GFS2: Use RCU for glock hash table

This has a number of advantages:

 - Reduces contention on the hash table lock
 - Makes the code smaller and simpler
 - Should speed up glock dumps when under load
 - Removes ref count changing in examine_bucket
 - No longer need hash chain lock in glock_put() in common case

There are some further changes which this enables and which
we may do in the future. One is to look at using SLAB_RCU,
and another is to look at using a per-cpu counter for the
per-sb glock counter, since that is touched twice in the
lifetime of each glock (but only used at umount time).

Signed-off-by: Steven Whitehouse <swhiteho@redhat.com>
Cc: Paul E. McKenney <paulmck@linux.vnet.ibm.com>
This commit is contained in:
Steven Whitehouse
2011-01-19 09:30:01 +00:00
parent 2b1caf6ed7
commit bc015cb841
8 changed files with 190 additions and 297 deletions

View File

@@ -14,6 +14,8 @@
#include <linux/module.h>
#include <linux/init.h>
#include <linux/gfs2_ondisk.h>
#include <linux/rcupdate.h>
#include <linux/rculist_bl.h>
#include <asm/atomic.h>
#include "gfs2.h"
@@ -45,7 +47,7 @@ static void gfs2_init_glock_once(void *foo)
{
struct gfs2_glock *gl = foo;
INIT_HLIST_NODE(&gl->gl_list);
INIT_HLIST_BL_NODE(&gl->gl_list);
spin_lock_init(&gl->gl_spin);
INIT_LIST_HEAD(&gl->gl_holders);
INIT_LIST_HEAD(&gl->gl_lru);
@@ -198,6 +200,8 @@ static void __exit exit_gfs2_fs(void)
unregister_filesystem(&gfs2meta_fs_type);
destroy_workqueue(gfs_recovery_wq);
rcu_barrier();
kmem_cache_destroy(gfs2_quotad_cachep);
kmem_cache_destroy(gfs2_rgrpd_cachep);
kmem_cache_destroy(gfs2_bufdata_cachep);