https://gcc.gnu.org/bugzilla/show_bug.cgi?id=91419
--- Comment #4 from Hans-Peter Nilsson <hp at gcc dot gnu.org> --- (In reply to rguent...@suse.de from comment #3) > OK, so all testcases in this PR use 'int' which means disabling > for !natural_alignment_32 would be enough (unless 'int' is not 32bit ;)) I'd assume alert target maintainers will notice and take action. > I agree that using target lists is imperfect (but it has the lowest > testsuite time overhead). JFTR, I'd say lower maintenance time beats lower test-run time ;) and the attribute is suitable for caching. But now that it's there I'd probably also just pile-on. > I'll amend the testcases with xfails for > !natural_alignment_32. Can you amend the target lists appropriately? > I guess cris-*-* (not only cris-*-elf) would need to be added. It seems you already did this in your patch, that's right, thanks!