- stunnel-users - stunnel.org

Vasudha Upadhyaya is out of the office.
by vasudha.upadhyaya＠convergys.com 16 Nov '05

16 Nov '05

I will be out of the office starting 11/16/2005 and will not return until 12/05/2005. I will respond to your message when I return. -- "NOTICE: The information contained in this electronic mail transmission is intended by Convergys Corporation for the use of the named individual or entity to which it is directed and may contain information that is privileged or otherwise confidential. If you have received this electronic mail transmission in error, please delete it from your system without copying or forwarding it, and notify the sender of the error by reply email or by telephone (collect), so that the sender's address records can be corrected."

1 0

more stunnel problems
by Louis Young 10 Nov '05

10 Nov '05

Hi guys Ok, as I've said in a previous mail, I'm now merely trying to get the linux client to work. In the config file: [ssmtp] accept = 465 connect = 25 But now: louislaptop:/usr/local/etc/stunnel # telnet localhost 25 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. 220 louislaptop.kpho.local ESMTP Postfix Works fine. I get the SMTP prompt. And now: louislaptop:/usr/local/etc/stunnel # telnet localhost 465 Trying 127.0.0.1... Connected to localhost. Escape character is '^]'. No SMTP prompt. In the log files: 2005.11.09 16:32:47 LOG5[32348:1]: stunnel 4.14 on i686-pc-linux-gnu UCONTEXT+POLL+IPv4 with OpenSSL 0.9.7g 11 Apr 2005 2005.11.09 16:32:47 LOG7[32348:1]: Snagged 64 random bytes from /root/.rnd 2005.11.09 16:32:47 LOG7[32348:1]: Wrote 1024 new random bytes to /root/.rnd 2005.11.09 16:32:47 LOG7[32348:1]: RAND_status claims sufficient entropy for the PRNG 2005.11.09 16:32:47 LOG6[32348:1]: PRNG seeded successfully 2005.11.09 16:32:47 LOG7[32348:1]: Certificate: /usr/local/etc/stunnel/stunnel.pem 2005.11.09 16:32:47 LOG7[32348:1]: Key file: /usr/local/etc/stunnel/stunnel.pem 2005.11.09 16:32:47 LOG6[32348:1]: file ulimit = 1024 (can be changed with 'ulimit -n') 2005.11.09 16:32:47 LOG6[32348:1]: poll() used - no FD_SETSIZE limit for file descriptors 2005.11.09 16:32:47 LOG5[32348:1]: 500 clients allowed 2005.11.09 16:32:47 LOG7[32348:1]: FD 4 in non-blocking mode 2005.11.09 16:32:47 LOG7[32348:1]: FD 5 in non-blocking mode 2005.11.09 16:32:47 LOG7[32348:1]: FD 6 in non-blocking mode 2005.11.09 16:32:47 LOG7[32348:1]: SO_REUSEADDR option set on accept socket 2005.11.09 16:32:47 LOG7[32348:1]: ssmtp bound to 0.0.0.0:7777 2005.11.09 16:32:47 LOG7[32349:1]: Created pid file /stunnel.pid 2005.11.09 16:32:47 LOG7[32349:0]: Waiting -1 second(s) for 2 file descriptor(s) 2005.11.09 16:32:53 LOG7[32349:0]: CONTEXT 1, FD=4, (IN)->() 2005.11.09 16:32:53 LOG7[32349:0]: CONTEXT 1, FD=6, (IN)->(IN) 2005.11.09 16:32:53 LOG7[32349:1]: ssmtp accepted FD=7 from 127.0.0.1:15767 2005.11.09 16:32:53 LOG7[32349:1]: Creating a new context 2005.11.09 16:32:53 LOG7[32349:1]: Context 2 created 2005.11.09 16:32:53 LOG7[32349:2]: Context swap: 1 -> 2 2005.11.09 16:32:53 LOG7[32349:2]: ssmtp started 2005.11.09 16:32:53 LOG7[32349:2]: FD 7 in non-blocking mode 2005.11.09 16:32:53 LOG7[32349:2]: TCP_NODELAY option set on local socket 2005.11.09 16:32:53 LOG5[32349:2]: ssmtp connected from 127.0.0.1:15767 2005.11.09 16:32:53 LOG7[32349:2]: SSL state (accept): before/accept initialization 2005.11.09 16:32:53 LOG7[32349:0]: Waiting 300 second(s) for 3 file descriptor(s) 2005.11.09 16:37:53 LOG7[32349:0]: CONTEXT 1, FD=4, (IN)->() 2005.11.09 16:37:53 LOG7[32349:0]: CONTEXT 1, FD=6, (IN)->() 2005.11.09 16:37:53 LOG7[32349:0]: CONTEXT 2, FD=7, (IN)->() 2005.11.09 16:37:53 LOG6[32349:2]: init_ssl: s_poll_wait timeout 2005.11.09 16:37:53 LOG7[32349:2]: ssmtp finished (0 left) 2005.11.09 16:37:53 LOG3[32349:2]: STACK_RESERVE is too high 2005.11.09 16:37:53 LOG7[32349:2]: Context 2 closed 2005.11.09 16:37:53 LOG7[32349:0]: Waiting -1 second(s) for 2 file descriptor(s) 2005.11.09 16:58:51 LOG5[32349:0]: Received signal 15; terminating 2005.11.09 16:58:51 LOG7[32349:0]: removing pid file /stunnel.pid Any ideas? -- Regards Louis Young Systems Administrator Kwikpay SA (Pty) Ltd

