Re: SSE2 disassembler

[Jan Hubicka <jh at suse dot cz>]

> On Tue, 9 Jan 2001, Jan Hubicka wrote:
> 
> > @@ -1907,14 +1922,12 @@ static const struct dis386 dis386_twobyt
> >    { "femms" , XX, XX, XX},
> >    { "", MX, EM, OPSUF }, /* See OP_3DNowSuffix */
> >    /* 10 */
> > -  { PREGRP8 },
> > -  { PREGRP9 },
> > -  { "movlps", XM, EX, SIMD_Fixup, 'h' },  /* really only 2 operands */
> > -  { "movlps", EX, XM, SIMD_Fixup, 'h' },
> > -  { "unpcklps", XM, EX, XX },
> > -  { "unpckhps", XM, EX, XX },
> > -  { "movhps", XM, EX, SIMD_Fixup, 'l' },
> > -  { "movhps", EX, XM, SIMD_Fixup, 'l' },
> > +  { "movlpX", XM, EX, SIMD_Fixup, 'h' },  /* really only 2 operands */
> > +  { "movlpX", EX, XM, SIMD_Fixup, 'h' },
> > +  { "unpcklpX", XM, EX, XX },
> > +  { "unpckhpX", XM, EX, XX },
> > +  { "movhpX", XM, EX, SIMD_Fixup, 'l' },
> > +  { "movhpX", EX, XM, SIMD_Fixup, 'l' },
> >    /* 18 */
> >    { GRP14 },
> >    { "(bad)", XX, XX, XX },
> 
> Replacing 6 instructions with eight???  Please fix this before committing.
> The rest of the patch looks OK, but please test it a little.
I've run the testsuite - interestingly enought it didn't catch...  I will
do some extra testing...

Honza
> 
> Alan Modra
> -- 
> Linuxcare.  Support for the Revolution.