With vhost kernel, to enable multiqueue, we need backend device
in kernel support multiqueue feature. Specifically, with tap
as the backend, as linux/Documentation/networking/tuntap.txt shows,
we check if tap supports IFF_MULTI_QUEUE feature.
And for vhost kernel, each queue pair has a vhost fd, a
When used with vhost kernel backend, we can offload at both directions.
- From vhost kernel to virtio_user, the offload is enabled so that
DPDK app can trust the flow is checksum-correct; and if DPDK app
sends it through another port, the checksum needs to be
recalculated or offloaded
This patch add support vhost kernel as the backend for virtio_user.
Three main hook functions are added:
- vhost_kernel_setup() to open char device, each vq pair needs one
vhostfd;
- vhost_kernel_ioctl() to communicate control messages with vhost
kernel module;
- vhost_kernel_enable_q
To support vhost kernel as the backend of net_virtio_user in coming
patches, we move vhost_user specific structs and macros into
vhost_user.c, and only keep common definitions in vhost.h.
Besides, remove VHOST_USER_MQ feature check, it will be added back
in following multiqueue patch.
Signed-off-
Add a struct virtio_user_backend_ops to abstract three kinds of backend
operations:
- setup, create the unix socket connection;
- send_request, sync messages with backend;
- enable_qp, enable some queue pair.
Signed-off-by: Jianfeng Tan
---
drivers/net/virtio/virtio_user/vhost.h
In driver probe phase, we obtain device information; and then virtio
driver will initialize device and stores info, like negotiated
features, in vhost_user layer; finally, vhost_user gets DRIVER_OK
notification from virtio driver, and then sync with backend device.
Previously, DRIVER_OK could be s
Before the commit 86d59b21468a ("net/virtio: support LRO"), features
in virtio PMD, is decided and properly set at device initialization
and will not be changed. But afterward, features could be changed in
virtio_dev_configure(), and will be re-negotiated if it's changed.
In virtio_user, device fe
v2: (Lots of them are from yuanhan's comment)
- Add offloding feature.
- Add multiqueue support.
- Add a new patch to postpone the sending of driver ok notification.
- Put fix patch ahead of the whole patch series.
- Split original 0001 patch into 0003 and 0004 patches.
- Remove the ori
Hi, Yigit
> -Original Message-
> From: Yigit, Ferruh
> Sent: Wednesday, December 21, 2016 1:25 AM
> To: Zhao1, Wei ; dev@dpdk.org
> Cc: Lu, Wenzhuo
> Subject: Re: [dpdk-dev] [PATCH 16/18] net/ixgbe: create consistent filter
>
> On 12/2/2016 10:43 AM, Wei Zhao wrote:
> > From: wei zhao1
Hi,
I have used this lpm6 speed improvement patch for BGPv6 by *Nikita
Kozlov*. it works great for fixed prefix length however, when i use
variable prefix length for routes after 2 iterations of addition and
deletion of routes (3 approx) lpm6 crashes in expand_rule due to
stack overflow.
When bind with uio_pci_generic in XL710, the result is failed.
uio_pci_generic is not supported by XL710.
Signed-off-by: Jeff Guo
---
doc/guides/rel_notes/known_issues.rst | 22 ++
1 file changed, 22 insertions(+)
diff --git a/doc/guides/rel_notes/known_issues.rst
b/doc/gui
On Thu, Dec 22, 2016 at 03:45:35PM +, Ferruh Yigit wrote:
> On 12/13/2016 1:08 AM, Michał Mirosław wrote:
> > Signed-off-by: Michał Mirosław
> > ---
> > drivers/net/i40e/i40e_ethdev.c | 5 +++--
> > lib/librte_eal/linuxapp/eal/eal_interrupts.c | 2 +-
> > 2 files changed, 4 inse
On Mon, Dec 19, 2016 at 06:48:52PM +, Ananyev, Konstantin wrote:
> Hi Michal,
>
> > -Original Message-
> > From: Michał Mirosław [mailto:mirq-li...@rere.qmqm.pl]
> > Sent: Wednesday, December 14, 2016 5:24 PM
> > To: dev@dpdk.org
> > Cc: Ananyev, Konstantin
> > Subject: [PATCH] acl: r
Hi Santosh,
On 22 December 2016 at 20:36, Santosh Shukla
wrote:
> Hi Jiangbo,
>
> On Thu, Dec 15, 2016 at 08:40:19PM -0800, Santosh Shukla wrote:
>> On Thu, Dec 15, 2016 at 04:37:12PM +0800, Jianbo Liu wrote:
>> > On 14 December 2016 at 09:55, Jerin Jacob
>> > wrote:
>> > > From: Santosh Shukla
Yes. One core is assigned.
