alamb commented on PR #9689:
URL: https://github.com/apache/arrow-rs/pull/9689#issuecomment-4329995075

   > Hmm, that's very annoying. I wouldn't have expected such a small 
inline-always helper function to hurt so much.
   > 
   > @alamb do we have any compiler/codegen experts who could weigh in on why 
the regression is so big?
   
   I do not know -- maybe @jhorstmann has some ideas
   
   Sometimes this stuff is super finicky / related to memory layout (e.g. 
getting unlucky with code changes, etc)
   
   it looks like the enchmark is running on a Graviton -- has anyone tried 
profiling directly on a graviton machine?


-- 
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]

Reply via email to