Re: 2.6.16-rc1-mm1

[Andrew Morton <akpm@osdl.org>]

Dave Jones <davej@redhat.com> wrote:
>
> With the patch above we'll mutex_unlock twice.

I was just testing you.

>  Is that allowed ? It sounds wrong to me.

It worked!


From: Andrew Morton <akpm@osdl.org>

Make the cpufreq code play nicely with the mutex debugging code: don't free a
held mutex.

Cc: Dave Jones <davej@codemonkey.org.uk>
Cc: Ingo Molnar <mingo@elte.hu>
Signed-off-by: Andrew Morton <akpm@osdl.org>
---

 drivers/cpufreq/cpufreq.c |    6 ++++--
 1 files changed, 4 insertions(+), 2 deletions(-)

diff -puN drivers/cpufreq/cpufreq.c~cpufreq-mutex-locking-fix drivers/cpufreq/cpufreq.c
--- devel/drivers/cpufreq/cpufreq.c~cpufreq-mutex-locking-fix	2006-01-18 13:37:09.000000000 -0800
+++ devel-akpm/drivers/cpufreq/cpufreq.c	2006-01-18 13:38:20.000000000 -0800
@@ -612,6 +612,7 @@ static int cpufreq_add_dev (struct sys_d
 	ret = cpufreq_driver->init(policy);
 	if (ret) {
 		dprintk("initialization failed\n");
+		mutex_unlock(&policy->lock);
 		goto err_out;
 	}
 
@@ -623,9 +624,10 @@ static int cpufreq_add_dev (struct sys_d
 	strlcpy(policy->kobj.name, "cpufreq", KOBJ_NAME_LEN);
 
 	ret = kobject_register(&policy->kobj);
-	if (ret)
+	if (ret) {
+		mutex_unlock(&policy->lock);
 		goto err_out_driver_exit;
-
+	}
 	/* set up files for this cpu device */
 	drv_attr = cpufreq_driver->attr;
 	while ((drv_attr) && (*drv_attr)) {
_