Re: [PATCH] pci/hotplug: work-around for missing _RMV on HP ZBook G2

[Jarod Wilson <jarod@redhat.com>]

On 5/18/2015 12:17 PM, Jarod Wilson wrote:
> On 5/18/2015 10:33 AM, Jarod Wilson wrote:
>> On 5/17/2015 8:26 PM, Rafael J. Wysocki wrote:
>>> On Saturday, May 16, 2015 09:41:55 AM Bjorn Helgaas wrote:
>>>> On Sat, May 16, 2015 at 09:37:50AM -0500, Bjorn Helgaas wrote:
>>>>> Hi Jarod,
>>>>>
>>>>> On Thu, May 14, 2015 at 03:33:58PM -0400, Jarod Wilson wrote:
>>>>>> The HP ZBook 15 and 17 Mobile Workstations, generation 2, up to and
>>>>>> including at least BIOS revision 01.07, do not have an ACPI _RMV
>>>>>> object
>>>>>> associated with their expresscard slots, so acpi-based
>>>>>> hotplug-capable
>>>>>> slot detection fails. If we fall back to pcie-based detection, the
>>>>>> systems
>>>>>> work just fine, so this uses dmi matching to do that. With luck, a
>>>>>> future
>>>>>> BIOS will remedy this (I've let someone at HP know about the
>>>>>> problem),
>>>>>> but for now, just use this for all existing versions.
>> ...
>>>>> Oh, my goodness.  I forgot how terrible this path is.  Can anyone
>>>>> write a
>>>>> simple explanation of how we choose to use acpiphp or pciehp?
>>>
>>> In theory, that should depend on the _OSC handshake in
>>> acpi_pci_root_add().
>>>
>>> If the firmware doesn't give us control of the PCIe features, we'll
>>> not use
>>> pciehp (or at least that's the idea).
>>>
>>> acpiphp is used if pciehp doesn't claim the device, AFAICS.
>>
>> [    4.013326] acpi PNP0A08:00: _OSC: OS supports [ExtendedConfig ASPM
>> ClockPM Segments MSI]
>> [    4.015860] acpi PNP0A08:00: _OSC: OS now controls [PCIeHotplug PME
>> AER PCIeCapability]
>>
>> So at a glance, it would appear that pciehp *should* be claiming it,
>> right? Something I noted in the bug I filed is that the device ID
>> reported there is PNP0A08, and the root_device_id table that associates
>> with acpi_pci_root_add() only includes PNP0A03 in it. Is that correct,
>> or should 08 also be in there, which might remedy this? (I can test this
>> out easily enough).
>
> Nope, makes no difference, seems those are just two different references
> to the same bus, based on a peek at the extracted dsdt:
>
> Name (_HID, EisaId ("PNP0A08") /* PCI Express Bus */)  // _HID: Hardware ID
> Name (_CID, EisaId ("PNP0A03") /* PCI Bus */)  // _CID: Compatible ID

Ah, I forgot some additional details. pciehp_probe() in 
drivers/pci/hotplug/pciehp_core.c fails on the 
pciehp_acpi_slot_detection_check() call for the expresscard slot, which 
is why the base pciehp doesn't bind. DEVICE_ACPI_HANDLE(&dev->dev) in 
the slot detection check is winding up with a NULL acpi device.

-- 
Jarod Wilson
jarod@redhat.com