On Fri, 23 Dec 2016 at 9:34 AM, Xu, Qian Q wrote:
> As far as I know, L2FWD only uses 1 core for all RX/TX, for all queues,
> but for ip_pipeline, you may use more cores.
>
> A simple question, are you using 1core in ip_pipeline or l3fwd test?
>
>
>
> > -Original M
As far as I know, L2FWD only uses 1 core for all RX/TX, for all queues, but for
ip_pipeline, you may use more cores.
A simple question, are you using 1core in ip_pipeline or l3fwd test?
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Royce Niu
> Sent: Thursda
Add a check and fail the VF's probe request if VF is already in
VF_ACQUIRED state.
Fixes: 22d07d939c3c ("net/qede/base: update")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/base/ecore_sriov.c | 12
1 file changed, 12 insertions(+)
diff --git a/drivers/net/qede/base/ecore_sri
Apparently VF over Legacy PF doesn't work, as VF would fail after
getting the initial rejection message [instead of sending an additional
one where it asks for a FW override and see if it works].
Fixes: 22d07d939c3c ("net/qede/base: update")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/base/
Fix to return error code ECORE_INVAL instead of 0 when EQ elements
is too large as done elsewhere in this function.
Fixes: 22d07d939c3c ("net/qede/base: update")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/base/ecore_dev.c |1 +
1 file changed, 1 insertion(+)
diff --git a/drivers/net/q
Fix OSAL_MUTEX_DEALLOC() in freeing the context manager.
Fixes: 22d07d939c3c ("net/qede/base: update")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/base/ecore_cxt.c |2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/net/qede/base/ecore_cxt.c
b/drivers/net/qede/ba
Rx queue access is still done prior to the index being validated by PF.
Hence move Rx queue and status block validation check before accessing
Rx queue to prevent malicious VFs from using out-of-bound queue indices.
Fixes: 98bc693e1938 ("net/qede/base: change queue start")
Signed-off-by: Rasesh M
Remove unmeaningful function ID value in print.
Don't print the number of Multicast filters as part of Acquire response,
as this is an obsolete field which isn't enforced by PF.
Fixes: 86a2265e59d7 ("qede: add SRIOV support")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/base/ecore_sriov.c |
There are certain conditions under which VF would infinitely send
ACQUIRE messages, as it will fail to understand that PF has rejected
the ACQUIRE request. Fix to reject multiple acquisition requests by VF.
Fixes: 22d07d939c3c ("net/qede/base: update")
Signed-off-by: Rasesh Mody
---
drivers/net
The vendor_id and device_id are used to determine device type. If you
don't have them, then check for determining device type fails and is
always set to default device type.
Fixes: ec94dbc57362 ("qede: add base driver")
Signed-off-by: Rasesh Mody
---
drivers/net/qede/qede_ethdev.c |4
qede PMD does not support SR-IOV PF driver functionality, so
max_vfs is set to 0 to imply the same.
Fixes: 2ea6f76aff40 ("qede: add core driver")
Signed-off-by: Harish Patil
---
drivers/net/qede/qede_ethdev.c |5 +
1 file changed, 1 insertion(+), 4 deletions(-)
diff --git a/drivers/net
This patch includes the following:
- Fix missing hash_key_size advertisement
- Fix RSS hash query function
- Update RSS offload flag
- Accept user provided RSS configuration params via rx_adv_conf
in dev_configure()
- Decouple RSS configuration from common qed_update_vport() and
instead make u
HW can support upto 128 queues based on the NIC config/personality.
But most of the testing is done with 32 queues for PF and 16 for VF
device across different qede devices, so change here is to advertise
only those many instead of returning max queues supported by HW.
