On Wed, Jun 12, 2013 at 10:09:10AM -0600, Stephen Warren wrote: > On 06/12/2013 06:30 AM, Thierry Reding wrote: > > On Mon, Apr 15, 2013 at 12:28:12PM -0600, Stephen Warren wrote: > >> On 04/03/2013 08:45 AM, Thierry Reding wrote: > >>> Move the PCIe driver from arch/arm/mach-tegra into the > >>> drivers/pci/host directory. The motivation is to collect > >>> various host controller drivers in the same location in order > >>> to facilitate refactoring. > >>> > >>> The Tegra PCIe driver has been largely rewritten, both in order > >>> to turn it into a proper platform driver and to add MSI (based > >>> on code by Krishna Kishore <kth...@nvidia.com>) as well as > >>> device tree support. > ... > >>> +static irqreturn_t tegra_pcie_msi_irq(int irq, void *data) > >> ... > >>> + return IRQ_HANDLED; > >> > >> Shouldn't this function return IRQ_NONE if no MSI status bits > >> were found set? > > > > The IRQ isn't marked IRQF_SHARED, so I don't think this is needed. > > Isn't it still useful to detect unexpected/stuck interrupts?
Yes, you're right. I can't think of a nicer way to do it, though, so I'll go with a processed IRQ counter or a flag. > >>> +static int tegra_pcie_probe(struct platform_device *pdev) > >> ... > >>> + pcibios_min_mem = 0; > >> > >> What does that mean/do? I wonder if that should be set to > >> 0x80000000 by the Tegra30 patches? > > > > ARM defines PCIBIOS_MIN_MEM to that variable. That macro in turn is > > only used by pci_bus_alloc_resource() AFAICT, which uses it to > > override the start of a resource when allocating if res->start == > > 0. As such it designates a lower-bound of valid PCI memory > > addresses, so 0 on Tegra20 and 0x80000000 on Tegra30 don't seem > > like good values. Maybe we need to set them to the lowest of the > > prefetchable and non-prefetchable memory areas as defined in the > > DT? > > > > It doesn't currently seem to matter at all, though, since we never > > pass in a range that's 0, so the start address of resources can > > never be 0 and therefore PCIBIOS_MIN_MEM is never used. > > Hmmm. I guess ignore it then. If the value won't ever be used, 0 is as > good a value as any? Alright, I'll do that then. Thierry
pgpFR6AxcDPmB.pgp
Description: PGP signature
_______________________________________________ devicetree-discuss mailing list devicetree-discuss@lists.ozlabs.org https://lists.ozlabs.org/listinfo/devicetree-discuss