Re: [Patch V3 00/16] Enable support of IOAPIC hotplug on x86 platforms

[Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>]

On Wed, Jul 16, 2014 at 11:01:45AM +0800, Jiang Liu wrote:
> Hi Konrad,

Hey Jiang,
> 	We have found the root cause with your great support. If BIOS
> assigns an IRQ number for a PCI device, acpi_pci_irq_enable() fails to

I would think that is a common scenario?

> assign IOAPIC IRQ then due to the check:
>         if (dev->irq > 0)
>                 return 0;

I should really look in the code to figure this out - but I think your
patch will have a detail analysis that will make tglrx happy, so
will just wait till then :-)

> 	For a quick workaround, just comment out above check. But we
> need some time to figure out a clean fixup for it. We will send out new
> version once it gets resolved.

Excellent. Once you do, please do give me the git tree so I can slurp
it in my testing

> 	Thanks for your great help again.

Of course! Thank you for your quick turn-around.
> Regards!
> Gerry
> On 2014/7/16 0:10, Konrad Rzeszutek Wilk wrote:
> > already assigned IRQ