This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: PATCH: Check all insns in fallthru to see if label is mentioned
- From: "John David Anglin" <dave at hiauly1 dot hia dot nrc dot ca>
- To: dave at hiauly1 dot hia dot nrc dot ca (John David Anglin)
- Cc: rth at redhat dot com, gcc-patches at gcc dot gnu dot org, hiller at redhat dot com
- Date: Sun, 9 Dec 2001 17:53:02 -0500 (EST)
- Subject: Re: PATCH: Check all insns in fallthru to see if label is mentioned
> There might be two regressions:
>
> FAIL: gcc.c-torture/execute/strct-pack-1.c execution, -O2
> FAIL: gcc.c-torture/execute/strct-pack-1.c execution, -Os
>
> These started failing in execution at some point after Dec. 1. It will
> be a few days before I can examine these in more detail.
It is my guess that these failures are due to the "enhanced structure
offset tracking" patch.
In the -O2 assembler output, we have
ldi 1,%r20
.stabn 68,0,20,L$M7-main
L$M7:
ldw -120(%r30),%r21
<manipulation of r21 deleted>
sth %r20,-120(%r30)
.stabn 68,0,20,L$M11-main
L$M11:
stw %r22,-112(%r30)
stw %r21,-120(%r30)
The halfword store of r20 gets moved downward, after the load of r21.
This occurs in the ce2 pass.
Dave
--
J. David Anglin dave.anglin@nrc.ca
National Research Council of Canada (613) 990-0752 (FAX: 952-6605)