All of lore.kernel.org
 help / color / mirror / Atom feed
* [PATCH v1 0/4] ASPEED sgpio driver enhancement.
@ 2021-05-26  9:46 ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
with 80 pins, AST2500/AST2400 SoC has 1 SGPIO master interface that
supports up to 80 pins.
In the current driver design, the max number of sgpio pins is hardcoded
in macro MAX_NR_HW_SGPIO and the value is 80.

For supporting sgpio master interfaces of AST2600 SoC, the patch series
contains the following enhancement:
- Convert txt dt-bindings to yaml.
- Update aspeed dtsi to support the enhanced sgpio.
- Get the max number of sgpio that SoC supported from dts.
- Support muiltiple SGPIO master interfaces.
- Support up to 128 pins.

Please help to review.

Thanks,
Steven


Steven Lee (4):
  dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  ARM: dts: aspeed-g6: Add SGPIO node.
  ARM: dts: aspeed-g5: Modify sgpio node for the enhanced sgpio driver.
  gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support

 .../bindings/gpio/aspeed,sgpio.yaml           |  91 +++++++++
 .../devicetree/bindings/gpio/sgpio-aspeed.txt |  46 -----
 arch/arm/boot/dts/aspeed-g5.dtsi              |   3 +-
 arch/arm/boot/dts/aspeed-g6.dtsi              |  32 +++
 drivers/gpio/gpio-aspeed-sgpio.c              | 193 ++++++++++++------
 5 files changed, 250 insertions(+), 115 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
 delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt

-- 
2.17.1


^ permalink raw reply	[flat|nested] 38+ messages in thread

* [PATCH v1 0/4] ASPEED sgpio driver enhancement.
@ 2021-05-26  9:46 ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
with 80 pins, AST2500/AST2400 SoC has 1 SGPIO master interface that
supports up to 80 pins.
In the current driver design, the max number of sgpio pins is hardcoded
in macro MAX_NR_HW_SGPIO and the value is 80.

For supporting sgpio master interfaces of AST2600 SoC, the patch series
contains the following enhancement:
- Convert txt dt-bindings to yaml.
- Update aspeed dtsi to support the enhanced sgpio.
- Get the max number of sgpio that SoC supported from dts.
- Support muiltiple SGPIO master interfaces.
- Support up to 128 pins.

Please help to review.

Thanks,
Steven


Steven Lee (4):
  dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  ARM: dts: aspeed-g6: Add SGPIO node.
  ARM: dts: aspeed-g5: Modify sgpio node for the enhanced sgpio driver.
  gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support

 .../bindings/gpio/aspeed,sgpio.yaml           |  91 +++++++++
 .../devicetree/bindings/gpio/sgpio-aspeed.txt |  46 -----
 arch/arm/boot/dts/aspeed-g5.dtsi              |   3 +-
 arch/arm/boot/dts/aspeed-g6.dtsi              |  32 +++
 drivers/gpio/gpio-aspeed-sgpio.c              | 193 ++++++++++++------
 5 files changed, 250 insertions(+), 115 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
 delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt

-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-26  9:46 ` Steven Lee
@ 2021-05-26  9:46   ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

SGPIO bindings should be converted as yaml format.
In addition to the file conversion, a new property max-ngpios is
added in the yaml file as well.
The new property is required by the enhanced sgpio driver for
making the configuration of the max number of gpio pins more flexible.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
 .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
 2 files changed, 91 insertions(+), 46 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
 delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt

diff --git a/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml b/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
new file mode 100644
index 000000000000..8b8588a29685
--- /dev/null
+++ b/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
@@ -0,0 +1,91 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/gpio/sgpio-aspeed.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Aspeed SGPIO controller
+
+maintainers:
+  - Andrew Jeffery <andrew@aj.id.au>
+
+description:
+  This SGPIO controller is for ASPEED AST2400, AST2500 and AST2600 SoC,
+  AST2600 have two sgpio master one with 128 pins another one with 80 pins,
+  AST2500/AST2400 have one sgpio master with 80 pins. Each of the Serial
+  GPIO pins can be programmed to support the following options
+  - Support interrupt option for each input port and various interrupt
+    sensitivity option (level-high, level-low, edge-high, edge-low)
+  - Support reset tolerance option for each output port
+  - Directly connected to APB bus and its shift clock is from APB bus clock
+    divided by a programmable value.
+  - Co-work with external signal-chained TTL components (74LV165/74LV595)
+
+properties:
+  compatible:
+    enum:
+      - aspeed,ast2400-sgpiom
+      - aspeed,ast2500-sgpiom
+      - aspeed,ast2600-sgpiom
+
+  reg:
+    maxItems: 1
+
+  gpio-controller: true
+
+  '#gpio-cells':
+    const: 2
+
+  interrupts:
+    maxItems: 1
+
+  interrupt-controller: true
+
+  clocks:
+    maxItems: 1
+
+  ngpios:
+    minimum: 0
+    maximum: 128
+
+  max-ngpios:
+    description:
+      represents the number of actual hardware-supported GPIOs (ie,
+      slots within the clocked serial GPIO data). Since each HW GPIO is both an
+      input and an output, we provide max_ngpios * 2 lines on our gpiochip
+      device. We also use it to define the split between the inputs and
+      outputs; the inputs start at line 0, the outputs start at max_ngpios.
+    minimum: 0
+    maximum: 128
+
+  bus-frequency: true
+
+required:
+  - compatible
+  - reg
+  - gpio-controller
+  - '#gpio-cells'
+  - interrupts
+  - interrupt-controller
+  - ngpios
+  - max-ngpios
+  - clocks
+  - bus-frequency
+
+additionalProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/clock/aspeed-clock.h>
+    sgpio: sgpio@1e780200 {
+        #gpio-cells = <2>;
+        compatible = "aspeed,ast2500-sgpiom";
+        gpio-controller;
+        interrupts = <40>;
+        reg = <0x1e780200 0x0100>;
+        clocks = <&syscon ASPEED_CLK_APB>;
+        interrupt-controller;
+        ngpios = <8>;
+        max-ngpios = <80>;
+        bus-frequency = <12000000>;
+    };
diff --git a/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt b/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
deleted file mode 100644
index be329ea4794f..000000000000
--- a/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
+++ /dev/null
@@ -1,46 +0,0 @@
-Aspeed SGPIO controller Device Tree Bindings
---------------------------------------------
-
-This SGPIO controller is for ASPEED AST2500 SoC, it supports up to 80 full
-featured Serial GPIOs. Each of the Serial GPIO pins can be programmed to
-support the following options:
-- Support interrupt option for each input port and various interrupt
-  sensitivity option (level-high, level-low, edge-high, edge-low)
-- Support reset tolerance option for each output port
-- Directly connected to APB bus and its shift clock is from APB bus clock
-  divided by a programmable value.
-- Co-work with external signal-chained TTL components (74LV165/74LV595)
-
-Required properties:
-
-- compatible : Should be one of
-  "aspeed,ast2400-sgpio", "aspeed,ast2500-sgpio"
-- #gpio-cells : Should be 2, see gpio.txt
-- reg : Address and length of the register set for the device
-- gpio-controller : Marks the device node as a GPIO controller
-- interrupts : Interrupt specifier, see interrupt-controller/interrupts.txt
-- interrupt-controller : Mark the GPIO controller as an interrupt-controller
-- ngpios : number of *hardware* GPIO lines, see gpio.txt. This will expose
-  2 software GPIOs per hardware GPIO: one for hardware input, one for hardware
-  output. Up to 80 pins, must be a multiple of 8.
-- clocks : A phandle to the APB clock for SGPM clock division
-- bus-frequency : SGPM CLK frequency
-
-The sgpio and interrupt properties are further described in their respective
-bindings documentation:
-
-- Documentation/devicetree/bindings/gpio/gpio.txt
-- Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
-
-  Example:
-	sgpio: sgpio@1e780200 {
-		#gpio-cells = <2>;
-		compatible = "aspeed,ast2500-sgpio";
-		gpio-controller;
-		interrupts = <40>;
-		reg = <0x1e780200 0x0100>;
-		clocks = <&syscon ASPEED_CLK_APB>;
-		interrupt-controller;
-		ngpios = <8>;
-		bus-frequency = <12000000>;
-	};
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-26  9:46   ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

SGPIO bindings should be converted as yaml format.
In addition to the file conversion, a new property max-ngpios is
added in the yaml file as well.
The new property is required by the enhanced sgpio driver for
making the configuration of the max number of gpio pins more flexible.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
 .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
 2 files changed, 91 insertions(+), 46 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
 delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt

diff --git a/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml b/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
new file mode 100644
index 000000000000..8b8588a29685
--- /dev/null
+++ b/Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
@@ -0,0 +1,91 @@
+# SPDX-License-Identifier: (GPL-2.0-only OR BSD-2-Clause)
+%YAML 1.2
+---
+$id: http://devicetree.org/schemas/gpio/sgpio-aspeed.yaml#
+$schema: http://devicetree.org/meta-schemas/core.yaml#
+
+title: Aspeed SGPIO controller
+
+maintainers:
+  - Andrew Jeffery <andrew@aj.id.au>
+
+description:
+  This SGPIO controller is for ASPEED AST2400, AST2500 and AST2600 SoC,
+  AST2600 have two sgpio master one with 128 pins another one with 80 pins,
+  AST2500/AST2400 have one sgpio master with 80 pins. Each of the Serial
+  GPIO pins can be programmed to support the following options
+  - Support interrupt option for each input port and various interrupt
+    sensitivity option (level-high, level-low, edge-high, edge-low)
+  - Support reset tolerance option for each output port
+  - Directly connected to APB bus and its shift clock is from APB bus clock
+    divided by a programmable value.
+  - Co-work with external signal-chained TTL components (74LV165/74LV595)
+
+properties:
+  compatible:
+    enum:
+      - aspeed,ast2400-sgpiom
+      - aspeed,ast2500-sgpiom
+      - aspeed,ast2600-sgpiom
+
+  reg:
+    maxItems: 1
+
+  gpio-controller: true
+
+  '#gpio-cells':
+    const: 2
+
+  interrupts:
+    maxItems: 1
+
+  interrupt-controller: true
+
+  clocks:
+    maxItems: 1
+
+  ngpios:
+    minimum: 0
+    maximum: 128
+
+  max-ngpios:
+    description:
+      represents the number of actual hardware-supported GPIOs (ie,
+      slots within the clocked serial GPIO data). Since each HW GPIO is both an
+      input and an output, we provide max_ngpios * 2 lines on our gpiochip
+      device. We also use it to define the split between the inputs and
+      outputs; the inputs start at line 0, the outputs start at max_ngpios.
+    minimum: 0
+    maximum: 128
+
+  bus-frequency: true
+
+required:
+  - compatible
+  - reg
+  - gpio-controller
+  - '#gpio-cells'
+  - interrupts
+  - interrupt-controller
+  - ngpios
+  - max-ngpios
+  - clocks
+  - bus-frequency
+
+additionalProperties: false
+
+examples:
+  - |
+    #include <dt-bindings/clock/aspeed-clock.h>
+    sgpio: sgpio@1e780200 {
+        #gpio-cells = <2>;
+        compatible = "aspeed,ast2500-sgpiom";
+        gpio-controller;
+        interrupts = <40>;
+        reg = <0x1e780200 0x0100>;
+        clocks = <&syscon ASPEED_CLK_APB>;
+        interrupt-controller;
+        ngpios = <8>;
+        max-ngpios = <80>;
+        bus-frequency = <12000000>;
+    };
diff --git a/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt b/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
deleted file mode 100644
index be329ea4794f..000000000000
--- a/Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
+++ /dev/null
@@ -1,46 +0,0 @@
-Aspeed SGPIO controller Device Tree Bindings
---------------------------------------------
-
-This SGPIO controller is for ASPEED AST2500 SoC, it supports up to 80 full
-featured Serial GPIOs. Each of the Serial GPIO pins can be programmed to
-support the following options:
-- Support interrupt option for each input port and various interrupt
-  sensitivity option (level-high, level-low, edge-high, edge-low)
-- Support reset tolerance option for each output port
-- Directly connected to APB bus and its shift clock is from APB bus clock
-  divided by a programmable value.
-- Co-work with external signal-chained TTL components (74LV165/74LV595)
-
-Required properties:
-
-- compatible : Should be one of
-  "aspeed,ast2400-sgpio", "aspeed,ast2500-sgpio"
-- #gpio-cells : Should be 2, see gpio.txt
-- reg : Address and length of the register set for the device
-- gpio-controller : Marks the device node as a GPIO controller
-- interrupts : Interrupt specifier, see interrupt-controller/interrupts.txt
-- interrupt-controller : Mark the GPIO controller as an interrupt-controller
-- ngpios : number of *hardware* GPIO lines, see gpio.txt. This will expose
-  2 software GPIOs per hardware GPIO: one for hardware input, one for hardware
-  output. Up to 80 pins, must be a multiple of 8.
-- clocks : A phandle to the APB clock for SGPM clock division
-- bus-frequency : SGPM CLK frequency
-
-The sgpio and interrupt properties are further described in their respective
-bindings documentation:
-
-- Documentation/devicetree/bindings/gpio/gpio.txt
-- Documentation/devicetree/bindings/interrupt-controller/interrupts.txt
-
-  Example:
-	sgpio: sgpio@1e780200 {
-		#gpio-cells = <2>;
-		compatible = "aspeed,ast2500-sgpio";
-		gpio-controller;
-		interrupts = <40>;
-		reg = <0x1e780200 0x0100>;
-		clocks = <&syscon ASPEED_CLK_APB>;
-		interrupt-controller;
-		ngpios = <8>;
-		bus-frequency = <12000000>;
-	};
-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
  2021-05-26  9:46 ` Steven Lee
@ 2021-05-26  9:46   ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
with 80 pins.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
index f96607b7b4e2..556ce9535c22 100644
--- a/arch/arm/boot/dts/aspeed-g6.dtsi
+++ b/arch/arm/boot/dts/aspeed-g6.dtsi
@@ -377,6 +377,38 @@
 				#interrupt-cells = <2>;
 			};
 
+			sgpiom0: sgpiom@1e780500 {
+				#gpio-cells = <2>;
+				gpio-controller;
+				compatible = "aspeed,ast2600-sgpiom";
+				reg = <0x1e780500 0x100>;
+				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
+				max-ngpios = <128>;
+				ngpios = <128>;
+				clocks = <&syscon ASPEED_CLK_APB2>;
+				interrupt-controller;
+				bus-frequency = <12000000>;
+				pinctrl-names = "default";
+				pinctrl-0 = <&pinctrl_sgpm1_default>;
+				status = "disabled";
+			};
+
+			sgpiom1: sgpiom@1e780600 {
+				#gpio-cells = <2>;
+				gpio-controller;
+				compatible = "aspeed,ast2600-sgpiom";
+				reg = <0x1e780600 0x100>;
+				interrupts = <GIC_SPI 70 IRQ_TYPE_LEVEL_HIGH>;
+				max-ngpios = <80>;
+				ngpios = <80>;
+				clocks = <&syscon ASPEED_CLK_APB2>;
+				interrupt-controller;
+				bus-frequency = <12000000>;
+				pinctrl-names = "default";
+				pinctrl-0 = <&pinctrl_sgpm2_default>;
+				status = "disabled";
+			};
+
 			gpio1: gpio@1e780800 {
 				#gpio-cells = <2>;
 				gpio-controller;
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
@ 2021-05-26  9:46   ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
with 80 pins.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
 1 file changed, 32 insertions(+)

diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
index f96607b7b4e2..556ce9535c22 100644
--- a/arch/arm/boot/dts/aspeed-g6.dtsi
+++ b/arch/arm/boot/dts/aspeed-g6.dtsi
@@ -377,6 +377,38 @@
 				#interrupt-cells = <2>;
 			};
 
+			sgpiom0: sgpiom@1e780500 {
+				#gpio-cells = <2>;
+				gpio-controller;
+				compatible = "aspeed,ast2600-sgpiom";
+				reg = <0x1e780500 0x100>;
+				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
+				max-ngpios = <128>;
+				ngpios = <128>;
+				clocks = <&syscon ASPEED_CLK_APB2>;
+				interrupt-controller;
+				bus-frequency = <12000000>;
+				pinctrl-names = "default";
+				pinctrl-0 = <&pinctrl_sgpm1_default>;
+				status = "disabled";
+			};
+
+			sgpiom1: sgpiom@1e780600 {
+				#gpio-cells = <2>;
+				gpio-controller;
+				compatible = "aspeed,ast2600-sgpiom";
+				reg = <0x1e780600 0x100>;
+				interrupts = <GIC_SPI 70 IRQ_TYPE_LEVEL_HIGH>;
+				max-ngpios = <80>;
+				ngpios = <80>;
+				clocks = <&syscon ASPEED_CLK_APB2>;
+				interrupt-controller;
+				bus-frequency = <12000000>;
+				pinctrl-names = "default";
+				pinctrl-0 = <&pinctrl_sgpm2_default>;
+				status = "disabled";
+			};
+
 			gpio1: gpio@1e780800 {
 				#gpio-cells = <2>;
 				gpio-controller;
-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 3/4] ARM: dts: aspeed-g5: Modify sgpio node for the enhanced sgpio driver.
  2021-05-26  9:46 ` Steven Lee
@ 2021-05-26  9:46   ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

