[PATCH 0/2] mm/ptdump: Drop assumption that pxd_val() is u64

[PATCH 0/2] mm/ptdump: Drop assumption that pxd_val() is u64

[Anshuman Khandual]

Last argument passed down in note_page() is u64 assuming pxd_val() returned
value (all page table levels) is 64 bit - which might not be the case going
ahead when D128 page tables is enabled on arm64 platform. Besides pxd_val()
is very platform specific and its type should not be assumed in generic MM.

This series splits note_page() into individual page table level specific
callbacks which accepts corresponding pxd_t page table entry as an argument
instead and later all subscribing platforms could derive pxd_val() from the
entries as required and proceed as before.

Page table entry's value, mask and protection are represented with pteval_t
not u64 that has been assumed while dumping the page table entries on arm64
platform. Replace such u64 instances with pteval_t instead as expected.

This series has been tested on arm64 platform but it does build on other
relevant platforms (v6.14-rc7).

Changes in V1:

- Added note_page_flush() callback and implemented the same on all
  subscribing platforms
- Moved note_page() argument change from u64 to pteval_t on arm64 platform
  from second patch to the first patch instead

Changes in RFC:

https://lore.kernel.org/all/20250310095902.390664-1-anshuman.khandual@arm.com/

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Steven Price <steven.price@arm.com>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-riscv@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: linux-mm@kvack.org

Anshuman Khandual (2):
  mm/ptdump: Split note_page() into level specific callbacks
  arm64/ptdump: Replace u64 with pteval_t

 arch/arm64/include/asm/ptdump.h | 24 ++++++++++++----
 arch/arm64/mm/ptdump.c          | 50 ++++++++++++++++++++++++++++++---
 arch/powerpc/mm/ptdump/ptdump.c | 46 ++++++++++++++++++++++++++++--
 arch/riscv/mm/ptdump.c          | 46 ++++++++++++++++++++++++++++--
 arch/s390/mm/dump_pagetables.c  | 46 ++++++++++++++++++++++++++++--
 arch/x86/mm/dump_pagetables.c   | 39 ++++++++++++++++++++++++-
 include/linux/ptdump.h          |  9 ++++--
 mm/ptdump.c                     | 42 ++++++++++++++++++++-------
 8 files changed, 272 insertions(+), 30 deletions(-)

-- 
2.25.1

[PATCH 1/2] mm/ptdump: Split note_page() into level specific callbacks

[Anshuman Khandual]

Last argument passed down in note_page() is u64 assuming pxd_val() returned
value (all page table levels) is 64 bit - which might not be the case going
ahead when D128 page tables is enabled on arm64 platform. Besides pxd_val()
is very platform specific and its type should not be assumed in generic MM.

Split note_page() into individual page table level specific callbacks which
accepts corresponding pxd_t argument instead and then subscribing platforms
just derive pxd_val() from the entries as required and proceed as earlier.

Also add a note_page_flush() callback for flushing the last page table page
that was being handled earlier via level = -1.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
Cc: Nicholas Piggin <npiggin@gmail.com>
Cc: Paul Walmsley <paul.walmsley@sifive.com>
Cc: Palmer Dabbelt <palmer@dabbelt.com>
Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
Cc: Heiko Carstens <hca@linux.ibm.com>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: Ingo Molnar <mingo@redhat.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linuxppc-dev@lists.ozlabs.org
Cc: linux-riscv@lists.infradead.org
Cc: linux-s390@vger.kernel.org
Cc: linux-mm@kvack.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
 arch/arm64/include/asm/ptdump.h | 16 +++++++++--
 arch/arm64/mm/ptdump.c          | 48 ++++++++++++++++++++++++++++++---
 arch/powerpc/mm/ptdump/ptdump.c | 46 +++++++++++++++++++++++++++++--
 arch/riscv/mm/ptdump.c          | 46 +++++++++++++++++++++++++++++--
 arch/s390/mm/dump_pagetables.c  | 46 +++++++++++++++++++++++++++++--
 arch/x86/mm/dump_pagetables.c   | 39 ++++++++++++++++++++++++++-
 include/linux/ptdump.h          |  9 ++++---
 mm/ptdump.c                     | 42 ++++++++++++++++++++++-------
 8 files changed, 267 insertions(+), 25 deletions(-)

diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
index 6cf4aae05219..e5da9ce8a515 100644
--- a/arch/arm64/include/asm/ptdump.h
+++ b/arch/arm64/include/asm/ptdump.h
@@ -59,7 +59,13 @@ struct ptdump_pg_state {
 
 void ptdump_walk(struct seq_file *s, struct ptdump_info *info);
 void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
-	       u64 val);
+	       pteval_t val);
+void note_page_pte(struct ptdump_state *st, unsigned long addr, pte_t pte);
+void note_page_pmd(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
+void note_page_pud(struct ptdump_state *st, unsigned long addr, pud_t pud);
+void note_page_p4d(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
+void note_page_pgd(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
+void note_page_flush(struct ptdump_state *st);
 #ifdef CONFIG_PTDUMP_DEBUGFS
 #define EFI_RUNTIME_MAP_END	DEFAULT_MAP_WINDOW_64
 void __init ptdump_debugfs_register(struct ptdump_info *info, const char *name);
@@ -69,7 +75,13 @@ static inline void ptdump_debugfs_register(struct ptdump_info *info,
 #endif /* CONFIG_PTDUMP_DEBUGFS */
 #else
 static inline void note_page(struct ptdump_state *pt_st, unsigned long addr,
-			     int level, u64 val) { }
+			     int level, pteval_t val) { }
+static inline void note_page_pte(struct ptdump_state *st, unsigned long addr, pte_t pte) { }
+static inline void note_page_pmd(struct ptdump_state *st, unsigned long addr, pmd_t pmd) { }
+static inline void note_page_pud(struct ptdump_state *st, unsigned long addr, pud_t pud) { }
+static inline void note_page_p4d(struct ptdump_state *st, unsigned long addr, p4d_t p4d) { }
+static inline void note_page_pgd(struct ptdump_state *st, unsigned long addr, pgd_t pgd) { }
+static inline void note_page_flush(struct ptdump_state *st) { }
 #endif /* CONFIG_PTDUMP_CORE */
 
 #endif /* __ASM_PTDUMP_H */
diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
index 688fbe0271ca..fd1610b4fd15 100644
--- a/arch/arm64/mm/ptdump.c
+++ b/arch/arm64/mm/ptdump.c
@@ -189,7 +189,7 @@ static void note_prot_wx(struct ptdump_pg_state *st, unsigned long addr)
 }
 
 void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
-	       u64 val)
+	       pteval_t val)
 {
 	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
 	struct ptdump_pg_level *pg_level = st->pg_level;
@@ -251,6 +251,38 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
 
