diff options
Diffstat (limited to 'arch/arm64/include/asm/kvm_mmu.h')
| -rw-r--r-- | arch/arm64/include/asm/kvm_mmu.h | 417 |
1 files changed, 341 insertions, 76 deletions
diff --git a/arch/arm64/include/asm/kvm_mmu.h b/arch/arm64/include/asm/kvm_mmu.h index efe609c6a3c9..2dc5e6e742bb 100644 --- a/arch/arm64/include/asm/kvm_mmu.h +++ b/arch/arm64/include/asm/kvm_mmu.h @@ -1,18 +1,7 @@ +/* SPDX-License-Identifier: GPL-2.0-only */ /* * Copyright (C) 2012,2013 - ARM Ltd * Author: Marc Zyngier <marc.zyngier@arm.com> - * - * This program is free software; you can redistribute it and/or modify - * it under the terms of the GNU General Public License version 2 as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it will be useful, - * but WITHOUT ANY WARRANTY; without even the implied warranty of - * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - * GNU General Public License for more details. - * - * You should have received a copy of the GNU General Public License - * along with this program. If not, see <http://www.gnu.org/licenses/>. */ #ifndef __ARM64_KVM_MMU_H__ @@ -20,116 +9,392 @@ #include <asm/page.h> #include <asm/memory.h> +#include <asm/mmu.h> +#include <asm/cpufeature.h> /* - * As we only have the TTBR0_EL2 register, we cannot express + * As ARMv8.0 only has the TTBR0_EL2 register, we cannot express * "negative" addresses. This makes it impossible to directly share * mappings with the kernel. * * Instead, give the HYP mode its own VA region at a fixed offset from * the kernel by just masking the top bits (which are all ones for a - * kernel address). + * kernel address). We need to find out how many bits to mask. + * + * We want to build a set of page tables that cover both parts of the + * idmap (the trampoline page used to initialize EL2), and our normal + * runtime VA space, at the same time. + * + * Given that the kernel uses VA_BITS for its entire address space, + * and that half of that space (VA_BITS - 1) is used for the linear + * mapping, we can also limit the EL2 space to (VA_BITS - 1). + * + * The main question is "Within the VA_BITS space, does EL2 use the + * top or the bottom half of that space to shadow the kernel's linear + * mapping?". As we need to idmap the trampoline page, this is + * determined by the range in which this page lives. + * + * If the page is in the bottom half, we have to use the top half. If + * the page is in the top half, we have to use the bottom half: + * + * T = __pa_symbol(__hyp_idmap_text_start) + * if (T & BIT(VA_BITS - 1)) + * HYP_VA_MIN = 0 //idmap in upper half + * else + * HYP_VA_MIN = 1 << (VA_BITS - 1) + * HYP_VA_MAX = HYP_VA_MIN + (1 << (VA_BITS - 1)) - 1 + * + * When using VHE, there are no separate hyp mappings and all KVM + * functionality is already mapped as part of the main kernel + * mappings, and none of this applies in that case. */ -#define HYP_PAGE_OFFSET_SHIFT VA_BITS -#define HYP_PAGE_OFFSET_MASK ((UL(1) << HYP_PAGE_OFFSET_SHIFT) - 1) -#define HYP_PAGE_OFFSET (PAGE_OFFSET & HYP_PAGE_OFFSET_MASK) + +#ifdef __ASSEMBLER__ + +#include <asm/alternative.h> /* - * Our virtual mapping for the idmap-ed MMU-enable code. Must be - * shared across all the page-tables. Conveniently, we use the last - * possible page, where no kernel mapping will ever exist. + * Convert a hypervisor VA to a PA + * reg: hypervisor address to be converted in place + * tmp: temporary register */ -#define TRAMPOLINE_VA (HYP_PAGE_OFFSET_MASK & PAGE_MASK) - -#ifdef __ASSEMBLY__ +.macro hyp_pa reg, tmp + ldr_l \tmp, hyp_physvirt_offset + add \reg, \reg, \tmp +.endm /* - * Convert a kernel VA into a HYP VA. - * reg: VA to be converted. + * Convert a hypervisor VA to a kernel image address + * reg: hypervisor address to be converted in place + * tmp: temporary register + * + * The actual code generation takes place in kvm_get_kimage_voffset, and + * the instructions below are only there to reserve the space and + * perform the register allocation (kvm_get_kimage_voffset uses the + * specific registers encoded in the instructions). */ -.macro kern_hyp_va reg - and \reg, \reg, #HYP_PAGE_OFFSET_MASK +.macro hyp_kimg_va reg, tmp + /* Convert hyp VA -> PA. */ + hyp_pa \reg, \tmp + + /* Load kimage_voffset. */ +alternative_cb ARM64_ALWAYS_SYSTEM, kvm_get_kimage_voffset + movz \tmp, #0 + movk \tmp, #0, lsl #16 + movk \tmp, #0, lsl #32 + movk \tmp, #0, lsl #48 +alternative_cb_end + + /* Convert PA -> kimg VA. */ + add \reg, \reg, \tmp .endm #else -#include <asm/cachetype.h> +#include <linux/pgtable.h> +#include <asm/pgalloc.h> +#include <asm/cache.h> #include <asm/cacheflush.h> +#include <asm/mmu_context.h> +#include <asm/kvm_emulate.h> +#include <asm/kvm_host.h> +#include <asm/kvm_nested.h> + +void kvm_update_va_mask(struct alt_instr *alt, + __le32 *origptr, __le32 *updptr, int nr_inst); +void kvm_compute_layout(void); +void kvm_apply_hyp_relocations(void); -#define KERN_TO_HYP(kva) ((unsigned long)kva - PAGE_OFFSET + HYP_PAGE_OFFSET) +#define __hyp_pa(x) (((phys_addr_t)(x)) + hyp_physvirt_offset) /* - * Align KVM with the kernel's view of physical memory. Should be - * 40bit IPA, with PGD being 8kB aligned in the 4KB page configuration. + * Convert a kernel VA into a HYP VA. + * + * Can be called from hyp or non-hyp context. + * + * The actual code generation takes place in kvm_update_va_mask(), and + * the instructions below are only there to reserve the space and + * perform the register allocation (kvm_update_va_mask() uses the + * specific registers encoded in the instructions). + */ +static __always_inline unsigned long __kern_hyp_va(unsigned long v) +{ +/* + * This #ifndef is an optimisation for when this is called from VHE hyp + * context. When called from a VHE non-hyp context, kvm_update_va_mask() will + * replace the instructions with `nop`s. */ -#define KVM_PHYS_SHIFT PHYS_MASK_SHIFT -#define KVM_PHYS_SIZE (1UL << KVM_PHYS_SHIFT) -#define KVM_PHYS_MASK (KVM_PHYS_SIZE - 1UL) +#ifndef __KVM_VHE_HYPERVISOR__ + asm volatile(ALTERNATIVE_CB("and %0, %0, #1\n" /* mask with va_mask */ + "ror %0, %0, #1\n" /* rotate to the first tag bit */ + "add %0, %0, #0\n" /* insert the low 12 bits of the tag */ + "add %0, %0, #0, lsl 12\n" /* insert the top 12 bits of the tag */ + "ror %0, %0, #63\n", /* rotate back */ + ARM64_ALWAYS_SYSTEM, + kvm_update_va_mask) + : "+r" (v)); +#endif + return v; +} -/* Make sure we get the right size, and thus the right alignment */ -#define PTRS_PER_S2_PGD (1 << (KVM_PHYS_SHIFT - PGDIR_SHIFT)) -#define S2_PGD_ORDER get_order(PTRS_PER_S2_PGD * sizeof(pgd_t)) +#define kern_hyp_va(v) ((typeof(v))(__kern_hyp_va((unsigned long)(v)))) -int create_hyp_mappings(void *from, void *to); -int create_hyp_io_mappings(void *from, void *to, phys_addr_t); -void free_boot_hyp_pgd(void); -void free_hyp_pgds(void); +extern u32 __hyp_va_bits; -int kvm_alloc_stage2_pgd(struct kvm *kvm); -void kvm_free_stage2_pgd(struct kvm *kvm); -int kvm_phys_addr_ioremap(struct kvm *kvm, phys_addr_t guest_ipa, - phys_addr_t pa, unsigned long size); +/* + * We currently support using a VM-specified IPA size. For backward + * compatibility, the default IPA size is fixed to 40bits. + */ +#define KVM_PHYS_SHIFT (40) + +#define kvm_phys_shift(mmu) VTCR_EL2_IPA((mmu)->vtcr) +#define kvm_phys_size(mmu) (_AC(1, ULL) << kvm_phys_shift(mmu)) +#define kvm_phys_mask(mmu) (kvm_phys_size(mmu) - _AC(1, ULL)) + +#include <asm/kvm_pgtable.h> +#include <asm/stage2_pgtable.h> + +int kvm_share_hyp(void *from, void *to); +void kvm_unshare_hyp(void *from, void *to); +int create_hyp_mappings(void *from, void *to, enum kvm_pgtable_prot prot); +int __create_hyp_mappings(unsigned long start, unsigned long size, + unsigned long phys, enum kvm_pgtable_prot prot); +int hyp_alloc_private_va_range(size_t size, unsigned long *haddr); +int create_hyp_io_mappings(phys_addr_t phys_addr, size_t size, + void __iomem **kaddr, + void __iomem **haddr); +int create_hyp_exec_mappings(phys_addr_t phys_addr, size_t size, + void **haddr); +int create_hyp_stack(phys_addr_t phys_addr, unsigned long *haddr); +void __init free_hyp_pgds(void); -int kvm_handle_guest_abort(struct kvm_vcpu *vcpu, struct kvm_run *run); +void kvm_stage2_unmap_range(struct kvm_s2_mmu *mmu, phys_addr_t start, + u64 size, bool may_block); +void kvm_stage2_flush_range(struct kvm_s2_mmu *mmu, phys_addr_t addr, phys_addr_t end); +void kvm_stage2_wp_range(struct kvm_s2_mmu *mmu, phys_addr_t addr, phys_addr_t end); -void kvm_mmu_free_memory_caches(struct kvm_vcpu *vcpu); +void stage2_unmap_vm(struct kvm *kvm); +int kvm_init_stage2_mmu(struct kvm *kvm, struct kvm_s2_mmu *mmu, unsigned long type); +void kvm_uninit_stage2_mmu(struct kvm *kvm); +void kvm_free_stage2_pgd(struct kvm_s2_mmu *mmu); +int kvm_phys_addr_ioremap(struct kvm *kvm, phys_addr_t guest_ipa, + phys_addr_t pa, unsigned long size, bool writable); + +int kvm_handle_guest_sea(struct kvm_vcpu *vcpu); +int kvm_handle_guest_abort(struct kvm_vcpu *vcpu); phys_addr_t kvm_mmu_get_httbr(void); -phys_addr_t kvm_mmu_get_boot_httbr(void); phys_addr_t kvm_get_idmap_vector(void); -int kvm_mmu_init(void); -void kvm_clear_hyp_idmap(void); +int __init kvm_mmu_init(u32 *hyp_va_bits); + +static inline void *__kvm_vector_slot2addr(void *base, + enum arm64_hyp_spectre_vector slot) +{ + int idx = slot - (slot != HYP_VECTOR_DIRECT); + + return base + (idx * SZ_2K); +} -#define kvm_set_pte(ptep, pte) set_pte(ptep, pte) +struct kvm; + +#define kvm_flush_dcache_to_poc(a,l) \ + dcache_clean_inval_poc((unsigned long)(a), (unsigned long)(a)+(l)) -static inline bool kvm_is_write_fault(unsigned long esr) +static inline bool vcpu_has_cache_enabled(struct kvm_vcpu *vcpu) { - unsigned long esr_ec = esr >> ESR_EL2_EC_SHIFT; + u64 cache_bits = SCTLR_ELx_M | SCTLR_ELx_C; + int reg; + + if (vcpu_is_el2(vcpu)) + reg = SCTLR_EL2; + else + reg = SCTLR_EL1; - if (esr_ec == ESR_EL2_EC_IABT) - return false; + return (vcpu_read_sys_reg(vcpu, reg) & cache_bits) == cache_bits; +} - if ((esr & ESR_EL2_ISV) && !