Re: loading new envstat properties reboots system

[Paul Goyette <paul%whooppee.com@localhost>]

Hmmm.

I only have one machine remaining with an amdtemp sensor.  I tried to 
reproduce this problem, but was unsuccessful.  My kernel is a bit newer 
than yours - 5.99.39 from 2010-10-28 16:51:58 UTC.  (The Aug. 11 kernel 
no longer exists on nyftp.netbsd.org as far as I can see.)

There was only one change to envsys code between your kernel and mine:

   Module Name:    src
   Committed By:   pgoyette
   Date:           Wed Sep 22 00:13:55 UTC 2010

   Modified Files:
        src/sys/dev/sysmon: sysmon_envsys_events.c

   Log Message:
   1. Don't try to call the driver's get_limits routine if it doesn't
      have one.  Fixes a panic reported by njoly@ in private Email

   2. Acquire the device before checking whether or not the get_limits
      routine exists.

   XXX There's still a bug in here somewhere that prevents setting of
   XXX limit values after using 'envstat -S'

   To generate a diff of this commit:
   cvs rdiff -u -r1.92 -r1.93 src/sys/dev/sysmon/sysmon_envsys_events.c


It is entirely possible that this is the source of your problem.

There have recently been some additional changes in this area to fix 
bugs the prevented limits from actually being removed (even though 
envstat seemed to indicate that they were removed).  If possible, you 
should update.

Otherwise, you should avoid using "envstat -S", which also means avoid 
using "/etc/rc.d/envsys stop" or "/etc/rc.d/envsys restart".


On Mon, 13 Dec 2010, Jeremy C. Reed wrote:

> The envstat output is:
>                  Current  CritMax  WarnMax  WarnMin  CritMin Unit
> [amdtemp0]
>  CPU0 Sensor0:    70.500   92.000                     85.000 degC
>  CPU0 Sensor1:    73.250   92.000                     85.000 degC
>  CPU1 Sensor0:    68.500   92.000                     85.000 degC
>  CPU1 Sensor1:    67.250   92.000                     85.000 degC
>
> The dmesg is at
> http://tx.reedmedia.net:17777/dmesg.boot
>
> NetBSD 5.99.38 NetBSD 5.99.38 (GENERIC) #0: Tue Aug 10 19:00:23 UTC 2010
> builds%b8.netbsd.org@localhost:/home/builds/ab/HEAD/amd64/201008110000Z-obj/home/builds/ab/HEAD/src/sys/arch/amd64/compile/GENERIC
> amd64 x86_64
>
> I set ddb.onpanic=1.
>
> I ran envstat -S and got:
>
> uvm_fault(0xffff80003eb3eba0, 0x0, 4) -> e
> fatal page fault in supervisor mode
> trap type 6 code 10 rip 0 cs 8 flags 10246 cr2 0 cp 0 rsp ffff8000003e1598c8
> kernel: page fault trap, code=0
> Stopped in pid 6850.1 (envstat) at     0:  invalid address
>
> bt resulted in
> Faulting DDB, continuing
>
> so did a sync and that dumped to dev 0,1 offset 267193
> ... 3 2 1 succeeded
>
> I have these in /var/crash/
>
> -rw-------  1 root  wheel  449947102 Dec 13 12:17 netbsd.1.core.gz
> -rw-------  1 root  wheel     479688 Dec 13 12:17 netbsd.1.gz
>
> (gdb) target kvm netbsd.1.core
> #0  0x00000000 in sigcode ()
>
> No problem running this envstat -S at initial startup though (via
> /etc/rc.d script).
>
> (By the way, I don't know why I didn't get this crash dump earlier. I
> thought it was automatic with kern.dump_on_panic=1.)
>
> !DSPAM:4d0667c92361176411142!

-------------------------------------------------------------------------
| Paul Goyette     | PGP Key fingerprint:     | E-mail addresses:       |
| Customer Service | FA29 0E3B 35AF E8AE 6651 | paul at whooppee.com    |
| Network Engineer | 0786 F758 55DE 53BA 7731 | pgoyette at juniper.net |
| Kernel Developer |                          | pgoyette at netbsd.org  |
-------------------------------------------------------------------------