2 3

Re: [stunnel-users] Solaris: new_context, makecontext
by Hans Werner Strube 09 Nov '05

09 Nov '05

Hans Werner Strube wrote: > 1. Line 92 of sthreads.c, > (char *)ctx->ctx.uc_stack.ss_sp+=STACK_SIZE-8; > causes a compiler warning: "a cast does not yield an lvalue." > > 2. The Solaris <9 SPARC peculiarity concerning the third argument of > makecontext might be taken into account. > > Both would be fixed by the following changes. > ***Caveat: I do not know how the third argument of makecontext behaves > in Solaris x86 for versions <9 vs. >=9. Have no machine to test this. Sorry, my patch contained a bad error: ctx->stack+=STACK_SIZE-8 is, of course, nonsense, since ctx->stack must not be changed. My patch to version 4.15b1 should read instead: *** sthreads.c.15b1 Wed Nov 9 12:18:18 2005 --- sthreads.c Wed Nov 9 17:07:57 2005 *************** *** 78,85 **** return NULL; } ctx->ctx.uc_link=NULL; /* it should never happen */ ! ctx->ctx.uc_stack.ss_sp=ctx->stack; ! #if defined(__sgi) || (defined(CPU_SPARC) && ( \ defined(OS_SOLARIS2_0) || \ defined(OS_SOLARIS2_1) || \ defined(OS_SOLARIS2_2) || \ --- 78,84 ---- return NULL; } ctx->ctx.uc_link=NULL; /* it should never happen */ ! #if defined(CPU_SPARC) && ( \ defined(OS_SOLARIS2_0) || \ defined(OS_SOLARIS2_1) || \ defined(OS_SOLARIS2_2) || \ *************** *** 88,96 **** defined(OS_SOLARIS2_5) || \ defined(OS_SOLARIS2_6) || \ defined(OS_SOLARIS2_7) || \ ! defined(OS_SOLARIS2_8))) ! (char *)ctx->ctx.uc_stack.ss_sp+=STACK_SIZE-8; #endif ctx->ctx.uc_stack.ss_size=STACK_SIZE; ctx->ctx.uc_stack.ss_flags=0; --- 87,102 ---- defined(OS_SOLARIS2_5) || \ defined(OS_SOLARIS2_6) || \ defined(OS_SOLARIS2_7) || \ ! defined(OS_SOLARIS2_8)) ! #define MCTXA 2 ! #else ! #define MCTXA 1 #endif + #if defined(__sgi) || (MCTXA == 2) + ctx->ctx.uc_stack.ss_sp=ctx->stack+STACK_SIZE-8; + #else + ctx->ctx.uc_stack.ss_sp=ctx->stack; + #endif ctx->ctx.uc_stack.ss_size=STACK_SIZE; ctx->ctx.uc_stack.ss_flags=0; *************** *** 121,127 **** if(!ctx) return -1; s_log(LOG_DEBUG, "Context %ld created", ctx->id); ! makecontext(&ctx->ctx, (void(*)(void))cli, 1, arg); return 0; } --- 127,133 ---- if(!ctx) return -1; s_log(LOG_DEBUG, "Context %ld created", ctx->id); ! makecontext(&ctx->ctx, (void(*)(void))cli, MCTXA, arg); return 0; }

1 0

Trouble between linux client and windows server
by Louis Young 09 Nov '05

09 Nov '05

Hi guys I'm trying to use stunnel between a windows server and a linux client. On the windows server: [VNC] accept = 5901 connect = 5900 and on the linux client: ; VNC server on 192.168.100.57 accept = 5901 connect = 192.168.100.65:5901 But, alas: Trying 127.0.0.1... telnet: connect to address 127.0.0.1: Connection refused Trying ::1... telnet: connect to address ::1: Connection refused It's almost as if stunnel isn't listening on that port on the linux client. Any ideas? -- Regards Louis Young Systems Administrator Kwikpay SA (Pty) Ltd

