har0ke/FFmpeg
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

x86inc: warn if XOP integer FMA instruction emulation is impossible

Browse Source 
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 <henrik@gramner.com>
Signed-off-by: Anton Khirnov <anton@khirnov.net>
This commit is contained in:
Anton Mitrofanov 2015-08-01 17:27:31 +02:00 committed by Anton Khirnov
parent 8f4a06faf4
commit 8c75ba55a4
1 changed files with 5 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
libavutil/x86/x86inc.asm
View File

@ -1410,15 +1410,18 @@ AVX_INSTR pfmul, 1, 0, 1
%macro %1 4-7 %1, %2, %3 %macro %1 4-7 %1, %2, %3
%if cpuflag(xop) %if cpuflag(xop)
v%5 %1, %2, %3, %4 v%5 %1, %2, %3, %4
%else %elifnidn %1, %4
%6 %1, %2, %3 %6 %1, %2, %3
%7 %1, %4 %7 %1, %4
%else
%error non-xop emulation of ``%5 %1, %2, %3, %4'' is not supported
%endif %endif
%endmacro %endmacro
%endmacro %endmacro
FMA_INSTR pmacsdd, pmulld, paddd
FMA_INSTR pmacsww, pmullw, paddw FMA_INSTR pmacsww, pmullw, paddw
FMA_INSTR pmacsdd, pmulld, paddd ; sse4 emulation
FMA_INSTR pmacsdql, pmuldq, paddq ; sse4 emulation
FMA_INSTR pmadcswd, pmaddwd, paddd FMA_INSTR pmadcswd, pmaddwd, paddd
; tzcnt is equivalent to "rep bsf" and is backwards-compatible with bsf. ; tzcnt is equivalent to "rep bsf" and is backwards-compatible with bsf.