kvm.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v4 0/4] KVM: selftests: add powerpc support
@ 2023-11-20 12:29 Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 1/4] KVM: selftests: Move pgd_created check into virt_pgd_alloc Nicholas Piggin
                   ` (3 more replies)
  0 siblings, 4 replies; 5+ messages in thread
From: Nicholas Piggin @ 2023-11-20 12:29 UTC (permalink / raw)
  To: kvm, Paolo Bonzini
  Cc: linuxppc-dev, Sean Christopherson, Aneesh Kumar K . V,
	Joel Stanley

This series adds initial KVM selftests support for powerpc
(64-bit, BookS, radix MMU).

Since v3:
https://lore.kernel.org/linuxppc-dev/20230608032425.59796-1-npiggin@gmail.com/
- Rebased to upstream (on top of Sean's ucall and guest assert
  rework).
- Drop powerpc specific tests for now, concentrate on base enablement.
- Fix a bunch of bugs and issues including the ones noticed by Joel
  in v3:
- Work around powerpc's max VCPU ID complexity to fix test case failure.
- Use TEST_REQUIRE for radix mode so hash hosts don't assert.
- Pack page table "fragments" in pages (like Linux kernel does), which
  fixes PTE memory consumption estimation and prevents the max memory
  test from failing with no memory for page tables.

Since v2:
- Added a couple of new tests (patch 5,6)
- Make default page size match host page size.
- Check for radix MMU capability.
- Build a few more of the generic tests.

Since v1:
- Update MAINTAINERS KVM PPC entry to include kvm selftests.
- Fixes and cleanups from Sean's review including new patch 1.
- Add 4K guest page support requiring new patch 2.

Thanks,
Nick

Nicholas Piggin (4):
  KVM: selftests: Move pgd_created check into virt_pgd_alloc
  KVM: selftests: Add aligned guest physical page allocator
  KVM: PPC: selftests: add support for powerpc
  KVM: PPC: selftests: powerpc enable kvm_create_max_vcpus test

 MAINTAINERS                                   |   2 +
 tools/testing/selftests/kvm/Makefile          |  20 +
 .../selftests/kvm/include/kvm_util_base.h     |  31 ++
 .../selftests/kvm/include/powerpc/hcall.h     |  17 +
 .../selftests/kvm/include/powerpc/ppc_asm.h   |  32 ++
 .../selftests/kvm/include/powerpc/processor.h |  39 ++
 .../selftests/kvm/include/powerpc/ucall.h     |  21 +
 .../selftests/kvm/kvm_create_max_vcpus.c      |   9 +
 .../selftests/kvm/lib/aarch64/processor.c     |   4 -
 tools/testing/selftests/kvm/lib/guest_modes.c |  27 +-
 tools/testing/selftests/kvm/lib/kvm_util.c    |  56 ++-
 .../selftests/kvm/lib/powerpc/handlers.S      |  93 ++++
 .../testing/selftests/kvm/lib/powerpc/hcall.c |  45 ++
 .../selftests/kvm/lib/powerpc/processor.c     | 468 ++++++++++++++++++
 .../testing/selftests/kvm/lib/powerpc/ucall.c |  21 +
 .../selftests/kvm/lib/riscv/processor.c       |   4 -
 .../selftests/kvm/lib/s390x/processor.c       |   4 -
 .../selftests/kvm/lib/x86_64/processor.c      |   7 +-
 tools/testing/selftests/kvm/powerpc/helpers.h |  46 ++
 19 files changed, 908 insertions(+), 38 deletions(-)
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/hcall.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/ppc_asm.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/processor.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/ucall.h
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/handlers.S
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/hcall.c
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/processor.c
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/ucall.c
 create mode 100644 tools/testing/selftests/kvm/powerpc/helpers.h

-- 
2.42.0


^ permalink raw reply	[flat|nested] 5+ messages in thread
* [PATCH v4 1/4] KVM: selftests: Move pgd_created check into virt_pgd_alloc
  2023-11-20 12:29 [PATCH v4 0/4] KVM: selftests: add powerpc support Nicholas Piggin
@ 2023-11-20 12:29 ` Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 2/4] KVM: selftests: Add aligned guest physical page allocator Nicholas Piggin
                   ` (2 subsequent siblings)
  3 siblings, 0 replies; 5+ messages in thread
From: Nicholas Piggin @ 2023-11-20 12:29 UTC (permalink / raw)
  To: kvm, Paolo Bonzini
  Cc: linuxppc-dev, Sean Christopherson, Aneesh Kumar K . V,
	Joel Stanley, Nicholas Piggin

virt_arch_pgd_alloc all do the same test and set of pgd_created. Move
this into common code.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 tools/testing/selftests/kvm/include/kvm_util_base.h | 5 +++++
 tools/testing/selftests/kvm/lib/aarch64/processor.c | 4 ----
 tools/testing/selftests/kvm/lib/riscv/processor.c   | 4 ----
 tools/testing/selftests/kvm/lib/s390x/processor.c   | 4 ----
 tools/testing/selftests/kvm/lib/x86_64/processor.c  | 7 ++-----
 5 files changed, 7 insertions(+), 17 deletions(-)

diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
index a18db6a7b3cf..e592a75ec052 100644
--- a/tools/testing/selftests/kvm/include/kvm_util_base.h
+++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
@@ -846,7 +846,12 @@ void virt_arch_pgd_alloc(struct kvm_vm *vm);
 
 static inline void virt_pgd_alloc(struct kvm_vm *vm)
 {
+	if (vm->pgd_created)
+		return;
+
 	virt_arch_pgd_alloc(vm);
+
+	vm->pgd_created = true;
 }
 
 /*
diff --git a/tools/testing/selftests/kvm/lib/aarch64/processor.c b/tools/testing/selftests/kvm/lib/aarch64/processor.c
index 6fe12e985ba5..69743d4c6748 100644
--- a/tools/testing/selftests/kvm/lib/aarch64/processor.c
+++ b/tools/testing/selftests/kvm/lib/aarch64/processor.c
@@ -96,13 +96,9 @@ void virt_arch_pgd_alloc(struct kvm_vm *vm)
 {
 	size_t nr_pages = page_align(vm, ptrs_per_pgd(vm) * 8) / vm->page_size;
 
-	if (vm->pgd_created)
-		return;
-
 	vm->pgd = vm_phy_pages_alloc(vm, nr_pages,
 				     KVM_GUEST_PAGE_TABLE_MIN_PADDR,
 				     vm->memslots[MEM_REGION_PT]);
-	vm->pgd_created = true;
 }
 
 static void _virt_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr,
diff --git a/tools/testing/selftests/kvm/lib/riscv/processor.c b/tools/testing/selftests/kvm/lib/riscv/processor.c
index d146ca71e0c0..7695ba2cd369 100644
--- a/tools/testing/selftests/kvm/lib/riscv/processor.c
+++ b/tools/testing/selftests/kvm/lib/riscv/processor.c
@@ -57,13 +57,9 @@ void virt_arch_pgd_alloc(struct kvm_vm *vm)
 {
 	size_t nr_pages = page_align(vm, ptrs_per_pte(vm) * 8) / vm->page_size;
 
-	if (vm->pgd_created)
-		return;
-
 	vm->pgd = vm_phy_pages_alloc(vm, nr_pages,
 				     KVM_GUEST_PAGE_TABLE_MIN_PADDR,
 				     vm->memslots[MEM_REGION_PT]);
-	vm->pgd_created = true;
 }
 
 void virt_arch_pg_map(struct kvm_vm *vm, uint64_t vaddr, uint64_t paddr)
diff --git a/tools/testing/selftests/kvm/lib/s390x/processor.c b/tools/testing/selftests/kvm/lib/s390x/processor.c
index 15945121daf1..358e03f09c7a 100644
--- a/tools/testing/selftests/kvm/lib/s390x/processor.c
+++ b/tools/testing/selftests/kvm/lib/s390x/processor.c
@@ -17,16 +17,12 @@ void virt_arch_pgd_alloc(struct kvm_vm *vm)
 	TEST_ASSERT(vm->page_size == 4096, "Unsupported page size: 0x%x",
 		    vm->page_size);
 
