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

Pull KVM updates from Paolo Bonzini:
 "ARM:
   - data abort report and injection
   - steal time support
   - GICv4 performance improvements
   - vgic ITS emulation fixes
   - simplify FWB handling
   - enable halt polling counters
   - make the emulated timer PREEMPT_RT compliant

  s390:
   - small fixes and cleanups
   - selftest improvements
   - yield improvements

  PPC:
   - add capability to tell userspace whether we can single-step the
     guest
   - improve the allocation of XIVE virtual processor IDs
   - rewrite interrupt synthesis code to deliver interrupts in virtual
     mode when appropriate.
   - minor cleanups and improvements.

  x86:
   - XSAVES support for AMD
   - more accurate report of nested guest TSC to the nested hypervisor
   - retpoline optimizations
   - support for nested 5-level page tables
   - PMU virtualization optimizations, and improved support for nested
     PMU virtualization
   - correct latching of INITs for nested virtualization
   - IOAPIC optimization
   - TSX_CTRL virtualization for more TAA happiness
   - improved allocation and flushing of SEV ASIDs
   - many bugfixes and cleanups"

* tag 'for-linus' of git://git.kernel.org/pub/scm/virt/kvm/kvm: (127 commits)
  kvm: nVMX: Relax guest IA32_FEATURE_CONTROL constraints
  KVM: x86: Grab KVM's srcu lock when setting nested state
  KVM: x86: Open code shared_msr_update() in its only caller
  KVM: Fix jump label out_free_* in kvm_init()
  KVM: x86: Remove a spurious export of a static function
  KVM: x86: create mmu/ subdirectory
  KVM: nVMX: Remove unnecessary TLB flushes on L1<->L2 switches when L1 use apic-access-page
  KVM: x86: remove set but not used variable 'called'
  KVM: nVMX: Do not mark vmcs02->apic_access_page as dirty when unpinning
  KVM: vmx: use MSR_IA32_TSX_CTRL to hard-disable TSX on guest that lack it
  KVM: vmx: implement MSR_IA32_TSX_CTRL disable RTM functionality
  KVM: x86: implement MSR_IA32_TSX_CTRL effect on CPUID
  KVM: x86: do not modify masked bits of shared MSRs
  KVM: x86: fix presentation of TSX feature in ARCH_CAPABILITIES
  KVM: PPC: Book3S HV: XIVE: Fix potential page leak on error path
  KVM: PPC: Book3S HV: XIVE: Free previous EQ page when setting up a new one
  KVM: nVMX: Assume TLB entries of L1 and L2 are tagged differently if L0 use EPT
  KVM: x86: Unexport kvm_vcpu_reload_apic_access_page()
  KVM: nVMX: add CR4_LA57 bit to nested CR4_FIXED1
  KVM: nVMX: Use semi-colon instead of comma for exit-handlers initialization
  ...

7 files changed, 158 insertions, 21 deletions

diff --git a/tools/testing/selftests/kvm/.gitignore b/tools/testing/selftests/kvm/.gitignore
index 409c1fa75e03..30072c3f52fb 100644
--- a/tools/testing/selftests/kvm/.gitignore
+++ b/tools/testing/selftests/kvm/.gitignore
@@ -13,6 +13,7 @@
 /x86_64/vmx_dirty_log_test
 /x86_64/vmx_set_nested_state_test
 /x86_64/vmx_tsc_adjust_test
+/x86_64/xss_msr_test
 /clear_dirty_log_test
 /dirty_log_test
 /kvm_create_max_vcpus
diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index c5ec868fa1e5..3138a916574a 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -25,6 +25,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/vmx_close_while_nested_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_dirty_log_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_set_nested_state_test
 TEST_GEN_PROGS_x86_64 += x86_64/vmx_tsc_adjust_test
+TEST_GEN_PROGS_x86_64 += x86_64/xss_msr_test
 TEST_GEN_PROGS_x86_64 += clear_dirty_log_test
 TEST_GEN_PROGS_x86_64 += dirty_log_test
 TEST_GEN_PROGS_x86_64 += kvm_create_max_vcpus
diff --git a/tools/testing/selftests/kvm/include/x86_64/processor.h b/tools/testing/selftests/kvm/include/x86_64/processor.h
index ff234018219c..635ee6c33ad2 100644
--- a/tools/testing/selftests/kvm/include/x86_64/processor.h
+++ b/tools/testing/selftests/kvm/include/x86_64/processor.h
@@ -308,6 +308,8 @@ struct kvm_x86_state *vcpu_save_state(struct kvm_vm *vm, uint32_t vcpuid);
 void vcpu_load_state(struct kvm_vm *vm, uint32_t vcpuid,
 		     struct kvm_x86_state *state);
 
