Thread (74 messages) 74 messages, 13 authors, 2014-03-21

Re: [RFC PATCH] [media]: of: move graph helpers from drivers/media/v4l2-core to drivers/of

From: Philipp Zabel <p.zabel@pengutronix.de>
Date: 2014-03-11 13:44:41
Also in: linux-media, lkml

Hi,

Am Dienstag, den 11.03.2014, 15:27 +0200 schrieb Tomi Valkeinen:
On 11/03/14 15:16, Laurent Pinchart wrote:
quoted
quoted
And if I gathered Grant's opinion correctly (correct me if I'm wrong),
he thinks things should be explicit, i.e. the bindings for, say, an
encoder should state that the encoder's output endpoint _must_ contain a
remote-endpoint property, whereas the encoder's input endpoint _must
not_ contain a remote-endpoint property.
Actually my understand was that DT links would have the same direction as the 
data flow. There would be no ambiguity in that case as the direction of the 
Ok. At least at some point in the discussion the rule of thumb was to
have the "master" point at the "slave", which are a bit vague terms, but
I think both display and camera controllers were given as examples of
"master".
quoted
data flow is known. What happens for bidirectional data flows still need to be 
discussed though. And if we want to use the of-graph bindings to describe 
graphs without a data flow, a decision will need to be taken there too.
Yep. My worry is that if the link direction is defined in the bindings
for the device, somebody will at some point have a complex board which
happens to use two devices in such a way, that either neither of them
point to each other, or both point to each other.

Maybe we can make sure that never happens, but I feel a bit nervous
about it especially for bi-directional cases. If the link has no clear
main-dataflow direction, it may be a bit difficult to say which way to link.

With double-linking all those concerns just disappear.
another possibility would be to just leave it open in the generic
bindings:

"Two endpoints are considered linked together if any of them contains a
 'remote-endpoint' phandle property that points to the other. If both
 contain a 'remote-endpoint' property, both must point at each other."

As long as we make sure that the generic code is able to generate the
undirected graph from this, we could then let more specific bindings
(like video-interfaces.txt) define that the phandle links always should
point in both directions, in data flow direction, north, or outwards
from the 'master'.

regards
Philipp
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help