gecko-dev/image/ProgressTracker.cpp

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

512 строки
16 KiB
C++
Исходник Обычный вид История

/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*-
*
2012-05-21 15:12:37 +04:00
* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
#include "ImageLogging.h"
#include "ProgressTracker.h"
#include "imgINotificationObserver.h"
#include "imgIRequest.h"
#include "Image.h"
#include "nsNetUtil.h"
#include "nsIObserverService.h"
#include "mozilla/AppShutdown.h"
#include "mozilla/Assertions.h"
#include "mozilla/SchedulerGroup.h"
#include "mozilla/Services.h"
using mozilla::WeakPtr;
namespace mozilla {
namespace image {
Bug 1321946. Adjust asserts about the progress of an image to deal with how multipart images are handled. r=aosmond For multipart images we create a MultipartImage which contains each part. Each part in turn is a VectorImage or RasterImage. The MultipartImage and each part image all have their own ProgressTracker. The ProgressTracker for the MultipartImage observes the notifications of each part image via the IProgressObserver interface. This interfaces notably has no way to notify about an image error. So when a part image has an error it never gets propagated to the MultipartImage's ProgressTracker. This confuses our assertions about consistency of progress notifications. In this case we expect that when we get the load complete notification then we either have the size of the image or we encountered an error. So if the first part of a multipart image is broken and we are unable to get a size from it we will trigger this assertion. There are two ways to fix this. One would be to propagate errors to the MultipartImage's ProgressTracker. This would put the ProgressTracker for the MultipartImage permanently into error state and prevent showing the images from the remaining parts if one part image had an error. So in this patch I create a way to tell a ProgressTracker that is is for a multipart image, and use that to relax the assertions. As far as I can tell our code should be able to handle "ignoring" an error in a bad part image. Addtionaly there is a way that an error flag can get propagated to the MultipartImage's tracker: in MultipartImage::FinishTransition we get the progress directly from the part image and notify for it. This seems like an oversight as the comment at https://dxr.mozilla.org/mozilla-central/rev/bfa85d23df57c8a1db17c99b267667becc1c4afd/image/imgRequest.cpp#989 indicates that we don't want one bad part to prevent later parts from displaying. So we add the error flag to the ones we filter out when we propagate progress.
2016-12-04 01:07:10 +03:00
static void CheckProgressConsistency(Progress aOldProgress,
Progress aNewProgress, bool aIsMultipart) {
// Check preconditions for every progress bit.
Bug 1321946. Adjust asserts about the progress of an image to deal with how multipart images are handled. r=aosmond For multipart images we create a MultipartImage which contains each part. Each part in turn is a VectorImage or RasterImage. The MultipartImage and each part image all have their own ProgressTracker. The ProgressTracker for the MultipartImage observes the notifications of each part image via the IProgressObserver interface. This interfaces notably has no way to notify about an image error. So when a part image has an error it never gets propagated to the MultipartImage's ProgressTracker. This confuses our assertions about consistency of progress notifications. In this case we expect that when we get the load complete notification then we either have the size of the image or we encountered an error. So if the first part of a multipart image is broken and we are unable to get a size from it we will trigger this assertion. There are two ways to fix this. One would be to propagate errors to the MultipartImage's ProgressTracker. This would put the ProgressTracker for the MultipartImage permanently into error state and prevent showing the images from the remaining parts if one part image had an error. So in this patch I create a way to tell a ProgressTracker that is is for a multipart image, and use that to relax the assertions. As far as I can tell our code should be able to handle "ignoring" an error in a bad part image. Addtionaly there is a way that an error flag can get propagated to the MultipartImage's tracker: in MultipartImage::FinishTransition we get the progress directly from the part image and notify for it. This seems like an oversight as the comment at https://dxr.mozilla.org/mozilla-central/rev/bfa85d23df57c8a1db17c99b267667becc1c4afd/image/imgRequest.cpp#989 indicates that we don't want one bad part to prevent later parts from displaying. So we add the error flag to the ones we filter out when we propagate progress.
