diff options
| -rw-r--r-- | Documentation/scheduler/sched-design.txt | 165 | ||||
| -rw-r--r-- | arch/x86/Kconfig | 1 | ||||
| -rw-r--r-- | include/linux/sched.h | 38 | ||||
| -rw-r--r-- | init/Kconfig | 11 | ||||
| -rw-r--r-- | init/main.c | 1 | ||||
| -rw-r--r-- | kernel/Makefile | 2 | ||||
| -rw-r--r-- | kernel/sched.c | 323 | ||||
| -rw-r--r-- | kernel/sched_clock.c | 236 | ||||
| -rw-r--r-- | kernel/sched_debug.c | 7 | ||||
| -rw-r--r-- | kernel/sched_fair.c | 39 | ||||
| -rw-r--r-- | kernel/sched_idletask.c | 2 | ||||
| -rw-r--r-- | kernel/sched_rt.c | 9 |
12 files changed, 428 insertions, 406 deletions
diff --git a/Documentation/scheduler/sched-design.txt b/Documentation/scheduler/sched-design.txt deleted file mode 100644 index 1605bf0cba8b..000000000000 --- a/Documentation/scheduler/sched-design.txt +++ /dev/null | |||
| @@ -1,165 +0,0 @@ | |||
| 1 | Goals, Design and Implementation of the | ||
| 2 | new ultra-scalable O(1) scheduler | ||
| 3 | |||
| 4 | |||
| 5 | This is an edited version of an email Ingo Molnar sent to | ||
| 6 | lkml on 4 Jan 2002. It describes the goals, design, and | ||
| 7 | implementation of Ingo's new ultra-scalable O(1) scheduler. | ||
| 8 | Last Updated: 18 April 2002. | ||
| 9 | |||
| 10 | |||
| 11 | Goal | ||
| 12 | ==== | ||
| 13 | |||
| 14 | The main goal of the new scheduler is to keep all the good things we know | ||
| 15 | and love about the current Linux scheduler: | ||
| 16 | |||
| 17 | - good interactive performance even during high load: if the user | ||
| 18 | types or clicks then the system must react instantly and must execute | ||
| 19 | the user tasks smoothly, even during considerable background load. | ||
| 20 | |||
| 21 | - good scheduling/wakeup performance with 1-2 runnable processes. | ||
| 22 | |||
| 23 | - fairness: no process should stay without any timeslice for any | ||
| 24 | unreasonable amount of time. No process should get an unjustly high | ||
| 25 | amount of CPU time. | ||
| 26 | |||
| 27 | - priorities: less important tasks can be started with lower priority, | ||
| 28 | more important tasks with higher priority. | ||
| 29 | |||
| 30 | - SMP efficiency: no CPU should stay idle if there is work to do. | ||
| 31 | |||
| 32 | - SMP affinity: processes which run on one CPU should stay affine to | ||
| 33 | that CPU. Processes should not bounce between CPUs too frequently. | ||
| 34 | |||
| 35 | - plus additional scheduler features: RT scheduling, CPU binding. | ||
| 36 | |||
| 37 | and the goal is also to add a few new things: | ||
| 38 | |||
| 39 | - fully O(1) scheduling. Are you tired of the recalculation loop | ||
| 40 | blowing the L1 cache away every now and then? Do you think the goodness | ||
| 41 | loop is taking a bit too long to finish if there are lots of runnable | ||
| 42 | processes? This new scheduler takes no prisoners: wakeup(), schedule(), | ||
| 43 | the timer interrupt are all O(1) algorithms. There is no recalculation | ||
| 44 | loop. There is no goodness loop either. | ||
| 45 | |||
| 46 | - 'perfect' SMP scalability. With the new scheduler there is no 'big' | ||
| 47 | runqueue_lock anymore - it's all per-CPU runqueues and locks - two | ||
| 48 | tasks on two separate CPUs can wake up, schedule and context-switch | ||
| 49 | completely in parallel, without any interlocking. All | ||
| 50 | scheduling-relevant data is structured for maximum scalability. | ||
| 51 | |||
| 52 | - better SMP affinity. The old scheduler has a particular weakness that | ||
| 53 | causes the random bouncing of tasks between CPUs if/when higher | ||
| 54 | priority/interactive tasks, this was observed and reported by many | ||
| 55 | people. The reason is that the timeslice recalculation loop first needs | ||
| 56 | every currently running task to consume its timeslice. But when this | ||
| 57 | happens on eg. an 8-way system, then this property starves an | ||
| 58 | increasing number of CPUs from executing any process. Once the last | ||
| 59 | task that has a timeslice left has finished using up that timeslice, | ||
| 60 | the recalculation loop is triggered and other CPUs can start executing | ||
| 61 | tasks again - after having idled around for a number of timer ticks. | ||
| 62 | The more CPUs, the worse this effect. | ||
| 63 | |||
| 64 | Furthermore, this same effect causes the bouncing effect as well: | ||
| 65 | whenever there is such a 'timeslice squeeze' of the global runqueue, | ||
| 66 | idle processors start executing tasks which are not affine to that CPU. | ||
| 67 | (because the affine tasks have finished off their timeslices already.) | ||
| 68 | |||
| 69 | The new scheduler solves this problem by distributing timeslices on a | ||
| 70 | per-CPU basis, without having any global synchronization or | ||
| 71 | recalculation. | ||
| 72 | |||
| 73 | - batch scheduling. A significant proportion of computing-intensive tasks | ||
| 74 | benefit from batch-scheduling, where timeslices are long and processes | ||
| 75 | are roundrobin scheduled. The new scheduler does such batch-scheduling | ||
| 76 | of the lowest priority tasks - so nice +19 jobs will get | ||
| 77 | 'batch-scheduled' automatically. With this scheduler, nice +19 jobs are | ||
| 78 | in essence SCHED_IDLE, from an interactiveness point of view. | ||
| 79 | |||
| 80 | - handle extreme loads more smoothly, without breakdown and scheduling | ||
| 81 | storms. | ||
| 82 | |||
| 83 | - O(1) RT scheduling. For those RT folks who are paranoid about the | ||
| 84 | O(nr_running) property of the goodness loop and the recalculation loop. | ||
| 85 | |||
| 86 | - run fork()ed children before the parent. Andrea has pointed out the | ||
| 87 | advantages of this a few months ago, but patches for this feature | ||
| 88 | do not work with the old scheduler as well as they should, | ||
| 89 | because idle processes often steal the new child before the fork()ing | ||
| 90 | CPU gets to execute it. | ||
| 91 | |||
| 92 | |||
| 93 | Design | ||
| 94 | ====== | ||
| 95 | |||
| 96 | The core of the new scheduler contains the following mechanisms: | ||
| 97 | |||
| 98 | - *two* priority-ordered 'priority arrays' per CPU. There is an 'active' | ||
| 99 | array and an 'expired' array. The active array contains all tasks that | ||
| 100 | are affine to this CPU and have timeslices left. The expired array | ||
| 101 | contains all tasks which have used up their timeslices - but this array | ||
| 102 | is kept sorted as well. The active and expired array is not accessed | ||
| 103 | directly, it's accessed through two pointers in the per-CPU runqueue | ||
| 104 | structure. If all active tasks are used up then we 'switch' the two | ||
| 105 | pointers and from now on the ready-to-go (former-) expired array is the | ||
| 106 | active array - and the empty active array serves as the new collector | ||
| 107 | for expired tasks. | ||
| 108 | |||
| 109 | - there is a 64-bit bitmap cache for array indices. Finding the highest | ||
| 110 | priority task is thus a matter of two x86 BSFL bit-search instructions. | ||
| 111 | |||
| 112 | the split-array solution enables us to have an arbitrary number of active | ||
| 113 | and expired tasks, and the recalculation of timeslices can be done | ||
| 114 | immediately when the timeslice expires. Because the arrays are always | ||
| 115 | access through the pointers in the runqueue, switching the two arrays can | ||
| 116 | be done very quickly. | ||
| 117 | |||
| 118 | this is a hybride priority-list approach coupled with roundrobin | ||
| 119 | scheduling and the array-switch method of distributing timeslices. | ||
| 120 | |||
| 121 | - there is a per-task 'load estimator'. | ||
| 122 | |||
| 123 | one of the toughest things to get right is good interactive feel during | ||
| 124 | heavy system load. While playing with various scheduler variants i found | ||
