linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: "Zhijian Li (Fujitsu)" <lizhijian@fujitsu.com>
To: "Huang, Ying" <ying.huang@intel.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Greg Kroah-Hartman <gregkh@linuxfoundation.org>,
	"rafael@kernel.org" <rafael@kernel.org>,
	"linux-mm@kvack.org" <linux-mm@kvack.org>,
	"Yasunori Gotou (Fujitsu)" <y-goto@fujitsu.com>,
	"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH RFC 3/4] mm/vmstat: rename pgdemote_* to pgdemote_dst_* and add pgdemote_src_*
Date: Thu, 2 Nov 2023 06:34:42 +0000	[thread overview]
Message-ID: <fbca99e1-40c4-4ffd-a0a1-89728dd0b900@fujitsu.com> (raw)
In-Reply-To: <87r0l81zfd.fsf@yhuang6-desk2.ccr.corp.intel.com>



On 02/11/2023 13:45, Huang, Ying wrote:
> Li Zhijian <lizhijian@fujitsu.com> writes:
> 
>> pgdemote_src_*: pages demoted from this node.
>> pgdemote_dst_*: pages demoted to this node.
>>
>> So that we are able to know their demotion per-node stats by checking this.
>>
>> In the environment, node0 and node1 are DRAM, node3 is PMEM.
>>
>> Global stats:
>> $ grep -E 'demote' /proc/vmstat
>> pgdemote_src_kswapd 130155
>> pgdemote_src_direct 113497
>> pgdemote_src_khugepaged 0
>> pgdemote_dst_kswapd 130155
>> pgdemote_dst_direct 113497
>> pgdemote_dst_khugepaged 0
>>
>> Per-node stats:
>> $ grep demote /sys/devices/system/node/node0/vmstat
>> pgdemote_src_kswapd 68454
>> pgdemote_src_direct 83431
>> pgdemote_src_khugepaged 0
>> pgdemote_dst_kswapd 0
>> pgdemote_dst_direct 0
>> pgdemote_dst_khugepaged 0
>>
>> $ grep demote /sys/devices/system/node/node1/vmstat
>> pgdemote_src_kswapd 185834
>> pgdemote_src_direct 30066
>> pgdemote_src_khugepaged 0
>> pgdemote_dst_kswapd 0
>> pgdemote_dst_direct 0
>> pgdemote_dst_khugepaged 0
>>
>> $ grep demote /sys/devices/system/node/node3/vmstat
>> pgdemote_src_kswapd 0
>> pgdemote_src_direct 0
>> pgdemote_src_khugepaged 0
>> pgdemote_dst_kswapd 254288
>> pgdemote_dst_direct 113497
>> pgdemote_dst_khugepaged 0
>>
>>  From above stats, we know node3 is the demotion destination which one
>> the node0 and node1 will demote to.
> 
> Why do we need these information?  Do you have some use case?

I recall our customers have mentioned that they want to know how much the memory is demoted
to the CXL memory device in a specific period.


>>>   	mod_node_page_state(NODE_DATA(target_nid),
>>> -		    PGDEMOTE_KSWAPD + reclaimer_offset(), nr_succeeded);
>>> +		    PGDEMOTE_DST_KSWAPD + reclaimer_offset(), nr_succeeded);

But if the *target_nid* is only indicate the preferred node, this accounting maybe not accurate.


