BUG: sleeping function called from invalid context at kernel/rwsem.c

Christian Kujau lists at nerdbynature.de
Sun May 20 04:51:37 EST 2007


Hi list,

I'm tracking vanilla 2.6-git on this iBook/G4 and sometimes, when 
closing the lid to put it to sleep and opening again to wake it up, the 
kernel barfs with:

[10606.436379] BUG: sleeping function called from invalid context at kernel/rwsem.c:20
[10606.436393] in_atomic():1, irqs_disabled():1
[10606.436399] Call Trace:
[10606.436405] [e79c9d10] [c0008d00] show_stack+0x3c/0x194 (unreliable)
[10606.436436] [e79c9d40] [c0027b28] __might_sleep+0xd4/0xe8
[10606.436463] [e79c9d50] [c00484b4] down_read+0x24/0x5c
[10606.436484] [e79c9d70] [c003ccec] __blocking_notifier_call_chain+0x38/0x74
[10606.436502] [e79c9d90] [c0167664] fb_notifier_call_chain+0x24/0x34
[10606.436531] [e79c9da0] [c0167cbc] fb_set_suspend+0x58/0x6c
[10606.436544] [e79c9dc0] [c0194968] radeonfb_pci_resume+0x1fc/0x3e8
[10606.436572] [e79c9de0] [c0194b78] radeonfb_early_resume+0x24/0x40
[10606.436585] [e79c9e00] [c001ac54] pmac_call_early_video_resume+0x2c/0x3c
[10606.436607] [e79c9e10] [c01f63dc] pmu_ioctl+0x7f8/0xc24
[10606.436629] [e79c9eb0] [c00924a0] do_ioctl+0x9c/0xa8
[10606.436645] [e79c9ed0] [c0092530] vfs_ioctl+0x84/0x490
[10606.436656] [e79c9f10] [c009297c] sys_ioctl+0x40/0x74
[10606.436668] [e79c9f40] [c0011930] ret_from_syscall+0x0/0x38
[10606.436690] --- Exception: c01 at 0xfee32a8
[10606.436722]     LR = 0xfee3240

This happened with 2.6.22-rc1 but has happened with 2.6.22-rc1-git7 
(reproduced just now). However, the suspend works and so does waking up. 
I just wonder if I should be worried by this message.

Please finde more details and config here: 
http://nerdbynature.de/bits/2.6.2x-powerpc/

Oh, speaking of "messages to worry about": lockdep says:

"142 out of 218 testcases failed, as expected." ...why is this 
"expected"?

Thanks,
Christian.
-- 
BOFH excuse #297:

Too many interrupts



More information about the Linuxppc-dev mailing list