Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

As far as I understand, performance is not a valid concern.

For example, from dvyukov's KTSAN wiki [1]:

  Given a sufficiently expressive atomic API and a good
  implementation, you pay only for what you really need (if
  you pay just a bit less, generated code becomes incorrect).
  So performance is not an argument here.
[1] https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONC...


You're talking about the sanitizer performance. We were talking about the kernel performance in the real world. Fixing these race conditions will have some effect.


No, that page is about kernel performance, not sanitizer performance.

The idea is that if you omit READ_ONCE/WRITE_ONCE when they are needed, it doesn't matter if the code is faster because it is wrong.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: