This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: Committed: verify_flow_info fixes exposed by libgfortran (at least on MMIX)
> Bummer. I won't be able to build for a few hours, so if you can
> bootstrap with my patch reverted on some target (but *with
> rtlchecking*), I'm happy with reverting it and will submit a
> better patch.
I bootstrapped with RTL checking a tree dated 10/17 on i486 yesterday.
> Or maybe it's safer to let barriers have a bb
> pointer?
I don't really know. If you think it's the best solution.
> BTW, I used configure with no enable/disable options when
> testing. I guess building with rtl-checking enabled for a few
> targets would show interesting things.
I regularly bootstrap with RTL checking on amd64. I'm under the impression
I'm not the only one (Andreas Jaeger from SuSE comes to mind).
--
Eric Botcazou