The enhanced sgpio driver has changed the compatible name to sgpiom
and reads the max number of sgpio pins that SoC supported from
max-ngpios in dts.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 arch/arm/boot/dts/aspeed-g5.dtsi | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/aspeed-g5.dtsi b/arch/arm/boot/dts/aspeed-g5.dtsi
index d733c1f161c1..730cbd48eedb 100644
--- a/arch/arm/boot/dts/aspeed-g5.dtsi
+++ b/arch/arm/boot/dts/aspeed-g5.dtsi
@@ -345,13 +345,14 @@
 
 			sgpio: sgpio@1e780200 {
 				#gpio-cells = <2>;
-				compatible = "aspeed,ast2500-sgpio";
+				compatible = "aspeed,ast2500-sgpiom";
 				gpio-controller;
 				interrupts = <40>;
 				reg = <0x1e780200 0x0100>;
 				clocks = <&syscon ASPEED_CLK_APB>;
 				interrupt-controller;
 				ngpios = <8>;
+				max-ngpios = <80>;
 				bus-frequency = <12000000>;
 				pinctrl-names = "default";
 				pinctrl-0 = <&pinctrl_sgpm_default>;
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 3/4] ARM: dts: aspeed-g5: Modify sgpio node for the enhanced sgpio driver.
@ 2021-05-26  9:46   ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

The enhanced sgpio driver has changed the compatible name to sgpiom
and reads the max number of sgpio pins that SoC supported from
max-ngpios in dts.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 arch/arm/boot/dts/aspeed-g5.dtsi | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/arch/arm/boot/dts/aspeed-g5.dtsi b/arch/arm/boot/dts/aspeed-g5.dtsi
index d733c1f161c1..730cbd48eedb 100644
--- a/arch/arm/boot/dts/aspeed-g5.dtsi
+++ b/arch/arm/boot/dts/aspeed-g5.dtsi
@@ -345,13 +345,14 @@
 
 			sgpio: sgpio@1e780200 {
 				#gpio-cells = <2>;
-				compatible = "aspeed,ast2500-sgpio";
+				compatible = "aspeed,ast2500-sgpiom";
 				gpio-controller;
 				interrupts = <40>;
 				reg = <0x1e780200 0x0100>;
 				clocks = <&syscon ASPEED_CLK_APB>;
 				interrupt-controller;
 				ngpios = <8>;
+				max-ngpios = <80>;
 				bus-frequency = <12000000>;
 				pinctrl-names = "default";
 				pinctrl-0 = <&pinctrl_sgpm_default>;
-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-26  9:46 ` Steven Lee
@ 2021-05-26  9:46   ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
supports up to 80 pins.
In the current driver design, the max number of sgpio pins is hardcoded
in macro MAX_NR_HW_SGPIO and the value is 80.
The patch makes the maximum gpio number *constraint of chip* comes from
the dts. The property name is max-ngpios.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
 1 file changed, 125 insertions(+), 68 deletions(-)

diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
index 64e54f8c30d2..f74bd7988b66 100644
--- a/drivers/gpio/gpio-aspeed-sgpio.c
+++ b/drivers/gpio/gpio-aspeed-sgpio.c
@@ -17,37 +17,28 @@
 #include <linux/spinlock.h>
 #include <linux/string.h>
 
-/*
- * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
- * slots within the clocked serial GPIO data). Since each HW GPIO is both an
- * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
- * device.
- *
- * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
- * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
- */
-#define MAX_NR_HW_SGPIO			80
-#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO
-
 #define ASPEED_SGPIO_CTRL		0x54
 
-#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
+#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)
 #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
 #define ASPEED_SGPIO_ENABLE		BIT(0)
 
 struct aspeed_sgpio {
 	struct gpio_chip chip;
+	struct irq_chip intc;
 	struct clk *pclk;
 	spinlock_t lock;
 	void __iomem *base;
 	int irq;
+	int max_ngpios;
 	int n_sgpio;
 };
 
 struct aspeed_sgpio_bank {
-	uint16_t    val_regs;
-	uint16_t    rdata_reg;
-	uint16_t    irq_regs;
+	u16    val_regs;
+	u16    rdata_reg;
+	u16    irq_regs;
+	u16    tolerance_regs;
 	const char  names[4][3];
 };
 
@@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank aspeed_sgpio_banks[] = {
 		.val_regs = 0x0000,
 		.rdata_reg = 0x0070,
 		.irq_regs = 0x0004,
+		.tolerance_regs = 0x0018,
 		.names = { "A", "B", "C", "D" },
 	},
 	{
 		.val_regs = 0x001C,
 		.rdata_reg = 0x0074,
 		.irq_regs = 0x0020,
+		.tolerance_regs = 0x0034,
 		.names = { "E", "F", "G", "H" },
 	},
 	{
 		.val_regs = 0x0038,
 		.rdata_reg = 0x0078,
 		.irq_regs = 0x003C,
-		.names = { "I", "J" },
+		.tolerance_regs = 0x0050,
+		.names = { "I", "J", "K", "L" },
+	},
+	{
+		.val_regs = 0x0090,
+		.rdata_reg = 0x007C,
+		.irq_regs = 0x0094,
+		.tolerance_regs = 0x00A8,
+		.names = { "M", "N", "O", "P" },
 	},
 };
 
@@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
 	reg_irq_type1,
 	reg_irq_type2,
 	reg_irq_status,
+	reg_tolerance,
 };
 
-#define GPIO_VAL_VALUE      0x00
-#define GPIO_IRQ_ENABLE     0x00
-#define GPIO_IRQ_TYPE0      0x04
-#define GPIO_IRQ_TYPE1      0x08
-#define GPIO_IRQ_TYPE2      0x0C
-#define GPIO_IRQ_STATUS     0x10
+#define GPIO_IRQ_OFFSET_ENABLE     0x00
+#define GPIO_IRQ_OFFSET_TYPE0      0x04
+#define GPIO_IRQ_OFFSET_TYPE1      0x08
+#define GPIO_IRQ_OFFSET_TYPE2      0x0C
+#define GPIO_IRQ_OFFSET_STATUS     0x10
 
 static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
 				     const struct aspeed_sgpio_bank *bank,
@@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
 {
 	switch (reg) {
 	case reg_val:
-		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
+		return gpio->base + bank->val_regs;
 	case reg_rdata:
 		return gpio->base + bank->rdata_reg;
 	case reg_irq_enable:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
 	case reg_irq_type0:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
 	case reg_irq_type1:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
 	case reg_irq_type2:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
 	case reg_irq_status:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
+	case reg_tolerance:
+		return gpio->base + bank->tolerance_regs;
 	default:
 		/* acturally if code runs to here, it's an error case */
 		BUG();
 	}
 }
 
-#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
-#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
+#define GPIO_BANK(x)    ((x) >> 5)
+/* modulo 32 */
+#define GPIO_OFFSET(x)  ((x) & 0x1f)
 #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
 
-static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
+static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, unsigned int max_ngpios)
 {
 	unsigned int bank;
 
-	bank = GPIO_BANK(offset);
+	bank = GPIO_BANK(offset % max_ngpios);
 
 	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
 	return &aspeed_sgpio_banks[bank];
@@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct gpio_chip *gc,
 		unsigned long *valid_mask, unsigned int ngpios)
 {
 	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
+	int max_ngpios = sgpio->max_ngpios;
 	int n = sgpio->n_sgpio;
-	int c = SGPIO_OUTPUT_OFFSET - n;
+	int c = max_ngpios - n;
 
-	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
+	WARN_ON(ngpios < max_ngpios * 2);
 
 	/* input GPIOs in the lower range */
 	bitmap_set(valid_mask, 0, n);
 	bitmap_clear(valid_mask, n, c);
 
-	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
-	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
-	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
+	/* output GPIOS above max_ngpios */
+	bitmap_set(valid_mask, max_ngpios, n);
+	bitmap_clear(valid_mask, max_ngpios + n, c);
 
 	return 0;
 }
@@ -161,30 +166,30 @@ static void aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
 	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
 	int n = sgpio->n_sgpio;
 
-	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
+	WARN_ON(ngpios < sgpio->max_ngpios * 2);
 
 	/* input GPIOs in the lower range */
 	bitmap_set(valid_mask, 0, n);
 	bitmap_clear(valid_mask, n, ngpios - n);
 }
 
-static bool aspeed_sgpio_is_input(unsigned int offset)
+static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
 {
-	return offset < SGPIO_OUTPUT_OFFSET;
+	return offset < max_ngpios;
 }
 
 static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
 {
+	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
 	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
-	const struct aspeed_sgpio_bank *bank = to_bank(offset);
 	unsigned long flags;
 	enum aspeed_sgpio_reg reg;
 	int rc = 0;
 
 	spin_lock_irqsave(&gpio->lock, flags);
 
-	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
-	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
+	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
+	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
 
 	spin_unlock_irqrestore(&gpio->lock, flags);
 
@@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
 
 static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
 {
+	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
 	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
-	const struct aspeed_sgpio_bank *bank = to_bank(offset);
 	void __iomem *addr_r, *addr_w;
 	u32 reg = 0;
 
-	if (aspeed_sgpio_is_input(offset))
+	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
 		return -EINVAL;
 
 	/* Since this is an output, read the cached value from rdata, then
@@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
 	reg = ioread32(addr_r);
 
 	if (val)
-		reg |= GPIO_BIT(offset);
+		reg |= GPIO_BIT(offset % gpio->max_ngpios);
 	else
-		reg &= ~GPIO_BIT(offset);
+		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
 
 	iowrite32(reg, addr_w);
 
@@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, unsigned int offset, int val)
 
 static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
 {
-	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
+	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
+
+	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
 }
 
 static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int offset, int val)
@@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int offset, int v
 
 static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned int offset)
 {
-	return !!aspeed_sgpio_is_input(offset);
+	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
+
+	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
 }
 
 static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
@@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
 	WARN_ON(!internal);
 
 	*gpio = internal;
-	*bank = to_bank(*offset);
-	*bit = GPIO_BIT(*offset);
+	*bank = to_bank(*offset, internal->max_ngpios);
+	*bit = GPIO_BIT(*offset % internal->max_ngpios);
 }
 
 static void aspeed_sgpio_irq_ack(struct irq_data *d)
@@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
 	chained_irq_exit(ic, desc);
 }
 
-static struct irq_chip aspeed_sgpio_irqchip = {
-	.name       = "aspeed-sgpio",
-	.irq_ack    = aspeed_sgpio_irq_ack,
-	.irq_mask   = aspeed_sgpio_irq_mask,
-	.irq_unmask = aspeed_sgpio_irq_unmask,
-	.irq_set_type   = aspeed_sgpio_set_type,
-};
-
 static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 				   struct platform_device *pdev)
 {
@@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
 	}
 
+	gpio->intc.name = dev_name(&pdev->dev);
+	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
+	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
+	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
+	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
+
 	irq = &gpio->chip.irq;
-	irq->chip = &aspeed_sgpio_irqchip;
+	irq->chip = &gpio->intc;
 	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
 	irq->handler = handle_bad_irq;
 	irq->default_type = IRQ_TYPE_NONE;
@@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 	return 0;
 }
 
+static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
+					unsigned int offset, bool enable)
+{
+	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
+	unsigned long flags;
+	void __iomem *reg;
+	u32 val;
+
+	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
+
+	spin_lock_irqsave(&gpio->lock, flags);
+
+	val = readl(reg);
+
+	if (enable)
+		val |= GPIO_BIT(offset % gpio->max_ngpios);
+	else
+		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
+
+	writel(val, reg);
+
+	spin_unlock_irqrestore(&gpio->lock, flags);
+
+	return 0;
+}
+
+static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
+				   unsigned long config)
+{
+	unsigned long param = pinconf_to_config_param(config);
+	u32 arg = pinconf_to_config_argument(config);
+
+	if (param == PIN_CONFIG_PERSIST_STATE)
+		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
+	else
+		return -EOPNOTSUPP;
+}
+
 static const struct of_device_id aspeed_sgpio_of_table[] = {
-	{ .compatible = "aspeed,ast2400-sgpio" },
-	{ .compatible = "aspeed,ast2500-sgpio" },
+	{ .compatible = "aspeed,ast2400-sgpiom" },
+	{ .compatible = "aspeed,ast2500-sgpiom" },
+	{ .compatible = "aspeed,ast2600-sgpiom" },
 	{}
 };
 
@@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
 
 static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 {
+	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
 	struct aspeed_sgpio *gpio;
-	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
 	int rc;
 	unsigned long apb_freq;
 
@@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	if (IS_ERR(gpio->base))
 		return PTR_ERR(gpio->base);
 
+	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", &max_ngpios);
+	if (rc < 0) {
+		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
+		return -EINVAL;
+	}
+	gpio->max_ngpios = max_ngpios;
+
 	rc = of_property_read_u32(pdev->dev.of_node, "ngpios", &nr_gpios);
 	if (rc < 0) {
 		dev_err(&pdev->dev, "Could not read ngpios property\n");
 		return -EINVAL;
-	} else if (nr_gpios > MAX_NR_HW_SGPIO) {
+	} else if (nr_gpios % 8) {
+		dev_err(&pdev->dev, "Number of GPIOs not multiple of 8: %d\n",
+			nr_gpios);
+		return -EINVAL;
+	} else if (nr_gpios > gpio->max_ngpios) {
 		dev_err(&pdev->dev, "Number of GPIOs exceeds the maximum of %d: %d\n",
-			MAX_NR_HW_SGPIO, nr_gpios);
+			gpio->max_ngpios, nr_gpios);
 		return -EINVAL;
 	}
 	gpio->n_sgpio = nr_gpios;
@@ -539,7 +596,7 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	spin_lock_init(&gpio->lock);
 
 	gpio->chip.parent = &pdev->dev;
-	gpio->chip.ngpio = MAX_NR_HW_SGPIO * 2;
+	gpio->chip.ngpio = gpio->max_ngpios * 2;
 	gpio->chip.init_valid_mask = aspeed_sgpio_init_valid_mask;
 	gpio->chip.direction_input = aspeed_sgpio_dir_in;
 	gpio->chip.direction_output = aspeed_sgpio_dir_out;
@@ -548,7 +605,7 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	gpio->chip.free = NULL;
 	gpio->chip.get = aspeed_sgpio_get;
 	gpio->chip.set = aspeed_sgpio_set;
-	gpio->chip.set_config = NULL;
+	gpio->chip.set_config = aspeed_sgpio_set_config;
 	gpio->chip.label = dev_name(&pdev->dev);
 	gpio->chip.base = -1;
 
-- 
2.17.1


^ permalink raw reply related	[flat|nested] 38+ messages in thread

* [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-26  9:46   ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-26  9:46 UTC (permalink / raw)
  To: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: steven_lee, Hongweiz, ryan_chen, billy_tsai

AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
supports up to 80 pins.
In the current driver design, the max number of sgpio pins is hardcoded
in macro MAX_NR_HW_SGPIO and the value is 80.
The patch makes the maximum gpio number *constraint of chip* comes from
the dts. The property name is max-ngpios.

Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
---
 drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
 1 file changed, 125 insertions(+), 68 deletions(-)

diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
index 64e54f8c30d2..f74bd7988b66 100644
--- a/drivers/gpio/gpio-aspeed-sgpio.c
+++ b/drivers/gpio/gpio-aspeed-sgpio.c
@@ -17,37 +17,28 @@
 #include <linux/spinlock.h>
 #include <linux/string.h>
 
-/*
- * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
- * slots within the clocked serial GPIO data). Since each HW GPIO is both an
- * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
- * device.
- *
- * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
- * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
- */
-#define MAX_NR_HW_SGPIO			80
-#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO
-
 #define ASPEED_SGPIO_CTRL		0x54
 
-#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
+#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)
 #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
 #define ASPEED_SGPIO_ENABLE		BIT(0)
 
 struct aspeed_sgpio {
 	struct gpio_chip chip;
+	struct irq_chip intc;
 	struct clk *pclk;
 	spinlock_t lock;
 	void __iomem *base;
 	int irq;
+	int max_ngpios;
 	int n_sgpio;
 };
 
 struct aspeed_sgpio_bank {
-	uint16_t    val_regs;
-	uint16_t    rdata_reg;
-	uint16_t    irq_regs;
+	u16    val_regs;
+	u16    rdata_reg;
+	u16    irq_regs;
+	u16    tolerance_regs;
 	const char  names[4][3];
 };
 
@@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank aspeed_sgpio_banks[] = {
 		.val_regs = 0x0000,
 		.rdata_reg = 0x0070,
 		.irq_regs = 0x0004,
+		.tolerance_regs = 0x0018,
 		.names = { "A", "B", "C", "D" },
 	},
 	{
 		.val_regs = 0x001C,
 		.rdata_reg = 0x0074,
 		.irq_regs = 0x0020,
+		.tolerance_regs = 0x0034,
 		.names = { "E", "F", "G", "H" },
 	},
 	{
 		.val_regs = 0x0038,
 		.rdata_reg = 0x0078,
 		.irq_regs = 0x003C,
-		.names = { "I", "J" },
+		.tolerance_regs = 0x0050,
+		.names = { "I", "J", "K", "L" },
+	},
+	{
+		.val_regs = 0x0090,
+		.rdata_reg = 0x007C,
+		.irq_regs = 0x0094,
+		.tolerance_regs = 0x00A8,
+		.names = { "M", "N", "O", "P" },
 	},
 };
 
@@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
 	reg_irq_type1,
 	reg_irq_type2,
 	reg_irq_status,
+	reg_tolerance,
 };
 
-#define GPIO_VAL_VALUE      0x00
-#define GPIO_IRQ_ENABLE     0x00
-#define GPIO_IRQ_TYPE0      0x04
-#define GPIO_IRQ_TYPE1      0x08
-#define GPIO_IRQ_TYPE2      0x0C
-#define GPIO_IRQ_STATUS     0x10
+#define GPIO_IRQ_OFFSET_ENABLE     0x00
+#define GPIO_IRQ_OFFSET_TYPE0      0x04
+#define GPIO_IRQ_OFFSET_TYPE1      0x08
+#define GPIO_IRQ_OFFSET_TYPE2      0x0C
+#define GPIO_IRQ_OFFSET_STATUS     0x10
 
 static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
 				     const struct aspeed_sgpio_bank *bank,
@@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
 {
 	switch (reg) {
 	case reg_val:
-		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
+		return gpio->base + bank->val_regs;
 	case reg_rdata:
 		return gpio->base + bank->rdata_reg;
 	case reg_irq_enable:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
 	case reg_irq_type0:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
 	case reg_irq_type1:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
 	case reg_irq_type2:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
 	case reg_irq_status:
-		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
+		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
+	case reg_tolerance:
+		return gpio->base + bank->tolerance_regs;
 	default:
 		/* acturally if code runs to here, it's an error case */
 		BUG();
 	}
 }
 
-#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
-#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
+#define GPIO_BANK(x)    ((x) >> 5)
+/* modulo 32 */
+#define GPIO_OFFSET(x)  ((x) & 0x1f)
 #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
 
-static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
+static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, unsigned int max_ngpios)
 {
 	unsigned int bank;
 
-	bank = GPIO_BANK(offset);
+	bank = GPIO_BANK(offset % max_ngpios);
 
 	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
 	return &aspeed_sgpio_banks[bank];
@@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct gpio_chip *gc,
 		unsigned long *valid_mask, unsigned int ngpios)
 {
 	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
+	int max_ngpios = sgpio->max_ngpios;
 	int n = sgpio->n_sgpio;
-	int c = SGPIO_OUTPUT_OFFSET - n;
+	int c = max_ngpios - n;
 
-	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
+	WARN_ON(ngpios < max_ngpios * 2);
 
 	/* input GPIOs in the lower range */
 	bitmap_set(valid_mask, 0, n);
 	bitmap_clear(valid_mask, n, c);
 
-	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
-	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
-	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
+	/* output GPIOS above max_ngpios */
+	bitmap_set(valid_mask, max_ngpios, n);
+	bitmap_clear(valid_mask, max_ngpios + n, c);
 
 	return 0;
 }
@@ -161,30 +166,30 @@ static void aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
 	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
 	int n = sgpio->n_sgpio;
 
-	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
+	WARN_ON(ngpios < sgpio->max_ngpios * 2);
 
 	/* input GPIOs in the lower range */
 	bitmap_set(valid_mask, 0, n);
 	bitmap_clear(valid_mask, n, ngpios - n);
 }
 
-static bool aspeed_sgpio_is_input(unsigned int offset)
+static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
 {
-	return offset < SGPIO_OUTPUT_OFFSET;
+	return offset < max_ngpios;
 }
 
 static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
 {
+	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
 	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
-	const struct aspeed_sgpio_bank *bank = to_bank(offset);
 	unsigned long flags;
 	enum aspeed_sgpio_reg reg;
 	int rc = 0;
 
 	spin_lock_irqsave(&gpio->lock, flags);
 
-	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
-	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
+	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
+	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
 
 	spin_unlock_irqrestore(&gpio->lock, flags);
 
@@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
 
 static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
 {
+	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
 	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
-	const struct aspeed_sgpio_bank *bank = to_bank(offset);
 	void __iomem *addr_r, *addr_w;
 	u32 reg = 0;
 
-	if (aspeed_sgpio_is_input(offset))
+	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
 		return -EINVAL;
 
 	/* Since this is an output, read the cached value from rdata, then
@@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
 	reg = ioread32(addr_r);
 
 	if (val)
-		reg |= GPIO_BIT(offset);
+		reg |= GPIO_BIT(offset % gpio->max_ngpios);
 	else
-		reg &= ~GPIO_BIT(offset);
+		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
 
 	iowrite32(reg, addr_w);
 
@@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, unsigned int offset, int val)
 
 static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
 {
-	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
+	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
+
+	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
 }
 
 static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int offset, int val)
@@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int offset, int v
 
 static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned int offset)
 {
-	return !!aspeed_sgpio_is_input(offset);
+	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
+
+	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
 }
 
 static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
@@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
 	WARN_ON(!internal);
 
 	*gpio = internal;
-	*bank = to_bank(*offset);
-	*bit = GPIO_BIT(*offset);
+	*bank = to_bank(*offset, internal->max_ngpios);
+	*bit = GPIO_BIT(*offset % internal->max_ngpios);
 }
 
 static void aspeed_sgpio_irq_ack(struct irq_data *d)
@@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
 	chained_irq_exit(ic, desc);
 }
 
-static struct irq_chip aspeed_sgpio_irqchip = {
-	.name       = "aspeed-sgpio",
-	.irq_ack    = aspeed_sgpio_irq_ack,
-	.irq_mask   = aspeed_sgpio_irq_mask,
-	.irq_unmask = aspeed_sgpio_irq_unmask,
-	.irq_set_type   = aspeed_sgpio_set_type,
-};
-
 static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 				   struct platform_device *pdev)
 {
@@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
 	}
 
+	gpio->intc.name = dev_name(&pdev->dev);
+	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
+	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
+	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
+	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
+
 	irq = &gpio->chip.irq;
-	irq->chip = &aspeed_sgpio_irqchip;
+	irq->chip = &gpio->intc;
 	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
 	irq->handler = handle_bad_irq;
 	irq->default_type = IRQ_TYPE_NONE;
@@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
 	return 0;
 }
 
+static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
+					unsigned int offset, bool enable)
+{
+	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
+	unsigned long flags;
+	void __iomem *reg;
+	u32 val;
+
+	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
+
+	spin_lock_irqsave(&gpio->lock, flags);
+
+	val = readl(reg);
+
+	if (enable)
+		val |= GPIO_BIT(offset % gpio->max_ngpios);
+	else
+		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
+
+	writel(val, reg);
+
+	spin_unlock_irqrestore(&gpio->lock, flags);
+
+	return 0;
+}
+
+static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
+				   unsigned long config)
+{
+	unsigned long param = pinconf_to_config_param(config);
+	u32 arg = pinconf_to_config_argument(config);
+
+	if (param == PIN_CONFIG_PERSIST_STATE)
+		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
+	else
+		return -EOPNOTSUPP;
+}
+
 static const struct of_device_id aspeed_sgpio_of_table[] = {
-	{ .compatible = "aspeed,ast2400-sgpio" },
-	{ .compatible = "aspeed,ast2500-sgpio" },
+	{ .compatible = "aspeed,ast2400-sgpiom" },
+	{ .compatible = "aspeed,ast2500-sgpiom" },
+	{ .compatible = "aspeed,ast2600-sgpiom" },
 	{}
 };
 
@@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
 
 static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 {
+	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
 	struct aspeed_sgpio *gpio;
-	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
 	int rc;
 	unsigned long apb_freq;
 
@@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	if (IS_ERR(gpio->base))
 		return PTR_ERR(gpio->base);
 
+	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", &max_ngpios);
+	if (rc < 0) {
+		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
+		return -EINVAL;
+	}
+	gpio->max_ngpios = max_ngpios;
+
 	rc = of_property_read_u32(pdev->dev.of_node, "ngpios", &nr_gpios);
 	if (rc < 0) {
 		dev_err(&pdev->dev, "Could not read ngpios property\n");
 		return -EINVAL;
-	} else if (nr_gpios > MAX_NR_HW_SGPIO) {
+	} else if (nr_gpios % 8) {
+		dev_err(&pdev->dev, "Number of GPIOs not multiple of 8: %d\n",
+			nr_gpios);
+		return -EINVAL;
+	} else if (nr_gpios > gpio->max_ngpios) {
 		dev_err(&pdev->dev, "Number of GPIOs exceeds the maximum of %d: %d\n",
-			MAX_NR_HW_SGPIO, nr_gpios);
+			gpio->max_ngpios, nr_gpios);
 		return -EINVAL;
 	}
 	gpio->n_sgpio = nr_gpios;
@@ -539,7 +596,7 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	spin_lock_init(&gpio->lock);
 
 	gpio->chip.parent = &pdev->dev;
-	gpio->chip.ngpio = MAX_NR_HW_SGPIO * 2;
+	gpio->chip.ngpio = gpio->max_ngpios * 2;
 	gpio->chip.init_valid_mask = aspeed_sgpio_init_valid_mask;
 	gpio->chip.direction_input = aspeed_sgpio_dir_in;
 	gpio->chip.direction_output = aspeed_sgpio_dir_out;
@@ -548,7 +605,7 @@ static int __init aspeed_sgpio_probe(struct platform_device *pdev)
 	gpio->chip.free = NULL;
 	gpio->chip.get = aspeed_sgpio_get;
 	gpio->chip.set = aspeed_sgpio_set;
-	gpio->chip.set_config = NULL;
+	gpio->chip.set_config = aspeed_sgpio_set_config;
 	gpio->chip.label = dev_name(&pdev->dev);
 	gpio->chip.base = -1;
 
-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply related	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-26  9:46   ` Steven Lee
@ 2021-05-26 12:56     ` Rob Herring
  -1 siblings, 0 replies; 38+ messages in thread
From: Rob Herring @ 2021-05-26 12:56 UTC (permalink / raw)
  To: Steven Lee
  Cc: Joel Stanley,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	open list, billy_tsai, moderated list:ARM/ASPEED MACHINE SUPPORT,
	Hongweiz, Bartosz Golaszewski, open list:GPIO SUBSYSTEM,
	Rob Herring, ryan_chen,
	moderated list:ARM/ASPEED MACHINE SUPPORT, Andrew Jeffery,
	Linus Walleij

On Wed, 26 May 2021 17:46:05 +0800, Steven Lee wrote:
> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more flexible.
> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
>  .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
>  2 files changed, 91 insertions(+), 46 deletions(-)
>  create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
>  delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
> 

My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):

yamllint warnings/errors:

dtschema/dtc warnings/errors:
./Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml: $id: relative path/filename doesn't match actual path or filename
	expected: http://devicetree.org/schemas/gpio/aspeed,sgpio.yaml#

See https://patchwork.ozlabs.org/patch/1483966

This check can fail if there are any dependencies. The base for a patch
series is generally the most recent rc1.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit.


^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-26 12:56     ` Rob Herring
  0 siblings, 0 replies; 38+ messages in thread
From: Rob Herring @ 2021-05-26 12:56 UTC (permalink / raw)
  To: Steven Lee
  Cc: Joel Stanley,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	open list, billy_tsai, moderated list:ARM/ASPEED MACHINE SUPPORT,
	Hongweiz, Bartosz Golaszewski, open list:GPIO SUBSYSTEM,
	Rob Herring, ryan_chen,
	moderated list:ARM/ASPEED MACHINE SUPPORT, Andrew Jeffery,
	Linus Walleij

On Wed, 26 May 2021 17:46:05 +0800, Steven Lee wrote:
> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more flexible.
> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
>  .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
>  2 files changed, 91 insertions(+), 46 deletions(-)
>  create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
>  delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
> 

My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
on your patch (DT_CHECKER_FLAGS is new in v5.13):

yamllint warnings/errors:

dtschema/dtc warnings/errors:
./Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml: $id: relative path/filename doesn't match actual path or filename
	expected: http://devicetree.org/schemas/gpio/aspeed,sgpio.yaml#

See https://patchwork.ozlabs.org/patch/1483966

This check can fail if there are any dependencies. The base for a patch
series is generally the most recent rc1.

If you already ran 'make dt_binding_check' and didn't see the above
error(s), then make sure 'yamllint' is installed and dt-schema is up to
date:

pip3 install dtschema --upgrade

Please check and re-submit.


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-26  9:46   ` Steven Lee
  (?)
@ 2021-05-26 14:04     ` kernel test robot
  -1 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:04 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: kbuild-all, steven_lee

[-- Attachment #1: Type: text/plain, Size: 3052 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: arc-randconfig-r023-20210526 (attached as .config)
compiler: arc-elf-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   drivers/gpio/gpio-aspeed-sgpio.c: In function 'aspeed_sgpio_get':
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: 'gpio' undeclared (first use in this function)
     183 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~
   drivers/gpio/gpio-aspeed-sgpio.c:183:57: note: each undeclared identifier is reported only once for each function it appears in
   drivers/gpio/gpio-aspeed-sgpio.c: In function 'sgpio_set_value':
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: 'gpio' undeclared (first use in this function)
     201 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~


vim +/gpio +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 33657 bytes --]

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-26 14:04     ` kernel test robot
  0 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:04 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: kbuild-all, steven_lee

[-- Attachment #1: Type: text/plain, Size: 3052 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: arc-randconfig-r023-20210526 (attached as .config)
compiler: arc-elf-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   drivers/gpio/gpio-aspeed-sgpio.c: In function 'aspeed_sgpio_get':
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: 'gpio' undeclared (first use in this function)
     183 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~
   drivers/gpio/gpio-aspeed-sgpio.c:183:57: note: each undeclared identifier is reported only once for each function it appears in
   drivers/gpio/gpio-aspeed-sgpio.c: In function 'sgpio_set_value':
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: 'gpio' undeclared (first use in this function)
     201 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~


vim +/gpio +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 33657 bytes --]

[-- Attachment #3: Type: text/plain, Size: 176 bytes --]

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-26 14:04     ` kernel test robot
  0 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:04 UTC (permalink / raw)
  To: kbuild-all

[-- Attachment #1: Type: text/plain, Size: 3119 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: arc-randconfig-r023-20210526 (attached as .config)
compiler: arc-elf-gcc (GCC) 9.3.0
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=gcc-9.3.0 make.cross ARCH=arc 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

   drivers/gpio/gpio-aspeed-sgpio.c: In function 'aspeed_sgpio_get':
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: 'gpio' undeclared (first use in this function)
     183 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~
   drivers/gpio/gpio-aspeed-sgpio.c:183:57: note: each undeclared identifier is reported only once for each function it appears in
   drivers/gpio/gpio-aspeed-sgpio.c: In function 'sgpio_set_value':
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: 'gpio' undeclared (first use in this function)
     201 |  const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
         |                                                         ^~~~


vim +/gpio +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

[-- Attachment #2: config.gz --]
[-- Type: application/gzip, Size: 33657 bytes --]

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-26  9:46   ` Steven Lee
  (?)
@ 2021-05-26 14:05     ` kernel test robot
  -1 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:05 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: kbuild-all, clang-built-linux, steven_lee

[-- Attachment #1: Type: text/plain, Size: 3307 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: powerpc64-randconfig-r021-20210526 (attached as .config)
compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 99155e913e9bad5f7f8a247f8bb3a3ff3da74af1)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install powerpc64 cross compiling tool for clang build
        # apt-get install binutils-powerpc64-linux-gnu
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: expected expression
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: expected expression
   4 errors generated.


vim +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 44443 bytes --]

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-26 14:05     ` kernel test robot
  0 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:05 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: kbuild-all, clang-built-linux, steven_lee

[-- Attachment #1: Type: text/plain, Size: 3307 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: powerpc64-randconfig-r021-20210526 (attached as .config)
compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 99155e913e9bad5f7f8a247f8bb3a3ff3da74af1)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install powerpc64 cross compiling tool for clang build
        # apt-get install binutils-powerpc64-linux-gnu
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: expected expression
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: expected expression
   4 errors generated.


vim +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all@lists.01.org

[-- Attachment #2: .config.gz --]
[-- Type: application/gzip, Size: 44443 bytes --]

[-- Attachment #3: Type: text/plain, Size: 176 bytes --]

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-26 14:05     ` kernel test robot
  0 siblings, 0 replies; 38+ messages in thread
From: kernel test robot @ 2021-05-26 14:05 UTC (permalink / raw)
  To: kbuild-all

[-- Attachment #1: Type: text/plain, Size: 3378 bytes --]

Hi Steven,

Thank you for the patch! Yet something to improve:

[auto build test ERROR on joel-aspeed/for-next]
[also build test ERROR on robh/for-next gpio/for-next v5.13-rc3 next-20210526]
[If your patch is applied to the wrong git tree, kindly drop us a note.
And when submitting patch, we suggest to use '--base' as documented in
https://git-scm.com/docs/git-format-patch]

url:    https://github.com/0day-ci/linux/commits/Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
base:   https://git.kernel.org/pub/scm/linux/kernel/git/joel/aspeed.git for-next
config: powerpc64-randconfig-r021-20210526 (attached as .config)
compiler: clang version 13.0.0 (https://github.com/llvm/llvm-project 99155e913e9bad5f7f8a247f8bb3a3ff3da74af1)
reproduce (this is a W=1 build):
        wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
        chmod +x ~/bin/make.cross
        # install powerpc64 cross compiling tool for clang build
        # apt-get install binutils-powerpc64-linux-gnu
        # https://github.com/0day-ci/linux/commit/72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        git remote add linux-review https://github.com/0day-ci/linux
        git fetch --no-tags linux-review Steven-Lee/ASPEED-sgpio-driver-enhancement/20210526-174912
        git checkout 72b31f2c47f6ab3ce9a7115dcc0dc01331bf45c0
        # save the attached .config to linux build tree
        COMPILER_INSTALL_PATH=$HOME/0day COMPILER=clang make.cross ARCH=powerpc64 

If you fix the issue, kindly add following tag as appropriate
Reported-by: kernel test robot <lkp@intel.com>

All errors (new ones prefixed by >>):

>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
>> drivers/gpio/gpio-aspeed-sgpio.c:183:57: error: expected expression
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: use of undeclared identifier 'gpio'; did you mean 'goto'?
           const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
                                                                  ^~~~
                                                                  goto
   drivers/gpio/gpio-aspeed-sgpio.c:201:57: error: expected expression
   4 errors generated.


vim +183 drivers/gpio/gpio-aspeed-sgpio.c

   180	
   181	static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
   182	{
 > 183		const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
   184		struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
   185		unsigned long flags;
   186		enum aspeed_sgpio_reg reg;
   187		int rc = 0;
   188	
   189		spin_lock_irqsave(&gpio->lock, flags);
   190	
   191		reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : reg_rdata;
   192		rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % gpio->max_ngpios));
   193	
   194		spin_unlock_irqrestore(&gpio->lock, flags);
   195	
   196		return rc;
   197	}
   198	

---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/kbuild-all(a)lists.01.org

[-- Attachment #2: config.gz --]
[-- Type: application/gzip, Size: 44443 bytes --]

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-26  9:46   ` Steven Lee
@ 2021-05-27  0:57     ` Andrew Jeffery
  -1 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  0:57 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more flexible.

