https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101632
--- Comment #5 from Steve Kargl <sgk at troutmask dot apl.washington.edu> --- On Tue, Aug 03, 2021 at 02:37:40PM +0000, jb at gcc dot gnu.org wrote: > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=101632 > > Janne Blomqvist <jb at gcc dot gnu.org> changed: > > What |Removed |Added > ---------------------------------------------------------------------------- > CC| |jb at gcc dot gnu.org > Depends on| |91413 > > --- Comment #4 from Janne Blomqvist <jb at gcc dot gnu.org> --- > In PR91413 we discussed the issue of stack overflows if recursive is made the > default. Some decision and solution on that issue is needed. > > Referenced Bugs: > > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91413 > [Bug 91413] [F2018]: Procedures are recursive by default; switching from stack > to static allocation is not safe Hi Janne, It's been awhile! I compiled an older version of the polyhedron testsuite. I did not see any impact while running the tests. Perhaps, Polyhedron does not contain a usage pattern that would cause a stack issue. I did run the test with my normal user privilege % limits | grep stack stacksize 524288 kB Perhaps, I should have limited the stack. Note, the patch does allow gfortran to compile standard conforming Fortran in that it will now allow a NON_RECURSIVE prefix. We could disable the actual setting of the recursive attribute until someone comes up with a deeper analysis of the handling of the stack. I suspect most current code does not use recursion unless RECURSIVE is explicitly used
