* changes:
hadamard highbd ssse3: use tran_low_t for coeff
hadamard highbd neon: use tran_low_t for coeff
hadamard highbd sse2: use tran_low_t for coeff
This commit resolves the compression performance regression in
real-time encoding setting when high bit-depth mode is enabled.
The current solution temporarily disables the SIMD implementations
of vpx_satd, hadamard8x8, and hadamard16x16 in high bit-depth mode.
The commit makes the coding results bit-wise identical between
regular coding pipeline and high bit-depth at profile 0.
BUG=webm:1365
Change-Id: Icfb900821733749685370460a1a5a7e07f76f4bf
Clears a clang static analyzer warning where 'cols' is assumed to be
less than 0, preventing the for loop from executing.
The assembly already requires that the size be 8 or 16 (U/V or Y plane)
and cols is a multiple of 8.
Change-Id: Ica4612690ead1638c94cfe56b306e87f8ce644f9
This commit reworks the SSSE3 implementation of the forward 8x8
2D-DCT. It uses a cyclic rotation approach to the temporary xmm
registers. It reduces the average cycles from 158 to 154. The SSE2
version uses 169 cycles.
Change-Id: I1b79b9642aae0ed3fb3cefb5b70246e6de5d5caa
This was much more amenable to optimization than the across filter.
Speedup of almost 2.5x
BUG=webm:1320
Change-Id: I49acc0f9cb2e7642303df90132cbc938acade4c4
Speed test shows 25% gain on vpx_idct16x16_256_add_neon(),
and vpx_idct16x16_10_add_neon() got trippled.
Change-Id: If8518d9b6a3efab74031297b8d40cd83c4a49541
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
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
Implement vpx_post_proc_down_and_across_mb_row in NEON.
Runs about 6-7x faster than C.
BUG=webm:1320
Change-Id: Ic5c7d3552a88cfcf999ec5bf2bd46fee460642c2
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
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
* 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
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
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
dst += stride behaving better with gcc/clang
Expanding inline function dc_SIZExSIZE() save intructions for
vpx_dc_predictor_SIZExSIZE_neon().
Change-Id: Id0ccbd58b6a31df539141fd33bdf28633339150d
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
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)
Updated code to process in 8bit as saturation/clipping takes care of
overflow
Removed unused macro
Change-Id: I113df60286fb28b216df800d95b2d3695ef71440
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
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
these functions are incompatible currently and unreferenced in rtcd,
exclude them from the build.
BUG=webm:1294
Change-Id: I7790c195a91e1b142f56c04d2a5e305d9133b896
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
to get_binary_prob(). the only other caller mode_mv_merge_probs() does
its own test on 0.
BUG=chromium:639712
Change-Id: I1178688706baeca2883f7aadbc254abb219a44ce
+ 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
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
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
removes some unnecessary casts and adds a few explicit uint32 ones for
larger sizes to quiet -Wshorten-64-to-32 warnings
Change-Id: I63c5fce8e62c426d5cf5c10a66a113c119a43518
assume __clang_major__==0 has the latest version of
_mm256_broadcastsi128_si256. fixes builds with custom clang toolchains.
BUG=b/30970831
Change-Id: I90becd56278e4716bd46e2ba9d910af977e8dfa6
* 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__
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
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
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
* 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
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
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
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
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
Due to rounding, hbd variance may become negative. This commit put in
check and clamp of negative values to 0.
Change-Id: I610d9c8aa2d4eebe7bc5f2c5624a9e3cadad4c94
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
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: I37e3e9c5694737d9134a6bce6698d3e43f8fc962
Due to rounding used computation, HDB variance computation may produce
slightly negative values. This commit adds clamping to make sure
output variance values for 10 and 12 to be non-negative.
Change-Id: Id679aa55a4c201958c4c7d28cd8733b9246a71c8
development has moved to the nextgenv2 branch and a snapshot from here
was used to seed aomedia
BUG=b/29457125
Change-Id: Iedaca11ec7870fb3a4e50b2c9ea0c2b056a0d3c0
This commit change to use int64_t to represent the sum of pixel
differences, which can be negative.
This fixes a number of ubsan warnings.
