summaryrefslogtreecommitdiffstats
path: root/libavutil/x86
Commit message (Collapse)AuthorAgeFilesLines
* x86inc: Various minor backports from x264Henrik Gramner2015-08-131-11/+21
| | | | Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: Drop SECTION_TEXT macroHenrik Gramner2015-08-111-12/+0
| | | | | | | The .text section is already 16-byte aligned by default on all supported platforms so `SECTION_TEXT` isn't any different from `SECTION .text`. Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: Disable vpbroadcastq workaround in newer yasm versionsHenrik Gramner2015-08-111-9/+11
| | | | | | The bug was fixed in 1.3.0, so only perform the workaround in earlier versions. Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: Fix instantiation of YMM registersChristophe Gisquet2015-08-111-1/+1
| | | | | Signed-off-by: Henrik Gramner <[email protected]> Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: warn when instructions incompatible with current cpuflags are usedAnton Mitrofanov2015-08-111-288/+299
| | | | | Signed-off-by: Henrik Gramner <[email protected]> Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: Support arbitrary stack alignmentsHenrik Gramner2015-08-111-22/+40
| | | | | | | | Change ALLOC_STACK to always align the stack before allocating stack space for consistency. Previously alignment would occur either before or after allocating stack space depending on whether manual alignment was required or not. Signed-off-by: Anton Khirnov <[email protected]>
* x86inc: warn if XOP integer FMA instruction emulation is impossibleAnton Mitrofanov2015-08-111-2/+5
| | | | | | | | | | | Emulation requires a temporary register if arguments 1 and 4 are the same; this doesn't obey the semantics of the original instruction, so we can't emulate that in x86inc. Also add pmacsdql emulation. Signed-off-by: Henrik Gramner <[email protected]> Signed-off-by: Anton Khirnov <[email protected]>
* x86: Serialize rdtsc in read_time()Henrik Gramner2015-07-091-1/+6
| | | | | | | | | | | | | | | | | Improves the accuracy of measurements, especially in short sections. To quote the Intel 64 and IA-32 Architectures Software Developer's Manual: "The RDTSC instruction is not a serializing instruction. It does not necessarily wait until all previous instructions have been executed before reading the counter. Similarly, subsequent instructions may begin execution before the read operation is performed. If software requires RDTSC to be executed only after all previous instructions have completed locally, it can either use RDTSCP (if the processor supports that instruction) or execute the sequence LFENCE;RDTSC." SSE2 is a requirement for lfence so only use it on SSE2-capable systems. Prefer lfence;rdtsc over rdtscp since rdtscp is supported on fewer systems. Signed-off-by: Luca Barbato <[email protected]>
* x86: check for AV_CPU_FLAG_AVXSLOW where usefulJames Almer2015-05-312-2/+2
| | | | | Signed-off-by: James Almer <[email protected]> Signed-off-by: Luca Barbato <[email protected]>
* x86: Add helper macros to check for slow cpuflagsJames Almer2015-05-311-0/+18
| | | | | Signed-off-by: James Almer <[email protected]> Signed-off-by: Luca Barbato <[email protected]>
* x86: add AV_CPU_FLAG_AVXSLOW flagJames Almer2015-05-311-3/+14
| | | | | Signed-off-by: James Almer <[email protected]> Signed-off-by: Luca Barbato <[email protected]>
* x86inc: Clear __SECT__Timothy Gu2015-05-281-1/+5
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | Silences warning(s) like: libavcodec/x86/fft.asm:93: warning: section flags ignored on section redeclaration The cause of this warning is that because `struc` and `endstruc` attempts to revert to the previous section state [1]. The section state is stored in the macro __SECT__, defined by x86inc.asm to be `.note.GNU-stack ...`, through the `SECTION` directive [2]. Thus, the `.note.GNU-stack` section is defined twice (once in x86inc.asm, once during `endstruc`), causing the warning. That is the first part of the commit: using the primitive `[section]` format for .note.GNU-stack etc., which does not update `__SECT__` [2]. That fixes only half of the problem. Even without any `SECTION` directives, `__SECT__` is predefined as `.text`, which conflicting with the later `SECTION_TEXT` (which expands to `.text align=16`). [1]: http://www.nasm.us/doc/nasmdoc6.html#section-6.4 [2]: http://www.nasm.us/doc/nasmdoc6.html#section-6.3 Signed-off-by: Luca Barbato <[email protected]>
