Date: Wed, 15 Dec 2004 21:34:52 -0500 (EST) From: Protius <protius@teryx.bobdbob.com> To: freebsd-firewire@freebsd.org, tcpdump23@hotmail.com Subject: Re: Firewire SAN Message-ID: <200412160234.iBG2YqFi071556@teryx.bobdbob.com> In-Reply-To: <BAY24-F193178173B8D0C147124B8A5A80@phx.gbl>
next in thread | previous in thread | raw e-mail | index | archive | help
>I'm considering taking up the idea of firewire-based SAN as a graduation >project. As I was looking at T13 (http://www.t13.org) and have found a >document about firewire-ATA attachment, so I was wondering if anybody out >there is already working on it as an opensource project (I know commercial >implementations of this idea exist). >More importantly, I was seeking the insight of the expertise in this mailing >list regarding the viability of this project and the difficulty (or lack >thereof) associated with it. Its not firewire <-> ATA, but you may want to look at scsi_target (in /usr/share/examples/scsi_target) and sbp_targ(4). (On a FreeBSD 5.3 box) There is a bug in scsi_target. I see it most easily when trying to write more than one block at a time onto a scsi_target. What happens is that there is a completed io queue (cmplt_io), and the same node gets inserted into it more than once, such that the TAILQ_FOREACH_REVERSE loop at line 697 of scsi_target.c (version 1.16) never exits. I havn't had time (day job, etc.) to finish hunting down exactly whats going on... But theres a heads up, if you look at it. -- -Tommy KE4ILZ "Microsoft wants shareholders. Open source programmers tjohnson@bobdbob.com "want the damn thing to work." - Cliff Sarginson Bottom posting makes you more readable. http://www.caliburn.nl/topposting.html
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200412160234.iBG2YqFi071556>