DPDK: dpdk-19.11-rc2/ File: doc/guides/sample_app_ug/l2_forward_event.rst CMD: `./build/l2fwd-event -l 0-7 -s 0-3 -n 4 --vdev event_sw0 --q 8 -p ffff --mode=eventdev --eventq-sched=ordered` Logs: EAL: Support maximum 128 logical core(s) by configuration. EAL: Detected 88 lcore(s) EAL: Detected 2 NUMA nodes EAL: invalid service coremask Usage: ./build/l2fwd-event [options] As I recollect, there is a check which prevents service cores to be independent of data-plane cores. Patchwork: https://patches.dpdk.org/patch/43407/ Can you share the correct core mask to test?
as per EAL args ` -s SERVICE COREMASK Hexadecimal bitmask of cores to be used as service cores` so will the command be `./build/l2fwd-event -l 0-7 -s 0x3 -n 4 --vdev event_sw0 --q 8 -p ffff --mode=eventdev --eventq-sched=ordered`
result of converting to hexadecimal core mask is $ ./build/l2fwd-event -l 0-7 -s 0x3 -n 4 --vdev event_sw0 -- -q 8 -p ffff --mode=eventdev --eventq-sched=ordered EAL: Detected 88 lcore(s) EAL: Detected 2 NUMA nodes EAL: Multi-process socket /run/user/1000/dpdk/rte/mp_socket EAL: FATAL: Cannot use IOVA as 'PA' since physical addresses are not available EAL: Cannot use IOVA as 'PA' since physical addresses are not available PANIC in main(): Invalid EAL arguments 5: [./build/l2fwd-event(_start+0x2a) [0x56495cc6888a]] 4: [/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7) [0x7f039e3f6b97]] 3: [./build/l2fwd-event(main+0xddb) [0x56495cbbc96b]] 2: [./build/l2fwd-event(__rte_panic+0xc5) [0x56495cbaf7e3]] 1: [./build/l2fwd-event(rte_dump_stack+0x2e) [0x56495cd7e92e]] Aborted (core dumped) Can you please suggest the right parameters?
I got the application to invoke rte_eal_init, after reboot NICS: i40e NIC CMD: ./build/l2fwd-event -l 0-7 -s 0x3 -n 4 --vdev event_sw0 -- -q 8 -p ffff --mode=eventdev --eventq-sched=ordered Logs-1: Initializing port 0... Ethdev port_id=0 invalid rss_hf: 0xa38c, valid value: 0x7ef8 PANIC in l2fwd_event_init_ports(): Cannot configure device: err=-22, port=0 6: [./build/l2fwd-event(_start+0x2a) [0x5632290b388a]] 5: [/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7) [0x7fe2af0e6b97]] 4: [./build/l2fwd-event(main+0x633) [0x5632290071c3]] 3: [./build/l2fwd-event(l2fwd_event_init_ports+0x3fa) [0x5632290b5f7a]] 2: [./build/l2fwd-event(__rte_panic+0xc5) [0x563228ffa7e3]] 1: [./build/l2fwd-event(rte_dump_stack+0x2e) [0x5632291c992e]] Aborted After white list same error. CMD: ./build/l2fwd-event -l 0-7 -s 0x3 -n 4 --vdev event_sw0 -w 08:00.2 -w 83:00.2 -- -q 8 -p 3 --mode=eventdev --eventq-sched=ordered Logs-2: EAL: Detected 88 lcore(s) EAL: Detected 2 NUMA nodes EAL: Multi-process socket /var/run/dpdk/rte/mp_socket EAL: Selected IOVA mode 'PA' EAL: Probing VFIO support... EAL: PCI device 0000:08:00.2 on NUMA socket 0 EAL: probe driver: 8086:1572 net_i40e EAL: PCI device 0000:83:00.2 on NUMA socket 1 EAL: probe driver: 8086:1572 net_i40e MAC updating enabled Initializing port 0... Ethdev port_id=0 invalid rss_hf: 0xa38c, valid value: 0x7ef8 PANIC in l2fwd_event_init_ports(): Cannot configure device: err=-22, port=0 6: [./build/l2fwd-event(_start+0x2a) [0x55a197ac488a]] 5: [/lib/x86_64-linux-gnu/libc.so.6(__libc_start_main+0xe7) [0x7f0d2e4ebb97]] 4: [./build/l2fwd-event(main+0x633) [0x55a197a181c3]] 3: [./build/l2fwd-event(l2fwd_event_init_ports+0x3fa) [0x55a197ac6f7a]] 2: [./build/l2fwd-event(__rte_panic+0xc5) [0x55a197a0b7e3]] 1: [./build/l2fwd-event(rte_dump_stack+0x2e) [0x55a197bda92e]] Aborted
Should we check the default hash capability of the port, then assign?
Update to the ticket Running with `./build/l2fwd-event -l 0-7 -s 0x3 -n 4 --vdev event_sw0 -- -q 8 -p ffff --mode=poll --eventq-sched=ordered`, results this in ```MAC updating enabled Initializing port 0... Port 0 failed to disable Rx offload RSS_HASH Port 0,MAC address: 3C:FD:FE:9F:3A:02 Initializing port 1... Port 1 failed to disable Rx offload RSS_HASH Port 1,MAC address: 3C:FD:FE:A0:11:9A Lcore 0: RX port 0 Lcore 0: RX port 1 Checking link status...done Port0 Link Up. Speed 10000 Mbps - full-duplex Port1 Link Up. Speed 10000 Mbps - full-duplex lcore 2 has nothing to do lcore 3 has nothing to do lcore 4 has nothing to do lcore 5 has nothing to do lcore 6 has nothing to do lcore 7 has nothing to do Port statistics ==================================== Statistics for port 0 ------------------------------ Packets sent: 0 Packets received: 0 Packets dropped: 0 Statistics for port 1 ------------------------------ Packets sent: 0 Packets received: 0 Packets dropped: 0 Aggregate lcore statistics ========================= Total packets sent: 0 Total packets received: 0 Total packets dropped: 0 ==================================================== ```
Question on x86 Xeon with SW eventdev should the eventmode be `poll`?
Pavan, Can you take a look? Thanks