From mboxrd@z Thu Jan 1 00:00:00 1970 Message-ID: <4CA1CD00.3080508@domain.hid> Date: Tue, 28 Sep 2010 13:09:52 +0200 From: Gilles Chanteperdrix MIME-Version: 1.0 References: <4C9D0A94.9030501@domain.hid> <4C9DB383.1050606@domain.hid> <4C9E6C21.8010403@domain.hid> <4C9F3163.4050106@domain.hid> <4C9F3F9F.9090509@domain.hid> <4CA0A60D.2010708@domain.hid> <4CA0D305.6030306@domain.hid> <4CA105E3.5090309@domain.hid> In-Reply-To: Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 7bit Subject: Re: [Xenomai-help] switchtest/xeon-test fails: Help needed (Ubuntu/Core i7) List-Id: Help regarding installation and common use of Xenomai List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Gajamohan Mohanarajah Cc: xenomai@xenomai.org, Jan Kiszka Gajamohan Mohanarajah wrote: > Hi Gilles, > > Here is the full test result. > > acpi=off -> works > acpi=hf -> negative (Disables all of ACPI except just enough to enable > Hyper Threading) > pci=noacpi -> negative (Disables ACPI for IRQ routing and PCI scanning) > acpi=noirq -> negative (Disables ACPI for IRQ routing) > pnpacpi=off -> negative (Disables the ACPI component of the Linux Plug > and Play code) > noapic -> negative (Disables the IO-APIC for IRQ routing or PCI scanning) > nolapic -> ***works*** (Disables the local APIC) > > This PC is to control a kuka-lwr arm, with a Orocos/Xenomai setup and > using the Kuka FRI (Fast Research Interface, UDP /1~10 ms sampling > times). What do you think of going ahead with acpi=off. I would like > to hear your comments. I think if Linux alone works with ACPI on and Linux + I-pipe does not, then there is a problem somewhere. We already saw in the past cases where disabling ACPI resulted in IRQ being routed incorrectly, indpendently of the I-pipe patch, (the kind of issue you seem to have). So, setting ACPI to off may result in such issues. The other issue, is that with acpi=off you get 4 cpus detected instead of 8, so you are not using the hardware you have completely. But of course, the choice is yours. -- Gilles.