From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andrew Morton Subject: [patch 016/127] powerpc: add support for folded p4d page tables Date: Thu, 04 Jun 2020 16:46:44 -0700 Message-ID: <20200604234644.sGr62a5Gp%akpm@linux-foundation.org> References: <20200604164523.e15f3177f4b69dcb4f2534a1@linux-foundation.org> Reply-To: linux-kernel@vger.kernel.org Return-path: Received: from mail.kernel.org ([198.145.29.99]:42040 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725955AbgFDXqs (ORCPT ); Thu, 4 Jun 2020 19:46:48 -0400 In-Reply-To: <20200604164523.e15f3177f4b69dcb4f2534a1@linux-foundation.org> Sender: mm-commits-owner@vger.kernel.org List-Id: mm-commits@vger.kernel.org To: akpm@linux-foundation.org, arnd@arndb.de, bcain@codeaurora.org, benh@kernel.crashing.org, catalin.marinas@arm.com, christophe.leroy@c-s.fr, dalias@libc.org, fenghua.yu@intel.com, geert+renesas@glider.be, gxt@pku.edu.cn, james.morse@arm.com, jonas@southpole.se, julien.thierry.kdev@gmail.com, ley.foon.tan@intel.com, linux-mm@kvack.org, linux@armlinux.org.uk, maz@kernel.org, mm-commits@vger.kernel.org, mpe@ellerman.id.au, paulus@samba.org, rppt@linux.ibm.com, shorne@gmail.com, stefan.kristiansson@saunalahti.fi, suzuki.poulose@arm.com, tony.luck@intel.com, torvalds@linux-foundation.org, will@kernel.org, ysato@users.sourceforge.jp From: Mike Rapoport Subject: 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 Link: http://lkml.kernel.org/r/20200414153455.21744-9-rppt@kernel.org Signed-off-by: Mike Rapoport Tested-by: Christophe Leroy # 8xx and 83xx Cc: Arnd Bergmann Cc: Benjamin Herrenschmidt Cc: Brian Cain Cc: Catalin Marinas Cc: Fenghua Yu Cc: Geert Uytterhoeven Cc: Guan Xuetao Cc: James Morse Cc: Jonas Bonn Cc: Julien Thierry Cc: Ley Foon Tan Cc: Marc Zyngier Cc: Michael Ellerman Cc: Paul Mackerras Cc: Rich Felker Cc: Russell King Cc: Stafford Horne Cc: Stefan Kristiansson Cc: Suzuki K Poulose Cc: Tony Luck Cc: Will Deacon Cc: Yoshinori Sato Signed-off-by: Andrew Morton --- arch/powerpc/include/asm/book3s/32/pgtable.h | 1 arch/powerpc/include/asm/book3s/64/hash.h | 4 arch/powerpc/include/asm/book3s/64/pgalloc.h | 4 arch/powerpc/include/asm/book3s/64/pgtable.h | 60 +++++++------- arch/powerpc/include/asm/book3s/64/radix.h | 6 - arch/powerpc/include/asm/nohash/32/pgtable.h | 1 arch/powerpc/include/asm/nohash/64/pgalloc.h | 2 arch/powerpc/include/asm/nohash/64/pgtable-4k.h | 32 +++---- arch/powerpc/include/asm/nohash/64/pgtable.h | 6 - arch/powerpc/include/asm/pgtable.h | 10 +- arch/powerpc/kvm/book3s_64_mmu_radix.c | 32 ++++--- arch/powerpc/lib/code-patching.c | 7 + arch/powerpc/mm/book3s64/hash_pgtable.c | 4 arch/powerpc/mm/book3s64/radix_pgtable.c | 26 +++--- arch/powerpc/mm/book3s64/subpage_prot.c | 6 - arch/powerpc/mm/hugetlbpage.c | 28 +++--- arch/powerpc/mm/kasan/kasan_init_32.c | 2 arch/powerpc/mm/nohash/book3e_pgtable.c | 15 +-- arch/powerpc/mm/pgtable.c | 30 ++++--- arch/powerpc/mm/pgtable_64.c | 10 +- arch/powerpc/mm/ptdump/hashpagetable.c | 20 +++- arch/powerpc/mm/ptdump/ptdump.c | 12 +- arch/powerpc/xmon/xmon.c | 27 +++--- 23 files changed, 200 insertions(+), 145 deletions(-) --- a/arch/powerpc/include/asm/book3s/32/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 #include --- a/arch/powerpc/include/asm/book3s/64/hash.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/include/asm/book3s/64/hash.h @@ -134,9 +134,9 @@ static inline int get_region_id(unsigned #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); --- a/arch/powerpc/include/asm/book3s/64/pgalloc.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/include/asm/book3s/64/pgalloc.h @@ -85,9 +85,9 @@ static inline void pgd_free(struct mm_st 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) --- a/arch/powerpc/include/asm/book3s/64/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 +#include #ifndef __ASSEMBLY__ #include @@ -251,7 +251,7 @@ extern unsigned long __pmd_frag_size_shi /* 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( 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__ */ --- a/arch/powerpc/include/asm/book3s/64/radix.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 p } -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 --- a/arch/powerpc/include/asm/nohash/32/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 #ifndef __ASSEMBLY__ --- a/arch/powerpc/include/asm/nohash/64/pgalloc.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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) { --- a/arch/powerpc/include/asm/nohash/64/pgtable-4k.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 +#include /* * 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) \ --- a/arch/powerpc/include/asm/nohash/64/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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); } /* --- a/arch/powerpc/include/asm/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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; } --- a/arch/powerpc/kvm/book3s_64_mmu_radix.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/kvm/book3s_64_mmu_radix.c @@ -499,13 +499,14 @@ void kvmppc_free_pgtable_radix(struct kv 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); } } @@ -566,6 +567,7 @@ int kvmppc_create_pte(struct kvm *kvm, p 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; @@ -573,9 +575,11 @@ int kvmppc_create_pte(struct kvm *kvm, p /* 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, p /* 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 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 } 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; --- a/arch/powerpc/lib/code-patching.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/lib/code-patching.c @@ -107,13 +107,18 @@ static inline int unmap_patch_area(unsig pte_t *ptep; pmd_t *pmdp; pud_t *pudp; + p4d_t *p4dp; pgd_t *pgdp; pgdp = pgd_offset_k(addr); if (unlikely(!pgdp)) return -EINVAL; - pudp = pud_offset(pgdp, addr); + p4dp = p4d_offset(pgdp, addr); + if (unlikely(!p4dp)) + return -EINVAL; + + pudp = pud_offset(p4dp, addr); if (unlikely(!pudp)) return -EINVAL; --- a/arch/powerpc/mm/book3s64/hash_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/hash_pgtable.c @@ -148,6 +148,7 @@ void hash__vmemmap_remove_mapping(unsign int hash__map_kernel_page(unsigned long ea, unsigned long pa, pgprot_t prot) { pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -155,7 +156,8 @@ int hash__map_kernel_page(unsigned long BUILD_BUG_ON(TASK_SIZE_USER64 > H_PGTABLE_RANGE); if (slab_is_available()) { pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; pmdp = pmd_alloc(&init_mm, pudp, ea); --- a/arch/powerpc/mm/book3s64/radix_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/radix_pgtable.c @@ -65,17 +65,19 @@ static int early_map_kernel_page(unsigne { unsigned long pfn = pa >> PAGE_SHIFT; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; pgdp = pgd_offset_k(ea); - if (pgd_none(*pgdp)) { + p4dp = p4d_offset(pgdp, ea); + if (p4d_none(*p4dp)) { pudp = early_alloc_pgtable(PUD_TABLE_SIZE, nid, region_start, region_end); - pgd_populate(&init_mm, pgdp, pudp); + p4d_populate(&init_mm, p4dp, pudp); } - pudp = pud_offset(pgdp, ea); + pudp = pud_offset(p4dp, ea); if (map_page_size == PUD_SIZE) { ptep = (pte_t *)pudp; goto set_the_pte; @@ -115,6 +117,7 @@ static int __map_kernel_page(unsigned lo { unsigned long pfn = pa >> PAGE_SHIFT; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -137,7 +140,8 @@ static int __map_kernel_page(unsigned lo * boot. */ pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; if (map_page_size == PUD_SIZE) { @@ -174,6 +178,7 @@ void radix__change_memory_range(unsigned { unsigned long idx; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -186,7 +191,8 @@ void radix__change_memory_range(unsigned for (idx = start; idx < end; idx += PAGE_SIZE) { pgdp = pgd_offset_k(idx); - pudp = pud_alloc(&init_mm, pgdp, idx); + p4dp = p4d_offset(pgdp, idx); + pudp = pud_alloc(&init_mm, p4dp, idx); if (!pudp) continue; if (pud_is_leaf(*pudp)) { @@ -850,6 +856,7 @@ static void __meminit remove_pagetable(u unsigned long addr, next; pud_t *pud_base; pgd_t *pgd; + p4d_t *p4d; spin_lock(&init_mm.page_table_lock); @@ -857,15 +864,16 @@ static void __meminit remove_pagetable(u next = pgd_addr_end(addr, end); pgd = pgd_offset_k(addr); - if (!pgd_present(*pgd)) + p4d = p4d_offset(pgd, addr); + if (!p4d_present(*p4d)) continue; - if (pgd_is_leaf(*pgd)) { - split_kernel_mapping(addr, end, PGDIR_SIZE, (pte_t *)pgd); + if (p4d_is_leaf(*p4d)) { + split_kernel_mapping(addr, end, P4D_SIZE, (pte_t *)p4d); continue; } - pud_base = (pud_t *)pgd_page_vaddr(*pgd); + pud_base = (pud_t *)p4d_page_vaddr(*p4d); remove_pud_table(pud_base, addr, next); } --- a/arch/powerpc/mm/book3s64/subpage_prot.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/subpage_prot.c @@ -54,15 +54,17 @@ static void hpte_flush_range(struct mm_s int npages) { pgd_t *pgd; + p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *pte; spinlock_t *ptl; pgd = pgd_offset(mm, addr); - if (pgd_none(*pgd)) + p4d = p4d_offset(pgd, addr); + if (p4d_none(*p4d)) return; - pud = pud_offset(pgd, addr); + pud = pud_offset(p4d, addr); if (pud_none(*pud)) return; pmd = pmd_offset(pud, addr); --- a/arch/powerpc/mm/hugetlbpage.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/hugetlbpage.c @@ -119,6 +119,7 @@ static int __hugepte_alloc(struct mm_str pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr, unsigned long sz) { pgd_t *pg; + p4d_t *p4; pud_t *pu; pmd_t *pm; hugepd_t *hpdp = NULL; @@ -128,20 +129,21 @@ pte_t *huge_pte_alloc(struct mm_struct * addr &= ~(sz-1); pg = pgd_offset(mm, addr); + p4 = p4d_offset(pg, addr); #ifdef CONFIG_PPC_BOOK3S_64 if (pshift == PGDIR_SHIFT) /* 16GB huge page */ - return (pte_t *) pg; + return (pte_t *) p4; else if (pshift > PUD_SHIFT) { /* * We need to use hugepd table */ ptl = &mm->page_table_lock; - hpdp = (hugepd_t *)pg; + hpdp = (hugepd_t *)p4; } else { pdshift = PUD_SHIFT; - pu = pud_alloc(mm, pg, addr); + pu = pud_alloc(mm, p4, addr); if (!pu) return NULL; if (pshift == PUD_SHIFT) @@ -166,10 +168,10 @@ pte_t *huge_pte_alloc(struct mm_struct * #else if (pshift >= PGDIR_SHIFT) { ptl = &mm->page_table_lock; - hpdp = (hugepd_t *)pg; + hpdp = (hugepd_t *)p4; } else { pdshift = PUD_SHIFT; - pu = pud_alloc(mm, pg, addr); + pu = pud_alloc(mm, p4, addr); if (!pu) return NULL; if (pshift >= PUD_SHIFT) { @@ -390,7 +392,7 @@ static void hugetlb_free_pmd_range(struc mm_dec_nr_pmds(tlb->mm); } -static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, +static void hugetlb_free_pud_range(struct mmu_gather *tlb, p4d_t *p4d, unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) { @@ -400,7 +402,7 @@ static void hugetlb_free_pud_range(struc start = addr; do { - pud = pud_offset(pgd, addr); + pud = pud_offset(p4d, addr); next = pud_addr_end(addr, end); if (!is_hugepd(__hugepd(pud_val(*pud)))) { if (pud_none_or_clear_bad(pud)) @@ -435,8 +437,8 @@ static void hugetlb_free_pud_range(struc if (end - 1 > ceiling - 1) return; - pud = pud_offset(pgd, start); - pgd_clear(pgd); + pud = pud_offset(p4d, start); + p4d_clear(p4d); pud_free_tlb(tlb, pud, start); mm_dec_nr_puds(tlb->mm); } @@ -449,6 +451,7 @@ void hugetlb_free_pgd_range(struct mmu_g unsigned long floor, unsigned long ceiling) { pgd_t *pgd; + p4d_t *p4d; unsigned long next; /* @@ -471,10 +474,11 @@ void hugetlb_free_pgd_range(struct mmu_g do { next = pgd_addr_end(addr, end); pgd = pgd_offset(tlb->mm, addr); + p4d = p4d_offset(pgd, addr); if (!is_hugepd(__hugepd(pgd_val(*pgd)))) { - if (pgd_none_or_clear_bad(pgd)) + if (p4d_none_or_clear_bad(p4d)) continue; - hugetlb_free_pud_range(tlb, pgd, addr, next, floor, ceiling); + hugetlb_free_pud_range(tlb, p4d, addr, next, floor, ceiling); } else { unsigned long more; /* @@ -487,7 +491,7 @@ void hugetlb_free_pgd_range(struct mmu_g if (more > next) next = more; - free_hugepd_range(tlb, (hugepd_t *)pgd, PGDIR_SHIFT, + free_hugepd_range(tlb, (hugepd_t *)p4d, PGDIR_SHIFT, addr, next, floor, ceiling); } } while (addr = next, addr != end); --- a/arch/powerpc/mm/kasan/kasan_init_32.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/kasan/kasan_init_32.c @@ -121,7 +121,7 @@ static void __init kasan_unmap_early_sha phys_addr_t pa = __pa(kasan_early_shadow_page); for (k_cur = k_start & PAGE_MASK; k_cur < k_end; k_cur += PAGE_SIZE) { - pmd_t *pmd = pmd_offset(pud_offset(pgd_offset_k(k_cur), k_cur), k_cur); + pmd_t *pmd = pmd_ptr_k(k_cur); pte_t *ptep = pte_offset_kernel(pmd, k_cur); if ((pte_val(*ptep) & PTE_RPN_MASK) != pa) --- a/arch/powerpc/mm/nohash/book3e_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/nohash/book3e_pgtable.c @@ -73,6 +73,7 @@ static void __init *early_alloc_pgtable( int __ref map_kernel_page(unsigned long ea, unsigned long pa, pgprot_t prot) { pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -80,7 +81,8 @@ int __ref map_kernel_page(unsigned long BUILD_BUG_ON(TASK_SIZE_USER64 > PGTABLE_RANGE); if (slab_is_available()) { pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; pmdp = pmd_alloc(&init_mm, pudp, ea); @@ -91,13 +93,12 @@ int __ref map_kernel_page(unsigned long return -ENOMEM; } else { pgdp = pgd_offset_k(ea); -#ifndef __PAGETABLE_PUD_FOLDED - if (pgd_none(*pgdp)) { - pudp = early_alloc_pgtable(PUD_TABLE_SIZE); - pgd_populate(&init_mm, pgdp, pudp); + p4dp = p4d_offset(pgdp, ea); + if (p4d_none(*p4dp)) { + pmdp = early_alloc_pgtable(PMD_TABLE_SIZE); + p4d_populate(&init_mm, p4dp, pmdp); } -#endif /* !__PAGETABLE_PUD_FOLDED */ - pudp = pud_offset(pgdp, ea); + pudp = pud_offset(p4dp, ea); if (pud_none(*pudp)) { pmdp = early_alloc_pgtable(PMD_TABLE_SIZE); pud_populate(&init_mm, pudp, pmdp); --- a/arch/powerpc/mm/pgtable_64.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/pgtable_64.c @@ -101,13 +101,13 @@ EXPORT_SYMBOL(__pte_frag_size_shift); #ifndef __PAGETABLE_PUD_FOLDED /* 4 level page table */ -struct page *pgd_page(pgd_t pgd) +struct page *p4d_page(p4d_t p4d) { - if (pgd_is_leaf(pgd)) { - VM_WARN_ON(!pgd_huge(pgd)); - return pte_page(pgd_pte(pgd)); + if (p4d_is_leaf(p4d)) { + VM_WARN_ON(!p4d_huge(p4d)); + return pte_page(p4d_pte(p4d)); } - return virt_to_page(pgd_page_vaddr(pgd)); + return virt_to_page(p4d_page_vaddr(p4d)); } #endif --- a/arch/powerpc/mm/pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/pgtable.c @@ -265,6 +265,7 @@ int huge_ptep_set_access_flags(struct vm void assert_pte_locked(struct mm_struct *mm, unsigned long addr) { pgd_t *pgd; + p4d_t *p4d; pud_t *pud; pmd_t *pmd; @@ -272,7 +273,9 @@ void assert_pte_locked(struct mm_struct return; pgd = mm->pgd + pgd_index(addr); BUG_ON(pgd_none(*pgd)); - pud = pud_offset(pgd, addr); + p4d = p4d_offset(pgd, addr); + BUG_ON(p4d_none(*p4d)); + pud = pud_offset(p4d, addr); BUG_ON(pud_none(*pud)); pmd = pmd_offset(pud, addr); /* @@ -312,12 +315,13 @@ EXPORT_SYMBOL_GPL(vmalloc_to_phys); pte_t *__find_linux_pte(pgd_t *pgdir, unsigned long ea, bool *is_thp, unsigned *hpage_shift) { - pgd_t pgd, *pgdp; + pgd_t *pgdp; + p4d_t p4d, *p4dp; pud_t pud, *pudp; pmd_t pmd, *pmdp; pte_t *ret_pte; hugepd_t *hpdp = NULL; - unsigned pdshift = PGDIR_SHIFT; + unsigned pdshift; if (hpage_shift) *hpage_shift = 0; @@ -325,24 +329,28 @@ pte_t *__find_linux_pte(pgd_t *pgdir, un if (is_thp) *is_thp = false; - pgdp = pgdir + pgd_index(ea); - pgd = READ_ONCE(*pgdp); /* * Always operate on the local stack value. This make sure the * value don't get updated by a parallel THP split/collapse, * page