[dpdk-dev] irq_to_desc undefined when compiling igb_uio

I agree the kernel is old but it's what we're stuck with at the moment. It appears that at least some thought went into this, as the code in question is actually compatibility code added for older kernel versions such as mine: #if LINUX_VERSION_CODE < KERNEL_VERSION(2, 6, 37) /* Compatability wrap

[dpdk-dev] overcommitting CPUs

The way to handle switch between out of poll mode is to use IRQ coalescing parameters. You want to hold off IRQ until there are a couple packets or a short delay. Going out of poll mode is harder to determine. On Tue, Aug 26, 2014 at 9:59 AM, Zhou, Danny wrote: > > > -Original Message-

[dpdk-dev] irq_to_desc undefined when compiling igb_uio

All, I am building the latest DPDK version from dpdk.org (1.7.0), last commit 8777aa. I am building with essentially the target defconfig_x86_64-native-linuxapp-gcc (renamed), using a gcc 4.4.6/eglibc 2.15 toolchain and a 2.6.34 linux kernel. During compilation, compiling the igb_uio kernel mo

[dpdk-dev] [PATCH v2 0/7] fix build errors

> > This patch series resolves build errors I encountered after disabling some > > parts > > of the dpdk. > > > > - simplified app/*/Makefile, > > - fixed some build error when disabling refcnt, > > - removed dead code (reference to pm_autotest), > > - reworked all the tests from app/test to use

[dpdk-dev] VMXNET 3

I forgot to mention that the performance expectation I gave was for "big packets (>512 bytes size) consumed by multiple VMs" because ESXi is optimized for enterprise kind of workloads. -Original Message- From: Patel, Rashmin N Sent: Tuesday, August 26, 2014 9:30 AM To: 'Alex Markuze';

[dpdk-dev] overcommitting CPUs

> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger > Sent: Wednesday, August 27, 2014 12:39 AM > To: Michael Marchetti > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] overcommitting CPUs > > On Tue, 26 Aug 2014 16:27:14 + > "Michael March

[dpdk-dev] [PATCH 01/11] ixgbe: clean log messages

Hello Jay, On Tue, Aug 26, 2014 at 4:23 PM, Jay Rolette wrote: > Why are you adding newlines to log message strings? Shouldn't that be up > to whatever the messages end up getting routed to? > Actually, I wanted to have consistent log formats in the PMDs so that the log messages displayed at ru

[dpdk-dev] testpmd: not seeing any traffic on 1g (pair) or 10g pair

Hi, Make sure to add iommu=pt inte_iommu=on to kernel parameters if VT-d is enabled. Otherwise, the PF may not work well with igb_uio because of DMAR. Please also refer to the comments in the DPDK getting started guide. In addition, to run the Intel? DPDK with Intel? VT-d, the iommu=pt kernel

[dpdk-dev] overcommitting CPUs

I have a prototype that works on Niantic to enable NIC rx interrupt and allow interrupt and polling mode switch according to real traffic load on the rx queue. It is designed for DPDK power management, and can apply to CPU resource sharing as well. It only works for non-virtualized environment

[dpdk-dev] VMXNET 3

As far as I know, some of the us (developers) are submitting patches for the VMXNET3 driver, which is inside the DPDK package in librte_pmd_vmxnet3. I have used the other version for benchmarking a long back but as it was limited to Linux kernel and I found the librte_pmd_vmxnet3 version more el

[dpdk-dev] overcommitting CPUs

Hi, has there been any consideration to introduce a non-spinning network driver (interrupt based), for the purpose of overcommitting CPUs in a virtualized environment? This would obviously have reduced high-end performance but would allow for increased guest density (sharing of physical CPUs) o

[dpdk-dev] [PATCH 3/3] eal: probe pci devices at rte_eal_init time

Restore call to rte_eal_pci_probe() from commit a155d430119d947d3cb03136ce50924a642dbfe0 so that applications do not have to know that pci devices should be probed. Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c |4 lib/librte_eal/linuxapp/eal/eal.c |4 2 f

[dpdk-dev] [PATCH 2/3] Revert "eal: support link bonding device initialization"

This reverts commit a155d430119d947d3cb03136ce50924a642dbfe0. Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c | 10 + lib/librte_eal/common/eal_common_dev.c | 58 +-- lib/librte_eal/common/eal_common_pci.c |3 -- lib/lib

[dpdk-dev] [PATCH 1/3] bond: move param parsing in dev_configure

Rework bond pmd initialisation so that we don't need to modify eal for this pmd to work. Signed-off-by: David Marchand --- lib/librte_pmd_bond/rte_eth_bond_pmd.c | 33 ++-- lib/librte_pmd_bond/rte_eth_bond_private.h |2 ++ 2 files changed, 28 insertions(+), 7 de

[dpdk-dev] [PATCH 0/3] eal / bonding pmd cleanup

This patchset reworks the bonding pmd so that we don't need to modify the eal for this pmd to work. Basically, the arguments parsed at bond_init are stored in the bond private structure to be used at dev_configure time. If no argument are present, we suppose that the bonding api has been called.

