unstable 5.3 boxes

[Doug White]

On Wed, 5 Jan 2005, Mitch Parks wrote:

> It seems both my test and production boxes are not stable with 5.3-Release.
> I'm hoping for some insight in how to make at least the production box
> not crash.

[...]

> #7  0xe93a0018 in ?? ()
> #8  0xc0580010 in kvprintf (fmt=0xc7b04c60 "?6w?\v6t?\v6t?", func=0x59,
>      arg=0xe93ad9bc, radix=-1067875706, ap=0xc3839674 "\001")
>      at /usr/src/sys/kern/subr_prf.c:643
> #9  0xc058940a in selwakeuppri (sip=0x0, pri=0)
>      at /usr/src/sys/kern/sys_generic.c:1096
> #10 0xc0598286 in ttwakeup (tp=0xc05842e8) at /usr/src/sys/kern/tty.c:2366

Something weird is going on here. selwakeuppri() doesn't call any of the
printf functions -- it calls doselwakeup() and thats it. I don't see
anything in doselwakeup() that would call a printf function offhand.  The
null arguments in frame 9 are also odd and would certainly trip up
doselwakeup() since it immediately dereferences sid.

This was opening a serial device it looks like.  Do you have syslog set to
output to a serial terminal, or have getty enabled on ttyd*, or something
plugged into any serial port, or something of that nature?  The later call
to ttwakeup() occurs in a block where carrier has been raised.

-- 
Doug White                    |  FreeBSD: The Power to Serve
dwhite at gumbysoft.com          |  www.FreeBSD.org