fault or a page unmap. The return pte_t * is still not * stable. So should be checked there for above conditions. + * Top level is an exception because it is folded into p4d. */ - if (pgd_none(pgd)) + pgdp = pgdir + pgd_index(ea); + p4dp = p4d_offset(pgdp, ea); + p4d = READ_ONCE(*p4dp); + pdshift = P4D_SHIFT; + + if (p4d_none(p4d)) return NULL; - if (pgd_is_leaf(pgd)) { - ret_pte = (pte_t *)pgdp; + if (p4d_is_leaf(p4d)) { + ret_pte = (pte_t *)p4dp; goto out; } - if (is_hugepd(__hugepd(pgd_val(pgd)))) { - hpdp = (hugepd_t *)&pgd; + if (is_hugepd(__hugepd(p4d_val(p4d)))) { + hpdp = (hugepd_t *)&p4d; goto out_huge; } @@ -352,7 +360,7 @@ pte_t *__find_linux_pte(pgd_t *pgdir, un * irq disabled */ pdshift = PUD_SHIFT; - pudp = pud_offset(&pgd, ea); + pudp = pud_offset(&p4d, ea); pud = READ_ONCE(*pudp); if (pud_none(pud)) --- a/arch/powerpc/mm/ptdump/hashpagetable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/ptdump/hashpagetable.c @@ -417,9 +417,9 @@ static void walk_pmd(struct pg_state *st } } -static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) +static void walk_pud(struct pg_state *st, p4d_t *p4d, unsigned long start) { - pud_t *pud = pud_offset(pgd, 0); + pud_t *pud = pud_offset(p4d, 0); unsigned long addr; unsigned int i; @@ -431,6 +431,20 @@ static void walk_pud(struct pg_state *st } } +static void walk_p4d(struct pg_state *st, pgd_t *pgd, unsigned long start) +{ + p4d_t *p4d = p4d_offset(pgd, 0); + unsigned long addr; + unsigned int i; + + for (i = 0; i < PTRS_PER_P4D; i++, p4d++) { + addr = start + i * P4D_SIZE; + if (!p4d_none(*p4d)) + /* p4d exists */ + walk_pud(st, p4d, addr); + } +} + static void walk_pagetables(struct pg_state *st) { pgd_t *pgd = pgd_offset_k(0UL); @@ -445,7 +459,7 @@ static void walk_pagetables(struct pg_st addr = KERN_VIRT_START + i * PGDIR_SIZE; if (!pgd_none(*pgd)) /* pgd exists */ - walk_pud(st, pgd, addr); + walk_p4d(st, pgd, addr); } } --- a/arch/powerpc/mm/ptdump/ptdump.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/ptdump/ptdump.c @@ -277,9 +277,9 @@ static void walk_pmd(struct pg_state *st } } -static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) +static void walk_pud(struct pg_state *st, p4d_t *p4d, unsigned long start) { - pud_t *pud = pud_offset(pgd, 0); + pud_t *pud = pud_offset(p4d, 0); unsigned long addr; unsigned int i; @@ -304,11 +304,13 @@ static void walk_pagetables(struct pg_st * the hash pagetable. */ for (i = pgd_index(addr); i < PTRS_PER_PGD; i++, pgd++, addr += PGDIR_SIZE) { - if (!pgd_none(*pgd) && !pgd_is_leaf(*pgd)) + p4d_t *p4d = p4d_offset(pgd, 0); + + if (!p4d_none(*p4d) && !p4d_is_leaf(*p4d)) /* pgd exists */ - walk_pud(st, pgd, addr); + walk_pud(st, p4d, addr); else - note_page(st, addr, 1, pgd_val(*pgd)); + note_page(st, addr, 1, p4d_val(*p4d)); } } --- a/arch/powerpc/xmon/xmon.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/xmon/xmon.c @@ -3135,7 +3135,8 @@ static void show_pte(unsigned long addr) unsigned long tskv = 0; struct task_struct *tsk = NULL; struct mm_struct *mm; - pgd_t *pgdp, *pgdir; + pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -3159,28 +3160,26 @@ static void show_pte(unsigned long addr) catch_memory_errors = 1; sync(); - if (mm == &init_mm) { + if (mm == &init_mm) pgdp = pgd_offset_k(addr); - pgdir = pgd_offset_k(0); - } else { + else pgdp = pgd_offset(mm, addr); - pgdir = pgd_offset(mm, 0); - } - if (pgd_none(*pgdp)) { - printf("no linux page table for address\n"); + p4dp = p4d_offset(pgdp, addr); + + if (p4d_none(*p4dp)) { + printf("No valid P4D\n"); return; } - printf("pgd @ 0x%px\n", pgdir); From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-3.8 required=3.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED, DKIM_VALID,HEADER_FROM_DIFFERENT_DOMAINS,MAILING_LIST_MULTI,SIGNED_OFF_BY, SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=no autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id 71CF7C433DF for ; Thu, 4 Jun 2020 23:46:48 +0000 (UTC) Received: from kanga.kvack.org (kanga.kvack.org [205.233.56.17]) by mail.kernel.org (Postfix) with ESMTP id F3741207D0 for ; Thu, 4 Jun 2020 23:46:47 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=kernel.org header.i=@kernel.org header.b="Uzh3w3aU" DMARC-Filter: OpenDMARC Filter v1.3.2 mail.kernel.org F3741207D0 Authentication-Results: mail.kernel.org; dmarc=none (p=none dis=none) header.from=linux-foundation.org Authentication-Results: mail.kernel.org; spf=pass smtp.mailfrom=owner-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix) id 9B89128000E; Thu, 4 Jun 2020 19:46:47 -0400 (EDT) Received: by kanga.kvack.org (Postfix, from userid 40) id 96848280002; Thu, 4 Jun 2020 19:46:47 -0400 (EDT) X-Delivered-To: int-list-linux-mm@kvack.org Received: by kanga.kvack.org (Postfix, from userid 63042) id 808AB28000E; Thu, 4 Jun 2020 19:46:47 -0400 (EDT) X-Delivered-To: linux-mm@kvack.org Received: from forelay.hostedemail.com (smtprelay0235.hostedemail.com [216.40.44.235]) by kanga.kvack.org (Postfix) with ESMTP id 63783280002 for ; Thu, 4 Jun 2020 19:46:47 -0400 (EDT) Received: from smtpin28.hostedemail.com (10.5.19.251.rfc1918.com [10.5.19.251]) by