On Thu, Feb 8, 2018 at 11:54 PM, Michael Ellerman <mpe@xxxxxxxxxxxxxx> wrote:
> Alexey Kardashevskiy <aik@xxxxxxxxx> writes:
>
>> Commit 59f47eff03a0 ("powerpc/pci: Use of_irq_parse_and_map_pci() helper")
>> replaced of_irq_parse_pci() + irq_create_of_mapping() with
>> of_irq_parse_and_map_pci() but this change lost virq returned by
>> irq_create_of_mapping() so virq remained zero causing INTx
>> misconfiguration.
>>
>> This fixes pci_read_irq_line() not to loose a virq returned by
>> of_irq_parse_and_map_pci().
>>
>> Fixes: 59f47eff03a0 "powerpc/pci: Use of_irq_parse_and_map_pci() helper"
>> Signed-off-by: Alexey Kardashevskiy <aik@xxxxxxxxx>
>> ---
>> Changes:
>> v2:
>> * changed the condition from <=0 to !=0 as by design
>> of_irq_parse_and_map_pci() can only return 0 for an error and virq>0.
>
> It returns int, so you should store the result in an int, and check the
> result for <= 0.
>
> Otherwise if it starts returning a negative error value this code will
> break.
That won't happen until we're confident all callers expect NO_IRQ is
only 0 and don't treat negative as NO_IRQ. I expect that to be never,
but maybe of_irq_parse_and_map_pci calls can be audited more easily
than of_irq_parse_and_map. We ended up with of_irq_get() because we
need to return error codes.
Rob
