The action for L2 tunnel should be VF or PF, not QUEUE.
Fixes: 99e7003831c ("net/ixgbe: parse L2 tunnel filter")
Signed-off-by: Wei Zhao
Acked-by: Wei Dai
---
v2:
-add vf id check.
v3:
-add action support for PF.
---
drivers/net/ixgbe/ixgbe_flow.c | 29 -
1 file
Hi, Jingjing
Thanks for your feedback.
> > Subject: [PATCH 2/2] net/i40e: fix Rx queue interrupt mapping in VF
> >
> > When a VF port is bound to VFIO-PCI, miscellaneous interrupt is mapped
> > to MSI-X vector 0 and Rx queues interrupt are mapped to other vectors
> > in vfio_enable_msix( ). To s
Hi, Jingjing
Thanks for your feedback.
> > Subject: [PATCH 1/2] net/i40e: fix VFIO interrupt mapping in VF
> >
> > When a VF port is bound to VFIO-PIC, only miscellaneous interrupt is
> > mapped to VFIO vector 0 in i40evf_dev_init( ).
> > In ixgbevf_dev_interrupt_handler( ), if previous VFIO inter
On Fri, Nov 03, 2017 at 11:21:43AM +0800, jianbo@arm.com wrote:
> The 11/02/2017 15:52, Ananyev, Konstantin wrote:
> >
> >
> > > -Original Message-
> > > From: Guduri Prathyusha [mailto:gprathyu...@caviumnetworks.com]
> > > Sent: Thursday, November 2, 2017 3:34 PM
> > > To: Ananyev, Kon
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Peng, Yuan
> Sent: Friday, November 3, 2017 11:39 AM
> To: Zhao1, Wei ; dev@dpdk.org
> Cc: Zhao1, Wei
> Subject: Re: [dpdk-dev] [PATCH v2] net/ixgbe: fix filter parser error in L2
> tunnel
>
> Tested-by: Peng Yu
The 11/02/2017 15:52, Ananyev, Konstantin wrote:
>
>
> > -Original Message-
> > From: Guduri Prathyusha [mailto:gprathyu...@caviumnetworks.com]
> > Sent: Thursday, November 2, 2017 3:34 PM
> > To: Ananyev, Konstantin
> > Cc: dev@dpdk.org; jianbo@arm.com; guduriprathyu...@gmail.com; Kan
This patch adds limitation of VF VLAN strip function.
Signed-off-by: Beilei Xing
---
doc/guides/nics/i40e.rst | 5 +
1 file changed, 5 insertions(+)
diff --git a/doc/guides/nics/i40e.rst b/doc/guides/nics/i40e.rst
index 05c1875..214fada 100644
--- a/doc/guides/nics/i40e.rst
+++ b/doc/guides
Tested-by: Peng Yuan
- Tested Branch: dpdk-master
- Tested Commit: 8ced1542f7a356097c0b24bd1e08db670ff31b92
- OS: 4.8.6-300.fc25.x86_64
- GCC: gcc (GCC) 6.4.1 20170727 (Red Hat 6.4.1-1)
- CPU: Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz
- NIC: Intel Corporation Ethernet Connection X552/X557-AT 10GBASE-
Tested-by: Peng Yuan
- Tested Branch: dpdk-master
- Tested Commit: commit 8ced1542f7a356097c0b24bd1e08db670ff31b92
- OS: 4.8.6-300.fc25.x86_64
- GCC: gcc (GCC) 6.4.1 20170727 (Red Hat 6.4.1-1)
- CPU: Intel(R) Xeon(R) CPU D-1541 @ 2.10GHz
- NIC: X552/X557-AT 10GBASE-T [8086:15ad]
- Default x86_64-
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Hemant Agrawal
> Sent: Thursday, November 2, 2017 6:09 PM
> To: dev@dpdk.org
> Cc: Yigit, Ferruh ; Liu, Yong ;
> sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH 4/5] net/i40e: fix compilation on ARM BE
>
> This patch
On 11/3/2017 8:51 AM, Ferruh Yigit wrote:
Adding a compile time configuration option to control device reset done
during DPDK application exit.
