Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
Date:   Mon, 11 Jul 2022 22:13:19 -0700
From:   Christoph Hellwig <hch@infradead.org>
To:     Uros Bizjak <ubizjak@gmail.com>
Cc:     linux-block@vger.kernel.org, linux-kernel@vger.kernel.org,
        Jens Axboe <axboe@kernel.dk>
Subject: Re: [PATCH] block: Use try_cmpxchg some more
Message-ID: <Ys0C72unFFlsWomq@infradead.org>
References: <20220711153301.2388-1-ubizjak@gmail.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20220711153301.2388-1-ubizjak@gmail.com>
Precedence: bulk

On Mon, Jul 11, 2022 at 05:33:01PM +0200, Uros Bizjak wrote:
> Use try_cmpxchg family of functions instead of cmpxchg (*ptr, old, new) == old.
> x86 CMPXCHG instruction returns success in ZF flag, so this change saves a
> compare after cmpxchg (and related move instruction in front of cmpxchg).
> 
> Also, try_cmpxchg implicitly assigns old *ptr value to "old" when
> cmpxchg fails, enabling further code simplifications.
> 
> No functional change intended.

You might want to split this into a patch per caller as it might
attact different reviewers.

> +	do {
> +	} while (old && !atomic_try_cmpxchg(&blkg->use_delay, &old, old - 1));

It might just be me, but for loops with an empty body this do { } while
construct looks odd.  Why not:

	while (old && !atomic_try_cmpxchg(&blkg->use_delay, &old, old - 1))
		;

?

The the use of the atomic on ->use_delay looks really whacky to start
with.  To me it sounds like it really wants to use a proper lock
instead of all this magic.

>  	else
>  		return;
>  
> -	old = atomic_cmpxchg(&iolat->scale_cookie, our_cookie, cur_cookie);
> -
> -	/* Somebody beat us to the punch, just bail. */
> -	if (old != our_cookie)
> +	if (!atomic_try_cmpxchg(&iolat->scale_cookie, &our_cookie, cur_cookie)) {
> +		/* Somebody beat us to the punch, just bail. */
>  		return;
> +	}


	/* If somebody beat us to the punch, just bail. */
	if (!atomic_try_cmpxchg(&iolat->scale_cookie, &our_cookie, cur_cookie))
		return;