From owner-freebsd-stable Fri Aug 23 15:54:12 2002 Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 4750A37B400 for ; Fri, 23 Aug 2002 15:54:09 -0700 (PDT) Received: from sinfonix.rz.tu-clausthal.de (sinfonix.rz.tu-clausthal.de [139.174.2.33]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0D8C043E7B for ; Fri, 23 Aug 2002 15:54:08 -0700 (PDT) (envelope-from Oliver.Biermann@tu-clausthal.de) Received: from interruptus.heim6.tu-clausthal.de (obie@interruptus.heim6.tu-clausthal.de [139.174.246.166]) by sinfonix.rz.tu-clausthal.de (8.9.3/8.9.3) with ESMTP id AAA27021 for ; Sat, 24 Aug 2002 00:54:06 +0200 (MET DST) Subject: Re: ATA driver is doing a fallback to pio mode4 on Promise Ultra66 From: Oliver Biermann To: freebsd-stable@freebsd.org Reply-To: obie@bmarkt.de In-Reply-To: <3D65350A.42B4331B@pantherdragon.org> References: <1030020619.1710.35.camel@interruptus.heim6.tu-clausthal.de> <3D65350A.42B4331B@pantherdragon.org> Content-Type: text/plain Content-Transfer-Encoding: 7bit X-Mailer: Ximian Evolution 1.0.8 Date: 24 Aug 2002 00:54:06 +0200 Message-Id: <1030143246.3029.30.camel@interruptus.heim6.tu-clausthal.de> Mime-Version: 1.0 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Thu, 2002-08-22 at 21:01, Darren Pilgrim wrote: > Oliver Biermann wrote: > > > > Hi, > > I've Problems with the ATA driver in 4-STABLE (cvsupped the source > > today, but the problem exists for me since back to 4.6-RELEASE as far > > as I remember). The driver is doing a fallback from UDMA66 mode to PIO4 > > mode on my Promise Ultra66 controller. (On the onboard controller of my > > mainboard [UDMA100] and on my HighPoint UDMA66 controller this does not > > happen). > > > > Here are the messages I receive: > > > > Aug 22 14:05:56 interruptus /kernel: ad8: READ command timeout tag=0 > > serv=0 - resetting > > Aug 22 14:05:56 interruptus /kernel: ata4: resetting devices .. done > > > Aug 22 14:06:26 interruptus /kernel: ad8: trying fallback to PIO mode > > Aug 22 14:06:26 interruptus /kernel: ata4: resetting devices .. done > > Aug 22 14:06:36 interruptus /kernel: ad8: READ command timeout tag=0 > > serv=0 - resetting > > Aug 22 14:06:36 interruptus /kernel: ata4: resetting devices .. done > > > > The machine was booted 25 minutes ago, and I was listening to some music > > from this hdd via samba 2.2.5 on my windows box... > > I had this behaviour quite often in the last days: > > booting, ata-driver in UDMA66 mode and then after some time using the > > hdd a fallback to PIO4... > > I would say it's not a cable-problem, because I checked my cables and > > run the same configuration on the other controllers where all seems ok. > > I've had similar problems myself from 4.2-R on, but only because I > didn't have any ATA66+ hardware until after I was using at least 4.2-R > everywhere. As long as the drive itself isn't a problem, switching > controller/cable combinations has always removed the problem. It seems > that heterogenous ATA66+ configurations with lots of drives is a > crap-shoot at best. Have you tried swaping the cable out with a > premium non-rounded ATA/100 cable (the kind you buy seperately for $20+ > each)? Ok, now I think I the ATA driver on FreeBSD4-STABLE has really a problem: Aug 23 03:14:25 interruptus /kernel: ata4: resetting devices .. ad9: DMA limited to UDMA33, non-ATA66 cable or device Aug 23 03:14:25 interruptus /kernel: done Aug 23 03:14:35 interruptus /kernel: ad9: READ command timeout tag=0 serv=0 - resetting Aug 23 03:14:35 interruptus /kernel: ata4: resetting devices .. ad9: DMA limited to UDMA33, non-ATA66 cable or device Aug 23 03:14:35 interruptus /kernel: done Aug 23 03:14:45 interruptus /kernel: ad9: READ command timeout tag=0 serv=0 - resetting Aug 23 03:14:45 interruptus /kernel: ad9: trying fallback to PIO mode Aug 23 03:14:45 interruptus /kernel: ata4: resetting devices .. Aug 23 03:14:45 interruptus /kernel: done I changed my configuration so, that now an UDMA33 hdd is running on the Promise Ultra66 (because the hdd is only capable of UDMA33 the message "non ATA66 cable or device" is coming), but I still get the fallback to PIO4 mode. So I now really think it's a problem with the FreeBSD-ATA-driver and the Promise Ultra66 (can't say if it's a problem with other Promise controllers). Btw: I changed the PCI-Slot of the card, too, and it all run perfect with 4.5 some time ago... Ok, thats it for now, Thanks guys obie ps: sorry I forget to cc to the mailinglist yesterday... -- --------------------------------------- - Oliver Biermann - - Leibnizstrasse 20 Zimmer 139 - - 38678 Clausthal-Zellerfeld - - mobile: 0171-1174913 - - mail: obie@bmarkt.de - --------------------------------------- To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message