Config option is CONFIG_RTE_EAL_IGB_UIO_RESET and enabled by default,
so by default reset will happen. Having this reset is safer to be sure
device le
Hi Jerin
On 11/2/2017 4:57 PM, Jia He Wrote:
Hi， Jerin
please see my performance test below
On 11/2/2017 3:04 AM, Jerin Jacob Wrote:
[...]
Should it be like instead?
+#else
+ *old_head = __atomic_load_n(&r->cons.head, __ATOMIC_ACQUIRE);
+ const uint32_t prod_tail = __atomic_loa
> -Original Message-
> From: Dai, Wei
> Sent: Thursday, November 2, 2017 9:11 PM
> To: Wu, Jingjing ; Xing, Beilei
> ;
> Liang, Cunming
> Cc: dev@dpdk.org; Dai, Wei ; sta...@dpdk.org
> Subject: [PATCH 1/2] net/i40e: fix VFIO interrupt mapping in VF
>
> When a VF port is bound to VFIO-P
> -Original Message-
> From: Dai, Wei
> Sent: Thursday, November 2, 2017 9:11 PM
> To: Wu, Jingjing ; Xing, Beilei
> ;
> Liang, Cunming
> Cc: dev@dpdk.org; Dai, Wei ; sta...@dpdk.org
> Subject: [PATCH 2/2] net/i40e: fix Rx queue interrupt mapping in VF
>
> When a VF port is bound to VF
Hi Jerin
On 11/3/2017 1:23 AM, Jerin Jacob Wrote:
-Original Message-
Date: Thu, 2 Nov 2017 08:43:30 +
From: Jia He
To: jerin.ja...@caviumnetworks.com, dev@dpdk.org, olivier.m...@6wind.com
Cc: konstantin.anan...@intel.com, bruce.richard...@intel.com,
jianbo@arm.com, hemant.a
Adding a compile time configuration option to control device reset done
during DPDK application exit.
Config option is CONFIG_RTE_EAL_IGB_UIO_RESET and enabled by default,
so by default reset will happen. Having this reset is safer to be sure
device left in a proper case.
But for special cases [1
On 11/2/2017 11:45 AM, Mody, Rasesh wrote:
>> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
>> Sent: Thursday, November 02, 2017 11:10 AM
>>
>> On 11/2/2017 10:34 AM, Mody, Rasesh wrote:
From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
Sent: Thursday, November 02, 2017 1:55 AM
Adding a compile time configuration option to control device reset done
during DPDK application exit.
Config option is CONFIG_RTE_EAL_IGB_UIO_RESET and enabled by default,
so by defualt reset will happen. Having this reset is safer to be sure
device left in a proper case.
But for special cases [1
On 11/2/2017 3:03 AM, Luca Boccassi wrote:
> On Wed, 2017-11-01 at 16:38 -0400, Chas Williams wrote:
>> From: Chas Williams
>>
>> At the end of the queue release, we can free the containers for the
>> queue objects.
>>
>> Fixes: dfaff37fc46d ("vmxnet3: import new vmxnet3 poll mode driver
>> implem
It is easier to find all constructor functions when they use
the same macros RTE_INIT or RTE_INIT_PRIO.
The macro definitions are moved from rte_eal.h to rte_common.h.
Signed-off-by: Thomas Monjalon
---
examples/performance-thread/common/lthread_diag.c | 3 +--
examples/performance-thread/com
The file rte_config.h is generated and automatically included
with -include option.
The explicit includes in drivers and libraries are useless.
Signed-off-by: Thomas Monjalon
---
drivers/bus/dpaa/dpaa_bus.c | 1 -
drivers/crypto/aesni_gcm/aesni_gcm_pmd.c | 1 -
drivers/cry
On 11/2/2017 9:42 AM, Matan Azrad wrote:
> 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
>
> v3:
> Rebase to 17.11-rc2.
> Remove "net/mlx4: save l
On 11/2/2017 6:41 AM, Adrien Mazarguil wrote:
> Fixes: 76df01ff622f ("net/mlx4: separate debugging macros")
>
> Signed-off-by: Adrien Mazarguil
Applied to dpdk-next-net/master, thanks.
