powerpc: add support for folded p4d page tables

Implement primitives necessary for the 4th level folding, add walks of p4d
level where appropriate and replace 5level-fixup.h with pgtable-nop4d.h.

[rppt@linux.ibm.com: powerpc/xmon: drop unused pgdir varialble in show_pte() function]
  Link: http://lkml.kernel.org/r/20200519181454.GI1059226@linux.ibm.com
[rppt@linux.ibm.com; build fix]
  Link: http://lkml.kernel.org/r/20200423141845.GI13521@linux.ibm.com
Signed-off-by: Mike Rapoport <rppt@linux.ibm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Tested-by: Christophe Leroy <christophe.leroy@c-s.fr> # 8xx and 83xx
Cc: Arnd Bergmann <arnd@arndb.de>
Cc: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Cc: Brian Cain <bcain@codeaurora.org>
Cc: Catalin Marinas <catalin.marinas@arm.com>
Cc: Fenghua Yu <fenghua.yu@intel.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Guan Xuetao <gxt@pku.edu.cn>
Cc: James Morse <james.morse@arm.com>
Cc: Jonas Bonn <jonas@southpole.se>
Cc: Julien Thierry <julien.thierry.kdev@gmail.com>
Cc: Ley Foon Tan <ley.foon.tan@intel.com>
Cc: Marc Zyngier <maz@kernel.org>
Cc: Michael Ellerman <mpe@ellerman.id.au>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Rich Felker <dalias@libc.org>
Cc: Russell King <linux@armlinux.org.uk>
Cc: Stafford Horne <shorne@gmail.com>
Cc: Stefan Kristiansson <stefan.kristiansson@saunalahti.fi>
Cc: Suzuki K Poulose <suzuki.poulose@arm.com>
Cc: Tony Luck <tony.luck@intel.com>
Cc: Will Deacon <will@kernel.org>
Cc: Yoshinori Sato <ysato@users.sourceforge.jp>
Link: http://lkml.kernel.org/r/20200414153455.21744-9-rppt@kernel.org
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>

arch/powerpc/include/asm/book3s/32/pgtable.h

@@ -2,7 +2,6 @@
 #ifndef _ASM_POWERPC_BOOK3S_32_PGTABLE_H
 #define _ASM_POWERPC_BOOK3S_32_PGTABLE_H
 
-#define __ARCH_USE_5LEVEL_HACK
 #include <asm-generic/pgtable-nopmd.h>
 
 #include <asm/book3s/32/hash.h>

arch/powerpc/include/asm/book3s/64/hash.h

@@ -134,9 +134,9 @@ static inline int get_region_id(unsigned long ea)
 
 #define	hash__pmd_bad(pmd)		(pmd_val(pmd) & H_PMD_BAD_BITS)
 #define	hash__pud_bad(pud)		(pud_val(pud) & H_PUD_BAD_BITS)
-static inline int hash__pgd_bad(pgd_t pgd)
+static inline int hash__p4d_bad(p4d_t p4d)
 {
-	return (pgd_val(pgd) == 0);
+	return (p4d_val(p4d) == 0);
 }
 #ifdef CONFIG_STRICT_KERNEL_RWX
 extern void hash__mark_rodata_ro(void);

arch/powerpc/include/asm/book3s/64/pgalloc.h

@@ -85,9 +85,9 @@ static inline void pgd_free(struct mm_struct *mm, pgd_t *pgd)
 	kmem_cache_free(PGT_CACHE(PGD_INDEX_SIZE), pgd);
 }
 
-static inline void pgd_populate(struct mm_struct *mm, pgd_t *pgd, pud_t *pud)
+static inline void p4d_populate(struct mm_struct *mm, p4d_t *pgd, pud_t *pud)
 {
-	*pgd =  __pgd(__pgtable_ptr_val(pud) | PGD_VAL_BITS);
+	*pgd =  __p4d(__pgtable_ptr_val(pud) | PGD_VAL_BITS);
 }
 
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)

arch/powerpc/include/asm/book3s/64/pgtable.h

@@ -2,7 +2,7 @@
 #ifndef _ASM_POWERPC_BOOK3S_64_PGTABLE_H_
 #define _ASM_POWERPC_BOOK3S_64_PGTABLE_H_
 
