[LU-18335] interop: conf-sanity test_136: RIP: 0010:ls_device_get+0x1e3/0x3b0 [obdclass]

• Type: Bug
• Resolution: Fixed
• Priority: Minor
• Fix Version/s: Lustre 2.16.0
• Affects Version/s: Lustre 2.16.0
• Labels: mtg
• Severity: 3
• Assignee: Andreas Dilger
• Reporter: Maloo

Description

This issue was created by maloo for jianyu <[e-mail]>

This issue relates to the following test suite run: https://testing.whamcloud.com/test_sets/31e964e5-1404-4a3c-b868-30ad5dd3fcc6

test_136 failed with the following error:

mount lustre  on /mnt/lustre.....
Starting client: trevis-81vm1.trevis.whamcloud.com:  -o user_xattr,flock 10.240.42.201@tcp:/lustre /mnt/lustre
CMD: trevis-81vm1.trevis.whamcloud.com mkdir -p /mnt/lustre
CMD: trevis-81vm1.trevis.whamcloud.com mount -t lustre -o user_xattr,flock 10.240.42.201@tcp:/lustre /mnt/lustre
CMD: trevis-81vm7 /usr/sbin/lctl attach echo_client ec ec_uuid
CMD: trevis-81vm7 /usr/sbin/lctl --device ec setup lustre-MDT0001 mdt

trevis-81vm7 crashed during conf-sanity test_136

Test session details:
clients: https://build.whamcloud.com/job/lustre-master/4581 - 4.18.0-553.16.1.el8_10.x86_64
servers: https://build.whamcloud.com/job/lustre-b2_15/94 - 4.18.0-553.5.1.el8_lustre.x86_64

Lustre: DEBUG MARKER: /usr/sbin/lctl attach echo_client ec ec_uuid
Lustre: DEBUG MARKER: /usr/sbin/lctl --device ec setup lustre-MDT0001 mdt
BUG: unable to handle kernel NULL pointer dereference at 0000000000000018
Oops: 0000 [#1] SMP PTI
CPU: 0 PID: 1409966 Comm: lctl 4.18.0-553.5.1.el8_lustre.x86_64 #1
RIP: 0010:ls_device_get+0x1e3/0x3b0 [obdclass]
Call Trace:
 local_oid_storage_init+0xb8/0x16f0 [obdclass]
 echo_device_alloc+0x6d0/0x1930 [obdecho]
 obd_setup+0x119/0x2e0 [obdclass]
 class_setup+0x587/0x790 [obdclass]
 class_process_config+0xfc8/0x2080 [obdclass]
 class_handle_ioctl+0x1b0/0x1e40 [obdclass]
 obd_class_ioctl+0x13b/0x190 [obdclass]
 do_vfs_ioctl+0xa4/0x690
 ksys_ioctl+0x64/0xa0
 __x64_sys_ioctl+0x16/0x20
 do_syscall_64+0x5b/0x1b0

VVVVVVV DO NOT REMOVE LINES BELOW, Added by Maloo for auto-association VVVVVVV
conf-sanity test_136 - trevis-81vm7 crashed during conf-sanity test_136

Issue Links

• is related to: LU-15784 md echo should don't panic with remote object

Comments

[Peter Jones]

Merged for 2.16

[Gerrit Updater]

"Oleg Drokin <[e-mail]>" merged in patch https://review.whamcloud.com/c/fs/lustre-release/+/56638/
Subject: LU-18335 tests: skip conf-sanity/136 in interop
Project: fs/lustre-release
Branch: master
Current Patch Set:
Commit: cb1e0174f7c7a2b0c62067af1cdfa09e8a7cc63e

[Jian Yu]

No problem, Andreas. Thank you for pushing a patch.

[Andreas Dilger]

yujian, sorry, I didn't see you had already assigned this to yourself. I've already pushed a patch to skip this subtest.

[Andreas Dilger]

This has only crashed 3x and only in the most recent build (2024-10-03). It looks like this is an interop testing issue, since test_136 was added in patch https://review.whamcloud.com/47147 "LU-15784 obdecho: don't panic with run on second mdt" to verify a bug that is crashing the MDS.

[Gerrit Updater]

"Andreas Dilger <[e-mail]>" uploaded a new patch: https://review.whamcloud.com/c/fs/lustre-release/+/56638
Subject: LU-18335 tests: skip conf-sanity/136 in interop
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: fd1bb8bb6bc640a28b64c02f63ff1442efd13ed8

[Andreas Dilger]

It looks like this is for conf-sanity test_136. There is also a crash in sanity test_136...