[dm-devel] Re: [PATCH 2.6.19 5/5] fs: freeze_bdev with semaphore not mutex

[Andrew Morton]

On Tue, 7 Nov 2006 18:34:59 +0000
Alasdair G Kergon <agk at redhat.com> wrote:

> From: Srinivasa Ds <srinivasa at in.ibm.com>
> 
> On debugging I found out that,"dmsetup suspend <device name>" calls
> "freeze_bdev()",which locks "bd_mount_mutex" to make sure that no new mounts
> happen on bdev until thaw_bdev() is called.  This "thaw_bdev()" is getting
> called when we resume the device through "dmsetup resume <device-name>".
> Hence we have 2 processes,one of which locks "bd_mount_mutex"(dmsetup
> suspend) and another(dmsetup resume) unlocks it.                                               

So...  what does this have to do with switching from mutex to semaphore?

Perhaps this works around the debugging code which gets offended if a mutex
is unlocked by a process which didn't do the lock?

If so, it's a bit sad to switch to semaphore just because of some errant
debugging code.  Perhaps it would be better to create a new
mutex_unlock_stfu() which suppresses the warning?


> --- linux-2.6.19-rc4.orig/fs/buffer.c	2006-11-07 17:06:20.000000000 +0000
> +++ linux-2.6.19-rc4/fs/buffer.c	2006-11-07 17:26:04.000000000 +0000
> @@ -188,7 +188,9 @@ struct super_block *freeze_bdev(struct b
>  {
>  	struct super_block *sb;
>  
> -	mutex_lock(&bdev->bd_mount_mutex);
> +	if (down_trylock(&bdev->bd_mount_sem))
> +		return -EBUSY;
> +

This is a functional change which isn't described in the changelog.  What's
happening here?