On Sun, Oct 22, 2017 at 01:00:22AM -0700, Yongseok Koh wrote:
>[...]
> --- a/drivers/net/mlx5/mlx5_rxtx.h
> +++ b/drivers/net/mlx5/mlx5_rxtx.h
> @@ -584,9 +584,11 @@ mlx5_tx_mb2mr(struct mlx5_txq_data *txq, struct rte_mbuf
> *mb)
> * Pointer to TX queue structure.
> * @param wqe
> * Poin
Hi Gaetan,
Thanks for your quick reply. Please see comments inline.
Regards,
Ophir
> -Original Message-
> From: Gaëtan Rivet [mailto:gaetan.ri...@6wind.com]
> Sent: Friday, October 20, 2017 1:35 PM
> To: Ophir Munk
> Cc: dev@dpdk.org; Thomas Monjalon ; Olga Shern
> ; sta...@dpdk.org
> Su
When number of DCB traffic class is 4, user priority should be
mapped to traffic class 0/1/2/3.
Fixes: cb60ede6e3b6 ("ethdev: rename DCB field in config structs")
Cc: sta...@dpdk.org
Signed-off-by: Wei Dai
---
app/test-pmd/testpmd.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
dif
Yongseok, Sagi, my small contribution to this discussion,
On Sun, Oct 22, 2017 at 03:01:04PM -0700, Yongseok Koh wrote:
> On Sun, Oct 22, 2017 at 12:46:53PM +0300, Sagi Grimberg wrote:
> >
> > > Configuring UAR as IO-mapped makes maximum throughput decline by
> > > noticeable
> > > amount. If UA
> From: Pavan Nikhilesh Bhagavatula [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Friday, October 20, 2017 8:09 PM
> To: Van Haaren, Harry
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/3] evendev: fix inconsistency in event
> queue config
>
> On Fri, Oct 20, 2017 at 04:38:57PM +, V
On 21/10/2017 18:12, Akhil Goyal wrote:
On 10/17/2017 6:47 PM, De Lara Guarch, Pablo wrote:
-Original Message-
From: Hemant Agrawal [mailto:hemant.agra...@nxp.com]
Sent: Tuesday, October 17, 2017 1:37 PM
To: De Lara Guarch, Pablo ; Aviad
Yehezkel ; Gonzalez Monroy, Sergio
; dev@dpdk.o
> -Original Message-
> From: alok.makhar...@nxp.com [mailto:alok.makhar...@nxp.com]
> Sent: Sunday, October 22, 2017 8:36 PM
> To: dev@dpdk.org
> Cc: De Lara Guarch, Pablo ;
> hemant.agra...@nxp.com; Alok Makhariya
> Subject: [PATCH v2 1/2] crypto/dpaa2_sec: add support for out of place
On Mon, Oct 23, 2017 at 07:17:41AM +, Ophir Munk wrote:
> Hi Gaetan,
> Thanks for your quick reply. Please see comments inline.
>
> Regards,
> Ophir
>
> > -Original Message-
> > From: Gaëtan Rivet [mailto:gaetan.ri...@6wind.com]
> > Sent: Friday, October 20, 2017 1:35 PM
> > To: Ophir
On Mon, Oct 23, 2017 at 08:04:26AM +, Van Haaren, Harry wrote:
> > From: Pavan Nikhilesh Bhagavatula [mailto:pbhagavat...@caviumnetworks.com]
> > Sent: Friday, October 20, 2017 8:09 PM
> > To: Van Haaren, Harry
> > Cc: dev@dpdk.org
> > Subject: Re: [dpdk-dev] [PATCH 1/3] evendev: fix inconsist
Hi Gaetan,
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Gaëtan Rivet
> Sent: Wednesday, October 11, 2017 3:32 PM
> To: Doherty, Declan
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH v2 00/14] Move PCI away from the EAL
>
> On Wed, Oct 11, 2017 at 03:19
Thanks Matan,
On Sat, Oct 21, 2017 at 08:54:45PM +, Matan Azrad wrote:
> The stats_get API was changed to signal a potential failure to read
> stats. Furthermore, some PMDs are able to provide statistics even
> after a removal event occurred.
