This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug target/62025] [4.9/4.10 Regression] Miscompilation of openssl sha512.c
- From: "jakub at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Mon, 11 Aug 2014 16:47:06 +0000
- Subject: [Bug target/62025] [4.9/4.10 Regression] Miscompilation of openssl sha512.c
- Auto-submitted: auto-generated
- References: <bug-62025-4 at http dot gcc dot gnu dot org/bugzilla/>
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=62025
--- Comment #4 from Jakub Jelinek <jakub at gcc dot gnu.org> ---
I've applied on top of r207604 just the
https://gcc.gnu.org/viewcvs/gcc/trunk/gcc/config/s390/s390.h?r1=207605&r2=207604&pathrev=207605
change and the testcase was miscompiled too, so the bug is just triggered by
different RA decisions from having reg 14 live on the epilogue.
The miscompilation goes away with -fno-schedule-insns2, but also with
-fno-schedule-insns -fschedule-insns2, but as I said earlier, appart from
different live in/out lr in/out (including r14 there) *.sched1 dump doesn't
contain any real insn changes.
Note only -mtune=z10 is miscompiled, other tuning options are ok.