2012-06-16 17:57:37 +04:00
# include <linux/export.h>
# include <linux/sched.h>
# include <linux/tsacct_kern.h>
# include <linux/kernel_stat.h>
# include <linux/static_key.h>
2012-07-25 09:56:04 +04:00
# include <linux/context_tracking.h>
2012-06-16 17:57:37 +04:00
# include "sched.h"
# ifdef CONFIG_IRQ_TIME_ACCOUNTING
/*
* There are no locks covering percpu hardirq / softirq time .
2012-09-08 17:23:11 +04:00
* They are only modified in vtime_account , on corresponding CPU
2012-06-16 17:57:37 +04:00
* with interrupts disabled . So , writes are safe .
* They are read and saved off onto struct rq in update_rq_clock ( ) .
* This may result in other CPU reading this CPU ' s irq time and can
2012-09-08 17:23:11 +04:00
* race with irq / vtime_account on this CPU . We would either get old
2012-06-16 17:57:37 +04:00
* or new value with a side effect of accounting a slice of irq time to wrong
* task when irq is in progress while we read rq - > clock . That is a worthy
* compromise in place of having locks on each irq in account_system_time .
*/
DEFINE_PER_CPU ( u64 , cpu_hardirq_time ) ;
DEFINE_PER_CPU ( u64 , cpu_softirq_time ) ;
static DEFINE_PER_CPU ( u64 , irq_start_time ) ;
static int sched_clock_irqtime ;
void enable_sched_clock_irqtime ( void )
{
sched_clock_irqtime = 1 ;
}
void disable_sched_clock_irqtime ( void )
{
sched_clock_irqtime = 0 ;
}
# ifndef CONFIG_64BIT
DEFINE_PER_CPU ( seqcount_t , irq_time_seq ) ;
# endif /* CONFIG_64BIT */
/*
* Called before incrementing preempt_count on { soft , } irq_enter
* and before decrementing preempt_count on { soft , } irq_exit .
*/
2012-10-06 07:23:22 +04:00
void irqtime_account_irq ( struct task_struct * curr )
2012-06-16 17:57:37 +04:00
{
unsigned long flags ;
s64 delta ;
int cpu ;
if ( ! sched_clock_irqtime )
return ;
local_irq_save ( flags ) ;
cpu = smp_processor_id ( ) ;
delta = sched_clock_cpu ( cpu ) - __this_cpu_read ( irq_start_time ) ;
__this_cpu_add ( irq_start_time , delta ) ;
irq_time_write_begin ( ) ;
/*
* We do not account for softirq time from ksoftirqd here .
* We want to continue accounting softirq time to ksoftirqd thread
* in that case , so as not to confuse scheduler with a special task
* that do not consume any time , but still wants to run .
*/
if ( hardirq_count ( ) )
__this_cpu_add ( cpu_hardirq_time , delta ) ;
else if ( in_serving_softirq ( ) & & curr ! = this_cpu_ksoftirqd ( ) )
__this_cpu_add ( cpu_softirq_time , delta ) ;
irq_time_write_end ( ) ;
local_irq_restore ( flags ) ;
}
2012-10-06 07:23:22 +04:00
EXPORT_SYMBOL_GPL ( irqtime_account_irq ) ;
2012-06-16 17:57:37 +04:00
static int irqtime_account_hi_update ( void )
{
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
unsigned long flags ;
u64 latest_ns ;
int ret = 0 ;
local_irq_save ( flags ) ;
latest_ns = this_cpu_read ( cpu_hardirq_time ) ;
if ( nsecs_to_cputime64 ( latest_ns ) > cpustat [ CPUTIME_IRQ ] )
ret = 1 ;
local_irq_restore ( flags ) ;
return ret ;
}
static int irqtime_account_si_update ( void )
{
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
unsigned long flags ;
u64 latest_ns ;
int ret = 0 ;
local_irq_save ( flags ) ;
latest_ns = this_cpu_read ( cpu_softirq_time ) ;
if ( nsecs_to_cputime64 ( latest_ns ) > cpustat [ CPUTIME_SOFTIRQ ] )
ret = 1 ;
local_irq_restore ( flags ) ;
return ret ;
}
# else /* CONFIG_IRQ_TIME_ACCOUNTING */
# define sched_clock_irqtime (0)
# endif /* !CONFIG_IRQ_TIME_ACCOUNTING */
static inline void task_group_account_field ( struct task_struct * p , int index ,
u64 tmp )
{
/*
* Since all updates are sure to touch the root cgroup , we
* get ourselves ahead and touch it first . If the root cgroup
* is the only cgroup , then nothing else should be necessary .
