93f1633abc
Platforms that "useDrawInsteadOfClear" take a different path in GrRenderTargetContext::internalClear. The different path involves a lot of comparisons of the drawn rect with the bounds of the RenderTarget. Since the RenderTargets are now deferred (and the instantiated version might be larger than the proxied size) case must be taken in prematurely optimizing away clears and draws. GOLD_TRYBOT_URL= https://gold.skia.org/search?issue=5148 TBR=bsalomon@google.com Change-Id: If63b6393fb66328175bc92de211e7242e903f25e Reviewed-on: https://skia-review.googlesource.com/5148 Reviewed-by: Robert Phillips <robertphillips@google.com> Reviewed-by: Greg Daniel <egdaniel@google.com> Commit-Queue: Robert Phillips <robertphillips@google.com>
320 lines
10 KiB
C++
320 lines
10 KiB
C++
/*
|
|
* Copyright 2016 Google Inc.
|
|
*
|
|
* Use of this source code is governed by a BSD-style license that can be
|
|
* found in the LICENSE file.
|
|
*/
|
|
|
|
#ifndef GrSurfaceProxy_DEFINED
|
|
#define GrSurfaceProxy_DEFINED
|
|
|
|
#include "GrGpuResource.h"
|
|
#include "GrSurface.h"
|
|
|
|
#include "SkRect.h"
|
|
|
|
class GrCaps;
|
|
class GrRenderTargetOpList;
|
|
class GrRenderTargetProxy;
|
|
class GrTextureOpList;
|
|
class GrTextureProvider;
|
|
class GrTextureProxy;
|
|
|
|
// This class replicates the functionality GrIORef<GrSurface> but tracks the
|
|
// utilitization for later resource allocation (for the deferred case) and
|
|
// forwards on the utilization in the wrapped case
|
|
class GrIORefProxy : public SkNoncopyable {
|
|
public:
|
|
void ref() const {
|
|
this->validate();
|
|
|
|
++fRefCnt;
|
|
if (fTarget) {
|
|
fTarget->ref();
|
|
}
|
|
}
|
|
|
|
void unref() const {
|
|
this->validate();
|
|
|
|
if (fTarget) {
|
|
fTarget->unref();
|
|
}
|
|
|
|
if (!(--fRefCnt)) {
|
|
delete this;
|
|
return;
|
|
}
|
|
|
|
this->validate();
|
|
}
|
|
|
|
void validate() const {
|
|
#ifdef SK_DEBUG
|
|
SkASSERT(fRefCnt >= 1);
|
|
SkASSERT(fPendingReads >= 0);
|
|
SkASSERT(fPendingWrites >= 0);
|
|
SkASSERT(fRefCnt + fPendingReads + fPendingWrites >= 1);
|
|
|
|
if (fTarget) {
|
|
SkASSERT(!fPendingReads && !fPendingWrites);
|
|
// The backing GrSurface can have more refs than the proxy if the proxy
|
|
// started off wrapping an external resource (that came in with refs).
|
|
// The GrSurface should never have fewer refs than the proxy however.
|
|
SkASSERT(fTarget->fRefCnt >= fRefCnt);
|
|
}
|
|
#endif
|
|
}
|
|
|
|
int32_t getProxyRefCnt_TestOnly() const;
|
|
int32_t getBackingRefCnt_TestOnly() const;
|
|
int32_t getPendingReadCnt_TestOnly() const;
|
|
int32_t getPendingWriteCnt_TestOnly() const;
|
|
|
|
protected:
|
|
GrIORefProxy() : fTarget(nullptr), fRefCnt(1), fPendingReads(0), fPendingWrites(0) {}
|
|
GrIORefProxy(sk_sp<GrSurface> surface) : fRefCnt(1), fPendingReads(0), fPendingWrites(0) {
|
|
// Since we're manually forwarding on refs & unrefs we don't want sk_sp doing
|
|
// anything extra.
|
|
fTarget = surface.release();
|
|
}
|
|
virtual ~GrIORefProxy() {
|
|
// We don't unref 'fTarget' here since the 'unref' method will already
|
|
// have forwarded on the unref call that got use here.
|
|
}
|
|
|
|
// This GrIORefProxy was deferred before but has just been instantiated. To
|
|
// make all the reffing & unreffing work out we now need to transfer any deferred
|
|
// refs & unrefs to the new GrSurface
|
|
void transferRefs() {
|
|
SkASSERT(fTarget);
|
|
|
|
fTarget->fRefCnt += (fRefCnt-1); // don't xfer the proxy's creation ref
|
|
fTarget->fPendingReads += fPendingReads;
|
|
fTarget->fPendingWrites += fPendingWrites;
|
|
|
|
fPendingReads = 0;
|
|
fPendingWrites = 0;
|
|
}
|
|
|
|
// For deferred proxies this will be null. For wrapped proxies it will point to the
|
|
// wrapped resource.
