Re: disabling IRQs cause nobody cares (incl. oops)

[Bill Davidsen]

H.Rosmanith (Kernel Mailing List) wrote:
> good evening,
>
> anyone knows what's the cause of this:
>
> Dec 30 01:13:45 burn kernel: irq 19: nobody cared!
> Dec 30 01:13:45 burn kernel:  [<c01318a3>] __report_bad_irq+0x33/0x98
> Dec 30 01:13:45 burn kernel:  [<c0131980>] note_interrupt+0x50/0x78 
> Dec 30 01:13:45 burn kernel:  [<c0131469>] __do_IRQ+0x10d/0x144
> Dec 30 01:13:45 burn kernel:  [<c0104cab>] do_IRQ+0x1b/0x28
> Dec 30 01:13:45 burn kernel:  [<c01036a2>] common_interrupt+0x1a/0x20
> Dec 30 01:13:45 burn kernel:  [<c02d007b>] packet_recvmsg+0xef/0x140
> Dec 30 01:13:45 burn kernel:  [<c0101049>] default_idle+0x29/0x38
> Dec 30 01:13:45 burn kernel:  [<c01010d9>] cpu_idle+0x39/0x58
> Dec 30 01:13:45 burn kernel:  [<c01002be>] rest_init+0x1e/0x20
> Dec 30 01:13:45 burn kernel:  [<c03b4994>] start_kernel+0x14c/0x154
> Dec 30 01:13:45 burn kernel: handlers:
> Dec 30 01:13:45 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
> Dec 30 01:13:45 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
> Dec 30 01:13:45 burn kernel: Disabling IRQ #19
> Dec 30 01:13:51 burn kernel: irq 17: nobody cared!
> Dec 30 01:13:51 burn kernel:  [<c01318a3>] __report_bad_irq+0x33/0x98
> Dec 30 01:13:51 burn kernel:  [<c0131980>] note_interrupt+0x50/0x78
> Dec 30 01:13:51 burn kernel:  [<c0131469>] __do_IRQ+0x10d/0x144
> Dec 30 01:13:51 burn kernel:  [<c0104cab>] do_IRQ+0x1b/0x28
> Dec 30 01:13:51 burn kernel:  [<c01036a2>] common_interrupt+0x1a/0x20
> Dec 30 01:13:51 burn kernel:  [<c02d007b>] packet_recvmsg+0xef/0x140
> Dec 30 01:13:51 burn kernel:  [<c0101049>] default_idle+0x29/0x38
> Dec 30 01:13:51 burn kernel:  [<c01010d9>] cpu_idle+0x39/0x58
> Dec 30 01:13:51 burn kernel:  [<c01002be>] rest_init+0x1e/0x20
> Dec 30 01:13:51 burn kernel:  [<c03b4994>] start_kernel+0x14c/0x154
> Dec 30 01:13:51 burn kernel: handlers:
> Dec 30 01:13:51 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
> Dec 30 01:13:51 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
> Dec 30 01:13:51 burn kernel: Disabling IRQ #17
>
>
>
> or even this:
>
>
>
> Dec 30 01:16:24 burn kernel: Disabling IRQ #19
> Dec 30 01:16:24 burn kernel: Unable to handle kernel NULL pointer dereference at virtual address 00000000
> Dec 30 01:16:24 burn kernel:  printing eip:
> Dec 30 01:16:24 burn kernel: c014fdb3
> Dec 30 01:16:24 burn kernel: *pde = 00000000
> Dec 30 01:16:24 burn kernel: Oops: 0002 [#1]
> Dec 30 01:16:24 burn kernel: PREEMPT SMP
> Dec 30 01:16:24 burn kernel: Modules linked in:
> Dec 30 01:16:24 burn kernel: CPU:    1
> Dec 30 01:16:24 burn kernel: EIP:    0060:[<c014fdb3>]    Not tainted VLI
> Dec 30 01:16:24 burn kernel: EFLAGS: 00010292   (2.6.10)
> Dec 30 01:16:24 burn kernel: EIP is at create_empty_buffers+0x1f/0x88
