Re: port-evbarm/48117: disklabel(8) and mount_msdos(8) on evbarm don't work with USB pen drives that work on other architectures

[David H.Gutteridge <dhgutteridge%sympatico.ca@localhost>]

The following reply was made to PR port-evbarm/48117; it has been noted by 
GNATS.

From: David H.Gutteridge <dhgutteridge%sympatico.ca@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: port-evbarm/48117: disklabel(8) and mount_msdos(8) on evbarm don't 
work with USB pen drives that work on other architectures
Date: Sun, 13 Jul 2014 23:25:57 -0400

 On 2014-07-13, at 11:07 PM, David H. Gutteridge wrote:
 > On 2014-07-13, at 1:10 AM, David Holland wrote:
 >> The following reply was made to PR port-evbarm/48117; it has been =
 noted by GNATS.
 >>=20
 >> From: David Holland <dholland-bugs%netbsd.org@localhost>
 >> To: gnats-bugs%NetBSD.org@localhost
 >> Cc:=20
 >> Subject: Re: port-evbarm/48117: disklabel(8) and mount_msdos(8) on =
 evbarm
 >> don't work with USB pen drives that work on other architectures
 >> Date: Sun, 13 Jul 2014 05:08:10 +0000
 >>=20
 >> On Tue, Aug 13, 2013 at 02:10:00AM +0000, 
 >> dhgutteridge%sympatico.ca@localhost =
 wrote:
 >>> If I try to mount a USB pen drive with an FAT12 or FAT16 file system
 >>> on a Raspberry Pi running 6.99.23/evbarm, the actual partition with
 >>> data on it isn't being recognized. The same examples work on other
 >>> architectures I've tried (amd64, i386, macppc).
 >>=20
 >> Is this related to 47463? (Or, the fix for 47463?)
 >=20
 > It presumably is: from looking at the other PR, it seems related. The
 > change that was committed doesn't address my problem. I just tested
 > with a kernel and userland from July 8th, and the problem still
 > persists. There's a difference now in terms of what disklabel(8)
 > reports, but it's not an improvement. For some reason, it now makes
 > reference to a 4.2BSD fstype entry, which is not present on the
 > drive.
 >=20
 > # /dev/rsd0c:
 > type: SCSI
 > disk: Store 'n' Go   =20
 > label: fictitious
 > flags: removable
 > bytes/sector: 512
 > sectors/track: 32
 > tracks/cylinder: 16
 > sectors/cylinder: 512
 > cylinders: 980
 > total sectors: 501760
 > rpm: 3600
 > interleave: 1
 > trackskew: 0
 > cylinderskew: 0
 > headswitch: 0           # microseconds
 > track-to-track seek: 0  # microseconds
 > drivedata: 0=20
 >=20
 > 3 partitions:
 > #        size    offset     fstype [fsize bsize cpg/sgs]
 > a:    501760         0     4.2BSD      0     0     0  # (Cyl.      0 - =
    979)
 > c:    501760         0     unused      0     0        # (Cyl.      0 - =
    979)
 > disklabel: boot block size 0
 > disklabel: super block size 0
 >=20
 > If you want to close this PR and track against the other, that's fine
 > with me.
 >=20
 > Dave
 
 I should add, for comparison, fdisk(8) does show correct information:
 
 Disk: /dev/rsd0c
 NetBSD disklabel disk geometry:
 cylinders: 980, heads: 16, sectors/track: 32 (512 sectors/cylinder)
 total sectors: 501760, bytes/sector: 512
 
 BIOS disk geometry:
 cylinders: 980, heads: 16, sectors/track: 32 (512 sectors/cylinder)
 total sectors: 501760
 
 Partitions aligned to 2048 sector boundaries, offset 32
 
 Partition table:
 0: DOS (16-bit FAT) - LBA (sysid 14)
     start 32, size 501728 (245 MB, Cyls 0-979), Active
 1: <UNUSED>
 2: <UNUSED>
 3: <UNUSED>
 First active partition: 0
 Drive serial number: 2759040381 (0xa4739d7d)
 
 Dave