You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We had an internal issue (T22633912) which manifested itself as an infinite nested update loop triggering our invariant.
@trueadm investigated that it was introduced in #10426 and then fixed in #11187. We don’t, however, have any failing tests related to that.
Even though we removed that coalescing implementation, I feel like we need to investigate what happened and add a regression test to prevent the same issue from occurring in the future.
It might also be valuable to think about any additional information we can surface in the error, at least in development. For example, component names for last ten fibers that scheduled an update.
The text was updated successfully, but these errors were encountered:
If this has anything to do with #11113 or a "Maximum update depth exceeded .." error, than I think it might.
To give a brief summary of the physical symptoms I see when encountering my problem, my component essentially tries to setState when it enters or leaves the viewport, when the onLeave event fires, it will completely freeze up the entire browser and become unresponsive, starting to utilize a significant amount of memory until it hits the upper limit of events, then the error will be logged. the duration of this freeze and the logging of the error varies from seemingly never to a few seconds.
I figured they might be related due to the infinite-loop nature of my own problem and the fact that I haven't experienced this problem previously.
@WardenUnleashed Infinite loop bugs could be caused by any number of things. Could you open a separate issue with more details, and ideally an isolated repro case?
We had an internal issue (T22633912) which manifested itself as an infinite nested update loop triggering our invariant.
@trueadm investigated that it was introduced in #10426 and then fixed in #11187. We don’t, however, have any failing tests related to that.
Even though we removed that coalescing implementation, I feel like we need to investigate what happened and add a regression test to prevent the same issue from occurring in the future.
It might also be valuable to think about any additional information we can surface in the error, at least in development. For example, component names for last ten fibers that scheduled an update.
The text was updated successfully, but these errors were encountered: