Re: Ok to commit ubd patch ?

[Sven Luther]

On Sun, Dec 19, 2004 at 09:43:25AM +1100, Andrew Clausen wrote:
> On Sat, Dec 18, 2004 at 09:31:03PM +0100, Sven Luther wrote:
> > It doesn't seems to be part of 1.6.20 though, an oversight ? 
> 
> It was an oversight.

Ok. I had a look yesterday, and this patch is somehow dependent on the s390
patch, which includes the DASD type. I suppose your discussion below would
also concern this one ? I remember you telling me that the s390 author should
do the necessary FSF paperwork to get his stuff included.

> However, looking at it again, I have decided to change
> _device_probe_type() to set PED_DEVICE_UNKNOWN rather than
> PED_DEVICE_FILE for block devices that are not in one of the
> existing major categories.  This means the kernel gets notified
> correctly of partition table changes.

Ok, seems fair to me, but i will ask the patch author about this. Matt, do you
have any comment on this ? It is about the :

## ubd.dpatch by Matt Zimmerman <address@hidden>
## DP: Recognize UML UBD devices.
## DP: Closes: #258188

patch to parted.

> I'm not sure how useful it is having a long list of PedDeviceType...
> Are there any arguments in favour of PED_DEVICE_UBD?  Remember,
> libparted is supposed to be portable...

Not that i know of, especially as they are not used anywhere in the rest of
the code. But then i don't really understand that part of the code, nor what
the devicetype is used for in the first place.

> (Actually, it wasn't my idea in the first place.  If it were me,
> I would have left PedDeviceType out all together)

Ah,

Friendly,

Sven Luther