On 11/2/2017 10:07 AM, Adrien Mazarguil wrote:
> On Thu, Nov 02, 2017 at 04:42:43PM +, Matan Azrad wrote:
>> 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
This patch adds vhost_crypto sample application to DPDK.
Signed-off-by: Fan Zhang
---
examples/vhost_crypto/Makefile | 59
examples/vhost_crypto/main.c | 617 +
2 files changed, 676 insertions(+)
create mode 100644 examples/vhost_crypto/Makefile
Signed-off-by: Fan Zhang
---
lib/librte_vhost/Makefile | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile
index e1084ab..c6ec7ff 100644
--- a/lib/librte_vhost/Makefile
+++ b/lib/librte_vhost/Makefile
@@ -48,9 +48,8 @@ end
This patch adds vhost_crypto device driver support to vhost_user.
Several new APIs are introduced for user to create and delete
vhost_crypto devices, fetch virtio queue descriptors and translate
to DPDK crypto operations for enqueuing to target cryptodev in the
backend. After dequeuing from the cry
This patch adds the vhost_crypto related compile configuration items
to config/common_base. A new item
"CONFIG_RTE_LIBRTE_VHOST_CRYPTO_DATA_QUEUE_NO_COPY" is introduced
to turn on/off data queue no copy feature to vhost_crypto. This
feature is in experimental level so no gurantee it would work,
esp
This patch adds virtio-crypto support to vhost_user.
Signed-off-by: Fan Zhang
---
lib/librte_vhost/vhost_user.h | 27 ++-
1 file changed, 26 insertions(+), 1 deletion(-)
diff --git a/lib/librte_vhost/vhost_user.h b/lib/librte_vhost/vhost_user.h
index 4052dbf..793cdc2 100
This patch adds a vhost_user_dev_priv structure and a vhost_user
message handler function prototype to vhost_user. This allows
different types of devices to add private information and their
device-specific vhost-user message function handlers to
virtio_net structure. The change to vhost_user_msg_h
This is a RFC patch, so please review. Thanks.
This RFC patchset adds crypto backend suppport to vhost_user library,
including a proof-of-concept sample application. The implementation
follows the virtio-crypto specification and have been tested
with qemu 2.9.50 (with several patches applied, deta
This patch adds a private data pointer to virtio_net structure, to
allow more virtio device driver information.
Signed-off-by: Fan Zhang
---
lib/librte_vhost/vhost.h | 4 +++-
1 file changed, 3 insertions(+), 1 deletion(-)
diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h
index 0
Signed-off-by: Thomas Monjalon
---
drivers/net/mlx5/mlx5_rxtx_vec_neon.h | 2 +-
drivers/net/mlx5/mlx5_rxtx_vec_sse.h | 2 +-
drivers/net/mlx5/mlx5_socket.c| 2 +-
3 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/net/mlx5/mlx5_rxtx_vec_neon.h
b/drivers/net/mlx5/mlx
Signed-off-by: Thomas Monjalon
---
lib/librte_eal/common/include/arch/arm/rte_vect.h | 2 +-
lib/librte_eventdev/rte_event_eth_rx_adapter.c| 2 +-
lib/librte_eventdev/rte_eventdev_pmd.h| 2 +-
3 files changed, 3 insertions(+), 3 deletions(-)
diff --git a/lib/librte_eal/common/inc
Hi Shally,
> -Original Message-
> From: Verma, Shally [mailto:shally.ve...@cavium.com]
> Sent: Friday, October 20, 2017 6:40 PM
> To: Trahe, Fiona ; dev@dpdk.org; Athreya, Narayana
> Prasad
> ; Challa, Mahipal
>
> Subject: RE: [dpdk-dev] [RFC] Compression API in DPDK
>
>
> > -Orig
On 11/2/2017 6:30 AM, Nelio Laranjeiro wrote:
> VLAN id is limited to MLX5_MAX_VLAN_IDS which is not verified by the code
> before trying to add a new VLAN filter.
>
> Fixes: 272733b5ebfd ("net/mlx5: use flow to enable unicast traffic")
>
> Signed-off-by: Nelio Laranjeiro
Applied to dpdk-next-n
On 11/2/2017 1:37 AM, Tomasz Duszynski wrote:
> In case errno contains some non-zero value and call to strtoul() is
> successful -2 error will be returned.
