APPLIED: [SRU][Trusty][PATCH 0/1] BUG: soft lockup - CPU#0 stuck for 23s! [kworker/0:1:1119] (LP: #1788817)

Kleber Souza kleber.souza at canonical.com
Wed Sep 5 07:06:38 UTC 2018


On 08/27/18 16:15, Juerg Haefliger wrote:
> This fixes stuck CPU warning on a server with an Intel S2600WFT mainboard.
> The patch looks more complicated than it is because of indentation changes.
> 'git diff -w' shows that it is actually fairly simple.
> 
> BugLink: https://bugs.launchpad.net/bugs/1788817
> 
> == SRU Justification ==
> 
> Booting Trusty with kernel 3.13.0-156-generic on a fairly new Intel box results in stuck CPU warnings:
> 
> [   33.551942] BUG: soft lockup - CPU#0 stuck for 23s! [kworker/0:1:1119]
> [   33.558560] Modules linked in: hid_generic ast(+) crct10dif_pclmul syscopyarea crc32_pclmul sysfillrect ghash_clmulni_intel sysimgblt aesni_intel i2c_algo_bit aes_x86_64 lrw ttm gf128mul glue_helper drm_kms_helper ablk_helper usbhid cryptd ahci drm hid libahci wmi
> [   33.583971] CPU: 0 PID: 1119 Comm: kworker/0:1 Not tainted 3.13.0-156-generic #206-Ubuntu
> [   33.592270] Hardware name: Intel Corporation S2600WFT/S2600WFT, BIOS SE5C620.86B.01.00.0294.101220161543 10/12/2016
> [   33.602836] Workqueue: events work_for_cpu_fn
> [   33.607351] task: ffff88084ddbc6b0 ti: ffff88084dd98000 task.ti: ffff88084dd98000
> [   33.614957] RIP: 0010:[<ffffffff81383bb2>]  [<ffffffff81383bb2>] ioread32+0x42/0x50
> [   33.622805] RSP: 0018:ffff88084dd99d18  EFLAGS: 00000292
> [   33.628201] RAX: 00000000ffffffff RBX: ffff880851265000 RCX: 00000000000028a6
> [   33.635429] RDX: ffffc90016bd0000 RSI: ffffc90016bd0000 RDI: ffffc90016bd0000
> [   33.642653] RBP: ffff88084dd99d40 R08: 0000000000000092 R09: 000000000000079e
> [   33.649877] R10: ffffffff813df570 R11: ffff88084dd99a46 R12: 0000000016bc0000
> [   33.657104] R13: ffffffffa01b2ab0 R14: ffff88084dd99cc8 R15: 0000000000000000
> [   33.664329] FS:  0000000000000000(0000) GS:ffff88085f200000(0000) knlGS:0000000000000000
> [   33.672541] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
> [   33.678380] CR2: 00007f3613be4000 CR3: 0000000001c0e000 CR4: 0000000000360770
> [   33.685605] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
> [   33.692829] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
> [   33.700052] Stack:
> [   33.702150]  ffffffffa01a89e0 ffff880851265000 0000000000000000 0000000000000000
> [   33.709948]  ffffffffa01b4100 ffff88084dd99d68 ffffffffa0073e88 ffff880851265000
> [   33.717751]  ffff880850dee000 ffff880850dee098 ffff88084dd99db8 ffffffffa0075d92
> [   33.725550] Call Trace:
> [   33.728085]  [<ffffffffa01a89e0>] ? ast_driver_load+0x320/0x540 [ast]
> [   33.734634]  [<ffffffffa0073e88>] drm_dev_register+0xa8/0x1f0 [drm]
> [   33.741000]  [<ffffffffa0075d92>] drm_get_pci_dev+0x92/0x140 [drm]
> [   33.747268]  [<ffffffffa01a81e5>] ast_pci_probe+0x15/0x20 [ast]
> [   33.753282]  [<ffffffff813b5f5a>] local_pci_probe+0x4a/0xb0
> [   33.758947]  [<ffffffff8108746a>] work_for_cpu_fn+0x1a/0x30
> [   33.764613]  [<ffffffff8108a5fe>] process_one_work+0x17e/0x480
> [   33.770536]  [<ffffffff8108b5ab>] worker_thread+0x29b/0x410
> [   33.776205]  [<ffffffff8108b310>] ? rescuer_thread+0x430/0x430
> [   33.782130]  [<ffffffff810922cb>] kthread+0xcb/0xf0
> [   33.787101]  [<ffffffff81092200>] ? kthread_create_on_node+0x1c0/0x1c0
> [   33.793721]  [<ffffffff8174758e>] ret_from_fork+0x6e/0xa0
> [   33.799211]  [<ffffffff81092200>] ? kthread_create_on_node+0x1c0/0x1c0
> [   33.805828] Code: 66 0f 1f 84 00 00 00 00 00 55 48 c7 c6 80 57 a9 81 48 89 e5 e8 f0 fe ff ff b8 ff ff ff ff 5d c3 66 0f 1f 84 00 00 00 00 00 8b 07 <c3> 66 66 66 66 2e 0f 1f 84 00 00 00 00 00 48 81 fe ff ff 03 00
> 
> == Fix ==
> 
> Backport commit 6c971c09f387 ("drm/ast: Fixed system hanged if disable P2A")
> 
> 
> == Regression Potential ==
> 
> Low. The patch only touches the ast driver and the modifications are fairly small. The patch is also in Xenial (via a stable upstream update) and received testing there.
> 
> == Test Case ==
> 
> Booted a patched kernel on HW to verify it comes up without stuck CPU warnings.
> 
> Signed-off-by: Juerg Haefliger <juergh at canonical.com>
> 
> 
> Y.C. Chen (1):
>   drm/ast: Fixed system hanged if disable P2A
> 
>  drivers/gpu/drm/ast/ast_drv.h  |   1 +
>  drivers/gpu/drm/ast/ast_main.c | 134 +++++++++++++++++----------------
>  drivers/gpu/drm/ast/ast_post.c |  10 ++-
>  3 files changed, 77 insertions(+), 68 deletions(-)
> 

Applied to trusty/master-next branch.

Thanks,
Kleber




More information about the kernel-team mailing list