generic-library/vpx
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity
209def2d72
vpx/vp8/decoder/threading.c

989 lines
37 KiB
C
Raw Normal View History

Initial WebM release
2010-05-18 17:58:33 +02:00
/*
Use WebM in copyright notice for consistency Changes 'The VP8 project' to 'The WebM project', for consistency with other webmproject.org repositories. Fixes issue #97. Change-Id: I37c13ed5fbdb9d334ceef71c6350e9febed9bbba
2010-09-09 14:16:39 +02:00
* Copyright (c) 2010 The WebM project authors. All Rights Reserved.
Initial WebM release
2010-05-18 17:58:33 +02:00
*
Remove secondary mv clamping from decode stage This patch removes the secondary MV clamping from the MV decoder. This behavior was consistent with limits placed on non-split MVs by the reference encoder, but was inconsistent with the MVs generated in the split case. The purpose of this secondary clamping was only to prevent crashes on invalid data. It was not intended to be a behaviour an encoder could or should rely on. Instead of doing additional clamping in a way that changes the entropy context, the secondary clamp is removed and the border handling is made implmentation specific. With respect to the spec, the border is treated as essentially infinite, limited only by the clamping performed on the near/nearest reference and the maximum encodable magnitude of the residual MV. This does not affect any currently produced streams. Change-Id: I68d35a2fbb51570d6569eab4ad233961405230a3
2010-06-09 17:29:20 +02:00
* Use of this source code is governed by a BSD-style license
LICENSE: update with latest text Change-Id: Ieebea089095d9073b3a94932791099f614ce120c
2010-06-04 22:19:40 +02:00
* that can be found in the LICENSE file in the root of the source
* tree. An additional intellectual property rights grant can be found
Remove secondary mv clamping from decode stage This patch removes the secondary MV clamping from the MV decoder. This behavior was consistent with limits placed on non-split MVs by the reference encoder, but was inconsistent with the MVs generated in the split case. The purpose of this secondary clamping was only to prevent crashes on invalid data. It was not intended to be a behaviour an encoder could or should rely on. Instead of doing additional clamping in a way that changes the entropy context, the secondary clamp is removed and the border handling is made implmentation specific. With respect to the spec, the border is treated as essentially infinite, limited only by the clamping performed on the near/nearest reference and the maximum encodable magnitude of the residual MV. This does not affect any currently produced streams. Change-Id: I68d35a2fbb51570d6569eab4ad233961405230a3
2010-06-09 17:29:20 +02:00
* in the file PATENTS. All contributing project authors may
LICENSE: update with latest text Change-Id: Ieebea089095d9073b3a94932791099f614ce120c
2010-06-04 22:19:40 +02:00
* be found in the AUTHORS file in the root of the source tree.
Initial WebM release
2010-05-18 17:58:33 +02:00
*/
Adds "armvX-none-rvct" targets Adds following targets to configure script to support RVCT compilation without operating system support (for Profiler or bare metal images). - armv5te-none-rvct - armv6-none-rvct - armv7-none-rvct To strip OS specific parts from the code "os_support"-config was added to script and CONFIG_OS_SUPPORT flag is used in the code to exclude OS specific parts such as OS specific includes and function calls for timers and threads etc. This was done to enable RVCT compilation for profiling purposes or running the image on bare metal target with Lauterbach. Removed separate AREA directives for READONLY data in armv6 and neon assembly files to fix the RVCT compilation. Otherwise "ldr <reg>, =label" syntax would have been needed to prevent linker errors. This syntax is not supported by older gnu assemblers. Change-Id: I14f4c68529e8c27397502fbc3010a54e505ddb43
2011-01-24 10:21:40 +01:00
#if !defined(WIN32) && CONFIG_OS_SUPPORT == 1
Initial WebM release
2010-05-18 17:58:33 +02:00
# include <unistd.h>
#endif
apple: include proper mach primatives Fixes implicit declaration warning for 'mach_task_self'. Patch courtesy of timeless at gmail.com Change-Id: I9991dedd1ccfddc092eca86705ecbc3b764b799d
2010-07-29 23:04:39 +02:00
#ifdef __APPLE__
#include <mach/mach_init.h>
#endif
Initial WebM release
2010-05-18 17:58:33 +02:00
#include "onyxd_int.h"
#include "vpx_mem/vpx_mem.h"
#include "threading.h"
#include "loopfilter.h"
#include "extend.h"
#include "vpx_ports/vpx_timer.h"
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
#include "detokenize.h"
#include "reconinter.h"
#include "reconintra_mt.h"
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
extern void mb_init_dequantizer(VP8D_COMP *pbi, MACROBLOCKD *xd);
extern void clamp_mvs(MACROBLOCKD *xd);
Initial WebM release
2010-05-18 17:58:33 +02:00
extern void vp8_build_uvmvs(MACROBLOCKD *x, int fullpixel);
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
#if CONFIG_RUNTIME_CPU_DETECT
#define RTCD_VTABLE(x) (&(pbi)->common.rtcd.x)
#else
#define RTCD_VTABLE(x) NULL
#endif
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Initial WebM release
2010-05-18 17:58:33 +02:00
void vp8_setup_decoding_thread_data(VP8D_COMP *pbi, MACROBLOCKD *xd, MB_ROW_DEC *mbrd, int count)
{
#if CONFIG_MULTITHREAD
VP8_COMMON *const pc = & pbi->common;
int i, j;
for (i = 0; i < count; i++)
{
MACROBLOCKD *mbd = &mbrd[i].mbd;
#if CONFIG_RUNTIME_CPU_DETECT
mbd->rtcd = xd->rtcd;
#endif
mbd->subpixel_predict = xd->subpixel_predict;
mbd->subpixel_predict8x4 = xd->subpixel_predict8x4;
mbd->subpixel_predict8x8 = xd->subpixel_predict8x8;
mbd->subpixel_predict16x16 = xd->subpixel_predict16x16;
mbd->mode_info_context = pc->mi + pc->mode_info_stride * (i + 1);
mbd->mode_info_stride = pc->mode_info_stride;
mbd->frame_type = pc->frame_type;
mbd->frames_since_golden = pc->frames_since_golden;
mbd->frames_till_alt_ref_frame = pc->frames_till_alt_ref_frame;
Swap alt/gold/new/last frame buffer ptrs instead of copying. At the end of the decode, frame buffers were being copied. The frames are not updated after the copy, they are just for reference on later frames. This change allows multiple references to the same frame buffer instead of copying it. Changes needed to be made to the encoder to handle this. The encoder is still doing frame buffer copies in similar places where pointer reference could be done. Change-Id: I7c38be4d23979cc49b5f17241ca3a78703803e66
