On 1/26/25 8:20 PM, NeilBrown wrote:
There are at least three locks in the kernel which are initialised as
spin_Lock_init(&l->lock);
This makes them hard to differential in /proc/lock_stat.
For the lock in nfsd/filecache.c introduce a variable with a more
descriptve name so we can:
spin_lock_init(&nfsd_fcache_disposal->lock);
and easily identify this in /proc/lock_stat.
Signed-off-by: NeilBrown <neilb@xxxxxxx>
---
fs/nfsd/filecache.c | 5 +++--
1 file changed, 3 insertions(+), 2 deletions(-)
diff --git a/fs/nfsd/filecache.c b/fs/nfsd/filecache.c
index eb95a53f806f..af95bc381753 100644
--- a/fs/nfsd/filecache.c
+++ b/fs/nfsd/filecache.c
@@ -867,12 +867,13 @@ __nfsd_file_cache_purge(struct net *net)
static struct nfsd_fcache_disposal *
nfsd_alloc_fcache_disposal(void)
{
- struct nfsd_fcache_disposal *l;
+ struct nfsd_fcache_disposal *l, *nfsd_fcache_disposal;
l = kmalloc(sizeof(*l), GFP_KERNEL);
if (!l)
return NULL;
- spin_lock_init(&l->lock);
+ nfsd_fcache_disposal = l;
+ spin_lock_init(&nfsd_fcache_disposal->lock);
timer_setup(&l->timer, nfsd_file_gc_worker, 0);
INIT_LIST_HEAD(&l->recent);
INIT_LIST_HEAD(&l->older);
My concern was there would be multiple dynamically allocated locks
in order to make the locking fine-grained, but I guess that was a
mistaken assumption.
No objection to this approach.
--
Chuck Lever
