linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH] sched/core: really pretend early bootup to be a normal task
@ 2016-07-05 11:23 Konstantin Khlebnikov
  2016-07-05 12:41 ` Peter Zijlstra
  0 siblings, 1 reply; 3+ messages in thread
From: Konstantin Khlebnikov @ 2016-07-05 11:23 UTC (permalink / raw)
  To: Peter Zijlstra, Ingo Molnar, linux-kernel; +Cc: Sasha Levin, Yao Dongdong

Commit 1b537c7d1e58 ("sched/core: Remove check of p->sched_class") placed
"current->sched_class = &fair_sched_class" before call of init_idle() which
immediately set sched_class back to idle_sched_class.

This patch reverts this piece because after commit de9b8f5dcbd9 ("sched:
Fix crash trying to dequeue/enqueue the idle thread") init_idle() doesn't
call do_set_cpus_allowed() and nobody will dereference ->sched_class.

[ Probably this pretending is no longer required and could be removed,
  or maybe it helps in catching bugs like sleeping during early bootup? ]

Signed-off-by: Konstantin Khlebnikov <khlebnikov@yandex-team.ru>
Fixes: 1b537c7d1e58 ("sched/core: Remove check of p->sched_class")
Cc: Yao Dongdong <yaodongdong@huawei.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Ingo Molnar <mingo@kernel.org>
Cc: Sasha Levin <sasha.levin@oracle.com>
---
 kernel/sched/core.c |   10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)

diff --git a/kernel/sched/core.c b/kernel/sched/core.c
index 51d7105f529a..c23aee77d10d 100644
--- a/kernel/sched/core.c
+++ b/kernel/sched/core.c
@@ -7481,11 +7481,6 @@ void __init sched_init(void)
 	enter_lazy_tlb(&init_mm, current);
 
 	/*
-	 * During early bootup we pretend to be a normal task:
-	 */
-	current->sched_class = &fair_sched_class;
-
-	/*
 	 * Make us the idle thread. Technically, schedule() should not be
 	 * called from this thread, however somewhere below it might be,
 	 * but because we are the idle thread, we just pick up running again
@@ -7495,6 +7490,11 @@ void __init sched_init(void)
 
 	calc_load_update = jiffies + LOAD_FREQ;
 
+	/*
+	 * During early bootup we pretend to be a normal task:
+	 */
+	current->sched_class = &fair_sched_class;
+
 #ifdef CONFIG_SMP
 	zalloc_cpumask_var(&sched_domains_tmpmask, GFP_NOWAIT);
 	/* May be allocated at isolcpus cmdline parse time */

^ permalink raw reply related	[flat|nested] 3+ messages in thread

* Re: [PATCH] sched/core: really pretend early bootup to be a normal task
  2016-07-05 11:23 [PATCH] sched/core: really pretend early bootup to be a normal task Konstantin Khlebnikov
@ 2016-07-05 12:41 ` Peter Zijlstra
  2016-07-05 12:47   ` Konstantin Khlebnikov
  0 siblings, 1 reply; 3+ messages in thread
From: Peter Zijlstra @ 2016-07-05 12:41 UTC (permalink / raw)
  To: Konstantin Khlebnikov
  Cc: Ingo Molnar, linux-kernel, Sasha Levin, Yao Dongdong

On Tue, Jul 05, 2016 at 02:23:23PM +0300, Konstantin Khlebnikov wrote:
> Commit 1b537c7d1e58 ("sched/core: Remove check of p->sched_class") placed
> "current->sched_class = &fair_sched_class" before call of init_idle() which
> immediately set sched_class back to idle_sched_class.
> 
> This patch reverts this piece because after commit de9b8f5dcbd9 ("sched:
> Fix crash trying to dequeue/enqueue the idle thread") init_idle() doesn't
> call do_set_cpus_allowed() and nobody will dereference ->sched_class.

I don't understand, what is broken?

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] sched/core: really pretend early bootup to be a normal task
  2016-07-05 12:41 ` Peter Zijlstra
@ 2016-07-05 12:47   ` Konstantin Khlebnikov
  0 siblings, 0 replies; 3+ messages in thread
From: Konstantin Khlebnikov @ 2016-07-05 12:47 UTC (permalink / raw)
  To: Peter Zijlstra; +Cc: Ingo Molnar, linux-kernel, Sasha Levin, Yao Dongdong

On 05.07.2016 15:41, Peter Zijlstra wrote:
> On Tue, Jul 05, 2016 at 02:23:23PM +0300, Konstantin Khlebnikov wrote:
>> Commit 1b537c7d1e58 ("sched/core: Remove check of p->sched_class") placed
>> "current->sched_class = &fair_sched_class" before call of init_idle() which
>> immediately set sched_class back to idle_sched_class.
>>
>> This patch reverts this piece because after commit de9b8f5dcbd9 ("sched:
>> Fix crash trying to dequeue/enqueue the idle thread") init_idle() doesn't
>> call do_set_cpus_allowed() and nobody will dereference ->sched_class.
>
> I don't understand, what is broken?
>

"pretending" hack is broken. As I wrote maybe it is not needed any more.

-- 
Konstantin

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2016-07-05 12:48 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2016-07-05 11:23 [PATCH] sched/core: really pretend early bootup to be a normal task Konstantin Khlebnikov
2016-07-05 12:41 ` Peter Zijlstra
2016-07-05 12:47   ` Konstantin Khlebnikov

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).