nfs: fix kernel warning when removing proc entry
Subsystems (2)
PATCH v2 1/1 nfs: fix kernel warning when removing proc entry
I saw the following kernel warning:
[ 1852.321222] ------------[ cut here ]------------ [ 1852.326527] WARNING: CPU: 0 PID: 118 at fs/proc/generic.c:521 remove_proc_entry+0x154/0x16b() [ 1852.335630] remove_proc_entry: removing non-empty directory 'fs/nfsfs', leaking at least 'volumes' [ 1852.344084] CPU: 0 PID: 118 Comm: kworker/u8:2 Not tainted 3.16.0+ #540 [ 1852.350036] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 1852.354992] Workqueue: netns cleanup_net [ 1852.358701] 0000000000000000 ffff880116f2fbd0 ffffffff819c03e9 ffff880116f2fc18 [ 1852.366474] ffff880116f2fc08 ffffffff810744ee ffffffff811e0e6e ffff8800d4e96238 [ 1852.373507] ffffffff81dbe665 ffff8800d46a5948 0000000000000005 ffff880116f2fc68 [ 1852.380224] Call Trace: [ 1852.381976] [<ffffffff819c03e9>] dump_stack+0x4d/0x66 [ 1852.385495] [<ffffffff810744ee>] warn_slowpath_common+0x7a/0x93 [ 1852.389869] [<ffffffff811e0e6e>] ? remove_proc_entry+0x154/0x16b [ 1852.393987] [<ffffffff8107457b>] warn_slowpath_fmt+0x4c/0x4e [ 1852.397999] [<ffffffff811e0e6e>] remove_proc_entry+0x154/0x16b [ 1852.402034] [<ffffffff8129c73d>] nfs_fs_proc_net_exit+0x53/0x56 [ 1852.406136] [<ffffffff812a103b>] nfs_net_exit+0x12/0x1d [ 1852.409774] [<ffffffff81785bc9>] ops_exit_list+0x44/0x55 [ 1852.413529] [<ffffffff81786389>] cleanup_net+0xee/0x182 [ 1852.417198] [<ffffffff81088c9e>] process_one_work+0x209/0x40d [ 1852.502320] [<ffffffff81088bf7>] ? process_one_work+0x162/0x40d [ 1852.587629] [<ffffffff810890c1>] worker_thread+0x1f0/0x2c7 [ 1852.673291] [<ffffffff81088ed1>] ? process_scheduled_works+0x2f/0x2f [ 1852.759470] [<ffffffff8108e079>] kthread+0xc9/0xd1 [ 1852.843099] [<ffffffff8109427f>] ? finish_task_switch+0x3a/0xce [ 1852.926518] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.008565] [<ffffffff819cbeac>] ret_from_fork+0x7c/0xb0 [ 1853.076477] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.140653] ---[ end trace 69c4c6617f78e32d ]---
It looks wrong that we add "/proc/net/nfsfs" in nfs_fs_proc_net_init() while remove "/proc/fs/nfsfs" in nfs_fs_proc_net_exit().
