dolphin/Source/Core/VideoBackends/Vulkan/VKGfx.cpp

610 lines
21 KiB
C++

// Copyright 2016 Dolphin Emulator Project
// SPDX-License-Identifier: GPL-2.0-or-later
#include "VideoBackends/Vulkan/VKGfx.h"
#include <algorithm>
#include <cstddef>
#include <cstdio>
#include <limits>
#include <string>
#include <tuple>
#include "Common/Assert.h"
#include "Common/CommonTypes.h"
#include "Common/Logging/Log.h"
#include "Common/MsgHandler.h"
#include "VideoBackends/Vulkan/CommandBufferManager.h"
#include "VideoBackends/Vulkan/ObjectCache.h"
#include "VideoBackends/Vulkan/StateTracker.h"
#include "VideoBackends/Vulkan/VKPipeline.h"
#include "VideoBackends/Vulkan/VKShader.h"
#include "VideoBackends/Vulkan/VKSwapChain.h"
#include "VideoBackends/Vulkan/VKTexture.h"
#include "VideoBackends/Vulkan/VKVertexFormat.h"
#include "VideoCommon/DriverDetails.h"
#include "VideoCommon/FramebufferManager.h"
#include "VideoCommon/Present.h"
#include "VideoCommon/RenderState.h"
#include "VideoCommon/VideoConfig.h"
namespace Vulkan
{
VKGfx::VKGfx(std::unique_ptr<SwapChain> swap_chain, float backbuffer_scale)
: m_swap_chain(std::move(swap_chain)), m_backbuffer_scale(backbuffer_scale)
{
UpdateActiveConfig();
for (SamplerState& sampler_state : m_sampler_states)
sampler_state = RenderState::GetPointSamplerState();
// Various initialization routines will have executed commands on the command buffer.
// Execute what we have done before beginning the first frame.
ExecuteCommandBuffer(true, false);
}
VKGfx::~VKGfx() = default;
bool VKGfx::IsHeadless() const
{
return m_swap_chain == nullptr;
}
std::unique_ptr<AbstractTexture> VKGfx::CreateTexture(const TextureConfig& config,
std::string_view name)
{
return VKTexture::Create(config, name);
}
std::unique_ptr<AbstractStagingTexture> VKGfx::CreateStagingTexture(StagingTextureType type,
const TextureConfig& config)
{
return VKStagingTexture::Create(type, config);
}
std::unique_ptr<AbstractShader>
VKGfx::CreateShaderFromSource(ShaderStage stage, std::string_view source, std::string_view name)
{
return VKShader::CreateFromSource(stage, source, name);
}
std::unique_ptr<AbstractShader> VKGfx::CreateShaderFromBinary(ShaderStage stage, const void* data,
size_t length, std::string_view name)
{
return VKShader::CreateFromBinary(stage, data, length, name);
}
std::unique_ptr<NativeVertexFormat>
VKGfx::CreateNativeVertexFormat(const PortableVertexDeclaration& vtx_decl)
{
return std::make_unique<VertexFormat>(vtx_decl);
}
std::unique_ptr<AbstractPipeline> VKGfx::CreatePipeline(const AbstractPipelineConfig& config,
const void* cache_data,
size_t cache_data_length)
{
return VKPipeline::Create(config);
}
std::unique_ptr<AbstractFramebuffer> VKGfx::CreateFramebuffer(AbstractTexture* color_attachment,
AbstractTexture* depth_attachment)
{
return VKFramebuffer::Create(static_cast<VKTexture*>(color_attachment),
static_cast<VKTexture*>(depth_attachment));
}
void VKGfx::SetPipeline(const AbstractPipeline* pipeline)
{
StateTracker::GetInstance()->SetPipeline(static_cast<const VKPipeline*>(pipeline));
}
void VKGfx::ClearRegion(const MathUtil::Rectangle<int>& rc,
const MathUtil::Rectangle<int>& target_rc, bool color_enable,
bool alpha_enable, bool z_enable, u32 color, u32 z)
{
VkRect2D target_vk_rc = {
{target_rc.left, target_rc.top},
{static_cast<uint32_t>(target_rc.GetWidth()), static_cast<uint32_t>(target_rc.GetHeight())}};
// Convert RGBA8 -> floating-point values.
