Add support for AST2700 INTC driver
Subsystems (3)
cover Add support for AST2700 INTC driver
Introduce to the AST27XX INTC modules, which contain two conponents in CPU die(12nm) and IO die(40mm) comunicating by SLI or LTPI protocol.
There are lots of device in IO die, which need to be serviced in requested interrupt handler. As two die ICs, combine 32 interrupt source in IO die into 1 interrupt in CPU die.
soc0_intc11 represent CPU die INTC, which each bit mapping to soc1_intcX. soc1_intcX represent IO die INTC, which combines 32 interrupt sources.
Kevin Chen (2):
dt-bindings: interrupt-controller: Add support for ASPEED AST27XX INTC irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX
platforms
.../aspeed,ast2700-intc.yaml | 120 +++++++++++ drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198 ++++++++++++++++++ 3 files changed, 319 insertions(+) create mode 100644 Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml create mode 100644 drivers/irqchip/irq-aspeed-intc.c
Signature · 1 line
2.34.1
PATCH 1/2 dt-bindings: interrupt-controller: Add support for ASPEED AST27XX INTC
The ASPEED AST27XX interrupt controller(INTC) combines 32 interrupt sources into 1 interrupt into GIC from CPU die to CPU die. The INTC design contains soc0_intc and soc1_intc module doing hand shake between CPU die and IO die INTC.
In soc0_intc11, each bit represent 1 GIC_SPI interrupt from soc1_intcX. In soc1_intcX, each bit represent 1 device interrupt in IO die.
By soc1_intcX in IO die, AST27XX INTC combines 32 interrupt sources to 1 interrupt source in soc0_intc11 in CPU die, which achieve the interrupt passing between the different die in AST27XX.
--- .../aspeed,ast2700-intc.yaml | 120 ++++++++++++++++++ 1 file changed, 120 insertions(+) create mode 100644 Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +120 -0
@@ -0,0 +1,120 @@ +# SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/interrupt-controller/aspeed,ast2700-intc.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Aspeed Interrupt Controller driver + +description: + These bindings are for the Aspeed interrupt controller. The AST2700 + SoC families include a legacy register layout before a re-designed + layout, but the bindings do not prescribe the use of one or the other. + +maintainers: + - Kevin Chen <[email protected]> + +allOf: + - $ref: /schemas/interrupt-controller.yaml# + +properties: + compatible: + oneOf: + - items: + - enum: + - aspeed,ast2700-intc-ic + - aspeed,ast2700-intc-icv2 + description: | + Use "aspeed,ast2700-intc-ic" for soc1 INTC in IO die + Use "aspeed,ast2700-intc-icv2" for soc0 INTC in CPU die + + interrupt-controller: true + + interrupts-extended: + minItems: 1 + maxItems: 3 + description: + Specifies which contexts are connected to the INTC, with "-1" specifying + that a context is not present. Each node pointed to should be a + aspeed,ast2700-intc-ic or aspeed,ast2700-intc-icv2 nodes which are pointed + to gic node. + + "#address-cells": + const: 2 + "#size-cells": + const: 2 + + '#interrupt-cells': + const: 2 + description: | + The first cell cell is the interrupt source IRQ number, and the second cell + is the trigger type as defined in interrupt.txt in this directory. + + reg: + minItems: 1 + maxItems: 2 + description: | + The first cell cell is the interrupt enable register, and the second cell + is the status register. + + ranges: true + + interrupts: + minItems: 1 + maxItems: 10 + description: | + Interrupt source of the CPU interrupts. In soc0_intc in CPU die INTC each bit + represent soc1_intc interrupt source. soc0_intc take care 10 interrupt source + from soc1_intc0~5 and ltpi0/1_soc1_intc0/1. + +required: + - compatible + - reg + - interrupt-controller + - '#interrupt-cells' + +additionalProperties: false + +example: + - | + soc0_intc: interrupt-controller at 12100000 { + compatible = "simple-mfd"; + reg = <0 0x12100000 0 0x4000>; + #address-cells = <2>; + #size-cells = <2>; + ranges = <0x0 0x0 0x0 0x12100000 0x0 0x4000>; + + soc0_intc11: interrupt-controller at 1b00 { + compatible = "aspeed,ast2700-intc-icv2"; + interrupts = <GIC_SPI 192 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 193 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 194 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 195 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 196 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 197 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 198 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 199 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 200 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 201 IRQ_TYPE_LEVEL_HIGH>; + #interrupt-cells = <2>; + interrupt-controller; + reg = <0x0 0x1b00 0x0 0x10>; + }; + }; + + - | + soc1_intc: interrupt-controller at 14c18000 { + compatible = "simple-mfd"; + reg = <0 0x14c18000 0 0x400>; + #address-cells = <2>; + #size-cells = <2>; + ranges = <0x0 0x0 0x0 0x14c18000 0x0 0x400>; + + soc1_intc0: interrupt-controller at 100 { + compatible = "aspeed,ast2700-intc-ic"; + interrupts-extended = <&soc0_intc11 0 IRQ_TYPE_LEVEL_HIGH>; + #interrupt-cells = <2>; + interrupt-controller; + reg = <0x0 0x100 0x0 0x10>; + }; + };
Signature · 1 line
2.34.1
Re: [PATCH v1 1/2] dt-bindings: interrupt-controller: Add support for ASPEED AST27XX INTC
Quoted 5 lines · Kevin Chen
The ASPEED AST27XX interrupt controller(INTC) combines 32 interrupt sources into 1 interrupt into GIC from CPU die to CPU die. The INTC design contains
soc0_intcandsoc1_intcmodule doing hand shake between CPU die and IO die INTC.
Quoted 7 lines
In
soc0_intc11, each bit represent 1GIC_SPIinterrupt fromsoc1_intcX. Insoc1_intcX, each bit represent 1 device interrupt in IO die.By
soc1_intcXin IO die, AST27XX INTC combines 32 interrupt sources to 1 interrupt source insoc0_intc11in CPU die, which achieve the interrupt passing between the different die in AST27XX.
---
This was never tested. Please do not send untested code.
It does not look like you tested the bindings, at least after quick look. Please run make dt_binding_check (see Documentation/devicetree/bindings/writing-schema.rst for instructions). Maybe you need to update your dtschema and yamllint.
Limited review follows due to lack of basic testing.
Quoted 10 lines
.../aspeed,ast2700-intc.yaml | 120 ++++++++++++++++++ 1 file changed, 120 insertions(+) create mode 100644 Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yamlDocumentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +0
-0
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 7
Quoted 7 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +7
-0@@ -0,0 +1,120 @@ +# SPDX-License-Identifier: GPL-2.0 OR BSD-2-Clause +%YAML 1.2 +--- +$id: http://devicetree.org/schemas/interrupt-controller/aspeed,ast2700-intc.yaml# +$schema: http://devicetree.org/meta-schemas/core.yaml# + +title: Aspeed Interrupt Controller driver
Drop driver, describe hardware. Aspeed or some specific SoC?
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 10
Quoted 3 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +3
-0+ +description: + These bindings are for the Aspeed interrupt controller. The AST2700
Drop first sentence. Pointless.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 12
Quoted 2 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +2
-0+ SoC families include a legacy register layout before a re-designed + layout, but the bindings do not prescribe the use of one or the other.
Entire description is pointless - you do not say anything valuable here. Describe this hardware instead.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 22
Quoted 10 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +10
-0+ +maintainers: + - Kevin Chen <[email protected]> + +allOf: + - $ref: /schemas/interrupt-controller.yaml# + +properties: + compatible: + oneOf:
Drop
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 23
Quoted 1 line
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +1
-0+ - items:
Drop
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 29
Quoted 6 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +6
-0+ - enum: + - aspeed,ast2700-intc-ic + - aspeed,ast2700-intc-icv2 + description: | + Use "aspeed,ast2700-intc-ic" for soc1 INTC in IO die + Use "aspeed,ast2700-intc-icv2" for soc0 INTC in CPU die
Use consistent naming. Isn't your other block called 0 and 1? Why using different namings?
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 33
Quoted 4 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +4
-0+ + interrupt-controller: true + + interrupts-extended:
interrupts instead.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 40
Quoted 7 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +7
-0+ minItems: 1 + maxItems: 3 + description: + Specifies which contexts are connected to the INTC, with "-1" specifying + that a context is not present. Each node pointed to should be a + aspeed,ast2700-intc-ic or aspeed,ast2700-intc-icv2 nodes which are pointed + to gic node.
Don't repeat constraints in free form text. Describe items instead.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 43
Quoted 3 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +3
-0+ + "#address-cells": + const: 2
Blank line
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 49
Quoted 6 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +6
-0+ "#size-cells": + const: 2 + + '#interrupt-cells': + const: 2 + description: |
Do not need '|' unless you need to preserve formatting.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 58
Quoted 9 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +9
-0+ The first cell cell is the interrupt source IRQ number, and the second cell + is the trigger type as defined in interrupt.txt in this directory. + + reg: + minItems: 1 + maxItems: 2 + description: | + The first cell cell is the interrupt enable register, and the second cell + is the status register.
List and describe the items instead.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 68
Quoted 10 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +10
-0+ + ranges: true + + interrupts: + minItems: 1 + maxItems: 10 + description: | + Interrupt source of the CPU interrupts. In soc0_intc in CPU die INTC each bit + represent soc1_intc interrupt source. soc0_intc take care 10 interrupt source + from soc1_intc0~5 and ltpi0/1_soc1_intc0/1.
No, you cannot have both. That's total mess. Anyway, standard comment applies - list and describe items.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 80
Quoted 12 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +12
-0+ +required: + - compatible + - reg + - interrupt-controller + - '#interrupt-cells' + +additionalProperties: false + +example: + - | + soc0_intc: interrupt-controller at 12100000 {
Drop label
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 81
Quoted 1 line
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +1
-0+ compatible = "simple-mfd";
No, it's not. Drop. Look how other bindings do it.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 85
Quoted 4 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +4
-0+ reg = <0 0x12100000 0 0x4000>; + #address-cells = <2>; + #size-cells = <2>; + ranges = <0x0 0x0 0x0 0x12100000 0x0 0x4000>;
Read DTS coding style.
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 87
Quoted 2 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +2
-0+ + soc0_intc11: interrupt-controller at 1b00 {
Drop label
Quoted 4 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +4
-0+ compatible = "aspeed,ast2700-intc-icv2"; + interrupts = <GIC_SPI 192 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 193 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 194 IRQ_TYPE_LEVEL_HIGH>,
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 101
Quoted 10 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +10
-0+ <GIC_SPI 195 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 196 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 197 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 198 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 199 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 200 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 201 IRQ_TYPE_LEVEL_HIGH>; + #interrupt-cells = <2>; + interrupt-controller; + reg = <0x0 0x1b00 0x0 0x10>;
DTS coding style
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml line 114
Quoted 13 lines
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml new file +13
-0+ }; + }; + + - | + soc1_intc: interrupt-controller at 14c18000 { + compatible = "simple-mfd"; + reg = <0 0x14c18000 0 0x400>; + #address-cells = <2>; + #size-cells = <2>; + ranges = <0x0 0x0 0x0 0x14c18000 0x0 0x400>; + + soc1_intc0: interrupt-controller at 100 { + compatible = "aspeed,ast2700-intc-ic";
Drop this example, almost no differences.
Best regards,
Krzysztof
Rob Herring (Arm)
maintainerRe: [PATCH v1 1/2] dt-bindings: interrupt-controller: Add support for ASPEED AST27XX INTC
In reply to Kevin ChenQuoted 8 lines · Kevin Chen
The ASPEED AST27XX interrupt controller(INTC) combines 32 interrupt sources into 1 interrupt into GIC from CPU die to CPU die. The INTC design contains
soc0_intcandsoc1_intcmodule doing hand shake between CPU die and IO die INTC.In
soc0_intc11, each bit represent 1GIC_SPIinterrupt fromsoc1_intcX. Insoc1_intcX, each bit represent 1 device interrupt in IO die.
Quoted 8 lines
By
soc1_intcXin IO die, AST27XX INTC combines 32 interrupt sources to 1 interrupt source insoc0_intc11in CPU die, which achieve the interrupt passing between the different die in AST27XX.--- .../aspeed,ast2700-intc.yaml | 120 ++++++++++++++++++ 1 file changed, 120 insertions(+) create mode 100644 Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml
My bot found errors running 'make dt_binding_check' on your patch:
yamllint warnings/errors:./Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml:24:9: [warning] wrong indentation: expected 10 but found 8 (indentation) ./Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml:25:11: [warning] wrong indentation: expected 12 but found 10 (indentation)
dtschema/dtc warnings/errors:
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml: 'example' is not one of ['$id', '$schema', 'title', 'description', 'examples', 'required', 'allOf', 'anyOf', 'oneOf', 'definitions', '$defs', 'additionalProperties', 'dependencies', 'dependentRequired', 'dependentSchemas', 'patternProperties', 'properties', 'not', 'if', 'then', 'else', 'unevaluatedProperties', 'deprecated', 'maintainers', 'select', '$ref']
from schema $id: http://devicetree.org/meta-schemas/base.yaml#
doc reference errors (make refcheckdocs):
See https://patchwork.ozlabs.org/project/devicetree-bindings/patch/20240813074338.969883-2-kevin_chen at aspeedtech.com
The base for the series is generally the latest rc1. A different dependency should be noted in *this* patch.
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 after running the above command yourself. Note that DT_SCHEMA_FILES can be set to your schema file to speed up checking your schema. However, it must be unset to test all examples with your schema.
