Hi Thomas,
Please consider pulling following virtio changes for 17.11-rc1 at
git://dpdk.org/next/dpdk-next-virtiomaster
Thanks.
--yliu
---
Daniel Mrzyglod (1):
net/virtio: fix of untrusted scalar value
Jay Zhou (1):
net/virtio: fix a typo
Maxime Coquelin (19):
On 09/11/2017 05:13 PM, Olivier Matz wrote:
The compilation with gcc-6.3.0 and EXTRA_CFLAGS=-Og gives the following
error:
CC virtio_rxtx.o
virtio_rxtx.c: In function ‘virtio_rx_offload’:
virtio_rxtx.c:680:10: error: ‘csum’ may be used uninitialized in
this fu
On Thu, Oct 05, 2017 at 06:29:12PM +0100, Ferruh Yigit wrote:
> On 10/5/2017 9:43 AM, Tomasz Duszynski wrote:
> > On Wed, Oct 04, 2017 at 05:59:11PM +0100, Ferruh Yigit wrote:
> >> On 10/4/2017 9:59 AM, Tomasz Duszynski wrote:
> >>> On Wed, Oct 04, 2017 at 01:24:27AM +0100, Ferruh Yigit wrote:
> >>
On Sat, Sep 23, 2017 at 08:31:37PM +, Jan Scheurich wrote:
...
> +int rte_vhost_tx_interrupt_requested(int vid, uint16_t qid)
> +{
> + struct virtio_net *dev;
> + struct vhost_virtqueue *vq;
> +
> + dev = get_device(vid);
> + if (dev == NULL)
> + return 0;
> +
> +
On Thu, Oct 05, 2017 at 09:31:25AM +, Zhang, Roy Fan wrote:
>
>
> > -Original Message-
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Zhiyong Yang
> > Sent: Friday, August 11, 2017 3:13 AM
> > To: dev@dpdk.org
> > Cc: maxime.coque...@redhat.com; y...@fridaylinux.org; Yang, Z
On Mon, Sep 04, 2017 at 01:26:00PM +, Rybalchenko, Kirill wrote:
>
>
> > -Original Message-
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Jay Zhou
> > Sent: Tuesday 22 August 2017 03:35
> > To: dev@dpdk.org; y...@fridaylinux.org; maxime.coque...@redhat.com
> > Cc: weidong.h
Series applied to dpdk-next-virtio.
Thanks.
--yliu
On Thu, Oct 05, 2017 at 10:36:08AM +0200, Maxime Coquelin wrote:
> This v3 lists the feature in the release note, and fixes the bug in
> is_vring_iotlb_update() reported by Yuanhan.
>
> The purpose of this series is to add support for
>
On Thu, Oct 05, 2017 at 02:50:02PM +0200, Nelio Laranjeiro wrote:
> Isolated works exclusively with the generic flow API, this patch adds a new
> set of operations valid in this mode.
>
> - promiscuous*()
> - allmulticast*()
> - reta*()
> - rss*()
>
> are not supported in this mode as it is f
On Thu, Oct 05, 2017 at 02:50:01PM +0200, Nelio Laranjeiro wrote:
> Support same functionalities as in
> commit cf521eaa3c76 ("net/mlx5: remove flow director support")
>
> This implementation is done on top of the generic flow API.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
On Thu, Oct 05, 2017 at 02:49:59PM +0200, Nelio Laranjeiro wrote:
> Moves ibv_attr containing the specification of the flow from Verbs point of
> view also with the verbs flow itself near the related verbs objects making
> the flow.
