Re: [PATCH 7/7] driver-core : convert semaphore to mutex in struct class

[Dave Young]

On Jan 22, 2008 5:16 AM, Jarek Poplawski <jarkao2@xxxxxxxxx> wrote:
> Dave Young wrote, On 01/21/2008 09:44 AM:
> ...
> > I applied it in my kernel, built and run without warnings, but it need
> > more testing.
> > I will be very glad to see the test result about this if you could, thanks.
>
> Bad news. (Alas I won't be able to check this today.)

Hi, thanks your effort. Now I think we should stop this thread and
waiting the class_device going away :)

Hope the iteration patches 1-6/7 could be applied.

>
> =============================================
> [ INFO: possible recursive locking detected ]
> 2.6.24-rc8 #5
> ---------------------------------------------
> modprobe/536 is trying to acquire lock:
>  (&cls->mutex/1){--..}, at: [<c0217e43>] class_device_add+0x2c6/0x322
>
> but task is already holding lock:
>  (&cls->mutex/1){--..}, at: [<c0217e43>] class_device_add+0x2c6/0x322
>
> other info that might help us debug this:
> 2 locks held by modprobe/536:
>  #0:  (&shost->scan_mutex){--..}, at: [<ce82795b>] __scsi_add_device+0x60/0xc9 [scsi_mod]
>  #1:  (&cls->mutex/1){--..}, at: [<c0217e43>] class_device_add+0x2c6/0x322
>
> stack backtrace:
> Pid: 536, comm: modprobe Not tainted 2.6.24-rc8 #5
>  [<c0138edf>] __lock_acquire+0x962/0x10d4
>  [<c028be37>] _spin_unlock_irqrestore+0x34/0x39
>  [<c015e693>] kfree+0x87/0xad
>  [<c01380cb>] trace_hardirqs_on+0xba/0x15b
>  [<c01396c2>] lock_acquire+0x71/0x8b
>  [<c0217e43>] class_device_add+0x2c6/0x322
>  [<c028a87f>] mutex_lock_nested+0x92/0x2bc
>  [<c0217e43>] class_device_add+0x2c6/0x322
>  [<c0217e43>] class_device_add+0x2c6/0x322
>  [<c0217e43>] class_device_add+0x2c6/0x322
>  [<c01afd42>] kobject_get+0xf/0x13
>  [<c01afe68>] kobject_init+0x29/0x38
>  [<c0217f2c>] class_device_create+0x7d/0x9e
>  [<cebfe5b4>] sg_add+0x144/0x39f [sg]
>  [<c0217e71>] class_device_add+0x2f4/0x322
>  [<ce8289f7>] scsi_sysfs_add_sdev+0x64/0x1ac [scsi_mod]
>  [<ce840ece>] ata_scsi_dev_config+0x14/0x91 [libata]
>  [<ce826b5c>] scsi_probe_and_add_lun+0x9c0/0x9e0 [scsi_mod]
>  [<ce82795b>] __scsi_add_device+0x60/0xc9 [scsi_mod]
>  [<ce8279c2>] __scsi_add_device+0xc7/0xc9 [scsi_mod]
>  [<ce841401>] ata_scsi_scan_host+0xd3/0x26a [libata]
>  [<ce83e7e6>] ata_host_register+0x205/0x280 [libata]
>  [<ce83f430>] ata_interrupt+0x0/0x200 [libata]
>  [<ce83e8ec>] ata_host_activate+0x8b/0xf2 [libata]
>  [<ceb95504>] svia_init_one+0x2e2/0x511 [sata_via]
>  [<c01ca42c>] pci_match_device+0xa5/0xb6
>  [<ceb95222>] svia_init_one+0x0/0x511 [sata_via]
>  [<c01ca4e9>] pci_device_probe+0x40/0x5f
>  [<c02173ae>] driver_probe_device+0x7c/0x175
>  [<c02175ed>] __driver_attach+0xa2/0xa4
>  [<c02168dd>] bus_for_each_dev+0x3c/0x5a
>  [<c0217263>] driver_attach+0x16/0x1a
>  [<c021754b>] __driver_attach+0x0/0xa4
>  [<c0216bea>] bus_add_driver+0x72/0x1c0
>  [<c01ca642>] __pci_register_driver+0x56/0x89
>  [<c013f11d>] sys_init_module+0xf8/0x1891
>  [<ce83901f>] ata_port_start+0x0/0x65 [libata]
>  [<c01380cb>] trace_hardirqs_on+0xba/0x15b
>  [<c0102936>] syscall_call+0x7/0xb
>  =======================
> scsi 2:0:0:0: Attached scsi generic sg1 type 0
>
> Regards,
> Jarek P.
>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/