-	if (vm->pgd_created)
-		return;
-
 	paddr = vm_phy_pages_alloc(vm, PAGES_PER_REGION,
 				   KVM_GUEST_PAGE_TABLE_MIN_PADDR,
 				   vm->memslots[MEM_REGION_PT]);
 	memset(addr_gpa2hva(vm, paddr), 0xff, PAGES_PER_REGION * vm->page_size);
 
 	vm->pgd = paddr;
-	vm->pgd_created = true;
 }
 
 /*
diff --git a/tools/testing/selftests/kvm/lib/x86_64/processor.c b/tools/testing/selftests/kvm/lib/x86_64/processor.c
index d8288374078e..2ec64bb45db2 100644
--- a/tools/testing/selftests/kvm/lib/x86_64/processor.c
+++ b/tools/testing/selftests/kvm/lib/x86_64/processor.c
@@ -127,11 +127,8 @@ void virt_arch_pgd_alloc(struct kvm_vm *vm)
 	TEST_ASSERT(vm->mode == VM_MODE_PXXV48_4K, "Attempt to use "
 		"unknown or unsupported guest mode, mode: 0x%x", vm->mode);
 
-	/* If needed, create page map l4 table. */
-	if (!vm->pgd_created) {
-		vm->pgd = vm_alloc_page_table(vm);
-		vm->pgd_created = true;
-	}
+	/* Create page map l4 table. */
+	vm->pgd = vm_alloc_page_table(vm);
 }
 
 static void *virt_get_pte(struct kvm_vm *vm, uint64_t *parent_pte,
-- 
2.42.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread
* [PATCH v4 2/4] KVM: selftests: Add aligned guest physical page allocator
  2023-11-20 12:29 [PATCH v4 0/4] KVM: selftests: add powerpc support Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 1/4] KVM: selftests: Move pgd_created check into virt_pgd_alloc Nicholas Piggin
@ 2023-11-20 12:29 ` Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 3/4] KVM: PPC: selftests: add support for powerpc Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 4/4] KVM: PPC: selftests: powerpc enable kvm_create_max_vcpus test Nicholas Piggin
  3 siblings, 0 replies; 5+ messages in thread
From: Nicholas Piggin @ 2023-11-20 12:29 UTC (permalink / raw)
  To: kvm, Paolo Bonzini
  Cc: linuxppc-dev, Sean Christopherson, Aneesh Kumar K . V,
	Joel Stanley, Nicholas Piggin

powerpc will require this to allocate MMU tables in guest memory that
are larger than guest base page size.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 .../selftests/kvm/include/kvm_util_base.h     |  2 +
 tools/testing/selftests/kvm/lib/kvm_util.c    | 44 ++++++++++++-------
 2 files changed, 29 insertions(+), 17 deletions(-)

diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
index e592a75ec052..46edefabac85 100644
--- a/tools/testing/selftests/kvm/include/kvm_util_base.h
+++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
@@ -703,6 +703,8 @@ const char *exit_reason_str(unsigned int exit_reason);
 
 vm_paddr_t vm_phy_page_alloc(struct kvm_vm *vm, vm_paddr_t paddr_min,
 			     uint32_t memslot);
+vm_paddr_t vm_phy_pages_alloc_align(struct kvm_vm *vm, size_t num, size_t align,
+			      vm_paddr_t paddr_min, uint32_t memslot);
 vm_paddr_t vm_phy_pages_alloc(struct kvm_vm *vm, size_t num,
 			      vm_paddr_t paddr_min, uint32_t memslot);
 vm_paddr_t vm_alloc_page_table(struct kvm_vm *vm);
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index 7a8af1821f5d..22cf6c0ca89f 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -1924,6 +1924,7 @@ const char *exit_reason_str(unsigned int exit_reason)
  * Input Args:
  *   vm - Virtual Machine
  *   num - number of pages
+ *   align - pages alignment
  *   paddr_min - Physical address minimum
  *   memslot - Memory region to allocate page from
  *
@@ -1937,7 +1938,7 @@ const char *exit_reason_str(unsigned int exit_reason)
  * and their base address is returned. A TEST_ASSERT failure occurs if
  * not enough pages are available at or above paddr_min.
  */
-vm_paddr_t vm_phy_pages_alloc(struct kvm_vm *vm, size_t num,
+vm_paddr_t vm_phy_pages_alloc_align(struct kvm_vm *vm, size_t num, size_t align,
 			      vm_paddr_t paddr_min, uint32_t memslot)
 {
 	struct userspace_mem_region *region;
@@ -1951,24 +1952,27 @@ vm_paddr_t vm_phy_pages_alloc(struct kvm_vm *vm, size_t num,
 		paddr_min, vm->page_size);
 
 	region = memslot2region(vm, memslot);
-	base = pg = paddr_min >> vm->page_shift;
-
-	do {
-		for (; pg < base + num; ++pg) {
-			if (!sparsebit_is_set(region->unused_phy_pages, pg)) {
-				base = pg = sparsebit_next_set(region->unused_phy_pages, pg);
-				break;
+	base = paddr_min >> vm->page_shift;
+
+again:
+	base = (base + align - 1) & ~(align - 1);
+	for (pg = base; pg < base + num; ++pg) {
+		if (!sparsebit_is_set(region->unused_phy_pages, pg)) {
+			base = sparsebit_next_set(region->unused_phy_pages, pg);
+			if (!base) {
+				fprintf(stderr, "No guest physical pages "
+					"available, paddr_min: 0x%lx "
+					"page_size: 0x%x memslot: %u "
+					"num_pages: %lu align: %lu\n",
+					paddr_min, vm->page_size, memslot,
+					num, align);
+				fputs("---- vm dump ----\n", stderr);
+				vm_dump(stderr, vm, 2);
+				TEST_ASSERT(false, "false");
+				abort();
 			}
+			goto again;
 		}
-	} while (pg && pg != base + num);
-
-	if (pg == 0) {
-		fprintf(stderr, "No guest physical page available, "
-			"paddr_min: 0x%lx page_size: 0x%x memslot: %u\n",
-			paddr_min, vm->page_size, memslot);
-		fputs("---- vm dump ----\n", stderr);
-		vm_dump(stderr, vm, 2);
-		abort();
 	}
 
 	for (pg = base; pg < base + num; ++pg)
@@ -1977,6 +1981,12 @@ vm_paddr_t vm_phy_pages_alloc(struct kvm_vm *vm, size_t num,
 	return base * vm->page_size;
 }
 
+vm_paddr_t vm_phy_pages_alloc(struct kvm_vm *vm, size_t num,
+			      vm_paddr_t paddr_min, uint32_t memslot)
+{
+	return vm_phy_pages_alloc_align(vm, num, 1, paddr_min, memslot);
+}
+
 vm_paddr_t vm_phy_page_alloc(struct kvm_vm *vm, vm_paddr_t paddr_min,
 			     uint32_t memslot)
 {
-- 
2.42.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread
* [PATCH v4 3/4] KVM: PPC: selftests: add support for powerpc
  2023-11-20 12:29 [PATCH v4 0/4] KVM: selftests: add powerpc support Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 1/4] KVM: selftests: Move pgd_created check into virt_pgd_alloc Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 2/4] KVM: selftests: Add aligned guest physical page allocator Nicholas Piggin
@ 2023-11-20 12:29 ` Nicholas Piggin
  2023-11-20 12:29 ` [PATCH v4 4/4] KVM: PPC: selftests: powerpc enable kvm_create_max_vcpus test Nicholas Piggin
  3 siblings, 0 replies; 5+ messages in thread
From: Nicholas Piggin @ 2023-11-20 12:29 UTC (permalink / raw)
  To: kvm, Paolo Bonzini
  Cc: linuxppc-dev, Sean Christopherson, Aneesh Kumar K . V,
	Joel Stanley, Nicholas Piggin, Michael Ellerman

Implement KVM selftests support for powerpc (Book3S-64).

ucalls are implemented with an unsuppored PAPR hcall number which will
always cause KVM to exit to userspace.

Virtual memory is implemented for the radix MMU, and only a base page
size is supported (both 4K and 64K).

