On Mon, Nov 20, 2017 at 10:40:29AM -0800, Ferruh Yigit wrote:
> On 11/17/2017 12:47 AM, Pavan Nikhilesh Bhagavatula wrote:
> > On Thu, Nov 16, 2017 at 11:48:14AM -0800, Ferruh Yigit wrote:
> >> On 11/16/2017 2:31 AM, Pavan Nikhilesh wrote:
> >>> Dynamic logs fail to print if the global log level is
Hi Ferruh,
Tuesday, November 21, 2017 3:43 AM, Ferruh Yigit:
> To track modification:
> ce17eddefc20 ("ethdev: introduce Rx queue offloads API") cba7f53b717d
> ("ethdev: introduce Tx queue offloads API")
>
> Proposed deadline for PMDs is v18.05
The deprecation of the old offloads API is set to 1
On 11/20/2017 7:09 PM, Anoob Joseph wrote:
Hi
See inline.
On 20-11-2017 23:19, Radu Nicolau wrote:
Hi
On 11/20/2017 3:32 PM, Anoob wrote:
Hi,
Having something like "get_pkt_metadata" should be fine for inline
protocol usage. But we will still need a "get cookie" call to get
the cookie
20/11/2017 19:40, Ferruh Yigit:
> And, since you are touching these files :),
> for better regex support, it may be good to have log types hierarchical, like
> "lib.mbuf", "lib.mempool", "lib.eal.malloc", as done in member library but
> with
> "lib" prefix instead of "librte", if this makes sense
Hi David,
I forgot to add V2 to the topic (sorry for that).
But did you had time to look at V2 of this patch?
On 11 November 2017 at 19:55, Radoslaw Biernacki <
radoslaw.bierna...@linaro.org> wrote:
> This series of patches is fixing bug in power ACPI subsystem code as
> well as improve the deb
On Tue, Nov 21, 2017 at 11:55:00AM +0100, Thomas Monjalon wrote:
> 20/11/2017 19:40, Ferruh Yigit:
> > And, since you are touching these files :),
> > for better regex support, it may be good to have log types hierarchical,
> > like
> > "lib.mbuf", "lib.mempool", "lib.eal.malloc", as done in membe
21/11/2017 12:14, Pavan Nikhilesh Bhagavatula:
> On Tue, Nov 21, 2017 at 11:55:00AM +0100, Thomas Monjalon wrote:
> > 20/11/2017 19:40, Ferruh Yigit:
> > > And, since you are touching these files :),
> > > for better regex support, it may be good to have log types hierarchical,
> > > like
> > > "l
On Tue, Nov 21, 2017 at 02:30:48PM +0100, Thomas Monjalon wrote:
> 21/11/2017 12:14, Pavan Nikhilesh Bhagavatula:
> > On Tue, Nov 21, 2017 at 11:55:00AM +0100, Thomas Monjalon wrote:
> > > 20/11/2017 19:40, Ferruh Yigit:
> >
> > I will be sending out the patch set soon.
>
> These are only the lega
Hi all,
Here is a list of patches targeted for LTS release 16.11.4. Please
help review and test. The planned date for the final release is December the
1st.
Before that, please shout if anyone has objections with these patches being
applied.
These patches are located at branch 16.11 of dpdk-sta
Following the refactoring of the entire mlx4 PMD for DPDK 17.11, UDP RSS
support was left in a disabled state. Likewise, inner VXLAN RSS used to be
performed automatically by default but it's not the case anymore.
This series brings back both features by relying on the latest API updates
and bugfi
A couple of structure fields are not Doxygen-friendly.
Fixes: 5db1d364086e ("net/mlx4: restore Tx checksum offloads")
Cc: Moti Haimovsky
Cc: sta...@dpdk.org
Signed-off-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drive
Fixes: a2ce2121c01c ("net/mlx4: separate Tx configuration functions")
Cc: sta...@dpdk.org
Signed-off-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_txq.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/drivers/net/mlx4/mlx4_txq.c b/drivers/net/mlx4/mlx4_txq.c
index 7882a4d..17f3b00 100644
---
Supported RSS hash fields are listed in function mlx4_conv_rss_hf() and
duplicated in mlx4_flow_prepare(); the latter are used when RSS is
requested without specifying any parameters.
This commit standardizes on mlx4_conv_rss_hf().