2016-12-04 01:07:10 +03:00
// Error's do not get propagated from the tracker for each image part to the
// tracker for the multipart image because we don't want one bad part to
// prevent the remaining parts from showing. So we need to consider whether
// this is a tracker for a multipart image for these assertions to work.
if (aNewProgress & FLAG_SIZE_AVAILABLE) {
// No preconditions.
}
if (aNewProgress & FLAG_DECODE_COMPLETE) {
MOZ_ASSERT(aNewProgress & FLAG_SIZE_AVAILABLE);
Bug 1321946. Adjust asserts about the progress of an image to deal with how multipart images are handled. r=aosmond For multipart images we create a MultipartImage which contains each part. Each part in turn is a VectorImage or RasterImage. The MultipartImage and each part image all have their own ProgressTracker. The ProgressTracker for the MultipartImage observes the notifications of each part image via the IProgressObserver interface. This interfaces notably has no way to notify about an image error. So when a part image has an error it never gets propagated to the MultipartImage's ProgressTracker. This confuses our assertions about consistency of progress notifications. In this case we expect that when we get the load complete notification then we either have the size of the image or we encountered an error. So if the first part of a multipart image is broken and we are unable to get a size from it we will trigger this assertion. There are two ways to fix this. One would be to propagate errors to the MultipartImage's ProgressTracker. This would put the ProgressTracker for the MultipartImage permanently into error state and prevent showing the images from the remaining parts if one part image had an error. So in this patch I create a way to tell a ProgressTracker that is is for a multipart image, and use that to relax the assertions. As far as I can tell our code should be able to handle "ignoring" an error in a bad part image. Addtionaly there is a way that an error flag can get propagated to the MultipartImage's tracker: in MultipartImage::FinishTransition we get the progress directly from the part image and notify for it. This seems like an oversight as the comment at https://dxr.mozilla.org/mozilla-central/rev/bfa85d23df57c8a1db17c99b267667becc1c4afd/image/imgRequest.cpp#989 indicates that we don't want one bad part to prevent later parts from displaying. So we add the error flag to the ones we filter out when we propagate progress.
2016-12-04 01:07:10 +03:00
MOZ_ASSERT(aIsMultipart ||
aNewProgress & (FLAG_FRAME_COMPLETE | FLAG_HAS_ERROR));
}
if (aNewProgress & FLAG_FRAME_COMPLETE) {
MOZ_ASSERT(aNewProgress & FLAG_SIZE_AVAILABLE);
}
if (aNewProgress & FLAG_LOAD_COMPLETE) {
Bug 1321946. Adjust asserts about the progress of an image to deal with how multipart images are handled. r=aosmond For multipart images we create a MultipartImage which contains each part. Each part in turn is a VectorImage or RasterImage. The MultipartImage and each part image all have their own ProgressTracker. The ProgressTracker for the MultipartImage observes the notifications of each part image via the IProgressObserver interface. This interfaces notably has no way to notify about an image error. So when a part image has an error it never gets propagated to the MultipartImage's ProgressTracker. This confuses our assertions about consistency of progress notifications. In this case we expect that when we get the load complete notification then we either have the size of the image or we encountered an error. So if the first part of a multipart image is broken and we are unable to get a size from it we will trigger this assertion. There are two ways to fix this. One would be to propagate errors to the MultipartImage's ProgressTracker. This would put the ProgressTracker for the MultipartImage permanently into error state and prevent showing the images from the remaining parts if one part image had an error. So in this patch I create a way to tell a ProgressTracker that is is for a multipart image, and use that to relax the assertions. As far as I can tell our code should be able to handle "ignoring" an error in a bad part image. Addtionaly there is a way that an error flag can get propagated to the MultipartImage's tracker: in MultipartImage::FinishTransition we get the progress directly from the part image and notify for it. This seems like an oversight as the comment at https://dxr.mozilla.org/mozilla-central/rev/bfa85d23df57c8a1db17c99b267667becc1c4afd/image/imgRequest.cpp#989 indicates that we don't want one bad part to prevent later parts from displaying. So we add the error flag to the ones we filter out when we propagate progress.