VkClearValue clear_color_value = {};
VkClearValue clear_depth_value = {};
clear_color_value.color.float32[0] = static_cast<float>((color >> 16) & 0xFF) / 255.0f;
clear_color_value.color.float32[1] = static_cast<float>((color >> 8) & 0xFF) / 255.0f;
clear_color_value.color.float32[2] = static_cast<float>((color >> 0) & 0xFF) / 255.0f;
clear_color_value.color.float32[3] = static_cast<float>((color >> 24) & 0xFF) / 255.0f;
clear_depth_value.depthStencil.depth = static_cast<float>(z & 0xFFFFFF) / 16777216.0f;
if (!g_ActiveConfig.backend_info.bSupportsReversedDepthRange)
clear_depth_value.depthStencil.depth = 1.0f - clear_depth_value.depthStencil.depth;
// If we're not in a render pass (start of the frame), we can use a clear render pass
// to discard the data, rather than loading and then clearing.
bool use_clear_attachments = (color_enable && alpha_enable) || z_enable;
bool use_clear_render_pass =
!StateTracker::GetInstance()->InRenderPass() && color_enable && alpha_enable && z_enable;
// The NVIDIA Vulkan driver causes the GPU to lock up, or throw exceptions if MSAA is enabled,
// a non-full clear rect is specified, and a clear loadop or vkCmdClearAttachments is used.
if (g_ActiveConfig.iMultisamples > 1 &&
DriverDetails::HasBug(DriverDetails::BUG_BROKEN_MSAA_CLEAR))
{
use_clear_render_pass = false;
use_clear_attachments = false;
}
// This path cannot be used if the driver implementation doesn't guarantee pixels with no drawn
// geometry in "this" renderpass won't be cleared
if (DriverDetails::HasBug(DriverDetails::BUG_BROKEN_CLEAR_LOADOP_RENDERPASS))
use_clear_render_pass = false;
// Fastest path: Use a render pass to clear the buffers.
if (use_clear_render_pass)
{
const std::array<VkClearValue, 2> clear_values = {{clear_color_value, clear_depth_value}};
StateTracker::GetInstance()->BeginClearRenderPass(target_vk_rc, clear_values.data(),
static_cast<u32>(clear_values.size()));
return;
}
// Fast path: Use vkCmdClearAttachments to clear the buffers within a render path
// We can't use this when preserving alpha but clearing color.
if (use_clear_attachments)
{
VkClearAttachment clear_attachments[2];
uint32_t num_clear_attachments = 0;
if (color_enable && alpha_enable)
{
clear_attachments[num_clear_attachments].aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
clear_attachments[num_clear_attachments].colorAttachment = 0;
clear_attachments[num_clear_attachments].clearValue = clear_color_value;
num_clear_attachments++;
color_enable = false;
alpha_enable = false;
}
if (z_enable)
{
clear_attachments[num_clear_attachments].aspectMask = VK_IMAGE_ASPECT_DEPTH_BIT;
clear_attachments[num_clear_attachments].colorAttachment = 0;
clear_attachments[num_clear_attachments].clearValue = clear_depth_value;
num_clear_attachments++;
z_enable = false;
}
if (num_clear_attachments > 0)
{
VkClearRect vk_rect = {target_vk_rc, 0, g_framebuffer_manager->GetEFBLayers()};
if (!StateTracker::GetInstance()->IsWithinRenderArea(
target_vk_rc.offset.x, target_vk_rc.offset.y, target_vk_rc.extent.width,
target_vk_rc.extent.height))
{
StateTracker::GetInstance()->EndClearRenderPass();
}
StateTracker::GetInstance()->BeginRenderPass();
vkCmdClearAttachments(g_command_buffer_mgr->GetCurrentCommandBuffer(), num_clear_attachments,
clear_attachments, 1, &vk_rect);
}
}
// Anything left over for the slow path?
