riscv: mm: Control p4d's folding by pgtable_l5_enabled

To determine pgtable level at boot time, we can not use helper functions
in include/asm-generic/pgtable-nop4d.h and must implement these
functions. This patch uses pgtable_l5_enabled variable instead of
including pgtable-nop4d.h to controle p4d's folding, and implements
corresponding helper functions.
Signed-off-by: Qinglin Pan <panqinglin2020@iscas.ac.cn>
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

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>

arch/riscv/mm/init.c

@@ -45,7 +45,9 @@ u64 satp_mode = SATP_MODE_32;
 EXPORT_SYMBOL(satp_mode);
 
 bool pgtable_l4_enabled = IS_ENABLED(CONFIG_64BIT) && !IS_ENABLED(CONFIG_XIP_KERNEL);
+bool pgtable_l5_enabled = false;
 EXPORT_SYMBOL(pgtable_l4_enabled);
+EXPORT_SYMBOL(pgtable_l5_enabled);
 
 phys_addr_t phys_ram_base __ro_after_init;
 EXPORT_SYMBOL(phys_ram_base);