RISC-V: Add Sv57 page table support

This implements Sv57 support at runtime. The kernel will try to boot
with 5-level page table firstly , and will fallback to 4-level if the HW
does not support it. And it will finally fallback to 3-level if the HW
alse does not support sv48.

* riscv-sv57:
  riscv: mm: Support kasan for sv57
  riscv: mm: Set sv57 on defaultly
  riscv: mm: Prepare pt_ops helper functions for sv57
  riscv: mm: Control p4d's folding by pgtable_l5_enabled

arch/riscv/Kconfig

@@ -151,7 +151,7 @@ config PAGE_OFFSET
 	hex
 	default 0xC0000000 if 32BIT
 	default 0x80000000 if 64BIT && !MMU
-	default 0xffffaf8000000000 if 64BIT
+	default 0xff60000000000000 if 64BIT
 
 config KASAN_SHADOW_OFFSET
 	hex
@@ -202,7 +202,7 @@ config FIX_EARLYCON_MEM
 
 config PGTABLE_LEVELS
 	int
-	default 4 if 64BIT
+	default 5 if 64BIT
 	default 2
 
 config LOCKDEP_SUPPORT

arch/riscv/include/asm/csr.h

@@ -47,6 +47,7 @@
 #define SATP_PPN	_AC(0x00000FFFFFFFFFFF, UL)
 #define SATP_MODE_39	_AC(0x8000000000000000, UL)
 #define SATP_MODE_48	_AC(0x9000000000000000, UL)
+#define SATP_MODE_57	_AC(0xa000000000000000, UL)
 #define SATP_ASID_BITS	16
 #define SATP_ASID_SHIFT	44
 #define SATP_ASID_MASK	_AC(0xFFFF, UL)

arch/riscv/include/asm/fixmap.h

@@ -25,6 +25,7 @@ enum fixed_addresses {
 	FIX_PTE,
 	FIX_PMD,
 	FIX_PUD,
+	FIX_P4D,
 	FIX_TEXT_POKE1,
 	FIX_TEXT_POKE0,
 	FIX_EARLYCON_MEM_BASE,

arch/riscv/include/asm/page.h

@@ -41,6 +41,7 @@
  * By default, CONFIG_PAGE_OFFSET value corresponds to SV48 address space so
  * define the PAGE_OFFSET value for SV39.
  */
+#define PAGE_OFFSET_L4		_AC(0xffffaf8000000000, UL)
 #define PAGE_OFFSET_L3		_AC(0xffffffd800000000, UL)
 #else
 #define PAGE_OFFSET		_AC(CONFIG_PAGE_OFFSET, UL)

arch/riscv/include/asm/pgalloc.h

@@ -59,6 +59,26 @@ static inline void p4d_populate_safe(struct mm_struct *mm, p4d_t *p4d,
 	}
 }
 
+static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, p4d_t *p4d)
+{
+	if (pgtable_l5_enabled) {
+		unsigned long pfn = virt_to_pfn(p4d);
+
+		set_pgd(pgd, __pgd((pfn << _PAGE_PFN_SHIFT) | _PAGE_TABLE));
+	}
+}
+
+static inline void pgd_populate_safe(struct mm_struct *mm, pgd_t *pgd,
+				     p4d_t *p4d)
+{
+	if (pgtable_l5_enabled) {
+		unsigned long pfn = virt_to_pfn(p4d);
+
+		set_pgd_safe(pgd,
+			     __pgd((pfn << _PAGE_PFN_SHIFT) | _PAGE_TABLE));
+	}
+}
+
 #define pud_alloc_one pud_alloc_one
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
 {
@@ -76,6 +96,35 @@ static inline void pud_free(struct mm_struct *mm, pud_t *pud)
 }
 
 #define __pud_free_tlb(tlb, pud, addr)  pud_free((tlb)->mm, pud)
+
+#define p4d_alloc_one p4d_alloc_one
+static inline p4d_t *p4d_alloc_one(struct mm_struct *mm, unsigned long addr)
+{
+	if (pgtable_l5_enabled) {
+		gfp_t gfp = GFP_PGTABLE_USER;
+
+		if (mm == &init_mm)
+			gfp = GFP_PGTABLE_KERNEL;
+		return (p4d_t *)get_zeroed_page(gfp);
+	}
+
+	return NULL;
+}
+
+static inline void __p4d_free(struct mm_struct *mm, p4d_t *p4d)
+{
+	BUG_ON((unsigned long)p4d & (PAGE_SIZE-1));
+	free_page((unsigned long)p4d);
+}
+
+#define p4d_free p4d_free
+static inline void p4d_free(struct mm_struct *mm, p4d_t *p4d)
+{
+	if (pgtable_l5_enabled)
+		__p4d_free(mm, p4d);
+}
+
+#define __p4d_free_tlb(tlb, p4d, addr)  p4d_free((tlb)->mm, p4d)
 #endif /* __PAGETABLE_PMD_FOLDED */
 
 static inline pgd_t *pgd_alloc(struct mm_struct *mm)

arch/riscv/include/asm/pgtable-64.h

@@ -9,16 +9,24 @@
 #include <linux/const.h>
 
 extern bool pgtable_l4_enabled;
+extern bool pgtable_l5_enabled;
 
