On Sun 12 Oct 2025 at 15:38:29 +0100, Robert Swindells wrote: > I am using a Radeon HD-5450 with -current, works fine including playing > videos. And how about viewing https://webglsamples.org/aquarium/aquarium.html in Firefox (I use 142.0.1nb1 but I first saw it with one from a year ago, 123 or so I think it was). In particular *switching away* from the tab (or switching to a different workspace in ctwm) seems to trigger it for me. It takes 20-30 seconds maybe before it happens. I am starting to wonder about my setup with 10.1. Maybe it doesn't actually use accelleration, because this page https://webglreport.com/?v=2 reports Platform: Linux x86_64 Browser User Agent: Mozilla/5.0 (X11; NetBSD x86_64; rv:142.0) Gecko/20100101 Firefox/142.0 Context Name: webgl2 GL Version: WebGL 2.0 Shading Language Version: WebGL GLSL ES 3.00 Vendor: Mozilla Renderer: llvmpipe, or similar Unmasked Vendor: Mesa/X.org Unmasked Renderer: llvmpipe, or similar Antialiasing: Available ANGLE: No Major Performance Caveat: No and mentions llvmpipe. On the other hand it may be inaccurate since the Platform value is wrong. On the test machine it shows "Radeon HD 3200 Graphics, or similar". Firefox on the working 10.1 also reports Crash Annotation GraphicsCriticalError: |[0][GFX1-]: glxtest: DRM device has no render node (t=0.512481) [GFX1-]: glxtest: DRM device has no render node Crash Annotation GraphicsCriticalError: |[0][GFX1-]: glxtest: DRM device has no render node (t=0.512481) |[1][GFX1-]: glxtest: Cannot find DRM device (t=0.512559) [GFX1-]: glxtest: Cannot find DRM device ATTENTION: default value of option mesa_glthread overridden by environment. ATTENTION: default value of option mesa_glthread overridden by environment. On the test machine it only mentions "ATTENTION: default value of option mesa_glthread overridden by environment." so it may be trying to do something more which my working environment doesn't attempt. I do have /dev/dri/renderD128 though. On the test machine I have /dev/dri/renderD128 and ...D129. Moving both aside gets me the similar messages about "DRM device has no render node" and the webglreport then mentions llvmpipe too. Video playing on youtube still seems better than when I disable all accelleration in Firefox. The problem with the WebGL Aquarium doesn't happen. Moving back renderD128 gets me back to "Radeon HD 3200 Graphics, or similar", and the problem with the Aquarium. I notice a big diffrence in the reported frames per second: about 14 with llvmpipe, about 40 otherwise. So it starts to look like this isn't actually a regression at all, just some bug that's already in 10.x which I somehow avoided so far. I do have /dev/dri/renderD128, so there must be something else that causes the message "DRM device has no render node". And other programs do seem to use GL, such as emulators/vice. -Olaf. -- ___ Olaf 'Rhialto' Seibert <rhialto/at/falu.nl> \X/ There is no AI. There is just someone else's work. --I. Rose
Attachment:
signature.asc
Description: PGP signature