>
> Considering this, the fail-safe can try to acc
On Sat, Oct 21, 2017 at 08:54:46PM +, Matan Azrad wrote:
> Fail-safe attempts to read an ultimate statistics on removal time; if
> that fails, it uses the latest recorded snapshot.
>
> This patch adds timestamp for each stats snapshot to allow a time report
> since the last snapshot in case of
Hi Jerin
On 10/20/2017 1:43 PM, Jerin Jacob Wrote:
-Original Message-
[...]
dependant on each other.
Thus a memory barrier is neccessary.
Yes. The barrier is necessary.
In fact, upstream freebsd fixed this issue for arm64. DPDK ring
implementation is derived from freebsd's buf_ring
Hello Pablo,
On Mon, Oct 23, 2017 at 08:44:31AM +, De Lara Guarch, Pablo wrote:
> Hi Gaetan,
>
> > -Original Message-
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Gaëtan Rivet
> > Sent: Wednesday, October 11, 2017 3:32 PM
> > To: Doherty, Declan
> > Cc: dev@dpdk.org
> > S
Hi, Adrien
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Adrien Mazarguil
> Sent: Saturday, December 17, 2016 12:25 AM
> To: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2 01/25] ethdev: introduce generic flow API
>
> This new API supersedes all the legacy fi
Hi,
> On 10/20/2017 1:43 PM, Jerin Jacob Wrote:
> > -Original Message-
> >>
> [...]
> >> dependant on each other.
> >> Thus a memory barrier is neccessary.
> > Yes. The barrier is necessary.
> > In fact, upstream freebsd fixed this issue for arm64. DPDK ring
> > implementation is derived
On Mon, Oct 23, 2017 at 09:05:24AM +, Kuusisaari, Juhamatti wrote:
>
> Hi,
>
> > On 10/20/2017 1:43 PM, Jerin Jacob Wrote:
> > > -Original Message-
> > >>
> > [...]
> > >> dependant on each other. Thus a memory barrier is neccessary.
> > > Yes. The barrier is necessary. In fact, ups
Hi,
The next Technical Board meeting will happen on IRC
this Thursday 26 at 2pm UTC.
You will find links to the agenda and previous meetings on this page:
http://dpdk.org/dev/techboard-minutes
Hey folks,
to assist in the work of moving the PCI
(http://dpdk.org/ml/archives/dev/2017-October/078803.html) and virtual
bus (http://dpdk.org/ml/archives/dev/2017-October/077858.html,
http://dpdk.org/ml/archives/dev/2017-October/078783.html) out of EAL
I've put together this patch set which
Changes in v2:
* Split last second commit in two, one for testpmd the second one for the PMD
* Re-work "net/mlx5: fix RSS hash update" to also fix a bug introduced in the
hash_get() blocking the update.
Nelio Laranjeiro (3):
net/mlx5: fix flow priority for queue action
app/testpmd: fix R
Priority is wrongly configured when the action is queue, using the Ethernet
layer priority instead of the most specific layer found.
Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Signed-off-by: Nelio Laranjeiro
Acked-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_flow
Struct rss_conf.rss_key_len is not initialised forcing the user to verify
the rss_conf.rss_key pointer to know if the key is present or not.
rss_conf.rss_key_len should have a valid length according to the size of
the rss_key pointed.
Fixes: 560e02ee5237 ("app/testpmd: configure RSS without restar
Few bugs fixes in both configuration get and hash update where inputs are
not handled as expected by the ethdev layer.
RSS structure may not be totally usable, the PMD should try to take as most
information from it has it can when it is an hash update or it should try
to fill as most as possible i
Remove rte_set_log_level(), rte_get_log_level(),
rte_set_log_type(), and rte_get_log_type().
Also update librte_eal.so version in docuementation.
