From: Matthew Auld <matthew.auld@intel.com>
To: igt-dev@lists.freedesktop.org
Cc: Matthew Brost <matthew.brost@intel.com>
Subject: [PATCH i-g-t] tests/xe/xe_fault_injection: add split/skip bind coverage
Date: Tue, 17 Mar 2026 18:47:36 +0000 [thread overview]
Message-ID: <20260317184735.67051-2-matthew.auld@intel.com> (raw)
We were missing something for having to unwind a vector of binds where
there is splitting, as well as skipping.
Should explode with a dmesg WARN without the required kernel fix.
Link: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/7602
Signed-off-by: Matthew Auld <matthew.auld@intel.com>
Cc: Matthew Brost <matthew.brost@intel.com>
---
tests/intel/xe_fault_injection.c | 87 ++++++++++++++++++++++++++++++++
1 file changed, 87 insertions(+)
diff --git a/tests/intel/xe_fault_injection.c b/tests/intel/xe_fault_injection.c
index 4f4a7922c..c69303ba6 100644
--- a/tests/intel/xe_fault_injection.c
+++ b/tests/intel/xe_fault_injection.c
@@ -478,6 +478,89 @@ vm_bind_fail(int fd, const char pci_slot[], const char function_name[])
igt_assert_eq(simple_vm_bind(fd, vm), 0);
}
+static int
+simple_vm_bind_null(int fd, uint32_t vm)
+{
+ struct drm_xe_sync syncobj = {
+ .type = DRM_XE_SYNC_TYPE_SYNCOBJ,
+ .flags = DRM_XE_SYNC_FLAG_SIGNAL,
+ .handle = syncobj_create(fd, 0),
+ };
+ struct drm_xe_vm_bind bind = {
+ .vm_id = vm,
+ .num_binds = 1,
+ .bind.range = SZ_2G,
+ .bind.addr = 0,
+ .bind.op = DRM_XE_VM_BIND_OP_MAP,
+ .bind.pat_index = intel_get_pat_idx_uc(fd),
+ .bind.flags = DRM_XE_VM_BIND_FLAG_NULL,
+ .num_syncs = 1,
+ .syncs = (uintptr_t)&syncobj,
+ };
+
+ return igt_ioctl(fd, DRM_IOCTL_XE_VM_BIND, &bind);
+}
+
+static int
+vm_unmap_map_split_vector_null(int fd, uint32_t vm)
+{
+ struct drm_xe_sync syncobj = {
+ .type = DRM_XE_SYNC_TYPE_SYNCOBJ,
+ .flags = DRM_XE_SYNC_FLAG_SIGNAL,
+ .handle = syncobj_create(fd, 0),
+ };
+ struct drm_xe_vm_bind_op ops[] = {
+ {
+ .range = SZ_2G,
+ .addr = 0,
+ .op = DRM_XE_VM_BIND_OP_MAP,
+ .pat_index = intel_get_pat_idx_wb(fd),
+ .flags = DRM_XE_VM_BIND_FLAG_NULL,
+ },
+ {
+ .range = SZ_2M,
+ .addr = SZ_1G - SZ_2M,
+ .op = DRM_XE_VM_BIND_OP_UNMAP,
+ },
+ };
+ struct drm_xe_vm_bind bind = {
+ .vm_id = vm,
+ .num_binds = sizeof(ops)/sizeof(struct drm_xe_vm_bind_op),
+ .vector_of_binds = (uintptr_t)&ops,
+ .num_syncs = 1,
+ .syncs = (uintptr_t)&syncobj,
+ };
+
+ return igt_ioctl(fd, DRM_IOCTL_XE_VM_BIND, &bind);
+}
+/**
+ * SUBTEST: vm-bind-split-unwind-fail-%s
+ * Description: inject an error in function %arg[1] used in vm bind IOCTL
+ * to make it fail
+ * Functionality: fault
+ *
+ * arg[1]:
+ * @vm_bind_ioctl_ops_create: vm_bind_ioctl_ops_create
+ * @vm_bind_ioctl_ops_execute: vm_bind_ioctl_ops_execute
+ * @xe_pt_update_ops_prepare: xe_pt_update_ops_prepare
+ * @xe_pt_update_ops_run: xe_pt_update_ops_run
+ * @xe_vma_ops_alloc: xe_vma_ops_alloc
+ * @xe_sync_entry_parse: xe_sync_entry_parse
+ */
+static void
+vm_bind_split_unwind_fail(int fd, const char pci_slot[], const char function_name[])
+{
+ uint32_t vm = xe_vm_create(fd, 0, 0);
+
+ igt_assert_eq(simple_vm_bind_null(fd, vm), 0);
+
+ /* See: https://gitlab.freedesktop.org/drm/xe/kernel/-/issues/7602 */
+ injection_list_add(function_name);
+ set_retval(function_name, INJECT_ERRNO);
+ igt_assert(vm_unmap_map_split_vector_null(fd, vm) != 0);
+ injection_list_remove(function_name);
+}
+
/**
* SUBTEST: oa-add-config-fail-%s
* Description: inject an error in function %arg[1] used in oa add config IOCTL to make it fail
@@ -644,6 +727,10 @@ int igt_main_args("I:", NULL, help_str, opt_handler, NULL)
igt_subtest_f("vm-bind-fail-%s", s->name)
vm_bind_fail(fd, pci_slot, s->name);
+ for (const struct section *s = vm_bind_fail_functions; s->name; s++)
+ igt_subtest_f("vm-bind-split-unwind-fail-%s", s->name)
+ vm_bind_split_unwind_fail(fd, pci_slot, s->name);
+
for (const struct section *s = exec_queue_create_fail_functions; s->name; s++)
igt_subtest_f("exec-queue-create-fail-%s", s->name)
xe_for_each_engine(fd, hwe)
--
2.53.0
next reply other threads:[~2026-03-17 18:48 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-03-17 18:47 Matthew Auld [this message]
2026-03-18 1:19 ` ✓ Xe.CI.BAT: success for tests/xe/xe_fault_injection: add split/skip bind coverage Patchwork
2026-03-18 1:35 ` ✓ i915.CI.BAT: " Patchwork
2026-03-19 14:16 ` ✓ Xe.CI.FULL: " Patchwork
2026-03-26 22:54 ` ✓ Xe.CI.BAT: success for tests/xe/xe_fault_injection: add split/skip bind coverage (rev2) Patchwork
2026-03-26 22:58 ` ✓ i915.CI.BAT: " Patchwork
2026-03-27 15:59 ` ✓ Xe.CI.FULL: " Patchwork
2026-03-27 23:03 ` ✗ i915.CI.Full: failure " Patchwork
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20260317184735.67051-2-matthew.auld@intel.com \
--to=matthew.auld@intel.com \
--cc=igt-dev@lists.freedesktop.org \
--cc=matthew.brost@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox