GitLab

Allow selectively building just the intrinsics for armv8

Change-Id: I2f29b2e4508b8b8e5649c2906b3159ad1d4ec477

Some version of clang refuse 'asm volatile'

Change-Id: I79d909ac8ae3c04b608f20c6f10fa79b7f9fc8e0

This reverts commit e516a425

Change-Id: I7c78712acc737ad5f580181cdab3aa76b23f3ca5

Both GCC and RVCT/ARMCC support __builtin_prefetch, but MSVC
doesn't.

Change-Id: I44e1eecead61bc88d8fdfd3fef03d76d4f5afe08

This fixes building with MSVC.

Change-Id: I763ba8855c8083d82c8b477d3a297e310e93a335

This reverts commit c500fc22

There is an issue with gcc 4.6 in the Android NDK:
loopfiltersimpleverticaledge_neon.c: In function 'vp8_loop_filter_bvs_neon':
loopfiltersimpleverticaledge_neon.c:176:1: error: insn does not satisfy its constraints:

Change-Id: I95b6509d12f075890308914cc691b813d2e5cd9f

This reverts commit a5d79f43

There is an issue with gcc 4.6 in the Android NDK:
loopfilter_neon.c: In function 'vp8_loop_filter_vertical_edge_y_neon':
loopfilter_neon.c:394:1: error: insn does not satisfy its constraints:

Change-Id: I2b8c6ee3fa595c152ac3a5c08dd79bd9770c7b52

Added macro to conditionally compile some of the
post-processing functions only when CONFIG_POSTPROC
is defined.

This was causing the build for the generic-gnu target
to fail.

Change-Id: Ibfa447feceb7a0528135025f105be48f97e9965c

The rounding of the ARNR filter output prior to
normalization by the filter strength was incorrect
when strength = 0.

In this case 1 << (strength - 1) would not create the
required rounding of 0, rather it would outrange. This
patch fixes this issue.

Change-Id: I771809ba34d6052b17d34c870ea11ff67b418dab

This fixes compilation with MSVC.

Change-Id: Iaf7d6b0a0134968a6addf315fde6d852f298db8c

MSVC doesn't support casting neon vector types but requires using
vreinterpret.

Change-Id: Iece0bf5632567efd7f37f527abea38afeab4926d

Add variance_neon.c
- vp8_variance16x16_neon
- vp8_variance16x8_neon
- vp8_variance8x16_neon
- vp8_variance8x8_neon

Change-Id: Idfb9c96134a1c6a696a98ce68b4f7ed593a00660
Signed-off-by: James Yu <james.yu@linaro.org>

Add idct_dequant_0_2x_neon.c
- idct_dequant_0_2x_neon

Change-Id: I8e129172ef1b2517cf72ff267788921f1a792586
Signed-off-by: James Yu <james.yu@linaro.org>

Add sixtappredict_neon.c
- vp8_sixtap_predict16x16_neon
- vp8_sixtap_predict8x8_neon
- vp8_sixtap_predict8x4_neon
- vp8_sixtap_predict4x4_neon

Change-Id: I3b02fce48ae2e6c6099041ba5ddd7b090f1463b9
Signed-off-by: James Yu <james.yu@linaro.org>

Add shortidct4x4llm_neon.c
- vp8_short_idct4x4llm_neon

Change-Id: I5a734bbffca8dacf8633c2b0ff07b98aa2f438ba
Signed-off-by: James Yu <james.yu@linaro.org>

Add sad_neon.c
- vp8_sad16x16_neon
- vp8_sad16x8_neon
- vp8_sad8x8_neon
- vp8_sad8x16_neon
- vp8_sad4x4_neon

Change-Id: I08eaae49ec03fb91b394354660a5df0367cea311
Signed-off-by: James Yu <james.yu@linaro.org>

Add mbloopfilter_neon.c
- vp8_mbloop_filter_horizontal_edge_y_neon
- vp8_mbloop_filter_horizontal_edge_uv_neon
- vp8_mbloop_filter_vertical_edge_y_neon
- vp8_mbloop_filter_vertical_edge_uv_neon

Change-Id: Ia9084e0892d4d49412d9cf2b165a0f719f2382d7
Signed-off-by: James Yu <james.yu@linaro.org>

Add loopfiltersimpleverticaledge_neon.c
- vp8_loop_filter_bvs_neon
- vp8_loop_filter_mbvs_neon

Change-Id: I7cf0a161ad4ae37c881b94cc0122f895d3baae79
Signed-off-by: James Yu <james.yu@linaro.org>

Add loopfiltersimplehorizontaledge_neon.c
- vp8_loop_filter_bhs_neon
- vp8_loop_filter_mbhs_neon

Change-Id: I77f9721b20585da8bf3869a3850ff0ae4b4bfeea
Signed-off-by: James Yu <james.yu@linaro.org>

This eliminates the asm_offsets dependency for future
all-assembly versions of this function.

Change-Id: I3227073ecfcb8ee6e593934fab941e9081abdda0

Add loopfilter_neon.c
- vp8_loop_filter_horizontal_edge_y_neon
- vp8_loop_filter_horizontal_edge_uv_neon
- vp8_loop_filter_vertical_edge_y_neon
- vp8_loop_filter_vertical_edge_uv_neon

Change-Id: I50b57dedabd42d2a3c183c1738cc5346f0e71ed8
Signed-off-by: James Yu <james.yu@linaro.org>

Add iwalsh_neon.c
- vp8_short_inv_walsh4x4_neon

Change-Id: I8beda6ce11ad8ce9e80cc0a38d40161938359162
Signed-off-by: James Yu <james.yu@linaro.org>

Replace it with some intrinsic code and inline assembly.

