Okay, I'm pretty sure this isn't a hardware problem now. I was seeing
periodic crashes with 4.5-RELEASE on a new laptop I was building.
Various posts on this list convinced me that I had a hardware problem.
I tried swapping out the RAM and it seemed to crash less (and in a
different place). Crashes after swapping the RAM were always in the xl
routines.
So, being convinced that I had a hardware problem (mother board or cpu
or something like that), I swapped out the *entire machine*.
I'm still seeing the crashes.
So, I have /usr/ports nfs mounted from another freebsd box.
I cd /usr/ports/misc/instant-workstation; make install
The machine panics (sometimes right away, sometimes after several
minutes). Here's the panic message:
Fatal trap 12: page fault while in kernel mode
fault virtual address = 0xc0fcec39
fault code = supervisor write, page not present
instruction pointer = 0x8:0xc0205243
stack pointer = 0x10:0xcde4fbd0
frame pointer = 0x10:0xcde4fbe4
code segment = base 0x0, limit 0xfffff, type 0x1b
= DPL 0, pres 1, def32 1, gran 1
processor eflags = interrupt enabled, resume, IOPL = 0
current process = 80 (nfsiod)
interrupt mask = net tty
kernel: type 12 trap, code=0
Stopped at nfs_realign+0xdb: incb 0(%eax,%edx,1)
db> trace
nfs_realign(c0ea5200,14,0,1,cdeb0cc0) at nfs_realign+0xdb
nfs_receive(c17ca280,cde4fc6c,cde4fc70,0,1) at nfs_receive+0x45d
nfs_reply(c17ca280,cde4fdc0,0,20,10) at nfs_reply+0x45
nfs_request(ce0b3c80,c0e5bf00,6,0,c16dcc80) at nfs_request+0x39a
nfs_readrpc(ce0b3c80,cde4fdc0,c16dcc80,c68a25b8,cdeb0cc0) at nfs_readrpc+0x538
nfs_doio(c68a25b8,c16dcc80,0) at nfs_doio+0x14b
nfssvc_iod(cc379740,cc379740,2,cde4ff80,c37d880) at nfssvc_iod+0x1c2
nfssvc(cc379740,cde4ff80,2,0,bfbffdd8) at nfssvc+0x6f
syscall2(2f,2f,2f,bfbffdd8,0) at syscall2+0x1f5
Xint0x80_syscall() at Xint0x80_syscall+0x25
db>
FYI, this machine is an HP Omnibook 6000. It has the "xl" ethernet
built in and this is the interface I'm using.
In this particular case, my entire sequence of events was:
boot
login
su
mount /usr/ports (from the nfs server)
cd /usr/ports/misc/instant-workstation
make clean
make install
<kaboom>
Just had another crash, panic this time was again supervisor read, page
not present, but current process was syncer. Traceback shows:
malloc(...)
initiate_write_inodeblock(...)
softdep_disk_io_initiation(...)
spec_strategy(...)
bwrite(...)
vop_stdbwrite(...)
vop_defaultop(...)
vfs_bio_awrite(...)
spec_fsync(...)
sched_sync(...)
fork_trampoline(...)
I do have soft updates turned on on the single / filesystem on this
box. Should I turn them off?
I would be happy to supply my kernel config file to anyone interested.
P.S. I also have the problem with these Omnibooks that if you tell them
to "reboot" (or "halt" followed by pressing any key to reboot), they
hang in a weird kind of half power on state. The display goes black
(the backlight turns off), but if you shine a flashlight on the LCD,
you can still see text there. The Power light on the front does not
turn off, and the machine still generates plenty of heat, but the cpu
cooling fan will not turn itself on. Any suggestions for that?
-jan-
--
Jan L. Peterson
<[EMAIL PROTECTED]>
To Unsubscribe: send mail to [EMAIL PROTECTED]
with "unsubscribe freebsd-stable" in the body of the message
