diff options
Diffstat (limited to 'usr/src/uts/i86pc/sys/vmm.h')
| -rw-r--r-- | usr/src/uts/i86pc/sys/vmm.h | 565 |
1 files changed, 565 insertions, 0 deletions
diff --git a/usr/src/uts/i86pc/sys/vmm.h b/usr/src/uts/i86pc/sys/vmm.h new file mode 100644 index 0000000000..e876ce748f --- /dev/null +++ b/usr/src/uts/i86pc/sys/vmm.h @@ -0,0 +1,565 @@ +/*- + * Copyright (c) 2011 NetApp, Inc. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * + * THIS SOFTWARE IS PROVIDED BY NETAPP, INC ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL NETAPP, INC OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * $FreeBSD: head/sys/amd64/include/vmm.h 273375 2014-10-21 07:10:43Z neel $ + */ +/* + * This file and its contents are supplied under the terms of the + * Common Development and Distribution License ("CDDL"), version 1.0. + * You may only use this file in accordance with the terms of version + * 1.0 of the CDDL. + * + * A full copy of the text of the CDDL should have accompanied this + * source. A copy of the CDDL is also available via the Internet at + * http://www.illumos.org/license/CDDL. + * + * Copyright 2015 Pluribus Networks Inc. + */ + +#ifndef _VMM_H_ +#define _VMM_H_ + +#include <x86/segments.h> + +enum vm_suspend_how { + VM_SUSPEND_NONE, + VM_SUSPEND_RESET, + VM_SUSPEND_POWEROFF, + VM_SUSPEND_HALT, + VM_SUSPEND_LAST +}; + +/* + * Identifiers for architecturally defined registers. + */ +enum vm_reg_name { + VM_REG_GUEST_RAX, + VM_REG_GUEST_RBX, + VM_REG_GUEST_RCX, + VM_REG_GUEST_RDX, + VM_REG_GUEST_RSI, + VM_REG_GUEST_RDI, + VM_REG_GUEST_RBP, + VM_REG_GUEST_R8, + VM_REG_GUEST_R9, + VM_REG_GUEST_R10, + VM_REG_GUEST_R11, + VM_REG_GUEST_R12, + VM_REG_GUEST_R13, + VM_REG_GUEST_R14, + VM_REG_GUEST_R15, + VM_REG_GUEST_CR0, + VM_REG_GUEST_CR3, + VM_REG_GUEST_CR4, + VM_REG_GUEST_DR7, + VM_REG_GUEST_RSP, + VM_REG_GUEST_RIP, + VM_REG_GUEST_RFLAGS, + VM_REG_GUEST_ES, + VM_REG_GUEST_CS, + VM_REG_GUEST_SS, + VM_REG_GUEST_DS, + VM_REG_GUEST_FS, + VM_REG_GUEST_GS, + VM_REG_GUEST_LDTR, + VM_REG_GUEST_TR, + VM_REG_GUEST_IDTR, + VM_REG_GUEST_GDTR, + VM_REG_GUEST_EFER, + VM_REG_GUEST_CR2, + VM_REG_LAST +}; + +enum x2apic_state { + X2APIC_DISABLED, + X2APIC_ENABLED, + X2APIC_STATE_LAST +}; + +#define VM_INTINFO_VECTOR(info) ((info) & 0xff) +#define VM_INTINFO_DEL_ERRCODE 0x800 +#define VM_INTINFO_RSVD 0x7ffff000 +#define VM_INTINFO_VALID 0x80000000 +#define VM_INTINFO_TYPE 0x700 +#define VM_INTINFO_HWINTR (0 << 8) +#define VM_INTINFO_NMI (2 << 8) +#define VM_INTINFO_HWEXCEPTION (3 << 8) +#define VM_INTINFO_SWINTR (4 << 8) + +#define VM_MAX_NAMELEN 32 + +#ifdef _KERNEL + +struct vm; +struct vm_exception; +struct vm_memory_segment; +struct seg_desc; +struct vm_exit; +struct vm_run; +struct vhpet; +struct vioapic; +struct vlapic; +struct vm_guest_paging; + +typedef int (*vmm_init_func_t)(void); +typedef int (*vmm_cleanup_func_t)(void); +typedef void * (*vmi_init_func_t)(struct vm *vm); /* instance specific apis */ +typedef int (*vmi_run_func_t)(void *vmi, int vcpu, register_t rip); +typedef void (*vmi_cleanup_func_t)(void *vmi); +typedef int (*vmi_mmap_set_func_t)(void *vmi, vm_paddr_t gpa, + vm_paddr_t hpa, size_t length, + vm_memattr_t