[Bug rtl-optimization/46522] New: ICE: in reset_sched_cycles_in_current_ebb, at sel-sched.c:7092 with -O3 -fsel-sched-pipelining -fselective-scheduling2 -funroll-loops

zsojka at seznam dot cz gcc-bugzilla@gcc.gnu.org
Wed Nov 17 16:14:00 GMT 2010


http://gcc.gnu.org/bugzilla/show_bug.cgi?id=46522

           Summary: ICE: in reset_sched_cycles_in_current_ebb, at
                    sel-sched.c:7092 with -O3 -fsel-sched-pipelining
                    -fselective-scheduling2 -funroll-loops
           Product: gcc
           Version: 4.6.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: rtl-optimization
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: zsojka@seznam.cz
              Host: x86_64-pc-linux-gnu
            Target: x86_64-pc-linux-gnu


Created attachment 22432
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=22432
reduced testcase (originally PR45352 comment #19)

I am sorry for the amount of compiler flags, removing any of them causes the
ICE to disappear. Even manual partial inlining, unrolling, loop unswitching,
... either removes the ICE or doesn't allow me to remove further flags.

Flags are either:
-O3 -fkeep-inline-functions -fsel-sched-pipelining -fselective-scheduling2
-funroll-loops
or
-O2 -fkeep-inline-functions -fsel-sched-pipelining -fselective-scheduling2
-funroll-loops -funswitch-loops
or
-O1 -fcse-follow-jumps -fkeep-inline-functions -fpartial-inlining
-freorder-blocks -frerun-cse-after-loop -fschedule-insns2
-fsel-sched-pipelining -fselective-scheduling2 -funroll-loops -funswitch-loops

Compiler output:
$ gcc -O3 -fkeep-inline-functions -fsel-sched-pipelining
-fselective-scheduling2 -funroll-loops pr46522.c
pr46522.c: In function 'foo2':
pr46522.c:30:1: internal compiler error: in reset_sched_cycles_in_current_ebb,
at sel-sched.c:7092
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

Tested revisions (for the -O1 case):
r166795 - crash
r165699 - OK
4.5 - doesn't know -fpartial-inlining



More information about the Gcc-bugs mailing list