xiaomi-sm8450/android_kernel_xiaomi_sm8450
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm

Browse Source 
Pull kvm updates from Paolo Bonzini:
 "ARM:
   - VHE optimizations

   - EL2 address space randomization

   - speculative execution mitigations ("variant 3a", aka execution past
     invalid privilege register access)

   - bugfixes and cleanups

  PPC:
   - improvements for the radix page fault handler for HV KVM on POWER9

  s390:
   - more kvm stat counters

   - virtio gpu plumbing

   - documentation

   - facilities improvements

  x86:
   - support for VMware magic I/O port and pseudo-PMCs

   - AMD pause loop exiting

   - support for AMD core performance extensions

   - support for synchronous register access

   - expose nVMX capabilities to userspace

   - support for Hyper-V signaling via eventfd

   - use Enlightened VMCS when running on Hyper-V

   - allow userspace to disable MWAIT/HLT/PAUSE vmexits

   - usual roundup of optimizations and nested virtualization bugfixes

  Generic:
   - API selftest infrastructure (though the only tests are for x86 as
     of now)"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (174 commits)
  kvm: x86: fix a prototype warning
  kvm: selftests: add sync_regs_test
  kvm: selftests: add API testing infrastructure
  kvm: x86: fix a compile warning
  KVM: X86: Add Force Emulation Prefix for "emulate the next instruction"
  KVM: X86: Introduce handle_ud()
  KVM: vmx: unify adjacent #ifdefs
  x86: kvm: hide the unused 'cpu' variable
  KVM: VMX: remove bogus WARN_ON in handle_ept_misconfig
  Revert "KVM: X86: Fix SMRAM accessing even if VM is shutdown"
  kvm: Add emulation for movups/movupd
  KVM: VMX: raise internal error for exception during invalid protected mode state
  KVM: nVMX: Optimization: Dont set KVM_REQ_EVENT when VMExit with nested_run_pending
  KVM: nVMX: Require immediate-exit when event reinjected to L2 and L1 event pending
  KVM: x86: Fix misleading comments on handling pending exceptions
  KVM: x86: Rename interrupt.pending to interrupt.injected
  KVM: VMX: No need to clear pending NMI/interrupt on inject realmode interrupt
  x86/kvm: use Enlightened VMCS when running on Hyper-V
  x86/hyper-v: detect nested features
  x86/hyper-v: define struct hv_enlightened_vmcs and clean field bits
  ...
This commit is contained in:
Linus Torvalds
2018-04-09 11:42:31 -07:00
parent e9092d0d97 e01bca2fc6
commit d8312a3f61
150 changed files with 11930 additions and 2012 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
tools/testing/selftests/Makefile
View File

@@ -15,6 +15,7 @@ TARGETS += gpio
TARGETS += intel_pstate
TARGETS += ipc
TARGETS += kcmp
TARGETS += kvm
TARGETS += lib
TARGETS += membarrier
TARGETS += memfd

39
tools/testing/selftests/kvm/Makefile Normal file
View File

@@ -0,0 +1,39 @@
all:
top_srcdir = ../../../../
UNAME_M := $(shell uname -m)
LIBKVM = lib/assert.c lib/elf.c lib/io.c lib/kvm_util.c lib/sparsebit.c
LIBKVM_x86_64 = lib/x86.c
TEST_GEN_PROGS_x86_64 = set_sregs_test
TEST_GEN_PROGS_x86_64 += sync_regs_test
TEST_GEN_PROGS += $(TEST_GEN_PROGS_$(UNAME_M))
LIBKVM += $(LIBKVM_$(UNAME_M))
INSTALL_HDR_PATH = $(top_srcdir)/usr
LINUX_HDR_PATH = $(INSTALL_HDR_PATH)/include/
CFLAGS += -O2 -g -I$(LINUX_HDR_PATH) -Iinclude -I$(<D)
# After inclusion, $(OUTPUT) is defined and
# $(TEST_GEN_PROGS) starts with $(OUTPUT)/
include ../lib.mk
STATIC_LIBS := $(OUTPUT)/libkvm.a
LIBKVM_OBJ := $(patsubst %.c, $(OUTPUT)/%.o, $(LIBKVM))
EXTRA_CLEAN += $(LIBKVM_OBJ) $(STATIC_LIBS)
x := $(shell mkdir -p $(sort $(dir $(LIBKVM_OBJ))))
$(LIBKVM_OBJ): $(OUTPUT)/%.o: %.c
$(CC) $(CFLAGS) $(CPPFLAGS) $(TARGET_ARCH) -c $< -o $@
$(OUTPUT)/libkvm.a: $(LIBKVM_OBJ)
$(AR) crs $@ $^
$(LINUX_HDR_PATH):
make -C $(top_srcdir) headers_install
all: $(STATIC_LIBS) $(LINUX_HDR_PATH)
$(TEST_GEN_PROGS): $(STATIC_LIBS)
$(TEST_GEN_PROGS) $(LIBKVM_OBJ): | $(LINUX_HDR_PATH)

142
tools/testing/selftests/kvm/include/kvm_util.h Normal file
View File

@@ -0,0 +1,142 @@
/*
* tools/testing/selftests/kvm/include/kvm_util.h
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*
*/
#ifndef SELFTEST_KVM_UTIL_H
#define SELFTEST_KVM_UTIL_H 1
#include "test_util.h"
#include "asm/kvm.h"
#include "linux/kvm.h"
#include <sys/ioctl.h>
#include "sparsebit.h"
/*
* Memslots can't cover the gfn starting at this gpa otherwise vCPUs can't be
* created. Only applies to VMs using EPT.
*/
#define KVM_DEFAULT_IDENTITY_MAP_ADDRESS 0xfffbc000ul
/* Callers of kvm_util only have an incomplete/opaque description of the
* structure kvm_util is using to maintain the state of a VM.
*/
struct kvm_vm;
typedef uint64_t vm_paddr_t; /* Virtual Machine (Guest) physical address */
typedef uint64_t vm_vaddr_t; /* Virtual Machine (Guest) virtual address */
/* Minimum allocated guest virtual and physical addresses */
#define KVM_UTIL_MIN_VADDR 0x2000
#define DEFAULT_GUEST_PHY_PAGES 512
#define DEFAULT_GUEST_STACK_VADDR_MIN 0xab6000
#define DEFAULT_STACK_PGS 5
enum vm_guest_mode {
VM_MODE_FLAT48PG,
};
enum vm_mem_backing_src_type {
VM_MEM_SRC_ANONYMOUS,
VM_MEM_SRC_ANONYMOUS_THP,
VM_MEM_SRC_ANONYMOUS_HUGETLB,
};
int kvm_check_cap(long cap);
struct kvm_vm *vm_create(enum vm_guest_mode mode, uint64_t phy_pages, int perm);
void kvm_vm_free(struct kvm_vm *vmp);
int kvm_memcmp_hva_gva(void *hva,
struct kvm_vm *vm, const vm_vaddr_t gva, size_t len);
void kvm_vm_elf_load(struct kvm_vm *vm, const char *filename,
uint32_t data_memslot, uint32_t pgd_memslot);
void vm_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent);
void vcpu_dump(FILE *stream, struct kvm_vm *vm,
uint32_t vcpuid, uint8_t indent);
void vm_create_irqchip(struct kvm_vm *vm);
void vm_userspace_mem_region_add(struct kvm_vm *vm,
enum vm_mem_backing_src_type src_type,
uint64_t guest_paddr, uint32_t slot, uint64_t npages,
uint32_t flags);
void vcpu_ioctl(struct kvm_vm *vm,
uint32_t vcpuid, unsigned long ioctl, void *arg);
void vm_ioctl(struct kvm_vm *vm, unsigned long ioctl, void *arg);
void vm_mem_region_set_flags(struct kvm_vm *vm, uint32_t slot, uint32_t flags);
void vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpuid);
vm_vaddr_t vm_vaddr_alloc(struct kvm_vm *vm, size_t sz, vm_vaddr_t vaddr_min,
uint32_t data_memslot, uint32_t pgd_memslot);
void *addr_gpa2hva(struct kvm_vm *vm, vm_paddr_t gpa);
void *addr_gva2hva(struct kvm_vm *vm, vm_vaddr_t gva);
vm_paddr_t addr_hva2gpa(struct kvm_vm *vm, void *hva);
vm_paddr_t addr_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva);
struct kvm_run *vcpu_state(struct kvm_vm *vm, uint32_t vcpuid);
void vcpu_run(struct kvm_vm *vm, uint32_t vcpuid);
int _vcpu_run(struct kvm_vm *vm, uint32_t vcpuid);
void vcpu_set_mp_state(struct kvm_vm *vm, uint32_t vcpuid,
struct kvm_mp_state *mp_state);
void vcpu_regs_get(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_regs *regs);
void vcpu_regs_set(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_regs *regs);
void vcpu_args_set(struct kvm_vm *vm, uint32_t vcpuid, unsigned int num, ...);
void vcpu_sregs_get(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_sregs *sregs);
void vcpu_sregs_set(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_sregs *sregs);
int _vcpu_sregs_set(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_sregs *sregs);
void vcpu_events_get(struct kvm_vm *vm, uint32_t vcpuid,
struct kvm_vcpu_events *events);
void vcpu_events_set(struct kvm_vm *vm, uint32_t vcpuid,
struct kvm_vcpu_events *events);
const char *exit_reason_str(unsigned int exit_reason);
void virt_pgd_alloc(struct kvm_vm *vm, uint32_t pgd_memslot);
void virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
uint32_t pgd_memslot);
vm_paddr_t vm_phy_page_alloc(struct kvm_vm *vm,
vm_paddr_t paddr_min, uint32_t memslot);
void kvm_get_supported_cpuid(struct kvm_cpuid2 *cpuid);
void vcpu_set_cpuid(
struct kvm_vm *vm, uint32_t vcpuid, struct kvm_cpuid2 *cpuid);
struct kvm_cpuid2 *allocate_kvm_cpuid2(void);
struct kvm_cpuid_entry2 *
find_cpuid_index_entry(struct kvm_cpuid2 *cpuid, uint32_t function,
uint32_t index);
static inline struct kvm_cpuid_entry2 *
find_cpuid_entry(struct kvm_cpuid2 *cpuid, uint32_t function)
{
return find_cpuid_index_entry(cpuid, function, 0);
}
struct kvm_vm *vm_create_default(uint32_t vcpuid, void *guest_code);
void vm_vcpu_add_default(struct kvm_vm *vm, uint32_t vcpuid, void *guest_code);
struct kvm_userspace_memory_region *
kvm_userspace_memory_region_find(struct kvm_vm *vm, uint64_t start,
uint64_t end);
struct kvm_dirty_log *
allocate_kvm_dirty_log(struct kvm_userspace_memory_region *region);
int vm_create_device(struct kvm_vm *vm, struct kvm_create_device *cd);
#endif /* SELFTEST_KVM_UTIL_H */