Please add the new property in a separate patch to the YAML conversion. 
Having changes to the properties in addition to switching to dt schema 
makes it harder to catch errors, and it seems Rob's bot has already 
picked up some.

Reviewing the changes as separate patches means its easier to give you 
a Reviewed-by tag on that patches I'm happy with.

Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-27  0:57     ` Andrew Jeffery
  0 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  0:57 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more flexible.

Please add the new property in a separate patch to the YAML conversion. 
Having changes to the properties in addition to switching to dt schema 
makes it harder to catch errors, and it seems Rob's bot has already 
picked up some.

Reviewing the changes as separate patches means its easier to give you 
a Reviewed-by tag on that patches I'm happy with.

Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-27  0:57     ` Andrew Jeffery
@ 2021-05-27  1:03       ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  1:03 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai

The 05/27/2021 08:57, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more flexible.
> 
> Please add the new property in a separate patch to the YAML conversion. 
> Having changes to the properties in addition to switching to dt schema 
> makes it harder to catch errors, and it seems Rob's bot has already 
> picked up some.
> 
> Reviewing the changes as separate patches means its easier to give you 
> a Reviewed-by tag on that patches I'm happy with.
> 

Thanks for your suggestion, I will add the new property in a
separate patch in v3 patch series.

> Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-27  1:03       ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  1:03 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai

The 05/27/2021 08:57, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more flexible.
> 
> Please add the new property in a separate patch to the YAML conversion. 
> Having changes to the properties in addition to switching to dt schema 
> makes it harder to catch errors, and it seems Rob's bot has already 
> picked up some.
> 
> Reviewing the changes as separate patches means its easier to give you 
> a Reviewed-by tag on that patches I'm happy with.
> 

Thanks for your suggestion, I will add the new property in a
separate patch in v3 patch series.

> Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-26  9:46   ` Steven Lee
@ 2021-05-27  1:26     ` Andrew Jeffery
  -1 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  1:26 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
> with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
> supports up to 80 pins.
> In the current driver design, the max number of sgpio pins is hardcoded
> in macro MAX_NR_HW_SGPIO and the value is 80.
> The patch makes the maximum gpio number *constraint of chip* comes from
> the dts. The property name is max-ngpios.
> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
>  1 file changed, 125 insertions(+), 68 deletions(-)
> 
> diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
> index 64e54f8c30d2..f74bd7988b66 100644
> --- a/drivers/gpio/gpio-aspeed-sgpio.c
> +++ b/drivers/gpio/gpio-aspeed-sgpio.c
> @@ -17,37 +17,28 @@
>  #include <linux/spinlock.h>
>  #include <linux/string.h>
>  
> -/*
> - * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
> - * slots within the clocked serial GPIO data). Since each HW GPIO is both an
> - * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
> - * device.
> - *
> - * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
> - * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
> - */
> -#define MAX_NR_HW_SGPIO			80
> -#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO

I expect you want to keep these as fall-back values for old devicetrees 
that don't have your new properties.

> -
>  #define ASPEED_SGPIO_CTRL		0x54
>  
> -#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
> +#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)

Is this valid on the 2400 and 2500? This might be better defined in 
platform data so we don't go writing reserved bits on older hardware?

>  #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
>  #define ASPEED_SGPIO_ENABLE		BIT(0)
>  
>  struct aspeed_sgpio {
>  	struct gpio_chip chip;
> +	struct irq_chip intc;

My gut instinct is this could be in a separate patch.

>  	struct clk *pclk;
>  	spinlock_t lock;
>  	void __iomem *base;
>  	int irq;
> +	int max_ngpios;
>  	int n_sgpio;
>  };
>  
>  struct aspeed_sgpio_bank {
> -	uint16_t    val_regs;
> -	uint16_t    rdata_reg;
> -	uint16_t    irq_regs;
> +	u16    val_regs;
> +	u16    rdata_reg;
> +	u16    irq_regs;
> +	u16    tolerance_regs;

I suggest splitting this out into a cleanup patch.

>  	const char  names[4][3];
>  };
>  
> @@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank 
> aspeed_sgpio_banks[] = {
>  		.val_regs = 0x0000,
>  		.rdata_reg = 0x0070,
>  		.irq_regs = 0x0004,
> +		.tolerance_regs = 0x0018,

Reset tolerance is supported on the 2500. Your patch subject claims its 
for supporting the 2600, so I think the addition of reset tolerance 
capabilities should be a separate patch.

By splitting out the different patches as requested, I think this patch 
can just become about supporting max_ngpios, which will help focus the 
discussion.

>  		.names = { "A", "B", "C", "D" },
>  	},
>  	{
>  		.val_regs = 0x001C,
>  		.rdata_reg = 0x0074,
>  		.irq_regs = 0x0020,
> +		.tolerance_regs = 0x0034,
>  		.names = { "E", "F", "G", "H" },
>  	},
>  	{
>  		.val_regs = 0x0038,
>  		.rdata_reg = 0x0078,
>  		.irq_regs = 0x003C,
> -		.names = { "I", "J" },
> +		.tolerance_regs = 0x0050,
> +		.names = { "I", "J", "K", "L" },
> +	},
> +	{
> +		.val_regs = 0x0090,
> +		.rdata_reg = 0x007C,
> +		.irq_regs = 0x0094,
> +		.tolerance_regs = 0x00A8,
> +		.names = { "M", "N", "O", "P" },
>  	},
>  };
>  
> @@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
>  	reg_irq_type1,
>  	reg_irq_type2,
>  	reg_irq_status,
> +	reg_tolerance,
>  };
>  
> -#define GPIO_VAL_VALUE      0x00
> -#define GPIO_IRQ_ENABLE     0x00
> -#define GPIO_IRQ_TYPE0      0x04
> -#define GPIO_IRQ_TYPE1      0x08
> -#define GPIO_IRQ_TYPE2      0x0C
> -#define GPIO_IRQ_STATUS     0x10
> +#define GPIO_IRQ_OFFSET_ENABLE     0x00
> +#define GPIO_IRQ_OFFSET_TYPE0      0x04
> +#define GPIO_IRQ_OFFSET_TYPE1      0x08
> +#define GPIO_IRQ_OFFSET_TYPE2      0x0C
> +#define GPIO_IRQ_OFFSET_STATUS     0x10

I don't think this change is necessary. It looks like it produces a 
bunch of noise in the diff.

>  
>  static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
>  				     const struct aspeed_sgpio_bank *bank,
> @@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
>  {
>  	switch (reg) {
>  	case reg_val:
> -		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
> +		return gpio->base + bank->val_regs;
>  	case reg_rdata:
>  		return gpio->base + bank->rdata_reg;
>  	case reg_irq_enable:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
>  	case reg_irq_type0:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
>  	case reg_irq_type1:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
>  	case reg_irq_type2:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
>  	case reg_irq_status:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
> +	case reg_tolerance:
> +		return gpio->base + bank->tolerance_regs;
>  	default:
>  		/* acturally if code runs to here, it's an error case */
>  		BUG();
>  	}
>  }
>  
> -#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
> -#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
> +#define GPIO_BANK(x)    ((x) >> 5)
> +/* modulo 32 */
> +#define GPIO_OFFSET(x)  ((x) & 0x1f)
>  #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
>  
> -static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
> +static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, 
> unsigned int max_ngpios)
>  {
>  	unsigned int bank;
>  
> -	bank = GPIO_BANK(offset);
> +	bank = GPIO_BANK(offset % max_ngpios);
>  
>  	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
>  	return &aspeed_sgpio_banks[bank];
> @@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct 
> gpio_chip *gc,
>  		unsigned long *valid_mask, unsigned int ngpios)
>  {
>  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> +	int max_ngpios = sgpio->max_ngpios;
>  	int n = sgpio->n_sgpio;
> -	int c = SGPIO_OUTPUT_OFFSET - n;
> +	int c = max_ngpios - n;
>  
> -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> +	WARN_ON(ngpios < max_ngpios * 2);
>  
>  	/* input GPIOs in the lower range */
>  	bitmap_set(valid_mask, 0, n);
>  	bitmap_clear(valid_mask, n, c);
>  
> -	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
> -	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
> -	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
> +	/* output GPIOS above max_ngpios */
> +	bitmap_set(valid_mask, max_ngpios, n);
> +	bitmap_clear(valid_mask, max_ngpios + n, c);
>  
>  	return 0;
>  }
> @@ -161,30 +166,30 @@ static void 
> aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
>  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
>  	int n = sgpio->n_sgpio;
>  
> -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> +	WARN_ON(ngpios < sgpio->max_ngpios * 2);
>  
>  	/* input GPIOs in the lower range */
>  	bitmap_set(valid_mask, 0, n);
>  	bitmap_clear(valid_mask, n, ngpios - n);
>  }
>  
> -static bool aspeed_sgpio_is_input(unsigned int offset)
> +static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
>  {
> -	return offset < SGPIO_OUTPUT_OFFSET;
> +	return offset < max_ngpios;
>  }
>  
>  static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
>  {
> +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);

This references gpio before it's defined, as the build bot noted.

>  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
>  	unsigned long flags;
>  	enum aspeed_sgpio_reg reg;
>  	int rc = 0;
>  
>  	spin_lock_irqsave(&gpio->lock, flags);
>  
> -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> reg_rdata;

We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
'const struct aspeed_sgpio *' parameter), rather than open-coding 
gpio->max_ngpios. This approach will make it easier to refactor the 
implementation in the future (if necessary).

> +	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % 
> gpio->max_ngpios));
>  
>  	spin_unlock_irqrestore(&gpio->lock, flags);
>  
> @@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, 
> unsigned int offset)
>  
>  static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
>  {
> +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);

References gpio before it's defined.

>  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
>  	void __iomem *addr_r, *addr_w;
>  	u32 reg = 0;
>  
> -	if (aspeed_sgpio_is_input(offset))
> +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
>  		return -EINVAL;
>  
>  	/* Since this is an output, read the cached value from rdata, then
> @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> unsigned int offset, int val)
>  	reg = ioread32(addr_r);
>  
>  	if (val)
> -		reg |= GPIO_BIT(offset);
> +		reg |= GPIO_BIT(offset % gpio->max_ngpios);

Pass gpio through GPIO_BIT() too.

>  	else
> -		reg &= ~GPIO_BIT(offset);
> +		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
>  
>  	iowrite32(reg, addr_w);
>  
> @@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, 
> unsigned int offset, int val)
>  
>  static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
>  {
> -	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> +
> +	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
>  }
>  
>  static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int 
> offset, int val)
> @@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip 
> *gc, unsigned int offset, int v
>  
>  static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned 
> int offset)
>  {
> -	return !!aspeed_sgpio_is_input(offset);
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> +
> +	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
>  }
>  
>  static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> @@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
>  	WARN_ON(!internal);
>  
>  	*gpio = internal;
> -	*bank = to_bank(*offset);
> -	*bit = GPIO_BIT(*offset);
> +	*bank = to_bank(*offset, internal->max_ngpios);

Make to_bank take a 'const struct aspeed_sgpio *'.

> +	*bit = GPIO_BIT(*offset % internal->max_ngpios);
>  }
>  
>  static void aspeed_sgpio_irq_ack(struct irq_data *d)
> @@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
>  	chained_irq_exit(ic, desc);
>  }
>  
> -static struct irq_chip aspeed_sgpio_irqchip = {
> -	.name       = "aspeed-sgpio",
> -	.irq_ack    = aspeed_sgpio_irq_ack,
> -	.irq_mask   = aspeed_sgpio_irq_mask,
> -	.irq_unmask = aspeed_sgpio_irq_unmask,
> -	.irq_set_type   = aspeed_sgpio_set_type,
> -};
> -
>  static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
>  				   struct platform_device *pdev)
>  {
> @@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct 
> aspeed_sgpio *gpio,
>  		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
>  	}
>  
> +	gpio->intc.name = dev_name(&pdev->dev);
> +	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
> +	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
> +	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
> +	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
> +
>  	irq = &gpio->chip.irq;
> -	irq->chip = &aspeed_sgpio_irqchip;
> +	irq->chip = &gpio->intc;

As I said earlier, I think it would be best to split the IRQ changes 
out into a separate patch ordered before this one.

>  	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
>  	irq->handler = handle_bad_irq;
>  	irq->default_type = IRQ_TYPE_NONE;
> @@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct 
> aspeed_sgpio *gpio,
>  	return 0;
>  }
>  
> +static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
> +					unsigned int offset, bool enable)
> +{
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
> +	unsigned long flags;
> +	void __iomem *reg;
> +	u32 val;
> +
> +	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
> +
> +	spin_lock_irqsave(&gpio->lock, flags);
> +
> +	val = readl(reg);
> +
> +	if (enable)
> +		val |= GPIO_BIT(offset % gpio->max_ngpios);
> +	else
> +		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
> +
> +	writel(val, reg);
> +
> +	spin_unlock_irqrestore(&gpio->lock, flags);
> +
> +	return 0;
> +}
> +
> +static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
> +				   unsigned long config)
> +{
> +	unsigned long param = pinconf_to_config_param(config);
> +	u32 arg = pinconf_to_config_argument(config);
> +
> +	if (param == PIN_CONFIG_PERSIST_STATE)
> +		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
> +	else
> +		return -EOPNOTSUPP;
> +}
> +

Again, lets split reset tolerance out to a separate patch as well.

>  static const struct of_device_id aspeed_sgpio_of_table[] = {
> -	{ .compatible = "aspeed,ast2400-sgpio" },
> -	{ .compatible = "aspeed,ast2500-sgpio" },
> +	{ .compatible = "aspeed,ast2400-sgpiom" },
> +	{ .compatible = "aspeed,ast2500-sgpiom" },
> +	{ .compatible = "aspeed,ast2600-sgpiom" },

No. You'll need to support both while we transition between the two as 
the devicetree patch and the driver patch are separate, and this will 
break bisectability when tracking down runtime issues. Also, the 
devicetree patch should be ordered after the driver change for the same 
reason.

You should continue to support the 'aspeed,ast2{4,5}00-sgpio' style 
compatible here to handle old devicetrees.

So in summary, just add the 'aspeed,ast2{4,5,6}00-sgpiom' compatibles 
for now, don't remove the 'aspeed,ast2{4,5}00-sgpio' compatibles.

>  	{}
>  };
>  
> @@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
>  
>  static int __init aspeed_sgpio_probe(struct platform_device *pdev)
>  {
> +	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
>  	struct aspeed_sgpio *gpio;
> -	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
>  	int rc;
>  	unsigned long apb_freq;
>  
> @@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct 
> platform_device *pdev)
>  	if (IS_ERR(gpio->base))
>  		return PTR_ERR(gpio->base);
>  
> +	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", 
> &max_ngpios);
> +	if (rc < 0) {
> +		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
> +		return -EINVAL;

I don't think this is right. You should just fall back to using the 
hard-coded value.

Cheers,

Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-27  1:26     ` Andrew Jeffery
  0 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  1:26 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
> with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
> supports up to 80 pins.
> In the current driver design, the max number of sgpio pins is hardcoded
> in macro MAX_NR_HW_SGPIO and the value is 80.
> The patch makes the maximum gpio number *constraint of chip* comes from
> the dts. The property name is max-ngpios.
> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
>  1 file changed, 125 insertions(+), 68 deletions(-)
> 
> diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
> index 64e54f8c30d2..f74bd7988b66 100644
> --- a/drivers/gpio/gpio-aspeed-sgpio.c
> +++ b/drivers/gpio/gpio-aspeed-sgpio.c
> @@ -17,37 +17,28 @@
>  #include <linux/spinlock.h>
>  #include <linux/string.h>
>  
> -/*
> - * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
> - * slots within the clocked serial GPIO data). Since each HW GPIO is both an
> - * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
> - * device.
> - *
> - * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
> - * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
> - */
> -#define MAX_NR_HW_SGPIO			80
> -#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO

I expect you want to keep these as fall-back values for old devicetrees 
that don't have your new properties.

> -
>  #define ASPEED_SGPIO_CTRL		0x54
>  
> -#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
> +#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)

Is this valid on the 2400 and 2500? This might be better defined in 
platform data so we don't go writing reserved bits on older hardware?

>  #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
>  #define ASPEED_SGPIO_ENABLE		BIT(0)
>  
>  struct aspeed_sgpio {
>  	struct gpio_chip chip;
> +	struct irq_chip intc;

My gut instinct is this could be in a separate patch.

>  	struct clk *pclk;
>  	spinlock_t lock;
>  	void __iomem *base;
>  	int irq;
> +	int max_ngpios;
>  	int n_sgpio;
>  };
>  
>  struct aspeed_sgpio_bank {
> -	uint16_t    val_regs;
> -	uint16_t    rdata_reg;
> -	uint16_t    irq_regs;
> +	u16    val_regs;
> +	u16    rdata_reg;
> +	u16    irq_regs;
> +	u16    tolerance_regs;

I suggest splitting this out into a cleanup patch.

>  	const char  names[4][3];
>  };
>  
> @@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank 
> aspeed_sgpio_banks[] = {
>  		.val_regs = 0x0000,
>  		.rdata_reg = 0x0070,
>  		.irq_regs = 0x0004,
> +		.tolerance_regs = 0x0018,

Reset tolerance is supported on the 2500. Your patch subject claims its 
for supporting the 2600, so I think the addition of reset tolerance 
capabilities should be a separate patch.

By splitting out the different patches as requested, I think this patch 
can just become about supporting max_ngpios, which will help focus the 
discussion.

