iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

nfsd: move blocked lock handling under a dedicated spinlock

Browse Source 
Bruce was hitting some lockdep warnings in testing, showing that we
could hit a deadlock with the new CB_NOTIFY_LOCK handling, involving a
rather complex situation involving four different spinlocks.

The crux of the matter is that we end up taking the nn->client_lock in
the lm_notify handler. The simplest fix is to just declare a new
per-nfsd_net spinlock to protect the new CB_NOTIFY_LOCK structures.

Signed-off-by: Jeff Layton <jlayton@redhat.com>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
This commit is contained in:
Jeff Layton
2016-10-20 09:34:31 -04:00
committed by J. Bruce Fields
parent 07d9a38068
commit 0cc11a61b8
2 changed files with 20 additions and 13 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
fs/nfsd/netns.h
View File

@ -84,6 +84,8 @@ struct nfsd_net {
struct list_head client_lru;
struct list_head close_lru;
struct list_head del_recall_lru;
/* protected by blocked_locks_lock */
struct list_head blocked_locks_lru;
struct delayed_work laundromat_work;
@ -91,6 +93,9 @@ struct nfsd_net {
/* client_lock protects the client lru list and session hash table */
spinlock_t client_lock;
/* protects blocked_locks_lru */
spinlock_t blocked_locks_lock;
struct file *rec_file;
bool in_grace;
const struct nfsd4_client_tracking_ops *client_tracking_ops;