* v210enc: Add SIMD optimised 8-bit and 10-bit encodersKieran Kunhya2014-12-051-0/+5
| | | | | Signed-off-by: Michael Niedermayer <[email protected]> Signed-off-by: Vittorio Giovara <[email protected]>
* x86inc: Make INIT_CPUFLAGS support an arbitrary number of cpuflagsHenrik Gramner2014-09-091-19/+22
| | | | | | Previously there was a limit of two cpuflags. Signed-off-by: Diego Biurrun <[email protected]>
* x86inc: Free up variable name "n" in global namespaceLoren Merritt2014-09-091-9/+9
| | | | Signed-off-by: Diego Biurrun <[email protected]>
* x86inc: Make ym# behave the same way as xm#Henrik Gramner2014-09-091-4/+4
| | | | | | This makes more sense for future implementations of templates with zmm registers. Signed-off-by: Diego Biurrun <[email protected]>
* Update Fiona's name in copyright statements.Diego Biurrun2014-07-012-2/+2
|
* x86: add detection for Bit Manipulation Instruction setsJames Almer2014-02-231-6/+11
| | | | | | Based on x264 code Signed-off-by: James Almer <[email protected]>
* x86: add detection for FMA3 instruction setJames Almer2014-02-232-1/+7
| | | | | | Based on x264 code Signed-off-by: James Almer <[email protected]>
* x86: add missing XOP checks and macrosJames Almer2014-02-231-0/+3
| | | | Signed-off-by: James Almer <[email protected]>
* x86: float dsp: unroll SSE versionsChristophe Gisquet2014-02-201-16/+24
| | | | | | | | | | vector_fmul and vector_fmac_scalar are guaranteed that they can process in batch of 16 elements, but their SSE versions only does 8 at a time. Therefore, unroll them a bit. 299 to 261c for 256 elements in vector_fmac_scalar on Arrandale/Win64. Signed-off-by: Janne Grunau <[email protected]>
* x86inc: Speed up assembling with YasmLoren Merritt2014-01-261-23/+23
| | | | | | | Work around Yasm's inefficiency with handling large numbers of variables in the global scope. Signed-off-by: Diego Biurrun <[email protected]>
* libavutil: x86: Add AVX2 capable CPU detection.Kieran Kunhya2013-10-252-0/+11
| | | | | | Patch based on x264's AVX2 detection Signed-off-by: Derek Buitenhuis <[email protected]>
* x86: more AVX2 frameworkJason Garrett-Glaser2013-10-141-0/+11
| | | | Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: FMA3/4 SupportJason Garrett-Glaser2013-10-142-1/+45
| | | | Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Remove our FMA4 supportDerek Buitenhuis2013-10-142-11/+16
| | | | | | | | This is so we can sync to x264's version of FMA4 support. This partialy reverts commit 79687079a97a039c325ab79d7a95920d800b791f. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Use VEX-encoded instructions in AVX functionsHenrik Gramner2013-10-141-84/+169
| | | | | | | | | | | | | Automatically use VEX-encoding in AVX/AVX2/XOP/FMA3/FMA4 functions for all instructions that exists in a VEX-encoded version. This change makes it easier to extend existing code to use AVX2. Also add support for AVX emulation of a few instructions that were missing before. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Remove .rodata kludgesHenrik Gramner2013-10-091-11/+5
| | | | | | | The Mach-O bug was fixed in yasm 0.8.0 and we don't support versions that old anymore. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: remove misaligned cpu flagHenrik Gramner2013-10-071-5/+4
| | | | | | | | | | | | Prevents a crash if the misaligned exception mask bit is cleared for some reason. Misaligned SSE functions are only used on AMD Phenom CPUs and the benefit is miniscule. They also require modifying the MXCSR control register and by removing those functions we can get rid of that complexity altogether. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: various minor backports from x264Jason Garrett-Glaser2013-10-071-4/+3
| | | | | | Small backports that sneaked into other asm commits in x264. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Check for __OUTPUT_FORMAT__ having a value of "x64"Derek Buitenhuis2013-10-071-0/+2
| | | | | | This is also a valid value for WIN64. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Utilize the shadow space on 64-bit WindowsHenrik Gramner2013-10-071-37/+46
| | | | | | | | | Store XMM6 and XMM7 in the shadow space in functions that clobbers them. This way we don't have to adjust the stack pointer as often, reducing the number of instructions as well as code size. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: create xm# and ym#, analagous to m#Loren Merritt2013-10-071-1/+25
| | | | | | For when we want to mix simd sizes within one function. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: fix some corner cases of SWAPLoren Merritt2013-10-071-27/+27
| | | | | | | | SWAP with >=3 named (rather than numbered) args PERMUTE followed by SWAP with 2 named args used to produce the wrong permutation Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Use SSE instead of SSE2 for copying dataHenrik Gramner2013-10-071-2/+2
| | | | | | | Reduces code size because movaps/movups is one byte shorter than movdqa/movdqu. Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: Set ELF hidden visibility for global constantsHenrik Gramner2013-10-071-1/+5
| | | | Signed-off-by: Derek Buitenhuis <[email protected]>
* x86inc: activate REP_RET automaticallyLoren Merritt2013-10-071-5/+31
| | | | | | | | | | | | Now RET checks whether it immediately follows a branch, so the programmer dosen't have to keep track of that condition. REP_RET is still needed manually when it's a branch target, but that's much rarer. The implementation involves lots of spurious labels, but that's OK because we strip them. Signed-off-by: Derek Buitenhuis <[email protected]>
* avutil: Fix compilation with inline asm disabled on mingwAlex Smith2013-09-221-0/+1
| | | | | | Because of -Werror=implicit-function-declaration the build will fail. Signed-off-by: Martin Storsjö <[email protected]>
* x86: Add and use more convenience macros to check CPU extension availabilityDiego Biurrun2013-08-291-0/+13
|
* avutil: Refactor CPU extension availability macrosDiego Biurrun2013-08-281-28/+26
|
* avutil: Move internal CPU detection function declarations to private headerDiego Biurrun2013-08-281-0/+1
|
* Consistently use "cpu_flags" as variable/parameter name for CPU flagsDiego Biurrun2013-07-181-6/+6
|
* lls/x86: use 3-operator vaddpd in ADDPD_MEMLoren Merritt2013-07-021-1/+1
| | | | | | Fixes build with yasm-1.1 Signed-off-by: Anton Khirnov <[email protected]>
* x86: lpc: fix a segfault in av_evaluate_lls_sse2()Loren Merritt2013-06-301-1/+1
|
* x86: lpc: simd av_evaluate_llsLoren Merritt2013-06-292-0/+41
| | | | | | 1.5x-1.8x faster on sandybridge Signed-off-by: Luca Barbato <[email protected]>
* x86: lpc: simd av_update_llsLoren Merritt2013-06-293-0/+236
| | | | | | 4x-6x faster on sandybridge Signed-off-by: Luca Barbato <[email protected]>
* avutil: Add av_cold attributes to init functions missing themDiego Biurrun2013-05-041-1/+2
|
* x86: float dsp: butterflies_float SSEChristophe Gisquet2013-05-032-0/+29
| | | | | 97c -> 49c Some codecs could benefit from more unrolling, but AAC doesn't.
* dsputil: Make dsputil selectableRonald S. Bultje2013-04-101-0/+2
| | | | Signed-off-by: Martin Storsjö <[email protected]>
* x86inc: Fix number of operands for cmp* instructionsChristophe Gisquet2013-04-091-4/+4
| | | | | | cmp{p,s}{s,d} instructions do take an imm8 operand. Signed-off-by: Diego Biurrun <[email protected]>
generated by cgit v1.2.3 (git 2.25.1) at 2025-09-16 19:27:38 +0000