[dpdk-dev] DPDK doesn't work with iommu=pt

> Subject: Re: [dpdk-dev] DPDK doesn't work with iommu=pt > > > > On Mon, Sep 29, 2014 at 2:53 AM, Hiroshi Shimamoto ct.jp.nec.com> wrote: > > Hi, > > > >> Subject: Re: [dpdk-dev] DPDK doesn't work with iommu=pt > >> > >> iommu=pt effectively disables iommu for the kernel and iommu is > >>

[dpdk-dev] [PATCH v5 01/11] lib/librte_vhost: move src files in vhost example to vhost lib directory

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com] > Sent: Tuesday, September 30, 2014 3:42 AM > To: Xie, Huawei > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH v5 01/11] lib/librte_vhost: move src files in > vhost > example to vhost lib directory

[dpdk-dev] [PATCH 1/7] Split atomic operations to architecture specific

Bruce and Neil, Thanks for your comments! Actually, the compiler hides the difference with different architecture. I'll submit another patch to correct this! Best Regards! -- Chao Zhu (??) Research Staff Member Cloud Infrastructure and Technology Group IBM China

[dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf change

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com] > Sent: Tuesday, September 30, 2014 3:44 AM > To: Xie, Huawei > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf > change > > > There is no rte_pktmbuf

[dpdk-dev] [PATCH] llib/ibrte_net: workaround to avoid macro conflict

Macros such as IPPROTO_TCP, IPPROTO_UDP are already defined in . If user's application includes and rte_ip.h at the same time, there will be conflict error. This patch uses the way "#ifndef #endif" to avoid the conflict. Signed-off-by: Jingjing Wu --- lib/librte_net/rte_ip.h | 5 + 1

[dpdk-dev] Hi all, does Amazon VMs supported DPDK or not?

Hi Shawn, Thanks a lot. Any more information of the c3.8xlarges? eg some public detail spec of this c3.8xlarges VM; Is it an Amazon EC2 VM in following link? http://aws.amazon.com/cn/blogs/aws/a-generation-of-ec2-instances-for-compute-intensive-workloads/ http://aws.amazon.com/ec2/

[dpdk-dev] VMDq Sample Application on Virtual Machines

> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of ANKIT BATRA > Sent: Monday, September 29, 2014 8:27 PM > To: dev at dpdk.org > Subject: [dpdk-dev] VMDq Sample Application on Virtual Machines > > Hi, > > I am running VMDq sample application on host

[dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf change

2014-09-30 02:41, Xie, Huawei: > I would rework the patch according to your comment. > I don't get clear about this comment. Do you mean that recreate the patch set > based on the example that already has this mbuf change? Yes > Some of the background you might not know: > I fully understand

[dpdk-dev] rc1 / call for review

2014-09-29 14:33, Matthew Hall: > On Mon, Sep 29, 2014 at 10:23:58PM +0200, Thomas Monjalon wrote: > > - mbuf rework > > - logs rework > > - some eal cleanups > > I was curious, did we happen to know if any of these three changes affected > the external API's much? > > It would help

[dpdk-dev] [PATCH 0/3] fix of lsc interrupt in i40e PF

Tested-by: Min Cao This patch has been verified on FC20 with Eagle Fountain: 4*10G . The i40e base driver update patch works well on FC20 with basic function. The test environment detail information as the following: HOST environment: CPU: Intel(R) Xeon(R) CPU E5-2680 v2 @ 2.80GHz OS: Linux

[dpdk-dev] [PATCH v2] bond: Add mode 4 support.

This patch adds support mode 4 of link bonding. It depend on Delcan Doherty patches v3 and rte alarms patch v2 or above. New version handles race issues with setting/cancelin callbacks, fixes promiscus mode setting in mode 4 and some other minor errors in mode 4 implementation. Signed-off-by:

[dpdk-dev] [PATCH v3 1/7] ethdev: add more annotations

Add more annotations about packet classification type. v3 changes: * Remove renamings of RSS 'SHIFT's. * Add more annotations for RSS 'SHIFT's. Signed-off-by: Helin Zhang Acked-by: Jingjing Wu --- lib/librte_ether/rte_ethdev.h | 5 - 1 file changed, 4 insertions(+), 1 deletion(-) diff

[dpdk-dev] [PATCH v3 2/7] ethdev: add interfaces and relevant for filter control

To support flexible filter control, 'rte_eth_dev_filter_ctrl()' and 'rte_eth_dev_filter_supported()' are added. In addition, filter types and operations are defined in a newly added header file. v3 changes: * Interfaces to be added have been re-designed. * Header file has been renamed.

