On Thu, Feb 21, 2019 at 02:23:10PM +0200, Laurent Pinchart wrote: > On Thu, Feb 21, 2019 at 12:19:13PM +0000, Brian Starkey wrote: [snip] > > > > I used a pre-existing internal tool which does exactly that. > > Any hope of sharing the sources ? > Not in a timescale or form which would be useful to you. I'm convinced people only ask questions like this to make us look like Bad Guys. Opening everything up is a process, and it's going to take us time. Sure we could be doing better, but I also think there's a lot of people who do worse. > > I appreciate that we don't have upstream tooling for writeback. As you > > say, it's a young API (well, not by date, but certainly by usage). > > > > I also do appreciate you taking the time to consider it, identifying > > issues which we did not, and for fixing them. The only way it stops > > being a young API, with bugs and no tooling, is if people adopt it. > > If the developers who initially pushed the API upstream without an > open-source test tool could spend a bit of time on this issue, I'm sure > it would help too :-) > No one suggested a test test tool before. In fact, the DRM subsystem explicitly requires that features land with something that isn't only a test tool, hence why we did drm_hwcomposer. That said, yes, we should be trying harder to get the igts landed. I personally think igts are far more useful than a random example C file, but I guess opinions differ. Thanks, -Brian
