* [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment
@ 2013-09-16 22:43 Kevin Hilman
2013-09-16 22:43 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
2013-11-18 21:42 ` [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
0 siblings, 2 replies; 8+ messages in thread
From: Kevin Hilman @ 2013-09-16 22:43 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: linux-arm-kernel, linaro-kernel, Paul McKenney, linux-kernel
Allow debugfs override of sched_tick_max_deferment in order to ease
finding/fixing the remaining issues with full nohz.
The value to be written is in jiffies, and -1 means the max deferment
is disabled (scheduler_tick_max_deferment() returns KTIME_MAX.)
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Kevin Hilman <khilman@linaro.org>
---
kernel/sched/core.c | 16 +++++++++++++++-
1 file changed, 15 insertions(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 5ac63c9..4b1fe3e 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2175,6 +2175,8 @@ void scheduler_tick(void)
}
#ifdef CONFIG_NO_HZ_FULL
+static u32 sched_tick_max_deferment = HZ;
+
/**
* scheduler_tick_max_deferment
*
@@ -2193,13 +2195,25 @@ u64 scheduler_tick_max_deferment(void)
struct rq *rq = this_rq();
unsigned long next, now = ACCESS_ONCE(jiffies);
- next = rq->last_sched_tick + HZ;
+ if (sched_tick_max_deferment == -1)
+ return KTIME_MAX;
+
+ next = rq->last_sched_tick + sched_tick_max_deferment;
if (time_before_eq(next, now))
return 0;
return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
}
+
+static __init int sched_nohz_full_init_debug(void)
+{
+ debugfs_create_u32("sched_tick_max_deferment", 0644, NULL,
+ &sched_tick_max_deferment);
+
+ return 0;
+}
+late_initcall(sched_nohz_full_init_debug);
#endif
notrace unsigned long get_parent_ip(unsigned long addr)
--
1.8.3
^ permalink raw reply related [flat|nested] 8+ messages in thread
* [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2013-09-16 22:43 [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
@ 2013-09-16 22:43 ` Kevin Hilman
2013-11-18 21:42 ` [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
1 sibling, 0 replies; 8+ messages in thread
From: Kevin Hilman @ 2013-09-16 22:43 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: linux-arm-kernel, linaro-kernel, Paul McKenney, linux-kernel
The conversion of the max deferment from usecs to nsecs can easily
overflow on platforms where a long is 32-bits. To fix, cast the usecs
value to u64 before multiplying by NSECS_PER_USEC.
This was discovered on 32-bit ARM platform when extending the max
deferment value.
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Kevin Hilman <khilman@linaro.org>
---
kernel/sched/core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 4b1fe3e..3d7c80e 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2203,7 +2203,7 @@ u64 scheduler_tick_max_deferment(void)
if (time_before_eq(next, now))
return 0;
- return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
+ return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
}
static __init int sched_nohz_full_init_debug(void)
--
1.8.3
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment
2013-09-16 22:43 [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
2013-09-16 22:43 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
@ 2013-11-18 21:42 ` Kevin Hilman
1 sibling, 0 replies; 8+ messages in thread
From: Kevin Hilman @ 2013-11-18 21:42 UTC (permalink / raw)
To: Frederic Weisbecker; +Cc: linux-arm-kernel, linaro-kernel, Paul McKenney, LKML
Frederic,
On Mon, Sep 16, 2013 at 3:43 PM, Kevin Hilman <khilman@linaro.org> wrote:
> Allow debugfs override of sched_tick_max_deferment in order to ease
> finding/fixing the remaining issues with full nohz.
>
> The value to be written is in jiffies, and -1 means the max deferment
> is disabled (scheduler_tick_max_deferment() returns KTIME_MAX.)
>
> Cc: Frederic Weisbecker <fweisbec@gmail.com>
> Signed-off-by: Kevin Hilman <khilman@linaro.org>
Any objections to this series for ease of debugging and tracking down
issues with the residual tick?
