[dpdk-dev] Could DPDK be used for Openstack?
Dear all, I'm a new who is getting started DPDK, and I wonder that the current supported version of DPDK vSwitch could replace Open vSwitch (ovs) in OpenStack. Thanks in advance. Best regards, Byeong-Gi KIM
[dpdk-dev] [Dpdk-ovs] Questions about the OpenStack Neutron with Intel Architecture document
Thank you for the reply. I installed DPDK 1.7.0 on RedHat 6.5 with Kernel 3.7.10 and it seems compiled successfully. When execute the HelloWorld example, however, there was a panic error as below: 1024 hugepages of size 2097152 reserved, but no mounted hugetlbfs found for that size PANIC in rte_eal_init(): Cannot get hugepage information The PC has Intel ATOM Processor and I binded a NIC to igb_uio module. I think the hugepage reservation is also correctly done since I put "echo 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages" What should I check to solve this problem? Thanks in advance. Best regards Byeong-Gi KIM 2014-08-05 18:19 GMT+09:00 Czesnowicz, Przemyslaw < przemyslaw.czesnowicz at intel.com>: > Hi BYEONG-GI KIM > > 1. This is only needed if you are working from a machine behind a > proxy. > > 2. By Accelerated OVS we mean ovs from openvswitch.org compiled and > running with dpdk support. > > You don?t have to deploy the ?Accelerated OVS? compute node. > > 3. We didn?t try running it on RH 6.5, let us know if you have any > issues. > > > > Regards > > Przemek > > > > *From:* Dpdk-ovs [mailto:dpdk-ovs-bounces at lists.01.org] *On Behalf Of > *BYEONG-GI > KIM > *Sent:* Monday, August 04, 2014 8:00 AM > *To:* Dpdk-ovs at lists.01.org > *Subject:* [Dpdk-ovs] Questions about the OpenStack Neutron with Intel > Architecture document > > > > Hello. > > > > I've noticed that the new version of both DPDK and corrresponding OVDK > finally has been released and the document for OpenStack Neutron with those > technologies also has been updated. > > > > The latest guide seems changed from the past guide. Here are several > questions about those changes. > > > > 1. The guide includes "Configuring Proxies", but I'm not sure this is > essential part. If it is essential, which and should be > set? And the is also confused. Does it mean I need > to deploy another proxy PC in addition to both Compute Node and Controller > Node for OpenStack? > > > > 2. Another Sample Compute local.conf is added at 6.1.1.3, but I wonder the > difference between past one and the new one. It is described as OpenStack > Accelerated OVS Compute Node. Does it mean the new one accelerates packet > processing but the past one does not? > > > > 3. Finally, I assume that the deployment would be fit at kernel 3.1.4 > version of Fedora (actually we are trying to install it on RedHat 6.5) > since the Fedora 20 is recommended. Is it right? > > > > Best regards > > > > Byeong-Gi KIM > > > > -- > Intel Shannon Limited > Registered in Ireland > Registered Office: Collinstown Industrial Park, Leixlip, County Kildare > Registered Number: 308263 > Business address: Dromore House, East Park, Shannon, Co. Clare > > This e-mail and any attachments may contain confidential material for the > sole use of the intended recipient(s). Any review or distribution by others > is strictly prohibited. If you are not the intended recipient, please > contact the sender and delete all copies. > >
[dpdk-dev] [Dpdk-ovs] Questions about the OpenStack Neutron with Intel Architecture document
The command shows "nodev on /mnt/huge type hugetlbfs (rw)", so I think it has been mounted correctly. By the way, I could execute HelloWorld example file by setting all the things through ./tools/setup.sh I'm still not sure which part was not set correctly though. Thanks anyway. Best regards Byeong-Gi KIM 2014-08-05 20:48 GMT+09:00 Czesnowicz, Przemyslaw < przemyslaw.czesnowicz at intel.com>: > Hi, > > > > From the error msg looks like You didn?t mount the hugetlbfs. > > To confirm do: > > mount | grep huge > > if there is no output mount hugetlbfs with: > > sudo mount -t hugetlbfs nodev /mnt/huge > > > > Regards > > Przemek > > *From:* BYEONG-GI KIM [mailto:kimbyeonggi at gmail.com] > *Sent:* Tuesday, August 05, 2014 12:28 PM > *To:* Czesnowicz, Przemyslaw; dev at dpdk.org > *Subject:* Re: [Dpdk-ovs] Questions about the OpenStack Neutron with > Intel Architecture document > > > > Thank you for the reply. > > > > I installed DPDK 1.7.0 on RedHat 6.5 with Kernel 3.7.10 and it seems > compiled successfully. > > > > When execute the HelloWorld example, however, there was a panic error as > below: > > 1024 hugepages of size 2097152 reserved, but no mounted hugetlbfs found > for that size > > PANIC in rte_eal_init(): > > Cannot get hugepage information > > > > The PC has Intel ATOM Processor and I binded a NIC to igb_uio module. I > think the hugepage reservation is also correctly done since I put "echo > 1024 > /sys/kernel/mm/hugepages/hugepages-2048kB/nr_hugepages" > > > > What should I check to solve this problem? > > > > Thanks in advance. > > > > Best regards > > > > Byeong-Gi KIM > > > > > > 2014-08-05 18:19 GMT+09:00 Czesnowicz, Przemyslaw < > przemyslaw.czesnowicz at intel.com>: > > Hi BYEONG-GI KIM > > 1. This is only needed if you are working from a machine behind a > proxy. > > 2. By Accelerated OVS we mean ovs from openvswitch.org compiled and > running with dpdk support. > > You don?t have to deploy the ?Accelerated OVS? compute node. > > 3. We didn?t try running it on RH 6.5, let us know if you have any > issues. > > > > Regards > > Przemek > > > > *From:* Dpdk-ovs [mailto:dpdk-ovs-bounces at lists.01.org] *On Behalf Of > *BYEONG-GI > KIM > *Sent:* Monday, August 04, 2014 8:00 AM > *To:* Dpdk-ovs at lists.01.org > *Subject:* [Dpdk-ovs] Questions