qemu-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Qemu-devel] [PATCH v2 3/7] device_tree: introduce qemu_fdt_node_pat


From: David Gibson
Subject: Re: [Qemu-devel] [PATCH v2 3/7] device_tree: introduce qemu_fdt_node_path
Date: Wed, 13 Jan 2016 10:07:32 +1100
User-agent: Mutt/1.5.24 (2015-08-30)

On Tue, Jan 12, 2016 at 06:02:00PM +0100, Eric Auger wrote:
> Hi David,
> On 01/12/2016 05:28 AM, David Gibson wrote:
> > On Mon, Jan 11, 2016 at 11:35:50AM +0100, Eric Auger wrote:
> >> Hi David,
> >> On 01/11/2016 03:38 AM, David Gibson wrote:
> >>> On Wed, Jan 06, 2016 at 03:13:21PM +0000, Eric Auger wrote:
> >>>> This new helper routine returns the node path of a device
> >>>> referred to by its node name and compat string.
> >>>
> >>> What if there are multiple nodes matching the name and compat?
> >> The function would return the first one. I can improve the doc comment.
> >> Do you think it is a problem stopping at the first one? Is it a real
> >> life test case I have to handle here?
> > 
> > Well, I don't know of a specific system which will have this, but it's
> > absolutely possible to get this situation:  e.g. two different PCI
> > busses, both of which have their own slot 0 populated with different
> > instances of the same device.
> > 
> > Whether it's possible for platform devices will depend on the
> > platform's specific bus toplogies, but you certainly can't rule it out
> > in general.
> OK I will handle that case then. I hope I will be able to test it.

One case where this might occur in practice for platform devices is if
you have a system built with multiple instances of a SoC.  A
peripheral on SoC 0 is likely to have the same name and compatible as
the same peripheral on Soc 1, just inside a different parent node.

> > I could consider adding a new libfdt function like
> > fdt_node_offset_by_compatible() that searches by name as well.  It's
> > just I'm not sure that matching by name and compatible isn't a sign of
> > a poor approach in the caller.
> well I can't really comment. That looked the most straightforward to me
> given the current libfdt API. But not sure it's worth to invest in a new
> function in libfdt
> 
> Thanks!
> 
> Eric
> > 
> 

-- 
David Gibson                    | I'll have my music baroque, and my code
david AT gibson.dropbear.id.au  | minimalist, thank you.  NOT _the_ _other_
                                | _way_ _around_!
http://www.ozlabs.org/~dgibson

Attachment: signature.asc
Description: PGP signature


reply via email to

[Prev in Thread] Current Thread [Next in Thread]