From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: From: Martin Steigerwald Subject: Re: "Invalidate" not functional on Windows? Date: Tue, 14 Aug 2012 19:50:30 +0200 References: <502A4D53.7080801@kernel.dk> <201208141934.00289.Martin@lichtvoll.de> (sfid-20120814_194942_050719_5AEE14C7) In-Reply-To: <201208141934.00289.Martin@lichtvoll.de> MIME-Version: 1.0 Content-Type: Text/Plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Message-Id: <201208141950.30531.Martin@lichtvoll.de> To: Jens Axboe Cc: Greg Sullivan , fio@vger.kernel.org, Bruce Cran List-ID: Am Dienstag, 14. August 2012 schrieb Martin Steigerwald: > Am Dienstag, 14. August 2012 schrieb Jens Axboe: > > On 08/14/2012 06:57 AM, Greg Sullivan wrote: > > > I need to perform reads without using any caching. I have the line > > > "invalidate=3D1" in my job file, however it is still using the cache.= I > > > am running Windows 7 32-bit. Is the "invalidate" command supposed to > > > be functional on Windows? If so, what am I doing wrong? > > >=20 > > > That aside, what a great utility this is - thankyou!! > >=20 > > As far as what I remember from talking/mailing with Bruce (CC'ed), > > there's no way to invalidate the cache in Windows. >=20 > invalidate discards the buffer cache before the job group is started. > During the job execution caches will be in use anyway. Unless using > direct I/O (direct=3D1). Don=B4t know whether direct I/O will work with > "windowsaio" I/O engine or in any other way on Windows. Hey, its even documented. Except from HOWTO: direct=3Dbool If value is true, use non-buffered io. This is usually O_DIRECT. Note that ZFS on Solaris doesn't support direct i= o. On Windows the synchronous ioengines don't support direct i= o. =2D-=20 Martin 'Helios' Steigerwald - http://www.Lichtvoll.de GPG: 03B0 0D6C 0040 0710 4AFA B82F 991B EAAC A599 84C7