From owner-freebsd-java Thu Nov 1 7:48:43 2001 Delivered-To: freebsd-java@freebsd.org Received: from smg.is.titech.ac.jp (smg.is.titech.ac.jp [131.112.35.1]) by hub.freebsd.org (Postfix) with ESMTP id 2F01937B405 for ; Thu, 1 Nov 2001 07:48:37 -0800 (PST) Received: from tripper.private (smg [131.112.35.1]) by smg.is.titech.ac.jp (Postfix) with ESMTP id 679E719886; Fri, 2 Nov 2001 00:47:24 +0900 (JST) Date: Fri, 02 Nov 2001 00:52:19 +0900 Message-ID: <55u1wea3po.wl@tripper.private> From: Fuyuhiko Maruyama To: SUGIMURA Takashi Cc: freebsd-java@FreeBSD.org Subject: Re: J2SE 1.3.1 patchset 5 /w JSSE 1.0.2 core dumps In-Reply-To: <20011101.234258.730605272.sugimura@YasudaKei.org> References: <20011101.234258.730605272.sugimura@YasudaKei.org> User-Agent: Wanderlust/2.7.4 (Too Funky) on XEmacs/21.5.3 (asparagus) MIME-Version: 1.0 (generated by SEMI 1.14.3 - "Ushinoya") Content-Type: text/plain; charset=US-ASCII Sender: owner-freebsd-java@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.org On Thu, 01 Nov 2001 23:42:58 +0900 (JST), SUGIMURA Takashi wrote: > > [1 ] > Hello, > > I am developing a commercial network software with FreeBSD by Java, > I often get the core on threaded programs using JSSE 1.0.2. > > When a SSL connection begins, it often makes SIGBUS and SIGABRT > as the attached log. SeedGenerator class is called from java.net.* classes, > and it calls System.gc() to fail. This failure occures not always, > at 1 time of a rate to about 10 times. > > On such a threaded program, can System.gc() be called safely? > And can I avoid this problem? At least with green_threads, System.gc() should never have any problems timing related to timing. So if gc causes a problem, it shows that there is something wrong with FreeBSD's port. Your logs seem to show java.lang.Runtime.gc(Native Method) causes SIGBUS as you think, but it is quite difficult to understand what's exactly happen on your program only with your logs. Can you try to check the stack trace with gdb? $ setenv DEBUG_PROG=gdb $ java_g gdb> r ARGUMENTS_YOU_WANT_TO_GIVE_JAVA_COMMAND ...when SIGBUS is caused gdb> bt > My environment is followings: > > FreeBSD 4.4-RC - 4.4-STABLE (from August 2001 to 22nd November 2001) > J2SE 1.3.1 patchset 3 - 5 (now using only patchset 5) > JSSE (Java Secure Socket Extensions) 1.0.2 > > > Thanks for all, > > --- > SUGIMURA Takashi > GnuPG Key fingerprint = AB4D 2DB4 E305 3ACB A0A8 9417 F4C3 1191 E866 65AC To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-java" in the body of the message