diff options
Diffstat (limited to 'arch/arm')
83 files changed, 4675 insertions, 920 deletions
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig index 48c19d44be9a..8cfc3ebdcf03 100644 --- a/arch/arm/Kconfig +++ b/arch/arm/Kconfig @@ -1179,12 +1179,6 @@ config XSCALE_PMU depends on CPU_XSCALE default y -config CPU_HAS_PMU - depends on (CPU_V6 || CPU_V6K || CPU_V7 || XSCALE_PMU) && \ - (!ARCH_OMAP3 || OMAP3_EMU) - default y - bool - config MULTI_IRQ_HANDLER bool help @@ -1563,6 +1557,53 @@ config SCHED_SMT MultiThreading at a cost of slightly increased overhead in some places. If unsure say N here. +config DISABLE_CPU_SCHED_DOMAIN_BALANCE + bool "(EXPERIMENTAL) Disable CPU level scheduler load-balancing" + help + Disables scheduler load-balancing at CPU sched domain level. + +config SCHED_HMP + bool "(EXPERIMENTAL) Heterogenous multiprocessor scheduling" + depends on DISABLE_CPU_SCHED_DOMAIN_BALANCE && SCHED_MC && FAIR_GROUP_SCHED && !SCHED_AUTOGROUP + help + Experimental scheduler optimizations for heterogeneous platforms. + Attempts to introspectively select task affinity to optimize power + and performance. Basic support for multiple (>2) cpu types is in place, + but it has only been tested with two types of cpus. + There is currently no support for migration of task groups, hence + !SCHED_AUTOGROUP. Furthermore, normal load-balancing must be disabled + between cpus of different type (DISABLE_CPU_SCHED_DOMAIN_BALANCE). + +config SCHED_HMP_PRIO_FILTER + bool "(EXPERIMENTAL) Filter HMP migrations by task priority" + depends on SCHED_HMP + default y + help + Enables task priority based HMP migration filter. Any task with + a NICE value above the threshold will always be on low-power cpus + with less compute capacity. + +config SCHED_HMP_PRIO_FILTER_VAL + int "NICE priority threshold" + default 5 + depends on SCHED_HMP_PRIO_FILTER + +config HMP_FAST_CPU_MASK + string "HMP scheduler fast CPU mask" + depends on SCHED_HMP + help + Leave empty to use device tree information. + Specify the cpuids of the fast CPUs in the system as a list string, + e.g. cpuid 0+1 should be specified as 0-1. + +config HMP_SLOW_CPU_MASK + string "HMP scheduler slow CPU mask" + depends on SCHED_HMP + help + Leave empty to use device tree information. + Specify the cpuids of the slow CPUs in the system as a list string, + e.g. cpuid 0+1 should be specified as 0-1. + config HAVE_ARM_SCU bool help @@ -1767,7 +1808,7 @@ config HIGHPTE config HW_PERF_EVENTS bool "Enable hardware performance counter support for perf events" - depends on PERF_EVENTS && CPU_HAS_PMU + depends on PERF_EVENTS default y help Enable hardware performance counter support for perf events. If @@ -1907,6 +1948,15 @@ config DEPRECATED_PARAM_STRUCT This was deprecated in 2001 and announced to live on for 5 years. Some old boot loaders still use this way. +config ARM_FLUSH_CONSOLE_ON_RESTART + bool "Force flush the console on restart" + help + If the console is locked while the system is rebooted, the messages + in the temporary logbuffer would not have propogated to all the + console drivers. This option forces the console lock to be + released if it failed to be acquired, which will cause all the + pending messages to be flushed. + endmenu menu "Boot options" @@ -2337,6 +2387,8 @@ source "net/Kconfig" source "drivers/Kconfig" +source "ubuntu/Kconfig" + source "fs/Kconfig" source "arch/arm/Kconfig.debug" diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug index e968a52e4881..8df0327909ac 100644 --- a/arch/arm/Kconfig.debug +++ b/arch/arm/Kconfig.debug @@ -63,6 +63,27 @@ config DEBUG_USER 8 - SIGSEGV faults 16 - SIGBUS faults +config DEBUG_RODATA + bool "Write protect kernel text section" + default n + depends on DEBUG_KERNEL && MMU + ---help--- + Mark the kernel text section as write-protected in the pagetables, + in order to catch accidental (and incorrect) writes to such const + data. This will cause the size of the kernel, plus up to 4MB, to + be mapped as pages instead of sections, which will increase TLB + pressure. + If in doubt, say "N". + +config DEBUG_RODATA_TEST + bool "Testcase for the DEBUG_RODATA feature" + depends on DEBUG_RODATA + default n + ---help--- + This option enables a testcase for the DEBUG_RODATA + feature. + If in doubt, say "N" + # These options are only for real kernel hackers who want to get their hands dirty. config DEBUG_LL bool "Kernel low-level debugging functions (read help!)" diff --git a/arch/arm/boot/compressed/head.S b/arch/arm/boot/compressed/head.S index bc67cbff3944..713ee96cd68d 100644 --- a/arch/arm/boot/compressed/head.S +++ b/arch/arm/boot/compressed/head.S @@ -755,6 +755,8 @@ call_cache_fn: adr r12, proc_types .align 2 .type proc_types,#object proc_types: +#if !defined(CONFIG_CPU_V7) + /* This collides with some V7 IDs, preventing correct detection */ .word 0x00000000 @ old ARM ID .word 0x0000f000 mov pc, lr @@ -763,6 +765,7 @@ proc_types: THUMB( nop ) mov pc, lr THUMB( nop ) +#endif .word 0x41007000 @ ARM7/710 .word 0xfff8fe00 diff --git a/arch/arm/boot/dts/genesi-efikamx.dts b/arch/arm/boot/dts/genesi-efikamx.dts new file mode 100644 index 000000000000..e81ffcc8443b --- /dev/null +++ b/arch/arm/boot/dts/genesi-efikamx.dts @@ -0,0 +1,22 @@ +/* + * Copyright 2011 Linaro Ltd. + * + * The code contained herein is licensed under the GNU General Public + * License. You may obtain a copy of the GNU General Public License + * Version 2 or later at the following locations: + * + * http://www.opensource.org/licenses/gpl-license.html + * http://www.gnu.org/copyleft/gpl.html + */ + +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "Genesi EfikaMX nettop"; + compatible = "genesi,efikamx", "fsl,mx51"; + + memory { + reg = <0x90000000 0x20000000>; + }; +}; diff --git a/arch/arm/boot/dts/genesi-efikasb.dts b/arch/arm/boot/dts/genesi-efikasb.dts new file mode 100644 index 000000000000..9fda6ae314e4 --- /dev/null +++ b/arch/arm/boot/dts/genesi-efikasb.dts @@ -0,0 +1,22 @@ +/* + * Copyright 2011 Linaro Ltd. + * + * The code contained herein is licensed under the GNU General Public + * License. You may obtain a copy of the GNU General Public License + * Version 2 or later at the following locations: + * + * http://www.opensource.org/licenses/gpl-license.html + * http://www.gnu.org/copyleft/gpl.html + */ + +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "Genesi Efika Smartbook"; + compatible = "genesi,efikasb", "fsl,mx51"; + + memory { + reg = <0x90000000 0x20000000>; + }; +}; diff --git a/arch/arm/boot/dts/highbank.dts b/arch/arm/boot/dts/highbank.dts index 9fecf1ae777b..d7c3e29f7663 100644 --- a/arch/arm/boot/dts/highbank.dts +++ b/arch/arm/boot/dts/highbank.dts @@ -144,6 +144,10 @@ clock-names = "apb_pclk"; }; + ipc-keys { + compatible = "calxeda,hb-keys"; + }; + gpioe: gpio@fff30000 { #gpio-cells = <2>; compatible = "arm,pl061", "arm,primecell"; diff --git a/arch/arm/boot/dts/isee-igep-v2.dts b/arch/arm/boot/dts/isee-igep-v2.dts new file mode 100644 index 000000000000..72caabb85b72 --- /dev/null +++ b/arch/arm/boot/dts/isee-igep-v2.dts @@ -0,0 +1,7 @@ +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "ISSE IGEPv2 Board"; + compatible = "ISEE,igep-v2"; +}; diff --git a/arch/arm/boot/dts/isee-igep-v3.dts b/arch/arm/boot/dts/isee-igep-v3.dts new file mode 100644 index 000000000000..f40886fef69f --- /dev/null +++ b/arch/arm/boot/dts/isee-igep-v3.dts @@ -0,0 +1,7 @@ +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "ISSE IGEPv3 Module"; + compatible = "ISEE,igep-v3"; +}; diff --git a/arch/arm/boot/dts/mx51-babbage.dts b/arch/arm/boot/dts/mx51-babbage.dts new file mode 100644 index 000000000000..e5e9c8913d01 --- /dev/null +++ b/arch/arm/boot/dts/mx51-babbage.dts @@ -0,0 +1,22 @@ +/* + * Copyright 2011 Linaro Ltd. + * + * The code contained herein is licensed under the GNU General Public + * License. You may obtain a copy of the GNU General Public License + * Version 2 or later at the following locations: + * + * http://www.opensource.org/licenses/gpl-license.html + * http://www.gnu.org/copyleft/gpl.html + */ + +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "Freescale i.MX51 Babbage"; + compatible = "fsl,mx51-babbage", "fsl,mx51"; + + memory { + reg = <0x90000000 0x20000000>; + }; +}; diff --git a/arch/arm/boot/dts/mx53-loco.dts b/arch/arm/boot/dts/mx53-loco.dts new file mode 100644 index 000000000000..8426c6206149 --- /dev/null +++ b/arch/arm/boot/dts/mx53-loco.dts @@ -0,0 +1,22 @@ +/* + * Copyright 2011 Linaro Ltd. + * + * The code contained herein is licensed under the GNU General Public + * License. You may obtain a copy of the GNU General Public License + * Version 2 or later at the following locations: + * + * http://www.opensource.org/licenses/gpl-license.html + * http://www.gnu.org/copyleft/gpl.html + */ + +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "Freescale i.MX53 LOCO"; + compatible = "fsl,mx53-loco", "fsl,mx53"; + + memory { + reg = <0x70000000 0x40000000>; + }; +}; diff --git a/arch/arm/boot/dts/omap3-overo.dts b/arch/arm/boot/dts/omap3-overo.dts new file mode 100644 index 000000000000..c61f0112bf3a --- /dev/null +++ b/arch/arm/boot/dts/omap3-overo.dts @@ -0,0 +1,7 @@ +/dts-v1/; +/include/ "skeleton.dtsi" + +/ { + model = "Gumstix Overo"; + compatible = "gumstix,omap3-overo"; +}; diff --git a/arch/arm/common/Kconfig b/arch/arm/common/Kconfig index 283fa1d804f4..271dd136753c 100644 --- a/arch/arm/common/Kconfig +++ b/arch/arm/common/Kconfig @@ -40,3 +40,53 @@ config SHARP_PARAM config SHARP_SCOOP bool + +config FIQ_GLUE + bool + select FIQ + +config FIQ_DEBUGGER + bool "FIQ Mode Serial Debugger" + select FIQ + select FIQ_GLUE + default n + help + The FIQ serial debugger can accept commands even when the + kernel is unresponsive due to being stuck with interrupts + disabled. + + +config FIQ_DEBUGGER_NO_SLEEP + bool "Keep serial debugger active" + depends on FIQ_DEBUGGER + default n + help + Enables the serial debugger at boot. Passing + fiq_debugger.no_sleep on the kernel commandline will + override this config option. + +config FIQ_DEBUGGER_WAKEUP_IRQ_ALWAYS_ON + bool "Don't disable wakeup IRQ when debugger is active" + depends on FIQ_DEBUGGER + default n + help + Don't disable the wakeup irq when enabling the uart clock. This will + cause extra interrupts, but it makes the serial debugger usable with + on some MSM radio builds that ignore the uart clock request in power + collapse. + +config FIQ_DEBUGGER_CONSOLE + bool "Console on FIQ Serial Debugger port" + depends on FIQ_DEBUGGER + default n + help + Enables a console so that printk messages are displayed on + the debugger serial port as the occur. + +config FIQ_DEBUGGER_CONSOLE_DEFAULT_ENABLE + bool "Put the FIQ debugger into console mode by default" + depends on FIQ_DEBUGGER_CONSOLE + default n + help + If enabled, this puts the fiq debugger into console mode by default. + Otherwise, the fiq debugger will start out in debug mode. diff --git a/arch/arm/common/Makefile b/arch/arm/common/Makefile index e8a4e58f1b82..091576aabf0e 100644 --- a/arch/arm/common/Makefile +++ b/arch/arm/common/Makefile @@ -13,3 +13,5 @@ obj-$(CONFIG_SHARP_PARAM) += sharpsl_param.o obj-$(CONFIG_SHARP_SCOOP) += scoop.o obj-$(CONFIG_PCI_HOST_ITE8152) += it8152.o obj-$(CONFIG_ARM_TIMER_SP804) += timer-sp.o +obj-$(CONFIG_FIQ_GLUE) += fiq_glue.o fiq_glue_setup.o +obj-$(CONFIG_FIQ_DEBUGGER) += fiq_debugger.o diff --git a/arch/arm/common/fiq_debugger.c b/arch/arm/common/fiq_debugger.c new file mode 100644 index 000000000000..1908ba95a900 --- /dev/null +++ b/arch/arm/common/fiq_debugger.c @@ -0,0 +1,1322 @@ +/* + * arch/arm/common/fiq_debugger.c + * + * Serial Debugger Interface accessed through an FIQ interrupt. + * + * Copyright (C) 2008 Google, Inc. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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 <stdarg.h> +#include <linux/module.h> +#include <linux/io.h> +#include <linux/console.h> +#include <linux/interrupt.h> +#include <linux/clk.h> +#include <linux/platform_device.h> +#include <linux/kernel_stat.h> +#include <linux/irq.h> +#include <linux/delay.h> +#include <linux/reboot.h> +#include <linux/sched.h> +#include <linux/slab.h> +#include <linux/smp.h> +#include <linux/timer.h> +#include <linux/tty.h> +#include <linux/tty_flip.h> +#include <linux/wakelock.h> + +#include <asm/fiq_debugger.h> +#include <asm/fiq_glue.h> +#include <asm/stacktrace.h> + +#include <linux/uaccess.h> + +#include "fiq_debugger_ringbuf.h" + +#define DEBUG_MAX 64 +#define MAX_UNHANDLED_FIQ_COUNT 1000000 + +#define MAX_FIQ_DEBUGGER_PORTS 4 + +#define THREAD_INFO(sp) ((struct thread_info *) \ + ((unsigned long)(sp) & ~(THREAD_SIZE - 1))) + +struct fiq_debugger_state { + struct fiq_glue_handler handler; + + int fiq; + int uart_irq; + int signal_irq; + int wakeup_irq; + bool wakeup_irq_no_set_wake; + struct clk *clk; + struct fiq_debugger_pdata *pdata; + struct platform_device *pdev; + + char debug_cmd[DEBUG_MAX]; + int debug_busy; + int debug_abort; + + char debug_buf[DEBUG_MAX]; + int debug_count; + + bool no_sleep; + bool debug_enable; + bool ignore_next_wakeup_irq; + struct timer_list sleep_timer; + spinlock_t sleep_timer_lock; + bool uart_enabled; + struct wake_lock debugger_wake_lock; + bool console_enable; + int current_cpu; + atomic_t unhandled_fiq_count; + bool in_fiq; + +#ifdef CONFIG_FIQ_DEBUGGER_CONSOLE + struct console console; + struct tty_struct *tty; + int tty_open_count; + struct fiq_debugger_ringbuf *tty_rbuf; + bool syslog_dumping; +#endif + + unsigned int last_irqs[NR_IRQS]; + unsigned int last_local_timer_irqs[NR_CPUS]; +}; + +#ifdef CONFIG_FIQ_DEBUGGER_CONSOLE +struct tty_driver *fiq_tty_driver; +#endif + +#ifdef CONFIG_FIQ_DEBUGGER_NO_SLEEP +static bool initial_no_sleep = true; +#else +static bool initial_no_sleep; +#endif + +#ifdef CONFIG_FIQ_DEBUGGER_CONSOLE_DEFAULT_ENABLE +static bool initial_debug_enable = true; +static bool initial_console_enable = true; +#else +static bool initial_debug_enable; +static bool initial_console_enable; +#endif + +static bool fiq_kgdb_enable; + +module_param_named(no_sleep, initial_no_sleep, bool, 0644); +module_param_named(debug_enable, initial_debug_enable, bool, 0644); +module_param_named(console_enable, initial_console_enable, bool, 0644); +module_param_named(kgdb_enable, fiq_kgdb_enable, bool, 0644); + +#ifdef CONFIG_FIQ_DEBUGGER_WAKEUP_IRQ_ALWAYS_ON +static inline void enable_wakeup_irq(struct fiq_debugger_state *state) {} +static inline void disable_wakeup_irq(struct fiq_debugger_state *state) {} +#else +static inline void enable_wakeup_irq(struct fiq_debugger_state *state) +{ + if (state->wakeup_irq < 0) + return; + enable_irq(state->wakeup_irq); + if (!state->wakeup_irq_no_set_wake) + enable_irq_wake(state->wakeup_irq); +} +static inline void disable_wakeup_irq(struct fiq_debugger_state *state) +{ + if (state->wakeup_irq < 0) + return; + disable_irq_nosync(state->wakeup_irq); + if (!state->wakeup_irq_no_set_wake) + disable_irq_wake(state->wakeup_irq); +} +#endif + +static bool inline debug_have_fiq(struct fiq_debugger_state *state) +{ + return (state->fiq >= 0); +} + +static void debug_force_irq(struct fiq_debugger_state *state) +{ + unsigned int irq = state->signal_irq; + + if (WARN_ON(!debug_have_fiq(state))) + return; + if (state->pdata->force_irq) { + state->pdata->force_irq(state->pdev, irq); + } else { + struct irq_chip *chip = irq_get_chip(irq); + if (chip && chip->irq_retrigger) + chip->irq_retrigger(irq_get_irq_data(irq)); + } +} + +static void debug_uart_enable(struct fiq_debugger_state *state) +{ + if (state->clk) + clk_enable(state->clk); + if (state->pdata->uart_enable) + state->pdata->uart_enable(state->pdev); +} + +static void debug_uart_disable(struct fiq_debugger_state *state) +{ + if (state->pdata->uart_disable) + state->pdata->uart_disable(state->pdev); + if (state->clk) + clk_disable(state->clk); +} + +static void debug_uart_flush(struct fiq_debugger_state *state) +{ + if (state->pdata->uart_flush) + state->pdata->uart_flush(state->pdev); +} + +static void debug_putc(struct fiq_debugger_state *state, char c) +{ + state->pdata->uart_putc(state->pdev, c); +} + +static void debug_puts(struct fiq_debugger_state *state, char *s) +{ + unsigned c; + while ((c = *s++)) { + if (c == '\n') + debug_putc(state, '\r'); + debug_putc(state, c); + } +} + +static void debug_prompt(struct fiq_debugger_state *state) +{ + debug_puts(state, "debug> "); +} + +int log_buf_copy(char *dest, int idx, int len); +static void dump_kernel_log(struct fiq_debugger_state *state) +{ + char buf[1024]; + int idx = 0; + int ret; + int saved_oip; + + /* setting oops_in_progress prevents log_buf_copy() + * from trying to take a spinlock which will make it + * very unhappy in some cases... + */ + saved_oip = oops_in_progress; + oops_in_progress = 1; + for (;;) { + ret = log_buf_copy(buf, idx, 1023); + if (ret <= 0) + break; + buf[ret] = 0; + debug_puts(state, buf); + idx += ret; + } + oops_in_progress = saved_oip; +} + +static char *mode_name(unsigned cpsr) +{ + switch (cpsr & MODE_MASK) { + case USR_MODE: return "USR"; + case FIQ_MODE: return "FIQ"; + case IRQ_MODE: return "IRQ"; + case SVC_MODE: return "SVC"; + case ABT_MODE: return "ABT"; + case UND_MODE: return "UND"; + case SYSTEM_MODE: return "SYS"; + default: return "???"; + } +} + +static int debug_printf(void *cookie, const char *fmt, ...) +{ + struct fiq_debugger_state *state = cookie; + char buf[256]; + va_list ap; + + va_start(ap, fmt); + vsnprintf(buf, sizeof(buf), fmt, ap); + va_end(ap); + + debug_puts(state, buf); + return state->debug_abort; +} + +/* Safe outside fiq context */ +static int debug_printf_nfiq(void *cookie, const char *fmt, ...) +{ + struct fiq_debugger_state *state = cookie; + char buf[256]; + va_list ap; + unsigned long irq_flags; + + va_start(ap, fmt); + vsnprintf(buf, 128, fmt, ap); + va_end(ap); + + local_irq_save(irq_flags); + debug_puts(state, buf); + debug_uart_flush(state); + local_irq_restore(irq_flags); + return state->debug_abort; +} + +static void dump_regs(struct fiq_debugger_state *state, unsigned *regs) +{ + debug_printf(state, " r0 %08x r1 %08x r2 %08x r3 %08x\n", + regs[0], regs[1], regs[2], regs[3]); + debug_printf(state, " r4 %08x r5 %08x r6 %08x r7 %08x\n", + regs[4], regs[5], regs[6], regs[7]); + debug_printf(state, " r8 %08x r9 %08x r10 %08x r11 %08x mode %s\n", + regs[8], regs[9], regs[10], regs[11], + mode_name(regs[16])); + if ((regs[16] & MODE_MASK) == USR_MODE) + debug_printf(state, " ip %08x sp %08x lr %08x pc %08x " + "cpsr %08x\n", regs[12], regs[13], regs[14], + regs[15], regs[16]); + else + debug_printf(state, " ip %08x sp %08x lr %08x pc %08x " + "cpsr %08x spsr %08x\n", regs[12], regs[13], + regs[14], regs[15], regs[16], regs[17]); +} + +struct mode_regs { + unsigned long sp_svc; + unsigned long lr_svc; + unsigned long spsr_svc; + + unsigned long sp_abt; + unsigned long lr_abt; + unsigned long spsr_abt; + + unsigned long sp_und; + unsigned long lr_und; + unsigned long spsr_und; + + unsigned long sp_irq; + unsigned long lr_irq; + unsigned long spsr_irq; + + unsigned long r8_fiq; + unsigned long r9_fiq; + unsigned long r10_fiq; + unsigned long r11_fiq; + unsigned long r12_fiq; + unsigned long sp_fiq; + unsigned long lr_fiq; + unsigned long spsr_fiq; +}; + +void __naked get_mode_regs(struct mode_regs *regs) +{ + asm volatile ( + "mrs r1, cpsr\n" + "msr cpsr_c, #0xd3 @(SVC_MODE | PSR_I_BIT | PSR_F_BIT)\n" + "stmia r0!, {r13 - r14}\n" + "mrs r2, spsr\n" + "msr cpsr_c, #0xd7 @(ABT_MODE | PSR_I_BIT | PSR_F_BIT)\n" + "stmia r0!, {r2, r13 - r14}\n" + "mrs r2, spsr\n" + "msr cpsr_c, #0xdb @(UND_MODE | PSR_I_BIT | PSR_F_BIT)\n" + "stmia r0!, {r2, r13 - r14}\n" + "mrs r2, spsr\n" + "msr cpsr_c, #0xd2 @(IRQ_MODE | PSR_I_BIT | PSR_F_BIT)\n" + "stmia r0!, {r2, r13 - r14}\n" + "mrs r2, spsr\n" + "msr cpsr_c, #0xd1 @(FIQ_MODE | PSR_I_BIT | PSR_F_BIT)\n" + "stmia r0!, {r2, r8 - r14}\n" + "mrs r2, spsr\n" + "stmia r0!, {r2}\n" + "msr cpsr_c, r1\n" + "bx lr\n"); +} + + +static void dump_allregs(struct fiq_debugger_state *state, unsigned *regs) +{ + struct mode_regs mode_regs; + dump_regs(state, regs); + get_mode_regs(&mode_regs); + debug_printf(state, " svc: sp %08x lr %08x spsr %08x\n", + mode_regs.sp_svc, mode_regs.lr_svc, mode_regs.spsr_svc); + debug_printf(state, " abt: sp %08x lr %08x spsr %08x\n", + mode_regs.sp_abt, mode_regs.lr_abt, mode_regs.spsr_abt); + debug_printf(state, " und: sp %08x lr %08x spsr %08x\n", + mode_regs.sp_und, mode_regs.lr_und, mode_regs.spsr_und); + debug_printf(state, " irq: sp %08x lr %08x spsr %08x\n", + mode_regs.sp_irq, mode_regs.lr_irq, mode_regs.spsr_irq); + debug_printf(state, " fiq: r8 %08x r9 %08x r10 %08x r11 %08x " + "r12 %08x\n", + mode_regs.r8_fiq, mode_regs.r9_fiq, mode_regs.r10_fiq, + mode_regs.r11_fiq, mode_regs.r12_fiq); + debug_printf(state, " fiq: sp %08x lr %08x spsr %08x\n", + mode_regs.sp_fiq, mode_regs.lr_fiq, mode_regs.spsr_fiq); +} + +static void dump_irqs(struct fiq_debugger_state *state) +{ + int n; + + debug_printf(state, "irqnr total since-last status name\n"); + for (n = 0; n < NR_IRQS; n++) { + struct irqaction *act = irq_desc[n].action; + if (!act && !kstat_irqs(n)) + continue; + debug_printf(state, "%5d: %10u %11u %8x %s\n", n, + kstat_irqs(n), + kstat_irqs(n) - state->last_irqs[n], + irq_desc[n].status_use_accessors, + (act && act->name) ? act->name : "???"); + state->last_irqs[n] = kstat_irqs(n); + } +} + +struct stacktrace_state { + struct fiq_debugger_state *state; + unsigned int depth; +}; + +static int report_trace(struct stackframe *frame, void *d) +{ + struct stacktrace_state *sts = d; + + if (sts->depth) { + debug_printf(sts->state, + " pc: %p (%pF), lr %p (%pF), sp %p, fp %p\n", + frame->pc, frame->pc, frame->lr, frame->lr, + frame->sp, frame->fp); + sts->depth--; + return 0; + } + debug_printf(sts->state, " ...\n"); + + return sts->depth == 0; +} + +struct frame_tail { + struct frame_tail *fp; + unsigned long sp; + unsigned long lr; +} __attribute__((packed)); + +static struct frame_tail *user_backtrace(struct fiq_debugger_state *state, + struct frame_tail *tail) +{ + struct frame_tail buftail[2]; + + /* Also check accessibility of one struct frame_tail beyond */ + if (!access_ok(VERIFY_READ, tail, sizeof(buftail))) { + debug_printf(state, " invalid frame pointer %p\n", tail); + return NULL; + } + if (__copy_from_user_inatomic(buftail, tail, sizeof(buftail))) { + debug_printf(state, + " failed to copy frame pointer %p\n", tail); + return NULL; + } + + debug_printf(state, " %p\n", buftail[0].lr); + + /* frame pointers should strictly progress back up the stack + * (towards higher addresses) */ + if (tail >= buftail[0].fp) + return NULL; + + return buftail[0].fp-1; +} + +void dump_stacktrace(struct fiq_debugger_state *state, + struct pt_regs * const regs, unsigned int depth, void *ssp) +{ + struct frame_tail *tail; + struct thread_info *real_thread_info = THREAD_INFO(ssp); + struct stacktrace_state sts; + + sts.depth = depth; + sts.state = state; + *current_thread_info() = *real_thread_info; + + if (!current) + debug_printf(state, "current NULL\n"); + else + debug_printf(state, "pid: %d comm: %s\n", + current->pid, current->comm); + dump_regs(state, (unsigned *)regs); + + if (!user_mode(regs)) { + struct stackframe frame; + frame.fp = regs->ARM_fp; + frame.sp = regs->ARM_sp; + frame.lr = regs->ARM_lr; + frame.pc = regs->ARM_pc; + debug_printf(state, + " pc: %p (%pF), lr %p (%pF), sp %p, fp %p\n", + regs->ARM_pc, regs->ARM_pc, regs->ARM_lr, regs->ARM_lr, + regs->ARM_sp, regs->ARM_fp); + walk_stackframe(&frame, report_trace, &sts); + return; + } + + tail = ((struct frame_tail *) regs->ARM_fp) - 1; + while (depth-- && tail && !((unsigned long) tail & 3)) + tail = user_backtrace(state, tail); +} + +static void do_ps(struct fiq_debugger_state *state) +{ + struct task_struct *g; + struct task_struct *p; + unsigned task_state; + static const char stat_nam[] = "RSDTtZX"; + + debug_printf(state, "pid ppid prio task pc\n"); + read_lock(&tasklist_lock); + do_each_thread(g, p) { + task_state = p->state ? __ffs(p->state) + 1 : 0; + debug_printf(state, + "%5d %5d %4d ", p->pid, p->parent->pid, p->prio); + debug_printf(state, "%-13.13s %c", p->comm, + task_state >= sizeof(stat_nam) ? '?' : stat_nam[task_state]); + if (task_state == TASK_RUNNING) + debug_printf(state, " running\n"); + else + debug_printf(state, " %08lx\n", thread_saved_pc(p)); + } while_each_thread(g, p); + read_unlock(&tasklist_lock); +} + +#ifdef CONFIG_FIQ_DEBUGGER_CONSOLE +static void begin_syslog_dump(struct fiq_debugger_state *state) +{ + state->syslog_dumping = true; +} + +static void end_syslog_dump(struct fiq_debugger_state *state) +{ + state->syslog_dumping = false; +} +#else +extern int do_syslog(int type, char __user *bug, int count); +static void begin_syslog_dump(struct fiq_debugger_state *state) +{ + do_syslog(5 /* clear */, NULL, 0); +} + +static void end_syslog_dump(struct fiq_debugger_state *state) +{ + char buf[128]; + int ret; + int idx = 0; + + while (1) { + ret = log_buf_copy(buf, idx, sizeof(buf) - 1); + if (ret <= 0) + break; + buf[ret] = 0; + debug_printf(state, "%s", buf); + idx += ret; + } +} +#endif + +static void do_sysrq(struct fiq_debugger_state *state, char rq) +{ + if ((rq == 'g' || rq == 'G') && !fiq_kgdb_enable) { + debug_printf(state, "sysrq-g blocked\n"); + return; + } + begin_syslog_dump(state); + handle_sysrq(rq); + end_syslog_dump(state); +} + +#ifdef CONFIG_KGDB +static void do_kgdb(struct fiq_debugger_state *state) +{ + if (!fiq_kgdb_enable) { + debug_printf(state, "kgdb through fiq debugger not enabled\n"); + return; + } + + debug_printf(state, "enabling console and triggering kgdb\n"); + state->console_enable = true; + handle_sysrq('g'); +} +#endif + +/* This function CANNOT be called in FIQ context */ +static void debug_irq_exec(struct fiq_debugger_state *state, char *cmd) +{ + if (!strcmp(cmd, "ps")) + do_ps(state); + if (!strcmp(cmd, "sysrq")) + do_sysrq(state, 'h'); + if (!strncmp(cmd, "sysrq ", 6)) + do_sysrq(state, cmd[6]); +#ifdef CONFIG_KGDB + if (!strcmp(cmd, "kgdb")) + do_kgdb(state); +#endif +} + +static void debug_help(struct fiq_debugger_state *state) +{ + debug_printf(state, "FIQ Debugger commands:\n" + " pc PC status\n" + " regs Register dump\n" + " allregs Extended Register dump\n" + " bt Stack trace\n" + " reboot Reboot\n" + " irqs Interupt status\n" + " kmsg Kernel log\n" + " version Kernel version\n"); + debug_printf(state, " sleep Allow sleep while in FIQ\n" + " nosleep Disable sleep while in FIQ\n" + " console Switch terminal to console\n" + " cpu Current CPU\n" + " cpu <number> Switch to CPU<number>\n"); + debug_printf(state, " ps Process list\n" + " sysrq sysrq options\n" + " sysrq <param> Execute sysrq with <param>\n"); +#ifdef CONFIG_KGDB + debug_printf(state, " kgdb Enter kernel debugger\n"); +#endif +} + +static void take_affinity(void *info) +{ + struct fiq_debugger_state *state = info; + struct cpumask cpumask; + + cpumask_clear(&cpumask); + cpumask_set_cpu(get_cpu(), &cpumask); + + irq_set_affinity(state->uart_irq, &cpumask); +} + +static void switch_cpu(struct fiq_debugger_state *state, int cpu) +{ + if (!debug_have_fiq(state)) + smp_call_function_single(cpu, take_affinity, state, false); + state->current_cpu = cpu; +} + +static bool debug_fiq_exec(struct fiq_debugger_state *state, + const char *cmd, unsigned *regs, void *svc_sp) +{ + bool signal_helper = false; + + if (!strcmp(cmd, "help") || !strcmp(cmd, "?")) { + debug_help(state); + } else if (!strcmp(cmd, "pc")) { + debug_printf(state, " pc %08x cpsr %08x mode %s\n", + regs[15], regs[16], mode_name(regs[16])); + } else if (!strcmp(cmd, "regs")) { + dump_regs(state, regs); + } else if (!strcmp(cmd, "allregs")) { + dump_allregs(state, regs); + } else if (!strcmp(cmd, "bt")) { + dump_stacktrace(state, (struct pt_regs *)regs, 100, svc_sp); + } else if (!strncmp(cmd, "reboot", 6)) { + cmd += 6; + while (*cmd == ' ') + cmd++; + if (*cmd) { + char tmp_cmd[32]; + strlcpy(tmp_cmd, cmd, sizeof(tmp_cmd)); + kernel_restart(tmp_cmd); + } else { + kernel_restart(NULL); + } + } else if (!strcmp(cmd, "irqs")) { + dump_irqs(state); + } else if (!strcmp(cmd, "kmsg")) { + dump_kernel_log(state); + } else if (!strcmp(cmd, "version")) { + debug_printf(state, "%s\n", linux_banner); + } else if (!strcmp(cmd, "sleep")) { + state->no_sleep = false; + debug_printf(state, "enabling sleep\n"); + } else if (!strcmp(cmd, "nosleep")) { + state->no_sleep = true; + debug_printf(state, "disabling sleep\n"); + } else if (!strcmp(cmd, "console")) { + state->console_enable = true; + debug_printf(state, "console mode\n"); + } else if (!strcmp(cmd, "cpu")) { + debug_printf(state, "cpu %d\n", state->current_cpu); + } else if (!strncmp(cmd, "cpu ", 4)) { + unsigned long cpu = 0; + if (strict_strtoul(cmd + 4, 10, &cpu) == 0) + switch_cpu(state, cpu); + else + debug_printf(state, "invalid cpu\n"); + debug_printf(state, "cpu %d\n", state->current_cpu); + } else { + if (state->debug_busy) { + debug_printf(state, + "command processor busy. trying to abort.