* [PATCH 01/37] jiffies: Reuse TICK_NSEC instead of NSEC_PER_JIFFY
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 02/37] time: Introduce jiffies64_to_nsecs() Frederic Weisbecker
` (35 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
NSEC_PER_JIFFY is an ad-hoc redefinition of TICK_NSEC. Let's rather
use a unique and well maintained version.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/time/jiffies.c | 32 ++++++++++++++++----------------
1 file changed, 16 insertions(+), 16 deletions(-)
diff --git a/kernel/time/jiffies.c b/kernel/time/jiffies.c
index a4a0e47..7906b3f 100644
--- a/kernel/time/jiffies.c
+++ b/kernel/time/jiffies.c
@@ -27,19 +27,8 @@
#include "timekeeping.h"
-/* The Jiffies based clocksource is the lowest common
- * denominator clock source which should function on
- * all systems. It has the same coarse resolution as
- * the timer interrupt frequency HZ and it suffers
- * inaccuracies caused by missed or lost timer
- * interrupts and the inability for the timer
- * interrupt hardware to accuratly tick at the
- * requested HZ value. It is also not recommended
- * for "tick-less" systems.
- */
-#define NSEC_PER_JIFFY ((NSEC_PER_SEC+HZ/2)/HZ)
-/* Since jiffies uses a simple NSEC_PER_JIFFY multiplier
+/* Since jiffies uses a simple TICK_NSEC multiplier
* conversion, the .shift value could be zero. However
* this would make NTP adjustments impossible as they are
* in units of 1/2^.shift. Thus we use JIFFIES_SHIFT to
@@ -47,8 +36,8 @@
* amount, and give ntp adjustments in units of 1/2^8
*
* The value 8 is somewhat carefully chosen, as anything
- * larger can result in overflows. NSEC_PER_JIFFY grows as
- * HZ shrinks, so values greater than 8 overflow 32bits when
+ * larger can result in overflows. TICK_NSEC grows as HZ
+ * shrinks, so values greater than 8 overflow 32bits when
* HZ=100.
*/
#if HZ < 34
@@ -64,12 +53,23 @@ static u64 jiffies_read(struct clocksource *cs)
return (u64) jiffies;
}
+/*
+ * The Jiffies based clocksource is the lowest common
+ * denominator clock source which should function on
+ * all systems. It has the same coarse resolution as
+ * the timer interrupt frequency HZ and it suffers
+ * inaccuracies caused by missed or lost timer
+ * interrupts and the inability for the timer
+ * interrupt hardware to accuratly tick at the
+ * requested HZ value. It is also not recommended
+ * for "tick-less" systems.
+ */
static struct clocksource clocksource_jiffies = {
.name = "jiffies",
.rating = 1, /* lowest valid rating*/
.read = jiffies_read,
.mask = CLOCKSOURCE_MASK(32),
- .mult = NSEC_PER_JIFFY << JIFFIES_SHIFT, /* details above */
+ .mult = TICK_NSEC << JIFFIES_SHIFT, /* details above */
.shift = JIFFIES_SHIFT,
.max_cycles = 10,
};
@@ -125,7 +125,7 @@ int register_refined_jiffies(long cycles_per_second)
shift_hz += cycles_per_tick/2;
do_div(shift_hz, cycles_per_tick);
/* Calculate nsec_per_tick using shift_hz */
- nsec_per_tick = (u64)NSEC_PER_SEC << 8;
+ nsec_per_tick = (u64)TICK_NSEC << 8;
nsec_per_tick += (u32)shift_hz/2;
do_div(nsec_per_tick, (u32)shift_hz);
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 02/37] time: Introduce jiffies64_to_nsecs()
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 01/37] jiffies: Reuse TICK_NSEC instead of NSEC_PER_JIFFY Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 03/37] sched: Remove unused INIT_CPUTIME macro Frederic Weisbecker
` (34 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This will be needed for the cputime_t to nsec conversion.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/jiffies.h | 2 ++
kernel/time/time.c | 10 ++++++++++
kernel/time/timeconst.bc | 6 ++++++
3 files changed, 18 insertions(+)
diff --git a/include/linux/jiffies.h b/include/linux/jiffies.h
index 589d14e..624215c 100644
--- a/include/linux/jiffies.h
+++ b/include/linux/jiffies.h
@@ -293,6 +293,8 @@ static inline u64 jiffies_to_nsecs(const unsigned long j)
return (u64)jiffies_to_usecs(j) * NSEC_PER_USEC;
}
+extern u64 jiffies64_to_nsecs(u64 j);
+
extern unsigned long __msecs_to_jiffies(const unsigned int m);
#if HZ <= MSEC_PER_SEC && !(MSEC_PER_SEC % HZ)
/*
diff --git a/kernel/time/time.c b/kernel/time/time.c
index a3a9a8a..25bdd25 100644
--- a/kernel/time/time.c
+++ b/kernel/time/time.c
@@ -702,6 +702,16 @@ u64 nsec_to_clock_t(u64 x)
#endif
}
+u64 jiffies64_to_nsecs(u64 j)
+{
+#if !(NSEC_PER_SEC % HZ)
+ return (NSEC_PER_SEC / HZ) * j;
+# else
+ return div_u64(j * HZ_TO_NSEC_NUM, HZ_TO_NSEC_DEN);
+#endif
+}
+EXPORT_SYMBOL(jiffies64_to_nsecs);
+
/**
* nsecs_to_jiffies64 - Convert nsecs in u64 to jiffies64
*
diff --git a/kernel/time/timeconst.bc b/kernel/time/timeconst.bc
index c486889..f83bbb8 100644
--- a/kernel/time/timeconst.bc
+++ b/kernel/time/timeconst.bc
@@ -98,6 +98,12 @@ define timeconst(hz) {
print "#define HZ_TO_USEC_DEN\t\t", hz/cd, "\n"
print "#define USEC_TO_HZ_NUM\t\t", hz/cd, "\n"
print "#define USEC_TO_HZ_DEN\t\t", 1000000/cd, "\n"
+
+ cd=gcd(hz,1000000000)
+ print "#define HZ_TO_NSEC_NUM\t\t", 1000000000/cd, "\n"
+ print "#define HZ_TO_NSEC_DEN\t\t", hz/cd, "\n"
+ print "#define NSEC_TO_HZ_NUM\t\t", hz/cd, "\n"
+ print "#define NSEC_TO_HZ_DEN\t\t", 1000000000/cd, "\n"
print "\n"
print "#endif /* KERNEL_TIMECONST_H */\n"
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 03/37] sched: Remove unused INIT_CPUTIME macro
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 01/37] jiffies: Reuse TICK_NSEC instead of NSEC_PER_JIFFY Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 02/37] time: Introduce jiffies64_to_nsecs() Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 04/37] cputime: Convert kcpustat to nsecs Frederic Weisbecker
` (33 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
It's a leftover from removed code.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/sched.h | 7 -------
1 file changed, 7 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index a8daed9..195f919 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -620,13 +620,6 @@ struct task_cputime {
#define prof_exp stime
#define sched_exp sum_exec_runtime
-#define INIT_CPUTIME \
- (struct task_cputime) { \
- .utime = 0, \
- .stime = 0, \
- .sum_exec_runtime = 0, \
- }
-
/*
* This is the atomic variant of task_cputime, which can be used for
* storing and updating task_cputime statistics without locking.
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 04/37] cputime: Convert kcpustat to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (2 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 03/37] sched: Remove unused INIT_CPUTIME macro Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 05/37] macintosh/rack-meter: Remove cputime_t internal use Frederic Weisbecker
` (32 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Kernel cpu stats are stored in cputime_t which is an architecture
defined type, and hence a bit opaque and requiring accessors and mutators
for any operation.
Converting them to nsecs simplifies the code and is one step toward
the removal of cputime_t in the core code.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/s390/appldata/appldata_os.c | 16 ++++-----
drivers/cpufreq/cpufreq.c | 6 ++--
drivers/cpufreq/cpufreq_governor.c | 2 +-
drivers/cpufreq/cpufreq_stats.c | 1 -
drivers/macintosh/rack-meter.c | 2 +-
fs/proc/stat.c | 68 +++++++++++++++++++-------------------
fs/proc/uptime.c | 7 ++--
kernel/sched/cpuacct.c | 2 +-
kernel/sched/cputime.c | 22 ++++++------
9 files changed, 61 insertions(+), 65 deletions(-)
diff --git a/arch/s390/appldata/appldata_os.c b/arch/s390/appldata/appldata_os.c
index 69b23b2..08b9e94 100644
--- a/arch/s390/appldata/appldata_os.c
+++ b/arch/s390/appldata/appldata_os.c
@@ -113,21 +113,21 @@ static void appldata_get_os_data(void *data)
j = 0;
for_each_online_cpu(i) {
os_data->os_cpu[j].per_cpu_user =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_USER]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_USER]);
os_data->os_cpu[j].per_cpu_nice =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_NICE]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_NICE]);
os_data->os_cpu[j].per_cpu_system =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SYSTEM]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SYSTEM]);
os_data->os_cpu[j].per_cpu_idle =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IDLE]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IDLE]);
os_data->os_cpu[j].per_cpu_irq =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IRQ]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IRQ]);
os_data->os_cpu[j].per_cpu_softirq =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SOFTIRQ]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_SOFTIRQ]);
os_data->os_cpu[j].per_cpu_iowait =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IOWAIT]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_IOWAIT]);
os_data->os_cpu[j].per_cpu_steal =
- cputime_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_STEAL]);
+ nsecs_to_jiffies(kcpustat_cpu(i).cpustat[CPUTIME_STEAL]);
os_data->os_cpu[j].cpu_id = i;
j++;
}
diff --git a/drivers/cpufreq/cpufreq.c b/drivers/cpufreq/cpufreq.c
index cc475ef..3e9b319 100644
--- a/drivers/cpufreq/cpufreq.c
+++ b/drivers/cpufreq/cpufreq.c
@@ -132,7 +132,7 @@ static inline u64 get_cpu_idle_time_jiffy(unsigned int cpu, u64 *wall)
u64 cur_wall_time;
u64 busy_time;
- cur_wall_time = jiffies64_to_cputime64(get_jiffies_64());
+ cur_wall_time = jiffies64_to_nsecs(get_jiffies_64());
busy_time = kcpustat_cpu(cpu).cpustat[CPUTIME_USER];
busy_time += kcpustat_cpu(cpu).cpustat[CPUTIME_SYSTEM];
@@ -143,9 +143,9 @@ static inline u64 get_cpu_idle_time_jiffy(unsigned int cpu, u64 *wall)
idle_time = cur_wall_time - busy_time;
if (wall)
- *wall = cputime_to_usecs(cur_wall_time);
+ *wall = div_u64(cur_wall_time, NSEC_PER_USEC);
- return cputime_to_usecs(idle_time);
+ return div_u64(idle_time, NSEC_PER_USEC);
}
u64 get_cpu_idle_time(unsigned int cpu, u64 *wall, int io_busy)
diff --git a/drivers/cpufreq/cpufreq_governor.c b/drivers/cpufreq/cpufreq_governor.c
index 0196467..631bd2c 100644
--- a/drivers/cpufreq/cpufreq_governor.c
+++ b/drivers/cpufreq/cpufreq_governor.c
@@ -152,7 +152,7 @@ unsigned int dbs_update(struct cpufreq_policy *policy)
if (ignore_nice) {
u64 cur_nice = kcpustat_cpu(j).cpustat[CPUTIME_NICE];
- idle_time += cputime_to_usecs(cur_nice - j_cdbs->prev_cpu_nice);
+ idle_time += div_u64(cur_nice - j_cdbs->prev_cpu_nice, NSEC_PER_USEC);
j_cdbs->prev_cpu_nice = cur_nice;
}
diff --git a/drivers/cpufreq/cpufreq_stats.c b/drivers/cpufreq/cpufreq_stats.c
index ac284e6..17048bb 100644
--- a/drivers/cpufreq/cpufreq_stats.c
+++ b/drivers/cpufreq/cpufreq_stats.c
@@ -13,7 +13,6 @@
#include <linux/cpufreq.h>
#include <linux/module.h>
#include <linux/slab.h>
-#include <linux/cputime.h>
static DEFINE_SPINLOCK(cpufreq_stats_lock);
diff --git a/drivers/macintosh/rack-meter.c b/drivers/macintosh/rack-meter.c
index 7755271..c114594 100644
--- a/drivers/macintosh/rack-meter.c
+++ b/drivers/macintosh/rack-meter.c
@@ -91,7 +91,7 @@ static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
if (rackmeter_ignore_nice)
retval += kcpustat_cpu(cpu).cpustat[CPUTIME_NICE];
- return retval;
+ return nsecs_to_cputime64(retval);
}
static void rackmeter_setup_i2s(struct rackmeter *rm)
diff --git a/fs/proc/stat.c b/fs/proc/stat.c
index d700c42..44475a4 100644
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@ -21,23 +21,23 @@
#ifdef arch_idle_time
-static cputime64_t get_idle_time(int cpu)
+static u64 get_idle_time(int cpu)
{
- cputime64_t idle;
+ u64 idle;
idle = kcpustat_cpu(cpu).cpustat[CPUTIME_IDLE];
if (cpu_online(cpu) && !nr_iowait_cpu(cpu))
- idle += arch_idle_time(cpu);
+ idle += cputime_to_nsecs(arch_idle_time(cpu));
return idle;
}
-static cputime64_t get_iowait_time(int cpu)
+static u64 get_iowait_time(int cpu)
{
- cputime64_t iowait;
+ u64 iowait;
iowait = kcpustat_cpu(cpu).cpustat[CPUTIME_IOWAIT];
if (cpu_online(cpu) && nr_iowait_cpu(cpu))
- iowait += arch_idle_time(cpu);
+ iowait += cputime_to_nsecs(arch_idle_time(cpu));
return iowait;
}
@@ -45,32 +45,32 @@ static cputime64_t get_iowait_time(int cpu)
static u64 get_idle_time(int cpu)
{
- u64 idle, idle_time = -1ULL;
+ u64 idle, idle_usecs = -1ULL;
if (cpu_online(cpu))
- idle_time = get_cpu_idle_time_us(cpu, NULL);
+ idle_usecs = get_cpu_idle_time_us(cpu, NULL);
- if (idle_time == -1ULL)
+ if (idle_usecs == -1ULL)
/* !NO_HZ or cpu offline so we can rely on cpustat.idle */
idle = kcpustat_cpu(cpu).cpustat[CPUTIME_IDLE];
else
- idle = usecs_to_cputime64(idle_time);
+ idle = idle_usecs * NSEC_PER_USEC;
return idle;
}
static u64 get_iowait_time(int cpu)
{
- u64 iowait, iowait_time = -1ULL;
+ u64 iowait, iowait_usecs = -1ULL;
if (cpu_online(cpu))
- iowait_time = get_cpu_iowait_time_us(cpu, NULL);
+ iowait_usecs = get_cpu_iowait_time_us(cpu, NULL);
- if (iowait_time == -1ULL)
+ if (iowait_usecs == -1ULL)
/* !NO_HZ or cpu offline so we can rely on cpustat.iowait */
iowait = kcpustat_cpu(cpu).cpustat[CPUTIME_IOWAIT];
else
- iowait = usecs_to_cputime64(iowait_time);
+ iowait = iowait_usecs * NSEC_PER_USEC;
return iowait;
}
@@ -115,16 +115,16 @@ static int show_stat(struct seq_file *p, void *v)
}
sum += arch_irq_stat();
- seq_put_decimal_ull(p, "cpu ", cputime64_to_clock_t(user));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(nice));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(system));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(idle));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(iowait));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(irq));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(softirq));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(steal));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(guest));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(guest_nice));
+ seq_put_decimal_ull(p, "cpu ", nsec_to_clock_t(user));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(nice));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(system));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(idle));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(iowait));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(irq));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(softirq));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(steal));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(guest));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(guest_nice));
seq_putc(p, '\n');
for_each_online_cpu(i) {
@@ -140,16 +140,16 @@ static int show_stat(struct seq_file *p, void *v)
guest = kcpustat_cpu(i).cpustat[CPUTIME_GUEST];
guest_nice = kcpustat_cpu(i).cpustat[CPUTIME_GUEST_NICE];
seq_printf(p, "cpu%d", i);
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(user));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(nice));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(system));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(idle));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(iowait));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(irq));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(softirq));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(steal));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(guest));
- seq_put_decimal_ull(p, " ", cputime64_to_clock_t(guest_nice));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(user));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(nice));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(system));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(idle));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(iowait));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(irq));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(softirq));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(steal));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(guest));
+ seq_put_decimal_ull(p, " ", nsec_to_clock_t(guest_nice));
seq_putc(p, '\n');
}
seq_put_decimal_ull(p, "intr ", (unsigned long long)sum);
diff --git a/fs/proc/uptime.c b/fs/proc/uptime.c
index 33de567..7981c4f 100644
--- a/fs/proc/uptime.c
+++ b/fs/proc/uptime.c
@@ -5,23 +5,20 @@
#include <linux/seq_file.h>
#include <linux/time.h>
#include <linux/kernel_stat.h>
-#include <linux/cputime.h>
static int uptime_proc_show(struct seq_file *m, void *v)
{
struct timespec uptime;
struct timespec idle;
- u64 idletime;
u64 nsec;
u32 rem;
int i;
- idletime = 0;
+ nsec = 0;
for_each_possible_cpu(i)
- idletime += (__force u64) kcpustat_cpu(i).cpustat[CPUTIME_IDLE];
+ nsec += (__force u64) kcpustat_cpu(i).cpustat[CPUTIME_IDLE];
get_monotonic_boottime(&uptime);
- nsec = cputime64_to_jiffies64(idletime) * TICK_NSEC;
idle.tv_sec = div_u64_rem(nsec, NSEC_PER_SEC, &rem);
idle.tv_nsec = rem;
seq_printf(m, "%lu.%02lu %lu.%02lu\n",
diff --git a/kernel/sched/cpuacct.c b/kernel/sched/cpuacct.c
index 9add206..f95ab29 100644
--- a/kernel/sched/cpuacct.c
+++ b/kernel/sched/cpuacct.c
@@ -297,7 +297,7 @@ static int cpuacct_stats_show(struct seq_file *sf, void *v)
for (stat = 0; stat < CPUACCT_STAT_NSTATS; stat++) {
seq_printf(sf, "%s %lld\n",
cpuacct_stat_desc[stat],
- (long long)cputime64_to_clock_t(val[stat]));
+ (long long)nsec_to_clock_t(val[stat]));
}
return 0;
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index f7c14cc..61e2709 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -75,9 +75,9 @@ static cputime_t irqtime_account_update(u64 irqtime, int idx, cputime_t maxtime)
u64 *cpustat = kcpustat_this_cpu->cpustat;
cputime_t irq_cputime;
- irq_cputime = nsecs_to_cputime64(irqtime) - cpustat[idx];
+ irq_cputime = nsecs_to_cputime64(irqtime - cpustat[idx]);
irq_cputime = min(irq_cputime, maxtime);
- cpustat[idx] += irq_cputime;
+ cpustat[idx] += cputime_to_nsecs(irq_cputime);
return irq_cputime;
}
@@ -140,7 +140,7 @@ void account_user_time(struct task_struct *p, cputime_t cputime)
index = (task_nice(p) > 0) ? CPUTIME_NICE : CPUTIME_USER;
/* Add user time to cpustat. */
- task_group_account_field(p, index, (__force u64) cputime);
+ task_group_account_field(p, index, cputime_to_nsecs(cputime));
/* Account for user time used */
acct_account_cputime(p);
@@ -162,11 +162,11 @@ void account_guest_time(struct task_struct *p, cputime_t cputime)
/* Add guest time to cpustat. */
if (task_nice(p) > 0) {
- cpustat[CPUTIME_NICE] += (__force u64) cputime;
- cpustat[CPUTIME_GUEST_NICE] += (__force u64) cputime;
+ cpustat[CPUTIME_NICE] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_GUEST_NICE] += cputime_to_nsecs(cputime);
} else {
- cpustat[CPUTIME_USER] += (__force u64) cputime;
- cpustat[CPUTIME_GUEST] += (__force u64) cputime;
+ cpustat[CPUTIME_USER] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_GUEST] += cputime_to_nsecs(cputime);
}
}
@@ -184,7 +184,7 @@ void account_system_index_time(struct task_struct *p,
account_group_system_time(p, cputime);
/* Add system time to cpustat. */
- task_group_account_field(p, index, (__force u64) cputime);
+ task_group_account_field(p, index, cputime_to_nsecs(cputime));
/* Account for system time used */
acct_account_cputime(p);
@@ -224,7 +224,7 @@ void account_steal_time(cputime_t cputime)
{
u64 *cpustat = kcpustat_this_cpu->cpustat;
- cpustat[CPUTIME_STEAL] += (__force u64) cputime;
+ cpustat[CPUTIME_STEAL] += cputime_to_nsecs(cputime);
}
/*
@@ -237,9 +237,9 @@ void account_idle_time(cputime_t cputime)
struct rq *rq = this_rq();
if (atomic_read(&rq->nr_iowait) > 0)
- cpustat[CPUTIME_IOWAIT] += (__force u64) cputime;
+ cpustat[CPUTIME_IOWAIT] += cputime_to_nsecs(cputime);
else
- cpustat[CPUTIME_IDLE] += (__force u64) cputime;
+ cpustat[CPUTIME_IDLE] += cputime_to_nsecs(cputime);
}
/*
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 05/37] macintosh/rack-meter: Remove cputime_t internal use
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (3 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 04/37] cputime: Convert kcpustat to nsecs Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 06/37] cputime: Convert guest time accounting to nsecs Frederic Weisbecker
` (31 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
cputime_t is being obsolete and replaced by nsecs units.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
drivers/macintosh/rack-meter.c | 30 +++++++++++++++---------------
1 file changed, 15 insertions(+), 15 deletions(-)
diff --git a/drivers/macintosh/rack-meter.c b/drivers/macintosh/rack-meter.c
index c114594..e199fd6 100644
--- a/drivers/macintosh/rack-meter.c
+++ b/drivers/macintosh/rack-meter.c
@@ -52,8 +52,8 @@ struct rackmeter_dma {
struct rackmeter_cpu {
struct delayed_work sniffer;
struct rackmeter *rm;
- cputime64_t prev_wall;
- cputime64_t prev_idle;
+ u64 prev_wall;
+ u64 prev_idle;
int zero;
} ____cacheline_aligned;
@@ -81,7 +81,7 @@ static int rackmeter_ignore_nice;
/* This is copied from cpufreq_ondemand, maybe we should put it in
* a common header somewhere
*/
-static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
+static inline u64 get_cpu_idle_time(unsigned int cpu)
{
u64 retval;
@@ -91,7 +91,7 @@ static inline cputime64_t get_cpu_idle_time(unsigned int cpu)
if (rackmeter_ignore_nice)
retval += kcpustat_cpu(cpu).cpustat[CPUTIME_NICE];
- return nsecs_to_cputime64(retval);
+ return retval;
}
static void rackmeter_setup_i2s(struct rackmeter *rm)
@@ -217,23 +217,23 @@ static void rackmeter_do_timer(struct work_struct *work)
container_of(work, struct rackmeter_cpu, sniffer.work);
struct rackmeter *rm = rcpu->rm;
unsigned int cpu = smp_processor_id();
- cputime64_t cur_jiffies, total_idle_ticks;
- unsigned int total_ticks, idle_ticks;
+ u64 cur_nsecs, total_idle_nsecs;
+ u64 total_nsecs, idle_nsecs;
int i, offset, load, cumm, pause;
- cur_jiffies = jiffies64_to_cputime64(get_jiffies_64());
- total_ticks = (unsigned int) (cur_jiffies - rcpu->prev_wall);
- rcpu->prev_wall = cur_jiffies;
+ cur_nsecs = jiffies64_to_nsecs(get_jiffies_64());
+ total_nsecs = cur_nsecs - rcpu->prev_wall;
+ rcpu->prev_wall = cur_nsecs;
- total_idle_ticks = get_cpu_idle_time(cpu);
- idle_ticks = (unsigned int) (total_idle_ticks - rcpu->prev_idle);
- idle_ticks = min(idle_ticks, total_ticks);
- rcpu->prev_idle = total_idle_ticks;
+ total_idle_nsecs = get_cpu_idle_time(cpu);
+ idle_nsecs = total_idle_nsecs - rcpu->prev_idle;
+ idle_nsecs = min(idle_nsecs, total_nsecs);
+ rcpu->prev_idle = total_idle_nsecs;
/* We do a very dumb calculation to update the LEDs for now,
* we'll do better once we have actual PWM implemented
*/
- load = (9 * (total_ticks - idle_ticks)) / total_ticks;
+ load = div64_u64(9 * (total_nsecs - idle_nsecs), total_nsecs);
offset = cpu << 3;
cumm = 0;
@@ -278,7 +278,7 @@ static void rackmeter_init_cpu_sniffer(struct rackmeter *rm)
continue;
rcpu = &rm->cpu[cpu];
rcpu->prev_idle = get_cpu_idle_time(cpu);
- rcpu->prev_wall = jiffies64_to_cputime64(get_jiffies_64());
+ rcpu->prev_wall = jiffies64_to_nsecs(get_jiffies_64());
schedule_delayed_work_on(cpu, &rm->cpu[cpu].sniffer,
msecs_to_jiffies(CPU_SAMPLING_RATE));
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 06/37] cputime: Convert guest time accounting to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (4 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 05/37] macintosh/rack-meter: Remove cputime_t internal use Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 07/37] cputime: Special API to return old-typed cputime Frederic Weisbecker
` (30 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
cputime_t is being obsolete and replaced by nsecs units in order to make
internal timestamps less opaque and more granular.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
fs/proc/array.c | 6 +++---
include/linux/sched.h | 10 +++++-----
kernel/sched/cputime.c | 8 ++++----
3 files changed, 12 insertions(+), 12 deletions(-)
diff --git a/fs/proc/array.c b/fs/proc/array.c
index 51a4213..25b54cf 100644
--- a/fs/proc/array.c
+++ b/fs/proc/array.c
@@ -402,7 +402,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns,
unsigned long cmin_flt = 0, cmaj_flt = 0;
unsigned long min_flt = 0, maj_flt = 0;
cputime_t cutime, cstime, utime, stime;
- cputime_t cgtime, gtime;
+ u64 cgtime, gtime;
unsigned long rsslim = 0;
char tcomm[sizeof(task->comm)];
unsigned long flags;
@@ -542,8 +542,8 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns,
seq_put_decimal_ull(m, " ", task->rt_priority);
seq_put_decimal_ull(m, " ", task->policy);
seq_put_decimal_ull(m, " ", delayacct_blkio_ticks(task));
- seq_put_decimal_ull(m, " ", cputime_to_clock_t(gtime));
- seq_put_decimal_ll(m, " ", cputime_to_clock_t(cgtime));
+ seq_put_decimal_ull(m, " ", nsec_to_clock_t(gtime));
+ seq_put_decimal_ll(m, " ", nsec_to_clock_t(cgtime));
if (mm && permitted) {
seq_put_decimal_ull(m, " ", mm->start_data);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 195f919..7d268ce 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -774,8 +774,8 @@ struct signal_struct {
*/
seqlock_t stats_lock;
cputime_t utime, stime, cutime, cstime;
- cputime_t gtime;
- cputime_t cgtime;
+ u64 gtime;
+ u64 cgtime;
struct prev_cputime prev_cputime;
unsigned long nvcsw, nivcsw, cnvcsw, cnivcsw;
unsigned long min_flt, maj_flt, cmin_flt, cmaj_flt;
@@ -1658,7 +1658,7 @@ struct task_struct {
#ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
cputime_t utimescaled, stimescaled;
#endif
- cputime_t gtime;
+ u64 gtime;
struct prev_cputime prev_cputime;
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
seqcount_t vtime_seqcount;
@@ -2254,7 +2254,7 @@ struct task_struct *try_get_task_struct(struct task_struct **ptask);
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
extern void task_cputime(struct task_struct *t,
cputime_t *utime, cputime_t *stime);
-extern cputime_t task_gtime(struct task_struct *t);
+extern u64 task_gtime(struct task_struct *t);
#else
static inline void task_cputime(struct task_struct *t,
cputime_t *utime, cputime_t *stime)
@@ -2263,7 +2263,7 @@ static inline void task_cputime(struct task_struct *t,
*stime = t->stime;
}
-static inline cputime_t task_gtime(struct task_struct *t)
+static inline u64 task_gtime(struct task_struct *t)
{
return t->gtime;
}
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 61e2709..8bcd98e 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -158,7 +158,7 @@ void account_guest_time(struct task_struct *p, cputime_t cputime)
/* Add guest time to process. */
p->utime += cputime;
account_group_user_time(p, cputime);
- p->gtime += cputime;
+ p->gtime += cputime_to_nsecs(cputime);
/* Add guest time to cpustat. */
if (task_nice(p) > 0) {
@@ -824,10 +824,10 @@ void vtime_init_idle(struct task_struct *t, int cpu)
local_irq_restore(flags);
}
-cputime_t task_gtime(struct task_struct *t)
+u64 task_gtime(struct task_struct *t)
{
unsigned int seq;
- cputime_t gtime;
+ u64 gtime;
if (!vtime_accounting_enabled())
return t->gtime;
@@ -837,7 +837,7 @@ cputime_t task_gtime(struct task_struct *t)
gtime = t->gtime;
if (t->vtime_snap_whence == VTIME_SYS && t->flags & PF_VCPU)
- gtime += vtime_delta(t);
+ gtime += cputime_to_nsecs(vtime_delta(t));
} while (read_seqcount_retry(&t->vtime_seqcount, seq));
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 07/37] cputime: Special API to return old-typed cputime
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (5 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 06/37] cputime: Convert guest time accounting to nsecs Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 08/37] cputime: Convert task/group cputime to nsecs Frederic Weisbecker
` (29 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This API returns the tasks cputime in cputime_t in order to ease the
conversion of cputime internals to use nsecs units instead. Blindly
converting all cputime readers to use this API now will later let us
convert more smoothly and step by step all these places to use the
new nsec based cputime.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/alpha/kernel/osf_sys.c | 2 +-
arch/x86/kernel/apm_32.c | 2 +-
drivers/isdn/mISDN/stack.c | 2 +-
fs/binfmt_elf.c | 6 +++---
fs/binfmt_elf_fdpic.c | 6 +++---
include/linux/sched.h | 32 ++++++++++++++++++++++++++---
kernel/acct.c | 2 +-
kernel/delayacct.c | 4 ++--
kernel/signal.c | 4 ++--
kernel/time/itimer.c | 2 +-
kernel/time/posix-cpu-timers.c | 46 +++++++++++++++++++++---------------------
kernel/tsacct.c | 6 +++---
12 files changed, 70 insertions(+), 44 deletions(-)
diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c
index 54d86166..0f92438 100644
--- a/arch/alpha/kernel/osf_sys.c
+++ b/arch/alpha/kernel/osf_sys.c
@@ -1154,7 +1154,7 @@ SYSCALL_DEFINE2(osf_getrusage, int, who, struct rusage32 __user *, ru)
memset(&r, 0, sizeof(r));
switch (who) {
case RUSAGE_SELF:
- task_cputime(current, &utime, &stime);
+ task_cputime_t(current, &utime, &stime);
utime_jiffies = cputime_to_jiffies(utime);
stime_jiffies = cputime_to_jiffies(stime);
jiffies_to_timeval32(utime_jiffies, &r.ru_utime);
diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c
index 45d44c1..89c84fc 100644
--- a/arch/x86/kernel/apm_32.c
+++ b/arch/x86/kernel/apm_32.c
@@ -913,7 +913,7 @@ static int apm_cpu_idle(struct cpuidle_device *dev,
unsigned int bucket;
recalc:
- task_cputime(current, &utime, &stime);
+ task_cputime_t(current, &utime, &stime);
if (jiffies_since_last_check > IDLE_CALC_LIMIT) {
use_apm_idle = 0;
} else if (jiffies_since_last_check > idle_period) {
diff --git a/drivers/isdn/mISDN/stack.c b/drivers/isdn/mISDN/stack.c
index 9cb4b62..0a36617 100644
--- a/drivers/isdn/mISDN/stack.c
+++ b/drivers/isdn/mISDN/stack.c
@@ -306,7 +306,7 @@ mISDNStackd(void *data)
"msg %d sleep %d stopped\n",
dev_name(&st->dev->dev), st->msg_cnt, st->sleep_cnt,
st->stopped_cnt);
- task_cputime(st->thread, &utime, &stime);
+ task_cputime_t(st->thread, &utime, &stime);
printk(KERN_DEBUG
"mISDNStackd daemon for %s utime(%ld) stime(%ld)\n",
dev_name(&st->dev->dev), utime, stime);
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 29a02da..58b9b09 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1421,19 +1421,19 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
/*
* This is the record for the group leader. It shows the
* group-wide total, not its individual thread total.
