[PATCH v2 0/3] mm/memory: remove unused vmf_insert_mixed_mkwrite()

[PATCH v2 0/3] mm/memory: remove unused vmf_insert_mixed_mkwrite()

[Sang-Heon Jeon]

Hello,

Since commit 38607c62b34b ("fs/dax: properly refcount fs dax pages"),
vmf_insert_mixed_mkwrite() has no callers. This series removes it and
the mkwrite handling that only it used.

Patch 1 removes the function.

Patch 2 removes the mkwrite argument, which is then always false. It
also removes the unreachable code in insert_pfn().

Patch 3 merges __vm_insert_mixed() into vmf_insert_mixed().

No functional change.

---
Changes from v1 [1]
- Split into three patches for easier review
- Move the comment about -EBUSY handling above the check in
  __vm_insert_mixed()

[1] https://lore.kernel.org/all/20260929143707.450805-1-ekffu200098@gmail.com/
---

Sang-Heon Jeon (3):
  mm/memory: remove unused vmf_insert_mixed_mkwrite()
  mm/memory: remove the mkwrite argument from __vm_insert_mixed()
  mm/memory: merge __vm_insert_mixed() into vmf_insert_mixed()

 include/linux/mm.h |  2 --
 mm/memory.c        | 81 ++++++++++++++--------------------------------
 2 files changed, 24 insertions(+), 59 deletions(-)


base-commit: 1c2b8d2725f84b43fabe3b3e9628c91db8ca6c65
-- 
2.43.0

[PATCH v2 1/3] mm/memory: remove unused vmf_insert_mixed_mkwrite()

[Sang-Heon Jeon]

Since commit 38607c62b34b ("fs/dax: properly refcount fs dax pages"),
vmf_insert_mixed_mkwrite() has no callers.

So remove the function and its declaration.

The comment above the function describes the -EBUSY handling in
__vm_insert_mixed(). Move it above that check.

No functional change.

Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>
---
 include/linux/mm.h |  2 --
 mm/memory.c        | 16 +++++-----------
 2 files changed, 5 insertions(+), 13 deletions(-)

diff --git a/include/linux/mm.h b/include/linux/mm.h
index c038d06825c3..2ac734efd4ec 100644
--- a/include/linux/mm.h
+++ b/include/linux/mm.h
@@ -5008,8 +5008,6 @@ vm_fault_t vmf_insert_pfn_prot(struct vm_area_struct *vma, unsigned long addr,
 			unsigned long pfn, pgprot_t pgprot);
 vm_fault_t vmf_insert_mixed(struct vm_area_struct *vma, unsigned long addr,
 			unsigned long pfn);
-vm_fault_t vmf_insert_mixed_mkwrite(struct vm_area_struct *vma,
-		unsigned long addr, unsigned long pfn);
 int vm_iomap_memory(struct vm_area_struct *vma, phys_addr_t start, unsigned long len);
 
 static inline vm_fault_t vmf_insert_page(struct vm_area_struct *vma,
diff --git a/mm/memory.c b/mm/memory.c
index 330cde31bf8b..4b2e4be28031 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -3052,6 +3052,11 @@ static vm_fault_t __vm_insert_mixed(struct vm_area_struct *vma,
 
 	if (err == -ENOMEM)
 		return VM_FAULT_OOM;
+	/*
+	 *  If the insertion of PTE failed because someone else already added a
+	 *  different entry in the mean time, we treat that as success as we assume
+	 *  the same entry was actually inserted.
+	 */
 	if (err < 0 && err != -EBUSY)
 		return VM_FAULT_SIGBUS;
 
@@ -3085,17 +3090,6 @@ vm_fault_t vmf_insert_mixed(struct vm_area_struct *vma, unsigned long addr,
 }
 EXPORT_SYMBOL(vmf_insert_mixed);
 
-/*
- *  If the insertion of PTE failed because someone else already added a
- *  different entry in the mean time, we treat that as success as we assume
- *  the same entry was actually inserted.
- */
-vm_fault_t vmf_insert_mixed_mkwrite(struct vm_area_struct *vma,
-		unsigned long addr, unsigned long pfn)
-{
-	return __vm_insert_mixed(vma, addr, pfn, true);
-}
-
 /*
  * maps a range of physical memory into the requested pages. the old
  * mappings are removed. any references to nonexistent pages results
-- 
2.43.0

Re: [PATCH v2 1/3] mm/memory: remove unused vmf_insert_mixed_mkwrite()

[SJ Park]

On Sat,  3 Oct 2026 01:05:32 +0900 Sang-Heon Jeon <ekffu200098@gmail.com> wrote:

> Since commit 38607c62b34b ("fs/dax: properly refcount fs dax pages"),
> vmf_insert_mixed_mkwrite() has no callers.
> 
> So remove the function and its declaration.
> 
> The comment above the function describes the -EBUSY handling in
> __vm_insert_mixed(). Move it above that check.
> 
> No functional change.

