Skip to contentSkip to conversation

mm/memremap: fix spurious large folio warning for FS-DAX

PATCH Superseded John Groves 4 participants 6 messages Dec 17, 2025 – Dec 19, 2025

Subsystems (3)

v1 of 2 Superseded Latest is v2 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 mm/memremap: fix spurious large folio warning for FS-DAX

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

The warning was introduced by commit that added support for large zone device private folios. However, that commit did not account for FS-DAX file-backed folios, which have always supported large (PMD-sized) mappings.

The check distinguishes between anonymous folios (which clear AnonExclusive flags for each sub-page) and file-backed folios. For file-backed folios, it assumes large folios are unexpected - but this assumption is incorrect for FS-DAX.

The fix is to exempt MEMORY_DEVICE_FS_DAX from the large folio warning, allowing FS-DAX to continue using PMD mappings without triggering false warnings.

---
=== How to reproduce ===

A reproducer is available at:

git clone https://github.com/jagalactic/dax-pmd-test.git
cd xfs-dax-test
make
sudo make test

This will set up XFS on pmem with 2MB stripe alignment and run a test that triggers the warning.

Alternatively, follow the manual steps below.

Prerequisites:
- Linux kernel with FS-DAX support and CONFIG_DEBUG_VM=y
- A pmem device (real or emulated)
- An fsdax namespace configured via ndctl as /dev/pmem0

Manual steps:

1. Create an fsdax namespace (if not already present):

   # ndctl create-namespace -m fsdax -e namespace0.0

2. Create XFS with 2MB stripe alignment:

   # mkfs.xfs -f -d su=2m,sw=1 /dev/pmem0
   # mount -o dax /dev/pmem0 /mnt/pmem

3. Compile and run the reproducer:

   # gcc -Wall -O2 -o dax_pmd_test dax_pmd_test.c
   # ./dax_pmd_test /mnt/pmem/testfile

4. Check dmesg for the warning:

   WARNING: mm/memremap.c:431 at free_zone_device_folio+0x.../0x...

Note: The 2MB stripe alignment (-d su=2m,sw=1) is critical. XFS normally allocates blocks at arbitrary offsets, causing PMD faults to fall back to PTE faults. The stripe alignment forces 2MB-aligned allocations, allowing PMD faults to succeed and exposing this bug.

=== Proposed fix ===

mm/memremap.c | 7 ++++++-

 1 file changed, 6 insertions(+), 1 deletion(-)
mm/memremap.c +6 -1
index 4c2e0d68eb27..af37c3b4e39b 100644
@@ -428,7 +428,12 @@ void free_zone_device_folio(struct folio *folio)
 		for (i = 0; i < nr; i++)
 			__ClearPageAnonExclusive(folio_page(folio, i));
 	} else {
-		VM_WARN_ON_ONCE(folio_test_large(folio));
+		/*
+		 * FS_DAX legitimately uses large file-mapped folios for
+		 * PMD mappings, so only warn for other device types.
+		 */
+		VM_WARN_ON_ONCE(pgmap->type != MEMORY_DEVICE_FS_DAX &&
+				folio_test_large(folio));
 	}
 
 	/*

base-commit: 8f0b4cce4481fb22653697cced8d0d04027cb1e8

Signature · 1 line
2.49.0
#1 lore raw

On 2025-12-18 at 08:13 +1100, John Groves <[email protected]> wrote...

Quoted 14 lines

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

Quoted 5 lines

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

Yep, and I'm pretty sure devdax can also create large folios so we might need a similar fix there. In fact looking at old vs. new code it seems we only ever used to have this warning for anon folios, which I think could only ever be true for DEVICE_PRIVATE or DEVICE_COHERENT folios.

So I suspect the proper fix is to just remove the warning entirely now that they also support compound sizes.

Quoted 4 lines

The warning was introduced by commit that added support for large zone device private folios. However, that commit did not account for FS-DAX file-backed folios, which have always supported large (PMD-sized) mappings.

Right, one of the nice side-effects (other than delaying fam-fs, sorry! :-/) of fixing the refcounting was that these started looking like normal large folios.

Quoted 8 lines

The check distinguishes between anonymous folios (which clear AnonExclusive flags for each sub-page) and file-backed folios. For file-backed folios, it assumes large folios are unexpected - but this assumption is incorrect for FS-DAX.

The fix is to exempt MEMORY_DEVICE_FS_DAX from the large folio warning, allowing FS-DAX to continue using PMD mappings without triggering false warnings.

As this is a fix you will want a "Fixes:" tag.

