11:25 a.m.
Red Hat has always been at the forefront of GCC development and deployment
to its userbase. Are there plans to migrate from current GCC 4.1.1 to GCC
4.2 snapshots for FC6(T.)?
The branch has reached a fairly mature status named "stage 3" which means
that only regressions and documentation are subject to fixes.
1:47 p.m.
>>>> "Joachim" == Joachim Frieben
<jfrieben(a)freesurf.fr> writes:
Joachim> Red Hat has always been at the forefront of GCC development
Joachim> and deployment to its userbase. Are there plans to migrate
Joachim> from current GCC 4.1.1 to GCC 4.2 snapshots for FC6(T.)?
No, I'm afraid not.
Tom
1:58 p.m.
On Wed, 2006-07-19 at 18:25 +0200, Joachim Frieben wrote:
Red Hat has always been at the forefront of GCC development and
deployment
to its userbase. Are there plans to migrate from current GCC 4.1.1 to GCC
4.2 snapshots for FC6(T.)?
The branch has reached a fairly mature status named "stage 3" which means
that only regressions and documentation are subject to fixes.
But we can't integrate it without a good feeling that the actual release
of gcc 4.2 will be out prior to the FC6 feature freeze[1]. And with
where gcc is in its schedule, that's not the case
Jeremy
[1] Which is the test 2 freeze, so now-ish.
2:01 p.m.
Joachim Frieben writes:
Red Hat has always been at the forefront of GCC development and
deployment
to its userbase. Are there plans to migrate from current GCC 4.1.1 to GCC
4.2 snapshots for FC6(T.)?
The branch has reached a fairly mature status named "stage 3" which means
that only regressions and documentation are subject to fixes.
Are you sure? As far as I can see the branch hasn't even been created
yet.
3:11 p.m.
Hi Andrew,
no, I'm not sure. I was simply referring to information available at the GCC
web site http://gcc.gnu.org. Quote:
"Active development: GCC 4.2.0 (changes)
Status: Stage 3; 2006-06-04 (regression fixes and docs only)."
I would be surprised if they were ahead of the factual state.
-Joachim
Are you sure? As far as I can see the branch hasn't even been created
yet.
3:17 p.m.
On Wed, Jul 19, 2006 at 10:11:04PM +0200, Joachim Frieben wrote:
no, I'm not sure. I was simply referring to information available
at the GCC
web site http://gcc.gnu.org. Quote:
"Active development: GCC 4.2.0 (changes)
Status: Stage 3; 2006-06-04 (regression fixes and docs only)."
I would be surprised if they were ahead of the factual state.
Please read
http://gcc.gnu.org/develop.html
When gcc-4_2-branch will be created, active development will talk about GCC
4.3.0.
Jakub
2:03 p.m.
On Wed, Jul 19, 2006 at 06:25:26PM +0200, Joachim Frieben wrote:
Red Hat has always been at the forefront of GCC development and
deployment
to its userbase. Are there plans to migrate from current GCC 4.1.1 to GCC
4.2 snapshots for FC6(T.)?
Not for FC6. The thing is, the end of stage3 phase is not close, there are
currently 160 important regressions. Even when gcc-4_2-branches after it, it
will take another couple of months before it is released. If you compare
it with the FC6 tentative schedule, it is clear that it is too late.
Not to mention that several of the interesting 4.2 features are backported
on redhat/gcc-4_1-branch (OpenMP, reassoc rewrite, C++ visibility, likely
Java improvements, various Fortran improvements just to name a few).
Jakub
2:16 p.m.
On Wed, 2006-07-19 at 18:25 +0200, Joachim Frieben wrote:
Red Hat has always been at the forefront of GCC development and
deployment
to its userbase. Are there plans to migrate from current GCC 4.1.1 to GCC
4.2 snapshots for FC6(T.)?
The branch has reached a fairly mature status named "stage 3" which means
that only regressions and documentation are subject to fixes.
Wouldn't using GCC 4.2 mean recompiling everything again, with the risk
it introduces compiler depended bugs that don't show up when things are
compiled with GCC 4.1.x. And wouldn't it than better to do that at the
start of the FC7 cycle instead of the end of the FC6 cycle?
- Erwin
2:30 p.m.
On Wednesday 19 July 2006 15:16, Erwin Rol wrote:
Wouldn't using GCC 4.2 mean recompiling everything again, with
the risk
it introduces compiler depended bugs that don't show up when things are
compiled with GCC 4.1.x. And wouldn't it than better to do that at the
start of the FC7 cycle instead of the end of the FC6 cycle?
Yes, one mass rebuild per release is plenty, thank you.
--
Jesse Keating
Release Engineer: Fedora
6515
days inactive
6515
days old
8 comments
7 participants
participants (7)
-
Andrew Haley -
Erwin Rol -
Jakub Jelinek -
Jeremy Katz -
Jesse Keating -
Joachim Frieben -
Tom Tromey