Hi Paul,
In
@@ -1335,19 +1340,49 @@ get_class_info_from_ss (stmtblock_t * pre, gfc_ss *ss,
tree *eltype) rhs_function = true;
}
}
+ else if (cntnr != NULL_TREE)
+{
+ tmp = gfc_class_vptr_get (rhs_class_expr);
+ gfc_add_modify (pre, tmp, fold_convert (TREE_TYPE (tmp),
+
Hi Andre,
Thank you for the review.
> ...snip...
>
> I am confused here, because you are assigning to rhs. When that is
> correct, why
> is there no else assigning zero to the rhs->_len when arg1 is not
> unlimited?
'rhs_class_expr' is highly confusing and came from the original use of this
pa
Hi Paul,
yes, please go ahead with the merge. To my astonishment, I had no conflicts
with your patch. Mine is addressing copy-in/(out) aka packing/unpacking of
derived-type to class-type arguments.
Thanks for the patch.
- Andre
On Tue, 2 Jul 2024 09:21:26 +0100
Paul Richard Thomas wrote:
> H
Hi Andre,
I have to hold back on the commit until the business below is sorted out.
Cheers
Paul
-- Forwarded message -
From: Paul Richard Thomas
Date: Tue, 2 Jul 2024 at 08:48
Subject: [Linaro-TCWG-CI] gcc patch #93154: FAIL: 1 regressions on arm
To:
Hi there,
You detected
Hi Paul,
please, please, please let me know what the cause was. I have a similar error
that my testcase fails, but for -O2, -O3 and -Os, which I have not figured yet.
- Andre
On Tue, 2 Jul 2024 11:10:37 +0100
Paul Richard Thomas wrote:
> Hi Andre,
>
> I have to hold back on the commit until th
Andre,
I am sorry but I don't have the slightest idea. I don't have an ARM system
to test it on and so I am waiting for some indications from them. As soon
as I know, you will too :-) As I indicated in my email to them, I think
that the FAIL at -O1 is bogus.
Cheers
Paul
On Tue, 2 Jul 2024 at 1
Dear all,
the attached patch addresses an effectively bogus warning about
uninitialized temporary string lengths of associate selectors.
The primary reason is that the array descriptor for a character
array is created before the corresponding string length is set.
Moving the setting of the string