forelay02.hostedemail.com (Postfix) with ESMTP id 2D6343752 for ; Thu, 4 Jun 2020 23:46:47 +0000 (UTC) X-FDA: 76893166854.28.crack54_0e1122a26d9b Received: from filter.hostedemail.com (10.5.16.251.rfc1918.com [10.5.16.251]) by smtpin28.hostedemail.com (Postfix) with ESMTP id 102906C33 for ; Thu, 4 Jun 2020 23:46:47 +0000 (UTC) X-HE-Tag: crack54_0e1122a26d9b X-Filterd-Recvd-Size: 34412 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by imf01.hostedemail.com (Postfix) with ESMTP for ; Thu, 4 Jun 2020 23:46:46 +0000 (UTC) Received: from localhost.localdomain (c-73-231-172-41.hsd1.ca.comcast.net [73.231.172.41]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 81897206A2; Thu, 4 Jun 2020 23:46:44 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1591314405; bh=TnGwqi7dtJG2hs7E0OB8+e111XWIQVLsOEGO6maLyVI=; h=Date:From:To:Subject:In-Reply-To:From; b=Uzh3w3aUjMa5DHfDZ39/hKfzPWszSfAKQUsGM+34Q1nMibDHpPsW5bT+14HZHrYmS SP1rRgyJISwCwPtsQ6GmY6NvKZQoY2KGgGk95biUDLELz33i0QvpMWY0/a9WYs0dBm nHg1JCbmjzojWD5kxPwAsTdVNSkArnU2J9LlMgsk= Date: Thu, 04 Jun 2020 16:46:44 -0700 From: Andrew Morton To: akpm@linux-foundation.org, arnd@arndb.de, bcain@codeaurora.org, benh@kernel.crashing.org, catalin.marinas@arm.com, christophe.leroy@c-s.fr, dalias@libc.org, fenghua.yu@intel.com, geert+renesas@glider.be, gxt@pku.edu.cn, james.morse@arm.com, jonas@southpole.se, julien.thierry.kdev@gmail.com, ley.foon.tan@intel.com, linux-mm@kvack.org, linux@armlinux.org.uk, maz@kernel.org, mm-commits@vger.kernel.org, mpe@ellerman.id.au, paulus@samba.org, rppt@linux.ibm.com, shorne@gmail.com, stefan.kristiansson@saunalahti.fi, suzuki.poulose@arm.com, tony.luck@intel.com, torvalds@linux-foundation.org, will@kernel.org, ysato@users.sourceforge.jp Subject: [patch 016/127] powerpc: add support for folded p4d page tables Message-ID: <20200604234644.sGr62a5Gp%akpm@linux-foundation.org> In-Reply-To: <20200604164523.e15f3177f4b69dcb4f2534a1@linux-foundation.org> User-Agent: s-nail v14.8.16 X-Rspamd-Queue-Id: 102906C33 X-Spamd-Result: default: False [0.00 / 100.00] X-Rspamd-Server: rspam02 X-Bogosity: Ham, tests=bogofilter, spamicity=0.000000, version=1.2.4 Sender: owner-linux-mm@kvack.org Precedence: bulk X-Loop: owner-majordomo@kvack.org List-ID: From: Mike Rapoport Subject: 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 Link: http://lkml.kernel.org/r/20200414153455.21744-9-rppt@kernel.org Signed-off-by: Mike Rapoport Tested-by: Christophe Leroy # 8xx and 83xx Cc: Arnd Bergmann Cc: Benjamin Herrenschmidt Cc: Brian Cain Cc: Catalin Marinas Cc: Fenghua Yu Cc: Geert Uytterhoeven Cc: Guan Xuetao Cc: James Morse Cc: Jonas Bonn Cc: Julien Thierry Cc: Ley Foon Tan Cc: Marc Zyngier Cc: Michael Ellerman Cc: Paul Mackerras Cc: Rich Felker Cc: Russell King Cc: Stafford Horne Cc: Stefan Kristiansson Cc: Suzuki K Poulose Cc: Tony Luck Cc: Will Deacon Cc: Yoshinori Sato Signed-off-by: Andrew Morton --- arch/powerpc/include/asm/book3s/32/pgtable.h | 1 arch/powerpc/include/asm/book3s/64/hash.h | 4 arch/powerpc/include/asm/book3s/64/pgalloc.h | 4 arch/powerpc/include/asm/book3s/64/pgtable.h | 60 +++++++------- arch/powerpc/include/asm/book3s/64/radix.h | 6 - arch/powerpc/include/asm/nohash/32/pgtable.h | 1 arch/powerpc/include/asm/nohash/64/pgalloc.h | 2 arch/powerpc/include/asm/nohash/64/pgtable-4k.h | 32 +++---- arch/powerpc/include/asm/nohash/64/pgtable.h | 6 - arch/powerpc/include/asm/pgtable.h | 10 +- arch/powerpc/kvm/book3s_64_mmu_radix.c | 32 ++++--- arch/powerpc/lib/code-patching.c | 7 + arch/powerpc/mm/book3s64/hash_pgtable.c | 4 arch/powerpc/mm/book3s64/radix_pgtable.c | 26 +++--- arch/powerpc/mm/book3s64/subpage_prot.c | 6 - arch/powerpc/mm/hugetlbpage.c | 28 +++--- arch/powerpc/mm/kasan/kasan_init_32.c | 2 arch/powerpc/mm/nohash/book3e_pgtable.c | 15 +-- arch/powerpc/mm/pgtable.c | 30 ++++--- arch/powerpc/mm/pgtable_64.c | 10 +- arch/powerpc/mm/ptdump/hashpagetable.c | 20 +++- arch/powerpc/mm/ptdump/ptdump.c | 12 +- arch/powerpc/xmon/xmon.c | 27 +++--- 23 files changed, 200 insertions(+), 145 deletions(-) --- a/arch/powerpc/include/asm/book3s/32/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 #include --- a/arch/powerpc/include/asm/book3s/64/hash.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/include/asm/book3s/64/hash.h @@ -134,9 +134,9 @@ static inline int get_region_id(unsigned #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); --- a/arch/powerpc/include/asm/book3s/64/pgalloc.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/include/asm/book3s/64/pgalloc.h @@ -85,9 +85,9 @@ static inline void pgd_free(struct mm_st 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) --- a/arch/powerpc/include/asm/book3s/64/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 +#include #ifndef __ASSEMBLY__ #include @@ -251,7 +251,7 @@ extern unsigned long __pmd_frag_size_shi /* 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( 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__ */ --- a/arch/powerpc/include/asm/book3s/64/radix.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 p } -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 --- a/arch/powerpc/include/asm/nohash/32/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 #ifndef __ASSEMBLY__ --- a/arch/powerpc/include/asm/nohash/64/pgalloc.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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) { --- a/arch/powerpc/include/asm/nohash/64/pgtable-4k.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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 +#include /* * 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) \ --- a/arch/powerpc/include/asm/nohash/64/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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); } /* --- a/arch/powerpc/include/asm/pgtable.h~powerpc-add-support-for-folded-p4d-page-tables +++ a/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; } --- a/arch/powerpc/kvm/book3s_64_mmu_radix.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/kvm/book3s_64_mmu_radix.c @@ -499,13 +499,14 @@ void kvmppc_free_pgtable_radix(struct kv 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); } } @@ -566,6 +567,7 @@ int kvmppc_create_pte(struct kvm *kvm, p 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; @@ -573,9 +575,11 @@ int kvmppc_create_pte(struct kvm *kvm, p /* 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, p /* 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 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 } 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; --- a/arch/powerpc/lib/code-patching.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/lib/code-patching.c @@ -107,13 +107,18 @@ static inline int unmap_patch_area(unsig pte_t *ptep; pmd_t *pmdp; pud_t *pudp; + p4d_t *p4dp; pgd_t *pgdp; pgdp = pgd_offset_k(addr); if (unlikely(!pgdp)) return -EINVAL; - pudp = pud_offset(pgdp, addr); + p4dp = p4d_offset(pgdp, addr); + if (unlikely(!p4dp)) + return -EINVAL; + + pudp = pud_offset(p4dp, addr); if (unlikely(!pudp)) return -EINVAL; --- a/arch/powerpc/mm/book3s64/hash_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/hash_pgtable.c @@ -148,6 +148,7 @@ void hash__vmemmap_remove_mapping(unsign int hash__map_kernel_page(unsigned long ea, unsigned long pa, pgprot_t prot) { pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -155,7 +156,8 @@ int hash__map_kernel_page(unsigned long BUILD_BUG_ON(TASK_SIZE_USER64 > H_PGTABLE_RANGE); if (slab_is_available()) { pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; pmdp = pmd_alloc(&init_mm, pudp, ea); --- a/arch/powerpc/mm/book3s64/radix_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/radix_pgtable.c @@ -65,17 +65,19 @@ static int early_map_kernel_page(unsigne { unsigned long pfn = pa >> PAGE_SHIFT; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; pgdp = pgd_offset_k(ea); - if (pgd_none(*pgdp)) { + p4dp = p4d_offset(pgdp, ea); + if (p4d_none(*p4dp)) { pudp = early_alloc_pgtable(PUD_TABLE_SIZE, nid, region_start, region_end); - pgd_populate(&init_mm, pgdp, pudp); + p4d_populate(&init_mm, p4dp, pudp); } - pudp = pud_offset(pgdp, ea); + pudp = pud_offset(p4dp, ea); if (map_page_size == PUD_SIZE) { ptep = (pte_t *)pudp; goto set_the_pte; @@ -115,6 +117,7 @@ static int __map_kernel_page(unsigned lo { unsigned long pfn = pa >> PAGE_SHIFT; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -137,7 +140,8 @@ static int __map_kernel_page(unsigned lo * boot. */ pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; if (map_page_size == PUD_SIZE) { @@ -174,6 +178,7 @@ void radix__change_memory_range(unsigned { unsigned long idx; pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -186,7 +191,8 @@ void radix__change_memory_range(unsigned for (idx = start; idx < end; idx += PAGE_SIZE) { pgdp = pgd_offset_k(idx); - pudp = pud_alloc(&init_mm, pgdp, idx); + p4dp = p4d_offset(pgdp, idx); + pudp = pud_alloc(&init_mm, p4dp, idx); if (!pudp) continue; if (pud_is_leaf(*pudp)) { @@ -850,6 +856,7 @@ static void __meminit remove_pagetable(u unsigned long addr, next; pud_t *pud_base; pgd_t *pgd; + p4d_t *p4d; spin_lock(&init_mm.page_table_lock); @@ -857,15 +864,16 @@ static void __meminit remove_pagetable(u next = pgd_addr_end(addr, end); pgd = pgd_offset_k(addr); - if (!pgd_present(*pgd)) + p4d = p4d_offset(pgd, addr); + if (!p4d_present(*p4d)) continue; - if (pgd_is_leaf(*pgd)) { - split_kernel_mapping(addr, end, PGDIR_SIZE, (pte_t *)pgd); + if (p4d_is_leaf(*p4d)) { + split_kernel_mapping(addr, end, P4D_SIZE, (pte_t *)p4d); continue; } - pud_base = (pud_t *)pgd_page_vaddr(*pgd); + pud_base = (pud_t *)p4d_page_vaddr(*p4d); remove_pud_table(pud_base, addr, next); } --- a/arch/powerpc/mm/book3s64/subpage_prot.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/book3s64/subpage_prot.c @@ -54,15 +54,17 @@ static void hpte_flush_range(struct mm_s int npages) { pgd_t *pgd; + p4d_t *p4d; pud_t *pud; pmd_t *pmd; pte_t *pte; spinlock_t *ptl; pgd = pgd_offset(mm, addr); - if (pgd_none(*pgd)) + p4d = p4d_offset(pgd, addr); + if (p4d_none(*p4d)) return; - pud = pud_offset(pgd, addr); + pud = pud_offset(p4d, addr); if (pud_none(*pud)) return; pmd = pmd_offset(pud, addr); --- a/arch/powerpc/mm/hugetlbpage.