Fixes: 2ea6f76aff40 ("qede:
- Support HW checksum and RSS offload for VXLAN traffic
- Identify inner/outer packet_types using lookup table
- Update documentation
Signed-off-by: Harish Patil
---
doc/guides/nics/features/qede.ini |3 +
doc/guides/nics/qede.rst |3 +-
drivers/net/qede/qede_eth_if.h|3
Move DP_NOTICE msg under CONFIG_RTE_LIBRTE_QEDE_DEBUG_INFO.
Signed-off-by: Harish Patil
---
drivers/net/qede/qede_logs.h | 13 -
1 file changed, 4 insertions(+), 9 deletions(-)
diff --git a/drivers/net/qede/qede_logs.h b/drivers/net/qede/qede_logs.h
index 45c4af0..13b76a7 100644
-
In qede_mac_addr_add() a check is added to differentiate between
unicast/multicast mac to prevent a multicast mac from being wrongly added
to unicast filter table. Secondly, two separate lists will be used to keep
track of unicast/multicast mac filters to prevent duplicate filter
programming. The o
- Enable/disable VXLAN tunneling
- Add/remove VXLAN classification rules
- Destination UDP port configuration
Signed-off-by: Harish Patil
---
drivers/net/qede/base/ecore_l2.c |3 +
drivers/net/qede/base/ecore_l2_api.h |1 +
drivers/net/qede/qede_ethdev.c | 400
On Thu, Dec 22, 2016 at 4:13 PM, Thomas Monjalon
wrote:
> 2016-12-21 16:09, Jan Blunck:
>> The struct rte_intr_handle is an abstraction layer for different types of
>> interrupt mechanisms. It is embedded in the low-level device (e.g. PCI).
>> On allocation of a struct rte_eth_dev a reference to t
On Thu, Dec 22, 2016 at 4:21 PM, Thomas Monjalon
wrote:
> 2016-12-21 16:09, Jan Blunck:
>> PCI drivers could use this helper instead of directly accessing fields of
>> rte_eth_dev to map to rte_pci_device.
> [...]
>> +/**
>> + * @internal
>> + * Helper for drivers that need to convert from rte_eth
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Tomasz Kulasek
> Sent: Thursday, December 22, 2016 5:05 AM
> To: dev@dpdk.org
> Cc: Ananyev, Konstantin
> Subject: [dpdk-dev] [PATCH v14 6/8] vmxnet3: add Tx preparation
>
> From: "Ananyev, Konstantin"
>
> Signe
On 12/21/2016 8:29 AM, Wenzhuo Lu wrote:
> When VF receives a message from PF, it should check the return
> value. But in i40evf_execute_vf_cmd the value is ignored and not
> returned to the caller.
>
> Fixes: 95cd21f45d1b ("i40evf: allocate virtchnl commands buffer per VF")
>
> Signed-off-by: We
In block cipher test cases, add checks that the source
and destination mbufs are not modified except where expected.
Signed-off-by: Fiona Trahe
---
app/test/test_cryptodev_blockcipher.c | 139 --
1 file changed, 134 insertions(+), 5 deletions(-)
diff --git a/app/
> -Original Message-
> From: Lu, Wenzhuo
> Sent: Wednesday, December 21, 2016 12:56 AM
> To: Iremonger, Bernard ; Yigit, Ferruh
> ; dev@dpdk.org
> Cc: Wu, Jingjing ; Zhang, Helin
> ; Zhang, Qi Z ; Chen, Jing D
>
> Subject: RE: [dpdk-dev] [PATCH v5 29/29] net/i40e: set/clear VF stats from
Thanks for the review, Jerin, I very much appreciate it. I'll address
all the minor comments, and I've a comment or two on the remaining
changes below.
On 22/12/2016 12:47 PM, Jerin Jacob wrote:
On Thu, Dec 22, 2016 at 04:37:04AM +, David Hunt wrote:
--snip--
+
+ /* set the
> -Original Message-
> From: Kusztal, ArkadiuszX
> Sent: Thursday, December 22, 2016 10:15 AM
> To: dev@dpdk.org
> Cc: Trahe, Fiona ; De Lara Guarch, Pablo
> ; Griffin, John ;
> Jain, Deepak K ; Kusztal, ArkadiuszX
>
> Subject: [PATCH v2 1/2] crypto/qat: add SGL capability to Intel Quick
On 12/13/2016 1:08 AM, Michał Mirosław wrote:
> Signed-off-by: Michał Mirosław
> ---
> drivers/net/i40e/i40e_ethdev.c | 5 +++--
> lib/librte_eal/linuxapp/eal/eal_interrupts.c | 2 +-
> 2 files changed, 4 insertions(+), 3 deletions(-)
>
> diff --git a/drivers/net/i40e/i40e_ethdev.c
2016-12-21 16:09, Jan Blunck:
> This is a partial merge of Stephens and my patches to make the rte_eth_dev
> independent of the rte_pci_device.
