Date: Tue, 8 Aug 2017 23:49:11 -0700
From: Davidlohr Bueso <dave@stgolabs.net>
To: Mel Gorman <mgorman@suse.de>
Cc: Thomas Gleixner <tglx@linutronix.de>,
        Mark Rutland <mark.rutland@arm.com>,
        Linus Torvalds <torvalds@linux-foundation.org>,
        Ingo Molnar <mingo@kernel.org>, Davidlohr Bueso <dbueso@suse.de>,
        Hugh Dickins <hughd@google.com>, Peter Zijlstra <peterz@infradead.org>,
        Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
        linux-kernel@vger.kernel.org
Subject: Re: [PATCH] futex: Remove unnecessary warning from get_futex_key
Message-ID: <20170809064911.GA31085@linux-80c1.suse>
References: <20170809062146.yqvn35t2jags65f3@suse.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Disposition: inline
In-Reply-To: <20170809062146.yqvn35t2jags65f3@suse.de>
User-Agent: Mutt/1.5.24 (2015-08-30)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1251
Lines: 32

On Wed, 09 Aug 2017, Mel Gorman wrote:

>@@ -676,7 +676,7 @@ get_futex_key(u32 __user *uaddr, int fshared, union futex_key *key, int rw)
> 		 * cases, therefore a successful atomic_inc return below will
> 		 * guarantee that get_futex_key() will still imply smp_mb(); (B).

You missed the comment above.

diff --git a/kernel/futex.c b/kernel/futex.c
index 16dbe4c93895..6b4a6a7cad3d 100644
--- a/kernel/futex.c
+++ b/kernel/futex.c
@@ -670,13 +670,14 @@ get_futex_key(u32 __user *uaddr, int fshared, union futex_key *key, int rw)
 		 * this reference was taken by ihold under the page lock
 		 * pinning the inode in place so i_lock was unnecessary. The
 		 * only way for this check to fail is if the inode was
-		 * truncated in parallel so warn for now if this happens.
+		 * truncated in parallel -- which is a bizarre scenario, in
+		 * any case, just retry.
 		 *
 		 * We are not calling into get_futex_key_refs() in file-backed
 		 * cases, therefore a successful atomic_inc return below will
 		 * guarantee that get_futex_key() will still imply smp_mb(); (B).
 		 */
-		if (WARN_ON_ONCE(!atomic_inc_not_zero(&inode->i_count))) {
+		if (!atomic_inc_not_zero(&inode->i_count)) {
 			rcu_read_unlock();
 			put_page(page);
 

Thanks,
Davidlohr