NetBSD Problem Report #23245

Received: (qmail 11410 invoked by uid 605); 23 Oct 2003 12:31:09 -0000
Message-Id: <200310231231.h9NCV17G003064@heiligenberg.nt.e-technik.tu-darmstadt.de>
Date: Thu, 23 Oct 2003 14:31:01 +0200 (CEST)
From: Hauke Fath <hf@spg.tu-darmstadt.de>
Sender: gnats-bugs-owner@NetBSD.org
Reply-To: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: gnats-bugs@gnats.netbsd.org
Cc: Hauke Fath <hf@spg.tu-darmstadt.de>
Subject: security/stunnel does not terminate
X-Send-Pr-Version: 3.95

>Number:         23245
>Category:       pkg
>Synopsis:       security/stunnel does not terminate properly.
>Confidential:   no
>Severity:       serious
>Priority:       low
>Responsible:    hauke
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Thu Oct 23 12:32:00 +0000 2003
>Closed-Date:    
>Last-Modified:  Mon Oct 19 18:42:35 +0000 2020
>Originator:     Hauke Fath <hf@spg.tu-darmstadt.de>
>Release:        NetBSD 1.6ZC
>Organization:
-- 
           Hauke Fath                   /~\  The ASCII Ribbon Campaign
Institut für Nachrichtentechnik         \ /    No HTML/RTF in email
          TU Darmstadt                   X     No Word docs in email
      Ruf +49-6151-16-3281              / \  Respect for open standards
>Environment:


System: NetBSD heiligenberg 1.6ZC NetBSD 1.6ZC (HEILIGENBERG) #1: Fri Sep 26 16:51:05 CEST 2003 hf@heiligenberg:/var/obj/netbsd-builds/i386/obj/sys/arch/i386/compile/HEILIGENBERG i386
Architecture: i386
Machine: i386

I see the same thing on a NetBSD/sparc 1.6.1 machine.

[hf@heiligenberg] ~ # stunnel -version
stunnel 4.04 on i386--netbsdelf FORK+LIBWRAP with OpenSSL 0.9.7b 10 Apr 2003

Global options
cert            = /etc/stunnel/stunnel.pem
ciphers         = ALL:!ADH:+RC4:@STRENGTH
debug           = 5
key             = /etc/stunnel/stunnel.pem
pid             = /usr/pkg/var/run/stunnel.pid
RNDbytes        = 64
RNDfile         = /dev/urandom
RNDoverwrite    = yes
session         = 300 seconds
verify          = none

Service-level options
TIMEOUTbusy     = 300 seconds
TIMEOUTclose    = 60 seconds
TIMEOUTidle     = 43200 seconds


[hf@heiligenberg] ~ # cat /etc/uucp/stunnel.conf 
# $Id: stunnel.conf,v 1.3 2003/09/08 21:16:52 hauke Exp hf $
#
# stunnel setup for uucp client

debug = mail.info

client = yes
pid = 

[uucico]
connect = uucp.XXXXX.YYY:940
exec = /usr/libexec/uucp/uucico
execargs = uucico --nodetach --nouuxqt --debug abnormal -S uucp


>Description:
	After updating stunnel 3.xx to version 4, stunnel processes
	keep hanging around indefinitely after the uucico job is done,
	both in the background and in the foreground (option
	'foreground = yes'). With two uucp connects per hour, this
	gets annoying quickly.

	This did not happen with stunnel 3.xx. Since the stunnel group
	mis-designed other things (removing the cli options, for one),
	can we have an stunnel3 package?

>How-To-Repeat:
	Do an 'stunnel /etc/uucp/stunnel.conf', see the job does not
	terminate when in the foreground, and processes keep hanging
	around when started as a daemon.

