Re: patch: x86 generic simd fixlet

[Andrew Pinski <pinskia at physics dot uc dot edu>]

On Tuesday, Jul 8, 2003, at 15:21 US/Eastern, Aldy Hernandez wrote:

> Janis' testcases expose a myraid of different ICE's on x86.
>
> This one in particular:
>
> 	typedef int __attribute__((mode(V4SI))) v4si;
> 	void
> 	pass_v4si (v4si v) {}
>
> ...aborts on ix86_function_arg_boundary.  The problem is that
> contains_128bit_aligned_vector_p() is only checking SSE_REG_MODE_P.
> It should also check TARGET_SSE because V4SI can be used with generic
> simd support.
>
> I also fixed a line wrap problem on the function comment for
> contains_128bit_aligned_vector_p.
>
> OK?
>
> Aldy
>
> 2003-07-08  Aldy Hernandez  <aldyh@redhat.com>
>
> 	* config/i386/i386.c (contains_128bit_aligned_vector_p): Check
> 	that SSE modes return true only if SSE is enabled.

This fixes PR11144, right?  If it does can you add PR markers to the 
changelog?

Thanks,
Andrew Pinski