 }
 
+void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
+{
+	note_page(pt_st, addr, 4, pte_val(pte));
+}
+
+void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
+{
+	note_page(pt_st, addr, 3, pmd_val(pmd));
+}
+
+void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
+{
+	note_page(pt_st, addr, 2, pud_val(pud));
+}
+
+void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
+{
+	note_page(pt_st, addr, 1, p4d_val(p4d));
+}
+
+void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
+{
+	note_page(pt_st, addr, 0, pgd_val(pgd));
+}
+
+void note_page_flush(struct ptdump_state *pt_st)
+{
+	pte_t pte_zero = {0};
+
+	note_page(pt_st, 0, -1, pte_val(pte_zero));
+}
+
 void ptdump_walk(struct seq_file *s, struct ptdump_info *info)
 {
 	unsigned long end = ~0UL;
@@ -266,7 +298,12 @@ void ptdump_walk(struct seq_file *s, struct ptdump_info *info)
 		.pg_level = &kernel_pg_levels[0],
 		.level = -1,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]){
 				{info->base_addr, end},
 				{0, 0}
@@ -303,7 +340,12 @@ bool ptdump_check_wx(void)
 		.level = -1,
 		.check_wx = true,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]) {
 				{_PAGE_OFFSET(vabits_actual), ~0UL},
 				{0, 0}
diff --git a/arch/powerpc/mm/ptdump/ptdump.c b/arch/powerpc/mm/ptdump/ptdump.c
index 9dc239967b77..b2358d794855 100644
--- a/arch/powerpc/mm/ptdump/ptdump.c
+++ b/arch/powerpc/mm/ptdump/ptdump.c
@@ -298,6 +298,38 @@ static void populate_markers(void)
 #endif
 }
 
+static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
+{
+	note_page(pt_st, addr, 4, pte_val(pte));
+}
+
+static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
+{
+	note_page(pt_st, addr, 3, pmd_val(pmd));
+}
+
+static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
+{
+	note_page(pt_st, addr, 2, pud_val(pud));
+}
+
+static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
+{
+	note_page(pt_st, addr, 1, p4d_val(p4d));
+}
+
+static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
+{
+	note_page(pt_st, addr, 0, pgd_val(pgd));
+}
+
+static void note_page_flush(struct ptdump_state *pt_st)
+{
+	pte_t pte_zero = {0};
+
+	note_page(pt_st, 0, -1, pte_val(pte_zero));
+}
+
 static int ptdump_show(struct seq_file *m, void *v)
 {
 	struct pg_state st = {
@@ -305,7 +337,12 @@ static int ptdump_show(struct seq_file *m, void *v)
 		.marker = address_markers,
 		.level = -1,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = ptdump_range,
 		}
 	};
@@ -338,7 +375,12 @@ bool ptdump_check_wx(void)
 		.level = -1,
 		.check_wx = true,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = ptdump_range,
 		}
 	};
diff --git a/arch/riscv/mm/ptdump.c b/arch/riscv/mm/ptdump.c
index 9d5f657a251b..32922550a50a 100644
--- a/arch/riscv/mm/ptdump.c
+++ b/arch/riscv/mm/ptdump.c
@@ -318,6 +318,38 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr,
 	}
 }
 
+static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
+{
+	note_page(pt_st, addr, 4, pte_val(pte));
+}
+
+static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
+{
+	note_page(pt_st, addr, 3, pmd_val(pmd));
+}
+
+static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
+{
+	note_page(pt_st, addr, 2, pud_val(pud));
+}
+
+static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
+{
+	note_page(pt_st, addr, 1, p4d_val(p4d));
+}
+
+static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
+{
+	note_page(pt_st, addr, 0, pgd_val(pgd));
+}
+
+static void note_page_flush(struct ptdump_state *pt_st)
+{
+	pte_t pte_zero = {0};
+
+	note_page(pt_st, 0, -1, pte_val(pte_zero));
+}
+
 static void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo)
 {
 	struct pg_state st = {
@@ -325,7 +357,12 @@ static void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo)
 		.marker = pinfo->markers,
 		.level = -1,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]) {
 				{pinfo->base_addr, pinfo->end},
 				{0, 0}
@@ -347,7 +384,12 @@ bool ptdump_check_wx(void)
 		.level = -1,
 		.check_wx = true,
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]) {
 				{KERN_VIRT_START, ULONG_MAX},
 				{0, 0}
diff --git a/arch/s390/mm/dump_pagetables.c b/arch/s390/mm/dump_pagetables.c
index fa54f3bc0c8d..e7cc52542f71 100644
--- a/arch/s390/mm/dump_pagetables.c
+++ b/arch/s390/mm/dump_pagetables.c
@@ -145,11 +145,48 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
 	}
 }
 
+static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
+{
+	note_page(pt_st, addr, 4, pte_val(pte));
+}
+
+static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
+{
+	note_page(pt_st, addr, 3, pmd_val(pmd));
+}
+
+static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
+{
+	note_page(pt_st, addr, 2, pud_val(pud));
+}
+
+static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
+{
+	note_page(pt_st, addr, 1, p4d_val(p4d));
+}
+
+static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
+{
+	note_page(pt_st, addr, 0, pgd_val(pgd));
+}
+
+static void note_page_flush(struct ptdump_state *pt_st)
+{
+	pte_t pte_zero = {0};
+
+	note_page(pt_st, 0, -1, pte_val(pte_zero));
+}
+
 bool ptdump_check_wx(void)
 {
 	struct pg_state st = {
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]) {
 				{.start = 0, .end = max_addr},
 				{.start = 0, .end = 0},
@@ -188,7 +225,12 @@ static int ptdump_show(struct seq_file *m, void *v)
 {
 	struct pg_state st = {
 		.ptdump = {
-			.note_page = note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.range = (struct ptdump_range[]) {
 				{.start = 0, .end = max_addr},
 				{.start = 0, .end = 0},
diff --git a/arch/x86/mm/dump_pagetables.c b/arch/x86/mm/dump_pagetables.c
index 89079ea73e65..2e1c2d006ace 100644
--- a/arch/x86/mm/dump_pagetables.c
+++ b/arch/x86/mm/dump_pagetables.c
@@ -362,6 +362,38 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
 	}
 }
 
+static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
+{
+	note_page(pt_st, addr, 4, pte_val(pte));
+}
+
+static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
+{
+	note_page(pt_st, addr, 3, pmd_val(pmd));
+}
+
+static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
+{
+	note_page(pt_st, addr, 2, pud_val(pud));
+}
+
+static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
+{
+	note_page(pt_st, addr, 1, p4d_val(p4d));
+}
+
+static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
+{
+	note_page(pt_st, addr, 0, pgd_val(pgd));
+}
+
+static void note_page_flush(struct ptdump_state *pt_st)
+{
+	pte_t pte_zero = {0};
+
+	note_page(pt_st, 0, -1, pte_val(pte_zero));
+}
+
 bool ptdump_walk_pgd_level_core(struct seq_file *m,
 				struct mm_struct *mm, pgd_t *pgd,
 				bool checkwx, bool dmesg)
@@ -378,7 +410,12 @@ bool ptdump_walk_pgd_level_core(struct seq_file *m,
 
 	struct pg_state st = {
 		.ptdump = {
-			.note_page	= note_page,
+			.note_page_pte = note_page_pte,
+			.note_page_pmd = note_page_pmd,
+			.note_page_pud = note_page_pud,
+			.note_page_p4d = note_page_p4d,
+			.note_page_pgd = note_page_pgd,
+			.note_page_flush = note_page_flush,
 			.effective_prot = effective_prot,
 			.range		= ptdump_ranges
 		},
diff --git a/include/linux/ptdump.h b/include/linux/ptdump.h
index 8dbd51ea8626..1c1eb1fae199 100644
--- a/include/linux/ptdump.h
+++ b/include/linux/ptdump.h
@@ -11,9 +11,12 @@ struct ptdump_range {
 };
 
 struct ptdump_state {
-	/* level is 0:PGD to 4:PTE, or -1 if unknown */
-	void (*note_page)(struct ptdump_state *st, unsigned long addr,
-			  int level, u64 val);
+	void (*note_page_pte)(struct ptdump_state *st, unsigned long addr, pte_t pte);
+	void (*note_page_pmd)(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
+	void (*note_page_pud)(struct ptdump_state *st, unsigned long addr, pud_t pud);
+	void (*note_page_p4d)(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
+	void (*note_page_pgd)(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
+	void (*note_page_flush)(struct ptdump_state *st);
 	void (*effective_prot)(struct ptdump_state *st, int level, u64 val);
 	const struct ptdump_range *range;
 };
diff --git a/mm/ptdump.c b/mm/ptdump.c
index 106e1d66e9f9..706cfc19439b 100644
--- a/mm/ptdump.c
+++ b/mm/ptdump.c
@@ -18,7 +18,7 @@ static inline int note_kasan_page_table(struct mm_walk *walk,
 {
 	struct ptdump_state *st = walk->private;
 
-	st->note_page(st, addr, 4, pte_val(kasan_early_shadow_pte[0]));
+	st->note_page_pte(st, addr, kasan_early_shadow_pte[0]);
 
 	walk->action = ACTION_CONTINUE;
 
@@ -42,7 +42,7 @@ static int ptdump_pgd_entry(pgd_t *pgd, unsigned long addr,
 		st->effective_prot(st, 0, pgd_val(val));
 
 	if (pgd_leaf(val)) {
-		st->note_page(st, addr, 0, pgd_val(val));
+		st->note_page_pgd(st, addr, val);
 		walk->action = ACTION_CONTINUE;
 	}
 
@@ -65,7 +65,7 @@ static int ptdump_p4d_entry(p4d_t *p4d, unsigned long addr,
 		st->effective_prot(st, 1, p4d_val(val));
 
 	if (p4d_leaf(val)) {
-		st->note_page(st, addr, 1, p4d_val(val));
+		st->note_page_p4d(st, addr, val);
 		walk->action = ACTION_CONTINUE;
 	}
 
@@ -88,7 +88,7 @@ static int ptdump_pud_entry(pud_t *pud, unsigned long addr,
 		st->effective_prot(st, 2, pud_val(val));
 
 	if (pud_leaf(val)) {
-		st->note_page(st, addr, 2, pud_val(val));
+		st->note_page_pud(st, addr, val);
 		walk->action = ACTION_CONTINUE;
 	}
 
@@ -109,7 +109,7 @@ static int ptdump_pmd_entry(pmd_t *pmd, unsigned long addr,
 	if (st->effective_prot)
 		st->effective_prot(st, 3, pmd_val(val));
 	if (pmd_leaf(val)) {
-		st->note_page(st, addr, 3, pmd_val(val));
+		st->note_page_pmd(st, addr, val);
 		walk->action = ACTION_CONTINUE;
 	}
 
@@ -125,7 +125,7 @@ static int ptdump_pte_entry(pte_t *pte, unsigned long addr,
 	if (st->effective_prot)
 		st->effective_prot(st, 4, pte_val(val));
 
-	st->note_page(st, addr, 4, pte_val(val));
+	st->note_page_pte(st, addr, val);
 
 	return 0;
 }
@@ -134,9 +134,31 @@ static int ptdump_hole(unsigned long addr, unsigned long next,
 		       int depth, struct mm_walk *walk)
 {
 	struct ptdump_state *st = walk->private;
-
-	st->note_page(st, addr, depth, 0);
-
+	pte_t pte_zero = {0};
+	pmd_t pmd_zero = {0};
+	pud_t pud_zero = {0};
+	p4d_t p4d_zero = {0};
+	pgd_t pgd_zero = {0};
+
+	switch (depth) {
+	case 4:
+		st->note_page_pte(st, addr, pte_zero);
+		break;
+	case 3:
+		st->note_page_pmd(st, addr, pmd_zero);
+		break;
+	case 2:
+		st->note_page_pud(st, addr, pud_zero);
+		break;
+	case 1:
+		st->note_page_p4d(st, addr, p4d_zero);
+		break;
+	case 0:
+		st->note_page_pgd(st, addr, pgd_zero);
+		break;
+	default:
+		break;
+	}
 	return 0;
 }
 
@@ -162,7 +184,7 @@ void ptdump_walk_pgd(struct ptdump_state *st, struct mm_struct *mm, pgd_t *pgd)
 	mmap_write_unlock(mm);
 
 	/* Flush out the last page */
-	st->note_page(st, 0, -1, 0);
+	st->note_page_flush(st);
 }
 
 static int check_wx_show(struct seq_file *m, void *v)
-- 
2.25.1

Re: [PATCH 1/2] mm/ptdump: Split note_page() into level specific callbacks

[Anshuman Khandual]

On 3/17/25 11:48, Anshuman Khandual wrote:
> Last argument passed down in note_page() is u64 assuming pxd_val() returned
> value (all page table levels) is 64 bit - which might not be the case going
> ahead when D128 page tables is enabled on arm64 platform. Besides pxd_val()
> is very platform specific and its type should not be assumed in generic MM.
> 
> Split note_page() into individual page table level specific callbacks which
> accepts corresponding pxd_t argument instead and then subscribing platforms
> just derive pxd_val() from the entries as required and proceed as earlier.
> 
> Also add a note_page_flush() callback for flushing the last page table page
> that was being handled earlier via level = -1.

Does this look good OR is there any concern OR an alternate approach ?

