summaryrefslogtreecommitdiffstats
path: root/src/video_core/memory_manager.cpp (follow)
Commit message (Collapse)AuthorAgeFilesLines
* prefer system reference over global accessorMichael Scire2019-07-091-7/+5
|
* Prevent merging of device mapped memory blocks.Michael Scire2019-07-091-1/+23
| | | | | | | 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.
* texture_cache: Handle uncontinuous surfaces.Fernando Sahmkow2019-06-211-2/+3
|
* Change texture_cache chaching from GPUAddr to CacheAddrFernando Sahmkow2019-06-211-1/+1
| | | | | This also reverses the changes to make invalidation and flushing through the GPU address.
* Deglobalize Memory Manager on texture cahe and Implement Invalidation and Flushing using GPUVAddrFernando Sahmkow2019-06-211-1/+1
|
* video_core/memory_manager: Mark IsBlockContinuous() as a const member functionLioncash2019-05-101-2/+2
| | | | | Corrects the typo in its name and marks the function as a const member function, given it doesn't actually modify memory manager state.
* video_core/memory_manager: Default the destructor within the cpp fileLioncash2019-05-101-0/+2
| | | | | | 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.
* make ReadBlockunsafe and WriteBlockunsafe, ignore invalid pages.Fernando Sahmkow2019-04-201-4/+12
|
* Implement IsBlockContinousFernando Sahmkow2019-04-171-0/+8
| | | | | This detects when a GPU Memory Block is not continous within host cpu memory.
* Document unsafe versions and add BlockCopyUnsafeFernando Sahmkow2019-04-161-5/+13
|
* Use ReadBlockUnsafe on TIC and TSC readingFernando Sahmkow2019-04-161-0/+2
| | | | | Use ReadBlockUnsafe on TIC and TSC reading as memory is never flushed from host GPU there.
* GPU MemoryManager: Implement ReadBlockUnsafe and WriteBlockUnsafeFernando Sahmkow2019-04-161-0/+32
|
* memory_manager: Improved implementation of read/write/copy block.bunnei2019-04-061-10/+77
| | | | | - Fixes graphical issues with Chocobo's Mystery Dungeon EVERY BUDDY! - Fixes a crash with Mario Tennis Aces
* video_core/memory_manager: Make Read() a const qualified member functionLioncash2019-04-061-5/+5
| | | | | Given this doesn't actually alter internal state, this can be made a const member function.
* video_core/memory_manager: Make ReadBlock() a const qualifier member functionLioncash2019-04-061-1/+1
| | | | | | | | 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.
* video_core/memory_manager: Add a const qualified variant of GetPointer()Lioncash2019-04-061-2/+16
| | | | Allows retrieving read-only pointers from a const context externally.
* video_core/memory_manager: Make FindFreeRegion() a const member functionLioncash2019-04-061-9/+10
| | | | | This doesn't modify internal state, so it can be made a const member function.
* video_core/memory_manager: Make GpuToCpuAddress() a const member functionLioncash2019-04-061-2/+2
| | | | | This doesn't modify any internal state, so it can be made a const member function to allow its use in const contexts.
* memory_manager: Cleanup FindFreeRegion.bunnei2019-03-211-10/+4
|
* memory_manager: Use Common::AlignUp in public interface as needed.bunnei2019-03-211-11/+22
|
* memory_manager: Bug fixes and further cleanup.bunnei2019-03-211-66/+65
|
* memory_manager: Add protections for invalid GPU addresses.bunnei2019-03-211-15/+35
| | | | - Avoid a crash in Xenoblade Chronicles 2.
* gpu: Rewrite virtual memory manager using PageTable.bunnei2019-03-211-134/+338
|
* video_core: Refactor to use MemoryManager interface for all memory access.bunnei2019-03-161-9/+46
| | | | | | | | | | | # 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
* memory_manager: Check for reserved page statusReinUsesLisp2019-02-031-1/+2
|
* memory_manager: Do not allow 0 to be a valid GPUVAddr.bunnei2018-11-231-0/+7
| | | | - Fixes a bug with Undertale using 0 for a render target.
* memory_manager: Do not MapBufferEx over already in use memory.bunnei2018-11-011-27/+48
| | | | - This fixes rendering when changing areas in Super Mario Odyssey.
* global: Use std::optional instead of boost::optional (#1578)Frederic L2018-10-301-4/+4
| | | | | | | | | | | | | | | | * 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
* memory_manager: Add a method for querying the end of a mapped GPU region.bunnei2018-10-161-0/+10
|
* video_core/memory_manager: Replace a loop with std::array's fill() function in PageSlot()Lioncash2018-07-241-3/+1
| | | | | We already have a function that does what this code was doing, so let's use that instead.
* video_core/memory_manager: Avoid repeated unnecessary page slot lookupsLioncash2018-07-241-11/+21
| | | | | | | We don't need to keep calling the same function over and over again in a loop, especially when the behavior is slightly non-trivial. We can just keep a reference to the looked up location and do all the checking and assignments based off it instead.
* GPU: Allow GpuToCpuAddress to return boost::none for unmapped addresses.Subv2018-07-021-2/+2
|
* GPU: Implemented nvhost-as-gpu's UnmapBuffer ioctl.Subv2018-05-201-0/+19
| | | | It removes a mapping previously created with the MapBufferEx ioctl.
* memory_manager: Add implement CpuToGpuAddress.bunnei2018-04-241-0/+17
|
* memory_manager: Make GpuToCpuAddress return an optional.bunnei2018-04-241-1/+6
|
* memory_manager: Use GPUVAdddr, not PAddr, for GPU addresses.bunnei2018-04-241-36/+36
|
* GPU: Make the GPU virtual memory manager use 16 page bits and 10 page table bits.Subv2018-04-231-33/+20
| | | | Also removed some dead code and added memory map consistency asserts.
* Make a GPU class in VideoCore to contain the GPU state.Subv2018-02-121-0/+110
Also moved the GPU MemoryManager class to video_core since it makes more sense for it to be there.