-#include <asm-generic/5level-fixup.h>
+#include <asm-generic/pgtable-nop4d.h>
 
 #ifndef __ASSEMBLY__
 #include <linux/mmdebug.h>
@@ -251,7 +251,7 @@ extern unsigned long __pmd_frag_size_shift;
 /* Bits to mask out from a PUD to get to the PMD page */
 #define PUD_MASKED_BITS		0xc0000000000000ffUL
 /* Bits to mask out from a PGD to get to the PUD page */
-#define PGD_MASKED_BITS		0xc0000000000000ffUL
+#define P4D_MASKED_BITS		0xc0000000000000ffUL
 
 /*
  * Used as an indicator for rcu callback functions
@@ -949,54 +949,60 @@ static inline bool pud_access_permitted(pud_t pud, bool write)
 	return pte_access_permitted(pud_pte(pud), write);
 }
 
-#define pgd_write(pgd)		pte_write(pgd_pte(pgd))
+#define __p4d_raw(x)	((p4d_t) { __pgd_raw(x) })
+static inline __be64 p4d_raw(p4d_t x)
+{
+	return pgd_raw(x.pgd);
+}
+
+#define p4d_write(p4d)		pte_write(p4d_pte(p4d))
 
-static inline void pgd_clear(pgd_t *pgdp)
+static inline void p4d_clear(p4d_t *p4dp)
 {
-	*pgdp = __pgd(0);
+	*p4dp = __p4d(0);
 }
 
-static inline int pgd_none(pgd_t pgd)
+static inline int p4d_none(p4d_t p4d)
 {
-	return !pgd_raw(pgd);
+	return !p4d_raw(p4d);
 }
 
-static inline int pgd_present(pgd_t pgd)
+static inline int p4d_present(p4d_t p4d)
 {
-	return !!(pgd_raw(pgd) & cpu_to_be64(_PAGE_PRESENT));
+	return !!(p4d_raw(p4d) & cpu_to_be64(_PAGE_PRESENT));
 }
 
-static inline pte_t pgd_pte(pgd_t pgd)
+static inline pte_t p4d_pte(p4d_t p4d)
 {
-	return __pte_raw(pgd_raw(pgd));
+	return __pte_raw(p4d_raw(p4d));
 }
 
-static inline pgd_t pte_pgd(pte_t pte)
+static inline p4d_t pte_p4d(pte_t pte)
 {
-	return __pgd_raw(pte_raw(pte));
+	return __p4d_raw(pte_raw(pte));
 }
 
-static inline int pgd_bad(pgd_t pgd)
+static inline int p4d_bad(p4d_t p4d)
 {
 	if (radix_enabled())
-		return radix__pgd_bad(pgd);
-	return hash__pgd_bad(pgd);
+		return radix__p4d_bad(p4d);
+	return hash__p4d_bad(p4d);
 }
 
-#define pgd_access_permitted pgd_access_permitted
-static inline bool pgd_access_permitted(pgd_t pgd, bool write)
+#define p4d_access_permitted p4d_access_permitted
+static inline bool p4d_access_permitted(p4d_t p4d, bool write)
 {
-	return pte_access_permitted(pgd_pte(pgd), write);
+	return pte_access_permitted(p4d_pte(p4d), write);
 }
 
-extern struct page *pgd_page(pgd_t pgd);
+extern struct page *p4d_page(p4d_t p4d);
 
 /* Pointers in the page table tree are physical addresses */
 #define __pgtable_ptr_val(ptr)	__pa(ptr)
 
 #define pmd_page_vaddr(pmd)	__va(pmd_val(pmd) & ~PMD_MASKED_BITS)
 #define pud_page_vaddr(pud)	__va(pud_val(pud) & ~PUD_MASKED_BITS)
-#define pgd_page_vaddr(pgd)	__va(pgd_val(pgd) & ~PGD_MASKED_BITS)
+#define p4d_page_vaddr(p4d)	__va(p4d_val(p4d) & ~P4D_MASKED_BITS)
 
 #define pgd_index(address) (((address) >> (PGDIR_SHIFT)) & (PTRS_PER_PGD - 1))
 #define pud_index(address) (((address) >> (PUD_SHIFT)) & (PTRS_PER_PUD - 1))
@@ -1010,8 +1016,8 @@ extern struct page *pgd_page(pgd_t pgd);
 