1 1

Re: [stunnel-users] Solaris: new_context, makecontext
by Hans Werner Strube 09 Nov '05

09 Nov '05

Michal Trojnara wrote: > Please try: ftp://stunnel.mirt.net/stunnel/stunnel-4.15b1.tar.gz The changes to common.h and sthreads.c with respect to new_context and makecontext are reasonable. However, there are two problems: 1. Line 92 of sthreads.c, (char *)ctx->ctx.uc_stack.ss_sp+=STACK_SIZE-8; causes a compiler warning: "a cast does not yield an lvalue." 2. The Solaris <9 SPARC peculiarity concerning the third argument of makecontext might be taken into account. Both would be fixed by the following changes. ***Caveat: I do not know how the third argument of makecontext behaves in Solaris x86 for versions <9 vs. >=9. Have no machine to test this. *** sthreads.c.15b1 Wed Nov 9 12:18:18 2005 --- sthreads.c Wed Nov 9 14:56:18 2005 *************** *** 78,85 **** return NULL; } ctx->ctx.uc_link=NULL; /* it should never happen */ ! ctx->ctx.uc_stack.ss_sp=ctx->stack; ! #if defined(__sgi) || (defined(CPU_SPARC) && ( \ defined(OS_SOLARIS2_0) || \ defined(OS_SOLARIS2_1) || \ defined(OS_SOLARIS2_2) || \ --- 78,84 ---- return NULL; } ctx->ctx.uc_link=NULL; /* it should never happen */ ! #if defined(CPU_SPARC) && ( \ defined(OS_SOLARIS2_0) || \ defined(OS_SOLARIS2_1) || \ defined(OS_SOLARIS2_2) || \ *************** *** 88,96 **** defined(OS_SOLARIS2_5) || \ defined(OS_SOLARIS2_6) || \ defined(OS_SOLARIS2_7) || \ ! defined(OS_SOLARIS2_8))) ! (char *)ctx->ctx.uc_stack.ss_sp+=STACK_SIZE-8; #endif ctx->ctx.uc_stack.ss_size=STACK_SIZE; ctx->ctx.uc_stack.ss_flags=0; --- 87,101 ---- defined(OS_SOLARIS2_5) || \ defined(OS_SOLARIS2_6) || \ defined(OS_SOLARIS2_7) || \ ! defined(OS_SOLARIS2_8)) ! #define MCtxA 2 ! #else ! #define MCtxA 1 #endif + #if defined(__sgi) || (MCtxA == 2) + ctx->stack+=STACK_SIZE-8; + #endif + ctx->ctx.uc_stack.ss_sp=ctx->stack; ctx->ctx.uc_stack.ss_size=STACK_SIZE; ctx->ctx.uc_stack.ss_flags=0; *************** *** 121,127 **** if(!ctx) return -1; s_log(LOG_DEBUG, "Context %ld created", ctx->id); ! makecontext(&ctx->ctx, (void(*)(void))cli, 1, arg); return 0; } --- 126,132 ---- if(!ctx) return -1; s_log(LOG_DEBUG, "Context %ld created", ctx->id); ! makecontext(&ctx->ctx, (void(*)(void))cli, MCtxA, arg); return 0; }

1 0

Re: [stunnel-users] Solaris: new_context, makecontext
by Hans Werner Strube 09 Nov '05

09 Nov '05

Alexander Lazic wrote: > The semantics of the uc_stack member of the ucontext_t structure have > changed as they apply to inputs to makecontext(). Prior to Solaris 10, > the ss_sp member of the uc_stack structure represented the high memory > address of the area reserved for the stack. The ss_sp member now > represents the base (low memory address), in keeping with other uses of > ss_sp. > > This change in the meaning of ss_sp is now the default behavior. The > -D__MAKECONTEXT_V2_SOURCE compilation flag used in Solaris 9 update > releases to access this behavior is obsolete. Solaris 9_x86 does not use __MAKECONTEXT_V2_SOURCE, see /usr/include/ucontext.h: *** Solaris 9: #if defined(__MAKECONTEXT_V2_SOURCE) && defined(__sparc) #ifdef __PRAGMA_REDEFINE_EXTNAME #pragma redefine_extname makecontext __makecontext_v2 #else #define makecontext __makecontext_v2 #endif #endif *** Solaris 10: #ifdef __sparc #ifdef __PRAGMA_REDEFINE_EXTNAME #pragma redefine_extname makecontext __makecontext_v2 #else #define makecontext __makecontext_v2 #endif In libc.so.1, __makecontext_v2 is the "new" function and makecontext the "old" one (binary compatible to Solaris 8 and earlier). For Solaris x86, there is only one function makecontext in all versions. Thus the behavior might not have been changed. A small test program in Solaris 9_x86 (have no earlier versions) indicated the "new" behavior.