2010-07-22 14:07:32 +02:00
mbd->pre = pc->yv12_fb[pc->lst_fb_idx];
mbd->dst = pc->yv12_fb[pc->new_fb_idx];
Initial WebM release
2010-05-18 17:58:33 +02:00
vp8_setup_block_dptrs(mbd);
vp8_build_block_doffsets(mbd);
mbd->segmentation_enabled = xd->segmentation_enabled;
mbd->mb_segement_abs_delta = xd->mb_segement_abs_delta;
vpx_memcpy(mbd->segment_feature_data, xd->segment_feature_data, sizeof(xd->segment_feature_data));
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*signed char ref_lf_deltas[MAX_REF_LF_DELTAS];*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vpx_memcpy(mbd->ref_lf_deltas, xd->ref_lf_deltas, sizeof(xd->ref_lf_deltas));
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*signed char mode_lf_deltas[MAX_MODE_LF_DELTAS];*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vpx_memcpy(mbd->mode_lf_deltas, xd->mode_lf_deltas, sizeof(xd->mode_lf_deltas));
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*unsigned char mode_ref_lf_delta_enabled;
unsigned char mode_ref_lf_delta_update;*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
mbd->mode_ref_lf_delta_enabled = xd->mode_ref_lf_delta_enabled;
mbd->mode_ref_lf_delta_update = xd->mode_ref_lf_delta_update;
Initial WebM release
2010-05-18 17:58:33 +02:00
mbd->current_bc = &pbi->bc2;
for (j = 0; j < 25; j++)
{
mbd->block[j].dequant = xd->block[j].dequant;
}
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (i=0; i< pc->mb_rows; i++)
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->mt_current_mb_col[i]=-1;
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
(void) pbi;
(void) xd;
(void) mbrd;
(void) count;
#endif
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
void vp8mt_decode_macroblock(VP8D_COMP *pbi, MACROBLOCKD *xd, int mb_row, int mb_col)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
#if CONFIG_MULTITHREAD
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int eobtotal = 0;
int i, do_clamp = xd->mode_info_context->mbmi.need_to_clamp_mvs;
VP8_COMMON *pc = &pbi->common;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->mode_info_context->mbmi.mb_skip_coeff)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8_reset_mb_tokens_context(xd);
}
else
{
eobtotal = vp8_decode_mb_tokens(pbi, xd);
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Perform temporary clamping of the MV to be used for prediction */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (do_clamp)
{
clamp_mvs(xd);
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
xd->mode_info_context->mbmi.dc_diff = 1;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->mode_info_context->mbmi.mode != B_PRED && xd->mode_info_context->mbmi.mode != SPLITMV && eobtotal == 0)
{
xd->mode_info_context->mbmi.dc_diff = 0;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*mt_skip_recon_mb(pbi, xd, mb_row, mb_col);*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->frame_type == KEY_FRAME || xd->mode_info_context->mbmi.ref_frame == INTRA_FRAME)
{
vp8mt_build_intra_predictors_mbuv_s(pbi, xd, mb_row, mb_col);
vp8mt_build_intra_predictors_mby_s(pbi, xd, mb_row, mb_col);
}
else
{
vp8_build_inter_predictors_mb_s(xd);
}
return;
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->segmentation_enabled)
mb_init_dequantizer(pbi, xd);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* do prediction */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->frame_type == KEY_FRAME || xd->mode_info_context->mbmi.ref_frame == INTRA_FRAME)
{
vp8mt_build_intra_predictors_mbuv(pbi, xd, mb_row, mb_col);
if (xd->mode_info_context->mbmi.mode != B_PRED)
{
vp8mt_build_intra_predictors_mby(pbi, xd, mb_row, mb_col);
} else {
vp8mt_intra_prediction_down_copy(pbi, xd, mb_row, mb_col);
}
}
else
{
vp8_build_inter_predictors_mb(xd);
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* dequantization and idct */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->mode_info_context->mbmi.mode != B_PRED && xd->mode_info_context->mbmi.mode != SPLITMV)
{
BLOCKD *b = &xd->block[24];
DEQUANT_INVOKE(&pbi->dequant, block)(b);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* do 2nd order transform on the dc block */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->eobs[24] > 1)
{
IDCT_INVOKE(RTCD_VTABLE(idct), iwalsh16)(&b->dqcoeff[0], b->diff);
((int *)b->qcoeff)[0] = 0;
((int *)b->qcoeff)[1] = 0;
((int *)b->qcoeff)[2] = 0;
((int *)b->qcoeff)[3] = 0;
((int *)b->qcoeff)[4] = 0;
((int *)b->qcoeff)[5] = 0;
((int *)b->qcoeff)[6] = 0;
((int *)b->qcoeff)[7] = 0;
}
else
{
IDCT_INVOKE(RTCD_VTABLE(idct), iwalsh1)(&b->dqcoeff[0], b->diff);
((int *)b->qcoeff)[0] = 0;
}
DEQUANT_INVOKE (&pbi->dequant, dc_idct_add_y_block)
Convert [4][4] matrices to [16] arrays. Most of the code that actually uses these matrices indexes them as if they were a single contiguous array, and coverity produces reports about the resulting accesses that overflow the static bounds of the first row. This is perfectly legal in C, but converting them to actual [16] arrays should eliminate the report, and removes a good deal of extraneous indexing and address operators from the code. Change-Id: Ibda479e2232b3e51f9edf3b355b8640520fdbf23
2010-10-22 02:04:30 +02:00
(xd->qcoeff, xd->block[0].dequant,
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
xd->predictor, xd->dst.y_buffer,
xd->dst.y_stride, xd->eobs, xd->block[24].diff);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
else if ((xd->frame_type == KEY_FRAME || xd->mode_info_context->mbmi.ref_frame == INTRA_FRAME) && xd->mode_info_context->mbmi.mode == B_PRED)
{
for (i = 0; i < 16; i++)
{
BLOCKD *b = &xd->block[i];
vp8mt_predict_intra4x4(pbi, xd, b->bmi.mode, b->predictor, mb_row, mb_col, i);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->eobs[i] > 1)
{
DEQUANT_INVOKE(&pbi->dequant, idct_add)
Convert [4][4] matrices to [16] arrays. Most of the code that actually uses these matrices indexes them as if they were a single contiguous array, and coverity produces reports about the resulting accesses that overflow the static bounds of the first row. This is perfectly legal in C, but converting them to actual [16] arrays should eliminate the report, and removes a good deal of extraneous indexing and address operators from the code. Change-Id: Ibda479e2232b3e51f9edf3b355b8640520fdbf23
2010-10-22 02:04:30 +02:00
(b->qcoeff, b->dequant, b->predictor,
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
*(b->base_dst) + b->dst, 16, b->dst_stride);
}
else
{