Thanks for the hard work.
> Jan Blunck (17):
> eal: define container_of macro
> eal: Allow passing const rte_intr_handle
> virtio: Don't fill dev_i
2016-12-22 15:05, Ferruh Yigit:
> On 12/22/2016 2:47 PM, Thomas Monjalon wrote:
> > 2016-12-22 14:36, Ferruh Yigit:
> >> On 12/22/2016 11:07 AM, Thomas Monjalon wrote:
> >>> I think it is OK to add a new dev_ops and a new API function for firmware
> >>> query. Generally speaking, it is a good thing
2016-12-22 15:16, Ferruh Yigit:
> On 12/22/2016 3:10 PM, Jan Blunck wrote:
> > On Thu, Dec 22, 2016 at 2:10 PM, Ferruh Yigit
> > wrote:
> >> - Grouped related items using empty lines
> >
> > Reordering fields of a struct is breaking ABI. We should bump the
> > library version now.
>
> You are r
2016-12-21 16:09, Jan Blunck:
> +/**
> + * @internal
> + * Helper macro for drivers that need to convert to struct rte_pci_device.
> + */
> +#define DEV_PCI_DEV(ptr) container_of(ptr, struct rte_pci_device, device)
Please use RTE_ prefix. RTE_PCI_DEV?
2016-12-21 16:09, Jan Blunck:
> PCI drivers could use this helper instead of directly accessing fields of
> rte_eth_dev to map to rte_pci_device.
[...]
> +/**
> + * @internal
> + * Helper for drivers that need to convert from rte_eth_dev to
> rte_pci_device.
> + */
> +static inline struct rte_pci_
On Thu, Dec 22, 2016 at 2:10 PM, Ferruh Yigit wrote:
> - Grouped related items using empty lines
Reordering fields of a struct is breaking ABI. We should bump the
library version now.
> - Aligned arguments to same column
> - All item comments that doesn't fit same line are placed blow the item
On 12/22/2016 3:10 PM, Jan Blunck wrote:
> On Thu, Dec 22, 2016 at 2:10 PM, Ferruh Yigit wrote:
>> - Grouped related items using empty lines
>
> Reordering fields of a struct is breaking ABI. We should bump the
> library version now.
You are right, sorry missed that.
Intention was not to break t
2016-12-21 16:09, Jan Blunck:
> The struct rte_intr_handle is an abstraction layer for different types of
> interrupt mechanisms. It is embedded in the low-level device (e.g. PCI).
