Date: Wed, 3 Sep 2008 16:01:49 +0300 From: Kostik Belousov <kostikbel@gmail.com> To: Igor Sysoev <is@rambler-co.ru> Cc: Tor Egge <tegge@freebsd.org>, freebsd-stable@freebsd.org Subject: Re: vfs.ffs.rawreadahead Message-ID: <20080903130149.GF2038@deviant.kiev.zoral.com.ua> In-Reply-To: <20080903124733.GH62541@rambler-co.ru> References: <20080903095352.GA62541@rambler-co.ru> <20080903123955.GE2038@deviant.kiev.zoral.com.ua> <20080903124733.GH62541@rambler-co.ru>
next in thread | previous in thread | raw e-mail | index | archive | help
--WCiRD21tF6xAtlSs Content-Type: text/plain; charset=us-ascii Content-Disposition: inline Content-Transfer-Encoding: quoted-printable On Wed, Sep 03, 2008 at 04:47:33PM +0400, Igor Sysoev wrote: > On Wed, Sep 03, 2008 at 03:39:55PM +0300, Kostik Belousov wrote: >=20 > > On Wed, Sep 03, 2008 at 01:53:52PM +0400, Igor Sysoev wrote: > > > Hi, > > >=20 > > > could anyone tell what does vfs.ffs.rawreadahead enable ? > > > As I understand it's used in DIRECTIO code that allows read data > > > directly to an userland buffer bypassing the buffer cache. > > > What I can not understand where the read ahead data can be placed in ? > >=20 > > The operation of the ffs_rawread is more accurately described as > > bypassing the page cache. It creates the physical buffer that maps > > the user pages. > >=20 > > The readahead is performed only when the supplied user memory region > > is bigger then blocksize. In this case, two reads are performed > > simultaneously, with both buffers mapping consequent blocks from > > user-supplied buffers. The read operation looks like footsteps. >=20 > Nice! >=20 > As I understand the size limit of one read operation is MAXPHYS, which is > equal to 128K due to LBA28 ATA limit. On SCSI, SATA, and LBA48 ATA this l= imit > can be increased. Is it safe ? As I understand, mnt_iosize_max value, that is topped at MAXPHYS, is used to limit the size of the buffer cluster used for clustered reads/writes. This readw clustering sometimes is called readahead too, but it seems to not be directly related to readahead done by ffs_rawread.c. --WCiRD21tF6xAtlSs Content-Type: application/pgp-signature Content-Disposition: inline -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.9 (FreeBSD) iEYEARECAAYFAki+ir0ACgkQC3+MBN1Mb4jP/QCeO//8yFyh+RvU8FfDnrH4MifQ k0YAoJoe4Mge7t3qehJJKqfbPPlqDC0g =MyOq -----END PGP SIGNATURE----- --WCiRD21tF6xAtlSs--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20080903130149.GF2038>