[dpdk-dev] [PATCH v3 4/7] i40e: add hash filter control implementation

Hash filter control has been implemented for i40e. It includes getting/setting - hash function type - symmetric hash enable per pctype (packet classification type) - symmetric hash enable per port - filter swap configurations v3 changes: * Remove public header file specific for i40e. * Use the

[dpdk-dev] [PATCH v3 5/7] i40e: add hardware initialization

As global registers will be reset only after a whole chip reset, those registers might not be in an initial state after each launching a physical port. The hardware initialization is added to put specific global registers into an initial state. v3 changes: * Renamed hardware initialization

[dpdk-dev] [PATCH v3 7/7] app/testpmd: add commands to support hash filter control

To demonstrate the hash filter control, commands are added. They are - get_sym_hash_ena_per_port - set_sym_hash_ena_per_port - get_sym_hash_ena_per_pctype - set_sym_hash_ena_per_pctype - get_filter_swap - set_filter_swap - get_hash_function - set_hash_function v3 changes: * Renamed the command

[dpdk-dev] [PATCH v3 3/7] ethdev: add structures and enum for hash filter control

Structures and enum are added in rte_eth_ctrl.h to support hash filter control. v3 changes: * Common structures are added in rte_eth_ctrl.h to support hash filter control. * Hash filter info types and hash function types are added in rte_eth_ctrl.h to support filter control. Signed-off-by:

[dpdk-dev] [PATCH v3 6/7] i40e: Use constant random hash keys

To be simpler, and remove the race condition, it uses prepared constant random hash keys to replace runtime generating the hash keys. v3 changes: * Use prepared random hash keys. Signed-off-by: Helin Zhang Acked-by: Jingjing Wu --- lib/librte_pmd_i40e/i40e_ethdev.c | 12 ++-- 1 file

[dpdk-dev] [PATCH] Fix for LRU corrupted returns

Sorry about the delay. The number 32 is not really a CACHE_LINE_SIZE but since __builtin_clz returns the number of leading 0's before the most significant set bit in a 32 bit number (entry_size is uint32_t), I subtract that number from 32 to get the number of trailing bits after the most

[dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf change

> -Original Message- > From: Thomas Monjalon [mailto:thomas.monjalon at 6wind.com] > Sent: Tuesday, September 30, 2014 12:46 PM > To: Xie, Huawei > Cc: dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf > change > > 2014-09-30 02:41, Xie,

[dpdk-dev] [PATCH v2] distributor_app: new sample app

Hi Konstantin, Any comments on below Pablos comment? If so please provide. Thanks, Reshma -Original Message- From: De Lara Guarch, Pablo Sent: Monday, September 29, 2014 2:35 PM To: Ananyev, Konstantin; Pattan, Reshma; dev at dpdk.org Subject: RE: [dpdk-dev] [PATCH v2]

[dpdk-dev] [PATCH v5 05/11] lib/librte_vhost: merge Oliver's mbuf change

2014-09-30 06:43, Xie, Huawei: > > 2014-09-30 02:41, Xie, Huawei: > > > I would rework the patch according to your comment. > > > I don't get clear about this comment. Do you mean that recreate the patch > > > set > > > based on the example that already has this mbuf change? > > > > Yes > > > >

[dpdk-dev] [PATCH 1/4 v3] compat: Add infrastructure to support symbol versioning

On Mon, Sep 29, 2014 at 11:44:03AM -0400, Neil Horman wrote: > Add initial pass header files to support symbol versioning. > > --- > Change notes > v2) > * Fixed ifdef in rte_compat.h to test for RTE_BUILD_SHARED_LIB instead of the > non-existant RTE_SYMBOL_VERSIONING > > * Fixed VERSION_SYMBOL

