From Bugzilla Helper: # rpm -Uvh up2date* Segmentation fault # ls up2date* up2date-2.9.16-7.x.9.i386.rpm up2date-gnome-2.9.16-7.x.9.i386.rpm up2date_script.txt # rpm -Uvh up2date_script.txt up2date_script.txt: not a package manifest: # rpm -Uvh up2date-2.9.16-7.x.9.i386.rpm up2date_script.txt Segmentation fault Steps to reproduce: 1. rpm -Uvh anypackage.rpm /etc/services Version-Release number of selected component (if applicable): rpm-4.1-0.50 How reproducible: Always Additional information: Found by accident.
Hmmm, nice catch. Uninititialized pointer happens to be NULL on 1st trip, happens to be non-NULL on remaining trips. Fix is in rpm-4.1-0.55 when built.
*** Bug 69758 has been marked as a duplicate of this bug. ***
I am not sure whether this is a new bug, or an old one, but with rpm-4.1-0.69 I still see a similar crash: % rpm -Kv up2date-2.9.33-7.x.9.i386.hdr up2date-2.9.33-7.x.9.i386.rpm up2date-2.9.38-7.x.9.i386.rpm up2date-gnome-2.9.33-7.x.9.i386.hdr up2date-gnome-2.9.33-7.x.9.i386.rpm up2date-gnome-2.9.38-7.x.9.i386.rpm coredumps in (gdb) bt #0 0x4008772a in rpmReadSignature () from /usr/lib/librpm-4.1.so #1 0x40076f62 in rpmVerifySignatures () from /usr/lib/librpm-4.1.so #2 0x40077eb4 in rpmcliSign () from /usr/lib/librpm-4.1.so #3 0x080499f7 in main () #4 0x42016664 in __libc_start_main () from /lib/i686/libc.so.6
Fixed in rpm-4.1-0.72: - fix: check for lead magic, better error message on failure (#69751). Please reopen if not.
Yes, rpm-4.1-0.75 handles it w/o any crash.