75
tools/testing/selftests/kvm/include/sparsebit.h Normal file
View File

@@ -0,0 +1,75 @@
/*
* tools/testing/selftests/kvm/include/sparsebit.h
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*
*
* Header file that describes API to the sparsebit library.
* This library provides a memory efficient means of storing
* the settings of bits indexed via a uint64_t. Memory usage
* is reasonable, significantly less than (2^64 / 8) bytes, as
* long as bits that are mostly set or mostly cleared are close
* to each other. This library is efficient in memory usage
* even in the case where most bits are set.
*/
#ifndef _TEST_SPARSEBIT_H_
#define _TEST_SPARSEBIT_H_
#include <stdbool.h>
#include <stdint.h>
#include <stdio.h>
#ifdef __cplusplus
extern "C" {
#endif
struct sparsebit;
typedef uint64_t sparsebit_idx_t;
typedef uint64_t sparsebit_num_t;
struct sparsebit *sparsebit_alloc(void);
void sparsebit_free(struct sparsebit **sbitp);
void sparsebit_copy(struct sparsebit *dstp, struct sparsebit *src);
bool sparsebit_is_set(struct sparsebit *sbit, sparsebit_idx_t idx);
bool sparsebit_is_set_num(struct sparsebit *sbit,
sparsebit_idx_t idx, sparsebit_num_t num);
bool sparsebit_is_clear(struct sparsebit *sbit, sparsebit_idx_t idx);
bool sparsebit_is_clear_num(struct sparsebit *sbit,
sparsebit_idx_t idx, sparsebit_num_t num);
sparsebit_num_t sparsebit_num_set(struct sparsebit *sbit);
bool sparsebit_any_set(struct sparsebit *sbit);
bool sparsebit_any_clear(struct sparsebit *sbit);
bool sparsebit_all_set(struct sparsebit *sbit);
bool sparsebit_all_clear(struct sparsebit *sbit);
sparsebit_idx_t sparsebit_first_set(struct sparsebit *sbit);
sparsebit_idx_t sparsebit_first_clear(struct sparsebit *sbit);
sparsebit_idx_t sparsebit_next_set(struct sparsebit *sbit, sparsebit_idx_t prev);
sparsebit_idx_t sparsebit_next_clear(struct sparsebit *sbit, sparsebit_idx_t prev);
sparsebit_idx_t sparsebit_next_set_num(struct sparsebit *sbit,
sparsebit_idx_t start, sparsebit_num_t num);
sparsebit_idx_t sparsebit_next_clear_num(struct sparsebit *sbit,
sparsebit_idx_t start, sparsebit_num_t num);
void sparsebit_set(struct sparsebit *sbitp, sparsebit_idx_t idx);
void sparsebit_set_num(struct sparsebit *sbitp, sparsebit_idx_t start,
sparsebit_num_t num);
void sparsebit_set_all(struct sparsebit *sbitp);
void sparsebit_clear(struct sparsebit *sbitp, sparsebit_idx_t idx);
void sparsebit_clear_num(struct sparsebit *sbitp,
sparsebit_idx_t start, sparsebit_num_t num);
void sparsebit_clear_all(struct sparsebit *sbitp);
void sparsebit_dump(FILE *stream, struct sparsebit *sbit,
unsigned int indent);
void sparsebit_validate_internal(struct sparsebit *sbit);
#ifdef __cplusplus
}
#endif
#endif /* _TEST_SPARSEBIT_H_ */

45
tools/testing/selftests/kvm/include/test_util.h Normal file
View File

@@ -0,0 +1,45 @@
/*
* tools/testing/selftests/kvm/include/test_util.h
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*
*/
#ifndef TEST_UTIL_H
#define TEST_UTIL_H 1
#include <stdlib.h>
#include <stdarg.h>
#include <stdbool.h>
#include <stdio.h>
#include <string.h>
#include <inttypes.h>
#include <errno.h>
#include <unistd.h>
#include <fcntl.h>
ssize_t test_write(int fd, const void *buf, size_t count);
ssize_t test_read(int fd, void *buf, size_t count);
int test_seq_read(const char *path, char **bufp, size_t *sizep);
void test_assert(bool exp, const char *exp_str,
const char *file, unsigned int line, const char *fmt, ...);
#define ARRAY_SIZE(array) (sizeof(array) / sizeof((array)[0]))
#define TEST_ASSERT(e, fmt, ...) \
test_assert((e), #e, __FILE__, __LINE__, fmt, ##__VA_ARGS__)
#define ASSERT_EQ(a, b) do { \
typeof(a) __a = (a); \
typeof(b) __b = (b); \
TEST_ASSERT(__a == __b, \
"ASSERT_EQ(%s, %s) failed.\n" \
"\t%s is %#lx\n" \
"\t%s is %#lx", \
#a, #b, #a, (unsigned long) __a, #b, (unsigned long) __b); \
} while (0)
#endif /* TEST_UTIL_H */

1043
tools/testing/selftests/kvm/include/x86.h Normal file
View File

File diff suppressed because it is too large Load Diff

87
tools/testing/selftests/kvm/lib/assert.c Normal file
View File

@@ -0,0 +1,87 @@
/*
* tools/testing/selftests/kvm/lib/assert.c
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*/
#define _GNU_SOURCE /* for getline(3) and strchrnul(3)*/
#include "test_util.h"
#include <execinfo.h>
#include <sys/syscall.h>
/* Dumps the current stack trace to stderr. */
static void __attribute__((noinline)) test_dump_stack(void);
static void test_dump_stack(void)
{
/*
* Build and run this command:
*
* addr2line -s -e /proc/$PPID/exe -fpai {backtrace addresses} | \
* grep -v test_dump_stack | cat -n 1>&2
*
* Note that the spacing is different and there's no newline.
*/
size_t i;
size_t n = 20;
void *stack[n];
const char *addr2line = "addr2line -s -e /proc/$PPID/exe -fpai";
const char *pipeline = "|cat -n 1>&2";
char cmd[strlen(addr2line) + strlen(pipeline) +
/* N bytes per addr * 2 digits per byte + 1 space per addr: */
n * (((sizeof(void *)) * 2) + 1) +
/* Null terminator: */
1];
char *c;
n = backtrace(stack, n);
c = &cmd[0];
c += sprintf(c, "%s", addr2line);
/*
* Skip the first 3 frames: backtrace, test_dump_stack, and
* test_assert. We hope that backtrace isn't inlined and the other two
* we've declared noinline.
*/
for (i = 2; i < n; i++)
c += sprintf(c, " %lx", ((unsigned long) stack[i]) - 1);
c += sprintf(c, "%s", pipeline);
#pragma GCC diagnostic push
#pragma GCC diagnostic ignored "-Wunused-result"
system(cmd);
#pragma GCC diagnostic pop
}
static pid_t gettid(void)
{
return syscall(SYS_gettid);
}
void __attribute__((noinline))
test_assert(bool exp, const char *exp_str,
const char *file, unsigned int line, const char *fmt, ...)
{
va_list ap;
if (!(exp)) {
va_start(ap, fmt);
fprintf(stderr, "==== Test Assertion Failure ====\n"
" %s:%u: %s\n"
" pid=%d tid=%d\n",
file, line, exp_str, getpid(), gettid());
test_dump_stack();
if (fmt) {
fputs(" ", stderr);
vfprintf(stderr, fmt, ap);
fputs("\n", stderr);
}
va_end(ap);
exit(254);
}
return;
}

197
tools/testing/selftests/kvm/lib/elf.c Normal file
View File

@@ -0,0 +1,197 @@
/*
* tools/testing/selftests/kvm/lib/elf.c
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*/
#include "test_util.h"
#include <bits/endian.h>
#include <linux/elf.h>
#include "kvm_util.h"
#include "kvm_util_internal.h"
static void elfhdr_get(const char *filename, Elf64_Ehdr *hdrp)
{
off_t offset_rv;
/* Open the ELF file. */
int fd;
fd = open(filename, O_RDONLY);
TEST_ASSERT(fd >= 0, "Failed to open ELF file,\n"
" filename: %s\n"
" rv: %i errno: %i", filename, fd, errno);
/* Read in and validate ELF Identification Record.
* The ELF Identification record is the first 16 (EI_NIDENT) bytes
* of the ELF header, which is at the beginning of the ELF file.
* For now it is only safe to read the first EI_NIDENT bytes. Once
* read and validated, the value of e_ehsize can be used to determine
* the real size of the ELF header.
*/
unsigned char ident[EI_NIDENT];
test_read(fd, ident, sizeof(ident));
TEST_ASSERT((ident[EI_MAG0] == ELFMAG0) && (ident[EI_MAG1] == ELFMAG1)
&& (ident[EI_MAG2] == ELFMAG2) && (ident[EI_MAG3] == ELFMAG3),
"ELF MAGIC Mismatch,\n"
" filename: %s\n"
" ident[EI_MAG0 - EI_MAG3]: %02x %02x %02x %02x\n"
" Expected: %02x %02x %02x %02x",
filename,
ident[EI_MAG0], ident[EI_MAG1], ident[EI_MAG2], ident[EI_MAG3],
ELFMAG0, ELFMAG1, ELFMAG2, ELFMAG3);
TEST_ASSERT(ident[EI_CLASS] == ELFCLASS64,
"Current implementation only able to handle ELFCLASS64,\n"
" filename: %s\n"
" ident[EI_CLASS]: %02x\n"
" expected: %02x",
filename,
ident[EI_CLASS], ELFCLASS64);
TEST_ASSERT(((BYTE_ORDER == LITTLE_ENDIAN)
&& (ident[EI_DATA] == ELFDATA2LSB))
|| ((BYTE_ORDER == BIG_ENDIAN)
&& (ident[EI_DATA] == ELFDATA2MSB)), "Current "
"implementation only able to handle\n"
"cases where the host and ELF file endianness\n"
"is the same:\n"
" host BYTE_ORDER: %u\n"
" host LITTLE_ENDIAN: %u\n"
" host BIG_ENDIAN: %u\n"
" ident[EI_DATA]: %u\n"
" ELFDATA2LSB: %u\n"
" ELFDATA2MSB: %u",
BYTE_ORDER, LITTLE_ENDIAN, BIG_ENDIAN,
ident[EI_DATA], ELFDATA2LSB, ELFDATA2MSB);
TEST_ASSERT(ident[EI_VERSION] == EV_CURRENT,
"Current implementation only able to handle current "
"ELF version,\n"
" filename: %s\n"
" ident[EI_VERSION]: %02x\n"
" expected: %02x",
filename, ident[EI_VERSION], EV_CURRENT);
/* Read in the ELF header.
* With the ELF Identification portion of the ELF header
* validated, especially that the value at EI_VERSION is
* as expected, it is now safe to read the entire ELF header.
*/
offset_rv = lseek(fd, 0, SEEK_SET);
TEST_ASSERT(offset_rv == 0, "Seek to ELF header failed,\n"
" rv: %zi expected: %i", offset_rv, 0);
test_read(fd, hdrp, sizeof(*hdrp));
TEST_ASSERT(hdrp->e_phentsize == sizeof(Elf64_Phdr),
"Unexpected physical header size,\n"
" hdrp->e_phentsize: %x\n"
" expected: %zx",
hdrp->e_phentsize, sizeof(Elf64_Phdr));
TEST_ASSERT(hdrp->e_shentsize == sizeof(Elf64_Shdr),
"Unexpected section header size,\n"
" hdrp->e_shentsize: %x\n"
" expected: %zx",
hdrp->e_shentsize, sizeof(Elf64_Shdr));
}
/* VM ELF Load
*
* Input Args:
* filename - Path to ELF file
*
* Output Args: None
*
* Input/Output Args:
* vm - Pointer to opaque type that describes the VM.
*
* Return: None, TEST_ASSERT failures for all error conditions
*
* Loads the program image of the ELF file specified by filename,
* into the virtual address space of the VM pointed to by vm. On entry
* the VM needs to not be using any of the virtual address space used
* by the image and it needs to have sufficient available physical pages, to
* back the virtual pages used to load the image.
*/
void kvm_vm_elf_load(struct kvm_vm *vm, const char *filename,
uint32_t data_memslot, uint32_t pgd_memslot)
{
off_t offset, offset_rv;
Elf64_Ehdr hdr;
/* Open the ELF file. */
int fd;
fd = open(filename, O_RDONLY);
TEST_ASSERT(fd >= 0, "Failed to open ELF file,\n"
" filename: %s\n"
" rv: %i errno: %i", filename, fd, errno);
/* Read in the ELF header. */
elfhdr_get(filename, &hdr);
/* For each program header.
* The following ELF header members specify the location
* and size of the program headers:
*
* e_phoff - File offset to start of program headers
* e_phentsize - Size of each program header
* e_phnum - Number of program header entries
*/
for (unsigned int n1 = 0; n1 < hdr.e_phnum; n1++) {
/* Seek to the beginning of the program header. */
offset = hdr.e_phoff + (n1 * hdr.e_phentsize);
offset_rv = lseek(fd, offset, SEEK_SET);
TEST_ASSERT(offset_rv == offset,
"Failed to seek to begining of program header %u,\n"
" filename: %s\n"
" rv: %jd errno: %i",
n1, filename, (intmax_t) offset_rv, errno);
/* Read in the program header. */
Elf64_Phdr phdr;
test_read(fd, &phdr, sizeof(phdr));
/* Skip if this header doesn't describe a loadable segment. */
if (phdr.p_type != PT_LOAD)
continue;
/* Allocate memory for this segment within the VM. */
TEST_ASSERT(phdr.p_memsz > 0, "Unexpected loadable segment "
"memsize of 0,\n"
" phdr index: %u p_memsz: 0x%" PRIx64,
n1, (uint64_t) phdr.p_memsz);
vm_vaddr_t seg_vstart = phdr.p_vaddr;
seg_vstart &= ~(vm_vaddr_t)(vm->page_size - 1);
vm_vaddr_t seg_vend = phdr.p_vaddr + phdr.p_memsz - 1;
seg_vend |= vm->page_size - 1;
size_t seg_size = seg_vend - seg_vstart + 1;
vm_vaddr_t vaddr = vm_vaddr_alloc(vm, seg_size, seg_vstart,
data_memslot, pgd_memslot);
TEST_ASSERT(vaddr == seg_vstart, "Unable to allocate "
"virtual memory for segment at requested min addr,\n"
" segment idx: %u\n"
" seg_vstart: 0x%lx\n"
" vaddr: 0x%lx",
n1, seg_vstart, vaddr);
memset(addr_gva2hva(vm, vaddr), 0, seg_size);
/* TODO(lhuemill): Set permissions of each memory segment
* based on the least-significant 3 bits of phdr.p_flags.
*/
/* Load portion of initial state that is contained within
* the ELF file.
*/
if (phdr.p_filesz) {
offset_rv = lseek(fd, phdr.p_offset, SEEK_SET);
TEST_ASSERT(offset_rv == phdr.p_offset,
"Seek to program segment offset failed,\n"
" program header idx: %u errno: %i\n"
" offset_rv: 0x%jx\n"
" expected: 0x%jx\n",
n1, errno, (intmax_t) offset_rv,
(intmax_t) phdr.p_offset);
test_read(fd, addr_gva2hva(vm, phdr.p_vaddr),
phdr.p_filesz);
}
}
}

