Marc Lehmann <marc@gimp.org> writes:
> On Sat, Jan 29, 2000 at 12:44:38PM -0500, Johan Kullstam <kullstam@ne.mediaone.net> wrote:
> > judging from lurking on the gcc list it doesn't look like any new
> > release is immanent either (since there hasn't been any calls for code
> > freezes and such).
>
> Indeed, there won't be a 2.96 (or 2.97, 3.0) very soon, for, ehrm, "minor
> lack of developer resources".
i am grateful for what i get. the egcs project has given my floating
point simulation stuff about 20-30% faster run times as compared to
2.7.2.3. this helps a lot when a run can last a week or more.
> There is currently discussion on wether to release a 2.95.3 with all the
> bugfixes found on the 2.95 release branch. (I'm all in favour of
> it).
> I haven't checked wether that particular bug was fixed in the 2.95 branch,
> though.
>
> > offers a patched version of 2.95.2. where do the rest of us go for
> > working gcc? are we expected to pull things off the cvs tree at
> > cygnus?
>
> There are many versions of gcc without that bug, and if all suffer from
> the same bug it is relatively unimportant ;) So up or downgrading should
> be an option. Or waiting for the next release.
the original topic which spurred this discussion was wanting to yank
-fno-strength-reduce from the kernel build flags. allowing strength
reduction (which, if i understand correctly, is a dubious optimization
on x86 due to the dearth of registers) would not be wise as long as
the latest gcc release suffers from bugs in the strength reduction
department.
the other solution to the bug is to just keep using
-fno-strenght-reduce until such time as the current gcc with a fixed
strength reducer becomes available.
-- J o h a n K u l l s t a m [kullstam@ne.mediaone.net] Don't Fear the Penguin!- To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to majordomo@vger.rutgers.edu Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Mon Feb 07 2000 - 21:00:05 EST