Epochs

2001-06-23 Thread Randolph Chung
(re ajt's comments to bug #101878) gcc-defaults doesn't build packages with epochs by default, so if we can keep it that way we'll be more consistent with the other architectures. But I think we just want to get something into the archive that works, so whatever we can agree to . cc'ing the

new gcc for potato?

2001-06-23 Thread Philip Blundell
I think our current gcc 2.95.4 is stable enough, and sufficiently better than the 2.95.2 in potato, that we should consider making new packages to go into 2.2r4 or whatever the next version is going to be. I guess this should be straightforward enough to achieve. Anybody object to this? If

new gcc for potato?

2001-06-23 Thread Matthias Klose
Philip Blundell writes: I think our current gcc 2.95.4 is stable enough, and sufficiently better than the 2.95.2 in potato, that we should consider making new packages to go into 2.2r4 or whatever the next version is going to be. I guess this should be straightforward enough to achieve.

Bug#101901: g++-3.0 and --use-cxa-atexit

2001-06-23 Thread Matthias Klose
Chris, any support in ld needed? laurent bonnaud writes: Package: g++-3.0 Version: 1:3.0-1 Severity: normal Hi, according to http://gcc.gnu.org/bugs.html#known, it would seem to be a good idea to compile g++-3.0 with the --use-cxa-atexit switch: Global destructors are

Re: new gcc for potato?

2001-06-23 Thread Philip Blundell
the current 2.95.4 doesn't builf on s390, but 2.95.3, so it might be necessary to add a reverse-diff (for woody as well). Is there any bug open for this? I couldn't find one from a quick look at the lists for gcc and gcc-2.95. In any case I don't think we have to worry about it for potato --

Bug#101901: g++-3.0 and --use-cxa-atexit

2001-06-23 Thread Daniel Jacobowitz
On Sat, Jun 23, 2001 at 02:47:51PM +0200, Matthias Klose wrote: Chris, any support in ld needed? laurent bonnaud writes: Package: g++-3.0 Version: 1:3.0-1 Severity: normal Hi, according to http://gcc.gnu.org/bugs.html#known, it would seem to be a good idea to

Re: new gcc for potato?

2001-06-23 Thread Philip Blundell
I sincerely doubt that this will ever get past the Release Manager unless you have a very good, very specific reason. I recommend talking to him before spending your time. I think it's worth making the packages even if the Release Manager (who is that for potato these days, anyway?) won't

gcc-3.0 doesn't build shared C++ library on i386-pc-gnu

2001-06-23 Thread Jeff Bailey
Submitter-Id: net-debian Originator:Jeff Bailey Organization: Confidential: no Synopsis: gcc-3.0 doesn't build shared C++ library on i386-pc-gnu Severity: serious Priority: medium Category: c++ Class: sw-bug Release: 3.0 20010426 (Debian prerelease)

Re: new gcc for potato?

2001-06-23 Thread Daniel Jacobowitz
On Sat, Jun 23, 2001 at 08:23:17PM +0100, Philip Blundell wrote: I sincerely doubt that this will ever get past the Release Manager unless you have a very good, very specific reason. I recommend talking to him before spending your time. I think it's worth making the packages even if the

Bug#101901: g++-3.0 and --use-cxa-atexit

2001-06-23 Thread Daniel Jacobowitz
On Sun, Jun 24, 2001 at 12:00:36AM +0200, Matthias Klose wrote: Ok to enable -fuse-cxa-atexit as default? According to Laurents citations this should be safe for Linux glibc. Safe for the Hurd as well? Should be. Might want to run it through a testsuite just to be sure, of course. -- Daniel