Initial version.

src/nxt_thread_mutex.c

@@ -0,0 +1,192 @@
+
+/*
+ * Copyright (C) Igor Sysoev
+ * Copyright (C) NGINX, Inc.
+ */
+
+#include <nxt_main.h>
+
+
+/*
+ * All modern pthread mutex implementations try to acquire a lock atomically
+ * in userland before going to sleep in kernel.  Some spins on SMP systems
+ * before the sleeping.
+ *
+ * In Solaris since version 8 all mutex types spin before sleeping.
+ * The default spin count is 1000.  It can be overridden using
+ * _THREAD_ADAPTIVE_SPIN=100 environment variable.
+ *
+ * In MacOSX all mutex types spin to acquire a lock protecting a mutex's
+ * internals.  If the mutex is busy, thread calls Mach semaphore_wait().
+ *
+ *
+ * PTHREAD_MUTEX_NORMAL lacks deadlock detection and is the fastest
+ * mutex type.
+ *
+ *   Linux:    No spinning.  The internal name PTHREAD_MUTEX_TIMED_NP
+ *             remains from the times when pthread_mutex_timedlock() was
+ *             non-standard extension.  Alias name: PTHREAD_MUTEX_FAST_NP.
+ *   FreeBSD:  No spinning.
+ *
+ *
+ * PTHREAD_MUTEX_ERRORCHECK is usually as fast as PTHREAD_MUTEX_NORMAL
+ * yet has lightweight deadlock detection.
+ *
+ *   Linux:    No spinning.  The internal name: PTHREAD_MUTEX_ERRORCHECK_NP.
+ *   FreeBSD:  No spinning.
+ *
+ *
+ * PTHREAD_MUTEX_RECURSIVE allows recursive locking.
+ *
+ *   Linux:    No spinning.  The internal name: PTHREAD_MUTEX_RECURSIVE_NP.
+ *   FreeBSD:  No spinning.
+ *
+ *
+ * PTHREAD_MUTEX_ADAPTIVE_NP spins on SMP systems before sleeping.
+ *
+ *   Linux:    No deadlock detection.  Dynamically changes a spin count
+ *             for each mutex from 10 to 100 based on spin count taken
+ *             previously.
+ *
+ *   FreeBSD:  Deadlock detection.  The default spin count is 2000.
+ *             It can be overriden using LIBPTHREAD_SPINLOOPS environment
+ *             variable or by pthread_mutex_setspinloops_np().  If a lock
+ *             is still busy, sched_yield() can be called on both UP and
+ *             SMP systems.  The default yield loop count is zero, but it
+ *             can be set by LIBPTHREAD_YIELDLOOPS environment variable or
+ *             by pthread_mutex_setyieldloops_np().  sched_yield() moves
+ *             a thread to the end of CPU scheduler run queue and this is
+ *             cheaper than removing the thread from the queue and sleeping.
+ *
+ *   Solaris:  No PTHREAD_MUTEX_ADAPTIVE_NP .
+ *   MacOSX:   No PTHREAD_MUTEX_ADAPTIVE_NP.
+ *
+ *
+ * PTHREAD_MUTEX_ELISION_NP is a Linux extension to elide locks using
+ * Intel Restricted Transactional Memory.  It is the most suitable for
+ * rwlock pattern access because it allows simultaneous reads without lock.
+ * Supported since glibc 2.18.
+ *
+ *
+ * PTHREAD_MUTEX_DEFAULT is default mutex type.
+ *
+ *   Linux:    PTHREAD_MUTEX_NORMAL.
+ *   FreeBSD:  PTHREAD_MUTEX_ERRORCHECK.
+ *   Solaris:  PTHREAD_MUTEX_NORMAL.
+ *   MacOSX:   PTHREAD_MUTEX_NORMAL.
+ */
+
+
+nxt_int_t
+nxt_thread_mutex_create(nxt_thread_mutex_t *mtx)
+{
+    nxt_err_t            err;
+    pthread_mutexattr_t  attr;
+
+    err = pthread_mutexattr_init(&attr);
+    if (err != 0) {
+        nxt_thread_log_emerg("pthread_mutexattr_init() failed %E", err);
+        return NXT_ERROR;
+    }
+
+    err = pthread_mutexattr_settype(&attr, PTHREAD_MUTEX_ERRORCHECK);
+    if (err != 0) {
+        nxt_thread_log_emerg("pthread_mutexattr_settype"
+                             "(PTHREAD_MUTEX_ERRORCHECK) failed %E", err);
+        return NXT_ERROR;
+    }
+
+    err = pthread_mutex_init(mtx, &attr);
+    if (err != 0) {
+        nxt_thread_log_emerg("pthread_mutex_init() failed %E", err);
+        return NXT_ERROR;
+    }
+
+    err = pthread_mutexattr_destroy(&attr);
+    if (err != 0) {
+        nxt_thread_log_alert("pthread_mutexattr_destroy() failed %E", err);
+    }
+
+    nxt_thread_log_debug("pthread_mutex_init(%p)", mtx);
+
+    return NXT_OK;
+}
+
+
+void
+nxt_thread_mutex_destroy(nxt_thread_mutex_t *mtx)
+{
+    nxt_err_t  err;
+
+    err = pthread_mutex_destroy(mtx);
+    if (nxt_slow_path(err != 0)) {
+        nxt_thread_log_alert("pthread_mutex_destroy() failed %E", err);
+    }
+
+    nxt_thread_log_debug("pthread_mutex_destroy(%p)", mtx);
+}
+
+
+nxt_int_t
+nxt_thread_mutex_lock(nxt_thread_mutex_t *mtx)
+{
+    nxt_err_t  err;
+
+    nxt_thread_log_debug("pthread_mutex_lock(%p) enter", mtx);
+
+    err = pthread_mutex_lock(mtx);
+    if (nxt_fast_path(err == 0)) {
+        return NXT_OK;
+    }
+
+    nxt_thread_log_alert("pthread_mutex_lock() failed %E", err);
+
+    return NXT_ERROR;
+}
+
+
+nxt_bool_t
+nxt_thread_mutex_trylock(nxt_thread_mutex_t *mtx)
+{
+    nxt_err_t  err;
+
+    nxt_thread_debug(thr);
+
+    nxt_log_debug(thr->log, "pthread_mutex_trylock(%p) enter", mtx);
+
+    err = pthread_mutex_trylock(mtx);
+    if (nxt_fast_path(err == 0)) {
+        return 1;
+    }
+
+    if (err == NXT_EBUSY) {
+        nxt_log_debug(thr->log, "pthread_mutex_trylock(%p) failed", mtx);
+
+    } else {
+        nxt_thread_log_alert("pthread_mutex_trylock() failed %E", err);
+    }
+
+    return 0;
+}
+
+
+nxt_int_t
+nxt_thread_mutex_unlock(nxt_thread_mutex_t *mtx)
+{
+    nxt_err_t     err;
+    nxt_thread_t  *thr;
+
+    err = pthread_mutex_unlock(mtx);
+
+    thr = nxt_thread();
+    nxt_thread_time_update(thr);
+
+    if (nxt_fast_path(err == 0)) {
+        nxt_log_debug(thr->log, "pthread_mutex_unlock(%p) exit", mtx);
+        return NXT_OK;
+    }
+
+    nxt_log_alert(thr->log, "pthread_mutex_unlock() failed %E", err);
+
+    return NXT_ERROR;
+}