*/
- thread_group_cputime(p, &cputime);
+ thread_group_cputime_t(p, &cputime);
cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
} else {
cputime_t utime, stime;
- task_cputime(p, &utime, &stime);
+ task_cputime_t(p, &utime, &stime);
cputime_to_timeval(utime, &prstatus->pr_utime);
cputime_to_timeval(stime, &prstatus->pr_stime);
}
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index d2e36f8..6ccd9df 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -1342,19 +1342,19 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
/*
* This is the record for the group leader. It shows the
* group-wide total, not its individual thread total.
*/
- thread_group_cputime(p, &cputime);
+ thread_group_cputime_t(p, &cputime);
cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
} else {
cputime_t utime, stime;
- task_cputime(p, &utime, &stime);
+ task_cputime_t(p, &utime, &stime);
cputime_to_timeval(utime, &prstatus->pr_utime);
cputime_to_timeval(stime, &prstatus->pr_stime);
}
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 7d268ce..10d2d03 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -615,6 +615,13 @@ struct task_cputime {
unsigned long long sum_exec_runtime;
};
+/* Temporary type to ease cputime_t to nsecs conversion */
+struct task_cputime_t {
+ cputime_t utime;
+ cputime_t stime;
+ unsigned long long sum_exec_runtime;
+};
+
/* Alternate field names when used to cache expirations. */
#define virt_exp utime
#define prof_exp stime
@@ -748,7 +755,7 @@ struct signal_struct {
struct thread_group_cputimer cputimer;
/* Earliest-expiration cache. */
- struct task_cputime cputime_expires;
+ struct task_cputime_t cputime_expires;
#ifdef CONFIG_NO_HZ_FULL
atomic_t tick_dep_mask;
@@ -1682,7 +1689,7 @@ struct task_struct {
/* mm fault and swap info: this can arguably be seen as either mm-specific or thread-specific */
unsigned long min_flt, maj_flt;
- struct task_cputime cputime_expires;
+ struct task_cputime_t cputime_expires;
struct list_head cpu_timers[3];
/* process credentials */
@@ -2286,6 +2293,19 @@ static inline void task_cputime_scaled(struct task_struct *t,
}
#endif
+static inline void task_cputime_t(struct task_struct *t,
+ cputime_t *utime, cputime_t *stime)
+{
+ task_cputime(t, utime, stime);
+}
+
+static inline void task_cputime_t_scaled(struct task_struct *t,
+ cputime_t *utimescaled,
+ cputime_t *stimescaled)
+{
+ task_cputime_scaled(t, utimescaled, stimescaled);
+}
+
extern void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st);
extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st);
@@ -3500,7 +3520,13 @@ static __always_inline bool need_resched(void)
* Thread group CPU time accounting.
*/
void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times);
-void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
+void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times);
+
+static inline void thread_group_cputime_t(struct task_struct *tsk,
+ struct task_cputime_t *times)
+{
+ thread_group_cputime(tsk, (struct task_cputime *)times);
+}
/*
* Reevaluate whether the task has signals pending delivery.
diff --git a/kernel/acct.c b/kernel/acct.c
index 74963d1..b9b190a8 100644
--- a/kernel/acct.c
+++ b/kernel/acct.c
@@ -559,7 +559,7 @@ void acct_collect(long exitcode, int group_dead)
pacct->ac_flag |= ACORE;
if (current->flags & PF_SIGNALED)
pacct->ac_flag |= AXSIG;
- task_cputime(current, &utime, &stime);
+ task_cputime_t(current, &utime, &stime);
pacct->ac_utime += utime;
pacct->ac_stime += stime;
pacct->ac_minflt += current->min_flt;
diff --git a/kernel/delayacct.c b/kernel/delayacct.c
index 435c14a..228640f 100644
--- a/kernel/delayacct.c
+++ b/kernel/delayacct.c
@@ -87,12 +87,12 @@ int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk)
unsigned long flags, t1;
s64 tmp;
- task_cputime(tsk, &utime, &stime);
+ task_cputime_t(tsk, &utime, &stime);
tmp = (s64)d->cpu_run_real_total;
tmp += cputime_to_nsecs(utime + stime);
d->cpu_run_real_total = (tmp < (s64)d->cpu_run_real_total) ? 0 : tmp;
- task_cputime_scaled(tsk, &utimescaled, &stimescaled);
+ task_cputime_t_scaled(tsk, &utimescaled, &stimescaled);
tmp = (s64)d->cpu_scaled_run_real_total;
tmp += cputime_to_nsecs(utimescaled + stimescaled);
d->cpu_scaled_run_real_total =
diff --git a/kernel/signal.c b/kernel/signal.c
index 3603d93..218048a 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1619,7 +1619,7 @@ bool do_notify_parent(struct task_struct *tsk, int sig)
task_uid(tsk));
rcu_read_unlock();
- task_cputime(tsk, &utime, &stime);
+ task_cputime_t(tsk, &utime, &stime);
info.si_utime = cputime_to_clock_t(utime + tsk->signal->utime);
info.si_stime = cputime_to_clock_t(stime + tsk->signal->stime);
@@ -1704,7 +1704,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk,
info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
rcu_read_unlock();
- task_cputime(tsk, &utime, &stime);
+ task_cputime_t(tsk, &utime, &stime);
info.si_utime = cputime_to_clock_t(utime);
info.si_stime = cputime_to_clock_t(stime);
diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c
index 8c89143..f2d5097 100644
--- a/kernel/time/itimer.c
+++ b/kernel/time/itimer.c
@@ -53,7 +53,7 @@ static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id,
cval = it->expires;
cinterval = it->incr;
if (cval) {
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
cputime_t t;
thread_group_cputimer(tsk, &cputime);
diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c
index e9e8c10..d53ff71 100644
--- a/kernel/time/posix-cpu-timers.c
+++ b/kernel/time/posix-cpu-timers.c
@@ -115,7 +115,7 @@ static void bump_cpu_timer(struct k_itimer *timer,
* Checks @cputime to see if all fields are zero. Returns true if all fields
* are zero, false if any field is nonzero.
*/
-static inline int task_cputime_zero(const struct task_cputime *cputime)
+static inline int task_cputime_zero(const struct task_cputime_t *cputime)
{
if (!cputime->utime && !cputime->stime && !cputime->sum_exec_runtime)
return 1;
@@ -126,7 +126,7 @@ static inline unsigned long long prof_ticks(struct task_struct *p)
{
cputime_t utime, stime;
- task_cputime(p, &utime, &stime);
+ task_cputime_t(p, &utime, &stime);
return cputime_to_expires(utime + stime);
}
@@ -134,7 +134,7 @@ static inline unsigned long long virt_ticks(struct task_struct *p)
{
cputime_t utime, stime;
- task_cputime(p, &utime, &stime);
+ task_cputime_t(p, &utime, &stime);
return cputime_to_expires(utime);
}
@@ -210,7 +210,7 @@ static inline void __update_gt_cputime(atomic64_t *cputime, u64 sum_cputime)
}
}
-static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct task_cputime *sum)
+static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct task_cputime_t *sum)
{
__update_gt_cputime(&cputime_atomic->utime, sum->utime);
__update_gt_cputime(&cputime_atomic->stime, sum->stime);
@@ -218,7 +218,7 @@ static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct
}
/* Sample task_cputime_atomic values in "atomic_timers", store results in "times". */
-static inline void sample_cputime_atomic(struct task_cputime *times,
+static inline void sample_cputime_atomic(struct task_cputime_t *times,
struct task_cputime_atomic *atomic_times)
{
times->utime = atomic64_read(&atomic_times->utime);
@@ -226,10 +226,10 @@ static inline void sample_cputime_atomic(struct task_cputime *times,
times->sum_exec_runtime = atomic64_read(&atomic_times->sum_exec_runtime);
}
-void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
+void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times)
{
struct thread_group_cputimer *cputimer = &tsk->signal->cputimer;
- struct task_cputime sum;
+ struct task_cputime_t sum;
/* Check if cputimer isn't running. This is accessed without locking. */
if (!READ_ONCE(cputimer->running)) {
@@ -238,7 +238,7 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
* values through the TIMER_ABSTIME flag, therefore we have
* to synchronize the timer to the clock every time we start it.
*/
- thread_group_cputime(tsk, &sum);
+ thread_group_cputime_t(tsk, &sum);
update_gt_cputime(&cputimer->cputime_atomic, &sum);
/*
@@ -262,21 +262,21 @@ static int cpu_clock_sample_group(const clockid_t which_clock,
struct task_struct *p,
unsigned long long *sample)
{
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
switch (CPUCLOCK_WHICH(which_clock)) {
default:
return -EINVAL;
case CPUCLOCK_PROF:
- thread_group_cputime(p, &cputime);
+ thread_group_cputime_t(p, &cputime);
*sample = cputime_to_expires(cputime.utime + cputime.stime);
break;
case CPUCLOCK_VIRT:
- thread_group_cputime(p, &cputime);
+ thread_group_cputime_t(p, &cputime);
*sample = cputime_to_expires(cputime.utime);
break;
case CPUCLOCK_SCHED:
- thread_group_cputime(p, &cputime);
+ thread_group_cputime_t(p, &cputime);
*sample = cputime.sum_exec_runtime;
break;
}
@@ -466,7 +466,7 @@ static void arm_timer(struct k_itimer *timer)
{
struct task_struct *p = timer->it.cpu.task;
struct list_head *head, *listpos;
- struct task_cputime *cputime_expires;
+ struct task_cputime_t *cputime_expires;
struct cpu_timer_list *const nt = &timer->it.cpu;
struct cpu_timer_list *next;
@@ -562,7 +562,7 @@ static int cpu_timer_sample_group(const clockid_t which_clock,
struct task_struct *p,
unsigned long long *sample)
{
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
thread_group_cputimer(p, &cputime);
switch (CPUCLOCK_WHICH(which_clock)) {
@@ -761,7 +761,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
/*
* Protect against sighand release/switch in exit/exec and
* also make timer sampling safe if it ends up calling
- * thread_group_cputime().
+ * thread_group_cputime_t().
*/
sighand = lock_task_sighand(p, &flags);
if (unlikely(sighand == NULL)) {
@@ -826,7 +826,7 @@ static void check_thread_timers(struct task_struct *tsk,
{
struct list_head *timers = tsk->cpu_timers;
struct signal_struct *const sig = tsk->signal;
- struct task_cputime *tsk_expires = &tsk->cputime_expires;
+ struct task_cputime_t *tsk_expires = &tsk->cputime_expires;
unsigned long long expires;
unsigned long soft;
@@ -934,7 +934,7 @@ static void check_process_timers(struct task_struct *tsk,
unsigned long long utime, ptime, virt_expires, prof_expires;
unsigned long long sum_sched_runtime, sched_expires;
struct list_head *timers = sig->cpu_timers;
- struct task_cputime cputime;
+ struct task_cputime_t cputime;
unsigned long soft;
/*
@@ -1037,7 +1037,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
} else {
/*
* Protect arm_timer() and timer sampling in case of call to
- * thread_group_cputime().
+ * thread_group_cputime_t().
*/
sighand = lock_task_sighand(p, &flags);
if (unlikely(sighand == NULL)) {
@@ -1080,8 +1080,8 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
* Returns true if any field of the former is greater than the corresponding
* field of the latter if the latter field is set. Otherwise returns false.
*/
-static inline int task_cputime_expired(const struct task_cputime *sample,
- const struct task_cputime *expires)
+static inline int task_cputime_expired(const struct task_cputime_t *sample,
+ const struct task_cputime_t *expires)
{
if (expires->utime && sample->utime >= expires->utime)
return 1;
@@ -1108,9 +1108,9 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
struct signal_struct *sig;
if (!task_cputime_zero(&tsk->cputime_expires)) {
- struct task_cputime task_sample;
+ struct task_cputime_t task_sample;
- task_cputime(tsk, &task_sample.utime, &task_sample.stime);
+ task_cputime_t(tsk, &task_sample.utime, &task_sample.stime);
task_sample.sum_exec_runtime = tsk->se.sum_exec_runtime;
if (task_cputime_expired(&task_sample, &tsk->cputime_expires))
return 1;
@@ -1133,7 +1133,7 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
*/
if (READ_ONCE(sig->cputimer.running) &&
!READ_ONCE(sig->cputimer.checking_timer)) {
- struct task_cputime group_sample;
+ struct task_cputime_t group_sample;
sample_cputime_atomic(&group_sample, &sig->cputimer.cputime_atomic);
diff --git a/kernel/tsacct.c b/kernel/tsacct.c
index f8e26ab..040d0a6 100644
--- a/kernel/tsacct.c
+++ b/kernel/tsacct.c
@@ -66,11 +66,11 @@ void bacct_add_tsk(struct user_namespace *user_ns,
task_tgid_nr_ns(rcu_dereference(tsk->real_parent), pid_ns) : 0;
rcu_read_unlock();
- task_cputime(tsk, &utime, &stime);
+ task_cputime_t(tsk, &utime, &stime);
stats->ac_utime = cputime_to_usecs(utime);
stats->ac_stime = cputime_to_usecs(stime);
- task_cputime_scaled(tsk, &utimescaled, &stimescaled);
+ task_cputime_t_scaled(tsk, &utimescaled, &stimescaled);
stats->ac_utimescaled = cputime_to_usecs(utimescaled);
stats->ac_stimescaled = cputime_to_usecs(stimescaled);
@@ -159,7 +159,7 @@ void acct_update_integrals(struct task_struct *tsk)
unsigned long flags;
local_irq_save(flags);
- task_cputime(tsk, &utime, &stime);
+ task_cputime_t(tsk, &utime, &stime);
__acct_update_integrals(tsk, utime, stime);
local_irq_restore(flags);
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 08/37] cputime: Convert task/group cputime to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (6 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 07/37] cputime: Special API to return old-typed cputime Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-28 11:57 ` Stanislaw Gruszka
2017-01-22 18:19 ` [PATCH 09/37] alpha: Convert obsolete cputime_t " Frederic Weisbecker
` (28 subsequent siblings)
36 siblings, 1 reply; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Now that most cputime readers use the transition API which return the
task cputime in old style cputime_t, we can safely store the cputime in
nsecs. This will eventually make cputime statistics less opaque and more
granular. Back and forth convertions between cputime_t and nsecs in order
to deal with cputime_t random granularity won't be needed anymore.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/alpha/kernel/osf_sys.c | 4 ++--
arch/powerpc/kernel/time.c | 4 ++--
arch/s390/kernel/vtime.c | 6 ++---
arch/x86/kvm/hyperv.c | 5 +++--
fs/binfmt_elf.c | 11 +++++++--
fs/binfmt_elf_fdpic.c | 4 ++--
fs/proc/array.c | 10 ++++-----
include/linux/sched.h | 55 ++++++++++++++++++++++++++++-----------------
kernel/exit.c | 4 ++--
kernel/sched/cputime.c | 35 ++++++++++++++---------------
kernel/signal.c | 4 ++--
kernel/sys.c | 16 ++++++-------
12 files changed, 89 insertions(+), 69 deletions(-)
diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c
index 0f92438..82ccb43 100644
--- a/arch/alpha/kernel/osf_sys.c
+++ b/arch/alpha/kernel/osf_sys.c
@@ -1163,8 +1163,8 @@ SYSCALL_DEFINE2(osf_getrusage, int, who, struct rusage32 __user *, ru)
r.ru_majflt = current->maj_flt;
break;
case RUSAGE_CHILDREN:
- utime_jiffies = cputime_to_jiffies(current->signal->cutime);
- stime_jiffies = cputime_to_jiffies(current->signal->cstime);
+ utime_jiffies = nsecs_to_jiffies(current->signal->cutime);
+ stime_jiffies = nsecs_to_jiffies(current->signal->cstime);
jiffies_to_timeval32(utime_jiffies, &r.ru_utime);
jiffies_to_timeval32(stime_jiffies, &r.ru_stime);
r.ru_minflt = current->signal->cmin_flt;
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 02e9730..3cca82e 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -396,7 +396,7 @@ void vtime_flush(struct task_struct *tsk)
account_user_time(tsk, acct->utime);
if (acct->utime_scaled)
- tsk->utimescaled += acct->utime_scaled;
+ tsk->utimescaled += cputime_to_nsecs(acct->utime_scaled);
if (acct->gtime)
account_guest_time(tsk, acct->gtime);
@@ -411,7 +411,7 @@ void vtime_flush(struct task_struct *tsk)
account_system_index_time(tsk, acct->stime, CPUTIME_SYSTEM);
if (acct->stime_scaled)
- tsk->stimescaled += acct->stime_scaled;
+ tsk->stimescaled += cputime_to_nsecs(acct->stime_scaled);
if (acct->hardirq_time)
account_system_index_time(tsk, acct->hardirq_time, CPUTIME_IRQ);
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index 0a9e5d6..f2fc274 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -114,7 +114,7 @@ static void account_system_index_scaled(struct task_struct *p,
cputime_t cputime, cputime_t scaled,
enum cpu_usage_stat index)
{
- p->stimescaled += scaled;
+ p->stimescaled += cputime_to_nsecs(scaled);
account_system_index_time(p, cputime, index);
}
@@ -167,12 +167,12 @@ static int do_account_vtime(struct task_struct *tsk)
/* Push account value */
if (user) {
account_user_time(tsk, user);
- tsk->utimescaled += scale_vtime(user);
+ tsk->utimescaled += cputime_to_nsecs(scale_vtime(user));
}
if (guest) {
account_guest_time(tsk, guest);
- tsk->utimescaled += scale_vtime(guest);
+ tsk->utimescaled += cputime_to_nsecs(scale_vtime(guest));
}
if (system)
diff --git a/arch/x86/kvm/hyperv.c b/arch/x86/kvm/hyperv.c
index 1572c35..2ecd7da 100644
--- a/arch/x86/kvm/hyperv.c
+++ b/arch/x86/kvm/hyperv.c
@@ -964,10 +964,11 @@ static int kvm_hv_set_msr_pw(struct kvm_vcpu *vcpu, u32 msr, u64 data,
/* Calculate cpu time spent by current task in 100ns units */
static u64 current_task_runtime_100ns(void)
{
- cputime_t utime, stime;
+ u64 utime, stime;
task_cputime_adjusted(current, &utime, &stime);
- return div_u64(cputime_to_nsecs(utime + stime), 100);
+
+ return div_u64(utime + stime, 100);
}
static int kvm_hv_set_msr(struct kvm_vcpu *vcpu, u32 msr, u64 data, bool host)
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 58b9b09..13cd15d 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1411,6 +1411,8 @@ static void fill_note(struct memelfnote *note, const char *name, int type,
static void fill_prstatus(struct elf_prstatus *prstatus,
struct task_struct *p, long signr)
{
+ struct timeval tv;
+
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
@@ -1437,8 +1439,13 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
cputime_to_timeval(utime, &prstatus->pr_utime);
cputime_to_timeval(stime, &prstatus->pr_stime);
}
- cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime);
- cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime);
+ tv = ns_to_timeval(p->signal->cutime);
+ prstatus->pr_cutime.tv_sec = tv.tv_sec;
+ prstatus->pr_cutime.tv_usec = tv.tv_usec;
+
+ tv = ns_to_timeval(p->signal->cstime);
+ prstatus->pr_cstime.tv_sec = tv.tv_sec;
+ prstatus->pr_cstime.tv_usec = tv.tv_usec;
}
static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index 6ccd9df..e1f3734 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -1358,8 +1358,8 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
cputime_to_timeval(utime, &prstatus->pr_utime);
cputime_to_timeval(stime, &prstatus->pr_stime);
}
- cputime_to_timeval(p->signal->cutime, &prstatus->pr_cutime);
- cputime_to_timeval(p->signal->cstime, &prstatus->pr_cstime);
+ prstatus->pr_cutime = ns_to_timeval(p->signal->cutime);
+ prstatus->pr_cstime = ns_to_timeval(p->signal->cstime);
prstatus->pr_exec_fdpic_loadmap = p->mm->context.exec_fdpic_loadmap;
prstatus->pr_interp_fdpic_loadmap = p->mm->context.interp_fdpic_loadmap;
diff --git a/fs/proc/array.c b/fs/proc/array.c
index 25b54cf..fe12b51 100644
--- a/fs/proc/array.c
+++ b/fs/proc/array.c
@@ -401,7 +401,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns,
unsigned long long start_time;
unsigned long cmin_flt = 0, cmaj_flt = 0;
unsigned long min_flt = 0, maj_flt = 0;
- cputime_t cutime, cstime, utime, stime;
+ u64 cutime, cstime, utime, stime;
u64 cgtime, gtime;
unsigned long rsslim = 0;
char tcomm[sizeof(task->comm)];
@@ -497,10 +497,10 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns,
seq_put_decimal_ull(m, " ", cmin_flt);
seq_put_decimal_ull(m, " ", maj_flt);
seq_put_decimal_ull(m, " ", cmaj_flt);
- seq_put_decimal_ull(m, " ", cputime_to_clock_t(utime));
- seq_put_decimal_ull(m, " ", cputime_to_clock_t(stime));
- seq_put_decimal_ll(m, " ", cputime_to_clock_t(cutime));
- seq_put_decimal_ll(m, " ", cputime_to_clock_t(cstime));
+ seq_put_decimal_ull(m, " ", nsec_to_clock_t(utime));
+ seq_put_decimal_ull(m, " ", nsec_to_clock_t(stime));
+ seq_put_decimal_ll(m, " ", nsec_to_clock_t(cutime));
+ seq_put_decimal_ll(m, " ", nsec_to_clock_t(cstime));
seq_put_decimal_ll(m, " ", priority);
seq_put_decimal_ll(m, " ", nice);
seq_put_decimal_ll(m, " ", num_threads);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 10d2d03..089a4ba 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -585,8 +585,8 @@ struct cpu_itimer {
*/
struct prev_cputime {
#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
- cputime_t utime;
- cputime_t stime;
+ u64 utime;
+ u64 stime;
raw_spinlock_t lock;
#endif
};
@@ -601,8 +601,8 @@ static inline void prev_cputime_init(struct prev_cputime *prev)
/**
* struct task_cputime - collected CPU time counts
- * @utime: time spent in user mode, in &cputime_t units
- * @stime: time spent in kernel mode, in &cputime_t units
+ * @utime: time spent in user mode, in nanoseconds
+ * @stime: time spent in kernel mode, in nanoseconds
* @sum_exec_runtime: total time spent on the CPU, in nanoseconds
*
* This structure groups together three kinds of CPU time that are tracked for
@@ -610,8 +610,8 @@ static inline void prev_cputime_init(struct prev_cputime *prev)
* these counts together and treat all three of them in parallel.
*/
struct task_cputime {
- cputime_t utime;
- cputime_t stime;
+ u64 utime;
+ u64 stime;
unsigned long long sum_exec_runtime;
};
@@ -780,7 +780,7 @@ struct signal_struct {
* in __exit_signal, except for the group leader.
*/
seqlock_t stats_lock;
- cputime_t utime, stime, cutime, cstime;
+ u64 utime, stime, cutime, cstime;
u64 gtime;
u64 cgtime;
struct prev_cputime prev_cputime;
@@ -1661,9 +1661,9 @@ struct task_struct {
int __user *set_child_tid; /* CLONE_CHILD_SETTID */
int __user *clear_child_tid; /* CLONE_CHILD_CLEARTID */
- cputime_t utime, stime;
+ u64 utime, stime;
#ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
- cputime_t utimescaled, stimescaled;
+ u64 utimescaled, stimescaled;
#endif
u64 gtime;
struct prev_cputime prev_cputime;
@@ -2260,11 +2260,11 @@ struct task_struct *try_get_task_struct(struct task_struct **ptask);
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
extern void task_cputime(struct task_struct *t,
- cputime_t *utime, cputime_t *stime);
+ u64 *utime, u64 *stime);
extern u64 task_gtime(struct task_struct *t);
#else
static inline void task_cputime(struct task_struct *t,
- cputime_t *utime, cputime_t *stime)
+ u64 *utime, u64 *stime)
{
*utime = t->utime;
*stime = t->stime;
@@ -2278,16 +2278,16 @@ static inline u64 task_gtime(struct task_struct *t)
#ifdef CONFIG_ARCH_HAS_SCALED_CPUTIME
static inline void task_cputime_scaled(struct task_struct *t,
- cputime_t *utimescaled,
- cputime_t *stimescaled)
+ u64 *utimescaled,
+ u64 *stimescaled)
{
*utimescaled = t->utimescaled;
*stimescaled = t->stimescaled;
}
#else
static inline void task_cputime_scaled(struct task_struct *t,
- cputime_t *utimescaled,
- cputime_t *stimescaled)
+ u64 *utimescaled,
+ u64 *stimescaled)
{
task_cputime(t, utimescaled, stimescaled);
}
@@ -2296,18 +2296,26 @@ static inline void task_cputime_scaled(struct task_struct *t,
static inline void task_cputime_t(struct task_struct *t,
cputime_t *utime, cputime_t *stime)
{
- task_cputime(t, utime, stime);
+ u64 ut, st;
+
+ task_cputime(t, &ut, &st);
+ *utime = nsecs_to_cputime(ut);
+ *stime = nsecs_to_cputime(st);
}
static inline void task_cputime_t_scaled(struct task_struct *t,
cputime_t *utimescaled,
cputime_t *stimescaled)
{
- task_cputime_scaled(t, utimescaled, stimescaled);
+ u64 ut, st;
+
+ task_cputime_scaled(t, &ut, &st);
+ *utimescaled = nsecs_to_cputime(ut);
+ *stimescaled = nsecs_to_cputime(st);
}
-extern void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st);
-extern void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st);
+extern void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st);
+extern void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st);
/*
* Per process flags
@@ -3523,9 +3531,14 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times);
void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times);
static inline void thread_group_cputime_t(struct task_struct *tsk,
- struct task_cputime_t *times)
+ struct task_cputime_t *cputime)
{
- thread_group_cputime(tsk, (struct task_cputime *)times);
+ struct task_cputime times;
+
+ thread_group_cputime(tsk, ×);
+ cputime->utime = nsecs_to_cputime(times.utime);
+ cputime->stime = nsecs_to_cputime(times.stime);
+ cputime->sum_exec_runtime = times.sum_exec_runtime;
}
/*
diff --git a/kernel/exit.c b/kernel/exit.c
index 8f14b86..8e5e213 100644
--- a/kernel/exit.c
+++ b/kernel/exit.c
@@ -86,7 +86,7 @@ static void __exit_signal(struct task_struct *tsk)
bool group_dead = thread_group_leader(tsk);
struct sighand_struct *sighand;
struct tty_struct *uninitialized_var(tty);
- cputime_t utime, stime;
+ u64 utime, stime;
sighand = rcu_dereference_check(tsk->sighand,
lockdep_tasklist_lock_is_held());
@@ -1091,7 +1091,7 @@ static int wait_task_zombie(struct wait_opts *wo, struct task_struct *p)
struct signal_struct *sig = p->signal;
struct signal_struct *psig = current->signal;
unsigned long maxrss;
- cputime_t tgutime, tgstime;
+ u64 tgutime, tgstime;
/*
* The resource counters for the group leader are in its
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 8bcd98e..0bdef50 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -134,7 +134,7 @@ void account_user_time(struct task_struct *p, cputime_t cputime)
int index;
/* Add user time to process. */
- p->utime += cputime;
+ p->utime += cputime_to_nsecs(cputime);
account_group_user_time(p, cputime);
index = (task_nice(p) > 0) ? CPUTIME_NICE : CPUTIME_USER;
@@ -156,7 +156,7 @@ void account_guest_time(struct task_struct *p, cputime_t cputime)
u64 *cpustat = kcpustat_this_cpu->cpustat;
/* Add guest time to process. */
- p->utime += cputime;
+ p->utime += cputime_to_nsecs(cputime);
account_group_user_time(p, cputime);
p->gtime += cputime_to_nsecs(cputime);
@@ -180,7 +180,7 @@ void account_system_index_time(struct task_struct *p,
cputime_t cputime, enum cpu_usage_stat index)
{
/* Add system time to process. */
- p->stime += cputime;
+ p->stime += cputime_to_nsecs(cputime);
account_group_system_time(p, cputime);
/* Add system time to cpustat. */
@@ -315,7 +315,7 @@ static u64 read_sum_exec_runtime(struct task_struct *t)
void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
{
struct signal_struct *sig = tsk->signal;
- cputime_t utime, stime;
+ u64 utime, stime;
struct task_struct *t;
unsigned int seq, nextseq;
unsigned long flags;
@@ -465,14 +465,14 @@ void vtime_account_irq_enter(struct task_struct *tsk)
EXPORT_SYMBOL_GPL(vtime_account_irq_enter);
#endif /* __ARCH_HAS_VTIME_ACCOUNT */
-void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
+void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
{
*ut = p->utime;
*st = p->stime;
}
EXPORT_SYMBOL_GPL(task_cputime_adjusted);
-void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
+void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
{
struct task_cputime cputime;
@@ -543,7 +543,7 @@ void account_idle_ticks(unsigned long ticks)
* Perform (stime * rtime) / total, but avoid multiplication overflow by
* loosing precision when the numbers are big.
*/
-static cputime_t scale_stime(u64 stime, u64 rtime, u64 total)
+static u64 scale_stime(u64 stime, u64 rtime, u64 total)
{
u64 scaled;
@@ -580,7 +580,7 @@ static cputime_t scale_stime(u64 stime, u64 rtime, u64 total)
* followed by a 64/32->64 divide.
*/
scaled = div_u64((u64) (u32) stime * (u64) (u32) rtime, (u32)total);
- return (__force cputime_t) scaled;
+ return scaled;
}
/*
@@ -605,14 +605,14 @@ static cputime_t scale_stime(u64 stime, u64 rtime, u64 total)
*/
static void cputime_adjust(struct task_cputime *curr,
struct prev_cputime *prev,
- cputime_t *ut, cputime_t *st)
+ u64 *ut, u64 *st)
{
- cputime_t rtime, stime, utime;
+ u64 rtime, stime, utime;
unsigned long flags;
/* Serialize concurrent callers such that we can honour our guarantees */
raw_spin_lock_irqsave(&prev->lock, flags);
- rtime = nsecs_to_cputime(curr->sum_exec_runtime);
+ rtime = curr->sum_exec_runtime;
/*
* This is possible under two circumstances:
@@ -643,8 +643,7 @@ static void cputime_adjust(struct task_cputime *curr,
goto update;
}
- stime = scale_stime((__force u64)stime, (__force u64)rtime,
- (__force u64)(stime + utime));
+ stime = scale_stime(stime, rtime, stime + utime);
update:
/*
@@ -677,7 +676,7 @@ static void cputime_adjust(struct task_cputime *curr,
raw_spin_unlock_irqrestore(&prev->lock, flags);
}
-void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
+void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
{
struct task_cputime cputime = {
.sum_exec_runtime = p->se.sum_exec_runtime,
@@ -688,7 +687,7 @@ void task_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
}
EXPORT_SYMBOL_GPL(task_cputime_adjusted);
-void thread_group_cputime_adjusted(struct task_struct *p, cputime_t *ut, cputime_t *st)
+void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
{
struct task_cputime cputime;
@@ -849,9 +848,9 @@ u64 task_gtime(struct task_struct *t)
* add up the pending nohz execution time since the last
* cputime snapshot.
