Re: [PATCH v3 07/20] kvm: arm/arm64: Prepare for VM specific stage2 translations

From: Marc Zyngier
Date: Mon Jul 02 2018 - 06:13:08 EST

On 29/06/18 12:15, Suzuki K Poulose wrote:
> Right now the stage2 page table for a VM is hard coded, assuming
> an IPA of 40bits. As we are about to add support for per VM IPA,
> prepare the stage2 page table helpers to accept the kvm instance
> to make the right decision for the VM. No functional changes.
> Adds stage2_pgd_size(kvm) to replace S2_PGD_SIZE. Also, moves
> some of the definitions dependent on kvm instance to asm/kvm_mmu.h
> for arm32. In that process drop the _AC() specifier constants
>
> Cc: Marc Zyngier <marc.zyngier@xxxxxxx>
> Cc: Christoffer Dall <cdall@xxxxxxxxxx>
> Signed-off-by: Suzuki K Poulose <suzuki.poulose@xxxxxxx>
> ---
> Changes since V2:
> - Update commit description abuot the movement to asm/kvm_mmu.h
> for arm32
> - Drop _AC() specifiers
> ---
> arch/arm/include/asm/kvm_arm.h | 3 +-
> arch/arm/include/asm/kvm_mmu.h | 15 +++-
> arch/arm/include/asm/stage2_pgtable.h | 42 ++++-----
> arch/arm64/include/asm/kvm_mmu.h | 7 +-
> arch/arm64/include/asm/stage2_pgtable-nopmd.h | 18 ++--
> arch/arm64/include/asm/stage2_pgtable-nopud.h | 16 ++--
> arch/arm64/include/asm/stage2_pgtable.h | 49 ++++++-----
> virt/kvm/arm/arm.c | 2 +-
> virt/kvm/arm/mmu.c | 119 +++++++++++++-------------
> virt/kvm/arm/vgic/vgic-kvm-device.c | 2 +-
> 10 files changed, 148 insertions(+), 125 deletions(-)
>
> diff --git a/arch/arm/include/asm/kvm_arm.h b/arch/arm/include/asm/kvm_arm.h
> index 3ab8b37..c3f1f9b 100644
> --- a/arch/arm/include/asm/kvm_arm.h
> +++ b/arch/arm/include/asm/kvm_arm.h
> @@ -133,8 +133,7 @@
> * space.
> */
> #define KVM_PHYS_SHIFT (40)
> -#define KVM_PHYS_SIZE (_AC(1, ULL) << KVM_PHYS_SHIFT)
> -#define KVM_PHYS_MASK (KVM_PHYS_SIZE - _AC(1, ULL))
> +
> #define PTRS_PER_S2_PGD (_AC(1, ULL) << (KVM_PHYS_SHIFT - 30))
>
> /* Virtualization Translation Control Register (VTCR) bits */
> diff --git a/arch/arm/include/asm/kvm_mmu.h b/arch/arm/include/asm/kvm_mmu.h
> index 8553d68..f36eb20 100644
> --- a/arch/arm/include/asm/kvm_mmu.h
> +++ b/arch/arm/include/asm/kvm_mmu.h
> @@ -36,15 +36,19 @@
> })
>
> /*
> - * KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation levels.
> + * kvm_mmu_cache_min_pages() is the number of stage2 page
> + * table translation levels, excluding the top level, for
> + * the given VM. Since we have a 3 level page-table, this
> + * is fixed.

I find this comment quite confusing: number of levels, but excluding the
top one? The original one was just as bad, to be honest.

Can't we just say: "kvm_mmu_cache_min_page() is the number of pages
required to install a stage-2 translation"?