+struct kvm_msr_list *kvm_get_msr_index_list(void);
+
 struct kvm_cpuid2 *kvm_get_supported_cpuid(void);
 void vcpu_set_cpuid(struct kvm_vm *vm, uint32_t vcpuid,
 		    struct kvm_cpuid2 *cpuid);
@@ -322,10 +324,13 @@ kvm_get_supported_cpuid_entry(uint32_t function)
 }
 
 uint64_t vcpu_get_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index);
+int _vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
+		  uint64_t msr_value);
 void vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
 	  	  uint64_t msr_value);
 
-uint32_t kvm_get_cpuid_max(void);
+uint32_t kvm_get_cpuid_max_basic(void);
+uint32_t kvm_get_cpuid_max_extended(void);
 void kvm_get_cpu_address_width(unsigned int *pa_bits, unsigned int *va_bits);
 
 /*
diff --git a/tools/testing/selftests/kvm/kvm_create_max_vcpus.c b/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
index 231d79e57774..6f38c3dc0d56 100644
--- a/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
+++ b/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
@@ -29,12 +29,9 @@ void test_vcpu_creation(int first_vcpu_id, int num_vcpus)
 
 	vm = vm_create(VM_MODE_DEFAULT, DEFAULT_GUEST_PHY_PAGES, O_RDWR);
 
-	for (i = 0; i < num_vcpus; i++) {
-		int vcpu_id = first_vcpu_id + i;
-
+	for (i = first_vcpu_id; i < first_vcpu_id + num_vcpus; i++)
 		/* This asserts that the vCPU was created. */
-		vm_vcpu_add(vm, vcpu_id);
-	}
+		vm_vcpu_add(vm, i);
 
 	kvm_vm_free(vm);
 }
diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
index 6698cb741e10..683d3bdb8f6a 100644
--- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
+++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
@@ -869,7 +869,7 @@ uint64_t vcpu_get_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index)
 	return buffer.entry.data;
 }
 
-/* VCPU Set MSR
+/* _VCPU Set MSR
  *
  * Input Args:
  *   vm - Virtual Machine
@@ -879,12 +879,12 @@ uint64_t vcpu_get_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index)
  *
  * Output Args: None
  *
- * Return: On success, nothing. On failure a TEST_ASSERT is produced.
+ * Return: The result of KVM_SET_MSRS.
  *
- * Set value of MSR for VCPU.
+ * Sets the value of an MSR for the given VCPU.
  */
-void vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
-	uint64_t msr_value)
+int _vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
+		  uint64_t msr_value)
 {
 	struct vcpu *vcpu = vcpu_find(vm, vcpuid);
 	struct {
@@ -899,6 +899,29 @@ void vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
 	buffer.entry.index = msr_index;
 	buffer.entry.data = msr_value;
 	r = ioctl(vcpu->fd, KVM_SET_MSRS, &buffer.header);
+	return r;
+}
+
+/* VCPU Set MSR
+ *
+ * Input Args:
+ *   vm - Virtual Machine
+ *   vcpuid - VCPU ID
+ *   msr_index - Index of MSR
+ *   msr_value - New value of MSR
+ *
+ * Output Args: None
+ *
+ * Return: On success, nothing. On failure a TEST_ASSERT is produced.
+ *
+ * Set value of MSR for VCPU.
+ */
+void vcpu_set_msr(struct kvm_vm *vm, uint32_t vcpuid, uint64_t msr_index,
+	uint64_t msr_value)
+{
+	int r;
+
+	r = _vcpu_set_msr(vm, vcpuid, msr_index, msr_value);
 	TEST_ASSERT(r == 1, "KVM_SET_MSRS IOCTL failed,\n"
 		"  rc: %i errno: %i", r, errno);
 }
@@ -1000,19 +1023,45 @@ struct kvm_x86_state {
 	struct kvm_msrs msrs;
 };
 
