[x265-commits] [x265] entropy: add check failure

Deepthi Nandakumar deepthi at multicorewareinc.com
Fri Mar 13 18:38:43 CET 2015


details:   http://hg.videolan.org/x265/rev/8f4fa5a24dac
branches:  
changeset: 9700:8f4fa5a24dac
user:      Deepthi Nandakumar <deepthi at multicorewareinc.com>
date:      Fri Mar 13 09:58:48 2015 +0530
description:
entropy: add check failure
Subject: [x265] asm: avx2 code for filter_vpp[16x4], filter_vps[16x4]: 303c->293c, 311c->253c

details:   http://hg.videolan.org/x265/rev/4029ab5bfc0f
branches:  
changeset: 9701:4029ab5bfc0f
user:      Divya Manivannan <divya at multicorewareinc.com>
date:      Fri Mar 13 09:52:41 2015 +0530
description:
asm: avx2 code for filter_vpp[16x4], filter_vps[16x4]: 303c->293c, 311c->253c
Subject: [x265] asm: improve ~5% on AVX2 interp_8tap_horiz_ps_4xN

details:   http://hg.videolan.org/x265/rev/9d680635d58c
branches:  
changeset: 9702:9d680635d58c
user:      Min Chen <chenm003 at 163.com>
date:      Thu Mar 12 20:46:46 2015 -0700
description:
asm: improve ~5% on AVX2 interp_8tap_horiz_ps_4xN
Subject: [x265] asm-intra_pred_ang8_11: improved, 317.84c -> 230.29c over SSE4 asm code

details:   http://hg.videolan.org/x265/rev/cfa8729f7d17
branches:  
changeset: 9703:cfa8729f7d17
user:      Praveen Tiwari <praveen at multicorewareinc.com>
date:      Thu Mar 12 18:40:23 2015 +0530
description:
asm-intra_pred_ang8_11: improved, 317.84c -> 230.29c over SSE4 asm code

AVX2:
intra_ang_8x8[11]       14.15x   230.29          3259.04

SSE4:
intra_ang_8x8[11]       10.25x   317.84          3258.71
Subject: [x265] asm-intra_pred_ang16_25: improved, 781.13c -> 466.16c

details:   http://hg.videolan.org/x265/rev/e99b176c0a61
branches:  
changeset: 9704:e99b176c0a61
user:      Praveen Tiwari <praveen at multicorewareinc.com>
date:      Thu Mar 12 19:08:34 2015 +0530
description:
asm-intra_pred_ang16_25: improved, 781.13c -> 466.16c

AVX2:
intra_ang_16x16[25]     18.09x   466.16          8434.26

SSE4:
intra_ang_16x16[25]     10.90x   781.13          8511.65
Subject: [x265] asm: filter_vsp[4x4], filter_vss[4x4] in avx2: 407c->198c, 361c->180c

details:   http://hg.videolan.org/x265/rev/9b873ad208ae
branches:  
changeset: 9705:9b873ad208ae
user:      Divya Manivannan <divya at multicorewareinc.com>
date:      Thu Mar 12 11:32:16 2015 +0530
description:
asm: filter_vsp[4x4], filter_vss[4x4] in avx2: 407c->198c, 361c->180c
Subject: [x265] asm: filter_vsp[8x8], filter_vss[8x8] in avx2: 887c->525c, 828c->524c

details:   http://hg.videolan.org/x265/rev/e2279723215a
branches:  
changeset: 9706:e2279723215a
user:      Divya Manivannan <divya at multicorewareinc.com>
date:      Thu Mar 12 14:02:18 2015 +0530
description:
asm: filter_vsp[8x8], filter_vss[8x8] in avx2: 887c->525c, 828c->524c
Subject: [x265] asm: filter_vsp[16x16, 32x16], filter_vss[16x16, 32x16]: 3042c->1875c, 5844c->3724c, 2646c->1988c, 4655c->4040c

details:   http://hg.videolan.org/x265/rev/eb89c6d5e259
branches:  
changeset: 9707:eb89c6d5e259
user:      Divya Manivannan <divya at multicorewareinc.com>
date:      Thu Mar 12 16:55:03 2015 +0530
description:
asm: filter_vsp[16x16, 32x16], filter_vss[16x16, 32x16]: 3042c->1875c, 5844c->3724c, 2646c->1988c, 4655c->4040c
Subject: [x265] asm: filter_vsp[16x32, 24x32, 32x32], filter_vss[16x32, 24x32, 32x32] in avx2