Quoted 34 lines

---
=== How to reproduce ===

A reproducer is available at:

git clone https://github.com/jagalactic/dax-pmd-test.git
cd xfs-dax-test
make
sudo make test

This will set up XFS on pmem with 2MB stripe alignment and run a test that triggers the warning.

Alternatively, follow the manual steps below.

Prerequisites:
- Linux kernel with FS-DAX support and CONFIG_DEBUG_VM=y
- A pmem device (real or emulated)
- An fsdax namespace configured via ndctl as /dev/pmem0

Manual steps:

1. Create an fsdax namespace (if not already present):

   # ndctl create-namespace -m fsdax -e namespace0.0

2. Create XFS with 2MB stripe alignment:

   # mkfs.xfs -f -d su=2m,sw=1 /dev/pmem0
   # mount -o dax /dev/pmem0 /mnt/pmem

3. Compile and run the reproducer:

   # gcc -Wall -O2 -o dax_pmd_test dax_pmd_test.c
   # ./dax_pmd_test /mnt/pmem/testfile
Quoted 7 lines

4. Check dmesg for the warning:

   WARNING: mm/memremap.c:431 at free_zone_device_folio+0x.../0x...

Note: The 2MB stripe alignment (-d su=2m,sw=1) is critical. XFS normally allocates blocks at arbitrary offsets, causing PMD faults to fall back to PTE faults. The stripe alignment forces 2MB-aligned allocations, allowing PMD faults to succeed and exposing this bug.

Thanks for the detailed repro instructions. Not always neccessary but definitely nice to have.

Quoted 28 lines

=== Proposed fix ===

mm/memremap.c | 7 ++++++-

 1 file changed, 6 insertions(+), 1 deletion(-)
mm/memremap.c +6 -1
index 4c2e0d68eb27..af37c3b4e39b 100644
@@ -428,7 +428,12 @@ void free_zone_device_folio(struct folio *folio)
 		for (i = 0; i < nr; i++)
 			__ClearPageAnonExclusive(folio_page(folio, i));
 	} else {
-		VM_WARN_ON_ONCE(folio_test_large(folio));
+		/*
+		 * FS_DAX legitimately uses large file-mapped folios for
+		 * PMD mappings, so only warn for other device types.
+		 */
+		VM_WARN_ON_ONCE(pgmap->type != MEMORY_DEVICE_FS_DAX &&
+				folio_test_large(folio));
 	}
 
 	/*

base-commit: 8f0b4cce4481fb22653697cced8d0d04027cb1e8

Signature · 1 line
2.49.0
maintainer
#2 lore raw
Quoted 21 lines · On Thu, 18 Dec 2025 09:58:02 +1100 Alistair Popple <[email protected]> wrote:

On 2025-12-18 at 08:13 +1100, John Groves <[email protected]> wrote...

Quoted 14 lines

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

Quoted 5 lines

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

Quoted 7 lines

Yep, and I'm pretty sure devdax can also create large folios so we might need a similar fix there. In fact looking at old vs. new code it seems we only ever used to have this warning for anon folios, which I think could only ever be true for DEVICE_PRIVATE or DEVICE_COHERENT folios.

So I suspect the proper fix is to just remove the warning entirely now that they also support compound sizes.

So I'm assuming we can expect an updated version of this fix.

Quoted 13 lines
Quoted 4 lines

The warning was introduced by commit that added support for large zone device private folios. However, that commit did not account for FS-DAX file-backed folios, which have always supported large (PMD-sized) mappings.

Right, one of the nice side-effects (other than delaying fam-fs, sorry! :-/) of fixing the refcounting was that these started looking like normal large folios.

Quoted 5 lines

The check distinguishes between anonymous folios (which clear AnonExclusive flags for each sub-page) and file-backed folios. For file-backed folios, it assumes large folios are unexpected - but this assumption is incorrect for FS-DAX.

Quoted 5 lines
Quoted 3 lines

The fix is to exempt MEMORY_DEVICE_FS_DAX from the large folio warning, allowing FS-DAX to continue using PMD mappings without triggering false warnings.

As this is a fix you will want a "Fixes:" tag.

Someone (possibly me) already added

Fixes d245f9b4ab80 ("mm/zone_device: support large zone device private folios")
#3 lore raw
Quoted 28 lines · Andrew Morton
Quoted 26 lines · On Thu, 18 Dec 2025 09:58:02 +1100 Alistair Popple <[email protected]> wrote:

On 2025-12-18 at 08:13 +1100, John Groves <[email protected]> wrote...