 #define PGDIR_SHIFT_L3  30
 #define PGDIR_SHIFT_L4  39
+#define PGDIR_SHIFT_L5  48
 #define PGDIR_SIZE_L3   (_AC(1, UL) << PGDIR_SHIFT_L3)
 
-#define PGDIR_SHIFT     (pgtable_l4_enabled ? PGDIR_SHIFT_L4 : PGDIR_SHIFT_L3)
+#define PGDIR_SHIFT     (pgtable_l5_enabled ? PGDIR_SHIFT_L5 : \
+		(pgtable_l4_enabled ? PGDIR_SHIFT_L4 : PGDIR_SHIFT_L3))
 /* Size of region mapped by a page global directory */
 #define PGDIR_SIZE      (_AC(1, UL) << PGDIR_SHIFT)
 #define PGDIR_MASK      (~(PGDIR_SIZE - 1))
 
+/* p4d is folded into pgd in case of 4-level page table */
+#define P4D_SHIFT      39
+#define P4D_SIZE       (_AC(1, UL) << P4D_SHIFT)
+#define P4D_MASK       (~(P4D_SIZE - 1))
+
 /* pud is folded into pgd in case of 3-level page table */
 #define PUD_SHIFT      30
 #define PUD_SIZE       (_AC(1, UL) << PUD_SHIFT)
@@ -29,6 +37,15 @@ extern bool pgtable_l4_enabled;
 #define PMD_SIZE        (_AC(1, UL) << PMD_SHIFT)
 #define PMD_MASK        (~(PMD_SIZE - 1))
 
+/* Page 4th Directory entry */
+typedef struct {
+	unsigned long p4d;
+} p4d_t;
+
+#define p4d_val(x)	((x).p4d)
+#define __p4d(x)	((p4d_t) { (x) })
+#define PTRS_PER_P4D	(PAGE_SIZE / sizeof(p4d_t))
+
 /* Page Upper Directory entry */
 typedef struct {
 	unsigned long pud;
@@ -99,6 +116,15 @@ static inline struct page *pud_page(pud_t pud)
 	return pfn_to_page(pud_val(pud) >> _PAGE_PFN_SHIFT);
 }
 
+#define mm_p4d_folded  mm_p4d_folded
+static inline bool mm_p4d_folded(struct mm_struct *mm)
+{
+	if (pgtable_l5_enabled)
+		return false;
+
+	return true;
+}
+
 #define mm_pud_folded  mm_pud_folded
 static inline bool mm_pud_folded(struct mm_struct *mm)
 {
@@ -128,6 +154,9 @@ static inline unsigned long _pmd_pfn(pmd_t pmd)
 #define pud_ERROR(e)   \
 	pr_err("%s:%d: bad pud %016lx.\n", __FILE__, __LINE__, pud_val(e))
 
+#define p4d_ERROR(e)   \
+	pr_err("%s:%d: bad p4d %016lx.\n", __FILE__, __LINE__, p4d_val(e))
+
 static inline void set_p4d(p4d_t *p4dp, p4d_t p4d)
 {
 	if (pgtable_l4_enabled)
@@ -166,6 +195,16 @@ static inline void p4d_clear(p4d_t *p4d)
 		set_p4d(p4d, __p4d(0));
 }
 
+static inline p4d_t pfn_p4d(unsigned long pfn, pgprot_t prot)
+{
+	return __p4d((pfn << _PAGE_PFN_SHIFT) | pgprot_val(prot));
+}
+
+static inline unsigned long _p4d_pfn(p4d_t p4d)
+{
+	return p4d_val(p4d) >> _PAGE_PFN_SHIFT;
+}
+
 static inline pud_t *p4d_pgtable(p4d_t p4d)
 {
 	if (pgtable_l4_enabled)
@@ -173,6 +212,7 @@ static inline pud_t *p4d_pgtable(p4d_t p4d)
 
 	return (pud_t *)pud_pgtable((pud_t) { p4d_val(p4d) });
 }
+#define p4d_page_vaddr(p4d)	((unsigned long)p4d_pgtable(p4d))
 
 static inline struct page *p4d_page(p4d_t p4d)
 {
@@ -190,4 +230,68 @@ static inline pud_t *pud_offset(p4d_t *p4d, unsigned long address)
 	return (pud_t *)p4d;
 }
 
