* [for-next][PATCH 0/2] tracing: Minor fixes to the start up tests
@ 2014-10-09 1:10 Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 1/2] tracing: Robustify wait loop Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread() Steven Rostedt
0 siblings, 2 replies; 4+ messages in thread
From: Steven Rostedt @ 2014-10-09 1:10 UTC (permalink / raw)
To: linux-kernel; +Cc: Ingo Molnar, Andrew Morton
git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace.git
for-next
Head SHA1: 0cd8f5614fbf8a598a0a47526ac7dc42ab2ba8cd
Peter Zijlstra (1):
tracing: Robustify wait loop
Steven Rostedt (1):
tracing: Clean up scheduling in trace_wakeup_test_thread()
----
kernel/trace/trace_events.c | 5 ++++-
kernel/trace/trace_selftest.c | 47 +++++++++++++++++++++++++++----------------
2 files changed, 34 insertions(+), 18 deletions(-)
^ permalink raw reply [flat|nested] 4+ messages in thread
* [for-next][PATCH 1/2] tracing: Robustify wait loop
2014-10-09 1:10 [for-next][PATCH 0/2] tracing: Minor fixes to the start up tests Steven Rostedt
@ 2014-10-09 1:10 ` Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread() Steven Rostedt
1 sibling, 0 replies; 4+ messages in thread
From: Steven Rostedt @ 2014-10-09 1:10 UTC (permalink / raw)
To: linux-kernel
Cc: Ingo Molnar, Andrew Morton, Fengguang Wu, Peter Zijlstra (Intel)
[-- Attachment #1: 0001-tracing-Robustify-wait-loop.patch --]
[-- Type: text/plain, Size: 1216 bytes --]
From: Peter Zijlstra <peterz@infradead.org>
The pending nested sleep debugging triggered on the potential stale
TASK_INTERRUPTIBLE in this code.
While there, fix the loop such that we won't revert to a while(1)
yield() 'spin' loop if we ever get a spurious wakeup.
And fix the actual issue by properly terminating the 'wait' loop by
setting TASK_RUNNING.
Link: http://lkml.kernel.org/p/20141008165110.GA14547@worktop.programming.kicks-ass.net
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
kernel/trace/trace_events.c | 5 ++++-
1 file changed, 4 insertions(+), 1 deletion(-)
diff --git a/kernel/trace/trace_events.c b/kernel/trace/trace_events.c
index ef06ce7e9cf8..0cc51edde3a8 100644
--- a/kernel/trace/trace_events.c
+++ b/kernel/trace/trace_events.c
@@ -2513,8 +2513,11 @@ static __init int event_test_thread(void *unused)
kfree(test_malloc);
set_current_state(TASK_INTERRUPTIBLE);
- while (!kthread_should_stop())
+ while (!kthread_should_stop()) {
schedule();
+ set_current_state(TASK_INTERRUPTIBLE);
+ }
+ __set_current_state(TASK_RUNNING);
return 0;
}
--
2.0.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread()
2014-10-09 1:10 [for-next][PATCH 0/2] tracing: Minor fixes to the start up tests Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 1/2] tracing: Robustify wait loop Steven Rostedt
@ 2014-10-09 1:10 ` Steven Rostedt
2014-10-09 9:39 ` Peter Zijlstra
1 sibling, 1 reply; 4+ messages in thread
From: Steven Rostedt @ 2014-10-09 1:10 UTC (permalink / raw)
To: linux-kernel; +Cc: Ingo Molnar, Andrew Morton, Peter Zijlstra
[-- Attachment #1: 0002-tracing-Clean-up-scheduling-in-trace_wakeup_test_thr.patch --]
[-- Type: text/plain, Size: 3537 bytes --]
From: Steven Rostedt <rostedt@goodmis.org>
Peter's new debugging tool triggers when tasks exit with !TASK_RUNNING.
The code in trace_wakeup_test_thread() also has a single schedule() call
that should be encompassed by a loop.
This cleans up the code a little to make it a bit more robust and
also makes the return exit properly with TASK_RUNNING.
Link: http://lkml.kernel.org/p/20141008135216.76142204@gandalf.local.home
Reported-by: Peter Zijlstra <peterz@infradead.org>
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
---
kernel/trace/trace_selftest.c | 47 +++++++++++++++++++++++++++----------------
1 file changed, 30 insertions(+), 17 deletions(-)
diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c
index 5ef60499dc8e..593f52b73551 100644
--- a/kernel/trace/trace_selftest.c
+++ b/kernel/trace/trace_selftest.c
@@ -1025,6 +1025,12 @@ trace_selftest_startup_nop(struct tracer *trace, struct trace_array *tr)
#endif
#ifdef CONFIG_SCHED_TRACER
+
+struct wakeup_test_data {
+ struct completion is_ready;
+ int go;
+};
+
static int trace_wakeup_test_thread(void *data)
{
/* Make this a -deadline thread */
@@ -1034,51 +1040,56 @@ static int trace_wakeup_test_thread(void *data)
.sched_deadline = 10000000ULL,
.sched_period = 10000000ULL
};
- struct completion *x = data;
+ struct wakeup_test_data *x = data;
sched_setattr(current, &attr);
/* Make it know we have a new prio */
- complete(x);
+ complete(&x->is_ready);
/* now go to sleep and let the test wake us up */
set_current_state(TASK_INTERRUPTIBLE);
- schedule();
+ while (!x->go) {
+ schedule();
+ set_current_state(TASK_INTERRUPTIBLE);
+ }
- complete(x);
+ complete(&x->is_ready);
+
+ set_current_state(TASK_INTERRUPTIBLE);
/* we are awake, now wait to disappear */
while (!kthread_should_stop()) {
- /*
- * This will likely be the system top priority
- * task, do short sleeps to let others run.