> */
> -#define KVM_MMU_CACHE_MIN_PAGES 2
> +#define kvm_mmu_cache_min_pages(kvm) 2
>
> #ifndef __ASSEMBLY__
>
> #include <linux/highmem.h>
> #include <asm/cacheflush.h>
> #include <asm/cputype.h>
> +#include <asm/kvm_arm.h>
> #include <asm/kvm_hyp.h>
> #include <asm/pgalloc.h>
> #include <asm/stage2_pgtable.h>
> @@ -52,6 +56,13 @@
> /* Ensure compatibility with arm64 */
> #define VA_BITS 32
>
> +#define kvm_phys_shift(kvm) KVM_PHYS_SHIFT
> +#define kvm_phys_size(kvm) (1ULL << kvm_phys_shift(kvm))
> +#define kvm_phys_mask(kvm) (kvm_phys_size(kvm) - 1ULL)
> +#define kvm_vttbr_baddr_mask(kvm) VTTBR_BADDR_MASK
> +
> +#define stage2_pgd_size(kvm) (PTRS_PER_S2_PGD * sizeof(pgd_t))
> +
> int create_hyp_mappings(void *from, void *to, pgprot_t prot);
> int create_hyp_io_mappings(phys_addr_t phys_addr, size_t size,
> void __iomem **kaddr,
> diff --git a/arch/arm/include/asm/stage2_pgtable.h b/arch/arm/include/asm/stage2_pgtable.h
> index 460d616..e22ae94 100644
> --- a/arch/arm/include/asm/stage2_pgtable.h
> +++ b/arch/arm/include/asm/stage2_pgtable.h
> @@ -19,43 +19,45 @@
> #ifndef __ARM_S2_PGTABLE_H_
> #define __ARM_S2_PGTABLE_H_
>
> -#define stage2_pgd_none(pgd) pgd_none(pgd)
> -#define stage2_pgd_clear(pgd) pgd_clear(pgd)
> -#define stage2_pgd_present(pgd) pgd_present(pgd)
> -#define stage2_pgd_populate(pgd, pud) pgd_populate(NULL, pgd, pud)
> -#define stage2_pud_offset(pgd, address) pud_offset(pgd, address)
> -#define stage2_pud_free(pud) pud_free(NULL, pud)
> +#define stage2_pgd_none(kvm, pgd) pgd_none(pgd)
> +#define stage2_pgd_clear(kvm, pgd) pgd_clear(pgd)
> +#define stage2_pgd_present(kvm, pgd) pgd_present(pgd)
> +#define stage2_pgd_populate(kvm, pgd, pud) pgd_populate(NULL, pgd, pud)
> +#define stage2_pud_offset(kvm, pgd, address) pud_offset(pgd, address)
> +#define stage2_pud_free(kvm, pud) pud_free(NULL, pud)
>
> -#define stage2_pud_none(pud) pud_none(pud)
> -#define stage2_pud_clear(pud) pud_clear(pud)
> -#define stage2_pud_present(pud) pud_present(pud)
> -#define stage2_pud_populate(pud, pmd) pud_populate(NULL, pud, pmd)
> -#define stage2_pmd_offset(pud, address) pmd_offset(pud, address)
> -#define stage2_pmd_free(pmd) pmd_free(NULL, pmd)
> +#define stage2_pud_none(kvm, pud) pud_none(pud)
> +#define stage2_pud_clear(kvm, pud) pud_clear(pud)
> +#define stage2_pud_present(kvm, pud) pud_present(pud)
> +#define stage2_pud_populate(kvm, pud, pmd) pud_populate(NULL, pud, pmd)
> +#define stage2_pmd_offset(kvm, pud, address) pmd_offset(pud, address)
> +#define stage2_pmd_free(kvm, pmd) pmd_free(NULL, pmd)
>
> -#define stage2_pud_huge(pud) pud_huge(pud)
> +#define stage2_pud_huge(kvm, pud) pud_huge(pud)
>
> /* Open coded p*d_addr_end that can deal with 64bit addresses */
> -static inline phys_addr_t stage2_pgd_addr_end(phys_addr_t addr, phys_addr_t end)
> +static inline phys_addr_t
> +stage2_pgd_addr_end(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> {
> phys_addr_t boundary = (addr + PGDIR_SIZE) & PGDIR_MASK;
>
> return (boundary - 1 < end - 1) ? boundary : end;
> }
>
> -#define stage2_pud_addr_end(addr, end) (end)
> +#define stage2_pud_addr_end(kvm, addr, end) (end)
>
> -static inline phys_addr_t stage2_pmd_addr_end(phys_addr_t addr, phys_addr_t end)
> +static inline phys_addr_t
> +stage2_pmd_addr_end(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> {
> phys_addr_t boundary = (addr + PMD_SIZE) & PMD_MASK;
>
> return (boundary - 1 < end - 1) ? boundary : end;
> }
>
> -#define stage2_pgd_index(addr) pgd_index(addr)
> +#define stage2_pgd_index(kvm, addr) pgd_index(addr)
>
> -#define stage2_pte_table_empty(ptep) kvm_page_empty(ptep)
> -#define stage2_pmd_table_empty(pmdp) kvm_page_empty(pmdp)
> -#define stage2_pud_table_empty(pudp) false
> +#define stage2_pte_table_empty(kvm, ptep) kvm_page_empty(ptep)
> +#define stage2_pmd_table_empty(kvm, pmdp) kvm_page_empty(pmdp)
> +#define stage2_pud_table_empty(kvm, pudp) false
>
> #endif /* __ARM_S2_PGTABLE_H_ */
> diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h
> index fb9a712..5da8f52 100644
> --- a/arch/arm64/include/asm/kvm_mmu.h
> +++ b/arch/arm64/include/asm/kvm_mmu.h
> @@ -141,8 +141,11 @@ static inline unsigned long __kern_hyp_va(unsigned long v)
> * We currently only support a 40bit IPA.
