[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: List of bugs that *must* be fixed before releasing Hamm

On Fri, May 29, 1998 at 09:59:43AM -0400, Brian White wrote:
> > > egcc              17768  libc6 bug in mathematical functions [114]  (Galen Hazelwood <galenh@micron.net>)
> > 
> > Fixing this bug exceeds my capabilities. I have send a message to the
> > egcs-bugs list asking them to look into the current reports, but have
> > received no responses.
> Would it be possible to remove "-force-mem" from the list of optimizations
> done by "-O2"?

I don't know. I'm not an expert on (e)gcc internals.

> > The bug only occurs with optimisation enabled, so it can be worked
> > around. egcc is the secondary compiler on the architectures for which
> > hamm will be released (i386 and m68k AFAIK). Personally, I don't think
> > this bug should delay the release of hamm as 2.0. Brian, please consider
> > downgrading this to Severity: normal .
> It wouldn't delay the release...  the package would be removed (ala the
> posting I made last Friday).

<mode = "posting after cooling down">
That is your priviledge as release manager. However, as a user and a
developer, I'd like to point out to you that this is a problem of the egcs
suite as a whole. As our current C++ compiler is the egcs one, you'd have to
remove it too (or switch back to the one, which is a lot more
broken). As packages in main must be compilable using packages in main,
you'd have to remove all packages compiled with it. You would end up with
something that fits the release criteria, but which would be still-born.

> > > ftp.debian.org    22390  Please move slink gcc packages to frozen [15]  (Guy Maor <ftpmaster@debian.org>)
> > 
> > This will be obsoleted when will be accepted into frozen.
> The changelog says "should fix #22292".  Are you not sure it does?

My development station is a laptop with severe disk space constraints. I was
not in a position to check this. I did the part that required my attention,
and stuck to the "testing is parallelisable" paradigm.
I had every reason to assume it does (and it does as Phil confirmed), but I
did not feel like downloading libc6 in the middle of the night and build it
(due to the disk constraints, I'd had to have the gcc packages finished and
their source tree cleaned up before I could attempt building libc6).
So I didn't test it, which is why it says "should fix".

LEADERSHIP  A form of self-preservation exhibited by people with auto-
destructive imaginations in order to ensure that when it comes to the crunch 
it'll be someone else's bones which go crack and not their own.       
- The Hipcrime Vocab by Chad C. Mulligan    

To UNSUBSCRIBE, email to debian-qa-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org

Reply to: