Escalated to Bugzilla from IssueTracker
I think this is due to de_thread() creating its own signal_struct and filling it in for itself rather than calling copy_signal() as it probably ought to. This means that it bypasses the extra setting up done to initialise the keyring pointers - but only in the case of exec'ing from a multithreaded program. Note that the key management facility is not available in U1.
An advisory has been issued which should help the problem described in this bug report. This report is therefore being closed with a resolution of ERRATA. For more information on the solution and/or where to find the updated files, please follow the link below. You may reopen this bug report if the solution does not work for you. http://rhn.redhat.com/errata/RHSA-2005-514.html