From: Ingo Molnar Subject: Re: 2.6.20-rc1-rt1 Panic on reboot after fresh kernel compile Date: Thu, 21 Dec 2006 16:31:45 +0100 Message-ID: <20061221153142.GA19242@elte.hu> References: <458AA2F4.4080601@atl.lmco.com> Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: QUOTED-PRINTABLE Cc: linux-rt-users@vger.kernel.org, "Gautam H. Thaker" Return-path: Received: from mx2.mail.elte.hu ([157.181.151.9]:46969 "EHLO mx2.mail.elte.hu" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1422868AbWLUPej (ORCPT ); Thu, 21 Dec 2006 10:34:39 -0500 To: Jonathan Walsh Content-Disposition: inline In-Reply-To: <458AA2F4.4080601@atl.lmco.com> Sender: linux-rt-users-owner@vger.kernel.org List-Id: linux-rt-users.vger.kernel.org * Jonathan Walsh wrote: > I also saw this yesterday in 2.6.19.1-rt15, retried today with the ne= w=20 > patch and saw the same result. This error only occurs with the RT=20 > patches applied, 2.6.20-rc1 (and 2.6.19.1) boots fine with the same=20 > config. After a clean compilation and install run, I reboot and=20 > receive the following console error message before hanging: >=20 > ... > isapnp: Scanning for PnP cards... > isapnp: No Plug & Play device found > Real Time Clock Driver v1.12ac > Serial: 8250/16550 driver $Revision: 1.90 $ 4 ports, IRQ sharing enab= led > =FFserial8250: ttyS0 at I/O 0x3f8 (irq =3D 4) is a NS16550A > [a few hundred junk characters] > Creating root deKernel panic - not syncing: Attempted to kill init! > [] panic+0x50/0x100 > [] do_exit+0x6d9/0xa50 > [] remove_vma+0x39/0x50 > [] do_munmap+0x19a/0x1f0 > [] do_group_exit+0x34/0xb0 > [] sys_munmap+0x48/0x60 > [] syscall_call+0x7/0xb > [] packet_seq_start+0x2b/0x80 >=20 > This is of course snipped to the last few relevant lines. This is a=20 > reproducible message so I can rerun and give the entire console log i= f=20 > it would be helpful, or if any other information at all would be=20 > useful. yeah, please try to capture a full (serial?) log of the incident. Also=20 please enable CONFIG_UNWIND_INFO and CONFIG_STACK_UNWIND to get easier=20 to read stack-traces. (the packet_seq_start() entry looks really out of= =20 place in the trace above) Ingo