| 125 | that the best interactive feel is achieved not by 'boosting' interactive | ||
| 126 | tasks, but by 'punishing' tasks that want to use more CPU time than there | ||
| 127 | is available. This method is also much easier to do in an O(1) fashion. | ||
| 128 | |||
| 129 | to establish the actual 'load' the task contributes to the system, a | ||
| 130 | complex-looking but pretty accurate method is used: there is a 4-entry | ||
| 131 | 'history' ringbuffer of the task's activities during the last 4 seconds. | ||
| 132 | This ringbuffer is operated without much overhead. The entries tell the | ||
| 133 | scheduler a pretty accurate load-history of the task: has it used up more | ||
| 134 | CPU time or less during the past N seconds. [the size '4' and the interval | ||
| 135 | of 4x 1 seconds was found by lots of experimentation - this part is | ||
| 136 | flexible and can be changed in both directions.] | ||
| 137 | |||
| 138 | the penalty a task gets for generating more load than the CPU can handle | ||
| 139 | is a priority decrease - there is a maximum amount to this penalty | ||
| 140 | relative to their static priority, so even fully CPU-bound tasks will | ||
| 141 | observe each other's priorities, and will share the CPU accordingly. | ||
| 142 | |||
| 143 | the SMP load-balancer can be extended/switched with additional parallel | ||
| 144 | computing and cache hierarchy concepts: NUMA scheduling, multi-core CPUs | ||
| 145 | can be supported easily by changing the load-balancer. Right now it's | ||
| 146 | tuned for my SMP systems. | ||
| 147 | |||
| 148 | i skipped the prev->mm == next->mm advantage - no workload i know of shows | ||
| 149 | any sensitivity to this. It can be added back by sacrificing O(1) | ||
| 150 | schedule() [the current and one-lower priority list can be searched for a | ||
| 151 | that->mm == current->mm condition], but costs a fair number of cycles | ||
| 152 | during a number of important workloads, so i wanted to avoid this as much | ||
| 153 | as possible. | ||
| 154 | |||
| 155 | - the SMP idle-task startup code was still racy and the new scheduler | ||
| 156 | triggered this. So i streamlined the idle-setup code a bit. We do not call | ||
| 157 | into schedule() before all processors have started up fully and all idle | ||
| 158 | threads are in place. | ||
| 159 | |||
| 160 | - the patch also cleans up a number of aspects of sched.c - moves code | ||
| 161 | into other areas of the kernel where it's appropriate, and simplifies | ||
| 162 | certain code paths and data constructs. As a result, the new scheduler's | ||
| 163 | code is smaller than the old one. | ||
| 164 | |||
| 165 | Ingo | ||
diff --git a/arch/x86/Kconfig b/arch/x86/Kconfig index 845ea2b2d487..bbcafaa160c0 100644 --- a/arch/x86/Kconfig +++ b/arch/x86/Kconfig | |||
| @@ -18,6 +18,7 @@ config X86_64 | |||
| 18 | ### Arch settings | 18 | ### Arch settings |
| 19 | config X86 | 19 | config X86 |
| 20 | def_bool y | 20 | def_bool y |
| 21 | select HAVE_UNSTABLE_SCHED_CLOCK | ||
| 21 | select HAVE_IDE | 22 | select HAVE_IDE |
| 22 | select HAVE_OPROFILE | 23 | select HAVE_OPROFILE |
| 23 | select HAVE_KPROBES | 24 | select HAVE_KPROBES |
diff --git a/include/linux/sched.h b/include/linux/sched.h index 03c238088aee..0c35b0343a76 100644 --- a/include/linux/sched.h +++ b/include/linux/sched.h | |||
| @@ -158,6 +158,8 @@ print_cfs_rq(struct seq_file *m, int cpu, struct cfs_rq *cfs_rq) | |||
| 158 | } | 158 | } |
| 159 | #endif | 159 | #endif |
| 160 | 160 | ||
| 161 | extern unsigned long long time_sync_thresh; | ||
| 162 | |||
| 161 | /* | 163 | /* |
| 162 | * Task state bitmask. NOTE! These bits are also | 164 | * Task state bitmask. NOTE! These bits are also |
| 163 | * encoded in fs/proc/array.c: get_task_state(). | 165 | * encoded in fs/proc/array.c: get_task_state(). |
| @@ -1551,6 +1553,35 @@ static inline int set_cpus_allowed(struct task_struct *p, cpumask_t new_mask) | |||
| 1551 | 1553 | ||
| 1552 | extern unsigned long long sched_clock(void); | 1554 | extern unsigned long long sched_clock(void); |
| 1553 | 1555 | ||
| 1556 | #ifndef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK | ||
| 1557 | static inline void sched_clock_init(void) | ||
| 1558 | { | ||
| 1559 | } | ||
| 1560 | |||
| 1561 | static inline u64 sched_clock_cpu(int cpu) | ||
| 1562 | { | ||
| 1563 | return sched_clock(); | ||
| 1564 | } | ||
| 1565 | |||
| 1566 | static inline void sched_clock_tick(void) | ||
| 1567 | { | ||
| 1568 | } | ||
| 1569 | |||
| 1570 | static inline void sched_clock_idle_sleep_event(void) | ||
| 1571 | { | ||
| 1572 | } | ||
| 1573 | |||
| 1574 | static inline void sched_clock_idle_wakeup_event(u64 delta_ns) | ||
| 1575 | { | ||
| 1576 | } | ||
| 1577 | #else | ||
| 1578 | extern void sched_clock_init(void); | ||
| 1579 | extern u64 sched_clock_cpu(int cpu); | ||
| 1580 | extern void sched_clock_tick(void); | ||
| 1581 | extern void sched_clock_idle_sleep_event(void); | ||
| 1582 | extern void sched_clock_idle_wakeup_event(u64 delta_ns); | ||
| 1583 | #endif | ||
| 1584 | |||
| 1554 | /* | 1585 | /* |
| 1555 | * For kernel-internal use: high-speed (but slightly incorrect) per-cpu | 1586 | * For kernel-internal use: high-speed (but slightly incorrect) per-cpu |
| 1556 | * clock constructed from sched_clock(): | 1587 | * clock constructed from sched_clock(): |
| @@ -1977,6 +2008,11 @@ static inline void clear_tsk_need_resched(struct task_struct *tsk) | |||
| 1977 | clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED); | 2008 | clear_tsk_thread_flag(tsk,TIF_NEED_RESCHED); |
| 1978 | } | 2009 | } |
| 1979 | 2010 | ||
| 2011 | static inline int test_tsk_need_resched(struct task_struct *tsk) | ||
| 2012 | { | ||
| 2013 | return unlikely(test_tsk_thread_flag(tsk,TIF_NEED_RESCHED)); | ||
| 2014 | } | ||
| 2015 | |||
| 1980 | static inline int signal_pending(struct task_struct *p) | 2016 | static inline int signal_pending(struct task_struct *p) |
| 1981 | { | 2017 | { |
| 1982 | return unlikely(test_tsk_thread_flag(p,TIF_SIGPENDING)); | 2018 | return unlikely(test_tsk_thread_flag(p,TIF_SIGPENDING)); |
| @@ -1991,7 +2027,7 @@ static inline int fatal_signal_pending(struct task_struct *p) | |||
| 1991 | 2027 | ||
| 1992 | static inline int need_resched(void) | 2028 | static inline int need_resched(void) |
| 1993 | { | 2029 | { |
| 1994 | return unlikely(test_thread_flag(TIF_NEED_RESCHED)); | 2030 | return unlikely(test_tsk_need_resched(current)); |
| 1995 | } | 2031 | } |
| 1996 | 2032 | ||
| 1997 | /* | 2033 | /* |
diff --git a/init/Kconfig b/init/Kconfig index f0e62e5ce0dc..4c33316743f5 100644 --- a/init/Kconfig +++ b/init/Kconfig | |||
| @@ -316,9 +316,16 @@ config CPUSETS | |||
| 316 | 316 | ||
| 317 | Say N if unsure. | 317 | Say N if unsure. |
| 318 | 318 | ||
| 319 | # | ||
| 320 | # Architectures with an unreliable sched_clock() should select this: | ||
| 321 | # | ||
| 322 | config HAVE_UNSTABLE_SCHED_CLOCK | ||
| 323 | bool | ||
| 324 | |||
| 319 | config GROUP_SCHED | 325 | config GROUP_SCHED |
| 320 | bool "Group CPU scheduler" | 326 | bool "Group CPU scheduler" |
| 321 | default y | 327 | depends on EXPERIMENTAL |
| 328 | default n | ||
| 322 | help | 329 | help |
| 323 | This feature lets CPU scheduler recognize task groups and control CPU | 330 | This feature lets CPU scheduler recognize task groups and control CPU |
| 324 | bandwidth allocation to such task groups. | 331 | bandwidth allocation to such task groups. |
| @@ -326,7 +333,7 @@ config GROUP_SCHED | |||
| 326 | config FAIR_GROUP_SCHED | 333 | config FAIR_GROUP_SCHED |
| 327 | bool "Group scheduling for SCHED_OTHER" | 334 | bool "Group scheduling for SCHED_OTHER" |
| 328 | depends on GROUP_SCHED | 335 | depends on GROUP_SCHED |
| 329 | default y | 336 | default GROUP_SCHED |
| 330 | 337 | ||
| 331 | config RT_GROUP_SCHED | 338 | config RT_GROUP_SCHED |
