hugopeixoto/yuzu
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
17,156 Commits 2 Branches 0 Tags 148 MiB
e9446d232f
Commit Graph

17156 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
ReinUsesLisp
ae5725b709 host_shaders: Add texture color blit fragment shader 2020-12-30 02:00:48 -03:00
ReinUsesLisp
64fbf319f1 host_shaders: Add shaders to present to the swapchain 2020-12-30 01:59:12 -03:00
ReinUsesLisp
82b7daed9c host_shaders: Add shaders to convert between depth and color images 2020-12-30 01:48:44 -03:00
ReinUsesLisp
dc81a90640 host_shaders: Add compute shader to copy BC4 as RG32UI to RGBA8 2020-12-30 01:47:08 -03:00
ReinUsesLisp
5169ce9fcd host_shaders: Add shader to render a full screen triangle 2020-12-30 01:44:09 -03:00
ReinUsesLisp
59c46f9de9 host_shaders: Add pitch linear upload compute shader 2020-12-30 01:41:42 -03:00
ReinUsesLisp
12d16248dd host_shaders: Add block linear upload compute shaders 2020-12-30 01:39:35 -03:00
ReinUsesLisp
f20e18f60d host_shaders: Add copyright headers to OpenGL present shaders 2020-12-30 01:35:56 -03:00
ReinUsesLisp
95d156a150 video_core/host_shaders: Add support for prebuilt SPIR-V shaders 
Add support for building SPIR-V shaders from GLSL and generating headers
to include the text of those same GLSL shaders to consume from OpenGL.
 2020-12-30 01:29:07 -03:00
ReinUsesLisp
b3587102d1 core/memory: Read and write page table atomically 
Squash attributes into the pointer's integer, making them an uintptr_t
pair containing 2 bits at the bottom and then the pointer. These bits
are currently unused thanks to alignment requirements.

Configure Dynarmic to mask out these bits on pointer reads.

While we are at it, remove some unused attributes carried over from
Citra.

Read/Write and other hot functions use a two step unpacking process that
is less readable to stop MSVC from emitting an extra AND instruction in
the hot path:

 mov         rdi,rcx
 shr         rdx,0Ch
 mov         r8,qword ptr [rax+8]
 mov         rax,qword ptr [r8+rdx*8]
 mov         rdx,rax
-and         al,3
 and         rdx,0FFFFFFFFFFFFFFFCh
 je          Core::Memory::Memory::Impl::Read<unsigned char>
 mov         rax,qword ptr [vaddr]
 movzx       eax,byte ptr [rdx+rax]
 2020-12-29 21:54:49 -03:00
bunnei
85cfd96f62
Merge pull request #5247 from comex/xx-concepts 
k_priority_queue: Fix concepts use
 2020-12-29 16:50:20 -08:00
bunnei
82e0eeed21 hle: kernel: service_thread: Make thread naming more consistent. 2020-12-29 16:46:29 -08:00
bunnei
a2a0f5318d hle: kernel: Manage service threads on another thread. 
- This is to allow service threads to defer destruction of themselves.
 2020-12-29 16:46:29 -08:00
bunnei
69e82d01d5 common: ThreadWorker: Add class to help do asynchronous work. 2020-12-29 16:46:29 -08:00
bunnei
b02464f685
Merge pull request #5246 from comex/xx-include 
Add missing include of "core/hle/kernel/kernel.h"
 2020-12-29 16:43:17 -08:00
bunnei
c192da3f82 hle: kernel: Manage host thread IDs using TLS. 
- Avoids the need to have a large map of host to guest thread IDs.
 2020-12-29 15:55:30 -08:00
LC
8d55c8c855
Merge pull request #5248 from ReinUsesLisp/update-dynarmic 
externals: Update Dynarmic
 2020-12-29 18:11:30 -05:00
ReinUsesLisp
3f048c8646 externals: Update Dynarmic 
Keeps yuzu up to date with the latest changes and introduces a change
needed for a lock-free optimization our side.
 2020-12-29 19:30:52 -03:00
comex
388cf58b31 k_priority_queue: Fix concepts use 
- For `std::same_as`, add missing include of `<concepts>`.

- For `std::convertible_to`, create a replacement in `common/concepts.h`
  and use that instead.

