glusterfs: Use gcc builtin ATOMIC operator to increase/decreate refcount.

Problem: In glusterfs code base we call mutex_lock/unlock to take
         reference/dereference for a object.Sometime it could be
         reason for lock contention also.

Solution: There is no need to use mutex to increase/decrease ref
          counter, instead of using mutex use gcc builtin ATOMIC
          operation.

Test:   I have not observed yet how much performance gain after apply
        this patch specific to glusterfs but i have tested same
        with below small program(mutex and atomic both) and
        get good difference.

static int numOuterLoops;
static void *
threadFunc(void *arg)
{
    int j;

    for (j = 0; j < numOuterLoops; j++) {
            __atomic_add_fetch (&glob, 1,__ATOMIC_ACQ_REL);
    }
    return NULL;
}

int
main(int argc, char *argv[])
{
    int opt, s, j;
    int numThreads;
    pthread_t *thread;
    int verbose;
    int64_t n = 0;

    if (argc < 2 ) {
     printf(" Please provide 2 args Num of threads && Outer Loop\n");
     exit (-1);
    }
    numThreads = atoi(argv[1]);
    numOuterLoops = atoi (argv[2]);

    if (1) {
        printf("\tthreads: %d; outer loops: %d;\n",
                numThreads, numOuterLoops);
    }

    thread = calloc(numThreads, sizeof(pthread_t));
    if (thread == NULL) {
        printf ("calloc error so exit\n");
        exit (-1);
    }

    __atomic_store (&glob, &n, __ATOMIC_RELEASE);
    for (j = 0; j < numThreads; j++) {
        s = pthread_create(&thread[j], NULL, threadFunc, NULL);
        if (s != 0) {
            printf ("pthread_create failed so exit\n");
            exit (-1);
        }
    }

    for (j = 0; j < numThreads; j++) {
        s = pthread_join(thread[j], NULL);
        if (s != 0) {
            printf ("pthread_join failed so exit\n");
            exit (-1);
        }
    }
    printf("glob value is %ld\n",__atomic_load_n (&glob,__ATOMIC_RELAXED));

    exit(0);
}

   time ./thr_count 800 800000
   threads: 800; outer loops: 800000;
   glob value is 640000000

real	1m10.288s
user	0m57.269s
sys	3m31.565s

time ./thr_count_atomic 800 800000
     threads: 800; outer loops: 800000;
glob value is 640000000

real	0m20.313s
user	1m20.558s
sys	0m0.028

Change-Id: Ie5030a52ea264875e002e108dd4b207b15ab7cc7
Signed-off-by: Mohit Agrawal <moagrawa@redhat.com>

2 files changed, 9 insertions, 15 deletions

diff --git a/xlators/features/changelog/src/changelog-ev-handle.h b/xlators/features/changelog/src/changelog-ev-handle.h
index eef0492a9ee..e89af8793a8 100644
--- a/xlators/features/changelog/src/changelog-ev-handle.h
+++ b/xlators/features/changelog/src/changelog-ev-handle.h
@@ -24,7 +24,7 @@ typedef struct changelog_rpc_clnt {
 
         gf_lock_t lock;
 
-        unsigned long ref;
+        gf_atomic_t   ref;
         gf_boolean_t  disconnected;
 
         unsigned int filter;
@@ -43,11 +43,7 @@ typedef struct changelog_rpc_clnt {
 static inline void
 changelog_rpc_clnt_ref (changelog_rpc_clnt_t *crpc)
 {
-        LOCK (&crpc->lock);
-        {
-                ++crpc->ref;
-        }
-        UNLOCK (&crpc->lock);
+        GF_ATOMIC_INC (crpc->ref);
 }
 
 static inline void
@@ -66,16 +62,14 @@ static inline void
 changelog_rpc_clnt_unref (changelog_rpc_clnt_t *crpc)
 {
         gf_boolean_t gone = _gf_false;
+        uint64_t     ref  = 0;
+
+        ref = GF_ATOMIC_DEC (crpc->ref);
 
-        LOCK (&crpc->lock);
-        {
-                if (!(--crpc->ref)
-                    && changelog_rpc_clnt_is_disconnected (crpc)) {
-                        list_del (&crpc->list);
-                        gone = _gf_true;
-                }
+        if (!ref && changelog_rpc_clnt_is_disconnected (crpc)) {
+                list_del (&crpc->list);
+                gone = _gf_true;
         }
-        UNLOCK (&crpc->lock);
 
         if (gone)
                 crpc->cleanup (crpc);
diff --git a/xlators/features/changelog/src/changelog-rpc.c b/xlators/features/changelog/src/changelog-rpc.c
index 5524e433cbb..72b5cdbc7cb 100644
--- a/xlators/features/changelog/src/changelog-rpc.c
+++ b/xlators/features/changelog/src/changelog-rpc.c
@@ -211,7 +211,7 @@ changelog_rpc_clnt_init (xlator_t *this,
         /* Take a ref, the last unref will be on RPC_CLNT_DESTROY
          * which comes as a result of last rpc_clnt_unref.
          */
-        crpc->ref = 1;
+        GF_ATOMIC_INIT (crpc->ref, 1);
         changelog_set_disconnect_flag (crpc, _gf_false);
 
         crpc->filter = rpc_req->filter;