From: Jens Axboe <jens.axboe@oracle.com>
Date: Wed, 1 Jul 2009 21:02:10 +0000 (+0200)
Subject: Add code to detect a task that exited prior to up'ing the startup mutex
X-Git-Tag: fio-1.29~1^2
X-Git-Url: https://git.kernel.dk/?a=commitdiff_plain;h=656b1393d43f9f22738404582ea14dec956aea83;p=fio.git

Add code to detect a task that exited prior to up'ing the startup mutex

Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
---

diff --git a/fio.c b/fio.c
index 63d9a1b7..fa5deabb 100644
--- a/fio.c
+++ b/fio.c
@@ -1492,7 +1492,13 @@ static void run_threads(void)
 					*fio_debug_jobp = pid;
 			}
 			dprint(FD_MUTEX, "wait on startup_mutex\n");
-			fio_mutex_down(startup_mutex);
+			if (fio_mutex_down_timeout(startup_mutex, 10)) {
+				log_err("fio: job startup hung? exiting.\n");
+				terminate_threads(TERMINATE_ALL);
+				fio_abort = 1;
+				nr_started--;
+				break;
+			}
 			dprint(FD_MUTEX, "done waiting on startup_mutex\n");
 		}
 
diff --git a/mutex.c b/mutex.c
index 1538f62c..a4474374 100644
--- a/mutex.c
+++ b/mutex.c
@@ -3,6 +3,7 @@
 #include <unistd.h>
 #include <stdlib.h>
 #include <fcntl.h>
+#include <time.h>
 #include <pthread.h>
 #include <sys/mman.h>
 
@@ -88,6 +89,30 @@ err:
 	return NULL;
 }
 
+int fio_mutex_down_timeout(struct fio_mutex *mutex, unsigned int seconds)
+{
+	struct timespec t;
+	int ret = 0;
+
+	clock_gettime(CLOCK_REALTIME, &t);
+	t.tv_sec += seconds;
+
+	pthread_mutex_lock(&mutex->lock);
+
+	while (!mutex->value && !ret) {
+		mutex->waiters++;
+		ret = pthread_cond_timedwait(&mutex->cond, &mutex->lock, &t);
+		mutex->waiters--;
+	}
+
+	if (!ret) {
+		mutex->value--;
+		pthread_mutex_unlock(&mutex->lock);
+	}
+
+	return ret;
+}
+
 void fio_mutex_down(struct fio_mutex *mutex)
 {
 	pthread_mutex_lock(&mutex->lock);
diff --git a/mutex.h b/mutex.h
index 7be0ab1b..ac79dc63 100644
--- a/mutex.h
+++ b/mutex.h
@@ -15,6 +15,7 @@ struct fio_mutex {
 extern struct fio_mutex *fio_mutex_init(int);
 extern void fio_mutex_remove(struct fio_mutex *);
 extern void fio_mutex_down(struct fio_mutex *);
+extern int fio_mutex_down_timeout(struct fio_mutex *, unsigned int);
 extern void fio_mutex_down_read(struct fio_mutex *);
 extern void fio_mutex_down_write(struct fio_mutex *);
 extern void fio_mutex_up(struct fio_mutex *);