*/
-void task_cputime(struct task_struct *t, cputime_t *utime, cputime_t *stime)
+void task_cputime(struct task_struct *t, u64 *utime, u64 *stime)
{
- cputime_t delta;
+ u64 delta;
unsigned int seq;
if (!vtime_accounting_enabled()) {
@@ -870,7 +869,7 @@ void task_cputime(struct task_struct *t, cputime_t *utime, cputime_t *stime)
if (t->vtime_snap_whence == VTIME_INACTIVE || is_idle_task(t))
continue;
- delta = vtime_delta(t);
+ delta = cputime_to_nsecs(vtime_delta(t));
/*
* Task runs either in user or kernel space, add pending nohz time to
diff --git a/kernel/signal.c b/kernel/signal.c
index 218048a..b635221 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1620,8 +1620,8 @@ bool do_notify_parent(struct task_struct *tsk, int sig)
rcu_read_unlock();
task_cputime_t(tsk, &utime, &stime);
- info.si_utime = cputime_to_clock_t(utime + tsk->signal->utime);
- info.si_stime = cputime_to_clock_t(stime + tsk->signal->stime);
+ info.si_utime = cputime_to_clock_t(utime + nsecs_to_cputime(tsk->signal->utime));
+ info.si_stime = cputime_to_clock_t(stime + nsecs_to_cputime(tsk->signal->stime));
info.si_status = tsk->exit_code & 0x7f;
if (tsk->exit_code & 0x80)
diff --git a/kernel/sys.c b/kernel/sys.c
index 842914e..7d4a9a6 100644
--- a/kernel/sys.c
+++ b/kernel/sys.c
@@ -881,15 +881,15 @@ SYSCALL_DEFINE0(getegid)
void do_sys_times(struct tms *tms)
{
- cputime_t tgutime, tgstime, cutime, cstime;
+ u64 tgutime, tgstime, cutime, cstime;
thread_group_cputime_adjusted(current, &tgutime, &tgstime);
cutime = current->signal->cutime;
cstime = current->signal->cstime;
- tms->tms_utime = cputime_to_clock_t(tgutime);
- tms->tms_stime = cputime_to_clock_t(tgstime);
- tms->tms_cutime = cputime_to_clock_t(cutime);
- tms->tms_cstime = cputime_to_clock_t(cstime);
+ tms->tms_utime = nsec_to_clock_t(tgutime);
+ tms->tms_stime = nsec_to_clock_t(tgstime);
+ tms->tms_cutime = nsec_to_clock_t(cutime);
+ tms->tms_cstime = nsec_to_clock_t(cstime);
}
SYSCALL_DEFINE1(times, struct tms __user *, tbuf)
@@ -1544,7 +1544,7 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r)
{
struct task_struct *t;
unsigned long flags;
- cputime_t tgutime, tgstime, utime, stime;
+ u64 tgutime, tgstime, utime, stime;
unsigned long maxrss = 0;
memset((char *)r, 0, sizeof (*r));
@@ -1600,8 +1600,8 @@ static void k_getrusage(struct task_struct *p, int who, struct rusage *r)
unlock_task_sighand(p, &flags);
out:
- cputime_to_timeval(utime, &r->ru_utime);
- cputime_to_timeval(stime, &r->ru_stime);
+ r->ru_utime = ns_to_timeval(utime);
+ r->ru_stime = ns_to_timeval(stime);
if (who != RUSAGE_CHILDREN) {
struct mm_struct *mm = get_task_mm(p);
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH 08/37] cputime: Convert task/group cputime to nsecs
2017-01-22 18:19 ` [PATCH 08/37] cputime: Convert task/group cputime to nsecs Frederic Weisbecker
@ 2017-01-28 11:57 ` Stanislaw Gruszka
2017-01-28 15:28 ` Frederic Weisbecker
0 siblings, 1 reply; 45+ messages in thread
From: Stanislaw Gruszka @ 2017-01-28 11:57 UTC (permalink / raw)
To: Frederic Weisbecker, Peter Zijlstra
Cc: LKML, Tony Luck, Wanpeng Li, Michael Ellerman, Heiko Carstens,
Benjamin Herrenschmidt, Thomas Gleixner, Paul Mackerras,
Ingo Molnar, Fenghua Yu, Rik van Riel, Martin Schwidefsky
Hi Frederic and sorry for late comment.
On Sun, Jan 22, 2017 at 07:19:44PM +0100, Frederic Weisbecker wrote:
> Now that most cputime readers use the transition API which return the
> task cputime in old style cputime_t, we can safely store the cputime in
> nsecs. This will eventually make cputime statistics less opaque and more
> granular. Back and forth convertions between cputime_t and nsecs in order
> to deal with cputime_t random granularity won't be needed anymore.
<snip>
> - cputime_t utime;
> - cputime_t stime;
> + u64 utime;
> + u64 stime;
> unsigned long long sum_exec_runtime;
> };
<snip>
> @@ -134,7 +134,7 @@ void account_user_time(struct task_struct *p, cputime_t cputime)
> int index;
>
> /* Add user time to process. */
> - p->utime += cputime;
> + p->utime += cputime_to_nsecs(cputime);
> account_group_user_time(p, cputime);
<snip>
> +void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
> {
> *ut = p->utime;
> *st = p->stime;
> o }
On 32 bit architectures 64bit store/load is not atomic and if not
protected - 64bit variables can be mangled. I do not see any protection
(lock) between utime/stime store and load in the patch and seems that
{u/s}time store/load can be performed at the same time. Though problem
is very very improbable it still can happen at least theoretically when
lower and upper 32 bits are changed at the same time i.e. process
{u,s}time become near to multiple of 2**32 nsec (aprox: 4sec) and
64bit {u,s}time is stored and loaded at the same time on different
cpus. As said this is very improbable situation, but eventually could
be possible on long lived processes.
BTW we have already similar problem with sum_exec_runtime. I posted
some patches to solve the problem, but non of them was good:
- https://lkml.org/lkml/2016/9/1/172
this one slow down scheduler hot path's and Peter hates it.
- https://lkml.org/lkml/2016/9/6/305
this one was fine for Peter, but I dislike it for taking
task_rq_lock() and do not push it forward.
I considering fixing problem of sum_exec_runtime possible mangling
by using prev_sum_exec_runtime:
u64 read_sum_exec_runtime(struct task_struct *t)
{
u64 ns, prev_ns;
do {
prev_ns = READ_ONCE(t->se.prev_sum_exec_runtime);
ns = READ_ONCE(t->se.sum_exec_runtime);
} while (ns < prev_ns || ns > (prev_ns + U32_MAX));
return ns;
}
This should work based on fact that prev_sum_exec_runtime and
sum_exec_runtime are not modified and stored at the same time, so only
one of those variabled can be mangled. Though I need to think about
correctnes of that a bit more.
Stanislaw
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH 08/37] cputime: Convert task/group cputime to nsecs
2017-01-28 11:57 ` Stanislaw Gruszka
@ 2017-01-28 15:28 ` Frederic Weisbecker
2017-01-30 13:54 ` Stanislaw Gruszka
0 siblings, 1 reply; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-28 15:28 UTC (permalink / raw)
To: Stanislaw Gruszka
Cc: Peter Zijlstra, LKML, Tony Luck, Wanpeng Li, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Martin Schwidefsky
On Sat, Jan 28, 2017 at 12:57:40PM +0100, Stanislaw Gruszka wrote:
> Hi Frederic and sorry for late comment.
>
> On Sun, Jan 22, 2017 at 07:19:44PM +0100, Frederic Weisbecker wrote:
> > Now that most cputime readers use the transition API which return the
> > task cputime in old style cputime_t, we can safely store the cputime in
> > nsecs. This will eventually make cputime statistics less opaque and more
> > granular. Back and forth convertions between cputime_t and nsecs in order
> > to deal with cputime_t random granularity won't be needed anymore.
> <snip>
> > - cputime_t utime;
> > - cputime_t stime;
> > + u64 utime;
> > + u64 stime;
> > unsigned long long sum_exec_runtime;
> > };
> <snip>
> > @@ -134,7 +134,7 @@ void account_user_time(struct task_struct *p, cputime_t cputime)
> > int index;
> >
> > /* Add user time to process. */
> > - p->utime += cputime;
> > + p->utime += cputime_to_nsecs(cputime);
> > account_group_user_time(p, cputime);
> <snip>
> > +void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
> > {
> > *ut = p->utime;
> > *st = p->stime;
> > o }
>
> On 32 bit architectures 64bit store/load is not atomic and if not
> protected - 64bit variables can be mangled. I do not see any protection
> (lock) between utime/stime store and load in the patch and seems that
> {u/s}time store/load can be performed at the same time. Though problem
> is very very improbable it still can happen at least theoretically when
> lower and upper 32 bits are changed at the same time i.e. process
> {u,s}time become near to multiple of 2**32 nsec (aprox: 4sec) and
> 64bit {u,s}time is stored and loaded at the same time on different
> cpus. As said this is very improbable situation, but eventually could
> be possible on long lived processes.
"Improbable situation" doesn't appply to Linux. With millions (billion?)
of machines using it, a rare issue in the core turns into likely to happen
somewhere in the planet every second.
So it's definetly a race we want to consider. Note it goes beyond the scope
of this patchset as the issue was already there before since cputime_t can already
map to u64 on 32 bits systems upstream. But this patchset definetly extends
the issue on all 32 bits configs.
kcpustat has the same issue upstream. It's is made of u64 on all configs.
>
> BTW we have already similar problem with sum_exec_runtime. I posted
> some patches to solve the problem, but non of them was good:
> - https://lkml.org/lkml/2016/9/1/172
> this one slow down scheduler hot path's and Peter hates it.
Note u64_stats_sync handles the ifdeffery for this solution.
>
> - https://lkml.org/lkml/2016/9/6/305
> this one was fine for Peter, but I dislike it for taking
> task_rq_lock() and do not push it forward.
Yeah taking rq lock looks too much.
>
> I considering fixing problem of sum_exec_runtime possible mangling
> by using prev_sum_exec_runtime:
>
> u64 read_sum_exec_runtime(struct task_struct *t)
> {
> u64 ns, prev_ns;
>
> do {
> prev_ns = READ_ONCE(t->se.prev_sum_exec_runtime);
> ns = READ_ONCE(t->se.sum_exec_runtime);
> } while (ns < prev_ns || ns > (prev_ns + U32_MAX));
>
> return ns;
> }
>
> This should work based on fact that prev_sum_exec_runtime and
> sum_exec_runtime are not modified and stored at the same time, so only
> one of those variabled can be mangled. Though I need to think about
> correctnes of that a bit more.
I'm not sure that would be enough. READ_ONCE prevents from reordering by the
compiler but not by the CPU. You'd need memory barriers between reads and
writes of prev_ns and ns.
WRITE ns READ prev_ns
smp_wmb() smp_rmb()
WRITE prev_ns READ ns
smp_wmb() smp_rmb()
It seems to be the only way to make sure that at least one of the reads
(prev_ns or ns) is correct.
But then I think that seqlock through u64_stats_sync would be more efficient.
It's indeed very complicated to solve such issue without bothering the fast path.
Irqtime accounting uses a per cpu u64_stats_sync which could be extended to the
whole cputime accounting. I fear that would only solve the issue for kcpustat
though as task cputime would need a per task u64_stats_sync.
Thanks.
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH 08/37] cputime: Convert task/group cputime to nsecs
2017-01-28 15:28 ` Frederic Weisbecker
@ 2017-01-30 13:54 ` Stanislaw Gruszka
2017-01-30 15:38 ` Frederic Weisbecker
0 siblings, 1 reply; 45+ messages in thread
From: Stanislaw Gruszka @ 2017-01-30 13:54 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: Peter Zijlstra, LKML, Tony Luck, Wanpeng Li, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Martin Schwidefsky
On Sat, Jan 28, 2017 at 04:28:13PM +0100, Frederic Weisbecker wrote:
> On Sat, Jan 28, 2017 at 12:57:40PM +0100, Stanislaw Gruszka wrote:
> > On 32 bit architectures 64bit store/load is not atomic and if not
> > protected - 64bit variables can be mangled. I do not see any protection
> > (lock) between utime/stime store and load in the patch and seems that
> > {u/s}time store/load can be performed at the same time. Though problem
> > is very very improbable it still can happen at least theoretically when
> > lower and upper 32 bits are changed at the same time i.e. process
> > {u,s}time become near to multiple of 2**32 nsec (aprox: 4sec) and
> > 64bit {u,s}time is stored and loaded at the same time on different
> > cpus. As said this is very improbable situation, but eventually could
> > be possible on long lived processes.
>
> "Improbable situation" doesn't appply to Linux. With millions (billion?)
> of machines using it, a rare issue in the core turns into likely to happen
> somewhere in the planet every second.
>
> So it's definetly a race we want to consider. Note it goes beyond the scope
> of this patchset as the issue was already there before since cputime_t can already
> map to u64 on 32 bits systems upstream. But this patchset definetly extends
> the issue on all 32 bits configs.
>
> kcpustat has the same issue upstream. It's is made of u64 on all configs.
I would like to add what are possible consequences if value will be
mangled. For sum_exec_runtime, utime and stime we could get wrong values
on cpu-clock related syscalls like clock_gettime() or clock_nanosleep()
and cpu-clock timers like timer_create(CLOCK_PROCESS_CPUTIME_ID) can be
triggered before or long after expected. For kcpustat this seems to be
wrong values read by procfs and 3 drivers (cpufreq, appldata, macintosh).
> > I considering fixing problem of sum_exec_runtime possible mangling
> > by using prev_sum_exec_runtime:
> >
> > u64 read_sum_exec_runtime(struct task_struct *t)
> > {
> > u64 ns, prev_ns;
> >
> > do {
> > prev_ns = READ_ONCE(t->se.prev_sum_exec_runtime);
> > ns = READ_ONCE(t->se.sum_exec_runtime);
> > } while (ns < prev_ns || ns > (prev_ns + U32_MAX));
> >
> > return ns;
> > }
> >
> > This should work based on fact that prev_sum_exec_runtime and
> > sum_exec_runtime are not modified and stored at the same time, so only
> > one of those variabled can be mangled. Though I need to think about
> > correctnes of that a bit more.
>
> I'm not sure that would be enough. READ_ONCE prevents from reordering by the
> compiler but not by the CPU. You'd need memory barriers between reads and
> writes of prev_ns and ns.
It will not be enough, this _suppose_ to work based on that sum_exec_runtime
and prev_sum_exec_runtime are not written at the same time. i.e. only
one variable can be mangled as another one sits already in the memory.
However "not written at the same time" is weak part of reasoning. Even
if those variables are stored at different part of code (sum_exec_runtime
on update_curr() and prev_sum_exec_runtime on set_next_entity()) we can
not assume store of one variable is finished before another one starts.
> WRITE ns READ prev_ns
> smp_wmb() smp_rmb()
> WRITE prev_ns READ ns
> smp_wmb() smp_rmb()
>
> It seems to be the only way to make sure that at least one of the reads
> (prev_ns or ns) is correct.
I think you have right, but seems on much code paths we have scenario:
WRITE ns READ prev_ns
smp_wmb() smp_rmb()
WRITE prev_ns READ ns
and we have already smp_wmb() after write of sum_exec_runtime on
update_min_vruntime().
Stanislaw
^ permalink raw reply [flat|nested] 45+ messages in thread* Re: [PATCH 08/37] cputime: Convert task/group cputime to nsecs
2017-01-30 13:54 ` Stanislaw Gruszka
@ 2017-01-30 15:38 ` Frederic Weisbecker
0 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-30 15:38 UTC (permalink / raw)
To: Stanislaw Gruszka
Cc: Peter Zijlstra, LKML, Tony Luck, Wanpeng Li, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Martin Schwidefsky
On Mon, Jan 30, 2017 at 02:54:51PM +0100, Stanislaw Gruszka wrote:
> On Sat, Jan 28, 2017 at 04:28:13PM +0100, Frederic Weisbecker wrote:
> > On Sat, Jan 28, 2017 at 12:57:40PM +0100, Stanislaw Gruszka wrote:
> > > On 32 bit architectures 64bit store/load is not atomic and if not
> > > protected - 64bit variables can be mangled. I do not see any protection
> > > (lock) between utime/stime store and load in the patch and seems that
> > > {u/s}time store/load can be performed at the same time. Though problem
> > > is very very improbable it still can happen at least theoretically when
> > > lower and upper 32 bits are changed at the same time i.e. process
> > > {u,s}time become near to multiple of 2**32 nsec (aprox: 4sec) and
> > > 64bit {u,s}time is stored and loaded at the same time on different
> > > cpus. As said this is very improbable situation, but eventually could
> > > be possible on long lived processes.
> >
> > "Improbable situation" doesn't appply to Linux. With millions (billion?)
> > of machines using it, a rare issue in the core turns into likely to happen
> > somewhere in the planet every second.
> >
> > So it's definetly a race we want to consider. Note it goes beyond the scope
> > of this patchset as the issue was already there before since cputime_t can already
> > map to u64 on 32 bits systems upstream. But this patchset definetly extends
> > the issue on all 32 bits configs.
> >
> > kcpustat has the same issue upstream. It's is made of u64 on all configs.
>
> I would like to add what are possible consequences if value will be
> mangled. For sum_exec_runtime, utime and stime we could get wrong values
> on cpu-clock related syscalls like clock_gettime() or clock_nanosleep()
> and cpu-clock timers like timer_create(CLOCK_PROCESS_CPUTIME_ID) can be
> triggered before or long after expected. For kcpustat this seems to be
> wrong values read by procfs and 3 drivers (cpufreq, appldata, macintosh).
Yep, all agreed.
>
> > > I considering fixing problem of sum_exec_runtime possible mangling
> > > by using prev_sum_exec_runtime:
> > >
> > > u64 read_sum_exec_runtime(struct task_struct *t)
> > > {
> > > u64 ns, prev_ns;
> > >
> > > do {
> > > prev_ns = READ_ONCE(t->se.prev_sum_exec_runtime);
> > > ns = READ_ONCE(t->se.sum_exec_runtime);
> > > } while (ns < prev_ns || ns > (prev_ns + U32_MAX));
> > >
> > > return ns;
> > > }
> > >
> > > This should work based on fact that prev_sum_exec_runtime and
> > > sum_exec_runtime are not modified and stored at the same time, so only
> > > one of those variabled can be mangled. Though I need to think about
> > > correctnes of that a bit more.
> >
> > I'm not sure that would be enough. READ_ONCE prevents from reordering by the
> > compiler but not by the CPU. You'd need memory barriers between reads and
> > writes of prev_ns and ns.
>
> It will not be enough, this _suppose_ to work based on that sum_exec_runtime
> and prev_sum_exec_runtime are not written at the same time. i.e. only
> one variable can be mangled as another one sits already in the memory.
> However "not written at the same time" is weak part of reasoning. Even
> if those variables are stored at different part of code (sum_exec_runtime
> on update_curr() and prev_sum_exec_runtime on set_next_entity()) we can
> not assume store of one variable is finished before another one starts.
Right.
>
> > WRITE ns READ prev_ns
> > smp_wmb() smp_rmb()
> > WRITE prev_ns READ ns
> > smp_wmb() smp_rmb()
> >
> > It seems to be the only way to make sure that at least one of the reads
> > (prev_ns or ns) is correct.
Well reading that again, I'm not 100% sure about the correctness guarantee.
But it might work.
>
> I think you have right, but seems on much code paths we have scenario:
>
> WRITE ns READ prev_ns
> smp_wmb() smp_rmb()
> WRITE prev_ns READ ns
>
> and we have already smp_wmb() after write of sum_exec_runtime on
> update_min_vruntime().
You still need a second barrier after the second write and read (or
before the first write and read, it's the same) to ensure that if you read
a mangled version of ns, prev_ns is ok.
Still I think u64_stats_sync is less trouble and more reliable.
Thanks.
^ permalink raw reply [flat|nested] 45+ messages in thread
* [PATCH 09/37] alpha: Convert obsolete cputime_t to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (7 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 08/37] cputime: Convert task/group cputime to nsecs Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 10/37] x86: Convert obsolete cputime type " Frederic Weisbecker
` (27 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/alpha/kernel/osf_sys.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/alpha/kernel/osf_sys.c b/arch/alpha/kernel/osf_sys.c
index 82ccb43..9d27a7d 100644
--- a/arch/alpha/kernel/osf_sys.c
+++ b/arch/alpha/kernel/osf_sys.c
@@ -1145,7 +1145,7 @@ struct rusage32 {
SYSCALL_DEFINE2(osf_getrusage, int, who, struct rusage32 __user *, ru)
{
struct rusage32 r;
- cputime_t utime, stime;
+ u64 utime, stime;
unsigned long utime_jiffies, stime_jiffies;
if (who != RUSAGE_SELF && who != RUSAGE_CHILDREN)
@@ -1154,9 +1154,9 @@ SYSCALL_DEFINE2(osf_getrusage, int, who, struct rusage32 __user *, ru)
memset(&r, 0, sizeof(r));
switch (who) {
case RUSAGE_SELF:
- task_cputime_t(current, &utime, &stime);
- utime_jiffies = cputime_to_jiffies(utime);
- stime_jiffies = cputime_to_jiffies(stime);
+ task_cputime(current, &utime, &stime);
+ utime_jiffies = nsecs_to_jiffies(utime);
+ stime_jiffies = nsecs_to_jiffies(stime);
jiffies_to_timeval32(utime_jiffies, &r.ru_utime);
jiffies_to_timeval32(stime_jiffies, &r.ru_stime);
r.ru_minflt = current->min_flt;
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 10/37] x86: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (8 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 09/37] alpha: Convert obsolete cputime_t " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 11/37] isdn: " Frederic Weisbecker
` (26 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/x86/kernel/apm_32.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff --git a/arch/x86/kernel/apm_32.c b/arch/x86/kernel/apm_32.c
index 89c84fc..4a7080c 100644
--- a/arch/x86/kernel/apm_32.c
+++ b/arch/x86/kernel/apm_32.c
@@ -905,21 +905,21 @@ static int apm_cpu_idle(struct cpuidle_device *dev,
{
static int use_apm_idle; /* = 0 */
static unsigned int last_jiffies; /* = 0 */
- static unsigned int last_stime; /* = 0 */
- cputime_t stime, utime;
+ static u64 last_stime; /* = 0 */
+ u64 stime, utime;
int apm_idle_done = 0;
unsigned int jiffies_since_last_check = jiffies - last_jiffies;
unsigned int bucket;
recalc:
- task_cputime_t(current, &utime, &stime);
+ task_cputime(current, &utime, &stime);
if (jiffies_since_last_check > IDLE_CALC_LIMIT) {
use_apm_idle = 0;
} else if (jiffies_since_last_check > idle_period) {
unsigned int idle_percentage;
- idle_percentage = cputime_to_jiffies(stime - last_stime);
+ idle_percentage = nsecs_to_jiffies(stime - last_stime);
idle_percentage *= 100;
idle_percentage /= jiffies_since_last_check;
use_apm_idle = (idle_percentage > idle_threshold);
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 11/37] isdn: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (9 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 10/37] x86: Convert obsolete cputime type " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 12/37] binfmt: " Frederic Weisbecker
` (25 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Not sure if MISDN stats are ABI but it displays task cputime in cputime_t
raw value regardless of what type cputime_t wraps which could be either
jiffies, nsecs, usecs, or whatever random time unit. Plus it wrongly
assumes that cputime_t is long.
Given that this dump is broken anyway, lets just display the nanosec
value and stick with that.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
drivers/isdn/mISDN/stack.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/isdn/mISDN/stack.c b/drivers/isdn/mISDN/stack.c
index 0a36617..b324474 100644
--- a/drivers/isdn/mISDN/stack.c
+++ b/drivers/isdn/mISDN/stack.c
@@ -203,7 +203,7 @@ mISDNStackd(void *data)
{
struct mISDNstack *st = data;
#ifdef MISDN_MSG_STATS
- cputime_t utime, stime;
+ u64 utime, stime;
#endif
int err = 0;
@@ -306,9 +306,9 @@ mISDNStackd(void *data)
"msg %d sleep %d stopped\n",
dev_name(&st->dev->dev), st->msg_cnt, st->sleep_cnt,
st->stopped_cnt);
- task_cputime_t(st->thread, &utime, &stime);
+ task_cputime(st->thread, &utime, &stime);
printk(KERN_DEBUG
- "mISDNStackd daemon for %s utime(%ld) stime(%ld)\n",
+ "mISDNStackd daemon for %s utime(%llu) stime(%llu)\n",
dev_name(&st->dev->dev), utime, stime);
printk(KERN_DEBUG
"mISDNStackd daemon for %s nvcsw(%ld) nivcsw(%ld)\n",
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 12/37] binfmt: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (10 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 11/37] isdn: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 13/37] acct: " Frederic Weisbecker
` (24 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/mips/kernel/binfmt_elfn32.c | 12 ++----------
arch/mips/kernel/binfmt_elfo32.c | 12 ++----------
arch/parisc/kernel/binfmt_elf32.c | 11 ++---------
fs/binfmt_elf.c | 26 ++++++++++----------------
fs/binfmt_elf_fdpic.c | 16 ++++++++--------
fs/compat_binfmt_elf.c | 18 ++----------------
include/linux/compat.h | 20 +++++++++++++++++++-
7 files changed, 45 insertions(+), 70 deletions(-)
diff --git a/arch/mips/kernel/binfmt_elfn32.c b/arch/mips/kernel/binfmt_elfn32.c
index 9c7f3e1..4a2ff39 100644
--- a/arch/mips/kernel/binfmt_elfn32.c
+++ b/arch/mips/kernel/binfmt_elfn32.c
@@ -99,15 +99,7 @@ jiffies_to_compat_timeval(unsigned long jiffies, struct compat_timeval *value)
#undef TASK_SIZE
#define TASK_SIZE TASK_SIZE32
-#undef cputime_to_timeval
-#define cputime_to_timeval cputime_to_compat_timeval
-static __inline__ void
-cputime_to_compat_timeval(const cputime_t cputime, struct compat_timeval *value)
-{
- unsigned long jiffies = cputime_to_jiffies(cputime);
-
- value->tv_usec = (jiffies % HZ) * (1000000L / HZ);
- value->tv_sec = jiffies / HZ;
-}
+#undef ns_to_timeval
+#define ns_to_timeval ns_to_compat_timeval
#include "../../../fs/binfmt_elf.c"
diff --git a/arch/mips/kernel/binfmt_elfo32.c b/arch/mips/kernel/binfmt_elfo32.c
index 1ab3432..3916404 100644
--- a/arch/mips/kernel/binfmt_elfo32.c
+++ b/arch/mips/kernel/binfmt_elfo32.c
@@ -102,15 +102,7 @@ jiffies_to_compat_timeval(unsigned long jiffies, struct compat_timeval *value)
#undef TASK_SIZE
#define TASK_SIZE TASK_SIZE32
-#undef cputime_to_timeval
-#define cputime_to_timeval cputime_to_compat_timeval
-static __inline__ void
-cputime_to_compat_timeval(const cputime_t cputime, struct compat_timeval *value)
-{
- unsigned long jiffies = cputime_to_jiffies(cputime);
-
- value->tv_usec = (jiffies % HZ) * (1000000L / HZ);
- value->tv_sec = jiffies / HZ;
-}
+#undef ns_to_timeval
+#define ns_to_timeval ns_to_compat_timeval
#include "../../../fs/binfmt_elf.c"
diff --git a/arch/parisc/kernel/binfmt_elf32.c b/arch/parisc/kernel/binfmt_elf32.c
index 00dc66f..f2adcf3 100644
--- a/arch/parisc/kernel/binfmt_elf32.c
+++ b/arch/parisc/kernel/binfmt_elf32.c
@@ -91,14 +91,7 @@ struct elf_prpsinfo32
current->thread.map_base = DEFAULT_MAP_BASE32; \
current->thread.task_size = DEFAULT_TASK_SIZE32 \
-#undef cputime_to_timeval
-#define cputime_to_timeval cputime_to_compat_timeval
-static __inline__ void
-cputime_to_compat_timeval(const cputime_t cputime, struct compat_timeval *value)
-{
- unsigned long jiffies = cputime_to_jiffies(cputime);
- value->tv_usec = (jiffies % HZ) * (1000000L / HZ);
- value->tv_sec = jiffies / HZ;
-}
+#undef ns_to_timeval
+#define ns_to_timeval ns_to_compat_timeval
#include "../../../fs/binfmt_elf.c"
diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c
index 13cd15d..f28595f 100644
--- a/fs/binfmt_elf.c
+++ b/fs/binfmt_elf.c
@@ -1411,8 +1411,6 @@ static void fill_note(struct memelfnote *note, const char *name, int type,
static void fill_prstatus(struct elf_prstatus *prstatus,
struct task_struct *p, long signr)
{
- struct timeval tv;
-
prstatus->pr_info.si_signo = prstatus->pr_cursig = signr;
prstatus->pr_sigpend = p->pending.signal.sig[0];
prstatus->pr_sighold = p->blocked.sig[0];
@@ -1423,29 +1421,25 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
/*
* This is the record for the group leader. It shows the
* group-wide total, not its individual thread total.
*/
- thread_group_cputime_t(p, &cputime);
- cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
- cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
+ thread_group_cputime(p, &cputime);
+ prstatus->pr_utime = ns_to_timeval(cputime.utime);
+ prstatus->pr_stime = ns_to_timeval(cputime.stime);
} else {
- cputime_t utime, stime;
+ u64 utime, stime;
- task_cputime_t(p, &utime, &stime);
- cputime_to_timeval(utime, &prstatus->pr_utime);
- cputime_to_timeval(stime, &prstatus->pr_stime);
+ task_cputime(p, &utime, &stime);
+ prstatus->pr_utime = ns_to_timeval(utime);
+ prstatus->pr_stime = ns_to_timeval(stime);
}
- tv = ns_to_timeval(p->signal->cutime);
- prstatus->pr_cutime.tv_sec = tv.tv_sec;
- prstatus->pr_cutime.tv_usec = tv.tv_usec;
- tv = ns_to_timeval(p->signal->cstime);
- prstatus->pr_cstime.tv_sec = tv.tv_sec;
- prstatus->pr_cstime.tv_usec = tv.tv_usec;
+ prstatus->pr_cutime = ns_to_timeval(p->signal->cutime);
+ prstatus->pr_cstime = ns_to_timeval(p->signal->cstime);
}
static int fill_psinfo(struct elf_prpsinfo *psinfo, struct task_struct *p,
diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c
index e1f3734..ffca4bb 100644
--- a/fs/binfmt_elf_fdpic.c
+++ b/fs/binfmt_elf_fdpic.c
@@ -1342,21 +1342,21 @@ static void fill_prstatus(struct elf_prstatus *prstatus,
prstatus->pr_pgrp = task_pgrp_vnr(p);
prstatus->pr_sid = task_session_vnr(p);
if (thread_group_leader(p)) {
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
/*
* This is the record for the group leader. It shows the
* group-wide total, not its individual thread total.
*/
- thread_group_cputime_t(p, &cputime);
- cputime_to_timeval(cputime.utime, &prstatus->pr_utime);
- cputime_to_timeval(cputime.stime, &prstatus->pr_stime);
+ thread_group_cputime(p, &cputime);
+ prstatus->pr_utime = ns_to_timeval(cputime.utime);
+ prstatus->pr_stime = ns_to_timeval(cputime.stime);
} else {
- cputime_t utime, stime;
+ u64 utime, stime;
- task_cputime_t(p, &utime, &stime);
- cputime_to_timeval(utime, &prstatus->pr_utime);
- cputime_to_timeval(stime, &prstatus->pr_stime);
+ task_cputime(p, &utime, &stime);
+ prstatus->pr_utime = ns_to_timeval(utime);
+ prstatus->pr_stime = ns_to_timeval(stime);
}
prstatus->pr_cutime = ns_to_timeval(p->signal->cutime);
prstatus->pr_cstime = ns_to_timeval(p->signal->cstime);
diff --git a/fs/compat_binfmt_elf.c b/fs/compat_binfmt_elf.c
index 4d24d17..504b3c3 100644
--- a/fs/compat_binfmt_elf.c
+++ b/fs/compat_binfmt_elf.c
@@ -51,22 +51,8 @@
#define elf_prstatus compat_elf_prstatus
#define elf_prpsinfo compat_elf_prpsinfo
-/*
- * Compat version of cputime_to_compat_timeval, perhaps this
- * should be an inline in <linux/compat.h>.
- */
-static void cputime_to_compat_timeval(const cputime_t cputime,
- struct compat_timeval *value)
-{
- struct timeval tv;
- cputime_to_timeval(cputime, &tv);
- value->tv_sec = tv.tv_sec;
- value->tv_usec = tv.tv_usec;
-}
-
-#undef cputime_to_timeval
-#define cputime_to_timeval cputime_to_compat_timeval
-
+#undef ns_to_timeval
+#define ns_to_timeval ns_to_compat_timeval
/*
* To use this file, asm/elf.h must define compat_elf_check_arch.
diff --git a/include/linux/compat.h b/include/linux/compat.h
index 6360939..9e40be5 100644
--- a/include/linux/compat.h
+++ b/include/linux/compat.h
@@ -731,7 +731,25 @@ asmlinkage long compat_sys_fanotify_mark(int, unsigned int, __u32, __u32,
static inline bool in_compat_syscall(void) { return is_compat_task(); }
#endif
-#else
+/**
+ * ns_to_compat_timeval - Compat version of ns_to_timeval
+ * @nsec: the nanoseconds value to be converted
+ *
+ * Returns the compat_timeval representation of the nsec parameter.
