/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2; -*- */ /* 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 "HeapSnapshot.h" #include #include #include #include "js/Array.h" // JS::NewArrayObject #include "js/Debug.h" #include "js/PropertyAndElement.h" // JS_DefineProperty #include "js/TypeDecls.h" #include "js/UbiNodeBreadthFirst.h" #include "js/UbiNodeCensus.h" #include "js/UbiNodeDominatorTree.h" #include "js/UbiNodeShortestPaths.h" #include "mozilla/Attributes.h" #include "mozilla/CycleCollectedJSContext.h" #include "mozilla/devtools/AutoMemMap.h" #include "mozilla/devtools/CoreDump.pb.h" #include "mozilla/devtools/DeserializedNode.h" #include "mozilla/devtools/DominatorTree.h" #include "mozilla/devtools/FileDescriptorOutputStream.h" #include "mozilla/devtools/HeapSnapshotTempFileHelperChild.h" #include "mozilla/devtools/ZeroCopyNSIOutputStream.h" #include "mozilla/dom/ChromeUtils.h" #include "mozilla/dom/ContentChild.h" #include "mozilla/dom/HeapSnapshotBinding.h" #include "mozilla/RangedPtr.h" #include "mozilla/Telemetry.h" #include "mozilla/Unused.h" #include "jsapi.h" #include "jsfriendapi.h" #include "js/MapAndSet.h" #include "js/Object.h" // JS::GetCompartment #include "nsComponentManagerUtils.h" // do_CreateInstance #include "nsCycleCollectionParticipant.h" #include "nsCRTGlue.h" #include "nsIFile.h" #include "nsIOutputStream.h" #include "nsISupportsImpl.h" #include "nsNetUtil.h" #include "nsPrintfCString.h" #include "prerror.h" #include "prio.h" #include "prtypes.h" #include "SpecialSystemDirectory.h" namespace mozilla { namespace devtools { using namespace JS; using namespace dom; using ::google::protobuf::io::ArrayInputStream; using ::google::protobuf::io::CodedInputStream; using ::google::protobuf::io::GzipInputStream; using ::google::protobuf::io::ZeroCopyInputStream; using JS::ubi::AtomOrTwoByteChars; using JS::ubi::ShortestPaths; MallocSizeOf GetCurrentThreadDebuggerMallocSizeOf() { auto ccjscx = CycleCollectedJSContext::Get(); MOZ_ASSERT(ccjscx); auto cx = ccjscx->Context(); MOZ_ASSERT(cx); auto mallocSizeOf = JS::dbg::GetDebuggerMallocSizeOf(cx); MOZ_ASSERT(mallocSizeOf); return mallocSizeOf; } /*** Cycle Collection Boilerplate *********************************************/ NS_IMPL_CYCLE_COLLECTION_WRAPPERCACHE(HeapSnapshot, mParent) NS_IMPL_CYCLE_COLLECTING_ADDREF(HeapSnapshot) NS_IMPL_CYCLE_COLLECTING_RELEASE(HeapSnapshot) NS_INTERFACE_MAP_BEGIN_CYCLE_COLLECTION(HeapSnapshot) NS_WRAPPERCACHE_INTERFACE_MAP_ENTRY NS_INTERFACE_MAP_ENTRY(nsISupports) NS_INTERFACE_MAP_END /* virtual */ JSObject* HeapSnapshot::WrapObject(JSContext* aCx, JS::Handle aGivenProto) { return HeapSnapshot_Binding::Wrap(aCx, this, aGivenProto); } /*** Reading Heap Snapshots ***************************************************/ /* static */ already_AddRefed HeapSnapshot::Create(JSContext* cx, GlobalObject& global, const uint8_t* buffer, uint32_t size, ErrorResult& rv) { RefPtr snapshot = new HeapSnapshot(cx, global.GetAsSupports()); if (!snapshot->init(cx, buffer, size)) { rv.Throw(NS_ERROR_UNEXPECTED); return nullptr; } return snapshot.forget(); } template static bool parseMessage(ZeroCopyInputStream& stream, uint32_t sizeOfMessage, MessageType& message) { // We need to create a new `CodedInputStream` for each message so that the // 64MB limit is applied per-message rather than to the whole stream. CodedInputStream codedStream(&stream); // The protobuf message nesting that core dumps exhibit is dominated by // allocation stacks' frames. In the most deeply nested case, each frame has // two messages: a StackFrame message and a StackFrame::Data message. These // frames are on top of a small constant of other messages. There are a // MAX_STACK_DEPTH number of frames, so we multiply this by 3 to make room for // the two messages per frame plus some head room for the constant number of // non-dominating messages. codedStream.SetRecursionLimit(HeapSnapshot::MAX_STACK_DEPTH * 3); auto limit = codedStream.PushLimit(sizeOfMessage); if (NS_WARN_IF(!message.ParseFromCodedStream(&codedStream)) || NS_WARN_IF(!codedStream.ConsumedEntireMessage()) || NS_WARN_IF(codedStream.BytesUntilLimit() != 0)) { return false; } codedStream.PopLimit(limit); return true; } template struct GetOrInternStringMatcher { InternedStringSet& internedStrings; explicit GetOrInternStringMatcher(InternedStringSet& strings) : internedStrings(strings) {} const CharT* operator()(const std::string* str) { MOZ_ASSERT(str); size_t length = str->length() / sizeof(CharT); auto tempString = reinterpret_cast(str->data()); UniqueFreePtr owned(NS_xstrndup(tempString, length)); if (!internedStrings.append(std::move(owned))) return nullptr; return internedStrings.back().get(); } const CharT* operator()(uint64_t ref) { if (MOZ_LIKELY(ref < internedStrings.length())) { auto& string = internedStrings[ref]; MOZ_ASSERT(string); return string.get(); } return nullptr; } }; template < // Either char or char16_t. typename CharT, // A reference to either `internedOneByteStrings` or // `internedTwoByteStrings` if CharT is char or char16_t respectively. typename InternedStringSet> const CharT* HeapSnapshot::getOrInternString( InternedStringSet& internedStrings, Maybe& maybeStrOrRef) { // Incomplete message: has neither a string nor a reference to an already // interned string. if (MOZ_UNLIKELY(maybeStrOrRef.isNothing())) return nullptr; GetOrInternStringMatcher m(internedStrings); return maybeStrOrRef->match(m); } // Get a de-duplicated string as a Maybe from the given `msg`. #define GET_STRING_OR_REF_WITH_PROP_NAMES(msg, strPropertyName, \ refPropertyName) \ (msg.has_##refPropertyName() ? Some(StringOrRef(msg.refPropertyName())) \ : msg.has_##strPropertyName() ? Some(StringOrRef(&msg.strPropertyName())) \ : Nothing()) #define GET_STRING_OR_REF(msg, property) \ (msg.has_##property##ref() ? Some(StringOrRef(msg.property##ref())) \ : msg.has_##property() ? Some(StringOrRef(&msg.property())) \ : Nothing()) bool HeapSnapshot::saveNode(const protobuf::Node& node, NodeIdSet& edgeReferents) { // NB: de-duplicated string properties must be read back and interned in the // same order here as they are written and serialized in // `CoreDumpWriter::writeNode` or else indices in references to already // serialized strings will be off. if (NS_WARN_IF(!node.has_id())) return false; NodeId id = node.id(); // NodeIds are derived from pointers (at most 48 bits) and we rely on them // fitting into JS numbers (IEEE 754 doubles, can precisely store 53 bit // integers) despite storing them on disk as 64 bit integers. if (NS_WARN_IF(!JS::Value::isNumberRepresentable(id))) return false; // Should only deserialize each node once. if (NS_WARN_IF(nodes.has(id))) return false; if (NS_WARN_IF(!JS::ubi::Uint32IsValidCoarseType(node.coarsetype()))) return false; auto coarseType = JS::ubi::Uint32ToCoarseType(node.coarsetype()); Maybe typeNameOrRef = GET_STRING_OR_REF_WITH_PROP_NAMES(node, typename_, typenameref); auto typeName = getOrInternString(internedTwoByteStrings, typeNameOrRef); if (NS_WARN_IF(!typeName)) return false; if (NS_WARN_IF(!node.has_size())) return false; uint64_t size = node.size(); auto edgesLength = node.edges_size(); DeserializedNode::EdgeVector edges; if (NS_WARN_IF(!edges.reserve(edgesLength))) return false; for (decltype(edgesLength) i = 0; i < edgesLength; i++) { auto& protoEdge = node.edges(i); if (NS_WARN_IF(!protoEdge.has_referent())) return false; NodeId referent = protoEdge.referent(); if (NS_WARN_IF(!edgeReferents.put(referent))) return false; const char16_t* edgeName = nullptr; if (protoEdge.EdgeNameOrRef_case() != protobuf::Edge::EDGENAMEORREF_NOT_SET) { Maybe edgeNameOrRef = GET_STRING_OR_REF(protoEdge, name); edgeName = getOrInternString(internedTwoByteStrings, edgeNameOrRef); if (NS_WARN_IF(!edgeName)) return false; } edges.infallibleAppend(DeserializedEdge(referent, edgeName)); } Maybe allocationStack; if (node.has_allocationstack()) { StackFrameId id = 0; if (NS_WARN_IF(!saveStackFrame(node.allocationstack(), id))) return false; allocationStack.emplace(id); } MOZ_ASSERT(allocationStack.isSome() == node.has_allocationstack()); const char* jsObjectClassName = nullptr; if (node.JSObjectClassNameOrRef_case() != protobuf::Node::JSOBJECTCLASSNAMEORREF_NOT_SET) { Maybe clsNameOrRef = GET_STRING_OR_REF(node, jsobjectclassname); jsObjectClassName = getOrInternString(internedOneByteStrings, clsNameOrRef); if (NS_WARN_IF(!jsObjectClassName)) return false; } const char* scriptFilename = nullptr; if (node.ScriptFilenameOrRef_case() != protobuf::Node::SCRIPTFILENAMEORREF_NOT_SET) { Maybe scriptFilenameOrRef = GET_STRING_OR_REF(node, scriptfilename); scriptFilename = getOrInternString(internedOneByteStrings, scriptFilenameOrRef); if (NS_WARN_IF(!scriptFilename)) return false; } const char16_t* descriptiveTypeName = nullptr; if (node.descriptiveTypeNameOrRef_case() != protobuf::Node::DESCRIPTIVETYPENAMEORREF_NOT_SET) { Maybe descriptiveTypeNameOrRef = GET_STRING_OR_REF(node, descriptivetypename); descriptiveTypeName = getOrInternString(internedTwoByteStrings, descriptiveTypeNameOrRef); if (NS_WARN_IF(!descriptiveTypeName)) return false; } if (NS_WARN_IF(!nodes.putNew( id, DeserializedNode(id, coarseType, typeName, size, std::move(edges), allocationStack, jsObjectClassName, scriptFilename, descriptiveTypeName, *this)))) { return false; }; return true; } bool HeapSnapshot::saveStackFrame(const protobuf::StackFrame& frame, StackFrameId& outFrameId) { // NB: de-duplicated string properties must be read in the same order here as // they are written in `CoreDumpWriter::getProtobufStackFrame` or else indices // in references to already serialized strings will be off. if (frame.has_ref()) { // We should only get a reference to the previous frame if we have already // seen the previous frame. if (!frames.has(frame.ref())) return false; outFrameId = frame.ref(); return true; } // Incomplete message. if (!frame.has_data()) return false; auto data = frame.data(); if (!data.has_id()) return false; StackFrameId id = data.id(); // This should be the first and only time we see this frame. if (frames.has(id)) return false; if (!data.has_line()) return false; uint32_t line = data.line(); if (!data.has_column()) return false; uint32_t column = data.column(); if (!data.has_issystem()) return false; bool isSystem = data.issystem(); if (!data.has_isselfhosted()) return false; bool isSelfHosted = data.isselfhosted(); Maybe sourceOrRef = GET_STRING_OR_REF(data, source); auto source = getOrInternString(internedTwoByteStrings, sourceOrRef); if (!source) return false; const char16_t* functionDisplayName = nullptr; if (data.FunctionDisplayNameOrRef_case() != protobuf::StackFrame_Data::FUNCTIONDISPLAYNAMEORREF_NOT_SET) { Maybe nameOrRef = GET_STRING_OR_REF(data, functiondisplayname); functionDisplayName = getOrInternString(internedTwoByteStrings, nameOrRef); if (!functionDisplayName) return false; } Maybe parent; if (data.has_parent()) { StackFrameId parentId = 0; if (!saveStackFrame(data.parent(), parentId)) return false; parent = Some(parentId); } if (!frames.putNew(id, DeserializedStackFrame(id, parent, line, column, source, functionDisplayName, isSystem, isSelfHosted, *this))) { return false; } outFrameId = id; return true; } #undef