From owner-freebsd-current Fri Aug 2 10:43:39 2002 Delivered-To: freebsd-current@freebsd.org Received: from mx1.FreeBSD.org (mx1.FreeBSD.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id 0427037B409 for ; Fri, 2 Aug 2002 10:43:32 -0700 (PDT) Received: from www.contentmedia.de (www.contentmedia.de [213.61.138.91]) by mx1.FreeBSD.org (Postfix) with SMTP id 2900A43E81 for ; Fri, 2 Aug 2002 10:43:31 -0700 (PDT) (envelope-from marc@informatik.uni-bremen.de) Received: (qmail 16562 invoked from network); 2 Aug 2002 17:41:32 -0000 Received: from unknown (HELO leeloo.intern.geht.de) (217.82.121.175) by www.contentmedia.de with SMTP; 2 Aug 2002 17:41:32 -0000 Date: Fri, 2 Aug 2002 19:39:24 +0200 From: Marc Recht To: current@freebsd.org Subject: Re: mysql-3.23.51 with Current for the last few weeks. Message-Id: <20020802193924.3c73ae68.marc@informatik.uni-bremen.de> In-Reply-To: References: <20020802190041.658814fc.marc@informatik.uni-bremen.de> X-Mailer: Sylpheed version 0.8.1 (GTK+ 1.2.10; i386-portbld-freebsd5.0) Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Sender: owner-freebsd-current@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG On Fri, 2 Aug 2002 10:08:47 -0700 (PDT) Julian Elischer wrote: > there are several possibilities: > 1/ do a ktrace on teh spinning process for a couple of seconds and see > what syscalls it is doing. stopped jikes with ctrl-c looks always like: 90998 sh RET read 650/0x28a 90998 sh CALL fork 90998 sh RET fork 91009/0x16381 90998 sh CALL getpgrp 90998 sh RET getpgrp 90998/0x16376 90998 sh CALL wait4(0xffffffff,0xbfbff498,0x2,0) 90998 sh RET wait4 -1 errno 4 Interrupted system call 90998 sh PSIG SIGINT caught handler=0x80595b0 mask=0x0 code=0x0 90998 sh CALL sigreturn(0xbfbff2b8) 90998 sh RET sigreturn JUSTRETURN 90998 sh CALL wait4(0xffffffff,0xbfbff498,0x2,0) 90998 sh RET wait4 91009/0x16381 90998 sh CALL read(0xa,0x80f6e80,0x3ff) 90998 sh GIO fd 10 read 0 bytes "" 90998 sh RET read 0 90998 sh CALL exit(0x82) > 2/ if it's spinning in the kernel, you can drop into ddb > and do a > tr > to see where it is in the kernel... > c to continue.. I always get these two: Debugger("manual escape to debugger") Stopped at Debugger+0x4d: xchgl %ebx,in_Debugger.0 db> tr 3402 mi_switch(c30f36c0,c01f23e8,c30f5cc0,f13c8d3c,c0237b5e) at mi_switch+0x15a ithread_schedule(c30dd500,1,f13c8d3c,c79309c0,0) at ithread_schedule+0x11f sched_ithd(5) at sched_ithd+0x43 Xintr5() at Xintr5+0x67 --- interrupt, eip = 0x80decb0, esp = 0xbfbfe4e0, ebp = 0xbfbfe548 --- db> c mi_switch(c30e3d40,0,5,0,800) at mi_switch+0x15a ast(f16ddd48) at ast+0x35a doreti_ast() at doreti_ast+0x1a > do this several times to ge a sortof 'profile' of where you see it.. Done that. Hope this helps. Please let me know if I could do more to find this problem. Marc To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-current" in the body of the message