|
|
GrSurface* fTarget;
|
|
|
|
private:
|
|
// This class is used to manage conversion of refs to pending reads/writes.
|
|
friend class GrGpuResourceRef;
|
|
template <typename, GrIOType> friend class GrPendingIOResource;
|
|
|
|
void addPendingRead() const {
|
|
this->validate();
|
|
|
|
if (fTarget) {
|
|
fTarget->addPendingRead();
|
|
return;
|
|
}
|
|
|
|
++fPendingReads;
|
|
}
|
|
|
|
void completedRead() const {
|
|
this->validate();
|
|
|
|
if (fTarget) {
|
|
fTarget->completedRead();
|
|
return;
|
|
}
|
|
|
|
SkFAIL("How was the read completed if the Proxy hasn't been instantiated?");
|
|
}
|
|
|
|
void addPendingWrite() const {
|
|
this->validate();
|
|
|
|
if (fTarget) {
|
|
fTarget->addPendingWrite();
|
|
return;
|
|
}
|
|
|
|
++fPendingWrites;
|
|
}
|
|
|
|
void completedWrite() const {
|
|
this->validate();
|
|
|
|
if (fTarget) {
|
|
fTarget->completedWrite();
|
|
return;
|
|
}
|
|
|
|
SkFAIL("How was the write completed if the Proxy hasn't been instantiated?");
|
|
}
|
|
|
|
mutable int32_t fRefCnt;
|
|
mutable int32_t fPendingReads;
|
|
mutable int32_t fPendingWrites;
|
|
};
|
|
|
|
class GrSurfaceProxy : public GrIORefProxy {
|
|
public:
|
|
static sk_sp<GrSurfaceProxy> MakeWrapped(sk_sp<GrSurface>);
|
|
|
|
static sk_sp<GrSurfaceProxy> MakeDeferred(const GrCaps&, const GrSurfaceDesc&,
|
|
SkBackingFit, SkBudgeted);
|
|
|
|
// TODO: need to refine ownership semantics of 'srcData' if we're in completely
|
|
// deferred mode
|
|
static sk_sp<GrSurfaceProxy> MakeDeferred(const GrCaps&, GrTextureProvider*,
|
|
const GrSurfaceDesc&, SkBudgeted,
|
|
const void* srcData, size_t rowBytes);
|
|
|
|
const GrSurfaceDesc& desc() const { return fDesc; }
|
|
|
|
GrSurfaceOrigin origin() const {
|
|
SkASSERT(kTopLeft_GrSurfaceOrigin == fDesc.fOrigin ||
|
|
kBottomLeft_GrSurfaceOrigin == fDesc.fOrigin);
|
|
return fDesc.fOrigin;
|
|
}
|
|
int width() const { return fDesc.fWidth; }
|
|
int height() const { return fDesc.fHeight; }
|
|
GrPixelConfig config() const { return fDesc.fConfig; }
|
|
|
|
class UniqueID {
|
|
public:
|
|
// wrapped
|
|
explicit UniqueID(const GrGpuResource::UniqueID& id) : fID(id.asUInt()) { }
|
|
// deferred
|
|
UniqueID() : fID(GrGpuResource::CreateUniqueID()) { }
|
|
|
|
uint32_t asUInt() const { return fID; }
|
|
|
|
bool operator==(const UniqueID& other) const {
|
|
return fID == other.fID;
|
|
}
|
|
bool operator!=(const UniqueID& other) const {
|
|
return !(*this == other);
|
|
}
|
|
|
|
bool isInvalid() const { return SK_InvalidUniqueID == fID; }
|
|
|
|
private:
|
|
const uint32_t fID;
|
|
};
|
|
|
|
/*
|
|
* The contract for the uniqueID is:
|
|
* for wrapped resources:
|
|
* the uniqueID will match that of the wrapped resource
|
|
*
|
|
* for deferred resources:
|
|
* the uniqueID will be different from the real resource, when it is allocated
|
|
* the proxy's uniqueID will not change across the instantiate call
|
|
*
|
|
* the uniqueIDs of the proxies and the resources draw from the same pool
|
|
*
|
|
* What this boils down to is that the uniqueID of a proxy can be used to consistently
|
|
* track/identify a proxy but should never be used to distinguish between
|
|
* resources and proxies - beware!