> */
> #define KVM_PHYS_SHIFT (40)
> -#define KVM_PHYS_SIZE (1UL << KVM_PHYS_SHIFT)
> -#define KVM_PHYS_MASK (KVM_PHYS_SIZE - 1UL)
> +
> +#define kvm_phys_shift(kvm) KVM_PHYS_SHIFT
> +#define kvm_phys_size(kvm) (_AC(1, ULL) << kvm_phys_shift(kvm))
> +#define kvm_phys_mask(kvm) (kvm_phys_size(kvm) - _AC(1, ULL))
> +#define kvm_vttbr_baddr_mask(kvm) VTTBR_BADDR_MASK
>
> #include <asm/stage2_pgtable.h>
>
> diff --git a/arch/arm64/include/asm/stage2_pgtable-nopmd.h b/arch/arm64/include/asm/stage2_pgtable-nopmd.h
> index 2656a0f..0280ded 100644
> --- a/arch/arm64/include/asm/stage2_pgtable-nopmd.h
> +++ b/arch/arm64/include/asm/stage2_pgtable-nopmd.h
> @@ -26,17 +26,17 @@
> #define S2_PMD_SIZE (1UL << S2_PMD_SHIFT)
> #define S2_PMD_MASK (~(S2_PMD_SIZE-1))
>
> -#define stage2_pud_none(pud) (0)
> -#define stage2_pud_present(pud) (1)
> -#define stage2_pud_clear(pud) do { } while (0)
> -#define stage2_pud_populate(pud, pmd) do { } while (0)
> -#define stage2_pmd_offset(pud, address) ((pmd_t *)(pud))
> +#define stage2_pud_none(kvm, pud) (0)
> +#define stage2_pud_present(kvm, pud) (1)
> +#define stage2_pud_clear(kvm, pud) do { } while (0)
> +#define stage2_pud_populate(kvm, pud, pmd) do { } while (0)
> +#define stage2_pmd_offset(kvm, pud, address) ((pmd_t *)(pud))
>
> -#define stage2_pmd_free(pmd) do { } while (0)
> +#define stage2_pmd_free(kvm, pmd) do { } while (0)
>
> -#define stage2_pmd_addr_end(addr, end) (end)
> +#define stage2_pmd_addr_end(kvm, addr, end) (end)
>
> -#define stage2_pud_huge(pud) (0)
> -#define stage2_pmd_table_empty(pmdp) (0)
> +#define stage2_pud_huge(kvm, pud) (0)
> +#define stage2_pmd_table_empty(kvm, pmdp) (0)
>
> #endif
> diff --git a/arch/arm64/include/asm/stage2_pgtable-nopud.h b/arch/arm64/include/asm/stage2_pgtable-nopud.h
> index 5ee87b5..cd6304e 100644
> --- a/arch/arm64/include/asm/stage2_pgtable-nopud.h
> +++ b/arch/arm64/include/asm/stage2_pgtable-nopud.h
> @@ -24,16 +24,16 @@
> #define S2_PUD_SIZE (_AC(1, UL) << S2_PUD_SHIFT)
> #define S2_PUD_MASK (~(S2_PUD_SIZE-1))
>
> -#define stage2_pgd_none(pgd) (0)
> -#define stage2_pgd_present(pgd) (1)
> -#define stage2_pgd_clear(pgd) do { } while (0)
> -#define stage2_pgd_populate(pgd, pud) do { } while (0)
> +#define stage2_pgd_none(kvm, pgd) (0)
> +#define stage2_pgd_present(kvm, pgd) (1)
> +#define stage2_pgd_clear(kvm, pgd) do { } while (0)
> +#define stage2_pgd_populate(kvm, pgd, pud) do { } while (0)
>
> -#define stage2_pud_offset(pgd, address) ((pud_t *)(pgd))
> +#define stage2_pud_offset(kvm, pgd, address) ((pud_t *)(pgd))
>
> -#define stage2_pud_free(x) do { } while (0)
> +#define stage2_pud_free(kvm, x) do { } while (0)
>
> -#define stage2_pud_addr_end(addr, end) (end)
> -#define stage2_pud_table_empty(pmdp) (0)
> +#define stage2_pud_addr_end(kvm, addr, end) (end)
> +#define stage2_pud_table_empty(kvm, pmdp) (0)
>
> #endif
> diff --git a/arch/arm64/include/asm/stage2_pgtable.h b/arch/arm64/include/asm/stage2_pgtable.h
> index 8b68099..057a405 100644
> --- a/arch/arm64/include/asm/stage2_pgtable.h
> +++ b/arch/arm64/include/asm/stage2_pgtable.h
> @@ -65,10 +65,10 @@
> #define PTRS_PER_S2_PGD (1 << (KVM_PHYS_SHIFT - S2_PGDIR_SHIFT))
>
> /*
> - * KVM_MMU_CACHE_MIN_PAGES is the number of stage2 page table translation
> + * kvm_mmmu_cache_min_pages is the number of stage2 page table translation
> * levels in addition to the PGD.
