freebsd-dev/usr.sbin/bhyveload/bhyveload.c
Neel Natu 318224bbe6 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

640 lines
12 KiB
C

/*-
* 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.
*
* $FreeBSD$
*/
/*-
* Copyright (c) 2011 Google, 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 THE AUTHOR AND CONTRIBUTORS ``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 THE AUTHOR 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/ioctl.h>
#include <sys/stat.h>
#include <sys/disk.h>
#include <machine/specialreg.h>
#include <machine/vmm.h>
#include <dirent.h>
#include <dlfcn.h>
#include <errno.h>
#include <fcntl.h>
#include <getopt.h>
#include <limits.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <termios.h>
#include <unistd.h>
#include <vmmapi.h>
#include "userboot.h"
#define MB (1024 * 1024UL)
#define GB (1024 * 1024 * 1024UL)
#define BSP 0
static char *host_base = "/";
static struct termios term, oldterm;
static int disk_fd = -1;
static char *vmname, *progname;
static struct vmctx *ctx;
static uint64_t gdtbase, cr3, rsp;
static void cb_exit(void *arg, int v);
/*
* Console i/o callbacks
*/
static void
cb_putc(void *arg, int ch)
{
char c = ch;
write(1, &c, 1);
}
static int
cb_getc(void *arg)
{
char c;
if (read(0, &c, 1) == 1)
return (c);
return (-1);
}
static int
cb_poll(void *arg)
{
int n;
if (ioctl(0, FIONREAD, &n) >= 0)
return (n > 0);
return (0);
}
/*
* Host filesystem i/o callbacks
*/
struct cb_file {
int cf_isdir;
size_t cf_size;
struct stat cf_stat;
union {
int fd;
DIR *dir;
} cf_u;
};
static int
cb_open(void *arg, const char *filename, void **hp)
{
struct stat st;
struct cb_file *cf;
char path[PATH_MAX];
if (!host_base)
return (ENOENT);
strlcpy(path, host_base, PATH_MAX);
if (path[strlen(path) - 1] == '/')
path[strlen(path) - 1] = 0;
strlcat(path, filename, PATH_MAX);
cf = malloc(sizeof(struct cb_file));
if (stat(path, &cf->cf_stat) < 0) {
free(cf);
return (errno);
}
cf->cf_size = st.st_size;
if (S_ISDIR(cf->cf_stat.st_mode)) {
cf->cf_isdir = 1;
cf->cf_u.dir = opendir(path);
if (!cf->cf_u.dir)
goto out;
*hp = cf;
return (0);
}
if (S_ISREG(cf->cf_stat.st_mode)) {
cf->cf_isdir = 0;
cf->cf_u.fd = open(path, O_RDONLY);
if (cf->cf_u.fd < 0)
goto out;
*hp = cf;
return (0);
}
out:
free(cf);
return (EINVAL);
}
static int
cb_close(void *arg, void *h)
{
struct cb_file *cf = h;
if (cf->cf_isdir)
closedir(cf->cf_u.dir);
else
close(cf->cf_u.fd);
free(cf);
return (0);
}
static int
cb_isdir(void *arg, void *h)
{
struct cb_file *cf = h;
return (cf->cf_isdir);
}
static int
cb_read(void *arg, void *h, void *buf, size_t size, size_t *resid)
{
struct cb_file *cf = h;
ssize_t sz;
if (cf->cf_isdir)
return (EINVAL);
sz = read(cf->cf_u.fd, buf, size);
if (sz < 0)
return (EINVAL);
*resid = size - sz;
return (0);
}
static int
cb_readdir(void *arg, void *h, uint32_t *fileno_return, uint8_t *type_return,
size_t *namelen_return, char *name)
{
struct cb_file *cf = h;
struct dirent *dp;
if (!cf->cf_isdir)
return (EINVAL);
dp = readdir(cf->cf_u.dir);
if (!dp)
return (ENOENT);
/*
* Note: d_namlen is in the range 0..255 and therefore less
* than PATH_MAX so we don't need to test before copying.
