Skip to contentSkip to conversation

ARM: shmobile: r8a73a4 SoC and APE6EVM board support

PATCH New Magnus Damm 3 participants 23 messages Mar 12, 2013 – Mar 22, 2013

Subsystems (6)
lore
Share & tools
lore

Magnus Damm

author maintainer
#0 lore raw

cover ARM: shmobile: r8a73a4 SoC and APE6EVM board support

ARM: shmobile: r8a73a4 SoC and APE6EVM board support

[PATCH 01/04] ARM: shmobile: Initial r8a73a4 SoC support [PATCH 02/04] ARM: shmobile: r8a73a4 SCIF support [PATCH 03/04] ARM: shmobile: r8a73a4 IRQC support [PATCH 04/04] ARM: shmobile: APE6EVM support

This series adds initial support for r8a73a4 and APE6EVM. The SoC is known as r8a73a4 which is built around Cortex-A15 and makes use of SCIF for serial and IRQC interrupt controller for external IRQs.

This initial SoC and board support is providing limited DT-only support and a very basic C-version of board support. In the future improved PINCTRL and clock framework support will be added, followed by SMP support.

---

Written against all+next in renesas.git.
[PATCH 03/04] requires the IRQC driver to build

 arch/arm/boot/dts/Makefile                    |    1 
 arch/arm/boot/dts/r8a73a4-ape6evm.dts         |   26 ++
 arch/arm/boot/dts/r8a73a4.dtsi                |   87 ++++++++
 arch/arm/mach-shmobile/Kconfig                |   16 +
 arch/arm/mach-shmobile/Makefile               |    2 
 arch/arm/mach-shmobile/board-ape6evm.c        |   40 +++
 arch/arm/mach-shmobile/clock-r8a73a4.c        |  104 +++++++++
 arch/arm/mach-shmobile/include/mach/r8a73a4.h |    7 
 arch/arm/mach-shmobile/setup-r8a73a4.c        |  268 +++++++++++++++++++++++++
 9 files changed, 549 insertions(+), 2 deletions(-)
author maintainer
#1 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

From: Magnus Damm <[email protected]>

Add initial support for the r8a73a4 SoC including:
- Single Cortex-A15 CPU Core
- GIC
- Architecture timer

No static virtual mappings are used, all the components make use of ioremap().

---

 arch/arm/boot/dts/r8a73a4.dtsi                |   55 +++++++++++++++
 arch/arm/mach-shmobile/Kconfig                |    7 +
 arch/arm/mach-shmobile/Makefile               |    1 
 arch/arm/mach-shmobile/clock-r8a73a4.c        |   89 +++++++++++++++++++++++++
 arch/arm/mach-shmobile/include/mach/r8a73a4.h |    7 +
 arch/arm/mach-shmobile/setup-r8a73a4.c        |   56 +++++++++++++++
 6 files changed, 215 insertions(+)
arch/arm/boot/dts/r8a73a4.dtsi new file +55 -0
@@ -0,0 +1,55 @@
+/*
+ * Device Tree Source for the r8a73a4 SoC
+ *
+ * Copyright (C) 2013 Renesas Solutions Corp.
+ * Copyright (C) 2013 Magnus Damm
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2.  This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+/include/ "skeleton.dtsi"
+
+/ {
+	compatible = "renesas,r8a73a4";
+	interrupt-parent = <&gic>;
+
+	cpus {
+		#address-cells = <1>;
+		#size-cells = <0>;
+
+		cpu0: cpu at 0 {
+			device_type = "cpu";
+			compatible = "arm,cortex-a15";
+			reg = <0>;
+			clock-frequency = <1500000000>;
+		};
+	};
+
+	gic: interrupt-controller at f1001000 {
+		compatible = "arm,cortex-a15-gic";
+		#interrupt-cells = <3>;
+		#address-cells = <0>;
+		interrupt-controller;
+		reg = <0xf1001000 0x1000>,
+			<0xf1002000 0x1000>,
+			<0xf1004000 0x2000>,
+			<0xf1006000 0x2000>;
+		interrupts = <1 9 0xf04>;
+
+		gic-cpuif at 4 {
+			compatible = "arm,gic-cpuif";
+			cpuif-id = <4>;
+			cpu = <&cpu0>;
+		};
+	};
+
+	timer {
+		compatible = "arm,armv7-timer";
+		interrupts = <1 13 0xf08>,
+				<1 14 0xf08>,
+				<1 11 0xf08>,
+				<1 10 0xf08>;
+	};
+};
arch/arm/mach-shmobile/Kconfig +7 -0
@@ -18,6 +18,13 @@ config ARCH_SH73A0
 	select SH_CLK_CPG
 	select RENESAS_INTC_IRQPIN
 
+config ARCH_R8A73A4
+	bool "R-Mobile APE6 (R8A73A40)"
+	select ARM_GIC
+	select CPU_V7
+	select ARM_ARCH_TIMER
+	select SH_CLK_CPG
+
 config ARCH_R8A7740
 	bool "R-Mobile A1 (R8A77400)"
 	select ARCH_WANT_OPTIONAL_GPIOLIB
arch/arm/mach-shmobile/Makefile +1 -0
@@ -8,6 +8,7 @@ obj-y				:= timer.o console.o clock.o
 # CPU objects
 obj-$(CONFIG_ARCH_SH7372)	+= setup-sh7372.o clock-sh7372.o intc-sh7372.o
 obj-$(CONFIG_ARCH_SH73A0)	+= setup-sh73a0.o clock-sh73a0.o intc-sh73a0.o
+obj-$(CONFIG_ARCH_R8A73A4)	+= setup-r8a73a4.o clock-r8a73a4.o
 obj-$(CONFIG_ARCH_R8A7740)	+= setup-r8a7740.o clock-r8a7740.o intc-r8a7740.o
 obj-$(CONFIG_ARCH_R8A7779)	+= setup-r8a7779.o clock-r8a7779.o intc-r8a7779.o
 obj-$(CONFIG_ARCH_EMEV2)	+= setup-emev2.o clock-emev2.o
arch/arm/mach-shmobile/clock-r8a73a4.c new file +89 -0

89 lines not on this page · raw message

arch/arm/mach-shmobile/include/mach/r8a73a4.h new file +7 -0
@@ -0,0 +1,7 @@
+#ifndef __ASM_R8A73A4_H__
+#define __ASM_R8A73A4_H__
+
+void r8a73a4_add_standard_devices(void);
+void r8a73a4_clock_init(void);
+
+#endif /* __ASM_R8A73A4_H__ */
arch/arm/mach-shmobile/setup-r8a73a4.c new file +56 -0