> */
> -#define KVM_MMU_CACHE_MIN_PAGES (STAGE2_PGTABLE_LEVELS - 1)
> +#define kvm_mmu_cache_min_pages(kvm) (STAGE2_PGTABLE_LEVELS - 1)

Same comment as for the 32bit case.

>
>
> #if STAGE2_PGTABLE_LEVELS > 3
> @@ -77,16 +77,17 @@
> #define S2_PUD_SIZE (_AC(1, UL) << S2_PUD_SHIFT)
> #define S2_PUD_MASK (~(S2_PUD_SIZE - 1))
>
> -#define stage2_pgd_none(pgd) pgd_none(pgd)
> -#define stage2_pgd_clear(pgd) pgd_clear(pgd)
> -#define stage2_pgd_present(pgd) pgd_present(pgd)
> -#define stage2_pgd_populate(pgd, pud) pgd_populate(NULL, pgd, pud)
> -#define stage2_pud_offset(pgd, address) pud_offset(pgd, address)
> -#define stage2_pud_free(pud) pud_free(NULL, pud)
> +#define stage2_pgd_none(kvm, pgd) pgd_none(pgd)
> +#define stage2_pgd_clear(kvm, pgd) pgd_clear(pgd)
> +#define stage2_pgd_present(kvm, pgd) pgd_present(pgd)
> +#define stage2_pgd_populate(kvm, pgd, pud) pgd_populate(NULL, pgd, pud)
> +#define stage2_pud_offset(kvm, pgd, address) pud_offset(pgd, address)
> +#define stage2_pud_free(kvm, pud) pud_free(NULL, pud)
>
> -#define stage2_pud_table_empty(pudp) kvm_page_empty(pudp)
> +#define stage2_pud_table_empty(kvm, pudp) kvm_page_empty(pudp)
>
> -static inline phys_addr_t stage2_pud_addr_end(phys_addr_t addr, phys_addr_t end)
> +static inline phys_addr_t
> +stage2_pud_addr_end(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> {
> phys_addr_t boundary = (addr + S2_PUD_SIZE) & S2_PUD_MASK;
>
> @@ -102,17 +103,18 @@ static inline phys_addr_t stage2_pud_addr_end(phys_addr_t addr, phys_addr_t end)
> #define S2_PMD_SIZE (_AC(1, UL) << S2_PMD_SHIFT)
> #define S2_PMD_MASK (~(S2_PMD_SIZE - 1))
>
> -#define stage2_pud_none(pud) pud_none(pud)
> -#define stage2_pud_clear(pud) pud_clear(pud)
> -#define stage2_pud_present(pud) pud_present(pud)
> -#define stage2_pud_populate(pud, pmd) pud_populate(NULL, pud, pmd)
> -#define stage2_pmd_offset(pud, address) pmd_offset(pud, address)
> -#define stage2_pmd_free(pmd) pmd_free(NULL, pmd)
> +#define stage2_pud_none(kvm, pud) pud_none(pud)
> +#define stage2_pud_clear(kvm, pud) pud_clear(pud)
> +#define stage2_pud_present(kvm, pud) pud_present(pud)
> +#define stage2_pud_populate(kvm, pud, pmd) pud_populate(NULL, pud, pmd)
> +#define stage2_pmd_offset(kvm, pud, address) pmd_offset(pud, address)
> +#define stage2_pmd_free(kvm, pmd) pmd_free(NULL, pmd)
>
> -#define stage2_pud_huge(pud) pud_huge(pud)
> -#define stage2_pmd_table_empty(pmdp) kvm_page_empty(pmdp)
> +#define stage2_pud_huge(kvm, pud) pud_huge(pud)
> +#define stage2_pmd_table_empty(kvm, pmdp) kvm_page_empty(pmdp)
>
> -static inline phys_addr_t stage2_pmd_addr_end(phys_addr_t addr, phys_addr_t end)
> +static inline phys_addr_t
> +stage2_pmd_addr_end(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> {
> phys_addr_t boundary = (addr + S2_PMD_SIZE) & S2_PMD_MASK;
>
> @@ -121,7 +123,7 @@ static inline phys_addr_t stage2_pmd_addr_end(phys_addr_t addr, phys_addr_t end)
>
> #endif /* STAGE2_PGTABLE_LEVELS > 2 */
>
> -#define stage2_pte_table_empty(ptep) kvm_page_empty(ptep)
> +#define stage2_pte_table_empty(kvm, ptep) kvm_page_empty(ptep)
>
> #if STAGE2_PGTABLE_LEVELS == 2
> #include <asm/stage2_pgtable-nopmd.h>
> @@ -129,10 +131,13 @@ static inline phys_addr_t stage2_pmd_addr_end(phys_addr_t addr, phys_addr_t end)
> #include <asm/stage2_pgtable-nopud.h>
> #endif
>
