summaryrefslogtreecommitdiff
path: root/tools
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2021-03-04 11:26:17 -0800
committerLinus Torvalds <torvalds@linux-foundation.org>2021-03-04 11:26:17 -0800
commitcee407c5cc427a7d9b21ee964fbda613e368bdff (patch)
treed5a5e4a0ec37fd4dc156495389fc5a834f483079 /tools
parentc5a58f877ca645a3303f7a57476f2de837fdb97a (diff)
parent9e46f6c6c959d9bb45445c2e8f04a75324a0dfd0 (diff)
Merge tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm
Pull KVM fixes from Paolo Bonzini: - Doc fixes - selftests fixes - Add runstate information to the new Xen support - Allow compiling out the Xen interface - 32-bit PAE without EPT bugfix - NULL pointer dereference bugfix * tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: KVM: SVM: Clear the CR4 register on reset KVM: x86/xen: Add support for vCPU runstate information KVM: x86/xen: Fix return code when clearing vcpu_info and vcpu_time_info selftests: kvm: Mmap the entire vcpu mmap area KVM: Documentation: Fix index for KVM_CAP_PPC_DAWR1 KVM: x86: allow compiling out the Xen hypercall interface KVM: xen: flush deferred static key before checking it KVM: x86/mmu: Set SPTE_AD_WRPROT_ONLY_MASK if and only if PML is enabled KVM: x86: hyper-v: Fix Hyper-V context null-ptr-deref KVM: x86: remove misplaced comment on active_mmu_pages KVM: Documentation: rectify rst markup in kvm_run->flags Documentation: kvm: fix messy conversion from .txt to .rst
Diffstat (limited to 'tools')
-rw-r--r--tools/testing/selftests/kvm/lib/kvm_util.c6
-rw-r--r--tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c159
2 files changed, 160 insertions, 5 deletions
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index d787cb802b4a..e5fbf16f725b 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -21,6 +21,8 @@
#define KVM_UTIL_PGS_PER_HUGEPG 512
#define KVM_UTIL_MIN_PFN 2
+static int vcpu_mmap_sz(void);
+
/* Aligns x up to the next multiple of size. Size must be a power of 2. */
static void *align(void *x, size_t size)
{
@@ -509,7 +511,7 @@ static void vm_vcpu_rm(struct kvm_vm *vm, struct vcpu *vcpu)
vcpu->dirty_gfns = NULL;
}
- ret = munmap(vcpu->state, sizeof(*vcpu->state));
+ ret = munmap(vcpu->state, vcpu_mmap_sz());
TEST_ASSERT(ret == 0, "munmap of VCPU fd failed, rc: %i "
"errno: %i", ret, errno);
close(vcpu->fd);
@@ -978,7 +980,7 @@ void vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpuid)
TEST_ASSERT(vcpu_mmap_sz() >= sizeof(*vcpu->state), "vcpu mmap size "
"smaller than expected, vcpu_mmap_sz: %i expected_min: %zi",
vcpu_mmap_sz(), sizeof(*vcpu->state));
- vcpu->state = (struct kvm_run *) mmap(NULL, sizeof(*vcpu->state),
+ vcpu->state = (struct kvm_run *) mmap(NULL, vcpu_mmap_sz(),
PROT_READ | PROT_WRITE, MAP_SHARED, vcpu->fd, 0);
TEST_ASSERT(vcpu->state != MAP_FAILED, "mmap vcpu_state failed, "
"vcpu id: %u errno: %i", vcpuid, errno);
diff --git a/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c b/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c
index 9246ea310587..804ff5ff022d 100644
--- a/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c
+++ b/tools/testing/selftests/kvm/x86_64/xen_shinfo_test.c
@@ -13,19 +13,27 @@
#include <stdint.h>
#include <time.h>
+#include <sched.h>
+#include <sys/syscall.h>
#define VCPU_ID 5
+#define SHINFO_REGION_GVA 0xc0000000ULL
