[RFC V3 PATCH 0/6] selftests: KVM: selftests for fd-based private memory

[RFC V3 PATCH 0/6] selftests: KVM: selftests for fd-based private memory

[Vishal Annapurve]

This v3 series implements selftests targeting the feature floated by Chao
via:
https://lore.kernel.org/linux-mm/20220706082016.2603916-12-chao.p.peng@linux.intel.com/T/

Below changes aim to test the fd based approach for guest private memory
in context of normal (non-confidential) VMs executing on non-confidential
platforms.

private_mem_test.c file adds selftest to access private memory from the
guest via private/shared accesses and checking if the contents can be
leaked to/accessed by vmm via shared memory view before/after conversions.

Updates in V3:
1) Series is based on v7 series from Chao
2) Changes are introduced in KVM to help execute private mem selftests
3) Selftests are executing from private memory
4) Test implementation is simplified to contain implicit/explicit memory
conversion paths according to feedback from Sean.
5) Addressed comments from Sean and Shuah.

This series has dependency on following patches:
1) V7 series patches from Chao mentioned above.
2) https://lore.kernel.org/lkml/20220810152033.946942-1-pgonda@google.com/T/#u
  - Series posted by Peter containing patches from Michael and Sean.

Github link for the patches posted as part of this series:
https://github.com/vishals4gh/linux/commits/priv_memfd_selftests_rfc_v3

Vishal Annapurve (6):
  kvm: x86: Add support for testing private memory
  selftests: kvm: Add support for private memory
  selftests: kvm: ucall: Allow querying ucall pool gpa
  selftests: kvm: x86: Execute hypercall as per the cpu
  selftests: kvm: x86: Execute VMs with private memory
  sefltests: kvm: x86: Add selftest for private memory

 arch/x86/include/uapi/asm/kvm_para.h          |   2 +
 arch/x86/kvm/Kconfig                          |   1 +
 arch/x86/kvm/mmu/mmu.c                        |  19 ++
 arch/x86/kvm/mmu/mmu_internal.h               |   2 +-
 arch/x86/kvm/x86.c                            |  67 +++-
 include/linux/kvm_host.h                      |  12 +
 tools/testing/selftests/kvm/.gitignore        |   1 +
 tools/testing/selftests/kvm/Makefile          |   2 +
 .../selftests/kvm/include/kvm_util_base.h     |  12 +-
 .../selftests/kvm/include/ucall_common.h      |   2 +
 .../kvm/include/x86_64/private_mem.h          |  51 +++
 tools/testing/selftests/kvm/lib/kvm_util.c    |  40 ++-
 .../testing/selftests/kvm/lib/ucall_common.c  |  12 +
 .../selftests/kvm/lib/x86_64/private_mem.c    | 297 ++++++++++++++++++
 .../selftests/kvm/lib/x86_64/processor.c      |  15 +-
 .../selftests/kvm/x86_64/private_mem_test.c   | 262 +++++++++++++++
 virt/kvm/Kconfig                              |   9 +
 virt/kvm/kvm_main.c                           |  90 +++++-
 18 files changed, 887 insertions(+), 9 deletions(-)
 create mode 100644 tools/testing/selftests/kvm/include/x86_64/private_mem.h
 create mode 100644 tools/testing/selftests/kvm/lib/x86_64/private_mem.c
 create mode 100644 tools/testing/selftests/kvm/x86_64/private_mem_test.c

-- 
2.37.1.595.g718a3a8f04-goog

[RFC V3 PATCH 1/6] kvm: x86: Add support for testing private memory

[Vishal Annapurve]

Introduce HAVE_KVM_PRIVATE_MEM_TESTING config to be able to test fd based
approach to support private memory with non-confidential selftest VMs.
To support this testing few important aspects need to be considered from
the perspective of selftests -
* Non-confidential VM should be able to boot from private memory:
	ENCRYPT_REG_REGION is modified to support populating initial
	private memory contents.
* Non-confidential VM needs to convey its access type to KVM:
	Introduce HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING to allow guest to
	convey the gpa ranges that need to be considered as accessed via
	private/shared access.
	memslots are modified to carry a shared bitmap that will be
	toggled when guest issues HC_MAP_GPA_RANGE hypercall with newly
	introduced attributes.
* arch_private_mem_supported and kvm_is_fault_private are added/updated
to allow private memory logic to work with non-confidential vm selftests.

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 arch/x86/include/uapi/asm/kvm_para.h |  2 +
 arch/x86/kvm/Kconfig                 |  1 +
 arch/x86/kvm/mmu/mmu.c               | 19 ++++++
 arch/x86/kvm/mmu/mmu_internal.h      |  2 +-
 arch/x86/kvm/x86.c                   | 67 ++++++++++++++++++++-
 include/linux/kvm_host.h             | 12 ++++
 virt/kvm/Kconfig                     |  9 +++
 virt/kvm/kvm_main.c                  | 90 +++++++++++++++++++++++++++-
 8 files changed, 198 insertions(+), 4 deletions(-)

diff --git a/arch/x86/include/uapi/asm/kvm_para.h b/arch/x86/include/uapi/asm/kvm_para.h
index 6e64b27b2c1e..541b7cbb3167 100644
--- a/arch/x86/include/uapi/asm/kvm_para.h
+++ b/arch/x86/include/uapi/asm/kvm_para.h
@@ -102,6 +102,8 @@ struct kvm_clock_pairing {
 #define KVM_MAP_GPA_RANGE_PAGE_SZ_2M	(1 << 0)
 #define KVM_MAP_GPA_RANGE_PAGE_SZ_1G	(1 << 1)
 #define KVM_MAP_GPA_RANGE_ENC_STAT(n)	(n << 4)
+#define KVM_MARK_GPA_RANGE_ENC_ACCESS	(1 << 8)
+#define KVM_CLR_GPA_RANGE_ENC_ACCESS	(1 << 9)
 #define KVM_MAP_GPA_RANGE_ENCRYPTED	KVM_MAP_GPA_RANGE_ENC_STAT(1)
 #define KVM_MAP_GPA_RANGE_DECRYPTED	KVM_MAP_GPA_RANGE_ENC_STAT(0)
 
diff --git a/arch/x86/kvm/Kconfig b/arch/x86/kvm/Kconfig
index 05861b9656a4..49b3d527e793 100644
--- a/arch/x86/kvm/Kconfig
+++ b/arch/x86/kvm/Kconfig
@@ -49,6 +49,7 @@ config KVM
 	select INTERVAL_TREE
 	select HAVE_KVM_PM_NOTIFIER if PM
 	select HAVE_KVM_PRIVATE_MEM if X86_64
+	select HAVE_KVM_PRIVATE_MEM_TESTING if HAVE_KVM_PRIVATE_MEM
 	select MEMFILE_NOTIFIER if HAVE_KVM_PRIVATE_MEM
 	select XARRAY_MULTI if HAVE_KVM_PRIVATE_MEM
 	help
diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index 27dbdd4fe8d1..aee6f07e7a8d 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4236,6 +4236,24 @@ static bool is_page_fault_stale(struct kvm_vcpu *vcpu,
 	       mmu_updating_retry_gfn(vcpu->kvm, mmu_seq, fault->gfn);
 }
 
+static bool kvm_is_fault_private(struct kvm *kvm, struct kvm_page_fault *fault)
+{
+	struct kvm_memory_slot *slot = fault->slot;
+	bool is_private;
+
+	if (!kvm_slot_can_be_private(slot))
+		return false;
+
+	mutex_lock(&kvm->slots_lock);
+	if (slot->shared_bitmap)
+		is_private = !test_bit(fault->gfn - slot->base_gfn, slot->shared_bitmap);
+	else
+		is_private = false;
+	mutex_unlock(&kvm->slots_lock);
+
+	return is_private;
+}
+
 static int direct_page_fault(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault)
 {
 	bool is_tdp_mmu_fault = is_tdp_mmu(vcpu->arch.mmu);
@@ -4245,6 +4263,7 @@ static int direct_page_fault(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault
 
 	fault->gfn = fault->addr >> PAGE_SHIFT;
 	fault->slot = kvm_vcpu_gfn_to_memslot(vcpu, fault->gfn);
+	fault->is_private = kvm_is_fault_private(vcpu->kvm, fault);
 
 	if (page_fault_handle_page_track(vcpu, fault))
 		return RET_PF_EMULATE;
diff --git a/arch/x86/kvm/mmu/mmu_internal.h b/arch/x86/kvm/mmu/mmu_internal.h
index fb9c298abcf0..c198fe90f421 100644
--- a/arch/x86/kvm/mmu/mmu_internal.h
+++ b/arch/x86/kvm/mmu/mmu_internal.h
@@ -188,7 +188,7 @@ struct kvm_page_fault {
 
 	/* Derived from mmu and global state.  */
 	const bool is_tdp;
-	const bool is_private;
+	bool is_private;
 	const bool nx_huge_page_workaround_enabled;
 
 	/*
diff --git a/arch/x86/kvm/x86.c b/arch/x86/kvm/x86.c
index 77d16b90045c..08a5e5b63119 100644
--- a/arch/x86/kvm/x86.c
+++ b/arch/x86/kvm/x86.c
@@ -9436,6 +9436,46 @@ static int complete_hypercall_exit(struct kvm_vcpu *vcpu)
 	return kvm_skip_emulated_instruction(vcpu);
 }
 
+#ifdef CONFIG_HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+static int kvm_update_guest_access_tracking(struct kvm *kvm, u64 attrs,
+					gfn_t gfn, u64 npages)
+{
+	int ret = 0;
+	struct kvm_memory_slot *memslot = gfn_to_memslot(kvm, gfn);
+
+	if (!memslot) {
+		pr_err("memslot doesn't exist for 0x%lx\n", gfn);
+		return -KVM_EINVAL;
+	}
+
+	mutex_lock(&kvm->slots_lock);
+	if (!kvm_slot_can_be_private(memslot)) {
+		pr_err("memslot not private for 0x%lx\n", gfn);
+		ret = -KVM_EINVAL;
+		goto err;
+	}
+
+	if (memslot->npages - (gfn - memslot->base_gfn) < npages) {
+		pr_err("memslot length insufficient for gfn 0x%lx pages 0x%lx\n",
+			gfn, npages);
+		ret = -KVM_EINVAL;
+		goto err;
+	}
+
+	if (attrs & KVM_MAP_GPA_RANGE_ENCRYPTED) {
+		bitmap_clear(memslot->shared_bitmap, (gfn - memslot->base_gfn),
+			npages);
+	} else {
+		bitmap_set(memslot->shared_bitmap, (gfn - memslot->base_gfn),
+			npages);
+	}
+
+err:
+	mutex_unlock(&kvm->slots_lock);
+	return ret;
+}
+#endif
+
 int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
 {
 	unsigned long nr, a0, a1, a2, a3, ret;
@@ -9503,17 +9543,36 @@ int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
 		break;
 	case KVM_HC_MAP_GPA_RANGE: {
 		u64 gpa = a0, npages = a1, attrs = a2;
+		bool exit_to_userspace = true;
+		gfn_t gfn;
 
 		ret = -KVM_ENOSYS;
 		if (!(vcpu->kvm->arch.hypercall_exit_enabled & (1 << KVM_HC_MAP_GPA_RANGE)))
 			break;
 