\n"); + state->debug_abort = -1; + } else { + strcpy(state->debug_cmd, cmd); + state->debug_busy = 1; + } + + return true; + } + if (!state->console_enable) + debug_prompt(state); + + return signal_helper; +} + +static void sleep_timer_expired(unsigned long data) +{ + struct fiq_debugger_state *state = (struct fiq_debugger_state *)data; + unsigned long flags; + + spin_lock_irqsave(&state->sleep_timer_lock, flags); + if (state->uart_enabled && !state->no_sleep) { + if (state->debug_enable && !state->console_enable) { + state->debug_enable = false; + debug_printf_nfiq(state, "suspending fiq debugger\n"); + } + state->ignore_next_wakeup_irq = true; + debug_uart_disable(state); + state->uart_enabled = false; + enable_wakeup_irq(state); + } + wake_unlock(&state->debugger_wake_lock); + spin_unlock_irqrestore(&state->sleep_timer_lock, flags); +} + +static void handle_wakeup(struct fiq_debugger_state *state) +{ + unsigned long flags; + + spin_lock_irqsave(&state->sleep_timer_lock, flags); + if (state->wakeup_irq >= 0 && state->ignore_next_wakeup_irq) { + state->ignore_next_wakeup_irq = false; + } else if (!state->uart_enabled) { + wake_lock(&state->debugger_wake_lock); + debug_uart_enable(state); + state->uart_enabled = true; + disable_wakeup_irq(state); + mod_timer(&state->sleep_timer, jiffies + HZ / 2); + } + spin_unlock_irqrestore(&state->sleep_timer_lock, flags); +} + +static irqreturn_t wakeup_irq_handler(int irq, void *dev) +{ + struct fiq_debugger_state *state = dev; + + if (!state->no_sleep) + debug_puts(state, "WAKEUP\n"); + handle_wakeup(state); + + return IRQ_HANDLED; +} + + +static void debug_handle_irq_context(struct fiq_debugger_state *state) +{ + if (!state->no_sleep) { + unsigned long flags; + + spin_lock_irqsave(&state->sleep_timer_lock, flags); + wake_lock(&state->debugger_wake_lock); + mod_timer(&state->sleep_timer, jiffies + HZ * 5); + spin_unlock_irqrestore(&state->sleep_timer_lock, flags); + } +#if defined(CONFIG_FIQ_DEBUGGER_CONSOLE) + if (state->tty) { + int i; + int count = fiq_debugger_ringbuf_level(state->tty_rbuf); + for (i = 0; i < count; i++) { + int c = fiq_debugger_ringbuf_peek(state->tty_rbuf, 0); + tty_insert_flip_char(state->tty, c, TTY_NORMAL); + if (!fiq_debugger_ringbuf_consume(state->tty_rbuf, 1)) + pr_warn("fiq tty failed to consume byte\n"); + } + tty_flip_buffer_push(state->tty); + } +#endif + if (state->debug_busy) { + debug_irq_exec(state, state->debug_cmd); + if (!state->console_enable) + debug_prompt(state); + state->debug_busy = 0; + } +} + +static int debug_getc(struct fiq_debugger_state *state) +{ + return state->pdata->uart_getc(state->pdev); +} + +static bool debug_handle_uart_interrupt(struct fiq_debugger_state *state, + int this_cpu, void *regs, void *svc_sp) +{ + int c; + static int last_c; + int count = 0; + bool signal_helper = false; + + if (this_cpu != state->current_cpu) { + if (state->in_fiq) + return false; + + if (atomic_inc_return(&state->unhandled_fiq_count) != + MAX_UNHANDLED_FIQ_COUNT) + return false; + + debug_printf(state, "fiq_debugger: cpu %d not responding, " + "reverting to cpu %d\n", state->current_cpu, + this_cpu); + + atomic_set(&state->unhandled_fiq_count, 0); + switch_cpu(state, this_cpu); + return false; + } + + state->in_fiq = true; + + while ((c = debug_getc(state)) != FIQ_DEBUGGER_NO_CHAR) { + count++; + if (!state->debug_enable) { + if ((c == 13) || (c == 10)) { + state->debug_enable = true; + state->debug_count = 0; + debug_prompt(state); + } + } else if (c == FIQ_DEBUGGER_BREAK) { + state->console_enable = false; + debug_puts(state, "fiq debugger mode\n"); + state->debug_count = 0; + debug_prompt(state); +#ifdef CONFIG_FIQ_DEBUGGER_CONSOLE + } else if (state->console_enable && state->tty_rbuf) { + fiq_debugger_ringbuf_push(state->tty_rbuf, c); + signal_helper = true; +#endif + } else if ((c >= ' ') && (c < 127)) { + if (state->debug_count < (DEBUG_MAX - 1)) { + state->debug_buf[state->debug_count++] = c; + debug_putc(state, c); + } + } else if ((c == 8) || (c == 127)) { + if (state->debug_count > 0) { + state->debug_count--; + debug_putc(state, 8); + debug_putc(state, ' '); + debug_putc(state, 8); + } + } else if ((c == 13) || (c == 10)) { + if (c == '\r' || (c == '\n' && last_c != '\r')) { + debug_putc(state, '\r'); + debug_putc(state, '\n'); + } + if (state->debug_count) { + state->debug_buf[state->debug_count] = 0; + state->debug_count = 0; + signal_helper |= + debug_fiq_exec(state, state->debug_buf, + regs, svc_sp); + } else { + debug_prompt(state); + } + } + last_c = c; + } + debug_uart_flush(state); + if (state->pdata->fiq_ack) + state->pdata->fiq_ack(state->pdev, state->fiq); + + /* poke sleep timer if necessary */ + if (state->debug_enable && !state->no_sleep) + signal_helper = true; + + atomic_set(&state->unhandled_fiq_count, 0); + state->in_fiq = false; + + return signal_helper; +} + +static void debug_fiq(struct fiq_glue_handler *h, void *regs, void *svc_sp) +{ + struct fiq_debugger_state *state = + container_of(h, struct fiq_debugger_state, handler); + unsigned int this_cpu = THREAD_INFO(svc_sp)->cpu; + bool need_irq; + + need_irq = debug_handle_uart_interrupt(state, this_cpu, regs, svc_sp); + if (need_irq) + debug_force_irq(state); +} + +/* + * When not using FIQs, we only use this single interrupt as an entry point. + * This just effectively takes over the UART interrupt and does all the work + * in this context. + */ +static irqreturn_t debug_uart_irq(int irq, void *dev) +{ + struct fiq_debugger_state *state = dev; + bool not_done; + + handle_wakeup(state); + + /* handle the debugger irq in regular context */ + not_done = debug_handle_uart_interrupt(state, smp_processor_id(), + get_irq_regs(), + current_thread_info()); + if (not_done) + debug_handle_irq_context(state); + + return IRQ_HANDLED; +} + +/* + * If FIQs are used, not everything can happen in fiq context. + * FIQ handler does what it can and then signals this interrupt to finish the + * job in irq context. + */ +static irqreturn_t debug_signal_irq(int irq, void *dev) +{ + struct fiq_debugger_state *state = dev; + + if (state->pdata->force_irq_ack) + state->pdata->force_irq_ack(state->pdev, state->signal_irq); + + debug_handle_irq_context(state); + + return IRQ_HANDLED; +} + +static void debug_resume(struct fiq_glue_handler *h) +{ + struct fiq_debugger_state *state = + container_of(h, struct fiq_debugger_state, handler); + if (state->pdata->uart_resume) + state->pdata->uart_resume(state->pdev); +} + +#if defined(CONFIG_FIQ_DEBUGGER_CONSOLE) +struct tty_driver *debug_console_device(struct console *co, int *index) +{ + *index = co->index; + return fiq_tty_driver; +} + +static void debug_console_write(struct console *co, + const char *s, unsigned int count) +{ + struct fiq_debugger_state *state; + + state = container_of(co, struct fiq_debugger_state, console); + + if (!state->console_enable && !state->syslog_dumping) + return; + + debug_uart_enable(state); + while (count--) { + if (*s == '\n') + debug_putc(state, '\r'); + debug_putc(state, *s++); + } + debug_uart_flush(state); + debug_uart_disable(state); +} + +static struct console fiq_debugger_console = { + .name = "ttyFIQ", + .device = debug_console_device, + .write = debug_console_write, + .flags = CON_PRINTBUFFER | CON_ANYTIME | CON_ENABLED, +}; + +int fiq_tty_open(struct tty_struct *tty, struct file *filp) +{ + int line = tty->index; + struct fiq_debugger_state **states = tty->driver->driver_state; + struct fiq_debugger_state *state = states[line]; + if (state->tty_open_count++) + return 0; + + tty->driver_data = state; + state->tty = tty; + return 0; +} + +void fiq_tty_close(struct tty_struct *tty, struct file *filp) +{ + struct fiq_debugger_state *state = tty->driver_data; + if (--state->tty_open_count) + return; + state->tty = NULL; +} + +int fiq_tty_write(struct tty_struct *tty, const unsigned char *buf, int count) +{ + int i; + struct fiq_debugger_state *state = tty->driver_data; + + if (!state->console_enable) + return count; + + debug_uart_enable(state); + for (i = 0; i < count; i++) + debug_putc(state, *buf++); + debug_uart_disable(state); + + return count; +} + +int fiq_tty_write_room(struct tty_struct *tty) +{ + return 1024; +} + +#ifdef CONFIG_CONSOLE_POLL +static int fiq_tty_poll_init(struct tty_driver *driver, int line, char *options) +{ + return 0; +} + +static int fiq_tty_poll_get_char(struct tty_driver *driver, int line) +{ + struct fiq_debugger_state *state = driver->ttys[line]->driver_data; + int c = NO_POLL_CHAR; + + debug_uart_enable(state); + if (debug_have_fiq(state)) { + int count = fiq_debugger_ringbuf_level(state->tty_rbuf); + if (count > 0) { + c = fiq_debugger_ringbuf_peek(state->tty_rbuf, 0); + fiq_debugger_ringbuf_consume(state->tty_rbuf, 1); + } + } else { + c = debug_getc(state); + if (c == FIQ_DEBUGGER_NO_CHAR) + c = NO_POLL_CHAR; + } + debug_uart_disable(state); + + return c; +} + +static void fiq_tty_poll_put_char(struct tty_driver *driver, int line, char ch) +{ + struct fiq_debugger_state *state = driver->ttys[line]->driver_data; + debug_uart_enable(state); + debug_putc(state, ch); + debug_uart_disable(state); +} +#endif + +static const struct tty_operations fiq_tty_driver_ops = { + .write = fiq_tty_write, + .write_room = fiq_tty_write_room, + .open = fiq_tty_open, + .close = fiq_tty_close, +#ifdef CONFIG_CONSOLE_POLL + .poll_init = fiq_tty_poll_init, + .poll_get_char = fiq_tty_poll_get_char, + .poll_put_char = fiq_tty_poll_put_char, +#endif +}; + +static int fiq_debugger_tty_init(void) +{ + int ret; + struct fiq_debugger_state **states = NULL; + + states = kzalloc(sizeof(*states) * MAX_FIQ_DEBUGGER_PORTS, GFP_KERNEL); + if (!states) { + pr_err("Failed to allocate fiq debugger state structres\n"); + return -ENOMEM; + } + + fiq_tty_driver = alloc_tty_driver(MAX_FIQ_DEBUGGER_PORTS); + if (!fiq_tty_driver) { + pr_err("Failed to allocate fiq debugger tty\n"); + ret = -ENOMEM; + goto err_free_state; + } + + fiq_tty_driver->owner = THIS_MODULE; + fiq_tty_driver->driver_name = "fiq-debugger"; + fiq_tty_driver->name = "ttyFIQ"; + fiq_tty_driver->type = TTY_DRIVER_TYPE_SERIAL; + fiq_tty_driver->subtype = SERIAL_TYPE_NORMAL; + fiq_tty_driver->init_termios = tty_std_termios; + fiq_tty_driver->flags = TTY_DRIVER_REAL_RAW | + TTY_DRIVER_DYNAMIC_DEV; + fiq_tty_driver->driver_state = states; + + fiq_tty_driver->init_termios.c_cflag = + B115200 | CS8 | CREAD | HUPCL | CLOCAL; + fiq_tty_driver->init_termios.c_ispeed = 115200; + fiq_tty_driver->init_termios.c_ospeed = 115200; + + tty_set_operations(fiq_tty_driver, &fiq_tty_driver_ops); + + ret = tty_register_driver(fiq_tty_driver); + if (ret) { + pr_err("Failed to register fiq tty: %d\n", ret); + goto err_free_tty; + } + + pr_info("Registered FIQ tty driver\n"); + return 0; + +err_free_tty: + put_tty_driver(fiq_tty_driver); + fiq_tty_driver = NULL; +err_free_state: + kfree(states); + return ret; +} + +static int fiq_debugger_tty_init_one(struct fiq_debugger_state *state) +{ + int ret; + struct device *tty_dev; + struct fiq_debugger_state **states = fiq_tty_driver->driver_state; + + states[state->pdev->id] = state; + + state->tty_rbuf = fiq_debugger_ringbuf_alloc(1024); + if (!state->tty_rbuf) { + pr_err("Failed to allocate fiq debugger ringbuf\n"); + ret = -ENOMEM; + goto err; + } + + tty_dev = tty_register_device(fiq_tty_driver, state->pdev->id, + &state->pdev->dev); + if (IS_ERR(tty_dev)) { + pr_err("Failed to register fiq debugger tty device\n"); + ret = PTR_ERR(tty_dev); + goto err; + } + + device_set_wakeup_capable(tty_dev, 1); + + pr_info("Registered fiq debugger ttyFIQ%d\n", state->pdev->id); + + return 0; + +err: + fiq_debugger_ringbuf_free(state->tty_rbuf); + state->tty_rbuf = NULL; + return ret; +} +#endif + +static int fiq_debugger_dev_suspend(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct fiq_debugger_state *state = platform_get_drvdata(pdev); + + if (state->pdata->uart_dev_suspend) + return state->pdata->uart_dev_suspend(pdev); + return 0; +} + +static int fiq_debugger_dev_resume(struct device *dev) +{ + struct platform_device *pdev = to_platform_device(dev); + struct fiq_debugger_state *state = platform_get_drvdata(pdev); + + if (state->pdata->uart_dev_resume) + return state->pdata->uart_dev_resume(pdev); + return 0; +} + +static int fiq_debugger_probe(struct platform_device *pdev) +{ + int ret; + struct fiq_debugger_pdata *pdata = dev_get_platdata(&pdev->dev); + struct fiq_debugger_state *state; + int fiq; + int uart_irq; + + if (pdev->id >= MAX_FIQ_DEBUGGER_PORTS) + return -EINVAL; + + if (!pdata->uart_getc || !pdata->uart_putc) + return -EINVAL; + if ((pdata->uart_enable && !pdata->uart_disable) || + (!pdata->uart_enable && pdata->uart_disable)) + return -EINVAL; + + fiq = platform_get_irq_byname(pdev, "fiq"); + uart_irq = platform_get_irq_byname(pdev, "uart_irq"); + + /* uart_irq mode and fiq mode are mutually exclusive, but one of them + * is required */ + if ((uart_irq < 0 && fiq < 0) || (uart_irq >= 0 && fiq >= 0)) + return -EINVAL; + if (fiq >= 0 && !pdata->fiq_enable) + return -EINVAL; + + state = kzalloc(sizeof(*state), GFP_KERNEL); + setup_timer(&state->sleep_timer, sleep_timer_expired, + (unsigned long)state); + state->pdata = pdata; + state->pdev = pdev; + state->no_sleep = initial_no_sleep; + state->debug_enable = initial_debug_enable; + state->console_enable = initial_console_enable; + + state->fiq = fiq; + state->uart_irq = uart_irq; + state->signal_irq = platform_get_irq_byname(pdev, "signal"); + state->wakeup_irq = platform_get_irq_byname(pdev, "wakeup"); + + platform_set_drvdata(pdev, state); + + spin_lock_init(&state->sleep_timer_lock); + + if (state->wakeup_irq < 0 && debug_have_fiq(state)) + state->no_sleep = true; + state->ignore_next_wakeup_irq = !state->no_sleep; + + wake_lock_init(&state->debugger_wake_lock, + WAKE_LOCK_SUSPEND, "serial-debug"); + + state->clk = clk_get(&pdev->dev, NULL); + if (IS_ERR(state->clk)) + state->clk = NULL; + + /* do not call pdata->uart_enable here since uart_init may still + * need to do some initialization before uart_enable can work. + * So, only try to manage the clock during init. + */ + if (state->clk) + clk_enable(state->clk); + + if (pdata->uart_init) { + ret = pdata->uart_init(pdev); + if (ret) + goto err_uart_init; + } + + debug_printf_nfiq(state, "<hit enter %sto activate fiq debugger>\n", + state->no_sleep ? "" : "twice "); + + if (debug_have_fiq(state)) { + state->handler.fiq = debug_fiq; + state->handler.resume = debug_resume; + ret = fiq_glue_register_handler(&state->handler); + if (ret) { + pr_err("%s: could not install fiq handler\n", __func__); + goto err_register_fiq; + } + + pdata->fiq_enable(pdev, state->fiq, 1); + } else { + ret = request_irq(state->uart_irq, debug_uart_irq, + IRQF_NO_SUSPEND, "debug", state); + if (ret) { + pr_err("%s: could not install irq handler\n", __func__); + goto err_register_irq; + } + + /* for irq-only mode, we want this irq to wake us up, if it + * can. + */ + enable_irq_wake(state->uart_irq); + } + + if (state->clk) + clk_disable(state->clk); + + if (state->signal_irq >= 0) { + ret = request_irq(state->signal_irq, debug_signal_irq, + IRQF_TRIGGER_RISING, "debug-signal", state); + if (ret) + pr_err("serial_debugger: could not install signal_irq"); + } + + if (state->wakeup_irq >= 0) { + ret = request_irq(state->wakeup_irq, wakeup_irq_handler, + IRQF_TRIGGER_FALLING | IRQF_DISABLED, + "debug-wakeup", state); + if (ret) { + pr_err("serial_debugger: " + "could not install wakeup irq\n"); + state->wakeup_irq = -1; + } else { + ret = enable_irq_wake(state->wakeup_irq); + if (ret) { + pr_err("serial_debugger: " + "could not enable wakeup\n"); + state->wakeup_irq_no_set_wake = true; + } + } + } + if (state->no_sleep) + handle_wakeup(state); + +#if defined(CONFIG_FIQ_DEBUGGER_CONSOLE) + state->console = fiq_debugger_console; + state->console.index = pdev->id; + if (!console_set_on_cmdline) + add_preferred_console(state->console.name, + state->console.index, NULL); + register_console(&state->console); + fiq_debugger_tty_init_one(state); +#endif + return 0; + +err_register_irq: +err_register_fiq: + if (pdata->uart_free) + pdata->uart_free(pdev); +err_uart_init: + if (state->clk) + clk_disable(state->clk); + if (state->clk) + clk_put(state->clk); + wake_lock_destroy(&state->debugger_wake_lock); + platform_set_drvdata(pdev, NULL); + kfree(state); + return ret; +} + +static const struct dev_pm_ops fiq_debugger_dev_pm_ops = { + .suspend = fiq_debugger_dev_suspend, + .resume = fiq_debugger_dev_resume, +}; + +static struct platform_driver fiq_debugger_driver = { + .probe = fiq_debugger_probe, + .driver = { + .name = "fiq_debugger", + .pm = &fiq_debugger_dev_pm_ops, + }, +}; + +static int __init fiq_debugger_init(void) +{ + fiq_debugger_tty_init(); + return platform_driver_register(&fiq_debugger_driver); +} + +postcore_initcall(fiq_debugger_init); diff --git a/arch/arm/common/fiq_debugger_ringbuf.h b/arch/arm/common/fiq_debugger_ringbuf.h new file mode 100644 index 000000000000..2649b5581088 --- /dev/null +++ b/arch/arm/common/fiq_debugger_ringbuf.h @@ -0,0 +1,94 @@ +/* + * arch/arm/common/fiq_debugger_ringbuf.c + * + * simple lockless ringbuffer + * + * Copyright (C) 2010 Google, Inc. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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/kernel.h> +#include <linux/slab.h> + +struct fiq_debugger_ringbuf { + int len; + int head; + int tail; + u8 buf[]; +}; + + +static inline struct fiq_debugger_ringbuf *fiq_debugger_ringbuf_alloc(int len) +{ + struct fiq_debugger_ringbuf *rbuf; + + rbuf = kzalloc(sizeof(*rbuf) + len, GFP_KERNEL); + if (rbuf == NULL) + return NULL; + + rbuf->len = len; + rbuf->head = 0; + rbuf->tail = 0; + smp_mb(); + + return rbuf; +} + +static inline void fiq_debugger_ringbuf_free(struct fiq_debugger_ringbuf *rbuf) +{ + kfree(rbuf); +} + +static inline int fiq_debugger_ringbuf_level(struct fiq_debugger_ringbuf *rbuf) +{ + int level = rbuf->head - rbuf->tail; + + if (level < 0) + level = rbuf->len + level; + + return level; +} + +static inline int fiq_debugger_ringbuf_room(struct fiq_debugger_ringbuf *rbuf) +{ + return rbuf->len - fiq_debugger_ringbuf_level(rbuf) - 1; +} + +static inline u8 +fiq_debugger_ringbuf_peek(struct fiq_debugger_ringbuf *rbuf, int i) +{ + return rbuf->buf[(rbuf->tail + i) % rbuf->len]; +} + +static inline int +fiq_debugger_ringbuf_consume(struct fiq_debugger_ringbuf *rbuf, int count) +{ + count = min(count, fiq_debugger_ringbuf_level(rbuf)); + + rbuf->tail = (rbuf->tail + count) % rbuf->len; + smp_mb(); + + return count; +} + +static inline int +fiq_debugger_ringbuf_push(struct fiq_debugger_ringbuf *rbuf, u8 datum) +{ + if (fiq_debugger_ringbuf_room(rbuf) == 0) + return 0; + + rbuf->buf[rbuf->head] = datum; + smp_mb(); + rbuf->head = (rbuf->head + 1) % rbuf->len; + smp_mb(); + + return 1; +} diff --git a/arch/arm/common/fiq_glue.S b/arch/arm/common/fiq_glue.S new file mode 100644 index 000000000000..9e3455a09f8f --- /dev/null +++ b/arch/arm/common/fiq_glue.S @@ -0,0 +1,111 @@ +/* + * Copyright (C) 2008 Google, Inc. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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/linkage.h> +#include <asm/assembler.h> + + .text + + .global fiq_glue_end + + /* fiq stack: r0-r15,cpsr,spsr of interrupted mode */ + +ENTRY(fiq_glue) + /* store pc, cpsr from previous mode */ + mrs r12, spsr + sub r11, lr, #4 + subs r10, #1 + bne nested_fiq + + stmfd sp!, {r11-r12, lr} + + /* store r8-r14 from previous mode */ + sub sp, sp, #(7 * 4) + stmia sp, {r8-r14}^ + nop + + /* store r0-r7 from previous mode */ + stmfd sp!, {r0-r7} + + /* setup func(data,regs) arguments */ + mov r0, r9 + mov r1, sp + mov r3, r8 + + mov r7, sp + + /* Get sp and lr from non-user modes */ + and r4, r12, #MODE_MASK + cmp r4, #USR_MODE + beq fiq_from_usr_mode + + mov r7, sp + orr r4, r4, #(PSR_I_BIT | PSR_F_BIT) + msr cpsr_c, r4 + str sp, [r7, #(4 * 13)] + str lr, [r7, #(4 * 14)] + mrs r5, spsr + str r5, [r7, #(4 * 17)] + + cmp r4, #(SVC_MODE | PSR_I_BIT | PSR_F_BIT) + /* use fiq stack if we reenter this mode */ + subne sp, r7, #(4 * 3) + +fiq_from_usr_mode: + msr cpsr_c, #(SVC_MODE | PSR_I_BIT | PSR_F_BIT) + mov r2, sp + sub sp, r7, #12 + stmfd sp!, {r2, ip, lr} + /* call func(data,regs) */ + blx r3 + ldmfd sp, {r2, ip, lr} + mov sp, r2 + + /* restore/discard saved state */ + cmp r4, #USR_MODE + beq fiq_from_usr_mode_exit + + msr cpsr_c, r4 + ldr sp, [r7, #(4 * 13)] + ldr lr, [r7, #(4 * 14)] + msr spsr_cxsf, r5 + +fiq_from_usr_mode_exit: + msr cpsr_c, #(FIQ_MODE | PSR_I_BIT | PSR_F_BIT) + + ldmfd sp!, {r0-r7} + add sp, sp, #(7 * 4) + ldmfd sp!, {r11-r12, lr} +exit_fiq: + msr spsr_cxsf, r12 + add r10, #1 + movs pc, r11 + +nested_fiq: + orr r12, r12, #(PSR_F_BIT) + b exit_fiq + +fiq_glue_end: + +ENTRY(fiq_glue_setup) /* func, data, sp */ + mrs r3, cpsr + msr cpsr_c, #(FIQ_MODE | PSR_I_BIT | PSR_F_BIT) + movs r8, r0 + mov r9, r1 + mov sp, r2 + moveq r10, #0 + movne r10, #1 + msr cpsr_c, r3 + bx lr + diff --git a/arch/arm/common/fiq_glue_setup.c b/arch/arm/common/fiq_glue_setup.c new file mode 100644 index 000000000000..4044c7db95c8 --- /dev/null +++ b/arch/arm/common/fiq_glue_setup.c @@ -0,0 +1,100 @@ +/* + * Copyright (C) 2010 Google, Inc. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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/kernel.h> +#include <linux/percpu.h> +#include <linux/slab.h> +#include <asm/fiq.h> +#include <asm/fiq_glue.h> + +extern unsigned char fiq_glue, fiq_glue_end; +extern void fiq_glue_setup(void *func, void *data, void *sp); + +static struct fiq_handler fiq_debbuger_fiq_handler = { + .name = "fiq_glue", +}; +DEFINE_PER_CPU(void *, fiq_stack); +static struct fiq_glue_handler *current_handler; +static DEFINE_MUTEX(fiq_glue_lock); + +static void fiq_glue_setup_helper(void *info) +{ + struct fiq_glue_handler *handler = info; + fiq_glue_setup(handler->fiq, handler, + __get_cpu_var(fiq_stack) + THREAD_START_SP); +} + +int fiq_glue_register_handler(struct fiq_glue_handler *handler) +{ + int ret; + int cpu; + + if (!handler || !handler->fiq) + return -EINVAL; + + mutex_lock(&fiq_glue_lock); + if (fiq_stack) { + ret = -EBUSY; + goto err_busy; + } + + for_each_possible_cpu(cpu) { + void *stack; + stack = (void *)__get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER); + if (WARN_ON(!stack)) { + ret = -ENOMEM; + goto err_alloc_fiq_stack; + } + per_cpu(fiq_stack, cpu) = stack; + } + + ret = claim_fiq(&fiq_debbuger_fiq_handler); + if (WARN_ON(ret)) + goto err_claim_fiq; + + current_handler = handler; + on_each_cpu(fiq_glue_setup_helper, handler, true); + set_fiq_handler(&fiq_glue, &fiq_glue_end - &fiq_glue); + + mutex_unlock(&fiq_glue_lock); + return 0; + +err_claim_fiq: +err_alloc_fiq_stack: + for_each_possible_cpu(cpu) { + __free_pages(per_cpu(fiq_stack, cpu), THREAD_SIZE_ORDER); + per_cpu(fiq_stack, cpu) = NULL; + } +err_busy: + mutex_unlock(&fiq_glue_lock); + return ret; +} + +/** + * fiq_glue_resume - Restore fiqs after suspend or low power idle states + * + * This must be called before calling local_fiq_enable after returning from a + * power state where the fiq mode registers were lost. If a driver provided + * a resume hook when it registered the handler it will be called. + */ + +void fiq_glue_resume(void) +{ + if (!current_handler) + return; + fiq_glue_setup(current_handler->fiq, current_handler, + __get_cpu_var(fiq_stack) + THREAD_START_SP); + if (current_handler->resume) + current_handler->resume(current_handler); +} + diff --git a/arch/arm/include/asm/cacheflush.h b/arch/arm/include/asm/cacheflush.h index e4448e16046d..c6e2ed906a7b 100644 --- a/arch/arm/include/asm/cacheflush.h +++ b/arch/arm/include/asm/cacheflush.h @@ -16,6 +16,7 @@ #include <asm/shmparam.h> #include <asm/cachetype.h> #include <asm/outercache.h> +#include <asm/rodata.h> #define CACHE_COLOUR(vaddr) ((vaddr & (SHMLBA - 1)) >> PAGE_SHIFT) diff --git a/arch/arm/include/asm/div64.h b/arch/arm/include/asm/div64.h index fe92ccf1d0b0..2dedab7ae089 100644 --- a/arch/arm/include/asm/div64.h +++ b/arch/arm/include/asm/div64.h @@ -73,6 +73,7 @@ #define do_div(n, base) \ ({ \ unsigned int __r, __b = (base); \ + asm("" : "+r" (__b)); \ if (!