Serial Synchronization #24

[tommywatson]

Fix for issue #24

I am unable to test it as I don't have any real hardware, I have attached a test program that runs it through short reads and invalid data.

main.c.txt

[johnny9]

Edit: I think it looks to solve the issues.

[tommywatson]

I don't believe that this will resolve the issue. I think you still have an issue when receiving a smaller response everything will be off by a couple of bytes. i think potentially an easier solution might be to tighten the timeout to reduce the chance of getting multiple nonce responses but we also need to handle when we receive partial nonce response.

The test program does random short reads/0 reads and it resyncs with the incoming data.

[johnny9]

I don't believe that this will resolve the issue. I think you still have an issue when receiving a smaller response everything will be off by a couple of bytes. i think potentially an easier solution might be to tighten the timeout to reduce the chance of getting multiple nonce responses but we also need to handle when we receive partial nonce response.

The test program does random short reads/0 reads and it resyncs with the incoming data.

yeah i think i understand it now.

[johnny9]

Concept ACK.

I will try to replicate the issue on actual hardware to verify before merging.

[johnny9]

Please squash commits as well and give a better commit message. Commit header should be clear with what was worked on and in the imperative mood. So something like, Ensure bm1366 serial data is the expected size. Then the description can include more details about the error and why. Issue number should go after the description/body of the commit.

Something like this for commit template.

<scope>: <short summary>

<optional body/description>

<related issue numbers>

[johnny9]

Sorry for the delays here. I'll do my best to get this tested soon. This shouldnt have been stuck in PR hell this long.