[dpdk-dev] vmxnet3 pmd dev restart

Hi I am using DPDKv1.7.0 running on Vmware Esxi 5.1 and am trying to reset the port which uses pmd_vmnet3 library functions from below function calls. rte_eth_dev_stop rte_eth_dev_start Doing this, i face panic while rte_free(ring->buf_info) in Vmxnet3_cmd_ring_release(). I have gone through

[dpdk-dev] [PATCH v2] distributor_app: new sample app

> -Original Message- > From: Pattan, Reshma > Sent: Tuesday, September 30, 2014 9:03 AM > To: Ananyev, Konstantin; De Lara Guarch, Pablo; dev at dpdk.org > Subject: RE: [dpdk-dev] [PATCH v2] distributor_app: new sample app > > Hi Konstantin, > > Any comments on below Pablos comment? If

[dpdk-dev] Building current 1.8.1-rc1 with clang

On Mon, Sep 29, 2014 at 09:50:34PM +, Wiles, Roger Keith wrote: > I just pulled the current repo and stated a build with ?make install > T=x86_64-native-linuxapp-clang? which produced the following error. I do not > think I am allowed to modify this file, correct? If that is the case then >

[dpdk-dev] [PATCH] ixgbe: Fix clang compilation issue

Issue reported by Keith Wiles. Clang fails with an error about a variable being used uninitialized: CC ixgbe_rxtx_vec.o /home/keithw/projects/dpdk-code/org-dpdk/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c:67:30: error: variable 'dma_addr0' is uninitialized when used here

[dpdk-dev] [PATCH 1/2] librte_pmd_null: Add null PMD

From: Tetsuya Mukawa 'null PMD' is a driver of the virtual device particulary designed to measure performance of DPDK PMDs. When an application call rx, null PMD just allocates mbufs and returns those. Also tx, the PMD just frees mbufs. The PMD has following options. - size:

[dpdk-dev] [PATCH 2/2] librte_pmd_null: Enable librte_pmd_null

From: Tetsuya Mukawa Signed-off-by: Tetsuya Mukawa --- mk/rte.app.mk | 4 1 file changed, 4 insertions(+) diff --git a/mk/rte.app.mk b/mk/rte.app.mk index 34dff2a..f059290 100644 --- a/mk/rte.app.mk +++ b/mk/rte.app.mk @@ -179,6 +179,10 @@ LDLIBS += -lrte_pmd_xenvirt

[dpdk-dev] [PATCH v4 2/8] bond: removing switch statement from rx burst method

Signed-off-by: Declan Doherty --- lib/librte_pmd_bond/rte_eth_bond_pmd.c | 62 +++--- 1 file changed, 35 insertions(+), 27 deletions(-) diff --git a/lib/librte_pmd_bond/rte_eth_bond_pmd.c b/lib/librte_pmd_bond/rte_eth_bond_pmd.c index aca2dcf..348e28f 100644 ---

[dpdk-dev] [PATCH v4 1/8] bond: link status interrupt support

Adding support for lsc interrupt from bonded device to link bonding library with supporting unit tests in the test application. Signed-off-by: Declan Doherty --- app/test/test_link_bonding.c | 213 +++-- lib/librte_pmd_bond/rte_eth_bond_api.c | 4 +

[dpdk-dev] [PATCH v4 6/8] testpmd: adding parameter to reconfig method to set socket_id when adding new port to portlist

Signed-off-by: Declan Doherty --- app/test-pmd/cmdline.c | 2 +- app/test-pmd/testpmd.c | 3 ++- app/test-pmd/testpmd.h | 2 +- 3 files changed, 4 insertions(+), 3 deletions(-) diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c index 225f669..15ca493 100644 ---

[dpdk-dev] [PATCH v4 4/8] bond: free mbufs if transmission fails in bonding tx_burst functions

Signed-off-by: Declan Doherty --- app/test/test_link_bonding.c | 393 - app/test/virtual_pmd.c | 80 +-- app/test/virtual_pmd.h | 7 + lib/librte_pmd_bond/rte_eth_bond_pmd.c | 83 +-- 4 files changed, 525