---
v2 - fix nfs_fs_proc_net_init() as well
fs/nfs/client.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
fs/nfs/client.c +2 -2
@@ -1418,7 +1418,7 @@ int nfs_fs_proc_net_init(struct net *net) error_2:remove_proc_entry ("servers" , nn->proc_nfsfs); error_1: -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_entry ("nfsfs" , net->proc_net); error_0:return -ENOMEM ; }
@@ -1429,7 +1429,7 @@ void nfs_fs_proc_net_exit(struct net *net)remove_proc_entry ("volumes" , nn->proc_nfsfs);remove_proc_entry ("servers" , nn->proc_nfsfs); -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_entry ("nfsfs" , net->proc_net); }/*
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
Eric
Quoted 66 lines
[ 1852.321222] ------------[ cut here ]------------ [ 1852.326527] WARNING: CPU: 0 PID: 118 at fs/proc/generic.c:521 remove_proc_entry+0x154/0x16b() [ 1852.335630] remove_proc_entry: removing non-empty directory 'fs/nfsfs', leaking at least 'volumes' [ 1852.344084] CPU: 0 PID: 118 Comm: kworker/u8:2 Not tainted 3.16.0+ #540 [ 1852.350036] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 1852.354992] Workqueue: netns cleanup_net [ 1852.358701] 0000000000000000 ffff880116f2fbd0 ffffffff819c03e9 ffff880116f2fc18 [ 1852.366474] ffff880116f2fc08 ffffffff810744ee ffffffff811e0e6e ffff8800d4e96238 [ 1852.373507] ffffffff81dbe665 ffff8800d46a5948 0000000000000005 ffff880116f2fc68 [ 1852.380224] Call Trace: [ 1852.381976] [<ffffffff819c03e9>] dump_stack+0x4d/0x66 [ 1852.385495] [<ffffffff810744ee>] warn_slowpath_common+0x7a/0x93 [ 1852.389869] [<ffffffff811e0e6e>] ? remove_proc_entry+0x154/0x16b [ 1852.393987] [<ffffffff8107457b>] warn_slowpath_fmt+0x4c/0x4e [ 1852.397999] [<ffffffff811e0e6e>] remove_proc_entry+0x154/0x16b [ 1852.402034] [<ffffffff8129c73d>] nfs_fs_proc_net_exit+0x53/0x56 [ 1852.406136] [<ffffffff812a103b>] nfs_net_exit+0x12/0x1d [ 1852.409774] [<ffffffff81785bc9>] ops_exit_list+0x44/0x55 [ 1852.413529] [<ffffffff81786389>] cleanup_net+0xee/0x182 [ 1852.417198] [<ffffffff81088c9e>] process_one_work+0x209/0x40d [ 1852.502320] [<ffffffff81088bf7>] ? process_one_work+0x162/0x40d [ 1852.587629] [<ffffffff810890c1>] worker_thread+0x1f0/0x2c7 [ 1852.673291] [<ffffffff81088ed1>] ? process_scheduled_works+0x2f/0x2f [ 1852.759470] [<ffffffff8108e079>] kthread+0xc9/0xd1 [ 1852.843099] [<ffffffff8109427f>] ? finish_task_switch+0x3a/0xce [ 1852.926518] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.008565] [<ffffffff819cbeac>] ret_from_fork+0x7c/0xb0 [ 1853.076477] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.140653] ---[ end trace 69c4c6617f78e32d ]---It looks wrong that we add "
/proc/net/nfsfs" innfs_fs_proc_net_init()while remove "/proc/fs/nfsfs" innfs_fs_proc_net_exit().Fixes commit 65b38851a17 (NFS: Fix /proc/fs/nfsfs/servers and /proc/fs/nfsfs/volumes)Cc Eric W. Biederman <[email protected]>Cc Trond Myklebust <[email protected]>Cc Dan Aloni <[email protected]>Signed-off-by Cong Wang <[email protected]>---
v2 - fixnfs_fs_proc_net_init()as wellfs/nfs/client.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)fs/nfs/client.c +2
-2@@ -1418,7 +1418,7 @@ int nfs_fs_proc_net_init(struct net *net) error_2:remove_proc_entry ("servers" , nn->proc_nfsfs); error_1: -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_entry ("nfsfs" , net->proc_net); error_0:return -ENOMEM ; }@@ -1429,7 +1429,7 @@ void nfs_fs_proc_net_exit(struct net *net)remove_proc_entry ("volumes" , nn->proc_nfsfs);remove_proc_entry ("servers" , nn->proc_nfsfs); -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_entry ("nfsfs" , net->proc_net); }/*
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 15 lines · Christian Kujau
Quoted 9 lines · Eric W. Biederman
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
This made the boot warning go away for me as well:
http://lkml.iu.edu/hypermail/linux/kernel/1408.2/01792.html
Quoted 8 lines
Tested-by Christian Kujau <[email protected]>Thanks!
Christian.
--
BOFH excuse #182:endothermal recalibration
It is customary (and expected!) to send these patches through the maintainer; not to push them yourself.
Signature · 5 lines
Trond Myklebust Linux NFS client maintainer, PrimaryData [email protected]
Quoted 10 lines · Eric W. Biederman
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
This is precisely _WHY_ it is supposed to go through the maintainer, and not through arbitrary namespace trees. I'm tired of the namespace folks pushing crap and just assuming that a "Cc:" is OK. A "Cc:
maintainer" means this is an RFC, not something that is to be committed.
Trond
Quoted 9 lines · Eric W. Biederman
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
Why does that code wank with one-by-one remove_proc_entry(), BTW?remove_proc_subtree("nfsfs", net->proc_net) will take care of the whole pile just fine, TYVM... While we are it, there's no need to keep ->proc_nfsfs at all - just have it in a local variable in nfs_fs_proc_net_init().
Quoted 8 lines · Al Viro
Quoted 7 lines · Eric W. Biederman
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
Quoted 7 lines
Quoted 2 lines
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
Why does that code wank with one-by-one
remove_proc_entry(), BTW?remove_proc_subtree("nfsfs", net->proc_net)will take care of the whole pile just fine, TYVM... While we are it, there's no need to keep ->proc_nfsfs at all - just have it in a local variable innfs_fs_proc_net_init().
Since nobody sent me an updated version with the remove_proc_subtree fix, I went ahead and edited the patch myself (see attachment). Cong, please let me know if you disagree with that change, otherwise, that will be the final patch sent upstream and Cc: stable # 3.4+.
I'll schedule cleanup patches to make the same changes to the original nfs_fs_proc_exit() and nfs_fs_proc_init() and to remove (struct nfs_net)->proc_nfsfs for merging in 3.18.
Signature · 5 lines
Trond Myklebust Linux NFS client maintainer, PrimaryData [email protected]
From 21e81002f9788a3af591416b6dec60d7b67f2fb2 Mon Sep 17 00:00:00 2001
From: Cong Wang <[email protected]>
Date: Mon, 8 Sep 2014 16:17:55 -0700
Subject: [PATCH] nfs: fix kernel warning when removing proc entry
I saw the following kernel warning:
[ 1852.321222] ------------[ cut here ]------------ [ 1852.326527] WARNING: CPU: 0 PID: 118 at fs/proc/generic.c:521 remove_proc_entry+0x154/0x16b() [ 1852.335630] remove_proc_entry: removing non-empty directory 'fs/nfsfs', leaking at least 'volumes' [ 1852.344084] CPU: 0 PID: 118 Comm: kworker/u8:2 Not tainted 3.16.0+ #540 [ 1852.350036] Hardware name: Bochs Bochs, BIOS Bochs 01/01/2011 [ 1852.354992] Workqueue: netns cleanup_net [ 1852.358701] 0000000000000000 ffff880116f2fbd0 ffffffff819c03e9 ffff880116f2fc18 [ 1852.366474] ffff880116f2fc08 ffffffff810744ee ffffffff811e0e6e ffff8800d4e96238 [ 1852.373507] ffffffff81dbe665 ffff8800d46a5948 0000000000000005 ffff880116f2fc68 [ 1852.380224] Call Trace: [ 1852.381976] [<ffffffff819c03e9>] dump_stack+0x4d/0x66 [ 1852.385495] [<ffffffff810744ee>] warn_slowpath_common+0x7a/0x93 [ 1852.389869] [<ffffffff811e0e6e>] ? remove_proc_entry+0x154/0x16b [ 1852.393987] [<ffffffff8107457b>] warn_slowpath_fmt+0x4c/0x4e [ 1852.397999] [<ffffffff811e0e6e>] remove_proc_entry+0x154/0x16b [ 1852.402034] [<ffffffff8129c73d>] nfs_fs_proc_net_exit+0x53/0x56 [ 1852.406136] [<ffffffff812a103b>] nfs_net_exit+0x12/0x1d [ 1852.409774] [<ffffffff81785bc9>] ops_exit_list+0x44/0x55 [ 1852.413529] [<ffffffff81786389>] cleanup_net+0xee/0x182 [ 1852.417198] [<ffffffff81088c9e>] process_one_work+0x209/0x40d [ 1852.502320] [<ffffffff81088bf7>] ? process_one_work+0x162/0x40d [ 1852.587629] [<ffffffff810890c1>] worker_thread+0x1f0/0x2c7 [ 1852.673291] [<ffffffff81088ed1>] ? process_scheduled_works+0x2f/0x2f [ 1852.759470] [<ffffffff8108e079>] kthread+0xc9/0xd1 [ 1852.843099] [<ffffffff8109427f>] ? finish_task_switch+0x3a/0xce [ 1852.926518] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.008565] [<ffffffff819cbeac>] ret_from_fork+0x7c/0xb0 [ 1853.076477] [<ffffffff8108dfb0>] ? __kthread_parkme+0x61/0x61 [ 1853.140653] ---[ end trace 69c4c6617f78e32d ]---
It looks wrong that we add "/proc/net/nfsfs" in nfs_fs_proc_net_init() while remove "/proc/fs/nfsfs" in nfs_fs_proc_net_exit().
