[PATCH] Fix gfortran.dg/forall_1.f90

[Tobias Schlüter]

Jakub Jelinek wrote:
> Of course we only want to increment it in the innermost loop which
> contains the load of the mask.  Here is a fix, tested with no check-f95
> failures on x86_64-linux.  Ok for HEAD and 4.0.2?
> 
> Alternatively, the increment could be moved to gfc_trans_nested_forall_loop
> and body passed as stmtblock_t instead of tree.  This change is shorter
> though...

The patch is ok for both provided you add more tests, the testcase I added is
not really sufficient, because it doesn't have multi-dimensional masks or
non-trivial dependencies between the mask and the assignments.

I don't think the alternative you proposed makes more sense:
gfc_trans_forall_loop is where the loops are built, let's keep it all in one
place.

- Tobi