Philip Rebohle

We get available display modes from win32 APIs; doesn't really sound like a DXVK bug. Does this work properly in other games? Edit: We do actually set the correct display...

Unless proven to also happen on Linux and Nvidia drivers, I'm going to **strongly** suspect just another AMD driver bug here. Game is known to work fine at least with...

And here's another reason why we don't support Windows. It's a known problem. EnumDisplaySettingsW just doesn't seem to work as expected but I'm not aware of any other way to...

No, and I don't plan to add one. I'd rather like to know why this just doesn't work on Windows. I don't have a high refresh rate display and no...

Ah, sorry, I should have been clearer - please run it *with DXVK*, i.e. with dxgi.dll next to the executable. Also, multiline code blocks on github work with three backticks,...

Yeah, so, it does actually detect the 120 and 144 Hz modes correctly. Note that we currently only report one monitor because wine used to have issues with multi-monitor setups...

Likely another issue caused by DXVK not supporting single-sampled rasterization to multisampled render targets. We can't really implement this in any resasonable way since the only Vulkan extension that would...

Can't reproduce here (6900XT, both mesa 22.0 and current git). It also doesn't make much sense for this to happen since the AO texture itself is computed via a compute...

Figured out what's wrong, D3D and Vulkan use different rules for computing texture coordinates when sampling with `NEAREST` filtering, and since the game uses a fullscreen quad to apply the...

> The game does not softlock without DXVK. the whole point of an apitrace is that we can see what the game is doing and perhaps reproduce the bug by...