* [PATCH v2] mm: hugetlb: allocate larger early CMA pages first
@ 2026-11-22 6:55 Qi Zheng
2026-11-22 13:27 ` Jose A. Perez de Azpillaga
2026-11-22 22:17 ` + mm-hugetlb-allocate-larger-early-cma-pages-first.patch added to mm-new branch Andrew Morton
0 siblings, 2 replies; 3+ messages in thread
From: Qi Zheng @ 2026-11-22 6:55 UTC (permalink / raw)
To: muchun.song, osalvador, david, akpm, fvdl
Cc: linux-mm, linux-kernel, Qi Zheng, stable
From: Qi Zheng <zhengqi.arch@bytedance.com>
With hugetlb_cma_only=1, early HugeTLB CMA allocations follow hstate
registration order, which can depend on the order of hugepagesz= and
default_hugepagesz= parameters. cma_reserve_early() advances early_pfn by
the requested size without aligning the start of each allocation.
On arm64 with 4 KiB base pages, 32 MiB and 1 GiB pages share a CMA pool
aligned to 1 GiB. In a 2 GiB CMA range starting at 4 GiB, allocating a
32 MiB page followed by a 1 GiB page places the latter at 4 GiB + 32 MiB,
violating its required 1 GiB physical alignment.
Simply rounding early_pfn up before the 1 GiB allocation would leave a
992 MiB gap and advance the cursor to 6 GiB. cma_activate_area() marks the
entire prefix below early_pfn as allocated and only frees pageblocks after
it. The gap would remain reserved with its bitmap bits set, and releasing
the two HugeTLB folios would not reclaim it, permanently losing 992 MiB
of usable CMA capacity. An alignment-only fix would therefore also need
to track and reclaim such gaps.
The same alignment issue can affect non-LPAR PowerPC64 hash MMU systems
supporting 16 MiB and 16 GiB pages, with 64 KiB base pages and
CONFIG_ARCH_FORCE_MAX_ORDER=7. The 16 MiB size then has order 8 and is
also gigantic, sharing the 16 GiB-aligned CMA pool.
Allocate early CMA hstates in descending order. Each range starts aligned
to the largest supported size, so this preserves alignment for subsequent
smaller allocations without creating gaps.
Fixes: d2d786714080 ("mm/hugetlb: enable bootmem allocation from CMA areas")
Cc: stable@vger.kernel.org
Suggested-by: David Hildenbrand (Arm) <david@kernel.org>
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Acked-by: Muchun Song (ByteDance) <muchun.song@linux.dev>
---
mm/hugetlb.c | 29 ++++++++++++++++++++++++++++-
1 file changed, 28 insertions(+), 1 deletion(-)
diff --git a/mm/hugetlb.c b/mm/hugetlb.c
index 4a448c5909d36..96b97f012ab49 100644
--- a/mm/hugetlb.c
+++ b/mm/hugetlb.c
@@ -4600,8 +4600,25 @@ void __init hugetlb_bootmem_set_nodes(void)
node_set(nid, hugetlb_bootmem_nodes);
}
+/* Return the largest early-CMA hstate below the given order. */
+static struct hstate *__init hugetlb_next_early_cma_hstate(unsigned int order)
+{
+ struct hstate *h, *next = NULL;
+
+ for_each_hstate(h) {
+ if (!hugetlb_early_cma(h) || huge_page_order(h) >= order)
+ continue;
+
+ if (!next || huge_page_order(h) > huge_page_order(next))
+ next = h;
+ }
+
+ return next;
+}
+
void __init hugetlb_bootmem_alloc(void)
{
+ unsigned int order;
struct hstate *h;
int i;
@@ -4615,9 +4632,19 @@ void __init hugetlb_bootmem_alloc(void)
for_each_hstate(h) {
h->next_nid_to_alloc = first_online_node;
- if (hstate_is_gigantic(h))
+ if (hstate_is_gigantic(h) && !hugetlb_early_cma(h))
hugetlb_hstate_alloc_pages(h);
}
+
+ /*
+ * CMA ranges are aligned to the largest supported size. Allocate larger
+ * sizes first to keep the early reservation cursor aligned without gaps.
