3.8p1 added the following to main() in sshd.c: #ifndef HAVE_CYGWIN /* Clear environment */ environ[0] = NULL; #endif This breaks the getenv("TZ") in session.c and causes logins to occur in GMT time. It also causes any sshd syslog messages to be written in GMT time. I'm on SCO Openserver 5.0.7, but this looks like it should affect all platforms. Am I missing something? I haven't seen it reported before. Thanks. -- Roger Cornelius rac at tenzing.org
On Tue, 2 Mar 2004, Roger Cornelius wrote:> > 3.8p1 added the following to main() in sshd.c: > > #ifndef HAVE_CYGWIN > /* Clear environment */ > environ[0] = NULL; > #endif > > This breaks the getenv("TZ") in session.c and causes logins to occur in > GMT time. It also causes any sshd syslog messages to be written in GMT > time. I'm on SCO Openserver 5.0.7, but this looks like it should affect > all platforms. Am I missing something? I haven't seen it reported > before.Perhaps it should be --- sshd.c.old 2004-02-29 15:36:01.351247001 -0800 +++ sshd.c 2004-03-02 14:41:32.551939053 -0800 @@ -1106,7 +1106,7 @@ unmounted if desired. */ chdir("/"); -#ifndef HAVE_CYGWIN +#ifndef DISABLE_FD_PASSING /* Clear environment */ environ[0] = NULL; #endif Wendy, are Crays having this problem? Can someone on DEC OSF check this?> > Thanks. > >-- Tim Rice Multitalents (707) 887-1469 tim at multitalents.net
Roger Cornelius wrote:> 3.8p1 added the following to main() in sshd.c: > > #ifndef HAVE_CYGWIN > /* Clear environment */ > environ[0] = NULL; > #endif > > This breaks the getenv("TZ") in session.cGood point - this probably needs to happen a fair bit later, in session.c. -d
Roger Cornelius wrote:> 3.8p1 added the following to main() in sshd.c: > > #ifndef HAVE_CYGWIN > /* Clear environment */ > environ[0] = NULL; > #endif > > This breaks the getenv("TZ") in session.c and causes logins to occur in > GMT time. It also causes any sshd syslog messages to be written in GMT > time. I'm on SCO Openserver 5.0.7, but this looks like it should affect > all platforms. Am I missing something? I haven't seen it reported > before.That was an attempt to fix issues with certain authentication types on AIX, but it causes other problems, as you found. The change has been backed out, and an alternative fix for the AIX issues has been implemented. (Both will be in the next release, and are in the snapshots now). Thanks for the report. -- Darren Tucker (dtucker at zip.com.au) GPG key 8FF4FA69 / D9A3 86E9 7EEE AF4B B2D4 37C9 C982 80C7 8FF4 FA69 Good judgement comes with experience. Unfortunately, the experience usually comes from bad judgement.