158
tools/testing/selftests/kvm/lib/io.c Normal file
View File

@@ -0,0 +1,158 @@
/*
* tools/testing/selftests/kvm/lib/io.c
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*/
#include "test_util.h"
/* Test Write
*
* A wrapper for write(2), that automatically handles the following
* special conditions:
*
* + Interrupted system call (EINTR)
* + Write of less than requested amount
* + Non-block return (EAGAIN)
*
* For each of the above, an additional write is performed to automatically
* continue writing the requested data.
* There are also many cases where write(2) can return an unexpected
* error (e.g. EIO). Such errors cause a TEST_ASSERT failure.
*
* Note, for function signature compatibility with write(2), this function
* returns the number of bytes written, but that value will always be equal
* to the number of requested bytes. All other conditions in this and
* future enhancements to this function either automatically issue another
* write(2) or cause a TEST_ASSERT failure.
*
* Args:
* fd - Opened file descriptor to file to be written.
* count - Number of bytes to write.
*
* Output:
* buf - Starting address of data to be written.
*
* Return:
* On success, number of bytes written.
* On failure, a TEST_ASSERT failure is caused.
*/
ssize_t test_write(int fd, const void *buf, size_t count)
{
ssize_t rc;
ssize_t num_written = 0;
size_t num_left = count;
const char *ptr = buf;
/* Note: Count of zero is allowed (see "RETURN VALUE" portion of
* write(2) manpage for details.
*/
TEST_ASSERT(count >= 0, "Unexpected count, count: %li", count);
do {
rc = write(fd, ptr, num_left);
switch (rc) {
case -1:
TEST_ASSERT(errno == EAGAIN || errno == EINTR,
"Unexpected write failure,\n"
" rc: %zi errno: %i", rc, errno);
continue;
case 0:
TEST_ASSERT(false, "Unexpected EOF,\n"
" rc: %zi num_written: %zi num_left: %zu",
rc, num_written, num_left);
break;
default:
TEST_ASSERT(rc >= 0, "Unexpected ret from write,\n"
" rc: %zi errno: %i", rc, errno);
num_written += rc;
num_left -= rc;
ptr += rc;
break;
}
} while (num_written < count);
return num_written;
}
/* Test Read
*
* A wrapper for read(2), that automatically handles the following
* special conditions:
*
* + Interrupted system call (EINTR)
* + Read of less than requested amount
* + Non-block return (EAGAIN)
*
* For each of the above, an additional read is performed to automatically
* continue reading the requested data.
* There are also many cases where read(2) can return an unexpected
* error (e.g. EIO). Such errors cause a TEST_ASSERT failure. Note,
* it is expected that the file opened by fd at the current file position
* contains at least the number of requested bytes to be read. A TEST_ASSERT
* failure is produced if an End-Of-File condition occurs, before all the
* data is read. It is the callers responsibility to assure that sufficient
* data exists.
*
* Note, for function signature compatibility with read(2), this function
* returns the number of bytes read, but that value will always be equal
* to the number of requested bytes. All other conditions in this and
* future enhancements to this function either automatically issue another
* read(2) or cause a TEST_ASSERT failure.
*
* Args:
* fd - Opened file descriptor to file to be read.
* count - Number of bytes to read.
*
* Output:
* buf - Starting address of where to write the bytes read.
*
* Return:
* On success, number of bytes read.
* On failure, a TEST_ASSERT failure is caused.
*/
ssize_t test_read(int fd, void *buf, size_t count)
{
ssize_t rc;
ssize_t num_read = 0;
size_t num_left = count;
char *ptr = buf;
/* Note: Count of zero is allowed (see "If count is zero" portion of
* read(2) manpage for details.
*/
TEST_ASSERT(count >= 0, "Unexpected count, count: %li", count);
do {
rc = read(fd, ptr, num_left);
switch (rc) {
case -1:
TEST_ASSERT(errno == EAGAIN || errno == EINTR,
"Unexpected read failure,\n"
" rc: %zi errno: %i", rc, errno);
break;
case 0:
TEST_ASSERT(false, "Unexpected EOF,\n"
" rc: %zi num_read: %zi num_left: %zu",
rc, num_read, num_left);
break;
default:
TEST_ASSERT(rc > 0, "Unexpected ret from read,\n"
" rc: %zi errno: %i", rc, errno);
num_read += rc;
num_left -= rc;
ptr += rc;
break;
}
} while (num_read < count);
return num_read;
}

1480
tools/testing/selftests/kvm/lib/kvm_util.c Normal file
View File

File diff suppressed because it is too large Load Diff

67
tools/testing/selftests/kvm/lib/kvm_util_internal.h Normal file
View File

@@ -0,0 +1,67 @@
/*
* tools/testing/selftests/kvm/lib/kvm_util.c
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*/
#ifndef KVM_UTIL_INTERNAL_H
#define KVM_UTIL_INTERNAL_H 1
#include "sparsebit.h"
#ifndef BITS_PER_BYTE
#define BITS_PER_BYTE 8
#endif
#ifndef BITS_PER_LONG
#define BITS_PER_LONG (BITS_PER_BYTE * sizeof(long))
#endif
#define DIV_ROUND_UP(n, d) (((n) + (d) - 1) / (d))
#define BITS_TO_LONGS(nr) DIV_ROUND_UP(nr, BITS_PER_LONG)
/* Concrete definition of struct kvm_vm. */
struct userspace_mem_region {
struct userspace_mem_region *next, *prev;
struct kvm_userspace_memory_region region;
struct sparsebit *unused_phy_pages;
int fd;
off_t offset;
void *host_mem;
void *mmap_start;
size_t mmap_size;
};
struct vcpu {
struct vcpu *next, *prev;
uint32_t id;
int fd;
struct kvm_run *state;
};
struct kvm_vm {
int mode;
int fd;
unsigned int page_size;
unsigned int page_shift;
uint64_t max_gfn;
struct vcpu *vcpu_head;
struct userspace_mem_region *userspace_mem_region_head;
struct sparsebit *vpages_valid;
struct sparsebit *vpages_mapped;
bool pgd_created;
vm_paddr_t pgd;
};
struct vcpu *vcpu_find(struct kvm_vm *vm,
uint32_t vcpuid);
void vcpu_setup(struct kvm_vm *vm, int vcpuid);
void virt_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent);
void regs_dump(FILE *stream, struct kvm_regs *regs,
uint8_t indent);
void sregs_dump(FILE *stream, struct kvm_sregs *sregs,
uint8_t indent);
#endif

