David S. Miller wrote:
From: "Michael Chan" <[EMAIL PROTECTED]>
Date: Tue, 31 Jan 2006 15:30:05 -0800
Agreed in general. That's why the default is still 100 with one ring of
255 descriptors. Most users will not change this default setting.
How about we compromise and cap the limit at 512 (or 511) max
RX descriptors?
It may not be completely germane, but the cap on rx descriptors was ringing a
bell. I'm doing aggregate netperf TCP_RR stuff on some systems (e1000 and tg3)
and I believe I may be hitting the limit in that when I have reasonably high
burst counts (TOT netperf2 bits "first burst" functionality) - this is with tg3
at least. I think I'm also losing packets elsewhere.
Like I said, this limit is coming anyways.
via channels?
Some stats
lnx20:~/netperf-2.4.2pre1 # ethtool -S eth2
NIC statistics:
rx_octets: 58323851315
rx_fragments: 0
rx_ucast_packets: 821374025
rx_mcast_packets: 1
rx_bcast_packets: 108
rx_fcs_errors: 0
rx_align_errors: 0
rx_xon_pause_rcvd: 0
rx_xoff_pause_rcvd: 0
rx_mac_ctrl_rcvd: 0
rx_xoff_entered: 0
rx_frame_too_long_errors: 0
rx_jabbers: 0
rx_undersize_packets: 0
rx_in_length_errors: 0
rx_out_length_errors: 0
rx_64_or_less_octet_packets: 86
rx_65_to_127_octet_packets: 821370383
rx_128_to_255_octet_packets: 630
rx_256_to_511_octet_packets: 3035
rx_512_to_1023_octet_packets: 0
rx_1024_to_1522_octet_packets: 0
rx_1523_to_2047_octet_packets: 0
rx_2048_to_4095_octet_packets: 0
rx_4096_to_8191_octet_packets: 0
rx_8192_to_9022_octet_packets: 0
tx_octets: 79462277775
tx_collisions: 0
tx_xon_sent: 0
tx_xoff_sent: 0
tx_flow_control: 0
tx_mac_errors: 0
tx_single_collisions: 0
tx_mult_collisions: 0
tx_deferred: 0
tx_excessive_collisions: 0
tx_late_collisions: 0
tx_collide_2times: 0
tx_collide_3times: 0
tx_collide_4times: 0
tx_collide_5times: 0
tx_collide_6times: 0
tx_collide_7times: 0
tx_collide_8times: 0
tx_collide_9times: 0
tx_collide_10times: 0
tx_collide_11times: 0
tx_collide_12times: 0
tx_collide_13times: 0
tx_collide_14times: 0
tx_collide_15times: 0
tx_ucast_packets: 764939889
tx_mcast_packets: 0
tx_bcast_packets: 16
tx_carrier_sense_errors: 0
tx_discards: 0
tx_errors: 0
dma_writeq_full: 140800
dma_write_prioq_full: 0
rxbds_empty: 0
rx_discards: 0
rx_errors: 0
rx_threshold_hit: 361032793
dma_readq_full: 2239136
dma_read_prioq_full: 220760
tx_comp_queue_full: 0
ring_set_send_prod_index: 764900921
ring_status_update: 367962331
nic_irqs: 241853275
nic_avoided_irqs: 126109056
nic_tx_threshold_hit: 56357689
lnx20:~/netperf-2.4.2pre1 # ethtool -g eth2
Ring parameters for eth2:
Pre-set maximums:
RX: 511
RX Mini: 0
RX Jumbo: 255
TX: 0
Current hardware settings:
RX: 511
RX Mini: 0
RX Jumbo: 100
TX: 511
lnx20:~/netperf-2.4.2pre1 # ethtool -i eth2
driver: tg3
version: 3.37
firmware-version:
bus-info: 0000:05:01.0
BTW, is that Pre-set maximum for TX of 0 correct? I ass-u-me-d that meant there
was not a limit, but I could not take it beyond 511
rick jones
onlist, no need to cc
-
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
