Commit Graph

16568 Commits

Author SHA1 Message Date
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