freebsd-dev/lib/libvmmapi/vmmapi.c

869 lines
18 KiB
C
Raw Normal View History

/*-
* 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.
*
* $FreeBSD$
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/types.h>
#include <sys/sysctl.h>
#include <sys/ioctl.h>
#include <sys/mman.h>
#include <machine/specialreg.h>
#include <stdio.h>
#include <stdlib.h>
#include <assert.h>
#include <string.h>
#include <fcntl.h>
#include <unistd.h>
#include <libutil.h>
#include <machine/vmm.h>
#include <machine/vmm_dev.h>
#include "vmmapi.h"
#define MB (1024 * 1024UL)
#define GB (1024 * 1024 * 1024UL)
struct vmctx {
int fd;
uint32_t lowmem_limit;
enum vm_mmap_style vms;
size_t lowmem;
char *lowmem_addr;
size_t highmem;
char *highmem_addr;
char *name;
};
#define CREATE(x) sysctlbyname("hw.vmm.create", NULL, NULL, (x), strlen((x)))
#define DESTROY(x) sysctlbyname("hw.vmm.destroy", NULL, NULL, (x), strlen((x)))
static int
vm_device_open(const char *name)
{
int fd, len;
char *vmfile;
len = strlen("/dev/vmm/") + strlen(name) + 1;
vmfile = malloc(len);
assert(vmfile != NULL);
snprintf(vmfile, len, "/dev/vmm/%s", name);
/* Open the device file */
fd = open(vmfile, O_RDWR, 0);
free(vmfile);
return (fd);
}
int
vm_create(const char *name)
{
return (CREATE((char *)name));
}
struct vmctx *
vm_open(const char *name)
{
struct vmctx *vm;
vm = malloc(sizeof(struct vmctx) + strlen(name) + 1);
assert(vm != NULL);
vm->fd = -1;
vm->lowmem_limit = 3 * GB;
vm->name = (char *)(vm + 1);
strcpy(vm->name, name);
if ((vm->fd = vm_device_open(vm->name)) < 0)
goto err;
return (vm);
err:
vm_destroy(vm);
return (NULL);
}
void
vm_destroy(struct vmctx *vm)
{
assert(vm != NULL);
if (vm->fd >= 0)
close(vm->fd);
DESTROY(vm->name);
free(vm);
}
int
vm_parse_memsize(const char *optarg, size_t *ret_memsize)
{
char *endptr;
size_t optval;
int error;
optval = strtoul(optarg, &endptr, 0);
if (*optarg != '\0' && *endptr == '\0') {
/*
* For the sake of backward compatibility if the memory size
* specified on the command line is less than a megabyte then
* it is interpreted as being in units of MB.
*/
if (optval < MB)
optval *= MB;
*ret_memsize = optval;
error = 0;
} else
error = expand_number(optarg, ret_memsize);
return (error);
}
int
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
vm_get_memory_seg(struct vmctx *ctx, vm_paddr_t gpa, size_t *ret_len,
int *wired)
{
int error;
struct vm_memory_segment seg;
bzero(&seg, sizeof(seg));
seg.gpa = gpa;
error = ioctl(ctx->fd, VM_GET_MEMORY_SEG, &seg);
*ret_len = seg.len;
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
if (wired != NULL)
*wired = seg.wired;
return (error);
}
uint32_t
vm_get_lowmem_limit(struct vmctx *ctx)
{
return (ctx->lowmem_limit);
}
void
vm_set_lowmem_limit(struct vmctx *ctx, uint32_t limit)
{
ctx->lowmem_limit = limit;
}
static int
setup_memory_segment(struct vmctx *ctx, vm_paddr_t gpa, size_t len, char **addr)
{
int error;
struct vm_memory_segment seg;
/*
* Create and optionally map 'len' bytes of memory at guest
* physical address 'gpa'
*/
bzero(&seg, sizeof(seg));
seg.gpa = gpa;
seg.len = len;
error = ioctl(ctx->fd, VM_MAP_MEMORY, &seg);
if (error == 0 && addr != NULL) {
*addr = mmap(NULL, len, PROT_READ | PROT_WRITE, MAP_SHARED,
ctx->fd, gpa);
}
return (error);
}
int
vm_setup_memory(struct vmctx *ctx, size_t memsize, enum vm_mmap_style vms)
{
char **addr;
int error;
/* XXX VM_MMAP_SPARSE not implemented yet */
assert(vms == VM_MMAP_NONE || vms == VM_MMAP_ALL);
ctx->vms = vms;
/*
* If 'memsize' cannot fit entirely in the 'lowmem' segment then
* create another 'highmem' segment above 4GB for the remainder.