-		if (!PAGE_ALIGNED(gpa) || !npages ||
-		    gpa_to_gfn(gpa) + npages <= gpa_to_gfn(gpa)) {
+		gfn = gpa_to_gfn(gpa);
+		if (!PAGE_ALIGNED(gpa) || (gfn + npages < gfn) || !npages) {
 			ret = -KVM_EINVAL;
 			break;
 		}
 
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+		if (attrs & (KVM_MARK_GPA_RANGE_ENC_ACCESS | KVM_CLR_GPA_RANGE_ENC_ACCESS)) {
+			exit_to_userspace = false;
+			if (attrs & KVM_MARK_GPA_RANGE_ENC_ACCESS)
+				attrs |= KVM_MAP_GPA_RANGE_ENCRYPTED;
+		}
+#endif
+
+#ifdef CONFIG_HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+		ret = kvm_update_guest_access_tracking(vcpu->kvm, attrs,
+				gfn, npages);
+		if (ret)
+			break;
+#endif
+		if (!exit_to_userspace)
+			break;
+
 		vcpu->run->exit_reason        = KVM_EXIT_HYPERCALL;
 		vcpu->run->hypercall.nr       = KVM_HC_MAP_GPA_RANGE;
 		vcpu->run->hypercall.args[0]  = gpa;
@@ -10388,6 +10447,10 @@ static int vcpu_enter_guest(struct kvm_vcpu *vcpu)
 
 	preempt_disable();
 
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+	vcpu->kvm->vm_entry_attempted = true;
+#endif
+
 	static_call(kvm_x86_prepare_switch_to_guest)(vcpu);
 
 	/*
diff --git a/include/linux/kvm_host.h b/include/linux/kvm_host.h
index 4e5a0db68799..4508fa0e8fb6 100644
--- a/include/linux/kvm_host.h
+++ b/include/linux/kvm_host.h
@@ -328,6 +328,9 @@ struct kvm_vcpu {
 	u64 requests;
 	unsigned long guest_debug;
 
+	uint64_t priv_gfn;
+	uint64_t priv_pages;
+
 	struct mutex mutex;
 	struct kvm_run *run;
 
@@ -576,6 +579,7 @@ struct kvm_memory_slot {
 	gfn_t base_gfn;
 	unsigned long npages;
 	unsigned long *dirty_bitmap;
+	unsigned long *shared_bitmap;
 	struct kvm_arch_memory_slot arch;
 	unsigned long userspace_addr;
 	u32 flags;
@@ -602,6 +606,11 @@ static inline unsigned long kvm_dirty_bitmap_bytes(struct kvm_memory_slot *memsl
 	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
 }
 
+static inline unsigned long kvm_shared_bitmap_bytes(struct kvm_memory_slot *memslot)
+{
+	return ALIGN(memslot->npages, BITS_PER_LONG) / 8;
+}
+
 static inline unsigned long *kvm_second_dirty_bitmap(struct kvm_memory_slot *memslot)
 {
 	unsigned long len = kvm_dirty_bitmap_bytes(memslot);
@@ -799,6 +808,9 @@ struct kvm {
 	u32 dirty_ring_size;
 	bool vm_bugged;
 	bool vm_dead;
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+	bool vm_entry_attempted;
+#endif
 
 #ifdef CONFIG_HAVE_KVM_PM_NOTIFIER
 	struct notifier_block pm_notifier;
diff --git a/virt/kvm/Kconfig b/virt/kvm/Kconfig
index ccaff13cc5b8..f28353c237ac 100644
--- a/virt/kvm/Kconfig
+++ b/virt/kvm/Kconfig
@@ -75,3 +75,12 @@ config HAVE_KVM_PM_NOTIFIER
 
 config HAVE_KVM_PRIVATE_MEM
        bool
+
+config HAVE_KVM_PRIVATE_MEM_TESTING
+       bool
+       select HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+       depends on HAVE_KVM_PRIVATE_MEM
+
+config HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+       bool
+       depends on HAVE_KVM_PRIVATE_MEM
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 00cdfda6915f..7597949fe031 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -916,6 +916,54 @@ static int kvm_init_mmu_notifier(struct kvm *kvm)
 
 #ifdef CONFIG_HAVE_KVM_PRIVATE_MEM
 #define KVM_MEM_ATTR_PRIVATE	0x0001
+
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+static int kvm_vm_populate_private_mem(struct kvm *kvm, unsigned long gfn_start,
+	unsigned long gfn_end)
+{
+	gfn_t gfn;
+	int ret = 0;
+	kvm_pfn_t pfn;
+	struct kvm_memory_slot *memslot = gfn_to_memslot(kvm, gfn_start);
+
+	if (!memslot || !kvm_slot_can_be_private(memslot)) {
+		pr_err("Private memslot not registered for gfn 0x%lx\n",
+			gfn_start);
+		return -EINVAL;
+	}
+
+	mutex_lock(&kvm->slots_lock);
+	for (gfn = gfn_start; gfn <= gfn_end; gfn++) {
+		int order;
+		void *kvaddr;
+
+		ret = kvm_private_mem_get_pfn(memslot, gfn, &pfn, &order);
+		if (ret) {
+			pr_err("pfn not found for 0x%lx\n", gfn);
+			goto err_ret;
+		}
+
+		kvaddr = pfn_to_kaddr(pfn);
+		if (!virt_addr_valid(kvaddr)) {
+			pr_err("Invalid kvaddr 0x%lx\n", (uint64_t)kvaddr);
+			ret = -EINVAL;
+			goto err_ret;
+		}
+
+		ret = kvm_read_guest_page(kvm, gfn, kvaddr, 0, PAGE_SIZE);
+		if (ret) {
+			pr_err("guest read failed 0x%lx\n", ret);
+			goto err_ret;
+		}
+
+		kvm_private_mem_put_pfn(memslot, pfn);
+	}
+err_ret:
+	mutex_unlock(&kvm->slots_lock);
+	return ret;
+}
+#endif
+
 static int kvm_vm_ioctl_set_encrypted_region(struct kvm *kvm, unsigned int ioctl,
 					     struct kvm_enc_region *region)
 {
@@ -944,6 +992,11 @@ static int kvm_vm_ioctl_set_encrypted_region(struct kvm *kvm, unsigned int ioctl
 
 	kvm_zap_gfn_range(kvm, start, end + 1);
 
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+	if (!kvm->vm_entry_attempted && (ioctl == KVM_MEMORY_ENCRYPT_REG_REGION))
+		r = kvm_vm_populate_private_mem(kvm, start, end);
+#endif
+
 	return r;
 }
 
@@ -1046,12 +1099,22 @@ static void kvm_destroy_dirty_bitmap(struct kvm_memory_slot *memslot)
 	memslot->dirty_bitmap = NULL;
 }
 
+static void kvm_destroy_shared_bitmap(struct kvm_memory_slot *memslot)
+{
+	if (!memslot->shared_bitmap)
+		return;
+
+	kvfree(memslot->shared_bitmap);
+	memslot->shared_bitmap = NULL;
+}
+
 /* This does not remove the slot from struct kvm_memslots data structures */
 static void kvm_free_memslot(struct kvm *kvm, struct kvm_memory_slot *slot)
 {
 	if (slot->flags & KVM_MEM_PRIVATE) {
 		kvm_private_mem_unregister(slot);
 		fput(slot->private_file);
+		kvm_destroy_shared_bitmap(slot);
 	}
 
 	kvm_destroy_dirty_bitmap(slot);
@@ -1477,6 +1540,19 @@ static int kvm_alloc_dirty_bitmap(struct kvm_memory_slot *memslot)
 	return 0;
 }
 
+#ifdef CONFIG_HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+static int kvm_alloc_shared_bitmap(struct kvm_memory_slot *memslot)
+{
+	unsigned long bitmap_bytes = kvm_shared_bitmap_bytes(memslot);
+
+	memslot->shared_bitmap = __vcalloc(2, bitmap_bytes, GFP_KERNEL_ACCOUNT);
+	if (!memslot->shared_bitmap)
+		return -ENOMEM;
+
+	return 0;
+}
+#endif
+
 static struct kvm_memslots *kvm_get_inactive_memslots(struct kvm *kvm, int as_id)
 {
 	struct kvm_memslots *active = __kvm_memslots(kvm, as_id);
@@ -1612,7 +1688,11 @@ static void kvm_replace_memslot(struct kvm *kvm,
 
 bool __weak kvm_arch_private_mem_supported(struct kvm *kvm)
 {
+#ifdef CONFIG_HAVE_KVM_PRIVATE_MEM_TESTING
+	return true;
+#else
 	return false;
+#endif
 }
 
 static int check_memory_region_flags(struct kvm *kvm,
@@ -1701,6 +1781,12 @@ static int kvm_prepare_memory_region(struct kvm *kvm,
 	int r;
 
 	if (change == KVM_MR_CREATE && new->flags & KVM_MEM_PRIVATE) {
+#ifdef CONFIG_HAVE_KVM_GUEST_PRIVATE_ACCESS_TRACKING
+		r = kvm_alloc_shared_bitmap(new);
+		if (r)
+			return r;
+#endif
+
 		r = kvm_private_mem_register(new);
 		if (r)
 			return r;
@@ -1734,8 +1820,10 @@ static int kvm_prepare_memory_region(struct kvm *kvm,
 	if (r && new && new->dirty_bitmap && (!old || !old->dirty_bitmap))
 		kvm_destroy_dirty_bitmap(new);
 
-	if (r && change == KVM_MR_CREATE && new->flags & KVM_MEM_PRIVATE)
+	if (r && change == KVM_MR_CREATE && new->flags & KVM_MEM_PRIVATE) {
 		kvm_private_mem_unregister(new);
+		kvm_destroy_shared_bitmap(new);
+	}
 
 	return r;
 }
-- 
2.37.1.595.g718a3a8f04-goog

[RFC V3 PATCH 2/6] selftests: kvm: Add support for private memory

[Vishal Annapurve]

Add support for registering private memory with kvm using
KVM_SET_USER_MEMORY_REGION ioctl.

Helper function to query extended userspace mem region is introduced to
allow memory conversion.

