Commit Graph

60 Commits

Author SHA1 Message Date
bunnei
5861bacafd
Merge pull request #5795 from ReinUsesLisp/bytes-to-map-end
video_core/memory_manager: Add BytesToMapEnd
2021-01-29 22:56:29 -08:00
ReinUsesLisp
bda177ef40 video_core/memory_manager: Add BytesToMapEnd
Track map address sizes in a flat ordered map and add a method to query
the number of bytes until the end of a map in a given address.
2021-01-22 18:31:12 -03:00
ReinUsesLisp
b7febb5625 video_core/memory_manager: Remove unused CopyBlockUnsafe
This function was not being used.
2021-01-21 19:16:06 -03:00
ReinUsesLisp
0e9a6759f9 video_core/memory_manager: Flush destination buffer on CopyBlock
When we copy into a buffer, it might contain data modified from the GPU
on the same pages. Because of this, we have to flush the contents before
writing new data.

An alternative approach would be to write the data in place, but games
can also write data in other ways, invalidating our contents.

Fixes geometry in Zombie Panic in Wonderland DX.
2021-01-21 19:16:06 -03:00
ReinUsesLisp
dd790abab0 video_core/memory_manager: Add GPU address based flush method
Allow flushing rasterizer contents based on a GPU address.
2021-01-21 19:16:05 -03:00
ReinUsesLisp
9764c13d6d video_core: Rewrite the texture cache
The current texture cache has several points that hurt maintainability
and performance. It's easy to break unrelated parts of the cache
when doing minor changes. The cache can easily forget valuable
information about the cached textures by CPU writes or simply by its
normal usage.The current texture cache has several points that hurt
maintainability and performance. It's easy to break unrelated parts
of the cache when doing minor changes. The cache can easily forget
valuable information about the cached textures by CPU writes or simply
by its normal usage.

This commit aims to address those issues.
2020-12-30 03:38:50 -03:00
ameerj
eb67a45ca8 video_core: NVDEC Implementation
This commit aims to implement the NVDEC (Nvidia Decoder) functionality, with video frame decoding being handled by the FFmpeg library.

The process begins with Ioctl commands being sent to the NVDEC and VIC (Video Image Composer) emulated devices. These allocate the necessary GPU buffers for the frame data, along with providing information on the incoming video data. A Submit command then signals the GPU to process and decode the frame data.

To decode the frame, the respective codec's header must be manually composed from the information provided by NVDEC, then sent with the raw frame data to the ffmpeg library.

Currently, H264 and VP9 are supported, with VP9 having some minor artifacting issues related mainly to the reference frame composition in its uncompressed header.

Async GPU is not properly implemented at the moment.

Co-Authored-By: David <25727384+ogniK5377@users.noreply.github.com>
2020-10-26 23:07:36 -04:00
Lioncash
ff45c39578 General: Make use of std::nullopt where applicable
Allows some implementations to avoid completely zeroing out the internal
buffer of the optional, and instead only set the validity byte within
the structure.

This also makes it consistent how we return empty optionals.
2020-09-22 17:32:33 -04:00
bunnei
902bf6d37d
Merge pull request #4574 from lioncash/const-fn
memory_manager: Mark IsGranularRange() as a const member function
2020-08-25 11:24:13 -04:00
Lioncash
5bce81c3d6 memory_manager: Mark IsGranularRange() as a const member function
This doesn't modify internal member state, so it can be marked as const.
2020-08-24 00:37:57 -04:00
ReinUsesLisp
da53bcee60 video_core: Initialize renderer with a GPU
Add an extra step in GPU initialization to be able to initialize render
backends with a valid GPU instance.
2020-08-22 01:51:45 -03:00
bunnei
05def61398 hle: nvdrv: Rewrite of GPU memory management. 2020-07-26 00:49:43 -04:00
Lioncash
479605b3e5 memory_manager: Eliminate variable shadowing
Renames some variables to prevent ones in inner scopes from shadowing
outer-scoped variables.