*/
if (memsize > ctx->lowmem_limit) {
ctx->lowmem = ctx->lowmem_limit;
ctx->highmem = memsize - ctx->lowmem;
} else {
ctx->lowmem = memsize;
ctx->highmem = 0;
}
if (ctx->lowmem > 0) {
addr = (vms == VM_MMAP_ALL) ? &ctx->lowmem_addr : NULL;
error = setup_memory_segment(ctx, 0, ctx->lowmem, addr);
if (error)
return (error);
}
if (ctx->highmem > 0) {
addr = (vms == VM_MMAP_ALL) ? &ctx->highmem_addr : NULL;
error = setup_memory_segment(ctx, 4*GB, ctx->highmem, addr);
if (error)
return (error);
}
return (0);
}
void *
vm_map_gpa(struct vmctx *ctx, vm_paddr_t gaddr, size_t len)
{
/* XXX VM_MMAP_SPARSE not implemented yet */
assert(ctx->vms == VM_MMAP_ALL);
if (gaddr < ctx->lowmem && gaddr + len <= ctx->lowmem)
return ((void *)(ctx->lowmem_addr + gaddr));
if (gaddr >= 4*GB) {
gaddr -= 4*GB;
if (gaddr < ctx->highmem && gaddr + len <= ctx->highmem)
return ((void *)(ctx->highmem_addr + gaddr));
}
return (NULL);
}
int
vm_set_desc(struct vmctx *ctx, int vcpu, int reg,
uint64_t base, uint32_t limit, uint32_t access)
{
int error;
struct vm_seg_desc vmsegdesc;
bzero(&vmsegdesc, sizeof(vmsegdesc));
vmsegdesc.cpuid = vcpu;
vmsegdesc.regnum = reg;
vmsegdesc.desc.base = base;
vmsegdesc.desc.limit = limit;
vmsegdesc.desc.access = access;
error = ioctl(ctx->fd, VM_SET_SEGMENT_DESCRIPTOR, &vmsegdesc);
return (error);
}
int
vm_get_desc(struct vmctx *ctx, int vcpu, int reg,
uint64_t *base, uint32_t *limit, uint32_t *access)
{
int error;
struct vm_seg_desc vmsegdesc;
bzero(&vmsegdesc, sizeof(vmsegdesc));
vmsegdesc.cpuid = vcpu;
vmsegdesc.regnum = reg;
error = ioctl(ctx->fd, VM_GET_SEGMENT_DESCRIPTOR, &vmsegdesc);
if (error == 0) {
*base = vmsegdesc.desc.base;
*limit = vmsegdesc.desc.limit;
*access = vmsegdesc.desc.access;
}
return (error);
}
int
vm_set_register(struct vmctx *ctx, int vcpu, int reg, uint64_t val)
{
int error;
struct vm_register vmreg;
bzero(&vmreg, sizeof(vmreg));
vmreg.cpuid = vcpu;
vmreg.regnum = reg;
vmreg.regval = val;
error = ioctl(ctx->fd, VM_SET_REGISTER, &vmreg);
return (error);
}
int
vm_get_register(struct vmctx *ctx, int vcpu, int reg, uint64_t *ret_val)
{
int error;
struct vm_register vmreg;
bzero(&vmreg, sizeof(vmreg));
vmreg.cpuid = vcpu;
vmreg.regnum = reg;
error = ioctl(ctx->fd, VM_GET_REGISTER, &vmreg);
*ret_val = vmreg.regval;
return (error);
}
int
vm_run(struct vmctx *ctx, int vcpu, uint64_t rip, struct vm_exit *vmexit)
{
int error;
struct vm_run vmrun;
bzero(&vmrun, sizeof(vmrun));
vmrun.cpuid = vcpu;
vmrun.rip = rip;
error = ioctl(ctx->fd, VM_RUN, &vmrun);
bcopy(&vmrun.vm_exit, vmexit, sizeof(struct vm_exit));
return (error);
}
static int
vm_inject_exception_real(struct vmctx *ctx, int vcpu, int vector,
int error_code, int error_code_valid)
{
struct vm_exception exc;
bzero(&exc, sizeof(exc));
exc.cpuid = vcpu;
exc.vector = vector;
exc.error_code = error_code;
exc.error_code_valid = error_code_valid;
return (ioctl(ctx->fd, VM_INJECT_EXCEPTION, &exc));
}
int
vm_inject_exception(struct vmctx *ctx, int vcpu, int vector)
{
return (vm_inject_exception_real(ctx, vcpu, vector, 0, 0));
}
int
vm_inject_exception2(struct vmctx *ctx, int vcpu, int vector, int errcode)
{
return (vm_inject_exception_real(ctx, vcpu, vector, errcode, 1));
}
int
vm_apicid2vcpu(struct vmctx *ctx, int apicid)
{
/*
* The apic id associated with the 'vcpu' has the same numerical value
* as the 'vcpu' itself.
