Darryl Okahata <darrylo(a)soco.agilent.com> writes:
Vin Shelton <acs(a)xemacs.org> wrote:
> Thank you, but I agree with Daryl Okahata's assessment if gcc-3.4.2
> with certain aggressive optimizations requires code like this, that's
> a bug in the compiler and we should advise users not to use that
> compiler with a particular set of options (perhaps on a particular
> platform).
I have not verified this, but the OP commented that the compiler is
buggy even without any optimization flags (even though the subject line
implies otherwise). If so, then perhaps configure.in should be modified
to disallow building with that compiler.
If that claim can be verified, I would (at the least) approve an entry
for the PROBLEMS file. Disallowing that compiler in configure is
another possibility, but that seems fairly drastic, as I'm pretty sure
that some people have gotten useful builds out of gcc-3.4.2.
- Vin