sunrpc/cache: allow threads to block while waiting for cache update.

The current practice of waiting for cache updates by queueing the
whole request to be retried has (at least) two problems.

1/ With NFSv4, requests can be quite complex and re-trying a whole
  request when a later part fails should only be a last-resort, not a
  normal practice.

2/ Large requests, and in particular any 'write' request, will not be
  queued by the current code and doing so would be undesirable.

In many cases only a very sort wait is needed before the cache gets
valid data.

So, providing the underlying transport permits it by setting
 ->thread_wait,
arrange to wait briefly for an upcall to be completed (as reflected in
the clearing of CACHE_PENDING).
If the short wait was not long enough and CACHE_PENDING is still set,
fall back on the old approach.

The 'thread_wait' value is set to 5 seconds when there are spare
threads, and 1 second when there are no spare threads.

These values are probably much higher than needed, but will ensure
some forward progress.

Note that as we only request an update for a non-valid item, and as
non-valid items are updated in place it is extremely unlikely that
cache_check will return -ETIMEDOUT.  Normally cache_defer_req will
sleep for a short while and then find that the item is_valid.

Signed-off-by: NeilBrown <neilb@suse.de>
Signed-off-by: J. Bruce Fields <bfields@redhat.com>

2 files changed, 69 insertions, 1 deletions

diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
index 8dc121955fdc..2c5297f245b4 100644
--- a/net/sunrpc/cache.c
+++ b/net/sunrpc/cache.c
@@ -509,10 +509,22 @@ static LIST_HEAD(cache_defer_list);
 static struct list_head cache_defer_hash[DFR_HASHSIZE];
 static int cache_defer_cnt;
 
+struct thread_deferred_req {
+        struct cache_deferred_req handle;
+        struct completion completion;
+};
+static void cache_restart_thread(struct cache_deferred_req *dreq, int too_many)
+{
+        struct thread_deferred_req *dr =
+                container_of(dreq, struct thread_deferred_req, handle);
+        complete(&dr->completion);
+}
+
 static int cache_defer_req(struct cache_req *req, struct cache_head *item)
 {
         struct cache_deferred_req *dreq, *discard;
         int hash = DFR_HASH(item);
+        struct thread_deferred_req sleeper;
 
         if (cache_defer_cnt >= DFR_MAX) {
                 /* too much in the cache, randomly drop this one,
@@ -521,7 +533,15 @@ static int cache_defer_req(struct cache_req *req, struct cache_head *item)
                 if (net_random()&1)
                         return -ENOMEM;
         }
-        dreq = req->defer(req);
+        if (req->thread_wait) {
+                dreq = &sleeper.handle;
+                sleeper.completion =
+                        COMPLETION_INITIALIZER_ONSTACK(sleeper.completion);
+                dreq->revisit = cache_restart_thread;
+        } else
+                dreq = req->defer(req);
+
+ retry:
         if (dreq == NULL)
                 return -ENOMEM;
 
@@ -555,6 +575,43 @@ static int cache_defer_req(struct cache_req *req, struct cache_head *item)
                 cache_revisit_request(item);
                 return -EAGAIN;
         }
+
+        if (dreq == &sleeper.handle) {
+                if (wait_for_completion_interruptible_timeout(
+                            &sleeper.completion, req->thread_wait) <= 0) {
+                        /* The completion wasn't completed, so we need
+                         * to clean up
+                         */
+                        spin_lock(&cache_defer_lock);
+                        if (!list_empty(&sleeper.handle.hash)) {
+                                list_del_init(&sleeper.handle.recent);
+                                list_del_init(&sleeper.handle.hash);
+                                cache_defer_cnt--;
+                                spin_unlock(&cache_defer_lock);
+                        } else {
+                                /* cache_revisit_request already removed
+                                 * this from the hash table, but hasn't
+                                 * called ->revisit yet.  It will very soon
+                                 * and we need to wait for it.
+                                 */
+                                spin_unlock(&cache_defer_lock);
+                                wait_for_completion(&sleeper.completion);
+                        }
+                }
+                if (test_bit(CACHE_PENDING, &item->flags)) {
+                        /* item is still pending, try request
+                         * deferral
+                         */
+                        dreq = req->defer(req);
+                        goto retry;
+                }
+                /* only return success if we actually deferred the
+                 * request.  In this case we waited until it was
+                 * answered so no deferral has happened - rather
+                 * an answer already exists.
+                 */
+                return -EEXIST;
+        }
         return 0;
 }
 
diff --git a/net/sunrpc/svc_xprt.c b/net/sunrpc/svc_xprt.c
index cbc084939dd8..8ff6840866fa 100644
--- a/net/sunrpc/svc_xprt.c
+++ b/net/sunrpc/svc_xprt.c
@@ -651,6 +651,11 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
         if (signalled() || kthread_should_stop())
                 return -EINTR;
 
+        /* Normally we will wait up to 5 seconds for any required
+         * cache information to be provided.
+         */
+        rqstp->rq_chandle.thread_wait = 5*HZ;
+
         spin_lock_bh(&pool->sp_lock);
         xprt = svc_xprt_dequeue(pool);
         if (xprt) {
@@ -658,6 +663,12 @@ int svc_recv(struct svc_rqst *rqstp, long timeout)
                 svc_xprt_get(xprt);
                 rqstp->rq_reserved = serv->sv_max_mesg;
                 atomic_add(rqstp->rq_reserved, &xprt->xpt_reserved);
+
+                /* As there is a shortage of threads and this request
+                 * had to be queue, don't allow the thread to wait so
+                 * long for cache updates.
+                 */
+                rqstp->rq_chandle.thread_wait = 1*HZ;
         } else {
                 /* No data pending. Go to sleep */
                 svc_thread_enqueue(pool, rqstp);