I just recently upgraded from mdadm-2.1 to mdadm-2.2. $server: mdadm -As /dev/md/0 segmentation fault I found this when I rebooted me server to update my kernel. I was able to re-start my raid array via: $server: mdadm -A /dev/md/0 /dev/hde1 /dev/hdg1 ... 2 drives started ... The problem is (I believe) with the scan option. The init.d/mdadm uses the scan option and fails to activate the array. Here is a discussion on this topic: http://www.mail-archive.com/linux-raid@vger.kernel.org/msg03242.html Reverting to mdadm 2.1 fixed the problem.
Created attachment 77308 [details, diff] mdadm-2.2-update-segv.patch try this patch and see if it fixes things for you
(In reply to comment #1) > Created an attachment (id=77308) [edit] > mdadm-2.2-update-segv.patch > > try this patch and see if it fixes things for you > Sorry, I am relatively new to Gentoo and the portage system. How would I apply this and make portage not replace it?
http://www.gentoo.org/doc/en/bugzilla-howto.xml#doc_chap8 should cover it
(In reply to comment #3) > http://www.gentoo.org/doc/en/bugzilla-howto.xml#doc_chap8 should cover it > Thanks for your help. That patch and the howto doc helped out great. mdadm-2.2 built with the patch and I was able to execute an assemble/scan with it. The array started and seems to be functioning fine.
added to cvs then as 2.2-r1, thanks for testing