[dpdk-dev] [PATCH v4 0/8] link bonding

v4: - Rebased to account for changes in master. - Fix for rte_eth_bond_slaves_get() introduced in v3 patch set - Addressed issue around disabling/enabling link status polling around adding/ removing slaves devices. v3 : - Typo fix for the bond free mbufs patch. - Rebased to account for changes

[dpdk-dev] [PATCH v4 5/8] test app: adding support for generating variable sized packet bursts

Signed-off-by: Declan Doherty --- app/test/packet_burst_generator.c | 25 - app/test/packet_burst_generator.h | 6 +- app/test/test_link_bonding.c | 14 +- 3 files changed, 22 insertions(+), 23 deletions(-) diff --git

[dpdk-dev] [PATCH v4 7/8] bond: lsc polling support

Signed-off-by: Declan Doherty --- app/test-pmd/cmdline.c | 63 + app/test/test.h| 7 +- app/test/test_link_bonding.c | 258 --- app/test/virtual_pmd.c | 17 +- app/test/virtual_pmd.h

[dpdk-dev] [PATCH v4 3/8] bond: fix naming inconsistency in tx_burst_round_robin

Signed-off-by: Declan Doherty --- lib/librte_pmd_bond/rte_eth_bond_pmd.c | 10 +- 1 file changed, 5 insertions(+), 5 deletions(-) diff --git a/lib/librte_pmd_bond/rte_eth_bond_pmd.c b/lib/librte_pmd_bond/rte_eth_bond_pmd.c index 348e28f..66f1650 100644 ---

[dpdk-dev] [PATCH v4 8/8] bond: unit test test macro refactor

Signed-off-by: Declan Doherty --- app/test/test_link_bonding.c | 2574 +- 1 file changed, 1036 insertions(+), 1538 deletions(-) diff --git a/app/test/test_link_bonding.c b/app/test/test_link_bonding.c index c32b685..c4fcaf7 100644 ---

[dpdk-dev] [PATCH v3] distributor_app: new sample app

From: Reshma Pattan A new sample app that shows the usage of the distributor library. This app works as follows: * An RX thread runs which pulls packets from each ethernet port in turn and passes those packets to worker using a distributor component. * The workers

[dpdk-dev] [PATCH v2] Change alarm cancel function to thread-safe:

Pawe? > -Original Message- > From: Richardson, Bruce > Sent: Monday, September 29, 2014 12:33 > To: Wodkowski, PawelX > Cc: Ananyev, Konstantin; Neil Horman; dev at dpdk.org > Subject: Re: [dpdk-dev] [PATCH v2] Change alarm cancel function to thread- > safe: > > On Mon, Sep 29, 2014 at

[dpdk-dev] [memnic PATCH v2 0/7] MEMNIC PMD performance improvement

From: Hiroshi Shimamoto This patchset improves MEMNIC PMD performance. The first patch introduces a new benchmark test run in guest, and will be used to evaluate the following patch effects. This patchset improves the throughput results of memnic-tester. Using Xeon

[dpdk-dev] [memnic PATCH v2 1/7] guest: memnic-tester: PMD benchmark in guest

From: Hiroshi Shimamoto Introduce memnic-tester which benchmarks MEMNIC PMD performance in guest. It starts with two threads, one thread produces and consumes packets, other thread receives packets and directly transmits the received packets. This evaluates MEMNIC PMD

[dpdk-dev] [memnic PATCH v2 2/7] pmd: remove needless assignment

From: Hiroshi Shimamoto Because these assignment are done in rte_pktmbuf_alloc(), get rid of them. Signed-off-by: Hiroshi Shimamoto Reviewed-by: Hayato Momma --- pmd/pmd_memnic.c | 2 -- 1 file changed, 2 deletions(-) diff --git a/pmd/pmd_memnic.c

[dpdk-dev] [memnic PATCH v2 3/7] pmd: use helper macros

From: Hiroshi Shimamoto Do not touch pktmbuf directly. Intead of direct access, use rte_pktmbuf_pkt_len() and rte_pktmbuf_data_len() to access the property. Signed-off-by: Hiroshi Shimamoto Reviewed-by: Hayato Momma --- pmd/pmd_memnic.c | 6 +++--- 1 file changed,

[dpdk-dev] [PATCH v2] bond: Add mode 4 support.

