Re: GCC Release Status (2003-06-20)

Re: GCC Release Status (2003-06-20)

[Wolfgang Bangerth]

> (I do not know how to create a 3.5 target, but we should have one for
> this kind of bug.  Or perhaps a tree-ssa target.  I will pay one
> gnudit to whoever adds such a thing!)

:-) I think this one will have to be spent in Texas then :-)

I created a tree-ssa target. We will have a 3.5 target when we branch 3.4 or 
ealier if you think it's useful. A milestone tree-ssa actually makes sense as 
the milestone to be reached before the branch can be merged.

Is there any interest in an update of the list I sent earlier this week, or 
will this be a byproduct of your triaging? Or is there anything else one 
could help you in this respect?

W.

-------------------------------------------------------------------------
Wolfgang Bangerth              email:            bangerth@ices.utexas.edu
                               www: http://www.ices.utexas.edu/~bangerth/

Re: GCC Release Status (2003-06-20)

[Mark Mitchell]

On Fri, 2003-06-20 at 17:49, Wolfgang Bangerth wrote:
> 
> > (I do not know how to create a 3.5 target, but we should have one for
> > this kind of bug.  Or perhaps a tree-ssa target.  I will pay one
> > gnudit to whoever adds such a thing!)
> 
> :-) I think this one will have to be spent in Texas then :-)
> 
> I created a tree-ssa target. We will have a 3.5 target when we branch 3.4 or 
> ealier if you think it's useful. A milestone tree-ssa actually makes sense as 
> the milestone to be reached before the branch can be merged.

Thanks!  Here's your gnudit. :-)

> Is there any interest in an update of the list I sent earlier this week, or 
> will this be a byproduct of your triaging? Or is there anything else one 
> could help you in this respect?

Yes, an update of your list would be useful to me.  I've used it to
cherry-pick some low-hanging fruit.  (In the usual way, I now get to
iterate on a couple of those changes -- I've got regressions on Darwin
and the something still wrong with the type-hashing code I added to fix
before I get to any *new* PRs...)
 
Thanks,

-- 
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com

GCC Release Status (2003-06-20)

[Mark Mitchell]

I'm going to try to get back in the habit of sending weekly mail
summarizing our progress towards releases.

--
Mark Mitchell
CodeSourcery, LLC
mark@codesourcery.com

GCC 3.3.1
=========

GCC 3.3.1 is 28 days away.

There are 122 regressions targeted for GCC 3.3.1.

Of these, 49 are C++ PRs, which clearly makes the C++ front end the
biggest area of focus.

There are also about 20 optimization PRs and 20 target-specific PRs.

I will be triaging the 3.3.1 PRs this weekend, or early next week.
Many of the less-important PRs will be re-targeted for 3.3.2.

Despite the large number of regressions, I think the 3.3.1 release is
in relatively good shape.  Many of the PRs are regressions relate to
warnings, ICEs on invalid code, and other not-too-critical areas.

GCC 3.4
=======

GCC 3.4 Stage 2 starts in 14 days.

There are 268 PRs targeted for GCC 3.4 at present.  Of course, many of
thse are the same as the 3.3.1 PRs mentioned above.  C++ again
represents the biggest share of the PRs, with 106.  

Some PRs are clearly mislabeled; for example PR 11266 is a tree-ssa
bug, but has 3.4 in its target release.  Since tree-ssa will not be in
3.4, this labeling is in error.  

(I do not know how to create a 3.5 target, but we should have one for
this kind of bug.  Or perhaps a tree-ssa target.  I will pay one
gnudit to whoever adds such a thing!)