From mboxrd@z Thu Jan 1 00:00:00 1970 From: "tj@kernel.org" Subject: Re: ATA failure regression Date: Thu, 29 Sep 2016 13:33:11 +0200 Message-ID: <20160929113311.GA26566@mtj.duckdns.org> References: <8520D5D51A55D047800579B094147198258E4869@XAP-PVEXMBX01.xlnx.xilinx.com> <20160929082229.GA11087@mtj.duckdns.org> <20160929112558.GA24350@khazad-dum.debian.net> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: <20160929112558.GA24350@khazad-dum.debian.net> Sender: linux-pci-owner@vger.kernel.org To: Henrique de Moraes Holschuh Cc: "Shah, Nehal-bakulchandra" , Bharat Kumar Gogada , "Deucher, Alexander" , "linux-pci@vger.kernel.org" , "holler@ahsoftware.de" , "linux-ide@vger.kernel.org" , "linux-kernel@vger.kernel.org" , Dan Williams List-Id: linux-ide@vger.kernel.org Hello, (cc'ing Dan, hi!) On Thu, Sep 29, 2016 at 08:25:58AM -0300, Henrique de Moraes Holschuh wrote: > Actually, according to Shah's post from the 26th to this thread, > everything works just fine even with the IOMMU enabled, as long as the > system is configured to NOT auto-shutdown unused SATA ports. > > That seems to give some extra avenues at fixing/working around the > issue, maybe? Hmm... it could be that the multiple msix routing gets screwed when the bios disables unoccupied ports. Dan, can you please look into this? Shah is seeing ahci failing during boot on an AMD platform with msi enabled. If disabling of unoccupied ports is turned off in bios, the problem goes away. Maybe we're mapping the irqs incorrectly in such cases? Thanks. -- tejun