56 lines not on this page · raw message

maintainer
#2 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

arch/arm/mach-shmobile/setup-r8a73a4.c line 37

Quoted 9 lines · Magnus Damm
arch/arm/mach-shmobile/setup-r8a73a4.c new file +9 -0
+static struct platform_device *r8a73a4_devices[] __initdata = {
+};
+
+void __init r8a73a4_add_standard_devices(void)
+{
+       r8a73a4_clock_init();
+
+       platform_add_devices(r8a73a4_devices, ARRAY_SIZE(r8a73a4_devices));
+}

I would suggest doing the platform_add_devices() only when you actually add devices to the array, unless you have a number of conflicting patches that each want to add their own devices.

arch/arm/mach-shmobile/setup-r8a73a4.c line 43

Quoted 5 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +5 -0
+#ifdef CONFIG_USE_OF
+void __init r8a73a4_add_standard_devices_dt(void)
+{
+       of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}

I have a patch that will make this function definition the default, so you no longer have to provide an init_machine callback if you don't do anything special. It's ok to leave it in for now, but we might want to do a follow up patch to remove it once both patches are merged.

arch/arm/mach-shmobile/setup-r8a73a4.c line 51

Quoted 7 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +7 -0
+static const char *r8a73a4_boards_compat_dt[] __initdata = {
+       "renesas,r8a73a4",
+       NULL,
+};
+
+DT_MACHINE_START(R8A73A4_DT, "Generic R8A73A4 (Flattened Device Tree)")
+       .init_irq       = irqchip_init,

Same thing for the default irqchip_init.

arch/arm/mach-shmobile/setup-r8a73a4.c line 54

Quoted 3 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +3 -0
+       .init_machine   = r8a73a4_add_standard_devices_dt,
+       .init_time      = shmobile_timer_init,
+       .dt_compat      = r8a73a4_boards_compat_dt,

Have you looked into using clocksource_of_init() here? Since you are using the ARM architected timers, I would expect that they soon will get probed using that function, which means we have to be careful crossing patches if someone wants to convert over all the existing users and you add a new one here.

Arnd

author maintainer
#3 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

Hi Arnd,

Quoted 14 lines · Arnd Bergmann

arch/arm/mach-shmobile/setup-r8a73a4.c line 37

Quoted 9 lines · Magnus Damm
arch/arm/mach-shmobile/setup-r8a73a4.c new file +9 -0
+static struct platform_device *r8a73a4_devices[] __initdata = {
+};
+
+void __init r8a73a4_add_standard_devices(void)
+{
+       r8a73a4_clock_init();
+
+       platform_add_devices(r8a73a4_devices, ARRAY_SIZE(r8a73a4_devices));
+}

I would suggest doing the platform_add_devices() only when you actually add devices to the array, unless you have a number of conflicting patches that each want to add their own devices.

So you would prefer that I fold this portion into a patch later in the series?

Quoted 10 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 43

Quoted 5 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +5 -0
+#ifdef CONFIG_USE_OF
+void __init r8a73a4_add_standard_devices_dt(void)
+{
+       of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}

I have a patch that will make this function definition the default, so you no longer have to provide an init_machine callback if you don't do anything special. It's ok to leave it in for now, but we might want to do a follow up patch to remove it once both patches are merged.

That's very nice! I am happy to hear that. Is that patch targeting v3.10?

Quoted 9 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 51

Quoted 7 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +7 -0
+static const char *r8a73a4_boards_compat_dt[] __initdata = {
+       "renesas,r8a73a4",
+       NULL,
+};
+
+DT_MACHINE_START(R8A73A4_DT, "Generic R8A73A4 (Flattened Device Tree)")
+       .init_irq       = irqchip_init,

Same thing for the default irqchip_init.

Excellent! I believe the easiest is to leave it as-is for now and have a clean-up series that can be merged late in the cycle after your patch is in. I am not sure if that fits with your way of dealing with the pull requests. At least that's how we used to do it for SH. If you have any suggestion when to remove it please let me know.

Quoted 9 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 54

Quoted 3 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +3 -0
+       .init_machine   = r8a73a4_add_standard_devices_dt,
+       .init_time      = shmobile_timer_init,
+       .dt_compat      = r8a73a4_boards_compat_dt,

Have you looked into using clocksource_of_init() here? Since you are using the ARM architected timers, I would expect that they soon will get probed using that function, which means we have to be careful crossing patches if someone wants to convert over all the existing users and you add a new one here.

Thanks for this suggestion. Right now we call shmobile_timer_init() that includes these calls:

	arch_timer_of_register();
	arch_timer_sched_clock_init();

For SH and mach-shmobile we have our own timers in drivers/clocksource as platform regular devices. They export clock source and clock event interfaces. Today we simply use the regular driver model together with SoC code to register platform devices during SoC setup - this including timers but excluding the arch timer.

I agree we need to find some way to make them work together with the ARM architected timer.

Thanks,

/ magnus

maintainer
#4 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

Quoted 16 lines · Magnus Damm
Quoted 14 lines · Arnd Bergmann

arch/arm/mach-shmobile/setup-r8a73a4.c line 37

Quoted 9 lines · Magnus Damm
arch/arm/mach-shmobile/setup-r8a73a4.c new file +9 -0
+static struct platform_device *r8a73a4_devices[] __initdata = {
+};
+
+void __init r8a73a4_add_standard_devices(void)
+{
+       r8a73a4_clock_init();
+
+       platform_add_devices(r8a73a4_devices, ARRAY_SIZE(r8a73a4_devices));
+}

I would suggest doing the platform_add_devices() only when you actually add devices to the array, unless you have a number of conflicting patches that each want to add their own devices.

Quoted 1 line

So you would prefer that I fold this portion into a patch later in the series?

If you are adding the devices in a later patch of the same series, it's probably better to leave it like it is. I would just like to avoid having dead code in a release if you don't get around to also add the devices.

Quoted 11 lines
Quoted 10 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 43

Quoted 5 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +5 -0
+#ifdef CONFIG_USE_OF
+void __init r8a73a4_add_standard_devices_dt(void)
+{
+       of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}

I have a patch that will make this function definition the default, so you no longer have to provide an init_machine callback if you don't do anything special. It's ok to leave it in for now, but we might want to do a follow up patch to remove it once both patches are merged.

Quoted 1 line

That's very nice! I am happy to hear that. Is that patch targeting v3.10?

Yes, I just need to produce a new version based on 3.9-rc.

Quoted 10 lines
Quoted 9 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 51

Quoted 7 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +7 -0
+static const char *r8a73a4_boards_compat_dt[] __initdata = {
+       "renesas,r8a73a4",
+       NULL,
+};
+
+DT_MACHINE_START(R8A73A4_DT, "Generic R8A73A4 (Flattened Device Tree)")
+       .init_irq       = irqchip_init,

Same thing for the default irqchip_init.