RE: [PATCH v1 1/2] dt-bindings: interrupt-controller: Add support for ASPEED AST27XX INTC
Quoted 45 lines
Quoted 5 lines
The ASPEED AST27XX interrupt controller(INTC) combines 32 interrupts sources into 1 interrupt into GIC from CPU die to CPU die. The INTC design contains
soc0_intcandsoc1_intcmodule doing hand shake between CPU die and IO die INTC.Quoted 8 lines
In
soc0_intc11, each bit represent 1GIC_SPIinterrupt fromsoc1_intcX. Insoc1_intcX, each bit represent 1 device interrupt in IO die.By
soc1_intcXin IO die, AST27XX INTC combines 32 interrupt sources to 1 interrupt source insoc0_intc11in CPU die, which achieve the interrupt passing between the different die in AST27XX.--- .../aspeed,ast2700-intc.yaml | 120++++++++++++++++++
Quoted 5 lines
1 file changed, 120 insertions(+) create mode 100644
Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yamlMy bot found errors running 'make
dt_binding_check' on your patch:yamllint warnings/errors:
./Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml:24:9: [warning] wrong indentation: expected 10 but found 8 (indentation)./Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml:25:11: [warning] wrong indentation: expected 12 but found 10 (indentation)dtschema/dtc warnings/errors:
/builds/robherring/dt-review-ci/linux/Documentation/devicetree/bindings/interrupt-controller/aspeed,ast2700-intc.yaml: 'example' is not one of ['$id', '$schema', 'title', 'description', 'examples', 'required', 'allOf', 'anyOf', 'oneOf', 'definitions', '$defs', 'additionalProperties', 'dependencies', 'dependentRequired', 'dependentSchemas', 'patternProperties', 'properties', 'not', 'if', 'then', 'else', 'unevaluatedProperties', 'deprecated', 'maintainers', 'select', '$ref']
from schema $id: http://devicetree.org/meta-schemas/base.yaml#doc reference errors (make refcheckdocs):
See
https://patchwork.ozlabs.org/project/devicetree-bindings/patch/202408130743
38.969883-2-kevin_chen at aspeedtech.com
Quoted 8 lines
The base for the series is generally the latest rc1. A different dependency should be noted in *this* patch.
If you already ran 'make
dt_binding_check' and didn't see the above error(s), then make sure 'yamllint' is installed and dt-schema is up to date:pip3 install dtschema --upgrade
Agree.
Quoted 4 lines
Please check and re-submit after running the above command yourself. Note that
DT_SCHEMA_FILEScan be set to your schema file to speed up checking your schema. However, it must be unset to test all examples with your schema.
Agree.
I will fix all the warnings in ' make dt_binding_check '.
Thanks a lot.
PATCH 2/2 irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
In reply to Kevin ChenThere are 10 interrupt source of soc0_intc in CPU die INTC.
1. 6 interrupt sources in IO die of soc1_intc0~soc1_intc5.
2. 2 interrupt sources in LTPI of ltpi0_intc0 and ltpi0_intc1.
3. 2 interrupt sources in LTPI of ltpi1_intc0 and ltpi1_intc1.
Request GIC interrupt to check each bit in status register to do next level INTC handler.
In next level INTC handler of IO die or LTPI INTC using soc1_intcX combining 32 interrupt sources into soc0_intc11 in CPU die. In soc1_intcX, handler would check 32 bit of status register to do the requested device handler.
--- drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198 ++++++++++++++++++++++++++++++ 2 files changed, 199 insertions(+) create mode 100644 drivers/irqchip/irq-aspeed-intc.c
drivers/irqchip/Makefile +1 -0
@@ -84,6 +84,7 @@ obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o obj-$(CONFIG_LS_EXTIRQ) += irq-ls-extirq.o obj-$(CONFIG_LS_SCFG_MSI) += irq-ls-scfg-msi.o obj-$(CONFIG_ARCH_ASPEED) += irq-aspeed-vic.o irq-aspeed-i2c-ic.o irq-aspeed-scu-ic.o +obj-$(CONFIG_MACH_ASPEED_G7) += irq-aspeed-intc.o obj-$(CONFIG_STM32MP_EXTI) += irq-stm32mp-exti.o obj-$(CONFIG_STM32_EXTI) += irq-stm32-exti.o obj-$(CONFIG_QCOM_IRQ_COMBINER) += qcom-irq-combiner.o
drivers/irqchip/irq-aspeed-intc.c new file +198 -0
198 lines not on this page · raw message
Signature · 1 line
2.34.1
Re: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Quoted 22 lines · Kevin Chen
There are 10 interrupt source of
soc0_intcin CPU die INTC.
1. 6 interrupt sources in IO die of soc1_intc0~soc1_intc5.
2. 2 interrupt sources in LTPI ofltpi0_intc0andltpi0_intc1.
3. 2 interrupt sources in LTPI ofltpi1_intc0andltpi1_intc1.
Request GIC interrupt to check each bit in status register to do next level INTC handler.In next level INTC handler of IO die or LTPI INTC using
soc1_intcXcombining 32 interrupt sources intosoc0_intc11in CPU die. Insoc1_intcX, handler would check 32 bit of status register to do the requested device handler.--- drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198 ++++++++++++++++++++++++++++++ 2 files changed, 199 insertions(+) create mode 100644 drivers/irqchip/irq-aspeed-intc.cdrivers/irqchip/Makefile +0
-0
drivers/irqchip/Makefile obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o · line 87
Quoted 4 lines
drivers/irqchip/Makefile +1
-0@@ -84,6 +84,7 @@ obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o obj-$(CONFIG_LS_EXTIRQ) += irq-ls-extirq.o obj-$(CONFIG_LS_SCFG_MSI) += irq-ls-scfg-msi.o obj-$(CONFIG_ARCH_ASPEED) += irq-aspeed-vic.o irq-aspeed-i2c-ic.o irq-aspeed-scu-ic.o +obj-$(CONFIG_MACH_ASPEED_G7) += irq-aspeed-intc.o
There is no such thing as CONFIG_MACH_ASPEED_G7. And there will never be.
You already received feedback on this, so why do you keep pushing your solution? You did not respond to any feedback given, just send the same and the same till we agree?
NAK.
Quoted 6 lines
drivers/irqchip/Makefile +0
-0obj-$(CONFIG_STM32MP_EXTI) += irq-stm32mp-exti.o obj-$(CONFIG_STM32_EXTI) += irq-stm32-exti.o obj-$(CONFIG_QCOM_IRQ_COMBINER) += qcom-irq-combiner.odrivers/irqchip/irq-aspeed-intc.c new file +0
-0
...
Quoted 41 lines
drivers/irqchip/irq-aspeed-intc.c new file +41
-0+static int __init aspeed_intc_ic_of_init_v2 (struct device_node *node, +struct device_node *parent) +{ +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq, i; + + intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + intc_ic->irq_domain =irq_domain_add_linear (node,32 , + &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap; + } + +raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ; + +for (i =0 ; i <of_irq_count (node); i++) { + irq =irq_of_parse_and_map (node, i); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap; + }else { +irq_set_chained_handler_and_data (irq, aspeed_intc_ic_irq_handler, intc_ic); + }
drivers/irqchip/irq-aspeed-intc.c line 195
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ } + +return 0 ; + +err_iounmap: +iounmap (intc_ic->base); +err_free_ic: +kfree (intc_ic); +return ret; +}
Don't duplicate code. These are almost the same, so define one function which is then called by aspeed_intc_ic_of_init and aspeed_intc_ic_of_init_v2.
drivers/irqchip/irq-aspeed-intc.c line 198
Quoted 3 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+ +IRQCHIP_DECLARE (ast2700_intc_ic,"aspeed,ast2700-intc-1.0" , aspeed_intc_ic_of_init); +IRQCHIP_DECLARE (ast2700_intc_icv2,"aspeed,ast2700-intc-2.0" , aspeed_intc_ic_of_init_v2);
Best regards,
Krzysztof
Kevin Chen
NAK回覆: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Hi Krzk,
In ASPEED, ast2400/2500/2600 use arm architecture with KCONFIG_ARCH_ASPEED which slect MACH_ASPEED_G4/G5/G6 in arch/arm/mach-aspeed/Kconfig. In the fureture, there would be ast2800/2900/... using arm64. We need to clarify the IC generation between 7th/8th/9th/....
