thr3ads.net - openssh unix dev - OSF/1 or Tru64 patch for Privsep [Aug 2002]

If this information is useful, please help other people find it:
Share via:

Ben Lindstrom

2002-Aug-11 22:37 UTC

OSF/1 or Tru64 patch for Privsep

Either this never made it to the list or no one cares about Tru64.  This
is the last time I'll send this patch to the list.  If no one steps up and
finishes it or provides me with enough information to fix any remaining
bugs (one being complaint that 'ssh site cmd' does not work right).

If there is no activity on this for a week.  I'll post it to bugzilla and
will ignore any request for me to fix Tru64.

I do thank everyone that has privately contributed help in regards to
testing and such.

This patch may apply to 3.4p1, but I'd perfer testing to be done on -cvs.

- Ben

diff -ur openssh-3.4p1/auth-sia.c openssh-3.4p1+/auth-sia.c
--- openssh-3.4p1/auth-sia.c	Fri Apr 12 11:36:08 2002
+++ openssh-3.4p1+/auth-sia.c	Mon Jul 22 23:34:37 2002
@@ -41,9 +41,11 @@
 #include <unistd.h>
 #include <string.h>

+/* import */
 extern ServerOptions options;
 extern int saved_argc;
 extern char **saved_argv;
+extern int use_privsep;

 extern int errno;

@@ -77,7 +79,7 @@
 }

 void
-session_setup_sia(char *user, char *tty)
+setup_sia(char *user, char *tty)
 {
 	struct passwd *pw;
 	SIAENTITY *ent = NULL;
@@ -86,9 +88,8 @@
 	host = get_canonical_hostname (options.verify_reverse_mapping);

 	if (sia_ses_init(&ent, saved_argc, saved_argv, host, user, tty, 0,
-	    NULL) != SIASUCCESS) {
+	    NULL) != SIASUCCESS)
 		fatal("sia_ses_init failed");
-	}

 	if ((pw = getpwnam(user)) == NULL) {
 		sia_ses_release(&ent);
@@ -100,25 +101,24 @@
 	}

 	ent->authtype = SIA_A_NONE;
-	if (sia_ses_estab(sia_collect_trm, ent) != SIASUCCESS) {
+	if (sia_ses_estab(sia_collect_trm, ent) != SIASUCCESS)
 		fatal("Couldn't establish session for %s from %s", user,
 		    host);
-	}

 	if (setpriority(PRIO_PROCESS, 0, 0) == -1) {
 		sia_ses_release(&ent);
 		fatal("setpriority: %s", strerror (errno));
 	}

-	if (sia_ses_launch(sia_collect_trm, ent) != SIASUCCESS) {
+	if (sia_ses_launch(sia_collect_trm, ent) != SIASUCCESS)
 		fatal("Couldn't launch session for %s from %s", user, host);
-	}
-
+
 	sia_ses_release(&ent);

-	if (setreuid(geteuid(), geteuid()) < 0) {
-		fatal("setreuid: %s", strerror(errno));
+	/* XXX: Should this be be around a if (!use_privsep) ? */
+	if (!use_privsep) {
+		if (setreuid(geteuid(), geteuid()) < 0)
+			fatal("setreuid: %s", strerror(errno));
 	}
 }
-
 #endif /* HAVE_OSF_SIA */
diff -ur openssh-3.4p1/auth-sia.h openssh-3.4p1+/auth-sia.h
--- openssh-3.4p1/auth-sia.h	Fri Apr 12 11:36:08 2002
+++ openssh-3.4p1+/auth-sia.h	Sun Jul 21 22:53:14 2002
@@ -27,6 +27,6 @@
 #ifdef HAVE_OSF_SIA

 int	auth_sia_password(Authctxt *authctxt, char *pass);
-void	session_setup_sia(char *user, char *tty);
+void	setup_sia(char *user, char *tty);

 #endif /* HAVE_OSF_SIA */
diff -ur openssh-3.4p1/monitor.c openssh-3.4p1+/monitor.c
--- openssh-3.4p1/monitor.c	Wed Jun 26 09:27:11 2002
+++ openssh-3.4p1+/monitor.c	Sun Jul 21 22:53:14 2002
@@ -120,6 +120,10 @@
 int mm_answer_pam_start(int, Buffer *);
 #endif

+#ifdef HAVE_OSF_SIA
+int mm_answer_setup_sia(int, Buffer *);
+#endif
+
 static Authctxt *authctxt;
 static BIGNUM *ssh1_challenge = NULL;	/* used for ssh1 rsa auth */