about the OpenStack Neutron with Intel > Architecture document > > > > Hello. > > > > I've noticed that the new version of both DPDK and corrresponding OVDK > finally has been released and the document for OpenStack Neutron with those > technologies also has been updated. > > > > The latest guide seems changed from the past guide. Here are several > questions about those changes. > > > > 1. The guide includes "Configuring Proxies", but I'm not sure this is > essential part. If it is essential, which and should be > set? And the is also confused. Does it mean I need > to deploy another proxy PC in addition to both Compute Node and Controller > Node for OpenStack? > > > > 2. Another Sample Compute local.conf is added at 6.1.1.3, but I wonder the > difference between past one and the new one. It is described as OpenStack > Accelerated OVS Compute Node. Does it mean the new one accelerates packet > processing but the past one does not? > > > > 3. Finally, I assume that the deployment would be fit at kernel 3.1.4 > version of Fedora (actually we are trying to install it on RedHat 6.5) > since the Fedora 20 is recommended. Is it right? > > > > Best regards > > > > Byeong-Gi KIM > > > > -- > Intel Shannon Limited > Registered in Ireland > Registered Office: Collinstown Industrial Park, Leixlip, County Kildare > Registered Number: 308263 > Business address: Dromore House, East Park, Shannon, Co. Clare > > This e-mail and any attachments may contain confidential material for the > sole use of the intended recipient(s). Any review or distribution by others > is strictly prohibited. If you are not the intended recipient, please > contact the sender and delete all copies. > > > > -- > Intel Shannon Limited > Registered in Ireland > Registered Office: Collinstown Industrial Park, Leixlip, County Kildare > Registered Number: 308263 > Business address: Dromore House, East Park, Shannon, Co. Clare > > This e-mail and any attachments may contain confidential material for the > sole use of the intended recipient(s). Any review or distribution by others > is strictly prohibited. If you are not the intended recipient, please > contact the sender and delete all copies. > >
[dpdk-dev] a NIC binding to igb_uio module of DPDK is down
Hello. I deployed openstack with dpdk-ovs and I'm now testing it. The problem is that a NIC which is binding to igb_uio module of DPDK is physically down so that it seems not communicate with the other hosts through the NIC. The test environment is as below: OS: RHEL 6 with kernel 3.7.10 Hardware information: Intel Atom Processor C2758, 64GB DDR3, 4x RJ45 Gigabit Ethernet LAN ports, etc. DPDK version: 1.7.0 ovdk version: 1.1 I checked the NIC's status via ./tools/dpdk_nic_bind.py --status, and it doesn't snow *Active* status: ./tools/dpdk_nic_bind.py --status Network devices using DPDK-compatible driver :00:14.1 'Ethernet Connection I354' drv=igb_uio unused=igb Network devices using kernel driver === :00:14.0 'Ethernet Connection I354' if=eth0 drv=igb unused=igb_uio *Active* :00:14.2 'Ethernet Connection I354' if=eth2 drv=igb unused=igb_uio *Active* :00:14.3 'Ethernet Connection I354' if=eth3 drv=igb unused=igb_uio Other network devices = The NIC port on the PC doesn't glitter physically, which means it seems not connected physically. Doesn't the processor support DPDK, or what is causing the problem? Any comments would be really appreciated. Best Regards Byeong-Gi KIM
[dpdk-dev] DPDK supported processor
Hello. I'm now using Intel Atom processor C2758 on the DPDK testing machines, but it seems not working. Which sample application would be recommended to test the NIC actually works well with DPDK? Best regards Byeong-Gi KIM
[dpdk-dev] DPDK supported processor
Thank you for the reply. I tested l3forwarding sample application, and the results were as below: sudo ./build/l3fwd -c 0x0F -n 4 -- -p 0x03 --config="(0,0,0),(0,1,1),(1,0,2),(1,1,3)" EAL: Detected lcore 0 as core 0 on socket 0 EAL: Detected lcore 1 as core 1 on socket 0 EAL: Detected lcore 2 as core 2 on socket 0 EAL: Detected lcore 3 as core 3 on socket 0 EAL: Detected lcore 4 as core 4 on socket 0 EAL: Detected lcore 5 as core 5 on socket 0 EAL: Detected lcore 6 as core 6 on socket 0 EAL: Detected lcore 7 as core 7 on socket 0 EAL: Support maximum 64 logical core(s) by configuration. EAL: Detected 8 lcore(s) EAL: Searching for IVSHMEM devices... EAL: No IVSHMEM configuration found! EAL: Setting up memory... EAL: Ask a virtual area of 0x5ffc0 bytes EAL: Virtual area found at 0x2aa4aae0 (size = 0x5ffc0) EAL: Ask a virtual area of 0x20 bytes EAL: Virtual area found at 0x73e0 (size = 0x20) EAL: Ask a virtual area of 0x20 bytes EAL: Virtual area found at 0x73a0 (size = 0x20) EAL: Requesting 12288 pages of size 2MB from socket 0 EAL: TSC frequency is ~238 KHz EAL: Master core 0 is ready (tid=f7fe9800) EAL: Core 3 is ready (tid=f5998700) EAL: Core 2 is ready (tid=f6199700) EAL: Core 1 is ready (tid=f699a700) EAL: PCI device :00:14.0 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: PCI memory mapped at 0x77f8d000 EAL: PCI memory mapped at 0x77f89000 EAL: PCI device :00:14.1 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: PCI memory mapped at 0x77f69000 EAL: PCI memory mapped at 0x77f65000 EAL: PCI device :00:14.2 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: :00:14.2 not managed by UIO driver, skipping EAL: PCI device :00:14.3 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: :00:14.3 not managed by UIO driver, skipping EAL: PCI device :00:14.2 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: :00:14.2 not managed by UIO driver, skipping EAL: PCI device :00:14.3 on NUMA socket -1 EAL: probe driver: 8086:1f41 rte_igb_pmd EAL: :00:14.3 