> 
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will@kernel.org>
> Cc: Madhavan Srinivasan <maddy@linux.ibm.com>
> Cc: Nicholas Piggin <npiggin@gmail.com>
> Cc: Paul Walmsley <paul.walmsley@sifive.com>
> Cc: Palmer Dabbelt <palmer@dabbelt.com>
> Cc: Gerald Schaefer <gerald.schaefer@linux.ibm.com>
> Cc: Heiko Carstens <hca@linux.ibm.com>
> Cc: Peter Zijlstra <peterz@infradead.org>
> Cc: Thomas Gleixner <tglx@linutronix.de>
> Cc: Ingo Molnar <mingo@redhat.com>
> Cc: Andrew Morton <akpm@linux-foundation.org>
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: linux-kernel@vger.kernel.org
> Cc: linuxppc-dev@lists.ozlabs.org
> Cc: linux-riscv@lists.infradead.org
> Cc: linux-s390@vger.kernel.org
> Cc: linux-mm@kvack.org
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
> ---
>  arch/arm64/include/asm/ptdump.h | 16 +++++++++--
>  arch/arm64/mm/ptdump.c          | 48 ++++++++++++++++++++++++++++++---
>  arch/powerpc/mm/ptdump/ptdump.c | 46 +++++++++++++++++++++++++++++--
>  arch/riscv/mm/ptdump.c          | 46 +++++++++++++++++++++++++++++--
>  arch/s390/mm/dump_pagetables.c  | 46 +++++++++++++++++++++++++++++--
>  arch/x86/mm/dump_pagetables.c   | 39 ++++++++++++++++++++++++++-
>  include/linux/ptdump.h          |  9 ++++---
>  mm/ptdump.c                     | 42 ++++++++++++++++++++++-------
>  8 files changed, 267 insertions(+), 25 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
> index 6cf4aae05219..e5da9ce8a515 100644
> --- a/arch/arm64/include/asm/ptdump.h
> +++ b/arch/arm64/include/asm/ptdump.h
> @@ -59,7 +59,13 @@ struct ptdump_pg_state {
>  
>  void ptdump_walk(struct seq_file *s, struct ptdump_info *info);
>  void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
> -	       u64 val);
> +	       pteval_t val);
> +void note_page_pte(struct ptdump_state *st, unsigned long addr, pte_t pte);
> +void note_page_pmd(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
> +void note_page_pud(struct ptdump_state *st, unsigned long addr, pud_t pud);
> +void note_page_p4d(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
> +void note_page_pgd(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
> +void note_page_flush(struct ptdump_state *st);
>  #ifdef CONFIG_PTDUMP_DEBUGFS
>  #define EFI_RUNTIME_MAP_END	DEFAULT_MAP_WINDOW_64
>  void __init ptdump_debugfs_register(struct ptdump_info *info, const char *name);
> @@ -69,7 +75,13 @@ static inline void ptdump_debugfs_register(struct ptdump_info *info,
>  #endif /* CONFIG_PTDUMP_DEBUGFS */
>  #else
>  static inline void note_page(struct ptdump_state *pt_st, unsigned long addr,
> -			     int level, u64 val) { }
> +			     int level, pteval_t val) { }
> +static inline void note_page_pte(struct ptdump_state *st, unsigned long addr, pte_t pte) { }
> +static inline void note_page_pmd(struct ptdump_state *st, unsigned long addr, pmd_t pmd) { }
> +static inline void note_page_pud(struct ptdump_state *st, unsigned long addr, pud_t pud) { }
> +static inline void note_page_p4d(struct ptdump_state *st, unsigned long addr, p4d_t p4d) { }
> +static inline void note_page_pgd(struct ptdump_state *st, unsigned long addr, pgd_t pgd) { }
> +static inline void note_page_flush(struct ptdump_state *st) { }
>  #endif /* CONFIG_PTDUMP_CORE */
>  
>  #endif /* __ASM_PTDUMP_H */
> diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
> index 688fbe0271ca..fd1610b4fd15 100644
> --- a/arch/arm64/mm/ptdump.c
> +++ b/arch/arm64/mm/ptdump.c
> @@ -189,7 +189,7 @@ static void note_prot_wx(struct ptdump_pg_state *st, unsigned long addr)
>  }
>  
>  void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
> -	       u64 val)
> +	       pteval_t val)
>  {
>  	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
>  	struct ptdump_pg_level *pg_level = st->pg_level;
> @@ -251,6 +251,38 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
>  
>  }
>  
> +void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> +{
> +	note_page(pt_st, addr, 4, pte_val(pte));
> +}
> +
> +void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> +{
> +	note_page(pt_st, addr, 3, pmd_val(pmd));
> +}
> +
> +void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> +{
> +	note_page(pt_st, addr, 2, pud_val(pud));
> +}
> +
> +void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> +{
> +	note_page(pt_st, addr, 1, p4d_val(p4d));
> +}
> +
> +void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> +{
> +	note_page(pt_st, addr, 0, pgd_val(pgd));
> +}
> +
> +void note_page_flush(struct ptdump_state *pt_st)
> +{
> +	pte_t pte_zero = {0};
> +
> +	note_page(pt_st, 0, -1, pte_val(pte_zero));
> +}
> +
>  void ptdump_walk(struct seq_file *s, struct ptdump_info *info)
>  {
>  	unsigned long end = ~0UL;
> @@ -266,7 +298,12 @@ void ptdump_walk(struct seq_file *s, struct ptdump_info *info)
>  		.pg_level = &kernel_pg_levels[0],
>  		.level = -1,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]){
>  				{info->base_addr, end},
>  				{0, 0}
> @@ -303,7 +340,12 @@ bool ptdump_check_wx(void)
>  		.level = -1,
>  		.check_wx = true,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]) {
>  				{_PAGE_OFFSET(vabits_actual), ~0UL},
>  				{0, 0}
> diff --git a/arch/powerpc/mm/ptdump/ptdump.c b/arch/powerpc/mm/ptdump/ptdump.c
> index 9dc239967b77..b2358d794855 100644
> --- a/arch/powerpc/mm/ptdump/ptdump.c
> +++ b/arch/powerpc/mm/ptdump/ptdump.c
> @@ -298,6 +298,38 @@ static void populate_markers(void)
>  #endif
>  }
>  
> +static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> +{
> +	note_page(pt_st, addr, 4, pte_val(pte));
> +}
> +
> +static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> +{
> +	note_page(pt_st, addr, 3, pmd_val(pmd));
> +}
> +
> +static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> +{
> +	note_page(pt_st, addr, 2, pud_val(pud));
> +}
> +
> +static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> +{
> +	note_page(pt_st, addr, 1, p4d_val(p4d));
> +}
> +
> +static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> +{
> +	note_page(pt_st, addr, 0, pgd_val(pgd));
> +}
> +
> +static void note_page_flush(struct ptdump_state *pt_st)
> +{
> +	pte_t pte_zero = {0};
> +
> +	note_page(pt_st, 0, -1, pte_val(pte_zero));
> +}
> +
>  static int ptdump_show(struct seq_file *m, void *v)
>  {
>  	struct pg_state st = {
> @@ -305,7 +337,12 @@ static int ptdump_show(struct seq_file *m, void *v)
>  		.marker = address_markers,
>  		.level = -1,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = ptdump_range,
>  		}
>  	};
> @@ -338,7 +375,12 @@ bool ptdump_check_wx(void)
>  		.level = -1,
>  		.check_wx = true,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = ptdump_range,
>  		}
>  	};
> diff --git a/arch/riscv/mm/ptdump.c b/arch/riscv/mm/ptdump.c
> index 9d5f657a251b..32922550a50a 100644
> --- a/arch/riscv/mm/ptdump.c
> +++ b/arch/riscv/mm/ptdump.c
> @@ -318,6 +318,38 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr,
>  	}
>  }
>  
> +static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> +{
> +	note_page(pt_st, addr, 4, pte_val(pte));
> +}
> +
> +static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> +{
> +	note_page(pt_st, addr, 3, pmd_val(pmd));
> +}
> +
> +static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> +{
> +	note_page(pt_st, addr, 2, pud_val(pud));
> +}
> +
> +static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> +{
> +	note_page(pt_st, addr, 1, p4d_val(p4d));
> +}
> +
> +static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> +{
> +	note_page(pt_st, addr, 0, pgd_val(pgd));
> +}
> +
> +static void note_page_flush(struct ptdump_state *pt_st)
> +{
> +	pte_t pte_zero = {0};
> +
> +	note_page(pt_st, 0, -1, pte_val(pte_zero));
> +}
> +
>  static void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo)
