This is the mail archive of the
mailing list for the GCC project.
Re: PR optimization/9786
- From: kenner at vlsi1 dot ultra dot nyu dot edu (Richard Kenner)
- To: mrs at apple dot com
- Cc: gcc at gcc dot gnu dot org
- Date: Tue, 16 Sep 03 21:12:57 EDT
- Subject: Re: PR optimization/9786
The question is simple, but probably not specific enough. As it
stands, the answer is probably no. But, in reading the referenced
message, it sounds like we know that the no-op doesn't trap, and
therefore, we can change it, optimize it based on the knowledge that it
doesn't trap and therefore, delete it.
I disagree in reason, but not in effect. The cited insn may indeed trap
if the input were a signaling NaN, but if we delete it, then some *other*
insn will trap instead, which is fine.
However, note that this question is somewhat vacuuous because the
point is that the reg-stack code seems to feel the insn actually shouldn't
be emitted. So that if it indeed *shouldn't* be deleted, then what *do*
we do, since the code seems to require it to be deleted?