not managed by UIO driver, skipping Initializing port 0 ... Creating queues: nb_rxq=2 nb_txq=4... Address:0C:C4:7A:05:52:7A, Allocated mbuf pool on socket 0 LPM: Adding route 0x01010100 / 24 (0) LPM: Adding route 0x02010100 / 24 (1) LPM: Adding route IPV6 / 48 (0) LPM: Adding route IPV6 / 48 (1) txq=0,0,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. Initializing port 1 ... Creating queues: nb_rxq=2 nb_txq=4... Address:0C:C4:7A:05:52:7B, txq=0,0,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX WTHRESH value to 4, 8, or 16. Initializing rx queues on lcore 0 ... rxq=0,0,0 Initializing rx queues on lcore 1 ... rxq=0,1,0 Initializing rx queues on lcore 2 ... rxq=1,0,0 Initializing rx queues on lcore 3 ... rxq=1,1,0 Checking link status.done Port 0 Link Up - speed 100 Mbps - full-duplex Port 1 Link Up - speed 100 Mbps - full-duplex L3FWD: entering main loop on lcore 1 L3FWD: entering main loop on lcore 3 L3FWD: -- lcoreid=1 portid=0 rxqueueid=1 L3FWD: -- lcoreid=3 portid=1 rxqueueid=1 L3FWD: entering main loop on lcore 0 L3FWD: -- lcoreid=0 portid=0 rxqueueid=0 L3FWD: entering main loop on lcore 2 L3FWD: -- lcoreid=2 portid=1 rxqueueid=0 Anyway, I'll also try to test the testpmd application. Best regards Byeong-Gi KIM 2014-08-21 17:56 GMT+09:00 De Lara Guarch, Pablo < pablo.de.lara.guarch at intel.com>: > > -Original Message- > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of BYEONG-GI KIM > > Sent: Thursday, August 21, 2014 7:04 AM > > To: dev at dpdk.org > > Subject: [dpdk-dev] DPDK supported processor > > > > Hello. > > > > I'm now using Intel Atom processor C2758 on the DPDK testing machines, > but > > it seems not working. > > > > Which sample application would be recommended to test the NIC actually > > works well with DPDK? > > > > Best regards > > > > Byeong-Gi KIM > > Hi, > > Testpmd is the best application you can use for that. > > Best regards, > Pablo >
[dpdk-dev] DPDK supported processor
Hello. Hmm, the problem is a little complicated. Actually, I'm trying to use dpdk-ovs with openstack deployment, but it seems not working correctly. I've thus been trying to keep track of the primary problem of this. I've thought my machine's hardware requirements may not be satisfied to support DPDK, but I think it is solved now. Best regards Byeong-Gi KIM 2014-08-21 18:13 GMT+09:00 Masaru Oki : > Hi, > What is the problem in l3fwd? link speed? > > > 2014-08-21 18:00 GMT+09:00 BYEONG-GI KIM : > > Thank you for the reply. >> >> I tested l3forwarding sample application, and the results were as below: >> >> sudo ./build/l3fwd -c 0x0F -n 4 -- -p 0x03 >> --config="(0,0,0),(0,1,1),(1,0,2),(1,1,3)" >> EAL: Detected lcore 0 as core 0 on socket 0 >> EAL: Detected lcore 1 as core 1 on socket 0 >> EAL: Detected lcore 2 as core 2 on socket 0 >> EAL: Detected lcore 3 as core 3 on socket 0 >> EAL: Detected lcore 4 as core 4 on socket 0 >> EAL: Detected lcore 5 as core 5 on socket 0 >> EAL: Detected lcore 6 as core 6 on socket 0 >> EAL: Detected lcore 7 as core 7 on socket 0 >> EAL: Support maximum 64 logical core(s) by configuration. >> EAL: Detected 8 lcore(s) >> EAL: Searching for IVSHMEM devices... >> EAL: No IVSHMEM configuration found! >> EAL: Setting up memory... >> EAL: Ask a virtual area of 0x5ffc0 bytes >> EAL: Virtual area found at 0x2aa4aae0 (size = 0x5ffc0) >> >> EAL: Ask a virtual area of 0x20 bytes >> EAL: Virtual area found at 0x73e0 (size = 0x20) >> EAL: Ask a virtual area of 0x20 bytes >> EAL: Virtual area found at 0x73a0 (size = 0x20) >> EAL: Requesting 12288 pages of size 2MB from socket 0 >> EAL: TSC frequency is ~238 KHz >> EAL: Master core 0 is ready (tid=f7fe9800) >> EAL: Core 3 is ready (tid=f5998700) >> EAL: Core 2 is ready (tid=f6199700) >> EAL: Core 1 is ready (tid=f699a700) >> EAL: PCI device :00:14.0 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: PCI memory mapped at 0x77f8d000 >> EAL: PCI memory mapped at 0x77f89000 >> EAL: PCI device :00:14.1 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: PCI memory mapped at 0x77f69000 >> EAL: PCI memory mapped at 0x77f65000 >> EAL: PCI device :00:14.2 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.2 not managed by UIO driver, skipping >> EAL: PCI device :00:14.3 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.3 not managed by UIO driver, skipping >> EAL: PCI device :00:14.2 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.2 not managed by UIO driver, skipping >> EAL: PCI device :00:14.3 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.3 not managed by UIO driver, skipping >> Initializing port 0 ... Creating queues: nb_rxq=2 nb_txq=4... >> Address:0C:C4:7A:05:52:7A, Allocated mbuf pool on socket 0 >> LPM: Adding route 0x01010100 / 24 (0) >> LPM: Adding route 0x02010100 / 24 (1) >> LPM: Adding route IPV6 / 48 (0) >> LPM: Adding route IPV6 / 48 (1) >> txq=0,0,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> >> Initializing port 1 ... Creating queues: nb_rxq=2 nb_txq=4... >> Address:0C:C4:7A:05:52:7B, txq=0,0,0 PMD: To improve 1G driver >> performance, consider setting the TX WTHRESH value to 4, 8, or 16. >> txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> >> >> Initializing rx queues on lcore 0 ... rxq=0,0,0 >> Initializing rx queues on lcore 1 ... rxq=0,1,0 >> Initializing rx queues on lcore 2 ... rxq=1,0,0 >> Initializing rx queues on lcore 3 ... rxq=1,1,0 >> >> Checking link status.............done >> Port 0 Link Up - speed 100 Mbps - full-d
[dpdk-dev] DPDK supported processor
