[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-Id: <20171108.134053.127884187686446469.davem@davemloft.net>
Date: Wed, 08 Nov 2017 13:40:53 +0900 (KST)
From: David Miller <davem@...emloft.net>
To: jon.maloy@...csson.com
Cc: netdev@...r.kernel.org,
mohan.krishna.ghanta.krishnamurthy@...csson.com,
tung.q.nguyen@...tech.com.au, hoang.h.le@...tech.com.au,
canh.d.luu@...tech.com.au, ying.xue@...driver.com,
tipc-discussion@...ts.sourceforge.net
Subject: Re: [net-next 1/1] tipc: improve link resiliency when rps is
activated
From: Jon Maloy <jon.maloy@...csson.com>
Date: Mon, 6 Nov 2017 12:44:11 +0100
> + return core & (num_online_cpus() - 1);
Well, this isn't exactly correct. The number of online cpus is not
necessarily a power of two.
And furthermore it shouldn't even be necessary. Just pass the whole
key around, it will have the modulus applied to it when needed
elsewhere in the call chains.
Powered by blists - more mailing lists