details:   http://hg.videolan.org/x265/rev/56fcfb6e49ac
branches:  
changeset: 9708:56fcfb6e49ac
user:      Divya Manivannan <divya at multicorewareinc.com>
date:      Thu Mar 12 17:31:45 2015 +0530
description:
asm: filter_vsp[16x32, 24x32, 32x32], filter_vss[16x32, 24x32, 32x32] in avx2

filter_vsp[16x32, 24x32, 32x32]: 6015c->3693c, 8710c->5692c, 11284c->7731c
filter_vss[16x32, 24x32, 32x32]: 4702c->4024c, 7013c->6132c, 9046c->7926c
Subject: [x265] encoder: set frame thread count correctly when no-wpp is enabled

details:   http://hg.videolan.org/x265/rev/6d05b68034cd
branches:  
changeset: 9709:6d05b68034cd
user:      Deepthi Nandakumar <deepthi at multicorewareinc.com>
date:      Fri Mar 13 11:52:43 2015 +0530
description:
encoder: set frame thread count correctly when no-wpp is enabled
Subject: [x265] asm : chroma_hps[4x2] for i420 avx2 - improved 329c->233c

details:   http://hg.videolan.org/x265/rev/0cb840697d81
branches:  
changeset: 9710:0cb840697d81
user:      Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
date:      Wed Mar 11 16:45:15 2015 +0530
description:
asm : chroma_hps[4x2] for i420 avx2 - improved 329c->233c
Subject: [x265] asm : chroma_hps[4x8 , 4x16] for i420 avx2 - improved 557c->377c, 873c->588c

details:   http://hg.videolan.org/x265/rev/b828ac0ca6b9
branches:  
changeset: 9711:b828ac0ca6b9
user:      Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
date:      Wed Mar 11 17:20:32 2015 +0530
description:
asm : chroma_hps[4x8 , 4x16] for i420 avx2 - improved 557c->377c, 873c->588c
Subject: [x265] asm: chroma_hps[16x4, 16x8, 16x12, 16x32] for i420 avx2 - improved 743c->468c, 1065c->681c, 1399c->894c, 2961c->1844c

details:   http://hg.videolan.org/x265/rev/9254dc264ba3
branches:  
changeset: 9712:9254dc264ba3
user:      Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
date:      Thu Mar 12 13:55:24 2015 +0530
description:
asm: chroma_hps[16x4, 16x8, 16x12, 16x32] for i420 avx2 - improved 743c->468c, 1065c->681c, 1399c->894c, 2961c->1844c
Subject: [x265] asm: chroma_hps[32x8, 32x16, 32x24] for i420 avx2 - improved 1843c->1210c, 3149c->2001c, 4440c->2906c

details:   http://hg.videolan.org/x265/rev/c6fc1bf05a0d
branches:  
changeset: 9713:c6fc1bf05a0d
user:      Aasaipriya Chandran <aasaipriya at multicorewareinc.com>
date:      Thu Mar 12 14:00:13 2015 +0530
description:
asm: chroma_hps[32x8, 32x16, 32x24] for i420 avx2 - improved 1843c->1210c, 3149c->2001c, 4440c->2906c
Subject: [x265] rc: recompute planned frame size when using vbv with --qpfile

details:   http://hg.videolan.org/x265/rev/62c1052a7c95
branches:  
changeset: 9714:62c1052a7c95
user:      Aarthi Thirumalai
date:      Wed Mar 11 23:05:55 2015 +0530
description:
rc: recompute planned frame size when using vbv with --qpfile
Subject: [x265] rc: clip qp after initial vbv-lookahead estimation.

details:   http://hg.videolan.org/x265/rev/e0f834c69cb2
branches:  
changeset: 9715:e0f834c69cb2
user:      Aarthi Thirumalai
date:      Wed Mar 11 23:26:27 2015 +0530
description:
rc: clip qp after initial vbv-lookahead estimation.

avoid drastic qp changes caused due to possible mispredictions from initial vbv predictors.
Subject: [x265] presets[OUTPUT CHANGE]: change superfast and ultrafast presets

details:   http://hg.videolan.org/x265/rev/c1a8eef8be14
branches:  
changeset: 9716:c1a8eef8be14
user:      Deepthi Nandakumar <deepthi at multicorewareinc.com>
date:      Fri Mar 13 16:40:06 2015 +0530
description:
presets[OUTPUT CHANGE]: change superfast and ultrafast presets

