From mboxrd@z Thu Jan 1 00:00:00 1970 From: Diego Calleja Subject: CONFIG_ACPI_PROCESSOR=y confuses the cpu scheduler Date: Thu, 19 Jan 2006 01:20:40 +0100 Message-ID: <20060119012040.733335f4.diegocg@gmail.com> Mime-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Return-path: Received: from uproxy.gmail.com ([66.249.92.207]:1417 "EHLO uproxy.gmail.com") by vger.kernel.org with ESMTP id S1161074AbWASAU7 (ORCPT ); Wed, 18 Jan 2006 19:20:59 -0500 Received: by uproxy.gmail.com with SMTP id s2so27991uge for ; Wed, 18 Jan 2006 16:20:58 -0800 (PST) Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: linux-acpi@vger.kernel.org Cc: linux-kernel@vger.kernel.org If I compile CONFIG_ACPI_PROCESSOR in the kernel, one of the two cpus doesn't get scheduled any process. The CPU works and everything, it services interrupts and I can force processes to run on that CPU with taskset, but they won't get scheduled in that CPU no matter how much processes and load you put on the machine. However, when I compile it as a module everything works fine. I can't say when this started happening; I noticed it in the current linus git tree from a couple of days ago, but testing 2.6.15-rc7 didn't help. The machine is a dual P3 machine. Here's a working dmesg: http://terra.es/personal/diegocg/dmesg-acpi a dmesg of the machine after setting CONFIG_ACPI_PROCESSOR to y: http://terra.es/personal/diegocg/dmesg-acpi-no and a acpidump output: http://terra.es/personal/diegocg/acpidump