Start using GrGpuCommandBuffer in GrDrawTarget.

src/gpu/vk/GrVkGpuCommandBuffer.cpp

Index: src/gpu/vk/GrVkGpuCommandBuffer.cpp
diff --git a/src/gpu/vk/GrVkGpuCommandBuffer.cpp b/src/gpu/vk/GrVkGpuCommandBuffer.cpp
index f6104a33083693f28cca8a91c17b47bddaa260a7..1ec82778c42733848a2c889511c04b9eba47f890 100644
--- a/src/gpu/vk/GrVkGpuCommandBuffer.cpp
+++ b/src/gpu/vk/GrVkGpuCommandBuffer.cpp
@@ -7,11 +7,18 @@
 
 #include "GrVkGpuCommandBuffer.h"
 
+#include "GrMesh.h"
+#include "GrPipeline.h"
+#include "GrRenderTargetPriv.h"
+#include "GrTextureAccess.h"
+#include "GrTexturePriv.h"
 #include "GrVkCommandBuffer.h"
 #include "GrVkGpu.h"
+#include "GrVkPipeline.h"
 #include "GrVkRenderPass.h"
 #include "GrVkRenderTarget.h"
 #include "GrVkResourceProvider.h"
+#include "GrVkTexture.h"
 
 void get_vk_load_store_ops(GrGpuCommandBuffer::LoadAndStoreOp op,
                            VkAttachmentLoadOp* loadOp, VkAttachmentStoreOp* storeOp) {
@@ -49,10 +56,11 @@ void get_vk_load_store_ops(GrGpuCommandBuffer::LoadAndStoreOp op,
 }
 
 GrVkGpuCommandBuffer::GrVkGpuCommandBuffer(GrVkGpu* gpu,
-                                           const GrVkRenderTarget& target,
+                                           GrVkRenderTarget* target,
                                            LoadAndStoreOp colorOp, GrColor colorClear,
                                            LoadAndStoreOp stencilOp, GrColor stencilClear)
-    : fGpu(gpu) {
+    : fGpu(gpu)
+    , fRenderTarget(target) {

[jvanverth1, 2016/06/17 15:25:16]

Why a non-const pointer rather than a const reference?

[egdaniel, 2016/06/17 15:57:35]

setImageLayout call on the RT is non const

[jvanverth1, 2016/06/17 16:49:13]

Acknowledged.

     VkAttachmentLoadOp vkLoadOp;
     VkAttachmentStoreOp vkStoreOp;
 
@@ -65,21 +73,21 @@ GrVkGpuCommandBuffer::GrVkGpuCommandBuffer(GrVkGpu* gpu,
     GrVkRenderPass::LoadStoreOps vkResolveOps(VK_ATTACHMENT_LOAD_OP_LOAD,
                                               VK_ATTACHMENT_STORE_OP_STORE);
 
-    const GrVkResourceProvider::CompatibleRPHandle& rpHandle = target.compatibleRenderPassHandle();
+    const GrVkResourceProvider::CompatibleRPHandle& rpHandle = target->compatibleRenderPassHandle();
     if (rpHandle.isValid()) {
         fRenderPass = fGpu->resourceProvider().findRenderPass(rpHandle,
                                                               vkColorOps,
                                                               vkResolveOps,
                                                               vkStencilOps);
     } else {
-        fRenderPass = fGpu->resourceProvider().findRenderPass(target,
+        fRenderPass = fGpu->resourceProvider().findRenderPass(*target,
                                                               vkColorOps,
                                                               vkResolveOps,
                                                               vkStencilOps);
     }
 
     fCommandBuffer = GrVkSecondaryCommandBuffer::Create(gpu, gpu->cmdPool(), fRenderPass);
-    fCommandBuffer->begin(gpu, target.framebuffer());
+    fCommandBuffer->begin(gpu, target->framebuffer());
 }
 
 GrVkGpuCommandBuffer::~GrVkGpuCommandBuffer() {
@@ -87,11 +95,322 @@ GrVkGpuCommandBuffer::~GrVkGpuCommandBuffer() {
     fRenderPass->unref(fGpu);
 }
 
+GrGpu* GrVkGpuCommandBuffer::gpu() { return fGpu; }
+
 void GrVkGpuCommandBuffer::end() {
     fCommandBuffer->end(fGpu);
 }
 