*
*/
__get_cpu_var ( kernel_cpustat ) . cpustat [ index ] + = tmp ;
2013-03-29 10:37:06 +04:00
cpuacct_account_field ( p , index , tmp ) ;
2012-06-16 17:57:37 +04:00
}
/*
* Account user cpu time to a process .
* @ p : the process that the cpu time gets accounted to
* @ cputime : the cpu time spent in user space since the last update
* @ cputime_scaled : cputime scaled by cpu frequency
*/
void account_user_time ( struct task_struct * p , cputime_t cputime ,
cputime_t cputime_scaled )
{
int index ;
/* Add user time to process. */
p - > utime + = cputime ;
p - > utimescaled + = cputime_scaled ;
account_group_user_time ( p , cputime ) ;
index = ( TASK_NICE ( p ) > 0 ) ? CPUTIME_NICE : CPUTIME_USER ;
/* Add user time to cpustat. */
task_group_account_field ( p , index , ( __force u64 ) cputime ) ;
/* Account for user time used */
2012-11-13 17:20:55 +04:00
acct_account_cputime ( p ) ;
2012-06-16 17:57:37 +04:00
}
/*
* Account guest cpu time to a process .
* @ p : the process that the cpu time gets accounted to
* @ cputime : the cpu time spent in virtual machine since the last update
* @ cputime_scaled : cputime scaled by cpu frequency
*/
static void account_guest_time ( struct task_struct * p , cputime_t cputime ,
cputime_t cputime_scaled )
{
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
/* Add guest time to process. */
p - > utime + = cputime ;
p - > utimescaled + = cputime_scaled ;
account_group_user_time ( p , cputime ) ;
p - > gtime + = cputime ;
/* Add guest time to cpustat. */
if ( TASK_NICE ( p ) > 0 ) {
cpustat [ CPUTIME_NICE ] + = ( __force u64 ) cputime ;
cpustat [ CPUTIME_GUEST_NICE ] + = ( __force u64 ) cputime ;
} else {
cpustat [ CPUTIME_USER ] + = ( __force u64 ) cputime ;
cpustat [ CPUTIME_GUEST ] + = ( __force u64 ) cputime ;
}
}
/*
* Account system cpu time to a process and desired cpustat field
* @ p : the process that the cpu time gets accounted to
* @ cputime : the cpu time spent in kernel space since the last update
* @ cputime_scaled : cputime scaled by cpu frequency
* @ target_cputime64 : pointer to cpustat field that has to be updated
*/
static inline
void __account_system_time ( struct task_struct * p , cputime_t cputime ,
cputime_t cputime_scaled , int index )
{
/* Add system time to process. */
p - > stime + = cputime ;
p - > stimescaled + = cputime_scaled ;
account_group_system_time ( p , cputime ) ;
/* Add system time to cpustat. */
task_group_account_field ( p , index , ( __force u64 ) cputime ) ;
/* Account for system time used */
2012-11-13 17:20:55 +04:00
acct_account_cputime ( p ) ;
2012-06-16 17:57:37 +04:00
}
/*
* Account system cpu time to a process .
* @ p : the process that the cpu time gets accounted to
* @ hardirq_offset : the offset to subtract from hardirq_count ( )
* @ cputime : the cpu time spent in kernel space since the last update
* @ cputime_scaled : cputime scaled by cpu frequency
*/
void account_system_time ( struct task_struct * p , int hardirq_offset ,
cputime_t cputime , cputime_t cputime_scaled )
{
int index ;
if ( ( p - > flags & PF_VCPU ) & & ( irq_count ( ) - hardirq_offset = = 0 ) ) {
account_guest_time ( p , cputime , cputime_scaled ) ;
return ;
}
if ( hardirq_count ( ) - hardirq_offset )
index = CPUTIME_IRQ ;
else if ( in_serving_softirq ( ) )
index = CPUTIME_SOFTIRQ ;
else
index = CPUTIME_SYSTEM ;
__account_system_time ( p , cputime , cputime_scaled , index ) ;
}
/*
* Account for involuntary wait time .
