linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Vlastimil Babka <vbabka@suse.cz>
To: Charan Teja Reddy <charante@codeaurora.org>,
	akpm@linux-foundation.org, mcgrof@kernel.org,
	keescook@chromium.org, yzaikin@google.com, nigupta@nvidia.com,
	bhe@redhat.com, mateusznosek0@gmail.com, sh_def@163.com,
	iamjoonsoo.kim@lge.com, vinmenon@codeaurora.org
Cc: linux-kernel@vger.kernel.org, linux-mm@kvack.org,
	linux-fsdevel@vger.kernel.org,
	Linux API <linux-api@vger.kernel.org>
Subject: Re: [PATCH V2] mm: compaction: support triggering of proactive compaction by user
Date: Fri, 28 May 2021 17:19:16 +0200	[thread overview]
Message-ID: <a29f0cf6-d007-ea17-25b7-642168b6efdd@suse.cz> (raw)
In-Reply-To: <1621345058-26676-1-git-send-email-charante@codeaurora.org>

+CC linux-api

On 5/18/21 3:37 PM, Charan Teja Reddy wrote:
> The proactive compaction[1] gets triggered for every 500msec and run
> compaction on the node for COMPACTION_HPAGE_ORDER (usually order-9)
> pages based on the value set to sysctl.compaction_proactiveness.
> Triggering the compaction for every 500msec in search of
> COMPACTION_HPAGE_ORDER pages is not needed for all applications,
> especially on the embedded system usecases which may have few MB's of
> RAM. Enabling the proactive compaction in its state will endup in
> running almost always on such systems.
> 
> Other side, proactive compaction can still be very much useful for
> getting a set of higher order pages in some controllable
> manner(controlled by using the sysctl.compaction_proactiveness). Thus on
> systems where enabling the proactive compaction always may proove not
> required, can trigger the same from user space on write to its sysctl
> interface. As an example, say app launcher decide to launch the memory
> heavy application which can be launched fast if it gets more higher
> order pages thus launcher can prepare the system in advance by
> triggering the proactive compaction from userspace.
> 
> This triggering of proactive compaction is done on a write to
> sysctl.compaction_proactiveness by user.
> 
> [1]https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit?id=facdaa917c4d5a376d09d25865f5a863f906234a
> 
> Signed-off-by: Charan Teja Reddy <charante@codeaurora.org>

Cancelling all current sleeps immediately when the controlling variable changes
doesn't sound wrong to me.
A question below:

