[Intel-gfx] [CI 2/2] drm/i915/selftets: Check random hang recovery

Chris Wilson chris at chris-wilson.co.uk
Fri Apr 24 08:33:03 UTC 2020


Userspace is untrusted and allowed to submit anything to the GPU for
execution, including random garbage. Our recovery should do the right
thing.

Signed-off-by: Chris Wilson <chris at chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala at linux.intel.com>
---
 .../drm/i915/gem/selftests/i915_gem_context.c | 157 ++++++++++++++++++
 1 file changed, 157 insertions(+)

diff --git a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
index c5c3433174dc..4cda46cfbe2a 100644
--- a/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
+++ b/drivers/gpu/drm/i915/gem/selftests/i915_gem_context.c
@@ -2305,6 +2305,162 @@ static int igt_reg_isolation(void *arg)
 	return err;
 }
 
+static struct i915_request *
+exec_obj(struct i915_gem_context *ctx,
+	 struct intel_engine_cs *engine,
+	 struct drm_i915_gem_object *obj)
+{
+	struct intel_context *ce;
+	struct i915_request *rq;
+	struct i915_vma *vma;
+	int err;
+
+	ce = i915_gem_context_get_engine(ctx, engine->legacy_idx);
+	if (IS_ERR(ce))
+		return ERR_CAST(ce);
+
+	vma = i915_vma_instance(obj, ce->vm, NULL);
+	if (IS_ERR(vma)) {
+		err = PTR_ERR(vma);
+		goto err_ce;
+	}
+
+	err = i915_vma_pin(vma, 0, 0, PIN_USER);
+	if (err)
+		goto err_ce;
+
+	rq = intel_context_create_request(ce);
+	if (IS_ERR(rq))
+		goto err_unpin;
+
+	if (engine->emit_init_breadcrumb) {
+		err = engine->emit_init_breadcrumb(rq);
+		if (err)
+			goto err_rq;
+	}
+
+	i915_vma_lock(vma);
+	err = i915_request_await_object(rq, vma->obj, false);
+	if (err == 0)
+		err = i915_vma_move_to_active(vma, rq, 0);
+	i915_vma_unlock(vma);
+	if (err)
+		goto err_rq;
+
+	err = engine->emit_bb_start(rq, vma->node.start, vma->node.size, 0);
+
+err_rq:
+	if (!err)
+		i915_request_get(rq);
+	else
+		i915_request_set_error_once(rq, err);
+	i915_request_add(rq);
+
+err_unpin:
+	i915_vma_unpin(vma);
+err_ce:
+	intel_context_put(ce);
+	return err ? ERR_PTR(err) : rq;
+}
+
+static int randomise_object(struct drm_i915_gem_object *obj,
+			    struct rnd_state *prng)
+{
+	unsigned long i;
+	u32 *cs;
+
+	cs = i915_gem_object_pin_map(obj, I915_MAP_WC);
+	if (IS_ERR(cs))
+		return PTR_ERR(cs);
+
+	for (i = 0; i < obj->base.size / sizeof(*cs); i++)
+		cs[i] = prandom_u32_state(prng);
+
+	i915_gem_object_flush_map(obj);
+	i915_gem_object_unpin_map(obj);
+
+	return 0;
+}
+
+static bool skip_garbage(const struct intel_engine_cs *engine)
+{
+	if (IS_GEN(engine->i915, 6))
+		return true;
+
+	return false;
+}
+
+static int igt_ctx_garbage(void *arg)
+{
+	struct drm_i915_private *i915 = arg;
+	struct drm_i915_gem_object *obj;
+	struct intel_engine_cs *engine;
+	struct i915_gem_context *ctx;
+	I915_RND_STATE(prng);
+	struct file *file;
+	int err;
+
+	/*
+	 * Imagine if userspace went crazy and submitted a batch of nothing
+	 * but random junk. The GPU may hang and we may be forced to kill
+	 * the context; but the driver will go on!
+	 */
+
+	file = mock_file(i915);
+	if (IS_ERR(file))
+		return PTR_ERR(file);
+
+	ctx = live_context(i915, file);
+	if (IS_ERR(ctx)) {
+		err = PTR_ERR(ctx);
+		goto out_file;
+	}
+
+	obj = i915_gem_object_create_internal(i915, 4096);
+	if (IS_ERR(obj)) {
+		err = PTR_ERR(obj);
+		goto out_file;
+	}
+
+	err = randomise_object(obj, &prng);
+	if (err)
+		goto out_obj;
+
+	i915_gem_context_clear_bannable(ctx);
+	for_each_uabi_engine(engine, i915) {
+		struct i915_request *rq;
+
+		if (!IS_ENABLED(CONFIG_DRM_I915_SELFTEST_BROKEN) &&
+		    skip_garbage(engine))
+			continue;
+
+		rq = exec_obj(ctx, engine, obj);
+		if (IS_ERR(rq)) {
+			err = PTR_ERR(rq);
+			break;
+		}
+
+		if (i915_request_wait(rq,
+				      I915_WAIT_INTERRUPTIBLE,
+				      HZ / 2) < 0) {
+			intel_gt_handle_error(engine->gt,
+					      engine->mask, 0,
+					      NULL);
+		}
+
+		i915_request_put(rq);
+	}
+
+	if (igt_flush_test(i915))
+		err = -EIO;
+
+out_obj:
+	i915_gem_object_put(obj);
+out_file:
+	fput(file);
+	return err;
+}
+
 static bool skip_unused_engines(struct intel_context *ce, void *data)
 {
 	return !ce->state;
@@ -2441,6 +2597,7 @@ int i915_gem_context_live_selftests(struct drm_i915_private *i915)
 		SUBTEST(igt_shared_ctx_exec),
 		SUBTEST(igt_vm_isolation),
 		SUBTEST(igt_reg_isolation),
+		SUBTEST(igt_ctx_garbage),
 	};
 
 	if (intel_gt_is_wedged(&i915->gt))
-- 
2.20.1



More information about the Intel-gfx mailing list