From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from c17870.thoms1.vic.optusnet.com.au ([210.49.248.224] helo=mail.kolivas.org) by pentafluge.infradead.org with esmtp (Exim 4.14 #3 (Red Hat Linux)) id 19M1bF-0002S2-0e for ; Sat, 31 May 2003 09:20:25 +0100 From: Con Kolivas To: =?iso-8859-1?q?J=F6rn=20Engel?= , "David S. Miller" Date: Sat, 31 May 2003 18:22:21 +1000 References: <20030530174319.GA16687@wohnheim.fh-wedel.de> <20030530.235505.23020750.davem@redhat.com> <20030531075615.GA25089@wohnheim.fh-wedel.de> In-Reply-To: <20030531075615.GA25089@wohnheim.fh-wedel.de> MIME-Version: 1.0 Content-Disposition: inline Message-Id: <200305311822.21823.kernel@kolivas.org> Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable cc: linux-mtd@lists.infradead.org cc: linux-kernel@vger.kernel.org cc: jmorris@intercode.com.au Subject: Re: [PATCH RFC] 1/2 central workspace for zlib List-Id: Linux MTD discussion mailing list List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , On Sat, 31 May 2003 17:56, J=F6rn Engel wrote: > CC List pruned a little. > > On Fri, 30 May 2003 23:55:05 -0700, David S. Miller wrote: > > From: J=F6rn Engel > > > > How about preemption? zlib operations take their time, so at least = on > > up, it makes sense to preempt them, when not in softirq context. Can > > this still be done lockless? > > > > You'll need to disable preemption. > > My gut feeling claims that this would hurt interactivity. Con, would > contest on a jffs2 (zlib compressed) filesystem be able to show > interactivity problems wrt zlib? The only way I could think of was perhaps using a load on another disk=20 (io_other which is in contest) that is using jffs2 when the contest baselin= e=20 is running on a normal filesystem - this has shown very little differences= =20 between filesystems normally. Otherwise if everything in contest is run on= =20 jffs2 it would affect every layer and hard to be sure you had a control to= =20 compare with. Con