if (!color_enable && !alpha_enable && !z_enable)
return;
g_framebuffer_manager->ClearEFB(rc, color_enable, alpha_enable, z_enable, color, z);
}
void VKGfx::Flush()
{
ExecuteCommandBuffer(true, false);
}
void VKGfx::WaitForGPUIdle()
{
ExecuteCommandBuffer(false, true);
}
void VKGfx::BindBackbuffer(const ClearColor& clear_color)
{
StateTracker::GetInstance()->EndRenderPass();
if (!g_command_buffer_mgr->CheckLastPresentDone())
g_command_buffer_mgr->WaitForWorkerThreadIdle();
// Handle host window resizes.
CheckForSurfaceChange();
CheckForSurfaceResize();
// Check for exclusive fullscreen request.
if (m_swap_chain->GetCurrentFullscreenState() != m_swap_chain->GetNextFullscreenState() &&
!m_swap_chain->SetFullscreenState(m_swap_chain->GetNextFullscreenState()))
{
// if it fails, don't keep trying
m_swap_chain->SetNextFullscreenState(m_swap_chain->GetCurrentFullscreenState());
}
const bool present_fail = g_command_buffer_mgr->CheckLastPresentFail();
VkResult res = present_fail ? g_command_buffer_mgr->GetLastPresentResult() :
m_swap_chain->AcquireNextImage();
if (res == VK_ERROR_FULL_SCREEN_EXCLUSIVE_MODE_LOST_EXT &&
!m_swap_chain->GetCurrentFullscreenState())
{
// AMD's binary driver as of 21.3 seems to return exclusive fullscreen lost even when it was
// never requested, so long as the caller requested it to be application controlled. Handle
// this ignoring the lost result and just continuing as normal if we never acquired it.
res = VK_SUCCESS;
if (present_fail)
{
// We still need to acquire an image.
res = m_swap_chain->AcquireNextImage();
}
}
if (res != VK_SUCCESS)
{
// Execute cmdbuffer before resizing, as the last frame could still be presenting.
ExecuteCommandBuffer(false, true);
// Was this a lost exclusive fullscreen?
if (res == VK_ERROR_FULL_SCREEN_EXCLUSIVE_MODE_LOST_EXT)
{
// The present keeps returning exclusive mode lost unless we re-create the swap chain.
INFO_LOG_FMT(VIDEO, "Lost exclusive fullscreen.");
m_swap_chain->RecreateSwapChain();
}
else if (res == VK_SUBOPTIMAL_KHR || res == VK_ERROR_OUT_OF_DATE_KHR)
{
INFO_LOG_FMT(VIDEO, "Resizing swap chain due to suboptimal/out-of-date");
m_swap_chain->ResizeSwapChain();
}
else
{
ERROR_LOG_FMT(VIDEO, "Unknown present error {:#010X} {}, please report.",
static_cast<u32>(res), VkResultToString(res));
m_swap_chain->RecreateSwapChain();
}
res = m_swap_chain->AcquireNextImage();
if (res != VK_SUCCESS)
PanicAlertFmt("Failed to grab image from swap chain: {:#010X} {}", static_cast<u32>(res),
VkResultToString(res));
}
// Transition from undefined (or present src, but it can be substituted) to
// color attachment ready for writing. These transitions must occur outside
// a render pass, unless the render pass declares a self-dependency.
m_swap_chain->GetCurrentTexture()->OverrideImageLayout(VK_IMAGE_LAYOUT_UNDEFINED);
m_swap_chain->GetCurrentTexture()->TransitionToLayout(
g_command_buffer_mgr->GetCurrentCommandBuffer(), VK_IMAGE_LAYOUT_COLOR_ATTACHMENT_OPTIMAL);
SetAndClearFramebuffer(m_swap_chain->GetCurrentFramebuffer(),
ClearColor{{0.0f, 0.0f, 0.0f, 1.0f}});
}
void VKGfx::PresentBackbuffer()
{
// End drawing to backbuffer
StateTracker::GetInstance()->EndRenderPass();
// Transition the backbuffer to PRESENT_SRC to ensure all commands drawing
// to it have finished before present.