>
> Setting errno to zero before calling strtoul() makes sure
> errors will be handled correctly.
>
> Fixes: 0ddc9b815b11 ("net/mrvl: add net
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Thursday, November 02, 2017 11:10 AM
>
> On 11/2/2017 10:34 AM, Mody, Rasesh wrote:
> >> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> >> Sent: Thursday, November 02, 2017 1:55 AM
> >>
> >> On 11/2/2017 1:03 AM, Mody, Rasesh wrote
On 11/1/2017 10:47 PM, John Daley wrote:
> A check was previously added to drop Tx packets greater than what the Nic
> is capable of sending since such packets can freeze the send queue. The
> check did not account for TSO packets however, so TSO was limited to 9208
> bytes.
>
> Check packet lengt
This is done for consistency with the rest of the code.
Fixes: 078b8b452e6b ("net/mlx4: add RSS flow rule action support")
Signed-off-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxq.c | 9 ++---
1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxq.c b/d
Memory regions assigned to hardware and used during Tx/Rx are mapped to
mbuf pools. Each Rx queue creates its own MR based on the mempool provided
during queue setup, while each Tx queue looks up and registers MRs for all
existing mbuf pools instead.
Since most applications use few large mbuf pool
On 11/2/2017 10:34 AM, Mody, Rasesh wrote:
>> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
>> Sent: Thursday, November 02, 2017 1:55 AM
>>
>> On 11/2/2017 1:03 AM, Mody, Rasesh wrote:
From: Stephen Hemminger [mailto:step...@networkplumber.org]
Sent: Wednesday, November 01, 2017 7:12
On 11/1/2017 10:53 PM, Matan Azrad wrote:
> Hi Ferruh, Gaetan
>
>> -Original Message-
>> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
>> Sent: Thursday, November 2, 2017 12:34 AM
>> To: Matan Azrad ; Gaetan Rivet
>>
>> Cc: dev@dpdk.org; john.mcnam...@intel.com
>> Subject: Re: [dpdk-
Hi,
Some issue with my mutt command line and hence apologies for unpleasant
formatting in the mail. Please see inline
From: dev on behalf of Ananyev, Konstantin
Sent: Thursday, November 2, 2017 9:22 PM
To: Prathyusha, Guduri
Cc: dev@dpdk.org; jianbo@arm.c
> From: Ferruh Yigit [mailto:ferruh.yi...@intel.com]
> Sent: Thursday, November 02, 2017 1:55 AM
>
> On 11/2/2017 1:03 AM, Mody, Rasesh wrote:
> >> From: Stephen Hemminger [mailto:step...@networkplumber.org]
> >> Sent: Wednesday, November 01, 2017 7:12 AM
> >>
> >> On Wed, 1 Nov 2017 06:58:53 +000
Hi John,
> > Subject: [PATCH 2/2] doc: flow classify sample app guide
> >
> > This file documents the flow_classify sample application which is used
> > to demonstate use of the Flow Classify library, librte_flow_classify.
> >
> > Updated MAINTAINERS file
> >
> > Signed-off-by: Bernard Iremonger
Hi John,
> > Subject: [PATCH 1/2] doc: flow classify library prog guide
> >
> > This file documents the Flow Classification library, librte_flow_classify.
> >
> > Updated MAINTAINERS file
> >
> > Signed-off-by: Bernard Iremonger
>
> Hi Bernard,
>
> Thanks for the doc.
>
> There is a doc buil
failsafe device has vlan stripping configured at startup however once
a sub device is found as non-capable of vlan-stripping failsafe
updates it configuration and removes vlan stripping from it.