>  		.names = { "A", "B", "C", "D" },
>  	},
>  	{
>  		.val_regs = 0x001C,
>  		.rdata_reg = 0x0074,
>  		.irq_regs = 0x0020,
> +		.tolerance_regs = 0x0034,
>  		.names = { "E", "F", "G", "H" },
>  	},
>  	{
>  		.val_regs = 0x0038,
>  		.rdata_reg = 0x0078,
>  		.irq_regs = 0x003C,
> -		.names = { "I", "J" },
> +		.tolerance_regs = 0x0050,
> +		.names = { "I", "J", "K", "L" },
> +	},
> +	{
> +		.val_regs = 0x0090,
> +		.rdata_reg = 0x007C,
> +		.irq_regs = 0x0094,
> +		.tolerance_regs = 0x00A8,
> +		.names = { "M", "N", "O", "P" },
>  	},
>  };
>  
> @@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
>  	reg_irq_type1,
>  	reg_irq_type2,
>  	reg_irq_status,
> +	reg_tolerance,
>  };
>  
> -#define GPIO_VAL_VALUE      0x00
> -#define GPIO_IRQ_ENABLE     0x00
> -#define GPIO_IRQ_TYPE0      0x04
> -#define GPIO_IRQ_TYPE1      0x08
> -#define GPIO_IRQ_TYPE2      0x0C
> -#define GPIO_IRQ_STATUS     0x10
> +#define GPIO_IRQ_OFFSET_ENABLE     0x00
> +#define GPIO_IRQ_OFFSET_TYPE0      0x04
> +#define GPIO_IRQ_OFFSET_TYPE1      0x08
> +#define GPIO_IRQ_OFFSET_TYPE2      0x0C
> +#define GPIO_IRQ_OFFSET_STATUS     0x10

I don't think this change is necessary. It looks like it produces a 
bunch of noise in the diff.

>  
>  static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
>  				     const struct aspeed_sgpio_bank *bank,
> @@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
>  {
>  	switch (reg) {
>  	case reg_val:
> -		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
> +		return gpio->base + bank->val_regs;
>  	case reg_rdata:
>  		return gpio->base + bank->rdata_reg;
>  	case reg_irq_enable:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
>  	case reg_irq_type0:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
>  	case reg_irq_type1:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
>  	case reg_irq_type2:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
>  	case reg_irq_status:
> -		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
> +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
> +	case reg_tolerance:
> +		return gpio->base + bank->tolerance_regs;
>  	default:
>  		/* acturally if code runs to here, it's an error case */
>  		BUG();
>  	}
>  }
>  
> -#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
> -#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
> +#define GPIO_BANK(x)    ((x) >> 5)
> +/* modulo 32 */
> +#define GPIO_OFFSET(x)  ((x) & 0x1f)
>  #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
>  
> -static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
> +static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, 
> unsigned int max_ngpios)
>  {
>  	unsigned int bank;
>  
> -	bank = GPIO_BANK(offset);
> +	bank = GPIO_BANK(offset % max_ngpios);
>  
>  	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
>  	return &aspeed_sgpio_banks[bank];
> @@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct 
> gpio_chip *gc,
>  		unsigned long *valid_mask, unsigned int ngpios)
>  {
>  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> +	int max_ngpios = sgpio->max_ngpios;
>  	int n = sgpio->n_sgpio;
> -	int c = SGPIO_OUTPUT_OFFSET - n;
> +	int c = max_ngpios - n;
>  
> -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> +	WARN_ON(ngpios < max_ngpios * 2);
>  
>  	/* input GPIOs in the lower range */
>  	bitmap_set(valid_mask, 0, n);
>  	bitmap_clear(valid_mask, n, c);
>  
> -	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
> -	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
> -	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
> +	/* output GPIOS above max_ngpios */
> +	bitmap_set(valid_mask, max_ngpios, n);
> +	bitmap_clear(valid_mask, max_ngpios + n, c);
>  
>  	return 0;
>  }
> @@ -161,30 +166,30 @@ static void 
> aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
>  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
>  	int n = sgpio->n_sgpio;
>  
> -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> +	WARN_ON(ngpios < sgpio->max_ngpios * 2);
>  
>  	/* input GPIOs in the lower range */
>  	bitmap_set(valid_mask, 0, n);
>  	bitmap_clear(valid_mask, n, ngpios - n);
>  }
>  
> -static bool aspeed_sgpio_is_input(unsigned int offset)
> +static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
>  {
> -	return offset < SGPIO_OUTPUT_OFFSET;
> +	return offset < max_ngpios;
>  }
>  
>  static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
>  {
> +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);

This references gpio before it's defined, as the build bot noted.

>  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
>  	unsigned long flags;
>  	enum aspeed_sgpio_reg reg;
>  	int rc = 0;
>  
>  	spin_lock_irqsave(&gpio->lock, flags);
>  
> -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> reg_rdata;

We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
'const struct aspeed_sgpio *' parameter), rather than open-coding 
gpio->max_ngpios. This approach will make it easier to refactor the 
implementation in the future (if necessary).

> +	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % 
> gpio->max_ngpios));
>  
>  	spin_unlock_irqrestore(&gpio->lock, flags);
>  
> @@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, 
> unsigned int offset)
>  
>  static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
>  {
> +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);

References gpio before it's defined.

>  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
>  	void __iomem *addr_r, *addr_w;
>  	u32 reg = 0;
>  
> -	if (aspeed_sgpio_is_input(offset))
> +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
>  		return -EINVAL;
>  
>  	/* Since this is an output, read the cached value from rdata, then
> @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> unsigned int offset, int val)
>  	reg = ioread32(addr_r);
>  
>  	if (val)
> -		reg |= GPIO_BIT(offset);
> +		reg |= GPIO_BIT(offset % gpio->max_ngpios);

Pass gpio through GPIO_BIT() too.

>  	else
> -		reg &= ~GPIO_BIT(offset);
> +		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
>  
>  	iowrite32(reg, addr_w);
>  
> @@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, 
> unsigned int offset, int val)
>  
>  static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
>  {
> -	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> +
> +	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
>  }
>  
>  static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int 
> offset, int val)
> @@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip 
> *gc, unsigned int offset, int v
>  
>  static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned 
> int offset)
>  {
> -	return !!aspeed_sgpio_is_input(offset);
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> +
> +	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
>  }
>  
>  static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> @@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
>  	WARN_ON(!internal);
>  
>  	*gpio = internal;
> -	*bank = to_bank(*offset);
> -	*bit = GPIO_BIT(*offset);
> +	*bank = to_bank(*offset, internal->max_ngpios);

Make to_bank take a 'const struct aspeed_sgpio *'.

> +	*bit = GPIO_BIT(*offset % internal->max_ngpios);
>  }
>  
>  static void aspeed_sgpio_irq_ack(struct irq_data *d)
> @@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
>  	chained_irq_exit(ic, desc);
>  }
>  
> -static struct irq_chip aspeed_sgpio_irqchip = {
> -	.name       = "aspeed-sgpio",
> -	.irq_ack    = aspeed_sgpio_irq_ack,
> -	.irq_mask   = aspeed_sgpio_irq_mask,
> -	.irq_unmask = aspeed_sgpio_irq_unmask,
> -	.irq_set_type   = aspeed_sgpio_set_type,
> -};
> -
>  static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
>  				   struct platform_device *pdev)
>  {
> @@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct 
> aspeed_sgpio *gpio,
>  		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
>  	}
>  
> +	gpio->intc.name = dev_name(&pdev->dev);
> +	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
> +	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
> +	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
> +	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
> +
>  	irq = &gpio->chip.irq;
> -	irq->chip = &aspeed_sgpio_irqchip;
> +	irq->chip = &gpio->intc;

As I said earlier, I think it would be best to split the IRQ changes 
out into a separate patch ordered before this one.

>  	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
>  	irq->handler = handle_bad_irq;
>  	irq->default_type = IRQ_TYPE_NONE;
> @@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct 
> aspeed_sgpio *gpio,
>  	return 0;
>  }
>  
> +static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
> +					unsigned int offset, bool enable)
> +{
> +	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
> +	unsigned long flags;
> +	void __iomem *reg;
> +	u32 val;
> +
> +	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
> +
> +	spin_lock_irqsave(&gpio->lock, flags);
> +
> +	val = readl(reg);
> +
> +	if (enable)
> +		val |= GPIO_BIT(offset % gpio->max_ngpios);
> +	else
> +		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
> +
> +	writel(val, reg);
> +
> +	spin_unlock_irqrestore(&gpio->lock, flags);
> +
> +	return 0;
> +}
> +
> +static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
> +				   unsigned long config)
> +{
> +	unsigned long param = pinconf_to_config_param(config);
> +	u32 arg = pinconf_to_config_argument(config);
> +
> +	if (param == PIN_CONFIG_PERSIST_STATE)
> +		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
> +	else
> +		return -EOPNOTSUPP;
> +}
> +

Again, lets split reset tolerance out to a separate patch as well.

>  static const struct of_device_id aspeed_sgpio_of_table[] = {
> -	{ .compatible = "aspeed,ast2400-sgpio" },
> -	{ .compatible = "aspeed,ast2500-sgpio" },
> +	{ .compatible = "aspeed,ast2400-sgpiom" },
> +	{ .compatible = "aspeed,ast2500-sgpiom" },
> +	{ .compatible = "aspeed,ast2600-sgpiom" },

No. You'll need to support both while we transition between the two as 
the devicetree patch and the driver patch are separate, and this will 
break bisectability when tracking down runtime issues. Also, the 
devicetree patch should be ordered after the driver change for the same 
reason.

You should continue to support the 'aspeed,ast2{4,5}00-sgpio' style 
compatible here to handle old devicetrees.

So in summary, just add the 'aspeed,ast2{4,5,6}00-sgpiom' compatibles 
for now, don't remove the 'aspeed,ast2{4,5}00-sgpio' compatibles.

>  	{}
>  };
>  
> @@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
>  
>  static int __init aspeed_sgpio_probe(struct platform_device *pdev)
>  {
> +	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
>  	struct aspeed_sgpio *gpio;
> -	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
>  	int rc;
>  	unsigned long apb_freq;
>  
> @@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct 
> platform_device *pdev)
>  	if (IS_ERR(gpio->base))
>  		return PTR_ERR(gpio->base);
>  
> +	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", 
> &max_ngpios);
> +	if (rc < 0) {
> +		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
> +		return -EINVAL;

I don't think this is right. You should just fall back to using the 
hard-coded value.

Cheers,

Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
  2021-05-26  9:46   ` Steven Lee
@ 2021-05-27  1:27     ` Andrew Jeffery
  -1 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  1:27 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
> with 80 pins.

Is there any chance the serial GPIO controllers can be explicitly 
listed in the Memory Space Allocation Table of the datasheet? Currently 
they're covered by the entry for "GPIO Controller (Parallel GPIO)" 
which is listed as ranging from 0x1e780000-0x1e7807ff.

Admittedly the details are listed in chapter 41 for the GPIO 
Controller, but it would be handy to not have to dig.

> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
>  1 file changed, 32 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
> index f96607b7b4e2..556ce9535c22 100644
> --- a/arch/arm/boot/dts/aspeed-g6.dtsi
> +++ b/arch/arm/boot/dts/aspeed-g6.dtsi
> @@ -377,6 +377,38 @@
>  				#interrupt-cells = <2>;
>  			};
>  
> +			sgpiom0: sgpiom@1e780500 {
> +				#gpio-cells = <2>;
> +				gpio-controller;
> +				compatible = "aspeed,ast2600-sgpiom";
> +				reg = <0x1e780500 0x100>;
> +				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
> +				max-ngpios = <128>;

I need to think more about this one.

Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
@ 2021-05-27  1:27     ` Andrew Jeffery
  0 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  1:27 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongwei Zhang, Ryan Chen, Billy Tsai

Hi Steven,

On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
> with 80 pins.

Is there any chance the serial GPIO controllers can be explicitly 
listed in the Memory Space Allocation Table of the datasheet? Currently 
they're covered by the entry for "GPIO Controller (Parallel GPIO)" 
which is listed as ranging from 0x1e780000-0x1e7807ff.

Admittedly the details are listed in chapter 41 for the GPIO 
Controller, but it would be handy to not have to dig.

> 
> Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> ---
>  arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
>  1 file changed, 32 insertions(+)
> 
> diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
> index f96607b7b4e2..556ce9535c22 100644
> --- a/arch/arm/boot/dts/aspeed-g6.dtsi
> +++ b/arch/arm/boot/dts/aspeed-g6.dtsi
> @@ -377,6 +377,38 @@
>  				#interrupt-cells = <2>;
>  			};
>  
> +			sgpiom0: sgpiom@1e780500 {
> +				#gpio-cells = <2>;
> +				gpio-controller;
> +				compatible = "aspeed,ast2600-sgpiom";
> +				reg = <0x1e780500 0x100>;
> +				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
> +				max-ngpios = <128>;

I need to think more about this one.

Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-26  9:46   ` Steven Lee
@ 2021-05-27  1:42     ` Jeremy Kerr
  -1 siblings, 0 replies; 38+ messages in thread
From: Jeremy Kerr @ 2021-05-27  1:42 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongweiz, ryan_chen, billy_tsai

Hi Steven,

> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more
> flexible.

There are a number of things going on here - you're doing the YAML
conversion, introducing the max-gpios property, and changing the
compatible value.

The first two make sense, but may be better split into separate
changes, so that the YAML conversion is a "linear" change.

I'm not clear on why you're changing the compatible value though,
particularly as you're dropping support for the existing compatible
value anyway. How about we keep the old one, and use the default of 128
for cases where max-ngpios is absent? That way, we retain support for
the existing device trees.

> +  max-ngpios:
> +    description:
> +      represents the number of actual hardware-supported GPIOs (ie,
> +      slots within the clocked serial GPIO data). Since each HW GPIO is both an
> +      input and an output, we provide max_ngpios * 2 lines on our gpiochip
> +      device. We also use it to define the split between the inputs and
> +      outputs; the inputs start at line 0, the outputs start at max_ngpios.

Most of this description is better suited to the ngpios property, which
controls the number of lines that the gpiochip will expose.

Also, minor nit: max_ngpios -> max-ngpios.

How about something like:

  ngpios:
    description:
      Number of GPIO lines to expose. Since each HW GPIO is an input and an
      output, we provide ngpios * 2 lines on our chip device. We also use it
      to define the split between the inputs and outputs; the inputs start at
      line 0, the outputs start at ngpios.

  max-ngpios:
    description:
      Represents the number of actual hardware-supported GPIOs (ie, slots within
      the clocked serial GPIO data), and therefore the maximum value for
      the ngpios property

> +    minimum: 0
> +    maximum: 128

Will this be the case for all (future) hardware? Can we leave this
unbound?

Cheers,


Jeremy


^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-27  1:42     ` Jeremy Kerr
  0 siblings, 0 replies; 38+ messages in thread
From: Jeremy Kerr @ 2021-05-27  1:42 UTC (permalink / raw)
  To: Steven Lee, Linus Walleij, Bartosz Golaszewski, Rob Herring,
	Joel Stanley, Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list
  Cc: Hongweiz, ryan_chen, billy_tsai

Hi Steven,

> SGPIO bindings should be converted as yaml format.
> In addition to the file conversion, a new property max-ngpios is
> added in the yaml file as well.
> The new property is required by the enhanced sgpio driver for
> making the configuration of the max number of gpio pins more
> flexible.

There are a number of things going on here - you're doing the YAML
conversion, introducing the max-gpios property, and changing the
compatible value.

The first two make sense, but may be better split into separate
changes, so that the YAML conversion is a "linear" change.

I'm not clear on why you're changing the compatible value though,
particularly as you're dropping support for the existing compatible
value anyway. How about we keep the old one, and use the default of 128
for cases where max-ngpios is absent? That way, we retain support for
the existing device trees.

> +  max-ngpios:
> +    description:
> +      represents the number of actual hardware-supported GPIOs (ie,
> +      slots within the clocked serial GPIO data). Since each HW GPIO is both an
> +      input and an output, we provide max_ngpios * 2 lines on our gpiochip
> +      device. We also use it to define the split between the inputs and
> +      outputs; the inputs start at line 0, the outputs start at max_ngpios.

Most of this description is better suited to the ngpios property, which
controls the number of lines that the gpiochip will expose.

Also, minor nit: max_ngpios -> max-ngpios.

How about something like:

  ngpios:
    description:
      Number of GPIO lines to expose. Since each HW GPIO is an input and an
      output, we provide ngpios * 2 lines on our chip device. We also use it
      to define the split between the inputs and outputs; the inputs start at
      line 0, the outputs start at ngpios.

  max-ngpios:
    description:
      Represents the number of actual hardware-supported GPIOs (ie, slots within
      the clocked serial GPIO data), and therefore the maximum value for
      the ngpios property

> +    minimum: 0
> +    maximum: 128

Will this be the case for all (future) hardware? Can we leave this
unbound?

Cheers,


Jeremy


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-27  1:26     ` Andrew Jeffery
@ 2021-05-27  2:34       ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:34 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

The 05/27/2021 09:26, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
> > with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
> > supports up to 80 pins.
> > In the current driver design, the max number of sgpio pins is hardcoded
> > in macro MAX_NR_HW_SGPIO and the value is 80.
> > The patch makes the maximum gpio number *constraint of chip* comes from
> > the dts. The property name is max-ngpios.
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
> >  1 file changed, 125 insertions(+), 68 deletions(-)
> > 
> > diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
> > index 64e54f8c30d2..f74bd7988b66 100644
> > --- a/drivers/gpio/gpio-aspeed-sgpio.c
> > +++ b/drivers/gpio/gpio-aspeed-sgpio.c
> > @@ -17,37 +17,28 @@
> >  #include <linux/spinlock.h>
> >  #include <linux/string.h>
> >  
> > -/*
> > - * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
> > - * slots within the clocked serial GPIO data). Since each HW GPIO is both an
> > - * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
> > - * device.
> > - *
> > - * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
> > - * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
> > - */
> > -#define MAX_NR_HW_SGPIO			80
> > -#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO
> 
> I expect you want to keep these as fall-back values for old devicetrees 
> that don't have your new properties.
> 

I will add them back as fall-back values.