> Dec 30 01:16:24 burn kernel: eax: 00000000   ebx: c1321e20   ecx: 00000010   edx: 00000000
> Dec 30 01:16:24 burn kernel: esi: 00000000   edi: 00000000   ebp: c1321e20   esp: e2acdcd0
> Dec 30 01:16:24 burn kernel: ds: 007b   es: 007b   ss: 0068
> Dec 30 01:16:24 burn kernel: Process rip-cd (pid: 396, threadinfo=e2acc000 task=c1a16020)
> Dec 30 01:16:24 burn kernel: Stack: c1321e38 00010000 e2acdd7c c0150649 c1321e20 00010000 00000000 c1321e38
> Dec 30 01:16:25 burn kernel:        c1321e20 e2acdd7c 00000000 c192d180 00000046 00000001 c01df315 00000001
> Dec 30 01:16:25 burn kernel:        00000020 00010000 e93ee440 c01df593 dc93aa94 dc93aaa0 00000000 dc93a9ec
> Dec 30 01:16:25 burn kernel: Call Trace:
> Dec 30 01:16:25 burn kernel:  [<c0150649>] block_read_full_page+0x61/0x2dc
> Dec 30 01:16:25 burn kernel:  [<c01df315>] radix_tree_node_alloc+0x15/0x60
> Dec 30 01:16:25 burn kernel:  [<c01df593>] radix_tree_insert+0x6f/0xf0
> Dec 30 01:16:25 burn kernel:  [<c013240e>] add_to_page_cache+0x3e/0xb8
> Dec 30 01:16:25 burn kernel:  [<c0132469>] add_to_page_cache+0x99/0xb8
> Dec 30 01:16:25 burn kernel:  [<c01540af>] blkdev_readpage+0xf/0x14
> Dec 30 01:16:25 burn kernel:  [<c0153f5c>] blkdev_get_block+0x0/0x54
> Dec 30 01:16:25 burn kernel:  [<c0138e4e>] read_pages+0xba/0x13c
> Dec 30 01:16:25 burn kernel:  [<c013679b>] __alloc_pages+0x2ff/0x30c
> Dec 30 01:16:25 burn kernel:  [<c0139218>] do_page_cache_readahead+0x16c/0x194
> Dec 30 01:16:25 burn kernel:  [<c0139386>] page_cache_readahead+0x146/0x1c8
> Dec 30 01:16:25 burn kernel:  [<c0132aa6>] do_generic_mapping_read+0xe6/0x464
> Dec 30 01:16:25 burn kernel:  [<c01330c8>] __generic_file_aio_read+0x1d0/0x1f4
> Dec 30 01:16:25 burn kernel:  [<c0132e24>] file_read_actor+0x0/0xd4
> Dec 30 01:16:25 burn kernel:  [<c01331ed>] generic_file_read+0xad/0xc8
> Dec 30 01:16:25 burn kernel:  [<c014e5f0>] file_move+0x38/0x3c
> Dec 30 01:16:25 burn kernel:  [<c014cdaa>] dentry_open+0x12a/0x208
> Dec 30 01:16:25 burn kernel:  [<c014cc76>] filp_open+0x52/0x5c
> Dec 30 01:16:25 burn kernel:  [<c012b81c>] autoremove_wake_function+0x0/0x40
> Dec 30 01:16:25 burn kernel:  [<c01540ec>] block_llseek+0x0/0xc8
> Dec 30 01:16:25 burn kernel:  [<c014d75c>] vfs_read+0x9c/0xcc
> Dec 30 01:16:25 burn kernel:  [<c014d97c>] sys_read+0x40/0x6c
> Dec 30 01:16:25 burn kernel:  [<c0102d33>] syscall_call+0x7/0xb
> Dec 30 01:16:25 burn kernel: Code: 08 8b 44 24 10 5b 5e 5f 5d 59 c3 90 57 56 53 8b 5c 24 10 8b 44 24 14 8b 7c 24 18 6a 01 50 53 e8 c8 f6 ff ff 89 c6 89 f2 83 c4 0c <09> 3a 89 d0 8b 52 04 85 d2 75 f5 89 70 04 8b 43 10 83 c0 44 e8
> Dec 30 01:16:25 burn kernel:  <3>irq 19: nobody cared!
> Dec 30 01:16:25 burn kernel:  [<c01318a3>] __report_bad_irq+0x33/0x98
> Dec 30 01:16:25 burn kernel:  [<c0131980>] note_interrupt+0x50/0x78
> Dec 30 01:16:25 burn kernel:  [<c0131469>] __do_IRQ+0x10d/0x144
> Dec 30 01:16:25 burn kernel:  [<c0104cab>] do_IRQ+0x1b/0x28