+static inline void set_pgd(pgd_t *pgdp, pgd_t pgd)
+{
+	if (pgtable_l5_enabled)
+		*pgdp = pgd;
+	else
+		set_p4d((p4d_t *)pgdp, (p4d_t){ pgd_val(pgd) });
+}
+
+static inline int pgd_none(pgd_t pgd)
+{
+	if (pgtable_l5_enabled)
+		return (pgd_val(pgd) == 0);
+
+	return 0;
+}
+
+static inline int pgd_present(pgd_t pgd)
+{
+	if (pgtable_l5_enabled)
+		return (pgd_val(pgd) & _PAGE_PRESENT);
+
+	return 1;
+}
+
+static inline int pgd_bad(pgd_t pgd)
+{
+	if (pgtable_l5_enabled)
+		return !pgd_present(pgd);
+
+	return 0;
+}
+
+static inline void pgd_clear(pgd_t *pgd)
+{
+	if (pgtable_l5_enabled)
+		set_pgd(pgd, __pgd(0));
+}
+
+static inline p4d_t *pgd_pgtable(pgd_t pgd)
+{
+	if (pgtable_l5_enabled)
+		return (p4d_t *)pfn_to_virt(pgd_val(pgd) >> _PAGE_PFN_SHIFT);
+
+	return (p4d_t *)p4d_pgtable((p4d_t) { pgd_val(pgd) });
+}
+#define pgd_page_vaddr(pgd)	((unsigned long)pgd_pgtable(pgd))
+
+static inline struct page *pgd_page(pgd_t pgd)
+{
+	return pfn_to_page(pgd_val(pgd) >> _PAGE_PFN_SHIFT);
+}
+#define pgd_page(pgd)	pgd_page(pgd)
+
+#define p4d_index(addr) (((addr) >> P4D_SHIFT) & (PTRS_PER_P4D - 1))
+
+#define p4d_offset p4d_offset
+static inline p4d_t *p4d_offset(pgd_t *pgd, unsigned long address)
+{
+	if (pgtable_l5_enabled)
+		return pgd_pgtable(*pgd) + p4d_index(address);
+
+	return (p4d_t *)pgd;
+}
+
 #endif /* _ASM_RISCV_PGTABLE_64_H */

arch/riscv/include/asm/pgtable.h

@@ -62,7 +62,8 @@
  * position vmemmap directly below the VMALLOC region.
  */
 #ifdef CONFIG_64BIT
-#define VA_BITS		(pgtable_l4_enabled ? 48 : 39)
+#define VA_BITS		(pgtable_l5_enabled ? \
+				57 : (pgtable_l4_enabled ? 48 : 39))
 #else
 #define VA_BITS		32
 #endif
@@ -102,7 +103,6 @@
 
 #ifndef __ASSEMBLY__
 
-#include <asm-generic/pgtable-nop4d.h>
 #include <asm/page.h>
 #include <asm/tlbflush.h>
 #include <linux/mm_types.h>
@@ -133,6 +133,8 @@ struct pt_alloc_ops {
 	phys_addr_t (*alloc_pmd)(uintptr_t va);
 	pud_t *(*get_pud_virt)(phys_addr_t pa);
 	phys_addr_t (*alloc_pud)(uintptr_t va);
+	p4d_t *(*get_p4d_virt)(phys_addr_t pa);
+	phys_addr_t (*alloc_p4d)(uintptr_t va);
 #endif
 };
 

arch/riscv/kernel/cpu.c

@@ -79,7 +79,9 @@ static void print_mmu(struct seq_file *f)
 #if defined(CONFIG_32BIT)
 	strncpy(sv_type, "sv32", 5);
 #elif defined(CONFIG_64BIT)
-	if (pgtable_l4_enabled)
+	if (pgtable_l5_enabled)
+		strncpy(sv_type, "sv57", 5);
+	else if (pgtable_l4_enabled)
 		strncpy(sv_type, "sv48", 5);
 	else
 		strncpy(sv_type, "sv39", 5);

arch/riscv/mm/init.c

@@ -38,14 +38,16 @@ EXPORT_SYMBOL(kernel_map);
 #endif
 
 #ifdef CONFIG_64BIT
-u64 satp_mode __ro_after_init = !IS_ENABLED(CONFIG_XIP_KERNEL) ? SATP_MODE_48 : SATP_MODE_39;
+u64 satp_mode __ro_after_init = !IS_ENABLED(CONFIG_XIP_KERNEL) ? SATP_MODE_57 : SATP_MODE_39;
 #else
 u64 satp_mode __ro_after_init = SATP_MODE_32;
 #endif
 EXPORT_SYMBOL(satp_mode);
 
 bool pgtable_l4_enabled = IS_ENABLED(CONFIG_64BIT) && !IS_ENABLED(CONFIG_XIP_KERNEL);
+bool pgtable_l5_enabled = IS_ENABLED(CONFIG_64BIT) && !IS_ENABLED(CONFIG_XIP_KERNEL);
 EXPORT_SYMBOL(pgtable_l4_enabled);
+EXPORT_SYMBOL(pgtable_l5_enabled);
 
 phys_addr_t phys_ram_base __ro_after_init;
 EXPORT_SYMBOL(phys_ram_base);
@@ -227,6 +229,7 @@ pgd_t trampoline_pg_dir[PTRS_PER_PGD] __page_aligned_bss;
 static pte_t fixmap_pte[PTRS_PER_PTE] __page_aligned_bss;
 
 pgd_t early_pg_dir[PTRS_PER_PGD] __initdata __aligned(PAGE_SIZE);
+static p4d_t __maybe_unused early_dtb_p4d[PTRS_PER_P4D] __initdata __aligned(PAGE_SIZE);
 static pud_t __maybe_unused early_dtb_pud[PTRS_PER_PUD] __initdata __aligned(PAGE_SIZE);
 static pmd_t __maybe_unused early_dtb_pmd[PTRS_PER_PMD] __initdata __aligned(PAGE_SIZE);
 