> > -
> >  #define ASPEED_SGPIO_CTRL		0x54
> >  
> > -#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
> > +#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)
> 
> Is this valid on the 2400 and 2500? This might be better defined in 
> platform data so we don't go writing reserved bits on older hardware?
> 

I will add them in platform data.
> >  #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
> >  #define ASPEED_SGPIO_ENABLE		BIT(0)
> >  
> >  struct aspeed_sgpio {
> >  	struct gpio_chip chip;
> > +	struct irq_chip intc;
> 
> My gut instinct is this could be in a separate patch.
> 

I will move irq releated modification to a separate patch.
> >  	struct clk *pclk;
> >  	spinlock_t lock;
> >  	void __iomem *base;
> >  	int irq;
> > +	int max_ngpios;
> >  	int n_sgpio;
> >  };
> >  
> >  struct aspeed_sgpio_bank {
> > -	uint16_t    val_regs;
> > -	uint16_t    rdata_reg;
> > -	uint16_t    irq_regs;
> > +	u16    val_regs;
> > +	u16    rdata_reg;
> > +	u16    irq_regs;
> > +	u16    tolerance_regs;
> 
> I suggest splitting this out into a cleanup patch.
> 

Will do it.
> >  	const char  names[4][3];
> >  };
> >  
> > @@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank 
> > aspeed_sgpio_banks[] = {
> >  		.val_regs = 0x0000,
> >  		.rdata_reg = 0x0070,
> >  		.irq_regs = 0x0004,
> > +		.tolerance_regs = 0x0018,
> 
> Reset tolerance is supported on the 2500. Your patch subject claims its 
> for supporting the 2600, so I think the addition of reset tolerance 
> capabilities should be a separate patch.
> 
> By splitting out the different patches as requested, I think this patch 
> can just become about supporting max_ngpios, which will help focus the 
> discussion.
> 

Ok, I will separate max_ngpios, irq_chip and tolerance_reg to 3 patches
in V3.

> >  		.names = { "A", "B", "C", "D" },
> >  	},
> >  	{
> >  		.val_regs = 0x001C,
> >  		.rdata_reg = 0x0074,
> >  		.irq_regs = 0x0020,
> > +		.tolerance_regs = 0x0034,
> >  		.names = { "E", "F", "G", "H" },
> >  	},
> >  	{
> >  		.val_regs = 0x0038,
> >  		.rdata_reg = 0x0078,
> >  		.irq_regs = 0x003C,
> > -		.names = { "I", "J" },
> > +		.tolerance_regs = 0x0050,
> > +		.names = { "I", "J", "K", "L" },
> > +	},
> > +	{
> > +		.val_regs = 0x0090,
> > +		.rdata_reg = 0x007C,
> > +		.irq_regs = 0x0094,
> > +		.tolerance_regs = 0x00A8,
> > +		.names = { "M", "N", "O", "P" },
> >  	},
> >  };
> >  
> > @@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
> >  	reg_irq_type1,
> >  	reg_irq_type2,
> >  	reg_irq_status,
> > +	reg_tolerance,
> >  };
> >  
> > -#define GPIO_VAL_VALUE      0x00
> > -#define GPIO_IRQ_ENABLE     0x00
> > -#define GPIO_IRQ_TYPE0      0x04
> > -#define GPIO_IRQ_TYPE1      0x08
> > -#define GPIO_IRQ_TYPE2      0x0C
> > -#define GPIO_IRQ_STATUS     0x10
> > +#define GPIO_IRQ_OFFSET_ENABLE     0x00
> > +#define GPIO_IRQ_OFFSET_TYPE0      0x04
> > +#define GPIO_IRQ_OFFSET_TYPE1      0x08
> > +#define GPIO_IRQ_OFFSET_TYPE2      0x0C
> > +#define GPIO_IRQ_OFFSET_STATUS     0x10
> 
> I don't think this change is necessary. It looks like it produces a 
> bunch of noise in the diff.
> 

Thanks, I will use the original naming.

> >  
> >  static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
> >  				     const struct aspeed_sgpio_bank *bank,
> > @@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
> >  {
> >  	switch (reg) {
> >  	case reg_val:
> > -		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
> > +		return gpio->base + bank->val_regs;
> >  	case reg_rdata:
> >  		return gpio->base + bank->rdata_reg;
> >  	case reg_irq_enable:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
> >  	case reg_irq_type0:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
> >  	case reg_irq_type1:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
> >  	case reg_irq_type2:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
> >  	case reg_irq_status:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
> > +	case reg_tolerance:
> > +		return gpio->base + bank->tolerance_regs;
> >  	default:
> >  		/* acturally if code runs to here, it's an error case */
> >  		BUG();
> >  	}
> >  }
> >  
> > -#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
> > -#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
> > +#define GPIO_BANK(x)    ((x) >> 5)
> > +/* modulo 32 */
> > +#define GPIO_OFFSET(x)  ((x) & 0x1f)
> >  #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
> >  
> > -static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
> > +static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, 
> > unsigned int max_ngpios)
> >  {
> >  	unsigned int bank;
> >  
> > -	bank = GPIO_BANK(offset);
> > +	bank = GPIO_BANK(offset % max_ngpios);
> >  
> >  	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
> >  	return &aspeed_sgpio_banks[bank];
> > @@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct 
> > gpio_chip *gc,
> >  		unsigned long *valid_mask, unsigned int ngpios)
> >  {
> >  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> > +	int max_ngpios = sgpio->max_ngpios;
> >  	int n = sgpio->n_sgpio;
> > -	int c = SGPIO_OUTPUT_OFFSET - n;
> > +	int c = max_ngpios - n;
> >  
> > -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> > +	WARN_ON(ngpios < max_ngpios * 2);
> >  
> >  	/* input GPIOs in the lower range */
> >  	bitmap_set(valid_mask, 0, n);
> >  	bitmap_clear(valid_mask, n, c);
> >  
> > -	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
> > -	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
> > -	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
> > +	/* output GPIOS above max_ngpios */
> > +	bitmap_set(valid_mask, max_ngpios, n);
> > +	bitmap_clear(valid_mask, max_ngpios + n, c);
> >  
> >  	return 0;
> >  }
> > @@ -161,30 +166,30 @@ static void 
> > aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
> >  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> >  	int n = sgpio->n_sgpio;
> >  
> > -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> > +	WARN_ON(ngpios < sgpio->max_ngpios * 2);
> >  
> >  	/* input GPIOs in the lower range */
> >  	bitmap_set(valid_mask, 0, n);
> >  	bitmap_clear(valid_mask, n, ngpios - n);
> >  }
> >  
> > -static bool aspeed_sgpio_is_input(unsigned int offset)
> > +static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
> >  {
> > -	return offset < SGPIO_OUTPUT_OFFSET;
> > +	return offset < max_ngpios;
> >  }
> >  
> >  static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
> >  {
> > +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
> 
> This references gpio before it's defined, as the build bot noted.
> 

Sorry for that, I've sent v2 patch for this issue.

> >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> >  	unsigned long flags;
> >  	enum aspeed_sgpio_reg reg;
> >  	int rc = 0;
> >  
> >  	spin_lock_irqsave(&gpio->lock, flags);
> >  
> > -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> > -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> > +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> > reg_rdata;
> 
> We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
> 'const struct aspeed_sgpio *' parameter), rather than open-coding 
> gpio->max_ngpios. This approach will make it easier to refactor the 
> implementation in the future (if necessary).
> 

I will rewrite the function to use aspeed_sgpio struct.

> > +	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % 
> > gpio->max_ngpios));
> >  
> >  	spin_unlock_irqrestore(&gpio->lock, flags);
> >  
> > @@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, 
> > unsigned int offset)
> >  
> >  static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
> >  {
> > +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
> 
> References gpio before it's defined.
> 

Fixed in V2.

> >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> >  	void __iomem *addr_r, *addr_w;
> >  	u32 reg = 0;
> >  
> > -	if (aspeed_sgpio_is_input(offset))
> > +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
> >  		return -EINVAL;
> >  
> >  	/* Since this is an output, read the cached value from rdata, then
> > @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> > unsigned int offset, int val)
> >  	reg = ioread32(addr_r);
> >  
> >  	if (val)
> > -		reg |= GPIO_BIT(offset);
> > +		reg |= GPIO_BIT(offset % gpio->max_ngpios);
> 
> Pass gpio through GPIO_BIT() too.
> 

I don't understand this comment.
Could you describe more?
Thanks.

> >  	else
> > -		reg &= ~GPIO_BIT(offset);
> > +		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
> >  
> >  	iowrite32(reg, addr_w);
> >  
> > @@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, 
> > unsigned int offset, int val)
> >  
> >  static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
> >  {
> > -	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > +
> > +	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
> >  }
> >  
> >  static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int 
> > offset, int val)
> > @@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip 
> > *gc, unsigned int offset, int v
> >  
> >  static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned 
> > int offset)
> >  {
> > -	return !!aspeed_sgpio_is_input(offset);
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > +
> > +	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
> >  }
> >  
> >  static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> > @@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> >  	WARN_ON(!internal);
> >  
> >  	*gpio = internal;
> > -	*bank = to_bank(*offset);
> > -	*bit = GPIO_BIT(*offset);
> > +	*bank = to_bank(*offset, internal->max_ngpios);
> 
> Make to_bank take a 'const struct aspeed_sgpio *'.
> 

Will modify it.

> > +	*bit = GPIO_BIT(*offset % internal->max_ngpios);
> >  }
> >  
> >  static void aspeed_sgpio_irq_ack(struct irq_data *d)
> > @@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
> >  	chained_irq_exit(ic, desc);
> >  }
> >  
> > -static struct irq_chip aspeed_sgpio_irqchip = {
> > -	.name       = "aspeed-sgpio",
> > -	.irq_ack    = aspeed_sgpio_irq_ack,
> > -	.irq_mask   = aspeed_sgpio_irq_mask,
> > -	.irq_unmask = aspeed_sgpio_irq_unmask,
> > -	.irq_set_type   = aspeed_sgpio_set_type,
> > -};
> > -
> >  static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
> >  				   struct platform_device *pdev)
> >  {
> > @@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct 
> > aspeed_sgpio *gpio,
> >  		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
> >  	}
> >  
> > +	gpio->intc.name = dev_name(&pdev->dev);
> > +	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
> > +	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
> > +	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
> > +	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
> > +
> >  	irq = &gpio->chip.irq;
> > -	irq->chip = &aspeed_sgpio_irqchip;
> > +	irq->chip = &gpio->intc;
> 
> As I said earlier, I think it would be best to split the IRQ changes 
> out into a separate patch ordered before this one.
> 

Will separate irq changes to another patch.

> >  	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
> >  	irq->handler = handle_bad_irq;
> >  	irq->default_type = IRQ_TYPE_NONE;
> > @@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct 
> > aspeed_sgpio *gpio,
> >  	return 0;
> >  }
> >  
> > +static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
> > +					unsigned int offset, bool enable)
> > +{
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
> > +	unsigned long flags;
> > +	void __iomem *reg;
> > +	u32 val;
> > +
> > +	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
> > +
> > +	spin_lock_irqsave(&gpio->lock, flags);
> > +
> > +	val = readl(reg);
> > +
> > +	if (enable)
> > +		val |= GPIO_BIT(offset % gpio->max_ngpios);
> > +	else
> > +		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
> > +
> > +	writel(val, reg);
> > +
> > +	spin_unlock_irqrestore(&gpio->lock, flags);
> > +
> > +	return 0;
> > +}
> > +
> > +static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
> > +				   unsigned long config)
> > +{
> > +	unsigned long param = pinconf_to_config_param(config);
> > +	u32 arg = pinconf_to_config_argument(config);
> > +
> > +	if (param == PIN_CONFIG_PERSIST_STATE)
> > +		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
> > +	else
> > +		return -EOPNOTSUPP;
> > +}
> > +
> 
> Again, lets split reset tolerance out to a separate patch as well.
> 

Will separate tolerance to another patch.

> >  static const struct of_device_id aspeed_sgpio_of_table[] = {
> > -	{ .compatible = "aspeed,ast2400-sgpio" },
> > -	{ .compatible = "aspeed,ast2500-sgpio" },
> > +	{ .compatible = "aspeed,ast2400-sgpiom" },
> > +	{ .compatible = "aspeed,ast2500-sgpiom" },
> > +	{ .compatible = "aspeed,ast2600-sgpiom" },
> 
> No. You'll need to support both while we transition between the two as 
> the devicetree patch and the driver patch are separate, and this will 
> break bisectability when tracking down runtime issues. Also, the 
> devicetree patch should be ordered after the driver change for the same 
> reason.
> 
> You should continue to support the 'aspeed,ast2{4,5}00-sgpio' style 
> compatible here to handle old devicetrees.
> 
> So in summary, just add the 'aspeed,ast2{4,5,6}00-sgpiom' compatibles 
> for now, don't remove the 'aspeed,ast2{4,5}00-sgpio' compatibles.
> 

I will keep the original implementation and just add ast2600 compatible.

> >  	{}
> >  };
> >  
> > @@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
> >  
> >  static int __init aspeed_sgpio_probe(struct platform_device *pdev)
> >  {
> > +	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
> >  	struct aspeed_sgpio *gpio;
> > -	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
> >  	int rc;
> >  	unsigned long apb_freq;
> >  
> > @@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct 
> > platform_device *pdev)
> >  	if (IS_ERR(gpio->base))
> >  		return PTR_ERR(gpio->base);
> >  
> > +	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", 
> > &max_ngpios);
> > +	if (rc < 0) {
> > +		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
> > +		return -EINVAL;
> 
> I don't think this is right. You should just fall back to using the 
> hard-coded value.
> 

I will use fallback value rather than returning error. 

> Cheers,
> 
> Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-27  2:34       ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:34 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

The 05/27/2021 09:26, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > AST2600 SoC has 2 SGPIO master interfaces one with 128 pins another one
> > with 80 pins, and AST2500/AST2400 SoC has 1 SGPIO master interface that
> > supports up to 80 pins.
> > In the current driver design, the max number of sgpio pins is hardcoded
> > in macro MAX_NR_HW_SGPIO and the value is 80.
> > The patch makes the maximum gpio number *constraint of chip* comes from
> > the dts. The property name is max-ngpios.
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  drivers/gpio/gpio-aspeed-sgpio.c | 193 ++++++++++++++++++++-----------
> >  1 file changed, 125 insertions(+), 68 deletions(-)
> > 
> > diff --git a/drivers/gpio/gpio-aspeed-sgpio.c b/drivers/gpio/gpio-aspeed-sgpio.c
> > index 64e54f8c30d2..f74bd7988b66 100644
> > --- a/drivers/gpio/gpio-aspeed-sgpio.c
> > +++ b/drivers/gpio/gpio-aspeed-sgpio.c
> > @@ -17,37 +17,28 @@
> >  #include <linux/spinlock.h>
> >  #include <linux/string.h>
> >  
> > -/*
> > - * MAX_NR_HW_GPIO represents the number of actual hardware-supported GPIOs (ie,
> > - * slots within the clocked serial GPIO data). Since each HW GPIO is both an
> > - * input and an output, we provide MAX_NR_HW_GPIO * 2 lines on our gpiochip
> > - * device.
> > - *
> > - * We use SGPIO_OUTPUT_OFFSET to define the split between the inputs and
> > - * outputs; the inputs start at line 0, the outputs start at OUTPUT_OFFSET.
> > - */
> > -#define MAX_NR_HW_SGPIO			80
> > -#define SGPIO_OUTPUT_OFFSET		MAX_NR_HW_SGPIO
> 
> I expect you want to keep these as fall-back values for old devicetrees 
> that don't have your new properties.
> 

I will add them back as fall-back values.

> > -
> >  #define ASPEED_SGPIO_CTRL		0x54
> >  
> > -#define ASPEED_SGPIO_PINS_MASK		GENMASK(9, 6)
> > +#define ASPEED_SGPIO_PINS_MASK		GENMASK(10, 6)
> 
> Is this valid on the 2400 and 2500? This might be better defined in 
> platform data so we don't go writing reserved bits on older hardware?
> 

I will add them in platform data.
> >  #define ASPEED_SGPIO_CLK_DIV_MASK	GENMASK(31, 16)
> >  #define ASPEED_SGPIO_ENABLE		BIT(0)
> >  
> >  struct aspeed_sgpio {
> >  	struct gpio_chip chip;
> > +	struct irq_chip intc;
> 
> My gut instinct is this could be in a separate patch.
> 

I will move irq releated modification to a separate patch.
> >  	struct clk *pclk;
> >  	spinlock_t lock;
> >  	void __iomem *base;
> >  	int irq;
> > +	int max_ngpios;
> >  	int n_sgpio;
> >  };
> >  
> >  struct aspeed_sgpio_bank {
> > -	uint16_t    val_regs;
> > -	uint16_t    rdata_reg;
> > -	uint16_t    irq_regs;
> > +	u16    val_regs;
> > +	u16    rdata_reg;
> > +	u16    irq_regs;
> > +	u16    tolerance_regs;
> 
> I suggest splitting this out into a cleanup patch.
> 

Will do it.
> >  	const char  names[4][3];
> >  };
> >  
> > @@ -63,19 +54,29 @@ static const struct aspeed_sgpio_bank 
> > aspeed_sgpio_banks[] = {
> >  		.val_regs = 0x0000,
> >  		.rdata_reg = 0x0070,
> >  		.irq_regs = 0x0004,
> > +		.tolerance_regs = 0x0018,
> 
> Reset tolerance is supported on the 2500. Your patch subject claims its 
> for supporting the 2600, so I think the addition of reset tolerance 
> capabilities should be a separate patch.
> 
> By splitting out the different patches as requested, I think this patch 
> can just become about supporting max_ngpios, which will help focus the 
> discussion.
> 

Ok, I will separate max_ngpios, irq_chip and tolerance_reg to 3 patches
in V3.

