On Wed, Mar 26, 2025 at 10:10:07AM -0700, Andi Kleen wrote:
> Jakub Jelinek <ja...@redhat.com> writes:
>
> > --- gcc/testsuite/g++.dg/opt/musttail2.C.jj 2025-03-24 13:27:44.329204196
> > +0100
> > +++ gcc/testsuite/g++.dg/opt/musttail2.C 2025-03-24 13:28:08.975867389
> > +0100
> > @@ -0,0 +1,14 @@
> > +// PR ipa/119376
> > +// { dg-do compile { target musttail } }
>
> I think this needs to be target external_tailcall, otherwise you will
> fail on targets that don't support that.
>
> > +// { dg-options "-O2 -fno-early-inlining -fdump-tree-optimized" }
> > +// { dg-final { scan-tree-dump-times " \[^\n\r]* = foo \\\(\[^\n\r]*\\\);
> > \\\[tail call\\\] \\\[must tail call\\\]" 1 "optimized" } }
> > +
> > +struct S { S () {} };
> > +char *foo (S);
>
> Or alternatively make this not extern.
You're right (although I don't remember which targets are
non-external_musttail).
Here is a patch to define the function, tested on x86_64-linux, ok for trunk?
2025-03-26 Jakub Jelinek <ja...@redhat.com>
* g++.dg/opt/musttail2.C (foo): Define the function instead of
just declaring it, add [[gnu::noipa]] attribute to it.
--- gcc/testsuite/g++.dg/opt/musttail2.C.jj 2025-03-25 09:36:31.502487137
+0100
+++ gcc/testsuite/g++.dg/opt/musttail2.C 2025-03-26 18:17:59.827281263
+0100
@@ -4,7 +4,12 @@
// { dg-final { scan-tree-dump-times " \[^\n\r]* = foo \\\(\[^\n\r]*\\\);
\\\[tail call\\\] \\\[must tail call\\\]" 1 "optimized" } }
struct S { S () {} };
-char *foo (S);
+
+[[gnu::noipa]] char *
+foo (S)
+{
+ return 0;
+}
char *
bar (S)
Jakub
