trying to understand what's going on with timers and boot process

[Mark Seger <Mark dot Seger at hp dot com>]

I use the perl module Time::HiRes to set alarms.  It had been working 
just fine for years and years.  Then a few years ago it starting failing 
on systems with glibc 2.4 installed and timers > 4.4 seconds because it 
was passing the whole thing in microseconds instead of passing the 
seconds portion as seconds.  I reported this to the author and he fixed 
it  and it works just fine now.  BUT I've found modern distros, such as 
RHEL5, are still shipping the old HiRes module and they're also bundling 
in glibc 2.5.

Now for my question - I've observed that 2.5 does appear to work 
correctly even with the incorrect parameters being passed so clearly 
something has changed since 2.4.  BUT the other thing I've observed is 
if I try to set a timer >1 second during boot, I get setitimer errors 
and once the system boots I don't.  In other words, if I start a perl 
script via /etc/init.d which sets a .5 second timer and sleeps, then 
increment it by a second and do it again, etc  It generates a setitimer 
error for every call from 1.5-10.5 seconds BUT then stops as the time 
gets up to about 15 and the script exits.  I don't know if there is 
something magic about the numbers 1.5-10.5 OR if some system resource 
has changed by the time the 11.5 second alarm is set.

My issue is I have an open source program that is dependent on HiRes and 
is being distributed on Fedora 10 and I fear I'm going to start getting 
a lot of support requests over this.  I do intend to release a new 
version that checks the version of glibc (I recently discovered I could 
just run /lib/libc.6.so and it tells me its version) and warn people 
when there is a problem, but I'm still stumped with the fact that I've 
only seen errors on 2.5 systems during boot.  Can someone shed any light 
on what I'm seeing?  Should I be worrying about ALL versions of glibc > 2.3?

-mark