2016-12-04 01:07:10 +03:00
MOZ_ASSERT(aIsMultipart ||
aNewProgress & (FLAG_SIZE_AVAILABLE | FLAG_HAS_ERROR));
}
if (aNewProgress & FLAG_IS_ANIMATED) {
// No preconditions; like FLAG_HAS_TRANSPARENCY, we should normally never
// discover this *after* FLAG_SIZE_AVAILABLE, but unfortunately some corrupt
// GIFs may fool us.
}
if (aNewProgress & FLAG_HAS_TRANSPARENCY) {
// XXX We'd like to assert that transparency is only set during metadata
// decode but we don't have any way to assert that until bug 1254892 is
// fixed.
}
if (aNewProgress & FLAG_LAST_PART_COMPLETE) {
MOZ_ASSERT(aNewProgress & FLAG_LOAD_COMPLETE);
}
if (aNewProgress & FLAG_HAS_ERROR) {
// No preconditions.
}
}
ProgressTracker::ProgressTracker()
: mMutex("ProgressTracker::mMutex"),
mImage(nullptr),
mObservers(new ObserverTable),
mProgress(NoProgress),
mIsMultipart(false) {}
void ProgressTracker::SetImage(Image* aImage) {
MutexAutoLock lock(mMutex);
MOZ_ASSERT(aImage, "Setting null image");
MOZ_ASSERT(!mImage, "Setting image when we already have one");
mImage = aImage;
}
void ProgressTracker::ResetImage() {
MutexAutoLock lock(mMutex);
MOZ_ASSERT(mImage, "Resetting image when it's already null!");
mImage = nullptr;
}
uint32_t ProgressTracker::GetImageStatus() const {
uint32_t status = imgIRequest::STATUS_NONE;
// Translate our current state to a set of imgIRequest::STATE_* flags.
if (mProgress & FLAG_SIZE_AVAILABLE) {
status |= imgIRequest::STATUS_SIZE_AVAILABLE;
}
if (mProgress & FLAG_DECODE_COMPLETE) {
status |= imgIRequest::STATUS_DECODE_COMPLETE;
}
if (mProgress & FLAG_FRAME_COMPLETE) {
status |= imgIRequest::STATUS_FRAME_COMPLETE;
}
if (mProgress & FLAG_LOAD_COMPLETE) {
status |= imgIRequest::STATUS_LOAD_COMPLETE;
}
if (mProgress & FLAG_IS_ANIMATED) {
status |= imgIRequest::STATUS_IS_ANIMATED;
}
if (mProgress & FLAG_HAS_TRANSPARENCY) {
status |= imgIRequest::STATUS_HAS_TRANSPARENCY;
}
if (mProgress & FLAG_HAS_ERROR) {
status |= imgIRequest::STATUS_ERROR;
}
return status;
}
// A helper class to allow us to call SyncNotify asynchronously.
class AsyncNotifyRunnable : public Runnable {
public:
AsyncNotifyRunnable(ProgressTracker* aTracker, IProgressObserver* aObserver)
: Runnable("ProgressTracker::AsyncNotifyRunnable"), mTracker(aTracker) {
MOZ_ASSERT(NS_IsMainThread(), "Should be created on the main thread");
MOZ_ASSERT(aTracker, "aTracker should not be null");
MOZ_ASSERT(aObserver, "aObserver should not be null");
mObservers.AppendElement(aObserver);
}
NS_IMETHOD Run() override {
MOZ_ASSERT(NS_IsMainThread(), "Should be running on the main thread");
MOZ_ASSERT(mTracker, "mTracker should not be null");
for (uint32_t i = 0; i < mObservers.Length(); ++i) {
mObservers[i]->ClearPendingNotify();
mTracker->SyncNotify(mObservers[i]);
}
mTracker->mRunnable = nullptr;
return NS_OK;
}
void AddObserver(IProgressObserver* aObserver) {
mObservers.AppendElement(aObserver);
}
void RemoveObserver(IProgressObserver* aObserver) {
mObservers.RemoveElement(aObserver);
}
private:
friend class ProgressTracker;
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<ProgressTracker> mTracker;
nsTArray<RefPtr<IProgressObserver>> mObservers;
};
ProgressTracker::RenderBlockingRunnable::RenderBlockingRunnable(
already_AddRefed<AsyncNotifyRunnable>&& aEvent)
: PrioritizableRunnable(std::move(aEvent),
nsIRunnablePriority::PRIORITY_RENDER_BLOCKING) {}
void ProgressTracker::RenderBlockingRunnable::AddObserver(
IProgressObserver* aObserver) {
static_cast<AsyncNotifyRunnable*>(mRunnable.get())->AddObserver(aObserver);
}
void ProgressTracker::RenderBlockingRunnable::RemoveObserver(
