linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: Michal Hocko <mhocko@suse.com>
To: Frederic Weisbecker <frederic@kernel.org>
Cc: LKML <linux-kernel@vger.kernel.org>,
	Vlastimil Babka <vbabka@suse.cz>,
	Andrew Morton <akpm@linux-foundation.org>,
	linux-mm@kvack.org, Peter Zijlstra <peterz@infradead.org>,
	Thomas Gleixner <tglx@linutronix.de>
Subject: Re: [PATCH 13/19] mm: Create/affine kcompactd to its preferred node
Date: Tue, 17 Sep 2024 08:04:44 +0200	[thread overview]
Message-ID: <Zukb_FEVRllBDhKO@tiehlicka> (raw)
In-Reply-To: <20240916224925.20540-14-frederic@kernel.org>

On Tue 17-09-24 00:49:17, Frederic Weisbecker wrote:
> Kcompactd is dedicated to a specific node. As such it wants to be
> preferrably affine to it, memory and CPUs-wise.
> 
> Use the proper kthread API to achieve that. As a bonus it takes care of
> CPU-hotplug events and CPU-isolation on its behalf.
> 
> Acked-by: Vlastimil Babka <vbabka@suse.cz>
> Signed-off-by: Frederic Weisbecker <frederic@kernel.org>

Acked-by: Michal Hocko <mhocko@suse.com>
Clear simplification, thanks!

> ---
>  mm/compaction.c | 43 +++----------------------------------------
>  1 file changed, 3 insertions(+), 40 deletions(-)
> 
> diff --git a/mm/compaction.c b/mm/compaction.c
> index eb95e9b435d0..69742555f2e5 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -3179,15 +3179,9 @@ void wakeup_kcompactd(pg_data_t *pgdat, int order, int highest_zoneidx)
>  static int kcompactd(void *p)
>  {
>  	pg_data_t *pgdat = (pg_data_t *)p;
> -	struct task_struct *tsk = current;
>  	long default_timeout = msecs_to_jiffies(HPAGE_FRAG_CHECK_INTERVAL_MSEC);
>  	long timeout = default_timeout;
>  
> -	const struct cpumask *cpumask = cpumask_of_node(pgdat->node_id);
> -
> -	if (!cpumask_empty(cpumask))
> -		set_cpus_allowed_ptr(tsk, cpumask);
> -
>  	set_freezable();
>  
>  	pgdat->kcompactd_max_order = 0;
> @@ -3258,10 +3252,12 @@ void __meminit kcompactd_run(int nid)
>  	if (pgdat->kcompactd)
>  		return;
>  
> -	pgdat->kcompactd = kthread_run(kcompactd, pgdat, "kcompactd%d", nid);
> +	pgdat->kcompactd = kthread_create_on_node(kcompactd, pgdat, nid, "kcompactd%d", nid);
>  	if (IS_ERR(pgdat->kcompactd)) {
>  		pr_err("Failed to start kcompactd on node %d\n", nid);
>  		pgdat->kcompactd = NULL;
> +	} else {
> +		wake_up_process(pgdat->kcompactd);
>  	}
>  }
>  
> @@ -3279,30 +3275,6 @@ void __meminit kcompactd_stop(int nid)
>  	}
>  }
>  
> -/*
> - * It's optimal to keep kcompactd on the same CPUs as their memory, but
> - * not required for correctness. So if the last cpu in a node goes
> - * away, we get changed to run anywhere: as the first one comes back,
> - * restore their cpu bindings.
> - */
> -static int kcompactd_cpu_online(unsigned int cpu)
> -{
> -	int nid;
> -
> -	for_each_node_state(nid, N_MEMORY) {
> -		pg_data_t *pgdat = NODE_DATA(nid);
> -		const struct cpumask *mask;
> -
> -		mask = cpumask_of_node(pgdat->node_id);
> -
> -		if (cpumask_any_and(cpu_online_mask, mask) < nr_cpu_ids)
> -			/* One of our CPUs online: restore mask */
> -			if (pgdat->kcompactd)
> -				set_cpus_allowed_ptr(pgdat->kcompactd, mask);
> -	}
> -	return 0;
> -}
> -
>  static int proc_dointvec_minmax_warn_RT_change(const struct ctl_table *table,
>  		int write, void *buffer, size_t *lenp, loff_t *ppos)
>  {
> @@ -3362,15 +3334,6 @@ static struct ctl_table vm_compaction[] = {
>  static int __init kcompactd_init(void)
>  {
>  	int nid;
> -	int ret;
> -
> -	ret = cpuhp_setup_state_nocalls(CPUHP_AP_ONLINE_DYN,
> -					"mm/compaction:online",
> -					kcompactd_cpu_online, NULL);
> -	if (ret < 0) {
> -		pr_err("kcompactd: failed to register hotplug callbacks.\n");
> -		return ret;
> -	}
>  
>  	for_each_node_state(nid, N_MEMORY)
>  		kcompactd_run(nid);
> -- 
> 2.46.0

-- 
Michal Hocko
SUSE Labs


  reply	other threads:[~2024-09-17  6:04 UTC|newest]

Thread overview: 16+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <20240916224925.20540-1-frederic@kernel.org>
2024-09-16 22:49 ` [PATCH 11/19] kthread: Make sure kthread hasn't started while binding it Frederic Weisbecker
2024-09-16 22:49 ` [PATCH 12/19] kthread: Default affine kthread to its preferred NUMA node Frederic Weisbecker
2024-09-17  6:26   ` Michal Hocko
2024-09-17  7:01     ` Vlastimil Babka
2024-09-17  7:05       ` Michal Hocko
2024-09-17  7:14         ` Vlastimil Babka
2024-09-17 10:34     ` Frederic Weisbecker
2024-09-17 11:07       ` Michal Hocko
2024-09-18  9:37         ` Frederic Weisbecker
2024-09-18 11:17           ` Michal Hocko
2024-09-16 22:49 ` [PATCH 13/19] mm: Create/affine kcompactd to its preferred node Frederic Weisbecker
2024-09-17  6:04   ` Michal Hocko [this message]
2024-09-16 22:49 ` [PATCH 14/19] mm: Create/affine kswapd " Frederic Weisbecker
2024-09-17  6:05   ` Michal Hocko
2024-09-16 22:49 ` [PATCH 15/19] kthread: Implement preferred affinity Frederic Weisbecker
     [not found] <20241211154035.75565-1-frederic@kernel.org>
2024-12-11 15:40 ` [PATCH 13/19] mm: Create/affine kcompactd to its preferred node Frederic Weisbecker

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=Zukb_FEVRllBDhKO@tiehlicka \
    --to=mhocko@suse.com \
    --cc=akpm@linux-foundation.org \
    --cc=frederic@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=peterz@infradead.org \
    --cc=tglx@linutronix.de \
    --cc=vbabka@suse.cz \
    /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