Date: Sat, 25 Apr 1998 01:53:28 -0500 (EST) From: "John S. Dyson" <toor@dyson.iquest.net> To: gibbs@narnia.plutotech.com (Justin T. Gibbs) Cc: dkelly@HiWAAY.net, scsi@FreeBSD.ORG Subject: Re: does CAM do this? Message-ID: <199804250653.BAA07149@dyson.iquest.net> In-Reply-To: <199804250607.AAA03997@narnia.plutotech.com> from "Justin T. Gibbs" at "Apr 25, 98 00:07:00 am"
next in thread | previous in thread | raw e-mail | index | archive | help
> > There are two issues here. > > 1) MAX_BSIZE is 64k. This is the largest a buffer can be in the > FreeBSD kernel unless you take advantage of some "modifications" > John Dyson did that can allow larger buffers to be constructed. > We really need to swith to a "buffer chaining" scheme (which > appears to be the solution BSDI chose) to allow larger transactions. > The modifications are in the code, and only require that they be enabled. I agree that a buffer chaining scheme is probably more desirable, but right now, in -current, IDE transfers are no longer limited to 64K. The variable is MAXPHYS, and is currently 128K. I would not suggest growing the maximum transfer size to much greater than this without Justin's change though. There is NO additional kernel cost for the SCSI code to support 128K transfer sizes as of today (actually a few months ago.) I don't mind if the current scheme is changed, as long as someone else does it :-). (We don't do the bogus buffer pagemove thing like the original code does, trying to use buffer cache buffers for clustering. We use a scheme, where we have physio buffers that can be any size of KVA, currently fixed due to lazyness to 128K (MAXPHYS). Physio buffers are used for all raw-type (and clustered ) disk I/O.) John To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-scsi" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?199804250653.BAA07149>