-static int kvm_get_num_msrs(struct kvm_vm *vm)
+static int kvm_get_num_msrs_fd(int kvm_fd)
 {
 	struct kvm_msr_list nmsrs;
 	int r;
 
 	nmsrs.nmsrs = 0;
-	r = ioctl(vm->kvm_fd, KVM_GET_MSR_INDEX_LIST, &nmsrs);
+	r = ioctl(kvm_fd, KVM_GET_MSR_INDEX_LIST, &nmsrs);
 	TEST_ASSERT(r == -1 && errno == E2BIG, "Unexpected result from KVM_GET_MSR_INDEX_LIST probe, r: %i",
 		r);
 
 	return nmsrs.nmsrs;
 }
 
+static int kvm_get_num_msrs(struct kvm_vm *vm)
+{
+	return kvm_get_num_msrs_fd(vm->kvm_fd);
+}
+
+struct kvm_msr_list *kvm_get_msr_index_list(void)
+{
+	struct kvm_msr_list *list;
+	int nmsrs, r, kvm_fd;
+
+	kvm_fd = open(KVM_DEV_PATH, O_RDONLY);
+	if (kvm_fd < 0)
+		exit(KSFT_SKIP);
+
+	nmsrs = kvm_get_num_msrs_fd(kvm_fd);
+	list = malloc(sizeof(*list) + nmsrs * sizeof(list->indices[0]));
+	list->nmsrs = nmsrs;
+	r = ioctl(kvm_fd, KVM_GET_MSR_INDEX_LIST, list);
+	close(kvm_fd);
+
+	TEST_ASSERT(r == 0, "Unexpected result from KVM_GET_MSR_INDEX_LIST, r: %i",
+		r);
+
+	return list;
+}
+
 struct kvm_x86_state *vcpu_save_state(struct kvm_vm *vm, uint32_t vcpuid)
 {
 	struct vcpu *vcpu = vcpu_find(vm, vcpuid);
@@ -1158,7 +1207,12 @@ bool is_intel_cpu(void)
 	return (ebx == chunk[0] && edx == chunk[1] && ecx == chunk[2]);
 }
 
-uint32_t kvm_get_cpuid_max(void)
+uint32_t kvm_get_cpuid_max_basic(void)
+{
+	return kvm_get_supported_cpuid_entry(0)->eax;
+}
+
+uint32_t kvm_get_cpuid_max_extended(void)
 {
 	return kvm_get_supported_cpuid_entry(0x80000000)->eax;
 }
@@ -1169,7 +1223,7 @@ void kvm_get_cpu_address_width(unsigned int *pa_bits, unsigned int *va_bits)
 	bool pae;
 
 	/* SDM 4.1.4 */
-	if (kvm_get_cpuid_max() < 0x80000008) {
+	if (kvm_get_cpuid_max_extended() < 0x80000008) {
 		pae = kvm_get_supported_cpuid_entry(1)->edx & (1 << 6);
 		*pa_bits = pae ? 36 : 32;
 		*va_bits = 32;
diff --git a/tools/testing/selftests/kvm/s390x/sync_regs_test.c b/tools/testing/selftests/kvm/s390x/sync_regs_test.c
index d5290b4ad636..b705637ca14b 100644
--- a/tools/testing/selftests/kvm/s390x/sync_regs_test.c
+++ b/tools/testing/selftests/kvm/s390x/sync_regs_test.c
@@ -25,12 +25,15 @@
 
 static void guest_code(void)
 {
-	register u64 stage asm("11") = 0;
-
-	for (;;) {
-		GUEST_SYNC(0);
-		asm volatile ("ahi %0,1" : : "r"(stage));
-	}
+	/*
+	 * We embed diag 501 here instead of doing a ucall to avoid that
+	 * the compiler has messed with r11 at the time of the ucall.
+	 */
+	asm volatile (
+		"0:	diag 0,0,0x501\n"
+		"	ahi 11,1\n"
+		"	j 0b\n"
+	);
 }
 
 #define REG_COMPARE(reg) \
diff --git a/tools/testing/selftests/kvm/x86_64/xss_msr_test.c b/tools/testing/selftests/kvm/x86_64/xss_msr_test.c
new file mode 100644
index 000000000000..851ea81b9d9f
--- /dev/null
+++ b/tools/testing/selftests/kvm/x86_64/xss_msr_test.c
@@ -0,0 +1,76 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2019, Google LLC.
+ *
+ * Tests for the IA32_XSS MSR.
+ */
+
+#define _GNU_SOURCE /* for program_invocation_short_name */
+#include <sys/ioctl.h>
+
+#include "test_util.h"
+#include "kvm_util.h"
+#include "vmx.h"
+
+#define VCPU_ID	      1
+#define MSR_BITS      64
+
+#define X86_FEATURE_XSAVES	(1<<3)
+
+bool is_supported_msr(u32 msr_index)
+{
+	struct kvm_msr_list *list;
+	bool found = false;
+	int i;
+
+	list = kvm_get_msr_index_list();
+	for (i = 0; i < list->nmsrs; ++i) {
+		if (list->indices[i] == msr_index) {
+			found = true;
+			break;
+		}
+	}
+
+	free(list);
+	return found;
+}
+
+int main(int argc, char *argv[])
+{
+	struct kvm_cpuid_entry2 *entry;
+	bool xss_supported = false;
+	struct kvm_vm *vm;
+	uint64_t xss_val;
+	int i, r;
+
+	/* Create VM */
+	vm = vm_create_default(VCPU_ID, 0, 0);
+
+	if (kvm_get_cpuid_max_basic() >= 0xd) {
+		entry = kvm_get_supported_cpuid_index(0xd, 1);
+		xss_supported = entry && !!(entry->eax & X86_FEATURE_XSAVES);
+	}
+	if (!xss_supported) {
+		printf("IA32_XSS is not supported by the vCPU.\n");
+		exit(KSFT_SKIP);
+	}
+
+	xss_val = vcpu_get_msr(vm, VCPU_ID, MSR_IA32_XSS);
+	TEST_ASSERT(xss_val == 0,
+		    "MSR_IA32_XSS should be initialized to zero\n");
+
+	vcpu_set_msr(vm, VCPU_ID, MSR_IA32_XSS, xss_val);
+	/*
+	 * At present, KVM only supports a guest IA32_XSS value of 0. Verify
+	 * that trying to set the guest IA32_XSS to an unsupported value fails.
+	 * Also, in the future when a non-zero value succeeds check that
+	 * IA32_XSS is in the KVM_GET_MSR_INDEX_LIST.
+	 */
+	for (i = 0; i < MSR_BITS; ++i) {
+		r = _vcpu_set_msr(vm, VCPU_ID, MSR_IA32_XSS, 1ull << i);
+		TEST_ASSERT(r == 0 || is_supported_msr(MSR_IA32_XSS),
+			    "IA32_XSS was able to be set, but was not found in KVM_GET_MSR_INDEX_LIST.\n");
+	}
+
+	kvm_vm_free(vm);
+}