> On allocation of a struct rte_eth_dev a reference to the intr_handle
> should be stored for devices supporting inter
On 12/22/2016 2:47 PM, Thomas Monjalon wrote:
> 2016-12-22 14:36, Ferruh Yigit:
>> On 12/22/2016 11:07 AM, Thomas Monjalon wrote:
>>> I think it is OK to add a new dev_ops and a new API function for firmware
>>> query. Generally speaking, it is a good thing to avoid putting all
>>> informations in
2016-12-22 13:10, Ferruh Yigit:
> - Grouped related items using empty lines
> - Aligned arguments to same column
> - All item comments that doesn't fit same line are placed blow the item
> itself
> - Moved some comments to same line if overall line < 100 chars
>
> Signed-off-by: Ferruh Yigit
A
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
MAINTAINERS | 6 ++
1 file changed, 6 insertions(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index 2f072b5..2ced843 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -457,6 +457,12 @@ M: Declan Doherty
F: drivers/crypto/null/
F: doc/gu
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
app/test/test_cryptodev_perf.c | 12
1 file changed, 12 insertions(+)
diff --git a/app/test/test_cryptodev_perf.c b/app/test/test_cryptodev_perf.c
index 59a6891..2d9b78e 100644
--- a/app/test/test_cryptodev_perf.c
+++ b/app
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
drivers/crypto/Makefile| 1 +
drivers/crypto/dpaa2_sec/Makefile | 74 ++
.../crypto/dpaa2_sec/rte_pmd_dpaa2_sec_version.map | 4 ++
mk/rte.app.mk
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
drivers/crypto/dpaa2_sec/dpaa2_sec_dpseci.c | 76 +
1 file changed, 76 insertions(+)
diff --git a/drivers/crypto/dpaa2_sec/dpaa2_sec_dpseci.c
b/drivers/crypto/dpaa2_sec/dpaa2_sec_dpseci.c
index 6c9895f..f4a4
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
config/defconfig_arm64-dpaa2-linuxapp-gcc | 12
1 file changed, 12 insertions(+)
diff --git a/config/defconfig_arm64-dpaa2-linuxapp-gcc
b/config/defconfig_arm64-dpaa2-linuxapp-gcc
index 18c9589..30fd4e3 100644
--- a/config
Signed-off-by: Akhil Goyal
Signed-off-by: Hemant Agrawal
---
drivers/bus/fslmc/fslmc_vfio.c |3 +-
drivers/bus/fslmc/rte_fslmc.h |5 +-
drivers/common/dpaa2/dpio/dpaa2_hw_pvt.h| 25 +
drivers/crypto/dpaa2_sec/dpaa2_sec_dpseci.c | 1592
Signed-off-by: Akhil Goyal
---
lib/librte_cryptodev/rte_cryptodev.h | 3 +++
1 file changed, 3 insertions(+)
diff --git a/lib/librte_cryptodev/rte_cryptodev.h
b/lib/librte_cryptodev/rte_cryptodev.h
index bb5f41c..bacf893 100644
--- a/lib/librte_cryptodev/rte_cryptodev.h
+++ b/lib/librte_cryptod
Signed-off-by: Akhil Goyal
Reviewed-by: Hemant Agrawal
---
doc/guides/cryptodevs/dpaa2_sec.rst | 233
doc/guides/cryptodevs/index.rst | 1 +
2 files changed, 234 insertions(+)
create mode 100644 doc/guides/cryptodevs/dpaa2_sec.rst
diff --git a/doc/gui
Based over the DPAA2 PMD driver [1], this series of patches introduces the
DPAA2_SEC PMD which provides DPDK crypto driver for NXP's DPAA2 CAAM
Hardware accelerator.
SEC is NXP DPAA2 SoC's security engine for cryptographic acceleration and
offloading. It implements block encryption, stream cipher,
This patch will not be required as some parallel work is going
on to add it across all crypto devices.
Signed-off-by: Akhil Goyal
---
lib/librte_cryptodev/rte_cryptodev.h | 1 +
1 file changed, 1 insertion(+)
diff --git a/lib/librte_cryptodev/rte_cryptodev.h
b/lib/librte_cryptodev/rte_cryptode
2016-12-22 12:00, Ferruh Yigit:
> Fixes: 9d41beed24b0 ("lib: provide initial versioning")
>
> Signed-off-by: Ferruh Yigit
Applied, thanks
2016-12-22 14:36, Ferruh Yigit:
> On 12/22/2016 11:07 AM, Thomas Monjalon wrote:
> > I think it is OK to add a new dev_ops and a new API function for firmware
> > query. Generally speaking, it is a good thing to avoid putting all
> > informations in the same structure (e.g. rte_eth_dev_info).
>
>
Hi Ferruh,
> -Original Message-
> From: Yigit, Ferruh
> Sent: Thursday, December 22, 2016 10:57 AM
> To: Lu, Wenzhuo ; dev@dpdk.org
> Cc: Iremonger, Bernard
> Subject: Re: [dpdk-dev] [PATCH v6 23/25] app/testpmd: handle i40e in VF
> VLAN filter command
>
> On 12/21/2016 6:34 AM, Wenzhuo
On 12/22/2016 11:07 AM, Thomas Monjalon wrote:
> 2016-12-08 16:34, Remy Horton:
>>
>> On 06/12/2016 15:16, Qiming Yang wrote:
>> [..]