BUG=webm:1219
Change-Id: I885f245ae895ab92ca5f3b9848d37024b07aac98
Provides more comprehensive coverage for --enable-coefficient-checking.
The intent is to make the --enable-coefficient-checking option
consistent with the VP9 spec.
Change-Id: I12d0120756d17572ca2b2d7e6a2ab9d8071d8d58
Replaced vpx_d45_predictor_4x4_ssse3(), vpx_d45_predictor_8x8_ssse3()
and vpx_d207_predictor_4x4_ssse3() with
created vpx_d45_predictor_4x4_sse2(), vpx_d45_predictor_8x8_sse2()
and vpx_d207_predictor_4x4_sse2() respectively.
It's mostly neutral or slightly worse than ssse3 in good cases and
better than ssse3 in the bad cases (but still worse than using the mmx
regs).
Change-Id: Ib0237ceb71d2c57b8a93fd3170330cfed9d56bdd
Followed the code style of other lpf fuctions.
These 2 functions put 2 rows of data in a single xmm register,
so they have similar but not identical filter operations,
and cannot share the same macros.
Change-Id: I3bab55a5d1a1232926ac8fd1f03251acc38302bc
This reverts commit 2468163e07.
causes valgrind errors for overread of buffer in SubpelVarianceTest
Change-Id: I448e52c76f815ac199305b71f7d169f2bc167679
This commit clarifies integer value range for vairables used in
several variance functions, also change to use proper type
conversion to reflect the value ranges.
Change-Id: Ic3234b83a912ce1ad12d1b254f3378763e15cc5c
In motion estimation stage for subpel motion, subpel variance is
computed use bilinear interpolation. The motion vector precision
used is at 1/8 pel and three bits are used to represent the x and y
subpel offsets. Based on this, the half pel check should be against
4, not 8.
Change-Id: I1f56fa1fa3f2f5e19a20d27983efe628557f170e
there are sse2 equivalents which is a reasonable modern baseline
Removed mmx variance functions:
vpx_get_mb_ss_mmx()
vpx_get8x8var_mmx()
vpx_get4x4var_mmx()
vpx_variance4x4_mmx()
vpx_variance8x8_mmx()
vpx_mse16x16_mmx()
vpx_variance16x16_mmx()
vpx_variance16x8_mmx()
vpx_variance8x16_mmx()
Change-Id: Iffaf85344c6676a3dd337c0645a2dd5deb2f86a1
This change makes the c match the assembly and removes the todo's
associated with getting this to work.
Change-Id: Ie32e9ebb584a9d60399662d8bcb71b74fbd19d1e
* changes:
vp9_frame_scale_ssse3.c: make 2 functions static
vp9_pickmode.c: make function static
vp9_noise_estimate.c: make function static
vp9_aq_360.c: add missing include
vp9_idct_intrin_sse2: add missing vp9_rtcd.h include
vpx_dsp/*.[hc]: add missing vpx_dsp_rtcd.h include
In so doing this fixes a couple of bugs:
vpx_plane_add_noise.c needed to subtract a clamp instead of add.
And the assembly (mmx sse) had assumptions that parameters were
continuous in memory which was not true.
Change-Id: I76f2c43cf54bfc838eb2edf8a443eaaa7565d7b5
The product always fits in uint32_t, but the operands don't.
An optimizing compiler should generate the wraparound code.
(Verified with clang).
Change-Id: I25eb64df99152992bc898b8ccbb01d55c8d16e3c
This was never hooked up for the 32x32_34 case as the neon_asm version
in 3f7c12da, when the intrinsics version was added.
Change-Id: Ic7db4ce5850c637315f9fe9e2de93a4f8cf9e320
only output[0] needs to be set, store_output is more involved than a
movdqa in the high bitdepth case
Change-Id: I2cbd85d7cf74688bdf47eb767934fe42e02bff67
replace with vpx_highbd_lpf_horizontal_edge_16 and
vpx_highbd_lpf_horizontal_edge_8 to avoid passing a count parameter
Change-Id: I551f8cec0fce57032cb2652584bb802e2248644d
replace with vpx_lpf_horizontal_edge_16 and vpx_lpf_horizontal_edge_8 to
avoid passing a count parameter
Change-Id: I848c95c02a3c6ebaa6c2bdf0983dce05cd645271
the results along the top and left border are then stored with a moving
window into the vector.
