Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 23 Oct 2019 06:37:37 +0000
From:      bugzilla-noreply@freebsd.org
To:        mono@FreeBSD.org
Subject:   [Bug 198727] [PATCH] lang/mono FileSystemWatcher (kevent) deadlock problem
Message-ID:  <bug-198727-23699-kIAfnCQNV9@https.bugs.freebsd.org/bugzilla/>
In-Reply-To: <bug-198727-23699@https.bugs.freebsd.org/bugzilla/>
References:  <bug-198727-23699@https.bugs.freebsd.org/bugzilla/>

next in thread | previous in thread | raw e-mail | index | archive | help
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D198727

perry3035 <perry3035@gmail.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
                 CC|                            |perry3035@gmail.com

--- Comment #34 from perry3035 <perry3035@gmail.com> ---
There were 2 issues with mono's unique execution of kevent FileSystemWatche=
r on
FreeBSD:=20

* FileSystemWatcher usage in mono's HEAD for kevent was depending on Darwin=
's
particular conduct (when kevent is executed with NULL break, if kqueue FD is
shut kevent approach Darwin comes back with mistake, anyway FreeBSD doesn't
carry on that way - in truth as indicated by criticism I got from bit desig=
ners
on FreeBSD close would square hanging tight for kevent to finish therefore
causing gridlock)=20

* All watcher occasions were produced from a similar string which recognized
them and that exhibited plausibility of halt if change handler was to
cripple/reset watcher (this showed in ASP.Net applications not being restar=
ted
after dll changes, and not recompiling singular parts on aspx/ascx document
changes)=20

Connected fix fixes those issues (I had the option to check that in the wak=
e of
applying it XSP and mono's fastcgi servers act in anticipated manner).=20

Joined fix completes a few things:=20

* renames individual from imported timespec structure from deluding tv_usec=
 to
appropriate tv_nsec (esteem is in nanoseconds, not microseconds)=20

* includes watcher string prematurely end on the off chance that it doesn't
effortlessly shutdown inside 2 seconds (I think not so much required, howev=
er
better to be as cautious as possible)=20

* moves real occasion dispatching to separate string (accordingly killing
potential for halt portrayed before)=20

Connected fix is to be applied on most recent rendition of port as of now
(3.12.1)=20

https://www.britishassignmentwriters.co.uk

--=20
You are receiving this mail because:
You are the assignee for the bug.=



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?bug-198727-23699-kIAfnCQNV9>