[Bug middle-end/54099] New: [4.8 Regression] gfortran.dg/pr44691.f fails with ICE in free_regset_pool, at sel-sched-ir.c:994

["burnus at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54099

             Bug #: 54099
           Summary: [4.8 Regression]  gfortran.dg/pr44691.f fails with ICE
                    in free_regset_pool, at sel-sched-ir.c:994
    Classification: Unclassified
           Product: gcc
           Version: 4.8.0
            Status: UNCONFIRMED
          Keywords: ice-on-valid-code
          Severity: normal
          Priority: P3
         Component: middle-end
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: burnus@gcc.gnu.org


I get the following failure on x86-64-gnu-linux:

FAIL: gfortran.dg/pr44691.f  -O  (internal compiler error)
FAIL: gfortran.dg/pr44691.f  -O  (test for excess errors)

And running it manually, I get:

$ gfortran -O2 -fselective-scheduling2 -fsel-sched-pipelining
-funroll-all-loops gfortran.dg/pr44691.f

gfortran.dg/pr44691.f: In function âorienâ:
gfortran.dg/pr44691.f:39:0: internal compiler error: in free_regset_pool, at
sel-sched-ir.c:994


Sometimes, I get the same failure for:
  FAIL: gfortran.dg/pr42294.f  -O  (internal compiler error)
  FAIL: gfortran.dg/pr42294.f  -O  (test for excess errors)

but I cannot reproduce it at the moment. However, both failures can be seen at
http://gcc.gnu.org/ml/gcc-testresults/2012-07/msg02161.html

Looking at the test results, no failure occurs for:

http://gcc.gnu.org/ml/gcc-testresults/2012-07/msg02099.html
4.8.0 20120724 [trunk revision 189815] (GCC) on x86_64-unknown-linux-gnu

but it does starting from

http://gcc.gnu.org/ml/gcc-testresults/2012-07/msg02109.html
4.8.0 20120724 [trunk revision 189820] (GCC) on x86_64-unknown-linux-gnu