Looks good to me.

> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

Reviewed-by: SJ Park <sj@kernel.org>


Thanks,
SJ

[...]

Re: [PATCH v2 1/3] mm/memory: remove unused vmf_insert_mixed_mkwrite()

[Jose A. Perez de Azpillaga]

On Sat, Oct 03, 2026 at 01:05:32AM +0900, Sang-Heon Jeon wrote:
> Since commit 38607c62b34b ("fs/dax: properly refcount fs dax pages"),
> vmf_insert_mixed_mkwrite() has no callers.
> 
> So remove the function and its declaration.
> 
> The comment above the function describes the -EBUSY handling in
> __vm_insert_mixed(). Move it above that check.
> 
> No functional change.
> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

all three LGMT so I'll RB them :)

Reviewed-by: Jose A. Perez de Azpillaga <azpijr@gmail.com>

-- 
cheers,
jose a. p-a

[PATCH v2 2/3] mm/memory: remove the mkwrite argument from __vm_insert_mixed()

[Sang-Heon Jeon]

__vm_insert_mixed() is only called from vmf_insert_mixed() with
mkwrite == false, so mkwrite is also always false in vm_mixed_ok()
and insert_pfn().

So remove the argument and the unreachable code in insert_pfn().

No functional change.

Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>
---
 mm/memory.c | 57 +++++++++++++++++------------------------------------
 1 file changed, 18 insertions(+), 39 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 4b2e4be28031..6c1d38eba7b2 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2453,7 +2453,14 @@ static int insert_page_into_pte_locked(struct vm_area_struct *vma, pte_t *pte,
 		if (!mkwrite)
 			return -EBUSY;
 
-		/* see insert_pfn(). */
+		/*
+		 * For read faults on private mappings the PFN passed in may
+		 * not match the PFN we have mapped if the mapped PFN is a
+		 * writeable COW page.  In the mkwrite case we are creating a
+		 * writable PTE for a shared mapping and we expect the PFNs to
+		 * match. If they don't match, we are likely racing with block
+		 * allocation and mapping invalidation.
+		 */
 		if (pte_pfn(pteval) != page_to_pfn(page)) {
 			WARN_ON_ONCE(!is_zero_pfn(pte_pfn(pteval)));
 			return -EFAULT;
@@ -2858,7 +2865,7 @@ int vm_map_pages_zero(struct vm_area_struct *vma, struct page **pages,
 EXPORT_SYMBOL(vm_map_pages_zero);
 
 static vm_fault_t insert_pfn(struct vm_area_struct *vma, unsigned long addr,
-			unsigned long pfn, pgprot_t prot, bool mkwrite)
+			unsigned long pfn, pgprot_t prot)
 {
 	struct mm_struct *mm = vma->vm_mm;
 	pte_t *pte, entry;
@@ -2868,38 +2875,12 @@ static vm_fault_t insert_pfn(struct vm_area_struct *vma, unsigned long addr,
 	if (!pte)
 		return VM_FAULT_OOM;
 	entry = ptep_get(pte);
-	if (!pte_none(entry)) {
-		if (mkwrite) {
-			/*
-			 * For read faults on private mappings the PFN passed
-			 * in may not match the PFN we have mapped if the
-			 * mapped PFN is a writeable COW page.  In the mkwrite
-			 * case we are creating a writable PTE for a shared
-			 * mapping and we expect the PFNs to match. If they
-			 * don't match, we are likely racing with block
-			 * allocation and mapping invalidation so just skip the
-			 * update.
-			 */
-			if (pte_pfn(entry) != pfn) {
-				WARN_ON_ONCE(!is_zero_pfn(pte_pfn(entry)));
-				goto out_unlock;
-			}
-			entry = pte_mkyoung(entry);
-			entry = maybe_mkwrite(pte_mkdirty(entry), vma);
-			if (ptep_set_access_flags(vma, addr, pte, entry, 1))
-				update_mmu_cache(vma, addr, pte);
-		}
+	if (!pte_none(entry))
 		goto out_unlock;
-	}
 
