From ec7f68e07bf10198717b7824c78201b46bbf1956 Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Wed, 30 Oct 2013 22:48:00 +0000 Subject: iio: hid_Sensors: fix crash during trigger unregister We can't store the trigger instance created by iio_trigger_alloc, in trig field of iio_device structure. This needs to be stored in the driver private data. Othewise it can result in crash during module unload. Hence created a trig_ptr in the common data structure for each HID sensor IIO driver and storing here. Signed-off-by: Srinivas Pandruvada Signed-off-by: Jonathan Cameron --- include/linux/hid-sensor-hub.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include/linux') diff --git a/include/linux/hid-sensor-hub.h b/include/linux/hid-sensor-hub.h index a265af294ea4..206a2af6b62b 100644 --- a/include/linux/hid-sensor-hub.h +++ b/include/linux/hid-sensor-hub.h @@ -21,6 +21,8 @@ #include #include +#include +#include /** * struct hid_sensor_hub_attribute_info - Attribute info @@ -184,6 +186,7 @@ struct hid_sensor_common { struct platform_device *pdev; unsigned usage_id; bool data_ready; + struct iio_trigger *trigger; struct hid_sensor_hub_attribute_info poll; struct hid_sensor_hub_attribute_info report_state; struct hid_sensor_hub_attribute_info power_state; -- cgit v1.3-14-g43fede From d5b5f391d434c5cc8bcb1ab2d759738797b85f52 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 14 Nov 2013 16:23:04 +0100 Subject: ftrace, perf: Avoid infinite event generation loop Vince's perf-trinity fuzzer found yet another 'interesting' problem. When we sample the irq_work_exit tracepoint with period==1 (or PERF_SAMPLE_PERIOD) and we add an fasync SIGNAL handler we create an infinite event generation loop: ,-> | irq_work_exit() -> | trace_irq_work_exit() -> | ... | __perf_event_overflow() -> (due to fasync) | irq_work_queue() -> (irq_work_list must be empty) '--------- arch_irq_work_raise() Similar things can happen due to regular poll() wakeups if we exceed the ring-buffer wakeup watermark, or have an event_limit. To avoid this, dis-allow sampling this particular tracepoint. In order to achieve this, create a special perf_perm function pointer for each event and call this (when set) on trying to create a tracepoint perf event. [ roasted: use expr... to allow for ',' in your expression ] Reported-by: Vince Weaver Tested-by: Vince Weaver Signed-off-by: Peter Zijlstra Cc: Steven Rostedt Cc: Dave Jones Cc: Frederic Weisbecker Link: http://lkml.kernel.org/r/20131114152304.GC5364@laptop.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- arch/x86/include/asm/trace/irq_vectors.h | 11 +++++++++++ include/linux/ftrace_event.h | 16 ++++++++++++++++ include/linux/tracepoint.h | 4 ++++ include/trace/ftrace.h | 7 +++++++ kernel/trace/trace_event_perf.c | 6 ++++++ 5 files changed, 44 insertions(+) (limited to 'include/linux') diff --git a/arch/x86/include/asm/trace/irq_vectors.h b/arch/x86/include/asm/trace/irq_vectors.h index 2874df24e7a4..4cab890007a7 100644 --- a/arch/x86/include/asm/trace/irq_vectors.h +++ b/arch/x86/include/asm/trace/irq_vectors.h @@ -71,6 +71,17 @@ DEFINE_IRQ_VECTOR_EVENT(x86_platform_ipi); */ DEFINE_IRQ_VECTOR_EVENT(irq_work); +/* + * We must dis-allow sampling irq_work_exit() because perf event sampling + * itself can cause irq_work, which would lead to an infinite loop; + * + * 1) irq_work_exit happens + * 2) generates perf sample + * 3) generates irq_work + * 4) goto 1 + */ +TRACE_EVENT_PERF_PERM(irq_work_exit, is_sampling_event(p_event) ? -EPERM : 0); + /* * call_function - called when entering/exiting a call function interrupt * vector handler diff --git a/include/linux/ftrace_event.h b/include/linux/ftrace_event.h index 9abbe630c456..8c9b7a1c4138 100644 --- a/include/linux/ftrace_event.h +++ b/include/linux/ftrace_event.h @@ -248,6 +248,9 @@ struct ftrace_event_call { #ifdef CONFIG_PERF_EVENTS int perf_refcount; struct hlist_head __percpu *perf_events; + + int (*perf_perm)(struct ftrace_event_call *, + struct perf_event *); #endif }; @@ -317,6 +320,19 @@ struct ftrace_event_file { } \ early_initcall(trace_init_flags_##name); +#define __TRACE_EVENT_PERF_PERM(name, expr...) \ + static int perf_perm_##name(struct ftrace_event_call *tp_event, \ + struct perf_event *p_event) \ + { \ + return ({ expr; }); \ + } \ + static int __init trace_init_perf_perm_##name(void) \ + { \ + event_##name.perf_perm = &perf_perm_##name; \ + return 0; \ + } \ + early_initcall(trace_init_perf_perm_##name); + #define PERF_MAX_TRACE_SIZE 2048 #define MAX_FILTER_STR_VAL 256 /* Should handle KSYM_SYMBOL_LEN */ diff --git a/include/linux/tracepoint.h b/include/linux/tracepoint.h index ebeab360d851..f16dc0a40049 100644 --- a/include/linux/tracepoint.h +++ b/include/linux/tracepoint.h @@ -267,6 +267,8 @@ static inline void tracepoint_synchronize_unregister(void) #define TRACE_EVENT_FLAGS(event, flag) +#define TRACE_EVENT_PERF_PERM(event, expr...) + #endif /* DECLARE_TRACE */ #ifndef TRACE_EVENT @@ -399,4 +401,6 @@ static inline void tracepoint_synchronize_unregister(void) #define TRACE_EVENT_FLAGS(event, flag) +#define TRACE_EVENT_PERF_PERM(event, expr...) + #endif /* ifdef TRACE_EVENT (see note above) */ diff --git a/include/trace/ftrace.h b/include/trace/ftrace.h index 52594b20179e..6b852f60f8ae 100644 --- a/include/trace/ftrace.h +++ b/include/trace/ftrace.h @@ -90,6 +90,10 @@ #define TRACE_EVENT_FLAGS(name, value) \ __TRACE_EVENT_FLAGS(name, value) +#undef TRACE_EVENT_PERF_PERM +#define TRACE_EVENT_PERF_PERM(name, expr...) \ + __TRACE_EVENT_PERF_PERM(name, expr) + #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) @@ -140,6 +144,9 @@ #undef TRACE_EVENT_FLAGS #define TRACE_EVENT_FLAGS(event, flag) +#undef TRACE_EVENT_PERF_PERM +#define TRACE_EVENT_PERF_PERM(event, expr...) + #include TRACE_INCLUDE(TRACE_INCLUDE_FILE) /* diff --git a/kernel/trace/trace_event_perf.c b/kernel/trace/trace_event_perf.c index 78e27e3b52ac..630889f68b1d 100644 --- a/kernel/trace/trace_event_perf.c +++ b/kernel/trace/trace_event_perf.c @@ -24,6 +24,12 @@ static int total_ref_count; static int perf_trace_event_perm(struct ftrace_event_call *tp_event, struct perf_event *p_event) { + if (tp_event->perf_perm) { + int ret = tp_event->perf_perm(tp_event, p_event); + if (ret) + return ret; + } + /* The ftrace function trace is allowed only for root. */ if (ftrace_event_is_function(tp_event) && perf_paranoid_tracepoint_raw() && !capable(CAP_SYS_ADMIN)) -- cgit v1.3-14-g43fede From b972fc308c2763096b61b62169f2167ee0ca5a19 Mon Sep 17 00:00:00 2001 From: Alex Shi Date: Tue, 19 Nov 2013 17:21:52 +0800 Subject: sched: Remove unused variable in 'struct sched_domain' The 'u64 last_update' variable isn't used now, remove it to save a bit of space. Signed-off-by: Alex Shi Signed-off-by: Peter Zijlstra Cc: Morten.Rasmussen@arm.com Cc: linaro-kernel@lists.linaro.org Link: http://lkml.kernel.org/r/1384852912-24791-1-git-send-email-alex.shi@linaro.org Signed-off-by: Ingo Molnar --- include/linux/sched.h | 2 -- 1 file changed, 2 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index f7efc8604652..b122395bf5f4 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -823,8 +823,6 @@ struct sched_domain { unsigned int balance_interval; /* initialise to 1. units in ms. */ unsigned int nr_balance_failed; /* initialise to 0 */ - u64 last_update; - /* idle_balance() stats */ u64 max_newidle_lb_cost; unsigned long next_decay_max_lb_cost; -- cgit v1.3-14-g43fede From 6bab2c613d7fa70bb8514f89ab7455ede717142b Mon Sep 17 00:00:00 2001 From: Geert Uytterhoeven Date: Fri, 15 Nov 2013 14:15:07 -0800 Subject: genirq: Correct fuzzy and fragile IRQ_RETVAL() definition commit bedd30d986a0 ("genirq: make irqreturn_t an enum") blindly replaced "0" by "IRQ_NONE" in the "IRQ_RETVAL(x)" macro definition. However, as "x" is a condition, "0" meant "boolean false", not an irqreturn_t value. All of this worked, and kept working after the addition of IRQ_WAKE_THREAD, as - both "boolean false" and "IRQ_NONE" are "0" (for the comparison), - "boolean true" and "boolean false" nicely map to the correct values of "IRQ_HANDLED" and "IRQ_NONE" (for the return value). Correct the macro definition for clarity and future-proofness. Signed-off-by: Geert Uytterhoeven Signed-off-by: Andrew Morton Signed-off-by: Thomas Gleixner --- include/linux/irqreturn.h | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'include/linux') diff --git a/include/linux/irqreturn.h b/include/linux/irqreturn.h index 714ba08dc092..e374e369fb2f 100644 --- a/include/linux/irqreturn.h +++ b/include/linux/irqreturn.h @@ -14,6 +14,6 @@ enum irqreturn { }; typedef enum irqreturn irqreturn_t; -#define IRQ_RETVAL(x) ((x) != IRQ_NONE) +#define IRQ_RETVAL(x) ((x) ? IRQ_HANDLED : IRQ_NONE) #endif -- cgit v1.3-14-g43fede From 694e096fd7c2a7d40760fc9c9dfc826bdd495ea4 Mon Sep 17 00:00:00 2001 From: Anna Schumaker Date: Wed, 13 Nov 2013 12:29:08 -0500 Subject: NFS: Enabling v4.2 should not recompile nfsd and lockd When CONFIG_NFS_V4_2 is toggled nfsd and lockd will be recompiled, instead of only the nfs client. This patch moves a small amount of code into the client directory to avoid unnecessary recompiles. Signed-off-by: Anna Schumaker Signed-off-by: Trond Myklebust --- fs/nfs/blocklayout/blocklayout.h | 1 + fs/nfs/dns_resolve.c | 2 ++ fs/nfs/internal.h | 15 +++++++++++++++ fs/nfs/nfs4_fs.h | 8 ++++++++ include/linux/nfs4.h | 10 ---------- include/linux/nfs_fs.h | 18 ------------------ 6 files changed, 26 insertions(+), 28 deletions(-) (limited to 'include/linux') diff --git a/fs/nfs/blocklayout/blocklayout.h b/fs/nfs/blocklayout/blocklayout.h index 8485978993e8..9838fb020473 100644 --- a/fs/nfs/blocklayout/blocklayout.h +++ b/fs/nfs/blocklayout/blocklayout.h @@ -36,6 +36,7 @@ #include #include +#include "../nfs4_fs.h" #include "../pnfs.h" #include "../netns.h" diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c index fc0f95ec7358..d25f10fb4926 100644 --- a/fs/nfs/dns_resolve.c +++ b/fs/nfs/dns_resolve.c @@ -46,7 +46,9 @@ ssize_t nfs_dns_resolve_name(struct net *net, char *name, size_t namelen, #include #include #include +#include +#include "nfs4_fs.h" #include "dns_resolve.h" #include "cache_lib.h" #include "netns.h" diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index bca6a3e3c49c..8b5cc04a8611 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -269,6 +269,21 @@ extern const u32 nfs41_maxgetdevinfo_overhead; extern struct rpc_procinfo nfs4_procedures[]; #endif +#ifdef CONFIG_NFS_V4_SECURITY_LABEL +extern struct nfs4_label *nfs4_label_alloc(struct nfs_server *server, gfp_t flags); +static inline void nfs4_label_free(struct nfs4_label *label) +{ + if (label) { + kfree(label->label); + kfree(label); + } + return; +} +#else +static inline struct nfs4_label *nfs4_label_alloc(struct nfs_server *server, gfp_t flags) { return NULL; } +static inline void nfs4_label_free(void *label) {} +#endif /* CONFIG_NFS_V4_SECURITY_LABEL */ + /* proc.c */ void nfs_close_context(struct nfs_open_context *ctx, int is_sync); extern struct nfs_client *nfs_init_client(struct nfs_client *clp, diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index 3ce79b04522e..5609edc742a0 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -9,6 +9,14 @@ #ifndef __LINUX_FS_NFS_NFS4_FS_H #define __LINUX_FS_NFS_NFS4_FS_H +#if defined(CONFIG_NFS_V4_2) +#define NFS4_MAX_MINOR_VERSION 2 +#elif defined(CONFIG_NFS_V4_1) +#define NFS4_MAX_MINOR_VERSION 1 +#else +#define NFS4_MAX_MINOR_VERSION 0 +#endif + #if IS_ENABLED(CONFIG_NFS_V4) #define NFS4_MAX_LOOP_ON_RECOVER (10) diff --git a/include/linux/nfs4.h b/include/linux/nfs4.h index c6f41b616965..ddd84871c0ad 100644 --- a/include/linux/nfs4.h +++ b/include/linux/nfs4.h @@ -410,16 +410,6 @@ enum lock_type4 { #define NFS4_VERSION 4 #define NFS4_MINOR_VERSION 0 -#if defined(CONFIG_NFS_V4_2) -#define NFS4_MAX_MINOR_VERSION 2 -#else -#if defined(CONFIG_NFS_V4_1) -#define NFS4_MAX_MINOR_VERSION 1 -#else -#define NFS4_MAX_MINOR_VERSION 0 -#endif /* CONFIG_NFS_V4_1 */ -#endif /* CONFIG_NFS_V4_2 */ - #define NFS4_DEBUG 1 /* Index of predefined Linux client operations */ diff --git a/include/linux/nfs_fs.h b/include/linux/nfs_fs.h index 14a48207a304..48997374eaf0 100644 --- a/include/linux/nfs_fs.h +++ b/include/linux/nfs_fs.h @@ -506,24 +506,6 @@ extern const struct inode_operations nfs_referral_inode_operations; extern int nfs_mountpoint_expiry_timeout; extern void nfs_release_automount_timer(void); -/* - * linux/fs/nfs/nfs4proc.c - */ -#ifdef CONFIG_NFS_V4_SECURITY_LABEL -extern struct nfs4_label *nfs4_label_alloc(struct nfs_server *server, gfp_t flags); -static inline void nfs4_label_free(struct nfs4_label *label) -{ - if (label) { - kfree(label->label); - kfree(label); - } - return; -} -#else -static inline struct nfs4_label *nfs4_label_alloc(struct nfs_server *server, gfp_t flags) { return NULL; } -static inline void nfs4_label_free(void *label) {} -#endif - /* * linux/fs/nfs/unlink.c */ -- cgit v1.3-14-g43fede From 53e7cac35db5941f42221314c33693e71ffa496b Mon Sep 17 00:00:00 2001 From: Alexandre Courbot Date: Sat, 16 Nov 2013 21:44:52 +0900 Subject: gpiolib: use dedicated flags for GPIO properties GPIO mapping properties were defined using the GPIOF_* flags, which are declared in linux/gpio.h. This file is not included when using the GPIO descriptor interface. This patch declares the flags that can be used as GPIO mappings properties in linux/gpio/driver.h, and uses them in gpiolib, so that no deprecated declarations are used by the GPIO descriptor interface. This patch also allows GPIO_OPEN_DRAIN and GPIO_OPEN_SOURCE to be specified as GPIO mapping properties. Signed-off-by: Alexandre Courbot Signed-off-by: Linus Walleij --- drivers/gpio/gpiolib.c | 22 +++++++++++++++------- include/linux/gpio/driver.h | 11 +++++++++-- 2 files changed, 24 insertions(+), 9 deletions(-) (limited to 'include/linux') diff --git a/drivers/gpio/gpiolib.c b/drivers/gpio/gpiolib.c index d8938b9b794a..490198365ce4 100644 --- a/drivers/gpio/gpiolib.c +++ b/drivers/gpio/gpiolib.c @@ -14,6 +14,7 @@ #include #include #include +#include #define CREATE_TRACE_POINTS #include @@ -2274,7 +2275,8 @@ void gpiod_add_table(struct gpiod_lookup *table, size_t size) #ifdef CONFIG_OF static struct gpio_desc *of_find_gpio(struct device *dev, const char *con_id, - unsigned int idx, unsigned long *flags) + unsigned int idx, + enum gpio_lookup_flags *flags) { char prop_name[32]; /* 32 is max size of property name */ enum of_gpio_flags of_flags; @@ -2292,7 +2294,7 @@ static struct gpio_desc *of_find_gpio(struct device *dev, const char *con_id, return desc; if (of_flags & OF_GPIO_ACTIVE_LOW) - *flags |= GPIOF_ACTIVE_LOW; + *flags |= GPIO_ACTIVE_LOW; return desc; } @@ -2305,7 +2307,8 @@ static struct gpio_desc *of_find_gpio(struct device *dev, const char *con_id, #endif static struct gpio_desc *acpi_find_gpio(struct device *dev, const char *con_id, - unsigned int idx, unsigned long *flags) + unsigned int idx, + enum gpio_lookup_flags *flags) { struct acpi_gpio_info info; struct gpio_desc *desc; @@ -2315,13 +2318,14 @@ static struct gpio_desc *acpi_find_gpio(struct device *dev, const char *con_id, return desc; if (info.gpioint && info.active_low) - *flags |= GPIOF_ACTIVE_LOW; + *flags |= GPIO_ACTIVE_LOW; return desc; } static struct gpio_desc *gpiod_find(struct device *dev, const char *con_id, - unsigned int idx, unsigned long *flags) + unsigned int idx, + enum gpio_lookup_flags *flags) { const char *dev_id = dev ? dev_name(dev) : NULL; struct gpio_desc *desc = ERR_PTR(-ENODEV); @@ -2413,7 +2417,7 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, { struct gpio_desc *desc; int status; - unsigned long flags = 0; + enum gpio_lookup_flags flags = 0; dev_dbg(dev, "GPIO lookup for consumer %s\n", con_id); @@ -2439,8 +2443,12 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, if (status < 0) return ERR_PTR(status); - if (flags & GPIOF_ACTIVE_LOW) + if (flags & GPIO_ACTIVE_LOW) set_bit(FLAG_ACTIVE_LOW, &desc->flags); + if (flags & GPIO_OPEN_DRAIN) + set_bit(FLAG_OPEN_DRAIN, &desc->flags); + if (flags & GPIO_OPEN_SOURCE) + set_bit(FLAG_OPEN_SOURCE, &desc->flags); return desc; } diff --git a/include/linux/gpio/driver.h b/include/linux/gpio/driver.h index 656a27efb2c8..82eac610ce1a 100644 --- a/include/linux/gpio/driver.h +++ b/include/linux/gpio/driver.h @@ -125,6 +125,13 @@ extern struct gpio_chip *gpiochip_find(void *data, int gpiod_lock_as_irq(struct gpio_desc *desc); void gpiod_unlock_as_irq(struct gpio_desc *desc); +enum gpio_lookup_flags { + GPIO_ACTIVE_HIGH = (0 << 0), + GPIO_ACTIVE_LOW = (1 << 0), + GPIO_OPEN_DRAIN = (1 << 1), + GPIO_OPEN_SOURCE = (1 << 2), +}; + /** * Lookup table for associating GPIOs to specific devices and functions using * platform data. @@ -152,9 +159,9 @@ struct gpiod_lookup { */ unsigned int idx; /* - * mask of GPIOF_* values + * mask of GPIO_* values */ - unsigned long flags; + enum gpio_lookup_flags flags; }; /* -- cgit v1.3-14-g43fede From 9886e1fd096b8004a25bb8b7e5690ad1408ebc6f Mon Sep 17 00:00:00 2001 From: Thierry Reding Date: Mon, 25 Nov 2013 11:49:47 -0700 Subject: ARM: tegra: Provide dummy powergate implementation In order to support increased build test coverage for drivers, implement dummies for the powergate implementation. This will allow the drivers to be built without requiring support for Tegra to be selected. This patch solves the following build errors, which can be triggered in v3.13-rc1 by selecting DRM_TEGRA without ARCH_TEGRA: drivers/built-in.o: In function `gr3d_remove': drivers/gpu/drm/tegra/gr3d.c:321: undefined reference to `tegra_powergate_power_off' drivers/gpu/drm/tegra/gr3d.c:325: undefined reference to `tegra_powergate_power_off' drivers/built-in.o: In function `gr3d_probe': drivers/gpu/drm/tegra/gr3d.c:266: undefined reference to `tegra_powergate_sequence_power_up' drivers/gpu/drm/tegra/gr3d.c:273: undefined reference to `tegra_powergate_sequence_power_up' Signed-off-by: Thierry Reding [swarren, updated commit description] Signed-off-by: Stephen Warren Signed-off-by: Olof Johansson --- include/linux/tegra-powergate.h | 27 +++++++++++++++++++++++++++ 1 file changed, 27 insertions(+) (limited to 'include/linux') diff --git a/include/linux/tegra-powergate.h b/include/linux/tegra-powergate.h index c98cfa406952..fd4498329c7c 100644 --- a/include/linux/tegra-powergate.h +++ b/include/linux/tegra-powergate.h @@ -45,6 +45,7 @@ struct clk; #define TEGRA_POWERGATE_3D0 TEGRA_POWERGATE_3D +#ifdef CONFIG_ARCH_TEGRA int tegra_powergate_is_powered(int id); int tegra_powergate_power_on(int id); int tegra_powergate_power_off(int id); @@ -52,5 +53,31 @@ int tegra_powergate_remove_clamping(int id); /* Must be called with clk disabled, and returns with clk enabled */ int tegra_powergate_sequence_power_up(int id, struct clk *clk); +#else +static inline int tegra_powergate_is_powered(int id) +{ + return -ENOSYS; +} + +static inline int tegra_powergate_power_on(int id) +{ + return -ENOSYS; +} + +static inline int tegra_powergate_power_off(int id) +{ + return -ENOSYS; +} + +static inline int tegra_powergate_remove_clamping(int id) +{ + return -ENOSYS; +} + +static inline int tegra_powergate_sequence_power_up(int id, struct clk *clk) +{ + return -ENOSYS; +} +#endif #endif /* _MACH_TEGRA_POWERGATE_H_ */ -- cgit v1.3-14-g43fede From 5a87182aa21d6d5d306840feab9321818dd3e2a3 Mon Sep 17 00:00:00 2001 From: Viresh Kumar Date: Wed, 27 Nov 2013 09:09:42 +0530 Subject: cpufreq: suspend governors on system suspend/hibernate This patch adds cpufreq suspend/resume calls to dpm_{suspend|resume}_noirq() for handling