* @ cputime : the cpu time spent in involuntary wait
*/
void account_steal_time ( cputime_t cputime )
{
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
cpustat [ CPUTIME_STEAL ] + = ( __force u64 ) cputime ;
}
/*
* Account for idle time .
* @ cputime : the cpu time spent in idle wait
*/
void account_idle_time ( cputime_t cputime )
{
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
struct rq * rq = this_rq ( ) ;
if ( atomic_read ( & rq - > nr_iowait ) > 0 )
cpustat [ CPUTIME_IOWAIT ] + = ( __force u64 ) cputime ;
else
cpustat [ CPUTIME_IDLE ] + = ( __force u64 ) cputime ;
}
static __always_inline bool steal_account_process_tick ( void )
{
# ifdef CONFIG_PARAVIRT
if ( static_key_false ( & paravirt_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 ;
}
2012-11-21 18:55:59 +04:00
/*
* Accumulate raw cputime values of dead tasks ( sig - > [ us ] time ) and live
* tasks ( sum on group iteration ) belonging to @ tsk ' s group .
*/
void thread_group_cputime ( struct task_struct * tsk , struct task_cputime * times )
{
struct signal_struct * sig = tsk - > signal ;
2012-11-13 17:20:55 +04:00
cputime_t utime , stime ;
2012-11-21 18:55:59 +04:00
struct task_struct * t ;
times - > utime = sig - > utime ;
times - > stime = sig - > stime ;
times - > sum_exec_runtime = sig - > sum_sched_runtime ;
rcu_read_lock ( ) ;
/* make sure we can trust tsk->thread_group list */
if ( ! likely ( pid_alive ( tsk ) ) )
goto out ;
t = tsk ;
do {
2013-04-04 12:57:48 +04:00
task_cputime ( t , & utime , & stime ) ;
2012-11-13 17:20:55 +04:00
times - > utime + = utime ;
times - > stime + = stime ;
2012-11-21 18:55:59 +04:00
times - > sum_exec_runtime + = task_sched_runtime ( t ) ;
} while_each_thread ( tsk , t ) ;
out :
rcu_read_unlock ( ) ;
}
2012-06-16 17:57:37 +04:00
# ifdef CONFIG_IRQ_TIME_ACCOUNTING
/*
* Account a tick to a process and cpustat
* @ p : the process that the cpu time gets accounted to
* @ user_tick : is the tick from userspace
* @ rq : the pointer to rq
*
* Tick demultiplexing follows the order
* - pending hardirq update
* - pending softirq update
* - user_time
* - idle_time
* - system time
* - check for guest_time
* - else account as system_time
*
* Check for hardirq is done both for system and user time as there is
* no timer going off while we are on hardirq and hence we may never get an
* opportunity to update it solely in system time .
* p - > stime and friends are only updated on system time and not on irq
* softirq as those do not count in task exec_runtime any more .
*/
static void irqtime_account_process_tick ( struct task_struct * p , int user_tick ,
struct rq * rq )
{
cputime_t one_jiffy_scaled = cputime_to_scaled ( cputime_one_jiffy ) ;
u64 * cpustat = kcpustat_this_cpu - > cpustat ;
if ( steal_account_process_tick ( ) )
return ;
if ( irqtime_account_hi_update ( ) ) {
cpustat [ CPUTIME_IRQ ] + = ( __force u64 ) cputime_one_jiffy ;
} else if ( irqtime_account_si_update ( ) ) {
cpustat [ CPUTIME_SOFTIRQ ] + = ( __force u64 ) cputime_one_jiffy ;
} else if ( this_cpu_ksoftirqd ( ) = = p ) {
/*
* ksoftirqd time do not get accounted in cpu_softirq_time .
* So , we have to handle it separately here .
* Also , p - > stime needs to be updated for ksoftirqd .
