Date: Wed, 10 Jul 2002 18:58:57 +0200 From: Thomas Seck <tmseck-lists@netcologne.de> To: stable@FreeBSD.ORG Subject: Re: strange ATA behavior with -STABLE Message-ID: <20020710165857.GB996@laurel.tmseck.homedns.org> In-Reply-To: <20020710015649.0592a44a.sheep.killer@cultdeadsheep.org> References: <20020710015649.0592a44a.sheep.killer@cultdeadsheep.org>
next in thread | previous in thread | raw e-mail | index | archive | help
* Clement Laforet (sheep.killer@cultdeadsheep.org): > Hi guys :) > > I "recently" upgraded one of my server from 4.5 (RELENG_4_5) to stable (RELENG_4) on the 4th july (nearly 9am CEST). > And I noticed few hours ago, a little difference with the ata drivers. > I know they have changed but there is something which "disturbs" me. > Here a piece of log I have when "ATA failled". I don't know why. Disks seems to be ok, maybe the heat ;) > so... > > Jul 4 09:00:40 goofy /kernel: ad0: READ command timeout tag=0 serv=0 - resetting > Jul 4 09:00:40 goofy /kernel: ata0: resetting devices .. done > Jul 4 09:00:50 goofy /kernel: ad0: READ command timeout tag=0 serv=0 - resetting > Jul 4 09:01:10 goofy last message repeated 2 times > Jul 4 09:01:10 goofy /kernel: ata0-master: timeout waiting for command=ef s=00 e=00 > Jul 4 09:01:10 goofy /kernel: ad0: trying fallback to PIO mode > Jul 4 09:01:20 goofy /kernel: ad0: READ command timeout tag=0 serv=0 - resetting (Not quite) the same here (Intel PIIX3, HX chipset with a 4 GB Fujitsu on ad0 and a 19 GB Maxtor on ad2). The driver does not fall back from UDMA 2 to PIO 4. In my opinion, this is a problem with the new ATA driver. I plan to do some testing with an older FreeBSD version to be sure that this is not a controller issue. > After this, the system was still up, even if I was unable to access > ad0 and ad2, processes "simply became dead". (system an 'sensible > data' are on SCSI disks) Without happiness, I deal with this ;) This > was before I upgrade ;) > > I thought (bad ?) that upgrading to newer version of ata driver, > allows me to reset ata buses, and repair this "on the fly". I was > wrong ! Now my FreeBSD system freezes. The last console info is > (like): ad2: read command timeout tag=0 serv=0 - resetting ata-1 > resetting *FREEZE* and then hard reboot Nothing in syslog :) Fortunately, the system always recovers here. I can trigger the timeouts easily when I do a dump of a filesystem on ad0 to ad2. They rarely occur during normal operation though. -- Thomas Seck This message was sent to a mailinglist I am subscribed to. Please send your replies to the list only and do *not* CC me. Thank you. To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020710165857.GB996>