Re: CONFIG_INTEL_IDLE causing crashes?

From: J. Bruce Fields
Date: Tue Oct 16 2012 - 10:28:57 EST

On Fri, Oct 05, 2012 at 06:23:57PM -0400, bfields wrote:
> I find that I can reliably crash 3.6 by booting to a kernel with
> CONFIG_INTEL_IDLE set and running
>
> dd if=/dev/zero of=BIG

For any archives: see further discussion at

https://bugzilla.redhat.com/show_bug.cgi?id=715485

Upgrading the BIOS or booting with intel_idle.max_cstate=0 work around
this particular problem for me.

--b.

>
> If I turn off CONFIG_INTEL_IDLE there's no crash. (dd just hits ENOSPC
> eventually.)
>
> This isn't new--I originally saw it on 2.6.35.6. (An upgrade to F14
> failed, and by trial and error I found CONFIG_INTEL_IDLE seemed to make
> the difference.)
>
> I only see it on this one machine, which has an i7-870 and discrete ATI
> graphics. Another machine with an i3-530 and integrated graphics, but
> otherwise similar hardware (they were built from the same barebones), is
> not crashing.
>
> The first oops messages from one such crash follows. (For what it's
> worth, my memory is that they aren't always in ext4--but I don't have
> any others in front of me.)
>
> Any ideas? Let me know of anything else that might be useful.
>
> --b.
>
> [ 32.010103] stack segment: 0000 [#1] SMP
> [ 32.010135] Modules linked in: ipv6 cpufreq_ondemand acpi_cpufreq freq_table mperf uinput snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_seq iTCO_wdt snd_seq_device iTCO_vendor_support snd_pcm snd_timer snd microcode serio_raw lpc_ich i2c_i801 r8169 soundcore mii mfd_core snd_page_alloc usb_storage radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core
> [ 32.010401] CPU 1
> [ 32.010412] Pid: 1305, comm: dd Not tainted 3.6.0 #3 Shuttle Inc. SH55J/SH55J
> [ 32.010439] RIP: 0010:[<ffffffff811a87d8>] [<ffffffff811a87d8>] ext4_test_inode_state+0xd/0xf
> [ 32.010476] RSP: 0018:ffff880234f6d958 EFLAGS: 00010246
> [ 32.010495] RAX: 0000000000000000 RBX: ffff88022e35d570 RCX: ffff88022e168340
> [ 32.010519] RDX: ffff880234f6d9e8 RSI: 0000000000000001 RDI: ffff88022e35d570
> [ 32.010543] RBP: ffff880234f6d958 R08: ffff88022e10d168 R09: ffff88022e168340
> [ 32.010566] R10: ffff88022e10d168 R11: ffff8802315b1dc0 R12: ffff88022e35d4c0
> [ 32.010590] R13: ffff88022e35d4c0 R14: ffff88022a4c8900 R15: ffff88022e10d168
> [ 32.010614] FS: 00007f8e5f83d700(0000) GS:ffff88023fc40000(0000) knlGS:0000000000000000
> [ 32.010641] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b
> [ 32.010661] CR2: ffffffffff600400 CR3: 0000000234d43000 CR4: 00000000000007e0
> [ 32.010684] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [ 32.010708] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
> [ 32.010732] Process dd (pid: 1305, threadinfo ffff880234f6c000, task ffff8802315b16f0)
> [ 32.010758] Stack:
> [ 32.010767] ffff880234f6d9b8 ffffffff811a89c4 ffff880234f6d9e8 ffff88022e35d570
> [ 32.010806] ffff88022e168340 ffff880234f6d9e8 0000000000000286 ffff88022e35d570
> [ 32.010845] ffff88022f302800 ffff88022e35d4c0 ffffffff811ab66e ffff88022e10d168
> [ 32.010883] Call Trace:
