summaryrefslogtreecommitdiffstats
path: root/libavcodec/x86
Commit message (Collapse)AuthorAgeFilesLines
* x86: dsputil: Rename dsputil_mmx.h --> dsputil_x86.hDiego Biurrun2013-05-1221-20/+20
| | | | The header is not (anymore) MMX-specific.
* x86: dsputil: Split inline assembly from init codeDiego Biurrun2013-05-124-727/+759
| | | | Also remove some pointless comments.
* x86: dsputil: Refactor pixels16 wrapper functions with a macroDiego Biurrun2013-05-128-143/+41
|
* configure: Rename cmov processor capability to i686Diego Biurrun2013-05-121-4/+2
| | | | | The goal is to make the capapility slightly more general and have it cover the availability of the nopl instruction in addition to cmov.
* x86: sbrdsp: implement SSE2 qmf_pre_shuffleChristophe Gisquet2013-05-102-0/+32
| | | | | | | From 253 to 51 cycles on Arrandale and Win64. 44 cycles on SandyBridge. Signed-off-by: Anton Khirnov <anton@khirnov.net>
* x86: dsputil: Remove unused argument from QPEL_OP macroDiego Biurrun2013-05-081-4/+4
|
* x86: dsputil: Move TRANSPOSE4 macro to the only place it is usedDiego Biurrun2013-05-082-11/+11
|
* x86: dsputil: Move constant declarations into separate headerDiego Biurrun2013-05-086-24/+57
|
* x86: dsputil: Group all assembly constants together in constants.cDiego Biurrun2013-05-082-15/+11
|
* x86: dsputil: Move ff_pd assembly constants to the only place they are usedDiego Biurrun2013-05-083-13/+11
|
* x86: dsputil: Remove unused ff_pb_3F constantDiego Biurrun2013-05-072-2/+0
|
* x86: dsputil: Remove unused MOVQ_BONE macroDiego Biurrun2013-05-072-9/+0
|
* x86: dsputil: Move rv40-specific functions where they belongDiego Biurrun2013-05-073-26/+27
|
* x86: dsputil hpeldsp: Move shared template functions into separate objectDiego Biurrun2013-05-077-26/+69
|
* x86: rnd_template: Eliminate pointless OP_AVG macro indirectionDiego Biurrun2013-05-074-12/+8
|
* x86: hpeldsp: Move avg_pixels8_x2_mmx() out of hpeldsp_rnd_template.cDiego Biurrun2013-05-065-25/+58
| | | | | The function is only instantiated once, so there is no point in keeping it in a template file.
* x86: hpeldsp: Only compile MMX hpeldsp code if MMX is enabledDiego Biurrun2013-05-061-2/+2
|
* x86: More specific ifdefs for dsputil/hpeldsp init functionsDiego Biurrun2013-05-062-16/+16
|
* avcodec: Add av_cold attributes to init functions missing themDiego Biurrun2013-05-042-2/+4
|
* silly typo fixesDiego Biurrun2013-05-031-1/+1
|
* x86: sbrdsp: Implement SSE2 qmf_deint_bflyChristophe Gisquet2013-05-032-0/+33
| | | | | | | | | | Sandybridge: 47 cycles Having a loop counter is a 7 cycle gain. Unrolling is another 7 cycle gain. Working in reverse scan is another 6 cycles. Signed-off-by: Diego Biurrun <diego@biurrun.de>
* x86: dsputil: Move cavs and vc1-specific functions where they belongDiego Biurrun2013-05-024-40/+35
|
* x86: dsputil: Move avg_pixels16_mmx() out of rnd_template.cDiego Biurrun2013-05-025-24/+29
| | | | | The function does not do any rounding, so there is no point in keeping it in a round template file.
* x86: dsputil: Move avg_pixels8_mmx() out of rnd_template.cDiego Biurrun2013-05-025-23/+25
| | | | | The function is only instantiated once, so there is no point in keeping it in a template file.
