[dpdk-dev] [PATCH v4 0/5] vhost: generalize buffer vectors

This series is again preliminray work to ease packed ring layout integration. Main changes are using vector buffres also in the dequeue path, and perform IOVA to HVA translation at vectors fill time. I still have to run more benchmarks, but PVP benchmarks does not show performance changes. Good

[dpdk-dev] [PATCH v4 1/5] vhost: use shadow used ring in dequeue path

Relax used ring contention by reusing the shadow used ring feature used by enqueue path. Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 50 +-- 1 file changed, 10 insertions(+), 40 deletions(-) diff --git a/lib/librte_vhost/virtio_net.

[dpdk-dev] [PATCH v4 2/5] vhost: use buffer vectors in dequeue path

To ease packed ring layout integration, this patch makes the dequeue path to re-use buffer vectors implemented for enqueue path. Doing this, copy_desc_to_mbuf() is now ring layout type agnostic. Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.h | 1 + lib/librte_vhost/virtio_ne

[dpdk-dev] [PATCH v4 3/5] vhost: improve prefetching in dequeue path

This is an optimization to prefetch next buffer while the current one is being processed. Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 10 +- 1 file changed, 9 insertions(+), 1 deletion(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c i

[dpdk-dev] [PATCH v4 4/5] vhost: prefetch first descriptor in dequeue path

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 2cfd8585c..2662a1d32 100644 --- a/lib/librte_vhost/virtio_net.c +++ b/lib/librte_vhost/virtio_net.c @@ -1083,6

[dpdk-dev] [PATCH v4 5/5] vhost: improve prefetching in enqueue path

This is an optimization to prefetch next buffer while the current one is being processed. Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 12 +--- 1 file changed, 9 insertions(+), 3 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.

[dpdk-dev] [PATCH v9 03/15] vhost: vring address setup for packed queues

From: Yuanhan Liu Add code to set up packed queues when enabled. Signed-off-by: Yuanhan Liu Signed-off-by: Jens Freimann Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.c | 43 ++- lib/librte_vhost/vhost.h | 7 ++- lib/librte_v

[dpdk-dev] [PATCH v9 01/15] vhost: add virtio packed virtqueue defines

From: Jens Freimann Signed-off-by: Jens Freimann --- lib/librte_vhost/vhost.h | 20 1 file changed, 20 insertions(+) diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h index 79e3117d2..83c028a1b 100644 --- a/lib/librte_vhost/vhost.h +++ b/lib/librte_vhost/vho

[dpdk-dev] [PATCH v9 00/15] Vhost: add support to packed ring layout

This series is a handover from Jen's "[PATCH v4 00/20] implement packed virtqueues", which only implements the vhost side. Virtio PMD implementation will follow in a next series. The series applies on top of previous reworks I posted during this cycle that merges mergeable and non-mergeable recei

[dpdk-dev] [PATCH v9 02/15] vhost: add helpers for packed virtqueues

From: Jens Freimann Add some helper functions to check descriptor flags and check if a vring is of type packed. Signed-off-by: Jens Freimann Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.h | 13 + 1 file changed, 13 insertions(+) diff --git a/lib/librte_vhost/vhost.h

[dpdk-dev] [PATCH v9 04/15] vhost: clear shadow used table index at flush time

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 82d5d9e17..d6b30899f 100644 --- a/lib/librte_vhost/virtio_net.c +++ b/lib/librte_vhost/virtio

[dpdk-dev] [PATCH v9 06/15] vhost: clear batch copy index at copy time

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 8 1 file changed, 4 insertions(+), 4 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 3bed77eec..5cc3138d0 100644 --- a/lib/librte_vhost/virtio_net.c +++ b/lib/librte_vhost/v

[dpdk-dev] [PATCH v9 05/15] vhost: make indirect desc table copy desc type agnostic

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 16 1 file changed, 8 insertions(+), 8 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index d6b30899f..3bed77eec 100644 --- a/lib/librte_vhost/virtio_net.c +++ b/lib/librt

[dpdk-dev] [PATCH v9 08/15] vhost: append shadow used ring function names with split

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 28 +++- 1 file changed, 15 insertions(+), 13 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index bdfd6ebef..ae256e062 100644 --- a/lib/librte_vhost/virtio_net.c +

[dpdk-dev] [PATCH v9 07/15] vhost: extract split ring handling from Rx and Tx functions

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 238 +++--- 1 file changed, 129 insertions(+), 109 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 5cc3138d0..bdfd6ebef 100644 --- a/lib/librte_vho

