glibc-2.3.2 make check error

Kelledin kelledin+LFS at skarpsey.dyndns.org
Tue Mar 4 16:54:11 PST 2003


On Tuesday 04 March 2003 05:13 pm, Greg Schafer wrote:
> > >There are some warnings about some builds breaking using
> > > that option
> >
> > I thought Greg mentioned it was broken in older builds
> > (glibc-2.2.xish) but is now actually recommended by Drepper
> > since it removes a lot of compatibility cruft.
>
> No, it has never been broken. But way back then it acted as a
> trigger to expose other problems due to dodgey build methods
> and so forth.
>
> Kelledin recently mentioned a problem that seemed related to
> this, but in actual fact, it was a straight out bug in glibc
> that is fixed in 2.3.2.

Yup, I can confirm this.  Greg's advice actually fixed the issue 
for me.

The danger isn't actually in the --enable-kernel switch itself 
being broken; the danger is that it enables glibc codepaths that 
aren't enabled by default, thus aren't so well tested.  This 
doesn't make the --enable-kernel=current switch any less risky, 
of course. ;)

So far, my glibc-2.3.1 hasn't exhibited any brokenness with 
--enable-kernel=current now that I've got it patched.  I haven't 
tried glibc-2.3.2 yet, though.  In 2.3.1, the 
--enable-kernel=current switch didn't enable any specialized 
codepaths beyond what --enable-kernel=2.4.1 would enable, but 
this may have changed in 2.3.2.  As always, YMMV. ;)

-- 
Kelledin
"If a server crashes in a server farm and no one pings it, does 
it still cost four figures to fix?"
-- 
Unsubscribe: send email to listar at linuxfromscratch.org
and put 'unsubscribe lfs-dev' in the subject header of the message



More information about the lfs-dev mailing list