The Copy* functions have no shadowing, but we rename them anyways to
remain consistent with the other functions.
2020-06-19 22:02:58 -04:00
bunnei
ff0c49e1ce
kernel: memory: Improve implementation of device shared memory. (#3707)
* kernel: memory: Improve implementation of device shared memory.

* fixup! kernel: memory: Improve implementation of device shared memory.

* fixup! kernel: memory: Improve implementation of device shared memory.
2020-04-23 11:37:12 -04:00
bunnei
32fc2aae3c video_core: memory_manager: Updates for Common::PageTable changes. 2020-04-17 00:59:34 -04:00
Fernando Sahmkow
1c18dc6577 Memory: Correct GCC errors. 2020-04-08 18:09:16 -04:00
Fernando Sahmkow
e00d992848 GPUMemoryManager: Improve safety of memory reads. 2020-04-08 12:08:06 -04:00
Fernando Sahmkow
9c0f40a1f5 GPU: Setup Flush/Invalidate to use VAddr instead of CacheAddr 2020-04-06 09:21:46 -04:00
bunnei
d65fa7d65c video_core: memory_manager: Flush/invalidate asynchronously on Unmap.
- Minor perf improvement.
2020-02-19 20:03:52 -05:00
bunnei
0f70f68fb3
Revert "video_core: memory_manager: Use GPU interface for cache functions." 2020-02-15 17:47:15 -05:00
bunnei
7cacb08cdf video_core: memory_manager: Use GPU interface for cache functions. 2020-02-07 22:59:35 -05:00
Lioncash
3f08e8d8d4 core/memory: Migrate over GetPointer()
With all of the interfaces ready for migration, it's trivial to migrate
over GetPointer().
2019-11-26 21:55:38 -05:00
Michael Scire
a1845d1dd3 prefer system reference over global accessor 2019-07-09 08:11:35 -07:00
Michael Scire
697206092e Prevent merging of device mapped memory blocks.
This sets the DeviceMapped attribute for GPU-mapped memory blocks,
and prevents merging device mapped blocks. This prevents memory
mapped from the gpu from having its backing address changed by
block coalesce.
2019-07-08 22:52:05 -07:00
Fernando Sahmkow
bdf9faab33 texture_cache: Handle uncontinuous surfaces. 2019-06-20 21:38:33 -03:00
Fernando Sahmkow
d86f9cd709 Change texture_cache chaching from GPUAddr to CacheAddr
This also reverses the changes to make invalidation and flushing through
the GPU address.
2019-06-20 21:36:12 -03:00
Fernando Sahmkow
6b0695b3cd Deglobalize Memory Manager on texture cahe and Implement Invalidation and Flushing using GPUVAddr 2019-06-20 21:36:11 -03:00
Lioncash
716fbaef74 video_core/memory_manager: Mark IsBlockContinuous() as a const member function
Corrects the typo in its name and marks the function as a const member
function, given it doesn't actually modify memory manager state.
2019-05-09 19:14:36 -04:00
Lioncash
fd12788967 video_core/memory_manager: Default the destructor within the cpp file
Makes the class less surprising when it comes to forward declaring the
type, and also prevents inlining the destruction code of the class,
given it contains non-trivial types.
2019-05-09 19:10:13 -04:00
Fernando Sahmkow
db4b2bc798 make ReadBlockunsafe and WriteBlockunsafe, ignore invalid pages. 2019-04-19 20:35:54 -04:00
Fernando Sahmkow
d0082de82a Implement IsBlockContinous
This detects when a GPU Memory Block is not continous within host cpu
memory.
2019-04-16 18:49:35 -04:00
Fernando Sahmkow
06d1c5a991 Document unsafe versions and add BlockCopyUnsafe 2019-04-16 10:11:35 -04:00
Fernando Sahmkow
ef381e6924 Use ReadBlockUnsafe on TIC and TSC reading
Use ReadBlockUnsafe on TIC and TSC reading as memory is never flushed
from host GPU there.
2019-04-15 23:10:24 -04:00
Fernando Sahmkow
367704aa82 GPU MemoryManager: Implement ReadBlockUnsafe and WriteBlockUnsafe 2019-04-15 23:01:35 -04:00
bunnei
20be92d5e6 memory_manager: Improved implementation of read/write/copy block.
- Fixes graphical issues with Chocobo's Mystery Dungeon EVERY BUDDY!
- Fixes a crash with Mario Tennis Aces
2019-04-05 23:43:34 -04:00
Lioncash
c13fbe6a41 video_core/memory_manager: Make Read() a const qualified member function
Given this doesn't actually alter internal state, this can be made a
const member function.
2019-04-05 20:30:48 -04:00
Lioncash
76ef6e5c2b video_core/memory_manager: Make ReadBlock() a const qualifier member function
Now, since we have a const qualified variant of GetPointer(), we can put
it to use in ReadBlock() to retrieve the source pointer that is passed
into memcpy.

