Date: Thu, 26 Sep 2002 21:10:23 +0900 From: Katsushi Kobayashi <ikob@koganei.wide.ad.jp> To: exile@chamber.ee Cc: freebsd-firewire@FreeBSD.ORG Subject: Re: 28bit address limitation on HD Message-ID: <EF7B9685-D148-11D6-8F7C-0030657D3460@koganei.wide.ad.jp> In-Reply-To: <20020926105811.envmzsns5m0ogk@.chamber.ee>
next in thread | previous in thread | raw e-mail | index | archive | help
Hello, By the basic policy of current SBP-II driver, the different device No. is assigned if the device has different EUI-64 ID. If you had once attached other SBP-II SCSI disk or IDE/Firewire bridges, the device No. is not zero. Even if the first device has been removed already. In the case of dynamic attached disk system, it is difficult to manipulate legacy UNIX way. This issue is not only SBP-II. I thought some mechanism or policy is needed which disk should bind which ids. Also, I am not sure, why happen busy device messages. On 2002.Sep.26, at 05:58 PM, Sten Poldma wrote: > > Hi, > > A follow-up om the case previously reported.. > > [excuse me for the rant] > > A new box: MAPOWER (MAP-35F), Unfortunately it does not appear on the > manufacturer homepage anymore: http://www.mapower.com.tw/ieee_1394.htm > But is similar to the MAP-J51F although not much details are offered > there. > > The box works fine Under MacOS X, when plugging into the FreeBSD and > loading the > firewire drivers, I see the Hd fine.. > > /kernel: da1: <Mapower 0032> Fixed Simplified Direct Access SCSI-4 > device > /kernel: da1: 50.000MB/s transfers > /kernel: da1: 156334MB (320173056 512 byte sectors: 255H 63S/T 19929C) > > As you can see, it is the same Hd, and the geometry looks fine.. > The HD was formatted to UFS in Mac before plugging it in, still, I did > not find > the partition there when I plugged it to the BSD machine. > > Anyhow, I partitioned, Labeled it and newfs'it through the > /stand/sysinstall > with no freaky settings. So far so good, and created the da1s1e > partition. > > When trying to mount it, I get device busy. I try to switch the ports > of the > firewire etc, Still unable to mount it with the same error. I check the > partition and label, they look to be OK. > > So what seems to be the problem now ? The sytem can see the box and > the disk, I > can fdisk,label,newfs it. But I can't mount it. > > Also, why is the device recognized as da1 ? It is the only device on > the chain. > > Anyway, I seemed to solve the problem with rebooting the entire box > (firewire > driver unload\load cycle did not seem to help). The reboot phase gave > me a Page > Fault 12 error as well, but after the system came up, the device was > recognized > as da0 and mounting the disk was fine. > > I find it peculiar behaviour, do you happen to have any idea of what > might have > caused this ? > > Anyway, I hope this little experience might come in handy to someone > in the > future. > > Regards, > > Sten > > Sten Poldma <exile@chamber.ee> said: > >> >> Hi, >> >> Thanks for the response, that's probably it, The bridge uses the >> Oxford chipset, >> which probably does not support the 48 bit mode. >> >> I'm sure a new box will remedy the problem. >> >> Thanks again, >> >> Sten >> >> Katsushi Kobayashi <ikob@koganei.wide.ad.jp> said: >> >>> >>> I have not had any experience on the such disk size of IEEE1394 disk. >>> >>> If you are using IEEE1394/IDE bridge or similar one, let me know >>> the product information. Or if you have an experience on other OS, >>> let me know whether to success to detect size in correct. >>> >>> >>> On 2002.Sep.23, at 09:19 PM, Sten Poldma wrote: >>> >>>> >>>> Hi, >>>> >>>> I'm Sorry if the following is not firewire related but depends on >>>> other system >>>> components instead. >>>> >>>> I'm using the newest firewire driver >>>> (http://people.freebsd.org/~simokawa/firewire-20020918.tar.gz) >>>> >>>> To test a 160GB IDE drive (in a box), through the CAM and >>>> passthrough >>>> drivers I >>>> have no problem locating the device (da0) but the geometry of the >>>> drive found >>>> makes me believe that there is a 28bit addressing mode bottleneck >>>> somewhere as >>>> the log reports that 268435455 512 byte sectors found. (~131GB) (I >>>> believe >>>> 268435456 being the 28bit limit?) >>>> >>>> Although using it is OK I'm still wondering if it would be possible >>>> to >>>> locate >>>> the missing 20GB before the disk actually goes into active use. >>>> >>>> I'd like to know what is causing this limitation? >>>> >>>>> From the hardware I have: >>>> >>>> FreeBSD 4.4 -> make buildworld to 4.6.2, latest >>>> kernel: 4.7-RC FreeBSD 4.7-RC #0: Sun Sep 22 01:39:47 EET 2002 >>>> Firewire PCI card: fwohci0: <Texas Instruments TSB12LV26> >>>> and a Western Digital 160GB drive in a standard firewire box. >>>> >>>> Again, accept my apologies if this is the wrong place to request >>>> help\clarification on the matter but any help or hint is >>>> appreciated. >>>> >>>> Have a nice day, >>>> >>>> Sten >>>> >>>> To Unsubscribe: send mail to majordomo@FreeBSD.org >>>> with "unsubscribe freebsd-firewire" in the body of the message >>> >> >> >> >> >> To Unsubscribe: send mail to majordomo@FreeBSD.org >> with "unsubscribe freebsd-firewire" in the body of the message >> > > To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-firewire" in the body of the message
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?EF7B9685-D148-11D6-8F7C-0030657D3460>