Re: [PATCH -next] pci: add function stub for pci_get_domain_bus_and_slot()

[Sinan Kaya]

On 12/5/2017 3:39 PM, Bjorn Helgaas wrote:
> [+cc Sinan]
> 
> On Tue, Nov 28, 2017 at 02:31:21PM -0800, Randy Dunlap wrote:
>> From: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
>>
>> The coretemp driver build fails when CONFIG_PCI is not enabled
>> because it uses a function that does not have a stub for that
>> config case, so add the function stub.
>>
>> ../drivers/hwmon/coretemp.c: In function 'adjust_tjmax':
>> ../drivers/hwmon/coretemp.c:250:9: error: implicit declaration of function 'pci_get_domain_bus_and_slot' [-Werror=implicit-function-declaration]
>>   struct pci_dev *host_bridge = pci_get_domain_bus_and_slot(0, 0, devfn);
>> ../drivers/hwmon/coretemp.c:250:32: warning: initialization makes pointer from integer without a cast [enabled by default]
>>   struct pci_dev *host_bridge = pci_get_domain_bus_and_slot(0, 0, devfn);
>>
>> Signed-off-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
> Applied with Guenter's ack to pci/enumeration for v4.16, thanks!
> 
> Apparently no other code calls pci_get_domain_bus_and_slot() when
> CONFIG_PCI isn't enabled.
> 
> 15ab6790f628 ("hwmon: (coretemp) deprecate pci_get_bus_and_slot()") is
> already in linux-next via hwmon-staging.
> 
> The pci tree is merged into next slightly before hwmon-staging, so
> this should be resolved as soon as the 0-day builder builds my
> pci/enumeration branch and I merge it into my next branch.
> 

Thanks for catching this. I assumed we do have a stub. It turns out we
don't.

-- 
Sinan Kaya
Qualcomm Datacenter Technologies, Inc. as an affiliate of Qualcomm Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project.