[PATCH v3 0/6] Open HugeTLB allocation routine for more generic use

[PATCH v3 0/6] Open HugeTLB allocation routine for more generic use

[Ackerley Tng via B4 Relay]

Hi,

The motivation for this patch series is guest_memfd, which would like
to use HugeTLB as a generic source of huge pages but not adopt
HugeTLB's reservation at mmap() time.

By refactoring alloc_hugetlb_folio() and some dependent functions,
there is now an option to allocate HugeTLB folios without providing a
VMA. Specifically, HugeTLB allocation used to be dependent on the VMA
to

1. Look up reservations in the resv_map
2. Get mpol, stored at vma->vm_policy

This refactoring provides hugetlb_alloc_folio(), which focuses on just
the allocation itself, and associated memory and HugeTLB charging
(cgroups). alloc_hugetlb_folio() still handles reservations in the
resv_map and subpools.

Regarding naming, I'm definitely open to alternative names :) I chose
hugetlb_alloc_folio() because I'm seeing this function as a general
allocation function that is provided by the HugeTLB subsystem (hence
the hugetlb_ prefix). I'm intending for alloc_hugetlb_folio() to be
later refactored as a static function for use just by HugeTLB, and
HugeTLBfs should probably use hugetlb_alloc_folio() directly.

To see how hugetlb_alloc_folio() is used by guest_memfd, the most
recent patch series that uses this more generic HugeTLB allocation
routine is at [1], and a newer revision of that patch series is at
[2].

Independently of guest_memfd, I believe this change is useful in
simplifying alloc_hugetlb_folio(). alloc_hugetlb_folio() was so
coupled to a VMA that even HugeTLBfs allocates HugeTLB folios using a
pseudo-VMA.

Testing:

+ libhugetlbfs tests pass
+ ./tools/testing/selftests/mm/ksft_hugetlb.sh passes

Changes in this revision:

+ Clarified comment on when alloc_hugetlb_folio() chooses to dequeue a page
  as requested by Oscar
    + After refactoring out hugetlb_alloc_folio(), hugetlb_alloc_folio()
      just acts on alloc_flags, so I think the reason for dequeueing is
      sufficiently clear, so I dropped the comment
+ Introduced concept of an interpreted mempolicy (struct mempolicy_interpreted).
    + dequeue_hugetlb_folio() and alloc_buddy_hugetlb_folio() are actually
      operating on an interpreted mempolicy, which comprises nid, nodemask, and
      a mempolicy_mode. mempolicy_mode tells the allocation function how to
      interpret the provided nid and nodemask.
    + I think mempolicy_interpreted could be generalized further outside of
      hugetlb. I think every caller of policy_nodemask() will eventually use the
      information in interpreted mempolicy to make the allocation.
    + Let me know if this makes sense, and if I should go ahead to do more
      refactoring!

Oscar, in my responses to v2, I agreed with your idea of an allocation ctxt, but
after doing some initial refactoring, I think the ctxt isn't really an
allocation context, I felt like I was just putting all the parameters for
hugetlb_alloc_folio() into a struct to get named parameters in C.

Here's my rationale for hugetlb_alloc_folio()'s parameters:

+ hstate
+ subpool
   + I think it makes sense to explicitly provide above two as parameters since
     these are always required for allocation of a hugetlb folio. subpool can be
     NULL, but we do want the caller to explicitly pass something in so subpool
     accounting can be performed within hugetlb_alloc_folio().
+ interpreted mempolicy
   + This summarizes what node the folio should be allocated from.
+ gfp
   + gfp is used in a few ways, relating to the actual allocation and zone to
     allocate from:
       1. It was used to interpret the mempolicy, if there was something
          requested in the gfp.
       2. It is used to determine which zone to allocate from in
          dequeue_hugetlb_folio()
       3. A modified version is used in alloc_buddy_hugetlb_folio() if
          MPOL_PREFERRED_MANY, otherwise it is used as-is to allocate a hugetlb
          folio.
    + I retained gfp as a separate parameter, and not part of
      mempolicy_interpreted, because after the mempolicy is interpreted, gfp is
      no longer used in relation to mempolicy to determine which node to
      allocate on.
+ allocation flags
    + I implemented Oscar's suggestion partially in allocation flags, which will
      allow us to add more allocation knobs without adding more parameters to
      hugetlb_alloc_folio().

[1] https://lore.kernel.org/all/cover.1747264138.git.ackerleytng@google.com/T/
[2] https://github.com/googleprodkernel/linux-cc/tree/wip-gmem-conversions-hugetlb-restructuring-12-08-25

RFC v1: https://lore.kernel.org/all/bb35a69a-5be9-45f5-a557-1902487a1bc2@linux.dev/
v2: https://lore.kernel.org/r/20260506-hugetlb-open-up-v2-0-826a0c5f28fc@google.com

---
Ackerley Tng (6):
      mm: hugetlb: Consolidate interpretation of gbl_chg within alloc_hugetlb_folio()
      mm: hugetlb: Move mpol interpretation out of alloc_buddy_hugetlb_folio_with_mpol()
      mm: hugetlb: Move mpol interpretation out of dequeue_hugetlb_folio_vma()
      mm: hugetlb: Use error variable in alloc_hugetlb_folio
      mm: hugetlb: Move mem_cgroup_charge_hugetlb() earlier in allocation
      mm: hugetlb: Refactor out hugetlb_alloc_folio()

 include/linux/hugetlb.h        |  19 ++++
 include/uapi/linux/mempolicy.h |   2 +-
 mm/hugetlb.c                   | 246 +++++++++++++++++++++++------------------
 3 files changed, 157 insertions(+), 110 deletions(-)
---
base-commit: 4d3a2a466b8d68d852a1f3bbf11204b718428dc4
change-id: 20260504-hugetlb-open-up-eaba80571b09

Best regards,
--
Ackerley Tng <ackerleytng@google.com>

[PATCH v3 1/6] mm: hugetlb: Consolidate interpretation of gbl_chg within alloc_hugetlb_folio()

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

The dequeue_hugetlb_folio_vma() function currently handles the gbl_chg
parameter to determine if a folio can be dequeued based on global page
availability. This leaks reservation-specific logic into the dequeueing
path.

Relocate this logic to alloc_hugetlb_folio() so that
dequeue_hugetlb_folio_vma() focuses solely on selecting and dequeuing a
folio. In alloc_hugetlb_folio(), only attempt to dequeue a folio if a
reservation exists (gbl_chg == 0) or if there are available huge pages in
the global pool.

No functional change intended.

Reviewed-by: James Houghton <jthoughton@google.com>
Acked-by: Oscar Salvador <osalvador@suse.de>
Reviewed-by: Joshua Hahn <joshua.hahnjy@gmail.com>
Signed-off-by: Ackerley Tng <ackerleytng@google.com>
---
 mm/hugetlb.c | 25 ++++++++++---------------
 1 file changed, 10 insertions(+), 15 deletions(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index f24bf49be047e..190ab539a97d4 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1336,7 +1336,7 @@ static unsigned long available_huge_pages(struct hstate *h)
 
 static struct folio *dequeue_hugetlb_folio_vma(struct hstate *h,
 				struct vm_area_struct *vma,
-				unsigned long address, long gbl_chg)
+				unsigned long address)
 {
 	struct folio *folio = NULL;
 	struct mempolicy *mpol;
@@ -1344,13 +1344,6 @@ static struct folio *dequeue_hugetlb_folio_vma(struct hstate *h,
 	nodemask_t *nodemask;
 	int nid;
 
-	/*
-	 * gbl_chg==1 means the allocation requires a new page that was not
-	 * reserved before.  Making sure there's at least one free page.
-	 */
-	if (gbl_chg && !available_huge_pages(h))
-		goto err;
-
 	gfp_mask = htlb_alloc_mask(h);
 	nid = huge_node(vma, address, gfp_mask, &mpol, &nodemask);
 
@@ -1368,9 +1361,6 @@ static struct folio *dequeue_hugetlb_folio_vma(struct hstate *h,
 
 	mpol_cond_put(mpol);
 	return folio;
-
-err:
-	return NULL;
 }
 
 #if defined(CONFIG_ARCH_HAS_GIGANTIC_PAGE) && defined(CONFIG_CONTIG_ALLOC)
@@ -2939,12 +2929,17 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		goto out_uncharge_cgroup_reservation;
 
 	spin_lock_irq(&hugetlb_lock);
+
 	/*
-	 * glb_chg is passed to indicate whether or not a page must be taken
-	 * from the global free pool (global change).  gbl_chg == 0 indicates
-	 * a reservation exists for the allocation.
+	 * gbl_chg == 0 indicates a reservation exists for the
+	 * allocation, so try dequeuing a page. In case there was no
+	 * reservation, try dequeuing a page if there are available
+	 * pages in the global pool.
 	 */
-	folio = dequeue_hugetlb_folio_vma(h, vma, addr, gbl_chg);
+	folio = NULL;
+	if (!gbl_chg || available_huge_pages(h))
+		folio = dequeue_hugetlb_folio_vma(h, vma, addr);
+
 	if (!folio) {
 		spin_unlock_irq(&hugetlb_lock);
 		folio = alloc_buddy_hugetlb_folio_with_mpol(h, vma, addr);

-- 
2.54.0.563.g4f69b47b94-goog

[PATCH v3 2/6] mm: hugetlb: Move mpol interpretation out of alloc_buddy_hugetlb_folio_with_mpol()

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

Move memory policy interpretation out of
alloc_buddy_hugetlb_folio_with_mpol() and into alloc_hugetlb_folio() to
separate reading and interpretation of memory policy from actual
allocation.

This will later allow memory policy to be interpreted outside of the
process of allocating a hugetlb folio entirely. This opens doors for other
callers of the HugeTLB folio allocation function, such as guest_memfd,
where memory may not always be mapped and hence may not have an associated
vma.

Introduce struct mempolicy_interpreted to hold all the components of an
interpreted memory policy.

Rename alloc_buddy_hugetlb_folio_with_mpol() to alloc_buddy_hugetlb_folio()
since the function no longer interprets memory policy.

No functional change intended.

Reviewed-by: James Houghton <jthoughton@google.com>
Acked-by: Oscar Salvador <osalvador@suse.de>
Signed-off-by: Ackerley Tng <ackerleytng@google.com>
---
 include/uapi/linux/mempolicy.h |  2 +-
 mm/hugetlb.c                   | 50 +++++++++++++++++++++++++++---------------
 2 files changed, 33 insertions(+), 19 deletions(-)

diff --git a/include/uapi/linux/mempolicy.h b/include/uapi/linux/mempolicy.h
index 6c962d866e864..7f6fc9599693b 100644
--- a/include/uapi/linux/mempolicy.h
+++ b/include/uapi/linux/mempolicy.h
@@ -16,7 +16,7 @@
  */
 
 /* Policies */
-enum {
+enum mempolicy_mode {
 	MPOL_DEFAULT,
 	MPOL_PREFERRED,
 	MPOL_BIND,
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 190ab539a97d4..6a5f69b3b1cb4 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1334,6 +1334,12 @@ static unsigned long available_huge_pages(struct hstate *h)
 	return h->free_huge_pages - h->resv_huge_pages;
 }
 
+struct mempolicy_interpreted {
+	int nid;
+	nodemask_t *nodemask;
+	enum mempolicy_mode mode;
+};
+
 static struct folio *dequeue_hugetlb_folio_vma(struct hstate *h,
 				struct vm_area_struct *vma,
 				unsigned long address)
@@ -2155,32 +2161,28 @@ static struct folio *alloc_migrate_hugetlb_folio(struct hstate *h, gfp_t gfp_mas
 	return folio;
 }
 