GET_STRING_OR_REF_WITH_PROP_NAMES #undef GET_STRING_OR_REF // Because protobuf messages aren't self-delimiting, we serialize each message // preceded by its size in bytes. When deserializing, we read this size and then // limit reading from the stream to the given byte size. If we didn't, then the // first message would consume the entire stream. static bool readSizeOfNextMessage(ZeroCopyInputStream& stream, uint32_t* sizep) { MOZ_ASSERT(sizep); CodedInputStream codedStream(&stream); return codedStream.ReadVarint32(sizep) && *sizep > 0; } bool HeapSnapshot::init(JSContext* cx, const uint8_t* buffer, uint32_t size) { ArrayInputStream stream(buffer, size); GzipInputStream gzipStream(&stream); uint32_t sizeOfMessage = 0; // First is the metadata. protobuf::Metadata metadata; if (NS_WARN_IF(!readSizeOfNextMessage(gzipStream, &sizeOfMessage))) return false; if (!parseMessage(gzipStream, sizeOfMessage, metadata)) return false; if (metadata.has_timestamp()) timestamp.emplace(metadata.timestamp()); // Next is the root node. protobuf::Node root; if (NS_WARN_IF(!readSizeOfNextMessage(gzipStream, &sizeOfMessage))) return false; if (!parseMessage(gzipStream, sizeOfMessage, root)) return false; // Although the id is optional in the protobuf format for future proofing, we // can't currently do anything without it. if (NS_WARN_IF(!root.has_id())) return false; rootId = root.id(); // The set of all node ids we've found edges pointing to. NodeIdSet edgeReferents(cx); if (NS_WARN_IF(!saveNode(root, edgeReferents))) return false; // Finally, the rest of the nodes in the core dump. // Test for the end of the stream. The protobuf library gives no way to tell // the difference between an underlying read error and the stream being // done. All we can do is attempt to read the size of the next message and // extrapolate guestimations from the result of that operation. while (readSizeOfNextMessage(gzipStream, &sizeOfMessage)) { protobuf::Node node; if (!parseMessage(gzipStream, sizeOfMessage, node)) return false; if (NS_WARN_IF(!saveNode(node, edgeReferents))) return false; } // Check the set of node ids referred to by edges we found and ensure that we // have the node corresponding to each id. If we don't have all of them, it is // unsafe to perform analyses of this heap snapshot. for (auto iter = edgeReferents.iter(); !iter.done(); iter.next()) { if (NS_WARN_IF(!nodes.has(iter.get()))) return false; } return true; } /*** Heap Snapshot Analyses ***************************************************/ void HeapSnapshot::TakeCensus(JSContext* cx, JS::Handle options, JS::MutableHandle rval, ErrorResult& rv) { JS::ubi::Census census(cx); JS::ubi::CountTypePtr rootType; if (NS_WARN_IF(!JS::ubi::ParseCensusOptions(cx, census, options, rootType))) { rv.Throw(NS_ERROR_UNEXPECTED); return; } JS::ubi::RootedCount rootCount(cx, rootType->makeCount()); if (NS_WARN_IF(!rootCount)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } JS::ubi::CensusHandler handler(census, rootCount, GetCurrentThreadDebuggerMallocSizeOf()); { JS::AutoCheckCannotGC nogc; JS::ubi::CensusTraversal traversal(cx, handler, nogc); if (NS_WARN_IF(!traversal.addStart(getRoot()))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } if (NS_WARN_IF(!traversal.traverse())) { rv.Throw(NS_ERROR_UNEXPECTED); return; } } if (NS_WARN_IF(!handler.report(cx, rval))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } } void HeapSnapshot::DescribeNode(JSContext* cx, JS::Handle breakdown, uint64_t nodeId, JS::MutableHandle rval, ErrorResult& rv) { MOZ_ASSERT(breakdown); JS::Rooted breakdownVal(cx, JS::ObjectValue(*breakdown)); JS::ubi::CountTypePtr rootType = JS::ubi::ParseBreakdown(cx, breakdownVal); if (NS_WARN_IF(!rootType)) { rv.Throw(NS_ERROR_UNEXPECTED); return; } JS::ubi::RootedCount rootCount(cx, rootType->makeCount()); if (NS_WARN_IF(!rootCount)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } JS::ubi::Node::Id id(nodeId); Maybe node = getNodeById(id); if (NS_WARN_IF(node.isNothing())) { rv.Throw(NS_ERROR_INVALID_ARG); return; } MallocSizeOf mallocSizeOf = GetCurrentThreadDebuggerMallocSizeOf(); if (NS_WARN_IF(!rootCount->count(mallocSizeOf, *node))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } if (NS_WARN_IF(!rootCount->report(cx, rval))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } } already_AddRefed HeapSnapshot::ComputeDominatorTree( ErrorResult& rv) { Maybe maybeTree; { auto ccjscx = CycleCollectedJSContext::Get(); MOZ_ASSERT(ccjscx); auto cx = ccjscx->Context(); MOZ_ASSERT(cx); JS::AutoCheckCannotGC nogc(cx); maybeTree = JS::ubi::DominatorTree::Create(cx, nogc, getRoot()); } if (NS_WARN_IF(maybeTree.isNothing())) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return nullptr; } return MakeAndAddRef(std::move(*maybeTree), this, mParent); } void HeapSnapshot::ComputeShortestPaths(JSContext* cx, uint64_t start, const Sequence& targets, uint64_t maxNumPaths, JS::MutableHandle results, ErrorResult& rv) { // First ensure that our inputs are valid. if (NS_WARN_IF(maxNumPaths == 0)) { rv.Throw(NS_ERROR_INVALID_ARG); return; } Maybe startNode = getNodeById(start); if (NS_WARN_IF(startNode.isNothing())) { rv.Throw(NS_ERROR_INVALID_ARG); return; } if (NS_WARN_IF(targets.Length() == 0)) { rv.Throw(NS_ERROR_INVALID_ARG); return; } // Aggregate the targets into a set and make sure that they exist in the heap // snapshot. JS::ubi::NodeSet targetsSet; for (const auto& target : targets) { Maybe targetNode = getNodeById(target); if (NS_WARN_IF(targetNode.isNothing())) { rv.Throw(NS_ERROR_INVALID_ARG); return; } if (NS_WARN_IF(!targetsSet.put(*targetNode))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } } // Walk