|
|
*/
|
|
UniqueID uniqueID() const { return fUniqueID; }
|
|
|
|
GrSurface* instantiate(GrTextureProvider* texProvider);
|
|
|
|
/**
|
|
* Helper that gets the width and height of the surface as a bounding rectangle.
|
|
*/
|
|
SkRect getBoundsRect() const { return SkRect::MakeIWH(this->width(), this->height()); }
|
|
|
|
int worstCaseWidth() const;
|
|
int worstCaseHeight() const;
|
|
|
|
/**
|
|
* @return the texture proxy associated with the surface proxy, may be NULL.
|
|
*/
|
|
virtual GrTextureProxy* asTextureProxy() { return nullptr; }
|
|
virtual const GrTextureProxy* asTextureProxy() const { return nullptr; }
|
|
|
|
/**
|
|
* @return the render target proxy associated with the surface proxy, may be NULL.
|
|
*/
|
|
virtual GrRenderTargetProxy* asRenderTargetProxy() { return nullptr; }
|
|
virtual const GrRenderTargetProxy* asRenderTargetProxy() const { return nullptr; }
|
|
|
|
/**
|
|
* Does the resource count against the resource budget?
|
|
*/
|
|
SkBudgeted isBudgeted() const { return fBudgeted; }
|
|
|
|
void setLastOpList(GrOpList* opList);
|
|
GrOpList* getLastOpList() { return fLastOpList; }
|
|
|
|
GrRenderTargetOpList* getLastRenderTargetOpList();
|
|
GrTextureOpList* getLastTextureOpList();
|
|
|
|
/**
|
|
* Retrieves the amount of GPU memory that will be or currently is used by this resource
|
|
* in bytes. It is approximate since we aren't aware of additional padding or copies made
|
|
* by the driver.
|
|
*
|
|
* @return the amount of GPU memory used in bytes
|
|
*/
|
|
size_t gpuMemorySize() const {
|
|
if (kInvalidGpuMemorySize == fGpuMemorySize) {
|
|
fGpuMemorySize = this->onGpuMemorySize();
|
|
SkASSERT(kInvalidGpuMemorySize != fGpuMemorySize);
|
|
}
|
|
return fGpuMemorySize;
|
|
}
|
|
|
|
bool isWrapped_ForTesting() const;
|
|
|
|
SkDEBUGCODE(void validate(GrContext*) const;)
|
|
|
|
protected:
|
|
// Deferred version
|
|
GrSurfaceProxy(const GrSurfaceDesc& desc, SkBackingFit fit, SkBudgeted budgeted)
|
|
: fDesc(desc)
|
|
, fFit(fit)
|
|
, fBudgeted(budgeted)
|
|
, fGpuMemorySize(kInvalidGpuMemorySize)
|
|
, fLastOpList(nullptr) {
|
|
// Note: this ctor pulls a new uniqueID from the same pool at the GrGpuResources
|
|
}
|
|
|
|
// Wrapped version
|
|
GrSurfaceProxy(sk_sp<GrSurface> surface, SkBackingFit fit);
|
|
|
|
virtual ~GrSurfaceProxy();
|
|
|
|
// For wrapped resources, 'fDesc' will always be filled in from the wrapped resource.
|
|
const GrSurfaceDesc fDesc;
|
|
const SkBackingFit fFit; // always exact for wrapped resources
|
|
const SkBudgeted fBudgeted; // set from the backing resource for wrapped resources
|
|
const UniqueID fUniqueID; // set from the backing resource for wrapped resources
|
|
|
|
static const size_t kInvalidGpuMemorySize = ~static_cast<size_t>(0);
|
|
SkDEBUGCODE(size_t getRawGpuMemorySize_debugOnly() const { return fGpuMemorySize; })
|
|
|
|
private:
|
|
virtual size_t onGpuMemorySize() const = 0;
|
|
|
|
// This entry is lazily evaluated so, when the proxy wraps a resource, the resource
|
|
// will be called but, when the proxy is deferred, it will compute the answer itself.
|
|
// If the proxy computes its own answer that answer is checked (in debug mode) in
|
|
// the instantiation method.
|
|
mutable size_t fGpuMemorySize;
|
|
|
|
// The last opList that wrote to or is currently going to write to this surface
|
|
// The opList can be closed (e.g., no render target context is currently bound
|
|
// to this renderTarget).
|
|
// This back-pointer is required so that we can add a dependancy between
|
|
// the opList used to create the current contents of this surface
|
|
// and the opList of a destination surface to which this one is being drawn or copied.
|
|
GrOpList* fLastOpList;
|
|
|
|
|
|
typedef GrIORefProxy INHERITED;
|
|
};
|
|
|
|
#endif
|