Commit bedd4b61 authored by Trond Myklebust's avatar Trond Myklebust Committed by J. Bruce Fields

nfsd: convert the lru list into a per-bucket thing

Signed-off-by: default avatarTrond Myklebust <trond.myklebust@primarydata.com>
Signed-off-by: default avatarJ. Bruce Fields <bfields@redhat.com>
parent 7142b98d
...@@ -29,10 +29,10 @@ ...@@ -29,10 +29,10 @@
struct nfsd_drc_bucket { struct nfsd_drc_bucket {
struct hlist_head cache_hash; struct hlist_head cache_hash;
struct list_head lru_head;
}; };
static struct nfsd_drc_bucket *drc_hashtbl; static struct nfsd_drc_bucket *drc_hashtbl;
static struct list_head lru_head;
static struct kmem_cache *drc_slab; static struct kmem_cache *drc_slab;
/* max number of entries allowed in the cache */ /* max number of entries allowed in the cache */
...@@ -40,6 +40,7 @@ static unsigned int max_drc_entries; ...@@ -40,6 +40,7 @@ static unsigned int max_drc_entries;
/* number of significant bits in the hash value */ /* number of significant bits in the hash value */
static unsigned int maskbits; static unsigned int maskbits;
static unsigned int drc_hashsize;
/* /*
* Stats and other tracking of on the duplicate reply cache. All of these and * Stats and other tracking of on the duplicate reply cache. All of these and
...@@ -167,8 +168,8 @@ nfsd_reply_cache_free(struct svc_cacherep *rp) ...@@ -167,8 +168,8 @@ nfsd_reply_cache_free(struct svc_cacherep *rp)
int nfsd_reply_cache_init(void) int nfsd_reply_cache_init(void)
{ {
unsigned int hashsize; unsigned int hashsize;
unsigned int i;
INIT_LIST_HEAD(&lru_head);
max_drc_entries = nfsd_cache_size_limit(); max_drc_entries = nfsd_cache_size_limit();
num_drc_entries = 0; num_drc_entries = 0;
hashsize = nfsd_hashsize(max_drc_entries); hashsize = nfsd_hashsize(max_drc_entries);
...@@ -183,6 +184,9 @@ int nfsd_reply_cache_init(void) ...@@ -183,6 +184,9 @@ int nfsd_reply_cache_init(void)
drc_hashtbl = kcalloc(hashsize, sizeof(*drc_hashtbl), GFP_KERNEL); drc_hashtbl = kcalloc(hashsize, sizeof(*drc_hashtbl), GFP_KERNEL);
if (!drc_hashtbl) if (!drc_hashtbl)
goto out_nomem; goto out_nomem;
for (i = 0; i < hashsize; i++)
INIT_LIST_HEAD(&drc_hashtbl[i].lru_head);
drc_hashsize = hashsize;
return 0; return 0;
out_nomem: out_nomem:
...@@ -194,17 +198,22 @@ int nfsd_reply_cache_init(void) ...@@ -194,17 +198,22 @@ int nfsd_reply_cache_init(void)
void nfsd_reply_cache_shutdown(void) void nfsd_reply_cache_shutdown(void)
{ {
struct svc_cacherep *rp; struct svc_cacherep *rp;
unsigned int i;
unregister_shrinker(&nfsd_reply_cache_shrinker); unregister_shrinker(&nfsd_reply_cache_shrinker);
cancel_delayed_work_sync(&cache_cleaner); cancel_delayed_work_sync(&cache_cleaner);
while (!list_empty(&lru_head)) { for (i = 0; i < drc_hashsize; i++) {
rp = list_entry(lru_head.next, struct svc_cacherep, c_lru); struct list_head *head = &drc_hashtbl[i].lru_head;
while (!list_empty(head)) {
rp = list_first_entry(head, struct svc_cacherep, c_lru);
nfsd_reply_cache_free_locked(rp); nfsd_reply_cache_free_locked(rp);
} }
}
kfree (drc_hashtbl); kfree (drc_hashtbl);
drc_hashtbl = NULL; drc_hashtbl = NULL;
drc_hashsize = 0;
if (drc_slab) { if (drc_slab) {
kmem_cache_destroy(drc_slab); kmem_cache_destroy(drc_slab);
...@@ -217,10 +226,10 @@ void nfsd_reply_cache_shutdown(void) ...@@ -217,10 +226,10 @@ void nfsd_reply_cache_shutdown(void)
* not already scheduled. * not already scheduled.
*/ */
static void static void
lru_put_end(struct svc_cacherep *rp) lru_put_end(struct nfsd_drc_bucket *b, struct svc_cacherep *rp)
{ {
rp->c_timestamp = jiffies; rp->c_timestamp = jiffies;
list_move_tail(&rp->c_lru, &lru_head); list_move_tail(&rp->c_lru, &b->lru_head);
schedule_delayed_work(&cache_cleaner, RC_EXPIRE); schedule_delayed_work(&cache_cleaner, RC_EXPIRE);
} }
...@@ -234,17 +243,13 @@ hash_refile(struct nfsd_drc_bucket *b, struct svc_cacherep *rp) ...@@ -234,17 +243,13 @@ hash_refile(struct nfsd_drc_bucket *b, struct svc_cacherep *rp)
hlist_add_head(&rp->c_hash, &b->cache_hash); hlist_add_head(&rp->c_hash, &b->cache_hash);
} }
/*
* Walk the LRU list and prune off entries that are older than RC_EXPIRE.