*/
return (apicid);
}
int
vm_lapic_irq(struct vmctx *ctx, int vcpu, int vector)
{
struct vm_lapic_irq vmirq;
bzero(&vmirq, sizeof(vmirq));
vmirq.cpuid = vcpu;
vmirq.vector = vector;
return (ioctl(ctx->fd, VM_LAPIC_IRQ, &vmirq));
}
int
vm_lapic_local_irq(struct vmctx *ctx, int vcpu, int vector)
{
struct vm_lapic_irq vmirq;
bzero(&vmirq, sizeof(vmirq));
vmirq.cpuid = vcpu;
vmirq.vector = vector;
return (ioctl(ctx->fd, VM_LAPIC_LOCAL_IRQ, &vmirq));
}
int
vm_lapic_msi(struct vmctx *ctx, uint64_t addr, uint64_t msg)
{
struct vm_lapic_msi vmmsi;
bzero(&vmmsi, sizeof(vmmsi));
vmmsi.addr = addr;
vmmsi.msg = msg;
return (ioctl(ctx->fd, VM_LAPIC_MSI, &vmmsi));
}
int
vm_ioapic_assert_irq(struct vmctx *ctx, int irq)
{
struct vm_ioapic_irq ioapic_irq;
bzero(&ioapic_irq, sizeof(struct vm_ioapic_irq));
ioapic_irq.irq = irq;
return (ioctl(ctx->fd, VM_IOAPIC_ASSERT_IRQ, &ioapic_irq));
}
int
vm_ioapic_deassert_irq(struct vmctx *ctx, int irq)
{
struct vm_ioapic_irq ioapic_irq;
bzero(&ioapic_irq, sizeof(struct vm_ioapic_irq));
ioapic_irq.irq = irq;
return (ioctl(ctx->fd, VM_IOAPIC_DEASSERT_IRQ, &ioapic_irq));
}
int
vm_ioapic_pulse_irq(struct vmctx *ctx, int irq)
{
struct vm_ioapic_irq ioapic_irq;
bzero(&ioapic_irq, sizeof(struct vm_ioapic_irq));
ioapic_irq.irq = irq;
return (ioctl(ctx->fd, VM_IOAPIC_PULSE_IRQ, &ioapic_irq));
}
int
vm_ioapic_pincount(struct vmctx *ctx, int *pincount)
{
return (ioctl(ctx->fd, VM_IOAPIC_PINCOUNT, pincount));
}
int
vm_inject_nmi(struct vmctx *ctx, int vcpu)
{
struct vm_nmi vmnmi;
bzero(&vmnmi, sizeof(vmnmi));
vmnmi.cpuid = vcpu;
return (ioctl(ctx->fd, VM_INJECT_NMI, &vmnmi));
}
static struct {
const char *name;
int type;
} capstrmap[] = {
{ "hlt_exit", VM_CAP_HALT_EXIT },
{ "mtrap_exit", VM_CAP_MTRAP_EXIT },
{ "pause_exit", VM_CAP_PAUSE_EXIT },
{ "unrestricted_guest", VM_CAP_UNRESTRICTED_GUEST },
{ "enable_invpcid", VM_CAP_ENABLE_INVPCID },
{ 0 }
};
int
vm_capability_name2type(const char *capname)
{
int i;
for (i = 0; capstrmap[i].name != NULL && capname != NULL; i++) {
if (strcmp(capstrmap[i].name, capname) == 0)
return (capstrmap[i].type);
}
return (-1);
}
const char *
vm_capability_type2name(int type)
{
int i;
for (i = 0; capstrmap[i].name != NULL; i++) {
if (capstrmap[i].type == type)
return (capstrmap[i].name);
}
return (NULL);
}
int
vm_get_capability(struct vmctx *ctx, int vcpu, enum vm_cap_type cap,
int *retval)
{
int error;
struct vm_capability vmcap;
bzero(&vmcap, sizeof(vmcap));
vmcap.cpuid = vcpu;
vmcap.captype = cap;
error = ioctl(ctx->fd, VM_GET_CAPABILITY, &vmcap);
*retval = vmcap.capval;
return (error);
}
int
vm_set_capability(struct vmctx *ctx, int vcpu, enum vm_cap_type cap, int val)
{
struct vm_capability vmcap;
bzero(&vmcap, sizeof(vmcap));
vmcap.cpuid = vcpu;
vmcap.captype = cap;
vmcap.capval = val;
return (ioctl(ctx->fd, VM_SET_CAPABILITY, &vmcap));
}
int
vm_assign_pptdev(struct vmctx *ctx, int bus, int slot, int func)
{
struct vm_pptdev pptdev;
bzero(&pptdev, sizeof(pptdev));
pptdev.bus = bus;
pptdev.slot = slot;
pptdev.func = func;
return (ioctl(ctx->fd, VM_BIND_PPTDEV, &pptdev));
}
int
vm_unassign_pptdev(struct vmctx *ctx, int bus, int slot, int func)
{
struct vm_pptdev pptdev;
bzero(&pptdev, sizeof(pptdev));
pptdev.bus = bus;
pptdev.slot = slot;
pptdev.func = func;
return (ioctl(ctx->fd, VM_UNBIND_PPTDEV, &pptdev));
}
int
vm_map_pptdev_mmio(struct vmctx *ctx, int bus, int slot, int func,
vm_paddr_t gpa, size_t len, vm_paddr_t hpa)
{
struct vm_pptdev_mmio pptmmio;
bzero(&pptmmio, sizeof(pptmmio));
pptmmio.bus = bus;
pptmmio.slot = slot;
pptmmio.func = func;
pptmmio.gpa = gpa;
pptmmio.len = len;
pptmmio.hpa = hpa;
return (ioctl(ctx->fd, VM_MAP_PPTDEV_MMIO, &pptmmio));
}
int
vm_setup_pptdev_msi(struct vmctx *ctx, int vcpu, int bus, int slot, int func,
uint64_t addr, uint64_t msg, int numvec)
{
struct vm_pptdev_msi pptmsi;
bzero(&pptmsi, sizeof(pptmsi));
pptmsi.vcpu = vcpu;
pptmsi.bus = bus;
pptmsi.slot = slot;
pptmsi.func = func;
pptmsi.msg = msg;
pptmsi.addr = addr;
pptmsi.numvec = numvec;
return (ioctl(ctx->fd, VM_PPTDEV_MSI, &pptmsi));
}
int
vm_setup_pptdev_msix(struct vmctx *ctx, int vcpu, int bus, int slot, int func,
int idx, uint64_t addr, uint64_t msg, uint32_t vector_control)
{
struct vm_pptdev_msix pptmsix;
bzero(&pptmsix, sizeof(pptmsix));
pptmsix.vcpu = vcpu;
pptmsix.bus = bus;
pptmsix.slot = slot;
pptmsix.func = func;
pptmsix.idx = idx;
pptmsix.msg = msg;
pptmsix.addr = addr;
pptmsix.vector_control = vector_control;
return ioctl(ctx->fd, VM_PPTDEV_MSIX, &pptmsix);
}
uint64_t *
vm_get_stats(struct vmctx *ctx, int vcpu, struct timeval *ret_tv,
int *ret_entries)
{
int error;
static struct vm_stats vmstats;
vmstats.cpuid = vcpu;
error = ioctl(ctx->fd, VM_STATS, &vmstats);
if (error == 0) {
if (ret_entries)
*ret_entries = vmstats.num_entries;
if (ret_tv)
*ret_tv = vmstats.tv;
return (vmstats.statbuf);
} else
return (NULL);
}
const char *
vm_get_stat_desc(struct vmctx *ctx, int index)
{
static struct vm_stat_desc statdesc;
statdesc.index = index;
if (ioctl(ctx->fd, VM_STAT_DESC, &statdesc) == 0)
return (statdesc.desc);
else
return (NULL);
}
int
vm_get_x2apic_state(struct vmctx *ctx, int vcpu, enum x2apic_state *state)
{
int error;
struct vm_x2apic x2apic;
bzero(&x2apic, sizeof(x2apic));
x2apic.cpuid = vcpu;
error = ioctl(ctx->fd, VM_GET_X2APIC_STATE, &x2apic);
*state = x2apic.state;
return (error);
}
int
vm_set_x2apic_state(struct vmctx *ctx, int vcpu, enum x2apic_state state)
{
int error;
struct vm_x2apic x2apic;
bzero(&x2apic, sizeof(x2apic));
x2apic.cpuid = vcpu;
x2apic.state = state;
error = ioctl(ctx->fd, VM_SET_X2APIC_STATE, &x2apic);
return (error);
}
/*
* From Intel Vol 3a:
* Table 9-1. IA-32 Processor States Following Power-up, Reset or INIT
*/
int
vcpu_reset(struct vmctx *vmctx, int vcpu)
{
int error;
uint64_t rflags, rip, cr0, cr4, zero, desc_base, rdx;
uint32_t desc_access, desc_limit;
uint16_t sel;
zero = 0;
rflags = 0x2;
error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RFLAGS, rflags);
if (error)
goto done;
rip = 0xfff0;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RIP, rip)) != 0)
goto done;
cr0 = CR0_NE;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_CR0, cr0)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_CR3, zero)) != 0)
goto done;
cr4 = 0;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_CR4, cr4)) != 0)
goto done;
/*
* CS: present, r/w, accessed, 16-bit, byte granularity, usable
*/
desc_base = 0xffff0000;
desc_limit = 0xffff;
desc_access = 0x0093;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_CS,
desc_base, desc_limit, desc_access);
if (error)
goto done;
sel = 0xf000;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_CS, sel)) != 0)
goto done;
/*
* SS,DS,ES,FS,GS: present, r/w, accessed, 16-bit, byte granularity
*/
desc_base = 0;
desc_limit = 0xffff;
desc_access = 0x0093;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_SS,
desc_base, desc_limit, desc_access);
if (error)
goto done;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_DS,