[Trond: replace uses of remove_proc_entry() with remove_proc_subtree() as suggested by Al Viro]
--- fs/nfs/client.c | 12 +++--------- 1 file changed, 3 insertions(+), 9 deletions(-)
fs/nfs/client.c +3 -9
@@ -1412,24 +1412,18 @@ int nfs_fs_proc_net_init(struct net *net) p =proc_create ("volumes" ,S_IFREG |S_IRUGO , nn->proc_nfsfs, &nfs_volume_list_fops);if (!p) - goto error_2; + goto error_1;return 0 ; -error_2: -remove_proc_entry ("servers" , nn->proc_nfsfs); error_1: -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_subtree ("nfsfs" , net->proc_net); error_0:return -ENOMEM ; }void nfs_fs_proc_net_exit (struct net *net) { -struct nfs_net *nn =net_generic (net, nfs_net_id); - -remove_proc_entry ("volumes" , nn->proc_nfsfs); -remove_proc_entry ("servers" , nn->proc_nfsfs); -remove_proc_entry ("fs/nfsfs" ,NULL ); +remove_proc_subtree ("nfsfs" , net->proc_net); }/*
Signature · 1 line
1.9.3
Quoted 17 lines · Trond Myklebust
Quoted 8 lines · Al Viro
Quoted 7 lines · Eric W. Biederman
Quoted 1 line · Cong Wang
I saw the following kernel warning:
Cong thanks for finding and tracking this. I was clearly asleep at the switch when I was testing my fix to the nfs client code :(
Quoted 7 lines
Quoted 2 lines
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
Why does that code wank with one-by-one
remove_proc_entry(), BTW?remove_proc_subtree("nfsfs", net->proc_net)will take care of the whole pile just fine, TYVM... While we are it, there's no need to keep ->proc_nfsfs at all - just have it in a local variable innfs_fs_proc_net_init().
Quoted 9 lines
Since nobody sent me an updated version with the
remove_proc_subtreefix, I went ahead and edited the patch myself (see attachment). Cong, please let me know if you disagree with that change, otherwise, that will be the final patch sent upstream and Cc: stable # 3.4+.I'll schedule cleanup patches to make the same changes to the original
nfs_fs_proc_exit()andnfs_fs_proc_init()and to remove (struct nfs_net)->proc_nfsfs for merging in 3.18.
Oops, I missed Al's reply and didn't know remove_proc_subtree() either.
Thanks for the update and it definitely looks good to me!
Matt Mullins
Tested-byQuoted 2 lines · Eric W. Biederman
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
I also fairly reliably hit this warning with trinity, and with Cong's patch, it seems to be gone.
If it helps get it pushed toward Linus:
Quoted 3 lines · Matt Mullins
Quoted 2 lines · Eric W. Biederman
I have applied this patch and will push it to Linus after it has a little bit to sit in linux-next.
what's the status of this patch?
It's rc4 already and I keep seeing it every boot:
[ 7.730034] WARNING: CPU: 3 PID: 103 at ../fs/proc/generic.c:521 remove_proc_entry+0x1a4/0x1b0() [ 7.730036] remove_proc_entry: removing non-empty directory 'fs/nfsfs', leaking at least 'volumes' [ 7.730037] Modules linked in: ... [ 7.730063] CPU: 3 PID: 103 Comm: kworker/u8:15 Not tainted 3.17.0-rc4+ #196