+ */
+static inline struct compat_timeval ns_to_compat_timeval(s64 nsec)
+{
+ struct timeval tv;
+ struct compat_timeval ctv;
+
+ tv = ns_to_timeval(nsec);
+ ctv.tv_sec = tv.tv_sec;
+ ctv.tv_usec = tv.tv_usec;
+
+ return ctv;
+}
+
+#else /* !CONFIG_COMPAT */
#define is_compat_task() (0)
static inline bool in_compat_syscall(void) { return false; }
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 13/37] acct: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (11 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 12/37] binfmt: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 14/37] delaycct: " Frederic Weisbecker
` (23 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/sched.h | 2 +-
kernel/acct.c | 9 +++++----
2 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 089a4ba..547b6f0 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -563,7 +563,7 @@ struct pacct_struct {
int ac_flag;
long ac_exitcode;
unsigned long ac_mem;
- cputime_t ac_utime, ac_stime;
+ u64 ac_utime, ac_stime;
unsigned long ac_minflt, ac_majflt;
};
diff --git a/kernel/acct.c b/kernel/acct.c
index b9b190a8..ca9cb55 100644
--- a/kernel/acct.c
+++ b/kernel/acct.c
@@ -453,8 +453,8 @@ static void fill_ac(acct_t *ac)
spin_lock_irq(¤t->sighand->siglock);
tty = current->signal->tty; /* Safe as we hold the siglock */
ac->ac_tty = tty ? old_encode_dev(tty_devnum(tty)) : 0;
- ac->ac_utime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_utime)));
- ac->ac_stime = encode_comp_t(jiffies_to_AHZ(cputime_to_jiffies(pacct->ac_stime)));
+ ac->ac_utime = encode_comp_t(nsec_to_AHZ(pacct->ac_utime));
+ ac->ac_stime = encode_comp_t(nsec_to_AHZ(pacct->ac_stime));
ac->ac_flag = pacct->ac_flag;
ac->ac_mem = encode_comp_t(pacct->ac_mem);
ac->ac_minflt = encode_comp_t(pacct->ac_minflt);
@@ -530,7 +530,7 @@ static void do_acct_process(struct bsd_acct_struct *acct)
void acct_collect(long exitcode, int group_dead)
{
struct pacct_struct *pacct = ¤t->signal->pacct;
- cputime_t utime, stime;
+ u64 utime, stime;
unsigned long vsize = 0;
if (group_dead && current->mm) {
@@ -559,7 +559,8 @@ void acct_collect(long exitcode, int group_dead)
pacct->ac_flag |= ACORE;
if (current->flags & PF_SIGNALED)
pacct->ac_flag |= AXSIG;
- task_cputime_t(current, &utime, &stime);
+
+ task_cputime(current, &utime, &stime);
pacct->ac_utime += utime;
pacct->ac_stime += stime;
pacct->ac_minflt += current->min_flt;
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 14/37] delaycct: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (12 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 13/37] acct: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 15/37] tsacct: " Frederic Weisbecker
` (22 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/delayacct.c | 10 +++++-----
1 file changed, 5 insertions(+), 5 deletions(-)
diff --git a/kernel/delayacct.c b/kernel/delayacct.c
index 228640f..6605496 100644
--- a/kernel/delayacct.c
+++ b/kernel/delayacct.c
@@ -82,19 +82,19 @@ void __delayacct_blkio_end(void)
int __delayacct_add_tsk(struct taskstats *d, struct task_struct *tsk)
{
- cputime_t utime, stime, stimescaled, utimescaled;
+ u64 utime, stime, stimescaled, utimescaled;
unsigned long long t2, t3;
unsigned long flags, t1;
s64 tmp;
- task_cputime_t(tsk, &utime, &stime);
+ task_cputime(tsk, &utime, &stime);
tmp = (s64)d->cpu_run_real_total;
- tmp += cputime_to_nsecs(utime + stime);
+ tmp += utime + stime;
d->cpu_run_real_total = (tmp < (s64)d->cpu_run_real_total) ? 0 : tmp;
- task_cputime_t_scaled(tsk, &utimescaled, &stimescaled);
+ task_cputime_scaled(tsk, &utimescaled, &stimescaled);
tmp = (s64)d->cpu_scaled_run_real_total;
- tmp += cputime_to_nsecs(utimescaled + stimescaled);
+ tmp += utimescaled + stimescaled;
d->cpu_scaled_run_real_total =
(tmp < (s64)d->cpu_scaled_run_real_total) ? 0 : tmp;
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 15/37] tsacct: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (13 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 14/37] delaycct: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 16/37] signal: " Frederic Weisbecker
` (21 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/sched.h | 2 +-
kernel/tsacct.c | 27 ++++++++++++---------------
2 files changed, 13 insertions(+), 16 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 547b6f0..7fcbffa 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -1815,7 +1815,7 @@ struct task_struct {
#if defined(CONFIG_TASK_XACCT)
u64 acct_rss_mem1; /* accumulated rss usage */
u64 acct_vm_mem1; /* accumulated virtual memory usage */
- cputime_t acct_timexpd; /* stime + utime since last update */
+ u64 acct_timexpd; /* stime + utime since last update */
#endif
#ifdef CONFIG_CPUSETS
nodemask_t mems_allowed; /* Protected by alloc_lock */
diff --git a/kernel/tsacct.c b/kernel/tsacct.c
index 040d0a6..5c21f05 100644
--- a/kernel/tsacct.c
+++ b/kernel/tsacct.c
@@ -31,7 +31,7 @@ void bacct_add_tsk(struct user_namespace *user_ns,
struct taskstats *stats, struct task_struct *tsk)
{
const struct cred *tcred;
- cputime_t utime, stime, utimescaled, stimescaled;
+ u64 utime, stime, utimescaled, stimescaled;
u64 delta;
BUILD_BUG_ON(TS_COMM_LEN < TASK_COMM_LEN);
@@ -66,13 +66,13 @@ void bacct_add_tsk(struct user_namespace *user_ns,
task_tgid_nr_ns(rcu_dereference(tsk->real_parent), pid_ns) : 0;
rcu_read_unlock();
- task_cputime_t(tsk, &utime, &stime);
- stats->ac_utime = cputime_to_usecs(utime);
- stats->ac_stime = cputime_to_usecs(stime);
+ task_cputime(tsk, &utime, &stime);
+ stats->ac_utime = div_u64(utime, NSEC_PER_USEC);
+ stats->ac_stime = div_u64(stime, NSEC_PER_USEC);
- task_cputime_t_scaled(tsk, &utimescaled, &stimescaled);
- stats->ac_utimescaled = cputime_to_usecs(utimescaled);
- stats->ac_stimescaled = cputime_to_usecs(stimescaled);
+ task_cputime_scaled(tsk, &utimescaled, &stimescaled);
+ stats->ac_utimescaled = div_u64(utimescaled, NSEC_PER_USEC);
+ stats->ac_stimescaled = div_u64(stimescaled, NSEC_PER_USEC);
stats->ac_minflt = tsk->min_flt;
stats->ac_majflt = tsk->maj_flt;
@@ -123,18 +123,15 @@ void xacct_add_tsk(struct taskstats *stats, struct task_struct *p)
#undef MB
static void __acct_update_integrals(struct task_struct *tsk,
- cputime_t utime, cputime_t stime)
+ u64 utime, u64 stime)
{
- cputime_t time, dtime;
- u64 delta;
+ u64 time, delta;
if (!likely(tsk->mm))
return;
time = stime + utime;
- dtime = time - tsk->acct_timexpd;
- /* Avoid division: cputime_t is often in nanoseconds already. */
- delta = cputime_to_nsecs(dtime);
+ delta = time - tsk->acct_timexpd;
if (delta < TICK_NSEC)
return;
@@ -155,11 +152,11 @@ static void __acct_update_integrals(struct task_struct *tsk,
*/
void acct_update_integrals(struct task_struct *tsk)
{
- cputime_t utime, stime;
+ u64 utime, stime;
unsigned long flags;
local_irq_save(flags);
- task_cputime_t(tsk, &utime, &stime);
+ task_cputime(tsk, &utime, &stime);
__acct_update_integrals(tsk, utime, stime);
local_irq_restore(flags);
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 16/37] signal: Convert obsolete cputime type to nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (14 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 15/37] tsacct: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 17/37] cputime: Increment kcpustat directly on irqtime account Frederic Weisbecker
` (20 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/signal.c | 16 ++++++++--------
1 file changed, 8 insertions(+), 8 deletions(-)
diff --git a/kernel/signal.c b/kernel/signal.c
index b635221..13f9def 100644
--- a/kernel/signal.c
+++ b/kernel/signal.c
@@ -1581,7 +1581,7 @@ bool do_notify_parent(struct task_struct *tsk, int sig)
unsigned long flags;
struct sighand_struct *psig;
bool autoreap = false;
- cputime_t utime, stime;
+ u64 utime, stime;
BUG_ON(sig == -1);
@@ -1619,9 +1619,9 @@ bool do_notify_parent(struct task_struct *tsk, int sig)
task_uid(tsk));
rcu_read_unlock();
- task_cputime_t(tsk, &utime, &stime);
- info.si_utime = cputime_to_clock_t(utime + nsecs_to_cputime(tsk->signal->utime));
- info.si_stime = cputime_to_clock_t(stime + nsecs_to_cputime(tsk->signal->stime));
+ task_cputime(tsk, &utime, &stime);
+ info.si_utime = nsec_to_clock_t(utime + tsk->signal->utime);
+ info.si_stime = nsec_to_clock_t(stime + tsk->signal->stime);
info.si_status = tsk->exit_code & 0x7f;
if (tsk->exit_code & 0x80)
@@ -1685,7 +1685,7 @@ static void do_notify_parent_cldstop(struct task_struct *tsk,
unsigned long flags;
struct task_struct *parent;
struct sighand_struct *sighand;
- cputime_t utime, stime;
+ u64 utime, stime;
if (for_ptracer) {
parent = tsk->parent;
@@ -1704,9 +1704,9 @@ static void do_notify_parent_cldstop(struct task_struct *tsk,
info.si_uid = from_kuid_munged(task_cred_xxx(parent, user_ns), task_uid(tsk));
rcu_read_unlock();
- task_cputime_t(tsk, &utime, &stime);
- info.si_utime = cputime_to_clock_t(utime);
- info.si_stime = cputime_to_clock_t(stime);
+ task_cputime(tsk, &utime, &stime);
+ info.si_utime = nsec_to_clock_t(utime);
+ info.si_stime = nsec_to_clock_t(stime);
info.si_code = why;
switch (why) {
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 17/37] cputime: Increment kcpustat directly on irqtime account
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (15 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 16/37] signal: " Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 18/37] posix-timers: Use TICK_NSEC instead of a dynamically ad-hoc calculated version Frederic Weisbecker
` (19 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
The irqtime is accounted is nsecs and stored in
cpu_irq_time.hardirq_time and cpu_irq_time.softirq_time. Once the
accumulated amount reaches a new jiffy, this one gets accounted to the
kcpustat.
This was necessary when kcpustat was stored in cputime_t, which could at
worst have a jiffies granularity. But now kcpustat is stored in nsecs
so this whole discretization game with temporary irqtime storage has
become unnecessary.
We can now directly account the irqtime to the kcpustat.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/sched/cputime.c | 50 +++++++++++++++++---------------------------------
kernel/sched/sched.h | 7 ++++---
2 files changed, 21 insertions(+), 36 deletions(-)
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 0bdef50..bee6c97 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -44,6 +44,7 @@ void disable_sched_clock_irqtime(void)
void irqtime_account_irq(struct task_struct *curr)
{
struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
+ u64 *cpustat = kcpustat_this_cpu->cpustat;
s64 delta;
int cpu;
@@ -61,49 +62,35 @@ void irqtime_account_irq(struct task_struct *curr)
* 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())
- irqtime->hardirq_time += delta;
- else if (in_serving_softirq() && curr != this_cpu_ksoftirqd())
- irqtime->softirq_time += delta;
+ if (hardirq_count()) {
+ cpustat[CPUTIME_IRQ] += delta;
+ irqtime->tick_delta += delta;
+ } else if (in_serving_softirq() && curr != this_cpu_ksoftirqd()) {
+ cpustat[CPUTIME_SOFTIRQ] += delta;
+ irqtime->tick_delta += delta;
+ }
u64_stats_update_end(&irqtime->sync);
}
EXPORT_SYMBOL_GPL(irqtime_account_irq);
-static cputime_t irqtime_account_update(u64 irqtime, int idx, cputime_t maxtime)
+static cputime_t irqtime_tick_accounted(cputime_t maxtime)
{
- u64 *cpustat = kcpustat_this_cpu->cpustat;
- cputime_t irq_cputime;
+ struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
+ cputime_t delta;
- irq_cputime = nsecs_to_cputime64(irqtime - cpustat[idx]);
- irq_cputime = min(irq_cputime, maxtime);
- cpustat[idx] += cputime_to_nsecs(irq_cputime);
+ delta = nsecs_to_cputime(irqtime->tick_delta);
+ delta = min(delta, maxtime);
+ irqtime->tick_delta -= cputime_to_nsecs(delta);
- return irq_cputime;
-}
-
-static cputime_t irqtime_account_hi_update(cputime_t maxtime)
-{
- return irqtime_account_update(__this_cpu_read(cpu_irqtime.hardirq_time),
- CPUTIME_IRQ, maxtime);
-}
-
-static cputime_t irqtime_account_si_update(cputime_t maxtime)
-{
- return irqtime_account_update(__this_cpu_read(cpu_irqtime.softirq_time),
- CPUTIME_SOFTIRQ, maxtime);
+ return delta;
}
#else /* CONFIG_IRQ_TIME_ACCOUNTING */
#define sched_clock_irqtime (0)
-static cputime_t irqtime_account_hi_update(cputime_t dummy)
-{
- return 0;
-}
-
-static cputime_t irqtime_account_si_update(cputime_t dummy)
+static cputime_t irqtime_tick_accounted(cputime_t dummy)
{
return 0;
}
@@ -280,10 +267,7 @@ static inline cputime_t account_other_time(cputime_t max)
accounted = steal_account_process_time(max);
if (accounted < max)
- accounted += irqtime_account_hi_update(max - accounted);
-
- if (accounted < max)
- accounted += irqtime_account_si_update(max - accounted);
+ accounted += irqtime_tick_accounted(max - accounted);
return accounted;
}
diff --git a/kernel/sched/sched.h b/kernel/sched/sched.h
index 6eeae7e..8ff5cc5 100644
--- a/kernel/sched/sched.h
+++ b/kernel/sched/sched.h
@@ -4,6 +4,7 @@
#include <linux/sched/rt.h>
#include <linux/u64_stats_sync.h>
#include <linux/sched/deadline.h>
+#include <linux/kernel_stat.h>
#include <linux/binfmts.h>
#include <linux/mutex.h>
#include <linux/spinlock.h>
@@ -1827,8 +1828,7 @@ static inline void nohz_balance_exit_idle(unsigned int cpu) { }
#ifdef CONFIG_IRQ_TIME_ACCOUNTING
struct irqtime {
- u64 hardirq_time;
- u64 softirq_time;
+ u64 tick_delta;
u64 irq_start_time;
struct u64_stats_sync sync;
};
@@ -1838,12 +1838,13 @@ DECLARE_PER_CPU(struct irqtime, cpu_irqtime);
static inline u64 irq_time_read(int cpu)
{
struct irqtime *irqtime = &per_cpu(cpu_irqtime, cpu);
+ u64 *cpustat = kcpustat_cpu(cpu).cpustat;
unsigned int seq;
u64 total;
do {
seq = __u64_stats_fetch_begin(&irqtime->sync);
- total = irqtime->softirq_time + irqtime->hardirq_time;
+ total = cpustat[CPUTIME_SOFTIRQ] + cpustat[CPUTIME_IRQ];
} while (__u64_stats_fetch_retry(&irqtime->sync, seq));
return total;
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 18/37] posix-timers: Use TICK_NSEC instead of a dynamically ad-hoc calculated version
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (16 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 17/37] cputime: Increment kcpustat directly on irqtime account Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 19/37] posix-timers: Convert internals to use nsecs Frederic Weisbecker
` (18 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/time/posix-cpu-timers.c | 11 ++---------
1 file changed, 2 insertions(+), 9 deletions(-)
diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c
index d53ff71..8349e02 100644
--- a/kernel/time/posix-cpu-timers.c
+++ b/kernel/time/posix-cpu-timers.c
@@ -890,8 +890,6 @@ static inline void stop_process_timers(struct signal_struct *sig)
tick_dep_clear_signal(sig, TICK_DEP_BIT_POSIX_TIMER);
}
-static u32 onecputick;
-
static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
unsigned long long *expires,
unsigned long long cur_time, int signo)
@@ -903,9 +901,9 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
if (it->incr) {
it->expires += it->incr;
it->error += it->incr_error;
- if (it->error >= onecputick) {
+ if (it->error >= TICK_NSEC) {
it->expires -= cputime_one_jiffy;
- it->error -= onecputick;
+ it->error -= TICK_NSEC;
}
} else {
it->expires = 0;
@@ -1476,15 +1474,10 @@ static __init int init_posix_cpu_timers(void)
.clock_get = thread_cpu_clock_get,
.timer_create = thread_cpu_timer_create,
};
- struct timespec ts;
posix_timers_register_clock(CLOCK_PROCESS_CPUTIME_ID, &process);
posix_timers_register_clock(CLOCK_THREAD_CPUTIME_ID, &thread);
- cputime_to_timespec(cputime_one_jiffy, &ts);
- onecputick = ts.tv_nsec;
- WARN_ON(ts.tv_sec != 0);
-
return 0;
}
__initcall(init_posix_cpu_timers);
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 19/37] posix-timers: Convert internals to use nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (17 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 18/37] posix-timers: Use TICK_NSEC instead of a dynamically ad-hoc calculated version Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 20/37] itimer: Convert internal cputime_t units to nsec Frederic Weisbecker
` (17 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Also convert posix-cpu-timers to use nsec based internal counters to
simplify it.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/posix-timers.h | 12 +--
include/linux/sched.h | 6 +-
kernel/fork.c | 2 +-
kernel/sched/cputime.c | 6 +-
kernel/sched/stats.h | 4 +-
kernel/time/itimer.c | 6 +-
kernel/time/posix-cpu-timers.c | 210 +++++++++++++++++------------------------
7 files changed, 100 insertions(+), 146 deletions(-)
diff --git a/include/linux/posix-timers.h b/include/linux/posix-timers.h
index 62d44c1..890de52 100644
--- a/include/linux/posix-timers.h
+++ b/include/linux/posix-timers.h
@@ -8,19 +8,9 @@
#include <linux/alarmtimer.h>
-static inline unsigned long long cputime_to_expires(cputime_t expires)
-{
- return (__force unsigned long long)expires;
-}
-
-static inline cputime_t expires_to_cputime(unsigned long long expires)
-{
- return (__force cputime_t)expires;
-}
-
struct cpu_timer_list {
struct list_head entry;
- unsigned long long expires, incr;
+ u64 expires, incr;
struct task_struct *task;
int firing;
};
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 7fcbffa..2768f94 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -755,7 +755,7 @@ struct signal_struct {
struct thread_group_cputimer cputimer;
/* Earliest-expiration cache. */
- struct task_cputime_t cputime_expires;
+ struct task_cputime cputime_expires;
#ifdef CONFIG_NO_HZ_FULL
atomic_t tick_dep_mask;
@@ -1689,7 +1689,7 @@ struct task_struct {
/* mm fault and swap info: this can arguably be seen as either mm-specific or thread-specific */
unsigned long min_flt, maj_flt;
- struct task_cputime_t cputime_expires;
+ struct task_cputime cputime_expires;
struct list_head cpu_timers[3];
/* process credentials */
@@ -3528,7 +3528,7 @@ static __always_inline bool need_resched(void)
* Thread group CPU time accounting.
*/
void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times);
-void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times);
+void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
static inline void thread_group_cputime_t(struct task_struct *tsk,
struct task_cputime_t *cputime)
diff --git a/kernel/fork.c b/kernel/fork.c
index 11c5c8a..09992ff 100644
--- a/kernel/fork.c
+++ b/kernel/fork.c
@@ -1313,7 +1313,7 @@ static void posix_cpu_timers_init_group(struct signal_struct *sig)
cpu_limit = READ_ONCE(sig->rlim[RLIMIT_CPU].rlim_cur);
if (cpu_limit != RLIM_INFINITY) {
- sig->cputime_expires.prof_exp = secs_to_cputime(cpu_limit);
+ sig->cputime_expires.prof_exp = cpu_limit * NSEC_PER_SEC;
sig->cputimer.running = true;
}
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index bee6c97..f7b9624 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -122,7 +122,7 @@ void account_user_time(struct task_struct *p, cputime_t cputime)
/* Add user time to process. */
p->utime += cputime_to_nsecs(cputime);
- account_group_user_time(p, cputime);
+ account_group_user_time(p, cputime_to_nsecs(cputime));
index = (task_nice(p) > 0) ? CPUTIME_NICE : CPUTIME_USER;
@@ -144,7 +144,7 @@ void account_guest_time(struct task_struct *p, cputime_t cputime)
/* Add guest time to process. */
p->utime += cputime_to_nsecs(cputime);
- account_group_user_time(p, cputime);
+ account_group_user_time(p, cputime_to_nsecs(cputime));
p->gtime += cputime_to_nsecs(cputime);
/* Add guest time to cpustat. */
@@ -168,7 +168,7 @@ void account_system_index_time(struct task_struct *p,
{
/* Add system time to process. */
p->stime += cputime_to_nsecs(cputime);
- account_group_system_time(p, cputime);
+ account_group_system_time(p, cputime_to_nsecs(cputime));
/* Add system time to cpustat. */
task_group_account_field(p, index, cputime_to_nsecs(cputime));
diff --git a/kernel/sched/stats.h b/kernel/sched/stats.h
index 34659a8..9788478 100644
--- a/kernel/sched/stats.h
+++ b/kernel/sched/stats.h
@@ -216,7 +216,7 @@ static inline bool cputimer_running(struct task_struct *tsk)
* running CPU and update the utime field there.
*/
static inline void account_group_user_time(struct task_struct *tsk,
- cputime_t cputime)
+ u64 cputime)
{
struct thread_group_cputimer *cputimer = &tsk->signal->cputimer;
@@ -237,7 +237,7 @@ static inline void account_group_user_time(struct task_struct *tsk,
* running CPU and update the stime field there.
*/
static inline void account_group_system_time(struct task_struct *tsk,
- cputime_t cputime)
+ u64 cputime)
{
struct thread_group_cputimer *cputimer = &tsk->signal->cputimer;
diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c
index f2d5097..bb01ff4 100644
--- a/kernel/time/itimer.c
+++ b/kernel/time/itimer.c
@@ -53,15 +53,15 @@ static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id,
cval = it->expires;
cinterval = it->incr;
if (cval) {
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
cputime_t t;
thread_group_cputimer(tsk, &cputime);
if (clock_id == CPUCLOCK_PROF)
- t = cputime.utime + cputime.stime;
+ t = nsecs_to_cputime(cputime.utime + cputime.stime);
else
/* CPUCLOCK_VIRT */
- t = cputime.utime;
+ t = nsecs_to_cputime(cputime.utime);
if (cval < t)
/* about to fire */
diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c
index 8349e02..45be3ce 100644
--- a/kernel/time/posix-cpu-timers.c
+++ b/kernel/time/posix-cpu-timers.c
@@ -50,39 +50,14 @@ static int check_clock(const clockid_t which_clock)
return error;
}
-static inline unsigned long long
-timespec_to_sample(const clockid_t which_clock, const struct timespec *tp)
-{
- unsigned long long ret;
-
- ret = 0; /* high half always zero when .cpu used */
- if (CPUCLOCK_WHICH(which_clock) == CPUCLOCK_SCHED) {
- ret = (unsigned long long)tp->tv_sec * NSEC_PER_SEC + tp->tv_nsec;
- } else {
- ret = cputime_to_expires(timespec_to_cputime(tp));
- }
- return ret;
-}
-
-static void sample_to_timespec(const clockid_t which_clock,
- unsigned long long expires,
- struct timespec *tp)
-{
- if (CPUCLOCK_WHICH(which_clock) == CPUCLOCK_SCHED)
- *tp = ns_to_timespec(expires);
- else
- cputime_to_timespec((__force cputime_t)expires, tp);
-}
-
/*
* Update expiry time from increment, and increase overrun count,
* given the current clock sample.
*/
-static void bump_cpu_timer(struct k_itimer *timer,
- unsigned long long now)
+static void bump_cpu_timer(struct k_itimer *timer, u64 now)
{
int i;
- unsigned long long delta, incr;
+ u64 delta, incr;
if (timer->it.cpu.incr == 0)
return;
@@ -115,28 +90,28 @@ static void bump_cpu_timer(struct k_itimer *timer,
* Checks @cputime to see if all fields are zero. Returns true if all fields
* are zero, false if any field is nonzero.
*/
-static inline int task_cputime_zero(const struct task_cputime_t *cputime)
+static inline int task_cputime_zero(const struct task_cputime *cputime)
{
if (!cputime->utime && !cputime->stime && !cputime->sum_exec_runtime)
return 1;
return 0;
}
-static inline unsigned long long prof_ticks(struct task_struct *p)
+static inline u64 prof_ticks(struct task_struct *p)
{
- cputime_t utime, stime;
+ u64 utime, stime;
- task_cputime_t(p, &utime, &stime);
+ task_cputime(p, &utime, &stime);
- return cputime_to_expires(utime + stime);
+ return utime + stime;
}
-static inline unsigned long long virt_ticks(struct task_struct *p)
+static inline u64 virt_ticks(struct task_struct *p)
{
- cputime_t utime, stime;
+ u64 utime, stime;
- task_cputime_t(p, &utime, &stime);
+ task_cputime(p, &utime, &stime);
- return cputime_to_expires(utime);
+ return utime;
}
static int
@@ -176,8 +151,8 @@ posix_cpu_clock_set(const clockid_t which_clock, const struct timespec *tp)
/*
* Sample a per-thread clock for the given task.
*/
-static int cpu_clock_sample(const clockid_t which_clock, struct task_struct *p,
- unsigned long long *sample)
+static int cpu_clock_sample(const clockid_t which_clock,
+ struct task_struct *p, u64 *sample)
{
switch (CPUCLOCK_WHICH(which_clock)) {
default:
@@ -210,7 +185,7 @@ static inline void __update_gt_cputime(atomic64_t *cputime, u64 sum_cputime)
}
}
-static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct task_cputime_t *sum)
+static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct task_cputime *sum)
{
__update_gt_cputime(&cputime_atomic->utime, sum->utime);
__update_gt_cputime(&cputime_atomic->stime, sum->stime);
@@ -218,7 +193,7 @@ static void update_gt_cputime(struct task_cputime_atomic *cputime_atomic, struct
}
/* Sample task_cputime_atomic values in "atomic_timers", store results in "times". */
-static inline void sample_cputime_atomic(struct task_cputime_t *times,
+static inline void sample_cputime_atomic(struct task_cputime *times,
struct task_cputime_atomic *atomic_times)
{
times->utime = atomic64_read(&atomic_times->utime);
@@ -226,10 +201,10 @@ static inline void sample_cputime_atomic(struct task_cputime_t *times,
times->sum_exec_runtime = atomic64_read(&atomic_times->sum_exec_runtime);
}
-void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times)
+void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times)
{
struct thread_group_cputimer *cputimer = &tsk->signal->cputimer;
- struct task_cputime_t sum;
+ struct task_cputime sum;
/* Check if cputimer isn't running. This is accessed without locking. */
if (!READ_ONCE(cputimer->running)) {
@@ -238,7 +213,7 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times
* values through the TIMER_ABSTIME flag, therefore we have
* to synchronize the timer to the clock every time we start it.
*/
- thread_group_cputime_t(tsk, &sum);
+ thread_group_cputime(tsk, &sum);
update_gt_cputime(&cputimer->cputime_atomic, &sum);
/*
@@ -260,23 +235,23 @@ void thread_group_cputimer(struct task_struct *tsk, struct task_cputime_t *times
*/
static int cpu_clock_sample_group(const clockid_t which_clock,
struct task_struct *p,
- unsigned long long *sample)
+ u64 *sample)
{
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
switch (CPUCLOCK_WHICH(which_clock)) {
default:
return -EINVAL;
case CPUCLOCK_PROF:
- thread_group_cputime_t(p, &cputime);
- *sample = cputime_to_expires(cputime.utime + cputime.stime);
+ thread_group_cputime(p, &cputime);
+ *sample = cputime.utime + cputime.stime;
break;
case CPUCLOCK_VIRT:
- thread_group_cputime_t(p, &cputime);
- *sample = cputime_to_expires(cputime.utime);
+ thread_group_cputime(p, &cputime);
+ *sample = cputime.utime;
break;
case CPUCLOCK_SCHED:
- thread_group_cputime_t(p, &cputime);
+ thread_group_cputime(p, &cputime);
*sample = cputime.sum_exec_runtime;
break;
}
@@ -288,7 +263,7 @@ static int posix_cpu_clock_get_task(struct task_struct *tsk,
struct timespec *tp)
{
int err = -EINVAL;
- unsigned long long rtn;
+ u64 rtn;
if (CPUCLOCK_PERTHREAD(which_clock)) {
if (same_thread_group(tsk, current))
@@ -299,7 +274,7 @@ static int posix_cpu_clock_get_task(struct task_struct *tsk,
}
if (!err)
- sample_to_timespec(which_clock, rtn, tp);
+ *tp = ns_to_timespec(rtn);
return err;
}
@@ -453,7 +428,7 @@ void posix_cpu_timers_exit_group(struct task_struct *tsk)
cleanup_timers(tsk->signal->cpu_timers);
}
-static inline int expires_gt(cputime_t expires, cputime_t new_exp)
+static inline int expires_gt(u64 expires, u64 new_exp)
{
return expires == 0 || expires > new_exp;
}
@@ -466,7 +441,7 @@ static void arm_timer(struct k_itimer *timer)
{
struct task_struct *p = timer->it.cpu.task;
struct list_head *head, *listpos;
- struct task_cputime_t *cputime_expires;
+ struct task_cputime *cputime_expires;
struct cpu_timer_list *const nt = &timer->it.cpu;
struct cpu_timer_list *next;
@@ -488,7 +463,7 @@ static void arm_timer(struct k_itimer *timer)
list_add(&nt->entry, listpos);
if (listpos == head) {
- unsigned long long exp = nt->expires;
+ u64 exp = nt->expires;
/*
* We are the new earliest-expiring POSIX 1.b timer, hence
@@ -499,16 +474,15 @@ static void arm_timer(struct k_itimer *timer)
switch (CPUCLOCK_WHICH(timer->it_clock)) {
case CPUCLOCK_PROF:
- if (expires_gt(cputime_expires->prof_exp, expires_to_cputime(exp)))
- cputime_expires->prof_exp = expires_to_cputime(exp);
+ if (expires_gt(cputime_expires->prof_exp, exp))
+ cputime_expires->prof_exp = exp;
break;
case CPUCLOCK_VIRT:
- if (expires_gt(cputime_expires->virt_exp, expires_to_cputime(exp)))
- cputime_expires->virt_exp = expires_to_cputime(exp);
+ if (expires_gt(cputime_expires->virt_exp, exp))
+ cputime_expires->virt_exp = exp;
break;
case CPUCLOCK_SCHED:
- if (cputime_expires->sched_exp == 0 ||
- cputime_expires->sched_exp > exp)
+ if (expires_gt(cputime_expires->sched_exp, exp))
cputime_expires->sched_exp = exp;
break;
}
@@ -559,20 +533,19 @@ static void cpu_timer_fire(struct k_itimer *timer)
* traversal.
*/
static int cpu_timer_sample_group(const clockid_t which_clock,
- struct task_struct *p,
- unsigned long long *sample)
+ struct task_struct *p, u64 *sample)
{
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
thread_group_cputimer(p, &cputime);
switch (CPUCLOCK_WHICH(which_clock)) {
default:
return -EINVAL;
case CPUCLOCK_PROF:
- *sample = cputime_to_expires(cputime.utime + cputime.stime);
+ *sample = cputime.utime + cputime.stime;
break;
case CPUCLOCK_VIRT:
- *sample = cputime_to_expires(cputime.utime);
+ *sample = cputime.utime;
break;
case CPUCLOCK_SCHED:
*sample = cputime.sum_exec_runtime;
@@ -593,12 +566,12 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
unsigned long flags;
struct sighand_struct *sighand;
struct task_struct *p = timer->it.cpu.task;
- unsigned long long old_expires, new_expires, old_incr, val;
+ u64 old_expires, new_expires, old_incr, val;
int ret;
WARN_ON_ONCE(p == NULL);
- new_expires = timespec_to_sample(timer->it_clock, &new->it_value);
+ new_expires = timespec_to_ns(&new->it_value);
/*
* Protect against sighand release/switch in exit/exec and p->cpu_timers
@@ -659,9 +632,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
bump_cpu_timer(timer, val);
if (val < timer->it.cpu.expires) {
old_expires = timer->it.cpu.expires - val;
- sample_to_timespec(timer->it_clock,
- old_expires,
- &old->it_value);
+ old->it_value = ns_to_timespec(old_expires);
} else {
old->it_value.tv_nsec = 1;
old->it_value.tv_sec = 0;
@@ -699,8 +670,7 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
* Install the new reload setting, and
* set up the signal and overrun bookkeeping.