The LIBABIVER variable in eal has already been modified in
commit f26ab687a74f ("eal: remove Xen dom0 support").
Signed-off-by: Olivier Matz
---
doc/
Hi Akhil,
>
> Hi Konstantin,
>
> On 10/19/2017 2:53 PM, Ananyev, Konstantin wrote:
> > Hi guys,
> >
> >> -Original Message-
> >> From: Akhil Goyal [mailto:akhil.go...@nxp.com]
> >> Sent: Saturday, October 14, 2017 11:17 PM
> >> To: dev@dpdk.org
> >> Cc: Doherty, Declan ; De Lara Guarch,
v1:
This series is a follow-up of the new datapath implementation introduced in RC1.
It is dedicated to Tx and improves the performance.
Matan Azrad (4):
net/mlx4: remove error flows from Tx fast path
net/mlx4: merge Tx path functions
net/mlx4: remove unnecessary variables in Tx burst
net/
Change functions to inline on Tx fast path to improve performance
Inside the inline function call other functions to handle "unlikely"
cases such that the inline function code footprint is small.
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 43 ++-
From: Matan Azrad
Move unnecessary error flows to DEBUG mode for
performance improvements.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 16 ++--
1 file changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx4_rxtx
This commit optimizes handling of one segment and calls a
dedicated function for handling multi segments
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 284 +++
1 file changed, 154 insertions(+), 130 deletions(-)
diff --git a/drivers/net/mlx
mlx4 NIC is using lkey in big endian format. Save lkey in this format
in order to avoid conversions during Tx fast path
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 14 ++
drivers/net/mlx4/mlx4_rxtx.h | 10 ++
2 files changed, 12 insertions(+), 12 deletions(-)
From: Matan Azrad
Merge tx_burst and mlx4_post_send functions to prevent
double asking about WQ remain space.
This should improve performance.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 353 +--
1 file changed, 170 insertions(+), 183
From: Matan Azrad
Remove usage of variables which doesn't add new information for
performance improvement.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx
From: Matan Azrad
Since one segment shouldn't use additional memory to save segments
byte_count for writing them in different order we can prevent
additional memory unnecessary usage in this case.
By the way, prevent loop management.
All for performance improvement.
Signed-off-by: Matan Azrad
-Original Message-
> Date: Mon, 23 Oct 2017 16:49:01 +0800
> From: Jia He
> To: Jerin Jacob
> Cc: "Ananyev, Konstantin" , "Zhao, Bing"
> , Olivier MATZ ,
> "dev@dpdk.org" , "jia...@hxt-semitech.com"
> , "jie2@hxt-semitech.com"
> , "bing.z...@hxt-semitech.com"
> , "Richardson, Bru
Hi Jianfeng,
> -Original Message-
> From: Tan, Jianfeng
> Sent: Friday, October 13, 2017 12:52 PM
> To: dev@dpdk.org
> Cc: jblu...@infradead.org; Richardson, Bruce
> ; Ananyev, Konstantin
> ; De Lara Guarch, Pablo
> ; tho...@monjalon.net;
> y...@fridaylinux.org; maxime.coque...@redhat.com;
18/10/2017 22:57, Ferruh Yigit:
> On 10/13/2017 5:09 PM, Thomas Monjalon wrote:
> > The libraries which have their ABI version increased in this release
> > must be prepended with a + sign to make them appear clearly.
> >
> > Fixes: f8244c6399d9 ("ethdev: increase port id range")
> > Fixes: ec5144
The libraries which have their ABI version increased in this release
must be prepended with a + sign to make them appear clearly.
Fixes: f8244c6399d9 ("ethdev: increase port id range")
Fixes: ec51443cc99a ("gso: add Generic Segmentation Offload API framework")
Signed-off-by: Thomas Monjalon
Acke
Fix problem with uninitialized rx/tx deferred_start flags of the struct
rte_eth_rxconf/txconf by initializing with 0 value (deferred start of
the rx/tx queues is turned off). This setting allows device rx/tx
queues to start with rte_eth_dev_start().
