[Patch, Fortran] PR fortran/40850 derived type allocatables freed before their components

[Mikael Morin]

On Monday 21 February 2011 14:31:16 Dominique Dhumieres wrote:
> Mikael,
> 
> > I forgot to add that I would appreciate if the testcase was confirmed to
> > fail without the patch on various platforms (I have checked on FreeBSD
> > 8.2),
> 
> It does not on ppc/intel-darwin.
> 
> > and if it doesn't how big the allocatable has to be to produce failure.
> 
> I don't understand what you want;-(
There is an allocate statement in the test with a comment : ! Big enough to 
make it fail
The question is how big is enough (1024 * 1024 in the test), i.e. which amount 
has to be allocated so that the (wrong) code generated produces a segfault, 
and is detected by the testsuite. 

Mikael