Linuxydable/suyu
1
0
Fork 0
forked from suyu/suyu
Code Pull requests Activity
suyu/src/video_core/renderer_vulkan/vk_stream_buffer.cpp

165 lines
5.7 KiB
C++
Raw Normal View History

vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
// Copyright 2019 yuzu Emulator Project
// Licensed under GPLv2 or any later version
// Refer to the license.txt file included.
#include <algorithm>
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
#include <limits>
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
#include <optional>
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
#include <tuple>
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
#include <vector>
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
#include "common/alignment.h"
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
#include "common/assert.h"
#include "video_core/renderer_vulkan/vk_device.h"
#include "video_core/renderer_vulkan/vk_scheduler.h"
#include "video_core/renderer_vulkan/vk_stream_buffer.h"
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
#include "video_core/renderer_vulkan/wrapper.h"
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
namespace Vulkan {
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
namespace {
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
constexpr u64 WATCHES_INITIAL_RESERVE = 0x4000;
constexpr u64 WATCHES_RESERVE_CHUNK = 0x1000;
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
constexpr u64 PREFERRED_STREAM_BUFFER_SIZE = 256 * 1024 * 1024;
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
/// Find a memory type with the passed requirements
std::optional<u32> FindMemoryType(const VkPhysicalDeviceMemoryProperties& properties,
VkMemoryPropertyFlags wanted,
u32 filter = std::numeric_limits<u32>::max()) {
for (u32 i = 0; i < properties.memoryTypeCount; ++i) {
const auto flags = properties.memoryTypes[i].propertyFlags;
if ((flags & wanted) == wanted && (filter & (1U << i)) != 0) {
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
return i;
}
}
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
return std::nullopt;
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
}
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
/// Get the preferred host visible memory type.
u32 GetMemoryType(const VkPhysicalDeviceMemoryProperties& properties,
u32 filter = std::numeric_limits<u32>::max()) {
// Prefer device local host visible allocations. Both AMD and Nvidia now provide one.
// Otherwise search for a host visible allocation.
static constexpr auto HOST_MEMORY =
VK_MEMORY_PROPERTY_HOST_VISIBLE_BIT | VK_MEMORY_PROPERTY_HOST_COHERENT_BIT;
static constexpr auto DYNAMIC_MEMORY = HOST_MEMORY | VK_MEMORY_PROPERTY_DEVICE_LOCAL_BIT;
std::optional preferred_type = FindMemoryType(properties, DYNAMIC_MEMORY);
if (!preferred_type) {
preferred_type = FindMemoryType(properties, HOST_MEMORY);
ASSERT_MSG(preferred_type, "No host visible and coherent memory type found");
}
return preferred_type.value_or(0);
}
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
} // Anonymous namespace
video_core: Remove all Core::System references in renderer Now that the GPU is initialized when video backends are initialized, it's no longer needed to query components once the game is running: it can be done when yuzu is booting. This allows us to pass components between constructors and in the process remove all Core::System references in the video backend.
2020-06-12 02:24:45 +02:00
VKStreamBuffer::VKStreamBuffer(const VKDevice& device_, VKScheduler& scheduler_,
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
VkBufferUsageFlags usage)
video_core: Remove all Core::System references in renderer Now that the GPU is initialized when video backends are initialized, it's no longer needed to query components once the game is running: it can be done when yuzu is booting. This allows us to pass components between constructors and in the process remove all Core::System references in the video backend.
2020-06-12 02:24:45 +02:00
: device{device_}, scheduler{scheduler_} {
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
CreateBuffers(usage);
ReserveWatches(current_watches, WATCHES_INITIAL_RESERVE);
ReserveWatches(previous_watches, WATCHES_INITIAL_RESERVE);
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
VKStreamBuffer::~VKStreamBuffer() = default;
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
std::tuple<u8*, u64, bool> VKStreamBuffer::Map(u64 size, u64 alignment) {
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
ASSERT(size <= stream_buffer_size);
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
mapped_size = size;
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
if (alignment > 0) {
offset = Common::AlignUp(offset, alignment);
}
WaitPendingOperations(offset);
bool invalidated = false;
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
if (offset + size > stream_buffer_size) {
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
// The buffer would overflow, save the amount of used watches and reset the state.
invalidation_mark = current_watch_cursor;
current_watch_cursor = 0;
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
offset = 0;
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
// Swap watches and reset waiting cursors.
std::swap(previous_watches, current_watches);
wait_cursor = 0;
wait_bound = 0;
// Ensure that we don't wait for uncommitted fences.
scheduler.Flush();
invalidated = true;
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
return {memory.Map(offset, size), offset, invalidated};
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
void VKStreamBuffer::Unmap(u64 size) {
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
ASSERT_MSG(size <= mapped_size, "Reserved size is too small");
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
memory.Unmap();
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
offset += size;
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
if (current_watch_cursor + 1 >= current_watches.size()) {
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
// Ensure that there are enough watches.
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
ReserveWatches(current_watches, WATCHES_RESERVE_CHUNK);
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
auto& watch = current_watches[current_watch_cursor++];
watch.upper_bound = offset;
renderer_vulkan: Make unconditional use of VK_KHR_timeline_semaphore This reworks how host<->device synchronization works on the Vulkan backend. Instead of "protecting" resources with a fence and signalling these as free when the fence is known to be signalled by the host GPU, use timeline semaphores. Vulkan timeline semaphores allow use to work on a subset of D3D12 fences. As far as we are concerned, timeline semaphores are a value set by the host or the device that can be waited by either of them. Taking advantange of this, we can have a monolithically increasing atomic value for each submission to the graphics queue. Instead of protecting resources with a fence, we simply store the current logical tick (the atomic value stored in CPU memory). When we want to know if a resource is free, it can be compared to the current GPU tick. This greatly simplifies resource management code and the free status of resources should have less false negatives. To workaround bugs in validation layers, when these are attached there's a thread waiting for timeline semaphores.