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 .../selftests/kvm/include/kvm_util_base.h     | 12 +++++-
 tools/testing/selftests/kvm/lib/kvm_util.c    | 40 ++++++++++++++++++-
 2 files changed, 49 insertions(+), 3 deletions(-)

diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
index ff2533a32af9..dfe454f228e7 100644
--- a/tools/testing/selftests/kvm/include/kvm_util_base.h
+++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
@@ -30,7 +30,10 @@ typedef uint64_t vm_paddr_t; /* Virtual Machine (Guest) physical address */
 typedef uint64_t vm_vaddr_t; /* Virtual Machine (Guest) virtual address */
 
 struct userspace_mem_region {
-	struct kvm_userspace_memory_region region;
+	union {
+		struct kvm_userspace_memory_region region;
+		struct kvm_userspace_memory_region_ext region_ext;
+	};
 	struct sparsebit *unused_phy_pages;
 	struct sparsebit *encrypted_phy_pages;
 	int fd;
@@ -200,7 +203,7 @@ static inline bool kvm_has_cap(long cap)
 
 #define kvm_do_ioctl(fd, cmd, arg)						\
 ({										\
-	static_assert(!_IOC_SIZE(cmd) || sizeof(*arg) == _IOC_SIZE(cmd), "");	\
+	static_assert(!_IOC_SIZE(cmd) || sizeof(*arg) >= _IOC_SIZE(cmd), "");	\
 	ioctl(fd, cmd, arg);							\
 })
 
@@ -388,6 +391,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 void vm_mem_region_set_flags(struct kvm_vm *vm, uint32_t slot, uint32_t flags);
 void vm_mem_region_move(struct kvm_vm *vm, uint32_t slot, uint64_t new_gpa);
 void vm_mem_region_delete(struct kvm_vm *vm, uint32_t slot);
+
 struct kvm_vcpu *__vm_vcpu_add(struct kvm_vm *vm, uint32_t vcpu_id);
 vm_vaddr_t vm_vaddr_alloc(struct kvm_vm *vm, size_t sz, vm_vaddr_t vaddr_min);
 vm_vaddr_t vm_vaddr_alloc_shared(struct kvm_vm *vm, size_t sz, vm_vaddr_t vaddr_min);
@@ -717,6 +721,10 @@ struct kvm_userspace_memory_region *
 kvm_userspace_memory_region_find(struct kvm_vm *vm, uint64_t start,
 				 uint64_t end);
 
+struct kvm_userspace_memory_region_ext *
+kvm_userspace_memory_region_ext_find(struct kvm_vm *vm, uint64_t start,
+				 uint64_t end);
+
 #define sync_global_to_guest(vm, g) ({				\
 	typeof(g) *_p = addr_gva2hva(vm, (vm_vaddr_t)&(g));	\
 	memcpy(_p, &(g), sizeof(g));				\
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index 171bc20a097a..f153c71d6988 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -17,6 +17,7 @@
 #include <unistd.h>
 #include <linux/kernel.h>
 
+#define MFD_INACCESSIBLE  0x0008U
 #define KVM_UTIL_MIN_PFN	2
 
 static int vcpu_mmap_sz(void);
@@ -466,6 +467,35 @@ kvm_userspace_memory_region_find(struct kvm_vm *vm, uint64_t start,
 	return &region->region;
 }
 
+/*
+ * KVM Userspace Memory Region Ext Find
+ *
+ * Input Args:
+ *   vm - Virtual Machine
+ *   start - Starting VM physical address
+ *   end - Ending VM physical address, inclusive.
+ *
+ * Output Args: None
+ *
+ * Return:
+ *   Pointer to overlapping ext region, NULL if no such region.
+ *
+ * Public interface to userspace_mem_region_find. Allows tests to look up
+ * the memslot datastructure for a given range of guest physical memory.
+ */
+struct kvm_userspace_memory_region_ext *
+kvm_userspace_memory_region_ext_find(struct kvm_vm *vm, uint64_t start,
+				 uint64_t end)
+{
+	struct userspace_mem_region *region;
+
+	region = userspace_mem_region_find(vm, start, end);
+	if (!region)
+		return NULL;
+
+	return &region->region_ext;
+}
+
 /*
  * VM VCPU Remove
  *
@@ -764,6 +794,7 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 	struct userspace_mem_region *region;
 	size_t backing_src_pagesz = get_backing_src_pagesz(src_type);
 	size_t alignment;
+	int priv_memfd = -1;
 
 	TEST_ASSERT(vm_adjust_num_guest_pages(vm->mode, npages) == npages,
 		"Number of guest pages is not compatible with the host. "
@@ -869,6 +900,11 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 			    vm_mem_backing_src_alias(src_type)->name);
 	}
 
+	if (flags & KVM_MEM_PRIVATE) {
+		priv_memfd = memfd_create("vm_private_mem_", MFD_INACCESSIBLE);
+		TEST_ASSERT(priv_memfd != -1, "Failed to create private memfd");
+	}
+
 	region->unused_phy_pages = sparsebit_alloc();
 	region->encrypted_phy_pages = sparsebit_alloc();
 	sparsebit_set_num(region->unused_phy_pages,
@@ -878,7 +914,9 @@ void vm_userspace_mem_region_add(struct kvm_vm *vm,
 	region->region.guest_phys_addr = guest_paddr;
 	region->region.memory_size = npages * vm->page_size;
 	region->region.userspace_addr = (uintptr_t) region->host_mem;
-	ret = __vm_ioctl(vm, KVM_SET_USER_MEMORY_REGION, &region->region);
+	region->region_ext.private_fd = priv_memfd;
+	region->region_ext.private_offset = 0;
+	ret = ioctl(vm->fd, KVM_SET_USER_MEMORY_REGION, &region->region_ext);
 	TEST_ASSERT(ret == 0, "KVM_SET_USER_MEMORY_REGION IOCTL failed,\n"
 		"  rc: %i errno: %i\n"
 		"  slot: %u flags: 0x%x\n"
-- 
2.37.1.595.g718a3a8f04-goog

[RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Vishal Annapurve]

Add a helper to query guest physical address for ucall pool
so that guest can mark the page as accessed shared or private.

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 tools/testing/selftests/kvm/include/ucall_common.h |  2 ++
 tools/testing/selftests/kvm/lib/ucall_common.c     | 12 ++++++++++++
 2 files changed, 14 insertions(+)

diff --git a/tools/testing/selftests/kvm/include/ucall_common.h b/tools/testing/selftests/kvm/include/ucall_common.h
index 279bbab011c7..2c6e5c4df012 100644
--- a/tools/testing/selftests/kvm/include/ucall_common.h
+++ b/tools/testing/selftests/kvm/include/ucall_common.h
@@ -31,6 +31,8 @@ void ucall_arch_uninit(struct kvm_vm *vm);
 void ucall_arch_do_ucall(vm_vaddr_t uc);
 void *ucall_arch_get_ucall(struct kvm_vcpu *vcpu);
 
+vm_paddr_t get_ucall_pool_paddr(void);
+
 void ucall(uint64_t cmd, int nargs, ...);
 uint64_t get_ucall(struct kvm_vcpu *vcpu, struct ucall *uc);
 
diff --git a/tools/testing/selftests/kvm/lib/ucall_common.c b/tools/testing/selftests/kvm/lib/ucall_common.c
index 5a15fa39cd51..4d2abef8ee77 100644
--- a/tools/testing/selftests/kvm/lib/ucall_common.c
+++ b/tools/testing/selftests/kvm/lib/ucall_common.c
@@ -11,6 +11,7 @@ struct ucall_header {
 
 static bool use_ucall_pool;
 static struct ucall_header *ucall_pool;
+static vm_paddr_t ucall_page_paddr;
 
 void ucall_init(struct kvm_vm *vm, void *arg)
 {
@@ -35,7 +36,10 @@ void ucall_init(struct kvm_vm *vm, void *arg)
 	}
 
 	ucall_pool = (struct ucall_header *)vaddr;
+	ucall_page_paddr = addr_gva2gpa(vm, vaddr);
 	sync_global_to_guest(vm, ucall_pool);
+	sync_global_to_guest(vm, ucall_page_paddr);
+	printf("ucall_page_paddr 0x%lx\n", ucall_page_paddr);
 
 out:
 	ucall_arch_init(vm, arg);
@@ -54,6 +58,14 @@ void ucall_uninit(struct kvm_vm *vm)
 	ucall_arch_uninit(vm);
 }
 
+vm_paddr_t get_ucall_pool_paddr(void)
+{
+	if (!use_ucall_pool)
+		return 0;
+
+	return ucall_page_paddr;
+}
+
 static struct ucall *ucall_alloc(void)
 {
 	struct ucall *uc = NULL;
-- 
2.37.1.595.g718a3a8f04-goog

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Sean Christopherson]

On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> Add a helper to query guest physical address for ucall pool
> so that guest can mark the page as accessed shared or private.
> 
> Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> ---

This should be handled by the SEV series[*].  Can you provide feedback on that
series if having a generic way to map the ucall address as shared won't work?

[*] https://lore.kernel.org/all/20220829171021.701198-1-pgonda@google.com

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Vishal Annapurve]

On Fri, Oct 7, 2022 at 1:32 AM Sean Christopherson <seanjc@google.com> wrote:
>
> On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > Add a helper to query guest physical address for ucall pool
> > so that guest can mark the page as accessed shared or private.
> >
> > Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> > ---
>
> This should be handled by the SEV series[*].  Can you provide feedback on that
> series if having a generic way to map the ucall address as shared won't work?
>
> [*] https://lore.kernel.org/all/20220829171021.701198-1-pgonda@google.com

Based on the SEV series you referred to, selftests are capable of
accessing ucall pool memory by having encryption bit cleared (as set
by guest pagetables) as allowed by generic API vm_vaddr_alloc_shared.
This change is needed in the context of fd based private memory where
guest (specifically non-confidential/sev guests) code in the selftests
will have to explicitly indicate that ucall pool address range will be
accessed by guest as shared.

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Sean Christopherson]

On Fri, Oct 14, 2022, Vishal Annapurve wrote:
> On Fri, Oct 7, 2022 at 1:32 AM Sean Christopherson <seanjc@google.com> wrote:
> >
> > On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > > Add a helper to query guest physical address for ucall pool
> > > so that guest can mark the page as accessed shared or private.
> > >
> > > Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> > > ---
> >
> > This should be handled by the SEV series[*].  Can you provide feedback on that
> > series if having a generic way to map the ucall address as shared won't work?
> >
> > [*] https://lore.kernel.org/all/20220829171021.701198-1-pgonda@google.com
> 
> Based on the SEV series you referred to, selftests are capable of
> accessing ucall pool memory by having encryption bit cleared (as set
> by guest pagetables) as allowed by generic API vm_vaddr_alloc_shared.
> This change is needed in the context of fd based private memory where
> guest (specifically non-confidential/sev guests) code in the selftests
> will have to explicitly indicate that ucall pool address range will be
> accessed by guest as shared.

Ah, right, the conversion needs an explicit hypercall, which gets downright
annoying because auto-converting shared pages would effectivfely require injecting
code into the start of every guest.

Ha!  I think we got too fancy.  This is purely for testing UPM, not any kind of
trust model, i.e. there's no need for KVM to treat userspace as untrusted.  Rather
than jump through hoops just to let the guest dictate private vs. shared, simply
"trust" userspace when determining whether a page should be mapped private.  Then
the selftests can invoke the repurposed KVM_MEMORY_ENCRYPT_(UN)REG_REGION ioctls
as appropriate when allocating/remapping guest private memory.

