Subject: kern/28897: reproducible diagnostic assertion failure in kern_synch.c
To: None <kern-bug-people@netbsd.org, gnats-admin@netbsd.org,>
From: Denis Lagno <dlagno@smtp.ru>
List: netbsd-bugs
Date: 01/07/2005 22:06:00
>Number:         28897
>Category:       kern
>Synopsis:       reproducible diagnostic assertion failure in kern_synch.c
>Confidential:   no
>Severity:       critical
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Fri Jan 07 22:06:00 +0000 2005
>Originator:     Denis Lagno
>Release:        NetBSD 2.99.12
>Organization:
Brigada
>Environment:
System: NetBSD chup.gado 2.99.12 NetBSD 2.99.12 (CHUP) #0: Fri Jan  7 05:24:22 MSK 2005 dina@flam.gado:/volatile/worksrc/netbsd-current/src/sys/arch/i386/compile/CHUP i386
Architecture: i386
Machine: i386
>Description:
Kernel panics with:
panic: kernel diagnostic assertion "p != NULL" failed: file ".../sys/kern/kern_synch.c", line 413

backtrace:

cpu_Debugger
panic
__assert
ltsleep
uvm_wait
uvm_kmapent_alloc
uvm_mapent_reserve
uvm_unmap
free
cgdiodone
biodone
wddone
wdc_ata_bio_done
wdc_ata_bio_intr
wdcintr
Xintr_legacy14
cpu_switch
ltsleep
uvm_scheduler
setrootfstime

>How-To-Repeat:
This panic is pretty reproducible on my laptop.
I have pristine ffsv1 partition residing on cgd device.
If I launch untarring of big tarball stuffed with a lot of
small files in it, kernel panics in a 20-60 seconds.

It first happened several days back, I rebuild from freshest
sources and it still panics.

I suspect it may be somehow related with cgd -- because when I
tried to untar the same tarball into non-cgd filesystem
no panic raised.
>Fix:
N/A

>Unformatted:
 X-send-pr-version: 3.95