Re: internal compiler error

[Roland McGrath]

> Do you see any value in testing the version 3 branch for these types of 
> problems, or should we test the 2.95 branch?

I guess I only answered half your question.  We should definitely make sure
the bugs triggered by the Hurd sources under -O3 get reported and fixed
before a GCC release is made.  But if the GCC folks are quickly fixing lots
of related bugs now and are hearing as many reports as they can look into,
then there's no reason we need to burden them now rather than just trying
again when the GCC 3 code base has settled down a bit more.