diff options
Diffstat (limited to 'kernel')
82 files changed, 3855 insertions, 2364 deletions
diff --git a/kernel/Kconfig.preempt b/kernel/Kconfig.preempt index bf987b95b35..24e7cb0ba26 100644 --- a/kernel/Kconfig.preempt +++ b/kernel/Kconfig.preempt @@ -35,6 +35,7 @@ config PREEMPT_VOLUNTARY config PREEMPT bool "Preemptible Kernel (Low-Latency Desktop)" + select PREEMPT_COUNT help This option reduces the latency of the kernel by making all kernel code (that is not executing in a critical section) @@ -52,3 +53,5 @@ config PREEMPT endchoice +config PREEMPT_COUNT + bool
\ No newline at end of file diff --git a/kernel/Makefile b/kernel/Makefile index 2d64cfcc8b4..d06467fc8f7 100644 --- a/kernel/Makefile +++ b/kernel/Makefile @@ -125,11 +125,10 @@ targets += config_data.gz $(obj)/config_data.gz: $(KCONFIG_CONFIG) FORCE $(call if_changed,gzip) -quiet_cmd_ikconfiggz = IKCFG $@ - cmd_ikconfiggz = (echo "static const char kernel_config_data[] __used = MAGIC_START"; cat $< | scripts/bin2c; echo "MAGIC_END;") > $@ + filechk_ikconfiggz = (echo "static const char kernel_config_data[] __used = MAGIC_START"; cat $< | scripts/bin2c; echo "MAGIC_END;") targets += config_data.h $(obj)/config_data.h: $(obj)/config_data.gz FORCE - $(call if_changed,ikconfiggz) + $(call filechk,ikconfiggz) $(obj)/time.o: $(obj)/timeconst.h diff --git a/kernel/async.c b/kernel/async.c index cd9dbb913c7..d5fe7af0de2 100644 --- a/kernel/async.c +++ b/kernel/async.c @@ -49,12 +49,13 @@ asynchronous and synchronous parts of the kernel. */ #include <linux/async.h> +#include <linux/atomic.h> +#include <linux/ktime.h> #include <linux/module.h> #include <linux/wait.h> #include <linux/sched.h> #include <linux/slab.h> #include <linux/workqueue.h> -#include <asm/atomic.h> static async_cookie_t next_cookie = 1; @@ -128,7 +129,8 @@ static void async_run_entry_fn(struct work_struct *work) /* 2) run (and print duration) */ if (initcall_debug && system_state == SYSTEM_BOOTING) { - printk("calling %lli_%pF @ %i\n", (long long)entry->cookie, + printk(KERN_DEBUG "calling %lli_%pF @ %i\n", + (long long)entry->cookie, entry->func, task_pid_nr(current)); calltime = ktime_get(); } @@ -136,7 +138,7 @@ static void async_run_entry_fn(struct work_struct *work) if (initcall_debug && system_state == SYSTEM_BOOTING) { rettime = ktime_get(); delta = ktime_sub(rettime, calltime); - printk("initcall %lli_%pF returned 0 after %lld usecs\n", + printk(KERN_DEBUG "initcall %lli_%pF returned 0 after %lld usecs\n", (long long)entry->cookie, entry->func, (long long)ktime_to_ns(delta) >> 10); @@ -270,7 +272,7 @@ void async_synchronize_cookie_domain(async_cookie_t cookie, ktime_t starttime, delta, endtime; if (initcall_debug && system_state == SYSTEM_BOOTING) { - printk("async_waiting @ %i\n", task_pid_nr(current)); + printk(KERN_DEBUG "async_waiting @ %i\n", task_pid_nr(current)); starttime = ktime_get(); } @@ -280,7 +282,7 @@ void async_synchronize_cookie_domain(async_cookie_t cookie, endtime = ktime_get(); delta = ktime_sub(endtime, starttime); - printk("async_continuing @ %i after %lli usec\n", + printk(KERN_DEBUG "async_continuing @ %i after %lli usec\n", task_pid_nr(current), (long long)ktime_to_ns(delta) >> 10); } diff --git a/kernel/audit.c b/kernel/audit.c index 93950031706..52501b5d490 100644 --- a/kernel/audit.c +++ b/kernel/audit.c @@ -55,6 +55,9 @@ #include <net/sock.h> #include <net/netlink.h> #include <linux/skbuff.h> +#ifdef CONFIG_SECURITY +#include <linux/security.h> +#endif #include <linux/netlink.h> #include <linux/freezer.h> #include <linux/tty.h> @@ -1502,6 +1505,32 @@ void audit_log(struct audit_context *ctx, gfp_t gfp_mask, int type, } } +#ifdef CONFIG_SECURITY +/** + * audit_log_secctx - Converts and logs SELinux context + * @ab: audit_buffer + * @secid: security number + * + * This is a helper function that calls security_secid_to_secctx to convert + * secid to secctx and then adds the (converted) SELinux context to the audit + * log by calling audit_log_format, thus also preventing leak of internal secid + * to userspace. If secid cannot be converted audit_panic is called. + */ +void audit_log_secctx(struct audit_buffer *ab, u32 secid) +{ + u32 len; + char *secctx; + + if (security_secid_to_secctx(secid, &secctx, &len)) { + audit_panic("Cannot convert secid to context"); + } else { + audit_log_format(ab, " obj=%s", secctx); + security_release_secctx(secctx, len); + } +} +EXPORT_SYMBOL(audit_log_secctx); +#endif + EXPORT_SYMBOL(audit_log_start); EXPORT_SYMBOL(audit_log_end); EXPORT_SYMBOL(audit_log_format); diff --git a/kernel/audit_tree.c b/kernel/audit_tree.c index e99dda04b12..5bf0790497e 100644 --- a/kernel/audit_tree.c +++ b/kernel/audit_tree.c @@ -93,16 +93,10 @@ static inline void get_tree(struct audit_tree *tree) atomic_inc(&tree->count); } -static void __put_tree(struct rcu_head *rcu) -{ - struct audit_tree *tree = container_of(rcu, struct audit_tree, head); - kfree(tree); -} - static inline void put_tree(struct audit_tree *tree) { if (atomic_dec_and_test(&tree->count)) - call_rcu(&tree->head, __put_tree); + kfree_rcu(tree, head); } /* to avoid bringing the entire thing in audit.h */ diff --git a/kernel/cgroup.c b/kernel/cgroup.c index 2731d115d72..a63507b92ca 100644 --- a/kernel/cgroup.c +++ b/kernel/cgroup.c @@ -1697,7 +1697,6 @@ int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen) { char *start; struct dentry *dentry = rcu_dereference_check(cgrp->dentry, - rcu_read_lock_held() || cgroup_lock_is_held()); if (!dentry || cgrp == dummytop) { @@ -1723,7 +1722,6 @@ int cgroup_path(const struct cgroup *cgrp, char *buf, int buflen) break; dentry = rcu_dereference_check(cgrp->dentry, - rcu_read_lock_held() || cgroup_lock_is_held()); if (!cgrp->parent) continue; @@ -3542,7 +3540,8 @@ static int cgroup_write_event_control(struct cgroup *cgrp, struct cftype *cft, } /* the process need read permission on control file */ - ret = file_permission(cfile, MAY_READ); + /* AV: shouldn't we check that it's been opened for read instead? */ + ret = inode_permission(cfile->f_path.dentry->d_inode, MAY_READ); if (ret < 0) goto fail; @@ -4813,8 +4812,7 @@ unsigned short css_id(struct cgroup_subsys_state *css) * on this or this is under rcu_read_lock(). Once css->id is allocated, * it's unchanged until freed. */ - cssid = rcu_dereference_check(css->id, - rcu_read_lock_held() || atomic_read(&css->refcnt)); + cssid = rcu_dereference_check(css->id, atomic_read(&css->refcnt)); if (cssid) return cssid->id; @@ -4826,8 +4824,7 @@ unsigned short css_depth(struct cgroup_subsys_state *css) { struct css_id *cssid; - cssid = rcu_dereference_check(css->id, - rcu_read_lock_held() || atomic_read(&css->refcnt)); + cssid = rcu_dereference_check(css->id, atomic_read(&css->refcnt)); if (cssid) return cssid->depth; diff --git a/kernel/compat.c b/kernel/compat.c index fc9eb093acd..18197ae2d46 100644 --- a/kernel/compat.c +++ b/kernel/compat.c @@ -890,6 +890,7 @@ sigset_from_compat (sigset_t *set, compat_sigset_t *compat) case 1: set->sig[0] = compat->sig[0] | (((long)compat->sig[1]) << 32 ); } } +EXPORT_SYMBOL_GPL(sigset_from_compat); asmlinkage long compat_sys_rt_sigtimedwait (compat_sigset_t __user *uthese, diff --git a/kernel/delayacct.c b/kernel/delayacct.c index ead9b610aa7..418b3f7053a 100644 --- a/kernel/delayacct.c +++ b/kernel/delayacct.c @@ -19,8 +19,10 @@ #include <linux/time.h> #include <linux/sysctl.h> #include <linux/delayacct.h> +#include <linux/module.h> int delayacct_on __read_mostly = 1; /* Delay accounting turned on/off */ +EXPORT_SYMBOL_GPL(delayacct_on); struct kmem_cache *delayacct_cache; static int __init delayacct_setup_disable(char *str) diff --git a/kernel/events/Makefile b/kernel/events/Makefile index 1ce23d3d839..89e5e8aa4c3 100644 --- a/kernel/events/Makefile +++ b/kernel/events/Makefile @@ -2,5 +2,5 @@ ifdef CONFIG_FUNCTION_TRACER CFLAGS_REMOVE_core.o = -pg endif -obj-y := core.o +obj-y := core.o ring_buffer.o obj-$(CONFIG_HAVE_HW_BREAKPOINT) += hw_breakpoint.o diff --git a/kernel/events/core.c b/kernel/events/core.c index d863b3c057b..b8785e26ee1 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -36,6 +36,8 @@ #include <linux/ftrace_event.h> #include <linux/hw_breakpoint.h> +#include "internal.h" + #include <asm/irq_regs.h> struct remote_function_call { @@ -200,6 +202,22 @@ __get_cpu_context(struct perf_event_context *ctx) return this_cpu_ptr(ctx->pmu->pmu_cpu_context); } +static void perf_ctx_lock(struct perf_cpu_context *cpuctx, + struct perf_event_context *ctx) +{ + raw_spin_lock(&cpuctx->ctx.lock); + if (ctx) + raw_spin_lock(&ctx->lock); +} + +static void perf_ctx_unlock(struct perf_cpu_context *cpuctx, + struct perf_event_context *ctx) +{ + if (ctx) + raw_spin_unlock(&ctx->lock); + raw_spin_unlock(&cpuctx->ctx.lock); +} + #ifdef CONFIG_CGROUP_PERF /* @@ -340,11 +358,8 @@ void perf_cgroup_switch(struct task_struct *task, int mode) rcu_read_lock(); list_for_each_entry_rcu(pmu, &pmus, entry) { - cpuctx = this_cpu_ptr(pmu->pmu_cpu_context); - perf_pmu_disable(cpuctx->ctx.pmu); - /* * perf_cgroup_events says at least one * context on this CPU has cgroup events. @@ -353,6 +368,8 @@ void perf_cgroup_switch(struct task_struct *task, int mode) * events for a context. */ if (cpuctx->ctx.nr_cgroups > 0) { + perf_ctx_lock(cpuctx, cpuctx->task_ctx); + perf_pmu_disable(cpuctx->ctx.pmu); if (mode & PERF_CGROUP_SWOUT) { cpu_ctx_sched_out(cpuctx, EVENT_ALL); @@ -372,9 +389,9 @@ void perf_cgroup_switch(struct task_struct *task, int mode) cpuctx->cgrp = perf_cgroup_from_task(task); cpu_ctx_sched_in(cpuctx, EVENT_ALL, task); } + perf_pmu_enable(cpuctx->ctx.pmu); + perf_ctx_unlock(cpuctx, cpuctx->task_ctx); } - - perf_pmu_enable(cpuctx->ctx.pmu); } rcu_read_unlock(); @@ -731,6 +748,7 @@ static u64 perf_event_time(struct perf_event *event) /* * Update the total_time_enabled and total_time_running fields for a event. + * The caller of this function needs to hold the ctx->lock. */ static void update_event_times(struct perf_event *event) { @@ -1105,6 +1123,10 @@ static int __perf_remove_from_context(void *info) raw_spin_lock(&ctx->lock); event_sched_out(event, cpuctx, ctx); list_del_event(event, ctx); + if (!ctx->nr_events && cpuctx->task_ctx == ctx) { + ctx->is_active = 0; + cpuctx->task_ctx = NULL; + } raw_spin_unlock(&ctx->lock); return 0; @@ -1454,8 +1476,24 @@ static void add_event_to_ctx(struct perf_event *event, event->tstamp_stopped = tstamp; } -static void perf_event_context_sched_in(struct perf_event_context *ctx, - struct task_struct *tsk); +static void task_ctx_sched_out(struct perf_event_context *ctx); +static void +ctx_sched_in(struct perf_event_context *ctx, + struct perf_cpu_context *cpuctx, + enum event_type_t event_type, + struct task_struct *task); + +static void perf_event_sched_in(struct perf_cpu_context *cpuctx, + struct perf_event_context *ctx, + struct task_struct *task) +{ + cpu_ctx_sched_in(cpuctx, EVENT_PINNED, task); + if (ctx) + ctx_sched_in(ctx, cpuctx, EVENT_PINNED, task); + cpu_ctx_sched_in(cpuctx, EVENT_FLEXIBLE, task); + if (ctx) + ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE, task); +} /* * Cross CPU call to install and enable a performance event @@ -1466,20 +1504,37 @@ static int __perf_install_in_context(void *info) { struct perf_event *event = info; struct perf_event_context *ctx = event->ctx; - struct perf_event *leader = event->group_leader; struct perf_cpu_context *cpuctx = __get_cpu_context(ctx); - int err; + struct perf_event_context *task_ctx = cpuctx->task_ctx; + struct task_struct *task = current; + + perf_ctx_lock(cpuctx, task_ctx); + perf_pmu_disable(cpuctx->ctx.pmu); /* - * In case we're installing a new context to an already running task, - * could also happen before perf_event_task_sched_in() on architectures - * which do context switches with IRQs enabled. + * If there was an active task_ctx schedule it out. */ - if (ctx->task && !cpuctx->task_ctx) - perf_event_context_sched_in(ctx, ctx->task); + if (task_ctx) + task_ctx_sched_out(task_ctx); + + /* + * If the context we're installing events in is not the + * active task_ctx, flip them. + */ + if (ctx->task && task_ctx != ctx) { + if (task_ctx) + raw_spin_unlock(&task_ctx->lock); + raw_spin_lock(&ctx->lock); + task_ctx = ctx; + } + + if (task_ctx) { + cpuctx->task_ctx = task_ctx; + task = task_ctx->task; + } + + cpu_ctx_sched_out(cpuctx, EVENT_ALL); - raw_spin_lock(&ctx->lock); - ctx->is_active = 1; update_context_time(ctx); /* * update cgrp time only if current cgrp @@ -1490,43 +1545,13 @@ static int __perf_install_in_context(void *info) add_event_to_ctx(event, ctx); - if (!event_filter_match(event)) - goto unlock; - - /* - * Don't put the event on if it is disabled or if - * it is in a group and the group isn't on. - */ - if (event->state != PERF_EVENT_STATE_INACTIVE || - (leader != event && leader->state != PERF_EVENT_STATE_ACTIVE)) - goto unlock; - /* - * An exclusive event can't go on if there are already active - * hardware events, and no hardware event can go on if there - * is already an exclusive event on. + * Schedule everything back in */ - if (!group_can_go_on(event, cpuctx, 1)) - err = -EEXIST; - else - err = event_sched_in(event, cpuctx, ctx); - - if (err) { - /* - * This event couldn't go on. If it is in a group - * then we have to pull the whole group off. - * If the event group is pinned then put it in error state. - */ - if (leader != event) - group_sched_out(leader, cpuctx, ctx); - if (leader->attr.pinned) { - update_group_times(leader); - leader->state = PERF_EVENT_STATE_ERROR; - } - } + perf_event_sched_in(cpuctx, task_ctx, task); -unlock: - raw_spin_unlock(&ctx->lock); + perf_pmu_enable(cpuctx->ctx.pmu); + perf_ctx_unlock(cpuctx, task_ctx); return 0; } @@ -1739,7 +1764,7 @@ out: raw_spin_unlock_irq(&ctx->lock); } -static int perf_event_refresh(struct perf_event *event, int refresh) +int perf_event_refresh(struct perf_event *event, int refresh) { /* * not supported on inherited events @@ -1752,36 +1777,35 @@ static int perf_event_refresh(struct perf_event *event, int refresh) return 0; } +EXPORT_SYMBOL_GPL(perf_event_refresh); static void ctx_sched_out(struct perf_event_context *ctx, struct perf_cpu_context *cpuctx, enum event_type_t event_type) { struct perf_event *event; + int is_active = ctx->is_active; - raw_spin_lock(&ctx->lock); - perf_pmu_disable(ctx->pmu); - ctx->is_active = 0; + ctx->is_active &= ~event_type; if (likely(!ctx->nr_events)) - goto out; + return; + update_context_time(ctx); update_cgrp_time_from_cpuctx(cpuctx); - if (!ctx->nr_active) - goto out; + return; - if (event_type & EVENT_PINNED) { + perf_pmu_disable(ctx->pmu); + if ((is_active & EVENT_PINNED) && (event_type & EVENT_PINNED)) { list_for_each_entry(event, &ctx->pinned_groups, group_entry) group_sched_out(event, cpuctx, ctx); } - if (event_type & EVENT_FLEXIBLE) { + if ((is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE)) { list_for_each_entry(event, &ctx->flexible_groups, group_entry) group_sched_out(event, cpuctx, ctx); } -out: perf_pmu_enable(ctx->pmu); - raw_spin_unlock(&ctx->lock); } /* @@ -1929,8 +1953,10 @@ static void perf_event_context_sched_out(struct task_struct *task, int ctxn, rcu_read_unlock(); if (do_switch) { + raw_spin_lock(&ctx->lock); ctx_sched_out(ctx, cpuctx, EVENT_ALL); cpuctx->task_ctx = NULL; + raw_spin_unlock(&ctx->lock); } } @@ -1965,8 +1991,7 @@ void __perf_event_task_sched_out(struct task_struct *task, perf_cgroup_sched_out(task); } -static void task_ctx_sched_out(struct perf_event_context *ctx, - enum event_type_t event_type) +static void task_ctx_sched_out(struct perf_event_context *ctx) { struct perf_cpu_context *cpuctx = __get_cpu_context(ctx); @@ -1976,7 +2001,7 @@ static void task_ctx_sched_out(struct perf_event_context *ctx, if (WARN_ON_ONCE(ctx != cpuctx->task_ctx)) return; - ctx_sched_out(ctx, cpuctx, event_type); + ctx_sched_out(ctx, cpuctx, EVENT_ALL); cpuctx->task_ctx = NULL; } @@ -2055,11 +2080,11 @@ ctx_sched_in(struct perf_event_context *ctx, struct task_struct *task) { u64 now; + int is_active = ctx->is_active; - raw_spin_lock(&ctx->lock); - ctx->is_active = 1; + ctx->is_active |= event_type; if (likely(!ctx->nr_events)) - goto out; + return; now = perf_clock(); ctx->timestamp = now; @@ -2068,15 +2093,12 @@ ctx_sched_in(struct perf_event_context *ctx, * First go through the list and put on any pinned groups * in order to give them the best chance of going on. */ - if (event_type & EVENT_PINNED) + if (!(is_active & EVENT_PINNED) && (event_type & EVENT_PINNED)) ctx_pinned_sched_in(ctx, cpuctx); /* Then walk through the lower prio flexible groups */ - if (event_type & EVENT_FLEXIBLE) + if (!(is_active & EVENT_FLEXIBLE) && (event_type & EVENT_FLEXIBLE)) ctx_flexible_sched_in(ctx, cpuctx); - -out: - raw_spin_unlock(&ctx->lock); } static void cpu_ctx_sched_in(struct perf_cpu_context *cpuctx, @@ -2088,19 +2110,6 @@ static void cpu_ctx_sched_in(struct perf_cpu_context *cpuctx, ctx_sched_in(ctx, cpuctx, event_type, task); } -static void task_ctx_sched_in(struct perf_event_context *ctx, - enum event_type_t event_type) -{ - struct perf_cpu_context *cpuctx; - - cpuctx = __get_cpu_context(ctx); - if (cpuctx->task_ctx == ctx) - return; - - ctx_sched_in(ctx, cpuctx, event_type, NULL); - cpuctx->task_ctx = ctx; -} - static void perf_event_context_sched_in(struct perf_event_context *ctx, struct task_struct *task) { @@ -2110,6 +2119,7 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx, if (cpuctx->task_ctx == ctx) return; + perf_ctx_lock(cpuctx, ctx); perf_pmu_disable(ctx->pmu); /* * We want to keep the following priority order: @@ -2118,18 +2128,18 @@ static void perf_event_context_sched_in(struct perf_event_context *ctx, */ cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE); - ctx_sched_in(ctx, cpuctx, EVENT_PINNED, task); - cpu_ctx_sched_in(cpuctx, EVENT_FLEXIBLE, task); - ctx_sched_in(ctx, cpuctx, EVENT_FLEXIBLE, task); + perf_event_sched_in(cpuctx, ctx, task); cpuctx->task_ctx = ctx; + perf_pmu_enable(ctx->pmu); + perf_ctx_unlock(cpuctx, ctx); + /* * Since these rotations are per-cpu, we need to ensure the * cpu-context we got scheduled on is actually rotating. */ perf_pmu_rotate_start(ctx->pmu); - perf_pmu_enable(ctx->pmu); } /* @@ -2269,7 +2279,6 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx, u64 period) u64 interrupts, now; s64 delta; - raw_spin_lock(&ctx->lock); list_for_each_entry_rcu(event, &ctx->event_list, event_entry) { if (event->state != PERF_EVENT_STATE_ACTIVE) continue; @@ -2301,7 +2310,6 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx, u64 period) if (delta > 0) perf_adjust_period(event, period, delta); } - raw_spin_unlock(&ctx->lock); } /* @@ -2309,16 +2317,12 @@ static void perf_ctx_adjust_freq(struct perf_event_context *ctx, u64 period) */ static void rotate_ctx(struct perf_event_context *ctx) { - raw_spin_lock(&ctx->lock); - /* * Rotate the first entry last of non-pinned groups. Rotation might be * disabled by the inheritance code. */ if (!ctx->rotate_disable) list_rotate_left(&ctx->flexible_groups); - - raw_spin_unlock(&ctx->lock); } /* @@ -2345,6 +2349,7 @@ static void perf_rotate_context(struct perf_cpu_context *cpuctx) rotate = 1; } + perf_ctx_lock(cpuctx, cpuctx->task_ctx); perf_pmu_disable(cpuctx->ctx.pmu); perf_ctx_adjust_freq(&cpuctx->ctx, interval); if (ctx) @@ -2355,21 +2360,20 @@ static void perf_rotate_context(struct perf_cpu_context *cpuctx) cpu_ctx_sched_out(cpuctx, EVENT_FLEXIBLE); if (ctx) - task_ctx_sched_out(ctx, EVENT_FLEXIBLE); + ctx_sched_out(ctx, cpuctx, EVENT_FLEXIBLE); rotate_ctx(&cpuctx->ctx); if (ctx) rotate_ctx(ctx); - cpu_ctx_sched_in(cpuctx, EVENT_FLEXIBLE, current); - if (ctx) - task_ctx_sched_in(ctx, EVENT_FLEXIBLE); + perf_event_sched_in(cpuctx, ctx, current); done: if (remove) list_del_init(&cpuctx->rotation_list); perf_pmu_enable(cpuctx->ctx.pmu); + perf_ctx_unlock(cpuctx, cpuctx->task_ctx); } void perf_event_task_tick(void) @@ -2424,9 +2428,9 @@ static void perf_event_enable_on_exec(struct perf_event_context *ctx) * in. */ perf_cgroup_sched_out(current); - task_ctx_sched_out(ctx, EVENT_ALL); raw_spin_lock(&ctx->lock); + task_ctx_sched_out(ctx); list_for_each_entry(event, &ctx->pinned_groups, group_entry) { ret = event_enable_on_exec(event, ctx); @@ -2835,16 +2839,12 @@ retry: unclone_ctx(ctx); ++ctx->pin_count; raw_spin_unlock_irqrestore(&ctx->lock, flags); - } - - if (!ctx) { + } else { ctx = alloc_perf_context(pmu, task); err = -ENOMEM; if (!ctx) goto errout; - get_ctx(ctx); - err = 0; mutex_lock(&task->perf_event_mutex); /* @@ -2856,14 +2856,14 @@ retry: else if (task->perf_event_ctxp[ctxn]) err = -EAGAIN; else { + get_ctx(ctx); ++ctx->pin_count; rcu_assign_pointer(task->perf_event_ctxp[ctxn], ctx); } mutex_unlock(&task->perf_event_mutex); if (unlikely(err)) { - put_task_struct(task); - kfree(ctx); + put_ctx(ctx); if (err == -EAGAIN) goto retry; @@ -2890,7 +2890,7 @@ static void free_event_rcu(struct rcu_head *head) kfree(event); } -static void perf_buffer_put(struct perf_buffer *buffer); +static void ring_buffer_put(struct ring_buffer *rb); static void free_event(struct perf_event *event) { @@ -2913,9 +2913,9 @@ static void free_event(struct perf_event *event) } } - if (event->buffer) { - perf_buffer_put(event->buffer); - event->buffer = NULL; + if (event->rb) { + ring_buffer_put(event->rb); + event->rb = NULL; } if (is_cgroup_event(event)) @@ -2934,12 +2934,6 @@ int perf_event_release_kernel(struct perf_event *event) { struct perf_event_context *ctx = event->ctx; - /* - * Remove from the PMU, can't get re-enabled since we got - * here because the last ref went. - */ - perf_event_disable(event); - WARN_ON_ONCE(ctx->parent_ctx); /* * There are two ways this annotation is useful: @@ -2956,8 +2950,8 @@ int perf_event_release_kernel(struct perf_event *event) mutex_lock_nested(&ctx->mutex, SINGLE_DEPTH_NESTING); raw_spin_lock_irq(&ctx->lock); perf_group_detach(event); - list_del_event(event, ctx); raw_spin_unlock_irq(&ctx->lock); + perf_remove_from_context(event); mutex_unlock(&ctx->mutex); free_event(event); @@ -3149,13 +3143,13 @@ perf_read(struct file *file, char __user *buf, size_t count, loff_t *ppos) static unsigned int perf_poll(struct file *file, poll_table *wait) { struct perf_event *event = file->private_data; - struct perf_buffer *buffer; + struct ring_buffer *rb; unsigned int events = POLL_HUP; rcu_read_lock(); - buffer = rcu_dereference(event->buffer); - if (buffer) - events = atomic_xchg(&buffer->poll, 0); + rb = rcu_dereference(event->rb); + if (rb) + events = atomic_xchg(&rb->poll, 0); rcu_read_unlock(); poll_wait(file, &event->waitq, wait); @@ -3358,6 +3352,18 @@ static int perf_event_index(struct perf_event *event) return event->hw.idx + 1 - PERF_EVENT_INDEX_OFFSET; } +static void calc_timer_values(struct perf_event *event, + u64 *running, + u64 *enabled) +{ + u64 now, ctx_time; + + now = perf_clock(); + ctx_time = event->shadow_ctx_time + now; + *enabled = ctx_time - event->tstamp_enabled; + *running = ctx_time - event->tstamp_running; +} + /* * Callers need to ensure there can be no nesting of this function, otherwise * the seqlock logic goes bad. We can not serialize this because the arch @@ -3366,14 +3372,25 @@ static int perf_event_index(struct perf_event *event) void perf_event_update_userpage(struct perf_event *event) { struct perf_event_mmap_page *userpg; - struct perf_buffer *buffer; + struct ring_buffer *rb; + u64 enabled, running; rcu_read_lock(); - buffer = rcu_dereference(event->buffer); - if (!buffer) + /* + * compute total_time_enabled, total_time_running + * based on snapshot values taken when the event + * was last scheduled in. + * + * we cannot simply called update_context_time() + * because of locking issue as we can be called in + * NMI context + */ + calc_timer_values(event, &enabled, &running); + rb = rcu_dereference(event->rb); + if (!rb) goto unlock; - userpg = buffer->user_page; + userpg = rb->user_page; /* * Disable preemption so as to not let the corresponding user-space @@ -3387,10 +3404,10 @@ void perf_event_update_userpage(struct perf_event *event) if (event->state == PERF_EVENT_STATE_ACTIVE) userpg->offset -= local64_read(&event->hw.prev_count); - userpg->time_enabled = event->total_time_enabled + + userpg->time_enabled = enabled + atomic64_read(&event->child_total_time_enabled); - userpg->time_running = event->total_time_running + + userpg->time_running = running + atomic64_read(&event->child_total_time_running); barrier(); @@ -3400,220 +3417,10 @@ unlock: rcu_read_unlock(); } -static unsigned long perf_data_size(struct perf_buffer *buffer); - -static void -perf_buffer_init(struct perf_buffer *buffer, long watermark, int flags) -{ - long max_size = perf_data_size(buffer); - - if (watermark) - buffer->watermark = min(max_size, watermark); - - if (!buffer->watermark) - buffer->watermark = max_size / 2; - - if (flags & PERF_BUFFER_WRITABLE) - buffer->writable = 1; - - atomic_set(&buffer->refcount, 1); -} - -#ifndef CONFIG_PERF_USE_VMALLOC - -/* - * Back perf_mmap() with regular GFP_KERNEL-0 pages. - */ - -static struct page * -perf_mmap_to_page(struct perf_buffer *buffer, unsigned long pgoff) -{ - if (pgoff > buffer->nr_pages) - return NULL; - - if (pgoff == 0) - return virt_to_page(buffer->user_page); - - return virt_to_page(buffer->data_pages[pgoff - 1]); -} - -static void *perf_mmap_alloc_page(int cpu) -{ - struct page *page; - int node; - - node = (cpu == -1) ? cpu : cpu_to_node(cpu); - page = alloc_pages_node(node, GFP_KERNEL | __GFP_ZERO, 0); - if (!page) - return NULL; - - return page_address(page); -} - -static struct perf_buffer * -perf_buffer_alloc(int nr_pages, long watermark, int cpu, int flags) -{ - struct perf_buffer *buffer; - unsigned long size; - int i; - - size = sizeof(struct perf_buffer); - size += nr_pages * sizeof(void *); - - buffer = kzalloc(size, GFP_KERNEL); - if (!buffer) - goto fail; - - buffer->user_page = perf_mmap_alloc_page(cpu); - if (!buffer->user_page) - goto fail_user_page; - - for (i = 0; i < nr_pages; i++) { - buffer->data_pages[i] = perf_mmap_alloc_page(cpu); - if (!buffer->data_pages[i]) - goto fail_data_pages; - } - - buffer->nr_pages = nr_pages; - - perf_buffer_init(buffer, watermark, flags); - - return buffer; - -fail_data_pages: - for (i--; i >= 0; i--) - free_page((unsigned long)buffer->data_pages[i]); - - free_page((unsigned long)buffer->user_page); - -fail_user_page: - kfree(buffer); - -fail: - return NULL; -} - -static void perf_mmap_free_page(unsigned long addr) -{ - struct page *page = virt_to_page((void *)addr); - - page->mapping = NULL; - __free_page(page); -} - -static void perf_buffer_free(struct perf_buffer *buffer) -{ - int i; - - perf_mmap_free_page((unsigned long)buffer->user_page); - for (i = 0; i < buffer->nr_pages; i++) - perf_mmap_free_page((unsigned long)buffer->data_pages[i]); - kfree(buffer); -} - -static inline int page_order(struct perf_buffer *buffer) -{ - return 0; -} - -#else - -/* - * Back perf_mmap() with vmalloc memory. - * - * Required for architectures that have d-cache aliasing issues. - */ - -static inline int page_order(struct perf_buffer *buffer) -{ - return buffer->page_order; -} - -static struct page * -perf_mmap_to_page(struct perf_buffer *buffer, unsigned long pgoff) -{ - if (pgoff > (1UL << page_order(buffer))) - return NULL; - - return vmalloc_to_page((void *)buffer->user_page + pgoff * PAGE_SIZE); -} - -static void perf_mmap_unmark_page(void *addr) -{ - struct page *page = vmalloc_to_page(addr); - - page->mapping = NULL; -} - -static void perf_buffer_free_work(struct work_struct *work) -{ - struct perf_buffer *buffer; - void *base; - int i, nr; - - buffer = container_of(work, struct perf_buffer, work); - nr = 1 << page_order(buffer); - - base = buffer->user_page; - for (i = 0; i < nr + 1; i++) - perf_mmap_unmark_page(base + (i * PAGE_SIZE)); - - vfree(base); - kfree(buffer); -} - -static void perf_buffer_free(struct perf_buffer *buffer) -{ - schedule_work(&buffer->work); -} - -static struct perf_buffer * -perf_buffer_alloc(int nr_pages, long watermark, int cpu, int flags) -{ - struct perf_buffer *buffer; - unsigned long size; - void *all_buf; - - size = sizeof(struct perf_buffer); - size += sizeof(void *); - - buffer = kzalloc(size, GFP_KERNEL); - if (!buffer) - goto fail; - - INIT_WORK(&buffer->work, perf_buffer_free_work); - - all_buf = vmalloc_user((nr_pages + 1) * PAGE_SIZE); - if (!all_buf) - goto fail_all_buf; - - buffer->user_page = all_buf; - buffer->data_pages[0] = all_buf + PAGE_SIZE; - buffer->page_order = ilog2(nr_pages); - buffer->nr_pages = 1; - - perf_buffer_init(buffer, watermark, flags); - - return buffer; - -fail_all_buf: - kfree(buffer); - -fail: - return NULL; -} - -#endif - -static unsigned long perf_data_size(struct perf_buffer *buffer) -{ - return buffer->nr_pages << (PAGE_SHIFT + page_order(buffer)); -} - static int perf_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) { struct perf_event *event = vma->vm_file->private_data; - struct perf_buffer *buffer; + struct ring_buffer *rb; int ret = VM_FAULT_SIGBUS; if (vmf->flags & FAULT_FLAG_MKWRITE) { @@ -3623,14 +3430,14 @@ static int perf_mmap_fault(struct vm_area_struct *vma, struct vm_fault *vmf) } rcu_read_lock(); - buffer = rcu_dereference(event->buffer); - if (!buffer) + rb = rcu_dereference(event->rb); + if (!rb) goto unlock; if (vmf->pgoff && (vmf->flags & FAULT_FLAG_WRITE)) goto unlock; - vmf->page = perf_mmap_to_page(buffer, vmf->pgoff); + vmf->page = perf_mmap_to_page(rb, vmf->pgoff); if (!vmf->page) goto unlock; @@ -3645,35 +3452,35 @@ unlock: return ret; } -static void perf_buffer_free_rcu(struct rcu_head *rcu_head) +static void rb_free_rcu(struct rcu_head *rcu_head) { - struct perf_buffer *buffer; + struct ring_buffer *rb; - buffer = container_of(rcu_head, struct perf_buffer, rcu_head); - perf_buffer_free(buffer); + rb = container_of(rcu_head, struct ring_buffer, rcu_head); + rb_free(rb); } -static struct perf_buffer *perf_buffer_get(struct perf_event *event) +static struct ring_buffer *ring_buffer_get(struct perf_event *event) { - struct perf_buffer *buffer; + struct ring_buffer *rb; rcu_read_lock(); - buffer = rcu_dereference(event->buffer); - if (buffer) { - if (!atomic_inc_not_zero(&buffer->refcount)) - buffer = NULL; + rb = rcu_dereference(event->rb); + if (rb) { + if (!atomic_inc_not_zero(&rb->refcount)) + rb = NULL; } rcu_read_unlock(); - return buffer; + return rb; } -static void perf_buffer_put(struct perf_buffer *buffer) +static void ring_buffer_put(struct ring_buffer *rb) { - if (!atomic_dec_and_test(&buffer->refcount)) + if (!atomic_dec_and_test(&rb->refcount)) return; - call_rcu(&buffer->rcu_head, perf_buffer_free_rcu); + call_rcu(&rb->rcu_head, rb_free_rcu); } static void perf_mmap_open(struct vm_area_struct *vma) @@ -3688,16 +3495,16 @@ static void perf_mmap_close(struct vm_area_struct *vma) struct perf_event *event = vma->vm_file->private_data; if (atomic_dec_and_mutex_lock(&event->mmap_count, &event->mmap_mutex)) { - unsigned long size = perf_data_size(event->buffer); + unsigned long size = perf_data_size(event->rb); struct user_struct *user = event->mmap_user; - struct perf_buffer *buffer = event->buffer; + struct ring_buffer *rb = event->rb; atomic_long_sub((size >> PAGE_SHIFT) + 1, &user->locked_vm); vma->vm_mm->locked_vm -= event->mmap_locked; - rcu_assign_pointer(event->buffer, NULL); + rcu_assign_pointer(event->rb, NULL); mutex_unlock(&event->mmap_mutex); - perf_buffer_put(buffer); + ring_buffer_put(rb); free_uid(user); } } @@ -3715,7 +3522,7 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) unsigned long user_locked, user_lock_limit; struct user_struct *user = current_user(); unsigned long locked, lock_limit; - struct perf_buffer *buffer; + struct ring_buffer *rb; unsigned long vma_size; unsigned long nr_pages; long user_extra, extra; @@ -3724,7 +3531,7 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) /* * Don't allow mmap() of inherited per-task counters. This would * create a performance issue due to all children writing to the - * same buffer. + * same rb. */ if (event->cpu == -1 && event->attr.inherit) return -EINVAL; @@ -3736,7 +3543,7 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) nr_pages = (vma_size / PAGE_SIZE) - 1; /* - * If we have buffer pages ensure they're a power-of-two number, so we + * If we have rb pages ensure they're a power-of-two number, so we * can do bitmasks instead of modulo. */ if (nr_pages != 0 && !is_power_of_2(nr_pages)) @@ -3750,9 +3557,9 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) WARN_ON_ONCE(event->ctx->parent_ctx); mutex_lock(&event->mmap_mutex); - if (event->buffer) { - if (event->buffer->nr_pages == nr_pages) - atomic_inc(&event->buffer->refcount); + if (event->rb) { + if (event->rb->nr_pages == nr_pages) + atomic_inc(&event->rb->refcount); else ret = -EINVAL; goto unlock; @@ -3782,18 +3589,20 @@ static int perf_mmap(struct file *file, struct vm_area_struct *vma) goto unlock; } - WARN_ON(event->buffer); + WARN_ON(event->rb); if (vma->vm_flags & VM_WRITE) - flags |= PERF_BUFFER_WRITABLE; + flags |= RING_BUFFER_WRITABLE; + + rb = rb_alloc(nr_pages, + event->attr.watermark ? event->attr.wakeup_watermark : 0, + event->cpu, flags); - buffer = perf_buffer_alloc(nr_pages, event->attr.wakeup_watermark, - event->cpu, flags); - if (!buffer) { + if (!rb) { ret = -ENOMEM; goto unlock; } - rcu_assign_pointer(event->buffer, buffer); + rcu_assign_pointer(event->rb, rb); atomic_long_add(user_extra, &user->locked_vm); event->mmap_locked = extra; @@ -3892,117 +3701,6 @@ int perf_unregister_guest_info_callbacks(struct perf_guest_info_callbacks *cbs) } EXPORT_SYMBOL_GPL(perf_unregister_guest_info_callbacks); -/* - * Output - */ -static bool perf_output_space(struct perf_buffer *buffer, unsigned long tail, - unsigned long offset, unsigned long head) -{ - unsigned long mask; - - if (!buffer->writable) - return true; - - mask = perf_data_size(buffer) - 1; - - offset = (offset - tail) & mask; - head = (head - tail) & mask; - - if ((int)(head - offset) < 0) - return false; - - return true; -} - -static void perf_output_wakeup(struct perf_output_handle *handle) -{ - atomic_set(&handle->buffer->poll, POLL_IN); - - if (handle->nmi) { - handle->event->pending_wakeup = 1; - irq_work_queue(&handle->event->pending); - } else - perf_event_wakeup(handle->event); -} - -/* - * We need to ensure a later event_id doesn't publish a head when a former - * event isn't done writing. However since we need to deal with NMIs we - * cannot fully serialize things. - * - * We only publish the head (and generate a wakeup) when the outer-most - * event completes. - */ -static void perf_output_get_handle(struct perf_output_handle *handle) -{ - struct perf_buffer *buffer = handle->buffer; - - preempt_disable(); - local_inc(&buffer->nest); - handle->wakeup = local_read(&buffer->wakeup); -} - -static void perf_output_put_handle(struct perf_output_handle *handle) -{ - struct perf_buffer *buffer = handle->buffer; - unsigned long head; - -again: - head = local_read(&buffer->head); - - /* - * IRQ/NMI can happen here, which means we can miss a head update. - */ - - if (!local_dec_and_test(&buffer->nest)) - goto out; - - /* - * Publish the known good head. Rely on the full barrier implied - * by atomic_dec_and_test() order the buffer->head read and this - * write. - */ - buffer->user_page->data_head = head; - - /* - * Now check if we missed an update, rely on the (compiler) - * barrier in atomic_dec_and_test() to re-read buffer->head. - */ - if (unlikely(head != local_read(&buffer->head))) { - local_inc(&buffer->nest); - goto again; - } - - if (handle->wakeup != local_read(&buffer->wakeup)) - perf_output_wakeup(handle); - -out: - preempt_enable(); -} - -__always_inline void perf_output_copy(struct perf_output_handle *handle, - const void *buf, unsigned int len) -{ - do { - unsigned long size = min_t(unsigned long, handle->size, len); - - memcpy(handle->addr, buf, size); - - len -= size; - handle->addr += size; - buf += size; - handle->size -= size; - if (!handle->size) { - struct perf_buffer *buffer = handle->buffer; - - handle->page++; - handle->page &= buffer->nr_pages - 1; - handle->addr = buffer->data_pages[handle->page]; - handle->size = PAGE_SIZE << page_order(buffer); - } - } while (len); -} - static void __perf_event_header__init_id(struct perf_event_header *header, struct perf_sample_data *data, struct perf_event *event) @@ -4033,9 +3731,9 @@ static void __perf_event_header__init_id(struct perf_event_header *header, } } -static void perf_event_header__init_id(struct perf_event_header *header, - struct perf_sample_data *data, - struct perf_event *event) +void perf_event_header__init_id(struct perf_event_header *header, + struct perf_sample_data *data, + struct perf_event *event) { if (event->attr.sample_id_all) __perf_event_header__init_id(header, data, event); @@ -4062,121 +3760,14 @@ static void __perf_event__output_id_sample(struct perf_output_handle *handle, perf_output_put(handle, data->cpu_entry); } -static void perf_event__output_id_sample(struct perf_event *event, - struct perf_output_handle *handle, - struct perf_sample_data *sample) +void perf_event__output_id_sample(struct perf_event *event, + struct perf_output_handle *handle, + struct perf_sample_data *sample) { if (event->attr.sample_id_all) __perf_event__output_id_sample(handle, sample); } -int perf_output_begin(struct perf_output_handle *handle, - struct perf_event *event, unsigned int size, - int nmi, int sample) -{ - struct perf_buffer *buffer; - unsigned long tail, offset, head; - int have_lost; - struct perf_sample_data sample_data; - struct { - struct perf_event_header header; - u64 id; - u64 lost; - } lost_event; - - rcu_read_lock(); - /* - * For inherited events we send all the output towards the parent. - */ - if (event->parent) - event = event->parent; - - buffer = rcu_dereference(event->buffer); - if (!buffer) - goto out; - - handle->buffer = buffer; - handle->event = event; - handle->nmi = nmi; - handle->sample = sample; - - if (!buffer->nr_pages) - goto out; - - have_lost = local_read(&buffer->lost); - if (have_lost) { - lost_event.header.size = sizeof(lost_event); - perf_event_header__init_id(&lost_event.header, &sample_data, - event); - size += lost_event.header.size; - } - - perf_output_get_handle(handle); - - do { - /* - * Userspace could choose to issue a mb() before updating the - * tail pointer. So that all reads will be completed before the - * write is issued. - */ - tail = ACCESS_ONCE(buffer->user_page->data_tail); - smp_rmb(); - offset = head = local_read(&buffer->head); - head += size; - if (unlikely(!perf_output_space(buffer, tail, offset, head))) - goto fail; - } while (local_cmpxchg(&buffer->head, offset, head) != offset); - - if (head - local_read(&buffer->wakeup) > buffer->watermark) - local_add(buffer->watermark, &buffer->wakeup); - - handle->page = offset >> (PAGE_SHIFT + page_order(buffer)); - handle->page &= buffer->nr_pages - 1; - handle->size = offset & ((PAGE_SIZE << page_order(buffer)) - 1); - handle->addr = buffer->data_pages[handle->page]; - handle->addr += handle->size; - handle->size = (PAGE_SIZE << page_order(buffer)) - handle->size; - - if (have_lost) { - lost_event.header.type = PERF_RECORD_LOST; - lost_event.header.misc = 0; - lost_event.id = event->id; - lost_event.lost = local_xchg(&buffer->lost, 0); - - perf_output_put(handle, lost_event); - perf_event__output_id_sample(event, handle, &sample_data); - } - - return 0; - -fail: - local_inc(&buffer->lost); - perf_output_put_handle(handle); -out: - rcu_read_unlock(); - - return -ENOSPC; -} - -void perf_output_end(struct perf_output_handle *handle) -{ - struct perf_event *event = handle->event; - struct perf_buffer *buffer = handle->buffer; - - int wakeup_events = event->attr.wakeup_events; - - if (handle->sample && wakeup_events) { - int events = local_inc_return(&buffer->events); - if (events >= wakeup_events) { - local_sub(wakeup_events, &buffer->events); - local_inc(&buffer->wakeup); - } - } - - perf_output_put_handle(handle); - rcu_read_unlock(); -} - static void perf_output_read_one(struct perf_output_handle *handle, struct perf_event *event, u64 enabled, u64 running) @@ -4197,7 +3788,7 @@ static void perf_output_read_one(struct perf_output_handle *handle, if (read_format & PERF_FORMAT_ID) values[n++] = primary_event_id(event); - perf_output_copy(handle, values, n * sizeof(u64)); + __output_copy(handle, values, n * sizeof(u64)); } /* @@ -4227,7 +3818,7 @@ static void perf_output_read_group(struct perf_output_handle *handle, if (read_format & PERF_FORMAT_ID) values[n++] = primary_event_id(leader); - perf_output_copy(handle, values, n * sizeof(u64)); + __output_copy(handle, values, n * sizeof(u64)); list_for_each_entry(sub, &leader->sibling_list, group_entry) { n = 0; @@ -4239,7 +3830,7 @@ static void perf_output_read_group(struct perf_output_handle *handle, if (read_format & PERF_FORMAT_ID) values[n++] = primary_event_id(sub); - perf_output_copy(handle, values, n * sizeof(u64)); + __output_copy(handle, values, n * sizeof(u64)); } } @@ -4249,7 +3840,7 @@ static void perf_output_read_group(struct perf_output_handle *handle, static void perf_output_read(struct perf_output_handle *handle, struct perf_event *event) { - u64 enabled = 0, running = 0, now, ctx_time; + u64 enabled = 0, running = 0; u64 read_format = event->attr.read_format; /* @@ -4261,12 +3852,8 @@ static void perf_output_read(struct perf_output_handle *handle, * because of locking issue as we are called in * NMI context */ - if (read_format & PERF_FORMAT_TOTAL_TIMES) { - now = perf_clock(); - ctx_time = event->shadow_ctx_time + now; - enabled = ctx_time - event->tstamp_enabled; - running = ctx_time - event->tstamp_running; - } + if (read_format & PERF_FORMAT_TOTAL_TIMES) + calc_timer_values(event, &enabled, &running); if (event->attr.read_format & PERF_FORMAT_GROUP) perf_output_read_group(handle, event, enabled, running); @@ -4319,7 +3906,7 @@ void perf_output_sample(struct perf_output_handle *handle, size *= sizeof(u64); - perf_output_copy(handle, data->callchain, size); + __output_copy(handle, data->callchain, size); } else { u64 nr = 0; perf_output_put(handle, nr); @@ -4329,8 +3916,8 @@ void perf_output_sample(struct perf_output_handle *handle, if (sample_type & PERF_SAMPLE_RAW) { if (data->raw) { perf_output_put(handle, data->raw->size); - perf_output_copy(handle, data->raw->data, - data->raw->size); + __output_copy(handle, data->raw->data, + data->raw->size); } else { struct { u32 size; @@ -4342,6 +3929,20 @@ void perf_output_sample(struct perf_output_handle *handle, perf_output_put(handle, raw); } } + + if (!event->attr.watermark) { + int wakeup_events = event->attr.wakeup_events; + + if (wakeup_events) { + struct ring_buffer *rb = handle->rb; + int events = local_inc_return(&rb->events); + + if (events >= wakeup_events) { + local_sub(wakeup_events, &rb->events); + local_inc(&rb->wakeup); + } + } + } } void perf_prepare_sample(struct perf_event_header *header, @@ -4386,7 +3987,7 @@ void perf_prepare_sample(struct perf_event_header *header, } } -static void perf_event_output(struct perf_event *event, int nmi, +static void perf_event_output(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs) { @@ -4398,7 +3999,7 @@ static void perf_event_output(struct perf_event *event, int nmi, perf_prepare_sample(&header, data, event, regs); - if (perf_output_begin(&handle, event, header.size, nmi, 1)) + if (perf_output_begin(&handle, event, header.size)) goto exit; perf_output_sample(&handle, &header, data, event); @@ -4438,7 +4039,7 @@ perf_event_read_event(struct perf_event *event, int ret; perf_event_header__init_id(&read_event.header, &sample, event); - ret = perf_output_begin(&handle, event, read_event.header.size, 0, 0); + ret = perf_output_begin(&handle, event, read_event.header.size); if (ret) return; @@ -4481,7 +4082,7 @@ static void perf_event_task_output(struct perf_event *event, perf_event_header__init_id(&task_event->event_id.header, &sample, event); ret = perf_output_begin(&handle, event, - task_event->event_id.header.size, 0, 0); + task_event->event_id.header.size); if (ret) goto out; @@ -4618,7 +4219,7 @@ static void perf_event_comm_output(struct perf_event *event, perf_event_header__init_id(&comm_event->event_id.header, &sample, event); ret = perf_output_begin(&handle, event, - comm_event->event_id.header.size, 0, 0); + comm_event->event_id.header.size); if (ret) goto out; @@ -4627,7 +4228,7 @@ static void perf_event_comm_output(struct perf_event *event, comm_event->event_id.tid = perf_event_tid(event, comm_event->task); perf_output_put(&handle, comm_event->event_id); - perf_output_copy(&handle, comm_event->comm, + __output_copy(&handle, comm_event->comm, comm_event->comm_size); perf_event__output_id_sample(event, &handle, &sample); @@ -4765,7 +4366,7 @@ static void perf_event_mmap_output(struct perf_event *event, perf_event_header__init_id(&mmap_event->event_id.header, &sample, event); ret = perf_output_begin(&handle, event, - mmap_event->event_id.header.size, 0, 0); + mmap_event->event_id.header.size); if (ret) goto out; @@ -4773,7 +4374,7 @@ static void perf_event_mmap_output(struct perf_event *event, mmap_event->event_id.tid = perf_event_tid(event, current); perf_output_put(&handle, mmap_event->event_id); - perf_output_copy(&handle, mmap_event->file_name, + __output_copy(&handle, mmap_event->file_name, mmap_event->file_size); perf_event__output_id_sample(event, &handle, &sample); @@ -4829,7 +4430,7 @@ static void perf_event_mmap_event(struct perf_mmap_event *mmap_event) if (file) { /* - * d_path works from the end of the buffer backwards, so we + * d_path works from the end of the rb backwards, so we * need to add enough zero bytes after the string to handle * the 64bit alignment we do later. */ @@ -4960,7 +4561,7 @@ static void perf_log_throttle(struct perf_event *event, int enable) perf_event_header__init_id(&throttle_event.header, &sample, event); ret = perf_output_begin(&handle, event, - throttle_event.header.size, 1, 0); + throttle_event.header.size); if (ret) return; @@ -4973,7 +4574,7 @@ static void perf_log_throttle(struct perf_event *event, int enable) * Generic event overflow handling, sampling. */ -static int __perf_event_overflow(struct perf_event *event, int nmi, +static int __perf_event_overflow(struct perf_event *event, int throttle, struct perf_sample_data *data, struct pt_regs *regs) { @@ -5016,34 +4617,28 @@ static int __perf_event_overflow(struct perf_event *event, int nmi, if (events && atomic_dec_and_test(&event->event_limit)) { ret = 1; event->pending_kill = POLL_HUP; - if (nmi) { - event->pending_disable = 1; - irq_work_queue(&event->pending); - } else - perf_event_disable(event); + event->pending_disable = 1; + irq_work_queue(&event->pending); } if (event->overflow_handler) - event->overflow_handler(event, nmi, data, regs); + event->overflow_handler(event, data, regs); else - perf_event_output(event, nmi, data, regs); + perf_event_output(event, data, regs); if (event->fasync && event->pending_kill) { - if (nmi) { - event->pending_wakeup = 1; - irq_work_queue(&event->pending); - } else - perf_event_wakeup(event); + event->pending_wakeup = 1; + irq_work_queue(&event->pending); } return ret; } -int perf_event_overflow(struct perf_event *event, int nmi, +int perf_event_overflow(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs) { - return __perf_event_overflow(event, nmi, 1, data, regs); + return __perf_event_overflow(event, 1, data, regs); } /* @@ -5092,7 +4687,7 @@ again: } static void perf_swevent_overflow(struct perf_event *event, u64 overflow, - int nmi, struct perf_sample_data *data, + struct perf_sample_data *data, struct pt_regs *regs) { struct hw_perf_event *hwc = &event->hw; @@ -5106,7 +4701,7 @@ static void perf_swevent_overflow(struct perf_event *event, u64 overflow, return; for (; overflow; overflow--) { - if (__perf_event_overflow(event, nmi, throttle, + if (__perf_event_overflow(event, throttle, data, regs)) { /* * We inhibit the overflow from happening when @@ -5119,7 +4714,7 @@ static void perf_swevent_overflow(struct perf_event *event, u64 overflow, } static void perf_swevent_event(struct perf_event *event, u64 nr, - int nmi, struct perf_sample_data *data, + struct perf_sample_data *data, struct pt_regs *regs) { struct hw_perf_event *hwc = &event->hw; @@ -5133,12 +4728,12 @@ static void perf_swevent_event(struct perf_event *event, u64 nr, return; if (nr == 1 && hwc->sample_period == 1 && !event->attr.freq) - return perf_swevent_overflow(event, 1, nmi, data, regs); + return perf_swevent_overflow(event, 1, data, regs); if (local64_add_negative(nr, &hwc->period_left)) return; - perf_swevent_overflow(event, 0, nmi, data, regs); + perf_swevent_overflow(event, 0, data, regs); } static int perf_exclude_event(struct perf_event *event, @@ -5226,7 +4821,7 @@ find_swevent_head(struct swevent_htable *swhash, struct perf_event *event) } static void do_perf_sw_event(enum perf_type_id type, u32 event_id, - u64 nr, int nmi, + u64 nr, struct perf_sample_data *data, struct pt_regs *regs) { @@ -5242,7 +4837,7 @@ static void do_perf_sw_event(enum perf_type_id type, u32 event_id, hlist_for_each_entry_rcu(event, node, head, hlist_entry) { if (perf_swevent_match(event, type, event_id, data, regs)) - perf_swevent_event(event, nr, nmi, data, regs); + perf_swevent_event(event, nr, data, regs); } end: rcu_read_unlock(); @@ -5263,8 +4858,7 @@ inline void perf_swevent_put_recursion_context(int rctx) put_recursion_context(swhash->recursion, rctx); } -void __perf_sw_event(u32 event_id, u64 nr, int nmi, - struct pt_regs *regs, u64 addr) +void __perf_sw_event(u32 event_id, u64 nr, struct pt_regs *regs, u64 addr) { struct perf_sample_data data; int rctx; @@ -5276,7 +4870,7 @@ void __perf_sw_event(u32 event_id, u64 nr, int nmi, perf_sample_data_init(&data, addr); - do_perf_sw_event(PERF_TYPE_SOFTWARE, event_id, nr, nmi, &data, regs); + do_perf_sw_event(PERF_TYPE_SOFTWARE, event_id, nr, &data, regs); perf_swevent_put_recursion_context(rctx); preempt_enable_notrace(); @@ -5524,7 +5118,7 @@ void perf_tp_event(u64 addr, u64 count, void *record, int entry_size, hlist_for_each_entry_rcu(event, node, head, hlist_entry) { if (perf_tp_event_match(event, &data, regs)) - perf_swevent_event(event, count, 1, &data, regs); + perf_swevent_event(event, count, &data, regs); } perf_swevent_put_recursion_context(rctx); @@ -5617,7 +5211,7 @@ void perf_bp_event(struct perf_event *bp, void *data) perf_sample_data_init(&sample, bp->attr.bp_addr); if (!bp->hw.state && !perf_exclude_event(bp, regs)) - perf_swevent_event(bp, 1, 1, &sample, regs); + perf_swevent_event(bp, 1, &sample, regs); } #endif @@ -5646,7 +5240,7 @@ static enum hrtimer_restart perf_swevent_hrtimer(struct hrtimer *hrtimer) if (regs && !perf_exclude_event(event, regs)) { if (!(event->attr.exclude_idle && current->pid == 0)) - if (perf_event_overflow(event, 0, &data, regs)) + if (perf_event_overflow(event, &data, regs)) ret = HRTIMER_NORESTART; } @@ -5986,6 +5580,7 @@ free_dev: } static struct lock_class_key cpuctx_mutex; +static struct lock_class_key cpuctx_lock; int perf_pmu_register(struct pmu *pmu, char *name, int type) { @@ -6036,6 +5631,7 @@ skip_type: cpuctx = per_cpu_ptr(pmu->pmu_cpu_context, cpu); __perf_event_init_context(&cpuctx->ctx); lockdep_set_class(&cpuctx->ctx.mutex, &cpuctx_mutex); + lockdep_set_class(&cpuctx->ctx.lock, &cpuctx_lock); cpuctx->ctx.type = cpu_context; cpuctx->ctx.pmu = pmu; cpuctx->jiffies_interval = 1; @@ -6150,7 +5746,8 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, struct task_struct *task, struct perf_event *group_leader, struct perf_event *parent_event, - perf_overflow_handler_t overflow_handler) + perf_overflow_handler_t overflow_handler, + void *context) { struct pmu *pmu; struct perf_event *event; @@ -6208,10 +5805,13 @@ perf_event_alloc(struct perf_event_attr *attr, int cpu, #endif } - if (!overflow_handler && parent_event) + if (!overflow_handler && parent_event) { overflow_handler = parent_event->overflow_handler; + context = parent_event->overflow_handler_context; + } event->overflow_handler = overflow_handler; + event->overflow_handler_context = context; if (attr->disabled) event->state = PERF_EVENT_STATE_OFF; @@ -6326,13 +5926,6 @@ static int perf_copy_attr(struct perf_event_attr __user *uattr, if (ret) return -EFAULT; - /* - * If the type exists, the corresponding creation will verify - * the attr->config. - */ - if (attr->type >= PERF_TYPE_MAX) - return -EINVAL; - if (attr->__reserved_1) return -EINVAL; @@ -6354,7 +5947,7 @@ err_size: static int perf_event_set_output(struct perf_event *event, struct perf_event *output_event) { - struct perf_buffer *buffer = NULL, *old_buffer = NULL; + struct ring_buffer *rb = NULL, *old_rb = NULL; int ret = -EINVAL; if (!output_event) @@ -6371,7 +5964,7 @@ perf_event_set_output(struct perf_event *event, struct perf_event *output_event) goto out; /* - * If its not a per-cpu buffer, it must be the same task. + * If its not a per-cpu rb, it must be the same task. */ if (output_event->cpu == -1 && output_event->ctx != event->ctx) goto out; @@ -6383,20 +5976,20 @@ set: goto unlock; if (output_event) { - /* get the buffer we want to redirect to */ - buffer = perf_buffer_get(output_event); - if (!buffer) + /* get the rb we want to redirect to */ + rb = ring_buffer_get(output_event); + if (!rb) goto unlock; } - old_buffer = event->buffer; - rcu_assign_pointer(event->buffer, buffer); + old_rb = event->rb; + rcu_assign_pointer(event->rb, rb); ret = 0; unlock: mutex_unlock(&event->mmap_mutex); - if (old_buffer) - perf_buffer_put(old_buffer); + if (old_rb) + ring_buffer_put(old_rb); out: return ret; } @@ -6478,7 +6071,8 @@ SYSCALL_DEFINE5(perf_event_open, } } - event = perf_event_alloc(&attr, cpu, task, group_leader, NULL, NULL); + event = perf_event_alloc(&attr, cpu, task, group_leader, NULL, + NULL, NULL); if (IS_ERR(event)) { err = PTR_ERR(event); goto err_task; @@ -6663,7 +6257,8 @@ err_fd: struct perf_event * perf_event_create_kernel_counter(struct perf_event_attr *attr, int cpu, struct task_struct *task, - perf_overflow_handler_t overflow_handler) + perf_overflow_handler_t overflow_handler, + void *context) { struct perf_event_context *ctx; struct perf_event *event; @@ -6673,7 +6268,8 @@ perf_event_create_kernel_counter(struct perf_event_attr *attr, int cpu, * Get the target context (task or percpu): */ - event = perf_event_alloc(attr, cpu, task, NULL, NULL, overflow_handler); + event = perf_event_alloc(attr, cpu, task, NULL, NULL, + overflow_handler, context); if (IS_ERR(event)) { err = PTR_ERR(event); goto err; @@ -6780,7 +6376,6 @@ static void perf_event_exit_task_context(struct task_struct *child, int ctxn) * our context. */ child_ctx = rcu_dereference_raw(child->perf_event_ctxp[ctxn]); - task_ctx_sched_out(child_ctx, EVENT_ALL); /* * Take the context lock here so that if find_get_context is @@ -6788,6 +6383,7 @@ static void perf_event_exit_task_context(struct task_struct *child, int ctxn) * incremented the context's refcount before we do put_ctx below. */ raw_spin_lock(&child_ctx->lock); + task_ctx_sched_out(child_ctx); child->perf_event_ctxp[ctxn] = NULL; /* * If this context is a clone; unclone it so it can't get @@ -6957,7 +6553,7 @@ inherit_event(struct perf_event *parent_event, parent_event->cpu, child, group_leader, parent_event, - NULL); + NULL, NULL); if (IS_ERR(child_event)) return child_event; get_ctx(child_ctx); @@ -6984,6 +6580,8 @@ inherit_event(struct perf_event *parent_event, child_event->ctx = child_ctx; child_event->overflow_handler = parent_event->overflow_handler; + child_event->overflow_handler_context + = parent_event->overflow_handler_context; /* * Precalculate sample_data sizes @@ -7402,26 +7000,12 @@ static int __perf_cgroup_move(void *info) return 0; } -static void perf_cgroup_move(struct task_struct *task) +static void +perf_cgroup_attach_task(struct cgroup *cgrp, struct task_struct *task) { task_function_call(task, __perf_cgroup_move, task); } -static void perf_cgroup_attach(struct cgroup_subsys *ss, struct cgroup *cgrp, - struct cgroup *old_cgrp, struct task_struct *task, - bool threadgroup) -{ - perf_cgroup_move(task); - if (threadgroup) { - struct task_struct *c; - rcu_read_lock(); - list_for_each_entry_rcu(c, &task->thread_group, thread_group) { - perf_cgroup_move(c); - } - rcu_read_unlock(); - } -} - static void perf_cgroup_exit(struct cgroup_subsys *ss, struct cgroup *cgrp, struct cgroup *old_cgrp, struct task_struct *task) { @@ -7433,7 +7017,7 @@ static void perf_cgroup_exit(struct cgroup_subsys *ss, struct cgroup *cgrp, if (!(task->flags & PF_EXITING)) return; - perf_cgroup_move(task); + perf_cgroup_attach_task(cgrp, task); } struct cgroup_subsys perf_subsys = { @@ -7442,6 +7026,6 @@ struct cgroup_subsys perf_subsys = { .create = perf_cgroup_create, .destroy = perf_cgroup_destroy, .exit = perf_cgroup_exit, - .attach = perf_cgroup_attach, + .attach_task = perf_cgroup_attach_task, }; #endif /* CONFIG_CGROUP_PERF */ diff --git a/kernel/events/hw_breakpoint.c b/kernel/events/hw_breakpoint.c index 086adf25a55..b7971d6f38b 100644 --- a/kernel/events/hw_breakpoint.c +++ b/kernel/events/hw_breakpoint.c @@ -431,9 +431,11 @@ int register_perf_hw_breakpoint(struct perf_event *bp) struct perf_event * register_user_hw_breakpoint(struct perf_event_attr *attr, perf_overflow_handler_t triggered, + void *context, struct task_struct *tsk) { - return perf_event_create_kernel_counter(attr, -1, tsk, triggered); + return perf_event_create_kernel_counter(attr, -1, tsk, triggered, + context); } EXPORT_SYMBOL_GPL(register_user_hw_breakpoint); @@ -502,7 +504,8 @@ EXPORT_SYMBOL_GPL(unregister_hw_breakpoint); */ struct perf_event * __percpu * register_wide_hw_breakpoint(struct perf_event_attr *attr, - perf_overflow_handler_t triggered) + perf_overflow_handler_t triggered, + void *context) { struct perf_event * __percpu *cpu_events, **pevent, *bp; long err; @@ -515,7 +518,8 @@ register_wide_hw_breakpoint(struct perf_event_attr *attr, get_online_cpus(); for_each_online_cpu(cpu) { pevent = per_cpu_ptr(cpu_events, cpu); - bp = perf_event_create_kernel_counter(attr, cpu, NULL, triggered); + bp = perf_event_create_kernel_counter(attr, cpu, NULL, + triggered, context); *pevent = bp; diff --git a/kernel/events/internal.h b/kernel/events/internal.h new file mode 100644 index 00000000000..09097dd8116 --- /dev/null +++ b/kernel/events/internal.h @@ -0,0 +1,96 @@ +#ifndef _KERNEL_EVENTS_INTERNAL_H +#define _KERNEL_EVENTS_INTERNAL_H + +#define RING_BUFFER_WRITABLE 0x01 + +struct ring_buffer { + atomic_t refcount; + struct rcu_head rcu_head; +#ifdef CONFIG_PERF_USE_VMALLOC + struct work_struct work; + int page_order; /* allocation order */ +#endif + int nr_pages; /* nr of data pages */ + int writable; /* are we writable */ + + atomic_t poll; /* POLL_ for wakeups */ + + local_t head; /* write position */ + local_t nest; /* nested writers */ + local_t events; /* event limit */ + local_t wakeup; /* wakeup stamp */ + local_t lost; /* nr records lost */ + + long watermark; /* wakeup watermark */ + + struct perf_event_mmap_page *user_page; + void *data_pages[0]; +}; + +extern void rb_free(struct ring_buffer *rb); +extern struct ring_buffer * +rb_alloc(int nr_pages, long watermark, int cpu, int flags); +extern void perf_event_wakeup(struct perf_event *event); + +extern void +perf_event_header__init_id(struct perf_event_header *header, + struct perf_sample_data *data, + struct perf_event *event); +extern void +perf_event__output_id_sample(struct perf_event *event, + struct perf_output_handle *handle, + struct perf_sample_data *sample); + +extern struct page * +perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff); + +#ifdef CONFIG_PERF_USE_VMALLOC +/* + * Back perf_mmap() with vmalloc memory. + * + * Required for architectures that have d-cache aliasing issues. + */ + +static inline int page_order(struct ring_buffer *rb) +{ + return rb->page_order; +} + +#else + +static inline int page_order(struct ring_buffer *rb) +{ + return 0; +} +#endif + +static unsigned long perf_data_size(struct ring_buffer *rb) +{ + return rb->nr_pages << (PAGE_SHIFT + page_order(rb)); +} + +static inline void +__output_copy(struct perf_output_handle *handle, + const void *buf, unsigned int len) +{ + do { + unsigned long size = min_t(unsigned long, handle->size, len); + + memcpy(handle->addr, buf, size); + + len -= size; + handle->addr += size; + buf += size; + handle->size -= size; + if (!handle->size) { + struct ring_buffer *rb = handle->rb; + + handle->page++; + handle->page &= rb->nr_pages - 1; + handle->addr = rb->data_pages[handle->page]; + handle->size = PAGE_SIZE << page_order(rb); + } + } while (len); +} + +#endif /* _KERNEL_EVENTS_INTERNAL_H */ diff --git a/kernel/events/ring_buffer.c b/kernel/events/ring_buffer.c new file mode 100644 index 00000000000..a2a29205cc0 --- /dev/null +++ b/kernel/events/ring_buffer.c @@ -0,0 +1,380 @@ +/* + * Performance events ring-buffer code: + * + * Copyright (C) 2008 Thomas Gleixner <tglx@linutronix.de> + * Copyright (C) 2008-2011 Red Hat, Inc., Ingo Molnar + * Copyright (C) 2008-2011 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> + * Copyright © 2009 Paul Mackerras, IBM Corp. <paulus@au1.ibm.com> + * + * For licensing details see kernel-base/COPYING + */ + +#include <linux/perf_event.h> +#include <linux/vmalloc.h> +#include <linux/slab.h> + +#include "internal.h" + +static bool perf_output_space(struct ring_buffer *rb, unsigned long tail, + unsigned long offset, unsigned long head) +{ + unsigned long mask; + + if (!rb->writable) + return true; + + mask = perf_data_size(rb) - 1; + + offset = (offset - tail) & mask; + head = (head - tail) & mask; + + if ((int)(head - offset) < 0) + return false; + + return true; +} + +static void perf_output_wakeup(struct perf_output_handle *handle) +{ + atomic_set(&handle->rb->poll, POLL_IN); + + handle->event->pending_wakeup = 1; + irq_work_queue(&handle->event->pending); +} + +/* + * We need to ensure a later event_id doesn't publish a head when a former + * event isn't done writing. However since we need to deal with NMIs we + * cannot fully serialize things. + * + * We only publish the head (and generate a wakeup) when the outer-most + * event completes. + */ +static void perf_output_get_handle(struct perf_output_handle *handle) +{ + struct ring_buffer *rb = handle->rb; + + preempt_disable(); + local_inc(&rb->nest); + handle->wakeup = local_read(&rb->wakeup); +} + +static void perf_output_put_handle(struct perf_output_handle *handle) +{ + struct ring_buffer *rb = handle->rb; + unsigned long head; + +again: + head = local_read(&rb->head); + + /* + * IRQ/NMI can happen here, which means we can miss a head update. + */ + + if (!local_dec_and_test(&rb->nest)) + goto out; + + /* + * Publish the known good head. Rely on the full barrier implied + * by atomic_dec_and_test() order the rb->head read and this + * write. + */ + rb->user_page->data_head = head; + + /* + * Now check if we missed an update, rely on the (compiler) + * barrier in atomic_dec_and_test() to re-read rb->head. + */ + if (unlikely(head != local_read(&rb->head))) { + local_inc(&rb->nest); + goto again; + } + + if (handle->wakeup != local_read(&rb->wakeup)) + perf_output_wakeup(handle); + +out: + preempt_enable(); +} + +int perf_output_begin(struct perf_output_handle *handle, + struct perf_event *event, unsigned int size) +{ + struct ring_buffer *rb; + unsigned long tail, offset, head; + int have_lost; + struct perf_sample_data sample_data; + struct { + struct perf_event_header header; + u64 id; + u64 lost; + } lost_event; + + rcu_read_lock(); + /* + * For inherited events we send all the output towards the parent. + */ + if (event->parent) + event = event->parent; + + rb = rcu_dereference(event->rb); + if (!rb) + goto out; + + handle->rb = rb; + handle->event = event; + + if (!rb->nr_pages) + goto out; + + have_lost = local_read(&rb->lost); + if (have_lost) { + lost_event.header.size = sizeof(lost_event); + perf_event_header__init_id(&lost_event.header, &sample_data, + event); + size += lost_event.header.size; + } + + perf_output_get_handle(handle); + + do { + /* + * Userspace could choose to issue a mb() before updating the + * tail pointer. So that all reads will be completed before the + * write is issued. + */ + tail = ACCESS_ONCE(rb->user_page->data_tail); + smp_rmb(); + offset = head = local_read(&rb->head); + head += size; + if (unlikely(!perf_output_space(rb, tail, offset, head))) + goto fail; + } while (local_cmpxchg(&rb->head, offset, head) != offset); + + if (head - local_read(&rb->wakeup) > rb->watermark) + local_add(rb->watermark, &rb->wakeup); + + handle->page = offset >> (PAGE_SHIFT + page_order(rb)); + handle->page &= rb->nr_pages - 1; + handle->size = offset & ((PAGE_SIZE << page_order(rb)) - 1); + handle->addr = rb->data_pages[handle->page]; + handle->addr += handle->size; + handle->size = (PAGE_SIZE << page_order(rb)) - handle->size; + + if (have_lost) { + lost_event.header.type = PERF_RECORD_LOST; + lost_event.header.misc = 0; + lost_event.id = event->id; + lost_event.lost = local_xchg(&rb->lost, 0); + + perf_output_put(handle, lost_event); + perf_event__output_id_sample(event, handle, &sample_data); + } + + return 0; + +fail: + local_inc(&rb->lost); + perf_output_put_handle(handle); +out: + rcu_read_unlock(); + + return -ENOSPC; +} + +void perf_output_copy(struct perf_output_handle *handle, + const void *buf, unsigned int len) +{ + __output_copy(handle, buf, len); +} + +void perf_output_end(struct perf_output_handle *handle) +{ + perf_output_put_handle(handle); + rcu_read_unlock(); +} + +static void +ring_buffer_init(struct ring_buffer *rb, long watermark, int flags) +{ + long max_size = perf_data_size(rb); + + if (watermark) + rb->watermark = min(max_size, watermark); + + if (!rb->watermark) + rb->watermark = max_size / 2; + + if (flags & RING_BUFFER_WRITABLE) + rb->writable = 1; + + atomic_set(&rb->refcount, 1); +} + +#ifndef CONFIG_PERF_USE_VMALLOC + +/* + * Back perf_mmap() with regular GFP_KERNEL-0 pages. + */ + +struct page * +perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff) +{ + if (pgoff > rb->nr_pages) + return NULL; + + if (pgoff == 0) + return virt_to_page(rb->user_page); + + return virt_to_page(rb->data_pages[pgoff - 1]); +} + +static void *perf_mmap_alloc_page(int cpu) +{ + struct page *page; + int node; + + node = (cpu == -1) ? cpu : cpu_to_node(cpu); + page = alloc_pages_node(node, GFP_KERNEL | __GFP_ZERO, 0); + if (!page) + return NULL; + + return page_address(page); +} + +struct ring_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags) +{ + struct ring_buffer *rb; + unsigned long size; + int i; + + size = sizeof(struct ring_buffer); + size += nr_pages * sizeof(void *); + + rb = kzalloc(size, GFP_KERNEL); + if (!rb) + goto fail; + + rb->user_page = perf_mmap_alloc_page(cpu); + if (!rb->user_page) + goto fail_user_page; + + for (i = 0; i < nr_pages; i++) { + rb->data_pages[i] = perf_mmap_alloc_page(cpu); + if (!rb->data_pages[i]) + goto fail_data_pages; + } + + rb->nr_pages = nr_pages; + + ring_buffer_init(rb, watermark, flags); + + return rb; + +fail_data_pages: + for (i--; i >= 0; i--) + free_page((unsigned long)rb->data_pages[i]); + + free_page((unsigned long)rb->user_page); + +fail_user_page: + kfree(rb); + +fail: + return NULL; +} + +static void perf_mmap_free_page(unsigned long addr) +{ + struct page *page = virt_to_page((void *)addr); + + page->mapping = NULL; + __free_page(page); +} + +void rb_free(struct ring_buffer *rb) +{ + int i; + + perf_mmap_free_page((unsigned long)rb->user_page); + for (i = 0; i < rb->nr_pages; i++) + perf_mmap_free_page((unsigned long)rb->data_pages[i]); + kfree(rb); +} + +#else + +struct page * +perf_mmap_to_page(struct ring_buffer *rb, unsigned long pgoff) +{ + if (pgoff > (1UL << page_order(rb))) + return NULL; + + return vmalloc_to_page((void *)rb->user_page + pgoff * PAGE_SIZE); +} + +static void perf_mmap_unmark_page(void *addr) +{ + struct page *page = vmalloc_to_page(addr); + + page->mapping = NULL; +} + +static void rb_free_work(struct work_struct *work) +{ + struct ring_buffer *rb; + void *base; + int i, nr; + + rb = container_of(work, struct ring_buffer, work); + nr = 1 << page_order(rb); + + base = rb->user_page; + for (i = 0; i < nr + 1; i++) + perf_mmap_unmark_page(base + (i * PAGE_SIZE)); + + vfree(base); + kfree(rb); +} + +void rb_free(struct ring_buffer *rb) +{ + schedule_work(&rb->work); +} + +struct ring_buffer *rb_alloc(int nr_pages, long watermark, int cpu, int flags) +{ + struct ring_buffer *rb; + unsigned long size; + void *all_buf; + + size = sizeof(struct ring_buffer); + size += sizeof(void *); + + rb = kzalloc(size, GFP_KERNEL); + if (!rb) + goto fail; + + INIT_WORK(&rb->work, rb_free_work); + + all_buf = vmalloc_user((nr_pages + 1) * PAGE_SIZE); + if (!all_buf) + goto fail_all_buf; + + rb->user_page = all_buf; + rb->data_pages[0] = all_buf + PAGE_SIZE; + rb->page_order = ilog2(nr_pages); + rb->nr_pages = 1; + + ring_buffer_init(rb, watermark, flags); + + return rb; + +fail_all_buf: + kfree(rb); + +fail: + return NULL; +} + +#endif diff --git a/kernel/exit.c b/kernel/exit.c index 20a40647152..9ee58bb9e60 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -85,7 +85,6 @@ static void __exit_signal(struct task_struct *tsk) struct tty_struct *uninitialized_var(tty); sighand = rcu_dereference_check(tsk->sighand, - rcu_read_lock_held() || lockdep_tasklist_lock_is_held()); spin_lock(&sighand->siglock); @@ -169,7 +168,6 @@ void release_task(struct task_struct * p) struct task_struct *leader; int zap_leader; repeat: - tracehook_prepare_release_task(p); /* don't need to get the RCU readlock here - the process is dead and * can't be modifying its own credentials. But shut RCU-lockdep up */ rcu_read_lock(); @@ -179,7 +177,7 @@ repeat: proc_flush_task(p); write_lock_irq(&tasklist_lock); - tracehook_finish_release_task(p); + ptrace_release_task(p); __exit_signal(p); /* @@ -190,22 +188,12 @@ repeat: zap_leader = 0; leader = p->group_leader; if (leader != p && thread_group_empty(leader) && leader->exit_state == EXIT_ZOMBIE) { - BUG_ON(task_detached(leader)); - do_notify_parent(leader, leader->exit_signal); /* * If we were the last child thread and the leader has * exited already, and the leader's parent ignores SIGCHLD, * then we are the one who should release the leader. - * - * do_notify_parent() will have marked it self-reaping in - * that case. - */ - zap_leader = task_detached(leader); - - /* - * This maintains the invariant that release_task() - * only runs on a task in EXIT_DEAD, just for sanity. */ + zap_leader = do_notify_parent(leader, leader->exit_signal); if (zap_leader) leader->exit_state = EXIT_DEAD; } @@ -277,18 +265,16 @@ int is_current_pgrp_orphaned(void) return retval; } -static int has_stopped_jobs(struct pid *pgrp) +static bool has_stopped_jobs(struct pid *pgrp) { - int retval = 0; struct task_struct *p; do_each_pid_task(pgrp, PIDTYPE_PGID, p) { - if (!task_is_stopped(p)) - continue; - retval = 1; - break; + if (p->signal->flags & SIGNAL_STOP_STOPPED) + return true; } while_each_pid_task(pgrp, PIDTYPE_PGID, p); - return retval; + + return false; } /* @@ -561,29 +547,28 @@ void exit_files(struct task_struct *tsk) #ifdef CONFIG_MM_OWNER /* - * Task p is exiting and it owned mm, lets find a new owner for it + * A task is exiting. If it owned this mm, find a new owner for the mm. */ -static inline int -mm_need_new_owner(struct mm_struct *mm, struct task_struct *p) -{ - /* - * If there are other users of the mm and the owner (us) is exiting - * we need to find a new owner to take on the responsibility. - */ - if (atomic_read(&mm->mm_users) <= 1) - return 0; - if (mm->owner != p) - return 0; - return 1; -} - void mm_update_next_owner(struct mm_struct *mm) { struct task_struct *c, *g, *p = current; retry: - if (!mm_need_new_owner(mm, p)) + /* + * If the exiting or execing task is not the owner, it's + * someone else's problem. + */ + if (mm->owner != p) return; + /* + * The current owner is exiting/execing and there are no other + * candidates. Do not leave the mm pointing to a possibly + * freed task structure. + */ + if (atomic_read(&mm->mm_users) <= 1) { + mm->owner = NULL; + return; + } read_lock(&tasklist_lock); /* @@ -752,7 +737,7 @@ static void reparent_leader(struct task_struct *father, struct task_struct *p, { list_move_tail(&p->sibling, &p->real_parent->children); - if (task_detached(p)) + if (p->exit_state == EXIT_DEAD) return; /* * If this is a threaded reparent there is no need to @@ -765,10 +750,9 @@ static void reparent_leader(struct task_struct *father, struct task_struct *p, p->exit_signal = SIGCHLD; /* If it has exited notify the new parent about this child's death. */ - if (!task_ptrace(p) && + if (!p->ptrace && p->exit_state == EXIT_ZOMBIE && thread_group_empty(p)) { - do_notify_parent(p, p->exit_signal); - if (task_detached(p)) { + if (do_notify_parent(p, p->exit_signal)) { p->exit_state = EXIT_DEAD; list_move_tail(&p->sibling, dead); } @@ -795,7 +779,7 @@ static void forget_original_parent(struct task_struct *father) do { t->real_parent = reaper; if (t->parent == father) { - BUG_ON(task_ptrace(t)); + BUG_ON(t->ptrace); t->parent = t->real_parent; } if (t->pdeath_signal) @@ -820,8 +804,7 @@ static void forget_original_parent(struct task_struct *father) */ static void exit_notify(struct task_struct *tsk, int group_dead) { - int signal; - void *cookie; + bool autoreap; /* * This does two things: @@ -852,26 +835,33 @@ static void exit_notify(struct task_struct *tsk, int group_dead) * we have changed execution domain as these two values started * the same after a fork. */ - if (tsk->exit_signal != SIGCHLD && !task_detached(tsk) && + if (thread_group_leader(tsk) && tsk->exit_signal != SIGCHLD && (tsk->parent_exec_id != tsk->real_parent->self_exec_id || tsk->self_exec_id != tsk->parent_exec_id)) tsk->exit_signal = SIGCHLD; - signal = tracehook_notify_death(tsk, &cookie, group_dead); - if (signal >= 0) - signal = do_notify_parent(tsk, signal); + if (unlikely(tsk->ptrace)) { + int sig = thread_group_leader(tsk) && + thread_group_empty(tsk) && + !ptrace_reparented(tsk) ? + tsk->exit_signal : SIGCHLD; + autoreap = do_notify_parent(tsk, sig); + } else if (thread_group_leader(tsk)) { + autoreap = thread_group_empty(tsk) && + do_notify_parent(tsk, tsk->exit_signal); + } else { + autoreap = true; + } - tsk->exit_state = signal == DEATH_REAP ? EXIT_DEAD : EXIT_ZOMBIE; + tsk->exit_state = autoreap ? EXIT_DEAD : EXIT_ZOMBIE; /* mt-exec, de_thread() is waiting for group leader */ if (unlikely(tsk->signal->notify_count < 0)) wake_up_process(tsk->signal->group_exit_task); write_unlock_irq(&tasklist_lock); - tracehook_report_death(tsk, signal, cookie, group_dead); - /* If the process is dead, release it - nobody will wait for it */ - if (signal == DEATH_REAP) + if (autoreap) release_task(tsk); } @@ -907,7 +897,6 @@ NORET_TYPE void do_exit(long code) profile_task_exit(tsk); - WARN_ON(atomic_read(&tsk->fs_excl)); WARN_ON(blk_needs_flush_plug(tsk)); if (unlikely(in_interrupt())) @@ -924,7 +913,7 @@ NORET_TYPE void do_exit(long code) */ set_fs(USER_DS); - tracehook_report_exit(&code); + ptrace_event(PTRACE_EVENT_EXIT, code); validate_creds_for_do_exit(tsk); @@ -1236,9 +1225,9 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p) traced = ptrace_reparented(p); /* * It can be ptraced but not reparented, check - * !task_detached() to filter out sub-threads. + * thread_group_leader() to filter out sub-threads. */ - if (likely(!traced) && likely(!task_detached(p))) { + if (likely(!traced) && thread_group_leader(p)) { struct signal_struct *psig; struct signal_struct *sig; unsigned long maxrss; @@ -1346,16 +1335,13 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p) /* We dropped tasklist, ptracer could die and untrace */ ptrace_unlink(p); /* - * If this is not a detached task, notify the parent. - * If it's still not detached after that, don't release - * it now. + * If this is not a sub-thread, notify the parent. + * If parent wants a zombie, don't release it now. */ - if (!task_detached(p)) { - do_notify_parent(p, p->exit_signal); - if (!task_detached(p)) { - p->exit_state = EXIT_ZOMBIE; - p = NULL; - } + if (thread_group_leader(p) && + !do_notify_parent(p, p->exit_signal)) { + p->exit_state = EXIT_ZOMBIE; + p = NULL; } write_unlock_irq(&tasklist_lock); } @@ -1368,7 +1354,8 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p) static int *task_stopped_code(struct task_struct *p, bool ptrace) { if (ptrace) { - if (task_is_stopped_or_traced(p)) + if (task_is_stopped_or_traced(p) && + !(p->jobctl & JOBCTL_LISTENING)) return &p->exit_code; } else { if (p->signal->flags & SIGNAL_STOP_STOPPED) @@ -1564,7 +1551,7 @@ static int wait_consider_task(struct wait_opts *wo, int ptrace, * Notification and reaping will be cascaded to the real * parent when the ptracer detaches. */ - if (likely(!ptrace) && unlikely(task_ptrace(p))) { + if (likely(!ptrace) && unlikely(p->ptrace)) { /* it will become visible, clear notask_error */ wo->notask_error = 0; return 0; @@ -1607,8 +1594,7 @@ static int wait_consider_task(struct wait_opts *wo, int ptrace, * own children, it should create a separate process which * takes the role of real parent. */ - if (likely(!ptrace) && task_ptrace(p) && - same_thread_group(p->parent, p->real_parent)) + if (likely(!ptrace) && p->ptrace && !ptrace_reparented(p)) return 0; /* diff --git a/kernel/fork.c b/kernel/fork.c index 0276c30401a..17bf7c8d651 100644 --- a/kernel/fork.c +++ b/kernel/fork.c @@ -37,7 +37,6 @@ #include <linux/swap.h> #include <linux/syscalls.h> #include <linux/jiffies.h> -#include <linux/tracehook.h> #include <linux/futex.h> #include <linux/compat.h> #include <linux/kthread.h> @@ -291,7 +290,6 @@ static struct task_struct *dup_task_struct(struct task_struct *orig) /* One for us, one for whoever does the "release_task()" (usually parent) */ atomic_set(&tsk->usage,2); - atomic_set(&tsk->fs_excl, 0); #ifdef CONFIG_BLK_DEV_IO_TRACE tsk->btrace_seq = 0; #endif @@ -1013,7 +1011,7 @@ static void rt_mutex_init_task(struct task_struct *p) { raw_spin_lock_init(&p->pi_lock); #ifdef CONFIG_RT_MUTEXES - plist_head_init_raw(&p->pi_waiters, &p->pi_lock); + plist_head_init(&p->pi_waiters); p->pi_blocked_on = NULL; #endif } @@ -1340,7 +1338,7 @@ static struct task_struct *copy_process(unsigned long clone_flags, } if (likely(p->pid)) { - tracehook_finish_clone(p, clone_flags, trace); + ptrace_init_task(p, (clone_flags & CLONE_PTRACE) || trace); if (thread_group_leader(p)) { if (is_child_reaper(pid)) @@ -1481,10 +1479,22 @@ long do_fork(unsigned long clone_flags, } /* - * When called from kernel_thread, don't do user tracing stuff. + * Determine whether and which event to report to ptracer. When + * called from kernel_thread or CLONE_UNTRACED is explicitly + * requested, no event is reported; otherwise, report if the event + * for the type of forking is enabled. */ - if (likely(user_mode(regs))) - trace = tracehook_prepare_clone(clone_flags); + if (likely(user_mode(regs)) && !(clone_flags & CLONE_UNTRACED)) { + if (clone_flags & CLONE_VFORK) + trace = PTRACE_EVENT_VFORK; + else if ((clone_flags & CSIGNAL) != SIGCHLD) + trace = PTRACE_EVENT_CLONE; + else + trace = PTRACE_EVENT_FORK; + + if (likely(!ptrace_event_enabled(current, trace))) + trace = 0; + } p = copy_process(clone_flags, stack_start, regs, stack_size, child_tidptr, NULL, trace); @@ -1508,26 +1518,26 @@ long do_fork(unsigned long clone_flags, } audit_finish_fork(p); - tracehook_report_clone(regs, clone_flags, nr, p); /* * We set PF_STARTING at creation in case tracing wants to * use this to distinguish a fully live task from one that - * hasn't gotten to tracehook_report_clone() yet. Now we - * clear it and set the child going. + * hasn't finished SIGSTOP raising yet. Now we clear it + * and set the child going. */ p->flags &= ~PF_STARTING; wake_up_new_task(p); - tracehook_report_clone_complete(trace, regs, - clone_flags, nr, p); + /* forking complete and child started to run, tell ptracer */ + if (unlikely(trace)) + ptrace_event(trace, nr); if (clone_flags & CLONE_VFORK) { freezer_do_not_count(); wait_for_completion(&vfork); freezer_count(); - tracehook_report_vfork_done(p, nr); + ptrace_event(PTRACE_EVENT_VFORK_DONE, nr); } } else { nr = PTR_ERR(p); @@ -1574,6 +1584,7 @@ void __init proc_caches_init(void) SLAB_HWCACHE_ALIGN|SLAB_PANIC|SLAB_NOTRACK, NULL); vm_area_cachep = KMEM_CACHE(vm_area_struct, SLAB_PANIC); mmap_init(); + nsproxy_cache_init(); } /* diff --git a/kernel/futex.c b/kernel/futex.c index fe28dc282ea..3fbc76cbb9a 100644 --- a/kernel/futex.c +++ b/kernel/futex.c @@ -2697,7 +2697,7 @@ static int __init futex_init(void) futex_cmpxchg_enabled = 1; for (i = 0; i < ARRAY_SIZE(futex_queues); i++) { - plist_head_init(&futex_queues[i].chain, &futex_queues[i].lock); + plist_head_init(&futex_queues[i].chain); spin_lock_init(&futex_queues[i].lock); } diff --git a/kernel/gcov/Kconfig b/kernel/gcov/Kconfig index b8cadf70b1f..5bf924d80b5 100644 --- a/kernel/gcov/Kconfig +++ b/kernel/gcov/Kconfig @@ -2,7 +2,8 @@ menu "GCOV-based kernel profiling" config GCOV_KERNEL bool "Enable gcov-based kernel profiling" - depends on DEBUG_FS && CONSTRUCTORS + depends on DEBUG_FS + select CONSTRUCTORS default n ---help--- This option enables gcov-based code profiling (e.g. for code coverage diff --git a/kernel/irq/generic-chip.c b/kernel/irq/generic-chip.c index 31a9db71190..3a2cab407b9 100644 --- a/kernel/irq/generic-chip.c +++ b/kernel/irq/generic-chip.c @@ -101,10 +101,10 @@ void irq_gc_unmask_enable_reg(struct irq_data *d) } /** - * irq_gc_ack - Ack pending interrupt + * irq_gc_ack_set_bit - Ack pending interrupt via setting bit * @d: irq_data */ -void irq_gc_ack(struct irq_data *d) +void irq_gc_ack_set_bit(struct irq_data *d) { struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d); u32 mask = 1 << (d->irq - gc->irq_base); @@ -115,6 +115,20 @@ void irq_gc_ack(struct irq_data *d) } /** + * irq_gc_ack_clr_bit - Ack pending interrupt via clearing bit + * @d: irq_data + */ +void irq_gc_ack_clr_bit(struct irq_data *d) +{ + struct irq_chip_generic *gc = irq_data_get_irq_chip_data(d); + u32 mask = ~(1 << (d->irq - gc->irq_base)); + + irq_gc_lock(gc); + irq_reg_writel(mask, gc->reg_base + cur_regs(d)->ack); + irq_gc_unlock(gc); +} + +/** * irq_gc_mask_disable_reg_and_ack- Mask and ack pending interrupt * @d: irq_data */ diff --git a/kernel/irq/handle.c b/kernel/irq/handle.c index 90cb55f6d7e..470d08c82bb 100644 --- a/kernel/irq/handle.c +++ b/kernel/irq/handle.c @@ -133,12 +133,6 @@ handle_irq_event_percpu(struct irq_desc *desc, struct irqaction *action) switch (res) { case IRQ_WAKE_THREAD: /* - * Set result to handled so the spurious check - * does not trigger. - */ - res = IRQ_HANDLED; - - /* * Catch drivers which return WAKE_THREAD but * did not set up a thread function */ diff --git a/kernel/irq/irqdesc.c b/kernel/irq/irqdesc.c index 886e80347b3..4c60a50e66b 100644 --- a/kernel/irq/irqdesc.c +++ b/kernel/irq/irqdesc.c @@ -257,13 +257,11 @@ int __init early_irq_init(void) count = ARRAY_SIZE(irq_desc); for (i = 0; i < count; i++) { - desc[i].irq_data.irq = i; - desc[i].irq_data.chip = &no_irq_chip; desc[i].kstat_irqs = alloc_percpu(unsigned int); - irq_settings_clr_and_set(desc, ~0, _IRQ_DEFAULT_INIT_FLAGS); - alloc_masks(desc + i, GFP_KERNEL, node); - desc_smp_init(desc + i, node); + alloc_masks(&desc[i], GFP_KERNEL, node); + raw_spin_lock_init(&desc[i].lock); lockdep_set_class(&desc[i].lock, &irq_desc_lock_class); + desc_set_defaults(i, &desc[i], node); } return arch_early_irq_init(); } @@ -346,6 +344,12 @@ irq_alloc_descs(int irq, unsigned int from, unsigned int cnt, int node) if (!cnt) return -EINVAL; + if (irq >= 0) { + if (from > irq) + return -EINVAL; + from = irq; + } + mutex_lock(&sparse_irq_lock); start = bitmap_find_next_zero_area(allocated_irqs, IRQ_BITMAP_BITS, diff --git a/kernel/irq/manage.c b/kernel/irq/manage.c index f7ce0021e1c..0a7840aeb0f 100644 --- a/kernel/irq/manage.c +++ b/kernel/irq/manage.c @@ -491,6 +491,9 @@ int irq_set_irq_wake(unsigned int irq, unsigned int on) struct irq_desc *desc = irq_get_desc_buslock(irq, &flags); int ret = 0; + if (!desc) + return -EINVAL; + /* wakeup-capable irqs can be shared between drivers that * don't need to have the same sleep mode behaviors. */ @@ -723,13 +726,16 @@ irq_thread_check_affinity(struct irq_desc *desc, struct irqaction *action) { } * context. So we need to disable bh here to avoid deadlocks and other * side effects. */ -static void +static irqreturn_t irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action) { + irqreturn_t ret; + local_bh_disable(); - action->thread_fn(action->irq, action->dev_id); + ret = action->thread_fn(action->irq, action->dev_id); irq_finalize_oneshot(desc, action, false); local_bh_enable(); + return ret; } /* @@ -737,10 +743,14 @@ irq_forced_thread_fn(struct irq_desc *desc, struct irqaction *action) * preemtible - many of them need to sleep and wait for slow busses to * complete. */ -static void irq_thread_fn(struct irq_desc *desc, struct irqaction *action) +static irqreturn_t irq_thread_fn(struct irq_desc *desc, + struct irqaction *action) { - action->thread_fn(action->irq, action->dev_id); + irqreturn_t ret; + + ret = action->thread_fn(action->irq, action->dev_id); irq_finalize_oneshot(desc, action, false); + return ret; } /* @@ -753,7 +763,8 @@ static int irq_thread(void *data) }; struct irqaction *action = data; struct irq_desc *desc = irq_to_desc(action->irq); - void (*handler_fn)(struct irq_desc *desc, struct irqaction *action); + irqreturn_t (*handler_fn)(struct irq_desc *desc, + struct irqaction *action); int wake; if (force_irqthreads & test_bit(IRQTF_FORCED_THREAD, @@ -783,8 +794,12 @@ static int irq_thread(void *data) desc->istate |= IRQS_PENDING; raw_spin_unlock_irq(&desc->lock); } else { + irqreturn_t action_ret; + raw_spin_unlock_irq(&desc->lock); - handler_fn(desc, action); + action_ret = handler_fn(desc, action); + if (!noirqdebug) + note_interrupt(action->irq, desc, action_ret); } wake = atomic_dec_and_test(&desc->threads_active); diff --git a/kernel/irq/spurious.c b/kernel/irq/spurious.c index dfbd550401b..aa57d5da18c 100644 --- a/kernel/irq/spurious.c +++ b/kernel/irq/spurious.c @@ -167,6 +167,13 @@ out: jiffies + POLL_SPURIOUS_IRQ_INTERVAL); } +static inline int bad_action_ret(irqreturn_t action_ret) +{ + if (likely(action_ret <= (IRQ_HANDLED | IRQ_WAKE_THREAD))) + return 0; + return 1; +} + /* * If 99,900 of the previous 100,000 interrupts have not been handled * then assume that the IRQ is stuck in some manner. Drop a diagnostic @@ -182,7 +189,7 @@ __report_bad_irq(unsigned int irq, struct irq_desc *desc, struct irqaction *action; unsigned long flags; - if (action_ret != IRQ_HANDLED && action_ret != IRQ_NONE) { + if (bad_action_ret(action_ret)) { printk(KERN_ERR "irq event %d: bogus return value %x\n", irq, action_ret); } else { @@ -201,10 +208,11 @@ __report_bad_irq(unsigned int irq, struct irq_desc *desc, raw_spin_lock_irqsave(&desc->lock, flags); action = desc->action; while (action) { - printk(KERN_ERR "[<%p>]", action->handler); - print_symbol(" (%s)", - (unsigned long)action->handler); - printk("\n"); + printk(KERN_ERR "[<%p>] %pf", action->handler, action->handler); + if (action->thread_fn) + printk(KERN_CONT " threaded [<%p>] %pf", + action->thread_fn, action->thread_fn); + printk(KERN_CONT "\n"); action = action->next; } raw_spin_unlock_irqrestore(&desc->lock, flags); @@ -262,7 +270,16 @@ void note_interrupt(unsigned int irq, struct irq_desc *desc, if (desc->istate & IRQS_POLL_INPROGRESS) return; - if (unlikely(action_ret != IRQ_HANDLED)) { + /* we get here again via the threaded handler */ + if (action_ret == IRQ_WAKE_THREAD) + return; + + if (bad_action_ret(action_ret)) { + report_bad_irq(irq, desc, action_ret); + return; + } + + if (unlikely(action_ret == IRQ_NONE)) { /* * If we are seeing only the odd spurious IRQ caused by * bus asynchronicity then don't eventually trigger an error, @@ -274,8 +291,6 @@ void note_interrupt(unsigned int irq, struct irq_desc *desc, else desc->irqs_unhandled++; desc->last_unhandled = jiffies; - if (unlikely(action_ret != IRQ_NONE)) - report_bad_irq(irq, desc, action_ret); } if (unlikely(try_misrouted_irq(irq, desc, action_ret))) { diff --git a/kernel/jump_label.c b/kernel/jump_label.c index fa27e750dbc..a8ce45097f3 100644 --- a/kernel/jump_label.c +++ b/kernel/jump_label.c @@ -375,15 +375,19 @@ int jump_label_text_reserved(void *start, void *end) static void jump_label_update(struct jump_label_key *key, int enable) { - struct jump_entry *entry = key->entries; - - /* if there are no users, entry can be NULL */ - if (entry) - __jump_label_update(key, entry, __stop___jump_table, enable); + struct jump_entry *entry = key->entries, *stop = __stop___jump_table; #ifdef CONFIG_MODULES + struct module *mod = __module_address((jump_label_t)key); + __jump_label_mod_update(key, enable); + + if (mod) + stop = mod->jump_entries + mod->num_jump_entries; #endif + /* if there are no users, entry can be NULL */ + if (entry) + __jump_label_update(key, entry, stop, enable); } #endif diff --git a/kernel/kexec.c b/kernel/kexec.c index 8d814cbc810..296fbc84d65 100644 --- a/kernel/kexec.c +++ b/kernel/kexec.c @@ -1095,7 +1095,7 @@ size_t crash_get_memory_size(void) size_t size = 0; mutex_lock(&kexec_mutex); if (crashk_res.end != crashk_res.start) - size = crashk_res.end - crashk_res.start + 1; + size = resource_size(&crashk_res); mutex_unlock(&kexec_mutex); return size; } diff --git a/kernel/kmod.c b/kernel/kmod.c index ad6a81c58b4..47613dfb7b2 100644 --- a/kernel/kmod.c +++ b/kernel/kmod.c @@ -156,12 +156,6 @@ static int ____call_usermodehelper(void *data) */ set_user_nice(current, 0); - if (sub_info->init) { - retval = sub_info->init(sub_info); - if (retval) - goto fail; - } - retval = -ENOMEM; new = prepare_kernel_cred(current); if (!new) @@ -173,6 +167,14 @@ static int ____call_usermodehelper(void *data) new->cap_inheritable); spin_unlock(&umh_sysctl_lock); + if (sub_info->init) { + retval = sub_info->init(sub_info, new); + if (retval) { + abort_creds(new); + goto fail; + } + } + commit_creds(new); retval = kernel_execve(sub_info->path, @@ -388,7 +390,7 @@ EXPORT_SYMBOL(call_usermodehelper_setup); * context in which call_usermodehelper_exec is called. */ void call_usermodehelper_setfns(struct subprocess_info *info, - int (*init)(struct subprocess_info *info), + int (*init)(struct subprocess_info *info, struct cred *new), void (*cleanup)(struct subprocess_info *info), void *data) { diff --git a/kernel/kprobes.c b/kernel/kprobes.c index 77981813a1e..b30fd54eb98 100644 --- a/kernel/kprobes.c +++ b/kernel/kprobes.c @@ -1255,19 +1255,29 @@ static int __kprobes in_kprobes_functions(unsigned long addr) /* * If we have a symbol_name argument, look it up and add the offset field * to it. This way, we can specify a relative address to a symbol. + * This returns encoded errors if it fails to look up symbol or invalid + * combination of parameters. */ static kprobe_opcode_t __kprobes *kprobe_addr(struct kprobe *p) { kprobe_opcode_t *addr = p->addr; + + if ((p->symbol_name && p->addr) || + (!p->symbol_name && !p->addr)) + goto invalid; + if (p->symbol_name) { - if (addr) - return NULL; kprobe_lookup_name(p->symbol_name, addr); + if (!addr) + return ERR_PTR(-ENOENT); } - if (!addr) - return NULL; - return (kprobe_opcode_t *)(((char *)addr) + p->offset); + addr = (kprobe_opcode_t *)(((char *)addr) + p->offset); + if (addr) + return addr; + +invalid: + return ERR_PTR(-EINVAL); } /* Check passed kprobe is valid and return kprobe in kprobe_table. */ @@ -1311,8 +1321,8 @@ int __kprobes register_kprobe(struct kprobe *p) kprobe_opcode_t *addr; addr = kprobe_addr(p); - if (!addr) - return -EINVAL; + if (IS_ERR(addr)) + return PTR_ERR(addr); p->addr = addr; ret = check_kprobe_rereg(p); @@ -1335,6 +1345,8 @@ int __kprobes register_kprobe(struct kprobe *p) */ probed_mod = __module_text_address((unsigned long) p->addr); if (probed_mod) { + /* Return -ENOENT if fail. */ + ret = -ENOENT; /* * We must hold a refcount of the probed module while updating * its code to prohibit unexpected unloading. @@ -1351,6 +1363,7 @@ int __kprobes register_kprobe(struct kprobe *p) module_put(probed_mod); goto fail_with_jump_label; } + /* ret will be updated by following code */ } preempt_enable(); jump_label_unlock(); @@ -1399,7 +1412,7 @@ out: fail_with_jump_label: preempt_enable(); jump_label_unlock(); - return -EINVAL; + return ret; } EXPORT_SYMBOL_GPL(register_kprobe); @@ -1686,8 +1699,8 @@ int __kprobes register_kretprobe(struct kretprobe *rp) if (kretprobe_blacklist_size) { addr = kprobe_addr(&rp->kp); - if (!addr) - return -EINVAL; + if (IS_ERR(addr)) + return PTR_ERR(addr); for (i = 0; kretprobe_blacklist[i].name != NULL; i++) { if (kretprobe_blacklist[i].addr == addr) diff --git a/kernel/lockdep.c b/kernel/lockdep.c index 63437d065ac..3956f5149e2 100644 --- a/kernel/lockdep.c +++ b/kernel/lockdep.c @@ -2468,6 +2468,9 @@ mark_held_locks(struct task_struct *curr, enum mark_type mark) BUG_ON(usage_bit >= LOCK_USAGE_STATES); + if (hlock_class(hlock)->key == &__lockdep_no_validate__) + continue; + if (!mark_lock(curr, hlock, usage_bit)) return 0; } @@ -2478,15 +2481,10 @@ mark_held_locks(struct task_struct *curr, enum mark_type mark) /* * Hardirqs will be enabled: */ -void trace_hardirqs_on_caller(unsigned long ip) +static void __trace_hardirqs_on_caller(unsigned long ip) { struct task_struct *curr = current; - time_hardirqs_on(CALLER_ADDR0, ip); - - if (unlikely(!debug_locks || current->lockdep_recursion)) - return; - if (DEBUG_LOCKS_WARN_ON(unlikely(early_boot_irqs_disabled))) return; @@ -2502,8 +2500,6 @@ void trace_hardirqs_on_caller(unsigned long ip) /* we'll do an OFF -> ON transition: */ curr->hardirqs_enabled = 1; - if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) - return; if (DEBUG_LOCKS_WARN_ON(current->hardirq_context)) return; /* @@ -2525,6 +2521,21 @@ void trace_hardirqs_on_caller(unsigned long ip) curr->hardirq_enable_event = ++curr->irq_events; debug_atomic_inc(hardirqs_on_events); } + +void trace_hardirqs_on_caller(unsigned long ip) +{ + time_hardirqs_on(CALLER_ADDR0, ip); + + if (unlikely(!debug_locks || current->lockdep_recursion)) + return; + + if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) + return; + + current->lockdep_recursion = 1; + __trace_hardirqs_on_caller(ip); + current->lockdep_recursion = 0; +} EXPORT_SYMBOL(trace_hardirqs_on_caller); void trace_hardirqs_on(void) @@ -2574,7 +2585,7 @@ void trace_softirqs_on(unsigned long ip) { struct task_struct *curr = current; - if (unlikely(!debug_locks)) + if (unlikely(!debug_locks || current->lockdep_recursion)) return; if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) @@ -2585,6 +2596,7 @@ void trace_softirqs_on(unsigned long ip) return; } + current->lockdep_recursion = 1; /* * We'll do an OFF -> ON transition: */ @@ -2599,6 +2611,7 @@ void trace_softirqs_on(unsigned long ip) */ if (curr->hardirqs_enabled) mark_held_locks(curr, SOFTIRQ); + current->lockdep_recursion = 0; } /* @@ -2608,7 +2621,7 @@ void trace_softirqs_off(unsigned long ip) { struct task_struct *curr = current; - if (unlikely(!debug_locks)) + if (unlikely(!debug_locks || current->lockdep_recursion)) return; if (DEBUG_LOCKS_WARN_ON(!irqs_disabled())) @@ -3426,7 +3439,7 @@ int lock_is_held(struct lockdep_map *lock) int ret = 0; if (unlikely(current->lockdep_recursion)) - return ret; + return 1; /* avoid false negative lockdep_assert_held() */ raw_local_irq_save(flags); check_flags(flags); diff --git a/kernel/module.c b/kernel/module.c index 795bdc7f5c3..04379f92f84 100644 --- a/kernel/module.c +++ b/kernel/module.c @@ -545,9 +545,9 @@ static void setup_modinfo_##field(struct module *mod, const char *s) \ mod->field = kstrdup(s, GFP_KERNEL); \ } \ static ssize_t show_modinfo_##field(struct module_attribute *mattr, \ - struct module *mod, char *buffer) \ + struct module_kobject *mk, char *buffer) \ { \ - return sprintf(buffer, "%s\n", mod->field); \ + return sprintf(buffer, "%s\n", mk->mod->field); \ } \ static int modinfo_##field##_exists(struct module *mod) \ { \ @@ -902,9 +902,9 @@ void symbol_put_addr(void *addr) EXPORT_SYMBOL_GPL(symbol_put_addr); static ssize_t show_refcnt(struct module_attribute *mattr, - struct module *mod, char *buffer) + struct module_kobject *mk, char *buffer) { - return sprintf(buffer, "%u\n", module_refcount(mod)); + return sprintf(buffer, "%u\n", module_refcount(mk->mod)); } static struct module_attribute refcnt = { @@ -952,11 +952,11 @@ static inline int module_unload_init(struct module *mod) #endif /* CONFIG_MODULE_UNLOAD */ static ssize_t show_initstate(struct module_attribute *mattr, - struct module *mod, char *buffer) + struct module_kobject *mk, char *buffer) { const char *state = "unknown"; - switch (mod->state) { + switch (mk->mod->state) { case MODULE_STATE_LIVE: state = "live"; break; @@ -975,10 +975,27 @@ static struct module_attribute initstate = { .show = show_initstate, }; +static ssize_t store_uevent(struct module_attribute *mattr, + struct module_kobject *mk, + const char *buffer, size_t count) +{ + enum kobject_action action; + + if (kobject_action_type(buffer, count, &action) == 0) + kobject_uevent(&mk->kobj, action); + return count; +} + +struct module_attribute module_uevent = { + .attr = { .name = "uevent", .mode = 0200 }, + .store = store_uevent, +}; + static struct module_attribute *modinfo_attrs[] = { &modinfo_version, &modinfo_srcversion, &initstate, + &module_uevent, #ifdef CONFIG_MODULE_UNLOAD &refcnt, #endif @@ -1187,7 +1204,7 @@ struct module_sect_attrs }; static ssize_t module_sect_show(struct module_attribute *mattr, - struct module *mod, char *buf) + struct module_kobject *mk, char *buf) { struct module_sect_attr *sattr = container_of(mattr, struct module_sect_attr, mattr); @@ -1697,6 +1714,15 @@ static void unset_module_core_ro_nx(struct module *mod) { } static void unset_module_init_ro_nx(struct module *mod) { } #endif +void __weak module_free(struct module *mod, void *module_region) +{ + vfree(module_region); +} + +void __weak module_arch_cleanup(struct module *mod) +{ +} + /* Free a module, remove from lists, etc. */ static void free_module(struct module *mod) { @@ -1851,6 +1877,26 @@ static int simplify_symbols(struct module *mod, const struct load_info *info) return ret; } +int __weak apply_relocate(Elf_Shdr *sechdrs, + const char *strtab, + unsigned int symindex, + unsigned int relsec, + struct module *me) +{ + pr_err("module %s: REL relocation unsupported\n", me->name); + return -ENOEXEC; +} + +int __weak apply_relocate_add(Elf_Shdr *sechdrs, + const char *strtab, + unsigned int symindex, + unsigned int relsec, + struct module *me) +{ + pr_err("module %s: RELA relocation unsupported\n", me->name); + return -ENOEXEC; +} + static int apply_relocations(struct module *mod, const struct load_info *info) { unsigned int i; @@ -2235,6 +2281,11 @@ static void dynamic_debug_remove(struct _ddebug *debug) ddebug_remove_module(debug->modname); } +void * __weak module_alloc(unsigned long size) +{ + return size == 0 ? NULL : vmalloc_exec(size); +} + static void *module_alloc_update_bounds(unsigned long size) { void *ret = module_alloc(size); @@ -2645,6 +2696,14 @@ static void flush_module_icache(const struct module *mod) set_fs(old_fs); } +int __weak module_frob_arch_sections(Elf_Ehdr *hdr, + Elf_Shdr *sechdrs, + char *secstrings, + struct module *mod) +{ + return 0; +} + static struct module *layout_and_allocate(struct load_info *info) { /* Module within temporary copy. */ @@ -2716,6 +2775,13 @@ static void module_deallocate(struct module *mod, struct load_info *info) module_free(mod, mod->module_core); } +int __weak module_finalize(const Elf_Ehdr *hdr, + const Elf_Shdr *sechdrs, + struct module *me) +{ + return 0; +} + static int post_relocation(struct module *mod, const struct load_info *info) { /* Sort exception table now relocations are done. */ diff --git a/kernel/nsproxy.c b/kernel/nsproxy.c index d6a00f3de15..9aeab4b98c6 100644 --- a/kernel/nsproxy.c +++ b/kernel/nsproxy.c @@ -271,10 +271,8 @@ out: return err; } -static int __init nsproxy_cache_init(void) +int __init nsproxy_cache_init(void) { nsproxy_cachep = KMEM_CACHE(nsproxy, SLAB_PANIC); return 0; } - -module_init(nsproxy_cache_init); diff --git a/kernel/params.c b/kernel/params.c index ed72e133086..22df3e0d142 100644 --- a/kernel/params.c +++ b/kernel/params.c @@ -225,8 +225,8 @@ int parse_args(const char *name, int ret; \ \ ret = strtolfn(val, 0, &l); \ - if (ret == -EINVAL || ((type)l != l)) \ - return -EINVAL; \ + if (ret < 0 || ((type)l != l)) \ + return ret < 0 ? ret : -EINVAL; \ *((type *)kp->arg) = l; \ return 0; \ } \ @@ -511,7 +511,7 @@ struct module_param_attrs #define to_param_attr(n) container_of(n, struct param_attribute, mattr) static ssize_t param_attr_show(struct module_attribute *mattr, - struct module *mod, char *buf) + struct module_kobject *mk, char *buf) { int count; struct param_attribute *attribute = to_param_attr(mattr); @@ -531,7 +531,7 @@ static ssize_t param_attr_show(struct module_attribute *mattr, /* sysfs always hands a nul-terminated string in buf. We rely on that. */ static ssize_t param_attr_store(struct module_attribute *mattr, - struct module *owner, + struct module_kobject *km, const char *buf, size_t len) { int err; @@ -730,6 +730,10 @@ static struct module_kobject * __init locate_module_kobject(const char *name) mk->kobj.kset = module_kset; err = kobject_init_and_add(&mk->kobj, &module_ktype, NULL, "%s", name); +#ifdef CONFIG_MODULES + if (!err) + err = sysfs_create_file(&mk->kobj, &module_uevent.attr); +#endif if (err) { kobject_put(&mk->kobj); printk(KERN_ERR @@ -807,7 +811,7 @@ static void __init param_sysfs_builtin(void) } ssize_t __modver_version_show(struct module_attribute *mattr, - struct module *mod, char *buf) + struct module_kobject *mk, char *buf) { struct module_version_attribute *vattr = container_of(mattr, struct module_version_attribute, mattr); @@ -852,7 +856,7 @@ static ssize_t module_attr_show(struct kobject *kobj, if (!attribute->show) return -EIO; - ret = attribute->show(attribute, mk->mod, buf); + ret = attribute->show(attribute, mk, buf); return ret; } @@ -871,7 +875,7 @@ static ssize_t module_attr_store(struct kobject *kobj, if (!attribute->store) return -EIO; - ret = attribute->store(attribute, mk->mod, buf, len); + ret = attribute->store(attribute, mk, buf, len); return ret; } diff --git a/kernel/pid.c b/kernel/pid.c index 57a8346a270..e432057f3b2 100644 --- a/kernel/pid.c +++ b/kernel/pid.c @@ -405,7 +405,6 @@ struct task_struct *pid_task(struct pid *pid, enum pid_type type) if (pid) { struct hlist_node *first; first = rcu_dereference_check(hlist_first_rcu(&pid->tasks[type]), - rcu_read_lock_held() || lockdep_tasklist_lock_is_held()); if (first) result = hlist_entry(first, struct task_struct, pids[(type)].node); diff --git a/kernel/pm_qos_params.c b/kernel/pm_qos_params.c index 6824ca7d4d0..37f05d0f079 100644 --- a/kernel/pm_qos_params.c +++ b/kernel/pm_qos_params.c @@ -74,7 +74,7 @@ static DEFINE_SPINLOCK(pm_qos_lock); static struct pm_qos_object null_pm_qos; static BLOCKING_NOTIFIER_HEAD(cpu_dma_lat_notifier); static struct pm_qos_object cpu_dma_pm_qos = { - .requests = PLIST_HEAD_INIT(cpu_dma_pm_qos.requests, pm_qos_lock), + .requests = PLIST_HEAD_INIT(cpu_dma_pm_qos.requests), .notifiers = &cpu_dma_lat_notifier, .name = "cpu_dma_latency", .target_value = PM_QOS_CPU_DMA_LAT_DEFAULT_VALUE, @@ -84,7 +84,7 @@ static struct pm_qos_object cpu_dma_pm_qos = { static BLOCKING_NOTIFIER_HEAD(network_lat_notifier); static struct pm_qos_object network_lat_pm_qos = { - .requests = PLIST_HEAD_INIT(network_lat_pm_qos.requests, pm_qos_lock), + .requests = PLIST_HEAD_INIT(network_lat_pm_qos.requests), .notifiers = &network_lat_notifier, .name = "network_latency", .target_value = PM_QOS_NETWORK_LAT_DEFAULT_VALUE, @@ -95,7 +95,7 @@ static struct pm_qos_object network_lat_pm_qos = { static BLOCKING_NOTIFIER_HEAD(network_throughput_notifier); static struct pm_qos_object network_throughput_pm_qos = { - .requests = PLIST_HEAD_INIT(network_throughput_pm_qos.requests, pm_qos_lock), + .requests = PLIST_HEAD_INIT(network_throughput_pm_qos.requests), .notifiers = &network_throughput_notifier, .name = "network_throughput", .target_value = PM_QOS_NETWORK_THROUGHPUT_DEFAULT_VALUE, diff --git a/kernel/power/Kconfig b/kernel/power/Kconfig index 87f4d24b55b..b1914cb9095 100644 --- a/kernel/power/Kconfig +++ b/kernel/power/Kconfig @@ -193,8 +193,8 @@ config APM_EMULATION notification of APM "events" (e.g. battery status change). In order to use APM, you will need supporting software. For location - and more information, read <file:Documentation/power/pm.txt> and the - Battery Powered Linux mini-HOWTO, available from + and more information, read <file:Documentation/power/apm-acpi.txt> + and the Battery Powered Linux mini-HOWTO, available from <http://www.tldp.org/docs.html#howto>. This driver does not spin down disk drives (see the hdparm(8) @@ -224,6 +224,10 @@ config PM_OPP implementations a ready to use framework to manage OPPs. For more information, read <file:Documentation/power/opp.txt> -config PM_RUNTIME_CLK +config PM_CLK def_bool y - depends on PM_RUNTIME && HAVE_CLK + depends on PM && HAVE_CLK + +config PM_GENERIC_DOMAINS + bool + depends on PM diff --git a/kernel/power/main.c b/kernel/power/main.c index 2981af4ce7c..6c601f87196 100644 --- a/kernel/power/main.c +++ b/kernel/power/main.c @@ -37,8 +37,9 @@ EXPORT_SYMBOL_GPL(unregister_pm_notifier); int pm_notifier_call_chain(unsigned long val) { - return (blocking_notifier_call_chain(&pm_chain_head, val, NULL) - == NOTIFY_BAD) ? -EINVAL : 0; + int ret = blocking_notifier_call_chain(&pm_chain_head, val, NULL); + + return notifier_to_errno(ret); } /* If set, devices may be suspended and resumed asynchronously. */ diff --git a/kernel/power/snapshot.c b/kernel/power/snapshot.c index ace55889f70..06efa54f93d 100644 --- a/kernel/power/snapshot.c +++ b/kernel/power/snapshot.c @@ -1211,7 +1211,11 @@ static void free_unnecessary_pages(void) to_free_highmem = alloc_highmem - save; } else { to_free_highmem = 0; - to_free_normal -= save - alloc_highmem; + save -= alloc_highmem; + if (to_free_normal > save) + to_free_normal -= save; + else + to_free_normal = 0; } memory_bm_position_reset(©_bm); diff --git a/kernel/power/suspend.c b/kernel/power/suspend.c index 1c41ba21541..b6b71ad2208 100644 --- a/kernel/power/suspend.c +++ b/kernel/power/suspend.c @@ -44,6 +44,7 @@ void suspend_set_ops(const struct platform_suspend_ops *ops) suspend_ops = ops; mutex_unlock(&pm_mutex); } +EXPORT_SYMBOL_GPL(suspend_set_ops); bool valid_state(suspend_state_t state) { @@ -65,6 +66,7 @@ int suspend_valid_only_mem(suspend_state_t state) { return state == PM_SUSPEND_MEM; } +EXPORT_SYMBOL_GPL(suspend_valid_only_mem); static int suspend_test(int level) { @@ -126,12 +128,13 @@ void __attribute__ ((weak)) arch_suspend_enable_irqs(void) } /** - * suspend_enter - enter the desired system sleep state. - * @state: state to enter + * suspend_enter - enter the desired system sleep state. + * @state: State to enter + * @wakeup: Returns information that suspend should not be entered again. * - * This function should be called after devices have been suspended. + * This function should be called after devices have been suspended. */ -static int suspend_enter(suspend_state_t state) +static int suspend_enter(suspend_state_t state, bool *wakeup) { int error; @@ -165,7 +168,8 @@ static int suspend_enter(suspend_state_t state) error = syscore_suspend(); if (!error) { - if (!(suspend_test(TEST_CORE) || pm_wakeup_pending())) { + *wakeup = pm_wakeup_pending(); + if (!(suspend_test(TEST_CORE) || *wakeup)) { error = suspend_ops->enter(state); events_check_enabled = false; } @@ -199,6 +203,7 @@ static int suspend_enter(suspend_state_t state) int suspend_devices_and_enter(suspend_state_t state) { int error; + bool wakeup = false; if (!suspend_ops) return -ENOSYS; @@ -220,7 +225,10 @@ int suspend_devices_and_enter(suspend_state_t state) if (suspend_test(TEST_DEVICES)) goto Recover_platform; - error = suspend_enter(state); + do { + error = suspend_enter(state, &wakeup); + } while (!error && !wakeup + && suspend_ops->suspend_again && suspend_ops->suspend_again()); Resume_devices: suspend_test_start(); diff --git a/kernel/power/user.c b/kernel/power/user.c index 7d02d33be69..42ddbc6f0de 100644 --- a/kernel/power/user.c +++ b/kernel/power/user.c @@ -113,8 +113,10 @@ static int snapshot_open(struct inode *inode, struct file *filp) if (error) pm_notifier_call_chain(PM_POST_RESTORE); } - if (error) + if (error) { + free_basic_memory_bitmaps(); atomic_inc(&snapshot_device_available); + } data->frozen = 0; data->ready = 0; data->platform_support = 0; diff --git a/kernel/printk.c b/kernel/printk.c index 35185392173..37dff3429ad 100644 --- a/kernel/printk.c +++ b/kernel/printk.c @@ -782,7 +782,7 @@ static inline int can_use_console(unsigned int cpu) static int console_trylock_for_printk(unsigned int cpu) __releases(&logbuf_lock) { - int retval = 0; + int retval = 0, wake = 0; if (console_trylock()) { retval = 1; @@ -795,12 +795,14 @@ static int console_trylock_for_printk(unsigned int cpu) */ if (!can_use_console(cpu)) { console_locked = 0; - up(&console_sem); + wake = 1; retval = 0; } } printk_cpu = UINT_MAX; spin_unlock(&logbuf_lock); + if (wake) + up(&console_sem); return retval; } static const char recursion_bug_msg [] = @@ -1242,7 +1244,7 @@ void console_unlock(void) { unsigned long flags; unsigned _con_start, _log_end; - unsigned wake_klogd = 0; + unsigned wake_klogd = 0, retry = 0; if (console_suspended) { up(&console_sem); @@ -1251,6 +1253,7 @@ void console_unlock(void) console_may_schedule = 0; +again: for ( ; ; ) { spin_lock_irqsave(&logbuf_lock, flags); wake_klogd |= log_start - log_end; @@ -1271,8 +1274,23 @@ void console_unlock(void) if (unlikely(exclusive_console)) exclusive_console = NULL; + spin_unlock(&logbuf_lock); + up(&console_sem); + + /* + * Someone could have filled up the buffer again, so re-check if there's + * something to flush. In case we cannot trylock the console_sem again, + * there's a new owner and the console_unlock() from them will do the + * flush, no worries. + */ + spin_lock(&logbuf_lock); + if (con_start != log_end) + retry = 1; spin_unlock_irqrestore(&logbuf_lock, flags); + if (retry && console_trylock()) + goto again; + if (wake_klogd) wake_up_klogd(); } diff --git a/kernel/ptrace.c b/kernel/ptrace.c index 2df115790cd..9de3ecfd20f 100644 --- a/kernel/ptrace.c +++ b/kernel/ptrace.c @@ -23,8 +23,15 @@ #include <linux/uaccess.h> #include <linux/regset.h> #include <linux/hw_breakpoint.h> +#include <linux/cn_proc.h> +static int ptrace_trapping_sleep_fn(void *flags) +{ + schedule(); + return 0; +} + /* * ptrace a task: make the debugger its new parent and * move it to the ptrace list. @@ -77,13 +84,20 @@ void __ptrace_unlink(struct task_struct *child) spin_lock(&child->sighand->siglock); /* - * Reinstate GROUP_STOP_PENDING if group stop is in effect and + * Clear all pending traps and TRAPPING. TRAPPING should be + * cleared regardless of JOBCTL_STOP_PENDING. Do it explicitly. + */ + task_clear_jobctl_pending(child, JOBCTL_TRAP_MASK); + task_clear_jobctl_trapping(child); + + /* + * Reinstate JOBCTL_STOP_PENDING if group stop is in effect and * @child isn't dead. */ if (!(child->flags & PF_EXITING) && (child->signal->flags & SIGNAL_STOP_STOPPED || child->signal->group_stop_count)) - child->group_stop |= GROUP_STOP_PENDING; + child->jobctl |= JOBCTL_STOP_PENDING; /* * If transition to TASK_STOPPED is pending or in TASK_TRACED, kick @@ -91,16 +105,30 @@ void __ptrace_unlink(struct task_struct *child) * is in TASK_TRACED; otherwise, we might unduly disrupt * TASK_KILLABLE sleeps. */ - if (child->group_stop & GROUP_STOP_PENDING || task_is_traced(child)) + if (child->jobctl & JOBCTL_STOP_PENDING || task_is_traced(child)) signal_wake_up(child, task_is_traced(child)); spin_unlock(&child->sighand->siglock); } -/* - * Check that we have indeed attached to the thing.. +/** + * ptrace_check_attach - check whether ptracee is ready for ptrace operation + * @child: ptracee to check for + * @ignore_state: don't check whether @child is currently %TASK_TRACED + * + * Check whether @child is being ptraced by %current and ready for further + * ptrace operations. If @ignore_state is %false, @child also should be in + * %TASK_TRACED state and on return the child is guaranteed to be traced + * and not executing. If @ignore_state is %true, @child can be in any + * state. + * + * CONTEXT: + * Grabs and releases tasklist_lock and @child->sighand->siglock. + * + * RETURNS: + * 0 on success, -ESRCH if %child is not ready. */ -int ptrace_check_attach(struct task_struct *child, int kill) +int ptrace_check_attach(struct task_struct *child, bool ignore_state) { int ret = -ESRCH; @@ -119,13 +147,14 @@ int ptrace_check_attach(struct task_struct *child, int kill) */ spin_lock_irq(&child->sighand->siglock); WARN_ON_ONCE(task_is_stopped(child)); - if (task_is_traced(child) || kill) + if (ignore_state || (task_is_traced(child) && + !(child->jobctl & JOBCTL_LISTENING))) ret = 0; spin_unlock_irq(&child->sighand->siglock); } read_unlock(&tasklist_lock); - if (!ret && !kill) + if (!ret && !ignore_state) ret = wait_task_inactive(child, TASK_TRACED) ? 0 : -ESRCH; /* All systems go.. */ @@ -182,11 +211,28 @@ bool ptrace_may_access(struct task_struct *task, unsigned int mode) return !err; } -static int ptrace_attach(struct task_struct *task) +static int ptrace_attach(struct task_struct *task, long request, + unsigned long flags) { - bool wait_trap = false; + bool seize = (request == PTRACE_SEIZE); int retval; + /* + * SEIZE will enable new ptrace behaviors which will be implemented + * gradually. SEIZE_DEVEL is used to prevent applications + * expecting full SEIZE behaviors trapping on kernel commits which + * are still in the process of implementing them. + * + * Only test programs for new ptrace behaviors being implemented + * should set SEIZE_DEVEL. If unset, SEIZE will fail with -EIO. + * + * Once SEIZE behaviors are completely implemented, this flag and + * the following test will be removed. + */ + retval = -EIO; + if (seize && !(flags & PTRACE_SEIZE_DEVEL)) + goto out; + audit_ptrace(task); retval = -EPERM; @@ -218,16 +264,21 @@ static int ptrace_attach(struct task_struct *task) goto unlock_tasklist; task->ptrace = PT_PTRACED; + if (seize) + task->ptrace |= PT_SEIZED; if (task_ns_capable(task, CAP_SYS_PTRACE)) task->ptrace |= PT_PTRACE_CAP; __ptrace_link(task, current); - send_sig_info(SIGSTOP, SEND_SIG_FORCED, task); + + /* SEIZE doesn't trap tracee on attach */ + if (!seize) + send_sig_info(SIGSTOP, SEND_SIG_FORCED, task); spin_lock(&task->sighand->siglock); /* - * If the task is already STOPPED, set GROUP_STOP_PENDING and + * If the task is already STOPPED, set JOBCTL_TRAP_STOP and * TRAPPING, and kick it so that it transits to TRACED. TRAPPING * will be cleared if the child completes the transition or any * event which clears the group stop states happens. We'll wait @@ -243,11 +294,9 @@ static int ptrace_attach(struct task_struct *task) * The following task_is_stopped() test is safe as both transitions * in and out of STOPPED are protected by siglock. */ - if (task_is_stopped(task)) { - task->group_stop |= GROUP_STOP_PENDING | GROUP_STOP_TRAPPING; + if (task_is_stopped(task) && + task_set_jobctl_pending(task, JOBCTL_TRAP_STOP | JOBCTL_TRAPPING)) signal_wake_up(task, 1); - wait_trap = true; - } spin_unlock(&task->sighand->siglock); @@ -257,9 +306,12 @@ unlock_tasklist: unlock_creds: mutex_unlock(&task->signal->cred_guard_mutex); out: - if (wait_trap) - wait_event(current->signal->wait_chldexit, - !