From: Fiona Glaser <fiona@x264.com>
Date: Sun, 6 Jul 2008 17:15:19 +0000 (-0600)
Subject: Resolve floating point exception with frame_init_lowres mmx
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=653249521805b21564c00148f7db1e4b28e6e15c;p=libx264

Resolve floating point exception with frame_init_lowres mmx
In some cases, the mmx version of frame_init_lowres could leave the FPU uninitialized for use in ratecontrol, resulting in floating point exceptions.
Since frame_init_lowres is such a time-consuming function, an emms was just put at the end, since it costs almost nothing compared to the total time of frame_init_lowres.
---

diff --git a/common/x86/mc-a2.asm b/common/x86/mc-a2.asm
index 47da6a29..50aa63e3 100644
--- a/common/x86/mc-a2.asm
+++ b/common/x86/mc-a2.asm
@@ -643,6 +643,7 @@ cglobal x264_frame_init_lowres_core_%1, 6,7
     dec    dword r8m
     jg .vloop
     ADD      rsp, 2*push_size
+    emms
     RET
 %endmacro ; FRAME_INIT_LOWRES