IDCT_INVOKE(RTCD_VTABLE(idct), idct1_scalar_add)
Convert [4][4] matrices to [16] arrays. Most of the code that actually uses these matrices indexes them as if they were a single contiguous array, and coverity produces reports about the resulting accesses that overflow the static bounds of the first row. This is perfectly legal in C, but converting them to actual [16] arrays should eliminate the report, and removes a good deal of extraneous indexing and address operators from the code. Change-Id: Ibda479e2232b3e51f9edf3b355b8640520fdbf23
2010-10-22 02:04:30 +02:00
(b->qcoeff[0] * b->dequant[0], b->predictor,
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
*(b->base_dst) + b->dst, 16, b->dst_stride);
((int *)b->qcoeff)[0] = 0;
}
}
}
else
{
DEQUANT_INVOKE (&pbi->dequant, idct_add_y_block)
Convert [4][4] matrices to [16] arrays. Most of the code that actually uses these matrices indexes them as if they were a single contiguous array, and coverity produces reports about the resulting accesses that overflow the static bounds of the first row. This is perfectly legal in C, but converting them to actual [16] arrays should eliminate the report, and removes a good deal of extraneous indexing and address operators from the code. Change-Id: Ibda479e2232b3e51f9edf3b355b8640520fdbf23
2010-10-22 02:04:30 +02:00
(xd->qcoeff, xd->block[0].dequant,
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
xd->predictor, xd->dst.y_buffer,
xd->dst.y_stride, xd->eobs);
}
DEQUANT_INVOKE (&pbi->dequant, idct_add_uv_block)
Convert [4][4] matrices to [16] arrays. Most of the code that actually uses these matrices indexes them as if they were a single contiguous array, and coverity produces reports about the resulting accesses that overflow the static bounds of the first row. This is perfectly legal in C, but converting them to actual [16] arrays should eliminate the report, and removes a good deal of extraneous indexing and address operators from the code. Change-Id: Ibda479e2232b3e51f9edf3b355b8640520fdbf23
2010-10-22 02:04:30 +02:00
(xd->qcoeff+16*16, xd->block[16].dequant,
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
xd->predictor+16*16, xd->dst.u_buffer, xd->dst.v_buffer,
xd->dst.uv_stride, xd->eobs+16);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
#else
(void) pbi;
(void) xd;
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
(void) mb_row;
(void) mb_col;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
#endif
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
Initial WebM release
2010-05-18 17:58:33 +02:00
THREAD_FUNCTION vp8_thread_decoding_proc(void *p_data)
{
#if CONFIG_MULTITHREAD
int ithread = ((DECODETHREAD_DATA *)p_data)->ithread;
VP8D_COMP *pbi = (VP8D_COMP *)(((DECODETHREAD_DATA *)p_data)->ptr1);
MB_ROW_DEC *mbrd = (MB_ROW_DEC *)(((DECODETHREAD_DATA *)p_data)->ptr2);
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
ENTROPY_CONTEXT_PLANES mb_row_left_context;
Initial WebM release
2010-05-18 17:58:33 +02:00
while (1)
{
if (pbi->b_multithreaded_rd == 0)
break;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*if(WaitForSingleObject(pbi->h_event_start_decoding[ithread], INFINITE) == WAIT_OBJECT_0)*/
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
if (sem_wait(&pbi->h_event_start_decoding[ithread]) == 0)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
if (pbi->b_multithreaded_rd == 0)
break;
else
{
VP8_COMMON *pc = &pbi->common;
MACROBLOCKD *xd = &mbrd->mbd;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
int mb_row;
int num_part = 1 << pbi->common.multi_token_partition;
Initial WebM release
2010-05-18 17:58:33 +02:00
volatile int *last_row_current_mb_col;
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
int nsync = pbi->sync_range;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (mb_row = ithread+1; mb_row < pc->mb_rows; mb_row += (pbi->decoding_thread_count + 1))
Initial WebM release
2010-05-18 17:58:33 +02:00
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
int i;
int recon_yoffset, recon_uvoffset;
int mb_col;
int ref_fb_idx = pc->lst_fb_idx;
int dst_fb_idx = pc->new_fb_idx;
int recon_y_stride = pc->yv12_fb[ref_fb_idx].y_stride;
int recon_uv_stride = pc->yv12_fb[ref_fb_idx].uv_stride;
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int filter_level;
loop_filter_info *lfi = pc->lf_info;
int alt_flt_enabled = xd->segmentation_enabled;
int Segment;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
pbi->mb_row_di[ithread].mb_row = mb_row;
pbi->mb_row_di[ithread].mbd.current_bc = &pbi->mbc[mb_row%num_part];
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
last_row_current_mb_col = &pbi->mt_current_mb_col[mb_row -1];
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
recon_yoffset = mb_row * recon_y_stride * 16;
recon_uvoffset = mb_row * recon_uv_stride * 8;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* reset above block coeffs */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
xd->above_context = pc->above_context;
xd->left_context = &mb_row_left_context;
vpx_memset(&mb_row_left_context, 0, sizeof(mb_row_left_context));
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->up_available = (mb_row != 0);
xd->mb_to_top_edge = -((mb_row * 16)) << 3;
xd->mb_to_bottom_edge = ((pc->mb_rows - 1 - mb_row) * 16) << 3;
for (mb_col = 0; mb_col < pc->mb_cols; mb_col++)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
if ((mb_col & (nsync-1)) == 0)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
while (mb_col > (*last_row_current_mb_col - nsync) && *last_row_current_mb_col != pc->mb_cols - 1)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
x86_pause_hint();
thread_sleep(0);
}
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
if (xd->mode_info_context->mbmi.mode == SPLITMV || xd->mode_info_context->mbmi.mode == B_PRED)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (i = 0; i < 16; i++)
{
BLOCKD *d = &xd->block[i];
vpx_memcpy(&d->bmi, &xd->mode_info_context->bmi[i], sizeof(B_MODE_INFO));
}
Initial WebM release
2010-05-18 17:58:33 +02:00
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if(pbi->common.filter_level)
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*update loopfilter info*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
Segment = (alt_flt_enabled) ? xd->mode_info_context->mbmi.segment_id : 0;
filter_level = pbi->mt_baseline_filter_level[Segment];
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Distance of Mb to the various image edges.