Quoted 5 lines

Excellent! I believe the easiest is to leave it as-is for now and have a clean-up series that can be merged late in the cycle after your patch is in. I am not sure if that fits with your way of dealing with the pull requests. At least that's how we used to do it for SH. If you have any suggestion when to remove it please let me know.

I don't want to get new patches during the merge window, but we can arrange a branch in the arm-soc tree that has both changes merged and the cleanup on top. We always send out pull requests for 10 to 15 branches in the merge window, and the later branches can be based on the earlier ones to deal with dependencies like this.

In this particular case, it would also not hurt to do the cleanup in 3.11, since the patch is only cosmetic.

Arnd

author maintainer
#5 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

Quoted 17 lines · Arnd Bergmann
Quoted 16 lines · Magnus Damm
Quoted 14 lines · Arnd Bergmann

arch/arm/mach-shmobile/setup-r8a73a4.c line 37

Quoted 9 lines · Magnus Damm
arch/arm/mach-shmobile/setup-r8a73a4.c new file +9 -0
+static struct platform_device *r8a73a4_devices[] __initdata = {
+};
+
+void __init r8a73a4_add_standard_devices(void)
+{
+       r8a73a4_clock_init();
+
+       platform_add_devices(r8a73a4_devices, ARRAY_SIZE(r8a73a4_devices));
+}

I would suggest doing the platform_add_devices() only when you actually add devices to the array, unless you have a number of conflicting patches that each want to add their own devices.

Quoted 5 lines
Quoted 1 line

So you would prefer that I fold this portion into a patch later in the series?

If you are adding the devices in a later patch of the same series, it's probably better to leave it like it is. I would just like to avoid having dead code in a release if you don't get around to also add the devices.

Sure. As you may have seen, I ended up moving away from static platform devices in V2. So this portion of the code is gone there.

Quoted 11 lines
Quoted 11 lines
Quoted 10 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 43

Quoted 5 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +5 -0
+#ifdef CONFIG_USE_OF
+void __init r8a73a4_add_standard_devices_dt(void)
+{
+       of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}

I have a patch that will make this function definition the default, so you no longer have to provide an init_machine callback if you don't do anything special. It's ok to leave it in for now, but we might want to do a follow up patch to remove it once both patches are merged.

Quoted 3 lines
Quoted 1 line

That's very nice! I am happy to hear that. Is that patch targeting v3.10?

Yes, I just need to produce a new version based on 3.9-rc.

That is fine, but as you suggest below, perhaps it is easier to wait until post-v3.10?

Quoted 16 lines
Quoted 10 lines
Quoted 9 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 51

Quoted 7 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +7 -0
+static const char *r8a73a4_boards_compat_dt[] __initdata = {
+       "renesas,r8a73a4",
+       NULL,
+};
+
+DT_MACHINE_START(R8A73A4_DT, "Generic R8A73A4 (Flattened Device Tree)")
+       .init_irq       = irqchip_init,

Same thing for the default irqchip_init.

Quoted 5 lines

Excellent! I believe the easiest is to leave it as-is for now and have a clean-up series that can be merged late in the cycle after your patch is in. I am not sure if that fits with your way of dealing with the pull requests. At least that's how we used to do it for SH. If you have any suggestion when to remove it please let me know.

Quoted 8 lines

I don't want to get new patches during the merge window, but we can arrange a branch in the arm-soc tree that has both changes merged and the cleanup on top. We always send out pull requests for 10 to 15 branches in the merge window, and the later branches can be based on the earlier ones to deal with dependencies like this.

In this particular case, it would also not hurt to do the cleanup in 3.11, since the patch is only cosmetic.

I agree that targeting v3.11 for this cleanup makes most sense. So I kept the functions in V2.

Thanks,

/ magnus

maintainer
#6 lore raw

PATCH 1/4 ARM: shmobile: Initial r8a73a4 SoC support

Quoted 11 lines · Magnus Damm
Quoted 11 lines
Quoted 11 lines
Quoted 10 lines

arch/arm/mach-shmobile/setup-r8a73a4.c line 43

Quoted 5 lines
arch/arm/mach-shmobile/setup-r8a73a4.c new file +5 -0
+#ifdef CONFIG_USE_OF
+void __init r8a73a4_add_standard_devices_dt(void)
+{
+       of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL);
+}

I have a patch that will make this function definition the default, so you no longer have to provide an init_machine callback if you don't do anything special. It's ok to leave it in for now, but we might want to do a follow up patch to remove it once both patches are merged.

Quoted 6 lines
Quoted 3 lines
Quoted 1 line

That's very nice! I am happy to hear that. Is that patch targeting v3.10?

Yes, I just need to produce a new version based on 3.9-rc.

That is fine, but as you suggest below, perhaps it is easier to wait until post-v3.10?

I'll try to do a cleanup of all the remaining machine_desc callbacks in arch/arm/mach-*/ as the final step of the 3.10 pull requests. Anything that I miss can be left for 3.11, so don't worry about it.

Arnd

author maintainer
#7 lore raw

PATCH 2/4 ARM: shmobile: r8a73a4 SCIF support

In reply to Magnus Damm

From: Magnus Damm <[email protected]>

Add SCIF serial port support to the r8a73a4 SoC by adding platform devices for SCIFA0 -> SCIFA1 as well as SCIFB0 -> SCIFB3 together with clock bindings. DT device description is excluded at this point since such bindings are still under development.

---

 arch/arm/mach-shmobile/clock-r8a73a4.c |   15 +++-
 arch/arm/mach-shmobile/setup-r8a73a4.c |  115 ++++++++++++++++++++++++++++++++
 2 files changed, 129 insertions(+), 1 deletion(-)
arch/arm/mach-shmobile/clock-r8a73a4.c +14 -1
@@ -28,6 +28,7 @@
 #define CPG_LEN 0x270
 
 #define MPCKCR 0xe6150080
+#define SMSTPCR2 0xe6150138
 
 static struct clk_mapping cpg_mapping = {
 	.phys   = CPG_BASE,
@@ -53,11 +54,23 @@ static struct clk *main_clks[] = {
 	&extal2_clk,
 };
 