Thanks
Zhijian

> 
> --
> Best Regards,
> Huang, Ying
> 
>> Signed-off-by: Li Zhijian <lizhijian@fujitsu.com>
>> ---
>> RFC: their names are open to discussion, maybe pgdemote_from/to_*
>> Another defect of this patch is that, SUM(pgdemote_src_*) is always same
>> as SUM(pgdemote_dst_*) in the global stats, shall we hide one of them.
>> ---
>>   include/linux/mmzone.h |  9 ++++++---
>>   mm/vmscan.c            | 13 ++++++++++---
>>   mm/vmstat.c            |  9 ++++++---
>>   3 files changed, 22 insertions(+), 9 deletions(-)
>>
>> diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
>> index ad0309eea850..a6140d894bec 100644
>> --- a/include/linux/mmzone.h
>> +++ b/include/linux/mmzone.h
>> @@ -207,9 +207,12 @@ enum node_stat_item {
>>   	PGPROMOTE_SUCCESS,	/* promote successfully */
>>   	PGPROMOTE_CANDIDATE,	/* candidate pages to promote */
>>   	/* PGDEMOTE_*: pages demoted */
>> -	PGDEMOTE_KSWAPD,
>> -	PGDEMOTE_DIRECT,
>> -	PGDEMOTE_KHUGEPAGED,
>> +	PGDEMOTE_SRC_KSWAPD,
>> +	PGDEMOTE_SRC_DIRECT,
>> +	PGDEMOTE_SRC_KHUGEPAGED,
>> +	PGDEMOTE_DST_KSWAPD,
>> +	PGDEMOTE_DST_DIRECT,
>> +	PGDEMOTE_DST_KHUGEPAGED,
>>   #endif
>>   	NR_VM_NODE_STAT_ITEMS
>>   };
>> diff --git a/mm/vmscan.c b/mm/vmscan.c
>> index 2f1fb4ec3235..55d2287d7150 100644
>> --- a/mm/vmscan.c
>> +++ b/mm/vmscan.c
>> @@ -1111,13 +1111,18 @@ void drop_slab(void)
>>   static int reclaimer_offset(void)
>>   {
>>   	BUILD_BUG_ON(PGSTEAL_DIRECT - PGSTEAL_KSWAPD !=
>> -			PGDEMOTE_DIRECT - PGDEMOTE_KSWAPD);
>> +			PGDEMOTE_SRC_DIRECT - PGDEMOTE_SRC_KSWAPD);
>>   	BUILD_BUG_ON(PGSTEAL_DIRECT - PGSTEAL_KSWAPD !=
>>   			PGSCAN_DIRECT - PGSCAN_KSWAPD);
>>   	BUILD_BUG_ON(PGSTEAL_KHUGEPAGED - PGSTEAL_KSWAPD !=
>> -			PGDEMOTE_KHUGEPAGED - PGDEMOTE_KSWAPD);
>> +			PGDEMOTE_SRC_KHUGEPAGED - PGDEMOTE_SRC_KSWAPD);
>>   	BUILD_BUG_ON(PGSTEAL_KHUGEPAGED - PGSTEAL_KSWAPD !=
>>   			PGSCAN_KHUGEPAGED - PGSCAN_KSWAPD);
>> +	BUILD_BUG_ON(PGDEMOTE_SRC_DIRECT - PGDEMOTE_SRC_KSWAPD !=
>> +			PGDEMOTE_DST_DIRECT - PGDEMOTE_DST_KSWAPD);
>> +	BUILD_BUG_ON(PGDEMOTE_SRC_KHUGEPAGED - PGDEMOTE_SRC_KSWAPD !=
>> +			PGDEMOTE_DST_KHUGEPAGED - PGDEMOTE_DST_KSWAPD);
>> +
>>   
>>   	if (current_is_kswapd())
>>   		return 0;
>> @@ -1678,8 +1683,10 @@ static unsigned int demote_folio_list(struct list_head *demote_folios,
>>   		      (unsigned long)&mtc, MIGRATE_ASYNC, MR_DEMOTION,
>>   		      &nr_succeeded);
>>   
>> +	mod_node_page_state(pgdat,
>> +		    PGDEMOTE_SRC_KSWAPD + reclaimer_offset(), nr_succeeded);
>>   	mod_node_page_state(NODE_DATA(target_nid),
>> -		    PGDEMOTE_KSWAPD + reclaimer_offset(), nr_succeeded);
>> +		    PGDEMOTE_DST_KSWAPD + reclaimer_offset(), nr_succeeded);
>>   
>>   	return nr_succeeded;
>>   }
>> diff --git a/mm/vmstat.c b/mm/vmstat.c
>> index f141c48c39e4..63f106a5e008 100644
>> --- a/mm/vmstat.c
>> +++ b/mm/vmstat.c
>> @@ -1244,9 +1244,12 @@ const char * const vmstat_text[] = {
>>   #ifdef CONFIG_NUMA_BALANCING
>>   	"pgpromote_success",
>>   	"pgpromote_candidate",
>> -	"pgdemote_kswapd",
>> -	"pgdemote_direct",
>> -	"pgdemote_khugepaged",
>> +	"pgdemote_src_kswapd",
>> +	"pgdemote_src_direct",
>> +	"pgdemote_src_khugepaged",
>> +	"pgdemote_dst_kswapd",
>> +	"pgdemote_dst_direct",
>> +	"pgdemote_dst_khugepaged",
>>   #endif
>>   
>>   	/* enum writeback_stat_item counters */