* These are specified to 8th pel as they are always compared to values that are in 1/8th pel units
* Apply any context driven MB level adjustment
*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8_adjust_mb_lf_value(xd, &filter_level);
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Distance of Mb to the various image edges.
* These are specified to 8th pel as they are always compared to values that are in 1/8th pel units
*/
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->mb_to_left_edge = -((mb_col * 16) << 3);
xd->mb_to_right_edge = ((pc->mb_cols - 1 - mb_col) * 16) << 3;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->dst.y_buffer = pc->yv12_fb[dst_fb_idx].y_buffer + recon_yoffset;
xd->dst.u_buffer = pc->yv12_fb[dst_fb_idx].u_buffer + recon_uvoffset;
xd->dst.v_buffer = pc->yv12_fb[dst_fb_idx].v_buffer + recon_uvoffset;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->left_available = (mb_col != 0);
Swap alt/gold/new/last frame buffer ptrs instead of copying. At the end of the decode, frame buffers were being copied. The frames are not updated after the copy, they are just for reference on later frames. This change allows multiple references to the same frame buffer instead of copying it. Changes needed to be made to the encoder to handle this. The encoder is still doing frame buffer copies in similar places where pointer reference could be done. Change-Id: I7c38be4d23979cc49b5f17241ca3a78703803e66
2010-07-22 14:07:32 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Select the appropriate reference frame for this MB */
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
if (xd->mode_info_context->mbmi.ref_frame == LAST_FRAME)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
ref_fb_idx = pc->lst_fb_idx;
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
else if (xd->mode_info_context->mbmi.ref_frame == GOLDEN_FRAME)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
ref_fb_idx = pc->gld_fb_idx;
else
ref_fb_idx = pc->alt_fb_idx;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->pre.y_buffer = pc->yv12_fb[ref_fb_idx].y_buffer + recon_yoffset;
xd->pre.u_buffer = pc->yv12_fb[ref_fb_idx].u_buffer + recon_uvoffset;
xd->pre.v_buffer = pc->yv12_fb[ref_fb_idx].v_buffer + recon_uvoffset;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
vp8_build_uvmvs(xd, pc->full_pixel);
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8mt_decode_macroblock(pbi, xd, mb_row, mb_col);
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->common.filter_level)
{
if( mb_row != pc->mb_rows-1 )
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Save decoded MB last row data for next-row decoding */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vpx_memcpy((pbi->mt_yabove_row[mb_row + 1] + 32 + mb_col*16), (xd->dst.y_buffer + 15 * recon_y_stride), 16);
vpx_memcpy((pbi->mt_uabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.u_buffer + 7 * recon_uv_stride), 8);
vpx_memcpy((pbi->mt_vabove_row[mb_row + 1] + 16 + mb_col*8), (xd->dst.v_buffer + 7 * recon_uv_stride), 8);
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* save left_col for next MB decoding */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if(mb_col != pc->mb_cols-1)
{
MODE_INFO *next = xd->mode_info_context +1;
if (xd->frame_type == KEY_FRAME || next->mbmi.ref_frame == INTRA_FRAME)
{
for (i = 0; i < 16; i++)
pbi->mt_yleft_col[mb_row][i] = xd->dst.y_buffer [i* recon_y_stride + 15];
for (i = 0; i < 8; i++)
{
pbi->mt_uleft_col[mb_row][i] = xd->dst.u_buffer [i* recon_uv_stride + 7];
pbi->mt_vleft_col[mb_row][i] = xd->dst.v_buffer [i* recon_uv_stride + 7];
}
}
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* loopfilter on this macroblock. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (filter_level)
{
if (mb_col > 0)
pc->lf_mbv(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
if (xd->mode_info_context->mbmi.dc_diff > 0)
pc->lf_bv(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* don't apply across umv border */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (mb_row > 0)
pc->lf_mbh(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
if (xd->mode_info_context->mbmi.dc_diff > 0)
pc->lf_bh(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
}
}
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
recon_yoffset += 16;
recon_uvoffset += 8;
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
++xd->mode_info_context; /* next mb */
Initial WebM release
2010-05-18 17:58:33 +02:00
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
xd->above_context++;
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*pbi->mb_row_di[ithread].current_mb_col = mb_col;*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->mt_current_mb_col[mb_row] = mb_col;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* adjust to the next row of mbs */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->common.filter_level)
{
if(mb_row != pc->mb_rows-1)
{
int lasty = pc->yv12_fb[ref_fb_idx].y_width + VP8BORDERINPIXELS;
int lastuv = (pc->yv12_fb[ref_fb_idx].y_width>>1) + (VP8BORDERINPIXELS>>1);
for (i = 0; i < 4; i++)
{
pbi->mt_yabove_row[mb_row +1][lasty + i] = pbi->mt_yabove_row[mb_row +1][lasty -1];
pbi->mt_uabove_row[mb_row +1][lastuv + i] = pbi->mt_uabove_row[mb_row +1][lastuv -1];
pbi->mt_vabove_row[mb_row +1][lastuv + i] = pbi->mt_vabove_row[mb_row +1][lastuv -1];
}
}
} else
vp8_extend_mb_row(&pc->yv12_fb[dst_fb_idx], xd->dst.y_buffer + 16, xd->dst.u_buffer + 8, xd->dst.v_buffer + 8);
Initial WebM release
2010-05-18 17:58:33 +02:00
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
++xd->mode_info_context; /* skip prediction column */
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* since we have multithread */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->mode_info_context += xd->mode_info_stride * pbi->decoding_thread_count;
Initial WebM release
2010-05-18 17:58:33 +02:00
}
}
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* add this to each frame */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
if ((mbrd->mb_row == pbi->common.mb_rows-1) || ((mbrd->mb_row == pbi->common.mb_rows-2) && (pbi->common.mb_rows % (pbi->decoding_thread_count+1))==1))
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*SetEvent(pbi->h_event_end_decoding);*/
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
sem_post(&pbi->h_event_end_decoding);
}
}
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
(void) p_data;
#endif
return 0 ;
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
void vp8_decoder_create_threads(VP8D_COMP *pbi)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
#if CONFIG_MULTITHREAD