Signed-off-by: Adrien Mazarguil
---
drivers/net/mlx4/mlx4_flow.
Until now, UDP RSS support could not be relied on due to a problem in the
Linux kernel implementation and mlx4 RSS capabilities were not reported at
all, hence the PMD had to make assumptions.
Since both issues will be addressed simultaneously in Linux 4.15 (related
patches already upstream) and l
Inner VXLAN RSS was supported and performed by default prior to the entire
mlx4 refactoring that occurred in DPDK 17.11, however so far the new Verbs
RSS API did not provide means to enable it. This will be addressed in
Linux 4.15 and in RDMA core.
Thanks to RSS capabilities, the PMD can now probe
On Tue, Nov 21, 2017 at 07:41:00AM +, Hu, Jiayu wrote:
> Hi Stephen,
>
> > -Original Message-
> > From: Stephen Hemminger [mailto:step...@networkplumber.org]
> > Sent: Tuesday, November 21, 2017 3:13 PM
> > To: Hu, Jiayu
> > Cc: dev@dpdk.org; y...@fridaylinux.org; Tan, Jianfeng
> >
>
On Tue, Nov 21, 2017 at 03:27:22PM +0100, Adrien Mazarguil wrote:
> Following the refactoring of the entire mlx4 PMD for DPDK 17.11, UDP RSS
> support was left in a disabled state. Likewise, inner VXLAN RSS used to be
> performed automatically by default but it's not the case anymore.
>
> This ser
On 11/21/2017 12:32 AM, Shahaf Shuler wrote:
> Hi Ferruh,
>
> Tuesday, November 21, 2017 3:43 AM, Ferruh Yigit:
>> To track modification:
>> ce17eddefc20 ("ethdev: introduce Rx queue offloads API") cba7f53b717d
>> ("ethdev: introduce Tx queue offloads API")
>>
>> Proposed deadline for PMDs is v18.
On 11/21/2017 6:04 AM, Pavan Nikhilesh Bhagavatula wrote:
> On Tue, Nov 21, 2017 at 02:30:48PM +0100, Thomas Monjalon wrote:
>> 21/11/2017 12:14, Pavan Nikhilesh Bhagavatula:
>>> On Tue, Nov 21, 2017 at 11:55:00AM +0100, Thomas Monjalon wrote:
20/11/2017 19:40, Ferruh Yigit:
>
>>>
>>> I will
Changing some part of the libraries but not updating all dependent code
cause maintenance problems.
The update in the contribution guide practically suggest updating all
dependent code.
With the counter argument that sometimes it is not possible to know
details of the dependent code and this prev
No config option changed, added or removed.
Only reshuffle PMD config options mostly to help new PMDs where to put
their new config option.
Ordered as physical, paravirtual and virtual groups. Alphabetical order
within a group.
Also tried to group vendor devices together which breaks alphabetical
> On Nov 20, 2017, at 7:35 AM, Nelio Laranjeiro
> wrote:
>
> Tx checksum offloads are correctly handled in a single Tx burst function
> whereas the capability is always set.
> This causes VXLAN packet with checksum offloads request to be ignored when
> the (E)MPS Tx functions are selected.
>
>
Hi Ferruh,
Not that it really mattered much, but I couldn't help but notice several
irregularities in the order of options that could have been avoided.
In particular, I believe it would make sense to put all xxx_DEBUG_xxx
options after the real control options, thus making it two classes of
optio
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> Adaptive Virtual Function (AVF) Driver is VF driver which supports
> for all future Intel devices without requiring a VM update.
> It promises the basic high speed connectivity. And since this happens
> to be an adaptive VF driver, every new drop of the V
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> Signed-off-by: Jingjing Wu
<...>
> #
> +# Compile burst-oriented AVF PMD driver
> +#
> +CONFIG_RTE_LIBRTE_AVF_PMD=y
Lets start PMD disabled and enable it after it become functional.