>
> This is also a preparation to handle correctly the RSS hash c
On Thu, Oct 05, 2017 at 02:49:58PM +0200, Nelio Laranjeiro wrote:
> Move mlx5_flow_validate/create/flush/isolate() to the end of the file.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On Thu, Oct 05, 2017 at 02:49:57PM +0200, Nelio Laranjeiro wrote:
> struct mlx5_flow_parse was commonly used with the name "flow" confusing
> sometimes the development. The variable name is replaced by parser to
> reflect its use.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
On Thu, Oct 05, 2017 at 02:49:56PM +0200, Nelio Laranjeiro wrote:
> struct mlx5_flow_parse now embed fields from struct mlx5_flow_action.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On Thu, Oct 05, 2017 at 02:49:55PM +0200, Nelio Laranjeiro wrote:
> In case the pattern contains an RSS actions, the RSS configuration to use
> is the one provided by the user. To make the correct conversion from DPDK
> RSS hash fields to Verbs ones according to the users requests the actions
> mu
On Thu, Oct 05, 2017 at 02:49:54PM +0200, Nelio Laranjeiro wrote:
> Validation of flows is only making few verifications on the pattern, in
> some situation the validate action could end by with success whereas the
> pattern could not be converted correctly.
>
> This brings this conversion verific
On Thu, Oct 05, 2017 at 02:49:53PM +0200, Nelio Laranjeiro wrote:
> From this commit the RSS support becomes un-available until it is replaced
> by the generic flow implementation.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On Thu, Oct 05, 2017 at 02:49:52PM +0200, Nelio Laranjeiro wrote:
> Since RSS configuration can also be used by flow API, there is no more
> necessity to keep a list of RSS configurable for each protocol.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On Thu, Oct 05, 2017 at 02:49:51PM +0200, Nelio Laranjeiro wrote:
[...]
> -int
> -mlx5_vlan_filter_set(struct rte_eth_dev *dev, uint16_t vlan_id, int on)
> -{
> - struct priv *priv = dev->data->dev_private;
> - int ret;
> -
> - priv_lock(priv);
> - ret = vlan_filter_set(dev, vlan_id
On Thu, Oct 05, 2017 at 02:49:50PM +0200, Nelio Laranjeiro wrote:
> RSS hash configuration is currently ignored by the PMD, this commits
> removes the RSS feature on promiscuous mode.
A mistake? "promiscuous" -> "all multi"
>
> This functionality will be added in a later commit.
>
> Signed-off-by
On Thu, Oct 05, 2017 at 02:49:49PM +0200, Nelio Laranjeiro wrote:
> @@ -1630,3 +1646,90 @@ priv_flow_verify(struct priv *priv)
> }
> return ret;
> }
> +
> +/**
> + * Enable/disable a flow control configured from the control plane.
"flow control" -> "control flow"
Acked-by: Yongseok Ko
On Thu, Oct 05, 2017 at 02:49:48PM +0200, Nelio Laranjeiro wrote:
> drivers/net/mlx5/mlx5_rxq.c:606:6: error: comparison of constant 4
> with expression of type 'enum hash_rxq_flow_type' is always true
> [-Werror,-Wtautological-constant-out-of-range-compare]
> i
On Thu, Oct 05, 2017 at 02:49:47PM +0200, Nelio Laranjeiro wrote:
[...]
> +struct mlx5_hrxq*
> +mlx5_priv_hrxq_get(struct priv *priv, uint8_t *rss_key, uint8_t rss_key_len,
> +uint64_t hash_fields, uint16_t queues[], uint16_t queues_n)
> +{
> + struct mlx5_hrxq *hrxq;
> +
> +
On Thu, Oct 05, 2017 at 02:49:46PM +0200, Nelio Laranjeiro wrote:
> Indirection table in verbs side resides in a list of final work queues to
> spread the packets according to an higher level queue. This indirection
> table can be shared among the hash Rx queues which points to them.