__builtin_constant_p(__b) || __b == 0 || \ (__LINUX_ARM_ARCH__ < 4 && (__b & (__b - 1)) != 0)) { \ /* non-constant divisor (or zero): slow path */ \ diff --git a/arch/arm/include/asm/fiq_debugger.h b/arch/arm/include/asm/fiq_debugger.h new file mode 100644 index 000000000000..4d274883ba6a --- /dev/null +++ b/arch/arm/include/asm/fiq_debugger.h @@ -0,0 +1,64 @@ +/* + * arch/arm/include/asm/fiq_debugger.h + * + * Copyright (C) 2010 Google, Inc. + * Author: Colin Cross <ccross@android.com> + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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. + * + */ + +#ifndef _ARCH_ARM_MACH_TEGRA_FIQ_DEBUGGER_H_ +#define _ARCH_ARM_MACH_TEGRA_FIQ_DEBUGGER_H_ + +#include <linux/serial_core.h> + +#define FIQ_DEBUGGER_NO_CHAR NO_POLL_CHAR +#define FIQ_DEBUGGER_BREAK 0x00ff0100 + +#define FIQ_DEBUGGER_FIQ_IRQ_NAME "fiq" +#define FIQ_DEBUGGER_SIGNAL_IRQ_NAME "signal" +#define FIQ_DEBUGGER_WAKEUP_IRQ_NAME "wakeup" + +/** + * struct fiq_debugger_pdata - fiq debugger platform data + * @uart_resume: used to restore uart state right before enabling + * the fiq. + * @uart_enable: Do the work necessary to communicate with the uart + * hw (enable clocks, etc.). This must be ref-counted. + * @uart_disable: Do the work necessary to disable the uart hw + * (disable clocks, etc.). This must be ref-counted. + * @uart_dev_suspend: called during PM suspend, generally not needed + * for real fiq mode debugger. + * @uart_dev_resume: called during PM resume, generally not needed + * for real fiq mode debugger. + */ +struct fiq_debugger_pdata { + int (*uart_init)(struct platform_device *pdev); + void (*uart_free)(struct platform_device *pdev); + int (*uart_resume)(struct platform_device *pdev); + int (*uart_getc)(struct platform_device *pdev); + void (*uart_putc)(struct platform_device *pdev, unsigned int c); + void (*uart_flush)(struct platform_device *pdev); + void (*uart_enable)(struct platform_device *pdev); + void (*uart_disable)(struct platform_device *pdev); + + int (*uart_dev_suspend)(struct platform_device *pdev); + int (*uart_dev_resume)(struct platform_device *pdev); + + void (*fiq_enable)(struct platform_device *pdev, unsigned int fiq, + bool enable); + void (*fiq_ack)(struct platform_device *pdev, unsigned int fiq); + + void (*force_irq)(struct platform_device *pdev, unsigned int irq); + void (*force_irq_ack)(struct platform_device *pdev, unsigned int irq); +}; + +#endif diff --git a/arch/arm/include/asm/fiq_glue.h b/arch/arm/include/asm/fiq_glue.h new file mode 100644 index 000000000000..d54c29db97a8 --- /dev/null +++ b/arch/arm/include/asm/fiq_glue.h @@ -0,0 +1,30 @@ +/* + * Copyright (C) 2010 Google, Inc. + * + * This software is licensed under the terms of the GNU General Public + * License version 2, as published by the Free Software Foundation, and + * may be copied, distributed, and modified under those terms. + * + * 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. + */ + +#ifndef __ASM_FIQ_GLUE_H +#define __ASM_FIQ_GLUE_H + +struct fiq_glue_handler { + void (*fiq)(struct fiq_glue_handler *h, void *regs, void *svc_sp); + void (*resume)(struct fiq_glue_handler *h); +}; + +int fiq_glue_register_handler(struct fiq_glue_handler *handler); + +#ifdef CONFIG_FIQ_GLUE +void fiq_glue_resume(void); +#else +static inline void fiq_glue_resume(void) {} +#endif + +#endif diff --git a/arch/arm/include/asm/hardirq.h b/arch/arm/include/asm/hardirq.h index 436e60b2cf7a..2740c2a2df63 100644 --- a/arch/arm/include/asm/hardirq.h +++ b/arch/arm/include/asm/hardirq.h @@ -5,7 +5,7 @@ #include <linux/threads.h> #include <asm/irq.h> -#define NR_IPI 5 +#define NR_IPI 6 typedef struct { unsigned int __softirq_pending; diff --git a/arch/arm/include/asm/hardware/cache-l2x0.h b/arch/arm/include/asm/hardware/cache-l2x0.h index c4c87bc12231..bd2c6a530a58 100644 --- a/arch/arm/include/asm/hardware/cache-l2x0.h +++ b/arch/arm/include/asm/hardware/cache-l2x0.h @@ -66,6 +66,7 @@ #define L2X0_STNDBY_MODE_EN (1 << 0) /* Registers shifts and masks */ +#define L2X0_CACHE_ID_REV_MASK (0x3f) #define L2X0_CACHE_ID_PART_MASK (0xf << 6) #define L2X0_CACHE_ID_PART_L210 (1 << 6) #define L2X0_CACHE_ID_PART_L310 (3 << 6) @@ -102,6 +103,8 @@ #define L2X0_ADDR_FILTER_EN 1 +#define REV_PL310_R2P0 4 + #ifndef __ASSEMBLY__ extern void __init l2x0_init(void __iomem *base, u32 aux_val, u32 aux_mask); #if defined(CONFIG_CACHE_L2X0) && defined(CONFIG_OF) diff --git a/arch/arm/include/asm/hardware/coresight.h b/arch/arm/include/asm/hardware/coresight.h index 7ecd793b8f5a..dcf74d715f2b 100644 --- a/arch/arm/include/asm/hardware/coresight.h +++ b/arch/arm/include/asm/hardware/coresight.h @@ -17,15 +17,23 @@ #define TRACER_ACCESSED_BIT 0 #define TRACER_RUNNING_BIT 1 #define TRACER_CYCLE_ACC_BIT 2 +#define TRACER_TRACE_DATA_BIT 3 +#define TRACER_TIMESTAMP_BIT 4 +#define TRACER_BRANCHOUTPUT_BIT 5 +#define TRACER_RETURN_STACK_BIT 6 #define TRACER_ACCESSED BIT(TRACER_ACCESSED_BIT) #define TRACER_RUNNING BIT(TRACER_RUNNING_BIT) #define TRACER_CYCLE_ACC BIT(TRACER_CYCLE_ACC_BIT) +#define TRACER_TRACE_DATA BIT(TRACER_TRACE_DATA_BIT) +#define TRACER_TIMESTAMP BIT(TRACER_TIMESTAMP_BIT) +#define TRACER_BRANCHOUTPUT BIT(TRACER_BRANCHOUTPUT_BIT) +#define TRACER_RETURN_STACK BIT(TRACER_RETURN_STACK_BIT) #define TRACER_TIMEOUT 10000 -#define etm_writel(t, v, x) \ - (__raw_writel((v), (t)->etm_regs + (x))) -#define etm_readl(t, x) (__raw_readl((t)->etm_regs + (x))) +#define etm_writel(t, id, v, x) \ + (__raw_writel((v), (t)->etm_regs[(id)] + (x))) +#define etm_readl(t, id, x) (__raw_readl((t)->etm_regs[(id)] + (x))) /* CoreSight Management Registers */ #define CSMR_LOCKACCESS 0xfb0 @@ -43,7 +51,7 @@ #define ETMCTRL_POWERDOWN 1 #define ETMCTRL_PROGRAM (1 << 10) #define ETMCTRL_PORTSEL (1 << 11) -#define ETMCTRL_DO_CONTEXTID (3 << 14) +#define ETMCTRL_CONTEXTIDSIZE(x) (((x) & 3) << 14) #define ETMCTRL_PORTMASK1 (7 << 4) #define ETMCTRL_PORTMASK2 (1 << 21) #define ETMCTRL_PORTMASK (ETMCTRL_PORTMASK1 | ETMCTRL_PORTMASK2) @@ -55,9 +63,12 @@ #define ETMCTRL_DATA_DO_BOTH (ETMCTRL_DATA_DO_DATA | ETMCTRL_DATA_DO_ADDR) #define ETMCTRL_BRANCH_OUTPUT (1 << 8) #define ETMCTRL_CYCLEACCURATE (1 << 12) +#define ETMCTRL_TIMESTAMP_EN (1 << 28) +#define ETMCTRL_RETURN_STACK_EN (1 << 29) /* ETM configuration code register */ #define ETMR_CONFCODE (0x04) +#define ETMCCR_ETMIDR_PRESENT BIT(31) /* ETM trace start/stop resource control register */ #define ETMR_TRACESSCTRL (0x18) @@ -113,10 +124,25 @@ #define ETMR_TRACEENCTRL 0x24 #define ETMTE_INCLEXCL BIT(24) #define ETMR_TRACEENEVT 0x20 -#define ETMCTRL_OPTS (ETMCTRL_DO_CPRT | \ - ETMCTRL_DATA_DO_ADDR | \ - ETMCTRL_BRANCH_OUTPUT | \ - ETMCTRL_DO_CONTEXTID) + +#define ETMR_VIEWDATAEVT 0x30 +#define ETMR_VIEWDATACTRL1 0x34 +#define ETMR_VIEWDATACTRL2 0x38 +#define ETMR_VIEWDATACTRL3 0x3c +#define ETMVDC3_EXCLONLY BIT(16) + +#define ETMCTRL_OPTS (ETMCTRL_DO_CPRT) + +#define ETMR_ID 0x1e4 +#define ETMIDR_VERSION(x) (((x) >> 4) & 0xff) +#define ETMIDR_VERSION_3_1 0x21 +#define ETMIDR_VERSION_PFT_1_0 0x30 + +#define ETMR_CCE 0x1e8 +#define ETMCCER_RETURN_STACK_IMPLEMENTED BIT(23) +#define ETMCCER_TIMESTAMPING_IMPLEMENTED BIT(22) + +#define ETMR_TRACEIDR 0x200 /* ETM management registers, "ETM Architecture", 3.5.24 */ #define ETMMR_OSLAR 0x300 @@ -140,14 +166,16 @@ #define ETBFF_TRIGIN BIT(8) #define ETBFF_TRIGEVT BIT(9) #define ETBFF_TRIGFL BIT(10) +#define ETBFF_STOPFL BIT(12) #define etb_writel(t, v, x) \ (__raw_writel((v), (t)->etb_regs + (x))) #define etb_readl(t, x) (__raw_readl((t)->etb_regs + (x))) -#define etm_lock(t) do { etm_writel((t), 0, CSMR_LOCKACCESS); } while (0) -#define etm_unlock(t) \ - do { etm_writel((t), UNLOCK_MAGIC, CSMR_LOCKACCESS); } while (0) +#define etm_lock(t, id) \ + do { etm_writel((t), (id), 0, CSMR_LOCKACCESS); } while (0) +#define etm_unlock(t, id) \ + do { etm_writel((t), (id), UNLOCK_MAGIC, CSMR_LOCKACCESS); } while (0) #define etb_lock(t) do { etb_writel((t), 0, CSMR_LOCKACCESS); } while (0) #define etb_unlock(t) \ diff --git a/arch/arm/include/asm/irq.h b/arch/arm/include/asm/irq.h index 35c21c375d81..3e0857a6248e 100644 --- a/arch/arm/include/asm/irq.h +++ b/arch/arm/include/asm/irq.h @@ -30,6 +30,9 @@ extern void asm_do_IRQ(unsigned int, struct pt_regs *); void handle_IRQ(unsigned int, struct pt_regs *); void init_IRQ(void); +void arch_trigger_all_cpu_backtrace(void); +#define arch_trigger_all_cpu_backtrace arch_trigger_all_cpu_backtrace + #endif #endif diff --git a/arch/arm/include/asm/mach/mmc.h b/arch/arm/include/asm/mach/mmc.h new file mode 100644 index 000000000000..bca864ac945f --- /dev/null +++ b/arch/arm/include/asm/mach/mmc.h @@ -0,0 +1,28 @@ +/* + * arch/arm/include/asm/mach/mmc.h + */ +#ifndef ASMARM_MACH_MMC_H +#define ASMARM_MACH_MMC_H + +#include <linux/mmc/host.h> +#include <linux/mmc/card.h> +#include <linux/mmc/sdio_func.h> + +struct embedded_sdio_data { + struct sdio_cis cis; + struct sdio_cccr cccr; + struct sdio_embedded_func *funcs; + int num_funcs; +}; + +struct mmc_platform_data { + unsigned int ocr_mask; /* available voltages */ + int built_in; /* built-in device flag */ + int card_present; /* card detect state */ + u32 (*translate_vdd)(struct device *, unsigned int); + unsigned int (*status)(struct device *); + struct embedded_sdio_data *embedded_sdio; + int (*register_status_notify)(void (*callback)(int card_present, void *dev_id), void *dev_id); +}; + +#endif diff --git a/arch/arm/include/asm/perf_event.h b/arch/arm/include/asm/perf_event.h index e074948d8143..00416edecead 100644 --- a/arch/arm/include/asm/perf_event.h +++ b/arch/arm/include/asm/perf_event.h @@ -12,6 +12,18 @@ #ifndef __ARM_PERF_EVENT_H__ #define __ARM_PERF_EVENT_H__ -/* Nothing to see here... */ +/* + * The ARMv7 CPU PMU supports up to 32 event counters. + */ +#define ARMPMU_MAX_HWEVENTS 32 + +#define HW_OP_UNSUPPORTED 0xFFFF +#define C(_x) PERF_COUNT_HW_CACHE_##_x +#define CACHE_OP_UNSUPPORTED 0xFFFF + +struct pt_regs; +extern unsigned long perf_instruction_pointer(struct pt_regs *regs); +extern unsigned long perf_misc_flags(struct pt_regs *regs); +#define perf_misc_flags(regs) perf_misc_flags(regs) #endif /* __ARM_PERF_EVENT_H__ */ diff --git a/arch/arm/include/asm/pmu.h b/arch/arm/include/asm/pmu.h index 4432305f4a2a..0cd7824ca762 100644 --- a/arch/arm/include/asm/pmu.h +++ b/arch/arm/include/asm/pmu.h @@ -16,69 +16,30 @@ #include <linux/perf_event.h> /* - * Types of PMUs that can be accessed directly and require mutual - * exclusion between profiling tools. - */ -enum arm_pmu_type { - ARM_PMU_DEVICE_CPU = 0, - ARM_NUM_PMU_DEVICES, -}; - -/* * struct arm_pmu_platdata - ARM PMU platform data * * @handle_irq: an optional handler which will be called from the * interrupt and passed the address of the low level handler, * and can be used to implement any platform specific handling * before or after calling it. - * @enable_irq: an optional handler which will be called after - * request_irq and be used to handle some platform specific - * irq enablement - * @disable_irq: an optional handler which will be called before - * free_irq and be used to handle some platform specific - * irq disablement + * @runtime_resume: an optional handler which will be called by the + * runtime PM framework following a call to pm_runtime_get(). + * Note that if pm_runtime_get() is called more than once in + * succession this handler will only be called once. + * @runtime_suspend: an optional handler which will be called by the + * runtime PM framework following a call to pm_runtime_put(). + * Note that if pm_runtime_get() is called more than once in + * succession this handler will only be called following the + * final call to pm_runtime_put() that actually disables the + * hardware. */ struct arm_pmu_platdata { irqreturn_t (*handle_irq)(int irq, void *dev, irq_handler_t pmu_handler); - void (*enable_irq)(int irq); - void (*disable_irq)(int irq); + int (*runtime_resume)(struct device *dev); + int (*runtime_suspend)(struct device *dev); }; -#ifdef CONFIG_CPU_HAS_PMU - -/** - * reserve_pmu() - reserve the hardware performance counters - * - * Reserve the hardware performance counters in the system for exclusive use. - * Returns 0 on success or -EBUSY if the lock is already held. - */ -extern int -reserve_pmu(enum arm_pmu_type type); - -/** - * release_pmu() - Relinquish control of the performance counters - * - * Release the performance counters and allow someone else to use them. - */ -extern void -release_pmu(enum arm_pmu_type type); - -#else /* CONFIG_CPU_HAS_PMU */ - -#include <linux/err.h> - -static inline int -reserve_pmu(enum arm_pmu_type type) -{ - return -ENODEV; -} - -static inline void -release_pmu(enum arm_pmu_type type) { } - -#endif /* CONFIG_CPU_HAS_PMU */ - #ifdef CONFIG_HW_PERF_EVENTS /* The events for a given PMU register set. */ @@ -101,24 +62,37 @@ struct pmu_hw_events { raw_spinlock_t pmu_lock; }; +struct cpupmu_regs { + u32 pmc; + u32 pmcntenset; + u32 pmuseren; + u32 pmintenset; + u32 pmxevttype[8]; + u32 pmxevtcnt[8]; +}; + struct arm_pmu { struct pmu pmu; - enum arm_pmu_type type; cpumask_t active_irqs; + cpumask_t valid_cpus; char *name; irqreturn_t (*handle_irq)(int irq_num, void *dev); - void (*enable)(struct hw_perf_event *evt, int idx); - void (*disable)(struct hw_perf_event *evt, int idx); + void (*enable)(struct perf_event *event); + void (*disable)(struct perf_event *event); int (*get_event_idx)(struct pmu_hw_events *hw_events, - struct hw_perf_event *hwc); + struct perf_event *event); int (*set_event_filter)(struct hw_perf_event *evt, struct perf_event_attr *attr); - u32 (*read_counter)(int idx); - void (*write_counter)(int idx, u32 val); - void (*start)(void); - void (*stop)(void); + u32 (*read_counter)(struct perf_event *event); + void (*write_counter)(struct perf_event *event, u32 val); + void (*start)(struct arm_pmu *); + void (*stop)(struct arm_pmu *); void (*reset)(void *); + int (*request_irq)(struct arm_pmu *, irq_handler_t handler); + void (*free_irq)(struct arm_pmu *); int (*map_event)(struct perf_event *event); + void (*save_regs)(struct arm_pmu *, struct cpupmu_regs *); + void (*restore_regs)(struct arm_pmu *, struct cpupmu_regs *); int num_events; atomic_t active_events; struct mutex reserve_mutex; @@ -129,15 +103,20 @@ struct arm_pmu { #define to_arm_pmu(p) (container_of(p, struct arm_pmu, pmu)) -int __init armpmu_register(struct arm_pmu *armpmu, char *name, int type); +extern const struct dev_pm_ops armpmu_dev_pm_ops; + +int armpmu_register(struct arm_pmu *armpmu, int type); + +u64 armpmu_event_update(struct perf_event *event); -u64 armpmu_event_update(struct perf_event *event, - struct hw_perf_event *hwc, - int idx); +int armpmu_event_set_period(struct perf_event *event); -int armpmu_event_set_period(struct perf_event *event, - struct hw_perf_event *hwc, - int idx); +int armpmu_map_event(struct perf_event *event, + const unsigned (*event_map)[PERF_COUNT_HW_MAX], + const unsigned (*cache_map)[PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX], + u32 raw_event_mask); #endif /* CONFIG_HW_PERF_EVENTS */ diff --git a/arch/arm/include/asm/rodata.h b/arch/arm/include/asm/rodata.h new file mode 100644 index 000000000000..8c8add87bbc5 --- /dev/null +++ b/arch/arm/include/asm/rodata.h @@ -0,0 +1,32 @@ +/* + * arch/arm/include/asm/rodata.h + * + * Copyright (C) 2011 Google, Inc. + * + * Author: Colin Cross <ccross@android.com> + * + * 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. + */ +#ifndef _ASMARM_RODATA_H +#define _ASMARM_RODATA_H + +#ifndef __ASSEMBLY__ + +#ifdef CONFIG_DEBUG_RODATA + +int set_memory_rw(unsigned long virt, int numpages); +int set_memory_ro(unsigned long virt, int numpages); + +void mark_rodata_ro(void); +void set_kernel_text_rw(void); +void set_kernel_text_ro(void); +#else +static inline void set_kernel_text_rw(void) { } +static inline void set_kernel_text_ro(void) { } +#endif + +#endif + +#endif diff --git a/arch/arm/include/asm/smp.h b/arch/arm/include/asm/smp.h index ae29293270a3..7f74b59f02a4 100644 --- a/arch/arm/include/asm/smp.h +++ b/arch/arm/include/asm/smp.h @@ -93,4 +93,6 @@ extern void platform_cpu_enable(unsigned int cpu); extern void arch_send_call_function_single_ipi(int cpu); extern void arch_send_call_function_ipi_mask(const struct cpumask *mask); +extern void smp_send_all_cpu_backtrace(void); + #endif /* ifndef __ASM_ARM_SMP_H */ diff --git a/arch/arm/include/asm/topology.h b/arch/arm/include/asm/topology.h index 58b8b84adcd2..167597d40e26 100644 --- a/arch/arm/include/asm/topology.h +++ b/arch/arm/include/asm/topology.h @@ -26,11 +26,46 @@ extern struct cputopo_arm cpu_topology[NR_CPUS]; void init_cpu_topology(void); void store_cpu_topology(unsigned int cpuid); const struct cpumask *cpu_coregroup_mask(int cpu); +int cluster_to_logical_mask(unsigned int socket_id, cpumask_t *cluster_mask); + +#ifdef CONFIG_DISABLE_CPU_SCHED_DOMAIN_BALANCE +/* Common values for CPUs */ +#ifndef SD_CPU_INIT +#define SD_CPU_INIT (struct sched_domain) { \ + .min_interval = 1, \ + .max_interval = 4, \ + .busy_factor = 64, \ + .imbalance_pct = 125, \ + .cache_nice_tries = 1, \ + .busy_idx = 2, \ + .idle_idx = 1, \ + .newidle_idx = 0, \ + .wake_idx = 0, \ + .forkexec_idx = 0, \ + \ + .flags = 0*SD_LOAD_BALANCE \ + | 1*SD_BALANCE_NEWIDLE \ + | 1*SD_BALANCE_EXEC \ + | 1*SD_BALANCE_FORK \ + | 0*SD_BALANCE_WAKE \ + | 1*SD_WAKE_AFFINE \ + | 0*SD_PREFER_LOCAL \ + | 0*SD_SHARE_CPUPOWER \ + | 0*SD_SHARE_PKG_RESOURCES \ + | 0*SD_SERIALIZE \ + , \ + .last_balance = jiffies, \ + .balance_interval = 1, \ +} +#endif +#endif /* CONFIG_DISABLE_CPU_SCHED_DOMAIN_BALANCE */ #else static inline void init_cpu_topology(void) { } static inline void store_cpu_topology(unsigned int cpuid) { } +static inline int cluster_to_logical_mask(unsigned int socket_id, + cpumask_t *cluster_mask) { return -EINVAL; } #endif diff --git a/arch/arm/kernel/Makefile b/arch/arm/kernel/Makefile index 7ad2d5cf7008..1c4321430737 100644 --- a/arch/arm/kernel/Makefile +++ b/arch/arm/kernel/Makefile @@ -69,8 +69,7 @@ obj-$(CONFIG_CPU_XSC3) += xscale-cp0.o obj-$(CONFIG_CPU_MOHAWK) += xscale-cp0.o obj-$(CONFIG_CPU_PJ4) += pj4-cp0.o obj-$(CONFIG_IWMMXT) += iwmmxt.o -obj-$(CONFIG_CPU_HAS_PMU) += pmu.o -obj-$(CONFIG_HW_PERF_EVENTS) += perf_event.o +obj-$(CONFIG_HW_PERF_EVENTS) += perf_event.o perf_event_cpu.o AFLAGS_iwmmxt.o := -Wa,-mcpu=iwmmxt obj-$(CONFIG_ARM_CPU_TOPOLOGY) += topology.o diff --git a/arch/arm/kernel/etm.c b/arch/arm/kernel/etm.c index 36d20bd50120..c5fb6c9f35e7 100644 --- a/arch/arm/kernel/etm.c +++ b/arch/arm/kernel/etm.c @@ -15,6 +15,7 @@ #include <linux/init.h> #include <linux/types.h> #include <linux/io.h> +#include <linux/slab.h> #include <linux/sysrq.h> #include <linux/device.h> #include <linux/clk.h> @@ -37,26 +38,37 @@ MODULE_AUTHOR("Alexander Shishkin"); struct tracectx { unsigned int etb_bufsz; void __iomem *etb_regs; - void __iomem *etm_regs; + void __iomem **etm_regs; + int etm_regs_count; unsigned long flags; int ncmppairs; int etm_portsz; + int etm_contextid_size; + u32 etb_fc; + unsigned long range_start; + unsigned long range_end; + unsigned long data_range_start; + unsigned long data_range_end; + bool dump_initial_etb; struct device *dev; struct clk *emu_clk; struct mutex mutex; }; -static struct tracectx tracer; +static struct tracectx tracer = { + .range_start = (unsigned long)_stext, + .range_end = (unsigned long)_etext, +}; static inline bool trace_isrunning(struct tracectx *t) { return !!(t->flags & TRACER_RUNNING); } -static int etm_setup_address_range(struct tracectx *t, int n, +static int etm_setup_address_range(struct tracectx *t, int id, int n, unsigned long start, unsigned long end, int exclude, int data) { - u32 flags = ETMAAT_ARM | ETMAAT_IGNCONTEXTID | ETMAAT_NSONLY | \ + u32 flags = ETMAAT_ARM | ETMAAT_IGNCONTEXTID | ETMAAT_IGNSECURITY | ETMAAT_NOVALCMP; if (n < 1 || n > t->ncmppairs) @@ -72,95 +84,185 @@ static int etm_setup_address_range(struct tracectx *t, int n, flags |= ETMAAT_IEXEC; /* first comparator for the range */ - etm_writel(t, flags, ETMR_COMP_ACC_TYPE(n * 2)); - etm_writel(t, start, ETMR_COMP_VAL(n * 2)); + etm_writel(t, id, flags, ETMR_COMP_ACC_TYPE(n * 2)); + etm_writel(t, id, start, ETMR_COMP_VAL(n * 2)); /* second comparator is right next to it */ - etm_writel(t, flags, ETMR_COMP_ACC_TYPE(n * 2 + 1)); - etm_writel(t, end, ETMR_COMP_VAL(n * 2 + 1)); - - flags = exclude ? ETMTE_INCLEXCL : 0; - etm_writel(t, flags | (1 << n), ETMR_TRACEENCTRL); + etm_writel(t, id, flags, ETMR_COMP_ACC_TYPE(n * 2 + 1)); + etm_writel(t, id, end, ETMR_COMP_VAL(n * 2 + 1)); + + if (data) { + flags = exclude ? ETMVDC3_EXCLONLY : 0; + if (exclude) + n += 8; + etm_writel(t, id, flags | BIT(n), ETMR_VIEWDATACTRL3); + } else { + flags = exclude ? ETMTE_INCLEXCL : 0; + etm_writel(t, id, flags | (1 << n), ETMR_TRACEENCTRL); + } return 0; } -static int trace_start(struct tracectx *t) +static int trace_start_etm(struct tracectx *t, int id) { u32 v; unsigned long timeout = TRACER_TIMEOUT; - etb_unlock(t); - - etb_writel(t, 0, ETBR_FORMATTERCTRL); - etb_writel(t, 1, ETBR_CTRL); - - etb_lock(t); - - /* configure etm */ v = ETMCTRL_OPTS | ETMCTRL_PROGRAM | ETMCTRL_PORTSIZE(t->etm_portsz); + v |= ETMCTRL_CONTEXTIDSIZE(t->etm_contextid_size); if (t->flags & TRACER_CYCLE_ACC) v |= ETMCTRL_CYCLEACCURATE; - etm_unlock(t); + if (t->flags & TRACER_BRANCHOUTPUT) + v |= ETMCTRL_BRANCH_OUTPUT; + + if (t->flags & TRACER_TRACE_DATA) + v |= ETMCTRL_DATA_DO_ADDR; + + if (t->flags & TRACER_TIMESTAMP) + v |= ETMCTRL_TIMESTAMP_EN; + + if (t->flags & TRACER_RETURN_STACK) + v |= ETMCTRL_RETURN_STACK_EN; - etm_writel(t, v, ETMR_CTRL); + etm_unlock(t, id); - while (!(etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout) + etm_writel(t, id, v, ETMR_CTRL); + + while (!(etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout) ; if (!timeout) { dev_dbg(t->dev, "Waiting for progbit to assert timed out\n"); - etm_lock(t); + etm_lock(t, id); return -EFAULT; } - etm_setup_address_range(t, 1, (unsigned long)_stext, - (unsigned long)_etext, 0, 0); - etm_writel(t, 0, ETMR_TRACEENCTRL2); - etm_writel(t, 0, ETMR_TRACESSCTRL); - etm_writel(t, 0x6f, ETMR_TRACEENEVT); + if (t->range_start || t->range_end) + etm_setup_address_range(t, id, 1, + t->range_start, t->range_end, 0, 0); + else + etm_writel(t, id, ETMTE_INCLEXCL, ETMR_TRACEENCTRL); + + etm_writel(t, id, 0, ETMR_TRACEENCTRL2); + etm_writel(t, id, 0, ETMR_TRACESSCTRL); + etm_writel(t, id, 0x6f, ETMR_TRACEENEVT); + + etm_writel(t, id, 0, ETMR_VIEWDATACTRL1); + etm_writel(t, id, 0, ETMR_VIEWDATACTRL2); + + if (t->data_range_start || t->data_range_end) + etm_setup_address_range(t, id, 2, t->data_range_start, + t->data_range_end, 0, 1); + else + etm_writel(t, id, ETMVDC3_EXCLONLY, ETMR_VIEWDATACTRL3); + + etm_writel(t, id, 0x6f, ETMR_VIEWDATAEVT); v &= ~ETMCTRL_PROGRAM; v |= ETMCTRL_PORTSEL; - etm_writel(t, v, ETMR_CTRL); + etm_writel(t, id, v, ETMR_CTRL); timeout = TRACER_TIMEOUT; - while (etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM && --timeout) + while (etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM && --timeout) ; if (!timeout) { dev_dbg(t->dev, "Waiting for progbit to deassert timed out\n"); - etm_lock(t); + etm_lock(t, id); return -EFAULT; } - etm_lock(t); + etm_lock(t, id); + return 0; +} + +static int trace_start(struct tracectx *t) +{ + int ret; + int id; + u32 etb_fc = t->etb_fc; + + etb_unlock(t); + + t->dump_initial_etb = false; + etb_writel(t, 0, ETBR_WRITEADDR); + etb_writel(t, etb_fc, ETBR_FORMATTERCTRL); + etb_writel(t, 1, ETBR_CTRL); + + etb_lock(t); + + /* configure etm(s) */ + for (id = 0; id < t->etm_regs_count; id++) { + ret = trace_start_etm(t, id); + if (ret) + return ret; + } t->flags |= TRACER_RUNNING; return 0; } -static int trace_stop(struct tracectx *t) +static int trace_stop_etm(struct tracectx *t, int id) { unsigned long timeout = TRACER_TIMEOUT; - etm_unlock(t); + etm_unlock(t, id); - etm_writel(t, 0x440, ETMR_CTRL); - while (!(etm_readl(t, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout) + etm_writel(t, id, 0x440, ETMR_CTRL); + while (!(etm_readl(t, id, ETMR_CTRL) & ETMCTRL_PROGRAM) && --timeout) ; if (!timeout) { - dev_dbg(t->dev, "Waiting for progbit to assert timed out\n"); - etm_lock(t); + dev_err(t->dev, + "etm%d: Waiting for progbit to assert timed out\n", + id); + etm_lock(t, id); return -EFAULT; } - etm_lock(t); + etm_lock(t, id); + return 0; +} + +static int trace_power_down_etm(struct tracectx *t, int id) +{ + unsigned long timeout = TRACER_TIMEOUT; + etm_unlock(t, id); + while (!(etm_readl(t, id, ETMR_STATUS) & ETMST_PROGBIT) && --timeout) + ; + if (!timeout) { + dev_err(t->dev, "etm%d: Waiting for status progbit to assert timed out\n", + id); + etm_lock(t, id); + return -EFAULT; + } + + etm_writel(t, id, 0x441, ETMR_CTRL); + + etm_lock(t, id); + return 0; +} + +static int trace_stop(struct tracectx *t) +{ + int id; + unsigned long timeout = TRACER_TIMEOUT; + u32 etb_fc = t->etb_fc; + + for (id = 0; id < t->etm_regs_count; id++) + trace_stop_etm(t, id); + + for (id = 0; id < t->etm_regs_count; id++) + trace_power_down_etm(t, id); etb_unlock(t); - etb_writel(t, ETBFF_MANUAL_FLUSH, ETBR_FORMATTERCTRL); + if (etb_fc) { + etb_fc |= ETBFF_STOPFL; + etb_writel(t, t->etb_fc, ETBR_FORMATTERCTRL); + } + etb_writel(t, etb_fc | ETBFF_MANUAL_FLUSH, ETBR_FORMATTERCTRL); timeout = TRACER_TIMEOUT; while (etb_readl(t, ETBR_FORMATTERCTRL) & @@ -185,24 +287,15 @@ static int trace_stop(struct tracectx *t) static int etb_getdatalen(struct tracectx *t) { u32 v; - int rp, wp; + int wp; v = etb_readl(t, ETBR_STATUS); if (v & 1) return t->etb_bufsz; - rp = etb_readl(t, ETBR_READADDR); wp = etb_readl(t, ETBR_WRITEADDR); - - if (rp > wp) { - etb_writel(t, 0, ETBR_READADDR); - etb_writel(t, 0, ETBR_WRITEADDR); - - return 0; - } - - return wp - rp; + return wp; } /* sysrq+v will always stop the running trace and leave it at that */ @@ -235,21 +328,18 @@ static void etm_dump(void) printk("%08x", cpu_to_be32(etb_readl(t, ETBR_READMEM))); printk(KERN_INFO "\n--- ETB buffer end ---\n"); - /* deassert the overflow bit */ - etb_writel(t, 1, ETBR_CTRL); - etb_writel(t, 0, ETBR_CTRL); - - etb_writel(t, 0, ETBR_TRIGGERCOUNT); - etb_writel(t, 0, ETBR_READADDR); - etb_writel(t, 0, ETBR_WRITEADDR); - etb_lock(t); } static void sysrq_etm_dump(int key) { + if (!mutex_trylock(&tracer.mutex)) { + printk(KERN_INFO "Tracing hardware busy\n"); + return; + } dev_dbg(tracer.dev, "Dumping ETB buffer\n"); etm_dump(); + mutex_unlock(&tracer.mutex); } static struct sysrq_key_op sysrq_etm_op = { @@ -276,6 +366,10 @@ static ssize_t etb_read(struct file *file, char __user *data, struct tracectx *t = file->private_data; u32 first = 0; u32 *buf; + int wpos; + int skip; + long wlength; + loff_t pos = *ppos; mutex_lock(&t->mutex); @@ -287,31 +381,39 @@ static ssize_t etb_read(struct file *file, char __user *data, etb_unlock(t); total = etb_getdatalen(t); + if (total == 0 && t->dump_initial_etb) + total = t->etb_bufsz; if (total == t->etb_bufsz) first = etb_readl(t, ETBR_WRITEADDR); + if (pos > total * 4) { + skip = 0; + wpos = total; + } else { + skip = (int)pos % 4; + wpos = (int)pos / 4; + } + total -= wpos; + first = (first + wpos) % t->etb_bufsz; + etb_writel(t, first, ETBR_READADDR); - length = min(total * 4, (int)len); - buf = vmalloc(length); + wlength = min(total, DIV_ROUND_UP(skip + (int)len, 4)); + length = min(total * 4 - skip, (int)len); + buf = vmalloc(wlength * 4); - dev_dbg(t->dev, "ETB buffer length: %d\n", total); + dev_dbg(t->dev, "ETB read %ld bytes to %lld from %ld words at %d\n", + length, pos, wlength, first); + dev_dbg(t->dev, "ETB buffer length: %d\n", total + wpos); dev_dbg(t->dev, "ETB status reg: %x\n", etb_readl(t, ETBR_STATUS)); - for (i = 0; i < length / 4; i++) + for (i = 0; i < wlength; i++) buf[i] = etb_readl(t, ETBR_READMEM); - /* the only way to deassert overflow bit in ETB status is this */ - etb_writel(t, 1, ETBR_CTRL); - etb_writel(t, 0, ETBR_CTRL); - - etb_writel(t, 0, ETBR_WRITEADDR); - etb_writel(t, 0, ETBR_READADDR); - etb_writel(t, 0, ETBR_TRIGGERCOUNT); - etb_lock(t); - length -= copy_to_user(data, buf, length); + length -= copy_to_user(data, (u8 *)buf + skip, length); vfree(buf); + *ppos = pos + length; out: mutex_unlock(&t->mutex); @@ -348,28 +450,17 @@ static int __devinit etb_probe(struct amba_device *dev, const struct amba_id *id if (ret) goto out; + mutex_lock(&t->mutex); t->etb_regs = ioremap_nocache(dev->res.start, resource_size(&dev->res)); if (!t->etb_regs) { ret = -ENOMEM; goto out_release; } + t->dev = &dev->dev; + t->dump_initial_etb = true; amba_set_drvdata(dev, t); - etb_miscdev.parent = &dev->dev; - - ret = misc_register(&etb_miscdev); - if (ret) - goto out_unmap; - - t->emu_clk = clk_get(&dev->dev, "emu_src_ck"); - if (IS_ERR(t->emu_clk)) { - dev_dbg(&dev->dev, "Failed to obtain emu_src_ck.