> ---
> changes in V2: 
>     - remove /proc interface trigger for proactive compaction
>     - Intention is same that add a way to trigger proactive compaction by user.
> 
> changes in V1:
>     -  https://lore.kernel.org/lkml/1619098678-8501-1-git-send-email-charante@codeaurora.org/
> 
>  include/linux/compaction.h |  2 ++
>  include/linux/mmzone.h     |  1 +
>  kernel/sysctl.c            |  2 +-
>  mm/compaction.c            | 35 ++++++++++++++++++++++++++++++++---
>  4 files changed, 36 insertions(+), 4 deletions(-)
> 
> diff --git a/include/linux/compaction.h b/include/linux/compaction.h
> index 4221888..04d5d9f 100644
> --- a/include/linux/compaction.h
> +++ b/include/linux/compaction.h
> @@ -84,6 +84,8 @@ static inline unsigned long compact_gap(unsigned int order)
>  extern unsigned int sysctl_compaction_proactiveness;
>  extern int sysctl_compaction_handler(struct ctl_table *table, int write,
>  			void *buffer, size_t *length, loff_t *ppos);
> +extern int compaction_proactiveness_sysctl_handler(struct ctl_table *table,
> +		int write, void *buffer, size_t *length, loff_t *ppos);
>  extern int sysctl_extfrag_threshold;
>  extern int sysctl_compact_unevictable_allowed;
>  
> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
> index 0d53eba..9455809 100644
> --- a/include/linux/mmzone.h
> +++ b/include/linux/mmzone.h
> @@ -815,6 +815,7 @@ typedef struct pglist_data {
>  	enum zone_type kcompactd_highest_zoneidx;
>  	wait_queue_head_t kcompactd_wait;
>  	struct task_struct *kcompactd;
> +	bool proactive_compact_trigger;
>  #endif
>  	/*
>  	 * This is a per-node reserve of pages that are not available
> diff --git a/kernel/sysctl.c b/kernel/sysctl.c
> index 14edf84..bed2fad 100644
> --- a/kernel/sysctl.c
> +++ b/kernel/sysctl.c
> @@ -2840,7 +2840,7 @@ static struct ctl_table vm_table[] = {
>  		.data		= &sysctl_compaction_proactiveness,
>  		.maxlen		= sizeof(sysctl_compaction_proactiveness),
>  		.mode		= 0644,
> -		.proc_handler	= proc_dointvec_minmax,
> +		.proc_handler	= compaction_proactiveness_sysctl_handler,
>  		.extra1		= SYSCTL_ZERO,
>  		.extra2		= &one_hundred,
>  	},
> diff --git a/mm/compaction.c b/mm/compaction.c
> index 84fde27..9056693 100644
> --- a/mm/compaction.c
> +++ b/mm/compaction.c
> @@ -2708,6 +2708,30 @@ static void compact_nodes(void)
>   */
>  unsigned int __read_mostly sysctl_compaction_proactiveness = 20;
>  
> +int compaction_proactiveness_sysctl_handler(struct ctl_table *table, int write,
> +		void *buffer, size_t *length, loff_t *ppos)
> +{
> +	int rc, nid;
> +
> +	rc = proc_dointvec_minmax(table, write, buffer, length, ppos);
> +	if (rc)
> +		return rc;
> +
> +	if (write && sysctl_compaction_proactiveness) {
> +		for_each_online_node(nid) {
> +			pg_data_t *pgdat = NODE_DATA(nid);
> +
> +			if (pgdat->proactive_compact_trigger)
> +				continue;
> +
> +			pgdat->proactive_compact_trigger = true;
> +			wake_up_interruptible(&pgdat->kcompactd_wait);
> +		}
> +	}
> +
> +	return 0;
> +}
> +
>  /*
>   * This is the entry point for compacting all nodes via
>   * /proc/sys/vm/compact_memory
> @@ -2752,7 +2776,8 @@ void compaction_unregister_node(struct node *node)
>  
>  static inline bool kcompactd_work_requested(pg_data_t *pgdat)
>  {
> -	return pgdat->kcompactd_max_order > 0 || kthread_should_stop();
> +	return pgdat->kcompactd_max_order > 0 || kthread_should_stop() ||
> +		pgdat->proactive_compact_trigger;
>  }
>  
>  static bool kcompactd_node_suitable(pg_data_t *pgdat)
> @@ -2905,7 +2930,8 @@ static int kcompactd(void *p)
>  		trace_mm_compaction_kcompactd_sleep(pgdat->node_id);
>  		if (wait_event_freezable_timeout(pgdat->kcompactd_wait,
>  			kcompactd_work_requested(pgdat),
> -			msecs_to_jiffies(HPAGE_FRAG_CHECK_INTERVAL_MSEC))) {
> +			msecs_to_jiffies(HPAGE_FRAG_CHECK_INTERVAL_MSEC)) &&
> +			!pgdat->proactive_compact_trigger) {
>  
>  			psi_memstall_enter(&pflags);
>  			kcompactd_do_work(pgdat);
> @@ -2919,7 +2945,7 @@ static int kcompactd(void *p)
>  
>  			if (proactive_defer) {
>  				proactive_defer--;
> -				continue;
> +				goto loop;

I don't understand this part. If we kick kcompactd from the sysctl handler
because we are changing proactiveness, shouldn't we also discard any accumulated
defer score?

>  			}
>  			prev_score = fragmentation_score_node(pgdat);
>  			proactive_compact_node(pgdat);
> @@ -2931,6 +2957,9 @@ static int kcompactd(void *p)
>  			proactive_defer = score < prev_score ?
>  					0 : 1 << COMPACT_MAX_DEFER_SHIFT;
>  		}
> +loop:
> +		if (pgdat->proactive_compact_trigger)
> +			pgdat->proactive_compact_trigger = false;
>  	}
>  
>  	return 0;
> 


  parent reply	other threads:[~2021-05-28 15:19 UTC|newest]

Thread overview: 10+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2021-05-18 13:37 [PATCH V2] mm: compaction: support triggering of proactive compaction by user Charan Teja Reddy
2021-05-19  1:41 ` 答复: " Chu,Kaiping
2021-05-19 10:19   ` Charan Teja Kalla
2021-05-25 14:20 ` Charan Teja Kalla
2021-05-25 20:35 ` Nitin Gupta
2021-05-27  9:28   ` Charan Teja Kalla
2021-05-27 23:52     ` Nitin Gupta
2021-05-29  2:57       ` Charan Teja Kalla
2021-05-28 15:19 ` Vlastimil Babka [this message]
2021-05-28 17:02   ` Charan Teja Kalla

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=a29f0cf6-d007-ea17-25b7-642168b6efdd@suse.cz \
    --to=vbabka@suse.cz \
    --cc=akpm@linux-foundation.org \
    --cc=bhe@redhat.com \
    --cc=charante@codeaurora.org \
    --cc=iamjoonsoo.kim@lge.com \
    --cc=keescook@chromium.org \
    --cc=linux-api@vger.kernel.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=mateusznosek0@gmail.com \
    --cc=mcgrof@kernel.org \
    --cc=nigupta@nvidia.com \
    --cc=sh_def@163.com \
    --cc=vinmenon@codeaurora.org \
    --cc=yzaikin@google.com \
    /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).