>
> Signed-o
On Thu, Oct 05, 2017 at 02:49:45PM +0200, Nelio Laranjeiro wrote:
> Use the same design for DPDK queue as for Verbs queue for symmetry, this
> also helps in fixing some issues like the DPDK release queue API which is
> not expected to fail. With such design, the queue is released when the
> refere
On Thu, Oct 05, 2017 at 02:49:44PM +0200, Nelio Laranjeiro wrote:
> Use the same design for DPDK queue as for Verbs queue for symmetry, this
> also helps in fixing some issues like the DPDK release queue API which is
> not expected to fail. With such design, the queue is released when the
> refere
On Thu, Oct 05, 2017 at 02:49:43PM +0200, Nelio Laranjeiro wrote:
> Move verbs object to their own functions to allocate/release them
> independently from the DPDK queue. At the same time a reference counter is
> added to help in issues detections when the queue is being release but
> still in use
On Friday 06 October 2017 06:00 AM, Thomas Monjalon wrote:
> 01/10/2017 11:14, Santosh Shukla:
>> --- a/lib/librte_ether/rte_ethdev_version.map
>> +++ b/lib/librte_ether/rte_ethdev_version.map
>> @@ -192,5 +192,6 @@ DPDK_17.11 {
>> global:
>>
>> rte_eth_dev_reset;
>> + rte_
On Friday 06 October 2017 05:59 AM, Thomas Monjalon wrote:
> 01/10/2017 11:14, Santosh Shukla:
>> --- a/lib/librte_eal/common/eal_common_options.c
>> +++ b/lib/librte_eal/common/eal_common_options.c
>> @@ -98,6 +98,7 @@ eal_long_options[] = {
>> {OPT_VFIO_INTR, 1, NULL, OPT_VFIO_INTR_
On Thu, Oct 05, 2017 at 02:49:42PM +0200, Nelio Laranjeiro wrote:
[...]
> +struct mlx5_rxq_ibv*
> +mlx5_priv_rxq_ibv_get(struct priv *priv, uint16_t idx)
> +{
> + struct mlx5_rxq_data *rxq_data = (*priv->rxqs)[idx];
> + struct mlx5_rxq_ctrl *rxq_ctrl;
> +
> + if (idx >= priv->rxqs_n)
>
On Friday 06 October 2017 05:49 AM, Thomas Monjalon wrote:
> 20/09/2017 13:23, Santosh Shukla:
>> For auto detection purpose:
>> * Below calls moved up in the eal initialization order:
>> - eal_option_device_parse
>> - rte_bus_scan
>>
>> Based on the result of rte_bus_scan_iommu_class -
On Friday 06 October 2017 05:47 AM, Thomas Monjalon wrote:
> 20/09/2017 13:23, Santosh Shukla:
>> +/** Device driver supports iova as va */
>> +#define RTE_PCI_DRV_IOVA_AS_VA 0X0040
> This flag name is surprizing and the comment does not help.
> For the comment:
> "Device driver supports I/O
Thomas,
You comment is annoying and infuriating both.
Patch is their for more than 4month, had enough time for you to comment
and understand the topic. Thorough review and testing has happened both.
NOTE: You have already delayed this series by one release and
I'm guessing that you intent to push
On 9/29/2017 8:17 AM, Zhiyong Yang wrote:
> port_id is currently defined as uint8_t, which is limited to the range
> 0 to 255. A larger range is required for vdev scalability.
>
> It is necessary for a redefinition of port_id to extend it from
> 1 bytes to 2 bytes. All ethdev APIs and usages relat
On 9/16/2017 11:52 AM, Hemant Agrawal wrote:
> This patchset includes the hw driver upgrades and additional
> nic feature implementations.
>
> patches 1..8 - upgrades the qbman hw driver
> patches 9..10 - adds the support for LX2160 platform
> patches 11..27 - adds various features and cleanups in
On 10/5/2017 10:49 AM, Adrien Mazarguil wrote:
> This series brings enhancements to various rte_flow helpers:
>
> - Allow applications to use rte_flow_error_set() by making it part of the
> public interface and documenting it as such.
>
> - Address rte_flow_copy()'s limitations by replacing it
On Thu, Oct 05, 2017 at 02:49:41PM +0200, Nelio Laranjeiro wrote:
[...]
> @@ -180,12 +133,14 @@ mlx5_txq_mp2mr_reg(struct mlx5_txq_data *txq, struct
> rte_mempool *mp,
> {
> struct mlx5_txq_ctrl *txq_ctrl =
> container_of(txq, struct mlx5_txq_ctrl, txq);
> - struct ibv_mr
On 10/3/2017 12:00 PM, Raslan Darawsheh wrote:
> Add a new offload capability flag for Rx HW
> timestamp and enabling/disabling this via rte_eth_rxmode.
