From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <54F0D46F.1070006@siemens.com> Date: Fri, 27 Feb 2015 21:32:47 +0100 From: Jan Kiszka MIME-Version: 1.0 References: <54EEF08B.6040905@triphase.com> <20150226102010.GA24003@hermes.click-hack.org> <54EF0790.3040607@triphase.com> <54F07AC2.6000902@triphase.com> In-Reply-To: <54F07AC2.6000902@triphase.com> Content-Type: text/plain; charset=windows-1252 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai] xeno3_rc3 - Watchdog detected hard LOCKUP List-Id: Discussions about the Xenomai project List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Niels Wellens , xenomai@xenomai.org On 2015-02-27 15:10, Niels Wellens wrote: > >> >>>> Hi, >>>> >>>> Yesterday I installed a Core-i7 based pc (4770 CPU - C226 chipset) with >>>> software configuration: >>>> - Debian Jessie RC1 >>>> - kernel 3.16.0-rc7 (.config in attachment) >>> Well, please try again with a kernel which is a real release and not >>> a release candidate, to rule out any mainline kernel issue. >>> -- >>> Gilles. >> >> Thanks Gilles, sorry I overlooked the fact that I was using an rc. The >> system is now up and running with an ipipe patched 3.16.0 kernel, I >> will let you know if I experience the same issues with this kernel. >> -- >> Niels > > Hi, > > I experience the same kind of lockup with the ipipe patched 3.16.0 > kernel (used repo: > git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable.git). > I first thought the problem was solved but after more than 20 hours of > operation (latency test + dohell load) it occurred again. This afternoon > it occurred again after 1 hour of operation (syslog + config file used > during kernel build in attachement). > > Any advice on what to test/change next? Your machine gets stuck on a spinlock involved in page allocations, it seems. We need to find out who's holding it and why he's not releasing it anymore. I'll try to reproduce the issue with your config next week. If you want to collect further data: CONFIG_FRAME_POINTER may give better backtraces. You could also try to send a l (show-backtrace-all-active-cpus) via the serial console to the system. Maybe we can get a better pictures that way. Jan -- Siemens AG, Corporate Technology, CT RTC ITP SES-DE Corporate Competence Center Embedded Linux