Date: Sat, 8 Nov 2003 14:11:31 +1300 From: Jonathan Chen <jonc@chen.org.nz> To: Alexey Zelkin <phantom@freebsd.org> Cc: java@freebsd.org Subject: Re: jdk14 fork() problem fix Message-ID: <20031108011131.GA59596@grimoire.chen.org.nz> In-Reply-To: <20031105232543.A99393@phantom.cris.net> References: <20031023122839.A75570@phantom.cris.net> <20031105232543.A99393@phantom.cris.net>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wed, Nov 05, 2003 at 11:25:43PM +0200, Alexey Zelkin wrote:
> hi,
>
> Though patch is working it has some limitations. Under high load
> java applications which utilize many parallel fork()'s may hang.
>
> It's reproducible with testcase then least 5 threads calling Runtime.exec()
>
> After finding a reason I had to rethink logic of 'critical section'
> while calling to fork. We can't use pthread_{suspend,resume}_all_np()
> in order to not to create race conditions.
>
> Updated version of patch to this fork problem is attached.
I've tried the updated patch and haven't had any problems for a while
now. I'll report again on a heavier loaded machine next week.
Cheers.
--
Jonathan Chen <jonc@chen.org.nz>
----------------------------------------------------------------------
"You can get farther with a kind word and a gun
than you can with a kind word alone" - Al Capone
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20031108011131.GA59596>