*/
__account_system_time ( p , cputime_one_jiffy , one_jiffy_scaled ,
CPUTIME_SOFTIRQ ) ;
} else if ( user_tick ) {
account_user_time ( p , cputime_one_jiffy , one_jiffy_scaled ) ;
} else if ( p = = rq - > idle ) {
account_idle_time ( cputime_one_jiffy ) ;
} else if ( p - > flags & PF_VCPU ) { /* System time or guest time */
account_guest_time ( p , cputime_one_jiffy , one_jiffy_scaled ) ;
} else {
__account_system_time ( p , cputime_one_jiffy , one_jiffy_scaled ,
CPUTIME_SYSTEM ) ;
}
}
static void irqtime_account_idle_ticks ( int ticks )
{
int i ;
struct rq * rq = this_rq ( ) ;
for ( i = 0 ; i < ticks ; i + + )
irqtime_account_process_tick ( current , 0 , rq ) ;
}
# else /* CONFIG_IRQ_TIME_ACCOUNTING */
2012-07-16 20:00:34 +04:00
static inline void irqtime_account_idle_ticks ( int ticks ) { }
static inline void irqtime_account_process_tick ( struct task_struct * p , int user_tick ,
2012-06-16 17:57:37 +04:00
struct rq * rq ) { }
# endif /* CONFIG_IRQ_TIME_ACCOUNTING */
/*
* Use precise platform statistics if available :
*/
# ifdef CONFIG_VIRT_CPU_ACCOUNTING
2012-09-08 18:14:02 +04:00
2012-11-14 03:24:25 +04:00
# ifndef __ARCH_HAS_VTIME_TASK_SWITCH
2013-07-12 05:10:15 +04:00
void vtime_common_task_switch ( struct task_struct * prev )
2012-11-14 03:24:25 +04:00
{
if ( is_idle_task ( prev ) )
vtime_account_idle ( prev ) ;
else
vtime_account_system ( prev ) ;
2012-07-25 09:56:04 +04:00
# ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
2012-11-14 03:24:25 +04:00
vtime_account_user ( prev ) ;
2012-07-25 09:56:04 +04:00
# endif
2012-11-14 03:24:25 +04:00
arch_vtime_task_switch ( prev ) ;
}
# endif
2012-10-24 20:05:51 +04:00
2012-09-08 18:14:02 +04:00
/*
* Archs that account the whole time spent in the idle task
* ( outside irq ) as idle time can rely on this and just implement
2012-11-13 21:21:22 +04:00
* vtime_account_system ( ) and vtime_account_idle ( ) . Archs that
2012-09-08 18:14:02 +04:00
* have other meaning of the idle time ( s390 only includes the
* time spent by the CPU when it ' s in low power mode ) must override
* vtime_account ( ) .
*/
# ifndef __ARCH_HAS_VTIME_ACCOUNT
2013-07-12 05:10:15 +04:00
void vtime_common_account_irq_enter ( struct task_struct * tsk )
2012-09-08 18:14:02 +04:00
{
2012-07-25 09:56:04 +04:00
if ( ! in_interrupt ( ) ) {
/*
* If we interrupted user , context_tracking_in_user ( )
* is 1 because the context tracking don ' t hook
* on irq entry / exit . This way we know if
* we need to flush user time on kernel entry .
*/
if ( context_tracking_in_user ( ) ) {
vtime_account_user ( tsk ) ;
return ;
}
if ( is_idle_task ( tsk ) ) {
vtime_account_idle ( tsk ) ;
return ;
}
}
vtime_account_system ( tsk ) ;
2012-09-08 18:14:02 +04:00
}
2013-07-12 05:10:15 +04:00
EXPORT_SYMBOL_GPL ( vtime_common_account_irq_enter ) ;
2012-09-08 18:14:02 +04:00
# endif /* __ARCH_HAS_VTIME_ACCOUNT */
2013-02-25 20:25:39 +04:00
# endif /* CONFIG_VIRT_CPU_ACCOUNTING */
# ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
void task_cputime_adjusted ( struct task_struct * p , cputime_t * ut , cputime_t * st )
{
* ut = p - > utime ;
* st = p - > stime ;
}
2012-09-08 18:14:02 +04:00
2013-02-25 20:25:39 +04:00
void thread_group_cputime_adjusted ( struct task_struct * p , cputime_t * ut , cputime_t * st )
{
struct task_cputime cputime ;
2012-06-16 17:57:37 +04:00
2013-02-25 20:25:39 +04:00
thread_group_cputime ( p , & cputime ) ;
* ut = cputime . utime ;
* st = cputime . stime ;
}
# else /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
/*
* Account a single tick of cpu time .