m_swap_chain->GetCurrentTexture()->TransitionToLayout(
g_command_buffer_mgr->GetCurrentCommandBuffer(), VK_IMAGE_LAYOUT_PRESENT_SRC_KHR);
// Submit the current command buffer, signaling rendering finished semaphore when it's done
// Because this final command buffer is rendering to the swap chain, we need to wait for
// the available semaphore to be signaled before executing the buffer. This final submission
// can happen off-thread in the background while we're preparing the next frame.
g_command_buffer_mgr->SubmitCommandBuffer(true, false, m_swap_chain->GetSwapChain(),
m_swap_chain->GetCurrentImageIndex());
// New cmdbuffer, so invalidate state.
StateTracker::GetInstance()->InvalidateCachedState();
}
void VKGfx::SetFullscreen(bool enable_fullscreen)
{
if (!m_swap_chain->IsFullscreenSupported())
return;
m_swap_chain->SetNextFullscreenState(enable_fullscreen);
}
bool VKGfx::IsFullscreen() const
{
return m_swap_chain && m_swap_chain->GetCurrentFullscreenState();
}
void VKGfx::ExecuteCommandBuffer(bool submit_off_thread, bool wait_for_completion)
{
StateTracker::GetInstance()->EndRenderPass();
g_command_buffer_mgr->SubmitCommandBuffer(submit_off_thread, wait_for_completion);
StateTracker::GetInstance()->InvalidateCachedState();
}
void VKGfx::CheckForSurfaceChange()
{
if (!g_presenter->SurfaceChangedTestAndClear() || !m_swap_chain)
return;
// Submit the current draws up until rendering the XFB.
ExecuteCommandBuffer(false, true);
// Clear the present failed flag, since we don't want to resize after recreating.
g_command_buffer_mgr->CheckLastPresentFail();
// Recreate the surface. If this fails we're in trouble.
if (!m_swap_chain->RecreateSurface(g_presenter->GetNewSurfaceHandle()))
PanicAlertFmt("Failed to recreate Vulkan surface. Cannot continue.");
// Handle case where the dimensions are now different.
OnSwapChainResized();
}
void VKGfx::CheckForSurfaceResize()
{
if (!g_presenter->SurfaceResizedTestAndClear())
return;
// If we don't have a surface, how can we resize the swap chain?
// CheckForSurfaceChange should handle this case.
if (!m_swap_chain)
{
WARN_LOG_FMT(VIDEO, "Surface resize event received without active surface, ignoring");
return;
}
// Wait for the GPU to catch up since we're going to destroy the swap chain.
ExecuteCommandBuffer(false, true);
// Clear the present failed flag, since we don't want to resize after recreating.
g_command_buffer_mgr->CheckLastPresentFail();
// Resize the swap chain.
m_swap_chain->RecreateSwapChain();
OnSwapChainResized();
}
void VKGfx::OnConfigChanged(u32 bits)
{
if (bits & CONFIG_CHANGE_BIT_HOST_CONFIG)
g_object_cache->ReloadPipelineCache();
// For vsync, we need to change the present mode, which means recreating the swap chain.
if (m_swap_chain && bits & CONFIG_CHANGE_BIT_VSYNC)
{
ExecuteCommandBuffer(false, true);
m_swap_chain->SetVSync(g_ActiveConfig.bVSyncActive);
}
// For quad-buffered stereo we need to change the layer count, so recreate the swap chain.
if (m_swap_chain && bits & CONFIG_CHANGE_BIT_STEREO_MODE)
{
ExecuteCommandBuffer(false, true);
m_swap_chain->RecreateSwapChain();
}
// Wipe sampler cache if force texture filtering or anisotropy changes.