-/*
- * Use the VMA's mpolicy to allocate a huge page from the buddy.
- */
 static
-struct folio *alloc_buddy_hugetlb_folio_with_mpol(struct hstate *h,
-		struct vm_area_struct *vma, unsigned long addr)
+struct folio *alloc_buddy_hugetlb_folio(struct hstate *h,
+		gfp_t gfp_mask, struct mempolicy_interpreted *mpoli)
 {
 	struct folio *folio = NULL;
-	struct mempolicy *mpol;
-	gfp_t gfp_mask = htlb_alloc_mask(h);
-	int nid;
-	nodemask_t *nodemask;
+	nodemask_t *nodemask = mpoli->nodemask;
 
-	nid = huge_node(vma, addr, gfp_mask, &mpol, &nodemask);
-	if (mpol_is_preferred_many(mpol)) {
+	if (mpoli->mode == MPOL_PREFERRED_MANY) {
 		gfp_t gfp = gfp_mask & ~(__GFP_DIRECT_RECLAIM | __GFP_NOFAIL);
 
-		folio = alloc_surplus_hugetlb_folio(h, gfp, nid, nodemask);
+		folio = alloc_surplus_hugetlb_folio(h, gfp, mpoli->nid,
+						    nodemask);
 
 		/* Fallback to all nodes if page==NULL */
 		nodemask = NULL;
 	}
 
-	if (!folio)
-		folio = alloc_surplus_hugetlb_folio(h, gfp_mask, nid, nodemask);
-	mpol_cond_put(mpol);
+	if (!folio) {
+		folio = alloc_surplus_hugetlb_folio(h, gfp_mask, mpoli->nid,
+						    nodemask);
+	}
+
 	return folio;
 }
 
@@ -2869,7 +2871,7 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	map_chg_state map_chg;
 	int ret, idx;
 	struct hugetlb_cgroup *h_cg = NULL;
-	gfp_t gfp = htlb_alloc_mask(h) | __GFP_RETRY_MAYFAIL;
+	gfp_t gfp = htlb_alloc_mask(h);
 
 	idx = hstate_index(h);
 
@@ -2941,8 +2943,20 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		folio = dequeue_hugetlb_folio_vma(h, vma, addr);
 
 	if (!folio) {
+		struct mempolicy_interpreted mpoli;
+		struct mempolicy *mpol;
+		nodemask_t *nodemask;
+		int nid;
+
 		spin_unlock_irq(&hugetlb_lock);
-		folio = alloc_buddy_hugetlb_folio_with_mpol(h, vma, addr);
+		nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
+		mpoli = (struct mempolicy_interpreted){
+			.nid = nid,
+			.mode = mpol->mode,
+			.nodemask = nodemask,
+		};
+		folio = alloc_buddy_hugetlb_folio(h, gfp, &mpoli);
+		mpol_cond_put(mpol);
 		if (!folio)
 			goto out_uncharge_cgroup;
 		spin_lock_irq(&hugetlb_lock);
@@ -2998,7 +3012,7 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		}
 	}
 
-	ret = mem_cgroup_charge_hugetlb(folio, gfp);
+	ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
 	/*
 	 * Unconditionally increment NR_HUGETLB here. If it turns out that
 	 * mem_cgroup_charge_hugetlb failed, then immediately free the page and

-- 
2.54.0.563.g4f69b47b94-goog

[PATCH v3 3/6] mm: hugetlb: Move mpol interpretation out of dequeue_hugetlb_folio_vma()

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

Move memory policy interpretation out of dequeue_hugetlb_folio_vma() and
into alloc_hugetlb_folio() to separate reading and interpretation of memory
policy from actual allocation.

Also rename dequeue_hugetlb_folio_vma() to
dequeue_hugetlb_folio_with_mpol() to remove association with vma and to
align with alloc_buddy_hugetlb_folio_with_mpol().

This will later allow memory policy to be interpreted outside of the
process of allocating a hugetlb folio entirely. This opens doors for other
callers of the HugeTLB folio allocation function, such as guest_memfd,
where memory may not always be mapped and hence may not have an associated
vma.

No functional change intended.

Signed-off-by: Ackerley Tng <ackerleytng@google.com>
Reviewed-by: James Houghton <jthoughton@google.com>
---
 mm/hugetlb.c | 57 ++++++++++++++++++++++++++++-----------------------------
 1 file changed, 28 insertions(+), 29 deletions(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 6a5f69b3b1cb4..9807bbe0d70df 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1340,32 +1340,26 @@ struct mempolicy_interpreted {
 	enum mempolicy_mode mode;
 };
 
-static struct folio *dequeue_hugetlb_folio_vma(struct hstate *h,
-				struct vm_area_struct *vma,
-				unsigned long address)
+static struct folio *dequeue_hugetlb_folio(struct hstate *h, gfp_t gfp_mask,
+					   struct mempolicy_interpreted *mpoli)
 {
+	nodemask_t *nodemask = mpoli->nodemask;
 	struct folio *folio = NULL;
-	struct mempolicy *mpol;
-	gfp_t gfp_mask;
-	nodemask_t *nodemask;
-	int nid;
 
-	gfp_mask = htlb_alloc_mask(h);
-	nid = huge_node(vma, address, gfp_mask, &mpol, &nodemask);
-
-	if (mpol_is_preferred_many(mpol)) {
+	if (mpoli->mode == MPOL_PREFERRED_MANY) {
 		folio = dequeue_hugetlb_folio_nodemask(h, gfp_mask,
-							nid, nodemask);
+						       mpoli->nid,
+						       nodemask);
 