* @ p : the process that the cpu time gets accounted to
* @ user_tick : indicates if the tick is a user or a system tick
*/
void account_process_tick ( struct task_struct * p , int user_tick )
2012-06-16 17:57:37 +04:00
{
2013-02-25 20:25:39 +04:00
cputime_t one_jiffy_scaled = cputime_to_scaled ( cputime_one_jiffy ) ;
struct rq * rq = this_rq ( ) ;
2012-06-16 17:57:37 +04:00
2013-02-25 20:25:39 +04:00
if ( vtime_accounting_enabled ( ) )
return ;
if ( sched_clock_irqtime ) {
irqtime_account_process_tick ( p , user_tick , rq ) ;
return ;
}
if ( steal_account_process_tick ( ) )
return ;
2012-06-16 17:57:37 +04:00
2013-02-25 20:25:39 +04:00
if ( user_tick )
account_user_time ( p , cputime_one_jiffy , one_jiffy_scaled ) ;
else if ( ( p ! = rq - > idle ) | | ( irq_count ( ) ! = HARDIRQ_OFFSET ) )
account_system_time ( p , HARDIRQ_OFFSET , cputime_one_jiffy ,
one_jiffy_scaled ) ;
2012-06-16 17:57:37 +04:00
else
2013-02-25 20:25:39 +04:00
account_idle_time ( cputime_one_jiffy ) ;
}
2012-06-16 17:57:37 +04:00
2013-02-25 20:25:39 +04:00
/*
* Account multiple ticks of steal time .
* @ p : the process from which the cpu time has been stolen
* @ ticks : number of stolen ticks
*/
void account_steal_ticks ( unsigned long ticks )
{
account_steal_time ( jiffies_to_cputime ( ticks ) ) ;
}
/*
* Account multiple ticks of idle time .
* @ ticks : number of stolen ticks
*/
void account_idle_ticks ( unsigned long ticks )
{
if ( sched_clock_irqtime ) {
irqtime_account_idle_ticks ( ticks ) ;
return ;
}
account_idle_time ( jiffies_to_cputime ( ticks ) ) ;
}
2012-06-16 17:57:37 +04:00
2013-02-20 21:54:55 +04:00
/*
2013-04-30 19:14:42 +04:00
* Perform ( stime * rtime ) / total , but avoid multiplication overflow by
* loosing precision when the numbers are big .
2013-02-20 21:54:55 +04:00
*/
static cputime_t scale_stime ( u64 stime , u64 rtime , u64 total )
2012-06-16 17:57:37 +04:00
{
2013-04-30 19:14:42 +04:00
u64 scaled ;
2012-06-16 17:57:37 +04:00
2013-04-30 19:14:42 +04:00
for ( ; ; ) {
/* Make sure "rtime" is the bigger of stime/rtime */
2013-05-02 17:34:33 +04:00
if ( stime > rtime )
swap ( rtime , stime ) ;
2013-04-30 19:14:42 +04:00
/* Make sure 'total' fits in 32 bits */
if ( total > > 32 )
goto drop_precision ;
/* Does rtime (and thus stime) fit in 32 bits? */
if ( ! ( rtime > > 32 ) )
break ;
/* Can we just balance rtime/stime rather than dropping bits? */
if ( stime > > 31 )
goto drop_precision ;
/* We can grow stime and shrink rtime and try to make them both fit */
stime < < = 1 ;
rtime > > = 1 ;
continue ;
drop_precision :
/* We drop from rtime, it has more bits than stime */
rtime > > = 1 ;
total > > = 1 ;
2013-02-20 21:54:55 +04:00
}
2012-06-16 17:57:37 +04:00
2013-04-30 19:14:42 +04:00
/*
* Make sure gcc understands that this is a 32 x32 - > 64 multiply ,
* followed by a 64 / 32 - > 64 divide .
*/
scaled = div_u64 ( ( u64 ) ( u32 ) stime * ( u64 ) ( u32 ) rtime , ( u32 ) total ) ;
2013-02-20 21:54:55 +04:00
return ( __force cputime_t ) scaled ;
2012-06-16 17:57:37 +04:00
}
2012-11-28 20:00:57 +04:00
/*
* Adjust tick based cputime random precision against scheduler
* runtime accounting .