>  {
>  	struct pg_state st = {
> @@ -325,7 +357,12 @@ static void ptdump_walk(struct seq_file *s, struct ptd_mm_info *pinfo)
>  		.marker = pinfo->markers,
>  		.level = -1,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]) {
>  				{pinfo->base_addr, pinfo->end},
>  				{0, 0}
> @@ -347,7 +384,12 @@ bool ptdump_check_wx(void)
>  		.level = -1,
>  		.check_wx = true,
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]) {
>  				{KERN_VIRT_START, ULONG_MAX},
>  				{0, 0}
> diff --git a/arch/s390/mm/dump_pagetables.c b/arch/s390/mm/dump_pagetables.c
> index fa54f3bc0c8d..e7cc52542f71 100644
> --- a/arch/s390/mm/dump_pagetables.c
> +++ b/arch/s390/mm/dump_pagetables.c
> @@ -145,11 +145,48 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
>  	}
>  }
>  
> +static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> +{
> +	note_page(pt_st, addr, 4, pte_val(pte));
> +}
> +
> +static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> +{
> +	note_page(pt_st, addr, 3, pmd_val(pmd));
> +}
> +
> +static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> +{
> +	note_page(pt_st, addr, 2, pud_val(pud));
> +}
> +
> +static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> +{
> +	note_page(pt_st, addr, 1, p4d_val(p4d));
> +}
> +
> +static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> +{
> +	note_page(pt_st, addr, 0, pgd_val(pgd));
> +}
> +
> +static void note_page_flush(struct ptdump_state *pt_st)
> +{
> +	pte_t pte_zero = {0};
> +
> +	note_page(pt_st, 0, -1, pte_val(pte_zero));
> +}
> +
>  bool ptdump_check_wx(void)
>  {
>  	struct pg_state st = {
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]) {
>  				{.start = 0, .end = max_addr},
>  				{.start = 0, .end = 0},
> @@ -188,7 +225,12 @@ static int ptdump_show(struct seq_file *m, void *v)
>  {
>  	struct pg_state st = {
>  		.ptdump = {
> -			.note_page = note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.range = (struct ptdump_range[]) {
>  				{.start = 0, .end = max_addr},
>  				{.start = 0, .end = 0},
> diff --git a/arch/x86/mm/dump_pagetables.c b/arch/x86/mm/dump_pagetables.c
> index 89079ea73e65..2e1c2d006ace 100644
> --- a/arch/x86/mm/dump_pagetables.c
> +++ b/arch/x86/mm/dump_pagetables.c
> @@ -362,6 +362,38 @@ static void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
>  	}
>  }
>  
> +static void note_page_pte(struct ptdump_state *pt_st, unsigned long addr, pte_t pte)
> +{
> +	note_page(pt_st, addr, 4, pte_val(pte));
> +}
> +
> +static void note_page_pmd(struct ptdump_state *pt_st, unsigned long addr, pmd_t pmd)
> +{
> +	note_page(pt_st, addr, 3, pmd_val(pmd));
> +}
> +
> +static void note_page_pud(struct ptdump_state *pt_st, unsigned long addr, pud_t pud)
> +{
> +	note_page(pt_st, addr, 2, pud_val(pud));
> +}
> +
> +static void note_page_p4d(struct ptdump_state *pt_st, unsigned long addr, p4d_t p4d)
> +{
> +	note_page(pt_st, addr, 1, p4d_val(p4d));
> +}
> +
> +static void note_page_pgd(struct ptdump_state *pt_st, unsigned long addr, pgd_t pgd)
> +{
> +	note_page(pt_st, addr, 0, pgd_val(pgd));
> +}
> +
> +static void note_page_flush(struct ptdump_state *pt_st)
> +{
> +	pte_t pte_zero = {0};
> +
> +	note_page(pt_st, 0, -1, pte_val(pte_zero));
> +}
> +
>  bool ptdump_walk_pgd_level_core(struct seq_file *m,
>  				struct mm_struct *mm, pgd_t *pgd,
>  				bool checkwx, bool dmesg)
> @@ -378,7 +410,12 @@ bool ptdump_walk_pgd_level_core(struct seq_file *m,
>  
>  	struct pg_state st = {
>  		.ptdump = {
> -			.note_page	= note_page,
> +			.note_page_pte = note_page_pte,
> +			.note_page_pmd = note_page_pmd,
> +			.note_page_pud = note_page_pud,
> +			.note_page_p4d = note_page_p4d,
> +			.note_page_pgd = note_page_pgd,
> +			.note_page_flush = note_page_flush,
>  			.effective_prot = effective_prot,
>  			.range		= ptdump_ranges
>  		},
> diff --git a/include/linux/ptdump.h b/include/linux/ptdump.h
> index 8dbd51ea8626..1c1eb1fae199 100644
> --- a/include/linux/ptdump.h
> +++ b/include/linux/ptdump.h
> @@ -11,9 +11,12 @@ struct ptdump_range {
>  };
>  
>  struct ptdump_state {
> -	/* level is 0:PGD to 4:PTE, or -1 if unknown */
> -	void (*note_page)(struct ptdump_state *st, unsigned long addr,
> -			  int level, u64 val);
> +	void (*note_page_pte)(struct ptdump_state *st, unsigned long addr, pte_t pte);
> +	void (*note_page_pmd)(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
> +	void (*note_page_pud)(struct ptdump_state *st, unsigned long addr, pud_t pud);
> +	void (*note_page_p4d)(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
> +	void (*note_page_pgd)(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
> +	void (*note_page_flush)(struct ptdump_state *st);
>  	void (*effective_prot)(struct ptdump_state *st, int level, u64 val);
>  	const struct ptdump_range *range;
>  };
> diff --git a/mm/ptdump.c b/mm/ptdump.c
> index 106e1d66e9f9..706cfc19439b 100644
> --- a/mm/ptdump.c
> +++ b/mm/ptdump.c
> @@ -18,7 +18,7 @@ static inline int note_kasan_page_table(struct mm_walk *walk,
>  {
>  	struct ptdump_state *st = walk->private;
>  
> -	st->note_page(st, addr, 4, pte_val(kasan_early_shadow_pte[0]));
> +	st->note_page_pte(st, addr, kasan_early_shadow_pte[0]);
>  
>  	walk->action = ACTION_CONTINUE;
>  
> @@ -42,7 +42,7 @@ static int ptdump_pgd_entry(pgd_t *pgd, unsigned long addr,
>  		st->effective_prot(st, 0, pgd_val(val));
>  
>  	if (pgd_leaf(val)) {
> -		st->note_page(st, addr, 0, pgd_val(val));
> +		st->note_page_pgd(st, addr, val);
>  		walk->action = ACTION_CONTINUE;
>  	}
>  
> @@ -65,7 +65,7 @@ static int ptdump_p4d_entry(p4d_t *p4d, unsigned long addr,
>  		st->effective_prot(st, 1, p4d_val(val));
>  
>  	if (p4d_leaf(val)) {
> -		st->note_page(st, addr, 1, p4d_val(val));
> +		st->note_page_p4d(st, addr, val);
>  		walk->action = ACTION_CONTINUE;
>  	}
>  
> @@ -88,7 +88,7 @@ static int ptdump_pud_entry(pud_t *pud, unsigned long addr,
>  		st->effective_prot(st, 2, pud_val(val));
>  
>  	if (pud_leaf(val)) {
> -		st->note_page(st, addr, 2, pud_val(val));
> +		st->note_page_pud(st, addr, val);
>  		walk->action = ACTION_CONTINUE;
>  	}
>  
> @@ -109,7 +109,7 @@ static int ptdump_pmd_entry(pmd_t *pmd, unsigned long addr,
>  	if (st->effective_prot)
>  		st->effective_prot(st, 3, pmd_val(val));
>  	if (pmd_leaf(val)) {
> -		st->note_page(st, addr, 3, pmd_val(val));
> +		st->note_page_pmd(st, addr, val);
>  		walk->action = ACTION_CONTINUE;
>  	}
>  
> @@ -125,7 +125,7 @@ static int ptdump_pte_entry(pte_t *pte, unsigned long addr,
>  	if (st->effective_prot)
>  		st->effective_prot(st, 4, pte_val(val));
>  
> -	st->note_page(st, addr, 4, pte_val(val));
> +	st->note_page_pte(st, addr, val);
>  
>  	return 0;
>  }
> @@ -134,9 +134,31 @@ static int ptdump_hole(unsigned long addr, unsigned long next,
>  		       int depth, struct mm_walk *walk)
>  {
>  	struct ptdump_state *st = walk->private;
> -
> -	st->note_page(st, addr, depth, 0);
> -
> +	pte_t pte_zero = {0};
> +	pmd_t pmd_zero = {0};
> +	pud_t pud_zero = {0};
> +	p4d_t p4d_zero = {0};
> +	pgd_t pgd_zero = {0};
> +
> +	switch (depth) {
> +	case 4:
> +		st->note_page_pte(st, addr, pte_zero);
> +		break;
> +	case 3:
> +		st->note_page_pmd(st, addr, pmd_zero);
> +		break;
> +	case 2:
> +		st->note_page_pud(st, addr, pud_zero);
> +		break;
> +	case 1:
> +		st->note_page_p4d(st, addr, p4d_zero);
> +		break;
> +	case 0:
> +		st->note_page_pgd(st, addr, pgd_zero);
> +		break;
> +	default:
> +		break;
> +	}
>  	return 0;
>  }
>  
> @@ -162,7 +184,7 @@ void ptdump_walk_pgd(struct ptdump_state *st, struct mm_struct *mm, pgd_t *pgd)
>  	mmap_write_unlock(mm);
>  
>  	/* Flush out the last page */
> -	st->note_page(st, 0, -1, 0);
> +	st->note_page_flush(st);
>  }
>  
>  static int check_wx_show(struct seq_file *m, void *v)

