -178,7 +171,7 @@ int release_vp9_frame_buffer(void *user_priv,
int get_vp9_zero_frame_buffer(void *user_priv, size_t min_size,
vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferList *const fb_list =
reinterpret_cast<ExternalFrameBufferList*>(user_priv);
reinterpret_cast<ExternalFrameBufferList *>(user_priv);
return fb_list->GetZeroFrameBuffer(min_size, fb);
}
@@ -186,7 +179,7 @@ int get_vp9_zero_frame_buffer(void *user_priv, size_t min_size,
int get_vp9_one_less_byte_frame_buffer(void *user_priv, size_t min_size,
vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferList *const fb_list =
reinterpret_cast<ExternalFrameBufferList*>(user_priv);
reinterpret_cast<ExternalFrameBufferList *>(user_priv);
return fb_list->GetFreeFrameBuffer(min_size - 1, fb);
}
@@ -203,16 +196,14 @@ int do_not_release_vp9_frame_buffer(void *user_priv,
// Class for testing passing in external frame buffers to libvpx.
class ExternalFrameBufferMD5Test
: public ::libvpx_test::DecoderTest,
public ::libvpx_test::CodecTestWithParam<const char*> {
public ::libvpx_test::CodecTestWithParam<const char *> {
protected:
ExternalFrameBufferMD5Test()
: DecoderTest(GET_PARAM(::libvpx_test::kCodecFactoryParam)),
md5_file_(NULL),
num_buffers_(0) {}
md5_file_(NULL), num_buffers_(0) {}
virtual ~ExternalFrameBufferMD5Test() {
if (md5_file_ != NULL)
fclose(md5_file_);
if (md5_file_ != NULL) fclose(md5_file_);
}
virtual void PreDecodeFrameHook(
@@ -222,15 +213,15 @@ class ExternalFrameBufferMD5Test
// Have libvpx use frame buffers we create.
ASSERT_TRUE(fb_list_.CreateBufferList(num_buffers_));
ASSERT_EQ(VPX_CODEC_OK,
decoder->SetFrameBufferFunctions(
GetVP9FrameBuffer, ReleaseVP9FrameBuffer, this));
decoder->SetFrameBufferFunctions(GetVP9FrameBuffer,
ReleaseVP9FrameBuffer, this));
}
}
void OpenMD5File(const std::string &md5_file_name_) {
md5_file_ = libvpx_test::OpenTestDataFile(md5_file_name_);
ASSERT_TRUE(md5_file_ != NULL) << "Md5 file open failed. Filename: "
<< md5_file_name_;
<< md5_file_name_;
}
virtual void DecompressedFrameHook(const vpx_image_t &img,
@@ -258,7 +249,7 @@ class ExternalFrameBufferMD5Test
static int GetVP9FrameBuffer(void *user_priv, size_t min_size,
vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferMD5Test *const md5Test =
reinterpret_cast<ExternalFrameBufferMD5Test*>(user_priv);
reinterpret_cast<ExternalFrameBufferMD5Test *>(user_priv);
return md5Test->fb_list_.GetFreeFrameBuffer(min_size, fb);
}
@@ -267,7 +258,7 @@ class ExternalFrameBufferMD5Test
static int ReleaseVP9FrameBuffer(void *user_priv,
vpx_codec_frame_buffer_t *fb) {
ExternalFrameBufferMD5Test *const md5Test =
reinterpret_cast<ExternalFrameBufferMD5Test*>(user_priv);
reinterpret_cast<ExternalFrameBufferMD5Test *>(user_priv);
return md5Test->fb_list_.ReturnFrameBuffer(fb);
}
@@ -286,10 +277,7 @@ const char kVP9TestFile[] = "vp90-2-02-size-lf-1920x1080.webm";
// Class for testing passing in external frame buffers to libvpx.
class ExternalFrameBufferTest : public ::testing::Test {
protected:
ExternalFrameBufferTest()
: video_(NULL),
decoder_(NULL),
num_buffers_(0) {}
ExternalFrameBufferTest() : video_(NULL), decoder_(NULL), num_buffers_(0) {}
virtual void SetUp() {
video_ = new libvpx_test::WebMVideoSource(kVP9TestFile);
@@ -309,8 +297,7 @@ class ExternalFrameBufferTest : public ::testing::Test {
// Passes the external frame buffer information to libvpx.
vpx_codec_err_t SetFrameBufferFunctions(
int num_buffers,
vpx_get_frame_buffer_cb_fn_t cb_get,
int num_buffers, vpx_get_frame_buffer_cb_fn_t cb_get,
vpx_release_frame_buffer_cb_fn_t cb_release) {
if (num_buffers > 0) {
num_buffers_ = num_buffers;
@@ -324,8 +311,7 @@ class ExternalFrameBufferTest : public ::testing::Test {
const vpx_codec_err_t res =
decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
CheckDecodedFrames();
if (res == VPX_CODEC_OK)
video_->Next();
if (res == VPX_CODEC_OK) video_->Next();
return res;
}
@@ -333,8 +319,7 @@ class ExternalFrameBufferTest : public ::testing::Test {
for (; video_->cxdata() != NULL; video_->Next()) {
const vpx_codec_err_t res =
decoder_->DecodeFrame(video_->cxdata(), video_->frame_size());
if (res != VPX_CODEC_OK)
return res;
if (res != VPX_CODEC_OK) return res;
CheckDecodedFrames();
}
return VPX_CODEC_OK;
@@ -365,7 +350,6 @@ class ExternalFrameBufferTest : public ::testing::Test {
// Otherwise, the test failed.
TEST_P(ExternalFrameBufferMD5Test, ExtFBMD5Match) {
const std::string filename = GET_PARAM(kVideoNameParam);
libvpx_test::CompressedVideoSource *video = NULL;
// Number of buffers equals #VP9_MAXIMUM_REF_BUFFERS +
// #VPX_MAXIMUM_WORK_BUFFERS + four jitter buffers.
@@ -380,18 +364,19 @@ TEST_P(ExternalFrameBufferMD5Test, ExtFBMD5Match) {
#endif
// Open compressed video file.
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
if (filename.substr(filename.length() - 3, 3) == "ivf") {
video = new libvpx_test::IVFVideoSource(filename);
video.reset(new libvpx_test::IVFVideoSource(filename));
} else {
#if CONFIG_WEBM_IO
video = new libvpx_test::WebMVideoSource(filename);
video.reset(new libvpx_test::WebMVideoSource(filename));
#else
fprintf(stderr, "WebM IO is disabled, skipping test vector %s\n",
filename.c_str());
return;
#endif
}
ASSERT_TRUE(video != NULL);
ASSERT_TRUE(video.get() != NULL);
video->Init();
// Construct md5 file name.
@@ -399,8 +384,7 @@ TEST_P(ExternalFrameBufferMD5Test, ExtFBMD5Match) {
OpenMD5File(md5_filename);
// Decode frame, and check the md5 matching.
ASSERT_NO_FATAL_FAILURE(RunLoop(video));
delete video;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get()));
}
#if CONFIG_WEBM_IO
@@ -409,8 +393,8 @@ TEST_F(ExternalFrameBufferTest, MinFrameBuffers) {
// #VP9_MAXIMUM_REF_BUFFERS + #VPX_MAXIMUM_WORK_BUFFERS.
const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
ASSERT_EQ(VPX_CODEC_OK,
SetFrameBufferFunctions(
num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer,
release_vp9_frame_buffer));
ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames());
}
@@ -421,8 +405,8 @@ TEST_F(ExternalFrameBufferTest, EightJitterBuffers) {
const int num_buffers =
VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS + jitter_buffers;
ASSERT_EQ(VPX_CODEC_OK,
SetFrameBufferFunctions(
num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer,
release_vp9_frame_buffer));
ASSERT_EQ(VPX_CODEC_OK, DecodeRemainingFrames());
}
@@ -432,8 +416,8 @@ TEST_F(ExternalFrameBufferTest, NotEnoughBuffers) {
// only use 5 frame buffers at one time.
const int num_buffers = 2;
ASSERT_EQ(VPX_CODEC_OK,
SetFrameBufferFunctions(
num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer,
release_vp9_frame_buffer));
ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame());
ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeRemainingFrames());
}
@@ -457,18 +441,17 @@ TEST_F(ExternalFrameBufferTest, NullRealloc) {
TEST_F(ExternalFrameBufferTest, ReallocOneLessByte) {
const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
ASSERT_EQ(VPX_CODEC_OK,
SetFrameBufferFunctions(
num_buffers, get_vp9_one_less_byte_frame_buffer,
release_vp9_frame_buffer));
ASSERT_EQ(VPX_CODEC_OK, SetFrameBufferFunctions(
num_buffers, get_vp9_one_less_byte_frame_buffer,
release_vp9_frame_buffer));
ASSERT_EQ(VPX_CODEC_MEM_ERROR, DecodeOneFrame());
}
TEST_F(ExternalFrameBufferTest, NullGetFunction) {
const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
ASSERT_EQ(VPX_CODEC_INVALID_PARAM,
SetFrameBufferFunctions(num_buffers, NULL,
release_vp9_frame_buffer));
ASSERT_EQ(
VPX_CODEC_INVALID_PARAM,
SetFrameBufferFunctions(num_buffers, NULL, release_vp9_frame_buffer));
}
TEST_F(ExternalFrameBufferTest, NullReleaseFunction) {
@@ -481,13 +464,14 @@ TEST_F(ExternalFrameBufferTest, SetAfterDecode) {
const int num_buffers = VP9_MAXIMUM_REF_BUFFERS + VPX_MAXIMUM_WORK_BUFFERS;
ASSERT_EQ(VPX_CODEC_OK, DecodeOneFrame());
ASSERT_EQ(VPX_CODEC_ERROR,
SetFrameBufferFunctions(
num_buffers, get_vp9_frame_buffer, release_vp9_frame_buffer));
SetFrameBufferFunctions(num_buffers, get_vp9_frame_buffer,
release_vp9_frame_buffer));
}
#endif // CONFIG_WEBM_IO
VP9_INSTANTIATE_TEST_CASE(ExternalFrameBufferMD5Test,
::testing::ValuesIn(libvpx_test::kVP9TestVectors,
libvpx_test::kVP9TestVectors +
libvpx_test::kNumVP9TestVectors));
VP9_INSTANTIATE_TEST_CASE(
ExternalFrameBufferMD5Test,
::testing::ValuesIn(libvpx_test::kVP9TestVectors,
libvpx_test::kVP9TestVectors +
libvpx_test::kNumVP9TestVectors));
} // namespace

View File

@@ -128,14 +128,14 @@ class Trans4x4TestBase {
}
}
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(test_input_block,
test_temp_block, pitch_));
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(test_input_block, test_temp_block, pitch_));
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block,
CONVERT_TO_BYTEPTR(dst16), pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(test_temp_block, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif
}
@@ -148,15 +148,13 @@ class Trans4x4TestBase {
const int diff = dst[j] - src[j];
#endif
const uint32_t error = diff * diff;
if (max_error < error)
max_error = error;
if (max_error < error) max_error = error;
total_error += error;
}
}
EXPECT_GE(static_cast<uint32_t>(limit), max_error)
<< "Error: 4x4 FHT/IHT has an individual round trip error > "
<< limit;
<< "Error: 4x4 FHT/IHT has an individual round trip error > " << limit;
EXPECT_GE(count_test_block * limit, total_error)
<< "Error: 4x4 FHT/IHT has average round trip error > " << limit
@@ -172,8 +170,9 @@ class Trans4x4TestBase {
for (int i = 0; i < count_test_block; ++i) {
// Initialize a test block with input range [-mask_, mask_].
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
input_block[j] = (rnd.Rand16() & mask_) - (rnd.Rand16() & mask_);
}
fwd_txfm_ref(input_block, output_ref_block, pitch_, tx_type_);
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_block, output_block, pitch_));
@@ -197,16 +196,14 @@ class Trans4x4TestBase {
input_extreme_block[j] = rnd.Rand8() % 2 ? mask_ : -mask_;
}
if (i == 0) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = mask_;
} else if (i == 1) {
for (int j = 0; j < kNumCoeffs; ++j)
input_extreme_block[j] = -mask_;
for (int j = 0; j < kNumCoeffs; ++j) input_extreme_block[j] = -mask_;
}
fwd_txfm_ref(input_extreme_block, output_ref_block, pitch_, tx_type_);
ASM_REGISTER_STATE_CHECK(RunFwdTxfm(input_extreme_block,
output_block, pitch_));
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(input_extreme_block, output_block, pitch_));
// The minimum quant value is 4.
for (int j = 0; j < kNumCoeffs; ++j) {
@@ -251,8 +248,8 @@ class Trans4x4TestBase {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif
}
@@ -265,8 +262,7 @@ class Trans4x4TestBase {
#endif
const uint32_t error = diff * diff;
EXPECT_GE(static_cast<uint32_t>(limit), error)
<< "Error: 4x4 IDCT has error " << error
<< " at index " << j;
<< "Error: 4x4 IDCT has error " << error << " at index " << j;
}
}
}
@@ -278,17 +274,16 @@ class Trans4x4TestBase {
int mask_;
};
class Trans4x4DCT
: public Trans4x4TestBase,
public ::testing::TestWithParam<Dct4x4Param> {
class Trans4x4DCT : public Trans4x4TestBase,
public ::testing::TestWithParam<Dct4x4Param> {
public:
virtual ~Trans4x4DCT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
pitch_ = 4;
tx_type_ = GET_PARAM(2);
pitch_ = 4;
fwd_txfm_ref = fdct4x4_ref;
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
@@ -307,33 +302,24 @@ class Trans4x4DCT
IdctFunc inv_txfm_;
};
TEST_P(Trans4x4DCT, AccuracyCheck) {
RunAccuracyCheck(1);
}
TEST_P(Trans4x4DCT, AccuracyCheck) { RunAccuracyCheck(1); }
TEST_P(Trans4x4DCT, CoeffCheck) {
RunCoeffCheck();
}
TEST_P(Trans4x4DCT, CoeffCheck) { RunCoeffCheck(); }
TEST_P(Trans4x4DCT, MemCheck) {
RunMemCheck();
}
TEST_P(Trans4x4DCT, MemCheck) { RunMemCheck(); }
TEST_P(Trans4x4DCT, InvAccuracyCheck) {
RunInvAccuracyCheck(1);
}
TEST_P(Trans4x4DCT, InvAccuracyCheck) { RunInvAccuracyCheck(1); }
class Trans4x4HT
: public Trans4x4TestBase,
public ::testing::TestWithParam<Ht4x4Param> {
class Trans4x4HT : public Trans4x4TestBase,
public ::testing::TestWithParam<Ht4x4Param> {
public:
virtual ~Trans4x4HT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
pitch_ = 4;
tx_type_ = GET_PARAM(2);
pitch_ = 4;
fwd_txfm_ref = fht4x4_ref;
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
@@ -353,33 +339,24 @@ class Trans4x4HT
IhtFunc inv_txfm_;
};
TEST_P(Trans4x4HT, AccuracyCheck) {
RunAccuracyCheck(1);
}
TEST_P(Trans4x4HT, AccuracyCheck) { RunAccuracyCheck(1); }
TEST_P(Trans4x4HT, CoeffCheck) {
RunCoeffCheck();
}
TEST_P(Trans4x4HT, CoeffCheck) { RunCoeffCheck(); }
TEST_P(Trans4x4HT, MemCheck) {
RunMemCheck();
}
TEST_P(Trans4x4HT, MemCheck) { RunMemCheck(); }
TEST_P(Trans4x4HT, InvAccuracyCheck) {
RunInvAccuracyCheck(1);
}
TEST_P(Trans4x4HT, InvAccuracyCheck) { RunInvAccuracyCheck(1); }
class Trans4x4WHT
: public Trans4x4TestBase,
public ::testing::TestWithParam<Dct4x4Param> {
class Trans4x4WHT : public Trans4x4TestBase,
public ::testing::TestWithParam<Dct4x4Param> {
public:
virtual ~Trans4x4WHT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
pitch_ = 4;
tx_type_ = GET_PARAM(2);
pitch_ = 4;
fwd_txfm_ref = fwht4x4_ref;
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
@@ -398,21 +375,13 @@ class Trans4x4WHT
IdctFunc inv_txfm_;
};
TEST_P(Trans4x4WHT, AccuracyCheck) {
RunAccuracyCheck(0);
}
TEST_P(Trans4x4WHT, AccuracyCheck) { RunAccuracyCheck(0); }
TEST_P(Trans4x4WHT, CoeffCheck) {
RunCoeffCheck();
}
TEST_P(Trans4x4WHT, CoeffCheck) { RunCoeffCheck(); }
TEST_P(Trans4x4WHT, MemCheck) {
RunMemCheck();
}
TEST_P(Trans4x4WHT, MemCheck) { RunMemCheck(); }
TEST_P(Trans4x4WHT, InvAccuracyCheck) {
RunInvAccuracyCheck(0);
}
TEST_P(Trans4x4WHT, InvAccuracyCheck) { RunInvAccuracyCheck(0); }
using std::tr1::make_tuple;
#if CONFIG_VP9_HIGHBITDEPTH
@@ -423,10 +392,10 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_12, 0, VPX_BITS_12),
make_tuple(&vpx_fdct4x4_c, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8)));
#else
INSTANTIATE_TEST_CASE_P(
C, Trans4x4DCT,
::testing::Values(
make_tuple(&vpx_fdct4x4_c, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(C, Trans4x4DCT,
::testing::Values(make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_c, 0,
VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if CONFIG_VP9_HIGHBITDEPTH
@@ -463,21 +432,18 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vp9_highbd_fwht4x4_c, &iwht4x4_12, 0, VPX_BITS_12),
make_tuple(&vp9_fwht4x4_c, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8)));
#else
INSTANTIATE_TEST_CASE_P(
C, Trans4x4WHT,
::testing::Values(
make_tuple(&vp9_fwht4x4_c, &vpx_iwht4x4_16_add_c, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(C, Trans4x4WHT,
::testing::Values(make_tuple(&vp9_fwht4x4_c,
&vpx_iwht4x4_16_add_c, 0,
VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
NEON, Trans4x4DCT,
::testing::Values(
make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_neon, 0, VPX_BITS_8)));
#endif // HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_NEON && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(NEON, Trans4x4DCT,
::testing::Values(make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_neon,
0, VPX_BITS_8)));
#if !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
NEON, Trans4x4HT,
::testing::Values(
@@ -485,9 +451,10 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 1, VPX_BITS_8),
make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 2, VPX_BITS_8),
make_tuple(&vp9_fht4x4_c, &vp9_iht4x4_16_add_neon, 3, VPX_BITS_8)));
#endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#endif // !CONFIG_VP9_HIGHBITDEPTH
#endif // HAVE_NEON && !CONFIG_EMULATE_HARDWARE
#if CONFIG_USE_X86INC && HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, Trans4x4WHT,
::testing::Values(
@@ -496,11 +463,10 @@ INSTANTIATE_TEST_CASE_P(
#endif
#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, Trans4x4DCT,
::testing::Values(
make_tuple(&vpx_fdct4x4_sse2,
&vpx_idct4x4_16_add_sse2, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(SSE2, Trans4x4DCT,
::testing::Values(make_tuple(&vpx_fdct4x4_sse2,
&vpx_idct4x4_16_add_sse2,
0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
SSE2, Trans4x4HT,
::testing::Values(
@@ -514,12 +480,11 @@ INSTANTIATE_TEST_CASE_P(
INSTANTIATE_TEST_CASE_P(
SSE2, Trans4x4DCT,
::testing::Values(
make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_10_sse2, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_10_sse2, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_10_sse2, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_12_sse2, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct4x4_c, &idct4x4_12_sse2, 0, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct4x4_sse2, &idct4x4_12_sse2, 0, VPX_BITS_12),
make_tuple(&vpx_fdct4x4_sse2, &vpx_idct4x4_16_add_c, 0,
VPX_BITS_8)));
make_tuple(&vpx_fdct4x4_sse2, &vpx_idct4x4_16_add_c, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
SSE2, Trans4x4HT,
@@ -531,10 +496,10 @@ INSTANTIATE_TEST_CASE_P(
#endif // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
MSA, Trans4x4DCT,
::testing::Values(
make_tuple(&vpx_fdct4x4_msa, &vpx_idct4x4_16_add_msa, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(MSA, Trans4x4DCT,
::testing::Values(make_tuple(&vpx_fdct4x4_msa,
&vpx_idct4x4_16_add_msa, 0,
VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
MSA, Trans4x4HT,
::testing::Values(