Guest interrupts are taken in real-mode, so require a page allocated at
gRA 0x0. Interrupt entry is complicated because gVA:gRA is not 1:1 mapped
(like the kernel is), so the MMU can not just just be switched on and
off.

Acked-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 MAINTAINERS                                   |   2 +
 tools/testing/selftests/kvm/Makefile          |  19 +
 .../selftests/kvm/include/kvm_util_base.h     |  24 +
 .../selftests/kvm/include/powerpc/hcall.h     |  17 +
 .../selftests/kvm/include/powerpc/ppc_asm.h   |  32 ++
 .../selftests/kvm/include/powerpc/processor.h |  39 ++
 .../selftests/kvm/include/powerpc/ucall.h     |  21 +
 tools/testing/selftests/kvm/lib/guest_modes.c |  27 +-
 tools/testing/selftests/kvm/lib/kvm_util.c    |  12 +
 .../selftests/kvm/lib/powerpc/handlers.S      |  93 ++++
 .../testing/selftests/kvm/lib/powerpc/hcall.c |  45 ++
 .../selftests/kvm/lib/powerpc/processor.c     | 468 ++++++++++++++++++
 .../testing/selftests/kvm/lib/powerpc/ucall.c |  21 +
 tools/testing/selftests/kvm/powerpc/helpers.h |  46 ++
 14 files changed, 862 insertions(+), 4 deletions(-)
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/hcall.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/ppc_asm.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/processor.h
 create mode 100644 tools/testing/selftests/kvm/include/powerpc/ucall.h
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/handlers.S
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/hcall.c
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/processor.c
 create mode 100644 tools/testing/selftests/kvm/lib/powerpc/ucall.c
 create mode 100644 tools/testing/selftests/kvm/powerpc/helpers.h

diff --git a/MAINTAINERS b/MAINTAINERS
index 97f51d5ec1cf..3f430d517462 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -11705,6 +11705,8 @@ F:	arch/powerpc/include/asm/kvm*
 F:	arch/powerpc/include/uapi/asm/kvm*
 F:	arch/powerpc/kernel/kvm*
 F:	arch/powerpc/kvm/
+F:	tools/testing/selftests/kvm/*/powerpc/
+F:	tools/testing/selftests/kvm/powerpc/
 
 KERNEL VIRTUAL MACHINE FOR RISC-V (KVM/riscv)
 M:	Anup Patel <anup@brainfault.org>
diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index a5963ab9215b..a198fe6136c8 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -65,6 +65,11 @@ LIBKVM_s390x += lib/s390x/ucall.c
 LIBKVM_riscv += lib/riscv/processor.c
 LIBKVM_riscv += lib/riscv/ucall.c
 
+LIBKVM_powerpc += lib/powerpc/handlers.S
+LIBKVM_powerpc += lib/powerpc/processor.c
+LIBKVM_powerpc += lib/powerpc/ucall.c
+LIBKVM_powerpc += lib/powerpc/hcall.c
+
 # Non-compiled test targets
 TEST_PROGS_x86_64 += x86_64/nx_huge_pages_test.sh
 
@@ -200,6 +205,20 @@ TEST_GEN_PROGS_riscv += kvm_page_table_test
 TEST_GEN_PROGS_riscv += set_memory_region_test
 TEST_GEN_PROGS_riscv += kvm_binary_stats_test
 
+TEST_GEN_PROGS_powerpc += access_tracking_perf_test
+TEST_GEN_PROGS_powerpc += demand_paging_test
+TEST_GEN_PROGS_powerpc += dirty_log_test
+TEST_GEN_PROGS_powerpc += dirty_log_perf_test
+TEST_GEN_PROGS_powerpc += guest_print_test
+TEST_GEN_PROGS_powerpc += hardware_disable_test
+TEST_GEN_PROGS_powerpc += kvm_page_table_test
+TEST_GEN_PROGS_powerpc += max_guest_memory_test
+TEST_GEN_PROGS_powerpc += memslot_modification_stress_test
+TEST_GEN_PROGS_powerpc += memslot_perf_test
+TEST_GEN_PROGS_powerpc += rseq_test
+TEST_GEN_PROGS_powerpc += set_memory_region_test
+TEST_GEN_PROGS_powerpc += kvm_binary_stats_test
+
 SPLIT_TESTS += get-reg-list
 
 TEST_PROGS += $(TEST_PROGS_$(ARCH_DIR))
diff --git a/tools/testing/selftests/kvm/include/kvm_util_base.h b/tools/testing/selftests/kvm/include/kvm_util_base.h
index 46edefabac85..3ce0bd0299fa 100644
--- a/tools/testing/selftests/kvm/include/kvm_util_base.h
+++ b/tools/testing/selftests/kvm/include/kvm_util_base.h
@@ -87,6 +87,12 @@ enum kvm_mem_region_type {
 	NR_MEM_REGIONS,
 };
 
+/* Page table fragment cache for guest page table < page size */
+struct vm_pt_frag_cache {
+	vm_paddr_t page;
+	size_t page_nr_used;
+};
+
 struct kvm_vm {
 	int mode;
 	unsigned long type;
@@ -106,6 +112,10 @@ struct kvm_vm {
 	bool pgd_created;
 	vm_paddr_t ucall_mmio_addr;
 	vm_paddr_t pgd;
+#if defined(__powerpc64__)
+	vm_paddr_t prtb; // process table
+	struct vm_pt_frag_cache pt_frag_cache[2]; // 256B and 4KB PT caches
+#endif
 	vm_vaddr_t gdt;
 	vm_vaddr_t tss;
 	vm_vaddr_t idt;
@@ -181,6 +191,8 @@ enum vm_guest_mode {
 	VM_MODE_PXXV48_4K,	/* For 48bits VA but ANY bits PA */
 	VM_MODE_P47V64_4K,
 	VM_MODE_P44V64_4K,
+	VM_MODE_P52V52_4K,
+	VM_MODE_P52V52_64K,
 	VM_MODE_P36V48_4K,
 	VM_MODE_P36V48_16K,
 	VM_MODE_P36V48_64K,
@@ -218,6 +230,18 @@ extern enum vm_guest_mode vm_mode_default;
 #define MIN_PAGE_SHIFT			12U
 #define ptes_per_page(page_size)	((page_size) / 8)
 
+#elif defined(__powerpc64__)
+
+extern enum vm_guest_mode vm_mode_default;
+
+#define VM_MODE_DEFAULT			vm_mode_default
+#define MIN_PAGE_SHIFT			12U
+#define ptes_per_page(page_size)	((page_size) / 8)
+
+#else
+
+#error "KVM selftests not implemented for architecture"
+
 #endif
 