@@ -317,6 +320,16 @@ static pmd_t early_pmd[PTRS_PER_PMD] __initdata __aligned(PAGE_SIZE);
 #define early_pmd      ((pmd_t *)XIP_FIXUP(early_pmd))
 #endif /* CONFIG_XIP_KERNEL */
 
+static p4d_t trampoline_p4d[PTRS_PER_P4D] __page_aligned_bss;
+static p4d_t fixmap_p4d[PTRS_PER_P4D] __page_aligned_bss;
+static p4d_t early_p4d[PTRS_PER_P4D] __initdata __aligned(PAGE_SIZE);
+
+#ifdef CONFIG_XIP_KERNEL
+#define trampoline_p4d ((p4d_t *)XIP_FIXUP(trampoline_p4d))
+#define fixmap_p4d     ((p4d_t *)XIP_FIXUP(fixmap_p4d))
+#define early_p4d      ((p4d_t *)XIP_FIXUP(early_p4d))
+#endif /* CONFIG_XIP_KERNEL */
+
 static pud_t trampoline_pud[PTRS_PER_PUD] __page_aligned_bss;
 static pud_t fixmap_pud[PTRS_PER_PUD] __page_aligned_bss;
 static pud_t early_pud[PTRS_PER_PUD] __initdata __aligned(PAGE_SIZE);
@@ -431,6 +444,44 @@ static phys_addr_t alloc_pud_late(uintptr_t va)
 	return __pa(vaddr);
 }
 
+static p4d_t *__init get_p4d_virt_early(phys_addr_t pa)
+{
+	return (p4d_t *)((uintptr_t)pa);
+}
+
+static p4d_t *__init get_p4d_virt_fixmap(phys_addr_t pa)
+{
+	clear_fixmap(FIX_P4D);
+	return (p4d_t *)set_fixmap_offset(FIX_P4D, pa);
+}
+
+static p4d_t *__init get_p4d_virt_late(phys_addr_t pa)
+{
+	return (p4d_t *)__va(pa);
+}
+
+static phys_addr_t __init alloc_p4d_early(uintptr_t va)
+{
+	/* Only one P4D is available for early mapping */
+	BUG_ON((va - kernel_map.virt_addr) >> PGDIR_SHIFT);
+
+	return (uintptr_t)early_p4d;
+}
+
+static phys_addr_t __init alloc_p4d_fixmap(uintptr_t va)
+{
+	return memblock_phys_alloc(PAGE_SIZE, PAGE_SIZE);
+}
+
+static phys_addr_t alloc_p4d_late(uintptr_t va)
+{
+	unsigned long vaddr;
+
+	vaddr = __get_free_page(GFP_KERNEL);
+	BUG_ON(!vaddr);
+	return __pa(vaddr);
+}
+
 static void __init create_pud_mapping(pud_t *pudp,
 				      uintptr_t va, phys_addr_t pa,
 				      phys_addr_t sz, pgprot_t prot)
@@ -458,21 +509,55 @@ static void __init create_pud_mapping(pud_t *pudp,
 	create_pmd_mapping(nextp, va, pa, sz, prot);
 }
 