+ */
+ order = MAX_FOLIO_ORDER + 1;
+ while ((h = hugetlb_next_early_cma_hstate(order))) {
+ hugetlb_hstate_alloc_pages(h);
+ order = huge_page_order(h);
+ }
}
/*
--
2.54.0
^ permalink raw reply related [flat|nested] 3+ messages in thread* Re: [PATCH v2] mm: hugetlb: allocate larger early CMA pages first
2026-11-22 6:55 [PATCH v2] mm: hugetlb: allocate larger early CMA pages first Qi Zheng
@ 2026-11-22 13:27 ` Jose A. Perez de Azpillaga
2026-11-22 22:17 ` + mm-hugetlb-allocate-larger-early-cma-pages-first.patch added to mm-new branch Andrew Morton
1 sibling, 0 replies; 3+ messages in thread
From: Jose A. Perez de Azpillaga @ 2026-11-22 13:27 UTC (permalink / raw)
To: Qi Zheng, muchun.song, osalvador, david, akpm, fvdl
Cc: linux-mm, linux-kernel, Qi Zheng, stable
On 11/22/26 7:55 AM, Qi Zheng wrote:
> From: Qi Zheng <zhengqi.arch@bytedance.com>
>
> With hugetlb_cma_only=1, early HugeTLB CMA allocations follow hstate
> registration order, which can depend on the order of hugepagesz= and
> default_hugepagesz= parameters. cma_reserve_early() advances early_pfn by
> the requested size without aligning the start of each allocation.
>
> On arm64 with 4 KiB base pages, 32 MiB and 1 GiB pages share a CMA pool
> aligned to 1 GiB. In a 2 GiB CMA range starting at 4 GiB, allocating a
> 32 MiB page followed by a 1 GiB page places the latter at 4 GiB + 32 MiB,
> violating its required 1 GiB physical alignment.
>
> Simply rounding early_pfn up before the 1 GiB allocation would leave a
> 992 MiB gap and advance the cursor to 6 GiB. cma_activate_area() marks the
> entire prefix below early_pfn as allocated and only frees pageblocks after
> it. The gap would remain reserved with its bitmap bits set, and releasing
> the two HugeTLB folios would not reclaim it, permanently losing 992 MiB
> of usable CMA capacity. An alignment-only fix would therefore also need
> to track and reclaim such gaps.
>
> The same alignment issue can affect non-LPAR PowerPC64 hash MMU systems
> supporting 16 MiB and 16 GiB pages, with 64 KiB base pages and
> CONFIG_ARCH_FORCE_MAX_ORDER=7. The 16 MiB size then has order 8 and is
> also gigantic, sharing the 16 GiB-aligned CMA pool.
>
> Allocate early CMA hstates in descending order. Each range starts aligned
> to the largest supported size, so this preserves alignment for subsequent
> smaller allocations without creating gaps.
>
> Fixes: d2d786714080 ("mm/hugetlb: enable bootmem allocation from CMA areas")
> Cc: stable@vger.kernel.org
> Suggested-by: David Hildenbrand (Arm) <david@kernel.org>
> Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
> Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
> Acked-by: Muchun Song (ByteDance) <muchun.song@linux.dev>
LGTM:
Reviewed-by: Jose A. Perez de Azpillaga <azpijr@gmail.com>
--
cheers,
jose a. p-a
^ permalink raw reply [flat|nested] 3+ messages in thread* + mm-hugetlb-allocate-larger-early-cma-pages-first.patch added to mm-new branch
2026-11-22 6:55 [PATCH v2] mm: hugetlb: allocate larger early CMA pages first Qi Zheng
2026-11-22 13:27 ` Jose A. Perez de Azpillaga
@ 2026-11-22 22:17 ` Andrew Morton
1 sibling, 0 replies; 3+ messages in thread
From: Andrew Morton @ 2026-11-22 22:17 UTC (permalink / raw)
To: mm-commits, stable, osalvador, muchun.song, fvdl, david, azpijr,
zhengqi.arch, akpm
The patch titled
Subject: mm: hugetlb: allocate larger early CMA pages first
has been added to the -mm mm-new branch. Its filename is
mm-hugetlb-allocate-larger-early-cma-pages-first.patch
This patch will shortly appear at
https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patches/mm-hugetlb-allocate-larger-early-cma-pages-first.patch
This patch will later appear in the mm-new branch at
git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
Note, mm-new is a provisional staging ground for work-in-progress
patches, and acceptance into mm-new is a notification for others take
notice and to finish up reviews. Please do not hesitate to respond to
review feedback and post updated versions to replace or incrementally
fixup patches in mm-new.
The mm-new branch of mm.git is not included in linux-next
If a few days of testing in mm-new is successful, the patch will me moved
into mm.git's mm-unstable branch, which is included in linux-next
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next via various
branches at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm
and is updated there most days
------------------------------------------------------
From: Qi Zheng <zhengqi.arch@bytedance.com>
Subject: mm: hugetlb: allocate larger early CMA pages first
Date: Sun, 22 Nov 2026 14:55:26 +0800
With hugetlb_cma_only=1, early HugeTLB CMA allocations follow hstate
registration order, which can depend on the order of hugepagesz= and
default_hugepagesz= parameters. cma_reserve_early() advances early_pfn by
the requested size without aligning the start of each allocation.