attr, int prot, + boolean_t superpages_ok); +typedef vm_paddr_t (*vmi_mmap_get_func_t)(void *vmi, vm_paddr_t gpa); +typedef int (*vmi_get_register_t)(void *vmi, int vcpu, int num, + uint64_t *retval); +typedef int (*vmi_set_register_t)(void *vmi, int vcpu, int num, + uint64_t val); +typedef int (*vmi_get_desc_t)(void *vmi, int vcpu, int num, + struct seg_desc *desc); +typedef int (*vmi_set_desc_t)(void *vmi, int vcpu, int num, + struct seg_desc *desc); +typedef int (*vmi_get_cap_t)(void *vmi, int vcpu, int num, int *retval); +typedef int (*vmi_set_cap_t)(void *vmi, int vcpu, int num, int val); +typedef struct vlapic * (*vmi_vlapic_init)(void *vmi, int vcpu); +typedef void (*vmi_vlapic_cleanup)(void *vmi, struct vlapic *vlapic); + +struct vmm_ops { + vmm_init_func_t init; /* module wide initialization */ + vmm_cleanup_func_t cleanup; + + vmi_init_func_t vminit; /* vm-specific initialization */ + vmi_run_func_t vmrun; + vmi_cleanup_func_t vmcleanup; + vmi_mmap_set_func_t vmmmap_set; + vmi_mmap_get_func_t vmmmap_get; + vmi_get_register_t vmgetreg; + vmi_set_register_t vmsetreg; + vmi_get_desc_t vmgetdesc; + vmi_set_desc_t vmsetdesc; + vmi_get_cap_t vmgetcap; + vmi_set_cap_t vmsetcap; + vmi_vlapic_init vlapic_init; + vmi_vlapic_cleanup vlapic_cleanup; +}; + +extern struct vmm_ops vmm_ops_intel; +extern struct vmm_ops vmm_ops_amd; + +int vm_create(const char *name, struct vm **retvm); +void vm_destroy(struct vm *vm); +const char *vm_name(struct vm *vm); +int vm_malloc(struct vm *vm, vm_paddr_t gpa, size_t len); +#ifdef __FreeBSD__ +int vm_map_mmio(struct vm *vm, vm_paddr_t gpa, size_t len, vm_paddr_t hpa); +#endif +int vm_unmap_mmio(struct vm *vm, vm_paddr_t gpa, size_t len); +#ifndef __FreeBSD__ +vm_paddr_t vm_gpa2hpa(struct vm *vm, vm_paddr_t gpa, size_t size); +#endif +void *vm_gpa_hold(struct vm *, vm_paddr_t gpa, size_t len, int prot, + void **cookie); +void vm_gpa_release(void *cookie); +int vm_gpabase2memseg(struct vm *vm, vm_paddr_t gpabase, + struct vm_memory_segment *seg); +int vm_get_register(struct vm *vm, int vcpu, int reg, uint64_t *retval); +int vm_set_register(struct vm *vm, int vcpu, int reg, uint64_t val); +int vm_get_seg_desc(struct vm *vm, int vcpu, int reg, + struct seg_desc *ret_desc); +int vm_set_seg_desc(struct vm *vm, int vcpu, int reg, + struct seg_desc *desc); +int vm_run(struct vm *vm, struct vm_run *vmrun); +int vm_inject_nmi(struct vm *vm, int vcpu); +int vm_nmi_pending(struct vm *vm, int vcpuid); +void vm_nmi_clear(struct vm *vm, int vcpuid); +int vm_inject_extint(struct vm *vm, int vcpu); +int vm_extint_pending(struct vm *vm, int vcpuid); +void vm_extint_clear(struct vm *vm, int vcpuid); +struct vlapic *vm_lapic(struct vm *vm, int cpu); +struct vioapic *vm_ioapic(struct vm *vm); +struct vhpet *vm_hpet(struct vm *vm); +int vm_get_capability(struct vm *vm, int vcpu, int type, int *val); +int vm_set_capability(struct vm *vm, int vcpu, int type, int val); +int vm_get_x2apic_state(struct vm *vm, int vcpu, enum x2apic_state *state); +int vm_set_x2apic_state(struct vm *vm, int vcpu, enum x2apic_state state); +int vm_apicid2vcpuid(struct vm *vm, int apicid); +int vm_activate_cpu(struct vm *vm, int