Re: [PING **3]: i386 backend patches waiting for review...

[Ian Lance Taylor <ian at airs dot com>]

Uros Bizjak <ubizjak@gmail.com> writes:

I'm sorry to take so long to look at these.

> - Peepholes to optimize fix_trunc?f?i_sse code for -mfpmath=387
> (Partially fixes the problems in PR 19398):
> http://gcc.gnu.org/ml/gcc-patches/2005-11/msg00519.html

This is OK when we go back to stage 1.

> - Optimize SSE float <-> double conversions:
> http://gcc.gnu.org/ml/gcc-patches/2005-11/msg00243.html

I see that this patch was updated here:
    http://gcc.gnu.org/ml/gcc-patches/2006-04/msg00137.html
That patch is OK when we go back to stage 1.

> - [RFC] patch for PR target/17390 to CSE some i387 specific compare insns:
> http://gcc.gnu.org/ml/gcc-patches/2005-10/msg01298.html

This doesn't seem like the right approach to me.  For one thing, it
should ideally be rewritten to use the new df framework.  Also, I
think we need to more seriously consider Jan's comment #2 in PR 17390,
and split the insn earlier and see if CSE can eliminate the duplicate
comparison.

Sorry again that this has been sitting for so long.

Ian