*/
2012-11-22 03:58:35 +04:00
static void cputime_adjust ( struct task_cputime * curr ,
struct cputime * prev ,
cputime_t * ut , cputime_t * st )
2012-06-16 17:57:37 +04:00
{
2013-04-30 13:35:06 +04:00
cputime_t rtime , stime , utime , total ;
2012-06-16 17:57:37 +04:00
2013-01-26 20:19:42 +04:00
stime = curr - > stime ;
total = stime + curr - > utime ;
2012-11-28 20:00:57 +04:00
2012-06-16 17:57:37 +04:00
/*
2012-11-28 20:00:57 +04:00
* Tick based cputime accounting depend on random scheduling
* timeslices of a task to be interrupted or not by the timer .
* Depending on these circumstances , the number of these interrupts
* may be over or under - optimistic , matching the real user and system
* cputime with a variable precision .
*
* Fix this by scaling these tick based values against the total
* runtime accounted by the CFS scheduler .
2012-06-16 17:57:37 +04:00
*/
2012-11-22 03:58:35 +04:00
rtime = nsecs_to_cputime ( curr - > sum_exec_runtime ) ;
2012-06-16 17:57:37 +04:00
2013-04-30 13:35:05 +04:00
/*
* Update userspace visible utime / stime values only if actual execution
* time is bigger than already exported . Note that can happen , that we
* provided bigger values due to scaling inaccuracy on big numbers .
*/
if ( prev - > stime + prev - > utime > = rtime )
goto out ;
2013-04-30 13:35:06 +04:00
if ( total ) {
2013-02-20 21:54:55 +04:00
stime = scale_stime ( ( __force u64 ) stime ,
( __force u64 ) rtime , ( __force u64 ) total ) ;
2013-04-30 13:35:06 +04:00
utime = rtime - stime ;
} else {
stime = rtime ;
utime = 0 ;
2013-02-20 21:54:55 +04:00
}
2012-06-16 17:57:37 +04:00
/*
2012-11-28 20:00:57 +04:00
* If the tick based count grows faster than the scheduler one ,
* the result of the scaling may go backward .
* Let ' s enforce monotonicity .
2012-06-16 17:57:37 +04:00
*/
2013-01-26 20:19:42 +04:00
prev - > stime = max ( prev - > stime , stime ) ;
2013-04-30 13:35:06 +04:00
prev - > utime = max ( prev - > utime , utime ) ;
2012-11-22 03:58:35 +04:00
2013-04-30 13:35:05 +04:00
out :
2012-11-22 03:58:35 +04:00
* ut = prev - > utime ;
* st = prev - > stime ;
}
2012-06-16 17:57:37 +04:00
2012-11-22 03:58:35 +04:00
void task_cputime_adjusted ( struct task_struct * p , cputime_t * ut , cputime_t * st )
{
struct task_cputime cputime = {
. sum_exec_runtime = p - > se . sum_exec_runtime ,
} ;
2012-11-13 17:20:55 +04:00
task_cputime ( p , & cputime . utime , & cputime . stime ) ;
2012-11-22 03:58:35 +04:00
cputime_adjust ( & cputime , & p - > prev_cputime , ut , st ) ;
2012-06-16 17:57:37 +04:00
}
/*
* Must be called with siglock held .
*/
2012-11-21 19:26:44 +04:00
void thread_group_cputime_adjusted ( struct task_struct * p , cputime_t * ut , cputime_t * st )
2012-06-16 17:57:37 +04:00
{
struct task_cputime cputime ;
thread_group_cputime ( p , & cputime ) ;
2012-11-22 03:58:35 +04:00
cputime_adjust ( & cputime , & p - > signal - > prev_cputime , ut , st ) ;
2012-06-16 17:57:37 +04:00
}
2013-02-25 20:25:39 +04:00
# endif /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
2012-07-25 09:56:04 +04:00
# ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
2012-12-16 23:00:34 +04:00
static unsigned long long vtime_delta ( struct task_struct * tsk )
{
unsigned long long clock ;
2013-02-23 20:28:45 +04:00