If you need to run a git bisect in the future on this commit, and y
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> enable device and queue setup ops like:
>
> - dev_configure
> - dev_start
> - dev_stop
> - dev_close
> - dev_infos_get
> - rx_queue_start
> - rx_queue_stop
> - tx_queue_start
> - tx_queue_stop
> - rx_queue_setup
> - rx_queue_release
> - tx_qu
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> Signed-off-by: Wenzhuo Lu
<...>
> @@ -214,6 +214,9 @@ CONFIG_RTE_LIBRTE_FM10K_INC_VECTOR=y
> # Compile burst-oriented AVF PMD driver
> #
> CONFIG_RTE_LIBRTE_AVF_PMD=y
> +CONFIG_RTE_LIBRTE_AVF_RX_DUMP=n
> +CONFIG_RTE_LIBRTE_AVF_TX_DUMP=n
Are these c
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
+ mtu_set which both commit log and patch title doesn't mention.
> Signed-off-by: Jingjing Wu
<...>
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> - promiscuous_enable
> - promiscuous_disable
> - allmulticast_enable
> - allmulticast_disable
> - mac_addr_add
> - mac_addr_remove
> - mac_addr_set
+ .vlan_filter_set= avf_dev_vlan_filter_set,
+ .vlan_offload_set
On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> - rxq_info_get
> - txq_info_get
> - rx_queue_count
> - rx_descriptor_done
> - rx_descriptor_status
> - tx_descriptor_status
+ some documentation.
<...>
> @@ -28,8 +28,8 @@
> (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE U
On 11/21/2017 3:17 PM, Aleksey Baulin wrote:
> Hi Ferruh,
>
> Not that it really mattered much, but I couldn't help but notice several
> irregularities in the order of options that could have been avoided.
We are on same boat, couldn't help on not noticing :)
> In particular, I believe it would
On Tue, 21 Nov 2017 16:06:24 -0800
Ferruh Yigit wrote:
> On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> > Signed-off-by: Wenzhuo Lu
>
> <...>
>
> > @@ -214,6 +214,9 @@ CONFIG_RTE_LIBRTE_FM10K_INC_VECTOR=y
> > # Compile burst-oriented AVF PMD driver
> > #
> > CONFIG_RTE_LIBRTE_AVF_PMD=y
> > +
> -Original Message-
> From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Wenzhuo Lu
> Sent: Monday, November 20, 2017 11:38 AM
> To: dev@dpdk.org
> Cc: Lu, Wenzhuo ; sta...@dpdk.org
> Subject: [dpdk-dev] [PATCH] net/ixgbe: fix VF RX hang
>
> The datasheet says, if using MSI-X mode, th
Thank you, Konstantin for your quick feedback.
Please see my inline comments below.
> -Original Message-
> From: Ananyev, Konstantin
> Sent: Monday, November 20, 2017 7:32 PM
> To: Dai, Wei ; Wu, Jingjing ;
> Xing, Beilei
> Cc: dev@dpdk.org; Dai, Wei
> Subject: RE: [dpdk-dev] [PATCH v3]
In virtio, Explicit Congestion Notification (ECN) includes two parts:
guest ECN and host ECN. Guest ECN means the frontend can handle TSO
packets which have ECN set, and host ECN means the backend can handle
TSO packets which have ECN set.
The ECN features are rarely used. However, virtio-net enab
Hi Everyone,
It's been some time i have been searching for DPDK architecture and
design for better understanding. So, finally i thought to get
confirmation on my queries.
I am working on a custom hypervisior similar to Qemu in Linux where
VNIC's operate in Para-virtualization like KVM/virtio-net
Adding support for inline protocol processing
In ingress side, application will receive regular IP packets, without
any IPsec related info. Application will do a selector check (SP-SA
check) by making use of the metadata from the packet.
In egress side, the plain packet would be submitted to the
In case of inline protocol processed ingress traffic, the packet may not
have enough information to determine the security parameters with which
the packet was processed. For such cases, application could register a
64 bit metadata in security session, which could be retrieved from the
packet using
The series adds inline protocol support in ipsec-secgw application.
First patch introduces changes in lib to enable applications to save a
64 bit metadata in security session. For inline processed packets,
application could call "rte_security_get_pkt_metadata" API to retrieve
this application regi
> -Original Message-
> From: Yigit, Ferruh
> Sent: Wednesday, November 22, 2017 8:06 AM
> To: Wu, Jingjing ; dev@dpdk.org
> Cc: Lu, Wenzhuo
> Subject: Re: [dpdk-dev] [RFC 4/9] net/avf: enable basic Rx Tx func
>
> On 10/20/2017 1:26 AM, Jingjing Wu wrote:
> > Signed-off-by: Wenzhuo Lu
>
40 matches
Mail list logo