E.g. on top of UPM v8, I think the test hook boils down to:

diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
index d68944f07b4b..d42d0e6bdd8c 100644
--- a/arch/x86/kvm/mmu/mmu.c
+++ b/arch/x86/kvm/mmu/mmu.c
@@ -4279,6 +4279,9 @@ static int direct_page_fault(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault
 
        fault->gfn = fault->addr >> PAGE_SHIFT;
        fault->slot = kvm_vcpu_gfn_to_memslot(vcpu, fault->gfn);
+       fault->is_private = IS_ENABLED(CONFIG_KVM_PRIVATE_MEM_TESTING) &&
+                           kvm_slot_can_be_private(fault->slot) &&
+                           kvm_mem_is_private(vcpu->kvm, fault->gfn);
 
        if (page_fault_handle_page_track(vcpu, fault))
                return RET_PF_EMULATE;
diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
index 8ffd4607c7d8..0dc5d0bf647c 100644
--- a/virt/kvm/kvm_main.c
+++ b/virt/kvm/kvm_main.c
@@ -1653,7 +1653,7 @@ static void kvm_replace_memslot(struct kvm *kvm,
 
 bool __weak kvm_arch_has_private_mem(struct kvm *kvm)
 {
-       return false;
+       return IS_ENABLED(CONFIG_KVM_PRIVATE_MEM_TESTING);
 }
 
 static int check_memory_region_flags(struct kvm *kvm,

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Vishal Annapurve]

On Sat, Oct 15, 2022 at 12:17 AM Sean Christopherson <seanjc@google.com> wrote:
>
> On Fri, Oct 14, 2022, Vishal Annapurve wrote:
> > On Fri, Oct 7, 2022 at 1:32 AM Sean Christopherson <seanjc@google.com> wrote:
> > >
> > > On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > > > Add a helper to query guest physical address for ucall pool
> > > > so that guest can mark the page as accessed shared or private.
> > > >
> > > > Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> > > > ---
> > >
> > > This should be handled by the SEV series[*].  Can you provide feedback on that
> > > series if having a generic way to map the ucall address as shared won't work?
> > >
> > > [*] https://lore.kernel.org/all/20220829171021.701198-1-pgonda@google.com
> >
> > Based on the SEV series you referred to, selftests are capable of
> > accessing ucall pool memory by having encryption bit cleared (as set
> > by guest pagetables) as allowed by generic API vm_vaddr_alloc_shared.
> > This change is needed in the context of fd based private memory where
> > guest (specifically non-confidential/sev guests) code in the selftests
> > will have to explicitly indicate that ucall pool address range will be
> > accessed by guest as shared.
>
> Ah, right, the conversion needs an explicit hypercall, which gets downright
> annoying because auto-converting shared pages would effectivfely require injecting
> code into the start of every guest.
>
Ack.

> Ha!  I think we got too fancy.  This is purely for testing UPM, not any kind of
> trust model, i.e. there's no need for KVM to treat userspace as untrusted.  Rather
> than jump through hoops just to let the guest dictate private vs. shared, simply
> "trust" userspace when determining whether a page should be mapped private.  Then
> the selftests can invoke the repurposed KVM_MEMORY_ENCRYPT_(UN)REG_REGION ioctls
> as appropriate when allocating/remapping guest private memory.
>
> E.g. on top of UPM v8, I think the test hook boils down to:
>
> diff --git a/arch/x86/kvm/mmu/mmu.c b/arch/x86/kvm/mmu/mmu.c
> index d68944f07b4b..d42d0e6bdd8c 100644
> --- a/arch/x86/kvm/mmu/mmu.c
> +++ b/arch/x86/kvm/mmu/mmu.c
> @@ -4279,6 +4279,9 @@ static int direct_page_fault(struct kvm_vcpu *vcpu, struct kvm_page_fault *fault
>
>         fault->gfn = fault->addr >> PAGE_SHIFT;
>         fault->slot = kvm_vcpu_gfn_to_memslot(vcpu, fault->gfn);
> +       fault->is_private = IS_ENABLED(CONFIG_KVM_PRIVATE_MEM_TESTING) &&
> +                           kvm_slot_can_be_private(fault->slot) &&
> +                           kvm_mem_is_private(vcpu->kvm, fault->gfn);
>
>         if (page_fault_handle_page_track(vcpu, fault))
>                 return RET_PF_EMULATE;
> diff --git a/virt/kvm/kvm_main.c b/virt/kvm/kvm_main.c
> index 8ffd4607c7d8..0dc5d0bf647c 100644
> --- a/virt/kvm/kvm_main.c
> +++ b/virt/kvm/kvm_main.c
> @@ -1653,7 +1653,7 @@ static void kvm_replace_memslot(struct kvm *kvm,
>
>  bool __weak kvm_arch_has_private_mem(struct kvm *kvm)
>  {
> -       return false;
> +       return IS_ENABLED(CONFIG_KVM_PRIVATE_MEM_TESTING);
>  }
>
>  static int check_memory_region_flags(struct kvm *kvm,

This is much sleeker and will avoid hacking KVM for testing. Only
caveat here is that these tests will not be able to exercise implicit
conversion path if we go this route.

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Sean Christopherson]

On Mon, Oct 17, 2022, Vishal Annapurve wrote:
> This is much sleeker and will avoid hacking KVM for testing. Only
> caveat here is that these tests will not be able to exercise implicit
> conversion path if we go this route.

Yeah, I think that's a perfectly fine tradeoff.  Implicit conversion isn't strictly
a UPM feature, e.g. if TDX and SNP "architecturally" disallowed implicit conversions,
then KVM wouldn't need to support implicit conversions at all, i.e. that testing can
be punted to SNP and/or TDX selftests.

Re: [RFC V3 PATCH 3/6] selftests: kvm: ucall: Allow querying ucall pool gpa

[Vishal Annapurve]

On Mon, Oct 17, 2022 at 11:38 PM Sean Christopherson <seanjc@google.com> wrote:
>
> On Mon, Oct 17, 2022, Vishal Annapurve wrote:
> > This is much sleeker and will avoid hacking KVM for testing. Only
> > caveat here is that these tests will not be able to exercise implicit
> > conversion path if we go this route.
>
> Yeah, I think that's a perfectly fine tradeoff.  Implicit conversion isn't strictly
> a UPM feature, e.g. if TDX and SNP "architecturally" disallowed implicit conversions,
> then KVM wouldn't need to support implicit conversions at all, i.e. that testing can
> be punted to SNP and/or TDX selftests.

Ack. Will address this feedback in the next series.

[RFC V3 PATCH 4/6] selftests: kvm: x86: Execute hypercall as per the cpu

[Vishal Annapurve]

Add support for executing vmmcall/vmcall instruction on amd/intel cpus.
In general kvm patches the instruction according to the cpu
implementation at runtime. While executing selftest vms from private
memory KVM will not be able to update the private memory of the guest.

Hypercall parameters are fixed to explicitly populate hypercall number
in eax. Otherwise inlined function calls to kvm_hypercall would call
vmmcall/vmcall instruction without updating eax with hypercall number.

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 .../testing/selftests/kvm/lib/x86_64/processor.c  | 15 +++++++++++++--
 1 file changed, 13 insertions(+), 2 deletions(-)

diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
index 53b115876417..09d757a0b148 100644
--- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
+++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
@@ -1254,10 +1254,21 @@ uint64_t kvm_hypercall(uint64_t nr, uint64_t a0, uint64_t a1, uint64_t a2,
 		       uint64_t a3)
 {
 	uint64_t r;
+	static bool is_cpu_checked;
+	static bool is_cpu_amd;
 
-	asm volatile("vmcall"
+	if (!is_cpu_checked)
+		is_cpu_amd = is_amd_cpu();
+
+	if (is_cpu_amd) {
+		asm volatile("vmmcall"
+		     : "=a"(r)
+		     : "a"(nr), "b"(a0), "c"(a1), "d"(a2), "S"(a3));
+	} else {
+		asm volatile("vmcall"
 		     : "=a"(r)
-		     : "b"(a0), "c"(a1), "d"(a2), "S"(a3));
+		     : "a"(nr), "b"(a0), "c"(a1), "d"(a2), "S"(a3));
+	}
 	return r;
 }
 
-- 
2.37.1.595.g718a3a8f04-goog

Re: [RFC V3 PATCH 4/6] selftests: kvm: x86: Execute hypercall as per the cpu

[Sean Christopherson]

On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> Add support for executing vmmcall/vmcall instruction on amd/intel cpus.
> In general kvm patches the instruction according to the cpu
> implementation at runtime. While executing selftest vms from private
> memory KVM will not be able to update the private memory of the guest.
> 
> Hypercall parameters are fixed to explicitly populate hypercall number
> in eax. Otherwise inlined function calls to kvm_hypercall would call
> vmmcall/vmcall instruction without updating eax with hypercall number.

Can you send a seperate non-RFC series to clean up the selftests mess?  kvm_hypercall()
isn't the only culprit.

  git grep \"vmcall | wc -l
  16

I'm pretty sure things work only because of KVM's dubious behavior of patching
VMCALL/VMMCALL by default.

Note, svm_vmcall_test.c intentionally uses the wrong instructions and shouldn't
be converted.  Actually, we can and should just delete that test, it's superseded
by the wonderfully named fix_hypercall_test.

> Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> ---
>  .../testing/selftests/kvm/lib/x86_64/processor.c  | 15 +++++++++++++--
>  1 file changed, 13 insertions(+), 2 deletions(-)
> 
> diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> index 53b115876417..09d757a0b148 100644
> --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
> +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> @@ -1254,10 +1254,21 @@ uint64_t kvm_hypercall(uint64_t nr, uint64_t a0, uint64_t a1, uint64_t a2,
>  		       uint64_t a3)
>  {
>  	uint64_t r;
> +	static bool is_cpu_checked;
> +	static bool is_cpu_amd;
>  
> -	asm volatile("vmcall"
> +	if (!is_cpu_checked)
> +		is_cpu_amd = is_amd_cpu();

This can be done using a single int, e.g.

	static bool is_cpu_amd = -1;

	if (is_cpu_amd < 0)
		is_cpu_amd = is_amd_cpu();

Although... what if we declare main() in lib/kvm_util.c (or maybe a dedicated
file?) and rename all tests to use __main()?  Then add an arch hook to do global
initialization and avoid the "did we check CPUID?!?!?" altogether.

That would allow us to dedup all of the hilarious copy paste:

	/* Tell stdout not to buffer its content */
	setbuf(stdout, NULL);

and we could turn is_amd_cpu() and is_intel_cpu() into bools.

E.g.

int main(int argc, char *argv[])
{
	/* Tell stdout not to buffer its content */
	setbuf(stdout, NULL);

	kvm_arch_main();

	return __main(argc, argv);
}

void kvm_arch_main(void)
{
	is_cpu_amd = cpu_vendor_string_is("AuthenticAMD");
	is_cpu_intel = cpu_vendor_string_is("AuthenticAMD");
}


And then we just need macro magic to emit the right VMCALL/VMMCALL instruction.

