common.c:get_parent_sysfs_path() does not get the parent device
Kay Sievers
kay.sievers at vrfy.org
Fri Nov 5 20:03:30 PST 2004
Hey David,
hald is unable to recognize all scsi bus devices of my usb stick.
A whole chain of parent object lookup fails for:
devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0/1:0:0:0
cause:
devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0
is not a device we get an event for and therefore can't be found as a
parent in the store.
sysfs:
[kay at pim ~]$ tree /sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1
/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1
|-- detach_state
|-- power
| `-- state
`-- target1:0:0
|-- 1:0:0:0
| |-- block -> ../../../../../../../../../block/sda
| |-- delete
| |-- detach_state
| |-- device_blocked
| |-- driver -> ../../../../../../../../../bus/scsi/drivers/sd
| |-- generic -> ../../../../../../../../../class/scsi_generic/sg0
| |-- max_sectors
| |-- model
| |-- power
| | `-- state
| |-- queue_depth
| |-- rescan
| |-- rev
| |-- scsi_level
| |-- state
| |-- timeout
| |-- type
| `-- vendor
|-- detach_state
`-- power
`-- state
Here is the hald debug output with the attached patch applied,
which makes it work again by a dirty hack.
linux/osspec.c:1727: Queing up seqnum=997, sysfspath=/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0/1:0:0:0, subsys=scsi
linux/osspec.c:1337: action=add seqnum=997 subsystem=scsi sysfs_path=/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0/1:0:0:0
linux/bus_device.c:111: parent_sysfs_path='/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0/1:0:0:0' parent_sysfs_path='/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0'
linux/bus_device.c:132: XXX parent=(nil) parent_sysfs_path='/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1/target1:0:0'
linux/bus_device.c:144: XXX grandparent=0x80bd718 parent_sysfs_path='/sys/devices/pci0000:00/0000:00:1d.1/usb3/3-1/3-1:1.0/host1'
I'm not sure if we really can assume that every directory in sysfs
is a device we got an earlier event for?
Any ideas?
Thanks,
Kay
-------------- next part --------------
A non-text attachment was scrubbed...
Name: hal-devices-grandparent.patch
Type: text/x-patch
Size: 1208 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/hal/attachments/20041106/95b625a6/hal-devices-grandparent.bin
-------------- next part --------------
_______________________________________________
hal mailing list
hal at freedesktop.org
http://freedesktop.org/mailman/listinfo/hal
More information about the Hal
mailing list