I previously posted about more or less this same problem four months ago in the following thread: http://www.linuxquestions.org/questi...hreadid=312903

I am running Slackware 10.0 2.6.7 kernel. As stated previously I had attempted to use swaret to update my server. However, I'm not sure if I had this problem before then. I recently installed my 10.0 CDs on something else and Sendmail is doing the same thing.

I tried what was suggested and I've even gone as far as to reinstall all libraries from the Slackware 10.0 CD as well as build a new package of Sendmail 8.13.4 and installed it. Same issues. Along with the things I said in the Link above I now have the following things I've noticed or have been pointed out by my users. One has nothing to do with sendmail itself or atleast to my knowledge.

When an email faults or for whatever reason gets stuck in sending mode (which is what I posted about in my other thread as being the main problem) I get the following info in syslog:

Aug 4 13:09:30 gator kernel: <1>Unable to handle kernel NULL pointer dereference at virtual address 0000026a
Aug 4 13:09:30 gator kernel: printing eip:
Aug 4 13:09:30 gator kernel: f9cea0ad
Aug 4 13:09:30 gator kernel: *pde = 00000000
Aug 4 13:09:30 gator kernel: Oops: 0000 [#601]
Aug 4 13:09:30 gator kernel: SMP
Aug 4 13:09:30 gator kernel: Modules linked in: fzvusm
Aug 4 13:09:30 gator kernel: CPU: 1
Aug 4 13:09:30 gator kernel: EIP: 0060:[<f9cea0ad>] Not tainted
Aug 4 13:09:30 gator kernel: EFLAGS: 00010246 (2.6.7)
Aug 4 13:09:30 gator kernel: EIP is at ocultarProceso+0x5d/0xc0 [fzvusm]
Aug 4 13:09:30 gator kernel: eax: 0000004e ebx: c2354a88 ecx: c205a460 edx: 00000000
Aug 4 13:09:30 gator kernel: esi: 0000026a edi: c2354a68 ebp: 0000026a esp: e4bd8f70
Aug 4 13:09:30 gator kernel: ds: 007b es: 007b ss: 0068
Aug 4 13:09:30 gator kernel: Process write (pid: 28442, threadinfo=e4bd8000 task=f408cdd0)
Aug 4 13:09:30 gator kernel: Stack: 0000000d 00000000 00000018 00000078 f2386908 f238691b f9ce8b92 c2354ce8
Aug 4 13:09:30 gator kernel: f238691b 00000046 f2386858 f23868d8 00000001 f2386888 00000048 00000030
Aug 4 13:09:30 gator kernel: 00000003 0000000d 00000003 e4bd8000 c010401f 00000003 0804d6d8 00000400
Aug 4 13:09:30 gator kernel: Call Trace:
Aug 4 13:09:30 gator kernel: [<f9ce8b92>] modGetdents64+0x1a2/0x1b0 [fzvusm]
Aug 4 13:09:30 gator kernel: [<c010401f>] syscall_call+0x7/0xb
Aug 4 13:09:30 gator kernel:
Aug 4 13:09:30 gator kernel: Code: ac ae 75 08 84 c0 75 f8 31 c0 eb 04 19 c0 0c 01 85 c0 75 d8

If I type sendmail -v to see the version (New issue) I get this stuff:
Aug 4 13:14:03 gator kernel: <1>Unable to handle kernel NULL pointer dereference at virtual address 0000026a
Aug 4 13:14:03 gator kernel: printing eip:
Aug 4 13:14:03 gator kernel: f9cea0ad
Aug 4 13:14:03 gator kernel: *pde = 00000000
Aug 4 13:14:03 gator kernel: Oops: 0000 [#604]
Aug 4 13:14:03 gator kernel: SMP
Aug 4 13:14:03 gator kernel: Modules linked in: fzvusm
Aug 4 13:14:03 gator kernel: CPU: 0
Aug 4 13:14:03 gator kernel: EIP: 0060:[<f9cea0ad>] Not tainted
Aug 4 13:14:03 gator kernel: EFLAGS: 00010246 (2.6.7)
Aug 4 13:14:03 gator kernel: EIP is at ocultarProceso+0x5d/0xc0 [fzvusm]
Aug 4 13:14:03 gator kernel: eax: 0000004e ebx: c2354a88 ecx: c205a460 edx: 00000000
Aug 4 13:14:03 gator kernel: esi: 0000026a edi: c2354a68 ebp: 0000026a esp: ee087f70
Aug 4 13:14:03 gator kernel: ds: 007b es: 007b ss: 0068
Aug 4 13:14:03 gator kernel: Process sendmail (pid: 29962, threadinfo=ee087000 task=f2089360)
Aug 4 13:14:03 gator kernel: Stack: 0000000d 00000000 00000018 00000078 d1786988 d178699b f9ce8b92 c2354ce8
Aug 4 13:14:03 gator kernel: d178699b 00000046 d17860d8 d1786958 00000001 d1786108 00000048 00000030
Aug 4 13:14:03 gator kernel: 00000003 0000000d 00000003 ee087000 c010401f 00000003 08129a48 00000400
Aug 4 13:14:03 gator kernel: Call Trace:
Aug 4 13:14:03 gator kernel: [<f9ce8b92>] modGetdents64+0x1a2/0x1b0 [fzvusm]
Aug 4 13:14:03 gator kernel: [<c010401f>] syscall_call+0x7/0xb
Aug 4 13:14:03 gator kernel:
Aug 4 13:14:03 gator kernel: Code: ac ae 75 08 84 c0 75 f8 31 c0 eb 04 19 c0 0c 01 85 c0 75 d8