| 332 | bool "Group scheduling for SCHED_RR/FIFO" | 339 | bool "Group scheduling for SCHED_RR/FIFO" |
diff --git a/init/main.c b/init/main.c index a87d4ca5c36c..ddada7acf363 100644 --- a/init/main.c +++ b/init/main.c | |||
| @@ -602,6 +602,7 @@ asmlinkage void __init start_kernel(void) | |||
| 602 | softirq_init(); | 602 | softirq_init(); |
| 603 | timekeeping_init(); | 603 | timekeeping_init(); |
| 604 | time_init(); | 604 | time_init(); |
| 605 | sched_clock_init(); | ||
| 605 | profile_init(); | 606 | profile_init(); |
| 606 | if (!irqs_disabled()) | 607 | if (!irqs_disabled()) |
| 607 | printk("start_kernel(): bug: interrupts were enabled early\n"); | 608 | printk("start_kernel(): bug: interrupts were enabled early\n"); |
diff --git a/kernel/Makefile b/kernel/Makefile index 188c43223f52..1c9938addb9d 100644 --- a/kernel/Makefile +++ b/kernel/Makefile | |||
| @@ -9,7 +9,7 @@ obj-y = sched.o fork.o exec_domain.o panic.o printk.o profile.o \ | |||
| 9 | rcupdate.o extable.o params.o posix-timers.o \ | 9 | rcupdate.o extable.o params.o posix-timers.o \ |
| 10 | kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \ | 10 | kthread.o wait.o kfifo.o sys_ni.o posix-cpu-timers.o mutex.o \ |
| 11 | hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \ | 11 | hrtimer.o rwsem.o nsproxy.o srcu.o semaphore.o \ |
| 12 | notifier.o ksysfs.o pm_qos_params.o | 12 | notifier.o ksysfs.o pm_qos_params.o sched_clock.o |
| 13 | 13 | ||
| 14 | obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o | 14 | obj-$(CONFIG_SYSCTL_SYSCALL_CHECK) += sysctl_check.o |
| 15 | obj-$(CONFIG_STACKTRACE) += stacktrace.o | 15 | obj-$(CONFIG_STACKTRACE) += stacktrace.o |
diff --git a/kernel/sched.c b/kernel/sched.c index 34bcc5bc120e..58fb8af15776 100644 --- a/kernel/sched.c +++ b/kernel/sched.c | |||
| @@ -75,16 +75,6 @@ | |||
| 75 | #include <asm/irq_regs.h> | 75 | #include <asm/irq_regs.h> |
| 76 | 76 | ||
| 77 | /* | 77 | /* |
| 78 | * Scheduler clock - returns current time in nanosec units. | ||
| 79 | * This is default implementation. | ||
| 80 | * Architectures and sub-architectures can override this. | ||
| 81 | */ | ||
| 82 | unsigned long long __attribute__((weak)) sched_clock(void) | ||
| 83 | { | ||
| 84 | return (unsigned long long)jiffies * (NSEC_PER_SEC / HZ); | ||
| 85 | } | ||
| 86 | |||
| 87 | /* | ||
| 88 | * Convert user-nice values [ -20 ... 0 ... 19 ] | 78 | * Convert user-nice values [ -20 ... 0 ... 19 ] |
| 89 | * to static priority [ MAX_RT_PRIO..MAX_PRIO-1 ], | 79 | * to static priority [ MAX_RT_PRIO..MAX_PRIO-1 ], |
| 90 | * and back. | 80 | * and back. |
| @@ -242,6 +232,12 @@ static void destroy_rt_bandwidth(struct rt_bandwidth *rt_b) | |||
| 242 | } | 232 | } |
| 243 | #endif | 233 | #endif |
| 244 | 234 | ||
| 235 | /* | ||
| 236 | * sched_domains_mutex serializes calls to arch_init_sched_domains, | ||
| 237 | * detach_destroy_domains and partition_sched_domains. | ||
| 238 | */ | ||
| 239 | static DEFINE_MUTEX(sched_domains_mutex); | ||
| 240 | |||
| 245 | #ifdef CONFIG_GROUP_SCHED | 241 | #ifdef CONFIG_GROUP_SCHED |
| 246 | 242 | ||
| 247 | #include <linux/cgroup.h> | 243 | #include <linux/cgroup.h> |
| @@ -308,9 +304,6 @@ static DEFINE_PER_CPU(struct rt_rq, init_rt_rq) ____cacheline_aligned_in_smp; | |||
| 308 | */ | 304 | */ |
| 309 | static DEFINE_SPINLOCK(task_group_lock); | 305 | static DEFINE_SPINLOCK(task_group_lock); |
| 310 | 306 | ||
| 311 | /* doms_cur_mutex serializes access to doms_cur[] array */ | ||
| 312 | static DEFINE_MUTEX(doms_cur_mutex); | ||
| 313 | |||
| 314 | #ifdef CONFIG_FAIR_GROUP_SCHED | 307 | #ifdef CONFIG_FAIR_GROUP_SCHED |
| 315 | #ifdef CONFIG_USER_SCHED | 308 | #ifdef CONFIG_USER_SCHED |
| 316 | # define INIT_TASK_GROUP_LOAD (2*NICE_0_LOAD) | 309 | # define INIT_TASK_GROUP_LOAD (2*NICE_0_LOAD) |
| @@ -318,7 +311,13 @@ static DEFINE_MUTEX(doms_cur_mutex); | |||
| 318 | # define INIT_TASK_GROUP_LOAD NICE_0_LOAD | 311 | # define INIT_TASK_GROUP_LOAD NICE_0_LOAD |
| 319 | #endif | 312 | #endif |
| 320 | 313 | ||
| 314 | /* | ||
| 315 | * A weight of 0, 1 or ULONG_MAX can cause arithmetics problems. | ||
| 316 | * (The default weight is 1024 - so there's no practical | ||
| 317 | * limitation from this.) | ||
| 318 | */ | ||
| 321 | #define MIN_SHARES 2 | 319 | #define MIN_SHARES 2 |
| 320 | #define MAX_SHARES (ULONG_MAX - 1) | ||
| 322 | 321 | ||
| 323 | static int init_task_group_load = INIT_TASK_GROUP_LOAD; | 322 | static int init_task_group_load = INIT_TASK_GROUP_LOAD; |
| 324 | #endif | 323 | #endif |
| @@ -358,21 +357,9 @@ static inline void set_task_rq(struct task_struct *p, unsigned int cpu) | |||
| 358 | #endif | 357 | #endif |
| 359 | } | 358 | } |
| 360 | 359 | ||
| 361 | static inline void lock_doms_cur(void) | ||
| 362 | { | ||
| 363 | mutex_lock(&doms_cur_mutex); | ||
| 364 | } | ||
| 365 | |||
| 366 | static inline void unlock_doms_cur(void) | ||
| 367 | { | ||
| 368 | mutex_unlock(&doms_cur_mutex); | ||
| 369 | } | ||
| 370 | |||
| 371 | #else | 360 | #else |
| 372 | 361 | ||
| 373 | static inline void set_task_rq(struct task_struct *p, unsigned int cpu) { } | 362 | static inline void set_task_rq(struct task_struct *p, unsigned int cpu) { } |
| 374 | static inline void lock_doms_cur(void) { } | ||
| 375 | static inline void unlock_doms_cur(void) { } | ||
| 376 | 363 | ||
| 377 | #endif /* CONFIG_GROUP_SCHED */ | 364 | #endif /* CONFIG_GROUP_SCHED */ |
| 378 | 365 | ||
| @@ -560,13 +547,7 @@ struct rq { | |||
| 560 | unsigned long next_balance; | 547 | unsigned long next_balance; |
| 561 | struct mm_struct *prev_mm; | 548 | struct mm_struct *prev_mm; |
| 562 | 549 | ||
| 563 | u64 clock, prev_clock_raw; | 550 | u64 clock; |
| 564 | s64 clock_max_delta; | ||
| 565 | |||
| 566 | unsigned int clock_warps, clock_overflows, clock_underflows; | ||
| 567 | u64 idle_clock; | ||
| 568 | unsigned int clock_deep_idle_events; | ||
| 569 | u64 tick_timestamp; | ||
| 570 | 551 | ||
| 571 | atomic_t nr_iowait; | 552 | atomic_t nr_iowait; |
| 572 | 553 | ||
| @@ -631,82 +612,6 @@ static inline int cpu_of(struct rq *rq) | |||
| 631 | #endif | 612 | #endif |
| 632 | } | 613 | } |
| 633 | 614 | ||
| 634 | #ifdef CONFIG_NO_HZ | ||
| 635 | static inline bool nohz_on(int cpu) | ||
| 636 | { | ||
| 637 | return tick_get_tick_sched(cpu)->nohz_mode != NOHZ_MODE_INACTIVE; | ||
| 638 | } | ||
| 639 | |||
| 640 | static inline u64 max_skipped_ticks(struct rq *rq) | ||
| 641 | { | ||
| 642 | return nohz_on(cpu_of(rq)) ? jiffies - rq->last_tick_seen + 2 : 1; | ||
| 643 | } | ||
| 644 | |||
| 645 | static inline void update_last_tick_seen(struct rq *rq) | ||
| 646 | { | ||
| 647 | rq->last_tick_seen = jiffies; | ||
| 648 | } | ||
| 649 | #else | ||
| 650 | static inline u64 max_skipped_ticks(struct rq *rq) | ||
| 651 | { | ||
| 652 | return 1; | ||
| 653 | } | ||
| 654 | |||
| 655 | static inline void update_last_tick_seen(struct rq *rq) | ||
| 656 | { | ||
| 657 | } | ||
| 658 | #endif | ||
| 659 | |||
| 660 | /* | ||
| 661 | * Update the per-runqueue clock, as finegrained as the platform can give | ||
| 662 | * us, but without assuming monotonicity, etc.: | ||
| 663 | */ | ||
| 664 | static void __update_rq_clock(struct rq *rq) | ||
| 665 | { | ||
| 666 | u64 prev_raw = rq->prev_clock_raw; | ||
| 667 | u64 now = sched_clock(); | ||
| 668 | s64 delta = now - prev_raw; | ||
| 669 | u64 clock = rq->clock; | ||
| 670 | |||
| 671 | #ifdef CONFIG_SCHED_DEBUG | ||
| 672 | WARN_ON_ONCE(cpu_of(rq) != smp_processor_id()); | ||
| 673 | #endif | ||
| 674 | /* | ||
| 675 | * Protect against sched_clock() occasionally going backwards: | ||
| 676 | */ | ||
| 677 | if (unlikely(delta < 0)) { | ||
| 678 | clock++; | ||
| 679 | rq->clock_warps++; | ||
| 680 | } else { | ||
| 681 | /* | ||
| 682 | * Catch too large forward jumps too: | ||
| 683 | */ | ||
| 684 | u64 max_jump = max_skipped_ticks(rq) * TICK_NSEC; | ||
| 685 | u64 max_time = rq->tick_timestamp + max_jump; | ||