-void GrVkGpuCommandBuffer::submit() {
-    fGpu->submitSecondaryCommandBuffer(fCommandBuffer);
+void GrVkGpuCommandBuffer::onSubmit(const SkIRect& bounds) {
+    // Change layout of our render target so it can be used as the color attachment
+    fRenderTarget->setImageLayout(fGpu,
+                                  VK_IMAGE_LAYOUT_COLOR_ATTACHMENT_OPTIMAL,
+                                  VK_ACCESS_COLOR_ATTACHMENT_WRITE_BIT,
+                                  VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT,
+                                  false);
+
+    // If we are using a stencil attachment we also need to update its layout
+    if (GrStencilAttachment* stencil = fRenderTarget->renderTargetPriv().getStencilAttachment()) {
+        GrVkStencilAttachment* vkStencil = (GrVkStencilAttachment*)stencil;
+        vkStencil->setImageLayout(fGpu,
+                                  VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL,
+                                  VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT |
+                                  VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_READ_BIT,
+                                  VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT,
+                                  false);
+    }
+
+    for (int i = 0; i < fSampledImages.count(); ++i) {
+        fSampledImages[i]->setImageLayout(fGpu,
+                                          VK_IMAGE_LAYOUT_SHADER_READ_ONLY_OPTIMAL,
+                                          VK_ACCESS_SHADER_READ_BIT,
+                                          VK_PIPELINE_STAGE_ALL_GRAPHICS_BIT,
+                                          false);
+    }
+
+    fGpu->submitSecondaryCommandBuffer(fCommandBuffer, fRenderPass, fRenderTarget, bounds);
+}
+
+void GrVkGpuCommandBuffer::onClearStencilClip(GrRenderTarget* target,
+                                              const SkIRect& rect,
+                                              bool insideClip) {
+    SkASSERT(target);
+
+    GrVkRenderTarget* vkRT = static_cast<GrVkRenderTarget*>(target);
+    GrStencilAttachment* sb = target->renderTargetPriv().getStencilAttachment();
+    // this should only be called internally when we know we have a
+    // stencil buffer.
+    SkASSERT(sb);
+    int stencilBitCount = sb->bits();
+
+    // The contract with the callers does not guarantee that we preserve all bits in the stencil
+    // during this clear. Thus we will clear the entire stencil to the desired value.
+
+    VkClearDepthStencilValue vkStencilColor;
+    memset(&vkStencilColor, 0, sizeof(VkClearDepthStencilValue));
+    if (insideClip) {
+        vkStencilColor.stencil = (1 << (stencilBitCount - 1));
+    } else {
+        vkStencilColor.stencil = 0;
+    }
+
+    VkClearRect clearRect;
+    // Flip rect if necessary
+    SkIRect vkRect = rect;
+
+    if (kBottomLeft_GrSurfaceOrigin == vkRT->origin()) {
+        vkRect.fTop = vkRT->height() - rect.fBottom;
+        vkRect.fBottom = vkRT->height() - rect.fTop;
+    }
+
+    clearRect.rect.offset = { vkRect.fLeft, vkRect.fTop };
+    clearRect.rect.extent = { (uint32_t)vkRect.width(), (uint32_t)vkRect.height() };
+
+    clearRect.baseArrayLayer = 0;
+    clearRect.layerCount = 1;
+
+    uint32_t stencilIndex;
+    SkAssertResult(fRenderPass->stencilAttachmentIndex(&stencilIndex));
+
+    VkClearAttachment attachment;
+    attachment.aspectMask = VK_IMAGE_ASPECT_STENCIL_BIT;
+    attachment.colorAttachment = 0; // this value shouldn't matter
+    attachment.clearValue.depthStencil = vkStencilColor;
+
+    fCommandBuffer->clearAttachments(fGpu, 1, &attachment, 1, &clearRect);
+}
+
+void GrVkGpuCommandBuffer::onClear(GrRenderTarget* target, const SkIRect& rect, GrColor color) {
+    // Currently we should not see clears in vulkan since we are converting them all to draws.
+    // We do this since some clears currently can come in the must happen outside a render pass

[jvanverth1, 2016/06/17 15:25:16]

Nit: that must happen

Also: seems like this might be a temporary perf hit, as a clear on a tiled
renderer is practically free.

[egdaniel, 2016/06/17 15:57:35]

Yes I believe this is a perf hit (though it hopefully will at least be grouped
up in a render pass with other draws to the same RT). Once we get clears set up
on the load/stores I expect to see a perf gain.

[jvanverth1, 2016/06/17 16:49:13]

Acknowledged.

[egdaniel, 2016/06/17 18:49:14]

Done.

