From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andreas Hirstius Date: Wed, 06 Apr 2005 08:33:07 +0000 Subject: Re: [Gelato-technical] Serious performance degradation on a RAID Message-Id: <42539EC3.9000305@cern.ch> List-Id: References: <16978.62532.841151.100745@napali.hpl.hp.com> In-Reply-To: <16978.62532.841151.100745@napali.hpl.hp.com> MIME-Version: 1.0 Content-Type: text/plain; charset="windows-1252" Content-Transfer-Encoding: quoted-printable To: linux-ia64@vger.kernel.org Chen, Kenneth W wrote: >On Tue, Apr 05, 2005 at 09:45:34PM -0700, David Mosberger wrote: > =20 > >>Did you compile the driver as a module? >> =20 >> > > >Duraid Madina wrote on Tuesday, April 05, 2005 9:53 PM > =20 > >>Nope. You're not seeing the problem? >> =20 >> > >That is the reason for the performance problem. And this is definitely >*not* ia64 specific. See this discussion which started with performance >problem seen on opteron and reproduced on DELL machine as well. >http://marc.theaimsgroup.com/?t=111028927300002&r=3D1&w=3D2 > >The changeset that david pointed out is the right fix. Or you can just >take the small patch I posted in the above thread. > > =20 > The problems mentioned in this discussion about the Fusion-MPT I can=20 perfectly reproduce on my boxes with 2.6.11, but only with pages sizes=20 4k and 8k; With 16k and 64k pages I don't see any problem at all !!! Actually I see block size <-> page size correlations which are similiar=20 to the correlations I see in my setup... i.e. 2.6.11 and 4k pages: dd ...bs@96: 40MB/s dd ...bs=8192: 74MB/s i.e. 2.6.11 and 8k pages: dd ...bs=8192: 56MB/s dd ...bs=16384: 74MB/s =20 with the driver as module I don't see any change (??). And I also don't see any change in performance for 4k/8k pages, when=20 applying either your small patch or the cset david pointed to... Since the fusion-mpt driver in 2.6.12-rc2 has this cset applied and no=20 other changes, I also don't see improvements with this kernel... (Again with 16k or 64k pages no performance problem in 2.6.12-rc2 with=20 the fusion-mpt driver...) But the driver in questions is 3w-9xxx and there have been no chances at=20 all in this driver between 2.6.10 and 2.6.12-rc1. The changed driver in =20 -rc2 still has the problem. And I've tried the driver built-in and as a module (default)... Andreas >- Ken > > > > =20 >