Alan Stern wrote:
On Fri, 7 Sep 2007, Peter Rasmussen wrote:
I'm sorry if this is a little late, but I had once had access to a Z6
and believe I had it successfully connected to my Linux host, and was
therefore puzzled by this message exchange.
I have now borrowed the device again for the weekend to check it out a
little more.
Using Alan's patch below, otherwise my kernel is a standard 2.6.23-rc5
(CONFIG_USB_STORAGE_DEBUG=y ) I got the following result after reboot
and connecting the Z6.:
...
In fact there was another patch from earlier in the email thread, which
was needed to work around the PQ = 1 problem. Since you didn't apply
that patch, the SCSI disk driver wasn't bound to your Z6.
Now I applied that patch, too, and it did make a difference, in drivers/scsi/scsi_sysfs.c I have:
static int scsi_bus_match(struct device *dev, struct device_driver *gendrv)
{
struct scsi_device *sdp = to_scsi_device(dev);
if (sdp->no_uld_attach)
return 0;
return 1;
// return (sdp->inq_periph_qual == SCSI_INQ_PQ_CON)? 1: 0;
}
Before posting a lot of USB debug output, I'll explain a shorter version :-)
1.After reboot I connected the Z6 (incl. a micro-SD card) and got what looked good, if not the usual output from dmesg:
<snip>
scsi 0:0:0:0: Direct-Access Motorola MSnc. PQ: 1 ANSI: 0 CCS
<snip>
sd 0:0:0:0: [sda] Attached SCSI removable disk
sd 0:0:0:0: Attached scsi generic sg0 type 0
<snip>
scsi 0:0:0:1: Direct-Access Motorola MSnc. PQ: 1 ANSI: 0 CCS
<snip>
sd 0:0:0:1: [sdb] Attached SCSI removable disk
sd 0:0:0:1: Attached scsi generic sg1 type 0
<snip>
2.I could then mount the SCSI devices like this:
# mount /dev/sda /mnt/usb
# mount /dev/sdb /mnt/usb2
When I normally with a USB mass storage device do, eg.:
# mount /dev/sda1 /mnt/usb
3.I couldn't however, always mount the mentioned devices as I would sometimes get a 'mount: No medium found' error, and a message on the mobile will sometimes show "USB connected" and sometimes "USB disconnected" even though it didn't seem to correlate to successful connection between my PC and the Z6.
4.So, something seems to be inconsistent. I have tried it with a few different cables and I can't detect a difference.
5.Should I post a full log of eg. connecting the Z6, mounting it and the SD card, accessing it and then unmounting?
A full log may then include failed attempts, which will perhaps give some hints at what is wrong?
Or do you already have a pretty good idea as to what is wrong, and find that such an exercise is not really useful?
What I can't understand, however, is that as this mobile phone
supposedly runs Linux. I can't see how Motorola could break it so badly
that the USB connection now doesn't work with a PC running Linux?
:-) There's no necessary relation between the OS running on a device
like your phone and its interoperability with computers running the
same OS.
Well, I would have expected that they have similar methods to connect, even though different Linux kernel version of course may have introduced some differences. My A780 has a 2.4.20 kernel and I wouldn't be surprised if other Motorola mobiles also use the 2.4.* series of Linux kernels.
And you say that this seems to be a more widespread problem with
Motorola devices? Do you remember which ones, and do they run Linux as well?
There were two problems. First was the PQ = 1 problem; I have never
seen it before now (so only on the Z6). The other problem was the
capacity, or last sector number; we know that the RAZR V3i and V3x both
suffer from it as well. (I have no idea whether they run Linux.)
Possibly other devices do too, and we just don't know about them.
I know that those two don't run Linux, but can they connect to a Linux host, or do they also have problems?
The following is from connecting a Z3 that supposedly isn't based on
Linux, but works all right (and this may have been the one I thought was
the Z6 that I claimed worked):
...
It looks normal.
I hope perhaps it can provide some comparison info now with a device
that works and one that doesn't.
If Motorola actualy has made changes to the Linux kernel they use with
the Z6, I suppose we should be able to get the actual code. Should I
investigate that?
Sure, go ahead and try. Maybe you can convince them to fix their bugs!
Although that won't help all the units that have already been
manufactured...
Well, if they could provide an update, then people that have one and want to use it with Linux should be able to get some relief.
Alan Stern
Peter
-
To unsubscribe from this list: 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
