[PATCH] pinctrl: document the pinctrl PM states

Tony Lindgren tony at atomide.com
Thu Jun 20 02:38:24 EDT 2013


* Stephen Warren <swarren at wwwdotorg.org> [130619 13:08]:
> On 06/17/2013 01:20 AM, Tony Lindgren wrote:
> ...
> > First, I think the concept of remuxing (or even checking) _all_ the pins
> > for a consumer device is wrong on most if not all hardware. For past 10
> > years I have not seen a case where _all_ the pins for a device would need
> > to be remuxed for any reason.
> > 
> > This means the named states "default", "idle" and "sleep" cannot represent
> > the needs of hardware. So need to have a bit finer granularity for this.
> 
> Well, I don't think that's quite true.
> 
> We can certainly list configurations for all pins in each of those 3
> states without any issue.
> 
> The only issue here is whether the data is normalized or not. By listing
> the entire configuration in each state, it's not really normalized.
> 
> By separating the static and dynamic parts into separate states as you
> propose below, it is more normalized.
> 
> However, the final configuration of HW is the same either way, and hence
> the overall configuration.

Right, and we maintain compability with existing binding. The "static"
we may need to keep as "default" though to avoid breaking existing
bindings. I think that can be covered by documenting it for the dynamic
cases.
 
> > Below are the pin remuxing cases I'm aware of:
> ...
> > Then for dealing with the above examples, I think we already have a
> > pretty good setup in pinctrl framework to deal with this with the
> > named modes. But I think that to do this properly with the named
> > modes we should have named modes like the following:
> > 
> > "static" && ("active" || "idle")
> > "static" && ("rx" || "tx")
> > 
> > Here the "static" pins would be set during driver probe and never
> > need to change until the driver is unloaded. This is close to what we
> > currently call "default". But we should probably make it clear that
> > these will not change to avoid confusion. See below for more info.
> > 
> > The the non-static states like "active"/"idle", or "rx"/"tx", can
> > be set in addition to "static", but they should not be subsets of
> > "static" to avoid the issues Stephen described earlier. This way we
> > allow the named modes to do the work for us while protecting the
> > claimed pins.
> 
> Yes, I think this can certainly work conceptually. It's equivalent to
> pre-computing which parts of the overall state don't change between the
> currently-defined "global" active/idle states and then applying the
> diffs at runtime - rather like what I suggested before, but without the
> pinctrl code having to do the diff at runtime. I'm not sure if I have
> (yet) a strong opinion on whether allowing multiple states to be active
> at once (i.e. static plus active) is the correct way to go. Maybe once
> I've finished reading the thread...

I don't think there's any issue for having multiple sets active the same
is an issue, we're already doing it quite a bit although for different
device drivers so we have the framework ready for that already.

For the dynamic muxing and reconfiguring of the pins we need to keep
the code down to minimum as that can happen every time we enter and exit
idle. So the checking of pins and functions to set should be only done
once during the driver probe.

Regards,

Tony



More information about the linux-arm-kernel mailing list