On Mar 29, 2026, Alexandre Oliva <[email protected]> wrote:
> On Mar 26, 2026, Alexandre Oliva <[email protected]> wrote:
>> diff --git a/gcc/testsuite/gcc.target/powerpc/safe-indirect-jump-1.c
>> b/gcc/testsuite/gcc.target/powerpc/safe-indirect-jump-1.c
>> index b9ad8c1dd36a6..adecb3cd67233 100644
>> --- a/gcc/testsuite/gcc.target/powerpc/safe-indirect-jump-1.c
>> +++ b/gcc/testsuite/gcc.target/powerpc/safe-indirect-jump-1.c
>> @@ -18,5 +18,5 @@ int bar ()
>> /* { dg-final { scan-assembler "beqctrl-" { target { lp64 || {
>> powerpc*-*-aix* } } } } } */
>> /* The other ABIs do allow a sibcall. */
>> -/* { dg-final { scan-assembler "beqctr-" { target { ilp32 &&
>> !powerpc*-*-aix* } } } } */
>> -/* { dg-final { scan-assembler {b \$} { target { ilp32 && !powerpc*-*-aix*
>> } } } } */
>> +/* { dg-final { scan-assembler "beqctr-" { target { ilp32 && { !
>> powerpc*-*-aix* } } } } } */
>> +/* { dg-final { scan-assembler {b \$} { target { ilp32 && { !
>> powerpc*-*-aix* } } } } } */
> FTR, this exposed a latent bug in the testcase. Some 32-bit ppc targets
> are now visibly failing because they have long failed to output the
> expected (?) ilp32 patterns, but we didn't check for them because
> '!powerpc*-*-aix*' never matched anything.
I went all the way back to r8-6397, where safe-indirect-jump-8.c was
folded into -1.c, and even to r8-6198, where -8.c was introduced, and
couldn't get a sibcall on powerpc-elf, powerpc-eabi, or
powerpc-linux-gnu. In desperation :-) I even tried -mabi=elfv{1,2} and
some -mcpu options, to no avail.
Could you please let me know what ilp32 non-AIX ABI you likely had in
mind when you made those changes?
Failing that, should we make the test for beqctrl- unconditional and
wait for reports to find out where that breaks?
--
Alexandre Oliva, happy hacker https://blog.lx.oliva.nom.br/
Free Software Activist FSFLA co-founder GNU Toolchain Engineer
More tolerance and less prejudice are key for inclusion and diversity.
Excluding neuro-others for not behaving ""normal"" is *not* inclusive!
