[GitHub] [arrow] zagto commented on pull request #12699: ARROW-16013: [C++][Python] Signed overflow when using negative stride in NumPyStridedConverter

[GitBox <gi...@apache.org>]

zagto commented on pull request #12699:
URL: https://github.com/apache/arrow/pull/12699#issuecomment-1077532368


   Not in a way I know of.
   - To reliably break it, we could use it on a numpy type, that is not a power of 2 in sizeof(). But that does not seem to exist x86_64. But for example long double on 32-bit x86 is 96-bit sizeof() (?)
   - As long is the size is a power of two it depends on the behavior of signed overflow and the binary representation of negative numbers to work/break, which happens to make it work on common platforms.
   - It could still break because of compiler optimizations, but then again that is something we can't really control in a test


-- 
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: github-unsubscribe@arrow.apache.org

For queries about this service, please contact Infrastructure at:
users@infra.apache.org