 		/* Fallback to all nodes if page==NULL */
 		nodemask = NULL;
 	}
 
-	if (!folio)
+	if (!folio) {
 		folio = dequeue_hugetlb_folio_nodemask(h, gfp_mask,
-							nid, nodemask);
-
-	mpol_cond_put(mpol);
+						       mpoli->nid,
+						       nodemask);
+	}
 	return folio;
 }
 
@@ -2871,7 +2865,11 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	map_chg_state map_chg;
 	int ret, idx;
 	struct hugetlb_cgroup *h_cg = NULL;
+	struct mempolicy_interpreted mpoli;
 	gfp_t gfp = htlb_alloc_mask(h);
+	struct mempolicy *mpol;
+	nodemask_t *nodemask;
+	int nid;
 
 	idx = hstate_index(h);
 
@@ -2930,6 +2928,14 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	if (ret)
 		goto out_uncharge_cgroup_reservation;
 
+	/* Takes reference on mpol. */
+	nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
+	mpoli = (struct mempolicy_interpreted){
+		.nid = nid,
+		.mode = mpol->mode,
+		.nodemask = nodemask,
+	};
+
 	spin_lock_irq(&hugetlb_lock);
 
 	/*
@@ -2940,31 +2946,24 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	 */
 	folio = NULL;
 	if (!gbl_chg || available_huge_pages(h))
-		folio = dequeue_hugetlb_folio_vma(h, vma, addr);
+		folio = dequeue_hugetlb_folio(h, gfp, &mpoli);
 
 	if (!folio) {
-		struct mempolicy_interpreted mpoli;
-		struct mempolicy *mpol;
-		nodemask_t *nodemask;
-		int nid;
-
 		spin_unlock_irq(&hugetlb_lock);
-		nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
-		mpoli = (struct mempolicy_interpreted){
-			.nid = nid,
-			.mode = mpol->mode,
-			.nodemask = nodemask,
-		};
 		folio = alloc_buddy_hugetlb_folio(h, gfp, &mpoli);
 		mpol_cond_put(mpol);
-		if (!folio)
+		if (!folio) {
+			mpol_cond_put(mpol);
 			goto out_uncharge_cgroup;
+		}
 		spin_lock_irq(&hugetlb_lock);
 		list_add(&folio->lru, &h->hugepage_activelist);
 		folio_ref_unfreeze(folio, 1);
 		/* Fall through */
 	}
 
+	mpol_cond_put(mpol);
+
 	/*
 	 * Either dequeued or buddy-allocated folio needs to add special
 	 * mark to the folio when it consumes a global reservation.

-- 
2.54.0.563.g4f69b47b94-goog

[PATCH v3 4/6] mm: hugetlb: Use error variable in alloc_hugetlb_folio

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

Refactor alloc_hugetlb_folio to use a local variable for returning error
codes. Instead of returning ERR_PTR(-ENOSPC) at the end of the error
path, assign -ENOSPC to a return variable at each failure point and
return that variable at the end.

This allows the cleanup goto targets to be used with other errors in a
later patch.

No functional change intended.

Signed-off-by: Ackerley Tng <ackerleytng@google.com>
---
 mm/hugetlb.c | 15 +++++++++++----
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 9807bbe0d70df..ad07e72d6fac3 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -2903,8 +2903,10 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	 */
 	if (map_chg) {
 		gbl_chg = hugepage_subpool_get_pages(spool, 1);
-		if (gbl_chg < 0)
+		if (gbl_chg < 0) {
+			ret = -ENOSPC;
 			goto out_end_reservation;
+		}
 	} else {
 		/*
 		 * If we have the vma reservation ready, no need for extra
@@ -2920,13 +2922,17 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	if (map_chg) {
 		ret = hugetlb_cgroup_charge_cgroup_rsvd(
 			idx, pages_per_huge_page(h), &h_cg);
-		if (ret)
+		if (ret) {
+			ret = -ENOSPC;
 			goto out_subpool_put;
+		}
 	}
 
 	ret = hugetlb_cgroup_charge_cgroup(idx, pages_per_huge_page(h), &h_cg);
-	if (ret)
+	if (ret) {
+		ret = -ENOSPC;
 		goto out_uncharge_cgroup_reservation;
+	}
 
 	/* Takes reference on mpol. */
 	nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
@@ -2954,6 +2960,7 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		mpol_cond_put(mpol);
 		if (!folio) {
 			mpol_cond_put(mpol);
+			ret = -ENOSPC;
 			goto out_uncharge_cgroup;
 		}
 		spin_lock_irq(&hugetlb_lock);
@@ -3046,7 +3053,7 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 out_end_reservation:
 	if (map_chg != MAP_CHG_ENFORCED)
 		vma_end_reservation(h, vma, addr);
-	return ERR_PTR(-ENOSPC);
+	return ERR_PTR(ret);
 }
 
 static __init void *alloc_bootmem(struct hstate *h, int nid, bool node_exact)

-- 
2.54.0.563.g4f69b47b94-goog

[PATCH v3 5/6] mm: hugetlb: Move mem_cgroup_charge_hugetlb() earlier in allocation

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

Move mem_cgroup_charge_hugetlb() earlier in the folio allocation
process. This change draws a cleaner line between memcg charging and the
subsequent hugetlb-specific reservation logic for VMAs and subpools.

While it would be ideal to make all accounting and reservations perfectly
symmetric, mem_cgroup_charge_hugetlb() is a complex operation that cannot
be performed under the hugetlb_lock. Moving the charge to this earlier
point ensures that memcg charging is handled before the code begins
manipulating subpool and VMA-specific state. These two types of accounting
will be separated in a future patch.

If mem_cgroup_charge_hugetlb() fails, the code now branches to
out_subpool_put to ensure the folio is freed and the subpool references are
handled correctly.