> Dec 30 01:16:25 burn kernel:  [<c01036a2>] common_interrupt+0x1a/0x20
> Dec 30 01:16:25 burn kernel:  [<c02e007b>] interruptible_sleep_on_timeout+0x2b/0x90
> Dec 30 01:16:25 burn kernel:  [<c025007b>] e1000_tbi_adjust_stats+0x197/0x1d4
> Dec 30 01:16:25 burn kernel:  [<c013259e>] end_page_writeback+0xe/0x50
> Dec 30 01:16:25 burn kernel:  [<c016c1f4>] mpage_end_io_write+0x4c/0x64
> Dec 30 01:16:25 burn kernel:  [<c01529a1>] bio_endio+0x51/0x58
> Dec 30 01:16:25 burn kernel:  [<c0240b41>] __end_that_request_first+0xf1/0x1dc
> Dec 30 01:16:25 burn kernel:  [<c0240c5c>] end_that_request_chunk+0x14/0x18
> Dec 30 01:16:25 burn kernel:  [<c0270b44>] scsi_end_request+0x24/0xb4
> Dec 30 01:16:25 burn kernel:  [<c0270f30>] scsi_io_completion+0x1d4/0x40c
> Dec 30 01:16:25 burn kernel:  [<c0278386>] sd_rw_intr+0x20e/0x218
> Dec 30 01:16:25 burn kernel:  [<c026d39c>] scsi_finish_command+0x88/0x90
> Dec 30 01:16:25 burn kernel:  [<c026d2cd>] scsi_softirq+0xc9/0xe0
> Dec 30 01:16:25 burn kernel:  [<c011e25a>] __do_softirq+0x6a/0xd4
> Dec 30 01:16:25 burn kernel:  [<c011e2ec>] do_softirq+0x28/0x30
> Dec 30 01:16:25 burn kernel:  [<c01312f5>] irq_exit+0x2d/0x38
> Dec 30 01:16:25 burn kernel:  [<c0110f32>] smp_apic_timer_interrupt+0xce/0xd4
> Dec 30 01:16:25 burn kernel:  [<c0103730>] apic_timer_interrupt+0x1c/0x24
> Dec 30 01:16:25 burn kernel:  [<c011007b>] machine_real_restart+0x7b/0xac
> Dec 30 01:16:25 burn kernel:  [<c02e0c3d>] _spin_unlock_irq+0x9/0x20
> Dec 30 01:16:25 burn kernel:  [<c0115559>] finish_task_switch+0x35/0x78
> Dec 30 01:16:25 burn kernel:  [<c02dfe5e>] schedule+0xa4e/0xaa0
> Dec 30 01:16:25 burn kernel:  [<c02e034c>] schedule_timeout+0x8c/0xb4
> Dec 30 01:16:25 burn kernel:  [<c02e0354>] schedule_timeout+0x94/0xb4
> Dec 30 01:16:25 burn kernel:  [<c0121cd4>] process_timeout+0x0/0xc
> Dec 30 01:16:25 burn kernel:  [<c015e189>] do_select+0x295/0x2d8
> Dec 30 01:16:25 burn kernel:  [<c015ddb4>] __pollwait+0x0/0x9c
> Dec 30 01:16:25 burn kernel:  [<c015e52a>] sys_select+0x336/0x494
> Dec 30 01:16:25 burn kernel:  [<c0102d33>] syscall_call+0x7/0xb
> Dec 30 01:16:25 burn kernel: handlers:
> Dec 30 01:16:25 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
> Dec 30 01:16:25 burn kernel: [<c025eeb4>] (ide_intr+0x0/0x134)
>
>
> kernel is 2.6.10.
>
> anyone interested in tracking this down?
>
> it goes on for several minutes, filling the logfile with thousands
> of lines like that.

I reported the same thing, IRQ 18, no preempt. Try with acpi=off or 
acpi=ht if you need it.

The code to disable the IRQ isn't working, this is the 4th instance of 
this I've seen reported, as well as a few off-list "me too" notes.

--
bill davidsen <davidsen@xxxxxxx>
  CTO TMR Associates, Inc
  Doing interesting things with small computers since 1979
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/