linux-mm.kvack.org archive mirror
 help / color / mirror / Atom feed
From: JP Kobryn <inwardvessel@gmail.com>
To: linux-mm@kvack.org
Cc: apopple@nvidia.com, akpm@linux-foundation.org,
	axelrasmussen@google.com, byungchul@sk.com,
	cgroups@vger.kernel.org, david@kernel.org, eperezma@redhat.com,
	gourry@gourry.net, jasowang@redhat.com, hannes@cmpxchg.org,
	joshua.hahnjy@gmail.com, Liam.Howlett@oracle.com,
	linux-kernel@vger.kernel.org, lorenzo.stoakes@oracle.com,
	matthew.brost@intel.com, mst@redhat.com, mhocko@suse.com,
	rppt@kernel.org, muchun.song@linux.dev,
	zhengqi.arch@bytedance.com, rakie.kim@sk.com,
	roman.gushchin@linux.dev, shakeel.butt@linux.dev,
	surenb@google.com, virtualization@lists.linux.dev,
	vbabka@suse.cz, weixugc@google.com, xuanzhuo@linux.alibaba.com,
	ying.huang@linux.alibaba.com, yuanchu@google.com, ziy@nvidia.com,
	kernel-team@meta.com
Subject: [PATCH 1/2] mm/mempolicy: track page allocations per mempolicy
Date: Wed, 11 Feb 2026 20:51:08 -0800	[thread overview]
Message-ID: <20260212045109.255391-2-inwardvessel@gmail.com> (raw)
In-Reply-To: <20260212045109.255391-1-inwardvessel@gmail.com>

It would be useful to see a breakdown of allocations to understand which
NUMA policies are driving them. For example, when investigating memory
pressure, having policy-specific counts could show that allocations were
bound to the affected node (via MPOL_BIND).

Add per-policy page allocation counters as new node stat items. These
counters can provide correlation between a mempolicy and pressure on a
given node.

Signed-off-by: JP Kobryn <inwardvessel@gmail.com>
Suggested-by: Johannes Weiner <hannes@cmpxchg.org>
---
 include/linux/mmzone.h |  9 +++++++++
 mm/mempolicy.c         | 30 ++++++++++++++++++++++++++++--
 mm/vmstat.c            |  9 +++++++++
 3 files changed, 46 insertions(+), 2 deletions(-)

diff --git a/include/linux/mmzone.h b/include/linux/mmzone.h
index fc5d6c88d2f0..762609d5f0af 100644
--- a/include/linux/mmzone.h
+++ b/include/linux/mmzone.h
@@ -255,6 +255,15 @@ enum node_stat_item {
 	PGDEMOTE_DIRECT,
 	PGDEMOTE_KHUGEPAGED,
 	PGDEMOTE_PROACTIVE,
+#ifdef CONFIG_NUMA
+	PGALLOC_MPOL_DEFAULT,
+	PGALLOC_MPOL_PREFERRED,
+	PGALLOC_MPOL_BIND,
+	PGALLOC_MPOL_INTERLEAVE,
+	PGALLOC_MPOL_LOCAL,
+	PGALLOC_MPOL_PREFERRED_MANY,
+	PGALLOC_MPOL_WEIGHTED_INTERLEAVE,
+#endif
 #ifdef CONFIG_HUGETLB_PAGE
 	NR_HUGETLB,
 #endif
diff --git a/mm/mempolicy.c b/mm/mempolicy.c
index 68a98ba57882..3c64784af761 100644
--- a/mm/mempolicy.c
+++ b/mm/mempolicy.c
@@ -217,6 +217,21 @@ static void reduce_interleave_weights(unsigned int *bw, u8 *new_iw)
 		new_iw[nid] /= iw_gcd;
 }
 
+#define CHECK_MPOL_NODE_STAT_OFFSET(mpol) \
+	BUILD_BUG_ON(PGALLOC_##mpol - mpol != PGALLOC_MPOL_DEFAULT)
+
+static enum node_stat_item mpol_node_stat(unsigned short mode)
+{
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_PREFERRED);
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_BIND);
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_INTERLEAVE);
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_LOCAL);
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_PREFERRED_MANY);
+	CHECK_MPOL_NODE_STAT_OFFSET(MPOL_WEIGHTED_INTERLEAVE);
+
+	return PGALLOC_MPOL_DEFAULT + mode;
+}
+
 int mempolicy_set_node_perf(unsigned int node, struct access_coordinate *coords)
 {
 	struct weighted_interleave_state *new_wi_state, *old_wi_state = NULL;
@@ -2446,8 +2461,14 @@ static struct page *alloc_pages_mpol(gfp_t gfp, unsigned int order,
 
 	nodemask = policy_nodemask(gfp, pol, ilx, &nid);
 
-	if (pol->mode == MPOL_PREFERRED_MANY)
-		return alloc_pages_preferred_many(gfp, order, nid, nodemask);
+	if (pol->mode == MPOL_PREFERRED_MANY) {
+		page = alloc_pages_preferred_many(gfp, order, nid, nodemask);
+		if (page)
+			__mod_node_page_state(page_pgdat(page),
+					mpol_node_stat(MPOL_PREFERRED_MANY), 1 << order);
+
+		return page;
+	}
 
 	if (IS_ENABLED(CONFIG_TRANSPARENT_HUGEPAGE) &&
 	    /* filter "hugepage" allocation, unless from alloc_pages() */
@@ -2472,6 +2493,9 @@ static struct page *alloc_pages_mpol(gfp_t gfp, unsigned int order,
 			page = __alloc_frozen_pages_noprof(
 				gfp | __GFP_THISNODE | __GFP_NORETRY, order,
 				nid, NULL);
+			if (page)
+				__mod_node_page_state(page_pgdat(page),
+						mpol_node_stat(pol->mode), 1 << order);
 			if (page || !(gfp & __GFP_DIRECT_RECLAIM))
 				return page;
 			/*
@@ -2484,6 +2508,8 @@ static struct page *alloc_pages_mpol(gfp_t gfp, unsigned int order,
 	}
 
