Date: Thu, 31 Jul 2003 16:14:27 +0200 From: Andre Albsmeier <andre.albsmeier@siemens.com> To: "Marc G. Fournier" <scrappy@hub.org> Cc: aic7xxx@freebsd.org Subject: Re: AIC7902 w/ seagate U320 drive issue on releng-4 (and current) Message-ID: <20030731141427.GC9802@curry.mchp.siemens.de> In-Reply-To: <20030729134931.Q5355@hub.org> References: <FE045D4D9F7AED4CBFF1B3B813C853370274203A@mail.sandvine.com> <20030729134931.Q5355@hub.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Tue, 29-Jul-2003 at 13:50:47 -0300, Marc G. Fournier wrote: > On Mon, 28 Jul 2003, Don Bowman wrote: > > > > > > From: The Hermit Hacker [mailto:scrappy@hub.org] > > > > > > Just as an FYI, we just downloaded the 006 firmware this afternoon ... > > > > > > > Can you please confirm this is 006? Seagate has just re-confirmed > > that 005 is the latest firmware available for these drives. > > just upgraded all 6 drives: > > mars# camcontrol devlist > <SEAGATE ST336607LC 0006> at scbus1 target 0 lun 0 (pass0,da0) > <SEAGATE ST336607LC 0006> at scbus1 target 1 lun 0 (pass1,da1) > <SEAGATE ST336607LC 0006> at scbus1 target 2 lun 0 (pass2,da2) > <SEAGATE ST336607LC 0006> at scbus1 target 3 lun 0 (pass3,da3) > <SEAGATE ST336607LC 0006> at scbus1 target 4 lun 0 (pass4,da4) > <SEAGATE ST336607LC 0006> at scbus1 target 5 lun 0 (pass5,da5) > > fwprog works great, but warning for anyone else using it ... pass1-5 went > fine, but since pass0 was my 'root drive', it required a cold boot after > the upgrade, as it seemed to have locked up the bus ... Yes. If have seen this also when playing around. It is generally a good idea to unmount the drive in question. When updating the root drive, single user mode is preferrable. The code was a private, experimental hack. I wanted to understand what Marc Frajola is doing there and how it works. Later I added support for other vendors. Maybe it would be a good idea to spit out a warning message at the start -- reminding the user that nothing else should access the drive during the upgrade. However, the nicest thing would be if I could, let's say, lock the device for exclusive use but I have no idea how this could be done. I don't think I can simply pass O_EXLOCK to the cam_open_spec_device call :-). -Andre
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20030731141427.GC9802>