@@ -176,6 +180,9 @@
     {MONITOR_REQ_PTY, 0, mm_answer_pty},
     {MONITOR_REQ_PTYCLEANUP, 0, mm_answer_pty_cleanup},
     {MONITOR_REQ_TERM, 0, mm_answer_term},
+#ifdef HAVE_OSF_SIA
+    {MONITOR_REQ_SETUP_SIA, 0, mm_answer_setup_sia},
+#endif
     {0, 0, NULL}
 };

@@ -206,6 +213,9 @@
     {MONITOR_REQ_PTY, MON_ONCE, mm_answer_pty},
     {MONITOR_REQ_PTYCLEANUP, MON_ONCE, mm_answer_pty_cleanup},
     {MONITOR_REQ_TERM, 0, mm_answer_term},
+#ifdef HAVE_OSF_SIA
+    {MONITOR_REQ_SETUP_SIA, MON_ONCE, mm_answer_setup_sia},
+#endif
     {0, 0, NULL}
 };

@@ -307,10 +317,16 @@
 		monitor_permit(mon_dispatch, MONITOR_REQ_MODULI, 1);
 		monitor_permit(mon_dispatch, MONITOR_REQ_SIGN, 1);
 		monitor_permit(mon_dispatch, MONITOR_REQ_TERM, 1);
+#ifdef HAVE_OSF_SIA
+		monitor_permit(mon_dispatch, MONITOR_REQ_SETUP_SIA, 1);
+#endif

 	} else {
 		mon_dispatch = mon_dispatch_postauth15;
 		monitor_permit(mon_dispatch, MONITOR_REQ_TERM, 1);
+#ifdef HAVE_OSF_SIA
+		monitor_permit(mon_dispatch, MONITOR_REQ_SETUP_SIA, 1);
+#endif
 	}
 	if (!no_pty_flag) {
 		monitor_permit(mon_dispatch, MONITOR_REQ_PTY, 1);
@@ -717,6 +733,22 @@

 	return (authok != 0);
 }
+#endif
+
+#ifdef HAVE_OSF_SIA
+int
+mm_answer_setup_sia(int socket, Buffer *m)
+{
+	char *user, *tty;
+
+	user = buffer_get_string(m, NULL);
+	tty = buffer_get_string(m, NULL);
+
+	setup_sia(user, tty);
+
+	xfree(user);
+	xfree(tty);
+}
 #endif

 #ifdef USE_PAM
diff -ur openssh-3.4p1/monitor.h openssh-3.4p1+/monitor.h
--- openssh-3.4p1/monitor.h	Tue Jun 11 12:42:49 2002
+++ openssh-3.4p1+/monitor.h	Sun Jul 21 22:53:14 2002
@@ -50,6 +50,7 @@
 	MONITOR_REQ_RSACHALLENGE, MONITOR_ANS_RSACHALLENGE,
 	MONITOR_REQ_RSARESPONSE, MONITOR_ANS_RSARESPONSE,
 	MONITOR_REQ_PAM_START,
+	MONITOR_REQ_SETUP_SIA,
 	MONITOR_REQ_TERM
 };

diff -ur openssh-3.4p1/monitor_wrap.c openssh-3.4p1+/monitor_wrap.c
--- openssh-3.4p1/monitor_wrap.c	Thu Jun 20 20:43:43 2002
+++ openssh-3.4p1+/monitor_wrap.c	Sun Jul 21 22:53:14 2002
@@ -649,6 +649,24 @@
 	s->ttyfd = -1;
 }

+#ifdef HAVE_OSF_SIA
+void
+mm_setup_sia(char *name, char *tty)
+{
+	Buffer m;
+
+	debug3("%s entering", __func__);
+
+	buffer_init(&m);
+	buffer_put_cstring(&m, name);
+	buffer_put_cstring(&m, tty);
+
+	mm_request_send(pmonitor->m_recvfd, MONITOR_REQ_SETUP_SIA, &m);
+
+	buffer_free(&m);
+}
+#endif
+
 #ifdef USE_PAM
 void
 mm_start_pam(char *user)
diff -ur openssh-3.4p1/monitor_wrap.h openssh-3.4p1+/monitor_wrap.h
--- openssh-3.4p1/monitor_wrap.h	Sun May 12 21:07:42 2002
+++ openssh-3.4p1+/monitor_wrap.h	Sun Jul 21 22:53:14 2002
@@ -59,6 +59,10 @@
 void mm_start_pam(char *);
 #endif

+#ifdef HAVE_OSF_SIA
+void mm_setup_sia(char *, char *);
+#endif
+
 void mm_terminate(void);
 int mm_pty_allocate(int *, int *, char *, int);
 void mm_session_pty_cleanup2(void *);
diff -ur openssh-3.4p1/session.c openssh-3.4p1+/session.c
--- openssh-3.4p1/session.c	Wed Jun 26 09:51:06 2002
+++ openssh-3.4p1+/session.c	Sun Jul 21 23:26:49 2002
@@ -57,6 +57,7 @@
 #include "canohost.h"
 #include "session.h"
 #include "monitor_wrap.h"
+#include "auth-sia.h"

 #ifdef HAVE_CYGWIN
 #include <windows.h>
@@ -1269,7 +1270,7 @@
 	 */
 	if (!options.use_login) {
 #ifdef HAVE_OSF_SIA
-		session_setup_sia(pw->pw_name, s->ttyfd == -1 ? NULL : s->tty);
+		PRIVSEP(setup_sia(pw->pw_name, s->ttyfd == -1 ? NULL : s->tty));
 		if (!check_quietlogin(s, command))
 			do_motd();
 #else /* HAVE_OSF_SIA */

SCHINCKE, KEITH (JSC-SM) (LM)

2002-Aug-12 15:50 UTC

head link

OSF/1 or Tru64 patch for Privsep

Hey Ben,

I pulled the latest snap shot (20020801) and got your patch to apply with
only one chunk rejected. 

I have been able to login with the new code using the OpenSSH client,
ssh.com 3.1.0 and putty 0.52.

"It works for me."

Any other data points you are interested in?

Thanks for the work you put into this.

Keith

-----Original Message-----
From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
Sent: Sunday, August 11, 2002 5:38 PM
To: OpenSSH Development
Subject: OSF/1 or Tru64 patch for Privsep



Either this never made it to the list or no one cares about Tru64.  This
is the last time I'll send this patch to the list.  If no one steps up and
finishes it or provides me with enough information to fix any remaining
bugs (one being complaint that 'ssh site cmd' does not work right).

If there is no activity on this for a week.  I'll post it to bugzilla and
will ignore any request for me to fix Tru64.

I do thank everyone that has privately contributed help in regards to
testing and such.

This patch may apply to 3.4p1, but I'd perfer testing to be done on -cvs.

- Ben

SCHINCKE, KEITH (JSC-SM) (LM)

2002-Aug-12 18:26 UTC

head link

OSF/1 or Tru64 patch for Privsep

HAVE_OSF_SIA yes
X11 forward yes
remote command execution fails with "buffer_put_cstring: s == NULL"

I am looking into where things failing at.

-----Original Message-----
From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
Sent: Monday, August 12, 2002 11:08 AM
To: SCHINCKE, KEITH (JSC-SM) (LM)
Cc: OpenSSH Development
Subject: RE: OSF/1 or Tru64 patch for Privsep



try everything... X11 forward, port forwarding, executing remote commands.

Everything needs to be tested.. Also check to ensure HAVE_OSF_SIA is set
in your config.h.

- Ben

On Mon, 12 Aug 2002, SCHINCKE, KEITH (JSC-SM) (LM) wrote:
> Hey Ben,
>
> I pulled the latest snap shot (20020801) and got your patch to apply with
> only one chunk rejected.
>
> I have been able to login with the new code using the OpenSSH client,
> ssh.com 3.1.0 and putty 0.52.
>
> "It works for me."
>
> Any other data points you are interested in?
>
> Thanks for the work you put into this.
>
> Keith
>
> -----Original Message-----
> From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
> Sent: Sunday, August 11, 2002 5:38 PM
> To: OpenSSH Development
> Subject: OSF/1 or Tru64 patch for Privsep
>
>
>
> Either this never made it to the list or no one cares about Tru64.  This
> is the last time I'll send this patch to the list.  If no one steps up
and
> finishes it or provides me with enough information to fix any remaining
> bugs (one being complaint that 'ssh site cmd' does not work right).
>
> If there is no activity on this for a week.  I'll post it to bugzilla
and
> will ignore any request for me to fix Tru64.
>
> I do thank everyone that has privately contributed help in regards to
> testing and such.
>
> This patch may apply to 3.4p1, but I'd perfer testing to be done on
-cvs.
>
> - Ben
>

SCHINCKE, KEITH (JSC-SM) (LM)

2002-Aug-12 19:08 UTC

head link

OSF/1 or Tru64 patch for Privsep

Remote execution works with privsep turned off.

Ok, I have traced some things down to where the program stops. 

session.c:do_child() calls the privsep wrapper for setup_sia (mm_setup_sia
is actuall called)
	s->ttyfd is -1 so NULL passed in as the tty.
monitor_wrap.c:mm_setup_sia calls buffer_put_cstring(&m, tty)
bufaux.c:buffer_put_cstring calls fatal() when s is NULL. 


I am still new to looking at the code but why not do
	'buffer_put_string(&m,tty,strlen(tty));'
in monitor_wrap.c:mm_setup_sia?


-----Original Message-----
From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
Sent: Monday, August 12, 2002 1:31 PM
To: SCHINCKE, KEITH (JSC-SM) (LM)
Cc: OpenSSH Development
Subject: RE: OSF/1 or Tru64 patch for Privsep



Every also test w/ and w/out Privsep so we don't break the non-privsep
case. (That would be a *REAL* shame if we did. <weak smile>)

- Ben


On Mon, 12 Aug 2002, SCHINCKE, KEITH (JSC-SM) (LM) wrote:
> HAVE_OSF_SIA yes
> X11 forward yes
> remote command execution fails with "buffer_put_cstring: s ==
NULL"
>
> I am looking into where things failing at.
>
> -----Original Message-----
> From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
> Sent: Monday, August 12, 2002 11:08 AM
> To: SCHINCKE, KEITH (JSC-SM) (LM)
> Cc: OpenSSH Development
> Subject: RE: OSF/1 or Tru64 patch for Privsep
>
>
>
> try everything... X11 forward, port forwarding, executing remote commands.
>
> Everything needs to be tested.. Also check to ensure HAVE_OSF_SIA is set
> in your config.h.
>
> - Ben
>
> On Mon, 12 Aug 2002, SCHINCKE, KEITH (JSC-SM) (LM) wrote:
>
> > Hey Ben,
> >
> > I pulled the latest snap shot (20020801) and got your patch to apply
with> > only one chunk rejected.
> >
> > I have been able to login with the new code using the OpenSSH client,
> > ssh.com 3.1.0 and putty 0.52.
> >
> > "It works for me."
> >
> > Any other data points you are interested in?
> >
> > Thanks for the work you put into this.
> >
> > Keith
> >
> > -----Original Message-----
> > From: Ben Lindstrom [mailto:mouring at etoh.eviladmin.org]
> > Sent: Sunday, August 11, 2002 5:38 PM
> > To: OpenSSH Development
> > Subject: OSF/1 or Tru64 patch for Privsep
> >
> >
> >
> > Either this never made it to the list or no one cares about Tru64. 
This
> > is the last time I'll send this patch to the list.  If no one
steps up
and> > finishes it or provides me with enough information to fix any
remaining
> > bugs (one being complaint that 'ssh site cmd' does not work
right).
> >
> > If there is no activity on this for a week.  I'll post it to
bugzilla
and> > will ignore any request for me to fix Tru64.
> >
> > I do thank everyone that has privately contributed help in regards to
> > testing and such.
> >
> > This patch may apply to 3.4p1, but I'd perfer testing to be done
on
-cvs.> >
> > - Ben
> >
>

Damien Miller

2002-Aug-12 22:54 UTC

head link

OSF/1 or Tru64 patch for Privsep

On Tue, 2002-08-13 at 01:50, SCHINCKE, KEITH (JSC-SM) (LM)
wrote:> Hey Ben,
> 
> I pulled the latest snap shot (20020801) and got your patch to apply with
> only one chunk rejected. 
> 
> I have been able to login with the new code using the OpenSSH client,
> ssh.com 3.1.0 and putty 0.52.
> 
> "It works for me."
> 
> Any other data points you are interested in?
If you could run the regression tests in regress/ You do need Berkeley
make though.

-d

Apparently Analagous Threads

Search for more reasonably related threads

openssh unix dev - Aug 2002 - OSF/1 or Tru64 patch for Privsep

OSF/1 or Tru64 patch for Privsep

OSF/1 or Tru64 patch for Privsep

OSF/1 or Tru64 patch for Privsep

OSF/1 or Tru64 patch for Privsep

OSF/1 or Tru64 patch for Privsep

Apparently Analagous Threads