	Attach gdb:

(gdb) bt
#0  0x481c454b in select () from /usr/lib/libc.so.12
#1  0x8051032 in sselect (n=4, readfds=0xbfbff344, writefds=0x0, 
    exceptfds=0x0, timeout=0x0) at sselect.c:85
#2  0x8052bc0 in daemon_loop () at stunnel.c:195
#3  0x80527b9 in main_execute () at stunnel.c:105
#4  0x80526db in main (argc=2, argv=0xbfbff424) at stunnel.c:72
#5  0x804a594 in ___start ()
(gdb) up
#1  0x8051032 in sselect (n=4, readfds=0xbfbff344, writefds=0x0, 
    exceptfds=0x0, timeout=0x0) at sselect.c:85
85                  retval=select(n, readfds, writefds, exceptfds, NULL);
(gdb) info frame
Stack level 1, frame at 0xbfbff2e4:
 eip = 0x8051032 in sselect (sselect.c:85); saved eip 0x8052bc0
 called by frame at 0xbfbff394, caller of frame at 0xbfbff2e4
 source language c.
 Arglist at 0xbfbff2e4, args: n=4, readfds=0xbfbff344, writefds=0x0, 
    exceptfds=0x0, timeout=0x0
 Locals at 0xbfbff2e4, Previous frame's sp is 0x0
 Saved registers:
  ebx at 0xbfbff2cc, ebp at 0xbfbff2e4, eip at 0xbfbff2e8
(gdb)

>Fix:
	Set up a cron job that kills the rogue stunnel processes, or
	roll back to stunnel 3.

	Or, fix stunnel.

>Release-Note:
>Audit-Trail:
Responsible-Changed-From-To: pkg-manager->martti 
Responsible-Changed-By: wiz 
Responsible-Changed-When: Tue Jul 6 10:48:41 UTC 2004 
Responsible-Changed-Why:  
over to maintainer 
State-Changed-From-To: open->feedback
State-Changed-By: reed@netbsd.org
State-Changed-When: Sat, 15 Oct 2005 02:50:02 +0000
State-Changed-Why:
Packaged updated a few times since then. Asked if problem still exists.


From: "Jeremy C. Reed" <reed@reedmedia.net>
To: Hauke Fath <hf@spg.tu-darmstadt.de>
Cc: gnats-bugs@netbsd.org
Subject: Re: pkg/23245
Date: Fri, 14 Oct 2005 19:49:38 -0700 (PDT)

 I am looking at old PR 23245.

   http://www.NetBSD.org/cgi-bin/query-pr-single.pl?number=23245

 I didn't check if this is a problem anymore myself.

 I saw your test platform was 1.6ZC and 1.6.1 which is now quite old.

 And stunnel has been updated a few times since then.

 Does this problem still exist?


   Jeremy C. Reed

   	  	 	 technical support & remote administration
  	  	 	 http://www.pugetsoundtechnology.com/

From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: reed@netbsd.org
Cc: martti@netbsd.org, pkgsrc-bugs@netbsd.org,
	gnats-admin@netbsd.org, reed@netbsd.org,
	Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
Subject: Re: pkg/23245
Date: Tue, 18 Oct 2005 00:31:36 +0200

 At 2:50 Uhr +0000 15.10.2005, reed@netbsd.org wrote:
 >Packaged updated a few times since then. Asked if problem still exists.

 I switched to wip/stunnel3 long ago because of the issue, but the stunnel 4
 config should still be there. I'll have a look at it.

 	hauke

 --
 "It's never straight up and down"     (DEVO)

From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: reed@netbsd.org
Cc: martti@netbsd.org, gnats-bugs@netbsd.org, gnats-admin@netbsd.org,
	Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
Subject: Re: pkg/23245
Date: Sun, 15 Jan 2006 23:06:52 +0100

 At 2:50 Uhr +0000 15.10.2005, reed@netbsd.org wrote:
 >State-Changed-From-To: open->feedback
 >State-Changed-By: reed@netbsd.org
 >State-Changed-When: Sat, 15 Oct 2005 02:50:02 +0000
 >State-Changed-Why:
 >Packaged updated a few times since then. Asked if problem still exists.

 Yes, it does.


