wpleonardo commented on PR #1375: URL: https://github.com/apache/orc/pull/1375#issuecomment-1448186916
Hi @wgtmac , I found that macOS doesn't fully support AVX512. For example, if we want to convert data type uint64_t to double with AVX512 compiler option, there will be "Illegal instruction" error in the running time. macOS also has another issue to support AVX512, and it has a different way dealing with AVX512 than Windows and Linux. We can find the description about it in https://github.com/apple/darwin-xnu/blob/0a798f6738bc1db01281fc08ae024145e84df927/osfmk/i386/fpu.c#L176 By default AVX512 is off in the newly created thread, which means CPUID flags will indicate AVX512 availability, but OS support check (XCR0) will not succeed. AVX512 can be enabled either by calling thread_set_state() or by executing any AVX512 instruction, which would cause #UD exception handled by the OS. So I choose to skip macOS to support AVX512 decode. -- 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]
