From 1593baab910da72480d651ea7bf2ce6e3a25a484 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:09:26 +0200 Subject: [PATCH 1/9] sched/debug: Implement consistent task-state printing Currently get_task_state() and task_state_to_char() report different states, create a number of common helpers and unify the reported state space. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- fs/proc/array.c | 15 ++------------- include/linux/sched.h | 26 +++++++++++++++++++------- 2 files changed, 21 insertions(+), 20 deletions(-) diff --git a/fs/proc/array.c b/fs/proc/array.c index 525157ca25cb..01196d3ad452 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -130,19 +130,8 @@ static const char * const task_state_array[] = { static inline const char *get_task_state(struct task_struct *tsk) { - unsigned int state = (tsk->state | tsk->exit_state) & TASK_REPORT; - - /* - * Parked tasks do not run; they sit in __kthread_parkme(). - * Without this check, we would report them as running, which is - * clearly wrong, so we report them as sleeping instead. - */ - if (tsk->state == TASK_PARKED) - state = TASK_INTERRUPTIBLE; - - BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != ARRAY_SIZE(task_state_array)-1); - - return task_state_array[fls(state)]; + BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != ARRAY_SIZE(task_state_array) - 1); + return task_state_array[__get_task_state(tsk)]; } static inline int get_task_umask(struct task_struct *tsk) diff --git a/include/linux/sched.h b/include/linux/sched.h index 92fb8dd5a9e4..163a0b738908 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1243,17 +1243,29 @@ static inline pid_t task_pgrp_nr(struct task_struct *tsk) return task_pgrp_nr_ns(tsk, &init_pid_ns); } -static inline char task_state_to_char(struct task_struct *task) +static inline unsigned int __get_task_state(struct task_struct *tsk) { - const char stat_nam[] = TASK_STATE_TO_CHAR_STR; - unsigned long state = task->state; + unsigned int tsk_state = READ_ONCE(tsk->state); + unsigned int state = (tsk_state | tsk->exit_state) & TASK_REPORT; - state = state ? __ffs(state) + 1 : 0; + if (tsk_state == TASK_PARKED) + state = TASK_INTERRUPTIBLE; - /* Make sure the string lines up properly with the number of task states: */ - BUILD_BUG_ON(sizeof(TASK_STATE_TO_CHAR_STR)-1 != ilog2(TASK_STATE_MAX)+1); + return fls(state); +} - return state < sizeof(stat_nam) - 1 ? stat_nam[state] : '?'; +static inline char __task_state_to_char(unsigned int state) +{ + static const char state_char[] = "RSDTtXZ"; + + BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != sizeof(state_char) - 2); + + return state_char[state]; +} + +static inline char task_state_to_char(struct task_struct *tsk) +{ + return __task_state_to_char(__get_task_state(tsk)); } /** From 92c4bc9f9cd92a8581e36bc5105f03b569f37e36 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:13:36 +0200 Subject: [PATCH 2/9] sched/debug: Convert TASK_state to hex Bit patterns are easier in hex. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- include/linux/sched.h | 28 ++++++++++++++-------------- 1 file changed, 14 insertions(+), 14 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 163a0b738908..69bed5339ffa 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -65,23 +65,23 @@ struct task_group; */ /* Used in tsk->state: */ -#define TASK_RUNNING 0 -#define TASK_INTERRUPTIBLE 1 -#define TASK_UNINTERRUPTIBLE 2 -#define __TASK_STOPPED 4 -#define __TASK_TRACED 8 +#define TASK_RUNNING 0x0000 +#define TASK_INTERRUPTIBLE 0x0001 +#define TASK_UNINTERRUPTIBLE 0x0002 +#define __TASK_STOPPED 0x0004 +#define __TASK_TRACED 0x0008 /* Used in