> +#define stage2_pgd_size(kvm) (PTRS_PER_S2_PGD * sizeof(pgd_t))
>
> -#define stage2_pgd_index(addr) (((addr) >> S2_PGDIR_SHIFT) & (PTRS_PER_S2_PGD - 1))
> +#define stage2_pgd_index(kvm, addr) \
> + (((addr) >> S2_PGDIR_SHIFT) & (PTRS_PER_S2_PGD - 1))
>
> -static inline phys_addr_t stage2_pgd_addr_end(phys_addr_t addr, phys_addr_t end)
> +static inline phys_addr_t
> +stage2_pgd_addr_end(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> {
> phys_addr_t boundary = (addr + S2_PGDIR_SIZE) & S2_PGDIR_MASK;
>
> diff --git a/virt/kvm/arm/arm.c b/virt/kvm/arm/arm.c
> index 04e554c..d2637bb 100644
> --- a/virt/kvm/arm/arm.c
> +++ b/virt/kvm/arm/arm.c
> @@ -538,7 +538,7 @@ static void update_vttbr(struct kvm *kvm)
>
> /* update vttbr to be used with the new vmid */
> pgd_phys = virt_to_phys(kvm->arch.pgd);
> - BUG_ON(pgd_phys & ~VTTBR_BADDR_MASK);
> + BUG_ON(pgd_phys & ~kvm_vttbr_baddr_mask(kvm));
> vmid = ((u64)(kvm->arch.vmid) << VTTBR_VMID_SHIFT) & VTTBR_VMID_MASK(kvm_vmid_bits);
> kvm->arch.vttbr = kvm_phys_to_vttbr(pgd_phys) | vmid;
>
> diff --git a/virt/kvm/arm/mmu.c b/virt/kvm/arm/mmu.c
> index 308171c..82dd571 100644
> --- a/virt/kvm/arm/mmu.c
> +++ b/virt/kvm/arm/mmu.c
> @@ -45,7 +45,6 @@ static phys_addr_t hyp_idmap_vector;
>
> static unsigned long io_map_base;
>
> -#define S2_PGD_SIZE (PTRS_PER_S2_PGD * sizeof(pgd_t))
> #define hyp_pgd_order get_order(PTRS_PER_PGD * sizeof(pgd_t))
>
> #define KVM_S2PTE_FLAG_IS_IOMAP (1UL << 0)
> @@ -150,20 +149,20 @@ static void *mmu_memory_cache_alloc(struct kvm_mmu_memory_cache *mc)
>
> static void clear_stage2_pgd_entry(struct kvm *kvm, pgd_t *pgd, phys_addr_t addr)
> {
> - pud_t *pud_table __maybe_unused = stage2_pud_offset(pgd, 0UL);
> - stage2_pgd_clear(pgd);
> + pud_t *pud_table __maybe_unused = stage2_pud_offset(kvm, pgd, 0UL);
> + stage2_pgd_clear(kvm, pgd);
> kvm_tlb_flush_vmid_ipa(kvm, addr);
> - stage2_pud_free(pud_table);
> + stage2_pud_free(kvm, pud_table);
> put_page(virt_to_page(pgd));
> }
>
> static void clear_stage2_pud_entry(struct kvm *kvm, pud_t *pud, phys_addr_t addr)
> {
> - pmd_t *pmd_table __maybe_unused = stage2_pmd_offset(pud, 0);
> - VM_BUG_ON(stage2_pud_huge(*pud));
> - stage2_pud_clear(pud);
> + pmd_t *pmd_table __maybe_unused = stage2_pmd_offset(kvm, pud, 0);
> + VM_BUG_ON(stage2_pud_huge(kvm, *pud));
> + stage2_pud_clear(kvm, pud);
> kvm_tlb_flush_vmid_ipa(kvm, addr);
> - stage2_pmd_free(pmd_table);
> + stage2_pmd_free(kvm, pmd_table);
> put_page(virt_to_page(pud));
> }
>
> @@ -219,7 +218,7 @@ static void unmap_stage2_ptes(struct kvm *kvm, pmd_t *pmd,
> }
> } while (pte++, addr += PAGE_SIZE, addr != end);
>
> - if (stage2_pte_table_empty(start_pte))
> + if (stage2_pte_table_empty(kvm, start_pte))
> clear_stage2_pmd_entry(kvm, pmd, start_addr);
> }
>
> @@ -229,9 +228,9 @@ static void unmap_stage2_pmds(struct kvm *kvm, pud_t *pud,
> phys_addr_t next, start_addr = addr;
> pmd_t *pmd, *start_pmd;
>
> - start_pmd = pmd = stage2_pmd_offset(pud, addr);
> + start_pmd = pmd = stage2_pmd_offset(kvm, pud, addr);
> do {
> - next = stage2_pmd_addr_end(addr, end);
> + next = stage2_pmd_addr_end(kvm, addr, end);
> if (!pmd_none(*pmd)) {
> if (pmd_thp_or_huge(*pmd)) {
> pmd_t old_pmd = *pmd;
> @@ -248,7 +247,7 @@ static void unmap_stage2_pmds(struct kvm *kvm, pud_t *pud,
> }
> } while (pmd++, addr = next, addr != end);
>
> - if (stage2_pmd_table_empty(start_pmd))
