[dpdk-dev] [PATCH] bus/pci: fix probing for non-netuio bound devices

[Tal Shnaiderman]

> Subject: [PATCH] bus/pci: fix probing for non-netuio bound devices
> 
> External email: Use caution opening links or attachments
> 
> 
> Implement rte_pci_map_device() to distinguish between the devices bound
> to netuio and NDIS devices.
> Only return success for the netuio devices.
> 
> Suggested-by: Dmitry Kozlyuk <dmitry.kozliuk at gmail.com>
> Signed-off-by: Pallavi Kadam <pallavi.kadam at intel.com>
> Reviewed-by: Ranjit Menon <ranjit.menon at intel.com>
> ---
>  drivers/bus/pci/windows/pci.c | 15 +++++++--------
>  1 file changed, 7 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/bus/pci/windows/pci.c b/drivers/bus/pci/windows/pci.c
> index f66258452..16edaa769 100644
> --- a/drivers/bus/pci/windows/pci.c
> +++ b/drivers/bus/pci/windows/pci.c
> @@ -23,20 +23,19 @@
> DEFINE_DEVPROPKEY(DEVPKEY_Device_Numa_Node, 0x540b947e, 0x8b40,
> 0x45bc,
>   * the registry hive for PCI devices.
>   */
> 
> -/* The functions below are not implemented on Windows,
> +/* Some of the functions below are not implemented on Windows,
>   * but need to be defined for compilation purposes
>   */
> 
>  /* Map pci device */
>  int
> -rte_pci_map_device(struct rte_pci_device *dev __rte_unused)
> +rte_pci_map_device(struct rte_pci_device *dev)
>  {
> -       /* This function is not implemented on Windows.
> -        * We really should short-circuit the call to these functions by
> -        * clearing the RTE_PCI_DRV_NEED_MAPPING flag
> -        * in the rte_pci_driver flags.
> -        */
> -       return 0;
> +       /* Only return success for devices bound to netuio */

Maybe extend this comment to mention where the mapping is actually done?

> +       if (dev->kdrv == RTE_PCI_KDRV_NIC_UIO)
> +               return 0;
> +       else
> +               return -EINVAL;	

The function should return -1 here to match the behavior with other OS implementations.

>  }
> 
>  /* Unmap pci device */
> --
> 2.18.0.windows.1