\n"); - return -EFAULT; - } - - clk_enable(t->emu_clk); - etb_unlock(t); t->etb_bufsz = etb_readl(t, ETBR_DEPTH); dev_dbg(&dev->dev, "Size: %x\n", t->etb_bufsz); @@ -378,6 +469,20 @@ static int __devinit etb_probe(struct amba_device *dev, const struct amba_id *id etb_writel(t, 0, ETBR_CTRL); etb_writel(t, 0x1000, ETBR_FORMATTERCTRL); etb_lock(t); + mutex_unlock(&t->mutex); + + etb_miscdev.parent = &dev->dev; + + ret = misc_register(&etb_miscdev); + if (ret) + goto out_unmap; + + /* Get optional clock. Currently used to select clock source on omap3 */ + t->emu_clk = clk_get(&dev->dev, "emu_src_ck"); + if (IS_ERR(t->emu_clk)) + dev_dbg(&dev->dev, "Failed to obtain emu_src_ck.\n"); + else + clk_enable(t->emu_clk); dev_dbg(&dev->dev, "ETB AMBA driver initialized.\n"); @@ -385,10 +490,13 @@ out: return ret; out_unmap: + mutex_lock(&t->mutex); amba_set_drvdata(dev, NULL); iounmap(t->etb_regs); + t->etb_regs = NULL; out_release: + mutex_unlock(&t->mutex); amba_release_regions(dev); return ret; @@ -403,8 +511,10 @@ static int etb_remove(struct amba_device *dev) iounmap(t->etb_regs); t->etb_regs = NULL; - clk_disable(t->emu_clk); - clk_put(t->emu_clk); + if (!IS_ERR(t->emu_clk)) { + clk_disable(t->emu_clk); + clk_put(t->emu_clk); + } amba_release_regions(dev); @@ -448,7 +558,10 @@ static ssize_t trace_running_store(struct kobject *kobj, return -EINVAL; mutex_lock(&tracer.mutex); - ret = value ? trace_start(&tracer) : trace_stop(&tracer); + if (!tracer.etb_regs) + ret = -ENODEV; + else + ret = value ? trace_start(&tracer) : trace_stop(&tracer); mutex_unlock(&tracer.mutex); return ret ? : n; @@ -463,36 +576,50 @@ static ssize_t trace_info_show(struct kobject *kobj, { u32 etb_wa, etb_ra, etb_st, etb_fc, etm_ctrl, etm_st; int datalen; + int id; + int ret; - etb_unlock(&tracer); - datalen = etb_getdatalen(&tracer); - etb_wa = etb_readl(&tracer, ETBR_WRITEADDR); - etb_ra = etb_readl(&tracer, ETBR_READADDR); - etb_st = etb_readl(&tracer, ETBR_STATUS); - etb_fc = etb_readl(&tracer, ETBR_FORMATTERCTRL); - etb_lock(&tracer); - - etm_unlock(&tracer); - etm_ctrl = etm_readl(&tracer, ETMR_CTRL); - etm_st = etm_readl(&tracer, ETMR_STATUS); - etm_lock(&tracer); + mutex_lock(&tracer.mutex); + if (tracer.etb_regs) { + etb_unlock(&tracer); + datalen = etb_getdatalen(&tracer); + etb_wa = etb_readl(&tracer, ETBR_WRITEADDR); + etb_ra = etb_readl(&tracer, ETBR_READADDR); + etb_st = etb_readl(&tracer, ETBR_STATUS); + etb_fc = etb_readl(&tracer, ETBR_FORMATTERCTRL); + etb_lock(&tracer); + } else { + etb_wa = etb_ra = etb_st = etb_fc = ~0; + datalen = -1; + } - return sprintf(buf, "Trace buffer len: %d\nComparator pairs: %d\n" + ret = sprintf(buf, "Trace buffer len: %d\nComparator pairs: %d\n" "ETBR_WRITEADDR:\t%08x\n" "ETBR_READADDR:\t%08x\n" "ETBR_STATUS:\t%08x\n" - "ETBR_FORMATTERCTRL:\t%08x\n" - "ETMR_CTRL:\t%08x\n" - "ETMR_STATUS:\t%08x\n", + "ETBR_FORMATTERCTRL:\t%08x\n", datalen, tracer.ncmppairs, etb_wa, etb_ra, etb_st, - etb_fc, + etb_fc + ); + + for (id = 0; id < tracer.etm_regs_count; id++) { + etm_unlock(&tracer, id); + etm_ctrl = etm_readl(&tracer, id, ETMR_CTRL); + etm_st = etm_readl(&tracer, id, ETMR_STATUS); + etm_lock(&tracer, id); + ret += sprintf(buf + ret, "ETMR_CTRL:\t%08x\n" + "ETMR_STATUS:\t%08x\n", etm_ctrl, etm_st ); + } + mutex_unlock(&tracer.mutex); + + return ret; } static struct kobj_attribute trace_info_attr = @@ -531,42 +658,260 @@ static ssize_t trace_mode_store(struct kobject *kobj, static struct kobj_attribute trace_mode_attr = __ATTR(trace_mode, 0644, trace_mode_show, trace_mode_store); +static ssize_t trace_contextid_size_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + /* 0: No context id tracing, 1: One byte, 2: Two bytes, 3: Four bytes */ + return sprintf(buf, "%d\n", (1 << tracer.etm_contextid_size) >> 1); +} + +static ssize_t trace_contextid_size_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned int contextid_size; + + if (sscanf(buf, "%u", &contextid_size) != 1) + return -EINVAL; + + if (contextid_size == 3 || contextid_size > 4) + return -EINVAL; + + mutex_lock(&tracer.mutex); + tracer.etm_contextid_size = fls(contextid_size); + mutex_unlock(&tracer.mutex); + + return n; +} + +static struct kobj_attribute trace_contextid_size_attr = + __ATTR(trace_contextid_size, 0644, + trace_contextid_size_show, trace_contextid_size_store); + +static ssize_t trace_branch_output_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + return sprintf(buf, "%d\n", !!(tracer.flags & TRACER_BRANCHOUTPUT)); +} + +static ssize_t trace_branch_output_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned int branch_output; + + if (sscanf(buf, "%u", &branch_output) != 1) + return -EINVAL; + + mutex_lock(&tracer.mutex); + if (branch_output) { + tracer.flags |= TRACER_BRANCHOUTPUT; + /* Branch broadcasting is incompatible with the return stack */ + tracer.flags &= ~TRACER_RETURN_STACK; + } else { + tracer.flags &= ~TRACER_BRANCHOUTPUT; + } + mutex_unlock(&tracer.mutex); + + return n; +} + +static struct kobj_attribute trace_branch_output_attr = + __ATTR(trace_branch_output, 0644, + trace_branch_output_show, trace_branch_output_store); + +static ssize_t trace_return_stack_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + return sprintf(buf, "%d\n", !!(tracer.flags & TRACER_RETURN_STACK)); +} + +static ssize_t trace_return_stack_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned int return_stack; + + if (sscanf(buf, "%u", &return_stack) != 1) + return -EINVAL; + + mutex_lock(&tracer.mutex); + if (return_stack) { + tracer.flags |= TRACER_RETURN_STACK; + /* Return stack is incompatible with branch broadcasting */ + tracer.flags &= ~TRACER_BRANCHOUTPUT; + } else { + tracer.flags &= ~TRACER_RETURN_STACK; + } + mutex_unlock(&tracer.mutex); + + return n; +} + +static struct kobj_attribute trace_return_stack_attr = + __ATTR(trace_return_stack, 0644, + trace_return_stack_show, trace_return_stack_store); + +static ssize_t trace_timestamp_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + return sprintf(buf, "%d\n", !!(tracer.flags & TRACER_TIMESTAMP)); +} + +static ssize_t trace_timestamp_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned int timestamp; + + if (sscanf(buf, "%u", ×tamp) != 1) + return -EINVAL; + + mutex_lock(&tracer.mutex); + if (timestamp) + tracer.flags |= TRACER_TIMESTAMP; + else + tracer.flags &= ~TRACER_TIMESTAMP; + mutex_unlock(&tracer.mutex); + + return n; +} + +static struct kobj_attribute trace_timestamp_attr = + __ATTR(trace_timestamp, 0644, + trace_timestamp_show, trace_timestamp_store); + +static ssize_t trace_range_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + return sprintf(buf, "%08lx %08lx\n", + tracer.range_start, tracer.range_end); +} + +static ssize_t trace_range_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned long range_start, range_end; + + if (sscanf(buf, "%lx %lx", &range_start, &range_end) != 2) + return -EINVAL; + + mutex_lock(&tracer.mutex); + tracer.range_start = range_start; + tracer.range_end = range_end; + mutex_unlock(&tracer.mutex); + + return n; +} + + +static struct kobj_attribute trace_range_attr = + __ATTR(trace_range, 0644, trace_range_show, trace_range_store); + +static ssize_t trace_data_range_show(struct kobject *kobj, + struct kobj_attribute *attr, + char *buf) +{ + unsigned long range_start; + u64 range_end; + mutex_lock(&tracer.mutex); + range_start = tracer.data_range_start; + range_end = tracer.data_range_end; + if (!range_end && (tracer.flags & TRACER_TRACE_DATA)) + range_end = 0x100000000ULL; + mutex_unlock(&tracer.mutex); + return sprintf(buf, "%08lx %08llx\n", range_start, range_end); +} + +static ssize_t trace_data_range_store(struct kobject *kobj, + struct kobj_attribute *attr, + const char *buf, size_t n) +{ + unsigned long range_start; + u64 range_end; + + if (sscanf(buf, "%lx %llx", &range_start, &range_end) != 2) + return -EINVAL; + + mutex_lock(&tracer.mutex); + tracer.data_range_start = range_start; + tracer.data_range_end = (unsigned long)range_end; + if (range_end) + tracer.flags |= TRACER_TRACE_DATA; + else + tracer.flags &= ~TRACER_TRACE_DATA; + mutex_unlock(&tracer.mutex); + + return n; +} + + +static struct kobj_attribute trace_data_range_attr = + __ATTR(trace_data_range, 0644, + trace_data_range_show, trace_data_range_store); + static int __devinit etm_probe(struct amba_device *dev, const struct amba_id *id) { struct tracectx *t = &tracer; int ret = 0; + void __iomem **new_regs; + int new_count; + u32 etmccr; + u32 etmidr; + u32 etmccer = 0; + u8 etm_version = 0; + + mutex_lock(&t->mutex); + new_count = t->etm_regs_count + 1; + new_regs = krealloc(t->etm_regs, + sizeof(t->etm_regs[0]) * new_count, GFP_KERNEL); - if (t->etm_regs) { - dev_dbg(&dev->dev, "ETM already initialized\n"); - ret = -EBUSY; + if (!new_regs) { + dev_dbg(&dev->dev, "Failed to allocate ETM register array\n"); + ret = -ENOMEM; goto out; } + t->etm_regs = new_regs; ret = amba_request_regions(dev, NULL); if (ret) goto out; - t->etm_regs = ioremap_nocache(dev->res.start, resource_size(&dev->res)); - if (!t->etm_regs) { + t->etm_regs[t->etm_regs_count] = + ioremap_nocache(dev->res.start, resource_size(&dev->res)); + if (!t->etm_regs[t->etm_regs_count]) { ret = -ENOMEM; goto out_release; } - amba_set_drvdata(dev, t); + amba_set_drvdata(dev, t->etm_regs[t->etm_regs_count]); - mutex_init(&t->mutex); - t->dev = &dev->dev; - t->flags = TRACER_CYCLE_ACC; + t->flags = TRACER_CYCLE_ACC | TRACER_TRACE_DATA | TRACER_BRANCHOUTPUT; t->etm_portsz = 1; + t->etm_contextid_size = 3; - etm_unlock(t); - (void)etm_readl(t, ETMMR_PDSR); + etm_unlock(t, t->etm_regs_count); + (void)etm_readl(t, t->etm_regs_count, ETMMR_PDSR); /* dummy first read */ - (void)etm_readl(&tracer, ETMMR_OSSRR); - - t->ncmppairs = etm_readl(t, ETMR_CONFCODE) & 0xf; - etm_writel(t, 0x440, ETMR_CTRL); - etm_lock(t); + (void)etm_readl(&tracer, t->etm_regs_count, ETMMR_OSSRR); + + etmccr = etm_readl(t, t->etm_regs_count, ETMR_CONFCODE); + t->ncmppairs = etmccr & 0xf; + if (etmccr & ETMCCR_ETMIDR_PRESENT) { + etmidr = etm_readl(t, t->etm_regs_count, ETMR_ID); + etm_version = ETMIDR_VERSION(etmidr); + if (etm_version >= ETMIDR_VERSION_3_1) + etmccer = etm_readl(t, t->etm_regs_count, ETMR_CCE); + } + etm_writel(t, t->etm_regs_count, 0x441, ETMR_CTRL); + etm_writel(t, t->etm_regs_count, new_count, ETMR_TRACEIDR); + etm_lock(t, t->etm_regs_count); ret = sysfs_create_file(&dev->dev.kobj, &trace_running_attr.attr); @@ -582,36 +927,101 @@ static int __devinit etm_probe(struct amba_device *dev, const struct amba_id *id if (ret) dev_dbg(&dev->dev, "Failed to create trace_mode in sysfs\n"); - dev_dbg(t->dev, "ETM AMBA driver initialized.\n"); + ret = sysfs_create_file(&dev->dev.kobj, + &trace_contextid_size_attr.attr); + if (ret) + dev_dbg(&dev->dev, + "Failed to create trace_contextid_size in sysfs\n"); + + ret = sysfs_create_file(&dev->dev.kobj, + &trace_branch_output_attr.attr); + if (ret) + dev_dbg(&dev->dev, + "Failed to create trace_branch_output in sysfs\n"); + + if (etmccer & ETMCCER_RETURN_STACK_IMPLEMENTED) { + ret = sysfs_create_file(&dev->dev.kobj, + &trace_return_stack_attr.attr); + if (ret) + dev_dbg(&dev->dev, + "Failed to create trace_return_stack in sysfs\n"); + } + + if (etmccer & ETMCCER_TIMESTAMPING_IMPLEMENTED) { + ret = sysfs_create_file(&dev->dev.kobj, + &trace_timestamp_attr.attr); + if (ret) + dev_dbg(&dev->dev, + "Failed to create trace_timestamp in sysfs\n"); + } + + ret = sysfs_create_file(&dev->dev.kobj, &trace_range_attr.attr); + if (ret) + dev_dbg(&dev->dev, "Failed to create trace_range in sysfs\n"); + + if (etm_version < ETMIDR_VERSION_PFT_1_0) { + ret = sysfs_create_file(&dev->dev.kobj, + &trace_data_range_attr.attr); + if (ret) + dev_dbg(&dev->dev, + "Failed to create trace_data_range in sysfs\n"); + } else { + tracer.flags &= ~TRACER_TRACE_DATA; + } + + dev_dbg(&dev->dev, "ETM AMBA driver initialized.\n"); + + /* Enable formatter if there are multiple trace sources */ + if (new_count > 1) + t->etb_fc = ETBFF_ENFCONT | ETBFF_ENFTC; + + t->etm_regs_count = new_count; out: + mutex_unlock(&t->mutex); return ret; out_unmap: amba_set_drvdata(dev, NULL); - iounmap(t->etm_regs); + iounmap(t->etm_regs[t->etm_regs_count]); out_release: amba_release_regions(dev); + mutex_unlock(&t->mutex); return ret; } static int etm_remove(struct amba_device *dev) { - struct tracectx *t = amba_get_drvdata(dev); + int i; + struct tracectx *t = &tracer; + void __iomem *etm_regs = amba_get_drvdata(dev); + + sysfs_remove_file(&dev->dev.kobj, &trace_running_attr.attr); + sysfs_remove_file(&dev->dev.kobj, &trace_info_attr.attr); + sysfs_remove_file(&dev->dev.kobj, &trace_mode_attr.attr); + sysfs_remove_file(&dev->dev.kobj, &trace_range_attr.attr); + sysfs_remove_file(&dev->dev.kobj, &trace_data_range_attr.attr); amba_set_drvdata(dev, NULL); - iounmap(t->etm_regs); - t->etm_regs = NULL; + mutex_lock(&t->mutex); + for (i = 0; i < t->etm_regs_count; i++) + if (t->etm_regs[i] == etm_regs) + break; + for (; i < t->etm_regs_count - 1; i++) + t->etm_regs[i] = t->etm_regs[i + 1]; + t->etm_regs_count--; + if (!t->etm_regs_count) { + kfree(t->etm_regs); + t->etm_regs = NULL; + } + mutex_unlock(&t->mutex); + iounmap(etm_regs); amba_release_regions(dev); - sysfs_remove_file(&dev->dev.kobj, &trace_running_attr.attr); - sysfs_remove_file(&dev->dev.kobj, &trace_info_attr.attr); - sysfs_remove_file(&dev->dev.kobj, &trace_mode_attr.attr); - return 0; } @@ -620,6 +1030,10 @@ static struct amba_id etm_ids[] = { .id = 0x0003b921, .mask = 0x0007ffff, }, + { + .id = 0x0003b950, + .mask = 0x0007ffff, + }, { 0, 0 }, }; @@ -637,6 +1051,8 @@ static int __init etm_init(void) { int retval; + mutex_init(&tracer.mutex); + retval = amba_driver_register(&etb_driver); if (retval) { printk(KERN_ERR "Failed to register etb\n"); diff --git a/arch/arm/kernel/ftrace.c b/arch/arm/kernel/ftrace.c index 34e56647dcee..6a740a93f4bb 100644 --- a/arch/arm/kernel/ftrace.c +++ b/arch/arm/kernel/ftrace.c @@ -13,6 +13,7 @@ */ #include <linux/ftrace.h> +#include <linux/module.h> #include <linux/uaccess.h> #include <asm/cacheflush.h> @@ -63,6 +64,20 @@ static unsigned long adjust_address(struct dyn_ftrace *rec, unsigned long addr) } #endif +int ftrace_arch_code_modify_prepare(void) +{ + set_kernel_text_rw(); + set_all_modules_text_rw(); + return 0; +} + +int ftrace_arch_code_modify_post_process(void) +{ + set_all_modules_text_ro(); + set_kernel_text_ro(); + return 0; +} + static unsigned long ftrace_call_replace(unsigned long pc, unsigned long addr) { return arm_gen_branch_link(pc, addr); diff --git a/arch/arm/kernel/hw_breakpoint.c b/arch/arm/kernel/hw_breakpoint.c index 281bf3301241..eed4d0cdd748 100644 --- a/arch/arm/kernel/hw_breakpoint.c +++ b/arch/arm/kernel/hw_breakpoint.c @@ -28,6 +28,7 @@ #include <linux/perf_event.h> #include <linux/hw_breakpoint.h> #include <linux/smp.h> +#include <linux/cpu_pm.h> #include <asm/cacheflush.h> #include <asm/cputype.h> @@ -42,6 +43,11 @@ static DEFINE_PER_CPU(struct perf_event *, bp_on_reg[ARM_MAX_BRP]); /* Watchpoint currently in use for each WRP. */ static DEFINE_PER_CPU(struct perf_event *, wp_on_reg[ARM_MAX_WRP]); +#ifdef CONFIG_CPU_PM +/* Storage for OS Save and Restore. */ +static DEFINE_PER_CPU(u32, cpu_dscr); +#endif + /* Number of BRP/WRP registers on this CPU. */ static int core_num_brps; static int core_num_wrps; @@ -990,6 +996,55 @@ static struct notifier_block __cpuinitdata dbg_reset_nb = { .notifier_call = dbg_reset_notify, }; +#ifdef CONFIG_CPU_PM +static void os_save(int cpu) +{ + /* Set OS Lock. */ + asm volatile("mcr p14, 0, %0, c1, c0, 4" : : "r" (0xC5ACCE55)); + isb(); + + /* Save DSCRext. */ + ARM_DBG_READ(c2, 2, per_cpu(cpu_dscr, cpu)); +} + +static void os_restore(int cpu) +{ + /* Restore DSCRext. */ + ARM_DBG_WRITE(c2, 2, per_cpu(cpu_dscr, cpu)); + + /* Clear OS Lock. */ + asm volatile("mcr p14, 0, %0, c1, c0, 4" : : "r" (0)); + isb(); +} + +static int dbg_cpu_pm_notify(struct notifier_block *self, unsigned long action, + void *v) +{ + int cpu = smp_processor_id(); + + if (action == CPU_PM_ENTER) + os_save(cpu); + else if (action == CPU_PM_EXIT) + os_restore(cpu); + + return NOTIFY_OK; +} + +static struct notifier_block __cpuinitdata dbg_cpu_pm_nb = { + .notifier_call = dbg_cpu_pm_notify, +}; + +static void __init pm_init(void) +{ + if (get_debug_arch() == ARM_DEBUG_ARCH_V7_1) + cpu_pm_register_notifier(&dbg_cpu_pm_nb); +} +#else +static inline void pm_init(void) +{ +} +#endif + static int __init arch_hw_breakpoint_init(void) { u32 dscr; @@ -1048,6 +1103,8 @@ static int __init arch_hw_breakpoint_init(void) /* Register hotplug notifier. */ register_cpu_notifier(&dbg_reset_nb); + + pm_init(); return 0; } arch_initcall(arch_hw_breakpoint_init); diff --git a/arch/arm/kernel/kprobes-test-arm.c b/arch/arm/kernel/kprobes-test-arm.c index 38c1a3b103a0..29d9ce019c2b 100644 --- a/arch/arm/kernel/kprobes-test-arm.c +++ b/arch/arm/kernel/kprobes-test-arm.c @@ -456,6 +456,8 @@ void kprobe_arm_test_cases(void) TEST_UNSUPPORTED(".word 0xe1700090") /* Unallocated space */ #if __LINUX_ARM_ARCH__ >= 6 TEST_UNSUPPORTED("ldrex r2, [sp]") +#endif +#if __LINUX_ARM_ARCH__ >= 7 TEST_UNSUPPORTED("strexd r0, r2, r3, [sp]") TEST_UNSUPPORTED("ldrexd r2, r3, [sp]") TEST_UNSUPPORTED("strexb r0, r2, [sp]") diff --git a/arch/arm/kernel/leds.c b/arch/arm/kernel/leds.c index 1911dae19e4f..2050399ebd08 100644 --- a/arch/arm/kernel/leds.c +++ b/arch/arm/kernel/leds.c @@ -10,6 +10,8 @@ #include <linux/export.h> #include <linux/init.h> #include <linux/device.h> +#include <linux/notifier.h> +#include <linux/cpu.h> #include <linux/syscore_ops.h> #include <linux/string.h> @@ -103,6 +105,25 @@ static struct syscore_ops leds_syscore_ops = { .resume = leds_resume, }; +static int leds_idle_notifier(struct notifier_block *nb, unsigned long val, + void *data) +{ + switch (val) { + case IDLE_START: + leds_event(led_idle_start); + break; + case IDLE_END: + leds_event(led_idle_end); + break; + } + + return 0; +} + +static struct notifier_block leds_idle_nb = { + .notifier_call = leds_idle_notifier, +}; + static int __init leds_init(void) { int ret; @@ -111,8 +132,11 @@ static int __init leds_init(void) ret = device_register(&leds_device); if (ret == 0) ret = device_create_file(&leds_device, &dev_attr_event); - if (ret == 0) + if (ret == 0) { register_syscore_ops(&leds_syscore_ops); + idle_notifier_register(&leds_idle_nb); + } + return ret; } diff --git a/arch/arm/kernel/perf_event.c b/arch/arm/kernel/perf_event.c index ab243b87118d..e90b2a7e8c25 100644 --- a/arch/arm/kernel/perf_event.c +++ b/arch/arm/kernel/perf_event.c @@ -12,68 +12,16 @@ */ #define pr_fmt(fmt) "hw perfevents: " fmt -#include <linux/bitmap.h> -#include <linux/interrupt.h> +#include <linux/cpumask.h> #include <linux/kernel.h> -#include <linux/export.h> -#include <linux/perf_event.h> #include <linux/platform_device.h> -#include <linux/spinlock.h> +#include <linux/pm_runtime.h> #include <linux/uaccess.h> -#include <asm/cputype.h> -#include <asm/irq.h> #include <asm/irq_regs.h> #include <asm/pmu.h> #include <asm/stacktrace.h> -/* - * ARMv6 supports a maximum of 3 events, starting from index 0. If we add - * another platform that supports more, we need to increase this to be the - * largest of all platforms. - * - * ARMv7 supports up to 32 events: - * cycle counter CCNT + 31 events counters CNT0..30. - * Cortex-A8 has 1+4 counters, Cortex-A9 has 1+6 counters. - */ -#define ARMPMU_MAX_HWEVENTS 32 - -static DEFINE_PER_CPU(struct perf_event * [ARMPMU_MAX_HWEVENTS], hw_events); -static DEFINE_PER_CPU(unsigned long [BITS_TO_LONGS(ARMPMU_MAX_HWEVENTS)], used_mask); -static DEFINE_PER_CPU(struct pmu_hw_events, cpu_hw_events); - -#define to_arm_pmu(p) (container_of(p, struct arm_pmu, pmu)) - -/* Set at runtime when we know what CPU type we are. */ -static struct arm_pmu *cpu_pmu; - -const char *perf_pmu_name(void) -{ - if (!cpu_pmu) - return NULL; - - return cpu_pmu->pmu.name; -} -EXPORT_SYMBOL_GPL(perf_pmu_name); - -int perf_num_counters(void) -{ - int max_events = 0; - - if (cpu_pmu != NULL) - max_events = cpu_pmu->num_events; - - return max_events; -} -EXPORT_SYMBOL_GPL(perf_num_counters); - -#define HW_OP_UNSUPPORTED 0xFFFF - -#define C(_x) \ - PERF_COUNT_HW_CACHE_##_x - -#define CACHE_OP_UNSUPPORTED 0xFFFF - static int armpmu_map_cache_event(const unsigned (*cache_map) [PERF_COUNT_HW_CACHE_MAX] @@ -104,7 +52,7 @@ armpmu_map_cache_event(const unsigned (*cache_map) } static int -armpmu_map_event(const unsigned (*event_map)[PERF_COUNT_HW_MAX], u64 config) +armpmu_map_hw_event(const unsigned (*event_map)[PERF_COUNT_HW_MAX], u64 config) { int mapping = (*event_map)[config]; return mapping == HW_OP_UNSUPPORTED ? -ENOENT : mapping; @@ -116,34 +64,36 @@ armpmu_map_raw_event(u32 raw_event_mask, u64 config) return (int)(config & raw_event_mask); } -static int map_cpu_event(struct perf_event *event, - const unsigned (*event_map)[PERF_COUNT_HW_MAX], - const unsigned (*cache_map) - [PERF_COUNT_HW_CACHE_MAX] - [PERF_COUNT_HW_CACHE_OP_MAX] - [PERF_COUNT_HW_CACHE_RESULT_MAX], - u32 raw_event_mask) +int +armpmu_map_event(struct perf_event *event, + const unsigned (*event_map)[PERF_COUNT_HW_MAX], + const unsigned (*cache_map) + [PERF_COUNT_HW_CACHE_MAX] + [PERF_COUNT_HW_CACHE_OP_MAX] + [PERF_COUNT_HW_CACHE_RESULT_MAX], + u32 raw_event_mask) { u64 config = event->attr.config; switch (event->attr.type) { case PERF_TYPE_HARDWARE: - return armpmu_map_event(event_map, config); + return armpmu_map_hw_event(event_map, config); case PERF_TYPE_HW_CACHE: return armpmu_map_cache_event(cache_map, config); case PERF_TYPE_RAW: return armpmu_map_raw_event(raw_event_mask, config); + default: + if (event->attr.type >= PERF_TYPE_MAX) + return armpmu_map_raw_event(raw_event_mask, config); } return -ENOENT; } -int -armpmu_event_set_period(struct perf_event *event, - struct hw_perf_event *hwc, - int idx) +int armpmu_event_set_period(struct perf_event *event) { struct arm_pmu *armpmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; s64 left = local64_read(&hwc->period_left); s64 period = hwc->sample_period; int ret = 0; @@ -167,24 +117,22 @@ armpmu_event_set_period(struct perf_event *event, local64_set(&hwc->prev_count, (u64)-left); - armpmu->write_counter(idx, (u64)(-left) & 0xffffffff); + armpmu->write_counter(event, (u64)(-left) & 0xffffffff); perf_event_update_userpage(event); return ret; } -u64 -armpmu_event_update(struct perf_event *event, - struct hw_perf_event *hwc, - int idx) +u64 armpmu_event_update(struct perf_event *event) { struct arm_pmu *armpmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; u64 delta, prev_raw_count, new_raw_count; again: prev_raw_count = local64_read(&hwc->prev_count); - new_raw_count = armpmu->read_counter(idx); + new_raw_count = armpmu->read_counter(event); if (local64_cmpxchg(&hwc->prev_count, prev_raw_count, new_raw_count) != prev_raw_count) @@ -207,7 +155,7 @@ armpmu_read(struct perf_event *event) if (hwc->idx < 0) return; - armpmu_event_update(event, hwc, hwc->idx); + armpmu_event_update(event); } static void @@ -216,24 +164,26 @@ armpmu_stop(struct perf_event *event, int flags) struct arm_pmu *armpmu = to_arm_pmu(event->pmu); struct hw_perf_event *hwc = &event->hw; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; /* * ARM pmu always has to update the counter, so ignore * PERF_EF_UPDATE, see comments in armpmu_start(). */ if (!