On Thu, Feb 18, 2016 at 09:54:40AM +0100, Maarten Lankhorst wrote: > This is a small change to make encoder stealing work more as expected. I think the first step here is to type in unit tests with specific cases so that we can actually have a reasonable discussion about what's "expected". Currently there's lots of shuffling going on in this code. I think a simple drm_atomic_tests.ko driver which sets up a mostly-faked drm_device instance would be the fastest way to get there. Later on we could move some of that into userspace perhaps, but as long as it's a separate module reloading to re-run tests is pretty fast. -Daniel > First patch is a small cleanup, second patch is a small fix > for a crash I was hitting on my skylake with MST. The third patch > refuses to steal encoders from connectors not part of the state. > I had an alternate version that would restart in that case, but I feel > that in those cases the connector <-> encoder mapping would stay the same > regardless, so it would be easier to understand if we would instead fail > right away. > > Maarten Lankhorst (3): > drm/atomic: Always call steal_encoder. > drm/atomic: Refuse to steal encoders with index < conn_idx. > drm/atomic: Refuse to steal encoders from connectors not part of the > state. > > drivers/gpu/drm/drm_atomic_helper.c | 121 +++++++++++++----------------------- > 1 file changed, 44 insertions(+), 77 deletions(-) > > -- > 2.1.0 > > _______________________________________________ > Intel-gfx mailing list > Intel-gfx@xxxxxxxxxxxxxxxxxxxxx > https://lists.freedesktop.org/mailman/listinfo/intel-gfx -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ dri-devel mailing list dri-devel@xxxxxxxxxxxxxxxxxxxxx https://lists.freedesktop.org/mailman/listinfo/dri-devel