~40-67% faster on ARM, ~40-77+% on x86 depending on the block size.
Change-Id: Iab369aa2946a3ae4eb7290d512868fe5db92dbc8
This commit changes SSSE3 optimized idct8x8 functions to work with
highbitdepth build.
With this commit and the previous one that enabled SSSE3 idct32x32
functions, tests showed virtually no difference on decoding speed for
file fdJc1_IBKJA.248.webm for the build with -enable-vp9-highbitdpeth
option and the build without the option.
Change-Id: Ibe0634149ec70e8b921e6b30171664b8690a9c45
This commit changes the SSSE3 assembly functions for idct32x32 to
support highbitdepth build.
On test clip fdJc1_IBKJA.248.webm, this cuts the speed difference
between hbd and lbd build from between 3-4% to 1-2%.
Change-Id: Ic3390e0113bc1ca5bba8ec80d1795ad31b484fca
This reverts commit ea48370a50, reversing
changes made to 15939cb2d7.
The commit was insufficiently tested and causes failures.
Change-Id: I623d6fc2cd3ae6fd42d0abab1f8eada465ae57a7
reduce the register count by 1 to avoid xmm6 and unnecessarily
penalizing the other users of the base macro
Change-Id: I59605c9a41a31c1b74f67ec06a40d1a7f92c4699
In 32-bit build with --enable-shared, there is a lot of
register pressure and register src_strideq is reused.
The code needs to use the stack based version of src_stride,
but this doesn't compile when used in an lea instruction.
This patch also fixes a related segmentation fault caused by the
implementation using src_strideq even though it has been
reused.
This patch also fixes the HBD subpel variance tests that fail
when compiled without disable-optimizations.
These failures were caused by local variables in the assembler
routines colliding with the caller's stack frame.
Change-Id: Ice9d4dafdcbdc6038ad5ee7c1c09a8f06deca362
Reallocate the xmm register usage so that no ARCH_X86_64 required.
Reduce memory access to the left neighbor by half.
Speed up by single digit on big core machine.
Change-Id: I392515ed8e8aeb02e6a717b3966b1ba13f5be990
GET_GOT modifies the stack pointer so the offset for left's address will
be wrong if loaded afterword.
Change-Id: Iff9433aec45f5f6fe1a59ed8080c589bad429536
Relocate the function from SSSE3 to SSE2, Unroll loop from 16 to 8,
and reduce mem access to left.
Speed up by single digit in ./test_intra_pred_speed on big core
machines.
Change-Id: I2b7fc95ffc0c42145be2baca4dc77116dff1c960
4x4 Intra predictor implemented with MMX is replaced with SSE2.
Segfault in change 315561 when decoding vp8 is taken care of.
Change-Id: I083a7cb4eb8982954c20865160f91ebec777ec76
Relocate the function from SSSE3 to SSE2, Unroll loop from 8 to 4,
and reduce mem access to left.
Speed up by >20% in ./test_intra_pred_speed.
Change-Id: Ie48229c2e32404706b722442942c84983bda74cc
Relocate the function from SSSE3 to SSE2, Unroll loop from 4 to 2,
and reduce mem access to left.
Speed up by >20% in ./test_intra_pred_speed.
Change-Id: Ib9f1846819783b6e05e2a310c930eb844b2b4d2e
Reduce mem access to left. Speed up by 10% in ./test_intra_pred_speed
with the same instruction size.
Change-Id: Ia33689d62476972cc82ebb06b50415aeccc95d15
Relocate h_predictor_4x4 from SSSE3 to SSE2 with XMM registers.
Speed up by ~25% in ./test_intra_pred_speed.
Change-Id: I64e14c13b482a471449be3559bfb0da45cf88d9d
Always round sum error and sum square error toward zero in variance
calculations. This prevents variance from becoming negative.
Avoiding rounding variance at all might be better but would be far
more invasive.
Change-Id: Icf24e0e75ff94952fc026ba6a4d26adf8d373f1c
tm_predictor_4x4 is implemented with SSE2 using XMM registers.
Speed up by ~25% in ./test_intra_pred_speed.
