From mboxrd@z Thu Jan 1 00:00:00 1970 From: "Kevin P. Fleming" Subject: How to begin debugging ACPI/network driver interrupt problem? Date: Wed, 09 Jul 2003 10:34:06 -0700 Sender: netdev-bounce@oss.sgi.com Message-ID: <3F0C520E.2080809@cox.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit Return-path: To: acpi-support@lists.sourceforge.net, netdev@oss.sgi.com Errors-to: netdev-bounce@oss.sgi.com List-Id: netdev.vger.kernel.org I've got a system here using an MSI KT266-based motherboard that's been running 2.5.X successfully for months now. However, 2.5.73/74 have introduced a problem that (so far) I have only been able to cure by either not compiling in ACPI or using acpi=off on the kernel command line. The system three NICs: eth0 is a National Semiconductor DP83815 eth1 is a National Semiconductor DP83820 eth2 is a Lite-On 82C168 All three NICs work fine. However, on the newer kernels, with ACPI compiled in, during initialization of eth0 the kernel generates "IRQ 17 and nobody cared!" messages and disables IRQ 17, which causes eth0 to be useless :-) With older kernels, or ACPI not turned on, eth0 still gets assigned IRQ 17, and works just fine. IRQ 17 is not shared with any other devices. I don't think I can handle the additional message traffic of these two mailing lists, so I haven't subscribed... If anyone can suggest a course of action for how to debug this problem, please cc me on any responses.