linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Steven Rostedt <rostedt@goodmis.org>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Peter Zijlstra <peterz@infradead.org>,
	Ingo Molnar <mingo@kernel.org>,
	Mark Rutland <mark.rutland@arm.com>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Subject: Re: [patch V2 01/17] init: Pin init task to boot cpu initially
Date: Tue, 16 May 2017 15:06:21 -0400	[thread overview]
Message-ID: <20170516150621.548862a2@gandalf.local.home> (raw)
In-Reply-To: <20170516184734.943149935@linutronix.de>

On Tue, 16 May 2017 20:42:32 +0200
Thomas Gleixner <tglx@linutronix.de> wrote:

> Some of the boot code in init_kernel_freeable() which runs before SMP
> bringup assumes (rightfully) that it runs on the boot cpu and therefor can
> use smp_processor_id() in preemptible context.
> 
> That works so far because the smp_processor_id() check starts to be
> effective after smp bringup. That's just wrong. Starting with SMP bringup
> and the ability to move threads around, smp_processor_id() in preemptible
> context is broken.
> 
> Aside of that it does not make sense to allow init to run on all cpus
> before sched_smp_init() has been run.
> 
> Pin the init to the boot cpu so the existing code can continue to use
> smp_processor_id() without triggering the checks when the enabling of those
> checks starts earlier.
> 
> Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
> ---
>  init/main.c |   17 ++++++++++++-----
>  1 file changed, 12 insertions(+), 5 deletions(-)
> 
> --- a/init/main.c
> +++ b/init/main.c
> @@ -389,6 +389,7 @@ static __initdata DECLARE_COMPLETION(kth
>  
>  static noinline void __ref rest_init(void)
>  {
> +	struct task_struct *tsk;
>  	int pid;
>  
>  	rcu_scheduler_starting();
> @@ -397,7 +398,17 @@ static noinline void __ref rest_init(voi
>  	 * the init task will end up wanting to create kthreads, which, if
>  	 * we schedule it before we create kthreadd, will OOPS.
>  	 */
> -	kernel_thread(kernel_init, NULL, CLONE_FS);
> +	pid = kernel_thread(kernel_init, NULL, CLONE_FS);
> +	/*
> +	 * Pin init on the boot cpu. Task migration is not properly working
> +	 * until sched_init_smp() has been run. It will set the allowed
> +	 * cpus for init to the non isolated cpus.
> +	 */
> +	rcu_read_lock();
> +	tsk = find_task_by_pid_ns(pid, &init_pid_ns);

Should we have a:

	BUG_ON(!tsk);

Just to be paranoid?

-- Steve

> +	set_cpus_allowed_ptr(tsk, cpumask_of(smp_processor_id()));
> +	rcu_read_unlock();
> +
>  	numa_default_policy();
>  	pid = kernel_thread(kthreadd, NULL, CLONE_FS | CLONE_FILES);
>  	rcu_read_lock();
> @@ -1015,10 +1026,6 @@ static noinline void __init kernel_init_
>  	 * init can allocate pages on any node
>  	 */
>  	set_mems_allowed(node_states[N_MEMORY]);
> -	/*
> -	 * init can run on any cpu.
> -	 */
> -	set_cpus_allowed_ptr(current, cpu_all_mask);
>  
>  	cad_pid = task_pid(current);
>  
> 

  reply	other threads:[~2017-05-16 19:06 UTC|newest]

Thread overview: 47+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-05-16 18:42 [patch V2 00/17] init: Enable might_sleep() and smp_processor_id() debugging early Thomas Gleixner
2017-05-16 18:42 ` [patch V2 01/17] init: Pin init task to boot cpu initially Thomas Gleixner
2017-05-16 19:06   ` Steven Rostedt [this message]
2017-05-23  8:47   ` [tip:sched/core] init: Pin init task to the boot CPU, initially tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 02/17] arm: Adjust system_state check Thomas Gleixner
2017-05-23  8:48   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 03/17] arm64: " Thomas Gleixner
2017-05-17 10:09   ` Mark Rutland
2017-05-17 14:23   ` Catalin Marinas
2017-05-23  8:48   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 04/17] x86/smp: " Thomas Gleixner
2017-05-23  8:49   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 05/17] metag: " Thomas Gleixner
2017-05-23  8:50   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 06/17] powerpc: " Thomas Gleixner
2017-05-23  8:50   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 07/17] ACPI: " Thomas Gleixner
2017-05-16 19:07   ` Steven Rostedt
2017-05-23  8:51   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 08/17] mm: " Thomas Gleixner
2017-05-23  8:51   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 09/17] cpufreq/pasemi: " Thomas Gleixner
2017-05-23  8:52   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 10/17] iommu/vt-d: Adjust system_state checks Thomas Gleixner
2017-05-23  8:52   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 11/17] iommu/of: Adjust system_state check Thomas Gleixner
2017-05-23  8:53   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 12/17] async: Adjust system_state checks Thomas Gleixner
2017-05-23  8:53   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 13/17] extable: " Thomas Gleixner
2017-05-23  8:54   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 14/17] printk: " Thomas Gleixner
2017-05-23  8:55   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 15/17] mm/vmscan: " Thomas Gleixner
2017-05-17  6:56   ` Vlastimil Babka
2017-05-23  8:55   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 16/17] init: Introduce SYSTEM_SCHEDULING state Thomas Gleixner
2017-05-17 10:58   ` Mark Rutland
2017-05-17 21:15   ` Boris Ostrovsky
2017-05-23  8:56   ` [tip:sched/core] " tip-bot for Thomas Gleixner
2017-05-16 18:42 ` [patch V2 17/17] sched: Enable might_sleep() and smp_processor_id() checks early Thomas Gleixner
2017-05-16 19:12   ` Steven Rostedt
2017-05-16 22:46     ` Thomas Gleixner
2017-05-16 22:50       ` Steven Rostedt
2017-05-17 11:02   ` Mark Rutland
2017-05-23  8:56   ` [tip:sched/core] sched/core: " tip-bot for Thomas Gleixner
2017-05-17 10:58 ` [patch V2 00/17] init: Enable might_sleep() and smp_processor_id() debugging early Mark Rutland

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=20170516150621.548862a2@gandalf.local.home \
    --to=rostedt@goodmis.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mark.rutland@arm.com \
    --cc=mingo@kernel.org \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    /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).