IProgressObserver* aObserver) {
static_cast<AsyncNotifyRunnable*>(mRunnable.get())->RemoveObserver(aObserver);
}
/* static */
already_AddRefed<ProgressTracker::RenderBlockingRunnable>
ProgressTracker::RenderBlockingRunnable::Create(
already_AddRefed<AsyncNotifyRunnable>&& aEvent) {
MOZ_ASSERT(NS_IsMainThread());
RefPtr<ProgressTracker::RenderBlockingRunnable> event(
new ProgressTracker::RenderBlockingRunnable(std::move(aEvent)));
return event.forget();
}
void ProgressTracker::Notify(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread());
if (aObserver->NotificationsDeferred()) {
// There is a pending notification, or the observer isn't ready yet.
return;
}
if (MOZ_LOG_TEST(gImgLog, LogLevel::Debug)) {
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<Image> image = GetImage();
LOG_FUNC_WITH_PARAM(gImgLog, "ProgressTracker::Notify async", "uri", image);
}
aObserver->MarkPendingNotify();
// If we have an existing runnable that we can use, we just append this
// observer to its list of observers to be notified. This ensures we don't
// unnecessarily delay onload.
if (mRunnable) {
mRunnable->AddObserver(aObserver);
} else if (!AppShutdown::IsInOrBeyond(ShutdownPhase::XPCOMShutdownThreads)) {
// Avoid dispatch if we are late in shutdown.
RefPtr<AsyncNotifyRunnable> ev = new AsyncNotifyRunnable(this, aObserver);
mRunnable = ProgressTracker::RenderBlockingRunnable::Create(ev.forget());
SchedulerGroup::Dispatch(do_AddRef(mRunnable));
}
}
// A helper class to allow us to call SyncNotify asynchronously for a given,
// fixed, state.
class AsyncNotifyCurrentStateRunnable : public Runnable {
public:
AsyncNotifyCurrentStateRunnable(ProgressTracker* aProgressTracker,
IProgressObserver* aObserver)
: Runnable("image::AsyncNotifyCurrentStateRunnable"),
mProgressTracker(aProgressTracker),
mObserver(aObserver) {
MOZ_ASSERT(NS_IsMainThread(), "Should be created on the main thread");
MOZ_ASSERT(mProgressTracker, "mProgressTracker should not be null");
MOZ_ASSERT(mObserver, "mObserver should not be null");
mImage = mProgressTracker->GetImage();
}
NS_IMETHOD Run() override {
MOZ_ASSERT(NS_IsMainThread(), "Should be running on the main thread");
mObserver->ClearPendingNotify();
mProgressTracker->SyncNotify(mObserver);
return NS_OK;
}
private:
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<ProgressTracker> mProgressTracker;
RefPtr<IProgressObserver> mObserver;
// We have to hold on to a reference to the tracker's image, just in case
// it goes away while we're in the event queue.
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<Image> mImage;
};
void ProgressTracker::NotifyCurrentState(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread());
if (aObserver->NotificationsDeferred()) {
// There is a pending notification, or the observer isn't ready yet.
return;
}
if (MOZ_LOG_TEST(gImgLog, LogLevel::Debug)) {
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<Image> image = GetImage();
LOG_FUNC_WITH_PARAM(gImgLog, "ProgressTracker::NotifyCurrentState", "uri",
image);
}
aObserver->MarkPendingNotify();
// Avoid dispatch if we are late in shutdown.
if (!AppShutdown::IsInOrBeyond(ShutdownPhase::XPCOMShutdownThreads)) {
nsCOMPtr<nsIRunnable> ev =
new AsyncNotifyCurrentStateRunnable(this, aObserver);
SchedulerGroup::Dispatch(ev.forget());
}
}
/**
* ImageObserverNotifier is a helper type that abstracts over the difference
* between sending notifications to all of the observers in an ObserverTable,
* and sending them to a single observer. This allows the same notification code
* to be used for both cases.
