On 08/06/2009 02:11 PM, David Lehman wrote:
>> diff --git a/storage/devicelibs/dm.py b/storage/devicelibs/dm.py
>> index 29df126..a4a4ca9 100644
>> --- a/storage/devicelibs/dm.py
>> +++ b/storage/devicelibs/dm.py
>> @@ -105,3 +113,13 @@ def get_backing_devs_from_name(map_name):
>> slave_devs = os.listdir("/sys/block/virtual/%s" % dm_node)
>> return slave_devs
>>
>> +def dm_is_multipath_disk(map_name):
>> + return False
>> +
>> +def dm_get_mutlipath_partition_disk(map_name):
>> + return None
>> +
>> +def dm_is_multipath_partition(map_name):
>> + return False
>> +
>> +
>
> Why add the above three functions if they don't do anything?
Yeah, they eventually got done differently. I'll remove them.
>
>> diff --git a/storage/devices.py b/storage/devices.py
>> index 3ca05aa..a866f14 100644
>> --- a/storage/devices.py
>> +++ b/storage/devices.py
>> @@ -96,6 +96,7 @@
>> import os
>> import math
>> import copy
>> +import string
>
> If we can avoid this by using the string instances' methods we should do
> so.
Indeed.
>> @@ -762,14 +763,15 @@ class DiskDevice(StorageDevice):
>> # When the device has no partition table but it has a FS, it
>> # will be created with label type loop. Treat the same as if
>> # the device had no label (cause it really doesn't).
>> - if self._partedDisk.type == "loop":
>> + if self.partedDisk.type == "loop":
>> if self._initcb is not None and self._initcb():
>> - self._partedDisk = parted.freshDisk( \
>> - device=self.partedDevice, \
>> + self._partedDisk = parted.freshDisk(device=self.partedDevice, \
>> ty = platform.getPlatform(None).diskType)
>> else:
>> raise DeviceUserDeniedFormatError("User prefered to not format.")
>>
>> + return self._partedDisk
>> +
>
> It looks like some of the previous patch is being reverted in the above
> hunk (self._partedDisk -v- self.partedDisk).
Ah, that and another problem -- the "return" here should have been in the
previous commit. I'll push the previous commit with that merged into it.
>>
>> + @property
>> + def up(self):
>> + return self._isUp
>
> Is this different from status?
Hrm. No, probably not.
>
>> +
>> + @property
>> + def path(self):
>> + """ Device node representing this device. """
>> + return "/dev/mapper/%s" % (self.name,)
>
> This should not be needed. StorageDevice already does this, except
> without hard-coding /dev/mapper.
Gone.
>> +
>> + @property
>> + def wwid(self):
>> + serial = self.identity
>> + ret = []
>> + while serial:
>> + ret.append(serial[:2])
>> + serial = serial[2:]
>> + return string.join(ret, ':')
>
> You could avoid the string import above by instead doing this:
>
> return ":".join(ret)
>
> Weird looking, I know.
Yeah.
>> + def handleMultipathMemberFormat(self, info, device):
>> + log_method_call(self, name=device.name, type=device.format.type)
>> +
>> + serial = udev_device_get_serial(info)
>> + found = False
>> + if self.__multipaths.has_key(serial):
>> + mp = self.__multipaths[serial]
>> + mp.addParent(device)
>> + else:
>> + name = generateMultipathDeviceName()
>> + devname = "/dev/mapper/%s" % (name,)
>> +
>> + if self.zeroMbr:
>> + cb = lambda: True
>> + else:
>> + desc = []
>> + serialtmp = serial
>> + while serialtmp:
>> + desc.append(serialtmp[:2])
>> + serialtmp = serialtmp[2:]
>> + desc = "WWID %s" % (string.join(desc, ':'),)
>
> Again, you can do ":".join(desc) here instead.
okay.
>> diff --git a/storage/udev.py b/storage/udev.py
>> index 450b54a..adb69e6 100644
>> --- a/storage/udev.py
>> +++ b/storage/udev.py
>> @@ -22,6 +22,7 @@
>>
>> import os
>> import stat
>> +import block
>>
>> import iutil
>> from errors import *
>> @@ -134,6 +135,10 @@ def udev_device_get_label(udev_info):
>> """ Get the label from the device's format as reported by udev. """
>> return udev_info.get("ID_FS_LABEL")
>>
>> +def udev_device_is_multipath_member(info):
>> + """ Return True if the device is part of a multipath. """
>> + return info.get("ID_FS_TYPE") == "multipath_member"
>
> It kind of sucks that we have this string both in the format class and
> here.
I can see that, yeah, but I don't really have a better way.
Another patch will hit your mailbox soon.
--
Peter
The Shuttle is now going five times the sound of speed.
-- Dan Rather, first landing of Columbia
_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@xxxxxxxxxx
https://www.redhat.com/mailman/listinfo/anaconda-devel-list
