[Bug c/49091] New: ice in evaluate_conditions_for_known_args

[Bug c/49091] New: ice in evaluate_conditions_for_known_args

[dcb314 at hotmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49091

           Summary: ice in evaluate_conditions_for_known_args
           Product: gcc
           Version: 4.7.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c
        AssignedTo: unassigned@gcc.gnu.org
        ReportedBy: dcb314@hotmail.com


Created attachment 24309
  --> http://gcc.gnu.org/bugzilla/attachment.cgi?id=24309
C source code

I just tried to compile the package cernlib-2006 with the latest 4.7
snapshot 20110514 on a Fedora Linux x86_64 box.

The compiler said

/home/dcb/rpmbuild/BUILD/cernlib-2006/2006/src/packlib/cspack/tcpaw/tcpaw.c:3568:1:
internal compiler error: vector VEC(tree,base) index domain error, in
evaluate_conditions_for_known_args at ipa-inline-analysis.c:558
Please submit a full bug report,
with preprocessed source if appropriate.
See <http://gcc.gnu.org/bugs.html> for instructions.

Preprocessed source code attached. Flag -O2 required.

[Bug c/49091] ice in evaluate_conditions_for_known_args

[hjl.tools at gmail dot com]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49091

H.J. Lu <hjl.tools at gmail dot com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|UNCONFIRMED                 |NEW
   Last reconfirmed|                            |2011.05.27 04:56:25
                 CC|                            |hubicka at gcc dot gnu.org
   Target Milestone|---                         |4.7.0
     Ever Confirmed|0                           |1

--- Comment #1 from H.J. Lu <hjl.tools at gmail dot com> 2011-05-27 04:56:25 UTC ---
It is caused by revision 172873:

http://gcc.gnu.org/ml/gcc-cvs/2011-04/msg01069.html

[Bug c/49091] ice in evaluate_conditions_for_known_args

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49091

--- Comment #2 from Jan Hubicka <hubicka at gcc dot gnu.org> 2011-06-04 13:56:00 UTC ---
OK, the problem is K&R style definition:
int ssetup(in, out, port)
int *in, *out;
int *port;
{
        struct hostent *hp;
        struct servent *sp;

        struct sockaddr_in myaddr_in;
        struct sockaddr_in peeraddr_in;


later used as:
int server_sock_setup(in, out)
int *in, *out;
{
   return(ssetup(in, out));
}

that is bug, too.

I assumed that here we won't inline because of our type check mismatch.
edge->call_stmt_cannot_inline_p is 0. This is because gimple_check_call_args
actually allows decl to have more arguments than caller.
Well, will fortify this at ipa-analysis side.

Honza

[Bug c/49091] ice in evaluate_conditions_for_known_args

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49091

--- Comment #3 from Jan Hubicka <hubicka at gcc dot gnu.org> 2011-06-04 18:00:51 UTC ---
Author: hubicka
Date: Sat Jun  4 18:00:47 2011
New Revision: 174648

URL: http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=174648
Log:
    PR tree-optimization/48893
    PR tree-optimization/49091
    PR tree-optimization/49179
    * ipa-inline-analysis.c (evaluate_conditions_for_known_args):
    Bounds check.

    * gfortran.dg/pr49179.f90: New testcase

Added:
    trunk/gcc/testsuite/gfortran.dg/pr49179.f90
Modified:
    trunk/gcc/ChangeLog
    trunk/gcc/ipa-inline-analysis.c
    trunk/gcc/testsuite/ChangeLog

[Bug c/49091] ice in evaluate_conditions_for_known_args

[hubicka at gcc dot gnu.org]

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=49091

Jan Hubicka <hubicka at gcc dot gnu.org> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
             Status|NEW                         |RESOLVED
         Resolution|                            |FIXED

--- Comment #4 from Jan Hubicka <hubicka at gcc dot gnu.org> 2011-06-04 18:06:40 UTC ---
Fixed.