[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <20180216171013.GI25201@hirez.programming.kicks-ass.net>
Date: Fri, 16 Feb 2018 18:10:13 +0100
From: Peter Zijlstra <peterz@...radead.org>
To: John Ogness <john.ogness@...utronix.de>
Cc: linux-fsdevel@...r.kernel.org, Al Viro <viro@...iv.linux.org.uk>,
Linus Torvalds <torvalds@...ux-foundation.org>,
Christoph Hellwig <hch@....de>,
Thomas Gleixner <tglx@...utronix.de>,
Sebastian Andrzej Siewior <bigeasy@...utronix.de>,
linux-kernel@...r.kernel.org
Subject: Re: [PATCH 3/4] fs/dcache: Avoid the try_lock loop in d_delete()
On Fri, Feb 16, 2018 at 04:09:32PM +0100, John Ogness wrote:
> +static bool dentry_lock_inode(struct dentry *dentry)
> +{
> + struct inode *inode = dentry->d_inode;
> +
> + lockdep_assert_held(&dentry->d_lock);
> +
> + if (unlikely(!spin_trylock(&inode->i_lock))) {
if (likely(spin_trylock(&inode->i_lock)))
return true;
and then unindent by 1 stop the below code:
> + rcu_read_lock();
> + spin_unlock(&dentry->d_lock);
> + spin_lock(&inode->i_lock);
> + spin_lock(&dentry->d_lock);
> + rcu_read_unlock();
> +
> + /*
> + * @dentry->d_inode might have changed after dropping
> + * @dentry->d_lock. If so, release @inode->i_lock and
> + * signal the caller to restart the operation.
> + */
> + if (unlikely(inode != dentry->d_inode)) {
> + spin_unlock(&inode->i_lock);
> + return false;
> + }
> + }
> + return true;
> +}
Powered by blists - more mailing lists