From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756960Ab0EaS5P (ORCPT ); Mon, 31 May 2010 14:57:15 -0400 Received: from mail.virtall.com ([178.63.195.102]:43664 "EHLO mail.virtall.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756893Ab0EaS5O (ORCPT ); Mon, 31 May 2010 14:57:14 -0400 X-Greylist: delayed 579 seconds by postgrey-1.27 at vger.kernel.org; Mon, 31 May 2010 14:57:14 EDT DomainKey-Signature: a=rsa-sha1; c=nofws; d=wpkg.org; h=message-id:date :from:mime-version:to:cc:subject:content-type :content-transfer-encoding; q=dns; s=default; b=Vzy2//120efCTpMA VS+6zn02hhblG4XYhWmeulfNj6GcrmNCK7IYUHBnkyuAwuOn5jCowlxqNwSjryKK 8gb6VSLZWNCbh0gUg4A26H3N0QO4YsdWcZntu95uN4q2Vsntprqcc4m9UWFStdqi 6xhyYZcvXqF+PtQhB82UfTdqoBM= Message-ID: <4C040441.9020107@wpkg.org> Date: Mon, 31 May 2010 20:47:29 +0200 From: Tomasz Chmielewski User-Agent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.9) Gecko/20100525 Thunderbird/3.0.4 MIME-Version: 1.0 To: linux-kernel@vger.kernel.org CC: mbroz@redhat.com, Andi Kleen , Eric Dumazet Subject: Re: [PATCH] DM-CRYPT: Scale to multiple CPUs Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Milan Broz wrote: > On 05/31/2010 06:04 PM, Andi Kleen wrote: >> DM-CRYPT: Scale to multiple CPUs >> >> Currently dm-crypt does all encryption work per dmcrypt mapping in a >> single workqueue. This does not scale well when multiple CPUs >> are submitting IO at a high rate. The single CPU running the single >> thread cannot keep up with the encryption and encrypted IO performance >> tanks. > > This is true only if encryption run on the CPU synchronously. > > (Usually it is high speed SSD or dm-crypt above striped RAID where > underlying device throughput is higher than CPU encryption speed.) Quite common to hit on a netbook/laptop with encrypted root these days, where storage can be relatively fast when compared to the speed of one of the cores of the (usually) multicore CPU. -- Tomasz Chmielewski http://wpkg.org