Better R-D curves for superfast and ultrafast and introducing minCUSize in ultrafast.
Subject: [x265] encoder: set frame thread count correctly when no-wpp is enabled

details:   http://hg.videolan.org/x265/rev/e4634aa7fdbe
branches:  stable
changeset: 9717:e4634aa7fdbe
user:      Deepthi Nandakumar <deepthi at multicorewareinc.com>
date:      Fri Mar 13 11:52:43 2015 +0530
description:
encoder: set frame thread count correctly when no-wpp is enabled
Subject: [x265] analysis: add logic for calculate qp for a given cu size

details:   http://hg.videolan.org/x265/rev/8dc54fd66706
branches:  
changeset: 9718:8dc54fd66706
user:      Sreelakshmy V G <sreelakshmy at multicorewareinc.com>
date:      Fri Mar 13 09:06:39 2015 +0530
description:
analysis: add logic for calculate qp for a given cu size

A new function calculateQpforCuSize() is added to calculate the qp for any
given cu.
Subject: [x265] rc: fix bug in CRF caused by e0f834c69cb2

details:   http://hg.videolan.org/x265/rev/1bed2e325efc
branches:  
changeset: 9719:1bed2e325efc
user:      Aarthi Thirumalai
date:      Fri Mar 13 21:57:56 2015 +0530
description:
rc: fix bug in CRF caused by e0f834c69cb2
Subject: [x265] asm: intra pred planar32 sse2

details:   http://hg.videolan.org/x265/rev/5e7519f25ad5
branches:  
changeset: 9720:5e7519f25ad5
user:      David T Yuen <dtyx265 at gmail.com>
date:      Thu Mar 12 19:19:24 2015 -0700
description:
asm: intra pred planar32 sse2

This replaces c code for systems using ssse3 to sse2 processors
The code is backported from intrapred planar32 sse4

There are essentially are two versions here.
One for x86_64 and one for x86_32.  It would have been too ugly
to conditionally code the differences in a single primitive.

64-bit

./test/TestBench --testbench intrapred | grep intra_planar_32x32
intra_planar_32x32	10.92x 	 11107.49 	 121282.19

32-bit

./test/TestBench --testbench intrapred | grep intra_planar_32x32
intra_planar_32x32	10.01x 	 9918.94  	 99315.12
Subject: [x265] asm: intra pred planar32 sse2 high bit

details:   http://hg.videolan.org/x265/rev/22ff3e631f65
branches:  
changeset: 9721:22ff3e631f65
user:      David T Yuen <dtyx265 at gmail.com>
date:      Thu Mar 12 19:24:11 2015 -0700
description:
asm: intra pred planar32 sse2 high bit

This replaces c code for systems using ssse3 to sse2 processors
The code is backported from intrapred planar32 sse4

Unlike the sse4 high bit version which operates with 32-bit values
this version uses 16-bits to calculate the prediction.  16-bits
is just enough for 10 bit luma depth, or at least it's enough for
the testbench.  Anything more and this primitive will overflow.

./test/TestBench --testbench intrapred | grep intra_planar_32x32
intra_planar_32x32	11.06x 	 10337.52 	 114374.60
Subject: [x265] Merge with stable

details:   http://hg.videolan.org/x265/rev/5ebd5d7c0a76
branches:  
changeset: 9722:5ebd5d7c0a76
user:      Steve Borho <steve at borho.org>
date:      Fri Mar 13 12:16:47 2015 -0500
description:
Merge with stable

diffstat:

 doc/reST/presets.rst                 |     8 +-
 source/common/param.cpp              |     7 +-
 source/common/x86/asm-primitives.cpp |    35 +
 source/common/x86/intrapred.h        |     3 +
 source/common/x86/intrapred16.asm    |   115 +++
 source/common/x86/intrapred8.asm     |   319 +++++++++
 source/common/x86/ipfilter8.asm      |  1103 +++++++++++++++++++++++++++++++++-
 source/common/x86/ipfilter8.h        |     3 +
 source/encoder/analysis.cpp          |    35 +
 source/encoder/analysis.h            |     2 +
 source/encoder/encoder.cpp           |     2 +-
 source/encoder/entropy.cpp           |     1 +
 source/encoder/ratecontrol.cpp       |    26 +-
 13 files changed, 1617 insertions(+), 42 deletions(-)

