Possible fix for audio.onComplete #657
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
…nd buffer and assignment logic Next up, prepare details from graphics.defineEffect()
…from Program Slight redesign of TimeTransform with caching (to synchronize multiple invocations with a frame) and slightly less heavyweight call site
Err, was calling Modulo for sine method, heh That said, time transform seems to work in basic test
Fixed error detection for positive number time transform inputs Relaxed amplitude positivity requirement
Maintenance Revert Test
My mistake adding back changes made
This reverts commit 0b5e1e9.
Shchvova
approved these changes
Dec 5, 2023
|
I've converted the scheduler to use atomic variables rather than the If this is needed sooner one of us can make a PR. There are other issues: the Lua state being used on the audio thread, or rather the smart pointer to it, is not safe at all. The cleanup of |
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Outdated suggestions cannot be applied.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.
I took the sample in this issue and seemed to pin down where the crash itself was happening. A description and analysis are mentioned in my comments here.
The comment here, right before the task was issued, was on the right track. The
Append()is being performed in the audio thread but theRun()is happening in the main thread. Since there was no synchronization anAppend()would occasionally happen during aRun().This adds a couple mutexes to guard against this. (
Rtt_Schedulerseems to be a singleton, so I made themstaticrather than add<mutex>in the header file, but they could just as well be astaticor even regular member.)Since you could have patterns like in the test, I added
Tasks to a list and synchronized them at certain points. That way theRun()shouldn't be stallingAppend()s repeatedly. One possible issue is thatDelete()could miss between extracting the list and synchronizing it, due to the two mutexes, but I didn't see any evidence of its use in the codebase. (Looks like apart from playing sounds, only the shell executing at startup and audio recording use the scheduler.) I think the sync in the destructor will account for any last-minute additions, since it happens in the same thread asRun().