suspend/resume of cpufreq governors. Lan Tianyu (Intel) & Jinhyuk Choi (Broadcom) found anr issue where tunables configuration for clusters/sockets with non-boot CPUs was getting lost after suspend/resume, as we were notifying governors with CPUFREQ_GOV_POLICY_EXIT on removal of the last cpu for that policy and so deallocating memory for tunables. This is fixed by this patch as we don't allow any operation on governors after device suspend and before device resume now. Reported-and-tested-by: Lan Tianyu Reported-by: Jinhyuk Choi Signed-off-by: Viresh Kumar [rjw: Changelog, minor cleanups] Signed-off-by: Rafael J. Wysocki --- drivers/base/power/main.c | 3 +++ drivers/cpufreq/cpufreq.c | 43 +++++++++++++++++++++++++++++++++++++++++++ include/linux/cpufreq.h | 8 ++++++++ 3 files changed, 54 insertions(+) (limited to 'include/linux') diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index 9f098a82cf04..10c3510d72a9 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -29,6 +29,7 @@ #include #include #include +#include #include #include "../base.h" #include "power.h" @@ -473,6 +474,7 @@ static void dpm_resume_noirq(pm_message_t state) dpm_show_time(starttime, state, "noirq"); resume_device_irqs(); cpuidle_resume(); + cpufreq_resume(); } /** @@ -885,6 +887,7 @@ static int dpm_suspend_noirq(pm_message_t state) ktime_t starttime = ktime_get(); int error = 0; + cpufreq_suspend(); cpuidle_pause(); suspend_device_irqs(); mutex_lock(&dpm_list_mtx); diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 02d534da22dd..606224a8abc2 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -26,6 +26,7 @@ #include #include #include +#include #include #include #include @@ -47,6 +48,9 @@ static LIST_HEAD(cpufreq_policy_list); static DEFINE_PER_CPU(char[CPUFREQ_NAME_LEN], cpufreq_cpu_governor); #endif +/* Flag to suspend/resume CPUFreq governors */ +static bool cpufreq_suspended; + static inline bool has_target(void) { return cpufreq_driver->target_index || cpufreq_driver->target; @@ -1462,6 +1466,41 @@ static struct subsys_interface cpufreq_interface = { .remove_dev = cpufreq_remove_dev, }; +void cpufreq_suspend(void) +{ + struct cpufreq_policy *policy; + + if (!has_target()) + return; + + pr_debug("%s: Suspending Governors\n", __func__); + + list_for_each_entry(policy, &cpufreq_policy_list, policy_list) + if (__cpufreq_governor(policy, CPUFREQ_GOV_STOP)) + pr_err("%s: Failed to stop governor for policy: %p\n", + __func__, policy); + + cpufreq_suspended = true; +} + +void cpufreq_resume(void) +{ + struct cpufreq_policy *policy; + + if (!has_target()) + return; + + pr_debug("%s: Resuming Governors\n", __func__); + + cpufreq_suspended = false; + + list_for_each_entry(policy, &cpufreq_policy_list, policy_list) + if (__cpufreq_governor(policy, CPUFREQ_GOV_START) + || __cpufreq_governor(policy, CPUFREQ_GOV_LIMITS)) + pr_err("%s: Failed to start governor for policy: %p\n", + __func__, policy); +} + /** * cpufreq_bp_suspend - Prepare the boot CPU for system suspend. * @@ -1764,6 +1803,10 @@ static int __cpufreq_governor(struct cpufreq_policy *policy, struct cpufreq_governor *gov = NULL; #endif + /* Don't start any governor operations if we are entering suspend */ + if (cpufreq_suspended) + return 0; + if (policy->governor->max_transition_latency && policy->cpuinfo.transition_latency > policy->governor->max_transition_latency) { diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index 5bd6ab9b0c27..91716658e9a1 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -272,6 +272,14 @@ cpufreq_verify_within_cpu_limits(struct cpufreq_policy *policy) policy->cpuinfo.max_freq); } +#ifdef CONFIG_CPU_FREQ +void cpufreq_suspend(void); +void cpufreq_resume(void); +#else +static inline void cpufreq_suspend(void) {} +static inline void cpufreq_resume(void) {} +#endif + /********************************************************************* * CPUFREQ NOTIFIER INTERFACE * *********************************************************************/ -- cgit v1.3-14-g43fede From e0d59733f6b1796b8d6692642c87d7dd862c3e3a Mon Sep 17 00:00:00 2001 From: Seiji Aguchi Date: Wed, 30 Oct 2013 15:27:26 -0400 Subject: efivars, efi-pstore: Hold off deletion of sysfs entry until the scan is completed Currently, when mounting pstore file system, a read callback of efi_pstore driver runs mutiple times as below. - In the first read callback, scan efivar_sysfs_list from head and pass a kmsg buffer of a entry to an upper pstore layer. - In the second read callback, rescan efivar_sysfs_list from the entry and pass another kmsg buffer to it. - Repeat the scan and pass until the end of efivar_sysfs_list. In this process, an entry is read across the multiple read function calls. To avoid race between the read and erasion, the whole process above is protected by a spinlock, holding in open() and releasing in close(). At the same time, kmemdup() is called to pass the buffer to pstore filesystem during it. And then, it causes a following lockdep warning. To make the dynamic memory allocation runnable without taking spinlock, holding off a deletion of sysfs entry if it happens while scanning it via efi_pstore, and deleting it after the scan is completed. To implement it, this patch introduces two flags, scanning and deleting, to efivar_entry. On the code basis, it seems that all the scanning and deleting logic is not needed because __efivars->lock are not dropped when reading from the EFI variable store. But, the scanning and deleting logic is still needed because an efi-pstore and a pstore filesystem works as follows. In case an entry(A) is found, the pointer is saved to psi->data. And efi_pstore_read() passes the entry(A) to a pstore filesystem by releasing __efivars->lock. And then, the pstore filesystem calls efi_pstore_read() again and the same entry(A), which is saved to psi->data, is used for resuming to scan a sysfs-list. So, to protect the entry(A), the logic is needed. [ 1.143710] ------------[ cut here ]------------ [ 1.144058] WARNING: CPU: 1 PID: 1 at kernel/lockdep.c:2740 lockdep_trace_alloc+0x104/0x110() [ 1.144058] DEBUG_LOCKS_WARN_ON(irqs_disabled_flags(flags)) [ 1.144058] Modules linked in: [ 1.144058] CPU: 1 PID: 1 Comm: systemd Not tainted 3.11.0-rc5 #2 [ 1.144058] 0000000000000009 ffff8800797e9ae0 ffffffff816614a5 ffff8800797e9b28 [ 1.144058] ffff8800797e9b18 ffffffff8105510d 0000000000000080 0000000000000046 [ 1.144058] 00000000000000d0 00000000000003af ffffffff81ccd0c0 ffff8800797e9b78 [ 1.144058] Call Trace: [ 1.144058] [] dump_stack+0x54/0x74 [ 1.144058] [] warn_slowpath_common+0x7d/0xa0 [ 1.144058] [] warn_slowpath_fmt+0x4c/0x50 [ 1.144058] [] ? vsscanf+0x57f/0x7b0 [ 1.144058] [] lockdep_trace_alloc+0x104/0x110 [ 1.144058] [] __kmalloc_track_caller+0x50/0x280 [ 1.144058] [] ? efi_pstore_read_func.part.1+0x12b/0x170 [ 1.144058] [] kmemdup+0x20/0x50 [ 1.144058] [] efi_pstore_read_func.part.1+0x12b/0x170 [ 1.144058] [] ? efi_pstore_read_func.part.1+0x170/0x170 [ 1.144058] [] efi_pstore_read_func+0xb4/0xe0 [ 1.144058] [] __efivar_entry_iter+0xfb/0x120 [ 1.144058] [] efi_pstore_read+0x3f/0x50 [ 1.144058] [] pstore_get_records+0x9a/0x150 [ 1.158207] [] ? selinux_d_instantiate+0x1c/0x20 [ 1.158207] [] ? parse_options+0x80/0x80 [ 1.158207] [] pstore_fill_super+0xa5/0xc0 [ 1.158207] [] mount_single+0xa2/0xd0 [ 1.158207] [] pstore_mount+0x18/0x20 [ 1.158207] [] mount_fs+0x39/0x1b0 [ 1.158207] [] ? __alloc_percpu+0x10/0x20 [ 1.158207] [] vfs_kern_mount+0x63/0xf0 [ 1.158207] [] do_mount+0x23e/0xa20 [ 1.158207] [] ? strndup_user+0x4b/0xf0 [ 1.158207] [] SyS_mount+0x83/0xc0 [ 1.158207] [] system_call_fastpath+0x16/0x1b [ 1.158207] ---[ end trace 61981bc62de9f6f4 ]--- Signed-off-by: Seiji Aguchi Tested-by: Madper Xie Cc: stable@kernel.org Signed-off-by: Matt Fleming --- drivers/firmware/efi/efi-pstore.c | 143 +++++++++++++++++++++++++++++++++++--- drivers/firmware/efi/efivars.c | 12 ++-- drivers/firmware/efi/vars.c | 12 +++- include/linux/efi.h | 4 ++ 4 files changed, 155 insertions(+), 16 deletions(-) (limited to 'include/linux') diff --git a/drivers/firmware/efi/efi-pstore.c b/drivers/firmware/efi/efi-pstore.c index 5002d50e3781..6ce31e93da55 100644 --- a/drivers/firmware/efi/efi-pstore.c +++ b/drivers/firmware/efi/efi-pstore.c @@ -18,14 +18,12 @@ module_param_named(pstore_disable, efivars_pstore_disable, bool, 0644); static int efi_pstore_open(struct pstore_info *psi) { - efivar_entry_iter_begin(); psi->data = NULL; return 0; } static int efi_pstore_close(struct pstore_info *psi) { - efivar_entry_iter_end(); psi->data = NULL; return 0; } @@ -91,19 +89,125 @@ static int efi_pstore_read_func(struct efivar_entry *entry, void *data) __efivar_entry_get(entry, &entry->var.Attributes, &entry->var.DataSize, entry->var.Data); size = entry->var.DataSize; + memcpy(*cb_data->buf, entry->var.Data, + (size_t)min_t(unsigned long, EFIVARS_DATA_SIZE_MAX, size)); - *cb_data->buf = kmemdup(entry->var.Data, size, GFP_KERNEL); - if (*cb_data->buf == NULL) - return -ENOMEM; return size; } +/** + * efi_pstore_scan_sysfs_enter + * @entry: scanning entry + * @next: next entry + * @head: list head + */ +static void efi_pstore_scan_sysfs_enter(struct efivar_entry *pos, + struct efivar_entry *next, + struct list_head *head) +{ + pos->scanning = true; + if (&next->list != head) + next->scanning = true; +} + +/** + * __efi_pstore_scan_sysfs_exit + * @entry: deleting entry + * @turn_off_scanning: Check if a scanning flag should be turned off + */ +static inline void __efi_pstore_scan_sysfs_exit(struct efivar_entry *entry, + bool turn_off_scanning) +{ + if (entry->deleting) { + list_del(&entry->list); + efivar_entry_iter_end(); + efivar_unregister(entry); + efivar_entry_iter_begin(); + } else if (turn_off_scanning) + entry->scanning = false; +} + +/** + * efi_pstore_scan_sysfs_exit + * @pos: scanning entry + * @next: next entry + * @head: list head + * @stop: a flag checking if scanning will stop + */ +static void efi_pstore_scan_sysfs_exit(struct efivar_entry *pos, + struct efivar_entry *next, + struct list_head *head, bool stop) +{ + __efi_pstore_scan_sysfs_exit(pos, true); + if (stop) + __efi_pstore_scan_sysfs_exit(next, &next->list != head); +} + +/** + * efi_pstore_sysfs_entry_iter + * + * @data: function-specific data to pass to callback + * @pos: entry to begin iterating from + * + * You MUST call efivar_enter_iter_begin() before this function, and + * efivar_entry_iter_end() afterwards. + * + * It is possible to begin iteration from an arbitrary entry within + * the list by passing @pos. @pos is updated on return to point to + * the next entry of the last one passed to efi_pstore_read_func(). + * To begin iterating from the beginning of the list @pos must be %NULL. + */ +static int efi_pstore_sysfs_entry_iter(void *data, struct efivar_entry **pos) +{ + struct efivar_entry *entry, *n; + struct list_head *head = &efivar_sysfs_list; + int size = 0; + + if (!