CID 194999 (#1 of 1): Uninitialized scalar varia
On Tue, Oct 17, 2017 at 05:41:02PM +0100, Bruce Richardson wrote:
> For static builds, even though the drivers are installed in the lib folder
> that does not mean that that folder should be set as the EAL driver search
> path. If it is, EAL will try loading all shared libraries as drivers and
> fa
> > The libraries reorder, distributor, EFD, ACL and member seem to belong to
> > the same category.
> > The previous categories "layers" and "containers" do not fit well.
> > It is proposed to group them under "classification" category.
> >
> > Signed-off-by: Thomas Monjalon
>
> Acked-by: John
For HMAC algorithms (MD5-HMAC, SHAx-HMAC), the supported
digest sizes are not a fixed value, but a range between
1 and the maximum digest size for those algorithms.
Fixes: 26c2e4ad5ad4 ("cryptodev: add capabilities discovery")
Cc: sta...@dpdk.org
Signed-off-by: Pablo de Lara
---
drivers/crypto/
On Wed, Oct 18, 2017 at 06:11:51PM +0100, Luca Boccassi wrote:
> On Tue, 2017-10-17 at 17:12 +0100, Bruce Richardson wrote:
> > This patchset enables sample apps to be used with the new meson build
> > system. The work is really three-fold:
> >
> > * fix issues and pre-requisites to get the sample
Hi Aviad,
> -Original Message-
> From: Gonzalez Monroy, Sergio
> Sent: Friday, October 20, 2017 6:56 AM
> To: avia...@dev.mellanox.co.il; dev@dpdk.org; De Lara Guarch, Pablo
> ; avia...@mellanox.com
> Cc: bor...@mellanox.com; akhil.go...@nxp.com;
> hemant.agra...@nxp.com; Nicolau, Radu ;
>
On 23-Oct-17 7:33 AM, Hemant Agrawal wrote:
Hi Anatoly,
Will you please review the following patch?
Regards,
Hemant
Hi Hemant,
My apologies, slipped through the cracks. Reviewing now.
--
Thanks,
Anatoly
Priority is wrongly configured when the action is queue, using the Ethernet
layer priority instead of the most specific layer found.
Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Signed-off-by: Nelio Laranjeiro
Acked-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_flow
Struct rss_conf.rss_key_len is not initialised forcing the user to verify
the rss_conf.rss_key pointer to know if the key is present or not.
rss_conf.rss_key_len should have a valid length according to the size of
the rss_key pointed.
Fixes: 560e02ee5237 ("app/testpmd: configure RSS without restar
Changes in v3:
* Add missing "fixes" tag in the last commit.
Changes in v2:
* Split last second commit in two, one for testpmd the second one for the PMD
* Re-work "net/mlx5: fix RSS hash update" to also fix a bug introduced in the
hash_get() blocking the update.
Nelio Laranjeiro (3):
n
Few bugs fixes in both configuration get and hash update where inputs are
not handled as expected by the ethdev layer.
RSS structure may not be totally usable, the PMD should try to take as most
information from it has it can when it is an hash update or it should try
to fill as most as possible i
num parameter for strncpy() function should be smaller than
actual destination buffer size to allow null termination.
Fixes: 40d1324423a4 ("net/i40e: get ddp profile protocol info")
Signed-off-by: Kirill Rybalchenko
---
drivers/net/i40e/rte_pmd_i40e.c | 2 +-
1 file changed, 1 insertion(+), 1 d
I talked with Akhil. I will send v3 of patches by EOD and hold back this
patch.
The reason I don't have time to deep dive regression tests and get back
with an example in the upcoming days.
Thanks!