>
> Signed-off-by: Raslan Darawsheh
<...>
> +++ b/doc/guides/nics/features.rst
> @@ -567,6 +567,17 @@ Supports L4 checksum offload.
> * **[pr
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> The number of queues in DPDK does not means that the array of queue will be
> totally filled, those information are uncorrelated. The number of queues
> is provided in the port configuration whereas the array is filled by
> calling tx
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> Reta update needs to stop/start the port but stopping the port does not
> disable the polling functions which may end in a segfault if a core is
> polling the queue while the control thread is modifying it.
>
> This patch changes the
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> This flag is already present in the Ethernet device.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> Debug tools to verify all flows are be un-register from the NIC.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> Prefix struct txq_ctrl and associated function with mlx5.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> Prefix struct rxq_ctrl and associated functions with mlx5.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
> On Oct 5, 2017, at 5:49 AM, Nelio Laranjeiro
> wrote:
>
> Generic flow API should be use for flow steering as is provides a better
> and easier way to configure flows.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On Thu, Oct 05, 2017 at 02:49:33PM +0200, Nelio Laranjeiro wrote:
> mlx5_flow_create() and mlx5_flow_validate() are making common checks.
>
> Signed-off-by: Nelio Laranjeiro
> ---
Acked-by: Yongseok Koh
Thanks
On 10/6/2017 12:42 AM, Roger B. Melton wrote:
> Hi Everyone,
>
> As soon as I submitted the patch, I realized I neglected to signoff.
> What's the recommended procedure, resubmit the original signed off, or
> bump to v1?
Please send a new one increasing version, since this was v1 already next
01/10/2017 11:14, Santosh Shukla:
> --- a/lib/librte_ether/rte_ethdev_version.map
> +++ b/lib/librte_ether/rte_ethdev_version.map
> @@ -192,5 +192,6 @@ DPDK_17.11 {
> global:
>
> rte_eth_dev_reset;
> + rte_eth_dev_pool_ops_supported;
You missed the alphabetical order here :
01/10/2017 11:14, Santosh Shukla:
> --- a/lib/librte_eal/common/eal_common_options.c
> +++ b/lib/librte_eal/common/eal_common_options.c
> @@ -98,6 +98,7 @@ eal_long_options[] = {
> {OPT_VFIO_INTR, 1, NULL, OPT_VFIO_INTR_NUM},
> {OPT_VMWARE_TSC_MAP,0, NULL, OPT_VMWARE
On 9/11/2017 4:13 PM, Olivier Matz wrote:
> In developer mode (RTE_DEVEL_BUILD=y) where -Werror is passed in the
> CFLAGS, the compilation fails with gcc-6.3.0 and EXTRA_CFLAGS=-Og. Some
> errors are real bugs (but not critical), while some are false positives
> (gcc bugs?).
>
> The solution often
20/09/2017 13:23, Santosh Shukla:
> For auto detection purpose:
> * Below calls moved up in the eal initialization order:
> - eal_option_device_parse
> - rte_bus_scan
>
> Based on the result of rte_bus_scan_iommu_class - select iova
> mapping mode.
It does not explain why you need to
On 9/11/2017 4:13 PM, Olivier Matz wrote:
> The compilation with gcc-6.3.0 and EXTRA_CFLAGS=-Og gives the following
> error:
>
> CC rte_lpm6.o
> rte_lpm6.c: In function ‘rte_lpm6_add_v1705’:
> rte_lpm6.c:442:11: error: ‘tbl_next’ may be used uninitialized in
> th
20/09/2017 13:23, Santosh Shukla:
> +/** Device driver supports iova as va */
> +#define RTE_PCI_DRV_IOVA_AS_VA 0X0040
This flag name is surprizing and the comment does not help.
For the comment:
"Device driver supports I/O virtual addressing" ?
For the flag:
RTE_PCI_DRV_IOVA ?
[.
This patch is introducing a new abstraction.