| 686 | |||
| 687 | if (unlikely(clock + delta > max_time)) { | ||
| 688 | if (clock < max_time) | ||
| 689 | clock = max_time; | ||
| 690 | else | ||
| 691 | clock++; | ||
| 692 | rq->clock_overflows++; | ||
| 693 | } else { | ||
| 694 | if (unlikely(delta > rq->clock_max_delta)) | ||
| 695 | rq->clock_max_delta = delta; | ||
| 696 | clock += delta; | ||
| 697 | } | ||
| 698 | } | ||
| 699 | |||
| 700 | rq->prev_clock_raw = now; | ||
| 701 | rq->clock = clock; | ||
| 702 | } | ||
| 703 | |||
| 704 | static void update_rq_clock(struct rq *rq) | ||
| 705 | { | ||
| 706 | if (likely(smp_processor_id() == cpu_of(rq))) | ||
| 707 | __update_rq_clock(rq); | ||
| 708 | } | ||
| 709 | |||
| 710 | /* | 615 | /* |
| 711 | * The domain tree (rq->sd) is protected by RCU's quiescent state transition. | 616 | * The domain tree (rq->sd) is protected by RCU's quiescent state transition. |
| 712 | * See detach_destroy_domains: synchronize_sched for details. | 617 | * See detach_destroy_domains: synchronize_sched for details. |
| @@ -722,6 +627,11 @@ static void update_rq_clock(struct rq *rq) | |||
| 722 | #define task_rq(p) cpu_rq(task_cpu(p)) | 627 | #define task_rq(p) cpu_rq(task_cpu(p)) |
| 723 | #define cpu_curr(cpu) (cpu_rq(cpu)->curr) | 628 | #define cpu_curr(cpu) (cpu_rq(cpu)->curr) |
| 724 | 629 | ||
| 630 | static inline void update_rq_clock(struct rq *rq) | ||
| 631 | { | ||
| 632 | rq->clock = sched_clock_cpu(cpu_of(rq)); | ||
| 633 | } | ||
| 634 | |||
| 725 | /* | 635 | /* |
| 726 | * Tunables that become constants when CONFIG_SCHED_DEBUG is off: | 636 | * Tunables that become constants when CONFIG_SCHED_DEBUG is off: |
| 727 | */ | 637 | */ |
| @@ -757,14 +667,14 @@ const_debug unsigned int sysctl_sched_features = | |||
| 757 | #define SCHED_FEAT(name, enabled) \ | 667 | #define SCHED_FEAT(name, enabled) \ |
| 758 | #name , | 668 | #name , |
| 759 | 669 | ||
| 760 | __read_mostly char *sched_feat_names[] = { | 670 | static __read_mostly char *sched_feat_names[] = { |
| 761 | #include "sched_features.h" | 671 | #include "sched_features.h" |
| 762 | NULL | 672 | NULL |
| 763 | }; | 673 | }; |
| 764 | 674 | ||
| 765 | #undef SCHED_FEAT | 675 | #undef SCHED_FEAT |
| 766 | 676 | ||
| 767 | int sched_feat_open(struct inode *inode, struct file *filp) | 677 | static int sched_feat_open(struct inode *inode, struct file *filp) |
| 768 | { | 678 | { |
| 769 | filp->private_data = inode->i_private; | 679 | filp->private_data = inode->i_private; |
| 770 | return 0; | 680 | return 0; |
| @@ -899,7 +809,7 @@ static inline u64 global_rt_runtime(void) | |||
| 899 | return (u64)sysctl_sched_rt_runtime * NSEC_PER_USEC; | 809 | return (u64)sysctl_sched_rt_runtime * NSEC_PER_USEC; |
| 900 | } | 810 | } |
| 901 | 811 | ||
| 902 | static const unsigned long long time_sync_thresh = 100000; | 812 | unsigned long long time_sync_thresh = 100000; |
| 903 | 813 | ||
| 904 | static DEFINE_PER_CPU(unsigned long long, time_offset); | 814 | static DEFINE_PER_CPU(unsigned long long, time_offset); |
| 905 | static DEFINE_PER_CPU(unsigned long long, prev_cpu_time); | 815 | static DEFINE_PER_CPU(unsigned long long, prev_cpu_time); |
| @@ -913,11 +823,14 @@ static DEFINE_PER_CPU(unsigned long long, prev_cpu_time); | |||
| 913 | static DEFINE_SPINLOCK(time_sync_lock); | 823 | static DEFINE_SPINLOCK(time_sync_lock); |
| 914 | static unsigned long long prev_global_time; | 824 | static unsigned long long prev_global_time; |
| 915 | 825 | ||
| 916 | static unsigned long long __sync_cpu_clock(cycles_t time, int cpu) | 826 | static unsigned long long __sync_cpu_clock(unsigned long long time, int cpu) |
| 917 | { | 827 | { |
| 918 | unsigned long flags; | 828 | /* |
| 919 | 829 | * We want this inlined, to not get tracer function calls | |
| 920 | spin_lock_irqsave(&time_sync_lock, flags); | 830 | * in this critical section: |
| 831 | */ | ||
| 832 | spin_acquire(&time_sync_lock.dep_map, 0, 0, _THIS_IP_); | ||
| 833 | __raw_spin_lock(&time_sync_lock.raw_lock); | ||
| 921 | 834 | ||
| 922 | if (time < prev_global_time) { | 835 | if (time < prev_global_time) { |
| 923 | per_cpu(time_offset, cpu) += prev_global_time - time; | 836 | per_cpu(time_offset, cpu) += prev_global_time - time; |
| @@ -926,7 +839,8 @@ static unsigned long long __sync_cpu_clock(cycles_t time, int cpu) | |||
| 926 | prev_global_time = time; | 839 | prev_global_time = time; |
| 927 | } | 840 | } |
| 928 | 841 | ||
| 929 | spin_unlock_irqrestore(&time_sync_lock, flags); | 842 | __raw_spin_unlock(&time_sync_lock.raw_lock); |
| 843 | spin_release(&time_sync_lock.dep_map, 1, _THIS_IP_); | ||
| 930 | 844 | ||
| 931 | return time; | 845 | return time; |
| 932 | } | 846 | } |
| @@ -934,8 +848,6 @@ static unsigned long long __sync_cpu_clock(cycles_t time, int cpu) | |||
| 934 | static unsigned long long __cpu_clock(int cpu) | 848 | static unsigned long long __cpu_clock(int cpu) |
| 935 | { | 849 | { |
| 936 | unsigned long long now; | 850 | unsigned long long now; |
| 937 | unsigned long flags; | ||
| 938 | struct rq *rq; | ||
| 939 | 851 | ||
| 940 | /* | 852 | /* |
| 941 | * Only call sched_clock() if the scheduler has already been | 853 | * Only call sched_clock() if the scheduler has already been |
| @@ -944,11 +856,7 @@ static unsigned long long __cpu_clock(int cpu) | |||
| 944 | if (unlikely(!scheduler_running)) | 856 | if (unlikely(!scheduler_running)) |
| 945 | return 0; | 857 | return 0; |
| 946 | 858 | ||
| 947 | local_irq_save(flags); | 859 | now = sched_clock_cpu(cpu); |
| 948 | rq = cpu_rq(cpu); | ||
| 949 | update_rq_clock(rq); | ||
| 950 | now = rq->clock; | ||
| 951 | local_irq_restore(flags); | ||
| 952 | 860 | ||
| 953 | return now; | 861 | return now; |
| 954 | } | 862 | } |
| @@ -960,13 +868,18 @@ static unsigned long long __cpu_clock(int cpu) | |||
| 960 | unsigned long long cpu_clock(int cpu) | 868 | unsigned long long cpu_clock(int cpu) |
| 961 | { | 869 | { |
| 962 | unsigned long long prev_cpu_time, time, delta_time; | 870 | unsigned long long prev_cpu_time, time, delta_time; |
| 871 | unsigned long flags; | ||
| 963 | 872 | ||
| 873 | local_irq_save(flags); | ||
| 964 | prev_cpu_time = per_cpu(prev_cpu_time, cpu); | 874 | prev_cpu_time = per_cpu(prev_cpu_time, cpu); |
| 965 | time = __cpu_clock(cpu) + per_cpu(time_offset, cpu); | 875 | time = __cpu_clock(cpu) + per_cpu(time_offset, cpu); |
| 966 | delta_time = time-prev_cpu_time; | 876 | delta_time = time-prev_cpu_time; |
| 967 | 877 | ||
| 968 | if (unlikely(delta_time > time_sync_thresh)) | 878 | if (unlikely(delta_time > time_sync_thresh)) { |
| 969 | time = __sync_cpu_clock(time, cpu); | 879 | time = __sync_cpu_clock(time, cpu); |
| 880 | per_cpu(prev_cpu_time, cpu) = time; | ||
| 881 | } | ||
| 882 | local_irq_restore(flags); | ||
| 970 | 883 | ||
| 971 | return time; | 884 | return time; |
| 972 | } | 885 | } |
| @@ -1117,43 +1030,6 @@ static struct rq *this_rq_lock(void) | |||
| 1117 | return rq; | 1030 | return rq; |
| 1118 | } | 1031 | } |
| 1119 | 1032 | ||
| 1120 | /* | ||
| 1121 | * We are going deep-idle (irqs are disabled): | ||
| 1122 | */ | ||
| 1123 | void sched_clock_idle_sleep_event(void) | ||
| 1124 | { | ||
| 1125 | struct rq *rq = cpu_rq(smp_processor_id()); | ||
| 1126 | |||
| 1127 | spin_lock(&rq->lock); | ||
| 1128 | __update_rq_clock(rq); | ||
| 1129 | spin_unlock(&rq->lock); | ||
| 1130 | rq->clock_deep_idle_events++; | ||
| 1131 | } | ||
| 1132 | EXPORT_SYMBOL_GPL(sched_clock_idle_sleep_event); | ||
| 1133 | |||
| 1134 | /* | ||
| 1135 | * We just idled delta nanoseconds (called with irqs disabled): | ||
| 1136 | */ | ||
| 1137 | void sched_clock_idle_wakeup_event(u64 delta_ns) | ||
| 1138 | { | ||
| 1139 | struct rq *rq = cpu_rq(smp_processor_id()); | ||
| 1140 | u64 now = sched_clock(); | ||
| 1141 | |||
| 1142 | rq->idle_clock += delta_ns; | ||
| 1143 | /* | ||
| 1144 | * Override the previous timestamp and ignore all | ||