 #define pgd_offset(mm, address)	 ((mm)->pgd + pgd_index(address))
 
-#define pud_offset(pgdp, addr)	\
-	(((pud_t *) pgd_page_vaddr(*(pgdp))) + pud_index(addr))
+#define pud_offset(p4dp, addr)	\
+	(((pud_t *) p4d_page_vaddr(*(p4dp))) + pud_index(addr))
 #define pmd_offset(pudp,addr) \
 	(((pmd_t *) pud_page_vaddr(*(pudp))) + pmd_index(addr))
 #define pte_offset_kernel(dir,addr) \
@@ -1366,11 +1372,11 @@ static inline bool pud_is_leaf(pud_t pud)
 	return !!(pud_raw(pud) & cpu_to_be64(_PAGE_PTE));
 }
 
-#define pgd_is_leaf pgd_is_leaf
-#define pgd_leaf pgd_is_leaf
-static inline bool pgd_is_leaf(pgd_t pgd)
+#define p4d_is_leaf p4d_is_leaf
+#define p4d_leaf p4d_is_leaf
+static inline bool p4d_is_leaf(p4d_t p4d)
 {
-	return !!(pgd_raw(pgd) & cpu_to_be64(_PAGE_PTE));
+	return !!(p4d_raw(p4d) & cpu_to_be64(_PAGE_PTE));
 }
 
 #endif /* __ASSEMBLY__ */

arch/powerpc/include/asm/book3s/64/radix.h

@@ -30,7 +30,7 @@
 /* Don't have anything in the reserved bits and leaf bits */
 #define RADIX_PMD_BAD_BITS		0x60000000000000e0UL
 #define RADIX_PUD_BAD_BITS		0x60000000000000e0UL
-#define RADIX_PGD_BAD_BITS		0x60000000000000e0UL
+#define RADIX_P4D_BAD_BITS		0x60000000000000e0UL
 
 #define RADIX_PMD_SHIFT		(PAGE_SHIFT + RADIX_PTE_INDEX_SIZE)
 #define RADIX_PUD_SHIFT		(RADIX_PMD_SHIFT + RADIX_PMD_INDEX_SIZE)
@@ -227,9 +227,9 @@ static inline int radix__pud_bad(pud_t pud)
 }
 
 
-static inline int radix__pgd_bad(pgd_t pgd)
+static inline int radix__p4d_bad(p4d_t p4d)
 {
-	return !!(pgd_val(pgd) & RADIX_PGD_BAD_BITS);
+	return !!(p4d_val(p4d) & RADIX_P4D_BAD_BITS);
 }
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE

arch/powerpc/include/asm/nohash/32/pgtable.h

@@ -2,7 +2,6 @@
 #ifndef _ASM_POWERPC_NOHASH_32_PGTABLE_H
 #define _ASM_POWERPC_NOHASH_32_PGTABLE_H
 
-#define __ARCH_USE_5LEVEL_HACK
 #include <asm-generic/pgtable-nopmd.h>
 
 #ifndef __ASSEMBLY__

arch/powerpc/include/asm/nohash/64/pgalloc.h

@@ -15,7 +15,7 @@ struct vmemmap_backing {
 };
 extern struct vmemmap_backing *vmemmap_list;
 
-#define pgd_populate(MM, PGD, PUD)	pgd_set(PGD, (unsigned long)PUD)
+#define p4d_populate(MM, P4D, PUD)	p4d_set(P4D, (unsigned long)PUD)
 
 static inline pud_t *pud_alloc_one(struct mm_struct *mm, unsigned long addr)
 {

arch/powerpc/include/asm/nohash/64/pgtable-4k.h

@@ -2,7 +2,7 @@
 #ifndef _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H
 #define _ASM_POWERPC_NOHASH_64_PGTABLE_4K_H
 
