Closed ruihe774 closed 3 months ago
Download the artifacts for this pull request:
cc @sfan5 @haasn
question: what's the usecase for float formats? where is it supported?
question: what's the usecase for float formats? where is it supported? because my AMD system doesn't
It's not AMD; Wayland is required to use float formats; X11 EGL does not support them.
float formats are used by Wayland color management protocol to support compositor-side color management and HDR processing; though it is not supported by mpv (and major compositors) now.
It's not AMD; Wayland is required to use float formats; X11 EGL does not support them.
I see.
I tested on a Wayland/Intel system I have and it supports rgba16f
but not rgba16
.
it supports
rgba16f
but notrgba16
.
Yep. Same as mine. Nothing bad to have support for some theoretically possible formats. The formats in this option list are also in the Vulkan surface format picking code of libplacebo, and I think they are there for reasons.
please rebase to move the documentation changes (docs: add docs for --egl-* options
) into the commit that adds the option
also squashing opengl: make format packing & unpacking beautiful
would make sense
This PR adds two options:
--egl-config-id
: select specific EGL config based on ID.--egl-output-format
: select EGL config based on format names.This allow users to customize swapchain format when using EGL.