*/
template <typename T>
struct ImageObserverNotifier;
template <>
struct MOZ_STACK_CLASS ImageObserverNotifier<const ObserverTable*> {
explicit ImageObserverNotifier(const ObserverTable* aObservers,
bool aIgnoreDeferral = false)
: mObservers(aObservers), mIgnoreDeferral(aIgnoreDeferral) {}
template <typename Lambda>
void operator()(Lambda aFunc) {
for (const auto& weakObserver : mObservers->Values()) {
RefPtr<IProgressObserver> observer = weakObserver.get();
if (observer && (mIgnoreDeferral || !observer->NotificationsDeferred())) {
aFunc(observer);
}
}
}
private:
const ObserverTable* mObservers;
const bool mIgnoreDeferral;
};
template <>
struct MOZ_STACK_CLASS ImageObserverNotifier<IProgressObserver*> {
explicit ImageObserverNotifier(IProgressObserver* aObserver)
: mObserver(aObserver) {}
template <typename Lambda>
void operator()(Lambda aFunc) {
if (mObserver && !mObserver->NotificationsDeferred()) {
aFunc(mObserver);
}
}
private:
IProgressObserver* mObserver;
};
template <typename T>
void SyncNotifyInternal(const T& aObservers, bool aHasImage, Progress aProgress,
const nsIntRect& aDirtyRect) {
MOZ_ASSERT(NS_IsMainThread());
typedef imgINotificationObserver I;
ImageObserverNotifier<T> notify(aObservers);
if (aProgress & FLAG_SIZE_AVAILABLE) {
notify([](IProgressObserver* aObs) { aObs->Notify(I::SIZE_AVAILABLE); });
}
2012-10-12 05:58:24 +04:00
if (aHasImage) {
// OnFrameUpdate
// If there's any content in this frame at all (always true for
// vector images, true for raster images that have decoded at
// least one frame) then send OnFrameUpdate.
if (!aDirtyRect.IsEmpty()) {
notify([&](IProgressObserver* aObs) {
aObs->Notify(I::FRAME_UPDATE, &aDirtyRect);
});
}
if (aProgress & FLAG_FRAME_COMPLETE) {
notify([](IProgressObserver* aObs) { aObs->Notify(I::FRAME_COMPLETE); });
}
if (aProgress & FLAG_HAS_TRANSPARENCY) {
notify(
[](IProgressObserver* aObs) { aObs->Notify(I::HAS_TRANSPARENCY); });
}
if (aProgress & FLAG_IS_ANIMATED) {
notify([](IProgressObserver* aObs) { aObs->Notify(I::IS_ANIMATED); });
}
}
if (aProgress & FLAG_DECODE_COMPLETE) {
MOZ_ASSERT(aHasImage, "Stopped decoding without ever having an image?");
notify([](IProgressObserver* aObs) { aObs->Notify(I::DECODE_COMPLETE); });
}
if (aProgress & FLAG_LOAD_COMPLETE) {
notify([=](IProgressObserver* aObs) {
aObs->OnLoadComplete(aProgress & FLAG_LAST_PART_COMPLETE);
});
}
}
void ProgressTracker::SyncNotifyProgress(Progress aProgress,
const nsIntRect& aInvalidRect
/* = nsIntRect() */) {
MOZ_ASSERT(NS_IsMainThread(), "Use mObservers on main thread only");
Progress progress = Difference(aProgress);
Bug 1321946. Adjust asserts about the progress of an image to deal with how multipart images are handled. r=aosmond For multipart images we create a MultipartImage which contains each part. Each part in turn is a VectorImage or RasterImage. The MultipartImage and each part image all have their own ProgressTracker. The ProgressTracker for the MultipartImage observes the notifications of each part image via the IProgressObserver interface. This interfaces notably has no way to notify about an image error. So when a part image has an error it never gets propagated to the MultipartImage's ProgressTracker. This confuses our assertions about consistency of progress notifications. In this case we expect that when we get the load complete notification then we either have the size of the image or we encountered an error. So if the first part of a multipart image is broken and we are unable to get a size from it we will trigger this assertion. There are two ways to fix this. One would be to propagate errors to the MultipartImage's ProgressTracker. This would put the ProgressTracker for the MultipartImage permanently into error state and prevent showing the images from the remaining parts if one part image had an error. So in this patch I create a way to tell a ProgressTracker that is is for a multipart image, and use that to relax the assertions. As far as I can tell our code should be able to handle "ignoring" an error in a bad part image. Addtionaly there is a way that an error flag can get propagated to the MultipartImage's tracker: in MultipartImage::FinishTransition we get the progress directly from the part image and notify for it. This seems like an oversight as the comment at https://dxr.mozilla.org/mozilla-central/rev/bfa85d23df57c8a1db17c99b267667becc1c4afd/image/imgRequest.cpp#989 indicates that we don't want one bad part to prevent later parts from displaying. So we add the error flag to the ones we filter out when we propagate progress.
