Subject: Re: kern/27067: cpu stuck in atabus3 thread, kernel won't finish booting. 2.0F works, 2.0H and 2.0Beta do not.
To: Manuel Bouyer <>
From: Chris Tribo <>
List: netbsd-bugs
Date: 09/30/2004 15:00:51
Could you provide me with a little insight as to how to do that? I read 
the ddb man page but it didn't shed a whole lot of light on the right 
thing to do. Trying to do break pdcintr results in symbol not found. I 
used userconf to disable pdcide and the kernel booted to asking for the 
root device, so it would seem the issue is somewhere close to the 
pdcide interrupt code. Do I need to have a kernel built that isn't 
stripped from the 2.0 Beta/RC1 sources in order to set a break on 

with pdcide on, show all procs gives

atabus2 waiting on atarst, atabus1 in atacnf, atabus0, atapibus0 and 
atabus3 aren't in wait, or ddb doesn't know what it is.

show sched_qs lists

4	3.1	(atabus0)
8	9.1	(pms0)
	10.1	(apm0)
	11.1	(atapibus0)

On Sep 30, 2004, at 2:14 PM, Manuel Bouyer wrote:

> On Tue, Sep 28, 2004 at 07:26:46PM +0000, 
> wrote:
>> Stopped in pid 8.1 (atabus3) at netbsd:cpu_Debugger+0x4: leave
>> db> bt
>> cpu_Debugger+0x4
>> internal_command+0x13c
>> wskbd_translate+0x6c
>> wskbd_input+0x137
>> pckbd_input+0x53
>> pckbportintr+0x3a
>> pckbcintr+0x94
>> Xintr_legacy1+0x4a
>> --- interrupt ---
>> DDB lost frame for netbsd:Xdoreti, trying 0xc9180c64
> This doesn't help much :(
> Could you see if it's stuck in an interrupt loop on pdcintr() ?
> -- 
> Manuel Bouyer <>
>      NetBSD: 26 ans d'experience feront toujours la difference
> --