It is important to explain it for future readers of this code.
20/09/2017 13:23, Santosh Shukla:
> +/**
> + * IOVA mapping mode.
> + */
Please explain what IOVA means and what is the purpose of
distinguish the different modes.
> +enum rte_iova_mode {
Hi Everyone,
As soon as I submitted the patch, I realized I neglected to signoff.
What's the recommended procedure, resubmit the original signed off, or
bump to v1?
Thanks,
Roger
On 10/5/17 3:11 PM, Roger B Melton wrote:
---
i40evf tx vector logic frees mbufs, but it does not remove the
13/07/2017 13:48, Hemant Agrawal:
> In case no_pci is configured, fslmc bus will still need the
> the vfio to be enabled.
>
> Signed-off-by: Hemant Agrawal
> ---
> --- a/lib/librte_eal/linuxapp/eal/eal.c
> +++ b/lib/librte_eal/linuxapp/eal/eal.c
> +#ifdef RTE_LIBRTE_FSLMC_BUS
> + if (!vfio_en
> Signed-off-by: Shreyansh Jain
> Signed-off-by: Hemant Agrawal
This is a huge patch without any comment!
It will be hard to find bugs in it when debugging in future.
And it will be also pointless to reference it in future fixes.
>From a quality point of view, it is bad.
As it is touching only
On 10/6/2017 12:17 AM, Ferruh Yigit wrote:
> On 9/11/2017 4:13 PM, Olivier Matz wrote:
>> The compilation with gcc-6.3.0 and EXTRA_CFLAGS=-Og gives the following
>> error:
>>
>> CC virtio_rxtx.o
>> virtio_rxtx.c: In function ‘virtio_rx_offload’:
>> virtio_rxtx.c:680:10: error: ‘csum’ may be u
On 9/11/2017 4:13 PM, Olivier Matz wrote:
> The compilation with gcc-6.3.0 and EXTRA_CFLAGS=-Og gives the following
> error:
>
> CC i40e_adminq.o
> i40e_adminq.c: In function ‘i40e_clean_arq_element’:
> i40e_adminq.c:1145:56: error: ‘ntu’ may be used uninitialized in
>
> On Oct 3, 2017, at 4:00 AM, Raslan Darawsheh wrote:
>
> Expose Rx HW timestamp to packet mbufs.
>
> Signed-off-by: Raslan Darawsheh
> ---
> Changes in v6:
> - Rebasing the work on top of 217b1421 ("net/mlx5: separate DPDK from Verbs
> Rx queue objects")
> ---
Raslan,
It looks like you als
19/09/2017 20:51, Jan Blunck:
> On Mon, Sep 18, 2017 at 1:36 PM, Hemant Agrawal
> wrote:
> > Tested-by: Hemant Agrawal
> >
> >
> > On 8/12/2017 3:52 PM, Shreyansh Jain wrote:
> >>
> >> Bus scan is responsible for finding devices over *all* buses.
> >> Some of these buses might not be able to sca
On 9/11/2017 4:13 PM, Olivier Matz wrote:
> The compilation with gcc-6.3.0 and EXTRA_CFLAGS=-Og gives the following
> error:
>
> CC virtio_rxtx.o
> virtio_rxtx.c: In function ‘virtio_rx_offload’:
> virtio_rxtx.c:680:10: error: ‘csum’ may be used uninitialized in
> thi
> On Oct 3, 2017, at 4:00 AM, Raslan Darawsheh wrote:
>
> Add enabling/disabling Rx HW timestamp from
> command line and parameter.
>
> Signed-off-by: Raslan Darawsheh
> ---
Acked-by: Yongseok Koh
Thanks
11/09/2017 16:06, Ferruh Yigit:
> On 8/25/2017 11:19 AM, Shreyansh Jain wrote:
> > Change Log:
> > ~~~
> > v2:
> >- Minor updates for logging (removed some logs and changed others
> > to make it cleaner when application starts)
> >
> > Brief:
> > ~~
> >
> > -- v1 is at [3]
Brings vectorization through NEON instructions.