Re: [PATCH 1/2] mm/ptdump: Split note_page() into level specific callbacks

[Alexander Gordeev]

On Mon, Mar 17, 2025 at 11:48:17AM +0530, Anshuman Khandual wrote:

Hi Anshuman,

...
> --- a/include/linux/ptdump.h
> +++ b/include/linux/ptdump.h
> @@ -11,9 +11,12 @@ struct ptdump_range {
>  };
>  
>  struct ptdump_state {
> -	/* level is 0:PGD to 4:PTE, or -1 if unknown */
> -	void (*note_page)(struct ptdump_state *st, unsigned long addr,
> -			  int level, u64 val);
> +	void (*note_page_pte)(struct ptdump_state *st, unsigned long addr, pte_t pte);
> +	void (*note_page_pmd)(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
> +	void (*note_page_pud)(struct ptdump_state *st, unsigned long addr, pud_t pud);
> +	void (*note_page_p4d)(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
> +	void (*note_page_pgd)(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
> +	void (*note_page_flush)(struct ptdump_state *st);
>  	void (*effective_prot)(struct ptdump_state *st, int level, u64 val);

Should you treat effective_prot() similarly?

>  	const struct ptdump_range *range;
>  };

Re: [PATCH 1/2] mm/ptdump: Split note_page() into level specific callbacks

[Anshuman Khandual]

On 3/31/25 14:29, Alexander Gordeev wrote:
> On Mon, Mar 17, 2025 at 11:48:17AM +0530, Anshuman Khandual wrote:
> 
> Hi Anshuman,
> 
> ...
>> --- a/include/linux/ptdump.h
>> +++ b/include/linux/ptdump.h
>> @@ -11,9 +11,12 @@ struct ptdump_range {
>>  };
>>  
>>  struct ptdump_state {
>> -	/* level is 0:PGD to 4:PTE, or -1 if unknown */
>> -	void (*note_page)(struct ptdump_state *st, unsigned long addr,
>> -			  int level, u64 val);
>> +	void (*note_page_pte)(struct ptdump_state *st, unsigned long addr, pte_t pte);
>> +	void (*note_page_pmd)(struct ptdump_state *st, unsigned long addr, pmd_t pmd);
>> +	void (*note_page_pud)(struct ptdump_state *st, unsigned long addr, pud_t pud);
>> +	void (*note_page_p4d)(struct ptdump_state *st, unsigned long addr, p4d_t p4d);
>> +	void (*note_page_pgd)(struct ptdump_state *st, unsigned long addr, pgd_t pgd);
>> +	void (*note_page_flush)(struct ptdump_state *st);
>>  	void (*effective_prot)(struct ptdump_state *st, int level, u64 val);
> 
> Should you treat effective_prot() similarly?

Agreed. effective_prot() also uses pxd_val() derived data type in generic MM and
might cause the same problem later going ahead. Will split the helper into level
specific callbacks as well.

> 
>>  	const struct ptdump_range *range;
>>  };

[PATCH 2/2] arm64/ptdump: Replace u64 with pteval_t

[Anshuman Khandual]

Page table entry's value, mask and protection are represented with pteval_t
data type format not u64 that has been assumed while dumping the page table
entries. Replace all such u64 instances with pteval_t instead as required.

Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Marc Zyngier <maz@kernel.org>
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
---
 arch/arm64/include/asm/ptdump.h | 8 ++++----
 arch/arm64/mm/ptdump.c          | 2 +-
 2 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
index e5da9ce8a515..476a870489b9 100644
--- a/arch/arm64/include/asm/ptdump.h
+++ b/arch/arm64/include/asm/ptdump.h
@@ -24,8 +24,8 @@ struct ptdump_info {
 };
 
 struct ptdump_prot_bits {
-	u64		mask;
-	u64		val;
+	pteval_t	mask;
+	pteval_t	val;
 	const char	*set;
 	const char	*clear;
 };
@@ -34,7 +34,7 @@ struct ptdump_pg_level {
 	const struct ptdump_prot_bits *bits;
 	char name[4];
 	int num;
-	u64 mask;
+	pteval_t mask;
 };
 
 /*
@@ -51,7 +51,7 @@ struct ptdump_pg_state {
 	const struct mm_struct *mm;
 	unsigned long start_address;
 	int level;
-	u64 current_prot;
+	pteval_t current_prot;
 	bool check_wx;
 	unsigned long wx_pages;
 	unsigned long uxn_pages;
diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
index fd1610b4fd15..a5651be95868 100644
--- a/arch/arm64/mm/ptdump.c
+++ b/arch/arm64/mm/ptdump.c
@@ -194,7 +194,7 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
 	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
 	struct ptdump_pg_level *pg_level = st->pg_level;
 	static const char units[] = "KMGTPE";
-	u64 prot = 0;
+	pteval_t prot = 0;
 
 	/* check if the current level has been folded dynamically */
 	if (st->mm && ((level == 1 && mm_p4d_folded(st->mm)) ||
-- 
2.25.1