| 1145 | * sched_clock() deltas that occured while we idled, | ||
| 1146 | * and use the PM-provided delta_ns to advance the | ||
| 1147 | * rq clock: | ||
| 1148 | */ | ||
| 1149 | spin_lock(&rq->lock); | ||
| 1150 | rq->prev_clock_raw = now; | ||
| 1151 | rq->clock += delta_ns; | ||
| 1152 | spin_unlock(&rq->lock); | ||
| 1153 | touch_softlockup_watchdog(); | ||
| 1154 | } | ||
| 1155 | EXPORT_SYMBOL_GPL(sched_clock_idle_wakeup_event); | ||
| 1156 | |||
| 1157 | static void __resched_task(struct task_struct *p, int tif_bit); | 1033 | static void __resched_task(struct task_struct *p, int tif_bit); |
| 1158 | 1034 | ||
| 1159 | static inline void resched_task(struct task_struct *p) | 1035 | static inline void resched_task(struct task_struct *p) |
| @@ -1189,6 +1065,7 @@ static inline void resched_rq(struct rq *rq) | |||
| 1189 | enum { | 1065 | enum { |
| 1190 | HRTICK_SET, /* re-programm hrtick_timer */ | 1066 | HRTICK_SET, /* re-programm hrtick_timer */ |
| 1191 | HRTICK_RESET, /* not a new slice */ | 1067 | HRTICK_RESET, /* not a new slice */ |
| 1068 | HRTICK_BLOCK, /* stop hrtick operations */ | ||
| 1192 | }; | 1069 | }; |
| 1193 | 1070 | ||
| 1194 | /* | 1071 | /* |
| @@ -1200,6 +1077,8 @@ static inline int hrtick_enabled(struct rq *rq) | |||
| 1200 | { | 1077 | { |
| 1201 | if (!sched_feat(HRTICK)) | 1078 | if (!sched_feat(HRTICK)) |
| 1202 | return 0; | 1079 | return 0; |
| 1080 | if (unlikely(test_bit(HRTICK_BLOCK, &rq->hrtick_flags))) | ||
| 1081 | return 0; | ||
| 1203 | return hrtimer_is_hres_active(&rq->hrtick_timer); | 1082 | return hrtimer_is_hres_active(&rq->hrtick_timer); |
| 1204 | } | 1083 | } |
| 1205 | 1084 | ||
| @@ -1275,14 +1154,70 @@ static enum hrtimer_restart hrtick(struct hrtimer *timer) | |||
| 1275 | WARN_ON_ONCE(cpu_of(rq) != smp_processor_id()); | 1154 | WARN_ON_ONCE(cpu_of(rq) != smp_processor_id()); |
| 1276 | 1155 | ||
| 1277 | spin_lock(&rq->lock); | 1156 | spin_lock(&rq->lock); |
| 1278 | __update_rq_clock(rq); | 1157 | update_rq_clock(rq); |
| 1279 | rq->curr->sched_class->task_tick(rq, rq->curr, 1); | 1158 | rq->curr->sched_class->task_tick(rq, rq->curr, 1); |
| 1280 | spin_unlock(&rq->lock); | 1159 | spin_unlock(&rq->lock); |
| 1281 | 1160 | ||
| 1282 | return HRTIMER_NORESTART; | 1161 | return HRTIMER_NORESTART; |
| 1283 | } | 1162 | } |
| 1284 | 1163 | ||
| 1285 | static inline void init_rq_hrtick(struct rq *rq) | 1164 | static void hotplug_hrtick_disable(int cpu) |
| 1165 | { | ||
| 1166 | struct rq *rq = cpu_rq(cpu); | ||
| 1167 | unsigned long flags; | ||
| 1168 | |||
| 1169 | spin_lock_irqsave(&rq->lock, flags); | ||
| 1170 | rq->hrtick_flags = 0; | ||
| 1171 | __set_bit(HRTICK_BLOCK, &rq->hrtick_flags); | ||
| 1172 | spin_unlock_irqrestore(&rq->lock, flags); | ||
| 1173 | |||
| 1174 | hrtick_clear(rq); | ||
| 1175 | } | ||
| 1176 | |||
| 1177 | static void hotplug_hrtick_enable(int cpu) | ||
| 1178 | { | ||
| 1179 | struct rq *rq = cpu_rq(cpu); | ||
| 1180 | unsigned long flags; | ||
| 1181 | |||
| 1182 | spin_lock_irqsave(&rq->lock, flags); | ||
| 1183 | __clear_bit(HRTICK_BLOCK, &rq->hrtick_flags); | ||
| 1184 | spin_unlock_irqrestore(&rq->lock, flags); | ||
| 1185 | } | ||
| 1186 | |||
| 1187 | static int | ||
| 1188 | hotplug_hrtick(struct notifier_block *nfb, unsigned long action, void *hcpu) | ||
| 1189 | { | ||
| 1190 | int cpu = (int)(long)hcpu; | ||
| 1191 | |||
| 1192 | switch (action) { | ||
| 1193 | case CPU_UP_CANCELED: | ||
| 1194 | case CPU_UP_CANCELED_FROZEN: | ||
| 1195 | case CPU_DOWN_PREPARE: | ||
| 1196 | case CPU_DOWN_PREPARE_FROZEN: | ||
| 1197 | case CPU_DEAD: | ||
| 1198 | case CPU_DEAD_FROZEN: | ||
| 1199 | hotplug_hrtick_disable(cpu); | ||
| 1200 | return NOTIFY_OK; | ||
| 1201 | |||
| 1202 | case CPU_UP_PREPARE: | ||
| 1203 | case CPU_UP_PREPARE_FROZEN: | ||
| 1204 | case CPU_DOWN_FAILED: | ||
| 1205 | case CPU_DOWN_FAILED_FROZEN: | ||
| 1206 | case CPU_ONLINE: | ||
| 1207 | case CPU_ONLINE_FROZEN: | ||
| 1208 | hotplug_hrtick_enable(cpu); | ||
| 1209 | return NOTIFY_OK; | ||
| 1210 | } | ||
| 1211 | |||
| 1212 | return NOTIFY_DONE; | ||
| 1213 | } | ||
| 1214 | |||
| 1215 | static void init_hrtick(void) | ||
| 1216 | { | ||
| 1217 | hotcpu_notifier(hotplug_hrtick, 0); | ||
| 1218 | } | ||
| 1219 | |||
| 1220 | static void init_rq_hrtick(struct rq *rq) | ||
| 1286 | { | 1221 | { |
| 1287 | rq->hrtick_flags = 0; | 1222 | rq->hrtick_flags = 0; |
| 1288 | hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); | 1223 | hrtimer_init(&rq->hrtick_timer, CLOCK_MONOTONIC, HRTIMER_MODE_REL); |
| @@ -1319,6 +1254,10 @@ static inline void init_rq_hrtick(struct rq *rq) | |||
| 1319 | void hrtick_resched(void) | 1254 | void hrtick_resched(void) |
| 1320 | { | 1255 | { |
| 1321 | } | 1256 | } |
| 1257 | |||
| 1258 | static inline void init_hrtick(void) | ||
| 1259 | { | ||
| 1260 | } | ||
| 1322 | #endif | 1261 | #endif |
| 1323 | 1262 | ||
| 1324 | /* | 1263 | /* |
| @@ -1438,8 +1377,8 @@ calc_delta_mine(unsigned long delta_exec, unsigned long weight, | |||
| 1438 | { | 1377 | { |
| 1439 | u64 tmp; | 1378 | u64 tmp; |
| 1440 | 1379 | ||
| 1441 | if (unlikely(!lw->inv_weight)) | 1380 | if (!lw->inv_weight) |
| 1442 | lw->inv_weight = (WMULT_CONST-lw->weight/2) / (lw->weight+1); | 1381 | lw->inv_weight = 1 + (WMULT_CONST-lw->weight/2)/(lw->weight+1); |
| 1443 | 1382 | ||
| 1444 | tmp = (u64)delta_exec * weight; | 1383 | tmp = (u64)delta_exec * weight; |
| 1445 | /* | 1384 | /* |
| @@ -1748,6 +1687,8 @@ __update_group_shares_cpu(struct task_group *tg, struct sched_domain *sd, | |||
| 1748 | 1687 | ||
| 1749 | if (shares < MIN_SHARES) | 1688 | if (shares < MIN_SHARES) |
| 1750 | shares = MIN_SHARES; | 1689 | shares = MIN_SHARES; |
| 1690 | else if (shares > MAX_SHARES) | ||
| 1691 | shares = MAX_SHARES; | ||
| 1751 | 1692 | ||
| 1752 | __set_se_shares(tg->se[tcpu], shares); | 1693 | __set_se_shares(tg->se[tcpu], shares); |
| 1753 | } | 1694 | } |
| @@ -4339,8 +4280,10 @@ void account_system_time(struct task_struct *p, int hardirq_offset, | |||
| 4339 | struct rq *rq = this_rq(); | 4280 | struct rq *rq = this_rq(); |
| 4340 | cputime64_t tmp; | 4281 | cputime64_t tmp; |
| 4341 | 4282 | ||
| 4342 | if ((p->flags & PF_VCPU) && (irq_count() - hardirq_offset == 0)) | 4283 | if ((p->flags & PF_VCPU) && (irq_count() - hardirq_offset == 0)) { |
| 4343 | return account_guest_time(p, cputime); | 4284 | account_guest_time(p, cputime); |
| 4285 | return; | ||
| 4286 | } | ||
| 4344 | 4287 | ||
| 4345 | p->stime = cputime_add(p->stime, cputime); | 4288 | p->stime = cputime_add(p->stime, cputime); |
| 4346 | 4289 | ||
| @@ -4404,19 +4347,11 @@ void scheduler_tick(void) | |||
| 4404 | int cpu = smp_processor_id(); | 4347 | int cpu = smp_processor_id(); |
| 4405 | struct rq *rq = cpu_rq(cpu); | 4348 | struct rq *rq = cpu_rq(cpu); |
| 4406 | struct task_struct *curr = rq->curr; | 4349 | struct task_struct *curr = rq->curr; |
| 4407 | u64 next_tick = rq->tick_timestamp + TICK_NSEC; | 4350 | |
| 4351 | sched_clock_tick(); | ||
| 4408 | 4352 | ||
| 4409 | spin_lock(&rq->lock); | 4353 | spin_lock(&rq->lock); |
| 4410 | __update_rq_clock(rq); | 4354 | update_rq_clock(rq); |
| 4411 | /* | ||
| 4412 | * Let rq->clock advance by at least TICK_NSEC: | ||
| 4413 | */ | ||
| 4414 | if (unlikely(rq->clock < next_tick)) { | ||
| 4415 | rq->clock = next_tick; | ||
| 4416 | rq->clock_underflows++; | ||
| 4417 | } | ||
| 4418 | rq->tick_timestamp = rq->clock; | ||
| 4419 | update_last_tick_seen(rq); | ||
| 4420 | update_cpu_load(rq); | 4355 | update_cpu_load(rq); |
| 4421 | curr->sched_class->task_tick(rq, curr, 0); | 4356 | curr->sched_class->task_tick(rq, curr, 0); |
| 4422 | spin_unlock(&rq->lock); | 4357 | spin_unlock(&rq->lock); |
| @@ -4570,7 +4505,7 @@ need_resched_nonpreemptible: | |||
| 4570 | * Do the rq-clock update outside the rq lock: | 4505 | * Do the rq-clock update outside the rq lock: |
