On Fri, 28 Feb 2025 at 09:18, Michael Bien <mbie...@gmail.com> wrote: > > On 24.02.25 10:43, Neil C Smith wrote: > > On Sun, 23 Feb 2025 at 01:35, Michael Bien <mbie...@gmail.com> wrote: > >> On 22.02.25 18:23, Jerome Lelasseux wrote: > >>> Is it the problem mentionned in the release notes "Windows/ARM is > >>> currently not a fully supported platform" ? BTW this message is really > >>> too vague to be useful... > >> it is intentionally kept vague since we have close to no coverage there > >> but know that a few things will likely not work. > >> > >> Make sure to file win/arm specific issues and consider help testing PRs > >> like https://github.com/apache/netbeans/pull/8224, participate in RC > >> testing etc. > > Well, afaik, what does work only works at all by accident, as we don't > > compile the launcher for aarch64. > > yep. I think the reason NB is starting at all is because the launcher is > emulated until the JVM process spawns. However any native lib which is loaded > by that (arm) JVM will likely fail to load as we can see in #8218.
I think it's a little more involved than that. It's been a while since I looked at the launcher code much, but as far as I remember it tries to launch the JVM in-process rather than spawning a new one. There's some fallback to spawning a new process to handle 32-bit / 64-bit switching, which should get logged somewhere. I assume that's also enabling the JVM process to launch from the emulated code in this case. How reliable that is I don't know. Best wishes, Neil --------------------------------------------------------------------- To unsubscribe, e-mail: users-unsubscr...@netbeans.apache.org For additional commands, e-mail: users-h...@netbeans.apache.org For further information about the NetBeans mailing lists, visit: https://cwiki.apache.org/confluence/display/NETBEANS/Mailing+lists
