Diamond75/Panda3DS
1
0
Fork 0
mirror of https://github.com/wheremyfoodat/Panda3DS.git synced 2025-04-18 11:41:31 +12:00
Code Issues Projects Releases Packages Wiki Activity Actions
762 commits 21 branches 10 tags 37 MiB
Commit graph

587 commits

Author SHA1 Message Date
fleroviux
13a53a9e9d [GPU] Disable the alpha combiner when the color combiner is in "Dot3 RGBA" configuration 2023-06-20 20:46:01 +02:00
fleroviux
a30f9ccd33 [GPU] Treat color and alpha sources as two separate RGBA values 
Previously we fetched the RGB channels of the color source and the alpha channel of the alpha source
and combined these into a single RGBA source value that the source operands are applied to.
This is incorrect. The correct behavior is to fetch the full RGBA color for both the color and alpha
sources and to apply the color operand to the color source and the alpha operand to the alpha source,
i.e. using the *alpha* source R-channel when the alpha operand is "One minus source red".
Only after that we combine both sources into a single RGBA color.
 2023-06-20 20:20:33 +02:00
wheremyfoodat
dbe293bb52
Merge pull request #43 from Wunkolo/pica-types 
Add PICA texel-format and topology types
 2023-06-20 13:26:58 +03:00
Wunkolo
1b9f270b19 Migrate IOFile implementation to io_file.cpp 
Makes the implementation of `IOFile` private, allowing inclusions and
defines such as `#define fseeko` and `#include <io.h>` to not poison
client-code or the global namespace.
 2023-06-19 19:37:36 -07:00
fleroviux
9b1ed2084f [GPU] Implement initial support for multi-texturing 2023-06-20 00:01:12 +02:00
fleroviux
170fe9a4b2 Merge branch 'GamingProcessingUnit' of github.com:fleroviux/Panda3DS into pica-tev-emulation 2023-06-19 19:54:03 +02:00
wheremyfoodat
6b5ec74aa4 [OutputVertex] Avoid unnamed struct 2023-06-18 23:59:41 +03:00
fleroviux
927bf9fe32 Merge branch 'GamingProcessingUnit' of github.com:fleroviux/Panda3DS into pica-tev-emulation 2023-06-18 22:27:24 +02:00
wheremyfoodat
952e2d6d22 More accurate VAO implementation 2023-06-18 23:20:28 +03:00
Sky
e9b3b7c2a2 Add workaround helper for bit_cast and missing #include <bit> 
Allows compilation on native macOS build system.
 2023-06-18 12:33:34 -07:00
Wunkolo
6ee3f73335 Bisect TextureFmt and ColorFmt 
Makes framebuffer-formats unrepresentable from texture formats while
allowing them to alias each other. Add utility functions as well that
just re-use the `TextureFmt` ones.
 2023-06-18 08:22:14 -07:00
Wunkolo
134e16d8ea Replace Texture::Formats with PICA::ColorFmt 
This enum was a duplicate of the PICA one. Now the PICA one is more
complete. Migrate `textureFormatToString` to be next to the enums
definition.
 2023-06-18 00:10:04 -07:00
Wunkolo
838d3f27f9 Migrate PICA-types into PICA namespace 
Rather than prefixing these types with `PICA`, a namespace is used instead.
 2023-06-17 18:23:47 -07:00
fleroviux
1416456035 [GPU] DOT3_RGB(A) modes should remap the operands from [0, 1] to [-1, +1] range 2023-06-18 01:32:32 +02:00
Wunkolo
78a3f9fa23 Add PICA texel-format and topology types 
Slowly stepping the codebase towards having renderer-agnostic types and keeping the translation of PICA-types to OpenGL/VK/DX/Software/etc to the renderer-backend.
 2023-06-17 12:26:41 -07:00
wheremyfoodat
d0ae5f0546
Merge pull request #37 from marysaka/feat/proper-hle-result 
hle: Add proper type for result code
 2023-06-17 18:39:36 +03:00
fleroviux
78827886fe Merge branch 'master' of github.com:fleroviux/Panda3DS into pica-tev-emulation 2023-06-17 12:39:48 +02:00
fleroviux
3aeef23b2e Merge branch 'bottom-screen' of github.com:fleroviux/Panda3DS into pica-tev-emulation 2023-06-17 12:38:33 +02:00
Mary
f3010f0909 misc: Switch from gl3w to glad 
Also add more checks on window and context creation.
 2023-06-17 11:26:18 +02:00
fleroviux
c9642292aa [GPU] Do scale on the final texture combiner output, not the individual sources 2023-06-17 02:06:05 +02:00
fleroviux
b854367676 [GPU] Move TEV uniforms update into its own method 2023-06-17 01:58:30 +02:00
fleroviux
a368d77005 [GPU] respect GPUREG_TEXENVi_SCALE 2023-06-17 01:42:53 +02:00
fleroviux
aef808cdfc [GPU] Cleanup TEV shader code to be less redundant 2023-06-17 01:35:35 +02:00
fleroviux
c3acf2b393 [GPU] cleanup and optimize the TEV shader functions 2023-06-17 01:27:26 +02:00
wheremyfoodat
5b3266be34 Fix merge conflict 2023-06-17 02:22:19 +03:00
wheremyfoodat
7b859ad602 Merge branch 'master' into bottom-screen 2023-06-17 02:19:30 +03:00
wheremyfoodat
1aab10bec7
Merge pull request #35 from Wunkolo/span-cpu-state 
Use `std::span` for CPU register-state
 2023-06-17 01:39:33 +03:00
fleroviux
76d6140797 [GPU] Fix order of operands to mix() in the interpolation combine mode. 2023-06-17 00:12:07 +02:00
fleroviux
bcf6401d4d [GPU] Output mangenta when an unimplemented TEV source is used 2023-06-16 23:39:05 +02:00
wheremyfoodat
936302da2a Raise clang-format column size, make std::span support in opengl.hpp version-dependent 2023-06-16 22:06:59 +03:00
Mary
122b1b2727 hle: Add proper type for result code 
This should clean up all HLE errorcode in the codebase.