*/
*fileno_return = dp->d_fileno;
*type_return = dp->d_type;
*namelen_return = dp->d_namlen;
memcpy(name, dp->d_name, dp->d_namlen);
name[dp->d_namlen] = 0;
return (0);
}
static int
cb_seek(void *arg, void *h, uint64_t offset, int whence)
{
struct cb_file *cf = h;
if (cf->cf_isdir)
return (EINVAL);
if (lseek(cf->cf_u.fd, offset, whence) < 0)
return (errno);
return (0);
}
static int
cb_stat(void *arg, void *h, int *mode, int *uid, int *gid, uint64_t *size)
{
struct cb_file *cf = h;
*mode = cf->cf_stat.st_mode;
*uid = cf->cf_stat.st_uid;
*gid = cf->cf_stat.st_gid;
*size = cf->cf_stat.st_size;
return (0);
}
/*
* Disk image i/o callbacks
*/
static int
cb_diskread(void *arg, int unit, uint64_t from, void *to, size_t size,
size_t *resid)
{
ssize_t n;
if (unit != 0 || disk_fd == -1)
return (EIO);
n = pread(disk_fd, to, size, from);
if (n < 0)
return (errno);
*resid = size - n;
return (0);
}
static int
cb_diskioctl(void *arg, int unit, u_long cmd, void *data)
{
struct stat sb;
if (unit != 0 || disk_fd == -1)
return (EBADF);
switch (cmd) {
case DIOCGSECTORSIZE:
*(u_int *)data = 512;
break;
case DIOCGMEDIASIZE:
if (fstat(disk_fd, &sb) == 0)
*(off_t *)data = sb.st_size;
else
return (ENOTTY);
break;
default:
return (ENOTTY);
}
return (0);
}
/*
* Guest virtual machine i/o callbacks
*/
static int
cb_copyin(void *arg, const void *from, uint64_t to, size_t size)
{
char *ptr;
to &= 0x7fffffff;
ptr = vm_map_gpa(ctx, to, size);
if (ptr == NULL)
return (EFAULT);
memcpy(ptr, from, size);
return (0);
}
static int
cb_copyout(void *arg, uint64_t from, void *to, size_t size)
{
char *ptr;
from &= 0x7fffffff;
ptr = vm_map_gpa(ctx, from, size);
if (ptr == NULL)
return (EFAULT);
memcpy(to, ptr, size);
return (0);
}
static void
cb_setreg(void *arg, int r, uint64_t v)
{
int error;
enum vm_reg_name vmreg;
vmreg = VM_REG_LAST;
switch (r) {
case 4:
vmreg = VM_REG_GUEST_RSP;
rsp = v;
break;
default:
break;
}
if (vmreg == VM_REG_LAST) {
printf("test_setreg(%d): not implemented\n", r);
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
error = vm_set_register(ctx, BSP, vmreg, v);
if (error) {
perror("vm_set_register");
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
}
static void
cb_setmsr(void *arg, int r, uint64_t v)
{
int error;
enum vm_reg_name vmreg;
vmreg = VM_REG_LAST;
switch (r) {
case MSR_EFER:
vmreg = VM_REG_GUEST_EFER;
break;
default:
break;
}
if (vmreg == VM_REG_LAST) {
printf("test_setmsr(%d): not implemented\n", r);
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
error = vm_set_register(ctx, BSP, vmreg, v);
if (error) {
perror("vm_set_msr");
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
}
static void
cb_setcr(void *arg, int r, uint64_t v)
{
int error;
enum vm_reg_name vmreg;
vmreg = VM_REG_LAST;
switch (r) {
case 0:
vmreg = VM_REG_GUEST_CR0;
break;
case 3:
vmreg = VM_REG_GUEST_CR3;
cr3 = v;
break;
case 4:
vmreg = VM_REG_GUEST_CR4;
break;
default:
break;
}
if (vmreg == VM_REG_LAST) {
printf("test_setcr(%d): not implemented\n", r);
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