(hwc->state & PERF_HES_STOPPED)) { - armpmu->disable(hwc, hwc->idx); - barrier(); /* why? */ - armpmu_event_update(event, hwc, hwc->idx); + armpmu->disable(event); + armpmu_event_update(event); hwc->state |= PERF_HES_STOPPED | PERF_HES_UPTODATE; } } -static void -armpmu_start(struct perf_event *event, int flags) +static void armpmu_start(struct perf_event *event, int flags) { struct arm_pmu *armpmu = to_arm_pmu(event->pmu); struct hw_perf_event *hwc = &event->hw; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; /* * ARM pmu always has to reprogram the period, so ignore * PERF_EF_RELOAD, see the comment below. @@ -249,8 +199,8 @@ armpmu_start(struct perf_event *event, int flags) * get an interrupt too soon or *way* too late if the overflow has * happened since disabling. */ - armpmu_event_set_period(event, hwc, hwc->idx); - armpmu->enable(hwc, hwc->idx); + armpmu_event_set_period(event); + armpmu->enable(event); } static void @@ -261,6 +211,9 @@ armpmu_del(struct perf_event *event, int flags) struct hw_perf_event *hwc = &event->hw; int idx = hwc->idx; + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return; + WARN_ON(idx < 0); armpmu_stop(event, PERF_EF_UPDATE); @@ -279,10 +232,14 @@ armpmu_add(struct perf_event *event, int flags) int idx; int err = 0; + /* An event following a process won't be stopped earlier */ + if (!cpumask_test_cpu(smp_processor_id(), &armpmu->valid_cpus)) + return 0; + perf_pmu_disable(event->pmu); /* If we don't have a space for the counter then finish early. */ - idx = armpmu->get_event_idx(hw_events, hwc); + idx = armpmu->get_event_idx(hw_events, event); if (idx < 0) { err = idx; goto out; @@ -293,7 +250,7 @@ armpmu_add(struct perf_event *event, int flags) * sure it is disabled. */ event->hw.idx = idx; - armpmu->disable(hwc, idx); + armpmu->disable(event); hw_events->events[idx] = event; hwc->state = PERF_HES_STOPPED | PERF_HES_UPTODATE; @@ -313,13 +270,12 @@ validate_event(struct pmu_hw_events *hw_events, struct perf_event *event) { struct arm_pmu *armpmu = to_arm_pmu(event->pmu); - struct hw_perf_event fake_event = event->hw; struct pmu *leader_pmu = event->group_leader->pmu; if (event->pmu != leader_pmu || event->state <= PERF_EVENT_STATE_OFF) return 1; - return armpmu->get_event_idx(hw_events, &fake_event) >= 0; + return armpmu->get_event_idx(hw_events, event) >= 0; } static int @@ -350,99 +306,41 @@ validate_group(struct perf_event *event) return 0; } -static irqreturn_t armpmu_platform_irq(int irq, void *dev) +static irqreturn_t armpmu_dispatch_irq(int irq, void *dev) { struct arm_pmu *armpmu = (struct arm_pmu *) dev; struct platform_device *plat_device = armpmu->plat_device; struct arm_pmu_platdata *plat = dev_get_platdata(&plat_device->dev); - return plat->handle_irq(irq, dev, armpmu->handle_irq); + if (plat && plat->handle_irq) + return plat->handle_irq(irq, dev, armpmu->handle_irq); + else + return armpmu->handle_irq(irq, dev); } static void armpmu_release_hardware(struct arm_pmu *armpmu) { - int i, irq, irqs; - struct platform_device *pmu_device = armpmu->plat_device; - struct arm_pmu_platdata *plat = - dev_get_platdata(&pmu_device->dev); - - irqs = min(pmu_device->num_resources, num_possible_cpus()); - - for (i = 0; i < irqs; ++i) { - if (!cpumask_test_and_clear_cpu(i, &armpmu->active_irqs)) - continue; - irq = platform_get_irq(pmu_device, i); - if (irq >= 0) { - if (plat && plat->disable_irq) - plat->disable_irq(irq); - free_irq(irq, armpmu); - } - } - - release_pmu(armpmu->type); + armpmu->free_irq(armpmu); + pm_runtime_put_sync(&armpmu->plat_device->dev); } static int armpmu_reserve_hardware(struct arm_pmu *armpmu) { - struct arm_pmu_platdata *plat; - irq_handler_t handle_irq; - int i, err, irq, irqs; + int err; struct platform_device *pmu_device = armpmu->plat_device; if (!pmu_device) return -ENODEV; - err = reserve_pmu(armpmu->type); + pm_runtime_get_sync(&pmu_device->dev); + err = armpmu->request_irq(armpmu, armpmu_dispatch_irq); if (err) { - pr_warning("unable to reserve pmu\n"); + armpmu_release_hardware(armpmu); return err; } - plat = dev_get_platdata(&pmu_device->dev); - if (plat && plat->handle_irq) - handle_irq = armpmu_platform_irq; - else - handle_irq = armpmu->handle_irq; - - irqs = min(pmu_device->num_resources, num_possible_cpus()); - if (irqs < 1) { - pr_err("no irqs for PMUs defined\n"); - return -ENODEV; - } - - for (i = 0; i < irqs; ++i) { - err = 0; - irq = platform_get_irq(pmu_device, i); - if (irq < 0) - continue; - - /* - * If we have a single PMU interrupt that we can't shift, - * assume that we're running on a uniprocessor machine and - * continue. Otherwise, continue without this interrupt. - */ - if (irq_set_affinity(irq, cpumask_of(i)) && irqs > 1) { - pr_warning("unable to set irq affinity (irq=%d, cpu=%u)\n", - irq, i); - continue; - } - - err = request_irq(irq, handle_irq, - IRQF_DISABLED | IRQF_NOBALANCING, - "arm-pmu", armpmu); - if (err) { - pr_err("unable to request IRQ%d for ARM PMU counters\n", - irq); - armpmu_release_hardware(armpmu); - return err; - } else if (plat && plat->enable_irq) - plat->enable_irq(irq); - - cpumask_set_cpu(i, &armpmu->active_irqs); - } - return 0; } @@ -536,6 +434,10 @@ static int armpmu_event_init(struct perf_event *event) int err = 0; atomic_t *active_events = &armpmu->active_events; + if (event->cpu != -1 && + !cpumask_test_cpu(event->cpu, &armpmu->valid_cpus)) + return -ENOENT; + /* does not support taken branch sampling */ if (has_branch_stack(event)) return -EOPNOTSUPP; @@ -572,15 +474,41 @@ static void armpmu_enable(struct pmu *pmu) int enabled = bitmap_weight(hw_events->used_mask, armpmu->num_events); if (enabled) - armpmu->start(); + armpmu->start(armpmu); } static void armpmu_disable(struct pmu *pmu) { struct arm_pmu *armpmu = to_arm_pmu(pmu); - armpmu->stop(); + armpmu->stop(armpmu); } +#ifdef CONFIG_PM_RUNTIME +static int armpmu_runtime_resume(struct device *dev) +{ + struct arm_pmu_platdata *plat = dev_get_platdata(dev); + + if (plat && plat->runtime_resume) + return plat->runtime_resume(dev); + + return 0; +} + +static int armpmu_runtime_suspend(struct device *dev) +{ + struct arm_pmu_platdata *plat = dev_get_platdata(dev); + + if (plat && plat->runtime_suspend) + return plat->runtime_suspend(dev); + + return 0; +} +#endif + +const struct dev_pm_ops armpmu_dev_pm_ops = { + SET_RUNTIME_PM_OPS(armpmu_runtime_suspend, armpmu_runtime_resume, NULL) +}; + static void __init armpmu_init(struct arm_pmu *armpmu) { atomic_set(&armpmu->active_events, 0); @@ -598,174 +526,14 @@ static void __init armpmu_init(struct arm_pmu *armpmu) }; } -int __init armpmu_register(struct arm_pmu *armpmu, char *name, int type) +int armpmu_register(struct arm_pmu *armpmu, int type) { armpmu_init(armpmu); - return perf_pmu_register(&armpmu->pmu, name, type); + pr_info("enabled with %s PMU driver, %d counters available\n", + armpmu->name, armpmu->num_events); + return perf_pmu_register(&armpmu->pmu, armpmu->name, type); } -/* Include the PMU-specific implementations. */ -#include "perf_event_xscale.c" -#include "perf_event_v6.c" -#include "perf_event_v7.c" - -/* - * Ensure the PMU has sane values out of reset. - * This requires SMP to be available, so exists as a separate initcall. - */ -static int __init -cpu_pmu_reset(void) -{ - if (cpu_pmu && cpu_pmu->reset) - return on_each_cpu(cpu_pmu->reset, NULL, 1); - return 0; -} -arch_initcall(cpu_pmu_reset); - -/* - * PMU platform driver and devicetree bindings. - */ -static struct of_device_id armpmu_of_device_ids[] = { - {.compatible = "arm,cortex-a9-pmu"}, - {.compatible = "arm,cortex-a8-pmu"}, - {.compatible = "arm,arm1136-pmu"}, - {.compatible = "arm,arm1176-pmu"}, - {}, -}; - -static struct platform_device_id armpmu_plat_device_ids[] = { - {.name = "arm-pmu"}, - {}, -}; - -static int __devinit armpmu_device_probe(struct platform_device *pdev) -{ - if (!cpu_pmu) - return -ENODEV; - - cpu_pmu->plat_device = pdev; - return 0; -} - -static struct platform_driver armpmu_driver = { - .driver = { - .name = "arm-pmu", - .of_match_table = armpmu_of_device_ids, - }, - .probe = armpmu_device_probe, - .id_table = armpmu_plat_device_ids, -}; - -static int __init register_pmu_driver(void) -{ - return platform_driver_register(&armpmu_driver); -} -device_initcall(register_pmu_driver); - -static struct pmu_hw_events *armpmu_get_cpu_events(void) -{ - return &__get_cpu_var(cpu_hw_events); -} - -static void __init cpu_pmu_init(struct arm_pmu *armpmu) -{ - int cpu; - for_each_possible_cpu(cpu) { - struct pmu_hw_events *events = &per_cpu(cpu_hw_events, cpu); - events->events = per_cpu(hw_events, cpu); - events->used_mask = per_cpu(used_mask, cpu); - raw_spin_lock_init(&events->pmu_lock); - } - armpmu->get_hw_events = armpmu_get_cpu_events; - armpmu->type = ARM_PMU_DEVICE_CPU; -} - -/* - * PMU hardware loses all context when a CPU goes offline. - * When a CPU is hotplugged back in, since some hardware registers are - * UNKNOWN at reset, the PMU must be explicitly reset to avoid reading - * junk values out of them. - */ -static int __cpuinit pmu_cpu_notify(struct notifier_block *b, - unsigned long action, void *hcpu) -{ - if ((action & ~CPU_TASKS_FROZEN) != CPU_STARTING) - return NOTIFY_DONE; - - if (cpu_pmu && cpu_pmu->reset) - cpu_pmu->reset(NULL); - - return NOTIFY_OK; -} - -static struct notifier_block __cpuinitdata pmu_cpu_notifier = { - .notifier_call = pmu_cpu_notify, -}; - -/* - * CPU PMU identification and registration. - */ -static int __init -init_hw_perf_events(void) -{ - unsigned long cpuid = read_cpuid_id(); - unsigned long implementor = (cpuid & 0xFF000000) >> 24; - unsigned long part_number = (cpuid & 0xFFF0); - - /* ARM Ltd CPUs. */ - if (0x41 == implementor) { - switch (part_number) { - case 0xB360: /* ARM1136 */ - case 0xB560: /* ARM1156 */ - case 0xB760: /* ARM1176 */ - cpu_pmu = armv6pmu_init(); - break; - case 0xB020: /* ARM11mpcore */ - cpu_pmu = armv6mpcore_pmu_init(); - break; - case 0xC080: /* Cortex-A8 */ - cpu_pmu = armv7_a8_pmu_init(); - break; - case 0xC090: /* Cortex-A9 */ - cpu_pmu = armv7_a9_pmu_init(); - break; - case 0xC050: /* Cortex-A5 */ - cpu_pmu = armv7_a5_pmu_init(); - break; - case 0xC0F0: /* Cortex-A15 */ - cpu_pmu = armv7_a15_pmu_init(); - break; - case 0xC070: /* Cortex-A7 */ - cpu_pmu = armv7_a7_pmu_init(); - break; - } - /* Intel CPUs [xscale]. */ - } else if (0x69 == implementor) { - part_number = (cpuid >> 13) & 0x7; - switch (part_number) { - case 1: - cpu_pmu = xscale1pmu_init(); - break; - case 2: - cpu_pmu = xscale2pmu_init(); - break; - } - } - - if (cpu_pmu) { - pr_info("enabled with %s PMU driver, %d counters available\n", - cpu_pmu->name, cpu_pmu->num_events); - cpu_pmu_init(cpu_pmu); - register_cpu_notifier(&pmu_cpu_notifier); - armpmu_register(cpu_pmu, cpu_pmu->name, PERF_TYPE_RAW); - } else { - pr_info("no hardware support available\n"); - } - - return 0; -} -early_initcall(init_hw_perf_events); - /* * Callchain handling code. */ @@ -817,6 +585,10 @@ perf_callchain_user(struct perf_callchain_entry *entry, struct pt_regs *regs) { struct frame_tail __user *tail; + if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) { + /* We don't support guest os callchain now */ + return; + } tail = (struct frame_tail __user *)regs->ARM_fp - 1; @@ -844,9 +616,41 @@ perf_callchain_kernel(struct perf_callchain_entry *entry, struct pt_regs *regs) { struct stackframe fr; + if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) { + /* We don't support guest os callchain now */ + return; + } + fr.fp = regs->ARM_fp; fr.sp = regs->ARM_sp; fr.lr = regs->ARM_lr; fr.pc = regs->ARM_pc; walk_stackframe(&fr, callchain_trace, entry); } + +unsigned long perf_instruction_pointer(struct pt_regs *regs) +{ + if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) + return perf_guest_cbs->get_guest_ip(); + + return instruction_pointer(regs); +} + +unsigned long perf_misc_flags(struct pt_regs *regs) +{ + int misc = 0; + + if (perf_guest_cbs && perf_guest_cbs->is_in_guest()) { + if (perf_guest_cbs->is_user_mode()) + misc |= PERF_RECORD_MISC_GUEST_USER; + else + misc |= PERF_RECORD_MISC_GUEST_KERNEL; + } else { + if (user_mode(regs)) + misc |= PERF_RECORD_MISC_USER; + else + misc |= PERF_RECORD_MISC_KERNEL; + } + + return misc; +} diff --git a/arch/arm/kernel/perf_event_cpu.c b/arch/arm/kernel/perf_event_cpu.c new file mode 100644 index 000000000000..66248ee706cb --- /dev/null +++ b/arch/arm/kernel/perf_event_cpu.c @@ -0,0 +1,380 @@ +/* + * 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. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + * + * Copyright (C) 2012 ARM Limited + * + * Author: Will Deacon <will.deacon@arm.com> + */ +#define pr_fmt(fmt) "CPU PMU: " fmt + +#include <linux/bitmap.h> +#include <linux/cpu_pm.h> +#include <linux/export.h> +#include <linux/kernel.h> +#include <linux/of.h> +#include <linux/platform_device.h> +#include <linux/slab.h> +#include <linux/spinlock.h> + +#include <asm/cputype.h> +#include <asm/irq_regs.h> +#include <asm/pmu.h> + +/* Set at runtime when we know what CPU type we are. */ +static DEFINE_PER_CPU(struct arm_pmu *, cpu_pmu); + +static DEFINE_PER_CPU(struct perf_event * [ARMPMU_MAX_HWEVENTS], hw_events); +static DEFINE_PER_CPU(unsigned long [BITS_TO_LONGS(ARMPMU_MAX_HWEVENTS)], used_mask); +static DEFINE_PER_CPU(struct pmu_hw_events, cpu_hw_events); + +static DEFINE_PER_CPU(struct cpupmu_regs, cpu_pmu_regs); + +/* + * Despite the names, these two functions are CPU-specific and are used + * by the OProfile/perf code. + */ +const char *perf_pmu_name(void) +{ + struct arm_pmu *pmu = per_cpu(cpu_pmu, 0); + if (!pmu) + return NULL; + + return pmu->name; +} +EXPORT_SYMBOL_GPL(perf_pmu_name); + +int perf_num_counters(void) +{ + struct arm_pmu *pmu = per_cpu(cpu_pmu, 0); + + if (!pmu) + return 0; + + return pmu->num_events; +} +EXPORT_SYMBOL_GPL(perf_num_counters); + +/* Include the PMU-specific implementations. */ +#include "perf_event_xscale.c" +#include "perf_event_v6.c" +#include "perf_event_v7.c" + +static struct pmu_hw_events *cpu_pmu_get_cpu_events(void) +{ + return &__get_cpu_var(cpu_hw_events); +} + +static void cpu_pmu_free_irq(struct arm_pmu *cpu_pmu) +{ + int i, irq, irqs; + struct platform_device *pmu_device = cpu_pmu->plat_device; + int cpu = -1; + + irqs = min(pmu_device->num_resources, num_possible_cpus()); + + for (i = 0; i < irqs; ++i) { + cpu = cpumask_next(cpu, &cpu_pmu->valid_cpus); + if (!cpumask_test_and_clear_cpu(cpu, &cpu_pmu->active_irqs)) + continue; + irq = platform_get_irq(pmu_device, i); + if (irq >= 0) + free_irq(irq, cpu_pmu); + } +} + +static int cpu_pmu_request_irq(struct arm_pmu *cpu_pmu, irq_handler_t handler) +{ + int i, err, irq, irqs; + struct platform_device *pmu_device = cpu_pmu->plat_device; + int cpu = -1; + + if (!pmu_device) + return -ENODEV; + + irqs = min(pmu_device->num_resources, num_possible_cpus()); + if (irqs < 1) { + pr_err("no irqs for PMUs defined\n"); + return -ENODEV; + } + + for (i = 0; i < irqs; ++i) { + err = 0; + cpu = cpumask_next(cpu, &cpu_pmu->valid_cpus); + irq = platform_get_irq(pmu_device, i); + if (irq < 0) + continue; + + /* + * If we have a single PMU interrupt that we can't shift, + * assume that we're running on a uniprocessor machine and + * continue. Otherwise, continue without this interrupt. + */ + if (irq_set_affinity(irq, cpumask_of(cpu)) && irqs > 1) { + pr_warning("unable to set irq affinity (irq=%d, cpu=%u)\n", + irq, i); + continue; + } + + err = request_irq(irq, handler, IRQF_NOBALANCING, "arm-pmu", + cpu_pmu); + if (err) { + pr_err("unable to request IRQ%d for ARM PMU counters\n", + irq); + return err; + } + + cpumask_set_cpu(cpu, &cpu_pmu->active_irqs); + } + + return 0; +} + +static void __devinit cpu_pmu_init(struct arm_pmu *cpu_pmu) +{ + int cpu; + for_each_cpu_mask(cpu, cpu_pmu->valid_cpus) { + struct pmu_hw_events *events = &per_cpu(cpu_hw_events, cpu); + events->events = per_cpu(hw_events, cpu); + events->used_mask = per_cpu(used_mask, cpu); + raw_spin_lock_init(&events->pmu_lock); + } + + cpu_pmu->get_hw_events = cpu_pmu_get_cpu_events; + cpu_pmu->request_irq = cpu_pmu_request_irq; + cpu_pmu->free_irq = cpu_pmu_free_irq; + + /* Ensure the PMU has sane values out of reset. */ + if (cpu_pmu && cpu_pmu->reset) + on_each_cpu_mask(&cpu_pmu->valid_cpus, cpu_pmu->reset, cpu_pmu, 1); +} + +/* + * PMU hardware loses all context when a CPU goes offline. + * When a CPU is hotplugged back in, since some hardware registers are + * UNKNOWN at reset, the PMU must be explicitly reset to avoid reading + * junk values out of them. + */ +static int __cpuinit cpu_pmu_notify(struct notifier_block *b, + unsigned long action, void *hcpu) +{ + struct arm_pmu *pmu = per_cpu(cpu_pmu, (long)hcpu); + + if ((action & ~CPU_TASKS_FROZEN) != CPU_STARTING) + return NOTIFY_DONE; + + if (pmu && pmu->reset) + pmu->reset(pmu); + else + return NOTIFY_DONE; + + return NOTIFY_OK; +} + +static int cpu_pmu_pm_notify(struct notifier_block *b, + unsigned long action, void *hcpu) +{ + int cpu = smp_processor_id(); + struct arm_pmu *pmu = per_cpu(cpu_pmu, cpu); + struct cpupmu_regs *pmuregs = &per_cpu(cpu_pmu_regs, cpu); + + if (!pmu) + return NOTIFY_DONE; + + if (action == CPU_PM_ENTER && pmu->save_regs) { + pmu->save_regs(pmu, pmuregs); + } else if (action == CPU_PM_EXIT && pmu->restore_regs) { + pmu->restore_regs(pmu, pmuregs); + } + + return NOTIFY_OK; +} + +static struct notifier_block __cpuinitdata cpu_pmu_hotplug_notifier = { + .notifier_call = cpu_pmu_notify, +}; + +static struct notifier_block __cpuinitdata cpu_pmu_pm_notifier = { + .notifier_call = cpu_pmu_pm_notify, +}; + +/* + * PMU platform driver and devicetree bindings. + */ +static struct of_device_id __devinitdata cpu_pmu_of_device_ids[] = { + {.compatible = "arm,cortex-a15-pmu", .data = armv7_a15_pmu_init}, + {.compatible = "arm,cortex-a9-pmu", .data = armv7_a9_pmu_init}, + {.compatible = "arm,cortex-a8-pmu", .data = armv7_a8_pmu_init}, + {.compatible = "arm,cortex-a7-pmu", .data = armv7_a7_pmu_init}, + {.compatible = "arm,cortex-a5-pmu", .data = armv7_a5_pmu_init}, + {.compatible = "arm,arm11mpcore-pmu", .data = armv6mpcore_pmu_init}, + {.compatible = "arm,arm1176-pmu", .data = armv6pmu_init}, + {.compatible = "arm,arm1136-pmu", .data = armv6pmu_init}, + {}, +}; + +static struct platform_device_id __devinitdata cpu_pmu_plat_device_ids[] = { + {.name = "arm-pmu"}, + {}, +}; + +/* + * CPU PMU identification and probing. + */ +static int __devinit probe_current_pmu(struct arm_pmu *pmu) +{ + int cpu = get_cpu(); + unsigned long cpuid = read_cpuid_id(); + unsigned long implementor = (cpuid & 0xFF000000) >> 24; + unsigned long part_number = (cpuid & 0xFFF0); + int ret = -ENODEV; + + pr_info("probing PMU on CPU %d\n", cpu); + + /* ARM Ltd CPUs. */ + if (0x41 == implementor) { + switch (part_number) { + case 0xB360: /* ARM1136 */ + case 0xB560: /* ARM1156 */ + case 0xB760: /* ARM1176 */ + ret = armv6pmu_init(pmu); + break; + case 0xB020: /* ARM11mpcore */ + ret = armv6mpcore_pmu_init(pmu); + break; + case 0xC080: /* Cortex-A8 */ + ret = armv7_a8_pmu_init(pmu); + break; + case 0xC090: /* Cortex-A9 */ + ret = armv7_a9_pmu_init(pmu); + break; + case 0xC050: /* Cortex-A5 */ + ret = armv7_a5_pmu_init(pmu); + break; + case 0xC0F0: /* Cortex-A15 */ + ret = armv7_a15_pmu_init(pmu); + break; + case 0xC070: /* Cortex-A7 */ + ret = armv7_a7_pmu_init(pmu); + break; + } + /* Intel CPUs [xscale]. */ + } else if (0x69 == implementor) { + part_number = (cpuid >> 13) & 0x7; + switch (part_number) { + case 1: + ret = xscale1pmu_init(pmu); + break; + case 2: + ret = xscale2pmu_init(pmu); + break; + } + } + + /* assume PMU support all the CPUs in this case */ + cpumask_setall(&pmu->valid_cpus); + + put_cpu(); + return ret; +} + +static int __devinit cpu_pmu_device_probe(struct platform_device *pdev) +{ + const struct of_device_id *of_id; + struct device_node *node = pdev->dev.of_node; + struct arm_pmu *pmu; + int ret = 0; + int cpu; + + pmu = kzalloc(sizeof(struct arm_pmu), GFP_KERNEL); + if (!pmu) { + pr_info("failed to allocate PMU device!"); + return -ENOMEM; + } + + if (node && (of_id = of_match_node(cpu_pmu_of_device_ids, pdev->dev.of_node))) { + smp_call_func_t init_fn = (smp_call_func_t)of_id->data; + struct device_node *ncluster; + int cluster = -1; + cpumask_t sibling_mask; + + ncluster = of_parse_phandle(node, "cluster", 0); + if (ncluster) { + int len; + const u32 *hwid; + hwid = of_get_property(ncluster, "reg", &len); + if (hwid && len == 4) + cluster = be32_to_cpup(hwid); + } + /* set sibling mask to all cpu mask if socket is not specified */ + if (cluster == -1 || + cluster_to_logical_mask(cluster, &sibling_mask)) + cpumask_setall(&sibling_mask); + + smp_call_function_any(&sibling_mask, init_fn, pmu, 1); + + /* now set the valid_cpus after init */ + cpumask_copy(&pmu->valid_cpus, &sibling_mask); + } else { + ret = probe_current_pmu(pmu); + } + + if (ret) { + pr_info("failed to register PMU devices!"); + kfree(pmu); + return ret; + } + + for_each_cpu_mask(cpu, pmu->valid_cpus) + per_cpu(cpu_pmu, cpu) = pmu; + + pmu->plat_device = pdev; + cpu_pmu_init(pmu); + armpmu_register(pmu, -1); + + return 0; +} + +static struct platform_driver cpu_pmu_driver = { + .driver = { + .name = "arm-pmu", + .pm = &armpmu_dev_pm_ops, + .of_match_table = cpu_pmu_of_device_ids, + }, + .probe = cpu_pmu_device_probe, + .id_table = cpu_pmu_plat_device_ids, +}; + +static int __init register_pmu_driver(void) +{ + int err; + + err = register_cpu_notifier(&cpu_pmu_hotplug_notifier); + if (err) + return err; + + err = cpu_pm_register_notifier(&cpu_pmu_pm_notifier); + if (err) { + unregister_cpu_notifier(&cpu_pmu_hotplug_notifier); + return err; + } + + err = platform_driver_register(&cpu_pmu_driver); + if (err) { + cpu_pm_unregister_notifier(&cpu_pmu_pm_notifier); + unregister_cpu_notifier(&cpu_pmu_hotplug_notifier); + } + + return err; +} +device_initcall(register_pmu_driver); diff --git a/arch/arm/kernel/perf_event_v6.c b/arch/arm/kernel/perf_event_v6.c index c90fcb2b6967..b4b0c0845116 100644 --- a/arch/arm/kernel/perf_event_v6.c +++ b/arch/arm/kernel/perf_event_v6.c @@ -401,9 +401,10 @@ armv6_pmcr_counter_has_overflowed(unsigned long pmcr, return ret; } -static inline u32 -armv6pmu_read_counter(int counter) +static inline u32 armv6pmu_read_counter(struct perf_event *event) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; unsigned long value = 0; if (ARMV6_CYCLE_COUNTER == counter) @@ -418,10 +419,11 @@ armv6pmu_read_counter(int counter) return value; } -static inline void -armv6pmu_write_counter(int counter, - u32 value) +static inline void armv6pmu_write_counter(struct perf_event *event, u32 value) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; + if (ARMV6_CYCLE_COUNTER == counter) asm volatile("mcr p15, 0, %0, c15, c12, 1" : : "r"(value)); else if (ARMV6_COUNTER0 == counter) @@ -432,12 +434,13 @@ armv6pmu_write_counter(int counter, WARN_ONCE(1, "invalid counter number (%d)\n", counter); } -static void -armv6pmu_enable_event(struct hw_perf_event *hwc, - int idx) +static void armv6pmu_enable_event(struct perf_event *event) { unsigned long val, mask, evt, flags; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; if (ARMV6_CYCLE_COUNTER == idx) { mask = 0; @@ -473,7 +476,8 @@ armv6pmu_handle_irq(int irq_num, { unsigned long pmcr = armv6_pmcr_read(); struct perf_sample_data data; - struct pmu_hw_events *cpuc; + struct arm_pmu *cpu_pmu = (struct arm_pmu *)dev; + struct pmu_hw_events *cpuc = cpu_pmu->get_hw_events(); struct pt_regs *regs; int idx; @@ -489,7 +493,6 @@ armv6pmu_handle_irq(int irq_num, */ armv6_pmcr_write(pmcr); - cpuc = &__get_cpu_var(cpu_hw_events); for (idx = 0; idx < cpu_pmu->num_events; ++idx) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -506,13 +509,13 @@ armv6pmu_handle_irq(int irq_num, continue; hwc = &event->hw; - armpmu_event_update(event, hwc, idx); + armpmu_event_update(event); perf_sample_data_init(&data, 0, hwc->last_period); - if (!armpmu_event_set_period(event, hwc, idx)) + if (!armpmu_event_set_period(event)) continue; if (perf_event_overflow(event, &data, regs)) - cpu_pmu->disable(hwc, idx); + cpu_pmu->disable(event); } /* @@ -527,8 +530,7 @@ armv6pmu_handle_irq(int irq_num, return IRQ_HANDLED; } -static void -armv6pmu_start(void) +static void armv6pmu_start(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -540,8 +542,7 @@ armv6pmu_start(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -armv6pmu_stop(void) +static void armv6pmu_stop(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -555,10 +556,11 @@ armv6pmu_stop(void) static int armv6pmu_get_event_idx(struct pmu_hw_events *cpuc, - struct hw_perf_event *event) + struct perf_event *event) { + struct hw_perf_event *hwc = &event->hw; /* Always place a cycle counter into the cycle counter. */ - if (ARMV6_PERFCTR_CPU_CYCLES == event->config_base) { + if (ARMV6_PERFCTR_CPU_CYCLES == hwc->config_base) { if (test_and_set_bit(ARMV6_CYCLE_COUNTER, cpuc->used_mask)) return -EAGAIN; @@ -579,12 +581,13 @@ armv6pmu_get_event_idx(struct pmu_hw_events *cpuc, } } -static void -armv6pmu_disable_event(struct hw_perf_event *hwc, - int idx) +static void armv6pmu_disable_event(struct perf_event *event) { unsigned long val, mask, evt, flags; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; if (ARMV6_CYCLE_COUNTER == idx) { mask = ARMV6_PMCR_CCOUNT_IEN; @@ -613,12 +616,13 @@ armv6pmu_disable_event(struct hw_perf_event *hwc, raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc, - int idx) +static void armv6mpcore_pmu_disable_event(struct perf_event *event) { unsigned long val, mask, flags, evt = 0; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; if (ARMV6_CYCLE_COUNTER == idx) { mask = ARMV6_PMCR_CCOUNT_IEN; @@ -645,28 +649,28 @@ armv6mpcore_pmu_disable_event(struct hw_perf_event *hwc, static int armv6_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv6_perf_map, + return armpmu_map_event(event, &armv6_perf_map, &armv6_perf_cache_map, 0xFF); } -static struct arm_pmu armv6pmu = { - .name = "v6", - .handle_irq = armv6pmu_handle_irq, - .enable = armv6pmu_enable_event, - .disable = armv6pmu_disable_event, - .read_counter = armv6pmu_read_counter, - .write_counter = armv6pmu_write_counter, - .get_event_idx = armv6pmu_get_event_idx, - .start = armv6pmu_start, - .stop = armv6pmu_stop, - .map_event = armv6_map_event, - .num_events = 3, - .max_period = (1LLU << 32) - 1, -}; - -static struct arm_pmu *__init armv6pmu_init(void) +static int __devinit armv6pmu_init(struct arm_pmu *cpu_pmu) { - return &armv6pmu; + *cpu_pmu = (struct arm_pmu) { + .name = "v6", + .handle_irq = armv6pmu_handle_irq, + .enable = armv6pmu_enable_event, + .disable = armv6pmu_disable_event, + .read_counter = armv6pmu_read_counter, + .write_counter = armv6pmu_write_counter, + .get_event_idx = armv6pmu_get_event_idx, + .start = armv6pmu_start, + .stop = armv6pmu_stop, + .map_event = armv6_map_event, + .num_events = 3, + .max_period = (1LLU << 32) - 1, + }; + + return 0; } /* @@ -679,37 +683,37 @@ static struct arm_pmu *__init armv6pmu_init(void) static int armv6mpcore_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv6mpcore_perf_map, + return armpmu_map_event(event, &armv6mpcore_perf_map, &armv6mpcore_perf_cache_map, 0xFF); } -static struct arm_pmu armv6mpcore_pmu = { - .name = "v6mpcore", - .handle_irq = armv6pmu_handle_irq, - .enable = armv6pmu_enable_event, - .disable = armv6mpcore_pmu_disable_event, - .read_counter = armv6pmu_read_counter, - .write_counter = armv6pmu_write_counter, - .get_event_idx = armv6pmu_get_event_idx, - .start = armv6pmu_start, - .stop = armv6pmu_stop, - .map_event = armv6mpcore_map_event, - .num_events = 3, - .max_period = (1LLU << 32) - 1, -}; - -static struct arm_pmu *__init armv6mpcore_pmu_init(void) +static int __devinit armv6mpcore_pmu_init(struct arm_pmu *cpu_pmu) { - return &armv6mpcore_pmu; + *cpu_pmu = (struct arm_pmu) { + .name = "v6mpcore", + .handle_irq = armv6pmu_handle_irq, + .enable = armv6pmu_enable_event, + .disable = armv6mpcore_pmu_disable_event, + .read_counter = armv6pmu_read_counter, + .write_counter = armv6pmu_write_counter, + .get_event_idx = armv6pmu_get_event_idx, + .start = armv6pmu_start, + .stop = armv6pmu_stop, + .map_event = armv6mpcore_map_event, + .num_events = 3, + .max_period = (1LLU << 32) - 1, + }; + + return 0; } #else -static struct arm_pmu *__init armv6pmu_init(void) +static int armv6pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init armv6mpcore_pmu_init(void) +static int armv6mpcore_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } #endif /* CONFIG_CPU_V6 || CONFIG_CPU_V6K */ diff --git a/arch/arm/kernel/perf_event_v7.c b/arch/arm/kernel/perf_event_v7.c index f04070bd2183..3565d8084d68 100644 --- a/arch/arm/kernel/perf_event_v7.c +++ b/arch/arm/kernel/perf_event_v7.c @@ -18,8 +18,6 @@ #ifdef CONFIG_CPU_V7 -static struct arm_pmu armv7pmu; - /* * Common ARMv7 event types * @@ -738,7 +736,8 @@ static const unsigned armv7_a7_perf_cache_map[PERF_COUNT_HW_CACHE_MAX] */ #define ARMV7_IDX_CYCLE_COUNTER 0 #define ARMV7_IDX_COUNTER0 1 -#define ARMV7_IDX_COUNTER_LAST (ARMV7_IDX_CYCLE_COUNTER + cpu_pmu->num_events - 1) +#define ARMV7_IDX_COUNTER_LAST(cpu_pmu) \ + (ARMV7_IDX_CYCLE_COUNTER + cpu_pmu->num_events - 1) #define ARMV7_MAX_COUNTERS 32 #define ARMV7_COUNTER_MASK (ARMV7_MAX_COUNTERS - 1) @@ -804,49 +803,34 @@ static inline int armv7_pmnc_has_overflowed(u32 pmnc) return pmnc & ARMV7_OVERFLOWED_MASK; } -static inline int armv7_pmnc_counter_valid(int idx) +static inline int armv7_pmnc_counter_valid(struct arm_pmu *cpu_pmu, int idx) { - return idx >= ARMV7_IDX_CYCLE_COUNTER && idx <= ARMV7_IDX_COUNTER_LAST; + return idx >= ARMV7_IDX_CYCLE_COUNTER && + idx <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); } static inline int armv7_pmnc_counter_has_overflowed(u32 pmnc, int idx) { - int ret = 0; - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u checking wrong counter %d overflow status\n", - smp_processor_id(), idx); - } else { - counter = ARMV7_IDX_TO_COUNTER(idx); - ret = pmnc & BIT(counter); - } - - return ret; + return pmnc & BIT(ARMV7_IDX_TO_COUNTER(idx)); } static inline int armv7_pmnc_select_counter(int idx) { - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u selecting wrong PMNC counter %d\n", - smp_processor_id(), idx); - return -EINVAL; - } - - counter = ARMV7_IDX_TO_COUNTER(idx); + u32 counter = ARMV7_IDX_TO_COUNTER(idx); asm volatile("mcr p15, 0, %0, c9, c12, 5" : : "r" (counter)); isb(); return idx; } -static inline u32 armv7pmu_read_counter(int idx) +static inline u32 armv7pmu_read_counter(struct perf_event *event) { + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; + int idx = hwc->idx; u32 value = 0; - if (!armv7_pmnc_counter_valid(idx)) + if (!armv7_pmnc_counter_valid(cpu_pmu, idx)) pr_err("CPU%u reading wrong counter %d\n", smp_processor_id(), idx); else if (idx == ARMV7_IDX_CYCLE_COUNTER) @@ -857,9 +841,13 @@ static inline u32 armv7pmu_read_counter(int idx) return value; } -static inline void armv7pmu_write_counter(int idx, u32 value) +static inline void armv7pmu_write_counter(struct perf_event *event, u32 value) { - if (!armv7_pmnc_counter_valid(idx)) + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; + int idx = hwc->idx; + + if (!armv7_pmnc_counter_valid(cpu_pmu, idx)) pr_err("CPU%u writing wrong counter %d\n", smp_processor_id(), idx); else if (idx == ARMV7_IDX_CYCLE_COUNTER) @@ -878,60 +866,28 @@ static inline void armv7_pmnc_write_evtsel(int idx, u32 val) static inline int armv7_pmnc_enable_counter(int idx) { - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u enabling wrong PMNC counter %d\n", - smp_processor_id(), idx); - return -EINVAL; - } - - counter = ARMV7_IDX_TO_COUNTER(idx); + u32 counter = ARMV7_IDX_TO_COUNTER(idx); asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (BIT(counter))); return idx; } static inline int armv7_pmnc_disable_counter(int idx) { - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u disabling wrong PMNC counter %d\n", - smp_processor_id(), idx); - return -EINVAL; - } - - counter = ARMV7_IDX_TO_COUNTER(idx); + u32 counter = ARMV7_IDX_TO_COUNTER(idx); asm volatile("mcr p15, 0, %0, c9, c12, 2" : : "r" (BIT(counter))); return idx; } static inline int armv7_pmnc_enable_intens(int idx) { - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u enabling wrong PMNC counter IRQ enable %d\n", - smp_processor_id(), idx); - return -EINVAL; - } - - counter = ARMV7_IDX_TO_COUNTER(idx); + u32 counter = ARMV7_IDX_TO_COUNTER(idx); asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (BIT(counter))); return idx; } static inline int armv7_pmnc_disable_intens(int idx) { - u32 counter; - - if (!armv7_pmnc_counter_valid(idx)) { - pr_err("CPU%u disabling wrong PMNC counter IRQ enable %d\n", - smp_processor_id(), idx); - return -EINVAL; - } - - counter = ARMV7_IDX_TO_COUNTER(idx); + u32 counter = ARMV7_IDX_TO_COUNTER(idx); asm volatile("mcr p15, 0, %0, c9, c14, 2" : : "r" (BIT(counter))); isb(); /* Clear the overflow flag in case an interrupt is pending. */ @@ -956,7 +912,7 @@ static inline u32 armv7_pmnc_getreset_flags(void) } #ifdef DEBUG -static void armv7_pmnc_dump_regs(void) +static void armv7_pmnc_dump_regs(struct arm_pmu *cpu_pmu) { u32 val; unsigned int cnt; @@ -981,7 +937,8 @@ static void armv7_pmnc_dump_regs(void) asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (val)); printk(KERN_INFO "CCNT =0x%08x\n", val); - for (cnt = ARMV7_IDX_COUNTER0; cnt <= ARMV7_IDX_COUNTER_LAST; cnt++) { + for (cnt = ARMV7_IDX_COUNTER0; + cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { armv7_pmnc_select_counter(cnt); asm volatile("mrc p15, 0, %0, c9, c13, 2" : "=r" (val)); printk(KERN_INFO "CNT[%d] count =0x%08x\n", @@ -993,10 +950,64 @@ static void armv7_pmnc_dump_regs(void) } #endif -static void armv7pmu_enable_event(struct hw_perf_event *hwc, int idx) +static void armv7pmu_save_regs(struct arm_pmu *cpu_pmu, + struct cpupmu_regs *regs) +{ + unsigned int cnt; + asm volatile("mrc p15, 0, %0, c9, c12, 0" : "=r" (regs->pmc)); + if (!(regs->pmc & ARMV7_PMNC_E)) + return; + + asm volatile("mrc p15, 0, %0, c9, c12, 1" : "=r" (regs->pmcntenset)); + asm volatile("mrc p15, 0, %0, c9, c14, 0" : "=r" (regs->pmuseren)); + asm volatile("mrc p15, 0, %0, c9, c14, 1" : "=r" (regs->pmintenset)); + asm volatile("mrc p15, 0, %0, c9, c13, 0" : "=r" (regs->pmxevtcnt[0])); + for (cnt = ARMV7_IDX_COUNTER0; + cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { + armv7_pmnc_select_counter(cnt); + asm volatile("mrc p15, 0, %0, c9, c13, 1" + : "=r"(regs->pmxevttype[cnt])); + asm volatile("mrc p15, 0, %0, c9, c13, 2" + : "=r"(regs->pmxevtcnt[cnt])); + } + return; +} + +static void armv7pmu_restore_regs(struct arm_pmu *cpu_pmu, + struct cpupmu_regs *regs) +{ + unsigned int cnt; + if (!(regs->pmc & ARMV7_PMNC_E)) + return; + + asm volatile("mcr p15, 0, %0, c9, c12, 1" : : "r" (regs->pmcntenset)); + asm volatile("mcr p15, 0, %0, c9, c14, 0" : : "r" (regs->pmuseren)); + asm volatile("mcr p15, 0, %0, c9, c14, 1" : : "r" (regs->pmintenset)); + asm volatile("mcr p15, 0, %0, c9, c13, 0" : : "r" (regs->pmxevtcnt[0])); + for (cnt = ARMV7_IDX_COUNTER0; + cnt <= ARMV7_IDX_COUNTER_LAST(cpu_pmu); cnt++) { + armv7_pmnc_select_counter(cnt); + asm volatile("mcr p15, 0, %0, c9, c13, 1" + : : "r"(regs->pmxevttype[cnt])); + asm volatile("mcr p15, 0, %0, c9, c13, 2" + : : "r"(regs->pmxevtcnt[cnt])); + } + asm volatile("mcr p15, 0, %0, c9, c12, 0" : : "r" (regs->pmc)); +} + +static void armv7pmu_enable_event(struct perf_event *event) { unsigned long flags; + struct hw_perf_event *hwc = &event->hw; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; + + if (!armv7_pmnc_counter_valid(cpu_pmu, idx)) { + pr_err("CPU%u enabling wrong PMNC counter IRQ enable %d\n", + smp_processor_id(), idx); + return; + } /* * Enable counter and interrupt, and set the counter to count @@ -1014,7 +1025,7 @@ static void armv7pmu_enable_event(struct hw_perf_event *hwc, int idx) * We only need to set the event for the cycle counter if we * have the ability to perform event filtering. */ - if (armv7pmu.set_event_filter || idx != ARMV7_IDX_CYCLE_COUNTER) + if (cpu_pmu->set_event_filter || idx != ARMV7_IDX_CYCLE_COUNTER) armv7_pmnc_write_evtsel(idx, hwc->config_base); /* @@ -1030,10 +1041,19 @@ static void armv7pmu_enable_event(struct hw_perf_event *hwc, int idx) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void armv7pmu_disable_event(struct hw_perf_event *hwc, int idx) +static void armv7pmu_disable_event(struct perf_event *event) { unsigned long flags; + struct hw_perf_event *hwc = &event->hw; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; + + if (!armv7_pmnc_counter_valid(cpu_pmu, idx)) { + pr_err("CPU%u disabling wrong PMNC counter IRQ enable %d\n", + smp_processor_id(), idx); + return; + } /* * Disable counter and interrupt @@ -1057,7 +1077,8 @@ static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev) { u32 pmnc; struct perf_sample_data data; - struct pmu_hw_events *cpuc; + struct arm_pmu *cpu_pmu = (struct arm_pmu *)dev; + struct pmu_hw_events *cpuc = cpu_pmu->get_hw_events(); struct pt_regs *regs; int idx; @@ -1077,7 +1098,6 @@ static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev) */ regs = get_irq_regs(); - cpuc = &__get_cpu_var(cpu_hw_events); for (idx = 0; idx < cpu_pmu->num_events; ++idx) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -1094,13 +1114,13 @@ static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev) continue; hwc = &event->hw; - armpmu_event_update(event, hwc, idx); + armpmu_event_update(event); perf_sample_data_init(&data, 0, hwc->last_period); - if (!armpmu_event_set_period(event, hwc, idx)) + if (!armpmu_event_set_period(event)) continue; if (perf_event_overflow(event, &data, regs)) - cpu_pmu->disable(hwc, idx); + cpu_pmu->disable(event); } /* @@ -1115,7 +1135,7 @@ static irqreturn_t armv7pmu_handle_irq(int irq_num, void *dev) return IRQ_HANDLED; } -static void armv7pmu_start(void) +static void armv7pmu_start(struct arm_pmu *cpu_pmu) { unsigned long flags; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -1126,7 +1146,7 @@ static void armv7pmu_start(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void armv7pmu_stop(void) +static void armv7pmu_stop(struct arm_pmu *cpu_pmu) { unsigned long flags; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -1138,10 +1158,12 @@ static void armv7pmu_stop(void) } static int armv7pmu_get_event_idx(struct pmu_hw_events *cpuc, - struct hw_perf_event *event) + struct perf_event *event) { int idx; - unsigned long evtype = event->config_base & ARMV7_EVTYPE_EVENT; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; + unsigned long evtype = hwc->config_base & ARMV7_EVTYPE_EVENT; /* Always place a cycle counter into the cycle counter. */ if (evtype == ARMV7_PERFCTR_CPU_CYCLES) { @@ -1192,11 +1214,14 @@ static int armv7pmu_set_event_filter(struct hw_perf_event *event, static void armv7pmu_reset(void *info) { + struct arm_pmu *cpu_pmu = (struct arm_pmu *)info; u32 idx, nb_cnt = cpu_pmu->num_events; /* The counter and interrupt enable registers are unknown at reset. */ - for (idx = ARMV7_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) - armv7pmu_disable_event(NULL, idx); + for (idx = ARMV7_IDX_CYCLE_COUNTER; idx < nb_cnt; ++idx) { + armv7_pmnc_disable_counter(idx); + armv7_pmnc_disable_intens(idx); + } /* Initialize & Reset PMNC: C and P bits */ armv7_pmnc_write(ARMV7_PMNC_P | ARMV7_PMNC_C); @@ -1204,48 +1229,53 @@ static void armv7pmu_reset(void *info) static int armv7_a8_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv7_a8_perf_map, + return armpmu_map_event(event, &armv7_a8_perf_map, &armv7_a8_perf_cache_map, 0xFF); } static int armv7_a9_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv7_a9_perf_map, + return armpmu_map_event(event, &armv7_a9_perf_map, &armv7_a9_perf_cache_map, 0xFF); } static int armv7_a5_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv7_a5_perf_map, + return armpmu_map_event(event, &armv7_a5_perf_map, &armv7_a5_perf_cache_map, 0xFF); } static int armv7_a15_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv7_a15_perf_map, + return armpmu_map_event(event, &armv7_a15_perf_map, &armv7_a15_perf_cache_map, 0xFF); } static int armv7_a7_map_event(struct perf_event *event) { - return map_cpu_event(event, &armv7_a7_perf_map, + return armpmu_map_event(event, &armv7_a7_perf_map, &armv7_a7_perf_cache_map, 0xFF); } -static struct arm_pmu armv7pmu = { - .handle_irq = armv7pmu_handle_irq, - .enable = armv7pmu_enable_event, - .disable = armv7pmu_disable_event, - .read_counter = armv7pmu_read_counter, - .write_counter = armv7pmu_write_counter, - .get_event_idx = armv7pmu_get_event_idx, - .start = armv7pmu_start, - .stop = armv7pmu_stop, - .reset = armv7pmu_reset, - .max_period = (1LLU << 32) - 1, +static void armv7pmu_init(struct arm_pmu *cpu_pmu) +{ + *cpu_pmu = (struct arm_pmu) { + .handle_irq = armv7pmu_handle_irq, + .enable = armv7pmu_enable_event, + .disable = armv7pmu_disable_event, + .read_counter = armv7pmu_read_counter, + .write_counter = armv7pmu_write_counter, + .get_event_idx = armv7pmu_get_event_idx, + .start = armv7pmu_start, + .stop = armv7pmu_stop, + .reset = armv7pmu_reset, + .save_regs = armv7pmu_save_regs, + .restore_regs = armv7pmu_restore_regs, + .max_period = (1LLU << 32) - 1, + }; }; -static u32 __init armv7_read_num_pmnc_events(void) +static u32 __devinit armv7_read_num_pmnc_events(void) { u32 nb_cnt; @@ -1256,70 +1286,75 @@ static u32 __init armv7_read_num_pmnc_events(void) return nb_cnt + 1; } -static struct arm_pmu *__init armv7_a8_pmu_init(void) +static int __devinit armv7_a8_pmu_init(struct arm_pmu *cpu_pmu) { - armv7pmu.name = "ARMv7 Cortex-A8"; - armv7pmu.map_event = armv7_a8_map_event; - armv7pmu.num_events = armv7_read_num_pmnc_events(); - return &armv7pmu; + armv7pmu_init(cpu_pmu); + cpu_pmu->name = "ARMv7_Cortex_A8"; + cpu_pmu->map_event = armv7_a8_map_event; + cpu_pmu->num_events = armv7_read_num_pmnc_events(); + return 0; } -static struct arm_pmu *__init armv7_a9_pmu_init(void) +static int __devinit armv7_a9_pmu_init(struct arm_pmu *cpu_pmu) { - armv7pmu.name = "ARMv7 Cortex-A9"; - armv7pmu.map_event = armv7_a9_map_event; - armv7pmu.num_events = armv7_read_num_pmnc_events(); - return &armv7pmu; + armv7pmu_init(cpu_pmu); + cpu_pmu->name = "ARMv7_Cortex_A9"; + cpu_pmu->map_event = armv7_a9_map_event; + cpu_pmu->num_events = armv7_read_num_pmnc_events(); + return 0; } -static struct arm_pmu *__init armv7_a5_pmu_init(void) +static int __devinit armv7_a5_pmu_init(struct arm_pmu *cpu_pmu) { - armv7pmu.name = "ARMv7 Cortex-A5"; - armv7pmu.map_event = armv7_a5_map_event; - armv7pmu.num_events = armv7_read_num_pmnc_events(); - return &armv7pmu; + armv7pmu_init(cpu_pmu); + cpu_pmu->name = "ARMv7_Cortex_A5"; + cpu_pmu->map_event = armv7_a5_map_event; + cpu_pmu->num_events = armv7_read_num_pmnc_events(); + return 0; } -static struct arm_pmu *__init armv7_a15_pmu_init(void) +static int __devinit armv7_a15_pmu_init(struct arm_pmu *cpu_pmu) { - armv7pmu.name = "ARMv7 Cortex-A15"; - armv7pmu.map_event = armv7_a15_map_event; - armv7pmu.num_events = armv7_read_num_pmnc_events(); - armv7pmu.set_event_filter = armv7pmu_set_event_filter; - return &armv7pmu; + armv7pmu_init(cpu_pmu); + cpu_pmu->name = "ARMv7_Cortex_A15"; + cpu_pmu->map_event = armv7_a15_map_event; + cpu_pmu->num_events = armv7_read_num_pmnc_events(); + cpu_pmu->set_event_filter = armv7pmu_set_event_filter; + return 0; } -static struct arm_pmu *__init armv7_a7_pmu_init(void) +static int __devinit armv7_a7_pmu_init(struct arm_pmu *cpu_pmu) { - armv7pmu.name = "ARMv7 Cortex-A7"; - armv7pmu.map_event = armv7_a7_map_event; - armv7pmu.num_events = armv7_read_num_pmnc_events(); - armv7pmu.set_event_filter = armv7pmu_set_event_filter; - return &armv7pmu; + armv7pmu_init(cpu_pmu); + cpu_pmu->name = "ARMv7_Cortex_A7"; + cpu_pmu->map_event = armv7_a7_map_event; + cpu_pmu->num_events = armv7_read_num_pmnc_events(); + cpu_pmu->set_event_filter = armv7pmu_set_event_filter; + return 0; } #else -static struct arm_pmu *__init armv7_a8_pmu_init(void) +static inline int armv7_a8_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init armv7_a9_pmu_init(void) +static inline int armv7_a9_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init armv7_a5_pmu_init(void) +static inline int armv7_a5_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init armv7_a15_pmu_init(void) +static inline int armv7_a15_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init armv7_a7_pmu_init(void) +static inline int armv7_a7_pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } #endif /* CONFIG_CPU_V7 */ diff --git a/arch/arm/kernel/perf_event_xscale.c b/arch/arm/kernel/perf_event_xscale.c index f759fe0bab63..1d3e1bf4865e 100644 --- a/arch/arm/kernel/perf_event_xscale.c +++ b/arch/arm/kernel/perf_event_xscale.c @@ -224,7 +224,8 @@ xscale1pmu_handle_irq(int irq_num, void *dev) { unsigned long pmnc; struct perf_sample_data data; - struct pmu_hw_events *cpuc; + struct arm_pmu *cpu_pmu = (struct arm_pmu *)dev; + struct pmu_hw_events *cpuc = cpu_pmu->get_hw_events(); struct pt_regs *regs; int idx; @@ -248,7 +249,6 @@ xscale1pmu_handle_irq(int irq_num, void *dev) regs = get_irq_regs(); - cpuc = &__get_cpu_var(cpu_hw_events); for (idx = 0; idx < cpu_pmu->num_events; ++idx) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -260,13 +260,13 @@ xscale1pmu_handle_irq(int irq_num, void *dev) continue; hwc = &event->hw; - armpmu_event_update(event, hwc, idx); + armpmu_event_update(event); perf_sample_data_init(&data, 0, hwc->last_period); - if (!armpmu_event_set_period(event, hwc, idx)) + if (!armpmu_event_set_period(event)) continue; if (perf_event_overflow(event, &data, regs)) - cpu_pmu->disable(hwc, idx); + cpu_pmu->disable(event); } irq_work_run(); @@ -280,11 +280,13 @@ xscale1pmu_handle_irq(int irq_num, void *dev) return IRQ_HANDLED; } -static void -xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx) +static void xscale1pmu_enable_event(struct perf_event *event) { unsigned long val, mask, evt, flags; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; switch (idx) { case XSCALE_CYCLE_COUNTER: @@ -314,11 +316,13 @@ xscale1pmu_enable_event(struct hw_perf_event *hwc, int idx) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx) +static void xscale1pmu_disable_event(struct perf_event *event) { unsigned long val, mask, evt, flags; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; switch (idx) { case XSCALE_CYCLE_COUNTER: @@ -348,9 +352,10 @@ xscale1pmu_disable_event(struct hw_perf_event *hwc, int idx) static int xscale1pmu_get_event_idx(struct pmu_hw_events *cpuc, - struct hw_perf_event *event) + struct perf_event *event) { - if (XSCALE_PERFCTR_CCNT == event->config_base) { + struct hw_perf_event *hwc = &event->hw; + if (XSCALE_PERFCTR_CCNT == hwc->config_base) { if (test_and_set_bit(XSCALE_CYCLE_COUNTER, cpuc->used_mask)) return -EAGAIN; @@ -366,8 +371,7 @@ xscale1pmu_get_event_idx(struct pmu_hw_events *cpuc, } } -static void -xscale1pmu_start(void) +static void xscale1pmu_start(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -379,8 +383,7 @@ xscale1pmu_start(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -xscale1pmu_stop(void) +static void xscale1pmu_stop(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -392,9 +395,10 @@ xscale1pmu_stop(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static inline u32 -xscale1pmu_read_counter(int counter) +static inline u32 xscale1pmu_read_counter(struct perf_event *event) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; u32 val = 0; switch (counter) { @@ -412,9 +416,11 @@ xscale1pmu_read_counter(int counter) return val; } -static inline void -xscale1pmu_write_counter(int counter, u32 val) +static inline void xscale1pmu_write_counter(struct perf_event *event, u32 val) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; + switch (counter) { case XSCALE_CYCLE_COUNTER: asm volatile("mcr p14, 0, %0, c1, c0, 0" : : "r" (val)); @@ -430,28 +436,28 @@ xscale1pmu_write_counter(int counter, u32 val) static int xscale_map_event(struct perf_event *event) { - return map_cpu_event(event, &xscale_perf_map, + return armpmu_map_event(event, &xscale_perf_map, &xscale_perf_cache_map, 0xFF); } -static struct arm_pmu xscale1pmu = { - .name = "xscale1", - .handle_irq = xscale1pmu_handle_irq, - .enable = xscale1pmu_enable_event, - .disable = xscale1pmu_disable_event, - .read_counter = xscale1pmu_read_counter, - .write_counter = xscale1pmu_write_counter, - .get_event_idx = xscale1pmu_get_event_idx, - .start = xscale1pmu_start, - .stop = xscale1pmu_stop, - .map_event = xscale_map_event, - .num_events = 3, - .max_period = (1LLU << 32) - 1, -}; - -static struct arm_pmu *__init xscale1pmu_init(void) +static int __devinit xscale1pmu_init(struct arm_pmu *cpu_pmu) { - return &xscale1pmu; + *cpu_pmu = (struct arm_pmu) { + .name = "xscale1", + .handle_irq = xscale1pmu_handle_irq, + .enable = xscale1pmu_enable_event, + .disable = xscale1pmu_disable_event, + .read_counter = xscale1pmu_read_counter, + .write_counter = xscale1pmu_write_counter, + .get_event_idx = xscale1pmu_get_event_idx, + .start = xscale1pmu_start, + .stop = xscale1pmu_stop, + .map_event = xscale_map_event, + .num_events = 3, + .max_period = (1LLU << 32) - 1, + }; + + return 0; } #define XSCALE2_OVERFLOWED_MASK 0x01f @@ -567,7 +573,8 @@ xscale2pmu_handle_irq(int irq_num, void *dev) { unsigned long pmnc, of_flags; struct perf_sample_data data; - struct pmu_hw_events *cpuc; + struct arm_pmu *cpu_pmu = (struct arm_pmu *)dev; + struct pmu_hw_events *cpuc = cpu_pmu->get_hw_events(); struct pt_regs *regs; int idx; @@ -585,7 +592,6 @@ xscale2pmu_handle_irq(int irq_num, void *dev) regs = get_irq_regs(); - cpuc = &__get_cpu_var(cpu_hw_events); for (idx = 0; idx < cpu_pmu->num_events; ++idx) { struct perf_event *event = cpuc->events[idx]; struct hw_perf_event *hwc; @@ -597,13 +603,13 @@ xscale2pmu_handle_irq(int irq_num, void *dev) continue; hwc = &event->hw; - armpmu_event_update(event, hwc, idx); + armpmu_event_update(event); perf_sample_data_init(&data, 0, hwc->last_period); - if (!armpmu_event_set_period(event, hwc, idx)) + if (!armpmu_event_set_period(event)) continue; if (perf_event_overflow(event, &data, regs)) - cpu_pmu->disable(hwc, idx); + cpu_pmu->disable(event); } irq_work_run(); @@ -617,11 +623,13 @@ xscale2pmu_handle_irq(int irq_num, void *dev) return IRQ_HANDLED; } -static void -xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx) +static void xscale2pmu_enable_event(struct perf_event *event) { unsigned long flags, ien, evtsel; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; ien = xscale2pmu_read_int_enable(); evtsel = xscale2pmu_read_event_select(); @@ -661,11 +669,13 @@ xscale2pmu_enable_event(struct hw_perf_event *hwc, int idx) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx) +static void xscale2pmu_disable_event(struct perf_event *event) { unsigned long flags, ien, evtsel, of_flags; + struct arm_pmu *cpu_pmu = to_arm_pmu(event->pmu); + struct hw_perf_event *hwc = &event->hw; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); + int idx = hwc->idx; ien = xscale2pmu_read_int_enable(); evtsel = xscale2pmu_read_event_select(); @@ -713,7 +723,7 @@ xscale2pmu_disable_event(struct hw_perf_event *hwc, int idx) static int xscale2pmu_get_event_idx(struct pmu_hw_events *cpuc, - struct hw_perf_event *event) + struct perf_event *event) { int idx = xscale1pmu_get_event_idx(cpuc, event); if (idx >= 0) @@ -727,8 +737,7 @@ out: return idx; } -static void -xscale2pmu_start(void) +static void xscale2pmu_start(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -740,8 +749,7 @@ xscale2pmu_start(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static void -xscale2pmu_stop(void) +static void xscale2pmu_stop(struct arm_pmu *cpu_pmu) { unsigned long flags, val; struct pmu_hw_events *events = cpu_pmu->get_hw_events(); @@ -753,9 +761,10 @@ xscale2pmu_stop(void) raw_spin_unlock_irqrestore(&events->pmu_lock, flags); } -static inline u32 -xscale2pmu_read_counter(int counter) +static inline u32 xscale2pmu_read_counter(struct perf_event *event) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; u32 val = 0; switch (counter) { @@ -779,9 +788,11 @@ xscale2pmu_read_counter(int counter) return val; } -static inline void -xscale2pmu_write_counter(int counter, u32 val) +static inline void xscale2pmu_write_counter(struct perf_event *event, u32 val) { + struct hw_perf_event *hwc = &event->hw; + int counter = hwc->idx; + switch (counter) { case XSCALE_CYCLE_COUNTER: asm volatile("mcr p14, 0, %0, c1, c1, 0" : : "r" (val)); @@ -801,33 +812,33 @@ xscale2pmu_write_counter(int counter, u32 val) } } -static struct arm_pmu xscale2pmu = { - .name = "xscale2", - .handle_irq = xscale2pmu_handle_irq, - .enable = xscale2pmu_enable_event, - .disable = xscale2pmu_disable_event, - .read_counter = xscale2pmu_read_counter, - .write_counter = xscale2pmu_write_counter, - .get_event_idx = xscale2pmu_get_event_idx, - .start = xscale2pmu_start, - .stop = xscale2pmu_stop, - .map_event = xscale_map_event, - .num_events = 5, - .max_period = (1LLU << 32) - 1, -}; - -static struct arm_pmu *__init xscale2pmu_init(void) +static int __devinit xscale2pmu_init(struct arm_pmu *cpu_pmu) { - return &xscale2pmu; + *cpu_pmu = (struct arm_pmu) { + .name = "xscale2", + .handle_irq = xscale2pmu_handle_irq, + .enable = xscale2pmu_enable_event, + .disable = xscale2pmu_disable_event, + .read_counter = xscale2pmu_read_counter, + .write_counter = xscale2pmu_write_counter, + .get_event_idx = xscale2pmu_get_event_idx, + .start = xscale2pmu_start, + .stop = xscale2pmu_stop, + .map_event = xscale_map_event, + .num_events = 5, + .max_period = (1LLU << 32) - 1, + }; + + return 0; } #else -static struct arm_pmu *__init xscale1pmu_init(void) +static inline int xscale1pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } -static struct arm_pmu *__init xscale2pmu_init(void) +static inline int xscale2pmu_init(struct arm_pmu *cpu_pmu) { - return NULL; + return -ENODEV; } #endif /* CONFIG_CPU_XSCALE */ diff --git a/arch/arm/kernel/pmu.c b/arch/arm/kernel/pmu.c deleted file mode 100644 index 2334bf8a650a..000000000000 --- a/arch/arm/kernel/pmu.c +++ /dev/null @@ -1,36 +0,0 @@ -/* - * linux/arch/arm/kernel/pmu.c - * - * Copyright (C) 2009 picoChip Designs Ltd, Jamie Iles - * Copyright (C) 2010 ARM Ltd, Will Deacon - * - * 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/err.h> -#include <linux/kernel.h> -#include <linux/module.h> - -#include <asm/pmu.h> - -/* - * PMU locking to ensure mutual exclusion between different subsystems. - */ -static unsigned long pmu_lock[BITS_TO_LONGS(ARM_NUM_PMU_DEVICES)]; - -int -reserve_pmu(enum arm_pmu_type type) -{ - return test_and_set_bit_lock(type, pmu_lock) ? -EBUSY : 0; -} -EXPORT_SYMBOL_GPL(reserve_pmu); - -void -release_pmu(enum arm_pmu_type type) -{ - clear_bit_unlock(type, pmu_lock); -} -EXPORT_SYMBOL_GPL(release_pmu); diff --git a/arch/arm/kernel/process.c b/arch/arm/kernel/process.c index 693b744fd572..4fff006675a0 100644 --- a/arch/arm/kernel/process.c +++ b/arch/arm/kernel/process.c @@ -31,9 +31,9 @@ #include <linux/random.h> #include <linux/hw_breakpoint.h> #include <linux/cpuidle.h> +#include <linux/console.h> #include <asm/cacheflush.h> -#include <asm/leds.h> #include <asm/processor.h> #include <asm/thread_notify.h> #include <asm/stacktrace.h> @@ -60,6 +60,18 @@ extern void setup_mm_for_reboot(void); static volatile int hlt_counter; +#ifdef CONFIG_SMP +void arch_trigger_all_cpu_backtrace(void) +{ + smp_send_all_cpu_backtrace(); +} +#else +void arch_trigger_all_cpu_backtrace(void) +{ + dump_stack(); +} +#endif + void disable_hlt(void) { hlt_counter++; @@ -92,6 +104,31 @@ __setup("hlt", hlt_setup); extern void call_with_stack(void (*fn)(void *), void *arg, void *sp); typedef void (*phys_reset_t)(unsigned long); +#ifdef CONFIG_ARM_FLUSH_CONSOLE_ON_RESTART +void arm_machine_flush_console(void) +{ + printk("\n"); + pr_emerg("Restarting %s\n", linux_banner); + if (console_trylock()) { + console_unlock(); + return; + } + + mdelay(50); + + local_irq_disable(); + if (!console_trylock()) + pr_emerg("arm_restart: Console was locked! Busting\n"); + else + pr_emerg("arm_restart: Console was locked!\n"); + console_unlock(); +} +#else +void arm_machine_flush_console(void) +{ +} +#endif + /* * A temporary stack to use for CPU reset. This is static so that we * don't clobber it with the identity mapping. When running with this @@ -187,9 +224,9 @@ void cpu_idle(void) /* endless idle loop with no priority at all */ while (1) { + idle_notifier_call_chain(IDLE_START); tick_nohz_idle_enter(); rcu_idle_enter(); - leds_event(led_idle_start); while (!need_resched()) { #ifdef CONFIG_HOTPLUG_CPU if (cpu_is_offline(smp_processor_id())) @@ -220,9 +257,9 @@ void cpu_idle(void) } else local_irq_enable(); } - leds_event(led_idle_end); rcu_idle_exit(); tick_nohz_idle_exit(); + idle_notifier_call_chain(IDLE_END); schedule_preempt_disabled(); } } @@ -262,6 +299,10 @@ void machine_restart(char *cmd) { machine_shutdown(); + /* Flush the console to make sure all the relevant messages make it + * out to the console drivers */ + arm_machine_flush_console(); + arm_pm_restart(reboot_mode, cmd); /* Give a grace period for failure to restart of 1s */ @@ -273,6 +314,77 @@ void machine_restart(char *cmd) while (1); } +/* + * dump a block of kernel memory from around the given address + */ +static void show_data(unsigned long addr, int nbytes, const char *name) +{ + int i, j; + int nlines; + u32 *p; + + /* + * don't attempt to dump non-kernel addresses or + * values that are probably just small negative numbers + */ + if (addr < PAGE_OFFSET || addr > -256UL) + return; + + printk("\n%s: %#lx:\n", name, addr); + + /* + * round address down to a 32 bit boundary + * and always dump a multiple of 32 bytes + */ + p = (u32 *)(addr & ~(sizeof(u32) - 1)); + nbytes += (addr & (sizeof(u32) - 1)); + nlines = (nbytes + 31) / 32; + + + for (i = 0; i < nlines; i++) { + /* + * just display low 16 bits of address to keep + * each line of the dump < 80 characters + */ + printk("%04lx ", (unsigned long)p & 0xffff); + for (j = 0; j < 8; j++) { + u32 data; + if (probe_kernel_address(p, data)) { + printk(" ********"); + } else { + printk(" %08x", data); + } + ++p; + } + printk("\n"); + } +} + +static void show_extra_register_data(struct pt_regs *regs, int nbytes) +{ + mm_segment_t fs; + + fs = get_fs(); + set_fs(KERNEL_DS); + show_data(regs->ARM_pc - nbytes, nbytes * 2, "PC"); + show_data(regs->ARM_lr - nbytes, nbytes * 2, "LR"); + show_data(regs->ARM_sp - nbytes, nbytes * 2, "SP"); + show_data(regs->ARM_ip - nbytes, nbytes * 2, "IP"); + show_data(regs->ARM_fp - nbytes, nbytes * 2, "FP"); + show_data(regs->ARM_r0 - nbytes, nbytes * 2, "R0"); + show_data(regs->ARM_r1 - nbytes, nbytes * 2, "R1"); + show_data(regs->ARM_r2 - nbytes, nbytes * 2, "R2"); + show_data(regs->ARM_r3 - nbytes, nbytes * 2, "R3"); + show_data(regs->ARM_r4 - nbytes, nbytes * 2, "R4"); + show_data(regs->ARM_r5 - nbytes, nbytes * 2, "R5"); + show_data(regs->ARM_r6 - nbytes, nbytes * 2, "R6"); + show_data(regs->ARM_r7 - nbytes, nbytes * 2, "R7"); + show_data(regs->ARM_r8 - nbytes, nbytes * 2, "R8"); + show_data(regs->ARM_r9 - nbytes, nbytes * 2, "R9"); + show_data(regs->ARM_r10 - nbytes, nbytes * 2, "R10"); + set_fs(fs); +} + void __show_regs(struct pt_regs *regs) { unsigned long flags; @@ -332,6 +444,8 @@ void __show_regs(struct pt_regs *regs) printk("Control: %08x%s\n", ctrl, buf); } #endif + + show_extra_register_data(regs, 128); } void show_regs(struct pt_regs * regs) diff --git a/arch/arm/kernel/smp.c b/arch/arm/kernel/smp.c index 0181f7efc53b..14ea5c52e0c4 100644 --- a/arch/arm/kernel/smp.c +++ b/arch/arm/kernel/smp.c @@ -56,6 +56,7 @@ enum ipi_msg_type { IPI_CALL_FUNC, IPI_CALL_FUNC_SINGLE, IPI_CPU_STOP, + IPI_CPU_BACKTRACE, }; static DECLARE_COMPLETION(cpu_running); @@ -359,6 +360,7 @@ static const char *ipi_types[NR_IPI] = { S(IPI_CALL_FUNC, "Function call interrupts"), S(IPI_CALL_FUNC_SINGLE, "Single function call interrupts"), S(IPI_CPU_STOP, "CPU stop interrupts"), + S(IPI_CPU_BACKTRACE, "CPU backtrace"), }; void show_ipi_list(struct seq_file *p, int prec) @@ -493,6 +495,58 @@ static void ipi_cpu_stop(unsigned int cpu) cpu_relax(); } +static cpumask_t backtrace_mask; +static DEFINE_RAW_SPINLOCK(backtrace_lock); + +/* "in progress" flag of arch_trigger_all_cpu_backtrace */ +static unsigned long backtrace_flag; + +void smp_send_all_cpu_backtrace(void) +{ + unsigned int this_cpu = smp_processor_id(); + int i; + + if (test_and_set_bit(0, &backtrace_flag)) + /* + * If there is already a trigger_all_cpu_backtrace() in progress + * (backtrace_flag == 1), don't output double cpu dump infos. + */ + return; + + cpumask_copy(&backtrace_mask, cpu_online_mask); + cpu_clear(this_cpu, backtrace_mask); + + pr_info("Backtrace for cpu %d (current):\n", this_cpu); + dump_stack(); + + pr_info("\nsending IPI to all other CPUs:\n"); + smp_cross_call(&backtrace_mask, IPI_CPU_BACKTRACE); + + /* Wait for up to 10 seconds for all other CPUs to do the backtrace */ + for (i = 0; i < 10 * 1000; i++) { + if (cpumask_empty(&backtrace_mask)) + break; + mdelay(1); + } + + clear_bit(0, &backtrace_flag); + smp_mb__after_clear_bit(); +} + +/* + * ipi_cpu_backtrace - handle IPI from smp_send_all_cpu_backtrace() + */ +static void ipi_cpu_backtrace(unsigned int cpu, struct pt_regs *regs) +{ + if (cpu_isset(cpu, backtrace_mask)) { + raw_spin_lock(&backtrace_lock); + pr_warning("IPI backtrace for cpu %d\n", cpu); + show_regs(regs); + raw_spin_unlock(&backtrace_lock); + cpu_clear(cpu, backtrace_mask); + } +} + /* * Main handler for inter-processor interrupts */ @@ -538,6 +592,10 @@ void handle_IPI(int ipinr, struct pt_regs *regs) irq_exit(); break; + case IPI_CPU_BACKTRACE: + ipi_cpu_backtrace(cpu, regs); + break; + default: printk(KERN_CRIT "CPU%u: Unknown IPI message 0x%x\n", cpu, ipinr); diff --git a/arch/arm/kernel/smp_twd.c b/arch/arm/kernel/smp_twd.c index e1f906989bb8..fd6ee41ad56e 100644 --- a/arch/arm/kernel/smp_twd.c +++ b/arch/arm/kernel/smp_twd.c @@ -357,6 +357,7 @@ const static struct of_device_id twd_of_match[] __initconst = { { .compatible = "arm,cortex-a9-twd-timer", }, { .compatible = "arm,cortex-a5-twd-timer", }, { .compatible = "arm,arm11mp-twd-timer", }, + { .compatible = "arm,smp-twd", }, { }, }; diff --git a/arch/arm/kernel/topology.c b/arch/arm/kernel/topology.c index 26c12c6440fc..0b512331c0f5 100644 --- a/arch/arm/kernel/topology.c +++ b/arch/arm/kernel/topology.c @@ -317,6 +317,126 @@ void store_cpu_topology(unsigned int cpuid) cpu_topology[cpuid].socket_id, mpidr); } + +#ifdef CONFIG_SCHED_HMP + +static const char * const little_cores[] = { + "arm,cortex-a7", + NULL, +}; + +static bool is_little_cpu(struct device_node *cn) +{ + const char * const *lc; + for (lc = little_cores; *lc; lc++) + if (of_device_is_compatible(cn, *lc)) + return true; + return false; +} + +void __init arch_get_fast_and_slow_cpus(struct cpumask *fast, + struct cpumask *slow) +{ + struct device_node *cn = NULL; + int cpu = 0; + + cpumask_clear(fast); + cpumask_clear(slow); + + /* + * Use the config options if they are given. This helps testing + * HMP scheduling on systems without a big.LITTLE architecture. + */ + if (strlen(CONFIG_HMP_FAST_CPU_MASK) && strlen(CONFIG_HMP_SLOW_CPU_MASK)) { + if (cpulist_parse(CONFIG_HMP_FAST_CPU_MASK, fast)) + WARN(1, "Failed to parse HMP fast cpu mask!\n"); + if (cpulist_parse(CONFIG_HMP_SLOW_CPU_MASK, slow)) + WARN(1, "Failed to parse HMP slow cpu mask!\n"); + return; + } + + /* + * Else, parse device tree for little cores. + */ + while ((cn = of_find_node_by_type(cn, "cpu"))) { + + if (cpu >= num_possible_cpus()) + break; + + if (is_little_cpu(cn)) + cpumask_set_cpu(cpu, slow); + else + cpumask_set_cpu(cpu, fast); + + cpu++; + } + + if (!cpumask_empty(fast) && !cpumask_empty(slow)) + return; + + /* + * We didn't find both big and little cores so let's call all cores + * fast as this will keep the system running, with all cores being + * treated equal. + */ + cpumask_setall(fast); + cpumask_clear(slow); +} + +void __init arch_get_hmp_domains(struct list_head *hmp_domains_list) +{ + struct cpumask hmp_fast_cpu_mask; + struct cpumask hmp_slow_cpu_mask; + struct hmp_domain *domain; + + arch_get_fast_and_slow_cpus(&hmp_fast_cpu_mask, &hmp_slow_cpu_mask); + + /* + * Initialize hmp_domains + * Must be ordered with respect to compute capacity. + * Fastest domain at head of list. + */ + if(!cpumask_empty(&hmp_slow_cpu_mask)) { + domain = (struct hmp_domain *) + kmalloc(sizeof(struct hmp_domain), GFP_KERNEL); + cpumask_copy(&domain->cpus, &hmp_slow_cpu_mask); + list_add(&domain->hmp_domains, hmp_domains_list); + } + domain = (struct hmp_domain *) + kmalloc(sizeof(struct hmp_domain), GFP_KERNEL); + cpumask_copy(&domain->cpus, &hmp_fast_cpu_mask); + list_add(&domain->hmp_domains, hmp_domains_list); +} +#endif /* CONFIG_SCHED_HMP */ + + +/* + * cluster_to_logical_mask - return cpu logical mask of CPUs in a cluster + * @socket_id: cluster HW identifier + * @cluster_mask: the cpumask location to be initialized, modified by the + * function only if return value == 0 + * + * Return: + * + * 0 on success + * -EINVAL if cluster_mask is NULL or there is no record matching socket_id + */ +int cluster_to_logical_mask(unsigned int socket_id, cpumask_t *cluster_mask) +{ + int cpu; + + if (!cluster_mask) + return -EINVAL; + + for_each_online_cpu(cpu) + if (socket_id == topology_physical_package_id(cpu)) { + cpumask_copy(cluster_mask, topology_core_cpumask(cpu)); + return 0; + } + + return -EINVAL; +} + /* * init_cpu_topology is called at boot when only one cpu is running * which prevent simultaneous write access to cpu_topology array diff --git a/arch/arm/mach-bcmring/arch.c b/arch/arm/mach-bcmring/arch.c index 45c97b1ee9b1..76e795323569 100644 --- a/arch/arm/mach-bcmring/arch.c +++ b/arch/arm/mach-bcmring/arch.c @@ -29,7 +29,6 @@ #include <asm/setup.h> #include <asm/mach-types.h> #include <asm/mach/time.h> -#include <asm/pmu.h> #include <asm/mach/arch.h> #include <mach/dma.h> @@ -116,7 +115,7 @@ static struct resource pmu_resource = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .resource = &pmu_resource, .num_resources = 1, }; diff --git a/arch/arm/mach-highbank/Makefile b/arch/arm/mach-highbank/Makefile index 3ec8bdd25d09..b8947086cf70 100644 --- a/arch/arm/mach-highbank/Makefile +++ b/arch/arm/mach-highbank/Makefile @@ -7,3 +7,5 @@ obj-$(CONFIG_DEBUG_HIGHBANK_UART) += lluart.o obj-$(CONFIG_SMP) += platsmp.o obj-$(CONFIG_HOTPLUG_CPU) += hotplug.o obj-$(CONFIG_PM_SLEEP) += pm.o + +obj-y += pl320-ipc.o diff --git a/arch/arm/mach-highbank/highbank.c b/arch/arm/mach-highbank/highbank.c index d75b0a78d88a..4ed8211b7fd9 100644 --- a/arch/arm/mach-highbank/highbank.c +++ b/arch/arm/mach-highbank/highbank.c @@ -15,6 +15,7 @@ */ #include <linux/clk.h> #include <linux/clkdev.h> +#include <linux/input.h> #include <linux/io.h> #include <linux/irq.h> #include <linux/irqdomain.h> @@ -22,6 +23,7 @@ #include <linux/of_irq.h> #include <linux/of_platform.h> #include <linux/of_address.h> +#include <linux/reboot.h> #include <linux/smp.h> #include <asm/cacheflush.h> @@ -35,6 +37,7 @@ #include <asm/mach/arch.h> #include <asm/mach/map.h> #include <asm/mach/time.h> +#include <mach/pl320-ipc.h> #include "core.h" #include "sysregs.h" @@ -149,9 +152,28 @@ static void highbank_power_off(void) cpu_do_idle(); } +static int hb_keys_notifier(struct notifier_block *nb, unsigned long event, void *data) +{ + u32 key = *(u32 *)data; + + if (event != 0x1000) + return 0; + + if (key == KEY_POWER) + orderly_poweroff(false); + else if (key == 0xffff) + ctrl_alt_del(); + + return 0; +} +static struct notifier_block hb_keys_nb = { + .notifier_call = hb_keys_notifier, +}; + static void __init highbank_init(void) { pm_power_off = highbank_power_off; + pl320_ipc_register_notifier(&hb_keys_nb); of_platform_populate(NULL, of_default_bus_match_table, NULL, NULL); } diff --git a/arch/arm/mach-highbank/include/mach/pl320-ipc.h b/arch/arm/mach-highbank/include/mach/pl320-ipc.h new file mode 100644 index 000000000000..a0e58eed4b0d --- /dev/null +++ b/arch/arm/mach-highbank/include/mach/pl320-ipc.h @@ -0,0 +1,20 @@ +/* + * Copyright 2010 Calxeda, Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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/>. + */ +int ipc_call_fast(u32 *data); +int ipc_call_slow(u32 *data); + +extern int pl320_ipc_register_notifier(struct notifier_block *nb); +extern int pl320_ipc_unregister_notifier(struct notifier_block *nb); diff --git a/arch/arm/mach-highbank/pl320-ipc.c b/arch/arm/mach-highbank/pl320-ipc.c new file mode 100644 index 000000000000..b68d85663954 --- /dev/null +++ b/arch/arm/mach-highbank/pl320-ipc.c @@ -0,0 +1,308 @@ +/* + * Copyright 2010 Calxeda, Inc. + * + * This program is free software; you can redistribute it and/or modify it + * under the terms and conditions of the GNU General Public License, + * version 2, as published by the Free Software Foundation. + * + * This program is distributed in the hope 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/types.h> +#include <linux/err.h> +#include <linux/delay.h> +#include <linux/export.h> +#include <linux/io.h> +#include <linux/interrupt.h> +#include <linux/completion.h> +#include <linux/mutex.h> +#include <linux/notifier.h> +#include <linux/spinlock.h> +#include <linux/device.h> +#include <linux/amba/bus.h> + +#include <mach/pl320-ipc.h> + +#define IPCMxSOURCE(m) ((m) * 0x40) +#define IPCMxDSET(m) (((m) * 0x40) + 0x004) +#define IPCMxDCLEAR(m) (((m) * 0x40) + 0x008) +#define IPCMxDSTATUS(m) (((m) * 0x40) + 0x00C) +#define IPCMxMODE(m) (((m) * 0x40) + 0x010) +#define IPCMxMSET(m) (((m) * 0x40) + 0x014) +#define IPCMxMCLEAR(m) (((m) * 0x40) + 0x018) +#define IPCMxMSTATUS(m) (((m) * 0x40) + 0x01C) +#define IPCMxSEND(m) (((m) * 0x40) + 0x020) +#define IPCMxDR(m, dr) (((m) * 0x40) + ((dr) * 4) + 0x024) + +#define IPCMMIS(irq) (((irq) * 8) + 0x800) +#define IPCMRIS(irq) (((irq) * 8) + 0x804) + +#define MBOX_MASK(n) (1 << (n)) +#define IPC_FAST_MBOX 0 +#define IPC_SLOW_MBOX 1 +#define IPC_RX_MBOX 2 + +#define CHAN_MASK(n) (1 << (n)) +#define A9_SOURCE 1 +#define M3_SOURCE 0 + +static void __iomem *ipc_base; +static int ipc_irq; +static DEFINE_SPINLOCK(ipc_m0_lock); +static DEFINE_MUTEX(ipc_m1_lock); +static DECLARE_COMPLETION(ipc_completion); +static ATOMIC_NOTIFIER_HEAD(ipc_notifier); + +static inline void set_destination(int source, int mbox) +{ + __raw_writel(CHAN_MASK(source), ipc_base + IPCMxDSET(mbox)); + __raw_writel(CHAN_MASK(source), ipc_base + IPCMxMSET(mbox)); +} + +static inline void clear_destination(int source, int mbox) +{ + __raw_writel(CHAN_MASK(source), ipc_base + IPCMxDCLEAR(mbox)); + __raw_writel(CHAN_MASK(source), ipc_base + IPCMxMCLEAR(mbox)); +} + +static void __ipc_send(int mbox, u32 *data) +{ + int i; + for (i = 0; i < 7; i++) + __raw_writel(data[i], ipc_base + IPCMxDR(mbox, i)); + __raw_writel(0x1, ipc_base + IPCMxSEND(mbox)); +} + +static u32 __ipc_rcv(int mbox, u32 *data) +{ + int i; + for (i = 0; i < 7; i++) + data[i] = __raw_readl(ipc_base + IPCMxDR(mbox, i)); + return data[1]; +} + +/* non-blocking implementation from the A9 side, interrupt safe in theory */ +int ipc_call_fast(u32 *data) +{ + int timeout, ret; + + spin_lock(&ipc_m0_lock); + + __ipc_send(IPC_FAST_MBOX, data); + + for (timeout = 5000; timeout > 0; timeout--) { + if (__raw_readl(ipc_base + IPCMxSEND(IPC_FAST_MBOX)) == 0x2) + break; + udelay(100); + } + if (timeout == 0) { + ret = -ETIMEDOUT; + goto out; + } + + ret = __ipc_rcv(IPC_FAST_MBOX, data); +out: + __raw_writel(0, ipc_base + IPCMxSEND(IPC_FAST_MBOX)); + spin_unlock(&ipc_m0_lock); + return ret; +} +EXPORT_SYMBOL(ipc_call_fast); + +/* blocking implmentation from the A9 side, not usuable in interrupts! */ +int ipc_call_slow(u32 *data) +{ + int ret; + + mutex_lock(&ipc_m1_lock); + + init_completion(&ipc_completion); + __ipc_send(IPC_SLOW_MBOX, data); + ret = wait_for_completion_timeout(&ipc_completion, + msecs_to_jiffies(1000)); + if (ret == 0) + goto out; + + ret = __ipc_rcv(IPC_SLOW_MBOX, data); +out: + mutex_unlock(&ipc_m1_lock); + return ret; +} +EXPORT_SYMBOL(ipc_call_slow); + +irqreturn_t ipc_handler(int irq, void *dev) +{ + u32 irq_stat; + u32 data[7]; + + irq_stat = __raw_readl(ipc_base + IPCMMIS(1)); + if (irq_stat & MBOX_MASK(IPC_SLOW_MBOX)) { + __raw_writel(0, ipc_base + IPCMxSEND(IPC_SLOW_MBOX)); + complete(&ipc_completion); + } + if (irq_stat & MBOX_MASK(IPC_RX_MBOX)) { + __ipc_rcv(IPC_RX_MBOX, data); + atomic_notifier_call_chain(&ipc_notifier, data[0], data + 1); + __raw_writel(2, ipc_base + IPCMxSEND(IPC_RX_MBOX)); + } + + return IRQ_HANDLED; +} + +int pl320_ipc_register_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_register(&ipc_notifier, nb); +} + +int pl320_ipc_unregister_notifier(struct notifier_block *nb) +{ + return atomic_notifier_chain_unregister(&ipc_notifier, nb); +} + +static int __devinit pl320_probe(struct amba_device *adev, const struct amba_id *id) +{ + int ret; + + ipc_base = ioremap(adev->res.start, resource_size(&adev->res)); + if (ipc_base == NULL) + return -ENOMEM; + + __raw_writel(0, ipc_base + IPCMxSEND(IPC_FAST_MBOX)); + __raw_writel(0, ipc_base + IPCMxSEND(IPC_SLOW_MBOX)); + + ipc_irq = adev->irq[0]; + ret = request_irq(ipc_irq, ipc_handler, 0, dev_name(&adev->dev), NULL); + if (ret < 0) + goto err; + + /* Init fast mailbox */ + __raw_writel(CHAN_MASK(A9_SOURCE), ipc_base + IPCMxSOURCE(IPC_FAST_MBOX)); + set_destination(M3_SOURCE, IPC_FAST_MBOX); + + /* Init slow mailbox */ + __raw_writel(CHAN_MASK(A9_SOURCE), ipc_base + IPCMxSOURCE(IPC_SLOW_MBOX)); + __raw_writel(CHAN_MASK(M3_SOURCE), ipc_base + IPCMxDSET(IPC_SLOW_MBOX)); + __raw_writel(CHAN_MASK(M3_SOURCE) | CHAN_MASK(A9_SOURCE), + ipc_base + IPCMxMSET(IPC_SLOW_MBOX)); + + /* Init receive mailbox */ + __raw_writel(CHAN_MASK(M3_SOURCE), ipc_base + IPCMxSOURCE(IPC_RX_MBOX)); + __raw_writel(CHAN_MASK(A9_SOURCE), ipc_base + IPCMxDSET(IPC_RX_MBOX)); + __raw_writel(CHAN_MASK(M3_SOURCE) | CHAN_MASK(A9_SOURCE), + ipc_base + IPCMxMSET(IPC_RX_MBOX)); + + return 0; +err: + iounmap(ipc_base); + return ret; +} + +static struct amba_id pl320_ids[] = { + { + .id = 0x00041320, + .mask = 0x000fffff, + }, + { 0, 0 }, +}; + +static struct amba_driver pl320_driver = { + .drv = { + .name = "pl320", + }, + .id_table = pl320_ids, + .probe = pl320_probe, +}; + +static int __init ipc_init(void) +{ + return amba_driver_register(&pl320_driver); +} +module_init(ipc_init); + +irqreturn_t ipc_test_handler(int irq, void *dev) +{ + u32 irq_stat; + + irq_stat = __raw_readl(ipc_base + IPCMMIS(irq - (ipc_irq - 1))); + if (irq_stat & MBOX_MASK(IPC_FAST_MBOX)) { + if ((__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 0)) == 0x900dbeef) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 1)) == 1) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 2)) == 2) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 3)) == 3) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 4)) == 4) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 5)) == 5) && + (__raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 6)) == 6)) { + printk(KERN_ERR "ipc fast mbox message %X received\n", __raw_readl(ipc_base + IPCMxDR(IPC_FAST_MBOX, 0))); + __raw_writel(0xBADBEEF, ipc_base + IPCMxDR(IPC_FAST_MBOX, 1)); + } + __raw_writel(0x2, ipc_base + IPCMxSEND(IPC_FAST_MBOX)); + } + if (irq_stat & MBOX_MASK(IPC_SLOW_MBOX)) { + if ((__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 0)) == 0x12345678) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 1)) == 6) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 2)) == 5) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 3)) == 4) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 4)) == 3) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 5)) == 2) && + (__raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 6)) == 1)) { + printk("slow mbox message %X received\n", __raw_readl(ipc_base + IPCMxDR(IPC_SLOW_MBOX, 0))); + __raw_writel(0x87654321, ipc_base + IPCMxDR(IPC_SLOW_MBOX, 1)); + } + __raw_writel(0x2, ipc_base + IPCMxSEND(IPC_SLOW_MBOX)); + } + if (irq_stat & MBOX_MASK(IPC_RX_MBOX)) { + __raw_writel(0, ipc_base + IPCMxSEND(IPC_RX_MBOX)); + // handle events + } + + return IRQ_HANDLED; +} + +static void __init ipc_test(void) +{ + int ret, i; + + printk("ipc test start\n"); + + for (i = 0; i < 8; i++) { + u32 data[7]; + int j; + if (i == 1) continue; + + if (request_irq(ipc_irq - 1 + i, ipc_test_handler, 0, "ipc", NULL) < 0) { + printk("ipc - request_irq failed - FAIL\n"); + return; + } + + set_destination(i, IPC_FAST_MBOX); + set_destination(i, IPC_SLOW_MBOX); + + for (j = 1; j < 7; j++) + data[j] = j; + data[0] = 0x900dbeef; + ret = ipc_call_fast(data); + if (ret == 0xbadbeef) + printk(KERN_ERR "ipc %d fast call - PASS\n", i); + else + printk(KERN_ERR "ipc %d fast call fail %d\n", i, ret); + for (j = 1; j < 7; j++) + data[j] = 7 - j; + data[0] = 0x12345678; + ret = ipc_call_slow(data); + if (ret == 0x87654321) + printk("ipc %d slow call - PASS\n", i); + + clear_destination(i, IPC_FAST_MBOX); + clear_destination(i, IPC_SLOW_MBOX); + + free_irq(ipc_irq - 1 + i, NULL); + } + set_destination(M3_SOURCE, IPC_FAST_MBOX); + set_destination(M3_SOURCE, IPC_SLOW_MBOX); +} +//late_initcall(ipc_test); + diff --git a/arch/arm/mach-highbank/sysregs.h b/arch/arm/mach-highbank/sysregs.h index 0e913389f445..6b10548801e8 100644 --- a/arch/arm/mach-highbank/sysregs.h +++ b/arch/arm/mach-highbank/sysregs.h @@ -31,22 +31,22 @@ extern void __iomem *sregs_base; static inline void hignbank_set_pwr_suspend(void) { - writel(HB_PWR_SUSPEND, sregs_base + HB_SREG_A9_PWR_REQ); + writel_relaxed(HB_PWR_SUSPEND, sregs_base + HB_SREG_A9_PWR_REQ); } static inline void hignbank_set_pwr_shutdown(void) { - writel(HB_PWR_SHUTDOWN, sregs_base + HB_SREG_A9_PWR_REQ); + writel_relaxed(HB_PWR_SHUTDOWN, sregs_base + HB_SREG_A9_PWR_REQ); } static inline void hignbank_set_pwr_soft_reset(void) { - writel(HB_PWR_SOFT_RESET, sregs_base + HB_SREG_A9_PWR_REQ); + writel_relaxed(HB_PWR_SOFT_RESET, sregs_base + HB_SREG_A9_PWR_REQ); } static inline void hignbank_set_pwr_hard_reset(void) { - writel(HB_PWR_HARD_RESET, sregs_base + HB_SREG_A9_PWR_REQ); + writel_relaxed(HB_PWR_HARD_RESET, sregs_base + HB_SREG_A9_PWR_REQ); } #endif diff --git a/arch/arm/mach-imx/Makefile.boot b/arch/arm/mach-imx/Makefile.boot index 05541cf4a878..2e63463178b7 100644 --- a/arch/arm/mach-imx/Makefile.boot +++ b/arch/arm/mach-imx/Makefile.boot @@ -30,10 +30,16 @@ zreladdr-$(CONFIG_SOC_IMX51) += 0x90008000 params_phys-$(CONFIG_SOC_IMX51) := 0x90000100 initrd_phys-$(CONFIG_SOC_IMX51) := 0x90800000 +dtb-$(CONFIG_MACH_MX51_BABBAGE) += mx51-babbage.dtb +dtb-$(CONFIG_MACH_MX51_EFIKAMX) += genesi-efikamx.dtb +dtb-$(CONFIG_MACH_MX51_EFIKASB) += genesi-efikasb.dtb + zreladdr-$(CONFIG_SOC_IMX53) += 0x70008000 params_phys-$(CONFIG_SOC_IMX53) := 0x70000100 initrd_phys-$(CONFIG_SOC_IMX53) := 0x70800000 +dtb-$(CONFIG_MACH_MX53_LOCO) += mx53-loco.dtb + zreladdr-$(CONFIG_SOC_IMX6Q) += 0x10008000 params_phys-$(CONFIG_SOC_IMX6Q) := 0x10000100 initrd_phys-$(CONFIG_SOC_IMX6Q) := 0x10800000 diff --git a/arch/arm/mach-imx/mach-mx51_babbage.c b/arch/arm/mach-imx/mach-mx51_babbage.c index 7b31cbde8775..87818d13f0e4 100644 --- a/arch/arm/mach-imx/mach-mx51_babbage.c +++ b/arch/arm/mach-imx/mach-mx51_babbage.c @@ -423,6 +423,11 @@ static struct sys_timer mx51_babbage_timer = { .init = mx51_babbage_timer_init, }; +static const char *mx51_babbage_dt_match[] __initdata = { + "fsl,mx51-babbage", + NULL +}; + MACHINE_START(MX51_BABBAGE, "Freescale MX51 Babbage Board") /* Maintainer: Amit Kucheria <amit.kucheria@canonical.com> */ .atag_offset = 0x100, @@ -434,4 +439,5 @@ MACHINE_START(MX51_BABBAGE, "Freescale MX51 Babbage Board") .init_machine = mx51_babbage_init, .init_late = imx51_init_late, .restart = mxc_restart, + .dt_compat = mx51_babbage_dt_match, MACHINE_END diff --git a/arch/arm/mach-imx/mach-mx51_efikamx.c b/arch/arm/mach-imx/mach-mx51_efikamx.c index 8d09c0126cab..928c98ecda39 100644 --- a/arch/arm/mach-imx/mach-mx51_efikamx.c +++ b/arch/arm/mach-imx/mach-mx51_efikamx.c @@ -287,7 +287,13 @@ static struct sys_timer mx51_efikamx_timer = { .init = mx51_efikamx_timer_init, }; +static const char *mx51_efikamx_dt_match[] __initdata = { + "genesi,efikamx", + NULL +}; + MACHINE_START(MX51_EFIKAMX, "Genesi Efika MX (Smarttop)") + /* Maintainer: Amit Kucheria <amit.kucheria@linaro.org> */ .atag_offset = 0x100, .map_io = mx51_map_io, .init_early = imx51_init_early, @@ -297,4 +303,5 @@ MACHINE_START(MX51_EFIKAMX, "Genesi Efika MX (Smarttop)") .init_machine = mx51_efikamx_init, .init_late = mx51_efikamx_init_late, .restart = mx51_efikamx_restart, + .dt_compat = mx51_efikamx_dt_match, MACHINE_END diff --git a/arch/arm/mach-imx/mach-mx51_efikasb.c b/arch/arm/mach-imx/mach-mx51_efikasb.c index fdbd181b97ef..37d52e00e434 100644 --- a/arch/arm/mach-imx/mach-mx51_efikasb.c +++ b/arch/arm/mach-imx/mach-mx51_efikasb.c @@ -283,6 +283,11 @@ static struct sys_timer mx51_efikasb_timer = { .init = mx51_efikasb_timer_init, }; +static const char *mx51_efikasb_dt_match[] __initdata = { + "genesi,efikasb", + NULL +}; + MACHINE_START(MX51_EFIKASB, "Genesi Efika MX (Smartbook)") .atag_offset = 0x100, .map_io = mx51_map_io, @@ -293,4 +298,5 @@ MACHINE_START(MX51_EFIKASB, "Genesi Efika MX (Smartbook)") .init_late = mx51_efikasb_init_late, .timer = &mx51_efikasb_timer, .restart = mxc_restart, + .dt_compat = mx51_efikasb_dt_match, MACHINE_END diff --git a/arch/arm/mach-imx/mach-mx53_loco.c b/arch/arm/mach-imx/mach-mx53_loco.c index 8abe23c1d3c8..67f5fc97d2a1 100644 --- a/arch/arm/mach-imx/mach-mx53_loco.c +++ b/arch/arm/mach-imx/mach-mx53_loco.c @@ -309,6 +309,10 @@ static struct sys_timer mx53_loco_timer = { .init = mx53_loco_timer_init, }; +static const char *mx53_loco_dt_match[] __initdata = { + "fsl,mx53-loco", + NULL +}; MACHINE_START(MX53_LOCO, "Freescale MX53 LOCO Board") .map_io = mx53_map_io, .init_early = imx53_init_early, @@ -318,4 +322,5 @@ MACHINE_START(MX53_LOCO, "Freescale MX53 LOCO Board") .init_machine = mx53_loco_board_init, .init_late = imx53_init_late, .restart = mxc_restart, + .dt_compat = mx53_loco_dt_match, MACHINE_END diff --git a/arch/arm/mach-omap2/Makefile.boot b/arch/arm/mach-omap2/Makefile.boot index b03e562acc60..eab4c41e41bd 100644 --- a/arch/arm/mach-omap2/Makefile.boot +++ b/arch/arm/mach-omap2/Makefile.boot @@ -1,3 +1,9 @@ zreladdr-y += 0x80008000 params_phys-y := 0x80000100 initrd_phys-y := 0x80800000 + +dtb-$(CONFIG_MACH_OMAP3_BEAGLE) += omap3-beagle.dtb +dtb-$(CONFIG_MACH_OMAP4_PANDA) += omap4-panda.dtb +dtb-$(CONFIG_MACH_OVERO) += omap3-overo.dtb +dtb-$(CONFIG_MACH_IGEP0020) += isee-igep-v2.dtb +dtb-$(CONFIG_MACH_IGEP0030) += isee-igep-v3.dtb diff --git a/arch/arm/mach-omap2/board-igep0020.c b/arch/arm/mach-omap2/board-igep0020.c index 28214483aaba..5dbc6993b0fe 100644 --- a/arch/arm/mach-omap2/board-igep0020.c +++ b/arch/arm/mach-omap2/board-igep0020.c @@ -644,6 +644,11 @@ static void __init igep_init(void) } } +static const char *igep2_dt_compat[] = { + "ISEE,igep-v2", + NULL +}; + MACHINE_START(IGEP0020, "IGEP v2 board") .atag_offset = 0x100, .reserve = omap_reserve, @@ -655,8 +660,14 @@ MACHINE_START(IGEP0020, "IGEP v2 board") .init_late = omap35xx_init_late, .timer = &omap3_timer, .restart = omap_prcm_restart, + .dt_compat = igep2_dt_compat, MACHINE_END +static const char *igep3_dt_compat[] = { + "ISEE,igep-v3", + NULL +}; + MACHINE_START(IGEP0030, "IGEP OMAP3 module") .atag_offset = 0x100, .reserve = omap_reserve, @@ -668,4 +679,5 @@ MACHINE_START(IGEP0030, "IGEP OMAP3 module") .init_late = omap35xx_init_late, .timer = &omap3_timer, .restart = omap_prcm_restart, + .dt_compat = &igep3_dt_compat, MACHINE_END diff --git a/arch/arm/mach-omap2/board-omap3beagle.c b/arch/arm/mach-omap2/board-omap3beagle.c index 6202fc76e490..8ead4c581021 100644 --- a/arch/arm/mach-omap2/board-omap3beagle.c +++ b/arch/arm/mach-omap2/board-omap3beagle.c @@ -530,6 +530,11 @@ static void __init omap3_beagle_init(void) beagle_opp_init(); } +static const char *omap3_beagle_dt_match[] = { + "ti,omap3-beagle", + NULL +}; + MACHINE_START(OMAP3_BEAGLE, "OMAP3 Beagle Board") /* Maintainer: Syed Mohammed Khasim - http://beagleboard.org */ .atag_offset = 0x100, @@ -542,4 +547,5 @@ MACHINE_START(OMAP3_BEAGLE, "OMAP3 Beagle Board") .init_late = omap3_init_late, .timer = &omap3_secure_timer, .restart = omap_prcm_restart, + .dt_compat = omap3_beagle_dt_match, MACHINE_END diff --git a/arch/arm/mach-omap2/board-omap4panda.c b/arch/arm/mach-omap2/board-omap4panda.c index 70f6d1d25463..613556c982fd 100644 --- a/arch/arm/mach-omap2/board-omap4panda.c +++ b/arch/arm/mach-omap2/board-omap4panda.c @@ -515,6 +515,11 @@ static void __init omap4_panda_init(void) omap4_panda_display_init(); } +static const char *omap4_panda_match[] = { + "ti,omap4-panda", + NULL, +}; + MACHINE_START(OMAP4_PANDA, "OMAP4 Panda board") /* Maintainer: David Anders - Texas Instruments Inc */ .atag_offset = 0x100, @@ -527,4 +532,5 @@ MACHINE_START(OMAP4_PANDA, "OMAP4 Panda board") .init_late = omap4430_init_late, .timer = &omap4_timer, .restart = omap_prcm_restart, + .dt_compat = omap4_panda_match, MACHINE_END diff --git a/arch/arm/mach-omap2/board-overo.c b/arch/arm/mach-omap2/board-overo.c index 779734d8ba37..64e90c61cd85 100644 --- a/arch/arm/mach-omap2/board-overo.c +++ b/arch/arm/mach-omap2/board-overo.c @@ -546,6 +546,11 @@ static void __init overo_init(void) "OVERO_GPIO_USBH_CPEN\n"); } +static const char *omap3_overo_dt_match[] = { + "gumstix,omap3-overo", + NULL +}; + MACHINE_START(OVERO, "Gumstix Overo") .atag_offset = 0x100, .reserve = omap_reserve, @@ -557,4 +562,5 @@ MACHINE_START(OVERO, "Gumstix Overo") .init_late = omap35xx_init_late, .timer = &omap3_timer, .restart = omap_prcm_restart, + .dt_compat = omap3_overo_dt_match, MACHINE_END diff --git a/arch/arm/mach-omap2/devices.c b/arch/arm/mach-omap2/devices.c index c00c68961bb8..02b9478b786f 100644 --- a/arch/arm/mach-omap2/devices.c +++ b/arch/arm/mach-omap2/devices.c @@ -23,7 +23,6 @@ #include <mach/irqs.h> #include <asm/mach-types.h> #include <asm/mach/map.h> -#include <asm/pmu.h> #include "iomap.h" #include <plat/board.h> @@ -448,7 +447,7 @@ static struct resource omap3_pmu_resource = { static struct platform_device omap_pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = 1, }; diff --git a/arch/arm/mach-pxa/devices.c b/arch/arm/mach-pxa/devices.c index 166eee5b8a70..c1f3b1279d97 100644 --- a/arch/arm/mach-pxa/devices.c +++ b/arch/arm/mach-pxa/devices.c @@ -6,7 +6,6 @@ #include <linux/spi/pxa2xx_spi.h> #include <linux/i2c/pxa-i2c.h> -#include <asm/pmu.h> #include <mach/udc.h> #include <mach/pxa3xx-u2d.h> #include <mach/pxafb.h> @@ -42,7 +41,7 @@ static struct resource pxa_resource_pmu = { struct platform_device pxa_device_pmu = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .resource = &pxa_resource_pmu, .num_resources = 1, }; diff --git a/arch/arm/mach-realview/realview_eb.c b/arch/arm/mach-realview/realview_eb.c index baf382c5e776..d7a6e9cebba4 100644 --- a/arch/arm/mach-realview/realview_eb.c +++ b/arch/arm/mach-realview/realview_eb.c @@ -32,7 +32,6 @@ #include <asm/irq.h> #include <asm/leds.h> #include <asm/mach-types.h> -#include <asm/pmu.h> #include <asm/pgtable.h> #include <asm/hardware/gic.h> #include <asm/hardware/cache-l2x0.h> @@ -297,7 +296,7 @@ static struct resource pmu_resources[] = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(pmu_resources), .resource = pmu_resources, }; diff --git a/arch/arm/mach-realview/realview_pb1176.c b/arch/arm/mach-realview/realview_pb1176.c index b1d7cafa1a6d..361f898884c8 100644 --- a/arch/arm/mach-realview/realview_pb1176.c +++ b/arch/arm/mach-realview/realview_pb1176.c @@ -34,7 +34,6 @@ #include <asm/irq.h> #include <asm/leds.h> #include <asm/mach-types.h> -#include <asm/pmu.h> #include <asm/pgtable.h> #include <asm/hardware/gic.h> #include <asm/hardware/cache-l2x0.h> @@ -280,7 +279,7 @@ static struct resource pmu_resource = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = 1, .resource = &pmu_resource, }; diff --git a/arch/arm/mach-realview/realview_pb11mp.c b/arch/arm/mach-realview/realview_pb11mp.c index a98c536e3327..c56bc8d4d11b 100644 --- a/arch/arm/mach-realview/realview_pb11mp.c +++ b/arch/arm/mach-realview/realview_pb11mp.c @@ -32,7 +32,6 @@ #include <asm/irq.h> #include <asm/leds.h> #include <asm/mach-types.h> -#include <asm/pmu.h> #include <asm/pgtable.h> #include <asm/hardware/gic.h> #include <asm/hardware/cache-l2x0.h> @@ -263,7 +262,7 @@ static struct resource pmu_resources[] = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(pmu_resources), .resource = pmu_resources, }; diff --git a/arch/arm/mach-realview/realview_pba8.c b/arch/arm/mach-realview/realview_pba8.c index 59650174e6ed..040937582453 100644 --- a/arch/arm/mach-realview/realview_pba8.c +++ b/arch/arm/mach-realview/realview_pba8.c @@ -31,7 +31,6 @@ #include <asm/irq.h> #include <asm/leds.h> #include <asm/mach-types.h> -#include <asm/pmu.h> #include <asm/pgtable.h> #include <asm/hardware/gic.h> @@ -241,7 +240,7 @@ static struct resource pmu_resource = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = 1, .resource = &pmu_resource, }; diff --git a/arch/arm/mach-realview/realview_pbx.c b/arch/arm/mach-realview/realview_pbx.c index 3f2f605624e9..97885dc11e8c 100644 --- a/arch/arm/mach-realview/realview_pbx.c +++ b/arch/arm/mach-realview/realview_pbx.c @@ -30,7 +30,6 @@ #include <asm/irq.h> #include <asm/leds.h> #include <asm/mach-types.h> -#include <asm/pmu.h> #include <asm/smp_twd.h> #include <asm/pgtable.h> #include <asm/hardware/gic.h> @@ -280,7 +279,7 @@ static struct resource pmu_resources[] = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(pmu_resources), .resource = pmu_resources, }; diff --git a/arch/arm/mach-tegra/devices.c b/arch/arm/mach-tegra/devices.c index c70e65ffa36b..61e9603744a7 100644 --- a/arch/arm/mach-tegra/devices.c +++ b/arch/arm/mach-tegra/devices.c @@ -23,7 +23,6 @@ #include <linux/fsl_devices.h> #include <linux/serial_8250.h> #include <linux/i2c-tegra.h> -#include <asm/pmu.h> #include <mach/irqs.h> #include <mach/iomap.h> #include <mach/dma.h> @@ -516,7 +515,7 @@ static struct resource tegra_pmu_resources[] = { struct platform_device tegra_pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(tegra_pmu_resources), .resource = tegra_pmu_resources, }; diff --git a/arch/arm/mach-ux500/cpu-db8500.c b/arch/arm/mach-ux500/cpu-db8500.c index db3c52d56ca4..7e6c384881a6 100644 --- a/arch/arm/mach-ux500/cpu-db8500.c +++ b/arch/arm/mach-ux500/cpu-db8500.c @@ -18,8 +18,8 @@ #include <linux/io.h> #include <linux/mfd/abx500/ab8500.h> -#include <asm/mach/map.h> #include <asm/pmu.h> +#include <asm/mach/map.h> #include <plat/gpio-nomadik.h> #include <mach/hardware.h> #include <mach/setup.h> @@ -122,7 +122,7 @@ struct arm_pmu_platdata db8500_pmu_platdata = { static struct platform_device db8500_pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(db8500_pmu_resources), .resource = db8500_pmu_resources, .dev.platform_data = &db8500_pmu_platdata, diff --git a/arch/arm/mach-versatile/Makefile.boot b/arch/arm/mach-versatile/Makefile.boot index ff0a4b5b0a82..6910b3144563 100644 --- a/arch/arm/mach-versatile/Makefile.boot +++ b/arch/arm/mach-versatile/Makefile.boot @@ -2,3 +2,5 @@ params_phys-y := 0x00000100 initrd_phys-y := 0x00800000 +dtb-$(CONFIG_MACH_VERSATILE_DT) += versatile-pb.dtb +dtb-$(CONFIG_MACH_VERSATILE_DT) += versatile-ab.dtb diff --git a/arch/arm/mach-vexpress/ct-ca9x4.c b/arch/arm/mach-vexpress/ct-ca9x4.c index 61c492403b05..e4073a60a864 100644 --- a/arch/arm/mach-vexpress/ct-ca9x4.c +++ b/arch/arm/mach-vexpress/ct-ca9x4.c @@ -13,7 +13,6 @@ #include <asm/hardware/arm_timer.h> #include <asm/hardware/cache-l2x0.h> #include <asm/hardware/gic.h> -#include <asm/pmu.h> #include <asm/smp_scu.h> #include <asm/smp_twd.h> @@ -144,7 +143,7 @@ static struct resource pmu_resources[] = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(pmu_resources), .resource = pmu_resources, }; diff --git a/arch/arm/mm/Kconfig b/arch/arm/mm/Kconfig index 101b9681c08c..d1d60b6d707f 100644 --- a/arch/arm/mm/Kconfig +++ b/arch/arm/mm/Kconfig @@ -851,7 +851,7 @@ config ARM_L1_CACHE_SHIFT config ARM_DMA_MEM_BUFFERABLE bool "Use non-cacheable memory for DMA" if (CPU_V6 || CPU_V6K) && !CPU_V7 depends on !(MACH_REALVIEW_PB1176 || REALVIEW_EB_ARM11MP || \ - MACH_REALVIEW_PB11MP) + MACH_REALVIEW_PB11MP || ARCH_HIGHBANK) default y if CPU_V6 || CPU_V6K || CPU_V7 help Historically, the kernel has used strongly ordered mappings to diff --git a/arch/arm/mm/Makefile b/arch/arm/mm/Makefile index 8a9c4cb50a93..9d32a253aac9 100644 --- a/arch/arm/mm/Makefile +++ b/arch/arm/mm/Makefile @@ -7,6 +7,7 @@ obj-y := dma-mapping.o extable.o fault.o init.o \ obj-$(CONFIG_MMU) += fault-armv.o flush.o idmap.o ioremap.o \ mmap.o pgd.o mmu.o vmregion.o +obj-$(CONFIG_DEBUG_RODATA) += rodata.o ifneq ($(CONFIG_MMU),y) obj-y += nommu.o diff --git a/arch/arm/mm/cache-l2x0.c b/arch/arm/mm/cache-l2x0.c index 2a8e380501e8..847fc742859f 100644 --- a/arch/arm/mm/cache-l2x0.c +++ b/arch/arm/mm/cache-l2x0.c @@ -33,6 +33,16 @@ static DEFINE_RAW_SPINLOCK(l2x0_lock); static u32 l2x0_way_mask; /* Bitmask of active ways */ static u32 l2x0_size; static unsigned long sync_reg_offset = L2X0_CACHE_SYNC; +static u32 l2x0_cache_id; +static unsigned int l2x0_sets; +static unsigned int l2x0_ways; + +static inline bool is_pl310_rev(int rev) +{ + return (l2x0_cache_id & + (L2X0_CACHE_ID_PART_MASK | L2X0_CACHE_ID_REV_MASK)) == + (L2X0_CACHE_ID_PART_L310 | rev); +} struct l2x0_regs l2x0_saved_regs; @@ -130,6 +140,23 @@ static void l2x0_cache_sync(void) raw_spin_unlock_irqrestore(&l2x0_lock, flags); } +#ifdef CONFIG_PL310_ERRATA_727915 +static void l2x0_for_each_set_way(void __iomem *reg) +{ + int set; + int way; + unsigned long flags; + + for (way = 0; way < l2x0_ways; way++) { + raw_spin_lock_irqsave(&l2x0_lock, flags); + for (set = 0; set < l2x0_sets; set++) + writel_relaxed((way << 28) | (set << 5), reg); + cache_sync(); + raw_spin_unlock_irqrestore(&l2x0_lock, flags); + } +} +#endif + static void __l2x0_flush_all(void) { debug_writel(0x03); @@ -143,6 +170,13 @@ static void l2x0_flush_all(void) { unsigned long flags; +#ifdef CONFIG_PL310_ERRATA_727915 + if (is_pl310_rev(REV_PL310_R2P0)) { + l2x0_for_each_set_way(l2x0_base + L2X0_CLEAN_INV_LINE_IDX); + return; + } +#endif + /* clean all ways */ raw_spin_lock_irqsave(&l2x0_lock, flags); __l2x0_flush_all(); @@ -153,11 +187,20 @@ static void l2x0_clean_all(void) { unsigned long flags; +#ifdef CONFIG_PL310_ERRATA_727915 + if (is_pl310_rev(REV_PL310_R2P0)) { + l2x0_for_each_set_way(l2x0_base + L2X0_CLEAN_LINE_IDX); + return; + } +#endif + /* clean all ways */ raw_spin_lock_irqsave(&l2x0_lock, flags); + debug_writel(0x03); writel_relaxed(l2x0_way_mask, l2x0_base + L2X0_CLEAN_WAY); cache_wait_way(l2x0_base + L2X0_CLEAN_WAY, l2x0_way_mask); cache_sync(); + debug_writel(0x00); raw_spin_unlock_irqrestore(&l2x0_lock, flags); } @@ -309,26 +352,24 @@ static void l2x0_unlock(u32 cache_id) void __init l2x0_init(void __iomem *base, u32 aux_val, u32 aux_mask) { u32 aux; - u32 cache_id; u32 way_size = 0; - int ways; const char *type; l2x0_base = base; - cache_id = readl_relaxed(l2x0_base + L2X0_CACHE_ID); + l2x0_cache_id = readl_relaxed(l2x0_base + L2X0_CACHE_ID); aux = readl_relaxed(l2x0_base + L2X0_AUX_CTRL); aux &= aux_mask; aux |= aux_val; /* Determine the number of ways */ - switch (cache_id & L2X0_CACHE_ID_PART_MASK) { + switch (l2x0_cache_id & L2X0_CACHE_ID_PART_MASK) { case L2X0_CACHE_ID_PART_L310: if (aux & (1 << 16)) - ways = 16; + l2x0_ways = 16; else - ways = 8; + l2x0_ways = 8; type = "L310"; #ifdef CONFIG_PL310_ERRATA_753970 /* Unmapped register. */ @@ -337,24 +378,25 @@ void __init l2x0_init(void __iomem *base, u32 aux_val, u32 aux_mask) outer_cache.set_debug = pl310_set_debug; break; case L2X0_CACHE_ID_PART_L210: - ways = (aux >> 13) & 0xf; + l2x0_ways = (aux >> 13) & 0xf; type = "L210"; break; default: /* Assume unknown chips have 8 ways */ - ways = 8; + l2x0_ways = 8; type = "L2x0 series"; break; } - l2x0_way_mask = (1 << ways) - 1; + l2x0_way_mask = (1 << l2x0_ways) - 1; /* * L2 cache Size = Way size * Number of ways */ way_size = (aux & L2X0_AUX_CTRL_WAY_SIZE_MASK) >> 17; - way_size = 1 << (way_size + 3); - l2x0_size = ways * way_size * SZ_1K; + way_size = SZ_1K << (way_size + 3); + l2x0_size = l2x0_ways * way_size; + l2x0_sets = way_size / CACHE_LINE_SIZE; /* * Check if l2x0 controller is already enabled. @@ -363,7 +405,7 @@ void __init l2x0_init(void __iomem *base, u32 aux_val, u32 aux_mask) */ if (!(readl_relaxed(l2x0_base + L2X0_CTRL) & 1)) { /* Make sure that I&D is not locked down when starting */ - l2x0_unlock(cache_id); + l2x0_unlock(l2x0_cache_id); /* l2x0 controller is disabled */ writel_relaxed(aux, l2x0_base + L2X0_AUX_CTRL); @@ -386,7 +428,7 @@ void __init l2x0_init(void __iomem *base, u32 aux_val, u32 aux_mask) printk(KERN_INFO "%s cache controller enabled\n", type); printk(KERN_INFO "l2x0: %d ways, CACHE_ID 0x%08x, AUX_CTRL 0x%08x, Cache size: %d B\n", - ways, cache_id, aux, l2x0_size); + l2x0_ways, l2x0_cache_id, aux, l2x0_size); } #ifdef CONFIG_OF diff --git a/arch/arm/mm/cache-v6.S b/arch/arm/mm/cache-v6.S index 4b10760c56d6..f4e60274e7a9 100644 --- a/arch/arm/mm/cache-v6.S +++ b/arch/arm/mm/cache-v6.S @@ -270,6 +270,11 @@ v6_dma_clean_range: * - end - virtual end address of region */ ENTRY(v6_dma_flush_range) +#ifdef CONFIG_CACHE_FLUSH_RANGE_LIMIT + sub r2, r1, r0 + cmp r2, #CONFIG_CACHE_FLUSH_RANGE_LIMIT + bhi v6_dma_flush_dcache_all +#endif #ifdef CONFIG_DMA_CACHE_RWFO ldrb r2, [r0] @ read for ownership strb r2, [r0] @ write for ownership @@ -292,6 +297,18 @@ ENTRY(v6_dma_flush_range) mcr p15, 0, r0, c7, c10, 4 @ drain write buffer mov pc, lr +#ifdef CONFIG_CACHE_FLUSH_RANGE_LIMIT +v6_dma_flush_dcache_all: + mov r0, #0 +#ifdef HARVARD_CACHE + mcr p15, 0, r0, c7, c14, 0 @ D cache clean+invalidate +#else + mcr p15, 0, r0, c7, c15, 0 @ Cache clean+invalidate +#endif + mcr p15, 0, r0, c7, c10, 4 @ drain write buffer + mov pc, lr +#endif + /* * dma_map_area(start, size, dir) * - start - kernel virtual start address diff --git a/arch/arm/mm/mmu.c b/arch/arm/mm/mmu.c index c2fa21d0103e..c61679932295 100644 --- a/arch/arm/mm/mmu.c +++ b/arch/arm/mm/mmu.c @@ -565,11 +565,25 @@ static void __init *early_alloc(unsigned long sz) return early_alloc_aligned(sz, sz); } -static pte_t * __init early_pte_alloc(pmd_t *pmd, unsigned long addr, unsigned long prot) +static pte_t * __init early_pte_alloc(pmd_t *pmd) +{ + if (pmd_none(*pmd) || pmd_bad(*pmd)) + return early_alloc(PTE_HWTABLE_OFF + PTE_HWTABLE_SIZE); + return pmd_page_vaddr(*pmd); +} + +static void __init early_pte_install(pmd_t *pmd, pte_t *pte, unsigned long prot) +{ + __pmd_populate(pmd, __pa(pte), prot); + BUG_ON(pmd_bad(*pmd)); +} + +static pte_t * __init early_pte_alloc_and_install(pmd_t *pmd, + unsigned long addr, unsigned long prot) { if (pmd_none(*pmd)) { - pte_t *pte = early_alloc(PTE_HWTABLE_OFF + PTE_HWTABLE_SIZE); - __pmd_populate(pmd, __pa(pte), prot); + pte_t *pte = early_pte_alloc(pmd); + early_pte_install(pmd, pte, prot); } BUG_ON(pmd_bad(*pmd)); return pte_offset_kernel(pmd, addr); @@ -579,16 +593,23 @@ static void __init alloc_init_pte(pmd_t *pmd, unsigned long addr, unsigned long end, unsigned long pfn, const struct mem_type *type) { - pte_t *pte = early_pte_alloc(pmd, addr, type->prot_l1); + pte_t *start_pte = early_pte_alloc(pmd); + pte_t *pte = start_pte + pte_index(addr); + + /* If replacing a section mapping, the whole section must be replaced */ + BUG_ON(pmd_bad(*pmd) && ((addr | end) & ~PMD_MASK)); + do { set_pte_ext(pte, pfn_pte(pfn, __pgprot(type->prot_pte)), 0); pfn++; } while (pte++, addr += PAGE_SIZE, addr != end); + early_pte_install(pmd, start_pte, type->prot_l1); } static void __init alloc_init_section(pud_t *pud, unsigned long addr, unsigned long end, phys_addr_t phys, - const struct mem_type *type) + const struct mem_type *type, + bool force_pages) { pmd_t *pmd = pmd_offset(pud, addr); @@ -598,7 +619,8 @@ static void __init alloc_init_section(pud_t *pud, unsigned long addr, * L1 entries, whereas PGDs refer to a group of L1 entries making * up one logical pointer to an L2 table. */ - if (type->prot_sect && ((addr | end | phys) & ~SECTION_MASK) == 0) { + if (type->prot_sect && ((addr | end | phys) & ~SECTION_MASK) == 0 + && !force_pages) { pmd_t *p = pmd; #ifndef CONFIG_ARM_LPAE @@ -621,15 +643,15 @@ static void __init alloc_init_section(pud_t *pud, unsigned long addr, } } -static void __init alloc_init_pud(pgd_t *pgd, unsigned long addr, - unsigned long end, unsigned long phys, const struct mem_type *type) +static void alloc_init_pud(pgd_t *pgd, unsigned long addr, unsigned long end, + unsigned long phys, const struct mem_type *type, bool force_pages) { pud_t *pud = pud_offset(pgd, addr); unsigned long next; do { next = pud_addr_end(addr, end); - alloc_init_section(pud, addr, next, phys, type); + alloc_init_section(pud, addr, next, phys, type, force_pages); phys += next - addr; } while (pud++, addr = next, addr != end); } @@ -703,7 +725,7 @@ static void __init create_36bit_mapping(struct map_desc *md, * offsets, and we take full advantage of sections and * supersections. */ -static void __init create_mapping(struct map_desc *md) +static void __init create_mapping(struct map_desc *md, bool force_pages) { unsigned long addr, length, end; phys_addr_t phys; @@ -753,7 +775,7 @@ static void __init create_mapping(struct map_desc *md) do { unsigned long next = pgd_addr_end(addr, end); - alloc_init_pud(pgd, addr, next, phys, type); + alloc_init_pud(pgd, addr, next, phys, type, force_pages); phys += next - addr; addr = next; @@ -774,7 +796,7 @@ void __init iotable_init(struct map_desc *io_desc, int nr) vm = early_alloc_aligned(sizeof(*vm) * nr, __alignof__(*vm)); for (md = io_desc; nr; md++, nr--) { - create_mapping(md); + create_mapping(md, false); vm->addr = (void *)(md->virtual & PAGE_MASK); vm->size = PAGE_ALIGN(md->length + (md->virtual & ~PAGE_MASK)); vm->phys_addr = __pfn_to_phys(md->pfn); @@ -1126,12 +1148,12 @@ static void __init devicemaps_init(struct machine_desc *mdesc) map.virtual = 0xffff0000; map.length = PAGE_SIZE; map.type = MT_HIGH_VECTORS; - create_mapping(&map); + create_mapping(&map, false); if (!vectors_high()) { map.virtual = 0; map.type = MT_LOW_VECTORS; - create_mapping(&map); + create_mapping(&map, false); } /* @@ -1154,20 +1176,23 @@ static void __init devicemaps_init(struct machine_desc *mdesc) static void __init kmap_init(void) { #ifdef CONFIG_HIGHMEM - pkmap_page_table = early_pte_alloc(pmd_off_k(PKMAP_BASE), + pkmap_page_table = early_pte_alloc_and_install(pmd_off_k(PKMAP_BASE), PKMAP_BASE, _PAGE_KERNEL_TABLE); #endif } + static void __init map_lowmem(void) { struct memblock_region *reg; + phys_addr_t start; + phys_addr_t end; + struct map_desc map; /* Map all the lowmem memory banks. */ for_each_memblock(memory, reg) { - phys_addr_t start = reg->base; - phys_addr_t end = start + reg->size; - struct map_desc map; + start = reg->base; + end = start + reg->size; if (end > arm_lowmem_limit) end = arm_lowmem_limit; @@ -1179,8 +1204,20 @@ static void __init map_lowmem(void) map.length = end - start; map.type = MT_MEMORY; - create_mapping(&map); + create_mapping(&map, false); } + +#ifdef CONFIG_DEBUG_RODATA + start = __pa(_stext) & PMD_MASK; + end = ALIGN(__pa(__end_rodata), PMD_SIZE); + + map.pfn = __phys_to_pfn(start); + map.virtual = __phys_to_virt(start); + map.length = end - start; + map.type = MT_MEMORY; + + create_mapping(&map, true); +#endif } /* diff --git a/arch/arm/mm/rodata.c b/arch/arm/mm/rodata.c new file mode 100644 index 000000000000..9a8eb841c428 --- /dev/null +++ b/arch/arm/mm/rodata.c @@ -0,0 +1,159 @@ +/* + * linux/arch/arm/mm/rodata.c + * + * Copyright (C) 2011 Google, Inc. + * + * Author: Colin Cross <ccross@android.com> + * + * Based on x86 implementation in arch/x86/mm/init_32.c + * + * 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/kernel.h> +#include <linux/mm.h> +#include <linux/module.h> + +#include <asm/cache.h> +#include <asm/pgtable.h> +#include <asm/rodata.h> +#include <asm/sections.h> +#include <asm/tlbflush.h> + +#include "mm.h" + +static int kernel_set_to_readonly __read_mostly; + +#ifdef CONFIG_DEBUG_RODATA_TEST +static const int rodata_test_data = 0xC3; + +static noinline void rodata_test(void) +{ + int result; + + pr_info("%s: attempting to write to read-only section:\n", __func__); + + if (*(volatile int *)&rodata_test_data != 0xC3) { + pr_err("read only data changed before test\n"); + return; + } + + /* + * Attempt to to write to rodata_test_data, trapping the expected + * data abort. If the trap executed, result will be 1. If it didn't, + * result will be 0xFF. + */ + asm volatile( + "0: str %[zero], [%[rodata_test_data]]\n" + " mov %[result], #0xFF\n" + " b 2f\n" + "1: mov %[result], #1\n" + "2:\n" + + /* Exception fixup - if store at label 0 faults, jumps to 1 */ + ".pushsection __ex_table, \"a\"\n" + " .long 0b, 1b\n" + ".popsection\n" + + : [result] "=r" (result) + : [rodata_test_data] "r" (&rodata_test_data), [zero] "r" (0) + : "memory" + ); + + if (result == 1) + pr_info("write to read-only section trapped, success\n"); + else + pr_err("write to read-only section NOT trapped, test failed\n"); + + if (*(volatile int *)&rodata_test_data != 0xC3) + pr_err("read only data changed during write\n"); +} +#else +static inline void rodata_test(void) { } +#endif + +static int set_page_attributes(unsigned long virt, int numpages, + pte_t (*f)(pte_t)) +{ + pmd_t *pmd; + pte_t *pte; + unsigned long start = virt; + unsigned long end = virt + (numpages << PAGE_SHIFT); + unsigned long pmd_end; + + while (virt < end) { + pmd = pmd_off_k(virt); + pmd_end = min(ALIGN(virt + 1, PMD_SIZE), end); + + if ((pmd_val(*pmd) & PMD_TYPE_MASK) != PMD_TYPE_TABLE) { + pr_err("%s: pmd %p=%08lx for %08lx not page table\n", + __func__, pmd, pmd_val(*pmd), virt); + virt = pmd_end; + continue; + } + + while (virt < pmd_end) { + pte = pte_offset_kernel(pmd, virt); + set_pte_ext(pte, f(*pte), 0); + virt += PAGE_SIZE; + } + } + + flush_tlb_kernel_range(start, end); + + return 0; +} + +int set_memory_ro(unsigned long virt, int numpages) +{ + return set_page_attributes(virt, numpages, pte_wrprotect); +} +EXPORT_SYMBOL(set_memory_ro); + +int set_memory_rw(unsigned long virt, int numpages) +{ + return set_page_attributes(virt, numpages, pte_mkwrite); +} +EXPORT_SYMBOL(set_memory_rw); + +void set_kernel_text_rw(void) +{ + unsigned long start = PAGE_ALIGN((unsigned long)_text); + unsigned long size = PAGE_ALIGN((unsigned long)__end_rodata) - start; + + if (!kernel_set_to_readonly) + return; + + pr_debug("Set kernel text: %lx - %lx to read-write\n", + start, start + size); + + set_memory_rw(start, size >> PAGE_SHIFT); +} + +void set_kernel_text_ro(void) +{ + unsigned long start = PAGE_ALIGN((unsigned long)_text); + unsigned long size = PAGE_ALIGN((unsigned long)__end_rodata) - start; + + if (!kernel_set_to_readonly) + return; + + pr_info_once("Write protecting the kernel text section %lx - %lx\n", + start, start + size); + + pr_debug("Set kernel text: %lx - %lx to read only\n", + start, start + size); + + set_memory_ro(start, size >> PAGE_SHIFT); +} + +void mark_rodata_ro(void) +{ + kernel_set_to_readonly = 1; + + set_kernel_text_ro(); + + rodata_test(); +} diff --git a/arch/arm/plat-iop/pmu.c b/arch/arm/plat-iop/pmu.c index a2024b8685a1..ad9f9744a82d 100644 --- a/arch/arm/plat-iop/pmu.c +++ b/arch/arm/plat-iop/pmu.c @@ -9,7 +9,6 @@ */ #include <linux/platform_device.h> -#include <asm/pmu.h> #include <mach/irqs.h> static struct resource pmu_resource = { @@ -26,7 +25,7 @@ static struct resource pmu_resource = { static struct platform_device pmu_device = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .resource = &pmu_resource, .num_resources = 1, }; diff --git a/arch/arm/plat-samsung/devs.c b/arch/arm/plat-samsung/devs.c index fc49f3dabd76..6ff45d53362c 100644 --- a/arch/arm/plat-samsung/devs.c +++ b/arch/arm/plat-samsung/devs.c @@ -35,7 +35,6 @@ #include <media/s5p_hdmi.h> #include <asm/irq.h> -#include <asm/pmu.h> #include <asm/mach/arch.h> #include <asm/mach/map.h> #include <asm/mach/irq.h> @@ -1132,7 +1131,7 @@ static struct resource s5p_pmu_resource[] = { static struct platform_device s5p_device_pmu = { .name = "arm-pmu", - .id = ARM_PMU_DEVICE_CPU, + .id = -1, .num_resources = ARRAY_SIZE(s5p_pmu_resource), .resource = s5p_pmu_resource, }; diff --git a/arch/arm/vfp/entry.S b/arch/arm/vfp/entry.S index cc926c985981..7788327044ff 100644 --- a/arch/arm/vfp/entry.S +++ b/arch/arm/vfp/entry.S @@ -28,6 +28,7 @@ ENTRY(do_vfp) str r11, [r10, #TI_PREEMPT] #endif enable_irq + str r2, [sp, #S_PC] @ update regs->ARM_pc for Thumb 2 case ldr r4, .LCvfp ldr r11, [r10, #TI_CPU] @ CPU number add r10, r10, #TI_VFPSTATE @ r10 = workspace |