Re: [PATCH 2/2] arm64/ptdump: Replace u64 with pteval_t

[Ryan Roberts]

On 17/03/2025 06:18, Anshuman Khandual wrote:
> Page table entry's value, mask and protection are represented with pteval_t
> data type format not u64 that has been assumed while dumping the page table
> entries. Replace all such u64 instances with pteval_t instead as required.
> 
> Cc: Catalin Marinas <catalin.marinas@arm.com>
> Cc: Will Deacon <will@kernel.org>
> Cc: Marc Zyngier <maz@kernel.org>
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: linux-kernel@vger.kernel.org
> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
> ---
>  arch/arm64/include/asm/ptdump.h | 8 ++++----
>  arch/arm64/mm/ptdump.c          | 2 +-
>  2 files changed, 5 insertions(+), 5 deletions(-)
> 
> diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
> index e5da9ce8a515..476a870489b9 100644
> --- a/arch/arm64/include/asm/ptdump.h
> +++ b/arch/arm64/include/asm/ptdump.h
> @@ -24,8 +24,8 @@ struct ptdump_info {
>  };
>  
>  struct ptdump_prot_bits {
> -	u64		mask;
> -	u64		val;
> +	pteval_t	mask;
> +	pteval_t	val;

Given Ard's suggestion of using "ptdesc" as a generic term for PTDESC_SHIFT (or
PTDESC_ORDER, or whatever we ended up calling it), I wonder if it would be
cleaner to do the same with the types? We could have a ptdesc_t, which is
typedef'ed as u64 (or u128), then pteval_t, pmdval_t, ..., could all be
typedef'ed as ptdesc_t. Then for code that just wants a generic pgtable
descriptor value, we can use that type to indicate that it can be at any level.

Thanks,
Ryan

>  	const char	*set;
>  	const char	*clear;
>  };
> @@ -34,7 +34,7 @@ struct ptdump_pg_level {
>  	const struct ptdump_prot_bits *bits;
>  	char name[4];
>  	int num;
> -	u64 mask;
> +	pteval_t mask;
>  };
>  
>  /*
> @@ -51,7 +51,7 @@ struct ptdump_pg_state {
>  	const struct mm_struct *mm;
>  	unsigned long start_address;
>  	int level;
> -	u64 current_prot;
> +	pteval_t current_prot;
>  	bool check_wx;
>  	unsigned long wx_pages;
>  	unsigned long uxn_pages;
> diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
> index fd1610b4fd15..a5651be95868 100644
> --- a/arch/arm64/mm/ptdump.c
> +++ b/arch/arm64/mm/ptdump.c
> @@ -194,7 +194,7 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
>  	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
>  	struct ptdump_pg_level *pg_level = st->pg_level;
>  	static const char units[] = "KMGTPE";
> -	u64 prot = 0;
> +	pteval_t prot = 0;
>  
>  	/* check if the current level has been folded dynamically */
>  	if (st->mm && ((level == 1 && mm_p4d_folded(st->mm)) ||

Re: [PATCH 2/2] arm64/ptdump: Replace u64 with pteval_t

[Anshuman Khandual]

On 3/17/25 14:58, Ryan Roberts wrote:
> On 17/03/2025 06:18, Anshuman Khandual wrote:
>> Page table entry's value, mask and protection are represented with pteval_t
>> data type format not u64 that has been assumed while dumping the page table
>> entries. Replace all such u64 instances with pteval_t instead as required.
>>
>> Cc: Catalin Marinas <catalin.marinas@arm.com>
>> Cc: Will Deacon <will@kernel.org>
>> Cc: Marc Zyngier <maz@kernel.org>
>> Cc: linux-arm-kernel@lists.infradead.org
>> Cc: linux-kernel@vger.kernel.org
>> Signed-off-by: Anshuman Khandual <anshuman.khandual@arm.com>
>> ---
>>  arch/arm64/include/asm/ptdump.h | 8 ++++----
>>  arch/arm64/mm/ptdump.c          | 2 +-
>>  2 files changed, 5 insertions(+), 5 deletions(-)
>>
>> diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
>> index e5da9ce8a515..476a870489b9 100644
>> --- a/arch/arm64/include/asm/ptdump.h
>> +++ b/arch/arm64/include/asm/ptdump.h
>> @@ -24,8 +24,8 @@ struct ptdump_info {
>>  };
>>  
>>  struct ptdump_prot_bits {
>> -	u64		mask;
>> -	u64		val;
>> +	pteval_t	mask;
>> +	pteval_t	val;
> 
> Given Ard's suggestion of using "ptdesc" as a generic term for PTDESC_SHIFT (or
> PTDESC_ORDER, or whatever we ended up calling it), I wonder if it would be
> cleaner to do the same with the types? We could have a ptdesc_t, which is
> typedef'ed as u64 (or u128), then pteval_t, pmdval_t, ..., could all be
> typedef'ed as ptdesc_t. Then for code that just wants a generic pgtable
> descriptor value, we can use that type to indicate that it can be at any level.

Something like the following ? Will cross check again if this might have
missed something which could be converted as ptdesc_t as well.

diff --git a/arch/arm64/include/asm/pgtable-types.h b/arch/arm64/include/asm/pgtable-types.h
index 6d6d4065b0cb..686541e986e3 100644
--- a/arch/arm64/include/asm/pgtable-types.h
+++ b/arch/arm64/include/asm/pgtable-types.h
@@ -11,11 +11,13 @@
 
 #include <asm/types.h>
 
-typedef u64 pteval_t;
-typedef u64 pmdval_t;
-typedef u64 pudval_t;
-typedef u64 p4dval_t;
-typedef u64 pgdval_t;
+typedef u64 ptdesc_t;
+
+typedef ptdesc_t pteval_t;
+typedef ptdesc_t pmdval_t;
+typedef ptdesc_t pudval_t;
+typedef ptdesc_t p4dval_t;
+typedef ptdesc_t pgdval_t;
 
 /*
  * These are used to make use of C type-checking..
@@ -46,7 +48,7 @@ typedef struct { pgdval_t pgd; } pgd_t;
 #define pgd_val(x)	((x).pgd)
 #define __pgd(x)	((pgd_t) { (x) } )
 
-typedef struct { pteval_t pgprot; } pgprot_t;
+typedef struct { ptdesc_t pgprot; } pgprot_t;
 #define pgprot_val(x)	((x).pgprot)
 #define __pgprot(x)	((pgprot_t) { (x) } )
 
diff --git a/arch/arm64/include/asm/ptdump.h b/arch/arm64/include/asm/ptdump.h
index e5da9ce8a515..9548813bc877 100644
--- a/arch/arm64/include/asm/ptdump.h
+++ b/arch/arm64/include/asm/ptdump.h
@@ -24,8 +24,8 @@ struct ptdump_info {
 };
 
 struct ptdump_prot_bits {
-	u64		mask;
-	u64		val;
+	ptdesc_t	mask;
+	ptdesc_t	val;
 	const char	*set;
 	const char	*clear;
 };
@@ -34,7 +34,7 @@ struct ptdump_pg_level {
 	const struct ptdump_prot_bits *bits;
 	char name[4];
 	int num;
-	u64 mask;
+	ptdesc_t mask;
 };
 
