https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122902
Paul Thomas <pault at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Assignee|unassigned at gcc dot gnu.org |pault at gcc dot gnu.org
--- Comment #4 from Paul Thomas <pault at gcc dot gnu.org> ---
Created attachment 63845
--> https://gcc.gnu.org/bugzilla/attachment.cgi?id=63845&action=edit
Fix for this PR
This fixes the PR and regtests OK.
It also shifts the fiats problem to a totally different place:
trainable_network_test_m.F90:232
do concurrent(batch=1:mini_batch_size, iter=1:num_iterations)
training_inputs(batch, iter) = tensor_t(merge(true, false,
harvest(:,batch,iter) < 0.5E0)) !!!!! This line fails in runtime
training_outputs(batch, iter) = or(training_inputs(batch, iter))
end do
@Damian - I seem to remember seeing a workaround for an expression using merge
somewhere else. I'll take a look later on but, right now, I have to sleep off
some jet-lag.
