Having lints would prove beneficial to prevent the same dark patterns from reoccuring over and over again in drivers. Add a TODO entry for that. Signed-off-by: Maxime Ripard <mripard@xxxxxxxxxx> --- Changes from v1: - Add allow_modeset and cec phys address handling --- Documentation/gpu/todo.rst | 34 ++++++++++++++++++++++++++++++++++ 1 file changed, 34 insertions(+) diff --git a/Documentation/gpu/todo.rst b/Documentation/gpu/todo.rst index 030ef6bec82a..9e0e2df3a59a 100644 --- a/Documentation/gpu/todo.rst +++ b/Documentation/gpu/todo.rst @@ -724,10 +724,44 @@ doesn't or document it if it does would be of great help. Contact: Maxime Ripard <mripard@xxxxxxxxxx> Level: Intermediate +Create lints for KMS drivers +---------------------------- + +Over time, we've accumulated a list of dark patterns in KMS drivers that +should be avoided. However, none of them are published anywhere, and not +all reviewers are aware of them. It creates a situation where we have +more drivers with problematic / deprecated code, even though we know +that they shouldn't. + +We should create a set of coccinelle scripts that match these patterns, +and make new drivers run that list. And possibly integrate them in CI. + +These patterns include: + + - Drivers using kzalloc() or devm_kzalloc() to allocate their memory, + instead of drmm_kzalloc(). + + - Drivers not protecting their device resources (MMIO, clocks, + regulators, etc.) by drm_dev_enter() and drm_dev_exit(). + + - Drivers using drm_dev_unregister() instead of drm_dev_unplug(). + + - Drivers not calling drm_atomic_helper_shutdown() at shutdown + + - Drivers using drm_dp_cec_set_edid(), cec_get_edid_phys_addr(), or + cec_s_phys_addr_from_edid() instead of drm_dp_cec_attach() or + cec_s_phys_addr(). + + - Drivers setting drm_atomic_state->allow_modeset manually. + +Contact: Maxime Ripard <mripard@xxxxxxxxxx> + +Level: Intermediate + Enable trinity for DRM ---------------------- And fix up the fallout. Should be really interesting ... -- 2.45.2
