This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
Home Products Support Community News


Re: [Xen-devel] Re: [PATCH] Modpost section mismatch fix

On Fri, Jul 08, 2011 at 01:18:51AM +0530, Raghavendra D Prabhu wrote:
> * On Thu, Jul 07, 2011 at 12:24:54PM -0400, Konrad Rzeszutek Wilk 
> <konrad.wilk@xxxxxxxxxx> wrote:
> >>Tested it. Works now.
> >
> >Ok, Stuck Tested-by on the patch. Hopefully Linus hasn't pulled it yet.
> Ok.
> >
> >>Reviewed-by: Raghavendra Prabhu <rprabhu@xxxxxxxxxxx>
> >>Also,
> >>The condition for acpi_gsi_to_irq can be removed since it always returns 
> >>zero.
> >
> >The function might in the future return something that is non-zero
> >and we should guard for it. Also you make 'irq' be unsigned which is not
> >good as the IRQ 0 is a valid value - and with making it unsigned if it is
> >set to -1 (the -1 is the invalid IRQ value) the check for 'irq != gsi'
> >will be true and and we will pass in -1 casted to unsigned. That is a
> >large value and not the right thing we want to pass to xen_register_gsi.
> My rationale for the unsigned part was that acpi_gsi_to_irq always
> assigns a positive value (>= 0) to the irq passed (as unsigned
> argument). But even otherwise that shouldn't make much of difference I guess.
> Also,
> I had sent another change (oneline) for the file
> arch/x86/xen/platform-pci-unplug.c for check_platform_magic, looks like that 
> has not gone into
> the pull request for Linus.

Oh, I didn't see it. Did you CC me on it? Can you bounce it to me please?

Xen-devel mailing list