Re: [RFC][PATCH 3/8] move mnt_want_write() out of may_open()

[Dave Hansen <haveblue@us.ibm.com>]

On Mon, 2007-10-01 at 21:55 +0200, Miklos Szeredi wrote:
> > @@ -1687,10 +1679,8 @@ static int open_namei_create(struct name
> >  	struct dentry *dir = nd->dentry;
> >  
> >  	/*
> > -	 * This ensures that the mnt stays writable
> > -	 * over the vfs_create() call to may_open(),
> > -	 * which takes a more persistent
> > -	 * mnt_want_write().
> > +	 * This mnt_want_write() is potentially persistent,
> > +	 * and balanced in __fput()
> >  	 */
> >  	error = mnt_want_write(nd->mnt);
> >  	if (error) {
> 
> I'm confused: isn't it the mnt_want_write() in __dentry_open(), that
> is balanced in __fput()?

This is broken.  I didn't realize that nameidata_to_filp() called
dentry_open.  I'll rework these.

-- Dave

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/