*/
- timer->it.cpu.incr = timespec_to_sample(timer->it_clock,
- &new->it_interval);
+ timer->it.cpu.incr = timespec_to_ns(&new->it_interval);
/*
* This acts as a modification timestamp for the timer,
@@ -723,17 +693,15 @@ static int posix_cpu_timer_set(struct k_itimer *timer, int timer_flags,
ret = 0;
out:
- if (old) {
- sample_to_timespec(timer->it_clock,
- old_incr, &old->it_interval);
- }
+ if (old)
+ old->it_interval = ns_to_timespec(old_incr);
return ret;
}
static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
{
- unsigned long long now;
+ u64 now;
struct task_struct *p = timer->it.cpu.task;
WARN_ON_ONCE(p == NULL);
@@ -741,8 +709,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
/*
* Easy part: convert the reload time.
*/
- sample_to_timespec(timer->it_clock,
- timer->it.cpu.incr, &itp->it_interval);
+ itp->it_interval = ns_to_timespec(timer->it.cpu.incr);
if (timer->it.cpu.expires == 0) { /* Timer not armed at all. */
itp->it_value.tv_sec = itp->it_value.tv_nsec = 0;
@@ -761,7 +728,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
/*
* Protect against sighand release/switch in exit/exec and
* also make timer sampling safe if it ends up calling
- * thread_group_cputime_t().
+ * thread_group_cputime().
*/
sighand = lock_task_sighand(p, &flags);
if (unlikely(sighand == NULL)) {
@@ -771,8 +738,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
* Call the timer disarmed, nothing else to do.
*/
timer->it.cpu.expires = 0;
- sample_to_timespec(timer->it_clock, timer->it.cpu.expires,
- &itp->it_value);
+ itp->it_value = ns_to_timespec(timer->it.cpu.expires);
return;
} else {
cpu_timer_sample_group(timer->it_clock, p, &now);
@@ -781,9 +747,7 @@ static void posix_cpu_timer_get(struct k_itimer *timer, struct itimerspec *itp)
}
if (now < timer->it.cpu.expires) {
- sample_to_timespec(timer->it_clock,
- timer->it.cpu.expires - now,
- &itp->it_value);
+ itp->it_value = ns_to_timespec(timer->it.cpu.expires - now);
} else {
/*
* The timer should have expired already, but the firing
@@ -826,8 +790,8 @@ static void check_thread_timers(struct task_struct *tsk,
{
struct list_head *timers = tsk->cpu_timers;
struct signal_struct *const sig = tsk->signal;
- struct task_cputime_t *tsk_expires = &tsk->cputime_expires;
- unsigned long long expires;
+ struct task_cputime *tsk_expires = &tsk->cputime_expires;
+ u64 expires;
unsigned long soft;
/*
@@ -838,10 +802,10 @@ static void check_thread_timers(struct task_struct *tsk,
return;
expires = check_timers_list(timers, firing, prof_ticks(tsk));
- tsk_expires->prof_exp = expires_to_cputime(expires);
+ tsk_expires->prof_exp = expires;
expires = check_timers_list(++timers, firing, virt_ticks(tsk));
- tsk_expires->virt_exp = expires_to_cputime(expires);
+ tsk_expires->virt_exp = expires;
tsk_expires->sched_exp = check_timers_list(++timers, firing,
tsk->se.sum_exec_runtime);
@@ -891,13 +855,12 @@ static inline void stop_process_timers(struct signal_struct *sig)
}
static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
- unsigned long long *expires,
- unsigned long long cur_time, int signo)
+ u64 *expires, u64 cur_time, int signo)
{
if (!it->expires)
return;
- if (cur_time >= it->expires) {
+ if (cur_time >= cputime_to_nsecs(it->expires)) {
if (it->incr) {
it->expires += it->incr;
it->error += it->incr_error;
@@ -915,8 +878,8 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
__group_send_sig_info(signo, SEND_SIG_PRIV, tsk);
}
- if (it->expires && (!*expires || it->expires < *expires)) {
- *expires = it->expires;
+ if (it->expires && (!*expires || cputime_to_nsecs(it->expires) < *expires)) {
+ *expires = cputime_to_nsecs(it->expires);
}
}
@@ -929,10 +892,10 @@ static void check_process_timers(struct task_struct *tsk,
struct list_head *firing)
{
struct signal_struct *const sig = tsk->signal;
- unsigned long long utime, ptime, virt_expires, prof_expires;
- unsigned long long sum_sched_runtime, sched_expires;
+ u64 utime, ptime, virt_expires, prof_expires;
+ u64 sum_sched_runtime, sched_expires;
struct list_head *timers = sig->cpu_timers;
- struct task_cputime_t cputime;
+ struct task_cputime cputime;
unsigned long soft;
/*
@@ -952,8 +915,8 @@ static void check_process_timers(struct task_struct *tsk,
* Collect the current process totals.
*/
thread_group_cputimer(tsk, &cputime);
- utime = cputime_to_expires(cputime.utime);
- ptime = utime + cputime_to_expires(cputime.stime);
+ utime = cputime.utime;
+ ptime = utime + cputime.stime;
sum_sched_runtime = cputime.sum_exec_runtime;
prof_expires = check_timers_list(timers, firing, ptime);
@@ -969,10 +932,10 @@ static void check_process_timers(struct task_struct *tsk,
SIGVTALRM);
soft = READ_ONCE(sig->rlim[RLIMIT_CPU].rlim_cur);
if (soft != RLIM_INFINITY) {
- unsigned long psecs = cputime_to_secs(ptime);
+ unsigned long psecs = div_u64(ptime, NSEC_PER_SEC);
unsigned long hard =
READ_ONCE(sig->rlim[RLIMIT_CPU].rlim_max);
- cputime_t x;
+ u64 x;
if (psecs >= hard) {
/*
* At the hard limit, we just die.
@@ -991,14 +954,13 @@ static void check_process_timers(struct task_struct *tsk,
sig->rlim[RLIMIT_CPU].rlim_cur = soft;
}
}
- x = secs_to_cputime(soft);
- if (!prof_expires || x < prof_expires) {
+ x = soft * NSEC_PER_SEC;
+ if (!prof_expires || x < prof_expires)
prof_expires = x;
- }
}
- sig->cputime_expires.prof_exp = expires_to_cputime(prof_expires);
- sig->cputime_expires.virt_exp = expires_to_cputime(virt_expires);
+ sig->cputime_expires.prof_exp = prof_expires;
+ sig->cputime_expires.virt_exp = virt_expires;
sig->cputime_expires.sched_exp = sched_expires;
if (task_cputime_zero(&sig->cputime_expires))
stop_process_timers(sig);
@@ -1015,7 +977,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
struct sighand_struct *sighand;
unsigned long flags;
struct task_struct *p = timer->it.cpu.task;
- unsigned long long now;
+ u64 now;
WARN_ON_ONCE(p == NULL);
@@ -1035,7 +997,7 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
} else {
/*
* Protect arm_timer() and timer sampling in case of call to
- * thread_group_cputime_t().
+ * thread_group_cputime().
*/
sighand = lock_task_sighand(p, &flags);
if (unlikely(sighand == NULL)) {
@@ -1078,8 +1040,8 @@ void posix_cpu_timer_schedule(struct k_itimer *timer)
* Returns true if any field of the former is greater than the corresponding
* field of the latter if the latter field is set. Otherwise returns false.
*/
-static inline int task_cputime_expired(const struct task_cputime_t *sample,
- const struct task_cputime_t *expires)
+static inline int task_cputime_expired(const struct task_cputime *sample,
+ const struct task_cputime *expires)
{
if (expires->utime && sample->utime >= expires->utime)
return 1;
@@ -1106,9 +1068,9 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
struct signal_struct *sig;
if (!task_cputime_zero(&tsk->cputime_expires)) {
- struct task_cputime_t task_sample;
+ struct task_cputime task_sample;
- task_cputime_t(tsk, &task_sample.utime, &task_sample.stime);
+ task_cputime(tsk, &task_sample.utime, &task_sample.stime);
task_sample.sum_exec_runtime = tsk->se.sum_exec_runtime;
if (task_cputime_expired(&task_sample, &tsk->cputime_expires))
return 1;
@@ -1131,7 +1093,7 @@ static inline int fastpath_timer_check(struct task_struct *tsk)
*/
if (READ_ONCE(sig->cputimer.running) &&
!READ_ONCE(sig->cputimer.checking_timer)) {
- struct task_cputime_t group_sample;
+ struct task_cputime group_sample;
sample_cputime_atomic(&group_sample, &sig->cputimer.cputime_atomic);
@@ -1214,7 +1176,7 @@ void run_posix_cpu_timers(struct task_struct *tsk)
void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx,
cputime_t *newval, cputime_t *oldval)
{
- unsigned long long now;
+ u64 now, new;
WARN_ON_ONCE(clock_idx == CPUCLOCK_SCHED);
cpu_timer_sample_group(clock_idx, tsk, &now);
@@ -1226,31 +1188,33 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx,
* it to be absolute.
*/
if (*oldval) {
- if (*oldval <= now) {
+ if (cputime_to_nsecs(*oldval) <= now) {
/* Just about to fire. */
*oldval = cputime_one_jiffy;
} else {
- *oldval -= now;
+ *oldval -= nsecs_to_cputime(now);
}
}
if (!*newval)
return;
- *newval += now;
+ *newval += nsecs_to_cputime(now);
}
+ new = cputime_to_nsecs(*newval);
+
/*
* Update expiration cache if we are the earliest timer, or eventually
* RLIMIT_CPU limit is earlier than prof_exp cpu timer expire.
*/
switch (clock_idx) {
case CPUCLOCK_PROF:
- if (expires_gt(tsk->signal->cputime_expires.prof_exp, *newval))
- tsk->signal->cputime_expires.prof_exp = *newval;
+ if (expires_gt(tsk->signal->cputime_expires.prof_exp, new))
+ tsk->signal->cputime_expires.prof_exp = new;
break;
case CPUCLOCK_VIRT:
- if (expires_gt(tsk->signal->cputime_expires.virt_exp, *newval))
- tsk->signal->cputime_expires.virt_exp = *newval;
+ if (expires_gt(tsk->signal->cputime_expires.virt_exp, new))
+ tsk->signal->cputime_expires.virt_exp = new;
break;
}
@@ -1308,7 +1272,7 @@ static int do_cpu_nanosleep(const clockid_t which_clock, int flags,
/*
* We were interrupted by a signal.
*/
- sample_to_timespec(which_clock, timer.it.cpu.expires, rqtp);
+ *rqtp = ns_to_timespec(timer.it.cpu.expires);
error = posix_cpu_timer_set(&timer, 0, &zero_it, it);
if (!error) {
/*
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 20/37] itimer: Convert internal cputime_t units to nsec
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (18 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 19/37] posix-timers: Convert internals to use nsecs Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 21/37] sched: Remove temporary cputime_t accessors Frederic Weisbecker
` (16 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Use the new nsec based cputime accessors as part of the whole cputime
conversion from cputime_t to nsecs.
Also convert itimers to use nsec based internal counters. This simplifies
it and remove the whole game with error/inc_error which served to deal
with cputime_t random granularity.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/posix-timers.h | 2 +-
include/linux/sched.h | 6 ++--
include/trace/events/timer.h | 26 ++++++++---------
kernel/time/itimer.c | 64 +++++++++++++++---------------------------
kernel/time/posix-cpu-timers.c | 43 +++++++++++-----------------
5 files changed, 55 insertions(+), 86 deletions(-)
diff --git a/include/linux/posix-timers.h b/include/linux/posix-timers.h
index 890de52..64aa189 100644
--- a/include/linux/posix-timers.h
+++ b/include/linux/posix-timers.h
@@ -119,7 +119,7 @@ void run_posix_cpu_timers(struct task_struct *task);
void posix_cpu_timers_exit(struct task_struct *task);
void posix_cpu_timers_exit_group(struct task_struct *task);
void set_process_cpu_timer(struct task_struct *task, unsigned int clock_idx,
- cputime_t *newval, cputime_t *oldval);
+ u64 *newval, u64 *oldval);
long clock_nanosleep_restart(struct restart_block *restart_block);
diff --git a/include/linux/sched.h b/include/linux/sched.h
index 2768f94..cc10cd5 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -568,10 +568,8 @@ struct pacct_struct {
};
struct cpu_itimer {
- cputime_t expires;
- cputime_t incr;
- u32 error;
- u32 incr_error;
+ u64 expires;
+ u64 incr;
};
/**
diff --git a/include/trace/events/timer.h b/include/trace/events/timer.h
index 1448637..1bca99d 100644
--- a/include/trace/events/timer.h
+++ b/include/trace/events/timer.h
@@ -269,17 +269,17 @@ DEFINE_EVENT(hrtimer_class, hrtimer_cancel,
TRACE_EVENT(itimer_state,
TP_PROTO(int which, const struct itimerval *const value,
- cputime_t expires),
+ unsigned long long expires),
TP_ARGS(which, value, expires),
TP_STRUCT__entry(
- __field( int, which )
- __field( cputime_t, expires )
- __field( long, value_sec )
- __field( long, value_usec )
- __field( long, interval_sec )
- __field( long, interval_usec )
+ __field( int, which )
+ __field( unsigned long long, expires )
+ __field( long, value_sec )
+ __field( long, value_usec )
+ __field( long, interval_sec )
+ __field( long, interval_usec )
),
TP_fast_assign(
@@ -292,7 +292,7 @@ TRACE_EVENT(itimer_state,
),
TP_printk("which=%d expires=%llu it_value=%ld.%ld it_interval=%ld.%ld",
- __entry->which, (unsigned long long)__entry->expires,
+ __entry->which, __entry->expires,
__entry->value_sec, __entry->value_usec,
__entry->interval_sec, __entry->interval_usec)
);
@@ -305,14 +305,14 @@ TRACE_EVENT(itimer_state,
*/
TRACE_EVENT(itimer_expire,
- TP_PROTO(int which, struct pid *pid, cputime_t now),
+ TP_PROTO(int which, struct pid *pid, unsigned long long now),
TP_ARGS(which, pid, now),
TP_STRUCT__entry(
- __field( int , which )
- __field( pid_t, pid )
- __field( cputime_t, now )
+ __field( int , which )
+ __field( pid_t, pid )
+ __field( unsigned long long, now )
),
TP_fast_assign(
@@ -322,7 +322,7 @@ TRACE_EVENT(itimer_expire,
),
TP_printk("which=%d pid=%d now=%llu", __entry->which,
- (int) __entry->pid, (unsigned long long)__entry->now)
+ (int) __entry->pid, __entry->now)
);
#ifdef CONFIG_NO_HZ_COMMON
diff --git a/kernel/time/itimer.c b/kernel/time/itimer.c
index bb01ff4..a95f13c 100644
--- a/kernel/time/itimer.c
+++ b/kernel/time/itimer.c
@@ -45,35 +45,35 @@ static struct timeval itimer_get_remtime(struct hrtimer *timer)
static void get_cpu_itimer(struct task_struct *tsk, unsigned int clock_id,
struct itimerval *const value)
{
- cputime_t cval, cinterval;
+ u64 val, interval;
struct cpu_itimer *it = &tsk->signal->it[clock_id];
spin_lock_irq(&tsk->sighand->siglock);
- cval = it->expires;
- cinterval = it->incr;
- if (cval) {
+ val = it->expires;
+ interval = it->incr;
+ if (val) {
struct task_cputime cputime;
- cputime_t t;
+ u64 t;
thread_group_cputimer(tsk, &cputime);
if (clock_id == CPUCLOCK_PROF)
- t = nsecs_to_cputime(cputime.utime + cputime.stime);
+ t = cputime.utime + cputime.stime;
else
/* CPUCLOCK_VIRT */
- t = nsecs_to_cputime(cputime.utime);
+ t = cputime.utime;
- if (cval < t)
+ if (val < t)
/* about to fire */
- cval = cputime_one_jiffy;
+ val = TICK_NSEC;
else
- cval = cval - t;
+ val -= t;
}
spin_unlock_irq(&tsk->sighand->siglock);
- cputime_to_timeval(cval, &value->it_value);
- cputime_to_timeval(cinterval, &value->it_interval);
+ value->it_value = ns_to_timeval(val);
+ value->it_interval = ns_to_timeval(interval);
}
int do_getitimer(int which, struct itimerval *value)
@@ -129,55 +129,35 @@ enum hrtimer_restart it_real_fn(struct hrtimer *timer)
return HRTIMER_NORESTART;
}
-static inline u32 cputime_sub_ns(cputime_t ct, s64 real_ns)
-{
- struct timespec ts;
- s64 cpu_ns;
-
- cputime_to_timespec(ct, &ts);
- cpu_ns = timespec_to_ns(&ts);
-
- return (cpu_ns <= real_ns) ? 0 : cpu_ns - real_ns;
-}
-
static void set_cpu_itimer(struct task_struct *tsk, unsigned int clock_id,
const struct itimerval *const value,
struct itimerval *const ovalue)
{
- cputime_t cval, nval, cinterval, ninterval;
- s64 ns_ninterval, ns_nval;
- u32 error, incr_error;
+ u64 oval, nval, ointerval, ninterval;
struct cpu_itimer *it = &tsk->signal->it[clock_id];
- nval = timeval_to_cputime(&value->it_value);
- ns_nval = timeval_to_ns(&value->it_value);
- ninterval = timeval_to_cputime(&value->it_interval);
- ns_ninterval = timeval_to_ns(&value->it_interval);
-
- error = cputime_sub_ns(nval, ns_nval);
- incr_error = cputime_sub_ns(ninterval, ns_ninterval);
+ nval = timeval_to_ns(&value->it_value);
+ ninterval = timeval_to_ns(&value->it_interval);
spin_lock_irq(&tsk->sighand->siglock);
- cval = it->expires;
- cinterval = it->incr;
- if (cval || nval) {
+ oval = it->expires;
+ ointerval = it->incr;
+ if (oval || nval) {
if (nval > 0)
- nval += cputime_one_jiffy;
- set_process_cpu_timer(tsk, clock_id, &nval, &cval);
+ nval += TICK_NSEC;
+ set_process_cpu_timer(tsk, clock_id, &nval, &oval);
}
it->expires = nval;
it->incr = ninterval;
- it->error = error;
- it->incr_error = incr_error;
trace_itimer_state(clock_id == CPUCLOCK_VIRT ?
ITIMER_VIRTUAL : ITIMER_PROF, value, nval);
spin_unlock_irq(&tsk->sighand->siglock);
if (ovalue) {
- cputime_to_timeval(cval, &ovalue->it_value);
- cputime_to_timeval(cinterval, &ovalue->it_interval);
+ ovalue->it_value = ns_to_timeval(oval);
+ ovalue->it_interval = ns_to_timeval(ointerval);
}
}
diff --git a/kernel/time/posix-cpu-timers.c b/kernel/time/posix-cpu-timers.c
index 45be3ce..b4377a5 100644
--- a/kernel/time/posix-cpu-timers.c
+++ b/kernel/time/posix-cpu-timers.c
@@ -20,10 +20,10 @@
*/
void update_rlimit_cpu(struct task_struct *task, unsigned long rlim_new)
{
- cputime_t cputime = secs_to_cputime(rlim_new);
+ u64 nsecs = rlim_new * NSEC_PER_SEC;
spin_lock_irq(&task->sighand->siglock);
- set_process_cpu_timer(task, CPUCLOCK_PROF, &cputime, NULL);
+ set_process_cpu_timer(task, CPUCLOCK_PROF, &nsecs, NULL);
spin_unlock_irq(&task->sighand->siglock);
}
@@ -860,17 +860,11 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
if (!it->expires)
return;
- if (cur_time >= cputime_to_nsecs(it->expires)) {
- if (it->incr) {
+ if (cur_time >= it->expires) {
+ if (it->incr)
it->expires += it->incr;
- it->error += it->incr_error;
- if (it->error >= TICK_NSEC) {
- it->expires -= cputime_one_jiffy;
- it->error -= TICK_NSEC;
- }
- } else {
+ else
it->expires = 0;
- }
trace_itimer_expire(signo == SIGPROF ?
ITIMER_PROF : ITIMER_VIRTUAL,
@@ -878,9 +872,8 @@ static void check_cpu_itimer(struct task_struct *tsk, struct cpu_itimer *it,
__group_send_sig_info(signo, SEND_SIG_PRIV, tsk);
}
- if (it->expires && (!*expires || cputime_to_nsecs(it->expires) < *expires)) {
- *expires = cputime_to_nsecs(it->expires);
- }
+ if (it->expires && (!*expires || it->expires < *expires))
+ *expires = it->expires;
}
/*
@@ -1174,9 +1167,9 @@ void run_posix_cpu_timers(struct task_struct *tsk)
* The tsk->sighand->siglock must be held by the caller.
*/
void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx,
- cputime_t *newval, cputime_t *oldval)
+ u64 *newval, u64 *oldval)
{
- u64 now, new;
+ u64 now;
WARN_ON_ONCE(clock_idx == CPUCLOCK_SCHED);
cpu_timer_sample_group(clock_idx, tsk, &now);
@@ -1188,33 +1181,31 @@ void set_process_cpu_timer(struct task_struct *tsk, unsigned int clock_idx,
* it to be absolute.
*/
if (*oldval) {
- if (cputime_to_nsecs(*oldval) <= now) {
+ if (*oldval <= now) {
/* Just about to fire. */
- *oldval = cputime_one_jiffy;
+ *oldval = TICK_NSEC;
} else {
- *oldval -= nsecs_to_cputime(now);
+ *oldval -= now;
}
}
if (!*newval)
return;
- *newval += nsecs_to_cputime(now);
+ *newval += now;
}
- new = cputime_to_nsecs(*newval);
-
/*
* Update expiration cache if we are the earliest timer, or eventually
* RLIMIT_CPU limit is earlier than prof_exp cpu timer expire.
*/
switch (clock_idx) {
case CPUCLOCK_PROF:
- if (expires_gt(tsk->signal->cputime_expires.prof_exp, new))
- tsk->signal->cputime_expires.prof_exp = new;
+ if (expires_gt(tsk->signal->cputime_expires.prof_exp, *newval))
+ tsk->signal->cputime_expires.prof_exp = *newval;
break;
case CPUCLOCK_VIRT:
- if (expires_gt(tsk->signal->cputime_expires.virt_exp, new))
- tsk->signal->cputime_expires.virt_exp = new;
+ if (expires_gt(tsk->signal->cputime_expires.virt_exp, *newval))
+ tsk->signal->cputime_expires.virt_exp = *newval;
break;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 21/37] sched: Remove temporary cputime_t accessors
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (19 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 20/37] itimer: Convert internal cputime_t units to nsec Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 22/37] cputime: Push time to account_user_time() in nsecs Frederic Weisbecker
` (15 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Now that the whole cputime conversion to nsec units is complete, we
can remove the compatibility accessors.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/sched.h | 40 ----------------------------------------
1 file changed, 40 deletions(-)
diff --git a/include/linux/sched.h b/include/linux/sched.h
index cc10cd5..762a0b6 100644
--- a/include/linux/sched.h
+++ b/include/linux/sched.h
@@ -29,7 +29,6 @@ struct sched_param {
#include <asm/page.h>
#include <asm/ptrace.h>
-#include <linux/cputime.h>
#include <linux/smp.h>
#include <linux/sem.h>
@@ -613,13 +612,6 @@ struct task_cputime {
unsigned long long sum_exec_runtime;
};
-/* Temporary type to ease cputime_t to nsecs conversion */
-struct task_cputime_t {
- cputime_t utime;
- cputime_t stime;
- unsigned long long sum_exec_runtime;
-};
-
/* Alternate field names when used to cache expirations. */
#define virt_exp utime
#define prof_exp stime
@@ -2291,27 +2283,6 @@ static inline void task_cputime_scaled(struct task_struct *t,
}
#endif
-static inline void task_cputime_t(struct task_struct *t,
- cputime_t *utime, cputime_t *stime)
-{
- u64 ut, st;
-
- task_cputime(t, &ut, &st);
- *utime = nsecs_to_cputime(ut);
- *stime = nsecs_to_cputime(st);
-}
-
-static inline void task_cputime_t_scaled(struct task_struct *t,
- cputime_t *utimescaled,
- cputime_t *stimescaled)
-{
- u64 ut, st;
-
- task_cputime_scaled(t, &ut, &st);
- *utimescaled = nsecs_to_cputime(ut);
- *stimescaled = nsecs_to_cputime(st);
-}
-
extern void task_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st);
extern void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st);
@@ -3528,17 +3499,6 @@ static __always_inline bool need_resched(void)
void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times);
void thread_group_cputimer(struct task_struct *tsk, struct task_cputime *times);
-static inline void thread_group_cputime_t(struct task_struct *tsk,
- struct task_cputime_t *cputime)
-{
- struct task_cputime times;
-
- thread_group_cputime(tsk, ×);
- cputime->utime = nsecs_to_cputime(times.utime);
- cputime->stime = nsecs_to_cputime(times.stime);
- cputime->sum_exec_runtime = times.sum_exec_runtime;
-}
-
/*
* Reevaluate whether the task has signals pending delivery.
* Wake the task if so.
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 22/37] cputime: Push time to account_user_time() in nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (20 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 21/37] sched: Remove temporary cputime_t accessors Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:19 ` [PATCH 23/37] cputime: Push time to account_steal_time() " Frederic Weisbecker
` (14 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This is one more step toward converting cputime accounting to pure nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/kernel/time.c | 2 +-
arch/powerpc/kernel/time.c | 2 +-
arch/s390/kernel/vtime.c | 2 +-
include/linux/kernel_stat.h | 2 +-
kernel/sched/cputime.c | 42 ++++++++++++++++++++++++------------------
5 files changed, 28 insertions(+), 22 deletions(-)
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index d040f12..ed98b26 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -67,7 +67,7 @@ void vtime_flush(struct task_struct *tsk)
cputime_t delta;
if (ti->utime)
- account_user_time(tsk, cycle_to_cputime(ti->utime));
+ account_user_time(tsk, cputime_to_nsecs(cycle_to_cputime(ti->utime)));
if (ti->gtime)
account_guest_time(tsk, cycle_to_cputime(ti->gtime));
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 3cca82e..c3931d8 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -393,7 +393,7 @@ void vtime_flush(struct task_struct *tsk)
struct cpu_accounting_data *acct = get_accounting(tsk);
if (acct->utime)
- account_user_time(tsk, acct->utime);
+ account_user_time(tsk, cputime_to_nsecs(acct->utime));
if (acct->utime_scaled)
tsk->utimescaled += cputime_to_nsecs(acct->utime_scaled);
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index f2fc274..ca206d1 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -166,7 +166,7 @@ static int do_account_vtime(struct task_struct *tsk)
/* Push account value */
if (user) {
- account_user_time(tsk, user);
+ account_user_time(tsk, cputime_to_nsecs(user));
tsk->utimescaled += cputime_to_nsecs(scale_vtime(user));
}
diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h
index c3e38de..b716001 100644
--- a/include/linux/kernel_stat.h
+++ b/include/linux/kernel_stat.h
@@ -78,7 +78,7 @@ static inline unsigned int kstat_cpu_irqs_sum(unsigned int cpu)
return kstat_cpu(cpu).irqs_sum;
}
-extern void account_user_time(struct task_struct *, cputime_t);
+extern void account_user_time(struct task_struct *, u64);
extern void account_guest_time(struct task_struct *, cputime_t);
extern void account_system_time(struct task_struct *, int, cputime_t);
extern void account_system_index_time(struct task_struct *, cputime_t,
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index f7b9624..55d31c3 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -116,18 +116,18 @@ static inline void task_group_account_field(struct task_struct *p, int index,
* @p: the process that the cpu time gets accounted to
* @cputime: the cpu time spent in user space since the last update
*/
-void account_user_time(struct task_struct *p, cputime_t cputime)
+void account_user_time(struct task_struct *p, u64 cputime)
{
int index;
/* Add user time to process. */
- p->utime += cputime_to_nsecs(cputime);
- account_group_user_time(p, cputime_to_nsecs(cputime));
+ p->utime += cputime;
+ 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, cputime_to_nsecs(cputime));
+ task_group_account_field(p, index, cputime);
/* Account for user time used */
acct_account_cputime(p);
@@ -363,8 +363,9 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
struct rq *rq, int ticks)
{
- u64 cputime = (__force u64) cputime_one_jiffy * ticks;
+ u64 old_cputime = (__force u64) cputime_one_jiffy * ticks;
cputime_t other;
+ u64 cputime;
/*
* When returning from idle, many ticks can get accounted at
@@ -374,9 +375,11 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
* other time can exceed ticks occasionally.
*/
other = account_other_time(ULONG_MAX);
- if (other >= cputime)
+ if (other >= old_cputime)
return;
- cputime -= other;
+
+ old_cputime -= other;
+ cputime = cputime_to_nsecs(old_cputime);
if (this_cpu_ksoftirqd() == p) {
/*
@@ -384,15 +387,16 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
* So, we have to handle it separately here.
* Also, p->stime needs to be updated for ksoftirqd.
*/
- account_system_index_time(p, cputime, CPUTIME_SOFTIRQ);
+ account_system_index_time(p, old_cputime, CPUTIME_SOFTIRQ);
} else if (user_tick) {
account_user_time(p, cputime);
} else if (p == rq->idle) {
- account_idle_time(cputime);
+ account_idle_time(old_cputime);
} else if (p->flags & PF_VCPU) { /* System time or guest time */
- account_guest_time(p, cputime);
+
+ account_guest_time(p, old_cputime);
} else {
- account_system_index_time(p, cputime, CPUTIME_SYSTEM);
+ account_system_index_time(p, old_cputime, CPUTIME_SYSTEM);
}
}
@@ -473,7 +477,8 @@ void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
*/
void account_process_tick(struct task_struct *p, int user_tick)
{
- cputime_t cputime, steal;
+ cputime_t old_cputime, steal;
+ u64 cputime;
struct rq *rq = this_rq();
if (vtime_accounting_cpu_enabled())
@@ -484,20 +489,21 @@ void account_process_tick(struct task_struct *p, int user_tick)
return;
}
- cputime = cputime_one_jiffy;
+ old_cputime = cputime_one_jiffy;
steal = steal_account_process_time(ULONG_MAX);
- if (steal >= cputime)
+ if (steal >= old_cputime)
return;
- cputime -= steal;
+ old_cputime -= steal;
+ cputime = cputime_to_nsecs(old_cputime);
if (user_tick)
account_user_time(p, cputime);
else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET))
- account_system_time(p, HARDIRQ_OFFSET, cputime);
+ account_system_time(p, HARDIRQ_OFFSET, old_cputime);
else
- account_idle_time(cputime);
+ account_idle_time(old_cputime);
}
/*
@@ -736,7 +742,7 @@ void vtime_account_user(struct task_struct *tsk)
tsk->vtime_snap_whence = VTIME_SYS;
if (vtime_delta(tsk)) {
delta_cpu = get_vtime_delta(tsk);
- account_user_time(tsk, delta_cpu);
+ account_user_time(tsk, cputime_to_nsecs(delta_cpu));
}
write_seqcount_end(&tsk->vtime_seqcount);
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 23/37] cputime: Push time to account_steal_time() in nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (21 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 22/37] cputime: Push time to account_user_time() in nsecs Frederic Weisbecker
@ 2017-01-22 18:19 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 24/37] cputime: Push time to account_idle_time() " Frederic Weisbecker
` (13 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:19 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This is one more step toward converting cputime accounting to pure nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/powerpc/kernel/time.c | 2 +-
arch/s390/kernel/vtime.c | 2 +-
include/linux/kernel_stat.h | 2 +-
kernel/sched/cputime.c | 11 ++++++-----
4 files changed, 9 insertions(+), 8 deletions(-)
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index c3931d8..53e5982 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -402,7 +402,7 @@ void vtime_flush(struct task_struct *tsk)
account_guest_time(tsk, acct->gtime);
if (acct->steal_time)
- account_steal_time(acct->steal_time);
+ account_steal_time(cputime_to_nsecs(acct->steal_time));
if (acct->idle_time)
account_idle_time(acct->idle_time);
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index ca206d1..1e7023c 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -188,7 +188,7 @@ static int do_account_vtime(struct task_struct *tsk)
steal = S390_lowcore.steal_timer;
if ((s64) steal > 0) {
S390_lowcore.steal_timer = 0;
- account_steal_time(steal);
+ account_steal_time(cputime_to_nsecs(steal));
}
return virt_timer_forward(user + guest + system + hardirq + softirq);
diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h
index b716001..1d55d10 100644
--- a/include/linux/kernel_stat.h
+++ b/include/linux/kernel_stat.h
@@ -83,7 +83,7 @@ extern void account_guest_time(struct task_struct *, cputime_t);
extern void account_system_time(struct task_struct *, int, cputime_t);
extern void account_system_index_time(struct task_struct *, cputime_t,
enum cpu_usage_stat);
-extern void account_steal_time(cputime_t);
+extern void account_steal_time(u64);
extern void account_idle_time(cputime_t);
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 55d31c3..9a80287 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -207,11 +207,11 @@ void account_system_time(struct task_struct *p, int hardirq_offset,
* Account for involuntary wait time.
