non-exhaustive (but exhausting) list of changes:
* base laid for multiple window support, but will likely require more work to work correctly
* encapsulation of frontend state for proper multi-instance support
* (JIT still needs a fix for the NDS::Current workaround but we can get there later)
* new, more flexible configuration system
* nothing works yet
* don't double buffer 3D framebuffers for the GL Renderer
looks like leftovers from when 3D+2D composition was done in the frontend
* oops
* it works!
* implement display capture for compute renderer
it's actually just all stolen from the regular OpenGL renderer
* fix bad indirect call
* handle cleanup properly
* add hires rendering to the compute shader renderer
* fix UB
also misc changes to use more unsigned multiplication
also fix framebuffer resize
* correct edge filling behaviour when AA is disabled
* fix full color textures
* fix edge marking (polygon id is 6-bit not 5)
also make the code a bit nicer
* take all edge cases into account for XMin/XMax calculation
* use hires coordinate again
* stop using fixed size buffers based on scale factor in shaders
this makes shader compile times tolerable on Wintel
- beginning of the shader cache
- increase size of tile idx in workdesc to 20 bits
* apparently & is not defined on bvec4
why does this even compile on Intel and Nvidia?
* put the texture cache into it's own file
* add compute shader renderer properly to the GUI
also add option to toggle using high resolution vertex coordinates
* unbind sampler object in compute shader renderer
* fix GetRangedBitMask for 64 bit aligned 64 bits
pretty embarassing
* convert NonStupidBitfield.h back to LF only new lines
* actually adapt to latest changes
* fix stupid merge
* actually make compute shader renderer work with newest changes
* show progress on shader compilation
* remove merge leftover
* fix the pu region's end point overflowing
According to gericom it cannot overflow at all
* set a minimum and a better maximum for the pu region size
* fix pu logging
* PU regions with a size of 31 always take up the entire address space
also tweak some logging a little more
* start is actually force aligned by size, oops
* small tweaks
* hopefully more clear code
* math is for nerds
Precompute all 16 5-bit RGB palette colours into 8-bit RGBA to avoid
repeated and superfluous calculation within the nested loop at the
point of index lookup.
A speedup was observed, from ~7ms, to a consistent 1ms
(i.e. now practically instantaneous) through timing with
std::chrono::high_resolution_clock.
Also improve comprehensibility, by using meaningful names, where
appropriate, for loop counter variables.
* implement channels
* rework power-down support, fixing bugs
* fix bug when W_BeaconInterval is zero
* fix potential missing IRQs when writing to W_IE
GitHub Actions' macOS runners have Python from homebrew installed and it's used by default instead of the Python that ships with macOS. Apparently Homebrew decided you shouldn't be able to install stuff with `pip3` anymore so our build broke since `setuptools` is no longer included by default and `glib` from vcpkg needs it to build.
Additionally,, the whole liblzma mess ended up breaking our builds too because libarchive (and its dependency libxml2) depends on it and the download is no longer available. The build option changes here should be reverted once this is sorted out because this is probably partially breaking archive support.
PS: Fuck you Jia Tan.
Should remove the artifacts caused by the previous nearest
resampling. May be worth replacing with something better in the
future, but this is an improvement for now.