diffs (truncated from 2024 to 300 lines):

diff -r 3187844f4a7f -r 5ebd5d7c0a76 doc/reST/presets.rst
--- a/doc/reST/presets.rst	Mon Mar 09 14:35:20 2015 +0530
+++ b/doc/reST/presets.rst	Fri Mar 13 12:16:47 2015 -0500
@@ -24,11 +24,13 @@ The presets adjust encoder parameters to
 +==============+===========+===========+==========+========+======+========+======+========+==========+=========+
 | ctu          |   32      |    32     |   32     |  64    |  64  |   64   |  64  |  64    |   64     |   64    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
-| bframes      |    4      |     4     |    4     |   4    |  4   |    4   |  4   |   8    |    8     |    8    |
+| min-cu-size  |   16      |     8     |    8     |   8    |   8  |    8   |   8  |   8    |    8     |    8    |
++--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
+| bframes      |    3      |     3     |    4     |   4    |  4   |    4   |  4   |   8    |    8     |    8    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
 | b-adapt      |    0      |     0     |    0     |   0    |  2   |    2   |  2   |   2    |    2     |    2    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
-| rc-lookahead |   10      |    10     |   15     |  15    |  15  |   20   |  25  |   30   |   40     |   60    |
+| rc-lookahead |    5      |    10     |   15     |  15    |  15  |   20   |  25  |   30   |   40     |   60    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
 | scenecut     |    0      |    40     |   40     |  40    |  40  |   40   |  40  |   40   |   40     |   40    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
@@ -36,7 +38,7 @@ The presets adjust encoder parameters to
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
 | me           |   dia     |   hex     |   hex    |  hex   | hex  |   hex  | star |  star  |   star   |   star  |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
-| merange      |   25      |    44     |   57     |  57    |  57  |   57   | 57   |  57    |   57     |   92    |
+| merange      |   57      |    57     |   57     |  57    |  57  |   57   | 57   |  57    |   57     |   92    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
 | subme        |    0      |     1     |    1     |   2    |  2   |    2   |  3   |   3    |    4     |    5    |
 +--------------+-----------+-----------+----------+--------+------+--------+------+--------+----------+---------+
diff -r 3187844f4a7f -r 5ebd5d7c0a76 source/common/param.cpp
--- a/source/common/param.cpp	Mon Mar 09 14:35:20 2015 +0530
+++ b/source/common/param.cpp	Fri Mar 13 12:16:47 2015 -0500
@@ -247,10 +247,11 @@ int x265_param_default_preset(x265_param
 
         if (!strcmp(preset, "ultrafast"))
         {
-            param->lookaheadDepth = 10;
+            param->lookaheadDepth = 5;
             param->scenecutThreshold = 0; // disable lookahead
             param->maxCUSize = 32;
-            param->searchRange = 25;
+            param->minCUSize = 16;
+            param->bframes = 3;
             param->bFrameAdaptive = 0;
             param->subpelRefine = 0;
             param->searchMethod = X265_DIA_SEARCH;
@@ -269,7 +270,7 @@ int x265_param_default_preset(x265_param
         {
             param->lookaheadDepth = 10;
             param->maxCUSize = 32;
-            param->searchRange = 44;
+            param->bframes = 3;
             param->bFrameAdaptive = 0;
             param->subpelRefine = 1;
             param->bEnableEarlySkip = 1;
diff -r 3187844f4a7f -r 5ebd5d7c0a76 source/common/x86/asm-primitives.cpp
--- a/source/common/x86/asm-primitives.cpp	Mon Mar 09 14:35:20 2015 +0530
+++ b/source/common/x86/asm-primitives.cpp	Fri Mar 13 12:16:47 2015 -0500
@@ -880,6 +880,7 @@ void setupAssemblyPrimitives(EncoderPrim
         p.cu[BLOCK_4x4].intra_pred[PLANAR_IDX] = x265_intra_pred_planar4_sse2;
         p.cu[BLOCK_8x8].intra_pred[PLANAR_IDX] = x265_intra_pred_planar8_sse2;
         p.cu[BLOCK_16x16].intra_pred[PLANAR_IDX] = x265_intra_pred_planar16_sse2;
+        p.cu[BLOCK_32x32].intra_pred[PLANAR_IDX] = x265_intra_pred_planar32_sse2;
 
