Re: Packet time delays on multi-core systems

[Jeff Kirsher]

On Fri, Oct 1, 2010 at 08:27, Eric Dumazet <eric.dumazet@xxxxxxxxx> wrote:
> Le vendredi 01 octobre 2010 Ã 18:18 +0400, Alexey Vlasov a Ãcrit :
>
>> NIC statistics:
>> Â Â Ârx_packets: 2973717440
>> Â Â Âtx_packets: 3032670910
>> Â Â Ârx_bytes: 1892633650741
>> Â Â Âtx_bytes: 2536130682695
>> Â Â Ârx_broadcast: 118773199
>> Â Â Âtx_broadcast: 68013
>> Â Â Ârx_multicast: 95257
>> Â Â Âtx_multicast: 0
>> Â Â Ârx_errors: 0
>> Â Â Âtx_errors: 0
>> Â Â Âtx_dropped: 0
>> Â Â Âmulticast: 95257
>> Â Â Âcollisions: 0
>> Â Â Ârx_length_errors: 0
>> Â Â Ârx_over_errors: 0
>> Â Â Ârx_crc_errors: 0
>> Â Â Ârx_frame_errors: 0
>> Â Â Ârx_no_buffer_count: 7939
>> Â Â Ârx_queue_drop_packet_count: 1324025520
>> Â Â Ârx_missed_errors: 146631
>> Â Â Âtx_aborted_errors: 0
>> Â Â Âtx_carrier_errors: 0
>> Â Â Âtx_fifo_errors: 0
>> Â Â Âtx_heartbeat_errors: 0
>> Â Â Âtx_window_errors: 0
>> Â Â Âtx_abort_late_coll: 0
>> Â Â Âtx_deferred_ok: 0
>> Â Â Âtx_single_coll_ok: 0
>> Â Â Âtx_multi_coll_ok: 0
>> Â Â Âtx_timeout_count: 0
>> Â Â Âtx_restart_queue: 50715
>> Â Â Ârx_long_length_errors: 0
>> Â Â Ârx_short_length_errors: 0
>> Â Â Ârx_align_errors: 0
>> Â Â Âtx_tcp_seg_good: 344724062
>> Â Â Âtx_tcp_seg_failed: 0
>> Â Â Ârx_flow_control_xon: 0
>> Â Â Ârx_flow_control_xoff: 0
>> Â Â Âtx_flow_control_xon: 0
>> Â Â Âtx_flow_control_xoff: 0
>> Â Â Ârx_long_byte_count: 1892633650741
>> Â Â Ârx_csum_offload_good: 2973697420
>> Â Â Ârx_csum_offload_errors: 6235
>> Â Â Âtx_dma_out_of_sync: 0
>> Â Â Âalloc_rx_buff_failed: 0
>> Â Â Âtx_smbus: 9327
>> Â Â Ârx_smbus: 118531661
>> Â Â Âdropped_smbus: 0
>> Â Â Âtx_queue_0_packets: 797617475
>> Â Â Âtx_queue_0_bytes: 630191908685
>> Â Â Âtx_queue_1_packets: 719681297
>> Â Â Âtx_queue_1_bytes: 625907304846
>> Â Â Âtx_queue_2_packets: 718841556
>> Â Â Âtx_queue_2_bytes: 620522418855
>> Â Â Âtx_queue_3_packets: 796521255
>> Â Â Âtx_queue_3_bytes: 646196024585
>> Â Â Ârx_queue_0_packets: 788885797
>> Â Â Ârx_queue_0_bytes: 458936338699
>> Â Â Ârx_queue_0_drops: 0
>> Â Â Ârx_queue_1_packets: 701354604
>> Â Â Ârx_queue_1_bytes: 457490536453
>> Â Â Ârx_queue_1_drops: 0
>> Â Â Ârx_queue_2_packets: 791887663
>> Â Â Ârx_queue_2_bytes: 534425333616
>> Â Â Ârx_queue_2_drops: 0
>> Â Â Ârx_queue_3_packets: 691579028
>> Â Â Ârx_queue_3_bytes: 429887244557
>> Â Â Ârx_queue_3_drops: 0
>> 11.111.80: R 1983626201:1983626201(0) win 0
>> > >
>
> OK
>
> IGB stats are wrong... for rx_queue_drop_packet_count field at least
>
> Here is a patch against 2.6.32.23, to get the idea...
>
> Dont trust it unless you patch your kernel ;)
>
> Thanks
>
> Note: current linux-2.6 tree doesnt have this bug.
>
> [PATCH] igb: rx_fifo_errors counter fix
>
> Alexey Vlasov reported insane rx_queue_drop_packet_count
> (rx_fifo_errors) values.
>
> IGB drivers is doing an accumulation for 82575, instead using a zero
> value for rqdpc_total.
>
> Reported-by: Alexey Vlasov <renton@xxxxxxxxxxx>
> Signed-off-by: Eric Dumazet <eric.dumazet@xxxxxxxxx>
> ---
> Âlinux-2.6.32.23/net/igb/igb_main.c | Â Â5 ++---
> Â1 file changed, 2 insertions(+), 3 deletions(-)
>

Thanks Eric!  I have added the patch to my queue.

-- 
Cheers,
Jeff
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/