  This would also be found in `<concepts>`, but unlike `std::same_as`,
  `std::convertible_to` is not yet implemented in libc++, LLVM's STL
  implementation - not even in master.  (In fact, `std::same_as` is the
  *only* concept currently implemented.  For some reason.)
 2020-12-29 14:33:41 -05:00
comex
b36896b90e Add missing include of "core/hle/kernel/kernel.h" 
This is needed as the header invokes methods on KernelCore.
 2020-12-29 14:22:35 -05:00
LC
aa87278bf0
Merge pull request #5245 from ameerj/sleepthread-log 
svc: demote SleepThread log to LOG_TRACE
 2020-12-29 14:03:24 -05:00
ameerj
0383363a8f svc: demote SleepThread log to LOG_TRACE 
This log is called often, and introduces a lot of noise when debug logging is enabled, making it difficult to see other debug logs.
 2020-12-29 14:01:56 -05:00
bunnei
22ba437aa4
Merge pull request #5236 from gal20/udp_client_patch 
input_common: process udp packets only for the correct pad
 2020-12-29 02:51:40 -08:00
bunnei
dfdac7d38a hle: kernel: Move ServiceThread ownership to KernelCore. 
- Fixes a circular dependency which prevented threads from being released on shutdown.
 2020-12-29 01:12:39 -08:00
bunnei
f57be2e626 hle: kernel: service_thread: Add thread name and take weak_ptr of ServerSession. 2020-12-29 01:06:39 -08:00
bunnei
7d77a3f88f hle: service: Acquire and release a lock on requests. 
- This makes it such that we can safely access service members from CoreTiming thread.
 2020-12-28 21:33:34 -08:00
bunnei
c7a06908ae audio_core: stream: Ensure buffer is valid before release. 2020-12-28 21:33:34 -08:00
bunnei
06f8c3dc01 core: Do not reset device_memory on shutdown. 
- This will be reset on initialization.
 2020-12-28 21:33:34 -08:00
bunnei
d0649d0971 core: hle: kernel: Clear process list on boot. 2020-12-28 21:33:34 -08:00
bunnei
954341763a gpu: gpu_thread: Ensure MicroProfile is shutdown on exit. 2020-12-28 21:33:34 -08:00
bunnei
994a9fec4e hle: service: vi: Refactor to grab buffer only once. 2020-12-28 21:33:34 -08:00
bunnei
6433b1dfd6 service: nvflinger: Improve synchronization for BufferQueue. 
- Use proper mechanisms for blocking on DequeueBuffer.
- Ensure service thread terminates on emulation Shutdown.
 2020-12-28 21:33:34 -08:00
bunnei
bea51d948d hle: service: Ensure system is powered on before writing IPC result. 2020-12-28 16:33:48 -08:00
bunnei
6d2f9428c5 core: kernel: Clear process list earlier. 2020-12-28 16:33:48 -08:00
bunnei
4991620f89 video_core: gpu_thread: Do not wait when system is powered down. 2020-12-28 16:33:48 -08:00
bunnei
916438a9de core: settings: Untangle multicore from asynchronous GPU. 
- Now that GPU is always threaded, we can support multicore with synchronous GPU.
 2020-12-28 16:33:48 -08:00
bunnei
40571c073f video_core: gpu: Implement synchronous mode using threaded GPU. 2020-12-28 16:33:48 -08:00
bunnei
14c825bd1c video_core: gpu: Refactor out synchronous/asynchronous GPU implementations. 
- We must always use a GPU thread now, even with synchronous GPU.
 2020-12-28 16:33:48 -08:00
bunnei
5d4715cc6a hle: kernel: hle_ipc: Remove SleepClientThread. 
- This was kind of hacky, and no longer is necessary with service threads.
 2020-12-28 16:33:48 -08:00
bunnei
87d6588cb5 hle: service: bsd: Update to work with service threads, removing SleepClientThread. 2020-12-28 16:33:48 -08:00
bunnei
0c81b83ca9 hle: service: nvdrv: Revert #4981 to remove usage of SleepClientThread. 
- Note, this always processes the ioctl right away, which fixes BotW 1.0.0 issues.
 2020-12-28 16:33:48 -08:00
bunnei
8bc3d66354 hle: kernel: service_thread: Add parameter for thread pool size. 2020-12-28 16:33:47 -08:00
bunnei
19a8f03ad5 hle: service: nvflinger: Refactor locking and interfaces. 2020-12-28 16:33:47 -08:00
bunnei
b377da042b hle: service: vi: Remove usage of SleepClientThread. 2020-12-28 16:33:47 -08:00
bunnei
28281ae250 core: hle: server_session: Use separate threads for each service connection. 2020-12-28 16:33:47 -08:00
bunnei
7dbdda908c
Merge pull request #5233 from german77/inverted 
InputCommon: Allow to invert analog axis with right click
 2020-12-28 14:06:21 -08:00
FearlessTobi
368b3ee227 .ci/templates: Enable QT translation for MSVC CI 
Previously this flag was missing, causing translation files not to be shipped with CI builds of yuzu.
 2020-12-28 15:54:02 +01:00
gal20
1defd0847a udp client: process packets only for the correct pad 2020-12-27 22:22:48 +02:00
german
80fece4e08 Allow to invert analog axis with right click 2020-12-26 17:46:14 -06:00
ReinUsesLisp
661483f313 renderer_vulkan/fixed_pipeline_state: Move enabled bindings to static state 
Without using VK_EXT_robustness2, we can't consider the 'enabled' (not
null) vertex buffers as dynamic state, as this leads to invalid Vulkan
state. Move this to static state that is always hashed and compared in
the pipeline key.

The bits for enabled vertex buffers are moved into the attribute state
bitfield. This is not 'correct' as it's not an attribute state, but that
struct has bits to spare, and it's used in an array of 32 elements (the
exact same number of vertex buffer bindings).
 2020-12-25 23:34:38 -03:00