 	/* Ok, finally just insert the thing.. */
 	entry = pte_mkspecial(pfn_pte(pfn, prot));
 
-	if (mkwrite) {
-		entry = pte_mkyoung(entry);
-		entry = maybe_mkwrite(pte_mkdirty(entry), vma);
-	}
-
 	set_pte_at(mm, addr, pte, entry);
 	update_mmu_cache(vma, addr, pte); /* XXX: why not for insert_page? */
 
@@ -2967,7 +2948,7 @@ vm_fault_t vmf_insert_pfn_prot(struct vm_area_struct *vma, unsigned long addr,
 
 	pfnmap_setup_cachemode_pfn(pfn, &pgprot);
 
-	return insert_pfn(vma, addr, pfn, pgprot, false);
+	return insert_pfn(vma, addr, pfn, pgprot);
 }
 EXPORT_SYMBOL(vmf_insert_pfn_prot);
 
@@ -2998,11 +2979,9 @@ vm_fault_t vmf_insert_pfn(struct vm_area_struct *vma, unsigned long addr,
 }
 EXPORT_SYMBOL(vmf_insert_pfn);
 
-static bool vm_mixed_ok(struct vm_area_struct *vma, unsigned long pfn,
-			bool mkwrite)
+static bool vm_mixed_ok(struct vm_area_struct *vma, unsigned long pfn)
 {
-	if (unlikely(is_zero_pfn(pfn)) &&
-	    (mkwrite || !vm_mixed_zeropage_allowed(vma)))
+	if (unlikely(is_zero_pfn(pfn)) && !vm_mixed_zeropage_allowed(vma))
 		return false;
 	/* these checks mirror the abort conditions in vm_normal_page */
 	if (vma->vm_flags & VM_MIXEDMAP)
@@ -3013,12 +2992,12 @@ static bool vm_mixed_ok(struct vm_area_struct *vma, unsigned long pfn,
 }
 
 static vm_fault_t __vm_insert_mixed(struct vm_area_struct *vma,
-		unsigned long addr, unsigned long pfn, bool mkwrite)
+		unsigned long addr, unsigned long pfn)
 {
 	pgprot_t pgprot = vma->vm_page_prot;
 	int err;
 
-	if (!vm_mixed_ok(vma, pfn, mkwrite))
+	if (!vm_mixed_ok(vma, pfn))
 		return VM_FAULT_SIGBUS;
 
 	if (addr < vma->vm_start || addr >= vma->vm_end)
@@ -3045,9 +3024,9 @@ static vm_fault_t __vm_insert_mixed(struct vm_area_struct *vma,
 		 * result in pfn_t_has_page() == false.
 		 */
 		page = pfn_to_page(pfn);
-		err = insert_page(vma, addr, page, pgprot, mkwrite);
+		err = insert_page(vma, addr, page, pgprot, false);
 	} else {
-		return insert_pfn(vma, addr, pfn, pgprot, mkwrite);
+		return insert_pfn(vma, addr, pfn, pgprot);
 	}
 
 	if (err == -ENOMEM)
@@ -3086,7 +3065,7 @@ EXPORT_SYMBOL_GPL(vmf_insert_page_mkwrite);
 vm_fault_t vmf_insert_mixed(struct vm_area_struct *vma, unsigned long addr,
 		unsigned long pfn)
 {
-	return __vm_insert_mixed(vma, addr, pfn, false);
+	return __vm_insert_mixed(vma, addr, pfn);
 }
 EXPORT_SYMBOL(vmf_insert_mixed);
 
-- 
2.43.0

Re: [PATCH v2 2/3] mm/memory: remove the mkwrite argument from __vm_insert_mixed()

[SJ Park]

On Sat,  3 Oct 2026 01:05:33 +0900 Sang-Heon Jeon <ekffu200098@gmail.com> wrote:

> __vm_insert_mixed() is only called from vmf_insert_mixed() with
> mkwrite == false, so mkwrite is also always false in vm_mixed_ok()
> and insert_pfn().
> 
> So remove the argument and the unreachable code in insert_pfn().
> 
> No functional change.

