<html> <head> <base href="https://bugs.freedesktop.org/"> </head> <body> <div> <a class="bz_bug_link bz_status_NEW " title="NEW - [CI][BAT] igt@i915_module_load@reload-with-fault-injection - incomplete - WARN_ON(!intel_irqs_enabled(dev_priv))" href="https://bugs.freedesktop.org/show_bug.cgi?id=109831#c3">Comment # 3</a> on <a class="bz_bug_link bz_status_NEW " title="NEW - [CI][BAT] igt@i915_module_load@reload-with-fault-injection - incomplete - WARN_ON(!intel_irqs_enabled(dev_priv))" href="https://bugs.freedesktop.org/show_bug.cgi?id=109831">bug 109831</a> from <a class="email" href="mailto:manasi.d.navare@intel.com" title="Manasi <manasi.d.navare@intel.com>"> Manasi</a> <pre>This failure is seen on KBL machines and last seen 1 day ago. Failure is in the igt@i915_module_load@reload-with-fault-injection test. This test injects a failure at various points during the driver load to test whether the unload-load(reload) can happen correctly. Failure analysis: Looking through the dmesg logs, the failure occurs at the failure injection checkpoint 9. This injects failure at a point when the worker threads are already created. Possible Cause: When we unload the driver due to the failure, the driver is not cleaning up/killing the worker threads correctly. Due to this drm_dp_mst_link_probe_work() continues to do the modeset and expects the vblank irqs in crtc_enable() while the IRqs are already disabled due to the unload causing the warnings about IRqs not enabled. Possible Fix: This could be fixed by properly handling the worker thread cleanup/kill in intel_modeset_cleanup() or earlier. Manasi</pre> </div> <hr> You are receiving this mail because: <ul> <li>You are the QA Contact for the bug.</li> <li>You are the assignee for the bug.</li> <li>You are on the CC list for the bug.</li> </ul> </body> </html>