Now block reading may be done from a const context.
2019-04-05 20:28:44 -04:00
Lioncash
34510bcda8 video_core/memory_manager: Add a const qualified variant of GetPointer()
Allows retrieving read-only pointers from a const context externally.
2019-04-05 20:25:28 -04:00
Lioncash
085b388a7a video_core/memory_manager: Make FindFreeRegion() a const member function
This doesn't modify internal state, so it can be made a const member
function.
2019-04-05 20:22:55 -04:00
Lioncash
9dec087fca video_core/memory_manager: Make GpuToCpuAddress() a const member function
This doesn't modify any internal state, so it can be made a const member
function to allow its use in const contexts.
2019-04-05 20:18:29 -04:00
bunnei
2117edd0f8 memory_manager: Cleanup FindFreeRegion. 2019-03-20 23:12:28 -04:00
bunnei
5a5fccaa23 memory_manager: Use Common::AlignUp in public interface as needed. 2019-03-20 22:58:49 -04:00
bunnei
72837e4b3d memory_manager: Bug fixes and further cleanup. 2019-03-20 22:36:03 -04:00
bunnei
197dcf0b5e memory_manager: Add protections for invalid GPU addresses.
- Avoid a crash in Xenoblade Chronicles 2.
2019-03-20 22:36:03 -04:00
bunnei
22d3dfbcd4 gpu: Rewrite virtual memory manager using PageTable. 2019-03-20 22:36:02 -04:00
bunnei
574e89d924 video_core: Refactor to use MemoryManager interface for all memory access.
# Conflicts:
#	src/video_core/engines/kepler_memory.cpp
#	src/video_core/engines/maxwell_3d.cpp
#	src/video_core/morton.cpp
#	src/video_core/morton.h
#	src/video_core/renderer_opengl/gl_global_cache.cpp
#	src/video_core/renderer_opengl/gl_global_cache.h
#	src/video_core/renderer_opengl/gl_rasterizer_cache.cpp
2019-03-16 00:38:48 -04:00
ReinUsesLisp
e01a9de35f memory_manager: Check for reserved page status 2019-02-03 04:58:24 -03:00
bunnei
0b1842294f memory_manager: Do not allow 0 to be a valid GPUVAddr.
- Fixes a bug with Undertale using 0 for a render target.
2018-11-23 12:58:55 -05:00
bunnei
4aa9779ae1 memory_manager: Do not MapBufferEx over already in use memory.
- This fixes rendering when changing areas in Super Mario Odyssey.
2018-11-01 18:57:59 -04:00
Frederic L
7a5eda5914 global: Use std::optional instead of boost::optional (#1578)
* get rid of boost::optional

* Remove optional references

* Use std::reference_wrapper for optional references

* Fix clang format

* Fix clang format part 2

* Adressed feedback

* Fix clang format and MacOS build
2018-10-30 00:03:25 -04:00