linux-nfs.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Jeff Layton <jlayton@primarydata.com>
To: bfields@fieldses.org
Cc: linux-nfs@vger.kernel.org, hch@infradead.org, Neil Brown <neilb@suse.de>
Subject: [PATCH v6 8/9] nfsd: give block_delegation and delegation_blocked its own spinlock
Date: Fri, 25 Jul 2014 07:34:26 -0400	[thread overview]
Message-ID: <1406288067-20663-9-git-send-email-jlayton@primarydata.com> (raw)
In-Reply-To: <1406288067-20663-1-git-send-email-jlayton@primarydata.com>

The state lock can be fairly heavily contended, and there's no reason
that nfs4_file lookups and delegation_blocked should be mutually
exclusive.  Let's give the new block_delegation code its own spinlock.
It does mean that we'll need to take a different lock in the delegation
break code, but that's not generally as critical to performance.

Cc: Neil Brown <neilb@suse.de>
Signed-off-by: Jeff Layton <jlayton@primarydata.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
---
 fs/nfsd/nfs4state.c | 21 +++++++++++----------
 1 file changed, 11 insertions(+), 10 deletions(-)

diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c
index 85d7ac664691..ecfddca9b841 100644
--- a/fs/nfsd/nfs4state.c
+++ b/fs/nfsd/nfs4state.c
@@ -517,10 +517,11 @@ static struct nfs4_ol_stateid * nfs4_alloc_stateid(struct nfs4_client *clp)
  * Each filter is 256 bits.  We hash the filehandle to 32bit and use the
  * low 3 bytes as hash-table indices.
  *
- * 'state_lock', which is always held when block_delegations() is called,
+ * 'blocked_delegations_lock', which is always taken in block_delegations(),
  * is used to manage concurrent access.  Testing does not need the lock
  * except when swapping the two filters.
  */
+static DEFINE_SPINLOCK(blocked_delegations_lock);
 static struct bloom_pair {
 	int	entries, old_entries;
 	time_t	swap_time;
@@ -536,7 +537,7 @@ static int delegation_blocked(struct knfsd_fh *fh)
 	if (bd->entries == 0)
 		return 0;
 	if (seconds_since_boot() - bd->swap_time > 30) {
-		spin_lock(&state_lock);
+		spin_lock(&blocked_delegations_lock);
 		if (seconds_since_boot() - bd->swap_time > 30) {
 			bd->entries -= bd->old_entries;
 			bd->old_entries = bd->entries;
@@ -545,7 +546,7 @@ static int delegation_blocked(struct knfsd_fh *fh)
 			bd->new = 1-bd->new;
 			bd->swap_time = seconds_since_boot();
 		}
-		spin_unlock(&state_lock);
+		spin_unlock(&blocked_delegations_lock);
 	}
 	hash = arch_fast_hash(&fh->fh_base, fh->fh_size, 0);
 	if (test_bit(hash&255, bd->set[0]) &&
@@ -566,16 +567,16 @@ static void block_delegations(struct knfsd_fh *fh)
 	u32 hash;
 	struct bloom_pair *bd = &blocked_delegations;
 
-	lockdep_assert_held(&state_lock);
-
 	hash = arch_fast_hash(&fh->fh_base, fh->fh_size, 0);
 
+	spin_lock(&blocked_delegations_lock);
 	__set_bit(hash&255, bd->set[bd->new]);
 	__set_bit((hash>>8)&255, bd->set[bd->new]);
 	__set_bit((hash>>16)&255, bd->set[bd->new]);
 	if (bd->entries == 0)
 		bd->swap_time = seconds_since_boot();
 	bd->entries += 1;
+	spin_unlock(&blocked_delegations_lock);
 }
 
 static struct nfs4_delegation *
@@ -3096,16 +3097,16 @@ void nfsd4_prepare_cb_recall(struct nfs4_delegation *dp)
 	struct nfs4_client *clp = dp->dl_stid.sc_client;
 	struct nfsd_net *nn = net_generic(clp->net, nfsd_net_id);
 
-	/*
-	 * We can't do this in nfsd_break_deleg_cb because it is
-	 * already holding inode->i_lock
-	 */
-	spin_lock(&state_lock);
 	block_delegations(&dp->dl_fh);
+
 	/*
+	 * We can't do this in nfsd_break_deleg_cb because it is
+	 * already holding inode->i_lock.
+	 *
 	 * If the dl_time != 0, then we know that it has already been
 	 * queued for a lease break. Don't queue it again.
 	 */
+	spin_lock(&state_lock);
 	if (dp->dl_time == 0) {
 		dp->dl_time = get_seconds();
 		list_add_tail(&dp->dl_recall_lru, &nn->del_recall_lru);
-- 
1.9.3


  parent reply	other threads:[~2014-07-25 11:34 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-07-25 11:34 [PATCH v6 0/9] nfsd: more delegation fixes to prepare for client_mutex removal Jeff Layton
2014-07-25 11:34 ` [PATCH v6 1/9] nfsd: simplify stateid allocation and file handling Jeff Layton
2014-07-25 11:34 ` [PATCH v6 2/9] nfsd: fully unhash delegations when revoking them Jeff Layton
2014-07-29 14:53   ` J. Bruce Fields
2014-07-29 16:41     ` Jeff Layton
2014-07-29 17:53       ` J. Bruce Fields
2014-07-25 11:34 ` [PATCH v6 3/9] nfsd: ensure that clp->cl_revoked list is protected by clp->cl_lock Jeff Layton
2014-07-25 11:34 ` [PATCH v6 4/9] nfsd: Convert delegation counter to an atomic_long_t type Jeff Layton
2014-07-25 11:34 ` [PATCH v6 5/9] nfsd: drop unused stp arg to alloc_init_deleg Jeff Layton
2014-07-25 11:34 ` [PATCH v6 6/9] nfsd: clean up arguments to nfs4_open_delegation Jeff Layton
2014-07-25 11:34 ` [PATCH v6 7/9] nfsd: clean up nfs4_set_delegation Jeff Layton
2014-07-25 11:34 ` Jeff Layton [this message]
2014-07-25 11:34 ` [PATCH v6 9/9] nfsd: remove dl_fh field from struct nfs4_delegation Jeff Layton
2014-07-29 18:59 ` [PATCH v6 0/9] nfsd: more delegation fixes to prepare for client_mutex removal J. Bruce Fields

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1406288067-20663-9-git-send-email-jlayton@primarydata.com \
    --to=jlayton@primarydata.com \
    --cc=bfields@fieldses.org \
    --cc=hch@infradead.org \
    --cc=linux-nfs@vger.kernel.org \
    --cc=neilb@suse.de \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).