public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
From: "hjl.tools at gmail dot com" <gcc-bugzilla@gcc.gnu.org>
To: gcc-bugs@gcc.gnu.org
Subject: [Bug rtl-optimization/67029] [5/6 regression] gcc-5.2.0 unable to find a register to spill with O3 fsched-pressure fschedule-insns
Date: Wed, 05 Aug 2015 19:14:00 -0000	[thread overview]
Message-ID: <bug-67029-4-DRl2ufCz40@http.gcc.gnu.org/bugzilla/> (raw)
In-Reply-To: <bug-67029-4@http.gcc.gnu.org/bugzilla/>

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67029

--- Comment #10 from H.J. Lu <hjl.tools at gmail dot com> ---
The problem is

/* The value of get_preferred_alternatives for the current instruction,
   supplemental to recog_data.  */
static alternative_mask preferred_alternatives;

ira_implicitly_set_insn_hard_regs may use the staled preferred_alternatives:

#0  ira_implicitly_set_insn_hard_regs (set=0x7fffffffd950)
    at /export/gnu/import/git/sources/gcc/gcc/ira-lives.c:847
#1  0x000000000156d604 in sched_analyze_insn (deps=0x7fffffffdba0, 
    x=0x7ffff18277d0, insn=0x7ffff1823240)
    at /export/gnu/import/git/sources/gcc/gcc/sched-deps.c:2901
#2  0x00000000015703f5 in deps_analyze_insn (deps=0x7fffffffdba0, 
    insn=0x7ffff1823240)
    at /export/gnu/import/git/sources/gcc/gcc/sched-deps.c:3675
#3  0x0000000001570aa6 in sched_analyze (deps=0x7fffffffdba0, 
    head=0x7ffff1823240, tail=0x7ffff1708990)
    at /export/gnu/import/git/sources/gcc/gcc/sched-deps.c:3822
#4  0x0000000000c4b3d0 in compute_block_dependences (bb=0)
    at /export/gnu/import/git/sources/gcc/gcc/sched-rgn.c:2743
#5  0x0000000000c4ccae in sched_rgn_compute_dependencies (rgn=0)
    at /export/gnu/import/git/sources/gcc/gcc/sched-rgn.c:3267
#6  0x0000000000c4c2bf in schedule_region (rgn=0)
    at /export/gnu/import/git/sources/gcc/gcc/sched-rgn.c:3023
#7  0x0000000000c4d3e6 in schedule_insns ()
    at /export/gnu/import/git/sources/gcc/gcc/sched-rgn.c:3426
#8  0x0000000000c4dca4 in rest_of_handle_sched ()


  parent reply	other threads:[~2015-08-05 19:14 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2015-07-27 15:52 [Bug c/67029] New: " ziebell_marco at posteo dot de
2015-07-27 16:06 ` [Bug c/67029] " trippels at gcc dot gnu.org
2015-07-27 16:12 ` ziebell_marco at posteo dot de
2015-07-27 18:58 ` trippels at gcc dot gnu.org
2015-07-27 19:12 ` ziebell_marco at posteo dot de
2015-07-27 19:29 ` trippels at gcc dot gnu.org
2015-07-27 19:29 ` trippels at gcc dot gnu.org
2015-07-27 19:55 ` [Bug rtl-optimization/67029] [5/6 regression] " trippels at gcc dot gnu.org
2015-08-03  7:18 ` ubizjak at gmail dot com
2015-08-03 14:51 ` hjl.tools at gmail dot com
2015-08-05 19:14 ` hjl.tools at gmail dot com [this message]
2015-08-05 20:27 ` hjl.tools at gmail dot com
2015-08-06  2:38 ` hjl.tools at gmail dot com
2015-08-07 16:18 ` hjl at gcc dot gnu.org
2015-08-07 16:50 ` [Bug rtl-optimization/67029] [5 Regression] " trippels at gcc dot gnu.org
2015-09-30 16:40 ` hjl.tools at gmail dot com
2015-09-30 16:40 ` hjl at gcc dot gnu.org

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=bug-67029-4-DRl2ufCz40@http.gcc.gnu.org/bugzilla/ \
    --to=gcc-bugzilla@gcc.gnu.org \
    --cc=gcc-bugs@gcc.gnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).