clock = local_clock ( ) ;
2012-12-16 23:00:34 +04:00
if ( clock < tsk - > vtime_snap )
return 0 ;
2012-07-25 09:56:04 +04:00
2012-12-16 23:00:34 +04:00
return clock - tsk - > vtime_snap ;
}
static cputime_t get_vtime_delta ( struct task_struct * tsk )
2012-07-25 09:56:04 +04:00
{
2012-12-16 23:00:34 +04:00
unsigned long long delta = vtime_delta ( tsk ) ;
2012-07-25 09:56:04 +04:00
2012-12-16 23:00:34 +04:00
WARN_ON_ONCE ( tsk - > vtime_snap_whence = = VTIME_SLEEPING ) ;
tsk - > vtime_snap + = delta ;
2012-07-25 09:56:04 +04:00
/* CHECKME: always safe to convert nsecs to cputime? */
return nsecs_to_cputime ( delta ) ;
}
2012-12-16 23:00:34 +04:00
static void __vtime_account_system ( struct task_struct * tsk )
{
cputime_t delta_cpu = get_vtime_delta ( tsk ) ;
account_system_time ( tsk , irq_count ( ) , delta_cpu , cputime_to_scaled ( delta_cpu ) ) ;
}
2012-07-25 09:56:04 +04:00
void vtime_account_system ( struct task_struct * tsk )
{
2012-12-16 23:00:34 +04:00
write_seqlock ( & tsk - > vtime_seqlock ) ;
__vtime_account_system ( tsk ) ;
write_sequnlock ( & tsk - > vtime_seqlock ) ;
}
2012-07-16 20:00:34 +04:00
2013-07-12 05:10:15 +04:00
void vtime_gen_account_irq_exit ( struct task_struct * tsk )
2012-12-16 23:00:34 +04:00
{
write_seqlock ( & tsk - > vtime_seqlock ) ;
if ( context_tracking_in_user ( ) )
tsk - > vtime_snap_whence = VTIME_USER ;
__vtime_account_system ( tsk ) ;
write_sequnlock ( & tsk - > vtime_seqlock ) ;
2012-07-25 09:56:04 +04:00
}
void vtime_account_user ( struct task_struct * tsk )
{
2013-07-13 19:10:18 +04:00
cputime_t delta_cpu ;
2012-07-25 09:56:04 +04:00
2012-12-16 23:00:34 +04:00
write_seqlock ( & tsk - > vtime_seqlock ) ;
2013-07-13 19:10:18 +04:00
delta_cpu = get_vtime_delta ( tsk ) ;
2012-12-16 23:00:34 +04:00
tsk - > vtime_snap_whence = VTIME_SYS ;
2012-07-25 09:56:04 +04:00
account_user_time ( tsk , delta_cpu , cputime_to_scaled ( delta_cpu ) ) ;
2012-12-16 23:00:34 +04:00
write_sequnlock ( & tsk - > vtime_seqlock ) ;
}
void vtime_user_enter ( struct task_struct * tsk )
{
write_seqlock ( & tsk - > vtime_seqlock ) ;
tsk - > vtime_snap_whence = VTIME_USER ;
__vtime_account_system ( tsk ) ;
write_sequnlock ( & tsk - > vtime_seqlock ) ;
}
void vtime_guest_enter ( struct task_struct * tsk )
{
2013-07-12 21:05:14 +04:00
/*
* The flags must be updated under the lock with
* the vtime_snap flush and update .
* That enforces a right ordering and update sequence
* synchronization against the reader ( task_gtime ( ) )
* that can thus safely catch up with a tickless delta .
*/
2012-12-16 23:00:34 +04:00
write_seqlock ( & tsk - > vtime_seqlock ) ;
__vtime_account_system ( tsk ) ;
current - > flags | = PF_VCPU ;
write_sequnlock ( & tsk - > vtime_seqlock ) ;
}
2013-07-10 04:44:35 +04:00
EXPORT_SYMBOL_GPL ( vtime_guest_enter ) ;
2012-12-16 23:00:34 +04:00
void vtime_guest_exit ( struct task_struct * tsk )
{
write_seqlock ( & tsk - > vtime_seqlock ) ;
__vtime_account_system ( tsk ) ;
current - > flags & = ~ PF_VCPU ;
write_sequnlock ( & tsk - > vtime_seqlock ) ;
2012-07-25 09:56:04 +04:00
}
2013-07-10 04:44:35 +04:00
EXPORT_SYMBOL_GPL ( vtime_guest_exit ) ;
2012-07-25 09:56:04 +04:00
void vtime_account_idle ( struct task_struct * tsk )
{
2012-12-16 23:00:34 +04:00