2016-12-04 01:07:10 +03:00
CheckProgressConsistency(mProgress, mProgress | progress, mIsMultipart);
// Apply the changes.
mProgress |= progress;
// Send notifications.
mObservers.Read([&](const ObserverTable* aTable) {
SyncNotifyInternal(aTable, HasImage(), progress, aInvalidRect);
});
if (progress & FLAG_HAS_ERROR) {
FireFailureNotification();
}
}
void ProgressTracker::SyncNotify(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread());
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<Image> image = GetImage();
LOG_SCOPE_WITH_PARAM(gImgLog, "ProgressTracker::SyncNotify", "uri", image);
nsIntRect rect;
if (image) {
int32_t width, height;
if (NS_FAILED(image->GetWidth(&width)) ||
NS_FAILED(image->GetHeight(&height))) {
// Either the image has no intrinsic size, or it has an error.
rect = GetMaxSizedIntRect();
} else {
rect.SizeTo(width, height);
}
}
SyncNotifyInternal(aObserver, !!image, mProgress, rect);
}
void ProgressTracker::EmulateRequestFinished(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread(),
"SyncNotifyState and mObservers are not threadsafe");
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<IProgressObserver> kungFuDeathGrip(aObserver);
if (!(mProgress & FLAG_LOAD_COMPLETE)) {
aObserver->OnLoadComplete(true);
}
}
already_AddRefed<nsIEventTarget> ProgressTracker::GetEventTarget() const {
return do_AddRef(GetMainThreadSerialEventTarget());
}
void ProgressTracker::AddObserver(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread());
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<IProgressObserver> observer = aObserver;
mObservers.Write([=](ObserverTable* aTable) {
MOZ_ASSERT(!aTable->Contains(observer),
"Adding duplicate entry for image observer");
WeakPtr<IProgressObserver> weakPtr = observer.get();
aTable->InsertOrUpdate(observer, weakPtr);
});
}
bool ProgressTracker::RemoveObserver(IProgressObserver* aObserver) {
MOZ_ASSERT(NS_IsMainThread());
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<IProgressObserver> observer = aObserver;
// Remove the observer from the list.
bool removed = mObservers.Write(
[observer](ObserverTable* aTable) { return aTable->Remove(observer); });
// Observers can get confused if they don't get all the proper teardown
// notifications. Part ways on good terms.
if (removed && !aObserver->NotificationsDeferred()) {
EmulateRequestFinished(aObserver);
}
// Make sure we don't give callbacks to an observer that isn't interested in
// them any more.
