PERFORCE change 18483 for review

[Alexander Kabaev]

On Wed, 2 Oct 2002 00:20:22 +0300
Giorgos Keramidas <keramida at FreeBSD.org> wrote:

> Alexander,
> 
> I've been running buildworlds with this patch for the past 2 weeks
> (not many buildworlds, I fear, but at least 4 of them).  It seems to
> work, but do you have any news from the GCC developers themselves?
> 
> Is this going to be committed to FreeBSD too, or are we supposed to
> wait for a new GCC version to be merged that includes it too?
> 
> Giorgos.

The current legend is this:

My patch has been applied on Sunday to both GCC mainline and
gcc-3_2-branch in FSF repository. I do not want to make an import though
because GCC developers are in final stages before GCC 3.2.1 release
which is supposed to happen mid-October, so doing a full import now
looks like a waste of time. Also, I have at least two more bug reports I
will try to investigate and get the patches committed before the
release. 

When GCC 3.2.1 release happens, GCC team will branch 3.3 and
preparations for 3.3 release will begin. At that point, we will have a
choice for importing either 3.2.1-release or 3.3-pre snapshot. David
O'Brien and I both think that the 3.3 snapshot is the way to go unless
we really want to get stuck with unsupported GCC release for the whole
FreeBSD 5.x series.

Given the potential for a great disturbance, I would really like to know
what re@ people think is the best. Before you do your judgment, please
take into account following points:

a) Some bugs which existed in 3.2.1 snapshot I imported first, were
fixed in GCC mainline even before GCC 3.1 was released, but newer were
backported into a stable branch.
b) once 3.2.1 branch is closed, the only fixes accepted there will be
fixes for confirmed regressions from previous GCC versions, that it, if
bug had existed in 3.0.x and 3.1.x series, it will not be fixed in 3.2
either.

--
Alexander Kabaev


-- 
Alexander Kabaev
To Unsubscribe: send mail to majordomo at trustedbsd.org
with "unsubscribe trustedbsd-cvs" in the body of the message