--- Comment #5 from jakub at gcc dot gnu dot org 2007-11-09 23:50 ---
Ok, reading the thread in there it is clear the stack size is the problem.
As it needs 16MB of memory on the stack, it should be run with
ulimit -s 16384
(note that ulimit -s unlimited doesn't imply unlimited thread st
--- Comment #4 from jakub at gcc dot gnu dot org 2007-11-09 21:30 ---
Given how big the array is it is very likely they just overflow the thread
stack.
What ulimit -s they are using?
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=34044
--- Comment #3 from burnus at gcc dot gnu dot org 2007-11-09 21:21 ---
> Yeah, the crash was clearly because j was not initialized.
My inability to generate test cases does not mean that there is not necessarily
no bug in gfortran. See http://gcc.gnu.org/ml/fortran/2007-11/msg00052.html
--- Comment #2 from jakub at gcc dot gnu dot org 2007-11-09 19:43 ---
Yeah, the crash was clearly because j was not initialized.
--
jakub at gcc dot gnu dot org changed:
What|Removed |Added
--
--- Comment #1 from burnus at gcc dot gnu dot org 2007-11-09 19:18 ---
(In reply to comment #0)
>y(j) = 0.0_dp
This part is nonsense and should be "y = 0.0_dp".
The original dump looks ok; I don't know whether something else goes wrong. -
It does not crash here anymore (x86-64 -m32/