[dpdk-dev] [PATCH RFC 3/3] eal: remove unused flags field

This field is not used anymore, remove it. Signed-off-by: David Marchand --- lib/librte_eal/common/include/rte_eal.h |3 --- 1 file changed, 3 deletions(-) diff --git a/lib/librte_eal/common/include/rte_eal.h b/lib/librte_eal/common/include/rte_eal.h index 8d39cba..f2c4b7d 100644 --- a/lib

[dpdk-dev] [PATCH RFC 2/3] eal: don't call rte_eal_iopl_init unconditionnally

There is no need for ioport access for applications that won't use virtio pmds. Make rte_eal_iopl_init() non-static so that it is called from pmds that need it. Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c |5 + lib/librte_eal/common/include/rte_eal.h | 14

[dpdk-dev] [PATCH RFC 1/3] eal/bsd: fix fd leak

Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c |1 + 1 file changed, 1 insertion(+) diff --git a/lib/librte_eal/bsdapp/eal/eal.c b/lib/librte_eal/bsdapp/eal/eal.c index a296da5..0697b05 100644 --- a/lib/librte_eal/bsdapp/eal/eal.c +++ b/lib/librte_eal/bsdapp/eal/eal.c @@ -

[dpdk-dev] [PATCH RFC 0/3] only call iopl when necessary

This patch series is just a little clean up to remove the unconditionnal call to iopl on linux. Rather than call iopl() at the eal level, let the PMD that needs it call rte_eal_iopl_init(). -- David Marchand David Marchand (3): eal/bsd: fix fd leak eal: don't call rte_eal_iopl_init unconditi

[dpdk-dev] [PATCH] eal: remove unused macros

Clean both linux and bsd implementations from unused macros. Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c | 18 -- lib/librte_eal/bsdapp/eal/eal_pci.c |7 --- lib/librte_eal/bsdapp/eal/eal_timer.c |2 -- lib/librte_eal/linuxa

[dpdk-dev] [PATCH] fix unix permissions for source files

No need for that 'x bit' on source files. Signed-off-by: David Marchand --- 0 files changed mode change 100755 => 100644 examples/l3fwd/main.c mode change 100755 => 100644 examples/qos_meter/main.c mode change 100755 => 100644 examples/qos_meter/main.h mode change 100755 => 100644 examples/q

[dpdk-dev] [PATCH 11/11] eal: set log level from command line

Add a --log-level option to set the default eal log level. Signed-off-by: David Marchand --- lib/librte_eal/bsdapp/eal/eal.c| 43 .../bsdapp/eal/include/eal_internal_cfg.h |1 + lib/librte_eal/linuxapp/eal/eal.c | 43

[dpdk-dev] [PATCH 10/11] e1000: add a message when forcing scatter mode

Signed-off-by: David Marchand --- lib/librte_pmd_e1000/em_rxtx.c |4 lib/librte_pmd_e1000/igb_rxtx.c | 14 ++ 2 files changed, 18 insertions(+) diff --git a/lib/librte_pmd_e1000/em_rxtx.c b/lib/librte_pmd_e1000/em_rxtx.c index fd082f8..19d18e7 100644 --- a/lib/librte_pmd_

[dpdk-dev] [PATCH 09/11] e1000: always log init messages

'init' messages should always be logged and filtered at runtime by rte_log. All the more so as these messages are not in the datapath. Signed-off-by: David Marchand --- lib/librte_pmd_e1000/e1000_logs.h |7 --- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/lib/librte_pmd_

[dpdk-dev] [PATCH 08/11] e1000: clean log messages

Clean log messages: - remove superfluous \n in log macros and add some \n where needed, - remove leading \n in some messages, - split multi lines messages, - introduce PMD_INIT_FUNC_TRACE and replace some PMD_INIT_LOG(DEBUG, "func\n"). Signed-off-by: David Marchand --- lib/librte_pmd_e1000/e1000

