On Thu, 2 Jan 2014 22:22:27 +0530 Raj Ravi <[email protected]> wrote:
> Hi, > Thanks. > > I just tried this : > modprobe ixgbe max_vfs=2,2 > But it creates 4 devices "ONLY " for the first port. > > # virsh nodedev-list | grep 25 > pci_0000_25_00_0 > pci_0000_25_00_1 > *pci_0000_25_10_0* > *pci_0000_25_10_1* > *pci_0000_25_10_2* > *pci_0000_25_10_3* No, those are for the two ports. The VF stride is 2 for dual port 82599 so that means 25:10.0 and 25:10.2 are for the first port and 25:10.1 and 25:10.3 are for the second port. - Greg > # > > dmesg output while carrying out this : > > modprobe ixgbe max_vfs=2,2 > > pps_core: LinuxPPS API ver. 1 registered > pps_core: Software ver. 5.3.6 - Copyright 2005-2007 Rodolfo Giometti < > [email protected]> > PTP clock support registered > Intel(R) 10 Gigabit PCI Express Network Driver - version 3.18.7 > Copyright (c) 1999-2013 Intel Corporation. > ixgbe 0000:25:00.0: PCI INT A -> GSI 32 (level, low) -> IRQ 32 > ixgbe: I/O Virtualization (IOV) set to 2 > ixgbe: 0000:25:00.0: ixgbe_check_options: FCoE Offload feature enabled > ixgbe 0000:25:00.0: (unregistered net_device): SR-IOV enabled with 2 > VFs ixgbevf: Intel(R) 10 Gigabit PCI Express Virtual Function Network > Driver - version 2.11.3 > Copyright (c) 2009-2012 Intel Corporation. > ixgbe 0000:25:00.0: (PCI Express:5.0GT/s:Width x8) 90:e2:ba:39:44:ac > ixgbe 0000:25:00.0: eth0: MAC: 2, PHY: 12, SFP+: 5, PBA No: E66560-000 > ixgbe 0000:25:00.0: PCI Express bandwidth of 32GT/s available > ixgbe 0000:25:00.0: (Speed:5.0GT/s, Width: x8, Encoding Loss:20%) > ixgbe 0000:25:00.0: eth0: Enabled Features: RxQ: 1 TxQ: 1 DCA > ixgbe 0000:25:00.0: eth0: Intel(R) 10 Gigabit Network Connection > ixgbevf 0000:25:10.0: enabling device (0000 -> 0002) > ixgbevf 0000:25:10.0: PF still in reset state. Is the PF interface > up? ixgbevf 0000:25:10.0: Assigning random MAC address > ixgbevf: eth%d: ixgbevf_init_interrupt_scheme: Multiqueue Disabled: Rx > Queue count = 1, Tx Queue count = 1 > ADDRCONF(NETDEV_UP): eth0: link is not ready > 8021q: adding VLAN 0 to HW filter on device eth0 > ixgbevf: eth1: ixgbevf_probe: Intel(R) 82599 Virtual Function > 4e:6b:80:df:b7:59 > ixgbevf: eth1: ixgbevf_probe: LRO is disabled > ixgbevf: eth1: ixgbevf_probe: GRO is enabled > ixgbevf: eth1: ixgbevf_probe: Intel(R) 10 Gigabit PCI Express Virtual > Function Network Driver > ixgbe 0000:25:00.1: PCI INT B -> GSI 36 (level, low) -> IRQ 36 > ixgbe: I/O Virtualization (IOV) set to 2 > ixgbe: 0000:25:00.1: ixgbe_check_options: FCoE Offload feature enabled > ixgbe 0000:25:00.0: eth0: detected SFP+: 5 > ixgbe 0000:25:00.1: (unregistered net_device): SR-IOV enabled with 2 > VFs ixgbe 0000:25:00.1: (PCI Express:5.0GT/s:Width x8) > 90:e2:ba:39:44:ad ixgbe 0000:25:00.1: eth2: MAC: 2, PHY: 12, SFP+: 6, > PBA No: E66560-000 udev: renamed network interface eth2 to rename49 > ixgbe 0000:25:00.1: PCI Express bandwidth of 32GT/s available > ixgbe 0000:25:00.1: (Speed:5.0GT/s, Width: x8, Encoding Loss:20%) > ixgbe 0000:25:00.1: rename49: Enabled Features: RxQ: 1 TxQ: 1 DCA > ixgbe 0000:25:00.1: rename49: Intel(R) 10 Gigabit Network Connection > ixgbe 0000:2d:00.0: PCI INT A -> GSI 40 (level, low) -> IRQ 40 > ixgbevf 0000:25:10.2: enabling device (0000 -> 0002) > ixgbe: 0000:2d:00.0: ixgbe_check_options: FCoE Offload feature enabled > ixgbe 0000:25:00.0: eth0: VF Reset msg received from vf 1 > ixgbe 0000:25:00.0: VF 1 has no MAC address assigned, you may have to > assign one manually > ixgbevf 0000:25:10.2: MAC address not assigned by administrator. > ixgbevf 0000:25:10.2: Assigning random MAC address > ixgbevf: eth%d: ixgbevf_init_interrupt_scheme: Multiqueue Enabled: Rx > Queue count = 2, Tx Queue count = 2 > ixgbevf: eth2: ixgbevf_probe: Intel(R) 82599 Virtual Function > c6:93:84:41:65:f1 > ixgbevf: eth2: ixgbevf_probe: LRO is disabled > ixgbevf: eth2: ixgbevf_probe: GRO is enabled > ixgbevf: eth2: ixgbevf_probe: Intel(R) 10 Gigabit PCI Express Virtual > Function Network Driver > ixgbevf 0000:25:10.1: enabling device (0000 -> 0002) > ixgbevf 0000:25:10.1: PF still in reset state. Is the PF interface > up? ixgbevf 0000:25:10.1: Assigning random MAC address > ixgbevf: eth%d: ixgbevf_init_interrupt_scheme: Multiqueue Disabled: Rx > Queue count = 1, Tx Queue count = 1 > ixgbe 0000:2d:00.0: (PCI Express:5.0GT/s:Width x8) 90:e2:ba:3e:7a:cc > ixgbevf: eth8: ixgbevf_probe: Intel(R) 82599 Virtual Function > 7e:ef:bb:24:eb:c1 > ixgbevf: eth8: ixgbevf_probe: LRO is disabled > ixgbevf: eth8: ixgbevf_probe: GRO is enabled > ixgbevf: eth8: ixgbevf_probe: Intel(R) 10 Gigabit PCI Express Virtual > Function Network Driver > ixgbevf 0000:25:10.3: enabling device (0000 -> 0002) > udev: renamed network interface eth3 to rename51 > ixgbe 0000:2d:00.0: rename51: MAC: 2, PHY: 12, SFP+: 5, PBA No: > E66560-000 ixgbe 0000:2d:00.0: PCI Express bandwidth of 32GT/s > available ixgbevf 0000:25:10.3: PF still in reset state. Is the PF > interface up? ixgbevf 0000:25:10.3: Assigning random MAC address > ixgbevf: eth%d: ixgbevf_init_interrupt_scheme: Multiqueue Disabled: Rx > Queue count = 1, Tx Queue count = 1 > ixgbevf: eth3: ixgbevf_probe: Intel(R) 82599 Virtual Function > 36:b8:12:2e:b4:82 > ixgbevf: eth3: ixgbevf_probe: LRO is disabled > ixgbevf: eth3: ixgbevf_probe: GRO is enabled > ixgbevf: eth3: ixgbevf_probe: Intel(R) 10 Gigabit PCI Express Virtual > Function Network Driver > udev: renamed network interface eth1 to rename48 > udev: renamed network interface eth2 to eth9 > udev: renamed network interface eth8 to eth10 > udev: renamed network interface eth3 to eth13 > udev: renamed network interface rename49 to eth1 > udev: renamed network interface rename51 to eth2 > udev: renamed network interface rename48 to eth8 > ixgbe 0000:2d:00.0: (Speed:5.0GT/s, Width: x8, Encoding Loss:20%) > ixgbe 0000:2d:00.0: eth2: Enabled Features: RxQ: 32 TxQ: 32 FdirHash > DCA ADDRCONF(NETDEV_UP): eth2: link is not ready > 8021q: adding VLAN 0 to HW filter on device eth2 > ixgbe 0000:2d:00.0: eth2: Intel(R) 10 Gigabit Network Connection > ixgbe 0000:2d:00.1: PCI INT B -> GSI 44 (level, low) -> IRQ 44 > ixgbe: 0000:2d:00.1: ixgbe_check_options: FCoE Offload feature enabled > ixgbe 0000:2d:00.0: eth2: detected SFP+: 5 > ADDRCONF(NETDEV_UP): eth1: link is not ready > 8021q: adding VLAN 0 to HW filter on device eth1 > ixgbe 0000:25:00.1: eth1: detected SFP+: 6 > ixgbe 0000:2d:00.1: (PCI Express:5.0GT/s:Width x8) 90:e2:ba:3e:7a:cd > ixgbe 0000:2d:00.1: eth3: MAC: 2, PHY: 12, SFP+: 6, PBA No: E66560-000 > ixgbe 0000:2d:00.1: PCI Express bandwidth of 32GT/s available > ixgbe 0000:2d:00.1: (Speed:5.0GT/s, Width: x8, Encoding Loss:20%) > ixgbe 0000:2d:00.1: eth3: Enabled Features: RxQ: 32 TxQ: 32 FdirHash > DCA ixgbe 0000:2d:00.1: eth3: Intel(R) 10 Gigabit Network Connection > [root@sv2-13 rules.d]# ixgbe 0000:2d:00.0: eth2: NIC Link is Up 10 > Gbps, Flow Control: RX/TX > ADDRCONF(NETDEV_UP): eth3: link