[FFmpeg-devel] [PATCH] hpeldsp: fix build without inline asm

Christophe Gisquet christophe.gisquet at gmail.com
Wed May 21 11:45:13 CEST 2014


Hi,

2014-05-21 11:19 GMT+02:00 Hendrik Leppkes <h.leppkes at gmail.com>:
> Also, it would be super great if people messing with inline-asm could run a
> --disable-inline-asm build once to make sure it doesn't break.

I tested with --disable-yasm. Does this mean 4 configs
(HAVE_MMX_EXTERNAL*HAVE_MMX_INLINE) should be tested then?

> +#if HAVE_MMX_INLINE
>  #define SET_HPEL_FUNCS(PFX, IDX, SIZE, CPU)                                     \
>      do {                                                                        \
> -        if (HAVE_MMX_EXTERNAL)                                                  \
> -        c->PFX ## _pixels_tab IDX [0] = PFX ## _pixels ## SIZE ## _     ## CPU; \
> +        SET_HPEL_FUNCS_EXT(PFX, IDX, SIZE, CPU)                                 \
>          if (HAVE_MMX_INLINE) {                                                  \
>          c->PFX ## _pixels_tab IDX [1] = PFX ## _pixels ## SIZE ## _x2_  ## CPU; \

I don't get what is happening. Is it a problem of the MMX inline
functions not being defined (seems so, the 16w-ide variants are
defined only through CALL_2X_PIXELS)? In which case, assuming dead
code elimination, this would be the actual problem.

Also, cavsdsp, diracdsp and vc1dsp probably have equivalent
requirements. Was the test that discovered the issue disabling them?

-- 
Christophe


More information about the ffmpeg-devel mailing list