From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-pb0-x22b.google.com (mail-pb0-x22b.google.com [IPv6:2607:f8b0:400e:c01::22b]) by dpdk.org (Postfix) with ESMTP id D168F232 for ; Tue, 18 Jun 2013 06:18:01 +0200 (CEST) Received: by mail-pb0-f43.google.com with SMTP id md12so3507872pbc.16 for ; Mon, 17 Jun 2013 21:18:12 -0700 (PDT) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=date:from:to:reply-to:subject:x-priority:x-has-attach:x-mailer :mime-version:message-id:content-type; bh=+FKR8xVbP15PIPzfapepIzqef4taY1KUveiDaSSpFqI=; b=uG2gmdarNIYkhPEx39HV7505HGtH8iQWDbq/5xUFG6CIBzSRBxb6t60C7eTcyBcSDH WnCblN2ppk0a/Dwor2Yn/X7wH21I7C5aeD0LYbh3ERDWs50UVCGBV9WEt/vpwGGJ7b5y Ig/29qWKR++TOaA9c0EcbRdkXazblgfToXWrxCE683DH3pcjGeXZESPAYrO8PUKpQ5D+ fZGj45ViCZZvu5AYThEyOrsfQjmB4yCoAGlWO4Gj9z6k7woQkfVTm/bn2PQy/B/T0WG0 SGEQ6lsxy0v/YrOY51AZcSVmO63aa157fHDN5pqKrTamQuhYC5TUfe19p9CgT1DcBbli 6FHw== X-Received: by 10.68.132.199 with SMTP id ow7mr15275918pbb.161.1371529092359; Mon, 17 Jun 2013 21:18:12 -0700 (PDT) Received: from zengxg-PC ([106.120.127.15]) by mx.google.com with ESMTPSA id eq5sm16403895pbc.15.2013.06.17.21.18.09 for (version=TLSv1 cipher=RC4-SHA bits=128/128); Mon, 17 Jun 2013 21:18:11 -0700 (PDT) Date: Tue, 18 Jun 2013 12:18:07 +0800 From: zengxg14 To: dev X-Priority: 3 X-Has-Attach: no X-Mailer: Foxmail 7.0.1.92[cn] Mime-Version: 1.0 Message-ID: <2013061812175449782514@gmail.com> Content-Type: multipart/alternative; boundary="----=_001_NextPart705057213766_=----" Subject: [dpdk-dev] testpmd drops many packets X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.15 Precedence: list Reply-To: zengxg14 List-Id: patches and discussions about DPDK List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Tue, 18 Jun 2013 04:18:02 -0000 This is a multi-part message in MIME format. ------=_001_NextPart705057213766_=---- Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 SGkuDQoNCkkgdXNlIHRlc3RwbWQgdG8gdGVzdCB0aGUgdGhyb3VnaHB1dCwgYnV0IGl0IGRyb3Bz IG1hbnkgcGFja2V0cy4NCg0KSGVyZSBpcyBteSBoYXJkd2FyZToNCkludGVsIFhlb24gRTU2NDUs ICAgMiBzb2NrZXRzLCAgSFQgaXMgY2xvc2VkLCB0aGUgdG90YWwgY29yZXMgbnVtYmVyIGlzIDEy Lg0KNjRHQiBtZW1vcnkNCkludGVsIDgyNTk5RUIgU0ZQKywgIG9ubHkgb25lIHBvcnQuDQpSZWRo YXQgNi4zICgyLjYuMzItMjc5LmVsNi54ODZfNjQpDQppeGlhIHNlbmQgcGFja2V0cywgbGluZSBy YXRlIHdpdGggMjAwIGJ5dGVzIHBhY2tldC4gUGFja2V0cyBhcmUgaGF2ZSByYW5kb20gc291cmNl IGlwIGFuZCBkZXN0IGlwLg0KDQpUaGUgY29tbWFuZCBpIHJ1bjoNCiAuL3Rlc3RwbWQgLWMgZmZm IC1uIDMgLS0gLWkgIC0tY29yZW1hc2s9MWUgLS1wb3J0bWFzaz0xIC0tcnNzLWlwIC0tcnhxIDQg LS10eHEgNCAtLXJ4ZCAyMDQ4IC0tdHhkIDIwNDggLS1idXJzdCA1MTINClRoZSBjb3JlbWFzayBz