Re: [Patch v4 01/17] net/mana: add basic driver, build environment and doc

[Ferruh Yigit <ferruh.yigit@xilinx.com>]

On 8/22/2022 4:03 PM, Ferruh Yigit wrote:
>> +                             struct rte_pci_device *pci_dev,
>> +                             struct rte_ether_addr *mac_addr)
>> +{
>> +       struct ibv_device **ibv_list;
>> +       int ibv_idx;
>> +       struct ibv_context *ctx;
>> +       struct ibv_device_attr_ex dev_attr;
>> +       int num_devices;
>> +       int ret = 0;
>> +       uint8_t port;
>> +       struct mana_priv *priv = NULL;
>> +       struct rte_eth_dev *eth_dev = NULL;
>> +       bool found_port;
>> +
>> +       ibv_list = ibv_get_device_list(&num_devices);
>> +       for (ibv_idx = 0; ibv_idx < num_devices; ibv_idx++) {
>> +               struct ibv_device *ibdev = ibv_list[ibv_idx];
>> +               struct rte_pci_addr pci_addr;
>> +
>> +               DRV_LOG(INFO, "Probe device name %s dev_name %s 
>> ibdev_path %s",
>> +                       ibdev->name, ibdev->dev_name, ibdev->ibdev_path);
>> +
>> +               if (mana_ibv_device_to_pci_addr(ibdev, &pci_addr))
>> +                       continue;
>> +
>> +               /* Ignore if this IB device is not this PCI device */
>> +               if (pci_dev->addr.domain != pci_addr.domain ||
>> +                   pci_dev->addr.bus != pci_addr.bus ||
>> +                   pci_dev->addr.devid != pci_addr.devid ||
>> +                   pci_dev->addr.function != pci_addr.function)
>> +                       continue;
>> +
> 
> As far as I understand, intention of this loop is to find 'ibdev' 
> matching this device, code gooes through all "ibv device list" for this, 
> I wonder if there is a easy way for doing this, like a sysfs entry to 
> help getting this information?
> And how mlx4/5 does this?

Since there are multiple RDMA devices now, does it make sense to have 
RDMA bus driver, which can hide some PCIe details under bus, and driver 
can get PCI and ibdev information during probe?