From: Jinmeng Zhou <jinmengzhou22@gmail.com>
To: Muchun Song <muchun.song@linux.dev>,
Oscar Salvador <osalvador@suse.de>,
David Hildenbrand <david@kernel.org>,
Johannes Weiner <hannes@cmpxchg.org>,
Michal Hocko <mhocko@kernel.org>,
Roman Gushchin <roman.gushchin@linux.dev>,
Shakeel Butt <shakeel.butt@linux.dev>,
Andrew Morton <akpm@linux-foundation.org>,
Nhat Pham <nphamcs@gmail.com>
Cc: linux-mm@kvack.org, linux-kernel@vger.kernel.org,
cgroups@vger.kernel.org, Jinmeng Zhou <zhoujinmeng@bytedance.com>,
stable@vger.kernel.org
Subject: [PATCH] mm/hugetlb: charge folios to the target mm's memcg
Date: Thu, 3 Sep 2026 15:50:48 +0800 [thread overview]
Message-ID: <20260903075048.3316-1-zhoujinmeng@bytedance.com> (raw)
HugeTLB folios are currently charged to the memcg of the allocating
task. This gives the wrong result when a userfaultfd handler populates a
HugeTLB VMA that belongs to another process. The UFFDIO_COPY ioctl
operates on the userfaultfd context's mm, but get_mem_cgroup_from_current()
charges the folio to the handler's memcg instead.
This can be reproduced by placing the faulting process and its userfaultfd
handler in different memory cgroups. Have the target process register a
HugeTLB mapping with userfaultfd, trigger a missing fault, and let the
handler resolve it with UFFDIO_COPY. The hugepage usage is then reported
in the handler's memory.current instead of the target's.
The generic userfaultfd population path avoids this problem by charging
folios to dst_vma->vm_mm.
Pass the target mm through hugetlb_alloc_folio() and charge the folio by
using get_mem_cgroup_from_mm(). This preserves the existing charge timing
and error handling while making HugeTLB userfaultfd population consistent
with the generic path.
Fixes: 8cba9576df60 ("hugetlb: memcg: account hugetlb-backed memory in memory controller")
Cc: stable@vger.kernel.org
Signed-off-by: Jinmeng Zhou <zhoujinmeng@bytedance.com>
---
include/linux/hugetlb.h | 3 ++-
include/linux/memcontrol.h | 8 +++++---
mm/hugetlb.c | 9 ++++++---
mm/memcontrol.c | 6 ++++--
4 files changed, 17 insertions(+), 9 deletions(-)
diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 16c4c4caa126..45ada75dc04e 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -699,7 +699,8 @@ enum hugetlb_alloc_flag {
#define HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS BIT(HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS_BIT)
struct folio *hugetlb_alloc_folio(struct hstate *h,
- struct mempolicy_interpreted *mpoli, u8 alloc_flags);
+ struct mempolicy_interpreted *mpoli, struct mm_struct *mm,
+ u8 alloc_flags);
struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
unsigned long addr, bool cow_from_owner);
struct folio *alloc_hugetlb_folio_nodemask(struct hstate *h, int preferred_nid,
diff --git a/include/linux/memcontrol.h b/include/linux/memcontrol.h
index 7d1c0ce189a8..362af58e50a4 100644
--- a/include/linux/memcontrol.h
+++ b/include/linux/memcontrol.h
@@ -662,7 +662,8 @@ static inline int mem_cgroup_charge(struct folio *folio, struct mm_struct *mm,
return __mem_cgroup_charge(folio, mm, gfp);
}
-int mem_cgroup_charge_hugetlb(struct folio* folio, gfp_t gfp);
+int mem_cgroup_charge_hugetlb(struct folio *folio, struct mm_struct *mm,
+ gfp_t gfp);
int mem_cgroup_swapin_charge_folio(struct folio *folio, unsigned short id,
struct mm_struct *mm, gfp_t gfp);
@@ -1156,9 +1157,10 @@ static inline int mem_cgroup_charge(struct folio *folio,
return 0;
}
-static inline int mem_cgroup_charge_hugetlb(struct folio* folio, gfp_t gfp)
+static inline int mem_cgroup_charge_hugetlb(struct folio *folio,
+ struct mm_struct *mm, gfp_t gfp)
{
- return 0;
+ return 0;
}
static inline int mem_cgroup_swapin_charge_folio(struct folio *folio,
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 785772845795..5ab5a5141574 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -2816,6 +2816,7 @@ void wait_for_freed_hugetlb_folios(void)
* hugetlb_alloc_folio - Allocate a hugetlb folio.