- */
- msleep(100);
+ schedule();
+ set_current_state(TASK_INTERRUPTIBLE);
}
+ __set_current_state(TASK_RUNNING);
+
return 0;
}
-
int
trace_selftest_startup_wakeup(struct tracer *trace, struct trace_array *tr)
{
unsigned long save_max = tr->max_latency;
struct task_struct *p;
- struct completion is_ready;
+ struct wakeup_test_data data;
unsigned long count;
int ret;
- init_completion(&is_ready);
+ memset(&data, 0, sizeof(data));
+
+ init_completion(&data.is_ready);
/* create a -deadline thread */
- p = kthread_run(trace_wakeup_test_thread, &is_ready, "ftrace-test");
+ p = kthread_run(trace_wakeup_test_thread, &data, "ftrace-test");
if (IS_ERR(p)) {
printk(KERN_CONT "Failed to create ftrace wakeup test thread ");
return -1;
}
/* make sure the thread is running at -deadline policy */
- wait_for_completion(&is_ready);
+ wait_for_completion(&data.is_ready);
/* start the tracing */
ret = tracer_init(trace, tr);
@@ -1099,18 +1110,20 @@ trace_selftest_startup_wakeup(struct tracer *trace, struct trace_array *tr)
msleep(100);
}
- init_completion(&is_ready);
+ init_completion(&data.is_ready);
+
+ data.go = 1;
+ /* memory barrier is in the wake_up_process() */
wake_up_process(p);
/* Wait for the task to wake up */
- wait_for_completion(&is_ready);
+ wait_for_completion(&data.is_ready);
/* stop the tracing. */
tracing_stop();
/* check both trace buffers */
ret = trace_test_buffer(&tr->trace_buffer, NULL);
- printk("ret = %d\n", ret);
if (!ret)
ret = trace_test_buffer(&tr->max_buffer, &count);
--
2.0.1
^ permalink raw reply related [flat|nested] 4+ messages in thread
* Re: [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread()
2014-10-09 1:10 ` [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread() Steven Rostedt
@ 2014-10-09 9:39 ` Peter Zijlstra
0 siblings, 0 replies; 4+ messages in thread
From: Peter Zijlstra @ 2014-10-09 9:39 UTC (permalink / raw)
To: Steven Rostedt; +Cc: linux-kernel, Ingo Molnar, Andrew Morton
On Wed, Oct 08, 2014 at 09:10:12PM -0400, Steven Rostedt wrote:
> From: Steven Rostedt <rostedt@goodmis.org>
>
> Peter's new debugging tool triggers when tasks exit with !TASK_RUNNING.
> The code in trace_wakeup_test_thread() also has a single schedule() call
> that should be encompassed by a loop.
>
> This cleans up the code a little to make it a bit more robust and
> also makes the return exit properly with TASK_RUNNING.
>
> Link: http://lkml.kernel.org/p/20141008135216.76142204@gandalf.local.home
>
> Reported-by: Peter Zijlstra <peterz@infradead.org>
> Signed-off-by: Steven Rostedt <rostedt@goodmis.org>
Acked-by: Peter Zijlstra <peterz@infreadead.org>
> ---
> kernel/trace/trace_selftest.c | 47 +++++++++++++++++++++++++++----------------
> 1 file changed, 30 insertions(+), 17 deletions(-)
>
> diff --git a/kernel/trace/trace_selftest.c b/kernel/trace/trace_selftest.c
> index 5ef60499dc8e..593f52b73551 100644
> --- a/kernel/trace/trace_selftest.c
> +++ b/kernel/trace/trace_selftest.c
> @@ -1025,6 +1025,12 @@ trace_selftest_startup_nop(struct tracer *trace, struct trace_array *tr)
> #endif
>
> #ifdef CONFIG_SCHED_TRACER
> +
> +struct wakeup_test_data {
> + struct completion is_ready;
> + int go;
> +};
Yeah that works..
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2014-10-09 9:39 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2014-10-09 1:10 [for-next][PATCH 0/2] tracing: Minor fixes to the start up tests Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 1/2] tracing: Robustify wait loop Steven Rostedt
2014-10-09 1:10 ` [for-next][PATCH 2/2] tracing: Clean up scheduling in trace_wakeup_test_thread() Steven Rostedt
2014-10-09 9:39 ` Peter Zijlstra
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).