Maybe change ARCH_ASPEED/MACH_ASPEEDG7 to ARCH_ASPEED first. Or, do you have better Kconfig usage?
Review
Quoted 9 lines
(quoted diff) +9
-0+config ARCH_ASPEED + bool "Aspeed SoC family" + select MACH_ASPEED_G7 + help + Say yes if you intend to run on an Aspeed ast2700 or similar + seventh generation Aspeed BMCs. + +config MACH_ASPEED_G7 + bool "Aspeed SoC AST2700"
There are no MACHines for arm64. Look at this code. Do you see MACH anywhere else? No. Then why Aspeed must be different?
--
Best Regards,
Kevin. Chen
________________________________
???: Krzysztof Kozlowski <[email protected]>
????: 2024?8?13? ?? 04:50
???: Kevin Chen <[email protected]>; tglx at linutronix.de <[email protected]>; robh at kernel.org <[email protected]>; krzk+dt at kernel.org <[email protected]>; conor+dt at kernel.org <[email protected]>; joel at jms.id.au <[email protected]>; andrew at codeconstruct.com.au <[email protected]>; linux-kernel at vger.kernel.org <[email protected]>; devicetree at vger.kernel.org <[email protected]>; linux-arm-kernel at lists.infradead.org <[email protected]>; linux-aspeed at lists.ozlabs.org <[email protected]> ??: Re: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Quoted 22 lines · Kevin Chen
There are 10 interrupt source of
soc0_intcin CPU die INTC.
1. 6 interrupt sources in IO die of soc1_intc0~soc1_intc5.
2. 2 interrupt sources in LTPI ofltpi0_intc0andltpi0_intc1.
3. 2 interrupt sources in LTPI ofltpi1_intc0andltpi1_intc1.
Request GIC interrupt to check each bit in status register to do next level INTC handler.In next level INTC handler of IO die or LTPI INTC using
soc1_intcXcombining 32 interrupt sources intosoc0_intc11in CPU die. Insoc1_intcX, handler would check 32 bit of status register to do the requested device handler.--- drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198 ++++++++++++++++++++++++++++++ 2 files changed, 199 insertions(+) create mode 100644 drivers/irqchip/irq-aspeed-intc.cdrivers/irqchip/Makefile +0
-0
drivers/irqchip/Makefile obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o · line 87
Quoted 4 lines
drivers/irqchip/Makefile +1
-0@@ -84,6 +84,7 @@ obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o obj-$(CONFIG_LS_EXTIRQ) += irq-ls-extirq.o obj-$(CONFIG_LS_SCFG_MSI) += irq-ls-scfg-msi.o obj-$(CONFIG_ARCH_ASPEED) += irq-aspeed-vic.o irq-aspeed-i2c-ic.o irq-aspeed-scu-ic.o +obj-$(CONFIG_MACH_ASPEED_G7) += irq-aspeed-intc.o
There is no such thing as CONFIG_MACH_ASPEED_G7. And there will never be.
You already received feedback on this, so why do you keep pushing your solution? You did not respond to any feedback given, just send the same and the same till we agree?
NAK.
Quoted 6 lines
drivers/irqchip/Makefile +0
-0obj-$(CONFIG_STM32MP_EXTI) += irq-stm32mp-exti.o obj-$(CONFIG_STM32_EXTI) += irq-stm32-exti.o obj-$(CONFIG_QCOM_IRQ_COMBINER) += qcom-irq-combiner.odrivers/irqchip/irq-aspeed-intc.c new file +0
-0
...
Quoted 41 lines
drivers/irqchip/irq-aspeed-intc.c new file +41
-0+static int __init aspeed_intc_ic_of_init_v2 (struct device_node *node, +struct device_node *parent) +{ +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq, i; + + intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + intc_ic->irq_domain =irq_domain_add_linear (node,32 , + &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap; + } + +raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ; + +for (i =0 ; i <of_irq_count (node); i++) { + irq =irq_of_parse_and_map (node, i); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap; + }else { +irq_set_chained_handler_and_data (irq, aspeed_intc_ic_irq_handler, intc_ic); + }
drivers/irqchip/irq-aspeed-intc.c line 195
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ } + +return 0 ; + +err_iounmap: +iounmap (intc_ic->base); +err_free_ic: +kfree (intc_ic); +return ret; +}
Don't duplicate code. These are almost the same, so define one function which is then called by aspeed_intc_ic_of_init and aspeed_intc_ic_of_init_v2.
drivers/irqchip/irq-aspeed-intc.c line 198
Quoted 3 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+ +IRQCHIP_DECLARE (ast2700_intc_ic,"aspeed,ast2700-intc-1.0" , aspeed_intc_ic_of_init); +IRQCHIP_DECLARE (ast2700_intc_icv2,"aspeed,ast2700-intc-2.0" , aspeed_intc_ic_of_init_v2);
Best regards,
Krzysztof
************* Email Confidentiality Notice ********************
????:
???(????)????????????????? ???????????????????????????, ??????????????????????????????!
DISCLAIMER:
This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you. -------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/linux-aspeed/attachments/20240813/82333479/attachment-0001.htm>
Re: 回覆: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Quoted 7 lines · Kevin Chen
Hi Krzk,
In ASPEED, ast2400/2500/2600 use arm architecture with
KCONFIG_ARCH_ASPEEDwhich slect MACH_ASPEED_G4/G5/G6 inarch/arm/mach-aspeed/Kconfig. In the fureture, there would be ast2800/2900/... using arm64. We need to clarify the IC generation between 7th/8th/9th/....Maybe change ARCH_ASPEED/MACH_ASPEEDG7 to
ARCH_ASPEEDfirst. Or, do you have betterKconfigusage?
Fix your quotes and do not top-post.
Please respond inline, instead of top-posting, because it makes your emails hard to follow.
https://elixir.bootlin.com/linux/v6.8-rc7/source/Documentation/process/submitting-patches.rst#L340
Quoted 14 lines
Review
Quoted 11 lines
(quoted diff) +9
-0+config ARCH_ASPEED + bool "Aspeed SoC family" + select MACH_ASPEED_G7 + help + Say yes if you intend to run on an Aspeed ast2700 or similar + seventh generation Aspeed BMCs. + +config MACH_ASPEED_G7 + bool "Aspeed SoC AST2700"There are no MACHines for arm64. Look at this code. Do you see MACH anywhere else? No. Then why Aspeed must be different?
What is this?
Quoted 11 lines
--
Best Regards,
Kevin. Chen________________________________
???: Krzysztof Kozlowski <[email protected]>
????: 2024?8?13? ?? 04:50
???: Kevin Chen <[email protected]>; tglx at linutronix.de <[email protected]>; robh at kernel.org <[email protected]>; krzk+dt at kernel.org <[email protected]>; conor+dt at kernel.org <[email protected]>; joel at jms.id.au <[email protected]>; andrew at codeconstruct.com.au <[email protected]>; linux-kernel at vger.kernel.org <[email protected]>; devicetree at vger.kernel.org <[email protected]>; linux-arm-kernel at lists.infradead.org <[email protected]>; linux-aspeed at lists.ozlabs.org <[email protected]> ??: Re: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
...
