Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 7 Oct 2002 02:05:13 -0700
From:      Bill Huey (Hui) <billh@gnuppy.monkey.org>
To:        Lutz Bichler <Lutz.Bichler@UniBw-Muenchen.de>
Cc:        freebsd-java@freebsd.org, freebsd-current@freebsd.org, "Bill Huey (Hui)" <billh@gnuppy.monkey.org>
Subject:   Re: Compilation of jdk with native threads failes
Message-ID:  <20021007090513.GA7539@gnuppy.monkey.org>
In-Reply-To: <200210070935.21180.Lutz.Bichler@unibw-muenchen.de>
References:  <200210070935.21180.Lutz.Bichler@unibw-muenchen.de>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, Oct 07, 2002 at 09:35:21AM +0200, Lutz Bichler wrote:
> I cannot find the CTX_ constants and/or their meaning. Any hints? 

I ran into this myself and it's because that stuff was delete recently
in -current's libc_r. Another patch release needs to happen because of
that to solve that problem. I have some changes in my tree, but it'll
break the stuff that's suppose to work in -stable too, which is not
recommend to be used.

My suggestion is to just comment all of that stuff out so that it'll
compile and use the HotSpot JIT instead. HotSpot is that only thing
that really works anyways, so you're not losing anything essential
by removing the ability to run -classic. -classic is dead anyways
for client/server side stuff.

The interruptable syscall (usleep(), read(), etc...) framework also
needs to be reintegrated into HotSpot, since programs like Tomcat3
do funny thing with Thread.sleep(). Having a non-interruptable usleep()
causes what looks like funny performance related problems, even though
our JIT compiler is pretty severly jamming and is as good as "gcc -O0"
for stuff like Sieve calculations.

bill


To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-current" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20021007090513.GA7539>