View File

@@ -51,10 +51,10 @@ void reference_8x8_dct_1d(const double in[8], double out[8]) {
const double kInvSqrt2 = 0.707106781186547524400844362104;
for (int k = 0; k < 8; k++) {
out[k] = 0.0;
for (int n = 0; n < 8; n++)
for (int n = 0; n < 8; n++) {
out[k] += in[n] * cos(kPi * (2 * n + 1) * k / 16.0);
if (k == 0)
out[k] = out[k] * kInvSqrt2;
}
if (k == 0) out[k] = out[k] * kInvSqrt2;
}
}
@@ -63,25 +63,20 @@ void reference_8x8_dct_2d(const int16_t input[kNumCoeffs],
// First transform columns
for (int i = 0; i < 8; ++i) {
double temp_in[8], temp_out[8];
for (int j = 0; j < 8; ++j)
temp_in[j] = input[j*8 + i];
for (int j = 0; j < 8; ++j) temp_in[j] = input[j * 8 + i];
reference_8x8_dct_1d(temp_in, temp_out);
for (int j = 0; j < 8; ++j)
output[j * 8 + i] = temp_out[j];
for (int j = 0; j < 8; ++j) output[j * 8 + i] = temp_out[j];
}
// Then transform rows
for (int i = 0; i < 8; ++i) {
double temp_in[8], temp_out[8];
for (int j = 0; j < 8; ++j)
temp_in[j] = output[j + i*8];
for (int j = 0; j < 8; ++j) temp_in[j] = output[j + i * 8];
reference_8x8_dct_1d(temp_in, temp_out);
// Scale by some magic number
for (int j = 0; j < 8; ++j)
output[j + i * 8] = temp_out[j] * 2;
for (int j = 0; j < 8; ++j) output[j + i * 8] = temp_out[j] * 2;
}
}
void fdct8x8_ref(const int16_t *in, tran_low_t *out, int stride,
int /*tx_type*/) {
vpx_fdct8x8_c(in, out, stride);
@@ -110,20 +105,20 @@ void iht8x8_12(const tran_low_t *in, uint8_t *out, int stride, int tx_type) {
#if HAVE_SSE2
void idct8x8_10_add_10_c(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_10_add_c(in, out, stride, 10);
void idct8x8_12_add_10_c(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_12_add_c(in, out, stride, 10);
}
void idct8x8_10_add_12_c(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_10_add_c(in, out, stride, 12);
void idct8x8_12_add_12_c(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_12_add_c(in, out, stride, 12);
}
void idct8x8_10_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_10_add_sse2(in, out, stride, 10);
void idct8x8_12_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_12_add_sse2(in, out, stride, 10);
}
void idct8x8_10_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_10_add_sse2(in, out, stride, 12);
void idct8x8_12_add_12_sse2(const tran_low_t *in, uint8_t *out, int stride) {
vpx_highbd_idct8x8_12_add_sse2(in, out, stride, 12);
}
void idct8x8_64_add_10_sse2(const tran_low_t *in, uint8_t *out, int stride) {
@@ -155,17 +150,19 @@ class FwdTrans8x8TestBase {
for (int i = 0; i < count_test_block; ++i) {
// Initialize a test block with input range [-255, 255].
for (int j = 0; j < 64; ++j)
for (int j = 0; j < 64; ++j) {
test_input_block[j] = ((rnd.Rand16() >> (16 - bit_depth_)) & mask_) -
((rnd.Rand16() >> (16 - bit_depth_)) & mask_);
}
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(test_input_block, test_output_block, pitch_));
for (int j = 0; j < 64; ++j) {
if (test_output_block[j] < 0)
if (test_output_block[j] < 0) {
++count_sign_block[j][0];
else if (test_output_block[j] > 0)
} else if (test_output_block[j] > 0) {
++count_sign_block[j][1];
}
}
}
@@ -177,25 +174,26 @@ class FwdTrans8x8TestBase {
<< 1. * max_diff / count_test_block * 100 << "%"
<< " for input range [-255, 255] at index " << j
<< " count0: " << count_sign_block[j][0]
<< " count1: " << count_sign_block[j][1]
<< " diff: " << diff;
<< " count1: " << count_sign_block[j][1] << " diff: " << diff;
}
memset(count_sign_block, 0, sizeof(count_sign_block));
for (int i = 0; i < count_test_block; ++i) {
// Initialize a test block with input range [-mask_ / 16, mask_ / 16].
for (int j = 0; j < 64; ++j)
test_input_block[j] = ((rnd.Rand16() & mask_) >> 4) -
((rnd.Rand16() & mask_) >> 4);
for (int j = 0; j < 64; ++j) {
test_input_block[j] =
((rnd.Rand16() & mask_) >> 4) - ((rnd.Rand16() & mask_) >> 4);
}
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(test_input_block, test_output_block, pitch_));
for (int j = 0; j < 64; ++j) {
if (test_output_block[j] < 0)
if (test_output_block[j] < 0) {
++count_sign_block[j][0];
else if (test_output_block[j] > 0)
} else if (test_output_block[j] > 0) {
++count_sign_block[j][1];
}
}
}
@@ -207,8 +205,7 @@ class FwdTrans8x8TestBase {
<< 1. * max_diff / count_test_block * 100 << "%"
<< " for input range [-15, 15] at index " << j
<< " count0: " << count_sign_block[j][0]
<< " count1: " << count_sign_block[j][1]
<< " diff: " << diff;
<< " count1: " << count_sign_block[j][1] << " diff: " << diff;
}
}
@@ -245,19 +242,18 @@ class FwdTrans8x8TestBase {
ASM_REGISTER_STATE_CHECK(
RunFwdTxfm(test_input_block, test_temp_block, pitch_));
for (int j = 0; j < 64; ++j) {
if (test_temp_block[j] > 0) {
test_temp_block[j] += 2;
test_temp_block[j] /= 4;
test_temp_block[j] *= 4;
} else {
test_temp_block[j] -= 2;
test_temp_block[j] /= 4;
test_temp_block[j] *= 4;
}
if (test_temp_block[j] > 0) {
test_temp_block[j] += 2;
test_temp_block[j] /= 4;
test_temp_block[j] *= 4;
} else {
test_temp_block[j] -= 2;
test_temp_block[j] /= 4;
test_temp_block[j] *= 4;
}
}
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(test_temp_block, dst, pitch_));
ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(
@@ -273,19 +269,18 @@ class FwdTrans8x8TestBase {
const int diff = dst[j] - src[j];
#endif
const int error = diff * diff;
if (max_error < error)
max_error = error;
if (max_error < error) max_error = error;
total_error += error;
}
}
EXPECT_GE(1 << 2 * (bit_depth_ - 8), max_error)
<< "Error: 8x8 FDCT/IDCT or FHT/IHT has an individual"
<< " roundtrip error > 1";
<< "Error: 8x8 FDCT/IDCT or FHT/IHT has an individual"
<< " roundtrip error > 1";
EXPECT_GE((count_test_block << 2 * (bit_depth_ - 8))/5, total_error)
<< "Error: 8x8 FDCT/IDCT or FHT/IHT has average roundtrip "
<< "error > 1/5 per block";
EXPECT_GE((count_test_block << 2 * (bit_depth_ - 8)) / 5, total_error)
<< "Error: 8x8 FDCT/IDCT or FHT/IHT has average roundtrip "
<< "error > 1/5 per block";
}
void RunExtremalCheck() {
@@ -341,8 +336,7 @@ class FwdTrans8x8TestBase {
ASM_REGISTER_STATE_CHECK(
fwd_txfm_ref(test_input_block, ref_temp_block, pitch_, tx_type_));
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(test_temp_block, dst, pitch_));
ASM_REGISTER_STATE_CHECK(RunInvTxfm(test_temp_block, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(
@@ -358,8 +352,7 @@ class FwdTrans8x8TestBase {
const int diff = dst[j] - src[j];
#endif
const int error = diff * diff;
if (max_error < error)
max_error = error;
if (max_error < error) max_error = error;
total_error += error;
const int coeff_diff = test_temp_block[j] - ref_temp_block[j];
@@ -370,7 +363,7 @@ class FwdTrans8x8TestBase {
<< "Error: Extremal 8x8 FDCT/IDCT or FHT/IHT has"
<< "an individual roundtrip error > 1";
EXPECT_GE((count_test_block << 2 * (bit_depth_ - 8))/5, total_error)
EXPECT_GE((count_test_block << 2 * (bit_depth_ - 8)) / 5, total_error)
<< "Error: Extremal 8x8 FDCT/IDCT or FHT/IHT has average"
<< " roundtrip error > 1/5 per block";
@@ -411,15 +404,16 @@ class FwdTrans8x8TestBase {
}
reference_8x8_dct_2d(in, out_r);
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
coeff[j] = static_cast<tran_low_t>(round(out_r[j]));
}
if (bit_depth_ == VPX_BITS_8) {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, dst, pitch_));
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif
}
@@ -432,8 +426,7 @@ class FwdTrans8x8TestBase {
#endif
const uint32_t error = diff * diff;
EXPECT_GE(1u << 2 * (bit_depth_ - 8), error)
<< "Error: 8x8 IDCT has error " << error
<< " at index " << j;
<< "Error: 8x8 IDCT has error " << error << " at index " << j;
}
}
}
@@ -449,25 +442,26 @@ class FwdTrans8x8TestBase {
double out_r[kNumCoeffs];
// Initialize a test block with input range [-mask_, mask_].
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
in[j] = rnd.Rand8() % 2 == 0 ? mask_ : -mask_;
}
RunFwdTxfm(in, coeff, pitch_);
reference_8x8_dct_2d(in, out_r);
for (int j = 0; j < kNumCoeffs; ++j)
for (int j = 0; j < kNumCoeffs; ++j) {
coeff_r[j] = static_cast<tran_low_t>(round(out_r[j]));
}
for (int j = 0; j < kNumCoeffs; ++j) {
const int32_t diff = coeff[j] - coeff_r[j];
const uint32_t error = diff * diff;
EXPECT_GE(9u << 2 * (bit_depth_ - 8), error)
<< "Error: 8x8 DCT has error " << error
<< " at index " << j;
<< "Error: 8x8 DCT has error " << error << " at index " << j;
}
}
}
void CompareInvReference(IdctFunc ref_txfm, int thresh) {
void CompareInvReference(IdctFunc ref_txfm, int thresh) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = 10000;
const int eob = 12;
@@ -484,7 +478,7 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
for (int j = 0; j < kNumCoeffs; ++j) {
if (j < eob) {
// Random values less than the threshold, either positive or negative
coeff[scan[j]] = rnd(thresh) * (1-2*(i%2));
coeff[scan[j]] = rnd(thresh) * (1 - 2 * (i % 2));
} else {
coeff[scan[j]] = 0;
}
@@ -504,8 +498,8 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
#if CONFIG_VP9_HIGHBITDEPTH
} else {
ref_txfm(coeff, CONVERT_TO_BYTEPTR(ref16), pitch_);
ASM_REGISTER_STATE_CHECK(RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16),
pitch_));
ASM_REGISTER_STATE_CHECK(
RunInvTxfm(coeff, CONVERT_TO_BYTEPTR(dst16), pitch_));
#endif
}
@@ -517,9 +511,8 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
const int diff = dst[j] - ref[j];
#endif
const uint32_t error = diff * diff;
EXPECT_EQ(0u, error)
<< "Error: 8x8 IDCT has error " << error
<< " at index " << j;
EXPECT_EQ(0u, error) << "Error: 8x8 IDCT has error " << error
<< " at index " << j;
}
}
}
@@ -530,17 +523,16 @@ void CompareInvReference(IdctFunc ref_txfm, int thresh) {
int mask_;
};
class FwdTrans8x8DCT
: public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Dct8x8Param> {
class FwdTrans8x8DCT : public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Dct8x8Param> {
public:
virtual ~FwdTrans8x8DCT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
pitch_ = 8;
tx_type_ = GET_PARAM(2);
pitch_ = 8;
fwd_txfm_ref = fdct8x8_ref;
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
@@ -560,37 +552,26 @@ class FwdTrans8x8DCT
IdctFunc inv_txfm_;
};
TEST_P(FwdTrans8x8DCT, SignBiasCheck) {
RunSignBiasCheck();
}
TEST_P(FwdTrans8x8DCT, SignBiasCheck) { RunSignBiasCheck(); }
TEST_P(FwdTrans8x8DCT, RoundTripErrorCheck) {
RunRoundTripErrorCheck();
}
TEST_P(FwdTrans8x8DCT, RoundTripErrorCheck) { RunRoundTripErrorCheck(); }
TEST_P(FwdTrans8x8DCT, ExtremalCheck) {
RunExtremalCheck();
}
TEST_P(FwdTrans8x8DCT, ExtremalCheck) { RunExtremalCheck(); }
TEST_P(FwdTrans8x8DCT, FwdAccuracyCheck) {
RunFwdAccuracyCheck();
}
TEST_P(FwdTrans8x8DCT, FwdAccuracyCheck) { RunFwdAccuracyCheck(); }
TEST_P(FwdTrans8x8DCT, InvAccuracyCheck) {
RunInvAccuracyCheck();
}
TEST_P(FwdTrans8x8DCT, InvAccuracyCheck) { RunInvAccuracyCheck(); }
class FwdTrans8x8HT
: public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Ht8x8Param> {
class FwdTrans8x8HT : public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Ht8x8Param> {
public:
virtual ~FwdTrans8x8HT() {}
virtual void SetUp() {
fwd_txfm_ = GET_PARAM(0);
inv_txfm_ = GET_PARAM(1);
tx_type_ = GET_PARAM(2);
pitch_ = 8;
tx_type_ = GET_PARAM(2);
pitch_ = 8;
fwd_txfm_ref = fht8x8_ref;
bit_depth_ = GET_PARAM(3);
mask_ = (1 << bit_depth_) - 1;
@@ -610,21 +591,14 @@ class FwdTrans8x8HT
IhtFunc inv_txfm_;
};
TEST_P(FwdTrans8x8HT, SignBiasCheck) {
RunSignBiasCheck();
}
TEST_P(FwdTrans8x8HT, SignBiasCheck) { RunSignBiasCheck(); }
TEST_P(FwdTrans8x8HT, RoundTripErrorCheck) {
RunRoundTripErrorCheck();
}
TEST_P(FwdTrans8x8HT, RoundTripErrorCheck) { RunRoundTripErrorCheck(); }
TEST_P(FwdTrans8x8HT, ExtremalCheck) {
RunExtremalCheck();
}
TEST_P(FwdTrans8x8HT, ExtremalCheck) { RunExtremalCheck(); }
class InvTrans8x8DCT
: public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Idct8x8Param> {
class InvTrans8x8DCT : public FwdTrans8x8TestBase,
public ::testing::TestWithParam<Idct8x8Param> {
public:
virtual ~InvTrans8x8DCT() {}
@@ -664,10 +638,10 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_10, 0, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_12, 0, VPX_BITS_12)));
#else
INSTANTIATE_TEST_CASE_P(
C, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_c, &vpx_idct8x8_64_add_c, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(C, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c, 0,
VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if CONFIG_VP9_HIGHBITDEPTH
@@ -696,15 +670,17 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_c, 3, VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#if HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
NEON, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_neon, &vpx_idct8x8_64_add_neon, 0,
VPX_BITS_8)));
#endif // HAVE_NEON_ASM && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_NEON && !CONFIG_EMULATE_HARDWARE
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(NEON, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_neon,
0, VPX_BITS_8)));
#else // !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(NEON, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_neon,
&vpx_idct8x8_64_add_neon,
0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
NEON, FwdTrans8x8HT,
::testing::Values(
@@ -712,14 +688,14 @@ INSTANTIATE_TEST_CASE_P(
make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 1, VPX_BITS_8),
make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 2, VPX_BITS_8),
make_tuple(&vp9_fht8x8_c, &vp9_iht8x8_64_add_neon, 3, VPX_BITS_8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_sse2, &vpx_idct8x8_64_add_sse2, 0,
VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(SSE2, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_sse2,
&vpx_idct8x8_64_add_sse2,
0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
SSE2, FwdTrans8x8HT,
::testing::Values(
@@ -732,16 +708,16 @@ INSTANTIATE_TEST_CASE_P(
#if HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_sse2, &vpx_idct8x8_64_add_c, 0, VPX_BITS_8),
make_tuple(&vpx_highbd_fdct8x8_c,
&idct8x8_64_add_10_sse2, 12, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct8x8_sse2,
&idct8x8_64_add_10_sse2, 12, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct8x8_c,
&idct8x8_64_add_12_sse2, 12, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct8x8_sse2,
&idct8x8_64_add_12_sse2, 12, VPX_BITS_12)));
::testing::Values(make_tuple(&vpx_fdct8x8_sse2, &vpx_idct8x8_64_add_c, 0,
VPX_BITS_8),
make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_64_add_10_sse2,
12, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct8x8_sse2,
&idct8x8_64_add_10_sse2, 12, VPX_BITS_10),
make_tuple(&vpx_highbd_fdct8x8_c, &idct8x8_64_add_12_sse2,
12, VPX_BITS_12),
make_tuple(&vpx_highbd_fdct8x8_sse2,
&idct8x8_64_add_12_sse2, 12, VPX_BITS_12)));
INSTANTIATE_TEST_CASE_P(
SSE2, FwdTrans8x8HT,
@@ -756,30 +732,27 @@ INSTANTIATE_TEST_CASE_P(
INSTANTIATE_TEST_CASE_P(
SSE2, InvTrans8x8DCT,
::testing::Values(
make_tuple(&idct8x8_10_add_10_c,
&idct8x8_10_add_10_sse2, 6225, VPX_BITS_10),
make_tuple(&idct8x8_10,
&idct8x8_64_add_10_sse2, 6225, VPX_BITS_10),
make_tuple(&idct8x8_10_add_12_c,
&idct8x8_10_add_12_sse2, 6225, VPX_BITS_12),
make_tuple(&idct8x8_12,
&idct8x8_64_add_12_sse2, 6225, VPX_BITS_12)));
make_tuple(&idct8x8_12_add_10_c, &idct8x8_12_add_10_sse2, 6225,
VPX_BITS_10),
make_tuple(&idct8x8_10, &idct8x8_64_add_10_sse2, 6225, VPX_BITS_10),
make_tuple(&idct8x8_12_add_12_c, &idct8x8_12_add_12_sse2, 6225,
VPX_BITS_12),
make_tuple(&idct8x8_12, &idct8x8_64_add_12_sse2, 6225, VPX_BITS_12)));
#endif // HAVE_SSE2 && CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 && \
!CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSSE3, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_ssse3, &vpx_idct8x8_64_add_ssse3, 0,
VPX_BITS_8)));
#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_VP9_HIGHBITDEPTH && \
!CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(SSSE3, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_ssse3,
&vpx_idct8x8_64_add_ssse3,
0, VPX_BITS_8)));
#endif
#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
MSA, FwdTrans8x8DCT,
::testing::Values(
make_tuple(&vpx_fdct8x8_msa, &vpx_idct8x8_64_add_msa, 0, VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(MSA, FwdTrans8x8DCT,
::testing::Values(make_tuple(&vpx_fdct8x8_msa,
&vpx_idct8x8_64_add_msa, 0,
VPX_BITS_8)));
INSTANTIATE_TEST_CASE_P(
MSA, FwdTrans8x8HT,
::testing::Values(

View File

@@ -13,12 +13,11 @@
namespace {
class VP9FrameSizeTestsLarge
: public ::libvpx_test::EncoderTest,
public ::testing::Test {
class VP9FrameSizeTestsLarge : public ::libvpx_test::EncoderTest,
public ::testing::Test {
protected:
VP9FrameSizeTestsLarge() : EncoderTest(&::libvpx_test::kVP9),
expected_res_(VPX_CODEC_OK) {}
VP9FrameSizeTestsLarge()
: EncoderTest(&::libvpx_test::kVP9), expected_res_(VPX_CODEC_OK) {}
virtual ~VP9FrameSizeTestsLarge() {}
virtual void SetUp() {
@@ -27,7 +26,7 @@ class VP9FrameSizeTestsLarge
}
virtual bool HandleDecodeResult(const vpx_codec_err_t res_dec,
const libvpx_test::VideoSource& /*video*/,
const libvpx_test::VideoSource & /*video*/,
libvpx_test::Decoder *decoder) {
EXPECT_EQ(expected_res_, res_dec) << decoder->DecodeError();
return !::testing::Test::HasFailure();
@@ -67,13 +66,13 @@ TEST_F(VP9FrameSizeTestsLarge, ValidSizes) {
expected_res_ = VPX_CODEC_OK;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
#else
// This test produces a pretty large single frame allocation, (roughly
// 25 megabits). The encoder allocates a good number of these frames
// one for each lag in frames (for 2 pass), and then one for each possible
// reference buffer (8) - we can end up with up to 30 buffers of roughly this
// size or almost 1 gig of memory.
// In total the allocations will exceed 2GiB which may cause a failure with
// mingw + wine, use a smaller size in that case.
// This test produces a pretty large single frame allocation, (roughly
// 25 megabits). The encoder allocates a good number of these frames
// one for each lag in frames (for 2 pass), and then one for each possible
// reference buffer (8) - we can end up with up to 30 buffers of roughly this
// size or almost 1 gig of memory.
// In total the allocations will exceed 2GiB which may cause a failure with
// mingw + wine, use a smaller size in that case.
#if defined(_WIN32) && !defined(_WIN64) || defined(__OS2__)
video.SetSize(4096, 3072);
#else

View File

@@ -80,8 +80,8 @@ void reference_hadamard16x16(const int16_t *a, int a_stride, int16_t *b) {
const int16_t b3 = (a2 - a3) >> 1;
/* Store a 16 bit value. */
b[ 0] = b0 + b2;
b[ 64] = b1 + b3;
b[0] = b0 + b2;
b[64] = b1 + b3;
b[128] = b0 - b2;
b[192] = b1 - b3;
@@ -152,10 +152,10 @@ INSTANTIATE_TEST_CASE_P(SSE2, Hadamard8x8Test,
::testing::Values(&vpx_hadamard_8x8_sse2));
#endif // HAVE_SSE2
#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64
#if HAVE_SSSE3 && ARCH_X86_64
INSTANTIATE_TEST_CASE_P(SSSE3, Hadamard8x8Test,
::testing::Values(&vpx_hadamard_8x8_ssse3));
#endif // HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64
#endif // HAVE_SSSE3 && ARCH_X86_64
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(NEON, Hadamard8x8Test,

View File

@@ -21,14 +21,11 @@ namespace libvpx_test {
// so that we can do actual file encodes.
class I420VideoSource : public YUVVideoSource {
public:
I420VideoSource(const std::string &file_name,
unsigned int width, unsigned int height,
int rate_numerator, int rate_denominator,
I420VideoSource(const std::string &file_name, unsigned int width,
unsigned int height, int rate_numerator, int rate_denominator,
unsigned int start, int limit)
: YUVVideoSource(file_name, VPX_IMG_FMT_I420,
width, height,
rate_numerator, rate_denominator,
start, limit) {}
: YUVVideoSource(file_name, VPX_IMG_FMT_I420, width, height,
rate_numerator, rate_denominator, start, limit) {}
};
} // namespace libvpx_test

View File

@@ -17,29 +17,21 @@
#include "./vpx_dsp_rtcd.h"
#include "test/acm_random.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/msvc.h" // for round()
using libvpx_test::ACMRandom;
namespace {
#ifdef _MSC_VER
static int round(double x) {
if (x < 0)
return static_cast<int>(ceil(x - 0.5));
else
return static_cast<int>(floor(x + 0.5));
}
#endif
void reference_dct_1d(double input[8], double output[8]) {
const double kPi = 3.141592653589793238462643383279502884;
const double kInvSqrt2 = 0.707106781186547524400844362104;
for (int k = 0; k < 8; k++) {
output[k] = 0.0;
for (int n = 0; n < 8; n++)
output[k] += input[n]*cos(kPi*(2*n+1)*k/16.0);
if (k == 0)
output[k] = output[k]*kInvSqrt2;
for (int n = 0; n < 8; n++) {
output[k] += input[n] * cos(kPi * (2 * n + 1) * k / 16.0);
}
if (k == 0) output[k] = output[k] * kInvSqrt2;
}
}
@@ -47,24 +39,19 @@ void reference_dct_2d(int16_t input[64], double output[64]) {
// First transform columns
for (int i = 0; i < 8; ++i) {
double temp_in[8], temp_out[8];
for (int j = 0; j < 8; ++j)
temp_in[j] = input[j*8 + i];
for (int j = 0; j < 8; ++j) temp_in[j] = input[j * 8 + i];
reference_dct_1d(temp_in, temp_out);
for (int j = 0; j < 8; ++j)
output[j*8 + i] = temp_out[j];
for (int j = 0; j < 8; ++j) output[j * 8 + i] = temp_out[j];
}
// Then transform rows
for (int i = 0; i < 8; ++i) {
double temp_in[8], temp_out[8];
for (int j = 0; j < 8; ++j)
temp_in[j] = output[j + i*8];
for (int j = 0; j < 8; ++j) temp_in[j] = output[j + i * 8];
reference_dct_1d(temp_in, temp_out);
for (int j = 0; j < 8; ++j)
output[j + i*8] = temp_out[j];
for (int j = 0; j < 8; ++j) output[j + i * 8] = temp_out[j];
}
// Scale by some magic number
for (int i = 0; i < 64; ++i)
output[i] *= 2;
for (int i = 0; i < 64; ++i) output[i] *= 2;
}
TEST(VP9Idct8x8Test, AccuracyCheck) {
@@ -81,19 +68,18 @@ TEST(VP9Idct8x8Test, AccuracyCheck) {
dst[j] = rnd.Rand8();
}
// Initialize a test block with input range [-255, 255].
for (int j = 0; j < 64; ++j)
input[j] = src[j] - dst[j];
for (int j = 0; j < 64; ++j) input[j] = src[j] - dst[j];
reference_dct_2d(input, output_r);
for (int j = 0; j < 64; ++j)
coeff[j] = round(output_r[j]);
for (int j = 0; j < 64; ++j) {
coeff[j] = static_cast<tran_low_t>(round(output_r[j]));
}
vpx_idct8x8_64_add_c(coeff, dst, 8);
for (int j = 0; j < 64; ++j) {
const int diff = dst[j] - src[j];
const int error = diff * diff;
EXPECT_GE(1, error)
<< "Error: 8x8 FDCT/IDCT has error " << error
<< " at index " << j;
EXPECT_GE(1, error) << "Error: 8x8 FDCT/IDCT has error " << error
<< " at index " << j;
}
}
}

View File

@@ -43,11 +43,12 @@ class IDCTTest : public ::testing::TestWithParam<IdctFunc> {
TEST_P(IDCTTest, TestGuardBlocks) {
int i;
for (i = 0; i < 256; i++)
for (i = 0; i < 256; i++) {
if ((i & 0xF) < 4 && i < 64)
EXPECT_EQ(0, output[i]) << i;
else
EXPECT_EQ(255, output[i]);
}
}
TEST_P(IDCTTest, TestAllZeros) {
@@ -55,11 +56,12 @@ TEST_P(IDCTTest, TestAllZeros) {
ASM_REGISTER_STATE_CHECK(UUT(input, output, 16, output, 16));
for (i = 0; i < 256; i++)
for (i = 0; i < 256; i++) {
if ((i & 0xF) < 4 && i < 64)
EXPECT_EQ(0, output[i]) << "i==" << i;
else
EXPECT_EQ(255, output[i]) << "i==" << i;
}
}
TEST_P(IDCTTest, TestAllOnes) {
@@ -68,11 +70,12 @@ TEST_P(IDCTTest, TestAllOnes) {
input[0] = 4;
ASM_REGISTER_STATE_CHECK(UUT(input, output, 16, output, 16));
for (i = 0; i < 256; i++)
for (i = 0; i < 256; i++) {
if ((i & 0xF) < 4 && i < 64)
EXPECT_EQ(1, output[i]) << "i==" << i;
else
EXPECT_EQ(255, output[i]) << "i==" << i;
}
}
TEST_P(IDCTTest, TestAddOne) {
@@ -82,11 +85,12 @@ TEST_P(IDCTTest, TestAddOne) {
input[0] = 4;
ASM_REGISTER_STATE_CHECK(UUT(input, predict, 16, output, 16));
for (i = 0; i < 256; i++)
for (i = 0; i < 256; i++) {
if ((i & 0xF) < 4 && i < 64)
EXPECT_EQ(i + 1, output[i]) << "i==" << i;
else
EXPECT_EQ(255, output[i]) << "i==" << i;
}
}
TEST_P(IDCTTest, TestWithData) {
@@ -96,7 +100,7 @@ TEST_P(IDCTTest, TestWithData) {
ASM_REGISTER_STATE_CHECK(UUT(input, output, 16, output, 16));
for (i = 0; i < 256; i++)
for (i = 0; i < 256; i++) {
if ((i & 0xF) > 3 || i > 63)
EXPECT_EQ(255, output[i]) << "i==" << i;
else if (i == 0)
@@ -107,9 +111,14 @@ TEST_P(IDCTTest, TestWithData) {
EXPECT_EQ(3, output[i]) << "i==" << i;
else
EXPECT_EQ(0, output[i]) << "i==" << i;
}
}
INSTANTIATE_TEST_CASE_P(C, IDCTTest, ::testing::Values(vp8_short_idct4x4llm_c));
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(NEON, IDCTTest,
::testing::Values(vp8_short_idct4x4llm_neon));
#endif
#if HAVE_MMX
INSTANTIATE_TEST_CASE_P(MMX, IDCTTest,
::testing::Values(vp8_short_idct4x4llm_mmx));