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int core_count = 0;
int ithread;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->b_multithreaded_rd = 0;
pbi->allocated_decoding_thread_count = 0;
core_count = (pbi->max_threads > 16) ? 16 : pbi->max_threads;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (core_count > 1)
{
pbi->b_multithreaded_rd = 1;
pbi->decoding_thread_count = core_count -1;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->h_decoding_thread, vpx_malloc(sizeof(pthread_t) * pbi->decoding_thread_count));
CHECK_MEM_ERROR(pbi->h_event_start_decoding, vpx_malloc(sizeof(sem_t) * pbi->decoding_thread_count));
CHECK_MEM_ERROR(pbi->mb_row_di, vpx_memalign(32, sizeof(MB_ROW_DEC) * pbi->decoding_thread_count));
vpx_memset(pbi->mb_row_di, 0, sizeof(MB_ROW_DEC) * pbi->decoding_thread_count);
CHECK_MEM_ERROR(pbi->de_thread_data, vpx_malloc(sizeof(DECODETHREAD_DATA) * pbi->decoding_thread_count));
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
for (ithread = 0; ithread < pbi->decoding_thread_count; ithread++)
{
sem_init(&pbi->h_event_start_decoding[ithread], 0, 0);
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->de_thread_data[ithread].ithread = ithread;
pbi->de_thread_data[ithread].ptr1 = (void *)pbi;
pbi->de_thread_data[ithread].ptr2 = (void *) &pbi->mb_row_di[ithread];
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pthread_create(&pbi->h_decoding_thread[ithread], 0, vp8_thread_decoding_proc, (&pbi->de_thread_data[ithread]));
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
sem_init(&pbi->h_event_end_decoding, 0, 0);
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->allocated_decoding_thread_count = pbi->decoding_thread_count;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
#else
(void) pbi;
#endif
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
void vp8mt_de_alloc_temp_buffers(VP8D_COMP *pbi, int mb_rows)
{
#if CONFIG_MULTITHREAD
VP8_COMMON *const pc = & pbi->common;
int i;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->b_multithreaded_rd)
{
if (pbi->mt_current_mb_col)
{
vpx_free(pbi->mt_current_mb_col);
pbi->mt_current_mb_col = NULL ;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Free above_row buffers. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_yabove_row)
{
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_yabove_row[i])
{
vpx_free(pbi->mt_yabove_row[i]);
pbi->mt_yabove_row[i] = NULL ;
}
}
vpx_free(pbi->mt_yabove_row);
pbi->mt_yabove_row = NULL ;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_uabove_row)
{
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_uabove_row[i])
{
vpx_free(pbi->mt_uabove_row[i]);
pbi->mt_uabove_row[i] = NULL ;
}
}
vpx_free(pbi->mt_uabove_row);
pbi->mt_uabove_row = NULL ;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_vabove_row)
{
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_vabove_row[i])
{
vpx_free(pbi->mt_vabove_row[i]);
pbi->mt_vabove_row[i] = NULL ;
}
}
vpx_free(pbi->mt_vabove_row);
pbi->mt_vabove_row = NULL ;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Free left_col buffers. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_yleft_col)
{
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_yleft_col[i])
{
vpx_free(pbi->mt_yleft_col[i]);
pbi->mt_yleft_col[i] = NULL ;
}
}
vpx_free(pbi->mt_yleft_col);
pbi->mt_yleft_col = NULL ;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_uleft_col)
{
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_uleft_col[i])
{
vpx_free(pbi->mt_uleft_col[i]);
pbi->mt_uleft_col[i] = NULL ;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vpx_free(pbi->mt_uleft_col);
pbi->mt_uleft_col = NULL ;
Initial WebM release
2010-05-18 17:58:33 +02:00
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->mt_vleft_col)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
for (i=0; i< mb_rows; i++)
{
if (pbi->mt_vleft_col[i])
{
vpx_free(pbi->mt_vleft_col[i]);
pbi->mt_vleft_col[i] = NULL ;
}
}
vpx_free(pbi->mt_vleft_col);
pbi->mt_vleft_col = NULL ;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
}
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
(void) pbi;
Initial WebM release
2010-05-18 17:58:33 +02:00
#endif
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8mt_alloc_temp_buffers: make prototype return void This function was never called in a context expecting a return value, the return value was always a constant, and the !CONFIG_MULTITHREAD path didn't have a return statement, which caused a compiler warning. This patch changes the function to return void instead. Fixes issue #231 Change-Id: I9ef7f56e54418b7265026c54fc4ed5660c1418d1
2010-11-17 15:13:54 +01:00
void vp8mt_alloc_temp_buffers(VP8D_COMP *pbi, int width, int prev_mb_rows)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
#if CONFIG_MULTITHREAD
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
VP8_COMMON *const pc = & pbi->common;
int i;
int uv_width;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->b_multithreaded_rd)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8mt_de_alloc_temp_buffers(pbi, prev_mb_rows);
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* our internal buffers are always multiples of 16 */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if ((width & 0xf) != 0)
width += 16 - (width & 0xf);
Initial WebM release
2010-05-18 17:58:33 +02:00
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
if (width < 640) pbi->sync_range = 1;
else if (width <= 1280) pbi->sync_range = 8;
else if (width <= 2560) pbi->sync_range =16;
else pbi->sync_range = 32;
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
uv_width = width >>1;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Allocate an int for each mb row. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_current_mb_col, vpx_malloc(sizeof(int) * pc->mb_rows));
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Allocate memory for above_row buffers. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_yabove_row, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_yabove_row[i], vpx_calloc(sizeof(unsigned char) * (width + (VP8BORDERINPIXELS<<1)), 1));
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_uabove_row, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_uabove_row[i], vpx_calloc(sizeof(unsigned char) * (uv_width + VP8BORDERINPIXELS), 1));
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_vabove_row, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_vabove_row[i], vpx_calloc(sizeof(unsigned char) * (uv_width + VP8BORDERINPIXELS), 1));
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Allocate memory for left_col buffers. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_yleft_col, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_yleft_col[i], vpx_calloc(sizeof(unsigned char) * 16, 1));
