https://gcc.gnu.org/bugzilla/show_bug.cgi?id=117657
--- Comment #9 from Andrew Stubbs <ams at gcc dot gnu.org> --- That commit should fix the build failure. However, I'm now seeing a wrong-code regression in gcc.dg/vect/vect-simd-17.c that I can't prove isn't related. The testcase now aborts, at least on gfx90a, where it used to pass. There's rather a lot going on in there, so I'll have to minimize the testcase. I'm not sure why I didn't see this issue before, but I did miss that the last version of the patch posted had extra new stuff in it, so that's probably why.