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/hugetlbpage.c @@ -119,6 +119,7 @@ static int __hugepte_alloc(struct mm_str pte_t *huge_pte_alloc(struct mm_struct *mm, unsigned long addr, unsigned long sz) { pgd_t *pg; + p4d_t *p4; pud_t *pu; pmd_t *pm; hugepd_t *hpdp = NULL; @@ -128,20 +129,21 @@ pte_t *huge_pte_alloc(struct mm_struct * addr &= ~(sz-1); pg = pgd_offset(mm, addr); + p4 = p4d_offset(pg, addr); #ifdef CONFIG_PPC_BOOK3S_64 if (pshift == PGDIR_SHIFT) /* 16GB huge page */ - return (pte_t *) pg; + return (pte_t *) p4; else if (pshift > PUD_SHIFT) { /* * We need to use hugepd table */ ptl = &mm->page_table_lock; - hpdp = (hugepd_t *)pg; + hpdp = (hugepd_t *)p4; } else { pdshift = PUD_SHIFT; - pu = pud_alloc(mm, pg, addr); + pu = pud_alloc(mm, p4, addr); if (!pu) return NULL; if (pshift == PUD_SHIFT) @@ -166,10 +168,10 @@ pte_t *huge_pte_alloc(struct mm_struct * #else if (pshift >= PGDIR_SHIFT) { ptl = &mm->page_table_lock; - hpdp = (hugepd_t *)pg; + hpdp = (hugepd_t *)p4; } else { pdshift = PUD_SHIFT; - pu = pud_alloc(mm, pg, addr); + pu = pud_alloc(mm, p4, addr); if (!pu) return NULL; if (pshift >= PUD_SHIFT) { @@ -390,7 +392,7 @@ static void hugetlb_free_pmd_range(struc mm_dec_nr_pmds(tlb->mm); } -static void hugetlb_free_pud_range(struct mmu_gather *tlb, pgd_t *pgd, +static void hugetlb_free_pud_range(struct mmu_gather *tlb, p4d_t *p4d, unsigned long addr, unsigned long end, unsigned long floor, unsigned long ceiling) { @@ -400,7 +402,7 @@ static void hugetlb_free_pud_range(struc start = addr; do { - pud = pud_offset(pgd, addr); + pud = pud_offset(p4d, addr); next = pud_addr_end(addr, end); if (!is_hugepd(__hugepd(pud_val(*pud)))) { if (pud_none_or_clear_bad(pud)) @@ -435,8 +437,8 @@ static void hugetlb_free_pud_range(struc if (end - 1 > ceiling - 1) return; - pud = pud_offset(pgd, start); - pgd_clear(pgd); + pud = pud_offset(p4d, start); + p4d_clear(p4d); pud_free_tlb(tlb, pud, start); mm_dec_nr_puds(tlb->mm); } @@ -449,6 +451,7 @@ void hugetlb_free_pgd_range(struct mmu_g unsigned long floor, unsigned long ceiling) { pgd_t *pgd; + p4d_t *p4d; unsigned long next; /* @@ -471,10 +474,11 @@ void hugetlb_free_pgd_range(struct mmu_g do { next = pgd_addr_end(addr, end); pgd = pgd_offset(tlb->mm, addr); + p4d = p4d_offset(pgd, addr); if (!is_hugepd(__hugepd(pgd_val(*pgd)))) { - if (pgd_none_or_clear_bad(pgd)) + if (p4d_none_or_clear_bad(p4d)) continue; - hugetlb_free_pud_range(tlb, pgd, addr, next, floor, ceiling); + hugetlb_free_pud_range(tlb, p4d, addr, next, floor, ceiling); } else { unsigned long more; /* @@ -487,7 +491,7 @@ void hugetlb_free_pgd_range(struct mmu_g if (more > next) next = more; - free_hugepd_range(tlb, (hugepd_t *)pgd, PGDIR_SHIFT, + free_hugepd_range(tlb, (hugepd_t *)p4d, PGDIR_SHIFT, addr, next, floor, ceiling); } } while (addr = next, addr != end); --- a/arch/powerpc/mm/kasan/kasan_init_32.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/kasan/kasan_init_32.c @@ -121,7 +121,7 @@ static void __init kasan_unmap_early_sha phys_addr_t pa = __pa(kasan_early_shadow_page); for (k_cur = k_start & PAGE_MASK; k_cur < k_end; k_cur += PAGE_SIZE) { - pmd_t *pmd = pmd_offset(pud_offset(pgd_offset_k(k_cur), k_cur), k_cur); + pmd_t *pmd = pmd_ptr_k(k_cur); pte_t *ptep = pte_offset_kernel(pmd, k_cur); if ((pte_val(*ptep) & PTE_RPN_MASK) != pa) --- a/arch/powerpc/mm/nohash/book3e_pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/nohash/book3e_pgtable.c @@ -73,6 +73,7 @@ static void __init *early_alloc_pgtable( int __ref map_kernel_page(unsigned long ea, unsigned long pa, pgprot_t prot) { pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -80,7 +81,8 @@ int __ref map_kernel_page(unsigned long BUILD_BUG_ON(TASK_SIZE_USER64 > PGTABLE_RANGE); if (slab_is_available()) { pgdp = pgd_offset_k(ea); - pudp = pud_alloc(&init_mm, pgdp, ea); + p4dp = p4d_offset(pgdp, ea); + pudp = pud_alloc(&init_mm, p4dp, ea); if (!pudp) return -ENOMEM; pmdp = pmd_alloc(&init_mm, pudp, ea); @@ -91,13 +93,12 @@ int __ref map_kernel_page(unsigned long return -ENOMEM; } else { pgdp = pgd_offset_k(ea); -#ifndef __PAGETABLE_PUD_FOLDED - if (pgd_none(*pgdp)) { - pudp = early_alloc_pgtable(PUD_TABLE_SIZE); - pgd_populate(&init_mm, pgdp, pudp); + p4dp = p4d_offset(pgdp, ea); + if (p4d_none(*p4dp)) { + pmdp = early_alloc_pgtable(PMD_TABLE_SIZE); + p4d_populate(&init_mm, p4dp, pmdp); } -#endif /* !__PAGETABLE_PUD_FOLDED */ - pudp = pud_offset(pgdp, ea); + pudp = pud_offset(p4dp, ea); if (pud_none(*pudp)) { pmdp = early_alloc_pgtable(PMD_TABLE_SIZE); pud_populate(&init_mm, pudp, pmdp); --- a/arch/powerpc/mm/pgtable_64.