reping (Arun/Yvan) - I'd love to get our buildbot running off upstream master.
On Sun, Oct 19, 2014 at 11:23 PM, C Bergström <[email protected]> wrote: > I thought it would be trivial to do, but here's the updated patch. I have > more patches for other things like Solaris and cmake. I hope that can get > upstreamed as well. > > Thanks > > On Fri, Oct 10, 2014 at 10:03 PM, Yvan Roux <[email protected]> wrote: > >> On 10 October 2014 16:51, Arun Sharma <[email protected]> wrote: >> > You wrote: >> > >> >> And yes, it seems we need to restore x0, x1, x2, x3, not only x0 and >> x1. >> > >> > Did I miss a patch sent to the list? >> >> I haven't seen it too. >> >> BTW, I may buy Christian's argument that not checking eh_valid_mask is >> slower than restoring the registers, but are we compliant with the >> specification of the function in that case ? and Arun do you >> understand IA64 code which restored the exception argument register >> when a signal is involved only when eh_valid_mask is set, and always >> restore them in the normal path ? My understanding of unw_resume >> documentation is that is should be the opposite. >> >> Yvan >> > >
_______________________________________________ Libunwind-devel mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/libunwind-devel
