kern/47486: XEN3_DOMU diag assert fail

[ci4ic4%gmail.com@localhost]

>Number:         47486
>Category:       kern
>Synopsis:       XEN3_DOMU diag assert fail
>Confidential:   no
>Severity:       serious
>Priority:       medium
>Responsible:    kern-bug-people
>State:          open
>Class:          sw-bug
>Submitter-Id:   net
>Arrival-Date:   Tue Jan 22 12:20:00 +0000 2013
>Originator:     Chavdar Ivanov
>Release:        6.99.16 amd64
>Organization:
Delcam Plc
>Environment:
support6# uname -a 
NetBSD support6.delcam.local 6.99.16 NetBSD 6.99.16 (XEN3_DOMU) #0: Mon Jan 21 
21:09:50 UTC 2013  
builds%b6.netbsd.org@localhost:/home/builds/ab/HEAD/amd64/201301211830Z-obj/home/builds/ab/HEAD/src/sys/arch/amd64/compile/XEN3_DOMU
 amd64
>Description:
Somewhere after the 13th of January 2012 my XEN3_DOMU fully pv host running 
under XenServer 6.1 started to panic; the first sign of the problem is fault 
with /usr/sbin/sshd as follows: 

# gdb /usr/sbin/sshd 
...
Reading symbols from /usr/sbin/sshd...(no debugging symbols found)...done.
(gdb) run 
Starting program: /usr/sbin/sshd 

Program received signal SIGSEGV, Segmentation fault.
0x00007f7ff434cddc in ?? () from /usr/lib/libcrypto.so.8
(gdb) bt
#0  0x00007f7ff434cddc in ?? () from /usr/lib/libcrypto.so.8
#1  0x00007f7ff432bbe6 in ?? () from /usr/lib/libcrypto.so.8
#2  0x00007f7ff435c7ae in BIO_set () from /usr/lib/libcrypto.so.8
#3  0x00007f7ff435c802 in ?? () from /usr/lib/libcrypto.so.8
#4  0x00007f7ff432bf6f in BIO_new_mem_buf () from /usr/lib/libcrypto.so.8
#5  0x00007f7ff4817b92 in ?? () from /usr/lib/libssh.so.20
#6  0x00007f7ff4817e26 in ?? () from /usr/lib/libssh.so.20
#7  0x00007f7ff4818bef in key_parse_private () from /usr/lib/libssh.so.20
#8  0x00007f7ff4818c80 in key_load_private () from /usr/lib/libssh.so.20
#9  0x000000000040de3d in main ()

Further I get a panic:

1]   Segmentation fault (core dumped) /usr/sbin/sshd
/etc/rc.d/sshd exited with code 1
Starting smbd.
panic: kernel diagnostic assertion "uvm_page_locked_p(pg)" failed: file 
"/home/builds/ab/HEAD/src/sys/arch/x86/x86/pmap.c", line 3306 
cpu3: Begin traceback...
kern_assert() at netbsd:kern_assert+0xae
pmap_remove_pte() at netbsd:pmap_remove_pte+0x312
pmap_remove() at netbsd:pmap_remove+0x1e9
ubc_alloc() at netbsd:ubc_alloc+0x5c3
ubc_uiomove() at netbsd:ubc_uiomove+0xe1
ffs_write() at netbsd:ffs_write+0x2c5
VOP_WRITE() at netbsd:VOP_WRITE+0x37
vn_write() at netbsd:vn_write+0xf9
dofilewrite() at netbsd:dofilewrite+0x7d
sys_write() at netbsd:sys_write+0x62
syscall() at netbsd:syscall+0x94
--- syscall (number 4) ---
7f7ff7438a7a:
cpu3: End traceback...

dumping to dev 142,17 (offset=4186991, size=0): not possible
rebooting...

The system works fine with a kernel from 07/01/2013 (I've lost subsequent 
kernels after a panic destroyed my root disk and I had to reinstall from 
scratch). The userland is current as of today (22/01/2013). 

 
>How-To-Repeat:
- Clean installation of NetBSD-current amd64 under XenServer 6.1
- switch system to PV by running 

xe vm-param-set  uuid=db74941f-2ccc-9c01-ecb2-bc2e26ffce19 HVM-boot-policy=
xe vm-param-set  uuid=db74941f-2ccc-9c01-ecb2-bc2e26ffce19 
PV-kernel=/boot/guest/netxen
xe vm-param-set  uuid=db74941f-2ccc-9c01-ecb2-bc2e26ffce19 
PV-args=boot_verbose=1,vfs.root.mountfrom=ffs:/dev/xbd1s1a,kern.hz=100

and modifying the /etc/fstab and ifconfig.xennet0 files
- copy the XEN3_DOMU kernel to /boot/guest/netxen on the XenServer 

>Fix:
None known to me at this time.