Hi Manikandan,
I am using buildroot 2024.04 sama5d3_eds_headless_defconfig config setup which leads to the following: 1. uboot from https://github.com/linux4microchip/u-boot-mchp.git 2. uboot version linux4microchip-2024.10-rc2 3. uboot config sama5d3_xplained_mmc
My board is sama5d3 EDS with DP83630 PHY connected to EMAC.
I confirm that using configuration above, I was not able to make Ethernet (EMAC) in uboot work because it was failing to configure EMAC pins to the required peripheral mode. The at91_pin_check_config function in pinctrl-at91 returned -EINVAL because of nbanks=0. After applying the patch, the Ethernet is working fine.
I also confirm that after applying the patch, I was able to toggle IO pin with gpio command (tried PIOC18 which is accessible through IO Connector #1), although I cannot tell you if this functionality also worked before applying the patch.
Regards, Alexey.
-----Original Message----- From: Manikandan.M@microchip.com Manikandan.M@microchip.com Sent: Monday, 21 October 2024 13:19 To: alexey@all4bambi.com; eugen.hristev@linaro.org; u-boot@lists.denx.de Cc: Varshini.Rajendran@microchip.com; Hari.PrasathGE@microchip.com Subject: Re: [PATCH 1/1] Fixed sama5d3 dts file so PIO sections are inside pinctrl as in kernel dts
Hi Alexey,
Can you confirm if you are using the u-boot-mchp repo from GitHub [1] to test the sama5d3_eds board where the driver changes to align U-Boot and Kernel DT are already present. The pinctrl driver probe fails with the proposed change and will cause regression on other boards that uses the same pinctrl PIO3 driver in the mainline repo.
[1] --> https://github.com/linux4microchip/u-boot-mchp
On 20/10/24 10:44 am, Alexey Tsirlin wrote:
EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
Hi Manikandan,
I have tested gpio cmd option on SAMA5D3 EDS board (BTW it wasn't enabled by default in sama5d3_xplained_mmc_defconfig which is used by this board) in u-boot with the DT change as I proposed and it seems to work fine, at least it detects all the 5 GPIO ports (A-E). pinmux cmd does not work too much, but this is because pinctrl-at91 does not implement get_pins_count function. Without the proposed change I was not able to make the Ethernet (EMAC) detect the PHY because MDIO interface was not working - the correct peripheral mode for the EMAC pins was not set as defined in DT.
Regards, Alexey.
-----Original Message----- From: Manikandan.M@microchip.com Manikandan.M@microchip.com Sent: Friday, 18 October 2024 12:30 To: eugen.hristev@linaro.org; alexey@all4bambi.com; u-boot@lists.denx.de Cc: Varshini.Rajendran@microchip.com; Hari.PrasathGE@microchip.com Subject: Re: [PATCH 1/1] Fixed sama5d3 dts file so PIO sections are inside pinctrl as in kernel dts
Hi Eugen,
On 18/10/24 12:45 pm, Eugen Hristev wrote:
[You don't often get email from eugen.hristev@linaro.org. Learn why this is important at https://aka.ms/LearnAboutSenderIdentification ]
EXTERNAL EMAIL: Do not click links or open attachments unless you know the content is safe
Hello Alexey,
Please fix the subject to adhere to the rules ARM: dts: .... etc, if you are unsure, please follow previous commits that touched this file.
On 10/17/24 11:51, Manikandan.M@microchip.com wrote:
Hi Alexey,
On 15/10/24 8:23 pm, Alexey Tsirlin wrote:
This allows setting the GPIO parameters from device tree, otherwise the at91_pin_check_config will fail because the priv->nbanks equal to zero
I remember these pin banks are outside of the pinctrl because the driver fails to probe them if they are inside. Is this no longer true ?
Indeed, you are correct.With the current code base the pinctrl fails to probe if they are defined inside. I started to review this code with an intention that my changes for pinctrl driver and DT to align U-Boot pinctrl DT node with the kernel had already been made upstream, however that is not the case. This patch is valid and necessary only after when my changes are up-streamed otherwise driver probe will fail
Since I don't own a board with this SoC, Alexey, could you kindly check the GPIO functions and determine whether this patch is actually necessary for the problem you're experiencing?If not, after incorporating the driver changes, you can send this as part of the DT alignment
Manikandan, is it possible to test this on the board? and use the gpio command in U-boot to toggle the pins , like e.g. for the LEDs and see if there is no regression ?
Thanks, Eugen
Signed-off-by: Alexey Tsirlin alexey@all4bambi.com
arch/arm/dts/sama5d3.dtsi | 111+++++++++++++++++++------------------- 1 file changed, 56 insertions(+), 55 deletions(-)
diff --git a/arch/arm/dts/sama5d3.dtsi b/arch/arm/dts/sama5d3.dtsi index 4c03a302ec..c671ea42f2 100644 --- a/arch/arm/dts/sama5d3.dtsi +++ b/arch/arm/dts/sama5d3.dtsi @@ -873,66 +873,67 @@ AT91_PIOE 17 AT91_PERIPH_B AT91_PINCTRL_NONE>; /* PE17 periph B, conflicts with A17 */ }; };
};pioA: gpio@fffff200 {compatible = "atmel,at91sam9x5-gpio","atmel,at91rm9200-gpio";
reg = <0xfffff200 0x100>;interrupts = <6 IRQ_TYPE_LEVEL_HIGH 1>;#gpio-cells = <2>;gpio-controller;interrupt-controller;har#interrupt-cells = <2>;clocks = <&pioA_clk>;bootph-all;};
pioA: gpio@fffff200 {compatible ="atmel,at91sam9x5-gpio", "atmel,at91rm9200-gpio";
Spaces instead of tab before 'compatible', should be consistent with the remaining properties of pioA node.
reg = <0xfffff200 0x100>;interrupts = <6IRQ_TYPE_LEVEL_HIGH 1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioA_clk>;bootph-all;};
pioB: gpio@fffff400 {compatible = "atmel,at91sam9x5-gpio","atmel,at91rm9200-gpio";
reg = <0xfffff400 0x100>;interrupts = <7 IRQ_TYPE_LEVEL_HIGH 1>;#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioB_clk>;bootph-all;};
pioB: gpio@fffff400 {compatible ="atmel,at91sam9x5-gpio", "atmel,at91rm9200-gpio";
Ditto
reg = <0xfffff400 0x100>;interrupts = <7IRQ_TYPE_LEVEL_HIGH 1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioB_clk>;bootph-all;};
pioC: gpio@fffff600 {compatible = "atmel,at91sam9x5-gpio","atmel,at91rm9200-gpio";
reg = <0xfffff600 0x100>;interrupts = <8 IRQ_TYPE_LEVEL_HIGH 1>;#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioC_clk>;bootph-all;};
pioC: gpio@fffff600 {compatible ="atmel,at91sam9x5-gpio", "atmel,at91rm9200-gpio";
Ditto
reg = <0xfffff600 0x100>;interrupts = <8IRQ_TYPE_LEVEL_HIGH 1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioC_clk>;bootph-all;};
pioD: gpio@fffff800 {compatible = "atmel,at91sam9x5-gpio","atmel,at91rm9200-gpio";
reg = <0xfffff800 0x100>;interrupts = <9 IRQ_TYPE_LEVEL_HIGH 1>;#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioD_clk>;bootph-all;};
pioD: gpio@fffff800 {compatible ="atmel,at91sam9x5-gpio", "atmel,at91rm9200-gpio";
Ditto
reg = <0xfffff800 0x100>;interrupts = <9IRQ_TYPE_LEVEL_HIGH 1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioD_clk>;bootph-all;};pioE: gpio@fffffa00 {compatible ="atmel,at91sam9x5-gpio", "atmel,at91rm9200-gpio";
Ditto
reg = <0xfffffa00 0x100>;interrupts = <10IRQ_TYPE_LEVEL_HIGH 1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioE_clk>;bootph-all;};Extra line
pioE: gpio@fffffa00 {compatible = "atmel,at91sam9x5-gpio","atmel,at91rm9200-gpio";
reg = <0xfffffa00 0x100>;interrupts = <10 IRQ_TYPE_LEVEL_HIGH1>;
#gpio-cells = <2>;gpio-controller;interrupt-controller;#interrupt-cells = <2>;clocks = <&pioE_clk>;bootph-all; }; pmc: pmc@fffffc00 {-- Thanks and Regards, Manikandan M.
-- Thanks and Regards, Manikandan M.