2 1

IPTables/Stunnel Help needed
by Alaric Dailey 08 Nov '05

08 Nov '05

Ok, I have a fairly complex firewall setup on my Linux box with 7 host IPs, 2 Internet connections, and a DMZ. I need to direct 1 IP to stunnel on the firewall to be forwarded to the mail server in the DMZ on ports 8383, 993, 995 etc.. but leave ports 25, and 3389 forwarded via iptables, can anyone lend a hand?

1 0

Re: [stunnel-users] Solaris: new_context, makecontext
by Hans Werner Strube 08 Nov '05

08 Nov '05

Michal Trojnara wrote: > I had no Solaris 10 to test it... So what's the correct way to detect > Solaris 10? This must probably be done by a modified configure and a new #if or #ifdef. uname -r gives 5.10 for Solaris 10. > Solaris 9 behaves correctly if __MAKECONTEXT_V2_SOURCE is defined (it > is defined in common.h). I've tested this case. Yes, of course. I have not stated anything different. > The manual claims: > The value of > argc must match the number of integer arguments passed to > func, otherwise the behavior is undefined. > (SunOS mion 5.9 Generic_112233-12 sun4u sparc SUNW,Ultra-250) > > I only pass single argument... Yes, but up to Solaris 8 it claimed: The integer value of argc must be one-greater-than the number of arguments that follow argc; otherwise, the behavior is undefined. For 5 arguments, the value of argc must be 6. AfaIk this would also hold for Solaris 9 without __MAKECONTEXT_V2_SOURCE, but this is deprecated. Btw, Solaris 9 and 10 can still handle binaries compiled under earlier versions.

1 0

Re: stunnel 4.14 and openssl 0.9.8a
by MeeAGhost 08 Nov '05

08 Nov '05

I am using stunnel to tunnel my VNC connection. As recent as version 4.12 I have been able to use the libeay32.dll and libssl32.dll files I created with openssl 0.9.X.X using Msys and Mingw with the steps below: 1. I comment out line 910 in the openssl Configure script 2. Create .def for dll's: perl util/mkdef.pl 32 libeay > ms/libeay32.def perl util/mkdef.pl 32 ssleay > ms/ssleay32.def 3. Execute the openssl configure script ./Configure mingw 4. Execute the make for openssl make 5. After the make completes I create the dlls dllwrap --dllname libeay32.dll --output-lib libeay32.a --def ms/libeay32.def libcrypto.a -lwsock32 -lgdi32 dllwrap --dllname libssl32.dll --output-lib libssl32.a --def ms/ssleay32.def libssl.a libeay32.a When I place the resulting dlls on the remote machine with version 4.14installed and establish a VNC connection to machine stunnel crashes and generates a Dr. Watson error. If I revert back to version 4.12 the dlls I created using the steps above --- everything works. If I install the 0.9.7idlls (which come as a part of the 4.14 Win Install package) everything works. I am trying to keep server and clients on the same version of openssl which is why I have been compling new dlls whenever I have upgraded the servers. Is there something unique being done in version 4.14 that prevents dlls created with 0.9.8a from being used? -- Chris

2 3

Solaris: new_context, makecontext
by Hans Werner Strube 08 Nov '05

08 Nov '05

Concerns all versions of sthreads.c using ucontext: As far as I know from the Solaris documentation and some experiments, the context treatment in new_context() is not universally correct. #if defined(__sgi) || (defined(__sparc) && !defined(HAVE___MAKECONTEXT_V2)) ctx->ctx.uc_stack.ss_sp=ctx->stack+STACK_SIZE-8; #else /* not an IRIX or old and buggy Solaris on Sparc */ ctx->ctx.uc_stack.ss_sp=ctx->stack; #endif The first case is wrong for Solaris 10, since HAVE___MAKECONTEXT_V2 is not defined there but the behavior is _always_ as with HAVE___MAKECONTEXT_V2 in Solaris 9 SPARC. (Solaris 9 SPARC behaves like Solaris 7 and 8 SPARC ["old and buggy"] if HAVE___MAKECONTEXT_V2 is not set, and like Solaris 10 if it is set.) In the "old and buggy" case, the third argument of makecontext() should be 2 instead of 1 (although the latter _may_ work). I do not know how this is in Solaris x86. At least, HAVE___MAKECONTEXT_V2 has no meaning there.

3 3