On Fri 08 Sep 2006 at 23:26:52 -0500, Jonathan A. Kollasch wrote:
> Does it also happen with USB?

No, but the machine in question only has USB 1 (which is so slow that I
really want to use the firewire; on my other computer I can use USB 2).

> It seems that the firewire portion of the Acomdata enclosure is Initio,
> like this one.  So, unless they changed something, or the firmware is
> different, I don't know.  Can you test the Acomdata one again?

The models that Acomdata is currently showing on their website differ
from my model (mine is plain brick-shaped), but when bought mine
(several months ago) I think they still showed it. Maybe they changed
the internals too.

> There seem to be no relevant changes in sd(4) since 3.0, perhaps some
> higher layer was changed.
> 
> Also, how is/was the reliability and performance of the FreeBSD/NetBSD
> firewire implementation?  I've been having a horrible time with
> fwip(4).  Just wondering if getting a combo enclosure is worth it
> if it's only going to be used on NetBSD.

I first used the Acomdata with 3.0 and fairly quickly the kernel logged
complaints about missed transactions or somesuch, then stalled. I then
tried a -current of around then (3.99.18, my kernel was compiled 11
may). That worked much, much better.

The Acomdata gives (3.99.18):

fwohci0: BUS reset
fwohci0: node_id=0x8800ffc0, gen=2, non CYCLEMASTER mode
ieee1394if0: 2 nodes, maxhop <= 1, cable IRM = 0 (me)
ieee1394if0: root node is not cycle master capable
ieee1394if0: bus manager 0 (me)
fwohci0: too many cycle lost, no cycle master presents?
ieee1394if0: New S400 device ID:0010100540018734
sbp0 at ieee1394if0: SBP-2/SCSI over IEEE1394
scsibus0 at sbp0: 1 target, 0 luns per target
sd0 at scsibus0 target 0 lun 0: <DMI, WD3000JB-00KFA0, 0243> disk fixed
sd0: fabricating a geometry
sd0: 279 GB, 286168 cyl, 64 head, 32 sec, 512 bytes/sect x 586072368 sectors

then, after a disklabel, the usual

sd0: fabricating a geometry

and then unplugging:

fwohci0: BUS reset
fwohci0: node_id=0xc800ffc0, gen=3, CYCLEMASTER mode
ieee1394if0: 1 nodes, maxhop <= 0, cable IRM = 0 (me)
ieee1394if0: bus manager 0 (me)

If I then plug in the "cooler master", I get this (different disk, 3.99.18):

fwohci0: BUS reset
fwohci0: node_id=0x8800ffc0, gen=4, non CYCLEMASTER mode
ieee1394if0: 2 nodes, maxhop <= 1, cable IRM = 0 (me)
ieee1394if0: root node is not cycle master capable
ieee1394if0: bus manager 0 (me)
fwohci0: too many cycle lost, no cycle master presents?
ieee1394if0: New S400 device ID:001010055000008c
sbp1 at ieee1394if0: SBP-2/SCSI over IEEE1394
scsibus1 at sbp1: 1 target, 0 luns per target
sd1 at scsibus1 target 0 lun 0: <Initio, WD2500JB-00REA0, 0312> disk fixed
sd1: fabricating a geometry
sd1: 3234 TB, 238475 cyl, 64 head, 32 sec, 7282720 bytes/sect x 488397168 sectors

I do wish I had a disk of this size (it's only 250 marketing-GB in reality).

A disklabel gives then a crash at the same place as before
(sd_get_parms_page4).

After reboot, I plug in the "cooler master" via usb:

uhub2 at uhub1 port 1
uhub2: Standard Microsystems product 0x2504, class 9/0, rev 2.00/0.01, addr 2
uhub2: 4 ports with 4 removable, self powered
umass0 at uhub2 port 1 configuration 1 interface 0
umass0: Initio WD2500JB-00REA0, rev 2.00/3.12, addr 3
umass0: using SCSI over Bulk-Only
scsibus0 at umass0: 2 targets, 1 lun per target
uhidev0 at uhub2 port 1 configuration 1 interface 1sd0 at scsibus0
target 0 lun 0: <Initio, WD2500JB-00REA0, 3.12> disk fixed

uhidev0: Initio WD2500JB-00REA0, rev 2.00/3.12, addr 3, iclass 3/0
uhid0 at uhidev0: input=2, output=2, feature=0
sd0: fabricating a geometry
sd0: 232 GB, 238475 cyl, 64 head, 32 sec, 512 bytes/sect x 488397168 sectors

disklabel results in

sd0: fabricating a geometry

Apparently it has an "uhidev0" too, it must be the "automatic backup"
button the user manual mentions. The messages get interlaced a bit.

> 	Jonathan Kollasch
-Olaf.
-- 
___ Olaf 'Rhialto' Seibert      -- You author it, and I'll reader it.
\X/ rhialto/at/xs4all.nl        -- Cetero censeo "authored" delendum esse.