similar to: Red: PAM auth with disabled user

Hi All. Has anyone else tried the current tree on Solaris 8? I installed a recommended patch cluster and now I get PAM errors, but only on a non-interactive (ie no TTY) login. I think this behaviour was introduced with the patch cluster. First thing is that in debug mode, the debug at auth-pam.c:534 derefs tty which is null, and segfaults. This occurs in debug mode only and is easy to fix.

Hi All. I (actually the tinderbox[1]) found a problem with the fix for bug #422: when PAM is enabled on a platform that uses /etc/shadow, the variable "passwd" in auth.c is used uninitialized. There's a simple patch attached to fix this. The question is: should the locked account test be done when PAM is enabled or should we rely on PAM to do the right thing? In theory they

Hi All. Today a patch was commited to OpenSSH that performs PAM password changes via SSH2 keyboard-interactive authentication. I should work fine with privsep, which some of the other solutions have problems with. While the patch itself is relatively small, it's bigger than it should have been due to differences in PAM implementations. I encourage anyone with a interest in this to try

bugzilla-daemon at mindrot.org wrote: >Summary: segfault if not using pam/keyboard-interactive mech and > password's expired I'm sorry to report that there is a bug in the PAM code in OpenSSH 3.8p1, and sorrier to say that I put it there. This is a NULL pointer dereference and is *not* considered to be a security vulnerability. When sshd is configured --with-pam, run with

Hi. If sshd is configured to use PAM and UsePrivilegeSeparation=no or you are logging is as root, any messages returned by PAM session modules are not displayed to the user. (Even when the config file has privsep=yes, logging in as root disables privsep anyway since there's no point, so it behaves the same way as privsep=no). I think I've figured out why: when privsep=no,

Hi All. Attached is another patch that attempts to do pam_chauthtok() via SSH2 keyboard-interactive authentication. It now passes the results from the authentication thread back to the monitor (based on a suggestion from djm). Because of this, it doesn't call do_pam_account twice and consequently now works on AIX 5.2, which the previous version didn't. I haven't tested it on any

Hi All. Attached is a patch that converts pam_chauthtok_conv into a generic pam_tty_conv, which is used rather than null_conv for do_pam_session. This allows, for example, display of messages from PAM session modules. The accumulation of PAM messages into loginmsg won't help until there is a way to collect loginmsg from the monitor (see, eg, the patches for bug #463). This is because the

Hi All. Damien merged a bunch of changes today which caused compile errors on a few platforms (which you can see live and in colour at [0]). a) Solaris, early AIX: ../crc32.c:100: `u_int32_t? undeclared (first use in this function) On these platforms u_int32_t is defined in defines.h which is not included by crc32.c. Fixed by attached patch. b) PAM platforms (Redhat, Solaris once a) is

Hi all. The old (~3.6.x) PAM code used to send PAM messages to stderr, whereas the new generic loginmsg code sends them to stdout, and it sends an extra newline. I think stderr is probably right, but the extra \n should probably be removed either way. -- Darren Tucker (dtucker at zip.com.au) GPG key 8FF4FA69 / D9A3 86E9 7EEE AF4B B2D4 37C9 C982 80C7 8FF4 FA69 Good judgement comes with

Olli Savia wrote: > The attached patch is a port of the current CVS (2005-08-11) version > of OpenSSH portable to LynxOS. Could you consider adding it to the > future releases of OpenSSH? If the patch needs additional work, please > let me know. Looks mostly reasonable, some comments and questions below. > + AC_DEFINE(LYNXOS_BROKEN_SETVBUF, 1, [LynxOS has broken setvbuf()

Is there a problem with the snapshots? The newest one on ftp.ca.openbsd.org is a week old. -Daz. -- 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.

Does anyone actually use sshd on a system that doesn't have socketpair? It's used elsewhere so the don't-have path seems like it'd never be exercised these days. Index: monitor.c =================================================================== RCS file: /usr/local/src/security/openssh/cvs/openssh/monitor.c,v retrieving revision 1.147 diff -u -p -r1.147 monitor.c --- monitor.c

Hi All. For a couple of weeks, the Portable snapshots have contained optional support for NetBSD's libedit in the sftp client, thanks to djm's work in OpenBSD. It's enabled with: ./configure --with-libedit. If enabled, sftp gains command history, recall and line editing (and probably other features too, I haven't looked into libedit's capabilities much). If not

I am just looking for a quick answer as to whether or not OpennSSH is compatible with Digital Unix Tru64 v 4.0F. Hing Fei Wong Systems Engineer Building 100, M1309 Valley Forge, PA Admin # 4-6242 -----Original Message----- From: Darren Tucker [mailto:dtucker at zip.com.au] Sent: Friday, June 23, 2006 3:53 AM To: Wong, Hing Fei Cc: www at openbsd.org Subject: Re: OpenSSH compatibility with

Hi All. As of now, openssh on AIX passes all regressions tests (and, yes, I just checked!), works with privsep, bugzilla has zero open AIX-specific bugs and IBM ship it essentially unmodified as a supported product. I think it's beyond "support underway" :-) -Daz. -- Darren Tucker (dtucker at zip.com.au) GPG key 8FF4FA69 / D9A3 86E9 7EEE AF4B B2D4 37C9 C982 80C7 8FF4 FA69

Hi all. I was getting something working over socks5 and was trying to figure out why it kept using socks4. It wasn't, it was just a misleading debug message.... Patch applies to either OpenBSD or Portable. -- 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

Hi All. Right now, if OpenSSH is configure'ed --with-kerberos5 and the system has a krb5-config that's not in /usr/local/bin then configure won't find it. The attached patch changes this so krb5-config will be used if it's anywhere in the path (although if it exists in the directory specified by --with-kerberos5= then the user-supplied path will take precedence). You will

Hi All. The attached patch speeds up the dynamic forwarding regression test: * moves starting the test sshd to the outer loop. * kills the sleep of when it's no longer required. -Daz. $ time PATH="`pwd`:$PATH" sh ../regress/test-exec.sh `pwd` \ ../regress/dynamic-forward.orig.sh ok dynamic forwarding real 0m54.585s user 0m5.760s sys 0m0.370s $ time

"DiNisco, Jeff" wrote: > I read in the change log that you fixed a bug that denies access to > accounts with locked passwords. My environment is dependent on public > key authentication. The account used does not have a person associated > with it but rather a service. I want to keep the password locked. Is > there a way to turn this fix off? What platform are we

Hi All. While working on something I noticed core dumps from sshd. They don't seem to be related to what I was working on. It's from the process forked to run the shell. Just after the fork, fatal_remove_all_cleanups() is called, which looks like: fatal_remove_all_cleanups(void) { struct fatal_cleanup *cu, *next_cu; for (cu = fatal_cleanups; cu; cu = next_cu) {