the heap graph and find the shortest paths. Maybe maybeShortestPaths; { JS::AutoCheckCannotGC nogc(cx); maybeShortestPaths = ShortestPaths::Create( cx, nogc, maxNumPaths, *startNode, std::move(targetsSet)); } if (NS_WARN_IF(maybeShortestPaths.isNothing())) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } auto& shortestPaths = *maybeShortestPaths; // Convert the results into a Map object mapping target node IDs to arrays of // paths found. JS::Rooted resultsMap(cx, JS::NewMapObject(cx)); if (NS_WARN_IF(!resultsMap)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } for (auto iter = shortestPaths.targetIter(); !iter.done(); iter.next()) { JS::Rooted key(cx, JS::NumberValue(iter.get().identifier())); JS::RootedVector paths(cx); bool ok = shortestPaths.forEachPath(iter.get(), [&](JS::ubi::Path& path) { JS::RootedVector pathValues(cx); for (JS::ubi::BackEdge* edge : path) { JS::Rooted pathPart(cx, JS_NewPlainObject(cx)); if (!pathPart) { return false; } JS::Rooted predecessor( cx, NumberValue(edge->predecessor().identifier())); if (!JS_DefineProperty(cx, pathPart, "predecessor", predecessor, JSPROP_ENUMERATE)) { return false; } JS::Rooted edgeNameVal(cx, NullValue()); if (edge->name()) { JS::Rooted edgeName( cx, JS_AtomizeUCString(cx, edge->name().get())); if (!edgeName) { return false; } edgeNameVal = StringValue(edgeName); } if (!JS_DefineProperty(cx, pathPart, "edge", edgeNameVal, JSPROP_ENUMERATE)) { return false; } if (!pathValues.append(ObjectValue(*pathPart))) { return false; } } JS::Rooted pathObj(cx, JS::NewArrayObject(cx, pathValues)); return pathObj && paths.append(ObjectValue(*pathObj)); }); if (NS_WARN_IF(!ok)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } JS::Rooted pathsArray(cx, JS::NewArrayObject(cx, paths)); if (NS_WARN_IF(!pathsArray)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } JS::Rooted pathsVal(cx, ObjectValue(*pathsArray)); if (NS_WARN_IF(!JS::MapSet(cx, resultsMap, key, pathsVal))) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return; } } results.set(resultsMap); } /*** Saving Heap Snapshots ****************************************************/ // If we are only taking a snapshot of the heap affected by the given set of // globals, find the set of compartments the globals are allocated // within. Returns false on OOM failure. static bool PopulateCompartmentsWithGlobals( CompartmentSet& compartments, JS::HandleVector globals) { unsigned length = globals.length(); for (unsigned i = 0; i < length; i++) { if (!compartments.put(JS::GetCompartment(globals[i]))) return false; } return true; } // Add the given set of globals as explicit roots in the given roots // list. Returns false on OOM failure. static bool AddGlobalsAsRoots(JS::HandleVector globals, ubi::RootList& roots) { unsigned length = globals.length(); for (unsigned i = 0; i < length; i++) { if (!roots.addRoot(ubi::Node(globals[i].get()), u"heap snapshot global")) { return false; } } return true; } // Choose roots and limits for a traversal, given `boundaries`. Set `roots` to // the set of nodes within the boundaries that are referred to by nodes // outside. If `boundaries` does not include all JS compartments, initialize // `compartments` to the set of included compartments; otherwise, leave // `compartments` uninitialized. (You can use compartments.initialized() to // check.) // // If `boundaries` is incoherent, or we encounter an error while trying to // handle it, or we run out of memory, set `rv` appropriately and return // `false`. // // Return value is a pair of the status and an AutoCheckCannotGC token, // forwarded from ubi::RootList::init(), to ensure that the caller does // not GC while the RootList is live and initialized. static std::pair EstablishBoundaries( JSContext* cx, ErrorResult& rv, const HeapSnapshotBoundaries& boundaries, ubi::RootList& roots, CompartmentSet& compartments) { MOZ_ASSERT(!roots.initialized()); MOZ_ASSERT(compartments.empty()); bool foundBoundaryProperty = false; if (boundaries.mRuntime.WasPassed()) { foundBoundaryProperty = true; if (!boundaries.mRuntime.Value()) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } auto [ok, nogc] = roots.init(); if (!ok) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return {false, nogc}; } } if (boundaries.mDebugger.WasPassed()) { if (foundBoundaryProperty) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } foundBoundaryProperty = true; JSObject* dbgObj = boundaries.mDebugger.Value(); if (!dbgObj || !dbg::IsDebugger(*dbgObj)) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } JS::RootedVector globals(cx); if (!dbg::GetDebuggeeGlobals(cx, *dbgObj, &globals) || !PopulateCompartmentsWithGlobals(compartments, globals) || !roots.init(compartments).first || !AddGlobalsAsRoots(globals, roots)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return {false, AutoCheckCannotGC(cx)}; } } if (boundaries.mGlobals.WasPassed()) { if (foundBoundaryProperty) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } foundBoundaryProperty = true; uint32_t length = boundaries.mGlobals.Value().Length(); if (length == 0) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } JS::RootedVector globals(cx); for (uint32_t i = 0; i < length; i++) { JSObject* global = boundaries.mGlobals.Value().ElementAt(i); if (!JS_IsGlobalObject(global)) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, AutoCheckCannotGC(cx)}; } if (!globals.append(global)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return {false, AutoCheckCannotGC(cx)}; } } if (!PopulateCompartmentsWithGlobals(compartments, globals) || !roots.init(compartments).first || !AddGlobalsAsRoots(globals, roots)) { rv.Throw(NS_ERROR_OUT_OF_MEMORY); return {false, AutoCheckCannotGC(cx)}; } } AutoCheckCannotGC nogc(cx); if (!foundBoundaryProperty) { rv.Throw(NS_ERROR_INVALID_ARG); return {false, nogc}; } MOZ_ASSERT(roots.initialized()); return {true, nogc}; } // A variant covering all the various two-byte strings that we can get from the // ubi::Node API. class TwoByteString : public Variant { using Base = Variant; struct CopyToBufferMatcher { RangedPtr destination; size_t maxLength; CopyToBufferMatcher(RangedPtr destination, size_t maxLength) : destination(destination), maxLength(maxLength) {} size_t operator()(JS::ubi::EdgeName& ptr) { return ptr ? operator()(ptr.get()) : 0; } size_t operator()(JSAtom* atom) { MOZ_ASSERT(atom); JS::ubi::AtomOrTwoByteChars s(atom); return s.copyToBuffer(destination, maxLength); } size_t operator()(const char16_t* chars) { MOZ_ASSERT(chars); JS::ubi::AtomOrTwoByteChars s(chars); return s.copyToBuffer(destination, maxLength); } }; public: template MOZ_IMPLICIT TwoByteString(T&& rhs) : Base(std::forward(rhs)) {} template TwoByteString& operator=(T&& rhs) { MOZ_ASSERT(this != &rhs, "self-move disallowed"); this->~TwoByteString(); new (this) TwoByteString(std::forward(rhs)); return *this; } TwoByteString(const TwoByteString&) = delete; TwoByteString& operator=(const TwoByteString&) = delete; // Rewrap the inner value of a JS::ubi::AtomOrTwoByteChars as a TwoByteString. static TwoByteString from(JS::ubi::AtomOrTwoByteChars&& s) { return s.match([](auto* a) { return TwoByteString(a); }); } // Returns true if the given TwoByteString is non-null, false otherwise. bool isNonNull() const { return match([](auto& t) { return t != nullptr; }); } // Return the length of the string, 0 if it is null. size_t length() const { return match( [](JSAtom* atom) -> size_t { MOZ_ASSERT(atom); JS::ubi::AtomOrTwoByteChars s(atom); return s.length(); }, [](const char16_t* chars) -> size_t { MOZ_ASSERT(chars); return NS_strlen(chars); }, [](const JS::ubi::EdgeName& ptr) -> size_t { MOZ_ASSERT(ptr); return NS_strlen(ptr.get()); }); } // Copy the contents of a TwoByteString into the provided buffer. The buffer // is NOT null terminated. The number of characters written is returned. size_t copyToBuffer(RangedPtr destination, size_t maxLength) { CopyToBufferMatcher m(destination, maxLength); return match(m); } struct HashPolicy; }; // A hashing policy for TwoByteString. // // Atoms are pointer hashed and use pointer equality, which means that we // tolerate some duplication across atoms and the other two types of two-byte // strings. In practice, we expect the amount of this duplication to be very low // because each type is generally a different semantic thing in addition to // having a slightly different representation. For example, the set of edge // names and the set stack frames' source names naturally tend not to overlap // very much if at all. struct TwoByteString::HashPolicy { using Lookup = TwoByteString; static js::HashNumber hash(const Lookup& l) { return l.match( [](const JSAtom* atom) { return js::DefaultHasher::hash(atom); }, [](const char16_t* chars) { MOZ_ASSERT(chars); auto length = NS_strlen(chars); return HashString(chars, length); }, [](const JS::ubi::EdgeName& ptr) { const char16_t* chars = ptr.get(); MOZ_ASSERT(chars); auto length = NS_strlen(chars); return HashString(chars, length); }); } struct EqualityMatcher { const TwoByteString& rhs; explicit EqualityMatcher(const TwoByteString& rhs) : rhs(rhs) {} bool operator()(const JSAtom* atom) { return rhs.is() && rhs.as() == atom; } bool operator()(const char16_t* chars) { MOZ_ASSERT(chars); const char16_t* rhsChars = nullptr; if (rhs.is()) rhsChars = rhs.as(); else if (rhs.is()) rhsChars = rhs.as().get(); else return false; MOZ_ASSERT(rhsChars); auto length = NS_strlen(chars); if (NS_strlen(rhsChars) != length) return false; return memcmp(chars, rhsChars, length * sizeof(char16_t)) == 0; } bool operator()(const JS::ubi::EdgeName& ptr) { MOZ_ASSERT(ptr); return operator()(ptr.get()); } }; static bool match(const TwoByteString& k, const Lookup& l) { EqualityMatcher eq(l); return k.match(eq); } static void rekey(TwoByteString& k, TwoByteString&& newKey) { k = std::move(newKey); } }; // Returns whether `edge` should be included in a heap snapshot of // `compartments`. The optional `policy` out-param is set to INCLUDE_EDGES // if we want to include the referent's edges, or EXCLUDE_EDGES if we don't // want to include them. static bool ShouldIncludeEdge(JS::CompartmentSet* compartments, const ubi::Node& origin, const ubi::Edge& edge, CoreDumpWriter::EdgePolicy* policy = nullptr) { if (policy) { *policy = CoreDumpWriter::INCLUDE_EDGES; } if (!compartments) { // We aren't targeting a particular set of compartments, so serialize all // the things! return true; } // We are targeting a particular set of compartments. If this node is in our // target set, serialize it and all of its edges. If this node is _not_ in our // target set, we also serialize under the assumption that it is a shared // resource being used by something in our target compartments since we // reached it by traversing the heap graph. However, we do not serialize its // outgoing edges and we abandon further traversal from this node. // // If the node does not belong to any compartment, we also serialize its // outgoing edges. This case is relevant for Shapes: they don't belong to a // specific compartment and contain edges to parent/kids Shapes we want to // include. Note that these Shapes may contain pointers into our target // compartment (the Shape's getter/setter