Created attachment 275 [details] l2fwd patch to demonstrate issue When a PCAP PMD Ethernet device is first configured, the selected number of RX and TX queues overrides the max values initially returned by rte_eth_dev_info_get() (max_rx_queues, max_tx_queues). Now, if one tries to reconfigure the device with a larger queue count(s), rte_eth_dev_configure() will fail to an internal capability check. Reconfiguring devices this way is useful for example in CI usage. The issue can be reproduced with the attached simple patch to l2fwd example. Example output from patched l2fwd (v22.11.4): $ sudo ./examples/dpdk-l2fwd --no-pci --vdev net_pcap0,iface=lo -- -p 0x1 EAL: Detected CPU lcores: 48 EAL: Detected NUMA nodes: 1 EAL: Detected static linkage of DPDK EAL: Multi-process socket /var/run/dpdk/rte/mp_socket EAL: Selected IOVA mode 'VA' EAL: VFIO support initialized TELEMETRY: No legacy callbacks, legacy socket not created MAC updating enabled Notice: odd number of ports in portmask. Lcore 0: RX port 0 TX port 0 Initializing port 0... orig. max_rx_queues: 1 orig. max_tx_queues: 1 after conf max_rx_queues: 0 after conf max_tx_queues: 1 Ethdev port_id=0 nb_rx_queues=1 > 0 EAL: Error - exiting with code: 1 Cause: Cannot configure device: err=-22, port=0