Looks correct to me.

> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

Reviewed-by: SJ Park <sj@kernel.org>


Thanks,
SJ

[...]

Re: [PATCH v2 2/3] mm/memory: remove the mkwrite argument from __vm_insert_mixed()

[Jose A. Perez de Azpillaga]

On Sat, Oct 03, 2026 at 01:05:33AM +0900, Sang-Heon Jeon wrote:
> __vm_insert_mixed() is only called from vmf_insert_mixed() with
> mkwrite == false, so mkwrite is also always false in vm_mixed_ok()
> and insert_pfn().
> 
> So remove the argument and the unreachable code in insert_pfn().
> 
> No functional change.
> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

Reviewed-by: Jose A. Perez de Azpillaga <azpijr@gmail.com>

-- 
cheers,
jose a. p-a

[PATCH v2 3/3] mm/memory: merge __vm_insert_mixed() into vmf_insert_mixed()

[Sang-Heon Jeon]

vmf_insert_mixed() just calls __vm_insert_mixed() with the same
arguments.

So merge __vm_insert_mixed() into vmf_insert_mixed().

No functional change.

Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>
---
 mm/memory.c | 12 +++---------
 1 file changed, 3 insertions(+), 9 deletions(-)

diff --git a/mm/memory.c b/mm/memory.c
index 6c1d38eba7b2..bde650e207c9 100644
--- a/mm/memory.c
+++ b/mm/memory.c
@@ -2991,8 +2991,8 @@ static bool vm_mixed_ok(struct vm_area_struct *vma, unsigned long pfn)
 	return false;
 }
 
-static vm_fault_t __vm_insert_mixed(struct vm_area_struct *vma,
-		unsigned long addr, unsigned long pfn)
+vm_fault_t vmf_insert_mixed(struct vm_area_struct *vma, unsigned long addr,
+		unsigned long pfn)
 {
 	pgprot_t pgprot = vma->vm_page_prot;
 	int err;
@@ -3041,6 +3041,7 @@ static vm_fault_t __vm_insert_mixed(struct vm_area_struct *vma,
 
 	return VM_FAULT_NOPAGE;
 }
+EXPORT_SYMBOL(vmf_insert_mixed);
 
 vm_fault_t vmf_insert_page_mkwrite(struct vm_fault *vmf, struct page *page,
 			bool write)
@@ -3062,13 +3063,6 @@ vm_fault_t vmf_insert_page_mkwrite(struct vm_fault *vmf, struct page *page,
 }
 EXPORT_SYMBOL_GPL(vmf_insert_page_mkwrite);
 
-vm_fault_t vmf_insert_mixed(struct vm_area_struct *vma, unsigned long addr,
-		unsigned long pfn)
-{
-	return __vm_insert_mixed(vma, addr, pfn);
-}
-EXPORT_SYMBOL(vmf_insert_mixed);
-
 /*
  * maps a range of physical memory into the requested pages. the old
  * mappings are removed. any references to nonexistent pages results
-- 
2.43.0

Re: [PATCH v2 3/3] mm/memory: merge __vm_insert_mixed() into vmf_insert_mixed()

[SJ Park]

On Sat,  3 Oct 2026 01:05:34 +0900 Sang-Heon Jeon <ekffu200098@gmail.com> wrote:

> vmf_insert_mixed() just calls __vm_insert_mixed() with the same
> arguments.
> 
> So merge __vm_insert_mixed() into vmf_insert_mixed().
> 
> No functional change.

Makes sense to me.

> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

Reviewed-by: SJ Park <sj@kernel.org>


Thanks,
SJ

[...]

Re: [PATCH v2 3/3] mm/memory: merge __vm_insert_mixed() into vmf_insert_mixed()

[Jose A. Perez de Azpillaga]

On Sat, Oct 03, 2026 at 01:05:34AM +0900, Sang-Heon Jeon wrote:
> vmf_insert_mixed() just calls __vm_insert_mixed() with the same
> arguments.
> 
> So merge __vm_insert_mixed() into vmf_insert_mixed().
> 
> No functional change.
> 
> Signed-off-by: Sang-Heon Jeon <ekffu200098@gmail.com>

Reviewed-by: Jose A. Perez de Azpillaga <azpijr@gmail.com>

-- 
cheers,
jose a. p-a