[PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[Baolin Wang]

According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to
ignore any THP or mTHP interface settings.

However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics
for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will
depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit"
(although that is the default), which can cause MADV_COLLAPSE to fail
unexpectedly.

Fix this by returning the result of shmem_huge_global_enabled() directly
when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring
shmem THP/mTHP settings.

Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem")
Reported-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
Closes: https://lore.kernel.org/all/20260908125105.1510704-7-kirill@shutemov.name/
Reviewed-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
Acked-by: Zi Yan <ziy@nvidia.com>
Reviewed-by: Lance Yang <lance.yang@linux.dev>
Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
---
Changes from v1:
 - Use the correct fixes tag (per Kiryl).
 - Collect reviewed tags from Kiryl, Zi and Lance. Thanks.
---
 mm/shmem.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/mm/shmem.c b/mm/shmem.c
index c92ed17dbc4a..b572c60f2af8 100644
--- a/mm/shmem.c
+++ b/mm/shmem.c
@@ -2049,8 +2049,11 @@ unsigned long shmem_allowable_huge_orders(struct inode *inode,
 
 	global_orders = shmem_huge_global_enabled(inode, index, write_end,
 						  shmem_huge_force, vma, vm_flags);
-	/* Tmpfs huge pages allocation */
-	if (!vma || !vma_is_anon_shmem(vma))
+	/*
+	 * Tmpfs huge pages allocation or forced collapse ignores
+	 * sysfs configs.
+	 */
+	if (!vma || !vma_is_anon_shmem(vma) || shmem_huge_force)
 		return global_orders;
 
 	/*
-- 
2.47.3

Re: [PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[Barry Song]

On Mon, Sep 14, 2026 at 10:13 AM Baolin Wang
<baolin.wang@linux.alibaba.com> wrote:
>
> According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to
> ignore any THP or mTHP interface settings.
>
> However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics
> for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will
> depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit"
> (although that is the default), which can cause MADV_COLLAPSE to fail
> unexpectedly.
>
> Fix this by returning the result of shmem_huge_global_enabled() directly
> when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring
> shmem THP/mTHP settings.
>
> Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem")
> Reported-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
> Closes: https://lore.kernel.org/all/20260908125105.1510704-7-kirill@shutemov.name/
> Reviewed-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
> Acked-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lance Yang <lance.yang@linux.dev>
> Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---

LGTM,

Reviewed-by: Barry Song <baohua@kernel.org>

Re: [PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[Andrew Morton]

On Mon, 14 Sep 2026 10:12:41 +0800 Baolin Wang <baolin.wang@linux.alibaba.com> wrote:

> According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to
> ignore any THP or mTHP interface settings.
> 
> However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics
> for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will
> depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit"
> (although that is the default), which can cause MADV_COLLAPSE to fail
> unexpectedly.

Userspace-visible runtime regression?

> Fix this by returning the result of shmem_huge_global_enabled() directly
> when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring
> shmem THP/mTHP settings.
> 
> Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem")

Should we backport?

Re: [PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[Baolin Wang]

On 9/14/26 10:43 AM, Andrew Morton wrote:
> On Mon, 14 Sep 2026 10:12:41 +0800 Baolin Wang <baolin.wang@linux.alibaba.com> wrote:
> 
>> According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to
>> ignore any THP or mTHP interface settings.
>>
>> However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics
>> for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will
>> depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit"
>> (although that is the default), which can cause MADV_COLLAPSE to fail
>> unexpectedly.
> 
> Userspace-visible runtime regression?

Yes, it could cause a userspace-visible performance regression.

>> Fix this by returning the result of shmem_huge_global_enabled() directly
>> when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring
>> shmem THP/mTHP settings.
>>
>> Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem")
> 
> Should we backport?

I'm not sure, since the issue is triggered by a rather uncommon 
configuration. But given that the fix is simple, I think it's fine to CC 
stable.

Re: [PATCH v2] mm: shmem: ignore sysfs configs for shmem forced collapse

[Qi Zheng]

On 9/14/26 10:12 AM, Baolin Wang wrote:
> According to Documentation/mm/transhuge.rst, MADV_COLLAPSE is expected to
> ignore any THP or mTHP interface settings.
> 
> However, after commit 26c7d8413aaf ("mm: thp: support "THPeligible" semantics
> for mTHP with anonymous shmem"), performing MADV_COLLAPSE on shmem will
> depend on /sys/.../hugepages-2048kB/shmem_enabled being set to "inherit"
> (although that is the default), which can cause MADV_COLLAPSE to fail
> unexpectedly.
> 
> Fix this by returning the result of shmem_huge_global_enabled() directly
> when MADV_COLLAPSE is requested, allowing PMD-order collapse while ignoring
> shmem THP/mTHP settings.
> 
> Fixes: 26c7d8413aaf ("mm: thp: support "THPeligible" semantics for mTHP with anonymous shmem")
> Reported-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
> Closes: https://lore.kernel.org/all/20260908125105.1510704-7-kirill@shutemov.name/
> Reviewed-by: Kiryl Shutsemau (Meta) <kas@kernel.org>
> Acked-by: Zi Yan <ziy@nvidia.com>
> Reviewed-by: Lance Yang <lance.yang@linux.dev>
> Signed-off-by: Baolin Wang <baolin.wang@linux.alibaba.com>
> ---
> Changes from v1:
>   - Use the correct fixes tag (per Kiryl).
>   - Collect reviewed tags from Kiryl, Zi and Lance. Thanks.
> ---
>   mm/shmem.c | 7 +++++--
>   1 file changed, 5 insertions(+), 2 deletions(-)

Acked-by: Qi Zheng <qi.zheng@linux.dev>

Thanks,
Qi

> 
> diff --git a/mm/shmem.c b/mm/shmem.c
> index c92ed17dbc4a..b572c60f2af8 100644
> --- a/mm/shmem.c
> +++ b/mm/shmem.c
> @@ -2049,8 +2049,11 @@ unsigned long shmem_allowable_huge_orders(struct inode *inode,
>   
>   	global_orders = shmem_huge_global_enabled(inode, index, write_end,
>   						  shmem_huge_force, vma, vm_flags);
> -	/* Tmpfs huge pages allocation */
> -	if (!vma || !vma_is_anon_shmem(vma))
> +	/*
> +	 * Tmpfs huge pages allocation or forced collapse ignores
> +	 * sysfs configs.
> +	 */
> +	if (!vma || !vma_is_anon_shmem(vma) || shmem_huge_force)
>   		return global_orders;
>   
>   	/*