felipecrv commented on PR #40647: URL: https://github.com/apache/arrow/pull/40647#issuecomment-2008524642
> > Does it actually change anything for x86? > > Wouldn't acq-rel cheaper than original total-ordering? Only on architectures with a weaker memory model (e.g. ARM). x86 guarantees that all the stores are ordered (not immediately visible, but they are ordered). [1] explains this much better than I ever could. `acq-rel` can enable the *compiler* to re-order operations if it decides that can unlock optimizations. Not the case here, so I'm manually experimenting with different orderings to mask the latency of fetch/load operations on atomic variables. [1] https://research.swtch.com/hwmm#x86 -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
