Date: Thu, 04 Aug 2005 21:13:55 +0300 From: Heikki Suonsivu <hsu@wlansystems.com> To: Heikki Suonsivu <hsu@mail.bbnetworks.net> Cc: Heikki Suonsivu <hsu@bbnetworks.net>, Hidetoshi Shimokawa <simokawa@FreeBSD.org>, freebsd-bugs@FreeBSD.org Subject: Re: kern/73313: Maxtor Onetouch drivers hang when used with firewire. Message-ID: <42F25AE3.1050305@wlansystems.com> In-Reply-To: <16811.23676.44163.220110@mail.bbnetworks.net> References: <200410301328.i9UDSQod067234@hsu.bbnetworks.net> <87653p2irf.wl@tora.nunu.org> <16811.23676.44163.220110@mail.bbnetworks.net>
next in thread | previous in thread | raw e-mail | index | archive | help
This is a multi-part message in MIME format. --------------090406050609040005030207 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit This problem is still present in 5-STABLE, with same kernel messages and same problems. I am often seeing major disk corruption and data written to random place (fsck reports paths which are obviously part of some text file, etc). I am also still seeing kernel puffing out messages with random characters in them, and other kernel loose pointer indications. Symptoms are usually kernel either getting more and more confused, a panic or a lockup. Usually just the firewire disks lock up, nothing else, though if I do not reboot within an hour of use or so, I will eventually get a full lockup or panic. The problems increase (or possibly appear only) when at least two disks are connected to one firewire port, but this could also be just an indication of load increase. Using two of these disks at the same time does not seem to correlate with the corruption. The current configuration include 4 Maxtor 250-300G disks, being used as storage for video recordings. I am very often getting a panic and major disk corruption if I copy a large video file from ATA disk to one of the Firewire disks. Corruption seems to happen to places which are being modified, like directories. If I stay off firewire disks, I do not see any problems (the computer has 3 ATA disks), so this looks like it is firewire related. I also tried using EHCI/USB2, but that was even more unstable. I have also seen this on another computer with 2 - 3 firewire disks, but rately. The usage pattern in this one is much less intensive and does not involve as large files. All disks are various Maxtor models. Aug 1 09:32:44 lelu kernel: fwohci0: <1394 Open Host Controller Interface> mem 0x4c300000-0x4c3007ff irq 11 at device 15.4 on pci0 Aug 1 09:32:44 lelu kernel: fwohci0: OHCI version 1.10 (ROM=1) Aug 1 09:32:44 lelu kernel: fwohci0: No. of Isochronous channels is 4. Aug 1 09:32:44 lelu kernel: fwohci0: EUI64 00:90:e6:00:00:00:02:e5 Aug 1 09:32:44 lelu kernel: fwohci0: Phy 1394a available S400, 2 ports. Aug 1 09:32:44 lelu kernel: fwohci0: Link S400, max_rec 2048 bytes. Aug 1 09:32:44 lelu kernel: firewire0: <IEEE1394(FireWire) bus> on fwohci0 Aug 1 09:32:44 lelu kernel: sbp0: <SBP-2/SCSI over FireWire> on firewire0 Aug 1 09:32:44 lelu kernel: fwe0: <Ethernet over FireWire> on firewire0 Aug 1 09:32:44 lelu kernel: if_fwe0: Fake Ethernet address: 02:90:e6:00:02:e5 Aug 1 09:32:44 lelu kernel: fwe0: Ethernet address: 02:90:e6:00:02:e5 Aug 1 09:32:44 lelu kernel: fwe0: if_start running deferred for Giant Aug 1 09:32:44 lelu kernel: fwohci0: Initiate bus reset Aug 1 09:32:44 lelu kernel: fwohci0: node_id=0xc800ffc4, gen=1, CYCLEMASTER mod e Aug 1 09:32:44 lelu kernel: firewire0: 5 nodes, maxhop <= 4, cable IRM = 4 (me) Aug 1 09:32:44 lelu kernel: firewire0: bus manager 4 (me) Aug 1 09:32:44 lelu kernel: firewire0: New S400 device ID:0010b92000d5db66 Aug 1 09:32:44 lelu kernel: firewire0: New S400 device ID:0010b9f7011466a0 Aug 1 09:32:44 lelu kernel: firewire0: New S400 device ID:0010b92100409c86 Aug 1 09:32:44 lelu kernel: firewire0: New S400 device ID:0010b9f701222bcc Aug 1 09:32:44 lelu kernel: da0 at sbp0 bus 0 target 0 lun 0 Aug 1 09:32:44 lelu kernel: da0: <Maxtor OneTouch 0000> Fixed Direct Access SCS I-4 device Aug 1 09:32:44 lelu kernel: da0: 50.000MB/s transfers Aug 1 09:32:44 lelu kernel: da0: 286103MB (585938944 512 byte sectors: 255H 63S /T 36473C) Aug 1 09:32:44 lelu kernel: da1 at sbp0 bus 0 target 1 lun 0 Aug 1 09:32:44 lelu kernel: da1: <Maxtor OneTouch II 0310> Fixed Direct Access SCSI-4 device Aug 1 09:32:44 lelu kernel: da1: 50.000MB/s transfers Aug 1 09:32:44 lelu kernel: da1: 286188MB (586114704 512 byte sectors: 255H 63S /T 36483C) Aug 1 09:32:44 lelu kernel: da2 at sbp0 bus 0 target 2 lun 0 Aug 1 09:32:44 lelu kernel: da2: <Maxtor OneTouch 0000> Fixed Direct Access SCS I-4 device Aug 1 09:32:44 lelu kernel: da2: 50.000MB/s transfers Aug 1 09:32:44 lelu kernel: da2: 286103MB (585938944 512 byte sectors: 255H 63S /T 36473C) Aug 1 09:32:44 lelu kernel: da3 at sbp0 bus 0 target 3 lun 0 Aug 1 09:32:44 lelu kernel: da3: <Maxtor OneTouch 0000> Fixed Direct Access SCS I-4 device Aug 1 09:32:44 lelu kernel: da3: 50.000MB/s transfers Aug 1 09:32:44 lelu kernel: da3: 286103MB (585938944 512 byte sectors: 255H 63S /T 36473C) > Hidetoshi Shimokawa writes: > > At Sat, 30 Oct 2004 16:28:26 +0300 (EEST), > > Heikki Suonsivu wrote: > > > Oct 30 15:30:45 hsu /kernel: Unknown service addr 0x0000:0x04f14800 WREQB(1) src=0xffc2 data=8 > > > Oct 30 15:32:39 hsu /kernel: Unknown service addr 0x0000:0x03b83800 WREQB(1) src=0xffc2 data=8 > > > Oct 30 15:32:54 hsu /kernel: Unknown service addr 0x0000:0x1761b800 WREQB(1) src=0xffc2 data=8 > > > Oct 30 15:32:54 hsu /kernel: Unknown service addr 0x0000:0x16242800 WREQB(1) src=0xffc2 data=8 > > > Oct 30 15:32:59 hsu /kernel: Unknown service addr 0x0000:0x0dddb000 WREQB(1) src=0xffc2 data=8 > > > > These address ranges should be handled as physical access and > > the kernel shouldn't be notified of it. > > It seems that your fwohci chip stop handling of physical access > > for unknown reason and this leads to timeout error. > > > > Can you swap your fwohci card with other cards with a different chip? > > I do not have any other cards, so not easily. The above errors do not seem > necessarily hang or crash the computer. I currently have messages full of > those messages, but the computer is still ok and firewire works (copying > large amount of data from firewire to ata disk - two crashes during this > 300G copy). I think there is an improvement though, as I now get a crash > instead of the computer hanging up. I enabled crash dumps. > > > /\ Hidetoshi Shimokawa > > \/ simokawa@FreeBSD.org > _______________________________________________ > freebsd-bugs@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-bugs > To unsubscribe, send any mail to "freebsd-bugs-unsubscribe@freebsd.org" --------------090406050609040005030207--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?42F25AE3.1050305>