Re: [RFC V3 PATCH 4/6] selftests: kvm: x86: Execute hypercall as per the cpu

[Vishal Annapurve]

On Wed, Aug 24, 2022 at 5:07 PM Sean Christopherson <seanjc@google.com> wrote:
>
> On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > Add support for executing vmmcall/vmcall instruction on amd/intel cpus.
> > In general kvm patches the instruction according to the cpu
> > implementation at runtime. While executing selftest vms from private
> > memory KVM will not be able to update the private memory of the guest.
> >
> > Hypercall parameters are fixed to explicitly populate hypercall number
> > in eax. Otherwise inlined function calls to kvm_hypercall would call
> > vmmcall/vmcall instruction without updating eax with hypercall number.
>
> Can you send a seperate non-RFC series to clean up the selftests mess?  kvm_hypercall()
> isn't the only culprit.
>
>   git grep \"vmcall | wc -l
>   16
>
> I'm pretty sure things work only because of KVM's dubious behavior of patching
> VMCALL/VMMCALL by default.
>
> Note, svm_vmcall_test.c intentionally uses the wrong instructions and shouldn't
> be converted.  Actually, we can and should just delete that test, it's superseded
> by the wonderfully named fix_hypercall_test.
>
> > Signed-off-by: Vishal Annapurve <vannapurve@google.com>
> > ---
> >  .../testing/selftests/kvm/lib/x86_64/processor.c  | 15 +++++++++++++--
> >  1 file changed, 13 insertions(+), 2 deletions(-)
> >
> > diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > index 53b115876417..09d757a0b148 100644
> > --- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > +++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
> > @@ -1254,10 +1254,21 @@ uint64_t kvm_hypercall(uint64_t nr, uint64_t a0, uint64_t a1, uint64_t a2,
> >                      uint64_t a3)
> >  {
> >       uint64_t r;
> > +     static bool is_cpu_checked;
> > +     static bool is_cpu_amd;
> >
> > -     asm volatile("vmcall"
> > +     if (!is_cpu_checked)
> > +             is_cpu_amd = is_amd_cpu();
>
> This can be done using a single int, e.g.
>
>         static bool is_cpu_amd = -1;
>
>         if (is_cpu_amd < 0)
>                 is_cpu_amd = is_amd_cpu();
>
> Although... what if we declare main() in lib/kvm_util.c (or maybe a dedicated
> file?) and rename all tests to use __main()?  Then add an arch hook to do global
> initialization and avoid the "did we check CPUID?!?!?" altogether.
>
> That would allow us to dedup all of the hilarious copy paste:
>
>         /* Tell stdout not to buffer its content */
>         setbuf(stdout, NULL);
>
> and we could turn is_amd_cpu() and is_intel_cpu() into bools.
>
> E.g.
>
> int main(int argc, char *argv[])
> {
>         /* Tell stdout not to buffer its content */
>         setbuf(stdout, NULL);
>
>         kvm_arch_main();
>
>         return __main(argc, argv);
> }
>
> void kvm_arch_main(void)
> {
>         is_cpu_amd = cpu_vendor_string_is("AuthenticAMD");
>         is_cpu_intel = cpu_vendor_string_is("AuthenticAMD");
> }
>
>
> And then we just need macro magic to emit the right VMCALL/VMMCALL instruction.

Thanks Sean for the feedback here. I have posted a separate series
addressing your comments:
https://lore.kernel.org/all/20220903012849.938069-4-vannapurve@google.com/T/

Regards,
Vishal

[RFC V3 PATCH 5/6] selftests: kvm: x86: Execute VMs with private memory

[Vishal Annapurve]

Introduce a set of APIs to execute VM with private memslots.

Host userspace APIs for:
1) Setting up and executing VM having private memslots
2) Backing/unbacking guest private memory

Guest APIs for:
1) Changing memory access type and memory type
2) Setting ucall page access type as shared

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 tools/testing/selftests/kvm/Makefile          |   1 +
 .../kvm/include/x86_64/private_mem.h          |  51 +++
 .../selftests/kvm/lib/x86_64/private_mem.c    | 297 ++++++++++++++++++
 3 files changed, 349 insertions(+)
 create mode 100644 tools/testing/selftests/kvm/include/x86_64/private_mem.h
 create mode 100644 tools/testing/selftests/kvm/lib/x86_64/private_mem.c

diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index 41fd1171fbd8..8fe72a60aef0 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -51,6 +51,7 @@ LIBKVM += lib/ucall_common.c
 LIBKVM_x86_64 += lib/x86_64/apic.c
 LIBKVM_x86_64 += lib/x86_64/handlers.S
 LIBKVM_x86_64 += lib/x86_64/perf_test_util.c
+LIBKVM_x86_64 += lib/x86_64/private_mem.c
 LIBKVM_x86_64 += lib/x86_64/processor.c
 LIBKVM_x86_64 += lib/x86_64/svm.c
 LIBKVM_x86_64 += lib/x86_64/ucall.c
diff --git a/tools/testing/selftests/kvm/include/x86_64/private_mem.h b/tools/testing/selftests/kvm/include/x86_64/private_mem.h
new file mode 100644
index 000000000000..00cab7b84f2c
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/x86_64/private_mem.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * Copyright (C) 2022, Google LLC.
+ */
+
+#ifndef SELFTEST_KVM_PRIVATE_MEM_H
+#define SELFTEST_KVM_PRIVATE_MEM_H
+
+#include <stdint.h>
+#include <kvm_util.h>
+
+enum mem_conversion_type {
+	TO_PRIVATE,
+	TO_SHARED
+};
+
+void guest_update_mem_access(enum mem_conversion_type type, uint64_t gpa,
+	uint64_t size);
+void guest_update_mem_map(enum mem_conversion_type type, uint64_t gpa,
+	uint64_t size);
+
+void guest_map_ucall_page_shared(void);
+
+enum mem_op {
+	ALLOCATE_MEM,
+	UNBACK_MEM
+};
+
+void vm_update_private_mem(struct kvm_vm *vm, uint64_t gpa, uint64_t size,
+	enum mem_op op);
+
+typedef void (*guest_code_fn)(void);
+typedef void (*io_exit_handler)(struct kvm_vm *vm, uint32_t uc_arg1);
+
+struct test_setup_info {
+	uint64_t test_area_gpa;
+	uint64_t test_area_size;
+	uint32_t test_area_slot;
+};
+
+struct vm_setup_info {
+	enum vm_mem_backing_src_type vm_mem_src;
+	uint32_t memslot0_pages;
+	struct test_setup_info test_info;
+	guest_code_fn guest_fn;
+	io_exit_handler ioexit_cb;
+};
+
+void execute_vm_with_private_mem(struct vm_setup_info *info);
+
+#endif /* SELFTEST_KVM_PRIVATE_MEM_H */
diff --git a/tools/testing/selftests/kvm/lib/x86_64/private_mem.c b/tools/testing/selftests/kvm/lib/x86_64/private_mem.c
new file mode 100644
index 000000000000..000584219045
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/x86_64/private_mem.c
@@ -0,0 +1,297 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * tools/testing/selftests/kvm/lib/kvm_util.c
+ *
+ * Copyright (C) 2022, Google LLC.
+ */
+#define _GNU_SOURCE /* for program_invocation_name */
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/ioctl.h>
+
+#include <linux/compiler.h>
+#include <linux/kernel.h>
+#include <linux/kvm_para.h>
+
+#include <test_util.h>
+#include <kvm_util.h>
+#include <private_mem.h>
+#include <processor.h>
+
+/*
+ * Execute KVM hypercall to change memory access type for a given gpa range.
+ *
+ * Input Args:
+ *   type - memory conversion type TO_SHARED/TO_PRIVATE
+ *   gpa - starting gpa address
+ *   size - size of the range starting from gpa for which memory access needs
+ *     to be changed
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Function called by guest logic in selftests to update the memory access type
+ * for a given gpa range. This API is useful in exercising implicit conversion
+ * path.
+ */
+void guest_update_mem_access(enum mem_conversion_type type, uint64_t gpa,
+	uint64_t size)
+{
+	int ret = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> MIN_PAGE_SHIFT,
+		type == TO_PRIVATE ? KVM_MARK_GPA_RANGE_ENC_ACCESS :
+			KVM_CLR_GPA_RANGE_ENC_ACCESS, 0);
+	GUEST_ASSERT_1(!ret, ret);
+}
+
+/*
+ * Execute KVM hypercall to change memory type for a given gpa range.
+ *
+ * Input Args:
+ *   type - memory conversion type TO_SHARED/TO_PRIVATE
+ *   gpa - starting gpa address
+ *   size - size of the range starting from gpa for which memory type needs
+ *     to be changed
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Function called by guest logic in selftests to update the memory type for a
+ * given gpa range. This API is useful in exercising explicit conversion path.
+ */
+void guest_update_mem_map(enum mem_conversion_type type, uint64_t gpa,
+	uint64_t size)
+{
+	int ret = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> MIN_PAGE_SHIFT,
+		type == TO_PRIVATE ? KVM_MAP_GPA_RANGE_ENCRYPTED :
+			KVM_MAP_GPA_RANGE_DECRYPTED, 0);
+	GUEST_ASSERT_1(!ret, ret);
+}
+
+/*
+ * Execute KVM hypercall to change memory access type for ucall page.
+ *
+ * Input Args: None
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Function called by guest logic in selftests to update the memory access type
+ * for ucall page since by default all the accesses from guest to private
+ * memslot are treated as private accesses.
+ */
+void guest_map_ucall_page_shared(void)
+{
+	vm_paddr_t ucall_paddr = get_ucall_pool_paddr();
+
+	guest_update_mem_access(TO_SHARED, ucall_paddr, 1 << MIN_PAGE_SHIFT);
+}
+
+/*
+ * Execute KVM ioctl to back/unback private memory for given gpa range.
+ *
+ * Input Args:
+ *   vm - kvm_vm handle
+ *   gpa - starting gpa address
+ *   size - size of the gpa range
+ *   op - mem_op indicating whether private memory needs to be allocated or
+ *     unbacked
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Function called by host userspace logic in selftests to back/unback private
+ * memory for gpa ranges. This function is useful to setup initial boot private
+ * memory and then convert memory during runtime.
+ */
+void vm_update_private_mem(struct kvm_vm *vm, uint64_t gpa, uint64_t size,
+	enum mem_op op)
+{
+	int priv_memfd;
+	uint64_t priv_offset, guest_phys_base, fd_offset;
+	struct kvm_enc_region enc_region;
+	struct kvm_userspace_memory_region_ext *region_ext;
+	struct kvm_userspace_memory_region *region;
+	int fallocate_mode = 0;
+	int ret;
+
+	region_ext = kvm_userspace_memory_region_ext_find(vm, gpa, gpa + size);
+	TEST_ASSERT(region_ext != NULL, "Region not found");
+	region = &region_ext->region;
+	TEST_ASSERT(region->flags & KVM_MEM_PRIVATE,
+		"Can not update private memfd for non-private memslot\n");
+	priv_memfd = region_ext->private_fd;
+	priv_offset = region_ext->private_offset;
+	guest_phys_base = region->guest_phys_addr;
+	fd_offset = priv_offset + (gpa - guest_phys_base);
+
+	if (op == UNBACK_MEM)
+		fallocate_mode = (FALLOC_FL_PUNCH_HOLE | FALLOC_FL_KEEP_SIZE);
+
+	printf("priv_memfd %d fallocate_mode 0x%x for offset 0x%lx size 0x%lx\n",
+		priv_memfd, fallocate_mode, fd_offset, size);
+	ret = fallocate(priv_memfd, fallocate_mode, fd_offset, size);
+	TEST_ASSERT(ret == 0, "fallocate failed\n");
+	enc_region.addr = gpa;
+	enc_region.size = size;
+	if (op == ALLOCATE_MEM) {
+		printf("doing encryption for gpa 0x%lx size 0x%lx\n", gpa, size);
+		vm_ioctl(vm, KVM_MEMORY_ENCRYPT_REG_REGION, &enc_region);
+	} else {
+		printf("undoing encryption for gpa 0x%lx size 0x%lx\n", gpa, size);
+		vm_ioctl(vm, KVM_MEMORY_ENCRYPT_UNREG_REGION, &enc_region);
+	}
+}
+
+static void handle_vm_exit_map_gpa_hypercall(struct kvm_vm *vm,
+				volatile struct kvm_run *run)
+{
+	uint64_t gpa, npages, attrs, size;
+
+	TEST_ASSERT(run->hypercall.nr == KVM_HC_MAP_GPA_RANGE,
+		"Unhandled Hypercall %lld\n", run->hypercall.nr);
+	gpa = run->hypercall.args[0];
+	npages = run->hypercall.args[1];
+	size = npages << MIN_PAGE_SHIFT;
+	attrs = run->hypercall.args[2];
+	pr_info("Explicit conversion off 0x%lx size 0x%lx to %s\n", gpa, size,
+		(attrs & KVM_MAP_GPA_RANGE_ENCRYPTED) ? "private" : "shared");
+
+	if (attrs & KVM_MAP_GPA_RANGE_ENCRYPTED)
+		vm_update_private_mem(vm, gpa, size, ALLOCATE_MEM);
+	else
+		vm_update_private_mem(vm, gpa, size, UNBACK_MEM);
+
+	run->hypercall.ret = 0;
+}
+
+static void handle_vm_exit_memory_error(struct kvm_vm *vm, volatile struct kvm_run *run)
+{
+	uint64_t gpa, size, flags;
+
+	gpa = run->memory.gpa;
+	size = run->memory.size;
+	flags = run->memory.flags;
+	pr_info("Implicit conversion off 0x%lx size 0x%lx to %s\n", gpa, size,
+		(flags & KVM_MEMORY_EXIT_FLAG_PRIVATE) ? "private" : "shared");
+	if (flags & KVM_MEMORY_EXIT_FLAG_PRIVATE)
+		vm_update_private_mem(vm, gpa, size, ALLOCATE_MEM);
+	else
+		vm_update_private_mem(vm, gpa, size, UNBACK_MEM);
+}
+
+static void vcpu_work(struct kvm_vm *vm, struct kvm_vcpu *vcpu,
+	struct vm_setup_info *info)
+{
+	volatile struct kvm_run *run;
+	struct ucall uc;
+	uint64_t cmd;
+
+	/*
+	 * Loop until the guest is done.
+	 */
+	run = vcpu->run;
+
+	while (true) {
+		vcpu_run(vcpu);
+
+		if (run->exit_reason == KVM_EXIT_IO) {
+			cmd = get_ucall(vcpu, &uc);
+			if (cmd != UCALL_SYNC)
+				break;
+
+			TEST_ASSERT(info->ioexit_cb, "ioexit cb not present");
+			info->ioexit_cb(vm, uc.args[1]);
+			continue;
+		}
+
+		if (run->exit_reason == KVM_EXIT_HYPERCALL) {
+			handle_vm_exit_map_gpa_hypercall(vm, run);
+			continue;
+		}
+
+		if (run->exit_reason == KVM_EXIT_MEMORY_FAULT) {
+			handle_vm_exit_memory_error(vm, run);
+			continue;
+		}
+
+		TEST_FAIL("Unhandled VCPU exit reason %d\n", run->exit_reason);
+		break;
+	}
+
+	if (run->exit_reason == KVM_EXIT_IO && cmd == UCALL_ABORT)
+		TEST_FAIL("%s at %s:%ld, val = %lu", (const char *)uc.args[0],
+			  __FILE__, uc.args[1], uc.args[2]);
+}
+
+/*
+ * Execute guest vm with private memory memslots.
+ *
+ * Input Args:
+ *   info - pointer to a structure containing information about setting up a VM
+ *     with private memslots
+ *
+ * Output Args: None
+ *
+ * Return: None
+ *
+ * Function called by host userspace logic in selftests to execute guest vm
+ * logic. It will install two memslots:
+ * 1) memslot 0 : containing all the boot code/stack pages
+ * 2) test_mem_slot : containing the region of memory that would be used to test
+ *   private/shared memory accesses to a memory backed by private memslots
+ */
+void execute_vm_with_private_mem(struct vm_setup_info *info)
+{
+	struct kvm_vm *vm;
+	struct kvm_enable_cap cap;
+	struct kvm_vcpu *vcpu;
+	uint32_t memslot0_pages = info->memslot0_pages;
+	uint64_t test_area_gpa, test_area_size;
+	struct test_setup_info *test_info = &info->test_info;
+
+	vm = vm_create_barebones();
+	vm_set_memory_encryption(vm, true, false, 0);
+	vm->use_ucall_pool = true;
+	vm_userspace_mem_region_add(vm, info->vm_mem_src, 0, 0,
+		memslot0_pages, KVM_MEM_PRIVATE);
+	kvm_vm_elf_load(vm, program_invocation_name);
+	vm_create_irqchip(vm);
+	TEST_ASSERT(info->guest_fn, "guest_fn not present");
+	vcpu = vm_vcpu_add(vm, 0, info->guest_fn);
+
+	vm_check_cap(vm, KVM_CAP_EXIT_HYPERCALL);
+	cap.cap = KVM_CAP_EXIT_HYPERCALL;
+	cap.flags = 0;
+	cap.args[0] = (1 << KVM_HC_MAP_GPA_RANGE);
+	vm_ioctl(vm, KVM_ENABLE_CAP, &cap);
+
+	TEST_ASSERT(test_info->test_area_size, "Test mem size not present");
+
+	test_area_size = test_info->test_area_size;
+	test_area_gpa = test_info->test_area_gpa;
+	vm_userspace_mem_region_add(vm, info->vm_mem_src, test_area_gpa,
+		test_info->test_area_slot, test_area_size / vm->page_size,
+		KVM_MEM_PRIVATE);
+	vm_update_private_mem(vm, test_area_gpa, test_area_size, ALLOCATE_MEM);
+
+	pr_info("Mapping test memory pages 0x%zx page_size 0x%x\n",
+		test_area_size/vm->page_size, vm->page_size);
+	virt_map(vm, test_area_gpa, test_area_gpa, test_area_size/vm->page_size);
+
+	ucall_init(vm, NULL);
+	vm_update_private_mem(vm, 0, (memslot0_pages << MIN_PAGE_SHIFT), ALLOCATE_MEM);
+
+	vcpu_work(vm, vcpu, info);
+
+	ucall_uninit(vm);
+	kvm_vm_free(vm);
+}
-- 
2.37.1.595.g718a3a8f04-goog

