Bug 208304

Summary: lockdep annotation not right?
Product: [Fedora] Fedora Reporter: Gary Lindstrom <gplindstrom>
Component: kernelAssignee: Peter Zijlstra <pzijlstr>
Status: CLOSED RAWHIDE QA Contact: Brian Brock <bbrock>
Severity: medium Docs Contact:
Priority: medium    
Version: rawhideCC: davej, lwang, wtogami
Target Milestone: ---   
Target Release: ---   
Hardware: i686   
OS: Linux   
Whiteboard:
Fixed In Version: Doc Type: Bug Fix
Doc Text:
Story Points: ---
Clone Of: Environment:
Last Closed: 2006-11-08 18:22:18 UTC Type: ---
Regression: --- Mount Type: ---
Documentation: --- CRM:
Verified Versions: Category: ---
oVirt Team: --- RHEL 7.3 requirements from Atomic Host:
Cloudforms Team: --- Target Upstream Version:
Embargoed:
Bug Depends On:    
Bug Blocks: 202141, 209199    

Description Gary Lindstrom 2006-09-27 19:39:06 UTC
Description of problem:

Get the following on bootup - most likely in the scsi driver for my compaq/hp
integrated raid controllers (both 5i and 6i on DL360/380 G2 and G4, maybe others):