On arm64 with 4 KiB base pages, 32 MiB and 1 GiB pages share a CMA pool
aligned to 1 GiB. In a 2 GiB CMA range starting at 4 GiB, allocating a 32
MiB page followed by a 1 GiB page places the latter at 4 GiB + 32 MiB,
violating its required 1 GiB physical alignment.
Simply rounding early_pfn up before the 1 GiB allocation would leave a 992
MiB gap and advance the cursor to 6 GiB. cma_activate_area() marks the
entire prefix below early_pfn as allocated and only frees pageblocks after
it. The gap would remain reserved with its bitmap bits set, and releasing
the two HugeTLB folios would not reclaim it, permanently losing 992 MiB of
usable CMA capacity. An alignment-only fix would therefore also need to
track and reclaim such gaps.
The same alignment issue can affect non-LPAR PowerPC64 hash MMU systems
supporting 16 MiB and 16 GiB pages, with 64 KiB base pages and
CONFIG_ARCH_FORCE_MAX_ORDER=7. The 16 MiB size then has order 8 and is
also gigantic, sharing the 16 GiB-aligned CMA pool.
Allocate early CMA hstates in descending order. Each range starts aligned
to the largest supported size, so this preserves alignment for subsequent
smaller allocations without creating gaps.
Link: https://patch.msgid.link/20261122065527.33123-1-qi.zheng@linux.dev
Fixes: d2d786714080 ("mm/hugetlb: enable bootmem allocation from CMA areas")
Signed-off-by: Qi Zheng <zhengqi.arch@bytedance.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Suggested-by: David Hildenbrand (Arm) <david@kernel.org>
Reviewed-by: David Hildenbrand (Arm) <david@kernel.org>
Acked-by: Muchun Song (ByteDance) <muchun.song@linux.dev>
Reviewed-by: Jose A. Perez de Azpillaga <azpijr@gmail.com>
Cc: Oscar Salvador <osalvador@suse.de>
Cc: Frank van der Linden <fvdl@google.com>
Cc: <stable@vger.kernel.org>
mm/hugetlb.c | 29 ++++++++++++++++++++++++++++-
1 file changed, 28 insertions(+), 1 deletion(-)
--- a/mm/hugetlb.c~mm-hugetlb-allocate-larger-early-cma-pages-first
+++ a/mm/hugetlb.c
@@ -4651,8 +4651,25 @@ void __init hugetlb_bootmem_set_nodes(vo
node_set(nid, hugetlb_bootmem_nodes);
}
+/* Return the largest early-CMA hstate below the given order. */
+static struct hstate *__init hugetlb_next_early_cma_hstate(unsigned int order)
+{
+ struct hstate *h, *next = NULL;
+
+ for_each_hstate(h) {
+ if (!hugetlb_early_cma(h) || huge_page_order(h) >= order)
+ continue;
+
+ if (!next || huge_page_order(h) > huge_page_order(next))
+ next = h;
+ }
+
+ return next;
+}
+
void __init hugetlb_bootmem_alloc(void)
{
+ unsigned int order;
struct hstate *h;
int i;
@@ -4666,9 +4683,19 @@ void __init hugetlb_bootmem_alloc(void)
for_each_hstate(h) {
h->next_nid_to_alloc = first_online_node;
- if (hstate_is_gigantic(h))
+ if (hstate_is_gigantic(h) && !hugetlb_early_cma(h))
hugetlb_hstate_alloc_pages(h);
}
+
+ /*
+ * CMA ranges are aligned to the largest supported size. Allocate larger
+ * sizes first to keep the early reservation cursor aligned without gaps.
+ */
+ order = MAX_FOLIO_ORDER + 1;
+ while ((h = hugetlb_next_early_cma_hstate(order))) {
+ hugetlb_hstate_alloc_pages(h);
+ order = huge_page_order(h);
+ }
}
/*
_
Patches currently in -mm which might be from zhengqi.arch@bytedance.com are
x86-hugetlb-disable-default-cma-reservations-on-32-bit-kernel.patch
mm-gup-allow-long-term-pins-of-cma-allocated-hugetlb-folios.patch
mm-hugetlb-fix-gfp-mask-propagation-to-cma.patch
mm-hugetlb-allocate-larger-early-cma-pages-first.patch
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2026-11-22 22:17 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2026-11-22 6:55 [PATCH v2] mm: hugetlb: allocate larger early CMA pages first Qi Zheng
2026-11-22 13:27 ` Jose A. Perez de Azpillaga
2026-11-22 22:17 ` + mm-hugetlb-allocate-larger-early-cma-pages-first.patch added to mm-new branch Andrew Morton
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox