On Thu, 10 Jun 2010, Hans de Goede wrote:


When an mdraid array with newer metadata versions is created with a name
of md#, like we do, the name returned from devicelibs.mdraid.examine
is in the form of md/#. However if we use that as device name in the
devicetree, then later on when we scan the mdraid array after it was
brought up by handleUdevMDMemberFormat, we will find a md# device as
that is the sysfs name, think it is a different device and add the
device to the tree a second time, this patch fixes this.
---
pyanaconda/storage/devicetree.py | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/pyanaconda/storage/devicetree.py b/pyanaconda/storage/devicetree.py
index 67c8f65..ef53db4 100644
--- a/pyanaconda/storage/devicetree.py
+++ b/pyanaconda/storage/devicetree.py
@@ -1621,6 +1621,7 @@ class DeviceTree(object):
# md_name can be either md# or md/#
if md_name.startswith("md/"):
minor = int(md_name[3:]) # strip off leading "md/"
+ md_name = "md%d" % minor # use a regular md# name
else:
minor = int(md_name[2:]) # strip off leading "md"
except (IndexError, ValueError):



Ack for master and rhel6-branch (with a Related rhbz).

--
David Cantrell <dcantrell@redhat.com>
Red Hat / Honolulu, HI

_______________________________________________
Anaconda-devel-list mailing list
Anaconda-devel-list@redhat.com
https://www.redhat.com/mailman/listinfo/anaconda-devel-list