RFR: 8293067: (fs) Implement WatchService using system library (macOS) [v10]

[Jaikiran Pai]

On Tue, 29 Nov 2022 10:47:03 GMT, Maxim Kartashev <mkartashev at openjdk.org> wrote:

> I just noticed a curious line in the documentation for `FSEventStreamStart()`:
> 
> > It ought to always succeed, but in the event it does not then your code should fall back to performing recursive scans of the directories of interest as appropriate.
> 
> So I added a check for the return value of `FSEventStreamStart()`; if anyone who has seen tests failing (@AlanBateman @jaikiran) could please re-run some of them to see if an `InternalError` gets thrown, we can at least establish that the fallback mechanism the documentation talks about actually has to be implemented, which will give us a way to move forward.
> 
> Thanks in advance!

I gave this latest state of the PR a try and the tests continue to fail across various versions of the macos, like previously. I even noticed a JVM crash:


#
# A fatal error has been detected by the Java Runtime Environment:
#
#  SIGSEGV (0xb) at pc=0x00007fff206a1b2e, pid=xxxx, tid=xxxx
#
# JRE version: Java(TM) SE Runtime Environment (20.0) (build 20-internal-2022-11-29)
# Java VM: Java HotSpot(TM) 64-Bit Server VM (20-internal-2022-11-29, mixed mode, sharing, tiered, compressed oops, compressed class ptrs, g1 gc, bsd-amd64)
# Problematic frame:
C [CoreFoundation+0x14ab2e]  _CFGetNonObjCTypeID+0xa

(it could be that the JVM crash was present even in previous version of this PR before your latest commit - I hadn't looked into all log files).

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

PR: https://git.openjdk.org/jdk/pull/10140