 #define MIN_PAGE_SIZE		(1U << MIN_PAGE_SHIFT)
diff --git a/tools/testing/selftests/kvm/include/powerpc/hcall.h b/tools/testing/selftests/kvm/include/powerpc/hcall.h
new file mode 100644
index 000000000000..4028baa6c5d8
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/powerpc/hcall.h
@@ -0,0 +1,17 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * powerpc hcall defines
+ */
+#ifndef SELFTEST_KVM_HCALL_H
+#define SELFTEST_KVM_HCALL_H
+
+#include <linux/compiler.h>
+
+/* Ucalls use unimplemented PAPR hcall 0 which exits KVM */
+#define H_UCALL	0
+
+int64_t hcall0(uint64_t token);
+int64_t hcall1(uint64_t token, uint64_t arg1);
+int64_t hcall2(uint64_t token, uint64_t arg1, uint64_t arg2);
+
+#endif
diff --git a/tools/testing/selftests/kvm/include/powerpc/ppc_asm.h b/tools/testing/selftests/kvm/include/powerpc/ppc_asm.h
new file mode 100644
index 000000000000..b9df64659792
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/powerpc/ppc_asm.h
@@ -0,0 +1,32 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * powerpc asm specific defines
+ */
+#ifndef SELFTEST_KVM_PPC_ASM_H
+#define SELFTEST_KVM_PPC_ASM_H
+
+#define STACK_FRAME_MIN_SIZE	112 /* Could be 32 on ELFv2 */
+#define STACK_REDZONE_SIZE	512
+
+#define INT_FRAME_SIZE		(STACK_FRAME_MIN_SIZE + STACK_REDZONE_SIZE)
+
+#define SPR_SRR0	0x01a
+#define SPR_SRR1	0x01b
+#define SPR_CFAR	0x01c
+
+#define MSR_SF		0x8000000000000000ULL
+#define MSR_HV		0x1000000000000000ULL
+#define MSR_VEC		0x0000000002000000ULL
+#define MSR_VSX		0x0000000000800000ULL
+#define MSR_EE		0x0000000000008000ULL
+#define MSR_PR		0x0000000000004000ULL
+#define MSR_FP		0x0000000000002000ULL
+#define MSR_ME		0x0000000000001000ULL
+#define MSR_IR		0x0000000000000020ULL
+#define MSR_DR		0x0000000000000010ULL
+#define MSR_RI		0x0000000000000002ULL
+#define MSR_LE		0x0000000000000001ULL
+
+#define LPCR_ILE	0x0000000002000000ULL
+
+#endif
diff --git a/tools/testing/selftests/kvm/include/powerpc/processor.h b/tools/testing/selftests/kvm/include/powerpc/processor.h
new file mode 100644
index 000000000000..ce5a23525dbd
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/powerpc/processor.h
@@ -0,0 +1,39 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+/*
+ * powerpc processor specific defines
+ */
+#ifndef SELFTEST_KVM_PROCESSOR_H
+#define SELFTEST_KVM_PROCESSOR_H
+
+#include <linux/compiler.h>
+#include "ppc_asm.h"
+
+extern unsigned char __interrupts_start[];
+extern unsigned char __interrupts_end[];
+
+struct kvm_vm;
+struct kvm_vcpu;
+extern bool (*interrupt_handler)(struct kvm_vcpu *vcpu, unsigned trap);
+
+struct ex_regs {
+	uint64_t	gprs[32];
+	uint64_t	nia;
+	uint64_t	msr;
+	uint64_t	cfar;
+	uint64_t	lr;
+	uint64_t	ctr;
+	uint64_t	xer;
+	uint32_t	cr;
+	uint32_t	trap;
+	uint64_t	vaddr; /* vaddr of this struct */
+};
+
+void vm_install_exception_handler(struct kvm_vm *vm, int vector,
+			void (*handler)(struct ex_regs *));
+
+static inline void cpu_relax(void)
+{
+	asm volatile("" ::: "memory");
+}
+
+#endif
diff --git a/tools/testing/selftests/kvm/include/powerpc/ucall.h b/tools/testing/selftests/kvm/include/powerpc/ucall.h
new file mode 100644
index 000000000000..6b1a40997f41
--- /dev/null
+++ b/tools/testing/selftests/kvm/include/powerpc/ucall.h
@@ -0,0 +1,21 @@
+/* SPDX-License-Identifier: GPL-2.0-only */
+#ifndef SELFTEST_KVM_UCALL_H
+#define SELFTEST_KVM_UCALL_H
+
+#include "hcall.h"
+
+#define UCALL_EXIT_REASON	KVM_EXIT_PAPR_HCALL
+
+#define UCALL_R4_UCALL	0x5715 // regular ucall, r5 contains ucall pointer
+#define UCALL_R4_SIMPLE	0x0000 // simple exit usable by asm with no ucall data
+
+static inline void ucall_arch_init(struct kvm_vm *vm, vm_paddr_t mmio_gpa)
+{
+}
+
+static inline void ucall_arch_do_ucall(vm_vaddr_t uc)
+{
+	hcall2(H_UCALL, UCALL_R4_UCALL, (uintptr_t)(uc));
+}
+
+#endif
diff --git a/tools/testing/selftests/kvm/lib/guest_modes.c b/tools/testing/selftests/kvm/lib/guest_modes.c
index 1df3ce4b16fd..4dfaed1706d9 100644
--- a/tools/testing/selftests/kvm/lib/guest_modes.c
+++ b/tools/testing/selftests/kvm/lib/guest_modes.c
@@ -4,7 +4,11 @@
  */
 #include "guest_modes.h"
 
-#ifdef __aarch64__
+#if defined(__powerpc__)
+#include <unistd.h>
+#endif
+
+#if defined(__aarch64__) || defined(__powerpc__)
 #include "processor.h"
 enum vm_guest_mode vm_mode_default;
 #endif
@@ -13,9 +17,7 @@ struct guest_mode guest_modes[NUM_VM_MODES];
 
 void guest_modes_append_default(void)
 {
-#ifndef __aarch64__
-	guest_mode_append(VM_MODE_DEFAULT, true, true);
-#else
+#ifdef __aarch64__
 	{
 		unsigned int limit = kvm_check_cap(KVM_CAP_ARM_VM_IPA_SIZE);
 		bool ps4k, ps16k, ps64k;
@@ -70,6 +72,8 @@ void guest_modes_append_default(void)
 				    KVM_S390_VM_CPU_PROCESSOR, &info);
 		close(vm_fd);
 		close(kvm_fd);
+
+		guest_mode_append(VM_MODE_DEFAULT, true, true);
 		/* Starting with z13 we have 47bits of physical address */
 		if (info.ibc >= 0x30)
 			guest_mode_append(VM_MODE_P47V64_4K, true, true);
@@ -79,12 +83,27 @@ void guest_modes_append_default(void)
 	{
 		unsigned int sz = kvm_check_cap(KVM_CAP_VM_GPA_BITS);
 
+		guest_mode_append(VM_MODE_DEFAULT, true, true);
 		if (sz >= 52)
 			guest_mode_append(VM_MODE_P52V48_4K, true, true);
 		if (sz >= 48)
 			guest_mode_append(VM_MODE_P48V48_4K, true, true);
 	}
 #endif
+#ifdef __powerpc__
+	{
+		TEST_ASSERT(kvm_check_cap(KVM_CAP_PPC_MMU_RADIX),
+			    "Radix MMU not available, KVM selftests "
+			    "does not support Hash MMU!");
+		/* Radix guest EA and RA are 52-bit on POWER9 and POWER10 */
+		if (sysconf(_SC_PAGESIZE) == 4096)
+			vm_mode_default = VM_MODE_P52V52_4K;
+		else
+			vm_mode_default = VM_MODE_P52V52_64K;
+		guest_mode_append(VM_MODE_P52V52_4K, true, true);
+		guest_mode_append(VM_MODE_P52V52_64K, true, true);
+	}
+#endif
 }
 