Sep 24 18:29:20 spool6 kernel: INFO: trying to register non-static key.
Sep 24 18:29:20 spool6 kernel: the code is fine but needs lockdep annotation.
Sep 24 18:29:20 spool6 kernel: turning off the locking correctness validator.
Sep 24 18:29:20 spool6 kernel:  [<c04051ee>] show_trace_log_lvl+0x58/0x171
Sep 24 18:29:20 spool6 kernel:  [<c0405802>] show_trace+0xd/0x10
Sep 24 18:29:20 spool6 kernel:  [<c040591b>] dump_stack+0x19/0x1b
Sep 24 18:29:20 spool6 kernel:  [<c043b366>] __lock_acquire+0xfd/0x99c
Sep 24 18:29:20 spool6 kernel:  [<c043c176>] lock_acquire+0x4b/0x6d
Sep 24 18:29:20 spool6 kernel:  [<c06155a4>] _spin_lock_irq+0x1f/0x2e
Sep 24 18:29:20 spool6 kernel:  [<c0613422>] wait_for_completion+0x29/0x9e
Sep 24 18:29:20 spool6 kernel:  [<f88cfc36>] sendcmd_withirq+0x173/0x289 [cciss]
Sep 24 18:29:20 spool6 kernel:  [<f88d0c9a>] cciss_read_capacity+0x2f/0xa6 [cciss]
Sep 24 18:29:20 spool6 kernel:  [<f88d2c32>] cciss_revalidate+0xaa/0xfa [cciss]
Sep 24 18:29:20 spool6 kernel:  [<c04a8380>] rescan_partitions+0x6b/0x1e7
Sep 24 18:29:20 spool6 kernel:  [<c047a5c8>] do_open+0x2e9/0x3e9
Sep 24 18:29:20 spool6 kernel:  [<c047a739>] blkdev_get+0x71/0x7c
Sep 24 18:29:20 spool6 kernel:  [<c04a82c3>] register_disk+0x115/0x167
Sep 24 18:29:20 spool6 kernel:  [<c04e099b>] add_disk+0x2e/0x3d
Sep 24 18:29:20 spool6 kernel:  [<f88d356a>] cciss_init_one+0x8e8/0xa5a [cciss]
Sep 24 18:29:20 spool6 kernel:  [<c04f4241>] pci_device_probe+0x39/0x5b
Sep 24 18:29:20 spool6 kernel:  [<c05552d6>] driver_probe_device+0x45/0x92
Sep 24 18:29:20 spool6 kernel:  [<c0555403>] __driver_attach+0x68/0x93
Sep 24 18:29:20 spool6 kernel:  [<c0554d4f>] bus_for_each_dev+0x3a/0x5f
Sep 24 18:29:20 spool6 kernel:  [<c0555231>] driver_attach+0x14/0x17
Sep 24 18:29:20 spool6 kernel:  [<c0554a26>] bus_add_driver+0x68/0x106
Sep 24 18:29:20 spool6 kernel:  [<c05556ac>] driver_register+0x78/0x7d
Sep 24 18:29:20 spool6 kernel:  [<c04f4392>] __pci_register_driver+0x4f/0x69
Sep 24 18:29:20 spool6 kernel:  [<f880401c>] cciss_init+0x1c/0x1e [cciss]
Sep 24 18:29:21 spool6 kernel:  [<c04426ce>] sys_init_module+0x16ad/0x1856
Sep 24 18:29:21 spool6 kernel:  [<c0403faf>] syscall_call+0x7/0xb
Sep 24 18:29:21 spool6 kernel: DWARF2 unwinder stuck at syscall_call+0x7/0xb
Sep 24 18:29:21 spool6 kernel: Leftover inexact backtrace:
Sep 24 18:29:21 spool6 kernel:  [<c0405802>] show_trace+0xd/0x10
Sep 24 18:29:21 spool6 kernel:  [<c040591b>] dump_stack+0x19/0x1b
Sep 24 18:29:21 spool6 kernel:  [<c043b366>] __lock_acquire+0xfd/0x99c
Sep 24 18:29:21 spool6 kernel:  [<c043c176>] lock_acquire+0x4b/0x6d
Sep 24 18:29:21 spool6 kernel:  [<c06155a4>] _spin_lock_irq+0x1f/0x2e
Sep 24 18:29:21 spool6 kernel:  [<c0613422>] wait_for_completion+0x29/0x9e
Sep 24 18:29:21 spool6 kernel:  [<f88cfc36>] sendcmd_withirq+0x173/0x289 [cciss]
Sep 24 18:29:21 spool6 kernel:  [<f88d0c9a>] cciss_read_capacity+0x2f/0xa6 [cciss]
Sep 24 18:29:21 spool6 kernel:  [<f88d2c32>] cciss_revalidate+0xaa/0xfa [cciss]
Sep 24 18:29:21 spool6 kernel:  [<c04a8380>] rescan_partitions+0x6b/0x1e7
Sep 24 18:29:21 spool6 kernel:  [<c047a5c8>] do_open+0x2e9/0x3e9
Sep 24 18:29:21 spool6 kernel:  [<c047a739>] blkdev_get+0x71/0x7c
Sep 24 18:29:21 spool6 kernel:  [<c04a82c3>] register_disk+0x115/0x167
Sep 24 18:29:21 spool6 kernel:  [<c04e099b>] add_disk+0x2e/0x3d
Sep 24 18:29:21 spool6 kernel:  [<f88d356a>] cciss_init_one+0x8e8/0xa5a [cciss]
Sep 24 18:29:21 spool6 kernel:  [<c04f4241>] pci_device_probe+0x39/0x5b
Sep 24 18:29:21 spool6 kernel:  [<c05552d6>] driver_probe_device+0x45/0x92
Sep 24 18:29:21 spool6 kernel:  [<c0555403>] __driver_attach+0x68/0x93
Sep 24 18:29:21 spool6 kernel:  [<c0554d4f>] bus_for_each_dev+0x3a/0x5f
Sep 24 18:29:21 spool6 kernel:  [<c0555231>] driver_attach+0x14/0x17
Sep 24 18:29:21 spool6 kernel:  [<c0554a26>] bus_add_driver+0x68/0x106
Sep 24 18:29:21 spool6 kernel:  [<c05556ac>] driver_register+0x78/0x7d
Sep 24 18:29:21 spool6 kernel:  [<c04f4392>] __pci_register_driver+0x4f/0x69
Sep 24 18:29:21 spool6 kernel:  [<f880401c>] cciss_init+0x1c/0x1e [cciss]
Sep 24 18:29:21 spool6 kernel:  [<c04426ce>] sys_init_module+0x16ad/0x1856
Sep 24 18:29:21 spool6 kernel:  [<c0403faf>] syscall_call+0x7/0xb

Version-Release number of selected component (if applicable):

All kernels tested with FC6 Test - maybe others.

How reproducible:

Occurs every boot...

Steps to Reproduce:
1. Boot server

Comment 1 Don Zickus 2006-09-28 22:31:34 UTC
in kernel-2.6.18-1.2707.el5.bz208456

Comment 2 Dave Jones 2006-09-28 22:54:10 UTC
will be in tomorrows rawhide.