2087
tools/testing/selftests/kvm/lib/sparsebit.c Normal file
View File

File diff suppressed because it is too large Load Diff

700
tools/testing/selftests/kvm/lib/x86.c Normal file
View File

@@ -0,0 +1,700 @@
/*
* tools/testing/selftests/kvm/lib/x86.c
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*/
#define _GNU_SOURCE /* for program_invocation_name */
#include "test_util.h"
#include "kvm_util.h"
#include "kvm_util_internal.h"
#include "x86.h"
/* Minimum physical address used for virtual translation tables. */
#define KVM_GUEST_PAGE_TABLE_MIN_PADDR 0x180000
/* Virtual translation table structure declarations */
struct pageMapL4Entry {
uint64_t present:1;
uint64_t writable:1;
uint64_t user:1;
uint64_t write_through:1;
uint64_t cache_disable:1;
uint64_t accessed:1;
uint64_t ignored_06:1;
uint64_t page_size:1;
uint64_t ignored_11_08:4;
uint64_t address:40;
uint64_t ignored_62_52:11;
uint64_t execute_disable:1;
};
struct pageDirectoryPointerEntry {
uint64_t present:1;
uint64_t writable:1;
uint64_t user:1;
uint64_t write_through:1;
uint64_t cache_disable:1;
uint64_t accessed:1;
uint64_t ignored_06:1;
uint64_t page_size:1;
uint64_t ignored_11_08:4;
uint64_t address:40;
uint64_t ignored_62_52:11;
uint64_t execute_disable:1;
};
struct pageDirectoryEntry {
uint64_t present:1;
uint64_t writable:1;
uint64_t user:1;
uint64_t write_through:1;
uint64_t cache_disable:1;
uint64_t accessed:1;
uint64_t ignored_06:1;
uint64_t page_size:1;
uint64_t ignored_11_08:4;
uint64_t address:40;
uint64_t ignored_62_52:11;
uint64_t execute_disable:1;
};
struct pageTableEntry {
uint64_t present:1;
uint64_t writable:1;
uint64_t user:1;
uint64_t write_through:1;
uint64_t cache_disable:1;
uint64_t accessed:1;
uint64_t dirty:1;
uint64_t reserved_07:1;
uint64_t global:1;
uint64_t ignored_11_09:3;
uint64_t address:40;
uint64_t ignored_62_52:11;
uint64_t execute_disable:1;
};
/* Register Dump
*
* Input Args:
* indent - Left margin indent amount
* regs - register
*
* Output Args:
* stream - Output FILE stream
*
* Return: None
*
* Dumps the state of the registers given by regs, to the FILE stream
* given by steam.
*/
void regs_dump(FILE *stream, struct kvm_regs *regs,
uint8_t indent)
{
fprintf(stream, "%*srax: 0x%.16llx rbx: 0x%.16llx "
"rcx: 0x%.16llx rdx: 0x%.16llx\n",
indent, "",
regs->rax, regs->rbx, regs->rcx, regs->rdx);
fprintf(stream, "%*srsi: 0x%.16llx rdi: 0x%.16llx "
"rsp: 0x%.16llx rbp: 0x%.16llx\n",
indent, "",
regs->rsi, regs->rdi, regs->rsp, regs->rbp);
fprintf(stream, "%*sr8: 0x%.16llx r9: 0x%.16llx "
"r10: 0x%.16llx r11: 0x%.16llx\n",
indent, "",
regs->r8, regs->r9, regs->r10, regs->r11);
fprintf(stream, "%*sr12: 0x%.16llx r13: 0x%.16llx "
"r14: 0x%.16llx r15: 0x%.16llx\n",
indent, "",
regs->r12, regs->r13, regs->r14, regs->r15);
fprintf(stream, "%*srip: 0x%.16llx rfl: 0x%.16llx\n",
indent, "",
regs->rip, regs->rflags);
}
/* Segment Dump
*
* Input Args:
* indent - Left margin indent amount
* segment - KVM segment
*
* Output Args:
* stream - Output FILE stream
*
* Return: None
*
* Dumps the state of the KVM segment given by segment, to the FILE stream
* given by steam.
*/
static void segment_dump(FILE *stream, struct kvm_segment *segment,
uint8_t indent)
{
fprintf(stream, "%*sbase: 0x%.16llx limit: 0x%.8x "
"selector: 0x%.4x type: 0x%.2x\n",
indent, "", segment->base, segment->limit,
segment->selector, segment->type);
fprintf(stream, "%*spresent: 0x%.2x dpl: 0x%.2x "
"db: 0x%.2x s: 0x%.2x l: 0x%.2x\n",
indent, "", segment->present, segment->dpl,
segment->db, segment->s, segment->l);
fprintf(stream, "%*sg: 0x%.2x avl: 0x%.2x "
"unusable: 0x%.2x padding: 0x%.2x\n",
indent, "", segment->g, segment->avl,
segment->unusable, segment->padding);
}
/* dtable Dump
*
* Input Args:
* indent - Left margin indent amount
* dtable - KVM dtable
*
* Output Args:
* stream - Output FILE stream
*
* Return: None
*
* Dumps the state of the KVM dtable given by dtable, to the FILE stream
* given by steam.
*/
static void dtable_dump(FILE *stream, struct kvm_dtable *dtable,
uint8_t indent)
{
fprintf(stream, "%*sbase: 0x%.16llx limit: 0x%.4x "
"padding: 0x%.4x 0x%.4x 0x%.4x\n",
indent, "", dtable->base, dtable->limit,
dtable->padding[0], dtable->padding[1], dtable->padding[2]);
}
/* System Register Dump
*
* Input Args:
* indent - Left margin indent amount
* sregs - System registers
*
* Output Args:
* stream - Output FILE stream
*
* Return: None
*
* Dumps the state of the system registers given by sregs, to the FILE stream
* given by steam.
*/
void sregs_dump(FILE *stream, struct kvm_sregs *sregs,
uint8_t indent)
{
unsigned int i;
fprintf(stream, "%*scs:\n", indent, "");
segment_dump(stream, &sregs->cs, indent + 2);
fprintf(stream, "%*sds:\n", indent, "");
segment_dump(stream, &sregs->ds, indent + 2);
fprintf(stream, "%*ses:\n", indent, "");
segment_dump(stream, &sregs->es, indent + 2);
fprintf(stream, "%*sfs:\n", indent, "");
segment_dump(stream, &sregs->fs, indent + 2);
fprintf(stream, "%*sgs:\n", indent, "");
segment_dump(stream, &sregs->gs, indent + 2);
fprintf(stream, "%*sss:\n", indent, "");
segment_dump(stream, &sregs->ss, indent + 2);
fprintf(stream, "%*str:\n", indent, "");
segment_dump(stream, &sregs->tr, indent + 2);
fprintf(stream, "%*sldt:\n", indent, "");
segment_dump(stream, &sregs->ldt, indent + 2);
fprintf(stream, "%*sgdt:\n", indent, "");
dtable_dump(stream, &sregs->gdt, indent + 2);
fprintf(stream, "%*sidt:\n", indent, "");
dtable_dump(stream, &sregs->idt, indent + 2);
fprintf(stream, "%*scr0: 0x%.16llx cr2: 0x%.16llx "
"cr3: 0x%.16llx cr4: 0x%.16llx\n",
indent, "",
sregs->cr0, sregs->cr2, sregs->cr3, sregs->cr4);
fprintf(stream, "%*scr8: 0x%.16llx efer: 0x%.16llx "
"apic_base: 0x%.16llx\n",
indent, "",
sregs->cr8, sregs->efer, sregs->apic_base);
fprintf(stream, "%*sinterrupt_bitmap:\n", indent, "");
for (i = 0; i < (KVM_NR_INTERRUPTS + 63) / 64; i++) {
fprintf(stream, "%*s%.16llx\n", indent + 2, "",
sregs->interrupt_bitmap[i]);
}
}
void virt_pgd_alloc(struct kvm_vm *vm, uint32_t pgd_memslot)
{
int rc;
TEST_ASSERT(vm->mode == VM_MODE_FLAT48PG, "Attempt to use "
"unknown or unsupported guest mode, mode: 0x%x", vm->mode);
/* If needed, create page map l4 table. */
if (!vm->pgd_created) {
vm_paddr_t paddr = vm_phy_page_alloc(vm,
KVM_GUEST_PAGE_TABLE_MIN_PADDR, pgd_memslot);
vm->pgd = paddr;
/* Set pointer to pgd tables in all the VCPUs that
* have already been created. Future VCPUs will have
* the value set as each one is created.
*/
for (struct vcpu *vcpu = vm->vcpu_head; vcpu;
vcpu = vcpu->next) {
struct kvm_sregs sregs;
/* Obtain the current system register settings */
vcpu_sregs_get(vm, vcpu->id, &sregs);
/* Set and store the pointer to the start of the
* pgd tables.
*/
sregs.cr3 = vm->pgd;
vcpu_sregs_set(vm, vcpu->id, &sregs);
}
vm->pgd_created = true;
}
}
/* VM Virtual Page Map
*
* Input Args:
* vm - Virtual Machine
* vaddr - VM Virtual Address
* paddr - VM Physical Address
* pgd_memslot - Memory region slot for new virtual translation tables
*
* Output Args: None
*
* Return: None
*
* Within the VM given by vm, creates a virtual translation for the page
* starting at vaddr to the page starting at paddr.
*/
void virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
uint32_t pgd_memslot)
{
uint16_t index[4];
struct pageMapL4Entry *pml4e;
TEST_ASSERT(vm->mode == VM_MODE_FLAT48PG, "Attempt to use "
"unknown or unsupported guest mode, mode: 0x%x", vm->mode);
TEST_ASSERT((vaddr % vm->page_size) == 0,
"Virtual address not on page boundary,\n"
" vaddr: 0x%lx vm->page_size: 0x%x",
vaddr, vm->page_size);
TEST_ASSERT(sparsebit_is_set(vm->vpages_valid,
(vaddr >> vm->page_shift)),
"Invalid virtual address, vaddr: 0x%lx",
vaddr);
TEST_ASSERT((paddr % vm->page_size) == 0,
"Physical address not on page boundary,\n"
" paddr: 0x%lx vm->page_size: 0x%x",
paddr, vm->page_size);
TEST_ASSERT((paddr >> vm->page_shift) <= vm->max_gfn,
"Physical address beyond beyond maximum supported,\n"
" paddr: 0x%lx vm->max_gfn: 0x%lx vm->page_size: 0x%x",
paddr, vm->max_gfn, vm->page_size);
index[0] = (vaddr >> 12) & 0x1ffu;
index[1] = (vaddr >> 21) & 0x1ffu;
index[2] = (vaddr >> 30) & 0x1ffu;
index[3] = (vaddr >> 39) & 0x1ffu;
/* Allocate page directory pointer table if not present. */
pml4e = addr_gpa2hva(vm, vm->pgd);
if (!pml4e[index[3]].present) {
pml4e[index[3]].address = vm_phy_page_alloc(vm,
KVM_GUEST_PAGE_TABLE_MIN_PADDR, pgd_memslot)
>> vm->page_shift;
pml4e[index[3]].writable = true;
pml4e[index[3]].present = true;
}
/* Allocate page directory table if not present. */
struct pageDirectoryPointerEntry *pdpe;
pdpe = addr_gpa2hva(vm, pml4e[index[3]].address * vm->page_size);
if (!pdpe[index[2]].present) {
pdpe[index[2]].address = vm_phy_page_alloc(vm,
KVM_GUEST_PAGE_TABLE_MIN_PADDR, pgd_memslot)
>> vm->page_shift;
pdpe[index[2]].writable = true;
pdpe[index[2]].present = true;
}
/* Allocate page table if not present. */
struct pageDirectoryEntry *pde;
