Add a test to xe_vm that determines if pagefaults are correctly tracked and reported by the DRM_IOCTL_XE_VM_GET_PROPERTY.
v2: - s/pageproperty/pagefaults Signed-off-by: Jonathan Cavitt <jonathan.cav...@intel.com> Suggested-by: Jianxun Zhang <jianxun.zh...@intel.com> Suggested-by: Stuart Summers <stuart.summ...@intel.com> --- tests/intel/xe_vm.c | 156 ++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 156 insertions(+) diff --git a/tests/intel/xe_vm.c b/tests/intel/xe_vm.c index 5c189f7af1..3269125686 100644 --- a/tests/intel/xe_vm.c +++ b/tests/intel/xe_vm.c @@ -2382,6 +2382,10 @@ static void invalid_vm_id(int fd) * SUBTEST: vm-get-property-invalid-property * Functionality: ioctl_input_validation * Description: Check query with invalid property returns expected error code + * + * SUBTEST: vm-get-property-exercise + * Functionality: drm_xe_vm_get_property + * Description: Check query correctly reports pagefaults on vm */ static void get_property_invalid_reserved(int fd, uint32_t vm) { @@ -2422,6 +2426,157 @@ static void get_property_invalid_property(int fd, uint32_t vm) do_ioctl_err(fd, DRM_IOCTL_XE_VM_GET_PROPERTY, &query, EINVAL); } +static void +gen_pf(int fd, uint32_t vm, struct drm_xe_engine_class_instance *eci) +{ + int n_exec_queues = 2; + int n_execs = 2; + uint64_t addr = 0x1a0000; + struct drm_xe_sync sync[2] = { + { .type = DRM_XE_SYNC_TYPE_SYNCOBJ, .flags = DRM_XE_SYNC_FLAG_SIGNAL, }, + { .type = DRM_XE_SYNC_TYPE_SYNCOBJ, .flags = DRM_XE_SYNC_FLAG_SIGNAL, }, + }; + struct drm_xe_exec exec = { + .num_batch_buffer = 1, + .num_syncs = 2, + .syncs = to_user_pointer(sync), + }; + uint32_t exec_queues[2]; + uint32_t syncobjs[2]; + size_t bo_size; + uint32_t bo = 0; + struct { + struct xe_spin spin; + uint32_t batch[16]; + uint64_t pad; + uint32_t data; + } *data; + struct xe_spin_opts spin_opts = { .preempt = false }; + int i, b; + + bo_size = sizeof(*data) * n_execs; + bo_size = xe_bb_size(fd, bo_size); + + bo = xe_bo_create(fd, vm, bo_size, + vram_if_possible(fd, eci->gt_id), + DRM_XE_GEM_CREATE_FLAG_NEEDS_VISIBLE_VRAM); + data = xe_bo_map(fd, bo, bo_size); + + for (i = 0; i < n_exec_queues; i++) { + exec_queues[i] = xe_exec_queue_create(fd, vm, eci, 0); + syncobjs[i] = syncobj_create(fd, 0); + }; + + sync[0].handle = syncobj_create(fd, 0); + xe_vm_bind_async(fd, vm, 0, bo, 0, addr, bo_size, sync, 1); + + for (i = 0; i < n_execs; i++) { + uint64_t base_addr = !i ? addr + bo_size * 128 : addr; + uint64_t batch_offset = (char *)&data[i].batch - (char *)data; + uint64_t batch_addr = base_addr + batch_offset; + uint64_t spin_offset = (char *)&data[i].spin - (char *)data; + uint64_t sdi_offset = (char *)&data[i].data - (char *)data; + uint64_t sdi_addr = base_addr + sdi_offset; + uint64_t exec_addr; + int e = i % n_exec_queues; + + if (!i) { + spin_opts.addr = base_addr + spin_offset; + xe_spin_init(&data[i].spin, &spin_opts); + exec_addr = spin_opts.addr; + } else { + b = 0; + data[i].batch[b++] = MI_STORE_DWORD_IMM_GEN4; + data[i].batch[b++] = sdi_addr; + data[i].batch[b++] = sdi_addr >> 32; + data[i].batch[b++] = 0xc0ffee; + data[i].batch[b++] = MI_BATCH_BUFFER_END; + igt_assert(b <= ARRAY_SIZE(data[i].batch)); + + exec_addr = batch_addr; + } + + sync[0].flags &= ~DRM_XE_SYNC_FLAG_SIGNAL; + sync[1].flags |= DRM_XE_SYNC_FLAG_SIGNAL; + sync[1].handle = syncobjs[e]; + + exec.exec_queue_id = exec_queues[e]; + exec.address = exec_addr; + if (e != i) + syncobj_reset(fd, &syncobjs[e], 1); + xe_exec(fd, &exec); + } + + for (i = 0; i < n_exec_queues && n_execs; i++) + igt_assert(syncobj_wait(fd, &syncobjs[i], 1, INT64_MAX, 0, + NULL)); + igt_assert(syncobj_wait(fd, &sync[0].handle, 1, INT64_MAX, 0, NULL)); + + sync[0].flags |= DRM_XE_SYNC_FLAG_SIGNAL; + xe_vm_unbind_async(fd, vm, 0, 0, addr, bo_size, sync, 1); + igt_assert(syncobj_wait(fd, &sync[0].handle, 1, INT64_MAX, 0, NULL)); + + syncobj_destroy(fd, sync[0].handle); + for (i = 0; i < n_exec_queues; i++) { + syncobj_destroy(fd, syncobjs[i]); + xe_exec_queue_destroy(fd, exec_queues[i]); + } + + munmap(data, bo_size); + gem_close(fd, bo); +} + +static void print_pf(struct xe_vm_fault *fault) +{ + igt_debug("FAULT:\n"); + igt_debug("address = 0x%08x%08x\n", + upper_32_bits(fault->address), + lower_32_bits(fault->address)); + igt_debug("address precision = %u\n", fault->address_precision); + igt_debug("access type = %u\n", fault->access_type); + igt_debug("fault type = %u\n", fault->fault_type); + igt_debug("fault level = %u\n", fault->fault_level); + igt_debug("\n"); +} + +static void get_property_exercise(int fd, uint32_t vm) +{ + struct drm_xe_engine_class_instance *hwe; + struct xe_vm_fault *faults, f0, f; + struct drm_xe_vm_get_property query = { + .vm_id = vm, + .property = DRM_XE_VM_GET_PROPERTY_FAULTS + }; + int i, fault_count; + + igt_assert_eq(igt_ioctl(fd, DRM_IOCTL_XE_VM_GET_PROPERTY, &query), 0); + + igt_assert_eq(query.size, 0); + + xe_for_each_engine(fd, hwe) + gen_pf(fd, vm, hwe); + + igt_assert_eq(igt_ioctl(fd, DRM_IOCTL_XE_VM_GET_PROPERTY, &query), 0); + igt_assert_lt(0, query.size); + + faults = malloc(query.size); + igt_assert(faults); + + query.data = to_user_pointer(faults); + igt_assert_eq(igt_ioctl(fd, DRM_IOCTL_XE_VM_GET_PROPERTY, &query), 0); + + fault_count = query.size / sizeof(struct xe_vm_fault); + f0 = faults[0]; + for (i = 0; i < fault_count; i++) { + f = faults[i]; + print_pf(&f); + igt_assert_eq(f.address, f0.address); + igt_assert_eq(f.access_type, f0.access_type); + igt_assert_eq(f.fault_type, f0.fault_type); + } + free(faults); +} + static void test_get_property(int fd, void (*func)(int fd, uint32_t vm)) { uint32_t vm; @@ -2551,6 +2706,7 @@ igt_main { "invalid-vm-id", get_property_invalid_vm_id }, { "invalid-size", get_property_invalid_size }, { "invalid-property", get_property_invalid_property }, + { "exercise", get_property_exercise }, { } }; -- 2.43.0