>>> Qiming Yang (5):
>>> ethdev: add firmware version get
>>> net/e1000: add firmware version get
>>> net/ixgbe: add firmware version get
>>> net/i40e: add f
2016-12-22 14:05, Tomasz Kulasek:
> Since all current drivers supports Tx preparation API, it is used
> in csum forwarding engine by default for all drivers.
[...]
> +/*
> + * Enable Tx preparation path in the "csum" engine.
> + */
> +uint8_t tx_prepare;
It seems this variable is not used.
Hi Tomasz,
2016-12-22 14:05, Tomasz Kulasek:
> Added API for `rte_eth_tx_prepare`
>
> uint16_t rte_eth_tx_prepare(uint8_t port_id, uint16_t queue_id,
> struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
As discussed earlier and agreed by Konstantin, please mark this API
as experimental.
We could
>
> 2016-12-12 11:51, Ananyev, Konstantin:
> > > > The application gets few information from tx_prepare() about what should
> > > > be done to make the packet accepted by the hw, and the actions will
> > > > probably be different depending on hardware.
> >
> > That's true.
> > I am open to suggest
Le 22/12/2016 à 09:10, Chen, Jing D a écrit :
In the meanwhile, we have some test models ongoing to validate combination of
Linux and
DPDK drivers for VF and PF. We'll fully support below 4 cases going forward.
1. DPDK PF + DPDK VF
2. DPDK PF + Linux VF
+ DPDK PF + FreeBSD VF
+ DPDK PF + Windo
On Thu, Dec 22, 2016 at 02:14:45PM +0100, Thomas Monjalon wrote:
> 2016-12-02 00:10, Ananyev, Konstantin:
> > I have absolutely no problem to remove the RTE_ETHDEV_TX_PREPARE and
> > associated logic.
> > I personally don't use ARM boxes and don't plan to,
> > and in theory users can still do cond
Dear Bruce,
Thanks for your kind explanation.
I will try to follow your suggestion and see the source code.
On Thu, Dec 22, 2016 at 9:25 PM, Bruce Richardson <
bruce.richard...@intel.com> wrote:
> On Thu, Dec 22, 2016 at 08:48:50PM +0800, Royce Niu wrote:
> > But, actually, L3FWD of IP_PIPELINE
2016-12-12 11:51, Ananyev, Konstantin:
> > > The application gets few information from tx_prepare() about what should
> > > be done to make the packet accepted by the hw, and the actions will
> > > probably be different depending on hardware.
>
> That's true.
> I am open to suggestions how in futu
On Thu, Dec 22, 2016 at 08:48:50PM +0800, Royce Niu wrote:
> But, actually, L3FWD of IP_PIPELINE is also faster than stock L2FWD, which
> also modifies mac addr. How can explain this?
>
> Actually, I want to know why IP_PIPELINE is much faster and I can learn
> from IP_PIPELINE and make our own pr
2016-12-02 00:10, Ananyev, Konstantin:
> I have absolutely no problem to remove the RTE_ETHDEV_TX_PREPARE and
> associated logic.
> I personally don't use ARM boxes and don't plan to,
> and in theory users can still do conditional compilation at the upper layer,
> if they want to.
Yes you're ri
2016-12-20 13:36, Ferruh Yigit:
> On 12/13/2016 5:41 PM, Tomasz Kulasek wrote:
> > From: "Ananyev, Konstantin"
> >
> > Signed-off-by: Konstantin Ananyev
> > ---
>
> <...>
>
> >
> > uint16_t
> > +vmxnet3_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts,
> > + uint16_t nb_pk
- Grouped related items using empty lines
- Aligned arguments to same column
- All item comments that doesn't fit same line are placed blow the item
itself
- Moved some comments to same line if overall line < 100 chars
Signed-off-by: Ferruh Yigit
---
- ! This patch has the problem of trashing
From: Maple
To:
Cc:
Subject: [PATCH] Load_balancer Tx Flush Bug
Date: Thu, 22 Dec 2016 09:57:48 +0800
Message-Id: <1482371868-19669-1-git-send-email-liuj...@raisecom.com>
X-Mailer: git-send-email 1.9.1
In-Reply-To: <2016122122394164225...@raisecom.com>
References: <2016122122394164225...@raiseco
From: Konstantin Ananyev
Signed-off-by: Konstantin Ananyev
---
drivers/net/ena/ena_ethdev.c | 51 ++
1 file changed, 51 insertions(+)
diff --git a/drivers/net/ena/ena_ethdev.c b/drivers/net/ena/ena_ethdev.c
index 555fb31..51af723 100644
--- a/drivers/n
Signed-off-by: Tomasz Kulasek
Acked-by: Konstantin Ananyev
---
drivers/net/ixgbe/ixgbe_ethdev.c |3 ++
drivers/net/ixgbe/ixgbe_ethdev.h |5 +++-
drivers/net/ixgbe/ixgbe_rxtx.c | 57 ++
drivers/net/ixgbe/ixgbe_rxtx.h |2 ++
4 files changed, 66
Since all current drivers supports Tx preparation API, it is used
in csum forwarding engine by default for all drivers.