pde = addr_gpa2hva(vm, pdpe[index[2]].address * vm->page_size);
if (!pde[index[1]].present) {
pde[index[1]].address = vm_phy_page_alloc(vm,
KVM_GUEST_PAGE_TABLE_MIN_PADDR, pgd_memslot)
>> vm->page_shift;
pde[index[1]].writable = true;
pde[index[1]].present = true;
}
/* Fill in page table entry. */
struct pageTableEntry *pte;
pte = addr_gpa2hva(vm, pde[index[1]].address * vm->page_size);
pte[index[0]].address = paddr >> vm->page_shift;
pte[index[0]].writable = true;
pte[index[0]].present = 1;
}
/* Virtual Translation Tables Dump
*
* Input Args:
* vm - Virtual Machine
* indent - Left margin indent amount
*
* Output Args:
* stream - Output FILE stream
*
* Return: None
*
* Dumps to the FILE stream given by stream, the contents of all the
* virtual translation tables for the VM given by vm.
*/
void virt_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent)
{
struct pageMapL4Entry *pml4e, *pml4e_start;
struct pageDirectoryPointerEntry *pdpe, *pdpe_start;
struct pageDirectoryEntry *pde, *pde_start;
struct pageTableEntry *pte, *pte_start;
if (!vm->pgd_created)
return;
fprintf(stream, "%*s "
" no\n", indent, "");
fprintf(stream, "%*s index hvaddr gpaddr "
"addr w exec dirty\n",
indent, "");
pml4e_start = (struct pageMapL4Entry *) addr_gpa2hva(vm,
vm->pgd);
for (uint16_t n1 = 0; n1 <= 0x1ffu; n1++) {
pml4e = &pml4e_start[n1];
if (!pml4e->present)
continue;
fprintf(stream, "%*spml4e 0x%-3zx %p 0x%-12lx 0x%-10lx %u "
" %u\n",
indent, "",
pml4e - pml4e_start, pml4e,
addr_hva2gpa(vm, pml4e), (uint64_t) pml4e->address,
pml4e->writable, pml4e->execute_disable);
pdpe_start = addr_gpa2hva(vm, pml4e->address
* vm->page_size);
for (uint16_t n2 = 0; n2 <= 0x1ffu; n2++) {
pdpe = &pdpe_start[n2];
if (!pdpe->present)
continue;
fprintf(stream, "%*spdpe 0x%-3zx %p 0x%-12lx 0x%-10lx "
"%u %u\n",
indent, "",
pdpe - pdpe_start, pdpe,
addr_hva2gpa(vm, pdpe),
(uint64_t) pdpe->address, pdpe->writable,
pdpe->execute_disable);
pde_start = addr_gpa2hva(vm,
pdpe->address * vm->page_size);
for (uint16_t n3 = 0; n3 <= 0x1ffu; n3++) {
pde = &pde_start[n3];
if (!pde->present)
continue;
fprintf(stream, "%*spde 0x%-3zx %p "
"0x%-12lx 0x%-10lx %u %u\n",
indent, "", pde - pde_start, pde,
addr_hva2gpa(vm, pde),
(uint64_t) pde->address, pde->writable,
pde->execute_disable);
pte_start = addr_gpa2hva(vm,
pde->address * vm->page_size);
for (uint16_t n4 = 0; n4 <= 0x1ffu; n4++) {
pte = &pte_start[n4];
if (!pte->present)
continue;
fprintf(stream, "%*spte 0x%-3zx %p "
"0x%-12lx 0x%-10lx %u %u "
" %u 0x%-10lx\n",
indent, "",
pte - pte_start, pte,
addr_hva2gpa(vm, pte),
(uint64_t) pte->address,
pte->writable,
pte->execute_disable,
pte->dirty,
((uint64_t) n1 << 27)
| ((uint64_t) n2 << 18)
| ((uint64_t) n3 << 9)
| ((uint64_t) n4));
}
}
}
}
}
/* Set Unusable Segment
*
* Input Args: None
*
* Output Args:
* segp - Pointer to segment register
*
* Return: None
*
* Sets the segment register pointed to by segp to an unusable state.
*/
static void kvm_seg_set_unusable(struct kvm_segment *segp)
{
memset(segp, 0, sizeof(*segp));
segp->unusable = true;
}
/* Set Long Mode Flat Kernel Code Segment
*
* Input Args:
* selector - selector value
*
* Output Args:
* segp - Pointer to KVM segment
*
* Return: None
*
* Sets up the KVM segment pointed to by segp, to be a code segment
* with the selector value given by selector.
*/
static void kvm_seg_set_kernel_code_64bit(uint16_t selector,
struct kvm_segment *segp)
{
memset(segp, 0, sizeof(*segp));
segp->selector = selector;
segp->limit = 0xFFFFFFFFu;
segp->s = 0x1; /* kTypeCodeData */
segp->type = 0x08 | 0x01 | 0x02; /* kFlagCode | kFlagCodeAccessed
* | kFlagCodeReadable
*/
segp->g = true;
segp->l = true;
segp->present = 1;
}
/* Set Long Mode Flat Kernel Data Segment
*
* Input Args:
* selector - selector value
*
* Output Args:
* segp - Pointer to KVM segment
*
* Return: None
*
* Sets up the KVM segment pointed to by segp, to be a data segment
* with the selector value given by selector.
*/
static void kvm_seg_set_kernel_data_64bit(uint16_t selector,
struct kvm_segment *segp)
{
memset(segp, 0, sizeof(*segp));
segp->selector = selector;
segp->limit = 0xFFFFFFFFu;
segp->s = 0x1; /* kTypeCodeData */
segp->type = 0x00 | 0x01 | 0x02; /* kFlagData | kFlagDataAccessed
* | kFlagDataWritable
*/
segp->g = true;
segp->present = true;
}
/* Address Guest Virtual to Guest Physical
*
* Input Args:
* vm - Virtual Machine
* gpa - VM virtual address
*
* Output Args: None
*
* Return:
* Equivalent VM physical address
*
* Translates the VM virtual address given by gva to a VM physical
* address and then locates the memory region containing the VM
* physical address, within the VM given by vm. When found, the host
* virtual address providing the memory to the vm physical address is returned.
* A TEST_ASSERT failure occurs if no region containing translated
* VM virtual address exists.
*/
vm_paddr_t addr_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva)
{
uint16_t index[4];
struct pageMapL4Entry *pml4e;
struct pageDirectoryPointerEntry *pdpe;
struct pageDirectoryEntry *pde;
struct pageTableEntry *pte;
void *hva;
TEST_ASSERT(vm->mode == VM_MODE_FLAT48PG, "Attempt to use "
"unknown or unsupported guest mode, mode: 0x%x", vm->mode);
index[0] = (gva >> 12) & 0x1ffu;
index[1] = (gva >> 21) & 0x1ffu;
index[2] = (gva >> 30) & 0x1ffu;
index[3] = (gva >> 39) & 0x1ffu;
if (!vm->pgd_created)
goto unmapped_gva;
pml4e = addr_gpa2hva(vm, vm->pgd);
if (!pml4e[index[3]].present)
goto unmapped_gva;
pdpe = addr_gpa2hva(vm, pml4e[index[3]].address * vm->page_size);
if (!pdpe[index[2]].present)
goto unmapped_gva;
pde = addr_gpa2hva(vm, pdpe[index[2]].address * vm->page_size);
if (!pde[index[1]].present)
goto unmapped_gva;
pte = addr_gpa2hva(vm, pde[index[1]].address * vm->page_size);
if (!pte[index[0]].present)
goto unmapped_gva;
return (pte[index[0]].address * vm->page_size) + (gva & 0xfffu);
unmapped_gva:
TEST_ASSERT(false, "No mapping for vm virtual address, "
"gva: 0x%lx", gva);
}
void vcpu_setup(struct kvm_vm *vm, int vcpuid)
{
struct kvm_sregs sregs;
/* Set mode specific system register values. */
vcpu_sregs_get(vm, vcpuid, &sregs);
switch (vm->mode) {
case VM_MODE_FLAT48PG:
sregs.cr0 = X86_CR0_PE | X86_CR0_NE | X86_CR0_PG;
sregs.cr4 |= X86_CR4_PAE;
sregs.efer |= (EFER_LME | EFER_LMA | EFER_NX);
kvm_seg_set_unusable(&sregs.ldt);
kvm_seg_set_kernel_code_64bit(0x8, &sregs.cs);
kvm_seg_set_kernel_data_64bit(0x10, &sregs.ds);
kvm_seg_set_kernel_data_64bit(0x10, &sregs.es);
break;
default:
TEST_ASSERT(false, "Unknown guest mode, mode: 0x%x", vm->mode);
}
vcpu_sregs_set(vm, vcpuid, &sregs);
/* If virtual translation table have been setup, set system register
* to point to the tables. It's okay if they haven't been setup yet,
* in that the code that sets up the virtual translation tables, will
* go back through any VCPUs that have already been created and set
* their values.
*/
if (vm->pgd_created) {
struct kvm_sregs sregs;
vcpu_sregs_get(vm, vcpuid, &sregs);
sregs.cr3 = vm->pgd;
vcpu_sregs_set(vm, vcpuid, &sregs);
}
}
/* Adds a vCPU with reasonable defaults (i.e., a stack)
*
* Input Args:
* vcpuid - The id of the VCPU to add to the VM.
* guest_code - The vCPU's entry point
*/
void vm_vcpu_add_default(struct kvm_vm *vm, uint32_t vcpuid, void *guest_code)
{
struct kvm_mp_state mp_state;
struct kvm_regs regs;
vm_vaddr_t stack_vaddr;
stack_vaddr = vm_vaddr_alloc(vm, DEFAULT_STACK_PGS * getpagesize(),
DEFAULT_GUEST_STACK_VADDR_MIN, 0, 0);
/* Create VCPU */
vm_vcpu_add(vm, vcpuid);
/* Setup guest general purpose registers */
vcpu_regs_get(vm, vcpuid, &regs);
regs.rflags = regs.rflags | 0x2;
regs.rsp = stack_vaddr + (DEFAULT_STACK_PGS * getpagesize());
regs.rip = (unsigned long) guest_code;
vcpu_regs_set(vm, vcpuid, &regs);
/* Setup the MP state */
mp_state.mp_state = 0;
vcpu_set_mp_state(vm, vcpuid, &mp_state);
}
/* VM VCPU CPUID Set
*
* Input Args:
* vm - Virtual Machine
* vcpuid - VCPU id
* cpuid - The CPUID values to set.
*
* Output Args: None
*
* Return: void
*
* Set the VCPU's CPUID.
*/
void vcpu_set_cpuid(struct kvm_vm *vm,
uint32_t vcpuid, struct kvm_cpuid2 *cpuid)
{
struct vcpu *vcpu = vcpu_find(vm, vcpuid);
int rc;
TEST_ASSERT(vcpu != NULL, "vcpu not found, vcpuid: %u", vcpuid);
rc = ioctl(vcpu->fd, KVM_SET_CPUID2, cpuid);
TEST_ASSERT(rc == 0, "KVM_SET_CPUID2 failed, rc: %i errno: %i",
rc, errno);
}
/* Create a VM with reasonable defaults
*
* Input Args:
* vcpuid - The id of the single VCPU to add to the VM.
* guest_code - The vCPU's entry point
*
* Output Args: None
*
* Return:
* Pointer to opaque structure that describes the created VM.
*/
struct kvm_vm *vm_create_default(uint32_t vcpuid, void *guest_code)
{
struct kvm_vm *vm;
/* Create VM */
vm = vm_create(VM_MODE_FLAT48PG, DEFAULT_GUEST_PHY_PAGES, O_RDWR);
/* Setup guest code */
kvm_vm_elf_load(vm, program_invocation_name, 0, 0);
/* Setup IRQ Chip */
vm_create_irqchip(vm);
/* Add the first vCPU. */
vm_vcpu_add_default(vm, vcpuid, guest_code);
return vm;
}

54
tools/testing/selftests/kvm/set_sregs_test.c Normal file
View File

@@ -0,0 +1,54 @@
/*
* KVM_SET_SREGS tests
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*
* This is a regression test for the bug fixed by the following commit:
* d3802286fa0f ("kvm: x86: Disallow illegal IA32_APIC_BASE MSR values")
*
* That bug allowed a user-mode program that called the KVM_SET_SREGS
* ioctl to put a VCPU's local APIC into an invalid state.
*
*/
#define _GNU_SOURCE /* for program_invocation_short_name */
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/ioctl.h>
#include "test_util.h"
#include "kvm_util.h"
#include "x86.h"
#define VCPU_ID 5
int main(int argc, char *argv[])
{
struct kvm_sregs sregs;
struct kvm_vm *vm;
int rc;
/* Tell stdout not to buffer its content */
setbuf(stdout, NULL);
/* Create VM */
vm = vm_create_default(VCPU_ID, NULL);
vcpu_sregs_get(vm, VCPU_ID, &sregs);
sregs.apic_base = 1 << 10;
rc = _vcpu_sregs_set(vm, VCPU_ID, &sregs);
TEST_ASSERT(rc, "Set IA32_APIC_BASE to %llx (invalid)",
sregs.apic_base);
sregs.apic_base = 1 << 11;
rc = _vcpu_sregs_set(vm, VCPU_ID, &sregs);
TEST_ASSERT(!rc, "Couldn't set IA32_APIC_BASE to %llx (valid)",
sregs.apic_base);
kvm_vm_free(vm);
return 0;
}

