FS#44323 - [linux] nfsv4.1-svc breaks suspend

Attached to Project: Arch Linux
Opened by Bjoern Franke (bjo) - Tuesday, 24 March 2015, 16:48 GMT
Last edited by Doug Newgard (Scimmia) - Sunday, 15 October 2017, 14:26 GMT
Task Type Bug Report
Category Kernel
Status Closed
Assigned To Tobias Powalowski (tpowa)
Architecture All
Severity Medium
Priority Normal
Reported Version
Due in Version Undecided
Due Date Undecided
Percent Complete 100%
Votes 6
Private No

Details

nfs4.1-svc hangs and breaks suspend:
[ 242.963957] PM: Syncing filesystems ... done.
[ 243.535088] PM: Preparing system for mem sleep
[ 243.730776] Freezing user space processes ... (elapsed 0.002 seconds) done.
[ 243.733034] Freezing remaining freezable tasks ...
[ 263.736906] Freezing of tasks failed after 20.007 seconds (1 tasks refusing to freeze, wq_busy=0):
[ 263.737154] nfsv4.1-svc D ffff8800a0fd7d78 0 951 2 0x00000000
[ 263.737165] ffff8800a0fd7d78 ffff8800a0f66d50 0000000000013f00 ffff8800a0fd7fd8
[ 263.737174] 0000000000013f00 ffff88012af86d50 ffff8800a0f66d50 ffffffffa0298aa0
[ 263.737181] ffff8800a0d70000 ffffffffa0298aa0 0000000000000000 ffff8800a0fd7cd8
[ 263.737188] Call Trace:
[ 263.737239] [<ffffffffa0298aa0>] ? rpc_destroy_wait_queue+0x20/0x20 [sunrpc]
[ 263.737262] [<ffffffffa0298aa0>] ? rpc_destroy_wait_queue+0x20/0x20 [sunrpc]
[ 263.737282] [<ffffffffa0298654>] ? rpc_release_resources_task+0x34/0x40 [sunrpc]
[ 263.737301] [<ffffffffa02997b8>] ? __rpc_execute+0x2c8/0x4a0 [sunrpc]
[ 263.737313] [<ffffffff810db9bb>] ? lock_timer_base.isra.37+0x2b/0x50
[ 263.737323] [<ffffffff8155e999>] schedule+0x29/0x70
[ 263.737332] [<ffffffff8156144b>] schedule_timeout+0x11b/0x250
[ 263.737340] [<ffffffff810db360>] ? migrate_timer_list+0xd0/0xd0
[ 263.737362] [<ffffffffa083e1ba>] nfs41_callback_svc+0x1aa/0x1e0 [nfsv4]
[ 263.737369] [<ffffffff810b4380>] ? wait_woken+0x90/0x90
[ 263.737387] [<ffffffffa083e010>] ? nfs4_callback_svc+0x60/0x60 [nfsv4]
[ 263.737396] [<ffffffff81091748>] kthread+0xd8/0xf0
[ 263.737404] [<ffffffff81091670>] ? kthread_create_on_node+0x1c0/0x1c0
[ 263.737412] [<ffffffff81562758>] ret_from_fork+0x58/0x90
[ 263.737419] [<ffffffff81091670>] ? kthread_create_on_node+0x1c0/0x1c0

This unclean suspend causes garbled Xorg output, or resume is just impossible.

There is a thread in the BBS for this issue: https://bbs.archlinux.org/viewtopic.php?id=195170

The issue occurs with linux 3.19.2, AFAIR it didn't appear with 3.19.1; 3.18.6 is unaffected.

According to https://www.kernel.org/pub/linux/kernel/v3.x/ChangeLog-3.18.9, 3.18.9 has a fix for it.
This task depends upon

Closed by  Doug Newgard (Scimmia)
Sunday, 15 October 2017, 14:26 GMT
Reason for closing:  No response
Comment by Bjoern Franke (bjo) - Wednesday, 25 March 2015, 11:32 GMT
Upstream: https://bugzilla.kernel.org/show_bug.cgi?id=95561
Comment by Bjoern Franke (bjo) - Wednesday, 25 March 2015, 18:02 GMT
The mentioned garbled screen output seems unrelated to this bug, but to https://bugzilla.kernel.org/show_bug.cgi?id=78661
Comment by Bjoern Franke (bjo) - Thursday, 26 March 2015, 22:19 GMT
Issue still exists with 3.19.3.
Comment by Bjoern Franke (bjo) - Thursday, 26 March 2015, 22:41 GMT
According to a post in the manjaro-forum, downgrading nfs-utils to 1.3.1-1 helps.
Comment by Pauli (pl) - Sunday, 29 March 2015, 06:50 GMT
Problem persists in 4.0rc5.

It's enough for the nfs client side services (rpcbind/nfs-client/remotefs, did not test them individually) to be started for it to occur -- nothing actually needs to be mounted.
Comment by Matt (madalu) - Sunday, 19 April 2015, 13:24 GMT
As an alternative to downgrading nfs-utils, forcing nfs-utils to use version 4.0 also works as a temporary workaround until this bug is fixed.

In /etc/nfsmount.conf, edit the line "Defaultvers=" so that it reads "Defaultvers=4.0"
Comment by witscher (henrythemouse) - Wednesday, 24 June 2015, 20:37 GMT
Is this is a kernel issue? I'm getting the same problem and I'm running 3.14.44-1-lts.
I get the freeze issue when trying to sleep and also a stop job loop that never ends when rebooting (related to an nfs mount).
The work around above did work for me, which is much appreciated.
Comment by mattia (nTia89) - Monday, 02 October 2017, 18:04 GMT
is this issue still valid?

Loading...