Subject: Re: Kernel hangs
To: None <>
From: Dave Huang <>
List: port-powerpc
Date: 03/24/2004 02:15:53
On Sun, Mar 21, 2004 at 12:19:51PM +0100, wrote:
> Aymeric Vincent <> wrote:
> > For several months, various people have reported kernel hangs on 
> > macppc. I would like to (help) track down the issue.
> > [...]
> > Especially, I would like to know if other PowerPC ports experience such 
> > "hangs" (you can only trigger the debugger from the console, and 
> > everything else is stuck), and on which processors (40x, 604, G3, G4?). 
> I got no report, but here is the answer to the question anyway: all ports using
> sys/arch/powerpc/oea/pmap.c were affected.
> rev. >= 1.22 should fix the problem.

My Powermac Dual G4 does seem more stable with a uniprocessor kernel,
but with an options MULTIPROCESSOR kernel, I'm still having the
lockups described in the "Recent macppc kernels hang under load"

Although instead of just hanging, a " -j2 distribution"
generally panics now:

mp_save_fpu_proc{1} pid = 3354.1, fcpu->ci_cpuid = 0
panic: mp_save_fpu_proc
Stoped in pid 3354.1 (nbmake) at netbsd:cpu_Debugger+0x10:  lwz  r0, r1, 0x14
db{1}> LK_SPIN spinout, excl 1, share 0
held by CPU 1
last locked at /usr/src.local/sys/kern/kern_lock.c:1355
last unlocked at /usr/src.local/sys/kern/kern_synch.c:858
0x0040cf30: at _lockmgr+5b0
0x0040cf80: at do_pending_int+c0
0x0040cfc0: at spllower+50
0x0040cfe0: at lcsplx+c
0x0040cff0: at Idle+18
Faulted in DDB; continuing...
dbtrap: pid 3354.1 (nbmake): kernel PGM trap @ 0x370000 (SRR1=0x80032)
(then it hangs)

The above was copied by hand, so there may be typos. Kernel is 1.6ZK
from March 21 23:00 GMT or so (it does have rev 1.22 of pmap.c).
Name: Dave Huang         |  Mammal, mammal / their names are called /
INet: |  they raise a paw / the bat, the cat /
FurryMUCK: Dahan         |  dolphin and dog / koala bear and hog -- TMBG
Dahan: Hani G Y+C 28 Y++ L+++ W- C++ T++ A+ E+ S++ V++ F- Q+++ P+ B+ PA+ PL++