* Also prune the oldest ones when the total exceeds the max number of entries.
*/
static long static long
prune_cache_entries(void) prune_bucket(struct nfsd_drc_bucket *b)
{ {
struct svc_cacherep *rp, *tmp; struct svc_cacherep *rp, *tmp;
long freed = 0; long freed = 0;
list_for_each_entry_safe(rp, tmp, &lru_head, c_lru) { list_for_each_entry_safe(rp, tmp, &b->lru_head, c_lru) {
/* /*
* Don't free entries attached to calls that are still * Don't free entries attached to calls that are still
* in-progress, but do keep scanning the list. * in-progress, but do keep scanning the list.
...@@ -257,16 +262,33 @@ prune_cache_entries(void) ...@@ -257,16 +262,33 @@ prune_cache_entries(void)
nfsd_reply_cache_free_locked(rp); nfsd_reply_cache_free_locked(rp);
freed++; freed++;
} }
return freed;
}
/*
* Walk the LRU list and prune off entries that are older than RC_EXPIRE.
* Also prune the oldest ones when the total exceeds the max number of entries.
*/
static long
prune_cache_entries(void)
{
unsigned int i;
long freed = 0;
bool cancel = true;
for (i = 0; i < drc_hashsize; i++) {
struct nfsd_drc_bucket *b = &drc_hashtbl[i];
freed += prune_bucket(b);
if (!list_empty(&b->lru_head))
cancel = false;
}
/* /*
* Conditionally rearm the job. If we cleaned out the list, then * Conditionally rearm the job to run in RC_EXPIRE since we just
* cancel any pending run (since there won't be any work to do). * ran the pruner.
* Otherwise, we rearm the job or modify the existing one to run in
* RC_EXPIRE since we just ran the pruner.
*/ */
if (list_empty(&lru_head)) if (!cancel)
cancel_delayed_work(&cache_cleaner);
else
mod_delayed_work(system_wq, &cache_cleaner, RC_EXPIRE); mod_delayed_work(system_wq, &cache_cleaner, RC_EXPIRE);
return freed; return freed;
} }
...@@ -458,7 +480,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) ...@@ -458,7 +480,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp)
rp->c_csum = csum; rp->c_csum = csum;
hash_refile(b, rp); hash_refile(b, rp);
lru_put_end(rp); lru_put_end(b, rp);
/* release any buffer */ /* release any buffer */
if (rp->c_type == RC_REPLBUFF) { if (rp->c_type == RC_REPLBUFF) {
...@@ -475,7 +497,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp) ...@@ -475,7 +497,7 @@ nfsd_cache_lookup(struct svc_rqst *rqstp)
nfsdstats.rchits++; nfsdstats.rchits++;
/* We found a matching entry which is either in progress or done. */ /* We found a matching entry which is either in progress or done. */
age = jiffies - rp->c_timestamp; age = jiffies - rp->c_timestamp;
lru_put_end(rp); lru_put_end(b, rp);
rtn = RC_DROPIT; rtn = RC_DROPIT;
/* Request being processed or excessive rexmits */ /* Request being processed or excessive rexmits */
...@@ -530,12 +552,17 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) ...@@ -530,12 +552,17 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp)
{ {
struct svc_cacherep *rp = rqstp->rq_cacherep; struct svc_cacherep *rp = rqstp->rq_cacherep;
struct kvec *resv = &rqstp->rq_res.head[0], *cachv; struct kvec *resv = &rqstp->rq_res.head[0], *cachv;
u32 hash;
struct nfsd_drc_bucket *b;
int len; int len;
size_t bufsize = 0; size_t bufsize = 0;
if (!rp) if (!rp)
return; return;
hash = nfsd_cache_hash(rp->c_xid);
b = &drc_hashtbl[hash];
len = resv->iov_len - ((char*)statp - (char*)resv->iov_base); len = resv->iov_len - ((char*)statp - (char*)resv->iov_base);
len >>= 2; len >>= 2;
...@@ -568,7 +595,7 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp) ...@@ -568,7 +595,7 @@ nfsd_cache_update(struct svc_rqst *rqstp, int cachetype, __be32 *statp)
} }
spin_lock(&cache_lock); spin_lock(&cache_lock);
drc_mem_usage += bufsize; drc_mem_usage += bufsize;
lru_put_end(rp); lru_put_end(b, rp);
rp->c_secure = rqstp->rq_secure; rp->c_secure = rqstp->rq_secure;
rp->c_type = cachetype; rp->c_type = cachetype;
rp->c_state = RC_DONE; rp->c_state = RC_DONE;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment