summaryrefslogtreecommitdiff
path: root/libavcodec
Commit message (Collapse)AuthorAgeFilesLines
* use left_xy[1] in mbaff QP loop filter check, this improves the amount that canMichael Niedermayer2010-01-261-1/+1
| | | | | | be skiped. Originally committed as revision 21465 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Optimize mv/ref cache init for left MB.Michael Niedermayer2010-01-261-8/+30
| | | | Originally committed as revision 21464 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Add UMULH()Måns Rullgård2010-01-261-0/+6
| | | | | | | This function/macro returns the high half of the unsigned 64-bit product of two unsigned 32-bit operands. Originally committed as revision 21463 to svn://svn.ffmpeg.org/ffmpeg/trunk
* optimize h264_loop_filter_strength_mmx2Loren Merritt2010-01-261-74/+70
| | | | | | 244->160 cycles on core2 Originally committed as revision 21462 to svn://svn.ffmpeg.org/ffmpeg/trunk
* remove unneeded checkMichael Niedermayer2010-01-261-2/+0
| | | | Originally committed as revision 21460 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Use left_mb_xy from fill_caches instead of recalculating it.Michael Niedermayer2010-01-261-5/+3
| | | | Originally committed as revision 21459 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Simplify left_xy content for the loop filter, this also makes it closer toMichael Niedermayer2010-01-261-0/+10
| | | | | | what is needed and its faster too. Originally committed as revision 21458 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Simplify loop filter a little by using top/left_type.Michael Niedermayer2010-01-261-11/+9
| | | | Originally committed as revision 21457 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Set top & left types for deblock in fill_caches().Michael Niedermayer2010-01-261-4/+13
| | | | Originally committed as revision 21456 to svn://svn.ffmpeg.org/ffmpeg/trunk
* 100l typo fix, mixed up +-1 forcing the loop filter skip to never skip.Michael Niedermayer2010-01-261-1/+1
| | | | Originally committed as revision 21455 to svn://svn.ffmpeg.org/ffmpeg/trunk
* cosmeticMichael Niedermayer2010-01-261-1/+1
| | | | Originally committed as revision 21454 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Fix qp_thres loop filter check for MBAFF.Michael Niedermayer2010-01-261-2/+5
| | | | Originally committed as revision 21453 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Optimize mb neighbor initialization for MBAFF in fill_caches().Michael Niedermayer2010-01-261-32/+29
| | | | | | ~10 cpu cycles speedup. Originally committed as revision 21452 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Metasoft MJPEG codec has Cb and Cr swapped, fixes issue 1611.Carl Eugen Hoyos2010-01-261-0/+4
| | | | Originally committed as revision 21451 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Fix heap overflow due to lack of nb_components check.Michael Niedermayer2010-01-251-0/+4
| | | | Originally committed as revision 21450 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Release buffer when decoding is finished.Carl Eugen Hoyos2010-01-251-1/+10
| | | | Originally committed as revision 21449 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Optimize decode_cabac_field_decoding_flag().Michael Niedermayer2010-01-251-12/+6
| | | | | | ~4 cpu cycles faster Originally committed as revision 21447 to svn://svn.ffmpeg.org/ffmpeg/trunk
* 100l, correct wrong H.264+adaptive MBAFF decodingAlexander Strange2010-01-251-2/+5
| | | | | | Fixes several FATE tests. Originally committed as revision 21445 to svn://svn.ffmpeg.org/ffmpeg/trunk
* H.264: Use 64-/128-bit write-combining macros for copiesAlexander Strange2010-01-252-52/+54
| | | | | | 2-3% faster decode on x86-32 core2. Originally committed as revision 21440 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Only check frame size if the header is valid.Justin Ruggles2010-01-241-1/+1
| | | | Originally committed as revision 21439 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Fix decoding ATI VCR2 codec, sample inCarl Eugen Hoyos2010-01-241-0/+3
| | | | | | | | | samples/V-codecs/ATI-VCR12/ATI/VCR2test.avi Software decoding was broken since r8865, when two new context variables were added, but not initialized in vcr2_init_sequence(). Another such context variable was added in r18381. Originally committed as revision 21438 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Branchless setting of MB_TYPE_8x8DCT.Michael Niedermayer2010-01-241-3/+1
| | | | | | Not benchmarked as i failed to find a sample that uses this one. But it should be faster. Originally committed as revision 21435 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove cruft.Michael Niedermayer2010-01-241-2/+0
| | | | Originally committed as revision 21434 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Mark index as unsigend so gcc knows it doesnt need a sign extension on x86_64.Michael Niedermayer2010-01-241-4/+6
| | | | | | | | Any tips on how i can convince gcc that it doesnt need a mov %eax, %eax in every get_bits() ? Originally committed as revision 21433 to svn://svn.ffmpeg.org/ffmpeg/trunk
* a[b-1] -> (a-1)[b].Michael Niedermayer2010-01-241-4/+4
| | | | | | Helps gcc not to add seperate -1 instructions. Originally committed as revision 21432 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Optimize suffix_length computation, 1 cpu cycle speedup.Michael Niedermayer2010-01-241-3/+1