Signed-off-by: Yongseok Koh
Acked-by: Nelio Laranjeiro
---
drivers/net/mlx5/Makefile |3 +-
drivers/net/mlx5/mlx5_rxtx_vec.c |4 +-
drivers/net/mlx5/mlx5_rxtx_vec.h |4 +
drivers/net/mlx5/mlx5_rxtx_vec_neon.h | 10
With the upstream rdma-core, to enable Rx CQE compression,
mlx5dv_create_cq() in Direct Verbs has to be used instead of regular Verbs
call (ibv_create_cq()). And if the size of CQE is 128 bytes, compression
is supported only by certain devices. Thus, it has to be decided by
checking the capabilitiy
Considering more architecture (e.g. ARM and PowerPC) will be added for
vectorized Rx/Tx burst, all the shareable functions which don't use any
vector instrinsics need to be separated from architecture-dependent
functions. All the vector functions for x86 SSE are moved to a new header
file - mlx5_rx
The size of Rx completion entry should match the size of a cacheline. This
is already reflected in struct mlx5_cqe by adding 64bytes padding if a
cacheline is 128bytes. Some ARM CPUs have 128bytes cacheline.
Signed-off-by: Yongseok Koh
---
drivers/net/mlx5/mlx5.c | 3 +++
1 file changed, 3 inser
Replace compile-time sanity check with static_assert() as c11 standard has
been set. Add mlx5_rxtx_vec.h and move the sanity checks to the file.
Signed-off-by: Yongseok Koh
---
drivers/net/mlx5/mlx5_rxtx_vec.c | 32 +--
drivers/net/mlx5/mlx5_rxtx_vec.h | 87 ++
Add dataplane functions using ARM NEON instructions. To modularize vectorized
functions for different architectures, the existing files having x86 SSE support
is reorganized.
This patchset has dependency with the following patches:
- net/mlx5: fix overflow of Rx SW ring
- Nelio's mlx5 flow cleanup
Rename mlx5_rxtx_vec_sse.c to mlx5_rxtx_vec.c to separate shareable vector
functions.
Signed-off-by: Yongseok Koh
Acked-by: Nelio Laranjeiro
---
drivers/net/mlx5/Makefile | 2 +-
drivers/net/mlx5/{mlx5_rxtx_vec_sse.c => mlx5_rxtx_vec.c} | 0
2 files changed, 1 in
Updating a consumer index to HW doesn't require a memory barrier in case
that there's no updated data to be posted to HW, but a compiler barrier is
sufficient. rte_wmb() is replaced with rte_io_wmb() when it makes changes
visible to HW, not other core.
Signed-off-by: Yongseok Koh
Acked-by: Shahaf
This commit prevents control path operations from failing after a sub
device removal.
Following are the failure steps:
1. The physical device is removed due to change in one of PF parameters
(e.g. MTU)
2. The interrupt thread flags the device
3. Within 2 seconds Interrupt thread initializes the ac
04/10/2017 21:24, Ferruh Yigit:
> On 9/22/2017 3:48 PM, Daniel Mrzyglod wrote:
> > Memory allocated in strdup is not free.
> >
> > Coverity issue: 143257
> > Fixes: d8a2bc71dfc2 ("log: remove app path from syslog id")
> > Cc: tho...@monjalon.net
> >
> > Signed-off-by: Daniel Mrzyglod
> > ---
> T
On 10/5/2017 10:37 PM, Yongseok Koh wrote:
> If vectorized Rx burst is short of mbufs in replenishment, Rx SW ring can
> overflow as the Rx burst handles 4 packets in a loop. This is because the
> function fills SW ring and its mbufs first and checks validity of
> each completion later. So, there s
> -Original Message-
> From: Kavanagh, Mark B
> Sent: Thursday, October 5, 2017 9:37 PM
> To: dev@dpdk.org
> Cc: Hu, Jiayu ; Tan, Jianfeng ;
> Ananyev, Konstantin ; Yigit,
> Ferruh ; tho...@monjalon.net; Kavanagh, Mark B
>
> Subject: [PATCH v9 0/6] Support TCP/IPv4, VxLAN, and GRE GSO
05/10/2017 23:55, Mokhtar, Amr:
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > 03/10/2017 16:29, Mokhtar, Amr:
> > > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > > > 25/08/2017 15:46, Amr Mokhtar:
> > > > > +int
> > > > > +rte_bbdev_configure(uint8_t dev_id, uint16_t num_queues,
28/09/2017 08:54, Lukasz Majczak:
> The assertion of return value from the open() function is done against
> 0, while it is a correct value - open() returns -1 in case of an error.