+    // and we assume all commands in this buffer are inside a renderpass.
+    SkASSERT(false);
+#if 0
+    // parent class should never let us get here with no RT
+    SkASSERT(target);
+
+    VkClearColorValue vkColor;
+    GrColorToRGBAFloat(color, vkColor.float32);
+
+    GrVkRenderTarget* vkRT = static_cast<GrVkRenderTarget*>(target);
+
+    if (rect.width() != target->width() || rect.height() != target->height()) {
+        vkRT->setImageLayout(fGpu,
+                             VK_IMAGE_LAYOUT_COLOR_ATTACHMENT_OPTIMAL,
+                             VK_ACCESS_COLOR_ATTACHMENT_WRITE_BIT,
+                             VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT,
+                             false);
+
+        // If we are using a stencil attachment we also need to change its layout to what the render
+        // pass is expecting.
+        if (GrStencilAttachment* stencil = vkRT->renderTargetPriv().getStencilAttachment()) {
+            GrVkStencilAttachment* vkStencil = (GrVkStencilAttachment*)stencil;
+            vkStencil->setImageLayout(fGpu,
+                                      VK_IMAGE_LAYOUT_DEPTH_STENCIL_ATTACHMENT_OPTIMAL,
+                                      VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_WRITE_BIT |
+                                      VK_ACCESS_DEPTH_STENCIL_ATTACHMENT_READ_BIT,
+                                      VK_PIPELINE_STAGE_TOP_OF_PIPE_BIT,
+                                      false);
+        }
+
+        VkClearRect clearRect;
+        // Flip rect if necessary
+        SkIRect vkRect = rect;
+        if (kBottomLeft_GrSurfaceOrigin == vkRT->origin()) {
+            vkRect.fTop = vkRT->height() - rect.fBottom;
+            vkRect.fBottom = vkRT->height() - rect.fTop;
+        }
+        clearRect.rect.offset = { vkRect.fLeft, vkRect.fTop };
+        clearRect.rect.extent = { (uint32_t)vkRect.width(), (uint32_t)vkRect.height() };
+        clearRect.baseArrayLayer = 0;
+        clearRect.layerCount = 1;
+
+        const GrVkRenderPass* renderPass = vkRT->simpleRenderPass();
+        SkASSERT(renderPass);
+        fCommandBuffer->beginRenderPass(fGpu, renderPass, *vkRT);
+
+        uint32_t colorIndex;
+        SkAssertResult(renderPass->colorAttachmentIndex(&colorIndex));
+
+        VkClearAttachment attachment;
+        attachment.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
+        attachment.colorAttachment = colorIndex;
+        attachment.clearValue.color = vkColor;
+
+        fCurrentCmdBuffer->clearAttachments(fGpu, 1, &attachment, 1, &clearRect);
+        fCurrentCmdBuffer->endRenderPass(fGpu);
+        return;
+    }
+
+    vkRT->setImageLayout(this,
+                         VK_IMAGE_LAYOUT_TRANSFER_DST_OPTIMAL,
+                         VK_ACCESS_TRANSFER_WRITE_BIT,
+                         VK_PIPELINE_STAGE_TRANSFER_BIT,
+                         false);
+
+    VkImageSubresourceRange subRange;
+    memset(&subRange, 0, sizeof(VkImageSubresourceRange));
+    subRange.aspectMask = VK_IMAGE_ASPECT_COLOR_BIT;
+    subRange.baseMipLevel = 0;
+    subRange.levelCount = 1;
+    subRange.baseArrayLayer = 0;
+    subRange.layerCount = 1;
+
+    // In the future we may not actually be doing this type of clear at all. If we are inside a
+    // render pass or doing a non full clear then we will use CmdClearColorAttachment. The more
+    // common use case will be clearing an attachment at the start of a render pass, in which case
+    // we will use the clear load ops.
+    fCommandBuffer->clearColorImage(this,
+                                    vkRT,
+                                    &vkColor,
+                                   1, &subRange);
+#endif
+}
+
+////////////////////////////////////////////////////////////////////////////////
+
+void GrVkGpuCommandBuffer::bindGeometry(const GrPrimitiveProcessor& primProc,
+                                        const GrNonInstancedMesh& mesh) {
+    // There is no need to put any memory barriers to make sure host writes have finished here.
+    // When a command buffer is submitted to a queue, there is an implicit memory barrier that
+    // occurs for all host writes. Additionally, BufferMemoryBarriers are not allowed inside of
+    // an active RenderPass.
+    GrVkVertexBuffer* vbuf;
+    vbuf = (GrVkVertexBuffer*)mesh.vertexBuffer();
+    SkASSERT(vbuf);
+    SkASSERT(!vbuf->isMapped());
+
+    fCommandBuffer->bindVertexBuffer(fGpu, vbuf);
+
+    if (mesh.isIndexed()) {
+        GrVkIndexBuffer* ibuf = (GrVkIndexBuffer*)mesh.indexBuffer();
+        SkASSERT(ibuf);
+        SkASSERT(!ibuf->isMapped());
+
+        fCommandBuffer->bindIndexBuffer(fGpu, ibuf);
+    }
+}
+
+sk_sp<GrVkPipelineState> GrVkGpuCommandBuffer::prepareDrawState(
+                                                               const GrPipeline& pipeline,
+                                                               const GrPrimitiveProcessor& primProc,
+                                                               GrPrimitiveType primitiveType,
+                                                               const GrVkRenderPass& renderPass) {
+    sk_sp<GrVkPipelineState> pipelineState =