| 4571 | */ | 4506 | */ |
| 4572 | local_irq_disable(); | 4507 | local_irq_disable(); |
| 4573 | __update_rq_clock(rq); | 4508 | update_rq_clock(rq); |
| 4574 | spin_lock(&rq->lock); | 4509 | spin_lock(&rq->lock); |
| 4575 | clear_tsk_need_resched(prev); | 4510 | clear_tsk_need_resched(prev); |
| 4576 | 4511 | ||
| @@ -4595,9 +4530,9 @@ need_resched_nonpreemptible: | |||
| 4595 | prev->sched_class->put_prev_task(rq, prev); | 4530 | prev->sched_class->put_prev_task(rq, prev); |
| 4596 | next = pick_next_task(rq, prev); | 4531 | next = pick_next_task(rq, prev); |
| 4597 | 4532 | ||
| 4598 | sched_info_switch(prev, next); | ||
| 4599 | |||
| 4600 | if (likely(prev != next)) { | 4533 | if (likely(prev != next)) { |
| 4534 | sched_info_switch(prev, next); | ||
| 4535 | |||
| 4601 | rq->nr_switches++; | 4536 | rq->nr_switches++; |
| 4602 | rq->curr = next; | 4537 | rq->curr = next; |
| 4603 | ++*switch_count; | 4538 | ++*switch_count; |
| @@ -7755,7 +7690,7 @@ void partition_sched_domains(int ndoms_new, cpumask_t *doms_new, | |||
| 7755 | { | 7690 | { |
| 7756 | int i, j; | 7691 | int i, j; |
| 7757 | 7692 | ||
| 7758 | lock_doms_cur(); | 7693 | mutex_lock(&sched_domains_mutex); |
| 7759 | 7694 | ||
| 7760 | /* always unregister in case we don't destroy any domains */ | 7695 | /* always unregister in case we don't destroy any domains */ |
| 7761 | unregister_sched_domain_sysctl(); | 7696 | unregister_sched_domain_sysctl(); |
| @@ -7804,7 +7739,7 @@ match2: | |||
| 7804 | 7739 | ||
| 7805 | register_sched_domain_sysctl(); | 7740 | register_sched_domain_sysctl(); |
| 7806 | 7741 | ||
| 7807 | unlock_doms_cur(); | 7742 | mutex_unlock(&sched_domains_mutex); |
| 7808 | } | 7743 | } |
| 7809 | 7744 | ||
| 7810 | #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) | 7745 | #if defined(CONFIG_SCHED_MC) || defined(CONFIG_SCHED_SMT) |
| @@ -7813,8 +7748,10 @@ int arch_reinit_sched_domains(void) | |||
| 7813 | int err; | 7748 | int err; |
| 7814 | 7749 | ||
| 7815 | get_online_cpus(); | 7750 | get_online_cpus(); |
| 7751 | mutex_lock(&sched_domains_mutex); | ||
| 7816 | detach_destroy_domains(&cpu_online_map); | 7752 | detach_destroy_domains(&cpu_online_map); |
| 7817 | err = arch_init_sched_domains(&cpu_online_map); | 7753 | err = arch_init_sched_domains(&cpu_online_map); |
| 7754 | mutex_unlock(&sched_domains_mutex); | ||
| 7818 | put_online_cpus(); | 7755 | put_online_cpus(); |
| 7819 | 7756 | ||
| 7820 | return err; | 7757 | return err; |
| @@ -7932,13 +7869,16 @@ void __init sched_init_smp(void) | |||
| 7932 | BUG_ON(sched_group_nodes_bycpu == NULL); | 7869 | BUG_ON(sched_group_nodes_bycpu == NULL); |
| 7933 | #endif | 7870 | #endif |
| 7934 | get_online_cpus(); | 7871 | get_online_cpus(); |
| 7872 | mutex_lock(&sched_domains_mutex); | ||
| 7935 | arch_init_sched_domains(&cpu_online_map); | 7873 | arch_init_sched_domains(&cpu_online_map); |
| 7936 | cpus_andnot(non_isolated_cpus, cpu_possible_map, cpu_isolated_map); | 7874 | cpus_andnot(non_isolated_cpus, cpu_possible_map, cpu_isolated_map); |
| 7937 | if (cpus_empty(non_isolated_cpus)) | 7875 | if (cpus_empty(non_isolated_cpus)) |
| 7938 | cpu_set(smp_processor_id(), non_isolated_cpus); | 7876 | cpu_set(smp_processor_id(), non_isolated_cpus); |
| 7877 | mutex_unlock(&sched_domains_mutex); | ||
| 7939 | put_online_cpus(); | 7878 | put_online_cpus(); |
| 7940 | /* XXX: Theoretical race here - CPU may be hotplugged now */ | 7879 | /* XXX: Theoretical race here - CPU may be hotplugged now */ |
| 7941 | hotcpu_notifier(update_sched_domains, 0); | 7880 | hotcpu_notifier(update_sched_domains, 0); |
| 7881 | init_hrtick(); | ||
| 7942 | 7882 | ||
| 7943 | /* Move init over to a non-isolated CPU */ | 7883 | /* Move init over to a non-isolated CPU */ |
| 7944 | if (set_cpus_allowed_ptr(current, &non_isolated_cpus) < 0) | 7884 | if (set_cpus_allowed_ptr(current, &non_isolated_cpus) < 0) |
| @@ -8025,7 +7965,7 @@ static void init_tg_cfs_entry(struct task_group *tg, struct cfs_rq *cfs_rq, | |||
| 8025 | 7965 | ||
| 8026 | se->my_q = cfs_rq; | 7966 | se->my_q = cfs_rq; |
| 8027 | se->load.weight = tg->shares; | 7967 | se->load.weight = tg->shares; |
| 8028 | se->load.inv_weight = div64_u64(1ULL<<32, se->load.weight); | 7968 | se->load.inv_weight = 0; |
| 8029 | se->parent = parent; | 7969 | se->parent = parent; |
| 8030 | } | 7970 | } |
| 8031 | #endif | 7971 | #endif |
| @@ -8149,8 +8089,6 @@ void __init sched_init(void) | |||
| 8149 | spin_lock_init(&rq->lock); | 8089 | spin_lock_init(&rq->lock); |
| 8150 | lockdep_set_class(&rq->lock, &rq->rq_lock_key); | 8090 | lockdep_set_class(&rq->lock, &rq->rq_lock_key); |
| 8151 | rq->nr_running = 0; | 8091 | rq->nr_running = 0; |
| 8152 | rq->clock = 1; | ||
| 8153 | update_last_tick_seen(rq); | ||
| 8154 | init_cfs_rq(&rq->cfs, rq); | 8092 | init_cfs_rq(&rq->cfs, rq); |
| 8155 | init_rt_rq(&rq->rt, rq); | 8093 | init_rt_rq(&rq->rt, rq); |
| 8156 | #ifdef CONFIG_FAIR_GROUP_SCHED | 8094 | #ifdef CONFIG_FAIR_GROUP_SCHED |
| @@ -8294,6 +8232,7 @@ EXPORT_SYMBOL(__might_sleep); | |||
| 8294 | static void normalize_task(struct rq *rq, struct task_struct *p) | 8232 | static void normalize_task(struct rq *rq, struct task_struct *p) |
| 8295 | { | 8233 | { |
| 8296 | int on_rq; | 8234 | int on_rq; |
| 8235 | |||
| 8297 | update_rq_clock(rq); | 8236 | update_rq_clock(rq); |
| 8298 | on_rq = p->se.on_rq; | 8237 | on_rq = p->se.on_rq; |
| 8299 | if (on_rq) | 8238 | if (on_rq) |
| @@ -8325,7 +8264,6 @@ void normalize_rt_tasks(void) | |||
| 8325 | p->se.sleep_start = 0; | 8264 | p->se.sleep_start = 0; |
| 8326 | p->se.block_start = 0; | 8265 | p->se.block_start = 0; |
| 8327 | #endif | 8266 | #endif |
| 8328 | task_rq(p)->clock = 0; | ||
| 8329 | 8267 | ||
| 8330 | if (!rt_task(p)) { | 8268 | if (!rt_task(p)) { |
| 8331 | /* | 8269 | /* |
| @@ -8692,7 +8630,7 @@ static void __set_se_shares(struct sched_entity *se, unsigned long shares) | |||
| 8692 | dequeue_entity(cfs_rq, se, 0); | 8630 | dequeue_entity(cfs_rq, se, 0); |
| 8693 | 8631 | ||
| 8694 | se->load.weight = shares; | 8632 | se->load.weight = shares; |
| 8695 | se->load.inv_weight = div64_u64((1ULL<<32), shares); | 8633 | se->load.inv_weight = 0; |
| 8696 | 8634 | ||
| 8697 | if (on_rq) | 8635 | if (on_rq) |
| 8698 | enqueue_entity(cfs_rq, se, 0); | 8636 | enqueue_entity(cfs_rq, se, 0); |
| @@ -8722,13 +8660,10 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares) | |||
| 8722 | if (!tg->se[0]) | 8660 | if (!tg->se[0]) |
| 8723 | return -EINVAL; | 8661 | return -EINVAL; |
| 8724 | 8662 | ||
| 8725 | /* | ||
| 8726 | * A weight of 0 or 1 can cause arithmetics problems. | ||
| 8727 | * (The default weight is 1024 - so there's no practical | ||
| 8728 | * limitation from this.) | ||
| 8729 | */ | ||
| 8730 | if (shares < MIN_SHARES) | 8663 | if (shares < MIN_SHARES) |
| 8731 | shares = MIN_SHARES; | 8664 | shares = MIN_SHARES; |
| 8665 | else if (shares > MAX_SHARES) | ||
| 8666 | shares = MAX_SHARES; | ||
| 8732 | 8667 | ||
| 8733 | mutex_lock(&shares_mutex); | 8668 | mutex_lock(&shares_mutex); |
| 8734 | if (tg->shares == shares) | 8669 | if (tg->shares == shares) |
| @@ -8753,7 +8688,7 @@ int sched_group_set_shares(struct task_group *tg, unsigned long shares) | |||
| 8753 | * force a rebalance | 8688 | * force a rebalance |
| 8754 | */ | 8689 | */ |
| 8755 | cfs_rq_set_shares(tg->cfs_rq[i], 0); | 8690 | cfs_rq_set_shares(tg->cfs_rq[i], 0); |
| 8756 | set_se_shares(tg->se[i], shares/nr_cpu_ids); | 8691 | set_se_shares(tg->se[i], shares); |
| 8757 | } | 8692 | } |
| 8758 | 8693 | ||
| 8759 | /* | 8694 | /* |
diff --git a/kernel/sched_clock.c b/kernel/sched_clock.c new file mode 100644 index 000000000000..9c597e37f7de --- /dev/null +++ b/kernel/sched_clock.c | |||
| @@ -0,0 +1,236 @@ | |||
| 1 | /* | ||
| 2 | * sched_clock for unstable cpu clocks | ||
| 3 | * | ||
| 4 | * Copyright (C) 2008 Red Hat, Inc., Peter Zijlstra <pzijlstr@redhat.com> | ||
| 5 | * | ||
| 6 | * Based on code by: | ||
| 7 | * Ingo Molnar <mingo@redhat.com> | ||
| 8 | * Guillaume Chazarain <guichaz@gmail.com> | ||
| 9 | * | ||
| 10 | * Create a semi stable clock from a mixture of other events, including: | ||
| 11 | * - gtod | ||
| 12 | * - jiffies | ||
| 13 | * - sched_clock() | ||
| 14 | * - explicit idle events | ||
| 15 | * | ||
| 16 | * We use gtod as base and the unstable clock deltas. The deltas are filtered, | ||
| 17 | * making it monotonic and keeping it within an expected window. This window | ||
| 18 | * is set up using jiffies. | ||
| 19 | * | ||
| 20 | * Furthermore, explicit sleep and wakeup hooks allow us to account for time | ||
| 21 | * that is otherwise invisible (TSC gets stopped). | ||
| 22 | * | ||
| 23 | * The clock: sched_clock_cpu() is monotonic per cpu, and should be somewhat | ||
| 24 | * consistent between cpus (never more than 1 jiffies difference). | ||
| 25 | */ | ||
| 26 | #include <linux/sched.h> | ||
| 27 | #include <linux/percpu.h> | ||
| 28 | #include <linux/spinlock.h> | ||
| 29 | #include <linux/ktime.h> | ||
| 30 | #include <linux/module.h> | ||
| 31 | |||
| 32 | |||
| 33 | #ifdef CONFIG_HAVE_UNSTABLE_SCHED_CLOCK | ||
| 34 | |||
| 35 | struct sched_clock_data { | ||
| 36 | /* | ||
| 37 | * Raw spinlock - this is a special case: this might be called | ||
| 38 | * from within instrumentation code so we dont want to do any | ||
| 39 | * instrumentation ourselves. | ||
| 40 | */ | ||
| 41 | raw_spinlock_t lock; | ||
| 42 | |||
| 43 | unsigned long prev_jiffies; | ||
| 44 | u64 prev_raw; | ||
| 45 | u64 tick_raw; | ||
| 46 | u64 tick_gtod; | ||
| 47 | u64 clock; | ||
| 48 | }; | ||
| 49 | |||
| 50 | static DEFINE_PER_CPU_SHARED_ALIGNED(struct sched_clock_data, sched_clock_data); | ||
| 51 | |||
| 52 | static inline struct sched_clock_data *this_scd(void) | ||
| 53 | { | ||
| 54 | return &__get_cpu_var(sched_clock_data); | ||
| 55 | } | ||
| 56 | |||
| 57 | static inline struct sched_clock_data *cpu_sdc(int cpu) | ||
| 58 | { | ||
| 59 | return &per_cpu(sched_clock_data, cpu); | ||
| 60 | } | ||
| 61 | |||
| 62 | void sched_clock_init(void) | ||
| 63 | { | ||
| 64 | u64 ktime_now = ktime_to_ns(ktime_get()); | ||
| 65 | u64 now = 0; | ||
| 66 | int cpu; | ||
| 67 | |||
| 68 | for_each_possible_cpu(cpu) { | ||
| 69 | struct sched_clock_data *scd = cpu_sdc(cpu); | ||
| 70 | |||
| 71 | scd->lock = (raw_spinlock_t)__RAW_SPIN_LOCK_UNLOCKED; | ||
| 72 | scd->prev_jiffies = jiffies; | ||
| 73 | scd->prev_raw = now; | ||
| 74 | scd->tick_raw = now; | ||
| 75 | scd->tick_gtod = ktime_now; | ||
| 76 | scd->clock = ktime_now; | ||
| 77 | } | ||
| 78 | } | ||
| 79 | |||
| 80 | /* | ||
| 81 | * update the percpu scd from the raw @now value | ||
| 82 | * | ||
| 83 | * - filter out backward motion | ||
| 84 | * - use jiffies to generate a min,max window to clip the raw values | ||
| 85 | */ | ||
| 86 | static void __update_sched_clock(struct sched_clock_data *scd, u64 now) | ||
| 87 | { | ||
| 88 | unsigned long now_jiffies = jiffies; | ||
| 89 | long delta_jiffies = now_jiffies - scd->prev_jiffies; | ||
| 90 | u64 clock = scd->clock; | ||
| 91 | u64 min_clock, max_clock; | ||
| 92 | s64 delta = now - scd->prev_raw; | ||
| 93 | |||
| 94 | WARN_ON_ONCE(!irqs_disabled()); | ||
| 95 | min_clock = scd->tick_gtod + delta_jiffies * TICK_NSEC; | ||
| 96 | |||
| 97 | if (unlikely(delta < 0)) { | ||
| 98 | clock++; | ||
| 99 | goto out; | ||
| 100 | } | ||
| 101 | |||
| 102 | max_clock = min_clock + TICK_NSEC; | ||
| 103 | |||
| 104 | if (unlikely(clock + delta > max_clock)) { | ||
| 105 | if (clock < max_clock) | ||
| 106 | clock = max_clock; | ||
| 107 | else | ||
| 108 | clock++; | ||
| 109 | } else { | ||
| 110 | clock += delta; | ||
| 111 | } | ||
| 112 | |||
| 113 | out: | ||
| 114 | if (unlikely(clock < min_clock)) | ||
| 115 | clock = min_clock; | ||
| 116 | |||
| 117 | scd->prev_raw = now; | ||
| 118 | scd->prev_jiffies = now_jiffies; | ||
| 119 | scd->clock = clock; | ||
| 120 | } | ||
| 121 | |||
| 122 | static void lock_double_clock(struct sched_clock_data *data1, | ||
| 123 | struct sched_clock_data *data2) | ||
| 124 | { | ||
| 125 | if (data1 < data2) { | ||
| 126 | __raw_spin_lock(&data1->lock); | ||
| 127 | __raw_spin_lock(&data2->lock); | ||
| 128 | } else { | ||
| 129 | __raw_spin_lock(&data2->lock); | ||
| 130 | __raw_spin_lock(&data1->lock); | ||
| 131 | } | ||
| 132 | } | ||
| 133 | |||
| 134 | u64 sched_clock_cpu(int cpu) | ||
| 135 | { | ||
| 136 | struct sched_clock_data *scd = cpu_sdc(cpu); | ||
| 137 | u64 now, clock; | ||
| 138 | |||
| 139 | WARN_ON_ONCE(!irqs_disabled()); | ||
| 140 | now = sched_clock(); | ||
| 141 | |||
| 142 | if (cpu != raw_smp_processor_id()) { | ||
| 143 | /* | ||
| 144 | * in order to update a remote cpu's clock based on our | ||
| 145 | * unstable raw time rebase it against: | ||
| 146 | * tick_raw (offset between raw counters) | ||
| 147 | * tick_gotd (tick offset between cpus) | ||
| 148 | */ | ||
| 149 | struct sched_clock_data *my_scd = this_scd(); | ||
| 150 | |||
| 151 | lock_double_clock(scd, my_scd); | ||
| 152 | |||
| 153 | now -= my_scd->tick_raw; | ||
| 154 | now += scd->tick_raw; | ||
| 155 | |||
| 156 | now -= my_scd->tick_gtod; | ||
| 157 | now += scd->tick_gtod; | ||
| 158 | |||
| 159 | __raw_spin_unlock(&my_scd->lock); | ||
| 160 | } else { | ||
| 161 | __raw_spin_lock(&scd->lock); | ||
| 162 | } | ||
| 163 | |||
| 164 | __update_sched_clock(scd, now); | ||
| 165 | clock = scd->clock; | ||
| 166 | |||
| 167 | __raw_spin_unlock(&scd->lock); | ||
| 168 | |||
| 169 | return clock; | ||
| 170 | } | ||
| 171 | |||
| 172 | void sched_clock_tick(void) | ||
| 173 | { | ||
| 174 | struct sched_clock_data *scd = this_scd(); | ||
| 175 | u64 now, now_gtod; | ||
| 176 | |||
| 177 | WARN_ON_ONCE(!irqs_disabled()); | ||
| 178 | |||
| 179 | now = sched_clock(); | ||
| 180 | now_gtod = ktime_to_ns(ktime_get()); | ||
| 181 | |||
| 182 | __raw_spin_lock(&scd->lock); | ||
| 183 | __update_sched_clock(scd, now); | ||
| 184 | /* | ||
| 185 | * update tick_gtod after __update_sched_clock() because that will | ||
| 186 | * already observe 1 new jiffy; adding a new tick_gtod to that would | ||
| 187 | * increase the clock 2 jiffies. | ||
| 188 | */ | ||
| 189 | scd->tick_raw = now; | ||
| 190 | scd->tick_gtod = now_gtod; | ||
| 191 | __raw_spin_unlock(&scd->lock); | ||
| 192 | } | ||
| 193 | |||
| 194 | /* | ||
| 195 | * We are going deep-idle (irqs are disabled): | ||
| 196 | */ | ||
| 197 | void sched_clock_idle_sleep_event(void) | ||
| 198 | { | ||
| 199 | sched_clock_cpu(smp_processor_id()); | ||
| 200 | } | ||
| 201 | EXPORT_SYMBOL_GPL(sched_clock_idle_sleep_event); | ||
| 202 | |||
| 203 | /* | ||
| 204 | * We just idled delta nanoseconds (called with irqs disabled): | ||
| 205 | */ | ||
| 206 | void sched_clock_idle_wakeup_event(u64 delta_ns) | ||
| 207 | { | ||
| 208 | struct sched_clock_data *scd = this_scd(); | ||
| 209 | u64 now = sched_clock(); | ||
| 210 | |||
| 211 | /* | ||
| 212 | * Override the previous timestamp and ignore all | ||
| 213 | * sched_clock() deltas that occured while we idled, | ||
| 214 | * and use the PM-provided delta_ns to advance the | ||
| 215 | * rq clock: | ||
| 216 | */ | ||
| 217 | __raw_spin_lock(&scd->lock); | ||
| 218 | scd->prev_raw = now; | ||
| 219 | scd->clock += delta_ns; | ||
| 220 | __raw_spin_unlock(&scd->lock); | ||
| 221 | |||
| 222 | touch_softlockup_watchdog(); | ||
| 223 | } | ||
| 224 | EXPORT_SYMBOL_GPL(sched_clock_idle_wakeup_event); | ||
| 225 | |||
| 226 | #endif | ||
| 227 | |||
| 228 | /* | ||
| 229 | * Scheduler clock - returns current time in nanosec units. | ||
| 230 | * This is default implementation. | ||
| 231 | * Architectures and sub-architectures can override this. | ||
| 232 | */ | ||
| 233 | unsigned long long __attribute__((weak)) sched_clock(void) | ||