Quoted 12 lines
drivers/irqchip/Makefile obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o · line 87
Quoted 9 lines
drivers/irqchip/Makefile +1
-0@@ -84,6 +84,7 @@ obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o obj-$(CONFIG_LS_EXTIRQ) += irq-ls-extirq.o obj-$(CONFIG_LS_SCFG_MSI) += irq-ls-scfg-msi.o obj-$(CONFIG_ARCH_ASPEED) += irq-aspeed-vic.o irq-aspeed-i2c-ic.o irq-aspeed-scu-ic.o +obj-$(CONFIG_MACH_ASPEED_G7) += irq-aspeed-intc.oThere is no such thing as
CONFIG_MACH_ASPEED_G7. And there will never be.
Quoted 5 lines
You already received feedback on this, so why do you keep pushing your solution? You did not respond to any feedback given, just send the same and the same till we agree?
NAK.
And this?
Quoted 8 lines
Quoted 7 lines
drivers/irqchip/Makefile +0
-0obj-$(CONFIG_STM32MP_EXTI) += irq-stm32mp-exti.o obj-$(CONFIG_STM32_EXTI) += irq-stm32-exti.o obj-$(CONFIG_QCOM_IRQ_COMBINER) += qcom-irq-combiner.odrivers/irqchip/irq-aspeed-intc.c new file +0
-0
...
Quoted 7 lines
************* Email Confidentiality Notice ********************
????:
???(????)????????????????? ???????????????????????????, ??????????????????????????????!DISCLAIMER:
This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you.
Maybe I am the intended recipient of your message, maybe not. I don't want to have any legal questions regarding upstream, public collaboration, thus probably I should just remove your messages.
Please talk with your IT that such disclaimers in open-source are not desired (and maybe even harmful).
If you do not understand why, please also see:
https://www.youtube.com/live/fMeH7wqOwXA?si=GY7igfbda6vnjXlJ&t=835
If you need to go around company SMTP server, then consider using b4
web-relay: https://b4.docs.kernel.org/en/latest/contributor/send.html
Please be informed that by responding to this email you agree that all communications from you and/or your company is made public. In other words, all messages originating from you and/or your company will be made public.
You already received exactly this feedback. Around three times. If you keep ignoring feedback, I will keep NAKing your patches.
Best regards,
Krzysztof
回覆: 回覆: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Hi Krzk,
Quoted 29 lines
Quoted 7 lines
Hi Krzk,
In ASPEED, ast2400/2500/2600 use arm architecture with
KCONFIG_ARCH_ASPEEDwhich slect MACH_ASPEED_G4/G5/G6 inarch/arm/mach-aspeed/Kconfig. In the fureture, there would be ast2800/2900/... using arm64. We need to clarify the IC generation between 7th/8th/9th/....Maybe change ARCH_ASPEED/MACH_ASPEEDG7 to
ARCH_ASPEEDfirst. Or, do you have betterKconfigusage?Fix your quotes and do not top-post.
Please respond inline, instead of top-posting, because it makes your emails hard to follow.
https://elixir.bootlin.com/linux/v6.8-rc7/source/Documentation/process/submitting-patches.rst#L340Quoted 14 lines
Review
Quoted 11 lines
(quoted diff) +9
-0+config ARCH_ASPEED + bool "Aspeed SoC family" + select MACH_ASPEED_G7 + help + Say yes if you intend to run on an Aspeed ast2700 or similar + seventh generation Aspeed BMCs. + +config MACH_ASPEED_G7 + bool "Aspeed SoC AST2700"There are no MACHines for arm64. Look at this code. Do you see MACH anywhere else? No. Then why Aspeed must be different?
Quoted 1 line
What is this?
OK. I will remove MACH_ASPEED_G7 and save ARCH_ASPEED for use.
--
Best Regards,
Kevin. Chen
________________________________
???: Krzysztof Kozlowski <[email protected]>
????: 2024?8?13? ?? 05:48
???: Kevin Chen <[email protected]>; tglx at linutronix.de <[email protected]>; robh at kernel.org <[email protected]>; krzk+dt at kernel.org <[email protected]>; conor+dt at kernel.org <[email protected]>; joel at jms.id.au <[email protected]>; andrew at codeconstruct.com.au <[email protected]>; linux-kernel at vger.kernel.org <[email protected]>; devicetree at vger.kernel.org <[email protected]>; linux-arm-kernel at lists.infradead.org <[email protected]>; linux-aspeed at lists.ozlabs.org <[email protected]> ??: Re: ??: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Quoted 7 lines · Kevin Chen
Hi Krzk,
In ASPEED, ast2400/2500/2600 use arm architecture with
KCONFIG_ARCH_ASPEEDwhich slect MACH_ASPEED_G4/G5/G6 inarch/arm/mach-aspeed/Kconfig. In the fureture, there would be ast2800/2900/... using arm64. We need to clarify the IC generation between 7th/8th/9th/....Maybe change ARCH_ASPEED/MACH_ASPEEDG7 to
ARCH_ASPEEDfirst. Or, do you have betterKconfigusage?
Fix your quotes and do not top-post.
Please respond inline, instead of top-posting, because it makes your emails hard to follow.
https://elixir.bootlin.com/linux/v6.8-rc7/source/Documentation/process/submitting-patches.rst#L340
Quoted 14 lines
Review
Quoted 11 lines
(quoted diff) +9
-0+config ARCH_ASPEED + bool "Aspeed SoC family" + select MACH_ASPEED_G7 + help + Say yes if you intend to run on an Aspeed ast2700 or similar + seventh generation Aspeed BMCs. + +config MACH_ASPEED_G7 + bool "Aspeed SoC AST2700"There are no MACHines for arm64. Look at this code. Do you see MACH anywhere else? No. Then why Aspeed must be different?
What is this?
Quoted 11 lines
--
Best Regards,
Kevin. Chen________________________________
???: Krzysztof Kozlowski <[email protected]>
????: 2024?8?13? ?? 04:50
???: Kevin Chen <[email protected]>; tglx at linutronix.de <[email protected]>; robh at kernel.org <[email protected]>; krzk+dt at kernel.org <[email protected]>; conor+dt at kernel.org <[email protected]>; joel at jms.id.au <[email protected]>; andrew at codeconstruct.com.au <[email protected]>; linux-kernel at vger.kernel.org <[email protected]>; devicetree at vger.kernel.org <[email protected]>; linux-arm-kernel at lists.infradead.org <[email protected]>; linux-aspeed at lists.ozlabs.org <[email protected]> ??: Re: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
...
Quoted 12 lines
drivers/irqchip/Makefile obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o · line 87
Quoted 9 lines
drivers/irqchip/Makefile +1
-0@@ -84,6 +84,7 @@ obj-$(CONFIG_MVEBU_SEI) += irq-mvebu-sei.o obj-$(CONFIG_LS_EXTIRQ) += irq-ls-extirq.o obj-$(CONFIG_LS_SCFG_MSI) += irq-ls-scfg-msi.o obj-$(CONFIG_ARCH_ASPEED) += irq-aspeed-vic.o irq-aspeed-i2c-ic.o irq-aspeed-scu-ic.o +obj-$(CONFIG_MACH_ASPEED_G7) += irq-aspeed-intc.oThere is no such thing as
CONFIG_MACH_ASPEED_G7. And there will never be.
Quoted 5 lines
You already received feedback on this, so why do you keep pushing your solution? You did not respond to any feedback given, just send the same and the same till we agree?
NAK.
And this?
Quoted 8 lines
Quoted 7 lines
drivers/irqchip/Makefile +0
-0obj-$(CONFIG_STM32MP_EXTI) += irq-stm32mp-exti.o obj-$(CONFIG_STM32_EXTI) += irq-stm32-exti.o obj-$(CONFIG_QCOM_IRQ_COMBINER) += qcom-irq-combiner.odrivers/irqchip/irq-aspeed-intc.c new file +0
-0
...
Quoted 7 lines
************* Email Confidentiality Notice ********************
????:
???(????)????????????????? ???????????????????????????, ??????????????????????????????!DISCLAIMER:
This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you.
Maybe I am the intended recipient of your message, maybe not. I don't want to have any legal questions regarding upstream, public collaboration, thus probably I should just remove your messages.
Please talk with your IT that such disclaimers in open-source are not desired (and maybe even harmful).
If you do not understand why, please also see:
https://www.youtube.com/live/fMeH7wqOwXA?si=GY7igfbda6vnjXlJ&t=835
If you need to go around company SMTP server, then consider using b4
web-relay: https://b4.docs.kernel.org/en/latest/contributor/send.html
Please be informed that by responding to this email you agree that all communications from you and/or your company is made public. In other words, all messages originating from you and/or your company will be made public.
You already received exactly this feedback. Around three times. If you keep ignoring feedback, I will keep NAKing your patches.
Best regards,
Krzysztof
************* Email Confidentiality Notice ********************
????:
???(????)????????????????? ???????????????????????????, ??????????????????????????????!
DISCLAIMER:
This message (and any attachments) may contain legally privileged and/or other confidential information. If you have received it in error, please notify the sender by reply e-mail and immediately delete the e-mail and any attachments without copying or disclosing the contents. Thank you. -------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ozlabs.org/pipermail/linux-aspeed/attachments/20240813/a51d2c53/attachment-0001.htm>
Re: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
In reply to Kevin ChenQuoted 11 lines · Kevin Chen
There are 10 interrupt source of
soc0_intcin CPU die INTC.
1. 6 interrupt sources in IO die of soc1_intc0~soc1_intc5.
2. 2 interrupt sources in LTPI ofltpi0_intc0andltpi0_intc1.
3. 2 interrupt sources in LTPI ofltpi1_intc0andltpi1_intc1.
Request GIC interrupt to check each bit in status register to do next level INTC handler.In next level INTC handler of IO die or LTPI INTC using
soc1_intcXcombining 32 interrupt sources intosoc0_intc11in CPU die. Insoc1_intcX, handler would check 32 bit of status register to do the requested device handler.
I can't figure out what this word salad is trying to tell me. Nothing in the code does any combining. The handler reads the very same INTC_INT_STATUS_REG.
Quoted 1 line
This lacks a Signed-off-by: tag. See Documentation/process/
Quoted 13 lines
--- drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198 ++++++++++++++++++++++++++++++drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ +#define INTC_INT_ENABLE_REG 0x00 +#define INTC_INT_STATUS_REG 0x04 + +struct aspeed_intc_ic { +void __iomem *base; +raw_spinlock_t gic_lock; +raw_spinlock_t intc_lock; +struct irq_domain *irq_domain; +};
drivers/irqchip/irq-aspeed-intc.c line 36
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ +static void aspeed_intc_ic_irq_handler (struct irq_desc *desc) +{ +struct aspeed_intc_ic *intc_ic =irq_desc_get_handler_data (desc); +struct irq_chip *chip =irq_desc_get_chip (desc); +unsigned long bit, status, flags; + +chained_irq_enter (chip, desc); + +raw_spin_lock_irqsave (&intc_ic->gic_lock, flags);
There is no point for irqsave(). This code is invoked with interrupts disabled and please convert to:
scoped_guard(raw_spinlock, &intc_ic->gic_lock) {drivers/irqchip/irq-aspeed-intc.c line 38
Quoted 2 lines
drivers/irqchip/irq-aspeed-intc.c new file +2
-0+ status =readl (intc_ic->base +INTC_INT_STATUS_REG ); +for_each_set_bit (bit, &status,32 ) {
Please use a define and not a hardcoded number.
drivers/irqchip/irq-aspeed-intc.c line 41
Quoted 3 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+generic_handle_domain_irq (intc_ic->irq_domain, bit); +writel (BIT (bit), intc_ic->base +INTC_INT_STATUS_REG ); + }
}
drivers/irqchip/irq-aspeed-intc.c line 51
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+raw_spin_unlock_irqrestore (&intc_ic->gic_lock, flags); + +chained_irq_exit (chip, desc); +} + +static void aspeed_intc_irq_mask (struct irq_data *data) +{ +struct aspeed_intc_ic *intc_ic =irq_data_get_irq_chip_data (data); +unsigned int mask =readl (intc_ic->base +INTC_INT_ENABLE_REG ) & ~BIT (data->hwirq); +unsigned long flags;
Invoked with interrupts disabled too.
drivers/irqchip/irq-aspeed-intc.c line 55
Quoted 3 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+raw_spin_lock_irqsave (&intc_ic->intc_lock, flags); +writel (mask, intc_ic->base +INTC_INT_ENABLE_REG ); +raw_spin_unlock_irqrestore (&intc_ic->intc_lock, flags);
guard(raw_spinlock)(&intc_ic->intc_lock); writel(mask, intc_ic->base + INTC_INT_ENABLE_REG);
drivers/irqchip/irq-aspeed-intc.c line 62
Quoted 7 lines
drivers/irqchip/irq-aspeed-intc.c new file +7
-0+} + +static void aspeed_intc_irq_unmask (struct irq_data *data) +{ +struct aspeed_intc_ic *intc_ic =irq_data_get_irq_chip_data (data); +unsigned int unmask =readl (intc_ic->base +INTC_INT_ENABLE_REG ) |BIT (data->hwirq); +unsigned long flags;
Ditto.
drivers/irqchip/irq-aspeed-intc.c line 73
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+raw_spin_lock_irqsave (&intc_ic->intc_lock, flags); +writel (unmask, intc_ic->base +INTC_INT_ENABLE_REG ); +raw_spin_unlock_irqrestore (&intc_ic->intc_lock, flags); +} + +static int aspeed_intc_irq_set_affinity (struct irq_data *data,const struct cpumask *dest, +bool force) +{ +return -EINVAL ; +}
No point for this stub, just leave irq_set_affinity uninitialized. The core code checks that pointer for NULL. Aside of that this stub and the assignment would need a #ifdef CONFIG_SMP guard.
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c new file +8
-0+static struct irq_chip aspeed_intc_chip = { + .name ="ASPEED INTC" , + .irq_mask = aspeed_intc_irq_mask, + .irq_unmask = aspeed_intc_irq_unmask, + .irq_set_affinity = aspeed_intc_irq_set_affinity, +}; + +static int aspeed_intc_ic_map_irq_domain(struct irq_domain *domain, unsigned int irq,
drivers/irqchip/irq-aspeed-intc.c line 92
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ irq_hw_number_t hwirq) +{ +irq_set_chip_and_handler (irq, &aspeed_intc_chip, handle_level_irq); +irq_set_chip_data (irq, domain->host_data); + +return 0 ; +} + +static const struct irq_domain_ops aspeed_intc_ic_irq_domain_ops = { + .map = aspeed_intc_ic_map_irq_domain,
.map = aspeed_intc_ic_map_irq_domain,
drivers/irqchip/irq-aspeed-intc.c line 99
Quoted 7 lines
drivers/irqchip/irq-aspeed-intc.c new file +7
-0+}; + +static int __init aspeed_intc_ic_of_init (struct device_node *node,struct device_node *parent) +{ +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq;
int irq, ret;
No point in initializing ret.
