From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (qmail 12912 invoked by alias); 16 Dec 2009 18:41:32 -0000 Received: (qmail 12884 invoked by uid 48); 16 Dec 2009 18:41:19 -0000 Date: Wed, 16 Dec 2009 18:41:00 -0000 From: "pterjan at mandriva dot com" To: glibc-bugs@sources.redhat.com Message-ID: <20091216184119.11099.pterjan@mandriva.com> Reply-To: sourceware-bugzilla@sourceware.org Subject: [Bug libc/11099] New: INT_FIELD_MAYBE_NULL changed behaviour on x86_64 X-Bugzilla-Reason: CC Mailing-List: contact glibc-bugs-help@sourceware.org; run by ezmlm Precedence: bulk List-Id: List-Subscribe: List-Post: List-Help: , Sender: glibc-bugs-owner@sourceware.org X-SW-Source: 2009-12/txt/msg00049.txt.bz2 We had password aging bugs on Mandriva and I found that when a field of /etc/shadow contains -1, it gets 32 bits value of -1 on x86_64 even if the field is long in the struct. According to the doc it should indeed be empty to disable instead of containing -1, but well the field is signed and nothing prevents from having a negative value It used to work and the behavior change broke tests like spw->sp_* != -1 in pam modules The change was introduced by http://repo.or.cz/w/glibc.git/commitdiff/1e545d0185072ac0d4646223d509c3692cbe550a with a commit message unrelated to this part of the commit -- Summary: INT_FIELD_MAYBE_NULL changed behaviour on x86_64 Product: glibc Version: 2.11 Status: NEW Severity: normal Priority: P2 Component: libc AssignedTo: drepper at redhat dot com ReportedBy: pterjan at mandriva dot com CC: glibc-bugs at sources dot redhat dot com http://sourceware.org/bugzilla/show_bug.cgi?id=11099 ------- You are receiving this mail because: ------- You are on the CC list for the bug, or are watching someone who is.