 	page = __alloc_frozen_pages_noprof(gfp, order, nid, nodemask);
+	if (page)
+		__mod_node_page_state(page_pgdat(page), mpol_node_stat(pol->mode), 1 << order);
 
 	if (unlikely(pol->mode == MPOL_INTERLEAVE ||
 		     pol->mode == MPOL_WEIGHTED_INTERLEAVE) && page) {
diff --git a/mm/vmstat.c b/mm/vmstat.c
index 65de88cdf40e..74e0ddde1e93 100644
--- a/mm/vmstat.c
+++ b/mm/vmstat.c
@@ -1291,6 +1291,15 @@ const char * const vmstat_text[] = {
 	[I(PGDEMOTE_DIRECT)]			= "pgdemote_direct",
 	[I(PGDEMOTE_KHUGEPAGED)]		= "pgdemote_khugepaged",
 	[I(PGDEMOTE_PROACTIVE)]			= "pgdemote_proactive",
+#ifdef CONFIG_NUMA
+	[I(PGALLOC_MPOL_DEFAULT)]		= "pgalloc_mpol_default",
+	[I(PGALLOC_MPOL_PREFERRED)]		= "pgalloc_mpol_preferred",
+	[I(PGALLOC_MPOL_BIND)]			= "pgalloc_mpol_bind",
+	[I(PGALLOC_MPOL_INTERLEAVE)]		= "pgalloc_mpol_interleave",
+	[I(PGALLOC_MPOL_LOCAL)]			= "pgalloc_mpol_local",
+	[I(PGALLOC_MPOL_PREFERRED_MANY)]	= "pgalloc_mpol_preferred_many",
+	[I(PGALLOC_MPOL_WEIGHTED_INTERLEAVE)]	= "pgalloc_mpol_weighted_interleave",
+#endif
 #ifdef CONFIG_HUGETLB_PAGE
 	[I(NR_HUGETLB)]				= "nr_hugetlb",
 #endif
-- 
2.47.3



  reply	other threads:[~2026-02-12  4:51 UTC|newest]

Thread overview: 27+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-02-12  4:51 [PATCH 0/2] improve per-node allocation and reclaim visibility JP Kobryn
2026-02-12  4:51 ` JP Kobryn [this message]
2026-02-12  7:29   ` [PATCH 1/2] mm/mempolicy: track page allocations per mempolicy Michal Hocko
2026-02-12 21:22     ` JP Kobryn
2026-02-16  8:26       ` Michal Hocko
2026-02-16 17:50         ` JP Kobryn (Meta)
2026-02-16 21:07           ` Michal Hocko
2026-02-17  7:48             ` JP Kobryn (Meta)
2026-02-17 12:37               ` Michal Hocko
2026-02-17 18:19                 ` JP Kobryn (Meta)
2026-02-17 18:52                   ` Michal Hocko
2026-02-12 15:07   ` Shakeel Butt
2026-02-12 21:23     ` JP Kobryn
2026-02-12 15:24   ` Vlastimil Babka
2026-02-12 21:25     ` JP Kobryn
2026-02-13  8:54       ` Vlastimil Babka
2026-02-13 19:56         ` JP Kobryn (Meta)
2026-02-18  4:25   ` kernel test robot
2026-02-12  4:51 ` [PATCH 2/2] mm: move pgscan and pgsteal to node stats JP Kobryn
2026-02-12  7:08   ` Michael S. Tsirkin
2026-02-12 21:23     ` JP Kobryn
2026-02-12  7:29   ` Michal Hocko
2026-02-12 21:20     ` JP Kobryn
2026-02-12  4:57 ` [PATCH 0/2] improve per-node allocation and reclaim visibility Matthew Wilcox
2026-02-12 21:22   ` JP Kobryn
2026-02-12 21:53     ` Matthew Wilcox
2026-02-12 18:08 ` [syzbot ci] " syzbot ci

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=20260212045109.255391-2-inwardvessel@gmail.com \
    --to=inwardvessel@gmail.com \
    --cc=Liam.Howlett@oracle.com \
    --cc=akpm@linux-foundation.org \
    --cc=apopple@nvidia.com \
    --cc=axelrasmussen@google.com \
    --cc=byungchul@sk.com \
    --cc=cgroups@vger.kernel.org \
    --cc=david@kernel.org \
    --cc=eperezma@redhat.com \
    --cc=gourry@gourry.net \
    --cc=hannes@cmpxchg.org \
    --cc=jasowang@redhat.com \
    --cc=joshua.hahnjy@gmail.com \
    --cc=kernel-team@meta.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mm@kvack.org \
    --cc=lorenzo.stoakes@oracle.com \
    --cc=matthew.brost@intel.com \
    --cc=mhocko@suse.com \
    --cc=mst@redhat.com \
    --cc=muchun.song@linux.dev \
    --cc=rakie.kim@sk.com \
    --cc=roman.gushchin@linux.dev \
    --cc=rppt@kernel.org \
    --cc=shakeel.butt@linux.dev \
    --cc=surenb@google.com \
    --cc=vbabka@suse.cz \
    --cc=virtualization@lists.linux.dev \
    --cc=weixugc@google.com \
    --cc=xuanzhuo@linux.alibaba.com \
    --cc=ying.huang@linux.alibaba.com \
    --cc=yuanchu@google.com \
    --cc=zhengqi.arch@bytedance.com \
    --cc=ziy@nvidia.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