From mboxrd@z Thu Jan 1 00:00:00 1970 From: Vineet Gupta Subject: Re: [PATCH v2 09/76] ARC: Checksum/byteorder/swab routines Date: Fri, 18 Jan 2013 19:56:12 +0530 Message-ID: <50F95B84.2010802@synopsys.com> References: <1358511930-7424-1-git-send-email-vgupta@synopsys.com> <1358511930-7424-10-git-send-email-vgupta@synopsys.com> <201301181421.09458.arnd@arndb.de> Mime-Version: 1.0 Content-Type: text/plain; charset="ISO-8859-15" Content-Transfer-Encoding: 7bit Return-path: Received: from us02smtp2.synopsys.com ([198.182.60.77]:61247 "EHLO alvesta.synopsys.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751915Ab3ARO15 (ORCPT ); Fri, 18 Jan 2013 09:27:57 -0500 In-Reply-To: <201301181421.09458.arnd@arndb.de> Sender: linux-arch-owner@vger.kernel.org List-ID: To: Arnd Bergmann Cc: linux-arch@vger.kernel.org, linux-kernel@vger.kernel.org On Friday 18 January 2013 07:51 PM, Arnd Bergmann wrote: > On Friday 18 January 2013, Vineet Gupta wrote: >> TBD: do_csum still needs to be written in asm > Do you actually expect a lot of improvement in do_csum? > I would hope that gcc can actually generate a pretty > good version of it, unless you have some 64-bit add-with-carry > instruction or something else that you could make use of. We do have 32bit add-with-carry and it has it's own set of micro-architectural hazards and stalls which deserve careful insn scheduling. So yes the mileage would vary but certainly deserves a try. > >> Signed-off-by: Vineet Gupta > Acked-by: Arnd Bergmann Thanks, -Vineet From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1756027Ab3ARO16 (ORCPT ); Fri, 18 Jan 2013 09:27:58 -0500 Received: from us02smtp2.synopsys.com ([198.182.60.77]:61247 "EHLO alvesta.synopsys.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751915Ab3ARO15 (ORCPT ); Fri, 18 Jan 2013 09:27:57 -0500 Message-ID: <50F95B84.2010802@synopsys.com> Date: Fri, 18 Jan 2013 19:56:12 +0530 From: Vineet Gupta User-Agent: Mozilla/5.0 (X11; Linux i686; rv:16.0) Gecko/20121011 Thunderbird/16.0.1 MIME-Version: 1.0 To: Arnd Bergmann CC: , Subject: Re: [PATCH v2 09/76] ARC: Checksum/byteorder/swab routines References: <1358511930-7424-1-git-send-email-vgupta@synopsys.com> <1358511930-7424-10-git-send-email-vgupta@synopsys.com> <201301181421.09458.arnd@arndb.de> In-Reply-To: <201301181421.09458.arnd@arndb.de> Content-Type: text/plain; charset="ISO-8859-15" Content-Transfer-Encoding: 7bit X-Originating-IP: [10.12.197.205] Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org On Friday 18 January 2013 07:51 PM, Arnd Bergmann wrote: > On Friday 18 January 2013, Vineet Gupta wrote: >> TBD: do_csum still needs to be written in asm > Do you actually expect a lot of improvement in do_csum? > I would hope that gcc can actually generate a pretty > good version of it, unless you have some 64-bit add-with-carry > instruction or something else that you could make use of. We do have 32bit add-with-carry and it has it's own set of micro-architectural hazards and stalls which deserve careful insn scheduling. So yes the mileage would vary but certainly deserves a try. > >> Signed-off-by: Vineet Gupta > Acked-by: Arnd Bergmann Thanks, -Vineet