JSObjects). However, we do not // serialize nodes in other compartments that are reachable from these // non-compartment nodes. JS::Compartment* compartment = edge.referent.compartment(); if (!compartment || compartments->has(compartment)) { return true; } if (policy) { *policy = CoreDumpWriter::EXCLUDE_EDGES; } return !!origin.compartment(); } // A `CoreDumpWriter` that serializes nodes to protobufs and writes them to the // given `ZeroCopyOutputStream`. class MOZ_STACK_CLASS StreamWriter : public CoreDumpWriter { using FrameSet = js::HashSet; using TwoByteStringMap = js::HashMap; using OneByteStringMap = js::HashMap; JSContext* cx; bool wantNames; // The set of |JS::ubi::StackFrame::identifier()|s that have already been // serialized and written to the core dump. FrameSet framesAlreadySerialized; // The set of two-byte strings that have already been serialized and written // to the core dump. TwoByteStringMap twoByteStringsAlreadySerialized; // The set of one-byte strings that have already been serialized and written // to the core dump. OneByteStringMap oneByteStringsAlreadySerialized; ::google::protobuf::io::ZeroCopyOutputStream& stream; JS::CompartmentSet* compartments; bool writeMessage(const ::google::protobuf::MessageLite& message) { // We have to create a new CodedOutputStream when writing each message so // that the 64MB size limit used by Coded{Output,Input}Stream to prevent // integer overflow is enforced per message rather than on the whole stream. ::google::protobuf::io::CodedOutputStream codedStream(&stream); codedStream.WriteVarint32(message.ByteSizeLong()); message.SerializeWithCachedSizes(&codedStream); return !codedStream.HadError(); } // Attach the full two-byte string or a reference to a two-byte string that // has already been serialized to a protobuf message. template bool attachTwoByteString(TwoByteString& string, SetStringFunction setString, SetRefFunction setRef) { auto ptr = twoByteStringsAlreadySerialized.lookupForAdd(string); if (ptr) { setRef(ptr->value()); return true; } auto length = string.length(); auto stringData = MakeUnique(length * sizeof(char16_t), '\0'); if (!stringData) return false; auto buf = const_cast( reinterpret_cast(stringData->data())); string.copyToBuffer(RangedPtr(buf, length), length); uint64_t ref = twoByteStringsAlreadySerialized.count(); if (!twoByteStringsAlreadySerialized.add(ptr, std::move(string), ref)) return false; setString(stringData.release()); return true; } // Attach the full one-byte string or a reference to a one-byte string that // has already been serialized to a protobuf message. template bool attachOneByteString(const char* string, SetStringFunction setString, SetRefFunction setRef) { auto ptr = oneByteStringsAlreadySerialized.lookupForAdd(string); if (ptr) { setRef(ptr->value()); return true; } auto length = strlen(string); auto stringData = MakeUnique(string, length); if (!stringData) return false; uint64_t ref = oneByteStringsAlreadySerialized.count(); if (!oneByteStringsAlreadySerialized.add(ptr, string, ref)) return false; setString(stringData.release()); return true; } protobuf::StackFrame* getProtobufStackFrame(JS::ubi::StackFrame& frame, size_t depth = 1) { // NB: de-duplicated string properties must be written in the same order // here as they are read in `HeapSnapshot::saveStackFrame` or else indices // in references to already serialized strings will be off. MOZ_ASSERT(frame, "null frames should be represented as the lack of a serialized " "stack frame"); auto id = frame.identifier(); auto protobufStackFrame = MakeUnique(); if (!protobufStackFrame) return nullptr; if (framesAlreadySerialized.has(id)) { protobufStackFrame->set_ref(id); return protobufStackFrame.release(); } auto data = MakeUnique(); if (!data) return nullptr; data->set_id(id); data->set_line(frame.line()); data->set_column(frame.column()); data->set_issystem(frame.isSystem()); data->set_isselfhosted(frame.isSelfHosted(cx)); auto dupeSource = TwoByteString::from(frame.source()); if (!attachTwoByteString( dupeSource, [&](std::string* source) { data->set_allocated_source(source); }, [&](uint64_t ref) { data->set_sourceref(ref); })) { return nullptr; } auto dupeName = TwoByteString::from(frame.functionDisplayName()); if (dupeName.isNonNull()) { if (!attachTwoByteString( dupeName, [&](std::string* name) { data->set_allocated_functiondisplayname(name); }, [&](uint64_t ref) { data->set_functiondisplaynameref(ref); })) { return nullptr; } } auto parent = frame.parent(); if (parent && depth < HeapSnapshot::MAX_STACK_DEPTH) { auto protobufParent = getProtobufStackFrame(parent, depth + 1); if (!protobufParent) return nullptr; data->set_allocated_parent(protobufParent); } protobufStackFrame->set_allocated_data(data.release()); if (!framesAlreadySerialized.put(id)) return nullptr; return protobufStackFrame.release(); } public: StreamWriter(JSContext* cx, ::google::protobuf::io::ZeroCopyOutputStream& stream, bool wantNames, JS::CompartmentSet* compartments) : cx(cx), wantNames(wantNames), framesAlreadySerialized(cx), twoByteStringsAlreadySerialized(cx), oneByteStringsAlreadySerialized(cx), stream(stream), compartments(compartments) {} ~StreamWriter() override {} bool writeMetadata(uint64_t timestamp) final { protobuf::Metadata metadata; metadata.set_timestamp(timestamp); return writeMessage(metadata); } bool writeNode(const JS::ubi::Node& ubiNode, EdgePolicy includeEdges) final { // NB: de-duplicated string properties must be written in the same order // here as they are read in `HeapSnapshot::saveNode` or else indices in // references to already serialized strings will be