vcpu); +cpuset_t vm_active_cpus(struct vm *vm); +struct vm_exit *vm_exitinfo(struct vm *vm, int vcpuid); + +typedef void (*vm_rendezvous_func_t)(struct vm *vm, int vcpuid, void *arg); + +/* + * Return 1 if device indicated by bus/slot/func is supposed to be a + * pci passthrough device. + * + * Return 0 otherwise. + */ +int vmm_is_pptdev(int bus, int slot, int func); + +void *vm_iommu_domain(struct vm *vm); + +enum vcpu_state { + VCPU_IDLE, + VCPU_FROZEN, + VCPU_RUNNING, + VCPU_SLEEPING, +}; + +int vcpu_set_state(struct vm *vm, int vcpu, enum vcpu_state state, + bool from_idle); +enum vcpu_state vcpu_get_state(struct vm *vm, int vcpu, int *hostcpu); + +static int __inline +vcpu_is_running(struct vm *vm, int vcpu, int *hostcpu) +{ + return (vcpu_get_state(vm, vcpu, hostcpu) == VCPU_RUNNING); +} + +void *vcpu_stats(struct vm *vm, int vcpu); +void vm_interrupt_hostcpu(struct vm *vm, int vcpu); +void vcpu_notify_event(struct vm *vm, int vcpuid, bool lapic_intr); +struct vatpic *vm_atpic(struct vm *vm); +struct vatpit *vm_atpit(struct vm *vm); + +/* + * Inject exception 'vme' into the guest vcpu. This function returns 0 on + * success and non-zero on failure. + * + * Wrapper functions like 'vm_inject_gp()' should be preferred to calling + * this function directly because they enforce the trap-like or fault-like + * behavior of an exception. + * + * This function should only be called in the context of the thread that is + * executing this vcpu. + */ +int vm_inject_exception(struct vm *vm, int vcpuid, struct vm_exception *vme); + +/* + * This function is called after a VM-exit that occurred during exception or + * interrupt delivery through the IDT. The format of 'intinfo' is described + * in Figure 15-1, "EXITINTINFO for All Intercepts", APM, Vol 2. + * + * If a VM-exit handler completes the event delivery successfully then it + * should call vm_exit_intinfo() to extinguish the pending event. For e.g., + * if the task switch emulation is triggered via a task gate then it should + * call this function with 'intinfo=0' to indicate that the external event + * is not pending anymore. + * + * Return value is 0 on success and non-zero on failure. + */ +int vm_exit_intinfo(struct vm *vm, int vcpuid, uint64_t intinfo); + +/* + * This function is called before every VM-entry to retrieve a pending + * event that should be injected into the guest. This function combines + * nested events into a double or triple fault. + * + * Returns 0 if there are no events that need to be injected into the guest + * and non-zero otherwise. + */ +int vm_entry_intinfo(struct vm *vm, int vcpuid, uint64_t *info); + +int vm_get_intinfo(struct vm *vm, int vcpuid, uint64_t *info1, uint64_t *info2); + +enum vm_reg_name vm_segment_name(int seg_encoding); + +struct vm_copyinfo { + uint64_t gpa; + size_t len; + void *hva; + void *cookie; +}; + +/* + * Set up 'copyinfo[]' to copy to/from guest linear address space starting + * at 'gla' and 'len' bytes long. The 'prot' should be set to PROT_READ for + * a copyin or PROT_WRITE for a copyout. + * + * Returns 0 on success. + * Returns 1 if an exception was injected into the guest. + * Returns -1 otherwise. + * + * The 