Fixed leak of NLLanguageRecognizer on worker thread exit. - #36
Open
haikesan wants to merge 1 commit into
Open
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Platform::Language::Recognizekeeps oneNLLanguageRecognizerper thread in astatic thread_localraw pointer (since c8ded8b7). Nothing ever releases it, so every time a thread that calledRecognizeexits, its recognizer is leaked together with the CoreML model it holds (MontrealNeuralNetwork/Espresso::net).On macOS
Spellchecker::CheckSpellingcallsRecognizefor every word, and it runs oncrl::asyncworker threads (platform_spellcheck_async.cpp). GCD spins those threads up and tears them down all the time, so the leak accumulates for as long as the app lives.Observed on a Telegram Desktop build after ~2 days of uptime,
leaks:and
heapshows the matching 1441std::__shared_ptr_emplace<Espresso::net>/ 1439MontrealNeuralNetworkinstances still alive.Standalone reproduction (200 short-lived threads, each calls
Recognizeonce):before -
15254 leaks for 13153648 total leaked bytes, 200xROOT LEAK: <NLLanguageRecognizer>after -
0 leaks for 0 total leaked bytesThe fix keeps the per-thread instance (same design as the cld3
thread_localin 3a47097) but wraps it in a small RAII holder, so the thread-local destructor releases the recognizer when the thread exits.repro.mm