* [PATCH 0/5] Add platform support for LSI AXM55xx
@ 2014-04-15 12:06 Anders Berg
2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg
` (4 more replies)
0 siblings, 5 replies; 17+ messages in thread
From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw)
To: linux-arm-kernel
Hi,
This set of patches introduces basic platform support for the LSI AXM5516 SoC.
This is another 16 core machine (4x4 cluster config). Please review and provide
feedback.
Regards,
Anders
Anders Berg (5):
ARM: Add platform support for LSI AXM55xx SoC
ARM: dts: Device tree for AXM55xx.
ARM: axxia: Adding defconfig for AXM55xx
power: reset: Add Axxia system reset driver
ARM: dts: axxia: Add reset controller
Documentation/devicetree/bindings/arm/axxia.txt | 12 +
.../devicetree/bindings/clock/clk-axxia.txt | 88 ++++++
.../bindings/power_supply/axxia-reset.txt | 20 ++
arch/arm/Kconfig | 2 +
arch/arm/Makefile | 2 +
arch/arm/boot/dts/Makefile | 1 +
arch/arm/boot/dts/axm5516-amarillo.dts | 51 ++++
arch/arm/boot/dts/axm5516-cpus.dtsi | 204 ++++++++++++++
arch/arm/boot/dts/axm55xx.dtsi | 311 +++++++++++++++++++++
arch/arm/configs/axm55xx_defconfig | 248 ++++++++++++++++
arch/arm/mach-axxia/Kconfig | 19 ++
arch/arm/mach-axxia/Makefile | 2 +
arch/arm/mach-axxia/Makefile.boot | 2 +
arch/arm/mach-axxia/axxia.c | 117 ++++++++
arch/arm/mach-axxia/axxia.h | 42 +++
arch/arm/mach-axxia/headsmp.S | 44 +++
arch/arm/mach-axxia/platsmp.c | 183 ++++++++++++
drivers/clk/Makefile | 1 +
drivers/clk/clk-axxia.c | 281 +++++++++++++++++++
drivers/power/reset/Kconfig | 8 +
drivers/power/reset/Makefile | 1 +
drivers/power/reset/axxia-reset.c | 97 +++++++
22 files changed, 1736 insertions(+)
create mode 100644 Documentation/devicetree/bindings/arm/axxia.txt
create mode 100644 Documentation/devicetree/bindings/clock/clk-axxia.txt
create mode 100644 Documentation/devicetree/bindings/power_supply/axxia-reset.txt
create mode 100644 arch/arm/boot/dts/axm5516-amarillo.dts
create mode 100644 arch/arm/boot/dts/axm5516-cpus.dtsi
create mode 100644 arch/arm/boot/dts/axm55xx.dtsi
create mode 100644 arch/arm/configs/axm55xx_defconfig
create mode 100644 arch/arm/mach-axxia/Kconfig
create mode 100644 arch/arm/mach-axxia/Makefile
create mode 100644 arch/arm/mach-axxia/Makefile.boot
create mode 100644 arch/arm/mach-axxia/axxia.c
create mode 100644 arch/arm/mach-axxia/axxia.h
create mode 100644 arch/arm/mach-axxia/headsmp.S
create mode 100644 arch/arm/mach-axxia/platsmp.c
create mode 100644 drivers/clk/clk-axxia.c
create mode 100644 drivers/power/reset/axxia-reset.c
--
1.8.3.2
^ permalink raw reply [flat|nested] 17+ messages in thread* [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg @ 2014-04-15 12:06 ` Anders Berg 2014-04-15 12:30 ` Arnd Bergmann 2014-04-24 9:04 ` Russell King - ARM Linux 2014-04-15 12:06 ` [PATCH 2/5] ARM: dts: Device tree for AXM55xx Anders Berg ` (3 subsequent siblings) 4 siblings, 2 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw) To: linux-arm-kernel The AXM55xx family consists of devices that may contain up to 16 ARM Cortex-A15 cores (in a 4x4 cluster configuration). The cores within each cluster share an L2 cache, and the clusters are connected to each other via a CCN-504 cache coherent interconnect. This machine requires CONFIG_ARM_LPAE enabled as all peripherals are located above 4GB in the memory map. Signed-off-by: Anders Berg <anders.berg@lsi.com> --- Documentation/devicetree/bindings/arm/axxia.txt | 12 + .../devicetree/bindings/clock/clk-axxia.txt | 88 +++++++ arch/arm/Kconfig | 2 + arch/arm/Makefile | 2 + arch/arm/mach-axxia/Kconfig | 19 ++ arch/arm/mach-axxia/Makefile | 2 + arch/arm/mach-axxia/Makefile.boot | 2 + arch/arm/mach-axxia/axxia.c | 117 +++++++++ arch/arm/mach-axxia/axxia.h | 42 +++ arch/arm/mach-axxia/headsmp.S | 44 ++++ arch/arm/mach-axxia/platsmp.c | 183 ++++++++++++++ drivers/clk/Makefile | 1 + drivers/clk/clk-axxia.c | 281 +++++++++++++++++++++ 13 files changed, 795 insertions(+) create mode 100644 Documentation/devicetree/bindings/arm/axxia.txt create mode 100644 Documentation/devicetree/bindings/clock/clk-axxia.txt create mode 100644 arch/arm/mach-axxia/Kconfig create mode 100644 arch/arm/mach-axxia/Makefile create mode 100644 arch/arm/mach-axxia/Makefile.boot create mode 100644 arch/arm/mach-axxia/axxia.c create mode 100644 arch/arm/mach-axxia/axxia.h create mode 100644 arch/arm/mach-axxia/headsmp.S create mode 100644 arch/arm/mach-axxia/platsmp.c create mode 100644 drivers/clk/clk-axxia.c diff --git a/Documentation/devicetree/bindings/arm/axxia.txt b/Documentation/devicetree/bindings/arm/axxia.txt new file mode 100644 index 0000000..7b4ef9c --- /dev/null +++ b/Documentation/devicetree/bindings/arm/axxia.txt @@ -0,0 +1,12 @@ +Axxia AXM55xx device tree bindings + +Boards using the AXM55xx SoC need to have the following properties: + +Required root node property: + + - compatible = "lsi,axm5516" + +Boards: + + LSI AXM5516 Validation board (Amarillo) + compatible = "lsi,axm5516-amarillo", "lsi,axm5516" diff --git a/Documentation/devicetree/bindings/clock/clk-axxia.txt b/Documentation/devicetree/bindings/clock/clk-axxia.txt new file mode 100644 index 0000000..94eb840 --- /dev/null +++ b/Documentation/devicetree/bindings/clock/clk-axxia.txt @@ -0,0 +1,88 @@ +* Axxia clock driver bindings + +This document contains bindings for the Axxia PLL, clock divider and mux. The +binding uses the common clock binding found in: + + Documentation/devicetree/bindings/clock/clock-bindings.txt + +* Axxia PLL + +Required properties: +- #clock-cells : shall be set to 0. +- compatible : shall be "lsi,axxia-pll-clock" +- clocks : parent clock phandle +- reg - specifies PLL control registers + +Optional properties: +- clock-output-names : From common clock binding. + +Example: + + clk_ref0: clk_ref0 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + clk_cpu_pll: clk_cpu_pll at 2010022000 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>; + clock-output-names = "clk_cpu_pll"; + reg = <0x20 0x10022000 0 0x2c>; + }; + + +* Axxia clock divider + +Required properties: +- #clock-cells : shall be set to 0. +- compatible : shall be "lsi,axxia-div-clock" +- clocks : parent clock phandle +- reg - clock divider control register +- lsi,bit-shift : bit position of the divider control field. +- lsi,bit-mask : width of the divider control field. + +Optional properties: +- clock-output-names : From common clock binding. + +Example: + + clk_cpu_div: cpu { + compatible = "lsi,axxia-div-clock"; + #clock-cells = <0>; + clocks = <&clk_cpu_pll>; + clock-output-names = "clk_cpu_div"; + reg = <0x20 0x10030008 0 0x4>; + lsi,bit-shift = <0>; + lsi,bit-width = <4>; + }; + + +* Axxia clock mux + +Required properties: +- #clock-cells : shall be set to 0. +- compatible : shall be "lsi,axxia-mux-clock" +- clocks : phandle of parent clocks (max 4) +- reg - clock divider control register +- lsi,bit-shift : bit position of the divider control field. +- lsi,bit-mask : width of the divider control field. + +Optional properties: +- clock-output-names : From common clock binding. + +Example: + + clk_cpu: clk_cpu0 at 2010030000 { + compatible = "lsi,axxia-mux-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>, + <&clk_cpu_pll>, + <&clk_cpu_div>, + <&clk_cpu_div>; + clock-output-names = "clk_cpu"; + reg = <0x20 0x10030000 0 0x4>; + lsi,bit-shift = <0>; + lsi,bit-width = <2>; + }; diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 1594945..ef87f59 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -934,6 +934,8 @@ source "arch/arm/mach-mvebu/Kconfig" source "arch/arm/mach-at91/Kconfig" +source "arch/arm/mach-axxia/Kconfig" + source "arch/arm/mach-bcm/Kconfig" source "arch/arm/mach-bcm2835/Kconfig" diff --git a/arch/arm/Makefile b/arch/arm/Makefile index 08a9ef5..b034687 100644 --- a/arch/arm/Makefile +++ b/arch/arm/Makefile @@ -138,10 +138,12 @@ endif textofs-$(CONFIG_ARCH_MSM7X30) := 0x00208000 textofs-$(CONFIG_ARCH_MSM8X60) := 0x00208000 textofs-$(CONFIG_ARCH_MSM8960) := 0x00208000 +textofs-$(CONFIG_ARCH_AXXIA) := 0x00308000 # Machine directory name. This list is sorted alphanumerically # by CONFIG_* macro name. machine-$(CONFIG_ARCH_AT91) += at91 +machine-$(CONFIG_ARCH_AXXIA) += axxia machine-$(CONFIG_ARCH_BCM) += bcm machine-$(CONFIG_ARCH_BCM2835) += bcm2835 machine-$(CONFIG_ARCH_BERLIN) += berlin diff --git a/arch/arm/mach-axxia/Kconfig b/arch/arm/mach-axxia/Kconfig new file mode 100644 index 0000000..336426a --- /dev/null +++ b/arch/arm/mach-axxia/Kconfig @@ -0,0 +1,19 @@ +config ARCH_AXXIA + bool "LSI Axxia platforms" if (ARCH_MULTI_V7 && ARM_LPAE) + select ARM_GIC + select HAVE_SMP + select MFD_SYSCON + select ARM_AMBA + select ARCH_WANT_OPTIONAL_GPIOLIB + select HAVE_ARM_ARCH_TIMER + select ARM_TIMER_SP804 + select ZONE_DMA + select ARCH_DMA_ADDR_T_64BIT + select ARCH_SUPPORTS_BIG_ENDIAN + select MIGHT_HAVE_PCI + select PCI_DOMAINS if PCI + help + This enables support for the LSI Axxia devices. + + The LSI Axxia platforms require a Flattened Device Tree to be passed + to the kernel. diff --git a/arch/arm/mach-axxia/Makefile b/arch/arm/mach-axxia/Makefile new file mode 100644 index 0000000..4daf118 --- /dev/null +++ b/arch/arm/mach-axxia/Makefile @@ -0,0 +1,2 @@ +obj-y += axxia.o +obj-$(CONFIG_SMP) += platsmp.o headsmp.o diff --git a/arch/arm/mach-axxia/Makefile.boot b/arch/arm/mach-axxia/Makefile.boot new file mode 100644 index 0000000..b3c724e --- /dev/null +++ b/arch/arm/mach-axxia/Makefile.boot @@ -0,0 +1,2 @@ + zreladdr-y += 0x00308000 +params_phys-y := 0x00300100 diff --git a/arch/arm/mach-axxia/axxia.c b/arch/arm/mach-axxia/axxia.c new file mode 100644 index 0000000..8d85926 --- /dev/null +++ b/arch/arm/mach-axxia/axxia.c @@ -0,0 +1,117 @@ +/* + * Support for the LSI Axxia SoC devices based on ARM cores. + * + * Copyright (C) 2012 LSI + * + * 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; either version 2 of the License, or + * (at your option) any later version. + * + * 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, see <http://www.gnu.org/licenses/>. + */ + +#include <linux/amba/bus.h> +#include <linux/device.h> +#include <linux/dma-mapping.h> +#include <linux/init.h> +#include <linux/io.h> +#include <linux/interrupt.h> +#include <linux/kernel.h> +#include <linux/of_platform.h> +#include <linux/sizes.h> +#include <asm/mach/arch.h> +#include <asm/pmu.h> +#include "axxia.h" + +/* + * The PMU IRQ lines of four cores are wired together into a single interrupt. + * Bounce the interrupt to other cores if it's not ours. + */ +static irqreturn_t axxia_pmu_handler(int irq, void *dev, irq_handler_t handler) +{ + irqreturn_t ret = handler(irq, dev); + + if (ret == IRQ_NONE) { + int cpu = smp_processor_id(); + int cluster = cpu / CORES_PER_CLUSTER; + int other; + + /* Look until we find another cpu that's online. */ + do { + other = (++cpu % CORES_PER_CLUSTER) + + (cluster * CORES_PER_CLUSTER); + } while (!cpu_online(other)); + + irq_set_affinity(irq, cpumask_of(other)); + } + + /* + * We should be able to get away with the amount of IRQ_NONEs we give, + * while still having the spurious IRQ detection code kick in if the + * interrupt really starts hitting spuriously. + */ + return ret; +} + +static struct arm_pmu_platdata pmu_pdata = { + .handle_irq = axxia_pmu_handler, +}; + +static struct of_dev_auxdata axxia_auxdata_lookup[] __initdata = { + OF_DEV_AUXDATA("arm,cortex-a15-pmu", 0, "pmu", &pmu_pdata), + {} +}; + + +static int +axxia_bus_notifier(struct notifier_block *nb, unsigned long event, void *obj) +{ + struct device *dev = obj; + + if (event != BUS_NOTIFY_ADD_DEVICE) + return NOTIFY_DONE; + + if (!of_property_read_bool(dev->of_node, "dma-coherent")) + return NOTIFY_DONE; + + set_dma_ops(dev, &arm_coherent_dma_ops); + + return NOTIFY_OK; +} + +static struct notifier_block axxia_platform_nb = { + .notifier_call = axxia_bus_notifier, +}; + +static struct notifier_block axxia_amba_nb = { + .notifier_call = axxia_bus_notifier, +}; + +static void __init axxia_dt_init(void) +{ + bus_register_notifier(&platform_bus_type, &axxia_platform_nb); + bus_register_notifier(&amba_bustype, &axxia_amba_nb); + + of_platform_populate(NULL, of_default_bus_match_table, + axxia_auxdata_lookup, NULL); +} + +static const char *axxia_dt_match[] __initconst = { + "lsi,axm55xx", + "lsi,axm55xx-sim", + "lsi,axm55xx-emu", + NULL +}; + +DT_MACHINE_START(AXXIA_DT, "LSI Axxia AXM55XX") + .dt_compat = axxia_dt_match, + .smp = smp_ops(axxia_smp_ops), + .init_machine = axxia_dt_init, +MACHINE_END diff --git a/arch/arm/mach-axxia/axxia.h b/arch/arm/mach-axxia/axxia.h new file mode 100644 index 0000000..594dd97 --- /dev/null +++ b/arch/arm/mach-axxia/axxia.h @@ -0,0 +1,42 @@ +/* + * Prototypes for platform functions. + * + * + * 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; either version 2 of the License, or + * (at your option) any later version. + * + * 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, see <http://www.gnu.org/licenses/>. + */ +#ifndef __AXXIA_H +#define __AXXIA_H + +#define CORES_PER_CLUSTER 4 + +#define AXXIA_PERIPH_PHYS 0x2000000000ULL +#define AXXIA_SYSCON_PHYS 0x2010030000ULL + +#if 0 +#define AXXIA_UART0_PHYS 0x2010080000ULL +#define AXXIA_UART1_PHYS 0x2010081000ULL +#define AXXIA_UART2_PHYS 0x2010082000ULL +#define AXXIA_UART3_PHYS 0x2010083000ULL + +#ifdef CONFIG_DEBUG_LL +#define AXXIA_DEBUG_UART_VIRT 0xf0080000 +#define AXXIA_DEBUG_UART_PHYS AXXIA_UART0_PHYS +#endif +#endif + +void axxia_secondary_startup(void); + +extern struct smp_operations axxia_smp_ops; + +#endif diff --git a/arch/arm/mach-axxia/headsmp.S b/arch/arm/mach-axxia/headsmp.S new file mode 100644 index 0000000..296cc53 --- /dev/null +++ b/arch/arm/mach-axxia/headsmp.S @@ -0,0 +1,44 @@ +/* + * linux/arch/arm/mach-axxia/headsmp.S + * + * Cloned from linux/arch/arm/mach-realview/headsmp.S + * + * Copyright (c) 2003 ARM Limited + * All Rights Reserved + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include <linux/linkage.h> +#include <linux/init.h> +#include <asm/assembler.h> + +/* + * Axxia specific entry point for secondary CPUs. This provides + * a "holding pen" into which all secondary cores are held until we're + * ready for them to initialise. + */ +ENTRY(axxia_secondary_startup) +ARM_BE8(setend be) /* go BE8 if entered LE */ + mrc p15, 0, r0, c0, c0, 5 + bic r0, #0xff000000 + adr r4, 2f + ldmia r4, {r5, r6} + sub r4, r4, r5 + add r6, r6, r4 +1: ldr r7, [r6] + cmp r7, r0 + wfene + bne 1b + + /* + * We've been released from the holding pen: secondary_stack + * should now contain the SVC stack for this core + */ + b secondary_startup +ENDPROC(axxia_secondary_startup) + + .align 2 +2: .long . + .long pen_release diff --git a/arch/arm/mach-axxia/platsmp.c b/arch/arm/mach-axxia/platsmp.c new file mode 100644 index 0000000..fd7f507 --- /dev/null +++ b/arch/arm/mach-axxia/platsmp.c @@ -0,0 +1,183 @@ +/* + * linux/arch/arm/mach-axxia/platsmp.c + * + * Copyright (C) 2012 LSI Corporation + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + */ +#include <linux/init.h> +#include <linux/errno.h> +#include <linux/smp.h> +#include <linux/delay.h> +#include <linux/device.h> +#include <linux/io.h> +#include <linux/jiffies.h> +#include <linux/of.h> +#include <linux/of_fdt.h> +#include <asm/smp_plat.h> +#include <asm/cacheflush.h> +#include <asm/mach/map.h> +#include <asm/virt.h> +#include "axxia.h" + +static int wfe_fixup; + +/* + * Check if we need to enable cross-cluster SEV workaround for a bug in + * revision 1.0 silicon (which could cause event signals (from SEV instruction) + * to get lost between clusters). As a workaround, we set the bit 7 in CP15 + * ACTLR (enable WFE as a no-operation) for each core. + */ +static void __init check_fixup_sev(void __iomem *syscon) +{ + u32 pfuse = readl(syscon + 0x34); + u32 chip_type = pfuse & 0x1f; + u32 chip_ver = (pfuse >> 5) & 0x3f; + + /* Set flag for secondary cores */ + wfe_fixup = (chip_type == 0 || chip_type == 9) && (chip_ver == 0); + pr_info("axxia: Cross-cluster SEV fixup: %s\n", wfe_fixup ? "yes":"no"); +} + +static void __init do_fixup_sev(void) +{ + u32 tmp; + + if (wfe_fixup) { + asm volatile("mrc\tp15, 0, %0, c1, c0, 1" : "=r"(tmp)); + tmp |= (1<<7); + asm volatile("mcr\tp15, 0, %0, c1, c0, 1" : : "r"(tmp)); + isb(); + } +} + +/* + * Write pen_release in a way that is guaranteed to be visible to all + * observers, irrespective of whether they're taking part in coherency + * or not. This is necessary for the hotplug code to work reliably. + */ +static void write_pen_release(int val) +{ + pen_release = val; + /* Make sure this store is visible to other CPUs */ + smp_wmb(); + __cpuc_flush_dcache_area((void *)&pen_release, sizeof(pen_release)); +} + +/* + * Write the entry point of the routine that the secondary cores will execute + * once they are released from their "holding pen". + */ +static void write_release_addr(u32 release_phys) +{ + u32 *virt = (u32 *) phys_to_virt(release_phys); + writel_relaxed(virt_to_phys(axxia_secondary_startup), virt); + /* Make sure this store is visible to other CPUs */ + smp_wmb(); + __cpuc_flush_dcache_area(virt, sizeof(u32)); +} + +static void axxia_secondary_init(unsigned int cpu) +{ + /* Fixup for cross-cluster SEV */ + do_fixup_sev(); + + /* + * Let the primary processor know we're out of the + * pen, then head off into the C entry point. + */ + write_pen_release(-1); +} + +static int axxia_boot_secondary(unsigned int cpu, + struct task_struct *idle) +{ + unsigned long timeout; + + /* Release the specified core */ + write_pen_release(cpu_logical_map(cpu)); + + /* Send a wakeup event to get the idled cpu out of WFE state */ + dsb_sev(); + + /* Wait for so long, then give up if nothing happens ... */ + timeout = jiffies + (1 * HZ); + while (time_before(jiffies, timeout)) { + /* Make sure stores to pen_release have completed */ + smp_rmb(); + if (pen_release == -1) + break; + udelay(1); + } + + return pen_release != -1 ? -ENOSYS : 0; +} + +static void __init axxia_smp_prepare_cpus(unsigned int max_cpus) +{ + void __iomem *syscon; + int cpu_count = 0; + int cpu; + + syscon = ioremap(AXXIA_SYSCON_PHYS, SZ_64K); + if (WARN_ON(!syscon)) + return; + + check_fixup_sev(syscon); + do_fixup_sev(); + + /* + * Initialise the present map, which describes the set of CPUs actually + * populated@the present time. + */ + for_each_possible_cpu(cpu) { + struct device_node *np; + u32 release_phys; + + np = of_get_cpu_node(cpu, NULL); + if (!np) + continue; + if (of_property_read_u32(np, "cpu-release-addr", &release_phys)) + continue; + + /* + * Release all physical cpus when not in hyp mode since we + * might want to bring them online later. + * + * Also we need to get the execution into kernel code (it's + * currently executing in u-boot). u-boot releases the cores + * from reset in hyp mode. + */ + if (!is_hyp_mode_available()) { + if (cpu != 0) { + u32 tmp = readl(syscon + 0x1010); + writel(0xab, syscon + 0x1000); + tmp &= ~(1 << cpu); + writel(tmp, syscon + 0x1010); + } + } + + if (cpu_count < max_cpus) { + set_cpu_present(cpu, true); + cpu_count++; + } + + /* + * This is the entry point of the routine that the secondary + * cores will execute once they are released from their + * "holding pen". + */ + if (release_phys != 0) + write_release_addr(release_phys); + } + + iounmap(syscon); +} + +struct smp_operations axxia_smp_ops __initdata = { + .smp_prepare_cpus = axxia_smp_prepare_cpus, + .smp_secondary_init = axxia_secondary_init, + .smp_boot_secondary = axxia_boot_secondary, +}; diff --git a/drivers/clk/Makefile b/drivers/clk/Makefile index a367a98..18507ad 100644 --- a/drivers/clk/Makefile +++ b/drivers/clk/Makefile @@ -12,6 +12,7 @@ obj-$(CONFIG_COMMON_CLK) += clk-composite.o # hardware specific clock types # please keep this section sorted lexicographically by file/directory path name obj-$(CONFIG_COMMON_CLK_AXI_CLKGEN) += clk-axi-clkgen.o +obj-$(CONFIG_ARCH_AXXIA) += clk-axxia.o obj-$(CONFIG_ARCH_BCM2835) += clk-bcm2835.o obj-$(CONFIG_ARCH_EFM32) += clk-efm32gg.o obj-$(CONFIG_ARCH_HIGHBANK) += clk-highbank.o diff --git a/drivers/clk/clk-axxia.c b/drivers/clk/clk-axxia.c new file mode 100644 index 0000000..996b8f2 --- /dev/null +++ b/drivers/clk/clk-axxia.c @@ -0,0 +1,281 @@ +/* + * arch/arm/mach-axxia/clock.c + * + * Provides clock implementations for three different types of clock devices on + * the Axxia device: PLL clock, a clock divider and a clock mux. + * + * Copyright (C) 2012 LSI + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of the GNU General Public License version 2 as published by + * the Free Software Foundation. + */ +#include <linux/module.h> +#include <linux/kernel.h> +#include <linux/slab.h> +#include <linux/of.h> +#include <linux/of_address.h> +#include <linux/clk-provider.h> + + +static void __iomem * +parse_and_map_reg(struct device_node *node, u32 *shift, u32 *width) +{ + void __iomem *reg; + + if (of_property_read_u32(node, "lsi,bit-shift", shift)) { + pr_err("%s: no 'lsi,bit-shift' property\n", node->name); + return NULL; + } + + if (of_property_read_u32(node, "lsi,bit-width", width)) { + pr_err("%s: no 'lsi,bit-width' property\n", node->name); + return NULL; + } + + reg = of_iomap(node, 0); + if (!reg) + pr_err("%s: failed to map register\n", node->name); + + return reg; +} + +/** + * struct axxia_pllclk - Axxia PLL generated clock. + * @hw: clk_hw for the PLL + * @regs: Pointer to control registers + */ +struct axxia_pllclk { + struct clk_hw hw; + void __iomem *regs; +}; + +#define to_axxia_pll(_hw) container_of(_hw, struct axxia_pllclk, hw) + +static unsigned long +axxia_pllclk_recalc(struct clk_hw *hw, unsigned long parent_rate) +{ + struct axxia_pllclk *pll = to_axxia_pll(hw); + unsigned long rate, fbdiv, refdiv, postdiv; + u32 control; + + control = readl(pll->regs); + postdiv = ((control >> 0) & 0xf) + 1; + fbdiv = ((control >> 4) & 0xfff) + 3; + refdiv = ((control >> 16) & 0x1f) + 1; + rate = (parent_rate / (refdiv * postdiv)) * fbdiv; + + return rate; +} + +static const struct clk_ops clk_pll_ops = { + .recalc_rate = axxia_pllclk_recalc, +}; + +static struct clk * +register_axxia_pllclk(struct device *dev, const char *name, + const char *parent_name, void __iomem *regs) +{ + struct clk_init_data init; + struct axxia_pllclk *pll; + struct clk *clk; + + pll = kzalloc(sizeof(*pll), GFP_KERNEL); + if (!pll) + return ERR_PTR(-ENOMEM); + + init.name = name; + init.ops = &clk_pll_ops; + init.flags = 0; + init.parent_names = &parent_name; + init.num_parents = 