> It causes problems while trying to run as a daemon, in which case, this
> call to open() will return 0 as a valid de
> > When compiled on Ubuntu with extra warnings enabled, the rte_strerror()
> > function triggered a warning about an unused return value from
> > strerror_r(). Rather than always have this warning disabled, we fix this,
> > and in the process do some cleanup of the code so as to reduce the
> > com
> -Original Message-
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Tuesday 3 October 2017 16:18
> To: Mokhtar, Amr
> Cc: dev@dpdk.org; f...@redhat.com; acon...@redhat.com; bl...@debian.org
> Subject: Re: [dpdk-dev] [RFC] Wireless Base Band Device (bbdev)
>
> 03/10/2017 16
Hi, Martin
Thanks for your thorough and valuable reporting. We could reproduce it. I found
a bug and fixed it. Please refer to the patch [1] I sent to the mailing list.
This might not be automatically applicable to v17.08 as I rebased it on top of
Nelio's flow cleanup patch. But as this is a simpl
21/09/2017 18:07, Patrick MacArthur:
> On 09/21/2017 09:49 AM, Michal Jastrzebski wrote:
> > From: Kuba Kozak
> >
> > Add file descriptor value check before calling close() function.
> >
> > Coverity issue: 141297
> > Fixes: 811b6b25060f ("vfio: fix file descriptor leak in multi-process")
> > Cc
If vectorized Rx burst is short of mbufs in replenishment, Rx SW ring can
overflow as the Rx burst handles 4 packets in a loop. This is because the
function fills SW ring and its mbufs first and checks validity of
each completion later. So, there should be some buffer slots at the tail of
the ring
> > Due to the uint32_t accesses in the hash computation, keys that aren't
> > aligned to a uint32_t boundary or multiples of uint32_t in length, may see
> > accesses beyond the end of the key. This may cross a page boundary.
> >
> > Signed-off-by: Chas Williams
>
> Acked-by: John McNamara
A
<...>
>> On 03/10/2017 16:55, Matan Azrad wrote:
>>> One of the main identified use cases for the tap PMD is to be used in
>>> combination with the fail-safe PMD as a fallback for a physical device.
>>>
>>> Fail-safe is very strict about making sure its current configuration is
>>> properly applie
On 10/4/2017 6:26 PM, Ferruh Yigit wrote:
> On 10/4/2017 9:15 AM, Pascal Mazon wrote:
>> Hi,
>>
>> I'm surprised there's only one place in the file with indent problem,
>> but I'm ok with the patch otherwise.
>>
>> Ferruh, I didn't get what you mean; Matan is modifying tap_flow.c in
>> patch 2/2, r
11/09/2017 16:57, Olivier Matz:
> The inner L2 length returned by rte_net_get_ptype() is not
> properly initialized. If the caller does not zero the header
> lengths structure, the inner_l2 field will be undefined.
>
> Fix it by initializing inner_l2 to 0 when parsing a inner layer.
>
> Fixes: 2c
20/09/2017 13:32, Olivier Matz:
> There is no reason to prevent ring from being larger than 0x0FFF.
> Increase the maximum size to 0x7FFF, which is the maximum possible
> without changing the code and the structure definition (size is stored
> on a uint32_t).
>
> Link: http://dpdk.org/ml/a
25/09/2017 12:25, Olivier MATZ:
> On Thu, Sep 21, 2017 at 08:44:09PM +0200, Radoslaw Biernacki wrote:
> > This patch fixes the dynamic log levels testing in logs_autotest.