* x86: Move duplicated put_pixels{8|16}_mmx functions into their own fileDiego Biurrun2013-05-025-134/+109
|
* x86: Drop unnecessary ff_ name prefixes from static functionsDiego Biurrun2013-04-305-53/+60
|
* mpegaudiosp: More consistent names for ppc/x86 optimization filesDiego Biurrun2013-04-302-1/+1
|
* x86: dsputil: Remove a set of pointless #ifs around function declarationsDiego Biurrun2013-04-301-2/+0
|
* x86: dsputil: cosmetics: Group ff_{avg|put}_pixels16_mmxext() declarationsDiego Biurrun2013-04-301-28/+14
|
* x86: hpeldsp: Remove unused macro definitionsDiego Biurrun2013-04-291-7/+0
|
* x86: ac3dsp: Remove 3dnow version of ff_ac3_extract_exponentsDiego Biurrun2013-04-262-37/+0
| | | | | | | The function requires increasing the fuzz factor for the ac3/eac3 encode tests and even so makes fate fail. It only provides a slight encoding speedup for legacy CPUs that do not support SS2. Thus its benefit is not worth the trouble it creates and fixing it would be a waste of time.
* x86: Rename dsputil_rnd_template.c to rnd_template.cMartin Storsjö2013-04-253-2/+2
| | | | | | | This makes it less confusing when this template is shared both by dsputil and by hpeldsp. Signed-off-by: Martin Storsjö <martin@martin.st>
* x86: Get rid of duplication between *_rnd_template.cMartin Storsjö2013-04-232-197/+5
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* x86: Factorize duplicated inline assembly snippetsMartin Storsjö2013-04-233-130/+76
| | | | Signed-off-by: Diego Biurrun <diego@biurrun.de>
* x86: Move some conditional code around to avoid unused variable warningsDiego Biurrun2013-04-223-17/+15
|
* x86: cavs: Refactor duplicate dspfunc macroDiego Biurrun2013-04-221-22/+14
|
* x86: cavs: Put mmx-specific code into its own init functionDiego Biurrun2013-04-223-15/+31
| | | | | Before, this code was labeled as mmxext and enabled both for the 3dnow and the mmxext case.
* x86: Remove some duplicate function declarationsDiego Biurrun2013-04-222-7/+0
|
* x86: Remove unused inline asm instruction definesMartin Storsjö2013-04-201-3/+0
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* x86: hpeldsp: Move half-pel assembly from dsputil to hpeldspRonald S. Bultje2013-04-198-657/+957
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* vp3: Use full transpose for all IDCTsRonald S. Bultje2013-04-152-43/+82
| | | | | | | | | | | This way, the special IDCT permutations are no longer needed. This is similar to how H264 does it, and removes the dsputil dependency imposed by the scantable code. Also remove the unused type == 0 cases from the plain C version of the idct. Signed-off-by: Martin Storsjö <martin@martin.st>
* x86: Move constants to the only place where they are usedRonald S. Bultje2013-04-153-9/+4
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* x86: dsputil: Move some ifdefs to avoid unused variable warningsDiego Biurrun2013-04-121-2/+2
|
* x86: dsputil: cosmetics: Remove two pointless variable indirectionsDiego Biurrun2013-04-121-4/+2
|
* x86: dsputil: Refactor some ff_{avg|put}_pixels function declarationsDiego Biurrun2013-04-123-15/+6
|
* x86: dsputil: ff_h263_*_loop_filter declarations to a more suitable placeDiego Biurrun2013-04-121-5/+3
|
* x86: h264qpel: int --> ptrdiff_t for some line_size parametersDiego Biurrun2013-04-122-6/+9
|
* Move misplaced file author information where it belongsDiego Biurrun2013-04-112-4/+4
|
* dsputil: Make dsputil selectableRonald S. Bultje2013-04-107-49/+60
| | | | Signed-off-by: Martin Storsjö <martin@martin.st>
* h264: Integrate clear_blocks calls with IDCTRonald S. Bultje2013-04-102-30/+131
| | | | | | | | | The non-intra-pcm branch in hl_decode_mb (simple, 8bpp) goes from 700 to 672 cycles, and the complete loop of decode_mb_cabac and hl_decode_mb (in the decode_slice loop) goes from 1759 to 1733 cycles on the clip tested (cathedral), i.e. almost 30 cycles per mb faster. Signed-off-by: Martin Storsjö <martin@martin.st>
OpenPOWER on IntegriCloud