Signed-off-by: Ackerley Tng <ackerleytng@google.com>
---
 mm/hugetlb.c | 31 ++++++++++++++++++-------------
 1 file changed, 18 insertions(+), 13 deletions(-)

diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index ad07e72d6fac3..81e73186dff09 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -2991,6 +2991,24 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 
 	spin_unlock_irq(&hugetlb_lock);
 
+	ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
+	/*
+	 * Unconditionally increment NR_HUGETLB here. If it turns out that
+	 * mem_cgroup_charge_hugetlb failed, then immediately free the page and
+	 * decrement NR_HUGETLB.
+	 */
+	lruvec_stat_mod_folio(folio, NR_HUGETLB, pages_per_huge_page(h));
+
+	if (ret == -ENOMEM) {
+		free_huge_folio(folio);
+		/*
+		 * Skip uncharging hugetlb_cgroup since the charges
+		 * were committed to the folio and freeing the folio
+		 * would have cleared those up.
+		 */
+		goto out_subpool_put;
+	}
+
 	hugetlb_set_folio_subpool(folio, spool);
 
 	if (map_chg != MAP_CHG_ENFORCED) {
@@ -3018,19 +3036,6 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		}
 	}
 
-	ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
-	/*
-	 * Unconditionally increment NR_HUGETLB here. If it turns out that
-	 * mem_cgroup_charge_hugetlb failed, then immediately free the page and
-	 * decrement NR_HUGETLB.
-	 */
-	lruvec_stat_mod_folio(folio, NR_HUGETLB, pages_per_huge_page(h));
-
-	if (ret == -ENOMEM) {
-		free_huge_folio(folio);
-		return ERR_PTR(-ENOMEM);
-	}
-
 	return folio;
 
 out_uncharge_cgroup:

-- 
2.54.0.563.g4f69b47b94-goog

[PATCH v3 6/6] mm: hugetlb: Refactor out hugetlb_alloc_folio()

[Ackerley Tng via B4 Relay]

From: Ackerley Tng <ackerleytng@google.com>

Refactor out hugetlb_alloc_folio() from alloc_hugetlb_folio(), which
handles allocation of a folio and memory and HugeTLB charging to cgroups.

This refactoring decouples the HugeTLB page allocation from VMAs,
specifically:

1. Reservations (as in resv_map) are stored in the vma
2. mpol is stored at vma->vm_policy
3. A vma must be used for allocation even if the pages are not meant to be
   used by host process.

Without this coupling, VMAs are no longer a requirement for
allocation. This opens up the allocation routine for usage without VMAs,
which will allow guest_memfd to use HugeTLB as a more generic allocator of
huge pages, since guest_memfd memory may not have any associated VMAs by
design. In addition, direct allocations from HugeTLB could possibly be
refactored to avoid the use of a pseudo-VMA.

Also, this decouples HugeTLB page allocation from HugeTLBfs, where the
subpool is stored at the fs mount. This is also a requirement for
guest_memfd, where the plan is to have a subpool created per-fd and stored
on the inode.

Provide and use alloc_flags to allow more allocation knobs in future
without expanding the number of parameters in hugetlb_alloc_folio().

No functional change intended.

Signed-off-by: Ackerley Tng <ackerleytng@google.com>
---
 include/linux/hugetlb.h |  19 +++++
 mm/hugetlb.c            | 188 +++++++++++++++++++++++++-----------------------
 2 files changed, 117 insertions(+), 90 deletions(-)

diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h
index 93418625d3c5f..9a0222851573d 100644
--- a/include/linux/hugetlb.h
+++ b/include/linux/hugetlb.h
@@ -2,6 +2,7 @@
 #ifndef _LINUX_HUGETLB_H
 #define _LINUX_HUGETLB_H
 
+#include <linux/mempolicy.h>
 #include <linux/mm.h>
 #include <linux/mm_types.h>
 #include <linux/mmdebug.h>
@@ -705,6 +706,24 @@ bool hugetlb_bootmem_page_zones_valid(int nid, struct huge_bootmem_page *m);
 int isolate_or_dissolve_huge_folio(struct folio *folio, struct list_head *list);
 int replace_free_hugepage_folios(unsigned long start_pfn, unsigned long end_pfn);
 void wait_for_freed_hugetlb_folios(void);
+
+struct mempolicy_interpreted {
+	int nid;
+	nodemask_t *nodemask;
+	enum mempolicy_mode mode;
+};
+
+enum hugetlb_alloc_flag {
+	HUGETLB_ALLOC_CHARGE_CGROUP_RSVD_BIT = 0,
+	HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS_BIT,
+};
+
+#define HUGETLB_ALLOC_CHARG_CGROUP_RSVD BIT(HUGETLB_ALLOC_CHARGE_CGROUP_RSVD_BIT)
+#define HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS BIT(HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS_BIT)
+
+struct folio *hugetlb_alloc_folio(struct hstate *h, struct hugepage_subpool *spool,
+		gfp_t gfp, struct mempolicy_interpreted *mpoli,
+		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/mm/hugetlb.c b/mm/hugetlb.c
index 81e73186dff09..abce2ca76fb9c 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -1334,12 +1334,6 @@ static unsigned long available_huge_pages(struct hstate *h)
 	return h->free_huge_pages - h->resv_huge_pages;
 }
 
-struct mempolicy_interpreted {
-	int nid;
-	nodemask_t *nodemask;
-	enum mempolicy_mode mode;
-};
-
 static struct folio *dequeue_hugetlb_folio(struct hstate *h, gfp_t gfp_mask,
 					   struct mempolicy_interpreted *mpoli)
 {
@@ -2829,6 +2823,90 @@ void wait_for_freed_hugetlb_folios(void)
 	flush_work(&free_hpage_work);
 }
 
+struct folio *hugetlb_alloc_folio(struct hstate *h, struct hugepage_subpool *spool,
+		gfp_t gfp, struct mempolicy_interpreted *mpoli,
+		u8 alloc_flags)
+{
+	bool charge_hugetlb_cgroup_rsvd = alloc_flags &
+					  HUGETLB_ALLOC_CHARG_CGROUP_RSVD;
+	bool use_global_reservation = alloc_flags &
+				      HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS;
+	size_t nr_pages = pages_per_huge_page(h);
+	struct hugetlb_cgroup *h_cg = NULL;
+	int idx = hstate_index(h);
+	struct folio *folio;
+	int ret;
+
+	if (charge_hugetlb_cgroup_rsvd &&
+	    hugetlb_cgroup_charge_cgroup_rsvd(idx, nr_pages, &h_cg))
+		return ERR_PTR(-ENOSPC);
+
+	if (hugetlb_cgroup_charge_cgroup(idx, nr_pages, &h_cg)) {
+		ret = -ENOSPC;
+		goto err_uncharge_hugetlb_cgroup_rsvd;
+	}
+
+	spin_lock_irq(&hugetlb_lock);
+
+	folio = NULL;
+	if (use_global_reservation || available_huge_pages(h))
+		folio = dequeue_hugetlb_folio(h, gfp, mpoli);
+
+	if (!folio) {
+		spin_unlock_irq(&hugetlb_lock);
+		folio = alloc_buddy_hugetlb_folio(h, gfp, mpoli);
+		if (!folio) {
+			ret = -ENOSPC;
+			goto err_uncharge_hugetlb_cgroup;
+		}
+		spin_lock_irq(&hugetlb_lock);
+		list_add(&folio->lru, &h->hugepage_activelist);
+		folio_ref_unfreeze(folio, 1);
+	}
+
+	if (use_global_reservation) {
+		folio_set_hugetlb_restore_reserve(folio);
+		h->resv_huge_pages--;
+	}
+
+	hugetlb_cgroup_commit_charge(idx, pages_per_huge_page(h), h_cg, folio);
+
+	if (charge_hugetlb_cgroup_rsvd) {
+		hugetlb_cgroup_commit_charge_rsvd(idx, pages_per_huge_page(h),
+						  h_cg, folio);
+	}
+
+	spin_unlock_irq(&hugetlb_lock);
+
+	ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
+	/*
+	 * Unconditionally increment NR_HUGETLB here because if
+	 * mem_cgroup_charge_hugetlb failed, freeing the page will
+	 * decrement NR_HUGETLB.
+	 */
+	lruvec_stat_mod_folio(folio, NR_HUGETLB, pages_per_huge_page(h));
+
+	if (ret == -ENOMEM) {
+		free_huge_folio(folio);
+		/*
+		 * Skip uncharging hugetlb_cgroup since the charges
+		 * were committed to the folio and freeing the folio
+		 * would have cleared those up.
+		 */
+		return ERR_PTR(ret);
+	}
+
+	return folio;
+
+ err_uncharge_hugetlb_cgroup:
+	hugetlb_cgroup_uncharge_cgroup(idx, nr_pages, h_cg);
+ err_uncharge_hugetlb_cgroup_rsvd:
+	if (charge_hugetlb_cgroup_rsvd)
+		hugetlb_cgroup_uncharge_cgroup_rsvd(idx, nr_pages, h_cg);
+
+	return ERR_PTR(ret);
+}
+
 typedef enum {
 	/*
 	 * For either 0/1: we checked the per-vma resv map, and one resv
@@ -2864,11 +2942,11 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	long retval, gbl_chg, gbl_reserve;
 	map_chg_state map_chg;
 	int ret, idx;
-	struct hugetlb_cgroup *h_cg = NULL;
 	struct mempolicy_interpreted mpoli;
 	gfp_t gfp = htlb_alloc_mask(h);
 	struct mempolicy *mpol;
 	nodemask_t *nodemask;
+	u8 alloc_flags = 0;
 	int nid;
 
 	idx = hstate_index(h);
@@ -2916,23 +2994,18 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 	}
 
 	/*
-	 * If this allocation is not consuming a per-vma reservation,
-	 * charge the hugetlb cgroup now.
+	 * If allocation doesn't reuse a reservation in the resv_map,
+	 * charge for the reservation.
 	 */
-	if (map_chg) {
-		ret = hugetlb_cgroup_charge_cgroup_rsvd(
-			idx, pages_per_huge_page(h), &h_cg);
-		if (ret) {
-			ret = -ENOSPC;
-			goto out_subpool_put;
-		}
-	}
+	if (map_chg != MAP_CHG_REUSE)
+		alloc_flags |= HUGETLB_ALLOC_CHARG_CGROUP_RSVD;
 
-	ret = hugetlb_cgroup_charge_cgroup(idx, pages_per_huge_page(h), &h_cg);
-	if (ret) {
-		ret = -ENOSPC;
-		goto out_uncharge_cgroup_reservation;
-	}
+	/*
+	 * gbl_chg == 0 indicates a reservation exists for this
+	 * allocation, so try to use it.
+	 */
+	if (gbl_chg == 0)
+		alloc_flags |= HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS;
 
 	/* Takes reference on mpol. */
 	nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
@@ -2942,70 +3015,12 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		.nodemask = nodemask,
 	};
 
-	spin_lock_irq(&hugetlb_lock);
-
-	/*
-	 * gbl_chg == 0 indicates a reservation exists for the
-	 * allocation, so try dequeuing a page. In case there was no
-	 * reservation, try dequeuing a page if there are available
-	 * pages in the global pool.
-	 */
-	folio = NULL;
-	if (!gbl_chg || available_huge_pages(h))
-		folio = dequeue_hugetlb_folio(h, gfp, &mpoli);
-
-	if (!folio) {
-		spin_unlock_irq(&hugetlb_lock);
-		folio = alloc_buddy_hugetlb_folio(h, gfp, &mpoli);
-		mpol_cond_put(mpol);
-		if (!folio) {
-			mpol_cond_put(mpol);
-			ret = -ENOSPC;
-			goto out_uncharge_cgroup;
-		}
-		spin_lock_irq(&hugetlb_lock);
-		list_add(&folio->lru, &h->hugepage_activelist);
-		folio_ref_unfreeze(folio, 1);
-		/* Fall through */
-	}
+	folio = hugetlb_alloc_folio(h, spool, gfp, &mpoli, alloc_flags);
 
 	mpol_cond_put(mpol);
 
