From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexander Graf Subject: Re: [Qemu-ppc] [RFC PATCH 12/17] PowerPC: booke64: Add DO_KVM kernel hooks Date: Thu, 12 Jul 2012 00:51:56 +0200 Message-ID: <52D81ED1-AEF7-43C8-84A4-98E1D895CD0B@suse.de> References: <1340627195-11544-1-git-send-email-mihai.caraman@freescale.com> <1340627195-11544-13-git-send-email-mihai.caraman@freescale.com> <1B2CBB56-7180-4A73-8E51-6538A725F710@suse.de> <1341440735.16808.42.camel@pasglop> <300B73AA675FCE4A93EB4FC1D42459FF15CDE6@039-SN2MPN1-013.039d.mgd.msft.net> ,<7305F50A-8E77-4E88-8EB8-4046A7E94DF9@suse.de> <300B73AA675FCE4A93EB4FC1D42459FF15CF83@039-SN2MPN1-013.039d.mgd.msft.net> <1342045680.18850.39.camel@pasglop> <1342046584.18850.40.camel@pasglop> Mime-Version: 1.0 (Apple Message framework v1278) Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 8BIT Cc: Caraman Mihai Claudiu-B02008 , "" , KVM list , linuxppc-dev , "qemu-ppc@nongnu.org List" To: Benjamin Herrenschmidt Return-path: In-Reply-To: <1342046584.18850.40.camel@pasglop> Sender: kvm-ppc-owner@vger.kernel.org List-Id: kvm.vger.kernel.org On 12.07.2012, at 00:43, Benjamin Herrenschmidt wrote: > On Thu, 2012-07-12 at 00:35 +0200, Alexander Graf wrote: >>> On Thu, 2012-07-12 at 00:25 +0200, Alexander Graf wrote: >>>> Or just free up another register early on, like here. >>> >>> If you're going to do that, you want to measure the impact on null >>> syscall performance though. >> >> That should hold true for any change in that code, no? > > Yes, but adding a gpr save is more invasive (you have to also load it > back later on & put it on the stack). What's the usual cache line size like on these boxes? If you align it properly with r10 and r11 on the paca so that all 3 registers are in the same cache line, the load should be almost for free, no? Alex From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mx2.suse.de (cantor2.suse.de [195.135.220.15]) (using TLSv1 with cipher DHE-RSA-AES256-SHA (256/256 bits)) (Client CN "mx2.suse.de", Issuer "CAcert Class 3 Root" (not verified)) by ozlabs.org (Postfix) with ESMTPS id D3D4B2C00E7 for ; Thu, 12 Jul 2012 08:52:04 +1000 (EST) Subject: Re: [Qemu-ppc] [RFC PATCH 12/17] PowerPC: booke64: Add DO_KVM kernel hooks Mime-Version: 1.0 (Apple Message framework v1278) Content-Type: text/plain; charset=us-ascii From: Alexander Graf In-Reply-To: <1342046584.18850.40.camel@pasglop> Date: Thu, 12 Jul 2012 00:51:56 +0200 Message-Id: <52D81ED1-AEF7-43C8-84A4-98E1D895CD0B@suse.de> References: <1340627195-11544-1-git-send-email-mihai.caraman@freescale.com> <1340627195-11544-13-git-send-email-mihai.caraman@freescale.com> <1B2CBB56-7180-4A73-8E51-6538A725F710@suse.de> <1341440735.16808.42.camel@pasglop> <300B73AA675FCE4A93EB4FC1D42459FF15CDE6@039-SN2MPN1-013.039d.mgd.msft.net> , <7305F50A-8E77-4E88-8EB8-4046A7E94DF9@suse.de> <300B73AA675FCE4A93EB4FC1D42459FF15CF83@039-SN2MPN1-013.039d.mgd.msft.net> <1342045680.18850.39.camel@pasglop> <1342046584.18850.40.camel@pasglop> To: Benjamin Herrenschmidt Cc: "qemu-ppc@nongnu.org List" , Caraman Mihai Claudiu-B02008 , linuxppc-dev , KVM list , "" List-Id: Linux on PowerPC Developers Mail List List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On 12.07.2012, at 00:43, Benjamin Herrenschmidt wrote: > On Thu, 2012-07-12 at 00:35 +0200, Alexander Graf wrote: >>> On Thu, 2012-07-12 at 00:25 +0200, Alexander Graf wrote: >>>> Or just free up another register early on, like here. >>>=20 >>> If you're going to do that, you want to measure the impact on null >>> syscall performance though. >>=20 >> That should hold true for any change in that code, no? >=20 > Yes, but adding a gpr save is more invasive (you have to also load it > back later on & put it on the stack). What's the usual cache line size like on these boxes? If you align it = properly with r10 and r11 on the paca so that all 3 registers are in the = same cache line, the load should be almost for free, no? Alex From mboxrd@z Thu Jan 1 00:00:00 1970 From: Alexander Graf Date: Wed, 11 Jul 2012 22:51:56 +0000 Subject: Re: [Qemu-ppc] [RFC PATCH 12/17] PowerPC: booke64: Add DO_KVM kernel hooks Message-Id: <52D81ED1-AEF7-43C8-84A4-98E1D895CD0B@suse.de> List-Id: References: <1340627195-11544-1-git-send-email-mihai.caraman@freescale.com> <1340627195-11544-13-git-send-email-mihai.caraman@freescale.com> <1B2CBB56-7180-4A73-8E51-6538A725F710@suse.de> <1341440735.16808.42.camel@pasglop> <300B73AA675FCE4A93EB4FC1D42459FF15CDE6@039-SN2MPN1-013.039d.mgd.msft.net> ,<7305F50A-8E77-4E88-8EB8-4046A7E94DF9@suse.de> <300B73AA675FCE4A93EB4FC1D42459FF15CF83@039-SN2MPN1-013.039d.mgd.msft.net> <1342045680.18850.39.camel@pasglop> <1342046584.18850.40.camel@pasglop> In-Reply-To: <1342046584.18850.40.camel@pasglop> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Benjamin Herrenschmidt Cc: Caraman Mihai Claudiu-B02008 , "" , KVM list , linuxppc-dev , "qemu-ppc@nongnu.org List" On 12.07.2012, at 00:43, Benjamin Herrenschmidt wrote: > On Thu, 2012-07-12 at 00:35 +0200, Alexander Graf wrote: >>> On Thu, 2012-07-12 at 00:25 +0200, Alexander Graf wrote: >>>> Or just free up another register early on, like here. >>> >>> If you're going to do that, you want to measure the impact on null >>> syscall performance though. >> >> That should hold true for any change in that code, no? > > Yes, but adding a gpr save is more invasive (you have to also load it > back later on & put it on the stack). What's the usual cache line size like on these boxes? If you align it properly with r10 and r11 on the paca so that all 3 registers are in the same cache line, the load should be almost for free, no? Alex