[dpdk-dev] [PATCH 07/11] i40e: add log messages when rx bulk mode is not usable

Signed-off-by: David Marchand --- lib/librte_pmd_i40e/i40e_rxtx.c | 29 - 1 file changed, 24 insertions(+), 5 deletions(-) diff --git a/lib/librte_pmd_i40e/i40e_rxtx.c b/lib/librte_pmd_i40e/i40e_rxtx.c index d592ad6..de3701d 100644 --- a/lib/librte_pmd_i40e/i40e_rxt

[dpdk-dev] [PATCH 06/11] i40e: always log init messages

'init' messages should always be logged and filtered at runtime by rte_log. All the more so as these messages are not in the datapath. Signed-off-by: David Marchand --- lib/librte_pmd_i40e/i40e_logs.h |7 --- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/lib/librte_pmd_i4

[dpdk-dev] [PATCH 05/11] i40e: clean log messages

Clean log messages: - remove superfluous \n in log macros and add some \n where needed, - remove leading \n in some messages, - split multi lines messages, - replace some PMD_INIT_LOG(DEBUG, "some_func\n") with PMD_INIT_FUNC_TRACE(). Signed-off-by: David Marchand --- lib/librte_pmd_i40e/i40e/i40

[dpdk-dev] [PATCH 04/11] ixgbe: add log messages when rx bulk mode is not usable

Signed-off-by: David Marchand --- lib/librte_pmd_ixgbe/ixgbe_rxtx.c | 29 - 1 file changed, 24 insertions(+), 5 deletions(-) diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c index e2804bd..7d6c7ea 100644 --- a/lib/librte_pmd_ixgbe/

[dpdk-dev] [PATCH 03/11] ixgbe: add a message when forcing scatter mode

Signed-off-by: David Marchand --- lib/librte_pmd_ixgbe/ixgbe_rxtx.c |8 1 file changed, 8 insertions(+) diff --git a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c b/lib/librte_pmd_ixgbe/ixgbe_rxtx.c index cbec821..e2804bd 100644 --- a/lib/librte_pmd_ixgbe/ixgbe_rxtx.c +++ b/lib/librte_pmd_ixgb

[dpdk-dev] [PATCH 02/11] ixgbe: always log init messages

'init' messages should always be logged and filtered at runtime by rte_log. All the more so as these messages are not in the datapath. Signed-off-by: David Marchand --- lib/librte_pmd_ixgbe/ixgbe_logs.h |7 --- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/lib/librte_pmd_

[dpdk-dev] [PATCH 01/11] ixgbe: clean log messages

Clean log messages: - remove superfluous \n in log macros and add some \n where needed, - remove leading \n in some messages, - split multi lines messages, - replace some PMD_INIT_LOG(DEBUG, "some_func\n") with PMD_INIT_FUNC_TRACE(). Signed-off-by: David Marchand --- lib/librte_pmd_ixgbe/ixgbe_e

[dpdk-dev] [PATCH 00/11] cleanup logs in main PMDs

Here is a patchset that reworks the log macro in e1000, ixgbe and i40e PMDs. The idea behind this is to make it easier to debug some init failures and to be sure of the datapath selected in these PMDs (rx / tx handlers selection). The PMDs changes involve adding more debug messages in the default

[dpdk-dev] irq_to_desc undefined when compiling igb_uio

On Tue, 26 Aug 2014 21:43:15 + "Habibi, Michael" wrote: > I agree the kernel is old but it's what we're stuck with at the moment. It > appears that at least some thought went into this, as the code in question > is actually compatibility code added for older kernel versions such as mine: > >

[dpdk-dev] [PATCH v2 7/7]app/testpmd:test VxLAN Tx checksum offload

Add test cases in testpmd to test VxLAN Tx Checksum offload, which include - IPv4 tunnel and IPv6 tunnel s - outer L3, inner L3 and L4 checksum offload for Tx side. Signed-off-by: jijiangl Acked-by: Helin Zhang Acked-by: Jingjing Wu Acked-by: Jing Chen --- app/test-pmd/config.c |6 +

[dpdk-dev] [PATCH v2 6/7]i40e:support VxLAN Tx checksum offload

Support VxLAN Tx checksum offload, which include - outer L3(IP) checksum offload - inner L3(IP) checksum offload - inner L4(UDP, TCP and SCTP) checksum offload Signed-off-by: jijiangl --- lib/librte_mbuf/rte_mbuf.h |4 +++ lib/librte_pmd_i40e/i40e_rxtx.c | 58 ++