Well, I did think that the NIC in my testing machine may not support DPDK rather than thinking l3fwd is not working. I just tested the l3fwd sample application to identify whether the NIC supports DPDK or not. Best regards Byeong-Gi KIM 2014-08-21 18:14 GMT+09:00 Chae-yong Chong : > Hi > > Could you give the details why you think the l3fwd is not working > > Best regards, > Chae-yong > > 2014? 8? 21? ???, BYEONG-GI KIM?? ??? ???: > > Thank you for the reply. >> >> I tested l3forwarding sample application, and the results were as below: >> >> sudo ./build/l3fwd -c 0x0F -n 4 -- -p 0x03 >> --config="(0,0,0),(0,1,1),(1,0,2),(1,1,3)" >> EAL: Detected lcore 0 as core 0 on socket 0 >> EAL: Detected lcore 1 as core 1 on socket 0 >> EAL: Detected lcore 2 as core 2 on socket 0 >> EAL: Detected lcore 3 as core 3 on socket 0 >> EAL: Detected lcore 4 as core 4 on socket 0 >> EAL: Detected lcore 5 as core 5 on socket 0 >> EAL: Detected lcore 6 as core 6 on socket 0 >> EAL: Detected lcore 7 as core 7 on socket 0 >> EAL: Support maximum 64 logical core(s) by configuration. >> EAL: Detected 8 lcore(s) >> EAL: Searching for IVSHMEM devices... >> EAL: No IVSHMEM configuration found! >> EAL: Setting up memory... >> EAL: Ask a virtual area of 0x5ffc0 bytes >> EAL: Virtual area found at 0x2aa4aae0 (size = 0x5ffc0) >> >> EAL: Ask a virtual area of 0x20 bytes >> EAL: Virtual area found at 0x73e0 (size = 0x20) >> EAL: Ask a virtual area of 0x20 bytes >> EAL: Virtual area found at 0x73a0 (size = 0x20) >> EAL: Requesting 12288 pages of size 2MB from socket 0 >> EAL: TSC frequency is ~238 KHz >> EAL: Master core 0 is ready (tid=f7fe9800) >> EAL: Core 3 is ready (tid=f5998700) >> EAL: Core 2 is ready (tid=f6199700) >> EAL: Core 1 is ready (tid=f699a700) >> EAL: PCI device :00:14.0 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: PCI memory mapped at 0x77f8d000 >> EAL: PCI memory mapped at 0x77f89000 >> EAL: PCI device :00:14.1 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: PCI memory mapped at 0x77f69000 >> EAL: PCI memory mapped at 0x77f65000 >> EAL: PCI device :00:14.2 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.2 not managed by UIO driver, skipping >> EAL: PCI device :00:14.3 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.3 not managed by UIO driver, skipping >> EAL: PCI device :00:14.2 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.2 not managed by UIO driver, skipping >> EAL: PCI device :00:14.3 on NUMA socket -1 >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> EAL: :00:14.3 not managed by UIO driver, skipping >> Initializing port 0 ... Creating queues: nb_rxq=2 nb_txq=4... >> Address:0C:C4:7A:05:52:7A, Allocated mbuf pool on socket 0 >> LPM: Adding route 0x01010100 / 24 (0) >> LPM: Adding route 0x02010100 / 24 (1) >> LPM: Adding route IPV6 / 48 (0) >> LPM: Adding route IPV6 / 48 (1) >> txq=0,0,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> >> Initializing port 1 ... Creating queues: nb_rxq=2 nb_txq=4... >> Address:0C:C4:7A:05:52:7B, txq=0,0,0 PMD: To improve 1G driver >> performance, consider setting the TX WTHRESH value to 4, 8, or 16. >> txq=1,1,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=2,2,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> txq=3,3,0 PMD: To improve 1G driver performance, consider setting the TX >> WTHRESH value to 4, 8, or 16. >> >> >> Initializing rx queues on lcore 0 ... rxq=0,0,0 >> Initializing rx queues on lcore 1 ... rxq=0,1,0 >> Initializing rx queues on lcore 2 ... rxq=1,0,0 >> Initializing rx queues on lcore 3 ... rxq=1,1,0 >> >> Checking link status.....done >> Port 0 Link Up - speed 100 Mbps - full-duplex >> Port 1 Link Up - speed 100 Mbps - full-duplex >> L3FWD: enterin
[dpdk-dev] DPDK supported processor
It works fine with DPDK at least. The problem is that OVDK, especially ovs-dpdk command supported by the OVDK, doesn't work correctly. I tested the DPDK application, l3fwd, in order to know where the incorrect operation comes from, either DPDK or OVDK itself. And I know now the issue seems belonging to OVDK. :) In the OVDK, the physical ports even could not detected... Best regards Byeong-Gi KIM 2014-08-21 19:00 GMT+09:00 Masaru Oki : > I have SUPERMICRO A1SRi-2758F mainboard (onboard Atom C2758), > and onboard NICs work fine with DPDK application. > > 2014-08-21 18:17 GMT+09:00 BYEONG-GI KIM : > > Well, I did think that the NIC in my testing machine may not support DPDK >> rather than thinking l3fwd is not working. >> >> I just tested the l3fwd sample application to identify whether the NIC >> supports DPDK or not. >> >> Best regards >> >> Byeong-Gi KIM >> >> >> 2014-08-21 18:14 GMT+09:00 Chae-yong Chong : >> >> > Hi >> > >> > Could you give the details why you think the l3fwd is not working >> > >> > Best regards, >> > Chae-yong >> > >> > 2014? 8? 21? ???, BYEONG-GI KIM?? ??? ???: >> > >> > Thank you for the reply. >> >> >> >> I tested l3forwarding sample application, and the results were as >> below: >> >> >> >> sudo ./build/l3fwd -c 0x0F -n 4 -- -p 0x03 >> >> --config="(0,0,0),(0,1,1),(1,0,2),(1,1,3)" >> >> EAL: Detected lcore 0 as core 0 on socket 0 >> >> EAL: Detected lcore 1 as core 1 on socket 0 >> >> EAL: Detected lcore 2 as core 2 on socket 0 >> >> EAL: Detected lcore 3 as core 3 on socket 0 >> >> EAL: Detected lcore 4 as core 4 on socket 0 >> >> EAL: Detected lcore 5 as core 5 on socket 0 >> >> EAL: Detected lcore 6 as core 6 on socket 0 >> >> EAL: Detected lcore 7 as core 7 on socket 0 >> >> EAL: Support maximum 64 logical core(s) by configuration. >> >> EAL: Detected 8 lcore(s) >> >> EAL: Searching for IVSHMEM devices... >> >> EAL: No IVSHMEM configuration found! >> >> EAL: Setting up memory... >> >> EAL: Ask a virtual area of 0x5ffc0 bytes >> >> EAL: Virtual area found at 0x2aa4aae0 (size = 0x5ffc0) >> >> >> >> EAL: Ask a virtual area of 0x20 bytes >> >> EAL: Virtual area found at 0x73e0 (size = 0x20) >> >> EAL: Ask a virtual area of 0x20 bytes >> >> EAL: Virtual area found at 0x73a0 (size = 0x20) >> >> EAL: Requesting 12288 pages of size 2MB from socket 0 >> >> EAL: TSC frequency is ~238 KHz >> >> EAL: Master core 0 is ready (tid=f7fe9800) >> >> EAL: Core 3 is ready (tid=f5998700) >> >> EAL: Core 2 is ready (tid=f6199700) >> >> EAL: Core 1 is ready (tid=f699a700) >> >> EAL: PCI device :00:14.0 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: PCI memory mapped at 0x77f8d000 >> >> EAL: PCI memory mapped at 0x77f89000 >> >> EAL: PCI device :00:14.1 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: PCI memory mapped at 0x77f69000 >> >> EAL: PCI memory mapped at 0x77f65000 >> >> EAL: PCI device :00:14.2 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: :00:14.2 not managed by UIO driver, skipping >> >> EAL: PCI device :00:14.3 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: :00:14.3 not managed by UIO driver, skipping >> >> EAL: PCI device :00:14.2 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: :00:14.2 not managed by UIO driver, skipping >> >> EAL: PCI device :00:14.3 on NUMA socket -1 >> >> EAL: probe driver: 8086:1f41 rte_igb_pmd >> >> EAL: :00:14.3 not managed by UIO driver, skipping >> >> Initializing port 0 ... Creating queues: nb_rxq=2 nb_txq=4... >> >> Address:0C:C4:7A:05:52:7A, Allocated mbuf pool on socket 0 >> >> LPM: Adding route 0x01010100 / 24 (0) >> >> LPM: Adding route 0x02010100 / 24 (1) >> >> LPM: Adding route IPV6 / 48 (0) >> >> LPM: Adding route IPV6 / 48 (1) >> >> txq=0,0,0 PMD: To improve 1G driver performance, consider setting the >> TX >> >> WTHRESH value to 4, 8, or 16. >>
[dpdk-dev] DPDK supported processor
Dear Cao Min Thank you for the information. Best regards Byeong-Gi KIM 2014-08-22 14:30 GMT+09:00 Cao, Min : > Hi Byeong-Gi KIM, > > We tested with platform as below. Pls take information below as reference. > > Cpu: Genuine Intel(R) CPU 4000 @ 2.41GHz > NIC: Niantic 82599 > Bios: EDVLCRB1.86B.0008.R00.1303061138 > OS: 3.4.18-yocto-standard(32/64 bit) or FC20 > > We validated the test cases list below: >pmd >Checksum >Ipv4frag >Hello_world >Flow director >Jumboframe and scatter >Ieee1588 >Multiprocess >Flow control >Whitelist >blacklist >vlan > > Regards, > Cao Min > > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of BYEONG-GI KIM > Sent: Thursday, August 21, 2014 2:04 PM > To: dev at dpdk.org > Subject: [dpdk-dev] DPDK supported processor > > Hello. > > I'm now using Intel Atom processor C2758 on the DPDK testing machines, but > it seems not working. > > Which sample application would be recommended to test the NIC actually > works well with DPDK? > > Best regards > > Byeong-Gi KIM > >
[dpdk-dev] Is it possible to install the current version of OVDK with DPDK 1.7.0?
Hello. I've noticed that DPDK 1.7.0 version has been released a few days ago. The DPDK vSwitch package, however, seems supporting old version, 1.6.x, only. Could the DPDK vSwitch be compiled by the latest DPDK? Or Do I need to use 1.6.x version of DPDK? Thanks in advance. Best regards Byeong-Gi KIM
[dpdk-dev] Error while compiling DPDK
Hello. An error occured while compiling DPDK 1.6.0r2 as below: [root at compute dpdk-1.6.0r2]# make install T=x86_64*gcc ...Installing... /ovdk/dpdk/dpdk-1.6.0r2/lib/librte_eal/bsdapp/eal/eal.c: In function ?rte_eal_init?: /ovdk/dpdk/dpdk-1.6.0r2/lib/librte_eal/bsdapp/eal/eal.c:873:2: error: format ?%p? expects argument of type ?void *?, but argument 5 has type ?pthread_t? [-Werror=format=] RTE_LOG(DEBUG, EAL, "Master core %u is ready (tid=%p)\n", ^ The current OS is Fedora20. What should I do?... Best regards Byeong-Gi KIM
[dpdk-dev] Error while compiling DPDK
Dear choi, Thank you for the reply. I've been trying to install DPDK while changing the command as you mentioned, i.e., I used T=x86_64-default-linuxapp-gcc, but error was still shown; [root at compute DPDK-1.6.0]# make install T=x86_64-default-linuxapp-gcc == Installing x86_64-default-linuxapp-gcc Configuration done == Build scripts == Build scripts/testhost HOSTCC testhost.o HOSTLD testhost INSTALL-HOSTAPP testhost == Build lib == Build lib/librte_eal == Build lib/librte_eal/common SYMLINK-FILE include/rte_atomic.h SYMLINK-FILE include/rte_branch_prediction.h SYMLINK-FILE include/rte_byteorder.h SYMLINK-FILE include/rte_common.h SYMLINK-FILE include/rte_cycles.h SYMLINK-FILE include/rte_debug.h SYMLINK-FILE include/rte_eal.h SYMLINK-FILE include/rte_errno.h SYMLINK-FILE include/rte_launch.h SYMLINK-FILE include/rte_lcore.h SYMLINK-FILE include/rte_log.h SYMLINK-FILE include/rte_memcpy.h SYMLINK-FILE include/rte_memory.h SYMLINK-FILE include/rte_memzone.h SYMLINK-FILE include/rte_pci.h SYMLINK-FILE include/rte_pci_dev_ids.h SYMLINK-FILE include/rte_per_lcore.h SYMLINK-FILE include/rte_prefetch.h SYMLINK-FILE include/rte_random.h SYMLINK-FILE include/rte_rwlock.h SYMLINK-FILE include/rte_spinlock.h SYMLINK-FILE include/rte_tailq.h SYMLINK-FILE include/rte_interrupts.h SYMLINK-FILE include/rte_alarm.h SYMLINK-FILE include/rte_string_fns.h SYMLINK-FILE include/rte_cpuflags.h SYMLINK-FILE include/rte_version.h SYMLINK-FILE include/rte_tailq_elem.h SYMLINK-FILE include/rte_eal_memconfig.h SYMLINK-FILE include/rte_malloc_heap.h SYMLINK-FILE include/rte_hexdump.h SYMLINK-FILE include/arch/rte_atomic.h == Build lib/librte_eal/linuxapp == Build lib/librte_eal/linuxapp/igb_uio LD /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/built-in.o CC [M] /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.o Building modules, stage 2. MODPOST 1 modules CC /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.mod.o LD [M] /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.ko INSTALL-MODULE igb_uio.ko == Build lib/librte_eal/linuxapp/eal CC eal.o gcc: error: ?g: ?? ? make[7]: *** [eal.o] ?? 