maxburke opened a new pull request #7693: URL: https://github.com/apache/arrow/pull/7693
This issue also applies to 32-bit integers miscast to 64-bit integers, however on little endian machines the cast + truncation results in a correct value. In the float32 case, the mis-read 64-bit float is denormal and so when it's converted to 32-bit it becomes zero. This only occurs in the case where there's one (4-byte) element. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