-#define pgd_next_t		pud_t
-#define alloc_pgd_next(__va)	(pgtable_l4_enabled ?			\
-		pt_ops.alloc_pud(__va) : pt_ops.alloc_pmd(__va))
-#define get_pgd_next_virt(__pa)	(pgtable_l4_enabled ?			\
-		pt_ops.get_pud_virt(__pa) : (pgd_next_t *)pt_ops.get_pmd_virt(__pa))
+static void __init create_p4d_mapping(p4d_t *p4dp,
+				      uintptr_t va, phys_addr_t pa,
+				      phys_addr_t sz, pgprot_t prot)
+{
+	pud_t *nextp;
+	phys_addr_t next_phys;
+	uintptr_t p4d_index = p4d_index(va);
+
+	if (sz == P4D_SIZE) {
+		if (p4d_val(p4dp[p4d_index]) == 0)
+			p4dp[p4d_index] = pfn_p4d(PFN_DOWN(pa), prot);
+		return;
+	}
+
+	if (p4d_val(p4dp[p4d_index]) == 0) {
+		next_phys = pt_ops.alloc_pud(va);
+		p4dp[p4d_index] = pfn_p4d(PFN_DOWN(next_phys), PAGE_TABLE);
+		nextp = pt_ops.get_pud_virt(next_phys);
+		memset(nextp, 0, PAGE_SIZE);
+	} else {
+		next_phys = PFN_PHYS(_p4d_pfn(p4dp[p4d_index]));
+		nextp = pt_ops.get_pud_virt(next_phys);
+	}
+
+	create_pud_mapping(nextp, va, pa, sz, prot);
+}
+
+#define pgd_next_t		p4d_t
+#define alloc_pgd_next(__va)	(pgtable_l5_enabled ?			\
+		pt_ops.alloc_p4d(__va) : (pgtable_l4_enabled ?		\
+		pt_ops.alloc_pud(__va) : pt_ops.alloc_pmd(__va)))
+#define get_pgd_next_virt(__pa)	(pgtable_l5_enabled ?			\
+		pt_ops.get_p4d_virt(__pa) : (pgd_next_t *)(pgtable_l4_enabled ?	\
+		pt_ops.get_pud_virt(__pa) : (pud_t *)pt_ops.get_pmd_virt(__pa)))
 #define create_pgd_next_mapping(__nextp, __va, __pa, __sz, __prot)	\
+				(pgtable_l5_enabled ?			\
+		create_p4d_mapping(__nextp, __va, __pa, __sz, __prot) : \
 				(pgtable_l4_enabled ?			\
-		create_pud_mapping(__nextp, __va, __pa, __sz, __prot) :	\
-		create_pmd_mapping((pmd_t *)__nextp, __va, __pa, __sz, __prot))
-#define fixmap_pgd_next		(pgtable_l4_enabled ?			\
-		(uintptr_t)fixmap_pud : (uintptr_t)fixmap_pmd)
-#define trampoline_pgd_next	(pgtable_l4_enabled ?			\
-		(uintptr_t)trampoline_pud : (uintptr_t)trampoline_pmd)
-#define early_dtb_pgd_next	(pgtable_l4_enabled ?			\
-		(uintptr_t)early_dtb_pud : (uintptr_t)early_dtb_pmd)
+		create_pud_mapping((pud_t *)__nextp, __va, __pa, __sz, __prot) :	\
+		create_pmd_mapping((pmd_t *)__nextp, __va, __pa, __sz, __prot)))
+#define fixmap_pgd_next		(pgtable_l5_enabled ?			\
+		(uintptr_t)fixmap_p4d : (pgtable_l4_enabled ?		\
+		(uintptr_t)fixmap_pud : (uintptr_t)fixmap_pmd))
+#define trampoline_pgd_next	(pgtable_l5_enabled ?			\
+		(uintptr_t)trampoline_p4d : (pgtable_l4_enabled ?	\
+		(uintptr_t)trampoline_pud : (uintptr_t)trampoline_pmd))
+#define early_dtb_pgd_next	(pgtable_l5_enabled ?			\
+		(uintptr_t)early_dtb_p4d : (pgtable_l4_enabled ?	\
+		(uintptr_t)early_dtb_pud : (uintptr_t)early_dtb_pmd))
 #else
 #define pgd_next_t		pte_t
 #define alloc_pgd_next(__va)	pt_ops.alloc_pte(__va)
@@ -481,6 +566,7 @@ static void __init create_pud_mapping(pud_t *pudp,
 	create_pte_mapping(__nextp, __va, __pa, __sz, __prot)
 #define fixmap_pgd_next		((uintptr_t)fixmap_pte)
 #define early_dtb_pgd_next	((uintptr_t)early_dtb_pmd)
+#define create_p4d_mapping(__pmdp, __va, __pa, __sz, __prot)
 #define create_pud_mapping(__pmdp, __va, __pa, __sz, __prot)
 #define create_pmd_mapping(__pmdp, __va, __pa, __sz, __prot)
 #endif /* __PAGETABLE_PMD_FOLDED */
@@ -573,6 +659,13 @@ static __init pgprot_t pgprot_from_va(uintptr_t va)
 #endif /* CONFIG_STRICT_KERNEL_RWX */
 
 #ifdef CONFIG_64BIT
+static void __init disable_pgtable_l5(void)
+{
+	pgtable_l5_enabled = false;
+	kernel_map.page_offset = PAGE_OFFSET_L4;
+	satp_mode = SATP_MODE_48;
+}
+
 static void __init disable_pgtable_l4(void)
 {
 	pgtable_l4_enabled = false;
@@ -589,12 +682,12 @@ static void __init disable_pgtable_l4(void)
 static __init void set_satp_mode(void)
 {
 	u64 identity_satp, hw_satp;
-	uintptr_t set_satp_mode_pmd;
+	uintptr_t set_satp_mode_pmd = ((unsigned long)set_satp_mode) & PMD_MASK;
+	bool check_l4 = false;
 
-	set_satp_mode_pmd = ((unsigned long)set_satp_mode) & PMD_MASK;
-	create_pgd_mapping(early_pg_dir,
-			   set_satp_mode_pmd, (uintptr_t)early_pud,
-			   PGDIR_SIZE, PAGE_TABLE);
+	create_p4d_mapping(early_p4d,
+			set_satp_mode_pmd, (uintptr_t)early_pud,
+			P4D_SIZE, PAGE_TABLE);
 	create_pud_mapping(early_pud,
 			   set_satp_mode_pmd, (uintptr_t)early_pmd,
 			   PUD_SIZE, PAGE_TABLE);
@@ -606,6 +699,11 @@ static __init void set_satp_mode(void)
 			   set_satp_mode_pmd + PMD_SIZE,
 			   set_satp_mode_pmd + PMD_SIZE,
 			   PMD_SIZE, PAGE_KERNEL_EXEC);
+retry:
+	create_pgd_mapping(early_pg_dir,
+			   set_satp_mode_pmd,
+			   check_l4 ? (uintptr_t)early_pud : (uintptr_t)early_p4d,
+			   PGDIR_SIZE, PAGE_TABLE);
 
 	identity_satp = PFN_DOWN((uintptr_t)&early_pg_dir) | satp_mode;
 
@@ -614,10 +712,17 @@ static __init void set_satp_mode(void)
 	hw_satp = csr_swap(CSR_SATP, 0ULL);
 	local_flush_tlb_all();
 
