Bug#434138: Re: Bug#434138: bug 434138

[Eric Lauriault]

Hi Martin,

Attached are bootlogs of the same system booting the same kernel with 
only one difference: the inclusion of the "sleep 5" patch in the initrd.
I generated the initrd's using :

mkinitramfs -o sleep.img 2.6.21-2-sparc64-smp
mkinitramfs -o nosleep.img 2.6.21-2-sparc64-smp

If you diff the logs, you can clearly see that the scsi bus scan for 
esp0 is still ongoing as local-top/mdadm tries to assemble all MD arrays 
which cause it to fail and the system to hang. All the sleep does really 
is give esp0 enough time to finish scanning it's bus and bring all 
devices up before having mdadm do it's thing.

I can't help but think that there must be a cleaner way to wait for the 
  bus scan to finish before handing control over to mdadm.

Cheers,

Eric
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: nosleep.log
Url: http://lists.alioth.debian.org/pipermail/pkg-mdadm-devel/attachments/20070724/26eb2290/attachment.txt 
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: sleep.log
Url: http://lists.alioth.debian.org/pipermail/pkg-mdadm-devel/attachments/20070724/26eb2290/attachment-0001.txt