cputime_t delta_cpu = get_vtime_delta ( tsk ) ;
2012-07-25 09:56:04 +04:00
account_idle_time ( delta_cpu ) ;
}
2012-07-16 20:00:34 +04:00
2012-12-16 23:00:34 +04:00
void arch_vtime_task_switch ( struct task_struct * prev )
{
write_seqlock ( & prev - > vtime_seqlock ) ;
prev - > vtime_snap_whence = VTIME_SLEEPING ;
write_sequnlock ( & prev - > vtime_seqlock ) ;
write_seqlock ( & current - > vtime_seqlock ) ;
current - > vtime_snap_whence = VTIME_SYS ;
2013-05-16 00:16:32 +04:00
current - > vtime_snap = sched_clock_cpu ( smp_processor_id ( ) ) ;
2012-12-16 23:00:34 +04:00
write_sequnlock ( & current - > vtime_seqlock ) ;
}
2013-05-16 00:16:32 +04:00
void vtime_init_idle ( struct task_struct * t , int cpu )
2012-12-16 23:00:34 +04:00
{
unsigned long flags ;
write_seqlock_irqsave ( & t - > vtime_seqlock , flags ) ;
t - > vtime_snap_whence = VTIME_SYS ;
2013-05-16 00:16:32 +04:00
t - > vtime_snap = sched_clock_cpu ( cpu ) ;
2012-12-16 23:00:34 +04:00
write_sequnlock_irqrestore ( & t - > vtime_seqlock , flags ) ;
}
cputime_t task_gtime ( struct task_struct * t )
{
unsigned int seq ;
cputime_t gtime ;
do {
2013-02-16 02:47:07 +04:00
seq = read_seqbegin ( & t - > vtime_seqlock ) ;
2012-12-16 23:00:34 +04:00
gtime = t - > gtime ;
if ( t - > flags & PF_VCPU )
gtime + = vtime_delta ( t ) ;
2013-02-16 02:47:07 +04:00
} while ( read_seqretry ( & t - > vtime_seqlock , seq ) ) ;
2012-12-16 23:00:34 +04:00
return gtime ;
}
/*
* Fetch cputime raw values from fields of task_struct and
* add up the pending nohz execution time since the last
* cputime snapshot .
*/
static void
fetch_task_cputime ( struct task_struct * t ,
cputime_t * u_dst , cputime_t * s_dst ,
cputime_t * u_src , cputime_t * s_src ,
cputime_t * udelta , cputime_t * sdelta )
{
unsigned int seq ;
unsigned long long delta ;
do {
* udelta = 0 ;
* sdelta = 0 ;
2013-02-16 02:47:07 +04:00
seq = read_seqbegin ( & t - > vtime_seqlock ) ;
2012-12-16 23:00:34 +04:00
if ( u_dst )
* u_dst = * u_src ;
if ( s_dst )
* s_dst = * s_src ;
/* Task is sleeping, nothing to add */
if ( t - > vtime_snap_whence = = VTIME_SLEEPING | |
is_idle_task ( t ) )
continue ;
delta = vtime_delta ( t ) ;
/*
* Task runs either in user or kernel space , add pending nohz time to
* the right place .
*/
if ( t - > vtime_snap_whence = = VTIME_USER | | t - > flags & PF_VCPU ) {
* udelta = delta ;
} else {
if ( t - > vtime_snap_whence = = VTIME_SYS )
* sdelta = delta ;
}
2013-02-16 02:47:07 +04:00
} while ( read_seqretry ( & t - > vtime_seqlock , seq ) ) ;
2012-12-16 23:00:34 +04:00
}
void task_cputime ( struct task_struct * t , cputime_t * utime , cputime_t * stime )
{
cputime_t udelta , sdelta ;
fetch_task_cputime ( t , utime , stime , & t - > utime ,
& t - > stime , & udelta , & sdelta ) ;
if ( utime )
* utime + = udelta ;
if ( stime )
* stime + = sdelta ;
}
void task_cputime_scaled ( struct task_struct * t ,
cputime_t * utimescaled , cputime_t * stimescaled )
{
cputime_t udelta , sdelta ;
fetch_task_cputime ( t , utimescaled , stimescaled ,
& t - > utimescaled , & t - > stimescaled , & udelta , & sdelta ) ;
if ( utimescaled )
* utimescaled + = cputime_to_scaled ( udelta ) ;
if ( stimescaled )
* stimescaled + = cputime_to_scaled ( sdelta ) ;
}
2012-07-25 09:56:04 +04:00
# endif /* CONFIG_VIRT_CPU_ACCOUNTING_GEN */