-	/*
-	 * Either dequeued or buddy-allocated folio needs to add special
-	 * mark to the folio when it consumes a global reservation.
-	 */
-	if (!gbl_chg) {
-		folio_set_hugetlb_restore_reserve(folio);
-		h->resv_huge_pages--;
-	}
-
-	hugetlb_cgroup_commit_charge(idx, pages_per_huge_page(h), h_cg, folio);
-	/* If allocation is not consuming a reservation, also store the
-	 * hugetlb_cgroup pointer on the page.
-	 */
-	if (map_chg) {
-		hugetlb_cgroup_commit_charge_rsvd(idx, pages_per_huge_page(h),
-						  h_cg, folio);
-	}
-
-	spin_unlock_irq(&hugetlb_lock);
-
-	ret = mem_cgroup_charge_hugetlb(folio, gfp | __GFP_RETRY_MAYFAIL);
-	/*
-	 * Unconditionally increment NR_HUGETLB here. If it turns out that
-	 * mem_cgroup_charge_hugetlb failed, then immediately free the page and
-	 * decrement NR_HUGETLB.
-	 */
-	lruvec_stat_mod_folio(folio, NR_HUGETLB, pages_per_huge_page(h));
-
-	if (ret == -ENOMEM) {
-		free_huge_folio(folio);
-		/*
-		 * Skip uncharging hugetlb_cgroup since the charges
-		 * were committed to the folio and freeing the folio
-		 * would have cleared those up.
-		 */
+	if (IS_ERR(folio)) {
+		ret = PTR_ERR(folio);
 		goto out_subpool_put;
 	}
 
@@ -3038,12 +3053,6 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 
 	return folio;
 
-out_uncharge_cgroup:
-	hugetlb_cgroup_uncharge_cgroup(idx, pages_per_huge_page(h), h_cg);
-out_uncharge_cgroup_reservation:
-	if (map_chg)
-		hugetlb_cgroup_uncharge_cgroup_rsvd(idx, pages_per_huge_page(h),
-						    h_cg);
 out_subpool_put:
 	/*
 	 * put page to subpool iff the quota of subpool's rsv_hpages is used
@@ -3054,7 +3063,6 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
 		hugetlb_acct_memory(h, -gbl_reserve);
 	}
 
-
 out_end_reservation:
 	if (map_chg != MAP_CHG_ENFORCED)
 		vma_end_reservation(h, vma, addr);

-- 
2.54.0.563.g4f69b47b94-goog

Re: [PATCH v3 6/6] mm: hugetlb: Refactor out hugetlb_alloc_folio()

[Oscar Salvador (SUSE)]

On Mon, May 18, 2026 at 05:19:45PM -0700, Ackerley Tng via B4 Relay wrote:
> From: Ackerley Tng <ackerleytng@google.com>
> 
> Refactor out hugetlb_alloc_folio() from alloc_hugetlb_folio(), which
> handles allocation of a folio and memory and HugeTLB charging to cgroups.
> 
> This refactoring decouples the HugeTLB page allocation from VMAs,
> specifically:
> 
> 1. Reservations (as in resv_map) are stored in the vma
> 2. mpol is stored at vma->vm_policy
> 3. A vma must be used for allocation even if the pages are not meant to be
>    used by host process.
> 
> Without this coupling, VMAs are no longer a requirement for
> allocation. This opens up the allocation routine for usage without VMAs,
> which will allow guest_memfd to use HugeTLB as a more generic allocator of
> huge pages, since guest_memfd memory may not have any associated VMAs by
> design. In addition, direct allocations from HugeTLB could possibly be
> refactored to avoid the use of a pseudo-VMA.
> 
> Also, this decouples HugeTLB page allocation from HugeTLBfs, where the
> subpool is stored at the fs mount. This is also a requirement for
> guest_memfd, where the plan is to have a subpool created per-fd and stored
> on the inode.
> 
> Provide and use alloc_flags to allow more allocation knobs in future
> without expanding the number of parameters in hugetlb_alloc_folio().
> 
> No functional change intended.
> 
> Signed-off-by: Ackerley Tng <ackerleytng@google.com>

I am traveling so I plan to do a more careful review in a few days but I
wanted to do some remark:

> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
...  
> +struct folio *hugetlb_alloc_folio(struct hstate *h, struct hugepage_subpool *spool,
> +		gfp_t gfp, struct mempolicy_interpreted *mpoli,
> +		u8 alloc_flags)
> +{
> +	bool charge_hugetlb_cgroup_rsvd = alloc_flags &
> +					  HUGETLB_ALLOC_CHARG_CGROUP_RSVD;
> +	bool use_global_reservation = alloc_flags &
> +				      HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS;
> +	size_t nr_pages = pages_per_huge_page(h);
> +	struct hugetlb_cgroup *h_cg = NULL;
> +	int idx = hstate_index(h);
> +	struct folio *folio;
> +	int ret;

IIUC, hugetlb_alloc_folio() will be the gateway function everyone will use from now
on to allocate hugetlb pages, that goes for hugetlb itself and external users like guest_memfd.

The gfp flag should really be defined here, because otherwise you are creating openings for
external users to start passing their own crafted gfp flags and that feels wrong.
Sure we can guard that down the road but we should not really need it.