         p.cu[BLOCK_4x4].sse_ss = x265_pixel_ssd_ss_4x4_mmx2;
         ALL_LUMA_CU(sse_ss, pixel_ssd_ss, sse2);
@@ -1193,6 +1194,7 @@ void setupAssemblyPrimitives(EncoderPrim
         p.cu[BLOCK_4x4].intra_pred[PLANAR_IDX] = x265_intra_pred_planar4_sse2;
         p.cu[BLOCK_8x8].intra_pred[PLANAR_IDX] = x265_intra_pred_planar8_sse2;
         p.cu[BLOCK_16x16].intra_pred[PLANAR_IDX] = x265_intra_pred_planar16_sse2;
+        p.cu[BLOCK_32x32].intra_pred[PLANAR_IDX] = x265_intra_pred_planar32_sse2;
 
         p.cu[BLOCK_4x4].calcresidual = x265_getResidual4_sse2;
         p.cu[BLOCK_8x8].calcresidual = x265_getResidual8_sse2;
@@ -1503,6 +1505,8 @@ void setupAssemblyPrimitives(EncoderPrim
         p.cu[BLOCK_8x8].intra_pred[25] = x265_intra_pred_ang8_25_avx2;
         p.cu[BLOCK_8x8].intra_pred[12] = x265_intra_pred_ang8_12_avx2;
         p.cu[BLOCK_8x8].intra_pred[24] = x265_intra_pred_ang8_24_avx2;
+        p.cu[BLOCK_8x8].intra_pred[11] = x265_intra_pred_ang8_11_avx2;
+        p.cu[BLOCK_16x16].intra_pred[25] = x265_intra_pred_ang16_25_avx2;
 
         // copy_sp primitives
         p.cu[BLOCK_16x16].copy_sp = x265_blockcopy_sp_16x16_avx2;
@@ -1582,6 +1586,19 @@ void setupAssemblyPrimitives(EncoderPrim
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_hps = x265_interp_4tap_horiz_ps_4x4_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x8].filter_hps = x265_interp_4tap_horiz_ps_8x8_avx2;
 
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_4x2].filter_hps = x265_interp_4tap_horiz_ps_4x2_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_4x8].filter_hps = x265_interp_4tap_horiz_ps_4x8_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_4x16].filter_hps = x265_interp_4tap_horiz_ps_4x16_avx2;
+
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x32].filter_hps = x265_interp_4tap_horiz_ps_16x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x12].filter_hps = x265_interp_4tap_horiz_ps_16x12_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_hps = x265_interp_4tap_horiz_ps_16x8_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_hps = x265_interp_4tap_horiz_ps_16x4_avx2;
+
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x16].filter_hps = x265_interp_4tap_horiz_ps_32x16_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x24].filter_hps = x265_interp_4tap_horiz_ps_32x24_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x8].filter_hps = x265_interp_4tap_horiz_ps_32x8_avx2;
+
         p.chroma[X265_CSP_I422].pu[CHROMA_422_4x4].filter_vpp = x265_interp_4tap_vert_pp_4x4_avx2;
 
