See JDK-8223777 for details on the glibc bug and the workaround.

I think we can remove this workaround now. From testing and source code I came to the conclusion that it had been fixed in glibc as side effect of

```
4b4d4056bb154603f36c6f8845757c1012758158

Author: Rasmus Villemoes <rv@rasmusvillemoes.dk>
Date: Tue Sep 27 17:52:17 2016 -0700

    linux: spawni.c: simplify error reporting to parent
```

This commit was part of glibc 2.25.

Which major Linux releases with long time support are affected:

Debian:

At the time of this writing (stable = bookworm = 12), old-old-stable already comes with glibc2.28, so Debian won't care.

RHEL:

The oldest RHEL still supported is RHEL 7.9, that runs with the still buggy glibc 2.17, but Java 26 won't target that. RHEL 8 comes with glibc 2.28. So we are good on RHEL, too.

Oracle Linux etc:

Oracle Linux seems to follow the exact same versioning scheme as Red Hat, and the same glibcs are supported. So, oldest glibc with this bug is 2.17 on Oracle Linux 7.9. But I very much doubt Java 26 will need to run on that.

Ubuntu:

Ubuntu 20.04 went EOL this May and already had glibc 2.31.

==============

Therefore I believe its safe to remove the workaround. Note also that getting this wrong and getting this bug is really life threatening; it just makes error analysis a bit more confusing in a rare corner case.

The sendAlivePing solution is also affected by this bug: https://bugs.openjdk.org/browse/JDK-8362267 , so let's remove it instead of just fixing it.