1; + + pll->regs = regs; + pll->hw.init = &init; + + clk = clk_register(NULL, &pll->hw); + if (IS_ERR(clk)) + kfree(pll); + + return clk; +} + + +/** + * struct axxia_divclk - Axxia clock divider + * @hw: clk_hw for the divider + * @regs: Pointer to control registers + */ +struct axxia_divclk { + struct clk_hw hw; + void __iomem *regs; + int shift; + int width; +}; + +#define to_axxia_divclk(_hw) container_of(_hw, struct axxia_divclk, hw) + +static unsigned long +axxia_divclk_recalc_rate(struct clk_hw *hw, unsigned long parent_rate) +{ + struct axxia_divclk *divclk = to_axxia_divclk(hw); + u32 ctrl, div; + + ctrl = readl(divclk->regs); + div = 1 + ((ctrl >> divclk->shift) & GENMASK(divclk->width-1, 0)); + + return parent_rate / div; +} + +const struct clk_ops axxia_divclk_ops = { + .recalc_rate = axxia_divclk_recalc_rate, +}; + +static struct clk * +register_axxia_divclk(struct device *dev, const char *name, + const char *parent_name, void __iomem *regs, + int shift, int width) +{ + struct axxia_divclk *div; + struct clk *clk; + struct clk_init_data init; + + div = kzalloc(sizeof(struct axxia_divclk), GFP_KERNEL); + if (!div) + return ERR_PTR(-ENOMEM); + + init.name = name; + init.ops = &axxia_divclk_ops; + init.flags = CLK_IS_BASIC; + init.parent_names = &parent_name; + init.num_parents = 1; + + div->regs = regs; + div->shift = shift; + div->width = width; + div->hw.init = &init; + + /* register the clock */ + clk = clk_register(dev, &div->hw); + if (IS_ERR(clk)) + kfree(div); + + return clk; +} + +/** + * axxia_pll_clk_init - Axxia PLL clock setup function + * @node: pointer to device tree node + */ +static void __init +axxia_pll_clk_init(struct device_node *node) +{ + struct clk *clk; + const char *parent_name; + void __iomem *regs; + const char *clk_name = node->name; + + of_property_read_string(node, "clock-output-names", &clk_name); + + parent_name = of_clk_get_parent_name(node, 0); + if (!parent_name) { + pr_err("%s: no parent clock\n", node->name); + return; + } + + regs = of_iomap(node, 0); + if (!regs) { + pr_err("%s: failed to map register\n", node->name); + return; + } + + clk = register_axxia_pllclk(NULL, clk_name, parent_name, regs); + if (IS_ERR(clk)) { + pr_err("%s: register failed (%ld)\n", clk_name, PTR_ERR(clk)); + iounmap(regs); + return; + } + + of_clk_add_provider(node, of_clk_src_simple_get, clk); +} +CLK_OF_DECLARE(axxia_pll_clk, "lsi,axxia-pll-clock", axxia_pll_clk_init); + + +/** + * axxia_div_clk_init - Axxia clock divider setup function + * @node: device tree node for this clock + */ +static void __init +axxia_div_clk_init(struct device_node *node) +{ + const char *parent_name; + void __iomem *reg; + u32 shift, width; + struct clk *clk; + const char *clk_name = node->name; + + of_property_read_string(node, "clock-output-names", &clk_name); + + parent_name = of_clk_get_parent_name(node, 0); + if (!parent_name) { + pr_err("%s: no parent clock\n", node->name); + return; + } + + reg = parse_and_map_reg(node, &shift, &width); + if (!reg) + return; + + clk = register_axxia_divclk(NULL, clk_name, parent_name, reg, + shift, width); + if (IS_ERR(clk)) { + pr_err("%s: register failed (%ld)\n", clk_name, PTR_ERR(clk)); + iounmap(reg); + return; + } + + of_clk_add_provider(node, of_clk_src_simple_get, clk); +} +CLK_OF_DECLARE(axxia_div_clk, "lsi,axxia-div-clock", axxia_div_clk_init); + + +/** + * axxia_mux_clk_init - Clock mux setup function + * @node: device tree node for this clock + */ +static void __init +axxia_mux_clk_init(struct device_node *node) +{ + void __iomem *reg; + struct clk *clk; + const char *parents[4]; + int num_parents; + const char *clk_name = node->name; + u32 shift, width; + int i; + + of_property_read_string(node, "clock-output-names", &clk_name); + + for (i = 0; i < ARRAY_SIZE(parents); ++i) { + parents[i] = of_clk_get_parent_name(node, i); + if (!parents[i]) + break; + } + num_parents = i; + + reg = parse_and_map_reg(node, &shift, &width); + if (!reg) + return; + + clk = clk_register_mux(NULL, clk_name, parents, + num_parents, 0, reg, shift, width, + CLK_MUX_READ_ONLY, NULL); + if (IS_ERR(clk)) { + pr_err("%s: register failed (%ld)\n", clk_name, PTR_ERR(clk)); + iounmap(reg); + return; + } + + of_clk_add_provider(node, of_clk_src_simple_get, clk); +} +CLK_OF_DECLARE(pll_mux_clock, "lsi,axxia-mux-clock", axxia_mux_clk_init); -- 1.8.3.2 ^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC 2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg @ 2014-04-15 12:30 ` Arnd Bergmann 2014-04-24 8:58 ` Anders Berg 2014-04-24 9:04 ` Russell King - ARM Linux 1 sibling, 1 reply; 17+ messages in thread From: Arnd Bergmann @ 2014-04-15 12:30 UTC (permalink / raw) To: linux-arm-kernel On Tuesday 15 April 2014 14:06:10 Anders Berg wrote: > diff --git a/arch/arm/mach-axxia/Kconfig b/arch/arm/mach-axxia/Kconfig > new file mode 100644 > index 0000000..336426a > --- /dev/null > +++ b/arch/arm/mach-axxia/Kconfig > @@ -0,0 +1,19 @@ > +config ARCH_AXXIA > + bool "LSI Axxia platforms" if (ARCH_MULTI_V7 && ARM_LPAE) > + select ARM_GIC > + select HAVE_SMP > + select MFD_SYSCON > + select ARM_AMBA > + select ARCH_WANT_OPTIONAL_GPIOLIB > + select HAVE_ARM_ARCH_TIMER > + select ARM_TIMER_SP804 > + select ZONE_DMA > + select ARCH_DMA_ADDR_T_64BIT > + select ARCH_SUPPORTS_BIG_ENDIAN > + select MIGHT_HAVE_PCI > + select PCI_DOMAINS if PCI No need to select HAVE_SMP or ARCH_WANT_OPTIONAL_GPIOLIB any more. We should rethink the ARCH_SUPPORTS_BIG_ENDIAN option I guess. It makes little sense to select that from one platform in a multiplatform build if other platforms don't support it. > --- /dev/null > +++ b/arch/arm/mach-axxia/Makefile.boot > @@ -0,0 +1,2 @@ > + zreladdr-y += 0x00308000 > +params_phys-y := 0x00300100 This won't be used, since AUTO_ZRELADDR is set by ARCH_MULTIPLATFORM. > diff --git a/arch/arm/mach-axxia/axxia.c b/arch/arm/mach-axxia/axxia.c > new file mode 100644 > index 0000000..8d85926 > --- /dev/null > +++ b/arch/arm/mach-axxia/axxia.c > @@ -0,0 +1,117 @@ > +/* > + * Support for the LSI Axxia SoC devices based on ARM cores. > + * > + * Copyright (C) 2012 LSI > + * > + * 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; either version 2 of the License, or > + * (at your option) any later version. > + * > + * 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, see <http://www.gnu.org/licenses/>. > + */ > + > +#include <linux/amba/bus.h> > +#include <linux/device.h> > +#include <linux/dma-mapping.h> > +#include <linux/init.h> > +#include <linux/io.h> > +#include <linux/interrupt.h> > +#include <linux/kernel.h> > +#include <linux/of_platform.h> > +#include <linux/sizes.h> > +#include <asm/mach/arch.h> > +#include <asm/pmu.h> > +#include "axxia.h" > + > +/* > + * The PMU IRQ lines of four cores are wired together into a single interrupt. > + * Bounce the interrupt to other cores if it's not ours. > + */ > +static irqreturn_t axxia_pmu_handler(int irq, void *dev, irq_handler_t handler) > +{ > + irqreturn_t ret = handler(irq, dev); > + > + if (ret == IRQ_NONE) { > + int cpu = smp_processor_id(); > + int cluster = cpu / CORES_PER_CLUSTER; > + int other; > + > + /* Look until we find another cpu that's online. */ > + do { > + other = (++cpu % CORES_PER_CLUSTER) + > + (cluster * CORES_PER_CLUSTER); > + } while (!cpu_online(other)); > + > + irq_set_affinity(irq, cpumask_of(other)); > + } > + > + /* > + * We should be able to get away with the amount of IRQ_NONEs we give, > + * while still having the spurious IRQ detection code kick in if the > + * interrupt really starts hitting spuriously. > + */ > + return ret; > +} > + > +static struct arm_pmu_platdata pmu_pdata = { > + .handle_irq = axxia_pmu_handler, > +}; > + > +static struct of_dev_auxdata axxia_auxdata_lookup[] __initdata = { > + OF_DEV_AUXDATA("arm,cortex-a15-pmu", 0, "pmu", &pmu_pdata), > + {} > +}; This looks similar to what we have in mach-ux500 as db8500_pmu_handler. To avoid duplication, I'd prefer moving support for this into the perf_event code itself, and get rid of the auxdata. > +static int > +axxia_bus_notifier(struct notifier_block *nb, unsigned long event, void *obj) > +{ > + struct device *dev = obj; > + > + if (event != BUS_NOTIFY_ADD_DEVICE) > + return NOTIFY_DONE; > + > + if (!of_property_read_bool(dev->of_node, "dma-coherent")) > + return NOTIFY_DONE; > + > + set_dma_ops(dev, &arm_coherent_dma_ops); > + > + return NOTIFY_OK; > +} > + > +static struct notifier_block axxia_platform_nb = { > + .notifier_call = axxia_bus_notifier, > +}; > + > +static struct notifier_block axxia_amba_nb = { > + .notifier_call = axxia_bus_notifier, > +}; And we definitely want to do this in a generic fashion. That should not be in platform specific code. > +static const char *axxia_dt_match[] __initconst = { > + "lsi,axm55xx", > + "lsi,axm55xx-sim", > + "lsi,axm55xx-emu", > + NULL > +}; Please no 'xx' in compatible strings. You can list all the relevant model numbers here, or require that machines list a specific model as compatible, e.g. compatible = "lsi,axm5510-sim", "lsi,axm5510", "lsi,axm5502"; > + > +DT_MACHINE_START(AXXIA_DT, "LSI Axxia AXM55XX") > + .dt_compat = axxia_dt_match, > + .smp = smp_ops(axxia_smp_ops), > + .init_machine = axxia_dt_init, > +MACHINE_END Please have a look at how smp ops are hooked up in mach-qcom and see if you can do it that way as well. > diff --git a/arch/arm/mach-axxia/axxia.h b/arch/arm/mach-axxia/axxia.h > new file mode 100644 > index 0000000..594dd97 > --- /dev/null > +++ b/arch/arm/mach-axxia/axxia.h > @@ -0,0 +1,42 @@ > +/* > + * Prototypes for platform functions. > + * > + * > + * 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; either version 2 of the License, or > + * (at your option) any later version. > + * > + * 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, see <http://www.gnu.org/licenses/>. > + */ > +#ifndef __AXXIA_H > +#define __AXXIA_H > + > +#define CORES_PER_CLUSTER 4 > + > +#define AXXIA_PERIPH_PHYS 0x2000000000ULL > +#define AXXIA_SYSCON_PHYS 0x2010030000ULL Remove these -- they have to come from DT. > +#if 0 > +#define AXXIA_UART0_PHYS 0x2010080000ULL > +#define AXXIA_UART1_PHYS 0x2010081000ULL > +#define AXXIA_UART2_PHYS 0x2010082000ULL > +#define AXXIA_UART3_PHYS 0x2010083000ULL > + > +#ifdef CONFIG_DEBUG_LL > +#define AXXIA_DEBUG_UART_VIRT 0xf0080000 > +#define AXXIA_DEBUG_UART_PHYS AXXIA_UART0_PHYS > +#endif > +#endif You can put these constants directly into debug code. > diff --git a/arch/arm/mach-axxia/platsmp.c b/arch/arm/mach-axxia/platsmp.c > new file mode 100644 > index 0000000..fd7f507 > --- /dev/null > +++ b/arch/arm/mach-axxia/platsmp.c > @@ -0,0 +1,183 @@ > +/* > + * linux/arch/arm/mach-axxia/platsmp.c > + * > + * Copyright (C) 2012 LSI Corporation > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License version 2 as > + * published by the Free Software Foundation. Have you considered using PSCI to bring up the secondary CPUs? > +static int axxia_boot_secondary(unsigned int cpu, > + struct task_struct *idle) > +{ > + unsigned long timeout; > + > + /* Release the specified core */ > + write_pen_release(cpu_logical_map(cpu)); > + > + /* Send a wakeup event to get the idled cpu out of WFE state */ > + dsb_sev(); > + > + /* Wait for so