-	if (hw_satp != identity_satp)
+	if (hw_satp != identity_satp) {
+		if (!check_l4) {
+			disable_pgtable_l5();
+			check_l4 = true;
+			goto retry;
+		}
 		disable_pgtable_l4();
+	}
 
 	memset(early_pg_dir, 0, PAGE_SIZE);
+	memset(early_p4d, 0, PAGE_SIZE);
 	memset(early_pud, 0, PAGE_SIZE);
 	memset(early_pmd, 0, PAGE_SIZE);
 }
@@ -691,10 +796,13 @@ static void __init create_fdt_early_page_table(pgd_t *pgdir, uintptr_t dtb_pa)
 			   PGDIR_SIZE,
 			   IS_ENABLED(CONFIG_64BIT) ? PAGE_TABLE : PAGE_KERNEL);
 
-	if (pgtable_l4_enabled) {
+	if (pgtable_l5_enabled)
+		create_p4d_mapping(early_dtb_p4d, DTB_EARLY_BASE_VA,
+				   (uintptr_t)early_dtb_pud, P4D_SIZE, PAGE_TABLE);
+
+	if (pgtable_l4_enabled)
 		create_pud_mapping(early_dtb_pud, DTB_EARLY_BASE_VA,
 				   (uintptr_t)early_dtb_pmd, PUD_SIZE, PAGE_TABLE);
-	}
 
 	if (IS_ENABLED(CONFIG_64BIT)) {
 		create_pmd_mapping(early_dtb_pmd, DTB_EARLY_BASE_VA,
@@ -730,6 +838,8 @@ void __init pt_ops_set_early(void)
 	pt_ops.get_pmd_virt = get_pmd_virt_early;
 	pt_ops.alloc_pud = alloc_pud_early;
 	pt_ops.get_pud_virt = get_pud_virt_early;
+	pt_ops.alloc_p4d = alloc_p4d_early;
+	pt_ops.get_p4d_virt = get_p4d_virt_early;
 #endif
 }
 
@@ -750,6 +860,8 @@ void __init pt_ops_set_fixmap(void)
 	pt_ops.get_pmd_virt = kernel_mapping_pa_to_va((uintptr_t)get_pmd_virt_fixmap);
 	pt_ops.alloc_pud = kernel_mapping_pa_to_va((uintptr_t)alloc_pud_fixmap);
 	pt_ops.get_pud_virt = kernel_mapping_pa_to_va((uintptr_t)get_pud_virt_fixmap);
+	pt_ops.alloc_p4d = kernel_mapping_pa_to_va((uintptr_t)alloc_p4d_fixmap);
+	pt_ops.get_p4d_virt = kernel_mapping_pa_to_va((uintptr_t)get_p4d_virt_fixmap);
 #endif
 }
 
@@ -766,6 +878,8 @@ void __init pt_ops_set_late(void)
 	pt_ops.get_pmd_virt = get_pmd_virt_late;
 	pt_ops.alloc_pud = alloc_pud_late;
 	pt_ops.get_pud_virt = get_pud_virt_late;
+	pt_ops.alloc_p4d = alloc_p4d_late;
+	pt_ops.get_p4d_virt = get_p4d_virt_late;
 #endif
 }
 
@@ -826,6 +940,10 @@ asmlinkage void __init setup_vm(uintptr_t dtb_pa)
 			   fixmap_pgd_next, PGDIR_SIZE, PAGE_TABLE);
 
 #ifndef __PAGETABLE_PMD_FOLDED
+	/* Setup fixmap P4D and PUD */
+	if (pgtable_l5_enabled)
+		create_p4d_mapping(fixmap_p4d, FIXADDR_START,
+				   (uintptr_t)fixmap_pud, P4D_SIZE, PAGE_TABLE);
 	/* Setup fixmap PUD and PMD */
 	if (pgtable_l4_enabled)
 		create_pud_mapping(fixmap_pud, FIXADDR_START,
@@ -835,6 +953,9 @@ asmlinkage void __init setup_vm(uintptr_t dtb_pa)
 	/* Setup trampoline PGD and PMD */
 	create_pgd_mapping(trampoline_pg_dir, kernel_map.virt_addr,
 			   trampoline_pgd_next, PGDIR_SIZE, PAGE_TABLE);
+	if (pgtable_l5_enabled)
+		create_p4d_mapping(trampoline_p4d, kernel_map.virt_addr,
+				   (uintptr_t)trampoline_pud, P4D_SIZE, PAGE_TABLE);
 	if (pgtable_l4_enabled)
 		create_pud_mapping(trampoline_pud, kernel_map.virt_addr,
 				   (uintptr_t)trampoline_pmd, PUD_SIZE, PAGE_TABLE);
@@ -936,6 +1057,7 @@ static void __init setup_vm_final(void)
 	clear_fixmap(FIX_PTE);
 	clear_fixmap(FIX_PMD);
 	clear_fixmap(FIX_PUD);
+	clear_fixmap(FIX_P4D);
 
