From mboxrd@z Thu Jan 1 00:00:00 1970 From: s.trumtrar@pengutronix.de (Steffen Trumtrar) Date: Thu, 25 Sep 2014 10:17:48 +0200 Subject: [RFC] pinctrl driver for Zynq In-Reply-To: References: Message-ID: <20140925081748.GB26991@pengutronix.de> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org Hi! On Wed, Sep 24, 2014 at 02:09:14PM -0700, S?ren Brinkmann wrote: > Hi, > > I think I have pinctrl driver that is covering the pinmux options of > Zynq and I also figured out how the DT bindings work. > > But there are a couple of things that probably could be done better. > > One thing making the DT bindings explode, seems to be all those single > pin functions that can be muxed to every pin. > Next to GPIO, this applies to SD card and write protect - which are even > present twice since Zynq has two SDIO cores. Just these functions > account for a couple of hundred nodes in the DT and a bunch of lines in > the driver. Is there a better way to do this? > > In particular for GPIO there seemed to be a better solution with > implementing gpio_request_enable(), but that seemed to allow GPIO in > parallel to request and mux the pin which does not work on Zynq. IOW: I > expected the core would reject a call of gpio_request_enable for a pin > that is already muxed to some other function, but that was not the case > in my testing. Am I missing something here? > > And finally, for SD card detect and write protect, we actually have to > disable the muxing. The problem with those functions is, that they have > a dedicated mux for that function which is in parallel to the "normal" > pinmuxes. So, muxing a "normal" function to a pin would not void the > muxing of the SD signals. I thought this would be easily resolved by > implementing the 'disable' op, but after I did that, I noticed that > there is only a stale documentation comment of this member of struct > pinmux_ops left, the actual function pointer is gone. > > Thanks, > S?ren > > ------------8<-----------------8<-------------------8<--------------8<---------- > Date: Mon, 15 Sep 2014 17:24:35 -0700 > Subject: [PATCH RFC] pinctrl: Add driver for Zynq > > Signed-off-by: Soren Brinkmann > --- > arch/arm/boot/dts/zynq-7000.dtsi | 3039 +++++++++++++++++++++++++++++++++++++- > arch/arm/boot/dts/zynq-zc706.dts | 13 + > arch/arm/mach-zynq/Kconfig | 1 + > drivers/pinctrl/Kconfig | 8 + > drivers/pinctrl/Makefile | 1 + > drivers/pinctrl/pinctrl-zynq.c | 927 ++++++++++++ > 6 files changed, 3988 insertions(+), 1 deletion(-) > create mode 100644 drivers/pinctrl/pinctrl-zynq.c > > diff --git a/arch/arm/boot/dts/zynq-7000.dtsi b/arch/arm/boot/dts/zynq-7000.dtsi > index 6cc83d4c6c76..814873da0392 100644 > --- a/arch/arm/boot/dts/zynq-7000.dtsi > +++ b/arch/arm/boot/dts/zynq-7000.dtsi > @@ -229,7 +229,7 @@ > slcr: slcr at f8000000 { > #address-cells = <1>; > #size-cells = <1>; > - compatible = "xlnx,zynq-slcr", "syscon"; > + compatible = "xlnx,zynq-slcr", "syscon", "simple-bus"; > reg = <0xF8000000 0x1000>; > ranges; > clkc: clkc at 100 { > @@ -250,6 +250,3043 @@ > "dbg_trc", "dbg_apb"; > reg = <0x100 0x100>; > }; > + > + pinctrl0: pinctrl at 700 { > + compatible = "xlnx,pinctrl-zynq"; > + reg = <0x700 0x200>; > + > + pinctrl_i2c0_0: pinctrl-i2c0 at 0 { > + i2c0-mux { > + function = "i2c0"; > + pins = "i2c0_0_grp"; > + }; > + }; > + (...) > + pinctrl_sdio1_cd_54: pinctrl-sdio1_cd at 54 { > + sdio1_cd-mux { > + function = "sdio1_cd"; > + pins = "sdio1_emio_cd_grp"; > + }; > + }; > + }; > }; > Wouldn't this reaaally bloat the dtb? I know that imx decided to remove all the pinctrls from the dtsis, because the dtbs got to big. Regards, Steffen -- Pengutronix e.K. | | Industrial Linux Solutions | http://www.pengutronix.de/ | Peiner Str. 6-8, 31137 Hildesheim, Germany | Phone: +49-5121-206917-0 | Amtsgericht Hildesheim, HRA 2686 | Fax: +49-5121-206917-5555 |