desc_base, desc_limit, desc_access);
if (error)
goto done;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_ES,
desc_base, desc_limit, desc_access);
if (error)
goto done;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_FS,
desc_base, desc_limit, desc_access);
if (error)
goto done;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_GS,
desc_base, desc_limit, desc_access);
if (error)
goto done;
sel = 0;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_SS, sel)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_DS, sel)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_ES, sel)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_FS, sel)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_GS, sel)) != 0)
goto done;
/* General purpose registers */
rdx = 0xf00;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RAX, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RBX, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RCX, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RDX, rdx)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RSI, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RDI, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RBP, zero)) != 0)
goto done;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_RSP, zero)) != 0)
goto done;
/* GDTR, IDTR */
desc_base = 0;
desc_limit = 0xffff;
desc_access = 0;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_GDTR,
desc_base, desc_limit, desc_access);
if (error != 0)
goto done;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_IDTR,
desc_base, desc_limit, desc_access);
if (error != 0)
goto done;
/* TR */
desc_base = 0;
desc_limit = 0xffff;
desc_access = 0x0000008b;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_TR, 0, 0, desc_access);
if (error)
goto done;
sel = 0;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_TR, sel)) != 0)
goto done;
/* LDTR */
desc_base = 0;
desc_limit = 0xffff;
desc_access = 0x00000082;
error = vm_set_desc(vmctx, vcpu, VM_REG_GUEST_LDTR, desc_base,
desc_limit, desc_access);
if (error)
goto done;
sel = 0;
if ((error = vm_set_register(vmctx, vcpu, VM_REG_GUEST_LDTR, 0)) != 0)
goto done;
/* XXX cr2, debug registers */
error = 0;
done:
return (error);
}
Merge projects/bhyve_npt_pmap into head. Make the amd64/pmap code aware of nested page table mappings used by bhyve guests. This allows bhyve to associate each guest with its own vmspace and deal with nested page faults in the context of that vmspace. This also enables features like accessed/dirty bit tracking, swapping to disk and transparent superpage promotions of guest memory. Guest vmspace: Each bhyve guest has a unique vmspace to represent the physical memory allocated to the guest. Each memory segment allocated by the guest is mapped into the guest's address space via the 'vmspace->vm_map' and is backed by an object of type OBJT_DEFAULT. pmap types: The amd64/pmap now understands two types of pmaps: PT_X86 and PT_EPT. The PT_X86 pmap type is used by the vmspace associated with the host kernel as well as user processes executing on the host. The PT_EPT pmap is used by the vmspace associated with a bhyve guest. Page Table Entries: The EPT page table entries as mostly similar in functionality to regular page table entries although there are some differences in terms of what bits are used to express that functionality. For e.g. the dirty bit is represented by bit 9 in the nested PTE as opposed to bit 6 in the regular x86 