Re: [PATCH] ACPI, PCI: Get PRT entry during acpi_pci_enable_irq()

[Bjorn Helgaas]

On Fri, Feb 15, 2013 at 6:37 PM, Yinghai Lu <yinghai@xxxxxxxxxx> wrote:
> On Fri, Feb 15, 2013 at 5:26 PM, Yinghai Lu <yinghai@xxxxxxxxxx> wrote:
>> On Fri, Feb 15, 2013 at 4:39 PM, Bjorn Helgaas <bhelgaas@xxxxxxxxxx> wrote:
>>> On Thu, Feb 14, 2013 at 5:50 PM, Yinghai Lu <yinghai@xxxxxxxxxx> wrote:
>>>> On Tue, Feb 12, 2013 at 12:22 PM, Rafael J. Wysocki <rjw@xxxxxxx> wrote:
>>>>> On Tuesday, February 12, 2013 11:11:23 AM Yinghai Lu wrote:
>>>>>> Peter Hurley found "irq 18 nobody cared" with pci-next, and dmesg has
>>>>>>
>>>>>> [    8.983246] pci 0000:00:1e.0: can't derive routing for PCI INT A
>>>>>> [    8.983600] snd_ctxfi 0000:09:02.0: PCI INT A: no GSI - using ISA IRQ 5
>>>>>>
>>>>>> bisect to
>>>>>> | commit 4f535093cf8f6da8cfda7c36c2c1ecd2e9586ee4
>>>>>> |     PCI: Put pci_dev in device tree as early as possible
>>>>>>
>>>>>> It turns out we need to call acpi_pci_irq_add_prt() after the pci bridges
>>>>>> are scanned.
>>>>>>
>>>>>> Bjorn said:
>>>>>>  The bus number binding means acpi_pci_irq_add_prt() has to happen
>>>>>>  after enumerating everything below a bridge, and it will prevent us
>>>>>>  from doing any bus number reassignment for hotplug.
>>>>>>
>>>>>>  I think we should remove the bus numbers from the cached _PRT (or
>>>>>>  maybe even remove the _PRT caching completely).  When we enable a PCI
>>>>>>  device's IRQ, we should search up the PCI device tree looking for a
>>>>>>  _PRT associated with each node, and applying normal PCI bridge
>>>>>>  swizzling when we don't find a _PRT.  I think this can be done without
>>>>>>  using PCI bus numbers at all.
>>>>>>
>>>>>> So here we try to remove _PRT caching completely.
>>>>>>
>>>>>> -v2: check !handle early.
>>>>>>
>>>>>> Reported-and-tested-by: Peter Hurley <peter@xxxxxxxxxxxxxxxxxx>
>>>>>> Suggested-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
>>>>>> Signed-off-by: Yinghai Lu <yinghai@xxxxxxxxxx>
>>>>>
>>>>> Acked-by: Rafael J. Wysocki <rafael.j.wysocki@xxxxxxxxx>
>>>>>
>>>>>> ---
>>>>>>  drivers/acpi/pci_irq.c      |   95 +++++++++++++++++---------------------------
>>>>>>  drivers/acpi/pci_root.c     |   18 --------
>>>>>>  drivers/pci/pci-acpi.c      |   24 -----------
>>>>>>  include/acpi/acpi_drivers.h |    5 --
>>>>>>  4 files changed, 38 insertions(+), 104 deletions(-)
>>>>
>>>> Bjorn,
>>>>
>>>> Can you put this one into pci/next?
>>>
>>> I'm not sure what this patch is based on or what the best way to merge
>>> it is.  It doesn't apply cleanly to my next or
>>> pci/yinghai-root-bus-hotplug branches.
>>
>> My fault, that is based on pci/next + pm/linux-next
>>
>> linux-next removed
>> acpi_power_resource_(un)register_device ...
>>
>>>
>>> I did apply it manually on top of pci/yinghai-root-bus-hotplug to try
>>> it out, but we need to tweak the messages a little bit.
>>>
>>> Previously we printed "ACPI: PCI Interrupt Routing Table [%s._PRT]"
>>> once when loading it, which was fine.  Now we print it every time we
>>> look at a _PRT, which is too much because it isn't really adding any
>>> information.
>>>
>>> We also print "ACPI Exception: AE_NOT_FOUND, Evaluating _PRT
>>> [AE_NOT_FOUND] (20121018/pci_irq-259)" if we find ACPI nodes without
>>> _PRTs, which we shouldn't do, because that's a common and normal
>>> situation.
>>
>> Sure. Can you have separated patch to do that ?
>>
>> Or want me to resend the patch.
>
> Please check attached updated version that remove print out ...
>
> and it could be applied cleanly on top of pci/yinghai-root-bus-hotplug

Thanks, I applied this to pci/yinghai-root-bus-hotplug and merged it
into my next branch.

Bjorn
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/