This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: [PATCH] Use tcc_comparison


Devang Patel <dpatel@apple.com> writes:

> When I added vector_compare_rtx() this one escaped because
> tcc_comparison work was being done in parallel. Only I am running into
> this because I have couple of patches waiting for review/feedback in
> my sources. I have following patchlet in my tree since weeks. If I do
> not hear any objection, I will commit this later today as obvious.
> 
> 2004-11-12  Devang Patel  <dpatel@apple.com>
> 
>       * optabs.c (vector_compare_rtx): Use tcc_comparison.

This is OK.

> Index: gcc/optabs.c
> ===================================================================
> RCS file: /cvs/gcc/gcc/gcc/optabs.c,v
> retrieving revision 1.247
> diff -Idpatel.pbxuser -c -3 -p -r1.247 optabs.c
> *** gcc/optabs.c        9 Nov 2004 17:34:03 -0000       1.247
> --- gcc/optabs.c        12 Nov 2004 20:39:47 -0000
> *************** vector_compare_rtx (tree cond, bool unsi
> *** 5082,5088 ****
>      tree t_op0, t_op1;
>      rtx rtx_op0, rtx_op1;
> 
> !   if (TREE_CODE_CLASS (TREE_CODE (cond)) != '<')
>        {
>          /* This is unlikely. While generating VEC_COND_EXPR,
>           auto vectorizer ensures that condition is a relational
> --- 5082,5088 ----
>      tree t_op0, t_op1;
>      rtx rtx_op0, rtx_op1;
> 
> !   if (TREE_CODE_CLASS (TREE_CODE (cond)) != tcc_comparison)
>        {
>          /* This is unlikely. While generating VEC_COND_EXPR,
>           auto vectorizer ensures that condition is a relational


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]