View File

@@ -34,21 +34,19 @@ std::ostream &operator<<(std::ostream &os, const DecodeParam &dp) {
return os << "threads: " << dp.threads << " file: " << dp.filename;
}
class InvalidFileTest
: public ::libvpx_test::DecoderTest,
public ::libvpx_test::CodecTestWithParam<DecodeParam> {
class InvalidFileTest : public ::libvpx_test::DecoderTest,
public ::libvpx_test::CodecTestWithParam<DecodeParam> {
protected:
InvalidFileTest() : DecoderTest(GET_PARAM(0)), res_file_(NULL) {}
virtual ~InvalidFileTest() {
if (res_file_ != NULL)
fclose(res_file_);
if (res_file_ != NULL) fclose(res_file_);
}
void OpenResFile(const std::string &res_file_name_) {
res_file_ = libvpx_test::OpenTestDataFile(res_file_name_);
ASSERT_TRUE(res_file_ != NULL) << "Result file open failed. Filename: "
<< res_file_name_;
<< res_file_name_;
}
virtual bool HandleDecodeResult(
@@ -72,8 +70,9 @@ class InvalidFileTest
EXPECT_TRUE(res_dec == expected_res_dec ||
res_dec == VPX_CODEC_CORRUPT_FRAME)
<< "Results don't match: frame number = " << video.frame_number()
<< ". (" << decoder->DecodeError() << "). Expected: "
<< expected_res_dec << " or " << VPX_CODEC_CORRUPT_FRAME;
<< ". (" << decoder->DecodeError()
<< "). Expected: " << expected_res_dec << " or "
<< VPX_CODEC_CORRUPT_FRAME;
} else {
EXPECT_EQ(expected_res_dec, res_dec)
<< "Results don't match: frame number = " << video.frame_number()
@@ -85,23 +84,24 @@ class InvalidFileTest
void RunTest() {
const DecodeParam input = GET_PARAM(1);
libvpx_test::CompressedVideoSource *video = NULL;
vpx_codec_dec_cfg_t cfg = vpx_codec_dec_cfg_t();
cfg.threads = input.threads;
const std::string filename = input.filename;
// Open compressed video file.
testing::internal::scoped_ptr<libvpx_test::CompressedVideoSource> video;
if (filename.substr(filename.length() - 3, 3) == "ivf") {
video = new libvpx_test::IVFVideoSource(filename);
video.reset(new libvpx_test::IVFVideoSource(filename));
} else if (filename.substr(filename.length() - 4, 4) == "webm") {
#if CONFIG_WEBM_IO
video = new libvpx_test::WebMVideoSource(filename);
video.reset(new libvpx_test::WebMVideoSource(filename));
#else
fprintf(stderr, "WebM IO is disabled, skipping test vector %s\n",
filename.c_str());
return;
#endif
}
ASSERT_TRUE(video.get() != NULL);
video->Init();
// Construct result file name. The file holds a list of expected integer
@@ -111,37 +111,45 @@ class InvalidFileTest
OpenResFile(res_filename);
// Decode frame, and check the md5 matching.
ASSERT_NO_FATAL_FAILURE(RunLoop(video, cfg));
delete video;
ASSERT_NO_FATAL_FAILURE(RunLoop(video.get(), cfg));
}
private:
FILE *res_file_;
};
TEST_P(InvalidFileTest, ReturnCode) {
RunTest();
}
TEST_P(InvalidFileTest, ReturnCode) { RunTest(); }
#if CONFIG_VP9_DECODER
const DecodeParam kVP9InvalidFileTests[] = {
{1, "invalid-vp90-02-v2.webm"},
{ 1, "invalid-vp90-02-v2.webm" },
#if CONFIG_VP9_HIGHBITDEPTH
{1, "invalid-vp90-2-00-quantizer-00.webm.ivf.s5861_r01-05_b6-.v2.ivf"},
{ 1, "invalid-vp90-2-00-quantizer-00.webm.ivf.s5861_r01-05_b6-.v2.ivf" },
#endif
{1, "invalid-vp90-03-v3.webm"},
{1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-.ivf"},
{1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-z.ivf"},
{1, "invalid-vp90-2-12-droppable_1.ivf.s3676_r01-05_b6-.ivf"},
{1, "invalid-vp90-2-05-resize.ivf.s59293_r01-05_b6-.ivf"},
{1, "invalid-vp90-2-09-subpixel-00.ivf.s20492_r01-05_b6-.v2.ivf"},
{1, "invalid-vp91-2-mixedrefcsp-444to420.ivf"},
{1, "invalid-vp90-2-12-droppable_1.ivf.s73804_r01-05_b6-.ivf"},
{1, "invalid-vp90-2-03-size-224x196.webm.ivf.s44156_r01-05_b6-.ivf"},
{1, "invalid-vp90-2-03-size-202x210.webm.ivf.s113306_r01-05_b6-.ivf"},
{ 1, "invalid-vp90-03-v3.webm" },
{ 1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-.ivf" },
{ 1, "invalid-vp90-2-00-quantizer-11.webm.ivf.s52984_r01-05_b6-z.ivf" },
// This file will cause a large allocation which is expected to fail in 32-bit
// environments. Test x86 for coverage purposes as the allocation failure will
// be in platform agnostic code.
#if ARCH_X86
{ 1, "invalid-vp90-2-00-quantizer-63.ivf.kf_65527x61446.ivf" },
#endif
{ 1, "invalid-vp90-2-12-droppable_1.ivf.s3676_r01-05_b6-.ivf" },
{ 1, "invalid-vp90-2-05-resize.ivf.s59293_r01-05_b6-.ivf" },
{ 1, "invalid-vp90-2-09-subpixel-00.ivf.s20492_r01-05_b6-.v2.ivf" },
{ 1, "invalid-vp91-2-mixedrefcsp-444to420.ivf" },
{ 1, "invalid-vp90-2-12-droppable_1.ivf.s73804_r01-05_b6-.ivf" },
{ 1, "invalid-vp90-2-03-size-224x196.webm.ivf.s44156_r01-05_b6-.ivf" },
{ 1, "invalid-vp90-2-03-size-202x210.webm.ivf.s113306_r01-05_b6-.ivf" },
{ 1,
"invalid-vp90-2-10-show-existing-frame.webm.ivf.s180315_r01-05_b6-.ivf" },
{ 1, "invalid-crbug-667044.webm" },
};
VP9_INSTANTIATE_TEST_CASE(InvalidFileTest,
::testing::ValuesIn(kVP9InvalidFileTests));
#endif // CONFIG_VP9_DECODER
// This class will include test vectors that are expected to fail
// peek. However they are still expected to have no fatal failures.
@@ -149,34 +157,46 @@ class InvalidFileInvalidPeekTest : public InvalidFileTest {
protected:
InvalidFileInvalidPeekTest() : InvalidFileTest() {}
virtual void HandlePeekResult(libvpx_test::Decoder *const /*decoder*/,
libvpx_test::CompressedVideoSource* /*video*/,
libvpx_test::CompressedVideoSource * /*video*/,
const vpx_codec_err_t /*res_peek*/) {}
};
TEST_P(InvalidFileInvalidPeekTest, ReturnCode) {
RunTest();
}
TEST_P(InvalidFileInvalidPeekTest, ReturnCode) { RunTest(); }
#if CONFIG_VP8_DECODER
const DecodeParam kVP8InvalidFileTests[] = {
{ 1, "invalid-vp80-00-comprehensive-018.ivf.2kf_0x6.ivf" },
};
VP8_INSTANTIATE_TEST_CASE(InvalidFileInvalidPeekTest,
::testing::ValuesIn(kVP8InvalidFileTests));
#endif // CONFIG_VP8_DECODER
#if CONFIG_VP9_DECODER
const DecodeParam kVP9InvalidFileInvalidPeekTests[] = {
{1, "invalid-vp90-01-v3.webm"},
{ 1, "invalid-vp90-01-v3.webm" },
};
VP9_INSTANTIATE_TEST_CASE(InvalidFileInvalidPeekTest,
::testing::ValuesIn(kVP9InvalidFileInvalidPeekTests));
const DecodeParam kMultiThreadedVP9InvalidFileTests[] = {
{4, "invalid-vp90-2-08-tile_1x4_frame_parallel_all_key.webm"},
{4, "invalid-"
"vp90-2-08-tile_1x2_frame_parallel.webm.ivf.s47039_r01-05_b6-.ivf"},
{4, "invalid-vp90-2-08-tile_1x8_frame_parallel.webm.ivf.s288_r01-05_b6-.ivf"},
{2, "invalid-vp90-2-09-aq2.webm.ivf.s3984_r01-05_b6-.v2.ivf"},
{4, "invalid-vp90-2-09-subpixel-00.ivf.s19552_r01-05_b6-.v2.ivf"},
{ 4, "invalid-vp90-2-08-tile_1x4_frame_parallel_all_key.webm" },
{ 4,
"invalid-"
"vp90-2-08-tile_1x2_frame_parallel.webm.ivf.s47039_r01-05_b6-.ivf" },
{ 4,
"invalid-vp90-2-08-tile_1x8_frame_parallel.webm.ivf.s288_r01-05_b6-.ivf" },
{ 2, "invalid-vp90-2-09-aq2.webm.ivf.s3984_r01-05_b6-.v2.ivf" },
{ 4, "invalid-vp90-2-09-subpixel-00.ivf.s19552_r01-05_b6-.v2.ivf" },
{ 2, "invalid-crbug-629481.webm" },
};
INSTANTIATE_TEST_CASE_P(
VP9MultiThreaded, InvalidFileTest,
::testing::Combine(
::testing::Values(
static_cast<const libvpx_test::CodecFactory*>(&libvpx_test::kVP9)),
static_cast<const libvpx_test::CodecFactory *>(&libvpx_test::kVP9)),
::testing::ValuesIn(kMultiThreadedVP9InvalidFileTests)));
#endif // CONFIG_VP9_DECODER
} // namespace

View File

@@ -29,19 +29,13 @@ static unsigned int MemGetLe32(const uint8_t *mem) {
class IVFVideoSource : public CompressedVideoSource {
public:
explicit IVFVideoSource(const std::string &file_name)
: file_name_(file_name),
input_file_(NULL),
compressed_frame_buf_(NULL),
frame_sz_(0),
frame_(0),
end_of_file_(false) {
}
: file_name_(file_name), input_file_(NULL), compressed_frame_buf_(NULL),
frame_sz_(0), frame_(0), end_of_file_(false) {}
virtual ~IVFVideoSource() {
delete[] compressed_frame_buf_;
if (input_file_)
fclose(input_file_);
if (input_file_) fclose(input_file_);
}
virtual void Init() {
@@ -54,15 +48,16 @@ class IVFVideoSource : public CompressedVideoSource {
virtual void Begin() {
input_file_ = OpenTestDataFile(file_name_);
ASSERT_TRUE(input_file_ != NULL) << "Input file open failed. Filename: "
<< file_name_;
<< file_name_;
// Read file header
uint8_t file_hdr[kIvfFileHdrSize];
ASSERT_EQ(kIvfFileHdrSize, fread(file_hdr, 1, kIvfFileHdrSize, input_file_))
<< "File header read failed.";
// Check file header
ASSERT_TRUE(file_hdr[0] == 'D' && file_hdr[1] == 'K' && file_hdr[2] == 'I'
&& file_hdr[3] == 'F') << "Input is not an IVF file.";
ASSERT_TRUE(file_hdr[0] == 'D' && file_hdr[1] == 'K' &&
file_hdr[2] == 'I' && file_hdr[3] == 'F')
<< "Input is not an IVF file.";
FillFrame();
}
@@ -76,8 +71,8 @@ class IVFVideoSource : public CompressedVideoSource {
ASSERT_TRUE(input_file_ != NULL);
uint8_t frame_hdr[kIvfFrameHdrSize];
// Check frame header and read a frame from input_file.
if (fread(frame_hdr, 1, kIvfFrameHdrSize, input_file_)
!= kIvfFrameHdrSize) {
if (fread(frame_hdr, 1, kIvfFrameHdrSize, input_file_) !=
kIvfFrameHdrSize) {
end_of_file_ = true;
} else {
end_of_file_ = false;

View File

@@ -17,8 +17,9 @@
namespace {
class KeyframeTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class KeyframeTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
KeyframeTest() : EncoderTest(GET_PARAM(0)) {}
virtual ~KeyframeTest() {}
@@ -34,10 +35,12 @@ class KeyframeTest : public ::libvpx_test::EncoderTest,
virtual void PreEncodeFrameHook(::libvpx_test::VideoSource *video,
::libvpx_test::Encoder *encoder) {
if (kf_do_force_kf_)
if (kf_do_force_kf_) {
frame_flags_ = (video->frame() % 3) ? 0 : VPX_EFLAG_FORCE_KF;
if (set_cpu_used_ && video->frame() == 1)
}
if (set_cpu_used_ && video->frame() == 1) {
encoder->Control(VP8E_SET_CPUUSED, set_cpu_used_);
}
}
virtual void FramePktHook(const vpx_codec_cx_pkt_t *pkt) {
@@ -65,8 +68,7 @@ TEST_P(KeyframeTest, TestRandomVideoSource) {
// In realtime mode - auto placed keyframes are exceedingly rare, don't
// bother with this check if(GetParam() > 0)
if (GET_PARAM(1) > 0)
EXPECT_GT(kf_count_, 1);
if (GET_PARAM(1) > 0) EXPECT_GT(kf_count_, 1);
}
TEST_P(KeyframeTest, TestDisableKeyframes) {
@@ -114,8 +116,7 @@ TEST_P(KeyframeTest, TestAutoKeyframe) {
// may not produce a keyframe like we expect. This is necessary when running
// on very slow environments (like Valgrind). The step -11 was determined
// experimentally as the fastest mode that still throws the keyframe.
if (deadline_ == VPX_DL_REALTIME)
set_cpu_used_ = -11;
if (deadline_ == VPX_DL_REALTIME) set_cpu_used_ = -11;
// This clip has a cut scene every 30 frames -> Frame 0, 30, 60, 90, 120.
// I check only the first 40 frames to make sure there's a keyframe at frame
@@ -135,7 +136,7 @@ TEST_P(KeyframeTest, TestAutoKeyframe) {
iter != kf_pts_list_.end(); ++iter) {
if (deadline_ == VPX_DL_REALTIME && *iter > 0)
EXPECT_EQ(0, (*iter - 1) % 30) << "Unexpected keyframe at frame "
<< *iter;
<< *iter;
else
EXPECT_EQ(0, *iter % 30) << "Unexpected keyframe at frame " << *iter;
}

View File

@@ -19,12 +19,9 @@ class LevelTest
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
protected:
LevelTest()
: EncoderTest(GET_PARAM(0)),
encoding_mode_(GET_PARAM(1)),
cpu_used_(GET_PARAM(2)),
min_gf_internal_(24),
target_level_(0),
level_(0) {}
: EncoderTest(GET_PARAM(0)), encoding_mode_(GET_PARAM(1)),
cpu_used_(GET_PARAM(2)), min_gf_internal_(24), target_level_(0),
level_(0) {}
virtual ~LevelTest() {}
virtual void SetUp() {
@@ -69,6 +66,36 @@ class LevelTest
int level_;
};
TEST_P(LevelTest, TestTargetLevel11) {
ASSERT_NE(encoding_mode_, ::libvpx_test::kRealTime);
::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
90);
target_level_ = 11;
cfg_.rc_target_bitrate = 150;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_EQ(target_level_, level_);
}
TEST_P(LevelTest, TestTargetLevel20) {
ASSERT_NE(encoding_mode_, ::libvpx_test::kRealTime);
::libvpx_test::I420VideoSource video("hantro_collage_w352h288.yuv", 352, 288,
30, 1, 0, 90);
target_level_ = 20;
cfg_.rc_target_bitrate = 1200;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_EQ(target_level_, level_);
}
TEST_P(LevelTest, TestTargetLevel31) {
ASSERT_NE(encoding_mode_, ::libvpx_test::kRealTime);
::libvpx_test::I420VideoSource video("niklas_1280_720_30.y4m", 1280, 720, 30,
1, 0, 60);
target_level_ = 31;
cfg_.rc_target_bitrate = 8000;
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
ASSERT_EQ(target_level_, level_);
}
// Test for keeping level stats only
TEST_P(LevelTest, TestTargetLevel0) {
::libvpx_test::I420VideoSource video("hantro_odd.yuv", 208, 144, 30, 1, 0,
@@ -97,6 +124,7 @@ TEST_P(LevelTest, TestTargetLevelApi) {
vpx_codec_ctx_t enc;
vpx_codec_enc_cfg_t cfg;
EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_config_default(codec, &cfg, 0));
cfg.rc_target_bitrate = 100;
EXPECT_EQ(VPX_CODEC_OK, vpx_codec_enc_init(&enc, codec, &cfg, 0));
for (int level = 0; level <= 256; ++level) {
if (level == 10 || level == 11 || level == 20 || level == 21 ||

View File

@@ -1,672 +0,0 @@
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <cmath>
#include <cstdlib>
#include <string>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vp9/common/vp9_entropy.h"
#include "vp9/common/vp9_loopfilter.h"
#include "vpx/vpx_integer.h"
using libvpx_test::ACMRandom;
namespace {
// Horizontally and Vertically need 32x32: 8 Coeffs preceeding filtered section
// 16 Coefs within filtered section
// 8 Coeffs following filtered section
const int kNumCoeffs = 1024;
const int number_of_iterations = 10000;
#if CONFIG_VP9_HIGHBITDEPTH
typedef void (*loop_op_t)(uint16_t *s, int p, const uint8_t *blimit,
const uint8_t *limit, const uint8_t *thresh,
int bd);
typedef void (*dual_loop_op_t)(uint16_t *s, int p, const uint8_t *blimit0,
const uint8_t *limit0, const uint8_t *thresh0,
const uint8_t *blimit1, const uint8_t *limit1,
const uint8_t *thresh1, int bd);
#else
typedef void (*loop_op_t)(uint8_t *s, int p, const uint8_t *blimit,
const uint8_t *limit, const uint8_t *thresh);
typedef void (*dual_loop_op_t)(uint8_t *s, int p, const uint8_t *blimit0,
const uint8_t *limit0, const uint8_t *thresh0,
const uint8_t *blimit1, const uint8_t *limit1,
const uint8_t *thresh1);
#endif // CONFIG_VP9_HIGHBITDEPTH
typedef std::tr1::tuple<loop_op_t, loop_op_t, int> loop8_param_t;
typedef std::tr1::tuple<dual_loop_op_t, dual_loop_op_t, int> dualloop8_param_t;
class Loop8Test6Param : public ::testing::TestWithParam<loop8_param_t> {
public:
virtual ~Loop8Test6Param() {}
virtual void SetUp() {
loopfilter_op_ = GET_PARAM(0);
ref_loopfilter_op_ = GET_PARAM(1);
bit_depth_ = GET_PARAM(2);
mask_ = (1 << bit_depth_) - 1;
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
int bit_depth_;
int mask_;
loop_op_t loopfilter_op_;
loop_op_t ref_loopfilter_op_;
};
class Loop8Test9Param : public ::testing::TestWithParam<dualloop8_param_t> {
public:
virtual ~Loop8Test9Param() {}
virtual void SetUp() {
loopfilter_op_ = GET_PARAM(0);
ref_loopfilter_op_ = GET_PARAM(1);
bit_depth_ = GET_PARAM(2);
mask_ = (1 << bit_depth_) - 1;
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
int bit_depth_;
int mask_;
dual_loop_op_t loopfilter_op_;
dual_loop_op_t ref_loopfilter_op_;
};
TEST_P(Loop8Test6Param, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
#if CONFIG_VP9_HIGHBITDEPTH
int32_t bd = bit_depth_;
DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
#else
DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
#endif // CONFIG_VP9_HIGHBITDEPTH
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
int32_t p = kNumCoeffs/32;
uint16_t tmp_s[kNumCoeffs];
int j = 0;
while (j < kNumCoeffs) {
uint8_t val = rnd.Rand8();
if (val & 0x80) { // 50% chance to choose a new value.
tmp_s[j] = rnd.Rand16();
j++;
} else { // 50% chance to repeat previous value in row X times
int k = 0;
while (k++ < ((val & 0x1f) + 1) && j < kNumCoeffs) {
if (j < 1) {
tmp_s[j] = rnd.Rand16();
} else if (val & 0x20) { // Increment by an value within the limit
tmp_s[j] = (tmp_s[j - 1] + (*limit - 1));
} else { // Decrement by an value within the limit
tmp_s[j] = (tmp_s[j - 1] - (*limit - 1));
}
j++;
}
}
}
for (j = 0; j < kNumCoeffs; j++) {
if (i % 2) {
s[j] = tmp_s[j] & mask_;
} else {
s[j] = tmp_s[p * (j % p) + j / p] & mask_;
}
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bd);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bd));
#else
ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test6Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test6Param, ValueCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
#if CONFIG_VP9_HIGHBITDEPTH
const int32_t bd = bit_depth_;
DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
#else
DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
#endif // CONFIG_VP9_HIGHBITDEPTH
int err_count_total = 0;
int first_failure = -1;
// NOTE: The code in vp9_loopfilter.c:update_sharpness computes mblim as a
// function of sharpness_lvl and the loopfilter lvl as:
// block_inside_limit = lvl >> ((sharpness_lvl > 0) + (sharpness_lvl > 4));
// ...
// memset(lfi->lfthr[lvl].mblim, (2 * (lvl + 2) + block_inside_limit),
// SIMD_WIDTH);
// This means that the largest value for mblim will occur when sharpness_lvl
// is equal to 0, and lvl is equal to its greatest value (MAX_LOOP_FILTER).
// In this case block_inside_limit will be equal to MAX_LOOP_FILTER and
// therefore mblim will be equal to (2 * (lvl + 2) + block_inside_limit) =
// 2 * (MAX_LOOP_FILTER + 2) + MAX_LOOP_FILTER = 3 * MAX_LOOP_FILTER + 4
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
int32_t p = kNumCoeffs / 32;
for (int j = 0; j < kNumCoeffs; ++j) {
s[j] = rnd.Rand16() & mask_;
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bd);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bd));
#else
ref_loopfilter_op_(ref_s+8+p*8, p, blimit, limit, thresh);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test6Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test9Param, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
#if CONFIG_VP9_HIGHBITDEPTH
const int32_t bd = bit_depth_;
DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
#else
DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
#endif // CONFIG_VP9_HIGHBITDEPTH
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
int32_t p = kNumCoeffs / 32;
uint16_t tmp_s[kNumCoeffs];
int j = 0;
const uint8_t limit = *limit0 < *limit1 ? *limit0 : *limit1;
while (j < kNumCoeffs) {
uint8_t val = rnd.Rand8();
if (val & 0x80) { // 50% chance to choose a new value.
tmp_s[j] = rnd.Rand16();
j++;
} else { // 50% chance to repeat previous value in row X times.
int k = 0;
while (k++ < ((val & 0x1f) + 1) && j < kNumCoeffs) {
if (j < 1) {
tmp_s[j] = rnd.Rand16();
} else if (val & 0x20) { // Increment by a value within the limit.
tmp_s[j] = (tmp_s[j - 1] + (limit - 1));
} else { // Decrement by an value within the limit.
tmp_s[j] = (tmp_s[j - 1] - (limit - 1));
}
j++;
}
}
}
for (j = 0; j < kNumCoeffs; j++) {
if (i % 2) {
s[j] = tmp_s[j] & mask_;
} else {
s[j] = tmp_s[p * (j % p) + j / p] & mask_;
}
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1, bd);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1, bd));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test9Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test9Param, ValueCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
#if CONFIG_VP9_HIGHBITDEPTH
DECLARE_ALIGNED(16, uint16_t, s[kNumCoeffs]);
DECLARE_ALIGNED(16, uint16_t, ref_s[kNumCoeffs]);
#else
DECLARE_ALIGNED(8, uint8_t, s[kNumCoeffs]);
DECLARE_ALIGNED(8, uint8_t, ref_s[kNumCoeffs]);
#endif // CONFIG_VP9_HIGHBITDEPTH
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh0[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t, blimit1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t, limit1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t, thresh1[16]) = {
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp
};
int32_t p = kNumCoeffs / 32; // TODO(pdlf) can we have non-square here?
for (int j = 0; j < kNumCoeffs; ++j) {
s[j] = rnd.Rand16() & mask_;
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
const int32_t bd = bit_depth_;
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1, bd);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0,
thresh0, blimit1, limit1, thresh1, bd));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0, thresh0,
blimit1, limit1, thresh1));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test9Param, C output doesn't match SSE2"
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
using std::tr1::make_tuple;
#if HAVE_SSE2
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
&vpx_highbd_lpf_horizontal_edge_8_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
&vpx_highbd_lpf_horizontal_edge_16_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
&vpx_highbd_lpf_horizontal_edge_8_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
&vpx_highbd_lpf_horizontal_edge_16_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_edge_8_sse2,
&vpx_highbd_lpf_horizontal_edge_8_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_edge_16_sse2,
&vpx_highbd_lpf_horizontal_edge_16_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 12)));
#else
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_sse2,
&vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_sse2,
&vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_8_sse2,
&vpx_lpf_horizontal_edge_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_16_sse2,
&vpx_lpf_horizontal_edge_16_c, 8),
make_tuple(&vpx_lpf_vertical_4_sse2,
&vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_sse2,
&vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_sse2,
&vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_sse2,
&vpx_lpf_vertical_16_dual_c, 8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif
#if HAVE_AVX2 && (!CONFIG_VP9_HIGHBITDEPTH)
INSTANTIATE_TEST_CASE_P(
AVX2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_edge_8_avx2,
&vpx_lpf_horizontal_edge_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_16_avx2,
&vpx_lpf_horizontal_edge_16_c, 8)));
#endif
#if HAVE_SSE2
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test9Param,
::testing::Values(
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 12)));
#else
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test9Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_dual_sse2,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_sse2,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_sse2,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_sse2,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif
#if HAVE_NEON
#if CONFIG_VP9_HIGHBITDEPTH
// No neon high bitdepth functions.
#else
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test6Param,
::testing::Values(
#if HAVE_NEON_ASM
// Using #if inside the macro is unsupported on MSVS but the tests are not
// currently built for MSVS with ARM and NEON.
make_tuple(&vpx_lpf_horizontal_edge_8_neon,
&vpx_lpf_horizontal_edge_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_16_neon,
&vpx_lpf_horizontal_edge_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_neon,
&vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_neon,
&vpx_lpf_vertical_16_dual_c, 8),
#endif // HAVE_NEON_ASM
make_tuple(&vpx_lpf_horizontal_8_neon,
&vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_vertical_8_neon,
&vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_horizontal_4_neon,
&vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_vertical_4_neon,
&vpx_lpf_vertical_4_c, 8)));
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test9Param,
::testing::Values(
#if HAVE_NEON_ASM
make_tuple(&vpx_lpf_horizontal_8_dual_neon,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_neon,
&vpx_lpf_vertical_8_dual_c, 8),
#endif // HAVE_NEON_ASM
make_tuple(&vpx_lpf_horizontal_4_dual_neon,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_neon,
&vpx_lpf_vertical_4_dual_c, 8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif // HAVE_NEON
#if HAVE_DSPR2 && !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
DSPR2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_dspr2,
&vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dspr2,
&vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_8,
&vpx_lpf_horizontal_edge_8, 8),
make_tuple(&vpx_lpf_horizontal_edge_16,
&vpx_lpf_horizontal_edge_16, 8),
make_tuple(&vpx_lpf_vertical_4_dspr2,
&vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_dspr2,
&vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_dspr2,
&vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_dspr2,
&vpx_lpf_vertical_16_dual_c, 8)));
INSTANTIATE_TEST_CASE_P(
DSPR2, Loop8Test9Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_dual_dspr2,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_dspr2,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_dspr2,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_dspr2,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // HAVE_DSPR2 && !CONFIG_VP9_HIGHBITDEPTH
#if HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
INSTANTIATE_TEST_CASE_P(
MSA, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_msa,
&vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_msa,
&vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_8_msa,
&vpx_lpf_horizontal_edge_8_c, 8),
make_tuple(&vpx_lpf_horizontal_edge_16_msa,
&vpx_lpf_horizontal_edge_16_c, 8),
make_tuple(&vpx_lpf_vertical_4_msa,
&vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_msa,
&vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_msa,
&vpx_lpf_vertical_16_c, 8)));
INSTANTIATE_TEST_CASE_P(
MSA, Loop8Test9Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_dual_msa,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_msa,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_msa,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_msa,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
} // namespace

