From mboxrd@z Thu Jan 1 00:00:00 1970 From: Bjorn Helgaas Subject: Re: [PATCH v1 2/4] PCI: Mark invalid BARs as unassigned Date: Wed, 15 Apr 2015 18:04:55 -0500 Message-ID: <20150415230455.GB27352@google.com> References: <20150312173201.1052.28341.stgit@bhelgaas-glaptop2.roam.corp.google.com> <20150312173533.1052.46464.stgit@bhelgaas-glaptop2.roam.corp.google.com> <20150415144819.GD19151@google.com> <20150415194226.GA27352@google.com> Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii Return-path: Content-Disposition: inline In-Reply-To: Sender: linux-pci-owner@vger.kernel.org To: Tony Luck Cc: "linux-pci@vger.kernel.org" , linux-acpi , "Rafael J . Wysocki" , Jiang Liu List-Id: linux-acpi@vger.kernel.org On Wed, Apr 15, 2015 at 02:28:08PM -0700, Tony Luck wrote: > On Wed, Apr 15, 2015 at 12:42 PM, Bjorn Helgaas wrote: > > I think we're about to conclude that the Producer/Consumer bit is useless > > and should be ignored. > > > > Can you drop the previous debug patch and try this one? > > That producer/consumer thing sounds horribly familiar ... did we venture along > this path before many years ago? 463eb297401e ("[IA64] respect ACPI producer/consumer flag for PCI root bridges") from 2005 added this check. There's also been recent discussion about Jiang's patches and how to handle that bit in some more generic code [1]. I think if we strictly follow the spec, it is correct to check the producer/consumer bit. If we don't check the bit, we are basically assuming that all Address Space Resource Descriptors are windows that are forwarded downstream, and I don't think there's a way for a host bridge to use those descriptors to describe its own CSR space. But either we haven't figured out how to interpret producer/consumer correctly, or BIOSes haven't used it consistently, and we end up with problems like this one. So I suspect we'll have to give up and just ignore the producer/consumer bit. > I applied the new patch on top of Linus latest (from this morning > HEAD=6c373ca89399c5a3f7ef210ad8f63dc3437da345) without any > reverts. System booted OK with both disks online and networking > functional). This looks better ... I think. Here's the change this patch makes: ACPI: PCI Root Bridge [PCI0] (domain 0000 [bus 00-7f]) acpi PNP0A08:00: host bridge window [io 0x0000-0x0cf7] acpi PNP0A08:00: host bridge window [io 0x1000-0x8fff] +acpi PNP0A08:00: host bridge window [mem 0x000a0000-0x000bffff] +acpi PNP0A08:00: host bridge window [mem 0x000c0000-0x000fffff] +acpi PNP0A08:00: host bridge window [mem 0xfec00000-0xfec3ffff] +acpi PNP0A08:00: host bridge window [mem 0xfed1c000-0xfed1c0ff] +acpi PNP0A08:00: host bridge window [mem 0xfed40000-0xfedfffff] acpi PNP0A08:00: host bridge window [mem 0x50000000-0x9fffffff] +acpi PNP0A08:00: host bridge window [mem 0x10000000000-0x100fffffffe] +acpi PNP0A08:00: host bridge window [mem 0x10100000000-0x101fffffffe] +acpi PNP0A08:00: host bridge window [mem 0x10200000000-0x102fffffffe] +acpi PNP0A08:00: host bridge window [mem 0x10300000000-0x103fffffffe] ACPI: PCI Root Bridge [PCI1] (domain 0000 [bus 80-ff]) acpi PNP0A08:01: host bridge window [io 0x9000-0xfffe] +acpi PNP0A08:01: host bridge window [mem 0xfec40000-0xfec7ffff] +acpi PNP0A08:01: host bridge window [mem 0xa0000000-0xefffffff] +acpi PNP0A08:01: host bridge window [mem 0x10400000000-0x104fffffffe] +acpi PNP0A08:01: host bridge window [mem 0x10500000000-0x105fffffffe] +acpi PNP0A08:01: host bridge window [mem 0x10600000000-0x106fffffffe] +acpi PNP0A08:01: host bridge window [mem 0x10700000000-0x107fffffffe] The addition of the [mem 0xa0000000-0xefffffff] window for PCI1, which we previously ignored, is what makes your igb and mptsas devices work again. It concerns me a little bit that the BIOS apparently did set the producer bit for the PCI0 [mem 0x50000000-0x9fffffff] window, but not for any of the other windows, not even the ones that fit in 32 bits. But maybe that inconsistency is just a little BIOS lint that doesn't mean anything. Anyway, I guess I'll package up this patch and post it. Bjorn [1] http://lkml.kernel.org/r/20150404030411.GG10892@google.com