 void for_each_guest_mode(void (*func)(enum vm_guest_mode, void *), void *arg)
diff --git a/tools/testing/selftests/kvm/lib/kvm_util.c b/tools/testing/selftests/kvm/lib/kvm_util.c
index 22cf6c0ca89f..db49abbc1918 100644
--- a/tools/testing/selftests/kvm/lib/kvm_util.c
+++ b/tools/testing/selftests/kvm/lib/kvm_util.c
@@ -158,6 +158,8 @@ const char *vm_guest_mode_string(uint32_t i)
 		[VM_MODE_PXXV48_4K]	= "PA-bits:ANY, VA-bits:48,  4K pages",
 		[VM_MODE_P47V64_4K]	= "PA-bits:47,  VA-bits:64,  4K pages",
 		[VM_MODE_P44V64_4K]	= "PA-bits:44,  VA-bits:64,  4K pages",
+		[VM_MODE_P52V52_4K]	= "PA-bits:52,  VA-bits:52,  4K pages",
+		[VM_MODE_P52V52_64K]	= "PA-bits:52,  VA-bits:52, 64K pages",
 		[VM_MODE_P36V48_4K]	= "PA-bits:36,  VA-bits:48,  4K pages",
 		[VM_MODE_P36V48_16K]	= "PA-bits:36,  VA-bits:48, 16K pages",
 		[VM_MODE_P36V48_64K]	= "PA-bits:36,  VA-bits:48, 64K pages",
@@ -183,6 +185,8 @@ const struct vm_guest_mode_params vm_guest_mode_params[] = {
 	[VM_MODE_PXXV48_4K]	= {  0,  0,  0x1000, 12 },
 	[VM_MODE_P47V64_4K]	= { 47, 64,  0x1000, 12 },
 	[VM_MODE_P44V64_4K]	= { 44, 64,  0x1000, 12 },
+	[VM_MODE_P52V52_4K]	= { 52, 52,  0x1000, 12 },
+	[VM_MODE_P52V52_64K]	= { 52, 52, 0x10000, 16 },
 	[VM_MODE_P36V48_4K]	= { 36, 48,  0x1000, 12 },
 	[VM_MODE_P36V48_16K]	= { 36, 48,  0x4000, 14 },
 	[VM_MODE_P36V48_64K]	= { 36, 48, 0x10000, 16 },
@@ -284,6 +288,14 @@ struct kvm_vm *____vm_create(enum vm_guest_mode mode)
 	case VM_MODE_P44V64_4K:
 		vm->pgtable_levels = 5;
 		break;
+#ifdef __powerpc__
+	case VM_MODE_P52V52_64K:
+		vm->pgtable_levels = 4;
+		break;
+	case VM_MODE_P52V52_4K:
+		vm->pgtable_levels = 4;
+		break;
+#endif
 	default:
 		TEST_FAIL("Unknown guest mode, mode: 0x%x", mode);
 	}
diff --git a/tools/testing/selftests/kvm/lib/powerpc/handlers.S b/tools/testing/selftests/kvm/lib/powerpc/handlers.S
new file mode 100644
index 000000000000..a68c187b835f
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/powerpc/handlers.S
@@ -0,0 +1,93 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+
+#include <ppc_asm.h>
+
+.macro INTERRUPT vec
+. = __interrupts_start + \vec
+	std	%r0,(0*8)(%r13)
+	std	%r3,(3*8)(%r13)
+	mfspr	%r0,SPR_CFAR
+	li	%r3,\vec
+	b	handle_interrupt
+.endm
+
+.balign 0x1000
+.global __interrupts_start
+__interrupts_start:
+INTERRUPT 0x100
+INTERRUPT 0x200
+INTERRUPT 0x300
+INTERRUPT 0x380
+INTERRUPT 0x400
+INTERRUPT 0x480
+INTERRUPT 0x500
+INTERRUPT 0x600
+INTERRUPT 0x700
+INTERRUPT 0x800
+INTERRUPT 0x900
+INTERRUPT 0xa00
+INTERRUPT 0xc00
+INTERRUPT 0xd00
+INTERRUPT 0xf00
+INTERRUPT 0xf20
+INTERRUPT 0xf40
+INTERRUPT 0xf60
+
+virt_handle_interrupt:
+	stdu	%r1,-INT_FRAME_SIZE(%r1)
+	mr	%r3,%r31
+	bl	route_interrupt
+	ld	%r4,(32*8)(%r31) /* NIA */
+	ld	%r5,(33*8)(%r31) /* MSR */
+	ld	%r6,(35*8)(%r31) /* LR */
+	ld	%r7,(36*8)(%r31) /* CTR */
+	ld	%r8,(37*8)(%r31) /* XER */
+	lwz	%r9,(38*8)(%r31) /* CR */
+	mtspr	SPR_SRR0,%r4
+	mtspr	SPR_SRR1,%r5
+	mtlr	%r6
+	mtctr	%r7
+	mtxer	%r8
+	mtcr	%r9
+reg=4
+	ld	%r0,(0*8)(%r31)
+	ld	%r3,(3*8)(%r31)
+.rept 28
+	ld	reg,(reg*8)(%r31)
+	reg=reg+1
+.endr
+	addi	%r1,%r1,INT_FRAME_SIZE
+	rfid
+
+virt_handle_interrupt_p:
+	.llong virt_handle_interrupt
+
+handle_interrupt:
+reg=4
+.rept 28
+	std	reg,(reg*8)(%r13)
+	reg=reg+1
+.endr
+	mfspr	%r4,SPR_SRR0
+	mfspr	%r5,SPR_SRR1
+	mflr	%r6
+	mfctr	%r7
+	mfxer	%r8
+	mfcr	%r9
+	std	%r4,(32*8)(%r13) /* NIA */
+	std	%r5,(33*8)(%r13) /* MSR */
+	std	%r0,(34*8)(%r13) /* CFAR */
+	std	%r6,(35*8)(%r13) /* LR */
+	std	%r7,(36*8)(%r13) /* CTR */
+	std	%r8,(37*8)(%r13) /* XER */
+	stw	%r9,(38*8)(%r13) /* CR */
+	stw	%r3,(38*8 + 4)(%r13) /* TRAP */
+
+	ld	%r31,(39*8)(%r13) /* vaddr */
+	ld	%r4,virt_handle_interrupt_p - __interrupts_start(0)
+	mtspr	SPR_SRR0,%r4
+	/* Reuse SRR1 */
+
+	rfid
+.global __interrupts_end
+__interrupts_end:
diff --git a/tools/testing/selftests/kvm/lib/powerpc/hcall.c b/tools/testing/selftests/kvm/lib/powerpc/hcall.c
new file mode 100644
index 000000000000..23a56aabad42
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/powerpc/hcall.c
@@ -0,0 +1,45 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * PAPR (pseries) hcall support.
+ */
+#include "kvm_util.h"
+#include "hcall.h"
+
+int64_t hcall0(uint64_t token)
+{
+	register uintptr_t r3 asm ("r3") = token;
+
+	asm volatile("sc 1" : "+r"(r3) :
+			    : "r0", "r4", "r5", "r6", "r7", "r8", "r9",
+			      "r10","r11", "r12", "ctr", "xer",
+			      "memory");
+
+	return r3;
+}
+
+int64_t hcall1(uint64_t token, uint64_t arg1)
+{
+	register uintptr_t r3 asm ("r3") = token;
+	register uintptr_t r4 asm ("r4") = arg1;
+
+	asm volatile("sc 1" : "+r"(r3), "+r"(r4) :
+			    : "r0", "r5", "r6", "r7", "r8", "r9",
+			      "r10","r11", "r12", "ctr", "xer",
+			      "memory");
+
+	return r3;
+}
+
+int64_t hcall2(uint64_t token, uint64_t arg1, uint64_t arg2)
+{
+	register uintptr_t r3 asm ("r3") = token;
+	register uintptr_t r4 asm ("r4") = arg1;
+	register uintptr_t r5 asm ("r5") = arg2;
+
+	asm volatile("sc 1" : "+r"(r3), "+r"(r4), "+r"(r5) :
+			    : "r0", "r6", "r7", "r8", "r9",
+			      "r10","r11", "r12", "ctr", "xer",
+			      "memory");
+
+	return r3;
+}
diff --git a/tools/testing/selftests/kvm/lib/powerpc/processor.c b/tools/testing/selftests/kvm/lib/powerpc/processor.c
new file mode 100644
index 000000000000..8f176cad3c7d
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/powerpc/processor.c
@@ -0,0 +1,468 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * KVM selftest powerpc library code - CPU-related functions (page tables...)
+ */
+
+#include <linux/sizes.h>
+
+#include "processor.h"
+#include "kvm_util.h"
+#include "kvm_util_base.h"
+#include "guest_modes.h"
+#include "hcall.h"
+
+#define RADIX_TREE_SIZE ((0x2UL << 61) | (0x5UL << 5)) // 52-bits
+#define RADIX_PGD_INDEX_SIZE 13
+
+static void set_proc_table(struct kvm_vm *vm, int pid, uint64_t dw0, uint64_t dw1)
+{
+	uint64_t *proc_table;
+
+	proc_table = addr_gpa2hva(vm, vm->prtb);
+	proc_table[pid * 2 + 0] = cpu_to_be64(dw0);
+	proc_table[pid * 2 + 1] = cpu_to_be64(dw1);
+}
+
+static void set_radix_proc_table(struct kvm_vm *vm, int pid, vm_paddr_t pgd)
+{
+	set_proc_table(vm, pid, pgd | RADIX_TREE_SIZE | RADIX_PGD_INDEX_SIZE, 0);
+}
+
+void virt_arch_pgd_alloc(struct kvm_vm *vm)
+{
+	struct kvm_ppc_mmuv3_cfg mmu_cfg;
+	vm_paddr_t prtb, pgtb;
+	size_t pgd_pages;
+
+	TEST_ASSERT((vm->mode == VM_MODE_P52V52_4K) ||
+		    (vm->mode == VM_MODE_P52V52_64K),
+		    "Unsupported guest mode, mode: 0x%x", vm->mode);
+
+	prtb = vm_phy_page_alloc(vm, KVM_GUEST_PAGE_TABLE_MIN_PADDR,
+				 vm->memslots[MEM_REGION_PT]);
+	vm->prtb = prtb;
+
+	pgd_pages = (1UL << (RADIX_PGD_INDEX_SIZE + 3)) >> vm->page_shift;
+	if (!pgd_pages)
+		pgd_pages = 1;
+	pgtb = vm_phy_pages_alloc_align(vm, pgd_pages, pgd_pages,
+					KVM_GUEST_PAGE_TABLE_MIN_PADDR,
+					vm->memslots[MEM_REGION_PT]);
+	vm->pgd = pgtb;
+
+	/* Set the base page directory in the proc table */
+	set_radix_proc_table(vm, 0, pgtb);
+
+	if (vm->mode == VM_MODE_P52V52_4K)
+		mmu_cfg.process_table = prtb | 0x8000000000000000UL | 0x0; // 4K size
+	else /* vm->mode == VM_MODE_P52V52_64K */
+		mmu_cfg.process_table = prtb | 0x8000000000000000UL | 0x4; // 64K size
+	mmu_cfg.flags = KVM_PPC_MMUV3_RADIX | KVM_PPC_MMUV3_GTSE;
+
+	vm_ioctl(vm, KVM_PPC_CONFIGURE_V3_MMU, &mmu_cfg);
+}
+
+static int pt_shift(struct kvm_vm *vm, int level)
+{
+	switch (level) {
+	case 1:
+		return 13;
+	case 2:
+	case 3:
+		return 9;
+	case 4:
+		if (vm->mode == VM_MODE_P52V52_4K)
+			return 9;
+		else /* vm->mode == VM_MODE_P52V52_64K */
+			return 5;
+	default:
+		TEST_ASSERT(false, "Invalid page table level %d\n", level);
+		return 0;
+	}
+}
+
+static uint64_t pt_entry_coverage(struct kvm_vm *vm, int level)
+{
+	uint64_t size = vm->page_size;
+
+	if (level == 4)
+		return size;
+	size <<= pt_shift(vm, 4);
+	if (level == 3)
+		return size;
+	size <<= pt_shift(vm, 3);
+	if (level == 2)
+		return size;
+	size <<= pt_shift(vm, 2);
+	return size;
+}
+
+static int pt_idx(struct kvm_vm *vm, uint64_t vaddr, int level, uint64_t *nls)
+{
+	switch (level) {
+	case 1:
+		*nls = 0x9;
+		return (vaddr >> 39) & 0x1fff;
+	case 2:
+		*nls = 0x9;
+		return (vaddr >> 30) & 0x1ff;
+	case 3:
+		if (vm->mode == VM_MODE_P52V52_4K)
+			*nls = 0x9;
+		else /* vm->mode == VM_MODE_P52V52_64K */
+			*nls = 0x5;
+		return (vaddr >> 21) & 0x1ff;
+	case 4:
+		if (vm->mode == VM_MODE_P52V52_4K)
+			return (vaddr >> 12) & 0x1ff;
+		else /* vm->mode == VM_MODE_P52V52_64K */
+			return (vaddr >> 16) & 0x1f;
+	default:
+		TEST_ASSERT(false, "Invalid page table level %d\n", level);
+		return 0;
+	}
+}
+
+static uint64_t *virt_get_pte(struct kvm_vm *vm, vm_paddr_t pt,
+			  uint64_t vaddr, int level, uint64_t *nls)
+{
+	int idx = pt_idx(vm, vaddr, level, nls);
+	uint64_t *ptep = addr_gpa2hva(vm, pt + idx*8);
+
+	return ptep;
+}
+
+#define PTE_VALID	0x8000000000000000ull
+#define PTE_LEAF	0x4000000000000000ull
+#define PTE_REFERENCED	0x0000000000000100ull
+#define PTE_CHANGED	0x0000000000000080ull
+#define PTE_PRIV	0x0000000000000008ull
+#define PTE_READ	0x0000000000000004ull
+#define PTE_RW		0x0000000000000002ull
+#define PTE_EXEC	0x0000000000000001ull
+#define PTE_PAGE_MASK	0x01fffffffffff000ull
+
+#define PDE_VALID	PTE_VALID
+#define PDE_NLS		0x0000000000000011ull
+#define PDE_PT_MASK	0x0fffffffffffff00ull
+
+static vm_paddr_t __vm_alloc_pt(struct kvm_vm *vm, uint64_t pt_shift)
+{
+	vm_paddr_t pt;
+
+	if (pt_shift >= vm->page_shift) {
+		size_t pt_pages = 1ULL << (pt_shift - vm->page_shift);
+		pt = vm_phy_pages_alloc_align(vm, pt_pages, pt_pages,
+					KVM_GUEST_PAGE_TABLE_MIN_PADDR,
+					vm->memslots[MEM_REGION_PT]);
+	} else {
+		struct vm_pt_frag_cache *pt_frag_cache;
+
+		if (pt_shift == 8) {
+			pt_frag_cache = &vm->pt_frag_cache[0];
+		} else if (pt_shift == 12) {
+			pt_frag_cache = &vm->pt_frag_cache[1];
+		} else {
+			TEST_ASSERT(0, "Invalid pt_shift:%lu\n", pt_shift);
+			return 0;
+		}
+
+		if (!pt_frag_cache->page) {
+			pt_frag_cache->page = vm_phy_pages_alloc_align(vm, 1, 1,
+						KVM_GUEST_PAGE_TABLE_MIN_PADDR,
+						vm->memslots[MEM_REGION_PT]);
+		}
+		pt = pt_frag_cache->page + pt_frag_cache->page_nr_used;
+		pt_frag_cache->page_nr_used += (1 << pt_shift);
+		if (pt_frag_cache->page_nr_used == vm->page_size) {
+			pt_frag_cache->page = 0;
+			pt_frag_cache->page_nr_used = 0;
+		}
+	}
+
+	return pt;
+}
+
+void virt_arch_pg_map(struct kvm_vm *vm, uint64_t gva, uint64_t gpa)
+{
+	vm_paddr_t pt = vm->pgd;
+	uint64_t *ptep, pte;
+	int level;
+
+	for (level = 1; level <= 3; level++) {
+		uint64_t nls;
+		uint64_t *pdep = virt_get_pte(vm, pt, gva, level, &nls);
+		uint64_t pde = be64_to_cpu(*pdep);
+
+		if (pde) {
+			TEST_ASSERT((pde & PDE_VALID) && !(pde & PTE_LEAF),
+				    "Invalid PDE at level: %u gva: 0x%lx pde:0x%lx\n",
+				    level, gva, pde);
+			pt = pde & PDE_PT_MASK;
+			continue;
+		}
+
+		pt = __vm_alloc_pt(vm, nls + 3);
+		pde = PDE_VALID | nls | pt;
+		*pdep = cpu_to_be64(pde);
+	}
+
+	ptep = virt_get_pte(vm, pt, gva, level, NULL);
+	pte = be64_to_cpu(*ptep);
+
+	TEST_ASSERT(!pte, "PTE already present at level: %u gva: 0x%lx pte:0x%lx\n",
+		    level, gva, pte);
+
+	pte = PTE_VALID | PTE_LEAF | PTE_REFERENCED | PTE_CHANGED |PTE_PRIV |
+	      PTE_READ | PTE_RW | PTE_EXEC | (gpa & PTE_PAGE_MASK);
+	*ptep = cpu_to_be64(pte);
+}
+
+vm_paddr_t addr_arch_gva2gpa(struct kvm_vm *vm, vm_vaddr_t gva)
+{
+	vm_paddr_t pt = vm->pgd;
+	uint64_t *ptep, pte;
+	int level;
+
+	for (level = 1; level <= 3; level++) {
+		uint64_t nls;
+		uint64_t *pdep = virt_get_pte(vm, pt, gva, level, &nls);
+		uint64_t pde = be64_to_cpu(*pdep);
+
+		TEST_ASSERT((pde & PDE_VALID) && !(pde & PTE_LEAF),
+			"PDE not present at level: %u gva: 0x%lx pde:0x%lx\n",
+			level, gva, pde);
+		pt = pde & PDE_PT_MASK;
+	}
+
+	ptep = virt_get_pte(vm, pt, gva, level, NULL);
+	pte = be64_to_cpu(*ptep);
+
+	TEST_ASSERT(pte,
+		"PTE not present at level: %u gva: 0x%lx pte:0x%lx\n",
+		level, gva, pte);
+
+	TEST_ASSERT((pte & PTE_VALID) && (pte & PTE_LEAF) &&
+		    (pte & PTE_READ) && (pte & PTE_RW) && (pte & PTE_EXEC),
+		    "PTE not valid at level: %u gva: 0x%lx pte:0x%lx\n",
+		    level, gva, pte);
+
+	return (pte & PTE_PAGE_MASK) + (gva & (vm->page_size - 1));
+}
+
+static void virt_dump_pt(FILE *stream, struct kvm_vm *vm, vm_paddr_t pt,
+			 vm_vaddr_t va, int level, uint8_t indent)
+{
+	int size, idx;
+
+	size = 1U << (pt_shift(vm, level) + 3);
+
+	for (idx = 0; idx < size; idx += 8, va += pt_entry_coverage(vm, level)) {
+		uint64_t *page_table = addr_gpa2hva(vm, pt + idx);
+		uint64_t pte = be64_to_cpu(*page_table);
+
+		if (!(pte & PTE_VALID))
+			continue;
+
+		if (pte & PTE_LEAF) {
+			fprintf(stream,
+				"%*s PTE[%d] gVA:0x%016lx -> gRA:0x%016llx\n",
+				indent, "", idx/8, va, pte & PTE_PAGE_MASK);
+		} else {
+			fprintf(stream, "%*sPDE%d[%d] gVA:0x%016lx\n",
+				indent, "", level, idx/8, va);
+			virt_dump_pt(stream, vm, pte & PDE_PT_MASK, va,
+				     level + 1, indent + 2);
+		}
+	}
+
+}
+
+void virt_arch_dump(FILE *stream, struct kvm_vm *vm, uint8_t indent)
+{
+	vm_paddr_t pt = vm->pgd;
+
+	if (!vm->pgd_created)
+		return;
+
+	virt_dump_pt(stream, vm, pt, 0, 1, indent);
+}
+
+static unsigned long get_r2(void)
+{
+	unsigned long r2;
+
+	asm("mr %0,%%r2" : "=r"(r2));
+
+	return r2;
+}
+
+struct kvm_vcpu *vm_arch_vcpu_add(struct kvm_vm *vm, uint32_t vcpu_id,
+				  void *guest_code)
+{
+	const size_t stack_size =  SZ_64K;
+	vm_vaddr_t stack_vaddr, ex_regs_vaddr;