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_vpp = x265_interp_4tap_vert_pp_4x4_avx2;
@@ -1597,6 +1614,7 @@ void setupAssemblyPrimitives(EncoderPrim
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x6].filter_vpp = x265_interp_4tap_vert_pp_8x6_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x16].filter_vpp = x265_interp_4tap_vert_pp_8x16_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x32].filter_vpp = x265_interp_4tap_vert_pp_8x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_vpp = x265_interp_4tap_vert_pp_16x4_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_vpp = x265_interp_4tap_vert_pp_16x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x12].filter_vpp = x265_interp_4tap_vert_pp_16x12_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x16].filter_vpp = x265_interp_4tap_vert_pp_16x16_avx2;
@@ -1619,6 +1637,7 @@ void setupAssemblyPrimitives(EncoderPrim
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x8].filter_vps = x265_interp_4tap_vert_ps_8x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x16].filter_vps = x265_interp_4tap_vert_ps_8x16_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_8x32].filter_vps = x265_interp_4tap_vert_ps_8x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x4].filter_vps = x265_interp_4tap_vert_ps_16x4_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x8].filter_vps = x265_interp_4tap_vert_ps_16x8_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_16x12].filter_vps = x265_interp_4tap_vert_ps_16x12_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_4x16].filter_vps = x265_interp_4tap_vert_ps_4x16_avx2;
@@ -1629,6 +1648,22 @@ void setupAssemblyPrimitives(EncoderPrim
         p.chroma[X265_CSP_I420].pu[CHROMA_420_32x24].filter_vps = x265_interp_4tap_vert_ps_32x24_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_32x16].filter_vps = x265_interp_4tap_vert_ps_32x16_avx2;
         p.chroma[X265_CSP_I420].pu[CHROMA_420_32x8].filter_vps = x265_interp_4tap_vert_ps_32x8_avx2;
+
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_vsp = x265_interp_4tap_vert_sp_4x4_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_8x8].filter_vsp = x265_interp_4tap_vert_sp_8x8_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x16].filter_vsp = x265_interp_4tap_vert_sp_16x16_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x32].filter_vsp = x265_interp_4tap_vert_sp_32x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x32].filter_vsp = x265_interp_4tap_vert_sp_16x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_24x32].filter_vsp = x265_interp_4tap_vert_sp_24x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x16].filter_vsp = x265_interp_4tap_vert_sp_32x16_avx2;
+
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_4x4].filter_vss = x265_interp_4tap_vert_ss_4x4_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_8x8].filter_vss = x265_interp_4tap_vert_ss_8x8_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x16].filter_vss = x265_interp_4tap_vert_ss_16x16_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x32].filter_vss = x265_interp_4tap_vert_ss_32x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_16x32].filter_vss = x265_interp_4tap_vert_ss_16x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_24x32].filter_vss = x265_interp_4tap_vert_ss_24x32_avx2;
+        p.chroma[X265_CSP_I420].pu[CHROMA_420_32x16].filter_vss = x265_interp_4tap_vert_ss_32x16_avx2;
     }
 #endif
 }
diff -r 3187844f4a7f -r 5ebd5d7c0a76 source/common/x86/intrapred.h
--- a/source/common/x86/intrapred.h	Mon Mar 09 14:35:20 2015 +0530
+++ b/source/common/x86/intrapred.h	Fri Mar 13 12:16:47 2015 -0500
@@ -38,6 +38,7 @@ void x265_intra_pred_dc32_sse4(pixel* ds
 void x265_intra_pred_planar4_sse2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
 void x265_intra_pred_planar8_sse2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
 void x265_intra_pred_planar16_sse2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
+void x265_intra_pred_planar32_sse2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
 void x265_intra_pred_planar4_sse4(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
 void x265_intra_pred_planar8_sse4(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
 void x265_intra_pred_planar16_sse4(pixel* dst, intptr_t dstStride, const pixel* srcPix, int, int);
@@ -181,6 +182,8 @@ void x265_intra_pred_ang8_27_avx2(pixel*
 void x265_intra_pred_ang8_25_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_intra_pred_ang8_12_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_intra_pred_ang8_24_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
+void x265_intra_pred_ang8_11_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
+void x265_intra_pred_ang16_25_avx2(pixel* dst, intptr_t dstStride, const pixel* srcPix, int dirMode, int bFilter);
 void x265_all_angs_pred_4x4_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
 void x265_all_angs_pred_8x8_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
 void x265_all_angs_pred_16x16_sse4(pixel *dest, pixel *refPix, pixel *filtPix, int bLuma);
diff -r 3187844f4a7f -r 5ebd5d7c0a76 source/common/x86/intrapred16.asm
--- a/source/common/x86/intrapred16.asm	Mon Mar 09 14:35:20 2015 +0530
+++ b/source/common/x86/intrapred16.asm	Fri Mar 13 12:16:47 2015 -0500
@@ -65,6 +65,10 @@ pw_planar16_0:        dw 15, 14, 13, 12,
 pw_planar16_1:        dw 15, 15, 15, 15, 15, 15, 15, 15
 pd_planar32_1:        dd 31, 31, 31, 31
 