(task->group_stop & GROUP_STOP_TRAPPING)); + if (!retval) { + wait_on_bit(&task->jobctl, JOBCTL_TRAPPING_BIT, + ptrace_trapping_sleep_fn, TASK_UNINTERRUPTIBLE); + proc_ptrace_connector(task, PTRACE_ATTACH); + } + return retval; } @@ -322,25 +374,27 @@ static int ignoring_children(struct sighand_struct *sigh) */ static bool __ptrace_detach(struct task_struct *tracer, struct task_struct *p) { + bool dead; + __ptrace_unlink(p); - if (p->exit_state == EXIT_ZOMBIE) { - if (!task_detached(p) && thread_group_empty(p)) { - if (!same_thread_group(p->real_parent, tracer)) - do_notify_parent(p, p->exit_signal); - else if (ignoring_children(tracer->sighand)) { - __wake_up_parent(p, tracer); - p->exit_signal = -1; - } - } - if (task_detached(p)) { - /* Mark it as in the process of being reaped. */ - p->exit_state = EXIT_DEAD; - return true; + if (p->exit_state != EXIT_ZOMBIE) + return false; + + dead = !thread_group_leader(p); + + if (!dead && thread_group_empty(p)) { + if (!same_thread_group(p->real_parent, tracer)) + dead = do_notify_parent(p, p->exit_signal); + else if (ignoring_children(tracer->sighand)) { + __wake_up_parent(p, tracer); + dead = true; } } - - return false; + /* Mark it as in the process of being reaped. */ + if (dead) + p->exit_state = EXIT_DEAD; + return dead; } static int ptrace_detach(struct task_struct *child, unsigned int data) @@ -365,6 +419,7 @@ static int ptrace_detach(struct task_struct *child, unsigned int data) } write_unlock_irq(&tasklist_lock); + proc_ptrace_connector(child, PTRACE_DETACH); if (unlikely(dead)) release_task(child); @@ -611,10 +666,12 @@ static int ptrace_regset(struct task_struct *task, int req, unsigned int type, int ptrace_request(struct task_struct *child, long request, unsigned long addr, unsigned long data) { + bool seized = child->ptrace & PT_SEIZED; int ret = -EIO; - siginfo_t siginfo; + siginfo_t siginfo, *si; void __user *datavp = (void __user *) data; unsigned long __user *datalp = datavp; + unsigned long flags; switch (request) { case PTRACE_PEEKTEXT: @@ -647,6 +704,62 @@ int ptrace_request(struct task_struct *child, long request, ret = ptrace_setsiginfo(child, &siginfo); break; + case PTRACE_INTERRUPT: + /* + * Stop tracee without any side-effect on signal or job + * control. At least one trap is guaranteed to happen + * after this request. If @child is already trapped, the + * current trap is not disturbed and another trap will + * happen after the current trap is ended with PTRACE_CONT. + * + * The actual trap might not be PTRACE_EVENT_STOP trap but + * the pending condition is cleared regardless. + */ + if (unlikely(!seized || !lock_task_sighand(child, &flags))) + break; + + /* + * INTERRUPT doesn't disturb existing trap sans one + * exception. If ptracer issued LISTEN for the current + * STOP, this INTERRUPT should clear LISTEN and re-trap + * tracee into STOP. + */ + if (likely(task_set_jobctl_pending(child, JOBCTL_TRAP_STOP))) + signal_wake_up(child, child->jobctl & JOBCTL_LISTENING); + + unlock_task_sighand(child, &flags); + ret = 0; + break; + + case PTRACE_LISTEN: + /* + * Listen for events. Tracee must be in STOP. It's not + * resumed per-se but is not considered to be in TRACED by + * wait(2) or ptrace(2). If an async event (e.g. group + * stop state change) happens, tracee will enter STOP trap + * again. Alternatively, ptracer can issue INTERRUPT to + * finish listening and re-trap tracee into STOP. + */ + if (unlikely(!seized || !lock_task_sighand(child, &flags))) + break; + + si = child->last_siginfo; + if (unlikely(!si || si->si_code >> 8 != PTRACE_EVENT_STOP)) + break; + + child->jobctl |= JOBCTL_LISTENING; + + /* + * If NOTIFY is set, it means event happened between start + * of this trap and now. Trigger re-trap immediately. + */ + if (child->jobctl & JOBCTL_TRAP_NOTIFY) + signal_wake_up(child, true); + + unlock_task_sighand(child, &flags); + ret = 0; + break; + case PTRACE_DETACH: /* detach a process that was attached. */ ret = ptrace_detach(child, data); break; @@ -761,8 +874,8 @@ SYSCALL_DEFINE4(ptrace, long, request, long, pid, unsigned long, addr, goto out; } - if (request == PTRACE_ATTACH) { - ret = ptrace_attach(child); + if (request == PTRACE_ATTACH || request == PTRACE_SEIZE) { + ret = ptrace_attach(child, request, data); /* * Some architectures need to do book-keeping after * a ptrace attach. @@ -772,7 +885,8 @@ SYSCALL_DEFINE4(ptrace, long, request, long, pid, unsigned long, addr, goto out_put_task_struct; } - ret = ptrace_check_attach(child, request == PTRACE_KILL); + ret = ptrace_check_attach(child, request == PTRACE_KILL || + request == PTRACE_INTERRUPT); if (ret < 0) goto out_put_task_struct; @@ -903,8 +1017,8 @@ asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid, goto out; } - if (request == PTRACE_ATTACH) { - ret = ptrace_attach(child); + if (request == PTRACE_ATTACH || request == PTRACE_SEIZE) { + ret = ptrace_attach(child, request, data); /* * Some architectures need to do book-keeping after * a ptrace attach. @@ -914,7 +1028,8 @@ asmlinkage long compat_sys_ptrace(compat_long_t request, compat_long_t pid, goto out_put_task_struct; } - ret = ptrace_check_attach(child, request == PTRACE_KILL); + ret = ptrace_check_attach(child, request == PTRACE_KILL || + request == PTRACE_INTERRUPT); if (!ret) ret = compat_arch_ptrace(child, request, addr, data); diff --git a/kernel/rcutorture.c b/kernel/rcutorture.c index 2e138db0338..ced72102adc 100644 --- a/kernel/rcutorture.c +++ b/kernel/rcutorture.c @@ -941,7 +941,6 @@ static void rcu_torture_timer(unsigned long unused) idx = cur_ops->readlock(); completed = cur_ops->completed(); p = rcu_dereference_check(rcu_torture_current, - rcu_read_lock_held() || rcu_read_lock_bh_held() || rcu_read_lock_sched_held() || srcu_read_lock_held(&srcu_ctl)); @@ -1002,7 +1001,6 @@ rcu_torture_reader(void *arg) idx = cur_ops->readlock(); completed = cur_ops->completed(); p = rcu_dereference_check(rcu_torture_current, - rcu_read_lock_held() || rcu_read_lock_bh_held() || rcu_read_lock_sched_held() || srcu_read_lock_held(&srcu_ctl)); diff --git a/kernel/rcutree.c b/kernel/rcutree.c index 77a7671dd14..ba06207b1dd 100644 --- a/kernel/rcutree.c +++ b/kernel/rcutree.c @@ -84,10 +84,35 @@ DEFINE_PER_CPU(struct rcu_data, rcu_bh_data); static struct rcu_state *rcu_state; +/* + * The rcu_scheduler_active variable transitions from zero to one just + * before the first task is spawned. So when this variable is zero, RCU + * can assume that there is but one task, allowing RCU to (for example) + * optimized synchronize_sched() to a simple barrier(). When this variable + * is one, RCU must actually do all the hard work required to detect real + * grace periods. This variable is also used to suppress boot-time false + * positives from lockdep-RCU error checking. + */ int rcu_scheduler_active __read_mostly; EXPORT_SYMBOL_GPL(rcu_scheduler_active); /* + * The rcu_scheduler_fully_active variable transitions from zero to one + * during the early_initcall() processing, which is after the scheduler + * is capable of creating new tasks. So RCU processing (for example, + * creating tasks for RCU priority boosting) must be delayed until after + * rcu_scheduler_fully_active transitions from zero to one. We also + * currently delay invocation of any RCU callbacks until after this point. + * + * It might later prove better for people registering RCU callbacks during + * early boot to take responsibility for these callbacks, but one step at + * a time. + */ +static int rcu_scheduler_fully_active __read_mostly; + +#ifdef CONFIG_RCU_BOOST + +/* * Control variables for per-CPU and per-rcu_node kthreads. These * handle all flavors of RCU. */ @@ -96,10 +121,12 @@ DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_status); DEFINE_PER_CPU(int, rcu_cpu_kthread_cpu); DEFINE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); DEFINE_PER_CPU(char, rcu_cpu_has_work); -static char rcu_kthreads_spawnable; + +#endif /* #ifdef CONFIG_RCU_BOOST */ static void rcu_node_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu); -static void invoke_rcu_cpu_kthread(void); +static void invoke_rcu_core(void); +static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp); #define RCU_KTHREAD_PRIO 1 /* RT priority for per-CPU kthreads. */ @@ -1088,14 +1115,8 @@ static void __rcu_offline_cpu(int cpu, struct rcu_state *rsp) int need_report = 0; struct rcu_data *rdp = per_cpu_ptr(rsp->rda, cpu); struct rcu_node *rnp; - struct task_struct *t; - /* Stop the CPU's kthread. */ - t = per_cpu(rcu_cpu_kthread_task, cpu); - if (t != NULL) { - per_cpu(rcu_cpu_kthread_task, cpu) = NULL; - kthread_stop(t); - } + rcu_stop_cpu_kthread(cpu); /* Exclude any attempts to start a new grace period. */ raw_spin_lock_irqsave(&rsp->onofflock, flags); @@ -1231,7 +1252,7 @@ static void rcu_do_batch(struct rcu_state *rsp, struct rcu_data *rdp) /* Re-raise the RCU softirq if there are callbacks remaining. */ if (cpu_has_callbacks_ready_to_invoke(rdp)) - invoke_rcu_cpu_kthread(); + invoke_rcu_core(); } /* @@ -1277,7 +1298,7 @@ void rcu_check_callbacks(int cpu, int user) } rcu_preempt_check_callbacks(cpu); if (rcu_pending(cpu)) - invoke_rcu_cpu_kthread(); + invoke_rcu_core(); } #ifdef CONFIG_SMP @@ -1442,13 +1463,14 @@ __rcu_process_callbacks(struct rcu_state *rsp, struct rcu_data *rdp) } /* If there are callbacks ready, invoke them. */ - rcu_do_batch(rsp, rdp); + if (cpu_has_callbacks_ready_to_invoke(rdp)) + invoke_rcu_callbacks(rsp, rdp); } /* * Do softirq processing for the current CPU. */ -static void rcu_process_callbacks(void) +static void rcu_process_callbacks(struct softirq_action *unused) { __rcu_process_callbacks(&rcu_sched_state, &__get_cpu_var(rcu_sched_data)); @@ -1465,330 +1487,22 @@ static void rcu_process_callbacks(void) * the current CPU with interrupts disabled, the rcu_cpu_kthread_task * cannot disappear out from under us. */ -static void invoke_rcu_cpu_kthread(void) +static void invoke_rcu_callbacks(struct rcu_state *rsp, struct rcu_data *rdp) { - unsigned long flags; - - local_irq_save(flags); - __this_cpu_write(rcu_cpu_has_work, 1); - if (__this_cpu_read(rcu_cpu_kthread_task) == NULL) { - local_irq_restore(flags); + if (unlikely(!ACCESS_ONCE(rcu_scheduler_fully_active))) return; - } - wake_up_process(__this_cpu_read(rcu_cpu_kthread_task)); - local_irq_restore(flags); -} - -/* - * Wake up the specified per-rcu_node-structure kthread. - * Because the per-rcu_node kthreads are immortal, we don't need - * to do anything to keep them alive. - */ -static void invoke_rcu_node_kthread(struct rcu_node *rnp) -{ - struct task_struct *t; - - t = rnp->node_kthread_task; - if (t != NULL) - wake_up_process(t); -} - -/* - * Set the specified CPU's kthread to run RT or not, as specified by - * the to_rt argument. The CPU-hotplug locks are held, so the task - * is not going away. - */ -static void rcu_cpu_kthread_setrt(int cpu, int to_rt) -{ - int policy; - struct sched_param sp; - struct task_struct *t; - - t = per_cpu(rcu_cpu_kthread_task, cpu); - if (t == NULL) + if (likely(!rsp->boost)) { + rcu_do_batch(rsp, rdp); return; - if (to_rt) { - policy = SCHED_FIFO; - sp.sched_priority = RCU_KTHREAD_PRIO; - } else { - policy = SCHED_NORMAL; - sp.sched_priority = 0; } - sched_setscheduler_nocheck(t, policy, &sp); + invoke_rcu_callbacks_kthread(); } -/* - * Timer handler to initiate the waking up of per-CPU kthreads that - * have yielded the CPU due to excess numbers of RCU callbacks. - * We wake up the per-rcu_node kthread, which in turn will wake up - * the booster kthread. - */ -static void rcu_cpu_kthread_timer(unsigned long arg) +static void invoke_rcu_core(void) { - struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, arg); - struct rcu_node *rnp = rdp->mynode; - - atomic_or(rdp->grpmask, &rnp->wakemask); - invoke_rcu_node_kthread(rnp); -} - -/* - * Drop to non-real-time priority and yield, but only after posting a - * timer that will cause us to regain our real-time priority if we - * remain preempted. Either way, we restore our real-time priority - * before returning. - */ -static void rcu_yield(void (*f)(unsigned long), unsigned long arg) -{ - struct sched_param sp; - struct timer_list yield_timer; - - setup_timer_on_stack(&yield_timer, f, arg); - mod_timer(&yield_timer, jiffies + 2); - sp.sched_priority = 0; - sched_setscheduler_nocheck(current, SCHED_NORMAL, &sp); - set_user_nice(current, 19); - schedule(); - sp.sched_priority = RCU_KTHREAD_PRIO; - sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); - del_timer(&yield_timer); -} - -/* - * Handle cases where the rcu_cpu_kthread() ends up on the wrong CPU. - * This can happen while the corresponding CPU is either coming online - * or going offline. We cannot wait until the CPU is fully online - * before starting the kthread, because the various notifier functions - * can wait for RCU grace periods. So we park rcu_cpu_kthread() until - * the corresponding CPU is online. - * - * Return 1 if the kthread needs to stop, 0 otherwise. - * - * Caller must disable bh. This function can momentarily enable it. - */ -static int rcu_cpu_kthread_should_stop(int cpu) -{ - while (cpu_is_offline(cpu) || - !cpumask_equal(¤t->cpus_allowed, cpumask_of(cpu)) || - smp_processor_id() != cpu) { - if (kthread_should_stop()) - return 1; - per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; - per_cpu(rcu_cpu_kthread_cpu, cpu) = raw_smp_processor_id(); - local_bh_enable(); - schedule_timeout_uninterruptible(1); - if (!cpumask_equal(¤t->cpus_allowed, cpumask_of(cpu))) - set_cpus_allowed_ptr(current, cpumask_of(cpu)); - local_bh_disable(); - } - per_cpu(rcu_cpu_kthread_cpu, cpu) = cpu; - return 0; + raise_softirq(RCU_SOFTIRQ); } -/* - * Per-CPU kernel thread that invokes RCU callbacks. This replaces the - * earlier RCU softirq. - */ -static int rcu_cpu_kthread(void *arg) -{ - int cpu = (int)(long)arg; - unsigned long flags; - int spincnt = 0; - unsigned int *statusp = &per_cpu(rcu_cpu_kthread_status, cpu); - char work; - char *workp = &per_cpu(rcu_cpu_has_work, cpu); - - for (;;) { - *statusp = RCU_KTHREAD_WAITING; - rcu_wait(*workp != 0 || kthread_should_stop()); - local_bh_disable(); - if (rcu_cpu_kthread_should_stop(cpu)) { - local_bh_enable(); - break; - } - *statusp = RCU_KTHREAD_RUNNING; - per_cpu(rcu_cpu_kthread_loops, cpu)++; - local_irq_save(flags); - work = *workp; - *workp = 0; - local_irq_restore(flags); - if (work) - rcu_process_callbacks(); - local_bh_enable(); - if (*workp != 0) - spincnt++; - else - spincnt = 0; - if (spincnt > 10) { - *statusp = RCU_KTHREAD_YIELDING; - rcu_yield(rcu_cpu_kthread_timer, (unsigned long)cpu); - spincnt = 0; - } - } - *statusp = RCU_KTHREAD_STOPPED; - return 0; -} - -/* - * Spawn a per-CPU kthread, setting up affinity and priority. - * Because the CPU hotplug lock is held, no other CPU will be attempting - * to manipulate rcu_cpu_kthread_task. There might be another CPU - * attempting to access it during boot, but the locking in kthread_bind() - * will enforce sufficient ordering. - */ -static int __cpuinit rcu_spawn_one_cpu_kthread(int cpu) -{ - struct sched_param sp; - struct task_struct *t; - - if (!rcu_kthreads_spawnable || - per_cpu(rcu_cpu_kthread_task, cpu) != NULL) - return 0; - t = kthread_create(rcu_cpu_kthread, (void *)(long)cpu, "rcuc%d", cpu); - if (IS_ERR(t)) - return PTR_ERR(t); - kthread_bind(t, cpu); - set_task_state(t, TASK_INTERRUPTIBLE); - per_cpu(rcu_cpu_kthread_cpu, cpu) = cpu; - WARN_ON_ONCE(per_cpu(rcu_cpu_kthread_task, cpu) != NULL); - per_cpu(rcu_cpu_kthread_task, cpu) = t; - sp.sched_priority = RCU_KTHREAD_PRIO; - sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); - return 0; -} - -/* - * Per-rcu_node kthread, which is in charge of waking up the per-CPU - * kthreads when needed. We ignore requests to wake up kthreads - * for offline CPUs, which is OK because force_quiescent_state() - * takes care of this case. - */ -static int rcu_node_kthread(void *arg) -{ - int cpu; - unsigned long flags; - unsigned long mask; - struct rcu_node *rnp = (struct rcu_node *)arg; - struct sched_param sp; - struct task_struct *t; - - for (;;) { - rnp->node_kthread_status = RCU_KTHREAD_WAITING; - rcu_wait(atomic_read(&rnp->wakemask) != 0); - rnp->node_kthread_status = RCU_KTHREAD_RUNNING; - raw_spin_lock_irqsave(&rnp->lock, flags); - mask = atomic_xchg(&rnp->wakemask, 0); - rcu_initiate_boost(rnp, flags); /* releases rnp->lock. */ - for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask >>= 1) { - if ((mask & 0x1) == 0) - continue; - preempt_disable(); - t = per_cpu(rcu_cpu_kthread_task, cpu); - if (!cpu_online(cpu) || t == NULL) { - preempt_enable(); - continue; - } - per_cpu(rcu_cpu_has_work, cpu) = 1; - sp.sched_priority = RCU_KTHREAD_PRIO; - sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); - preempt_enable(); - } - } - /* NOTREACHED */ - rnp->node_kthread_status = RCU_KTHREAD_STOPPED; - return 0; -} - -/* - * Set the per-rcu_node kthread's affinity to cover all CPUs that are - * served by the rcu_node in question. The CPU hotplug lock is still - * held, so the value of rnp->qsmaskinit will be stable. - * - * We don't include outgoingcpu in the affinity set, use -1 if there is - * no outgoing CPU. If there are no CPUs left in the affinity set, - * this function allows the kthread to execute on any CPU. - */ -static void rcu_node_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) -{ - cpumask_var_t cm; - int cpu; - unsigned long mask = rnp->qsmaskinit; - - if (rnp->node_kthread_task == NULL) - return; - if (!alloc_cpumask_var(&cm, GFP_KERNEL)) - return; - cpumask_clear(cm); - for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask >>= 1) - if ((mask & 0x1) && cpu != outgoingcpu) - cpumask_set_cpu(cpu, cm); - if (cpumask_weight(cm) == 0) { - cpumask_setall(cm); - for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++) - cpumask_clear_cpu(cpu, cm); - WARN_ON_ONCE(cpumask_weight(cm) == 0); - } - set_cpus_allowed_ptr(rnp->node_kthread_task, cm); - rcu_boost_kthread_setaffinity(rnp, cm); - free_cpumask_var(cm); -} - -/* - * Spawn a per-rcu_node kthread, setting priority and affinity. - * Called during boot before online/offline can happen, or, if - * during runtime, with the main CPU-hotplug locks held. So only - * one of these can be executing at a time. - */ -static int __cpuinit rcu_spawn_one_node_kthread(struct rcu_state *rsp, - struct rcu_node *rnp) -{ - unsigned long flags; - int rnp_index = rnp - &rsp->node[0]; - struct sched_param sp; - struct task_struct *t; - - if (!rcu_kthreads_spawnable || - rnp->qsmaskinit == 0) - return 0; - if (rnp->node_kthread_task == NULL) { - t = kthread_create(rcu_node_kthread, (void *)rnp, - "rcun%d", rnp_index); - if (IS_ERR(t)) - return PTR_ERR(t); - raw_spin_lock_irqsave(&rnp->lock, flags); - set_task_state(t, TASK_INTERRUPTIBLE); - rnp->node_kthread_task = t; - raw_spin_unlock_irqrestore(&rnp->lock, flags); - sp.sched_priority = 99; - sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); - } - return rcu_spawn_one_boost_kthread(rsp, rnp, rnp_index); -} - -/* - * Spawn all kthreads -- called as soon as the scheduler is running. - */ -static int __init rcu_spawn_kthreads(void) -{ - int cpu; - struct rcu_node *rnp; - - rcu_kthreads_spawnable = 1; - for_each_possible_cpu(cpu) { - per_cpu(rcu_cpu_has_work, cpu) = 0; - if (cpu_online(cpu)) - (void)rcu_spawn_one_cpu_kthread(cpu); - } - rnp = rcu_get_root(rcu_state); - (void)rcu_spawn_one_node_kthread(rcu_state, rnp); - if (NUM_RCU_NODES > 1) { - rcu_for_each_leaf_node(rcu_state, rnp) - (void)rcu_spawn_one_node_kthread(rcu_state, rnp); - } - return 0; -} -early_initcall(rcu_spawn_kthreads); - static void __call_rcu(struct rcu_head *head, void (*func)(struct rcu_head *rcu), struct rcu_state *rsp) @@ -2188,26 +1902,13 @@ rcu_init_percpu_data(int cpu, struct rcu_state *rsp, int preemptible) raw_spin_unlock_irqrestore(&rsp->onofflock, flags); } -static void __cpuinit rcu_online_cpu(int cpu) +static void __cpuinit rcu_prepare_cpu(int cpu) { rcu_init_percpu_data(cpu, &rcu_sched_state, 0); rcu_init_percpu_data(cpu, &rcu_bh_state, 0); rcu_preempt_init_percpu_data(cpu); } -static void __cpuinit rcu_online_kthreads(int cpu) -{ - struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, cpu); - struct rcu_node *rnp = rdp->mynode; - - /* Fire up the incoming CPU's kthread and leaf rcu_node kthread. */ - if (rcu_kthreads_spawnable) { - (void)rcu_spawn_one_cpu_kthread(cpu); - if (rnp->node_kthread_task == NULL) - (void)rcu_spawn_one_node_kthread(rcu_state, rnp); - } -} - /* * Handle CPU online/offline notification events. */ @@ -2221,8 +1922,8 @@ static int __cpuinit rcu_cpu_notify(struct notifier_block *self, switch (action) { case CPU_UP_PREPARE: case CPU_UP_PREPARE_FROZEN: - rcu_online_cpu(cpu); - rcu_online_kthreads(cpu); + rcu_prepare_cpu(cpu); + rcu_prepare_kthreads(cpu); break; case CPU_ONLINE: case CPU_DOWN_FAILED: @@ -2372,6 +2073,7 @@ void __init rcu_init(void) rcu_init_one(&rcu_sched_state, &rcu_sched_data); rcu_init_one(&rcu_bh_state, &rcu_bh_data); __rcu_init_preempt(); + open_softirq(RCU_SOFTIRQ, rcu_process_callbacks); /* * We don't need protection against CPU-hotplug here because diff --git a/kernel/rcutree.h b/kernel/rcutree.h index 7b9a08b4aae..01b2ccda26f 100644 --- a/kernel/rcutree.h +++ b/kernel/rcutree.h @@ -369,6 +369,7 @@ struct rcu_state { /* period because */ /* force_quiescent_state() */ /* was running. */ + u8 boost; /* Subject to priority boost. */ unsigned long gpnum; /* Current gp number. */ unsigned long completed; /* # of last completed gp. */ @@ -426,6 +427,7 @@ static int rcu_preempt_blocked_readers_cgp(struct rcu_node *rnp); #ifdef CONFIG_HOTPLUG_CPU static void rcu_report_unblock_qs_rnp(struct rcu_node *rnp, unsigned long flags); +static void rcu_stop_cpu_kthread(int cpu); #endif /* #ifdef CONFIG_HOTPLUG_CPU */ static void rcu_print_detail_task_stall(struct rcu_state *rsp); static void rcu_print_task_stall(struct rcu_node *rnp); @@ -450,11 +452,19 @@ static void rcu_preempt_send_cbs_to_online(void); static void __init __rcu_init_preempt(void); static void rcu_needs_cpu_flush(void); static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags); +static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); +static void invoke_rcu_callbacks_kthread(void); +#ifdef CONFIG_RCU_BOOST +static void rcu_preempt_do_callbacks(void); static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, cpumask_var_t cm); -static void rcu_preempt_boost_start_gp(struct rcu_node *rnp); static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp, struct rcu_node *rnp, int rnp_index); +static void invoke_rcu_node_kthread(struct rcu_node *rnp); +static void rcu_yield(void (*f)(unsigned long), unsigned long arg); +#endif /* #ifdef CONFIG_RCU_BOOST */ +static void rcu_cpu_kthread_setrt(int cpu, int to_rt); +static void __cpuinit rcu_prepare_kthreads(int cpu); #endif /* #ifndef RCU_TREE_NONCORE */ diff --git a/kernel/rcutree_plugin.h b/kernel/rcutree_plugin.h index a767b7dac36..8aafbb80b8b 100644 --- a/kernel/rcutree_plugin.h +++ b/kernel/rcutree_plugin.h @@ -68,6 +68,7 @@ struct rcu_state rcu_preempt_state = RCU_STATE_INITIALIZER(rcu_preempt_state); DEFINE_PER_CPU(struct rcu_data, rcu_preempt_data); static struct rcu_state *rcu_state = &rcu_preempt_state; +static void rcu_read_unlock_special(struct task_struct *t); static int rcu_preempted_readers_exp(struct rcu_node *rnp); /* @@ -147,7 +148,7 @@ static void rcu_preempt_note_context_switch(int cpu) struct rcu_data *rdp; struct rcu_node *rnp; - if (t->rcu_read_lock_nesting && + if (t->rcu_read_lock_nesting > 0 && (t->rcu_read_unlock_special & RCU_READ_UNLOCK_BLOCKED) == 0) { /* Possibly blocking in an RCU read-side critical section. */ @@ -190,6 +191,14 @@ static void rcu_preempt_note_context_switch(int cpu) rnp->gp_tasks = &t->rcu_node_entry; } raw_spin_unlock_irqrestore(&rnp->lock, flags); + } else if (t->rcu_read_lock_nesting < 0 && + t->rcu_read_unlock_special) { + + /* + * Complete exit from RCU read-side critical section on + * behalf of preempted instance of __rcu_read_unlock(). + */ + rcu_read_unlock_special(t); } /* @@ -284,7 +293,7 @@ static struct list_head *rcu_next_node_entry(struct task_struct *t, * notify RCU core processing or task having blocked during the RCU * read-side critical section. */ -static void rcu_read_unlock_special(struct task_struct *t) +static noinline void rcu_read_unlock_special(struct task_struct *t) { int empty; int empty_exp; @@ -309,7 +318,7 @@ static void rcu_read_unlock_special(struct task_struct *t) } /* Hardware IRQ handlers cannot block. */ - if (in_irq()) { + if (in_irq() || in_serving_softirq()) { local_irq_restore(flags); return; } @@ -342,6 +351,11 @@ static void rcu_read_unlock_special(struct task_struct *t) #ifdef CONFIG_RCU_BOOST if (&t->rcu_node_entry == rnp->boost_tasks) rnp->boost_tasks = np; + /* Snapshot and clear ->rcu_boosted with rcu_node lock held. */ + if (t->rcu_boosted) { + special |= RCU_READ_UNLOCK_BOOSTED; + t->rcu_boosted = 0; + } #endif /* #ifdef CONFIG_RCU_BOOST */ t->rcu_blocked_node = NULL; @@ -358,7 +372,6 @@ static void rcu_read_unlock_special(struct task_struct *t) #ifdef CONFIG_RCU_BOOST /* Unboost if we were boosted. */ if (special & RCU_READ_UNLOCK_BOOSTED) { - t->rcu_read_unlock_special &= ~RCU_READ_UNLOCK_BOOSTED; rt_mutex_unlock(t->rcu_boost_mutex); t->rcu_boost_mutex = NULL; } @@ -387,13 +400,22 @@ void __rcu_read_unlock(void) struct task_struct *t = current; barrier(); /* needed if we ever invoke rcu_read_unlock in rcutree.c */ - --t->rcu_read_lock_nesting; - barrier(); /* decrement before load of ->rcu_read_unlock_special */ - if (t->rcu_read_lock_nesting == 0 && - unlikely(ACCESS_ONCE(t->rcu_read_unlock_special))) - rcu_read_unlock_special(t); + if (t->rcu_read_lock_nesting != 1) + --t->rcu_read_lock_nesting; + else { + t->rcu_read_lock_nesting = INT_MIN; + barrier(); /* assign before ->rcu_read_unlock_special load */ + if (unlikely(ACCESS_ONCE(t->rcu_read_unlock_special))) + rcu_read_unlock_special(t); + barrier(); /* ->rcu_read_unlock_special load before assign */ + t->rcu_read_lock_nesting = 0; + } #ifdef CONFIG_PROVE_LOCKING - WARN_ON_ONCE(ACCESS_ONCE(t->rcu_read_lock_nesting) < 0); + { + int rrln = ACCESS_ONCE(t->rcu_read_lock_nesting); + + WARN_ON_ONCE(rrln < 0 && rrln > INT_MIN / 2); + } #endif /* #ifdef CONFIG_PROVE_LOCKING */ } EXPORT_SYMBOL_GPL(__rcu_read_unlock); @@ -589,7 +611,8 @@ static void rcu_preempt_check_callbacks(int cpu) rcu_preempt_qs(cpu); return; } - if (per_cpu(rcu_preempt_data, cpu).qs_pending) + if (t->rcu_read_lock_nesting > 0 && + per_cpu(rcu_preempt_data, cpu).qs_pending) t->rcu_read_unlock_special |= RCU_READ_UNLOCK_NEED_QS; } @@ -602,6 +625,15 @@ static void rcu_preempt_process_callbacks(void) &__get_cpu_var(rcu_preempt_data)); } +#ifdef CONFIG_RCU_BOOST + +static void rcu_preempt_do_callbacks(void) +{ + rcu_do_batch(&rcu_preempt_state, &__get_cpu_var(rcu_preempt_data)); +} + +#endif /* #ifdef CONFIG_RCU_BOOST */ + /* * Queue a preemptible-RCU callback for invocation after a grace period. */ @@ -686,9 +718,12 @@ static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp) raw_spin_lock_irqsave(&rnp->lock, flags); for (;;) { - if (!sync_rcu_preempt_exp_done(rnp)) + if (!sync_rcu_preempt_exp_done(rnp)) { + raw_spin_unlock_irqrestore(&rnp->lock, flags); break; + } if (rnp->parent == NULL) { + raw_spin_unlock_irqrestore(&rnp->lock, flags); wake_up(&sync_rcu_preempt_exp_wq); break; } @@ -698,7 +733,6 @@ static void rcu_report_exp_rnp(struct rcu_state *rsp, struct rcu_node *rnp) raw_spin_lock(&rnp->lock); /* irqs already disabled */ rnp->expmask &= ~mask; } - raw_spin_unlock_irqrestore(&rnp->lock, flags); } /* @@ -1165,7 +1199,7 @@ static int rcu_boost(struct rcu_node *rnp) t = container_of(tb, struct task_struct, rcu_node_entry); rt_mutex_init_proxy_locked(&mtx, t); t->rcu_boost_mutex = &mtx; - t->rcu_read_unlock_special |= RCU_READ_UNLOCK_BOOSTED; + t->rcu_boosted = 1; raw_spin_unlock_irqrestore(&rnp->lock, flags); rt_mutex_lock(&mtx); /* Side effect: boosts task t's priority. */ rt_mutex_unlock(&mtx); /* Keep lockdep happy. */ @@ -1249,6 +1283,23 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) } /* + * Wake up the per-CPU kthread to invoke RCU callbacks. + */ +static void invoke_rcu_callbacks_kthread(void) +{ + unsigned long flags; + + local_irq_save(flags); + __this_cpu_write(rcu_cpu_has_work, 1); + if (__this_cpu_read(rcu_cpu_kthread_task) == NULL) { + local_irq_restore(flags); + return; + } + wake_up_process(__this_cpu_read(rcu_cpu_kthread_task)); + local_irq_restore(flags); +} + +/* * Set the affinity of the boost kthread. The CPU-hotplug locks are * held, so no one should be messing with the existence of the boost * kthread. @@ -1288,6 +1339,7 @@ static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp, if (&rcu_preempt_state != rsp) return 0; + rsp->boost = 1; if (rnp->boost_kthread_task != NULL) return 0; t = kthread_create(rcu_boost_kthread, (void *)rnp, @@ -1295,14 +1347,378 @@ static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp, if (IS_ERR(t)) return PTR_ERR(t); raw_spin_lock_irqsave(&rnp->lock, flags); - set_task_state(t, TASK_INTERRUPTIBLE); rnp->boost_kthread_task = t; raw_spin_unlock_irqrestore(&rnp->lock, flags); sp.sched_priority = RCU_KTHREAD_PRIO; sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); + wake_up_process(t); /* get to TASK_INTERRUPTIBLE quickly. */ + return 0; +} + +#ifdef CONFIG_HOTPLUG_CPU + +/* + * Stop the RCU's per-CPU kthread when its CPU goes offline,. + */ +static void rcu_stop_cpu_kthread(int cpu) +{ + struct task_struct *t; + + /* Stop the CPU's kthread. */ + t = per_cpu(rcu_cpu_kthread_task, cpu); + if (t != NULL) { + per_cpu(rcu_cpu_kthread_task, cpu) = NULL; + kthread_stop(t); + } +} + +#endif /* #ifdef CONFIG_HOTPLUG_CPU */ + +static void rcu_kthread_do_work(void) +{ + rcu_do_batch(&rcu_sched_state, &__get_cpu_var(rcu_sched_data)); + rcu_do_batch(&rcu_bh_state, &__get_cpu_var(rcu_bh_data)); + rcu_preempt_do_callbacks(); +} + +/* + * Wake up the specified per-rcu_node-structure kthread. + * Because the per-rcu_node kthreads are immortal, we don't need + * to do anything to keep them alive. + */ +static void invoke_rcu_node_kthread(struct rcu_node *rnp) +{ + struct task_struct *t; + + t = rnp->node_kthread_task; + if (t != NULL) + wake_up_process(t); +} + +/* + * Set the specified CPU's kthread to run RT or not, as specified by + * the to_rt argument. The CPU-hotplug locks are held, so the task + * is not going away. + */ +static void rcu_cpu_kthread_setrt(int cpu, int to_rt) +{ + int policy; + struct sched_param sp; + struct task_struct *t; + + t = per_cpu(rcu_cpu_kthread_task, cpu); + if (t == NULL) + return; + if (to_rt) { + policy = SCHED_FIFO; + sp.sched_priority = RCU_KTHREAD_PRIO; + } else { + policy = SCHED_NORMAL; + sp.sched_priority = 0; + } + sched_setscheduler_nocheck(t, policy, &sp); +} + +/* + * Timer handler to initiate the waking up of per-CPU kthreads that + * have yielded the CPU due to excess numbers of RCU callbacks. + * We wake up the per-rcu_node kthread, which in turn will wake up + * the booster kthread. + */ +static void rcu_cpu_kthread_timer(unsigned long arg) +{ + struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, arg); + struct rcu_node *rnp = rdp->mynode; + + atomic_or(rdp->grpmask, &rnp->wakemask); + invoke_rcu_node_kthread(rnp); +} + +/* + * Drop to non-real-time priority and yield, but only after posting a + * timer that will cause us to regain our real-time priority if we + * remain preempted. Either way, we restore our real-time priority + * before returning. + */ +static void rcu_yield(void (*f)(unsigned long), unsigned long arg) +{ + struct sched_param sp; + struct timer_list yield_timer; + + setup_timer_on_stack(&yield_timer, f, arg); + mod_timer(&yield_timer, jiffies + 2); + sp.sched_priority = 0; + sched_setscheduler_nocheck(current, SCHED_NORMAL, &sp); + set_user_nice(current, 19); + schedule(); + sp.sched_priority = RCU_KTHREAD_PRIO; + sched_setscheduler_nocheck(current, SCHED_FIFO, &sp); + del_timer(&yield_timer); +} + +/* + * Handle cases where the rcu_cpu_kthread() ends up on the wrong CPU. + * This can happen while the corresponding CPU is either coming online + * or going offline. We cannot wait until the CPU is fully online + * before starting the kthread, because the various notifier functions + * can wait for RCU grace periods. So we park rcu_cpu_kthread() until + * the corresponding CPU is online. + * + * Return 1 if the kthread needs to stop, 0 otherwise. + * + * Caller must disable bh. This function can momentarily enable it. + */ +static int rcu_cpu_kthread_should_stop(int cpu) +{ + while (cpu_is_offline(cpu) || + !cpumask_equal(¤t->cpus_allowed, cpumask_of(cpu)) || + smp_processor_id() != cpu) { + if (kthread_should_stop()) + return 1; + per_cpu(rcu_cpu_kthread_status, cpu) = RCU_KTHREAD_OFFCPU; + per_cpu(rcu_cpu_kthread_cpu, cpu) = raw_smp_processor_id(); + local_bh_enable(); + schedule_timeout_uninterruptible(1); + if (!cpumask_equal(¤t->cpus_allowed, cpumask_of(cpu))) + set_cpus_allowed_ptr(current, cpumask_of(cpu)); + local_bh_disable(); + } + per_cpu(rcu_cpu_kthread_cpu, cpu) = cpu; return 0; } +/* + * Per-CPU kernel thread that invokes RCU callbacks. This replaces the + * earlier RCU softirq. + */ +static int rcu_cpu_kthread(void *arg) +{ + int cpu = (int)(long)arg; + unsigned long flags; + int spincnt = 0; + unsigned int *statusp = &per_cpu(rcu_cpu_kthread_status, cpu); + char work; + char *workp = &per_cpu(rcu_cpu_has_work, cpu); + + for (;;) { + *statusp = RCU_KTHREAD_WAITING; + rcu_wait(*workp != 0 || kthread_should_stop()); + local_bh_disable(); + if (rcu_cpu_kthread_should_stop(cpu)) { + local_bh_enable(); + break; + } + *statusp = RCU_KTHREAD_RUNNING; + per_cpu(rcu_cpu_kthread_loops, cpu)++; + local_irq_save(flags); + work = *workp; + *workp = 0; + local_irq_restore(flags); + if (work) + rcu_kthread_do_work(); + local_bh_enable(); + if (*workp != 0) + spincnt++; + else + spincnt = 0; + if (spincnt > 10) { + *statusp = RCU_KTHREAD_YIELDING; + rcu_yield(rcu_cpu_kthread_timer, (unsigned long)cpu); + spincnt = 0; + } + } + *statusp = RCU_KTHREAD_STOPPED; + return 0; +} + +/* + * Spawn a per-CPU kthread, setting up affinity and priority. + * Because the CPU hotplug lock is held, no other CPU will be attempting + * to manipulate rcu_cpu_kthread_task. There might be another CPU + * attempting to access it during boot, but the locking in kthread_bind() + * will enforce sufficient ordering. + * + * Please note that we cannot simply refuse to wake up the per-CPU + * kthread because kthreads are created in TASK_UNINTERRUPTIBLE state, + * which can result in softlockup complaints if the task ends up being + * idle for more than a couple of minutes. + * + * However, please note also that we cannot bind the per-CPU kthread to its + * CPU until that CPU is fully online. We also cannot wait until the + * CPU is fully online before we create its per-CPU kthread, as this would + * deadlock the system when CPU notifiers tried waiting for grace + * periods. So we bind the per-CPU kthread to its CPU only if the CPU + * is online. If its CPU is not yet fully online, then the code in + * rcu_cpu_kthread() will wait until it is fully online, and then do + * the binding. + */ +static int __cpuinit rcu_spawn_one_cpu_kthread(int cpu) +{ + struct sched_param sp; + struct task_struct *t; + + if (!rcu_scheduler_fully_active || + per_cpu(rcu_cpu_kthread_task, cpu) != NULL) + return 0; + t = kthread_create(rcu_cpu_kthread, (void *)(long)cpu, "rcuc%d", cpu); + if (IS_ERR(t)) + return PTR_ERR(t); + if (cpu_online(cpu)) + kthread_bind(t, cpu); + per_cpu(rcu_cpu_kthread_cpu, cpu) = cpu; + WARN_ON_ONCE(per_cpu(rcu_cpu_kthread_task, cpu) != NULL); + sp.sched_priority = RCU_KTHREAD_PRIO; + sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); + per_cpu(rcu_cpu_kthread_task, cpu) = t; + wake_up_process(t); /* Get to TASK_INTERRUPTIBLE quickly. */ + return 0; +} + +/* + * Per-rcu_node kthread, which is in charge of waking up the per-CPU + * kthreads when needed. We ignore requests to wake up kthreads + * for offline CPUs, which is OK because force_quiescent_state() + * takes care of this case. + */ +static int rcu_node_kthread(void *arg) +{ + int cpu; + unsigned long flags; + unsigned long mask; + struct rcu_node *rnp = (struct rcu_node *)arg; + struct sched_param sp; + struct task_struct *t; + + for (;;) { + rnp->node_kthread_status = RCU_KTHREAD_WAITING; + rcu_wait(atomic_read(&rnp->wakemask) != 0); + rnp->node_kthread_status = RCU_KTHREAD_RUNNING; + raw_spin_lock_irqsave(&rnp->lock, flags); + mask = atomic_xchg(&rnp->wakemask, 0); + rcu_initiate_boost(rnp, flags); /* releases rnp->lock. */ + for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask >>= 1) { + if ((mask & 0x1) == 0) + continue; + preempt_disable(); + t = per_cpu(rcu_cpu_kthread_task, cpu); + if (!cpu_online(cpu) || t == NULL) { + preempt_enable(); + continue; + } + per_cpu(rcu_cpu_has_work, cpu) = 1; + sp.sched_priority = RCU_KTHREAD_PRIO; + sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); + preempt_enable(); + } + } + /* NOTREACHED */ + rnp->node_kthread_status = RCU_KTHREAD_STOPPED; + return 0; +} + +/* + * Set the per-rcu_node kthread's affinity to cover all CPUs that are + * served by the rcu_node in question. The CPU hotplug lock is still + * held, so the value of rnp->qsmaskinit will be stable. + * + * We don't include outgoingcpu in the affinity set, use -1 if there is + * no outgoing CPU. If there are no CPUs left in the affinity set, + * this function allows the kthread to execute on any CPU. + */ +static void rcu_node_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) +{ + cpumask_var_t cm; + int cpu; + unsigned long mask = rnp->qsmaskinit; + + if (rnp->node_kthread_task == NULL) + return; + if (!alloc_cpumask_var(&cm, GFP_KERNEL)) + return; + cpumask_clear(cm); + for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++, mask >>= 1) + if ((mask & 0x1) && cpu != outgoingcpu) + cpumask_set_cpu(cpu, cm); + if (cpumask_weight(cm) == 0) { + cpumask_setall(cm); + for (cpu = rnp->grplo; cpu <= rnp->grphi; cpu++) + cpumask_clear_cpu(cpu, cm); + WARN_ON_ONCE(cpumask_weight(cm) == 0); + } + set_cpus_allowed_ptr(rnp->node_kthread_task, cm); + rcu_boost_kthread_setaffinity(rnp, cm); + free_cpumask_var(cm); +} + +/* + * Spawn a per-rcu_node kthread, setting priority and affinity. + * Called during boot before online/offline can happen, or, if + * during runtime, with the main CPU-hotplug locks held. So only + * one of these can be executing at a time. + */ +static int __cpuinit rcu_spawn_one_node_kthread(struct rcu_state *rsp, + struct rcu_node *rnp) +{ + unsigned long flags; + int rnp_index = rnp - &rsp->node[0]; + struct sched_param sp; + struct task_struct *t; + + if (!rcu_scheduler_fully_active || + rnp->qsmaskinit == 0) + return 0; + if (rnp->node_kthread_task == NULL) { + t = kthread_create(rcu_node_kthread, (void *)rnp, + "rcun%d", rnp_index); + if (IS_ERR(t)) + return PTR_ERR(t); + raw_spin_lock_irqsave(&rnp->lock, flags); + rnp->node_kthread_task = t; + raw_spin_unlock_irqrestore(&rnp->lock, flags); + sp.sched_priority = 99; + sched_setscheduler_nocheck(t, SCHED_FIFO, &sp); + wake_up_process(t); /* get to TASK_INTERRUPTIBLE quickly. */ + } + return rcu_spawn_one_boost_kthread(rsp, rnp, rnp_index); +} + +/* + * Spawn all kthreads -- called as soon as the scheduler is running. + */ +static int __init rcu_spawn_kthreads(void) +{ + int cpu; + struct rcu_node *rnp; + + rcu_scheduler_fully_active = 1; + for_each_possible_cpu(cpu) { + per_cpu(rcu_cpu_has_work, cpu) = 0; + if (cpu_online(cpu)) + (void)rcu_spawn_one_cpu_kthread(cpu); + } + rnp = rcu_get_root(rcu_state); + (void)rcu_spawn_one_node_kthread(rcu_state, rnp); + if (NUM_RCU_NODES > 1) { + rcu_for_each_leaf_node(rcu_state, rnp) + (void)rcu_spawn_one_node_kthread(rcu_state, rnp); + } + return 0; +} +early_initcall(rcu_spawn_kthreads); + +static void __cpuinit rcu_prepare_kthreads(int cpu) +{ + struct rcu_data *rdp = per_cpu_ptr(rcu_state->rda, cpu); + struct rcu_node *rnp = rdp->mynode; + + /* Fire up the incoming CPU's kthread and leaf rcu_node kthread. */ + if (rcu_scheduler_fully_active) { + (void)rcu_spawn_one_cpu_kthread(cpu); + if (rnp->node_kthread_task == NULL) + (void)rcu_spawn_one_node_kthread(rcu_state, rnp); + } +} + #else /* #ifdef CONFIG_RCU_BOOST */ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) @@ -1310,21 +1726,41 @@ static void rcu_initiate_boost(struct rcu_node *rnp, unsigned long flags) raw_spin_unlock_irqrestore(&rnp->lock, flags); } -static void rcu_boost_kthread_setaffinity(struct rcu_node *rnp, - cpumask_var_t cm) +static void invoke_rcu_callbacks_kthread(void) { + WARN_ON_ONCE(1); } static void rcu_preempt_boost_start_gp(struct rcu_node *rnp) { } -static int __cpuinit rcu_spawn_one_boost_kthread(struct rcu_state *rsp, - struct rcu_node *rnp, - int rnp_index) +#ifdef CONFIG_HOTPLUG_CPU + +static void rcu_stop_cpu_kthread(int cpu) +{ +} + +#endif /* #ifdef CONFIG_HOTPLUG_CPU */ + +static void rcu_node_kthread_setaffinity(struct rcu_node *rnp, int outgoingcpu) +{ +} + +static void rcu_cpu_kthread_setrt(int cpu, int to_rt) { +} + +static int __init rcu_scheduler_really_started(void) +{ + rcu_scheduler_fully_active = 1; return 0; } +early_initcall(rcu_scheduler_really_started); + +static void __cpuinit rcu_prepare_kthreads(int cpu) +{ +} #endif /* #else #ifdef CONFIG_RCU_BOOST */ @@ -1500,7 +1936,7 @@ static DEFINE_PER_CPU(unsigned long, rcu_dyntick_holdoff); * * Because it is not legal to invoke rcu_process_callbacks() with irqs * disabled, we do one pass of force_quiescent_state(), then do a - * invoke_rcu_cpu_kthread() to cause rcu_process_callbacks() to be invoked + * invoke_rcu_core() to cause rcu_process_callbacks() to be invoked * later. The per-cpu rcu_dyntick_drain variable controls the sequencing. */ int rcu_needs_cpu(int cpu) @@ -1551,7 +1987,7 @@ int rcu_needs_cpu(int cpu) /* If RCU callbacks are still pending, RCU still needs this CPU. */ if (c) - invoke_rcu_cpu_kthread(); + invoke_rcu_core(); return c; } diff --git a/kernel/rcutree_trace.c b/kernel/rcutree_trace.c index 9678cc3650f..4e144876dc6 100644 --- a/kernel/rcutree_trace.c +++ b/kernel/rcutree_trace.c @@ -46,6 +46,8 @@ #define RCU_TREE_NONCORE #include "rcutree.h" +#ifdef CONFIG_RCU_BOOST + DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_status); DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_cpu); DECLARE_PER_CPU(unsigned int, rcu_cpu_kthread_loops); @@ -58,6 +60,8 @@ static char convert_kthread_status(unsigned int kthread_status) return "SRWOY"[kthread_status]; } +#endif /* #ifdef CONFIG_RCU_BOOST */ + static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) { if (!rdp->beenonline) @@ -76,7 +80,7 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) rdp->dynticks_fqs); #endif /* #ifdef CONFIG_NO_HZ */ seq_printf(m, " of=%lu ri=%lu", rdp->offline_fqs, rdp->resched_ipi); - seq_printf(m, " ql=%ld qs=%c%c%c%c kt=%d/%c/%d ktl=%x b=%ld", + seq_printf(m, " ql=%ld qs=%c%c%c%c", rdp->qlen, ".N"[rdp->nxttail[RCU_NEXT_READY_TAIL] != rdp->nxttail[RCU_NEXT_TAIL]], @@ -84,13 +88,16 @@ static void print_one_rcu_data(struct seq_file *m, struct rcu_data *rdp) rdp->nxttail[RCU_NEXT_READY_TAIL]], ".W"[rdp->nxttail[RCU_DONE_TAIL] != rdp->nxttail[RCU_WAIT_TAIL]], - ".D"[&rdp->nxtlist != rdp->nxttail[RCU_DONE_TAIL]], + ".D"[&rdp->nxtlist != rdp->nxttail[RCU_DONE_TAIL]]); +#ifdef CONFIG_RCU_BOOST + seq_printf(m, " kt=%d/%c/%d ktl=%x", per_cpu(rcu_cpu_has_work, rdp->cpu), convert_kthread_status(per_cpu(rcu_cpu_kthread_status, rdp->cpu)), per_cpu(rcu_cpu_kthread_cpu, rdp->cpu), - per_cpu(rcu_cpu_kthread_loops, rdp->cpu) & 0xffff, - rdp->blimit); + per_cpu(rcu_cpu_kthread_loops, rdp->cpu) & 0xffff); +#endif /* #ifdef CONFIG_RCU_BOOST */ + seq_printf(m, " b=%ld", rdp->blimit); seq_printf(m, " ci=%lu co=%lu ca=%lu\n", rdp->n_cbs_invoked, rdp->n_cbs_orphaned, rdp->n_cbs_adopted); } @@ -147,18 +154,21 @@ static void print_one_rcu_data_csv(struct seq_file *m, struct rcu_data *rdp) rdp->dynticks_fqs); #endif /* #ifdef CONFIG_NO_HZ */ seq_printf(m, ",%lu,%lu", rdp->offline_fqs, rdp->resched_ipi); - seq_printf(m, ",%ld,\"%c%c%c%c\",%d,\"%c\",%ld", rdp->qlen, + seq_printf(m, ",%ld,\"%c%c%c%c\"", rdp->qlen, ".N"[rdp->nxttail[RCU_NEXT_READY_TAIL] != rdp->nxttail[RCU_NEXT_TAIL]], ".R"[rdp->nxttail[RCU_WAIT_TAIL] != rdp->nxttail[RCU_NEXT_READY_TAIL]], ".W"[rdp->nxttail[RCU_DONE_TAIL] != rdp->nxttail[RCU_WAIT_TAIL]], - ".D"[&rdp->nxtlist != rdp->nxttail[RCU_DONE_TAIL]], + ".D"[&rdp->nxtlist != rdp->nxttail[RCU_DONE_TAIL]]); +#ifdef CONFIG_RCU_BOOST + seq_printf(m, ",%d,\"%c\"", per_cpu(rcu_cpu_has_work, rdp->cpu), convert_kthread_status(per_cpu(rcu_cpu_kthread_status, - rdp->cpu)), - rdp->blimit); + rdp->cpu))); +#endif /* #ifdef CONFIG_RCU_BOOST */ + seq_printf(m, ",%ld", rdp->blimit); seq_printf(m, ",%lu,%lu,%lu\n", rdp->n_cbs_invoked, rdp->n_cbs_orphaned, rdp->n_cbs_adopted); } @@ -169,7 +179,11 @@ static int show_rcudata_csv(struct seq_file *m, void *unused) #ifdef CONFIG_NO_HZ seq_puts(m, "\"dt\",\"dt nesting\",\"dt NMI nesting\",\"df\","); #endif /* #ifdef CONFIG_NO_HZ */ - seq_puts(m, "\"of\",\"ri\",\"ql\",\"b\",\"ci\",\"co\",\"ca\"\n"); + seq_puts(m, "\"of\",\"ri\",\"ql\",\"qs\""); +#ifdef CONFIG_RCU_BOOST + seq_puts(m, "\"kt\",\"ktl\""); +#endif /* #ifdef CONFIG_RCU_BOOST */ + seq_puts(m, ",\"b\",\"ci\",\"co\",\"ca\"\n"); #ifdef CONFIG_TREE_PREEMPT_RCU seq_puts(m, "\"rcu_preempt:\"\n"); PRINT_RCU_DATA(rcu_preempt_data, print_one_rcu_data_csv, m); diff --git a/kernel/resource.c b/kernel/resource.c index 798e2fae2a0..3ff40178dce 100644 --- a/kernel/resource.c +++ b/kernel/resource.c @@ -38,6 +38,14 @@ struct resource iomem_resource = { }; EXPORT_SYMBOL(iomem_resource); +/* constraints to be met while allocating resources */ +struct resource_constraint { + resource_size_t min, max, align; + resource_size_t (*alignf)(void *, const struct resource *, + resource_size_t, resource_size_t); + void *alignf_data; +}; + static DEFINE_RWLOCK(resource_lock); static void *r_next(struct seq_file *m, void *v, loff_t *pos) @@ -384,16 +392,13 @@ static bool resource_contains(struct resource *res1, struct resource *res2) } /* - * Find empty slot in the resource tree given range and alignment. + * Find empty slot in the resource tree with the given range and + * alignment constraints */ -static int find_resource(struct resource *root, struct resource *new, - resource_size_t size, resource_size_t min, - resource_size_t max, resource_size_t align, - resource_size_t (*alignf)(void *, - const struct resource *, - resource_size_t, - resource_size_t), - void *alignf_data) +static int __find_resource(struct resource *root, struct resource *old, + struct resource *new, + resource_size_t size, + struct resource_constraint *constraint) { struct resource *this = root->child; struct resource tmp = *new, avail, alloc; @@ -404,25 +409,26 @@ static int find_resource(struct resource *root, struct resource *new, * Skip past an allocated resource that starts at 0, since the assignment * of this->start - 1 to tmp->end below would cause an underflow. */ - if (this && this->start == 0) { - tmp.start = this->end + 1; + if (this && this->start == root->start) { + tmp.start = (this == old) ? old->start : this->end + 1; this = this->sibling; } for(;;) { if (this) - tmp.end = this->start - 1; + tmp.end = (this == old) ? this->end : this->start - 1; else tmp.end = root->end; - resource_clip(&tmp, min, max); + resource_clip(&tmp, constraint->min, constraint->max); arch_remove_reservations(&tmp); /* Check for overflow after ALIGN() */ avail = *new; - avail.start = ALIGN(tmp.start, align); + avail.start = ALIGN(tmp.start, constraint->align); avail.end = tmp.end; if (avail.start >= tmp.start) { - alloc.start = alignf(alignf_data, &avail, size, align); + alloc.start = constraint->alignf(constraint->alignf_data, &avail, + size, constraint->align); alloc.end = alloc.start + size - 1; if (resource_contains(&avail, &alloc)) { new->start = alloc.start; @@ -432,14 +438,75 @@ static int find_resource(struct resource *root, struct resource *new, } if (!this) break; - tmp.start = this->end + 1; + if (this != old) + tmp.start = this->end + 1; this = this->sibling; } return -EBUSY; } +/* + * Find empty slot in the resource tree given range and alignment. + */ +static int find_resource(struct resource *root, struct resource *new, + resource_size_t size, + struct resource_constraint *constraint) +{ + return __find_resource(root, NULL, new, size, constraint); +} + /** - * allocate_resource - allocate empty slot in the resource tree given range & alignment + * reallocate_resource - allocate a slot in the resource tree given range & alignment. + * The resource will be relocated if the new size cannot be reallocated in the + * current location. + * + * @root: root resource descriptor + * @old: resource descriptor desired by caller + * @newsize: new size of the resource descriptor + * @constraint: the size and alignment constraints to be met. + */ +int reallocate_resource(struct resource *root, struct resource *old, + resource_size_t newsize, + struct resource_constraint *constraint) +{ + int err=0; + struct resource new = *old; + struct resource *conflict; + + write_lock(&resource_lock); + + if ((err = __find_resource(root, old, &new, newsize, constraint))) + goto out; + + if (resource_contains(&new, old)) { + old->start = new.start; + old->end = new.end; + goto out; + } + + if (old->child) { + err = -EBUSY; + goto out; + } + + if (resource_contains(old, &new)) { + old->start = new.start; + old->end = new.end; + } else { + __release_resource(old); + *old = new; + conflict = __request_resource(root, old); + BUG_ON(conflict); + } +out: + write_unlock(&resource_lock); + return err; +} + + +/** + * allocate_resource - allocate empty slot in the resource tree given range & alignment. + * The resource will be reallocated with a new size if it was already allocated * @root: root resource descriptor * @new: resource descriptor desired by caller * @size: requested resource region size @@ -459,12 +526,25 @@ int allocate_resource(struct resource *root, struct resource *new, void *alignf_data) { int err; + struct resource_constraint constraint; if (!alignf) alignf = simple_align_resource; + constraint.min = min; + constraint.max = max; + constraint.align = align; + constraint.alignf = alignf; + constraint.alignf_data = alignf_data; + + if ( new->parent ) { + /* resource is already allocated, try reallocating with + the new constraints */ + return reallocate_resource(root, new, size, &constraint); + } + write_lock(&resource_lock); - err = find_resource(root, new, size, min, max, align, alignf, alignf_data); + err = find_resource(root, new, size, &constraint); if (err >= 0 && __request_resource(root, new)) err = -EBUSY; write_unlock(&resource_lock); diff --git a/kernel/rtmutex.c b/kernel/rtmutex.c index ab449117aaf..255e1662acd 100644 --- a/kernel/rtmutex.c +++ b/kernel/rtmutex.c @@ -890,7 +890,7 @@ void __rt_mutex_init(struct rt_mutex *lock, const char *name) { lock->owner = NULL; raw_spin_lock_init(&lock->wait_lock); - plist_head_init_raw(&lock->wait_list, &lock->wait_lock); + plist_head_init(&lock->wait_list); debug_rt_mutex_init(lock, name); } diff --git a/kernel/rwsem.c b/kernel/rwsem.c index cae050b05f5..176e5e56ffa 100644 --- a/kernel/rwsem.c +++ b/kernel/rwsem.c @@ -117,15 +117,6 @@ void down_read_nested(struct rw_semaphore *sem, int subclass) EXPORT_SYMBOL(down_read_nested); -void down_read_non_owner(struct rw_semaphore *sem) -{ - might_sleep(); - - __down_read(sem); -} - -EXPORT_SYMBOL(down_read_non_owner); - void down_write_nested(struct rw_semaphore *sem, int subclass) { might_sleep(); @@ -136,13 +127,6 @@ void down_write_nested(struct rw_semaphore *sem, int subclass) EXPORT_SYMBOL(down_write_nested); -void up_read_non_owner(struct rw_semaphore *sem) -{ - __up_read(sem); -} - -EXPORT_SYMBOL(up_read_non_owner); - #endif diff --git a/kernel/sched.c b/kernel/sched.c index cbb3a0eee58..ccacdbdecf4 100644 --- a/kernel/sched.c +++ b/kernel/sched.c @@ -75,6 +75,9 @@ #include <asm/tlb.h> #include <asm/irq_regs.h> #include <asm/mutex.h> +#ifdef CONFIG_PARAVIRT +#include <asm/paravirt.h> +#endif #include "sched_cpupri.h" #include "workqueue_sched.h" @@ -124,7 +127,7 @@ static inline int rt_policy(int policy) { - if (unlikely(policy == SCHED_FIFO || policy == SCHED_RR)) + if (policy == SCHED_FIFO || policy == SCHED_RR) return 1; return 0; } @@ -292,8 +295,8 @@ static DEFINE_SPINLOCK(task_group_lock); * (The default weight is 1024 - so there's no practical * limitation from this.) */ -#define MIN_SHARES 2 -#define MAX_SHARES (1UL << (18 + SCHED_LOAD_RESOLUTION)) +#define MIN_SHARES (1UL << 1) +#define MAX_SHARES (1UL << 18) static int root_task_group_load = ROOT_TASK_GROUP_LOAD; #endif @@ -422,6 +425,7 @@ struct rt_rq { */ struct root_domain { atomic_t refcount; + atomic_t rto_count; struct rcu_head rcu; cpumask_var_t span; cpumask_var_t online; @@ -431,7 +435,6 @@ struct root_domain { * one runnable RT task. */ cpumask_var_t rto_mask; - atomic_t rto_count; struct cpupri cpupri; }; @@ -528,6 +531,12 @@ struct rq { #ifdef CONFIG_IRQ_TIME_ACCOUNTING u64 prev_irq_time; #endif +#ifdef CONFIG_PARAVIRT + u64 prev_steal_time; +#endif +#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING + u64 prev_steal_time_rq; +#endif /* calc_load related fields */ unsigned long calc_load_update; @@ -581,7 +590,6 @@ static inline int cpu_of(struct rq *rq) #define rcu_dereference_check_sched_domain(p) \ rcu_dereference_check((p), \ - rcu_read_lock_held() || \ lockdep_is_held(&sched_domains_mutex)) /* @@ -605,10 +613,10 @@ static inline int cpu_of(struct rq *rq) /* * Return the group to which this tasks belongs. * - * We use task_subsys_state_check() and extend the RCU verification - * with lockdep_is_held(&p->pi_lock) because cpu_cgroup_attach() - * holds that lock for each task it moves into the cgroup. Therefore - * by holding that lock, we pin the task to the current cgroup. + * We use task_subsys_state_check() and extend the RCU verification with + * pi->lock and rq->lock because cpu_cgroup_attach() holds those locks for each + * task it moves into the cgroup. Therefore by holding either of those locks, + * we pin the task to the current cgroup. */ static inline struct task_group *task_group(struct task_struct *p) { @@ -616,7 +624,8 @@ static inline struct task_group *task_group(struct task_struct *p) struct cgroup_subsys_state *css; css = task_subsys_state_check(p, cpu_cgroup_subsys_id, - lockdep_is_held(&p->pi_lock)); + lockdep_is_held(&p->pi_lock) || + lockdep_is_held(&task_rq(p)->lock)); tg = container_of(css, struct task_group, css); return autogroup_task_group(p, tg); @@ -1567,38 +1576,6 @@ static unsigned long cpu_avg_load_per_task(int cpu) return rq->avg_load_per_task; } -#ifdef CONFIG_FAIR_GROUP_SCHED - -/* - * Compute the cpu's hierarchical load factor for each task group. - * This needs to be done in a top-down fashion because the load of a child - * group is a fraction of its parents load. - */ -static int tg_load_down(struct task_group *tg, void *data) -{ - unsigned long load; - long cpu = (long)data; - - if (!tg->parent) { - load = cpu_rq(cpu)->load.weight; - } else { - load = tg->parent->cfs_rq[cpu]->h_load; - load *= tg->se[cpu]->load.weight; - load /= tg->parent->cfs_rq[cpu]->load.weight + 1; - } - - tg->cfs_rq[cpu]->h_load = load; - - return 0; -} - -static void update_h_load(long cpu) -{ - walk_tg_tree(tg_load_down, tg_nop, (void *)cpu); -} - -#endif - #ifdef CONFIG_PREEMPT static void double_rq_lock(struct rq *rq1, struct rq *rq2); @@ -1952,10 +1929,28 @@ void account_system_vtime(struct task_struct *curr) } EXPORT_SYMBOL_GPL(account_system_vtime); -static void update_rq_clock_task(struct rq *rq, s64 delta) +#endif /* CONFIG_IRQ_TIME_ACCOUNTING */ + +#ifdef CONFIG_PARAVIRT +static inline u64 steal_ticks(u64 steal) { - s64 irq_delta; + if (unlikely(steal > NSEC_PER_SEC)) + return div_u64(steal, TICK_NSEC); + return __iter_div_u64_rem(steal, TICK_NSEC, &steal); +} +#endif + +static void update_rq_clock_task(struct rq *rq, s64 delta) +{ +/* + * In theory, the compile should just see 0 here, and optimize out the call + * to sched_rt_avg_update. But I don't trust it... + */ +#if defined(CONFIG_IRQ_TIME_ACCOUNTING) || defined(CONFIG_PARAVIRT_TIME_ACCOUNTING) + s64 steal = 0, irq_delta = 0; +#endif +#ifdef CONFIG_IRQ_TIME_ACCOUNTING irq_delta = irq_time_read(cpu_of(rq)) - rq->prev_irq_time; /* @@ -1978,12 +1973,35 @@ static void update_rq_clock_task(struct rq *rq, s64 delta) rq->prev_irq_time += irq_delta; delta -= irq_delta; +#endif +#ifdef CONFIG_PARAVIRT_TIME_ACCOUNTING + if (static_branch((¶virt_steal_rq_enabled))) { + u64 st; + + steal = paravirt_steal_clock(cpu_of(rq)); + steal -= rq->prev_steal_time_rq; + + if (unlikely(steal > delta)) + steal = delta; + + st = steal_ticks(steal); + steal = st * TICK_NSEC; + + rq->prev_steal_time_rq += steal; + + delta -= steal; + } +#endif + rq->clock_task += delta; - if (irq_delta && sched_feat(NONIRQ_POWER)) - sched_rt_avg_update(rq, irq_delta); +#if defined(CONFIG_IRQ_TIME_ACCOUNTING) || defined(CONFIG_PARAVIRT_TIME_ACCOUNTING) + if ((irq_delta + steal) && sched_feat(NONTASK_POWER)) + sched_rt_avg_update(rq, irq_delta + steal); +#endif } +#ifdef CONFIG_IRQ_TIME_ACCOUNTING static int irqtime_account_hi_update(void) { struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; @@ -2018,12 +2036,7 @@ static int irqtime_account_si_update(void) #define sched_clock_irqtime (0) -static void update_rq_clock_task(struct rq *rq, s64 delta) -{ - rq->clock_task += delta; -} - -#endif /* CONFIG_IRQ_TIME_ACCOUNTING */ +#endif #include "sched_idletask.c" #include "sched_fair.c" @@ -2200,6 +2213,16 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) !(task_thread_info(p)->preempt_count & PREEMPT_ACTIVE)); #ifdef CONFIG_LOCKDEP + /* + * The caller should hold either p->pi_lock or rq->lock, when changing + * a task's CPU. ->pi_lock for waking tasks, rq->lock for runnable tasks. + * + * sched_move_task() holds both and thus holding either pins the cgroup, + * see set_task_rq(). + * + * Furthermore, all task_rq users should acquire both locks, see + * task_rq_lock(). + */ WARN_ON_ONCE(debug_locks && !(lockdep_is_held(&p->pi_lock) || lockdep_is_held(&task_rq(p)->lock))); #endif @@ -2209,7 +2232,7 @@ void set_task_cpu(struct task_struct *p, unsigned int new_cpu) if (task_cpu(p) != new_cpu) { p->se.nr_migrations++; - perf_sw_event(PERF_COUNT_SW_CPU_MIGRATIONS, 1, 1, NULL, 0); + perf_sw_event(PERF_COUNT_SW_CPU_MIGRATIONS, 1, NULL, 0); } __set_task_cpu(p, new_cpu); @@ -2447,6 +2470,10 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) } rcu_read_unlock(); } + + if (wake_flags & WF_MIGRATED) + schedstat_inc(p, se.statistics.nr_wakeups_migrate); + #endif /* CONFIG_SMP */ schedstat_inc(rq, ttwu_count); @@ -2455,9 +2482,6 @@ ttwu_stat(struct task_struct *p, int cpu, int wake_flags) if (wake_flags & WF_SYNC) schedstat_inc(p, se.statistics.nr_wakeups_sync); - if (cpu != task_cpu(p)) - schedstat_inc(p, se.statistics.nr_wakeups_migrate); - #endif /* CONFIG_SCHEDSTATS */ } @@ -2485,7 +2509,7 @@ ttwu_do_wakeup(struct rq *rq, struct task_struct *p, int wake_flags) if (p->sched_class->task_woken) p->sched_class->task_woken(rq, p); - if (unlikely(rq->idle_stamp)) { + if (rq->idle_stamp) { u64 delta = rq->clock - rq->idle_stamp; u64 max = 2*sysctl_sched_migration_cost; @@ -2532,13 +2556,9 @@ static int ttwu_remote(struct task_struct *p, int wake_flags) } #ifdef CONFIG_SMP -static void sched_ttwu_pending(void) +static void sched_ttwu_do_pending(struct task_struct *list) { struct rq *rq = this_rq(); - struct task_struct *list = xchg(&rq->wake_list, NULL); - - if (!list) - return; raw_spin_lock(&rq->lock); @@ -2551,9 +2571,45 @@ static void sched_ttwu_pending(void) raw_spin_unlock(&rq->lock); } +#ifdef CONFIG_HOTPLUG_CPU + +static void sched_ttwu_pending(void) +{ + struct rq *rq = this_rq(); + struct task_struct *list = xchg(&rq->wake_list, NULL); + + if (!list) + return; + + sched_ttwu_do_pending(list); +} + +#endif /* CONFIG_HOTPLUG_CPU */ + void scheduler_ipi(void) { - sched_ttwu_pending(); + struct rq *rq = this_rq(); + struct task_struct *list = xchg(&rq->wake_list, NULL); + + if (!list) + return; + + /* + * Not all reschedule IPI handlers call irq_enter/irq_exit, since + * traditionally all their work was done from the interrupt return + * path. Now that we actually do some work, we need to make sure + * we do call them. + * + * Some archs already do call them, luckily irq_enter/exit nest + * properly. + * + * Arguably we should visit all archs and update all handlers, + * however a fair share of IPIs are still resched only so this would + * somewhat pessimize the simple resched case. + */ + irq_enter(); + sched_ttwu_do_pending(list); + irq_exit(); } static void ttwu_queue_remote(struct task_struct *p, int cpu) @@ -2600,6 +2656,7 @@ static void ttwu_queue(struct task_struct *p, int cpu) #if defined(CONFIG_SMP) if (sched_feat(TTWU_QUEUE) && cpu != smp_processor_id()) { + sched_clock_cpu(cpu); /* sync clocks x-cpu */ ttwu_queue_remote(p, cpu); return; } @@ -2674,8 +2731,10 @@ try_to_wake_up(struct task_struct *p, unsigned int state, int wake_flags) p->sched_class->task_waking(p); cpu = select_task_rq(p, SD_BALANCE_WAKE, wake_flags); - if (task_cpu(p) != cpu) + if (task_cpu(p) != cpu) { + wake_flags |= WF_MIGRATED; set_task_cpu(p, cpu); + } #endif /* CONFIG_SMP */ ttwu_queue(p, cpu); @@ -2839,7 +2898,7 @@ void sched_fork(struct task_struct *p) #if defined(CONFIG_SMP) p->on_cpu = 0; #endif -#ifdef CONFIG_PREEMPT +#ifdef CONFIG_PREEMPT_COUNT /* Want to start with kernel preemption disabled. */ task_thread_info(p)->preempt_count = 1; #endif @@ -3830,6 +3889,25 @@ void account_idle_time(cputime_t cputime) cpustat->idle = cputime64_add(cpustat->idle, cputime64); } +static __always_inline bool steal_account_process_tick(void) +{ +#ifdef CONFIG_PARAVIRT + if (static_branch(¶virt_steal_enabled)) { + u64 steal, st = 0; + + steal = paravirt_steal_clock(smp_processor_id()); + steal -= this_rq()->prev_steal_time; + + st = steal_ticks(steal); + this_rq()->prev_steal_time += st * TICK_NSEC; + + account_steal_time(st); + return st; + } +#endif + return false; +} + #ifndef CONFIG_VIRT_CPU_ACCOUNTING #ifdef CONFIG_IRQ_TIME_ACCOUNTING @@ -3861,6 +3939,9 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick, cputime64_t tmp = cputime_to_cputime64(cputime_one_jiffy); struct cpu_usage_stat *cpustat = &kstat_this_cpu.cpustat; + if (steal_account_process_tick()) + return; + if (irqtime_account_hi_update()) { cpustat->irq = cputime64_add(cpustat->irq, tmp); } else if (irqtime_account_si_update()) { @@ -3914,6 +3995,9 @@ void account_process_tick(struct task_struct *p, int user_tick) return; } + if (steal_account_process_tick()) + return; + if (user_tick) account_user_time(p, cputime_one_jiffy, one_jiffy_scaled); else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET)) @@ -4291,11 +4375,8 @@ EXPORT_SYMBOL(schedule); static inline bool owner_running(struct mutex *lock, struct task_struct *owner) { - bool ret = false; - - rcu_read_lock(); if (lock->owner != owner) - goto fail; + return false; /* * Ensure we emit the owner->on_cpu, dereference _after_ checking @@ -4305,11 +4386,7 @@ static inline bool owner_running(struct mutex *lock, struct task_struct *owner) */ barrier(); - ret = owner->on_cpu; -fail: - rcu_read_unlock(); - - return ret; + return owner->on_cpu; } /* @@ -4321,21 +4398,21 @@ int mutex_spin_on_owner(struct mutex *lock, struct task_struct *owner) if (!sched_feat(OWNER_SPIN)) return 0; + rcu_read_lock(); while (owner_running(lock, owner)) { if (need_resched()) - return 0; + break; arch_mutex_cpu_relax(); } + rcu_read_unlock(); /* - * If the owner changed to another task there is likely - * heavy contention, stop spinning. + * We break out the loop above on need_resched() and when the + * owner changed, which is a sign for heavy contention. Return + * success only when lock->owner is NULL. */ - if (lock->owner) - return 0; - - return 1; + return lock->owner == NULL; } #endif @@ -6542,7 +6619,7 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level, break; } - if (!group->cpu_power) { + if (!group->sgp->power) { printk(KERN_CONT "\n"); printk(KERN_ERR "ERROR: domain->cpu_power not " "set\n"); @@ -6566,9 +6643,9 @@ static int sched_domain_debug_one(struct sched_domain *sd, int cpu, int level, cpulist_scnprintf(str, sizeof(str), sched_group_cpus(group)); printk(KERN_CONT " %s", str); - if (group->cpu_power != SCHED_POWER_SCALE) { + if (group->sgp->power != SCHED_POWER_SCALE) { printk(KERN_CONT " (cpu_power = %d)", - group->cpu_power); + group->sgp->power); } group = group->next; @@ -6759,11 +6836,39 @@ static struct root_domain *alloc_rootdomain(void) return rd; } +static void free_sched_groups(struct sched_group *sg, int free_sgp) +{ + struct sched_group *tmp, *first; + + if (!sg) + return; + + first = sg; + do { + tmp = sg->next; + + if (free_sgp && atomic_dec_and_test(&sg->sgp->ref)) + kfree(sg->sgp); + + kfree(sg); + sg = tmp; + } while (sg != first); +} + static void free_sched_domain(struct rcu_head *rcu) { struct sched_domain *sd = container_of(rcu, struct sched_domain, rcu); - if (atomic_dec_and_test(&sd->groups->ref)) + + /* + * If its an overlapping domain it has private groups, iterate and + * nuke them all. + */ + if (sd->flags & SD_OVERLAP) { + free_sched_groups(sd->groups, 1); + } else if (atomic_dec_and_test(&sd->groups->ref)) { + kfree(sd->groups->sgp); kfree(sd->groups); + } kfree(sd); } @@ -6930,6 +7035,7 @@ int sched_smt_power_savings = 0, sched_mc_power_savings = 0; struct sd_data { struct sched_domain **__percpu sd; struct sched_group **__percpu sg; + struct sched_group_power **__percpu sgp; }; struct s_data { @@ -6949,15 +7055,73 @@ struct sched_domain_topology_level; typedef struct sched_domain *(*sched_domain_init_f)(struct sched_domain_topology_level *tl, int cpu); typedef const struct cpumask *(*sched_domain_mask_f)(int cpu); +#define SDTL_OVERLAP 0x01 + struct sched_domain_topology_level { sched_domain_init_f init; sched_domain_mask_f mask; + int flags; struct sd_data data; }; -/* - * Assumes the sched_domain tree is fully constructed - */ +static int +build_overlap_sched_groups(struct sched_domain *sd, int cpu) +{ + struct sched_group *first = NULL, *last = NULL, *groups = NULL, *sg; + const struct cpumask *span = sched_domain_span(sd); + struct cpumask *covered = sched_domains_tmpmask; + struct sd_data *sdd = sd->private; + struct sched_domain *child; + int i; + + cpumask_clear(covered); + + for_each_cpu(i, span) { + struct cpumask *sg_span; + + if (cpumask_test_cpu(i, covered)) + continue; + + sg = kzalloc_node(sizeof(struct sched_group) + cpumask_size(), + GFP_KERNEL, cpu_to_node(i)); + + if (!sg) + goto fail; + + sg_span = sched_group_cpus(sg); + + child = *per_cpu_ptr(sdd->sd, i); + if (child->child) { + child = child->child; + cpumask_copy(sg_span, sched_domain_span(child)); + } else + cpumask_set_cpu(i, sg_span); + + cpumask_or(covered, covered, sg_span); + + sg->sgp = *per_cpu_ptr(sdd->sgp, cpumask_first(sg_span)); + atomic_inc(&sg->sgp->ref); + + if (cpumask_test_cpu(cpu, sg_span)) + groups = sg; + + if (!first) + first = sg; + if (last) + last->next = sg; + last = sg; + last->next = first; + } + sd->groups = groups; + + return 0; + +fail: + free_sched_groups(first, 0); + + return -ENOMEM; +} + static int get_group(int cpu, struct sd_data *sdd, struct sched_group **sg) { struct sched_domain *sd = *per_cpu_ptr(sdd->sd, cpu); @@ -6966,24 +7130,24 @@ static int get_group(int cpu, struct sd_data *sdd, struct sched_group **sg) if (child) cpu = cpumask_first(sched_domain_span(child)); - if (sg) + if (sg) { *sg = *per_cpu_ptr(sdd->sg, cpu); + (*sg)->sgp = *per_cpu_ptr(sdd->sgp, cpu); + atomic_set(&(*sg)->sgp->ref, 1); /* for claim_allocations */ + } return cpu; } /* - * build_sched_groups takes the cpumask we wish to span, and a pointer - * to a function which identifies what group(along with sched group) a CPU - * belongs to. The return value of group_fn must be a >= 0 and < nr_cpu_ids - * (due to the fact that we keep track of groups covered with a struct cpumask). - * * build_sched_groups will build a circular linked list of the groups * covered by the given span, and will set each group's ->cpumask correctly, * and ->cpu_power to 0. + * + * Assumes the sched_domain tree is fully constructed */ -static void -build_sched_groups(struct sched_domain *sd) +static int +build_sched_groups(struct sched_domain *sd, int cpu) { struct sched_group *first = NULL, *last = NULL; struct sd_data *sdd = sd->private; @@ -6991,6 +7155,12 @@ build_sched_groups(struct sched_domain *sd) struct cpumask *covered; int i; + get_group(cpu, sdd, &sd->groups); + atomic_inc(&sd->groups->ref); + + if (cpu != cpumask_first(sched_domain_span(sd))) + return 0; + lockdep_assert_held(&sched_domains_mutex); covered = sched_domains_tmpmask; @@ -7005,7 +7175,7 @@ build_sched_groups(struct sched_domain *sd) continue; cpumask_clear(sched_group_cpus(sg)); - sg->cpu_power = 0; + sg->sgp->power = 0; for_each_cpu(j, span) { if (get_group(j, sdd, NULL) != group) @@ -7022,6 +7192,8 @@ build_sched_groups(struct sched_domain *sd) last = sg; } last->next = first; + + return 0; } /* @@ -7036,12 +7208,17 @@ build_sched_groups(struct sched_domain *sd) */ static void init_sched_groups_power(int cpu, struct sched_domain *sd) { - WARN_ON(!sd || !sd->groups); + struct sched_group *sg = sd->groups; - if (cpu != group_first_cpu(sd->groups)) - return; + WARN_ON(!sd || !sg); - sd->groups->group_weight = cpumask_weight(sched_group_cpus(sd->groups)); + do { + sg->group_weight = cpumask_weight(sched_group_cpus(sg)); + sg = sg->next; + } while (sg != sd->groups); + + if (cpu != group_first_cpu(sg)) + return; update_group_power(sd, cpu); } @@ -7162,15 +7339,15 @@ static enum s_alloc __visit_domain_allocation_hell(struct s_data *d, static void claim_allocations(int cpu, struct sched_domain *sd) { struct sd_data *sdd = sd->private; - struct sched_group *sg = sd->groups; WARN_ON_ONCE(*per_cpu_ptr(sdd->sd, cpu) != sd); *per_cpu_ptr(sdd->sd, cpu) = NULL; - if (cpu == cpumask_first(sched_group_cpus(sg))) { - WARN_ON_ONCE(*per_cpu_ptr(sdd->sg, cpu) != sg); + if (atomic_read(&(*per_cpu_ptr(sdd->sg, cpu))->ref)) *per_cpu_ptr(sdd->sg, cpu) = NULL; - } + + if (atomic_read(&(*per_cpu_ptr(sdd->sgp, cpu))->ref)) + *per_cpu_ptr(sdd->sgp, cpu) = NULL; } #ifdef CONFIG_SCHED_SMT @@ -7195,7 +7372,7 @@ static struct sched_domain_topology_level default_topology[] = { #endif { sd_init_CPU, cpu_cpu_mask, }, #ifdef CONFIG_NUMA - { sd_init_NODE, cpu_node_mask, }, + { sd_init_NODE, cpu_node_mask, SDTL_OVERLAP, }, { sd_init_ALLNODES, cpu_allnodes_mask, }, #endif { NULL, }, @@ -7219,9 +7396,14 @@ static int __sdt_alloc(const struct cpumask *cpu_map) if (!sdd->sg) return -ENOMEM; + sdd->sgp = alloc_percpu(struct sched_group_power *); + if (!sdd->sgp) + return -ENOMEM; + for_each_cpu(j, cpu_map) { struct sched_domain *sd; struct sched_group *sg; + struct sched_group_power *sgp; sd = kzalloc_node(sizeof(struct sched_domain) + cpumask_size(), GFP_KERNEL, cpu_to_node(j)); @@ -7236,6 +7418,13 @@ static int __sdt_alloc(const struct cpumask *cpu_map) return -ENOMEM; *per_cpu_ptr(sdd->sg, j) = sg; + + sgp = kzalloc_node(sizeof(struct sched_group_power), + GFP_KERNEL, cpu_to_node(j)); + if (!sgp) + return -ENOMEM; + + *per_cpu_ptr(sdd->sgp, j) = sgp; } } @@ -7251,11 +7440,15 @@ static void __sdt_free(const struct cpumask *cpu_map) struct sd_data *sdd = &tl->data; for_each_cpu(j, cpu_map) { - kfree(*per_cpu_ptr(sdd->sd, j)); + struct sched_domain *sd = *per_cpu_ptr(sdd->sd, j); + if (sd && (sd->flags & SD_OVERLAP)) + free_sched_groups(sd->groups, 0); kfree(*per_cpu_ptr(sdd->sg, j)); + kfree(*per_cpu_ptr(sdd->sgp, j)); } free_percpu(sdd->sd); free_percpu(sdd->sg); + free_percpu(sdd->sgp); } } @@ -7301,8 +7494,13 @@ static int build_sched_domains(const struct cpumask *cpu_map, struct sched_domain_topology_level *tl; sd = NULL; - for (tl = sched_domain_topology; tl->init; tl++) + for (tl = sched_domain_topology; tl->init; tl++) { sd = build_sched_domain(tl, &d, cpu_map, attr, sd, i); + if (tl->flags & SDTL_OVERLAP || sched_feat(FORCE_SD_OVERLAP)) + sd->flags |= SD_OVERLAP; + if (cpumask_equal(cpu_map, sched_domain_span(sd))) + break; + } while (sd->child) sd = sd->child; @@ -7314,13 +7512,13 @@ static int build_sched_domains(const struct cpumask *cpu_map, for_each_cpu(i, cpu_map) { for (sd = *per_cpu_ptr(d.sd, i); sd; sd = sd->parent) { sd->span_weight = cpumask_weight(sched_domain_span(sd)); - get_group(i, sd->private, &sd->groups); - atomic_inc(&sd->groups->ref); - - if (i != cpumask_first(sched_domain_span(sd))) - continue; - - build_sched_groups(sd); + if (sd->flags & SD_OVERLAP) { + if (build_overlap_sched_groups(sd, i)) + goto error; + } else { + if (build_sched_groups(sd, i)) + goto error; + } } } @@ -7730,18 +7928,14 @@ int in_sched_functions(unsigned long addr) && addr < (unsigned long)__sched_text_end); } -static void init_cfs_rq(struct cfs_rq *cfs_rq, struct rq *rq) +static void init_cfs_rq(struct cfs_rq *cfs_rq) { cfs_rq->tasks_timeline = RB_ROOT; INIT_LIST_HEAD(&cfs_rq->tasks); -#ifdef CONFIG_FAIR_GROUP_SCHED - cfs_rq->rq = rq; - /* allow initial update_cfs_load() to truncate */ -#ifdef CONFIG_SMP - cfs_rq->load_stamp = 1; -#endif -#endif cfs_rq->min_vruntime = (u64)(-(1LL << 20)); +#ifndef CONFIG_64BIT + cfs_rq->min_vruntime_copy = cfs_rq->min_vruntime; +#endif } static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq) @@ -7757,27 +7951,18 @@ static void init_rt_rq(struct rt_rq *rt_rq, struct rq *rq) /* delimiter for bitsearch: */ __set_bit(MAX_RT_PRIO, array->bitmap); -#if defined CONFIG_SMP || defined CONFIG_RT_GROUP_SCHED +#if defined CONFIG_SMP rt_rq->highest_prio.curr = MAX_RT_PRIO; -#ifdef CONFIG_SMP rt_rq->highest_prio.next = MAX_RT_PRIO; -#endif -#endif -#ifdef CONFIG_SMP rt_rq->rt_nr_migratory = 0; rt_rq->overloaded = 0; - plist_head_init_raw(&rt_rq->pushable_tasks, &rq->lock); + plist_head_init(&rt_rq->pushable_tasks); #endif rt_rq->rt_time = 0; rt_rq->rt_throttled = 0; rt_rq->rt_runtime = 0; raw_spin_lock_init(&rt_rq->rt_runtime_lock); - -#ifdef CONFIG_RT_GROUP_SCHED - rt_rq->rt_nr_boosted = 0; - rt_rq->rq = rq; -#endif } #ifdef CONFIG_FAIR_GROUP_SCHED @@ -7786,11 +7971,17 @@ static void init_tg_cfs_entry(struct task_group *tg, struct cfs_rq *cfs_rq, struct sched_entity *parent) { struct rq *rq = cpu_rq(cpu); - tg->cfs_rq[cpu] = cfs_rq; - init_cfs_rq(cfs_rq, rq); + cfs_rq->tg = tg; + cfs_rq->rq = rq; +#ifdef CONFIG_SMP + /* allow initial update_cfs_load() to truncate */ + cfs_rq->load_stamp = 1; +#endif + tg->cfs_rq[cpu] = cfs_rq; tg->se[cpu] = se; + /* se could be NULL for root_task_group */ if (!se) return; @@ -7813,12 +8004,14 @@ static void init_tg_rt_entry(struct task_group *tg, struct rt_rq *rt_rq, { struct rq *rq = cpu_rq(cpu); - tg->rt_rq[cpu] = rt_rq; - init_rt_rq(rt_rq, rq); + rt_rq->highest_prio.curr = MAX_RT_PRIO; + rt_rq->rt_nr_boosted = 0; + rt_rq->rq = rq; rt_rq->tg = tg; - rt_rq->rt_runtime = tg->rt_bandwidth.rt_runtime; + tg->rt_rq[cpu] = rt_rq; tg->rt_se[cpu] = rt_se; + if (!rt_se) return; @@ -7900,7 +8093,7 @@ void __init sched_init(void) rq->nr_running = 0; rq->calc_load_active = 0; rq->calc_load_update = jiffies + LOAD_FREQ; - init_cfs_rq(&rq->cfs, rq); + init_cfs_rq(&rq->cfs); init_rt_rq(&rq->rt, rq); #ifdef CONFIG_FAIR_GROUP_SCHED root_task_group.shares = root_task_group_load; @@ -7971,7 +8164,7 @@ void __init sched_init(void) #endif #ifdef CONFIG_RT_MUTEXES - plist_head_init_raw(&init_task.pi_waiters, &init_task.pi_lock); + plist_head_init(&init_task.pi_waiters); #endif /* @@ -8014,7 +8207,7 @@ void __init sched_init(void) scheduler_running = 1; } -#ifdef CONFIG_DEBUG_SPINLOCK_SLEEP +#ifdef CONFIG_DEBUG_ATOMIC_SLEEP static inline int preempt_count_equals(int preempt_offset) { int nested = (preempt_count() & ~PREEMPT_ACTIVE) + rcu_preempt_depth(); @@ -8024,7 +8217,6 @@ static inline int preempt_count_equals(int preempt_offset) void __might_sleep(const char *file, int line, int preempt_offset) { -#ifdef in_atomic static unsigned long prev_jiffy; /* ratelimiting */ if ((preempt_count_equals(preempt_offset) && !irqs_disabled()) || @@ -8046,7 +8238,6 @@ void __might_sleep(const char *file, int line, int preempt_offset) if (irqs_disabled()) print_irqtrace_events(current); dump_stack(); -#endif } EXPORT_SYMBOL(__might_sleep); #endif @@ -8205,6 +8396,7 @@ int alloc_fair_sched_group(struct task_group *tg, struct task_group *parent) if (!se) goto err_free_rq; + init_cfs_rq(cfs_rq); init_tg_cfs_entry(tg, cfs_rq, se, i, parent->se[i]); } @@ -8232,7 +8424,7 @@ static inline void unregister_fair_sched_group(struct task_group *tg, int cpu) list_del_leaf_cfs_rq(tg->cfs_rq[cpu]); raw_spin_unlock_irqrestore(&rq->lock, flags); } -#else /* !CONFG_FAIR_GROUP_SCHED */ +#else /* !CONFIG_FAIR_GROUP_SCHED */ static inline void free_fair_sched_group(struct task_group *tg) { } @@ -8253,7 +8445,8 @@ static void free_rt_sched_group(struct task_group *tg) { int i; - destroy_rt_bandwidth(&tg->rt_bandwidth); + if (tg->rt_se) + destroy_rt_bandwidth(&tg->rt_bandwidth); for_each_possible_cpu(i) { if (tg->rt_rq) @@ -8294,6 +8487,8 @@ int alloc_rt_sched_group(struct task_group *tg, struct task_group *parent) if (!rt_se) goto err_free_rq; + init_rt_rq(rt_rq, cpu_rq(i)); + rt_rq->rt_runtime = tg->rt_bandwidth.rt_runtime; init_tg_rt_entry(tg, rt_rq, rt_se, i, parent->rt_se[i]); } @@ -8435,10 +8630,7 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares) if (!tg->se[0]) return -EINVAL; - if (shares < MIN_SHARES) - shares = MIN_SHARES; - else if (shares > MAX_SHARES) - shares = MAX_SHARES; + shares = clamp(shares, scale_load(MIN_SHARES), scale_load(MAX_SHARES)); mutex_lock(&shares_mutex); if (tg->shares == shares) diff --git a/kernel/sched_autogroup.h b/kernel/sched_autogroup.h index 05577055cfc..c2f0e7248dc 100644 --- a/kernel/sched_autogroup.h +++ b/kernel/sched_autogroup.h @@ -13,6 +13,7 @@ struct autogroup { int nice; }; +static inline bool task_group_is_autogroup(struct task_group *tg); static inline struct task_group * autogroup_task_group(struct task_struct *p, struct task_group *tg); diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 433491c2dc8..bc8ee999381 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c @@ -135,14 +135,6 @@ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp) return grp->my_q; } -/* Given a group's cfs_rq on one cpu, return its corresponding cfs_rq on - * another cpu ('this_cpu') - */ -static inline struct cfs_rq *cpu_cfs_rq(struct cfs_rq *cfs_rq, int this_cpu) -{ - return cfs_rq->tg->cfs_rq[this_cpu]; -} - static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq) { if (!cfs_rq->on_list) { @@ -271,11 +263,6 @@ static inline struct cfs_rq *group_cfs_rq(struct sched_entity *grp) return NULL; } -static inline struct cfs_rq *cpu_cfs_rq(struct cfs_rq *cfs_rq, int this_cpu) -{ - return &cpu_rq(this_cpu)->cfs; -} - static inline void list_add_leaf_cfs_rq(struct cfs_rq *cfs_rq) { } @@ -334,11 +321,6 @@ static inline int entity_before(struct sched_entity *a, return (s64)(a->vruntime - b->vruntime) < 0; } -static inline s64 entity_key(struct cfs_rq *cfs_rq, struct sched_entity *se) -{ - return se->vruntime - cfs_rq->min_vruntime; -} - static void update_min_vruntime(struct cfs_rq *cfs_rq) { u64 vruntime = cfs_rq->min_vruntime; @@ -372,7 +354,6 @@ static void __enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) struct rb_node **link = &cfs_rq->tasks_timeline.rb_node; struct rb_node *parent = NULL; struct sched_entity *entry; - s64 key = entity_key(cfs_rq, se); int leftmost = 1; /* @@ -385,7 +366,7 @@ static void __enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se) * We dont care about collisions. Nodes with * the same key stay together. */ - if (key < entity_key(cfs_rq, entry)) { + if (entity_before(se, entry)) { link = &parent->rb_left; } else { link = &parent->rb_right; @@ -1336,7 +1317,7 @@ enqueue_task_fair(struct rq *rq, struct task_struct *p, int flags) } for_each_sched_entity(se) { - struct cfs_rq *cfs_rq = cfs_rq_of(se); + cfs_rq = cfs_rq_of(se); update_cfs_load(cfs_rq, 0); update_cfs_shares(cfs_rq); @@ -1370,13 +1351,16 @@ static void dequeue_task_fair(struct rq *rq, struct task_struct *p, int flags) */ if (task_sleep && parent_entity(se)) set_next_buddy(parent_entity(se)); + + /* avoid re-evaluating load for this entity */ + se = parent_entity(se); break; } flags |= DEQUEUE_SLEEP; } for_each_sched_entity(se) { - struct cfs_rq *cfs_rq = cfs_rq_of(se); + cfs_rq = cfs_rq_of(se); update_cfs_load(cfs_rq, 0); update_cfs_shares(cfs_rq); @@ -1481,7 +1465,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) * effect of the currently running task from the load * of the current CPU: */ - rcu_read_lock(); if (sync) { tg = task_group(current); weight = current->se.load.weight; @@ -1517,7 +1500,6 @@ static int wake_affine(struct sched_domain *sd, struct task_struct *p, int sync) balanced = this_eff_load <= prev_eff_load; } else balanced = true; - rcu_read_unlock(); /* * If the currently running task will sleep within @@ -1585,7 +1567,7 @@ find_idlest_group(struct sched_domain *sd, struct task_struct *p, } /* Adjust by relative CPU power of the group */ - avg_load = (avg_load * SCHED_POWER_SCALE) / group->cpu_power; + avg_load = (avg_load * SCHED_POWER_SCALE) / group->sgp->power; if (local_group) { this_load = avg_load; @@ -1921,8 +1903,8 @@ static void check_preempt_wakeup(struct rq *rq, struct task_struct *p, int wake_ if (!sched_feat(WAKEUP_PREEMPT)) return; - update_curr(cfs_rq); find_matching_se(&se, &pse); + update_curr(cfs_rq_of(se)); BUG_ON(!pse); if (wakeup_preempt_entity(se, pse) == 1) { /* @@ -2231,11 +2213,43 @@ static void update_shares(int cpu) struct rq *rq = cpu_rq(cpu); rcu_read_lock(); + /* + * Iterates the task_group tree in a bottom up fashion, see + * list_add_leaf_cfs_rq() for details. + */ for_each_leaf_cfs_rq(rq, cfs_rq) update_shares_cpu(cfs_rq->tg, cpu); rcu_read_unlock(); } +/* + * Compute the cpu's hierarchical load factor for each task group. + * This needs to be done in a top-down fashion because the load of a child + * group is a fraction of its parents load. + */ +static int tg_load_down(struct task_group *tg, void *data) +{ + unsigned long load; + long cpu = (long)data; + + if (!tg->parent) { + load = cpu_rq(cpu)->load.weight; + } else { + load = tg->parent->cfs_rq[cpu]->h_load; + load *= tg->se[cpu]->load.weight; + load /= tg->parent->cfs_rq[cpu]->load.weight + 1; + } + + tg->cfs_rq[cpu]->h_load = load; + + return 0; +} + +static void update_h_load(long cpu) +{ + walk_tg_tree(tg_load_down, tg_nop, (void *)cpu); +} + static unsigned long load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, unsigned long max_load_move, @@ -2243,14 +2257,12 @@ load_balance_fair(struct rq *this_rq, int this_cpu, struct rq *busiest, int *all_pinned) { long rem_load_move = max_load_move; - int busiest_cpu = cpu_of(busiest); - struct task_group *tg; + struct cfs_rq *busiest_cfs_rq; rcu_read_lock(); - update_h_load(busiest_cpu); + update_h_load(cpu_of(busiest)); - list_for_each_entry_rcu(tg, &task_groups, list) { - struct cfs_rq *busiest_cfs_rq = tg->cfs_rq[busiest_cpu]; + for_each_leaf_cfs_rq(busiest, busiest_cfs_rq) { unsigned long busiest_h_load = busiest_cfs_rq->h_load; unsigned long busiest_weight = busiest_cfs_rq->load.weight; u64 rem_load, moved_load; @@ -2631,7 +2643,7 @@ static void update_cpu_power(struct sched_domain *sd, int cpu) power >>= SCHED_POWER_SHIFT; } - sdg->cpu_power_orig = power; + sdg->sgp->power_orig = power; if (sched_feat(ARCH_POWER)) power *= arch_scale_freq_power(sd, cpu); @@ -2647,7 +2659,7 @@ static void update_cpu_power(struct sched_domain *sd, int cpu) power = 1; cpu_rq(cpu)->cpu_power = power; - sdg->cpu_power = power; + sdg->sgp->power = power; } static void update_group_power(struct sched_domain *sd, int cpu) @@ -2665,11 +2677,11 @@ static void update_group_power(struct sched_domain *sd, int cpu) group = child->groups; do { - power += group->cpu_power; + power += group->sgp->power; group = group->next; } while (group != child->groups); - sdg->cpu_power = power; + sdg->sgp->power = power; } /* @@ -2691,7 +2703,7 @@ fix_small_capacity(struct sched_domain *sd, struct sched_group *group) /* * If ~90% of the cpu_power is still there, we're good. */ - if (group->cpu_power * 32 > group->cpu_power_orig * 29) + if (group->sgp->power * 32 > group->sgp->power_orig * 29) return 1; return 0; @@ -2771,7 +2783,7 @@ static inline void update_sg_lb_stats(struct sched_domain *sd, } /* Adjust by relative CPU power of the group */ - sgs->avg_load = (sgs->group_load*SCHED_POWER_SCALE) / group->cpu_power; + sgs->avg_load = (sgs->group_load*SCHED_POWER_SCALE) / group->sgp->power; /* * Consider the group unbalanced when the imbalance is larger @@ -2788,7 +2800,7 @@ static inline void update_sg_lb_stats(struct sched_domain *sd, if ((max_cpu_load - min_cpu_load) >= avg_load_per_task && max_nr_running > 1) sgs->group_imb = 1; - sgs->group_capacity = DIV_ROUND_CLOSEST(group->cpu_power, + sgs->group_capacity = DIV_ROUND_CLOSEST(group->sgp->power, SCHED_POWER_SCALE); if (!sgs->group_capacity) sgs->group_capacity = fix_small_capacity(sd, group); @@ -2877,7 +2889,7 @@ static inline void update_sd_lb_stats(struct sched_domain *sd, int this_cpu, return; sds->total_load += sgs.group_load; - sds->total_pwr += sg->cpu_power; + sds->total_pwr += sg->sgp->power; /* * In case the child domain prefers tasks go to siblings @@ -2962,7 +2974,7 @@ static int check_asym_packing(struct sched_domain *sd, if (this_cpu > busiest_cpu) return 0; - *imbalance = DIV_ROUND_CLOSEST(sds->max_load * sds->busiest->cpu_power, + *imbalance = DIV_ROUND_CLOSEST(sds->max_load * sds->busiest->sgp->power, SCHED_POWER_SCALE); return 1; } @@ -2993,7 +3005,7 @@ static inline void fix_small_imbalance(struct sd_lb_stats *sds, scaled_busy_load_per_task = sds->busiest_load_per_task * SCHED_POWER_SCALE; - scaled_busy_load_per_task /= sds->busiest->cpu_power; + scaled_busy_load_per_task /= sds->busiest->sgp->power; if (sds->max_load - sds->this_load + scaled_busy_load_per_task >= (scaled_busy_load_per_task * imbn)) { @@ -3007,28 +3019,28 @@ static inline void fix_small_imbalance(struct sd_lb_stats *sds, * moving them. */ - pwr_now += sds->busiest->cpu_power * + pwr_now += sds->busiest->sgp->power * min(sds->busiest_load_per_task, sds->max_load); - pwr_now += sds->this->cpu_power * + pwr_now += sds->this->sgp->power * min(sds->this_load_per_task, sds->this_load); pwr_now /= SCHED_POWER_SCALE; /* Amount of load we'd subtract */ tmp = (sds->busiest_load_per_task * SCHED_POWER_SCALE) / - sds->busiest->cpu_power; + sds->busiest->sgp->power; if (sds->max_load > tmp) - pwr_move += sds->busiest->cpu_power * + pwr_move += sds->busiest->sgp->power * min(sds->busiest_load_per_task, sds->max_load - tmp); /* Amount of load we'd add */ - if (sds->max_load * sds->busiest->cpu_power < + if (sds->max_load * sds->busiest->sgp->power < sds->busiest_load_per_task * SCHED_POWER_SCALE) - tmp = (sds->max_load * sds->busiest->cpu_power) / - sds->this->cpu_power; + tmp = (sds->max_load * sds->busiest->sgp->power) / + sds->this->sgp->power; else tmp = (sds->busiest_load_per_task * SCHED_POWER_SCALE) / - sds->this->cpu_power; - pwr_move += sds->this->cpu_power * + sds->this->sgp->power; + pwr_move += sds->this->sgp->power * min(sds->this_load_per_task, sds->this_load + tmp); pwr_move /= SCHED_POWER_SCALE; @@ -3074,7 +3086,7 @@ static inline void calculate_imbalance(struct sd_lb_stats *sds, int this_cpu, load_above_capacity *= (SCHED_LOAD_SCALE * SCHED_POWER_SCALE); - load_above_capacity /= sds->busiest->cpu_power; + load_above_capacity /= sds->busiest->sgp->power; } /* @@ -3090,8 +3102,8 @@ static inline void calculate_imbalance(struct sd_lb_stats *sds, int this_cpu, max_pull = min(sds->max_load - sds->avg_load, load_above_capacity); /* How much load to actually move to equalise the imbalance */ - *imbalance = min(max_pull * sds->busiest->cpu_power, - (sds->avg_load - sds->this_load) * sds->this->cpu_power) + *imbalance = min(max_pull * sds->busiest->sgp->power, + (sds->avg_load - sds->this_load) * sds->this->sgp->power) / SCHED_POWER_SCALE; /* diff --git a/kernel/sched_features.h b/kernel/sched_features.h index be40f7371ee..2e74677cb04 100644 --- a/kernel/sched_features.h +++ b/kernel/sched_features.h @@ -61,12 +61,14 @@ SCHED_FEAT(LB_BIAS, 1) SCHED_FEAT(OWNER_SPIN, 1) /* - * Decrement CPU power based on irq activity + * Decrement CPU power based on time not spent running tasks */ -SCHED_FEAT(NONIRQ_POWER, 1) +SCHED_FEAT(NONTASK_POWER, 1) /* * Queue remote wakeups on the target CPU and process them * using the scheduler IPI. Reduces rq->lock contention/bounces. */ SCHED_FEAT(TTWU_QUEUE, 1) + +SCHED_FEAT(FORCE_SD_OVERLAP, 0) diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index 88725c939e0..97540f0c9e4 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c @@ -185,11 +185,23 @@ static inline u64 sched_rt_period(struct rt_rq *rt_rq) typedef struct task_group *rt_rq_iter_t; -#define for_each_rt_rq(rt_rq, iter, rq) \ - for (iter = list_entry_rcu(task_groups.next, typeof(*iter), list); \ - (&iter->list != &task_groups) && \ - (rt_rq = iter->rt_rq[cpu_of(rq)]); \ - iter = list_entry_rcu(iter->list.next, typeof(*iter), list)) +static inline struct task_group *next_task_group(struct task_group *tg) +{ + do { + tg = list_entry_rcu(tg->list.next, + typeof(struct task_group), list); + } while (&tg->list != &task_groups && task_group_is_autogroup(tg)); + + if (&tg->list == &task_groups) + tg = NULL; + + return tg; +} + +#define for_each_rt_rq(rt_rq, iter, rq) \ + for (iter = container_of(&task_groups, typeof(*iter), list); \ + (iter = next_task_group(iter)) && \ + (rt_rq = iter->rt_rq[cpu_of(rq)]);) static inline void list_add_leaf_rt_rq(struct rt_rq *rt_rq) { @@ -1096,7 +1108,7 @@ static void check_preempt_curr_rt(struct rq *rq, struct task_struct *p, int flag * to move current somewhere else, making room for our non-migratable * task. */ - if (p->prio == rq->curr->prio && !need_resched()) + if (p->prio == rq->curr->prio && !test_tsk_need_resched(rq->curr)) check_preempt_equal_prio(rq, p); #endif } @@ -1126,7 +1138,7 @@ static struct task_struct *_pick_next_task_rt(struct rq *rq) rt_rq = &rq->rt; - if (unlikely(!rt_rq->rt_nr_running)) + if (!rt_rq->rt_nr_running) return NULL; if (rt_rq_throttled(rt_rq)) @@ -1239,6 +1251,10 @@ static int find_lowest_rq(struct task_struct *task) int this_cpu = smp_processor_id(); int cpu = task_cpu(task); + /* Make sure the mask is initialized first */ + if (unlikely(!lowest_mask)) + return -1; + if (task->rt.nr_cpus_allowed == 1) return -1; /* No other targets possible */ @@ -1544,7 +1560,7 @@ skip: static void pre_schedule_rt(struct rq *rq, struct task_struct *prev) { /* Try to pull RT tasks here if we lower this rq's prio */ - if (unlikely(rt_task(prev)) && rq->rt.highest_prio.curr > prev->prio) + if (rq->rt.highest_prio.curr > prev->prio) pull_rt_task(rq); } diff --git a/kernel/signal.c b/kernel/signal.c index 86c32b884f8..d7f70aed1cc 100644 --- a/kernel/signal.c +++ b/kernel/signal.c @@ -87,7 +87,7 @@ static int sig_ignored(struct task_struct *t, int sig, int from_ancestor_ns) /* * Tracers may want to know about even ignored signals. */ - return !tracehook_consider_ignored_signal(t, sig); + return !t->ptrace; } /* @@ -124,7 +124,7 @@ static inline int has_pending_signals(sigset_t *signal, sigset_t *blocked) static int recalc_sigpending_tsk(struct task_struct *t) { - if ((t->group_stop & GROUP_STOP_PENDING) || + if ((t->jobctl & JOBCTL_PENDING_MASK) || PENDING(&t->pending, &t->blocked) || PENDING(&t->signal->shared_pending, &t->blocked)) { set_tsk_thread_flag(t, TIF_SIGPENDING); @@ -150,9 +150,7 @@ void recalc_sigpending_and_wake(struct task_struct *t) void recalc_sigpending(void) { - if (unlikely(tracehook_force_sigpending())) - set_thread_flag(TIF_SIGPENDING); - else if (!recalc_sigpending_tsk(current) && !freezing(current)) + if (!recalc_sigpending_tsk(current) && !freezing(current)) clear_thread_flag(TIF_SIGPENDING); } @@ -224,47 +222,93 @@ static inline void print_dropped_signal(int sig) } /** - * task_clear_group_stop_trapping - clear group stop trapping bit + * task_set_jobctl_pending - set jobctl pending bits * @task: target task + * @mask: pending bits to set * - * If GROUP_STOP_TRAPPING is set, a ptracer is waiting for us. Clear it - * and wake up the ptracer. Note that we don't need any further locking. - * @task->siglock guarantees that @task->parent points to the ptracer. + * Clear @mask from @task->jobctl. @mask must be subset of + * %JOBCTL_PENDING_MASK | %JOBCTL_STOP_CONSUME | %JOBCTL_STOP_SIGMASK | + * %JOBCTL_TRAPPING. If stop signo is being set, the existing signo is + * cleared. If @task is already being killed or exiting, this function + * becomes noop. + * + * CONTEXT: + * Must be called with @task->sighand->siglock held. + * + * RETURNS: + * %true if @mask is set, %false if made noop because @task was dying. + */ +bool task_set_jobctl_pending(struct task_struct *task, unsigned int mask) +{ + BUG_ON(mask & ~(JOBCTL_PENDING_MASK | JOBCTL_STOP_CONSUME | + JOBCTL_STOP_SIGMASK | JOBCTL_TRAPPING)); + BUG_ON((mask & JOBCTL_TRAPPING) && !(mask & JOBCTL_PENDING_MASK)); + + if (unlikely(fatal_signal_pending(task) || (task->flags & PF_EXITING))) + return false; + + if (mask & JOBCTL_STOP_SIGMASK) + task->jobctl &= ~JOBCTL_STOP_SIGMASK; + + task->jobctl |= mask; + return true; +} + +/** + * task_clear_jobctl_trapping - clear jobctl trapping bit + * @task: target task + * + * If JOBCTL_TRAPPING is set, a ptracer is waiting for us to enter TRACED. + * Clear it and wake up the ptracer. Note that we don't need any further + * locking. @task->siglock guarantees that @task->parent points to the + * ptracer. * * CONTEXT: * Must be called with @task->sighand->siglock held. */ -static void task_clear_group_stop_trapping(struct task_struct *task) +void task_clear_jobctl_trapping(struct task_struct *task) { - if (unlikely(task->group_stop & GROUP_STOP_TRAPPING)) { - task->group_stop &= ~GROUP_STOP_TRAPPING; - __wake_up_sync_key(&task->parent->signal->wait_chldexit, - TASK_UNINTERRUPTIBLE, 1, task); + if (unlikely(task->jobctl & JOBCTL_TRAPPING)) { + task->jobctl &= ~JOBCTL_TRAPPING; + wake_up_bit(&task->jobctl, JOBCTL_TRAPPING_BIT); } } /** - * task_clear_group_stop_pending - clear pending group stop + * task_clear_jobctl_pending - clear jobctl pending bits * @task: target task + * @mask: pending bits to clear * - * Clear group stop states for @task. + * Clear @mask from @task->jobctl. @mask must be subset of + * %JOBCTL_PENDING_MASK. If %JOBCTL_STOP_PENDING is being cleared, other + * STOP bits are cleared together. + * + * If clearing of @mask leaves no stop or trap pending, this function calls + * task_clear_jobctl_trapping(). * * CONTEXT: * Must be called with @task->sighand->siglock held. */ -void task_clear_group_stop_pending(struct task_struct *task) +void task_clear_jobctl_pending(struct task_struct *task, unsigned int mask) { - task->group_stop &= ~(GROUP_STOP_PENDING | GROUP_STOP_CONSUME | - GROUP_STOP_DEQUEUED); + BUG_ON(mask & ~JOBCTL_PENDING_MASK); + + if (mask & JOBCTL_STOP_PENDING) + mask |= JOBCTL_STOP_CONSUME | JOBCTL_STOP_DEQUEUED; + + task->jobctl &= ~mask; + + if (!(task->jobctl & JOBCTL_PENDING_MASK)) + task_clear_jobctl_trapping(task); } /** * task_participate_group_stop - participate in a group stop * @task: task participating in a group stop * - * @task has GROUP_STOP_PENDING set and is participating in a group stop. + * @task has %JOBCTL_STOP_PENDING set and is participating in a group stop. * Group stop states are cleared and the group stop count is consumed if - * %GROUP_STOP_CONSUME was set. If the consumption completes the group + * %JOBCTL_STOP_CONSUME was set. If the consumption completes the group * stop, the appropriate %SIGNAL_* flags are set. * * CONTEXT: @@ -277,11 +321,11 @@ void task_clear_group_stop_pending(struct task_struct *task) static bool task_participate_group_stop(struct task_struct *task) { struct signal_struct *sig = task->signal; - bool consume = task->group_stop & GROUP_STOP_CONSUME; + bool consume = task->jobctl & JOBCTL_STOP_CONSUME; - WARN_ON_ONCE(!(task->group_stop & GROUP_STOP_PENDING)); + WARN_ON_ONCE(!(task->jobctl & JOBCTL_STOP_PENDING)); - task_clear_group_stop_pending(task); + task_clear_jobctl_pending(task, JOBCTL_STOP_PENDING); if (!consume) return false; @@ -449,7 +493,8 @@ int unhandled_signal(struct task_struct *tsk, int sig) return 1; if (handler != SIG_IGN && handler != SIG_DFL) return 0; - return !tracehook_consider_fatal_signal(tsk, sig); + /* if ptraced, let the tracer determine */ + return !tsk->ptrace; } /* @@ -604,7 +649,7 @@ int dequeue_signal(struct task_struct *tsk, sigset_t *mask, siginfo_t *info) * is to alert stop-signal processing code when another * processor has come along and cleared the flag. */ - current->group_stop |= GROUP_STOP_DEQUEUED; + current->jobctl |= JOBCTL_STOP_DEQUEUED; } if ((info->si_code & __SI_MASK) == __SI_TIMER && info->si_sys_private) { /* @@ -773,6 +818,32 @@ static int check_kill_permission(int sig, struct siginfo *info, return security_task_kill(t, info, sig, 0); } +/** + * ptrace_trap_notify - schedule trap to notify ptracer + * @t: tracee wanting to notify tracer + * + * This function schedules sticky ptrace trap which is cleared on the next + * TRAP_STOP to notify ptracer of an event. @t must have been seized by + * ptracer. + * + * If @t is running, STOP trap will be taken. If trapped for STOP and + * ptracer is listening for events, tracee is woken up so that it can + * re-trap for the new event. If trapped otherwise, STOP trap will be + * eventually taken without returning to userland after the existing traps + * are finished by PTRACE_CONT. + * + * CONTEXT: + * Must be called with @task->sighand->siglock held. + */ +static void ptrace_trap_notify(struct task_struct *t) +{ + WARN_ON_ONCE(!(t->ptrace & PT_SEIZED)); + assert_spin_locked(&t->sighand->siglock); + + task_set_jobctl_pending(t, JOBCTL_TRAP_NOTIFY); + signal_wake_up(t, t->jobctl & JOBCTL_LISTENING); +} + /* * Handle magic process-wide effects of stop/continue signals. Unlike * the signal actions, these happen immediately at signal-generation @@ -809,9 +880,12 @@ static int prepare_signal(int sig, struct task_struct *p, int from_ancestor_ns) rm_from_queue(SIG_KERNEL_STOP_MASK, &signal->shared_pending); t = p; do { - task_clear_group_stop_pending(t); + task_clear_jobctl_pending(t, JOBCTL_STOP_PENDING); rm_from_queue(SIG_KERNEL_STOP_MASK, &t->pending); - wake_up_state(t, __TASK_STOPPED); + if (likely(!(t->ptrace & PT_SEIZED))) + wake_up_state(t, __TASK_STOPPED); + else + ptrace_trap_notify(t); } while_each_thread(p, t); /* @@ -908,8 +982,7 @@ static void complete_signal(int sig, struct task_struct *p, int group) if (sig_fatal(p, sig) && !(signal->flags & (SIGNAL_UNKILLABLE | SIGNAL_GROUP_EXIT)) && !sigismember(&t->real_blocked, sig) && - (sig == SIGKILL || - !tracehook_consider_fatal_signal(t, sig))) { + (sig == SIGKILL || !t->ptrace)) { /* * This signal will be fatal to the whole group. */ @@ -925,7 +998,7 @@ static void complete_signal(int sig, struct task_struct *p, int group) signal->group_stop_count = 0; t = p; do { - task_clear_group_stop_pending(t); + task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK); sigaddset(&t->pending.signal, SIGKILL); signal_wake_up(t, 1); } while_each_thread(p, t); @@ -1160,7 +1233,7 @@ int zap_other_threads(struct task_struct *p) p->signal->group_stop_count = 0; while_each_thread(p, t) { - task_clear_group_stop_pending(t); + task_clear_jobctl_pending(t, JOBCTL_PENDING_MASK); count++; /* Don't bother with already dead threads */ @@ -1178,18 +1251,25 @@ struct sighand_struct *__lock_task_sighand(struct task_struct *tsk, { struct sighand_struct *sighand; - rcu_read_lock(); for (;;) { + local_irq_save(*flags); + rcu_read_lock(); sighand = rcu_dereference(tsk->sighand); - if (unlikely(sighand == NULL)) + if (unlikely(sighand == NULL)) { + rcu_read_unlock(); + local_irq_restore(*flags); break; + } - spin_lock_irqsave(&sighand->siglock, *flags); - if (likely(sighand == tsk->sighand)) + spin_lock(&sighand->siglock); + if (likely(sighand == tsk->sighand)) { + rcu_read_unlock(); break; - spin_unlock_irqrestore(&sighand->siglock, *flags); + } + spin_unlock(&sighand->siglock); + rcu_read_unlock(); + local_irq_restore(*flags); } - rcu_read_unlock(); return sighand; } @@ -1504,22 +1584,22 @@ ret: * Let a parent know about the death of a child. * For a stopped/continued status change, use do_notify_parent_cldstop instead. * - * Returns -1 if our parent ignored us and so we've switched to - * self-reaping, or else @sig. + * Returns true if our parent ignored us and so we've switched to + * self-reaping. */ -int do_notify_parent(struct task_struct *tsk, int sig) +bool do_notify_parent(struct task_struct *tsk, int sig) { struct siginfo info; unsigned long flags; struct sighand_struct *psig; - int ret = sig; + bool autoreap = false; BUG_ON(sig == -1); /* do_notify_parent_cldstop should have been called instead. */ BUG_ON(task_is_stopped_or_traced(tsk)); - BUG_ON(!task_ptrace(tsk) && + BUG_ON(!tsk->ptrace && (tsk->group_leader != tsk || !thread_group_empty(tsk))); info.si_signo = sig; @@ -1558,7 +1638,7 @@ int do_notify_parent(struct task_struct *tsk, int sig) psig = tsk->parent->sighand; spin_lock_irqsave(&psig->siglock, flags); - if (!task_ptrace(tsk) && sig == SIGCHLD && + if (!tsk->ptrace && sig == SIGCHLD && (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN || (psig->action[SIGCHLD-1].sa.sa_flags & SA_NOCLDWAIT))) { /* @@ -1576,16 +1656,16 @@ int do_notify_parent(struct task_struct *tsk, int sig) * is implementation-defined: we do (if you don't want * it, just use SIG_IGN instead). */ - ret = tsk->exit_signal = -1; + autoreap = true; if (psig->action[SIGCHLD-1].sa.sa_handler == SIG_IGN) - sig = -1; + sig = 0; } - if (valid_signal(sig) && sig > 0) + if (valid_signal(sig) && sig) __group_send_sig_info(sig, &info, tsk->parent); __wake_up_parent(tsk, tsk->parent); spin_unlock_irqrestore(&psig->siglock, flags); - return ret; + return autoreap; } /** @@ -1658,7 +1738,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk, static inline int may_ptrace_stop(void) { - if (!likely(task_ptrace(current))) + if (!likely(current->ptrace)) return 0; /* * Are we in the middle of do_coredump? @@ -1687,15 +1767,6 @@ static int sigkill_pending(struct task_struct *tsk) } /* - * Test whether the target task of the usual cldstop notification - the - * real_parent of @child - is in the same group as the ptracer. - */ -static bool real_parent_is_ptracer(struct task_struct *child) -{ - return same_thread_group(child->parent, child->real_parent); -} - -/* * This must be called with current->sighand->siglock held. * * This should be the path for all ptrace stops. @@ -1732,31 +1803,34 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) } /* - * If @why is CLD_STOPPED, we're trapping to participate in a group - * stop. Do the bookkeeping. Note that if SIGCONT was delievered - * while siglock was released for the arch hook, PENDING could be - * clear now. We act as if SIGCONT is received after TASK_TRACED - * is entered - ignore it. + * We're committing to trapping. TRACED should be visible before + * TRAPPING is cleared; otherwise, the tracer might fail do_wait(). + * Also, transition to TRACED and updates to ->jobctl should be + * atomic with respect to siglock and should be done after the arch + * hook as siglock is released and regrabbed across it. */ - if (why == CLD_STOPPED && (current->group_stop & GROUP_STOP_PENDING)) - gstop_done = task_participate_group_stop(current); + set_current_state(TASK_TRACED); current->last_siginfo = info; current->exit_code = exit_code; /* - * TRACED should be visible before TRAPPING is cleared; otherwise, - * the tracer might fail do_wait(). + * If @why is CLD_STOPPED, we're trapping to participate in a group + * stop. Do the bookkeeping. Note that if SIGCONT was delievered + * across siglock relocks since INTERRUPT was scheduled, PENDING + * could be clear now. We act as if SIGCONT is received after + * TASK_TRACED is entered - ignore it. */ - set_current_state(TASK_TRACED); + if (why == CLD_STOPPED && (current->jobctl & JOBCTL_STOP_PENDING)) + gstop_done = task_participate_group_stop(current); - /* - * We're committing to trapping. Clearing GROUP_STOP_TRAPPING and - * transition to TASK_TRACED should be atomic with respect to - * siglock. This hsould be done after the arch hook as siglock is - * released and regrabbed across it. - */ - task_clear_group_stop_trapping(current); + /* any trap clears pending STOP trap, STOP trap clears NOTIFY */ + task_clear_jobctl_pending(current, JOBCTL_TRAP_STOP); + if (info && info->si_code >> 8 == PTRACE_EVENT_STOP) + task_clear_jobctl_pending(current, JOBCTL_TRAP_NOTIFY); + + /* entering a trap, clear TRAPPING */ + task_clear_jobctl_trapping(current); spin_unlock_irq(¤t->sighand->siglock); read_lock(&tasklist_lock); @@ -1772,7 +1846,7 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) * separately unless they're gonna be duplicates. */ do_notify_parent_cldstop(current, true, why); - if (gstop_done && !real_parent_is_ptracer(current)) + if (gstop_done && ptrace_reparented(current)) do_notify_parent_cldstop(current, false, why); /* @@ -1792,9 +1866,9 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) * * If @gstop_done, the ptracer went away between group stop * completion and here. During detach, it would have set - * GROUP_STOP_PENDING on us and we'll re-enter TASK_STOPPED - * in do_signal_stop() on return, so notifying the real - * parent of the group stop completion is enough. + * JOBCTL_STOP_PENDING on us and we'll re-enter + * TASK_STOPPED in do_signal_stop() on return, so notifying + * the real parent of the group stop completion is enough. */ if (gstop_done) do_notify_parent_cldstop(current, false, why); @@ -1820,6 +1894,9 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) spin_lock_irq(¤t->sighand->siglock); current->last_siginfo = NULL; + /* LISTENING can be set only during STOP traps, clear it */ + current->jobctl &= ~JOBCTL_LISTENING; + /* * Queued signals ignored us while we were stopped for tracing. * So check for any that we should take before resuming user mode. @@ -1828,44 +1905,66 @@ static void ptrace_stop(int exit_code, int why, int clear_code, siginfo_t *info) recalc_sigpending_tsk(current); } -void ptrace_notify(int exit_code) +static void ptrace_do_notify(int signr, int exit_code, int why) { siginfo_t info; - BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP); - memset(&info, 0, sizeof info); - info.si_signo = SIGTRAP; + info.si_signo = signr; info.si_code = exit_code; info.si_pid = task_pid_vnr(current); info.si_uid = current_uid(); /* Let the debugger run. */ + ptrace_stop(exit_code, why, 1, &info); +} + +void ptrace_notify(int exit_code) +{ + BUG_ON((exit_code & (0x7f | ~0xffff)) != SIGTRAP); + spin_lock_irq(¤t->sighand->siglock); - ptrace_stop(exit_code, CLD_TRAPPED, 1, &info); + ptrace_do_notify(SIGTRAP, exit_code, CLD_TRAPPED); spin_unlock_irq(¤t->sighand->siglock); } -/* - * This performs the stopping for SIGSTOP and other stop signals. - * We have to stop all threads in the thread group. - * Returns non-zero if we've actually stopped and released the siglock. - * Returns zero if we didn't stop and still hold the siglock. +/** + * do_signal_stop - handle group stop for SIGSTOP and other stop signals + * @signr: signr causing group stop if initiating + * + * If %JOBCTL_STOP_PENDING is not set yet, initiate group stop with @signr + * and participate in it. If already set, participate in the existing + * group stop. If participated in a group stop (and thus slept), %true is + * returned with siglock released. + * + * If ptraced, this function doesn't handle stop itself. Instead, + * %JOBCTL_TRAP_STOP is scheduled and %false is returned with siglock + * untouched. The caller must ensure that INTERRUPT trap handling takes + * places afterwards. + * + * CONTEXT: + * Must be called with @current->sighand->siglock held, which is released + * on %true return. + * + * RETURNS: + * %false if group stop is already cancelled or ptrace trap is scheduled. + * %true if participated in group stop. */ -static int do_signal_stop(int signr) +static bool do_signal_stop(int signr) + __releases(¤t->sighand->siglock) { struct signal_struct *sig = current->signal; - if (!(current->group_stop & GROUP_STOP_PENDING)) { - unsigned int gstop = GROUP_STOP_PENDING | GROUP_STOP_CONSUME; + if (!(current->jobctl & JOBCTL_STOP_PENDING)) { + unsigned int gstop = JOBCTL_STOP_PENDING | JOBCTL_STOP_CONSUME; struct task_struct *t; - /* signr will be recorded in task->group_stop for retries */ - WARN_ON_ONCE(signr & ~GROUP_STOP_SIGMASK); + /* signr will be recorded in task->jobctl for retries */ + WARN_ON_ONCE(signr & ~JOBCTL_STOP_SIGMASK); - if (!likely(current->group_stop & GROUP_STOP_DEQUEUED) || + if (!likely(current->jobctl & JOBCTL_STOP_DEQUEUED) || unlikely(signal_group_exit(sig))) - return 0; + return false; /* * There is no group stop already in progress. We must * initiate one now. @@ -1888,28 +1987,32 @@ static int do_signal_stop(int signr) if (!(sig->flags & SIGNAL_STOP_STOPPED)) sig->group_exit_code = signr; else - WARN_ON_ONCE(!task_ptrace(current)); + WARN_ON_ONCE(!current->ptrace); + + sig->group_stop_count = 0; + + if (task_set_jobctl_pending(current, signr | gstop)) + sig->group_stop_count++; - current->group_stop &= ~GROUP_STOP_SIGMASK; - current->group_stop |= signr | gstop; - sig->group_stop_count = 1; for (t = next_thread(current); t != current; t = next_thread(t)) { - t->group_stop &= ~GROUP_STOP_SIGMASK; /* * Setting state to TASK_STOPPED for a group * stop is always done with the siglock held, * so this check has no races. */ - if (!(t->flags & PF_EXITING) && !task_is_stopped(t)) { - t->group_stop |= signr | gstop; + if (!task_is_stopped(t) && + task_set_jobctl_pending(t, signr | gstop)) { sig->group_stop_count++; - signal_wake_up(t, 0); + if (likely(!(t->ptrace & PT_SEIZED))) + signal_wake_up(t, 0); + else + ptrace_trap_notify(t); } } } -retry: - if (likely(!task_ptrace(current))) { + + if (likely(!current->ptrace)) { int notify = 0; /* @@ -1940,43 +2043,65 @@ retry: /* Now we don't run again until woken by SIGCONT or SIGKILL */ schedule(); - - spin_lock_irq(¤t->sighand->siglock); + return true; } else { - ptrace_stop(current->group_stop & GROUP_STOP_SIGMASK, - CLD_STOPPED, 0, NULL); - current->exit_code = 0; + /* + * While ptraced, group stop is handled by STOP trap. + * Schedule it and let the caller deal with it. + */ + task_set_jobctl_pending(current, JOBCTL_TRAP_STOP); + return false; } +} - /* - * GROUP_STOP_PENDING could be set if another group stop has - * started since being woken up or ptrace wants us to transit - * between TASK_STOPPED and TRACED. Retry group stop. - */ - if (current->group_stop & GROUP_STOP_PENDING) { - WARN_ON_ONCE(!(current->group_stop & GROUP_STOP_SIGMASK)); - goto retry; +/** + * do_jobctl_trap - take care of ptrace jobctl traps + * + * When PT_SEIZED, it's used for both group stop and explicit + * SEIZE/INTERRUPT traps. Both generate PTRACE_EVENT_STOP trap with + * accompanying siginfo. If stopped, lower eight bits of exit_code contain + * the stop signal; otherwise, %SIGTRAP. + * + * When !PT_SEIZED, it's used only for group stop trap with stop signal + * number as exit_code and no siginfo. + * + * CONTEXT: + * Must be called with @current->sighand->siglock held, which may be + * released and re-acquired before returning with intervening sleep. + */ +static void do_jobctl_trap(void) +{ + struct signal_struct *signal = current->signal; + int signr = current->jobctl & JOBCTL_STOP_SIGMASK; + + if (current->ptrace & PT_SEIZED) { + if (!signal->group_stop_count && + !(signal->flags & SIGNAL_STOP_STOPPED)) + signr = SIGTRAP; + WARN_ON_ONCE(!signr); + ptrace_do_notify(signr, signr | (PTRACE_EVENT_STOP << 8), + CLD_STOPPED); + } else { + WARN_ON_ONCE(!signr); + ptrace_stop(signr, CLD_STOPPED, 0, NULL); + current->exit_code = 0; } - - /* PTRACE_ATTACH might have raced with task killing, clear trapping */ - task_clear_group_stop_trapping(current); - - spin_unlock_irq(¤t->sighand->siglock); - - tracehook_finish_jctl(); - - return 1; } static int ptrace_signal(int signr, siginfo_t *info, struct pt_regs *regs, void *cookie) { - if (!task_ptrace(current)) - return signr; - ptrace_signal_deliver(regs, cookie); - - /* Let the debugger run. */ + /* + * We do not check sig_kernel_stop(signr) but set this marker + * unconditionally because we do not know whether debugger will + * change signr. This flag has no meaning unless we are going + * to stop after return from ptrace_stop(). In this case it will + * be checked in do_signal_stop(), we should only stop if it was + * not cleared by SIGCONT while we were sleeping. See also the + * comment in dequeue_signal(). + */ + current->jobctl |= JOBCTL_STOP_DEQUEUED; ptrace_stop(signr, CLD_TRAPPED, 0, info); /* We're back. Did the debugger cancel the sig? */ @@ -2032,7 +2157,6 @@ relock: * the CLD_ si_code into SIGNAL_CLD_MASK bits. */ if (unlikely(signal->flags & SIGNAL_CLD_MASK)) { - struct task_struct *leader; int why; if (signal->flags & SIGNAL_CLD_CONTINUED) @@ -2053,13 +2177,11 @@ relock: * a duplicate. */ read_lock(&tasklist_lock); - do_notify_parent_cldstop(current, false, why); - leader = current->group_leader; - if (task_ptrace(leader) && !real_parent_is_ptracer(leader)) - do_notify_parent_cldstop(leader, true, why); - + if (ptrace_reparented(current->group_leader)) + do_notify_parent_cldstop(current->group_leader, + true, why); read_unlock(&tasklist_lock); goto relock; @@ -2067,37 +2189,31 @@ relock: for (;;) { struct k_sigaction *ka; - /* - * Tracing can induce an artificial signal and choose sigaction. - * The return value in @signr determines the default action, - * but @info->si_signo is the signal number we will report. - */ - signr = tracehook_get_signal(current, regs, info, return_ka); - if (unlikely(signr < 0)) + + if (unlikely(current->jobctl & JOBCTL_STOP_PENDING) && + do_signal_stop(0)) goto relock; - if (unlikely(signr != 0)) - ka = return_ka; - else { - if (unlikely(current->group_stop & - GROUP_STOP_PENDING) && do_signal_stop(0)) - goto relock; - signr = dequeue_signal(current, ¤t->blocked, - info); + if (unlikely(current->jobctl & JOBCTL_TRAP_MASK)) { + do_jobctl_trap(); + spin_unlock_irq(&sighand->siglock); + goto relock; + } - if (!signr) - break; /* will return 0 */ + signr = dequeue_signal(current, ¤t->blocked, info); - if (signr != SIGKILL) { - signr = ptrace_signal(signr, info, - regs, cookie); - if (!signr) - continue; - } + if (!signr) + break; /* will return 0 */ - ka = &sighand->action[signr-1]; + if (unlikely(current->ptrace) && signr != SIGKILL) { + signr = ptrace_signal(signr, info, + regs, cookie); + if (!signr) + continue; } + ka = &sighand->action[signr-1]; + /* Trace actually delivered signals. */ trace_signal_deliver(signr, info, ka); @@ -2253,7 +2369,7 @@ void exit_signals(struct task_struct *tsk) signotset(&unblocked); retarget_shared_pending(tsk, &unblocked); - if (unlikely(tsk->group_stop & GROUP_STOP_PENDING) && + if (unlikely(tsk->jobctl & JOBCTL_STOP_PENDING) && task_participate_group_stop(tsk)) group_stop = CLD_STOPPED; out: @@ -2365,7 +2481,7 @@ int sigprocmask(int how, sigset_t *set, sigset_t *oldset) /** * sys_rt_sigprocmask - change the list of currently blocked signals * @how: whether to add, remove, or set signals - * @set: stores pending signals + * @nset: stores pending signals * @oset: previous value of signal mask if non-null * @sigsetsize: size of sigset_t type */ diff --git a/kernel/smp.c b/kernel/smp.c index 73a19519355..fb67dfa8394 100644 --- a/kernel/smp.c +++ b/kernel/smp.c @@ -74,7 +74,7 @@ static struct notifier_block __cpuinitdata hotplug_cfd_notifier = { .notifier_call = hotplug_cfd, }; -static int __cpuinit init_call_single_data(void) +void __init call_function_init(void) { void *cpu = (void *)(long)smp_processor_id(); int i; @@ -88,10 +88,7 @@ static int __cpuinit init_call_single_data(void) hotplug_cfd(&hotplug_cfd_notifier, CPU_UP_PREPARE, cpu); register_cpu_notifier(&hotplug_cfd_notifier); - - return 0; } -early_initcall(init_call_single_data); /* * csd_lock/csd_unlock used to serialize access to per-cpu csd resources diff --git a/kernel/softirq.c b/kernel/softirq.c index 13960170cad..fca82c32042 100644 --- a/kernel/softirq.c +++ b/kernel/softirq.c @@ -58,7 +58,7 @@ DEFINE_PER_CPU(struct task_struct *, ksoftirqd); char *softirq_to_name[NR_SOFTIRQS] = { "HI", "TIMER", "NET_TX", "NET_RX", "BLOCK", "BLOCK_IOPOLL", - "TASKLET", "SCHED", "HRTIMER" + "TASKLET", "SCHED", "HRTIMER", "RCU" }; /* @@ -315,16 +315,24 @@ static inline void invoke_softirq(void) { if (!force_irqthreads) __do_softirq(); - else + else { + __local_bh_disable((unsigned long)__builtin_return_address(0), + SOFTIRQ_OFFSET); wakeup_softirqd(); + __local_bh_enable(SOFTIRQ_OFFSET); + } } #else static inline void invoke_softirq(void) { if (!force_irqthreads) do_softirq(); - else + else { + __local_bh_disable((unsigned long)__builtin_return_address(0), + SOFTIRQ_OFFSET); wakeup_softirqd(); + __local_bh_enable(SOFTIRQ_OFFSET); + } } #endif diff --git a/kernel/stacktrace.c b/kernel/stacktrace.c index eb212f8f8bc..d20c6983aad 100644 --- a/kernel/stacktrace.c +++ b/kernel/stacktrace.c @@ -26,12 +26,18 @@ void print_stack_trace(struct stack_trace *trace, int spaces) EXPORT_SYMBOL_GPL(print_stack_trace); /* - * Architectures that do not implement save_stack_trace_tsk get this - * weak alias and a once-per-bootup warning (whenever this facility - * is utilized - for example by procfs): + * Architectures that do not implement save_stack_trace_tsk or + * save_stack_trace_regs get this weak alias and a once-per-bootup warning + * (whenever this facility is utilized - for example by procfs): */ __weak void save_stack_trace_tsk(struct task_struct *tsk, struct stack_trace *trace) { WARN_ONCE(1, KERN_INFO "save_stack_trace_tsk() not implemented yet.\n"); } + +__weak void +save_stack_trace_regs(struct pt_regs *regs, struct stack_trace *trace) +{ + WARN_ONCE(1, KERN_INFO "save_stack_trace_regs() not implemented yet.\n"); +} diff --git a/kernel/stop_machine.c b/kernel/stop_machine.c index e3516b29076..c1124752e1d 100644 --- a/kernel/stop_machine.c +++ b/kernel/stop_machine.c @@ -136,10 +136,11 @@ void stop_one_cpu_nowait(unsigned int cpu, cpu_stop_fn_t fn, void *arg, static DEFINE_MUTEX(stop_cpus_mutex); static DEFINE_PER_CPU(struct cpu_stop_work, stop_cpus_work); -int __stop_cpus(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg) +static void queue_stop_cpus_work(const struct cpumask *cpumask, + cpu_stop_fn_t fn, void *arg, + struct cpu_stop_done *done) { struct cpu_stop_work *work; - struct cpu_stop_done done; unsigned int cpu; /* initialize works and done */ @@ -147,9 +148,8 @@ int __stop_cpus(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg) work = &per_cpu(stop_cpus_work, cpu); work->fn = fn; work->arg = arg; - work->done = &done; + work->done = done; } - cpu_stop_init_done(&done, cpumask_weight(cpumask)); /* * Disable preemption while queueing to avoid getting @@ -161,7 +161,15 @@ int __stop_cpus(const struct cpumask *cpumask, cpu_stop_fn_t fn, void *arg) cpu_stop_queue_work(&per_cpu(cpu_stopper, cpu), &per_cpu(stop_cpus_work, cpu)); preempt_enable(); +} +static int __stop_cpus(const struct cpumask *cpumask, + cpu_stop_fn_t fn, void *arg) +{ + struct cpu_stop_done done; + + cpu_stop_init_done(&done, cpumask_weight(cpumask)); + queue_stop_cpus_work(cpumask, fn, arg, &done); wait_for_completion(&done.completion); return done.executed ? done.ret : -ENOENT; } @@ -431,8 +439,15 @@ static int stop_machine_cpu_stop(void *data) struct stop_machine_data *smdata = data; enum stopmachine_state curstate = STOPMACHINE_NONE; int cpu = smp_processor_id(), err = 0; + unsigned long flags; bool is_active; + /* + * When called from stop_machine_from_inactive_cpu(), irq might + * already be disabled. Save the state and restore it on exit. + */ + local_save_flags(flags); + if (!smdata->active_cpus) is_active = cpu == cpumask_first(cpu_online_mask); else @@ -460,7 +475,7 @@ static int stop_machine_cpu_stop(void *data) } } while (curstate != STOPMACHINE_EXIT); - local_irq_enable(); + local_irq_restore(flags); return err; } @@ -487,4 +502,57 @@ int stop_machine(int (*fn)(void *), void *data, const struct cpumask *cpus) } EXPORT_SYMBOL_GPL(stop_machine); +/** + * stop_machine_from_inactive_cpu - stop_machine() from inactive CPU + * @fn: the function to run + * @data: the data ptr for the @fn() + * @cpus: the cpus to run the @fn() on (NULL = any online cpu) + * + * This is identical to stop_machine() but can be called from a CPU which + * is not active. The local CPU is in the process of hotplug (so no other + * CPU hotplug can start) and not marked active and doesn't have enough + * context to sleep. + * + * This function provides stop_machine() functionality for such state by + * using busy-wait for synchronization and executing @fn directly for local + * CPU. + * + * CONTEXT: + * Local CPU is inactive. Temporarily stops all active CPUs. + * + * RETURNS: + * 0 if all executions of @fn returned 0, any non zero return value if any + * returned non zero. + */ +int stop_machine_from_inactive_cpu(int (*fn)(void *), void *data, + const struct cpumask *cpus) +{ + struct stop_machine_data smdata = { .fn = fn, .data = data, + .active_cpus = cpus }; + struct cpu_stop_done done; + int ret; + + /* Local CPU must be inactive and CPU hotplug in progress. */ + BUG_ON(cpu_active(raw_smp_processor_id())); + smdata.num_threads = num_active_cpus() + 1; /* +1 for local */ + + /* No proper task established and can't sleep - busy wait for lock. */ + while (!mutex_trylock(&stop_cpus_mutex)) + cpu_relax(); + + /* Schedule work on other CPUs and execute directly for local CPU */ + set_state(&smdata, STOPMACHINE_PREPARE); + cpu_stop_init_done(&done, num_active_cpus()); + queue_stop_cpus_work(cpu_active_mask, stop_machine_cpu_stop, &smdata, + &done); + ret = stop_machine_cpu_stop(&smdata); + + /* Busy wait for completion. */ + while (!completion_done(&done.completion)) + cpu_relax(); + + mutex_unlock(&stop_cpus_mutex); + return ret ?: done.ret; +} + #endif /* CONFIG_STOP_MACHINE */ diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 4fc92445a29..11d65b531e5 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -938,6 +938,12 @@ static struct ctl_table kern_table[] = { }, #endif #ifdef CONFIG_PERF_EVENTS + /* + * User-space scripts rely on the existence of this file + * as a feature check for perf_events being enabled. + * + * So it's an ABI, do not remove! + */ { .procname = "perf_event_paranoid", .data = &sysctl_perf_event_paranoid, @@ -1584,16 +1590,11 @@ void sysctl_head_get(struct ctl_table_header *head) spin_unlock(&sysctl_lock); } -static void free_head(struct rcu_head *rcu) -{ - kfree(container_of(rcu, struct ctl_table_header, rcu)); -} - void sysctl_head_put(struct ctl_table_header *head) { spin_lock(&sysctl_lock); if (!--head->count) - call_rcu(&head->rcu, free_head); + kfree_rcu(head, rcu); spin_unlock(&sysctl_lock); } @@ -1965,10 +1966,10 @@ void unregister_sysctl_table(struct ctl_table_header * header) start_unregistering(header); if (!--header->parent->count) { WARN_ON(1); - call_rcu(&header->parent->rcu, free_head); + kfree_rcu(header->parent, rcu); } if (!