1 make[6]: *** [eal] ?? 2 make[5]: *** [linuxapp] ?? 2 make[4]: *** [librte_eal] ?? 2 make[3]: *** [lib] ?? 2 make[2]: *** [all] ?? 2 make[1]: *** [x86_64-default-linuxapp-gcc_install] ?? 2 make: *** [install] ?? 2 I'm sorry for the Korean characteristics. It was shown because My Fedora OS is set to use Korean as a default language. I set two environmental variables following the intel-dpdk-programmers-guide.pdf as below: export RTE_SDK=/home/stack/dpdk/DPDK-1.6.0 export RTE_TARGET=x86_64-native-linuxapp-gcc What should I do in order to install DPDK successfully?... Thanks in advance. Best regards Byeong-Gi KIM 2014-07-14 10:00 GMT+09:00 Choi, Sy Jong : > Hi Mr. Kim, > > Please do not compile with "T=x86*gcc" because it will compile BSD app > too, which cause the error. > > Please complile only the configuration you need. This should remove the > error. > > > Regards, > Choi, Sy Jong > > > -Original Message- > *From: *BYEONG-GI KIM [kimbyeonggi at gmail.com] > *Sent: *Monday, July 14, 2014 08:25 AM Malay Peninsula Standard Time > *To: *Dpdk-ovs at lists.01.org; dev at dpdk.org > *Subject: *[dpdk-dev] Error while compiling DPDK > > Hello. > > An error occured while compiling DPDK 1.6.0r2 as below: > > [root at compute dpdk-1.6.0r2]# make install T=x86_64*gcc > ...Installing... > /ovdk/dpdk/dpdk-1.6.0r2/lib/librte_eal/bsdapp/eal/eal.c: In function > ?rte_eal_init?: > /ovdk/dpdk/dpdk-1.6.0r2/lib/librte_eal/bsdapp/eal/eal.c:873:2: error: > format ?%p? expects argument of type ?void *?, but argument 5 has type > ?pthread_t? [-Werror=format=] > RTE_LOG(DEBUG, EAL, "Master core %u is ready (tid=%p)\n", > ^ > > The current OS is Fedora20. What should I do?... > > Best regards > > Byeong-Gi KIM >
[dpdk-dev] Error while compiling DPDK
Here is the error message in English; == Build lib/librte_eal/linuxapp == Build lib/librte_eal/linuxapp/igb_uio LD /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/built-in.o CC [M] /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.o Building modules, stage 2. MODPOST 1 modules CC /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.mod.o LD [M] /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.ko INSTALL-MODULE igb_uio.ko == Build lib/librte_eal/linuxapp/eal CC eal.o gcc: error: ?g: No such file or directory make[7]: *** [eal.o] Error 1 make[6]: *** [eal] Error 2 make[5]: *** [linuxapp] Error 2 make[4]: *** [librte_eal] Error 2 make[3]: *** [lib] Error 2 make[2]: *** [all] Error 2 make[1]: *** [x86_64-default-linuxapp-gcc_install] Error 2 make: *** [install] Error 2 2014-07-14 11:02 GMT+09:00 Masaru Oki : > > I'm sorry for the Korean characteristics. It was shown because My Fedora > OS > > is set to use Korean as a default language. > > Try env LANG=C make install ... > > > 2014-07-14 10:24 GMT+09:00 BYEONG-GI KIM : > >> Dear choi, >> >> Thank you for the reply. >> >> I've been trying to install DPDK while changing the command as you >> mentioned, i.e., I used T=x86_64-default-linuxapp-gcc, but error was still >> shown; >> >> >> [root at compute DPDK-1.6.0]# make install T=x86_64-default-linuxapp-gcc >> == Installing x86_64-default-linuxapp-gcc >> Configuration done >> == Build scripts >> == Build scripts/testhost >> HOSTCC testhost.o >> HOSTLD testhost >> INSTALL-HOSTAPP testhost >> == Build lib >> == Build lib/librte_eal >> == Build lib/librte_eal/common >> SYMLINK-FILE include/rte_atomic.h >> SYMLINK-FILE include/rte_branch_prediction.h >> SYMLINK-FILE include/rte_byteorder.h >> SYMLINK-FILE include/rte_common.h >> SYMLINK-FILE include/rte_cycles.h >> SYMLINK-FILE include/rte_debug.h >> SYMLINK-FILE include/rte_eal.h >> SYMLINK-FILE include/rte_errno.h >> SYMLINK-FILE include/rte_launch.h >> SYMLINK-FILE include/rte_lcore.h >> SYMLINK-FILE include/rte_log.h >> SYMLINK-FILE include/rte_memcpy.h >> SYMLINK-FILE include/rte_memory.h >> SYMLINK-FILE include/rte_memzone.h >> SYMLINK-FILE include/rte_pci.h >> SYMLINK-FILE include/rte_pci_dev_ids.h >> SYMLINK-FILE include/rte_per_lcore.h >> SYMLINK-FILE include/rte_prefetch.h >> SYMLINK-FILE include/rte_random.h >> SYMLINK-FILE include/rte_rwlock.h >> SYMLINK-FILE include/rte_spinlock.h >> SYMLINK-FILE include/rte_tailq.h >> SYMLINK-FILE include/rte_interrupts.h >> SYMLINK-FILE include/rte_alarm.h >> SYMLINK-FILE include/rte_string_fns.h >> SYMLINK-FILE include/rte_cpuflags.h >> SYMLINK-FILE include/rte_version.h >> SYMLINK-FILE include/rte_tailq_elem.h >> SYMLINK-FILE include/rte_eal_memconfig.h >> SYMLINK-FILE include/rte_malloc_heap.h >> SYMLINK-FILE include/rte_hexdump.h >> SYMLINK-FILE include/arch/rte_atomic.h >> == Build lib/librte_eal/linuxapp >> == Build lib/librte_eal/linuxapp/igb_uio >> LD >> >> >> /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/built-in.o >> CC [M] >> >> >> /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.o >> Building modules, stage 2. >> MODPOST 1 modules >> CC >> >> >> /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.mod.o >> LD [M] >> >> >> /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.ko >> INSTALL-MODULE igb_uio.ko >> == Build lib/librte_eal/linuxapp/eal >> CC eal.o >> gcc: error: ?g: ?? ? >> make[7]: *** [eal.o] ?? 1 >> make[6]: *** [eal] ?? 2 >> make[5]: *** [linuxapp] ?? 2 >> make[4]: *** [librte_eal] ?? 2 >> make[3]: *** [lib] ?? 2 >> make[2]: *** [all] ?? 2 >> make[1]: *** [x86_64-default-linuxapp-gcc_install] ?? 2 >> make: *** [install] ?? 2 >> >> I'm sorry for the Korean characteristics. It was shown because My Fedora >> OS >> is set to use Korean as a default language. >> >> I set two environmental variables following the >> intel-dpdk-programmers-guide.pdf as below: >> >&
[dpdk-dev] Error while compiling DPDK