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
CHECK_MEM_ERROR(pbi->mt_uleft_col, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_uleft_col[i], vpx_calloc(sizeof(unsigned char) * 8, 1));
CHECK_MEM_ERROR(pbi->mt_vleft_col, vpx_malloc(sizeof(unsigned char *) * pc->mb_rows));
for (i=0; i< pc->mb_rows; i++)
CHECK_MEM_ERROR(pbi->mt_vleft_col[i], vpx_calloc(sizeof(unsigned char) * 8, 1));
}
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
(void) pbi;
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
(void) width;
Initial WebM release
2010-05-18 17:58:33 +02:00
#endif
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
Initial WebM release
2010-05-18 17:58:33 +02:00
void vp8_decoder_remove_threads(VP8D_COMP *pbi)
{
#if CONFIG_MULTITHREAD
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* shutdown MB Decoding thread; */
Initial WebM release
2010-05-18 17:58:33 +02:00
if (pbi->b_multithreaded_rd)
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
int i;
Initial WebM release
2010-05-18 17:58:33 +02:00
pbi->b_multithreaded_rd = 0;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* allow all threads to exit */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (i = 0; i < pbi->allocated_decoding_thread_count; i++)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
sem_post(&pbi->h_event_start_decoding[i]);
pthread_join(pbi->h_decoding_thread[i], NULL);
Initial WebM release
2010-05-18 17:58:33 +02:00
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (i = 0; i < pbi->allocated_decoding_thread_count; i++)
{
sem_destroy(&pbi->h_event_start_decoding[i]);
Initial WebM release
2010-05-18 17:58:33 +02:00
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
sem_destroy(&pbi->h_event_end_decoding);
Initial WebM release
2010-05-18 17:58:33 +02:00
if (pbi->h_decoding_thread)
{
vpx_free(pbi->h_decoding_thread);
pbi->h_decoding_thread = NULL;
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
if (pbi->h_event_start_decoding)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
vpx_free(pbi->h_event_start_decoding);
pbi->h_event_start_decoding = NULL;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
if (pbi->mb_row_di)
{
vpx_free(pbi->mb_row_di);
pbi->mb_row_di = NULL ;
}
if (pbi->de_thread_data)
{
vpx_free(pbi->de_thread_data);
pbi->de_thread_data = NULL;
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
(void) pbi;
#endif
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
void vp8mt_lpf_init( VP8D_COMP *pbi, int default_filt_lvl)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
#if CONFIG_MULTITHREAD
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
VP8_COMMON *cm = &pbi->common;
MACROBLOCKD *mbd = &pbi->mb;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*YV12_BUFFER_CONFIG *post = &cm->new_frame;*/ /*frame_to_show;*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
loop_filter_info *lfi = cm->lf_info;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
FRAME_TYPE frame_type = cm->frame_type;
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*int mb_row;
int mb_col;
int baseline_filter_level[MAX_MB_SEGMENTS];*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int alt_flt_enabled = mbd->segmentation_enabled;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int i;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/*unsigned char *y_ptr, *u_ptr, *v_ptr;*/
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Note the baseline filter values for each segment */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (alt_flt_enabled)
{
for (i = 0; i < MAX_MB_SEGMENTS; i++)
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Abs value */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (mbd->mb_segement_abs_delta == SEGMENT_ABSDATA)
pbi->mt_baseline_filter_level[i] = mbd->segment_feature_data[MB_LVL_ALT_LF][i];
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Delta Value */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
else
{
pbi->mt_baseline_filter_level[i] = default_filt_lvl + mbd->segment_feature_data[MB_LVL_ALT_LF][i];
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
pbi->mt_baseline_filter_level[i] = (pbi->mt_baseline_filter_level[i] >= 0) ? ((pbi->mt_baseline_filter_level[i] <= MAX_LOOP_FILTER) ? pbi->mt_baseline_filter_level[i] : MAX_LOOP_FILTER) : 0; /* Clamp to valid range */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
}
}
}
else
{
for (i = 0; i < MAX_MB_SEGMENTS; i++)
pbi->mt_baseline_filter_level[i] = default_filt_lvl;
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Initialize the loop filter for this frame. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if ((cm->last_filter_type != cm->filter_type) || (cm->last_sharpness_level != cm->sharpness_level))
vp8_init_loop_filter(cm);
else if (frame_type != cm->last_frame_type)
vp8_frame_init_loop_filter(lfi, frame_type);
Initial WebM release
2010-05-18 17:58:33 +02:00
#else
(void) pbi;
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
(void) default_filt_lvl;
Initial WebM release
2010-05-18 17:58:33 +02:00
#endif
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
void vp8mt_decode_mb_rows( VP8D_COMP *pbi, MACROBLOCKD *xd)
Initial WebM release
2010-05-18 17:58:33 +02:00
{
#if CONFIG_MULTITHREAD
int mb_row;
VP8_COMMON *pc = &pbi->common;
int ibc = 0;
int num_part = 1 << pbi->common.multi_token_partition;
Remove unused local variables Removing unused local variables causing compiler warnings in Visual Studio. Change-Id: I0e2096303be1fdbc01428a6e57cca9796bb32c8a
2011-01-11 13:01:24 +01:00
int i;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
volatile int *last_row_current_mb_col = NULL;
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
int nsync = pbi->sync_range;
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
int filter_level;
loop_filter_info *lfi = pc->lf_info;
int alt_flt_enabled = xd->segmentation_enabled;
int Segment;
if(pbi->common.filter_level)
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Set above_row buffer to 127 for decoding first MB row */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vpx_memset(pbi->mt_yabove_row[0] + VP8BORDERINPIXELS-1, 127, pc->yv12_fb[pc->lst_fb_idx].y_width + 5);
vpx_memset(pbi->mt_uabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (pc->yv12_fb[pc->lst_fb_idx].y_width>>1) +5);
vpx_memset(pbi->mt_vabove_row[0] + (VP8BORDERINPIXELS>>1)-1, 127, (pc->yv12_fb[pc->lst_fb_idx].y_width>>1) +5);
for (i=1; i<pc->mb_rows; i++)
{
vpx_memset(pbi->mt_yabove_row[i] + VP8BORDERINPIXELS-1, (unsigned char)129, 1);
