* Andi Kleen <andi@xxxxxxxxxxxxxx> wrote:
> > After this patch, reiserfs suffers from a slight performance
> > regression (for now).
>
> That might be related to the scheduling behaviour. Watch the
> reschedule counters in vmstat 1
FYI, Frederic is one of the main upstream tracing developers, and i
can assure you that when working on scalability he watches something
far more sophisticated than "vmstat 1" ;-)
See the commit below for example.
Thanks,
Ingo
--------------------->
>From 8f350bb74293dded6c51870c34a74645b054caba Mon Sep 17 00:00:00 2001
From: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Date: Mon, 6 Apr 2009 01:49:33 +0200
Subject: [PATCH] tracing/lockdep: report the time waited for a lock
While trying to optimize the new lock on reiserfs to replace
the bkl, I find the lock tracing very useful though it lacks
something important for performance (and latency) instrumentation:
the time a task waits for a lock.
That's what this patch implements:
bash-4816 [000] 202.652815: lock_contended: lock_contended: &sb->s_type->i_mutex_key
bash-4816 [000] 202.652819: lock_acquired: &rq->lock (0.000 us)
<...>-4787 [000] 202.652825: lock_acquired: &rq->lock (0.000 us)
<...>-4787 [000] 202.652829: lock_acquired: &rq->lock (0.000 us)
bash-4816 [000] 202.652833: lock_acquired: &sb->s_type->i_mutex_key (16.005 us)
As shown above, the "lock acquired" field is followed by the time
it has been waiting for the lock. Usually, a lock contended entry
is followed by a near lock_acquired entry with a non-zero time waited.
Signed-off-by: Frederic Weisbecker <fweisbec@xxxxxxxxx>
Acked-by: Peter Zijlstra <peterz@xxxxxxxxxxxxx>
Cc: Steven Rostedt <rostedt@xxxxxxxxxxx>
LKML-Reference: <1238975373-15739-1-git-send-email-fweisbec@xxxxxxxxx>
Signed-off-by: Ingo Molnar <mingo@xxxxxxx>
---
include/trace/lockdep_event_types.h | 23 ++++++++++++++++++-----
kernel/lockdep.c | 8 ++++----
2 files changed, 22 insertions(+), 9 deletions(-)
diff --git a/include/trace/lockdep_event_types.h b/include/trace/lockdep_event_types.h
index adccfcd..863f1e4 100644
--- a/include/trace/lockdep_event_types.h
+++ b/include/trace/lockdep_event_types.h
@@ -32,11 +32,24 @@ TRACE_FORMAT(lock_contended,
TP_FMT("%s", lock->name)
);
-TRACE_FORMAT(lock_acquired,
- TP_PROTO(struct lockdep_map *lock, unsigned long ip),
- TP_ARGS(lock, ip),
- TP_FMT("%s", lock->name)
- );
+TRACE_EVENT(lock_acquired,
+ TP_PROTO(struct lockdep_map *lock, unsigned long ip, s64 waittime),
+
+ TP_ARGS(lock, ip, waittime),
+
+ TP_STRUCT__entry(
+ __field(const char *, name)
+ __field(unsigned long, wait_usec)
+ __field(unsigned long, wait_nsec_rem)
+ ),
+ TP_fast_assign(
+ __entry->name = lock->name;
+ __entry->wait_nsec_rem = do_div(waittime, NSEC_PER_USEC);
+ __entry->wait_usec = (unsigned long) waittime;
+ ),
+ TP_printk("%s (%lu.%03lu us)", __entry->name, __entry->wait_usec,
+ __entry->wait_nsec_rem)
+);
#endif
#endif
diff --git a/kernel/lockdep.c b/kernel/lockdep.c
index b0f0118..c4582a6 100644
--- a/kernel/lockdep.c
+++ b/kernel/lockdep.c
@@ -3061,6 +3061,8 @@ found_it:
put_lock_stats(stats);
}
+DEFINE_TRACE(lock_acquired);
+
static void
__lock_acquired(struct lockdep_map *lock, unsigned long ip)
{
@@ -3099,6 +3101,8 @@ found_it:
hlock->holdtime_stamp = now;
}
+ trace_lock_acquired(lock, ip, waittime);
+
stats = get_lock_stats(hlock_class(hlock));
if (waittime) {
if (hlock->read)
@@ -3137,14 +3141,10 @@ void lock_contended(struct lockdep_map *lock, unsigned long ip)
}
EXPORT_SYMBOL_GPL(lock_contended);
-DEFINE_TRACE(lock_acquired);
-
void lock_acquired(struct lockdep_map *lock, unsigned long ip)
{
unsigned long flags;
- trace_lock_acquired(lock, ip);
-
if (unlikely(!lock_stat))
return;
--
To unsubscribe from this list: send the line "unsubscribe reiserfs-devel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
