linux-kernel - Re: [PATCH v2 2/2] MIPS: PCI: Pass controller's resources to pci_create_bus() in pcibios

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CAErSpo5n-FJXwOd2h179tctw2LX0bZ7xcyxU5Mam+tmVoqgprw@mail.gmail.com>
Date:	Thu, 25 Aug 2011 10:09:48 -0600
From:	Bjorn Helgaas <bhelgaas@...gle.com>
To:	Deng-Cheng Zhu <dczhu@...s.com>
Cc:	jbarnes@...tuousgeek.org, ralf@...ux-mips.org,
	linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
	linux-mips@...ux-mips.org, eyal@...s.com, zenon@...s.com,
	dengcheng.zhu@...il.com
Subject: Re: [PATCH v2 2/2] MIPS: PCI: Pass controller's resources to
 pci_create_bus() in pcibios_scanbus()

On Thu, Aug 25, 2011 at 5:18 AM, Deng-Cheng Zhu <dczhu@...s.com> wrote:
> Use the new interface of pci_create_bus() so that system controller's
> resources are added to the root bus upon bus creation, thereby avoiding
> conflicts with PCI quirks before pcibios_fixup_bus() gets the chance to do
> right things in pci_scan_child_bus(). Also, since we are passing resources
> to pci_create_bus() and setting them up in the bus->resources list as
> opposed to the bus->resource[] array, we have to adopt the list style while
> doing bus fixups later on, or else, for example in this MIPS case, system
> controller's resources will stay in both bus->resources and
> bus->resource[].

> +static void __devinit
> +pcibios_setup_root_resources(struct pci_bus *bus, struct pci_controller *ctrl)
> +{
> +       pci_bus_remove_resources(bus);
> +       pci_bus_add_resource(bus, ctrl->io_resource, 0);
> +       pci_bus_add_resource(bus, ctrl->mem_resource, 0);
> +}
> +
>  void __devinit pcibios_fixup_bus(struct pci_bus *bus)
>  {
>        /* Propagate hose info into the subordinate devices.  */
> @@ -269,10 +313,9 @@ void __devinit pcibios_fixup_bus(struct pci_bus *bus)
>        struct list_head *ln;
>        struct pci_dev *dev = bus->self;
>
> -       if (!dev) {
> -               bus->resource[0] = hose->io_resource;
> -               bus->resource[1] = hose->mem_resource;
> -       } else if (pci_probe_only &&
> +       if (!dev)
> +               pcibios_setup_root_resources(bus, hose);

As I mentioned in my other response, I think you can just drop this
whole "if (!dev)" block since the root bus resources should already be
correct.  There's no need to move them from the bus->resource[] array
to the bus->resources list.

Someday, if all arches adopt your nice pci_create_bus() extension to
make root bus resources correct from the beginning, there should be no
arch references to the bus->resource[] array left, and then we can
remove it altogether.  At least, that's what I was hoping when I added
the list :)

In any case, all the *readers* of bus resources already use the
pci_bus_for_each_resource() interface, which knows how to look at both
the array and the list.

Bjorn
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/