I didn't removed Rust::Result as this should be a cleanup for another
iteration.
 2023-06-16 20:09:37 +02:00
Wunkolo
553d23974a Use std::span to pass vertex data 
Starts utilizing
[std::span](https://en.cppreference.com/w/cpp/container/span) to
indicate a non-owning view of a contiguous array of elements rather than
`T* data, usize count`.
 2023-06-16 07:28:35 -07:00
Wunkolo
fde93381a5 Use std::span for CPU register-state 
Following the trend of #33: `std::span` provides some useful utility
functions like `size_bytes()` and `as_bytes()` and serves as a better
non-owning "chunk of data"-type over just passing around an
`std::array&`.
 2023-06-16 07:24:37 -07:00
wheremyfoodat
519fb35c5f Less UB 2023-06-16 14:22:29 +03:00
fleroviux
2e85f30c10 [GPU] attempt to emulate the previous TEV color buffer 2023-06-16 01:34:11 +02:00
fleroviux
d153133070 [GPU] fix missing break statements in 'previous' combiner mode implementation 2023-06-16 00:52:00 +02:00
fleroviux
1aa4dea63e [GPU] Add the remaining two TEV units 
Looks like I overlooked the fact there actually are fix texture combiners,
not just four. The last two texture combiners are easy to miss, because
there are other GPU registers between the first four combiners and the last two.
 2023-06-16 00:41:48 +02:00
fleroviux
e7435e3fdf [GPU] implement TEV combiner chaining and the remaining combine modes 2023-06-16 00:33:42 +02:00
fleroviux
525890bff7 [GPU] Upload configuration for all texture combiners to the shader program 
Upload the configuration register values for the remaining texture combiners to the
shader program's uniforms. This is in preparation for emulating those texture combiners next.
 2023-06-15 23:27:33 +02:00
fleroviux
e66370598a Fix the formatting which I broke (sorry) 2023-06-15 23:01:16 +02:00
fleroviux
4322ebda2c [GPU] Start texture combiner implementation 
This commit first implements experimental and incomplete
texture combiner support. Currently only the first texture combiner
is implemented. Many sources and combine modes are not implemented yet.
 2023-06-15 22:45:13 +02:00
wheremyfoodat
79c3cf2661 [Shader JIT] Less UB 2023-06-15 19:30:51 +03:00
wheremyfoodat
3cf4bdd95f [Shader JIT] Use rotl+xor for combining hashes 
Co-Authored-By: Wunk <wunkolo@gmail.com>
 2023-06-15 01:58:00 +03:00
wheremyfoodat
87ac4d5a0f {GL renderer] Better screen stub 2023-06-14 21:47:17 +03:00
wheremyfoodat
f0f7327b90 [PICA] Switch to xxh3 by default 2023-06-14 15:39:29 +03:00
wheremyfoodat
f82b27ddba Add xxhash 2023-06-14 14:46:11 +03:00
wheremyfoodat
3606c005c5 Merge branch 'master' into dynapica 2023-06-14 13:55:15 +03:00
wheremyfoodat
8284c3cda4 Remove printf 2023-06-14 13:49:47 +03:00
wheremyfoodat
8e90b8a103 [GPU] Bottom screen stub 2023-06-13 22:47:01 +03:00
wheremyfoodat
2520b2cb46 [GPU] Disable blend when rendering output texture 2023-06-13 20:38:17 +03:00