https://gcc.gnu.org/bugzilla/show_bug.cgi?id=116625
--- Comment #2 from Thiago Jung Bauermann <thiago.bauermann at linaro dot org>
---
(In reply to Andrew Pinski from comment #1)
> Since the difference is in main, I suspect if the function was named
> differently there would be no difference. That is main is known to be called
> once so the frequence of this is being taken into account.
>
> I suspect this is just a testcase issue.
Did I understand your suggestion correctly?
diff --git a/gcc/testsuite/gcc.target/arm/cmse/bitfield-4.x
b/gcc/testsuite/gcc.target/arm/cmse/bitfield-4.x
index 62e35cc3cb8e..a8ec77785a48 100644
--- a/gcc/testsuite/gcc.target/arm/cmse/bitfield-4.x
+++ b/gcc/testsuite/gcc.target/arm/cmse/bitfield-4.x
@@ -24,7 +24,7 @@ typedef void __attribute__ ((cmse_nonsecure_call)) (*foo_ns)
(test_st);
extern void foo (test_st st);
int
-main (void)
+blah (void)
{
read_st r;
foo_ns f;
I tried the above (also applied to the other tests reported here) but it didn't
work.