[dpdk-dev] [PATCH v2 5/7]]app/testpmd:test VxLAN packet filter API

Add tunnel_filter command in testpmd to test VxLAN packet filter API. Signed-off-by: jijiangl Acked-by: Helin Zhang Acked-by: Jingjing Wu Acked-by: Jing Chen --- app/test-pmd/cmdline.c | 149 +++- 1 files changed, 148 insertions(+), 1 deletions(-)

[dpdk-dev] [PATCH v2 4/7]i40e:implement API of VxLAN packet filter in librte_pmd_i40e

The implementation of VxLAN packet filter in librte_pmd_i40e. Signed-off-by: jijiangl Acked-by: Helin Zhang Acked-by: Jingjing Wu Acked-by: Jing Chen --- lib/librte_pmd_i40e/i40e_ethdev.c | 112 + 1 files changed, 112 insertions(+), 0 deletions(-) diff -

[dpdk-dev] [PATCH v2 3/7]rte_ether:add API of VxLAN packet filter in librte_ether

VxLAN packet filters, which use MAC and VLAN to point to a queue. The filter types supported are listed below: 1. Inner MAC and Inner VLAN ID 2. Inner MAC address, inner VLAN ID and tenant ID. 3. Inner MAC and tenant ID 4. Inner MAC address 5. Outer MAC address, tenant ID and inne

[dpdk-dev] [PATCH v2 2/7]]app/test-pmd:test VxLAN packet identification

Add commands to test VxLAN packet identification, which include - use commands to add/delete VxLAN UDP port. - use rxonly mode to check VxLAN packet. Signed-off-by: jijiangl Acked-by: Helin Zhang Acked-by: Jingjing Wu Acked-by: Jing Chen --- app/test-pmd/cmdline.c| 78 +++

[dpdk-dev] [PATCH v2 1/7]i40e:support VxLAN packet identification

VxLAN UDP port configuration on i40e, which include - VxLAN UDP port initialization - add APIs to configure VxLAN UDP port Signed-off-by: jijiangl Acked-by: Helin Zhang Acked-by: Jingjing Wu Acked-by: Jing Chen --- lib/librte_ether/rte_ethdev.c | 63 lib/librte_ether/

[dpdk-dev] [PATCH v2 0/7]]Support VxLAN on Fortville

The patch set supports VxLAN on Fortville based on current mbuf structure. When Bruce's new mbuf structure is done, there will be minor changes later. It includes: - Support VxLAN packet identification by configuring tunneling UDP port. - Support VxLAN packet filters. It uses MAC and VLAN to p

[dpdk-dev] [PATCHv3] librte_acl make it build/work for 'default' target

On Mon, Aug 25, 2014 at 04:30:05PM +, Ananyev, Konstantin wrote: > Hi Neil, > > > -Original Message- > > From: Neil Horman [mailto:nhorman at tuxdriver.com] > > Sent: Thursday, August 21, 2014 9:15 PM > > To: dev at dpdk.org > > Cc: Ananyev, Konstantin; thomas.monjalon at 6wind.com; Ne

[dpdk-dev] i212 controller with 0 queues per port

Hi. I've got a SCB-8973 machine equipped with i212 network controller (6 ports). My dpdk application (see the code snippet below) returns zeros for both dev_info.max_tx_queues and dev_info.max_rx_queues Have anybody got an experience of addressing this problem? Regards, Sergey for (portid

[dpdk-dev] VMXNET 3

Hi I'm looking for reasonable DPDK based solution in fully virtualised VMware environment. >From what I've seen there are several flavours of VMXNET 3 driver for dpdk not all of them seem to be alive - user map -last updated on may and doesn't compile on DPDK 1.7. So to my question what is the sta

[dpdk-dev] Fwd: ixgbe network card has dev_info.max_rx_queues == 0

Yep, I understand now. The manual [1] says that intel 827574 controller (controlled by e1000 driver) has only 2 ports per port. Thanks for the help. Sergey [1] - http://www.intel.com/content/www/us/en/ethernet-controllers/i211-ethernet-controller-datasheet.html PS I actually have another machine

[dpdk-dev] [PATCH 1/2] igb_uio: fix compability on old kernel

> >> This is what we came up with. It works for us. In our kernel headers' >> linux/pci.h, pci_num_vf is enclosed within "#ifdef >>CONFIG_PCI_IOV/#endif"; >> pci_intx_mask_supported and pci_check_and_mask_intx are enclosed within >> "#ifdef HAVE_PCI_SET_MWI/#endif". >> >> What do you think? > >Ma

[dpdk-dev] [RFC] lib/librte_vhost: qemu vhost-user support into DPDK vhost library

