From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tony Lindgren Subject: Re: [PATCH] omap4: enable L2 prefetching Date: Tue, 16 Nov 2010 10:11:57 -0800 Message-ID: <20101116181157.GY9264@atomide.com> References: <1289838044-32016-1-git-send-email-nm@ti.com> Mime-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: QUOTED-PRINTABLE Return-path: Received: from mho-02-ewr.mailhop.org ([204.13.248.72]:45399 "EHLO mho-02-ewr.mailhop.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1754817Ab0KPSMG (ORCPT ); Tue, 16 Nov 2010 13:12:06 -0500 Content-Disposition: inline In-Reply-To: Sender: linux-omap-owner@vger.kernel.org List-Id: linux-omap@vger.kernel.org To: =?utf-8?B?TcOlbnMgUnVsbGfDpXJk?= Cc: linux-omap@vger.kernel.org, linux-arm-kernel@lists.infradead.org * M=C3=A5ns Rullg=C3=A5rd [101115 09:01]: > Nishanth Menon writes: >=20 > > From: Mans Rullgard > > > > Enabling L2 prefetching improves performance as shown on Panda > > ES2.1 board with mem test, and it has measurable impact on > > performances. I think we should consider it, even though it damages > > "writes" a bit. (rebased to k.org) > > Usually the prefetch is used at both levels together L1 + L2, howev= er, > > to enable the CP15 prefetch engines, these are under security, and = on > > GP devices, we cannot enable it(e.g. on PandaBoard). However, just > > enabling PL310 prefetch seems to provide performance improvement, > > as shown in the data below (from Ubuntu) and would be a great thing > > to pull in. >=20 > What this does is enable automatic next line prefetching. With this > enabled, whenever the PL310 receives a cachable read request, it > automatically prefetches the following cache line as well. A larger > offset can be programmed in secure mode, but the TI ROM authors > neglected to include this. >=20 > Testing with FFmpeg showed a speedup of 10% with this patch in some > cases. M=C3=A5ns and Nishant, care to repost this with the updated comments? Regards, Tony -- To unsubscribe from this list: send the line "unsubscribe linux-omap" i= n the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html From mboxrd@z Thu Jan 1 00:00:00 1970 From: tony@atomide.com (Tony Lindgren) Date: Tue, 16 Nov 2010 10:11:57 -0800 Subject: [PATCH] omap4: enable L2 prefetching In-Reply-To: References: <1289838044-32016-1-git-send-email-nm@ti.com> Message-ID: <20101116181157.GY9264@atomide.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org * M?ns Rullg?rd [101115 09:01]: > Nishanth Menon writes: > > > From: Mans Rullgard > > > > Enabling L2 prefetching improves performance as shown on Panda > > ES2.1 board with mem test, and it has measurable impact on > > performances. I think we should consider it, even though it damages > > "writes" a bit. (rebased to k.org) > > Usually the prefetch is used at both levels together L1 + L2, however, > > to enable the CP15 prefetch engines, these are under security, and on > > GP devices, we cannot enable it(e.g. on PandaBoard). However, just > > enabling PL310 prefetch seems to provide performance improvement, > > as shown in the data below (from Ubuntu) and would be a great thing > > to pull in. > > What this does is enable automatic next line prefetching. With this > enabled, whenever the PL310 receives a cachable read request, it > automatically prefetches the following cache line as well. A larger > offset can be programmed in secure mode, but the TI ROM authors > neglected to include this. > > Testing with FFmpeg showed a speedup of 10% with this patch in some > cases. M?ns and Nishant, care to repost this with the updated comments? Regards, Tony