-enum { MSTP_NR };
+enum { MSTP217, MSTP216, MSTP207, MSTP206, MSTP204, MSTP203, MSTP_NR };
 static struct clk mstp_clks[MSTP_NR] = {
+	[MSTP204] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 4, 1), /* SCIFA0 */
+	[MSTP203] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 3, 1), /* SCIFA1 */
+	[MSTP206] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 6, 1), /* SCIFB0 */
+	[MSTP207] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 7, 1), /* SCIFB1 */
+	[MSTP216] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 16, 1), /* SCIFB2 */
+	[MSTP217] = SH_CLK_MSTP32(&extal2_clk, SMSTPCR2, 17, 1), /* SCIFB3 */
 };
 
 static struct clk_lookup lookups[] = {
+	CLKDEV_DEV_ID("sh-sci.0", &mstp_clks[MSTP204]),
+	CLKDEV_DEV_ID("sh-sci.1", &mstp_clks[MSTP203]),
+	CLKDEV_DEV_ID("sh-sci.2", &mstp_clks[MSTP206]),
+	CLKDEV_DEV_ID("sh-sci.3", &mstp_clks[MSTP207]),
+	CLKDEV_DEV_ID("sh-sci.4", &mstp_clks[MSTP216]),
+	CLKDEV_DEV_ID("sh-sci.5", &mstp_clks[MSTP217]),
 };
 
 void __init r8a73a4_clock_init(void)
arch/arm/mach-shmobile/setup-r8a73a4.c +115 -0

127 lines not on this page · raw message

author maintainer
#8 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

In reply to Magnus Damm

From: Magnus Damm <[email protected]>

Add IRQC interrupt controller support to r8a73a4 by hooking up two IRQC instances to handle 58 external IRQ signals. There IRQC controllers are tied to SPIs of the GIC. On r8a73a4 exact IRQ pin routing is handled by the PFC which is excluded from this patch.

Both platform devices and DT devices are added in this patch. The platform device versions are used to provide a static interrupt map configuration for board code written in C.

---

Depends on:
[PATCH] irqchip: Renesas IRQC driver
[PATCH] irqchip: irqc: Add DT support

 arch/arm/boot/dts/r8a73a4.dtsi         |   32 ++++++++++
 arch/arm/mach-shmobile/Kconfig         |    1 
 arch/arm/mach-shmobile/setup-r8a73a4.c |   97 ++++++++++++++++++++++++++++++++
 3 files changed, 130 insertions(+)
arch/arm/boot/dts/r8a73a4.dtsi +32 -0
@@ -51,4 +51,36 @@
 				<1 11 0xf08>,
 				<1 10 0xf08>;
 	};
+
+	irqc0: interrupt-controller at e61c0000 {
+		compatible = "renesas,irqc";
+		#interrupt-cells = <2>;
+		interrupt-controller;
+		reg = <0xe61c0000 0x200>;
+		interrupt-parent = <&gic>;
+		interrupts = <0 0 4>, <0 1 4>, <0 2 4>,	<0 3 4>,
+				<0 4 4>, <0 5 4>, <0 6 4>, <0 7 4>,
+				<0 8 4>, <0 9 4>, <0 10 4>, <0 11 4>,
+				<0 12 4>, <0 13 4>, <0 14 4>, <0 15 4>,
+				<0 16 4>, <0 17 4>, <0 18 4>, <0 19 4>,
+				<0 20 4>, <0 21 4>, <0 22 4>, <0 23 4>,
+				<0 24 4>, <0 25 4>, <0 26 4>, <0 27 4>,
+				<0 28 4>, <0 29 4>, <0 30 4>, <0 31 4>;
+	};
+
+	irqc1: interrupt-controller at e61c0200 {
+		compatible = "renesas,irqc";
+		#interrupt-cells = <2>;
+		interrupt-controller;
+		reg = <0xe61c0200 0x200>;
+		interrupt-parent = <&gic>;
+		interrupts = <0 32 4>, <0 33 4>, <0 34 4>, <0 35 4>,
+				<0 36 4>, <0 37 4>, <0 38 4>, <0 39 4>,
+				<0 40 4>, <0 41 4>, <0 42 4>, <0 43 4>,
+				<0 44 4>, <0 45 4>, <0 46 4>, <0 47 4>,
+				<0 48 4>, <0 49 4>, <0 50 4>, <0 51 4>,
+				<0 52 4>, <0 53 4>, <0 54 4>, <0 55 4>,
+				<0 56 4>, <0 57 4>;
+	};
+
 };
arch/arm/mach-shmobile/Kconfig +1 -0
@@ -24,6 +24,7 @@ config ARCH_R8A73A4
 	select CPU_V7
 	select ARM_ARCH_TIMER
 	select SH_CLK_CPG
+	select RENESAS_IRQC
 
 config ARCH_R8A7740
 	bool "R-Mobile A1 (R8A77400)"
arch/arm/mach-shmobile/setup-r8a73a4.c +97 -0

115 lines not on this page · raw message

maintainer
#9 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

Quoted 8 lines · Magnus Damm

From: Magnus Damm <[email protected]>

Add IRQC interrupt controller support to r8a73a4 by hooking up two IRQC instances to handle 58 external IRQ signals. There IRQC controllers are tied to SPIs of the GIC. On r8a73a4 exact IRQ pin routing is handled by the PFC which is excluded from this patch.

Quoted 6 lines

Both platform devices and DT devices are added in this patch. The platform device versions are used to provide a static interrupt map configuration for board code written in C.

What is the status of the IRQC DT support? I remember that I wasn't happy with a prior version, but I did not follow up on some of the questions that came up, sorry about that.

Did the patches end up getting merged anyway, or should we resume the discussion about those patches? I understand that lack of INTC bindings would make new SoC support particularly hard, and I don't want to be responsible for holding you up here.

Arnd

author maintainer
#10 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

Hi Arnd,

Thanks for your feedback, please see below for my reply.

Quoted 14 lines · Arnd Bergmann
Quoted 13 lines · Magnus Damm

From: Magnus Damm <[email protected]>

Add IRQC interrupt controller support to r8a73a4 by hooking up two IRQC instances to handle 58 external IRQ signals. There IRQC controllers are tied to SPIs of the GIC. On r8a73a4 exact IRQ pin routing is handled by the PFC which is excluded from this patch.

Both platform devices and DT devices are added in this patch. The platform device versions are used to provide a static interrupt map configuration for board code written in C.

Quoted 5 lines
Quoted 1 line

What is the status of the IRQC DT support? I remember that I wasn't happy with a prior version, but I did not follow up on some of the questions that came up, sorry about that.

Uhm, perhaps I misunderstand, but I wonder if you refer to INTC instead of IRQC?

This particular driver is for the IRQC hardware block. It is not compatible with INTC. A while ago I posted an incremental DT support patch for IRQC - "[PATCH] irqchip: irqc: Add DT support", please see
https://lkml.org/lkml/2013/3/6/50

Quoted 4 lines

Did the patches end up getting merged anyway, or should we resume the discussion about those patches? I understand that lack of INTC bindings would make new SoC support particularly hard, and I don't want to be responsible for holding you up here.