  reply	other threads:[~2023-11-02  6:34 UTC|newest]

Thread overview: 32+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2023-11-02  2:56 Subject: [PATCH RFC 0/4] Demotion Profiling Improvements Li Zhijian
2023-11-02  2:56 ` [PATCH RFC 1/4] drivers/base/node: Add demotion_nodes sys infterface Li Zhijian
2023-11-02  3:17   ` Huang, Ying
2023-11-02  3:39     ` Zhijian Li (Fujitsu)
2023-11-02  5:18       ` Huang, Ying
2023-11-02  5:54         ` Zhijian Li (Fujitsu)
2023-11-02  5:58           ` Huang, Ying
2023-11-03  3:05             ` Zhijian Li (Fujitsu)
2024-01-30  8:53     ` Li Zhijian
2024-01-31  1:13       ` Huang, Ying
2024-01-31  3:18         ` Zhijian Li (Fujitsu)
2024-02-02  7:43           ` Zhijian Li (Fujitsu)
2024-02-02  8:19             ` Huang, Ying
2024-02-05  7:31               ` Zhijian Li (Fujitsu)
2024-01-31  6:23         ` Yasunori Gotou (Fujitsu)
2024-01-31  6:52           ` Huang, Ying
2023-11-02  2:56 ` [PATCH RFC 2/4] mm/vmstat: Move pgdemote_* to per-node stats Li Zhijian
2023-11-02  4:56   ` Huang, Ying
2023-11-02  5:43   ` Huang, Ying
2023-11-02  5:57     ` Zhijian Li (Fujitsu)
2023-11-02  2:56 ` [PATCH RFC 3/4] mm/vmstat: rename pgdemote_* to pgdemote_dst_* and add pgdemote_src_* Li Zhijian
2023-11-02  5:45   ` Huang, Ying
2023-11-02  6:34     ` Zhijian Li (Fujitsu) [this message]
2023-11-02  6:56       ` Huang, Ying
2023-11-02  7:38       ` Yasunori Gotou (Fujitsu)
2023-11-02  7:46         ` Huang, Ying
2023-11-02  9:45           ` Yasunori Gotou (Fujitsu)
2023-11-03  6:14             ` Huang, Ying
2023-11-06  5:02               ` Yasunori Gotou (Fujitsu)
2023-11-02  2:56 ` [PATCH RFC 4/4] drivers/base/node: add demote_src and demote_dst to numastat Li Zhijian
2023-11-02  5:40   ` Greg Kroah-Hartman
2023-11-02  8:15     ` Zhijian Li (Fujitsu)

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=fbca99e1-40c4-4ffd-a0a1-89728dd0b900@fujitsu.com \
    --to=lizhijian@fujitsu.com \
    --cc=akpm@linux-foundation.org \
    --cc=gregkh@linuxfoundation.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=rafael@kernel.org \
    --cc=y-goto@fujitsu.com \
    --cc=ying.huang@intel.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