* @cputime: the cpu time spent in involuntary wait
*/
-void account_steal_time(cputime_t cputime)
+void account_steal_time(u64 cputime)
{
u64 *cpustat = kcpustat_this_cpu->cpustat;
- cpustat[CPUTIME_STEAL] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_STEAL] += cputime;
}
/*
@@ -239,14 +239,15 @@ static __always_inline cputime_t steal_account_process_time(cputime_t maxtime)
#ifdef CONFIG_PARAVIRT
if (static_key_false(¶virt_steal_enabled)) {
cputime_t steal_cputime;
- u64 steal;
+ u64 steal, rounded;
steal = paravirt_steal_clock(smp_processor_id());
steal -= this_rq()->prev_steal_time;
steal_cputime = min(nsecs_to_cputime(steal), maxtime);
- account_steal_time(steal_cputime);
- this_rq()->prev_steal_time += cputime_to_nsecs(steal_cputime);
+ rounded = cputime_to_nsecs(steal_cputime);
+ account_steal_time(rounded);
+ this_rq()->prev_steal_time += rounded;
return steal_cputime;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 24/37] cputime: Push time to account_idle_time() in nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (22 preceding siblings ...)
2017-01-22 18:19 ` [PATCH 23/37] cputime: Push time to account_steal_time() " Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 25/37] cputime: Push time to account_system_time() " Frederic Weisbecker
` (12 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This is one more step toward converting cputime accounting to pure nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/kernel/time.c | 2 +-
arch/powerpc/kernel/time.c | 2 +-
arch/s390/kernel/idle.c | 2 +-
include/linux/kernel_stat.h | 2 +-
kernel/sched/cputime.c | 18 +++++++++---------
5 files changed, 13 insertions(+), 13 deletions(-)
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index ed98b26..5dc801d 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -73,7 +73,7 @@ void vtime_flush(struct task_struct *tsk)
account_guest_time(tsk, cycle_to_cputime(ti->gtime));
if (ti->idle_time)
- account_idle_time(cycle_to_cputime(ti->idle_time));
+ account_idle_time(cputime_to_nsecs(cycle_to_cputime(ti->idle_time)));
if (ti->stime) {
delta = cycle_to_cputime(ti->stime);
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 53e5982..739897a 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -405,7 +405,7 @@ void vtime_flush(struct task_struct *tsk)
account_steal_time(cputime_to_nsecs(acct->steal_time));
if (acct->idle_time)
- account_idle_time(acct->idle_time);
+ account_idle_time(cputime_to_nsecs(acct->idle_time));
if (acct->stime)
account_system_index_time(tsk, acct->stime, CPUTIME_SYSTEM);
diff --git a/arch/s390/kernel/idle.c b/arch/s390/kernel/idle.c
index 7a55c29..99f1d81 100644
--- a/arch/s390/kernel/idle.c
+++ b/arch/s390/kernel/idle.c
@@ -43,7 +43,7 @@ void enabled_wait(void)
idle->clock_idle_enter = idle->clock_idle_exit = 0ULL;
idle->idle_time += idle_time;
idle->idle_count++;
- account_idle_time(idle_time);
+ account_idle_time(cputime_to_nsecs(idle_time));
write_seqcount_end(&idle->seqcount);
}
NOKPROBE_SYMBOL(enabled_wait);
diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h
index 1d55d10..e1cd897 100644
--- a/include/linux/kernel_stat.h
+++ b/include/linux/kernel_stat.h
@@ -84,7 +84,7 @@ extern void account_system_time(struct task_struct *, int, cputime_t);
extern void account_system_index_time(struct task_struct *, cputime_t,
enum cpu_usage_stat);
extern void account_steal_time(u64);
-extern void account_idle_time(cputime_t);
+extern void account_idle_time(u64);
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
static inline void account_process_tick(struct task_struct *tsk, int user)
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index 9a80287..fd5375f 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -218,15 +218,15 @@ void account_steal_time(u64 cputime)
* Account for idle time.
* @cputime: the cpu time spent in idle wait
*/
-void account_idle_time(cputime_t cputime)
+void account_idle_time(u64 cputime)
{
u64 *cpustat = kcpustat_this_cpu->cpustat;
struct rq *rq = this_rq();
if (atomic_read(&rq->nr_iowait) > 0)
- cpustat[CPUTIME_IOWAIT] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_IOWAIT] += cputime;
else
- cpustat[CPUTIME_IDLE] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_IDLE] += cputime;
}
/*
@@ -392,7 +392,7 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
} else if (user_tick) {
account_user_time(p, cputime);
} else if (p == rq->idle) {
- account_idle_time(old_cputime);
+ account_idle_time(cputime);
} else if (p->flags & PF_VCPU) { /* System time or guest time */
account_guest_time(p, old_cputime);
@@ -504,7 +504,7 @@ void account_process_tick(struct task_struct *p, int user_tick)
else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET))
account_system_time(p, HARDIRQ_OFFSET, old_cputime);
else
- account_idle_time(old_cputime);
+ account_idle_time(cputime);
}
/*
@@ -513,15 +513,15 @@ void account_process_tick(struct task_struct *p, int user_tick)
*/
void account_idle_ticks(unsigned long ticks)
{
- cputime_t cputime, steal;
+ u64 cputime, steal;
if (sched_clock_irqtime) {
irqtime_account_idle_ticks(ticks);
return;
}
- cputime = jiffies_to_cputime(ticks);
- steal = steal_account_process_time(ULONG_MAX);
+ cputime = ticks * TICK_NSEC;
+ steal = cputime_to_nsecs(steal_account_process_time(ULONG_MAX));
if (steal >= cputime)
return;
@@ -787,7 +787,7 @@ void vtime_account_idle(struct task_struct *tsk)
{
cputime_t delta_cpu = get_vtime_delta(tsk);
- account_idle_time(delta_cpu);
+ account_idle_time(cputime_to_nsecs(delta_cpu));
}
void arch_vtime_task_switch(struct task_struct *prev)
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 25/37] cputime: Push time to account_system_time() in nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (23 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 24/37] cputime: Push time to account_idle_time() " Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 26/37] cputime: Complete nsec conversion of tick based accounting Frederic Weisbecker
` (11 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This is one more step toward converting cputime accounting to pure nsecs.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/kernel/time.c | 11 ++++++-----
arch/powerpc/kernel/time.c | 15 ++++++++-------
arch/s390/kernel/idle.c | 2 +-
arch/s390/kernel/vtime.c | 6 +++---
include/linux/kernel_stat.h | 7 +++----
kernel/sched/cputime.c | 39 +++++++++++++++++++--------------------
6 files changed, 40 insertions(+), 40 deletions(-)
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index 5dc801d..f15bca4 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -21,6 +21,7 @@
#include <linux/timex.h>
#include <linux/timekeeper_internal.h>
#include <linux/platform_device.h>
+#include <linux/cputime.h>
#include <asm/machvec.h>
#include <asm/delay.h>
@@ -64,29 +65,29 @@ extern cputime_t cycle_to_cputime(u64 cyc);
void vtime_flush(struct task_struct *tsk)
{
struct thread_info *ti = task_thread_info(tsk);
- cputime_t delta;
+ u64 delta;
if (ti->utime)
account_user_time(tsk, cputime_to_nsecs(cycle_to_cputime(ti->utime)));
if (ti->gtime)
- account_guest_time(tsk, cycle_to_cputime(ti->gtime));
+ account_guest_time(tsk, cputime_to_nsecs(cycle_to_cputime(ti->gtime)));
if (ti->idle_time)
account_idle_time(cputime_to_nsecs(cycle_to_cputime(ti->idle_time)));
if (ti->stime) {
- delta = cycle_to_cputime(ti->stime);
+ delta = cputime_to_nsecs(cycle_to_cputime(ti->stime));
account_system_index_time(tsk, delta, CPUTIME_SYSTEM);
}
if (ti->hardirq_time) {
- delta = cycle_to_cputime(ti->hardirq_time);
+ delta = cputime_to_nsecs(cycle_to_cputime(ti->hardirq_time));
account_system_index_time(tsk, delta, CPUTIME_IRQ);
}
if (ti->softirq_time) {
- delta = cycle_to_cputime(ti->softirq_time);
+ delta = cputime_to_nsecs(cycle_to_cputime(ti->softirq_time));
account_system_index_time(tsk, delta, CPUTIME_SOFTIRQ);
}
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 739897a..01f53bf 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -57,6 +57,7 @@
#include <linux/clk-provider.h>
#include <linux/suspend.h>
#include <linux/rtc.h>
+#include <linux/cputime.h>
#include <asm/trace.h>
#include <asm/io.h>
@@ -72,7 +73,6 @@
#include <asm/smp.h>
#include <asm/vdso_datapage.h>
#include <asm/firmware.h>
-#include <asm/cputime.h>
#include <asm/asm-prototypes.h>
/* powerpc clocksource/clockevent code */
@@ -399,7 +399,7 @@ void vtime_flush(struct task_struct *tsk)
tsk->utimescaled += cputime_to_nsecs(acct->utime_scaled);
if (acct->gtime)
- account_guest_time(tsk, acct->gtime);
+ account_guest_time(tsk, cputime_to_nsecs(acct->gtime));
if (acct->steal_time)
account_steal_time(cputime_to_nsecs(acct->steal_time));
@@ -408,16 +408,17 @@ void vtime_flush(struct task_struct *tsk)
account_idle_time(cputime_to_nsecs(acct->idle_time));
if (acct->stime)
- account_system_index_time(tsk, acct->stime, CPUTIME_SYSTEM);
-
+ account_system_index_time(tsk, cputime_to_nsecs(acct->stime),
+ CPUTIME_SYSTEM);
if (acct->stime_scaled)
tsk->stimescaled += cputime_to_nsecs(acct->stime_scaled);
if (acct->hardirq_time)
- account_system_index_time(tsk, acct->hardirq_time, CPUTIME_IRQ);
-
+ account_system_index_time(tsk, cputime_to_nsecs(acct->hardirq_time),
+ CPUTIME_IRQ);
if (acct->softirq_time)
- account_system_index_time(tsk, acct->softirq_time, CPUTIME_SOFTIRQ);
+ account_system_index_time(tsk, cputime_to_nsecs(acct->softirq_time),
+ CPUTIME_SOFTIRQ);
acct->utime = 0;
acct->utime_scaled = 0;
diff --git a/arch/s390/kernel/idle.c b/arch/s390/kernel/idle.c
index 99f1d81..5c0e08e7 100644
--- a/arch/s390/kernel/idle.c
+++ b/arch/s390/kernel/idle.c
@@ -12,7 +12,7 @@
#include <linux/notifier.h>
#include <linux/init.h>
#include <linux/cpu.h>
-#include <asm/cputime.h>
+#include <linux/cputime.h>
#include <asm/nmi.h>
#include <asm/smp.h>
#include "entry.h"
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index 1e7023c..b4a3e9e 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -6,13 +6,13 @@
*/
#include <linux/kernel_stat.h>
+#include <linux/cputime.h>
#include <linux/export.h>
#include <linux/kernel.h>
#include <linux/timex.h>
#include <linux/types.h>
#include <linux/time.h>
-#include <asm/cputime.h>
#include <asm/vtimer.h>
#include <asm/vtime.h>
#include <asm/cpu_mf.h>
@@ -115,7 +115,7 @@ static void account_system_index_scaled(struct task_struct *p,
enum cpu_usage_stat index)
{
p->stimescaled += cputime_to_nsecs(scaled);
- account_system_index_time(p, cputime, index);
+ account_system_index_time(p, cputime_to_nsecs(cputime), index);
}
/*
@@ -171,7 +171,7 @@ static int do_account_vtime(struct task_struct *tsk)
}
if (guest) {
- account_guest_time(tsk, guest);
+ account_guest_time(tsk, cputime_to_nsecs(guest));
tsk->utimescaled += cputime_to_nsecs(scale_vtime(guest));
}
diff --git a/include/linux/kernel_stat.h b/include/linux/kernel_stat.h
index e1cd897..66be8b6 100644
--- a/include/linux/kernel_stat.h
+++ b/include/linux/kernel_stat.h
@@ -9,7 +9,6 @@
#include <linux/sched.h>
#include <linux/vtime.h>
#include <asm/irq.h>
-#include <linux/cputime.h>
/*
* 'kernel_stat.h' contains the definitions needed for doing
@@ -79,9 +78,9 @@ static inline unsigned int kstat_cpu_irqs_sum(unsigned int cpu)
}
extern void account_user_time(struct task_struct *, u64);
-extern void account_guest_time(struct task_struct *, cputime_t);
-extern void account_system_time(struct task_struct *, int, cputime_t);
-extern void account_system_index_time(struct task_struct *, cputime_t,
+extern void account_guest_time(struct task_struct *, u64);
+extern void account_system_time(struct task_struct *, int, u64);
+extern void account_system_index_time(struct task_struct *, u64,
enum cpu_usage_stat);
extern void account_steal_time(u64);
extern void account_idle_time(u64);
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index fd5375f..d28e9c5 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -4,6 +4,7 @@
#include <linux/kernel_stat.h>
#include <linux/static_key.h>
#include <linux/context_tracking.h>
+#include <linux/cputime.h>
#include "sched.h"
#ifdef CONFIG_PARAVIRT
#include <asm/paravirt.h>
@@ -138,22 +139,22 @@ void account_user_time(struct task_struct *p, u64 cputime)
* @p: the process that the cpu time gets accounted to
* @cputime: the cpu time spent in virtual machine since the last update
*/
-void account_guest_time(struct task_struct *p, cputime_t cputime)
+void account_guest_time(struct task_struct *p, u64 cputime)
{
u64 *cpustat = kcpustat_this_cpu->cpustat;
/* Add guest time to process. */
- p->utime += cputime_to_nsecs(cputime);
- account_group_user_time(p, cputime_to_nsecs(cputime));
- p->gtime += cputime_to_nsecs(cputime);
+ p->utime += cputime;
+ account_group_user_time(p, cputime);
+ p->gtime += cputime;
/* Add guest time to cpustat. */
if (task_nice(p) > 0) {
- cpustat[CPUTIME_NICE] += cputime_to_nsecs(cputime);
- cpustat[CPUTIME_GUEST_NICE] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_NICE] += cputime;
+ cpustat[CPUTIME_GUEST_NICE] += cputime;
} else {
- cpustat[CPUTIME_USER] += cputime_to_nsecs(cputime);
- cpustat[CPUTIME_GUEST] += cputime_to_nsecs(cputime);
+ cpustat[CPUTIME_USER] += cputime;
+ cpustat[CPUTIME_GUEST] += cputime;
}
}
@@ -164,14 +165,14 @@ void account_guest_time(struct task_struct *p, cputime_t cputime)
* @index: pointer to cpustat field that has to be updated
*/
void account_system_index_time(struct task_struct *p,
- cputime_t cputime, enum cpu_usage_stat index)
+ u64 cputime, enum cpu_usage_stat index)
{
/* Add system time to process. */
- p->stime += cputime_to_nsecs(cputime);
- account_group_system_time(p, cputime_to_nsecs(cputime));
+ p->stime += cputime;
+ account_group_system_time(p, cputime);
/* Add system time to cpustat. */
- task_group_account_field(p, index, cputime_to_nsecs(cputime));
+ task_group_account_field(p, index, cputime);
/* Account for system time used */
acct_account_cputime(p);
@@ -183,8 +184,7 @@ void account_system_index_time(struct task_struct *p,
* @hardirq_offset: the offset to subtract from hardirq_count()
* @cputime: the cpu time spent in kernel space since the last update
*/
-void account_system_time(struct task_struct *p, int hardirq_offset,
- cputime_t cputime)
+void account_system_time(struct task_struct *p, int hardirq_offset, u64 cputime)
{
int index;
@@ -388,16 +388,15 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
* So, we have to handle it separately here.
* Also, p->stime needs to be updated for ksoftirqd.
*/
- account_system_index_time(p, old_cputime, CPUTIME_SOFTIRQ);
+ account_system_index_time(p, cputime, CPUTIME_SOFTIRQ);
} else if (user_tick) {
account_user_time(p, cputime);
} else if (p == rq->idle) {
account_idle_time(cputime);
} else if (p->flags & PF_VCPU) { /* System time or guest time */
-
- account_guest_time(p, old_cputime);
+ account_guest_time(p, cputime);
} else {
- account_system_index_time(p, old_cputime, CPUTIME_SYSTEM);
+ account_system_index_time(p, cputime, CPUTIME_SYSTEM);
}
}
@@ -502,7 +501,7 @@ void account_process_tick(struct task_struct *p, int user_tick)
if (user_tick)
account_user_time(p, cputime);
else if ((p != rq->idle) || (irq_count() != HARDIRQ_OFFSET))
- account_system_time(p, HARDIRQ_OFFSET, old_cputime);
+ account_system_time(p, HARDIRQ_OFFSET, cputime);
else
account_idle_time(cputime);
}
@@ -722,7 +721,7 @@ 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);
+ account_system_time(tsk, irq_count(), cputime_to_nsecs(delta_cpu));
}
void vtime_account_system(struct task_struct *tsk)
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 26/37] cputime: Complete nsec conversion of tick based accounting
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (24 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 25/37] cputime: Push time to account_system_time() " Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 27/37] vtime: Return nsecs instead of cputime_t to account Frederic Weisbecker
` (10 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This is the final step toward tick based cputime conversion. Now that
whole cputime accounting engine accounts in nsecs, we can convert the
very source of the cputime to account into nsec.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/sched/cputime.c | 52 +++++++++++++++++++++-----------------------------
1 file changed, 22 insertions(+), 30 deletions(-)
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index d28e9c5..f3a56bf 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -75,14 +75,13 @@ void irqtime_account_irq(struct task_struct *curr)
}
EXPORT_SYMBOL_GPL(irqtime_account_irq);
-static cputime_t irqtime_tick_accounted(cputime_t maxtime)
+static u64 irqtime_tick_accounted(u64 maxtime)
{
struct irqtime *irqtime = this_cpu_ptr(&cpu_irqtime);
- cputime_t delta;
+ u64 delta;
- delta = nsecs_to_cputime(irqtime->tick_delta);
- delta = min(delta, maxtime);
- irqtime->tick_delta -= cputime_to_nsecs(delta);
+ delta = min(irqtime->tick_delta, maxtime);
+ irqtime->tick_delta -= delta;
return delta;
}
@@ -91,7 +90,7 @@ static cputime_t irqtime_tick_accounted(cputime_t maxtime)
#define sched_clock_irqtime (0)
-static cputime_t irqtime_tick_accounted(cputime_t dummy)
+static u64 irqtime_tick_accounted(u64 dummy)
{
return 0;
}
@@ -234,22 +233,19 @@ void account_idle_time(u64 cputime)
* ticks are not redelivered later. Due to that, this function may on
* occasion account more time than the calling functions think elapsed.
*/
-static __always_inline cputime_t steal_account_process_time(cputime_t maxtime)
+static __always_inline u64 steal_account_process_time(u64 maxtime)
{
#ifdef CONFIG_PARAVIRT
if (static_key_false(¶virt_steal_enabled)) {
- cputime_t steal_cputime;
- u64 steal, rounded;
+ u64 steal;
steal = paravirt_steal_clock(smp_processor_id());
steal -= this_rq()->prev_steal_time;
+ steal = min(steal, maxtime);
+ account_steal_time(steal);
+ this_rq()->prev_steal_time += steal;
- steal_cputime = min(nsecs_to_cputime(steal), maxtime);
- rounded = cputime_to_nsecs(steal_cputime);
- account_steal_time(rounded);
- this_rq()->prev_steal_time += rounded;
-
- return steal_cputime;
+ return steal;
}
#endif
return 0;
@@ -258,9 +254,9 @@ static __always_inline cputime_t steal_account_process_time(cputime_t maxtime)
/*
* Account how much elapsed time was spent in steal, irq, or softirq time.
*/
-static inline cputime_t account_other_time(cputime_t max)
+static inline u64 account_other_time(u64 max)
{
- cputime_t accounted;
+ u64 accounted;
/* Shall be converted to a lockdep-enabled lightweight check */
WARN_ON_ONCE(!irqs_disabled());
@@ -364,9 +360,7 @@ void thread_group_cputime(struct task_struct *tsk, struct task_cputime *times)
static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
struct rq *rq, int ticks)
{
- u64 old_cputime = (__force u64) cputime_one_jiffy * ticks;
- cputime_t other;
- u64 cputime;
+ u64 other, cputime = TICK_NSEC * ticks;
/*
* When returning from idle, many ticks can get accounted at
@@ -376,11 +370,10 @@ static void irqtime_account_process_tick(struct task_struct *p, int user_tick,
* other time can exceed ticks occasionally.
*/
other = account_other_time(ULONG_MAX);
- if (other >= old_cputime)
+ if (other >= cputime)
return;
- old_cputime -= other;
- cputime = cputime_to_nsecs(old_cputime);
+ cputime -= other;
if (this_cpu_ksoftirqd() == p) {
/*
@@ -477,8 +470,7 @@ void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
*/
void account_process_tick(struct task_struct *p, int user_tick)
{
- cputime_t old_cputime, steal;
- u64 cputime;
+ u64 cputime, steal;
struct rq *rq = this_rq();
if (vtime_accounting_cpu_enabled())
@@ -489,14 +481,13 @@ void account_process_tick(struct task_struct *p, int user_tick)
return;
}
- old_cputime = cputime_one_jiffy;
+ cputime = TICK_NSEC;
steal = steal_account_process_time(ULONG_MAX);
- if (steal >= old_cputime)
+ if (steal >= cputime)
return;
- old_cputime -= steal;
- cputime = cputime_to_nsecs(old_cputime);
+ cputime -= steal;
if (user_tick)
account_user_time(p, cputime);
@@ -520,7 +511,7 @@ void account_idle_ticks(unsigned long ticks)
}
cputime = ticks * TICK_NSEC;
- steal = cputime_to_nsecs(steal_account_process_time(ULONG_MAX));
+ steal = steal_account_process_time(ULONG_MAX);
if (steal >= cputime)
return;
@@ -741,6 +732,7 @@ void vtime_account_user(struct task_struct *tsk)
write_seqcount_begin(&tsk->vtime_seqcount);
tsk->vtime_snap_whence = VTIME_SYS;
if (vtime_delta(tsk)) {
+ u64 nsecs;
delta_cpu = get_vtime_delta(tsk);
account_user_time(tsk, cputime_to_nsecs(delta_cpu));
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 27/37] vtime: Return nsecs instead of cputime_t to account
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (25 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 26/37] cputime: Complete nsec conversion of tick based accounting Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 28/37] cputime: Remove jiffies based cputime Frederic Weisbecker
` (9 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Turn the full dynticks cputime clock source to return nsec while keeping
its very internals jiffies based for performance reasons.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
kernel/sched/cputime.c | 31 +++++++++++--------------------
1 file changed, 11 insertions(+), 20 deletions(-)
diff --git a/kernel/sched/cputime.c b/kernel/sched/cputime.c
index f3a56bf..2ecec3a 100644
--- a/kernel/sched/cputime.c
+++ b/kernel/sched/cputime.c
@@ -678,20 +678,20 @@ void thread_group_cputime_adjusted(struct task_struct *p, u64 *ut, u64 *st)
#endif /* !CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
-static cputime_t vtime_delta(struct task_struct *tsk)
+static u64 vtime_delta(struct task_struct *tsk)
{
unsigned long now = READ_ONCE(jiffies);
if (time_before(now, (unsigned long)tsk->vtime_snap))
return 0;
- return jiffies_to_cputime(now - tsk->vtime_snap);
+ return jiffies_to_nsecs(now - tsk->vtime_snap);
}
-static cputime_t get_vtime_delta(struct task_struct *tsk)
+static u64 get_vtime_delta(struct task_struct *tsk)
{
unsigned long now = READ_ONCE(jiffies);
- cputime_t delta, other;
+ u64 delta, other;
/*
* Unlike tick based timing, vtime based timing never has lost
@@ -700,7 +700,7 @@ static cputime_t get_vtime_delta(struct task_struct *tsk)
* elapsed time. Limit account_other_time to prevent rounding
* errors from causing elapsed vtime to go negative.
*/
- delta = jiffies_to_cputime(now - tsk->vtime_snap);
+ delta = jiffies_to_nsecs(now - tsk->vtime_snap);
other = account_other_time(delta);
WARN_ON_ONCE(tsk->vtime_snap_whence == VTIME_INACTIVE);
tsk->vtime_snap = now;
@@ -710,9 +710,7 @@ static cputime_t get_vtime_delta(struct task_struct *tsk)
static void __vtime_account_system(struct task_struct *tsk)
{
- cputime_t delta_cpu = get_vtime_delta(tsk);
-
- account_system_time(tsk, irq_count(), cputime_to_nsecs(delta_cpu));
+ account_system_time(tsk, irq_count(), get_vtime_delta(tsk));
}
void vtime_account_system(struct task_struct *tsk)
@@ -727,15 +725,10 @@ void vtime_account_system(struct task_struct *tsk)
void vtime_account_user(struct task_struct *tsk)
{
- cputime_t delta_cpu;
-
write_seqcount_begin(&tsk->vtime_seqcount);
tsk->vtime_snap_whence = VTIME_SYS;
- if (vtime_delta(tsk)) {
- u64 nsecs;
- delta_cpu = get_vtime_delta(tsk);
- account_user_time(tsk, cputime_to_nsecs(delta_cpu));
- }
+ if (vtime_delta(tsk))
+ account_user_time(tsk, get_vtime_delta(tsk));
write_seqcount_end(&tsk->vtime_seqcount);
}
@@ -776,9 +769,7 @@ EXPORT_SYMBOL_GPL(vtime_guest_exit);
void vtime_account_idle(struct task_struct *tsk)
{
- cputime_t delta_cpu = get_vtime_delta(tsk);
-
- account_idle_time(cputime_to_nsecs(delta_cpu));
+ account_idle_time(get_vtime_delta(tsk));
}
void arch_vtime_task_switch(struct task_struct *prev)
@@ -818,7 +809,7 @@ u64 task_gtime(struct task_struct *t)
gtime = t->gtime;
if (t->vtime_snap_whence == VTIME_SYS && t->flags & PF_VCPU)
- gtime += cputime_to_nsecs(vtime_delta(t));
+ gtime += vtime_delta(t);
} while (read_seqcount_retry(&t->vtime_seqcount, seq));
@@ -851,7 +842,7 @@ void task_cputime(struct task_struct *t, u64 *utime, u64 *stime)
if (t->vtime_snap_whence == VTIME_INACTIVE || is_idle_task(t))
continue;
- delta = cputime_to_nsecs(vtime_delta(t));
+ delta = vtime_delta(t);
/*
* Task runs either in user or kernel space, add pending nohz time to
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 28/37] cputime: Remove jiffies based cputime
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (26 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 27/37] vtime: Return nsecs instead of cputime_t to account Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 29/37] ia64: Move nsecs based cputime headers to the last arch using it Frederic Weisbecker
` (8 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This cputime_t implementation is now unused, we can remove it.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/asm-generic/cputime.h | 4 --
include/asm-generic/cputime_jiffies.h | 75 -----------------------------------
2 files changed, 79 deletions(-)
delete mode 100644 include/asm-generic/cputime_jiffies.h
diff --git a/include/asm-generic/cputime.h b/include/asm-generic/cputime.h
index 5196943..8a624b4 100644
--- a/include/asm-generic/cputime.h
+++ b/include/asm-generic/cputime.h
@@ -4,10 +4,6 @@
#include <linux/time.h>
#include <linux/jiffies.h>
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING
-# include <asm-generic/cputime_jiffies.h>
-#endif
-
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
# include <asm-generic/cputime_nsecs.h>
#endif
diff --git a/include/asm-generic/cputime_jiffies.h b/include/asm-generic/cputime_jiffies.h
deleted file mode 100644
index 6bb8cd4..0000000
--- a/include/asm-generic/cputime_jiffies.h
+++ /dev/null
@@ -1,75 +0,0 @@
-#ifndef _ASM_GENERIC_CPUTIME_JIFFIES_H
-#define _ASM_GENERIC_CPUTIME_JIFFIES_H
-
-typedef unsigned long __nocast cputime_t;
-
-#define cmpxchg_cputime(ptr, old, new) cmpxchg(ptr, old, new)
-
-#define cputime_one_jiffy jiffies_to_cputime(1)
-#define cputime_to_jiffies(__ct) (__force unsigned long)(__ct)
-#define jiffies_to_cputime(__hz) (__force cputime_t)(__hz)
-
-typedef u64 __nocast cputime64_t;
-
-#define cputime64_to_jiffies64(__ct) (__force u64)(__ct)
-#define jiffies64_to_cputime64(__jif) (__force cputime64_t)(__jif)
-
-
-/*
- * Convert nanoseconds <-> cputime
- */
-#define cputime_to_nsecs(__ct) \
- jiffies_to_nsecs(cputime_to_jiffies(__ct))
-#define nsecs_to_cputime64(__nsec) \
- jiffies64_to_cputime64(nsecs_to_jiffies64(__nsec))
-#define nsecs_to_cputime(__nsec) \
- jiffies_to_cputime(nsecs_to_jiffies(__nsec))
-
-
-/*
- * Convert cputime to microseconds and back.
- */
-#define cputime_to_usecs(__ct) \
- jiffies_to_usecs(cputime_to_jiffies(__ct))
-#define usecs_to_cputime(__usec) \
- jiffies_to_cputime(usecs_to_jiffies(__usec))
-#define usecs_to_cputime64(__usec) \
- jiffies64_to_cputime64(nsecs_to_jiffies64((__usec) * 1000))
-
-/*
- * Convert cputime to seconds and back.
- */
-#define cputime_to_secs(jif) (cputime_to_jiffies(jif) / HZ)
-#define secs_to_cputime(sec) jiffies_to_cputime((sec) * HZ)
-
-/*
- * Convert cputime to timespec and back.
- */
-#define timespec_to_cputime(__val) \
- jiffies_to_cputime(timespec_to_jiffies(__val))
-#define cputime_to_timespec(__ct,__val) \
- jiffies_to_timespec(cputime_to_jiffies(__ct),__val)
-
-/*
- * Convert cputime to timeval and back.
- */
-#define timeval_to_cputime(__val) \
- jiffies_to_cputime(timeval_to_jiffies(__val))
-#define cputime_to_timeval(__ct,__val) \
- jiffies_to_timeval(cputime_to_jiffies(__ct),__val)
-
-/*
- * Convert cputime to clock and back.
- */
-#define cputime_to_clock_t(__ct) \
- jiffies_to_clock_t(cputime_to_jiffies(__ct))
-#define clock_t_to_cputime(__x) \
- jiffies_to_cputime(clock_t_to_jiffies(__x))
-
-/*
- * Convert cputime64 to clock.