Thanks. I don't think the INTC patches went anywhere.

To zoom out a bit let me list different interrupt controllers:

A) INTC (drivers/sh/intc) [no DT yet]
B) GIC (drivers/irqchip/irq-gic.c) [DT]
C) INTC External IRQ Pin (drivers/irqchip/irq-renesas-intc-irqpin.c) [DT]
D) IRQC (drivers/irqchip/irq-renesas-irqc.c) [DT]

Simple use cases are:
- Legacy SH SoCs or ARM SoCs with Cortex-A8 or older make use of A).
- More recent ARM SoCs with Cortex-A9 or newer use B) and C) or B) and D).

On top of this we now and then have GPIO controllers that have built-in interrupt controllers.

Hope this helps,

/ magnus

maintainer
#11 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

Quoted 19 lines · Magnus Damm

Hi Arnd,

Thanks for your feedback, please see below for my reply.

Quoted 14 lines · Arnd Bergmann
Quoted 13 lines · Magnus Damm

From: Magnus Damm <[email protected]>

Add IRQC interrupt controller support to r8a73a4 by hooking up two IRQC instances to handle 58 external IRQ signals. There IRQC controllers are tied to SPIs of the GIC. On r8a73a4 exact IRQ pin routing is handled by the PFC which is excluded from this patch.

Both platform devices and DT devices are added in this patch. The platform device versions are used to provide a static interrupt map configuration for board code written in C.

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

What is the status of the IRQC DT support? I remember that I wasn't happy with a prior version, but I did not follow up on some of the questions that came up, sorry about that.

Uhm, perhaps I misunderstand, but I wonder if you refer to INTC instead of IRQC?

Yes, you are right.

Quoted 4 lines

This particular driver is for the IRQC hardware block. It is not compatible with INTC. A while ago I posted an incremental DT support patch for IRQC - "[PATCH] irqchip: irqc: Add DT support", please see
https://lkml.org/lkml/2013/3/6/50

Ok, I had missed that, but the driver certainly looks good to me.

Quoted 9 lines
Quoted 4 lines

Did the patches end up getting merged anyway, or should we resume the discussion about those patches? I understand that lack of INTC bindings would make new SoC support particularly hard, and I don't want to be responsible for holding you up here.

Thanks. I don't think the INTC patches went anywhere.

To zoom out a bit let me list different interrupt controllers:

Quoted 8 lines

A) INTC (drivers/sh/intc) [no DT yet]
B) GIC (drivers/irqchip/irq-gic.c) [DT]
C) INTC External IRQ Pin (drivers/irqchip/irq-renesas-intc-irqpin.c) [DT]
D) IRQC (drivers/irqchip/irq-renesas-irqc.c) [DT]

Simple use cases are:
- Legacy SH SoCs or ARM SoCs with Cortex-A8 or older make use of A).
- More recent ARM SoCs with Cortex-A9 or newer use B) and C) or B) and D).

Ok. And I guess the EMMA EV2 uses only GIC but not IRQC or INTC, right?

Quoted 2 lines

On top of this we now and then have GPIO controllers that have built-in interrupt controllers.

Yes, obviously.

Coming back to INTC, are you planning to use the same binding for A and C? Which of them the binding you posted earlier for?

When I looked at the existing code, I had the impression that doing a binding for just the SH-Mobile SoCs that have an ARM core in them (including those that also have an SH core) would be much easier than doing a binding that also covers the older SH SoCs, since those are much less uniform.

Arnd

author maintainer
#12 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

Quoted 22 lines · Arnd Bergmann
Quoted 21 lines · Magnus Damm

Hi Arnd,

Thanks for your feedback, please see below for my reply.

Quoted 16 lines · Arnd Bergmann
Quoted 14 lines · Magnus Damm

From: Magnus Damm <[email protected]>

Add IRQC interrupt controller support to r8a73a4 by hooking up two IRQC instances to handle 58 external IRQ signals. There IRQC controllers are tied to SPIs of the GIC. On r8a73a4 exact IRQ pin routing is handled by the PFC which is excluded from this patch.

Both platform devices and DT devices are added in this patch. The platform device versions are used to provide a static interrupt map configuration for board code written in C.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

What is the status of the IRQC DT support? I remember that I wasn't happy with a prior version, but I did not follow up on some of the questions that came up, sorry about that.

Uhm, perhaps I misunderstand, but I wonder if you refer to INTC instead of IRQC?

Yes, you are right.

I don't blame you for the misunderstanding - as you see there are a couple of interrupt controllers just from Renesas SoC division, and you probably get to see similar situations with every SoC vendor. Quite a few different interrupt controllers to keep track of! =)

Quoted 6 lines
Quoted 4 lines

This particular driver is for the IRQC hardware block. It is not compatible with INTC. A while ago I posted an incremental DT support patch for IRQC - "[PATCH] irqchip: irqc: Add DT support", please see
https://lkml.org/lkml/2013/3/6/50

Ok, I had missed that, but the driver certainly looks good to me.

Thanks.

Quoted 14 lines
Quoted 14 lines
Quoted 4 lines

Did the patches end up getting merged anyway, or should we resume the discussion about those patches? I understand that lack of INTC bindings would make new SoC support particularly hard, and I don't want to be responsible for holding you up here.

Thanks. I don't think the INTC patches went anywhere.

To zoom out a bit let me list different interrupt controllers:

A) INTC (drivers/sh/intc) [no DT yet]
B) GIC (drivers/irqchip/irq-gic.c) [DT]
C) INTC External IRQ Pin (drivers/irqchip/irq-renesas-intc-irqpin.c) [DT]
D) IRQC (drivers/irqchip/irq-renesas-irqc.c) [DT]

Quoted 5 lines
Quoted 3 lines

Simple use cases are:
- Legacy SH SoCs or ARM SoCs with Cortex-A8 or older make use of A).
- More recent ARM SoCs with Cortex-A9 or newer use B) and C) or B) and D).

Ok. And I guess the EMMA EV2 uses only GIC but not IRQC or INTC, right?

Yes, correct. GIC and per-GPIO interrupts provided by gpio-em.c.

Quoted 4 lines
Quoted 2 lines

On top of this we now and then have GPIO controllers that have built-in interrupt controllers.

Yes, obviously.

What may not be so obvious is the GPIO interface provided by the PFC code that is now located in drivers/pinctrl/sh-pfc/. There used to be a function-only GPIOs in the PFC code but we're now moving over to PINCTRL thanks to great effort by Laurent. So besides the old function-only GPIOs provided by PFC there are also regular GPIOs exported by the PFC code. Those regular GPIOs exported by the PFC do not have any built-in interrupt controller - instead the external IRQs handled by C) and D) above are used.