Kevin
> ---
> kernel/sched/core.c | 16 +++++++++++++++-
> 1 file changed, 15 insertions(+), 1 deletion(-)
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 5ac63c9..4b1fe3e 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -2175,6 +2175,8 @@ void scheduler_tick(void)
> }
>
> #ifdef CONFIG_NO_HZ_FULL
> +static u32 sched_tick_max_deferment = HZ;
> +
> /**
> * scheduler_tick_max_deferment
> *
> @@ -2193,13 +2195,25 @@ u64 scheduler_tick_max_deferment(void)
> struct rq *rq = this_rq();
> unsigned long next, now = ACCESS_ONCE(jiffies);
>
> - next = rq->last_sched_tick + HZ;
> + if (sched_tick_max_deferment == -1)
> + return KTIME_MAX;
> +
> + next = rq->last_sched_tick + sched_tick_max_deferment;
>
> if (time_before_eq(next, now))
> return 0;
>
> return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
> }
> +
> +static __init int sched_nohz_full_init_debug(void)
> +{
> + debugfs_create_u32("sched_tick_max_deferment", 0644, NULL,
> + &sched_tick_max_deferment);
> +
> + return 0;
> +}
> +late_initcall(sched_nohz_full_init_debug);
> #endif
>
> notrace unsigned long get_parent_ip(unsigned long addr)
> --
> 1.8.3
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2014-01-05 13:06 ` Frederic Weisbecker
@ 2014-01-06 18:27 ` Kevin Hilman
0 siblings, 0 replies; 8+ messages in thread
From: Kevin Hilman @ 2014-01-06 18:27 UTC (permalink / raw)
To: Frederic Weisbecker
Cc: Thomas Gleixner, linux-kernel, linaro-kernel, Peter Zijlstra,
Ingo Molnar
Frederic Weisbecker <fweisbec@gmail.com> writes:
> On Tue, Dec 17, 2013 at 01:23:08PM -0800, Kevin Hilman wrote:
>> The conversion of the max deferment from usecs to nsecs can easily
>> overflow on platforms where a long is 32-bits. To fix, cast the usecs
>> value to u64 before multiplying by NSECS_PER_USEC.
>>
>> This was discovered on 32-bit ARM platform when extending the max
>> deferment value.
>>
>> Cc: Frederic Weisbecker <fweisbec@gmail.com>
>> Signed-off-by: Kevin Hilman <khilman@linaro.org>
>> ---
>> kernel/sched/core.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
>> index 4b1fe3e69fe4..3d7c80e1c4d9 100644
>> --- a/kernel/sched/core.c
>> +++ b/kernel/sched/core.c
>> @@ -2203,7 +2203,7 @@ u64 scheduler_tick_max_deferment(void)
>> if (time_before_eq(next, now))
>> return 0;
>>
>> - return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
>> + return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
>
> Just to be sure I understand the issue. The problem is that jiffies_to_usecs()
> return an unsigned int which is then multiplied by NSEC_PER_USEC. If the result
> of the mul is too big to be stored in an unsigned int, we overflow and may loose
> some high part of the result. Right?
Correct.
Kevin
^ permalink raw reply [flat|nested] 8+ messages in thread
* Re: [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2013-12-17 21:23 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
@ 2014-01-05 13:06 ` Frederic Weisbecker
2014-01-06 18:27 ` Kevin Hilman
0 siblings, 1 reply; 8+ messages in thread
From: Frederic Weisbecker @ 2014-01-05 13:06 UTC (permalink / raw)
To: Kevin Hilman
Cc: Thomas Gleixner, linux-kernel, linaro-kernel, Peter Zijlstra,
Ingo Molnar
On Tue, Dec 17, 2013 at 01:23:08PM -0800, Kevin Hilman wrote:
> The conversion of the max deferment from usecs to nsecs can easily
> overflow on platforms where a long is 32-bits. To fix, cast the usecs
> value to u64 before multiplying by NSECS_PER_USEC.
>
> This was discovered on 32-bit ARM platform when extending the max
> deferment value.
>
> Cc: Frederic Weisbecker <fweisbec@gmail.com>
> Signed-off-by: Kevin Hilman <khilman@linaro.org>
> ---
> kernel/sched/core.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index 4b1fe3e69fe4..3d7c80e1c4d9 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -2203,7 +2203,7 @@ u64 scheduler_tick_max_deferment(void)
> if (time_before_eq(next, now))
> return 0;
>
> - return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
> + return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
Just to be sure I understand the issue. The problem is that jiffies_to_usecs()
return an unsigned int which is then multiplied by NSEC_PER_USEC. If the result
of the mul is too big to be stored in an unsigned int, we overflow and may loose
some high part of the result. Right?
> }
>
> static __init int sched_nohz_full_init_debug(void)
> --
> 1.8.3
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2013-12-17 21:23 Kevin Hilman
@ 2013-12-17 21:23 ` Kevin Hilman
2014-01-05 13:06 ` Frederic Weisbecker
0 siblings, 1 reply; 8+ messages in thread
From: Kevin Hilman @ 2013-12-17 21:23 UTC (permalink / raw)
To: Frederic Weisbecker, Thomas Gleixner; +Cc: linux-kernel, linaro-kernel
The conversion of the max deferment from usecs to nsecs can easily
overflow on platforms where a long is 32-bits. To fix, cast the usecs
value to u64 before multiplying by NSECS_PER_USEC.