 [hauke@pizza] /etc/uucp # uname -a
 NetBSD pizza.causeuse.org 3.0_STABLE NetBSD 3.0_STABLE (PIZZA) #1: Tue Jan
 3 22:52:41 CET 2006
 hauke@pizza.causeuse.org:/var/obj/netbsd-builds/netbsd-3/sparc/sys/arch/sparc/co
 mpile/PIZZA sparc
 [hauke@pizza] /etc/uucp # pkg_info | grep stunnel
 stunnel-4.07nb2     Universal SSL tunnel
 [hauke@pizza] /etc/uucp # /usr/pkg/sbin/stunnel -version
 stunnel 4.07 on sparc--netbsdelf FORK+POLL+IPv6+LIBWRAP with OpenSSL 0.9.7d
 17 Mar 2004

 Global options
 cert            = /etc/stunnel/stunnel.pem
 ciphers         = ALL:!ADH:+RC4:@STRENGTH
 debug           = 5
 key             = /etc/stunnel/stunnel.pem
 pid             = /var/run/stunnel.pid
 RNDbytes        = 64
 RNDfile         = /dev/urandom
 RNDoverwrite    = yes
 session         = 300 seconds
 verify          = none

 Service-level options
 TIMEOUTbusy     = 300 seconds
 TIMEOUTclose    = 60 seconds
 TIMEOUTconnect  = 10 seconds
 TIMEOUTidle     = 43200 seconds
 [hauke@pizza] /etc/uucp # ps ax | grep st\\unnel
 27190 ?     Is    0:00.03 /usr/pkg/sbin/stunnel /etc/uucp/stunnel.conf
 [hauke@pizza] /etc/uucp # gdb /usr/pkg/sbin/stunnel 27190

 [...]

 (gdb) bt
 #0  0x4026275c in poll () from /usr/lib/libc.so.12
 #1  0x0001c854 in s_poll_wait (fds=0xefffe208, timeout=-1) at network.c:170
 #2  0x000202b4 in daemon_loop () at stunnel.c:187
 #3  0x0001fe10 in main_execute () at stunnel.c:106
 #4  0x0001fca4 in main (argc=2, argv=0xefffe5d4) at stunnel.c:72
 #5  0x0001287c in ___start ()
 (gdb)

 	hauke

 --
 "It's never straight up and down"     (DEVO)


Responsible-Changed-From-To: martti->pkg-manager
Responsible-Changed-By: reed@netbsd.org
Responsible-Changed-When: Sat, 08 Apr 2006 05:58:30 +0000
Responsible-Changed-Why:
Changed because is no longer the maintainer and
said can't test anymore.


From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: gnats-bugs@netbsd.org
Cc: pkg-manager@netbsd.org
Subject: Re: pkg/23245
Date: Sat, 21 Oct 2006 01:32:47 +0200

 Things got worse.

 I have tried both stunnel 4.15 (the current pkgsrc version) and 4.18, the
 latter built with and without pthreads support.

 [hauke@pizza] /etc/uucp # stunnel -version
 stunnel 4.18 on sparc--netbsdelf with OpenSSL 0.9.7d 17 Mar 2004
 Threading:FORK SSL:ENGINE Sockets:POLL,IPv6 Auth:LIBWRAP

 Global options
 debug           = 5
 pid             = /var/run/stunnel/stunnel.pid
 RNDbytes        = 64
 RNDfile         = /dev/urandom
 RNDoverwrite    = yes

 Service-level options
 cert            = /etc/pkg/stunnel/stunnel.pem
 ciphers         = ALL:!ADH:+RC4:@STRENGTH
 key             = /etc/pkg/stunnel/stunnel.pem
 session         = 300 seconds
 sslVersion      = SSLv3 for client, all for server
 TIMEOUTbusy     = 300 seconds
 TIMEOUTclose    = 60 seconds
 TIMEOUTconnect  = 10 seconds
 TIMEOUTidle     = 43200 seconds
 verify          = none

 With the config file

 [hauke@pizza] /etc/uucp # cat stunnel.conf
 # $Id: stunnel.conf,v 1.3 2003/09/08 21:16:52 hauke Exp hauke $
 #
 # stunnel setup for uucp client

 debug = mail.debug
 foreground = yes
 client = yes
 pid =

 [uucico]
 connect = uucp.xxxxx.yyy:940
 exec = /usr/libexec/uucp/uucico
 execargs = uucico --nodetach --debug abnormal -S uucp


 all the builds _loop_, i.e. call uucico again and again after it has
 completed the data transfer.

