On Tue, Apr 26, 2016 at 2:55 PM, Jonathan Wakely <jwakely.gcc@xxxxxxxxx> wrote: > On 26 April 2016 at 13:24, Andrew Haley wrote: >> On 04/26/2016 10:51 AM, Mathieu Malaterre wrote: >>> When I >>> compile and run this code on sparc64 >> >> We'll need a complete and runnable test case. > > And the GCC options you use to compile it. OK. I was worried the code was explicitly invalid (UB). I'll incorporate some local changes I have for sparc64, push to github and update ASAP.