+        fGpu->resourceProvider().findOrCreateCompatiblePipelineState(pipeline,
+                                                                     primProc,
+                                                                     primitiveType,
+                                                                     renderPass);
+    if (!pipelineState) {
+        return pipelineState;
+    }
+
+    pipelineState->setData(fGpu, primProc, pipeline);
+
+    pipelineState->bind(fGpu, fCommandBuffer);
+
+    GrVkPipeline::SetDynamicState(fGpu, fCommandBuffer, pipeline);
+
+    return pipelineState;
+}
+
+static void append_sampled_images(const GrProcessor& processor,
+                                  const GrVkGpu* gpu,
+                                  SkTArray<GrVkImage*>* sampledImages) {
+    if (int numTextures = processor.numTextures()) {
+        GrVkImage** images = sampledImages->push_back_n(numTextures);
+        int i = 0;
+        do {
+            const GrTextureAccess& texAccess = processor.textureAccess(i);
+            GrVkTexture* vkTexture = static_cast<GrVkTexture*>(processor.texture(i));
+            SkASSERT(vkTexture);
+            const GrTextureParams& params = texAccess.getParams();
+            // Check if we need to regenerate any mip maps
+            if (GrTextureParams::kMipMap_FilterMode == params.filterMode()) {
+                if (vkTexture->texturePriv().mipMapsAreDirty()) {
+                    gpu->generateMipmap(vkTexture);
+                    vkTexture->texturePriv().dirtyMipMaps(false);
+                }
+            }
+
+            images[i] = vkTexture;
+        } while (++i < numTextures);
+
+    }
+}
+
+void GrVkGpuCommandBuffer::onDraw(const GrPipeline& pipeline,
+                                  const GrPrimitiveProcessor& primProc,
+                                  const GrMesh* meshes,
+                                  int meshCount) {
+    if (!meshCount) {
+        return;
+    }
+    GrRenderTarget* rt = pipeline.getRenderTarget();
+    GrVkRenderTarget* vkRT = static_cast<GrVkRenderTarget*>(rt);
+    const GrVkRenderPass* renderPass = vkRT->simpleRenderPass();
+    SkASSERT(renderPass);
+
+    GrPrimitiveType primitiveType = meshes[0].primitiveType();
+    sk_sp<GrVkPipelineState> pipelineState = this->prepareDrawState(pipeline,
+                                                                    primProc,
+                                                                    primitiveType,
+                                                                    *renderPass);
+    if (!pipelineState) {
+        return;
+    }
+
+    append_sampled_images(primProc, fGpu, &fSampledImages);
+    for (int i = 0; i < pipeline.numFragmentProcessors(); ++i) {
+        append_sampled_images(pipeline.getFragmentProcessor(i), fGpu, &fSampledImages);
+    }
+    append_sampled_images(pipeline.getXferProcessor(), fGpu, &fSampledImages);
+
+    for (int i = 0; i < meshCount; ++i) {
+        const GrMesh& mesh = meshes[i];
+        GrMesh::Iterator iter;
+        const GrNonInstancedMesh* nonIdxMesh = iter.init(mesh);
+        do {
+            if (nonIdxMesh->primitiveType() != primitiveType) {
+                // Technically we don't have to call this here (since there is a safety check in
+                // pipelineState:setData but this will allow for quicker freeing of resources if the
+                // pipelineState sits in a cache for a while.
+                pipelineState->freeTempResources(fGpu);
+                SkDEBUGCODE(pipelineState = nullptr);
+                primitiveType = nonIdxMesh->primitiveType();
+                pipelineState = this->prepareDrawState(pipeline,
+                                                       primProc,
+                                                       primitiveType,
+                                                       *renderPass);
+                if (!pipelineState) {
+                    return;
+                }
+            }
+            SkASSERT(pipelineState);
+            this->bindGeometry(primProc, *nonIdxMesh);
+
+            if (nonIdxMesh->isIndexed()) {
+                fCommandBuffer->drawIndexed(fGpu,
+                                            nonIdxMesh->indexCount(),
+                                            1,
+                                            nonIdxMesh->startIndex(),
+                                            nonIdxMesh->startVertex(),
+                                            0);
+            } else {
+                fCommandBuffer->draw(fGpu,
+                                     nonIdxMesh->vertexCount(),
+                                     1,
+                                     nonIdxMesh->startVertex(),
+                                     0);
+            }
+
+            fGpu->stats()->incNumDraws();
+        } while ((nonIdxMesh = iter.next()));
+    }
+
+    // Technically we don't have to call this here (since there is a safety check in
+    // pipelineState:setData but this will allow for quicker freeing of resources if the
+    // pipelineState sits in a cache for a while.
+    pipelineState->freeTempResources(fGpu);
 }