Hi all: We are implementing qemu official vhost-user interface into DPDK vhost library, so there would be two coexisting implementations for user space vhost backend. Pro and cons in my mind: Existing solution: Pros: works with qemu version before 2.1; Cons: depends on eventfd proxy kernel modu

[dpdk-dev] overcommitting CPUs

On Tue, 26 Aug 2014 16:27:14 + "Michael Marchetti" wrote: > Hi, has there been any consideration to introduce a non-spinning network > driver (interrupt based), for the purpose of overcommitting CPUs in a > virtualized environment? This would obviously have reduced high-end > performance

[dpdk-dev] [RFC 06/10] virtio: use software vlan stripping

On Tue, 26 Aug 2014 08:37:11 + "Ouyang, Changchun" wrote: > Hi Stephen, > > Would you please describe the use scenario for the front end rx vlan strip > and tx vlan insertion? > In our current implementation, backend will strip vlan tag for RX, and insert > vlan tag for TX. > > Thanks >

[dpdk-dev] [PATCH 01/11] ixgbe: clean log messages

Why are you adding newlines to log message strings? Shouldn't that be up to whatever the messages end up getting routed to? Jay On Tue, Aug 26, 2014 at 9:09 AM, David Marchand wrote: > Clean log messages: > - remove superfluous \n in log macros and add some \n where needed, > - remove leading

[dpdk-dev] [PATCH] vmxnet3: initialize receive mode correctly

2014-08-25 17:33, Stephen Hemminger: > Thomas Monjalon wrote: > > 2014-07-25 10:50, Stephen Hemminger: > > > - vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_UCAST | VMXNET3_RXM_ALL_MULTI, > > > 1); > > > + vmxnet3_dev_set_rxmode(hw, VMXNET3_RXM_UCAST | VMXNET3_RXM_BCAST, 1); > > > > It's also removing

[dpdk-dev] [RFC 08/10] virtio: remove redundant vq_alignment

Acked by: Changchun Ouyang > -Original Message- > From: Stephen Hemminger [mailto:stephen at networkplumber.org] > Sent: Tuesday, August 26, 2014 10:08 AM > To: Ouyang, Changchun > Cc: dev at dpdk.org; Stephen Hemminger > Subject: [RFC 08/10] virtio: remove redundant vq_alignment > > Sin

[dpdk-dev] [RFC 06/10] virtio: use software vlan stripping

Hi Stephen, Would you please describe the use scenario for the front end rx vlan strip and tx vlan insertion? In our current implementation, backend will strip vlan tag for RX, and insert vlan tag for TX. Thanks Changchun > -Original Message- > From: Stephen Hemminger [mailto:stephen a

[dpdk-dev] [RFC 01/10] virtio: rearrange resource initialization

Acked-by: Changchun Ouyang > -Original Message- > From: Stephen Hemminger [mailto:stephen at networkplumber.org] > Sent: Tuesday, August 26, 2014 10:08 AM > To: Ouyang, Changchun > Cc: dev at dpdk.org > Subject: [RFC 01/10] virtio: rearrange resource initialization > > For clarity make t

[dpdk-dev] [RFC 10/10] virtio: add support for promiscious and multicast

This patch is very similar with my previous patch: [PATCH 4/5] virtio: New API to enable/disable multicast and promisc mode So suggest applying only one of both. Thanks Changchun > -Original Message- > From: Stephen Hemminger [mailto:stephen at networkplumber.org] > Sent: Tuesday, Augu

[dpdk-dev] [RFC 07/10] virtio: remove unnecessary adapter structure

Acked-by: Changchun Ouyang > -Original Message- > From: Stephen Hemminger [mailto:stephen at networkplumber.org] > Sent: Tuesday, August 26, 2014 10:08 AM > To: Ouyang, Changchun > Cc: dev at dpdk.org > Subject: [RFC 07/10] virtio: remove unnecessary adapter structure > > Cleanup virtio

[dpdk-dev] [RFT] vmxnet3 testing needed

Hi Stephen, We have tested intel VMXNet3 driver before. If your patch is based on Intel DPDK VMXNet3 driver, we can try it in my platform. Regards Waterman >-Original Message- >From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Stephen Hemminger >Sent: Tuesday, August 26, 2014

[dpdk-dev] [PATCH 4/5] virtio: New API to enable/disable multicast and promisc mode

Hi Stephen, My response below. Thanks Changchun > -Original Message- > From: Stephen Hemminger [mailto:stephen at networkplumber.org] > Sent: Tuesday, August 26, 2014 8:13 AM > To: Ouyang, Changchun > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH 4/5] virtio: New API to enable/