 	/* Move to swapper page table */
 	csr_write(CSR_SATP, PFN_DOWN(__pa_symbol(swapper_pg_dir)) | satp_mode);

arch/riscv/mm/kasan_init.c

@@ -111,6 +111,8 @@ static void __init kasan_populate_pud(pgd_t *pgd,
 		 * pt_ops facility.
 		 */
 		base_pud = pt_ops.get_pud_virt(pfn_to_phys(_pgd_pfn(*pgd)));
+	} else if (pgd_none(*pgd)) {
+		base_pud = memblock_alloc(PTRS_PER_PUD * sizeof(pud_t), PAGE_SIZE);
 	} else {
 		base_pud = (pud_t *)pgd_page_vaddr(*pgd);
 		if (base_pud == lm_alias(kasan_early_shadow_pud))
@@ -149,13 +151,72 @@ static void __init kasan_populate_pud(pgd_t *pgd,
 		set_pgd(pgd, pfn_pgd(PFN_DOWN(__pa(base_pud)), PAGE_TABLE));
 }
 
-#define kasan_early_shadow_pgd_next			(pgtable_l4_enabled ?	\
+static void __init kasan_populate_p4d(pgd_t *pgd,
+				      unsigned long vaddr, unsigned long end,
+				      bool early)
+{
+	phys_addr_t phys_addr;
+	p4d_t *p4dp, *base_p4d;
+	unsigned long next;
+
+	if (early) {
+		/*
+		 * We can't use pgd_page_vaddr here as it would return a linear
+		 * mapping address but it is not mapped yet, but when populating
+		 * early_pg_dir, we need the physical address and when populating
+		 * swapper_pg_dir, we need the kernel virtual address so use
+		 * pt_ops facility.
+		 */
+		base_p4d = pt_ops.get_p4d_virt(pfn_to_phys(_pgd_pfn(*pgd)));
+	} else {
+		base_p4d = (p4d_t *)pgd_page_vaddr(*pgd);
+		if (base_p4d == lm_alias(kasan_early_shadow_p4d))
+			base_p4d = memblock_alloc(PTRS_PER_PUD * sizeof(p4d_t), PAGE_SIZE);
+	}
+
+	p4dp = base_p4d + p4d_index(vaddr);
+
+	do {
+		next = p4d_addr_end(vaddr, end);
+
+		if (p4d_none(*p4dp) && IS_ALIGNED(vaddr, P4D_SIZE) && (next - vaddr) >= P4D_SIZE) {
+			if (early) {
+				phys_addr = __pa(((uintptr_t)kasan_early_shadow_pud));
+				set_p4d(p4dp, pfn_p4d(PFN_DOWN(phys_addr), PAGE_TABLE));
+				continue;
+			} else {
+				phys_addr = memblock_phys_alloc(P4D_SIZE, P4D_SIZE);
+				if (phys_addr) {
+					set_p4d(p4dp, pfn_p4d(PFN_DOWN(phys_addr), PAGE_KERNEL));
+					continue;
+				}
+			}
+		}
+
+		kasan_populate_pud((pgd_t *)p4dp, vaddr, next, early);
+	} while (p4dp++, vaddr = next, vaddr != end);
+
+	/*
+	 * Wait for the whole P4D to be populated before setting the P4D in
+	 * the page table, otherwise, if we did set the P4D before populating
+	 * it entirely, memblock could allocate a page at a physical address
+	 * where KASAN is not populated yet and then we'd get a page fault.
+	 */
+	if (!early)
+		set_pgd(pgd, pfn_pgd(PFN_DOWN(__pa(base_p4d)), PAGE_TABLE));
+}
+
+#define kasan_early_shadow_pgd_next			(pgtable_l5_enabled ?	\
+				(uintptr_t)kasan_early_shadow_p4d :		\
+							(pgtable_l4_enabled ?	\
 				(uintptr_t)kasan_early_shadow_pud :		\
-				(uintptr_t)kasan_early_shadow_pmd)
+				(uintptr_t)kasan_early_shadow_pmd))
 #define kasan_populate_pgd_next(pgdp, vaddr, next, early)			\
+		(pgtable_l5_enabled ?						\
+		kasan_populate_p4d(pgdp, vaddr, next, early) :			\
 		(pgtable_l4_enabled ?						\
 			kasan_populate_pud(pgdp, vaddr, next, early) :		\
-			kasan_populate_pmd((pud_t *)pgdp, vaddr, next))
+			kasan_populate_pmd((pud_t *)pgdp, vaddr, next)))
 
 static void __init kasan_populate_pgd(pgd_t *pgdp,
 				      unsigned long vaddr, unsigned long end,
@@ -219,6 +280,14 @@ asmlinkage void __init kasan_early_init(void)
 					PAGE_TABLE));
 	}
 
+	if (pgtable_l5_enabled) {
+		for (i = 0; i < PTRS_PER_P4D; ++i)
+			set_p4d(kasan_early_shadow_p4d + i,
+				pfn_p4d(PFN_DOWN
+					(__pa(((uintptr_t)kasan_early_shadow_pud))),
+					PAGE_TABLE));
+	}
+
 	kasan_populate_pgd(early_pg_dir + pgd_index(KASAN_SHADOW_START),
 			   KASAN_SHADOW_START, KASAN_SHADOW_END, true);
 
@@ -244,9 +313,27 @@ static void __init kasan_populate(void *start, void *end)
 	memset(start, KASAN_SHADOW_INIT, end - start);
 }
 