Quoted 14 lines

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

Quoted 5 lines

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

Yep, and I'm pretty sure devdax can also create large folios so we might need a similar fix there. In fact looking at old vs. new code it seems we only ever used to have this warning for anon folios, which I think could only ever be true for DEVICE_PRIVATE or DEVICE_COHERENT folios.

Quoted 4 lines
Quoted 2 lines

So I suspect the proper fix is to just remove the warning entirely now that they also support compound sizes.

So I'm assuming we can expect an updated version of this fix.

I'll send an update Friday morning

<snip>

Thanks Alistair, Dan and Andrew!

John

#4 lore raw
In reply to John Groves
Quoted 20 lines · John Groves

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

Quoted 4 lines

The warning was introduced by commit that added support for large zone device private folios. However, that commit did not account for FS-DAX file-backed folios, which have always supported large (PMD-sized) mappings.

Oh, I was not copied on:

d245f9b4ab80 mm/zone_device: support large zone device private folios

...I should probably add myself as a reviewer to the MEMORY HOT(UN)PLUG entry in MAINTAINERS at least for the mm/mememap.c bits.

Now, why is the warning there in the first place?

I.e. what is the risk of just doing this fixup:

mm/memremap.c +0 -2
index 4c2e0d68eb27..63c6ab4fdf08 100644
@@ -427,8 +427,6 @@ void free_zone_device_folio(struct folio *folio)
        if (folio_test_anon(folio)) {
                for (i = 0; i < nr; i++)
                        __ClearPageAnonExclusive(folio_page(folio, i));
-       } else {
-               VM_WARN_ON_ONCE(folio_test_large(folio));
        }
 
        /*
#5 lore raw

On 2025-12-18 at 10:59 +1100, [email protected] wrote...

Quoted 26 lines
Quoted 20 lines · John Groves

From: John Groves <[email protected]>

This patch addresses a warning that I discovered while working on famfs, which is an fs-dax file system that virtually always does PMD faults (next famfs patch series coming after the holidays).

However, XFS also does PMD faults in fs-dax mode, and it also triggers the warning. It takes some effort to get XFS to do a PMD fault, but instructions to reproduce it are below.

The VM_WARN_ON_ONCE(folio_test_large(folio)) check in free_zone_device_folio() incorrectly triggers for MEMORY_DEVICE_FS_DAX when PMD (2MB) mappings are used.

FS-DAX legitimately creates large file-backed folios when handling PMD faults. This is a core feature of FS-DAX that provides significant performance benefits by mapping 2MB regions directly to persistent memory. When these mappings are unmapped, the large folios are freed through free_zone_device_folio(), which triggers the spurious warning.

Quoted 4 lines

The warning was introduced by commit that added support for large zone device private folios. However, that commit did not account for FS-DAX file-backed folios, which have always supported large (PMD-sized) mappings.

Quoted 8 lines

Oh, I was not copied on:

d245f9b4ab80 mm/zone_device: support large zone device private folios

...I should probably add myself as a reviewer to the MEMORY HOT(UN)PLUG entry in MAINTAINERS at least for the mm/mememap.c bits.

Now, why is the warning there in the first place?

Lets wait for Balbir to comment but I suspect it's just a mistake in d245f9b4ab80 ("mm/zone_device: support large zone device private folios"):

- /*

-        * Note: we don't expect anonymous compound pages yet. Once supported
-        * and we could PTE-map them similar to THP, we'd have to clear
-        * PG_anon_exclusive on all tail pages.
-        */
        if (folio_test_anon(folio)) {
-               VM_BUG_ON_FOLIO(folio_test_large(folio), folio);
-               __ClearPageAnonExclusive(folio_page(folio, 0));
+               for (i = 0; i < nr; i++)
+                       __ClearPageAnonExclusive(folio_page(folio, i));
+       } else {
+               VM_WARN_ON_ONCE(folio_test_large(folio));
        }

The warning never applied to !folio_test_anon() folios and was just a reminder for the comment above which was deleted because it was fixed. Therefore the warning should also have been deleted.

Quoted 1 line

I.e. what is the risk of just doing this fixup:

None, I think that is the correct fix.

Quoted 14 lines
mm/memremap.c +0 -2
index 4c2e0d68eb27..63c6ab4fdf08 100644
@@ -427,8 +427,6 @@ void free_zone_device_folio(struct folio *folio)
        if (folio_test_anon(folio)) {
                for (i = 0; i < nr; i++)
                        __ClearPageAnonExclusive(folio_page(folio, i));
-       } else {
-               VM_WARN_ON_ONCE(folio_test_large(folio));
        }
 
        /*