#define SHINFO_REGION_GPA 0xc0000000ULL
#define SHINFO_REGION_SLOT 10
#define PAGE_SIZE 4096
#define PVTIME_ADDR (SHINFO_REGION_GPA + PAGE_SIZE)
+#define RUNSTATE_ADDR (SHINFO_REGION_GPA + PAGE_SIZE + 0x20)
+
+#define RUNSTATE_VADDR (SHINFO_REGION_GVA + PAGE_SIZE + 0x20)
static struct kvm_vm *vm;
#define XEN_HYPERCALL_MSR 0x40000000
+#define MIN_STEAL_TIME 50000
+
struct pvclock_vcpu_time_info {
u32 version;
u32 pad0;
@@ -43,11 +51,67 @@ struct pvclock_wall_clock {
u32 nsec;
} __attribute__((__packed__));
+struct vcpu_runstate_info {
+ uint32_t state;
+ uint64_t state_entry_time;
+ uint64_t time[4];
+};
+
+#define RUNSTATE_running 0
+#define RUNSTATE_runnable 1
+#define RUNSTATE_blocked 2
+#define RUNSTATE_offline 3
+
static void guest_code(void)
{
+ struct vcpu_runstate_info *rs = (void *)RUNSTATE_VADDR;
+
+ /* Test having the host set runstates manually */
+ GUEST_SYNC(RUNSTATE_runnable);
+ GUEST_ASSERT(rs->time[RUNSTATE_runnable] != 0);
+ GUEST_ASSERT(rs->state == 0);
+
+ GUEST_SYNC(RUNSTATE_blocked);
+ GUEST_ASSERT(rs->time[RUNSTATE_blocked] != 0);
+ GUEST_ASSERT(rs->state == 0);
+
+ GUEST_SYNC(RUNSTATE_offline);
+ GUEST_ASSERT(rs->time[RUNSTATE_offline] != 0);
+ GUEST_ASSERT(rs->state == 0);
+
+ /* Test runstate time adjust */
+ GUEST_SYNC(4);
+ GUEST_ASSERT(rs->time[RUNSTATE_blocked] == 0x5a);
+ GUEST_ASSERT(rs->time[RUNSTATE_offline] == 0x6b6b);
+
+ /* Test runstate time set */
+ GUEST_SYNC(5);
+ GUEST_ASSERT(rs->state_entry_time >= 0x8000);
+ GUEST_ASSERT(rs->time[RUNSTATE_runnable] == 0);
+ GUEST_ASSERT(rs->time[RUNSTATE_blocked] == 0x6b6b);
+ GUEST_ASSERT(rs->time[RUNSTATE_offline] == 0x5a);
+
+ /* sched_yield() should result in some 'runnable' time */
+ GUEST_SYNC(6);
+ GUEST_ASSERT(rs->time[RUNSTATE_runnable] >= MIN_STEAL_TIME);
+
GUEST_DONE();
}
+static long get_run_delay(void)
+{
+ char path[64];
+ long val[2];
+ FILE *fp;
+
+ sprintf(path, "/proc/%ld/schedstat", syscall(SYS_gettid));
+ fp = fopen(path, "r");
+ fscanf(fp, "%ld %ld ", &val[0], &val[1]);
+ fclose(fp);
+
+ return val[1];
+}
+
static int cmp_timespec(struct timespec *a, struct timespec *b)
{
if (a->tv_sec > b->tv_sec)
@@ -66,12 +130,14 @@ int main(int argc, char *argv[])
{
struct timespec min_ts, max_ts, vm_ts;
- if (!(kvm_check_cap(KVM_CAP_XEN_HVM) &
- KVM_XEN_HVM_CONFIG_SHARED_INFO) ) {
+ int xen_caps = kvm_check_cap(KVM_CAP_XEN_HVM);
+ if (!(xen_caps & KVM_XEN_HVM_CONFIG_SHARED_INFO) ) {
print_skip("KVM_XEN_HVM_CONFIG_SHARED_INFO not available");
exit(KSFT_SKIP);
}
+ bool do_runstate_tests = !!(xen_caps & KVM_XEN_HVM_CONFIG_RUNSTATE);
+
clock_gettime(CLOCK_REALTIME, &min_ts);
vm = vm_create_default(VCPU_ID, 0, (void *) guest_code);
@@ -80,6 +146,7 @@ int main(int argc, char *argv[])
/* Map a region for the shared_info page */
vm_userspace_mem_region_add(vm, VM_MEM_SRC_ANONYMOUS,
SHINFO_REGION_GPA, SHINFO_REGION_SLOT, 2, 0);
+ virt_map(vm, SHINFO_REGION_GVA, SHINFO_REGION_GPA, 2, 0);
struct kvm_xen_hvm_config hvmc = {
.flags = KVM_XEN_HVM_CONFIG_INTERCEPT_HCALL,
@@ -111,6 +178,17 @@ int main(int argc, char *argv[])
};
vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_SET_ATTR, &pvclock);
+ if (do_runstate_tests) {
+ struct kvm_xen_vcpu_attr st = {
+ .type = KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADDR,
+ .u.gpa = RUNSTATE_ADDR,
+ };
+ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_SET_ATTR, &st);
+ }
+
+ struct vcpu_runstate_info *rs = addr_gpa2hva(vm, RUNSTATE_ADDR);;
+ rs->state = 0x5a;
+
for (;;) {
volatile struct kvm_run *run = vcpu_state(vm, VCPU_ID);
struct ucall uc;
@@ -126,8 +204,56 @@ int main(int argc, char *argv[])
case UCALL_ABORT:
TEST_FAIL("%s", (const char *)uc.args[0]);
/* NOT REACHED */
- case UCALL_SYNC:
+ case UCALL_SYNC: {
+ struct kvm_xen_vcpu_attr rst;
+ long rundelay;
+
+ /* If no runstate support, bail out early */
+ if (!do_runstate_tests)
+ goto done;
+
+ TEST_ASSERT(rs->state_entry_time == rs->time[0] +
+ rs->time[1] + rs->time[2] + rs->time[3],
+ "runstate times don't add up");
+
+ switch (uc.args[1]) {
+ case RUNSTATE_running...RUNSTATE_offline:
+ rst.type = KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_CURRENT;
+ rst.u.runstate.state = uc.args[1];
+ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_SET_ATTR, &rst);
+ break;
+ case 4:
+ rst.type = KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_ADJUST;
+ memset(&rst.u, 0, sizeof(rst.u));
+ rst.u.runstate.state = (uint64_t)-1;
+ rst.u.runstate.time_blocked =
+ 0x5a - rs->time[RUNSTATE_blocked];
+ rst.u.runstate.time_offline =
+ 0x6b6b - rs->time[RUNSTATE_offline];
+ rst.u.runstate.time_runnable = -rst.u.runstate.time_blocked -
+ rst.u.runstate.time_offline;
+ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_SET_ATTR, &rst);
+ break;
+
+ case 5:
+ rst.type = KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA;
+ memset(&rst.u, 0, sizeof(rst.u));
+ rst.u.runstate.state = RUNSTATE_running;
+ rst.u.runstate.state_entry_time = 0x6b6b + 0x5a;
+ rst.u.runstate.time_blocked = 0x6b6b;
+ rst.u.runstate.time_offline = 0x5a;
+ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_SET_ATTR, &rst);
+ break;
+ case 6:
+ /* Yield until scheduler delay exceeds target */
+ rundelay = get_run_delay() + MIN_STEAL_TIME;
+ do {
+ sched_yield();
+ } while (get_run_delay() < rundelay);
+ break;
+ }
break;
+ }
case UCALL_DONE:
goto done;
default:
@@ -162,6 +288,33 @@ int main(int argc, char *argv[])
TEST_ASSERT(ti2->version && !(ti2->version & 1),
"Bad time_info version %x", ti->version);
+ if (do_runstate_tests) {
+ /*
+ * Fetch runstate and check sanity. Strictly speaking in the
+ * general case we might not expect the numbers to be identical
+ * but in this case we know we aren't running the vCPU any more.
+ */
+ struct kvm_xen_vcpu_attr rst = {
+ .type = KVM_XEN_VCPU_ATTR_TYPE_RUNSTATE_DATA,
+ };
+ vcpu_ioctl(vm, VCPU_ID, KVM_XEN_VCPU_GET_ATTR, &rst);
+
+ TEST_ASSERT(rs->state == rst.u.runstate.state, "Runstate mismatch");
+ TEST_ASSERT(rs->state_entry_time == rst.u.runstate.state_entry_time,
+ "State entry time mismatch");
+ TEST_ASSERT(rs->time[RUNSTATE_running] == rst.u.runstate.time_running,
+ "Running time mismatch");
+ TEST_ASSERT(rs->time[RUNSTATE_runnable] == rst.u.runstate.time_runnable,
+ "Runnable time mismatch");
+ TEST_ASSERT(rs->time[RUNSTATE_blocked] == rst.u.runstate.time_blocked,
+ "Blocked time mismatch");
+ TEST_ASSERT(rs->time[RUNSTATE_offline] == rst.u.runstate.time_offline,
+ "Offline time mismatch");
+
+ TEST_ASSERT(rs->state_entry_time == rs->time[0] +
+ rs->time[1] + rs->time[2] + rs->time[3],
+ "runstate times don't add up");
+ }
kvm_vm_free(vm);
return 0;
}