vtk ctest has many failures which appear to be caused by its selection of a MSAA visual and a missing resolve on the read pixels - the image it gets is stretched 4x horizontally and 2x vertically. https://open.cdash.org/viewTest.php?onlyfailed&buildid=4530390 Tested with ubuntu-16.04 nouveau driver and a GeForce GTX 750 Ti (GM107).
As I mentioned on IRC, I strongly believe this is due to the following issue: https://trello.com/c/jzBpfrzL/134-resolve-on-transfer-map-upscale-on-transfer-unmap Basically we're supposed to do a resolve on transfer_map, return the resolved image, and if any of it is changed, copy back at unmap time while up-scaling. We do none of that right now, and 8x MSAA is laid out in a 4x2 grid for each pixel, so it makes sense that the output would be "scaled up" that way.
-- GitLab Migration Automatic Message -- This bug has been migrated to freedesktop.org's GitLab instance and has been closed from further activity. You can subscribe and participate further through the new bug through this link to our GitLab instance: https://gitlab.freedesktop.org/mesa/mesa/issues/1111.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.