From Bugzilla Helper:
User-Agent: Mozilla/5.0 (X11; U; SunOS i86pc; en-US; rv:1.2.1) Gecko/20030121

Description of problem:
Got a very similar problem on a fresh RedHat 9 system (rpm-4.2-0.69) while
updating glibc RPMs: hung, would not respond to signals, etc.  This time, the
strace showed:

# strace -p 8804
futex(0x405940cc, FUTEX_WAIT, 0, NULL <unfinished ...>

This looks like the same bug in 77562, 89554, and 92270.  The usual remedy
(sigKILL the RPM process, rm the __db lockfiles, rebuild the database) got me a
working RPM installation again, but trying the same installation again got the
same hang with the same trace.  

Those tickets keep being closed with some talk about signals, but no action
appears to be taken on the root problem-- the fact that RPM is hanging in the
first place!  Manually deleting lockfiles and rebuilding is fine on your home
workstation, but far from practical in a large enterprise environment where you
might be deploying an RPM on a hundred machines at once....  This is a very
serious problem and is potentially impacting our migration plans.

Version-Release number of selected component (if applicable):


How reproducible:
Sometimes

Steps to Reproduce:
1.In the redhat updates/i386 directory: rpm -Uvh glibc-utils-2.3.2-27.9.i386.rpm
 glibc-common-2.3.2-27.9.i386.rpm glibc-profile-2.3.2-27.9.i386.rpm
glibc-debug-2.3.2-27.9.i386.rpm glibc-devel-2.3.2-27.9.i386.rpm
../i686/glibc-2.3.2-27.9.i686.rpm ../i686/nptl-devel-2.3.2-27.9.i686.rpm
2. Voila!  It hangs.  Usually.  
3.
    

Actual Results:  RPM hangs, will not respond to signals.

Expected Results:  RPM installed-- or at least a clean failure with an error
message.

Additional info:

I am marking this "high" because it makes RedHat 9 (and 8.0, for that matter) an
unmaintainable distribution in my environment.

It seems to be more likely to hang if many RPMs are specified on the argument list.