--header->count) - call_rcu(&header->rcu, free_head); + kfree_rcu(header, rcu); spin_unlock(&sysctl_lock); } diff --git a/kernel/taskstats.c b/kernel/taskstats.c index 9ffea360a77..fc0f2200541 100644 --- a/kernel/taskstats.c +++ b/kernel/taskstats.c @@ -285,16 +285,18 @@ ret: static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd) { struct listener_list *listeners; - struct listener *s, *tmp; + struct listener *s, *tmp, *s2; unsigned int cpu; if (!cpumask_subset(mask, cpu_possible_mask)) return -EINVAL; + s = NULL; if (isadd == REGISTER) { for_each_cpu(cpu, mask) { - s = kmalloc_node(sizeof(struct listener), GFP_KERNEL, - cpu_to_node(cpu)); + if (!s) + s = kmalloc_node(sizeof(struct listener), + GFP_KERNEL, cpu_to_node(cpu)); if (!s) goto cleanup; s->pid = pid; @@ -303,9 +305,16 @@ static int add_del_listener(pid_t pid, const struct cpumask *mask, int isadd) listeners = &per_cpu(listener_array, cpu); down_write(&listeners->sem); + list_for_each_entry_safe(s2, tmp, &listeners->list, list) { + if (s2->pid == pid) + goto next_cpu; + } list_add(&s->list, &listeners->list); + s = NULL; +next_cpu: up_write(&listeners->sem); } + kfree(s); return 0; } diff --git a/kernel/time/alarmtimer.c b/kernel/time/alarmtimer.c index 2d966244ea6..59f369f98a0 100644 --- a/kernel/time/alarmtimer.c +++ b/kernel/time/alarmtimer.c @@ -42,15 +42,75 @@ static struct alarm_base { clockid_t base_clockid; } alarm_bases[ALARM_NUMTYPE]; +/* freezer delta & lock used to handle clock_nanosleep triggered wakeups */ +static ktime_t freezer_delta; +static DEFINE_SPINLOCK(freezer_delta_lock); + #ifdef CONFIG_RTC_CLASS /* rtc timer and device for setting alarm wakeups at suspend */ static struct rtc_timer rtctimer; static struct rtc_device *rtcdev; -#endif +static DEFINE_SPINLOCK(rtcdev_lock); -/* freezer delta & lock used to handle clock_nanosleep triggered wakeups */ -static ktime_t freezer_delta; -static DEFINE_SPINLOCK(freezer_delta_lock); +/** + * has_wakealarm - check rtc device has wakealarm ability + * @dev: current device + * @name_ptr: name to be returned + * + * This helper function checks to see if the rtc device can wake + * from suspend. + */ +static int has_wakealarm(struct device *dev, void *name_ptr) +{ + struct rtc_device *candidate = to_rtc_device(dev); + + if (!candidate->ops->set_alarm) + return 0; + if (!device_may_wakeup(candidate->dev.parent)) + return 0; + + *(const char **)name_ptr = dev_name(dev); + return 1; +} + +/** + * alarmtimer_get_rtcdev - Return selected rtcdevice + * + * This function returns the rtc device to use for wakealarms. + * If one has not already been chosen, it checks to see if a + * functional rtc device is available. + */ +static struct rtc_device *alarmtimer_get_rtcdev(void) +{ + struct device *dev; + char *str; + unsigned long flags; + struct rtc_device *ret; + + spin_lock_irqsave(&rtcdev_lock, flags); + if (!rtcdev) { + /* Find an rtc device and init the rtc_timer */ + dev = class_find_device(rtc_class, NULL, &str, has_wakealarm); + /* If we have a device then str is valid. See has_wakealarm() */ + if (dev) { + rtcdev = rtc_class_open(str); + /* + * Drop the reference we got in class_find_device, + * rtc_open takes its own. + */ + put_device(dev); + rtc_timer_init(&rtctimer, NULL, NULL); + } + } + ret = rtcdev; + spin_unlock_irqrestore(&rtcdev_lock, flags); + + return ret; +} +#else +#define alarmtimer_get_rtcdev() (0) +#define rtcdev (0) +#endif /** @@ -166,6 +226,7 @@ static int alarmtimer_suspend(struct device *dev) struct rtc_time tm; ktime_t min, now; unsigned long flags; + struct rtc_device *rtc; int i; spin_lock_irqsave(&freezer_delta_lock, flags); @@ -173,8 +234,9 @@ static int alarmtimer_suspend(struct device *dev) freezer_delta = ktime_set(0, 0); spin_unlock_irqrestore(&freezer_delta_lock, flags); + rtc = rtcdev; /* If we have no rtcdev, just return */ - if (!rtcdev) + if (!rtc) return 0; /* Find the soonest timer to expire*/ @@ -199,12 +261,12 @@ static int alarmtimer_suspend(struct device *dev) WARN_ON(min.tv64 < NSEC_PER_SEC); /* Setup an rtc timer to fire that far in the future */ - rtc_timer_cancel(rtcdev, &rtctimer); - rtc_read_time(rtcdev, &tm); + rtc_timer_cancel(rtc, &rtctimer); + rtc_read_time(rtc, &tm); now = rtc_tm_to_ktime(tm); now = ktime_add(now, min); - rtc_timer_start(rtcdev, &rtctimer, now, ktime_set(0, 0)); + rtc_timer_start(rtc, &rtctimer, now, ktime_set(0, 0)); return 0; } @@ -322,6 +384,9 @@ static int alarm_clock_getres(const clockid_t which_clock, struct timespec *tp) { clockid_t baseid = alarm_bases[clock2alarm(which_clock)].base_clockid; + if (!alarmtimer_get_rtcdev()) + return -ENOTSUPP; + return hrtimer_get_res(baseid, tp); } @@ -336,6 +401,9 @@ static int alarm_clock_get(clockid_t which_clock, struct timespec *tp) { struct alarm_base *base = &alarm_bases[clock2alarm(which_clock)]; + if (!alarmtimer_get_rtcdev()) + return -ENOTSUPP; + *tp = ktime_to_timespec(base->gettime()); return 0; } @@ -351,6 +419,9 @@ static int alarm_timer_create(struct k_itimer *new_timer) enum alarmtimer_type type; struct alarm_base *base; + if (!alarmtimer_get_rtcdev()) + return -ENOTSUPP; + if (!capable(CAP_WAKE_ALARM)) return -EPERM; @@ -385,6 +456,9 @@ static void alarm_timer_get(struct k_itimer *timr, */ static int alarm_timer_del(struct k_itimer *timr) { + if (!rtcdev) + return -ENOTSUPP; + alarm_cancel(&timr->it.alarmtimer); return 0; } @@ -402,6 +476,9 @@ static int alarm_timer_set(struct k_itimer *timr, int flags, struct itimerspec *new_setting, struct itimerspec *old_setting) { + if (!rtcdev) + return -ENOTSUPP; + /* Save old values */ old_setting->it_interval = ktime_to_timespec(timr->it.alarmtimer.period); @@ -541,6 +618,9 @@ static int alarm_timer_nsleep(const clockid_t which_clock, int flags, int ret = 0; struct restart_block *restart; + if (!alarmtimer_get_rtcdev()) + return -ENOTSUPP; + if (!capable(CAP_WAKE_ALARM)) return -EPERM; @@ -638,65 +718,3 @@ static int __init alarmtimer_init(void) } device_initcall(alarmtimer_init); -#ifdef CONFIG_RTC_CLASS -/** - * has_wakealarm - check rtc device has wakealarm ability - * @dev: current device - * @name_ptr: name to be returned - * - * This helper function checks to see if the rtc device can wake - * from suspend. - */ -static int __init has_wakealarm(struct device *dev, void *name_ptr) -{ - struct rtc_device *candidate = to_rtc_device(dev); - - if (!candidate->ops->set_alarm) - return 0; - if (!device_may_wakeup(candidate->dev.parent)) - return 0; - - *(const char **)name_ptr = dev_name(dev); - return 1; -} - -/** - * alarmtimer_init_late - Late initializing of alarmtimer code - * - * This function locates a rtc device to use for wakealarms. - * Run as late_initcall to make sure rtc devices have been - * registered. - */ -static int __init alarmtimer_init_late(void) -{ - struct device *dev; - char *str; - - /* Find an rtc device and init the rtc_timer */ - dev = class_find_device(rtc_class, NULL, &str, has_wakealarm); - /* If we have a device then str is valid. See has_wakealarm() */ - if (dev) { - rtcdev = rtc_class_open(str); - /* - * Drop the reference we got in class_find_device, - * rtc_open takes its own. - */ - put_device(dev); - } - if (!rtcdev) { - printk(KERN_WARNING "No RTC device found, ALARM timers will" - " not wake from suspend"); - } - rtc_timer_init(&rtctimer, NULL, NULL); - - return 0; -} -#else -static int __init alarmtimer_init_late(void) -{ - printk(KERN_WARNING "Kernel not built with RTC support, ALARM timers" - " will not wake from suspend"); - return 0; -} -#endif -late_initcall(alarmtimer_init_late); diff --git a/kernel/time/clockevents.c b/kernel/time/clockevents.c index c027d4f602f..e4c699dfa4e 100644 --- a/kernel/time/clockevents.c +++ b/kernel/time/clockevents.c @@ -182,7 +182,10 @@ void clockevents_register_device(struct clock_event_device *dev) unsigned long flags; BUG_ON(dev->mode != CLOCK_EVT_MODE_UNUSED); - BUG_ON(!dev->cpumask); + if (!dev->cpumask) { + WARN_ON(num_possible_cpus() > 1); + dev->cpumask = cpumask_of(smp_processor_id()); + } raw_spin_lock_irqsave(&clockevents_lock, flags); diff --git a/kernel/time/clocksource.c b/kernel/time/clocksource.c index 1c95fd67732..e0980f0d9a0 100644 --- a/kernel/time/clocksource.c +++ b/kernel/time/clocksource.c @@ -185,7 +185,6 @@ static struct clocksource *watchdog; static struct timer_list watchdog_timer; static DECLARE_WORK(watchdog_work, clocksource_watchdog_work); static DEFINE_SPINLOCK(watchdog_lock); -static cycle_t watchdog_last; static int watchdog_running; static int clocksource_watchdog_kthread(void *data); @@ -254,11 +253,6 @@ static void clocksource_watchdog(unsigned long data) if (!watchdog_running) goto out; - wdnow = watchdog->read(watchdog); - wd_nsec = clocksource_cyc2ns((wdnow - watchdog_last) & watchdog->mask, - watchdog->mult, watchdog->shift); - watchdog_last = wdnow; - list_for_each_entry(cs, &watchdog_list, wd_list) { /* Clocksource already marked unstable? */ @@ -268,19 +262,28 @@ static void clocksource_watchdog(unsigned long data) continue; } + local_irq_disable(); csnow = cs->read(cs); + wdnow = watchdog->read(watchdog); + local_irq_enable(); /* Clocksource initialized ? */ if (!(cs->flags & CLOCK_SOURCE_WATCHDOG)) { cs->flags |= CLOCK_SOURCE_WATCHDOG; - cs->wd_last = csnow; + cs->wd_last = wdnow; + cs->cs_last = csnow; continue; } - /* Check the deviation from the watchdog clocksource. */ - cs_nsec = clocksource_cyc2ns((csnow - cs->wd_last) & + wd_nsec = clocksource_cyc2ns((wdnow - cs->wd_last) & watchdog->mask, + watchdog->mult, watchdog->shift); + + cs_nsec = clocksource_cyc2ns((csnow - cs->cs_last) & cs->mask, cs->mult, cs->shift); - cs->wd_last = csnow; + cs->cs_last = csnow; + cs->wd_last = wdnow; + + /* Check the deviation from the watchdog clocksource. */ if (abs(cs_nsec - wd_nsec) > WATCHDOG_THRESHOLD) { clocksource_unstable(cs, cs_nsec - wd_nsec); continue; @@ -318,7 +321,6 @@ static inline void clocksource_start_watchdog(void) return; init_timer(&watchdog_timer); watchdog_timer.function = clocksource_watchdog; - watchdog_last = watchdog->read(watchdog); watchdog_timer.expires = jiffies + WATCHDOG_INTERVAL; add_timer_on(&watchdog_timer, cpumask_first(cpu_online_mask)); watchdog_running = 1; diff --git a/kernel/time/timekeeping.c b/kernel/time/timekeeping.c index 342408cf68d..2b021b0e850 100644 --- a/kernel/time/timekeeping.c +++ b/kernel/time/timekeeping.c @@ -604,6 +604,12 @@ static struct timespec timekeeping_suspend_time; */ static void __timekeeping_inject_sleeptime(struct timespec *delta) { + if (!timespec_valid(delta)) { + printk(KERN_WARNING "__timekeeping_inject_sleeptime: Invalid " + "sleep delta value!\n"); + return; + } + xtime = timespec_add(xtime, *delta); wall_to_monotonic = timespec_sub(wall_to_monotonic, *delta); total_sleep_time = timespec_add(total_sleep_time, *delta); @@ -686,12 +692,34 @@ static void timekeeping_resume(void) static int timekeeping_suspend(void) { unsigned long flags; + struct timespec delta, delta_delta; + static struct timespec old_delta; read_persistent_clock(&timekeeping_suspend_time); write_seqlock_irqsave(&xtime_lock, flags); timekeeping_forward_now(); timekeeping_suspended = 1; + + /* + * To avoid drift caused by repeated suspend/resumes, + * which each can add ~1 second drift error, + * try to compensate so the difference in system time + * and persistent_clock time stays close to constant. + */ + delta = timespec_sub(xtime, timekeeping_suspend_time); + delta_delta = timespec_sub(delta, old_delta); + if (abs(delta_delta.tv_sec) >= 2) { + /* + * if delta_delta is too large, assume time correction + * has occured and set old_delta to the current delta. + */ + old_delta = delta; + } else { + /* Otherwise try to adjust old_system to compensate */ + timekeeping_suspend_time = + timespec_add(timekeeping_suspend_time, delta_delta); + } write_sequnlock_irqrestore(&xtime_lock, flags); clockevents_notify(CLOCK_EVT_NOTIFY_SUSPEND, NULL); diff --git a/kernel/timer.c b/kernel/timer.c index fd6198692b5..8cff36119e4 100644 --- a/kernel/timer.c +++ b/kernel/timer.c @@ -749,16 +749,15 @@ unsigned long apply_slack(struct timer_list *timer, unsigned long expires) unsigned long expires_limit, mask; int bit; - expires_limit = expires; - if (timer->slack >= 0) { expires_limit = expires + timer->slack; } else { - unsigned long now = jiffies; + long delta = expires - jiffies; + + if (delta < 256) + return expires; - /* No slack, if already expired else auto slack 0.4% */ - if (time_after(expires, now)) - expires_limit = expires + (expires - now)/256; + expires_limit = expires + delta / 256; } mask = expires ^ expires_limit; if (mask == 0) @@ -795,6 +794,8 @@ unsigned long apply_slack(struct timer_list *timer, unsigned long expires) */ int mod_timer(struct timer_list *timer, unsigned long expires) { + expires = apply_slack(timer, expires); + /* * This is a common optimization triggered by the * networking code - if the timer is re-modified @@ -803,8 +804,6 @@ int mod_timer(struct timer_list *timer, unsigned long expires) if (timer_pending(timer) && timer->expires == expires) return 1; - expires = apply_slack(timer, expires); - return __mod_timer(timer, expires, false, TIMER_NOT_PINNED); } EXPORT_SYMBOL(mod_timer); diff --git a/kernel/trace/ftrace.c b/kernel/trace/ftrace.c index 1ee417fcbfa..c3e4575e782 100644 --- a/kernel/trace/ftrace.c +++ b/kernel/trace/ftrace.c @@ -32,7 +32,6 @@ #include <trace/events/sched.h> -#include <asm/ftrace.h> #include <asm/setup.h> #include "trace_output.h" @@ -82,14 +81,14 @@ static int ftrace_disabled __read_mostly; static DEFINE_MUTEX(ftrace_lock); -static struct ftrace_ops ftrace_list_end __read_mostly = -{ +static struct ftrace_ops ftrace_list_end __read_mostly = { .func = ftrace_stub, }; static struct ftrace_ops *ftrace_global_list __read_mostly = &ftrace_list_end; static struct ftrace_ops *ftrace_ops_list __read_mostly = &ftrace_list_end; ftrace_func_t ftrace_trace_function __read_mostly = ftrace_stub; +static ftrace_func_t __ftrace_trace_function_delay __read_mostly = ftrace_stub; ftrace_func_t __ftrace_trace_function __read_mostly = ftrace_stub; ftrace_func_t ftrace_pid_function __read_mostly = ftrace_stub; static struct ftrace_ops global_ops; @@ -148,9 +147,11 @@ void clear_ftrace_function(void) { ftrace_trace_function = ftrace_stub; __ftrace_trace_function = ftrace_stub; + __ftrace_trace_function_delay = ftrace_stub; ftrace_pid_function = ftrace_stub; } +#undef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST #ifndef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST /* * For those archs that do not test ftrace_trace_stop in their @@ -210,7 +211,12 @@ static void update_ftrace_function(void) #ifdef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST ftrace_trace_function = func; #else +#ifdef CONFIG_DYNAMIC_FTRACE + /* do not update till all functions have been modified */ + __ftrace_trace_function_delay = func; +#else __ftrace_trace_function = func; +#endif ftrace_trace_function = ftrace_test_stop_func; #endif } @@ -785,8 +791,7 @@ static void unregister_ftrace_profiler(void) unregister_ftrace_graph(); } #else -static struct ftrace_ops ftrace_profile_ops __read_mostly = -{ +static struct ftrace_ops ftrace_profile_ops __read_mostly = { .func = function_profile_call, }; @@ -806,19 +811,10 @@ ftrace_profile_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { unsigned long val; - char buf[64]; /* big enough to hold a number */ int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; val = !!val; @@ -1182,8 +1178,14 @@ alloc_and_copy_ftrace_hash(int size_bits, struct ftrace_hash *hash) return NULL; } +static void +ftrace_hash_rec_disable(struct ftrace_ops *ops, int filter_hash); +static void +ftrace_hash_rec_enable(struct ftrace_ops *ops, int filter_hash); + static int -ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src) +ftrace_hash_move(struct ftrace_ops *ops, int enable, + struct ftrace_hash **dst, struct ftrace_hash *src) { struct ftrace_func_entry *entry; struct hlist_node *tp, *tn; @@ -1193,9 +1195,16 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src) unsigned long key; int size = src->count; int bits = 0; + int ret; int i; /* + * Remove the current set, update the hash and add + * them back. + */ + ftrace_hash_rec_disable(ops, enable); + + /* * If the new source is empty, just free dst and assign it * the empty_hash. */ @@ -1215,9 +1224,10 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src) if (bits > FTRACE_HASH_MAX_BITS) bits = FTRACE_HASH_MAX_BITS; + ret = -ENOMEM; new_hash = alloc_ftrace_hash(bits); if (!new_hash) - return -ENOMEM; + goto out; size = 1 << src->size_bits; for (i = 0; i < size; i++) { @@ -1236,7 +1246,16 @@ ftrace_hash_move(struct ftrace_hash **dst, struct ftrace_hash *src) rcu_assign_pointer(*dst, new_hash); free_ftrace_hash_rcu(old_hash); - return 0; + ret = 0; + out: + /* + * Enable regardless of ret: + * On success, we enable the new hash. + * On failure, we re-enable the original hash. + */ + ftrace_hash_rec_enable(ops, enable); + + return ret; } /* @@ -1596,6 +1615,12 @@ static int __ftrace_modify_code(void *data) { int *command = data; + /* + * Do not call function tracer while we update the code. + * We are in stop machine, no worrying about races. + */ + function_trace_stop++; + if (*command & FTRACE_ENABLE_CALLS) ftrace_replace_code(1); else if (*command & FTRACE_DISABLE_CALLS) @@ -1609,6 +1634,18 @@ static int __ftrace_modify_code(void *data) else if (*command & FTRACE_STOP_FUNC_RET) ftrace_disable_ftrace_graph_caller(); +#ifndef CONFIG_HAVE_FUNCTION_TRACE_MCOUNT_TEST + /* + * For archs that call ftrace_test_stop_func(), we must + * wait till after we update all the function callers + * before we update the callback. This keeps different + * ops that record different functions from corrupting + * each other. + */ + __ftrace_trace_function = __ftrace_trace_function_delay; +#endif + function_trace_stop--; + return 0; } @@ -1744,10 +1781,36 @@ static cycle_t ftrace_update_time; static unsigned long ftrace_update_cnt; unsigned long ftrace_update_tot_cnt; +static int ops_traces_mod(struct ftrace_ops *ops) +{ + struct ftrace_hash *hash; + + hash = ops->filter_hash; + return !!(!hash || !hash->count); +} + static int ftrace_update_code(struct module *mod) { struct dyn_ftrace *p; cycle_t start, stop; + unsigned long ref = 0; + + /* + * When adding a module, we need to check if tracers are + * currently enabled and if they are set to trace all functions. + * If they are, we need to enable the module functions as well + * as update the reference counts for those function records. + */ + if (mod) { + struct ftrace_ops *ops; + + for (ops = ftrace_ops_list; + ops != &ftrace_list_end; ops = ops->next) { + if (ops->flags & FTRACE_OPS_FL_ENABLED && + ops_traces_mod(ops)) + ref++; + } + } start = ftrace_now(raw_smp_processor_id()); ftrace_update_cnt = 0; @@ -1760,7 +1823,7 @@ static int ftrace_update_code(struct module *mod) p = ftrace_new_addrs; ftrace_new_addrs = p->newlist; - p->flags = 0L; + p->flags = ref; /* * Do the initial record conversion from mcount jump @@ -1783,7 +1846,7 @@ static int ftrace_update_code(struct module *mod) * conversion puts the module to the correct state, thus * passing the ftrace_make_call check. */ - if (ftrace_start_up) { + if (ftrace_start_up && ref) { int failed = __ftrace_replace_code(p, 1); if (failed) { ftrace_bug(failed, p->ip); @@ -2407,10 +2470,9 @@ ftrace_match_module_records(struct ftrace_hash *hash, char *buff, char *mod) */ static int -ftrace_mod_callback(char *func, char *cmd, char *param, int enable) +ftrace_mod_callback(struct ftrace_hash *hash, + char *func, char *cmd, char *param, int enable) { - struct ftrace_ops *ops = &global_ops; - struct ftrace_hash *hash; char *mod; int ret = -EINVAL; @@ -2430,11 +2492,6 @@ ftrace_mod_callback(char *func, char *cmd, char *param, int enable) if (!strlen(mod)) return ret; - if (enable) - hash = ops->filter_hash; - else - hash = ops->notrace_hash; - ret = ftrace_match_module_records(hash, func, mod); if (!ret) ret = -EINVAL; @@ -2740,7 +2797,7 @@ static int ftrace_process_regex(struct ftrace_hash *hash, { char *func, *command, *next = buff; struct ftrace_func_command *p; - int ret; + int ret = -EINVAL; func = strsep(&next, ":"); @@ -2760,7 +2817,7 @@ static int ftrace_process_regex(struct ftrace_hash *hash, mutex_lock(&ftrace_cmd_mutex); list_for_each_entry(p, &ftrace_commands, list) { if (strcmp(p->name, command) == 0) { - ret = p->func(func, command, next, enable); + ret = p->func(hash, func, command, next, enable); goto out_unlock; } } @@ -2857,7 +2914,11 @@ ftrace_set_regex(struct ftrace_ops *ops, unsigned char *buf, int len, ftrace_match_records(hash, buf, len); mutex_lock(&ftrace_lock); - ret = ftrace_hash_move(orig_hash, hash); + ret = ftrace_hash_move(ops, enable, orig_hash, hash); + if (!ret && ops->flags & FTRACE_OPS_FL_ENABLED + && ftrace_enabled) + ftrace_run_update_code(FTRACE_ENABLE_CALLS); + mutex_unlock(&ftrace_lock); mutex_unlock(&ftrace_regex_lock); @@ -3040,18 +3101,12 @@ ftrace_regex_release(struct inode *inode, struct file *file) orig_hash = &iter->ops->notrace_hash; mutex_lock(&ftrace_lock); - /* - * Remove the current set, update the hash and add - * them back. - */ - ftrace_hash_rec_disable(iter->ops, filter_hash); - ret = ftrace_hash_move(orig_hash, iter->hash); - if (!ret) { - ftrace_hash_rec_enable(iter->ops, filter_hash); - if (iter->ops->flags & FTRACE_OPS_FL_ENABLED - && ftrace_enabled) - ftrace_run_update_code(FTRACE_ENABLE_CALLS); - } + ret = ftrace_hash_move(iter->ops, filter_hash, + orig_hash, iter->hash); + if (!ret && (iter->ops->flags & FTRACE_OPS_FL_ENABLED) + && ftrace_enabled) + ftrace_run_update_code(FTRACE_ENABLE_CALLS); + mutex_unlock(&ftrace_lock); } free_ftrace_hash(iter->hash); @@ -3330,6 +3385,7 @@ static int ftrace_process_locs(struct module *mod, { unsigned long *p; unsigned long addr; + unsigned long flags = 0; /* Shut up gcc */ mutex_lock(&ftrace_lock); p = start; @@ -3346,7 +3402,19 @@ static int ftrace_process_locs(struct module *mod, ftrace_record_ip(addr); } + /* + * We only need to disable interrupts on start up + * because we are modifying code that an interrupt + * may execute, and the modification is not atomic. + * But for modules, nothing runs the code we modify + * until we are finished with it, and there's no + * reason to cause large interrupt latencies while we do it. + */ + if (!mod) + local_irq_save(flags); ftrace_update_code(mod); + if (!mod) + local_irq_restore(flags); mutex_unlock(&ftrace_lock); return 0; diff --git a/kernel/trace/ring_buffer.c b/kernel/trace/ring_buffer.c index b0c7aa40794..731201bf4ac 100644 --- a/kernel/trace/ring_buffer.c +++ b/kernel/trace/ring_buffer.c @@ -997,15 +997,21 @@ static int rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer, unsigned nr_pages) { struct buffer_page *bpage, *tmp; - unsigned long addr; LIST_HEAD(pages); unsigned i; WARN_ON(!nr_pages); for (i = 0; i < nr_pages; i++) { + struct page *page; + /* + * __GFP_NORETRY flag makes sure that the allocation fails + * gracefully without invoking oom-killer and the system is + * not destabilized. + */ bpage = kzalloc_node(ALIGN(sizeof(*bpage), cache_line_size()), - GFP_KERNEL, cpu_to_node(cpu_buffer->cpu)); + GFP_KERNEL | __GFP_NORETRY, + cpu_to_node(cpu_buffer->cpu)); if (!bpage) goto free_pages; @@ -1013,10 +1019,11 @@ static int rb_allocate_pages(struct ring_buffer_per_cpu *cpu_buffer, list_add(&bpage->list, &pages); - addr = __get_free_page(GFP_KERNEL); - if (!addr) + page = alloc_pages_node(cpu_to_node(cpu_buffer->cpu), + GFP_KERNEL | __GFP_NORETRY, 0); + if (!page) goto free_pages; - bpage->page = (void *)addr; + bpage->page = page_address(page); rb_init_page(bpage->page); } @@ -1045,7 +1052,7 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu) { struct ring_buffer_per_cpu *cpu_buffer; struct buffer_page *bpage; - unsigned long addr; + struct page *page; int ret; cpu_buffer = kzalloc_node(ALIGN(sizeof(*cpu_buffer), cache_line_size()), @@ -1067,10 +1074,10 @@ rb_allocate_cpu_buffer(struct ring_buffer *buffer, int cpu) rb_check_bpage(cpu_buffer, bpage); cpu_buffer->reader_page = bpage; - addr = __get_free_page(GFP_KERNEL); - if (!addr) + page = alloc_pages_node(cpu_to_node(cpu), GFP_KERNEL, 0); + if (!page) goto fail_free_reader; - bpage->page = (void *)addr; + bpage->page = page_address(page); rb_init_page(bpage->page); INIT_LIST_HEAD(&cpu_buffer->reader_page->list); @@ -1314,7 +1321,6 @@ int ring_buffer_resize(struct ring_buffer *buffer, unsigned long size) unsigned nr_pages, rm_pages, new_pages; struct buffer_page *bpage, *tmp; unsigned long buffer_size; - unsigned long addr; LIST_HEAD(pages); int i, cpu; @@ -1375,16 +1381,24 @@ int ring_buffer_resize(struct ring_buffer *buffer, unsigned long size) for_each_buffer_cpu(buffer, cpu) { for (i = 0; i < new_pages; i++) { + struct page *page; + /* + * __GFP_NORETRY flag makes sure that the allocation + * fails gracefully without invoking oom-killer and + * the system is not destabilized. + */ bpage = kzalloc_node(ALIGN(sizeof(*bpage), cache_line_size()), - GFP_KERNEL, cpu_to_node(cpu)); + GFP_KERNEL | __GFP_NORETRY, + cpu_to_node(cpu)); if (!bpage) goto free_pages; list_add(&bpage->list, &pages); - addr = __get_free_page(GFP_KERNEL); - if (!addr) + page = alloc_pages_node(cpu_to_node(cpu), + GFP_KERNEL | __GFP_NORETRY, 0); + if (!page) goto free_pages; - bpage->page = (void *)addr; + bpage->page = page_address(page); rb_init_page(bpage->page); } } @@ -3730,16 +3744,17 @@ EXPORT_SYMBOL_GPL(ring_buffer_swap_cpu); * Returns: * The page allocated, or NULL on error. */ -void *ring_buffer_alloc_read_page(struct ring_buffer *buffer) +void *ring_buffer_alloc_read_page(struct ring_buffer *buffer, int cpu) { struct buffer_data_page *bpage; - unsigned long addr; + struct page *page; - addr = __get_free_page(GFP_KERNEL); - if (!addr) + page = alloc_pages_node(cpu_to_node(cpu), + GFP_KERNEL | __GFP_NORETRY, 0); + if (!page) return NULL; - bpage = (void *)addr; + bpage = page_address(page); rb_init_page(bpage); @@ -3978,20 +3993,11 @@ rb_simple_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { unsigned long *p = filp->private_data; - char buf[64]; unsigned long val; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; if (val) diff --git a/kernel/trace/ring_buffer_benchmark.c b/kernel/trace/ring_buffer_benchmark.c index 302f8a61463..a5457d577b9 100644 --- a/kernel/trace/ring_buffer_benchmark.c +++ b/kernel/trace/ring_buffer_benchmark.c @@ -106,7 +106,7 @@ static enum event_status read_page(int cpu) int inc; int i; - bpage = ring_buffer_alloc_read_page(buffer); + bpage = ring_buffer_alloc_read_page(buffer, cpu); if (!bpage) return EVENT_DROPPED; diff --git a/kernel/trace/trace.c b/kernel/trace/trace.c index ee9c921d7f2..e5df02c69b1 100644 --- a/kernel/trace/trace.c +++ b/kernel/trace/trace.c @@ -343,26 +343,27 @@ unsigned long trace_flags = TRACE_ITER_PRINT_PARENT | TRACE_ITER_PRINTK | static int trace_stop_count; static DEFINE_SPINLOCK(tracing_start_lock); +static void wakeup_work_handler(struct work_struct *work) +{ + wake_up(&trace_wait); +} + +static DECLARE_DELAYED_WORK(wakeup_work, wakeup_work_handler); + /** * trace_wake_up - wake up tasks waiting for trace input * - * Simply wakes up any task that is blocked on the trace_wait - * queue. These is used with trace_poll for tasks polling the trace. + * Schedules a delayed work to wake up any task that is blocked on the + * trace_wait queue. These is used with trace_poll for tasks polling the + * trace. */ void trace_wake_up(void) { - int cpu; + const unsigned long delay = msecs_to_jiffies(2); if (trace_flags & TRACE_ITER_BLOCK) return; - /* - * The runqueue_is_locked() can fail, but this is the best we - * have for now: - */ - cpu = get_cpu(); - if (!runqueue_is_locked(cpu)) - wake_up(&trace_wait); - put_cpu(); + schedule_delayed_work(&wakeup_work, delay); } static int __init set_buf_size(char *str) @@ -424,6 +425,7 @@ static const char *trace_options[] = { "graph-time", "record-cmd", "overwrite", + "disable_on_free", NULL }; @@ -1191,6 +1193,18 @@ void trace_nowake_buffer_unlock_commit(struct ring_buffer *buffer, } EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit); +void trace_nowake_buffer_unlock_commit_regs(struct ring_buffer *buffer, + struct ring_buffer_event *event, + unsigned long flags, int pc, + struct pt_regs *regs) +{ + ring_buffer_unlock_commit(buffer, event); + + ftrace_trace_stack_regs(buffer, flags, 0, pc, regs); + ftrace_trace_userstack(buffer, flags, pc); +} +EXPORT_SYMBOL_GPL(trace_nowake_buffer_unlock_commit_regs); + void trace_current_buffer_discard_commit(struct ring_buffer *buffer, struct ring_buffer_event *event) { @@ -1234,30 +1248,103 @@ ftrace(struct trace_array *tr, struct trace_array_cpu *data, } #ifdef CONFIG_STACKTRACE + +#define FTRACE_STACK_MAX_ENTRIES (PAGE_SIZE / sizeof(unsigned long)) +struct ftrace_stack { + unsigned long calls[FTRACE_STACK_MAX_ENTRIES]; +}; + +static DEFINE_PER_CPU(struct ftrace_stack, ftrace_stack); +static DEFINE_PER_CPU(int, ftrace_stack_reserve); + static void __ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags, - int skip, int pc) + int skip, int pc, struct pt_regs *regs) { struct ftrace_event_call *call = &event_kernel_stack; struct ring_buffer_event *event; struct stack_entry *entry; struct stack_trace trace; + int use_stack; + int size = FTRACE_STACK_ENTRIES; + + trace.nr_entries = 0; + trace.skip = skip; + + /* + * Since events can happen in NMIs there's no safe way to + * use the per cpu ftrace_stacks. We reserve it and if an interrupt + * or NMI comes in, it will just have to use the default + * FTRACE_STACK_SIZE. + */ + preempt_disable_notrace(); + + use_stack = ++__get_cpu_var(ftrace_stack_reserve); + /* + * We don't need any atomic variables, just a barrier. + * If an interrupt comes in, we don't care, because it would + * have exited and put the counter back to what we want. + * We just need a barrier to keep gcc from moving things + * around. + */ + barrier(); + if (use_stack == 1) { + trace.entries = &__get_cpu_var(ftrace_stack).calls[0]; + trace.max_entries = FTRACE_STACK_MAX_ENTRIES; + + if (regs) + save_stack_trace_regs(regs, &trace); + else + save_stack_trace(&trace); + + if (trace.nr_entries > size) + size = trace.nr_entries; + } else + /* From now on, use_stack is a boolean */ + use_stack = 0; + + size *= sizeof(unsigned long); event = trace_buffer_lock_reserve(buffer, TRACE_STACK, - sizeof(*entry), flags, pc); + sizeof(*entry) + size, flags, pc); if (!event) - return; - entry = ring_buffer_event_data(event); - memset(&entry->caller, 0, sizeof(entry->caller)); + goto out; + entry = ring_buffer_event_data(event); - trace.nr_entries = 0; - trace.max_entries = FTRACE_STACK_ENTRIES; - trace.skip = skip; - trace.entries = entry->caller; + memset(&entry->caller, 0, size); + + if (use_stack) + memcpy(&entry->caller, trace.entries, + trace.nr_entries * sizeof(unsigned long)); + else { + trace.max_entries = FTRACE_STACK_ENTRIES; + trace.entries = entry->caller; + if (regs) + save_stack_trace_regs(regs, &trace); + else + save_stack_trace(&trace); + } + + entry->size = trace.nr_entries; - save_stack_trace(&trace); if (!filter_check_discard(call, entry, buffer, event)) ring_buffer_unlock_commit(buffer, event); + + out: + /* Again, don't let gcc optimize things here */ + barrier(); + __get_cpu_var(ftrace_stack_reserve)--; + preempt_enable_notrace(); + +} + +void ftrace_trace_stack_regs(struct ring_buffer *buffer, unsigned long flags, + int skip, int pc, struct pt_regs *regs) +{ + if (!(trace_flags & TRACE_ITER_STACKTRACE)) + return; + + __ftrace_trace_stack(buffer, flags, skip, pc, regs); } void ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags, @@ -1266,13 +1353,13 @@ void ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags, if (!(trace_flags & TRACE_ITER_STACKTRACE)) return; - __ftrace_trace_stack(buffer, flags, skip, pc); + __ftrace_trace_stack(buffer, flags, skip, pc, NULL); } void __trace_stack(struct trace_array *tr, unsigned long flags, int skip, int pc) { - __ftrace_trace_stack(tr->buffer, flags, skip, pc); + __ftrace_trace_stack(tr->buffer, flags, skip, pc, NULL); } /** @@ -1288,7 +1375,7 @@ void trace_dump_stack(void) local_save_flags(flags); /* skipping 3 traces, seems to get us at the caller of this function */ - __ftrace_trace_stack(global_trace.buffer, flags, 3, preempt_count()); + __ftrace_trace_stack(global_trace.buffer, flags, 3, preempt_count(), NULL); } static DEFINE_PER_CPU(int, user_stack_count); @@ -1536,7 +1623,12 @@ peek_next_entry(struct trace_iterator *iter, int cpu, u64 *ts, ftrace_enable_cpu(); - return event ? ring_buffer_event_data(event) : NULL; + if (event) { + iter->ent_size = ring_buffer_event_length(event); + return ring_buffer_event_data(event); + } + iter->ent_size = 0; + return NULL; } static struct trace_entry * @@ -2051,6 +2143,9 @@ void trace_default_header(struct seq_file *m) { struct trace_iterator *iter = m->private; + if (!(trace_flags & TRACE_ITER_CONTEXT_INFO)) + return; + if (iter->iter_flags & TRACE_FILE_LAT_FMT) { /* print nothing if the buffers are empty */ if (trace_empty(iter)) @@ -2701,20 +2796,11 @@ tracing_ctrl_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { struct trace_array *tr = filp->private_data; - char buf[64]; unsigned long val; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; val = !!val; @@ -2767,7 +2853,7 @@ int tracer_init(struct tracer *t, struct trace_array *tr) return t->init(tr); } -static int tracing_resize_ring_buffer(unsigned long size) +static int __tracing_resize_ring_buffer(unsigned long size) { int ret; @@ -2819,6 +2905,41 @@ static int tracing_resize_ring_buffer(unsigned long size) return ret; } +static ssize_t tracing_resize_ring_buffer(unsigned long size) +{ + int cpu, ret = size; + + mutex_lock(&trace_types_lock); + + tracing_stop(); + + /* disable all cpu buffers */ + for_each_tracing_cpu(cpu) { + if (global_trace.data[cpu]) + atomic_inc(&global_trace.data[cpu]->disabled); + if (max_tr.data[cpu]) + atomic_inc(&max_tr.data[cpu]->disabled); + } + + if (size != global_trace.entries) + ret = __tracing_resize_ring_buffer(size); + + if (ret < 0) + ret = -ENOMEM; + + for_each_tracing_cpu(cpu) { + if (global_trace.data[cpu]) + atomic_dec(&global_trace.data[cpu]->disabled); + if (max_tr.data[cpu]) + atomic_dec(&max_tr.data[cpu]->disabled); + } + + tracing_start(); + mutex_unlock(&trace_types_lock); + + return ret; +} + /** * tracing_update_buffers - used by tracing facility to expand ring buffers @@ -2836,7 +2957,7 @@ int tracing_update_buffers(void) mutex_lock(&trace_types_lock); if (!ring_buffer_expanded) - ret = tracing_resize_ring_buffer(trace_buf_size); + ret = __tracing_resize_ring_buffer(trace_buf_size); mutex_unlock(&trace_types_lock); return ret; @@ -2860,7 +2981,7 @@ static int tracing_set_tracer(const char *buf) mutex_lock(&trace_types_lock); if (!ring_buffer_expanded) { - ret = tracing_resize_ring_buffer(trace_buf_size); + ret = __tracing_resize_ring_buffer(trace_buf_size); if (ret < 0) goto out; ret = 0; @@ -2966,20 +3087,11 @@ tracing_max_lat_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { unsigned long *ptr = filp->private_data; - char buf[64]; unsigned long val; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; *ptr = val * 1000; @@ -3434,67 +3546,54 @@ tracing_entries_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { unsigned long val; - char buf[64]; - int ret, cpu; - - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; + int ret; - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; /* must have at least 1 entry */ if (!val) return -EINVAL; - mutex_lock(&trace_types_lock); - - tracing_stop(); - - /* disable all cpu buffers */ - for_each_tracing_cpu(cpu) { - if (global_trace.data[cpu]) - atomic_inc(&global_trace.data[cpu]->disabled); - if (max_tr.data[cpu]) - atomic_inc(&max_tr.data[cpu]->disabled); - } - /* value is in KB */ val <<= 10; - if (val != global_trace.entries) { - ret = tracing_resize_ring_buffer(val); - if (ret < 0) { - cnt = ret; - goto out; - } - } + ret = tracing_resize_ring_buffer(val); + if (ret < 0) + return ret; *ppos += cnt; - /* If check pages failed, return ENOMEM */ - if (tracing_disabled) - cnt = -ENOMEM; - out: - for_each_tracing_cpu(cpu) { - if (global_trace.data[cpu]) - atomic_dec(&global_trace.data[cpu]->disabled); - if (max_tr.data[cpu]) - atomic_dec(&max_tr.data[cpu]->disabled); - } + return cnt; +} - tracing_start(); - mutex_unlock(&trace_types_lock); +static ssize_t +tracing_free_buffer_write(struct file *filp, const char __user *ubuf, + size_t cnt, loff_t *ppos) +{ + /* + * There is no need to read what the user has written, this function + * is just to make sure that there is no error when "echo" is used + */ + + *ppos += cnt; return cnt; } +static int +tracing_free_buffer_release(struct inode *inode, struct file *filp) +{ + /* disable tracing ? */ + if (trace_flags & TRACE_ITER_STOP_ON_FREE) + tracing_off(); + /* resize the ring buffer to 0 */ + tracing_resize_ring_buffer(0); + + return 0; +} + static int mark_printk(const char *fmt, ...) { int ret; @@ -3640,6 +3739,11 @@ static const struct file_operations tracing_entries_fops = { .llseek = generic_file_llseek, }; +static const struct file_operations tracing_free_buffer_fops = { + .write = tracing_free_buffer_write, + .release = tracing_free_buffer_release, +}; + static const struct file_operations tracing_mark_fops = { .open = tracing_open_generic, .write = tracing_mark_write, @@ -3696,7 +3800,7 @@ tracing_buffers_read(struct file *filp, char __user *ubuf, return 0; if (!info->spare) - info->spare = ring_buffer_alloc_read_page(info->tr->buffer); + info->spare = ring_buffer_alloc_read_page(info->tr->buffer, info->cpu); if (!info->spare) return -ENOMEM; @@ -3853,7 +3957,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, ref->ref = 1; ref->buffer = info->tr->buffer; - ref->page = ring_buffer_alloc_read_page(ref->buffer); + ref->page = ring_buffer_alloc_read_page(ref->buffer, info->cpu); if (!ref->page) { kfree(ref); break; @@ -3862,8 +3966,7 @@ tracing_buffers_splice_read(struct file *file, loff_t *ppos, r = ring_buffer_read_page(ref->buffer, &ref->page, len, info->cpu, 1); if (r < 0) { - ring_buffer_free_read_page(ref->buffer, - ref->page); + ring_buffer_free_read_page(ref->buffer, ref->page); kfree(ref); break; } @@ -4099,19 +4202,10 @@ trace_options_write(struct file *filp, const char __user *ubuf, size_t cnt, { struct trace_option_dentry *topt = filp->private_data; unsigned long val; - char buf[64]; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; if (val != 0 && val != 1) @@ -4159,20 +4253,11 @@ trace_options_core_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { long index = (long)filp->private_data; - char buf[64]; unsigned long val; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; if (val != 0 && val != 1) @@ -4365,6 +4450,9 @@ static __init int tracer_init_debugfs(void) trace_create_file("buffer_size_kb", 0644, d_tracer, &global_trace, &tracing_entries_fops); + trace_create_file("free_buffer", 0644, d_tracer, + &global_trace, &tracing_free_buffer_fops); + trace_create_file("trace_marker", 0220, d_tracer, NULL, &tracing_mark_fops); diff --git a/kernel/trace/trace.h b/kernel/trace/trace.h index 229f8591f61..3f381d0b20a 100644 --- a/kernel/trace/trace.h +++ b/kernel/trace/trace.h @@ -278,6 +278,29 @@ struct tracer { }; +/* Only current can touch trace_recursion */ +#define trace_recursion_inc() do { (current)->trace_recursion++; } while (0) +#define trace_recursion_dec() do { (current)->trace_recursion--; } while (0) + +/* Ring buffer has the 10 LSB bits to count */ +#define trace_recursion_buffer() ((current)->trace_recursion & 0x3ff) + +/* for function tracing recursion */ +#define TRACE_INTERNAL_BIT (1<<11) +#define TRACE_GLOBAL_BIT (1<<12) +/* + * Abuse of the trace_recursion. + * As we need a way to maintain state if we are tracing the function + * graph in irq because we want to trace a particular function that + * was called in irq context but we have irq tracing off. Since this + * can only be modified by current, we can reuse trace_recursion. + */ +#define TRACE_IRQ_BIT (1<<13) + +#define trace_recursion_set(bit) do { (current)->trace_recursion |= (bit); } while (0) +#define trace_recursion_clear(bit) do { (current)->trace_recursion &= ~(bit); } while (0) +#define trace_recursion_test(bit) ((current)->trace_recursion & (bit)) + #define TRACE_PIPE_ALL_CPU -1 int tracer_init(struct tracer *t, struct trace_array *tr); @@ -389,6 +412,9 @@ void update_max_tr_single(struct trace_array *tr, void ftrace_trace_stack(struct ring_buffer *buffer, unsigned long flags, int skip, int pc); +void ftrace_trace_stack_regs(struct ring_buffer *buffer, unsigned long flags, + int skip, int pc, struct pt_regs *regs); + void ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc); @@ -400,6 +426,12 @@ static inline void ftrace_trace_stack(struct ring_buffer *buffer, { } +static inline void ftrace_trace_stack_regs(struct ring_buffer *buffer, + unsigned long flags, int skip, + int pc, struct pt_regs *regs) +{ +} + static inline void ftrace_trace_userstack(struct ring_buffer *buffer, unsigned long flags, int pc) { @@ -507,8 +539,18 @@ static inline int ftrace_graph_addr(unsigned long addr) return 1; for (i = 0; i < ftrace_graph_count; i++) { - if (addr == ftrace_graph_funcs[i]) + if (addr == ftrace_graph_funcs[i]) { + /* + * If no irqs are to be traced, but a set_graph_function + * is set, and called by an interrupt handler, we still + * want to trace it. + */ + if (in_irq()) + trace_recursion_set(TRACE_IRQ_BIT); + else + trace_recursion_clear(TRACE_IRQ_BIT); return 1; + } } return 0; @@ -609,6 +651,7 @@ enum trace_iterator_flags { TRACE_ITER_GRAPH_TIME = 0x80000, TRACE_ITER_RECORD_CMD = 0x100000, TRACE_ITER_OVERWRITE = 0x200000, + TRACE_ITER_STOP_ON_FREE = 0x400000, }; /* @@ -677,6 +720,7 @@ struct event_subsystem { struct dentry *entry; struct event_filter *filter; int nr_events; + int ref_count; }; #define FILTER_PRED_INVALID ((unsigned short)-1) @@ -784,19 +828,4 @@ extern const char *__stop___trace_bprintk_fmt[]; FTRACE_ENTRY(call, struct_name, id, PARAMS(tstruct), PARAMS(print)) #include "trace_entries.h" -/* Only current can touch trace_recursion */ -#define trace_recursion_inc() do { (current)->trace_recursion++; } while (0) -#define trace_recursion_dec() do { (current)->trace_recursion--; } while (0) - -/* Ring buffer has the 10 LSB bits to count */ -#define trace_recursion_buffer() ((current)->trace_recursion & 0x3ff) - -/* for function tracing recursion */ -#define TRACE_INTERNAL_BIT (1<<11) -#define TRACE_GLOBAL_BIT (1<<12) - -#define trace_recursion_set(bit) do { (current)->trace_recursion |= (bit); } while (0) -#define trace_recursion_clear(bit) do { (current)->trace_recursion &= ~(bit); } while (0) -#define trace_recursion_test(bit) ((current)->trace_recursion & (bit)) - #endif /* _LINUX_KERNEL_TRACE_H */ diff --git a/kernel/trace/trace_entries.h b/kernel/trace/trace_entries.h index e32744c84d9..93365907f21 100644 --- a/kernel/trace/trace_entries.h +++ b/kernel/trace/trace_entries.h @@ -161,7 +161,8 @@ FTRACE_ENTRY(kernel_stack, stack_entry, TRACE_STACK, F_STRUCT( - __array( unsigned long, caller, FTRACE_STACK_ENTRIES ) + __field( int, size ) + __dynamic_array(unsigned long, caller ) ), F_printk("\t=> (%08lx)\n\t=> (%08lx)\n\t=> (%08lx)\n\t=> (%08lx)\n" diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c index 686ec399f2a..581876f9f38 100644 --- a/kernel/trace/trace_events.c +++ b/kernel/trace/trace_events.c @@ -244,6 +244,35 @@ static void ftrace_clear_events(void) mutex_unlock(&event_mutex); } +static void __put_system(struct event_subsystem *system) +{ + struct event_filter *filter = system->filter; + + WARN_ON_ONCE(system->ref_count == 0); + if (--system->ref_count) + return; + + if (filter) { + kfree(filter->filter_string); + kfree(filter); + } + kfree(system->name); + kfree(system); +} + +static void __get_system(struct event_subsystem *system) +{ + WARN_ON_ONCE(system->ref_count == 0); + system->ref_count++; +} + +static void put_system(struct event_subsystem *system) +{ + mutex_lock(&event_mutex); + __put_system(system); + mutex_unlock(&event_mutex); +} + /* * __ftrace_set_clr_event(NULL, NULL, NULL, set) will set/unset all events. */ @@ -486,20 +515,11 @@ event_enable_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { struct ftrace_event_call *call = filp->private_data; - char buf[64]; unsigned long val; int ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; ret = tracing_update_buffers(); @@ -528,7 +548,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) { const char set_to_char[4] = { '?', '0', '1', 'X' }; - const char *system = filp->private_data; + struct event_subsystem *system = filp->private_data; struct ftrace_event_call *call; char buf[2]; int set = 0; @@ -539,7 +559,7 @@ system_enable_read(struct file *filp, char __user *ubuf, size_t cnt, if (!call->name || !call->class || !call->class->reg) continue; - if (system && strcmp(call->class->system, system) != 0) + if (system && strcmp(call->class->system, system->name) != 0) continue; /* @@ -569,21 +589,13 @@ static ssize_t system_enable_write(struct file *filp, const char __user *ubuf, size_t cnt, loff_t *ppos) { - const char *system = filp->private_data; + struct event_subsystem *system = filp->private_data; + const char *name = NULL; unsigned long val; - char buf[64]; ssize_t ret; - if (cnt >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, cnt)) - return -EFAULT; - - buf[cnt] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, cnt, 10, &val); + if (ret) return ret; ret = tracing_update_buffers(); @@ -593,7 +605,14 @@ system_enable_write(struct file *filp, const char __user *ubuf, size_t cnt, if (val != 0 && val != 1) return -EINVAL; - ret = __ftrace_set_clr_event(NULL, system, NULL, val); + /* + * Opening of "enable" adds a ref count to system, + * so the name is safe to use. + */ + if (system) + name = system->name; + + ret = __ftrace_set_clr_event(NULL, name, NULL, val); if (ret) goto out; @@ -826,6 +845,52 @@ event_filter_write(struct file *filp, const char __user *ubuf, size_t cnt, return cnt; } +static LIST_HEAD(event_subsystems); + +static int subsystem_open(struct inode *inode, struct file *filp) +{ + struct event_subsystem *system = NULL; + int ret; + + if (!inode->i_private) + goto skip_search; + + /* Make sure the system still exists */ + mutex_lock(&event_mutex); + list_for_each_entry(system, &event_subsystems, list) { + if (system == inode->i_private) { + /* Don't open systems with no events */ + if (!system->nr_events) { + system = NULL; + break; + } + __get_system(system); + break; + } + } + mutex_unlock(&event_mutex); + + if (system != inode->i_private) + return -ENODEV; + + skip_search: + ret = tracing_open_generic(inode, filp); + if (ret < 0 && system) + put_system(system); + + return ret; +} + +static int subsystem_release(struct inode *inode, struct file *file) +{ + struct event_subsystem *system = inode->i_private; + + if (system) + put_system(system); + + return 0; +} + static ssize_t subsystem_filter_read(struct file *filp, char __user *ubuf, size_t cnt, loff_t *ppos) @@ -963,17 +1028,19 @@ static const struct file_operations ftrace_event_filter_fops = { }; static const struct file_operations ftrace_subsystem_filter_fops = { - .open = tracing_open_generic, + .open = subsystem_open, .read = subsystem_filter_read, .write = subsystem_filter_write, .llseek = default_llseek, + .release = subsystem_release, }; static const struct file_operations ftrace_system_enable_fops = { - .open = tracing_open_generic, + .open = subsystem_open, .read = system_enable_read, .write = system_enable_write, .llseek = default_llseek, + .release = subsystem_release, }; static const struct file_operations ftrace_show_header_fops = { @@ -1002,8 +1069,6 @@ static struct dentry *event_trace_events_dir(void) return d_events; } -static LIST_HEAD(event_subsystems); - static struct dentry * event_subsystem_dir(const char *name, struct dentry *d_events) { @@ -1013,6 +1078,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events) /* First see if we did not already create this dir */ list_for_each_entry(system, &event_subsystems, list) { if (strcmp(system->name, name) == 0) { + __get_system(system); system->nr_events++; return system->entry; } @@ -1035,6 +1101,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events) } system->nr_events = 1; + system->ref_count = 1; system->name = kstrdup(name, GFP_KERNEL); if (!system->name) { debugfs_remove(system->entry); @@ -1062,8 +1129,7 @@ event_subsystem_dir(const char *name, struct dentry *d_events) "'%s/filter' entry\n", name); } - trace_create_file("enable", 0644, system->entry, - (void *)system->name, + trace_create_file("enable", 0644, system->entry, system, &ftrace_system_enable_fops); return system->entry; @@ -1184,16 +1250,9 @@ static void remove_subsystem_dir(const char *name) list_for_each_entry(system, &event_subsystems, list) { if (strcmp(system->name, name) == 0) { if (!