This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] Fix REE (PR rtl-optimization/53160)
- From: Richard Henderson <rth at redhat dot com>
- To: Jakub Jelinek <jakub at redhat dot com>
- Cc: gcc-patches at gcc dot gnu dot org
- Date: Mon, 30 Apr 2012 11:58:48 -0700
- Subject: Re: [PATCH] Fix REE (PR rtl-optimization/53160)
- References: <20120430134308.GM16117@tyan-ft48-01.lab.bos.redhat.com>
On 04/30/2012 06:43 AM, Jakub Jelinek wrote:
As shown on the testcase below, if REE modifies some sign/zero extension
insn, which is on the candidate vector, as a def_insn of some other
extension, before combine_reaching_defs is called on that insn, we might
ignore the changes done to that insn and just assume it is only an
extension to the recorded mode. It might be an extension to a wider mode
Fixed thusly, bootstrapped/regtested on x86_64-linux and i686-linux,
ok for trunk/4.7?
2012-04-30 Jakub Jelinek<email@example.com>
* ree.c (combine_reaching_defs): Handle the case where cand->insn
has been modified by ree pass already.
* gcc.c-torture/execute/pr53160.c: New test.