Operating system: Linux-4.15.0-51-generic-x86_64-with-debian-buster-sid 64 Bits
Graphics card: Radeon RX Vega (VEGA10, DRM 3.23.0, 4.15.0-51-generic, LLVM 8.0.0) X.Org 4.5 (Core Profile) Mesa 19.0.2
Broken: version: 2.80 (sub 74), branch: master, commit date: 2019-07-11 13:50, hash: rB06312c6d2db8
Short description of error
Sometimes a 3D view can become corrupted and can't show camera view.
The viewport will not show anything.
Rendered viewport will flip between Path Tracing Sample counts but nothing will get done.
Rendering an image from the camera will work, suggesting it's not the camera itself that is erroneous.
Switching to other tabs with a 3D view will also work fine, suggesting it's that specific 3D view that is corrupted.
Exact steps for others to reproduce the error
In provided file:
- Observe that in Shading.002 tab, everything works - camera perspective and user perspective view, solid and rendered.
- Switch to Shading tab and you get a gray wall in camera perspective, while user perspective view works. Switching to rendered view while in camera perspective, the sample counter will just flip between a couple of numbers rather than complete.
I've had this happen a couple of times now, for no apparent reason - I can't reproduce on demand.
Clicking the add workspace quickly fixes it, but the first time I lost hours trying to figure out what was going on.
Trying to search for bugs here, all I got was the error message:
Unhandled Exception ("PhutilAggregateException")
Note: In my startup, I am using two linked math libraries (bunch of material node groups); one with failing image texture links and one with updated texture links. In the original work file (completely stripped down here), I was using linked blend containing objects in instanced collections and objects regularly instanced. Some objects have textures on multiple UVs (incl BastioniLab human). That's about the only "special" things I can remember about the real file.