This update occurs only once at startup. Following a later plugin
attempt and in case of vlan stripping
-Original Message-
> Date: Thu, 2 Nov 2017 08:43:30 +
> From: Jia He
> To: jerin.ja...@caviumnetworks.com, dev@dpdk.org, olivier.m...@6wind.com
> Cc: konstantin.anan...@intel.com, bruce.richard...@intel.com,
> jianbo@arm.com, hemant.agra...@nxp.com, Jia He ,
> jie2@hxt-semit
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Shahaf Shuler
> Sent: Monday, October 30, 2017 10:58 AM
> To: nelio.laranje...@6wind.com; adrien.mazarg...@6wind.com;
> ys...@mellanox.com
> Cc: dev@dpdk.org
> Subject: [dpdk-dev] [PATCH v2] doc: update mlx5 guide
> -Original Message-
> From: Shreyansh Jain [mailto:shreyansh.j...@nxp.com]
> Sent: Thursday, October 26, 2017 3:26 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; Mcnamara, John ;
> Shreyansh Jain
> Subject: [PATCH] doc: update release note for DPAA2 Event PMD
>
> Signed-off-by: Shrey
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of
> luca.bocca...@gmail.com
> Sent: Tuesday, October 31, 2017 4:29 PM
> To: dev@dpdk.org
> Cc: tho...@monjalon.net; Luca Boccassi
> Subject: [dpdk-dev] [PATCH] doc/guides: mention 17.11 LTS in
> contributing/stable
On Thu, Nov 02, 2017 at 04:42:43PM +, Matan Azrad wrote:
> 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
>
> v3:
> Rebase to 17.11-rc2.
> Remo
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Rami Rosen
> Sent: Saturday, October 28, 2017 4:37 AM
> To: dev@dpdk.org
> Cc: Rosen, Rami
> Subject: [dpdk-dev] [PATCH] doc: fix an error in DPDK programmers's guide
> (EAL)
>
> Fix an error in DPDK programmer'
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Friday, October 27, 2017 12:18 AM
> To: Mcnamara, John
> Cc: dev@dpdk.org; Ori Kam ; adrien.mazarg...@6wind.com
> Subject: Re: [dpdk-dev] [PATCH v2] doc: add basic howto for flow API
>
> 25/10/2017 17:27,
> -Original Message-
> From: Singh, Jasvinder
> Sent: Thursday, October 26, 2017 12:07 PM
> To: dev@dpdk.org
> Cc: Mcnamara, John
> Subject: [PATCH] doc: fix description of traffic management function in
> testpmd
>
> Fixes: 5b590fbe09b6 ("app/testpmd: add traffic management forwarding
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Rami Rosen
> Sent: Sunday, October 29, 2017 1:25 PM
> To: dev@dpdk.org
> Cc: Rosen, Rami
> Subject: [dpdk-dev] [PATCH] doc: fix a typo in DPDK programmer's guide
>
> This patch fixes a trivial typo in DPDK progr
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Rami Rosen
> Sent: Saturday, October 28, 2017 8:58 AM
> To: dev@dpdk.org
> Cc: Dumitrescu, Cristian ; Rosen, Rami
>
> Subject: [dpdk-dev] [PATCH] doc: fix a typo in ip pipeline app guide
>
> This patch fixes a t
-Original Message-
> Date: Thu, 2 Nov 2017 16:16:33 +
> From: "Ananyev, Konstantin"
> To: Jia He , "jerin.ja...@caviumnetworks.com"
> , "dev@dpdk.org" ,
> "olivier.m...@6wind.com"
> CC: "Richardson, Bruce" , "jianbo@arm.com"
> , "hemant.agra...@nxp.com" ,
> "jie2@hxt-semit
> -Original Message-
> From: Iremonger, Bernard
> Sent: Thursday, October 26, 2017 10:09 AM
> To: dev@dpdk.org; Mcnamara, John
> Cc: Iremonger, Bernard
> Subject: [PATCH 2/2] doc: flow classify sample app guide
>
> This file documents the flow_classify sample application which is used
> -Original Message-
> From: Iremonger, Bernard
> Sent: Thursday, October 26, 2017 10:09 AM
> To: dev@dpdk.org; Mcnamara, John
> Cc: Iremonger, Bernard
> Subject: [PATCH 1/2] doc: flow classify library prog guide
>
> This file documents the Flow Classification library, librte_flow_clas
Replace most of the memory barriers by IO memory barriers since they
are all targeted to the DRAM; This improves code efficiency for
systems which force store order between different addresses.
