https://gcc.gnu.org/bugzilla/show_bug.cgi?id=93365
--- Comment #8 from markeggleston at gcc dot gnu.org --- (In reply to markeggleston from comment #7) > (In reply to Steve Kargl from comment #6) > > On Fri, Jan 24, 2020 at 12:36:45PM +0000, markeggleston at gcc dot gnu.org > > wrote: > > > gfc_simplify_expr calls simplify_parameter_variable which for zero size > > > arrays it creates a new gfc_expr structure with the expr_type set to > > > EXPR_ARRAY > > > and replaces primary. When primary is replaced its chain of ref > > > structures is > > > deleted so tmp->u.i is invalid causing the ICE the following lines are > > > never > > > reached: > > > > The lines are reached in my debugging sessions. :-) > > I bootstrap with --enable-checking, which might > > prevent the ICE issue. > > Tried that to no effect, however, I found that the version of the compiler > used to compile gfortran does affect the outcome: > > bootstrap no ICE > gcc-8.2.0 no ICE checking the environment again: gcc 8.2.1, g++ 4.8.5 > gcc-6.3.0 ICE > > I haven't tried an other compilers. > > > > > > > then enters the switch statement and get to lines 2384-2385 > > > > > > > > primary->ts.type = BT_INTEGER; > > > > primary->ts.kind = gfc_default_integer_kind; > > > > > > > > This just resets a part of primary, but it does not fix up > > > > it up correctly. Likely, need to check for zero size > > > > arrays do extra work. > > > > > > I have modified gfc_simplify_expr: > > > > I'm beginning to think that calling gfc_simplify_expr may > > not be correct, here. But, don't have time to investigate > > more.