(esr & ESR_EL2_WNR)) - return false; +static inline void __clean_dcache_guest_page(void *va, size_t size) +{ + /* + * With FWB, we ensure that the guest always accesses memory using + * cacheable attributes, and we don't have to clean to PoC when + * faulting in pages. Furthermore, FWB implies IDC, so cleaning to + * PoU is not required either in this case. + */ + if (cpus_have_final_cap(ARM64_HAS_STAGE2_FWB)) + return; - return true; + kvm_flush_dcache_to_poc(va, size); } -static inline void kvm_clean_dcache_area(void *addr, size_t size) {} -static inline void kvm_clean_pgd(pgd_t *pgd) {} -static inline void kvm_clean_pmd_entry(pmd_t *pmd) {} -static inline void kvm_clean_pte(pte_t *pte) {} -static inline void kvm_clean_pte_entry(pte_t *pte) {} +static inline size_t __invalidate_icache_max_range(void) +{ + u8 iminline; + u64 ctr; + + asm volatile(ALTERNATIVE_CB("movz %0, #0\n" + "movk %0, #0, lsl #16\n" + "movk %0, #0, lsl #32\n" + "movk %0, #0, lsl #48\n", + ARM64_ALWAYS_SYSTEM, + kvm_compute_final_ctr_el0) + : "=r" (ctr)); + + iminline = SYS_FIELD_GET(CTR_EL0, IminLine, ctr) + 2; + return MAX_DVM_OPS << iminline; +} -static inline void kvm_set_s2pte_writable(pte_t *pte) +static inline void __invalidate_icache_guest_page(void *va, size_t size) { - pte_val(*pte) |= PTE_S2_RDWR; + /* + * Blow the whole I-cache if it is aliasing (i.e. VIPT) or the + * invalidation range exceeds our arbitrary limit on invadations by + * cache line. + */ + if (icache_is_aliasing() || size > __invalidate_icache_max_range()) + icache_inval_all_pou(); + else + icache_inval_pou((unsigned long)va, (unsigned long)va + size); } -struct kvm; +void kvm_set_way_flush(struct kvm_vcpu *vcpu); +void kvm_toggle_cache(struct kvm_vcpu *vcpu, bool was_enabled); + +static inline unsigned int kvm_get_vmid_bits(void) +{ + int reg = read_sanitised_ftr_reg(SYS_ID_AA64MMFR1_EL1); -static inline void coherent_icache_guest_page(struct kvm *kvm, gfn_t gfn) + return get_vmid_bits(reg); +} + +/* + * We are not in the kvm->srcu critical section most of the time, so we take + * the SRCU read lock here. Since we copy the data from the user page, we + * can immediately drop the lock again. + */ +static inline int kvm_read_guest_lock(struct kvm *kvm, + gpa_t gpa, void *data, unsigned long len) { - if (!icache_is_aliasing()) { /* PIPT */ - unsigned long hva = gfn_to_hva(kvm, gfn); - flush_icache_range(hva, hva + PAGE_SIZE); - } else if (!icache_is_aivivt()) { /* non ASID-tagged VIVT */ - /* any kind of VIPT cache */ - __flush_icache_all(); - } + int srcu_idx = srcu_read_lock(&kvm->srcu); + int ret = kvm_read_guest(kvm, gpa, data, len); + + srcu_read_unlock(&kvm->srcu, srcu_idx); + + return ret; } -#define kvm_flush_dcache_to_poc(a,l) __flush_dcache_area((a), (l)) +static inline int kvm_write_guest_lock(struct kvm *kvm, gpa_t gpa, + const void *data, unsigned long len) +{ + int srcu_idx = srcu_read_lock(&kvm->srcu); + int ret = kvm_write_guest(kvm, gpa, data, len); + + srcu_read_unlock(&kvm->srcu, srcu_idx); + + return ret; +} + +#define kvm_phys_to_vttbr(addr) phys_to_ttbr(addr) + +/* + * When this is (directly or indirectly) used on the TLB invalidation + * path, we rely on a previously issued DSB so that page table updates + * and VMID reads are correctly ordered. + */ +static __always_inline u64 kvm_get_vttbr(struct kvm_s2_mmu *mmu) +{ + struct kvm_vmid *vmid = &mmu->vmid; + u64 vmid_field, baddr; + u64 cnp = system_supports_cnp() ? VTTBR_CNP_BIT : 0; + + baddr = mmu->pgd_phys; + vmid_field = atomic64_read(&vmid->id) << VTTBR_VMID_SHIFT; + vmid_field &= VTTBR_VMID_MASK(kvm_arm_vmid_bits); + return kvm_phys_to_vttbr(baddr) | vmid_field | cnp; +} + +/* + * Must be called from hyp code running at EL2 with an updated VTTBR + * and interrupts disabled. + */ +static __always_inline void __load_stage2(struct kvm_s2_mmu *mmu, + struct kvm_arch *arch) +{ + write_sysreg(mmu->vtcr, vtcr_el2); + write_sysreg(kvm_get_vttbr(mmu), vttbr_el2); + + /* + * ARM errata 1165522 and 1530923 require the actual execution of the + * above before we can switch to the EL1/EL0 translation regime used by + * the guest. + */ + asm(ALTERNATIVE("nop", "isb", ARM64_WORKAROUND_SPECULATIVE_AT)); +} + +static inline struct kvm *kvm_s2_mmu_to_kvm(struct kvm_s2_mmu *mmu) +{ + return container_of(mmu->arch, struct kvm, arch); +} + +static inline u64 get_vmid(u64 vttbr) +{ + return (vttbr & VTTBR_VMID_MASK(kvm_get_vmid_bits())) >> + VTTBR_VMID_SHIFT; +} + +static inline bool kvm_s2_mmu_valid(struct kvm_s2_mmu *mmu) +{ + return !(mmu->tlb_vttbr & VTTBR_CNP_BIT); +} + +static inline bool kvm_is_nested_s2_mmu(struct kvm *kvm, struct kvm_s2_mmu *mmu) +{ + /* + * Be careful, mmu may not be fully initialised so do look at + * *any* of its fields. + */ + return &kvm->arch.mmu != mmu; +} + +static inline void kvm_fault_lock(struct kvm *kvm) +{ + if (is_protected_kvm_enabled()) + write_lock(&kvm->mmu_lock); + else + read_lock(&kvm->mmu_lock); +} + +static inline void kvm_fault_unlock(struct kvm *kvm) +{ + if (is_protected_kvm_enabled()) + write_unlock(&kvm->mmu_lock); + else + read_unlock(&kvm->mmu_lock); +} + +/* + * ARM64 KVM relies on a simple conversion from physaddr to a kernel + * virtual address (KVA) when it does cache maintenance as the CMO + * instructions work on virtual addresses. This is incompatible with + * VM_PFNMAP VMAs which may not have a kernel direct mapping to a + * virtual address. + * + * With S2FWB and CACHE DIC features, KVM need not do cache flushing + * and CMOs are NOP'd. This has the effect of no longer requiring a + * KVA for addresses mapped into the S2. The presence of these features + * are thus necessary to support cacheable S2 mapping of VM_PFNMAP. + */ +static inline bool kvm_supports_cacheable_pfnmap(void) +{ + return cpus_have_final_cap(ARM64_HAS_STAGE2_FWB) && + cpus_have_final_cap(ARM64_HAS_CACHE_DIC); +} + +#ifdef CONFIG_PTDUMP_STAGE2_DEBUGFS +void kvm_s2_ptdump_create_debugfs(struct kvm *kvm); +#else +static inline void kvm_s2_ptdump_create_debugfs(struct kvm *kvm) {} +#endif /* CONFIG_PTDUMP_STAGE2_DEBUGFS */ -#endif /* __ASSEMBLY__ */ +#endif /* __ASSEMBLER__ */ #endif /* __ARM64_KVM_MMU_H__ */ |