Login to iSCSI target results in Kernel oops inside Xen-Guest-Domain
andy...@gmail.com
I tried to use the iSCSI initiator from inside a xen-guest running on
a amd64 - Server (kernel 2.6.18-4-xen-amd64 on the Host (dom0) as well
as inside the guest) .
I compiled the current semi-stable (2.0.-865.8) against debian-kernel-
headers I can mount the devices (iscsitarget-0.4.15) from real
physical machines, but if I try to login from inside a VM I get the
following:
If you'd like to have access to a Xen vm for testing - no problem -
just ask...
Cheers
Andreas Balg
-------------------------SNIPP---------------------
Unable to handle kernel NULL pointer dereference at 0000000000000078
RIP:
[<ffffffff881473a2>] :scsi_mod:scsi_calculate_bounce_limit+0x15/0x49
PGD 1d814067 PUD 1d805067 PMD 0
Oops: 0000 [1] SMP
CPU 0
Modules linked in: crc32c libcrc32c rdma_cm ib_addr ib_cm ib_sa ib_mad
ib_core iscsi_tcp libiscsi scsi_transport_iscscpudp iptable_filter
ip_tables x_tables ipv6 8250 serial_core ext3 jbd mbcache dm_mirror
dm_snapshot dm_mod
Pid: 12741, comm: iscsid Not tainted 2.6.18-4-xen-amd64 #1
RIP: e030:[<ffffffff881473a2>]
[<ffffffff881473a2>] :scsi_mod:scsi_calculate_bounce_limit+0x15/0x49
RSP: e02b:ffff88001faf1c30 EFLAGS: 00010246
RAX: 0000000000000000 RBX: ffff88001f3a6048 RCX: 0000000000000071
RDX: 0000000000000067 RSI: 000000000000ffff RDI: ffff880002301000
RBP: ffff880002301000 R08: ffff88001f7c5000 R09: 0000000000000014
R10: ffffffff881488d0 R11: 0000000000000048 R12: ffff88001da25828
R13: ffff880002301000 R14: 0000000000000000 R15: ffff88001da25800
FS: 00002b582f45d6d0(0000) GS:ffffffff804c4000(0000) knlGS:
0000000000000000
CS: e033 DS: 0000 ES: 0000
Process iscsid (pid: 12741, threadinfo ffff88001faf0000, task
ffff88001f3ed7f0)
Stack: ffffffff881479f2 ffff88001df55000 ffff88001df55000
ffff88001da25800
ffffffff88149641 ffff880002301168 00000001881431fe
ffff880002301000
0000000000000000 0000000000000000
Call Trace:
[<ffffffff881479f2>] :scsi_mod:scsi_alloc_queue+0x65/0xb6
[<ffffffff88149641>] :scsi_mod:scsi_alloc_sdev+0x12e/0x1d2
[<ffffffff88149870>] :scsi_mod:scsi_probe_and_add_lun+0x10d/0x9c6
[<ffffffff8814a46c>] :scsi_mod:scsi_alloc_target+0x21e/0x327
[<ffffffff8814a6b8>] :scsi_mod:__scsi_scan_target+0xc3/0x5e7
[<ffffffff8020ba57>] do_page_fault+0xdce/0x112f
[<ffffffff8024c587>] sscanf+0x49/0x51
[<ffffffff8814ae86>] :scsi_mod:scsi_scan_target+0x6c/0x83
[<ffffffff8816a7a0>] :scsi_transport_iscsi:iscsi_user_scan+0x65/0x94
[<ffffffff8814b6f9>] :scsi_mod:store_scan+0x9b/0xc5
[<ffffffff802dad63>] sysfs_write_file+0xb4/0xe3
[<ffffffff80216ab9>] vfs_write+0xce/0x174
[<ffffffff802172dc>] sys_write+0x45/0x6e
[<ffffffff8025c72e>] system_call+0x86/0x8b
[<ffffffff8025c6a8>] system_call+0x0/0x8b
Code: 83 78 78 00 75 0c 48 8b 05 59 6a 40 f8 48 c1 e0 0c c3 48 8b
RIP [<ffffffff881473a2>] :scsi_mod:scsi_calculate_bounce_limit
+0x15/0x49
RSP <ffff88001faf1c30>
CR2: 0000000000000078
Message from syslogd@web01 at Thu Aug 2 18:04:09 2007 ...
web01 kernel: Oops: 0000 [1] SMP
Message from syslogd@web01 at Thu Aug 2 18:04:09 2007 ...
web01 kernel: CR2: 0000000000000078
Mike Christie
This is a bug with xen. I believe there is a updated kernel with that fixed.