From eb67a45ca82bc01ac843c853fd3c17f2a90e0250 Mon Sep 17 00:00:00 2001 From: ameerj Date: Mon, 26 Oct 2020 23:07:36 -0400 Subject: 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> --- src/video_core/memory_manager.cpp | 12 ++++++++++-- 1 file changed, 10 insertions(+), 2 deletions(-) (limited to 'src/video_core/memory_manager.cpp') diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp index 02cf53d15..6e70bd362 100644 --- a/src/video_core/memory_manager.cpp +++ b/src/video_core/memory_manager.cpp @@ -11,6 +11,7 @@ #include "video_core/gpu.h" #include "video_core/memory_manager.h" #include "video_core/rasterizer_interface.h" +#include "video_core/renderer_base.h" namespace Tegra { @@ -44,6 +45,12 @@ GPUVAddr MemoryManager::MapAllocate(VAddr cpu_addr, std::size_t size, std::size_ return Map(cpu_addr, *FindFreeRange(size, align), size); } +GPUVAddr MemoryManager::MapAllocate32(VAddr cpu_addr, std::size_t size) { + const std::optional gpu_addr = FindFreeRange(size, 1, true); + ASSERT(gpu_addr); + return Map(cpu_addr, *gpu_addr, size); +} + void MemoryManager::Unmap(GPUVAddr gpu_addr, std::size_t size) { if (!size) { return; @@ -108,7 +115,8 @@ void MemoryManager::SetPageEntry(GPUVAddr gpu_addr, PageEntry page_entry, std::s page_table[PageEntryIndex(gpu_addr)] = page_entry; } -std::optional MemoryManager::FindFreeRange(std::size_t size, std::size_t align) const { +std::optional MemoryManager::FindFreeRange(std::size_t size, std::size_t align, + bool start_32bit_address) const { if (!align) { align = page_size; } else { @@ -116,7 +124,7 @@ std::optional MemoryManager::FindFreeRange(std::size_t size, std::size } u64 available_size{}; - GPUVAddr gpu_addr{address_space_start}; + GPUVAddr gpu_addr{start_32bit_address ? address_space_start_low : address_space_start}; while (gpu_addr + available_size < address_space_size) { if (GetPageEntry(gpu_addr + available_size).IsUnmapped()) { available_size += page_size; -- cgit v1.2.3 From 9764c13d6d2977903f407761b27d847c0056e1c4 Mon Sep 17 00:00:00 2001 From: ReinUsesLisp Date: Wed, 30 Dec 2020 02:25:23 -0300 Subject: 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. --- src/video_core/memory_manager.cpp | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'src/video_core/memory_manager.cpp') diff --git a/src/video_core/memory_manager.cpp b/src/video_core/memory_manager.cpp index 6e70bd362..65feff588 100644 --- a/src/video_core/memory_manager.cpp +++ b/src/video_core/memory_manager.cpp @@ -57,7 +57,10 @@ void MemoryManager::Unmap(GPUVAddr gpu_addr, std::size_t size) { } // Flush and invalidate through the GPU interface, to be asynchronous if possible. - system.GPU().FlushAndInvalidateRegion(*GpuToCpuAddress(gpu_addr), size); + const std::optional cpu_addr = GpuToCpuAddress(gpu_addr); + ASSERT(cpu_addr); + + rasterizer->UnmapMemory(*cpu_addr, size); UpdateRange(gpu_addr, PageEntry::State::Unmapped, size); } -- cgit v1.2.3