Bug 30672 - Internal compiler error: segment violation
Summary: Internal compiler error: segment violation
Status: RESOLVED DUPLICATE of bug 30391
Alias: None
Product: gcc
Classification: Unclassified
Component: fortran (show other bugs)
Version: 4.3.0
: P3 normal
Target Milestone: ---
Assignee: Not yet assigned to anyone
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2007-02-01 20:46 UTC by Luis Guillermo Cota
Modified: 2007-02-02 00:28 UTC (History)
11 users (show)

See Also:
Host:
Target:
Build:
Known to work:
Known to fail:
Last reconfirmed:


Attachments

Note You need to log in before you can comment on or make changes to this bug.
Description Luis Guillermo Cota 2007-02-01 20:46:39 UTC
error interno del compilador: Violación de segmento

/home/lgc/irun/bin/../libexec/gcc/i386-pc-linux-gnu/4.3.0/f951 ../dynamic.f -ffixed-form -quiet -dumpbase dynamic.f -mtune=i386 -auxbase dynamic -O2 -version -ffixed-line-length-132 -I.. -fintrinsic-modules-path /home/lgc/irun/bin/../lib/gcc/i386-pc-linux-gnu/4.3.0/finclude -o dynamic.s
GNU F95 version 4.3.0 20070131 (experimental) (i386-pc-linux-gnu)
        compiled by GNU C version 4.3.0 20070131 (experimental).
GGC heurísticas: --param ggc-min-expand=100 --param ggc-min-heapsize=131072

------- code segment below -------

 else
        kvf(1:3,1:3) = kv(1:3,1:3)
      endif

----------------------------------

Please I don't want to annoy you but the instructions to file a bug are really hard to follow and lengthy. I haven't been able to produce the "*.i* file, but just the ".s" one. Please mail me if you want me to send you the whole package or else give me other instructions.

Thank you.
Comment 1 Andrew Pinski 2007-02-01 20:49:18 UTC
(In reply to comment #0)
> Please I don't want to annoy you but the instructions to file a bug are really
> hard to follow and lengthy. I haven't been able to produce the "*.i* file, but
> just the ".s" one. Please mail me if you want me to send you the whole package
> or else give me other instructions.

Those instrunctions are more for C/C++ bug filing.
Does the code use modules?
If not can you just attach dynamic.f if it is open source?
Comment 2 Luis Guillermo Cota 2007-02-01 20:54:20 UTC
I could attach the 'dynamic.f' file if you tell me how (or is it just to cut-and-paste it?). The file is NOT open source but is free for academic use. It DOES use modules.
Comment 3 kargls 2007-02-01 21:44:09 UTC
Luis,

You can email me the file at sgk@troutmask.apl.washington.edu.
I'll extract a testcase if the failure occurs on my systems.

Comment 4 kargls 2007-02-01 23:36:31 UTC
Just a quick note.  dynamic.f will compile without optimization.
I can reproduce the problem.  Now to reduce it to something managable.
Comment 5 kargls 2007-02-02 00:08:28 UTC
Reduced testcase


module current
   integer, save :: ndim
   real, save :: kv(3,3)
end module current

subroutine dynamic

   use current

   implicit none

   real :: kvf(3,3)

   if (ndim .eq. 3) then
      call kvector3Df(kvf)
   else
      kvf(1:3,1:3) = kv(1:3,1:3)
   end if

end subroutine dynamic



troutmask:sgk[372] gfc4x -c -O zxc.f90
zxc.f90: In function 'dynamic':
zxc.f90:17: internal compiler error: in operand_equal_p, at fold-const.c:2557
Please submit a full bug report,
with preprocessed source if appropriate.
See <URL:http://gcc.gnu.org/bugs.html> for instructions.
Comment 6 kargls 2007-02-02 00:12:05 UTC
Further reduction is possible.

subroutine dynamic(ndim, kv)
   implicit none
   integer :: ndim
   real :: kvf(3,3), kv(3,3)
   if (ndim .eq. 3) then
      call kvector3Df(kvf)
   else
      kvf(1:3,1:3) = kv(1:3,1:3)
   end if
end subroutine dynamic
Comment 7 Andrew Pinski 2007-02-02 00:28:34 UTC

*** This bug has been marked as a duplicate of 30391 ***