[Bug 292767] Re: BUG: scheduling with irqs disabled

[Sébastien Valette]

OK, this is much better with linux-image-2.6.27-4-rt, but during a
stress-test (recording from firewire+compiling some code, all under
powersave mode), this happened :

Nov 17 00:28:26 argo kernel: [ 3526.243951] BUG: scheduling with irqs disabled: IRQ-5/0x00000000/1223
Nov 17 00:28:26 argo kernel: [ 3526.243969] caller is rt_spin_lock_slowlock+0x92/0x250
Nov 17 00:28:26 argo kernel: [ 3526.243976] Pid: 1223, comm: IRQ-5 Not tainted 2.6.27-4-rt #1
Nov 17 00:28:26 argo kernel: [ 3526.243981]  [<c0351566>] ? printk+0x18/0x1a
Nov 17 00:28:26 argo kernel: [ 3526.243990]  [<c0351f20>] schedule+0xf0/0x100
Nov 17 00:28:26 argo kernel: [ 3526.243998]  [<c03538f2>] rt_spin_lock_slowlock+0x92/0x250
Nov 17 00:28:26 argo kernel: [ 3526.244007]  [<c035408a>] __rt_spin_lock+0x4a/0x50
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0354098>] rt_spin_lock+0x8/0x10
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c011cf9a>] __wake_up+0x1a/0x50
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c017eed7>] ? page_address+0xb7/0xd0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c017f446>] kunmap_high+0x66/0xb0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c011a2a9>] kunmap+0x59/0x60
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c011a306>] kunmap_virt+0x56/0x70
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c023b06d>] sg_miter_stop+0x3d/0x90
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c023b0ef>] sg_miter_next+0x2f/0x110
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c023b264>] sg_copy_buffer+0x94/0xe0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c023b2ca>] sg_copy_to_buffer+0x1a/0x20
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f890a77a>] atapi_qc_complete+0xaa/0x180 [libata]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f8902aa5>] __ata_qc_complete+0x65/0xc0 [libata]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f8903c25>] ata_qc_complete+0x95/0x1d0 [libata]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f8903fe0>] ata_qc_complete_multiple+0x70/0xb0 [libata]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c011d42b>] ? hrtick_start_fair+0x8b/0xe0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f88dcb55>] ahci_port_intr+0xb5/0x170 [ahci]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<f88dcc7b>] ahci_interrupt+0x6b/0xd8 [ahci]
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0167c7c>] handle_IRQ_event+0x5c/0x100
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0104fb3>] ? common_interrupt+0x23/0x30
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c01689a2>] thread_simple_irq+0x72/0x100
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0168c8f>] do_hardirq+0xaf/0x110
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0168cf0>] ? do_irqd+0x0/0xa0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0168d52>] do_irqd+0x62/0xa0
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c013a90a>] kthread+0x3a/0x70
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c013a8d0>] ? kthread+0x0/0x70
Nov 17 00:28:26 argo kernel: [ 3526.244014]  [<c0105107>] kernel_thread_helper+0x7/0x10
Nov 17 00:28:26 argo kernel: [ 3526.244014]  =======================
Nov 17 00:28:26 argo kernel: [ 3526.244490] ------------[ cut here ]------------

in the following messages, there seems to be a slaughter of processes
due to out-of-memory (and the windows eventually disappear one by one)
and the computer is finally irresponsive. This event also happened
during a real use case. I also join the full kern.log.


** Attachment added: "kernel messages"
   http://launchpadlibrarian.net/19707253/kern.log

-- 
BUG: scheduling with irqs disabled
https://bugs.launchpad.net/bugs/292767
You received this bug notification because you are a member of Ubuntu
Bugs, which is subscribed to Ubuntu.

-- 
ubuntu-bugs mailing list
ubuntu-bugs at lists.ubuntu.com
https://lists.ubuntu.com/mailman/listinfo/ubuntu-bugs