Dear Choi I used x86_64-default-linuxapp-gcc. I also checked x86_64-native-linuxapp-gcc after exporting RTE_TARGET=x86_64-native-linuxapp-gcc, but there was a message "make[1]: Nothing to be done for `install'." while trying to install it. Best regards Byeong-Gi KIM 2014-07-14 11:40 GMT+09:00 Choi, Sy Jong : > Hi Mr. Kim, > > May confirm which environment are you using? > x86_64-default-linuxapp-gcc or > x86_64-native-linuxapp-gcc > > > > > Regards, > Choi, Sy Jong > > -Original Message- > *From: *BYEONG-GI KIM [kimbyeonggi at gmail.com] > *Sent: *Monday, July 14, 2014 10:05 AM Malay Peninsula Standard Time > *To: *Masaru Oki; dev at dpdk.org; Dpdk-ovs at lists.01.org > *Subject: *Re: [dpdk-dev] Error while compiling DPDK > > Here is the error message in English; > > == Build lib/librte_eal/linuxapp > == Build lib/librte_eal/linuxapp/igb_uio > LD > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/built-in.o > CC [M] > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.o > Building modules, stage 2. > MODPOST 1 modules > CC > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.mod.o > LD [M] > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.ko > INSTALL-MODULE igb_uio.ko > == Build lib/librte_eal/linuxapp/eal > CC eal.o > gcc: error: ?g: No such file or directory > make[7]: *** [eal.o] Error 1 > make[6]: *** [eal] Error 2 > make[5]: *** [linuxapp] Error 2 > make[4]: *** [librte_eal] Error 2 > make[3]: *** [lib] Error 2 > make[2]: *** [all] Error 2 > make[1]: *** [x86_64-default-linuxapp-gcc_install] Error 2 > make: *** [install] Error 2 > > > > 2014-07-14 11:02 GMT+09:00 Masaru Oki : > > > > I'm sorry for the Korean characteristics. It was shown because My > Fedora > > OS > > > is set to use Korean as a default language. > > > > Try env LANG=C make install ... > > > > > > 2014-07-14 10:24 GMT+09:00 BYEONG-GI KIM : > > > >> Dear choi, > >> > >> Thank you for the reply. > >> > >> I've been trying to install DPDK while changing the command as you > >> mentioned, i.e., I used T=x86_64-default-linuxapp-gcc, but error was > still > >> shown; > >> > >> > >> [root at compute DPDK-1.6.0]# make install T=x86_64-default-linuxapp-gcc > >> == Installing x86_64-default-linuxapp-gcc > >> Configuration done > >> == Build scripts > >> == Build scripts/testhost > >> HOSTCC testhost.o > >> HOSTLD testhost > >> INSTALL-HOSTAPP testhost > >> == Build lib > >> == Build lib/librte_eal > >> == Build lib/librte_eal/common > >> SYMLINK-FILE include/rte_atomic.h > >> SYMLINK-FILE include/rte_branch_prediction.h > >> SYMLINK-FILE include/rte_byteorder.h > >> SYMLINK-FILE include/rte_common.h > >> SYMLINK-FILE include/rte_cycles.h > >> SYMLINK-FILE include/rte_debug.h > >> SYMLINK-FILE include/rte_eal.h > >> SYMLINK-FILE include/rte_errno.h > >> SYMLINK-FILE include/rte_launch.h > >> SYMLINK-FILE include/rte_lcore.h > >> SYMLINK-FILE include/rte_log.h > >> SYMLINK-FILE include/rte_memcpy.h > >> SYMLINK-FILE include/rte_memory.h > >> SYMLINK-FILE include/rte_memzone.h > >> SYMLINK-FILE include/rte_pci.h > >> SYMLINK-FILE include/rte_pci_dev_ids.h > >> SYMLINK-FILE include/rte_per_lcore.h > >> SYMLINK-FILE include/rte_prefetch.h > >> SYMLINK-FILE include/rte_random.h > >> SYMLINK-FILE include/rte_rwlock.h > >> SYMLINK-FILE include/rte_spinlock.h > >> SYMLINK-FILE include/rte_tailq.h > >> SYMLINK-FILE include/rte_interrupts.h > >> SYMLINK-FILE include/rte_alarm.h > >> SYMLINK-FILE include/rte_string_fns.h > >> SYMLINK-FILE include/rte_cpuflags.h > >> SYMLINK-FILE include/rte_version.h > >> SYMLINK-FILE include/rte_tailq_elem.h > >> SYMLINK-FILE include/rte_eal_memconfig.h > >> SYMLINK-FILE include/rte_malloc_heap.h > >> SYMLINK-FILE include/rte_hexdump.h > >> SYMLINK-FILE include/arch/rte_atomic.h > >> == Build lib/librte_eal/linuxapp > >> == Build lib/librte_eal/linuxapp/igb_uio > >> LD > >> > >> > /home/stack/dpdk/DPDK-1.6.0/x86_64-
[dpdk-dev] Error while compiling DPDK
Dear Choi, Of course I used "export RTE_TARGET=x86_64-default-linuxapp-gcc" when I was trying to compile it. By the way, the compilation went off to one more step, there was another error message though; == Build lib/librte_eal/linuxapp/kni LD /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni/built-in.o CC [M] /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni/ixgbe_main.o /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni/ixgbe_main.c: In function 'ixgbe_del_mac_filter': /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni/ixgbe_main.c:525:3: error: implicit declaration of function 'compare_ether_addr' [-Werror=implicit-function-declaration] if (!compare_ether_addr(addr, adapter->mac_table[i].addr) && ^ cc1: all warnings being treated as errors make[10]: *** [/home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni/ixgbe_main.o] Error 1 make[9]: *** [_module_/home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/kni] Error 2 make[8]: *** [sub-make] Error 2 make[7]: *** [rte_kni.ko] Error 2 make[6]: *** [kni] Error 2 make[5]: *** [linuxapp] Error 2 make[4]: *** [librte_eal] Error 2 make[3]: *** [lib] Error 2 make[2]: *** [all] Error 2 make[1]: *** [x86_64-default-linuxapp-gcc_install] Error 2 make: *** [install] Error 2 Honestly, I've never been able to compile DPDK successfully... I had been trying to compile it on CentOS 6.5, and was failed. Then I changed OS from CentOS to Fedora, but there are still lots of errors. Best regards Byeong-Gi KIM 2014-07-14 13:32 GMT+09:00 Choi, Sy Jong : > Hi Mr. Kim, > > > > If you are compiling x86_64-default-linuxapp-gcc > > The RTE_TARGET=x86_64-default-linuxapp-gcc > > > > I have manage to compile DPDK r1.6.0 with the following OS and kernel > version > > CentOS with customer kernel > > Linux localhost.localdomain 3.3.1 #1 SMP Tue Jan 7 