'copyinfo[]' can be passed to 'vm_copyin()' or 'vm_copyout()' only if + * the return value is 0. The 'copyinfo[]' resources should be freed by calling + * 'vm_copy_teardown()' after the copy is done. + */ +int vm_copy_setup(struct vm *vm, int vcpuid, struct vm_guest_paging *paging, + uint64_t gla, size_t len, int prot, struct vm_copyinfo *copyinfo, + int num_copyinfo); +void vm_copy_teardown(struct vm *vm, int vcpuid, struct vm_copyinfo *copyinfo, + int num_copyinfo); +void vm_copyin(struct vm *vm, int vcpuid, struct vm_copyinfo *copyinfo, + void *kaddr, size_t len); +void vm_copyout(struct vm *vm, int vcpuid, const void *kaddr, + struct vm_copyinfo *copyinfo, size_t len); +#endif /* KERNEL */ + +#define VM_MAXCPU 16 /* maximum virtual cpus */ + +/* + * Identifiers for optional vmm capabilities + */ +enum vm_cap_type { + VM_CAP_HALT_EXIT, + VM_CAP_MTRAP_EXIT, + VM_CAP_PAUSE_EXIT, + VM_CAP_UNRESTRICTED_GUEST, + VM_CAP_ENABLE_INVPCID, + VM_CAP_MAX +}; + +enum vm_intr_trigger { + EDGE_TRIGGER, + LEVEL_TRIGGER +}; + +/* + * The 'access' field has the format specified in Table 21-2 of the Intel + * Architecture Manual vol 3b. + * + * XXX The contents of the 'access' field are architecturally defined except + * bit 16 - Segment Unusable. + */ +struct seg_desc { + uint64_t base; + uint32_t limit; + uint32_t access; +}; + +#define SEG_DESC_TYPE(access) ((access) & 0x001f) +#define SEG_DESC_DPL(access) (((access) >> 5) & 0x3) +#define SEG_DESC_PRESENT(access) (((access) & 0x0080) ? 1 : 0) +#define SEG_DESC_DEF32(access) (((access) & 0x4000) ? 1 : 0) +#define SEG_DESC_GRANULARITY(access) (((access) & 0x8000) ? 1 : 0) +#define SEG_DESC_UNUSABLE(access) (((access) & 0x10000) ? 1 : 0) + +enum vm_cpu_mode { + CPU_MODE_REAL, + CPU_MODE_PROTECTED, + CPU_MODE_COMPATIBILITY, /* IA-32E mode (CS.L = 0) */ + CPU_MODE_64BIT, /* IA-32E mode (CS.L = 1) */ +}; + +enum vm_paging_mode { + PAGING_MODE_FLAT, + PAGING_MODE_32, + PAGING_MODE_PAE, + PAGING_MODE_64, +}; + +struct vm_guest_paging { + uint64_t cr3; + int cpl; + enum vm_cpu_mode cpu_mode; + enum vm_paging_mode paging_mode; +}; + +/* + * The data structures 'vie' and 'vie_op' are meant to be opaque to the + * consumers of instruction decoding. The only reason why their contents + * need to be exposed is because they are part of the 'vm_exit' structure. + */ +struct vie_op { + uint8_t op_byte; /* actual opcode byte */ + uint8_t op_type; /* type of operation (e.g. MOV) */ + uint16_t op_flags; +}; + +#define VIE_INST_SIZE 15 +struct vie { + uint8_t inst[VIE_INST_SIZE]; /* instruction bytes */ + uint8_t num_valid; /* size of the instruction */ + uint8_t num_processed; + + uint8_t addrsize:4, opsize:4; /* address and operand sizes */ + uint8_t rex_w:1, /* REX prefix */ + rex_r:1, + rex_x:1, + rex_b:1, + rex_present:1, + repz_present:1, /* REP/REPE/REPZ prefix */ + repnz_present:1, /* REPNE/REPNZ prefix */ + opsize_override:1, /* Operand size override */ + addrsize_override:1, /* Address size override */ + segment_override:1; /* Segment override */ + + uint8_t mod:2, /* ModRM byte */ + reg:4, + rm:4; + + uint8_t ss:2, /* SIB byte */ + index:4, + base:4; + + uint8_t disp_bytes; + uint8_t