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;
Message-ID: <e63e6fc511d9d515fcb8501f48f218192f36afd3.camel@redhat.com>
Subject: Re: [PATCH] dn_route: use time_is_before_jiffies(a) to replace
 "jiffies - a > 0"
From:   Paolo Abeni <pabeni@redhat.com>
To:     Yu Zhe <yuzhe@nfschina.com>, davem@davemloft.net, kuba@kernel.org
Cc:     linux-decnet-user@lists.sourceforge.net, netdev@vger.kernel.org,
        linux-kernel@vger.kernel.org, kernel-janitors@vger.kernel.org,
        liqiong@nfschina.com
Date:   Thu, 28 Jul 2022 15:11:04 +0200
In-Reply-To: <20220727024600.18413-1-yuzhe@nfschina.com>
References: <20220727024600.18413-1-yuzhe@nfschina.com>
Content-Type: text/plain; charset="UTF-8"
User-Agent: Evolution 3.42.4 (3.42.4-2.fc35) 
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Precedence: bulk

On Wed, 2022-07-27 at 10:46 +0800, Yu Zhe wrote:
> time_is_before_jiffies deals with timer wrapping correctly.
> 
> Signed-off-by: Yu Zhe <yuzhe@nfschina.com>
> ---
>  net/decnet/dn_route.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/net/decnet/dn_route.c b/net/decnet/dn_route.c
> index 552a53f1d5d0..0a4bed0bc2e3 100644
> --- a/net/decnet/dn_route.c
> +++ b/net/decnet/dn_route.c
> @@ -201,7 +201,7 @@ static void dn_dst_check_expire(struct timer_list *unused)
>  		}
>  		spin_unlock(&dn_rt_hash_table[i].lock);
>  
> -		if ((jiffies - now) > 0)
> +		if (time_is_before_jiffies(now))

Uhmm... it looks like the wrap-around condition can happen only in
theory: 'now' is initialized just before entering this loop, and we
will break as soon as jiffies increment. The wrap-around could happen
only if a single iteration of the loop takes more than LONG_MAX
jiffies.

If that happens, we have a completely different kind of much more
serious problem, I think ;)

So this change is mostly for core readability's sake. I personally find
more readable:

		if (jiffies != now)

Cheers,

Paolo

p.s. given the above I guess this is for the net-next tree, right?