-#include <asm-generic/5level-fixup.h>
+#include <asm-generic/pgtable-nop4d.h>
 
 /*
  * Entries per page directory level.  The PTE level must use a 64b record
@@ -45,41 +45,41 @@
 #define PMD_MASKED_BITS		0
 /* Bits to mask out from a PUD to get to the PMD page */
 #define PUD_MASKED_BITS		0
-/* Bits to mask out from a PGD to get to the PUD page */
-#define PGD_MASKED_BITS		0
+/* Bits to mask out from a P4D to get to the PUD page */
+#define P4D_MASKED_BITS		0
 
 
 /*
  * 4-level page tables related bits
  */
 
-#define pgd_none(pgd)		(!pgd_val(pgd))
-#define pgd_bad(pgd)		(pgd_val(pgd) == 0)
-#define pgd_present(pgd)	(pgd_val(pgd) != 0)
-#define pgd_page_vaddr(pgd)	(pgd_val(pgd) & ~PGD_MASKED_BITS)
+#define p4d_none(p4d)		(!p4d_val(p4d))
+#define p4d_bad(p4d)		(p4d_val(p4d) == 0)
+#define p4d_present(p4d)	(p4d_val(p4d) != 0)
+#define p4d_page_vaddr(p4d)	(p4d_val(p4d) & ~P4D_MASKED_BITS)
 
 #ifndef __ASSEMBLY__
 
-static inline void pgd_clear(pgd_t *pgdp)
+static inline void p4d_clear(p4d_t *p4dp)
 {
-	*pgdp = __pgd(0);
+	*p4dp = __p4d(0);
 }
 
-static inline pte_t pgd_pte(pgd_t pgd)
+static inline pte_t p4d_pte(p4d_t p4d)
 {
-	return __pte(pgd_val(pgd));
+	return __pte(p4d_val(p4d));
 }
 
-static inline pgd_t pte_pgd(pte_t pte)
+static inline p4d_t pte_p4d(pte_t pte)
 {
-	return __pgd(pte_val(pte));
+	return __p4d(pte_val(pte));
 }
-extern struct page *pgd_page(pgd_t pgd);
+extern struct page *p4d_page(p4d_t p4d);
 
 #endif /* !__ASSEMBLY__ */
 
-#define pud_offset(pgdp, addr)	\
-  (((pud_t *) pgd_page_vaddr(*(pgdp))) + \
+#define pud_offset(p4dp, addr)	\
+  (((pud_t *) p4d_page_vaddr(*(p4dp))) + \
     (((addr) >> PUD_SHIFT) & (PTRS_PER_PUD - 1)))
 
 #define pud_ERROR(e) \

arch/powerpc/include/asm/nohash/64/pgtable.h

@@ -175,11 +175,11 @@ static inline pud_t pte_pud(pte_t pte)
 	return __pud(pte_val(pte));
 }
 #define pud_write(pud)		pte_write(pud_pte(pud))
-#define pgd_write(pgd)		pte_write(pgd_pte(pgd))
+#define p4d_write(pgd)		pte_write(p4d_pte(p4d))
 
-static inline void pgd_set(pgd_t *pgdp, unsigned long val)
+static inline void p4d_set(p4d_t *p4dp, unsigned long val)
 {
-	*pgdp = __pgd(val);
+	*p4dp = __p4d(val);
 }
 
 /*

arch/powerpc/include/asm/pgtable.h

@@ -44,12 +44,12 @@ struct mm_struct;
 #ifdef CONFIG_PPC32
 static inline pmd_t *pmd_ptr(struct mm_struct *mm, unsigned long va)
 {
-	return pmd_offset(pud_offset(pgd_offset(mm, va), va), va);
+	return pmd_offset(pud_offset(p4d_offset(pgd_offset(mm, va), va), va), va);
 }
 
 static inline pmd_t *pmd_ptr_k(unsigned long va)
 {
-	return pmd_offset(pud_offset(pgd_offset_k(va), va), va);
+	return pmd_offset(pud_offset(p4d_offset(pgd_offset_k(va), va), va), va);
 }
 
 static inline pte_t *virt_to_kpte(unsigned long vaddr)
@@ -158,9 +158,9 @@ static inline bool pud_is_leaf(pud_t pud)
 }
 #endif
 
