[PATCH 2.6.23] ibmebus: Prevent bus_id collisions

Nathan Lynch ntl at pobox.com
Fri Aug 31 03:56:09 EST 2007


Hi Joachim-

Joachim Fenkes wrote:
> Nathan Lynch <ntl at pobox.com> wrote on 29.08.2007 20:12:32:
> > Will anything break?
> 
> Nope. Userspace programs should not depend on ibmebus' way of naming the 
> devices; especially since some overly long loc_codes tended to be 
> truncated and thus rendered useless. I have tested IBM's DLPAR tools 
> against the changed kernel, and they didn't break.

Okay.


> > Also, I dislike this approach of duplicating the firmware device tree
> > path in sysfs.
> 
> Why? Any specific reasons for your dislike?

struct device's bus_id field is but 20 bytes in size.  Too close for
comfort?


> > Are GX/ibmebus devices guaranteed to be children of
> > the same node in the OF device tree?  If so, their unit addresses will
> > be unique, and therefore suitable values for bus_id.  I believe this
> > is what the powerpc vio bus code does.
> 
> While there's no such guarantee (as in "officially signed document"), yes, 
> I expect future GX devices to also appear beneath the OFDT root node. For 
> the existing devices, the unit addresses are already part of the device 
> name, so I save the need to use sprintf() again. Plus, I rather like using 
> the full_name since it also contains a descriptive name as opposed to 
> being just nondescript numbers, helping the layman (ie user) to make sense 
> out of a dev_id.

Okay, but your layman isn't supposed to be relying on any
user-friendly properties of the name :) Hope he doesn't work on a
distro installer.

Anyway, if you're still confident in this approach, I relent.  :)



More information about the Linuxppc-dev mailing list