Change-Id: I81b4df146db3d01039059be7dae31083e2943b97

Add idct_dequant_full_2x_neon.c
- idct_dequant_full_2x_neon

==== Summary of apply VP8 decode patch series ====
Benchmark on Samsung Chromebook, Cortex-A15, 1.7GHz, Dual core
Toolchain: linaro-1.13.1-4.8-2014.01
Compile argument: CROSS=arm-linux-gnueabihf- ../libvpx/configure
                     --target=armv7-linux-gcc --prefix=$HOME/out
                     --enable-shared --cpu=cortex-a7
Test argument: vpxdec --summary --noblit ./tears_of_steel_1080p.webm

NEON assembly   46.68 (fps)
Apply patch 06  46.65, -0.03
Apply patch 07  46.86, +0.21
Apply patch 08  46.58, -0.28
Apply patch 09  46.57, -0.01
Apply patch 10  46.51, -0.06
Apply patch 11  46.13, -0.38
Apply patch 12  45.42, -0.71
Apply patch 13  46.06, +0.64
Apply patch 14  45.19, -0.87
Apply patch 15  45.93, +0.74
Apply patch 16  45.48, -0.45
Apply patch 17  45.84, +0.36
Apply patch 18  45.91, +0.07  <= With all NEON intrinsics patches
                 Total -0.77 fps, 1.65% performance regression

Change-Id: I77bfc9eaccfb97b8d401e949ceff8795e26ca6b7
Signed-off-by: James Yu <james.yu@linaro.org>

Used horizonal add instructions instead of adding
byte lanes.  The encoder performance improved by
~4% for the test clip used.

Change-Id: Iaddd10403fcffb5b3f53b1f591ab2fe0ff002c08

This patch did a cleanup following the commit "Save NEON registers
in VP8 NEON functions". The pushing/poping of callee-saved NEON
registers was moved into individual NEON functions. Therefore,
we don't need to save those registers at the beginning of codec.
The related code was removed.

Change-Id: I5648166514fc9beffb780aa138495597731f49ea

The recent compiler can generate optimized code that uses NEON registers
for various operations besides floating-point operations. Therefore,
only saving callee-saved registers d8 - d15 at the beginning of the
encoder/decoder is not enough anymore. This patch added register saving
code in VP8 NEON functions that use those registers.

Change-Id: Ie9e44f5188cf410990c8aaaac68faceee9dffd31

Change-Id: I88f86c8ff9af34e0b6531028b691921b54c2fc48

This patch fixed errors reported in Issue 746: "dr memory VP8
encode errors" and Issue 745: "dr memory VP8 decode errors".
The "UNINITIALIZED READ" errors were fixed in x86 assembly
code. The list of files fixed is
vp8_intra_pred_uv_tm_sse2
vp8_intra_pred_uv_tm_ssse3

vp8_intra_pred_uv_ho_mmx2
vp8_intra_pred_uv_ho_ssse3

vp8_intra_pred_y_tm_sse2
vp8_intra_pred_y_tm_ssse3

vp8_intra_pred_y_ho_sse2

Change-Id: Ib6df7bf1d442077fe534edfd90e50ad16fadacdd

This commit added a check of reference frame to make sure that pre
buffer pointers are initialized only when necessary and make them
to 0 if ref frame is intra, hence those buffer should never be used.

Change-Id: Ieb474fcd9feb759f02e2f9c282b7348a8fa31117

Added command line flags "resize-width" & "resize-height"
to allow the user to specify the frame size to encode at.

These two flags are ignored if the "resize-allowed" switch
is not set to 1.

All frames in the clip are then encoded at this size, which
must be smaller than the raw frame size.

Change-Id: I3d64bd9303d5c0bd678461a866a1ea621700d744

Change-Id: Ib9f7fd3fd56e304e5f587f790c97ac34a3077265

This fixes warning when compiling libvpx for PNaCl. PNaCl's version
of math.h defines M_LOG2_E.

Change-Id: Iba9450441538e9f82447ad2936bea94d21bafdf1

Change-Id: Ie0eb9ac4529db00a322511e5241a59b501c289b7

This patch fixed WebRTC Issue 3020: "Uninit error at
vp8_mbpost_proc_down_xmm". The first 8 values in d were not initialized,
but was accessed. This patch fixed c code as well as mmx and sse2 code.

Change-Id: Iaa5b41a4ed3bea971b15fb826ce34b7ab4e36fb1

eob is limited by GetCoeffs

Change-Id: Ie5c0d024796fe6c9b2db0374892544e421bd5d09

eob is limited by GetCoeffs

Change-Id: Id48a92e600375a1d4fb956757c93c91ebb5df59a

'number_of_layers' is range checked before assignment from the user
config.

Change-Id: Idefdaceb8736f126fa7c647da2b047dafb56ea52

significantly speeds up file generation.

the goal of this change is to convert rtcd.sh to perl as directly as
possible to allow for simple comparison. future changes can make it more
perl-like.

---
Linux
    [CREATE] vpx_scale_rtcd.h
real    0m0.485s ->    0m0.022s
    [CREATE] vp8_rtcd.h
real    0m4.619s ->    0m0.060s
    [CREATE] vp9_rtcd.h
real    0m10.102s ->    0m0.087s

Windows
    [CREATE] vpx_scale_rtcd.h
real    0m8.360s ->    0m0.080s
    [CREATE] vp8_rtcd.h
real    1m8.083s ->    0m0.160s
    [CREATE] vp9_rtcd.h
real    2m6.489s ->    0m0.233s

Change-Id: Idfb71188206c91237d6a3c3a81dfe00d103f11ee

Removing all copies of identical vp8_mse2psnr/vp9_mse2psnr functions.
Using vpx_sse_to_psnr() instead in all places.

Change-Id: I15beef9834d43d8fc8a8a7a2d1fc5de3d658fed8