Re: [2.5.75] S3 and S4

Andrew Morton (akpm@osdl.org)
Fri, 11 Jul 2003 13:48:33 -0700


Pavel Machek <pavel@suse.cz> wrote:
>
> > Suspending devices
> > Badness in local_bh_enable at kernel/softirq.c:113
> > Call Trace:
> > [<c0130078>] local_bh_enable+0x88/0x90
> > [<f0a44fa4>] e100_do_wol+0x14/0x60 [e100]
> > [<f0a461ee>] e100_suspend+0x3e/0xa0 [e100]
> > [<f0a461b0>] e100_suspend+0x0/0xa0 [e100]
> > [<c0212577>] pci_device_suspend+0x47/0x70
> > [<c029bc99>] device_suspend+0xd9/0x100
> > [<c023e047>] acpi_system_save_state+0x42/0x8c
> > [<c023e153>] acpi_suspend+0x5e/0xb3
> > [<c023e394>] acpi_system_write_sleep+0xe3/0x132
> > [<c0177de0>] filp_open+0x60/0x70
> > [<c017952d>] vfs_write+0xad/0x120
> > [<c017963f>] sys_write+0x3f/0x60
> > [<c010b10f>] syscall_call+0x7/0xb
> >
>
> If e100. You have the hardware...

No, it's acpi_system_save_state() illegally calling device_suspend() under
local_irq_disable().

> > S4 suspends fine, but it seems that it doesn't like preemption on
> > resume:
> >
> > ......[lots of dots]......[nosave c041c000]Enabling SEP on CPU 0
> > Freeing prev allocated pagedir
> > bad: scheduling while atomic!
> > Call Trace:
> > [<c012419e>] schedule+0x6ce/0x6e0
> > [<c02ded7b>] pci_read+0x3b/0x40
> > [<c01358e8>] schedule_timeout+0x88/0xe0
> > [<c02dedbc>] pci_write+0x3c/0x40
> > [<c0135850>] process_timeout+0x0/0x10
> > [<c020f9e4>] pci_set_power_state+0xe4/0x190
> > [<f0a46278>] e100_resume+0x28/0x70 [e100]
> > [<c02125cd>] pci_device_resume+0x2d/0x30
> > [<c029bd76>] device_resume+0xb6/0xd0
> > [<c014e60c>] drivers_resume+0x8c/0xa0
> > [<c014ea15>] do_magic_resume_2+0xc5/0x170
> > [<c011d410>] restore_processor_state+0x70/0x90
> > [<c011ff5f>] do_magic+0x11f/0x140
> > [<c014ef1d>] do_software_suspend+0x6d/0xa0
> > [<c023e3cc>] acpi_system_write_sleep+0x11b/0x132
> > [<c0177de0>] filp_open+0x60/0x70
> > [<c017952d>] vfs_write+0xad/0x120
> > [<c017963f>] sys_write+0x3f/0x60
> > [<c010b10f>] syscall_call+0x7/0xb
>
> That's e100 driver, again. Fix it.

No, e100_resume() looks OK to me. Something seems to have corrupted
preempt_count().

> > Fixing swap signatures... <3>bad: scheduling while atomic!
> > Call Trace:
> > [<c012419e>] schedule+0x6ce/0x6e0
> > [<c02a219d>] generic_make_request+0x17d/0x210
> > [<c01272d0>] autoremove_wake_function+0x0/0x50
> > [<c012641e>] io_schedule+0xe/0x20
> > [<c0150522>] wait_on_page_bit+0xa2/0xd0
> > [<c01272d0>] autoremove_wake_function+0x0/0x50
> > [<c01272d0>] autoremove_wake_function+0x0/0x50
> > [<c017194b>] swap_readpage+0x5b/0x80
> > [<c0171a2a>] rw_swap_page_sync+0xba/0x110
> > [<c014d8be>] mark_swapfiles+0x7e/0x1b0
> > [<c014ea35>] do_magic_resume_2+0xe5/0x170
> > [<c011d410>] restore_processor_state+0x70/0x90
> > [<c011ff5f>] do_magic+0x11f/0x140
> > [<c014ef1d>] do_software_suspend+0x6d/0xa0
> > [<c023e3cc>] acpi_system_write_sleep+0x11b/0x132
> > [<c0177de0>] filp_open+0x60/0x70
> > [<c017952d>] vfs_write+0xad/0x120
> > [<c017963f>] sys_write+0x3f/0x60
> > [<c010b10f>] syscall_call+0x7/0xb
>
> Ahha, this looks like generic problem. I'll probably take a look...

This will have the same cause as the earlier "scheduling while atomic"
problem.

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/