if (bits & (CONFIG_CHANGE_BIT_ANISOTROPY | CONFIG_CHANGE_BIT_FORCE_TEXTURE_FILTERING))
{
ExecuteCommandBuffer(false, true);
ResetSamplerStates();
}
}
void VKGfx::OnSwapChainResized()
{
g_presenter->SetBackbuffer(m_swap_chain->GetWidth(), m_swap_chain->GetHeight());
}
void VKGfx::BindFramebuffer(VKFramebuffer* fb)
{
StateTracker::GetInstance()->EndRenderPass();
// Shouldn't be bound as a texture.
if (fb->GetColorAttachment())
{
StateTracker::GetInstance()->UnbindTexture(
static_cast<VKTexture*>(fb->GetColorAttachment())->GetView());
}
if (fb->GetDepthAttachment())
{
StateTracker::GetInstance()->UnbindTexture(
static_cast<VKTexture*>(fb->GetDepthAttachment())->GetView());
}
fb->TransitionForRender();
StateTracker::GetInstance()->SetFramebuffer(fb);
m_current_framebuffer = fb;
}
void VKGfx::SetFramebuffer(AbstractFramebuffer* framebuffer)
{
if (m_current_framebuffer == framebuffer)
return;
VKFramebuffer* vkfb = static_cast<VKFramebuffer*>(framebuffer);
BindFramebuffer(vkfb);
}
void VKGfx::SetAndDiscardFramebuffer(AbstractFramebuffer* framebuffer)
{
if (m_current_framebuffer == framebuffer)
return;
VKFramebuffer* vkfb = static_cast<VKFramebuffer*>(framebuffer);
BindFramebuffer(vkfb);
// If we're discarding, begin the discard pass, then switch to a load pass.
// This way if the command buffer is flushed, we don't start another discard pass.
StateTracker::GetInstance()->BeginDiscardRenderPass();
}
void VKGfx::SetAndClearFramebuffer(AbstractFramebuffer* framebuffer, const ClearColor& color_value,
float depth_value)
{
VKFramebuffer* vkfb = static_cast<VKFramebuffer*>(framebuffer);
BindFramebuffer(vkfb);
std::array<VkClearValue, 2> clear_values;
u32 num_clear_values = 0;
if (vkfb->GetColorFormat() != AbstractTextureFormat::Undefined)
{
std::memcpy(clear_values[num_clear_values].color.float32, color_value.data(),
sizeof(clear_values[num_clear_values].color.float32));
num_clear_values++;
}
if (vkfb->GetDepthFormat() != AbstractTextureFormat::Undefined)
{
clear_values[num_clear_values].depthStencil.depth = depth_value;
clear_values[num_clear_values].depthStencil.stencil = 0;
num_clear_values++;
}
StateTracker::GetInstance()->BeginClearRenderPass(vkfb->GetRect(), clear_values.data(),
num_clear_values);
}
void VKGfx::SetTexture(u32 index, const AbstractTexture* texture)
{
// Texture should always be in SHADER_READ_ONLY layout prior to use.
// This is so we don't need to transition during render passes.
const VKTexture* tex = static_cast<const VKTexture*>(texture);
if (tex)
{
if (tex->GetLayout() != VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL)
{
if (StateTracker::GetInstance()->InRenderPass())
{
WARN_LOG_FMT(VIDEO, "Transitioning image in render pass in VKGfx::SetTexture()");
StateTracker::GetInstance()->EndRenderPass();
}
tex->TransitionToLayout(g_command_buffer_mgr->GetCurrentCommandBuffer(),
VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL);
}
StateTracker::GetInstance()->SetTexture(index, tex->GetView());
}
else
{
StateTracker::GetInstance()->SetTexture(0, VK_NULL_HANDLE);
}
}
void VKGfx::SetSamplerState(u32 index, const SamplerState& state)
{
// Skip lookup if the state hasn't changed.
if (m_sampler_states[index] == state)
return;
// Look up new state and replace in state tracker.