 /*
@@ -51,7 +51,7 @@ struct ptdump_pg_state {
 	const struct mm_struct *mm;
 	unsigned long start_address;
 	int level;
-	u64 current_prot;
+	ptdesc_t current_prot;
 	bool check_wx;
 	unsigned long wx_pages;
 	unsigned long uxn_pages;
diff --git a/arch/arm64/kernel/efi.c b/arch/arm64/kernel/efi.c
index 1d25d8899dbf..42e281c07c2f 100644
--- a/arch/arm64/kernel/efi.c
+++ b/arch/arm64/kernel/efi.c
@@ -29,7 +29,7 @@ static bool region_is_misaligned(const efi_memory_desc_t *md)
  * executable, everything else can be mapped with the XN bits
  * set. Also take the new (optional) RO/XP bits into account.
  */
-static __init pteval_t create_mapping_protection(efi_memory_desc_t *md)
+static __init ptdesc_t create_mapping_protection(efi_memory_desc_t *md)
 {
 	u64 attr = md->attribute;
 	u32 type = md->type;
@@ -83,7 +83,7 @@ static __init pteval_t create_mapping_protection(efi_memory_desc_t *md)
 
 int __init efi_create_mapping(struct mm_struct *mm, efi_memory_desc_t *md)
 {
-	pteval_t prot_val = create_mapping_protection(md);
+	ptdesc_t prot_val = create_mapping_protection(md);
 	bool page_mappings_only = (md->type == EFI_RUNTIME_SERVICES_CODE ||
 				   md->type == EFI_RUNTIME_SERVICES_DATA);
 
diff --git a/arch/arm64/kernel/pi/map_kernel.c b/arch/arm64/kernel/pi/map_kernel.c
index e57b043f324b..a00f57c73d81 100644
--- a/arch/arm64/kernel/pi/map_kernel.c
+++ b/arch/arm64/kernel/pi/map_kernel.c
@@ -159,7 +159,7 @@ static void noinline __section(".idmap.text") set_ttbr0_for_lpa2(u64 ttbr)
 static void __init remap_idmap_for_lpa2(void)
 {
 	/* clear the bits that change meaning once LPA2 is turned on */
-	pteval_t mask = PTE_SHARED;
+	ptdesc_t mask = PTE_SHARED;
 
 	/*
 	 * We have to clear bits [9:8] in all block or page descriptors in the
diff --git a/arch/arm64/kernel/pi/map_range.c b/arch/arm64/kernel/pi/map_range.c
index 2b69e3beeef8..30c6bc50844f 100644
--- a/arch/arm64/kernel/pi/map_range.c
+++ b/arch/arm64/kernel/pi/map_range.c
@@ -30,7 +30,7 @@ void __init map_range(u64 *pte, u64 start, u64 end, u64 pa, pgprot_t prot,
 		      int level, pte_t *tbl, bool may_use_cont, u64 va_offset)
 {
 	u64 cmask = (level == 3) ? CONT_PTE_SIZE - 1 : U64_MAX;
-	pteval_t protval = pgprot_val(prot) & ~PTE_TYPE_MASK;
+	ptdesc_t protval = pgprot_val(prot) & ~PTE_TYPE_MASK;
 	int lshift = (3 - level) * (PAGE_SHIFT - 3);
 	u64 lmask = (PAGE_SIZE << lshift) - 1;
 
@@ -87,7 +87,7 @@ void __init map_range(u64 *pte, u64 start, u64 end, u64 pa, pgprot_t prot,
 	}
 }
 
-asmlinkage u64 __init create_init_idmap(pgd_t *pg_dir, pteval_t clrmask)
+asmlinkage u64 __init create_init_idmap(pgd_t *pg_dir, ptdesc_t clrmask)
 {
 	u64 ptep = (u64)pg_dir + PAGE_SIZE;
 	pgprot_t text_prot = PAGE_KERNEL_ROX;
diff --git a/arch/arm64/kernel/pi/pi.h b/arch/arm64/kernel/pi/pi.h
index c91e5e965cd3..91dcb5b6bbd1 100644
--- a/arch/arm64/kernel/pi/pi.h
+++ b/arch/arm64/kernel/pi/pi.h
@@ -33,4 +33,4 @@ void map_range(u64 *pgd, u64 start, u64 end, u64 pa, pgprot_t prot,
 
 asmlinkage void early_map_kernel(u64 boot_status, void *fdt);
 
-asmlinkage u64 create_init_idmap(pgd_t *pgd, pteval_t clrmask);
+asmlinkage u64 create_init_idmap(pgd_t *pgd, ptdesc_t clrmask);
diff --git a/arch/arm64/mm/mmap.c b/arch/arm64/mm/mmap.c
index 07aeab8a7606..c86c348857c4 100644
--- a/arch/arm64/mm/mmap.c
+++ b/arch/arm64/mm/mmap.c
@@ -83,7 +83,7 @@ arch_initcall(adjust_protection_map);
 
 pgprot_t vm_get_page_prot(unsigned long vm_flags)
 {
-	pteval_t prot;
+	ptdesc_t prot;
 
 	/* Short circuit GCS to avoid bloating the table. */
 	if (system_supports_gcs() && (vm_flags & VM_SHADOW_STACK)) {
diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
index fd1610b4fd15..280e850f1688 100644
--- a/arch/arm64/mm/ptdump.c
+++ b/arch/arm64/mm/ptdump.c
@@ -194,7 +194,7 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
 	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
 	struct ptdump_pg_level *pg_level = st->pg_level;
 	static const char units[] = "KMGTPE";
-	u64 prot = 0;
+	ptdesc_t prot = 0;
 
 	/* check if the current level has been folded dynamically */
 	if (st->mm && ((level == 1 && mm_p4d_folded(st->mm)) ||
-- 
2.25.1

> 
> Thanks,
> Ryan
> 
>>  	const char	*set;
>>  	const char	*clear;
>>  };
>> @@ -34,7 +34,7 @@ struct ptdump_pg_level {
>>  	const struct ptdump_prot_bits *bits;
>>  	char name[4];
>>  	int num;
>> -	u64 mask;
>> +	pteval_t mask;
>>  };
>>  
>>  /*
>> @@ -51,7 +51,7 @@ struct ptdump_pg_state {
>>  	const struct mm_struct *mm;
>>  	unsigned long start_address;
>>  	int level;
>> -	u64 current_prot;
>> +	pteval_t current_prot;
>>  	bool check_wx;
>>  	unsigned long wx_pages;
>>  	unsigned long uxn_pages;
>> diff --git a/arch/arm64/mm/ptdump.c b/arch/arm64/mm/ptdump.c
>> index fd1610b4fd15..a5651be95868 100644
>> --- a/arch/arm64/mm/ptdump.c
>> +++ b/arch/arm64/mm/ptdump.c
>> @@ -194,7 +194,7 @@ void note_page(struct ptdump_state *pt_st, unsigned long addr, int level,
>>  	struct ptdump_pg_state *st = container_of(pt_st, struct ptdump_pg_state, ptdump);
>>  	struct ptdump_pg_level *pg_level = st->pg_level;
>>  	static const char units[] = "KMGTPE";
>> -	u64 prot = 0;
>> +	pteval_t prot = 0;
>>  
>>  	/* check if the current level has been folded dynamically */
>>  	if (st->mm && ((level == 1 && mm_p4d_folded(st->mm)) ||
>