[i-g-t,v6,17/24] tests/core_hotunplug: Also check health of render device node

Commit Message

Janusz Krzysztofik Sept. 11, 2020, 10:30 a.m. UTC

Failures of subsequent tests accessing the render device node have been
observed on CI after late close of a hot rebound device.  Extend our
health check over the render node to detect that condition and start
our recovery phase with unbinding the driver from the device found
faulty.  Also, check health of both device nodes before running any
subtests.

Signed-off-by: Janusz Krzysztofik <[e-mail]>
---
 tests/core_hotunplug.c | 42 +++++++++++++++++++++++++++++++-----------
 1 file changed, 31 insertions(+), 11 deletions(-)

Patch

diff --git a/tests/core_hotunplug.c b/tests/core_hotunplug.c
index 6cf56d047..5e9eba8e7 100644
--- a/tests/core_hotunplug.c
+++ b/tests/core_hotunplug.c
@@ -55,13 +55,15 @@  struct hotunplug {
  * use drm_open_driver() since in case of an i915 device it opens it
  * twice and keeps a second file descriptor open for exit handler use.
  */
-static int local_drm_open_driver(const char *when, const char *why)
+static int local_drm_open_driver(bool render, const char *when, const char *why)
 {
 	int fd_drm;
 
-	igt_debug("%sopening device%s\n", when, why);
+	igt_debug("%sopening %s device%s\n", when, render ? "render" : "DRM",
+		  why);
 
-	fd_drm = __drm_open_driver(DRIVER_ANY);
+	fd_drm = render ? __drm_open_driver_render(DRIVER_ANY) :
+			  __drm_open_driver(DRIVER_ANY);
 	igt_assert_fd(fd_drm);
 
 	return fd_drm;
@@ -200,17 +202,15 @@  static void cleanup(struct hotunplug *priv)
 	priv->fd.sysfs_dev = close_sysfs(priv->fd.sysfs_dev);
 }
 
-static void healthcheck(struct hotunplug *priv)
+static void node_healthcheck(struct hotunplug *priv, bool render)
 {
 	/* preserve potentially dirty device status stored in priv->fd.drm */
 	bool closed = priv->fd.drm == -1;
 	int fd_drm;
 
-	/* device name may have changed, rebuild IGT device list */
-	igt_devices_scan(true);
-
-	priv->failure = "Device reopen failure!";
-	fd_drm = local_drm_open_driver("re", " for health check");
+	priv->failure = render ? "Render device reopen failure!" :
+				 "DRM device reopen failure!";
+	fd_drm = local_drm_open_driver(render, "re", " for health check");
 	if (closed)	/* store fd for cleanup if not dirty */
 		priv->fd.drm = fd_drm;
 
@@ -226,6 +226,16 @@  static void healthcheck(struct hotunplug *priv)
 	fd_drm = close_device(fd_drm, "", "health checked ");
 	if (closed || fd_drm < -1)	/* update status for post_healthcheck */
 		priv->fd.drm = fd_drm;
+}
+
+static void healthcheck(struct hotunplug *priv)
+{
+	/* device name may have changed, rebuild IGT device list */
+	igt_devices_scan(true);
+
+	node_healthcheck(priv, false);
+	if (!priv->failure)
+		node_healthcheck(priv, true);
 
 	/* not only request igt_abort on failure, also fail the health check */
 	igt_fail_on_f(priv->failure, "%s\n", priv->failure);
@@ -235,6 +245,11 @@  static void recover(struct hotunplug *priv)
 {
 	cleanup(priv);
 
+	/* unbind the driver from a possibly hot rebound unhealthy device */
+	if (!faccessat(priv->fd.sysfs_drv, priv->dev_bus_addr, F_OK, 0) &&
+	    priv->fd.drm == -1 && priv->failure)
+		driver_unbind(priv, "post ", 60);
+
 	if (faccessat(priv->fd.sysfs_bus, priv->dev_bus_addr, F_OK, 0))
 		bus_rescan(priv, 60);
 
@@ -295,7 +310,7 @@  static void unplug_rescan(struct hotunplug *priv)
 static void hotunbind_lateclose(struct hotunplug *priv)
 {
 	igt_assert_eq(priv->fd.drm, -1);
-	priv->fd.drm = local_drm_open_driver("", " for hot unbind");
+	priv->fd.drm = local_drm_open_driver(false, "", " for hot unbind");
 
 	driver_unbind(priv, "hot ", 0);
 
@@ -310,7 +325,7 @@  static void hotunbind_lateclose(struct hotunplug *priv)
 static void hotunplug_lateclose(struct hotunplug *priv)
 {
 	igt_assert_eq(priv->fd.drm, -1);
-	priv->fd.drm = local_drm_open_driver("", " for hot unplug");
+	priv->fd.drm = local_drm_open_driver(false, "", " for hot unplug");
 
 	device_unplug(priv, "hot ", 0);
 
@@ -348,6 +363,11 @@  igt_main
 		igt_assert_eq(close_device(fd_drm, "", "selected "), -1);
 
 		prepare(&priv);
+
+		node_healthcheck(&priv, false);
+		if (!priv.failure)
+			node_healthcheck(&priv, true);
+		igt_skip_on_f(priv.failure, "%s\n", priv.failure);
 	}
 
 	igt_subtest_group {