> [ 32.010896] [<ffffffff811a89c4>] ext4_mark_iloc_dirty+0x73/0x5fb
> [ 32.010918] [<ffffffff811ab66e>] ? ext4_dirty_inode+0x33/0x4c
> [ 32.010939] [<ffffffff811ab608>] ext4_mark_inode_dirty+0x1bc/0x1ef
> [ 32.010962] [<ffffffff811e857b>] ? jbd2_journal_start+0x13/0x15
> [ 32.010984] [<ffffffff811ab66e>] ext4_dirty_inode+0x33/0x4c
> [ 32.011006] [<ffffffff8114c0aa>] __mark_inode_dirty+0x31/0x195
> [ 32.011027] [<ffffffff81152c20>] generic_write_end+0x67/0x79
> [ 32.011047] [<ffffffff811acf93>] ext4_da_write_end+0x1c4/0x245
> [ 32.011070] [<ffffffff810e5436>] ? iov_iter_copy_from_user_atomic+0x73/0xe3
> [ 32.011094] [<ffffffff810e5743>] generic_file_buffered_write+0x186/0x289
> [ 32.011117] [<ffffffff8114c0aa>] ? __mark_inode_dirty+0x31/0x195
> [ 32.011139] [<ffffffff810e693e>] __generic_file_aio_write+0x1c2/0x1f7
> [ 32.011162] [<ffffffff810e69eb>] generic_file_aio_write+0x78/0xdd
> [ 32.011185] [<ffffffff810e6979>] ? generic_file_aio_write+0x6/0xdd
> [ 32.011207] [<ffffffff811a6019>] ext4_file_write+0x3ec/0x43d
> [ 32.011229] [<ffffffff810713c0>] ? __dequeue_entity+0x33/0x38
> [ 32.011251] [<ffffffff8100e73b>] ? __switch_to+0x227/0x380
> [ 32.011271] [<ffffffff8107370e>] ? pick_next_task_fair+0xd4/0x126
> [ 32.011294] [<ffffffff8112b226>] do_sync_write+0x9d/0xda
> [ 32.011314] [<ffffffff8112bbc9>] vfs_write+0xae/0x10a
> [ 32.011333] [<ffffffff8112bce8>] sys_write+0x4a/0x71
> [ 32.011352] [<ffffffff814acca9>] system_call_fastpath+0x16/0x1b
> [ 32.011373] Code: e8 a7 fe ff ff 85 c0 74 10 48 89 df e8 a7 fe ff ff 83 e0 01 83 f0 01 eb 02 31 c0 5b c9 c3 55 8d 46 20 48 89 e5 0f a3 47 a0 19 c0 <c9> c3 55 48 89 e5 41 54 53 48 83 ec 10 66 66 66 66 90 4c 8b 26
> [ 32.011766] RIP [<ffffffff811a87d8>] ext4_test_inode_state+0xd/0xf
> [ 32.011792] RSP <ffff880234f6d958>
> [ 32.011909] ---[ end trace 9aa419bb9a9cdab0 ]---
> [ 32.011942] BUG: scheduling while atomic: dd/1305/0x10000001
> [ 32.011968] Modules linked in: ipv6 cpufreq_ondemand acpi_cpufreq freq_table mperf uinput snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_seq iTCO_wdt snd_seq_device iTCO_vendor_support snd_pcm snd_timer snd microcode serio_raw lpc_ich i2c_i801 r8169 soundcore mii mfd_core snd_page_alloc usb_storage radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core
> [ 32.012246] Pid: 1305, comm: dd Tainted: G D 3.6.0 #3
> [ 32.012266] Call Trace:
> [ 32.012277] <#SS> [<ffffffff8106ca92>] __schedule_bug+0x4d/0x59
> [ 32.012308] [<ffffffff814a4c6f>] __schedule+0xa2/0x569
> [ 32.012328] [<ffffffff8106feeb>] __cond_resched+0x2a/0x36
> [ 32.012349] [<ffffffff814a51b1>] _cond_resched+0x2c/0x37
> [ 32.012370] [<ffffffff814a485d>] down_read+0x16/0x28
> [ 32.012389] [<ffffffff810551df>] exit_signals+0x26/0x122
> [ 32.014303] [<ffffffff810655ac>] ? blocking_notifier_call_chain+0x14/0x16
> [ 32.016112] [<ffffffff81048164>] do_exit+0x113/0x7bb
> [ 32.017940] [<ffffffff814a3ed6>] ? printk+0x77/0x79
> [ 32.019769] [<ffffffff814a6fd7>] oops_end+0xc0/0xc8
> [ 32.021789] [<ffffffff81011c13>] die+0x5a/0x63
> [ 32.023537] [<ffffffff814a6be6>] do_trap+0x121/0x130
> [ 32.025270] [<ffffffff8100fcc5>] do_stack_segment+0x7a/0x9e
> [ 32.026952] [<ffffffff814a6415>] stack_segment+0x25/0x30
> [ 32.028933] [<ffffffff811a87d8>] ? ext4_test_inode_state+0xd/0xf
> [ 32.030593] <<EOE>> [<ffffffff811a89c4>] ext4_mark_iloc_dirty+0x73/0x5fb
> [ 32.032259] [<ffffffff811ab66e>] ? ext4_dirty_inode+0x33/0x4c
> [ 32.034125] [<ffffffff811ab608>] ext4_mark_inode_dirty+0x1bc/0x1ef
> [ 32.035763] [<ffffffff811e857b>] ? jbd2_journal_start+0x13/0x15
> [ 32.037417] [<ffffffff811ab66e>] ext4_dirty_inode+0x33/0x4c
> [ 32.039071] [<ffffffff8114c0aa>] __mark_inode_dirty+0x31/0x195
> [ 32.041029] [<ffffffff81152c20>] generic_write_end+0x67/0x79
> [ 32.042660] [<ffffffff811acf93>] ext4_da_write_end+0x1c4/0x245
> [ 32.044289] [<ffffffff810e5436>] ? iov_iter_copy_from_user_atomic+0x73/0xe3
> [ 32.046099] [<ffffffff810e5743>] generic_file_buffered_write+0x186/0x289
> [ 32.047672] [<ffffffff8114c0aa>] ? __mark_inode_dirty+0x31/0x195
> [ 32.049269] [<ffffffff810e693e>] __generic_file_aio_write+0x1c2/0x1f7
> [ 32.050862] [<ffffffff810e69eb>] generic_file_aio_write+0x78/0xdd
> [ 32.052430] [<ffffffff810e6979>] ? generic_file_aio_write+0x6/0xdd
> [ 32.054182] [<ffffffff811a6019>] ext4_file_write+0x3ec/0x43d
> [ 32.055723] [<ffffffff810713c0>] ? __dequeue_entity+0x33/0x38
> [ 32.057275] [<ffffffff8100e73b>] ? __switch_to+0x227/0x380
> [ 32.058825] [<ffffffff8107370e>] ? pick_next_task_fair+0xd4/0x126
> [ 32.060534] [<ffffffff8112b226>] do_sync_write+0x9d/0xda
> [ 32.062031] [<ffffffff8112bbc9>] vfs_write+0xae/0x10a
> [ 32.063548] [<ffffffff8112bce8>] sys_write+0x4a/0x71
> [ 32.065068] [<ffffffff814acca9>] system_call_fastpath+0x16/0x1b
> [ 32.066855] note: dd[1305] exited with preempt_count 1
> [ 32.068341] ------------[ cut here ]------------
> [ 32.069809] WARNING: at arch/x86/kernel/irq_64.c:70 handle_irq+0xf3/0x135()
> [ 32.071293] Hardware name: SH55J
> [ 32.072770] do_IRQ(): dd has overflown the kernel stack (cur:ffff880234f6c000,sp:ffff88023fc45cf8,irq stk top-bottom:ffff88023fc00080-ffff88023fc03fc0,exception stk top-bottom:ffff88023fc05080-ffff88023fc0b000)
> [ 32.075898] Modules linked in: ipv6 cpufreq_ondemand acpi_cpufreq freq_table mperf uinput snd_hda_codec_hdmi snd_hda_codec_realtek snd_hda_intel snd_hda_codec snd_hwdep snd_seq iTCO_wdt snd_seq_device iTCO_vendor_support snd_pcm snd_timer snd microcode serio_raw lpc_ich i2c_i801 r8169 soundcore mii mfd_core snd_page_alloc usb_storage radeon ttm drm_kms_helper drm i2c_algo_bit i2c_core
> [ 32.079345] Pid: 1305, comm: dd Tainted: G D W 3.6.0 #3
> [ 32.081004] Call Trace:
> [ 32.082677] <IRQ> [<ffffffff8104324a>] warn_slowpath_common+0x85/0x9d
> [ 32.084359] [<ffffffff81043305>] warn_slowpath_fmt+0x46/0x48
> [ 32.086039] [<ffffffff810707df>] ? sched_clock_idle_wakeup_event+0x12/0x14
> [ 32.087717] [<ffffffff81010cb6>] handle_irq+0xf3/0x135
> [ 32.089417] [<ffffffff814ae67d>] do_IRQ+0x4d/0xb4
> [ 32.091104] [<ffffffff814a622a>] common_interrupt+0x6a/0x6a
> [ 32.092784] <EOI>
> [ 32.092793] ---[ end trace 9aa419bb9a9cdab1 ]---
>
--
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/