ZXQgdG8gMWUsIHNvIGkgY2FuIHVzZSA0IGNvcmVzIHRvIHJlY2VpdmUgcGFja2V0cyBmcm9tIDQg cXVldWVzLg0KDQpTZXQgdGhlIGVuZ2luZSB0byByeF9vbmx5X2VuZ2luZS4NCiAgICAgICAgIHN0 cnVjdCBmd2RfZW5naW5lICpjdXJfZndkX2VuZyA9ICZyeF9vbmx5X2VuZ2luZTsNCg0KSSBhbHNv IGNoYW5nZWQgdGVzdHBtZC5jIGFuZCBpeGdiZV84MjU5OS5jIHRvIG1ha2UgaXQgcnVuLiAgDQpI ZXJlIGlzIHdoYXQgaSBjaGFuZ2VkOg0KKDEpIEluIG1haW4oKSwgSSBkZWxldGUgdGhlIGZ1bmN0 aW9uIHJ0ZV9pZ2JfcG1kX2luaXQoKSwgU1NIIG5lZWQgdG8gdXNlIGlnYiBuaWMuDQoNCigyKSBJ biBmdWNudGlvbiBpbml0X2NvbmZpZygpLCBpIHJlc2V0IHRoZSBuYl9tYnVmX3Blcl9wb29sLiBJ IGhhdmUgdG8sIG90aGVyd2lzZSB0aGUgdGVzdHBtZCB0ZWxscyBhbGxvYyByeCBtYnVmcyBmYWls ZWQuDQpuYl9tYnVmX3Blcl9wb29sID0gKG5iX3J4ZCArIChuYl9sY29yZXMgKiBtYl9tZW1wb29s X2NhY2hlKSArDQogICAgICAgICAgICAgICAgbmJfdHhkICsgTUFYX1BLVF9CVVJTVCkgKiBuYl9y eHE7DQoNCigzKSBJbiBsaWIvbGlicnRlX3BtZF9peGdiZS9peGdiZS9peGdiZV84MjU5OS5jLCBm dW5jdGlvbiAiaXhnYmVfc2V0dXBfc2ZwX21vZHVsZXNfODI1OTkiLg0KSSBjb21tZW50IHRoaXMg dGhyZWUgbGluZXMuDQoNCi8vICAgICAgICAgICAgICAgICAgICAgIERFQlVHT1VUKCJzZnAgbW9k dWxlIHNldHVwIG5vdCBjb21wbGV0ZVxuIik7DQovLyAgICAgICAgICAgICAgICAgICAgICByZXRf dmFsID0gSVhHQkVfRVJSX1NGUF9TRVRVUF9OT1RfQ09NUExFVEU7DQovLyAgICAgICAgICAgICAg ICAgICAgICBnb3RvIHNldHVwX3NmcF9vdXQ7DQoNCkkgZG9uJ3Qga25vdyB3aHkgdGhlICJzZnAg bW9kdWxlIHNldHVwIG5vdCBjb21wbGV0ZSIsIEkgc2ltcGx5IGNvbW1lbnQgdGhlbSB0byBtYWtl IGl4Z2JlIGNhbiBiZSAgaW5pdC4gDQoNClJ1biB0aGUgY29tbWFuZCwgdXNlIGl4aWEgc2VuZCBz b21lIHBhY2tldHMgYWZ0ZXIgc3RhcnQuIEhlcmUgaXMgdGhlIG91dHB1dC4NCg0KdGVzdHBtZD4g c3RhcnQNCiAgcnhvbmx5IHBhY2tldCBmb3J3YXJkaW5nIC0gQ1JDIHN0cmlwcGluZyBkaXNhYmxl ZCAtIHBhY2tldHMvYnVyc3Q9NTEyDQogIG5iIGZvcndhcmRpbmcgY29yZXM9NCAtIG5iIGZvcndh cmRpbmcgcG9ydHM9MQ0KICBSWCBxdWV1ZXM9NCAtIFJYIGRlc2M9MjA0OCAtIFJYIGZyZWUgdGhy ZXNob2xkPTANCiAgUlggdGhyZXNob2xkIHJlZ2lzdGVyczogcHRocmVzaD04IGh0aHJlc2g9OCB3 dGhyZXNoPTQNCiAgVFggcXVldWVzPTQgLSBUWCBkZXNjPTIwNDggLSBUWCBmcmVlIHRocmVzaG9s ZD0wDQogIFRYIHRocmVzaG9sZCByZWdpc3RlcnM6IHB0aHJlc2g9MzYgaHRocmVzaD0wIHd0aHJl c2g9MA0KICBUWCBSUyBiaXQgdGhyZXNob2xkPTANCnRlc3RwbWQ+IHNob3cgcG9ydCBzdGF0cyAw DQoNCiAgIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIE5JQyBzdGF0aXN0aWNzIGZvciBwb3J0IDAg ICMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIw0KICBSWC1wYWNrZXRzOiA3Mzc1NDM5OTIgIFJYLWVy cm9yczogMzkzOTg0NTc1IFJYLWJ5dGVzOiA2ODcxOTQ3NjczNQ0KICBUWC1wYWNrZXRzOiAwICAg ICAgICAgIFRYLWVycm9yczogMCAgICAgICAgIFRYLWJ5dGVzOiAwDQogICMjIyMjIyMjIyMjIyMj IyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMjIyMj IyMjIyMNCnRlc3RwbWQ+IHF1aXQNCg0KDQoNCg0KDQp6ZW5neGcxNA== ------=_001_NextPart705057213766_=---- Content-Type: text/html; charset="gb2312" Content-Transfer-Encoding: quoted-printable
Hi.
 