imm_bytes; + + uint8_t scale; + int base_register; /* VM_REG_GUEST_xyz */ + int index_register; /* VM_REG_GUEST_xyz */ + int segment_register; /* VM_REG_GUEST_xyz */ + + int64_t displacement; /* optional addr displacement */ + int64_t immediate; /* optional immediate operand */ + + uint8_t decoded; /* set to 1 if successfully decoded */ + + struct vie_op op; /* opcode description */ +}; + +enum vm_exitcode { + VM_EXITCODE_INOUT, + VM_EXITCODE_VMX, + VM_EXITCODE_BOGUS, + VM_EXITCODE_RDMSR, + VM_EXITCODE_WRMSR, + VM_EXITCODE_HLT, + VM_EXITCODE_MTRAP, + VM_EXITCODE_PAUSE, + VM_EXITCODE_PAGING, + VM_EXITCODE_INST_EMUL, + VM_EXITCODE_SPINUP_AP, + VM_EXITCODE_DEPRECATED1, /* used to be SPINDOWN_CPU */ + VM_EXITCODE_INOUT_STR, + VM_EXITCODE_MAX +}; + +struct vm_inout { + uint16_t bytes:3; /* 1 or 2 or 4 */ + uint16_t in:1; + uint16_t string:1; + uint16_t rep:1; + uint16_t port; + uint32_t eax; /* valid for out */ +}; + +struct vm_inout_str { + struct vm_inout inout; /* must be the first element */ + struct vm_guest_paging paging; + uint64_t rflags; + uint64_t cr0; + uint64_t index; + uint64_t count; /* rep=1 (%rcx), rep=0 (1) */ + int addrsize; + enum vm_reg_name seg_name; + struct seg_desc seg_desc; +}; + +struct vm_exit { + enum vm_exitcode exitcode; + int inst_length; /* 0 means unknown */ + uint64_t rip; + union { + struct vm_inout inout; + struct vm_inout_str inout_str; + struct { + uint64_t gpa; + int fault_type; + } paging; + struct { + uint64_t gpa; + uint64_t gla; + uint64_t cs_base; + int cs_d; /* CS.D */ + struct vm_guest_paging paging; + struct vie vie; + } inst_emul; + /* + * VMX specific payload. Used when there is no "better" + * exitcode to represent the VM-exit. + */ + struct { + int status; /* vmx inst status */ + /* + * 'exit_reason' and 'exit_qualification' are valid + * only if 'status' is zero. + */ + uint32_t exit_reason; + uint64_t exit_qualification; + /* + * 'inst_error' and 'inst_type' are valid + * only if 'status' is non-zero. + */ + int inst_type; + int inst_error; + } vmx; + struct { + uint32_t code; /* ecx value */ + uint64_t wval; + } msr; + struct { + int vcpu; + uint64_t rip; + } spinup_ap; + struct { + uint64_t rflags; + } hlt; + } u; +}; + +/* APIs to inject faults into the guest */ +void vm_inject_fault(void *vm, int vcpuid, int vector, int errcode_valid, + int errcode); + +static __inline void +vm_inject_ud(void *vm, int vcpuid) +{ + vm_inject_fault(vm, vcpuid, IDT_UD, 0, 0); +} + +static __inline void +vm_inject_gp(void *vm, int vcpuid) +{ + vm_inject_fault(vm, vcpuid, IDT_GP, 1, 0); +} + +static __inline void +vm_inject_ac(void *vm, int vcpuid, int errcode) +{ + vm_inject_fault(vm, vcpuid, IDT_AC, 1, errcode); +} + +static __inline void +vm_inject_ss(void *vm, int vcpuid, int errcode) +{ + vm_inject_fault(vm, vcpuid, IDT_SS, 1, errcode); +} + +void vm_inject_pf(void *vm, int vcpuid, int error_code, uint64_t cr2); + +int vm_restart_instruction(void *vm, int vcpuid); + +#ifndef __FreeBSD__ +#ifdef _KERNEL +extern void vmm_sol_glue_init(void); +extern void vmm_sol_glue_cleanup(void); + +extern int vmm_mod_load(void); +extern int vmm_mod_unload(void); +#endif +#endif + +#endif /* _VMM_H_ */ |