- */
-#define cputime64_to_clock_t(__ct) \
- jiffies_64_to_clock_t(cputime64_to_jiffies64(__ct))
-
-#endif
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 29/37] ia64: Move nsecs based cputime headers to the last arch using it
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (27 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 28/37] cputime: Remove jiffies based cputime Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 30/37] ia64: Convert vtime to use nsec units directly Frederic Weisbecker
` (7 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Initially, nsec based cputime_t implementation belonged to ia64. It got
exported later for CONFIG_VIRT_CPU_ACCOUNTING_GEN but now it is again
only used by ia64. So let's move it back there.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/include/asm/cputime.h | 108 +++++++++++++++++++++++++++++++-
include/asm-generic/cputime.h | 4 --
include/asm-generic/cputime_nsecs.h | 121 ------------------------------------
3 files changed, 106 insertions(+), 127 deletions(-)
delete mode 100644 include/asm-generic/cputime_nsecs.h
diff --git a/arch/ia64/include/asm/cputime.h b/arch/ia64/include/asm/cputime.h
index e2d3f5b..fee773c 100644
--- a/arch/ia64/include/asm/cputime.h
+++ b/arch/ia64/include/asm/cputime.h
@@ -20,9 +20,113 @@
#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
# include <asm-generic/cputime.h>
+
#else
-# include <asm/processor.h>
-# include <asm-generic/cputime_nsecs.h>
+
+#include <asm/processor.h>
+#include <linux/math64.h>
+
+typedef u64 __nocast cputime_t;
+typedef u64 __nocast cputime64_t;
+
+#define cmpxchg_cputime(ptr, old, new) cmpxchg64(ptr, old, new)
+
+#define cputime_one_jiffy jiffies_to_cputime(1)
+
+#define cputime_div(__ct, divisor) div_u64((__force u64)__ct, divisor)
+#define cputime_div_rem(__ct, divisor, remainder) \
+ div_u64_rem((__force u64)__ct, divisor, remainder);
+
+/*
+ * Convert cputime <-> jiffies (HZ)
+ */
+#define cputime_to_jiffies(__ct) \
+ cputime_div(__ct, NSEC_PER_SEC / HZ)
+#define jiffies_to_cputime(__jif) \
+ (__force cputime_t)((__jif) * (NSEC_PER_SEC / HZ))
+#define cputime64_to_jiffies64(__ct) \
+ cputime_div(__ct, NSEC_PER_SEC / HZ)
+#define jiffies64_to_cputime64(__jif) \
+ (__force cputime64_t)((__jif) * (NSEC_PER_SEC / HZ))
+
+
+/*
+ * Convert cputime <-> nanoseconds
+ */
+#define cputime_to_nsecs(__ct) \
+ (__force u64)(__ct)
+#define nsecs_to_cputime(__nsecs) \
+ (__force cputime_t)(__nsecs)
+#define nsecs_to_cputime64(__nsecs) \
+ (__force cputime64_t)(__nsecs)
+
+
+/*
+ * Convert cputime <-> microseconds
+ */
+#define cputime_to_usecs(__ct) \
+ cputime_div(__ct, NSEC_PER_USEC)
+#define usecs_to_cputime(__usecs) \
+ (__force cputime_t)((__usecs) * NSEC_PER_USEC)
+#define usecs_to_cputime64(__usecs) \
+ (__force cputime64_t)((__usecs) * NSEC_PER_USEC)
+
+/*
+ * Convert cputime <-> seconds
+ */
+#define cputime_to_secs(__ct) \
+ cputime_div(__ct, NSEC_PER_SEC)
+#define secs_to_cputime(__secs) \
+ (__force cputime_t)((__secs) * NSEC_PER_SEC)
+
+/*
+ * Convert cputime <-> timespec (nsec)
+ */
+static inline cputime_t timespec_to_cputime(const struct timespec *val)
+{
+ u64 ret = (u64)val->tv_sec * NSEC_PER_SEC + val->tv_nsec;
+ return (__force cputime_t) ret;
+}
+static inline void cputime_to_timespec(const cputime_t ct, struct timespec *val)
+{
+ u32 rem;
+
+ val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
+ val->tv_nsec = rem;
+}
+
+/*
+ * Convert cputime <-> timeval (msec)
+ */
+static inline cputime_t timeval_to_cputime(const struct timeval *val)
+{
+ u64 ret = (u64)val->tv_sec * NSEC_PER_SEC +
+ val->tv_usec * NSEC_PER_USEC;
+ return (__force cputime_t) ret;
+}
+static inline void cputime_to_timeval(const cputime_t ct, struct timeval *val)
+{
+ u32 rem;
+
+ val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
+ val->tv_usec = rem / NSEC_PER_USEC;
+}
+
+/*
+ * Convert cputime <-> clock (USER_HZ)
+ */
+#define cputime_to_clock_t(__ct) \
+ cputime_div(__ct, (NSEC_PER_SEC / USER_HZ))
+#define clock_t_to_cputime(__x) \
+ (__force cputime_t)((__x) * (NSEC_PER_SEC / USER_HZ))
+
+/*
+ * Convert cputime64 to clock.
+ */
+#define cputime64_to_clock_t(__ct) \
+ cputime_to_clock_t((__force cputime_t)__ct)
+
+
extern void arch_vtime_task_switch(struct task_struct *tsk);
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
diff --git a/include/asm-generic/cputime.h b/include/asm-generic/cputime.h
index 8a624b4..358e547 100644
--- a/include/asm-generic/cputime.h
+++ b/include/asm-generic/cputime.h
@@ -4,8 +4,4 @@
#include <linux/time.h>
#include <linux/jiffies.h>
-#ifdef CONFIG_VIRT_CPU_ACCOUNTING_GEN
-# include <asm-generic/cputime_nsecs.h>
-#endif
-
#endif
diff --git a/include/asm-generic/cputime_nsecs.h b/include/asm-generic/cputime_nsecs.h
deleted file mode 100644
index 4e3b18e..0000000
--- a/include/asm-generic/cputime_nsecs.h
+++ /dev/null
@@ -1,121 +0,0 @@
-/*
- * Definitions for measuring cputime in nsecs resolution.
- *
- * Based on <arch/ia64/include/asm/cputime.h>
- *
- * Copyright (C) 2007 FUJITSU LIMITED
- * Copyright (C) 2007 Hidetoshi Seto <seto.hidetoshi@jp.fujitsu.com>
- *
- * This program is free software; you can redistribute it and/or
- * modify it under the terms of the GNU General Public License
- * as published by the Free Software Foundation; either version
- * 2 of the License, or (at your option) any later version.
- *
- */
-
-#ifndef _ASM_GENERIC_CPUTIME_NSECS_H
-#define _ASM_GENERIC_CPUTIME_NSECS_H
-
-#include <linux/math64.h>
-
-typedef u64 __nocast cputime_t;
-typedef u64 __nocast cputime64_t;
-
-#define cmpxchg_cputime(ptr, old, new) cmpxchg64(ptr, old, new)
-
-#define cputime_one_jiffy jiffies_to_cputime(1)
-
-#define cputime_div(__ct, divisor) div_u64((__force u64)__ct, divisor)
-#define cputime_div_rem(__ct, divisor, remainder) \
- div_u64_rem((__force u64)__ct, divisor, remainder);
-
-/*
- * Convert cputime <-> jiffies (HZ)
- */
-#define cputime_to_jiffies(__ct) \
- cputime_div(__ct, NSEC_PER_SEC / HZ)
-#define jiffies_to_cputime(__jif) \
- (__force cputime_t)((__jif) * (NSEC_PER_SEC / HZ))
-#define cputime64_to_jiffies64(__ct) \
- cputime_div(__ct, NSEC_PER_SEC / HZ)
-#define jiffies64_to_cputime64(__jif) \
- (__force cputime64_t)((__jif) * (NSEC_PER_SEC / HZ))
-
-
-/*
- * Convert cputime <-> nanoseconds
- */
-#define cputime_to_nsecs(__ct) \
- (__force u64)(__ct)
-#define nsecs_to_cputime(__nsecs) \
- (__force cputime_t)(__nsecs)
-#define nsecs_to_cputime64(__nsecs) \
- (__force cputime64_t)(__nsecs)
-
-
-/*
- * Convert cputime <-> microseconds
- */
-#define cputime_to_usecs(__ct) \
- cputime_div(__ct, NSEC_PER_USEC)
-#define usecs_to_cputime(__usecs) \
- (__force cputime_t)((__usecs) * NSEC_PER_USEC)
-#define usecs_to_cputime64(__usecs) \
- (__force cputime64_t)((__usecs) * NSEC_PER_USEC)
-
-/*
- * Convert cputime <-> seconds
- */
-#define cputime_to_secs(__ct) \
- cputime_div(__ct, NSEC_PER_SEC)
-#define secs_to_cputime(__secs) \
- (__force cputime_t)((__secs) * NSEC_PER_SEC)
-
-/*
- * Convert cputime <-> timespec (nsec)
- */
-static inline cputime_t timespec_to_cputime(const struct timespec *val)
-{
- u64 ret = (u64)val->tv_sec * NSEC_PER_SEC + val->tv_nsec;
- return (__force cputime_t) ret;
-}
-static inline void cputime_to_timespec(const cputime_t ct, struct timespec *val)
-{
- u32 rem;
-
- val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
- val->tv_nsec = rem;
-}
-
-/*
- * Convert cputime <-> timeval (msec)
- */
-static inline cputime_t timeval_to_cputime(const struct timeval *val)
-{
- u64 ret = (u64)val->tv_sec * NSEC_PER_SEC +
- val->tv_usec * NSEC_PER_USEC;
- return (__force cputime_t) ret;
-}
-static inline void cputime_to_timeval(const cputime_t ct, struct timeval *val)
-{
- u32 rem;
-
- val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
- val->tv_usec = rem / NSEC_PER_USEC;
-}
-
-/*
- * Convert cputime <-> clock (USER_HZ)
- */
-#define cputime_to_clock_t(__ct) \
- cputime_div(__ct, (NSEC_PER_SEC / USER_HZ))
-#define clock_t_to_cputime(__x) \
- (__force cputime_t)((__x) * (NSEC_PER_SEC / USER_HZ))
-
-/*
- * Convert cputime64 to clock.
- */
-#define cputime64_to_clock_t(__ct) \
- cputime_to_clock_t((__force cputime_t)__ct)
-
-#endif
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 30/37] ia64: Convert vtime to use nsec units directly
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (28 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 29/37] ia64: Move nsecs based cputime headers to the last arch using it Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 31/37] ia64: Remove unused cputime definitions Frederic Weisbecker
` (6 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
There is no need anymore for this cputime_t midlayer. Let's use nsec
units directly.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/kernel/head.S | 4 ++--
arch/ia64/kernel/time.c | 14 +++++++-------
2 files changed, 9 insertions(+), 9 deletions(-)
diff --git a/arch/ia64/kernel/head.S b/arch/ia64/kernel/head.S
index c9b5e94..3204fdd 100644
--- a/arch/ia64/kernel/head.S
+++ b/arch/ia64/kernel/head.S
@@ -1031,7 +1031,7 @@ GLOBAL_ENTRY(ia64_native_sched_clock)
END(ia64_native_sched_clock)
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-GLOBAL_ENTRY(cycle_to_cputime)
+GLOBAL_ENTRY(cycle_to_nsec)
alloc r16=ar.pfs,1,0,0,0
addl r8=THIS_CPU(ia64_cpu_info) + IA64_CPUINFO_NSEC_PER_CYC_OFFSET,r0
;;
@@ -1047,7 +1047,7 @@ GLOBAL_ENTRY(cycle_to_cputime)
;;
shrp r8=r9,r8,IA64_NSEC_PER_CYC_SHIFT
br.ret.sptk.many rp
-END(cycle_to_cputime)
+END(cycle_to_nsec)
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
#ifdef CONFIG_IA64_BRL_EMU
diff --git a/arch/ia64/kernel/time.c b/arch/ia64/kernel/time.c
index f15bca4..faa1168 100644
--- a/arch/ia64/kernel/time.c
+++ b/arch/ia64/kernel/time.c
@@ -60,7 +60,7 @@ static struct clocksource *itc_clocksource;
#include <linux/kernel_stat.h>
-extern cputime_t cycle_to_cputime(u64 cyc);
+extern u64 cycle_to_nsec(u64 cyc);
void vtime_flush(struct task_struct *tsk)
{
@@ -68,26 +68,26 @@ void vtime_flush(struct task_struct *tsk)
u64 delta;
if (ti->utime)
- account_user_time(tsk, cputime_to_nsecs(cycle_to_cputime(ti->utime)));
+ account_user_time(tsk, cycle_to_nsec(ti->utime));
if (ti->gtime)
- account_guest_time(tsk, cputime_to_nsecs(cycle_to_cputime(ti->gtime)));
+ account_guest_time(tsk, cycle_to_nsec(ti->gtime));
if (ti->idle_time)
- account_idle_time(cputime_to_nsecs(cycle_to_cputime(ti->idle_time)));
+ account_idle_time(cycle_to_nsec(ti->idle_time));
if (ti->stime) {
- delta = cputime_to_nsecs(cycle_to_cputime(ti->stime));
+ delta = cycle_to_nsec(ti->stime);
account_system_index_time(tsk, delta, CPUTIME_SYSTEM);
}
if (ti->hardirq_time) {
- delta = cputime_to_nsecs(cycle_to_cputime(ti->hardirq_time));
+ delta = cycle_to_nsec(ti->hardirq_time);
account_system_index_time(tsk, delta, CPUTIME_IRQ);
}
if (ti->softirq_time) {
- delta = cputime_to_nsecs(cycle_to_cputime(ti->softirq_time));
+ delta = cycle_to_nsec(ti->softirq_time));
account_system_index_time(tsk, delta, CPUTIME_SOFTIRQ);
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 31/37] ia64: Remove unused cputime definitions
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (29 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 30/37] ia64: Convert vtime to use nsec units directly Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 32/37] s390: Make arch_cpu_idle_time() to return nsecs Frederic Weisbecker
` (5 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This nsec based cputime_t implementation isn't used anymore. We can
remove it.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/ia64/include/asm/cputime.h | 106 ----------------------------------------
1 file changed, 106 deletions(-)
diff --git a/arch/ia64/include/asm/cputime.h b/arch/ia64/include/asm/cputime.h
index fee773c..44bcffc 100644
--- a/arch/ia64/include/asm/cputime.h
+++ b/arch/ia64/include/asm/cputime.h
@@ -20,113 +20,7 @@
#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
# include <asm-generic/cputime.h>
-
#else
-
-#include <asm/processor.h>
-#include <linux/math64.h>
-
-typedef u64 __nocast cputime_t;
-typedef u64 __nocast cputime64_t;
-
-#define cmpxchg_cputime(ptr, old, new) cmpxchg64(ptr, old, new)
-
-#define cputime_one_jiffy jiffies_to_cputime(1)
-
-#define cputime_div(__ct, divisor) div_u64((__force u64)__ct, divisor)
-#define cputime_div_rem(__ct, divisor, remainder) \
- div_u64_rem((__force u64)__ct, divisor, remainder);
-
-/*
- * Convert cputime <-> jiffies (HZ)
- */
-#define cputime_to_jiffies(__ct) \
- cputime_div(__ct, NSEC_PER_SEC / HZ)
-#define jiffies_to_cputime(__jif) \
- (__force cputime_t)((__jif) * (NSEC_PER_SEC / HZ))
-#define cputime64_to_jiffies64(__ct) \
- cputime_div(__ct, NSEC_PER_SEC / HZ)
-#define jiffies64_to_cputime64(__jif) \
- (__force cputime64_t)((__jif) * (NSEC_PER_SEC / HZ))
-
-
-/*
- * Convert cputime <-> nanoseconds
- */
-#define cputime_to_nsecs(__ct) \
- (__force u64)(__ct)
-#define nsecs_to_cputime(__nsecs) \
- (__force cputime_t)(__nsecs)
-#define nsecs_to_cputime64(__nsecs) \
- (__force cputime64_t)(__nsecs)
-
-
-/*
- * Convert cputime <-> microseconds
- */
-#define cputime_to_usecs(__ct) \
- cputime_div(__ct, NSEC_PER_USEC)
-#define usecs_to_cputime(__usecs) \
- (__force cputime_t)((__usecs) * NSEC_PER_USEC)
-#define usecs_to_cputime64(__usecs) \
- (__force cputime64_t)((__usecs) * NSEC_PER_USEC)
-
-/*
- * Convert cputime <-> seconds
- */
-#define cputime_to_secs(__ct) \
- cputime_div(__ct, NSEC_PER_SEC)
-#define secs_to_cputime(__secs) \
- (__force cputime_t)((__secs) * NSEC_PER_SEC)
-
-/*
- * Convert cputime <-> timespec (nsec)
- */
-static inline cputime_t timespec_to_cputime(const struct timespec *val)
-{
- u64 ret = (u64)val->tv_sec * NSEC_PER_SEC + val->tv_nsec;
- return (__force cputime_t) ret;
-}
-static inline void cputime_to_timespec(const cputime_t ct, struct timespec *val)
-{
- u32 rem;
-
- val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
- val->tv_nsec = rem;
-}
-
-/*
- * Convert cputime <-> timeval (msec)
- */
-static inline cputime_t timeval_to_cputime(const struct timeval *val)
-{
- u64 ret = (u64)val->tv_sec * NSEC_PER_SEC +
- val->tv_usec * NSEC_PER_USEC;
- return (__force cputime_t) ret;
-}
-static inline void cputime_to_timeval(const cputime_t ct, struct timeval *val)
-{
- u32 rem;
-
- val->tv_sec = cputime_div_rem(ct, NSEC_PER_SEC, &rem);
- val->tv_usec = rem / NSEC_PER_USEC;
-}
-
-/*
- * Convert cputime <-> clock (USER_HZ)
- */
-#define cputime_to_clock_t(__ct) \
- cputime_div(__ct, (NSEC_PER_SEC / USER_HZ))
-#define clock_t_to_cputime(__x) \
- (__force cputime_t)((__x) * (NSEC_PER_SEC / USER_HZ))
-
-/*
- * Convert cputime64 to clock.
- */
-#define cputime64_to_clock_t(__ct) \
- cputime_to_clock_t((__force cputime_t)__ct)
-
-
extern void arch_vtime_task_switch(struct task_struct *tsk);
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 32/37] s390: Make arch_cpu_idle_time() to return nsecs
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (30 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 31/37] ia64: Remove unused cputime definitions Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 33/37] powerpc: Remove unused cputime definitions Frederic Weisbecker
` (4 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
This way we don't need to deal with cputime_t details from the core code.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/s390/include/asm/cputime.h | 2 +-
arch/s390/kernel/idle.c | 5 +++--
fs/proc/stat.c | 4 ++--
3 files changed, 6 insertions(+), 5 deletions(-)
diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h
index 221b454..9a94481 100644
--- a/arch/s390/include/asm/cputime.h
+++ b/arch/s390/include/asm/cputime.h
@@ -141,7 +141,7 @@ static inline clock_t cputime64_to_clock_t(cputime64_t cputime)
return clock;
}
-cputime64_t arch_cpu_idle_time(int cpu);
+u64 arch_cpu_idle_time(int cpu);
#define arch_idle_time(cpu) arch_cpu_idle_time(cpu)
diff --git a/arch/s390/kernel/idle.c b/arch/s390/kernel/idle.c
index 5c0e08e7..d3bf69e 100644
--- a/arch/s390/kernel/idle.c
+++ b/arch/s390/kernel/idle.c
@@ -84,7 +84,7 @@ static ssize_t show_idle_time(struct device *dev,
}
DEVICE_ATTR(idle_time_us, 0444, show_idle_time, NULL);
-cputime64_t arch_cpu_idle_time(int cpu)
+u64 arch_cpu_idle_time(int cpu)
{
struct s390_idle_data *idle = &per_cpu(s390_idle, cpu);
unsigned long long now, idle_enter, idle_exit;
@@ -96,7 +96,8 @@ cputime64_t arch_cpu_idle_time(int cpu)
idle_enter = ACCESS_ONCE(idle->clock_idle_enter);
idle_exit = ACCESS_ONCE(idle->clock_idle_exit);
} while (read_seqcount_retry(&idle->seqcount, seq));
- return idle_enter ? ((idle_exit ?: now) - idle_enter) : 0;
+
+ return cputime_to_nsecs(idle_enter ? ((idle_exit ?: now) - idle_enter) : 0);
}
void arch_cpu_idle_enter(void)
diff --git a/fs/proc/stat.c b/fs/proc/stat.c
index 44475a4..e47c3e8 100644
--- a/fs/proc/stat.c
+++ b/fs/proc/stat.c
@@ -27,7 +27,7 @@ static u64 get_idle_time(int cpu)
idle = kcpustat_cpu(cpu).cpustat[CPUTIME_IDLE];
if (cpu_online(cpu) && !nr_iowait_cpu(cpu))
- idle += cputime_to_nsecs(arch_idle_time(cpu));
+ idle += arch_idle_time(cpu);
return idle;
}
@@ -37,7 +37,7 @@ static u64 get_iowait_time(int cpu)
iowait = kcpustat_cpu(cpu).cpustat[CPUTIME_IOWAIT];
if (cpu_online(cpu) && nr_iowait_cpu(cpu))
- iowait += cputime_to_nsecs(arch_idle_time(cpu));
+ iowait += arch_idle_time(cpu);
return iowait;
}
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 33/37] powerpc: Remove unused cputime definitions
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (31 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 32/37] s390: Make arch_cpu_idle_time() to return nsecs Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 34/37] s390: " Frederic Weisbecker
` (3 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Since the core doesn't deal with cputime_t anymore, most of these APIs
have been left unused. Lets remove these.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/powerpc/include/asm/cputime.h | 173 -------------------------------------
arch/powerpc/kernel/time.c | 20 +----
2 files changed, 2 insertions(+), 191 deletions(-)
diff --git a/arch/powerpc/include/asm/cputime.h b/arch/powerpc/include/asm/cputime.h
index aa2e6a3..6ec0ba6 100644
--- a/arch/powerpc/include/asm/cputime.h
+++ b/arch/powerpc/include/asm/cputime.h
@@ -18,9 +18,6 @@
#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
#include <asm-generic/cputime.h>
-#ifdef __KERNEL__
-static inline void setup_cputime_one_jiffy(void) { }
-#endif
#else
#include <linux/types.h>
@@ -36,65 +33,6 @@ typedef u64 __nocast cputime64_t;
#define cmpxchg_cputime(ptr, old, new) cmpxchg(ptr, old, new)
#ifdef __KERNEL__
-
-/*
- * One jiffy in timebase units computed during initialization
- */
-extern cputime_t cputime_one_jiffy;
-
-/*
- * Convert cputime <-> jiffies
- */
-extern u64 __cputime_jiffies_factor;
-
-static inline unsigned long cputime_to_jiffies(const cputime_t ct)
-{
- return mulhdu((__force u64) ct, __cputime_jiffies_factor);
-}
-
-static inline cputime_t jiffies_to_cputime(const unsigned long jif)
-{
- u64 ct;
- unsigned long sec;
-
- /* have to be a little careful about overflow */
- ct = jif % HZ;
- sec = jif / HZ;
- if (ct) {
- ct *= tb_ticks_per_sec;
- do_div(ct, HZ);
- }
- if (sec)
- ct += (cputime_t) sec * tb_ticks_per_sec;
- return (__force cputime_t) ct;
-}
-
-static inline void setup_cputime_one_jiffy(void)
-{
- cputime_one_jiffy = jiffies_to_cputime(1);
-}
-
-static inline cputime64_t jiffies64_to_cputime64(const u64 jif)
-{
- u64 ct;
- u64 sec = jif;
-
- /* have to be a little careful about overflow */
- ct = do_div(sec, HZ);
- if (ct) {
- ct *= tb_ticks_per_sec;
- do_div(ct, HZ);
- }
- if (sec)
- ct += (u64) sec * tb_ticks_per_sec;
- return (__force cputime64_t) ct;
-}
-
-static inline u64 cputime64_to_jiffies64(const cputime_t ct)
-{
- return mulhdu((__force u64) ct, __cputime_jiffies_factor);
-}
-
/*
* Convert cputime <-> microseconds
*/
@@ -105,117 +43,6 @@ static inline unsigned long cputime_to_usecs(const cputime_t ct)
return mulhdu((__force u64) ct, __cputime_usec_factor);
}
-static inline cputime_t usecs_to_cputime(const unsigned long us)
-{
- u64 ct;
- unsigned long sec;
-
- /* have to be a little careful about overflow */
- ct = us % 1000000;
- sec = us / 1000000;
- if (ct) {
- ct *= tb_ticks_per_sec;
- do_div(ct, 1000000);
- }
- if (sec)
- ct += (cputime_t) sec * tb_ticks_per_sec;
- return (__force cputime_t) ct;
-}
-
-#define usecs_to_cputime64(us) usecs_to_cputime(us)
-
-/*
- * Convert cputime <-> seconds
- */
-extern u64 __cputime_sec_factor;
-
-static inline unsigned long cputime_to_secs(const cputime_t ct)
-{
- return mulhdu((__force u64) ct, __cputime_sec_factor);
-}
-
-static inline cputime_t secs_to_cputime(const unsigned long sec)
-{
- return (__force cputime_t)((u64) sec * tb_ticks_per_sec);
-}
-
-/*
- * Convert cputime <-> timespec
- */
-static inline void cputime_to_timespec(const cputime_t ct, struct timespec *p)
-{
- u64 x = (__force u64) ct;
- unsigned int frac;
-
- frac = do_div(x, tb_ticks_per_sec);
- p->tv_sec = x;
- x = (u64) frac * 1000000000;
- do_div(x, tb_ticks_per_sec);
- p->tv_nsec = x;
-}
-
-static inline cputime_t timespec_to_cputime(const struct timespec *p)
-{
- u64 ct;
-
- ct = (u64) p->tv_nsec * tb_ticks_per_sec;
- do_div(ct, 1000000000);
- return (__force cputime_t)(ct + (u64) p->tv_sec * tb_ticks_per_sec);
-}
-
-/*
- * Convert cputime <-> timeval
- */
-static inline void cputime_to_timeval(const cputime_t ct, struct timeval *p)
-{
- u64 x = (__force u64) ct;
- unsigned int frac;
-
- frac = do_div(x, tb_ticks_per_sec);
- p->tv_sec = x;
- x = (u64) frac * 1000000;
- do_div(x, tb_ticks_per_sec);
- p->tv_usec = x;
-}
-
-static inline cputime_t timeval_to_cputime(const struct timeval *p)
-{
- u64 ct;
-
- ct = (u64) p->tv_usec * tb_ticks_per_sec;
- do_div(ct, 1000000);
- return (__force cputime_t)(ct + (u64) p->tv_sec * tb_ticks_per_sec);
-}
-
-/*
- * Convert cputime <-> clock_t (units of 1/USER_HZ seconds)
- */
-extern u64 __cputime_clockt_factor;
-
-static inline unsigned long cputime_to_clock_t(const cputime_t ct)
-{
- return mulhdu((__force u64) ct, __cputime_clockt_factor);
-}
-
-static inline cputime_t clock_t_to_cputime(const unsigned long clk)
-{
- u64 ct;
- unsigned long sec;
-
- /* have to be a little careful about overflow */
- ct = clk % USER_HZ;
- sec = clk / USER_HZ;
- if (ct) {
- ct *= tb_ticks_per_sec;
- do_div(ct, USER_HZ);
- }
- if (sec)
- ct += (u64) sec * tb_ticks_per_sec;
- return (__force cputime_t) ct;
-}
-
-#define cputime64_to_clock_t(ct) cputime_to_clock_t((cputime_t)(ct))
-
/*
* PPC64 uses PACA which is task independent for storing accounting data while
* PPC32 uses struct thread_info, therefore at task switch the accounting data
diff --git a/arch/powerpc/kernel/time.c b/arch/powerpc/kernel/time.c
index 01f53bf..14e4855 100644
--- a/arch/powerpc/kernel/time.c
+++ b/arch/powerpc/kernel/time.c
@@ -152,20 +152,11 @@ EXPORT_SYMBOL_GPL(ppc_tb_freq);
#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
/*
- * Factors for converting from cputime_t (timebase ticks) to
- * jiffies, microseconds, seconds, and clock_t (1/USER_HZ seconds).
- * These are all stored as 0.64 fixed-point binary fractions.
+ * Factor for converting from cputime_t (timebase ticks) to
+ * microseconds. This is stored as 0.64 fixed-point binary fraction.
*/
-u64 __cputime_jiffies_factor;
-EXPORT_SYMBOL(__cputime_jiffies_factor);
u64 __cputime_usec_factor;
EXPORT_SYMBOL(__cputime_usec_factor);
-u64 __cputime_sec_factor;
-EXPORT_SYMBOL(__cputime_sec_factor);
-u64 __cputime_clockt_factor;
-EXPORT_SYMBOL(__cputime_clockt_factor);
-
-cputime_t cputime_one_jiffy;
#ifdef CONFIG_PPC_SPLPAR
void (*dtl_consumer)(struct dtl_entry *, u64);
@@ -181,14 +172,8 @@ static void calc_cputime_factors(void)
{
struct div_result res;
- div128_by_32(HZ, 0, tb_ticks_per_sec, &res);
- __cputime_jiffies_factor = res.result_low;
div128_by_32(1000000, 0, tb_ticks_per_sec, &res);
__cputime_usec_factor = res.result_low;
- div128_by_32(1, 0, tb_ticks_per_sec, &res);
- __cputime_sec_factor = res.result_low;
- div128_by_32(USER_HZ, 0, tb_ticks_per_sec, &res);
- __cputime_clockt_factor = res.result_low;
}
/*
@@ -1053,7 +1038,6 @@ void __init time_init(void)
tb_ticks_per_sec = ppc_tb_freq;
tb_ticks_per_usec = ppc_tb_freq / 1000000;
calc_cputime_factors();
- setup_cputime_one_jiffy();
/*
* Compute scale factor for sched_clock.
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 34/37] s390: Remove unused cputime definitions
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (32 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 33/37] powerpc: Remove unused cputime definitions Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 35/37] cputime: Remove unused nsec_to_cputime Frederic Weisbecker
` (2 subsequent siblings)
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
Since the core doesn't deal with cputime_t anymore, most of these APIs
have been left unused. Lets remove these.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/s390/include/asm/cputime.h | 107 ----------------------------------------
1 file changed, 107 deletions(-)
diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h
index 9a94481..d1c407d 100644
--- a/arch/s390/include/asm/cputime.h
+++ b/arch/s390/include/asm/cputime.h
@@ -25,33 +25,6 @@ static inline unsigned long __div(unsigned long long n, unsigned long base)
return n / base;
}
-#define cputime_one_jiffy jiffies_to_cputime(1)
-
-/*
- * Convert cputime to jiffies and back.
- */
-static inline unsigned long cputime_to_jiffies(const cputime_t cputime)
-{
- return __div((__force unsigned long long) cputime, CPUTIME_PER_SEC / HZ);
-}
-
-static inline cputime_t jiffies_to_cputime(const unsigned int jif)
-{
- return (__force cputime_t)(jif * (CPUTIME_PER_SEC / HZ));
-}
-
-static inline u64 cputime64_to_jiffies64(cputime64_t cputime)
-{
- unsigned long long jif = (__force unsigned long long) cputime;
- do_div(jif, CPUTIME_PER_SEC / HZ);
- return jif;
-}
-
-static inline cputime64_t jiffies64_to_cputime64(const u64 jif)
-{
- return (__force cputime64_t)(jif * (CPUTIME_PER_SEC / HZ));
-}
-
/*
* Convert cputime to microseconds and back.
*/
@@ -60,86 +33,6 @@ static inline unsigned int cputime_to_usecs(const cputime_t cputime)
return (__force unsigned long long) cputime >> 12;
}
-static inline cputime_t usecs_to_cputime(const unsigned int m)
-{
- return (__force cputime_t)(m * CPUTIME_PER_USEC);
-}
-
-#define usecs_to_cputime64(m) usecs_to_cputime(m)
-
-/*
- * Convert cputime to milliseconds and back.
- */
-static inline unsigned int cputime_to_secs(const cputime_t cputime)
-{
- return __div((__force unsigned long long) cputime, CPUTIME_PER_SEC / 2) >> 1;
-}
-
-static inline cputime_t secs_to_cputime(const unsigned int s)
-{
- return (__force cputime_t)(s * CPUTIME_PER_SEC);
-}
-
-/*
- * Convert cputime to timespec and back.
- */
-static inline cputime_t timespec_to_cputime(const struct timespec *value)
-{
- unsigned long long ret = value->tv_sec * CPUTIME_PER_SEC;
- return (__force cputime_t)(ret + __div(value->tv_nsec * CPUTIME_PER_USEC, NSEC_PER_USEC));
-}
-
-static inline void cputime_to_timespec(const cputime_t cputime,
- struct timespec *value)
-{
- unsigned long long __cputime = (__force unsigned long long) cputime;
- value->tv_nsec = (__cputime % CPUTIME_PER_SEC) * NSEC_PER_USEC / CPUTIME_PER_USEC;
- value->tv_sec = __cputime / CPUTIME_PER_SEC;
-}
-
-/*
- * Convert cputime to timeval and back.
- * Since cputime and timeval have the same resolution (microseconds)
- * this is easy.
