Data races and the limits of ThreadSanitizer in C and Go

Data races and the limits of ThreadSanitizer in C and Go(theconsensus.dev)

39 points by matt_d 3 days ago | 3 comments

camkego 20 hours ago |

I’ve often thought that single threaded architectures that can eliminate whole classes of bugs are often undervalued.

For example, the command execution design of Redis is purely single threaded. (Redis does have other threads for slower activities)

pjmlp 12 hours ago | |

Because we were already there with multi-processing and OS IPC before mainstream OSes added multi-threading support, and while it is better for stability and security, there are additional hardware resources to account for.

To use modern speak, we were doing microservices inside a single computer.

1over137 3 days ago |

Nice write up! I didn’t know about those TSan limits; it was informative.