On 5/15/23 16:19, Manivannan Sadhasivami wrote:
> On Mon, May 15, 2023 at 02:14:59PM +0900, Damien Le Moal wrote:
>> pci_epf_type_add_cfs() should not be called with an unbound EPF device,
>> that is, an epf device with epf->driver not set. For such case, replace
>> the NULL return in pci_epf_type_add_cfs() with a clear ERR_PTR(-EINVAL)
>> error return.
>>
>
> Shouldn't the error code be -ENODEV?
Yeah, I wondered about it and settled for EINVAL... Will change to ENODEV.
>
> - Mani
>
>> Signed-off-by: Damien Le Moal <dlemoal@xxxxxxxxxx>
>> ---
>> drivers/pci/endpoint/pci-ep-cfs.c | 2 +-
>> 1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/drivers/pci/endpoint/pci-ep-cfs.c b/drivers/pci/endpoint/pci-ep-cfs.c
>> index 0fb6c376166f..d8a6abe2c31c 100644
>> --- a/drivers/pci/endpoint/pci-ep-cfs.c
>> +++ b/drivers/pci/endpoint/pci-ep-cfs.c
>> @@ -516,7 +516,7 @@ static struct config_group *pci_epf_type_add_cfs(struct pci_epf *epf,
>>
>> if (!epf->driver) {
>> dev_err(&epf->dev, "epf device not bound to driver\n");
>> - return NULL;
>> + return ERR_PTR(-EINVAL);
>> }
>>
>> if (!epf->driver->ops->add_cfs)
>> --
>> 2.40.1
>>
>
--
Damien Le Moal
Western Digital Research
