https://gcc.gnu.org/bugzilla/show_bug.cgi?id=123377

--- Comment #4 from Hans-Peter Nilsson <hp at gcc dot gnu.org> ---
(In reply to Andrew Pinski from comment #3)
> The difference is `--param logical-op-non-short-circuit=0` vs `--param
> logical-op-non-short-circuit=1` checked on x86_64 and powerpc64-linux-gnu.

That certainly rings a bell!

> What I am going to do is create a new version which sets `--param
> logical-op-non-short-circuit=0` and the original will be for `--param
> logical-op-non-short-circuit=1`.
> 
> I don't think there is a target supports that is for
> logical-op-non-short-circuit either. I am not sure if there should be one
> which is why my suggestion is to create 2 testcases out of this one and
> xfail the one for `logical-op-non-short-circuit=1` case. At least then we
> have better control at when things fail/pass.

Sounds good.  Thanks!

Reply via email to