if (aObserver->NotificationsDeferred() && mRunnable) {
mRunnable->RemoveObserver(aObserver);
aObserver->ClearPendingNotify();
}
return removed;
}
uint32_t ProgressTracker::ObserverCount() const {
MOZ_ASSERT(NS_IsMainThread());
return mObservers.Read(
[](const ObserverTable* aTable) { return aTable->Count(); });
}
void ProgressTracker::OnUnlockedDraw() {
MOZ_ASSERT(NS_IsMainThread());
mObservers.Read([](const ObserverTable* aTable) {
ImageObserverNotifier<const ObserverTable*> notify(aTable);
notify([](IProgressObserver* aObs) {
aObs->Notify(imgINotificationObserver::UNLOCKED_DRAW);
});
});
}
void ProgressTracker::ResetForNewRequest() {
MOZ_ASSERT(NS_IsMainThread());
mProgress = NoProgress;
}
void ProgressTracker::OnDiscard() {
MOZ_ASSERT(NS_IsMainThread());
mObservers.Read([](const ObserverTable* aTable) {
ImageObserverNotifier<const ObserverTable*> notify(aTable);
notify([](IProgressObserver* aObs) {
aObs->Notify(imgINotificationObserver::DISCARD);
});
});
}
void ProgressTracker::OnImageAvailable() {
MOZ_ASSERT(NS_IsMainThread());
// Notify any imgRequestProxys that are observing us that we have an Image.
mObservers.Read([](const ObserverTable* aTable) {
ImageObserverNotifier<const ObserverTable*> notify(
aTable, /* aIgnoreDeferral = */ true);
notify([](IProgressObserver* aObs) { aObs->SetHasImage(); });
});
}
void ProgressTracker::FireFailureNotification() {
MOZ_ASSERT(NS_IsMainThread());
// Some kind of problem has happened with image decoding.
// Report the URI to net:failed-to-process-uri-conent observers.
Bug 1207245 - part 6 - rename nsRefPtr<T> to RefPtr<T>; r=ehsan; a=Tomcat The bulk of this commit was generated with a script, executed at the top level of a typical source code checkout. The only non-machine-generated part was modifying MFBT's moz.build to reflect the new naming. CLOSED TREE makes big refactorings like this a piece of cake. # The main substitution. find . -name '*.cpp' -o -name '*.cc' -o -name '*.h' -o -name '*.mm' -o -name '*.idl'| \ xargs perl -p -i -e ' s/nsRefPtr\.h/RefPtr\.h/g; # handle includes s/nsRefPtr ?</RefPtr</g; # handle declarations and variables ' # Handle a special friend declaration in gfx/layers/AtomicRefCountedWithFinalize.h. perl -p -i -e 's/::nsRefPtr;/::RefPtr;/' gfx/layers/AtomicRefCountedWithFinalize.h # Handle nsRefPtr.h itself, a couple places that define constructors # from nsRefPtr, and code generators specially. We do this here, rather # than indiscriminantly s/nsRefPtr/RefPtr/, because that would rename # things like nsRefPtrHashtable. perl -p -i -e 's/nsRefPtr/RefPtr/g' \ mfbt/nsRefPtr.h \ xpcom/glue/nsCOMPtr.h \ xpcom/base/OwningNonNull.h \ ipc/ipdl/ipdl/lower.py \ ipc/ipdl/ipdl/builtin.py \ dom/bindings/Codegen.py \ python/lldbutils/lldbutils/utils.py # In our indiscriminate substitution above, we renamed # nsRefPtrGetterAddRefs, the class behind getter_AddRefs. Fix that up. find . -name '*.cpp' -o -name '*.h' -o -name '*.idl' | \ xargs perl -p -i -e 's/nsRefPtrGetterAddRefs/RefPtrGetterAddRefs/g' if [ -d .git ]; then git mv mfbt/nsRefPtr.h mfbt/RefPtr.h else hg mv mfbt/nsRefPtr.h mfbt/RefPtr.h fi --HG-- rename : mfbt/nsRefPtr.h => mfbt/RefPtr.h
2015-10-18 08:24:48 +03:00
RefPtr<Image> image = GetImage();
if (image) {
// Should be on main thread, so ok to create a new nsIURI.
nsCOMPtr<nsIURI> uri = image->GetURI();
if (uri) {
nsCOMPtr<nsIObserverService> os = mozilla::services::GetObserverService();
if (os) {
os->NotifyObservers(uri, "net:failed-to-process-uri-content", nullptr);
}
}
}
}
} // namespace image
} // namespace mozilla