>Number:         15305
>Category:       kern
>Synopsis:       v6 enabled sendmail causes 1.5.3 crash
>Confidential:   no
>Severity:       critical
>Priority:       high
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Sat Jan 19 17:31:00 PST 2002
>Closed-Date:
>Last-Modified:
>Originator:     Michael Richardson
>Release:        NetBSD 1.5.2
>Organization:
Sandelman Software Works Corporation
>Environment:

Architecture: i386

NetBSD nox.sandelman.ottawa.on.ca 1.5.3_ALPHA NetBSD 1.5.3_ALPHA (LOX) #13: Wed Oct 17 15:31:26 EDT 2001     mcr@istari.sandelman.ottawa.on.ca:/j/netbsd/obj/LOX i386

>Description:

I am getting a periodic (every three or four days) crash on a 1.5.2 machine
running a v6 + TLS enabled copy of sendmail. I am running the identical
kernel on another machine (slightly different underlying hardware) 
that does not receive v6 SMTP connections and it stays up for weeks/months.

This has been going on since August. I have tried repeated kernel updates
along the 1.5 branch. (I am happy to update again, and will do soon)

>How-To-Repeat:

Unclear how to force the issue. Here is the traceback.

db> trace
soo_poll(c7738270,58,c7721e34) at soo_poll+0xb8
selscan(c7721e34,c794fe74,c794fe7c,14,c794ff78) at selscan+0xd0
sys_select(c7721e34,c794ff80,c794ff78,0,bfbfdcf4) at sys_select+0x259
syscall() at syscall+0x1e7                                           
--- syscall (number 93) ---
0x482ad7b3:                
db> ps     
 PID             PPID       PGRP        UID S   FLAGS          COMMAND    WAIT
 6918               1       6918          0 3    0x84 sendmail-8.11.6-   netio
 6911             232        232          0 3    0x84 sendmail-8.11.6-  netcon
 6874             257        257      32767 3   0x184            httpd   lockf
 6873             257        257      32767 3   0x184            httpd  select
 6872             257        257      32767 3   0x184            httpd   lockf
 6871             257        257      32767 3   0x184            httpd   lockf
 6870             257        257      32767 3   0x184            httpd   lockf
 6869             257        257      32767 3   0x184            httpd   lockf
 5869            5860       5869        179 3  0x4086              zsh   ttyin
 5860            5857        223        179 3  0x4184            xterm  select
 5857             223        223          0 3    0x84            sshd2  select
 261                1        261          0 3  0x4086            getty   ttyin
 260                1        260          0 3  0x4086            getty   ttyin
 259                1          1          0 3  0x4084            getty nanosle
 257                1        257          0 3    0x84            httpd  select
 256              235          5       5002 3  0x4186           mysqld  select
 241                1        241          0 3    0x84             cron nanosle
 238                1        238          0 3    0x84            inetd  select
 235                1          5          0 3  0x4086               sh    wait
 232                1        232          0 3    0x84 sendmail-8.11.6-  select
 223                1        223          0 3    0x84            sshd2  select
 209                1        209          0 3    0x84             ntpd   pause
 203                1        203          0 3    0x84           routed  select
>Fix:
	
>Release-Note:
>Audit-Trail:
>Unformatted:
 >150                1        150          0 7    0x44            named        
  141                1        141          0 2    0x84          syslogd
  4                  0          0          0 3 0x20204          ioflush  syncer
  3                  0          0          0 3 0x20204           reaper  reaper
  2                  0          0          0 3 0x20204       pagedaemon daemon_
  1                  0          1          0 3  0x4084             init    wait
  0                 -1          0          0 3 0x20204          swapper schedul
 
 
   If I knew how to get the netstat equivalent output from kdb, I'd do that
 too to see how many v6 connections are alive.