*pos) { + list_for_each_entry_safe(entry, n, head, list) { + efi_pstore_scan_sysfs_enter(entry, n, head); + + size = efi_pstore_read_func(entry, data); + efi_pstore_scan_sysfs_exit(entry, n, head, size < 0); + if (size) + break; + } + *pos = n; + return size; + } + + list_for_each_entry_safe_from((*pos), n, head, list) { + efi_pstore_scan_sysfs_enter((*pos), n, head); + + size = efi_pstore_read_func((*pos), data); + efi_pstore_scan_sysfs_exit((*pos), n, head, size < 0); + if (size) + break; + } + *pos = n; + return size; +} + +/** + * efi_pstore_read + * + * This function returns a size of NVRAM entry logged via efi_pstore_write(). + * The meaning and behavior of efi_pstore/pstore are as below. + * + * size > 0: Got data of an entry logged via efi_pstore_write() successfully, + * and pstore filesystem will continue reading subsequent entries. + * size == 0: Entry was not logged via efi_pstore_write(), + * and efi_pstore driver will continue reading subsequent entries. + * size < 0: Failed to get data of entry logging via efi_pstore_write(), + * and pstore will stop reading entry. + */ static ssize_t efi_pstore_read(u64 *id, enum pstore_type_id *type, int *count, struct timespec *timespec, char **buf, bool *compressed, struct pstore_info *psi) { struct pstore_read_data data; + ssize_t size; data.id = id; data.type = type; @@ -112,8 +216,17 @@ static ssize_t efi_pstore_read(u64 *id, enum pstore_type_id *type, data.compressed = compressed; data.buf = buf; - return __efivar_entry_iter(efi_pstore_read_func, &efivar_sysfs_list, &data, - (struct efivar_entry **)&psi->data); + *data.buf = kzalloc(EFIVARS_DATA_SIZE_MAX, GFP_KERNEL); + if (!*data.buf) + return -ENOMEM; + + efivar_entry_iter_begin(); + size = efi_pstore_sysfs_entry_iter(&data, + (struct efivar_entry **)&psi->data); + efivar_entry_iter_end(); + if (size <= 0) + kfree(*data.buf); + return size; } static int efi_pstore_write(enum pstore_type_id type, @@ -184,9 +297,17 @@ static int efi_pstore_erase_func(struct efivar_entry *entry, void *data) return 0; } + if (entry->scanning) { + /* + * Skip deletion because this entry will be deleted + * after scanning is completed. + */ + entry->deleting = true; + } else + list_del(&entry->list); + /* found */ __efivar_entry_delete(entry); - list_del(&entry->list); return 1; } @@ -214,10 +335,12 @@ static int efi_pstore_erase(enum pstore_type_id type, u64 id, int count, efivar_entry_iter_begin(); found = __efivar_entry_iter(efi_pstore_erase_func, &efivar_sysfs_list, &edata, &entry); - efivar_entry_iter_end(); - if (found) + if (found && !entry->scanning) { + efivar_entry_iter_end(); efivar_unregister(entry); + } else + efivar_entry_iter_end(); return 0; } diff --git a/drivers/firmware/efi/efivars.c b/drivers/firmware/efi/efivars.c index 933eb027d527..3dc248239197 100644 --- a/drivers/firmware/efi/efivars.c +++ b/drivers/firmware/efi/efivars.c @@ -383,12 +383,16 @@ static ssize_t efivar_delete(struct file *filp, struct kobject *kobj, else if (__efivar_entry_delete(entry)) err = -EIO; - efivar_entry_iter_end(); - - if (err) + if (err) { + efivar_entry_iter_end(); return err; + } - efivar_unregister(entry); + if (!entry->scanning) { + efivar_entry_iter_end(); + efivar_unregister(entry); + } else + efivar_entry_iter_end(); /* It's dead Jim.... */ return count; diff --git a/drivers/firmware/efi/vars.c b/drivers/firmware/efi/vars.c index 391c67b182d9..b22659cccca4 100644 --- a/drivers/firmware/efi/vars.c +++ b/drivers/firmware/efi/vars.c @@ -683,8 +683,16 @@ struct efivar_entry *efivar_entry_find(efi_char16_t *name, efi_guid_t guid, if (!found) return NULL; - if (remove) - list_del(&entry->list); + if (remove) { + if (entry->scanning) { + /* + * The entry will be deleted + * after scanning is completed. + */ + entry->deleting = true; + } else + list_del(&entry->list); + } return entry; } diff --git a/include/linux/efi.h b/include/linux/efi.h index bc5687d0f315..11ce6784a196 100644 --- a/include/linux/efi.h +++ b/include/linux/efi.h @@ -801,6 +801,8 @@ struct efivar_entry { struct efi_variable var; struct list_head list; struct kobject kobj; + bool scanning; + bool deleting; }; @@ -866,6 +868,8 @@ void efivar_run_worker(void); #if defined(CONFIG_EFI_VARS) || defined(CONFIG_EFI_VARS_MODULE) int efivars_sysfs_init(void); +#define EFIVARS_DATA_SIZE_MAX 1024 + #endif /* CONFIG_EFI_VARS */ #endif /* _LINUX_EFI_H */ -- cgit v1.3-14-g43fede From 35773dac5f862cb1c82ea151eba3e2f6de51ec3e Mon Sep 17 00:00:00 2001 From: David Laight Date: Mon, 11 Nov 2013 12:26:54 +0000 Subject: usb: xhci: Link TRB must not occur within a USB payload burst Section 4.11.7.1 of rev 1.0 of the xhci specification states that a link TRB can only occur at a boundary between underlying USB frames (512 bytes for high speed devices). If this isn't done the USB frames aren't formatted correctly and, for example, the USB3 ethernet ax88179_178a card will stop sending (while still receiving) when running a netperf tcp transmit test with (say) and 8k buffer. This should be a candidate for stable, the ax88179_178a driver defaults to gso and tso enabled so it passes a lot of fragmented skb to the USB stack. Notes from Sarah: Discussion: http://marc.info/?l=linux-usb&m=138384509604981&w=2 This patch fixes a long-standing xHCI driver bug that was revealed by a change in 3.12 in the usb-net driver. Commit 638c5115a794981441246fa8fa5d95c1875af5ba "USBNET: support DMA SG" added support to use bulk endpoint scatter-gather (urb->sg). Only the USB ethernet drivers trigger this bug, because the mass storage driver sends sg list entries in page-sized chunks. This patch only fixes the issue for bulk endpoint scatter-gather. The problem will still occur for periodic endpoints, because hosts will interpret no-op transfers as a request to skip a service interval, which is not what we want. Luckily, the USB core isn't set up for scatter-gather on isochronous endpoints, and no USB drivers use scatter-gather for interrupt endpoints. Document this known limitation so that developers won't try to use urb->sg for interrupt endpoints until this issue is fixed. The more comprehensive fix would be to allow link TRBs in the middle of the endpoint ring and revert this patch, but that fix would touch too much code to be allowed in for stable. This patch should be backported to kernels as old as 3.12, that contain the commit 638c5115a794981441246fa8fa5d95c1875af5ba "USBNET: support DMA SG". Without this patch, the USB network device gets wedged, and stops sending packets. Mark Lord confirms this patch fixes the regression: http://marc.info/?l=linux-netdev&m=138487107625966&w=2 Signed-off-by: David Laight Signed-off-by: Sarah Sharp Tested-by: Mark Lord Cc: stable@vger.kernel.org --- drivers/usb/host/xhci-ring.c | 54 ++++++++++++++++++++++++++++++++++++++++++-- include/linux/usb.h | 2 ++ 2 files changed, 54 insertions(+), 2 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/host/xhci-ring.c b/drivers/usb/host/xhci-ring.c index 1e2f3f495843..53c2e296467f 100644 --- a/drivers/usb/host/xhci-ring.c +++ b/drivers/usb/host/xhci-ring.c @@ -2973,8 +2973,58 @@ static int prepare_ring(struct xhci_hcd *xhci, struct xhci_ring *ep_ring, } while (1) { - if (room_on_ring(xhci, ep_ring, num_trbs)) - break; + if (room_on_ring(xhci, ep_ring, num_trbs)) { + union xhci_trb *trb = ep_ring->enqueue; + unsigned int usable = ep_ring->enq_seg->trbs + + TRBS_PER_SEGMENT - 1 - trb; + u32 nop_cmd; + + /* + * Section 4.11.7.1 TD Fragments states that a link + * TRB must only occur at the boundary between + * data bursts (eg 512 bytes for 480M). + * While it is possible to split a large fragment + * we don't know the size yet. + * Simplest solution is to fill the trb before the + * LINK with nop commands. + */ + if (num_trbs == 1 || num_trbs <= usable || usable == 0) + break; + + if (ep_ring->type != TYPE_BULK) + /* + * While isoc transfers might have a buffer that + * crosses a 64k boundary it is unlikely. + * Since we can't add NOPs without generating + * gaps in the traffic just hope it never + * happens at the end of the ring. + * This could be fixed by writing a LINK TRB + * instead of the first NOP - however the + * TRB_TYPE_LINK_LE32() calls would all need + * changing to check the ring length. + */ + break; + + if (num_trbs >= TRBS_PER_SEGMENT) { + xhci_err(xhci, "Too many fragments %d, max %d\n", + num_trbs, TRBS_PER_SEGMENT - 1); + return -ENOMEM; + } + + nop_cmd = cpu_to_le32(TRB_TYPE(TRB_TR_NOOP) | + ep_ring->cycle_state); + ep_ring->num_trbs_free -= usable; + do { + trb->generic.field[0] = 0; + trb->generic.field[1] = 0; + trb->generic.field[2] = 0; + trb->generic.field[3] = nop_cmd; + trb++; + } while (--usable); + ep_ring->enqueue = trb; + if (room_on_ring(xhci, ep_ring, num_trbs)) + break; + } if (ep_ring == xhci->cmd_ring) { xhci_err(xhci, "Do not support expand command ring\n"); diff --git a/include/linux/usb.h b/include/linux/usb.h index 7454865ad148..512ab162832c 100644 --- a/include/linux/usb.h +++ b/include/linux/usb.h @@ -1264,6 +1264,8 @@ typedef void (*usb_complete_t)(struct urb *); * @sg: scatter gather buffer list, the buffer size of each element in * the list (except the last) must be divisible by the endpoint's * max packet size if no_sg_constraint isn't set in 'struct usb_bus' + * (FIXME: scatter-gather under xHCI is broken for periodic transfers. + * Do not use urb->sg for interrupt endpoints for now, only bulk.) * @num_mapped_sgs: (internal) number of mapped sg entries * @num_sgs: number of entries in the sg list * @transfer_buffer_length: How big is transfer_buffer. The transfer may -- cgit v1.3-14-g43fede From 9f740ffa8134aaef770f964485dac3ed6780d8b7 Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Wed, 27 Nov 2013 22:19:00 +0000 Subject: HID: hid-sensor-hub: Add logical min and max Exporting logical minimum and maximum of HID fields as part of the hid sensor attribute info. This can be used for range checking and to calculate enumeration base for NAry fields of HID sensor hub. Signed-off-by: Srinivas Pandruvada Signed-off-by: Jonathan Cameron --- drivers/hid/hid-sensor-hub.c | 20 ++++++++------------ include/linux/hid-sensor-hub.h | 2 ++ 2 files changed, 10 insertions(+), 12 deletions(-) (limited to 'include/linux') diff --git a/drivers/hid/hid-sensor-hub.c b/drivers/hid/hid-sensor-hub.c index a184e1921c11..d87f7cb4bee5 100644 --- a/drivers/hid/hid-sensor-hub.c +++ b/drivers/hid/hid-sensor-hub.c @@ -112,13 +112,15 @@ static int sensor_hub_get_physical_device_count( static void sensor_hub_fill_attr_info( struct hid_sensor_hub_attribute_info *info, - s32 index, s32 report_id, s32 units, s32 unit_expo, s32 size) + s32 index, s32 report_id, struct hid_field *field) { info->index = index; info->report_id = report_id; - info->units = units; - info->unit_expo = unit_expo; - info->size = size/8; + info->units = field->unit; + info->unit_expo = field->unit_exponent; + info->size = (field->report_size * field->report_count)/8; + info->logical_minimum = field->logical_minimum; + info->logical_maximum = field->logical_maximum; } static struct hid_sensor_hub_callbacks *sensor_hub_get_callback( @@ -325,9 +327,7 @@ int sensor_hub_input_get_attribute_info(struct hid_sensor_hub_device *hsdev, if (field->physical == usage_id && field->logical == attr_usage_id) { sensor_hub_fill_attr_info(info, i, report->id, - field->unit, field->unit_exponent, - field->report_size * - field->report_count); + field); ret = 0; } else { for (j = 0; j < field->maxusage; ++j) { @@ -336,11 +336,7 @@ int sensor_hub_input_get_attribute_info(struct hid_sensor_hub_device *hsdev, field->usage[j].collection_index == collection_index) { sensor_hub_fill_attr_info(info, - i, report->id, - field->unit, - field->unit_exponent, - field->report_size * - field->report_count); + i, report->id, field); ret = 0; break; } diff --git a/include/linux/hid-sensor-hub.h b/include/linux/hid-sensor-hub.h index 206a2af6b62b..b914ca3f57ba 100644 --- a/include/linux/hid-sensor-hub.h +++ b/include/linux/hid-sensor-hub.h @@ -42,6 +42,8 @@ struct hid_sensor_hub_attribute_info { s32 units; s32 unit_expo; s32 size; + s32 logical_minimum; + s32 logical_maximum; }; /** -- cgit v1.3-14-g43fede From 751d17e23a9f7c8e0bca5c0b2e8d39af655ecd2a Mon Sep 17 00:00:00 2001 From: Srinivas Pandruvada Date: Wed, 27 Nov 2013 22:19:00 +0000 Subject: iio: hid-sensors: Fix power and report state In the original HID sensor hub firmwares all Named array enums were to 0-based. But the most recent hub implemented as 1-based, because of the implementation by one of the major OS vendor. Using logical minimum for the field as the base of enum. So we add logical minimum to the selector values before setting those fields. Some sensor hub FWs already changed logical minimum from 0 to 1 to reflect this and hope every other vendor will follow. There is no easy way to add a common HID quirk for NAry elements, even if the standard specifies these field as NAry, the collection used to describe selectors is still just "logical". Signed-off-by: Srinivas Pandruvada Signed-off-by: Jonathan Cameron --- drivers/iio/common/hid-sensors/Kconfig | 9 --------- drivers/iio/common/hid-sensors/hid-sensor-trigger.c | 20 +++++++++++++++----- include/linux/hid-sensor-ids.h | 12 ++++++++++++ 3 files changed, 27 insertions(+), 14 deletions(-) (limited to 'include/linux') diff --git a/drivers/iio/common/hid-sensors/Kconfig b/drivers/iio/common/hid-sensors/Kconfig index 1178121b55b0..39188b72cd3b 100644 --- a/drivers/iio/common/hid-sensors/Kconfig +++ b/drivers/iio/common/hid-sensors/Kconfig @@ -25,13 +25,4 @@ config HID_SENSOR_IIO_TRIGGER If this driver is compiled as a module, it will be named hid-sensor-trigger. -config HID_SENSOR_ENUM_BASE_QUIRKS - bool "ENUM base quirks for HID Sensor IIO drivers" - depends on HID_SENSOR_IIO_COMMON - help - Say yes here to build support for sensor hub FW using - enumeration, which is using 1 as base instead of 0. - Since logical minimum is still set 0 instead of 1, - there is no easy way to differentiate. - endmenu diff --git a/drivers/iio/common/hid-sensors/hid-sensor-trigger.c b/drivers/iio/common/hid-sensors/hid-sensor-trigger.c index bbd6426c9726..7dcf83998e6f 100644 --- a/drivers/iio/common/hid-sensors/hid-sensor-trigger.c +++ b/drivers/iio/common/hid-sensors/hid-sensor-trigger.c @@ -33,24 +33,34 @@ static int hid_sensor_data_rdy_trigger_set_state(struct iio_trigger *trig, { struct hid_sensor_common *st = iio_trigger_get_drvdata(trig); int state_val; + int report_val; if (state) { if (sensor_hub_device_open(st->hsdev)) return -EIO; - } else + state_val = + HID_USAGE_SENSOR_PROP_POWER_STATE_D0_FULL_POWER_ENUM; + report_val = + HID_USAGE_SENSOR_PROP_REPORTING_STATE_ALL_EVENTS_ENUM; + + } else { sensor_hub_device_close(st->hsdev); + state_val = + HID_USAGE_SENSOR_PROP_POWER_STATE_D4_POWER_OFF_ENUM; + report_val = + HID_USAGE_SENSOR_PROP_REPORTING_STATE_NO_EVENTS_ENUM; + } - state_val = state ? 1 : 0; - if (IS_ENABLED(CONFIG_HID_SENSOR_ENUM_BASE_QUIRKS)) - ++state_val; st->data_ready = state; + state_val += st->power_state.logical_minimum; + report_val += st->report_state.logical_minimum; sensor_hub_set_feature(st->hsdev, st->power_state.report_id, st->power_state.index, (s32)state_val); sensor_hub_set_feature(st->hsdev, st->report_state.report_id, st->report_state.index, - (s32)state_val); + (s32)report_val); return 0; } diff --git a/include/linux/hid-sensor-ids.h b/include/linux/hid-sensor-ids.h index 4f945d3ed49f..8323775ac21d 100644 --- a/include/linux/hid-sensor-ids.h +++ b/include/linux/hid-sensor-ids.h @@ -117,4 +117,16 @@ #define HID_USAGE_SENSOR_PROP_REPORT_STATE 0x200316 #define HID_USAGE_SENSOR_PROY_POWER_STATE 0x200319 +/* Power state enumerations */ +#define HID_USAGE_SENSOR_PROP_POWER_STATE_UNDEFINED_ENUM 0x00 +#define HID_USAGE_SENSOR_PROP_POWER_STATE_D0_FULL_POWER_ENUM 0x01 +#define HID_USAGE_SENSOR_PROP_POWER_STATE_D1_LOW_POWER_ENUM 0x02 +#define HID_USAGE_SENSOR_PROP_POWER_STATE_D2_STANDBY_WITH_WAKE_ENUM 0x03 +#define HID_USAGE_SENSOR_PROP_POWER_STATE_D3_SLEEP_WITH_WAKE_ENUM 0x04 +#define HID_USAGE_SENSOR_PROP_POWER_STATE_D4_POWER_OFF_ENUM 0x05 + +/* Report State enumerations */ +#define HID_USAGE_SENSOR_PROP_REPORTING_STATE_NO_EVENTS_ENUM 0x00 +#define HID_USAGE_SENSOR_PROP_REPORTING_STATE_ALL_EVENTS_ENUM 0x01 + #endif -- cgit v1.3-14-g43fede From 471e42ad148c05d091219096726d751684ebf918 Mon Sep 17 00:00:00 2001 From: Thomas Pugliese Date: Mon, 2 Dec 2013 15:39:45 -0600 Subject: usb: wusbcore: fix deadlock in wusbhc_gtk_rekey When multiple wireless USB devices are connected and one of the devices disconnects, the host will distribute a new group key to the remaining devicese using wusbhc_gtk_rekey. wusbhc_gtk_rekey takes the wusbhc->mutex and holds it while it submits a URB to set the new key. This causes a deadlock in wa_urb_enqueue when it calls a device lookup helper function that takes the same lock. This patch changes wusbhc_gtk_rekey to submit a work item to set the GTK so that the URB is submitted without holding wusbhc->mutex. Signed-off-by: Thomas Pugliese Signed-off-by: Greg Kroah-Hartman --- drivers/usb/wusbcore/devconnect.c | 24 +--------- drivers/usb/wusbcore/security.c | 98 +++++++++++++++++++++++---------------- drivers/usb/wusbcore/wusbhc.h | 6 +-- include/linux/usb/wusb.h | 2 + 4 files changed, 62 insertions(+), 68 deletions(-) (limited to 'include/linux') diff --git a/drivers/usb/wusbcore/devconnect.c b/drivers/usb/wusbcore/devconnect.c index 5107ca999d4c..f14e7929ba22 100644 --- a/drivers/usb/wusbcore/devconnect.c +++ b/drivers/usb/wusbcore/devconnect.c @@ -97,18 +97,12 @@ static void wusbhc_devconnect_acked_work(struct work_struct *work); static void wusb_dev_free(struct wusb_dev *wusb_dev) { - if (wusb_dev) { - kfree(wusb_dev->set_gtk_req); - usb_free_urb(wusb_dev->set_gtk_urb); - kfree(wusb_dev); - } + kfree(wusb_dev); } static struct wusb_dev *wusb_dev_alloc(struct wusbhc *wusbhc) { struct wusb_dev *wusb_dev; - struct urb *urb; - struct usb_ctrlrequest *req; wusb_dev = kzalloc(sizeof(*wusb_dev), GFP_KERNEL); if (wusb_dev == NULL) @@ -118,22 +112,6 @@ static struct wusb_dev *wusb_dev_alloc(struct wusbhc *wusbhc) INIT_WORK(&wusb_dev->devconnect_acked_work, wusbhc_devconnect_acked_work); - urb = usb_alloc_urb(0, GFP_KERNEL); - if (urb == NULL) - goto err; - wusb_dev->set_gtk_urb = urb; - - req = kmalloc(sizeof(*req), GFP_KERNEL); - if (req == NULL) - goto err; - wusb_dev->set_gtk_req = req; - - req->bRequestType = USB_DIR_OUT | USB_TYPE_STANDARD | USB_RECIP_DEVICE; - req->bRequest = USB_REQ_SET_DESCRIPTOR; - req->wValue = cpu_to_le16(USB_DT_KEY << 8 | wusbhc->gtk_index); - req->wIndex = 0; - req->wLength = cpu_to_le16(wusbhc->gtk.descr.bLength); - return wusb_dev; err: wusb_dev_free(wusb_dev); diff --git a/drivers/usb/wusbcore/security.c b/drivers/usb/wusbcore/security.c index dd88441c8f78..4c40d0dbf53d 100644 --- a/drivers/usb/wusbcore/security.c +++ b/drivers/usb/wusbcore/security.c @@ -29,19 +29,16 @@ #include #include "wusbhc.h" -static void wusbhc_set_gtk_callback(struct urb *urb); -static void wusbhc_gtk_rekey_done_work(struct work_struct *work); +static void wusbhc_gtk_rekey_work(struct work_struct *work); int wusbhc_sec_create(struct wusbhc *wusbhc) { wusbhc->gtk.descr.bLength = sizeof(wusbhc->gtk.descr) + sizeof(wusbhc->gtk.data); wusbhc->gtk.descr.bDescriptorType = USB_DT_KEY; wusbhc->gtk.descr.bReserved = 0; + wusbhc->gtk_index = 0; - wusbhc->gtk_index = wusb_key_index(0, WUSB_KEY_INDEX_TYPE_GTK, - WUSB_KEY_INDEX_ORIGINATOR_HOST); - - INIT_WORK(&wusbhc->gtk_rekey_done_work, wusbhc_gtk_rekey_done_work); + INIT_WORK(&wusbhc->gtk_rekey_work, wusbhc_gtk_rekey_work); return 0; } @@ -113,7 +110,7 @@ int wusbhc_sec_start(struct wusbhc *wusbhc) wusbhc_generate_gtk(wusbhc); result = wusbhc->set_gtk(wusbhc, wusbhc->gtk_tkid, - &wusbhc->gtk.descr.bKeyData, key_size); + &wusbhc->gtk.descr.bKeyData, key_size); if (result < 0) dev_err(wusbhc->dev, "cannot set GTK for the host: %d\n", result); @@ -129,7 +126,7 @@ int wusbhc_sec_start(struct wusbhc *wusbhc) */ void wusbhc_sec_stop(struct wusbhc *wusbhc) { - cancel_work_sync(&wusbhc->gtk_rekey_done_work); + cancel_work_sync(&wusbhc->gtk_rekey_work); } @@ -185,12 +182,14 @@ static int wusb_dev_set_encryption(struct usb_device *usb_dev, int value) static int wusb_dev_set_gtk(struct wusbhc *wusbhc, struct wusb_dev *wusb_dev) { struct usb_device *usb_dev = wusb_dev->usb_dev; + u8 key_index = wusb_key_index(wusbhc->gtk_index, + WUSB_KEY_INDEX_TYPE_GTK, WUSB_KEY_INDEX_ORIGINATOR_HOST); return usb_control_msg( usb_dev, usb_sndctrlpipe(usb_dev, 0), USB_REQ_SET_DESCRIPTOR, USB_DIR_OUT | USB_TYPE_STANDARD | USB_RECIP_DEVICE, - USB_DT_KEY << 8 | wusbhc->gtk_index, 0, + USB_DT_KEY << 8 | key_index, 0, &wusbhc->gtk.descr, wusbhc->gtk.descr.bLength, 1000); } @@ -520,24 +519,55 @@ error_kzalloc: * Once all connected and authenticated devices have received the new * GTK, switch the host to using it. */ -static void wusbhc_gtk_rekey_done_work(struct work_struct *work) +static void wusbhc_gtk_rekey_work(struct work_struct *work) { - struct wusbhc *wusbhc = container_of(work, struct wusbhc, gtk_rekey_done_work); + struct wusbhc *wusbhc = container_of(work, + struct wusbhc, gtk_rekey_work); size_t key_size = sizeof(wusbhc->gtk.data); + int port_idx; + struct wusb_dev *wusb_dev, *wusb_dev_next; + LIST_HEAD(rekey_list); mutex_lock(&wusbhc->mutex); + /* generate the new key */ + wusbhc_generate_gtk(wusbhc); + /* roll the gtk index. */ + wusbhc->gtk_index = (wusbhc->gtk_index + 1) % (WUSB_KEY_INDEX_MAX + 1); + /* + * Save all connected devices on a list while holding wusbhc->mutex and + * take a reference to each one. Then submit the set key request to + * them after releasing the lock in order to avoid a deadlock. + */ + for (port_idx = 0; port_idx < wusbhc->ports_max; port_idx++) { + wusb_dev = wusbhc->port[port_idx].wusb_dev; + if (!wusb_dev || !wusb_dev->usb_dev + || !wusb_dev->usb_dev->authenticated) + continue; - if (--wusbhc->pending_set_gtks == 0) - wusbhc->set_gtk(wusbhc, wusbhc->gtk_tkid, &wusbhc->gtk.descr.bKeyData, key_size); - + wusb_dev_get(wusb_dev); + list_add_tail(&wusb_dev->rekey_node, &rekey_list); + } mutex_unlock(&wusbhc->mutex); -} -static void wusbhc_set_gtk_callback(struct urb *urb) -{ - struct wusbhc *wusbhc = urb->context; + /* Submit the rekey requests without holding wusbhc->mutex. */ + list_for_each_entry_safe(wusb_dev, wusb_dev_next, &rekey_list, + rekey_node) { + list_del_init(&wusb_dev->rekey_node); + dev_dbg(&wusb_dev->usb_dev->dev, "%s: rekey device at port %d\n", + __func__, wusb_dev->port_idx); + + if (wusb_dev_set_gtk(wusbhc, wusb_dev) < 0) { + dev_err(&wusb_dev->usb_dev->dev, "%s: rekey device at port %d failed\n", + __func__, wusb_dev->port_idx); + } + wusb_dev_put(wusb_dev); + } - queue_work(wusbd, &wusbhc->gtk_rekey_done_work); + /* Switch the host controller to use the new GTK. */ + mutex_lock(&wusbhc->mutex); + wusbhc->set_gtk(wusbhc, wusbhc->gtk_tkid, + &wusbhc->gtk.descr.bKeyData, key_size); + mutex_unlock(&wusbhc->mutex); } /** @@ -553,26 +583,12 @@ static void wusbhc_set_gtk_callback(struct urb *urb) */ void wusbhc_gtk_rekey(struct wusbhc *wusbhc) { - static const size_t key_size = sizeof(wusbhc->gtk.data); - int p; - - wusbhc_generate_gtk(wusbhc); - - for (p = 0; p < wusbhc->ports_max; p++) { - struct wusb_dev *wusb_dev; - - wusb_dev = wusbhc->port[p].wusb_dev; - if (!wusb_dev || !wusb_dev->usb_dev || !wusb_dev->usb_dev->authenticated) - continue; - - usb_fill_control_urb(wusb_dev->set_gtk_urb, wusb_dev->usb_dev, - usb_sndctrlpipe(wusb_dev->usb_dev, 0), - (void *)wusb_dev->set_gtk_req, - &wusbhc->gtk.descr, wusbhc->gtk.descr.bLength, - wusbhc_set_gtk_callback, wusbhc); - if (usb_submit_urb(wusb_dev->set_gtk_urb, GFP_KERNEL) == 0) - wusbhc->pending_set_gtks++; - } - if (wusbhc->pending_set_gtks == 0) - wusbhc->set_gtk(wusbhc, wusbhc->gtk_tkid, &wusbhc->gtk.descr.bKeyData, key_size); + /* + * We need to submit a URB to the downstream WUSB devices in order to + * change the group key. This can't be done while holding the + * wusbhc->mutex since that is also taken in the urb_enqueue routine + * and will cause a deadlock. Instead, queue a work item to do + * it when the lock is not held + */ + queue_work(wusbd, &wusbhc->gtk_rekey_work); } diff --git a/drivers/usb/wusbcore/wusbhc.h b/drivers/usb/wusbcore/wusbhc.h index 711b1952b114..6bd3b819a6b5 100644 --- a/drivers/usb/wusbcore/wusbhc.h +++ b/drivers/usb/wusbcore/wusbhc.h @@ -97,6 +97,7 @@ struct wusb_dev { struct kref refcnt; struct wusbhc *wusbhc; struct list_head cack_node; /* Connect-Ack list */ + struct list_head rekey_node; /* GTK rekey list */ u8 port_idx; u8 addr; u8 beacon_type:4; @@ -107,8 +108,6 @@ struct wusb_dev { struct usb_wireless_cap_descriptor *wusb_cap_descr; struct uwb_mas_bm availability; struct work_struct devconnect_acked_work; - struct urb *set_gtk_urb; - struct usb_ctrlrequest *set_gtk_req; struct usb_device *usb_dev; }; @@ -296,8 +295,7 @@ struct wusbhc { } __attribute__((packed)) gtk; u8 gtk_index; u32 gtk_tkid; - struct work_struct gtk_rekey_done_work; - int pending_set_gtks; + struct work_struct gtk_rekey_work; struct usb_encryption_descriptor *ccm1_etd; }; diff --git a/include/linux/usb/wusb.h b/include/linux/usb/wusb.h index 0c4d4ca370ec..eeb28329fa3c 100644 --- a/include/linux/usb/wusb.h +++ b/include/linux/usb/wusb.h @@ -271,6 +271,8 @@ static inline u8 wusb_key_index(int index, int type, int originator) #define WUSB_KEY_INDEX_TYPE_GTK 2 #define WUSB_KEY_INDEX_ORIGINATOR_HOST 0 #define WUSB_KEY_INDEX_ORIGINATOR_DEVICE 1 +/* bits 0-3 used for the key index. */ +#define WUSB_KEY_INDEX_MAX 15 /* A CCM Nonce, defined in WUSB1.0[6.4.1] */ struct aes_ccm_nonce { -- cgit v1.3-14-g43fede From c9a9972b6f093e4e2f81f58892a7523df894144d Mon Sep 17 00:00:00 2001 From: Alexandre Courbot Date: Mon, 25 Nov 2013 18:34:24 +0900 Subject: gpiolib: add missing declarations Add declaration of 'struct of_phandle_args' to avoid the following warning: In file included from arch/arm/mach-tegra/board-paz00.c:21:0: include/linux/gpio/driver.h:102:17: warning: 'struct of_phandle_args' declared inside parameter list include/linux/gpio/driver.h:102:17: warning: its scope is only this definition or declaration, which is probably not what you want Also proactively add other definitions/includes that could be missing in other contexts. Signed-off-by: Alexandre Courbot Reported-by: Stephen Warren Reviewed-by: Mika Westerberg Signed-off-by: Linus Walleij --- include/linux/gpio/driver.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include/linux') diff --git a/include/linux/gpio/driver.h b/include/linux/gpio/driver.h index 82eac610ce1a..3ea2cf6b0e6c 100644 --- a/include/linux/gpio/driver.h +++ b/include/linux/gpio/driver.h @@ -2,9 +2,12 @@ #define __LINUX_GPIO_DRIVER_H #include +#include struct device; struct gpio_desc; +struct of_phandle_args; +struct device_node; struct seq_file; /** -- cgit v1.3-14-g43fede From 12205a4b79bef56ef618a4b7caa840c5c971cff2 Mon Sep 17 00:00:00 2001 From: "Rafael J. Wysocki" Date: Sun, 8 Dec 2013 01:04:17 +0100 Subject: Revert "cpufreq: suspend governors on system suspend/hibernate" MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit 5a87182aa21d (cpufreq: suspend governors on system suspend/hibernate) causes hibernation problems to happen on Bjørn Mork's and Paul Bolle's systems, so revert it. Fixes: 5a87182aa21d (cpufreq: suspend governors on system suspend/hibernate) Reported-by: Bjørn Mork Reported-by: Paul Bolle Signed-off-by: Rafael J. Wysocki --- drivers/base/power/main.c | 3 --- drivers/cpufreq/cpufreq.c | 43 ------------------------------------------- include/linux/cpufreq.h | 8 -------- 3 files changed, 54 deletions(-) (limited to 'include/linux') diff --git a/drivers/base/power/main.c b/drivers/base/power/main.c index e3219dfd736c..1b41fca3d65a 100644 --- a/drivers/base/power/main.c +++ b/drivers/base/power/main.c @@ -29,7 +29,6 @@ #include #include #include -#include #include #include @@ -541,7 +540,6 @@ static void dpm_resume_noirq(pm_message_t state) dpm_show_time(starttime, state, "noirq"); resume_device_irqs(); cpuidle_resume(); - cpufreq_resume(); } /** @@ -957,7 +955,6 @@ static int dpm_suspend_noirq(pm_message_t state) ktime_t starttime = ktime_get(); int error = 0; - cpufreq_suspend(); cpuidle_pause(); suspend_device_irqs(); mutex_lock(&dpm_list_mtx); diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c index 81e9d4412db8..b7c3b877da44 100644 --- a/drivers/cpufreq/cpufreq.c +++ b/drivers/cpufreq/cpufreq.c @@ -26,7 +26,6 @@ #include #include #include -#include #include #include #include @@ -48,9 +47,6 @@ static LIST_HEAD(cpufreq_policy_list); static DEFINE_PER_CPU(char[CPUFREQ_NAME_LEN], cpufreq_cpu_governor); #endif -/* Flag to suspend/resume CPUFreq governors */ -static bool cpufreq_suspended; - static inline bool has_target(void) { return cpufreq_driver->target_index || cpufreq_driver->target; @@ -1466,41 +1462,6 @@ static struct subsys_interface cpufreq_interface = { .remove_dev = cpufreq_remove_dev, }; -void cpufreq_suspend(void) -{ - struct cpufreq_policy *policy; - - if (!has_target()) - return; - - pr_debug("%s: Suspending Governors\n", __func__); - - list_for_each_entry(policy, &cpufreq_policy_list, policy_list) - if (__cpufreq_governor(policy, CPUFREQ_GOV_STOP)) - pr_err("%s: Failed to stop governor for policy: %p\n", - __func__, policy); - - cpufreq_suspended = true; -} - -void cpufreq_resume(void) -{ - struct cpufreq_policy *policy; - - if (!has_target()) - return; - - pr_debug("%s: Resuming Governors\n", __func__); - - cpufreq_suspended = false; - - list_for_each_entry(policy, &cpufreq_policy_list, policy_list) - if (__cpufreq_governor(policy, CPUFREQ_GOV_START) - || __cpufreq_governor(policy, CPUFREQ_GOV_LIMITS)) - pr_err("%s: Failed to start governor for policy: %p\n", - __func__, policy); -} - /** * cpufreq_bp_suspend - Prepare the boot CPU for system suspend. * @@ -1803,10 +1764,6 @@ static int __cpufreq_governor(struct cpufreq_policy *policy, struct cpufreq_governor *gov = NULL; #endif - /* Don't start any governor operations if we are entering suspend */ - if (cpufreq_suspended) - return 0; - if (policy->governor->max_transition_latency && policy->cpuinfo.transition_latency > policy->governor->max_transition_latency) { diff --git a/include/linux/cpufreq.h b/include/linux/cpufreq.h index ee5fe9d77ae8..dc196bbcf227 100644 --- a/include/linux/cpufreq.h +++ b/include/linux/cpufreq.h @@ -280,14 +280,6 @@ cpufreq_verify_within_cpu_limits(struct cpufreq_policy *policy) policy->cpuinfo.max_freq); } -#ifdef CONFIG_CPU_FREQ -void cpufreq_suspend(void); -void cpufreq_resume(void); -#else -static inline void cpufreq_suspend(void) {} -static inline void cpufreq_resume(void) {} -#endif - /********************************************************************* * CPUFREQ NOTIFIER INTERFACE * *********************************************************************/ -- cgit v1.3-14-g43fede From ba1f14fbe70965ae0fb1655a5275a62723f65b77 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Thu, 28 Nov 2013 14:26:41 +0100 Subject: sched: Remove PREEMPT_NEED_RESCHED from generic code While hunting a preemption issue with Alexander, Ben noticed that the currently generic PREEMPT_NEED_RESCHED stuff is horribly broken for load-store architectures. We currently rely on the IPI to fold TIF_NEED_RESCHED into PREEMPT_NEED_RESCHED, but when this IPI lands while we already have a load for the preempt-count but before the store, the store will erase the PREEMPT_NEED_RESCHED change. The current preempt-count only works on load-store archs because interrupts are assumed to be completely balanced wrt their preempt_count fiddling; the previous preempt_count load will match the preempt_count state after the interrupt and therefore nothing gets lost. This patch removes the PREEMPT_NEED_RESCHED usage from generic code and pushes it into x86 arch code; the generic code goes back to relying on TIF_NEED_RESCHED. Boot tested on x86_64 and compile tested on ppc64. Reported-by: Benjamin Herrenschmidt Reported-and-Tested-by: Alexander Graf Signed-off-by: Peter Zijlstra Cc: Linus Torvalds Link: http://lkml.kernel.org/r/20131128132641.GP10022@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- arch/x86/include/asm/preempt.h | 11 +++++++++++ include/asm-generic/preempt.h | 35 +++++++++++------------------------ include/linux/sched.h | 2 -- 3 files changed, 22 insertions(+), 26 deletions(-) (limited to 'include/linux') diff --git a/arch/x86/include/asm/preempt.h b/arch/x86/include/asm/preempt.h index 8729723636fd..c8b051933b1b 100644 --- a/arch/x86/include/asm/preempt.h +++ b/arch/x86/include/asm/preempt.h @@ -7,6 +7,12 @@ DECLARE_PER_CPU(int, __preempt_count); +/* + * We use the PREEMPT_NEED_RESCHED bit as an inverted NEED_RESCHED such + * that a decrement hitting 0 means we can and should reschedule. + */ +#define PREEMPT_ENABLED (0 + PREEMPT_NEED_RESCHED) + /* * We mask the PREEMPT_NEED_RESCHED bit so as not to confuse all current users * that think a non-zero value indicates we cannot preempt. @@ -74,6 +80,11 @@ static __always_inline void __preempt_count_sub(int val) __this_cpu_add_4(__preempt_count, -val); } +/* + * Because we keep PREEMPT_NEED_RESCHED set when we do _not_ need to reschedule + * a decrement which hits zero means we have no preempt_count and should + * reschedule. + */ static __always_inline bool __preempt_count_dec_and_test(void) { GEN_UNARY_RMWcc("decl", __preempt_count, __percpu_arg(0), "e"); diff --git a/include/asm-generic/preempt.h b/include/asm-generic/preempt.h index ddf2b420ac8f..1cd3f5d767a8 100644 --- a/include/asm-generic/preempt.h +++ b/include/asm-generic/preempt.h @@ -3,13 +3,11 @@ #include -/* - * We mask the PREEMPT_NEED_RESCHED bit so as not to confuse all current users - * that think a non-zero value indicates we cannot preempt. - */ +#define PREEMPT_ENABLED (0) + static __always_inline int preempt_count(void) { - return current_thread_info()->preempt_count & ~PREEMPT_NEED_RESCHED; + return current_thread_info()->preempt_count; } static __always_inline int *preempt_count_ptr(void) @@ -17,11 +15,6 @@ static __always_inline int *preempt_count_ptr(void) return ¤t_thread_info()->preempt_count; } -/* - * We now loose PREEMPT_NEED_RESCHED and cause an extra reschedule; however the - * alternative is loosing a reschedule. Better schedule too often -- also this - * should be a very rare operation. - */ static __always_inline void preempt_count_set(int pc) { *preempt_count_ptr() = pc; @@ -41,28 +34,17 @@ static __always_inline void preempt_count_set(int pc) task_thread_info(p)->preempt_count = PREEMPT_ENABLED; \ } while (0) -/* - * We fold the NEED_RESCHED bit into the preempt count such that - * preempt_enable() can decrement and test for needing to reschedule with a - * single instruction. - * - * We invert the actual bit, so that when the decrement hits 0 we know we both - * need to resched (the bit is cleared) and can resched (no preempt count). - */ - static __always_inline void set_preempt_need_resched(void) { - *preempt_count_ptr() &= ~PREEMPT_NEED_RESCHED; } static __always_inline void clear_preempt_need_resched(void) { - *preempt_count_ptr() |= PREEMPT_NEED_RESCHED; } static __always_inline bool test_preempt_need_resched(void) { - return !(*preempt_count_ptr() & PREEMPT_NEED_RESCHED); + return false; } /* @@ -81,7 +63,12 @@ static __always_inline void __preempt_count_sub(int val) static __always_inline bool __preempt_count_dec_and_test(void) { - return !--*preempt_count_ptr(); + /* + * Because of load-store architectures cannot do per-cpu atomic + * operations; we cannot use PREEMPT_NEED_RESCHED because it might get + * lost. + */ + return !--*preempt_count_ptr() && tif_need_resched(); } /* @@ -89,7 +76,7 @@ static __always_inline bool __preempt_count_dec_and_test(void) */ static __always_inline bool should_resched(void) { - return unlikely(!*preempt_count_ptr()); + return unlikely(!preempt_count() && tif_need_resched()); } #ifdef CONFIG_PREEMPT diff --git a/include/linux/sched.h b/include/linux/sched.h index 768b037dfacb..96d674ba3876 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -440,8 +440,6 @@ struct task_cputime { .sum_exec_runtime = 0, \ } -#define PREEMPT_ENABLED (PREEMPT_NEED_RESCHED) - #ifdef CONFIG_PREEMPT_COUNT #define PREEMPT_DISABLED (1 + PREEMPT_ENABLED) #else -- cgit v1.3-14-g43fede From be5e610c0fd6ef772cafb9e0bd4128134804aef3 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 18 Nov 2013 18:27:06 +0100 Subject: math64: Add mul_u64_u32_shr() Introduce mul_u64_u32_shr() as proposed by Andy a while back; it allows using 64x64->128 muls on 64bit archs and recent GCC which defines __SIZEOF_INT128__ and __int128. (This new method will be used by the scheduler.) Signed-off-by: Peter Zijlstra Cc: fweisbec@gmail.com Cc: Andy Lutomirski Cc: Linus Torvalds Link: http://lkml.kernel.org/n/tip-hxjoeuzmrcaumR0uZwjpe2pv@git.kernel.org Signed-off-by: Ingo Molnar --- arch/x86/Kconfig | 1 + include/linux/math64.h | 30 ++++++++++++++++++++++++++++++ init/Kconfig | 6 ++++++ 3 files changed, 37 insertions(+) (limited to 'include/linux') diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index e903c71f7e69..0952ecd60eca 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig @@ -26,6 +26,7 @@ config X86 select HAVE_AOUT if X86_32 select HAVE_UNSTABLE_SCHED_CLOCK select ARCH_SUPPORTS_NUMA_BALANCING + select ARCH_SUPPORTS_INT128 if X86_64 select ARCH_WANTS_PROT_NUMA_PROT_NONE select HAVE_IDE select HAVE_OPROFILE diff --git a/include/linux/math64.h b/include/linux/math64.h index 69ed5f5e9f6e..c45c089bfdac 100644 --- a/include/linux/math64.h +++ b/include/linux/math64.h @@ -133,4 +133,34 @@ __iter_div_u64_rem(u64 dividend, u32 divisor, u64 *remainder) return ret; } +#if defined(CONFIG_ARCH_SUPPORTS_INT128) && defined(__SIZEOF_INT128__) + +#ifndef mul_u64_u32_shr +static inline u64 mul_u64_u32_shr(u64 a, u32 mul, unsigned int shift) +{ + return (u64)(((unsigned __int128)a * mul) >> shift); +} +#endif /* mul_u64_u32_shr */ + +#else + +#ifndef mul_u64_u32_shr +static inline u64 mul_u64_u32_shr(u64 a, u32 mul, unsigned int shift) +{ + u32 ah, al; + u64 ret; + + al = a; + ah = a >> 32; + + ret = ((u64)al * mul) >> shift; + if (ah) + ret += ((u64)ah * mul) << (32 - shift); + + return ret; +} +#endif /* mul_u64_u32_shr */ + +#endif + #endif /* _LINUX_MATH64_H */ diff --git a/init/Kconfig b/init/Kconfig index 79383d3aa5dc..4e5d96ab2034 100644 --- a/init/Kconfig +++ b/init/Kconfig @@ -809,6 +809,12 @@ config GENERIC_SCHED_CLOCK config ARCH_SUPPORTS_NUMA_BALANCING bool +# +# For architectures that know their GCC __int128 support is sound +# +config ARCH_SUPPORTS_INT128 + bool + # For architectures that (ab)use NUMA to represent different memory regions # all cpu-local but of different latencies, such as SuperH. # -- cgit v1.3-14-g43fede From 9dbdb155532395ba000c5d5d187658b0e17e529f Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Mon, 18 Nov 2013 18:27:06 +0100 Subject: sched/fair: Rework sched_fair time accounting Christian suffers from a bad BIOS that wrecks his i5's TSC sync. This results in him occasionally seeing time going backwards - which crashes the scheduler ... Most of our time accounting can actually handle that except the most common one; the tick time update of sched_fair. There is a further problem with that code; previously we assumed that because we get a tick every TICK_NSEC our time delta could never exceed 32bits and math was simpler. However, ever since Frederic managed to get NO_HZ_FULL merged; this is no longer the case since now a task can run for a long time indeed without getting a tick. It only takes about ~4.2 seconds to overflow our u32 in nanoseconds. This means we not only need to better deal with time going backwards; but also means we need to be able to deal with large deltas. This patch reworks the entire code and uses mul_u64_u32_shr() as proposed by Andy a long while ago. We express our virtual time scale factor in a u32 multiplier and shift right and the 32bit mul_u64_u32_shr() implementation reduces to a single 32x32->64 multiply if the time delta is still short (common case). For 64bit a 64x64->128 multiply can be used if ARCH_SUPPORTS_INT128. Reported-and-Tested-by: Christian Engelmayer Signed-off-by: Peter Zijlstra Cc: fweisbec@gmail.com Cc: Paul Turner Cc: Stanislaw Gruszka Cc: Andy Lutomirski Cc: Linus Torvalds Cc: Andrew Morton Link: http://lkml.kernel.org/r/20131118172706.GI3866@twins.programming.kicks-ass.net Signed-off-by: Ingo Molnar --- include/linux/sched.h | 3 +- kernel/sched/fair.c | 144 ++++++++++++++++++++++---------------------------- 2 files changed, 66 insertions(+), 81 deletions(-) (limited to 'include/linux') diff --git a/include/linux/sched.h b/include/linux/sched.h index 96d674ba3876..53f97eb8dbc7 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -930,7 +930,8 @@ struct pipe_inode_info; struct uts_namespace; struct load_weight { - unsigned long weight, inv_weight; + unsigned long weight; + u32 inv_weight; }; struct sched_avg { diff --git a/kernel/sched/fair.c b/kernel/sched/fair.c index fd773ade1a31..9030da7bcb15 100644 --- a/kernel/sched/fair.c +++ b/kernel/sched/fair.c @@ -178,59 +178,61 @@ void sched_init_granularity(void) update_sysctl(); } -#if BITS_PER_LONG == 32 -# define WMULT_CONST (~0UL) -#else -# define WMULT_CONST (1UL << 32) -#endif - +#define WMULT_CONST (~0U) #define WMULT_SHIFT 32 -/* - * Shift right and round: - */ -#define SRR(x, y) (((x) + (1UL << ((y) - 1))) >> (y)) +static void __update_inv_weight(struct load_weight *lw) +{ + unsigned long w; + + if (likely(lw->inv_weight)) + return; + + w = scale_load_down(lw->weight); + + if (BITS_PER_LONG > 32 && unlikely(w >= WMULT_CONST)) + lw->inv_weight = 1; + else if (unlikely(!w)) + lw->inv_weight = WMULT_CONST; + else + lw->inv_weight = WMULT_CONST / w; +} /* - * delta *= weight / lw + * delta_exec * weight / lw.weight + * OR + * (delta_exec * (weight * lw->inv_weight)) >> WMULT_SHIFT + * + * Either weight := NICE_0_LOAD and lw \e prio_to_wmult[], in which case + * we're guaranteed shift stays positive because inv_weight is guaranteed to + * fit 32 bits, and NICE_0_LOAD gives another 10 bits; therefore shift >= 22. + * + * Or, weight =< lw.weight (because lw.weight is the runqueue weight), thus + * weight/lw.weight <= 1, and therefore our shift will also be positive. */ -static unsigned long -calc_delta_mine(unsigned long delta_exec, unsigned long weight, - struct load_weight *lw) +static u64 __calc_delta(u64 delta_exec, unsigned long weight, struct load_weight *lw) { - u64 tmp; + u64 fact = scale_load_down(weight); + int shift = WMULT_SHIFT; - /* - * weight can be less than 2^SCHED_LOAD_RESOLUTION for task group sched - * entities since MIN_SHARES = 2. Treat weight as 1 if less than - * 2^SCHED_LOAD_RESOLUTION. - */ - if (likely(weight > (1UL << SCHED_LOAD_RESOLUTION))) - tmp = (u64)delta_exec * scale_load_down(weight); - else - tmp = (u64)delta_exec; + __update_inv_weight(lw); - if (!lw->inv_weight) { - unsigned long w = scale_load_down(lw->weight); - - if (BITS_PER_LONG > 32 && unlikely(w >= WMULT_CONST)) - lw->inv_weight = 1; - else if (unlikely(!w)) - lw->inv_weight = WMULT_CONST; - else - lw->inv_weight = WMULT_CONST / w; + if (unlikely(fact >> 32)) { + while (fact >> 32) { + fact >>= 1; + shift--; + } } - /* - * Check whether we'd overflow the 64-bit multiplication: - */ - if (unlikely(tmp > WMULT_CONST)) - tmp = SRR(SRR(tmp, WMULT_SHIFT/2) * lw->inv_weight, - WMULT_SHIFT/2); - else - tmp = SRR(tmp * lw->inv_weight, WMULT_SHIFT); + /* hint to use a 32x32->64 mul */ + fact = (u64)(u32)fact * lw->inv_weight; + + while (fact >> 32) { + fact >>= 1; + shift--; + } - return (unsigned long)min(tmp, (u64)(unsigned long)LONG_MAX); + return mul_u64_u32_shr(delta_exec, fact, shift); } @@ -443,7 +445,7 @@ find_matching_se(struct sched_entity **se, struct sched_entity **pse) #endif /* CONFIG_FAIR_GROUP_SCHED */ static __always_inline -void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec); +void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec); /************************************************************** * Scheduling class tree data structure manipulation methods: @@ -612,11 +614,10 @@ int sched_proc_update_handler(struct ctl_table *table, int write, /* * delta /= w */ -static inline unsigned long -calc_delta_fair(unsigned long delta, struct sched_entity *se) +static inline u64 calc_delta_fair(u64 delta, struct sched_entity *se) { if (unlikely(se->load.weight != NICE_0_LOAD)) - delta = calc_delta_mine(delta, NICE_0_LOAD, &se->load); + delta = __calc_delta(delta, NICE_0_LOAD, &se->load); return delta; } @@ -665,7 +666,7 @@ static u64 sched_slice(struct cfs_rq *cfs_rq, struct sched_entity *se) update_load_add(&lw, se->load.weight); load = &lw; } - slice = calc_delta_mine(slice, se->load.weight, load); + slice = __calc_delta(slice, se->load.weight, load); } return slice; } @@ -703,47 +704,32 @@ void init_task_runnable_average(struct task_struct *p) #endif /* - * Update the current task's runtime statistics. Skip current tasks that - * are not in our scheduling class. + * Update the current task's runtime statistics. */ -static inline void -__update_curr(struct cfs_rq *cfs_rq, struct sched_entity *curr, - unsigned long delta_exec) -{ - unsigned long delta_exec_weighted; - - schedstat_set(curr->statistics.exec_max, - max((u64)delta_exec, curr->statistics.exec_max)); - - curr->sum_exec_runtime += delta_exec; - schedstat_add(cfs_rq, exec_clock, delta_exec); - delta_exec_weighted = calc_delta_fair(delta_exec, curr); - - curr->vruntime += delta_exec_weighted; - update_min_vruntime(cfs_rq); -} - static void update_curr(struct cfs_rq *cfs_rq) { struct sched_entity *curr = cfs_rq->curr; u64 now = rq_clock_task(rq_of(cfs_rq)); - unsigned long delta_exec; + u64 delta_exec; if (unlikely(!curr)) return; - /* - * Get the amount of time the current task was running - * since the last time we changed load (this cannot - * overflow on 32 bits): - */ - delta_exec = (unsigned long)(now - curr->exec_start); - if (!delta_exec) + delta_exec = now - curr->exec_start; + if (unlikely((s64)delta_exec <= 0)) return; - __update_curr(cfs_rq, curr, delta_exec); curr->exec_start = now; + schedstat_set(curr->statistics.exec_max, + max(delta_exec, curr->statistics.exec_max)); + + curr->sum_exec_runtime += delta_exec; + schedstat_add(cfs_rq, exec_clock, delta_exec); + + curr->vruntime += calc_delta_fair(delta_exec, curr); + update_min_vruntime(cfs_rq); + if (entity_is_task(curr)) { struct task_struct *curtask = task_of(curr); @@ -3015,8 +3001,7 @@ static void expire_cfs_rq_runtime(struct cfs_rq *cfs_rq) } } -static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, - unsigned long delta_exec) +static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) { /* dock delta_exec before expiring quota (as it could span periods) */ cfs_rq->runtime_remaining -= delta_exec; @@ -3034,7 +3019,7 @@ static void __account_cfs_rq_runtime(struct cfs_rq *cfs_rq, } static __always_inline -void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, unsigned long delta_exec) +void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) { if (!cfs_bandwidth_used() || !cfs_rq->runtime_enabled) return; @@ -3574,8 +3559,7 @@ static inline u64 cfs_rq_clock_task(struct cfs_rq *cfs_rq) return rq_clock_task(rq_of(cfs_rq)); } -static void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, - unsigned long delta_exec) {} +static void account_cfs_rq_runtime(struct cfs_rq *cfs_rq, u64 delta_exec) {} static void check_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} static void check_enqueue_throttle(struct cfs_rq *cfs_rq) {} static __always_inline void return_cfs_rq_runtime(struct cfs_rq *cfs_rq) {} -- cgit v1.3-14-g43fede