From: Herbert Xu Subject: Re: [RFC] per-CPU cryptd thread implementation based on workqueue Date: Thu, 22 Jan 2009 14:04:00 +1100 Message-ID: <20090122030400.GA10229@gondor.apana.org.au> References: <1232075437.13948.12.camel@yhuang-dev.sh.intel.com> <20090116033105.GB10390@gondor.apana.org.au> <1232591537.6101.3.camel@yhuang-dev.sh.intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Cc: "linux-kernel@vger.kernel.org" , "linux-crypto@vger.kernel.org" To: Huang Ying Return-path: Content-Disposition: inline In-Reply-To: <1232591537.6101.3.camel@yhuang-dev.sh.intel.com> Sender: linux-kernel-owner@vger.kernel.org List-Id: linux-crypto.vger.kernel.org On Thu, Jan 22, 2009 at 10:32:17AM +0800, Huang Ying wrote: > > This is the first attempt to use a dedicate workqueue for crypto. It is > not intended to be merged. Please feedback your comments, especially on > desgin. Thanks for the patch! > + spin_lock_init(&cpu_queue->lock); Since we're switching to per-cpu queues it would be good to just kill the spin lock. AFAICS the only place you really need it is in cryptd_tfm_in_queue. That's just used for debugging so we can just kill it and lose this spin lock. Cheers, -- Visit Openswan at http://www.openswan.org/ Email: Herbert Xu ~{PmV>HI~} Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt