Bug 214955 - running stress-ng cpu-online stressor triggers kworker preemptible bug
Summary: running stress-ng cpu-online stressor triggers kworker preemptible bug
Status: NEW
Alias: None
Product: Process Management
Classification: Unclassified
Component: Preemption (show other bugs)
Hardware: All Linux
: P1 normal
Assignee: Robert Love
URL:
Keywords:
Depends on:
Blocks:
 
Reported: 2021-11-05 18:57 UTC by Colin Ian King
Modified: 2021-11-05 18:59 UTC (History)
0 users

See Also:
Kernel Version: 5.15 - pre-rc1
Tree: Mainline
Regression: No


Attachments
dmesg (84.66 KB, text/plain)
2021-11-05 18:57 UTC, Colin Ian King
Details
config file (254.83 KB, text/plain)
2021-11-05 18:59 UTC, Colin Ian King
Details

Description Colin Ian King 2021-11-05 18:57:04 UTC
Created attachment 299463 [details]
dmesg

On a SMP system, the bug can be reproduced with 5.15 at commit d4439a1189f93d0ac1eaf0197db8e6b3e197d5c7

git clone https://github.com/ColinIanKing/stress-ng
cd stress-ng
make -j $(nproc)
sudo ./stress-ng --cpu-online 0 -t 15 --pathological 

We get the following issues in kernel log:

[ 2239.380169] kvm-clock: cpu 6, msr 79201181, secondary cpu clock
[ 2239.401652] ------------[ cut here ]------------
[ 2239.401658] DEAD callback error for CPU6
[ 2239.401721] WARNING: CPU: 3 PID: 1134 at kernel/cpu.c:1163 _cpu_down+0x20a/0x3a0
[ 2239.401856] Modules linked in: dm_multipath scsi_dh_rdac scsi_dh_emc scsi_dh_alua intel_rapl_msr intel_rapl_common kvm_intel snd_hda_codec_generic ledtrig_audio snd_hda_intel kvm snd_intel_dspcfg snd_intel_sdw_acpi snd_hda_codec rapl joydev input_leds snd_hda_core snd_hwdep snd_pcm snd_timer snd serio_raw soundcore qemu_fw_cfg mac_hid sch_fq_codel msr virtio_rng ip_tables x_tables autofs4 btrfs blake2b_generic zstd_compress raid10 raid456 async_raid6_recov async_memcpy async_pq async_xor async_tx xor raid6_pq libcrc32c raid1 raid0 multipath linear qxl hid_generic drm_ttm_helper ttm drm_kms_helper crct10dif_pclmul crc32_pclmul ghash_clmulni_intel virtio_net syscopyarea aesni_intel sysfillrect sysimgblt fb_sys_fops cec usbhid crypto_simd net_failover rc_core i2c_i801 ahci hid cryptd i2c_smbus psmouse libahci drm lpc_ich virtio_blk failover
[ 2239.402631] CPU: 3 PID: 1134 Comm: stress-ng Not tainted 5.15.0+ #1
[ 2239.402649] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 1.14.0-2 04/01/2014
[ 2239.402654] RIP: 0010:_cpu_down+0x20a/0x3a0
[ 2239.402664] Code: 8b 03 41 39 c6 7f 6f 39 45 c0 0f 8e 34 ff ff ff 3d 96 00 00 00 0f 84 be 00 00 00 44 89 ee 48 c7 c7 f3 38 9a 95 e8 26 a2 f9 ff <0f> 0b e9 13 ff ff ff e8 3a bd 52 ff e9 2a ff ff ff f0 48 0f b3 05
[ 2239.402744] RSP: 0018:ffffa30e008b7cc0 EFLAGS: 00010282
[ 2239.402755] RAX: 0000000000000000 RBX: ffff8ccb7bda0660 RCX: 0000000000000000
[ 2239.402760] RDX: 0000000000000001 RSI: ffffffff959b6099 RDI: 00000000ffffffff
[ 2239.402766] RBP: ffffa30e008b7d00 R08: 0000000000000000 R09: ffffa30e008b7ab0
[ 2239.402771] R10: ffffa30e008b7aa8 R11: ffffffff96356908 R12: 0000000000000000
[ 2239.402776] R13: 0000000000000006 R14: 0000000000000096 R15: 00000000ffffffea
[ 2239.402783] FS:  00007f5ee8713740(0000) GS:ffff8ccb7bcc0000(0000) knlGS:0000000000000000
[ 2239.402791] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2239.402796] CR2: 000055e786065c20 CR3: 000000010622e004 CR4: 0000000000370ee0
[ 2239.402811] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 2239.402816] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400

full log attached to the bug and config file

This does not occur with 5.13.
Comment 1 Colin Ian King 2021-11-05 18:59:23 UTC
Created attachment 299465 [details]
config file

Note You need to log in before you can comment on or make changes to this bug.