Re: [RFC V3 PATCH 5/6] selftests: kvm: x86: Execute VMs with private memory

[Sean Christopherson]

On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> +/*
> + * Execute KVM hypercall to change memory access type for a given gpa range.
> + *
> + * Input Args:
> + *   type - memory conversion type TO_SHARED/TO_PRIVATE
> + *   gpa - starting gpa address
> + *   size - size of the range starting from gpa for which memory access needs
> + *     to be changed
> + *
> + * Output Args: None
> + *
> + * Return: None
> + *
> + * Function called by guest logic in selftests to update the memory access type
> + * for a given gpa range. This API is useful in exercising implicit conversion
> + * path.
> + */
> +void guest_update_mem_access(enum mem_conversion_type type, uint64_t gpa,
> +	uint64_t size)

Provide wrappers to self-document what's going on, then the massive block comments
go away.  And the guts of this and guest_update_mem_map() are nearly identical.

Hmm, and we probably want to make it possible to do negative testing.

Then the one-off enums for TO_PRIVATE and whatnot go way too.

> +{
> +	int ret = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> MIN_PAGE_SHIFT,

Needs an assert that @size is page aligned.  And since these are x86-64 specific,
just use PAGE_SHIFT.  Huh, IS_ALIGNED() doesn't exist in selftests.  That should
be added, either by pulling in align.h or by adding the generic macros to
kvm_util_base.h.

And then x86-64's processor.h can defined IS_PAGE_ALIGNED().

E.g.

static inline void __kvm_hypercall_map_gpa_range(uint64_t gpa, uint64_t size,
						 uint64_t flags)
{
	return = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> PAGE_SHIFT, flags, 0);
}

static inline void kvm_hypercall_map_gpa_range(uint64_t gpa, uint64_t size,
					       uint64_t flags)
{
	int ret;

	GUEST_ASSERT_2(IS_PAGE_ALIGNED(gpa) && IS_PAGE_ALIGNED(size), gpa, size);

	ret = __kvm_hypercall_map_gpa_range(gpa, size, flags);
	GUEST_ASSERT_1(!ret, ret);
}

static inline kvm_hypercall_map_shared(uint64_t gpa, uint64_t size)
{
	kvm_hypercall_map_gpa_range(gpa, size, KVM_CLR_GPA_RANGE_ENC_ACCESS);
}

static inline kvm_hypercall_map_private(uint64_t gpa, uint64_t size)
{
	kvm_hypercall_map_gpa_range(gpa, size, KVM_MARK_GPA_RANGE_ENC_ACCESS);
}

> +static void handle_vm_exit_map_gpa_hypercall(struct kvm_vm *vm,
> +				volatile struct kvm_run *run)

Pass in @vcpu, not a vm+run.

> +{
> +	uint64_t gpa, npages, attrs, size;
> +
> +	TEST_ASSERT(run->hypercall.nr == KVM_HC_MAP_GPA_RANGE,
> +		"Unhandled Hypercall %lld\n", run->hypercall.nr);
> +	gpa = run->hypercall.args[0];
> +	npages = run->hypercall.args[1];
> +	size = npages << MIN_PAGE_SHIFT;
> +	attrs = run->hypercall.args[2];
> +	pr_info("Explicit conversion off 0x%lx size 0x%lx to %s\n", gpa, size,
> +		(attrs & KVM_MAP_GPA_RANGE_ENCRYPTED) ? "private" : "shared");
> +
> +	if (attrs & KVM_MAP_GPA_RANGE_ENCRYPTED)
> +		vm_update_private_mem(vm, gpa, size, ALLOCATE_MEM);
> +	else
> +		vm_update_private_mem(vm, gpa, size, UNBACK_MEM);
> +
> +	run->hypercall.ret = 0;
> +}
> +
> +static void handle_vm_exit_memory_error(struct kvm_vm *vm, volatile struct kvm_run *run)

Same  here, take a @vcpu.

Re: [RFC V3 PATCH 5/6] selftests: kvm: x86: Execute VMs with private memory

[Vishal Annapurve]