Quoted 21 lines
drivers/irqchip/irq-aspeed-intc.c new file +21
-0+ intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + irq =irq_of_parse_and_map (node,0 ); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap; + } + + intc_ic->irq_domain = irq_domain_add_linear(node, 32,
drivers/irqchip/irq-aspeed-intc.c line 131
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap; + } + +raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ;
See above.
Quoted 33 lines
drivers/irqchip/irq-aspeed-intc.c new file +33
-0+irq_set_chained_handler_and_data (irq, + aspeed_intc_ic_irq_handler, intc_ic); + +return 0 ; + +err_iounmap: +iounmap (intc_ic->base); +err_free_ic: +kfree (intc_ic); +return ret; +} + +static int __init aspeed_intc_ic_of_init_v2 (struct device_node *node, +struct device_node *parent) +{ +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq, i; + + intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + intc_ic->irq_domain = irq_domain_add_linear(node, 32,
drivers/irqchip/irq-aspeed-intc.c line 175
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap; + } + +raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ;
So up to this point aspeed_intc_ic_of_init_v2() is a verbatim copy of aspeed_intc_ic_of_init(). Why can't you reuse that function? It's not rocket science to make that work.
drivers/irqchip/irq-aspeed-intc.c line 182
Quoted 6 lines
drivers/irqchip/irq-aspeed-intc.c new file +6
-0+for (i =0 ; i <of_irq_count (node); i++) { + irq =irq_of_parse_and_map (node, i); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap;
Assume #0 and #1 succeed. #2 fails and leaves the chained handlers and the irqdomain around, but then unmaps the base and frees the data which the handler and the domain code needs. Seriously?
drivers/irqchip/irq-aspeed-intc.c line 183
Quoted 1 line
drivers/irqchip/irq-aspeed-intc.c new file +1
-0+ }else {
Pointless else as the if clause terminates with a goto.
drivers/irqchip/irq-aspeed-intc.c line 184
Quoted 1 line
drivers/irqchip/irq-aspeed-intc.c new file +1
-0+irq_set_chained_handler_and_data (irq, aspeed_intc_ic_irq_handler, intc_ic);
So if I understand the code correctly then the hardware coalesces the pending bits into a single status register, but depending on which part of the SoC raised the interrupt one of the demultiplex interrupts is raised in the GIC.
Any of those demultiplex interrupt handles _all_ pending bits and therefore you need gic_lock to serialize them, right?
Thanks,
tglx
RE: [PATCH v1 2/2] irqchip/aspeed-intc: Add support for 10 INTC interrupts on AST27XX platforms
Quoted 7 lines
Quoted 7 lines
There are 10 interrupt sources of
soc0_intcin CPU die INTC.
1. 6 interrupt sources in IO die of soc1_intc0~soc1_intc5.
2. 2 interrupt sources in LTPI ofltpi0_intc0andltpi0_intc1.
3. 2 interrupt sources in LTPI ofltpi1_intc0andltpi1_intc1.
Request GIC interrupt to check each bit in status register to do next level INTC handler.
Quoted 8 lines
Quoted 5 lines
In next level INTC handler of IO die or LTPI INTC using
soc1_intcXcombining
32 interrupt sources intosoc0_intc11in CPU die. Insoc1_intcX, handler would check 32 bit of status register to do the requested device handler.I can't figure out what this word salad is trying to tell me. Nothing in the code does any combining. The handler reads the very same
INTC_INT_STATUS_REG.
According to AST2700 datasheet, there are two kinds of interrupt controller with enable and raw status registers for use.
1. INTC0 is used to assert GIC(#192~#197) if interrupt in INTC1 asserted. There are 6 GIC interrupt number(#192~#197) used in one INTC0.
2. INTC1 is used to assert INTC0 if interrupt of modules asserted. There are 32 module interrupts used in one INTC1.
+------+ +---------+ +-----------+ ---module0
| GIC | -----|INTC0 | ---+----| INTC1_0|---module2...
+------+ +---------+ | +-----------+---module31
|
| +-----------+---module0
+-----| INTC1_0|---module2...
| +-----------+---module31
...
| +-----------+---module0
+-----| INTC1_5|---module2...
| +-----------+---module31Quoted 32 lines
Quoted 2 lines
This lacks a Signed-off-by: tag. See
Documentation/process/Quoted 13 lines
--- drivers/irqchip/Makefile | 1 + drivers/irqchip/irq-aspeed-intc.c | 198++++++++++++++++++++++++++++++
drivers/irqchip/irq-aspeed-intc.c new file +9
-0+ +#define INTC_INT_ENABLE_REG 0x00 +#define INTC_INT_STATUS_REG 0x04 + +struct aspeed_intc_ic { +void __iomem *base; +raw_spinlock_t gic_lock; +raw_spinlock_t intc_lock; +struct irq_domain *irq_domain;drivers/irqchip/irq-aspeed-intc.c line 36
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+}; + +static void aspeed_intc_ic_irq_handler (struct irq_desc *desc) { +struct aspeed_intc_ic *intc_ic =irq_desc_get_handler_data (desc); +struct irq_chip *chip =irq_desc_get_chip (desc); +unsigned long bit, status, flags; + +chained_irq_enter (chip, desc); + +raw_spin_lock_irqsave (&intc_ic->gic_lock, flags);There is no point for
irqsave(). This code is invoked with interrupts disabled and please convert to:
Quoted 1 line
scoped_guard(raw_spinlock, &intc_ic->gic_lock) {Agree.
Quoted 5 lines
drivers/irqchip/irq-aspeed-intc.c line 38
Quoted 3 lines
drivers/irqchip/irq-aspeed-intc.c new file +2
-0+ status =readl (intc_ic->base +INTC_INT_STATUS_REG ); +for_each_set_bit (bit, &status,32 ) {Please use a define and not a hardcoded number.
Agree.
Quoted 16 lines
drivers/irqchip/irq-aspeed-intc.c line 41
Quoted 4 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+generic_handle_domain_irq (intc_ic->irq_domain, bit); +writel (BIT (bit), intc_ic->base +INTC_INT_STATUS_REG ); + }}drivers/irqchip/irq-aspeed-intc.c
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c new file +8
-0+raw_spin_unlock_irqrestore (&intc_ic->gic_lock, flags); + +chained_irq_exit (chip, desc); +} + +static void aspeed_intc_irq_mask (struct irq_data *data) { +struct aspeed_intc_ic *intc_ic =irq_data_get_irq_chip_data (data); + unsigned int mask = readl(intc_ic->base + INTC_INT_ENABLE_REG) &~BIT(data->hwirq);
Quoted 3 lines
Review
Quoted 1 line
(quoted diff) +1
-0+ unsigned long flags;Invoked with interrupts disabled too.
Agree.
