Messages in this thread | | | Date | Tue, 14 May 2013 10:23:17 +0200 | Subject | Re: [PATCH 2/9] mfd: ab8500-core: Use the correct driver name when enabling gpio/pinctrl | From | Linus Walleij <> |
| |
On Wed, May 8, 2013 at 3:29 PM, Lee Jones <lee.jones@linaro.org> wrote:
> When we're using Device Tree to enable GPIO drivers we're forced to be > OS agnostic, thus we are forbidden use names like pinctrl as they are > specific only to Linux.
Oh that depends. I have hade lectures on pin control at the embedded systems conference and there I treat it as a neutral term referring to this kind of electronic constructions.
However how something is established in OF ontology I don't know so you might be right there, but you're making it sound like Documentation/devicetree/bindings/pinctrl/ should not exist and all those bindings be moved to gpio or something if the assumption that "pin control" or the abbreviation pinctrl is a Linux-specific term. (Which is not my interpretation.)
> However, when we are registering devices using > internal systems such as MFD or platform registration, we can use such > terminology. In this case we can and should use the platform device ID > mechanism to specify which device we wish to utilise by detailing > pinctrl-<device_name>. > > Cc: Samuel Ortiz <sameo@linux.intel.com> > Signed-off-by: Lee Jones <lee.jones@linaro.org>
The commit message should state what regression it is actually solving. Which is that when using a non-DT boot, the ABx500 pinctrl devices are not probed, right?
Apart from the commit message: Acked-by: Linus Walleij <linus.walleij@linaro.org>
On the patch as such.
Yours, Linus Walleij
| |