NPTL Wierdness, degraded?

[SpEnTBoY]

Ok I'm confused and hoping someone here can help or at least explain what would have happened.

I had a stock CentOS 4 machine with Zimbra installed. I figured I would document my own install as I'm going to install CentOS5 on this machine when it's released. I ran the installer to remove Zimbra (which worked fine) and then proceeded to re-install it.

I got the error mentioning that NPTL was not found I double checked on other CentOS 4 machines and got this:

#getconf GNU_LIBPTHREAD_VERSION
NPTL 2.3.4

But on my machine I'm now showing:
# /usr/bin/getconf GNU_LIBPTHREAD_VERSION
linuxthreads-0.10

Does anyone have any ideas how this would revert BACK? Here are some specifics:

# uname -rvmpio
2.6.9-42.0.10.EL #1 Tue Feb 27 09:24:42 EST 2007 i686 athlon i386 GNU/Linux

# cat /proc/cpuinfo | grep -i model
model : 10
model name : AMD Athlon(tm) XP 2800+

I'm just not seeing how this reverts. Everything I've read says it's the default threading now (since RHEL3 even) and every one of my machiens is showing it installed to some capacity. Thing is it had to be available for me to even install Zimbra the first time so I'm trying to track what happened that was different on the box.

This is a very vanilla install with only really two applications on it (or at least it DID have two working lol):

1. Zimbra
2. Bacula

Anyone out there have some suggestions, ideas, or anything else I could check? Sorry if this is off topic but I've been hunting for a while now and noticed that if it's a pre-req, maybe somenoe here could help.

Regards,
Lonny

[SpEnTBoY]

Another application had mentioned that the tls lib directory should be moved as there was an issue. In another part of the doc it mentions the issue was with tls on older 2.4 kernels. Unfortunatley, libraries were updated while the tls lib dir was moved and things broke horribly.

Hopefully this doesn't happen to anyone else