On Fri, Oct 7, 2022 at 1:47 AM Sean Christopherson <seanjc@google.com> wrote:
>
> On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > +/*
> > + * Execute KVM hypercall to change memory access type for a given gpa range.
> > + *
> > + * Input Args:
> > + *   type - memory conversion type TO_SHARED/TO_PRIVATE
> > + *   gpa - starting gpa address
> > + *   size - size of the range starting from gpa for which memory access needs
> > + *     to be changed
> > + *
> > + * Output Args: None
> > + *
> > + * Return: None
> > + *
> > + * Function called by guest logic in selftests to update the memory access type
> > + * for a given gpa range. This API is useful in exercising implicit conversion
> > + * path.
> > + */
> > +void guest_update_mem_access(enum mem_conversion_type type, uint64_t gpa,
> > +     uint64_t size)
>
> Provide wrappers to self-document what's going on, then the massive block comments
> go away.  And the guts of this and guest_update_mem_map() are nearly identical.
>
> Hmm, and we probably want to make it possible to do negative testing.
>
> Then the one-off enums for TO_PRIVATE and whatnot go way too.
>
> > +{
> > +     int ret = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> MIN_PAGE_SHIFT,
>
> Needs an assert that @size is page aligned.  And since these are x86-64 specific,
> just use PAGE_SHIFT.  Huh, IS_ALIGNED() doesn't exist in selftests.  That should
> be added, either by pulling in align.h or by adding the generic macros to
> kvm_util_base.h.
>
> And then x86-64's processor.h can defined IS_PAGE_ALIGNED().
>
> E.g.
>
> static inline void __kvm_hypercall_map_gpa_range(uint64_t gpa, uint64_t size,
>                                                  uint64_t flags)
> {
>         return = kvm_hypercall(KVM_HC_MAP_GPA_RANGE, gpa, size >> PAGE_SHIFT, flags, 0);
> }
>
> static inline void kvm_hypercall_map_gpa_range(uint64_t gpa, uint64_t size,
>                                                uint64_t flags)
> {
>         int ret;
>
>         GUEST_ASSERT_2(IS_PAGE_ALIGNED(gpa) && IS_PAGE_ALIGNED(size), gpa, size);
>
>         ret = __kvm_hypercall_map_gpa_range(gpa, size, flags);
>         GUEST_ASSERT_1(!ret, ret);
> }
>
> static inline kvm_hypercall_map_shared(uint64_t gpa, uint64_t size)
> {
>         kvm_hypercall_map_gpa_range(gpa, size, KVM_CLR_GPA_RANGE_ENC_ACCESS);
> }
>
> static inline kvm_hypercall_map_private(uint64_t gpa, uint64_t size)
> {
>         kvm_hypercall_map_gpa_range(gpa, size, KVM_MARK_GPA_RANGE_ENC_ACCESS);
> }
>
> > +static void handle_vm_exit_map_gpa_hypercall(struct kvm_vm *vm,
> > +                             volatile struct kvm_run *run)
>
> Pass in @vcpu, not a vm+run.
>
> > +{
> > +     uint64_t gpa, npages, attrs, size;
> > +
> > +     TEST_ASSERT(run->hypercall.nr == KVM_HC_MAP_GPA_RANGE,
> > +             "Unhandled Hypercall %lld\n", run->hypercall.nr);
> > +     gpa = run->hypercall.args[0];
> > +     npages = run->hypercall.args[1];
> > +     size = npages << MIN_PAGE_SHIFT;
> > +     attrs = run->hypercall.args[2];
> > +     pr_info("Explicit conversion off 0x%lx size 0x%lx to %s\n", gpa, size,
> > +             (attrs & KVM_MAP_GPA_RANGE_ENCRYPTED) ? "private" : "shared");
> > +
> > +     if (attrs & KVM_MAP_GPA_RANGE_ENCRYPTED)
> > +             vm_update_private_mem(vm, gpa, size, ALLOCATE_MEM);
> > +     else
> > +             vm_update_private_mem(vm, gpa, size, UNBACK_MEM);
> > +
> > +     run->hypercall.ret = 0;
> > +}
> > +
> > +static void handle_vm_exit_memory_error(struct kvm_vm *vm, volatile struct kvm_run *run)
>
> Same  here, take a @vcpu.

Ack. Will address these comments in the next series.

[RFC V3 PATCH 6/6] sefltests: kvm: x86: Add selftest for private memory

[Vishal Annapurve]

Add a selftest to exercise implicit/explicit conversion functionality
within KVM and verify:
1) Shared memory is visible to host userspace after conversion
2) Private memory is not visible to host userspace before/after conversion
3) Host userspace and guest can communicate over shared memory

Signed-off-by: Vishal Annapurve <vannapurve@google.com>
---
 tools/testing/selftests/kvm/.gitignore        |   1 +
 tools/testing/selftests/kvm/Makefile          |   1 +
 .../selftests/kvm/x86_64/private_mem_test.c   | 262 ++++++++++++++++++
 3 files changed, 264 insertions(+)
 create mode 100644 tools/testing/selftests/kvm/x86_64/private_mem_test.c

diff --git a/tools/testing/selftests/kvm/.gitignore b/tools/testing/selftests/kvm/.gitignore
index ec084a61a819..095b67dc632e 100644
--- a/tools/testing/selftests/kvm/.gitignore
+++ b/tools/testing/selftests/kvm/.gitignore
@@ -31,6 +31,7 @@
 /x86_64/monitor_mwait_test
 /x86_64/nx_huge_pages_test
 /x86_64/platform_info_test
+/x86_64/private_mem_test
 /x86_64/pmu_event_filter_test
 /x86_64/set_boot_cpu_id
 /x86_64/set_sregs_test
diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index 8fe72a60aef0..c5fc8ea2c843 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -94,6 +94,7 @@ TEST_GEN_PROGS_x86_64 += x86_64/mmu_role_test
 TEST_GEN_PROGS_x86_64 += x86_64/monitor_mwait_test
 TEST_GEN_PROGS_x86_64 += x86_64/platform_info_test
 TEST_GEN_PROGS_x86_64 += x86_64/pmu_event_filter_test
+TEST_GEN_PROGS_x86_64 += x86_64/private_mem_test
 TEST_GEN_PROGS_x86_64 += x86_64/set_boot_cpu_id
 TEST_GEN_PROGS_x86_64 += x86_64/set_sregs_test
 TEST_GEN_PROGS_x86_64 += x86_64/smm_test
diff --git a/tools/testing/selftests/kvm/x86_64/private_mem_test.c b/tools/testing/selftests/kvm/x86_64/private_mem_test.c
new file mode 100644
index 000000000000..9f491e2a16bb
--- /dev/null
+++ b/tools/testing/selftests/kvm/x86_64/private_mem_test.c
@@ -0,0 +1,262 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * tools/testing/selftests/kvm/lib/kvm_util.c
+ *
+ * Copyright (C) 2022, Google LLC.
+ */
+#define _GNU_SOURCE /* for program_invocation_short_name */
+#include <fcntl.h>
+#include <limits.h>
+#include <sched.h>
+#include <signal.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <string.h>
+#include <sys/ioctl.h>
+
+#include <linux/compiler.h>
+#include <linux/kernel.h>
+#include <linux/kvm_para.h>
+#include <linux/memfd.h>
+
+#include <test_util.h>
+#include <kvm_util.h>
+#include <private_mem.h>
+#include <processor.h>
+
+#define VM_MEMSLOT0_PAGES	(512 * 10)
+
+#define TEST_AREA_SLOT		10
+#define TEST_AREA_GPA		0xC0000000
+#define TEST_AREA_SIZE		(2 * 1024 * 1024)
+#define GUEST_TEST_MEM_OFFSET	(1 * 1024 * 1024)
+#define GUEST_TEST_MEM_SIZE	(10 * 4096)
+
+#define VM_STAGE_PROCESSED(x)	pr_info("Processed stage %s\n", #x)
+
+#define TEST_MEM_DATA_PAT1	0x66
+#define TEST_MEM_DATA_PAT2	0x99
+#define TEST_MEM_DATA_PAT3	0x33
+#define TEST_MEM_DATA_PAT4	0xaa
+#define TEST_MEM_DATA_PAT5	0x12
+
+static bool verify_mem_contents(void *mem, uint32_t size, uint8_t pat)
+{
+	uint8_t *buf = (uint8_t *)mem;
+
+	for (uint32_t i = 0; i < size; i++) {
+		if (buf[i] != pat)
+			return false;
+	}
+
+	return true;
+}
+
+/*
+ * Add custom implementation for memset to avoid using standard/builtin memset
+ * which may use features like SSE/GOT that don't work with guest vm execution
+ * within selftests.
+ */
+void *memset(void *mem, int byte, size_t size)
+{
+	uint8_t *buf = (uint8_t *)mem;
+
+	for (uint32_t i = 0; i < size; i++)
+		buf[i] = byte;
+
+	return buf;
+}
+
+static void populate_test_area(void *test_area_base, uint64_t pat)
+{
+	memset(test_area_base, pat, TEST_AREA_SIZE);
+}
+
+static void populate_guest_test_mem(void *guest_test_mem, uint64_t pat)
+{
+	memset(guest_test_mem, pat, GUEST_TEST_MEM_SIZE);
+}
+
+static bool verify_test_area(void *test_area_base, uint64_t area_pat,
+	uint64_t guest_pat)
+{
+	void *test_area1_base = test_area_base;
+	uint64_t test_area1_size = GUEST_TEST_MEM_OFFSET;
+	void *guest_test_mem = test_area_base + test_area1_size;
+	uint64_t guest_test_size = GUEST_TEST_MEM_SIZE;
+	void *test_area2_base = guest_test_mem + guest_test_size;
+	uint64_t test_area2_size = (TEST_AREA_SIZE - (GUEST_TEST_MEM_OFFSET +
+			GUEST_TEST_MEM_SIZE));
+
+	return (verify_mem_contents(test_area1_base, test_area1_size, area_pat) &&
+		verify_mem_contents(guest_test_mem, guest_test_size, guest_pat) &&
+		verify_mem_contents(test_area2_base, test_area2_size, area_pat));
+}
+
+#define GUEST_STARTED			0
+#define GUEST_PRIVATE_MEM_POPULATED	1
+#define GUEST_SHARED_MEM_POPULATED	2
+#define GUEST_PRIVATE_MEM_POPULATED2	3
+#define GUEST_IMPLICIT_MEM_CONV1	4
+#define GUEST_IMPLICIT_MEM_CONV2	5
+
+/*
+ * Run memory conversion tests supporting two types of conversion:
+ * 1) Explicit: Execute KVM hypercall to map/unmap gpa range which will cause
+ *   userspace exit to back/unback private memory. Subsequent accesses by guest
+ *   to the gpa range will not cause exit to userspace.
+ * 2) Implicit: Execute KVM hypercall to update memory access to a gpa range as
+ *   private/shared without exiting to userspace. Subsequent accesses by guest
+ *   to the gpa range will result in KVM EPT/NPT faults and then exit to
+ *   userspace for each page.
+ *
+ * Test memory conversion scenarios with following steps:
+ * 1) Access private memory using private access and verify that memory contents
+ *   are not visible to userspace.
+ * 2) Convert memory to shared using explicit/implicit conversions and ensure
+ *   that userspace is able to access the shared regions.
+ * 3) Convert memory back to private using explicit/implicit conversions and
+ *   ensure that userspace is again not able to access converted private
+ *   regions.
+ */
+static void guest_conv_test_fn(bool test_explicit_conv)
+{
+	void *test_area_base = (void *)TEST_AREA_GPA;
+	void *guest_test_mem = (void *)(TEST_AREA_GPA + GUEST_TEST_MEM_OFFSET);
+	uint64_t guest_test_size = GUEST_TEST_MEM_SIZE;
+
+	guest_map_ucall_page_shared();
+	GUEST_SYNC(GUEST_STARTED);
+
+	populate_test_area(test_area_base, TEST_MEM_DATA_PAT1);
+	GUEST_SYNC(GUEST_PRIVATE_MEM_POPULATED);
+	GUEST_ASSERT(verify_test_area(test_area_base, TEST_MEM_DATA_PAT1,
+		TEST_MEM_DATA_PAT1));
+
+	if (test_explicit_conv)
+		guest_update_mem_map(TO_SHARED, (uint64_t)guest_test_mem,
+			guest_test_size);
+	else {
+		guest_update_mem_access(TO_SHARED, (uint64_t)guest_test_mem,
+			guest_test_size);
+		GUEST_SYNC(GUEST_IMPLICIT_MEM_CONV1);
+	}
+
+	populate_guest_test_mem(guest_test_mem, TEST_MEM_DATA_PAT2);
+
+	GUEST_SYNC(GUEST_SHARED_MEM_POPULATED);
+	GUEST_ASSERT(verify_test_area(test_area_base, TEST_MEM_DATA_PAT1,
+		TEST_MEM_DATA_PAT5));
+
+	if (test_explicit_conv)
+		guest_update_mem_map(TO_PRIVATE, (uint64_t)guest_test_mem,
+			guest_test_size);
+	else {
+		guest_update_mem_access(TO_PRIVATE, (uint64_t)guest_test_mem,
+			guest_test_size);
+		GUEST_SYNC(GUEST_IMPLICIT_MEM_CONV2);
+	}
+
+	populate_guest_test_mem(guest_test_mem, TEST_MEM_DATA_PAT3);
+	GUEST_SYNC(GUEST_PRIVATE_MEM_POPULATED2);
+
+	GUEST_ASSERT(verify_test_area(test_area_base, TEST_MEM_DATA_PAT1,
+		TEST_MEM_DATA_PAT3));
+	GUEST_DONE();
+}
+
+static void conv_test_ioexit_fn(struct kvm_vm *vm, uint32_t uc_arg1)
+{
+	void *test_area_hva = addr_gpa2hva(vm, TEST_AREA_GPA);
+	void *guest_test_mem_hva = (test_area_hva + GUEST_TEST_MEM_OFFSET);
+	uint64_t guest_mem_gpa = (TEST_AREA_GPA + GUEST_TEST_MEM_OFFSET);
+	uint64_t guest_test_size = GUEST_TEST_MEM_SIZE;
+
+	switch (uc_arg1) {
+	case GUEST_STARTED:
+		populate_test_area(test_area_hva, TEST_MEM_DATA_PAT4);
+		VM_STAGE_PROCESSED(GUEST_STARTED);
+		break;
+	case GUEST_PRIVATE_MEM_POPULATED:
+		TEST_ASSERT(verify_test_area(test_area_hva, TEST_MEM_DATA_PAT4,
+				TEST_MEM_DATA_PAT4), "failed");
+		VM_STAGE_PROCESSED(GUEST_PRIVATE_MEM_POPULATED);
+		break;
+	case GUEST_SHARED_MEM_POPULATED:
+		TEST_ASSERT(verify_test_area(test_area_hva, TEST_MEM_DATA_PAT4,
+				TEST_MEM_DATA_PAT2), "failed");
+		populate_guest_test_mem(guest_test_mem_hva, TEST_MEM_DATA_PAT5);
+		VM_STAGE_PROCESSED(GUEST_SHARED_MEM_POPULATED);
+		break;
+	case GUEST_PRIVATE_MEM_POPULATED2:
+		TEST_ASSERT(verify_test_area(test_area_hva, TEST_MEM_DATA_PAT4,
+				TEST_MEM_DATA_PAT5), "failed");
+		VM_STAGE_PROCESSED(GUEST_PRIVATE_MEM_POPULATED2);
+		break;
+	case GUEST_IMPLICIT_MEM_CONV1:
+		/*
+		 * For first implicit conversion, memory is already private so
+		 * mark it private again just to zap the pte entries for the gpa
+		 * range, so that subsequent accesses from the guest will
+		 * generate ept/npt fault and memory conversion path will be
+		 * exercised by KVM.
+		 */
+		vm_update_private_mem(vm, guest_mem_gpa, guest_test_size,
+				ALLOCATE_MEM);
+		VM_STAGE_PROCESSED(GUEST_IMPLICIT_MEM_CONV1);
+		break;
+	case GUEST_IMPLICIT_MEM_CONV2:
+		/*
+		 * For second implicit conversion, memory is already shared so
+		 * mark it shared again just to zap the pte entries for the gpa
+		 * range, so that subsequent accesses from the guest will
+		 * generate ept/npt fault and memory conversion path will be
+		 * exercised by KVM.
+		 */
+		vm_update_private_mem(vm, guest_mem_gpa, guest_test_size,
+				UNBACK_MEM);
+		VM_STAGE_PROCESSED(GUEST_IMPLICIT_MEM_CONV2);
+		break;
+	default:
+		TEST_FAIL("Unknown stage %d\n", uc_arg1);
+		break;
+	}
+}
+
+static void guest_explicit_conv_test_fn(void)
+{
+	guest_conv_test_fn(true);
+}
+
+static void guest_implicit_conv_test_fn(void)
+{
+	guest_conv_test_fn(false);
+}
+
+static void execute_memory_conversion_test(void)
+{
+	struct vm_setup_info info;
+	struct test_setup_info *test_info = &info.test_info;
+
+	info.vm_mem_src = VM_MEM_SRC_ANONYMOUS;
+	info.memslot0_pages = VM_MEMSLOT0_PAGES;
+	test_info->test_area_gpa = TEST_AREA_GPA;
+	test_info->test_area_size = TEST_AREA_SIZE;
+	test_info->test_area_slot = TEST_AREA_SLOT;
+	info.ioexit_cb = conv_test_ioexit_fn;
+
+	info.guest_fn = guest_explicit_conv_test_fn;
+	execute_vm_with_private_mem(&info);
+
+	info.guest_fn = guest_implicit_conv_test_fn;
+	execute_vm_with_private_mem(&info);
+}
+
+int main(int argc, char *argv[])
+{
+	/* Tell stdout not to buffer its content */
+	setbuf(stdout, NULL);
+
+	execute_memory_conversion_test();
+	return 0;
+}
-- 
2.37.1.595.g718a3a8f04-goog