vpx_memset(pbi->mt_uabove_row[i] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
vpx_memset(pbi->mt_vabove_row[i] + (VP8BORDERINPIXELS>>1)-1, (unsigned char)129, 1);
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Set left_col to 129 initially */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
for (i=0; i<pc->mb_rows; i++)
{
vpx_memset(pbi->mt_yleft_col[i], (unsigned char)129, 16);
vpx_memset(pbi->mt_uleft_col[i], (unsigned char)129, 8);
vpx_memset(pbi->mt_vleft_col[i], (unsigned char)129, 8);
}
vp8mt_lpf_init(pbi, pc->filter_level);
}
Initial WebM release
2010-05-18 17:58:33 +02:00
vp8_setup_decoding_thread_data(pbi, xd, pbi->mb_row_di, pbi->decoding_thread_count);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
for (i = 0; i < pbi->decoding_thread_count; i++)
sem_post(&pbi->h_event_start_decoding[i]);
Initial WebM release
2010-05-18 17:58:33 +02:00
for (mb_row = 0; mb_row < pc->mb_rows; mb_row += (pbi->decoding_thread_count + 1))
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->current_bc = &pbi->mbc[mb_row%num_part];
Initial WebM release
2010-05-18 17:58:33 +02:00
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* vp8_decode_mb_row(pbi, pc, mb_row, xd); */
Initial WebM release
2010-05-18 17:58:33 +02:00
{
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
int i;
int recon_yoffset, recon_uvoffset;
int mb_col;
int ref_fb_idx = pc->lst_fb_idx;
int dst_fb_idx = pc->new_fb_idx;
int recon_y_stride = pc->yv12_fb[ref_fb_idx].y_stride;
int recon_uv_stride = pc->yv12_fb[ref_fb_idx].uv_stride;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* volatile int *last_row_current_mb_col = NULL; */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
if (mb_row > 0)
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
last_row_current_mb_col = &pbi->mt_current_mb_col[mb_row -1];
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
vpx_memset(&pc->left_context, 0, sizeof(pc->left_context));
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
recon_yoffset = mb_row * recon_y_stride * 16;
recon_uvoffset = mb_row * recon_uv_stride * 8;
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* reset above block coeffs */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
xd->above_context = pc->above_context;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->up_available = (mb_row != 0);
xd->mb_to_top_edge = -((mb_row * 16)) << 3;
xd->mb_to_bottom_edge = ((pc->mb_rows - 1 - mb_row) * 16) << 3;
for (mb_col = 0; mb_col < pc->mb_cols; mb_col++)
{
Adjust multi-thread sync ranges according to image sizes In multi-threaded decoder, set different sync ranges for different video resolutions. Change-Id: Iea48fd36f51919e0152c8ed3b1f10e1b723c0ca7
2010-09-23 19:53:09 +02:00
if ( mb_row > 0 && (mb_col & (nsync-1)) == 0){
while (mb_col > (*last_row_current_mb_col - nsync) && *last_row_current_mb_col != pc->mb_cols - 1)
Replace sleep(0) calls in multi-threaded decoder This is a workaround for gLucid problem. Change-Id: I188a016a07e4c2ea212444c5a6284ff3c48a5caa
2010-08-31 00:16:04 +02:00
{
x86_pause_hint();
thread_sleep(0);
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Initial WebM release
2010-05-18 17:58:33 +02:00
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
if (xd->mode_info_context->mbmi.mode == SPLITMV || xd->mode_info_context->mbmi.mode == B_PRED)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
{
for (i = 0; i < 16; i++)
{
BLOCKD *d = &xd->block[i];
vpx_memcpy(&d->bmi, &xd->mode_info_context->bmi[i], sizeof(B_MODE_INFO));
}
}
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if(pbi->common.filter_level)
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* update loopfilter info */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
Segment = (alt_flt_enabled) ? xd->mode_info_context->mbmi.segment_id : 0;
filter_level = pbi->mt_baseline_filter_level[Segment];
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Distance of Mb to the various image edges.
* These are specified to 8th pel as they are always compared to values that are in 1/8th pel units
* Apply any context driven MB level adjustment
*/
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8_adjust_mb_lf_value(xd, &filter_level);
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Distance of Mb to the various image edges.
* These are specified to 8th pel as they are always compared to values that are in 1/8th pel units
*/
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
xd->mb_to_left_edge = -((mb_col * 16) << 3);
xd->mb_to_right_edge = ((pc->mb_cols - 1 - mb_col) * 16) << 3;
xd->dst.y_buffer = pc->yv12_fb[dst_fb_idx].y_buffer + recon_yoffset;
xd->dst.u_buffer = pc->yv12_fb[dst_fb_idx].u_buffer + recon_uvoffset;
xd->dst.v_buffer = pc->yv12_fb[dst_fb_idx].v_buffer + recon_uvoffset;
xd->left_available = (mb_col != 0);
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Select the appropriate reference frame for this MB */
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
if (xd->mode_info_context->mbmi.ref_frame == LAST_FRAME)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
ref_fb_idx = pc->lst_fb_idx;
Removed unnecessary MB_MODE_INFO copies These copies occurred for each macroblock in the encoder and decoder. Thetemp MB_MODE_INFO mbmi was removed from MACROBLOCKD. As a result, a large number compile errors had to be fixed. Change-Id: I4cf0ffae3ce244f6db04a4c217d52dd256382cf3
2010-08-12 22:25:43 +02:00
else if (xd->mode_info_context->mbmi.ref_frame == GOLDEN_FRAME)
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
ref_fb_idx = pc->gld_fb_idx;
else
ref_fb_idx = pc->alt_fb_idx;
xd->pre.y_buffer = pc->yv12_fb[ref_fb_idx].y_buffer + recon_yoffset;
xd->pre.u_buffer = pc->yv12_fb[ref_fb_idx].u_buffer + recon_uvoffset;
xd->pre.v_buffer = pc->yv12_fb[ref_fb_idx].v_buffer + recon_uvoffset;
Implement error tracking in the decoder A new vpx_codec_control called VP8D_GET_FRAME_CORRUPTED. The output from the function is non-zero if the last decoded frame contains corruption due to packet losses. The decoder is also modified to accept encoded frames of zero length. A zero length frame indicates to the decoder that one or more frames have been completely lost. This will mark the last decoded reference buffer as corrupted. The data pointer can be NULL if the length is zero. Change-Id: Ic5902c785a281c6e05329deea958554b7a6c75ce
2010-12-16 16:46:31 +01:00
if (xd->mode_info_context->mbmi.ref_frame != INTRA_FRAME)
{
/* propagate errors from reference frames */
xd->corrupted |= pc->yv12_fb[ref_fb_idx].corrupted;
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
vp8_build_uvmvs(xd, pc->full_pixel);
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