232
tools/testing/selftests/kvm/sync_regs_test.c Normal file
View File

@@ -0,0 +1,232 @@
/*
* Test for x86 KVM_CAP_SYNC_REGS
*
* Copyright (C) 2018, Google LLC.
*
* This work is licensed under the terms of the GNU GPL, version 2.
*
* Verifies expected behavior of x86 KVM_CAP_SYNC_REGS functionality,
* including requesting an invalid register set, updates to/from values
* in kvm_run.s.regs when kvm_valid_regs and kvm_dirty_regs are toggled.
*/
#define _GNU_SOURCE /* for program_invocation_short_name */
#include <fcntl.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#include <sys/ioctl.h>
#include "test_util.h"
#include "kvm_util.h"
#include "x86.h"
#define VCPU_ID 5
#define PORT_HOST_SYNC 0x1000
static void __exit_to_l0(uint16_t port, uint64_t arg0, uint64_t arg1)
{
__asm__ __volatile__("in %[port], %%al"
:
: [port]"d"(port), "D"(arg0), "S"(arg1)
: "rax");
}
#define exit_to_l0(_port, _arg0, _arg1) \
__exit_to_l0(_port, (uint64_t) (_arg0), (uint64_t) (_arg1))
#define GUEST_ASSERT(_condition) do { \
if (!(_condition)) \
exit_to_l0(PORT_ABORT, "Failed guest assert: " #_condition, 0);\
} while (0)
void guest_code(void)
{
for (;;) {
exit_to_l0(PORT_HOST_SYNC, "hello", 0);
asm volatile ("inc %r11");
}
}
static void compare_regs(struct kvm_regs *left, struct kvm_regs *right)
{
#define REG_COMPARE(reg) \
TEST_ASSERT(left->reg == right->reg, \
"Register " #reg \
" values did not match: 0x%llx, 0x%llx\n", \
left->reg, right->reg)
REG_COMPARE(rax);
REG_COMPARE(rbx);
REG_COMPARE(rcx);
REG_COMPARE(rdx);
REG_COMPARE(rsi);
REG_COMPARE(rdi);
REG_COMPARE(rsp);
REG_COMPARE(rbp);
REG_COMPARE(r8);
REG_COMPARE(r9);
REG_COMPARE(r10);
REG_COMPARE(r11);
REG_COMPARE(r12);
REG_COMPARE(r13);
REG_COMPARE(r14);
REG_COMPARE(r15);
REG_COMPARE(rip);
REG_COMPARE(rflags);
#undef REG_COMPARE
}
static void compare_sregs(struct kvm_sregs *left, struct kvm_sregs *right)
{
}
static void compare_vcpu_events(struct kvm_vcpu_events *left,
struct kvm_vcpu_events *right)
{
}
int main(int argc, char *argv[])
{
struct kvm_vm *vm;
struct kvm_run *run;
struct kvm_regs regs;
struct kvm_sregs sregs;
struct kvm_vcpu_events events;
int rv, cap;
/* Tell stdout not to buffer its content */
setbuf(stdout, NULL);
cap = kvm_check_cap(KVM_CAP_SYNC_REGS);
TEST_ASSERT((unsigned long)cap == KVM_SYNC_X86_VALID_FIELDS,
"KVM_CAP_SYNC_REGS (0x%x) != KVM_SYNC_X86_VALID_FIELDS (0x%lx)\n",
cap, KVM_SYNC_X86_VALID_FIELDS);
/* Create VM */
vm = vm_create_default(VCPU_ID, guest_code);
run = vcpu_state(vm, VCPU_ID);
/* Request reading invalid register set from VCPU. */
run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS << 1;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(rv < 0 && errno == EINVAL,
"Invalid kvm_valid_regs did not cause expected KVM_RUN error: %d\n",
rv);
vcpu_state(vm, VCPU_ID)->kvm_valid_regs = 0;
/* Request setting invalid register set into VCPU. */
run->kvm_dirty_regs = KVM_SYNC_X86_VALID_FIELDS << 1;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(rv < 0 && errno == EINVAL,
"Invalid kvm_dirty_regs did not cause expected KVM_RUN error: %d\n",
rv);
vcpu_state(vm, VCPU_ID)->kvm_dirty_regs = 0;
/* Request and verify all valid register sets. */
/* TODO: BUILD TIME CHECK: TEST_ASSERT(KVM_SYNC_X86_NUM_FIELDS != 3); */
run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
"Unexpected exit reason: %u (%s),\n",
run->exit_reason,
exit_reason_str(run->exit_reason));
vcpu_regs_get(vm, VCPU_ID, &regs);
compare_regs(&regs, &run->s.regs.regs);
vcpu_sregs_get(vm, VCPU_ID, &sregs);
compare_sregs(&sregs, &run->s.regs.sregs);
vcpu_events_get(vm, VCPU_ID, &events);
compare_vcpu_events(&events, &run->s.regs.events);
/* Set and verify various register values. */
run->s.regs.regs.r11 = 0xBAD1DEA;
run->s.regs.sregs.apic_base = 1 << 11;
/* TODO run->s.regs.events.XYZ = ABC; */
run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS;
run->kvm_dirty_regs = KVM_SYNC_X86_REGS | KVM_SYNC_X86_SREGS;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
"Unexpected exit reason: %u (%s),\n",
run->exit_reason,
exit_reason_str(run->exit_reason));
TEST_ASSERT(run->s.regs.regs.r11 == 0xBAD1DEA + 1,
"r11 sync regs value incorrect 0x%llx.",
run->s.regs.regs.r11);
TEST_ASSERT(run->s.regs.sregs.apic_base == 1 << 11,
"apic_base sync regs value incorrect 0x%llx.",
run->s.regs.sregs.apic_base);
vcpu_regs_get(vm, VCPU_ID, &regs);
compare_regs(&regs, &run->s.regs.regs);
vcpu_sregs_get(vm, VCPU_ID, &sregs);
compare_sregs(&sregs, &run->s.regs.sregs);
vcpu_events_get(vm, VCPU_ID, &events);
compare_vcpu_events(&events, &run->s.regs.events);
/* Clear kvm_dirty_regs bits, verify new s.regs values are
* overwritten with existing guest values.
*/
run->kvm_valid_regs = KVM_SYNC_X86_VALID_FIELDS;
run->kvm_dirty_regs = 0;
run->s.regs.regs.r11 = 0xDEADBEEF;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
"Unexpected exit reason: %u (%s),\n",
run->exit_reason,
exit_reason_str(run->exit_reason));
TEST_ASSERT(run->s.regs.regs.r11 != 0xDEADBEEF,
"r11 sync regs value incorrect 0x%llx.",
run->s.regs.regs.r11);
/* Clear kvm_valid_regs bits and kvm_dirty_bits.
* Verify s.regs values are not overwritten with existing guest values
* and that guest values are not overwritten with kvm_sync_regs values.
*/
run->kvm_valid_regs = 0;
run->kvm_dirty_regs = 0;
run->s.regs.regs.r11 = 0xAAAA;
regs.r11 = 0xBAC0;
vcpu_regs_set(vm, VCPU_ID, &regs);
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
"Unexpected exit reason: %u (%s),\n",
run->exit_reason,
exit_reason_str(run->exit_reason));
TEST_ASSERT(run->s.regs.regs.r11 == 0xAAAA,
"r11 sync regs value incorrect 0x%llx.",
run->s.regs.regs.r11);
vcpu_regs_get(vm, VCPU_ID, &regs);
TEST_ASSERT(regs.r11 == 0xBAC0 + 1,
"r11 guest value incorrect 0x%llx.",
regs.r11);
/* Clear kvm_valid_regs bits. Verify s.regs values are not overwritten
* with existing guest values but that guest values are overwritten
* with kvm_sync_regs values.
*/
run->kvm_valid_regs = 0;
run->kvm_dirty_regs = KVM_SYNC_X86_VALID_FIELDS;
run->s.regs.regs.r11 = 0xBBBB;
rv = _vcpu_run(vm, VCPU_ID);
TEST_ASSERT(run->exit_reason == KVM_EXIT_IO,
"Unexpected exit reason: %u (%s),\n",
run->exit_reason,
exit_reason_str(run->exit_reason));
TEST_ASSERT(run->s.regs.regs.r11 == 0xBBBB,
"r11 sync regs value incorrect 0x%llx.",
run->s.regs.regs.r11);
vcpu_regs_get(vm, VCPU_ID, &regs);
TEST_ASSERT(regs.r11 == 0xBBBB + 1,
"r11 guest value incorrect 0x%llx.",
regs.r11);
kvm_vm_free(vm);
return 0;
}