tsk->exit_state: */ -#define EXIT_DEAD 16 -#define EXIT_ZOMBIE 32 +#define EXIT_DEAD 0x0010 +#define EXIT_ZOMBIE 0x0020 #define EXIT_TRACE (EXIT_ZOMBIE | EXIT_DEAD) /* Used in tsk->state again: */ -#define TASK_DEAD 64 -#define TASK_WAKEKILL 128 -#define TASK_WAKING 256 -#define TASK_PARKED 512 -#define TASK_NOLOAD 1024 -#define TASK_NEW 2048 -#define TASK_STATE_MAX 4096 +#define TASK_DEAD 0x0040 +#define TASK_WAKEKILL 0x0080 +#define TASK_WAKING 0x0100 +#define TASK_PARKED 0x0200 +#define TASK_NOLOAD 0x0400 +#define TASK_NEW 0x0800 +#define TASK_STATE_MAX 0x1000 #define TASK_STATE_TO_CHAR_STR "RSDTtXZxKWPNn" From 65d5dc47fe8530e17e318722fb4df676536c2bfc Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:14:08 +0200 Subject: [PATCH 3/9] sched/debug: Remove unused variable Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/debug.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/kernel/sched/debug.c b/kernel/sched/debug.c index 01217fb5a5de..2f93e4a2d9f6 100644 --- a/kernel/sched/debug.c +++ b/kernel/sched/debug.c @@ -466,8 +466,6 @@ static char *task_group_path(struct task_group *tg) } #endif -static const char stat_nam[] = TASK_STATE_TO_CHAR_STR; - static void print_task(struct seq_file *m, struct rq *rq, struct task_struct *p) { From efb40f588b4370ffaeffafbd50f6ff213d954254 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:19:53 +0200 Subject: [PATCH 4/9] sched/tracing: Fix trace_sched_switch task-state printing Convert trace_sched_switch to use the common task-state helpers and fix the "X" and "Z" order, possibly they ended up in the wrong order because TASK_REPORT has them in the wrong order too. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- include/linux/sched.h | 2 +- include/trace/events/sched.h | 18 +++++++++++------- 2 files changed, 12 insertions(+), 8 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index 69bed5339ffa..a2fe636b6825 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -99,7 +99,7 @@ struct task_group; /* get_task_state(): */ #define TASK_REPORT (TASK_RUNNING | TASK_INTERRUPTIBLE | \ TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ - __TASK_TRACED | EXIT_ZOMBIE | EXIT_DEAD) + __TASK_TRACED | EXIT_DEAD | EXIT_ZOMBIE) #define task_is_traced(task) ((task->state & __TASK_TRACED) != 0) diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index ae1409ffe99a..c63e20c9ef24 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -114,7 +114,10 @@ static inline long __trace_sched_switch_state(bool preempt, struct task_struct * * Preemption ignores task state, therefore preempted tasks are always * RUNNING (we will not have dequeued if state != RUNNING). */ - return preempt ? TASK_RUNNING | TASK_STATE_MAX : p->state; + if (preempt) + return TASK_STATE_MAX; + + return __get_task_state(p); } #endif /* CREATE_TRACE_POINTS */ @@ -152,12 +155,13 @@ TRACE_EVENT(sched_switch, TP_printk("prev_comm=%s prev_pid=%d prev_prio=%d prev_state=%s%s ==> next_comm=%s next_pid=%d next_prio=%d", __entry->prev_comm, __entry->prev_pid, __entry->prev_prio, - __entry->prev_state & (TASK_STATE_MAX-1) ? - __print_flags(__entry->prev_state & (TASK_STATE_MAX-1), "|", - { 1, "S"} , { 2, "D" }, { 4, "T" }, { 8, "t" }, - { 16, "Z" }, { 32, "X" }, { 64, "x" }, - { 128, "K" }, { 256, "W" }, { 512, "P" }, - { 1024, "N" }) : "R", + + (__entry->prev_state & TASK_REPORT) ? + __print_flags(__entry->prev_state & TASK_REPORT, "|", + { 0x01, "S" }, { 0x02, "D" }, { 0x04, "T" }, + { 0x08, "t" }, { 0x10, "X" }, { 0x20, "Z" }) : + "R", + __entry->prev_state & TASK_STATE_MAX ? "+" : "", __entry->next_comm, __entry->next_pid, __entry->next_prio) ); From 5f6ad26ea353fdf3dad2328052cbee49e0b9c5b4 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:23:31 +0200 Subject: [PATCH 5/9] sched/tracing: Use common task-state helpers Remove yet another task-state char instance. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- include/linux/sched.h | 2 -- kernel/trace/trace_output.c | 21 ++++++--------------- kernel/trace/trace_sched_wakeup.c | 8 ++++---- 3 files changed, 10 insertions(+), 21 deletions(-) diff --git a/include/linux/sched.h b/include/linux/sched.h index a2fe636b6825..bc7807933415 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -83,8 +83,6 @@ struct task_group; #define TASK_NEW 0x0800 #define TASK_STATE_MAX 0x1000 -#define TASK_STATE_TO_CHAR_STR "RSDTtXZxKWPNn" - /* Convenience macros for the sake of set_current_state: */ #define TASK_KILLABLE (TASK_WAKEKILL | TASK_UNINTERRUPTIBLE) #define TASK_STOPPED (TASK_WAKEKILL | __TASK_STOPPED) diff --git a/kernel/trace/trace_output.c b/kernel/trace/trace_output.c index bac629af2285..c738e764e2a5 100644 --- a/kernel/trace/trace_output.c +++ b/kernel/trace/trace_output.c @@ -656,15 +656,6 @@ int trace_print_lat_context(struct trace_iterator *iter) return !trace_seq_has_overflowed(s); } -static const char state_to_char[] = TASK_STATE_TO_CHAR_STR; - -static int task_state_char(unsigned long state) -{ - int bit = state ? __ffs(state) + 1 : 0; - - return bit < sizeof(state_to_char) - 1 ? state_to_char[bit] : '?'; -} - /** * ftrace_find_event - find a registered event * @type: the type of event to look for @@ -930,8 +921,8 @@ static enum print_line_t trace_ctxwake_print(struct trace_iterator *iter, trace_assign_type(field, iter->ent); - T = task_state_char(field->next_state); - S = task_state_char(field->prev_state); + T = __task_state_to_char(field->next_state); + S = __task_state_to_char(field->prev_state); trace_find_cmdline(field->next_pid, comm); trace_seq_printf(&iter->seq, " %5d:%3d:%c %s [%03d] %5d:%3d:%c %s\n", @@ -966,8 +957,8 @@ static int trace_ctxwake_raw(struct trace_iterator *iter, char S) trace_assign_type(field, iter->ent); if (!S) - S = task_state_char(field->prev_state); - T = task_state_char(field->next_state); + S = __task_state_to_char(field->prev_state); + T = __task_state_to_char(field->next_state); trace_seq_printf(&iter->seq, "%d %d %c %d %d %d %c\n", field->prev_pid, field->prev_prio, @@ -1002,8 +993,8 @@ static int trace_ctxwake_hex(struct trace_iterator *iter, char S) trace_assign_type(field, iter->ent); if (!S) - S = task_state_char(field->prev_state); - T = task_state_char(field->next_state); + S = __task_state_to_char(field->prev_state); + T = __task_state_to_char(field->next_state); SEQ_PUT_HEX_FIELD(s, field->prev_pid); SEQ_PUT_HEX_FIELD(s, field->prev_prio); diff --git a/kernel/trace/trace_sched_wakeup.c b/kernel/trace/trace_sched_wakeup.c index ddec53b67646..0c331978b1a6 100644 --- a/kernel/trace/trace_sched_wakeup.c +++ b/kernel/trace/trace_sched_wakeup.c @@ -397,10 +397,10 @@ tracing_sched_switch_trace(struct trace_array *tr, entry = ring_buffer_event_data(event); entry->prev_pid = prev->pid; entry->prev_prio = prev->prio; - entry->prev_state = prev->state; + entry->prev_state = __get_task_state(prev); entry->next_pid = next->pid; entry->next_prio = next->prio; - entry->next_state = next->state; + entry->next_state = __get_task_state(next); entry->next_cpu = task_cpu(next); if (!call_filter_check_discard(call, entry, buffer, event)) @@ -425,10 +425,10 @@ tracing_sched_wakeup_trace(struct trace_array *tr, entry = ring_buffer_event_data(event); entry->prev_pid = curr->pid; entry->prev_prio = curr->prio; - entry->prev_state = curr->state; + entry->prev_state = __get_task_state(curr); entry->next_pid = wakee->pid; entry->next_prio = wakee->prio; - entry->next_state = wakee->state; + entry->next_state = __get_task_state(wakee); entry->next_cpu = task_cpu(wakee); if (!call_filter_check_discard(call, entry, buffer, event)) From 06eb61844d841d0032a9950ce7f8e783ee49c0d0 Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:30:40 +0200 Subject: [PATCH 6/9] sched/debug: Add explicit TASK_IDLE printing Markus reported that kthreads that idle using TASK_IDLE instead of TASK_INTERRUPTIBLE are reported in as TASK_UNINTERRUPTIBLE and things like htop mark those red. This is undesirable, so add an explicit state for TASK_IDLE. Reported-by: Markus Trippelsdorf Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- fs/proc/array.c | 21 +++++++++++++-------- include/linux/sched.h | 12 ++++++++++-- include/trace/events/sched.h | 7 ++++--- 3 files changed, 27 insertions(+), 13 deletions(-) diff --git a/fs/proc/array.c b/fs/proc/array.c index 01196d3ad452..a120a4549d48 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -119,18 +119,23 @@ static inline void task_name(struct seq_file *m, struct task_struct *p) * simple bit tests. */ static const char * const task_state_array[] = { - "R (running)", /* 0 */ - "S (sleeping)", /* 1 */ - "D (disk sleep)", /* 2 */ - "T (stopped)", /* 4 */ - "t (tracing stop)", /* 8 */ - "X (dead)", /* 16 */ - "Z (zombie)", /* 32 */ + + /* states in TASK_REPORT: */ + "R (running)", /* 0x00 */ + "S (sleeping)", /* 0x01 */ + "D (disk sleep)", /* 0x02 */ + "T (stopped)", /* 0x04 */ + "t (tracing stop)", /* 0x08 */ + "X (dead)", /* 0x10 */ + "Z (zombie)", /* 0x20 */ + + /* states beyond TASK_REPORT: */ + "I (idle)", /* 0x40 */ }; static inline const char *get_task_state(struct task_struct *tsk) { - BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != ARRAY_SIZE(task_state_array) - 1); + BUILD_BUG_ON(1 + ilog2(TASK_REPORT_MAX) != ARRAY_SIZE(task_state_array)); return task_state_array[__get_task_state(tsk)]; } diff --git a/include/linux/sched.h b/include/linux/sched.h index bc7807933415..286fc1117046 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -1241,22 +1241,30 @@ static inline pid_t task_pgrp_nr(struct task_struct *tsk) return task_pgrp_nr_ns(tsk, &init_pid_ns); } +#define TASK_REPORT_IDLE (TASK_REPORT + 1) +#define TASK_REPORT_MAX (TASK_REPORT_IDLE << 1) + static inline unsigned int __get_task_state(struct task_struct *tsk) { unsigned int tsk_state = READ_ONCE(tsk->state); unsigned int state = (tsk_state | tsk->exit_state) & TASK_REPORT; + BUILD_BUG_ON_NOT_POWER_OF_2(TASK_REPORT_MAX); + if (tsk_state == TASK_PARKED) state = TASK_INTERRUPTIBLE; + if (tsk_state == TASK_IDLE) + state = TASK_REPORT_IDLE; + return fls(state); } static inline char __task_state_to_char(unsigned int state) { - static const char state_char[] = "RSDTtXZ"; + static const char state_char[] = "RSDTtXZI"; - BUILD_BUG_ON(1 + ilog2(TASK_REPORT) != sizeof(state_char) - 2); + BUILD_BUG_ON(1 + ilog2(TASK_REPORT_MAX) != sizeof(state_char) - 1); return state_char[state]; } diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index c63e20c9ef24..b371ef8206e1 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -156,10 +156,11 @@ TRACE_EVENT(sched_switch, TP_printk("prev_comm=%s prev_pid=%d prev_prio=%d prev_state=%s%s ==> next_comm=%s next_pid=%d next_prio=%d", __entry->prev_comm, __entry->prev_pid, __entry->prev_prio, - (__entry->prev_state & TASK_REPORT) ? - __print_flags(__entry->prev_state & TASK_REPORT, "|", + (__entry->prev_state & (TASK_REPORT_MAX - 1)) ? + __print_flags(__entry->prev_state & (TASK_REPORT_MAX - 1), "|", { 0x01, "S" }, { 0x02, "D" }, { 