> > Introduction of rte_log_set_level() in patch c1b5fa94a46f was done
> > with parameter RTE_LOG_EMERG which caused all RTE_LO
On 10/5/2017 9:14 AM, Beilei Xing wrote:
> This patch set enables RSS/FDIR/cloud filter for GPT-C and GTP-U.
>
<...>
>
> Beilei Xing (7):
> mbuf: support GTP in software packet type parser
> net/i40e: update ptype and pctype info
> ethdev: add GTP items to support flow API
> net/i40e: f
Hi Alan,
Comments inline, search for AGR>
Alan.
From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of alangordonde...@gmail.com
Sent: Thursday, October 05, 2017 10:21 AM
To: cristian.dumitre...@intel.com
Cc: dev@dpdk.org; Alan Dewar
Subject: [dpdk-dev] [RFC] sched: parameterize QoS traffic-class
v3:
Add description of raw flow type mode for flow_director_filter
command in testpmd.
Signed-off-by: Kirill Rybalchenko
---
doc/guides/testpmd_app_ug/testpmd_funcs.rst | 4
1 file changed, 4 insertions(+)
diff --git a/doc/guides/testpmd_app_ug/testpmd_funcs.rst
b/doc/guides/testpmd_app_u
When addidng flow director filter for raw flow type, instead
of constructing packet use buffer with pre-constructed packet.
v3:
Merge with dynamic pctype to flow type mapping patch.
Signed-off-by: Kirill Rybalchenko
---
drivers/net/i40e/i40e_fdir.c | 25 -
1 file changed
v3:
Add raw flow type support to flow_director_filter command
Signed-off-by: Kirill Rybalchenko
---
app/test-pmd/cmdline.c | 69 +++---
1 file changed, 66 insertions(+), 3 deletions(-)
diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index
For complex packets use raw flow type with pre-constructed packet buffer
instead of creating a packet internally in PMD.
v2:
Fixed code style, comment added
v3:
Merged with dunamic pctype to flow type mapping patch.
Added raw fode to flow_director_filter command in testpmd.
Kirill Rybalchenko (4
Add new structure rte_eth_raw_flow to the union rte_eth_fdir_flow
to support filter for raw flow type.
Signed-off-by: Kirill Rybalchenko
---
lib/librte_ether/rte_eth_ctrl.h | 10 ++
1 file changed, 10 insertions(+)
diff --git a/lib/librte_ether/rte_eth_ctrl.h b/lib/librte_ether/rte_eth_
05/10/2017 22:06, Mokhtar, Amr:
> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > 25/08/2017 15:46, Amr Mokhtar:
> > Did you think about the API required for inline processing (i.e. bbdev
> > combined
> > with ethdev Rx/Tx)?
>
> The current programming model is that ethdev is being used fo
On 10/5/2017 4:06 PM, Ajit Khaparde wrote:
> A continuation of the previous submission,
> this patchset adds support for ntuple filtering.
> It also fixes an issue found while testing VMDQ
> and updates bnxt.ini
> The last patch removes some unnecessary parentheses from the code.
>
> Ajit Khaparde
Add programmer's guide doc to explain the design and use of the
GSO library.
Signed-off-by: Mark Kavanagh
Signed-off-by: Jiayu Hu
---
MAINTAINERS| 6 +
.../generic_segmentation_offload_lib.rst | 256 +++
.../prog_guide/img/gso-output-s
From: Jiayu Hu
This patch adds GSO support for TCP/IPv4 packets. Supported packets
may include a single VLAN tag. TCP/IPv4 GSO doesn't check if input
packets have correct checksums, and doesn't update checksums for
output packets (the responsibility for this lies with the application).
Additional
From: Jiayu Hu
This patch adds GSO support to the csum forwarding engine. Oversized
packets transmitted over a GSO-enabled port will undergo segmentation
(with the exception of packet-types unsupported by the GSO library).
GSO support is disabled by default.
GSO support may be toggled on a per-p
1 - 100 of 360 matches
Mail list logo
