[Patch V1 0/6] Refine generic/PCI MSI irqodmian interfaces

Jiang Liu jiang.liu at linux.intel.com
Thu Nov 13 17:22:30 PST 2014


On 2014/11/14 9:09, Yijing Wang wrote:
> On 2014/11/14 8:25, Jiang Liu wrote:
>> On 2014/11/14 5:00, Marc Zyngier wrote:
>>> On 13/11/14 11:43, Jiang Liu wrote:
>>>> This patch set is based on tip/irq/irqdomain and tries to refine
>>>> interfaces to support irqdomain for generic MSI and PCI MSI.
>>>>
>>>> With this patch set applied, the generic MSI and PCI MSI interfaces
>>>> are much easier to use. For extreme case, you only need to define
>>>> a "struct msi_domain_info" and don't need to implement any callbacks,
>>>> just using the default callbacks is OK:)
>>>>
>>>> This patch set is also a preparation for:
>>>> 1) Kill all weak functions in drivers/pci/msi.c
>>>> 2) Implement support for non-PCI-compliant MSI device
>>>
>>> I've rebased (once more!) the GICv3 ITS driver on top of this, and this
>>> is definitely a major improvement. This is basically the first version
>>> I can use without having to hack into the core code (apart from the
>>> couple of nits I've mentioned earlier).
>> Sorry for the rebasing, but I hope it worthy rebasing:)
>>
>>>
>>> Now, Thomas' idea of putting the irq_domain close to the bus is very
>>> appealing, and I've tweaked an earlier patch in order to do this:
>> I feel that's the right direction. There are other threads discussing
>> associating an MSI controller structure with each PCI bus (at least
>> root bus).
>> http://www.spinics.net/lists/arm-kernel/msg376328.html
> 
> Associate the irq domain and PCI bus is not necessary, because all PCI buses under same host bridge
> always share same MSI chip/irq domain, we only need associate them and pci host bridge.
> I'm refactoring the pci_host_bridge, make it be a generic one, rip out of the pci root bus
> creation, so we could put the irq domain and pci domain etc.. in it. Finally, we could
> eliminate lots platform arch functions. I will post it out within one week.
Hi Yijing,
	Theoretically, it's not true on x86 when interrupt remapping
is enabled. There may be multiple IOMMU(interrupt remapping) units
serving the same host bridge, so we need different MSI domains
to serve different PCI devices.
Regards!
Gerry

> 
> Thanks!
> Yijing.
> 
>> Regards!
>> Gerry
>>
>> .
>>
> 
> 



More information about the linux-arm-kernel mailing list