From mboxrd@z Thu Jan 1 00:00:00 1970 From: Matthew Garrett Subject: [Regression] GPEs being enabled inappropriately Date: Fri, 10 Dec 2010 17:55:05 +0000 Message-ID: <20101210175505.GA12319@srcf.ucam.org> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Received: from cavan.codon.org.uk ([93.93.128.6]:33270 "EHLO cavan.codon.org.uk" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1756384Ab0LJRzL (ORCPT ); Fri, 10 Dec 2010 12:55:11 -0500 Content-Disposition: inline Sender: linux-acpi-owner@vger.kernel.org List-Id: linux-acpi@vger.kernel.org To: linux-acpi@vger.kernel.org Cc: rjw@sisk.pl I've just had a bug report from someone who's seeing high GPE-triggered load in .37. It turns out that GPEs 1B and 1E are constantly firing. Both have associated _L1B and _L1E methods, but they're also present in _PRW methods s should be disabled by default. However, part of the GPE rework means that we're now ignoring _PRW methods if the device they're associated with reports itself missing in _STA. I think we need to pay attention to _PRW regardless of whether or not the device is present, as otherwise we may be enabling GPEs that are tied to GPIOs that aren't connected to anything at the other end. -- Matthew Garrett | mjg59@srcf.ucam.org