+	vm_paddr_t ex_regs_paddr;
+	struct ex_regs *ex_regs;
+	struct kvm_regs regs;
+	struct kvm_vcpu *vcpu;
+	uint64_t lpcr;
+
+	stack_vaddr = __vm_vaddr_alloc(vm, stack_size,
+				       DEFAULT_GUEST_STACK_VADDR_MIN,
+				       MEM_REGION_DATA);
+
+	ex_regs_vaddr = __vm_vaddr_alloc(vm, stack_size,
+				       DEFAULT_GUEST_STACK_VADDR_MIN,
+				       MEM_REGION_DATA);
+	ex_regs_paddr = addr_gva2gpa(vm, ex_regs_vaddr);
+	ex_regs = addr_gpa2hva(vm, ex_regs_paddr);
+	ex_regs->vaddr = ex_regs_vaddr;
+
+	vcpu = __vm_vcpu_add(vm, vcpu_id);
+
+	vcpu_enable_cap(vcpu, KVM_CAP_PPC_PAPR, 1);
+
+	/* Setup guest registers */
+	vcpu_regs_get(vcpu, &regs);
+	vcpu_get_reg(vcpu, KVM_REG_PPC_LPCR_64, &lpcr);
+
+	regs.pc = (uintptr_t)guest_code;
+	regs.gpr[1] = stack_vaddr + stack_size - 256;
+	regs.gpr[2] = (uintptr_t)get_r2();
+	regs.gpr[12] = (uintptr_t)guest_code;
+	regs.gpr[13] = (uintptr_t)ex_regs_paddr;
+
+	regs.msr = MSR_SF | MSR_VEC | MSR_VSX | MSR_FP |
+		   MSR_ME | MSR_IR | MSR_DR | MSR_RI;
+
+	if (BYTE_ORDER == LITTLE_ENDIAN) {
+		regs.msr |= MSR_LE;
+		lpcr |= LPCR_ILE;
+	} else {
+		lpcr &= ~LPCR_ILE;
+	}
+
+	vcpu_regs_set(vcpu, &regs);
+	vcpu_set_reg(vcpu, KVM_REG_PPC_LPCR_64, lpcr);
+
+	return vcpu;
+}
+
+void vcpu_args_set(struct kvm_vcpu *vcpu, unsigned int num, ...)
+{
+	va_list ap;
+	struct kvm_regs regs;
+	int i;
+
+	TEST_ASSERT(num >= 1 && num <= 5, "Unsupported number of args: %u\n",
+		    num);
+
+	va_start(ap, num);
+	vcpu_regs_get(vcpu, &regs);
+
+	for (i = 0; i < num; i++)
+		regs.gpr[i + 3] = va_arg(ap, uint64_t);
+
+	vcpu_regs_set(vcpu, &regs);
+	va_end(ap);
+}
+
+void vcpu_arch_dump(FILE *stream, struct kvm_vcpu *vcpu, uint8_t indent)
+{
+	struct kvm_regs regs;
+
+	vcpu_regs_get(vcpu, &regs);
+
+	fprintf(stream, "%*sNIA: 0x%016llx  MSR: 0x%016llx\n",
+			indent, "", regs.pc, regs.msr);
+	fprintf(stream, "%*sLR:  0x%016llx  CTR :0x%016llx\n",
+			indent, "", regs.lr, regs.ctr);
+	fprintf(stream, "%*sCR:  0x%08llx          XER :0x%016llx\n",
+			indent, "", regs.cr, regs.xer);
+}
+
+void kvm_arch_vm_post_create(struct kvm_vm *vm)
+{
+	vm_paddr_t excp_paddr;
+	void *mem;
+
+	excp_paddr = vm_phy_page_alloc(vm, 0, vm->memslots[MEM_REGION_DATA]);
+
+	TEST_ASSERT(excp_paddr == 0,
+		    "Interrupt vectors not allocated at gPA address 0: (0x%lx)",
+		    excp_paddr);
+
+	mem = addr_gpa2hva(vm, excp_paddr);
+	memcpy(mem, __interrupts_start, __interrupts_end - __interrupts_start);
+}
+
+void assert_on_unhandled_exception(struct kvm_vcpu *vcpu)
+{
+	struct ucall uc;
+
+	if (get_ucall(vcpu, &uc) == UCALL_UNHANDLED) {
+		vm_paddr_t ex_regs_paddr;
+		struct ex_regs *ex_regs;
+		struct kvm_regs regs;
+
+		vcpu_regs_get(vcpu, &regs);
+		ex_regs_paddr = (vm_paddr_t)regs.gpr[13];
+		ex_regs = addr_gpa2hva(vcpu->vm, ex_regs_paddr);
+
+		TEST_FAIL("Unexpected interrupt in guest NIA:0x%016lx MSR:0x%016lx TRAP:0x%04x",
+			  ex_regs->nia, ex_regs->msr, ex_regs->trap);
+	}
+}
+
+struct handler {
+	void (*fn)(struct ex_regs *regs);
+	int trap;
+};
+
+#define NR_HANDLERS	10
+static struct handler handlers[NR_HANDLERS];
+
+void route_interrupt(struct ex_regs *regs)
+{
+	int i;
+
+	for (i = 0; i < NR_HANDLERS; i++) {
+		if (handlers[i].trap == regs->trap) {
+			handlers[i].fn(regs);
+			return;
+		}
+	}
+
+	ucall(UCALL_UNHANDLED, 0);
+}
+
+void vm_install_exception_handler(struct kvm_vm *vm, int trap,
+			       void (*fn)(struct ex_regs *))
+{
+	int i;
+
+	for (i = 0; i < NR_HANDLERS; i++) {
+		if (!handlers[i].trap || handlers[i].trap == trap) {
+			if (fn == NULL)
+				trap = 0; /* Clear handler */
+			handlers[i].trap = trap;
+			handlers[i].fn = fn;
+			sync_global_to_guest(vm, handlers[i]);
+			return;
+		}
+	}
+
+	TEST_FAIL("Out of exception handlers");
+}
+
+void kvm_selftest_arch_init(void)
+{
+	TEST_REQUIRE(kvm_has_cap(KVM_CAP_PPC_MMU_RADIX));
+
+	/*
+	 * powerpc default mode is set by host page size and not static,
+	 * so start by computing that early.
+	 */
+	guest_modes_append_default();
+}
diff --git a/tools/testing/selftests/kvm/lib/powerpc/ucall.c b/tools/testing/selftests/kvm/lib/powerpc/ucall.c
new file mode 100644
index 000000000000..3cb598a8b743
--- /dev/null
+++ b/tools/testing/selftests/kvm/lib/powerpc/ucall.c
@@ -0,0 +1,21 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * ucall support. A ucall is a "hypercall to host userspace".
+ */
+#include "kvm_util.h"
+#include "hcall.h"
+
+void *ucall_arch_get_ucall(struct kvm_vcpu *vcpu)
+{
+	struct kvm_run *run = vcpu->run;
+
+	if (run->exit_reason == UCALL_EXIT_REASON &&
+	    run->papr_hcall.nr == H_UCALL) {
+		struct kvm_regs regs;
+
+		vcpu_regs_get(vcpu, &regs);
+		if (regs.gpr[4] == UCALL_R4_UCALL)
+			return (void *)regs.gpr[5];
+	}
+	return NULL;
+}
diff --git a/tools/testing/selftests/kvm/powerpc/helpers.h b/tools/testing/selftests/kvm/powerpc/helpers.h
new file mode 100644
index 000000000000..8f60bb826830
--- /dev/null
+++ b/tools/testing/selftests/kvm/powerpc/helpers.h
@@ -0,0 +1,46 @@
+// SPDX-License-Identifier: GPL-2.0-only
+
+#ifndef SELFTEST_KVM_HELPERS_H
+#define SELFTEST_KVM_HELPERS_H
+
+#include "kvm_util.h"
+#include "processor.h"
+
+static inline void __handle_ucall(struct kvm_vcpu *vcpu, uint64_t expect, struct ucall *uc)
+{
+	uint64_t ret;
+	struct kvm_regs regs;
+
+	ret = get_ucall(vcpu, uc);
+	if (ret == expect)
+		return;
+
+	vcpu_regs_get(vcpu, &regs);
+	fprintf(stderr, "Guest failure at NIA:0x%016llx MSR:0x%016llx\n", regs.pc, regs.msr);
+	fprintf(stderr, "Expected ucall: %lu\n", expect);
+
+	if (ret == UCALL_ABORT)
+		REPORT_GUEST_ASSERT(*uc);
+	else
+		TEST_FAIL("Unexpected ucall: %lu exit_reason=%s",
+			ret, exit_reason_str(vcpu->run->exit_reason));
+}
+
+static inline void handle_ucall(struct kvm_vcpu *vcpu, uint64_t expect)
+{
+	struct ucall uc;
+
+	__handle_ucall(vcpu, expect, &uc);
+}
+
+static inline void host_sync(struct kvm_vcpu *vcpu, uint64_t sync)
+{
+	struct ucall uc;
+
+	__handle_ucall(vcpu, UCALL_SYNC, &uc);
+
+	TEST_ASSERT(uc.args[1] == (sync), "Sync failed host:%ld guest:%ld",
+						(long)sync, (long)uc.args[1]);
+}
+
+#endif
-- 
2.42.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread
* [PATCH v4 4/4] KVM: PPC: selftests: powerpc enable kvm_create_max_vcpus test
  2023-11-20 12:29 [PATCH v4 0/4] KVM: selftests: add powerpc support Nicholas Piggin
                   ` (2 preceding siblings ...)
  2023-11-20 12:29 ` [PATCH v4 3/4] KVM: PPC: selftests: add support for powerpc Nicholas Piggin
@ 2023-11-20 12:29 ` Nicholas Piggin
  3 siblings, 0 replies; 5+ messages in thread
From: Nicholas Piggin @ 2023-11-20 12:29 UTC (permalink / raw)
  To: kvm, Paolo Bonzini
  Cc: linuxppc-dev, Sean Christopherson, Aneesh Kumar K . V,
	Joel Stanley, Nicholas Piggin