Other GPIO blocks like gpio-em.c and gpio-rcar.c have built-in interrupt controllers and can do per-GPIO interrupts.

Quoted 1 line

Coming back to INTC, are you planning to use the same binding for A and C?

I was not planning on that, no. To be honest, at this point I am not sure which way forward is best for A).

Quoted 1 line

Which of them the binding you posted earlier for?

Regarding A), I wrote some local DT prototype patches for INTC a year or two ago, showing how things could be done and how we can start using DT. Then I handed the job over to other people. However, from there my advice of incremental development was ignored and instead more complete bindings were developed directly without much review. So I wouldn't say that I posted the bindings myself. Right now I'm very hands-off in that area.

As for C), those DT bindings were done by me. They are however not compatible with A). The hardware is different. C) is basically a special case subset of A).

Quoted 5 lines

When I looked at the existing code, I had the impression that doing a binding for just the SH-Mobile SoCs that have an ARM core in them (including those that also have an SH core) would be much easier than doing a binding that also covers the older SH SoCs, since those are much less uniform.

I agree that in some cases it may make sense to split the SH bits from ARM, but I wonder how much we would win for the INTC.

Right now I'm considering converting r8a7740 to use B) and C) (if possible), and if so then the only ARM SoC using A) for main interrupt controller left is sh7372.

For sh7372 we could simply try to use C) for board-level interrupts (and board level DT) but keep the SoC portion in C with A) until the SoC is being phased out. Or we could have a simple compatible "renesas,intc-sh7372" with tables in C using irq domain to support DT, but that would be exactly as my first incremental development task that wasn't followed...

What would you do?

Thanks for your help,

/ magnus

maintainer
#13 lore raw

PATCH 3/4 ARM: shmobile: r8a73a4 IRQC support

Quoted 10 lines · Magnus Damm
Quoted 3 lines · Arnd Bergmann

On Thursday 14 March 2013, Magnus Damm wrote:

Coming back to INTC, are you planning to use the same binding for A and C?

I was not planning on that, no. To be honest, at this point I am not sure which way forward is best for A).

Quoted 1 line

Which of them the binding you posted earlier for?

Quoted 7 lines

Regarding A), I wrote some local DT prototype patches for INTC a year or two ago, showing how things could be done and how we can start using DT. Then I handed the job over to other people. However, from there my advice of incremental development was ignored and instead more complete bindings were developed directly without much review. So I wouldn't say that I posted the bindings myself. Right now I'm very hands-off in that area.

Ok.

Quoted 3 lines

As for C), those DT bindings were done by me. They are however not compatible with A). The hardware is different. C) is basically a special case subset of A).

Makes sense.

Quoted 13 lines
Quoted 5 lines

When I looked at the existing code, I had the impression that doing a binding for just the SH-Mobile SoCs that have an ARM core in them (including those that also have an SH core) would be much easier than doing a binding that also covers the older SH SoCs, since those are much less uniform.

I agree that in some cases it may make sense to split the SH bits from ARM, but I wonder how much we would win for the INTC.

Right now I'm considering converting r8a7740 to use B) and C) (if possible), and if so then the only ARM SoC using A) for main interrupt controller left is sh7372.

Quoted 8 lines

For sh7372 we could simply try to use C) for board-level interrupts (and board level DT) but keep the SoC portion in C with A) until the SoC is being phased out. Or we could have a simple compatible "renesas,intc-sh7372" with tables in C using irq domain to support DT, but that would be exactly as my first incremental development task that wasn't followed...

What would you do?

Both of htese approaches sound fine. If there is only a single odd one out, it makes sense to have a hardwired implementation for that one. I rejected that approach originally because I wanted something more generic, and it seems you have done exactly that with the irq-renesas-intc-irqpin driver.

Arnd

author maintainer
#14 lore raw

PATCH 4/4 ARM: shmobile: APE6EVM support

In reply to Magnus Damm

From: Magnus Damm <[email protected]>

Add APE6EVM board support making use of 1 GiB of memory, the SCIFA0 serial port and the Architecture timer.

More devices will be added on top of this patch after PICNTRL and clock framework are in better shape.

---

 arch/arm/boot/dts/Makefile             |    1 
 arch/arm/boot/dts/r8a73a4-ape6evm.dts  |   26 ++++++++++++++++++++
 arch/arm/mach-shmobile/Kconfig         |    8 +++++-
 arch/arm/mach-shmobile/Makefile        |    1 
 arch/arm/mach-shmobile/board-ape6evm.c |   40 ++++++++++++++++++++++++++++++++
 5 files changed, 75 insertions(+), 1 deletion(-)
arch/arm/boot/dts/Makefile +1 -0
@@ -125,6 +125,7 @@ dtb-$(CONFIG_ARCH_SHMOBILE) += emev2-kzm
 	r8a7779-marzen-reference.dtb \
 	sh73a0-kzm9g.dtb \
 	sh73a0-kzm9g-reference.dtb \
+	r8a73a4-ape6evm.dtb \
 	sh7372-mackerel.dtb
 dtb-$(CONFIG_ARCH_SPEAR13XX) += spear1310-evb.dtb \
 	spear1340-evb.dtb
arch/arm/boot/dts/r8a73a4-ape6evm.dts new file +26 -0
@@ -0,0 +1,26 @@
+/*
+ * Device Tree Source for the APE6EVM board
+ *
+ * Copyright (C) 2013 Renesas Solutions Corp.
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2.  This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+/dts-v1/;
+/include/ "r8a73a4.dtsi"
+
+/ {
+	model = "APE6EVM";
+	compatible = "renesas,ape6evm", "renesas,r8a73a4";
+
+	chosen {
+		bootargs = "console=ttySC0,115200 ignore_loglevel";
+	};
+
+	memory at 40000000 {
+		device_type = "memory";
+		reg = <0x40000000 0x40000000>;
+	};
+};
arch/arm/mach-shmobile/Kconfig +7 -1
@@ -76,6 +76,11 @@ config MACH_AG5EVM
 	select REGULATOR_FIXED_VOLTAGE if REGULATOR
 	select SH_LCD_MIPI_DSI
 
+config MACH_APE6EVM
+	bool "APE6EVM board"
+	depends on ARCH_R8A73A4
+	select USE_OF
+
 config MACH_MACKEREL
 	bool "mackerel board"
 	depends on ARCH_SH7372
@@ -163,7 +168,7 @@ config MEMORY_START
 	hex "Physical memory start address"
 	default "0x40000000" if MACH_AP4EVB || MACH_AG5EVM || \
 				MACH_MACKEREL || MACH_BONITO || \
-				MACH_ARMADILLO800EVA
+				MACH_ARMADILLO800EVA || MACH_APE6EVM
 	default "0x41000000" if MACH_KOTA2
 	default "0x00000000"
 	---help---
@@ -173,6 +178,7 @@ config MEMORY_START
 
 config MEMORY_SIZE
 	hex "Physical memory size"
+	default "0x40000000" if MACH_APE6EVM
 	default "0x20000000" if MACH_AG5EVM || MACH_BONITO || \
 				MACH_ARMADILLO800EVA
 	default "0x1e000000" if MACH_KOTA2
arch/arm/mach-shmobile/Makefile +1 -0
@@ -35,6 +35,7 @@ obj-$(CONFIG_ARCH_SH73A0)	+= pm-sh73a0.o
 # Board objects
 obj-$(CONFIG_MACH_AP4EVB)	+= board-ap4evb.o
 obj-$(CONFIG_MACH_AG5EVM)	+= board-ag5evm.o
+obj-$(CONFIG_MACH_APE6EVM)	+= board-ape6evm.o
 obj-$(CONFIG_MACH_MACKEREL)	+= board-mackerel.o
 obj-$(CONFIG_MACH_KOTA2)	+= board-kota2.o
 obj-$(CONFIG_MACH_BONITO)	+= board-bonito.o
arch/arm/mach-shmobile/board-ape6evm.c new file +40 -0
@@ -0,0 +1,40 @@
+/*
+ * APE6EVM board support
+ *
+ * Copyright (C) 2013  Renesas Solutions Corp.
+ * Copyright (C) 2013  Magnus Damm
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/irqchip.h>
+#include <mach/common.h>
+#include <mach/r8a73a4.h>
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>
+
+static const char *ape6evm_boards_compat_dt[] __initdata = {
+	"renesas,ape6evm",
+	NULL,
+};
+
+DT_MACHINE_START(APE6EVM_DT, "ape6evm")
+	.init_irq	= irqchip_init,
+	.init_time	= shmobile_timer_init,
+	.init_machine	= r8a73a4_add_standard_devices,
+	.dt_compat	= ape6evm_boards_compat_dt,
+MACHINE_END
#15 lore raw

PATCH 4/4 ARM: shmobile: APE6EVM support

Hi Magnus

I already sent Reviewed-by, but...

Quoted 19 lines
arch/arm/mach-shmobile/board-ape6evm.c new file +18 -0
@@ -0,0 +1,40 @@
+/*
+ * APE6EVM board support
+ *
+ * Copyright (C) 2013  Renesas Solutions Corp.
+ * Copyright (C) 2013  Magnus Damm
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA

arch/arm/mach-shmobile/board-ape6evm.c line 28

Quoted 10 lines
arch/arm/mach-shmobile/board-ape6evm.c new file +10 -0
+ */
+
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
+#include <linux/irqchip.h>
+#include <mach/common.h>
+#include <mach/r8a73a4.h>
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>

There are too many include here,
and not in alphabetical order :)

Best regards
---
Kuninori Morimoto

author maintainer
#16 lore raw

PATCH 4/4 ARM: shmobile: APE6EVM support

Quoted 29 lines · Kuninori Morimoto

Hi Magnus

I already sent Reviewed-by, but...

Quoted 24 lines
arch/arm/mach-shmobile/board-ape6evm.c new file +23 -0
@@ -0,0 +1,40 @@
+/*
+ * APE6EVM board support
+ *
+ * Copyright (C) 2013  Renesas Solutions Corp.
+ * Copyright (C) 2013  Magnus Damm
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; version 2 of the License.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA  02110-1301  USA
+ */
+
+#include <linux/kernel.h>
+#include <linux/interrupt.h>
+#include <linux/platform_device.h>
Quoted 8 lines

arch/arm/mach-shmobile/board-ape6evm.c line 28

Quoted 5 lines
arch/arm/mach-shmobile/board-ape6evm.c new file +5 -0
+#include <linux/irqchip.h>
+#include <mach/common.h>
+#include <mach/r8a73a4.h>
+#include <asm/mach-types.h>
+#include <asm/mach/arch.h>

There are too many include here,
and not in alphabetical order :)

Yes. And what's next? Do I need to encode using Shift-JIS? =)

/ magnus

maintainer
#17 lore raw

PATCH 4/4 ARM: shmobile: APE6EVM support

In reply to Magnus Damm

arch/arm/mach-shmobile/board-ape6evm.c line 40

Quoted 12 lines · Magnus Damm
arch/arm/mach-shmobile/board-ape6evm.c new file +12 -0
+
+static const char *ape6evm_boards_compat_dt[] __initdata = {
+       "renesas,ape6evm",
+       NULL,
+};
+
+DT_MACHINE_START(APE6EVM_DT, "ape6evm")
+       .init_irq       = irqchip_init,
+       .init_time      = shmobile_timer_init,
+       .init_machine   = r8a73a4_add_standard_devices,
+       .dt_compat      = ape6evm_boards_compat_dt,
+MACHINE_END

Do you plan to add board-specific contents here in a follow-up patch soon?

If not, it seems the setup-r8a73a4.c already handles this board with its defaults.

Arnd

author maintainer
#18 lore raw

PATCH 4/4 ARM: shmobile: APE6EVM support

Hi Arnd,

Quoted 15 lines · Arnd Bergmann

arch/arm/mach-shmobile/board-ape6evm.c line 40

Quoted 12 lines · Magnus Damm
arch/arm/mach-shmobile/board-ape6evm.c new file +12 -0
+
+static const char *ape6evm_boards_compat_dt[] __initdata = {
+       "renesas,ape6evm",
+       NULL,
+};
+
+DT_MACHINE_START(APE6EVM_DT, "ape6evm")
+       .init_irq       = irqchip_init,
+       .init_time      = shmobile_timer_init,
+       .init_machine   = r8a73a4_add_standard_devices,
+       .dt_compat      = ape6evm_boards_compat_dt,
+MACHINE_END

Do you plan to add board-specific contents here in a follow-up patch soon?

Yes, correct. At this point we cannot configure the pin controller from DT.

Quoted 2 lines

If not, it seems the setup-r8a73a4.c already handles this board with its defaults.

Right, this portion is only made as a ground work for incremental patches.

Thanks,

/ magnus

#19 lore raw

cover ARM: shmobile: r8a73a4 SoC and APE6EVM board support

In reply to Magnus Damm

Hi

Quoted 22 lines

ARM: shmobile: r8a73a4 SoC and APE6EVM board support

[PATCH 01/04] ARM: shmobile: Initial r8a73a4 SoC support [PATCH 02/04] ARM: shmobile: r8a73a4 SCIF support [PATCH 03/04] ARM: shmobile: r8a73a4 IRQC support [PATCH 04/04] ARM: shmobile: APE6EVM support

This series adds initial support for r8a73a4 and APE6EVM. The SoC is known as r8a73a4 which is built around Cortex-A15 and makes use of SCIF for serial and IRQC interrupt controller for external IRQs.