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/pgtable_64.c @@ -101,13 +101,13 @@ EXPORT_SYMBOL(__pte_frag_size_shift); #ifndef __PAGETABLE_PUD_FOLDED /* 4 level page table */ -struct page *pgd_page(pgd_t pgd) +struct page *p4d_page(p4d_t p4d) { - if (pgd_is_leaf(pgd)) { - VM_WARN_ON(!pgd_huge(pgd)); - return pte_page(pgd_pte(pgd)); + if (p4d_is_leaf(p4d)) { + VM_WARN_ON(!p4d_huge(p4d)); + return pte_page(p4d_pte(p4d)); } - return virt_to_page(pgd_page_vaddr(pgd)); + return virt_to_page(p4d_page_vaddr(p4d)); } #endif --- a/arch/powerpc/mm/pgtable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/pgtable.c @@ -265,6 +265,7 @@ int huge_ptep_set_access_flags(struct vm void assert_pte_locked(struct mm_struct *mm, unsigned long addr) { pgd_t *pgd; + p4d_t *p4d; pud_t *pud; pmd_t *pmd; @@ -272,7 +273,9 @@ void assert_pte_locked(struct mm_struct return; pgd = mm->pgd + pgd_index(addr); BUG_ON(pgd_none(*pgd)); - pud = pud_offset(pgd, addr); + p4d = p4d_offset(pgd, addr); + BUG_ON(p4d_none(*p4d)); + pud = pud_offset(p4d, addr); BUG_ON(pud_none(*pud)); pmd = pmd_offset(pud, addr); /* @@ -312,12 +315,13 @@ EXPORT_SYMBOL_GPL(vmalloc_to_phys); pte_t *__find_linux_pte(pgd_t *pgdir, unsigned long ea, bool *is_thp, unsigned *hpage_shift) { - pgd_t pgd, *pgdp; + pgd_t *pgdp; + p4d_t p4d, *p4dp; pud_t pud, *pudp; pmd_t pmd, *pmdp; pte_t *ret_pte; hugepd_t *hpdp = NULL; - unsigned pdshift = PGDIR_SHIFT; + unsigned pdshift; if (hpage_shift) *hpage_shift = 0; @@ -325,24 +329,28 @@ pte_t *__find_linux_pte(pgd_t *pgdir, un if (is_thp) *is_thp = false; - pgdp = pgdir + pgd_index(ea); - pgd = READ_ONCE(*pgdp); /* * Always operate on the local stack value. This make sure the * value don't get updated by a parallel THP split/collapse, * page fault or a page unmap. The return pte_t * is still not * stable. So should be checked there for above conditions. + * Top level is an exception because it is folded into p4d. */ - if (pgd_none(pgd)) + pgdp = pgdir + pgd_index(ea); + p4dp = p4d_offset(pgdp, ea); + p4d = READ_ONCE(*p4dp); + pdshift = P4D_SHIFT; + + if (p4d_none(p4d)) return NULL; - if (pgd_is_leaf(pgd)) { - ret_pte = (pte_t *)pgdp; + if (p4d_is_leaf(p4d)) { + ret_pte = (pte_t *)p4dp; goto out; } - if (is_hugepd(__hugepd(pgd_val(pgd)))) { - hpdp = (hugepd_t *)&pgd; + if (is_hugepd(__hugepd(p4d_val(p4d)))) { + hpdp = (hugepd_t *)&p4d; goto out_huge; } @@ -352,7 +360,7 @@ pte_t *__find_linux_pte(pgd_t *pgdir, un * irq disabled */ pdshift = PUD_SHIFT; - pudp = pud_offset(&pgd, ea); + pudp = pud_offset(&p4d, ea); pud = READ_ONCE(*pudp); if (pud_none(pud)) --- a/arch/powerpc/mm/ptdump/hashpagetable.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/ptdump/hashpagetable.c @@ -417,9 +417,9 @@ static void walk_pmd(struct pg_state *st } } -static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) +static void walk_pud(struct pg_state *st, p4d_t *p4d, unsigned long start) { - pud_t *pud = pud_offset(pgd, 0); + pud_t *pud = pud_offset(p4d, 0); unsigned long addr; unsigned int i; @@ -431,6 +431,20 @@ static void walk_pud(struct pg_state *st } } +static void walk_p4d(struct pg_state *st, pgd_t *pgd, unsigned long start) +{ + p4d_t *p4d = p4d_offset(pgd, 0); + unsigned long addr; + unsigned int i; + + for (i = 0; i < PTRS_PER_P4D; i++, p4d++) { + addr = start + i * P4D_SIZE; + if (!p4d_none(*p4d)) + /* p4d exists */ + walk_pud(st, p4d, addr); + } +} + static void walk_pagetables(struct pg_state *st) { pgd_t *pgd = pgd_offset_k(0UL); @@ -445,7 +459,7 @@ static void walk_pagetables(struct pg_st addr = KERN_VIRT_START + i * PGDIR_SIZE; if (!pgd_none(*pgd)) /* pgd exists */ - walk_pud(st, pgd, addr); + walk_p4d(st, pgd, addr); } } --- a/arch/powerpc/mm/ptdump/ptdump.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/mm/ptdump/ptdump.c @@ -277,9 +277,9 @@ static void walk_pmd(struct pg_state *st } } -static void walk_pud(struct pg_state *st, pgd_t *pgd, unsigned long start) +static void walk_pud(struct pg_state *st, p4d_t *p4d, unsigned long start) { - pud_t *pud = pud_offset(pgd, 0); + pud_t *pud = pud_offset(p4d, 0); unsigned long addr; unsigned int i; @@ -304,11 +304,13 @@ static void walk_pagetables(struct pg_st * the hash pagetable. */ for (i = pgd_index(addr); i < PTRS_PER_PGD; i++, pgd++, addr += PGDIR_SIZE) { - if (!pgd_none(*pgd) && !pgd_is_leaf(*pgd)) + p4d_t *p4d = p4d_offset(pgd, 0); + + if (!p4d_none(*p4d) && !p4d_is_leaf(*p4d)) /* pgd exists */ - walk_pud(st, pgd, addr); + walk_pud(st, p4d, addr); else - note_page(st, addr, 1, pgd_val(*pgd)); + note_page(st, addr, 1, p4d_val(*p4d)); } } --- a/arch/powerpc/xmon/xmon.c~powerpc-add-support-for-folded-p4d-page-tables +++ a/arch/powerpc/xmon/xmon.c @@ -3135,7 +3135,8 @@ static void show_pte(unsigned long addr) unsigned long tskv = 0; struct task_struct *tsk = NULL; struct mm_struct *mm; - pgd_t *pgdp, *pgdir; + pgd_t *pgdp; + p4d_t *p4dp; pud_t *pudp; pmd_t *pmdp; pte_t *ptep; @@ -3159,28 +3160,26 @@ static void show_pte(unsigned long addr) catch_memory_errors = 1; sync(); - if (mm == &init_mm) { + if (mm == &init_mm) pgdp = pgd_offset_k(addr); - pgdir = pgd_offset_k(0); - } else { + else pgdp = pgd_offset(mm, addr); - pgdir = pgd_offset(mm, 0); - } - if (pgd_none(*pgdp)) { - printf("no linux page table for address\n"); + p4dp = p4d_offset(pgdp, addr); + + if (p4d_none(*p4dp)) { + printf("No valid P4D\n"); return; } - printf("pgd @ 0x%px\n", pgdir); - - if (pgd_is_leaf(*pgdp)) { - format_pte(pgdp, pgd_val(*pgdp)); + if (p4d_is_leaf(*p4dp)) { + format_pte(p4dp, p4d_val(*p4dp)); return; } - printf("pgdp @ 0x%px = 0x%016lx\n", pgdp, pgd_val(*pgdp)); - pudp = pud_offset(pgdp, addr); + printf("p4dp @ 0x%px = 0x%016lx\n", p4dp, p4d_val(*p4dp)); + + pudp = pud_offset(p4dp, addr); if (pud_none(*pudp)) { printf("No valid PUD\n"); _