2002-10-09 15:56:03

by Steven Cole

[permalink] [raw]
Subject: 2.5.41-bk2 yet another $SLEEPING at mm/slab.c:1374 report

Greetings all,

I looked through recent archives and although I saw several reports of
"mm/slabc:1374" debug messages, I didn't see this particular traceback.
Apologies in advance if this is a duplicate report.
Kernel is 2.5.41-bk2.

Steven

Debug: sleeping function called from illegal context at mm/slab.c:1374
Call Trace:
[<c0119c86>] __might_sleep+0x56/0x5d
[<c0135f75>] kmalloc+0x65/0x1b0
[<c0120478>] __request_region+0x18/0xc0
[<c021897f>] eata2x_detect+0x13f/0xd10
[<c01359bc>] kmem_cache_grow+0xfc/0x320
[<c010eeea>] timer_interrupt+0xfa/0x1b0
[<c010adca>] handle_IRQ_event+0x3a/0x60
[<c02064f4>] ahc_linux_alloc_device+0x14/0x70
[<c020571c>] ahc_linux_queue+0x16c/0x1a0
[<c01098e4>] common_interrupt+0x18/0x20
[<c01eeb80>] scsi_done+0x0/0x80
[<c01cb46e>] elv_queue_empty+0xe/0x20
[<c0118081>] schedule+0x321/0x3b0
[<c01cb46e>] elv_queue_empty+0xe/0x20
[<c01f5400>] scsi_request_fn+0x1f0/0x4c0
[<c01f4b4d>] scsi_queue_next_request+0x5d/0x170
[<c01ee26d>] scsi_release_command+0x27d/0x290
[<c0118160>] default_wake_function+0x0/0x40
[<c01ee612>] scsi_wait_req+0xa2/0xc0
[<c013610a>] kmem_cache_free+0x4a/0x80
[<c01cb302>] elevator_exit+0x12/0x20
[<c01ccd4f>] blk_cleanup_queue+0x4f/0x60
[<c01f73b6>] scan_scsis+0x96/0xa0
[<c01efa28>] scsi_register_host+0x48/0x330
[<c01b3d83>] put_bus+0x13/0x57
[<c01050b1>] init+0x51/0x1d0
[<c0105060>] init+0x0/0x1d0
[<c01070c9>] kernel_thread_helper+0x5/0xc





2002-10-09 17:51:29

by Andrew Morton

[permalink] [raw]
Subject: Re: 2.5.41-bk2 yet another $SLEEPING at mm/slab.c:1374 report

Steven Cole wrote:
>
> ...
> Debug: sleeping function called from illegal context at mm/slab.c:1374
> Call Trace:
> [<c0119c86>] __might_sleep+0x56/0x5d
> [<c0135f75>] kmalloc+0x65/0x1b0
> [<c0120478>] __request_region+0x18/0xc0
> [<c021897f>] eata2x_detect+0x13f/0xd10

Ton of stack gunk there. This is eata2x_detect() calling
port_detect() under driver_lock.