VkSampler sampler = g_object_cache->GetSampler(state);
if (sampler == VK_NULL_HANDLE)
{
ERROR_LOG_FMT(VIDEO, "Failed to create sampler");
sampler = g_object_cache->GetPointSampler();
}
StateTracker::GetInstance()->SetSampler(index, sampler);
m_sampler_states[index] = state;
}
void VKGfx::SetComputeImageTexture(AbstractTexture* texture, bool read, bool write)
{
VKTexture* vk_texture = static_cast<VKTexture*>(texture);
if (vk_texture)
{
StateTracker::GetInstance()->EndRenderPass();
StateTracker::GetInstance()->SetImageTexture(vk_texture->GetView());
vk_texture->TransitionToLayout(g_command_buffer_mgr->GetCurrentCommandBuffer(),
read ? (write ? VKTexture::ComputeImageLayout::ReadWrite :
VKTexture::ComputeImageLayout::ReadOnly) :
VKTexture::ComputeImageLayout::WriteOnly);
}
else
{
StateTracker::GetInstance()->SetImageTexture(VK_NULL_HANDLE);
}
}
void VKGfx::UnbindTexture(const AbstractTexture* texture)
{
StateTracker::GetInstance()->UnbindTexture(static_cast<const VKTexture*>(texture)->GetView());
}
void VKGfx::ResetSamplerStates()
{
// Invalidate all sampler states, next draw will re-initialize them.
for (u32 i = 0; i < m_sampler_states.size(); i++)
{
m_sampler_states[i] = RenderState::GetPointSamplerState();
StateTracker::GetInstance()->SetSampler(i, g_object_cache->GetPointSampler());
}
// Invalidate all sampler objects (some will be unused now).
g_object_cache->ClearSamplerCache();
}
void VKGfx::SetScissorRect(const MathUtil::Rectangle<int>& rc)
{
VkRect2D scissor = {{rc.left, rc.top},
{static_cast<u32>(rc.GetWidth()), static_cast<u32>(rc.GetHeight())}};
// See Vulkan spec for vkCmdSetScissor:
// The x and y members of offset must be greater than or equal to 0.
if (scissor.offset.x < 0)
{
scissor.extent.width -= -scissor.offset.x;
scissor.offset.x = 0;
}
if (scissor.offset.y < 0)
{
scissor.extent.height -= -scissor.offset.y;
scissor.offset.y = 0;
}
StateTracker::GetInstance()->SetScissor(scissor);
}
void VKGfx::SetViewport(float x, float y, float width, float height, float near_depth,
float far_depth)
{
VkViewport viewport = {x, y, width, height, near_depth, far_depth};
StateTracker::GetInstance()->SetViewport(viewport);
}
void VKGfx::Draw(u32 base_vertex, u32 num_vertices)
{
if (!StateTracker::GetInstance()->Bind())
return;
vkCmdDraw(g_command_buffer_mgr->GetCurrentCommandBuffer(), num_vertices, 1, base_vertex, 0);
}
void VKGfx::DrawIndexed(u32 base_index, u32 num_indices, u32 base_vertex)
{
if (!StateTracker::GetInstance()->Bind())
return;
vkCmdDrawIndexed(g_command_buffer_mgr->GetCurrentCommandBuffer(), num_indices, 1, base_index,
base_vertex, 0);
}
void VKGfx::DispatchComputeShader(const AbstractShader* shader, u32 groupsize_x, u32 groupsize_y,
u32 groupsize_z, u32 groups_x, u32 groups_y, u32 groups_z)
{
StateTracker::GetInstance()->SetComputeShader(static_cast<const VKShader*>(shader));
if (StateTracker::GetInstance()->BindCompute())
vkCmdDispatch(g_command_buffer_mgr->GetCurrentCommandBuffer(), groups_x, groups_y, groups_z);
}
SurfaceInfo VKGfx::GetSurfaceInfo() const
{
return {m_swap_chain ? m_swap_chain->GetWidth() : 1u,
m_swap_chain ? m_swap_chain->GetHeight() : 0u, m_backbuffer_scale,
m_swap_chain ? m_swap_chain->GetTextureFormat() : AbstractTextureFormat::Undefined};
}
} // namespace Vulkan