- */
-static inline cputime_t timeval_to_cputime(const struct timeval *value)
-{
- unsigned long long ret = value->tv_sec * CPUTIME_PER_SEC;
- return (__force cputime_t)(ret + value->tv_usec * CPUTIME_PER_USEC);
-}
-
-static inline void cputime_to_timeval(const cputime_t cputime,
- struct timeval *value)
-{
- unsigned long long __cputime = (__force unsigned long long) cputime;
- value->tv_usec = (__cputime % CPUTIME_PER_SEC) / CPUTIME_PER_USEC;
- value->tv_sec = __cputime / CPUTIME_PER_SEC;
-}
-
-/*
- * Convert cputime to clock and back.
- */
-static inline clock_t cputime_to_clock_t(cputime_t cputime)
-{
- unsigned long long clock = (__force unsigned long long) cputime;
- do_div(clock, CPUTIME_PER_SEC / USER_HZ);
- return clock;
-}
-
-static inline cputime_t clock_t_to_cputime(unsigned long x)
-{
- return (__force cputime_t)(x * (CPUTIME_PER_SEC / USER_HZ));
-}
-
-/*
- * Convert cputime64 to clock.
- */
-static inline clock_t cputime64_to_clock_t(cputime64_t cputime)
-{
- unsigned long long clock = (__force unsigned long long) cputime;
- do_div(clock, CPUTIME_PER_SEC / USER_HZ);
- return clock;
-}
u64 arch_cpu_idle_time(int cpu);
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 35/37] cputime: Remove unused nsec_to_cputime
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (33 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 34/37] s390: " Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 36/37] cputime: Remove asm generic headers Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 37/37] s390: Prevent from cputime leaks Frederic Weisbecker
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
It's unused now.
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
include/linux/cputime.h | 5 -----
1 file changed, 5 deletions(-)
diff --git a/include/linux/cputime.h b/include/linux/cputime.h
index f2eb2ee..a257d66 100644
--- a/include/linux/cputime.h
+++ b/include/linux/cputime.h
@@ -8,9 +8,4 @@
(cputime_to_usecs(__ct) * NSEC_PER_USEC)
#endif
-#ifndef nsecs_to_cputime
-# define nsecs_to_cputime(__nsecs) \
- usecs_to_cputime((__nsecs) / NSEC_PER_USEC)
-#endif
-
#endif /* __LINUX_CPUTIME_H */
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 36/37] cputime: Remove asm generic headers
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (34 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 35/37] cputime: Remove unused nsec_to_cputime Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-22 18:20 ` [PATCH 37/37] s390: Prevent from cputime leaks Frederic Weisbecker
36 siblings, 0 replies; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
cputime_t is now only used by two archs:
* powerpc (when CONFIG_VIRT_CPU_ACCOUNTING_NATIVE=y)
* s390
And since the core doesn't use it anymore, we don't need any arch support
from the others. So we can remove their stub implementations.
A final cleanup would be to provide an efficient pure arch
implementation of cputime_to_nsec() for s390 and powerpc and finally
remove include/linux/cputime.h .
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/alpha/include/asm/Kbuild | 1 -
arch/arc/include/asm/Kbuild | 1 -
arch/arm/include/asm/Kbuild | 1 -
arch/arm64/include/asm/Kbuild | 1 -
arch/avr32/include/asm/Kbuild | 1 -
arch/blackfin/include/asm/Kbuild | 1 -
arch/c6x/include/asm/Kbuild | 1 -
arch/cris/include/asm/Kbuild | 1 -
arch/frv/include/asm/Kbuild | 1 -
arch/h8300/include/asm/Kbuild | 1 -
arch/hexagon/include/asm/Kbuild | 1 -
arch/ia64/include/asm/cputime.h | 4 +---
arch/m32r/include/asm/Kbuild | 1 -
arch/m68k/include/asm/Kbuild | 1 -
arch/metag/include/asm/Kbuild | 1 -
arch/microblaze/include/asm/Kbuild | 1 -
arch/mips/include/asm/Kbuild | 1 -
arch/mn10300/include/asm/Kbuild | 1 -
arch/nios2/include/asm/Kbuild | 1 -
arch/openrisc/include/asm/Kbuild | 1 -
arch/parisc/include/asm/Kbuild | 1 -
arch/powerpc/include/asm/cputime.h | 4 +---
arch/score/include/asm/Kbuild | 1 -
arch/sh/include/asm/Kbuild | 1 -
arch/sparc/include/asm/Kbuild | 1 -
arch/tile/include/asm/Kbuild | 1 -
arch/um/include/asm/Kbuild | 1 -
arch/unicore32/include/asm/Kbuild | 1 -
arch/x86/include/asm/Kbuild | 1 -
arch/xtensa/include/asm/Kbuild | 1 -
include/asm-generic/cputime.h | 7 -------
include/linux/cputime.h | 2 ++
32 files changed, 4 insertions(+), 41 deletions(-)
delete mode 100644 include/asm-generic/cputime.h
diff --git a/arch/alpha/include/asm/Kbuild b/arch/alpha/include/asm/Kbuild
index bf8475c..baa152b 100644
--- a/arch/alpha/include/asm/Kbuild
+++ b/arch/alpha/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += exec.h
generic-y += export.h
generic-y += irq_work.h
diff --git a/arch/arc/include/asm/Kbuild b/arch/arc/include/asm/Kbuild
index c332604..63a0401 100644
--- a/arch/arc/include/asm/Kbuild
+++ b/arch/arc/include/asm/Kbuild
@@ -2,7 +2,6 @@ generic-y += auxvec.h
generic-y += bitsperlong.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += device.h
generic-y += div64.h
generic-y += emergency-restart.h
diff --git a/arch/arm/include/asm/Kbuild b/arch/arm/include/asm/Kbuild
index efb2175..b14e8c7 100644
--- a/arch/arm/include/asm/Kbuild
+++ b/arch/arm/include/asm/Kbuild
@@ -2,7 +2,6 @@
generic-y += bitsperlong.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += early_ioremap.h
generic-y += emergency-restart.h
diff --git a/arch/arm64/include/asm/Kbuild b/arch/arm64/include/asm/Kbuild
index 8365a84..a12f1af 100644
--- a/arch/arm64/include/asm/Kbuild
+++ b/arch/arm64/include/asm/Kbuild
@@ -1,6 +1,5 @@
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += delay.h
generic-y += div64.h
generic-y += dma.h
diff --git a/arch/avr32/include/asm/Kbuild b/arch/avr32/include/asm/Kbuild
index 241b9b9..3d7ef2c 100644
--- a/arch/avr32/include/asm/Kbuild
+++ b/arch/avr32/include/asm/Kbuild
@@ -1,6 +1,5 @@
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += delay.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/blackfin/include/asm/Kbuild b/arch/blackfin/include/asm/Kbuild
index 2fb67b5..d6fa60b 100644
--- a/arch/blackfin/include/asm/Kbuild
+++ b/arch/blackfin/include/asm/Kbuild
@@ -2,7 +2,6 @@
generic-y += auxvec.h
generic-y += bitsperlong.h
generic-y += bugs.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/c6x/include/asm/Kbuild b/arch/c6x/include/asm/Kbuild
index 64465e7..4e9f574 100644
--- a/arch/c6x/include/asm/Kbuild
+++ b/arch/c6x/include/asm/Kbuild
@@ -5,7 +5,6 @@ generic-y += barrier.h
generic-y += bitsperlong.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/cris/include/asm/Kbuild b/arch/cris/include/asm/Kbuild
index 1778805..9f19e19 100644
--- a/arch/cris/include/asm/Kbuild
+++ b/arch/cris/include/asm/Kbuild
@@ -4,7 +4,6 @@ generic-y += barrier.h
generic-y += bitsperlong.h
generic-y += clkdev.h
generic-y += cmpxchg.h
-generic-y += cputime.h
generic-y += device.h
generic-y += div64.h
generic-y += errno.h
diff --git a/arch/frv/include/asm/Kbuild b/arch/frv/include/asm/Kbuild
index 1fa084c..0f5b0d5 100644
--- a/arch/frv/include/asm/Kbuild
+++ b/arch/frv/include/asm/Kbuild
@@ -1,6 +1,5 @@
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += exec.h
generic-y += irq_work.h
generic-y += mcs_spinlock.h
diff --git a/arch/h8300/include/asm/Kbuild b/arch/h8300/include/asm/Kbuild
index 373cb23..5efd0c8 100644
--- a/arch/h8300/include/asm/Kbuild
+++ b/arch/h8300/include/asm/Kbuild
@@ -5,7 +5,6 @@ generic-y += bugs.h
generic-y += cacheflush.h
generic-y += checksum.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += delay.h
generic-y += device.h
diff --git a/arch/hexagon/include/asm/Kbuild b/arch/hexagon/include/asm/Kbuild
index db8ddab..a43a7c9 100644
--- a/arch/hexagon/include/asm/Kbuild
+++ b/arch/hexagon/include/asm/Kbuild
@@ -6,7 +6,6 @@ generic-y += barrier.h
generic-y += bug.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/ia64/include/asm/cputime.h b/arch/ia64/include/asm/cputime.h
index 44bcffc..3d665c0 100644
--- a/arch/ia64/include/asm/cputime.h
+++ b/arch/ia64/include/asm/cputime.h
@@ -18,9 +18,7 @@
#ifndef __IA64_CPUTIME_H
#define __IA64_CPUTIME_H
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-# include <asm-generic/cputime.h>
-#else
+#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
extern void arch_vtime_task_switch(struct task_struct *tsk);
#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
diff --git a/arch/m32r/include/asm/Kbuild b/arch/m32r/include/asm/Kbuild
index 860e440..652100b 100644
--- a/arch/m32r/include/asm/Kbuild
+++ b/arch/m32r/include/asm/Kbuild
@@ -1,6 +1,5 @@
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += exec.h
generic-y += irq_work.h
generic-y += kvm_para.h
diff --git a/arch/m68k/include/asm/Kbuild b/arch/m68k/include/asm/Kbuild
index 1f2e5d3..6c76d6c 100644
--- a/arch/m68k/include/asm/Kbuild
+++ b/arch/m68k/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += barrier.h
generic-y += bitsperlong.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += device.h
generic-y += emergency-restart.h
generic-y += errno.h
diff --git a/arch/metag/include/asm/Kbuild b/arch/metag/include/asm/Kbuild
index 167150c..d3731f0 100644
--- a/arch/metag/include/asm/Kbuild
+++ b/arch/metag/include/asm/Kbuild
@@ -2,7 +2,6 @@ generic-y += auxvec.h
generic-y += bitsperlong.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += dma.h
diff --git a/arch/microblaze/include/asm/Kbuild b/arch/microblaze/include/asm/Kbuild
index b0ae88c..6275eb0 100644
--- a/arch/microblaze/include/asm/Kbuild
+++ b/arch/microblaze/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += barrier.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += device.h
generic-y += exec.h
generic-y += irq_work.h
diff --git a/arch/mips/include/asm/Kbuild b/arch/mips/include/asm/Kbuild
index 3269b74..994b1c4 100644
--- a/arch/mips/include/asm/Kbuild
+++ b/arch/mips/include/asm/Kbuild
@@ -1,7 +1,6 @@
# MIPS headers
generic-(CONFIG_GENERIC_CSUM) += checksum.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += dma-contiguous.h
generic-y += emergency-restart.h
diff --git a/arch/mn10300/include/asm/Kbuild b/arch/mn10300/include/asm/Kbuild
index 1c8dd0f..97f64c7 100644
--- a/arch/mn10300/include/asm/Kbuild
+++ b/arch/mn10300/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += barrier.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += exec.h
generic-y += irq_work.h
generic-y += mcs_spinlock.h
diff --git a/arch/nios2/include/asm/Kbuild b/arch/nios2/include/asm/Kbuild
index d63330e..35b0e88 100644
--- a/arch/nios2/include/asm/Kbuild
+++ b/arch/nios2/include/asm/Kbuild
@@ -6,7 +6,6 @@ generic-y += bitsperlong.h
generic-y += bug.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/openrisc/include/asm/Kbuild b/arch/openrisc/include/asm/Kbuild
index 2832f03..ef8d1cc 100644
--- a/arch/openrisc/include/asm/Kbuild
+++ b/arch/openrisc/include/asm/Kbuild
@@ -12,7 +12,6 @@ generic-y += checksum.h
generic-y += clkdev.h
generic-y += cmpxchg-local.h
generic-y += cmpxchg.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/parisc/include/asm/Kbuild b/arch/parisc/include/asm/Kbuild
index 91f53c0..4e179d7 100644
--- a/arch/parisc/include/asm/Kbuild
+++ b/arch/parisc/include/asm/Kbuild
@@ -2,7 +2,6 @@
generic-y += auxvec.h
generic-y += barrier.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += device.h
generic-y += div64.h
generic-y += emergency-restart.h
diff --git a/arch/powerpc/include/asm/cputime.h b/arch/powerpc/include/asm/cputime.h
index 6ec0ba6..99b5418 100644
--- a/arch/powerpc/include/asm/cputime.h
+++ b/arch/powerpc/include/asm/cputime.h
@@ -16,9 +16,7 @@
#ifndef __POWERPC_CPUTIME_H
#define __POWERPC_CPUTIME_H
-#ifndef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
-#include <asm-generic/cputime.h>
-#else
+#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
#include <linux/types.h>
#include <linux/time.h>
diff --git a/arch/score/include/asm/Kbuild b/arch/score/include/asm/Kbuild
index a05218f..51970bb 100644
--- a/arch/score/include/asm/Kbuild
+++ b/arch/score/include/asm/Kbuild
@@ -4,7 +4,6 @@ header-y +=
generic-y += barrier.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += irq_work.h
generic-y += mcs_spinlock.h
generic-y += mm-arch-hooks.h
diff --git a/arch/sh/include/asm/Kbuild b/arch/sh/include/asm/Kbuild
index 751c337..cf2a750 100644
--- a/arch/sh/include/asm/Kbuild
+++ b/arch/sh/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += bitsperlong.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += delay.h
generic-y += div64.h
diff --git a/arch/sparc/include/asm/Kbuild b/arch/sparc/include/asm/Kbuild
index 0569bfa..e9e837b 100644
--- a/arch/sparc/include/asm/Kbuild
+++ b/arch/sparc/include/asm/Kbuild
@@ -2,7 +2,6 @@
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += div64.h
generic-y += emergency-restart.h
generic-y += exec.h
diff --git a/arch/tile/include/asm/Kbuild b/arch/tile/include/asm/Kbuild
index 2d1f563..51a339f 100644
--- a/arch/tile/include/asm/Kbuild
+++ b/arch/tile/include/asm/Kbuild
@@ -4,7 +4,6 @@ header-y += ../arch/
generic-y += bug.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += div64.h
generic-y += emergency-restart.h
generic-y += errno.h
diff --git a/arch/um/include/asm/Kbuild b/arch/um/include/asm/Kbuild
index 052f7f6..90c281c 100644
--- a/arch/um/include/asm/Kbuild
+++ b/arch/um/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += barrier.h
generic-y += bug.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += delay.h
generic-y += device.h
diff --git a/arch/unicore32/include/asm/Kbuild b/arch/unicore32/include/asm/Kbuild
index 256c45b..5d51ade 100644
--- a/arch/unicore32/include/asm/Kbuild
+++ b/arch/unicore32/include/asm/Kbuild
@@ -4,7 +4,6 @@ generic-y += auxvec.h
generic-y += bitsperlong.h
generic-y += bugs.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += current.h
generic-y += device.h
generic-y += div64.h
diff --git a/arch/x86/include/asm/Kbuild b/arch/x86/include/asm/Kbuild
index 2b892e2..5d6a53f 100644
--- a/arch/x86/include/asm/Kbuild
+++ b/arch/x86/include/asm/Kbuild
@@ -7,7 +7,6 @@ generated-y += unistd_64_x32.h
generated-y += xen-hypercalls.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += dma-contiguous.h
generic-y += early_ioremap.h
generic-y += mcs_spinlock.h
diff --git a/arch/xtensa/include/asm/Kbuild b/arch/xtensa/include/asm/Kbuild
index b7fbaa5..9e9760b 100644
--- a/arch/xtensa/include/asm/Kbuild
+++ b/arch/xtensa/include/asm/Kbuild
@@ -1,7 +1,6 @@
generic-y += bitsperlong.h
generic-y += bug.h
generic-y += clkdev.h
-generic-y += cputime.h
generic-y += div64.h
generic-y += dma-contiguous.h
generic-y += emergency-restart.h
diff --git a/include/asm-generic/cputime.h b/include/asm-generic/cputime.h
deleted file mode 100644
index 358e547..0000000
--- a/include/asm-generic/cputime.h
+++ /dev/null
@@ -1,7 +0,0 @@
-#ifndef _ASM_GENERIC_CPUTIME_H
-#define _ASM_GENERIC_CPUTIME_H
-
-#include <linux/time.h>
-#include <linux/jiffies.h>
-
-#endif
diff --git a/include/linux/cputime.h b/include/linux/cputime.h
index a257d66..a691dc4 100644
--- a/include/linux/cputime.h
+++ b/include/linux/cputime.h
@@ -1,6 +1,7 @@
#ifndef __LINUX_CPUTIME_H
#define __LINUX_CPUTIME_H
+#ifdef CONFIG_VIRT_CPU_ACCOUNTING_NATIVE
#include <asm/cputime.h>
#ifndef cputime_to_nsecs
@@ -8,4 +9,5 @@
(cputime_to_usecs(__ct) * NSEC_PER_USEC)
#endif
+#endif /* CONFIG_VIRT_CPU_ACCOUNTING_NATIVE */
#endif /* __LINUX_CPUTIME_H */
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* [PATCH 37/37] s390: Prevent from cputime leaks
2017-01-22 18:19 [PATCH 00/37] cputime: Convert core use of cputime_t to nsecs v3 Frederic Weisbecker
` (35 preceding siblings ...)
2017-01-22 18:20 ` [PATCH 36/37] cputime: Remove asm generic headers Frederic Weisbecker
@ 2017-01-22 18:20 ` Frederic Weisbecker
2017-01-23 9:44 ` Martin Schwidefsky
36 siblings, 1 reply; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-22 18:20 UTC (permalink / raw)
To: LKML
Cc: Frederic Weisbecker, Tony Luck, Wanpeng Li, Peter Zijlstra,
Michael Ellerman, Heiko Carstens, Benjamin Herrenschmidt,
Thomas Gleixner, Paul Mackerras, Ingo Molnar, Fenghua Yu,
Rik van Riel, Martin Schwidefsky, Stanislaw Gruszka
The s390 clock has a higher granularity than nanoseconds. 1 nanosec
equals 4.096 in s390 cputime_t. Therefore we leak a remainder while
flushing the cputime through cputime_to_nsecs().
For more precision, make sure we keep that remainder on cputime
accumulators for later accounting.
Reported-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Rik van Riel <riel@redhat.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Stanislaw Gruszka <sgruszka@redhat.com>
Cc: Wanpeng Li <wanpeng.li@hotmail.com>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
---
arch/s390/include/asm/cputime.h | 4 ++++
arch/s390/kernel/vtime.c | 33 +++++++++++++++++++++------------
2 files changed, 25 insertions(+), 12 deletions(-)
diff --git a/arch/s390/include/asm/cputime.h b/arch/s390/include/asm/cputime.h
index d1c407d..86b5e3f 100644
--- a/arch/s390/include/asm/cputime.h
+++ b/arch/s390/include/asm/cputime.h
@@ -33,6 +33,10 @@ static inline unsigned int cputime_to_usecs(const cputime_t cputime)
return (__force unsigned long long) cputime >> 12;
}
+static inline cputime_t nsecs_to_cputime(u64 cputime)
+{
+ return (__force cputime_t) cputime * CPUTIME_PER_USEC * NSEC_PER_USEC;
+}
u64 arch_cpu_idle_time(int cpu);
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index b4a3e9e..922f959 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -94,8 +94,14 @@ static inline u64 update_tsk_timer(unsigned long *tsk_vtime, u64 new)
{
u64 delta;
- delta = new - *tsk_vtime;
- *tsk_vtime = new;
+ /*
+ * Since nsecs is less granular than cputime_t in s390,
+ * the conversion to nsecs is rounded. Make sure we don't
+ * lose the remainder.
+ */
+ delta = cputime_to_nsecs(new - *tsk_vtime);
+ *tsk_vtime += nsecs_to_cputime(delta);
+
return delta;
}
@@ -124,7 +130,8 @@ static void account_system_index_scaled(struct task_struct *p,
*/
static int do_account_vtime(struct task_struct *tsk)
{
- u64 timer, clock, user, guest, system, hardirq, softirq, steal;
+ u64 timer, clock, delta;
+ u64 user, guest, system, hardirq, softirq, steal;
timer = S390_lowcore.last_update_timer;
clock = S390_lowcore.last_update_clock;
@@ -161,18 +168,19 @@ static int do_account_vtime(struct task_struct *tsk)
READ_ONCE(S390_lowcore.hardirq_timer));
softirq = update_tsk_timer(&tsk->thread.softirq_timer,
READ_ONCE(S390_lowcore.softirq_timer));
- S390_lowcore.steal_timer +=
- clock - user - guest - system - hardirq - softirq;
+
+ delta = nsecs_to_cputime(user + guest + system + hardirq + softirq);
+ S390_lowcore.steal_timer += clock - delta;
/* Push account value */
if (user) {
- account_user_time(tsk, cputime_to_nsecs(user));
- tsk->utimescaled += cputime_to_nsecs(scale_vtime(user));
+ account_user_time(tsk, user);
+ tsk->utimescaled += scale_vtime(user);
}
if (guest) {
- account_guest_time(tsk, cputime_to_nsecs(guest));
- tsk->utimescaled += cputime_to_nsecs(scale_vtime(guest));
+ account_guest_time(tsk, guest);
+ tsk->utimescaled += scale_vtime(guest);
}
if (system)
@@ -187,11 +195,12 @@ static int do_account_vtime(struct task_struct *tsk)
steal = S390_lowcore.steal_timer;
if ((s64) steal > 0) {
- S390_lowcore.steal_timer = 0;
- account_steal_time(cputime_to_nsecs(steal));
+ u64 nsecs = cputime_to_nsecs(steal);
+ S390_lowcore.steal_timer -= nsecs_to_cputime(nsecs);
+ account_steal_time(nsecs);
}
- return virt_timer_forward(user + guest + system + hardirq + softirq);
+ return virt_timer_forward(delta);
}
void vtime_task_switch(struct task_struct *prev)
--
2.7.4
^ permalink raw reply related [flat|nested] 45+ messages in thread* Re: [PATCH 37/37] s390: Prevent from cputime leaks
2017-01-22 18:20 ` [PATCH 37/37] s390: Prevent from cputime leaks Frederic Weisbecker
@ 2017-01-23 9:44 ` Martin Schwidefsky
2017-01-25 15:25 ` Frederic Weisbecker
0 siblings, 1 reply; 45+ messages in thread
From: Martin Schwidefsky @ 2017-01-23 9:44 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: LKML, Tony Luck, Wanpeng Li, Peter Zijlstra, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Stanislaw Gruszka
On Sun, 22 Jan 2017 19:20:13 +0100
Frederic Weisbecker <fweisbec@gmail.com> wrote:
> The s390 clock has a higher granularity than nanoseconds. 1 nanosec
> equals 4.096 in s390 cputime_t. Therefore we leak a remainder while
> flushing the cputime through cputime_to_nsecs().
>
> For more precision, make sure we keep that remainder on cputime
> accumulators for later accounting.
>
> Reported-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
> Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> Cc: Paul Mackerras <paulus@samba.org>
> Cc: Michael Ellerman <mpe@ellerman.id.au>
> Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
> Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
> Cc: Tony Luck <tony.luck@intel.com>
> Cc: Fenghua Yu <fenghua.yu@intel.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Rik van Riel <riel@redhat.com>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@kernel.org>
> Cc: Stanislaw Gruszka <sgruszka@redhat.com>
> Cc: Wanpeng Li <wanpeng.li@hotmail.com>
> Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
NAK. Good intention but the patch is just broken. with 36 of the 37
patches applied all looks good but the last one completely breaks the
accounting for s390. This is from an idle system:
top - 10:39:33 up 0 min, 1 user, load average: 0,00, 0,00, 0,00
Tasks: 106 total, 1 running, 105 sleeping, 0 stopped, 0 zombie
%Cpu0 : 8,9 us, 21,6 sy, 0,0 ni, 0,0 id, 0,0 wa, 10,8 hi, 4,3 si, 54,4 st
%Cpu1 : 0,0 us, 23,5 sy, 0,0 ni, 0,0 id, 0,0 wa, 19,0 hi, 13,1 si, 44,3 st
%Cpu2 : 0,0 us, 30,3 sy, 0,0 ni, 0,0 id, 0,0 wa, 14,7 hi, 14,8 si, 40,2 st
KiB Mem : 1009304 total, 818808 free, 57284 used, 133212 buff/cache
KiB Swap: 1048556 total, 1048556 free, 0 used. 917356 avail Mem
There is another issue that affects precision, there is no s390 specific
version of cputime_to_nsecs. The generic version uses cputime_to_usecs
and mulitplies by 1000 to get nano-seconds. That already looses precision.
For now just drop that last patch please.
--
blue skies,
Martin.
"Reality continues to ruin my life." - Calvin.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH 37/37] s390: Prevent from cputime leaks
2017-01-23 9:44 ` Martin Schwidefsky
@ 2017-01-25 15:25 ` Frederic Weisbecker
2017-01-25 15:40 ` Martin Schwidefsky
0 siblings, 1 reply; 45+ messages in thread
From: Frederic Weisbecker @ 2017-01-25 15:25 UTC (permalink / raw)
To: Martin Schwidefsky
Cc: LKML, Tony Luck, Wanpeng Li, Peter Zijlstra, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Stanislaw Gruszka
On Mon, Jan 23, 2017 at 10:44:56AM +0100, Martin Schwidefsky wrote:
> On Sun, 22 Jan 2017 19:20:13 +0100
> Frederic Weisbecker <fweisbec@gmail.com> wrote:
>
> > The s390 clock has a higher granularity than nanoseconds. 1 nanosec
> > equals 4.096 in s390 cputime_t. Therefore we leak a remainder while
> > flushing the cputime through cputime_to_nsecs().
> >
> > For more precision, make sure we keep that remainder on cputime
> > accumulators for later accounting.
> >
> > Reported-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
> > Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> > Cc: Paul Mackerras <paulus@samba.org>
> > Cc: Michael Ellerman <mpe@ellerman.id.au>
> > Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
> > Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
> > Cc: Tony Luck <tony.luck@intel.com>
> > Cc: Fenghua Yu <fenghua.yu@intel.com>
> > Cc: Peter Zijlstra <peterz@infradead.org>
> > Cc: Rik van Riel <riel@redhat.com>
> > Cc: Thomas Gleixner <tglx@linutronix.de>
> > Cc: Ingo Molnar <mingo@kernel.org>
> > Cc: Stanislaw Gruszka <sgruszka@redhat.com>
> > Cc: Wanpeng Li <wanpeng.li@hotmail.com>
> > Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
>
> NAK. Good intention but the patch is just broken. with 36 of the 37
> patches applied all looks good but the last one completely breaks the
> accounting for s390. This is from an idle system:
>
> top - 10:39:33 up 0 min, 1 user, load average: 0,00, 0,00, 0,00
> Tasks: 106 total, 1 running, 105 sleeping, 0 stopped, 0 zombie
> %Cpu0 : 8,9 us, 21,6 sy, 0,0 ni, 0,0 id, 0,0 wa, 10,8 hi, 4,3 si, 54,4 st
> %Cpu1 : 0,0 us, 23,5 sy, 0,0 ni, 0,0 id, 0,0 wa, 19,0 hi, 13,1 si, 44,3 st
> %Cpu2 : 0,0 us, 30,3 sy, 0,0 ni, 0,0 id, 0,0 wa, 14,7 hi, 14,8 si, 40,2 st
> KiB Mem : 1009304 total, 818808 free, 57284 used, 133212 buff/cache
> KiB Swap: 1048556 total, 1048556 free, 0 used. 917356 avail Mem
Oh ok. I must have done something wrong.
>
> There is another issue that affects precision, there is no s390 specific
> version of cputime_to_nsecs. The generic version uses cputime_to_usecs
> and mulitplies by 1000 to get nano-seconds. That already looses precision.
That's right. And that's the point of this patch. I'm not sure we can have a
more precise version of cputime_to_nsecs() if 1 nsec == 4.096 cputime_t
>
> For now just drop that last patch please.
Ok, I'm leaving it apart.
Thanks.
^ permalink raw reply [flat|nested] 45+ messages in thread
* Re: [PATCH 37/37] s390: Prevent from cputime leaks
2017-01-25 15:25 ` Frederic Weisbecker
@ 2017-01-25 15:40 ` Martin Schwidefsky
0 siblings, 0 replies; 45+ messages in thread
From: Martin Schwidefsky @ 2017-01-25 15:40 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: LKML, Tony Luck, Wanpeng Li, Peter Zijlstra, Michael Ellerman,
Heiko Carstens, Benjamin Herrenschmidt, Thomas Gleixner,
Paul Mackerras, Ingo Molnar, Fenghua Yu, Rik van Riel,
Stanislaw Gruszka
On Wed, 25 Jan 2017 16:25:20 +0100
Frederic Weisbecker <fweisbec@gmail.com> wrote:
> On Mon, Jan 23, 2017 at 10:44:56AM +0100, Martin Schwidefsky wrote:
> > On Sun, 22 Jan 2017 19:20:13 +0100
> > Frederic Weisbecker <fweisbec@gmail.com> wrote:
> >
> > > The s390 clock has a higher granularity than nanoseconds. 1 nanosec
> > > equals 4.096 in s390 cputime_t. Therefore we leak a remainder while
> > > flushing the cputime through cputime_to_nsecs().
> > >
> > > For more precision, make sure we keep that remainder on cputime
> > > accumulators for later accounting.
> > >
> > > Reported-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
> > > Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
> > > Cc: Paul Mackerras <paulus@samba.org>
> > > Cc: Michael Ellerman <mpe@ellerman.id.au>
> > > Cc: Heiko Carstens <heiko.carstens@de.ibm.com>
> > > Cc: Martin Schwidefsky <schwidefsky@de.ibm.com>
> > > Cc: Tony Luck <tony.luck@intel.com>
> > > Cc: Fenghua Yu <fenghua.yu@intel.com>
> > > Cc: Peter Zijlstra <peterz@infradead.org>
> > > Cc: Rik van Riel <riel@redhat.com>
> > > Cc: Thomas Gleixner <tglx@linutronix.de>
> > > Cc: Ingo Molnar <mingo@kernel.org>
> > > Cc: Stanislaw Gruszka <sgruszka@redhat.com>
> > > Cc: Wanpeng Li <wanpeng.li@hotmail.com>
> > > Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
> >
> > NAK. Good intention but the patch is just broken. with 36 of the 37
> > patches applied all looks good but the last one completely breaks the
> > accounting for s390. This is from an idle system:
> >
> > top - 10:39:33 up 0 min, 1 user, load average: 0,00, 0,00, 0,00
> > Tasks: 106 total, 1 running, 105 sleeping, 0 stopped, 0 zombie
> > %Cpu0 : 8,9 us, 21,6 sy, 0,0 ni, 0,0 id, 0,0 wa, 10,8 hi, 4,3 si, 54,4 st
> > %Cpu1 : 0,0 us, 23,5 sy, 0,0 ni, 0,0 id, 0,0 wa, 19,0 hi, 13,1 si, 44,3 st
> > %Cpu2 : 0,0 us, 30,3 sy, 0,0 ni, 0,0 id, 0,0 wa, 14,7 hi, 14,8 si, 40,2 st
> > KiB Mem : 1009304 total, 818808 free, 57284 used, 133212 buff/cache
> > KiB Swap: 1048556 total, 1048556 free, 0 used. 917356 avail Mem
>
> Oh ok. I must have done something wrong.
>
> >
> > There is another issue that affects precision, there is no s390 specific
> > version of cputime_to_nsecs. The generic version uses cputime_to_usecs
> > and mulitplies by 1000 to get nano-seconds. That already looses precision.
>
> That's right. And that's the point of this patch. I'm not sure we can have a
> more precise version of cputime_to_nsecs() if 1 nsec == 4.096 cputime_t
I am thinking about a version that does the calculation (cputime * 1000) >> 12
in several steps, like this
nsecs = (((((cputime * 5) >> 3) * 5) >> 3) * 5) >> 3
To multiply with 5 is a shift by 2 and an add. In total that would gives us
6 shifts and 3 adds for the conversion.
--
blue skies,
Martin.
"Reality continues to ruin my life." - Calvin.
^ permalink raw reply [flat|nested] 45+ messages in thread