powerpc's maximum permitted vCPU ID depends on the VM's SMT mode, and
the maximum reported by KVM_CAP_MAX_VCPU_ID exceeds a simple non-SMT
VM's limit.

The powerpc KVM selftest port uses non-SMT VMs, so add a workaround
to the kvm_create_max_vcpus test case to limit vCPU IDs to
KVM_CAP_MAX_VCPUS on powerpc. And enable the test case.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 tools/testing/selftests/kvm/Makefile               | 1 +
 tools/testing/selftests/kvm/kvm_create_max_vcpus.c | 9 +++++++++
 2 files changed, 10 insertions(+)

diff --git a/tools/testing/selftests/kvm/Makefile b/tools/testing/selftests/kvm/Makefile
index a198fe6136c8..1e904d8871d7 100644
--- a/tools/testing/selftests/kvm/Makefile
+++ b/tools/testing/selftests/kvm/Makefile
@@ -211,6 +211,7 @@ TEST_GEN_PROGS_powerpc += dirty_log_test
 TEST_GEN_PROGS_powerpc += dirty_log_perf_test
 TEST_GEN_PROGS_powerpc += guest_print_test
 TEST_GEN_PROGS_powerpc += hardware_disable_test
+TEST_GEN_PROGS_powerpc += kvm_create_max_vcpus
 TEST_GEN_PROGS_powerpc += kvm_page_table_test
 TEST_GEN_PROGS_powerpc += max_guest_memory_test
 TEST_GEN_PROGS_powerpc += memslot_modification_stress_test
diff --git a/tools/testing/selftests/kvm/kvm_create_max_vcpus.c b/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
index 31b3cb24b9a7..330ede73c147 100644
--- a/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
+++ b/tools/testing/selftests/kvm/kvm_create_max_vcpus.c
@@ -51,6 +51,15 @@ int main(int argc, char *argv[])
 	pr_info("KVM_CAP_MAX_VCPU_ID: %d\n", kvm_max_vcpu_id);
 	pr_info("KVM_CAP_MAX_VCPUS: %d\n", kvm_max_vcpus);
 
+#ifdef __powerpc64__
+	/*
+	 * powerpc has a particular format for the vcpu ID that depends on
+	 * the guest SMT mode, and the max ID cap is too large for non-SMT
+	 * modes, where the maximum ID is the same as the maximum vCPUs.
+	 */
+	kvm_max_vcpu_id = kvm_max_vcpus;
+#endif
+
 	/*
 	 * Check that we're allowed to open nr_fds_wanted file descriptors and
 	 * try raising the limits if needed.
-- 
2.42.0


^ permalink raw reply related	[flat|nested] 5+ messages in thread
end of thread, other threads:[~2023-11-20 12:29 UTC | newest]

Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2023-11-20 12:29 [PATCH v4 0/4] KVM: selftests: add powerpc support Nicholas Piggin
2023-11-20 12:29 ` [PATCH v4 1/4] KVM: selftests: Move pgd_created check into virt_pgd_alloc Nicholas Piggin
2023-11-20 12:29 ` [PATCH v4 2/4] KVM: selftests: Add aligned guest physical page allocator Nicholas Piggin
2023-11-20 12:29 ` [PATCH v4 3/4] KVM: PPC: selftests: add support for powerpc Nicholas Piggin
2023-11-20 12:29 ` [PATCH v4 4/4] KVM: PPC: selftests: powerpc enable kvm_create_max_vcpus test Nicholas Piggin

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).