Fixed patch version sent. Pawel

[dpdk-dev] [memnic PATCH v2 4/7] pmd: use compiler barrier

From: Hiroshi Shimamoto x86 can keep store ordering with standard operations. Using memory barrier is much expensive in main packet processing loop. Removing this improves xmit/recv packet performance. We can see performance improvements with memnic-tester. Using

[dpdk-dev] [memnic PATCH v2 5/7] pmd: packet receiving optimization with prefetch

From: Hiroshi Shimamoto Prefetch the next packet area to reduce memory stall cycles. Prefetching the next packet area could hide memory stall, because the next area will be accessed just after processing the current receive operations. We can see performance

[dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit

From: Hiroshi Shimamoto To reduce instruction cache miss, add branch condition hints into recv/xmit functions. This improves a bit performance. We can see performance improvements with memnic-tester. Using Xeon E5-2697 v2 @ 2.70GHz, 4 vCPU. size | before | after

[dpdk-dev] [memnic PATCH v2 7/7] pmd: burst mbuf freeing in xmit

From: Hiroshi Shimamoto In rte_pktmbuf_free(), there might be cache miss/memory stall issue. In small packet case, it could harm the performance. >From the result of memnic-tester, in less than 1024 frame size the performance could be improved. Using Xeon E5-2697 v2

[dpdk-dev] [PATCH 2/2] librte_pmd_null: Enable librte_pmd_null

On Tue, Sep 30, 2014 at 06:56:10PM +0900, mukawa at igel.co.jp wrote: > From: Tetsuya Mukawa > > Signed-off-by: Tetsuya Mukawa > --- > mk/rte.app.mk | 4 > 1 file changed, 4 insertions(+) > > diff --git a/mk/rte.app.mk b/mk/rte.app.mk > index 34dff2a..f059290 100644 > --- a/mk/rte.app.mk

[dpdk-dev] [PATCH v3] distributor_app: new sample app

On Tue, Sep 30, 2014 at 11:39:37AM +0100, reshmapa wrote: > From: Reshma Pattan > > A new sample app that shows the usage of the distributor library. This > app works as follows: > > * An RX thread runs which pulls packets from each ethernet port in turn > and passes those packets to worker

[dpdk-dev] [PATCH v2] Change alarm cancel function to thread-safe:

> -Original Message- > Pawe? > > > On Mon, Sep 29, 2014 at 10:11:38AM +, Wodkowski, PawelX wrote: > > > > > > > > > > Image how you will be damned by someone that not even notice you > > change > > > > > and he Is managing some kind of resource based on returned number of > > > > >

[dpdk-dev] [PATCH] Fix for LRU corrupted returns

On Tue, Sep 30, 2014 at 06:26:23AM +, Saha, Avik (AWS) wrote: > Sorry about the delay. The number 32 is not really a CACHE_LINE_SIZE but > since __builtin_clz returns the number of leading 0's before the most > significant set bit in a 32 bit number (entry_size is uint32_t), I subtract >

[dpdk-dev] [PATCH v2] Change alarm cancel function to thread-safe:

On Tue, Sep 30, 2014 at 12:30:08PM +, Ananyev, Konstantin wrote: > > > > -Original Message- > > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Wodkowski, PawelX > > Sent: Tuesday, September 30, 2014 1:05 PM > > To: Wodkowski, PawelX; Richardson, Bruce > > Cc: dev at dpdk.org

[dpdk-dev] [PATCH v2 14/18] ixgbe: Remove unnecessary delay

On Mon, Sep 29, 2014 at 03:16:22PM +0800, Ouyang Changchun wrote: > This patch removes unnecessary delay when setting up physical link > and negotiating in IXGBE share code. > Why was this there in the first place then? Was there some older hardware that required it? If so, is there a need to

[dpdk-dev] [PATCH 1/2] librte_pmd_null: Add null PMD

2014-09-30 18:56, mukawa at igel.co.jp: > --- /dev/null > +++ b/lib/librte_pmd_null/Makefile > @@ -0,0 +1,58 @@ > +# BSD LICENSE > +# > +# Copyright (C) 2014 Nippon Telegraph and Telephone Corporation. > +# All rights reserved. > +# > +# Redistribution and use in source and binary forms,