> + if (stage2_pmd_table_empty(kvm, start_pmd))
> clear_stage2_pud_entry(kvm, pud, start_addr);
> }
>
> @@ -258,14 +257,14 @@ static void unmap_stage2_puds(struct kvm *kvm, pgd_t *pgd,
> phys_addr_t next, start_addr = addr;
> pud_t *pud, *start_pud;
>
> - start_pud = pud = stage2_pud_offset(pgd, addr);
> + start_pud = pud = stage2_pud_offset(kvm, pgd, addr);
> do {
> - next = stage2_pud_addr_end(addr, end);
> - if (!stage2_pud_none(*pud)) {
> - if (stage2_pud_huge(*pud)) {
> + next = stage2_pud_addr_end(kvm, addr, end);
> + if (!stage2_pud_none(kvm, *pud)) {
> + if (stage2_pud_huge(kvm, *pud)) {
> pud_t old_pud = *pud;
>
> - stage2_pud_clear(pud);
> + stage2_pud_clear(kvm, pud);
> kvm_tlb_flush_vmid_ipa(kvm, addr);
> kvm_flush_dcache_pud(old_pud);
> put_page(virt_to_page(pud));
> @@ -275,7 +274,7 @@ static void unmap_stage2_puds(struct kvm *kvm, pgd_t *pgd,
> }
> } while (pud++, addr = next, addr != end);
>
> - if (stage2_pud_table_empty(start_pud))
> + if (stage2_pud_table_empty(kvm, start_pud))
> clear_stage2_pgd_entry(kvm, pgd, start_addr);
> }
>
> @@ -299,7 +298,7 @@ static void unmap_stage2_range(struct kvm *kvm, phys_addr_t start, u64 size)
> assert_spin_locked(&kvm->mmu_lock);
> WARN_ON(size & ~PAGE_MASK);
>
> - pgd = kvm->arch.pgd + stage2_pgd_index(addr);
> + pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
> do {
> /*
> * Make sure the page table is still active, as another thread
> @@ -308,8 +307,8 @@ static void unmap_stage2_range(struct kvm *kvm, phys_addr_t start, u64 size)
> */
> if (!READ_ONCE(kvm->arch.pgd))
> break;
> - next = stage2_pgd_addr_end(addr, end);
> - if (!stage2_pgd_none(*pgd))
> + next = stage2_pgd_addr_end(kvm, addr, end);
> + if (!stage2_pgd_none(kvm, *pgd))
> unmap_stage2_puds(kvm, pgd, addr, next);
> /*
> * If the range is too large, release the kvm->mmu_lock
> @@ -338,9 +337,9 @@ static void stage2_flush_pmds(struct kvm *kvm, pud_t *pud,
> pmd_t *pmd;
> phys_addr_t next;
>
> - pmd = stage2_pmd_offset(pud, addr);
> + pmd = stage2_pmd_offset(kvm, pud, addr);
> do {
> - next = stage2_pmd_addr_end(addr, end);
> + next = stage2_pmd_addr_end(kvm, addr, end);
> if (!pmd_none(*pmd)) {
> if (pmd_thp_or_huge(*pmd))
> kvm_flush_dcache_pmd(*pmd);
> @@ -356,11 +355,11 @@ static void stage2_flush_puds(struct kvm *kvm, pgd_t *pgd,
> pud_t *pud;
> phys_addr_t next;
>
> - pud = stage2_pud_offset(pgd, addr);
> + pud = stage2_pud_offset(kvm, pgd, addr);
> do {
> - next = stage2_pud_addr_end(addr, end);
> - if (!stage2_pud_none(*pud)) {
> - if (stage2_pud_huge(*pud))
> + next = stage2_pud_addr_end(kvm, addr, end);
> + if (!stage2_pud_none(kvm, *pud)) {
> + if (stage2_pud_huge(kvm, *pud))
> kvm_flush_dcache_pud(*pud);
> else
> stage2_flush_pmds(kvm, pud, addr, next);
> @@ -376,10 +375,10 @@ static void stage2_flush_memslot(struct kvm *kvm,
> phys_addr_t next;
> pgd_t *pgd;
>
> - pgd = kvm->arch.pgd + stage2_pgd_index(addr);
> + pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
> do {
> - next = stage2_pgd_addr_end(addr, end);
> - if (!stage2_pgd_none(*pgd))
> + next = stage2_pgd_addr_end(kvm, addr, end);
> + if (!stage2_pgd_none(kvm, *pgd))
> stage2_flush_puds(kvm, pgd, addr, next);
> } while (pgd++, addr = next, addr != end);
> }
> @@ -869,7 +868,7 @@ int kvm_alloc_stage2_pgd(struct kvm *kvm)
> }
>
> /* Allocate the HW PGD, making sure that each page gets its own refcount */
> - pgd = alloc_pages_exact(S2_PGD_SIZE, GFP_KERNEL | __GFP_ZERO);
> + pgd = alloc_pages_exact(stage2_pgd_size(kvm), GFP_KERNEL | __GFP_ZERO);
> if (!pgd)
> return -ENOMEM;
>
> @@ -958,7 +957,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm)
>
> spin_lock(&kvm->mmu_lock);
> if (kvm->arch.pgd) {
> - unmap_stage2_range(kvm, 0, KVM_PHYS_SIZE);
> + unmap_stage2_range(kvm, 0, kvm_phys_size(kvm));
> pgd = READ_ONCE(kvm->arch.pgd);
> kvm->arch.pgd = NULL;
> }
> @@ -966,7 +965,7 @@ void kvm_free_stage2_pgd(struct kvm *kvm)
>
> /* Free the HW pgd, one page at a time */
> if (pgd)
> - free_pages_exact(pgd, S2_PGD_SIZE);
> + free_pages_exact(pgd, stage2_pgd_size(kvm));
> }
>
> static pud_t *stage2_get_pud(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
> @@ -975,16 +974,16 @@ static pud_t *stage2_get_pud(struct kvm *kvm, struct kvm_mmu_memory_cache *cache
> pgd_t *pgd;
> pud_t *pud;
>
> - pgd = kvm->arch.pgd + stage2_pgd_index(addr);
> - if (stage2_pgd_none(*pgd)) {
> + pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
> + if (stage2_pgd_none(kvm, *pgd)) {
> if (!cache)
> return NULL;
> pud = mmu_memory_cache_alloc(cache);
> - stage2_pgd_populate(pgd, pud);
> + stage2_pgd_populate(kvm, pgd, pud);
> get_page(virt_to_page(pgd));
> }
>
> - return stage2_pud_offset(pgd, addr);
> + return stage2_pud_offset(kvm, pgd, addr);
> }
>
> static pmd_t *stage2_get_pmd(struct kvm *kvm, struct kvm_mmu_memory_cache *cache,
> @@ -997,15 +996,15 @@ static pmd_t *stage2_get_pmd(struct kvm *kvm, struct kvm_mmu_memory_cache *cache
> if (!pud)
> return NULL;
>
> - if (stage2_pud_none(*pud)) {
> + if (stage2_pud_none(kvm, *pud)) {
> if (!cache)
> return NULL;
> pmd = mmu_memory_cache_alloc(cache);
> - stage2_pud_populate(pud, pmd);
> + stage2_pud_populate(kvm, pud, pmd);
> get_page(virt_to_page(pud));
> }
>
> - return stage2_pmd_offset(pud, addr);
> + return stage2_pmd_offset(kvm, pud, addr);
> }
>
> static int stage2_set_pmd_huge(struct kvm *kvm, struct kvm_mmu_memory_cache
> @@ -1159,8 +1158,9 @@ int kvm_phys_addr_ioremap(struct kvm *kvm, phys_addr_t guest_ipa,
> if (writable)
> pte = kvm_s2pte_mkwrite(pte);
>
> - ret = mmu_topup_memory_cache(&cache, KVM_MMU_CACHE_MIN_PAGES,
> - KVM_NR_MEM_OBJS);
> + ret = mmu_topup_memory_cache(&cache,
> + kvm_mmu_cache_min_pages(kvm),
> + KVM_NR_MEM_OBJS);
> if (ret)
> goto out;
> spin_lock(&kvm->mmu_lock);
> @@ -1248,19 +1248,21 @@ static void stage2_wp_ptes(pmd_t *pmd, phys_addr_t addr, phys_addr_t end)
>
> /**
> * stage2_wp_pmds - write protect PUD range
> + * kvm: kvm instance for the VM
> * @pud: pointer to pud entry
> * @addr: range start address
> * @end: range end address
> */
> -static void stage2_wp_pmds(pud_t *pud, phys_addr_t addr, phys_addr_t end)
> +static void stage2_wp_pmds(struct kvm *kvm, pud_t *pud,
> + phys_addr_t addr, phys_addr_t end)
> {
> pmd_t *pmd;
> phys_addr_t next;
>
> - pmd = stage2_pmd_offset(pud, addr);
> + pmd = stage2_pmd_offset(kvm, pud, addr);
>
> do {
> - next = stage2_pmd_addr_end(addr, end);
> + next = stage2_pmd_addr_end(kvm, addr, end);
> if (!pmd_none(*pmd)) {
> if (pmd_thp_or_huge(*pmd)) {
> if (!kvm_s2pmd_readonly(pmd))
> @@ -1280,18 +1282,19 @@ static void stage2_wp_pmds(pud_t *pud, phys_addr_t addr, phys_addr_t end)
> *
> * Process PUD entries, for a huge PUD we cause a panic.