0x04, "T" }, - { 0x08, "t" }, { 0x10, "X" }, { 0x20, "Z" }) : + { 0x08, "t" }, { 0x10, "X" }, { 0x20, "Z" }, + { 0x40, "I" }) : "R", __entry->prev_state & TASK_STATE_MAX ? "+" : "", From 5d68cc95fb24b2f58060cc5340dd7402a11f054e Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:32:41 +0200 Subject: [PATCH 7/9] sched/debug: Ignore TASK_IDLE for SysRq-W Markus reported that tasks in TASK_IDLE state are reported by SysRq-W, which results in undesirable clutter. Reported-by: Markus Trippelsdorf Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- kernel/sched/core.c | 24 +++++++++++++++++++++++- 1 file changed, 23 insertions(+), 1 deletion(-) diff --git a/kernel/sched/core.c b/kernel/sched/core.c index 18a6966567da..d17c5da523a0 100644 --- a/kernel/sched/core.c +++ b/kernel/sched/core.c @@ -5166,6 +5166,28 @@ void sched_show_task(struct task_struct *p) put_task_stack(p); } +static inline bool +state_filter_match(unsigned long state_filter, struct task_struct *p) +{ + /* no filter, everything matches */ + if (!state_filter) + return true; + + /* filter, but doesn't match */ + if (!(p->state & state_filter)) + return false; + + /* + * When looking for TASK_UNINTERRUPTIBLE skip TASK_IDLE (allows + * TASK_KILLABLE). + */ + if (state_filter == TASK_UNINTERRUPTIBLE && p->state == TASK_IDLE) + return false; + + return true; +} + + void show_state_filter(unsigned long state_filter) { struct task_struct *g, *p; @@ -5188,7 +5210,7 @@ void show_state_filter(unsigned long state_filter) */ touch_nmi_watchdog(); touch_all_softlockup_watchdogs(); - if (!state_filter || (p->state & state_filter)) + if (state_filter_match(state_filter, p)) sched_show_task(p); } From 8ef9925b02c23e3838d5e593c5cf37984141150f Mon Sep 17 00:00:00 2001 From: Peter Zijlstra Date: Fri, 22 Sep 2017 18:37:28 +0200 Subject: [PATCH 8/9] sched/debug: Add explicit TASK_PARKED printing Currently TASK_PARKED is masqueraded as TASK_INTERRUPTIBLE, give it its own print state because it will not in fact get woken by regular wakeups and is a long-term state. This requires moving TASK_PARKED into the TASK_REPORT mask, and since that latter needs to be a contiguous bitmask, we need to shuffle the bits around a bit. Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: linux-kernel@vger.kernel.org Signed-off-by: Ingo Molnar --- fs/proc/array.c | 3 ++- include/linux/sched.h | 16 +++++++--------- include/trace/events/sched.h | 2 +- 3 files changed, 10 insertions(+), 11 deletions(-) diff --git a/fs/proc/array.c b/fs/proc/array.c index a120a4549d48..77a8eacbe032 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -128,9 +128,10 @@ static const char * const task_state_array[] = { "t (tracing stop)", /* 0x08 */ "X (dead)", /* 0x10 */ "Z (zombie)", /* 0x20 */ + "P (parked)", /* 0x40 */ /* states beyond TASK_REPORT: */ - "I (idle)", /* 0x40 */ + "I (idle)", /* 0x80 */ }; static inline const char *get_task_state(struct task_struct *tsk) diff --git a/include/linux/sched.h b/include/linux/sched.h index 286fc1117046..26a7df4e558c 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h @@ -75,10 +75,10 @@ struct task_group; #define EXIT_ZOMBIE 0x0020 #define EXIT_TRACE (EXIT_ZOMBIE | EXIT_DEAD) /* Used in tsk->state again: */ -#define TASK_DEAD 0x0040 -#define TASK_WAKEKILL 0x0080 -#define TASK_WAKING 0x0100 -#define TASK_PARKED 0x0200 +#define TASK_PARKED 0x0040 +#define TASK_DEAD 0x0080 +#define TASK_WAKEKILL 0x0100 +#define TASK_WAKING 0x0200 #define TASK_NOLOAD 0x0400 #define TASK_NEW 0x0800 #define TASK_STATE_MAX 0x1000 @@ -97,7 +97,8 @@ struct task_group; /* get_task_state(): */ #define TASK_REPORT (TASK_RUNNING | TASK_INTERRUPTIBLE | \ TASK_UNINTERRUPTIBLE | __TASK_STOPPED | \ - __TASK_TRACED | EXIT_DEAD | EXIT_ZOMBIE) + __TASK_TRACED | EXIT_DEAD | EXIT_ZOMBIE | \ + TASK_PARKED) #define task_is_traced(task) ((task->state & __TASK_TRACED) != 0) @@ -1251,9 +1252,6 @@ static inline unsigned int __get_task_state(struct task_struct *tsk) BUILD_BUG_ON_NOT_POWER_OF_2(TASK_REPORT_MAX); - if (tsk_state == TASK_PARKED) - state = TASK_INTERRUPTIBLE; - if (tsk_state == TASK_IDLE) state = TASK_REPORT_IDLE; @@ -1262,7 +1260,7 @@ static inline unsigned int __get_task_state(struct task_struct *tsk) static inline char __task_state_to_char(unsigned int state) { - static const char state_char[] = "RSDTtXZI"; + static const char state_char[] = "RSDTtXZPI"; BUILD_BUG_ON(1 + ilog2(TASK_REPORT_MAX) != sizeof(state_char) - 1); diff --git a/include/trace/events/sched.h b/include/trace/events/sched.h index b371ef8206e1..3c8b7f625670 100644 --- a/include/trace/events/sched.h +++ b/include/trace/events/sched.h @@ -160,7 +160,7 @@ TRACE_EVENT(sched_switch, __print_flags(__entry->prev_state & (TASK_REPORT_MAX - 1), "|", { 0x01, "S" }, { 0x02, "D" }, { 0x04, "T" }, { 0x08, "t" }, { 0x10, "X" }, { 0x20, "Z" }, - { 0x40, "I" }) : + { 0x40, "P" }, { 0x80, "I" }) : "R", __entry->prev_state & TASK_STATE_MAX ? "+" : "", From 5ccba44ba118a5000cccc50076b0344632459779 Mon Sep 17 00:00:00 2001 From: Ethan Zhao Date: Mon, 4 Sep 2017 13:59:34 +0800 Subject: [PATCH 9/9] sched/sysctl: Check user input value of sysctl_sched_time_avg System will hang if user set sysctl_sched_time_avg to 0: [root@XXX ~]# sysctl kernel.sched_time_avg_ms=0 Stack traceback for pid 0 0xffff883f6406c600 0 0 1 3 R 0xffff883f6406cf50 *swapper/3 ffff883f7ccc3ae8 0000000000000018 ffffffff810c4dd0 0000000000000000 0000000000017800 ffff883f7ccc3d78 0000000000000003 ffff883f7ccc3bf8 ffffffff810c4fc9 ffff883f7ccc3c08 00000000810c5043 ffff883f7ccc3c08 Call Trace: [] ? update_group_capacity+0x110/0x200 [] ? update_sd_lb_stats+0x109/0x600 [] ? find_busiest_group+0x47/0x530 [] ? load_balance+0x194/0x900 [] ? update_rq_clock.part.83+0x1a/0xe0 [] ? rebalance_domains+0x152/0x290 [] ? run_rebalance_domains+0xdc/0x1d0 [] ? __do_softirq+0xfb/0x320 [] ? irq_exit+0x125/0x130 [] ? scheduler_ipi+0x97/0x160 [] ? smp_reschedule_interrupt+0x29/0x30 [] ? reschedule_interrupt+0x6e/0x80 [] ? cpuidle_enter_state+0xcc/0x230 [] ? cpuidle_enter_state+0x9c/0x230 [] ? cpuidle_enter+0x17/0x20 [] ? cpu_startup_entry+0x38c/0x420 [] ? start_secondary+0x173/0x1e0 Because divide-by-zero error happens in function: update_group_capacity() update_cpu_capacity() scale_rt_capacity() { ... total = sched_avg_period() + delta; used = div_u64(avg, total); ... } To fix this issue, check user input value of sysctl_sched_time_avg, keep it unchanged when hitting invalid input, and set the minimum limit of sysctl_sched_time_avg to 1 ms. Reported-by: James Puthukattukaran Signed-off-by: Ethan Zhao Signed-off-by: Peter Zijlstra (Intel) Cc: Linus Torvalds Cc: Peter Zijlstra Cc: Thomas Gleixner Cc: efault@gmx.de Cc: ethan.kernel@gmail.com Cc: keescook@chromium.org Cc: mcgrof@kernel.org Cc: Link: http://lkml.kernel.org/r/1504504774-18253-1-git-send-email-ethan.zhao@oracle.com Signed-off-by: Ingo Molnar --- kernel/sysctl.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/kernel/sysctl.c b/kernel/sysctl.c index 6648fbbb8157..423554ad3610 100644 --- a/kernel/sysctl.c +++ b/kernel/sysctl.c @@ -367,7 +367,8 @@ static struct ctl_table kern_table[] = { .data = &sysctl_sched_time_avg, .maxlen = sizeof(unsigned int), .mode = 0644, - .proc_handler = proc_dointvec, + .proc_handler = proc_dointvec_minmax, + .extra1 = &one, }, #ifdef CONFIG_SCHEDSTATS {