> >  		.names = { "A", "B", "C", "D" },
> >  	},
> >  	{
> >  		.val_regs = 0x001C,
> >  		.rdata_reg = 0x0074,
> >  		.irq_regs = 0x0020,
> > +		.tolerance_regs = 0x0034,
> >  		.names = { "E", "F", "G", "H" },
> >  	},
> >  	{
> >  		.val_regs = 0x0038,
> >  		.rdata_reg = 0x0078,
> >  		.irq_regs = 0x003C,
> > -		.names = { "I", "J" },
> > +		.tolerance_regs = 0x0050,
> > +		.names = { "I", "J", "K", "L" },
> > +	},
> > +	{
> > +		.val_regs = 0x0090,
> > +		.rdata_reg = 0x007C,
> > +		.irq_regs = 0x0094,
> > +		.tolerance_regs = 0x00A8,
> > +		.names = { "M", "N", "O", "P" },
> >  	},
> >  };
> >  
> > @@ -87,14 +88,14 @@ enum aspeed_sgpio_reg {
> >  	reg_irq_type1,
> >  	reg_irq_type2,
> >  	reg_irq_status,
> > +	reg_tolerance,
> >  };
> >  
> > -#define GPIO_VAL_VALUE      0x00
> > -#define GPIO_IRQ_ENABLE     0x00
> > -#define GPIO_IRQ_TYPE0      0x04
> > -#define GPIO_IRQ_TYPE1      0x08
> > -#define GPIO_IRQ_TYPE2      0x0C
> > -#define GPIO_IRQ_STATUS     0x10
> > +#define GPIO_IRQ_OFFSET_ENABLE     0x00
> > +#define GPIO_IRQ_OFFSET_TYPE0      0x04
> > +#define GPIO_IRQ_OFFSET_TYPE1      0x08
> > +#define GPIO_IRQ_OFFSET_TYPE2      0x0C
> > +#define GPIO_IRQ_OFFSET_STATUS     0x10
> 
> I don't think this change is necessary. It looks like it produces a 
> bunch of noise in the diff.
> 

Thanks, I will use the original naming.

> >  
> >  static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
> >  				     const struct aspeed_sgpio_bank *bank,
> > @@ -102,34 +103,37 @@ static void __iomem *bank_reg(struct aspeed_sgpio *gpio,
> >  {
> >  	switch (reg) {
> >  	case reg_val:
> > -		return gpio->base + bank->val_regs + GPIO_VAL_VALUE;
> > +		return gpio->base + bank->val_regs;
> >  	case reg_rdata:
> >  		return gpio->base + bank->rdata_reg;
> >  	case reg_irq_enable:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_ENABLE;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_ENABLE;
> >  	case reg_irq_type0:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE0;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE0;
> >  	case reg_irq_type1:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE1;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE1;
> >  	case reg_irq_type2:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_TYPE2;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_TYPE2;
> >  	case reg_irq_status:
> > -		return gpio->base + bank->irq_regs + GPIO_IRQ_STATUS;
> > +		return gpio->base + bank->irq_regs + GPIO_IRQ_OFFSET_STATUS;
> > +	case reg_tolerance:
> > +		return gpio->base + bank->tolerance_regs;
> >  	default:
> >  		/* acturally if code runs to here, it's an error case */
> >  		BUG();
> >  	}
> >  }
> >  
> > -#define GPIO_BANK(x)    ((x % SGPIO_OUTPUT_OFFSET) >> 5)
> > -#define GPIO_OFFSET(x)  ((x % SGPIO_OUTPUT_OFFSET) & 0x1f)
> > +#define GPIO_BANK(x)    ((x) >> 5)
> > +/* modulo 32 */
> > +#define GPIO_OFFSET(x)  ((x) & 0x1f)
> >  #define GPIO_BIT(x)     BIT(GPIO_OFFSET(x))
> >  
> > -static const struct aspeed_sgpio_bank *to_bank(unsigned int offset)
> > +static const struct aspeed_sgpio_bank *to_bank(unsigned int offset, 
> > unsigned int max_ngpios)
> >  {
> >  	unsigned int bank;
> >  
> > -	bank = GPIO_BANK(offset);
> > +	bank = GPIO_BANK(offset % max_ngpios);
> >  
> >  	WARN_ON(bank >= ARRAY_SIZE(aspeed_sgpio_banks));
> >  	return &aspeed_sgpio_banks[bank];
> > @@ -139,18 +143,19 @@ static int aspeed_sgpio_init_valid_mask(struct 
> > gpio_chip *gc,
> >  		unsigned long *valid_mask, unsigned int ngpios)
> >  {
> >  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> > +	int max_ngpios = sgpio->max_ngpios;
> >  	int n = sgpio->n_sgpio;
> > -	int c = SGPIO_OUTPUT_OFFSET - n;
> > +	int c = max_ngpios - n;
> >  
> > -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> > +	WARN_ON(ngpios < max_ngpios * 2);
> >  
> >  	/* input GPIOs in the lower range */
> >  	bitmap_set(valid_mask, 0, n);
> >  	bitmap_clear(valid_mask, n, c);
> >  
> > -	/* output GPIOS above SGPIO_OUTPUT_OFFSET */
> > -	bitmap_set(valid_mask, SGPIO_OUTPUT_OFFSET, n);
> > -	bitmap_clear(valid_mask, SGPIO_OUTPUT_OFFSET + n, c);
> > +	/* output GPIOS above max_ngpios */
> > +	bitmap_set(valid_mask, max_ngpios, n);
> > +	bitmap_clear(valid_mask, max_ngpios + n, c);
> >  
> >  	return 0;
> >  }
> > @@ -161,30 +166,30 @@ static void 
> > aspeed_sgpio_irq_init_valid_mask(struct gpio_chip *gc,
> >  	struct aspeed_sgpio *sgpio = gpiochip_get_data(gc);
> >  	int n = sgpio->n_sgpio;
> >  
> > -	WARN_ON(ngpios < MAX_NR_HW_SGPIO * 2);
> > +	WARN_ON(ngpios < sgpio->max_ngpios * 2);
> >  
> >  	/* input GPIOs in the lower range */
> >  	bitmap_set(valid_mask, 0, n);
> >  	bitmap_clear(valid_mask, n, ngpios - n);
> >  }
> >  
> > -static bool aspeed_sgpio_is_input(unsigned int offset)
> > +static bool aspeed_sgpio_is_input(unsigned int offset, unsigned int max_ngpios)
> >  {
> > -	return offset < SGPIO_OUTPUT_OFFSET;
> > +	return offset < max_ngpios;
> >  }
> >  
> >  static int aspeed_sgpio_get(struct gpio_chip *gc, unsigned int offset)
> >  {
> > +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
> 
> This references gpio before it's defined, as the build bot noted.
> 

Sorry for that, I've sent v2 patch for this issue.

> >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> >  	unsigned long flags;
> >  	enum aspeed_sgpio_reg reg;
> >  	int rc = 0;
> >  
> >  	spin_lock_irqsave(&gpio->lock, flags);
> >  
> > -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> > -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> > +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> > reg_rdata;
> 
> We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
> 'const struct aspeed_sgpio *' parameter), rather than open-coding 
> gpio->max_ngpios. This approach will make it easier to refactor the 
> implementation in the future (if necessary).
> 

I will rewrite the function to use aspeed_sgpio struct.

> > +	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset % 
> > gpio->max_ngpios));
> >  
> >  	spin_unlock_irqrestore(&gpio->lock, flags);
> >  
> > @@ -193,12 +198,12 @@ static int aspeed_sgpio_get(struct gpio_chip *gc, 
> > unsigned int offset)
> >  
> >  static int sgpio_set_value(struct gpio_chip *gc, unsigned int offset, int val)
> >  {
> > +	const struct aspeed_sgpio_bank *bank = to_bank(offset, gpio->max_ngpios);
> 
> References gpio before it's defined.
> 

Fixed in V2.

> >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> >  	void __iomem *addr_r, *addr_w;
> >  	u32 reg = 0;
> >  
> > -	if (aspeed_sgpio_is_input(offset))
> > +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
> >  		return -EINVAL;
> >  
> >  	/* Since this is an output, read the cached value from rdata, then
> > @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> > unsigned int offset, int val)
> >  	reg = ioread32(addr_r);
> >  
> >  	if (val)
> > -		reg |= GPIO_BIT(offset);
> > +		reg |= GPIO_BIT(offset % gpio->max_ngpios);
> 
> Pass gpio through GPIO_BIT() too.
> 

I don't understand this comment.
Could you describe more?
Thanks.

> >  	else
> > -		reg &= ~GPIO_BIT(offset);
> > +		reg &= ~GPIO_BIT(offset % gpio->max_ngpios);
> >  
> >  	iowrite32(reg, addr_w);
> >  
> > @@ -232,7 +237,9 @@ static void aspeed_sgpio_set(struct gpio_chip *gc, 
> > unsigned int offset, int val)
> >  
> >  static int aspeed_sgpio_dir_in(struct gpio_chip *gc, unsigned int offset)
> >  {
> > -	return aspeed_sgpio_is_input(offset) ? 0 : -EINVAL;
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > +
> > +	return aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? 0 : -EINVAL;
> >  }
> >  
> >  static int aspeed_sgpio_dir_out(struct gpio_chip *gc, unsigned int 
> > offset, int val)
> > @@ -253,7 +260,9 @@ static int aspeed_sgpio_dir_out(struct gpio_chip 
> > *gc, unsigned int offset, int v
> >  
> >  static int aspeed_sgpio_get_direction(struct gpio_chip *gc, unsigned 
> > int offset)
> >  {
> > -	return !!aspeed_sgpio_is_input(offset);
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > +
> > +	return !!aspeed_sgpio_is_input(offset, gpio->max_ngpios);
> >  }
> >  
> >  static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> > @@ -268,8 +277,8 @@ static void irqd_to_aspeed_sgpio_data(struct irq_data *d,
> >  	WARN_ON(!internal);
> >  
> >  	*gpio = internal;
> > -	*bank = to_bank(*offset);
> > -	*bit = GPIO_BIT(*offset);
> > +	*bank = to_bank(*offset, internal->max_ngpios);
> 
> Make to_bank take a 'const struct aspeed_sgpio *'.
> 

Will modify it.

> > +	*bit = GPIO_BIT(*offset % internal->max_ngpios);
> >  }
> >  
> >  static void aspeed_sgpio_irq_ack(struct irq_data *d)
> > @@ -412,14 +421,6 @@ static void aspeed_sgpio_irq_handler(struct irq_desc *desc)
> >  	chained_irq_exit(ic, desc);
> >  }
> >  
> > -static struct irq_chip aspeed_sgpio_irqchip = {
> > -	.name       = "aspeed-sgpio",
> > -	.irq_ack    = aspeed_sgpio_irq_ack,
> > -	.irq_mask   = aspeed_sgpio_irq_mask,
> > -	.irq_unmask = aspeed_sgpio_irq_unmask,
> > -	.irq_set_type   = aspeed_sgpio_set_type,
> > -};
> > -
> >  static int aspeed_sgpio_setup_irqs(struct aspeed_sgpio *gpio,
> >  				   struct platform_device *pdev)
> >  {
> > @@ -442,8 +443,14 @@ static int aspeed_sgpio_setup_irqs(struct 
> > aspeed_sgpio *gpio,
> >  		iowrite32(0xffffffff, bank_reg(gpio, bank, reg_irq_status));
> >  	}
> >  
> > +	gpio->intc.name = dev_name(&pdev->dev);
> > +	gpio->intc.irq_ack = aspeed_sgpio_irq_ack;
> > +	gpio->intc.irq_mask = aspeed_sgpio_irq_mask;
> > +	gpio->intc.irq_unmask = aspeed_sgpio_irq_unmask;
> > +	gpio->intc.irq_set_type = aspeed_sgpio_set_type;
> > +
> >  	irq = &gpio->chip.irq;
> > -	irq->chip = &aspeed_sgpio_irqchip;
> > +	irq->chip = &gpio->intc;
> 
> As I said earlier, I think it would be best to split the IRQ changes 
> out into a separate patch ordered before this one.
> 

Will separate irq changes to another patch.

> >  	irq->init_valid_mask = aspeed_sgpio_irq_init_valid_mask;
> >  	irq->handler = handle_bad_irq;
> >  	irq->default_type = IRQ_TYPE_NONE;
> > @@ -466,9 +473,48 @@ static int aspeed_sgpio_setup_irqs(struct 
> > aspeed_sgpio *gpio,
> >  	return 0;
> >  }
> >  
> > +static int aspeed_sgpio_reset_tolerance(struct gpio_chip *chip,
> > +					unsigned int offset, bool enable)
> > +{
> > +	struct aspeed_sgpio *gpio = gpiochip_get_data(chip);
> > +	unsigned long flags;
> > +	void __iomem *reg;
> > +	u32 val;
> > +
> > +	reg = bank_reg(gpio, to_bank(offset, gpio->max_ngpios), reg_tolerance);
> > +
> > +	spin_lock_irqsave(&gpio->lock, flags);
> > +
> > +	val = readl(reg);
> > +
> > +	if (enable)
> > +		val |= GPIO_BIT(offset % gpio->max_ngpios);
> > +	else
> > +		val &= ~GPIO_BIT(offset % gpio->max_ngpios);
> > +
> > +	writel(val, reg);
> > +
> > +	spin_unlock_irqrestore(&gpio->lock, flags);
> > +
> > +	return 0;
> > +}
> > +
> > +static int aspeed_sgpio_set_config(struct gpio_chip *chip, unsigned int offset,
> > +				   unsigned long config)
> > +{
> > +	unsigned long param = pinconf_to_config_param(config);
> > +	u32 arg = pinconf_to_config_argument(config);
> > +
> > +	if (param == PIN_CONFIG_PERSIST_STATE)
> > +		return aspeed_sgpio_reset_tolerance(chip, offset, arg);
> > +	else
> > +		return -EOPNOTSUPP;
> > +}
> > +
> 
> Again, lets split reset tolerance out to a separate patch as well.
> 

Will separate tolerance to another patch.

> >  static const struct of_device_id aspeed_sgpio_of_table[] = {
> > -	{ .compatible = "aspeed,ast2400-sgpio" },
> > -	{ .compatible = "aspeed,ast2500-sgpio" },
> > +	{ .compatible = "aspeed,ast2400-sgpiom" },
> > +	{ .compatible = "aspeed,ast2500-sgpiom" },
> > +	{ .compatible = "aspeed,ast2600-sgpiom" },
> 
> No. You'll need to support both while we transition between the two as 
> the devicetree patch and the driver patch are separate, and this will 
> break bisectability when tracking down runtime issues. Also, the 
> devicetree patch should be ordered after the driver change for the same 
> reason.
> 
> You should continue to support the 'aspeed,ast2{4,5}00-sgpio' style 
> compatible here to handle old devicetrees.
> 
> So in summary, just add the 'aspeed,ast2{4,5,6}00-sgpiom' compatibles 
> for now, don't remove the 'aspeed,ast2{4,5}00-sgpio' compatibles.
> 

I will keep the original implementation and just add ast2600 compatible.

> >  	{}
> >  };
> >  
> > @@ -476,8 +522,8 @@ MODULE_DEVICE_TABLE(of, aspeed_sgpio_of_table);
> >  
> >  static int __init aspeed_sgpio_probe(struct platform_device *pdev)
> >  {
> > +	u32 nr_gpios, sgpio_freq, sgpio_clk_div, max_ngpios;
> >  	struct aspeed_sgpio *gpio;
> > -	u32 nr_gpios, sgpio_freq, sgpio_clk_div;
> >  	int rc;
> >  	unsigned long apb_freq;
> >  
> > @@ -489,13 +535,24 @@ static int __init aspeed_sgpio_probe(struct 
> > platform_device *pdev)
> >  	if (IS_ERR(gpio->base))
> >  		return PTR_ERR(gpio->base);
> >  
> > +	rc = of_property_read_u32(pdev->dev.of_node, "max-ngpios", 
> > &max_ngpios);
> > +	if (rc < 0) {
> > +		dev_err(&pdev->dev, "Could not read max-ngpios property\n");
> > +		return -EINVAL;
> 
> I don't think this is right. You should just fall back to using the 
> hard-coded value.
> 

I will use fallback value rather than returning error. 

> Cheers,
> 
> Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-26 12:56     ` Rob Herring
@ 2021-05-27  2:45       ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:45 UTC (permalink / raw)
  To: Rob Herring
  Cc: Joel Stanley,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	open list, Billy Tsai, moderated list:ARM/ASPEED MACHINE SUPPORT,
	Hongweiz, Bartosz Golaszewski, open list:GPIO SUBSYSTEM,
	Rob Herring, Ryan Chen,
	moderated list:ARM/ASPEED MACHINE SUPPORT, Andrew Jeffery,
	Linus Walleij

The 05/26/2021 20:56, Rob Herring wrote:
> On Wed, 26 May 2021 17:46:05 +0800, Steven Lee wrote:
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more flexible.
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
> >  .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
> >  2 files changed, 91 insertions(+), 46 deletions(-)
> >  create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
> >  delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
> > 
> 
> My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
> on your patch (DT_CHECKER_FLAGS is new in v5.13):
> 
> yamllint warnings/errors:
> 
> dtschema/dtc warnings/errors:
> ./Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml: $id: relative path/filename doesn't match actual path or filename
> 	expected: http://devicetree.org/schemas/gpio/aspeed,sgpio.yaml#
> 
> See https://patchwork.ozlabs.org/patch/1483966
> 
> This check can fail if there are any dependencies. The base for a patch
> series is generally the most recent rc1.
> 
> If you already ran 'make dt_binding_check' and didn't see the above
> error(s), then make sure 'yamllint' is installed and dt-schema is up to
> date:
> 
> pip3 install dtschema --upgrade
> 
> Please check and re-submit.
> 

Thanks, I will modify it.


^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-27  2:45       ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:45 UTC (permalink / raw)
  To: Rob Herring
  Cc: Joel Stanley,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	open list, Billy Tsai, moderated list:ARM/ASPEED MACHINE SUPPORT,
	Hongweiz, Bartosz Golaszewski, open list:GPIO SUBSYSTEM,
	Rob Herring, Ryan Chen,
	moderated list:ARM/ASPEED MACHINE SUPPORT, Andrew Jeffery,
	Linus Walleij

The 05/26/2021 20:56, Rob Herring wrote:
> On Wed, 26 May 2021 17:46:05 +0800, Steven Lee wrote:
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more flexible.
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  .../bindings/gpio/aspeed,sgpio.yaml           | 91 +++++++++++++++++++
> >  .../devicetree/bindings/gpio/sgpio-aspeed.txt | 46 ----------
> >  2 files changed, 91 insertions(+), 46 deletions(-)
> >  create mode 100644 Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml
> >  delete mode 100644 Documentation/devicetree/bindings/gpio/sgpio-aspeed.txt
> > 
> 
> My bot found errors running 'make DT_CHECKER_FLAGS=-m dt_binding_check'
> on your patch (DT_CHECKER_FLAGS is new in v5.13):
> 
> yamllint warnings/errors:
> 
> dtschema/dtc warnings/errors:
> ./Documentation/devicetree/bindings/gpio/aspeed,sgpio.yaml: $id: relative path/filename doesn't match actual path or filename
> 	expected: http://devicetree.org/schemas/gpio/aspeed,sgpio.yaml#
> 
> See https://patchwork.ozlabs.org/patch/1483966
> 
> This check can fail if there are any dependencies. The base for a patch
> series is generally the most recent rc1.
> 
> If you already ran 'make dt_binding_check' and didn't see the above
> error(s), then make sure 'yamllint' is installed and dt-schema is up to
> date:
> 
> pip3 install dtschema --upgrade
> 
> Please check and re-submit.
> 

Thanks, I will modify it.


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
  2021-05-27  1:42     ` Jeremy Kerr
@ 2021-05-27  2:58       ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:58 UTC (permalink / raw)
  To: Jeremy Kerr
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list, Hongweiz,
	Ryan Chen, Billy Tsai

The 05/27/2021 09:42, Jeremy Kerr wrote:
> Hi Steven,
> 
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more
> > flexible.
> 
> There are a number of things going on here - you're doing the YAML
> conversion, introducing the max-gpios property, and changing the
> compatible value.
> 
> The first two make sense, but may be better split into separate
> changes, so that the YAML conversion is a "linear" change.
> 

Thanks for your suggestion, I will split them into 2 patches.

> I'm not clear on why you're changing the compatible value though,
> particularly as you're dropping support for the existing compatible
> value anyway. How about we keep the old one, and use the default of 128
> for cases where max-ngpios is absent? That way, we retain support for
> the existing device trees.
> 

AST2600 support both SGPIO master and slave interfaces. So we decide to
rename sgpio to sgpiom as it is the driver for sgpio master interface.
Since Andrew also point out the compatibility issues, I will modify driver
to keep the original design.

> > +  max-ngpios:
> > +    description:
> > +      represents the number of actual hardware-supported GPIOs (ie,
> > +      slots within the clocked serial GPIO data). Since each HW GPIO is both an
> > +      input and an output, we provide max_ngpios * 2 lines on our gpiochip
> > +      device. We also use it to define the split between the inputs and
> > +      outputs; the inputs start at line 0, the outputs start at max_ngpios.
> 
> Most of this description is better suited to the ngpios property, which
> controls the number of lines that the gpiochip will expose.
> 
> Also, minor nit: max_ngpios -> max-ngpios.
> 
> How about something like:
> 
>   ngpios:
>     description:
>       Number of GPIO lines to expose. Since each HW GPIO is an input and an
>       output, we provide ngpios * 2 lines on our chip device. We also use it
>       to define the split between the inputs and outputs; the inputs start at
>       line 0, the outputs start at ngpios.
> 
>   max-ngpios:
>     description:
>       Represents the number of actual hardware-supported GPIOs (ie, slots within
>       the clocked serial GPIO data), and therefore the maximum value for
>       the ngpios property
> 

I will modify the description as you suggested.

> > +    minimum: 0
> > +    maximum: 128
> 
> Will this be the case for all (future) hardware? Can we leave this
> unbound?
> 

Since the future hardware may supports more gpios, I will remove it.

> Cheers,
> 
> 
> Jeremy
> 

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml.
@ 2021-05-27  2:58       ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  2:58 UTC (permalink / raw)
  To: Jeremy Kerr
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	Andrew Jeffery, open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list, Hongweiz,
	Ryan Chen, Billy Tsai

The 05/27/2021 09:42, Jeremy Kerr wrote:
> Hi Steven,
> 
> > SGPIO bindings should be converted as yaml format.
> > In addition to the file conversion, a new property max-ngpios is
> > added in the yaml file as well.
> > The new property is required by the enhanced sgpio driver for
> > making the configuration of the max number of gpio pins more
> > flexible.
> 
> There are a number of things going on here - you're doing the YAML
> conversion, introducing the max-gpios property, and changing the
> compatible value.
> 
> The first two make sense, but may be better split into separate
> changes, so that the YAML conversion is a "linear" change.
> 

Thanks for your suggestion, I will split them into 2 patches.

> I'm not clear on why you're changing the compatible value though,
> particularly as you're dropping support for the existing compatible
> value anyway. How about we keep the old one, and use the default of 128
> for cases where max-ngpios is absent? That way, we retain support for
> the existing device trees.
> 

AST2600 support both SGPIO master and slave interfaces. So we decide to
rename sgpio to sgpiom as it is the driver for sgpio master interface.
Since Andrew also point out the compatibility issues, I will modify driver
to keep the original design.

> > +  max-ngpios:
> > +    description:
> > +      represents the number of actual hardware-supported GPIOs (ie,
> > +      slots within the clocked serial GPIO data). Since each HW GPIO is both an
> > +      input and an output, we provide max_ngpios * 2 lines on our gpiochip
> > +      device. We also use it to define the split between the inputs and
> > +      outputs; the inputs start at line 0, the outputs start at max_ngpios.
> 
> Most of this description is better suited to the ngpios property, which
> controls the number of lines that the gpiochip will expose.
> 
> Also, minor nit: max_ngpios -> max-ngpios.
> 
> How about something like:
> 
>   ngpios:
>     description:
>       Number of GPIO lines to expose. Since each HW GPIO is an input and an
>       output, we provide ngpios * 2 lines on our chip device. We also use it
>       to define the split between the inputs and outputs; the inputs start at
>       line 0, the outputs start at ngpios.
> 
>   max-ngpios:
>     description:
>       Represents the number of actual hardware-supported GPIOs (ie, slots within
>       the clocked serial GPIO data), and therefore the maximum value for
>       the ngpios property
> 

I will modify the description as you suggested.

> > +    minimum: 0
> > +    maximum: 128
> 
> Will this be the case for all (future) hardware? Can we leave this
> unbound?
> 

Since the future hardware may supports more gpios, I will remove it.

> Cheers,
> 
> 
> Jeremy
> 

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
  2021-05-27  2:34       ` Steven Lee
@ 2021-05-27  3:04         ` Andrew Jeffery
  -1 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  3:04 UTC (permalink / raw)
  To: Steven Lee
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

On Thu, 27 May 2021, at 12:04, Steven Lee wrote:
> The 05/27/2021 09:26, Andrew Jeffery wrote:
> > Hi Steven,
> > 
> > On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> > >  	unsigned long flags;
> > >  	enum aspeed_sgpio_reg reg;
> > >  	int rc = 0;
> > >  
> > >  	spin_lock_irqsave(&gpio->lock, flags);
> > >  
> > > -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> > > -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> > > +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> > > reg_rdata;
> > 
> > We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
> > 'const struct aspeed_sgpio *' parameter), rather than open-coding 
> > gpio->max_ngpios. This approach will make it easier to refactor the 
> > implementation in the future (if necessary).
> > 
> 
> I will rewrite the function to use aspeed_sgpio struct.


> > >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> > >  	void __iomem *addr_r, *addr_w;
> > >  	u32 reg = 0;
> > >  
> > > -	if (aspeed_sgpio_is_input(offset))
> > > +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
> > >  		return -EINVAL;
> > >  
> > >  	/* Since this is an output, read the cached value from rdata, then
> > > @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> > > unsigned int offset, int val)
> > >  	reg = ioread32(addr_r);
> > >  
> > >  	if (val)
> > > -		reg |= GPIO_BIT(offset);
> > > +		reg |= GPIO_BIT(offset % gpio->max_ngpios);
> > 
> > Pass gpio through GPIO_BIT() too.
> > 
> 
> I don't understand this comment.
> Could you describe more?

Oh, sure, what I was suggesting was to make a similar change to 
GPIO_BIT() as I suggested for aspeed_sgpio_is_input() above.

Hope that helps.

Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support
@ 2021-05-27  3:04         ` Andrew Jeffery
  0 siblings, 0 replies; 38+ messages in thread
From: Andrew Jeffery @ 2021-05-27  3:04 UTC (permalink / raw)
  To: Steven Lee
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai, Jeremy Kerr

On Thu, 27 May 2021, at 12:04, Steven Lee wrote:
> The 05/27/2021 09:26, Andrew Jeffery wrote:
> > Hi Steven,
> > 
> > On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> > >  	unsigned long flags;
> > >  	enum aspeed_sgpio_reg reg;
> > >  	int rc = 0;
> > >  
> > >  	spin_lock_irqsave(&gpio->lock, flags);
> > >  
> > > -	reg = aspeed_sgpio_is_input(offset) ? reg_val : reg_rdata;
> > > -	rc = !!(ioread32(bank_reg(gpio, bank, reg)) & GPIO_BIT(offset));
> > > +	reg = aspeed_sgpio_is_input(offset, gpio->max_ngpios) ? reg_val : 
> > > reg_rdata;
> > 
> > We should just pass gpio here (i.e. make aspeed_sgpio_is_input take a 
> > 'const struct aspeed_sgpio *' parameter), rather than open-coding 
> > gpio->max_ngpios. This approach will make it easier to refactor the 
> > implementation in the future (if necessary).
> > 
> 
> I will rewrite the function to use aspeed_sgpio struct.


> > >  	struct aspeed_sgpio *gpio = gpiochip_get_data(gc);
> > > -	const struct aspeed_sgpio_bank *bank = to_bank(offset);
> > >  	void __iomem *addr_r, *addr_w;
> > >  	u32 reg = 0;
> > >  
> > > -	if (aspeed_sgpio_is_input(offset))
> > > +	if (aspeed_sgpio_is_input(offset, gpio->max_ngpios))
> > >  		return -EINVAL;
> > >  
> > >  	/* Since this is an output, read the cached value from rdata, then
> > > @@ -209,9 +214,9 @@ static int sgpio_set_value(struct gpio_chip *gc, 
> > > unsigned int offset, int val)
> > >  	reg = ioread32(addr_r);
> > >  
> > >  	if (val)
> > > -		reg |= GPIO_BIT(offset);
> > > +		reg |= GPIO_BIT(offset % gpio->max_ngpios);
> > 
> > Pass gpio through GPIO_BIT() too.
> > 
> 
> I don't understand this comment.
> Could you describe more?

Oh, sure, what I was suggesting was to make a similar change to 
GPIO_BIT() as I suggested for aspeed_sgpio_is_input() above.

Hope that helps.

Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
  2021-05-27  1:27     ` Andrew Jeffery
@ 2021-05-27  4:01       ` Steven Lee
  -1 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  4:01 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai

The 05/27/2021 09:27, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
> > with 80 pins.
> 
> Is there any chance the serial GPIO controllers can be explicitly 
> listed in the Memory Space Allocation Table of the datasheet? Currently 
> they're covered by the entry for "GPIO Controller (Parallel GPIO)" 
> which is listed as ranging from 0x1e780000-0x1e7807ff.
> 

I've forwarded your suggestion to designers.
Per the discussion with designers, they may change the
GPIO controller description of Memory Space Allocation Table to
"GPIO Controller (including Parallel and Serial GPIO)".

> Admittedly the details are listed in chapter 41 for the GPIO 
> Controller, but it would be handy to not have to dig.
> 
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
> >  1 file changed, 32 insertions(+)
> > 
> > diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
> > index f96607b7b4e2..556ce9535c22 100644
> > --- a/arch/arm/boot/dts/aspeed-g6.dtsi
> > +++ b/arch/arm/boot/dts/aspeed-g6.dtsi
> > @@ -377,6 +377,38 @@
> >  				#interrupt-cells = <2>;
> >  			};
> >  
> > +			sgpiom0: sgpiom@1e780500 {
> > +				#gpio-cells = <2>;
> > +				gpio-controller;
> > +				compatible = "aspeed,ast2600-sgpiom";
> > +				reg = <0x1e780500 0x100>;
> > +				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
> > +				max-ngpios = <128>;
> 
> I need to think more about this one.
> 
> Andrew

^ permalink raw reply	[flat|nested] 38+ messages in thread

* Re: [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node.
@ 2021-05-27  4:01       ` Steven Lee
  0 siblings, 0 replies; 38+ messages in thread
From: Steven Lee @ 2021-05-27  4:01 UTC (permalink / raw)
  To: Andrew Jeffery
  Cc: Linus Walleij, Bartosz Golaszewski, Rob Herring, Joel Stanley,
	open list:GPIO SUBSYSTEM,
	open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS,
	moderated list:ARM/ASPEED MACHINE SUPPORT,
	moderated list:ARM/ASPEED MACHINE SUPPORT, open list,
	Hongwei Zhang, Ryan Chen, Billy Tsai

The 05/27/2021 09:27, Andrew Jeffery wrote:
> Hi Steven,
> 
> On Wed, 26 May 2021, at 19:16, Steven Lee wrote:
> > AST2600 supports 2 SGPIO master interfaces one with 128 pins another one
> > with 80 pins.
> 
> Is there any chance the serial GPIO controllers can be explicitly 
> listed in the Memory Space Allocation Table of the datasheet? Currently 
> they're covered by the entry for "GPIO Controller (Parallel GPIO)" 
> which is listed as ranging from 0x1e780000-0x1e7807ff.
> 

I've forwarded your suggestion to designers.
Per the discussion with designers, they may change the
GPIO controller description of Memory Space Allocation Table to
"GPIO Controller (including Parallel and Serial GPIO)".

> Admittedly the details are listed in chapter 41 for the GPIO 
> Controller, but it would be handy to not have to dig.
> 
> > 
> > Signed-off-by: Steven Lee <steven_lee@aspeedtech.com>
> > ---
> >  arch/arm/boot/dts/aspeed-g6.dtsi | 32 ++++++++++++++++++++++++++++++++
> >  1 file changed, 32 insertions(+)
> > 
> > diff --git a/arch/arm/boot/dts/aspeed-g6.dtsi b/arch/arm/boot/dts/aspeed-g6.dtsi
> > index f96607b7b4e2..556ce9535c22 100644
> > --- a/arch/arm/boot/dts/aspeed-g6.dtsi
> > +++ b/arch/arm/boot/dts/aspeed-g6.dtsi
> > @@ -377,6 +377,38 @@
> >  				#interrupt-cells = <2>;
> >  			};
> >  
> > +			sgpiom0: sgpiom@1e780500 {
> > +				#gpio-cells = <2>;
> > +				gpio-controller;
> > +				compatible = "aspeed,ast2600-sgpiom";
> > +				reg = <0x1e780500 0x100>;
> > +				interrupts = <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>;
> > +				max-ngpios = <128>;
> 
> I need to think more about this one.
> 
> Andrew

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

^ permalink raw reply	[flat|nested] 38+ messages in thread

end of thread, other threads:[~2021-05-27  4:05 UTC | newest]

Thread overview: 38+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-05-26  9:46 [PATCH v1 0/4] ASPEED sgpio driver enhancement Steven Lee
2021-05-26  9:46 ` Steven Lee
2021-05-26  9:46 ` [PATCH v1 1/4] dt-bindings: aspeed-sgpio: Convert txt bindings to yaml Steven Lee
2021-05-26  9:46   ` Steven Lee
2021-05-26 12:56   ` Rob Herring
2021-05-26 12:56     ` Rob Herring
2021-05-27  2:45     ` Steven Lee
2021-05-27  2:45       ` Steven Lee
2021-05-27  0:57   ` Andrew Jeffery
2021-05-27  0:57     ` Andrew Jeffery
2021-05-27  1:03     ` Steven Lee
2021-05-27  1:03       ` Steven Lee
2021-05-27  1:42   ` Jeremy Kerr
2021-05-27  1:42     ` Jeremy Kerr
2021-05-27  2:58     ` Steven Lee
2021-05-27  2:58       ` Steven Lee
2021-05-26  9:46 ` [PATCH v1 2/4] ARM: dts: aspeed-g6: Add SGPIO node Steven Lee
2021-05-26  9:46   ` Steven Lee
2021-05-27  1:27   ` Andrew Jeffery
2021-05-27  1:27     ` Andrew Jeffery
2021-05-27  4:01     ` Steven Lee
2021-05-27  4:01       ` Steven Lee
2021-05-26  9:46 ` [PATCH v1 3/4] ARM: dts: aspeed-g5: Modify sgpio node for the enhanced sgpio driver Steven Lee
2021-05-26  9:46   ` Steven Lee
2021-05-26  9:46 ` [PATCH v1 4/4] gpio: gpio-aspeed-sgpio: Add AST2600 sgpio support Steven Lee
2021-05-26  9:46   ` Steven Lee
2021-05-26 14:04   ` kernel test robot
2021-05-26 14:04     ` kernel test robot
2021-05-26 14:04     ` kernel test robot
2021-05-26 14:05   ` kernel test robot
2021-05-26 14:05     ` kernel test robot
2021-05-26 14:05     ` kernel test robot
2021-05-27  1:26   ` Andrew Jeffery
2021-05-27  1:26     ` Andrew Jeffery
2021-05-27  2:34     ` Steven Lee
2021-05-27  2:34       ` Steven Lee
2021-05-27  3:04       ` Andrew Jeffery
2021-05-27  3:04         ` Andrew Jeffery

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.