--system->nr_events) { - struct event_filter *filter = system->filter; - debugfs_remove_recursive(system->entry); list_del(&system->list); - if (filter) { - kfree(filter->filter_string); - kfree(filter); - } - kfree(system->name); - kfree(system); + __put_system(system); } break; } diff --git a/kernel/trace/trace_events_filter.c b/kernel/trace/trace_events_filter.c index 8008ddcfbf2..256764ecccd 100644 --- a/kernel/trace/trace_events_filter.c +++ b/kernel/trace/trace_events_filter.c @@ -1886,6 +1886,12 @@ int apply_subsystem_event_filter(struct event_subsystem *system, mutex_lock(&event_mutex); + /* Make sure the system still has events */ + if (!system->nr_events) { + err = -ENODEV; + goto out_unlock; + } + if (!strcmp(strstrip(filter_string), "0")) { filter_free_subsystem_preds(system); remove_filter_string(system->filter); diff --git a/kernel/trace/trace_functions.c b/kernel/trace/trace_functions.c index 8d0e1cc4e97..c7b0c6a7db0 100644 --- a/kernel/trace/trace_functions.c +++ b/kernel/trace/trace_functions.c @@ -324,7 +324,8 @@ ftrace_trace_onoff_unreg(char *glob, char *cmd, char *param) } static int -ftrace_trace_onoff_callback(char *glob, char *cmd, char *param, int enable) +ftrace_trace_onoff_callback(struct ftrace_hash *hash, + char *glob, char *cmd, char *param, int enable) { struct ftrace_probe_ops *ops; void *count = (void *)-1; diff --git a/kernel/trace/trace_functions_graph.c b/kernel/trace/trace_functions_graph.c index 962cdb24ed8..a7d2a4c653d 100644 --- a/kernel/trace/trace_functions_graph.c +++ b/kernel/trace/trace_functions_graph.c @@ -74,6 +74,20 @@ static struct tracer_flags tracer_flags = { static struct trace_array *graph_array; +/* + * DURATION column is being also used to display IRQ signs, + * following values are used by print_graph_irq and others + * to fill in space into DURATION column. + */ +enum { + DURATION_FILL_FULL = -1, + DURATION_FILL_START = -2, + DURATION_FILL_END = -3, +}; + +static enum print_line_t +print_graph_duration(unsigned long long duration, struct trace_seq *s, + u32 flags); /* Add a function return address to the trace stack on thread info.*/ int @@ -213,7 +227,7 @@ int __trace_graph_entry(struct trace_array *tr, static inline int ftrace_graph_ignore_irqs(void) { - if (!ftrace_graph_skip_irqs) + if (!ftrace_graph_skip_irqs || trace_recursion_test(TRACE_IRQ_BIT)) return 0; return in_irq(); @@ -577,32 +591,6 @@ get_return_for_leaf(struct trace_iterator *iter, return next; } -/* Signal a overhead of time execution to the output */ -static int -print_graph_overhead(unsigned long long duration, struct trace_seq *s, - u32 flags) -{ - /* If duration disappear, we don't need anything */ - if (!(flags & TRACE_GRAPH_PRINT_DURATION)) - return 1; - - /* Non nested entry or return */ - if (duration == -1) - return trace_seq_printf(s, " "); - - if (flags & TRACE_GRAPH_PRINT_OVERHEAD) { - /* Duration exceeded 100 msecs */ - if (duration > 100000ULL) - return trace_seq_printf(s, "! "); - - /* Duration exceeded 10 msecs */ - if (duration > 10000ULL) - return trace_seq_printf(s, "+ "); - } - - return trace_seq_printf(s, " "); -} - static int print_graph_abs_time(u64 t, struct trace_seq *s) { unsigned long usecs_rem; @@ -625,34 +613,36 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, addr >= (unsigned long)__irqentry_text_end) return TRACE_TYPE_UNHANDLED; - /* Absolute time */ - if (flags & TRACE_GRAPH_PRINT_ABS_TIME) { - ret = print_graph_abs_time(iter->ts, s); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - } + if (trace_flags & TRACE_ITER_CONTEXT_INFO) { + /* Absolute time */ + if (flags & TRACE_GRAPH_PRINT_ABS_TIME) { + ret = print_graph_abs_time(iter->ts, s); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + } - /* Cpu */ - if (flags & TRACE_GRAPH_PRINT_CPU) { - ret = print_graph_cpu(s, cpu); - if (ret == TRACE_TYPE_PARTIAL_LINE) - return TRACE_TYPE_PARTIAL_LINE; - } + /* Cpu */ + if (flags & TRACE_GRAPH_PRINT_CPU) { + ret = print_graph_cpu(s, cpu); + if (ret == TRACE_TYPE_PARTIAL_LINE) + return TRACE_TYPE_PARTIAL_LINE; + } - /* Proc */ - if (flags & TRACE_GRAPH_PRINT_PROC) { - ret = print_graph_proc(s, pid); - if (ret == TRACE_TYPE_PARTIAL_LINE) - return TRACE_TYPE_PARTIAL_LINE; - ret = trace_seq_printf(s, " | "); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; + /* Proc */ + if (flags & TRACE_GRAPH_PRINT_PROC) { + ret = print_graph_proc(s, pid); + if (ret == TRACE_TYPE_PARTIAL_LINE) + return TRACE_TYPE_PARTIAL_LINE; + ret = trace_seq_printf(s, " | "); + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; + } } /* No overhead */ - ret = print_graph_overhead(-1, s, flags); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; + ret = print_graph_duration(DURATION_FILL_START, s, flags); + if (ret != TRACE_TYPE_HANDLED) + return ret; if (type == TRACE_GRAPH_ENT) ret = trace_seq_printf(s, "==========>"); @@ -662,9 +652,10 @@ print_graph_irq(struct trace_iterator *iter, unsigned long addr, if (!ret) return TRACE_TYPE_PARTIAL_LINE; - /* Don't close the duration column if haven't one */ - if (flags & TRACE_GRAPH_PRINT_DURATION) - trace_seq_printf(s, " |"); + ret = print_graph_duration(DURATION_FILL_END, s, flags); + if (ret != TRACE_TYPE_HANDLED) + return ret; + ret = trace_seq_printf(s, "\n"); if (!ret) @@ -716,9 +707,49 @@ trace_print_graph_duration(unsigned long long duration, struct trace_seq *s) } static enum print_line_t -print_graph_duration(unsigned long long duration, struct trace_seq *s) +print_graph_duration(unsigned long long duration, struct trace_seq *s, + u32 flags) { - int ret; + int ret = -1; + + if (!(flags & TRACE_GRAPH_PRINT_DURATION) || + !(trace_flags & TRACE_ITER_CONTEXT_INFO)) + return TRACE_TYPE_HANDLED; + + /* No real adata, just filling the column with spaces */ + switch (duration) { + case DURATION_FILL_FULL: + ret = trace_seq_printf(s, " | "); + return ret ? TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE; + case DURATION_FILL_START: + ret = trace_seq_printf(s, " "); + return ret ? TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE; + case DURATION_FILL_END: + ret = trace_seq_printf(s, " |"); + return ret ? TRACE_TYPE_HANDLED : TRACE_TYPE_PARTIAL_LINE; + } + + /* Signal a overhead of time execution to the output */ + if (flags & TRACE_GRAPH_PRINT_OVERHEAD) { + /* Duration exceeded 100 msecs */ + if (duration > 100000ULL) + ret = trace_seq_printf(s, "! "); + /* Duration exceeded 10 msecs */ + else if (duration > 10000ULL) + ret = trace_seq_printf(s, "+ "); + } + + /* + * The -1 means we either did not exceed the duration tresholds + * or we dont want to print out the overhead. Either way we need + * to fill out the space. + */ + if (ret == -1) + ret = trace_seq_printf(s, " "); + + /* Catching here any failure happenned above */ + if (!ret) + return TRACE_TYPE_PARTIAL_LINE; ret = trace_print_graph_duration(duration, s); if (ret != TRACE_TYPE_HANDLED) @@ -767,18 +798,11 @@ print_graph_entry_leaf(struct trace_iterator *iter, cpu_data->enter_funcs[call->depth] = 0; } - /* Overhead */ - ret = print_graph_overhead(duration, s, flags); - if (!ret) + /* Overhead and duration */ + ret = print_graph_duration(duration, s, flags); + if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; - /* Duration */ - if (flags & TRACE_GRAPH_PRINT_DURATION) { - ret = print_graph_duration(duration, s); - if (ret == TRACE_TYPE_PARTIAL_LINE) - return TRACE_TYPE_PARTIAL_LINE; - } - /* Function */ for (i = 0; i < call->depth * TRACE_GRAPH_INDENT; i++) { ret = trace_seq_printf(s, " "); @@ -815,17 +839,10 @@ print_graph_entry_nested(struct trace_iterator *iter, cpu_data->enter_funcs[call->depth] = call->func; } - /* No overhead */ - ret = print_graph_overhead(-1, s, flags); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - /* No time */ - if (flags & TRACE_GRAPH_PRINT_DURATION) { - ret = trace_seq_printf(s, " | "); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - } + ret = print_graph_duration(DURATION_FILL_FULL, s, flags); + if (ret != TRACE_TYPE_HANDLED) + return ret; /* Function */ for (i = 0; i < call->depth * TRACE_GRAPH_INDENT; i++) { @@ -865,6 +882,9 @@ print_graph_prologue(struct trace_iterator *iter, struct trace_seq *s, return TRACE_TYPE_PARTIAL_LINE; } + if (!(trace_flags & TRACE_ITER_CONTEXT_INFO)) + return 0; + /* Absolute time */ if (flags & TRACE_GRAPH_PRINT_ABS_TIME) { ret = print_graph_abs_time(iter->ts, s); @@ -1078,18 +1098,11 @@ print_graph_return(struct ftrace_graph_ret *trace, struct trace_seq *s, if (print_graph_prologue(iter, s, 0, 0, flags)) return TRACE_TYPE_PARTIAL_LINE; - /* Overhead */ - ret = print_graph_overhead(duration, s, flags); - if (!ret) + /* Overhead and duration */ + ret = print_graph_duration(duration, s, flags); + if (ret == TRACE_TYPE_PARTIAL_LINE) return TRACE_TYPE_PARTIAL_LINE; - /* Duration */ - if (flags & TRACE_GRAPH_PRINT_DURATION) { - ret = print_graph_duration(duration, s); - if (ret == TRACE_TYPE_PARTIAL_LINE) - return TRACE_TYPE_PARTIAL_LINE; - } - /* Closing brace */ for (i = 0; i < trace->depth * TRACE_GRAPH_INDENT; i++) { ret = trace_seq_printf(s, " "); @@ -1146,17 +1159,10 @@ print_graph_comment(struct trace_seq *s, struct trace_entry *ent, if (print_graph_prologue(iter, s, 0, 0, flags)) return TRACE_TYPE_PARTIAL_LINE; - /* No overhead */ - ret = print_graph_overhead(-1, s, flags); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - /* No time */ - if (flags & TRACE_GRAPH_PRINT_DURATION) { - ret = trace_seq_printf(s, " | "); - if (!ret) - return TRACE_TYPE_PARTIAL_LINE; - } + ret = print_graph_duration(DURATION_FILL_FULL, s, flags); + if (ret != TRACE_TYPE_HANDLED) + return ret; /* Indentation */ if (depth > 0) @@ -1207,7 +1213,7 @@ print_graph_comment(struct trace_seq *s, struct trace_entry *ent, enum print_line_t -__print_graph_function_flags(struct trace_iterator *iter, u32 flags) +print_graph_function_flags(struct trace_iterator *iter, u32 flags) { struct ftrace_graph_ent_entry *field; struct fgraph_data *data = iter->private; @@ -1270,18 +1276,7 @@ __print_graph_function_flags(struct trace_iterator *iter, u32 flags) static enum print_line_t print_graph_function(struct trace_iterator *iter) { - return __print_graph_function_flags(iter, tracer_flags.val); -} - -enum print_line_t print_graph_function_flags(struct trace_iterator *iter, - u32 flags) -{ - if (trace_flags & TRACE_ITER_LATENCY_FMT) - flags |= TRACE_GRAPH_PRINT_DURATION; - else - flags |= TRACE_GRAPH_PRINT_ABS_TIME; - - return __print_graph_function_flags(iter, flags); + return print_graph_function_flags(iter, tracer_flags.val); } static enum print_line_t @@ -1309,8 +1304,7 @@ static void print_lat_header(struct seq_file *s, u32 flags) seq_printf(s, "#%.*s / _----=> need-resched \n", size, spaces); seq_printf(s, "#%.*s| / _---=> hardirq/softirq \n", size, spaces); seq_printf(s, "#%.*s|| / _--=> preempt-depth \n", size, spaces); - seq_printf(s, "#%.*s||| / _-=> lock-depth \n", size, spaces); - seq_printf(s, "#%.*s|||| / \n", size, spaces); + seq_printf(s, "#%.*s||| / \n", size, spaces); } static void __print_graph_headers_flags(struct seq_file *s, u32 flags) @@ -1329,7 +1323,7 @@ static void __print_graph_headers_flags(struct seq_file *s, u32 flags) if (flags & TRACE_GRAPH_PRINT_PROC) seq_printf(s, " TASK/PID "); if (lat) - seq_printf(s, "|||||"); + seq_printf(s, "||||"); if (flags & TRACE_GRAPH_PRINT_DURATION) seq_printf(s, " DURATION "); seq_printf(s, " FUNCTION CALLS\n"); @@ -1343,7 +1337,7 @@ static void __print_graph_headers_flags(struct seq_file *s, u32 flags) if (flags & TRACE_GRAPH_PRINT_PROC) seq_printf(s, " | | "); if (lat) - seq_printf(s, "|||||"); + seq_printf(s, "||||"); if (flags & TRACE_GRAPH_PRINT_DURATION) seq_printf(s, " | | "); seq_printf(s, " | | | |\n"); @@ -1358,15 +1352,16 @@ void print_graph_headers_flags(struct seq_file *s, u32 flags) { struct trace_iterator *iter = s->private; + if (!(trace_flags & TRACE_ITER_CONTEXT_INFO)) + return; + if (trace_flags & TRACE_ITER_LATENCY_FMT) { /* print nothing if the buffers are empty */ if (trace_empty(iter)) return; print_trace_header(s, iter); - flags |= TRACE_GRAPH_PRINT_DURATION; - } else - flags |= TRACE_GRAPH_PRINT_ABS_TIME; + } __print_graph_headers_flags(s, flags); } diff --git a/kernel/trace/trace_irqsoff.c b/kernel/trace/trace_irqsoff.c index c77424be284..667aa8cc0cf 100644 --- a/kernel/trace/trace_irqsoff.c +++ b/kernel/trace/trace_irqsoff.c @@ -226,7 +226,9 @@ static void irqsoff_trace_close(struct trace_iterator *iter) } #define GRAPH_TRACER_FLAGS (TRACE_GRAPH_PRINT_CPU | \ - TRACE_GRAPH_PRINT_PROC) + TRACE_GRAPH_PRINT_PROC | \ + TRACE_GRAPH_PRINT_ABS_TIME | \ + TRACE_GRAPH_PRINT_DURATION) static enum print_line_t irqsoff_print_line(struct trace_iterator *iter) { diff --git a/kernel/trace/trace_kprobe.c b/kernel/trace/trace_kprobe.c index f925c45f0af..5fb3697bf0e 100644 --- a/kernel/trace/trace_kprobe.c +++ b/kernel/trace/trace_kprobe.c @@ -343,6 +343,14 @@ DEFINE_BASIC_FETCH_FUNCS(deref) DEFINE_FETCH_deref(string) DEFINE_FETCH_deref(string_size) +static __kprobes void update_deref_fetch_param(struct deref_fetch_param *data) +{ + if (CHECK_FETCH_FUNCS(deref, data->orig.fn)) + update_deref_fetch_param(data->orig.data); + else if (CHECK_FETCH_FUNCS(symbol, data->orig.fn)) + update_symbol_cache(data->orig.data); +} + static __kprobes void free_deref_fetch_param(struct deref_fetch_param *data) { if (CHECK_FETCH_FUNCS(deref, data->orig.fn)) @@ -377,6 +385,19 @@ DEFINE_BASIC_FETCH_FUNCS(bitfield) #define fetch_bitfield_string_size NULL static __kprobes void +update_bitfield_fetch_param(struct bitfield_fetch_param *data) +{ + /* + * Don't check the bitfield itself, because this must be the + * last fetch function. + */ + if (CHECK_FETCH_FUNCS(deref, data->orig.fn)) + update_deref_fetch_param(data->orig.data); + else if (CHECK_FETCH_FUNCS(symbol, data->orig.fn)) + update_symbol_cache(data->orig.data); +} + +static __kprobes void free_bitfield_fetch_param(struct bitfield_fetch_param *data) { /* @@ -389,6 +410,7 @@ free_bitfield_fetch_param(struct bitfield_fetch_param *data) free_symbol_cache(data->orig.data); kfree(data); } + /* Default (unsigned long) fetch type */ #define __DEFAULT_FETCH_TYPE(t) u##t #define _DEFAULT_FETCH_TYPE(t) __DEFAULT_FETCH_TYPE(t) @@ -536,6 +558,7 @@ struct probe_arg { /* Flags for trace_probe */ #define TP_FLAG_TRACE 1 #define TP_FLAG_PROFILE 2 +#define TP_FLAG_REGISTERED 4 struct trace_probe { struct list_head list; @@ -555,16 +578,49 @@ struct trace_probe { (sizeof(struct probe_arg) * (n))) -static __kprobes int probe_is_return(struct trace_probe *tp) +static __kprobes int trace_probe_is_return(struct trace_probe *tp) { return tp->rp.handler != NULL; } -static __kprobes const char *probe_symbol(struct trace_probe *tp) +static __kprobes const char *trace_probe_symbol(struct trace_probe *tp) { return tp->symbol ? tp->symbol : "unknown"; } +static __kprobes unsigned long trace_probe_offset(struct trace_probe *tp) +{ + return tp->rp.kp.offset; +} + +static __kprobes bool trace_probe_is_enabled(struct trace_probe *tp) +{ + return !!(tp->flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE)); +} + +static __kprobes bool trace_probe_is_registered(struct trace_probe *tp) +{ + return !!(tp->flags & TP_FLAG_REGISTERED); +} + +static __kprobes bool trace_probe_has_gone(struct trace_probe *tp) +{ + return !!(kprobe_gone(&tp->rp.kp)); +} + +static __kprobes bool trace_probe_within_module(struct trace_probe *tp, + struct module *mod) +{ + int len = strlen(mod->name); + const char *name = trace_probe_symbol(tp); + return strncmp(mod->name, name, len) == 0 && name[len] == ':'; +} + +static __kprobes bool trace_probe_is_on_module(struct trace_probe *tp) +{ + return !!strchr(trace_probe_symbol(tp), ':'); +} + static int register_probe_event(struct trace_probe *tp); static void unregister_probe_event(struct trace_probe *tp); @@ -646,6 +702,16 @@ error: return ERR_PTR(ret); } +static void update_probe_arg(struct probe_arg *arg) +{ + if (CHECK_FETCH_FUNCS(bitfield, arg->fetch.fn)) + update_bitfield_fetch_param(arg->fetch.data); + else if (CHECK_FETCH_FUNCS(deref, arg->fetch.fn)) + update_deref_fetch_param(arg->fetch.data); + else if (CHECK_FETCH_FUNCS(symbol, arg->fetch.fn)) + update_symbol_cache(arg->fetch.data); +} + static void free_probe_arg(struct probe_arg *arg) { if (CHECK_FETCH_FUNCS(bitfield, arg->fetch.fn)) @@ -671,7 +737,7 @@ static void free_trace_probe(struct trace_probe *tp) kfree(tp); } -static struct trace_probe *find_probe_event(const char *event, +static struct trace_probe *find_trace_probe(const char *event, const char *group) { struct trace_probe *tp; @@ -683,13 +749,96 @@ static struct trace_probe *find_probe_event(const char *event, return NULL; } +/* Enable trace_probe - @flag must be TP_FLAG_TRACE or TP_FLAG_PROFILE */ +static int enable_trace_probe(struct trace_probe *tp, int flag) +{ + int ret = 0; + + tp->flags |= flag; + if (trace_probe_is_enabled(tp) && trace_probe_is_registered(tp) && + !trace_probe_has_gone(tp)) { + if (trace_probe_is_return(tp)) + ret = enable_kretprobe(&tp->rp); + else + ret = enable_kprobe(&tp->rp.kp); + } + + return ret; +} + +/* Disable trace_probe - @flag must be TP_FLAG_TRACE or TP_FLAG_PROFILE */ +static void disable_trace_probe(struct trace_probe *tp, int flag) +{ + tp->flags &= ~flag; + if (!trace_probe_is_enabled(tp) && trace_probe_is_registered(tp)) { + if (trace_probe_is_return(tp)) + disable_kretprobe(&tp->rp); + else + disable_kprobe(&tp->rp.kp); + } +} + +/* Internal register function - just handle k*probes and flags */ +static int __register_trace_probe(struct trace_probe *tp) +{ + int i, ret; + + if (trace_probe_is_registered(tp)) + return -EINVAL; + + for (i = 0; i < tp->nr_args; i++) + update_probe_arg(&tp->args[i]); + + /* Set/clear disabled flag according to tp->flag */ + if (trace_probe_is_enabled(tp)) + tp->rp.kp.flags &= ~KPROBE_FLAG_DISABLED; + else + tp->rp.kp.flags |= KPROBE_FLAG_DISABLED; + + if (trace_probe_is_return(tp)) + ret = register_kretprobe(&tp->rp); + else + ret = register_kprobe(&tp->rp.kp); + + if (ret == 0) + tp->flags |= TP_FLAG_REGISTERED; + else { + pr_warning("Could not insert probe at %s+%lu: %d\n", + trace_probe_symbol(tp), trace_probe_offset(tp), ret); + if (ret == -ENOENT && trace_probe_is_on_module(tp)) { + pr_warning("This probe might be able to register after" + "target module is loaded. Continue.\n"); + ret = 0; + } else if (ret == -EILSEQ) { + pr_warning("Probing address(0x%p) is not an " + "instruction boundary.\n", + tp->rp.kp.addr); + ret = -EINVAL; + } + } + + return ret; +} + +/* Internal unregister function - just handle k*probes and flags */ +static void __unregister_trace_probe(struct trace_probe *tp) +{ + if (trace_probe_is_registered(tp)) { + if (trace_probe_is_return(tp)) + unregister_kretprobe(&tp->rp); + else + unregister_kprobe(&tp->rp.kp); + tp->flags &= ~TP_FLAG_REGISTERED; + /* Cleanup kprobe for reuse */ + if (tp->rp.kp.symbol_name) + tp->rp.kp.addr = NULL; + } +} + /* Unregister a trace_probe and probe_event: call with locking probe_lock */ static void unregister_trace_probe(struct trace_probe *tp) { - if (probe_is_return(tp)) - unregister_kretprobe(&tp->rp); - else - unregister_kprobe(&tp->rp.kp); + __unregister_trace_probe(tp); list_del(&tp->list); unregister_probe_event(tp); } @@ -702,41 +851,65 @@ static int register_trace_probe(struct trace_probe *tp) mutex_lock(&probe_lock); - /* register as an event */ - old_tp = find_probe_event(tp->call.name, tp->call.class->system); + /* Delete old (same name) event if exist */ + old_tp = find_trace_probe(tp->call.name, tp->call.class->system); if (old_tp) { - /* delete old event */ unregister_trace_probe(old_tp); free_trace_probe(old_tp); } + + /* Register new event */ ret = register_probe_event(tp); if (ret) { pr_warning("Failed to register probe event(%d)\n", ret); goto end; } - tp->rp.kp.flags |= KPROBE_FLAG_DISABLED; - if (probe_is_return(tp)) - ret = register_kretprobe(&tp->rp); - else - ret = register_kprobe(&tp->rp.kp); - - if (ret) { - pr_warning("Could not insert probe(%d)\n", ret); - if (ret == -EILSEQ) { - pr_warning("Probing address(0x%p) is not an " - "instruction boundary.\n", - tp->rp.kp.addr); - ret = -EINVAL; - } + /* Register k*probe */ + ret = __register_trace_probe(tp); + if (ret < 0) unregister_probe_event(tp); - } else + else list_add_tail(&tp->list, &probe_list); + end: mutex_unlock(&probe_lock); return ret; } +/* Module notifier call back, checking event on the module */ +static int trace_probe_module_callback(struct notifier_block *nb, + unsigned long val, void *data) +{ + struct module *mod = data; + struct trace_probe *tp; + int ret; + + if (val != MODULE_STATE_COMING) + return NOTIFY_DONE; + + /* Update probes on coming module */ + mutex_lock(&probe_lock); + list_for_each_entry(tp, &probe_list, list) { + if (trace_probe_within_module(tp, mod)) { + __unregister_trace_probe(tp); + ret = __register_trace_probe(tp); + if (ret) + pr_warning("Failed to re-register probe %s on" + "%s: %d\n", + tp->call.name, mod->name, ret); + } + } + mutex_unlock(&probe_lock); + + return NOTIFY_DONE; +} + +static struct notifier_block trace_probe_module_nb = { + .notifier_call = trace_probe_module_callback, + .priority = 1 /* Invoked after kprobe module callback */ +}; + /* Split symbol and offset. */ static int split_symbol_offset(char *symbol, unsigned long *offset) { @@ -962,8 +1135,8 @@ static int create_trace_probe(int argc, char **argv) { /* * Argument syntax: - * - Add kprobe: p[:[GRP/]EVENT] KSYM[+OFFS]|KADDR [FETCHARGS] - * - Add kretprobe: r[:[GRP/]EVENT] KSYM[+0] [FETCHARGS] + * - Add kprobe: p[:[GRP/]EVENT] [MOD:]KSYM[+OFFS]|KADDR [FETCHARGS] + * - Add kretprobe: r[:[GRP/]EVENT] [MOD:]KSYM[+0] [FETCHARGS] * Fetch args: * $retval : fetch return value * $stack : fetch stack address @@ -1025,7 +1198,7 @@ static int create_trace_probe(int argc, char **argv) return -EINVAL; } mutex_lock(&probe_lock); - tp = find_probe_event(event, group); + tp = find_trace_probe(event, group); if (!tp) { mutex_unlock(&probe_lock); pr_info("Event %s/%s doesn't exist.\n", group, event); @@ -1144,7 +1317,7 @@ error: return ret; } -static void cleanup_all_probes(void) +static void release_all_trace_probes(void) { struct trace_probe *tp; @@ -1158,7 +1331,6 @@ static void cleanup_all_probes(void) mutex_unlock(&probe_lock); } - /* Probes listing interfaces */ static void *probes_seq_start(struct seq_file *m, loff_t *pos) { @@ -1181,15 +1353,16 @@ static int probes_seq_show(struct seq_file *m, void *v) struct trace_probe *tp = v; int i; - seq_printf(m, "%c", probe_is_return(tp) ? 'r' : 'p'); + seq_printf(m, "%c", trace_probe_is_return(tp) ? 'r' : 'p'); seq_printf(m, ":%s/%s", tp->call.class->system, tp->call.name); if (!tp->symbol) seq_printf(m, " 0x%p", tp->rp.kp.addr); else if (tp->rp.kp.offset) - seq_printf(m, " %s+%u", probe_symbol(tp), tp->rp.kp.offset); + seq_printf(m, " %s+%u", trace_probe_symbol(tp), + tp->rp.kp.offset); else - seq_printf(m, " %s", probe_symbol(tp)); + seq_printf(m, " %s", trace_probe_symbol(tp)); for (i = 0; i < tp->nr_args; i++) seq_printf(m, " %s=%s", tp->args[i].name, tp->args[i].comm); @@ -1209,7 +1382,7 @@ static int probes_open(struct inode *inode, struct file *file) { if ((file->f_mode & FMODE_WRITE) && (file->f_flags & O_TRUNC)) - cleanup_all_probes(); + release_all_trace_probes(); return seq_open(file, &probes_seq_op); } @@ -1397,7 +1570,8 @@ static __kprobes void kprobe_trace_func(struct kprobe *kp, struct pt_regs *regs) store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize); if (!filter_current_check_discard(buffer, call, entry, event)) - trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); + trace_nowake_buffer_unlock_commit_regs(buffer, event, + irq_flags, pc, regs); } /* Kretprobe handler */ @@ -1429,7 +1603,8 @@ static __kprobes void kretprobe_trace_func(struct kretprobe_instance *ri, store_trace_args(sizeof(*entry), tp, regs, (u8 *)&entry[1], dsize); if (!filter_current_check_discard(buffer, call, entry, event)) - trace_nowake_buffer_unlock_commit(buffer, event, irq_flags, pc); + trace_nowake_buffer_unlock_commit_regs(buffer, event, + irq_flags, pc, regs); } /* Event entry printers */ @@ -1511,30 +1686,6 @@ partial: return TRACE_TYPE_PARTIAL_LINE; } -static int probe_event_enable(struct ftrace_event_call *call) -{ - struct trace_probe *tp = (struct trace_probe *)call->data; - - tp->flags |= TP_FLAG_TRACE; - if (probe_is_return(tp)) - return enable_kretprobe(&tp->rp); - else - return enable_kprobe(&tp->rp.kp); -} - -static void probe_event_disable(struct ftrace_event_call *call) -{ - struct trace_probe *tp = (struct trace_probe *)call->data; - - tp->flags &= ~TP_FLAG_TRACE; - if (!(tp->flags & (TP_FLAG_TRACE | TP_FLAG_PROFILE))) { - if (probe_is_return(tp)) - disable_kretprobe(&tp->rp); - else - disable_kprobe(&tp->rp.kp); - } -} - #undef DEFINE_FIELD #define DEFINE_FIELD(type, item, name, is_signed) \ do { \ @@ -1596,7 +1747,7 @@ static int __set_print_fmt(struct trace_probe *tp, char *buf, int len) const char *fmt, *arg; - if (!probe_is_return(tp)) { + if (!trace_probe_is_return(tp)) { fmt = "(%lx)"; arg = "REC->" FIELD_STRING_IP; } else { @@ -1713,49 +1864,25 @@ static __kprobes void kretprobe_perf_func(struct kretprobe_instance *ri, head = this_cpu_ptr(call->perf_events); perf_trace_buf_submit(entry, size, rctx, entry->ret_ip, 1, regs, head); } - -static int probe_perf_enable(struct ftrace_event_call *call) -{ - struct trace_probe *tp = (struct trace_probe *)call->data; - - tp->flags |= TP_FLAG_PROFILE; - - if (probe_is_return(tp)) - return enable_kretprobe(&tp->rp); - else - return enable_kprobe(&tp->rp.kp); -} - -static void probe_perf_disable(struct ftrace_event_call *call) -{ - struct trace_probe *tp = (struct trace_probe *)call->data; - - tp->flags &= ~TP_FLAG_PROFILE; - - if (!(tp->flags & TP_FLAG_TRACE)) { - if (probe_is_return(tp)) - disable_kretprobe(&tp->rp); - else - disable_kprobe(&tp->rp.kp); - } -} #endif /* CONFIG_PERF_EVENTS */ static __kprobes int kprobe_register(struct ftrace_event_call *event, enum trace_reg type) { + struct trace_probe *tp = (struct trace_probe *)event->data; + switch (type) { case TRACE_REG_REGISTER: - return probe_event_enable(event); + return enable_trace_probe(tp, TP_FLAG_TRACE); case TRACE_REG_UNREGISTER: - probe_event_disable(event); + disable_trace_probe(tp, TP_FLAG_TRACE); return 0; #ifdef CONFIG_PERF_EVENTS case TRACE_REG_PERF_REGISTER: - return probe_perf_enable(event); + return enable_trace_probe(tp, TP_FLAG_PROFILE); case TRACE_REG_PERF_UNREGISTER: - probe_perf_disable(event); + disable_trace_probe(tp, TP_FLAG_PROFILE); return 0; #endif } @@ -1805,7 +1932,7 @@ static int register_probe_event(struct trace_probe *tp) /* Initialize ftrace_event_call */ INIT_LIST_HEAD(&call->class->fields); - if (probe_is_return(tp)) { + if (trace_probe_is_return(tp)) { call->event.funcs = &kretprobe_funcs; call->class->define_fields = kretprobe_event_define_fields; } else { @@ -1844,6 +1971,9 @@ static __init int init_kprobe_trace(void) struct dentry *d_tracer; struct dentry *entry; + if (register_module_notifier(&trace_probe_module_nb)) + return -EINVAL; + d_tracer = tracing_init_dentry(); if (!d_tracer) return 0; @@ -1870,8 +2000,12 @@ fs_initcall(init_kprobe_trace); #ifdef CONFIG_FTRACE_STARTUP_TEST -static int kprobe_trace_selftest_target(int a1, int a2, int a3, - int a4, int a5, int a6) +/* + * The "__used" keeps gcc from removing the function symbol + * from the kallsyms table. + */ +static __used int kprobe_trace_selftest_target(int a1, int a2, int a3, + int a4, int a5, int a6) { return a1 + a2 + a3 + a4 + a5 + a6; } @@ -1893,12 +2027,12 @@ static __init int kprobe_trace_self_tests_init(void) warn++; } else { /* Enable trace point */ - tp = find_probe_event("testprobe", KPROBE_EVENT_SYSTEM); + tp = find_trace_probe("testprobe", KPROBE_EVENT_SYSTEM); if (WARN_ON_ONCE(tp == NULL)) { pr_warning("error on getting new probe.\n"); warn++; } else - probe_event_enable(&tp->call); + enable_trace_probe(tp, TP_FLAG_TRACE); } ret = command_trace_probe("r:testprobe2 kprobe_trace_selftest_target " @@ -1908,12 +2042,12 @@ static __init int kprobe_trace_self_tests_init(void) warn++; } else { /* Enable trace point */ - tp = find_probe_event("testprobe2", KPROBE_EVENT_SYSTEM); + tp = find_trace_probe("testprobe2", KPROBE_EVENT_SYSTEM); if (WARN_ON_ONCE(tp == NULL)) { pr_warning("error on getting new probe.\n"); warn++; } else - probe_event_enable(&tp->call); + enable_trace_probe(tp, TP_FLAG_TRACE); } if (warn) @@ -1934,7 +2068,7 @@ static __init int kprobe_trace_self_tests_init(void) } end: - cleanup_all_probes(); + release_all_trace_probes(); if (warn) pr_cont("NG: Some tests are failed. Please check them.\n"); else diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index e37de492a9e..51999309a6c 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -1107,19 +1107,20 @@ static enum print_line_t trace_stack_print(struct trace_iterator *iter, { struct stack_entry *field; struct trace_seq *s = &iter->seq; - int i; + unsigned long *p; + unsigned long *end; trace_assign_type(field, iter->ent); + end = (unsigned long *)((long)iter->ent + iter->ent_size); if (!trace_seq_puts(s, "<stack trace>\n")) goto partial; - for (i = 0; i < FTRACE_STACK_ENTRIES; i++) { - if (!field->caller[i] || (field->caller[i] == ULONG_MAX)) - break; + + for (p = field->caller; p && *p != ULONG_MAX && p < end; p++) { if (!trace_seq_puts(s, " => ")) goto partial; - if (!seq_print_ip_sym(s, field->caller[i], flags)) + if (!seq_print_ip_sym(s, *p, flags)) goto partial; if (!trace_seq_puts(s, "\n")) goto partial; diff --git a/kernel/trace/trace_printk.c b/kernel/trace/trace_printk.c index dff763b7baf..1f06468a10d 100644 --- a/kernel/trace/trace_printk.c +++ b/kernel/trace/trace_printk.c @@ -240,13 +240,10 @@ static const char **find_next(void *v, loff_t *pos) const char **fmt = v; int start_index; - if (!fmt) - fmt = __start___trace_bprintk_fmt + *pos; - start_index = __stop___trace_bprintk_fmt - __start___trace_bprintk_fmt; if (*pos < start_index) - return fmt; + return __start___trace_bprintk_fmt + *pos; return find_next_mod_format(start_index, v, fmt, pos); } diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index f029dd4fd2c..e4a70c0c71b 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -227,7 +227,9 @@ static void wakeup_trace_close(struct trace_iterator *iter) graph_trace_close(iter); } -#define GRAPH_TRACER_FLAGS (TRACE_GRAPH_PRINT_PROC) +#define GRAPH_TRACER_FLAGS (TRACE_GRAPH_PRINT_PROC | \ + TRACE_GRAPH_PRINT_ABS_TIME | \ + TRACE_GRAPH_PRINT_DURATION) static enum print_line_t wakeup_print_line(struct trace_iterator *iter) { diff --git a/kernel/trace/trace_stack.c b/kernel/trace/trace_stack.c index b0b53b8e4c2..77575b386d9 100644 --- a/kernel/trace/trace_stack.c +++ b/kernel/trace/trace_stack.c @@ -156,20 +156,11 @@ stack_max_size_write(struct file *filp, const char __user *ubuf, { long *ptr = filp->private_data; unsigned long val, flags; - char buf[64]; int ret; int cpu; - if (count >= sizeof(buf)) - return -EINVAL; - - if (copy_from_user(&buf, ubuf, count)) - return -EFAULT; - - buf[count] = 0; - - ret = strict_strtoul(buf, 10, &val); - if (ret < 0) + ret = kstrtoul_from_user(ubuf, count, 10, &val); + if (ret) return ret; local_irq_save(flags); diff --git a/kernel/watchdog.c b/kernel/watchdog.c index 3d0c56ad479..36491cd5b7d 100644 --- a/kernel/watchdog.c +++ b/kernel/watchdog.c @@ -200,6 +200,7 @@ static int is_softlockup(unsigned long touch_ts) } #ifdef CONFIG_HARDLOCKUP_DETECTOR + static struct perf_event_attr wd_hw_attr = { .type = PERF_TYPE_HARDWARE, .config = PERF_COUNT_HW_CPU_CYCLES, @@ -209,7 +210,7 @@ static struct perf_event_attr wd_hw_attr = { }; /* Callback function for perf event subsystem */ -static void watchdog_overflow_callback(struct perf_event *event, int nmi, +static void watchdog_overflow_callback(struct perf_event *event, struct perf_sample_data *data, struct pt_regs *regs) { @@ -368,10 +369,11 @@ static int watchdog_nmi_enable(int cpu) if (event != NULL) goto out_enable; - /* Try to register using hardware perf events */ wd_attr = &wd_hw_attr; wd_attr->sample_period = hw_nmi_get_sample_period(watchdog_thresh); - event = perf_event_create_kernel_counter(wd_attr, cpu, NULL, watchdog_overflow_callback); + + /* Try to register using hardware perf events */ + event = perf_event_create_kernel_counter(wd_attr, cpu, NULL, watchdog_overflow_callback, NULL); if (!IS_ERR(event)) { printk(KERN_INFO "NMI watchdog enabled, takes one hw-pmu counter.\n"); goto out_save; diff --git a/kernel/workqueue.c b/kernel/workqueue.c index 0400553f0d0..25fb1b0e53f 100644 --- a/kernel/workqueue.c +++ b/kernel/workqueue.c @@ -221,7 +221,7 @@ typedef unsigned long mayday_mask_t; * per-CPU workqueues: */ struct workqueue_struct { - unsigned int flags; /* I: WQ_* flags */ + unsigned int flags; /* W: WQ_* flags */ union { struct cpu_workqueue_struct __percpu *pcpu; struct cpu_workqueue_struct *single; @@ -240,6 +240,7 @@ struct workqueue_struct { mayday_mask_t mayday_mask; /* cpus requesting rescue */ struct worker *rescuer; /* I: rescue worker */ + int nr_drainers; /* W: drain in progress */ int saved_max_active; /* W: saved cwq max_active */ const char *name; /* I: workqueue name */ #ifdef CONFIG_LOCKDEP @@ -990,7 +991,7 @@ static void __queue_work(unsigned int cpu, struct workqueue_struct *wq, debug_work_activate(work); /* if dying, only works from the same workqueue are allowed */ - if (unlikely(wq->flags & WQ_DYING) && + if (unlikely(wq->flags & WQ_DRAINING) && WARN_ON_ONCE(!is_chained_work(wq))) return; @@ -2381,6 +2382,54 @@ out_unlock: } EXPORT_SYMBOL_GPL(flush_workqueue); +/** + * drain_workqueue - drain a workqueue + * @wq: workqueue to drain + * + * Wait until the workqueue becomes empty. While draining is in progress, + * only chain queueing is allowed. IOW, only currently pending or running + * work items on @wq can queue further work items on it. @wq is flushed + * repeatedly until it becomes empty. The number of flushing is detemined + * by the depth of chaining and should be relatively short. Whine if it + * takes too long. + */ +void drain_workqueue(struct workqueue_struct *wq) +{ + unsigned int flush_cnt = 0; + unsigned int cpu; + + /* + * __queue_work() needs to test whether there are drainers, is much + * hotter than drain_workqueue() and already looks at @wq->flags. + * Use WQ_DRAINING so that queue doesn't have to check nr_drainers. + */ + spin_lock(&workqueue_lock); + if (!wq->nr_drainers++) + wq->flags |= WQ_DRAINING; + spin_unlock(&workqueue_lock); +reflush: + flush_workqueue(wq); + + for_each_cwq_cpu(cpu, wq) { + struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq); + + if (!cwq->nr_active && list_empty(&cwq->delayed_works)) + continue; + + if (++flush_cnt == 10 || + (flush_cnt % 100 == 0 && flush_cnt <= 1000)) + pr_warning("workqueue %s: flush on destruction isn't complete after %u tries\n", + wq->name, flush_cnt); + goto reflush; + } + + spin_lock(&workqueue_lock); + if (!--wq->nr_drainers) + wq->flags &= ~WQ_DRAINING; + spin_unlock(&workqueue_lock); +} +EXPORT_SYMBOL_GPL(drain_workqueue); + static bool start_flush_work(struct work_struct *work, struct wq_barrier *barr, bool wait_executing) { @@ -3009,34 +3058,10 @@ EXPORT_SYMBOL_GPL(__alloc_workqueue_key); */ void destroy_workqueue(struct workqueue_struct *wq) { - unsigned int flush_cnt = 0; unsigned int cpu; - /* - * Mark @wq dying and drain all pending works. Once WQ_DYING is - * set, only chain queueing is allowed. IOW, only currently - * pending or running work items on @wq can queue further work - * items on it. @wq is flushed repeatedly until it becomes empty. - * The number of flushing is detemined by the depth of chaining and - * should be relatively short. Whine if it takes too long. - */ - wq->flags |= WQ_DYING; -reflush: - flush_workqueue(wq); - - for_each_cwq_cpu(cpu, wq) { - struct cpu_workqueue_struct *cwq = get_cwq(cpu, wq); - - if (!cwq->nr_active && list_empty(&cwq->delayed_works)) - continue; - - if (++flush_cnt == 10 || - (flush_cnt % 100 == 0 && flush_cnt <= 1000)) - printk(KERN_WARNING "workqueue %s: flush on " - "destruction isn't complete after %u tries\n", - wq->name, flush_cnt); - goto reflush; - } + /* drain it before proceeding with destruction */ + drain_workqueue(wq); /* * wq list is used to freeze wq, remove from list after |