From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <45DDBB86.2080105@domain.hid> Date: Thu, 22 Feb 2007 16:49:26 +0100 From: Stephan Zimmermann MIME-Version: 1.0 Subject: Re: [Xenomai-help] fata: removing non-linked element... References: <45CC651C.6060402@domain.hid> <45CC68F3.1000003@domain.hid> <45D1C033.9010002@domain.hid> <45D1C61B.7030503@domain.hid> <1171376742.885.3.camel@domain.hid> <45D1CAFA.6050408@domain.hid> <1171385716.885.17.camel@domain.hid> <45D200C3.1030001@domain.hid> <1171393088.885.37.camel@domain.hid> <45D59F98.8020006@domain.hid> <45D5E493.4080405@domain.hid> <45D5F0AD.9010404@domain.hid> <1171649724.10479.12.camel@domain.hid> <45D5F71F.60509@domain.hid> <1171650868.10479.21.camel@domain.hid> <45D5FA33.3030503@domain.hid> <1171651639.10479.24.camel@domain.hid> <45DD9A4E.9070709@domain.hid> <45DD9F8E.6070708@domain.hid> In-Reply-To: <45DD9F8E.6070708@domain.hid> Content-Type: text/plain; charset=ISO-8859-15; format=flowed Content-Transfer-Encoding: 7bit List-Id: Help regarding installation and common use of Xenomai List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: xenomai@xenomai.org Jan Kiszka schrieb: > Stephan Zimmermann wrote: >> Hm, updated my machines to 2.6.20 and 2.3.0 rev 2222. I still see ( less >> frequent?) some backtraces occuring on my console, either during reboot >> or sometimes when running the testprogram and reloading the modules.(see >> dmesg snippet below) The Celeron M still crashes often totally. >> >> My main-application still freezes the SMP mashines if I have 'priority >> coupling' enabled in kernel-config. Using trunk makes things even worse. >> The pc will crash during startup of my app, the testapp I sendt you >> works, but i can not shutdown the pc afterwards, because it freezes >> completely when terminating kdm. >> >> Stephan >> >> >> >> dmesg tells me the following (2.3.0.2222, 2.6.20, SMP): > > Latest I-pipe version? Kernel was freshly patched with: ./ksrc/arch/i386/patches/adeos-ipipe-2.6.20-i386-1.7-02.patch the one from svn rev. 2222 > >> --snip-- >> Bad page state in process 'kswapd0' >> page:c17f3060 flags:0x80010408 mapping:00000000 mapcount:0 count:0 >> Trying to fix it up, but a reboot is needed > > Something is corrupting kernel memory. > > I did some remark on your test, let's get back to it: You are running > the test at 100 kHz. That /may/ have side-effects like overload, or it > might actually trigger some stack overrun. Did you enable the stack > checks? Can you confirm the problem occurs also at slower tick rates? Test was run at 10KHz, but stack-debugging wasn't on, I will repeat it and post again... > Jan >