This was discovered on 32-bit ARM platform when extending the max
deferment value.
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Kevin Hilman <khilman@linaro.org>
---
kernel/sched/core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 4b1fe3e69fe4..3d7c80e1c4d9 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2203,7 +2203,7 @@ u64 scheduler_tick_max_deferment(void)
if (time_before_eq(next, now))
return 0;
- return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
+ return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
}
static __init int sched_nohz_full_init_debug(void)
--
1.8.3
^ permalink raw reply related [flat|nested] 8+ messages in thread
* Re: [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2013-06-18 23:58 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
@ 2013-06-19 19:06 ` Frederic Weisbecker
0 siblings, 0 replies; 8+ messages in thread
From: Frederic Weisbecker @ 2013-06-19 19:06 UTC (permalink / raw)
To: Kevin Hilman; +Cc: linaro-kernel, Ingo Molnar, Peter Zijlstra, open list
On Tue, Jun 18, 2013 at 04:58:29PM -0700, Kevin Hilman wrote:
> The conversion of the max deferment from usecs to nsecs can easily
> overflow on platforms where a long is 32-bits. To fix, cast the usecs
> value to u64 before multiplying by NSECS_PER_USEC.
>
> This was discovered on 32-bit ARM platform when extending the max
> deferment value.
>
> Cc: Frederic Weisbecker <fweisbec@gmail.com>
> Signed-off-by: Kevin Hilman <khilman@linaro.org>
Right, if we make it tunable we need that patch.
Thanks!
Acked-by: Frederic Weisbecker <fweisbec@gmail.com>
> ---
> kernel/sched/core.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/kernel/sched/core.c b/kernel/sched/core.c
> index b5d3f99..b506722 100644
> --- a/kernel/sched/core.c
> +++ b/kernel/sched/core.c
> @@ -2765,7 +2765,7 @@ u64 scheduler_tick_max_deferment(void)
> if (time_before_eq(next, now))
> return 0;
>
> - return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
> + return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
> }
> #endif
>
> --
> 1.8.3
>
^ permalink raw reply [flat|nested] 8+ messages in thread
* [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment()
2013-06-18 23:58 [PATCH 1/2] sched/nohz: add sysctl control over sched_tick_max_deferment Kevin Hilman
@ 2013-06-18 23:58 ` Kevin Hilman
2013-06-19 19:06 ` Frederic Weisbecker
0 siblings, 1 reply; 8+ messages in thread
From: Kevin Hilman @ 2013-06-18 23:58 UTC (permalink / raw)
To: Frederic Weisbecker; +Cc: linaro-kernel, Ingo Molnar, Peter Zijlstra, open list
The conversion of the max deferment from usecs to nsecs can easily
overflow on platforms where a long is 32-bits. To fix, cast the usecs
value to u64 before multiplying by NSECS_PER_USEC.
This was discovered on 32-bit ARM platform when extending the max
deferment value.
Cc: Frederic Weisbecker <fweisbec@gmail.com>
Signed-off-by: Kevin Hilman <khilman@linaro.org>
---
kernel/sched/core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index b5d3f99..b506722 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -2765,7 +2765,7 @@ u64 scheduler_tick_max_deferment(void)
if (time_before_eq(next, now))
return 0;
- return jiffies_to_usecs(next - now) * NSEC_PER_USEC;
+ return (u64)jiffies_to_usecs(next - now) * NSEC_PER_USEC;
}
#endif
--
1.8.3
^ permalink raw reply related [flat|nested] 8+ messages in thread
end of thread, other threads:[~2014-01-06 18:27 UTC | newest]
Thread overview: 8+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2013-09-16 22:43 [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
2013-09-16 22:43 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
2013-11-18 21:42 ` [PATCH 1/2] sched/nohz: add debugfs control over sched_tick_max_deferment Kevin Hilman
-- strict thread matches above, loose matches on Subject: below --
2013-12-17 21:23 Kevin Hilman
2013-12-17 21:23 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
2014-01-05 13:06 ` Frederic Weisbecker
2014-01-06 18:27 ` Kevin Hilman
2013-06-18 23:58 [PATCH 1/2] sched/nohz: add sysctl control over sched_tick_max_deferment Kevin Hilman
2013-06-18 23:58 ` [PATCH 2/2] sched/nohz: fix overflow error in scheduler_tick_max_deferment() Kevin Hilman
2013-06-19 19:06 ` Frederic Weisbecker
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).