From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <4DA433B7.5020608@domain.hid> Date: Tue, 12 Apr 2011 13:12:55 +0200 From: Gilles Chanteperdrix MIME-Version: 1.0 References: <4D9EB884.6020804@domain.hid> <205D3924-0110-4DA7-838B-DF40BDBE6702@domain.hid> <4D9FB431.6010203@domain.hid> <9948C373-EE10-42D0-A854-DC6A6C38B830@domain.hid> In-Reply-To: <9948C373-EE10-42D0-A854-DC6A6C38B830@domain.hid> Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai-help] Total system lockup doing anything real-time List-Id: Help regarding installation and common use of Xenomai List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Nolan Waite Cc: xenomai@xenomai.org Nolan Waite wrote: > Hello, > > Thanks for the help so far. > > On 2011-04-08, at 7:19 PM, Gilles Chanteperdrix wrote: > >> The question being, do you get the text of a kernel oops when the >> bug happens? Because if this is what happens, what we would like is >> the kernel oops text, and the serial console is the best way to get >> it. If you do not get any message after launching the latency test, >> then forget about the serial console, please try enabling the nmi >> watchdog as explained in Documentation/nmi_watchdog.txt in the >> Linux kernel sources. > > Enabling the nmi watchdog (had to add nmi_watchdog=2 to the boot > command line, not nmi_watchdog=1 like I had expected) got me some > panic text (in attached zip as panic.log) which I got via the > netconsole. Nothing else appeared on the screen after the panic log. That is an improvement. In the kernel configuration. Could you try changing the following configuration options: CONFIG_FTRACE to n CONFIG_SECURITY to n CONFIG_DEBUG_RODATA to n CONFIG_KGDB to n CONFIG_AUDIT to n Could you to try a 64 bits kernel (with the 32 bits compatibility layer, it should work with a 32 bits rootfs)? If you still get a oops, please send us a disassembly of the EIP where the NMI watchdog triggers.2 -- Gilles.