672
test/lpf_test.cc Normal file
View File

@@ -0,0 +1,672 @@
/*
* Copyright (c) 2014 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <cmath>
#include <cstdlib>
#include <string>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vp9/common/vp9_entropy.h"
#include "vp9/common/vp9_loopfilter.h"
#include "vpx/vpx_integer.h"
using libvpx_test::ACMRandom;
namespace {
// Horizontally and Vertically need 32x32: 8 Coeffs preceeding filtered section
// 16 Coefs within filtered section
// 8 Coeffs following filtered section
const int kNumCoeffs = 1024;
const int number_of_iterations = 10000;
#if CONFIG_VP9_HIGHBITDEPTH
typedef uint16_t Pixel;
#define PIXEL_WIDTH 16
typedef void (*loop_op_t)(Pixel *s, int p, const uint8_t *blimit,
const uint8_t *limit, const uint8_t *thresh, int bd);
typedef void (*dual_loop_op_t)(Pixel *s, int p, const uint8_t *blimit0,
const uint8_t *limit0, const uint8_t *thresh0,
const uint8_t *blimit1, const uint8_t *limit1,
const uint8_t *thresh1, int bd);
#else
typedef uint8_t Pixel;
#define PIXEL_WIDTH 8
typedef void (*loop_op_t)(Pixel *s, int p, const uint8_t *blimit,
const uint8_t *limit, const uint8_t *thresh);
typedef void (*dual_loop_op_t)(Pixel *s, int p, const uint8_t *blimit0,
const uint8_t *limit0, const uint8_t *thresh0,
const uint8_t *blimit1, const uint8_t *limit1,
const uint8_t *thresh1);
#endif // CONFIG_VP9_HIGHBITDEPTH
typedef std::tr1::tuple<loop_op_t, loop_op_t, int> loop8_param_t;
typedef std::tr1::tuple<dual_loop_op_t, dual_loop_op_t, int> dualloop8_param_t;
void InitInput(Pixel *s, Pixel *ref_s, ACMRandom *rnd, const uint8_t limit,
const int mask, const int32_t p, const int i) {
uint16_t tmp_s[kNumCoeffs];
for (int j = 0; j < kNumCoeffs;) {
const uint8_t val = rnd->Rand8();
if (val & 0x80) { // 50% chance to choose a new value.
tmp_s[j] = rnd->Rand16();
j++;
} else { // 50% chance to repeat previous value in row X times.
int k = 0;
while (k++ < ((val & 0x1f) + 1) && j < kNumCoeffs) {
if (j < 1) {
tmp_s[j] = rnd->Rand16();
} else if (val & 0x20) { // Increment by a value within the limit.
tmp_s[j] = tmp_s[j - 1] + (limit - 1);
} else { // Decrement by a value within the limit.
tmp_s[j] = tmp_s[j - 1] - (limit - 1);
}
j++;
}
}
}
for (int j = 0; j < kNumCoeffs;) {
const uint8_t val = rnd->Rand8();
if (val & 0x80) {
j++;
} else { // 50% chance to repeat previous value in column X times.
int k = 0;
while (k++ < ((val & 0x1f) + 1) && j < kNumCoeffs) {
if (j < 1) {
tmp_s[j] = rnd->Rand16();
} else if (val & 0x20) { // Increment by a value within the limit.
tmp_s[(j % 32) * 32 + j / 32] =
tmp_s[((j - 1) % 32) * 32 + (j - 1) / 32] + (limit - 1);
} else { // Decrement by a value within the limit.
tmp_s[(j % 32) * 32 + j / 32] =
tmp_s[((j - 1) % 32) * 32 + (j - 1) / 32] - (limit - 1);
}
j++;
}
}
}
for (int j = 0; j < kNumCoeffs; j++) {
if (i % 2) {
s[j] = tmp_s[j] & mask;
} else {
s[j] = tmp_s[p * (j % p) + j / p] & mask;
}
ref_s[j] = s[j];
}
}
class Loop8Test6Param : public ::testing::TestWithParam<loop8_param_t> {
public:
virtual ~Loop8Test6Param() {}
virtual void SetUp() {
loopfilter_op_ = GET_PARAM(0);
ref_loopfilter_op_ = GET_PARAM(1);
bit_depth_ = GET_PARAM(2);
mask_ = (1 << bit_depth_) - 1;
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
int bit_depth_;
int mask_;
loop_op_t loopfilter_op_;
loop_op_t ref_loopfilter_op_;
};
class Loop8Test9Param : public ::testing::TestWithParam<dualloop8_param_t> {
public:
virtual ~Loop8Test9Param() {}
virtual void SetUp() {
loopfilter_op_ = GET_PARAM(0);
ref_loopfilter_op_ = GET_PARAM(1);
bit_depth_ = GET_PARAM(2);
mask_ = (1 << bit_depth_) - 1;
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
int bit_depth_;
int mask_;
dual_loop_op_t loopfilter_op_;
dual_loop_op_t ref_loopfilter_op_;
};
TEST_P(Loop8Test6Param, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
const int32_t p = kNumCoeffs / 32;
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, s[kNumCoeffs]);
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, ref_s[kNumCoeffs]);
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
InitInput(s, ref_s, &rnd, *limit, mask_, p, i);
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bit_depth_);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bit_depth_));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test6Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test6Param, ValueCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, s[kNumCoeffs]);
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, ref_s[kNumCoeffs]);
int err_count_total = 0;
int first_failure = -1;
// NOTE: The code in vp9_loopfilter.c:update_sharpness computes mblim as a
// function of sharpness_lvl and the loopfilter lvl as:
// block_inside_limit = lvl >> ((sharpness_lvl > 0) + (sharpness_lvl > 4));
// ...
// memset(lfi->lfthr[lvl].mblim, (2 * (lvl + 2) + block_inside_limit),
// SIMD_WIDTH);
// This means that the largest value for mblim will occur when sharpness_lvl
// is equal to 0, and lvl is equal to its greatest value (MAX_LOOP_FILTER).
// In this case block_inside_limit will be equal to MAX_LOOP_FILTER and
// therefore mblim will be equal to (2 * (lvl + 2) + block_inside_limit) =
// 2 * (MAX_LOOP_FILTER + 2) + MAX_LOOP_FILTER = 3 * MAX_LOOP_FILTER + 4
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
int32_t p = kNumCoeffs / 32;
for (int j = 0; j < kNumCoeffs; ++j) {
s[j] = rnd.Rand16() & mask_;
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh, bit_depth_);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh, bit_depth_));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit, limit, thresh);
ASM_REGISTER_STATE_CHECK(
loopfilter_op_(s + 8 + p * 8, p, blimit, limit, thresh));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test6Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test9Param, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, s[kNumCoeffs]);
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, ref_s[kNumCoeffs]);
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
int32_t p = kNumCoeffs / 32;
const uint8_t limit = *limit0 < *limit1 ? *limit0 : *limit1;
InitInput(s, ref_s, &rnd, limit, mask_, p, i);
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0, blimit1,
limit1, thresh1, bit_depth_);
ASM_REGISTER_STATE_CHECK(loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0,
thresh0, blimit1, limit1, thresh1,
bit_depth_));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0, blimit1,
limit1, thresh1);
ASM_REGISTER_STATE_CHECK(loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0,
thresh0, blimit1, limit1, thresh1));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test9Param, C output doesn't match SSE2 "
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
TEST_P(Loop8Test9Param, ValueCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
const int count_test_block = number_of_iterations;
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, s[kNumCoeffs]);
DECLARE_ALIGNED(PIXEL_WIDTH, Pixel, ref_s[kNumCoeffs]);
int err_count_total = 0;
int first_failure = -1;
for (int i = 0; i < count_test_block; ++i) {
int err_count = 0;
uint8_t tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh0[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(3 * MAX_LOOP_FILTER + 4));
DECLARE_ALIGNED(16, const uint8_t,
blimit1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = static_cast<uint8_t>(rnd(MAX_LOOP_FILTER));
DECLARE_ALIGNED(16, const uint8_t,
limit1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
tmp = rnd.Rand8();
DECLARE_ALIGNED(16, const uint8_t,
thresh1[16]) = { tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp,
tmp, tmp, tmp, tmp, tmp, tmp, tmp, tmp };
int32_t p = kNumCoeffs / 32; // TODO(pdlf) can we have non-square here?
for (int j = 0; j < kNumCoeffs; ++j) {
s[j] = rnd.Rand16() & mask_;
ref_s[j] = s[j];
}
#if CONFIG_VP9_HIGHBITDEPTH
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0, blimit1,
limit1, thresh1, bit_depth_);
ASM_REGISTER_STATE_CHECK(loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0,
thresh0, blimit1, limit1, thresh1,
bit_depth_));
#else
ref_loopfilter_op_(ref_s + 8 + p * 8, p, blimit0, limit0, thresh0, blimit1,
limit1, thresh1);
ASM_REGISTER_STATE_CHECK(loopfilter_op_(s + 8 + p * 8, p, blimit0, limit0,
thresh0, blimit1, limit1, thresh1));
#endif // CONFIG_VP9_HIGHBITDEPTH
for (int j = 0; j < kNumCoeffs; ++j) {
err_count += ref_s[j] != s[j];
}
if (err_count && !err_count_total) {
first_failure = i;
}
err_count_total += err_count;
}
EXPECT_EQ(0, err_count_total)
<< "Error: Loop8Test9Param, C output doesn't match SSE2"
"loopfilter output. "
<< "First failed at test case " << first_failure;
}
using std::tr1::make_tuple;
#if HAVE_SSE2
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test6Param,
::testing::Values(make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
&vpx_highbd_lpf_horizontal_16_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_sse2,
&vpx_highbd_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
&vpx_highbd_lpf_horizontal_16_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_sse2,
&vpx_highbd_lpf_horizontal_16_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_sse2,
&vpx_highbd_lpf_horizontal_4_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_sse2,
&vpx_highbd_lpf_vertical_4_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_sse2,
&vpx_highbd_lpf_horizontal_8_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_16_sse2,
&vpx_highbd_lpf_horizontal_16_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_sse2,
&vpx_highbd_lpf_horizontal_16_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_sse2,
&vpx_highbd_lpf_vertical_8_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_sse2,
&vpx_highbd_lpf_vertical_16_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_sse2,
&vpx_highbd_lpf_vertical_16_dual_c, 12)));
#else
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_sse2, &vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_sse2, &vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_16_sse2, &vpx_lpf_horizontal_16_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dual_sse2,
&vpx_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_sse2, &vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_sse2, &vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_sse2, &vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_sse2, &vpx_lpf_vertical_16_dual_c,
8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif
#if HAVE_AVX2 && (!CONFIG_VP9_HIGHBITDEPTH)
INSTANTIATE_TEST_CASE_P(
AVX2, Loop8Test6Param,
::testing::Values(make_tuple(&vpx_lpf_horizontal_16_avx2,
&vpx_lpf_horizontal_16_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dual_avx2,
&vpx_lpf_horizontal_16_dual_c, 8)));
#endif
#if HAVE_SSE2
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_sse2,
&vpx_highbd_lpf_horizontal_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_sse2,
&vpx_highbd_lpf_horizontal_8_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_sse2,
&vpx_highbd_lpf_vertical_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_sse2,
&vpx_highbd_lpf_vertical_8_dual_c, 12)));
#else
INSTANTIATE_TEST_CASE_P(
SSE2, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_lpf_horizontal_4_dual_sse2,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_sse2,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_sse2,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_sse2,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif
#if HAVE_NEON
#if CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test6Param,
::testing::Values(make_tuple(&vpx_highbd_lpf_horizontal_4_neon,
&vpx_highbd_lpf_horizontal_4_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_neon,
&vpx_highbd_lpf_horizontal_4_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_neon,
&vpx_highbd_lpf_horizontal_4_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_neon,
&vpx_highbd_lpf_horizontal_8_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_neon,
&vpx_highbd_lpf_horizontal_8_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_neon,
&vpx_highbd_lpf_horizontal_8_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_16_neon,
&vpx_highbd_lpf_horizontal_16_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_16_neon,
&vpx_highbd_lpf_horizontal_16_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_16_neon,
&vpx_highbd_lpf_horizontal_16_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_neon,
&vpx_highbd_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_neon,
&vpx_highbd_lpf_horizontal_16_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_16_dual_neon,
&vpx_highbd_lpf_horizontal_16_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_neon,
&vpx_highbd_lpf_vertical_4_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_neon,
&vpx_highbd_lpf_vertical_4_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_neon,
&vpx_highbd_lpf_vertical_4_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_neon,
&vpx_highbd_lpf_vertical_8_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_neon,
&vpx_highbd_lpf_vertical_8_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_neon,
&vpx_highbd_lpf_vertical_8_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_neon,
&vpx_highbd_lpf_vertical_16_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_neon,
&vpx_highbd_lpf_vertical_16_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_neon,
&vpx_highbd_lpf_vertical_16_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_neon,
&vpx_highbd_lpf_vertical_16_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_neon,
&vpx_highbd_lpf_vertical_16_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_16_dual_neon,
&vpx_highbd_lpf_vertical_16_dual_c, 12)));
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_highbd_lpf_horizontal_4_dual_neon,
&vpx_highbd_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_neon,
&vpx_highbd_lpf_horizontal_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_4_dual_neon,
&vpx_highbd_lpf_horizontal_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_neon,
&vpx_highbd_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_neon,
&vpx_highbd_lpf_horizontal_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_horizontal_8_dual_neon,
&vpx_highbd_lpf_horizontal_8_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_neon,
&vpx_highbd_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_neon,
&vpx_highbd_lpf_vertical_4_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_4_dual_neon,
&vpx_highbd_lpf_vertical_4_dual_c, 12),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_neon,
&vpx_highbd_lpf_vertical_8_dual_c, 8),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_neon,
&vpx_highbd_lpf_vertical_8_dual_c, 10),
make_tuple(&vpx_highbd_lpf_vertical_8_dual_neon,
&vpx_highbd_lpf_vertical_8_dual_c, 12)));
#else
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_16_neon, &vpx_lpf_horizontal_16_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dual_neon,
&vpx_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_lpf_vertical_16_neon, &vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_neon, &vpx_lpf_vertical_16_dual_c,
8),
make_tuple(&vpx_lpf_horizontal_8_neon, &vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_vertical_8_neon, &vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_horizontal_4_neon, &vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_vertical_4_neon, &vpx_lpf_vertical_4_c, 8)));
INSTANTIATE_TEST_CASE_P(
NEON, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_lpf_horizontal_8_dual_neon,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_neon,
&vpx_lpf_vertical_8_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_4_dual_neon,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_neon,
&vpx_lpf_vertical_4_dual_c, 8)));
#endif // CONFIG_VP9_HIGHBITDEPTH
#endif // HAVE_NEON
#if HAVE_DSPR2 && !CONFIG_VP9_HIGHBITDEPTH
INSTANTIATE_TEST_CASE_P(
DSPR2, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_dspr2, &vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dspr2, &vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dspr2, &vpx_lpf_horizontal_16_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dual_dspr2,
&vpx_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dspr2, &vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_dspr2, &vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_dspr2, &vpx_lpf_vertical_16_c, 8),
make_tuple(&vpx_lpf_vertical_16_dual_dspr2, &vpx_lpf_vertical_16_dual_c,
8)));
INSTANTIATE_TEST_CASE_P(
DSPR2, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_lpf_horizontal_4_dual_dspr2,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_dspr2,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_dspr2,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_dspr2,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // HAVE_DSPR2 && !CONFIG_VP9_HIGHBITDEPTH
#if HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
INSTANTIATE_TEST_CASE_P(
MSA, Loop8Test6Param,
::testing::Values(
make_tuple(&vpx_lpf_horizontal_4_msa, &vpx_lpf_horizontal_4_c, 8),
make_tuple(&vpx_lpf_horizontal_8_msa, &vpx_lpf_horizontal_8_c, 8),
make_tuple(&vpx_lpf_horizontal_16_msa, &vpx_lpf_horizontal_16_c, 8),
make_tuple(&vpx_lpf_horizontal_16_dual_msa,
&vpx_lpf_horizontal_16_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_msa, &vpx_lpf_vertical_4_c, 8),
make_tuple(&vpx_lpf_vertical_8_msa, &vpx_lpf_vertical_8_c, 8),
make_tuple(&vpx_lpf_vertical_16_msa, &vpx_lpf_vertical_16_c, 8)));
INSTANTIATE_TEST_CASE_P(
MSA, Loop8Test9Param,
::testing::Values(make_tuple(&vpx_lpf_horizontal_4_dual_msa,
&vpx_lpf_horizontal_4_dual_c, 8),
make_tuple(&vpx_lpf_horizontal_8_dual_msa,
&vpx_lpf_horizontal_8_dual_c, 8),
make_tuple(&vpx_lpf_vertical_4_dual_msa,
&vpx_lpf_vertical_4_dual_c, 8),
make_tuple(&vpx_lpf_vertical_8_dual_msa,
&vpx_lpf_vertical_8_dual_c, 8)));
#endif // HAVE_MSA && (!CONFIG_VP9_HIGHBITDEPTH)
} // namespace

View File

@@ -17,9 +17,7 @@
namespace libvpx_test {
class MD5 {
public:
MD5() {
MD5Init(&md5_);
}
MD5() { MD5Init(&md5_); }
void Add(const vpx_image_t *img) {
for (int plane = 0; plane < 3; ++plane) {
@@ -30,10 +28,13 @@ class MD5 {
// This works only for chroma_shift of 0 and 1.
const int bytes_per_sample =
(img->fmt & VPX_IMG_FMT_HIGHBITDEPTH) ? 2 : 1;
const int h = plane ? (img->d_h + img->y_chroma_shift) >>
img->y_chroma_shift : img->d_h;
const int w = (plane ? (img->d_w + img->x_chroma_shift) >>
img->x_chroma_shift : img->d_w) * bytes_per_sample;
const int h =
plane ? (img->d_h + img->y_chroma_shift) >> img->y_chroma_shift
: img->d_h;
const int w =
(plane ? (img->d_w + img->x_chroma_shift) >> img->x_chroma_shift
: img->d_w) *
bytes_per_sample;
for (int y = 0; y < h; ++y) {
MD5Update(&md5_, buf, w);
@@ -56,8 +57,8 @@ class MD5 {
MD5Final(tmp, &ctx_tmp);
for (int i = 0; i < 16; i++) {
res_[i * 2 + 0] = hex[tmp[i] >> 4];
res_[i * 2 + 1] = hex[tmp[i] & 0xf];
res_[i * 2 + 0] = hex[tmp[i] >> 4];
res_[i * 2 + 1] = hex[tmp[i] & 0xf];
}
res_[32] = 0;

View File

@@ -23,9 +23,8 @@ namespace {
using ::libvpx_test::ACMRandom;
typedef void (*MinMaxFunc)(const uint8_t *a, int a_stride,
const uint8_t *b, int b_stride,
int *min, int *max);
typedef void (*MinMaxFunc)(const uint8_t *a, int a_stride, const uint8_t *b,
int b_stride, int *min, int *max);
class MinMaxTest : public ::testing::TestWithParam<MinMaxFunc> {
public:
@@ -39,9 +38,8 @@ class MinMaxTest : public ::testing::TestWithParam<MinMaxFunc> {
ACMRandom rnd_;
};
void reference_minmax(const uint8_t *a, int a_stride,
const uint8_t *b, int b_stride,
int *min_ret, int *max_ret) {
void reference_minmax(const uint8_t *a, int a_stride, const uint8_t *b,
int b_stride, int *min_ret, int *max_ret) {
int min = 255;
int max = 0;
for (int i = 0; i < 8; i++) {
@@ -110,9 +108,9 @@ TEST_P(MinMaxTest, CompareReferenceAndVaryStride) {
reference_minmax(a, a_stride, b, b_stride, &min_ref, &max_ref);
ASM_REGISTER_STATE_CHECK(mm_func_(a, a_stride, b, b_stride, &min, &max));
EXPECT_EQ(max_ref, max) << "when a_stride = " << a_stride
<< " and b_stride = " << b_stride;;
<< " and b_stride = " << b_stride;
EXPECT_EQ(min_ref, min) << "when a_stride = " << a_stride
<< " and b_stride = " << b_stride;;
<< " and b_stride = " << b_stride;
}
}
}

