logger: changed from RefCell to Mutex. #52

pca006132 · 2020-07-15T17:09:51+08:00

pca006132 commented

2020-07-15 17:09:51 +08:00

The original implementation uses RefCell as core1 was not supposed to log messages, but this is not the case for zynq. This is to fix the possible race condition when two cores attempt to log messages at the same time.

Using mutex may reduce the throughput a bit, but it should not be too slow.
If the logger becomes a bottleneck, we can use multiple buffer for each core and use AtomicPtr to swap out the buffer to send to client while minimizing the downtime for the logger...

The original implementation uses RefCell as core1 was not supposed to log messages, but this is not the case for zynq. This is to fix the possible race condition when two cores attempt to log messages at the same time. Using mutex may reduce the throughput a bit, but it should not be too slow. If the logger becomes a bottleneck, we can use multiple buffer for each core and use `AtomicPtr` to swap out the buffer to send to client while minimizing the downtime for the logger...

sb10q closed this pull request

2020-07-15 17:15:09 +08:00

Sign in to join this conversation.

No reviewers