Date: Sat, 13 Apr 2002 15:40:48 -0700 From: Paul Saab <ps@mu.org> To: Lukas Ertl <l.ertl@univie.ac.at> Cc: freebsd-scsi@freebsd.org Subject: Re: ciss driver and tagged queuing Message-ID: <20020413224048.GA6046@elvis.mu.org> In-Reply-To: <Pine.WNT.4.43.0204131315560.1784-100000@korben> References: <20020411091935.GA99318@elvis.mu.org> <Pine.WNT.4.43.0204131315560.1784-100000@korben>
next in thread | previous in thread | raw e-mail | index | archive | help
--HlL+5n6rz5pIUxbD Content-Type: text/plain; charset=iso-8859-1 Content-Disposition: inline Content-Transfer-Encoding: 8bit Yea.. Mike and I discovered that the CAM sim was being frozen but never being unfrozen. Can you try this patch? I still have not put this into FreeBSD because of other weirdness I am seeing on my DL380g2, but I am unable to reproduce the same problems w/ my DL360g2, so I'm going to put another DL380g2 in the rack on monday and see if it suffers from the lockups I am seeing or if that machine's array is just jacked. Lukas Ertl (l.ertl@univie.ac.at) wrote: > On Thu, 11 Apr 2002, Paul Saab wrote: > > > Can you try this patch? It should fix the performance problems you > > are having. > > Hi, > > I've seen that the patch already made it intp -STABLE, so I cvsupped > and rebuilt everything, but I'm afraid to say that things got worse. > Mounting the file system that is located on the Smart Array controller > works, but as soon as there is some heavier disk activity on the > controller (like the daily security runs etc.), the accessing processes > hang up and keep stuck in an "uninterruptable disk wait" state. I can't > even reboot the machine remotely, I have to go to the system room and > push the button. > > camcontrol now gives this output: > > [root@raidtest /home/le]# camcontrol tags da0 -v > (pass0:ciss0:0:0:0): dev_openings 1024 > (pass0:ciss0:0:0:0): dev_active 0 > (pass0:ciss0:0:0:0): devq_openings 1024 > (pass0:ciss0:0:0:0): devq_queued 0 > (pass0:ciss0:0:0:0): held 0 > (pass0:ciss0:0:0:0): mintags 2 > (pass0:ciss0:0:0:0): maxtags 255 > > (This is taken from the built-in SmartArray 5i controller, but the figures > were the same on the 5300 controller AFAIR. I can't verify it right now, > because "camcontrol tags da1" gets stuck too and I'm about 200km away > from the machine :-( ). > > Any hints? I can test more patches if you have them, but not before > monday, when I'm back to work. > > regards, > le > > -- > Lukas Ertl eMail: l.ertl@univie.ac.at > UNIX-Systemadministrator Tel.: (+43 1) 4277-14073 > Zentraler Informatikdienst (ZID) Fax.: (+43 1) 4277-9140 > der Universität Wien http://mailbox.univie.ac.at/~le/ -- Paul Saab Technical Yahoo ps@mu.org - ps@yahoo-inc.com - ps@freebsd.org Do You .. uhh .. Yahoo!? --HlL+5n6rz5pIUxbD Content-Type: text/plain; charset=us-ascii Content-Disposition: attachment; filename=diff Index: yahoo/ybsd_4/src/sys/dev/ciss/ciss.c =========================================================================== --- ciss.c 2002/04/13 15:36:17 #3 +++ yahoo/ybsd_4/src/sys/dev/ciss/ciss.c 2002/04/13 15:36:17 @@ -2000,7 +2000,7 @@ */ if ((sc->ciss_cam_sim = cam_sim_alloc(ciss_cam_action, ciss_cam_poll, "ciss", sc, device_get_unit(sc->ciss_dev), - sc->ciss_cfg->max_outstanding_commands, + sc->ciss_max_requests - 2, 1, sc->ciss_cam_devq)) == NULL) { ciss_printf(sc, "can't allocate CAM SIM\n"); @@ -2421,6 +2421,9 @@ /* handle post-command fixup */ ciss_cam_complete_fixup(sc, csio); + /* tell CAM we're ready for more commands */ + csio->ccb_h.status |= CAM_RELEASE_SIMQ; + xpt_done((union ccb *)csio); ciss_release_request(cr); } --HlL+5n6rz5pIUxbD-- 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?20020413224048.GA6046>