[LU-9106]  ASSERTION( vio->vui_tot_nrsegs >= vio->vui_iter->nr_segs ) Created: 10/Feb/17  Updated: 16/Jun/20  Resolved: 21/May/17

Status: Resolved
Project: Lustre
Component/s: None
Affects Version/s: None
Fix Version/s: None

Type: Bug Priority: Minor
Reporter: Andriy Skulysh Assignee: WC Triage
Resolution: Cannot Reproduce Votes: 0
Labels: None

Issue Links:
Duplicate
duplicates LU-7382 (vvp_io.c:573:vvp_io_update_iov()) AS... Reopened
Related
is related to LU-4257 parallel dds are slower than serial dds Resolved
Severity: 3
Rank (Obsolete): 9223372036854775807

 Description   

[ 4206.142919] LustreError: 663:0:(vvp_io.c:573:vvp_io_update_iov()) ASSERTION( vio->vui_tot_nrsegs >= vio->vui_iter->nr_segs ) failed: tot_nrsegs: 1, nrsegs: 2
[ 4206.149328] LustreError: 663:0:(vvp_io.c:573:vvp_io_update_iov()) LBUG
[ 4206.152489] Pid: 663, comm: nfsd
[ 4206.154854]
               Call Trace:
[ 4206.154877]  [<ffffffffa03f1853>] libcfs_debug_dumpstack+0x53/0x80 [libcfs]
[ 4206.154883]  [<ffffffffa03f1f95>] lbug_with_loc+0x45/0xc0 [libcfs]
[ 4206.154915]  [<ffffffffa0aab327>] vvp_io_rw_lock+0x6c7/0x770 [lustre]
[ 4206.154924]  [<ffffffffa0401d27>] ? libcfs_debug_msg+0x57/0x80 [libcfs]
[ 4206.154938]  [<ffffffffa0aab403>] vvp_io_write_lock+0x33/0x40 [lustre]
[ 4206.154984]  [<ffffffffa05707fc>] cl_io_lock+0x5c/0x410 [obdclass]
[ 4206.155014]  [<ffffffffa0570db2>] cl_io_loop+0x92/0x190 [obdclass]
[ 4206.155025]  [<ffffffffa0a46c91>] ll_file_io_generic+0x601/0xad0 [lustre]
[ 4206.155035]  [<ffffffffa0a4750d>] ll_file_aio_write+0x20d/0x870 [lustre]
[ 4206.155039]  [<ffffffff811dde39>] do_sync_readv_writev+0x79/0xd0
[ 4206.155040]  [<ffffffff811df40e>] do_readv_writev+0xce/0x260
[ 4206.155049]  [<ffffffffa0a47300>] ? ll_file_aio_write+0x0/0x870 [lustre]
[ 4206.155059]  [<ffffffffa0a47b70>] ? ll_file_write+0x0/0x1d0 [lustre]
[ 4206.155061]  [<ffffffff812a7383>] ? ima_get_action+0x23/0x30
[ 4206.155063]  [<ffffffff812a69ee>] ? process_measurement+0x8e/0x250
[ 4206.155064]  [<ffffffff812a6c13>] ? ima_file_check+0x63/0x1b0
[ 4206.155065]  [<ffffffff811df635>] vfs_writev+0x35/0x60
[ 4206.155076]  [<ffffffffa029923e>] nfsd_vfs_write.isra.13+0xae/0x380 [nfsd]
[ 4206.155081]  [<ffffffffa029cf0c>] nfsd_write+0xec/0x100 [nfsd]
[ 4206.155087]  [<ffffffffa02a3660>] nfsd3_proc_write+0xc0/0x160 [nfsd]
[ 4206.155091]  [<ffffffffa029512b>] nfsd_dispatch+0xbb/0x200 [nfsd]
[ 4206.155104]  [<ffffffffa025a183>] svc_process_common+0x453/0x6f0 [sunrpc]
[ 4206.155111]  [<ffffffffa025a523>] svc_process+0x103/0x170 [sunrpc]
[ 4206.155115]  [<ffffffffa0294aaf>] nfsd+0xdf/0x150 [nfsd]
[ 4206.155118]  [<ffffffffa02949d0>] ? nfsd+0x0/0x150 [nfsd]
[ 4206.155122]  [<ffffffff810a5acf>] kthread+0xcf/0xe0
[ 4206.155123]  [<ffffffff810a5a00>] ? kthread+0x0/0xe0
[ 4206.155127]  [<ffffffff81645bd8>] ret_from_fork+0x58/0x90
[ 4206.155128]  [<ffffffff810a5a00>] ? kthread+0x0/0xe0
[ 4206.155129]
[ 4206.155130] Kernel panic - not syncing: LBUG



 Comments   
Comment by Gerrit Updater [ 10/Feb/17 ]

Andriy Skulysh (andriy.skulysh@seagate.com) uploaded a new patch: https://review.whamcloud.com/25371
Subject: LU-9106 llite: ASSERTION(vio->vui_tot_nrsegs >= vio->vui_iter->nr_segs)
Project: fs/lustre-release
Branch: master
Current Patch Set: 1
Commit: 009f9e659e7af42b11fb5da315bab9136cc6cb1d

Comment by Dmitry Eremin (Inactive) [ 10/Feb/17 ]

This is not related to master. What version was used to get this assert?

Comment by Andriy Skulysh [ 10/Feb/17 ]

The assertion was hit before https://review.whamcloud.com/#/c/20256/  but u.normal.via_iter is still modified.

Comment by Dmitry Eremin (Inactive) [ 10/Feb/17 ]

Ok. This means not affected versions more than v2.8.54.
The u.normal.via_iter is just pointer. So, modification of it not affect the content.

Comment by Jinshan Xiong (Inactive) [ 10/Feb/17 ]

I agree with Dmitry - it's just a pointer.

Comment by Andreas Dilger [ 11/May/17 ]

Should this be closed as a duplicate of LU-7382, and the patch abandoned?

Comment by Andriy Skulysh [ 21/May/17 ]

LU-7382 is a different issue. This one reproduces with LU-7382 patch applied but on 2.7 branch. master doesn't need it now.

Comment by Peter Jones [ 21/May/17 ]

The LU branch relates to the community releases and the master branch. It sounds like this suggestion is something that should be tracked on the SEA project (which it is)

Generated at Sat Feb 10 02:23:17 UTC 2024 using Jira 9.4.14#940014-sha1:734e6822bbf0d45eff9af51f82432957f73aa32c.