View File

@@ -12,6 +12,8 @@
#include <stdlib.h>
#include <string.h>
#include <limits>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vp9_rtcd.h"
@@ -23,321 +25,668 @@
#include "vp9/common/vp9_blockd.h"
#include "vp9/common/vp9_scan.h"
#include "vpx/vpx_integer.h"
#include "vpx_ports/vpx_timer.h"
using libvpx_test::ACMRandom;
namespace {
typedef void (*FwdTxfmFunc)(const int16_t *in, tran_low_t *out, int stride);
typedef void (*InvTxfmFunc)(const tran_low_t *in, uint8_t *out, int stride);
typedef std::tr1::tuple<FwdTxfmFunc,
InvTxfmFunc,
InvTxfmFunc,
TX_SIZE, int> PartialInvTxfmParam;
typedef void (*InvTxfmWithBdFunc)(const tran_low_t *in, uint8_t *out,
int stride, int bd);
template <InvTxfmFunc fn>
void wrapper(const tran_low_t *in, uint8_t *out, int stride, int bd) {
(void)bd;
fn(in, out, stride);
}
#if CONFIG_VP9_HIGHBITDEPTH
template <InvTxfmWithBdFunc fn>
void highbd_wrapper(const tran_low_t *in, uint8_t *out, int stride, int bd) {
fn(in, CONVERT_TO_BYTEPTR(out), stride, bd);
}
#endif
typedef std::tr1::tuple<FwdTxfmFunc, InvTxfmWithBdFunc, InvTxfmWithBdFunc,
TX_SIZE, int, int, int>
PartialInvTxfmParam;
const int kMaxNumCoeffs = 1024;
const int kCountTestBlock = 1000;
// https://bugs.chromium.org/p/webm/issues/detail?id=1332
// The functions specified do not pass with INT16_MIN/MAX. They fail at the
// value specified, but pass when 1 is added/subtracted.
int16_t MaxSupportedCoeff(InvTxfmWithBdFunc a) {
#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_EMULATE_HARDWARE
if (a == &wrapper<vpx_idct8x8_64_add_ssse3> ||
a == &wrapper<vpx_idct8x8_12_add_ssse3>) {
return 23625 - 1;
}
#else
(void)a;
#endif
return std::numeric_limits<int16_t>::max();
}
int16_t MinSupportedCoeff(InvTxfmWithBdFunc a) {
#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_EMULATE_HARDWARE
if (a == &wrapper<vpx_idct8x8_64_add_ssse3> ||
a == &wrapper<vpx_idct8x8_12_add_ssse3>) {
return -23625 + 1;
}
#else
(void)a;
#endif
return std::numeric_limits<int16_t>::min();
}
class PartialIDctTest : public ::testing::TestWithParam<PartialInvTxfmParam> {
public:
virtual ~PartialIDctTest() {}
virtual void SetUp() {
rnd_.Reset(ACMRandom::DeterministicSeed());
ftxfm_ = GET_PARAM(0);
full_itxfm_ = GET_PARAM(1);
partial_itxfm_ = GET_PARAM(2);
tx_size_ = GET_PARAM(3);
tx_size_ = GET_PARAM(3);
last_nonzero_ = GET_PARAM(4);
bit_depth_ = GET_PARAM(5);
pixel_size_ = GET_PARAM(6);
mask_ = (1 << bit_depth_) - 1;
switch (tx_size_) {
case TX_4X4: size_ = 4; break;
case TX_8X8: size_ = 8; break;
case TX_16X16: size_ = 16; break;
case TX_32X32: size_ = 32; break;
default: FAIL() << "Wrong Size!"; break;
}
// Randomize stride_ to a value less than or equal to 1024
stride_ = rnd_(1024) + 1;
if (stride_ < size_) {
stride_ = size_;
}
// Align stride_ to 16 if it's bigger than 16.
if (stride_ > 16) {
stride_ &= ~15;
}
input_block_size_ = size_ * size_;
output_block_size_ = size_ * stride_;
input_block_ = reinterpret_cast<tran_low_t *>(
vpx_memalign(16, sizeof(*input_block_) * input_block_size_));
output_block_ = reinterpret_cast<uint8_t *>(
vpx_memalign(16, pixel_size_ * output_block_size_));
output_block_ref_ = reinterpret_cast<uint8_t *>(
vpx_memalign(16, pixel_size_ * output_block_size_));
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
virtual void TearDown() {
vpx_free(input_block_);
input_block_ = NULL;
vpx_free(output_block_);
output_block_ = NULL;
vpx_free(output_block_ref_);
output_block_ref_ = NULL;
libvpx_test::ClearSystemState();
}
void InitMem() {
memset(input_block_, 0, sizeof(*input_block_) * input_block_size_);
if (pixel_size_ == 1) {
for (int j = 0; j < output_block_size_; ++j) {
output_block_[j] = output_block_ref_[j] = rnd_.Rand16() & mask_;
}
} else {
ASSERT_EQ(2, pixel_size_);
uint16_t *const output = reinterpret_cast<uint16_t *>(output_block_);
uint16_t *const output_ref =
reinterpret_cast<uint16_t *>(output_block_ref_);
for (int j = 0; j < output_block_size_; ++j) {
output[j] = output_ref[j] = rnd_.Rand16() & mask_;
}
}
}
void InitInput() {
const int max_coeff = 32766 / 4;
int max_energy_leftover = max_coeff * max_coeff;
for (int j = 0; j < last_nonzero_; ++j) {
int16_t coeff = static_cast<int16_t>(sqrt(1.0 * max_energy_leftover) *
(rnd_.Rand16() - 32768) / 65536);
max_energy_leftover -= coeff * coeff;
if (max_energy_leftover < 0) {
max_energy_leftover = 0;
coeff = 0;
}
input_block_[vp9_default_scan_orders[tx_size_].scan[j]] = coeff;
}
}
protected:
int last_nonzero_;
TX_SIZE tx_size_;
tran_low_t *input_block_;
uint8_t *output_block_;
uint8_t *output_block_ref_;
int size_;
int stride_;
int pixel_size_;
int input_block_size_;
int output_block_size_;
int bit_depth_;
int mask_;
FwdTxfmFunc ftxfm_;
InvTxfmFunc full_itxfm_;
InvTxfmFunc partial_itxfm_;
InvTxfmWithBdFunc full_itxfm_;
InvTxfmWithBdFunc partial_itxfm_;
ACMRandom rnd_;
};
TEST_P(PartialIDctTest, RunQuantCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
int size;
switch (tx_size_) {
case TX_4X4:
size = 4;
break;
case TX_8X8:
size = 8;
break;
case TX_16X16:
size = 16;
break;
case TX_32X32:
size = 32;
break;
default:
FAIL() << "Wrong Size!";
break;
}
DECLARE_ALIGNED(16, tran_low_t, test_coef_block1[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, tran_low_t, test_coef_block2[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, uint8_t, dst1[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, uint8_t, dst2[kMaxNumCoeffs]);
const int count_test_block = 1000;
const int block_size = size * size;
DECLARE_ALIGNED(16, int16_t, input_extreme_block[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, tran_low_t, output_ref_block[kMaxNumCoeffs]);
int max_error = 0;
for (int i = 0; i < count_test_block; ++i) {
// clear out destination buffer
memset(dst1, 0, sizeof(*dst1) * block_size);
memset(dst2, 0, sizeof(*dst2) * block_size);
memset(test_coef_block1, 0, sizeof(*test_coef_block1) * block_size);
memset(test_coef_block2, 0, sizeof(*test_coef_block2) * block_size);
ACMRandom rnd(ACMRandom::DeterministicSeed());
for (int i = 0; i < count_test_block; ++i) {
// Initialize a test block with input range [-255, 255].
if (i == 0) {
for (int j = 0; j < block_size; ++j)
input_extreme_block[j] = 255;
} else if (i == 1) {
for (int j = 0; j < block_size; ++j)
input_extreme_block[j] = -255;
} else {
for (int j = 0; j < block_size; ++j) {
input_extreme_block[j] = rnd.Rand8() % 2 ? 255 : -255;
}
InitMem();
for (int i = 0; i < kCountTestBlock * kCountTestBlock; ++i) {
// Initialize a test block with input range [-mask_, mask_].
if (i == 0) {
for (int k = 0; k < input_block_size_; ++k) {
input_extreme_block[k] = mask_;
}
} else if (i == 1) {
for (int k = 0; k < input_block_size_; ++k) {
input_extreme_block[k] = -mask_;
}
} else {
for (int k = 0; k < input_block_size_; ++k) {
input_extreme_block[k] = rnd_.Rand8() % 2 ? mask_ : -mask_;
}
ftxfm_(input_extreme_block, output_ref_block, size);
// quantization with maximum allowed step sizes
test_coef_block1[0] = (output_ref_block[0] / 1336) * 1336;
for (int j = 1; j < last_nonzero_; ++j)
test_coef_block1[vp9_default_scan_orders[tx_size_].scan[j]]
= (output_ref_block[j] / 1828) * 1828;
}
ASM_REGISTER_STATE_CHECK(full_itxfm_(test_coef_block1, dst1, size));
ASM_REGISTER_STATE_CHECK(partial_itxfm_(test_coef_block1, dst2, size));
ftxfm_(input_extreme_block, output_ref_block, size_);
for (int j = 0; j < block_size; ++j) {
const int diff = dst1[j] - dst2[j];
const int error = diff * diff;
if (max_error < error)
max_error = error;
// quantization with minimum allowed step sizes
input_block_[0] = (output_ref_block[0] / 4) * 4;
for (int k = 1; k < last_nonzero_; ++k) {
const int pos = vp9_default_scan_orders[tx_size_].scan[k];
input_block_[pos] = (output_ref_block[pos] / 4) * 4;
}
ASM_REGISTER_STATE_CHECK(
full_itxfm_(input_block_, output_block_ref_, stride_, bit_depth_));
ASM_REGISTER_STATE_CHECK(
partial_itxfm_(input_block_, output_block_, stride_, bit_depth_));
ASSERT_EQ(0, memcmp(output_block_ref_, output_block_,
pixel_size_ * output_block_size_))
<< "Error: partial inverse transform produces different results";
}
EXPECT_EQ(0, max_error)
<< "Error: partial inverse transform produces different results";
}
TEST_P(PartialIDctTest, ResultsMatch) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
int size;
switch (tx_size_) {
case TX_4X4:
size = 4;
break;
case TX_8X8:
size = 8;
break;
case TX_16X16:
size = 16;
break;
case TX_32X32:
size = 32;
break;
default:
FAIL() << "Wrong Size!";
break;
for (int i = 0; i < kCountTestBlock; ++i) {
InitMem();
InitInput();
ASM_REGISTER_STATE_CHECK(
full_itxfm_(input_block_, output_block_ref_, stride_, bit_depth_));
ASM_REGISTER_STATE_CHECK(
partial_itxfm_(input_block_, output_block_, stride_, bit_depth_));
ASSERT_EQ(0, memcmp(output_block_ref_, output_block_,
pixel_size_ * output_block_size_))
<< "Error: partial inverse transform produces different results";
}
DECLARE_ALIGNED(16, tran_low_t, test_coef_block1[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, tran_low_t, test_coef_block2[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, uint8_t, dst1[kMaxNumCoeffs]);
DECLARE_ALIGNED(16, uint8_t, dst2[kMaxNumCoeffs]);
const int count_test_block = 1000;
const int max_coeff = 32766 / 4;
const int block_size = size * size;
int max_error = 0;
for (int i = 0; i < count_test_block; ++i) {
// clear out destination buffer
memset(dst1, 0, sizeof(*dst1) * block_size);
memset(dst2, 0, sizeof(*dst2) * block_size);
memset(test_coef_block1, 0, sizeof(*test_coef_block1) * block_size);
memset(test_coef_block2, 0, sizeof(*test_coef_block2) * block_size);
int max_energy_leftover = max_coeff * max_coeff;
}
TEST_P(PartialIDctTest, AddOutputBlock) {
for (int i = 0; i < kCountTestBlock; ++i) {
InitMem();
for (int j = 0; j < last_nonzero_; ++j) {
int16_t coef = static_cast<int16_t>(sqrt(1.0 * max_energy_leftover) *
(rnd.Rand16() - 32768) / 65536);
max_energy_leftover -= coef * coef;
if (max_energy_leftover < 0) {
max_energy_leftover = 0;
coef = 0;
}
test_coef_block1[vp9_default_scan_orders[tx_size_].scan[j]] = coef;
input_block_[vp9_default_scan_orders[tx_size_].scan[j]] = 10;
}
memcpy(test_coef_block2, test_coef_block1,
sizeof(*test_coef_block2) * block_size);
ASM_REGISTER_STATE_CHECK(
full_itxfm_(input_block_, output_block_ref_, stride_, bit_depth_));
ASM_REGISTER_STATE_CHECK(
partial_itxfm_(input_block_, output_block_, stride_, bit_depth_));
ASSERT_EQ(0, memcmp(output_block_ref_, output_block_,
pixel_size_ * output_block_size_))
<< "Error: Transform results are not correctly added to output.";
}
}
ASM_REGISTER_STATE_CHECK(full_itxfm_(test_coef_block1, dst1, size));
ASM_REGISTER_STATE_CHECK(partial_itxfm_(test_coef_block2, dst2, size));
TEST_P(PartialIDctTest, SingleExtremeCoeff) {
const int16_t max_coeff = MaxSupportedCoeff(partial_itxfm_);
const int16_t min_coeff = MinSupportedCoeff(partial_itxfm_);
for (int i = 0; i < last_nonzero_; ++i) {
memset(input_block_, 0, sizeof(*input_block_) * input_block_size_);
// Run once for min and once for max.
for (int j = 0; j < 2; ++j) {
const int coeff = j ? min_coeff : max_coeff;
for (int j = 0; j < block_size; ++j) {
const int diff = dst1[j] - dst2[j];
const int error = diff * diff;
if (max_error < error)
max_error = error;
memset(output_block_, 0, pixel_size_ * output_block_size_);
memset(output_block_ref_, 0, pixel_size_ * output_block_size_);
input_block_[vp9_default_scan_orders[tx_size_].scan[i]] = coeff;
ASM_REGISTER_STATE_CHECK(
full_itxfm_(input_block_, output_block_ref_, stride_, bit_depth_));
ASM_REGISTER_STATE_CHECK(
partial_itxfm_(input_block_, output_block_, stride_, bit_depth_));
ASSERT_EQ(0, memcmp(output_block_ref_, output_block_,
pixel_size_ * output_block_size_))
<< "Error: Fails with single coeff of " << coeff << " at " << i
<< ".";
}
}
}
EXPECT_EQ(0, max_error)
TEST_P(PartialIDctTest, DISABLED_Speed) {
// Keep runtime stable with transform size.
const int kCountSpeedTestBlock = 500000000 / input_block_size_;
InitMem();
InitInput();
for (int i = 0; i < kCountSpeedTestBlock; ++i) {
ASM_REGISTER_STATE_CHECK(
full_itxfm_(input_block_, output_block_ref_, stride_, bit_depth_));
}
vpx_usec_timer timer;
vpx_usec_timer_start(&timer);
for (int i = 0; i < kCountSpeedTestBlock; ++i) {
partial_itxfm_(input_block_, output_block_, stride_, bit_depth_);
}
libvpx_test::ClearSystemState();
vpx_usec_timer_mark(&timer);
const int elapsed_time =
static_cast<int>(vpx_usec_timer_elapsed(&timer) / 1000);
printf("idct%dx%d_%d (bitdepth %d) time: %5d ms ", size_, size_,
last_nonzero_, bit_depth_, elapsed_time);
ASSERT_EQ(0, memcmp(output_block_ref_, output_block_,
pixel_size_ * output_block_size_))
<< "Error: partial inverse transform produces different results";
}
using std::tr1::make_tuple;
INSTANTIATE_TEST_CASE_P(
C, PartialIDctTest,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_34_add_c,
TX_32X32, 34),
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_1_add_c,
TX_32X32, 1),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_10_add_c,
TX_16X16, 10),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_1_add_c,
TX_16X16, 1),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_12_add_c,
TX_8X8, 12),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_1_add_c,
TX_8X8, 1),
make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_c,
&vpx_idct4x4_1_add_c,
TX_4X4, 1)));
const PartialInvTxfmParam c_partial_idct_tests[] = {
#if CONFIG_VP9_HIGHBITDEPTH
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>, TX_32X32, 1024, 8, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>, TX_32X32, 1024, 10, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>, TX_32X32, 1024, 12, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_34_add_c>, TX_32X32, 34, 8, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_34_add_c>, TX_32X32, 34, 10, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_34_add_c>, TX_32X32, 34, 12, 2),
make_tuple(&vpx_highbd_fdct32x32_c,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_c>, TX_32X32, 1, 8, 2),
make_tuple(&vpx_highbd_fdct32x32_c,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_c>, TX_32X32, 1, 10, 2),
make_tuple(&vpx_highbd_fdct32x32_c,
&highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_c>, TX_32X32, 1, 12, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>, TX_16X16, 256, 8, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>, TX_16X16, 256, 10, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>, TX_16X16, 256, 12, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_c>, TX_16X16, 10, 8, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_c>, TX_16X16, 10, 10, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_c>, TX_16X16, 10, 12, 2),
make_tuple(&vpx_highbd_fdct16x16_c,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_1_add_c>, TX_16X16, 1, 8, 2),
make_tuple(&vpx_highbd_fdct16x16_c,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_1_add_c>, TX_16X16, 1, 10, 2),
make_tuple(&vpx_highbd_fdct16x16_c,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_1_add_c>, TX_16X16, 1, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>, TX_8X8, 64, 8, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>, TX_8X8, 64, 10, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>, TX_8X8, 64, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_c>, TX_8X8, 12, 8, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_c>, TX_8X8, 12, 10, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_c>, TX_8X8, 12, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_c>, TX_8X8, 1, 8, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_c>, TX_8X8, 1, 10, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_c>, TX_8X8, 1, 12, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>, TX_4X4, 16, 8, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>, TX_4X4, 16, 10, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>, TX_4X4, 16, 12, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_c>, TX_4X4, 1, 8, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_c>, TX_4X4, 1, 10, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_c>, TX_4X4, 1, 12, 2),
#endif // CONFIG_VP9_HIGHBITDEPTH
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_c>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_135_add_c>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_c>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1_add_c>, TX_32X32, 1, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_256_add_c>, TX_16X16, 256, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_10_add_c>, TX_16X16, 10, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_1_add_c>, TX_16X16, 1, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_c>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_c>, TX_8X8, 12, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_1_add_c>, TX_8X8, 1, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_16_add_c>, TX_4X4, 16, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_1_add_c>, TX_4X4, 1, 8, 1)
};
#if HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
NEON, PartialIDctTest,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_1_add_neon,
TX_32X32, 1),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_10_add_neon,
TX_16X16, 10),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_1_add_neon,
TX_16X16, 1),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_12_add_neon,
TX_8X8, 12),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_1_add_neon,
TX_8X8, 1),
make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_c,
&vpx_idct4x4_1_add_neon,
TX_4X4, 1)));
#endif // HAVE_NEON && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(C, PartialIDctTest,
::testing::ValuesIn(c_partial_idct_tests));
#if HAVE_SSE2 && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSE2, PartialIDctTest,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_34_add_sse2,
TX_32X32, 34),
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_1_add_sse2,
TX_32X32, 1),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_10_add_sse2,
TX_16X16, 10),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_1_add_sse2,
TX_16X16, 1),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_12_add_sse2,
TX_8X8, 12),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_1_add_sse2,
TX_8X8, 1),
make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_c,
&vpx_idct4x4_1_add_sse2,
TX_4X4, 1)));
#endif
#if HAVE_NEON && !CONFIG_EMULATE_HARDWARE
const PartialInvTxfmParam neon_partial_idct_tests[] = {
#if CONFIG_VP9_HIGHBITDEPTH
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_neon>, TX_8X8, 64, 8, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_neon>, TX_8X8, 64, 10, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_neon>, TX_8X8, 64, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_neon>, TX_8X8, 12, 8, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_12_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_neon>, TX_8X8, 12, 10, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_12_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_neon>, TX_8X8, 12, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_1_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_neon>, TX_8X8, 1, 8, 2),
make_tuple(&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_1_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_neon>, TX_8X8, 1, 10, 2),
make_tuple(&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_1_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_1_add_neon>, TX_8X8, 1, 12, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_neon>, TX_4X4, 16, 8, 2),
make_tuple(
&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_neon>, TX_4X4, 16, 10, 2),
make_tuple(
&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_neon>, TX_4X4, 16, 12, 2),
make_tuple(&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_1_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_neon>, TX_4X4, 1, 8, 2),
make_tuple(&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_1_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_neon>, TX_4X4, 1, 10, 2),
make_tuple(&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_1_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_1_add_neon>, TX_4X4, 1, 12, 2),
#endif // CONFIG_VP9_HIGHBITDEPTH
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_neon>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_135_add_neon>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_neon>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1_add_neon>, TX_32X32, 1, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_256_add_neon>, TX_16X16, 256, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_10_add_neon>, TX_16X16, 10, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_1_add_neon>, TX_16X16, 1, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_neon>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_neon>, TX_8X8, 12, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_1_add_neon>, TX_8X8, 1, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_16_add_neon>, TX_4X4, 16, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_1_add_neon>, TX_4X4, 1, 8, 1)
};
#if HAVE_SSSE3 && CONFIG_USE_X86INC && ARCH_X86_64 && \
!CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
SSSE3_64, PartialIDctTest,
::testing::Values(
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_12_add_ssse3,
TX_8X8, 12)));
#endif
INSTANTIATE_TEST_CASE_P(NEON, PartialIDctTest,
::testing::ValuesIn(neon_partial_idct_tests));
#endif // HAVE_NEON && !CONFIG_EMULATE_HARDWARE
#if HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
INSTANTIATE_TEST_CASE_P(
MSA, PartialIDctTest,
::testing::Values(
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_34_add_msa,
TX_32X32, 34),
make_tuple(&vpx_fdct32x32_c,
&vpx_idct32x32_1024_add_c,
&vpx_idct32x32_1_add_msa,
TX_32X32, 1),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_10_add_msa,
TX_16X16, 10),
make_tuple(&vpx_fdct16x16_c,
&vpx_idct16x16_256_add_c,
&vpx_idct16x16_1_add_msa,
TX_16X16, 1),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_12_add_msa,
TX_8X8, 10),
make_tuple(&vpx_fdct8x8_c,
&vpx_idct8x8_64_add_c,
&vpx_idct8x8_1_add_msa,
TX_8X8, 1),
make_tuple(&vpx_fdct4x4_c,
&vpx_idct4x4_16_add_c,
&vpx_idct4x4_1_add_msa,
TX_4X4, 1)));
#endif // HAVE_MSA && !CONFIG_VP9_HIGHBITDEPTH && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE
// 32x32_135_ is implemented using the 1024 version.
const PartialInvTxfmParam sse2_partial_idct_tests[] = {
#if CONFIG_VP9_HIGHBITDEPTH
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_sse2>, TX_32X32, 1, 8, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_sse2>, TX_32X32, 1, 10, 2),
make_tuple(
&vpx_highbd_fdct32x32_c, &highbd_wrapper<vpx_highbd_idct32x32_1024_add_c>,
&highbd_wrapper<vpx_highbd_idct32x32_1_add_sse2>, TX_32X32, 1, 12, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_sse2>, TX_16X16, 256, 8, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_sse2>, TX_16X16, 256, 10, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_256_add_sse2>, TX_16X16, 256, 12, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_sse2>, TX_16X16, 10, 8, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_sse2>, TX_16X16, 10, 10, 2),
make_tuple(
&vpx_highbd_fdct16x16_c, &highbd_wrapper<vpx_highbd_idct16x16_256_add_c>,
&highbd_wrapper<vpx_highbd_idct16x16_10_add_sse2>, TX_16X16, 10, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_sse2>, TX_8X8, 64, 8, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_sse2>, TX_8X8, 64, 10, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_sse2>, TX_8X8, 64, 12, 2),
make_tuple(&vpx_highbd_fdct8x8_c,
&highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_sse2>, TX_8X8, 12, 8, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_sse2>, TX_8X8, 12, 10, 2),
make_tuple(
&vpx_highbd_fdct8x8_c, &highbd_wrapper<vpx_highbd_idct8x8_64_add_c>,
&highbd_wrapper<vpx_highbd_idct8x8_12_add_sse2>, TX_8X8, 12, 12, 2),
make_tuple(&vpx_highbd_fdct4x4_c,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_sse2>, TX_4X4, 16, 8, 2),
make_tuple(
&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_sse2>, TX_4X4, 16, 10, 2),
make_tuple(
&vpx_highbd_fdct4x4_c, &highbd_wrapper<vpx_highbd_idct4x4_16_add_c>,
&highbd_wrapper<vpx_highbd_idct4x4_16_add_sse2>, TX_4X4, 16, 12, 2),
#endif // CONFIG_VP9_HIGHBITDEPTH
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_sse2>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_sse2>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_sse2>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1_add_sse2>, TX_32X32, 1, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_256_add_sse2>, TX_16X16, 256, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_10_add_sse2>, TX_16X16, 10, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_1_add_sse2>, TX_16X16, 1, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_sse2>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_sse2>, TX_8X8, 12, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_1_add_sse2>, TX_8X8, 1, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_16_add_sse2>, TX_4X4, 16, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_1_add_sse2>, TX_4X4, 1, 8, 1)
};
INSTANTIATE_TEST_CASE_P(SSE2, PartialIDctTest,
::testing::ValuesIn(sse2_partial_idct_tests));
#endif // HAVE_SSE2 && !CONFIG_EMULATE_HARDWARE
#if HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_EMULATE_HARDWARE
const PartialInvTxfmParam ssse3_partial_idct_tests[] = {
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_ssse3>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_135_add_ssse3>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_ssse3>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_ssse3>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_ssse3>, TX_8X8, 12, 8, 1)
};
INSTANTIATE_TEST_CASE_P(SSSE3, PartialIDctTest,
::testing::ValuesIn(ssse3_partial_idct_tests));
#endif // HAVE_SSSE3 && ARCH_X86_64 && !CONFIG_EMULATE_HARDWARE
#if HAVE_DSPR2 && !CONFIG_EMULATE_HARDWARE && !CONFIG_VP9_HIGHBITDEPTH
const PartialInvTxfmParam dspr2_partial_idct_tests[] = {
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_dspr2>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_dspr2>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_dspr2>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1_add_dspr2>, TX_32X32, 1, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_256_add_dspr2>, TX_16X16, 256, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_10_add_dspr2>, TX_16X16, 10, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_1_add_dspr2>, TX_16X16, 1, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_dspr2>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_dspr2>, TX_8X8, 12, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_1_add_dspr2>, TX_8X8, 1, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_16_add_dspr2>, TX_4X4, 16, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_1_add_dspr2>, TX_4X4, 1, 8, 1)
};
INSTANTIATE_TEST_CASE_P(DSPR2, PartialIDctTest,
::testing::ValuesIn(dspr2_partial_idct_tests));
#endif // HAVE_DSPR2 && !CONFIG_EMULATE_HARDWARE && !CONFIG_VP9_HIGHBITDEPTH
#if HAVE_MSA && !CONFIG_EMULATE_HARDWARE && !CONFIG_VP9_HIGHBITDEPTH
// 32x32_135_ is implemented using the 1024 version.
const PartialInvTxfmParam msa_partial_idct_tests[] = {
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_msa>, TX_32X32, 1024, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1024_add_msa>, TX_32X32, 135, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_34_add_msa>, TX_32X32, 34, 8, 1),
make_tuple(&vpx_fdct32x32_c, &wrapper<vpx_idct32x32_1024_add_c>,
&wrapper<vpx_idct32x32_1_add_msa>, TX_32X32, 1, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_256_add_msa>, TX_16X16, 256, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_10_add_msa>, TX_16X16, 10, 8, 1),
make_tuple(&vpx_fdct16x16_c, &wrapper<vpx_idct16x16_256_add_c>,
&wrapper<vpx_idct16x16_1_add_msa>, TX_16X16, 1, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_64_add_msa>, TX_8X8, 64, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_12_add_msa>, TX_8X8, 12, 8, 1),
make_tuple(&vpx_fdct8x8_c, &wrapper<vpx_idct8x8_64_add_c>,
&wrapper<vpx_idct8x8_1_add_msa>, TX_8X8, 1, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_16_add_msa>, TX_4X4, 16, 8, 1),
make_tuple(&vpx_fdct4x4_c, &wrapper<vpx_idct4x4_16_add_c>,
&wrapper<vpx_idct4x4_1_add_msa>, TX_4X4, 1, 8, 1)
};
INSTANTIATE_TEST_CASE_P(MSA, PartialIDctTest,
::testing::ValuesIn(msa_partial_idct_tests));
#endif // HAVE_MSA && !CONFIG_EMULATE_HARDWARE && !CONFIG_VP9_HIGHBITDEPTH
} // namespace