Quoted 6 lines
drivers/irqchip/irq-aspeed-intc.c line 55
Quoted 4 lines
drivers/irqchip/irq-aspeed-intc.c new file +3
-0+raw_spin_lock_irqsave (&intc_ic->intc_lock, flags); +writel (mask, intc_ic->base +INTC_INT_ENABLE_REG ); +raw_spin_unlock_irqrestore (&intc_ic->intc_lock, flags);guard(raw_spinlock)(&intc_ic->intc_lock);
Agree.
Quoted 11 lines
writel(mask, intc_ic->base + INTC_INT_ENABLE_REG);drivers/irqchip/irq-aspeed-intc.c
Quoted 5 lines
drivers/irqchip/irq-aspeed-intc.c new file +5
-0+} + +static void aspeed_intc_irq_unmask (struct irq_data *data) { +struct aspeed_intc_ic *intc_ic =irq_data_get_irq_chip_data (data); + unsigned int unmask = readl(intc_ic->base + INTC_INT_ENABLE_REG) |
BIT(data->hwirq);Review
Quoted 1 line
(quoted diff) +1
-0+ unsigned long flags;Ditto.
Agree.
Quoted 7 lines
drivers/irqchip/irq-aspeed-intc.c
Quoted 6 lines
drivers/irqchip/irq-aspeed-intc.c new file +5
-0+raw_spin_lock_irqsave (&intc_ic->intc_lock, flags); +writel (unmask, intc_ic->base +INTC_INT_ENABLE_REG ); +raw_spin_unlock_irqrestore (&intc_ic->intc_lock, flags); } + +static int aspeed_intc_irq_set_affinity(struct irq_data *data, const structcpumask *dest,
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c line 73
Quoted 4 lines
drivers/irqchip/irq-aspeed-intc.c new file +4
-0+ bool force) +{ +return -EINVAL ; +}No point for this stub, just leave
irq_set_affinityuninitialized. The core code checks that pointer for NULL. Aside of that this stub and the assignment would need a #ifdefCONFIG_SMPguard.
Agree.
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c
Quoted 9 lines
drivers/irqchip/irq-aspeed-intc.c new file +8
-0+static struct irq_chip aspeed_intc_chip = { + .name ="ASPEED INTC" , + .irq_mask = aspeed_intc_irq_mask, + .irq_unmask = aspeed_intc_irq_unmask, + .irq_set_affinity = aspeed_intc_irq_set_affinity, +}; + +static int aspeed_intc_ic_map_irq_domain(struct irq_domain *domain,unsigned int irq,
Quoted 12 lines
drivers/irqchip/irq-aspeed-intc.c line 92
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c new file +10
-0+ irq_hw_number_t hwirq) +{ +irq_set_chip_and_handler (irq, &aspeed_intc_chip, handle_level_irq); +irq_set_chip_data (irq, domain->host_data); + +return 0 ; +} + +static const struct irq_domain_ops aspeed_intc_ic_irq_domain_ops = { + .map = aspeed_intc_ic_map_irq_domain,.map =
aspeed_intc_ic_map_irq_domain,
Agree.
Quoted 10 lines
drivers/irqchip/irq-aspeed-intc.c line 99
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c new file +7
-0+}; + +static int __init aspeed_intc_ic_of_init (struct device_node *node, +struct device_node *parent) { +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq;int irq, ret;
Agree.
Quoted 2 lines
No point in initializing ret.
Agree.
Quoted 26 lines
Quoted 26 lines
drivers/irqchip/irq-aspeed-intc.c new file +25
-0+ intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + irq =irq_of_parse_and_map (node,0 ); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap; + } + + intc_ic->irq_domain =irq_domain_add_linear (node,32 , + &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap;
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c line 131
Quoted 6 lines
drivers/irqchip/irq-aspeed-intc.c new file +6
-0+ } + +raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ;See above.
Do you mean the name of "ASPEED INTC" would be covered by "aspeed-intc-doman"?
Quoted 40 lines
Quoted 40 lines
drivers/irqchip/irq-aspeed-intc.c new file +39
-0+irq_set_chained_handler_and_data (irq, + aspeed_intc_ic_irq_handler, intc_ic); + +return 0 ; + +err_iounmap: +iounmap (intc_ic->base); +err_free_ic: +kfree (intc_ic); +return ret; +} + +static int __init aspeed_intc_ic_of_init_v2 (struct device_node *node, +struct device_node *parent) +{ +struct aspeed_intc_ic *intc_ic; +int ret =0 ; +int irq, i; + + intc_ic =kzalloc (sizeof (*intc_ic),GFP_KERNEL ); +if (!intc_ic) +return -ENOMEM ; + + intc_ic->base =of_iomap (node,0 ); +if (!intc_ic->base) { +pr_err ("Failed to iomap intc_ic base\n" ); + ret = -ENOMEM ; + goto err_free_ic; + } +writel (0xffffffff , intc_ic->base +INTC_INT_STATUS_REG ); +writel (0x0 , intc_ic->base +INTC_INT_ENABLE_REG ); + + intc_ic->irq_domain =irq_domain_add_linear (node,32 , + &aspeed_intc_ic_irq_domain_ops, intc_ic); +if (!intc_ic->irq_domain) { + ret = -ENOMEM ; + goto err_iounmap; + } +
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c line 131
Quoted 4 lines
drivers/irqchip/irq-aspeed-intc.c new file +4
-0+raw_spin_lock_init (&intc_ic->gic_lock); +raw_spin_lock_init (&intc_ic->intc_lock); + + intc_ic->irq_domain->name ="aspeed-intc-domain" ;So up to this point
aspeed_intc_ic_of_init_v2()is a verbatim copy ofaspeed_intc_ic_of_init(). Why can't you reuse that function? It's not rocket science to make that work.
Agree.
Quoted 11 lines
drivers/irqchip/irq-aspeed-intc.c line 182
Quoted 7 lines
drivers/irqchip/irq-aspeed-intc.c new file +6
-0+for (i =0 ; i <of_irq_count (node); i++) { + irq =irq_of_parse_and_map (node, i); +if (!irq) { +pr_err ("Failed to get irq number\n" ); + ret = -EINVAL ; + goto err_iounmap;Assume #0 and #1 succeed. #2 fails and leaves the chained handlers and the irqdomain around, but then unmaps the base and frees the data which the handler and the domain code needs. Seriously?
So, do you recommend moving check irq out of for loop? And, irq_set_chained_hanlder_and_data in another for loop?
Quoted 4 lines
drivers/irqchip/irq-aspeed-intc.c line 183
Quoted 2 lines
drivers/irqchip/irq-aspeed-intc.c new file +1
-0+ }else {Pointless else as the if clause terminates with a goto.
Agree. I will remove the else
Quoted 8 lines
drivers/irqchip/irq-aspeed-intc.c line 133
Quoted 2 lines
drivers/irqchip/irq-aspeed-intc.c new file +1
-0+ irq_set_chained_handler_and_data(irq,
aspeed_intc_ic_irq_handler,Quoted 1 line
+intc_ic);
So if I understand the code correctly then the hardware coalesces the pending bits into a single status register, but depending on which part of the SoC raised the interrupt one of the demultiplex interrupts is raised in the GIC.
Yes.
Quoted 3 lines
Any of those demultiplex interrupt handles _all_ pending bits and therefore you need
gic_lockto serialize them, right?
Yes.
Quoted 4 lines
Thanks,
tglx
Thanks a lot for your review.