2009-07-05 18:55:08

by Pavel Machek

Subject: 2.6.31-rc0: sdb1 only starts to exist when sdb is accessed?

...happens with android phone doing usb storage emulation. I'm using
static /dev/ ... I hope!

Jul 5 20:06:02 amd udevd-event[17666]: mknod(/dev/bus/usb/001/073, 020664, (189,72) failed: Operation not permitted
Jul 5 20:06:07 amd kernel: scsi 24:0:0:0: Direct-Access HTC Android Phone 0100 PQ: 0 ANSI: 2
Jul 5 20:06:07 amd kernel: sd 24:0:0:0: Attached scsi generic sg1 type 0
Jul 5 20:06:07 amd kernel: sd 24:0:0:0: [sdb] Attached SCSI removable disk
Jul 5 20:13:08 amd pand[17815]: Bluetooth PAN daemon version 3.36
...
Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] 3862528 512-byte hardware sectors: (1.97 GB/1.84 GiB)
Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] Assuming drive cache: write through
Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] Assuming drive cache: write through
Jul 5 20:51:02 amd kernel: sdb: sdb1

You see? Partitions are only detected long after device is
detected. IIRC it did not behave like that before...
Pavel
--
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html

2009-07-05 19:30:23

by Andries E. Brouwer

Subject: Re: 2.6.31-rc0: sdb1 only starts to exist when sdb is accessed?

On Sun, Jul 05, 2009 at 08:54:47PM +0200, Pavel Machek wrote:
>
> ...happens with android phone doing usb storage emulation. I'm using
> static /dev/ ... I hope!
>
> Jul 5 20:06:02 amd udevd-event[17666]: mknod(/dev/bus/usb/001/073, 020664, (189,72) failed: Operation not permitted
> Jul 5 20:06:07 amd kernel: scsi 24:0:0:0: Direct-Access HTC Android Phone 0100 PQ: 0 ANSI: 2
> Jul 5 20:06:07 amd kernel: sd 24:0:0:0: Attached scsi generic sg1 type 0
> Jul 5 20:06:07 amd kernel: sd 24:0:0:0: [sdb] Attached SCSI removable disk
> Jul 5 20:13:08 amd pand[17815]: Bluetooth PAN daemon version 3.36
> ...
> Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] 3862528 512-byte hardware sectors: (1.97 GB/1.84 GiB)
> Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] Assuming drive cache: write through
> Jul 5 20:51:02 amd kernel: sd 24:0:0:0: [sdb] Assuming drive cache: write through
> Jul 5 20:51:02 amd kernel: sdb: sdb1
>
> You see? Partitions are only detected long after device is
> detected. IIRC it did not behave like that before...

Hi Pavel,

This doesn't seem to have anything to do with partition detection:
the partition line has the same time stamp as the scsi identify line
("3862528 512-byte hardware sectors").

Andries