Re: Missed gcc 3.2/3.3 patch?

[Joe Buck <jbuck at synopsys dot com>]

On Thu, Oct 16, 2003 at 10:26:59AM -0700, H. J. Lu wrote:
> On Thu, Oct 16, 2003 at 10:11:57AM -0700, Joe Buck wrote:
> > On Wed, Oct 15, 2003 at 09:34:50AM -0700, H. J. Lu wrote:
> > > Jakub submitted a patch for 3.2/3.3/trunk:
> > > 
> > > http://gcc.gnu.org/ml/gcc-patches/2003-04/msg00392.html
> > > 
> > > For some reason, it is not in 3.2/3.3. Without it, glibc won't work
> > > correctly on ia64. Can it be checked into 3.2/3.3?
> > 
> > 3.2 is gone and done.  For 3.3.2 it's up to Mark.  But this seems very odd -
> > people have been shipping distros with 3.2.x on ia64.  What have
> > they been doing about this problem?
> 
> Red Hat uses gcc-3_2-rhl8-branch, which has the patch. The same goes
> for gcc-3_3-rhl-branch. See:
> 
> http://sources.redhat.com/ml/libc-hacker/2003-10/msg00011.html
> 
> Any reason not to fix gcc 3.3?

That's not how it goes.  As advocate to get a patch in at the last minute,
it's up to you to explain why it is so vital that the 3.3.2 release, which
Mark might have been planning to ship TODAY, must be delayed; it is not up
to others to give a reason why not.

Sigh.  It's really too bad that the glibc people didn't bother to notify
the gcc list until way after the 3.3.2 hard freeze, even though the
message you point to shows Jakub raising this issue on the 9th, saying

> Might ask Mark if he allows that for 3.3.2...
> Certainly, without that not just backtrace, but lots of other things (like
> NPTL) don't work properly.

Unless someone asked Mark by private mail, it seems that the ball was
dropped, and a whole week passed before anyone did ask.  This kind of
thing is not fair to the RM.

It's also a concern if the glibc developers are using Red Hat versions of
gcc rather than FSF versions in their development, and not noticing
problems until too late.

Mark, it's your call.