On Mon, 24 Oct 2016 21:53:22 +0200 Henrik Gramner <hen...@gramner.com> wrote:
> On Mon, Oct 24, 2016 at 9:34 PM, wm4 <nfx...@googlemail.com> wrote: > > a ASM function must, according to the calling convention, reset the > > MMX state when returning. > > > > What FFmpeg does here was misdesigned from the very start. > > The decision to issue emms manually instead of after every MMX > function was a deliberate decision. I'd hardly call it "misdesigned" > to make SIMD code twice as fast at the cost of technically abusing the > ABI, which has worked flawlessly for years until very recently. It's still clearly not a good idea because it violates the ABI directly. OK let's get back to the "practical" argument that we have not much of a choice but to violate the ABI. Then at least the "weaker" assumption, that emms should be called before calling (or returning to) functions that don't explicitly support MMX, should be followed. _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
