Bug#526525: Kernel linux-image-2.6.26-2-amd64 does not see root RAID device /dev/md0

[Tony Godshall <togo@of.net>]

Ah, I guess you refer to bug 433905. Trying your suggestions.

rootdelay=10 does not help.  Neither does rootdelay=30

Other interesting things:
bnx2 0000:01:00.1: Cannot find PCI device base address, aborting.
mptbase: ioc0: ERROR - : ERROR - Unable to map adapter memory!
mptbase: ioc1: ERROR - : ERROR - Unable to map adapter memory!
mdadm: No devices listed in conf file were found.

Perhaps this bug is an mdadm bug rather than a kernel bug?  But why
then would it work then on 686 and 686-bigmem kernels?

Re the initramfs debug page: it talks mostly about scsi drive
detections, which doesnt seem to be the issue here.  But I did
determine that there's no /dev/md0.  Oddly, there *is* a /dev/md/
directory.  cat /proc/modules shows raid1 and md_mod loaded.

Best Regards.
Tony


On Sat, May 2, 2009 at 12:06 PM, maximilian attems <max@stro.at> wrote:
> On Fri, May 01, 2009 at 11:22:47AM -0700, Tony Godshall wrote:
>> Package: linux-image-2.6.26-2-amd64
>> Version: 2.6.26-15
>> Severity: important
>>
>>
>>
>> Kernel from linux-image-2.6.26-2-amd64 does not see
>> root RAID device /dev/md0 so boot sequence drops
>> busybox shell.  2.6.26-2-686-bigmem and 2.6.26-2-686
>> work properly.
>>
>> This is after an install with no issues from the std
>> 32-bit install iso image.
>>
>> Machine is a Dell T610 server with 8 GB RAM and
>> 8 cores (two four-core Xeons, I think)
>>
>
> check out wiki.debian.org/InitramfsDebug to provide info.
> most probably rootdelay will work for you or your rootdev
> got renamend, which is in either case no kernelbug.
>
>
>