vp8mt_decode_macroblock(pbi, xd, mb_row, mb_col);
Implement error tracking in the decoder A new vpx_codec_control called VP8D_GET_FRAME_CORRUPTED. The output from the function is non-zero if the last decoded frame contains corruption due to packet losses. The decoder is also modified to accept encoded frames of zero length. A zero length frame indicates to the decoder that one or more frames have been completely lost. This will mark the last decoded reference buffer as corrupted. The data pointer can be NULL if the length is zero. Change-Id: Ic5902c785a281c6e05329deea958554b7a6c75ce
2010-12-16 16:46:31 +01:00
/* check if the boolean decoder has suffered an error */
xd->corrupted |= vp8dx_bool_error(xd->current_bc);
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->common.filter_level)
{
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* Save decoded MB last row data for next-row decoding */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if(mb_row != pc->mb_rows-1)
{
vpx_memcpy((pbi->mt_yabove_row[mb_row +1] + 32 + mb_col*16), (xd->dst.y_buffer + 15 * recon_y_stride), 16);
vpx_memcpy((pbi->mt_uabove_row[mb_row +1] + 16 + mb_col*8), (xd->dst.u_buffer + 7 * recon_uv_stride), 8);
vpx_memcpy((pbi->mt_vabove_row[mb_row +1] + 16 + mb_col*8), (xd->dst.v_buffer + 7 * recon_uv_stride), 8);
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* save left_col for next MB decoding */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if(mb_col != pc->mb_cols-1)
{
MODE_INFO *next = xd->mode_info_context +1;
if (xd->frame_type == KEY_FRAME || next->mbmi.ref_frame == INTRA_FRAME)
{
for (i = 0; i < 16; i++)
pbi->mt_yleft_col[mb_row][i] = xd->dst.y_buffer [i* recon_y_stride + 15];
for (i = 0; i < 8; i++)
{
pbi->mt_uleft_col[mb_row][i] = xd->dst.u_buffer [i* recon_uv_stride + 7];
pbi->mt_vleft_col[mb_row][i] = xd->dst.v_buffer [i* recon_uv_stride + 7];
}
}
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* loopfilter on this macroblock. */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (filter_level)
{
if (mb_col > 0)
pc->lf_mbv(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (xd->mode_info_context->mbmi.dc_diff > 0)
pc->lf_bv(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* don't apply across umv border */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (mb_row > 0)
pc->lf_mbh(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
if (xd->mode_info_context->mbmi.dc_diff > 0)
pc->lf_bh(xd->dst.y_buffer, xd->dst.u_buffer, xd->dst.v_buffer, recon_y_stride, recon_uv_stride, &lfi[filter_level], pc->simpler_lpf);
}
}
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
recon_yoffset += 16;
recon_uvoffset += 8;
++xd->mode_info_context; /* next mb */
Changed above and left context data layout The main reason for the change was to reduce cycles in the token decoder. (~1.5% gain for 32 bit) This layout should be more cache friendly. As a result of this change, the encoder had to be updated. Change-Id: Id5e804169d8889da0378b3a519ac04dabd28c837 Note: dixie uses a similar layout
2010-08-31 16:49:57 +02:00
xd->above_context++;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
pbi->mt_current_mb_col[mb_row] = mb_col;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
/* adjust to the next row of mbs */
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
if (pbi->common.filter_level)
{
if(mb_row != pc->mb_rows-1)
{
int lasty = pc->yv12_fb[ref_fb_idx].y_width + VP8BORDERINPIXELS;
int lastuv = (pc->yv12_fb[ref_fb_idx].y_width>>1) + (VP8BORDERINPIXELS>>1);
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
for (i = 0; i < 4; i++)
{
pbi->mt_yabove_row[mb_row +1][lasty + i] = pbi->mt_yabove_row[mb_row +1][lasty -1];
pbi->mt_uabove_row[mb_row +1][lastuv + i] = pbi->mt_uabove_row[mb_row +1][lastuv -1];
pbi->mt_vabove_row[mb_row +1][lastuv + i] = pbi->mt_vabove_row[mb_row +1][lastuv -1];
}
}
}else
vp8_extend_mb_row(&pc->yv12_fb[dst_fb_idx], xd->dst.y_buffer + 16, xd->dst.u_buffer + 8, xd->dst.v_buffer + 8);
Initial WebM release
2010-05-18 17:58:33 +02:00
Restructure multi-threaded decoder On each MB, loopfiltering is done right after MB decoding. This combines two loops in multi-threaded code into one, which reduces number of synchronizations to half. The above-row/left-col data are saved in temp buffers for next-row/next MB decoding. Tests on 4-core gLucid machine showed 10% decoder performance gain with threads=4 (tulip clip). Testing on other platforms isn't done yet. Change-Id: Id18ea7c1e84965dabea65d4c01ca5bc056ddeac9
2010-09-16 20:08:52 +02:00
++xd->mode_info_context; /* skip prediction column */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Initial WebM release
2010-05-18 17:58:33 +02:00
xd->mode_info_context += xd->mode_info_stride * pbi->decoding_thread_count;
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
}
Eliminate more warnings. This eliminates a large set of warnings exposed by the Mozilla build system (Use of C++ comments in ISO C90 source, commas at the end of enum lists, a couple incomplete initializers, and signed/unsigned comparisons). It also eliminates many (but not all) of the warnings expose by newer GCC versions and _FORTIFY_SOURCE (e.g., calling fread and fwrite without checking the return values). There are a few spurious warnings left on my system: ../vp8/encoder/encodemb.c:274:9: warning: 'sz' may be used uninitialized in this function gcc seems to be unable to figure out that the value shortcut doesn't change between the two if blocks that test it here. ../vp8/encoder/onyx_if.c:5314:5: warning: comparison of unsigned expression >= 0 is always true ../vp8/encoder/onyx_if.c:5319:5: warning: comparison of unsigned expression >= 0 is always true This is true, so far as it goes, but it's comparing against an enum, and the C standard does not mandate that enums be unsigned, so the checks can't be removed. Change-Id: Iaf689ae3e3d0ddc5ade00faa474debe73b8d3395
2010-10-28 01:04:02 +02:00
sem_wait(&pbi->h_event_end_decoding); /* add back for each frame */
First modification of multi-thread decoder This is the first modification of VP8 multi-thread decoder, which uses same threads to decode macroblocks and then do loopfiltering for each frame. Inspired by Rob Clark, synchronization was done on every 8 macroblocks instead of every macroblock to reduce lock contention. Comparing with the original code, this implementation gave about 15%- 20% performance gain while decoding my test clips on a Core2 Quad platform (Linux). The work is not done yet. Test on other platforms are needed. Change-Id: Ice9ddb0b511af1359b9f71e65066143c04fef3b5
2010-07-29 22:24:26 +02:00
#else
(void) pbi;
(void) xd;
#endif
}
Reference in New Issue Copy Permalink