| | | | Originally committed as revision 21431 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Simplify suffix_length computation, same speed.Michael Niedermayer2010-01-241-3/+1
| | | | Originally committed as revision 21430 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Optimize level_code computation, 6cpu cycles speedup.Michael Niedermayer2010-01-241-1/+1
| | | | Originally committed as revision 21428 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Fix crash in MLP decoder due to integer overflow.Reimar Döffinger2010-01-241-1/+1
| | | | | | | Probably only DoS, init_get_bits sets buffer to NULL, thus causing a NULL-dereference directly after. Originally committed as revision 21426 to svn://svn.ffmpeg.org/ffmpeg/trunk
* 1 cpu cycle faster suffix_length calculation.Michael Niedermayer2010-01-241-1/+1
| | | | Originally committed as revision 21425 to svn://svn.ffmpeg.org/ffmpeg/trunk
* VC-1/WMV3 DXVA2 implementationLaurent Aimar2010-01-243-0/+294
| | | | | | | | It allows VLD VC-1/WMV3 decoding using DXVA2 (GPU assisted decoding API under VISTA and Windows 7). It is implemented by using AVHWAccel API. Originally committed as revision 21424 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove is_complex from loop_filter() its useless there in its current form.Michael Niedermayer2010-01-241-5/+3
| | | | Originally committed as revision 21423 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Added a few missing consts in dxva_h264.Laurent Aimar2010-01-241-5/+5
| | | | Originally committed as revision 21422 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Added a missing const to ff_h264_get_slice_type().Laurent Aimar2010-01-242-2/+2
| | | | Originally committed as revision 21421 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Added missing const to get_bits_count().Laurent Aimar2010-01-241-3/+3
| | | | Originally committed as revision 21419 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Prefer cbp over cbp_table.Michael Niedermayer2010-01-241-4/+4
| | | | Originally committed as revision 21418 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove all uses of slice_type* from the loop filter, also remove itsMichael Niedermayer2010-01-242-11/+5
| | | | | | initialization befre the loop filter. Originally committed as revision 21416 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove unneeded reset of non_zero_count_cache for deblock.Michael Niedermayer2010-01-231-13/+2
| | | | Originally committed as revision 21414 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove useless things from the deblock side of fill_caches().Michael Niedermayer2010-01-231-11/+3
| | | | Originally committed as revision 21413 to svn://svn.ffmpeg.org/ffmpeg/trunk
* make mv_cache init 64bit where possible.Michael Niedermayer2010-01-231-8/+4
| | | | Originally committed as revision 21412 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Moved reusable functions from dxva2_h264.c to dxva2.cLaurent Aimar2010-01-234-133/+205
| | | | Originally committed as revision 21407 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Move +52 from the loop filter to the alpha/beta offsets in the context.Michael Niedermayer2010-01-234-25/+30
| | | | | | | This should fix a segfault, also it might be faster on systems where the +52 wasnt free. Originally committed as revision 21406 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Release buffer when the codec is closed.Reimar Döffinger2010-01-231-0/+2
| | | | Originally committed as revision 21405 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Use qstride = 0 and allocate only a single line for qscale_table sinceReimar Döffinger2010-01-231-3/+3
| | | | | | for this codecs all quants are the same for a single frame. Originally committed as revision 21404 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Added ff_dxva2_ prefix to get_surface(_index) and commit_buffer functions.Laurent Aimar2010-01-231-22/+22
| | | | Originally committed as revision 21403 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Reindent the content of one if(){} in ff_dxva2_common_end_frame.Laurent Aimar2010-01-231-9/+9
| | | | Originally committed as revision 21402 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Moved reusable code from dxva2_h264.c:end_frame to ff_dxva2_common_end_frame.Laurent Aimar2010-01-231-21/+37
| | | | Originally committed as revision 21401 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Set edges based on cbp and mv partitioning, not just skiped MBs.Michael Niedermayer2010-01-231-2/+2
| | | | | | This is faster for videos that have lots of MBs that fall in this category. Originally committed as revision 21400 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Use AVPALETTE_COUNT instead of hard-coding 256.Reimar Döffinger2010-01-231-1/+1
| | | | | | | This is consistent with other codecs and will also avoid a crash on the memcpy to data[1] if AVPALETTE_SIZE ever increases. Originally committed as revision 21399 to svn://svn.ffmpeg.org/ffmpeg/trunk
* Remove redundant initialization of the palette, it is part of the contextReimar Döffinger2010-01-231-2/+0
| | | | | | and that one is always 0-initialized already. Originally committed as revision 21398 to svn://svn.ffmpeg.org/ffmpeg/trunk
View Lorry Controller Status