And lastly a user emailed me today saying that when he types the write command it also segfaults. If I as root do it it works, but not for normal users. Heres the syslog for that too.
Aug 4 13:09:36 gator kernel: <1>Unable to handle kernel NULL pointer dereference at virtual address 0000026a
Aug 4 13:09:36 gator kernel: printing eip:
Aug 4 13:09:36 gator kernel: f9cea0ad
Aug 4 13:09:36 gator kernel: *pde = 00000000
Aug 4 13:09:36 gator kernel: Oops: 0000 [#603]
Aug 4 13:09:36 gator kernel: SMP
Aug 4 13:09:36 gator kernel: Modules linked in: fzvusm
Aug 4 13:09:36 gator kernel: CPU: 1
Aug 4 13:09:36 gator kernel: EIP: 0060:[<f9cea0ad>] Not tainted
Aug 4 13:09:36 gator kernel: EFLAGS: 00010246 (2.6.7)
Aug 4 13:09:36 gator kernel: EIP is at ocultarProceso+0x5d/0xc0 [fzvusm]
Aug 4 13:09:36 gator kernel: eax: 0000004e ebx: c2354a88 ecx: c205a460 edx: 00000000
Aug 4 13:09:36 gator kernel: esi: 0000026a edi: c2354a68 ebp: 0000026a esp: d642ff70
Aug 4 13:09:36 gator kernel: ds: 007b es: 007b ss: 0068
Aug 4 13:09:36 gator kernel: Process write (pid: 28473, threadinfo=d642f000 task=cfb225c0)
Aug 4 13:09:36 gator kernel: Stack: 0000000d 00000000 00000018 00000078 c21cc388 c21cc39b f9ce8b92 c2354ce8
Aug 4 13:09:36 gator kernel: c21cc39b 00000046 c21cc558 c21cc358 00000001 c21cc588 00000048 00000030
Aug 4 13:09:36 gator kernel: 00000003 0000000d 00000003 d642f000 c010401f 00000003 0804d6d8 00000400
Aug 4 13:09:36 gator kernel: Call Trace:
Aug 4 13:09:36 gator kernel: [<f9ce8b92>] modGetdents64+0x1a2/0x1b0 [fzvusm]
Aug 4 13:09:36 gator kernel: [<c010401f>] syscall_call+0x7/0xb
Aug 4 13:09:36 gator kernel:
Aug 4 13:09:36 gator kernel: Code: ac ae 75 08 84 c0 75 f8 31 c0 eb 04 19 c0 0c 01 85 c0 75 d8


I am now getting presure from admin in regards to the repeative email sends and what am I going to do about it. At this point I think the only other option I may have is to DL Slackware 10.1 and either redo eveything or attempt an upgrade that way which I'm trying not to do.

As always any thoughts or suggestions would be greatly appriciated.
Thanks.
-Smillie

You configured sendmail in this machine or another person ?

When you start/stop sendmail, all goes fine ?
Did you consider upgrade the kernel or it is not an option ?

I configured sendmail on this machine. For the most part my sendmail.cf file is pretty standard. I just added stuff for DNSBL and BitDefender Milter agent.

When I stop and start sendmail VIA the RC script or use Webmin to do it I don't have any problems doing so.

I am trying to avoid upgrading the kernel, but if you guys think its a factor I'm will to try. I can't imagine things can get too much weirder around this place...:> If thats what you recommend where do you think I should start. Just get it off the Slackware site as a package or another alternative.
-Smillie

Frankly, I don't know if kernel upgrade will fix your problem :/

But I would try it just to see, as for how to upgrade it, for my
part, I always compile and install kernels from www.kernel.org
but it is up to you, choose the method you are the most familiar
with

OK I DLed the 2.6.12.3 kernel source from kernels.org without a problem. Unarc'd it and the first time through I tried using my previously saved config to configure and compile it. For whatever reason my crappy monitor in the MDF told me that the signal was out of range when I started up with the kernel in place so I rebooted back to my old junk and config'd from scratch without any frame buffers or any other stuff I really didn't have to have and then the machine booted right up without an issue.

I have tried the write command and used the sendmail command as a normal user and they do infact work now.

I am currently flushing the mail que to see if I get anything out of the ordinary. Unfortunately I most likely have to wait till some idiot here forwards an email to like 50 people before I know for sure if we have this totally worked out now or not.

I think I found a new problem though. When I login to SSH this pops up in my syslog:
Aug 4 17:42:02 gator sshd[6681]: error: open /dev/tty failed - could not set controlling tty: Is a directory
Aug 4 17:43:44 gator sshd[7197]: error: open /dev/tty failed - could not set controlling tty: Is a directory

But I can still login and work normally so I think that will be a problem for another day.

In the mean time thank you so much for your info & help so far. I know we knocked out 2 of 3 and it wouldn't surprise me now if the 3 is gone too. Again thank you.
-Smillie

You may find this interesting:

http://linuxbugs.coverity.com/linuxbugs.htm

41% of the bugs found in kernel 2.6.9 are null-pointer related. Most of these were fixed in 2.6.12, they say. Perhaps you should give it a try...

Thanks Ilgar, I think you just reassured me that I did the right thing then:>
-Smillie