On 2015/10/27 7:15, jakeo@xxxxxxxxxxxxx wrote:
> From: Jake Oshins <jakeo@xxxxxxxxxxxxx>
>
> The existing PCI code looks for an IRQ domain associated with a root PCI bus
> by looking in the Open Firmware tree. This patch introduces a second way
> to identify the associated IRQ domain, if the lookup in the OF tree fails.
> The handle used for the IRQ domain lookup was introduced in the previous patch
> in the series.
>
> Signed-off-by: Jake Oshins <jakeo@xxxxxxxxxxxxx>
> ---
> drivers/pci/probe.c | 13 +++++++++++++
> 1 file changed, 13 insertions(+)
>
> diff --git a/drivers/pci/probe.c b/drivers/pci/probe.c
> index c0f2e44..62c9ac7 100644
> --- a/drivers/pci/probe.c
> +++ b/drivers/pci/probe.c
> @@ -664,6 +664,7 @@ static void pci_set_bus_speed(struct pci_bus *bus)
> static struct irq_domain *pci_host_bridge_msi_domain(struct pci_bus *bus)
> {
> struct irq_domain *d;
> + struct pci_host_bridge *host_bridge;
>
> /*
> * Any firmware interface that can resolve the msi_domain
> @@ -671,6 +672,18 @@ static struct irq_domain *pci_host_bridge_msi_domain(struct pci_bus *bus)
> */
> d = pci_host_bridge_of_msi_domain(bus);
>
> + /*
> + * If no IRQ domain was found via the OF tree, try looking it up
> + * directly through the fwnode_handle.
> + */
> + if (!d) {
> + host_bridge = to_pci_host_bridge(bus->bridge);
> + if (host_bridge->fwnode) {
> + d = irq_find_matching_fwnode(host_bridge->fwnode,
> + DOMAIN_BUS_ANY);
Should we use DOMAIN_BUS_PCI_MSI instead here?
> + }
> + }
> +
> return d;
> }
>
>
_______________________________________________
devel mailing list
devel@xxxxxxxxxxxxxxxxxxxxxx
http://driverdev.linuxdriverproject.org/mailman/listinfo/driverdev-devel