-#ifndef pgd_is_leaf
-#define pgd_is_leaf pgd_is_leaf
-static inline bool pgd_is_leaf(pgd_t pgd)
+#ifndef p4d_is_leaf
+#define p4d_is_leaf p4d_is_leaf
+static inline bool p4d_is_leaf(p4d_t p4d)
 {
 	return false;
 }

arch/powerpc/kvm/book3s_64_mmu_radix.c

@@ -499,13 +499,14 @@ void kvmppc_free_pgtable_radix(struct kvm *kvm, pgd_t *pgd, unsigned int lpid)
 	unsigned long ig;
 
 	for (ig = 0; ig < PTRS_PER_PGD; ++ig, ++pgd) {
+		p4d_t *p4d = p4d_offset(pgd, 0);
 		pud_t *pud;
 
-		if (!pgd_present(*pgd))
+		if (!p4d_present(*p4d))
 			continue;
-		pud = pud_offset(pgd, 0);
+		pud = pud_offset(p4d, 0);
 		kvmppc_unmap_free_pud(kvm, pud, lpid);
-		pgd_clear(pgd);
+		p4d_clear(p4d);
 	}
 }
 
@@ -566,16 +567,19 @@ int kvmppc_create_pte(struct kvm *kvm, pgd_t *pgtable, pte_t pte,
 		      unsigned long *rmapp, struct rmap_nested **n_rmap)
 {
 	pgd_t *pgd;
+	p4d_t *p4d;
 	pud_t *pud, *new_pud = NULL;
 	pmd_t *pmd, *new_pmd = NULL;
 	pte_t *ptep, *new_ptep = NULL;
 	int ret;
 
 	/* Traverse the guest's 2nd-level tree, allocate new levels needed */
 	pgd = pgtable + pgd_index(gpa);
+	p4d = p4d_offset(pgd, gpa);
+
 	pud = NULL;
-	if (pgd_present(*pgd))
-		pud = pud_offset(pgd, gpa);
+	if (p4d_present(*p4d))
+		pud = pud_offset(p4d, gpa);
 	else
 		new_pud = pud_alloc_one(kvm->mm, gpa);
 
@@ -596,13 +600,13 @@ int kvmppc_create_pte(struct kvm *kvm, pgd_t *pgtable, pte_t pte,
 
 	/* Now traverse again under the lock and change the tree */
 	ret = -ENOMEM;
-	if (pgd_none(*pgd)) {
+	if (p4d_none(*p4d)) {
 		if (!new_pud)
 			goto out_unlock;
-		pgd_populate(kvm->mm, pgd, new_pud);
+		p4d_populate(kvm->mm, p4d, new_pud);
 		new_pud = NULL;
 	}
-	pud = pud_offset(pgd, gpa);
+	pud = pud_offset(p4d, gpa);
 	if (pud_is_leaf(*pud)) {
 		unsigned long hgpa = gpa & PUD_MASK;
 
@@ -1220,7 +1224,8 @@ static ssize_t debugfs_radix_read(struct file *file, char __user *buf,
 	unsigned long gpa;
 	pgd_t *pgt;
 	struct kvm_nested_guest *nested;
-	pgd_t pgd, *pgdp;
+	pgd_t *pgdp;
+	p4d_t p4d, *p4dp;
 	pud_t pud, *pudp;
 	pmd_t pmd, *pmdp;
 	pte_t *ptep;
@@ -1293,13 +1298,14 @@ static ssize_t debugfs_radix_read(struct file *file, char __user *buf,
 		}
 
 		pgdp = pgt + pgd_index(gpa);
-		pgd = READ_ONCE(*pgdp);
-		if (!(pgd_val(pgd) & _PAGE_PRESENT)) {
-			gpa = (gpa & PGDIR_MASK) + PGDIR_SIZE;
+		p4dp = p4d_offset(pgdp, gpa);
+		p4d = READ_ONCE(*p4dp);
+		if (!(p4d_val(p4d) & _PAGE_PRESENT)) {
+			gpa = (gpa & P4D_MASK) + P4D_SIZE;
 			continue;
 		}
 
-		pudp = pud_offset(&pgd, gpa);
+		pudp = pud_offset(&p4d, gpa);
 		pud = READ_ONCE(*pudp);
 		if (!(pud_val(pud) & _PAGE_PRESENT)) {
 			gpa = (gpa & PUD_MASK) + PUD_SIZE;