Description of problem:

On one of ours machines, there is big number of segfaults seen in dmesg, like:

ld-linux.so.2[7326]: segfault at 4 ip 0053ba35 sp bfcfd894 error 4 in ld-2.12.so[535000+1e000]
tar[11071]: segfault at 2e2e2e2e ip 0050ea38 sp bf817d8c error 4 in libc-2.12.so[3d9000+189000]
make[15896]: segfault at 0 ip 0805e862 sp bfdda5f0 error 4 in make[8048000+28000]

Big majority of these are about ld-linux.so.2, and all are stating 'error 4' except these two:

npviewer.bin[26200]: segfault at 218 ip 03182496 sp bffa1a78 error 6 in libflashplayer.so[2f04000+bc4000]
dropwatch[29080]: segfault at 13 ip 0017ec93 sp bfd785d8 error 6 in libc-2.12.so[110000+189000]

I had a problem with rebuilding of openjdk src rpm yesterday, it ended with segfaults on various places, and always there was a new dmesg segfault entry after running it.

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

.qa.[root@i386-6c-m1 ~]# uname -a
Linux *** 2.6.32-220.4.2.el6.i686 #1 SMP Mon Feb 6 16:46:31 EST 2012 i686 i686 i386 GNU/Linux
.qa.[root@i386-6c-m1 ~]# rpm -q kernel
kernel-2.6.32-220.4.1.el6.i686
kernel-2.6.32-220.4.2.el6.i686
kernel-2.6.32-220.7.1.el6.i686

.qa.[root@i386-6c-m1 ~]# rpm -q glibc
glibc-2.12-1.47.el6_2.9.i686


How reproducible:

Almost always by running complex rpmbuild command for java-1.6.0-openjdk

Steps to Reproduce:
1. rebuilding some openjdk package using rpmbuild --rebuild --target=i686
2. this is seen only on one specific i686 machine
3. 
  
Actual results:

segfaults causing tasks to fail

Expected results:

no segfaults seen

Additional info:

Access to this machine available on request