I use testpmd to test the throughput, but it drops many packets.
 
Here is my hardware:
Intel Xeon E5645,   2 sockets,  HT is closed, the tota= l=20 cores number is 12.
64GB memory
Intel 82599EB SFP+,  only one port.
Redhat 6.3 (2.6.32-279.el6.x86_64)
ixia send packets, line rate with 200 bytes packet. Packets are have = random=20 source ip and dest ip.
 
The command i run:
 ./testpmd -c fff -n 3 -- -i =20 --coremask=3D1e --portmask=3D1 --rss-ip --rxq 4 --txq 4 --rxd 2048 --txd 2= 048=20 --burst 512
The coremask set to 1e, so i can use 4 cores to receive packets from = 4=20 queues.
 
Set the engine to rx_only_engine.
         struct fwd_engine=20 *cur_fwd_eng =3D &rx_only_engine;
 
I also changed testpmd.c and ixgbe_82599.c to make it run.&= nbsp;=20
Here is what i changed:
(1) In main(), I delete the function rte_igb_pmd_init(), SSH need to = use=20 igb nic.
 
(2) In fucntion init_config(), i reset the nb_mbuf_per_pool. I have t= o,=20 otherwise the testpmd tells alloc rx mbufs failed.
nb_mbuf_per_pool =3D (nb_rxd + (nb_lcores *&= nbsp;mb_mempool_cache) +
           &nb= sp;    nb_txd + MAX_PKT_BURST) * n= b_rxq;
 
(3) In lib/librte_pmd_ixgbe/ixgbe/ixgbe_82599.c, function=20 "ixgbe_setup_sfp_modules_82599".
I comment this three lines.
 
//           &= nbsp;         =20 DEBUGOUT("sfp module setup not=20 complete\n");
//         &= nbsp;           =20 ret_val =3D=20 IXGBE_ERR_SFP_SETUP_NOT_COMPLETE;
//      = ;            &= nbsp;  =20 goto setup_sfp_out;
I don't know why the "sfp module setup not complete", I simply commen= t them=20 to make ixgbe can be  init.
 
Run the command, use ixia send some packets after start. Here is the=20 output.

testpmd> start
  rxonly packet forwarding - CRC stripp= ing=20 disabled - packets/burst=3D512
  nb forwarding cores=3D4 - nb forw= arding=20 ports=3D1
  RX queues=3D4 - RX desc=3D2048 - RX free threshold=3D0=
  RX=20 threshold registers: pthresh=3D8 hthresh=3D8 wthresh=3D4
  TX queu= es=3D4 - TX=20 desc=3D2048 - TX free threshold=3D0
  TX threshold registers: pthr= esh=3D36=20 hthresh=3D0 wthresh=3D0
  TX RS bit threshold=3D0
testpmd> s= how port=20 stats 0
 
  ######################## NIC statistics for port 0 =20 ########################
  RX-packets: 737543992  RX-errors:=20 393984575 RX-bytes: 68719476735
  TX-packets:=20 0          TX-errors:=20 0         TX-bytes: 0
 =20 ##########################################################################= ##
testpmd>=20 quit
 
zengxg14
------=_001_NextPart705057213766_=------