Re: [RFC V3 PATCH 6/6] sefltests: kvm: x86: Add selftest for private memory

[Sean Christopherson]

On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> +static bool verify_mem_contents(void *mem, uint32_t size, uint8_t pat)

As per feedback in v1[*], spell out "pattern".

[*] https://lore.kernel.org/all/YtiJx11AZHslcGnN@google.com

> +{
> +	uint8_t *buf = (uint8_t *)mem;
> +
> +	for (uint32_t i = 0; i < size; i++) {
> +		if (buf[i] != pat)
> +			return false;
> +	}
> +
> +	return true;
> +}
> +
> +/*
> + * Add custom implementation for memset to avoid using standard/builtin memset
> + * which may use features like SSE/GOT that don't work with guest vm execution
> + * within selftests.
> + */
> +void *memset(void *mem, int byte, size_t size)
> +{
> +	uint8_t *buf = (uint8_t *)mem;
> +
> +	for (uint32_t i = 0; i < size; i++)
> +		buf[i] = byte;
> +
> +	return buf;
> +}

memset(), memcpy(), and memcmp() are safe to use as of commit 6b6f71484bf4 ("KVM:
selftests: Implement memcmp(), memcpy(), and memset() for guest use").

Note the "fun" with gcc "optimizing" into infinite recursion... :-)

> +
> +static void populate_test_area(void *test_area_base, uint64_t pat)
> +{
> +	memset(test_area_base, pat, TEST_AREA_SIZE);
> +}
> +
> +static void populate_guest_test_mem(void *guest_test_mem, uint64_t pat)
> +{
> +	memset(guest_test_mem, pat, GUEST_TEST_MEM_SIZE);
> +}
> +
> +static bool verify_test_area(void *test_area_base, uint64_t area_pat,
> +	uint64_t guest_pat)

Again, avoid "pat".

> +{
> +	void *test_area1_base = test_area_base;
> +	uint64_t test_area1_size = GUEST_TEST_MEM_OFFSET;
> +	void *guest_test_mem = test_area_base + test_area1_size;
> +	uint64_t guest_test_size = GUEST_TEST_MEM_SIZE;
> +	void *test_area2_base = guest_test_mem + guest_test_size;
> +	uint64_t test_area2_size = (TEST_AREA_SIZE - (GUEST_TEST_MEM_OFFSET +
> +			GUEST_TEST_MEM_SIZE));

This is all amazingly hard to read.  AFAICT, the local variables are largely useless.
Actually, why even take in @test_area_base, isn't it hardcoded to TEST_AREA_GPA?
Then everything except the pattern can be hardcoded.

> +	return (verify_mem_contents(test_area1_base, test_area1_size, area_pat) &&
> +		verify_mem_contents(guest_test_mem, guest_test_size, guest_pat) &&
> +		verify_mem_contents(test_area2_base, test_area2_size, area_pat));
> +}

Re: [RFC V3 PATCH 6/6] sefltests: kvm: x86: Add selftest for private memory

[Vishal Annapurve]

On Fri, Oct 7, 2022 at 1:54 AM Sean Christopherson <seanjc@google.com> wrote:
>
> On Fri, Aug 19, 2022, Vishal Annapurve wrote:
> > +static bool verify_mem_contents(void *mem, uint32_t size, uint8_t pat)
>
> As per feedback in v1[*], spell out "pattern".
>
> [*] https://lore.kernel.org/all/YtiJx11AZHslcGnN@google.com
>
> > +{
> > +     uint8_t *buf = (uint8_t *)mem;
> > +
> > +     for (uint32_t i = 0; i < size; i++) {
> > +             if (buf[i] != pat)
> > +                     return false;
> > +     }
> > +
> > +     return true;
> > +}
> > +
> > +/*
> > + * Add custom implementation for memset to avoid using standard/builtin memset
> > + * which may use features like SSE/GOT that don't work with guest vm execution
> > + * within selftests.
> > + */
> > +void *memset(void *mem, int byte, size_t size)
> > +{
> > +     uint8_t *buf = (uint8_t *)mem;
> > +
> > +     for (uint32_t i = 0; i < size; i++)
> > +             buf[i] = byte;
> > +
> > +     return buf;
> > +}
>
> memset(), memcpy(), and memcmp() are safe to use as of commit 6b6f71484bf4 ("KVM:
> selftests: Implement memcmp(), memcpy(), and memset() for guest use").
>

This is much better. It made less sense to add a custom memset for a
single selftest.

> Note the "fun" with gcc "optimizing" into infinite recursion... :-)
>
> > +
> > +static void populate_test_area(void *test_area_base, uint64_t pat)
> > +{
> > +     memset(test_area_base, pat, TEST_AREA_SIZE);
> > +}
> > +
> > +static void populate_guest_test_mem(void *guest_test_mem, uint64_t pat)
> > +{
> > +     memset(guest_test_mem, pat, GUEST_TEST_MEM_SIZE);
> > +}
> > +
> > +static bool verify_test_area(void *test_area_base, uint64_t area_pat,
> > +     uint64_t guest_pat)
>
> Again, avoid "pat".
>
> > +{
> > +     void *test_area1_base = test_area_base;
> > +     uint64_t test_area1_size = GUEST_TEST_MEM_OFFSET;
> > +     void *guest_test_mem = test_area_base + test_area1_size;
> > +     uint64_t guest_test_size = GUEST_TEST_MEM_SIZE;
> > +     void *test_area2_base = guest_test_mem + guest_test_size;
> > +     uint64_t test_area2_size = (TEST_AREA_SIZE - (GUEST_TEST_MEM_OFFSET +
> > +                     GUEST_TEST_MEM_SIZE));
>
> This is all amazingly hard to read.  AFAICT, the local variables are largely useless.
> Actually, why even take in @test_area_base, isn't it hardcoded to TEST_AREA_GPA?
> Then everything except the pattern can be hardcoded.
>
> > +     return (verify_mem_contents(test_area1_base, test_area1_size, area_pat) &&
> > +             verify_mem_contents(guest_test_mem, guest_test_size, guest_pat) &&
> > +             verify_mem_contents(test_area2_base, test_area2_size, area_pat));
> > +}

Ack. Will address these comments in the next series.