View File

@@ -7,38 +7,50 @@
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <limits.h>
#include "./vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vp8_rtcd.h"
#include "vpx/vpx_integer.h"
#include "vpx_mem/vpx_mem.h"
typedef void (*PostProcFunc)(unsigned char *src_ptr,
unsigned char *dst_ptr,
int src_pixels_per_line,
int dst_pixels_per_line,
int cols,
unsigned char *flimit,
int size);
using libvpx_test::ACMRandom;
typedef void (*VpxPostProcDownAndAcrossMbRowFunc)(
unsigned char *src_ptr, unsigned char *dst_ptr, int src_pixels_per_line,
int dst_pixels_per_line, int cols, unsigned char *flimit, int size);
typedef void (*VpxMbPostProcAcrossIpFunc)(unsigned char *src, int pitch,
int rows, int cols, int flimit);
typedef void (*VpxMbPostProcDownFunc)(unsigned char *dst, int pitch, int rows,
int cols, int flimit);
namespace {
class VP8PostProcessingFilterTest
: public ::testing::TestWithParam<PostProcFunc> {
// Compute the filter level used in post proc from the loop filter strength
int q2mbl(int x) {
if (x < 20) x = 20;
x = 50 + (x - 50) * 10 / 8;
return x * x / 3;
}
class VpxPostProcDownAndAcrossMbRowTest
: public ::testing::TestWithParam<VpxPostProcDownAndAcrossMbRowFunc> {
public:
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
};
// Test routine for the VP8 post-processing function
// vp8_post_proc_down_and_across_mb_row_c.
// Test routine for the VPx post-processing function
// vpx_post_proc_down_and_across_mb_row_c.
TEST_P(VP8PostProcessingFilterTest, FilterOutputCheck) {
TEST_P(VpxPostProcDownAndAcrossMbRowTest, CheckFilterOutput) {
// Size of the underlying data block that will be filtered.
const int block_width = 16;
const int block_width = 16;
const int block_height = 16;
// 5-tap filter needs 2 padding rows above and below the block in the input.
@@ -53,14 +65,20 @@ TEST_P(VP8PostProcessingFilterTest, FilterOutputCheck) {
const int output_stride = output_width;
const int output_size = output_width * output_height;
uint8_t *const src_image =
reinterpret_cast<uint8_t*>(vpx_calloc(input_size, 1));
uint8_t *const dst_image =
reinterpret_cast<uint8_t*>(vpx_calloc(output_size, 1));
uint8_t *const src_image = new uint8_t[input_size];
ASSERT_TRUE(src_image != NULL);
// Though the left padding is only 8 bytes, the assembly code tries to
// read 16 bytes before the pointer.
uint8_t *const dst_image = new uint8_t[output_size + 8];
ASSERT_TRUE(dst_image != NULL);
// Pointers to top-left pixel of block in the input and output images.
uint8_t *const src_image_ptr = src_image + (input_stride << 1);
uint8_t *const dst_image_ptr = dst_image + 8;
// The assembly works in increments of 16. The first read may be offset by
// this amount.
uint8_t *const dst_image_ptr = dst_image + 16;
uint8_t *const flimits =
reinterpret_cast<uint8_t *>(vpx_memalign(16, block_width));
(void)memset(flimits, 255, block_width);
@@ -80,39 +98,518 @@ TEST_P(VP8PostProcessingFilterTest, FilterOutputCheck) {
// Initialize pixels in the output to 99.
(void)memset(dst_image, 99, output_size);
ASM_REGISTER_STATE_CHECK(
GetParam()(src_image_ptr, dst_image_ptr, input_stride,
output_stride, block_width, flimits, 16));
ASM_REGISTER_STATE_CHECK(GetParam()(src_image_ptr, dst_image_ptr,
input_stride, output_stride, block_width,
flimits, 16));
static const uint8_t expected_data[block_height] = {
static const uint8_t kExpectedOutput[block_height] = {
4, 3, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 3, 4
};
pixel_ptr = dst_image_ptr;
for (int i = 0; i < block_height; ++i) {
for (int j = 0; j < block_width; ++j) {
EXPECT_EQ(expected_data[i], pixel_ptr[j])
<< "VP8PostProcessingFilterTest failed with invalid filter output";
ASSERT_EQ(kExpectedOutput[i], pixel_ptr[j]) << "at (" << i << ", " << j
<< ")";
}
pixel_ptr += output_stride;
}
vpx_free(src_image);
vpx_free(dst_image);
delete[] src_image;
delete[] dst_image;
vpx_free(flimits);
};
INSTANTIATE_TEST_CASE_P(C, VP8PostProcessingFilterTest,
::testing::Values(vp8_post_proc_down_and_across_mb_row_c));
TEST_P(VpxPostProcDownAndAcrossMbRowTest, CheckCvsAssembly) {
// Size of the underlying data block that will be filtered.
// Y blocks are always a multiple of 16 wide and exactly 16 high. U and V
// blocks are always a multiple of 8 wide and exactly 8 high.
const int block_width = 136;
const int block_height = 16;
// 5-tap filter needs 2 padding rows above and below the block in the input.
// SSE2 reads in blocks of 16. Pad an extra 8 in case the width is not %16.
const int input_width = block_width;
const int input_height = block_height + 4 + 8;
const int input_stride = input_width;
const int input_size = input_stride * input_height;
// Filter extends output block by 8 samples at left and right edges.
// SSE2 reads in blocks of 16. Pad an extra 8 in case the width is not %16.
const int output_width = block_width + 24;
const int output_height = block_height;
const int output_stride = output_width;
const int output_size = output_stride * output_height;
uint8_t *const src_image = new uint8_t[input_size];
ASSERT_TRUE(src_image != NULL);
// Though the left padding is only 8 bytes, the assembly code tries to
// read 16 bytes before the pointer.
uint8_t *const dst_image = new uint8_t[output_size + 8];
ASSERT_TRUE(dst_image != NULL);
uint8_t *const dst_image_ref = new uint8_t[output_size + 8];
ASSERT_TRUE(dst_image_ref != NULL);
// Pointers to top-left pixel of block in the input and output images.
uint8_t *const src_image_ptr = src_image + (input_stride << 1);
// The assembly works in increments of 16. The first read may be offset by
// this amount.
uint8_t *const dst_image_ptr = dst_image + 16;
uint8_t *const dst_image_ref_ptr = dst_image + 16;
// Filter values are set in blocks of 16 for Y and 8 for U/V. Each macroblock
// can have a different filter. SSE2 assembly reads flimits in blocks of 16 so
// it must be padded out.
const int flimits_width = block_width % 16 ? block_width + 8 : block_width;
uint8_t *const flimits =
reinterpret_cast<uint8_t *>(vpx_memalign(16, flimits_width));
ACMRandom rnd;
rnd.Reset(ACMRandom::DeterministicSeed());
// Initialize pixels in the input:
// block pixels to random values.
// border pixels to value 10.
(void)memset(src_image, 10, input_size);
uint8_t *pixel_ptr = src_image_ptr;
for (int i = 0; i < block_height; ++i) {
for (int j = 0; j < block_width; ++j) {
pixel_ptr[j] = rnd.Rand8();
}
pixel_ptr += input_stride;
}
for (int blocks = 0; blocks < block_width; blocks += 8) {
(void)memset(flimits, 0, sizeof(*flimits) * flimits_width);
for (int f = 0; f < 255; f++) {
(void)memset(flimits + blocks, f, sizeof(*flimits) * 8);
(void)memset(dst_image, 0, output_size);
(void)memset(dst_image_ref, 0, output_size);
vpx_post_proc_down_and_across_mb_row_c(
src_image_ptr, dst_image_ref_ptr, input_stride, output_stride,
block_width, flimits, block_height);
ASM_REGISTER_STATE_CHECK(GetParam()(src_image_ptr, dst_image_ptr,
input_stride, output_stride,
block_width, flimits, 16));
for (int i = 0; i < block_height; ++i) {
for (int j = 0; j < block_width; ++j) {
ASSERT_EQ(dst_image_ref_ptr[j + i * output_stride],
dst_image_ptr[j + i * output_stride])
<< "at (" << i << ", " << j << ")";
}
}
}
}
delete[] src_image;
delete[] dst_image;
delete[] dst_image_ref;
vpx_free(flimits);
}
class VpxMbPostProcAcrossIpTest
: public ::testing::TestWithParam<VpxMbPostProcAcrossIpFunc> {
public:
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
void SetCols(unsigned char *s, int rows, int cols, int src_width) {
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
s[c] = c;
}
s += src_width;
}
}
void RunComparison(const unsigned char *expected_output, unsigned char *src_c,
int rows, int cols, int src_pitch) {
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
ASSERT_EQ(expected_output[c], src_c[c]) << "at (" << r << ", " << c
<< ")";
}
src_c += src_pitch;
}
}
void RunFilterLevel(unsigned char *s, int rows, int cols, int src_width,
int filter_level, const unsigned char *expected_output) {
ASM_REGISTER_STATE_CHECK(
GetParam()(s, src_width, rows, cols, filter_level));
RunComparison(expected_output, s, rows, cols, src_width);
}
};
TEST_P(VpxMbPostProcAcrossIpTest, CheckLowFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_left_padding = 8;
const int src_right_padding = 17;
const int src_width = cols + src_left_padding + src_right_padding;
const int src_size = rows * src_width;
unsigned char *const src = new unsigned char[src_size];
ASSERT_TRUE(src != NULL);
memset(src, 10, src_size);
unsigned char *const s = src + src_left_padding;
SetCols(s, rows, cols, src_width);
unsigned char *expected_output = new unsigned char[rows * cols];
ASSERT_TRUE(expected_output != NULL);
SetCols(expected_output, rows, cols, cols);
RunFilterLevel(s, rows, cols, src_width, q2mbl(0), expected_output);
delete[] src;
delete[] expected_output;
}
TEST_P(VpxMbPostProcAcrossIpTest, CheckMediumFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_left_padding = 8;
const int src_right_padding = 17;
const int src_width = cols + src_left_padding + src_right_padding;
const int src_size = rows * src_width;
unsigned char *const src = new unsigned char[src_size];
ASSERT_TRUE(src != NULL);
memset(src, 10, src_size);
unsigned char *const s = src + src_left_padding;
SetCols(s, rows, cols, src_width);
static const unsigned char kExpectedOutput[cols] = {
2, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 13
};
RunFilterLevel(s, rows, cols, src_width, q2mbl(70), kExpectedOutput);
delete[] src;
}
TEST_P(VpxMbPostProcAcrossIpTest, CheckHighFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_left_padding = 8;
const int src_right_padding = 17;
const int src_width = cols + src_left_padding + src_right_padding;
const int src_size = rows * src_width;
unsigned char *const src = new unsigned char[src_size];
ASSERT_TRUE(src != NULL);
unsigned char *const s = src + src_left_padding;
memset(src, 10, src_size);
SetCols(s, rows, cols, src_width);
static const unsigned char kExpectedOutput[cols] = {
2, 2, 3, 4, 4, 5, 6, 7, 8, 9, 10, 11, 11, 12, 13, 13
};
RunFilterLevel(s, rows, cols, src_width, INT_MAX, kExpectedOutput);
memset(src, 10, src_size);
SetCols(s, rows, cols, src_width);
RunFilterLevel(s, rows, cols, src_width, q2mbl(100), kExpectedOutput);
delete[] src;
}
TEST_P(VpxMbPostProcAcrossIpTest, CheckCvsAssembly) {
const int rows = 16;
const int cols = 16;
const int src_left_padding = 8;
const int src_right_padding = 17;
const int src_width = cols + src_left_padding + src_right_padding;
const int src_size = rows * src_width;
unsigned char *const c_mem = new unsigned char[src_size];
unsigned char *const asm_mem = new unsigned char[src_size];
ASSERT_TRUE(c_mem != NULL);
ASSERT_TRUE(asm_mem != NULL);
unsigned char *const src_c = c_mem + src_left_padding;
unsigned char *const src_asm = asm_mem + src_left_padding;
// When level >= 100, the filter behaves the same as the level = INT_MAX
// When level < 20, it behaves the same as the level = 0
for (int level = 0; level < 100; level++) {
memset(c_mem, 10, src_size);
memset(asm_mem, 10, src_size);
SetCols(src_c, rows, cols, src_width);
SetCols(src_asm, rows, cols, src_width);
vpx_mbpost_proc_across_ip_c(src_c, src_width, rows, cols, q2mbl(level));
ASM_REGISTER_STATE_CHECK(
GetParam()(src_asm, src_width, rows, cols, q2mbl(level)));
RunComparison(src_c, src_asm, rows, cols, src_width);
}
delete[] c_mem;
delete[] asm_mem;
}
class VpxMbPostProcDownTest
: public ::testing::TestWithParam<VpxMbPostProcDownFunc> {
public:
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
void SetRows(unsigned char *src_c, int rows, int cols) {
for (int r = 0; r < rows; r++) {
memset(src_c, r, cols);
src_c += cols;
}
}
void SetRandom(unsigned char *src_c, unsigned char *src_asm, int rows,
int cols, int src_pitch) {
ACMRandom rnd;
rnd.Reset(ACMRandom::DeterministicSeed());
// Add some random noise to the input
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
const int noise = rnd(4);
src_c[c] = r + noise;
src_asm[c] = r + noise;
}
src_c += src_pitch;
src_asm += src_pitch;
}
}
void SetRandomSaturation(unsigned char *src_c, unsigned char *src_asm,
int rows, int cols, int src_pitch) {
ACMRandom rnd;
rnd.Reset(ACMRandom::DeterministicSeed());
// Add some random noise to the input
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
const int noise = 3 * rnd(2);
src_c[c] = r + noise;
src_asm[c] = r + noise;
}
src_c += src_pitch;
src_asm += src_pitch;
}
}
void RunComparison(const unsigned char *expected_output, unsigned char *src_c,
int rows, int cols, int src_pitch) {
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
ASSERT_EQ(expected_output[r * rows + c], src_c[c]) << "at (" << r
<< ", " << c << ")";
}
src_c += src_pitch;
}
}
void RunComparison(unsigned char *src_c, unsigned char *src_asm, int rows,
int cols, int src_pitch) {
for (int r = 0; r < rows; r++) {
for (int c = 0; c < cols; c++) {
ASSERT_EQ(src_c[c], src_asm[c]) << "at (" << r << ", " << c << ")";
}
src_c += src_pitch;
src_asm += src_pitch;
}
}
void RunFilterLevel(unsigned char *s, int rows, int cols, int src_width,
int filter_level, const unsigned char *expected_output) {
ASM_REGISTER_STATE_CHECK(
GetParam()(s, src_width, rows, cols, filter_level));
RunComparison(expected_output, s, rows, cols, src_width);
}
};
TEST_P(VpxMbPostProcDownTest, CheckHighFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_pitch = cols;
const int src_top_padding = 8;
const int src_bottom_padding = 17;
const int src_size = cols * (rows + src_top_padding + src_bottom_padding);
unsigned char *const c_mem = new unsigned char[src_size];
ASSERT_TRUE(c_mem != NULL);
memset(c_mem, 10, src_size);
unsigned char *const src_c = c_mem + src_top_padding * src_pitch;
SetRows(src_c, rows, cols);
static const unsigned char kExpectedOutput[rows * cols] = {
2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 2,
2, 3, 2, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 3, 3, 3, 3, 3, 3,
3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 3, 4, 4, 3, 3, 3,
4, 4, 3, 4, 4, 3, 3, 4, 5, 4, 4, 4, 4, 4, 4, 4, 5, 4, 4,
4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5,
5, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 7, 7,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 9, 8, 9, 9, 8, 8, 8, 9,
9, 8, 9, 9, 8, 8, 8, 9, 9, 10, 10, 9, 9, 9, 10, 10, 9, 10, 10,
9, 9, 9, 10, 10, 10, 11, 10, 10, 10, 11, 10, 11, 10, 11, 10, 10, 10, 11,
10, 11, 11, 11, 11, 11, 11, 11, 12, 11, 11, 11, 11, 11, 11, 11, 12, 11, 12,
12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 13, 12,
13, 12, 13, 12, 12, 12, 13, 12, 13, 12, 13, 12, 13, 13, 13, 14, 13, 13, 13,
13, 13, 13, 13, 14, 13, 13, 13, 13
};
RunFilterLevel(src_c, rows, cols, src_pitch, INT_MAX, kExpectedOutput);
memset(c_mem, 10, src_size);
SetRows(src_c, rows, cols);
RunFilterLevel(src_c, rows, cols, src_pitch, q2mbl(100), kExpectedOutput);
delete[] c_mem;
}
TEST_P(VpxMbPostProcDownTest, CheckMediumFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_pitch = cols;
const int src_top_padding = 8;
const int src_bottom_padding = 17;
const int src_size = cols * (rows + src_top_padding + src_bottom_padding);
unsigned char *const c_mem = new unsigned char[src_size];
ASSERT_TRUE(c_mem != NULL);
memset(c_mem, 10, src_size);
unsigned char *const src_c = c_mem + src_top_padding * src_pitch;
SetRows(src_c, rows, cols);
static const unsigned char kExpectedOutput[rows * cols] = {
2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 1, 1, 2, 2, 2, 2, 2, 2, 2,
2, 3, 2, 2, 2, 2, 2, 2, 2, 3, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3,
3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4, 4,
4, 4, 4, 4, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5,
5, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 6, 7, 7,
7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 8, 8, 8, 8, 8,
8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 9, 9, 9, 9, 9, 9, 9, 9,
9, 9, 9, 9, 9, 9, 9, 9, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10, 10,
10, 10, 10, 10, 10, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11, 11,
11, 11, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 12, 13,
13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 13, 12, 12, 13, 12,
13, 12, 13, 12, 12, 12, 13, 12, 13, 12, 13, 12, 13, 13, 13, 14, 13, 13, 13,
13, 13, 13, 13, 14, 13, 13, 13, 13
};
RunFilterLevel(src_c, rows, cols, src_pitch, q2mbl(70), kExpectedOutput);
delete[] c_mem;
}
TEST_P(VpxMbPostProcDownTest, CheckLowFilterOutput) {
const int rows = 16;
const int cols = 16;
const int src_pitch = cols;
const int src_top_padding = 8;
const int src_bottom_padding = 17;
const int src_size = cols * (rows + src_top_padding + src_bottom_padding);
unsigned char *const c_mem = new unsigned char[src_size];
ASSERT_TRUE(c_mem != NULL);
memset(c_mem, 10, src_size);
unsigned char *const src_c = c_mem + src_top_padding * src_pitch;
SetRows(src_c, rows, cols);
unsigned char *expected_output = new unsigned char[rows * cols];
ASSERT_TRUE(expected_output != NULL);
SetRows(expected_output, rows, cols);
RunFilterLevel(src_c, rows, cols, src_pitch, q2mbl(0), expected_output);
delete[] c_mem;
delete[] expected_output;
}
TEST_P(VpxMbPostProcDownTest, CheckCvsAssembly) {
const int rows = 16;
const int cols = 16;
const int src_pitch = cols;
const int src_top_padding = 8;
const int src_bottom_padding = 17;
const int src_size = cols * (rows + src_top_padding + src_bottom_padding);
unsigned char *const c_mem = new unsigned char[src_size];
unsigned char *const asm_mem = new unsigned char[src_size];
ASSERT_TRUE(c_mem != NULL);
ASSERT_TRUE(asm_mem != NULL);
unsigned char *const src_c = c_mem + src_top_padding * src_pitch;
unsigned char *const src_asm = asm_mem + src_top_padding * src_pitch;
for (int level = 0; level < 100; level++) {
memset(c_mem, 10, src_size);
memset(asm_mem, 10, src_size);
SetRandom(src_c, src_asm, rows, cols, src_pitch);
vpx_mbpost_proc_down_c(src_c, src_pitch, rows, cols, q2mbl(level));
ASM_REGISTER_STATE_CHECK(
GetParam()(src_asm, src_pitch, rows, cols, q2mbl(level)));
RunComparison(src_c, src_asm, rows, cols, src_pitch);
memset(c_mem, 10, src_size);
memset(asm_mem, 10, src_size);
SetRandomSaturation(src_c, src_asm, rows, cols, src_pitch);
vpx_mbpost_proc_down_c(src_c, src_pitch, rows, cols, q2mbl(level));
ASM_REGISTER_STATE_CHECK(
GetParam()(src_asm, src_pitch, rows, cols, q2mbl(level)));
RunComparison(src_c, src_asm, rows, cols, src_pitch);
}
delete[] c_mem;
delete[] asm_mem;
}
INSTANTIATE_TEST_CASE_P(
C, VpxPostProcDownAndAcrossMbRowTest,
::testing::Values(vpx_post_proc_down_and_across_mb_row_c));
INSTANTIATE_TEST_CASE_P(C, VpxMbPostProcAcrossIpTest,
::testing::Values(vpx_mbpost_proc_across_ip_c));
INSTANTIATE_TEST_CASE_P(C, VpxMbPostProcDownTest,
::testing::Values(vpx_mbpost_proc_down_c));
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(SSE2, VP8PostProcessingFilterTest,
::testing::Values(vp8_post_proc_down_and_across_mb_row_sse2));
#endif
INSTANTIATE_TEST_CASE_P(
SSE2, VpxPostProcDownAndAcrossMbRowTest,
::testing::Values(vpx_post_proc_down_and_across_mb_row_sse2));
INSTANTIATE_TEST_CASE_P(SSE2, VpxMbPostProcAcrossIpTest,
::testing::Values(vpx_mbpost_proc_across_ip_sse2));
INSTANTIATE_TEST_CASE_P(SSE2, VpxMbPostProcDownTest,
::testing::Values(vpx_mbpost_proc_down_sse2));
#endif // HAVE_SSE2
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(
NEON, VpxPostProcDownAndAcrossMbRowTest,
::testing::Values(vpx_post_proc_down_and_across_mb_row_neon));
INSTANTIATE_TEST_CASE_P(NEON, VpxMbPostProcAcrossIpTest,
::testing::Values(vpx_mbpost_proc_across_ip_neon));
#endif // HAVE_NEON
#if HAVE_MSA
INSTANTIATE_TEST_CASE_P(MSA, VP8PostProcessingFilterTest,
::testing::Values(vp8_post_proc_down_and_across_mb_row_msa));
#endif
INSTANTIATE_TEST_CASE_P(
MSA, VpxPostProcDownAndAcrossMbRowTest,
::testing::Values(vpx_post_proc_down_and_across_mb_row_msa));
INSTANTIATE_TEST_CASE_P(MSA, VpxMbPostProcAcrossIpTest,
::testing::Values(vpx_mbpost_proc_across_ip_msa));
INSTANTIATE_TEST_CASE_P(MSA, VpxMbPostProcDownTest,
::testing::Values(vpx_mbpost_proc_down_msa));
#endif // HAVE_MSA
} // namespace

376
test/predict_test.cc Normal file
View File

@@ -0,0 +1,376 @@
/*
* Copyright (c) 2013 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <stdlib.h>
#include <string.h>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vp8_rtcd.h"
#include "./vpx_config.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vpx/vpx_integer.h"
#include "vpx_mem/vpx_mem.h"
namespace {
using libvpx_test::ACMRandom;
using std::tr1::make_tuple;
typedef void (*PredictFunc)(uint8_t *src_ptr, int src_pixels_per_line,
int xoffset, int yoffset, uint8_t *dst_ptr,
int dst_pitch);
typedef std::tr1::tuple<int, int, PredictFunc> PredictParam;
class PredictTestBase : public ::testing::TestWithParam<PredictParam> {
public:
PredictTestBase()
: width_(GET_PARAM(0)), height_(GET_PARAM(1)), predict_(GET_PARAM(2)),
src_(NULL), padded_dst_(NULL), dst_(NULL), dst_c_(NULL) {}
virtual void SetUp() {
src_ = new uint8_t[kSrcSize];
ASSERT_TRUE(src_ != NULL);
// padded_dst_ provides a buffer of kBorderSize around the destination
// memory to facilitate detecting out of bounds writes.
dst_stride_ = kBorderSize + width_ + kBorderSize;
padded_dst_size_ = dst_stride_ * (kBorderSize + height_ + kBorderSize);
padded_dst_ =
reinterpret_cast<uint8_t *>(vpx_memalign(16, padded_dst_size_));
ASSERT_TRUE(padded_dst_ != NULL);
dst_ = padded_dst_ + (kBorderSize * dst_stride_) + kBorderSize;
dst_c_ = new uint8_t[16 * 16];
ASSERT_TRUE(dst_c_ != NULL);
memset(src_, 0, kSrcSize);
memset(padded_dst_, 128, padded_dst_size_);
memset(dst_c_, 0, 16 * 16);
}
virtual void TearDown() {
delete[] src_;
src_ = NULL;
vpx_free(padded_dst_);
padded_dst_ = NULL;
dst_ = NULL;
delete[] dst_c_;
dst_c_ = NULL;
libvpx_test::ClearSystemState();
}
protected:
// Make reference arrays big enough for 16x16 functions. Six-tap filters need
// 5 extra pixels outside of the macroblock.
static const int kSrcStride = 21;
static const int kSrcSize = kSrcStride * kSrcStride;
static const int kBorderSize = 16;
int width_;
int height_;
PredictFunc predict_;
uint8_t *src_;
uint8_t *padded_dst_;
uint8_t *dst_;
int padded_dst_size_;
uint8_t *dst_c_;
int dst_stride_;
bool CompareBuffers(const uint8_t *a, int a_stride, const uint8_t *b,
int b_stride) const {
for (int height = 0; height < height_; ++height) {
EXPECT_EQ(0, memcmp(a + height * a_stride, b + height * b_stride,
sizeof(*a) * width_))
<< "Row " << height << " does not match.";
}
return !HasFailure();
}
// Given a block of memory 'a' with size 'a_size', determine if all regions
// excepting block 'b' described by 'b_stride', 'b_height', and 'b_width'
// match pixel value 'c'.
bool CheckBorder(const uint8_t *a, int a_size, const uint8_t *b, int b_width,
int b_height, int b_stride, uint8_t c) const {
const uint8_t *a_end = a + a_size;
const int b_size = (b_stride * b_height) + b_width;
const uint8_t *b_end = b + b_size;
const int left_border = (b_stride - b_width) / 2;
const int right_border = left_border + ((b_stride - b_width) % 2);
EXPECT_GE(b - left_border, a) << "'b' does not start within 'a'";
EXPECT_LE(b_end + right_border, a_end) << "'b' does not end within 'a'";
// Top border.
for (int pixel = 0; pixel < b - a - left_border; ++pixel) {
EXPECT_EQ(c, a[pixel]) << "Mismatch at " << pixel << " in top border.";
}
// Left border.
for (int height = 0; height < b_height; ++height) {
for (int width = left_border; width > 0; --width) {
EXPECT_EQ(c, b[height * b_stride - width])
<< "Mismatch at row " << height << " column " << left_border - width
<< " in left border.";
}
}
// Right border.
for (int height = 0; height < b_height; ++height) {
for (int width = b_width; width < b_width + right_border; ++width) {
EXPECT_EQ(c, b[height * b_stride + width])
<< "Mismatch at row " << height << " column " << width - b_width
<< " in right border.";
}
}
// Bottom border.
for (int pixel = static_cast<int>(b - a + b_size); pixel < a_size;
++pixel) {
EXPECT_EQ(c, a[pixel]) << "Mismatch at " << pixel << " in bottom border.";
}
return !HasFailure();
}
void TestWithRandomData(PredictFunc reference) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
// Run tests for almost all possible offsets.
for (int xoffset = 0; xoffset < 8; ++xoffset) {
for (int yoffset = 0; yoffset < 8; ++yoffset) {
if (xoffset == 0 && yoffset == 0) {
// This represents a copy which is not required to be handled by this
// module.
continue;
}
for (int i = 0; i < kSrcSize; ++i) {
src_[i] = rnd.Rand8();
}
reference(&src_[kSrcStride * 2 + 2], kSrcStride, xoffset, yoffset,
dst_c_, 16);
ASM_REGISTER_STATE_CHECK(predict_(&src_[kSrcStride * 2 + 2], kSrcStride,
xoffset, yoffset, dst_, dst_stride_));
ASSERT_TRUE(CompareBuffers(dst_c_, 16, dst_, dst_stride_));
ASSERT_TRUE(CheckBorder(padded_dst_, padded_dst_size_, dst_, width_,
height_, dst_stride_, 128));
}
}
}
void TestWithUnalignedDst(PredictFunc reference) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
// Only the 4x4 need to be able to handle unaligned writes.
if (width_ == 4 && height_ == 4) {
for (int xoffset = 0; xoffset < 8; ++xoffset) {
for (int yoffset = 0; yoffset < 8; ++yoffset) {
if (xoffset == 0 && yoffset == 0) {
continue;
}
for (int i = 0; i < kSrcSize; ++i) {
src_[i] = rnd.Rand8();
}
reference(&src_[kSrcStride * 2 + 2], kSrcStride, xoffset, yoffset,
dst_c_, 16);
for (int i = 1; i < 4; ++i) {
memset(padded_dst_, 128, padded_dst_size_);
ASM_REGISTER_STATE_CHECK(predict_(&src_[kSrcStride * 2 + 2],
kSrcStride, xoffset, yoffset,
dst_ + i, dst_stride_ + i));
ASSERT_TRUE(CompareBuffers(dst_c_, 16, dst_ + i, dst_stride_ + i));
ASSERT_TRUE(CheckBorder(padded_dst_, padded_dst_size_, dst_ + i,
width_, height_, dst_stride_ + i, 128));
}
}
}
}
}
};
class SixtapPredictTest : public PredictTestBase {};
TEST_P(SixtapPredictTest, TestWithRandomData) {
TestWithRandomData(vp8_sixtap_predict16x16_c);
}
TEST_P(SixtapPredictTest, TestWithUnalignedDst) {
TestWithUnalignedDst(vp8_sixtap_predict16x16_c);
}
TEST_P(SixtapPredictTest, TestWithPresetData) {
// Test input
static const uint8_t kTestData[kSrcSize] = {
184, 4, 191, 82, 92, 41, 0, 1, 226, 236, 172, 20, 182, 42, 226,
177, 79, 94, 77, 179, 203, 206, 198, 22, 192, 19, 75, 17, 192, 44,
233, 120, 48, 168, 203, 141, 210, 203, 143, 180, 184, 59, 201, 110, 102,
171, 32, 182, 10, 109, 105, 213, 60, 47, 236, 253, 67, 55, 14, 3,
99, 247, 124, 148, 159, 71, 34, 114, 19, 177, 38, 203, 237, 239, 58,
83, 155, 91, 10, 166, 201, 115, 124, 5, 163, 104, 2, 231, 160, 16,
234, 4, 8, 103, 153, 167, 174, 187, 26, 193, 109, 64, 141, 90, 48,
200, 174, 204, 36, 184, 114, 237, 43, 238, 242, 207, 86, 245, 182, 247,
6, 161, 251, 14, 8, 148, 182, 182, 79, 208, 120, 188, 17, 6, 23,
65, 206, 197, 13, 242, 126, 128, 224, 170, 110, 211, 121, 197, 200, 47,
188, 207, 208, 184, 221, 216, 76, 148, 143, 156, 100, 8, 89, 117, 14,
112, 183, 221, 54, 197, 208, 180, 69, 176, 94, 180, 131, 215, 121, 76,
7, 54, 28, 216, 238, 249, 176, 58, 142, 64, 215, 242, 72, 49, 104,
87, 161, 32, 52, 216, 230, 4, 141, 44, 181, 235, 224, 57, 195, 89,
134, 203, 144, 162, 163, 126, 156, 84, 185, 42, 148, 145, 29, 221, 194,
134, 52, 100, 166, 105, 60, 140, 110, 201, 184, 35, 181, 153, 93, 121,
243, 227, 68, 131, 134, 232, 2, 35, 60, 187, 77, 209, 76, 106, 174,
15, 241, 227, 115, 151, 77, 175, 36, 187, 121, 221, 223, 47, 118, 61,
168, 105, 32, 237, 236, 167, 213, 238, 202, 17, 170, 24, 226, 247, 131,
145, 6, 116, 117, 121, 11, 194, 41, 48, 126, 162, 13, 93, 209, 131,
154, 122, 237, 187, 103, 217, 99, 60, 200, 45, 78, 115, 69, 49, 106,
200, 194, 112, 60, 56, 234, 72, 251, 19, 120, 121, 182, 134, 215, 135,
10, 114, 2, 247, 46, 105, 209, 145, 165, 153, 191, 243, 12, 5, 36,
119, 206, 231, 231, 11, 32, 209, 83, 27, 229, 204, 149, 155, 83, 109,
35, 93, 223, 37, 84, 14, 142, 37, 160, 52, 191, 96, 40, 204, 101,
77, 67, 52, 53, 43, 63, 85, 253, 147, 113, 226, 96, 6, 125, 179,
115, 161, 17, 83, 198, 101, 98, 85, 139, 3, 137, 75, 99, 178, 23,
201, 255, 91, 253, 52, 134, 60, 138, 131, 208, 251, 101, 48, 2, 227,
228, 118, 132, 245, 202, 75, 91, 44, 160, 231, 47, 41, 50, 147, 220,
74, 92, 219, 165, 89, 16
};
// Expected results for xoffset = 2 and yoffset = 2.
static const int kExpectedDstStride = 16;
static const uint8_t kExpectedDst[256] = {
117, 102, 74, 135, 42, 98, 175, 206, 70, 73, 222, 197, 50, 24, 39,
49, 38, 105, 90, 47, 169, 40, 171, 215, 200, 73, 109, 141, 53, 85,
177, 164, 79, 208, 124, 89, 212, 18, 81, 145, 151, 164, 217, 153, 91,
154, 102, 102, 159, 75, 164, 152, 136, 51, 213, 219, 186, 116, 193, 224,
186, 36, 231, 208, 84, 211, 155, 167, 35, 59, 42, 76, 216, 149, 73,
201, 78, 149, 184, 100, 96, 196, 189, 198, 188, 235, 195, 117, 129, 120,
129, 49, 25, 133, 113, 69, 221, 114, 70, 143, 99, 157, 108, 189, 140,
78, 6, 55, 65, 240, 255, 245, 184, 72, 90, 100, 116, 131, 39, 60,
234, 167, 33, 160, 88, 185, 200, 157, 159, 176, 127, 151, 138, 102, 168,
106, 170, 86, 82, 219, 189, 76, 33, 115, 197, 106, 96, 198, 136, 97,
141, 237, 151, 98, 137, 191, 185, 2, 57, 95, 142, 91, 255, 185, 97,
137, 76, 162, 94, 173, 131, 193, 161, 81, 106, 72, 135, 222, 234, 137,
66, 137, 106, 243, 210, 147, 95, 15, 137, 110, 85, 66, 16, 96, 167,
147, 150, 173, 203, 140, 118, 196, 84, 147, 160, 19, 95, 101, 123, 74,
132, 202, 82, 166, 12, 131, 166, 189, 170, 159, 85, 79, 66, 57, 152,
132, 203, 194, 0, 1, 56, 146, 180, 224, 156, 28, 83, 181, 79, 76,
80, 46, 160, 175, 59, 106, 43, 87, 75, 136, 85, 189, 46, 71, 200,
90
};
ASM_REGISTER_STATE_CHECK(
predict_(const_cast<uint8_t *>(kTestData) + kSrcStride * 2 + 2,
kSrcStride, 2, 2, dst_, dst_stride_));
ASSERT_TRUE(
CompareBuffers(kExpectedDst, kExpectedDstStride, dst_, dst_stride_));
}
INSTANTIATE_TEST_CASE_P(
C, SixtapPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_sixtap_predict16x16_c),
make_tuple(8, 8, &vp8_sixtap_predict8x8_c),
make_tuple(8, 4, &vp8_sixtap_predict8x4_c),
make_tuple(4, 4, &vp8_sixtap_predict4x4_c)));
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(
NEON, SixtapPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_sixtap_predict16x16_neon),
make_tuple(8, 8, &vp8_sixtap_predict8x8_neon),
make_tuple(8, 4, &vp8_sixtap_predict8x4_neon),
make_tuple(4, 4, &vp8_sixtap_predict4x4_neon)));
#endif
#if HAVE_MMX
INSTANTIATE_TEST_CASE_P(
MMX, SixtapPredictTest,
::testing::Values(make_tuple(4, 4, &vp8_sixtap_predict4x4_mmx)));
#endif
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(
SSE2, SixtapPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_sixtap_predict16x16_sse2),
make_tuple(8, 8, &vp8_sixtap_predict8x8_sse2),
make_tuple(8, 4, &vp8_sixtap_predict8x4_sse2)));
#endif
#if HAVE_SSSE3
INSTANTIATE_TEST_CASE_P(
SSSE3, SixtapPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_sixtap_predict16x16_ssse3),
make_tuple(8, 8, &vp8_sixtap_predict8x8_ssse3),
make_tuple(8, 4, &vp8_sixtap_predict8x4_ssse3),
make_tuple(4, 4, &vp8_sixtap_predict4x4_ssse3)));
#endif
#if HAVE_MSA
INSTANTIATE_TEST_CASE_P(
MSA, SixtapPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_sixtap_predict16x16_msa),
make_tuple(8, 8, &vp8_sixtap_predict8x8_msa),
make_tuple(8, 4, &vp8_sixtap_predict8x4_msa),
make_tuple(4, 4, &vp8_sixtap_predict4x4_msa)));
#endif
class BilinearPredictTest : public PredictTestBase {};
TEST_P(BilinearPredictTest, TestWithRandomData) {
TestWithRandomData(vp8_bilinear_predict16x16_c);
}
TEST_P(BilinearPredictTest, TestWithUnalignedDst) {
TestWithUnalignedDst(vp8_bilinear_predict16x16_c);
}
INSTANTIATE_TEST_CASE_P(
C, BilinearPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_bilinear_predict16x16_c),
make_tuple(8, 8, &vp8_bilinear_predict8x8_c),
make_tuple(8, 4, &vp8_bilinear_predict8x4_c),
make_tuple(4, 4, &vp8_bilinear_predict4x4_c)));
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(
NEON, BilinearPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_bilinear_predict16x16_neon),
make_tuple(8, 8, &vp8_bilinear_predict8x8_neon),
make_tuple(8, 4, &vp8_bilinear_predict8x4_neon),
make_tuple(4, 4, &vp8_bilinear_predict4x4_neon)));
#endif
#if HAVE_MMX
INSTANTIATE_TEST_CASE_P(
MMX, BilinearPredictTest,
::testing::Values(make_tuple(8, 4, &vp8_bilinear_predict8x4_mmx),
make_tuple(4, 4, &vp8_bilinear_predict4x4_mmx)));
#endif
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(
SSE2, BilinearPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_bilinear_predict16x16_sse2),
make_tuple(8, 8, &vp8_bilinear_predict8x8_sse2)));
#endif
#if HAVE_SSSE3
INSTANTIATE_TEST_CASE_P(
SSSE3, BilinearPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_bilinear_predict16x16_ssse3),
make_tuple(8, 8, &vp8_bilinear_predict8x8_ssse3)));
#endif
#if HAVE_MSA
INSTANTIATE_TEST_CASE_P(
MSA, BilinearPredictTest,
::testing::Values(make_tuple(16, 16, &vp8_bilinear_predict16x16_msa),
make_tuple(8, 8, &vp8_bilinear_predict8x8_msa),
make_tuple(8, 4, &vp8_bilinear_predict8x4_msa),
make_tuple(4, 4, &vp8_bilinear_predict4x4_msa)));
#endif
} // namespace