On 10/23/2017 1:54 PM, De Lara Guarch, Pablo wrote:
Hi Aviad,
-Original Message-
F
On 10/22/2017 11:59 PM, Thomas Monjalon wrote:
22/10/2017 22:37, Akhil Goyal:
Hi All,
On 10/21/2017 9:43 PM, Akhil Goyal wrote:
Hi Thomas,
On 10/20/2017 3:02 PM, Thomas Monjalon wrote:
16/10/2017 12:44, Thomas Monjalon:
15/10/2017 00:17, Akhil Goyal:
This patchset introduce the rte_securit
On 11-Oct-17 7:42 AM, Hemant Agrawal wrote:
VFIO may be used by buses other than PCI. This patch enables
the VFIO on the basis of vfio root presence.
Since vfio_enable should be called only once, pci_vfio_enable
is also removed.
A debug print is added in case vfio_pci module is not present.
Si
Hi,
Here are the minutes of the last DPDK technical board meeting
held on 2017-10-13.
Attendees:
- Bruce Richardson
- Hemant Agrawal
- Jan Blunck
- Jerin Jacob
- Konstantin Ananyev
- Olivier Matz
- Stephen Hemminger
- Thomas Monjalon
- Yuanhan Liu
1) Vendor trees
The proposal to introduce ne
PKT_RX_VLAN_PKT and PKT_RX_QINQ_PKT are deprecated for a while.
As explained in [1], these flags were kept to let the applications and
PMDs move to the new flag. There is also a need to support Rx vlan
offload without vlan strip (at least for the ixgbe driver).
This patch renames the old flags for
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Nikhil Rao
> Sent: Wednesday, October 18, 2017 10:58 AM
> To: jerin.ja...@caviumnetworks.com
> Cc: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2] doc: add event eth Rx adapter programmer's
> guide
>
> Add programm
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Wenzhuo Lu
> Sent: Monday, October 23, 2017 6:38 AM
> To: dev@dpdk.org
> Cc: Lu, Wenzhuo
> Subject: [dpdk-dev] [PATCH] doc: how to set VF MAC address on ixgbe
>
> Add the description of how to add a MAC address
On Fri, Oct 20, 2017 at 09:57:38AM -0700, Ferruh Yigit wrote:
> On 10/20/2017 9:55 AM, Ferruh Yigit wrote:
> > Remove device reset during application start, the reset for application
> > exit still there.
> >
> > Reset in open removed because of following comments:
> > 1- Device reset not complete
On Mon, Oct 23, 2017 at 12:24:49PM +0100, Kirill Rybalchenko wrote:
> num parameter for strncpy() function should be smaller than
> actual destination buffer size to allow null termination.
>
> Fixes: 40d1324423a4 ("net/i40e: get ddp profile protocol info")
>
> Signed-off-by: Kirill Rybalchenko
> -Original Message-
> From: Richardson, Bruce
> Sent: Monday 23 October 2017 13:58
> To: Rybalchenko, Kirill
> Cc: dev@dpdk.org; Chilikin, Andrey ; Xing, Beilei
> ; Wu, Jingjing
> Subject: Re: [dpdk-dev] [PATCH] net/i40e: fix value of num parameter for
> strncpy function
>
> On Mon, O
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Xiao Wang
> Sent: Sunday, October 22, 2017 10:37 PM
> To: dev@dpdk.org
> Cc: Wu, Jingjing ; Chen, Jing D
> ; Wang, Xiao W
> Subject: [dpdk-dev] [PATCH 4/4] doc: add switch restart support to fm10k
> guide
>
> Th
> -Original Message-
> From: Ananyev, Konstantin
> Sent: Monday, October 23, 2017 10:57 AM
> To: Akhil Goyal ; dev@dpdk.org
> Cc: Doherty, Declan ; De Lara Guarch, Pablo
> ; hemant.agra...@nxp.com; Nicolau,
> Radu ; bor...@mellanox.com;
> avia...@mellanox.com; tho...@monjalon.net; sandeep
> -Original Message-
> From: Santosh Shukla [mailto:santosh.shu...@caviumnetworks.com]
> Sent: Friday, October 20, 2017 4:21 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; jerin.ja...@caviumnetworks.com; Mcnamara, John
> ; Santosh Shukla
>
> Subject: [PATCH v4 3/3] doc: use correct mem