+static void __init kasan_shallow_populate_pmd(pgd_t *pgdp,
+					      unsigned long vaddr, unsigned long end)
+{
+	unsigned long next;
+	pmd_t *pmdp, *base_pmd;
+	bool is_kasan_pte;
+
+	base_pmd = (pmd_t *)pgd_page_vaddr(*pgdp);
+	pmdp = base_pmd + pmd_index(vaddr);
+
+	do {
+		next = pmd_addr_end(vaddr, end);
+		is_kasan_pte = (pmd_pgtable(*pmdp) == lm_alias(kasan_early_shadow_pte));
+
+		if (is_kasan_pte)
+			pmd_clear(pmdp);
+	} while (pmdp++, vaddr = next, vaddr != end);
+}
+
 static void __init kasan_shallow_populate_pud(pgd_t *pgdp,
-					      unsigned long vaddr, unsigned long end,
-					      bool kasan_populate)
+					      unsigned long vaddr, unsigned long end)
 {
 	unsigned long next;
 	pud_t *pudp, *base_pud;
@@ -256,21 +343,60 @@ static void __init kasan_shallow_populate_pud(pgd_t *pgdp,
 	base_pud = (pud_t *)pgd_page_vaddr(*pgdp);
 	pudp = base_pud + pud_index(vaddr);
 
-	if (kasan_populate)
-		memcpy(base_pud, (void *)kasan_early_shadow_pgd_next,
-		       sizeof(pud_t) * PTRS_PER_PUD);
-
 	do {
 		next = pud_addr_end(vaddr, end);
 		is_kasan_pmd = (pud_pgtable(*pudp) == lm_alias(kasan_early_shadow_pmd));
 
-		if (is_kasan_pmd) {
-			base_pmd = memblock_alloc(PAGE_SIZE, PAGE_SIZE);
-			set_pud(pudp, pfn_pud(PFN_DOWN(__pa(base_pmd)), PAGE_TABLE));
-		}
+		if (!is_kasan_pmd)
+			continue;
+
+		base_pmd = memblock_alloc(PAGE_SIZE, PAGE_SIZE);
+		set_pud(pudp, pfn_pud(PFN_DOWN(__pa(base_pmd)), PAGE_TABLE));
+
+		if (IS_ALIGNED(vaddr, PUD_SIZE) && (next - vaddr) >= PUD_SIZE)
+			continue;
+
+		memcpy(base_pmd, (void *)kasan_early_shadow_pmd, PAGE_SIZE);
+		kasan_shallow_populate_pmd((pgd_t *)pudp, vaddr, next);
 	} while (pudp++, vaddr = next, vaddr != end);
 }
 
+static void __init kasan_shallow_populate_p4d(pgd_t *pgdp,
+					      unsigned long vaddr, unsigned long end)
+{
+	unsigned long next;
+	p4d_t *p4dp, *base_p4d;
+	pud_t *base_pud;
+	bool is_kasan_pud;
+
+	base_p4d = (p4d_t *)pgd_page_vaddr(*pgdp);
+	p4dp = base_p4d + p4d_index(vaddr);
+
+	do {
+		next = p4d_addr_end(vaddr, end);
+		is_kasan_pud = (p4d_pgtable(*p4dp) == lm_alias(kasan_early_shadow_pud));
+
+		if (!is_kasan_pud)
+			continue;
+
+		base_pud = memblock_alloc(PAGE_SIZE, PAGE_SIZE);
+		set_p4d(p4dp, pfn_p4d(PFN_DOWN(__pa(base_pud)), PAGE_TABLE));
+
+		if (IS_ALIGNED(vaddr, P4D_SIZE) && (next - vaddr) >= P4D_SIZE)
+			continue;
+
+		memcpy(base_pud, (void *)kasan_early_shadow_pud, PAGE_SIZE);
+		kasan_shallow_populate_pud((pgd_t *)p4dp, vaddr, next);
+	} while (p4dp++, vaddr = next, vaddr != end);
+}
+
+#define kasan_shallow_populate_pgd_next(pgdp, vaddr, next)			\
+		(pgtable_l5_enabled ?						\
+		kasan_shallow_populate_p4d(pgdp, vaddr, next) :			\
+		(pgtable_l4_enabled ?						\
+		kasan_shallow_populate_pud(pgdp, vaddr, next) :			\
+		kasan_shallow_populate_pmd(pgdp, vaddr, next)))
+
 static void __init kasan_shallow_populate_pgd(unsigned long vaddr, unsigned long end)
 {
 	unsigned long next;
@@ -291,7 +417,8 @@ static void __init kasan_shallow_populate_pgd(unsigned long vaddr, unsigned long
 		if (IS_ALIGNED(vaddr, PGDIR_SIZE) && (next - vaddr) >= PGDIR_SIZE)
 			continue;
 
-		kasan_shallow_populate_pud(pgd_k, vaddr, next, is_kasan_pgd_next);
+		memcpy(p, (void *)kasan_early_shadow_pgd_next, PAGE_SIZE);
+		kasan_shallow_populate_pgd_next(pgd_k, vaddr, next);
 	} while (pgd_k++, vaddr = next, vaddr != end);
 }