This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Where to test g++.old-deja/g++.pt/static11.C with -da?


g++.old-deja/g++.pt/static11.C compiled with -da (actually,
either -dj or -df) gets a SEGV when compiling for
mmix-knuth-mmixware, though I guess it's actually not
target-dependant:

X/g++.old-deja/g++.pt/static11.C: In constructor `C<T>::C() [with T = int]':
X/g++.old-deja/g++.pt/static11.C:18:   instantiated from `C<T>::C() [with T = int]'
X/g++.old-deja/g++.pt/static11.C:27:   instantiated from here
X/g++.old-deja/g++.pt/static11.C:18: internal error: Segmentation fault

What's the best way to deal with that in the test-suite?  I'm
tempted to just copy it to g++.dg/other, with this patch
applied:

*** ../../g++.old-deja/g++.pt/static11.C	Fri Sep 22 00:05:48 2000
--- ./static11.C	Fri Mar 29 22:22:44 2002
***************
*** 1,5 ****
! // Bug: g++ was failing to destroy C<int>::a because it was using two
! // different sentry variables for construction and destruction.

  extern "C" void _exit (int);

--- 1,8 ----
! // This is a copy of g++.old-deja/g++.pt/static11.C which at one
! // time got a SEGV for mmix-knuth-mmixware when compiled with
! // -da (or either -dj or -df).
! // { dg-do run }
! // { dg-options "-da" }

  extern "C" void _exit (int);

Maybe there are better ideas?  Perhaps I should also delete
g++.old-deja/g++.pt/static11.C and keep the old comment too in
g++.dg/other/static11.C?

brgds, H-P


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]