[Bug tree-optimization/81848] New: Segmentation fault at gimple-ssa-strength-reduction.c:2356

jamborm at gcc dot gnu.org gcc-bugzilla@gcc.gnu.org
Mon Aug 14 12:35:00 GMT 2017


https://gcc.gnu.org/bugzilla/show_bug.cgi?id=81848

            Bug ID: 81848
           Summary: Segmentation fault at
                    gimple-ssa-strength-reduction.c:2356
           Product: gcc
           Version: 8.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: tree-optimization
          Assignee: unassigned at gcc dot gnu.org
          Reporter: jamborm at gcc dot gnu.org
  Target Milestone: ---

Created attachment 41988
  --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=41988&action=edit
Testcase

Starting with r250416, spec 2017 benchmark 521.wrf_r started to ICE at
-Ofast optimization level.  I have extracted the attached (slightly
minimized) function from it which reproduces the ICE:

$ /home/mjambor/gcc/bisect/inst/bin/gfortran -S -Ofast bug.f90
during GIMPLE pass: slsr
bug.f90:4:0:

   integer function ice_in_me( pcols , pver   , ncol       , fixed_ubc  , mw   
 , ubc_mmr , &

internal compiler error: Segmentation fault
0xf4c4fb crash_signal
        /home/mjambor/gcc/bisect/src/gcc/toplev.c:338
0x190bdf4 create_phi_basis
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:2356
0x190bfa0 create_phi_basis
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:2386
0x190bfa0 create_phi_basis
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:2386
0x190bfa0 create_phi_basis
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:2386
0x190f7a5 replace_profitable_candidates
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:3670
0x190fa9e analyze_candidates_and_replace
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:3769
0x190fc82 execute
        /home/mjambor/gcc/bisect/src/gcc/gimple-ssa-strength-reduction.c:3843
Please submit a full bug report,


More information about the Gcc-bugs mailing list