From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from gate.crashing.org ([63.228.1.57]:48218 "EHLO gate.crashing.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S932134AbaJGXuX (ORCPT ); Tue, 7 Oct 2014 19:50:23 -0400 Message-ID: <1412725790.30859.214.camel@pasglop> Subject: Re: [PATCH v3 4/7] sound/radeon: Add quirk for broken 64-bit MSI From: Benjamin Herrenschmidt To: Alex Deucher Cc: Bjorn Helgaas , Dave Airlie , Brian King , Takashi Iwai , Linux PCI , Yijing Wang , Anton Blanchard , linuxppc-dev@ozlabs.org Date: Wed, 08 Oct 2014 10:49:50 +1100 In-Reply-To: References: <1412656696.30859.122.camel@pasglop> Content-Type: text/plain; charset="UTF-8" Mime-Version: 1.0 Sender: linux-pci-owner@vger.kernel.org List-ID: On Tue, 2014-10-07 at 19:47 -0400, Alex Deucher wrote: > > While recent ASICs have that problem fixed, they don't seem to > > be listed in the PCI IDs of the current driver, so let's quirk all > > the ATI HDMI for now. The consequences are nil on x86 anyway. > > > > Signed-off-by: Alex Deucher > > Signed-off-by: Benjamin Herrenschmidt > > CC: > > Further discussion with the hw teams have revealed that this is still > an issue on newer asics so I think your original patch is correct > after all. Just disable 64 bit MSIs on all AMD audio PCI ids. Ok. You confirm that this is however good on newer video side right ? Also, is there a known issue that if MSI were once enabled, the chip still shoots them even if we disable them in config space ? When testing my new patch set, I was trying to test the error path when the arch does *not* honor the limitation. The error went back up to the driver as far as I can tell (ie, pci_enable_msi() failed), but the system still got into error state as if the card had tried to shoot an MSI to the wrong address. (I suspect the graphics side) Cheers, Ben.