Only the doorbell register store should be protected by memory barrier
since it is targeted to the PCI m
Optimize single segment case by processing it in different block which
prevents checks, calculations and barriers relevant only for multi
segment case.
Call a dedicated function for handling multi segments case.
Signed-off-by: Matan Azrad
Signed-off-by: Ophir Munk
Acked-by: Adrien Mazarguil
--
Volatilize all Rx/Tx HW negotiation memories to be sure no compiler
optimization prevents either load or store commands.
Fixes: c3c977bbecbd ("net/mlx4: add Tx bypassing Verbs")
Fixes: 9f57340a8087 ("net/mlx4: restore Rx offloads")
Fixes: 6681b845034c ("net/mlx4: add Rx bypassing Verbs")
Fixes: 62
Remove usage of variable which count the packets for completion and
doesn't add more information than packets counter.
Remove no space in elements ring check which is already covered by
regular Tx flow.
Signed-off-by: Matan Azrad
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxtx.c | 6
From: Ophir Munk
Associate memory region to mempool (on data path) in a short function.
Handle the less common case of adding a new memory region to mempool
in a separate function.
Signed-off-by: Ophir Munk
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4.h | 2 ++
drivers/net/mlx4/
Move unnecessary error flows to DEBUG mode.
Signed-off-by: Matan Azrad
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxtx.c | 16 +++-
1 file changed, 7 insertions(+), 9 deletions(-)
diff --git a/drivers/net/mlx4/mlx4_rxtx.c b/drivers/net/mlx4/mlx4_rxtx.c
index 67dc712..79c7
Remove unlikely hint from WQ wraparound check because it is
expected case.
Fixes: c3c977bbecbd ("net/mlx4: add Tx bypassing Verbs")
Signed-off-by: Matan Azrad
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxtx.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers
Merge tx_burst and mlx4_post_send functions to prevent
double asking about WQ remain space.
Signed-off-by: Matan Azrad
Acked-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_rxtx.c | 355 +--
1 file changed, 170 insertions(+), 185 deletions(-)
diff --git a
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
v3:
Rebase to 17.11-rc2.
Remove "net/mlx4: save lkey in big-endian format" patch which degrade
performance.
Merg
> -Original Message-
> From: Akhil Goyal [mailto:akhil.go...@nxp.com]
> Sent: Wednesday, November 1, 2017 6:21 AM
> To: Thomas Monjalon ; Mcnamara, John
>
> Cc: dev@dpdk.org; Doherty, Declan ; De Lara
> Guarch, Pablo ; hemant.agra...@nxp.com;
> Nicolau, Radu ; bor...@mellanox.com;
> avia
> -Original Message-
> From: Jia He [mailto:hejia...@gmail.com]
> Sent: Thursday, November 2, 2017 3:43 PM
> To: Ananyev, Konstantin ;
> jerin.ja...@caviumnetworks.com; dev@dpdk.org; olivier.m...@6wind.com
> Cc: Richardson, Bruce ; jianbo@arm.com;
> hemant.agra...@nxp.com; jie2@
This patch adds vhost_crypto sample application to DPDK.
Signed-off-by: Fan Zhang
---
examples/vhost_crypto/Makefile | 59
examples/vhost_crypto/main.c | 617 +
2 files changed, 676 insertions(+)
create mode 100644 examples/vhost_crypto/Makefile
This patch adds vhost_crypto device driver support to vhost_user.
Several new APIs are introduced for user to create and delete
vhost_crypto devices, fetch virtio queue descriptors and translate
to DPDK crypto operations for enqueuing to target cryptodev in the
backend. After dequeuing from the cry
Signed-off-by: Fan Zhang
---
lib/librte_vhost/Makefile | 5 ++---
1 file changed, 2 insertions(+), 3 deletions(-)
diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile
index e1084ab..c6ec7ff 100644
--- a/lib/librte_vhost/Makefile
+++ b/lib/librte_vhost/Makefile
@@ -48,9 +48,8 @@ end
This patch adds the vhost_crypto related compile configuration items
to config/common_base. A new item
"CONFIG_RTE_LIBRTE_VHOST_CRYPTO_DATA_QUEUE_NO_COPY" is introduced
to turn on/off data queue no copy feature to vhost_crypto. This
feature is in experimental level so no gurantee it would work,
esp
This RFC patchset adds crypto backend suppport to vhost_user library,
including a proof-of-concept sample application. The implementation
follows the virtio-crypto specification and have been tested
with qemu 2.9.50 (with several patches applied, detailed later)
with Fedora 24 running in the fronte
This patch adds a vhost_user_dev_priv structure and a vhost_user
message handler function prototype to vhost_user. This allows
different types of devices to add private information and their
device-specific vhost-user message function handlers to
virtio_net structure. The change to vhost_user_msg_h
This patch adds virtio-crypto support to vhost_user.