[dpdk-dev] [PATCH v9 09/15] vhost: add shadow used ring support for packed rings

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.c | 9 -- lib/librte_vhost/vhost.h | 13 ++-- lib/librte_vhost/vhost_user.c | 64 -- lib/librte_vhost/virtio_net.c | 71 +-- 4 files changed, 13

[dpdk-dev] [PATCH v9 13/15] vhost: add Tx support for packed ring

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.h | 1 + lib/librte_vhost/virtio_net.c | 119 +- 2 files changed, 119 insertions(+), 1 deletion(-) diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h index 70f0eebcf..a7e602bec

[dpdk-dev] [PATCH v9 10/15] vhost: create descriptor mapping function

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 70 --- 1 file changed, 40 insertions(+), 30 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 44b9daf4a..058786871 100644 --- a/lib/librte_vhost

[dpdk-dev] [PATCH v9 11/15] vhost: add vector filling support for packed ring

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 111 ++ 1 file changed, 111 insertions(+) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 058786871..9171ee733 100644 --- a/lib/librte_vhost/virtio_net.c ++

[dpdk-dev] [PATCH v9 12/15] vhost: add Rx support for packed ring

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/virtio_net.c | 123 -- 1 file changed, 119 insertions(+), 4 deletions(-) diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c index 9171ee733..ef86e5b40 100644 --- a/lib/librte_vhost

[dpdk-dev] [PATCH v9 15/15] vhost: advertize packed ring layout support

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.h | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) diff --git a/lib/librte_vhost/vhost.h b/lib/librte_vhost/vhost.h index 760a09c0d..9b0ebb754 100644 --- a/lib/librte_vhost/vhost.h +++ b/lib/librte_vhost/vhost.h @@ -275,7 +275,8 @@

[dpdk-dev] [PATCH v9 14/15] vhost: add notification for packed ring

Signed-off-by: Maxime Coquelin --- lib/librte_vhost/vhost.c | 71 +++-- lib/librte_vhost/vhost.h | 73 +-- lib/librte_vhost/vhost_user.c | 24 ++ lib/librte_vhost/virtio_net.c | 12 +++ 4 files c

Re: [dpdk-dev] [PATCH v4 1/5] vhost: use shadow used ring in dequeue path

Hi Tiwei, On 07/06/2018 09:04 AM, Maxime Coquelin wrote: Relax used ring contention by reusing the shadow used ring feature used by enqueue path. Signed-off-by: Maxime Coquelin Just noticed I forgot to apply your: Reviewed-by: Tiwei Bie Regards, Maxime --- lib/librte_vhost/virtio_net.c |

[dpdk-dev] [PATCH v3 01/16] bus/dpaa: fix phandle support for kernel 4.16

From: Alok Makhariya Fixes: 2183c6f69d7e ("bus/dpaa: add OF parser for device scanning") Cc: Shreyansh Jain Cc: sta...@dpdk.org Signed-off-by: Alok Makhariya Acked-by: Shreyansh Jain --- drivers/bus/dpaa/base/fman/of.c | 5 + 1 file changed, 5 insertions(+) diff --git a/drivers/bus/dpaa

[dpdk-dev] [PATCH v3 02/16] bus/dpaa: fix svr id fetch location

Otherwise the SVR may not be avilable for dpaa init. Fixes: 3b59b73dea08 ("bus/dpaa: update platform SoC value register routines") Cc: sta...@dpdk.org Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/bus/dpaa/dpaa_bus.c | 14 +++--- 1 file changed, 7 insertions(+), 7

[dpdk-dev] [PATCH v3 04/16] net/dpaa: fix the queue err handling and logs

Fixes: 5e7455931442 ("net/dpaa: support Rx queue configurations with eventdev") Cc: sta...@dpdk.org Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/net/dpaa/dpaa_ethdev.c | 34 ++ 1 file changed, 26 insertions(+), 8 deletions(-) diff --git a/d

[dpdk-dev] [PATCH v3 03/16] bus/dpaa: fix the buffer offset setting in FMAN

The buffer offset was incorrectly being set at 64, thus not honoring the packet headroom. Fixes: 6d6b4f49a155 (bus/dpaa: add FMAN hardware operations") Cc: sta...@dpdk.org Signed-off-by: Hemant Agrawal --- drivers/bus/dpaa/base/fman/fman_hw.c | 20 +++- 1 file changed, 11 insert