Adding additional step to the csum engine costs about 3-4% of performance
drop, on my setup with ixgbe driver. It's caused mostly by the need
of reaccessing and modification of p
From: "Ananyev, Konstantin"
Signed-off-by: Konstantin Ananyev
---
drivers/net/vmxnet3/vmxnet3_ethdev.c |6
drivers/net/vmxnet3/vmxnet3_ethdev.h |2 ++
drivers/net/vmxnet3/vmxnet3_rxtx.c | 56 ++
3 files changed, 64 insertions(+)
diff --git a/dri
Signed-off-by: Tomasz Kulasek
Acked-by: Konstantin Ananyev
---
drivers/net/fm10k/fm10k.h|6 +
drivers/net/fm10k/fm10k_ethdev.c |5
drivers/net/fm10k/fm10k_rxtx.c | 50 +-
3 files changed, 60 insertions(+), 1 deletion(-)
diff --git
Signed-off-by: Tomasz Kulasek
Acked-by: Konstantin Ananyev
---
drivers/net/i40e/i40e_ethdev.c |3 ++
drivers/net/i40e/i40e_rxtx.c | 74 +++-
drivers/net/i40e/i40e_rxtx.h |8 +
3 files changed, 84 insertions(+), 1 deletion(-)
diff --git a/dri
Signed-off-by: Tomasz Kulasek
Acked-by: Konstantin Ananyev
---
drivers/net/e1000/e1000_ethdev.h | 11
drivers/net/e1000/em_ethdev.c|5 +++-
drivers/net/e1000/em_rxtx.c | 48 +-
drivers/net/e1000/igb_ethdev.c |4 +++
drivers/net/e100
Added API for `rte_eth_tx_prepare`
uint16_t rte_eth_tx_prepare(uint8_t port_id, uint16_t queue_id,
struct rte_mbuf **tx_pkts, uint16_t nb_pkts)
Added fields to the `struct rte_eth_desc_lim`:
uint16_t nb_seg_max;
/**< Max number of segments per whole packet. */
As discussed in that thread:
http://dpdk.org/ml/archives/dev/2015-September/023603.html
Different NIC models depending on HW offload requested might impose
different requirements on packets to be TX-ed in terms of:
- Max number of fragments per packet allowed
- Max number of fragments per TSO
But, actually, L3FWD of IP_PIPELINE is also faster than stock L2FWD, which
also modifies mac addr. How can explain this?
Actually, I want to know why IP_PIPELINE is much faster and I can learn
from IP_PIPELINE and make our own program.
But, the documentation of that is not detailed enough. if it
On Wed, Dec 21, 2016 at 05:19:16PM +0100, Simon Horman wrote:
> On Fri, Dec 16, 2016 at 05:24:57PM +0100, Adrien Mazarguil wrote:
> > As previously discussed in RFC v1 [1], RFC v2 [2], with changes
> > described in [3] (also pasted below), here is the first non-draft series
> > for this new API.
