Public SourceMage Discussion List

[Thomas Orgis <thomas-forum AT orgis.org>]

Am Thu, 4 Jan 2007 11:56:45 -0600
schrieb Jeremy Blosser <jblosser-smgl AT firinn.org>: 


> over that init.d issue speaks volumes on what we're currently actually
> [...]
> open that long?  It's even 99% diagnosed, it "just" needs fixing.
> Obviously I haven't had time, but I find it hard to belive *no one else*
> has had time either. 

I'd add that apparently noone is affected by the prognosted failure except 
perhaps the reporter of that old bug?
We didn't support any device mapper targets that were set up in 
initrd/initramfs until I thought it would be cool to use suspend2 with 
encrypted swap.
I encrypted my swap, fiddled around with busybox to get my initramfs and then 
added this dmsetup call that rediscovers the volumes on "real" boot before 
mount in the init script.
No rocket science (well, the init.d part at least;-), all it needed was a 
_use case_.
We at least need one LVM root setup that could be bitten by the bug to test 
the fix on. For me that means setting up some spare box with LVM root from 
scratch (since I didn't do any LVM stuff yet... only bare device-mapper for 
encryption).
Isn't there ANYONE among our devs who has LVM root or at least basic LVM 
experience and some hardware to quickly set up the test case?
If no, then perhaps we should just push stable out and wait for any complaint 
to happen (we could warn about LVM root in release notes... eh... if we had 
such a thing on scribe update).


Thomas.