From mboxrd@z Thu Jan 1 00:00:00 1970 From: Len Brown Subject: Re: [PATCH 4/12] PCI PM: Make it possible to force using INTx for PCIe PME signaling Date: Fri, 08 Jan 2010 15:08:03 -0500 (EST) Message-ID: References: <200912272057.10443.rjw@sisk.pl> <20100106135625.14b37f10@jbarnes-piketon> <20100106220214.GA25366@srcf.ucam.org> <201001070000.41807.rjw@sisk.pl> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Return-path: In-reply-to: <201001070000.41807.rjw@sisk.pl> Sender: linux-pci-owner@vger.kernel.org To: "Rafael J. Wysocki" Cc: Matthew Garrett , Jesse Barnes , LKML , pm list , Alan Stern , ACPI Devel Maling List , Linux PCI , Oliver Neukum , Bjorn Helgaas , Shaohua Li , Francois Romieu List-Id: linux-acpi@vger.kernel.org > > > This scares me a little... seems like we're not restoring some state on > > > resume that's needed for MSIs to work? Looks fine though. > > > > Yeah, that was my concern. We debugged this a little on IRC and didn't > > get anywhere, though, so probably need to find someone who knows more > > about MSIs... > > I'd rather say someone who knows more about ICH7. :-) > > MSIs work just fine after resume for the other devices in the same box, > so I'd say it's a chipset issue. Hmmm, do all the ICH7 boxes behave the same way? Yesterday MSI announced a netbook running Suse/Moblin, so there is hope for more technical work w/ MSI to support Linux... It is a bit worrysome to go out of the gate to need a DMI blacklist for relatively new hardware. cheers, Len Brown, Intel Open Source Technology Center