20/10/2017 16:39, Cristian Dumitrescu:
> The following changes since commit 6b9ed026a8704b9e5ee5da7997617ef7cc82e114:
>
> igb_uio: fix build with kernel <= 3.17 (2017-10-16 13:07:11 +0200)
>
> are available in the git repository at:
>
> http://dpdk.org/git/next/dpdk-next-tm
Pulled
I've ju
> --- a/lib/Makefile
> +++ b/lib/Makefile
> @@ -83,6 +83,9 @@ DIRS-$(CONFIG_RTE_LIBRTE_POWER) += librte_power
> DEPDIRS-librte_power := librte_eal
> DIRS-$(CONFIG_RTE_LIBRTE_METER) += librte_meter
> DEPDIRS-librte_meter := librte_eal
> +DIRS-$(CONFIG_RTE_LIBRTE_FLOW_CLASSIFY) += librte_flow_cla
On Mon, Oct 23, 2017 at 02:40:36PM +0800, Zhiyong Yang wrote:
> In the function virtqueue_enqueue_xmit(), when can_push is true,
> vtnet_hdr_size is added to pkt_len by calling rte_pktmbuf_prepend.
> So, virtio header length should be subtracted before calling stats
> function.
>
> Fixes: 58169a9c
On Mon, Oct 23, 2017 at 02:06:05PM +0100, Rybalchenko, Kirill wrote:
>
>
> > -Original Message-
> > From: Richardson, Bruce
> > Sent: Monday 23 October 2017 13:58
> > To: Rybalchenko, Kirill
> > Cc: dev@dpdk.org; Chilikin, Andrey ; Xing, Beilei
> > ; Wu, Jingjing
> > Subject: Re: [dpdk-
Hi Jasvinder,
Thanks for reviewing.
> -Original Message-
> From: Singh, Jasvinder
> Sent: Monday, October 23, 2017 2:22 PM
> To: Iremonger, Bernard ; dev@dpdk.org;
> Yigit, Ferruh ; Ananyev, Konstantin
> ; Dumitrescu, Cristian
> ; adrien.mazarg...@6wind.com
> Subject: RE: [PATCH v9 1/4] l
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Friday, October 20, 2017 10:11 PM
> To: Mcnamara, John ; Santosh Shukla
>
> Cc: dev@dpdk.org; jerin.ja...@caviumnetworks.com
> Subject: Re: [PATCH v4 0/3] Octeontx doc misc
>
> 20/10/2017 18:07, Mcnamara,
From: Matan Azrad
Move unnecessary error flows to DEBUG mode for
performance improvements.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 16 ++--
1 file changed, 6 insertions(+), 10 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx4_rxtx
v1:
This series is a follow-up of the new datapath implementation introduced in RC1.
It is dedicated to Tx and improves the performance.
v2:
Change uint32_t be_lkey to rte_be32_t lkey
Matan Azrad (4):
net/mlx4: remove error flows from Tx fast path
net/mlx4: merge Tx path functions
net/mlx4:
Change functions to inline on Tx fast path to improve performance
Inside the inline function call other functions to handle "unlikely"
cases such that the inline function code footprint is small.
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 43 ++-
mlx4 NIC is using lkey in big endian format. Save lkey in this format
in order to avoid conversions during Tx fast path
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 16 +++-
drivers/net/mlx4/mlx4_rxtx.h | 14 --
2 files changed, 15 insertions(+), 15 deleti
From: Matan Azrad
Merge tx_burst and mlx4_post_send functions to prevent
double asking about WQ remain space.
This should improve performance.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 353 +--
1 file changed, 170 insertions(+), 183
From: Matan Azrad
Remove usage of variables which doesn't add new information for
performance improvement.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4_rxtx.c | 8 ++--
1 file changed, 2 insertions(+), 6 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx
From: Matan Azrad
Since one segment shouldn't use additional memory to save segments
byte_count for writing them in different order we can prevent
additional memory unnecessary usage in this case.
