2.5.41-ac1: Debug: sleeping function called from illegal context at include/asm/semaphore.h:145

2.5.41-ac1: Debug: sleeping function called from illegal context at include/asm/semaphore.h:145

[Martin Dahl]

On my thinkpad a20m i get the following debug messages during boot:

At IDE initialization:

Debug: sleeping function called from illegal context at mm/slab.c:1374
Call Trace:
 [__kmem_cache_alloc+202/208] __kmem_cache_alloc+0xca/0xd0
 [blk_init_free_list+101/224] blk_init_free_list+0x65/0xe0
 [blk_init_queue+23/256] blk_init_queue+0x17/0x100
 [ide_init_queue+57/160] ide_init_queue+0x39/0xa0
 [do_ide_request+0/48] do_ide_request+0x0/0x30
 [init_irq+462/912] init_irq+0x1ce/0x390
 [ide_intr+0/384] ide_intr+0x0/0x180
 [hwif_init+216/608] hwif_init+0xd8/0x260
 [probe_hwif_init+36/112] probe_hwif_init+0x24/0x70
 [ide_setup_pci_device+80/128] ide_setup_pci_device+0x50/0x80
 [piix_init_one+54/64] piix_init_one+0x36/0x40
 [init+53/352] init+0x35/0x160
 [init+0/352] init+0x0/0x160
 [kernel_thread_helper+5/16] kernel_thread_helper+0x5/0x10

At USB hub initialization:

Debug: sleeping function called from illegal context at include/asm/semaphore.h:119
Call Trace:
 [<cc8485e2>] usb_hub_events+0x72/0x3c0 [usbcore]
 [<cc85300b>] .rodata.str1.1+0x42b/0x840 [usbcore]
 [schedule+402/768] schedule+0x192/0x300
 [reparent_to_init+228/384] reparent_to_init+0xe4/0x180
 [<cc848965>] usb_hub_thread+0x35/0xf0 [usbcore]
 [default_wake_function+0/64] default_wake_function+0x0/0x40
 [<cc85b1a0>] khubd_wait+0x0/0x8 [usbcore]
 [<cc85b1a0>] khubd_wait+0x0/0x8 [usbcore]
 [<cc848930>] usb_hub_thread+0x0/0xf0 [usbcore]
 [kernel_thread_helper+5/16] kernel_thread_helper+0x5/0x10

Also, i the following occurs when using the buttons for adjusting audio
volume or anything involving the Fn key.

Debug: sleeping function called from illegal context at include/asm/semaphore.h:145
Call Trace:
 [<c01a14f8>] acpi_os_wait_semaphore+0x1e8/0x240
 [<c01ce53d>] acpi_ut_acquire_mutex+0xdd/0x200
 [<c01b6334>] acpi_ex_enter_interpreter+0x44/0xa0
 [<c01ad33d>] acpi_acquire_global_lock+0x1d/0x60
 [<c01d5ed9>] acpi_ec_query+0x79/0x1c0
 [<c01d61bc>] acpi_ec_gpe_handler+0x8c/0xf0
 [<c01aa319>] acpi_ev_gpe_dispatch+0x229/0x2b0
 [<c01a9f6f>] acpi_ev_gpe_detect+0x14f/0x160
 [<c01aaf4e>] acpi_ev_sci_handler+0x9e/0xc0
 [<c01a0971>] acpi_irq+0x11/0x20
 [<c0108dc5>] handle_IRQ_event+0x45/0x70
 [<c01a0960>] acpi_irq+0x0/0x20
 [<c0108ff7>] do_IRQ+0x97/0x120
 [<c01053f0>] default_idle+0x0/0x40
 [<c01053f0>] default_idle+0x0/0x40
 [<c0107970>] common_interrupt+0x18/0x20
 [<c01053f0>] default_idle+0x0/0x40
 [<c01053f0>] default_idle+0x0/0x40
 [<c0105414>] default_idle+0x24/0x40
 [<c01054aa>] cpu_idle+0x3a/0x50
 [<c0105000>] stext+0x0/0x30

regards,
martin

-- 
Martin Dahl
dahlm@izno.net

Re: 2.5.41-ac1: Debug: sleeping function called from illegal context at include/asm/semaphore.h:145

[Andrew Morton]

Martin Dahl wrote:
> 
> ...
> Debug: sleeping function called from illegal context at include/asm/semaphore.h:145
> Call Trace:
>  [<c01a14f8>] acpi_os_wait_semaphore+0x1e8/0x240
>  [<c01ce53d>] acpi_ut_acquire_mutex+0xdd/0x200
>  [<c01b6334>] acpi_ex_enter_interpreter+0x44/0xa0
>  [<c01ad33d>] acpi_acquire_global_lock+0x1d/0x60
>  [<c01d5ed9>] acpi_ec_query+0x79/0x1c0
>  [<c01d61bc>] acpi_ec_gpe_handler+0x8c/0xf0
>  [<c01aa319>] acpi_ev_gpe_dispatch+0x229/0x2b0
>  [<c01a9f6f>] acpi_ev_gpe_detect+0x14f/0x160
>  [<c01aaf4e>] acpi_ev_sci_handler+0x9e/0xc0
>  [<c01a0971>] acpi_irq+0x11/0x20
>  [<c0108dc5>] handle_IRQ_event+0x45/0x70
>  [<c01a0960>] acpi_irq+0x0/0x20
>  [<c0108ff7>] do_IRQ+0x97/0x120
>  [<c01053f0>] default_idle+0x0/0x40
>  [<c01053f0>] default_idle+0x0/0x40
>  [<c0107970>] common_interrupt+0x18/0x20

ewww.

acpi_os_wait_semaphore() is doing down_interruptible() and
schedule_timeout() in hard interrupt context.

Re: 2.5.41-ac1: Debug: sleeping function called from illegal context at include/asm/semaphore.h:145

[Petr Konecny]

>>>>> Martin Dahl (Martin) said:

 Martin> On my thinkpad a20m i get the following debug messages during boot:
 Martin> At IDE initialization:

 Martin> Debug: sleeping function called from illegal context at mm/slab.c:1374
 Martin> Call Trace:
 Martin> [__kmem_cache_alloc+202/208] __kmem_cache_alloc+0xca/0xd0
 Martin> [blk_init_free_list+101/224] blk_init_free_list+0x65/0xe0
 Martin> [blk_init_queue+23/256] blk_init_queue+0x17/0x100
 Martin> [ide_init_queue+57/160] ide_init_queue+0x39/0xa0
 Martin> [do_ide_request+0/48] do_ide_request+0x0/0x30
 Martin> [init_irq+462/912] init_irq+0x1ce/0x390
 Martin> [ide_intr+0/384] ide_intr+0x0/0x180
 Martin> [hwif_init+216/608] hwif_init+0xd8/0x260
 Martin> [probe_hwif_init+36/112] probe_hwif_init+0x24/0x70
 Martin> [ide_setup_pci_device+80/128] ide_setup_pci_device+0x50/0x80
 Martin> [piix_init_one+54/64] piix_init_one+0x36/0x40
 Martin> [init+53/352] init+0x35/0x160
 Martin> [init+0/352] init+0x0/0x160
 Martin> [kernel_thread_helper+5/16] kernel_thread_helper+0x5/0x10
Same here on Dell Inspiron 5000.

Happens with CONFIG_PREEMPT=y. Disappears with CONFIG_PREEMPT=n.

                                                Petr