off. protobuf::Node protobufNode; protobufNode.set_id(ubiNode.identifier()); protobufNode.set_coarsetype( JS::ubi::CoarseTypeToUint32(ubiNode.coarseType())); auto typeName = TwoByteString(ubiNode.typeName()); if (NS_WARN_IF(!attachTwoByteString( typeName, [&](std::string* name) { protobufNode.set_allocated_typename_(name); }, [&](uint64_t ref) { protobufNode.set_typenameref(ref); }))) { return false; } mozilla::MallocSizeOf mallocSizeOf = dbg::GetDebuggerMallocSizeOf(cx); MOZ_ASSERT(mallocSizeOf); protobufNode.set_size(ubiNode.size(mallocSizeOf)); if (includeEdges) { auto edges = ubiNode.edges(cx, wantNames); if (NS_WARN_IF(!edges)) return false; for (; !edges->empty(); edges->popFront()) { ubi::Edge& ubiEdge = edges->front(); if (!ShouldIncludeEdge(compartments, ubiNode, ubiEdge)) { continue; } protobuf::Edge* protobufEdge = protobufNode.add_edges(); if (NS_WARN_IF(!protobufEdge)) { return false; } protobufEdge->set_referent(ubiEdge.referent.identifier()); if (wantNames && ubiEdge.name) { TwoByteString edgeName(std::move(ubiEdge.name)); if (NS_WARN_IF(!attachTwoByteString( edgeName, [&](std::string* name) { protobufEdge->set_allocated_name(name); }, [&](uint64_t ref) { protobufEdge->set_nameref(ref); }))) { return false; } } } } if (ubiNode.hasAllocationStack()) { auto ubiStackFrame = ubiNode.allocationStack(); auto protoStackFrame = getProtobufStackFrame(ubiStackFrame); if (NS_WARN_IF(!protoStackFrame)) return false; protobufNode.set_allocated_allocationstack(protoStackFrame); } if (auto className = ubiNode.jsObjectClassName()) { if (NS_WARN_IF(!attachOneByteString( className, [&](std::string* name) { protobufNode.set_allocated_jsobjectclassname(name); }, [&](uint64_t ref) { protobufNode.set_jsobjectclassnameref(ref); }))) { return false; } } if (auto scriptFilename = ubiNode.scriptFilename()) { if (NS_WARN_IF(!attachOneByteString( scriptFilename, [&](std::string* name) { protobufNode.set_allocated_scriptfilename(name); }, [&](uint64_t ref) { protobufNode.set_scriptfilenameref(ref); }))) { return false; } } if (ubiNode.descriptiveTypeName()) { auto descriptiveTypeName = TwoByteString(ubiNode.descriptiveTypeName()); if (NS_WARN_IF(!attachTwoByteString( descriptiveTypeName, [&](std::string* name) { protobufNode.set_allocated_descriptivetypename(name); }, [&](uint64_t ref) { protobufNode.set_descriptivetypenameref(ref); }))) { return false; } } return writeMessage(protobufNode); } }; // A JS::ubi::BreadthFirst handler that serializes a snapshot of the heap into a // core dump. class MOZ_STACK_CLASS HeapSnapshotHandler { CoreDumpWriter& writer; JS::CompartmentSet* compartments; public: // For telemetry. uint32_t nodeCount; uint32_t edgeCount; HeapSnapshotHandler(CoreDumpWriter& writer, JS::CompartmentSet* compartments) : writer(writer), compartments(compartments), nodeCount(0), edgeCount(0) {} // JS::ubi::BreadthFirst handler interface. class NodeData {}; typedef JS::ubi::BreadthFirst Traversal; bool operator()(Traversal& traversal, JS::ubi::Node origin, const JS::ubi::Edge& edge, NodeData*, bool first) { edgeCount++; // We're only interested in the first time we reach edge.referent, not in // every edge arriving at that node. "But, don't we want to serialize every // edge in the heap graph?" you ask. Don't worry! This edge is still // serialized into the core dump. Serializing a node also serializes each of // its edges, and if we are traversing a given edge, we must have already // visited and serialized the origin node and its edges. if (!first) return true; CoreDumpWriter::EdgePolicy policy; if (!ShouldIncludeEdge(compartments, origin, edge, &policy)) { // Because ShouldIncludeEdge considers the |origin| node as well, we don't // want to consider this node 'visited' until we write it to the core // dump. traversal.doNotMarkReferentAsVisited(); return true; } nodeCount++; if (policy == CoreDumpWriter::EXCLUDE_EDGES) traversal.abandonReferent(); return writer.writeNode(edge.referent, policy); } }; bool WriteHeapGraph(JSContext* cx, const JS::ubi::Node& node, CoreDumpWriter& writer, bool wantNames, JS::CompartmentSet* compartments, JS::AutoCheckCannotGC& noGC, uint32_t& outNodeCount, uint32_t& outEdgeCount) { // Serialize the starting node to the core dump. if (NS_WARN_IF(!writer.writeNode(node, CoreDumpWriter::INCLUDE_EDGES))) { return false; } // Walk the heap graph starting from the given node and serialize it into the // core dump. HeapSnapshotHandler handler(writer, compartments); HeapSnapshotHandler::Traversal traversal(cx, handler, noGC); traversal.wantNames = wantNames; bool ok = traversal.addStartVisited(node) && traversal.traverse(); if (ok) { outNodeCount = handler.nodeCount; outEdgeCount = handler.edgeCount; } return ok; } static unsigned long msSinceProcessCreation(const TimeStamp& now) { auto duration = now - TimeStamp::ProcessCreation(); return (unsigned long)duration.ToMilliseconds(); } /* static */ already_AddRefed HeapSnapshot::CreateUniqueCoreDumpFile( ErrorResult& rv, const TimeStamp& now, nsAString& outFilePath, nsAString& outSnapshotId) { MOZ_RELEASE_ASSERT(XRE_IsParentProcess()); nsCOMPtr file; rv = GetSpecialSystemDirectory(OS_TemporaryDirectory, getter_AddRefs(file)); if (NS_WARN_IF(rv.Failed())) return nullptr; nsAutoString tempPath; rv = file->GetPath(tempPath); if (NS_WARN_IF(rv.Failed())) return nullptr; auto ms = msSinceProcessCreation(now); rv = file->AppendNative(nsPrintfCString("%lu.fxsnapshot", ms)); if (NS_WARN_IF(rv.Failed())) return nullptr; rv = file->CreateUnique(nsIFile::NORMAL_FILE_TYPE, 0666); if (NS_WARN_IF(rv.Failed())) return nullptr; rv = file->GetPath(outFilePath); if (NS_WARN_IF(rv.Failed())) return