PTE. Therefore the bitmask representing the dirty bit is now computed at runtime based on the type of the pmap. Thus PG_M that was previously a macro now becomes a local variable that is initialized at runtime using 'pmap_modified_bit(pmap)'. An additional wrinkle associated with EPT mappings is that older Intel processors don't have hardware support for tracking accessed/dirty bits in the PTE. This means that the amd64/pmap code needs to emulate these bits to provide proper accounting to the VM subsystem. This is achieved by using the following mapping for EPT entries that need emulation of A/D bits: Bit Position Interpreted By PG_V 52 software (accessed bit emulation handler) PG_RW 53 software (dirty bit emulation handler) PG_A 0 hardware (aka EPT_PG_RD) PG_M 1 hardware (aka EPT_PG_WR) The idea to use the mapping listed above for A/D bit emulation came from Alan Cox (alc@). The final difference with respect to x86 PTEs is that some EPT implementations do not support superpage mappings. This is recorded in the 'pm_flags' field of the pmap. TLB invalidation: The amd64/pmap code has a number of ways to do invalidation of mappings that may be cached in the TLB: single page, multiple pages in a range or the entire TLB. All of these funnel into a single EPT invalidation routine called 'pmap_invalidate_ept()'. This routine bumps up the EPT generation number and sends an IPI to the host cpus that are executing the guest's vcpus. On a subsequent entry into the guest it will detect that the EPT has changed and invalidate the mappings from the TLB. Guest memory access: Since the guest memory is no longer wired we need to hold the host physical page that backs the guest physical page before we can access it. The helper functions 'vm_gpa_hold()/vm_gpa_release()' are available for this purpose. PCI passthru: Guest's with PCI passthru devices will wire the entire guest physical address space. The MMIO BAR associated with the passthru device is backed by a vm_object of type OBJT_SG. An IOMMU domain is created only for guest's that have one or more PCI passthru devices attached to them. Limitations: There isn't a way to map a guest physical page without execute permissions. This is because the amd64/pmap code interprets the guest physical mappings as user mappings since they are numerically below VM_MAXUSER_ADDRESS. Since PG_U shares the same bit position as EPT_PG_EXECUTE all guest mappings become automatically executable. Thanks to Alan Cox and Konstantin Belousov for their rigorous code reviews as well as their support and encouragement. Thanks for John Baldwin for reviewing the use of OBJT_SG as the backing object for pci passthru mmio regions. Special thanks to Peter Holm for testing the patch on short notice. Approved by: re Discussed with: grehan Reviewed by: alc, kib Tested by: pho
2013-10-05 21:22:35 +00:00
int
vm_get_gpa_pmap(struct vmctx *ctx, uint64_t gpa, uint64_t *pte, int *num)
{
int error, i;
struct vm_gpa_pte gpapte;
bzero(&gpapte, sizeof(gpapte));
gpapte.gpa = gpa;
error = ioctl(ctx->fd, VM_GET_GPA_PMAP, &gpapte);
if (error == 0) {
*num = gpapte.ptenum;
for (i = 0; i < gpapte.ptenum; i++)
pte[i] = gpapte.pte[i];
}
return (error);
}
int
vm_get_hpet_capabilities(struct vmctx *ctx, uint32_t *capabilities)
{
int error;
struct vm_hpet_cap cap;
bzero(&cap, sizeof(struct vm_hpet_cap));
error = ioctl(ctx->fd, VM_GET_HPET_CAPABILITIES, &cap);
if (capabilities != NULL)
*capabilities = cap.capabilities;
return (error);
}