long, then give up if nothing happens ... */ > + timeout = jiffies + (1 * HZ); > + while (time_before(jiffies, timeout)) { > + /* Make sure stores to pen_release have completed */ > + smp_rmb(); > + if (pen_release == -1) > + break; > + udelay(1); > + } > + > + return pen_release != -1 ? -ENOSYS : 0; > +} This is pretty sad. No hardware support for waking up CPUs? > +static void __init axxia_smp_prepare_cpus(unsigned int max_cpus) > +{ > + void __iomem *syscon; > + int cpu_count = 0; > + int cpu; > + > + syscon = ioremap(AXXIA_SYSCON_PHYS, SZ_64K); > + if (WARN_ON(!syscon)) > + return; > + > + check_fixup_sev(syscon); Please use the syscon driver and regmap. That driver might need a little help to get syscon_node_to_regmap() to work before driver initialization, but other people need the same thing, and we should just implement it. > diff --git a/drivers/clk/clk-axxia.c b/drivers/clk/clk-axxia.c > new file mode 100644 > index 0000000..996b8f2 > --- /dev/null > +++ b/drivers/clk/clk-axxia.c > @@ -0,0 +1,281 @@ > +/* > + * arch/arm/mach-axxia/clock.c This should be a separate patch, and get merged by the clk maintainer. Arnd ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC 2014-04-15 12:30 ` Arnd Bergmann @ 2014-04-24 8:58 ` Anders Berg 0 siblings, 0 replies; 17+ messages in thread From: Anders Berg @ 2014-04-24 8:58 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 02:30:55PM +0200, Arnd Bergmann wrote: > On Tuesday 15 April 2014 14:06:10 Anders Berg wrote: > > > diff --git a/arch/arm/mach-axxia/Kconfig b/arch/arm/mach-axxia/Kconfig > > new file mode 100644 > > index 0000000..336426a > > --- /dev/null > > +++ b/arch/arm/mach-axxia/Kconfig > > @@ -0,0 +1,19 @@ > > +config ARCH_AXXIA > > + bool "LSI Axxia platforms" if (ARCH_MULTI_V7 && ARM_LPAE) > > + select ARM_GIC > > + select HAVE_SMP > > + select MFD_SYSCON > > + select ARM_AMBA > > + select ARCH_WANT_OPTIONAL_GPIOLIB > > + select HAVE_ARM_ARCH_TIMER > > + select ARM_TIMER_SP804 > > + select ZONE_DMA > > + select ARCH_DMA_ADDR_T_64BIT > > + select ARCH_SUPPORTS_BIG_ENDIAN > > + select MIGHT_HAVE_PCI > > + select PCI_DOMAINS if PCI > > No need to select HAVE_SMP or ARCH_WANT_OPTIONAL_GPIOLIB any more. OK. > > We should rethink the ARCH_SUPPORTS_BIG_ENDIAN option I guess. It makes > little sense to select that from one platform in a multiplatform build > if other platforms don't support it. > > > --- /dev/null > > +++ b/arch/arm/mach-axxia/Makefile.boot > > @@ -0,0 +1,2 @@ > > + zreladdr-y += 0x00308000 > > +params_phys-y := 0x00300100 > > This won't be used, since AUTO_ZRELADDR is set by ARCH_MULTIPLATFORM. OK. > > > diff --git a/arch/arm/mach-axxia/axxia.c b/arch/arm/mach-axxia/axxia.c > > new file mode 100644 > > index 0000000..8d85926 > > --- /dev/null > > +++ b/arch/arm/mach-axxia/axxia.c > > @@ -0,0 +1,117 @@ > > +/* > > + * Support for the LSI Axxia SoC devices based on ARM cores. > > + * > > + * Copyright (C) 2012 LSI > > + * > > + * 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; either version 2 of the License, or > > + * (at your option) any later version. > > + * > > + * 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, see <http://www.gnu.org/licenses/>. > > + */ > > + > > +#include <linux/amba/bus.h> > > +#include <linux/device.h> > > +#include <linux/dma-mapping.h> > > +#include <linux/init.h> > > +#include <linux/io.h> > > +#include <linux/interrupt.h> > > +#include <linux/kernel.h> > > +#include <linux/of_platform.h> > > +#include <linux/sizes.h> > > +#include <asm/mach/arch.h> > > +#include <asm/pmu.h> > > +#include "axxia.h" > > + > > +/* > > + * The PMU IRQ lines of four cores are wired together into a single interrupt. > > + * Bounce the interrupt to other cores if it's not ours. > > + */ > > +static irqreturn_t axxia_pmu_handler(int irq, void *dev, irq_handler_t handler) > > +{ > > + irqreturn_t ret = handler(irq, dev); > > + > > + if (ret == IRQ_NONE) { > > + int cpu = smp_processor_id(); > > + int cluster = cpu / CORES_PER_CLUSTER; > > + int other; > > + > > + /* Look until we find another cpu that's online. */ > > + do { > > + other = (++cpu % CORES_PER_CLUSTER) + > > + (cluster * CORES_PER_CLUSTER); > > + } while (!cpu_online(other)); > > + > > + irq_set_affinity(irq, cpumask_of(other)); > > + } > > + > > + /* > > + * We should be able to get away with the amount of IRQ_NONEs we give, > > + * while still having the spurious IRQ detection code kick in if the > > + * interrupt really starts hitting spuriously. > > + */ > > + return ret; > > +} > > + > > +static struct arm_pmu_platdata pmu_pdata = { > > + .handle_irq = axxia_pmu_handler, > > +}; > > + > > +static struct of_dev_auxdata axxia_auxdata_lookup[] __initdata = { > > + OF_DEV_AUXDATA("arm,cortex-a15-pmu", 0, "pmu", &pmu_pdata), > > + {} > > +}; > > This looks similar to what we have in mach-ux500 as db8500_pmu_handler. > > To avoid duplication, I'd prefer moving support for this into the > perf_event code itself, and get rid of the auxdata. > I'll drop it fow now. > > +static int > > +axxia_bus_notifier(struct notifier_block *nb, unsigned long event, void *obj) > > +{ > > + struct device *dev = obj; > > + > > + if (event != BUS_NOTIFY_ADD_DEVICE) > > + return NOTIFY_DONE; > > + > > + if (!of_property_read_bool(dev->of_node, "dma-coherent")) > > + return NOTIFY_DONE; > > + > > + set_dma_ops(dev, &arm_coherent_dma_ops); > > + > > + return NOTIFY_OK; > > +} > > + > > +static struct notifier_block axxia_platform_nb = { > > + .notifier_call = axxia_bus_notifier, > > +}; > > + > > +static struct notifier_block axxia_amba_nb = { > > + .notifier_call = axxia_bus_notifier, > > +}; > > And we definitely want to do this in a generic fashion. That should not be in > platform specific code. I'll drop this too. I saw there was some work in progress on this. > > > +static const char *axxia_dt_match[] __initconst = { > > + "lsi,axm55xx", > > + "lsi,axm55xx-sim", > > + "lsi,axm55xx-emu", > > + NULL > > +}; > > Please no 'xx' in compatible strings. You can list all the relevant > model numbers here, or require that machines list a specific model > as compatible, e.g. > > compatible = "lsi,axm5510-sim", "lsi,axm5510", "lsi,axm5502"; > Got that. I'll use lsi,axm5516 as the complatible model number. > > + > > +DT_MACHINE_START(AXXIA_DT, "LSI Axxia AXM55XX") > > + .dt_compat = axxia_dt_match, > > + .smp = smp_ops(axxia_smp_ops), > > + .init_machine = axxia_dt_init, > > +MACHINE_END > > Please have a look at how smp ops are hooked up in mach-qcom and see > if you can do it that way as well. > Yes, that will work nicely. > > diff --git a/arch/arm/mach-axxia/axxia.h b/arch/arm/mach-axxia/axxia.h > > new file mode 100644 > > index 0000000..594dd97 > > --- /dev/null > > +++ b/arch/arm/mach-axxia/axxia.h > > @@ -0,0 +1,42 @@ > > +/* > > + * Prototypes for platform functions. > > + * > > + * > > + * 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; either version 2 of the License, or > > + * (at your option) any later version. > > + * > > + * 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, see <http://www.gnu.org/licenses/>. > > + */ > > +#ifndef __AXXIA_H > > +#define __AXXIA_H > > + > > +#define CORES_PER_CLUSTER 4 > > + > > +#define AXXIA_PERIPH_PHYS 0x2000000000ULL > > +#define AXXIA_SYSCON_PHYS 0x2010030000ULL > > Remove these -- they have to come from DT. > > > +#if 0 > > +#define AXXIA_UART0_PHYS 0x2010080000ULL > > +#define AXXIA_UART1_PHYS 0x2010081000ULL > > +#define AXXIA_UART2_PHYS 0x2010082000ULL > > +#define AXXIA_UART3_PHYS 0x2010083000ULL > > + > > +#ifdef CONFIG_DEBUG_LL > > +#define AXXIA_DEBUG_UART_VIRT 0xf0080000 > > +#define AXXIA_DEBUG_UART_PHYS AXXIA_UART0_PHYS > > +#endif > > +#endif > > You can put these constants directly into debug code. > Ok. > > > diff --git a/arch/arm/mach-axxia/platsmp.c b/arch/arm/mach-axxia/platsmp.c > > new file mode 100644 > > index 0000000..fd7f507 > > --- /dev/null > > +++ b/arch/arm/mach-axxia/platsmp.c > > @@ -0,0 +1,183 @@ > > +/* > > + * linux/arch/arm/mach-axxia/platsmp.c > > + * > > + * Copyright (C) 2012 LSI Corporation > > + * > > + * This program is free software; you can redistribute it and/or modify > > + * it under the terms of the GNU General Public License version 2 as > > + * published by the Free Software Foundation. > > Have you considered using PSCI to bring up the secondary CPUs? > The current u-boot is a bit dated... But I agree. I'll push for an update to this. > > +static int axxia_boot_secondary(unsigned int cpu, > > + struct task_struct *idle) > > +{ > > + unsigned long timeout; > > + > > + /* Release the specified core */ > > + write_pen_release(cpu_logical_map(cpu)); > > + > > + /* Send a wakeup event to get the idled cpu out of WFE state */ > > + dsb_sev(); > > + > > + /* Wait for so long, then give up if nothing happens ... */ > > + timeout = jiffies + (1 * HZ); > > + while (time_before(jiffies, timeout)) { > > + /* Make sure stores to pen_release have completed */ > > + smp_rmb(); > > + if (pen_release == -1) > > + break; > > + udelay(1); > > + } > > + > > + return pen_release != -1 ? -ENOSYS : 0; > > +} > > This is pretty sad. No hardware support for waking up CPUs? > I'll rework this. There are sw control of the resets to individual cores. And the holding pen is not needed (too much other platsmp implementations), I will have the secondary cores released directly into secondary_startup. > > +static void __init axxia_smp_prepare_cpus(unsigned int max_cpus) > > +{ > > + void __iomem *syscon; > > + int cpu_count = 0; > > + int cpu; > > + > > + syscon = ioremap(AXXIA_SYSCON_PHYS, SZ_64K); > > + if (WARN_ON(!syscon)) > > + return; > > + > > + check_fixup_sev(syscon); > > Please use the syscon driver and regmap. That driver might need a little > help to get syscon_node_to_regmap() to work before driver initialization, > but other people need the same thing, and we should just implement it. Ok, until then... I'll drop the hardcoded address and get the address from the syscon node. > > > diff --git a/drivers/clk/clk-axxia.c b/drivers/clk/clk-axxia.c > > new file mode 100644 > > index 0000000..996b8f2 > > --- /dev/null > > +++ b/drivers/clk/clk-axxia.c > > @@ -0,0 +1,281 @@ > > +/* > > + * arch/arm/mach-axxia/clock.c > > This should be a separate patch, and get merged by the clk maintainer. Ok. I'll submit a v2 with the corrections. /Anders > > Arnd > > ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC 2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg 2014-04-15 12:30 ` Arnd Bergmann @ 2014-04-24 9:04 ` Russell King - ARM Linux 1 sibling, 0 replies; 17+ messages in thread From: Russell King - ARM Linux @ 2014-04-24 9:04 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 02:06:10PM +0200, Anders Berg wrote: > + adr r4, 2f > + ldmia r4, {r5, r6} > + sub r4, r4, r5 > + add r6, r6, r4 > +1: ldr r7, [r6] > + cmp r7, r0 > + wfene > + bne 1b ... > +2: .long . > + .long pen_release > diff --git a/arch/arm/mach-axxia/platsmp.c b/arch/arm/mach-axxia/platsmp.c ... > +/* > + * Write pen_release in a way that is guaranteed to be visible to all > + * observers, irrespective of whether they're taking part in coherency > + * or not. This is necessary for the hotplug code to work reliably. > + */ > +static void write_pen_release(int val) > +{ > + pen_release = val; > + /* Make sure this store is visible to other CPUs */ > + smp_wmb(); > + __cpuc_flush_dcache_area((void *)&pen_release, sizeof(pen_release)); > +} Do you /really/ need to do this pen_release thing on your platform? > +static void __init axxia_smp_prepare_cpus(unsigned int max_cpus) > +{ ... > + /* > + * Release all physical cpus when not in hyp mode since we > + * might want to bring them online later. > + * > + * Also we need to get the execution into kernel code (it's > + * currently executing in u-boot). u-boot releases the cores > + * from reset in hyp mode. > + */ > + if (!is_hyp_mode_available()) { > + if (cpu != 0) { > + u32 tmp = readl(syscon + 0x1010); > + writel(0xab, syscon + 0x1000); > + tmp &= ~(1 << cpu); > + writel(tmp, syscon + 0x1010); > + } As this implies that you can control the release of each CPU from u-boot individually. -- FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly improving, and getting towards what was expected from it. ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg 2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg @ 2014-04-15 12:06 ` Anders Berg 2014-04-15 12:42 ` Marc Zyngier 2014-04-15 12:45 ` Arnd Bergmann 2014-04-15 12:06 ` [PATCH 3/5] ARM: axxia: Adding defconfig " Anders Berg ` (2 subsequent siblings) 4 siblings, 2 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw) To: linux-arm-kernel Add device tree for the Amarillo validation board with an AXM5516 SoC. Signed-off-by: Anders Berg <anders.berg@lsi.com> --- arch/arm/boot/dts/Makefile | 1 + arch/arm/boot/dts/axm5516-amarillo.dts | 51 ++++++ arch/arm/boot/dts/axm5516-cpus.dtsi | 204 ++++++++++++++++++++++ arch/arm/boot/dts/axm55xx.dtsi | 306 +++++++++++++++++++++++++++++++++ 4 files changed, 562 insertions(+) create mode 100644 arch/arm/boot/dts/axm5516-amarillo.dts create mode 100644 arch/arm/boot/dts/axm5516-cpus.dtsi create mode 100644 arch/arm/boot/dts/axm55xx.dtsi diff --git a/arch/arm/boot/dts/Makefile b/arch/arm/boot/dts/Makefile index 0320303..341a3e1 100644 --- a/arch/arm/boot/dts/Makefile +++ b/arch/arm/boot/dts/Makefile @@ -46,6 +46,7 @@ dtb-$(CONFIG_ARCH_AT91) += sama5d35ek.dtb dtb-$(CONFIG_ARCH_AT91) += sama5d36ek.dtb dtb-$(CONFIG_ARCH_ATLAS6) += atlas6-evb.dtb +dtb-$(CONFIG_ARCH_AXXIA) += axm5516-amarillo.dtb dtb-$(CONFIG_ARCH_BCM2835) += bcm2835-rpi-b.dtb dtb-$(CONFIG_ARCH_BCM_MOBILE) += bcm11351-brt.dtb \ bcm28155-ap.dtb diff --git a/arch/arm/boot/dts/axm5516-amarillo.dts b/arch/arm/boot/dts/axm5516-amarillo.dts new file mode 100644 index 0000000..1760d6c --- /dev/null +++ b/arch/arm/boot/dts/axm5516-amarillo.dts @@ -0,0 +1,51 @@ +/* + * arch/arm/boot/dts/axm5516-amarillo.dts + * + * Copyright (C) 2013 LSI + * + * 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; either version 2 of the License, or + * (at your option) any later version. + */ + +/dts-v1/; + +/memreserve/ 0x00000000 0x00400000; + +#include "axm55xx.dtsi" +#include "axm5516-cpus.dtsi" + +/ { + model = "Amarillo AXM5516"; + compatible = "lsi,axm5516-amarillo", "lsi,axm55xx"; + + memory { + device_type = "memory"; + reg = <0 0x00000000 0x02 0x00000000>; + }; +}; + +&serial0 { + status = "okay"; +}; + +&serial1 { + status = "okay"; +}; + +&serial2 { + status = "okay"; +}; + +&serial3 { + status = "okay"; +}; + +&gpio0 { + status = "okay"; +}; + +&gpio1 { + status = "okay"; +}; diff --git a/arch/arm/boot/dts/axm5516-cpus.dtsi b/arch/arm/boot/dts/axm5516-cpus.dtsi new file mode 100644 index 0000000..b85f360 --- /dev/null +++ b/arch/arm/boot/dts/axm5516-cpus.dtsi @@ -0,0 +1,204 @@ +/* + * arch/arm/boot/dts/axm5516-cpus.dtsi + * + * Copyright (C) 2013 LSI + * + * 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; either version 2 of the License, or + * (at your option) any later version. + */ + +/ { + cpus { + #address-cells = <1>; + #size-cells = <0>; + + cpu-map { + cluster0 { + core0 { + cpu = <&CPU0>; + }; + core1 { + cpu = <&CPU1>; + }; + core2 { + cpu = <&CPU2>; + }; + core3 { + cpu = <&CPU3>; + }; + }; + cluster1 { + core0 { + cpu = <&CPU4>; + }; + core1 { + cpu = <&CPU5>; + }; + core2 { + cpu = <&CPU6>; + }; + core3 { + cpu = <&CPU7>; + }; + }; + cluster2 { + core0 { + cpu = <&CPU8>; + }; + core1 { + cpu = <&CPU9>; + }; + core2 { + cpu = <&CPU10>; + }; + core3 { + cpu = <&CPU11>; + }; + }; + cluster3 { + core0 { + cpu = <&CPU12>; + }; + core1 { + cpu = <&CPU13>; + }; + core2 { + cpu = <&CPU14>; + }; + core3 { + cpu = <&CPU15>; + }; + }; + }; + + CPU0: cpu at 0 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x00>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU1: cpu at 1 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x01>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU2: cpu at 2 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x02>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU3: cpu at 3 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x03>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU4: cpu at 100 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x100>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU5: cpu at 101 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x101>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU6: cpu at 102 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x102>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU7: cpu at 103 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x103>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU8: cpu at 200 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x200>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU9: cpu at 201 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x201>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU10: cpu at 202 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x202>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU11: cpu at 203 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x203>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU12: cpu at 300 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x300>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU13: cpu at 301 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x301>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU14: cpu at 302 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x302>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + + CPU15: cpu at 303 { + device_type = "cpu"; + compatible = "arm,cortex-a15"; + reg = <0x303>; + clock-frequency= <1400000000>; + cpu-release-addr = <0>; // Fixed by the boot loader + }; + }; +}; diff --git a/arch/arm/boot/dts/axm55xx.dtsi b/arch/arm/boot/dts/axm55xx.dtsi new file mode 100644 index 0000000..25fc174 --- /dev/null +++ b/arch/arm/boot/dts/axm55xx.dtsi @@ -0,0 +1,306 @@ +/* + * arch/arm/boot/dts/axm55xx.dtsi + * + * Copyright (C) 2013 LSI + * + * 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; either version 2 of the License, or + * (at your option) any later version. + */ + +#include <dt-bindings/interrupt-controller/arm-gic.h> + +#include "skeleton64.dtsi" + +/ { + interrupt-parent = <&gic>; + + aliases { + serial0 = &serial0; + serial1 = &serial1; + serial2 = &serial2; + serial3 = &serial3; + timer = &timer0; + }; + + clocks { + compatible = "simple-bus"; + #address-cells = <2>; + #size-cells = <2>; + ranges; + + /* Reference clocks */ + + clk_ref0: clk_ref0 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + clk_ref1: clk_ref1 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + clk_ref2: clk_ref2 { + compatible = "fixed-clock"; + #clock-cells = <0>; + clock-frequency = <125000000>; + }; + + /* PLLs */ + + clk_fab_pll: clk_fab_pll at 2010021800 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>; + clock-output-names = "clk_fab_pll"; + reg = <0x20 0x10021800 0 0x2c>; + }; + + clk_cpu_pll: clk_cpu_pll at 2010022000 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>; + clock-output-names = "clk_cpu_pll"; + reg = <0x20 0x10022000 0 0x2c>; + }; + + clk_sys_pll: clk_sys_pll at 2010022800 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>; + clock-output-names = "clk_sys_pll"; + reg = <0x20 0x10022800 0 0x2c>; + }; + + clk_sm0_pll: sm0_pll at 2010023000 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref2>; + clock-output-names = "clk_sm0_pll"; + reg = <0x20 0x10023000 0 0x2c>; + }; + + clk_sm1_pll: sm1_pll at 2010023800 { + compatible = "lsi,axxia-pll-clock"; + #clock-cells = <0>; + clocks = <&clk_ref1>; + clock-output-names = "clk_sm1_pll"; + reg = <0x20 0x10023800 0 0x2c>; + }; + + /* CPU clock */ + + clk_cpu_div: cpu { + compatible = "lsi,axxia-div-clock"; + #clock-cells = <0>; + clocks = <&clk_cpu_pll>; + clock-output-names = "clk_cpu_div"; + reg = <0x20 0x10030008 0 0x4>; + lsi,bit-shift = <0>; + lsi,bit-width = <4>; + }; + + clk_cpu: clk_cpu0 at 2010030000 { + compatible = "lsi,axxia-mux-clock"; + #clock-cells = <0>; + clocks = <&clk_ref0>, + <&clk_cpu_pll>, + <&clk_cpu_div>, + <&clk_cpu_div>; + clock-output-names = "clk_cpu"; + reg = <0x20 0x10030000 0 0x4>; + lsi,bit-shift = <0>; + lsi,bit-width = <2>; + }; + + /* Peripheral clock */ + + clk_per_div: peripheral { + compatible = "lsi,axxia-div-clock"; + #clock-cells = <0>; + clocks = <&clk_sm1_pll>; + clock-output-names = "clk_per_div"; + reg = <0x20 0x1003000c 0 0x4>; + lsi,bit-shift = <12>; + lsi,bit-width = <4>; + }; + + clk_per: clk_per at 2010030004 { + compatible = "lsi,axxia-mux-clock"; + #clock-cells = <0>; + clocks = <&clk_ref1>, <&clk_per_div>; + clock-output-names = "clk_per"; + reg = <0x20 0x10030004 0 0x4>; + lsi,bit-shift = <6>; + lsi,bit-width = <1>; + }; + + /* MMC clock */ + + clk_mmc_div: emmc { + compatible = "lsi,axxia-div-clock"; + #clock-cells = <0>; + clocks = <&clk_sm1_pll>; + clock-output-names = "clk_mmc_div"; + reg = <0x20 0x1003000c 0 0x4>; + lsi,bit-shift = <16>; + lsi,bit-width = <4>; + }; + + clk_mmc: clk_mmc at 2010030004 { + compatible = "lsi,axxia-mux-clock"; + #clock-cells = <0>; + clocks = <&clk_ref1>, <&clk_mmc_div>; + clock-output-names = "clk_mmc"; + reg = <0x20 0x10030004 0 0x4>; + lsi,bit-shift = <9>; + lsi,bit-width = <1>; + }; + + }; + + gic: interrupt-controller at 2001001000 { + compatible = "arm,cortex-a15-gic"; + #interrupt-cells = <3>; + #address-cells = <0>; + interrupt-controller; + reg = <0x20 0x01001000 0 0x1000>, + <0x20 0x01002000 0 0x1000>, + <0x20 0x01004000 0 0x2000>, + <0x20 0x01006000 0 0x2000>; + interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(4) | + IRQ_TYPE_LEVEL_HIGH)>; + }; + + timer { + compatible = "arm,armv7-timer"; + interrupts = + <GIC_PPI 13 + (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_LOW)>, + <GIC_PPI 14 + (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_LOW)>, + <GIC_PPI 11 + (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_LOW)>, + <GIC_PPI 10 + (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_LOW)>; + }; + + + pmu { + compatible = "arm,cortex-a15-pmu"; + interrupts = <GIC_SPI 190 IRQ_TYPE_LEVEL_HIGH>; + }; + + soc { + compatible = "simple-bus"; + device_type = "soc"; + #address-cells = <2>; + #size-cells = <2>; + interrupt-parent = <&gic>; + ranges; + + syscon: syscon at 2010030000 { + compatible = "lsi,axxia-syscon", "syscon"; + reg = <0x20 0x10030000 0 0x2000>; + }; + + amba { + compatible = "arm,amba-bus"; + #address-cells = <2>; + #size-cells = <2>; + ranges; + + serial0: uart at 2010080000 { + compatible = "arm,pl011", "arm,primecell"; + reg = <0x20 0x10080000 0 0x1000>; + interrupts = <GIC_SPI 56 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>, <&clk_per>; + clock-names = "uartclk", "apb_pclk"; + status = "disabled"; + }; + + serial1: uart at 2010081000 { + compatible = "arm,pl011", "arm,primecell"; + reg = <0x20 0x10081000 0 0x1000>; + interrupts = <GIC_SPI 57 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>, <&clk_per>; + clock-names = "uartclk", "apb_pclk"; + status = "disabled"; + }; + + serial2: uart at 2010082000 { + compatible = "arm,pl011", "arm,primecell"; + reg = <0x20 0x10082000 0 0x1000>; + interrupts = <GIC_SPI 58 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>, <&clk_per>; + clock-names = "uartclk", "apb_pclk"; + status = "disabled"; + }; + + serial3: uart at 2010083000 { + compatible = "arm,pl011", "arm,primecell"; + reg = <0x20 0x10083000 0 0x1000>; + interrupts = <GIC_SPI 59 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>, <&clk_per>; + clock-names = "uartclk", "apb_pclk"; + status = "disabled"; + }; + + timer0: timer at 2010091000 { + compatible = "arm,sp804", "arm,primecell"; + reg = <0x20 0x10091000 0 0x1000>; + interrupts = <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 47 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 48 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 49 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 50 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 52 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 53 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>, <&clk_per>; + clock-names = "timclken1", "apb_pclk"; + status = "okay"; + }; + + gpio0: gpio at 2010092000 { + #gpio-cells = <2>; + compatible = "arm,pl061", "arm,primecell"; + gpio-controller; + reg = <0x20 0x10092000 0x00 0x1000>; + interrupts = <GIC_SPI 10 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 11 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 13 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 14 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 15 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>, + <GIC_SPI 17 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>; + clock-names = "apb_pclk"; + status = "disabled"; + }; + + gpio1: gpio at 2010093000 { + #gpio-cells = <2>; + compatible = "arm,pl061", "arm,primecell"; + gpio-controller; + reg = <0x20 0x10093000 0x00 0x1000>; + interrupts = <GIC_SPI 18 IRQ_TYPE_LEVEL_HIGH>; + clocks = <&clk_per>; + clock-names = "apb_pclk"; + status = "disabled"; + }; + }; + }; +}; + +/* + Local Variables: + mode: C + End: +*/ -- 1.8.3.2 ^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:06 ` [PATCH 2/5] ARM: dts: Device tree for AXM55xx Anders Berg @ 2014-04-15 12:42 ` Marc Zyngier 2014-04-15 14:56 ` Anders Berg 2014-04-15 12:45 ` Arnd Bergmann 1 sibling, 1 reply; 17+ messages in thread From: Marc Zyngier @ 2014-04-15 12:42 UTC (permalink / raw) To: linux-arm-kernel Hi Anders, On 15/04/14 13:06, Anders Berg wrote: > Add device tree for the Amarillo validation board with an AXM5516 SoC. > > Signed-off-by: Anders Berg <anders.berg@lsi.com> > --- > arch/arm/boot/dts/Makefile | 1 + > arch/arm/boot/dts/axm5516-amarillo.dts | 51 ++++++ > arch/arm/boot/dts/axm5516-cpus.dtsi | 204 ++++++++++++++++++++++ > arch/arm/boot/dts/axm55xx.dtsi | 306 +++++++++++++++++++++++++++++++++ > 4 files changed, 562 insertions(+) > create mode 100644 arch/arm/boot/dts/axm5516-amarillo.dts > create mode 100644 arch/arm/boot/dts/axm5516-cpus.dtsi > create mode 100644 arch/arm/boot/dts/axm55xx.dtsi > [...] > + gic: interrupt-controller at 2001001000 { > + compatible = "arm,cortex-a15-gic"; > + #interrupt-cells = <3>; > + #address-cells = <0>; > + interrupt-controller; > + reg = <0x20 0x01001000 0 0x1000>, > + <0x20 0x01002000 0 0x1000>, > + <0x20 0x01004000 0 0x2000>, > + <0x20 0x01006000 0 0x2000>; > + interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(4) | > + IRQ_TYPE_LEVEL_HIGH)>; > + }; Given how many CPUs this system has, what's the catch regarding the GIC? Is there a second one shadowing this one at the same address for another set of 8 CPUs? Is there an additional mechanism to IPI the other CPUs? Thanks, M. -- Jazz is not dead. It just smells funny... ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:42 ` Marc Zyngier @ 2014-04-15 14:56 ` Anders Berg 0 siblings, 0 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 14:56 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 01:42:53PM +0100, Marc Zyngier wrote: > Hi Anders, > > On 15/04/14 13:06, Anders Berg wrote: > > Add device tree for the Amarillo validation board with an AXM5516 SoC. > > > > Signed-off-by: Anders Berg <anders.berg@lsi.com> > > --- > > arch/arm/boot/dts/Makefile | 1 + > > arch/arm/boot/dts/axm5516-amarillo.dts | 51 ++++++ > > arch/arm/boot/dts/axm5516-cpus.dtsi | 204 ++++++++++++++++++++++ > > arch/arm/boot/dts/axm55xx.dtsi | 306 +++++++++++++++++++++++++++++++++ > > 4 files changed, 562 insertions(+) > > create mode 100644 arch/arm/boot/dts/axm5516-amarillo.dts > > create mode 100644 arch/arm/boot/dts/axm5516-cpus.dtsi > > create mode 100644 arch/arm/boot/dts/axm55xx.dtsi > > > > [...] > > > + gic: interrupt-controller at 2001001000 { > > + compatible = "arm,cortex-a15-gic"; > > + #interrupt-cells = <3>; > > + #address-cells = <0>; > > + interrupt-controller; > > + reg = <0x20 0x01001000 0 0x1000>, > > + <0x20 0x01002000 0 0x1000>, > > + <0x20 0x01004000 0 0x2000>, > > + <0x20 0x01006000 0 0x2000>; > > + interrupts = <GIC_PPI 9 (GIC_CPU_MASK_SIMPLE(4) | > > + IRQ_TYPE_LEVEL_HIGH)>; > > + }; > > Given how many CPUs this system has, what's the catch regarding the GIC? > Is there a second one shadowing this one at the same address for another > set of 8 CPUs? Is there an additional mechanism to IPI the other CPUs? > There is actually one GIC per cluster and SPIs are used to hook the clusters together for use as IPIs, so with this patch one cluster may be booted (using the standard SGIs still works within each cluster). Have a modified GIC driver to run all clusters, but this patch is a bit to invasive to submit at the moment... /Anders ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:06 ` [PATCH 2/5] ARM: dts: Device tree for AXM55xx Anders Berg 2014-04-15 12:42 ` Marc Zyngier @ 2014-04-15 12:45 ` Arnd Bergmann 2014-04-15 15:20 ` Anders Berg ` (2 more replies) 1 sibling, 3 replies; 17+ messages in thread From: Arnd Bergmann @ 2014-04-15 12:45 UTC (permalink / raw) To: linux-arm-kernel On Tuesday 15 April 2014 14:06:11 Anders Berg wrote: > diff --git a/arch/arm/boot/dts/axm5516-amarillo.dts b/arch/arm/boot/dts/axm5516-amarillo.dts > new file mode 100644 > index 0000000..1760d6c > --- /dev/null > +++ b/arch/arm/boot/dts/axm5516-amarillo.dts > @@ -0,0 +1,51 @@ > +/* > + * arch/arm/boot/dts/axm5516-amarillo.dts > + * > + * Copyright (C) 2013 LSI > + * > + * 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; either version 2 of the License, or > + * (at your option) any later version. > + */ It's better to use a permissive license such as the BSD license for dts files, so they can be shared with other OSs. The situation with the clocks is still very strange: either the bindings are all in need of update, or you got all devices wrong: > + serial0: uart at 2010080000 { > + compatible = "arm,pl011", "arm,primecell"; > + reg = <0x20 0x10080000 0 0x1000>; > + interrupts = <GIC_SPI 56 IRQ_TYPE_LEVEL_HIGH>; > + clocks = <&clk_per>, <&clk_per>; > + clock-names = "uartclk", "apb_pclk"; > + status = "disabled"; > + }; "uartclk" is not a valid string for pl011, as per binding: | - clocks: When present, must refer to exactly one clock named | "apb_pclk" I do see that a lot of platforms do the same thing you have here, not sure who is wrong. > + timer0: timer at 2010091000 { > + compatible = "arm,sp804", "arm,primecell"; > + reg = <0x20 0x10091000 0 0x1000>; > + interrupts = <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 47 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 48 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 49 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 50 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 52 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 53 IRQ_TYPE_LEVEL_HIGH>; > + clocks = <&clk_per>, <&clk_per>; > + clock-names = "timclken1", "apb_pclk"; Citing the binding: | - clocks: clocks driving the dual timer hardware. This list should be 1 or 3 | clocks. With 3 clocks, the order is timer0 clock, timer1 clock, | apb_pclk. A single clock can also be specified if the same clock is | used for all clock inputs. I think you only want to have one clock here and make that the "apb_pclk". > + gpio0: gpio at 2010092000 { > + #gpio-cells = <2>; > + compatible = "arm,pl061", "arm,primecell"; > + gpio-controller; > + reg = <0x20 0x10092000 0x00 0x1000>; > + interrupts = <GIC_SPI 10 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 11 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 13 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 14 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 15 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>, > + <GIC_SPI 17 IRQ_TYPE_LEVEL_HIGH>; > + clocks = <&clk_per>; > + clock-names = "apb_pclk"; > + status = "disabled"; The pl061 binding does not specify any clocks at all. Do we need to update that? Arnd ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:45 ` Arnd Bergmann @ 2014-04-15 15:20 ` Anders Berg 2014-04-15 15:23 ` Arnd Bergmann 2014-04-22 21:31 ` Linus Walleij 2014-04-22 21:39 ` Russell King - ARM Linux 2 siblings, 1 reply; 17+ messages in thread From: Anders Berg @ 2014-04-15 15:20 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 02:45:48PM +0200, Arnd Bergmann wrote: > On Tuesday 15 April 2014 14:06:11 Anders Berg wrote: > > diff --git a/arch/arm/boot/dts/axm5516-amarillo.dts b/arch/arm/boot/dts/axm5516-amarillo.dts > > new file mode 100644 > > index 0000000..1760d6c > > --- /dev/null > > +++ b/arch/arm/boot/dts/axm5516-amarillo.dts > > @@ -0,0 +1,51 @@ > > +/* > > + * arch/arm/boot/dts/axm5516-amarillo.dts > > + * > > + * Copyright (C) 2013 LSI > > + * > > + * 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; either version 2 of the License, or > > + * (at your option) any later version. > > + */ > > It's better to use a permissive license such as the BSD license for dts files, > so they can be shared with other OSs. Ok. > > The situation with the clocks is still very strange: either the bindings > are all in need of update, or you got all devices wrong: > > > + serial0: uart at 2010080000 { > > + compatible = "arm,pl011", "arm,primecell"; > > + reg = <0x20 0x10080000 0 0x1000>; > > + interrupts = <GIC_SPI 56 IRQ_TYPE_LEVEL_HIGH>; > > + clocks = <&clk_per>, <&clk_per>; > > + clock-names = "uartclk", "apb_pclk"; > > + status = "disabled"; > > + }; > > "uartclk" is not a valid string for pl011, as per binding: > > | - clocks: When present, must refer to exactly one clock named > | "apb_pclk" > > I do see that a lot of