 The tunnel is set up correctly, but stunnel mis-manages the local client.
 In addition, when "foreground = no" is set in the above configuration,
 non-terminating stunnel processes end up hanging in the background as
 before.

 I can reproduce the problem, and provide a debug log and ktrace if needed.

 The pkg update 4.15 -> 4.18 is here:
 http://la.causeuse.org/hauke/NetBSD/pkgsrc/pr23245.shar

 	hauke


 --
 "It's never straight up and down"     (DEVO)


State-Changed-From-To: feedback->open
State-Changed-By: minskim@netbsd.org
State-Changed-When: Thu, 21 Jun 2007 12:36:27 +0000
State-Changed-Why:
Feedback received.


State-Changed-From-To: open->feedback
State-Changed-By: asau@NetBSD.org
State-Changed-When: Mon, 02 Nov 2009 13:57:48 +0000
State-Changed-Why:
Does the problem persist with stunnel-4.27 (current) and supported release?


From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: gnats-bugs@NetBSD.org
Cc: pkg-manager@NetBSD.org, gnats-admin@NetBSD.org, asau@NetBSD.org
Subject: Re: pkg/23245 (security/stunnel does not terminate properly.)
Date: Sun, 4 Apr 2010 16:00:45 +0200

 At 13:57 Uhr +0000 2.11.2009, asau@NetBSD.org wrote:
 >Synopsis: security/stunnel does not terminate properly.
 >
 >Does the problem persist with stunnel-4.27 (current) and supported release?

 Sorry for the late reply. I just checked with stunnel 4.32 (easy pkg
 update, just adjust checksum)

 # uname -mrs
 NetBSD 4.0_STABLE sparc
 # stunnel -version
 stunnel 4.32 on sparc--netbsdelf with OpenSSL 0.9.8e 23 Feb 2007
 Threading:PTHREAD SSL:ENGINE Sockets:POLL,IPv6 Auth:LIBWRAP

 Global options
 debug           = daemon.notice
 pid             = /var/run/stunnel.pid
 RNDbytes        = 64
 RNDfile         = /dev/urandom
 RNDoverwrite    = yes

 Service-level options
 cert            = /etc/pkg/stunnel/stunnel.pem
 ciphers         = ALL:!ADH:+RC4:@STRENGTH
 session         = 300 seconds
 stack           = 65536 bytes
 sslVersion      = SSLv3 for client, all for server
 TIMEOUTbusy     = 300 seconds
 TIMEOUTclose    = 60 seconds
 TIMEOUTconnect  = 10 seconds
 TIMEOUTidle     = 43200 seconds
 verify          = none
 #

 and things are worse than ever. Using the following stunnel config file


 # $Id: stunnel.conf,v 1.4 2006/12/08 23:11:39 hauke Exp hauke $
 #
 # stunnel setup for uucp client

 syslog = no
 debug = mail.debug
 output = /var/log/stunnel.log

 foreground = no

 client = yes
 pid =

 [uucico]
 connect = uucp.rninet.net:940
 exec = /usr/pkg/sbin/uucico
 execargs = uucico --nodetach --debug abnormal -S uucp

 pty = yes


 with "foreground = yes" set, uucico is executed, then stunnel hangs there,
 not terminating. With "foreground = no", the log has