Signed-off-by: Fan Zhang
---
lib/librte_vhost/vhost_user.h | 27 ++-
1 file changed, 26 insertions(+), 1 deletion(-)
diff --git a/lib/librte_vhost/vhost_user.h b/lib/librte_vhost/vhost_user.h
index 4052dbf..793cdc2 100
On 11/02/2017 09:21 AM, Maxime Coquelin wrote:
Hi Lei,
On 11/02/2017 08:21 AM, Yao, Lei A wrote:
...
Hi, Maxime > I met one issue with your patch set during the v17.11 test.
Is it with v17.11-rc2 or -rc1?
The test scenario is following,
1. Bind one NIC, use test-pmd set vhost-user w
> -Original Message-
> From: Guduri Prathyusha [mailto:gprathyu...@caviumnetworks.com]
> Sent: Thursday, November 2, 2017 3:34 PM
> To: Ananyev, Konstantin
> Cc: dev@dpdk.org; jianbo@arm.com; guduriprathyu...@gmail.com; Kantecki,
> Tomasz
> Subject: Re: [dpdk-dev] [PATCH ] examples
Hi Ananyev
On 11/2/2017 9:26 PM, Ananyev, Konstantin Wrote:
Hi Jia,
-Original Message-
From: Jia He [mailto:hejia...@gmail.com]
Sent: Thursday, November 2, 2017 8:44 AM
To: jerin.ja...@caviumnetworks.com; dev@dpdk.org; olivier.m...@6wind.com
Cc: Ananyev, Konstantin ; Richardson, Bruce
Fail-safe PMD expects to get -ENODEV error value if sub PMD control
command fails because of device removal.
Make control callbacks return with -ENODEV when the device has
disappeared.
Signed-off-by: Matan Azrad
---
drivers/net/mlx5/mlx5.h| 1 +
drivers/net/mlx5/mlx5_ethdev.c | 39
There is time between the physical removal of the device until sub-device PMDs
get a RMV interrupt. At this time DPDK PMDs and applications still don't know
about the removal and may call sub-device control operation which should return
an error.
This series adjusts the -ENODEV error value to fa
There is time between the physical removal of the device until
sub-device PMDs get a RMV interrupt. At this time DPDK PMDs and
applications still don't know about the removal and may call sub-device
control operation which should return an error.
In previous code this error is reported to the appl
Fail-safe PMD expects to get -ENODEV error value if sub PMD control
command fails because of device removal.
Make control callbacks return with -ENODEV when the device has
disappeared.
