Browse Source

avcodec/x86/fmtconvert: Add emms to int32_to_float_fmul_array8_sse()

this should fix checkasm on x86_64-archlinux-gcc-valgrind

Signed-off-by: Michael Niedermayer <michael@niedermayer.cc>
tags/n3.0
Michael Niedermayer 10 years ago
parent
commit
da6f34516b
1 changed files with 8 additions and 1 deletions
  1. +8
    -1
      libavcodec/x86/fmtconvert.asm

+ 8
- 1
libavcodec/x86/fmtconvert.asm View File

@@ -107,7 +107,14 @@ cglobal int32_to_float_fmul_array8, 5, 5, 5, c, dst, src, mul, len
add mulq, 4
add lenq, 32
jl .loop
REP_RET
%if notcpuflag(sse2)
;; cvtpi2ps switches to MMX even if the source is a memory location
;; possible an error in documentation since every tested CPU disagrees with
;; that. Use emms anyway since the vast majority of machines will use the
;; SSE2 variant
emms
%endif
RET
%endmacro

INIT_XMM sse


Loading…
Cancel
Save