Re: Keyboard lockup with NetBSD5

[Alexander Coers <Alexander.Coers%gmx.de@localhost>]

Hi Frank,

this one is very long.
First this is the dmesg output running your provided NetBSD5 kernel.

<-----------------------------------------------------

Copyright (c) 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,  
2005,
    2006, 2007, 2008, 2009
    The NetBSD Foundation, Inc.  All rights reserved.
Copyright (c) 1982, 1986, 1989, 1991, 1993
    The Regents of the University of California.  All rights reserved.

NetBSD 5.99.20 (GENERIC) #3: Wed Oct  7 21:38:46 CEST 2009
        frank%cygnus.wendik.de@localhost:/home/frank/netbsd/current/src/sys/ 
arch/amiga/compile/obj/GENERIC
Amiga 1200 (68060 rev.6 CPU/MMU/FPU)
total memory = 255 MB
avail memory = 247 MB
memory segment 0 at 70000000 size 0ff80000
memory segment 1 at 00000000 size 00200000
timecounter: Timecounters tick every 10.000 msec
mainbus0 (root)
clock0 at mainbus0: CIA B system hz 100 hardware hz 709379
timecounter: Timecounter "CIA B" frequency 709379 Hz quality 100
a2kbbc0 at mainbus0
ser0 at mainbus0: input fifo 512 output fifo 32
par0 at mainbus0
kbd0 at mainbus0: CIA A type Amiga
ms0 at mainbus0
grfcc0 at mainbus0
grf0 at grfcc0: width 640 height 400 colors 4
ite0 at grf0: rows 50 cols 80 repeat at (30/100)s next at (10/100)s  
has keyboard
fdc0 at mainbus0: dmabuf pa 0x1e0c5c: dmabuf ka 0x636c5c
fd0 at fdc0 unit 0: 3.5dd 80 cyl, 2 head, 11 sec [9 sec], 512 bytes/sec
wdc0 at mainbus0
atabus0 at wdc0 channel 0
pccard0 at mainbus0
pcmcia0 at pccard0
ne0 at pcmcia0 function 0: <D-Link, DFE-670TXD, PC Card>
ne0: Ethernet address 00:13:46:34:0d:85
OUI 0x1000e8 model 0x0002 rev 0 at ne0 phy 1 not configured
ne0: WARNING: powerhook_establish is deprecated
aucc0 at mainbus0
audio0 at aucc0: half duplex, playback, capture
zbus0 at mainbus0
board at zbus0: pa 0xf00c34 man/pro 8512/110 not configured
timecounter: Timecounter "clockinterrupt" frequency 100 Hz quality 0
wd0 at atabus0 drive 0: <CF Card>
wd0: drive supports 1-sector PIO transfers, LBA addressing
wd0: 1923 MB, 3909 cyl, 16 head, 63 sec, 512 bytes/sect x 3940272  
sectors
wd0: drive supports PIO mode 4, DMA mode 2, Ultra-DMA mode 5 (Ultra/100)
4 views configured
Kernelized RAIDframe activated
root on wd0a dumps on wd0b
root file system type: ffs

----------------------------------------------------->

Output from log:

<-----------------------------------------------------
a1200# tail /var/log/messages
Oct 14 00:19:49 a1200 /netbsd: timecounter: Timecounter  
"clockinterrupt" frequency 100 Hz quality 0
Oct 14 00:19:49 a1200 /netbsd: wd0 at atabus0 drive 0: <CF Card>
Oct 14 00:19:49 a1200 /netbsd: wd0: drive supports 1-sector PIO  
transfers, LBA addressing
Oct 14 00:19:49 a1200 /netbsd: wd0: 1923 MB, 3909 cyl, 16 head, 63  
sec, 512 bytes/sect x 3940272 sectors
Oct 14 00:19:49 a1200 /netbsd: wd0: drive supports PIO mode 4, DMA  
mode 2, Ultra-DMA mode 5 (Ultra/100)
Oct 14 00:19:49 a1200 /netbsd: 4 views configured
Oct 14 00:19:49 a1200 /netbsd: Kernelized RAIDframe activated
Oct 14 00:19:49 a1200 /netbsd: root on wd0a dumps on wd0b
Oct 14 00:19:49 a1200 /netbsd: root file system type: ffs
Oct 14 00:19:51 a1200 savecore: can't find device 10/491552
----------------------------------------------------->


There is no difference between the load before and after the keyboard  
lockup.
(measured with top, ranging from 97% to 99% idle)
vmstat -i gives the following error:
vmstat: time makes no sense; namelist must be wrong.

On NetBSD4, vmstat works.

If I type on the keyboard the last char is repeated and after a pause  
it it again repeated (roughly 5-9 times).
But if I wait somewhat the keyboard comes back online and I can enter  
a few chars before it locks up again (at least I managed to get to  
this point for one time).

regards,
 Alex




Am 14.10.2009 um 14:09 schrieb Frank Wille:

> On Wed, 14 Oct 2009 11:16:17 +0200
> "Alexander Coers" <Alexander.Coers%gmx.de@localhost> wrote:
>
> > I have a DLink PCMCIA ethernet card, but while installing and
> > configuring sshd I stumbled on another "feature" of NetBSD 4.0.1, the
> > "can not change password" thing (Message is something like "Couldn't
> > generate salt Unable to change auth token: error in service module").
>
> Oh yes, that one...
>
>
> > Now I have to edit something in /etc/passwd.conf and hope that I can
> > change my password and get access to the amiga via ssh from my Mac.
> > At least ping works...
>
> That's promising. When you get remote access, please let us know about
> anything unusual, like system load or the kbd-interrupts (vmstat -i).
>
>
> Just another point to check: can you build and try a kernel with  
> wscons
> enabled, instead of ite? I don't remember if anybody ever tested it
> (I didn't), but I guess you have to enable in the config file:
> - amidisplaycc0, wsdisplay0 and the following options
> - wskbd0
> - wsmouse
> . pseudo-device wsmux (and probably wsfont as well)
> And disable all ite*.
>
> I will try that myself during the following days and prepare a kernel
> for you, if you like.
>
> That might be an interesting test, because the only modification in  
> kbd.c
> since NetBSD 4 was at the bottom of kbdstuffchar(), called by  
> kbdintr().
> It has to do with the timecounter/clock modifications, and now we are
> calling the MI firm_gettime() function there.
>
> But for wscons, this part is not even reached.
>
> --
> Frank Wille