> */
> -static void stage2_wp_puds(pgd_t *pgd, phys_addr_t addr, phys_addr_t end)
> +static void stage2_wp_puds(struct kvm *kvm, pgd_t *pgd,
> + phys_addr_t addr, phys_addr_t end)
> {
> pud_t *pud;
> phys_addr_t next;
>
> - pud = stage2_pud_offset(pgd, addr);
> + pud = stage2_pud_offset(kvm, pgd, addr);
> do {
> - next = stage2_pud_addr_end(addr, end);
> - if (!stage2_pud_none(*pud)) {
> + next = stage2_pud_addr_end(kvm, addr, end);
> + if (!stage2_pud_none(kvm, *pud)) {
> /* TODO:PUD not supported, revisit later if supported */
> - BUG_ON(stage2_pud_huge(*pud));
> - stage2_wp_pmds(pud, addr, next);
> + BUG_ON(stage2_pud_huge(kvm, *pud));
> + stage2_wp_pmds(kvm, pud, addr, next);
> }
> } while (pud++, addr = next, addr != end);
> }
> @@ -1307,7 +1310,7 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> pgd_t *pgd;
> phys_addr_t next;
>
> - pgd = kvm->arch.pgd + stage2_pgd_index(addr);
> + pgd = kvm->arch.pgd + stage2_pgd_index(kvm, addr);
> do {
> /*
> * Release kvm_mmu_lock periodically if the memory region is
> @@ -1321,9 +1324,9 @@ static void stage2_wp_range(struct kvm *kvm, phys_addr_t addr, phys_addr_t end)
> cond_resched_lock(&kvm->mmu_lock);
> if (!READ_ONCE(kvm->arch.pgd))
> break;
> - next = stage2_pgd_addr_end(addr, end);
> - if (stage2_pgd_present(*pgd))
> - stage2_wp_puds(pgd, addr, next);
> + next = stage2_pgd_addr_end(kvm, addr, end);
> + if (stage2_pgd_present(kvm, *pgd))
> + stage2_wp_puds(kvm, pgd, addr, next);
> } while (pgd++, addr = next, addr != end);
> }
>
> @@ -1472,7 +1475,7 @@ static int user_mem_abort(struct kvm_vcpu *vcpu, phys_addr_t fault_ipa,
> up_read(&current->mm->mmap_sem);
>
> /* We need minimum second+third level pages */
> - ret = mmu_topup_memory_cache(memcache, KVM_MMU_CACHE_MIN_PAGES,
> + ret = mmu_topup_memory_cache(memcache, kvm_mmu_cache_min_pages(kvm),
> KVM_NR_MEM_OBJS);
> if (ret)
> return ret;
> @@ -1715,7 +1718,7 @@ int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run)
> }
>
> /* Userspace should not be able to register out-of-bounds IPAs */
> - VM_BUG_ON(fault_ipa >= KVM_PHYS_SIZE);
> + VM_BUG_ON(fault_ipa >= kvm_phys_size(vcpu->kvm));
>
> if (fault_status == FSC_ACCESS) {
> handle_access_fault(vcpu, fault_ipa);
> @@ -2019,7 +2022,7 @@ int kvm_arch_prepare_memory_region(struct kvm *kvm,
> * space addressable by the KVM guest IPA space.
> */
> if (memslot->base_gfn + memslot->npages >=
> - (KVM_PHYS_SIZE >> PAGE_SHIFT))
> + (kvm_phys_size(kvm) >> PAGE_SHIFT))
> return -EFAULT;
>
> down_read(&current->mm->mmap_sem);
> diff --git a/virt/kvm/arm/vgic/vgic-kvm-device.c b/virt/kvm/arm/vgic/vgic-kvm-device.c
> index 6ada243..114dce9 100644
> --- a/virt/kvm/arm/vgic/vgic-kvm-device.c
> +++ b/virt/kvm/arm/vgic/vgic-kvm-device.c
> @@ -25,7 +25,7 @@
> int vgic_check_ioaddr(struct kvm *kvm, phys_addr_t *ioaddr,
> phys_addr_t addr, phys_addr_t alignment)
> {
> - if (addr & ~KVM_PHYS_MASK)
> + if (addr & ~kvm_phys_mask(kvm))
> return -E2BIG;
>
> if (!IS_ALIGNED(addr, alignment))
>

Otherwise:

Acked-by: Marc Zyngier <marc.zyngier@xxxxxxx>

M.
--
Jazz is not dead. It just smells funny...