* @h: Hugetlb state control block.
* @mpoli: Interpreted memory policy to use for allocation.
+ * @mm: Memory descriptor of the allocation target.
* @alloc_flags: Flags controlling the allocation behavior.
*
* Allocates a hugetlb folio and handles cgroup charging and global hstate
@@ -2826,7 +2827,8 @@ void wait_for_freed_hugetlb_folios(void)
* -ENOMEM if mem cgroup charging fails.
*/
struct folio *hugetlb_alloc_folio(struct hstate *h,
- struct mempolicy_interpreted *mpoli, u8 alloc_flags)
+ struct mempolicy_interpreted *mpoli, struct mm_struct *mm,
+ u8 alloc_flags)
{
bool charge_hugetlb_cgroup_rsvd = alloc_flags &
HUGETLB_ALLOC_CHARG_CGROUP_RSVD;
@@ -2881,7 +2883,8 @@ struct folio *hugetlb_alloc_folio(struct hstate *h,
spin_unlock_irq(&hugetlb_lock);
- ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
+ ret = mem_cgroup_charge_hugetlb(folio, mm,
+ gfp | __GFP_RETRY_MAYFAIL);
/*
* Unconditionally increment NR_HUGETLB here because if
* mem_cgroup_charge_hugetlb failed, freeing the page will
@@ -3020,7 +3023,7 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
.nodemask = nodemask,
};
- folio = hugetlb_alloc_folio(h, &mpoli, alloc_flags);
+ folio = hugetlb_alloc_folio(h, &mpoli, vma->vm_mm, alloc_flags);
mpol_cond_put(mpol);
diff --git a/mm/memcontrol.c b/mm/memcontrol.c
index 1271d390b617..0b795bf1e6cf 100644
--- a/mm/memcontrol.c
+++ b/mm/memcontrol.c
@@ -5233,6 +5233,7 @@ int __mem_cgroup_charge(struct folio *folio, struct mm_struct *mm, gfp_t gfp)
/**
* mem_cgroup_charge_hugetlb - charge the memcg for a hugetlb folio
* @folio: folio being charged
+ * @mm: mm context of the allocation target
* @gfp: reclaim mode
*
* This function is called when allocating a huge page folio, after the page has
@@ -5242,9 +5243,10 @@ int __mem_cgroup_charge(struct folio *folio, struct mm_struct *mm, gfp_t gfp)
* Returns ENOMEM if the memcg is already full.
* Returns 0 if either the charge was successful, or if we skip the charging.
*/
-int mem_cgroup_charge_hugetlb(struct folio *folio, gfp_t gfp)
+int mem_cgroup_charge_hugetlb(struct folio *folio, struct mm_struct *mm,
+ gfp_t gfp)
{
- struct mem_cgroup *memcg = get_mem_cgroup_from_current();
+ struct mem_cgroup *memcg = get_mem_cgroup_from_mm(mm);
int ret = 0;
/*
--
2.39.5
next reply other threads:[~2026-09-03 7:51 UTC|newest]
Thread overview: 3+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-09-03 7:50 Jinmeng Zhou [this message]
2026-09-03 11:09 ` [PATCH] mm/hugetlb: charge folios to the target mm's memcg Muchun Song
2026-09-04 2:27 ` Hongfu Li
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=20260903075048.3316-1-zhoujinmeng@bytedance.com \
--to=jinmengzhou22@gmail.com \
--cc=akpm@linux-foundation.org \
--cc=cgroups@vger.kernel.org \
--cc=david@kernel.org \
--cc=hannes@cmpxchg.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mm@kvack.org \
--cc=mhocko@kernel.org \
--cc=muchun.song@linux.dev \
--cc=nphamcs@gmail.com \
--cc=osalvador@suse.de \
--cc=roman.gushchin@linux.dev \
--cc=shakeel.butt@linux.dev \
--cc=stable@vger.kernel.org \
--cc=zhoujinmeng@bytedance.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