From: Frederic Weisbecker <fweisbec@gmail.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
Frederic Weisbecker <fweisbec@gmail.com>,
Jiri Olsa <jolsa@redhat.com>, Namhyung Kim <namhyung@kernel.org>,
Ingo Molnar <mingo@kernel.org>,
Arnaldo Carvalho de Melo <acme@redhat.com>,
Stephane Eranian <eranian@google.com>
Subject: [PATCH 7/8] perf: Finer grained full dynticks kick
Date: Tue, 23 Jul 2013 02:31:05 +0200 [thread overview]
Message-ID: <1374539466-4799-8-git-send-email-fweisbec@gmail.com> (raw)
In-Reply-To: <1374539466-4799-1-git-send-email-fweisbec@gmail.com>
Currently the full dynticks subsystem keep the
tick alive as long as there are perf events running.
This prevents the tick from being stopped as long as features
such that the lockup detectors are running. As a temporary fix,
the lockup detector is disabled by default when full dynticks
is built but this is not a long term viable solution.
To fix this, only keep the tick alive when an event configured
with a frequency rather than a period is running on the CPU,
or when an event throttles on the CPU.
These are the only purposes of the perf tick, especially now that
the rotation of flexible events is handled from a seperate hrtimer.
The tick can be shutdown the rest of the time.
Original-patch-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Frederic Weisbecker <fweisbec@gmail.com>
Cc: Jiri Olsa <jolsa@redhat.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Namhyung Kim <namhyung@kernel.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Stephane Eranian <eranian@google.com>
---
kernel/events/core.c | 17 +++++++++--------
1 files changed, 9 insertions(+), 8 deletions(-)
diff --git a/kernel/events/core.c b/kernel/events/core.c
index f9bd39b..61d70b7 100644
--- a/kernel/events/core.c
+++ b/kernel/events/core.c
@@ -870,12 +870,8 @@ static void perf_pmu_rotate_start(struct pmu *pmu)
WARN_ON(!irqs_disabled());
- if (list_empty(&cpuctx->rotation_list)) {
- int was_empty = list_empty(head);
+ if (list_empty(&cpuctx->rotation_list))
list_add(&cpuctx->rotation_list, head);
- if (was_empty)
- tick_nohz_full_kick();
- }
}
static void get_ctx(struct perf_event_context *ctx)
@@ -1875,6 +1871,9 @@ static int __perf_install_in_context(void *info)
perf_pmu_enable(cpuctx->ctx.pmu);
perf_ctx_unlock(cpuctx, task_ctx);
+ if (atomic_read(&__get_cpu_var(perf_freq_events)))
+ tick_nohz_full_kick();
+
return 0;
}
@@ -2812,10 +2811,11 @@ done:
#ifdef CONFIG_NO_HZ_FULL
bool perf_event_can_stop_tick(void)
{
- if (list_empty(&__get_cpu_var(rotation_list)))
- return true;
- else
+ if (atomic_read(&__get_cpu_var(perf_freq_events)) ||
+ __this_cpu_read(perf_throttled_count))
return false;
+ else
+ return true;
}
#endif
@@ -5201,6 +5201,7 @@ static int __perf_event_overflow(struct perf_event *event,
__this_cpu_inc(perf_throttled_count);
hwc->interrupts = MAX_INTERRUPTS;
perf_log_throttle(event, 0);
+ tick_nohz_full_kick();
ret = 1;
}
}
--
1.7.5.4
next prev parent reply other threads:[~2013-07-23 0:31 UTC|newest]
Thread overview: 49+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-07-23 0:30 [PATCH 0/8] perf: Finer grained full dynticks handling Frederic Weisbecker
2013-07-23 0:30 ` [PATCH 1/8] perf: Fix branch stack refcount leak on callchain init failure Frederic Weisbecker
2013-07-31 8:55 ` [tip:perf/core] " tip-bot for Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 2/8] perf: Sanitize get_callchain_buffer() Frederic Weisbecker
2013-07-31 8:56 ` [tip:perf/core] " tip-bot for Frederic Weisbecker
2013-08-01 13:01 ` [PATCH 2/8] " Jiri Olsa
2013-08-01 13:28 ` Frederic Weisbecker
2013-08-01 13:32 ` Jiri Olsa
2013-08-01 13:49 ` Frederic Weisbecker
2013-08-01 13:54 ` Jiri Olsa
2013-08-01 13:57 ` Frederic Weisbecker
2013-08-01 13:29 ` Jiri Olsa
2013-08-01 13:42 ` Frederic Weisbecker
2013-08-01 13:51 ` Jiri Olsa
2013-08-01 14:30 ` Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 3/8] perf: Gather event accounting code Frederic Weisbecker
2013-07-31 8:56 ` [tip:perf/core] perf: Factor out event accounting code to account_event()/__free_event() tip-bot for Frederic Weisbecker
2013-08-01 13:13 ` [PATCH 3/8] perf: Gather event accounting code Jiri Olsa
2013-08-01 13:30 ` Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 4/8] perf: Split per cpu " Frederic Weisbecker
2013-07-31 8:56 ` [tip:perf/core] perf: Split the per-cpu accounting part of the " tip-bot for Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 5/8] perf: Migrate per cpu event accounting Frederic Weisbecker
2013-07-31 8:56 ` [tip:perf/core] " tip-bot for Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 6/8] perf: Account freq events per cpu Frederic Weisbecker
2013-07-31 8:56 ` [tip:perf/core] " tip-bot for Frederic Weisbecker
2013-08-01 12:46 ` [PATCH 6/8] " Jiri Olsa
2013-08-01 12:48 ` Jiri Olsa
2013-08-01 13:31 ` Peter Zijlstra
2013-08-01 13:35 ` Peter Zijlstra
2013-08-01 13:39 ` Jiri Olsa
2013-08-01 13:56 ` Peter Zijlstra
2013-08-01 13:55 ` Frederic Weisbecker
2013-08-01 14:03 ` Peter Zijlstra
2013-08-01 14:06 ` Peter Zijlstra
2013-08-01 14:21 ` Frederic Weisbecker
2013-08-01 14:40 ` Peter Zijlstra
2013-08-02 16:25 ` Frederic Weisbecker
2013-08-01 14:19 ` Frederic Weisbecker
2013-07-23 0:31 ` Frederic Weisbecker [this message]
2013-07-31 8:56 ` [tip:perf/core] perf: Implement finer grained full dynticks kick tip-bot for Frederic Weisbecker
2013-07-23 0:31 ` [PATCH 8/8] watchdog: Remove hack to make full dynticks working Frederic Weisbecker
2013-07-23 12:33 ` Don Zickus
2013-07-23 12:44 ` Frederic Weisbecker
2013-07-23 12:45 ` Peter Zijlstra
2013-07-31 8:57 ` [tip:perf/core] watchdog: Make it work under full dynticks tip-bot for Frederic Weisbecker
2013-07-25 9:59 ` [PATCH 0/8] perf: Finer grained full dynticks handling Peter Zijlstra
2013-07-25 14:02 ` Frederic Weisbecker
2013-07-25 16:29 ` Peter Zijlstra
2013-07-25 20:07 ` Frederic Weisbecker
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1374539466-4799-8-git-send-email-fweisbec@gmail.com \
--to=fweisbec@gmail.com \
--cc=acme@redhat.com \
--cc=eranian@google.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=namhyung@kernel.org \
--cc=peterz@infradead.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).