platforms do the same thing you have here, not > sure who is wrong. The docs are probably correct. I used some vexpress dtsi as a reference, hence the uartclk reference. It's gone now. > > > + timer0: timer at 2010091000 { > > + compatible = "arm,sp804", "arm,primecell"; > > + reg = <0x20 0x10091000 0 0x1000>; > > + interrupts = <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 46 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 47 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 48 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 49 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 50 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 51 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 52 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 53 IRQ_TYPE_LEVEL_HIGH>; > > + clocks = <&clk_per>, <&clk_per>; > > + clock-names = "timclken1", "apb_pclk"; > > Citing the binding: > > | - clocks: clocks driving the dual timer hardware. This list should be 1 or 3 > | clocks. With 3 clocks, the order is timer0 clock, timer1 clock, > | apb_pclk. A single clock can also be specified if the same clock is > | used for all clock inputs. > > I think you only want to have one clock here and make that the "apb_pclk". Yepp, will be fixed. > > > + gpio0: gpio at 2010092000 { > > + #gpio-cells = <2>; > > + compatible = "arm,pl061", "arm,primecell"; > > + gpio-controller; > > + reg = <0x20 0x10092000 0x00 0x1000>; > > + interrupts = <GIC_SPI 10 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 11 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 13 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 14 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 15 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>, > > + <GIC_SPI 17 IRQ_TYPE_LEVEL_HIGH>; > > + clocks = <&clk_per>; > > + clock-names = "apb_pclk"; > > + status = "disabled"; > > The pl061 binding does not specify any clocks at all. Do we need to update > that? Doesn't all AMBA devices need at least one apb_pclk since the bus driver does clk_get(...,"apb_pclk") before calling probe()? /Anders ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 15:20 ` Anders Berg @ 2014-04-15 15:23 ` Arnd Bergmann 2014-04-22 21:21 ` Linus Walleij 0 siblings, 1 reply; 17+ messages in thread From: Arnd Bergmann @ 2014-04-15 15:23 UTC (permalink / raw) To: linux-arm-kernel On Tuesday 15 April 2014 17:20:47 Anders Berg wrote: > > > + gpio0: gpio at 2010092000 { > > > + #gpio-cells = <2>; > > > + compatible = "arm,pl061", "arm,primecell"; > > > + gpio-controller; > > > + reg = <0x20 0x10092000 0x00 0x1000>; > > > + interrupts = <GIC_SPI 10 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 11 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 13 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 14 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 15 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>, > > > + <GIC_SPI 17 IRQ_TYPE_LEVEL_HIGH>; > > > + clocks = <&clk_per>; > > > + clock-names = "apb_pclk"; > > > + status = "disabled"; > > > > The pl061 binding does not specify any clocks at all. Do we need to update > > that? > > Doesn't all AMBA devices need at least one apb_pclk since the bus driver does > clk_get(...,"apb_pclk") before calling probe()? Yes, I was mostly wondering whether we had a policy about whether this clock should also be listed in the binding or not. My feeling is that it's better to make that explicit. Arnd ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 15:23 ` Arnd Bergmann @ 2014-04-22 21:21 ` Linus Walleij 0 siblings, 0 replies; 17+ messages in thread From: Linus Walleij @ 2014-04-22 21:21 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 5:23 PM, Arnd Bergmann <arnd@arndb.de> wrote: > On Tuesday 15 April 2014 17:20:47 Anders Berg wrote: >> > > + gpio0: gpio at 2010092000 { >> > > + #gpio-cells = <2>; >> > > + compatible = "arm,pl061", "arm,primecell"; >> > > + gpio-controller; >> > > + reg = <0x20 0x10092000 0x00 0x1000>; >> > > + interrupts = <GIC_SPI 10 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 11 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 12 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 13 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 14 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 15 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 16 IRQ_TYPE_LEVEL_HIGH>, >> > > + <GIC_SPI 17 IRQ_TYPE_LEVEL_HIGH>; >> > > + clocks = <&clk_per>; >> > > + clock-names = "apb_pclk"; >> > > + status = "disabled"; >> > >> > The pl061 binding does not specify any clocks at all. Do we need to update >> > that? >> >> Doesn't all AMBA devices need at least one apb_pclk since the bus driver does >> clk_get(...,"apb_pclk") before calling probe()? > > Yes, I was mostly wondering whether we had a policy about whether this clock > should also be listed in the binding or not. My feeling is that it's better > to make that explicit. A separate patch to the GPIO maintainer for this will be accepted. :-) Yours, Linus Walleij ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:45 ` Arnd Bergmann 2014-04-15 15:20 ` Anders Berg @ 2014-04-22 21:31 ` Linus Walleij 2014-04-22 21:39 ` Russell King - ARM Linux 2 siblings, 0 replies; 17+ messages in thread From: Linus Walleij @ 2014-04-22 21:31 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 2:45 PM, Arnd Bergmann <arnd@arndb.de> wrote: > On Tuesday 15 April 2014 14:06:11 Anders Berg wrote: >> + serial0: uart at 2010080000 { >> + compatible = "arm,pl011", "arm,primecell"; >> + reg = <0x20 0x10080000 0 0x1000>; >> + interrupts = <GIC_SPI 56 IRQ_TYPE_LEVEL_HIGH>; >> + clocks = <&clk_per>, <&clk_per>; >> + clock-names = "uartclk", "apb_pclk"; >> + status = "disabled"; >> + }; > > "uartclk" is not a valid string for pl011, as per binding: > > | - clocks: When present, must refer to exactly one clock named > | "apb_pclk" > > I do see that a lot of platforms do the same thing you have here, not > sure who is wrong. OK the PL011 is tricksier. It has two clocks, the APB clock driving the logic and another clock input that is divided down to generate the external data clock (baud rate). The bus will do: pcdev->pclk = clk_get(&pcdev->dev, "apb_pclk"); The driver will do: uap->clk = devm_clk_get(&dev->dev, NULL); So the driver relies on the anonymous clock being "first" in a clock retrieveal operation, and it's a bit fragile indeed. That leads to that either you give this clock a sensible name or you list the empty string, as you have to have clock-names there to name the last entry so the bus can retrieve it. This clock is *indeed* named "UARTCLK" (capital letters) in the TRM for PL011. So I think this is perfectly reasonable. The situation is the same as with the PL022 SPI driver we have discussed before, it also retrieves the anonymous clock but we consistently named it "SSPCLK" in commits: dfc9832ca3f7a10d573e29e46d6ba03a54fbe580 80fbe30f63c9ba2ab62eb2c1f7ef607ce0721a95 As everyone else was using that. And it was the name from the TRM. Yours, Linus Walleij ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 2/5] ARM: dts: Device tree for AXM55xx. 2014-04-15 12:45 ` Arnd Bergmann 2014-04-15 15:20 ` Anders Berg 2014-04-22 21:31 ` Linus Walleij @ 2014-04-22 21:39 ` Russell King - ARM Linux 2 siblings, 0 replies; 17+ messages in thread From: Russell King - ARM Linux @ 2014-04-22 21:39 UTC (permalink / raw) To: linux-arm-kernel On Tue, Apr 15, 2014 at 02:45:48PM +0200, Arnd Bergmann wrote: > "uartclk" is not a valid string for pl011, as per binding: > > | - clocks: When present, must refer to exactly one clock named > | "apb_pclk" The binding document is basically *wrong*. -- FTTC broadband for 0.8mile line: now at 9.7Mbps down 460kbps up... slowly improving, and getting towards what was expected from it. ^ permalink raw reply [flat|nested] 17+ messages in thread
* [PATCH 3/5] ARM: axxia: Adding defconfig for AXM55xx 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg 2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg 2014-04-15 12:06 ` [PATCH 2/5] ARM: dts: Device tree for AXM55xx Anders Berg @ 2014-04-15 12:06 ` Anders Berg 2014-04-15 12:06 ` [PATCH 4/5] power: reset: Add Axxia system reset driver Anders Berg 2014-04-15 12:06 ` [PATCH 5/5] ARM: dts: axxia: Add reset controller Anders Berg 4 siblings, 0 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw) To: linux-arm-kernel Add a defconfig file for the LSI Axxia family of devices (CONFIG_ARCH_AXXIA). Signed-off-by: Anders Berg <anders.berg@lsi.com> --- arch/arm/configs/axm55xx_defconfig | 248 +++++++++++++++++++++++++++++++++++++ 1 file changed, 248 insertions(+) create mode 100644 arch/arm/configs/axm55xx_defconfig diff --git a/arch/arm/configs/axm55xx_defconfig b/arch/arm/configs/axm55xx_defconfig new file mode 100644 index 0000000..d3260d7 --- /dev/null +++ b/arch/arm/configs/axm55xx_defconfig @@ -0,0 +1,248 @@ +CONFIG_SYSVIPC=y +CONFIG_POSIX_MQUEUE=y +CONFIG_FHANDLE=y +CONFIG_AUDIT=y +CONFIG_HIGH_RES_TIMERS=y +CONFIG_BSD_PROCESS_ACCT=y +CONFIG_BSD_PROCESS_ACCT_V3=y +CONFIG_TASKSTATS=y +CONFIG_TASK_DELAY_ACCT=y +CONFIG_TASK_XACCT=y +CONFIG_TASK_IO_ACCOUNTING=y +CONFIG_IKCONFIG=y +CONFIG_IKCONFIG_PROC=y +CONFIG_LOG_BUF_SHIFT=16 +CONFIG_NAMESPACES=y +# CONFIG_UTS_NS is not set +# CONFIG_IPC_NS is not set +# CONFIG_PID_NS is not set +# CONFIG_NET_NS is not set +CONFIG_SCHED_AUTOGROUP=y +CONFIG_RELAY=y +CONFIG_BLK_DEV_INITRD=y +CONFIG_SYSCTL_SYSCALL=y +CONFIG_EMBEDDED=y +# CONFIG_COMPAT_BRK is not set +CONFIG_PROFILING=y +CONFIG_MODULES=y +CONFIG_MODULE_UNLOAD=y +# CONFIG_IOSCHED_DEADLINE is not set +CONFIG_ARCH_AXXIA=y +CONFIG_GPIO_PCA953X=y +CONFIG_ARM_LPAE=y +CONFIG_ARM_THUMBEE=y +CONFIG_ARM_ERRATA_430973=y +CONFIG_ARM_ERRATA_643719=y +CONFIG_ARM_ERRATA_720789=y +CONFIG_ARM_ERRATA_754322=y +CONFIG_ARM_ERRATA_754327=y +CONFIG_ARM_ERRATA_764369=y +CONFIG_ARM_ERRATA_775420=y +CONFIG_ARM_ERRATA_798181=y +CONFIG_PCI=y +CONFIG_PCI_MSI=y +CONFIG_PCIE_AXXIA=y +CONFIG_SMP=y +CONFIG_NR_CPUS=16 +CONFIG_HOTPLUG_CPU=y +CONFIG_PREEMPT=y +CONFIG_AEABI=y +CONFIG_OABI_COMPAT=y +CONFIG_HIGHMEM=y +CONFIG_KSM=y +CONFIG_ZBOOT_ROM_TEXT=0x0 +CONFIG_ZBOOT_ROM_BSS=0x0 +CONFIG_ARM_APPENDED_DTB=y +CONFIG_ARM_ATAG_DTB_COMPAT=y +CONFIG_VFP=y +CONFIG_NEON=y +# CONFIG_CORE_DUMP_DEFAULT_ELF_HEADERS is not set +CONFIG_BINFMT_MISC=y +# CONFIG_SUSPEND is not set +CONFIG_NET=y +CONFIG_PACKET=y +CONFIG_UNIX=y +CONFIG_XFRM_USER=y +CONFIG_XFRM_SUB_POLICY=y +CONFIG_XFRM_MIGRATE=y +CONFIG_XFRM_STATISTICS=y +CONFIG_NET_KEY=y +CONFIG_INET=y +CONFIG_IP_PNP=y +CONFIG_IP_PNP_DHCP=y +CONFIG_IP_PNP_BOOTP=y +CONFIG_INET_AH=y +CONFIG_INET_ESP=y +CONFIG_INET_IPCOMP=y +# CONFIG_INET_LRO is not set +# CONFIG_IPV6 is not set +CONFIG_NETWORK_PHY_TIMESTAMPING=y +CONFIG_BRIDGE=y +# CONFIG_WIRELESS is not set +CONFIG_UEVENT_HELPER_PATH="/sbin/hotplug" +CONFIG_DEVTMPFS=y +CONFIG_DEVTMPFS_MOUNT=y +CONFIG_MTD=y +CONFIG_MTD_CMDLINE_PARTS=y +CONFIG_MTD_AFS_PARTS=y +CONFIG_MTD_BLOCK=y +CONFIG_MTD_CFI=y +CONFIG_MTD_CFI_INTELEXT=y +CONFIG_MTD_CFI_AMDSTD=y +CONFIG_MTD_CFI_STAA=y +CONFIG_MTD_PHYSMAP=y +CONFIG_MTD_PHYSMAP_OF=y +CONFIG_MTD_M25P80=y +CONFIG_PROC_DEVICETREE=y +CONFIG_BLK_DEV_LOOP=y +CONFIG_BLK_DEV_RAM=y +CONFIG_EEPROM_AT24=y +CONFIG_EEPROM_AT25=y +CONFIG_BLK_DEV_SD=y +CONFIG_CHR_DEV_SG=y +CONFIG_ATA=y +CONFIG_PATA_PLATFORM=y +CONFIG_PATA_OF_PLATFORM=y +CONFIG_MD=y +CONFIG_BLK_DEV_DM=y +CONFIG_DM_UEVENT=y +CONFIG_NETDEVICES=y +CONFIG_TUN=y +CONFIG_VETH=y +CONFIG_VIRTIO_NET=y +# CONFIG_NET_CADENCE is not set +# CONFIG_NET_VENDOR_BROADCOM is not set +# CONFIG_NET_VENDOR_CIRRUS is not set +# CONFIG_NET_VENDOR_FARADAY is not set +# CONFIG_NET_VENDOR_INTEL is not set +# CONFIG_NET_VENDOR_MARVELL is not set +# CONFIG_NET_VENDOR_MICREL is not set +# CONFIG_NET_VENDOR_NATSEMI is not set +# CONFIG_NET_VENDOR_SEEQ is not set +# CONFIG_NET_VENDOR_SMSC is not set +# CONFIG_NET_VENDOR_STMICRO is not set +# CONFIG_NET_VENDOR_VIA is not set +# CONFIG_NET_VENDOR_WIZNET is not set +CONFIG_BROADCOM_PHY=y +# CONFIG_WLAN is not set +# CONFIG_MOUSE_PS2_ALPS is not set +# CONFIG_MOUSE_PS2_LOGIPS2PP is not set +# CONFIG_MOUSE_PS2_SYNAPTICS is not set +# CONFIG_MOUSE_PS2_TRACKPOINT is not set +# CONFIG_SERIO_SERPORT is not set +CONFIG_SERIO_AMBAKMI=y +CONFIG_LEGACY_PTY_COUNT=16 +CONFIG_SERIAL_AMBA_PL011=y +CONFIG_SERIAL_AMBA_PL011_CONSOLE=y +CONFIG_VIRTIO_CONSOLE=y +# CONFIG_HW_RANDOM is not set +CONFIG_I2C=y +CONFIG_I2C_CHARDEV=y +CONFIG_I2C_MUX=y +CONFIG_I2C_AXXIA=y +CONFIG_SPI=y +CONFIG_SPI_PL022=y +CONFIG_DP83640_PHY=y +CONFIG_GPIOLIB=y +CONFIG_GPIO_SYSFS=y +CONFIG_GPIO_PL061=y +CONFIG_POWER_SUPPLY=y +CONFIG_POWER_RESET=y +CONFIG_POWER_RESET_AXXIA=y +CONFIG_SENSORS_ADT7475=y +CONFIG_SENSORS_JC42=y +CONFIG_SENSORS_LM75=y +CONFIG_PMBUS=y +CONFIG_SENSORS_LTC2978=y +CONFIG_WATCHDOG=y +CONFIG_ARM_SP805_WATCHDOG=y +CONFIG_FB=y +CONFIG_FB_ARMCLCD=y +CONFIG_FRAMEBUFFER_CONSOLE=y +CONFIG_LOGO=y +# CONFIG_LOGO_LINUX_MONO is not set +# CONFIG_LOGO_LINUX_VGA16 is not set +CONFIG_HID_A4TECH=y +CONFIG_HID_APPLE=y +CONFIG_HID_BELKIN=y +CONFIG_HID_CHERRY=y +CONFIG_HID_CHICONY=y +CONFIG_HID_CYPRESS=y +CONFIG_HID_DRAGONRISE=y +CONFIG_HID_EZKEY=y +CONFIG_HID_KYE=y +CONFIG_HID_GYRATION=y +CONFIG_HID_TWINHAN=y +CONFIG_HID_KENSINGTON=y +CONFIG_HID_LOGITECH=y +CONFIG_HID_MICROSOFT=y +CONFIG_HID_MONTEREY=y +CONFIG_HID_NTRIG=y +CONFIG_HID_ORTEK=y +CONFIG_HID_PANTHERLORD=y +CONFIG_HID_PETALYNX=y +CONFIG_HID_SAMSUNG=y +CONFIG_HID_SUNPLUS=y +CONFIG_HID_GREENASIA=y +CONFIG_HID_SMARTJOYPLUS=y +CONFIG_HID_TOPSEED=y +CONFIG_HID_THRUSTMASTER=y +CONFIG_HID_ZEROPLUS=y +CONFIG_USB=y +CONFIG_USB_ANNOUNCE_NEW_DEVICES=y +CONFIG_USB_EHCI_HCD=y +# CONFIG_USB_EHCI_TT_NEWSCHED is not set +CONFIG_USB_EHCI_HCD_AXXIA=y +CONFIG_USB_STORAGE=y +CONFIG_MMC=y +CONFIG_MMC_ARMMMCI=y +CONFIG_DMADEVICES=y +CONFIG_PL330_DMA=y +CONFIG_VIRT_DRIVERS=y +CONFIG_VIRTIO_MMIO=y +CONFIG_MAILBOX=y +CONFIG_PL320_MBOX=y +# CONFIG_IOMMU_SUPPORT is not set +CONFIG_EXT2_FS=y +CONFIG_EXT3_FS=y +# CONFIG_EXT3_DEFAULTS_TO_ORDERED is not set +CONFIG_EXT4_FS=y +CONFIG_AUTOFS4_FS=y +CONFIG_FUSE_FS=y +CONFIG_CUSE=y +CONFIG_FSCACHE=y +CONFIG_FSCACHE_STATS=y +CONFIG_FSCACHE_HISTOGRAM=y +CONFIG_FSCACHE_DEBUG=y +CONFIG_FSCACHE_OBJECT_LIST=y +CONFIG_CACHEFILES=y +CONFIG_CACHEFILES_HISTOGRAM=y +CONFIG_ISO9660_FS=y +CONFIG_UDF_FS=y +CONFIG_MSDOS_FS=y +CONFIG_VFAT_FS=y +CONFIG_NTFS_FS=y +CONFIG_TMPFS=y +CONFIG_JFFS2_FS=y +CONFIG_CRAMFS=y +CONFIG_NFS_FS=y +CONFIG_NFS_V4=y +CONFIG_ROOT_NFS=y +CONFIG_NFS_FSCACHE=y +CONFIG_SUNRPC_DEBUG=y +CONFIG_NLS_CODEPAGE_437=y +CONFIG_NLS_ISO8859_1=y +CONFIG_PRINTK_TIME=y +CONFIG_DEBUG_INFO=y +CONFIG_DEBUG_FS=y +CONFIG_MAGIC_SYSRQ=y +# CONFIG_SCHED_DEBUG is not set +CONFIG_RCU_CPU_STALL_TIMEOUT=60 +# CONFIG_FTRACE is not set +CONFIG_DEBUG_USER=y +CONFIG_CRYPTO_GCM=y +CONFIG_CRYPTO_XCBC=y +CONFIG_CRYPTO_SHA256=y +# CONFIG_CRYPTO_ANSI_CPRNG is not set +CONFIG_VIRTUALIZATION=y +CONFIG_KVM=y -- 1.8.3.2 ^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH 4/5] power: reset: Add Axxia system reset driver 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg ` (2 preceding siblings ...) 2014-04-15 12:06 ` [PATCH 3/5] ARM: axxia: Adding defconfig " Anders Berg @ 2014-04-15 12:06 ` Anders Berg 2014-04-15 12:06 ` [PATCH 5/5] ARM: dts: axxia: Add reset controller Anders Berg 4 siblings, 0 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw) To: linux-arm-kernel Add Axxia (AXM55xx) SoC system reset driver. This driver handles only system reboot (and not power-off). Signed-off-by: Anders Berg <anders.berg@lsi.com> --- .../bindings/power_supply/axxia-reset.txt | 20 +++++ drivers/power/reset/Kconfig | 8 ++ drivers/power/reset/Makefile | 1 + drivers/power/reset/axxia-reset.c | 97 ++++++++++++++++++++++ 4 files changed, 126 insertions(+) create mode 100644 Documentation/devicetree/bindings/power_supply/axxia-reset.txt create mode 100644 drivers/power/reset/axxia-reset.c diff --git a/Documentation/devicetree/bindings/power_supply/axxia-reset.txt b/Documentation/devicetree/bindings/power_supply/axxia-reset.txt new file mode 100644 index 0000000..47e720d --- /dev/null +++ b/Documentation/devicetree/bindings/power_supply/axxia-reset.txt @@ -0,0 +1,20 @@ +Axxia Restart Driver + +This driver can do reset of the Axxia SoC. It uses the registers in the syscon +block to initiate a chip reset. + +Required Properties: + -compatible: "lsi,axm55xx-reset" + -syscon: phandle to the syscon node. + +Example: + + syscon: syscon at 2010030000 { + compatible = "lsi,axxia-syscon", "syscon"; + reg = <0x20 0x10030000 0 0x2000>; + }; + + reset: reset at 2010031000 { + compatible = "lsi,axm55xx-reset"; + syscon = <&syscon>; + }; diff --git a/drivers/power/reset/Kconfig b/drivers/power/reset/Kconfig index 6d452a7..7c39066 100644 --- a/drivers/power/reset/Kconfig +++ b/drivers/power/reset/Kconfig @@ -12,6 +12,14 @@ config POWER_RESET_AS3722 help This driver supports turning off board via a ams AS3722 power-off. +config POWER_RESET_AXXIA + bool "LSI Axxia reset driver" + depends on POWER_RESET && ARCH_AXXIA + help + This driver supports restart for Axxia SoC. + + Say Y if you have an Axxia family SoC. + config POWER_RESET_GPIO bool "GPIO power-off driver" depends on OF_GPIO && POWER_RESET diff --git a/drivers/power/reset/Makefile b/drivers/power/reset/Makefile index a5b4a77..16c0516 100644 --- a/drivers/power/reset/Makefile +++ b/drivers/power/reset/Makefile @@ -1,4 +1,5 @@ obj-$(CONFIG_POWER_RESET_AS3722) += as3722-poweroff.o +obj-$(CONFIG_POWER_RESET_AXXIA) += axxia-reset.o obj-$(CONFIG_POWER_RESET_GPIO) += gpio-poweroff.o obj-$(CONFIG_POWER_RESET_MSM) += msm-poweroff.o obj-$(CONFIG_POWER_RESET_QNAP) += qnap-poweroff.o diff --git a/drivers/power/reset/axxia-reset.c b/drivers/power/reset/axxia-reset.c new file mode 100644 index 0000000..a16eb50 --- /dev/null +++ b/drivers/power/reset/axxia-reset.c @@ -0,0 +1,97 @@ +/* + * Reset driver for Axxia devices + * + * Copyright (C) 2014 LSI + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 and + * only version 2 as published by the Free Software Foundation. + * + * 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. + * + */ +#include <linux/delay.h> +#include <linux/err.h> +#include <linux/init.h> +#include <linux/io.h> +#include <linux/kernel.h> +#include <linux/mfd/syscon.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/platform_device.h> +#include <linux/reboot.h> +#include <linux/regmap.h> + +#include <asm/system_misc.h> + + +#define SC_CRIT_WRITE_KEY 0x1000 +#define SC_LATCH_ON_RESET 0x1004 +#define SC_RESET_CONTROL 0x1008 +#define RSTCTL_RST_ZERO (1<<3) +#define RSTCTL_RST_FAB (1<<2) +#define RSTCTL_RST_CHIP (1<<1) +#define RSTCTL_RST_SYS (1<<0) +#define SC_EFUSE_INT_STATUS 0x180c +#define EFUSE_READ_DONE (1<<31) + +static struct regmap *syscon; + +static void do_axxia_restart(enum reboot_mode reboot_mode, const char *cmd) +{ + /* Access Key (0xab) */ + regmap_write(syscon, SC_CRIT_WRITE_KEY, 0xab); + /* Select internal boot from 0xffff0000 */ + regmap_write(syscon, SC_LATCH_ON_RESET, 0x00000040); + /* Assert ResetReadDone (to avoid hanging in boot ROM) */ + regmap_write(syscon, SC_EFUSE_INT_STATUS, EFUSE_READ_DONE); + /* Assert chip reset */ + regmap_update_bits(syscon, SC_RESET_CONTROL, + RSTCTL_RST_CHIP, RSTCTL_RST_CHIP); +} + +static int axxia_reset_probe(struct platform_device *pdev) +{ + struct device *dev = &pdev->dev; + struct device_node *dev_np = dev->of_node; + struct device_node *syscon_np; + + syscon_np = of_parse_phandle(dev_np, "syscon", 0); + if (!syscon_np) { + pr_err("%s: no handle to syscon node\n", dev_np->name); + return -ENODEV; + } + + syscon = syscon_node_to_regmap(syscon_np); + if (IS_ERR(syscon)) { + pr_err("%s: could not locate syscon regmap\n", dev_np->name); + return PTR_ERR(syscon); + } + + arm_pm_restart = do_axxia_restart; + + return 0; +} + +static const struct of_device_id of_axxia_reset_match[] = { + { .compatible = "lsi,axm55xx-reset", }, + {}, +}; +MODULE_DEVICE_TABLE(of, of_axxia_reset_match); + +static struct platform_driver axxia_reset_driver = { + .probe = axxia_reset_probe, + .driver = { + .name = "axxia-reset", + .of_match_table = of_match_ptr(of_axxia_reset_match), + }, +}; + +static int __init axxia_reset_init(void) +{ + return platform_driver_register(&axxia_reset_driver); +} +device_initcall(axxia_reset_init); -- 1.8.3.2 ^ permalink raw reply related [flat|nested] 17+ messages in thread
* [PATCH 5/5] ARM: dts: axxia: Add reset controller 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg ` (3 preceding siblings ...) 2014-04-15 12:06 ` [PATCH 4/5] power: reset: Add Axxia system reset driver Anders Berg @ 2014-04-15 12:06 ` Anders Berg 4 siblings, 0 replies; 17+ messages in thread From: Anders Berg @ 2014-04-15 12:06 UTC (permalink / raw) To: linux-arm-kernel Add the reset controller to the AXM5xx device tree. Signed-off-by: Anders Berg <anders.berg@lsi.com> --- arch/arm/boot/dts/axm55xx.dtsi | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/arch/arm/boot/dts/axm55xx.dtsi b/arch/arm/boot/dts/axm55xx.dtsi index 25fc174..9e33a8e 100644 --- a/arch/arm/boot/dts/axm55xx.dtsi +++ b/arch/arm/boot/dts/axm55xx.dtsi @@ -208,6 +208,11 @@ reg = <0x20 0x10030000 0 0x2000>; }; + reset: reset at 2010031000 { + compatible = "lsi,axm55xx-reset"; + syscon = <&syscon>; + }; + amba { compatible = "arm,amba-bus"; #address-cells = <2>; -- 1.8.3.2 ^ permalink raw reply related [flat|nested] 17+ messages in thread
end of thread, other threads:[~2014-04-24 9:04 UTC | newest] Thread overview: 17+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2014-04-15 12:06 [PATCH 0/5] Add platform support for LSI AXM55xx Anders Berg 2014-04-15 12:06 ` [PATCH 1/5] ARM: Add platform support for LSI AXM55xx SoC Anders Berg 2014-04-15 12:30 ` Arnd Bergmann 2014-04-24 8:58 ` Anders Berg 2014-04-24 9:04 ` Russell King - ARM Linux 2014-04-15 12:06 ` [PATCH 2/5] ARM: dts: Device tree for AXM55xx Anders Berg 2014-04-15 12:42 ` Marc Zyngier 2014-04-15 14:56 ` Anders Berg 2014-04-15 12:45 ` Arnd Bergmann 2014-04-15 15:20 ` Anders Berg 2014-04-15 15:23 ` Arnd Bergmann 2014-04-22 21:21 ` Linus Walleij 2014-04-22 21:31 ` Linus Walleij 2014-04-22 21:39 ` Russell King - ARM Linux 2014-04-15 12:06 ` [PATCH 3/5] ARM: axxia: Adding defconfig " Anders Berg 2014-04-15 12:06 ` [PATCH 4/5] power: reset: Add Axxia system reset driver Anders Berg 2014-04-15 12:06 ` [PATCH 5/5] ARM: dts: axxia: Add reset controller Anders Berg
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).