This initial SoC and board support is providing limited DT-only support and a very basic C-version of board support. In the future improved PINCTRL and clock framework support will be added, followed by SMP support.

---

Written against all+next in renesas.git.
[PATCH 03/04] requires the IRQC driver to build

Quoted 10 lines
 arch/arm/boot/dts/Makefile                    |    1 
 arch/arm/boot/dts/r8a73a4-ape6evm.dts         |   26 ++
 arch/arm/boot/dts/r8a73a4.dtsi                |   87 ++++++++
 arch/arm/mach-shmobile/Kconfig                |   16 +
 arch/arm/mach-shmobile/Makefile               |    2 
 arch/arm/mach-shmobile/board-ape6evm.c        |   40 +++
 arch/arm/mach-shmobile/clock-r8a73a4.c        |  104 +++++++++
 arch/arm/mach-shmobile/include/mach/r8a73a4.h |    7 
 arch/arm/mach-shmobile/setup-r8a73a4.c        |  268 +++++++++++++++++++++++++
 9 files changed, 549 insertions(+), 2 deletions(-)

For all patches

Best regards
---
Kuninori Morimoto

maintainer
#20 lore raw

cover ARM: shmobile: r8a73a4 SoC and APE6EVM board support

In reply to Magnus Damm
Quoted 11 lines · Magnus Damm

ARM: shmobile: r8a73a4 SoC and APE6EVM board support

[PATCH 01/04] ARM: shmobile: Initial r8a73a4 SoC support [PATCH 02/04] ARM: shmobile: r8a73a4 SCIF support [PATCH 03/04] ARM: shmobile: r8a73a4 IRQC support [PATCH 04/04] ARM: shmobile: APE6EVM support

This series adds initial support for r8a73a4 and APE6EVM. The SoC is known as r8a73a4 which is built around Cortex-A15 and makes use of SCIF for serial and IRQC interrupt controller for external IRQs.

Quoted 6 lines

This initial SoC and board support is providing limited DT-only support and a very basic C-version of board support. In the future improved PINCTRL and clock framework support will be added, followed by SMP support.

Patches look good so far. I assume you have made them in a way that lets you easily keep around non-DT support internally, right?

Arnd

author maintainer
#21 lore raw

cover ARM: shmobile: r8a73a4 SoC and APE6EVM board support

Quoted 17 lines · Arnd Bergmann
Quoted 16 lines · Magnus Damm

ARM: shmobile: r8a73a4 SoC and APE6EVM board support

[PATCH 01/04] ARM: shmobile: Initial r8a73a4 SoC support [PATCH 02/04] ARM: shmobile: r8a73a4 SCIF support [PATCH 03/04] ARM: shmobile: r8a73a4 IRQC support [PATCH 04/04] ARM: shmobile: APE6EVM support

This series adds initial support for r8a73a4 and APE6EVM. The SoC is known as r8a73a4 which is built around Cortex-A15 and makes use of SCIF for serial and IRQC interrupt controller for external IRQs.

This initial SoC and board support is providing limited DT-only support and a very basic C-version of board support. In the future improved PINCTRL and clock framework support will be added, followed by SMP support.

Quoted 4 lines
Quoted 1 line

Patches look good so far. I assume you have made them in a way that lets you easily keep around non-DT support internally, right?

Thanks. Regarding your question, please allow me try to answer a rather simple question in a complicated way. =)

I have some perhaps naive hope of not having to force people to write out of tree code. I believe it is possible to use upstream to deliver something short term and at the same time work with upstream towards long term goals. So I hope we won't be forced to keep non-DT support around internally.

For short term, in some cases where we don't have complete DT bindings (like clocks and pinctrl at the moment) we still have to use board code written in C to configure the hardware. Using DT only at this point would result in a non-working kernel without support for mandatory components like pinmux, clock support and power domains, so for short term that doesn't fly.

As I know you've seen, we are actively working on improving our DT coverage. Our approach to DT so far is to work our way towards more complete support in the -reference DT board support code. We keep the C version around as feature complete implementation and when the -reference DT version of the board code is good enough then we kill the C version and use DT only.

These patches for new SoCs and boards use a mix of DT and C. In practice that means that I plan on using platform devices and DT in parallel. These platform devices can easily go away upon your request, if so a new version of the patch will be submitted with the offensive hunks removed. The old version may however still be used for back porting.

So if there are some portions of the code that you would like me to rework then please let me know and I will do my best to follow your guidance. I am also open to clean up various portions of the code if you have some particular place that you feel needs extra attention.

Thanks for your help,

/ magnus

maintainer
#22 lore raw

cover ARM: shmobile: r8a73a4 SoC and APE6EVM board support

Quoted 6 lines · Magnus Damm

These patches for new SoCs and boards use a mix of DT and C. In practice that means that I plan on using platform devices and DT in parallel. These platform devices can easily go away upon your request, if so a new version of the patch will be submitted with the offensive hunks removed. The old version may however still be used for back porting.

Yes, that sounds like a good plan.

Quoted 4 lines

So if there are some portions of the code that you would like me to rework then please let me know and I will do my best to follow your guidance. I am also open to clean up various portions of the code if you have some particular place that you feel needs extra attention.

The only part that confused me was the use of "#ifdef CONFIG_USE_OF" in the setup-r8a73a4.c file. If the plan is to use a combination of DT with static platform_data (or platform_devices, more on that below), you always need that conditional code, and I think it would be more logical to remove the #ifdef. I understand that having the #ifdef in place makes your life easier for backporting to kernels that don't have enough DT support yet, but I think you are going to run into larger issues tha this the more those divert from current mainline over time.

Regarding how to best mix platform devices and DT, my recommendation is to use AUXDATA for the platform_data and device names while getting the IRQ and MMIO resources from DT for devices that have no binding yet, rather than creating the complete device in platform code. In particular, this should make it easier to deal with IRQ domains, and with devices that are disabled on some boards, but again it will make you backporting more work.

Regarding how platform devices are created, Greg KH has been trying to stop people from statically declaring the devices in platform code for many years, the recommended way to do it (if you have to) is calling one of platform_device_register_{full,resndata,data,simple}. We still have a large number of static platform devices in ARM, and I'm not asking anyone to actively convert the existing ones, but I should probably be a little stricter about using the proper interfaces in new code.

To summarize, the many ways to create platform devices from most preferred to least are:

1. DT without auxdata
2. DT with auxdata
3. platform_device_register_*
4. static definition with platform_device_register

Arnd