[dpdk-dev] [PATCH] llib/ibrte_net: workaround to avoid macro conflict

On Tue, Sep 30, 2014 at 10:49:08AM +0800, Jingjing Wu wrote: > Macros such as IPPROTO_TCP, IPPROTO_UDP are already defined in . > If user's application includes and rte_ip.h at the same time, > there will be conflict error. > > This patch uses the way "#ifndef #endif" to avoid the conflict. >

[dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit

The patch is ok. For the commit message, is it better "to reduce branch mispredication"? > -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Hiroshi Shimamoto > Sent: Tuesday, September 30, 2014 7:15 PM > To: dev at dpdk.org > Cc: Hayato Momma > Subject:

[dpdk-dev] [PATCH] ixgbe: Fix clang compilation issue

Acked-by: Keith Wiles On Sep 30, 2014, at 4:40 AM, Bruce Richardson wrote: > Issue reported by Keith Wiles. > Clang fails with an error about a variable being used uninitialized: > > CC ixgbe_rxtx_vec.o > /home/keithw/projects/dpdk-code/org-dpdk/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c:67:30:

[dpdk-dev] [memnic PATCH v2 0/7] MEMNIC PMD performance improvement

On Tue, Sep 30, 2014 at 11:10:45AM +, Hiroshi Shimamoto wrote: > From: Hiroshi Shimamoto > > This patchset improves MEMNIC PMD performance. > > The first patch introduces a new benchmark test run in guest, > and will be used to evaluate the following patch effects. > > This patchset

[dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit

On Tue, Sep 30, 2014 at 11:14:40AM +, Hiroshi Shimamoto wrote: > From: Hiroshi Shimamoto > > To reduce instruction cache miss, add branch condition hints into > recv/xmit functions. This improves a bit performance. > > We can see performance improvements with memnic-tester. > Using Xeon

[dpdk-dev] [memnic PATCH v2 0/7] MEMNIC PMD performance improvement

On 9/30/2014 7:29 AM, Neil Horman wrote: > On Tue, Sep 30, 2014 at 11:10:45AM +, Hiroshi Shimamoto wrote: >> From: Hiroshi Shimamoto >> >> This patchset improves MEMNIC PMD performance. >> >> The first patch introduces a new benchmark test run in guest, >> and will be used to evaluate the

[dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit

> -Original Message- > From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Hiroshi Shimamoto > Sent: Tuesday, September 30, 2014 7:15 PM > To: dev at dpdk.org > Cc: Hayato Momma > Subject: [dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit > > From: Hiroshi Shimamoto

[dpdk-dev] GSO support by PMD drivers

Hello Alex, Vadim, On 09/28/2014 09:19 AM, Alex Markuze wrote: > LSO/TSO support is an important feature, I'm surprised its not > supported in DPDK. > I personally would like to see these patches. > > On Fri, Sep 26, 2014 at 1:23 PM, Vadim Suraev > wrote: >> Hi, all, >> I found ixgbe in couple

[dpdk-dev] [memnic PATCH v2 6/7] pmd: add branch hint in recv/xmit

2014-09-30 14:38, Xie, Huawei: > > - if (++next >= MEMNIC_NR_PACKET) > > + if (unlikely(++next >= MEMNIC_NR_PACKET)) > > On IA, compiler can use add, cmp and cmov to avoid branch. > But If MEMNIC_NR_PACKET is always power of 2, > it is better just next = (next + 1) &

[dpdk-dev] GSO support by PMD drivers

Than you, Oliver On Sep 30, 2014 5:42 PM, "Olivier MATZ" wrote: > Hello Alex, Vadim, > > On 09/28/2014 09:19 AM, Alex Markuze wrote: > >> LSO/TSO support is an important feature, I'm surprised its not >> supported in DPDK. >> I personally would like to see these patches. >> >> On Fri, Sep 26,

[dpdk-dev] [PATCH 1/4 v4] compat: Add infrastructure to support symbol versioning

Add initial pass header files to support symbol versioning. --- Change notes v2) * Fixed ifdef in rte_compat.h to test for RTE_BUILD_SHARED_LIB instead of the non-existant RTE_SYMBOL_VERSIONING * Fixed VERSION_SYMBOL macro to add the needed extra @ to make versioning work properly *

