[SYCL] Fix hang in deallocateStreams function #2796
Merged
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
This patch fixes hang caused by data race in `scheduler.cpp`'s `deallocateStreams` function: data of `StreamsToDeallocate` vector is corrupted by another thread, and due to this, it was a hang in `Scheduler::deallocateStreamBuffers` function when `StreamBuffersPool[Impl]` had been being deleted.
vladimirlaz
previously approved these changes
Nov 19, 2020
|
There are some surprisingly failed LIT tests caused by this change. I'm working on resolving them. |
|
@vladimirlaz, fixed patch, all LIT tests pass now, updated description. |
vladimirlaz
approved these changes
Nov 23, 2020
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
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Outdated suggestions cannot be applied.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.
This patch fixes hang caused by deadlock in
Scheduler::deallocateStreamBuffersfunction. Deadlock washappened when
StreamBuffersPool[Impl]had been beingdeleted.
StreamBuffersPool[Impl]is a pointer toStreamBuffersstructobject which has
bufferobject as a field, so the dtor ofbufferclass was called. After that the call stack is the following:
~buffer_impl()->SYCLMemObjT::updateHostMemory()->Scheduler::removeMemoryObject->deallocateStreams->Scheduler::deallocateStreamBuffers-> deadlock inlock_guard, mutex was already locked by the same thread.Changed
std::mutextostd::recursive_mutex, so nowthe same thread can enter the critical section, locked by
itself before.