By the way, prevent loop management.
All for performance improvement.
Signed-off-by: Matan Azrad
This commit optimizes handling of one segment and calls a
dedicated function for handling multi segments
Signed-off-by: Ophir Munk
---
drivers/net/mlx4/mlx4_rxtx.c | 284 +++
1 file changed, 154 insertions(+), 130 deletions(-)
diff --git a/drivers/net/mlx
> > Coverity Issue: 195000
> > Fixes: 02fd6c744350 ("mempool/octeontx: support allocation")
> >
> > Signed-off-by: Santosh Shukla
>
> Acked-by: Jerin Jacob
Applied, thanks
Hi Jianfeng,
> -Original Message-
> From: Tan, Jianfeng
> Sent: Friday, October 13, 2017 12:52 PM
> To: dev@dpdk.org
> Cc: jblu...@infradead.org; Richardson, Bruce
> ; Ananyev, Konstantin
> ; De Lara Guarch, Pablo
> ; tho...@monjalon.net;
> y...@fridaylinux.org; maxime.coque...@redhat.com;
> -Original Message-
> From: Santosh Shukla [mailto:santosh.shu...@caviumnetworks.com]
> Sent: Friday, October 20, 2017 4:21 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; jerin.ja...@caviumnetworks.com; Mcnamara, John
> ; Santosh Shukla
>
> Subject: [PATCH v4 1/3] doc: add platform de
This change is rejected because this PMD has no API/ABI.
16/10/2017 15:42, Santosh Shukla:
> This commit bumps the library version to reflect the ABI change
> caused by removing the below function from event/octeontx:
> * octeontx_ssovf_info
> * octeontx_ssovf_bar
> * octeontx_ssovf_mbox_send
>
>
> -Original Message-
> From: Olivier Matz [mailto:olivier.m...@6wind.com]
> Sent: Thursday, October 12, 2017 11:04 AM
> To: tho...@monjalon.net
> Cc: Eads, Gage ; santosh.shu...@caviumnetworks.com;
> Richardson, Bruce ;
> jerin.ja...@caviumnetworks.com; dev@dpdk.org
> Subject: [PATCH 0/3]
> From: Pavan Nikhilesh Bhagavatula [mailto:pbhagavat...@caviumnetworks.com]
> Sent: Monday, October 23, 2017 9:42 AM
> To: Van Haaren, Harry ;
> jerin.ja...@caviumnetworks.com
> Cc: dev@dpdk.org
> Subject: Re: [dpdk-dev] [PATCH 1/3] evendev: fix inconsistency in event
> queue config
>
> On Mon, O
> -Original Message-
> From: Santosh Shukla [mailto:santosh.shu...@caviumnetworks.com]
> Sent: Friday, October 20, 2017 4:21 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; jerin.ja...@caviumnetworks.com; Mcnamara, John
> ; Santosh Shukla
>
> Subject: [PATCH v4 2/3] doc: add mempool and
Some fixes on bugs and wrong behaviors.
This series applies on top of:
http://dpdk.org/patch/30707
Changes in v2:
* Fix packets reception in allmulti mode and when VLAN are present.
* Fix a double flow insertion causing an infinite loop.
Nelio Laranjeiro (7):
net/mlx5: fix segfault on flow
When ports are stopped, the hash Rx queue should not be created.
Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Signed-off-by: Nelio Laranjeiro
Acked-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_flow.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/net
Indirection table size must be in log to communicate with verbs when the
number of queue is not a power of two, the maximum indirection table size
is use, but not converted to log2. This makes a memory corruption.
Fixes: 4c7a0f5ff876 ("net/mlx5: make indirection tables shareable")
Signed-off-by:
Flow director drop action as not been brought back with the new
implementation on top of rte flow.
Fixes: 4c3e9bcdd52e ("net/mlx5: support flow director")
Signed-off-by: Nelio Laranjeiro
Acked-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_flow.c | 25 -
1 file changed, 16
Fix the drop queue rule creation when the port is stopped.
