[igt-dev] [PATCH i-g-t] Add TODO.rst

Daniel Vetter daniel.vetter at ffwll.ch
Thu Oct 11 16:36:32 UTC 2018


First attempt at collecting a bunch of cleanup tasks. I think using
the kernel as an example, it's much better if we keep these here than
in our own internal $bug_db instance.

v2: Fix typo (Rodrigo).

Cc: Petri Latvala <petri.latvala at intel.com>
Cc: Arkadiusz Hiler <arkadiusz.hiler at intel.com>
Cc: Maarten Lankhorst <maarten.lankhorst at linux.intel.com>
Cc: Rodrigo Vivi <rodrigo.vivi at intel.com>
Signed-off-by: Daniel Vetter <daniel.vetter at intel.com>
---
 TODO.rst | 31 +++++++++++++++++++++++++++++++
 1 file changed, 31 insertions(+)
 create mode 100644 TODO.rst

diff --git a/TODO.rst b/TODO.rst
new file mode 100644
index 000000000000..eb3e50f14ed4
--- /dev/null
+++ b/TODO.rst
@@ -0,0 +1,31 @@
+TODO
+====
+
+This contains a list of refactoring, cleanup and getting-started tasks around
+the IGT library.
+
+Split up igt_kms.c/igt_kms.h
+----------------------------
+
+igt_kms contains both a low-level modeset library, with thin convenience
+wrappers around core kernel code and libdrm. These functions usually have a
+drmtest_ prefix (but not all of them).
+
+The other part is a higher-level library around the igt_display and related
+structures. Those usually come with an igt_ prefix.
+
+The task would be to split this up, and where necessary, fix up the prefixes to
+match the level a function operates at.
+
+Remove property enums from igt_kms
+----------------------------------
+
+These are just needless indirection for writing tests. We can keep the #defines
+(since those strings are defacto uapi), but everything else is best handled by
+runtime-sizing all the arrays.
+
+Documentation
+-------------
+
+igt documentation is full of warnings and fairly incomplete. Pick a library, and
+work together with its authors to fix things up.
-- 
2.19.1



More information about the igt-dev mailing list