View File

@@ -23,8 +23,7 @@ class RealtimeTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
RealtimeTest()
: EncoderTest(GET_PARAM(0)), frame_packets_(0) {}
RealtimeTest() : EncoderTest(GET_PARAM(0)), frame_packets_(0) {}
virtual ~RealtimeTest() {}
virtual void SetUp() {

View File

@@ -32,11 +32,13 @@
#undef NOMINMAX
#define NOMINMAX
#ifndef WIN32_LEAN_AND_MEAN
#define WIN32_LEAN_AND_MEAN
#endif
#include <windows.h>
#include <winnt.h>
inline bool operator==(const M128A& lhs, const M128A& rhs) {
inline bool operator==(const M128A &lhs, const M128A &rhs) {
return (lhs.Low == rhs.Low && lhs.High == rhs.High);
}
@@ -48,10 +50,10 @@ namespace libvpx_test {
class RegisterStateCheck {
public:
RegisterStateCheck() { initialized_ = StoreRegisters(&pre_context_); }
~RegisterStateCheck() { EXPECT_TRUE(Check()); }
~RegisterStateCheck() { Check(); }
private:
static bool StoreRegisters(CONTEXT* const context) {
static bool StoreRegisters(CONTEXT *const context) {
const HANDLE this_thread = GetCurrentThread();
EXPECT_TRUE(this_thread != NULL);
context->ContextFlags = CONTEXT_FLOATING_POINT;
@@ -61,34 +63,34 @@ class RegisterStateCheck {
}
// Compares the register state. Returns true if the states match.
bool Check() const {
if (!initialized_) return false;
void Check() const {
ASSERT_TRUE(initialized_);
CONTEXT post_context;
if (!StoreRegisters(&post_context)) return false;
ASSERT_TRUE(StoreRegisters(&post_context));
const M128A* xmm_pre = &pre_context_.Xmm6;
const M128A* xmm_post = &post_context.Xmm6;
const M128A *xmm_pre = &pre_context_.Xmm6;
const M128A *xmm_post = &post_context.Xmm6;
for (int i = 6; i <= 15; ++i) {
EXPECT_EQ(*xmm_pre, *xmm_post) << "xmm" << i << " has been modified!";
++xmm_pre;
++xmm_post;
}
return !testing::Test::HasNonfatalFailure();
}
bool initialized_;
CONTEXT pre_context_;
};
#define ASM_REGISTER_STATE_CHECK(statement) do { \
libvpx_test::RegisterStateCheck reg_check; \
statement; \
} while (false)
#define ASM_REGISTER_STATE_CHECK(statement) \
do { \
libvpx_test::RegisterStateCheck reg_check; \
statement; \
} while (false)
} // namespace libvpx_test
#elif defined(CONFIG_SHARED) && defined(HAVE_NEON_ASM) && defined(CONFIG_VP9) \
&& !CONFIG_SHARED && HAVE_NEON_ASM && CONFIG_VP9
#elif defined(CONFIG_SHARED) && defined(HAVE_NEON_ASM) && \
defined(CONFIG_VP9) && !CONFIG_SHARED && HAVE_NEON_ASM && CONFIG_VP9
extern "C" {
// Save the d8-d15 registers into store.
@@ -102,35 +104,28 @@ namespace libvpx_test {
// arm platform.
class RegisterStateCheck {
public:
RegisterStateCheck() { initialized_ = StoreRegisters(pre_store_); }
~RegisterStateCheck() { EXPECT_TRUE(Check()); }
RegisterStateCheck() { vpx_push_neon(pre_store_); }
~RegisterStateCheck() { Check(); }
private:
static bool StoreRegisters(int64_t store[8]) {
vpx_push_neon(store);
return true;
}
// Compares the register state. Returns true if the states match.
bool Check() const {
if (!initialized_) return false;
void Check() const {
int64_t post_store[8];
vpx_push_neon(post_store);
for (int i = 0; i < 8; ++i) {
EXPECT_EQ(pre_store_[i], post_store[i]) << "d"
<< i + 8 << " has been modified";
EXPECT_EQ(pre_store_[i], post_store[i]) << "d" << i + 8
<< " has been modified";
}
return !testing::Test::HasNonfatalFailure();
}
bool initialized_;
int64_t pre_store_[8];
};
#define ASM_REGISTER_STATE_CHECK(statement) do { \
libvpx_test::RegisterStateCheck reg_check; \
statement; \
} while (false)
#define ASM_REGISTER_STATE_CHECK(statement) \
do { \
libvpx_test::RegisterStateCheck reg_check; \
statement; \
} while (false)
} // namespace libvpx_test
@@ -156,12 +151,12 @@ class RegisterStateCheckMMX {
RegisterStateCheckMMX() {
__asm__ volatile("fstenv %0" : "=rm"(pre_fpu_env_));
}
~RegisterStateCheckMMX() { EXPECT_TRUE(Check()); }
~RegisterStateCheckMMX() { Check(); }
private:
// Checks the FPU tag word pre/post execution, returning false if not cleared
// to 0xffff.
bool Check() const {
void Check() const {
EXPECT_EQ(0xffff, pre_fpu_env_[4])
<< "FPU was in an inconsistent state prior to call";
@@ -169,16 +164,16 @@ class RegisterStateCheckMMX {
__asm__ volatile("fstenv %0" : "=rm"(post_fpu_env));
EXPECT_EQ(0xffff, post_fpu_env[4])
<< "FPU was left in an inconsistent state after call";
return !testing::Test::HasNonfatalFailure();
}
uint16_t pre_fpu_env_[14];
};
#define API_REGISTER_STATE_CHECK(statement) do { \
libvpx_test::RegisterStateCheckMMX reg_check; \
ASM_REGISTER_STATE_CHECK(statement); \
} while (false)
#define API_REGISTER_STATE_CHECK(statement) \
do { \
libvpx_test::RegisterStateCheckMMX reg_check; \
ASM_REGISTER_STATE_CHECK(statement); \
} while (false)
} // namespace libvpx_test

View File

@@ -44,9 +44,9 @@ static void write_ivf_file_header(const vpx_codec_enc_cfg_t *const cfg,
header[1] = 'K';
header[2] = 'I';
header[3] = 'F';
mem_put_le16(header + 4, 0); /* version */
mem_put_le16(header + 6, 32); /* headersize */
mem_put_le32(header + 8, 0x30395056); /* fourcc (vp9) */
mem_put_le16(header + 4, 0); /* version */
mem_put_le16(header + 6, 32); /* headersize */
mem_put_le32(header + 8, 0x30395056); /* fourcc (vp9) */
mem_put_le16(header + 12, cfg->g_w); /* width */
mem_put_le16(header + 14, cfg->g_h); /* height */
mem_put_le32(header + 16, cfg->g_timebase.den); /* rate */
@@ -68,8 +68,7 @@ static void write_ivf_frame_header(const vpx_codec_cx_pkt_t *const pkt,
char header[12];
vpx_codec_pts_t pts;
if (pkt->kind != VPX_CODEC_CX_FRAME_PKT)
return;
if (pkt->kind != VPX_CODEC_CX_FRAME_PKT) return;
pts = pkt->data.frame.pts;
mem_put_le32(header, static_cast<unsigned int>(pkt->data.frame.sz));
@@ -92,12 +91,9 @@ struct FrameInfo {
unsigned int h;
};
void ScaleForFrameNumber(unsigned int frame,
unsigned int initial_w,
unsigned int initial_h,
unsigned int *w,
unsigned int *h,
int flag_codec) {
void ScaleForFrameNumber(unsigned int frame, unsigned int initial_w,
unsigned int initial_h, unsigned int *w,
unsigned int *h, int flag_codec) {
if (frame < 10) {
*w = initial_w;
*h = initial_h;
@@ -219,7 +215,7 @@ void ScaleForFrameNumber(unsigned int frame,
return;
}
if (frame < 250) {
*w = initial_w / 2;
*w = initial_w / 2;
*h = initial_h / 2;
return;
}
@@ -268,8 +264,9 @@ class ResizingVideoSource : public ::libvpx_test::DummyVideoSource {
}
};
class ResizeTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
class ResizeTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWithParam<libvpx_test::TestMode> {
protected:
ResizeTest() : EncoderTest(GET_PARAM(0)) {}
@@ -285,7 +282,7 @@ class ResizeTest : public ::libvpx_test::EncoderTest,
frame_info_list_.push_back(FrameInfo(pts, img.d_w, img.d_h));
}
std::vector< FrameInfo > frame_info_list_;
std::vector<FrameInfo> frame_info_list_;
};
TEST_P(ResizeTest, TestExternalResizeWorks) {
@@ -299,12 +296,12 @@ TEST_P(ResizeTest, TestExternalResizeWorks) {
const unsigned int frame = static_cast<unsigned>(info->pts);
unsigned int expected_w;
unsigned int expected_h;
ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight,
&expected_w, &expected_h, 0);
EXPECT_EQ(expected_w, info->w)
<< "Frame " << frame << " had unexpected width";
EXPECT_EQ(expected_h, info->h)
<< "Frame " << frame << " had unexpected height";
ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight, &expected_w,
&expected_h, 0);
EXPECT_EQ(expected_w, info->w) << "Frame " << frame
<< " had unexpected width";
EXPECT_EQ(expected_h, info->h) << "Frame " << frame
<< " had unexpected height";
}
}
@@ -315,10 +312,7 @@ class ResizeInternalTest : public ResizeTest {
protected:
#if WRITE_COMPRESSED_STREAM
ResizeInternalTest()
: ResizeTest(),
frame0_psnr_(0.0),
outfile_(NULL),
out_frames_(0) {}
: ResizeTest(), frame0_psnr_(0.0), outfile_(NULL), out_frames_(0) {}
#else
ResizeInternalTest() : ResizeTest(), frame0_psnr_(0.0) {}
#endif
@@ -347,30 +341,29 @@ class ResizeInternalTest : public ResizeTest {
if (change_config_) {
int new_q = 60;
if (video->frame() == 0) {
struct vpx_scaling_mode mode = {VP8E_ONETWO, VP8E_ONETWO};
struct vpx_scaling_mode mode = { VP8E_ONETWO, VP8E_ONETWO };
encoder->Control(VP8E_SET_SCALEMODE, &mode);
}
if (video->frame() == 1) {
struct vpx_scaling_mode mode = {VP8E_NORMAL, VP8E_NORMAL};
struct vpx_scaling_mode mode = { VP8E_NORMAL, VP8E_NORMAL };
encoder->Control(VP8E_SET_SCALEMODE, &mode);
cfg_.rc_min_quantizer = cfg_.rc_max_quantizer = new_q;
encoder->Config(&cfg_);
}
} else {
if (video->frame() == kStepDownFrame) {
struct vpx_scaling_mode mode = {VP8E_FOURFIVE, VP8E_THREEFIVE};
struct vpx_scaling_mode mode = { VP8E_FOURFIVE, VP8E_THREEFIVE };
encoder->Control(VP8E_SET_SCALEMODE, &mode);
}
if (video->frame() == kStepUpFrame) {
struct vpx_scaling_mode mode = {VP8E_NORMAL, VP8E_NORMAL};
struct vpx_scaling_mode mode = { VP8E_NORMAL, VP8E_NORMAL };
encoder->Control(VP8E_SET_SCALEMODE, &mode);
}
}
}
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
if (!frame0_psnr_)
frame0_psnr_ = pkt->data.psnr.psnr[0];
if (frame0_psnr_ == 0.) frame0_psnr_ = pkt->data.psnr.psnr[0];
EXPECT_NEAR(pkt->data.psnr.psnr[0], frame0_psnr_, 2.0);
}
@@ -379,8 +372,7 @@ class ResizeInternalTest : public ResizeTest {
++out_frames_;
// Write initial file header if first frame.
if (pkt->data.frame.pts == 0)
write_ivf_file_header(&cfg_, 0, outfile_);
if (pkt->data.frame.pts == 0) write_ivf_file_header(&cfg_, 0, outfile_);
// Write frame header and data.
write_ivf_frame_header(pkt, outfile_);
@@ -434,8 +426,9 @@ TEST_P(ResizeInternalTest, TestInternalResizeChangeConfig) {
ASSERT_NO_FATAL_FAILURE(RunLoop(&video));
}
class ResizeRealtimeTest : public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
class ResizeRealtimeTest
: public ::libvpx_test::EncoderTest,
public ::libvpx_test::CodecTestWith2Params<libvpx_test::TestMode, int> {
protected:
ResizeRealtimeTest() : EncoderTest(GET_PARAM(0)) {}
virtual ~ResizeRealtimeTest() {}
@@ -465,16 +458,13 @@ class ResizeRealtimeTest : public ::libvpx_test::EncoderTest,
frame_info_list_.push_back(FrameInfo(pts, img.d_w, img.d_h));
}
virtual void MismatchHook(const vpx_image_t *img1,
const vpx_image_t *img2) {
virtual void MismatchHook(const vpx_image_t *img1, const vpx_image_t *img2) {
double mismatch_psnr = compute_psnr(img1, img2);
mismatch_psnr_ += mismatch_psnr;
++mismatch_nframes_;
}
unsigned int GetMismatchFrames() {
return mismatch_nframes_;
}
unsigned int GetMismatchFrames() { return mismatch_nframes_; }
void DefaultConfig() {
cfg_.rc_buf_initial_sz = 500;
@@ -491,14 +481,14 @@ class ResizeRealtimeTest : public ::libvpx_test::EncoderTest,
// Enable dropped frames.
cfg_.rc_dropframe_thresh = 1;
// Enable error_resilience mode.
cfg_.g_error_resilient = 1;
cfg_.g_error_resilient = 1;
// Enable dynamic resizing.
cfg_.rc_resize_allowed = 1;
// Run at low bitrate.
cfg_.rc_target_bitrate = 200;
}
std::vector< FrameInfo > frame_info_list_;
std::vector<FrameInfo> frame_info_list_;
int set_cpu_used_;
bool change_bitrate_;
double mismatch_psnr_;
@@ -521,12 +511,12 @@ TEST_P(ResizeRealtimeTest, TestExternalResizeWorks) {
const unsigned int frame = static_cast<unsigned>(info->pts);
unsigned int expected_w;
unsigned int expected_h;
ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight,
&expected_w, &expected_h, 1);
EXPECT_EQ(expected_w, info->w)
<< "Frame " << frame << " had unexpected width";
EXPECT_EQ(expected_h, info->h)
<< "Frame " << frame << " had unexpected height";
ScaleForFrameNumber(frame, kInitialWidth, kInitialHeight, &expected_w,
&expected_h, 1);
EXPECT_EQ(expected_w, info->w) << "Frame " << frame
<< " had unexpected width";
EXPECT_EQ(expected_h, info->h) << "Frame " << frame
<< " had unexpected height";
EXPECT_EQ(static_cast<unsigned int>(0), GetMismatchFrames());
}
}
@@ -618,10 +608,8 @@ TEST_P(ResizeRealtimeTest, TestInternalResizeDownUpChangeBitRate) {
}
vpx_img_fmt_t CspForFrameNumber(int frame) {
if (frame < 10)
return VPX_IMG_FMT_I420;
if (frame < 20)
return VPX_IMG_FMT_I444;
if (frame < 10) return VPX_IMG_FMT_I420;
if (frame < 20) return VPX_IMG_FMT_I444;
return VPX_IMG_FMT_I420;
}
@@ -629,10 +617,7 @@ class ResizeCspTest : public ResizeTest {
protected:
#if WRITE_COMPRESSED_STREAM
ResizeCspTest()
: ResizeTest(),
frame0_psnr_(0.0),
outfile_(NULL),
out_frames_(0) {}
: ResizeTest(), frame0_psnr_(0.0), outfile_(NULL), out_frames_(0) {}
#else
ResizeCspTest() : ResizeTest(), frame0_psnr_(0.0) {}
#endif
@@ -671,8 +656,7 @@ class ResizeCspTest : public ResizeTest {
}
virtual void PSNRPktHook(const vpx_codec_cx_pkt_t *pkt) {
if (!frame0_psnr_)
frame0_psnr_ = pkt->data.psnr.psnr[0];
if (frame0_psnr_ == 0.) frame0_psnr_ = pkt->data.psnr.psnr[0];
EXPECT_NEAR(pkt->data.psnr.psnr[0], frame0_psnr_, 2.0);
}
@@ -681,8 +665,7 @@ class ResizeCspTest : public ResizeTest {
++out_frames_;
// Write initial file header if first frame.
if (pkt->data.frame.pts == 0)
write_ivf_file_header(&cfg_, 0, outfile_);
if (pkt->data.frame.pts == 0) write_ivf_file_header(&cfg_, 0, outfile_);
// Write frame header and data.
write_ivf_frame_header(pkt, outfile_);

