Hi Tobias,
I have not yet looked at the patch, but I wonder whether that causes
invalid code for
if (any([a,b,c] < f()))
by evaluating f() multiple times.
This is avoided by this part of the patch:
+ if (op2->expr_type == EXPR_CONSTANT)
+ scalar = gfc_copy_expr (op2);
+ else
+ scalar = create_var (gfc_copy_expr (op2));
which results in code like this:
__var_1 = f ();
if ((a > __var_1 || b > __var_1) || c > __var_1)
{
{
Another question is: Why do we generate a temporary array for ANY/ALL?
The straightforward way of evaluating
if (any([a,b,c] > 0.)) then
is to
- create a temporary array [a,b,c]
- to create a second logical array and fill it with the
logical expressions,
- call a library function for ANY on it.
This was the original code and was not really efficient ;-)
Since Mikael's patch some time ago, we
- create a temporary array [a,b,c]
- loop over it, exiting early if a true value was found
With this patch (and my previous patch, which is already in),
with front-end optimization enabled, we now
- convert the expression into if (any[a>0, b>0, c>0])
- convert this into a>0 || b>0 || c>0 (effect of my previous patch)
A
scalar should be sufficient, i.e.
res = .false.
do i = 1, n
res = res .or. (array(i) < eps)
end do
if (res)
We have been doing that since Mikael's patch some years ago.
Regards
Thomas
