RFR: 8292016: Split Windows API error handling from errors passed through the runtime in the JDK [v22]

[Julian Waters]

On Tue, 27 Sep 2022 04:22:59 GMT, David Holmes <dholmes at openjdk.org> wrote:

> This seems to be trying to future-proof the reporting code by introducing a duality that is not actually needed at present in some cases. That means the code is untested. Are there planned uses of `RUNTIME` in the very near future?

The only use of it is to differentiate Windows API errors from the regular errno ones thus far, I doubt this will change any time soon. Honestly speaking I'm not very happy with the changes here, and it's been pretty tricky trying to come up with an elegant solution to this issue, but I'm not sure if the alternative of deleting all the error reporting utilities across the JDK and using what C already has (for instance perror) is acceptable in this case, because pretty much all of them have this same problem (and is also why I refrained from outright removing JLI_ReportErrorMessageSys initially like Thomas suggested). On the flip side, all of them don't seem to be used in _too many_ places, and can be rather easily replaced, so it is a little tempting to remove them and just sidestep the problem instead. Are any of these a part of the public JNI API?

-------------

PR: https://git.openjdk.org/jdk/pull/9870