Copilot commented on code in PR #13439: URL: https://github.com/apache/trafficserver/pull/13439#discussion_r3712792531
########## src/iocore/eventsystem/unit_tests/test_Lock.cc: ########## @@ -0,0 +1,381 @@ +/** @file + + Catch2 unit tests for the Lock.h mutex acquisition macros. + + @section license License + + Licensed to the Apache Software Foundation (ASF) under one + or more contributor license agreements. See the NOTICE file + distributed with this work for additional information + regarding copyright ownership. The ASF licenses this file + to you under the Apache License, Version 2.0 (the + "License"); you may not use this file except in compliance + with the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. + */ + +#define CATCH_CONFIG_THREAD_SAFE_ASSERTIONS +#include "inkevent_test_fixtures.h" + +using inkevent_test::AtomicFlag; +using inkevent_test::EventProcessorListener; + +CATCH_REGISTER_LISTENER(EventProcessorListener) + +namespace +{ + +class HoldOnEThread : public Continuation +{ +public: + HoldOnEThread(ProxyMutex *self_mutex, Ptr<ProxyMutex> &target) : Continuation(self_mutex), target_mutex(target) + { + SET_HANDLER(&HoldOnEThread::on_event); + this->callback_action = eventProcessor.schedule_imm(this, ET_CALL); + } + + // In case of an exception in a thread that would have set release, we set + // it here in order to unfreeze any threads that may be waiting on done. + ~HoldOnEThread() + { + this->release.set(); + this->cancel_callback(); + this->done.wait_until_set(); + } Review Comment: This schedules `this` onto another thread, but `HoldOnEThread` is stack-allocated in the tests. The destructor attempts to prevent use-after-free by canceling the action and waiting on `done`, but `cancel_callback()` sets `done` immediately after calling `cancel()`, without establishing that the callback cannot still run (cancellation may be best-effort depending on timing). If the callback executes after `done` is set and the destructor returns, it could dereference a freed object. Prefer making the continuation lifetime explicit (e.g., heap-allocate and ensure the callback owns/releases it, or only signal `done` after you have a guaranteed \"won't run\" outcome from cancellation / a definitive synchronization that the callback has not started). ########## src/iocore/eventsystem/unit_tests/test_Lock.cc: ########## @@ -0,0 +1,381 @@ +/** @file + + Catch2 unit tests for the Lock.h mutex acquisition macros. + + @section license License + + Licensed to the Apache Software Foundation (ASF) under one + or more contributor license agreements. See the NOTICE file + distributed with this work for additional information + regarding copyright ownership. The ASF licenses this file + to you under the Apache License, Version 2.0 (the + "License"); you may not use this file except in compliance + with the License. You may obtain a copy of the License at + + http://www.apache.org/licenses/LICENSE-2.0 + + Unless required by applicable law or agreed to in writing, software + distributed under the License is distributed on an "AS IS" BASIS, + WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. + See the License for the specific language governing permissions and + limitations under the License. + */ + +#define CATCH_CONFIG_THREAD_SAFE_ASSERTIONS +#include "inkevent_test_fixtures.h" + +using inkevent_test::AtomicFlag; +using inkevent_test::EventProcessorListener; + +CATCH_REGISTER_LISTENER(EventProcessorListener) + +namespace +{ + +class HoldOnEThread : public Continuation +{ +public: + HoldOnEThread(ProxyMutex *self_mutex, Ptr<ProxyMutex> &target) : Continuation(self_mutex), target_mutex(target) + { + SET_HANDLER(&HoldOnEThread::on_event); + this->callback_action = eventProcessor.schedule_imm(this, ET_CALL); + } + + // In case of an exception in a thread that would have set release, we set + // it here in order to unfreeze any threads that may be waiting on done. + ~HoldOnEThread() + { + this->release.set(); + this->cancel_callback(); + this->done.wait_until_set(); + } + + bool + wait_for_callback_start() + { + return this->held.wait_until_set(); + } + + bool + wait_for_callback_finish() + { + this->release.set(); + return this->done.wait_until_set(); + } + +private: + Action *callback_action{}; + Ptr<ProxyMutex> target_mutex; + AtomicFlag held; + AtomicFlag release; + AtomicFlag done; + + int + on_event(int /* event ATS_UNUSED */, void * /* data ATS_UNUSED */) + { + SCOPED_MUTEX_LOCK(guard, this->target_mutex, this_ethread()); + this->held.set(); + if (this->release.wait_until_set()) { + MUTEX_RELEASE(guard); + this->done.set(); + } + return 0; Review Comment: `done` is only set when `release.wait_until_set()` returns true. Since this method returns `bool`, a `false` return path appears possible; if that happens, `~HoldOnEThread()` will block forever on `done.wait_until_set()`, hanging the test process. To make the test robust, ensure `done` is always signaled (e.g., set `done` unconditionally before returning, or handle the `false` path explicitly). -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