22:44:17 MYT 2014 > x86_64 x86_64 x86_64 GNU/Linux > > > > Ubuntu 14.04 > > Linux choi-desktop 3.13.0-30-generic #55-Ubuntu SMP Fri Jul 4 21:40:53 UTC > 2014 x86_64 x86_64 x86_64 GNU/Linux > > > > May I know, have you been able to compile DPDK previously? > > > > Regards, > > Choi, Sy Jong > > Platform Application Engineer > > > > *From:* BYEONG-GI KIM [mailto:kimbyeonggi at gmail.com] > *Sent:* Monday, July 14, 2014 11:29 AM > *To:* Choi, Sy Jong; dev at dpdk.org; Dpdk-ovs at lists.01.org > > *Subject:* Re: [dpdk-dev] Error while compiling DPDK > > > > Dear Choi > > > > I used x86_64-default-linuxapp-gcc. > > > > I also checked x86_64-native-linuxapp-gcc after exporting > RTE_TARGET=x86_64-native-linuxapp-gcc, but there was a message "make[1]: > Nothing to be done for `install'." while trying to install it. > > > > Best regards > > > > Byeong-Gi KIM > > > > 2014-07-14 11:40 GMT+09:00 Choi, Sy Jong : > > Hi Mr. Kim, > > May confirm which environment are you using? > x86_64-default-linuxapp-gcc or > x86_64-native-linuxapp-gcc > > > > > > Regards, > Choi, Sy Jong > > -Original Message- > > *From: *BYEONG-GI KIM [kimbyeonggi at gmail.com] > *Sent: *Monday, July 14, 2014 10:05 AM Malay Peninsula Standard Time > *To: *Masaru Oki; dev at dpdk.org; Dpdk-ovs at lists.01.org > *Subject: *Re: [dpdk-dev] Error while compiling DPDK > > Here is the error message in English; > > == Build lib/librte_eal/linuxapp > == Build lib/librte_eal/linuxapp/igb_uio > LD > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/built-in.o > CC [M] > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.o > Building modules, stage 2. > MODPOST 1 modules > CC > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.mod.o > LD [M] > > > /home/stack/dpdk/DPDK-1.6.0/x86_64-default-linuxapp-gcc/build/lib/librte_eal/linuxapp/igb_uio/igb_uio.ko > INSTALL-MODULE igb_uio.ko > == Build lib/librte_eal/linuxapp/eal > CC eal.o > gcc: error: ?g: No such file or directory > make[7]: *** [eal.o] Error 1 > make[6]: *** [eal] Error 2 > make[5]: *** [linuxapp] Error 2 > make[4]: *** [librte_eal] Error 2 > make[3]: *** [lib] Error 2 > make[2]: *** [all] Error 2 > make[1]: *** [x86_64-default-linuxapp-gcc_install] Error 2 > make: *** [install] Error 2 > > > > 2014-07-14 11:02 GMT+09:00 Masaru Oki : > > > > I'm sorr
[dpdk-dev] No "pci_unbind.py" in tools subdirectory
Hello. I'm sorry for the repeated question. I finally compiled DPDK successfully (there was a problem in the source code what I downloaded), and I'm moving to compile dpdk-ovs. Before compiling it, I'm trying to binding a network port to igb_uio module. In the dpdk-getting-started-guide, "pci_unbind.py" utility script is used to provide a view of the current state of the network ports on the system, and to bind/unbind those ports from the different kernel modules. The script, however, was not there in tools directory; There were only "igb_uio_bind.py". Is it fine as it is, or do I miss something? I downloaded the DPDK source from http://dpdk.org/browse/dpdk/refs/, and the file was dpdk-1.6.0r2.zip. I compiled both x86-64-defaultlinuxapp-gcc and x86-64-ivshmem-linuxapp-gcc now. Thanks in advance. Best regards Byeong-Gi KIM
[dpdk-dev] No "pci_unbind.py" in tools subdirectory
Thank you for the reply. By the way, I got an error message while compiling Open vSwitch. Is it essential to bind a port to igb_uio module in order to compile Open vSwitch without error? I just wonder whether the errors attached below just happened because I did not bind the port to the module or not... make RTE_TARGET=x86_64-ivshmem-linuxapp-gcc ... /home/stack/dpdk/dpdk-1.6.0r2/x86_64-ivshmem-linuxapp-gcc/lib/librte_eal.a(eal.o): In function `rte_eal_init': eal.c:(.text+0xedf): undefined reference to `dlopen' eal.c:(.text+0xefe): undefined reference to `dlerror' collect2: error: ld returned 1 exit status make[2]: *** [utilities/ovs-controller] Error 1 make[2]: Leaving directory `/home/stack/dpdk-ovs/openvswitch' make[1]: *** [all-recursive] Error 1 make[1]: Leaving directory `/home/stack/dpdk-ovs/openvswitch' make: *** [all] Error 2 Best regards Byeong-Gi KIM 2014-07-14 17:10 GMT+09:00 De Lara Guarch, Pablo < pablo.de.lara.guarch at intel.com>: > > -Original Message- > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Aravind > > Sent: Monday, July 14, 2014 8:52 AM > > To: BYEONG-GI KIM > > Cc: dev at dpdk.org; Dpdk-ovs at lists.01.org > > Subject: Re: [dpdk-dev] No "pci_unbind.py" in tools subdirectory > > > > Hi, > > > > You should be using igb_uio_bind.py , have a look at the quick start > guide > > : http://dpdk.org/doc/quick-start > > > > Actually, now it is called dpdk_nic_bind.py :) > > > > > > > On Mon, Jul 14, 2014 at 1:09 PM, BYEONG-GI KIM > > wrote: > > > > > Hello. > > > > > > I'm sorry for the repeated question. > > > > > > I finally compiled DPDK successfully (there was a problem in the source > > > code what I downloaded), and I'm moving to compile dpdk-ovs. > > > > > > Before compiling it, I'm trying to binding a network port to igb_uio > > > module. In the dpdk-getting-started-guide, "pci_unbind.py" utility > script > > > is used to provide a view of the current state of the network ports on > the > > > system, and to bind/unbind those ports from the different kernel > modules. > > > > > > The script, however, was not there in tools directory; There were only > > > "igb_uio_bind.py". Is it fine as it is, or do I miss something? > > > > > > I downloaded the DPDK source from http://dpdk.org/browse/dpdk/refs/, > and > > > the file was dpdk-1.6.0r2.zip. I compiled both > x86-64-defaultlinuxapp-gcc > > > and x86-64-ivshmem-linuxapp-gcc now. > > > > > > Thanks in advance. > > > > > > Best regards > > > > > > Byeong-Gi KIM > > > > >