2020-09-10 08:43:30 +02:00
watch.tick = scheduler.CurrentTick();
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
void VKStreamBuffer::CreateBuffers(VkBufferUsageFlags usage) {
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
const auto memory_properties = device.GetPhysical().GetMemoryProperties();
const u32 preferred_type = GetMemoryType(memory_properties);
const u32 preferred_heap = memory_properties.memoryTypes[preferred_type].heapIndex;
// Substract from the preferred heap size some bytes to avoid getting out of memory.
const VkDeviceSize heap_size = memory_properties.memoryHeaps[preferred_heap].size;
vk_stream_buffer: set allocable_size to 9 MiB This solves the crash on Linux systems running the current Linux Long Lived branch nVidia driver.
2020-07-10 03:28:32 +02:00
const VkDeviceSize allocable_size = heap_size - 9 * 1024 * 1024;
vk_stream_buffer: Make use of designated initializers where applicable
2020-07-17 01:21:42 +02:00
buffer = device.GetLogical().CreateBuffer({
.sType = VK_STRUCTURE_TYPE_BUFFER_CREATE_INFO,
.pNext = nullptr,
.flags = 0,
.size = std::min(PREFERRED_STREAM_BUFFER_SIZE, allocable_size),
.usage = usage,
.sharingMode = VK_SHARING_MODE_EXCLUSIVE,
.queueFamilyIndexCount = 0,
.pQueueFamilyIndices = nullptr,
});
vk_stream_buffer: Fix out of memory on boot on recent Nvidia drivers Nvidia recently introduced a new memory type for data streaming (awesome!), but yuzu was assuming that all heaps had enough memory for the assumed stream buffer size (256 MiB). This worked fine on AMD but Nvidia's new memory heap was smaller than 256 MiB. This commit changes this assumption and allocates a bit less than the size of the preferred heap, with a maximum of 256 MiB (to avoid allocating all system memory on integrated devices). - Fixes a crash on NVIDIA 450.82.0.0
2020-04-17 10:13:35 +02:00
const auto requirements = device.GetLogical().GetBufferMemoryRequirements(*buffer);
const u32 required_flags = requirements.memoryTypeBits;
stream_buffer_size = static_cast<u64>(requirements.size);
vk_stream_buffer: Make use of designated initializers where applicable
2020-07-17 01:21:42 +02:00
memory = device.GetLogical().AllocateMemory({
.sType = VK_STRUCTURE_TYPE_MEMORY_ALLOCATE_INFO,
.pNext = nullptr,
.allocationSize = requirements.size,
.memoryTypeIndex = GetMemoryType(memory_properties, required_flags),
});
renderer_vulkan: Drop Vulkan-Hpp
2020-03-27 05:33:21 +01:00
buffer.BindMemory(*memory, 0);
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
void VKStreamBuffer::ReserveWatches(std::vector<Watch>& watches, std::size_t grow_size) {
watches.resize(watches.size() + grow_size);
}
void VKStreamBuffer::WaitPendingOperations(u64 requested_upper_bound) {
if (!invalidation_mark) {
return;
}
while (requested_upper_bound < wait_bound && wait_cursor < *invalidation_mark) {
auto& watch = previous_watches[wait_cursor];
wait_bound = watch.upper_bound;
renderer_vulkan: Make unconditional use of VK_KHR_timeline_semaphore This reworks how host<->device synchronization works on the Vulkan backend. Instead of "protecting" resources with a fence and signalling these as free when the fence is known to be signalled by the host GPU, use timeline semaphores. Vulkan timeline semaphores allow use to work on a subset of D3D12 fences. As far as we are concerned, timeline semaphores are a value set by the host or the device that can be waited by either of them. Taking advantange of this, we can have a monolithically increasing atomic value for each submission to the graphics queue. Instead of protecting resources with a fence, we simply store the current logical tick (the atomic value stored in CPU memory). When we want to know if a resource is free, it can be compared to the current GPU tick. This greatly simplifies resource management code and the free status of resources should have less false negatives. To workaround bugs in validation layers, when these are attached there's a thread waiting for timeline semaphores.
2020-09-10 08:43:30 +02:00
scheduler.Wait(watch.tick);
vk_stream_buffer/vk_buffer_cache: Avoid halting and use generic cache The stream buffer before this commit once it was full (no more bytes to write before looping) waiting for all previous operations to finish. This was a temporary solution and had a noticeable performance penalty in performance (from what a profiler showed). To avoid this mark with fences usages of the stream buffer and once it loops wait for them to be signaled. On average this will never wait. Each fence knows where its usage finishes, resulting in a non-paged stream buffer. On the other side, the buffer cache is reimplemented using the generic buffer cache. It makes use of the staging buffer pool and the new stream buffer.
2020-01-06 21:59:20 +01:00
++wait_cursor;
}
vk_stream_buffer: Implement a stream buffer This manages two kinds of streaming buffers: one for unified memory models and one for dedicated GPUs. The first one skips the copy from the staging buffer to the real buffer, since it creates an unified buffer. This implementation waits for all fences to finish their operation before "invalidating". This is suboptimal since it should allocate another buffer or start searching from the beginning. There is room for improvement here. This could also handle AMD's "pinned" memory (a heap with 256 MiB) that seems to be designed for buffer streaming.
2019-02-24 08:22:33 +01:00
}
} // namespace Vulkan
Copy permalink