From: "Chen, Kenneth W" <kenneth.w.chen@intel.com>
To: "'Nick Piggin'" <nickpiggin@yahoo.com.au>
Cc: "Con Kolivas" <kernel@kolivas.org>,
<linux-kernel@vger.kernel.org>, "'Chris Mason'" <mason@suse.com>,
"Ingo Molnar" <mingo@elte.hu>
Subject: RE: [PATCH RFC] smt nice introduces significant lock contention
Date: Fri, 2 Jun 2006 02:36:54 -0700 [thread overview]
Message-ID: <000201c68628$16f64550$0c4ce984@amr.corp.intel.com> (raw)
In-Reply-To: <447FFD35.9020909@yahoo.com.au>
Nick Piggin wrote on Friday, June 02, 2006 1:56 AM
> Chen, Kenneth W wrote:
>
> > Ha, you beat me by one minute. It did cross my mind to use try lock there as
> > well, take a look at my version, I think I have a better inner loop.
>
> Actually you *have* to use trylocks I think, because the current runqueue
> is already locked.
>
> And why do we lock all siblings in the other case, for that matter? (not
> that it makes much difference except on niagara today).
>
> Rolled up patch with everyone's changes attached.
OK, it's down to nit-picking now:
Remove this_rq argument from wake_sleeping_dependent() since it is not
used. Nick, you had that in your earlier version, but it got lost in
the woods.
I don't like cpumask being declared on the stack. Here is my version
to rid it out in wake_sleeping_dependent() and dependent_sleeper().
- Ken
--- ./kernel/sched.c.orig 2006-06-02 03:20:40.000000000 -0700
+++ ./kernel/sched.c 2006-06-02 03:20:59.000000000 -0700
@@ -2714,10 +2714,9 @@ static inline void wakeup_busy_runqueue(
/*
* Called with interrupts disabled and this_rq's runqueue locked.
*/
-static void wake_sleeping_dependent(int this_cpu, runqueue_t *this_rq)
+static void wake_sleeping_dependent(int this_cpu)
{
struct sched_domain *tmp, *sd = NULL;
- cpumask_t sibling_map;
int i;
for_each_domain(this_cpu, tmp)
@@ -2728,10 +2727,11 @@ static void wake_sleeping_dependent(int
if (!sd)
return;
- sibling_map = sd->span;
- cpu_clear(this_cpu, sibling_map);
- for_each_cpu_mask(i, sibling_map) {
+ for_each_cpu_mask(i, sd->span) {
runqueue_t *smt_rq = cpu_rq(i);
+
+ if (i == this_cpu)
+ continue;
if (unlikely(!spin_trylock(&smt_rq->lock)))
continue;
@@ -2761,7 +2761,6 @@ static int dependent_sleeper(int this_cp
struct task_struct *p)
{
struct sched_domain *tmp, *sd = NULL;
- cpumask_t sibling_map;
int ret = 0, i;
for_each_domain(this_cpu, tmp)
@@ -2772,12 +2771,13 @@ static int dependent_sleeper(int this_cp
if (!sd)
return 0;
- sibling_map = sd->span;
- cpu_clear(this_cpu, sibling_map);
- for_each_cpu_mask(i, sibling_map) {
+ for_each_cpu_mask(i, sd->span) {
runqueue_t *smt_rq;
task_t *smt_curr;
+ if (i == this_cpu)
+ continue;
+
smt_rq = cpu_rq(i);
if (unlikely(!spin_trylock(&smt_rq->lock)))
continue;
@@ -2842,7 +2842,7 @@ check_smt_task:
return ret;
}
#else
-static inline void wake_sleeping_dependent(int this_cpu, runqueue_t *this_rq)
+static inline void wake_sleeping_dependent(int this_cpu)
{
}
@@ -2973,7 +2973,7 @@ need_resched_nonpreemptible:
if (!rq->nr_running) {
next = rq->idle;
rq->expired_timestamp = 0;
- wake_sleeping_dependent(cpu, rq);
+ wake_sleeping_dependent(cpu);
goto switch_tasks;
}
}
next prev parent reply other threads:[~2006-06-02 9:37 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2006-06-01 22:55 Chris Mason
2006-06-01 23:57 ` Chen, Kenneth W
2006-06-02 1:59 ` Con Kolivas
2006-06-02 2:28 ` Con Kolivas
2006-06-02 3:55 ` Con Kolivas
2006-06-02 4:18 ` Nick Piggin
2006-06-02 6:08 ` Con Kolivas
2006-06-02 7:53 ` Nick Piggin
2006-06-02 8:17 ` Con Kolivas
2006-06-02 8:28 ` Nick Piggin
2006-06-02 8:34 ` Chen, Kenneth W
2006-06-02 8:56 ` Nick Piggin
2006-06-02 9:17 ` Chen, Kenneth W
2006-06-02 9:25 ` Con Kolivas
2006-06-02 9:31 ` Chen, Kenneth W
2006-06-02 9:34 ` Con Kolivas
2006-06-02 9:53 ` Chen, Kenneth W
2006-06-02 10:12 ` Con Kolivas
2006-06-02 20:53 ` Chen, Kenneth W
2006-06-02 22:15 ` Con Kolivas
2006-06-02 22:19 ` Chen, Kenneth W
2006-06-02 22:31 ` Con Kolivas
2006-06-02 22:58 ` Chen, Kenneth W
2006-06-03 0:02 ` Con Kolivas
2006-06-03 0:08 ` Chen, Kenneth W
2006-06-03 0:27 ` Con Kolivas
2006-06-02 9:36 ` Chen, Kenneth W [this message]
2006-06-02 10:30 ` Con Kolivas
2006-06-02 13:16 ` Con Kolivas
2006-06-02 21:54 ` Chen, Kenneth W
2006-06-02 22:04 ` Con Kolivas
2006-06-02 22:14 ` Chen, Kenneth W
2006-06-02 10:19 ` Con Kolivas
2006-06-02 20:59 ` Chen, Kenneth W
2006-06-02 8:38 ` Chen, Kenneth W
2006-06-02 8:24 ` Chen, Kenneth W
2006-06-02 8:31 ` Chen, Kenneth W
2006-06-02 8:50 ` Chen, Kenneth W
2006-06-02 2:35 ` Chen, Kenneth W
2006-06-02 3:04 ` Con Kolivas
2006-06-02 3:23 ` Con Kolivas
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='000201c68628$16f64550$0c4ce984@amr.corp.intel.com' \
--to=kenneth.w.chen@intel.com \
--cc=kernel@kolivas.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mason@suse.com \
--cc=mingo@elte.hu \
--cc=nickpiggin@yahoo.com.au \
/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
Powered by JetHome