Change-Id: I25074b78d476a2cb17f81cf654bdfd80df2070e0
Added optimization of the 8 bit assembly quantizer routines. This makes
these functions up to 100% faster, depending on encoding parameters.
This patch maskes the encoder faster in both the high bitdepth and 8bit
configurations. In the high bitdepth configuration, it effects profile 0
only.
Based on my profiling using 1080p input the net gain is between 1-3% for
the 8 bit config, and around 2.5-4.5% for the high bitdepth config,
depending on target bitrate. The difference between the 8 bit and high
bitdepth configurations for the same encoder run is reduced by 1% in all
cases I have profiled.
Change-Id: I86714a6b7364da20cd468cd784247009663a5140
This experiment allows using full above/right edges for all transform
sizes whenever available (for d45/d63), and adds bottom/left edges for
d207.
See issue 1043.
Change-Id: I5cf7f345e783e8539bb6b6d2c9972fb1d6d0a78b
When configured with high bit detpth enabled, the 8bit quantize
function stopped using optimised code. This made 8bit content
decode slowly. This commit re-enables the SSSE3 optimisations.
Change-Id: I194b505dd3f4c494e5c5e53e020f5d94534b16b5
When configured with high bit detpth enabled, the 8bit quantize
function stopped using optimised code. This made 8bit content
decode slowly. This commit re-enables the SSE2 optimisation
(but not the SSSE3 optimisation).
Change-Id: Id015fe3c1c44580a4bff3f4bd985170f2806a9d9
This is more a proof of concept than anything else. The problem here
isn't so much how to code it, but rather where to place the resulting
code. All intrapred DSP code lives in vpx_dsp, so do we want the vp10
specific intra pred functions to live there, or in vp10/?
See issue 1015.
Change-Id: I675f7badcc8e18fd99a9553910ecf3ddf81f0a05
I've added a few new functions (d45e, d63e, he, ve) to cover the
filtered h/v 4x4 predictors that are vp8-specific, the "correct"
d45 with the correctly filtered bottom-right pixel (as opposed to
the unfiltered version in vp9), and the "broken" d63 with weirdly
filtered bottom-right pixels (which is correctly filtered in vp9).
There may be a minor performance impact on all systems because we
have to do an extra copy of the Above pixel array to incorporate
the topleft pixel in the same array (thus fitting the vpx_dsp API).
In addition, armv6 will have a more serious performance impact b/c
I removed the armv6/vp8-specific assembly. I'm not sure anyone
cares...
Change-Id: I7f9e5ebee11d8e21aca2cd517a69eefc181b2e86
When configured with high bitdepth enabled, the 8bit transform
stopped using optimised code. This made 8bit content decode slowly.
Change-Id: I67d91f9b212921d5320f949fc0a0d3f32f90c0ea
fixes build on windows x64; previously 'heightq' i.e., the 64-bit register
was accessed when only the 32-bit value was needed. given this is from a
stack variable the upper bits were undefined.
+ bump register/xmm counts; users of SETUP_LOCAL_VARS touch xmm13 in
64-bit builds and filter_block1d16_v* uses one extra temp variable
Change-Id: I9c768c0b2047481d1d3b11c2e16b2f8de6eb0d80
For reading, this makes the operation branchless, although it still
requires two shifts. For writing, this makes the operation as fast
as writing an unsigned value, branchlessly. This is also how other
codecs typically code signed, non-arithcoded bitstream elements.
See issue 1039.
Change-Id: I6a8182cc88a16842fb431688c38f6b52d7f24ead
This is based on the original patch optimized for 32bit
platforms by Tamar/Ilya and now uses the x86inc style asm.
The assembly was also modified to support 64bit platforms.
Change-Id: Ice12f249bbbc162a7427e3d23fbf0cbe4135aff2
* changes:
Only build ssse3 filter functions on 64 bit
Clean up unused function warnings in vp8 encoder
Clean up unused function warnings in vp8 onyx_if.c
These were lost in the great sub pixel variance move of
6a82f0d7fb
Not having these functions caused a ~10% performance regression in
some realtime vp8 encodes.
Change-Id: I50658483d9198391806b27899f2c0d309233c4b5