[dpdk-dev] [memnic PATCH v2 0/7] MEMNIC PMD performance improvement

> This patchset improves MEMNIC PMD performance. > > Hiroshi Shimamoto (7): > guest: memnic-tester: PMD benchmark in guest > pmd: remove needless assignment > pmd: use helper macros > pmd: use compiler barrier > pmd: packet receiving optimization with prefetch > pmd: add branch hint

[dpdk-dev] [PATCH RFC] mbuf: Adjust TX flags to start at bit 32

This patch takes the existing TX flags defined for the mbuf and shifts each uniquely defined one left so that additional RX flags can be defined without having RX and TX flags mixed together. Signed-off-by: Bruce Richardson --- lib/librte_mbuf/rte_mbuf.h | 26 +- 1 file

[dpdk-dev] [PATCH RFC] mbuf: Adjust TX flags to start at bit 32

On Tue, Sep 30, 2014 at 04:26:02PM +0100, Bruce Richardson wrote: > This patch takes the existing TX flags defined for the mbuf and shifts > each uniquely defined one left so that additional RX flags can be > defined without having RX and TX flags mixed together. > > Signed-off-by: Bruce

[dpdk-dev] [PATCH RFC] mbuf: Adjust TX flags to start at bit 32

Hi Bruce, I like the idea of the split, which should make it easier to do the testing of those bits. One comment below. On Sep 30, 2014, at 10:33 AM, Bruce Richardson wrote: > On Tue, Sep 30, 2014 at 04:26:02PM +0100, Bruce Richardson wrote: >> This patch takes the existing TX flags defined

[dpdk-dev] [PATCH RFC] mbuf: Adjust TX flags to start at bit 32

On Tue, Sep 30, 2014 at 05:00:04PM +0100, Wiles, Roger Keith wrote: > Hi Bruce, > > I like the idea of the split, which should make it easier to do the testing > of those bits. > One comment below. > > On Sep 30, 2014, at 10:33 AM, Bruce Richardson intel.com> wrote: > > > On Tue, Sep 30, 2014

[dpdk-dev] [PATCH RFC] mbuf: Adjust TX flags to start at bit 32

On Tue, Sep 30, 2014 at 04:26:02PM +0100, Bruce Richardson wrote: > This patch takes the existing TX flags defined for the mbuf and shifts > each uniquely defined one left so that additional RX flags can be > defined without having RX and TX flags mixed together. > > Signed-off-by: Bruce

[dpdk-dev] [PATCH] ixgbe: Fix clang compilation issue

> > Issue reported by Keith Wiles. > > Clang fails with an error about a variable being used uninitialized: > > > > CC ixgbe_rxtx_vec.o > > /home/keithw/projects/dpdk-code/org-dpdk/lib/librte_pmd_ixgbe/ixgbe_rxtx_vec.c:67:30: > > error: variable 'dma_addr0' is uninitialized > > when used

[dpdk-dev] [PATCH] kni:fix build on Ubuntu 12.04.5 with current HWE

2014-09-30 13:10, Daniel Mrzyglod: > Recent Ubuntu 12.04.5 LTS is shipped with 3.13.0-36.63 as the only supported > kernel. > > Patch a09b359daca3d8af43dc22a57b34cf317f958236 describe the problem. > > Signed-off-by: Daniel Mrzyglod > #if ( LINUX_VERSION_CODE < KERNEL_VERSION(3,14,0) ) > #if

[dpdk-dev] Hi all, does Amazon VMs supported DPDK or not?

I have only experimented with C3.8xlarge instances with DPDK. You have to attach at least 2 ENIs to the instances since one of them would be taken over by DPDK. Based on the size, you would be able to attach 8, 16 or 32 ENIs to the instances (these will be visible as ethn devices of ifconfig)

[dpdk-dev] [PATCH] Fix for LRU corrupted returns

On Tue, Sep 30, 2014 at 06:14:46PM +, Saha, Avik (AWS) wrote: > I have to point out that I am commenting out the the power_of_2 check on > entry_size. I am not sure if this is the right way but I don't know why this > soft assumption is important (since I cannot find the power of 2