 2010.04.04 15:40:39 LOG5[9468:4018143232]: Reading configuration from file
 /etc/pkg/uucp/stunnel.conf
 2010.04.04 15:40:39 LOG7[9468:4018143232]: Snagged 64 random bytes from
 /dev/urandom
 2010.04.04 15:40:39 LOG7[9468:4018143232]: RAND_status claims sufficient
 entropy for the PRNG
 2010.04.04 15:40:39 LOG7[9468:4018143232]: PRNG seeded successfully
 2010.04.04 15:40:39 LOG7[9468:4018143232]: SSL context initialized for
 service uucico
 2010.04.04 15:40:39 LOG5[9468:4018143232]: Configuration successful
 2010.04.04 15:40:39 LOG5[9468:4018143232]: No limit detected for the number
 of clients
 2010.04.04 15:40:39 LOG7[9468:4018143232]: FD=8 in non-blocking mode
 2010.04.04 15:40:39 LOG7[9468:4018143232]: FD=9 in non-blocking mode
 2010.04.04 15:40:39 LOG7[8731:4018143232]: No pid file being created
 2010.04.04 15:40:39 LOG5[8731:4018143232]: stunnel 4.32 on sparc--netbsdelf
 with OpenSSL 0.9.8e 23 Feb 2007
 2010.04.04 15:40:39 LOG5[8731:4018143232]: Threading:PTHREAD SSL:ENGINE
 Sockets:POLL,IPv6 Auth:LIBWRAP


 -- stunnel just hangs there, uucico is not executed.

 I'll take the issue to the stunnel-users list, and see what comes of it. As
 usual, wip/stunnel3 just works.

 	hauke

 --
 "It's never straight up and down"     (DEVO)


From: Hauke Fath <hauke@Espresso.Rhein-Neckar.DE>
To: gnats-bugs@NetBSD.org
Cc: pkg-manager@NetBSD.org, gnats-admin@NetBSD.org
Subject: Re: pkg/23245 (security/stunnel does not terminate properly.)
Date: Sun, 4 Apr 2010 17:12:05 +0200

 At 14:20 Uhr +0000 4.4.2010, Hauke Fath wrote:
 > At 13:57 Uhr +0000 2.11.2009, asau@NetBSD.org wrote:
 > >Synopsis: security/stunnel does not terminate properly.
 > >
 > >Does the problem persist with stunnel-4.27 (current) and supported release?
 >
 > Sorry for the late reply. I just checked with stunnel 4.32

 Building stunnel with

 # See <http://mail-index.netbsd.org/pkgsrc-users/2008/06/03/msg007314.html>
 # pthreads have issues on sparc
 PKG_OPTIONS.stunnel +=  -threads

 makes the "foreground = no" case work "as before", i.e. the stunnel process
 runs uucico successfully, but then stays around forever after uucico has
 terminated.

 	hauke


 --
 "It's never straight up and down"     (DEVO)


State-Changed-From-To: feedback->open
State-Changed-By: hauke@NetBSD.org
State-Changed-When: Sun, 04 Apr 2010 15:54:10 +0000
State-Changed-Why:
I provided feedback. The problems with the package still exist 
on current releases.


State-Changed-From-To: open->feedback
State-Changed-By: maya@NetBSD.org
State-Changed-When: Sun, 28 May 2017 03:12:53 +0000
State-Changed-Why:
Requesting a re-test after PR kern/47569 (SOCK_NONBLOCK doesn't work)


From: coypu@sdf.org
To: gnats-bugs@netbsd.org
Cc: 
Subject: Re: pkg/23245: security/stunnel does not terminate properly.
Date: Sun, 28 May 2017 03:17:02 +0000

 Sorry for the excess feedback requests with no changes.
 PR 47569 seems very relevant for this and is fixed.
 not having non-block is also relevant-sounding.

 Thanks for your understanding

Responsible-Changed-From-To: pkg-manager->hauke
Responsible-Changed-By: bsiegert@NetBSD.org
Responsible-Changed-When: Mon, 19 Oct 2020 18:42:35 +0000
Responsible-Changed-Why:
Hauke is now a developer.


State-Changed-From-To: feedback->open
State-Changed-By: bsiegert@NetBSD.org
State-Changed-When: Mon, 19 Oct 2020 18:42:35 +0000
State-Changed-Why:
feedback timeout


>Unformatted:

Home	PR Database Search