Signed-off-by: Matan Azrad
---
drivers/net/mlx4/mlx4.h| 1 +
drivers/net/mlx4/mlx4_ethdev.c | 38
On Thu, Nov 02, 2017 at 02:46:43PM +, Ananyev, Konstantin wrote:
> Hi,
Hi
>
> > -Original Message-
> > From: Guduri Prathyusha [mailto:gprathyu...@caviumnetworks.com]
> > Sent: Thursday, November 2, 2017 2:31 PM
> > To: Kantecki, Tomasz
> > Cc: jianbo@arm.com; guduriprathyu...@gmai
Hi,
> -Original Message-
> From: Guduri Prathyusha [mailto:gprathyu...@caviumnetworks.com]
> Sent: Thursday, November 2, 2017 2:31 PM
> To: Kantecki, Tomasz
> Cc: jianbo@arm.com; guduriprathyu...@gmail.com; Ananyev, Konstantin
> ; dev@dpdk.org; Guduri
> Prathyusha
> Subject: [dpdk-d
With -f-strict-aliasing enabled by default from -O2, gcc > 5.x gives
undefined behavior in port_groupx4. 'pn' and 'pnum' are two different
pointers pointing to same chunk of memory and with -f-strict-aliasing the
pointers are assumed to be pointing to different memory and compiler
reorders instruct
> -Original Message-
> From: Jerin Jacob [mailto:jerin.ja...@caviumnetworks.com]
> Sent: Wednesday, November 1, 2017 11:12 PM
> To: Eads, Gage
> Cc: dev@dpdk.org; Richardson, Bruce ; Van
> Haaren, Harry ; Hemant Agrawal
> ; Nipun Gupta ; Rao,
> Nikhil ; Pavan Nikhilesh
> ; Thomas Monjalo
On Thu, Nov 02, 2017 at 02:52:16PM +0100, Gaëtan Rivet wrote:
> On Wed, Nov 01, 2017 at 08:12:38PM +, Ophir Munk wrote:
> > failsafe device has vlan stripping configured at startup however once
> > a sub device is found as non-capable of vlan-stripping failsafe
> > updates it configuration and
Hi,
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Yanglong Wu
> Sent: Thursday, November 2, 2017 5:05 PM
> To: dev@dpdk.org
> Cc: Wu, Yanglong
> Subject: [dpdk-dev] [PATCH 2/2] net/ixgbe: fix l3fwd start failed on PF
>
> which occurred when the SRIOV is acti
This reverts commit 284ae3e9ff9a92575c28c858efd2c85c8de6d440
("eal/ppc: fix mmap for memory initialization").The logic if the
address hint is respected on mmap is no different on POWER than
on x86 however the address layout of userspace applications might
be slightly different i.e. you might need t
On Wed, Nov 01, 2017 at 08:12:38PM +, Ophir Munk wrote:
> failsafe device has vlan stripping configured at startup however once
> a sub device is found as non-capable of vlan-stripping failsafe
> updates it configuration and removes vlan stripping from it.
> This update occurs only once at star
On Tue, Oct 31, 2017 at 06:21:28PM +, Matan Azrad wrote:
> Remove unlikely hint from WQ wraparound check because it is
> expected case.
>
> Fixes: c3c977bbecbd ("net/mlx4: add Tx bypassing Verbs")
>
> Signed-off-by: Matan Azrad
Acked-by: Adrien Mazarguil
--
Adrien Mazarguil
6WIND
On Tue, Oct 31, 2017 at 06:21:33PM +, Matan Azrad wrote:
> Replace most of the memory barriers by IO memory barriers since they
> are all targeted to the DRAM; This improves code efficiency for
> systems which force store order between different addresses.
>
> Only the doorbell register store
On Tue, Oct 31, 2017 at 06:21:31PM +, Matan Azrad wrote:
> Optimize single segment case by processing it in different block which
> prevents checks, calculations and barriers relevant only for multi
> segment case.
>
> Call a dedicated function for handling multi segments case.
>
> Signed-off
On Tue, Oct 31, 2017 at 06:21:32PM +, Matan Azrad wrote:
> Volatilize all Rx/Tx HW negotiation memories to be sure no compiler
> optimization prevents either load or store commands.
>
> Fixes: c3c977bbecbd ("net/mlx4: add Tx bypassing Verbs")
> Fixes: 9f57340a8087 ("net/mlx4: restore Rx offloa
On Tue, Oct 31, 2017 at 06:21:29PM +, Matan Azrad wrote:
> Merge tx_burst and mlx4_post_send functions to prevent
> double asking about WQ remain space.
>
> Signed-off-by: Matan Azrad
Ideally this patch and subsequent ones should have come after:
"net/mlx4: fix HW memory optimizations care
Hi Matan/Ophir,
On Tue, Oct 31, 2017 at 06:21:27PM +, Matan Azrad wrote:
> From: Ophir Munk
>
> Associate memory region to mempool (on data path) in a short function.
> Handle the less common case of adding a new memory region to mempool
> in a separate function.
>
> Signed-off-by: Ophir Mu
1 - 100 of 135 matches
Mail list logo
