diff options
Diffstat (limited to 'netwerk/base/Predictor.cpp')
-rw-r--r-- | netwerk/base/Predictor.cpp | 2445 |
1 files changed, 2445 insertions, 0 deletions
diff --git a/netwerk/base/Predictor.cpp b/netwerk/base/Predictor.cpp new file mode 100644 index 0000000000..95aeef06f4 --- /dev/null +++ b/netwerk/base/Predictor.cpp @@ -0,0 +1,2445 @@ +/* vim: set ts=2 sts=2 et sw=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 <algorithm> + +#include "Predictor.h" + +#include "nsAppDirectoryServiceDefs.h" +#include "nsICacheStorage.h" +#include "nsICachingChannel.h" +#include "nsICancelable.h" +#include "nsIChannel.h" +#include "nsContentUtils.h" +#include "nsIDNSService.h" +#include "mozilla/dom/Document.h" +#include "nsIFile.h" +#include "nsIHttpChannel.h" +#include "nsIInputStream.h" +#include "nsILoadContext.h" +#include "nsILoadContextInfo.h" +#include "nsILoadGroup.h" +#include "nsINetworkPredictorVerifier.h" +#include "nsIObserverService.h" +#include "nsISpeculativeConnect.h" +#include "nsITimer.h" +#include "nsIURI.h" +#include "nsNetUtil.h" +#include "nsServiceManagerUtils.h" +#include "nsStreamUtils.h" +#include "nsString.h" +#include "nsThreadUtils.h" +#include "mozilla/Logging.h" + +#include "mozilla/OriginAttributes.h" +#include "mozilla/Preferences.h" +#include "mozilla/SchedulerGroup.h" +#include "mozilla/StaticPrefs_network.h" +#include "mozilla/Telemetry.h" + +#include "mozilla/net/NeckoCommon.h" +#include "mozilla/net/NeckoParent.h" + +#include "LoadContextInfo.h" +#include "mozilla/ipc/URIUtils.h" +#include "SerializedLoadContext.h" +#include "mozilla/net/NeckoChild.h" + +#include "mozilla/dom/ContentParent.h" +#include "mozilla/ClearOnShutdown.h" + +#include "CacheControlParser.h" +#include "ReferrerInfo.h" + +using namespace mozilla; + +namespace mozilla { +namespace net { + +Predictor* Predictor::sSelf = nullptr; + +static LazyLogModule gPredictorLog("NetworkPredictor"); + +#define PREDICTOR_LOG(args) \ + MOZ_LOG(gPredictorLog, mozilla::LogLevel::Debug, args) + +#define NOW_IN_SECONDS() static_cast<uint32_t>(PR_Now() / PR_USEC_PER_SEC) + +// All these time values are in sec +static const uint32_t ONE_DAY = 86400U; +static const uint32_t ONE_WEEK = 7U * ONE_DAY; +static const uint32_t ONE_MONTH = 30U * ONE_DAY; +static const uint32_t ONE_YEAR = 365U * ONE_DAY; + +static const uint32_t STARTUP_WINDOW = 5U * 60U; // 5min + +// Version of metadata entries we expect +static const uint32_t METADATA_VERSION = 1; + +// Flags available in entries +// FLAG_PREFETCHABLE - we have determined that this item is eligible for +// prefetch +static const uint32_t FLAG_PREFETCHABLE = 1 << 0; + +// We save 12 bits in the "flags" section of our metadata for actual flags, the +// rest are to keep track of a rolling count of which loads a resource has been +// used on to determine if we can prefetch that resource or not; +static const uint8_t kRollingLoadOffset = 12; +static const int32_t kMaxPrefetchRollingLoadCount = 20; +static const uint32_t kFlagsMask = ((1 << kRollingLoadOffset) - 1); + +// ID Extensions for cache entries +#define PREDICTOR_ORIGIN_EXTENSION "predictor-origin" + +// Get the full origin (scheme, host, port) out of a URI (maybe should be part +// of nsIURI instead?) +static nsresult ExtractOrigin(nsIURI* uri, nsIURI** originUri) { + nsAutoCString s; + nsresult rv = nsContentUtils::GetASCIIOrigin(uri, s); + NS_ENSURE_SUCCESS(rv, rv); + + return NS_NewURI(originUri, s); +} + +// All URIs we get passed *must* be http or https if they're not null. This +// helps ensure that. +static bool IsNullOrHttp(nsIURI* uri) { + if (!uri) { + return true; + } + + return uri->SchemeIs("http") || uri->SchemeIs("https"); +} + +// Listener for the speculative DNS requests we'll fire off, which just ignores +// the result (since we're just trying to warm the cache). This also exists to +// reduce round-trips to the main thread, by being something threadsafe the +// Predictor can use. + +NS_IMPL_ISUPPORTS(Predictor::DNSListener, nsIDNSListener); + +NS_IMETHODIMP +Predictor::DNSListener::OnLookupComplete(nsICancelable* request, + nsIDNSRecord* rec, nsresult status) { + return NS_OK; +} + +// Class to proxy important information from the initial predictor call through +// the cache API and back into the internals of the predictor. We can't use the +// predictor itself, as it may have multiple actions in-flight, and each action +// has different parameters. +NS_IMPL_ISUPPORTS(Predictor::Action, nsICacheEntryOpenCallback); + +Predictor::Action::Action(bool fullUri, bool predict, Predictor::Reason reason, + nsIURI* targetURI, nsIURI* sourceURI, + nsINetworkPredictorVerifier* verifier, + Predictor* predictor) + : mFullUri(fullUri), + mPredict(predict), + mTargetURI(targetURI), + mSourceURI(sourceURI), + mVerifier(verifier), + mStackCount(0), + mPredictor(predictor) { + mStartTime = TimeStamp::Now(); + if (mPredict) { + mPredictReason = reason.mPredict; + } else { + mLearnReason = reason.mLearn; + } +} + +Predictor::Action::Action(bool fullUri, bool predict, Predictor::Reason reason, + nsIURI* targetURI, nsIURI* sourceURI, + nsINetworkPredictorVerifier* verifier, + Predictor* predictor, uint8_t stackCount) + : mFullUri(fullUri), + mPredict(predict), + mTargetURI(targetURI), + mSourceURI(sourceURI), + mVerifier(verifier), + mStackCount(stackCount), + mPredictor(predictor) { + mStartTime = TimeStamp::Now(); + if (mPredict) { + mPredictReason = reason.mPredict; + } else { + mLearnReason = reason.mLearn; + } +} + +NS_IMETHODIMP +Predictor::Action::OnCacheEntryCheck(nsICacheEntry* entry, + nsIApplicationCache* appCache, + uint32_t* result) { + *result = nsICacheEntryOpenCallback::ENTRY_WANTED; + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Action::OnCacheEntryAvailable(nsICacheEntry* entry, bool isNew, + nsIApplicationCache* appCache, + nsresult result) { + MOZ_ASSERT(NS_IsMainThread(), "Got cache entry off main thread!"); + + nsAutoCString targetURI, sourceURI; + mTargetURI->GetAsciiSpec(targetURI); + if (mSourceURI) { + mSourceURI->GetAsciiSpec(sourceURI); + } + PREDICTOR_LOG( + ("OnCacheEntryAvailable %p called. entry=%p mFullUri=%d mPredict=%d " + "mPredictReason=%d mLearnReason=%d mTargetURI=%s " + "mSourceURI=%s mStackCount=%d isNew=%d result=0x%08" PRIx32, + this, entry, mFullUri, mPredict, mPredictReason, mLearnReason, + targetURI.get(), sourceURI.get(), mStackCount, isNew, + static_cast<uint32_t>(result))); + if (NS_FAILED(result)) { + PREDICTOR_LOG( + ("OnCacheEntryAvailable %p FAILED to get cache entry (0x%08" PRIX32 + "). Aborting.", + this, static_cast<uint32_t>(result))); + return NS_OK; + } + Telemetry::AccumulateTimeDelta(Telemetry::PREDICTOR_WAIT_TIME, mStartTime); + if (mPredict) { + bool predicted = + mPredictor->PredictInternal(mPredictReason, entry, isNew, mFullUri, + mTargetURI, mVerifier, mStackCount); + Telemetry::AccumulateTimeDelta(Telemetry::PREDICTOR_PREDICT_WORK_TIME, + mStartTime); + if (predicted) { + Telemetry::AccumulateTimeDelta( + Telemetry::PREDICTOR_PREDICT_TIME_TO_ACTION, mStartTime); + } else { + Telemetry::AccumulateTimeDelta( + Telemetry::PREDICTOR_PREDICT_TIME_TO_INACTION, mStartTime); + } + } else { + mPredictor->LearnInternal(mLearnReason, entry, isNew, mFullUri, mTargetURI, + mSourceURI); + Telemetry::AccumulateTimeDelta(Telemetry::PREDICTOR_LEARN_WORK_TIME, + mStartTime); + } + + return NS_OK; +} + +NS_IMPL_ISUPPORTS(Predictor, nsINetworkPredictor, nsIObserver, + nsISpeculativeConnectionOverrider, nsIInterfaceRequestor, + nsICacheEntryMetaDataVisitor, nsINetworkPredictorVerifier) + +Predictor::Predictor() + : mInitialized(false), + mStartupTime(0), + mLastStartupTime(0), + mStartupCount(1) { + MOZ_ASSERT(!sSelf, "multiple Predictor instances!"); + sSelf = this; +} + +Predictor::~Predictor() { + if (mInitialized) Shutdown(); + + sSelf = nullptr; +} + +// Predictor::nsIObserver + +nsresult Predictor::InstallObserver() { + MOZ_ASSERT(NS_IsMainThread(), "Installing observer off main thread"); + + nsresult rv = NS_OK; + nsCOMPtr<nsIObserverService> obs = mozilla::services::GetObserverService(); + if (!obs) { + return NS_ERROR_NOT_AVAILABLE; + } + + rv = obs->AddObserver(this, NS_XPCOM_SHUTDOWN_OBSERVER_ID, false); + NS_ENSURE_SUCCESS(rv, rv); + + return rv; +} + +void Predictor::RemoveObserver() { + MOZ_ASSERT(NS_IsMainThread(), "Removing observer off main thread"); + + nsCOMPtr<nsIObserverService> obs = mozilla::services::GetObserverService(); + if (obs) { + obs->RemoveObserver(this, NS_XPCOM_SHUTDOWN_OBSERVER_ID); + } +} + +NS_IMETHODIMP +Predictor::Observe(nsISupports* subject, const char* topic, + const char16_t* data_unicode) { + nsresult rv = NS_OK; + MOZ_ASSERT(NS_IsMainThread(), + "Predictor observing something off main thread!"); + + if (!strcmp(NS_XPCOM_SHUTDOWN_OBSERVER_ID, topic)) { + Shutdown(); + } + + return rv; +} + +// Predictor::nsISpeculativeConnectionOverrider + +NS_IMETHODIMP +Predictor::GetIgnoreIdle(bool* ignoreIdle) { + *ignoreIdle = true; + return NS_OK; +} + +NS_IMETHODIMP +Predictor::GetParallelSpeculativeConnectLimit( + uint32_t* parallelSpeculativeConnectLimit) { + *parallelSpeculativeConnectLimit = 6; + return NS_OK; +} + +NS_IMETHODIMP +Predictor::GetIsFromPredictor(bool* isFromPredictor) { + *isFromPredictor = true; + return NS_OK; +} + +NS_IMETHODIMP +Predictor::GetAllow1918(bool* allow1918) { + *allow1918 = false; + return NS_OK; +} + +// Predictor::nsIInterfaceRequestor + +NS_IMETHODIMP +Predictor::GetInterface(const nsIID& iid, void** result) { + return QueryInterface(iid, result); +} + +// Predictor::nsICacheEntryMetaDataVisitor + +#define SEEN_META_DATA "predictor::seen" +#define RESOURCE_META_DATA "predictor::resource-count" +#define META_DATA_PREFIX "predictor::" + +static bool IsURIMetadataElement(const char* key) { + return StringBeginsWith(nsDependentCString(key), + nsLiteralCString(META_DATA_PREFIX)) && + !nsLiteralCString(SEEN_META_DATA).Equals(key) && + !nsLiteralCString(RESOURCE_META_DATA).Equals(key); +} + +nsresult Predictor::OnMetaDataElement(const char* asciiKey, + const char* asciiValue) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsURIMetadataElement(asciiKey)) { + // This isn't a bit of metadata we care about + return NS_OK; + } + + nsCString key, value; + key.AssignASCII(asciiKey); + value.AssignASCII(asciiValue); + mKeysToOperateOn.AppendElement(key); + mValuesToOperateOn.AppendElement(value); + + return NS_OK; +} + +// Predictor::nsINetworkPredictor + +nsresult Predictor::Init() { + MOZ_DIAGNOSTIC_ASSERT(!IsNeckoChild()); + + if (!NS_IsMainThread()) { + MOZ_ASSERT(false, "Predictor::Init called off the main thread!"); + return NS_ERROR_UNEXPECTED; + } + + nsresult rv = NS_OK; + + rv = InstallObserver(); + NS_ENSURE_SUCCESS(rv, rv); + + mLastStartupTime = mStartupTime = NOW_IN_SECONDS(); + + if (!mDNSListener) { + mDNSListener = new DNSListener(); + } + + mCacheStorageService = + do_GetService("@mozilla.org/netwerk/cache-storage-service;1", &rv); + NS_ENSURE_SUCCESS(rv, rv); + + mSpeculativeService = do_GetService("@mozilla.org/network/io-service;1", &rv); + NS_ENSURE_SUCCESS(rv, rv); + + rv = NS_NewURI(getter_AddRefs(mStartupURI), "predictor://startup"); + NS_ENSURE_SUCCESS(rv, rv); + + mDnsService = do_GetService("@mozilla.org/network/dns-service;1", &rv); + NS_ENSURE_SUCCESS(rv, rv); + + mInitialized = true; + + return rv; +} + +namespace { +class PredictorLearnRunnable final : public Runnable { + public: + PredictorLearnRunnable(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, + const OriginAttributes& oa) + : Runnable("PredictorLearnRunnable"), + mTargetURI(targetURI), + mSourceURI(sourceURI), + mReason(reason), + mOA(oa) { + MOZ_DIAGNOSTIC_ASSERT(targetURI, "Must have a target URI"); + } + + ~PredictorLearnRunnable() = default; + + NS_IMETHOD Run() override { + if (!gNeckoChild) { + // This may have gone away between when this runnable was dispatched and + // when it actually runs, so let's be safe here, even though we asserted + // earlier. + PREDICTOR_LOG(("predictor::learn (async) gNeckoChild went away")); + return NS_OK; + } + + PREDICTOR_LOG(("predictor::learn (async) forwarding to parent")); + gNeckoChild->SendPredLearn(mTargetURI, mSourceURI, mReason, mOA); + + return NS_OK; + } + + private: + nsCOMPtr<nsIURI> mTargetURI; + nsCOMPtr<nsIURI> mSourceURI; + PredictorLearnReason mReason; + const OriginAttributes mOA; +}; + +} // namespace + +void Predictor::Shutdown() { + if (!NS_IsMainThread()) { + MOZ_ASSERT(false, "Predictor::Shutdown called off the main thread!"); + return; + } + + RemoveObserver(); + + mInitialized = false; +} + +nsresult Predictor::Create(nsISupports* aOuter, const nsIID& aIID, + void** aResult) { + MOZ_ASSERT(NS_IsMainThread()); + + nsresult rv; + + if (aOuter != nullptr) { + return NS_ERROR_NO_AGGREGATION; + } + + RefPtr<Predictor> svc = new Predictor(); + if (IsNeckoChild()) { + NeckoChild::InitNeckoChild(); + + // Child threads only need to be call into the public interface methods + // so we don't bother with initialization + return svc->QueryInterface(aIID, aResult); + } + + rv = svc->Init(); + if (NS_FAILED(rv)) { + PREDICTOR_LOG(("Failed to initialize predictor, predictor will be a noop")); + } + + // We treat init failure the same as the service being disabled, since this + // is all an optimization anyway. No need to freak people out. That's why we + // gladly continue on QI'ing here. + rv = svc->QueryInterface(aIID, aResult); + + return rv; +} + +NS_IMETHODIMP +Predictor::Predict(nsIURI* targetURI, nsIURI* sourceURI, + PredictorPredictReason reason, + JS::HandleValue originAttributes, + nsINetworkPredictorVerifier* verifier, JSContext* aCx) { + OriginAttributes attrs; + + if (!originAttributes.isObject() || !attrs.Init(aCx, originAttributes)) { + return NS_ERROR_INVALID_ARG; + } + + return PredictNative(targetURI, sourceURI, reason, attrs, verifier); +} + +// Called from the main thread to initiate predictive actions +NS_IMETHODIMP +Predictor::PredictNative(nsIURI* targetURI, nsIURI* sourceURI, + PredictorPredictReason reason, + const OriginAttributes& originAttributes, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread(), + "Predictor interface methods must be called on the main thread"); + + PREDICTOR_LOG(("Predictor::Predict")); + + if (IsNeckoChild()) { + MOZ_DIAGNOSTIC_ASSERT(gNeckoChild); + + PREDICTOR_LOG((" called on child process")); + // If two different threads are predicting concurently, this will be + // overwritten. Thankfully, we only use this in tests, which will + // overwrite mVerifier perhaps multiple times for each individual test; + // however, within each test, the multiple predict calls should have the + // same verifier. + if (verifier) { + PREDICTOR_LOG((" was given a verifier")); + mChildVerifier = verifier; + } + PREDICTOR_LOG((" forwarding to parent process")); + gNeckoChild->SendPredPredict(targetURI, sourceURI, reason, originAttributes, + verifier); + return NS_OK; + } + + PREDICTOR_LOG((" called on parent process")); + + if (!mInitialized) { + PREDICTOR_LOG((" not initialized")); + return NS_OK; + } + + if (!StaticPrefs::network_predictor_enabled()) { + PREDICTOR_LOG((" not enabled")); + return NS_OK; + } + + if (originAttributes.mPrivateBrowsingId > 0) { + // Don't want to do anything in PB mode + PREDICTOR_LOG((" in PB mode")); + return NS_OK; + } + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + // Nothing we can do for non-HTTP[S] schemes + PREDICTOR_LOG((" got non-http[s] URI")); + return NS_OK; + } + + // Ensure we've been given the appropriate arguments for the kind of + // prediction we're being asked to do + nsCOMPtr<nsIURI> uriKey = targetURI; + nsCOMPtr<nsIURI> originKey; + switch (reason) { + case nsINetworkPredictor::PREDICT_LINK: + if (!targetURI || !sourceURI) { + PREDICTOR_LOG((" link invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + // Link hover is a special case where we can predict without hitting the + // db, so let's go ahead and fire off that prediction here. + PredictForLink(targetURI, sourceURI, originAttributes, verifier); + return NS_OK; + case nsINetworkPredictor::PREDICT_LOAD: + if (!targetURI || sourceURI) { + PREDICTOR_LOG((" load invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + break; + case nsINetworkPredictor::PREDICT_STARTUP: + if (targetURI || sourceURI) { + PREDICTOR_LOG((" startup invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + uriKey = mStartupURI; + originKey = mStartupURI; + break; + default: + PREDICTOR_LOG((" invalid reason")); + return NS_ERROR_INVALID_ARG; + } + + Predictor::Reason argReason; + argReason.mPredict = reason; + + // First we open the regular cache entry, to ensure we don't gum up the works + // waiting on the less-important predictor-only cache entry + RefPtr<Predictor::Action> uriAction = new Predictor::Action( + Predictor::Action::IS_FULL_URI, Predictor::Action::DO_PREDICT, argReason, + targetURI, nullptr, verifier, this); + nsAutoCString uriKeyStr; + uriKey->GetAsciiSpec(uriKeyStr); + PREDICTOR_LOG((" Predict uri=%s reason=%d action=%p", uriKeyStr.get(), + reason, uriAction.get())); + + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + RefPtr<LoadContextInfo> lci = new LoadContextInfo(false, originAttributes); + + nsresult rv = mCacheStorageService->DiskCacheStorage( + lci, false, getter_AddRefs(cacheDiskStorage)); + NS_ENSURE_SUCCESS(rv, rv); + + uint32_t openFlags = + nsICacheStorage::OPEN_READONLY | nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::OPEN_PRIORITY | nsICacheStorage::CHECK_MULTITHREADED; + cacheDiskStorage->AsyncOpenURI(uriKey, ""_ns, openFlags, uriAction); + + // Now we do the origin-only (and therefore predictor-only) entry + nsCOMPtr<nsIURI> targetOrigin; + rv = ExtractOrigin(uriKey, getter_AddRefs(targetOrigin)); + NS_ENSURE_SUCCESS(rv, rv); + if (!originKey) { + originKey = targetOrigin; + } + + RefPtr<Predictor::Action> originAction = new Predictor::Action( + Predictor::Action::IS_ORIGIN, Predictor::Action::DO_PREDICT, argReason, + targetOrigin, nullptr, verifier, this); + nsAutoCString originKeyStr; + originKey->GetAsciiSpec(originKeyStr); + PREDICTOR_LOG((" Predict origin=%s reason=%d action=%p", + originKeyStr.get(), reason, originAction.get())); + openFlags = nsICacheStorage::OPEN_READONLY | nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::CHECK_MULTITHREADED; + cacheDiskStorage->AsyncOpenURI(originKey, + nsLiteralCString(PREDICTOR_ORIGIN_EXTENSION), + openFlags, originAction); + + PREDICTOR_LOG((" predict returning")); + return NS_OK; +} + +bool Predictor::PredictInternal(PredictorPredictReason reason, + nsICacheEntry* entry, bool isNew, bool fullUri, + nsIURI* targetURI, + nsINetworkPredictorVerifier* verifier, + uint8_t stackCount) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::PredictInternal")); + bool rv = false; + + nsCOMPtr<nsILoadContextInfo> lci; + entry->GetLoadContextInfo(getter_AddRefs(lci)); + + if (!lci) { + return rv; + } + + if (reason == nsINetworkPredictor::PREDICT_LOAD) { + MaybeLearnForStartup(targetURI, fullUri, *lci->OriginAttributesPtr()); + } + + if (isNew) { + // nothing else we can do here + PREDICTOR_LOG((" new entry")); + return rv; + } + + switch (reason) { + case nsINetworkPredictor::PREDICT_LOAD: + rv = PredictForPageload(entry, targetURI, stackCount, fullUri, verifier); + break; + case nsINetworkPredictor::PREDICT_STARTUP: + rv = PredictForStartup(entry, fullUri, verifier); + break; + default: + PREDICTOR_LOG((" invalid reason")); + MOZ_ASSERT(false, "Got unexpected value for prediction reason"); + } + + return rv; +} + +void Predictor::PredictForLink(nsIURI* targetURI, nsIURI* sourceURI, + const OriginAttributes& originAttributes, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::PredictForLink")); + if (!mSpeculativeService) { + PREDICTOR_LOG((" missing speculative service")); + return; + } + + if (!StaticPrefs::network_predictor_enable_hover_on_ssl()) { + if (sourceURI->SchemeIs("https")) { + // We don't want to predict from an HTTPS page, to avoid info leakage + PREDICTOR_LOG((" Not predicting for link hover - on an SSL page")); + return; + } + } + + nsCOMPtr<nsIPrincipal> principal = + BasePrincipal::CreateContentPrincipal(targetURI, originAttributes); + + mSpeculativeService->SpeculativeConnect(targetURI, principal, nullptr); + if (verifier) { + PREDICTOR_LOG((" sending verification")); + verifier->OnPredictPreconnect(targetURI); + } +} + +// This is the driver for prediction based on a new pageload. +static const uint8_t MAX_PAGELOAD_DEPTH = 10; +bool Predictor::PredictForPageload(nsICacheEntry* entry, nsIURI* targetURI, + uint8_t stackCount, bool fullUri, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::PredictForPageload")); + + if (stackCount > MAX_PAGELOAD_DEPTH) { + PREDICTOR_LOG((" exceeded recursion depth!")); + return false; + } + + uint32_t lastLoad; + nsresult rv = entry->GetLastFetched(&lastLoad); + NS_ENSURE_SUCCESS(rv, false); + + int32_t globalDegradation = CalculateGlobalDegradation(lastLoad); + PREDICTOR_LOG((" globalDegradation = %d", globalDegradation)); + + int32_t loadCount; + rv = entry->GetFetchCount(&loadCount); + NS_ENSURE_SUCCESS(rv, false); + + nsCOMPtr<nsILoadContextInfo> lci; + + rv = entry->GetLoadContextInfo(getter_AddRefs(lci)); + NS_ENSURE_SUCCESS(rv, false); + + nsCOMPtr<nsIURI> redirectURI; + if (WouldRedirect(entry, loadCount, lastLoad, globalDegradation, + getter_AddRefs(redirectURI))) { + mPreconnects.AppendElement(redirectURI); + Predictor::Reason reason; + reason.mPredict = nsINetworkPredictor::PREDICT_LOAD; + RefPtr<Predictor::Action> redirectAction = new Predictor::Action( + Predictor::Action::IS_FULL_URI, Predictor::Action::DO_PREDICT, reason, + redirectURI, nullptr, verifier, this, stackCount + 1); + nsAutoCString redirectUriString; + redirectURI->GetAsciiSpec(redirectUriString); + + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + rv = mCacheStorageService->DiskCacheStorage( + lci, false, getter_AddRefs(cacheDiskStorage)); + NS_ENSURE_SUCCESS(rv, false); + + PREDICTOR_LOG((" Predict redirect uri=%s action=%p", + redirectUriString.get(), redirectAction.get())); + uint32_t openFlags = + nsICacheStorage::OPEN_READONLY | nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::OPEN_PRIORITY | nsICacheStorage::CHECK_MULTITHREADED; + cacheDiskStorage->AsyncOpenURI(redirectURI, ""_ns, openFlags, + redirectAction); + return RunPredictions(nullptr, *lci->OriginAttributesPtr(), verifier); + } + + CalculatePredictions(entry, targetURI, lastLoad, loadCount, globalDegradation, + fullUri); + + return RunPredictions(targetURI, *lci->OriginAttributesPtr(), verifier); +} + +// This is the driver for predicting at browser startup time based on pages that +// have previously been loaded close to startup. +bool Predictor::PredictForStartup(nsICacheEntry* entry, bool fullUri, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::PredictForStartup")); + + nsCOMPtr<nsILoadContextInfo> lci; + + nsresult rv = entry->GetLoadContextInfo(getter_AddRefs(lci)); + NS_ENSURE_SUCCESS(rv, false); + + int32_t globalDegradation = CalculateGlobalDegradation(mLastStartupTime); + CalculatePredictions(entry, nullptr, mLastStartupTime, mStartupCount, + globalDegradation, fullUri); + return RunPredictions(nullptr, *lci->OriginAttributesPtr(), verifier); +} + +// This calculates how much to degrade our confidence in our data based on +// the last time this top-level resource was loaded. This "global degradation" +// applies to *all* subresources we have associated with the top-level +// resource. This will be in addition to any reduction in confidence we have +// associated with a particular subresource. +int32_t Predictor::CalculateGlobalDegradation(uint32_t lastLoad) { + MOZ_ASSERT(NS_IsMainThread()); + + int32_t globalDegradation; + uint32_t delta = NOW_IN_SECONDS() - lastLoad; + if (delta < ONE_DAY) { + globalDegradation = StaticPrefs::network_predictor_page_degradation_day(); + } else if (delta < ONE_WEEK) { + globalDegradation = StaticPrefs::network_predictor_page_degradation_week(); + } else if (delta < ONE_MONTH) { + globalDegradation = StaticPrefs::network_predictor_page_degradation_month(); + } else if (delta < ONE_YEAR) { + globalDegradation = StaticPrefs::network_predictor_page_degradation_year(); + } else { + globalDegradation = StaticPrefs::network_predictor_page_degradation_max(); + } + + Telemetry::Accumulate(Telemetry::PREDICTOR_GLOBAL_DEGRADATION, + globalDegradation); + return globalDegradation; +} + +// This calculates our overall confidence that a particular subresource will be +// loaded as part of a top-level load. +// @param hitCount - the number of times we have loaded this subresource as part +// of this top-level load +// @param hitsPossible - the number of times we have performed this top-level +// load +// @param lastHit - the timestamp of the last time we loaded this subresource as +// part of this top-level load +// @param lastPossible - the timestamp of the last time we performed this +// top-level load +// @param globalDegradation - the degradation for this top-level load as +// determined by CalculateGlobalDegradation +int32_t Predictor::CalculateConfidence(uint32_t hitCount, uint32_t hitsPossible, + uint32_t lastHit, uint32_t lastPossible, + int32_t globalDegradation) { + MOZ_ASSERT(NS_IsMainThread()); + + Telemetry::AutoCounter<Telemetry::PREDICTOR_PREDICTIONS_CALCULATED> + predictionsCalculated; + ++predictionsCalculated; + + if (!hitsPossible) { + return 0; + } + + int32_t baseConfidence = (hitCount * 100) / hitsPossible; + int32_t maxConfidence = 100; + int32_t confidenceDegradation = 0; + + if (lastHit < lastPossible) { + // We didn't load this subresource the last time this top-level load was + // performed, so let's not bother preconnecting (at the very least). + maxConfidence = + StaticPrefs::network_predictor_preconnect_min_confidence() - 1; + + // Now calculate how much we want to degrade our confidence based on how + // long it's been between the last time we did this top-level load and the + // last time this top-level load included this subresource. + PRTime delta = lastPossible - lastHit; + if (delta == 0) { + confidenceDegradation = 0; + } else if (delta < ONE_DAY) { + confidenceDegradation = + StaticPrefs::network_predictor_subresource_degradation_day(); + } else if (delta < ONE_WEEK) { + confidenceDegradation = + StaticPrefs::network_predictor_subresource_degradation_week(); + } else if (delta < ONE_MONTH) { + confidenceDegradation = + StaticPrefs::network_predictor_subresource_degradation_month(); + } else if (delta < ONE_YEAR) { + confidenceDegradation = + StaticPrefs::network_predictor_subresource_degradation_year(); + } else { + confidenceDegradation = + StaticPrefs::network_predictor_subresource_degradation_max(); + maxConfidence = 0; + } + } + + // Calculate our confidence and clamp it to between 0 and maxConfidence + // (<= 100) + int32_t confidence = + baseConfidence - confidenceDegradation - globalDegradation; + confidence = std::max(confidence, 0); + confidence = std::min(confidence, maxConfidence); + + Telemetry::Accumulate(Telemetry::PREDICTOR_BASE_CONFIDENCE, baseConfidence); + Telemetry::Accumulate(Telemetry::PREDICTOR_SUBRESOURCE_DEGRADATION, + confidenceDegradation); + Telemetry::Accumulate(Telemetry::PREDICTOR_CONFIDENCE, confidence); + return confidence; +} + +static void MakeMetadataEntry(const uint32_t hitCount, const uint32_t lastHit, + const uint32_t flags, nsCString& newValue) { + newValue.Truncate(); + newValue.AppendInt(METADATA_VERSION); + newValue.Append(','); + newValue.AppendInt(hitCount); + newValue.Append(','); + newValue.AppendInt(lastHit); + newValue.Append(','); + newValue.AppendInt(flags); +} + +// On every page load, the rolling window gets shifted by one bit, leaving the +// lowest bit at 0, to indicate that the subresource in question has not been +// seen on the most recent page load. If, at some point later during the page +// load, the subresource is seen again, we will then set the lowest bit to 1. +// This is how we keep track of how many of the last n pageloads (for n <= 20) a +// particular subresource has been seen. The rolling window is kept in the upper +// 20 bits of the flags element of the metadata. This saves 12 bits for regular +// old flags. +void Predictor::UpdateRollingLoadCount(nsICacheEntry* entry, + const uint32_t flags, const char* key, + const uint32_t hitCount, + const uint32_t lastHit) { + // Extract just the rolling load count from the flags, shift it to clear the + // lowest bit, and put the new value with the existing flags. + uint32_t rollingLoadCount = flags & ~kFlagsMask; + rollingLoadCount <<= 1; + uint32_t newFlags = (flags & kFlagsMask) | rollingLoadCount; + + // Finally, update the metadata on the cache entry. + nsAutoCString newValue; + MakeMetadataEntry(hitCount, lastHit, newFlags, newValue); + entry->SetMetaDataElement(key, newValue.BeginReading()); +} + +uint32_t Predictor::ClampedPrefetchRollingLoadCount() { + int32_t n = StaticPrefs::network_predictor_prefetch_rolling_load_count(); + if (n < 0) { + return 0; + } + if (n > kMaxPrefetchRollingLoadCount) { + return kMaxPrefetchRollingLoadCount; + } + return n; +} + +void Predictor::CalculatePredictions(nsICacheEntry* entry, nsIURI* referrer, + uint32_t lastLoad, uint32_t loadCount, + int32_t globalDegradation, bool fullUri) { + MOZ_ASSERT(NS_IsMainThread()); + + // Since the visitor gets called under a cache lock, all we do there is get + // copies of the keys/values we care about, and then do the real work here + entry->VisitMetaData(this); + nsTArray<nsCString> keysToOperateOn = std::move(mKeysToOperateOn), + valuesToOperateOn = std::move(mValuesToOperateOn); + + MOZ_ASSERT(keysToOperateOn.Length() == valuesToOperateOn.Length()); + for (size_t i = 0; i < keysToOperateOn.Length(); ++i) { + const char* key = keysToOperateOn[i].BeginReading(); + const char* value = valuesToOperateOn[i].BeginReading(); + + nsCString uri; + uint32_t hitCount, lastHit, flags; + if (!ParseMetaDataEntry(key, value, uri, hitCount, lastHit, flags)) { + // This failed, get rid of it so we don't waste space + entry->SetMetaDataElement(key, nullptr); + continue; + } + + int32_t confidence = CalculateConfidence(hitCount, loadCount, lastHit, + lastLoad, globalDegradation); + if (fullUri) { + UpdateRollingLoadCount(entry, flags, key, hitCount, lastHit); + } + PREDICTOR_LOG(("CalculatePredictions key=%s value=%s confidence=%d", key, + value, confidence)); + PrefetchIgnoreReason reason = PREFETCH_OK; + if (!fullUri) { + // Not full URI - don't prefetch! No sense in it! + PREDICTOR_LOG((" forcing non-cacheability - not full URI")); + if (flags & FLAG_PREFETCHABLE) { + // This only applies if we had somehow otherwise marked this + // prefetchable. + reason = NOT_FULL_URI; + } + flags &= ~FLAG_PREFETCHABLE; + } else if (!referrer) { + // No referrer means we can't prefetch, so pretend it's non-cacheable, + // no matter what. + PREDICTOR_LOG((" forcing non-cacheability - no referrer")); + if (flags & FLAG_PREFETCHABLE) { + // This only applies if we had somehow otherwise marked this + // prefetchable. + reason = NO_REFERRER; + } + flags &= ~FLAG_PREFETCHABLE; + } else { + uint32_t expectedRollingLoadCount = + (1 << ClampedPrefetchRollingLoadCount()) - 1; + expectedRollingLoadCount <<= kRollingLoadOffset; + if ((flags & expectedRollingLoadCount) != expectedRollingLoadCount) { + PREDICTOR_LOG((" forcing non-cacheability - missed a load")); + if (flags & FLAG_PREFETCHABLE) { + // This only applies if we had somehow otherwise marked this + // prefetchable. + reason = MISSED_A_LOAD; + } + flags &= ~FLAG_PREFETCHABLE; + } + } + + PREDICTOR_LOG((" setting up prediction")); + SetupPrediction(confidence, flags, uri, reason); + } +} + +// (Maybe) adds a predictive action to the prediction runner, based on our +// calculated confidence for the subresource in question. +void Predictor::SetupPrediction(int32_t confidence, uint32_t flags, + const nsCString& uri, + PrefetchIgnoreReason earlyReason) { + MOZ_ASSERT(NS_IsMainThread()); + + nsresult rv = NS_OK; + PREDICTOR_LOG( + ("SetupPrediction enable-prefetch=%d prefetch-min-confidence=%d " + "preconnect-min-confidence=%d preresolve-min-confidence=%d " + "flags=%d confidence=%d uri=%s", + StaticPrefs::network_predictor_enable_prefetch(), + StaticPrefs::network_predictor_prefetch_min_confidence(), + StaticPrefs::network_predictor_preconnect_min_confidence(), + StaticPrefs::network_predictor_preresolve_min_confidence(), flags, + confidence, uri.get())); + + bool prefetchOk = !!(flags & FLAG_PREFETCHABLE); + PrefetchIgnoreReason reason = earlyReason; + if (prefetchOk && !StaticPrefs::network_predictor_enable_prefetch()) { + prefetchOk = false; + reason = PREFETCH_DISABLED; + } else if (prefetchOk && !ClampedPrefetchRollingLoadCount() && + confidence < + StaticPrefs::network_predictor_prefetch_min_confidence()) { + prefetchOk = false; + if (!ClampedPrefetchRollingLoadCount()) { + reason = PREFETCH_DISABLED_VIA_COUNT; + } else { + reason = CONFIDENCE_TOO_LOW; + } + } + + // prefetchOk == false and reason == PREFETCH_OK indicates that the reason + // we aren't prefetching this item is because it was marked un-prefetchable in + // our metadata. We already have separate telemetry on that decision, so we + // aren't going to accumulate more here. Right now we only care about why + // something we had marked prefetchable isn't being prefetched. + if (!prefetchOk && reason != PREFETCH_OK) { + Telemetry::Accumulate(Telemetry::PREDICTOR_PREFETCH_IGNORE_REASON, reason); + } + + if (prefetchOk) { + nsCOMPtr<nsIURI> prefetchURI; + rv = NS_NewURI(getter_AddRefs(prefetchURI), uri); + if (NS_SUCCEEDED(rv)) { + mPrefetches.AppendElement(prefetchURI); + } + } else if (confidence >= + StaticPrefs::network_predictor_preconnect_min_confidence()) { + nsCOMPtr<nsIURI> preconnectURI; + rv = NS_NewURI(getter_AddRefs(preconnectURI), uri); + if (NS_SUCCEEDED(rv)) { + mPreconnects.AppendElement(preconnectURI); + } + } else if (confidence >= + StaticPrefs::network_predictor_preresolve_min_confidence()) { + nsCOMPtr<nsIURI> preresolveURI; + rv = NS_NewURI(getter_AddRefs(preresolveURI), uri); + if (NS_SUCCEEDED(rv)) { + mPreresolves.AppendElement(preresolveURI); + } + } + + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" NS_NewURI returned 0x%" PRIx32, static_cast<uint32_t>(rv))); + } +} + +nsresult Predictor::Prefetch(nsIURI* uri, nsIURI* referrer, + const OriginAttributes& originAttributes, + nsINetworkPredictorVerifier* verifier) { + nsAutoCString strUri, strReferrer; + uri->GetAsciiSpec(strUri); + referrer->GetAsciiSpec(strReferrer); + PREDICTOR_LOG(("Predictor::Prefetch uri=%s referrer=%s verifier=%p", + strUri.get(), strReferrer.get(), verifier)); + nsCOMPtr<nsIChannel> channel; + nsresult rv = NS_NewChannel( + getter_AddRefs(channel), uri, nsContentUtils::GetSystemPrincipal(), + nsILoadInfo::SEC_ALLOW_CROSS_ORIGIN_SEC_CONTEXT_IS_NULL, + nsIContentPolicy::TYPE_OTHER, nullptr, /* nsICookieJarSettings */ + nullptr, /* aPerformanceStorage */ + nullptr, /* aLoadGroup */ + nullptr, /* aCallbacks */ + nsIRequest::LOAD_BACKGROUND); + + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" NS_NewChannel failed rv=0x%" PRIX32, static_cast<uint32_t>(rv))); + return rv; + } + + nsCOMPtr<nsILoadInfo> loadInfo = channel->LoadInfo(); + rv = loadInfo->SetOriginAttributes(originAttributes); + + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" Set originAttributes into loadInfo failed rv=0x%" PRIX32, + static_cast<uint32_t>(rv))); + return rv; + } + + nsCOMPtr<nsIHttpChannel> httpChannel; + httpChannel = do_QueryInterface(channel); + if (!httpChannel) { + PREDICTOR_LOG((" Could not get HTTP Channel from new channel!")); + return NS_ERROR_UNEXPECTED; + } + + nsCOMPtr<nsIReferrerInfo> referrerInfo = new dom::ReferrerInfo(referrer); + rv = httpChannel->SetReferrerInfoWithoutClone(referrerInfo); + NS_ENSURE_SUCCESS(rv, rv); + // XXX - set a header here to indicate this is a prefetch? + + nsCOMPtr<nsIStreamListener> listener = + new PrefetchListener(verifier, uri, this); + PREDICTOR_LOG((" calling AsyncOpen listener=%p channel=%p", listener.get(), + channel.get())); + rv = channel->AsyncOpen(listener); + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" AsyncOpen failed rv=0x%" PRIX32, static_cast<uint32_t>(rv))); + } + + return rv; +} + +// Runs predictions that have been set up. +bool Predictor::RunPredictions(nsIURI* referrer, + const OriginAttributes& originAttributes, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread(), "Running prediction off main thread"); + + PREDICTOR_LOG(("Predictor::RunPredictions")); + + bool predicted = false; + uint32_t len, i; + + nsTArray<nsCOMPtr<nsIURI>> prefetches = std::move(mPrefetches), + preconnects = std::move(mPreconnects), + preresolves = std::move(mPreresolves); + + Telemetry::AutoCounter<Telemetry::PREDICTOR_TOTAL_PREDICTIONS> + totalPredictions; + Telemetry::AutoCounter<Telemetry::PREDICTOR_TOTAL_PREFETCHES> totalPrefetches; + Telemetry::AutoCounter<Telemetry::PREDICTOR_TOTAL_PRECONNECTS> + totalPreconnects; + Telemetry::AutoCounter<Telemetry::PREDICTOR_TOTAL_PRERESOLVES> + totalPreresolves; + + len = prefetches.Length(); + for (i = 0; i < len; ++i) { + PREDICTOR_LOG((" doing prefetch")); + nsCOMPtr<nsIURI> uri = prefetches[i]; + if (NS_SUCCEEDED(Prefetch(uri, referrer, originAttributes, verifier))) { + ++totalPredictions; + ++totalPrefetches; + predicted = true; + } + } + + len = preconnects.Length(); + for (i = 0; i < len; ++i) { + PREDICTOR_LOG((" doing preconnect")); + nsCOMPtr<nsIURI> uri = preconnects[i]; + ++totalPredictions; + ++totalPreconnects; + nsCOMPtr<nsIPrincipal> principal = + BasePrincipal::CreateContentPrincipal(uri, originAttributes); + mSpeculativeService->SpeculativeConnect(uri, principal, this); + predicted = true; + if (verifier) { + PREDICTOR_LOG((" sending preconnect verification")); + verifier->OnPredictPreconnect(uri); + } + } + + len = preresolves.Length(); + for (i = 0; i < len; ++i) { + nsCOMPtr<nsIURI> uri = preresolves[i]; + ++totalPredictions; + ++totalPreresolves; + nsAutoCString hostname; + uri->GetAsciiHost(hostname); + PREDICTOR_LOG((" doing preresolve %s", hostname.get())); + nsCOMPtr<nsICancelable> tmpCancelable; + mDnsService->AsyncResolveNative( + hostname, nsIDNSService::RESOLVE_TYPE_DEFAULT, + (nsIDNSService::RESOLVE_PRIORITY_MEDIUM | + nsIDNSService::RESOLVE_SPECULATE), + nullptr, mDNSListener, nullptr, originAttributes, + getter_AddRefs(tmpCancelable)); + + // Fetch HTTPS RR if needed. + if (StaticPrefs::network_dns_upgrade_with_https_rr() || + StaticPrefs::network_dns_use_https_rr_as_altsvc()) { + mDnsService->AsyncResolveNative( + hostname, nsIDNSService::RESOLVE_TYPE_HTTPSSVC, + (nsIDNSService::RESOLVE_PRIORITY_MEDIUM | + nsIDNSService::RESOLVE_SPECULATE), + nullptr, mDNSListener, nullptr, originAttributes, + getter_AddRefs(tmpCancelable)); + } + + predicted = true; + if (verifier) { + PREDICTOR_LOG((" sending preresolve verification")); + verifier->OnPredictDNS(uri); + } + } + + return predicted; +} + +// Find out if a top-level page is likely to redirect. +bool Predictor::WouldRedirect(nsICacheEntry* entry, uint32_t loadCount, + uint32_t lastLoad, int32_t globalDegradation, + nsIURI** redirectURI) { + // TODO - not doing redirects for first go around + MOZ_ASSERT(NS_IsMainThread()); + + return false; +} + +NS_IMETHODIMP +Predictor::Learn(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, JS::HandleValue originAttributes, + JSContext* aCx) { + OriginAttributes attrs; + + if (!originAttributes.isObject() || !attrs.Init(aCx, originAttributes)) { + return NS_ERROR_INVALID_ARG; + } + + return LearnNative(targetURI, sourceURI, reason, attrs); +} + +// Called from the main thread to update the database +NS_IMETHODIMP +Predictor::LearnNative(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, + const OriginAttributes& originAttributes) { + MOZ_ASSERT(NS_IsMainThread(), + "Predictor interface methods must be called on the main thread"); + + PREDICTOR_LOG(("Predictor::Learn")); + + if (IsNeckoChild()) { + MOZ_DIAGNOSTIC_ASSERT(gNeckoChild); + + PREDICTOR_LOG((" called on child process")); + + RefPtr<PredictorLearnRunnable> runnable = new PredictorLearnRunnable( + targetURI, sourceURI, reason, originAttributes); + SchedulerGroup::Dispatch(TaskCategory::Other, runnable.forget()); + + return NS_OK; + } + + PREDICTOR_LOG((" called on parent process")); + + if (!mInitialized) { + PREDICTOR_LOG((" not initialized")); + return NS_OK; + } + + if (!StaticPrefs::network_predictor_enabled()) { + PREDICTOR_LOG((" not enabled")); + return NS_OK; + } + + if (originAttributes.mPrivateBrowsingId > 0) { + // Don't want to do anything in PB mode + PREDICTOR_LOG((" in PB mode")); + return NS_OK; + } + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + PREDICTOR_LOG((" got non-HTTP[S] URI")); + return NS_ERROR_INVALID_ARG; + } + + nsCOMPtr<nsIURI> targetOrigin; + nsCOMPtr<nsIURI> sourceOrigin; + nsCOMPtr<nsIURI> uriKey; + nsCOMPtr<nsIURI> originKey; + nsresult rv; + + switch (reason) { + case nsINetworkPredictor::LEARN_LOAD_TOPLEVEL: + if (!targetURI || sourceURI) { + PREDICTOR_LOG((" load toplevel invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + rv = ExtractOrigin(targetURI, getter_AddRefs(targetOrigin)); + NS_ENSURE_SUCCESS(rv, rv); + uriKey = targetURI; + originKey = targetOrigin; + break; + case nsINetworkPredictor::LEARN_STARTUP: + if (!targetURI || sourceURI) { + PREDICTOR_LOG((" startup invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + rv = ExtractOrigin(targetURI, getter_AddRefs(targetOrigin)); + NS_ENSURE_SUCCESS(rv, rv); + uriKey = mStartupURI; + originKey = mStartupURI; + break; + case nsINetworkPredictor::LEARN_LOAD_REDIRECT: + case nsINetworkPredictor::LEARN_LOAD_SUBRESOURCE: + if (!targetURI || !sourceURI) { + PREDICTOR_LOG((" redirect/subresource invalid URI state")); + return NS_ERROR_INVALID_ARG; + } + rv = ExtractOrigin(targetURI, getter_AddRefs(targetOrigin)); + NS_ENSURE_SUCCESS(rv, rv); + rv = ExtractOrigin(sourceURI, getter_AddRefs(sourceOrigin)); + NS_ENSURE_SUCCESS(rv, rv); + uriKey = sourceURI; + originKey = sourceOrigin; + break; + default: + PREDICTOR_LOG((" invalid reason")); + return NS_ERROR_INVALID_ARG; + } + + Telemetry::AutoCounter<Telemetry::PREDICTOR_LEARN_ATTEMPTS> learnAttempts; + ++learnAttempts; + + Predictor::Reason argReason; + argReason.mLearn = reason; + + // We always open the full uri (general cache) entry first, so we don't gum up + // the works waiting on predictor-only entries to open + RefPtr<Predictor::Action> uriAction = new Predictor::Action( + Predictor::Action::IS_FULL_URI, Predictor::Action::DO_LEARN, argReason, + targetURI, sourceURI, nullptr, this); + nsAutoCString uriKeyStr, targetUriStr, sourceUriStr; + uriKey->GetAsciiSpec(uriKeyStr); + targetURI->GetAsciiSpec(targetUriStr); + if (sourceURI) { + sourceURI->GetAsciiSpec(sourceUriStr); + } + PREDICTOR_LOG( + (" Learn uriKey=%s targetURI=%s sourceURI=%s reason=%d " + "action=%p", + uriKeyStr.get(), targetUriStr.get(), sourceUriStr.get(), reason, + uriAction.get())); + + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + RefPtr<LoadContextInfo> lci = new LoadContextInfo(false, originAttributes); + + rv = mCacheStorageService->DiskCacheStorage(lci, false, + getter_AddRefs(cacheDiskStorage)); + NS_ENSURE_SUCCESS(rv, rv); + + // For learning full URI things, we *always* open readonly and secretly, as we + // rely on actual pageloads to update the entry's metadata for us. + uint32_t uriOpenFlags = nsICacheStorage::OPEN_READONLY | + nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::CHECK_MULTITHREADED; + if (reason == nsINetworkPredictor::LEARN_LOAD_TOPLEVEL) { + // Learning for toplevel we want to open the full uri entry priority, since + // it's likely this entry will be used soon anyway, and we want this to be + // opened ASAP. + uriOpenFlags |= nsICacheStorage::OPEN_PRIORITY; + } + cacheDiskStorage->AsyncOpenURI(uriKey, ""_ns, uriOpenFlags, uriAction); + + // Now we open the origin-only (and therefore predictor-only) entry + RefPtr<Predictor::Action> originAction = new Predictor::Action( + Predictor::Action::IS_ORIGIN, Predictor::Action::DO_LEARN, argReason, + targetOrigin, sourceOrigin, nullptr, this); + nsAutoCString originKeyStr, targetOriginStr, sourceOriginStr; + originKey->GetAsciiSpec(originKeyStr); + targetOrigin->GetAsciiSpec(targetOriginStr); + if (sourceOrigin) { + sourceOrigin->GetAsciiSpec(sourceOriginStr); + } + PREDICTOR_LOG( + (" Learn originKey=%s targetOrigin=%s sourceOrigin=%s reason=%d " + "action=%p", + originKeyStr.get(), targetOriginStr.get(), sourceOriginStr.get(), reason, + originAction.get())); + uint32_t originOpenFlags; + if (reason == nsINetworkPredictor::LEARN_LOAD_TOPLEVEL) { + // This is the only case when we want to update the 'last used' metadata on + // the cache entry we're getting. This only applies to predictor-specific + // entries. + originOpenFlags = + nsICacheStorage::OPEN_NORMALLY | nsICacheStorage::CHECK_MULTITHREADED; + } else { + originOpenFlags = nsICacheStorage::OPEN_READONLY | + nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::CHECK_MULTITHREADED; + } + cacheDiskStorage->AsyncOpenURI(originKey, + nsLiteralCString(PREDICTOR_ORIGIN_EXTENSION), + originOpenFlags, originAction); + + PREDICTOR_LOG(("Predictor::Learn returning")); + return NS_OK; +} + +void Predictor::LearnInternal(PredictorLearnReason reason, nsICacheEntry* entry, + bool isNew, bool fullUri, nsIURI* targetURI, + nsIURI* sourceURI) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::LearnInternal")); + + nsCString junk; + if (!fullUri && reason == nsINetworkPredictor::LEARN_LOAD_TOPLEVEL && + NS_FAILED( + entry->GetMetaDataElement(SEEN_META_DATA, getter_Copies(junk)))) { + // This is an origin-only entry that we haven't seen before. Let's mark it + // as seen. + PREDICTOR_LOG((" marking new origin entry as seen")); + nsresult rv = entry->SetMetaDataElement(SEEN_META_DATA, "1"); + if (NS_FAILED(rv)) { + PREDICTOR_LOG((" failed to mark origin entry seen")); + return; + } + + // Need to ensure someone else can get to the entry if necessary + entry->MetaDataReady(); + return; + } + + switch (reason) { + case nsINetworkPredictor::LEARN_LOAD_TOPLEVEL: + // This case only exists to be used during tests - code outside the + // predictor tests should NEVER call Learn with LEARN_LOAD_TOPLEVEL. + // The predictor xpcshell test needs this branch, however, because we + // have no real page loads in xpcshell, and this is how we fake it up + // so that all the work that normally happens behind the scenes in a + // page load can be done for testing purposes. + if (fullUri && StaticPrefs::network_predictor_doing_tests()) { + PREDICTOR_LOG( + (" WARNING - updating rolling load count. " + "If you see this outside tests, you did it wrong")); + + // Since the visitor gets called under a cache lock, all we do there is + // get copies of the keys/values we care about, and then do the real + // work here + entry->VisitMetaData(this); + nsTArray<nsCString> keysToOperateOn = std::move(mKeysToOperateOn), + valuesToOperateOn = std::move(mValuesToOperateOn); + + MOZ_ASSERT(keysToOperateOn.Length() == valuesToOperateOn.Length()); + for (size_t i = 0; i < keysToOperateOn.Length(); ++i) { + const char* key = keysToOperateOn[i].BeginReading(); + const char* value = valuesToOperateOn[i].BeginReading(); + + nsCString uri; + uint32_t hitCount, lastHit, flags; + if (!ParseMetaDataEntry(key, value, uri, hitCount, lastHit, flags)) { + // This failed, get rid of it so we don't waste space + entry->SetMetaDataElement(key, nullptr); + continue; + } + UpdateRollingLoadCount(entry, flags, key, hitCount, lastHit); + } + } else { + PREDICTOR_LOG((" nothing to do for toplevel")); + } + break; + case nsINetworkPredictor::LEARN_LOAD_REDIRECT: + if (fullUri) { + LearnForRedirect(entry, targetURI); + } + break; + case nsINetworkPredictor::LEARN_LOAD_SUBRESOURCE: + LearnForSubresource(entry, targetURI); + break; + case nsINetworkPredictor::LEARN_STARTUP: + LearnForStartup(entry, targetURI); + break; + default: + PREDICTOR_LOG((" unexpected reason value")); + MOZ_ASSERT(false, "Got unexpected value for learn reason!"); + } +} + +NS_IMPL_ISUPPORTS(Predictor::SpaceCleaner, nsICacheEntryMetaDataVisitor) + +NS_IMETHODIMP +Predictor::SpaceCleaner::OnMetaDataElement(const char* key, const char* value) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsURIMetadataElement(key)) { + // This isn't a bit of metadata we care about + return NS_OK; + } + + nsCString uri; + uint32_t hitCount, lastHit, flags; + bool ok = + mPredictor->ParseMetaDataEntry(key, value, uri, hitCount, lastHit, flags); + + if (!ok) { + // Couldn't parse this one, just get rid of it + nsCString nsKey; + nsKey.AssignASCII(key); + mLongKeysToDelete.AppendElement(nsKey); + return NS_OK; + } + + uint32_t uriLength = uri.Length(); + if (uriLength > StaticPrefs::network_predictor_max_uri_length()) { + // Default to getting rid of URIs that are too long and were put in before + // we had our limit on URI length, in order to free up some space. + nsCString nsKey; + nsKey.AssignASCII(key); + mLongKeysToDelete.AppendElement(nsKey); + return NS_OK; + } + + if (!mLRUKeyToDelete || lastHit < mLRUStamp) { + mLRUKeyToDelete = key; + mLRUStamp = lastHit; + } + + return NS_OK; +} + +void Predictor::SpaceCleaner::Finalize(nsICacheEntry* entry) { + MOZ_ASSERT(NS_IsMainThread()); + + if (mLRUKeyToDelete) { + entry->SetMetaDataElement(mLRUKeyToDelete, nullptr); + } + + for (size_t i = 0; i < mLongKeysToDelete.Length(); ++i) { + entry->SetMetaDataElement(mLongKeysToDelete[i].BeginReading(), nullptr); + } +} + +// Called when a subresource has been hit from a top-level load. Uses the two +// helper functions above to update the database appropriately. +void Predictor::LearnForSubresource(nsICacheEntry* entry, nsIURI* targetURI) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG(("Predictor::LearnForSubresource")); + + uint32_t lastLoad; + nsresult rv = entry->GetLastFetched(&lastLoad); + NS_ENSURE_SUCCESS_VOID(rv); + + int32_t loadCount; + rv = entry->GetFetchCount(&loadCount); + NS_ENSURE_SUCCESS_VOID(rv); + + nsCString key; + key.AssignLiteral(META_DATA_PREFIX); + nsCString uri; + targetURI->GetAsciiSpec(uri); + key.Append(uri); + if (uri.Length() > StaticPrefs::network_predictor_max_uri_length()) { + // We do this to conserve space/prevent OOMs + PREDICTOR_LOG((" uri too long!")); + entry->SetMetaDataElement(key.BeginReading(), nullptr); + return; + } + + nsCString value; + rv = entry->GetMetaDataElement(key.BeginReading(), getter_Copies(value)); + + uint32_t hitCount, lastHit, flags; + bool isNewResource = + (NS_FAILED(rv) || + !ParseMetaDataEntry(key.BeginReading(), value.BeginReading(), uri, + hitCount, lastHit, flags)); + + int32_t resourceCount = 0; + if (isNewResource) { + // This is a new addition + PREDICTOR_LOG((" new resource")); + nsCString s; + rv = entry->GetMetaDataElement(RESOURCE_META_DATA, getter_Copies(s)); + if (NS_SUCCEEDED(rv)) { + resourceCount = atoi(s.BeginReading()); + } + if (resourceCount >= + StaticPrefs::network_predictor_max_resources_per_entry()) { + RefPtr<Predictor::SpaceCleaner> cleaner = + new Predictor::SpaceCleaner(this); + entry->VisitMetaData(cleaner); + cleaner->Finalize(entry); + } else { + ++resourceCount; + } + nsAutoCString count; + count.AppendInt(resourceCount); + rv = entry->SetMetaDataElement(RESOURCE_META_DATA, count.BeginReading()); + if (NS_FAILED(rv)) { + PREDICTOR_LOG((" failed to update resource count")); + return; + } + hitCount = 1; + flags = 0; + } else { + PREDICTOR_LOG((" existing resource")); + hitCount = std::min(hitCount + 1, static_cast<uint32_t>(loadCount)); + } + + // Update the rolling load count to mark this sub-resource as seen on the + // most-recent pageload so it can be eligible for prefetch (assuming all + // the other stars align). + flags |= (1 << kRollingLoadOffset); + + nsCString newValue; + MakeMetadataEntry(hitCount, lastLoad, flags, newValue); + rv = entry->SetMetaDataElement(key.BeginReading(), newValue.BeginReading()); + PREDICTOR_LOG( + (" SetMetaDataElement -> 0x%08" PRIX32, static_cast<uint32_t>(rv))); + if (NS_FAILED(rv) && isNewResource) { + // Roll back the increment to the resource count we made above. + PREDICTOR_LOG((" rolling back resource count update")); + --resourceCount; + if (resourceCount == 0) { + entry->SetMetaDataElement(RESOURCE_META_DATA, nullptr); + } else { + nsAutoCString count; + count.AppendInt(resourceCount); + entry->SetMetaDataElement(RESOURCE_META_DATA, count.BeginReading()); + } + } +} + +// This is called when a top-level loaded ended up redirecting to a different +// URI so we can keep track of that fact. +void Predictor::LearnForRedirect(nsICacheEntry* entry, nsIURI* targetURI) { + MOZ_ASSERT(NS_IsMainThread()); + + // TODO - not doing redirects for first go around + PREDICTOR_LOG(("Predictor::LearnForRedirect")); +} + +// This will add a page to our list of startup pages if it's being loaded +// before our startup window has expired. +void Predictor::MaybeLearnForStartup(nsIURI* uri, bool fullUri, + const OriginAttributes& originAttributes) { + MOZ_ASSERT(NS_IsMainThread()); + + // TODO - not doing startup for first go around + PREDICTOR_LOG(("Predictor::MaybeLearnForStartup")); +} + +// Add information about a top-level load to our list of startup pages +void Predictor::LearnForStartup(nsICacheEntry* entry, nsIURI* targetURI) { + MOZ_ASSERT(NS_IsMainThread()); + + // These actually do the same set of work, just on different entries, so we + // can pass through to get the real work done here + PREDICTOR_LOG(("Predictor::LearnForStartup")); + LearnForSubresource(entry, targetURI); +} + +bool Predictor::ParseMetaDataEntry(const char* key, const char* value, + nsCString& uri, uint32_t& hitCount, + uint32_t& lastHit, uint32_t& flags) { + MOZ_ASSERT(NS_IsMainThread()); + + PREDICTOR_LOG( + ("Predictor::ParseMetaDataEntry key=%s value=%s", key ? key : "", value)); + + const char* comma = strchr(value, ','); + if (!comma) { + PREDICTOR_LOG((" could not find first comma")); + return false; + } + + uint32_t version = static_cast<uint32_t>(atoi(value)); + PREDICTOR_LOG((" version -> %u", version)); + + if (version != METADATA_VERSION) { + PREDICTOR_LOG( + (" metadata version mismatch %u != %u", version, METADATA_VERSION)); + return false; + } + + value = comma + 1; + comma = strchr(value, ','); + if (!comma) { + PREDICTOR_LOG((" could not find second comma")); + return false; + } + + hitCount = static_cast<uint32_t>(atoi(value)); + PREDICTOR_LOG((" hitCount -> %u", hitCount)); + + value = comma + 1; + comma = strchr(value, ','); + if (!comma) { + PREDICTOR_LOG((" could not find third comma")); + return false; + } + + lastHit = static_cast<uint32_t>(atoi(value)); + PREDICTOR_LOG((" lastHit -> %u", lastHit)); + + value = comma + 1; + flags = static_cast<uint32_t>(atoi(value)); + PREDICTOR_LOG((" flags -> %u", flags)); + + if (key) { + const char* uriStart = key + (sizeof(META_DATA_PREFIX) - 1); + uri.AssignASCII(uriStart); + PREDICTOR_LOG((" uri -> %s", uriStart)); + } else { + uri.Truncate(); + } + + return true; +} + +NS_IMETHODIMP +Predictor::Reset() { + MOZ_ASSERT(NS_IsMainThread(), + "Predictor interface methods must be called on the main thread"); + + PREDICTOR_LOG(("Predictor::Reset")); + + if (IsNeckoChild()) { + MOZ_DIAGNOSTIC_ASSERT(gNeckoChild); + + PREDICTOR_LOG((" forwarding to parent process")); + gNeckoChild->SendPredReset(); + return NS_OK; + } + + PREDICTOR_LOG((" called on parent process")); + + if (!mInitialized) { + PREDICTOR_LOG((" not initialized")); + return NS_OK; + } + + if (!StaticPrefs::network_predictor_enabled()) { + PREDICTOR_LOG((" not enabled")); + return NS_OK; + } + + RefPtr<Predictor::Resetter> reset = new Predictor::Resetter(this); + PREDICTOR_LOG((" created a resetter")); + mCacheStorageService->AsyncVisitAllStorages(reset, true); + PREDICTOR_LOG((" Cache async launched, returning now")); + + return NS_OK; +} + +NS_IMPL_ISUPPORTS(Predictor::Resetter, nsICacheEntryOpenCallback, + nsICacheEntryMetaDataVisitor, nsICacheStorageVisitor); + +Predictor::Resetter::Resetter(Predictor* predictor) + : mEntriesToVisit(0), mPredictor(predictor) {} + +NS_IMETHODIMP +Predictor::Resetter::OnCacheEntryCheck(nsICacheEntry* entry, + nsIApplicationCache* appCache, + uint32_t* result) { + *result = nsICacheEntryOpenCallback::ENTRY_WANTED; + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Resetter::OnCacheEntryAvailable(nsICacheEntry* entry, bool isNew, + nsIApplicationCache* appCache, + nsresult result) { + MOZ_ASSERT(NS_IsMainThread()); + + if (NS_FAILED(result)) { + // This can happen when we've tried to open an entry that doesn't exist for + // some non-reset operation, and then get reset shortly thereafter (as + // happens in some of our tests). + --mEntriesToVisit; + if (!mEntriesToVisit) { + Complete(); + } + return NS_OK; + } + + entry->VisitMetaData(this); + nsTArray<nsCString> keysToDelete = std::move(mKeysToDelete); + + for (size_t i = 0; i < keysToDelete.Length(); ++i) { + const char* key = keysToDelete[i].BeginReading(); + entry->SetMetaDataElement(key, nullptr); + } + + --mEntriesToVisit; + if (!mEntriesToVisit) { + Complete(); + } + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Resetter::OnMetaDataElement(const char* asciiKey, + const char* asciiValue) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!StringBeginsWith(nsDependentCString(asciiKey), + nsLiteralCString(META_DATA_PREFIX))) { + // Not a metadata entry we care about, carry on + return NS_OK; + } + + nsCString key; + key.AssignASCII(asciiKey); + mKeysToDelete.AppendElement(key); + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Resetter::OnCacheStorageInfo(uint32_t entryCount, + uint64_t consumption, uint64_t capacity, + nsIFile* diskDirectory) { + MOZ_ASSERT(NS_IsMainThread()); + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Resetter::OnCacheEntryInfo(nsIURI* uri, const nsACString& idEnhance, + int64_t dataSize, int32_t fetchCount, + uint32_t lastModifiedTime, + uint32_t expirationTime, bool aPinned, + nsILoadContextInfo* aInfo) { + MOZ_ASSERT(NS_IsMainThread()); + + nsresult rv; + + // The predictor will only ever touch entries with no idEnhance ("") or an + // idEnhance of PREDICTOR_ORIGIN_EXTENSION, so we filter out any entries that + // don't match that to avoid doing extra work. + if (idEnhance.EqualsLiteral(PREDICTOR_ORIGIN_EXTENSION)) { + // This is an entry we own, so we can just doom it entirely + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + rv = mPredictor->mCacheStorageService->DiskCacheStorage( + aInfo, false, getter_AddRefs(cacheDiskStorage)); + + NS_ENSURE_SUCCESS(rv, rv); + cacheDiskStorage->AsyncDoomURI(uri, idEnhance, nullptr); + } else if (idEnhance.IsEmpty()) { + // This is an entry we don't own, so we have to be a little more careful and + // just get rid of our own metadata entries. Append it to an array of things + // to operate on and then do the operations later so we don't end up calling + // Complete() multiple times/too soon. + ++mEntriesToVisit; + mURIsToVisit.AppendElement(uri); + mInfosToVisit.AppendElement(aInfo); + } + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::Resetter::OnCacheEntryVisitCompleted() { + MOZ_ASSERT(NS_IsMainThread()); + + nsresult rv; + + nsTArray<nsCOMPtr<nsIURI>> urisToVisit = std::move(mURIsToVisit); + + MOZ_ASSERT(mEntriesToVisit == urisToVisit.Length()); + + nsTArray<nsCOMPtr<nsILoadContextInfo>> infosToVisit = + std::move(mInfosToVisit); + + MOZ_ASSERT(mEntriesToVisit == infosToVisit.Length()); + + if (!mEntriesToVisit) { + Complete(); + return NS_OK; + } + + uint32_t entriesToVisit = urisToVisit.Length(); + for (uint32_t i = 0; i < entriesToVisit; ++i) { + nsCString u; + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + rv = mPredictor->mCacheStorageService->DiskCacheStorage( + infosToVisit[i], false, getter_AddRefs(cacheDiskStorage)); + NS_ENSURE_SUCCESS(rv, rv); + + urisToVisit[i]->GetAsciiSpec(u); + cacheDiskStorage->AsyncOpenURI(urisToVisit[i], ""_ns, + nsICacheStorage::OPEN_READONLY | + nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::CHECK_MULTITHREADED, + this); + } + + return NS_OK; +} + +void Predictor::Resetter::Complete() { + MOZ_ASSERT(NS_IsMainThread()); + + nsCOMPtr<nsIObserverService> obs = mozilla::services::GetObserverService(); + if (!obs) { + PREDICTOR_LOG(("COULD NOT GET OBSERVER SERVICE!")); + return; + } + + obs->NotifyObservers(nullptr, "predictor-reset-complete", nullptr); +} + +// Helper functions to make using the predictor easier from native code + +static StaticRefPtr<nsINetworkPredictor> sPredictor; + +static nsresult EnsureGlobalPredictor(nsINetworkPredictor** aPredictor) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!sPredictor) { + nsresult rv; + nsCOMPtr<nsINetworkPredictor> predictor = + do_GetService("@mozilla.org/network/predictor;1", &rv); + NS_ENSURE_SUCCESS(rv, rv); + sPredictor = predictor; + ClearOnShutdown(&sPredictor); + } + + nsCOMPtr<nsINetworkPredictor> predictor = sPredictor.get(); + predictor.forget(aPredictor); + return NS_OK; +} + +nsresult PredictorPredict(nsIURI* targetURI, nsIURI* sourceURI, + PredictorPredictReason reason, + const OriginAttributes& originAttributes, + nsINetworkPredictorVerifier* verifier) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + return NS_OK; + } + + nsCOMPtr<nsINetworkPredictor> predictor; + nsresult rv = EnsureGlobalPredictor(getter_AddRefs(predictor)); + NS_ENSURE_SUCCESS(rv, rv); + + return predictor->PredictNative(targetURI, sourceURI, reason, + originAttributes, verifier); +} + +nsresult PredictorLearn(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, + const OriginAttributes& originAttributes) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + return NS_OK; + } + + nsCOMPtr<nsINetworkPredictor> predictor; + nsresult rv = EnsureGlobalPredictor(getter_AddRefs(predictor)); + NS_ENSURE_SUCCESS(rv, rv); + + return predictor->LearnNative(targetURI, sourceURI, reason, originAttributes); +} + +nsresult PredictorLearn(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, nsILoadGroup* loadGroup) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + return NS_OK; + } + + nsCOMPtr<nsINetworkPredictor> predictor; + nsresult rv = EnsureGlobalPredictor(getter_AddRefs(predictor)); + NS_ENSURE_SUCCESS(rv, rv); + + nsCOMPtr<nsILoadContext> loadContext; + OriginAttributes originAttributes; + + if (loadGroup) { + nsCOMPtr<nsIInterfaceRequestor> callbacks; + loadGroup->GetNotificationCallbacks(getter_AddRefs(callbacks)); + if (callbacks) { + loadContext = do_GetInterface(callbacks); + + if (loadContext) { + loadContext->GetOriginAttributes(originAttributes); + } + } + } + + return predictor->LearnNative(targetURI, sourceURI, reason, originAttributes); +} + +nsresult PredictorLearn(nsIURI* targetURI, nsIURI* sourceURI, + PredictorLearnReason reason, dom::Document* document) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + return NS_OK; + } + + nsCOMPtr<nsINetworkPredictor> predictor; + nsresult rv = EnsureGlobalPredictor(getter_AddRefs(predictor)); + NS_ENSURE_SUCCESS(rv, rv); + + OriginAttributes originAttributes; + + if (document) { + nsCOMPtr<nsIPrincipal> docPrincipal = document->NodePrincipal(); + + if (docPrincipal) { + originAttributes = docPrincipal->OriginAttributesRef(); + } + } + + return predictor->LearnNative(targetURI, sourceURI, reason, originAttributes); +} + +nsresult PredictorLearnRedirect(nsIURI* targetURI, nsIChannel* channel, + const OriginAttributes& originAttributes) { + MOZ_ASSERT(NS_IsMainThread()); + + nsCOMPtr<nsIURI> sourceURI; + nsresult rv = channel->GetOriginalURI(getter_AddRefs(sourceURI)); + NS_ENSURE_SUCCESS(rv, rv); + + bool sameUri; + rv = targetURI->Equals(sourceURI, &sameUri); + NS_ENSURE_SUCCESS(rv, rv); + + if (sameUri) { + return NS_OK; + } + + if (!IsNullOrHttp(targetURI) || !IsNullOrHttp(sourceURI)) { + return NS_OK; + } + + nsCOMPtr<nsINetworkPredictor> predictor; + rv = EnsureGlobalPredictor(getter_AddRefs(predictor)); + NS_ENSURE_SUCCESS(rv, rv); + + return predictor->LearnNative(targetURI, sourceURI, + nsINetworkPredictor::LEARN_LOAD_REDIRECT, + originAttributes); +} + +// nsINetworkPredictorVerifier + +/** + * Call through to the child's verifier (only during tests) + */ +NS_IMETHODIMP +Predictor::OnPredictPrefetch(nsIURI* aURI, uint32_t httpStatus) { + if (IsNeckoChild()) { + if (mChildVerifier) { + // Ideally, we'd assert here. But since we're slowly moving towards a + // world where we have multiple child processes, and only one child + // process will be likely to have a verifier, we have to play it safer. + return mChildVerifier->OnPredictPrefetch(aURI, httpStatus); + } + return NS_OK; + } + + MOZ_DIAGNOSTIC_ASSERT(aURI, "aURI must not be null"); + + for (auto* cp : dom::ContentParent::AllProcesses(dom::ContentParent::eLive)) { + PNeckoParent* neckoParent = SingleManagedOrNull(cp->ManagedPNeckoParent()); + if (!neckoParent) { + continue; + } + if (!neckoParent->SendPredOnPredictPrefetch(aURI, httpStatus)) { + return NS_ERROR_NOT_AVAILABLE; + } + } + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::OnPredictPreconnect(nsIURI* aURI) { + if (IsNeckoChild()) { + if (mChildVerifier) { + // Ideally, we'd assert here. But since we're slowly moving towards a + // world where we have multiple child processes, and only one child + // process will be likely to have a verifier, we have to play it safer. + return mChildVerifier->OnPredictPreconnect(aURI); + } + return NS_OK; + } + + MOZ_DIAGNOSTIC_ASSERT(aURI, "aURI must not be null"); + + for (auto* cp : dom::ContentParent::AllProcesses(dom::ContentParent::eLive)) { + PNeckoParent* neckoParent = SingleManagedOrNull(cp->ManagedPNeckoParent()); + if (!neckoParent) { + continue; + } + if (!neckoParent->SendPredOnPredictPreconnect(aURI)) { + return NS_ERROR_NOT_AVAILABLE; + } + } + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::OnPredictDNS(nsIURI* aURI) { + if (IsNeckoChild()) { + if (mChildVerifier) { + // Ideally, we'd assert here. But since we're slowly moving towards a + // world where we have multiple child processes, and only one child + // process will be likely to have a verifier, we have to play it safer. + return mChildVerifier->OnPredictDNS(aURI); + } + return NS_OK; + } + + MOZ_DIAGNOSTIC_ASSERT(aURI, "aURI must not be null"); + + for (auto* cp : dom::ContentParent::AllProcesses(dom::ContentParent::eLive)) { + PNeckoParent* neckoParent = SingleManagedOrNull(cp->ManagedPNeckoParent()); + if (!neckoParent) { + continue; + } + if (!neckoParent->SendPredOnPredictDNS(aURI)) { + return NS_ERROR_NOT_AVAILABLE; + } + } + + return NS_OK; +} + +// Predictor::PrefetchListener +// nsISupports +NS_IMPL_ISUPPORTS(Predictor::PrefetchListener, nsIStreamListener, + nsIRequestObserver) + +// nsIRequestObserver +NS_IMETHODIMP +Predictor::PrefetchListener::OnStartRequest(nsIRequest* aRequest) { + mStartTime = TimeStamp::Now(); + return NS_OK; +} + +NS_IMETHODIMP +Predictor::PrefetchListener::OnStopRequest(nsIRequest* aRequest, + nsresult aStatusCode) { + PREDICTOR_LOG(("OnStopRequest this=%p aStatusCode=0x%" PRIX32, this, + static_cast<uint32_t>(aStatusCode))); + NS_ENSURE_ARG(aRequest); + if (NS_FAILED(aStatusCode)) { + return aStatusCode; + } + Telemetry::AccumulateTimeDelta(Telemetry::PREDICTOR_PREFETCH_TIME, + mStartTime); + + nsCOMPtr<nsIHttpChannel> httpChannel = do_QueryInterface(aRequest); + if (!httpChannel) { + PREDICTOR_LOG((" Could not get HTTP Channel!")); + return NS_ERROR_UNEXPECTED; + } + nsCOMPtr<nsICachingChannel> cachingChannel = do_QueryInterface(httpChannel); + if (!cachingChannel) { + PREDICTOR_LOG((" Could not get caching channel!")); + return NS_ERROR_UNEXPECTED; + } + + nsresult rv = NS_OK; + uint32_t httpStatus; + rv = httpChannel->GetResponseStatus(&httpStatus); + if (NS_SUCCEEDED(rv) && httpStatus == 200) { + rv = cachingChannel->ForceCacheEntryValidFor( + StaticPrefs::network_predictor_prefetch_force_valid_for()); + PREDICTOR_LOG((" forcing entry valid for %d seconds rv=%" PRIX32, + StaticPrefs::network_predictor_prefetch_force_valid_for(), + static_cast<uint32_t>(rv))); + } else { + rv = cachingChannel->ForceCacheEntryValidFor(0); + PREDICTOR_LOG((" removing any forced validity rv=%" PRIX32, + static_cast<uint32_t>(rv))); + } + + nsAutoCString reqName; + rv = aRequest->GetName(reqName); + if (NS_FAILED(rv)) { + reqName.AssignLiteral("<unknown>"); + } + + PREDICTOR_LOG((" request %s status %u", reqName.get(), httpStatus)); + + if (mVerifier) { + mVerifier->OnPredictPrefetch(mURI, httpStatus); + } + + return rv; +} + +// nsIStreamListener +NS_IMETHODIMP +Predictor::PrefetchListener::OnDataAvailable(nsIRequest* aRequest, + nsIInputStream* aInputStream, + uint64_t aOffset, + const uint32_t aCount) { + uint32_t result; + return aInputStream->ReadSegments(NS_DiscardSegment, nullptr, aCount, + &result); +} + +// Miscellaneous Predictor + +void Predictor::UpdateCacheability(nsIURI* sourceURI, nsIURI* targetURI, + uint32_t httpStatus, + nsHttpRequestHead& requestHead, + nsHttpResponseHead* responseHead, + nsILoadContextInfo* lci, bool isTracking) { + MOZ_ASSERT(NS_IsMainThread()); + + if (lci && lci->IsPrivate()) { + PREDICTOR_LOG(("Predictor::UpdateCacheability in PB mode - ignoring")); + return; + } + + if (!sourceURI || !targetURI) { + PREDICTOR_LOG( + ("Predictor::UpdateCacheability missing source or target uri")); + return; + } + + if (!IsNullOrHttp(sourceURI) || !IsNullOrHttp(targetURI)) { + PREDICTOR_LOG(("Predictor::UpdateCacheability non-http(s) uri")); + return; + } + + RefPtr<Predictor> self = sSelf; + if (self) { + nsAutoCString method; + requestHead.Method(method); + + nsAutoCString vary; + Unused << responseHead->GetHeader(nsHttp::Vary, vary); + + nsAutoCString cacheControlHeader; + Unused << responseHead->GetHeader(nsHttp::Cache_Control, + cacheControlHeader); + CacheControlParser cacheControl(cacheControlHeader); + + self->UpdateCacheabilityInternal(sourceURI, targetURI, httpStatus, method, + *lci->OriginAttributesPtr(), isTracking, + !vary.IsEmpty(), cacheControl.NoStore()); + } +} + +void Predictor::UpdateCacheabilityInternal( + nsIURI* sourceURI, nsIURI* targetURI, uint32_t httpStatus, + const nsCString& method, const OriginAttributes& originAttributes, + bool isTracking, bool couldVary, bool isNoStore) { + PREDICTOR_LOG(("Predictor::UpdateCacheability httpStatus=%u", httpStatus)); + + nsresult rv; + + if (!mInitialized) { + PREDICTOR_LOG((" not initialized")); + return; + } + + if (!StaticPrefs::network_predictor_enabled()) { + PREDICTOR_LOG((" not enabled")); + return; + } + + nsCOMPtr<nsICacheStorage> cacheDiskStorage; + + RefPtr<LoadContextInfo> lci = new LoadContextInfo(false, originAttributes); + + rv = mCacheStorageService->DiskCacheStorage(lci, false, + getter_AddRefs(cacheDiskStorage)); + if (NS_FAILED(rv)) { + PREDICTOR_LOG((" cannot get disk cache storage")); + return; + } + + uint32_t openFlags = nsICacheStorage::OPEN_READONLY | + nsICacheStorage::OPEN_SECRETLY | + nsICacheStorage::CHECK_MULTITHREADED; + RefPtr<Predictor::CacheabilityAction> action = + new Predictor::CacheabilityAction(targetURI, httpStatus, method, + isTracking, couldVary, isNoStore, this); + nsAutoCString uri; + targetURI->GetAsciiSpec(uri); + PREDICTOR_LOG((" uri=%s action=%p", uri.get(), action.get())); + cacheDiskStorage->AsyncOpenURI(sourceURI, ""_ns, openFlags, action); +} + +NS_IMPL_ISUPPORTS(Predictor::CacheabilityAction, nsICacheEntryOpenCallback, + nsICacheEntryMetaDataVisitor); + +NS_IMETHODIMP +Predictor::CacheabilityAction::OnCacheEntryCheck(nsICacheEntry* entry, + nsIApplicationCache* appCache, + uint32_t* result) { + *result = nsICacheEntryOpenCallback::ENTRY_WANTED; + return NS_OK; +} + +namespace { +enum PrefetchDecisionReason { + PREFETCHABLE, + STATUS_NOT_200, + METHOD_NOT_GET, + URL_HAS_QUERY_STRING, + RESOURCE_IS_TRACKING, + RESOURCE_COULD_VARY, + RESOURCE_IS_NO_STORE +}; +} + +NS_IMETHODIMP +Predictor::CacheabilityAction::OnCacheEntryAvailable( + nsICacheEntry* entry, bool isNew, nsIApplicationCache* appCache, + nsresult result) { + MOZ_ASSERT(NS_IsMainThread()); + // This is being opened read-only, so isNew should always be false + MOZ_ASSERT(!isNew); + + PREDICTOR_LOG(("CacheabilityAction::OnCacheEntryAvailable this=%p", this)); + if (NS_FAILED(result)) { + // Nothing to do + PREDICTOR_LOG((" nothing to do result=%" PRIX32 " isNew=%d", + static_cast<uint32_t>(result), isNew)); + return NS_OK; + } + + nsCString strTargetURI; + nsresult rv = mTargetURI->GetAsciiSpec(strTargetURI); + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" GetAsciiSpec returned %" PRIx32, static_cast<uint32_t>(rv))); + return NS_OK; + } + + rv = entry->VisitMetaData(this); + if (NS_FAILED(rv)) { + PREDICTOR_LOG( + (" VisitMetaData returned %" PRIx32, static_cast<uint32_t>(rv))); + return NS_OK; + } + + nsTArray<nsCString> keysToCheck = std::move(mKeysToCheck), + valuesToCheck = std::move(mValuesToCheck); + + bool hasQueryString = false; + nsAutoCString query; + if (NS_SUCCEEDED(mTargetURI->GetQuery(query)) && !query.IsEmpty()) { + hasQueryString = true; + } + + MOZ_ASSERT(keysToCheck.Length() == valuesToCheck.Length()); + for (size_t i = 0; i < keysToCheck.Length(); ++i) { + const char* key = keysToCheck[i].BeginReading(); + const char* value = valuesToCheck[i].BeginReading(); + nsCString uri; + uint32_t hitCount, lastHit, flags; + + if (!mPredictor->ParseMetaDataEntry(key, value, uri, hitCount, lastHit, + flags)) { + PREDICTOR_LOG((" failed to parse key=%s value=%s", key, value)); + continue; + } + + if (strTargetURI.Equals(uri)) { + bool prefetchable = true; + PrefetchDecisionReason reason = PREFETCHABLE; + + if (mHttpStatus != 200) { + prefetchable = false; + reason = STATUS_NOT_200; + } else if (!mMethod.EqualsLiteral("GET")) { + prefetchable = false; + reason = METHOD_NOT_GET; + } else if (hasQueryString) { + prefetchable = false; + reason = URL_HAS_QUERY_STRING; + } else if (mIsTracking) { + prefetchable = false; + reason = RESOURCE_IS_TRACKING; + } else if (mCouldVary) { + prefetchable = false; + reason = RESOURCE_COULD_VARY; + } else if (mIsNoStore) { + // We don't set prefetchable = false yet, because we just want to know + // what kind of effect this would have on prefetching. + reason = RESOURCE_IS_NO_STORE; + } + + Telemetry::Accumulate(Telemetry::PREDICTOR_PREFETCH_DECISION_REASON, + reason); + + if (prefetchable) { + PREDICTOR_LOG((" marking %s cacheable", key)); + flags |= FLAG_PREFETCHABLE; + } else { + PREDICTOR_LOG((" marking %s uncacheable", key)); + flags &= ~FLAG_PREFETCHABLE; + } + nsCString newValue; + MakeMetadataEntry(hitCount, lastHit, flags, newValue); + entry->SetMetaDataElement(key, newValue.BeginReading()); + break; + } + } + + return NS_OK; +} + +NS_IMETHODIMP +Predictor::CacheabilityAction::OnMetaDataElement(const char* asciiKey, + const char* asciiValue) { + MOZ_ASSERT(NS_IsMainThread()); + + if (!IsURIMetadataElement(asciiKey)) { + return NS_OK; + } + + nsCString key, value; + key.AssignASCII(asciiKey); + value.AssignASCII(asciiValue); + mKeysToCheck.AppendElement(key); + mValuesToCheck.AppendElement(value); + + return NS_OK; +} + +} // namespace net +} // namespace mozilla |