nullptr; // The snapshot ID must be computed in the process that created the // temp file, because TmpD may not be the same in all processes. outSnapshotId.Assign(Substring( outFilePath, tempPath.Length() + 1, outFilePath.Length() - tempPath.Length() - sizeof(".fxsnapshot"))); return file.forget(); } // Deletion policy for cleaning up PHeapSnapshotTempFileHelperChild pointers. class DeleteHeapSnapshotTempFileHelperChild { public: constexpr DeleteHeapSnapshotTempFileHelperChild() {} void operator()(PHeapSnapshotTempFileHelperChild* ptr) const { Unused << NS_WARN_IF(!HeapSnapshotTempFileHelperChild::Send__delete__(ptr)); } }; // A UniquePtr alias to automatically manage PHeapSnapshotTempFileHelperChild // pointers. using UniqueHeapSnapshotTempFileHelperChild = UniquePtr; // Get an nsIOutputStream that we can write the heap snapshot to. In non-e10s // and in the e10s parent process, open a file directly and create an output // stream for it. In e10s child processes, we are sandboxed without access to // the filesystem. Use IPDL to request a file descriptor from the parent // process. static already_AddRefed getCoreDumpOutputStream( ErrorResult& rv, TimeStamp& start, nsAString& outFilePath, nsAString& outSnapshotId) { if (XRE_IsParentProcess()) { // Create the file and open the output stream directly. nsCOMPtr file = HeapSnapshot::CreateUniqueCoreDumpFile( rv, start, outFilePath, outSnapshotId); if (NS_WARN_IF(rv.Failed())) return nullptr; nsCOMPtr outputStream; rv = NS_NewLocalFileOutputStream(getter_AddRefs(outputStream), file, PR_WRONLY, -1, 0); if (NS_WARN_IF(rv.Failed())) return nullptr; return outputStream.forget(); } // Request a file descriptor from the parent process over IPDL. auto cc = ContentChild::GetSingleton(); if (!cc) { rv.Throw(NS_ERROR_UNEXPECTED); return nullptr; } UniqueHeapSnapshotTempFileHelperChild helper( cc->SendPHeapSnapshotTempFileHelperConstructor()); if (NS_WARN_IF(!helper)) { rv.Throw(NS_ERROR_UNEXPECTED); return nullptr; } OpenHeapSnapshotTempFileResponse response; if (!helper->SendOpenHeapSnapshotTempFile(&response)) { rv.Throw(NS_ERROR_UNEXPECTED); return nullptr; } if (response.type() == OpenHeapSnapshotTempFileResponse::Tnsresult) { rv.Throw(response.get_nsresult()); return nullptr; } auto opened = response.get_OpenedFile(); outFilePath = opened.path(); outSnapshotId = opened.snapshotId(); nsCOMPtr outputStream = FileDescriptorOutputStream::Create(opened.descriptor()); if (NS_WARN_IF(!outputStream)) { rv.Throw(NS_ERROR_UNEXPECTED); return nullptr; } return outputStream.forget(); } } // namespace devtools namespace dom { using namespace JS; using namespace devtools; /* static */ void ChromeUtils::SaveHeapSnapshotShared( GlobalObject& global, const HeapSnapshotBoundaries& boundaries, nsAString& outFilePath, nsAString& outSnapshotId, ErrorResult& rv) { auto start = TimeStamp::Now(); bool wantNames = true; CompartmentSet compartments; uint32_t nodeCount = 0; uint32_t edgeCount = 0; nsCOMPtr outputStream = getCoreDumpOutputStream(rv, start, outFilePath, outSnapshotId); if (NS_WARN_IF(rv.Failed())) return; ZeroCopyNSIOutputStream zeroCopyStream(outputStream); ::google::protobuf::io::GzipOutputStream gzipStream(&zeroCopyStream); JSContext* cx = global.Context(); { ubi::RootList rootList(cx, wantNames); auto [ok, nogc] = EstablishBoundaries(cx, rv, boundaries, rootList, compartments); if (!ok) { return; } StreamWriter writer(cx, gzipStream, wantNames, !compartments.empty() ? &compartments : nullptr); ubi::Node roots(&rootList); // Serialize the initial heap snapshot metadata to the core dump. if (!writer.writeMetadata(PR_Now()) || // Serialize the heap graph to the core dump, starting from our list of // roots. !WriteHeapGraph(cx, roots, writer, wantNames, !compartments.empty() ? &compartments : nullptr, nogc, nodeCount, edgeCount)) { rv.Throw(zeroCopyStream.failed() ? zeroCopyStream.result() : NS_ERROR_UNEXPECTED); return; } } Telemetry::AccumulateTimeDelta(Telemetry::DEVTOOLS_SAVE_HEAP_SNAPSHOT_MS, start); Telemetry::Accumulate(Telemetry::DEVTOOLS_HEAP_SNAPSHOT_NODE_COUNT, nodeCount); Telemetry::Accumulate(Telemetry::DEVTOOLS_HEAP_SNAPSHOT_EDGE_COUNT, edgeCount); } /* static */ uint64_t ChromeUtils::GetObjectNodeId(GlobalObject& global, JS::Handle val) { JS::Rooted obj(global.Context(), val); JS::ubi::Node node(obj); return node.identifier(); } /* static */ void ChromeUtils::SaveHeapSnapshot(GlobalObject& global, const HeapSnapshotBoundaries& boundaries, nsAString& outFilePath, ErrorResult& rv) { nsAutoString snapshotId; SaveHeapSnapshotShared(global, boundaries, outFilePath, snapshotId, rv); } /* static */ void ChromeUtils::SaveHeapSnapshotGetId( GlobalObject& global, const HeapSnapshotBoundaries& boundaries, nsAString& outSnapshotId, ErrorResult& rv) { nsAutoString filePath; SaveHeapSnapshotShared(global, boundaries, filePath, outSnapshotId, rv); } /* static */ already_AddRefed ChromeUtils::ReadHeapSnapshot( GlobalObject& global, const nsAString& filePath, ErrorResult& rv) { auto start = TimeStamp::Now(); nsresult nsrv; nsCOMPtr snapshotFile = do_CreateInstance("@mozilla.org/file/local;1", &nsrv); if (NS_FAILED(nsrv)) { rv = nsrv; return nullptr; } rv = snapshotFile->InitWithPath(filePath); if (rv.Failed()) { return nullptr; } AutoMemMap mm; rv = mm.init(snapshotFile); if (rv.Failed()) return nullptr; RefPtr snapshot = HeapSnapshot::Create( global.Context(), global, reinterpret_cast(mm.address()), mm.size(), rv); if (!rv.Failed()) Telemetry::AccumulateTimeDelta(Telemetry::DEVTOOLS_READ_HEAP_SNAPSHOT_MS, start); return snapshot.forget(); } } // namespace dom } // namespace mozilla