File diff suppressed because it is too large Load Diff

View File

@@ -8,7 +8,6 @@
* be found in the AUTHORS file in the root of the source tree.
*/
#include <math.h>
#include <stddef.h>
#include <stdio.h>
@@ -33,14 +32,10 @@ TEST(VP8RoiMapTest, ParameterCheck) {
unsigned int threshold[MAX_MB_SEGMENTS] = { 0, 100, 200, 300 };
const int internalq_trans[] = {
0, 1, 2, 3, 4, 5, 7, 8,
9, 10, 12, 13, 15, 17, 18, 19,
20, 21, 23, 24, 25, 26, 27, 28,
29, 30, 31, 33, 35, 37, 39, 41,
43, 45, 47, 49, 51, 53, 55, 57,
59, 61, 64, 67, 70, 73, 76, 79,
82, 85, 88, 91, 94, 97, 100, 103,
106, 109, 112, 115, 118, 121, 124, 127,
0, 1, 2, 3, 4, 5, 7, 8, 9, 10, 12, 13, 15, 17, 18, 19,
20, 21, 23, 24, 25, 26, 27, 28, 29, 30, 31, 33, 35, 37, 39, 41,
43, 45, 47, 49, 51, 53, 55, 57, 59, 61, 64, 67, 70, 73, 76, 79,
82, 85, 88, 91, 94, 97, 100, 103, 106, 109, 112, 115, 118, 121, 124, 127,
};
// Initialize elements of cpi with valid defaults.
@@ -60,17 +55,17 @@ TEST(VP8RoiMapTest, ParameterCheck) {
// Allocate memory for the source memory map.
unsigned char *roi_map =
reinterpret_cast<unsigned char *>(vpx_calloc(mbs, 1));
reinterpret_cast<unsigned char *>(vpx_calloc(mbs, 1));
memset(&roi_map[mbs >> 2], 1, (mbs >> 2));
memset(&roi_map[mbs >> 1], 2, (mbs >> 2));
memset(&roi_map[mbs -(mbs >> 2)], 3, (mbs >> 2));
memset(&roi_map[mbs - (mbs >> 2)], 3, (mbs >> 2));
// Do a test call with valid parameters.
int roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols, delta_q, delta_lf,
threshold);
int roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows, cpi.common.mb_cols,
delta_q, delta_lf, threshold);
EXPECT_EQ(0, roi_retval)
<< "vp8_set_roimap roi failed with default test parameters";
<< "vp8_set_roimap roi failed with default test parameters";
// Check that the values in the cpi structure get set as expected.
if (roi_retval == 0) {
@@ -83,9 +78,9 @@ TEST(VP8RoiMapTest, ParameterCheck) {
for (int i = 0; i < MAX_MB_SEGMENTS; ++i) {
const int transq = internalq_trans[abs(delta_q[i])];
if (abs(cpi.segment_feature_data[MB_LVL_ALT_Q][i]) != transq) {
EXPECT_EQ(transq, cpi.segment_feature_data[MB_LVL_ALT_Q][i])
<< "segment delta_q error";
break;
EXPECT_EQ(transq, cpi.segment_feature_data[MB_LVL_ALT_Q][i])
<< "segment delta_q error";
break;
}
}
@@ -93,7 +88,7 @@ TEST(VP8RoiMapTest, ParameterCheck) {
for (int i = 0; i < MAX_MB_SEGMENTS; ++i) {
if (cpi.segment_feature_data[MB_LVL_ALT_LF][i] != delta_lf[i]) {
EXPECT_EQ(delta_lf[i], cpi.segment_feature_data[MB_LVL_ALT_LF][i])
<< "segment delta_lf error";
<< "segment delta_lf error";
break;
}
}
@@ -101,23 +96,21 @@ TEST(VP8RoiMapTest, ParameterCheck) {
// Check the breakout thresholds
for (int i = 0; i < MAX_MB_SEGMENTS; ++i) {
unsigned int breakout =
static_cast<unsigned int>(cpi.segment_encode_breakout[i]);
static_cast<unsigned int>(cpi.segment_encode_breakout[i]);
if (threshold[i] != breakout) {
EXPECT_EQ(threshold[i], breakout)
<< "breakout threshold error";
EXPECT_EQ(threshold[i], breakout) << "breakout threshold error";
break;
}
}
// Segmentation, and segmentation update flages should be set.
EXPECT_EQ(1, cpi.mb.e_mbd.segmentation_enabled)
<< "segmentation_enabled error";
<< "segmentation_enabled error";
EXPECT_EQ(1, cpi.mb.e_mbd.update_mb_segmentation_map)
<< "update_mb_segmentation_map error";
<< "update_mb_segmentation_map error";
EXPECT_EQ(1, cpi.mb.e_mbd.update_mb_segmentation_data)
<< "update_mb_segmentation_data error";
<< "update_mb_segmentation_data error";
// Try a range of delta q and lf parameters (some legal, some not)
for (int i = 0; i < 1000; ++i) {
@@ -128,57 +121,54 @@ TEST(VP8RoiMapTest, ParameterCheck) {
rand_deltas[2] = rnd(160) - 80;
rand_deltas[3] = rnd(160) - 80;
deltas_valid = ((abs(rand_deltas[0]) <= 63) &&
(abs(rand_deltas[1]) <= 63) &&
(abs(rand_deltas[2]) <= 63) &&
(abs(rand_deltas[3]) <= 63)) ? 0 : -1;
deltas_valid =
((abs(rand_deltas[0]) <= 63) && (abs(rand_deltas[1]) <= 63) &&
(abs(rand_deltas[2]) <= 63) && (abs(rand_deltas[3]) <= 63))
? 0
: -1;
// Test with random delta q values.
roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols, rand_deltas,
delta_lf, threshold);
roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows, cpi.common.mb_cols,
rand_deltas, delta_lf, threshold);
EXPECT_EQ(deltas_valid, roi_retval) << "dq range check error";
// One delta_q error shown at a time
if (deltas_valid != roi_retval)
break;
if (deltas_valid != roi_retval) break;
// Test with random loop filter values.
roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols, delta_q,
rand_deltas, threshold);
roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows, cpi.common.mb_cols,
delta_q, rand_deltas, threshold);
EXPECT_EQ(deltas_valid, roi_retval) << "dlf range check error";
// One delta loop filter error shown at a time
if (deltas_valid != roi_retval)
break;
if (deltas_valid != roi_retval) break;
}
// Test that we report and error if cyclic refresh is enabled.
cpi.cyclic_refresh_mode_enabled = 1;
roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols, delta_q,
delta_lf, threshold);
roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows, cpi.common.mb_cols,
delta_q, delta_lf, threshold);
EXPECT_EQ(-1, roi_retval) << "cyclic refresh check error";
cpi.cyclic_refresh_mode_enabled = 0;
// Test invalid number of rows or colums.
roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows + 1,
cpi.common.mb_cols, delta_q,
delta_lf, threshold);
roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows + 1,
cpi.common.mb_cols, delta_q, delta_lf, threshold);
EXPECT_EQ(-1, roi_retval) << "MB rows bounds check error";
roi_retval = vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols - 1, delta_q,
delta_lf, threshold);
roi_retval =
vp8_set_roimap(&cpi, roi_map, cpi.common.mb_rows,
cpi.common.mb_cols - 1, delta_q, delta_lf, threshold);
EXPECT_EQ(-1, roi_retval) << "MB cols bounds check error";
}
// Free allocated memory
if (cpi.segmentation_map)
vpx_free(cpi.segmentation_map);
if (roi_map)
vpx_free(roi_map);
if (cpi.segmentation_map) vpx_free(cpi.segmentation_map);
if (roi_map) vpx_free(roi_map);
};
} // namespace

View File

@@ -1,233 +0,0 @@
/*
* Copyright (c) 2013 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <math.h>
#include <stdlib.h>
#include <string.h>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vp8_rtcd.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vpx/vpx_integer.h"
#include "vpx_mem/vpx_mem.h"
namespace {
typedef void (*SixtapPredictFunc)(uint8_t *src_ptr,
int src_pixels_per_line,
int xoffset,
int yoffset,
uint8_t *dst_ptr,
int dst_pitch);
typedef std::tr1::tuple<int, int, SixtapPredictFunc> SixtapPredictParam;
class SixtapPredictTest
: public ::testing::TestWithParam<SixtapPredictParam> {
public:
static void SetUpTestCase() {
src_ = reinterpret_cast<uint8_t*>(vpx_memalign(kDataAlignment, kSrcSize));
dst_ = reinterpret_cast<uint8_t*>(vpx_memalign(kDataAlignment, kDstSize));
dst_c_ = reinterpret_cast<uint8_t*>(vpx_memalign(kDataAlignment, kDstSize));
}
static void TearDownTestCase() {
vpx_free(src_);
src_ = NULL;
vpx_free(dst_);
dst_ = NULL;
vpx_free(dst_c_);
dst_c_ = NULL;
}
virtual void TearDown() {
libvpx_test::ClearSystemState();
}
protected:
// Make test arrays big enough for 16x16 functions. Six-tap filters
// need 5 extra pixels outside of the macroblock.
static const int kSrcStride = 21;
static const int kDstStride = 16;
static const int kDataAlignment = 16;
static const int kSrcSize = kSrcStride * kSrcStride + 1;
static const int kDstSize = kDstStride * kDstStride;
virtual void SetUp() {
width_ = GET_PARAM(0);
height_ = GET_PARAM(1);
sixtap_predict_ = GET_PARAM(2);
memset(src_, 0, kSrcSize);
memset(dst_, 0, kDstSize);
memset(dst_c_, 0, kDstSize);
}
int width_;
int height_;
SixtapPredictFunc sixtap_predict_;
// The src stores the macroblock we will filter on, and makes it 1 byte larger
// in order to test unaligned access. The result is stored in dst and dst_c(c
// reference code result).
static uint8_t* src_;
static uint8_t* dst_;
static uint8_t* dst_c_;
};
uint8_t* SixtapPredictTest::src_ = NULL;
uint8_t* SixtapPredictTest::dst_ = NULL;
uint8_t* SixtapPredictTest::dst_c_ = NULL;
TEST_P(SixtapPredictTest, TestWithPresetData) {
// Test input
static const uint8_t test_data[kSrcSize] = {
216, 184, 4, 191, 82, 92, 41, 0, 1, 226, 236, 172, 20, 182, 42, 226, 177,
79, 94, 77, 179, 203, 206, 198, 22, 192, 19, 75, 17, 192, 44, 233, 120,
48, 168, 203, 141, 210, 203, 143, 180, 184, 59, 201, 110, 102, 171, 32,
182, 10, 109, 105, 213, 60, 47, 236, 253, 67, 55, 14, 3, 99, 247, 124,
148, 159, 71, 34, 114, 19, 177, 38, 203, 237, 239, 58, 83, 155, 91, 10,
166, 201, 115, 124, 5, 163, 104, 2, 231, 160, 16, 234, 4, 8, 103, 153,
167, 174, 187, 26, 193, 109, 64, 141, 90, 48, 200, 174, 204, 36, 184,
114, 237, 43, 238, 242, 207, 86, 245, 182, 247, 6, 161, 251, 14, 8, 148,
182, 182, 79, 208, 120, 188, 17, 6, 23, 65, 206, 197, 13, 242, 126, 128,
224, 170, 110, 211, 121, 197, 200, 47, 188, 207, 208, 184, 221, 216, 76,
148, 143, 156, 100, 8, 89, 117, 14, 112, 183, 221, 54, 197, 208, 180, 69,
176, 94, 180, 131, 215, 121, 76, 7, 54, 28, 216, 238, 249, 176, 58, 142,
64, 215, 242, 72, 49, 104, 87, 161, 32, 52, 216, 230, 4, 141, 44, 181,
235, 224, 57, 195, 89, 134, 203, 144, 162, 163, 126, 156, 84, 185, 42,
148, 145, 29, 221, 194, 134, 52, 100, 166, 105, 60, 140, 110, 201, 184,
35, 181, 153, 93, 121, 243, 227, 68, 131, 134, 232, 2, 35, 60, 187, 77,
209, 76, 106, 174, 15, 241, 227, 115, 151, 77, 175, 36, 187, 121, 221,
223, 47, 118, 61, 168, 105, 32, 237, 236, 167, 213, 238, 202, 17, 170,
24, 226, 247, 131, 145, 6, 116, 117, 121, 11, 194, 41, 48, 126, 162, 13,
93, 209, 131, 154, 122, 237, 187, 103, 217, 99, 60, 200, 45, 78, 115, 69,
49, 106, 200, 194, 112, 60, 56, 234, 72, 251, 19, 120, 121, 182, 134, 215,
135, 10, 114, 2, 247, 46, 105, 209, 145, 165, 153, 191, 243, 12, 5, 36,
119, 206, 231, 231, 11, 32, 209, 83, 27, 229, 204, 149, 155, 83, 109, 35,
93, 223, 37, 84, 14, 142, 37, 160, 52, 191, 96, 40, 204, 101, 77, 67, 52,
53, 43, 63, 85, 253, 147, 113, 226, 96, 6, 125, 179, 115, 161, 17, 83,
198, 101, 98, 85, 139, 3, 137, 75, 99, 178, 23, 201, 255, 91, 253, 52,
134, 60, 138, 131, 208, 251, 101, 48, 2, 227, 228, 118, 132, 245, 202,
75, 91, 44, 160, 231, 47, 41, 50, 147, 220, 74, 92, 219, 165, 89, 16
};
// Expected result
static const uint8_t expected_dst[kDstSize] = {
117, 102, 74, 135, 42, 98, 175, 206, 70, 73, 222, 197, 50, 24, 39, 49, 38,
105, 90, 47, 169, 40, 171, 215, 200, 73, 109, 141, 53, 85, 177, 164, 79,
208, 124, 89, 212, 18, 81, 145, 151, 164, 217, 153, 91, 154, 102, 102,
159, 75, 164, 152, 136, 51, 213, 219, 186, 116, 193, 224, 186, 36, 231,
208, 84, 211, 155, 167, 35, 59, 42, 76, 216, 149, 73, 201, 78, 149, 184,
100, 96, 196, 189, 198, 188, 235, 195, 117, 129, 120, 129, 49, 25, 133,
113, 69, 221, 114, 70, 143, 99, 157, 108, 189, 140, 78, 6, 55, 65, 240,
255, 245, 184, 72, 90, 100, 116, 131, 39, 60, 234, 167, 33, 160, 88, 185,
200, 157, 159, 176, 127, 151, 138, 102, 168, 106, 170, 86, 82, 219, 189,
76, 33, 115, 197, 106, 96, 198, 136, 97, 141, 237, 151, 98, 137, 191,
185, 2, 57, 95, 142, 91, 255, 185, 97, 137, 76, 162, 94, 173, 131, 193,
161, 81, 106, 72, 135, 222, 234, 137, 66, 137, 106, 243, 210, 147, 95,
15, 137, 110, 85, 66, 16, 96, 167, 147, 150, 173, 203, 140, 118, 196,
84, 147, 160, 19, 95, 101, 123, 74, 132, 202, 82, 166, 12, 131, 166,
189, 170, 159, 85, 79, 66, 57, 152, 132, 203, 194, 0, 1, 56, 146, 180,
224, 156, 28, 83, 181, 79, 76, 80, 46, 160, 175, 59, 106, 43, 87, 75,
136, 85, 189, 46, 71, 200, 90
};
uint8_t *src = const_cast<uint8_t*>(test_data);
ASM_REGISTER_STATE_CHECK(
sixtap_predict_(&src[kSrcStride * 2 + 2 + 1], kSrcStride,
2, 2, dst_, kDstStride));
for (int i = 0; i < height_; ++i)
for (int j = 0; j < width_; ++j)
ASSERT_EQ(expected_dst[i * kDstStride + j], dst_[i * kDstStride + j])
<< "i==" << (i * width_ + j);
}
using libvpx_test::ACMRandom;
TEST_P(SixtapPredictTest, TestWithRandomData) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
for (int i = 0; i < kSrcSize; ++i)
src_[i] = rnd.Rand8();
// Run tests for all possible offsets.
for (int xoffset = 0; xoffset < 8; ++xoffset) {
for (int yoffset = 0; yoffset < 8; ++yoffset) {
// Call c reference function.
// Move start point to next pixel to test if the function reads
// unaligned data correctly.
vp8_sixtap_predict16x16_c(&src_[kSrcStride * 2 + 2 + 1], kSrcStride,
xoffset, yoffset, dst_c_, kDstStride);
// Run test.
ASM_REGISTER_STATE_CHECK(
sixtap_predict_(&src_[kSrcStride * 2 + 2 + 1], kSrcStride,
xoffset, yoffset, dst_, kDstStride));
for (int i = 0; i < height_; ++i)
for (int j = 0; j < width_; ++j)
ASSERT_EQ(dst_c_[i * kDstStride + j], dst_[i * kDstStride + j])
<< "i==" << (i * width_ + j);
}
}
}
using std::tr1::make_tuple;
INSTANTIATE_TEST_CASE_P(
C, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_c),
make_tuple(8, 8, &vp8_sixtap_predict8x8_c),
make_tuple(8, 4, &vp8_sixtap_predict8x4_c),
make_tuple(4, 4, &vp8_sixtap_predict4x4_c)));
#if HAVE_NEON
INSTANTIATE_TEST_CASE_P(
NEON, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_neon),
make_tuple(8, 8, &vp8_sixtap_predict8x8_neon),
make_tuple(8, 4, &vp8_sixtap_predict8x4_neon)));
#endif
#if HAVE_MMX
INSTANTIATE_TEST_CASE_P(
MMX, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_mmx),
make_tuple(8, 8, &vp8_sixtap_predict8x8_mmx),
make_tuple(8, 4, &vp8_sixtap_predict8x4_mmx),
make_tuple(4, 4, &vp8_sixtap_predict4x4_mmx)));
#endif
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(
SSE2, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_sse2),
make_tuple(8, 8, &vp8_sixtap_predict8x8_sse2),
make_tuple(8, 4, &vp8_sixtap_predict8x4_sse2)));
#endif
#if HAVE_SSSE3
INSTANTIATE_TEST_CASE_P(
SSSE3, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_ssse3),
make_tuple(8, 8, &vp8_sixtap_predict8x8_ssse3),
make_tuple(8, 4, &vp8_sixtap_predict8x4_ssse3),
make_tuple(4, 4, &vp8_sixtap_predict4x4_ssse3)));
#endif
#if HAVE_MSA
INSTANTIATE_TEST_CASE_P(
MSA, SixtapPredictTest, ::testing::Values(
make_tuple(16, 16, &vp8_sixtap_predict16x16_msa),
make_tuple(8, 8, &vp8_sixtap_predict8x8_msa),
make_tuple(8, 4, &vp8_sixtap_predict8x4_msa),
make_tuple(4, 4, &vp8_sixtap_predict4x4_msa)));
#endif
} // namespace

141
test/stress.sh Executable file
View File

@@ -0,0 +1,141 @@
#!/bin/sh
##
## Copyright (c) 2016 The WebM project authors. All Rights Reserved.
##
## Use of this source code is governed by a BSD-style license
## that can be found in the LICENSE file in the root of the source
## tree. An additional intellectual property rights grant can be found
## in the file PATENTS. All contributing project authors may
## be found in the AUTHORS file in the root of the source tree.
##
## This file performs a stress test. It runs 5 encodes and 30 decodes in
## parallel.
. $(dirname $0)/tools_common.sh
YUV="${LIBVPX_TEST_DATA_PATH}/niklas_1280_720_30.yuv"
VP8="${LIBVPX_TEST_DATA_PATH}/tos_vp8.webm"
VP9="${LIBVPX_TEST_DATA_PATH}/vp90-2-sintel_1920x818_tile_1x4_fpm_2279kbps.webm"
DATA_URL="http://downloads.webmproject.org/test_data/libvpx/"
SHA1_FILE="$(dirname $0)/test-data.sha1"
# Set sha1sum to proper sha program (sha1sum, shasum, sha1). This code is
# cribbed from libs.mk.
[ -x "$(which sha1sum)" ] && sha1sum=sha1sum
[ -x "$(which shasum)" ] && sha1sum=shasum
[ -x "$(which sha1)" ] && sha1sum=sha1
# Download a file from the url and check its sha1sum.
download_and_check_file() {
# Get the file from the file path.
local readonly root="${1#${LIBVPX_TEST_DATA_PATH}/}"
# Download the file using curl. Trap to insure non partial file.
(trap "rm -f $1" INT TERM \
&& eval "curl --retry 1 -L -o $1 ${DATA_URL}${root} ${devnull}")
# Check the sha1 sum of the file.
if [ -n "${sha1sum}" ]; then
set -e
grep ${root} ${SHA1_FILE} \
| (cd ${LIBVPX_TEST_DATA_PATH}; ${sha1sum} -c);
fi
}
# Environment check: Make sure input is available.
stress_verify_environment() {
if [ ! -e "${SHA1_FILE}" ] ; then
echo "Missing ${SHA1_FILE}"
return 1
fi
for file in "${YUV}" "${VP8}" "${VP9}"; do
if [ ! -e "${file}" ] ; then
download_and_check_file "${file}"
fi
done
if [ ! -e "${YUV}" ] || [ ! -e "${VP8}" ] || [ ! -e "${VP9}" ] ; then
elog "Libvpx test data must exist in LIBVPX_TEST_DATA_PATH."
return 1
fi
if [ -z "$(vpx_tool_path vpxenc)" ]; then
elog "vpxenc not found. It must exist in LIBVPX_BIN_PATH or its parent."
return 1
fi
if [ -z "$(vpx_tool_path vpxdec)" ]; then
elog "vpxdec not found. It must exist in LIBVPX_BIN_PATH or its parent."
return 1
fi
}
# This function runs tests on libvpx that run multiple encodes and decodes
# in parallel in hopes of catching synchronization and/or threading issues.
stress() {
local readonly decoder="$(vpx_tool_path vpxdec)"
local readonly encoder="$(vpx_tool_path vpxenc)"
local readonly codec="$1"
local readonly webm="$2"
local readonly decode_count="$3"
local pids=""
local rt_max_jobs=${STRESS_RT_MAX_JOBS:-5}
local twopass_max_jobs=${STRESS_TWOPASS_MAX_JOBS:-5}
# Enable job control, so we can run multiple processes.
set -m
# Start $twopass_max_jobs encode jobs in parallel.
for i in $(seq ${twopass_max_jobs}); do
bitrate=$(($i * 20 + 300))
eval "${VPX_TEST_PREFIX}" "${encoder}" "--codec=${codec} -w 1280 -h 720" \
"${YUV}" "-t 4 --limit=150 --test-decode=fatal " \
"--target-bitrate=${bitrate} -o ${VPX_TEST_OUTPUT_DIR}/${i}.webm" \
${devnull} &
pids="${pids} $!"
done
# Start $rt_max_jobs rt encode jobs in parallel.
for i in $(seq ${rt_max_jobs}); do
bitrate=$(($i * 20 + 300))
eval "${VPX_TEST_PREFIX}" "${encoder}" "--codec=${codec} -w 1280 -h 720" \
"${YUV}" "-t 4 --limit=150 --test-decode=fatal " \
"--target-bitrate=${bitrate} --lag-in-frames=0 --error-resilient=1" \
"--kf-min-dist=3000 --kf-max-dist=3000 --cpu-used=-6 --static-thresh=1" \
"--end-usage=cbr --min-q=2 --max-q=56 --undershoot-pct=100" \
"--overshoot-pct=15 --buf-sz=1000 --buf-initial-sz=500" \
"--buf-optimal-sz=600 --max-intra-rate=900 --resize-allowed=0" \
"--drop-frame=0 --passes=1 --rt --noise-sensitivity=4" \
"-o ${VPX_TEST_OUTPUT_DIR}/${i}.rt.webm" ${devnull} &
pids="${pids} $!"
done
# Start $decode_count decode jobs in parallel.
for i in $(seq "${decode_count}"); do
eval "${decoder}" "-t 4" "${webm}" "--noblit" ${devnull} &
pids="${pids} $!"
done
# Wait for all parallel jobs to finish.
fail=0
for job in "${pids}"; do
wait $job || fail=$(($fail + 1))
done
return $fail
}
vp8_stress_test() {
local vp8_max_jobs=${STRESS_VP8_DECODE_MAX_JOBS:-40}
if [ "$(vp8_decode_available)" = "yes" -a \
"$(vp8_encode_available)" = "yes" ]; then
stress vp8 "${VP8}" "${vp8_max_jobs}"
fi
}
vp9_stress_test() {
local vp9_max_jobs=${STRESS_VP9_DECODE_MAX_JOBS:-25}
if [ "$(vp9_decode_available)" = "yes" -a \
"$(vp9_encode_available)" = "yes" ]; then
stress vp9 "${VP9}" "${vp9_max_jobs}"
fi
}
run_tests stress_verify_environment "vp8_stress_test vp9_stress_test"

113
test/sum_squares_test.cc Normal file
View File

@@ -0,0 +1,113 @@
/*
* Copyright (c) 2016 The WebM project authors. All Rights Reserved.
*
* Use of this source code is governed by a BSD-style license
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
* in the file PATENTS. All contributing project authors may
* be found in the AUTHORS file in the root of the source tree.
*/
#include <cmath>
#include <cstdlib>
#include <string>
#include "third_party/googletest/src/include/gtest/gtest.h"
#include "./vpx_config.h"
#include "./vpx_dsp_rtcd.h"
#include "test/acm_random.h"
#include "test/clear_system_state.h"
#include "test/register_state_check.h"
#include "test/util.h"
#include "vpx_ports/mem.h"
using libvpx_test::ACMRandom;
namespace {
const int kNumIterations = 10000;
typedef uint64_t (*SSI16Func)(const int16_t *src, int stride, int size);
typedef std::tr1::tuple<SSI16Func, SSI16Func> SumSquaresParam;
class SumSquaresTest : public ::testing::TestWithParam<SumSquaresParam> {
public:
virtual ~SumSquaresTest() {}
virtual void SetUp() {
ref_func_ = GET_PARAM(0);
tst_func_ = GET_PARAM(1);
}
virtual void TearDown() { libvpx_test::ClearSystemState(); }
protected:
SSI16Func ref_func_;
SSI16Func tst_func_;
};
TEST_P(SumSquaresTest, OperationCheck) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
DECLARE_ALIGNED(16, int16_t, src[256 * 256]);
const int msb = 11; // Up to 12 bit input
const int limit = 1 << (msb + 1);
for (int k = 0; k < kNumIterations; k++) {
const int size = 4 << rnd(6); // Up to 128x128
int stride = 4 << rnd(7); // Up to 256 stride
while (stride < size) { // Make sure it's valid
stride = 4 << rnd(7);
}
for (int i = 0; i < size; ++i) {
for (int j = 0; j < size; ++j) {
src[i * stride + j] = rnd(2) ? rnd(limit) : -rnd(limit);
}
}
const uint64_t res_ref = ref_func_(src, stride, size);
uint64_t res_tst;
ASM_REGISTER_STATE_CHECK(res_tst = tst_func_(src, stride, size));
ASSERT_EQ(res_ref, res_tst) << "Error: Sum Squares Test"
<< " C output does not match optimized output.";
}
}
TEST_P(SumSquaresTest, ExtremeValues) {
ACMRandom rnd(ACMRandom::DeterministicSeed());
DECLARE_ALIGNED(16, int16_t, src[256 * 256]);
const int msb = 11; // Up to 12 bit input
const int limit = 1 << (msb + 1);
for (int k = 0; k < kNumIterations; k++) {
const int size = 4 << rnd(6); // Up to 128x128
int stride = 4 << rnd(7); // Up to 256 stride
while (stride < size) { // Make sure it's valid
stride = 4 << rnd(7);
}
const int val = rnd(2) ? limit - 1 : -(limit - 1);
for (int i = 0; i < size; ++i) {
for (int j = 0; j < size; ++j) {
src[i * stride + j] = val;
}
}
const uint64_t res_ref = ref_func_(src, stride, size);
uint64_t res_tst;
ASM_REGISTER_STATE_CHECK(res_tst = tst_func_(src, stride, size));
ASSERT_EQ(res_ref, res_tst) << "Error: Sum Squares Test"
<< " C output does not match optimized output.";
}
}
using std::tr1::make_tuple;
#if HAVE_SSE2
INSTANTIATE_TEST_CASE_P(
SSE2, SumSquaresTest,
::testing::Values(make_tuple(&vpx_sum_squares_2d_i16_c,
&vpx_sum_squares_2d_i16_sse2)));
#endif // HAVE_SSE2
} // namespace

Some files were not shown because too many files have changed in this diff Show More