 

-- 
Oscar Salvador
SUSE Labs

Re: [PATCH v3 6/6] mm: hugetlb: Refactor out hugetlb_alloc_folio()

[Ackerley Tng]

"Oscar Salvador (SUSE)" <osalvador@kernel.org> writes:

> On Mon, May 18, 2026 at 05:19:45PM -0700, Ackerley Tng via B4 Relay wrote:
>> From: Ackerley Tng <ackerleytng@google.com>
>>
>> Refactor out hugetlb_alloc_folio() from alloc_hugetlb_folio(), which
>> handles allocation of a folio and memory and HugeTLB charging to cgroups.
>>
>> This refactoring decouples the HugeTLB page allocation from VMAs,
>> specifically:
>>
>> 1. Reservations (as in resv_map) are stored in the vma
>> 2. mpol is stored at vma->vm_policy
>> 3. A vma must be used for allocation even if the pages are not meant to be
>>    used by host process.
>>
>> Without this coupling, VMAs are no longer a requirement for
>> allocation. This opens up the allocation routine for usage without VMAs,
>> which will allow guest_memfd to use HugeTLB as a more generic allocator of
>> huge pages, since guest_memfd memory may not have any associated VMAs by
>> design. In addition, direct allocations from HugeTLB could possibly be
>> refactored to avoid the use of a pseudo-VMA.
>>
>> Also, this decouples HugeTLB page allocation from HugeTLBfs, where the
>> subpool is stored at the fs mount. This is also a requirement for
>> guest_memfd, where the plan is to have a subpool created per-fd and stored
>> on the inode.
>>
>> Provide and use alloc_flags to allow more allocation knobs in future
>> without expanding the number of parameters in hugetlb_alloc_folio().
>>
>> No functional change intended.
>>
>> Signed-off-by: Ackerley Tng <ackerleytng@google.com>
>
> I am traveling so I plan to do a more careful review in a few days but I
> wanted to do some remark:
>

Looking forward to the rest of your reviews!

>> --- a/mm/hugetlb.c
>> +++ b/mm/hugetlb.c
> ...
>> +struct folio *hugetlb_alloc_folio(struct hstate *h, struct hugepage_subpool *spool,
>> +		gfp_t gfp, struct mempolicy_interpreted *mpoli,
>> +		u8 alloc_flags)
>> +{
>> +	bool charge_hugetlb_cgroup_rsvd = alloc_flags &
>> +					  HUGETLB_ALLOC_CHARG_CGROUP_RSVD;
>> +	bool use_global_reservation = alloc_flags &
>> +				      HUGETLB_ALLOC_USE_GLOBAL_RESERVATIONS;
>> +	size_t nr_pages = pages_per_huge_page(h);
>> +	struct hugetlb_cgroup *h_cg = NULL;
>> +	int idx = hstate_index(h);
>> +	struct folio *folio;
>> +	int ret;
>
> IIUC, hugetlb_alloc_folio() will be the gateway function everyone will use from now
> on to allocate hugetlb pages, that goes for hugetlb itself and external users like guest_memfd.
>
> The gfp flag should really be defined here, because otherwise you are creating openings for
> external users to start passing their own crafted gfp flags and that feels wrong.
> Sure we can guard that down the road but we should not really need it.
>

Makes sense. I'll define gfp with htlb_alloc_mask(h) within
hugetlb_alloc_folio() instead of letting gfp be a parameter in the next
version.

>
>
> --
> Oscar Salvador
> SUSE Labs

Re: [PATCH v3 2/6] mm: hugetlb: Move mpol interpretation out of alloc_buddy_hugetlb_folio_with_mpol()

[Ackerley Tng]

Ackerley Tng via B4 Relay <devnull+ackerleytng.google.com@kernel.org>
writes:

>
> [...snip...]
>
> diff --git a/mm/hugetlb.c b/mm/hugetlb.c
> index 190ab539a97d4..6a5f69b3b1cb4 100644
> --- a/mm/hugetlb.c
> +++ b/mm/hugetlb.c
> @@ -1334,6 +1334,12 @@ static unsigned long available_huge_pages(struct hstate *h)
>  	return h->free_huge_pages - h->resv_huge_pages;
>  }
>
> +struct mempolicy_interpreted {
> +	int nid;
> +	nodemask_t *nodemask;
> +	enum mempolicy_mode mode;
> +};
> +
>
> [...snip...]
>
> @@ -2941,8 +2943,20 @@ struct folio *alloc_hugetlb_folio(struct vm_area_struct *vma,
>  		folio = dequeue_hugetlb_folio_vma(h, vma, addr);
>
>  	if (!folio) {
> +		struct mempolicy_interpreted mpoli;
> +		struct mempolicy *mpol;
> +		nodemask_t *nodemask;
> +		int nid;
> +
>  		spin_unlock_irq(&hugetlb_lock);
> -		folio = alloc_buddy_hugetlb_folio_with_mpol(h, vma, addr);
> +		nid = huge_node(vma, addr, gfp, &mpol, &nodemask);
> +		mpoli = (struct mempolicy_interpreted){
> +			.nid = nid,
> +			.mode = mpol->mode,
> +			.nodemask = nodemask,
> +		};

This needs to take into account CONFIG_NUMA, so the next revision will
look something like this instead:

	mpoli = (struct mempolicy_interpreted){
		.nid = nid,
#ifdef CONFIG_NUMA
		.mode = mpol ? mpol->mode : MPOL_DEFAULT,
#else
		.mode = MPOL_DEFAULT,
#endif
		.nodemask = nodemask,
	};


> +		folio = alloc_buddy_hugetlb_folio(h, gfp, &mpoli);
> +		mpol_cond_put(mpol);
>  		if (!folio)
>  			goto out_uncharge_cgroup;
>  		spin_lock_irq(&hugetlb_lock);
>
> [...snip...]
>