Fixes: 8086cf08b2f0 ("net/mlx5: handle RSS hash configuration in RSS flow")
Signed-off-by: Nelio Laranjeiro
Acked-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_flow.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/dri
Flows are added by priv_flow_create() in the associated list, adding them a
second time corrupts the list causing an infinite loop when parsing it.
Fixes: 4c3e9bcdd52e ("net/mlx5: support flow director")
Signed-off-by: Nelio Laranjeiro
---
drivers/net/mlx5/mlx5_flow.c | 1 -
1 file changed, 1 d
Marking a packet which will not be received by the NIC is useless, even if
this action remains possible, it blocks the creation of the flow counter
which embed a mark action to a drop queue to be created.
Fixes: 31ba9997f11a ("net/mlx5: fully convert a flow to verbs in validate")
Signed-off-by: N
16/10/2017 15:42, Santosh Shukla:
> Octeontx mempool pmd need mailbox for pool setup.
> That mailbox was defined at drivers/event/octeontx.
> So mempool has dependency on event/octeontx driver and
> commit:8700239f7767 has added make rule which makes sure event/*
> get build before mempool, but thi
When VLAN is enabled in the Rx side, only packets matching this VLAN are
expected, this also includes the broadcast and all multicast packets.
Fixes: 272733b5ebfd ("net/mlx5: use flow to enable unicast traffic")
Fixes: 6a6b6828fe6a ("net/mlx5: use flow to enable all multi mode")
Signed-off-by: Ne
On Mon, Oct 23, 2017 at 02:45:32PM +, Van Haaren, Harry wrote:
> > From: Pavan Nikhilesh Bhagavatula [mailto:pbhagavat...@caviumnetworks.com]
> > Sent: Monday, October 23, 2017 9:42 AM
> > To: Van Haaren, Harry ;
> > jerin.ja...@caviumnetworks.com
> > Cc: dev@dpdk.org
> > Subject: Re: [dpdk-dev
20/10/2017 14:31, Santosh Shukla:
> v3:
> Include v2 review comment and rebased on
> top of upstream tip commit:6b9ed026a8704
>
>
> v2:
> Include build fixes reported in patchworks.
Please fix checkpatch issues, thanks
> -Original Message-
> From: Van Haaren, Harry
> Sent: Thursday, October 19, 2017 5:31 AM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; Mcnamara, John ;
> Eads, Gage ; Van Haaren, Harry
>
> Subject: [PATCH v2] example: add new service cores sample application
>
> This commit adds a new
DPDK works with packets, but some network administration tools work with flow
information.
This library is proposed to provide an API to classify received packets using
flow patterns.
Basically the library consists of APIs to create the classifier object, add a
table to the classifer,
add and
From: Ferruh Yigit
The following APIs's are implemented in the
librte_flow_classify library:
rte_flow_classifier_create
rte_flow_classifier_free
rte_flow_classifier_query
rte_flow_classify_table_create
rte_flow_classify_table_entry_add
rte_flow_classify_table_entry_delete
The following librte_t
Add flow_classify_autotest program.
Set up IPv4 ACL field definitions.
Create table_acl for use by librte_flow_classify API's.
Create an mbuf pool for use by rte_flow_classify_query.
For each of the librte_flow_classify API's:
test with invalid parameters
test with invalid patterns
test with inva
add initialize_tcp_header function
add initialize_stcp_header function
add initialize_ipv4_header_proto function
add generate_packet_burst_proto function
Signed-off-by: Bernard Iremonger
---
test/test/packet_burst_generator.c | 191 +
test/test/packet_burst_ge
The flow_classify sample application exercises the following
librte_flow_classify API's:
rte_flow_classifier_create
rte_flow_classifier_query
rte_flow_classify_table_create
rte_flow_classify_table_entry_add
rte_flow_classify_table_entry_delete
It sets up the IPv4 ACL field definitions.
It creates
1 - 100 of 217 matches
Mail list logo
