On Fri, May 11, 2012 at 02:42:11PM -0400, Justin Ruggles wrote:
> On 05/11/2012 10:58 AM, Diego Biurrun wrote:
> >
> > --- a/libavcodec/x86/ac3dsp.asm
> > +++ b/libavcodec/x86/ac3dsp.asm
> > @@ -137,10 +137,8 @@ cglobal ac3_max_msb_abs_int16_%1, 2,2,5, src, len
> >
> > INIT_MMX
> > -%define ABS2 ABS2_MMX
> > -%define PSHUFLW pshufw
> > -AC3_MAX_MSB_ABS_INT16 mmx, or_abs
> > %define ABS2 ABS2_MMX2
> > +%define PSHUFLW pshufw
> > AC3_MAX_MSB_ABS_INT16 mmxext, min_max
> > INIT_XMM
> > %define PSHUFLW pshuflw
> > --- a/libavcodec/x86/ac3dsp_mmx.c
> > +++ b/libavcodec/x86/ac3dsp_mmx.c
> > @@ -27,7 +27,6 @@ extern void ff_ac3_exponent_min_mmx (uint8_t *exp, int
> > num_reuse_blocks, int n
> >
> > -extern int ff_ac3_max_msb_abs_int16_mmx (const int16_t *src, int len);
> > extern int ff_ac3_max_msb_abs_int16_mmxext(const int16_t *src, int len);
> > extern int ff_ac3_max_msb_abs_int16_sse2 (const int16_t *src, int len);
> > @@ -55,7 +54,6 @@ av_cold void ff_ac3dsp_init_x86(AC3DSPContext *c, int
> > bit_exact)
> >
> > if (mm_flags & AV_CPU_FLAG_MMX) {
> > c->ac3_exponent_min = ff_ac3_exponent_min_mmx;
> > - c->ac3_max_msb_abs_int16 = ff_ac3_max_msb_abs_int16_mmx;
> > c->ac3_lshift_int16 = ff_ac3_lshift_int16_mmx;
> > c->ac3_rshift_int32 = ff_ac3_rshift_int32_mmx;
>
> You can drop the function for now or wait until Monday. I actually have
> a better fix for this, but I'll be gone soon for the weekend. I'll send
> the fix when I'm back on Monday (I have to split it into separate patches).
No hurry, this has been broken for a loooong time.
Diego
_______________________________________________
libav-devel mailing list
[email protected]
https://lists.libav.org/mailman/listinfo/libav-devel
