On Thu, May 02, 2019 at 08:41:29PM +0200, Rainer Orth wrote: > Hi Segher, > > > On Thu, May 02, 2019 at 03:47:33PM +0200, Rainer Orth wrote: > >> I've now tested this RC on i386-pc-solaris2.1[01] and > >> sparc-sun-solaris2.1[01]. The only issue (apart from the just-fixed > >> spellcheck-options-5.c testcase) is > >> > >> +FAIL: gcc.target/i386/pr90178.c scan-assembler-times xorl[\\\\t > >> ]*\\\\%eax,[\\\\t ]*%eax 1 > >> > >> While this has already been fixed on mainline, it's still present on the > >> gcc-9 branch on every x86 target. > > > > It's not a regression though (the testcase is new). > > true, but in a way that's even worse: a new target-specific testcase > failing on exactly that target (and a primary platform at that) doesn't > seem particularly convincing to me: we should be getting rid of > testsuite failures now, not introducing new ones at the 11th hour.
If you prefer to not know about the bug, sure, delete the testcase :-/ The problem is fixed on trunk already, fwiw. Segher