is not ready > 8021q: adding VLAN 0 to HW filter on device eth3 > ADDRCONF(NETDEV_CHANGE): eth2: link becomes ready > ixgbe 0000:2d:00.1: eth3: detected SFP+: 6 > ixgbe 0000:2d:00.1: eth3: NIC Link is Up 10 Gbps, Flow Control: RX/TX > ADDRCONF(NETDEV_CHANGE): eth3: link becomes ready > ------------------------------------------------------------------------------------------------------------------ > There are actually four ixgbe ports present : > [root@sv2-13 rules.d]# ethtool -i eth0 > driver: ixgbe > version: 3.18.7 > firmware-version: 0x5e0b0001 > bus-info: 0000:25:00.0 <== > > [root@sv2-13 rules.d]# ethtool -i eth1 > driver: ixgbe > version: 3.18.7 > firmware-version: 0x5e0b0001 > bus-info: 0000:25:00.1 > > [root@sv2-13 rules.d]# ethtool -i eth2 > driver: ixgbe > version: 3.18.7 > firmware-version: 0x5e0b0001 > bus-info: 0000:2d:00.0 > > [root@sv2-13 rules.d]# ethtool -i eth3 > driver: ixgbe > version: 3.18.7 > firmware-version: 0x5e0b0001 > bus-info: 0000:2d:00.1 > ------------------------------------------------------------------------------------------------------------------ > > All the Virtual functions are created for eth0 alone/ > Please refer bus-info for eth0. > Please let me know how to avoid this and get VFs created for all > interfaces. > > > -- > > > > > > > On Thu, Jan 2, 2014 at 10:11 PM, Greg Rose <[email protected]> > wrote: > > > On Thu, 2 Jan 2014 22:01:50 +0530 > > Raj Ravi <[email protected]> wrote: > > > > > Hi, > > > I am trying to update to these drivers(ixgbe-3.18.7.tar.gz > > > , ixgbevf-2.11.3.tar.gz) from existing versions of drivers. > > > > > > With my existing drivers, I can see multiple virtual function for > > > each physical port properly created. > > > > > > for eg: modprobe ixgbe max_vfs=2 > > > This creates two virtual functions for each Physical port. > > > eth0, eth2, eth3 every physical contains two virtual functions. I > > > can verify this with lspci, ethtool -i bus info. > > > > > > I have replaced existing drivers with ixgbe-3.18.7.tar.gz > > > , ixgbevf-2.11.3.tar.gz, (by updating them in /lib/modules/.. ) > > > then tried , modprobe ixgbe max_vfs=2 > > > > > > This creates two virtual function only for "FIRST PORT". > > > In my case eth0 only contains two virtual function. eth2, eth3 > > > are not showing any Virtual function . > > > Any idea about this ? Is there any issue with these drivers ?? I > > > need to load them in any other way ?? > > > > The out of tree drivers use positional paramater placement. If you > > have four ports then the command using the OOT drivers would be like > > this: > > > > modprobe ixgbe max_vfs=2,2,2,2 > > > > It's an additional feature of the OOT driver that you're able to > > create different numbers of VFs, or none at all, per port. > > > > - Greg > > > ------------------------------------------------------------------------------ Rapidly troubleshoot problems before they affect your business. Most IT organizations don't have a clear picture of how application performance affects their revenue. With AppDynamics, you get 100% visibility into your Java,.NET, & PHP application. Start your 15-day FREE TRIAL of AppDynamics Pro! http://pubads.g.doubleclick.net/gampad/clk?id=84349831&iu=/4140/ostg.clktrk _______________________________________________ E1000-devel mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/e1000-devel To learn more about Intel® Ethernet, visit http://communities.intel.com/community/wired
