From owner-freebsd-stable Wed Aug 30 15:36:58 2000 Delivered-To: freebsd-stable@freebsd.org Received: from duke.cs.duke.edu (duke.cs.duke.edu [152.3.140.1]) by hub.freebsd.org (Postfix) with ESMTP id DBEFA37B423 for ; Wed, 30 Aug 2000 15:36:54 -0700 (PDT) Received: from grasshopper.cs.duke.edu (grasshopper.cs.duke.edu [152.3.145.30]) by duke.cs.duke.edu (8.9.3/8.9.3) with ESMTP id SAA18705; Wed, 30 Aug 2000 18:36:53 -0400 (EDT) Received: (from gallatin@localhost) by grasshopper.cs.duke.edu (8.11.0/8.9.1) id e7UMarr08125; Wed, 30 Aug 2000 18:36:53 -0400 (EDT) (envelope-from gallatin@cs.duke.edu) From: Andrew Gallatin MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Date: Wed, 30 Aug 2000 18:36:53 -0400 (EDT) To: Mike Tancsa Cc: freebsd-stable@FreeBSD.ORG Subject: Re: 4.1 STABLE broken since today! In-Reply-To: <4.3.2.7.0.20000830170456.0397c740@marble.sentex.ca> References: <4.3.2.7.0.20000830170456.0397c740@marble.sentex.ca> X-Mailer: VM 6.43 under 20.4 "Emerald" XEmacs Lucid Message-ID: <14765.35667.455203.449245@grasshopper.cs.duke.edu> Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.ORG Mike Tancsa writes: > At 10:52 PM 8/30/00 +0200, O. Hartmann wrote: > >I received this afternoon the latest cvsupdate from cvsup.freebsd.org, > >made a buildworld, made an installworld, compiled a new kernel. The kernel > >on one machine booted fine, but then the keyboard stuck. No inputs ... but > >I could change the ttys by ALT-FX key. > > > >The other machine, our main server, worked fine for a while - the it > >crashed without > >any message and rebooted. Nice! I booted again. Tried ssh or telnet - no > >response, the > >other machine was dead. I booted with the old kernel and all things were > >fine! SSH > >worked, telnet worked perfectly. What's up with the Stable release? > > Wow! I am seeing the exact same behaviour. I thought it was just hardware > going bad or something :-( "me too" It happened prior to 4amEDT today, which is when my cvsup script runs. Can you bracket it better on the other end that 4amEDT on Aug 28th? Here's my description of the problem: A -stable machine with a new kernel will lock solid when setting up the first incoming network connection. If I break into the debugger, the curproc is always the relevant daemon (sshd or telnetd, for example). The callchain is always some variation on: db> tr siointr1(c0e2b000,c8b2ee5c,c02ddae6,c0e2b000,10) at siointr1+0xb1 siointr(c0e2b000,10,c8b2ee70,10,0) at siointr+0xb Xfastintr4(c7f8d780,0,c8b2eedc,0,0) at Xfastintr4+0x16 soo_write(c0eb7500,c8b2eedc,c0b1a880,0,c8145a40) at soo_write+0x24 dofilewrite(c8145a40,c0eb7500,0,805b080,3) at dofilewrite+0xb1 write(c8145a40,c8b2ef80,3,ffffffff,805bc2c) at write+0x33 syscall2(2f,2f,2f,805bc2c,ffffffff) at syscall2+0x1f1 Xint0x80_syscall() at Xint0x80_syscall+0x25 db> Sometimes I also see as much as the following above soo_write(): db> tr siointr1(c0e2b000,c8b2edc4,c02ddae6,c0e2b000,c0190010) at siointr1+0xb1 siointr(c0e2b000,c0190010,678490,c7f8d7f4,620000) at siointr+0xb Xfastintr4(c0b47700,c80b2d80,0,0,0) at Xfastintr4+0x16 sbcompress(c7f8d7f4,c0b47700,0,c8b2ee18,c01ece04) at sbcompress+0x54 sbappend(c7f8d7f4,c0b47700) at sbappend+0x4b tcp_usr_send(c7f8d780,0,c0b47700,0,0) at tcp_usr_send+0xc0 sosend(c7f8d780,0,c8b2eedc,c0b47700,0) at sosend+0x5df soo_write(c0eb7500,c8b2eedc,c0b1a880,0,c8145a40) at soo_write+0x24 dofilewrite(c8145a40,c0eb7500,0,805b080,3) at dofilewrite+0xb1 write(c8145a40,c8b2ef80,3,ffffffff,805bc2c) at write+0x33 syscall2(2f,2f,2f,805bc2c,ffffffff) at syscall2+0x1f1 Xint0x80_syscall() at Xint0x80_syscall+0x25 db> c A breakpoint in mi_switch() never triggers, so its probably really stuck looping somewhere in that callchain. ------------------------------------------------------------------------------ Andrew Gallatin, Sr Systems Programmer http://www.cs.duke.edu/~gallatin Duke University Email: gallatin@cs.duke.edu Department of Computer Science Phone: (919) 660-6590 To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message