>
On Thu, Dec 22, 2016 at 04:37:04AM +, David Hunt wrote:
> Now sends bursts of up to 8 mbufs to each worker, and tracks
> the in-flight flow-ids (atomic scheduling)
>
> New file with a new api, similar to the old API except with _burst
> at the end of the function names
>
> Signed-off-by: Davi
2016-12-22 11:53, Ferruh Yigit:
> To make it easy to comment to latest struct, copy-paste here:
> [With some extra notes]
>
> struct eth_dev_ops {
> eth_dev_configure_tdev_configure; /**< Configure device. */
> eth_dev_start_tdev_start; /**< Start device. */
>
Hi Jiangbo,
On Thu, Dec 15, 2016 at 08:40:19PM -0800, Santosh Shukla wrote:
> On Thu, Dec 15, 2016 at 04:37:12PM +0800, Jianbo Liu wrote:
> > On 14 December 2016 at 09:55, Jerin Jacob
> > wrote:
> > > From: Santosh Shukla
> > >
> >
> > memory barrier operation is put inside IXGBE_PCI_REG_READ/W
On Thu, Dec 22, 2016 at 04:37:06AM +, David Hunt wrote:
> Signed-off-by: David Hunt
> ---
> + * it does nothing but return packets and count them.
> + */
> +static int
> +handle_work_burst(void *arg)
> +{
> + //struct rte_mbuf *pkt = NULL;
Seems like their is lot test code with // in this
On 12/22/2016 11:32 AM, Andrew Rybchenko wrote:
> Fixes: 0880c40113ef ("drivers: advertise kmod dependencies in pmdinfo")
>
> Signed-off-by: Andrew Rybchenko
Applied to dpdk-next-net/master, thanks.
On 12/22/2016 03:07 PM, Ferruh Yigit wrote:
On 12/22/2016 11:35 AM, Andrew Rybchenko wrote:
On 12/22/2016 02:04 PM, Ferruh Yigit wrote:
On 12/21/2016 11:40 AM, Andrew Rybchenko wrote:
On 12/21/2016 02:37 PM, Neil Horman wrote:
On Wed, Dec 21, 2016 at 12:21:14PM +0300, Andrew Rybchenko wrote:
On 12/22/2016 11:35 AM, Andrew Rybchenko wrote:
> On 12/22/2016 02:04 PM, Ferruh Yigit wrote:
>> On 12/21/2016 11:40 AM, Andrew Rybchenko wrote:
>>> On 12/21/2016 02:37 PM, Neil Horman wrote:
On Wed, Dec 21, 2016 at 12:21:14PM +0300, Andrew Rybchenko wrote:
> On 12/20/2016 08:26 PM, Thomas
Fixes: 9d41beed24b0 ("lib: provide initial versioning")
Signed-off-by: Ferruh Yigit
---
lib/librte_ether/rte_ether_version.map | 1 -
1 file changed, 1 deletion(-)
diff --git a/lib/librte_ether/rte_ether_version.map
b/lib/librte_ether/rte_ether_version.map
index 209328c..b7ee80c 100644
--- a/l
- Grouped related items using empty lines
- Aligned arguments to same column
- All item comments that doesn't fit same line are placed blow the item
itself
- Moved some comments to same line if overall line < 100 chars
Signed-off-by: Ferruh Yigit
---
- ! This patch has the problem of trashing
Signed-off-by: David Hunt
---
doc/guides/prog_guide/packet_distrib_lib.rst | 4
1 file changed, 4 insertions(+)
diff --git a/doc/guides/prog_guide/packet_distrib_lib.rst
b/doc/guides/prog_guide/packet_distrib_lib.rst
index b5bdabb..dffd4ad 100644
--- a/doc/guides/prog_guide/packet_distrib_
Signed-off-by: David Hunt
---
app/test/test_distributor_perf.c | 133 +--
1 file changed, 127 insertions(+), 6 deletions(-)
diff --git a/app/test/test_distributor_perf.c b/app/test/test_distributor_perf.c
index 7947fe9..86285fd 100644
--- a/app/test/test_distr
Signed-off-by: David Hunt
---
examples/distributor/main.c | 505 ++--
1 file changed, 388 insertions(+), 117 deletions(-)
diff --git a/examples/distributor/main.c b/examples/distributor/main.c
index e7641d2..451e253 100644
--- a/examples/distributor/main.c
Signed-off-by: David Hunt
---
app/test/test_distributor.c | 500 ++--
1 file changed, 391 insertions(+), 109 deletions(-)
diff --git a/app/test/test_distributor.c b/app/test/test_distributor.c
index 85cb8f3..7738f04 100644
--- a/app/test/test_distributor.c
1 - 100 of 120 matches
Mail list logo
