Debian部署dpdk-1.8.0（VMWARE Fusion 7）
Debian部署dpdk-1.8.0（VMWARE Fusion 7）
环境
Debian版本： 7.4
内核：3.2.0
gcc：4.7.2
后来在安装的过程中发现没有sudo，所以如果没有的话最好先装一个：apt-get install sudo
给虚拟机分配3块网卡，注意别用NAT模式，否则测试程序无法收发包，统计结果都是0。
Debian环境下启用一块网卡即可，有实际地址，用来ssh登录进行操作，其他两块给dpdk折腾
编译&加载
先设置好环境变量：
export RTE_SDK=`pwd`
export RTE_TARGET=x86_64-native-linuxapp-gcc
进入源码目录执行：
root@debian:~/code/dpdk-1.8.0# ./tools/setup.sh
------------------------------------------------------------------------------
RTE_SDK exported as /root/code/dpdk-1.8.0
------------------------------------------------------------------------------
----------------------------------------------------------
Step 1: Select the DPDK environment to build
----------------------------------------------------------
[1] i686-native-linuxapp-gcc
[2] i686-native-linuxapp-icc
[3] ppc_64-power8-linuxapp-gcc
[4] x86_64-ivshmem-linuxapp-gcc
[5] x86_64-ivshmem-linuxapp-icc
[6] x86_64-native-bsdapp-clang
[7] x86_64-native-bsdapp-gcc
[8] x86_64-native-linuxapp-clang
[9] x86_64-native-linuxapp-gcc
[10] x86_64-native-linuxapp-icc
----------------------------------------------------------
Step 2: Setup linuxapp environment
----------------------------------------------------------
[11] Insert IGB UIO module
[12] Insert VFIO module
[13] Insert KNI module
[14] Setup hugepage mappings for non-NUMA systems
[15] Setup hugepage mappings for NUMA systems
[16] Display current Ethernet device settings
[17] Bind Ethernet device to IGB UIO module
[18] Bind Ethernet device to VFIO module
[19] Setup VFIO permissions
----------------------------------------------------------
Step 3: Run test application for linuxapp environment
----------------------------------------------------------
[20] Run test application ($RTE_TARGET/app/test)
[21] Run testpmd application in interactive mode ($RTE_TARGET/app/testpmd)
----------------------------------------------------------
Step 4: Other tools
----------------------------------------------------------
[22] List hugepage info from /proc/meminfo
----------------------------------------------------------
Step 5: Uninstall and system cleanup
----------------------------------------------------------
[23] Uninstall all targets
[24] Unbind NICs from IGB UIO driver
[25] Remove IGB UIO module
[26] Remove VFIO module
[27] Remove KNI module
[28] Remove hugepage mappings
[29] Exit Script
Option:
我是64位系统，选择[9]开始编译，一开头就碰到这么个错误：
/lib/module/`uname -r`/build: no such file or directory
即使手动创建对应的目录，同样会报错：No targets specified and no makefile found.这是因为正常情况build不是个目录，而是个软链接，指向/usr/src下对应的kernel头文件目录。因此手动创建个build的软链接即可，即/usr/src/linux-headers-`uname -r`/
如果没有安装kernel header，建议根据自己的内核版本下载：apt-get install linux-headers-`uname -r`
然后就是加载内核模块、分配大页内存、绑定网卡之类了
这里选[11],[14],[17], 大页内存可设置为128，网卡的话，需要填写网卡的PCIE的地址，如0000:02:05.0之类，操作过程的指导很清楚，可以照着提示信息选网卡。
测试程序
选择21，启动测试程序，我只有两个核，在选择 bitmask of cores时，输入了3
但是start后会不断打印错误日志：
EAL: Error reading from file descriptor
貌似是由于VMWare对PCIE的INTX中断模拟得比较差导致，
修改源码：
diff --git a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
index d1ca26e..c46a00f 100644
--- a/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
+++ b/lib/librte_eal/linuxapp/igb_uio/igb_uio.c
@@ -505,14 +505,11 @@ igbuio_pci_probe(struct pci_dev *dev, const struct pci_device_id *id)
}
/* fall back to INTX */
case RTE_INTR_MODE_LEGACY:
- if (pci_intx_mask_supported(dev)) {
- dev_dbg(&dev->dev, "using INTX");
- udev->info.irq_flags = IRQF_SHARED;
- udev->info.irq = dev->irq;
- udev->mode = RTE_INTR_MODE_LEGACY;
- break;
- }
- dev_notice(&dev->dev, "PCI INTX mask not supported\n");
+ dev_dbg(&dev->dev, "using INTX");
+ udev->info.irq_flags = IRQF_SHARED;
+ udev->info.irq = dev->irq;
+ udev->mode = RTE_INTR_MODE_LEGACY;
+ break;
/* fall back to no IRQ */
case RTE_INTR_MODE_NONE:
udev->mode = RTE_INTR_MODE_NONE;
不光是这块，由于修改后pci_intx_mask_supported()函数没有用到，编译还会报错（dpdk认为warning也是出错），得把头文件compat.h里这个函数的定义也去掉...
重新编译后一切ok：
testpmd> start
io packet forwarding - CRC stripping disabled - packets/burst=32
nb forwarding cores=1 - nb forwarding ports=2
RX queues=1 - RX desc=128 - RX free threshold=32
RX threshold registers: pthresh=8 hthresh=8 wthresh=0
TX queues=1 - TX desc=512 - TX free threshold=0
TX threshold registers: pthresh=32 hthresh=0 wthresh=0
TX RS bit threshold=0 - TXQ flags=0x0
testpmd> stop
Telling cores to stop...
Waiting for lcores to finish...
---------------------- Forward statistics for port 0 ----------------------
RX-packets: 829923 RX-dropped: 0 RX-total: 829923
TX-packets: 829856 TX-dropped: 0 TX-total: 829856
----------------------------------------------------------------------------
---------------------- Forward statistics for port 1 ----------------------
RX-packets: 829915 RX-dropped: 0 RX-total: 829915
TX-packets: 829856 TX-dropped: 0 TX-total: 829856
----------------------------------------------------------------------------
+++++++++++++++ Accumulated forward statistics for all ports+++++++++++++++
RX-packets: 1659838 RX-dropped: 0 RX-total: 1659838
TX-packets: 1659712 TX-dropped: 0 TX-total: 1659712
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Done.
不过我这里一旦start，CPU都占满了，虚拟机系统会变得很慢，敲stop都要等好一会
通过脚本加载
交互式加载不好自动化，可以写个脚本加载
首先编译安装dpdk：
make install T=x86_64-native-linuxapp-gcc
接下来的命令可以写到脚本里，PCI地址需要根据自己的情况设置：
echo 128 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages
mount -t hugetlbfs nodev /mnt/huge
modprobe uio
insmod x86_64-native-linuxapp-gcc/kmod/igb_uio.ko
./tools/dpdk_nic_bind.py -b igb_uio 0000:02:05.0
./tools/dpdk_nic_bind.py -b igb_uio 0000:02:06.0
执行测试程序：
./x86_64-native-linuxapp-gcc/app/testpmd -c 0x3 -n 2 -- -i
