[PATCH] Fix the two C regressions caused by my non-lowering patch

Jason Merrill jason@redhat.com
Tue Aug 24 15:10:00 GMT 2004


On Mon, 23 Aug 2004 20:21:35 -0700, Andrew Pinski <pinskia@physics.uc.edu> wrote:

> I was reminded by RTH that I also caused two C regressions which really
> is a gimplifier bug and can be fixed by having the gimplifier not convert
> array[x] "into the equivalent *(&array + offset) form."

What sort of bugs?  Again, both forms are correct, and any codegen errors
are a different bug.  The latter form may be less optimizable, but we
should still handle it correctly.

Jason



More information about the Gcc-patches mailing list