Hi, On Wed, Sep 30, 2015 at 5:27 PM, Henrik Gramner <hen...@gramner.com> wrote:
> Makes it possible to use them in arithmetic expressions. > --- > libavutil/x86/x86inc.asm | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/libavutil/x86/x86inc.asm b/libavutil/x86/x86inc.asm > index 6ad9785..929d0d2 100644 > --- a/libavutil/x86/x86inc.asm > +++ b/libavutil/x86/x86inc.asm > @@ -773,8 +773,8 @@ BRANCH_INSTR jz, je, jnz, jne, jl, jle, jnl, jnle, jg, > jge, jng, jnge, ja, jae, > %assign cpuflags_bmi1 (1<<22)|cpuflags_lzcnt > %assign cpuflags_bmi2 (1<<23)|cpuflags_bmi1 > > -%define cpuflag(x) ((cpuflags & (cpuflags_ %+ x)) == (cpuflags_ %+ x)) > -%define notcpuflag(x) ((cpuflags & (cpuflags_ %+ x)) != (cpuflags_ %+ x)) > +%define cpuflag(x) (((((cpuflags & (cpuflags_ %+ x)) ^ (cpuflags_ %+ > x)) - 1) >> 31) & 1) > +%define notcpuflag(x) (cpuflag(x) ^ 1) > > ; Takes an arbitrary number of cpuflags from the above list. > ; All subsequent functions (up to the next INIT_CPUFLAGS) is built for > the specified cpu. > -- > 1.9.1 Wow, many operands. Looks OK, although I have to admit this isn't exactly world-readable anymore :) I guess for this type of code that's OK... Ronald _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
