Ben Collins wrote:
So you're saying that generally, everything is working ok now, atleast where RBC is concerned?
Yes and no. The command and response conversions as they are now done by scsi seem to work. After the FX-3A I tried also the IceCube 800 which works too. Could not try the IceCube 400 yet.
So the sd driver doesn't implement something for the driver model to autoload it when it picks up a known type, without /etc/hotplug/scsi.agent entries?
Seems to be the case. And the last official hotplug scripts release maps type 14 to "enclosure" and to no module.
On Sun, Jul 10, 2005 at 10:13:37PM +0200, Stefan Richter wrote:
...
- Suddenly, the other problem reappeared. When I disconnect the disk, scsi_eh_0 stays there even though I am running the scsi_device_put modification in sbp2 (which helped before). This is accompanied by ieee1394's knodemgrd falling into D state.
Knodemgrd hanging in D state has _sometimes_ happened with older kernels too. At the moment it looks suspiciously like it always happens (with the new sbp2 code) if I disconnect a TYPE_RBC device, but it might be mere coincidence. Anyway, there is obviously still some deadlock in device management somewhere. I will try to find out whether that can happen with 1394 hardware other than SBP-2 too but it will take some time.
- As written in the other post, one of my 2.5" disks lets the system reboot shortly after the disk was attached. I have to find a way to identify the point of failure.
I let syslogd log synchronously and added printk(KERN_NOTICE "--- sleep 1s ---\n"); msleep_interruptible(1000); right at the end of print_inquiry() in scsi_scan.c. Yields: Jul 10 22:32:26 shuttle kernel: scsi0 : SCSI emulation for IEEE-1394 SBP-2 Devices Jul 10 22:32:26 shuttle kernel: ieee1394: sbp2: Node 1-00:1023: Using 36byte inquiry workaround Jul 10 22:32:27 shuttle kernel: ieee1394: sbp2: Logged into SBP-2 device Jul 10 22:32:27 shuttle kernel: Vendor: Initio Model: IC25N080ATMR04-0 Rev: 2.18 Jul 10 22:32:27 shuttle kernel: Type: Direct-Access ANSI SCSI revision: 02 Jul 10 22:33:28 shuttle syslogd 1.4.1: restart. Maybe the MODE_SENSE response needs to be mangled for this device as if it was an RBC device. (But remember, my older 2.5" disk says it's Direct-Access and it worked without MODE_SENSE response conversion.) Can unexpected MODE_SENSE responses lead to a reboot? Furthermore, I never noticed before that this disk also triggers the 36byte inquiry workaround. I will try it without the workaround. -- Stefan Richter -=====-=-=-= -=== -=-=- http://arcgraph.de/sr/ - : send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
