This is the mail archive of the
mailing list for the GCC project.
Re: DSE calls to builtins (memset, etc)
- From: Marc Glisse <marc dot glisse at inria dot fr>
- To: Richard Biener <richard dot guenther at gmail dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Wed, 20 Aug 2014 16:35:29 +0200 (CEST)
- Subject: Re: DSE calls to builtins (memset, etc)
- Authentication-results: sourceware.org; auth=none
- References: <alpine dot DEB dot 2 dot 02 dot 1408162253190 dot 20292 at stedding dot saclay dot inria dot fr> <CAFiYyc3aqS=5y9Rzk-senPMPfCgPh0a=O_RLb-JjWnc+NwdAig at mail dot gmail dot com> <alpine dot DEB dot 2 dot 11 dot 1408191540250 dot 2316 at stedding dot saclay dot inria dot fr> <alpine dot DEB dot 2 dot 02 dot 1408200901460 dot 5162 at stedding dot saclay dot inria dot fr> <CAFiYyc3sHS-RiP3mmEnMu92rMUUKEdUOktW7=xD+0t8sj0-xfg at mail dot gmail dot com> <alpine dot DEB dot 2 dot 11 dot 1408201619160 dot 1778 at laptop-mg dot saclay dot inria dot fr>
On Wed, 20 Aug 2014, Marc Glisse wrote:
On Wed, 20 Aug 2014, Richard Biener wrote:
On Wed, Aug 20, 2014 at 9:14 AM, Marc Glisse <email@example.com> wrote:
here is a new version of the patch which passed bootstrap+testsuite. I am
still removing the ref_maybe_used_by_stmt_p test, see the previous message
for a discussion.
Looks good to me. I'm still nervous about removing that check (though
the description looks odd - if the stmt kills the lhs and it also uses
it then if this is not an exact overlap it's undefined
I hadn't realized that. Note that this looks true if the reference is the lhs
of an assignment, but it can be wrong if the reference was computed from a
Obviously I swapped p and p+2...
memmove would kill memset, except that...
and if it is then the stmt is still dead...).
use_stmt would be dead in that case, not stmt.
stmt: a = 42;
use_stmt: a = a;
So - ok if it bootstraps/tests ok.