Dark Mod 2.09 - graphics issues

[noni]

1. Cu r_useNewBackend 1 = rău, și cu r_useNewBackend 0 = bun.
2. Nu pot seta o rezoluție mai mică (de exemplu, 1366x768) în fullscreen. Acesta rămâne pe rezoluția nativă a laptopului (1920x1080).
3. Specificații laptop.

[nbohr1more]

It's a graphics bug with AMD drivers.

Disable bindless textures:

r_useBindlessTextures 0

[V-Man339]

This worked for me, thank you.

[duzenko]

NHAT mission 1

view pos 1705 3835 -68 -13 -25 0

old backend: 32fps

new backend: 26 fps

What is the command name to show CPU and GPU times? Can't find/remember it.

 

[duzenko]

@cabalisticThere could be a a vertex cache issue here

r_skipModels 1 sometimes doubles FPS with little changes in draw call and tri count

The backend spends most of the time in RB_DrawElementsWithCounters

Curiously the zombie pit in the same map has higher FPS

[duzenko]

@stgatilov

How is it possible that r_usePortals 0 hides animated models?

 

[duzenko]

54 minutes ago, duzenko said:

@cabalisticThere could be a a vertex cache issue here

r_skipModels 1 sometimes doubles FPS with little changes in draw call and tri count

The backend spends most of the time in RB_DrawElementsWithCounters

Curiously the zombie pit in the same map has higher FPS

 

 

Never mind, it's particles overdrawing

[HMart]

On 2/15/2021 at 3:25 PM, nbohr1more said:

It's a graphics bug with AMD drivers.

Disable bindless textures:

r_useBindlessTextures 0

This person also add problems with BT on some AMD laptop GPU's and was able to solve it by doing the following 

Quote

So the solution in my case was to make sure to get "dynamically uniform" initialization of my bindless texture sampler. For example the simplest way is to turn the whole sampler or the texture handle into a uniform and set it via glUniform* commands from the client side. Of course this again makes the whole bindless texture extension less useful in my opinion. So i get the handle via a Uniform Buffer (instead of Texture Buffer - which apparently retains "dynamically uniform[ity]").

Is this something you guys are aware? 

[cabalistic]

We already get the handle via a uniform buffer. Our issue is almost certainly a bug in AMD's shader compiler. I have found a partial workaround, but not a full fix (as of yet).