error = vm_set_register(ctx, BSP, vmreg, v);
if (error) {
perror("vm_set_cr");
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
}
static void
cb_setgdt(void *arg, uint64_t base, size_t size)
{
int error;
error = vm_set_desc(ctx, BSP, VM_REG_GUEST_GDTR, base, size - 1, 0);
if (error != 0) {
perror("vm_set_desc(gdt)");
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
gdtbase = base;
}
static void
cb_exec(void *arg, uint64_t rip)
{
int error;
error = vm_setup_freebsd_registers(ctx, BSP, rip, cr3, gdtbase, rsp);
if (error) {
perror("vm_setup_freebsd_registers");
cb_exit(NULL, USERBOOT_EXIT_QUIT);
}
cb_exit(NULL, 0);
}
/*
* Misc
*/
static void
cb_delay(void *arg, int usec)
{
usleep(usec);
}
static void
cb_exit(void *arg, int v)
{
tcsetattr(0, TCSAFLUSH, &oldterm);
exit(v);
}
static void
cb_getmem(void *arg, uint64_t *ret_lowmem, uint64_t *ret_highmem)
{
vm_get_memory_seg(ctx, 0, ret_lowmem, NULL);
vm_get_memory_seg(ctx, 4 * GB, ret_highmem, NULL);
}
static const char *
cb_getenv(void *arg, int num)
{
int max;
static const char * var[] = {
"smbios.bios.vendor=BHYVE",
"boot_serial=1",
NULL
};
max = sizeof(var) / sizeof(var[0]);
if (num < max)
return (var[num]);
else
return (NULL);
}
static struct loader_callbacks cb = {
.getc = cb_getc,
.putc = cb_putc,
.poll = cb_poll,
.open = cb_open,
.close = cb_close,
.isdir = cb_isdir,
.read = cb_read,
.readdir = cb_readdir,
.seek = cb_seek,
.stat = cb_stat,
.diskread = cb_diskread,
.diskioctl = cb_diskioctl,
.copyin = cb_copyin,
.copyout = cb_copyout,
.setreg = cb_setreg,
.setmsr = cb_setmsr,
.setcr = cb_setcr,
.setgdt = cb_setgdt,
.exec = cb_exec,
.delay = cb_delay,
.exit = cb_exit,
.getmem = cb_getmem,
.getenv = cb_getenv,
};
static void
usage(void)
{
fprintf(stderr,
"usage: %s [-m mem-size][-d <disk-path>] [-h <host-path>] "
"<vmname>\n", progname);
exit(1);
}
int
main(int argc, char** argv)
{
void *h;
void (*func)(struct loader_callbacks *, void *, int, int);
uint64_t mem_size;
int opt, error;
char *disk_image;
progname = argv[0];
mem_size = 256 * MB;
disk_image = NULL;
while ((opt = getopt(argc, argv, "d:h:m:")) != -1) {
switch (opt) {
case 'd':
disk_image = optarg;
break;
case 'h':
host_base = optarg;
break;
case 'm':
mem_size = strtoul(optarg, NULL, 0) * MB;
break;
case '?':
usage();
}
}
argc -= optind;
argv += optind;
if (argc != 1)
usage();
vmname = argv[0];
error = vm_create(vmname);
if (error != 0 && errno != EEXIST) {
perror("vm_create");
exit(1);
}
ctx = vm_open(vmname);
if (ctx == NULL) {
perror("vm_open");
exit(1);
}
error = vm_setup_memory(ctx, mem_size, VM_MMAP_ALL);
if (error) {
perror("vm_setup_memory");
exit(1);
}
tcgetattr(0, &term);
oldterm = term;
term.c_lflag &= ~(ICANON|ECHO);
term.c_iflag &= ~ICRNL;
tcsetattr(0, TCSAFLUSH, &term);
h = dlopen("/boot/userboot.so", RTLD_LOCAL);
if (!h) {
printf("%s\n", dlerror());
return (1);
}
func = dlsym(h, "loader_main");
if (!func) {
printf("%s\n", dlerror());
return (1);
}
if (disk_image) {
disk_fd = open(disk_image, O_RDONLY);
}
func(&cb, NULL, USERBOOT_VERSION_3, disk_fd >= 0);
}