From mboxrd@z Thu Jan 1 00:00:00 1970 From: jbrunet@baylibre.com (Jerome Brunet) Date: Mon, 08 Jan 2018 09:42:54 +0100 Subject: [PATCH v2 3/5] ARM64: dts: meson-axg: uart: Add the pinctrl info description In-Reply-To: References: <20180106001044.108163-1-yixun.lan@amlogic.com> <20180106001044.108163-4-yixun.lan@amlogic.com> Message-ID: <1515400974.5048.58.camel@baylibre.com> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Mon, 2018-01-08 at 14:07 +0800, Yixun Lan wrote: > > (the following question just came up while I was looking at this > > patch, but I guess it's more a question towards the pinctrl driver) > > the name of the function looks a bit "weird" since below you are also > > using "uart_ao_b" > > you right here, it's a question related to pinctrl subsystem. > from my point of view, it's even weird from the hardware perspective: > that, the UART function of AO domain route the pin of EE domain.. > > > did you choose "uart_ao_b_gpioz" here because we cannot have the same > > function name for the periphs and AO pinctrl or is there some other > > reason? > > > > Current there is a conflict in the code level which both two pinctrl > tree (EE, AO) are using the same macro[1] to expand the definitions, so > there would be conflict symbol if we name both as 'uart_ao_b' > > I think your idea of having an uniform function 'uart_ao_b' for both > pinctrl subsystem is actually possible/positive.. > > I will think about your suggestion and come up with a patch later, > thanks a lot! > > > [1] drivers/pinctrl/meson/pinctrl-meson.h > > #define FUNCTION(fn) \ > { \ > .name = #fn, \ > .groups = fn ## _groups, \ > .num_groups = ARRAY_SIZE(fn ## _groups), \ > } The name feels weird because it should have been uart_ao_b_z ... We missed it in the initial review. Except for correcting the function name, I don't think this justify a change a pinctrl