| 234 | { | ||
| 235 | return (unsigned long long)jiffies * (NSEC_PER_SEC / HZ); | ||
| 236 | } | ||
diff --git a/kernel/sched_debug.c b/kernel/sched_debug.c index 6b4a12558e88..5f06118fbc31 100644 --- a/kernel/sched_debug.c +++ b/kernel/sched_debug.c | |||
| @@ -204,13 +204,6 @@ static void print_cpu(struct seq_file *m, int cpu) | |||
| 204 | PN(next_balance); | 204 | PN(next_balance); |
| 205 | P(curr->pid); | 205 | P(curr->pid); |
| 206 | PN(clock); | 206 | PN(clock); |
| 207 | PN(idle_clock); | ||
| 208 | PN(prev_clock_raw); | ||
| 209 | P(clock_warps); | ||
| 210 | P(clock_overflows); | ||
| 211 | P(clock_underflows); | ||
| 212 | P(clock_deep_idle_events); | ||
| 213 | PN(clock_max_delta); | ||
| 214 | P(cpu_load[0]); | 207 | P(cpu_load[0]); |
| 215 | P(cpu_load[1]); | 208 | P(cpu_load[1]); |
| 216 | P(cpu_load[2]); | 209 | P(cpu_load[2]); |
diff --git a/kernel/sched_fair.c b/kernel/sched_fair.c index 89fa32b4edf2..c863663d204d 100644 --- a/kernel/sched_fair.c +++ b/kernel/sched_fair.c | |||
| @@ -682,6 +682,7 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int wakeup) | |||
| 682 | * Update run-time statistics of the 'current'. | 682 | * Update run-time statistics of the 'current'. |
| 683 | */ | 683 | */ |
| 684 | update_curr(cfs_rq); | 684 | update_curr(cfs_rq); |
| 685 | account_entity_enqueue(cfs_rq, se); | ||
| 685 | 686 | ||
| 686 | if (wakeup) { | 687 | if (wakeup) { |
| 687 | place_entity(cfs_rq, se, 0); | 688 | place_entity(cfs_rq, se, 0); |
| @@ -692,7 +693,6 @@ enqueue_entity(struct cfs_rq *cfs_rq, struct sched_entity *se, int wakeup) | |||
| 692 | check_spread(cfs_rq, se); | 693 | check_spread(cfs_rq, se); |
| 693 | if (se != cfs_rq->curr) | 694 | if (se != cfs_rq->curr) |
| 694 | __enqueue_entity(cfs_rq, se); | 695 | __enqueue_entity(cfs_rq, se); |
| 695 | account_entity_enqueue(cfs_rq, se); | ||
| 696 | } | 696 | } |
| 697 | 697 | ||
| 698 | static void update_avg(u64 *avg, u64 sample) | 698 | static void update_avg(u64 *avg, u64 sample) |
| @@ -841,8 +841,10 @@ entity_tick(struct cfs_rq *cfs_rq, struct sched_entity *curr, int queued) | |||
| 841 | * queued ticks are scheduled to match the slice, so don't bother | 841 | * queued ticks are scheduled to match the slice, so don't bother |
| 842 | * validating it and just reschedule. | 842 | * validating it and just reschedule. |
| 843 | */ | 843 | */ |
| 844 | if (queued) | 844 | if (queued) { |
| 845 | return resched_task(rq_of(cfs_rq)->curr); | 845 | resched_task(rq_of(cfs_rq)->curr); |
| 846 | return; | ||
| 847 | } | ||
| 846 | /* | 848 | /* |
| 847 | * don't let the period tick interfere with the hrtick preemption | 849 | * don't let the period tick interfere with the hrtick preemption |
| 848 | */ | 850 | */ |
| @@ -957,7 +959,7 @@ static void yield_task_fair(struct rq *rq) | |||
| 957 | return; | 959 | return; |
| 958 | 960 | ||
| 959 | if (likely(!sysctl_sched_compat_yield) && curr->policy != SCHED_BATCH) { | 961 | if (likely(!sysctl_sched_compat_yield) && curr->policy != SCHED_BATCH) { |
| 960 | __update_rq_clock(rq); | 962 | update_rq_clock(rq); |
| 961 | /* | 963 | /* |
| 962 | * Update run-time statistics of the 'current'. | 964 | * Update run-time statistics of the 'current'. |
| 963 | */ | 965 | */ |
| @@ -1007,7 +1009,7 @@ static int wake_idle(int cpu, struct task_struct *p) | |||
| 1007 | * sibling runqueue info. This will avoid the checks and cache miss | 1009 | * sibling runqueue info. This will avoid the checks and cache miss |
| 1008 | * penalities associated with that. | 1010 | * penalities associated with that. |
| 1009 | */ | 1011 | */ |
| 1010 | if (idle_cpu(cpu) || cpu_rq(cpu)->nr_running > 1) | 1012 | if (idle_cpu(cpu) || cpu_rq(cpu)->cfs.nr_running > 1) |
| 1011 | return cpu; | 1013 | return cpu; |
| 1012 | 1014 | ||
| 1013 | for_each_domain(cpu, sd) { | 1015 | for_each_domain(cpu, sd) { |
| @@ -1611,30 +1613,6 @@ static const struct sched_class fair_sched_class = { | |||
| 1611 | }; | 1613 | }; |
| 1612 | 1614 | ||
| 1613 | #ifdef CONFIG_SCHED_DEBUG | 1615 | #ifdef CONFIG_SCHED_DEBUG |
| 1614 | static void | ||
| 1615 | print_cfs_rq_tasks(struct seq_file *m, struct cfs_rq *cfs_rq, int depth) | ||
| 1616 | { | ||
| 1617 | struct sched_entity *se; | ||
| 1618 | |||
| 1619 | if (!cfs_rq) | ||
| 1620 | return; | ||
| 1621 | |||
| 1622 | list_for_each_entry_rcu(se, &cfs_rq->tasks, group_node) { | ||
| 1623 | int i; | ||
| 1624 | |||
| 1625 | for (i = depth; i; i--) | ||
| 1626 | seq_puts(m, " "); | ||
| 1627 | |||
| 1628 | seq_printf(m, "%lu %s %lu\n", | ||
| 1629 | se->load.weight, | ||
| 1630 | entity_is_task(se) ? "T" : "G", | ||
| 1631 | calc_delta_weight(SCHED_LOAD_SCALE, se) | ||
| 1632 | ); | ||
| 1633 | if (!entity_is_task(se)) | ||
| 1634 | print_cfs_rq_tasks(m, group_cfs_rq(se), depth + 1); | ||
| 1635 | } | ||
| 1636 | } | ||
| 1637 | |||
| 1638 | static void print_cfs_stats(struct seq_file *m, int cpu) | 1616 | static void print_cfs_stats(struct seq_file *m, int cpu) |
| 1639 | { | 1617 | { |
| 1640 | struct cfs_rq *cfs_rq; | 1618 | struct cfs_rq *cfs_rq; |
| @@ -1642,9 +1620,6 @@ static void print_cfs_stats(struct seq_file *m, int cpu) | |||
| 1642 | rcu_read_lock(); | 1620 | rcu_read_lock(); |
| 1643 | for_each_leaf_cfs_rq(cpu_rq(cpu), cfs_rq) | 1621 | for_each_leaf_cfs_rq(cpu_rq(cpu), cfs_rq) |
| 1644 | print_cfs_rq(m, cpu, cfs_rq); | 1622 | print_cfs_rq(m, cpu, cfs_rq); |
| 1645 | |||
| 1646 | seq_printf(m, "\nWeight tree:\n"); | ||
| 1647 | print_cfs_rq_tasks(m, &cpu_rq(cpu)->cfs, 1); | ||
| 1648 | rcu_read_unlock(); | 1623 | rcu_read_unlock(); |
| 1649 | } | 1624 | } |
| 1650 | #endif | 1625 | #endif |
diff --git a/kernel/sched_idletask.c b/kernel/sched_idletask.c index 2bcafa375633..3a4f92dbbe66 100644 --- a/kernel/sched_idletask.c +++ b/kernel/sched_idletask.c | |||
| @@ -99,7 +99,7 @@ static void prio_changed_idle(struct rq *rq, struct task_struct *p, | |||
| 99 | /* | 99 | /* |
| 100 | * Simple, special scheduling class for the per-CPU idle tasks: | 100 | * Simple, special scheduling class for the per-CPU idle tasks: |
| 101 | */ | 101 | */ |
| 102 | const struct sched_class idle_sched_class = { | 102 | static const struct sched_class idle_sched_class = { |
| 103 | /* .next is NULL */ | 103 | /* .next is NULL */ |
| 104 | /* no enqueue/yield_task for idle tasks */ | 104 | /* no enqueue/yield_task for idle tasks */ |
| 105 | 105 | ||
diff --git a/kernel/sched_rt.c b/kernel/sched_rt.c index c2730a5a4f05..060e87b0cb1c 100644 --- a/kernel/sched_rt.c +++ b/kernel/sched_rt.c | |||
| @@ -1098,11 +1098,14 @@ static void post_schedule_rt(struct rq *rq) | |||
| 1098 | } | 1098 | } |
| 1099 | } | 1099 | } |
| 1100 | 1100 | ||
| 1101 | 1101 | /* | |
| 1102 | * If we are not running and we are not going to reschedule soon, we should | ||
| 1103 | * try to push tasks away now | ||
| 1104 | */ | ||
| 1102 | static void task_wake_up_rt(struct rq *rq, struct task_struct *p) | 1105 | static void task_wake_up_rt(struct rq *rq, struct task_struct *p) |
| 1103 | { | 1106 | { |
| 1104 | if (!task_running(rq, p) && | 1107 | if (!task_running(rq, p) && |
| 1105 | (p->prio >= rq->rt.highest_prio) && | 1108 | !test_tsk_need_resched(rq->curr) && |
| 1106 | rq->rt.overloaded) | 1109 | rq->rt.overloaded) |
| 1107 | push_rt_tasks(rq); | 1110 | push_rt_tasks(rq); |
| 1108 | } | 1111 | } |
| @@ -1309,7 +1312,7 @@ static void set_curr_task_rt(struct rq *rq) | |||
| 1309 | p->se.exec_start = rq->clock; | 1312 | p->se.exec_start = rq->clock; |
| 1310 | } | 1313 | } |
| 1311 | 1314 | ||
| 1312 | const struct sched_class rt_sched_class = { | 1315 | static const struct sched_class rt_sched_class = { |
| 1313 | .next = &fair_sched_class, | 1316 | .next = &fair_sched_class, |
| 1314 | .enqueue_task = enqueue_task_rt, | 1317 | .enqueue_task = enqueue_task_rt, |
| 1315 | .dequeue_task = dequeue_task_rt, | 1318 | .dequeue_task = dequeue_task_rt, |