Comment 3 Gary Lindstrom 2006-10-03 20:41:16 UTC
Not sure which kernel it was put in, but it seems to be fixed in
2.6.18-1.2724.fc6, at least on the DL380 G4...  Thanks!

Comment 4 Gary Lindstrom 2006-10-03 22:28:06 UTC
Hmmm... I just found this on another of my machines...  I am thinking this is
the Emulex FC HBA as listed:

Oct  3 14:54:56 virtual1b kernel: Emulex LightPulse Fibre Channel SCSI driver 8.1.9
Oct  3 14:54:56 virtual1b kernel: Copyright(c) 2004-2006 Emulex.  All rights
reserved.

I don't seem to see this on the machines with the Qlogic FC HBA.  Do you need
anything more from me to fix this?  Here is the error:

Oct  3 15:56:06 virtual1b kernel: scsi1 :  on PCI bus 07 device 20 irq 15
Oct  3 15:56:06 virtual1b kernel: INFO: trying to register non-static key.
Oct  3 15:56:06 virtual1b kernel: the code is fine but needs lockdep annotation.
Oct  3 15:56:06 virtual1b kernel: turning off the locking correctness validator.
Oct  3 15:56:06 virtual1b kernel:  [<c04051ed>] show_trace_log_lvl+0x58/0x16a
Oct  3 15:56:06 virtual1b kernel:  [<c04057fa>] show_trace+0xd/0x10
Oct  3 15:56:06 virtual1b kernel:  [<c0405913>] dump_stack+0x19/0x1b
Oct  3 15:56:06 virtual1b kernel:  [<c043b1e2>] __lock_acquire+0xf0/0x90d
Oct  3 15:56:06 virtual1b kernel:  [<c043bf70>] lock_acquire+0x4b/0x6b
Oct  3 15:56:06 virtual1b kernel:  [<c061472f>] _spin_lock_irqsave+0x22/0x32
Oct  3 15:56:07 virtual1b kernel:  [<c04363d3>] prepare_to_wait+0x17/0x4b
Oct  3 15:56:07 virtual1b kernel:  [<f89a24b6>] lpfc_do_work+0xdd/0xcc2 [lpfc]
Oct  3 15:56:07 virtual1b kernel:  [<c04361b9>] kthread+0xc3/0xf2
Oct  3 15:56:07 virtual1b kernel:  [<c0402005>] kernel_thread_helper+0x5/0xb
Oct  3 15:56:07 virtual1b kernel: DWARF2 unwinder stuck at
kernel_thread_helper+0x5/0xb
Oct  3 15:56:07 virtual1b kernel: Leftover inexact backtrace:
Oct  3 15:56:07 virtual1b kernel:  [<c04057fa>] show_trace+0xd/0x10
Oct  3 15:56:07 virtual1b kernel:  [<c0405913>] dump_stack+0x19/0x1b
Oct  3 15:56:07 virtual1b kernel:  [<c043b1e2>] __lock_acquire+0xf0/0x90d
Oct  3 15:56:07 virtual1b kernel:  [<c043bf70>] lock_acquire+0x4b/0x6b
Oct  3 15:56:07 virtual1b kernel:  [<c061472f>] _spin_lock_irqsave+0x22/0x32
Oct  3 15:56:07 virtual1b kernel:  [<c04363d3>] prepare_to_wait+0x17/0x4b
Oct  3 15:56:07 virtual1b kernel:  [<f89a24b6>] lpfc_do_work+0xdd/0xcc2 [lpfc]
Oct  3 15:56:07 virtual1b kernel:  [<c04361b9>] kthread+0xc3/0xf2
Oct  3 15:56:07 virtual1b kernel:  [<c0402005>] kernel_thread_helper+0x5/0xb


Comment 5 Dave Jones 2006-10-28 05:42:59 UTC
That's a new problem.  Peter ?


Comment 6 Peter Zijlstra 2006-11-06 10:41:01 UTC
Yeah it was:
 http://lkml.org/lkml/2006/10/28/75

Comment 7 Dave Jones 2006-11-08 18:22:18 UTC
ok, devel/ is rebased to .19rc5 which includes this, so I'll close this out.
FC6 has lockdep off, so this doesn't matter there.

I'll start doing rawhide builds again soon.