+pw_planar32_1:        dw 31, 31, 31, 31, 31, 31, 31, 31
+pw_planar32_L:        dw 31, 30, 29, 28, 27, 26, 25, 24
+pw_planar32_H:        dw 23, 22, 21, 20, 19, 18, 17, 16
+
 const planar32_table
 %assign x 31
 %rep 8
@@ -86,12 +90,15 @@ cextern pw_2
 cextern pw_4
 cextern pw_8
 cextern pw_16
+cextern pw_32
 cextern pw_1023
 cextern pd_16
 cextern pd_32
 cextern pw_4096
 cextern multiL
 cextern multiH
+cextern multiH2
+cextern multiH3
 cextern multi_2Row
 cextern pw_swap
 cextern pb_unpackwq1
@@ -575,6 +582,114 @@ cglobal intra_pred_planar16, 3,3,8
     INTRA_PRED_PLANAR_16 15
     RET
 
+;---------------------------------------------------------------------------------------
+; void intra_pred_planar(pixel* dst, intptr_t dstStride, pixel*srcPix, int, int filter)
+;---------------------------------------------------------------------------------------
+INIT_XMM sse2
+cglobal intra_pred_planar32, 3,3,16
+    movd            m3, [r2 + 66]               ; topRight   = above[32]
+
+    pshuflw         m3, m3, 0x00
+    pshufd          m3, m3, 0x44
+
+    pmullw          m0, m3, [multiL]            ; (x + 1) * topRight
+    pmullw          m1, m3, [multiH]            ; (x + 1) * topRight
+    pmullw          m2, m3, [multiH2]           ; (x + 1) * topRight
+    pmullw          m3, [multiH3]               ; (x + 1) * topRight
+
+    movd            m6, [r2 + 194]               ; bottomLeft = left[32]
+    pshuflw         m6, m6, 0x00
+    pshufd          m6, m6, 0x44
+    mova            m5, m6
+    paddw           m5, [pw_32]
+
+    paddw           m0, m5
+    paddw           m1, m5
+    paddw           m2, m5
+    paddw           m3, m5
+    mova            m8, m6
+    mova            m9, m6
+    mova            m10, m6
+
+    mova            m12, [pw_planar32_1]
+    movu            m4, [r2 + 2]
+    psubw           m8, m4
+    pmullw          m4, m12
+    paddw           m0, m4
+
+    movu            m5, [r2 + 18]
+    psubw           m9, m5
+    pmullw          m5, m12
+    paddw           m1, m5
+
+    movu            m4, [r2 + 34]
+    psubw           m10, m4
+    pmullw          m4, m12
+    paddw           m2, m4
+
+    movu            m5, [r2 + 50]
+    psubw           m6, m5
+    pmullw          m5, m12
+    paddw           m3, m5
+
+    mova            m12, [pw_planar32_L]
+    mova            m13, [pw_planar32_H]
+    mova            m14, [pw_planar16_0]
+    mova            m15, [pw_planar8_0]
+    add             r1, r1
+
+%macro PROCESS 1
+    pmullw          m5, %1, m12
+    pmullw          m11, %1, m13
+    paddw           m5, m0
+    paddw           m11, m1
+    psrlw           m5, 6
+    psrlw           m11, 6
+    movu            [r0], m5
+    movu            [r0 + 16], m11
+
+    pmullw          m5, %1, m14
+    pmullw          %1, m15
+    paddw           m5, m2
+    paddw           %1, m3
+    psrlw           m5, 6
+    psrlw           %1, 6
+    movu            [r0 + 32], m5
+    movu            [r0 + 48], %1
+%endmacro
+
+%macro  INCREMENT 0
+    paddw           m2, m10
+    paddw           m3, m6
+    paddw           m0, m8
+    paddw           m1, m9
+    add             r0, r1
+%endmacro
+
+    add             r2, 130             ;130 = 32*sizeof(pixel)*2 + 1*sizeof(pixel)
+%assign x 0
+%rep 4
+    movu            m4, [r2]
+    add             r2, 16
+%assign y 0
+%rep 8
+    %if y < 4
+    pshuflw         m7, m4, 0x55 * y
+    pshufd          m7, m7, 0x44
+    %else
+    pshufhw         m7, m4, 0x55 * (y - 4)
+    pshufd          m7, m7, 0xEE
+    %endif
+        PROCESS m7
+    %if x + y < 10
+    INCREMENT
+    %endif
+%assign y y+1
+%endrep
+%assign x x+1


More information about the x265-commits mailing list