[dpdk-dev] [PATCH v3 06/16] bus/dpaa: optimize the fq callback routine

Avoid array of fq as packets are dq only from a single q. Signed-off-by: Sunil Kumar Kori Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/bus/dpaa/base/qbman/qman.c | 15 +++ drivers/net/dpaa/dpaa_rxtx.c | 2 +- 2 files changed, 8 insertions(+), 9 deletion

[dpdk-dev] [PATCH v3 08/16] bus/dpaa: make vdqcr configurable

From: Nipun Gupta This patch add support for configurable vdqcr exact flag. This boost the performance, however this can give side effects for some extra packet fetch. Which has been taken care in the patch as well. Signed-off-by: Nipun Gupta Acked-by: Shreyansh Jain --- drivers/bus/dpaa/base

[dpdk-dev] [PATCH v3 09/16] net/dpaa: support default queue mode

In case DPAA FMAN configuration tool (FMC) is not available, the system can still work with default queue(1 queue per port). This patch also fixes some logs related to FQ ids, which were idetified while testing this support. Signed-off-by: Hemant Agrawal --- drivers/net/dpaa/dpaa_ethdev.c | 37

[dpdk-dev] [PATCH v3 05/16] net/dpaa2: fix the prefetch Rx to honor nb pkts

This patch fixes prefetch rx routine to set the next prefetch request to the size of nb_pkts. It assumes that next request would ideally be of same size. Fixes: 4bc5ab88dbd6 ("net/dpaa2: fix Tx only mode") Cc: sta...@dpdk.org Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/n

[dpdk-dev] [PATCH v3 07/16] bus/dpaa: implement new of API to get MAC address

From: Akhil Goyal Signed-off-by: Akhil Goyal Acked-by: Shreyansh Jain --- drivers/bus/dpaa/base/fman/of.c | 39 +++ drivers/bus/dpaa/include/of.h | 2 ++ drivers/bus/dpaa/rte_bus_dpaa_version.map | 8 +++ 3 files changed, 49 insertions(+)

[dpdk-dev] [PATCH v3 11/16] bus/dpaa: cleanup unnecessary global variables

Signed-off-by: Pavan Nikhilesh Signed-off-by: Hemant Agrawal --- drivers/bus/dpaa/base/fman/netcfg_layer.c | 5 - drivers/bus/dpaa/base/qbman/bman_driver.c | 4 ++-- drivers/bus/dpaa/base/qbman/qman.c| 2 +- drivers/bus/dpaa/base/qbman/qman_driver.c | 4 ++-- drivers/bus/dpaa/base/qb

[dpdk-dev] [PATCH v3 10/16] net/dpaa: remove experimental tag from PMD APIs

Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/net/dpaa/dpaa_ethdev.c| 6 +++--- drivers/net/dpaa/dpaa_ethdev.h| 8 +--- drivers/net/dpaa/rte_pmd_dpaa.h | 5 + drivers/net/dpaa/rte_pmd_dpaa_version.map | 4 ++-- 4 files changed, 11 in

[dpdk-dev] [PATCH v3 12/16] bus/fslmc: cleanup unnecessary global variables

Signed-off-by: Pavan Nikhilesh Signed-off-by: Hemant Agrawal --- drivers/bus/fslmc/qbman/qbman_portal.c | 3 +-- drivers/bus/fslmc/qbman/qbman_portal.h | 1 - 2 files changed, 1 insertion(+), 3 deletions(-) diff --git a/drivers/bus/fslmc/qbman/qbman_portal.c b/drivers/bus/fslmc/qbman/qbman_por

[dpdk-dev] [PATCH v3 14/16] net/dpaa: move the push queue set to global init

Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/net/dpaa/dpaa_ethdev.c | 24 ++-- 1 file changed, 14 insertions(+), 10 deletions(-) diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c index 00611f8..5c0aafb 100644 --- a/drivers/net

[dpdk-dev] [PATCH v3 15/16] bus/dpaa: add support for SG config

Signed-off-by: Hemant Agrawal --- drivers/bus/dpaa/base/fman/fman_hw.c | 42 +++ drivers/bus/dpaa/include/fsl_fman.h | 6 + drivers/bus/dpaa/rte_bus_dpaa_version.map | 2 ++ 3 files changed, 50 insertions(+) diff --git a/drivers/bus/dpaa/base/fman/fma

[dpdk-dev] [PATCH v3 16/16] net/dpaa: implement scatter offload support

Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/net/dpaa/dpaa_ethdev.c | 75 +- drivers/net/dpaa/dpaa_ethdev.h | 3 +- drivers/net/dpaa/dpaa_rxtx.c | 4 +-- drivers/net/dpaa/dpaa_rxtx.h | 2 -- 4 files changed, 70 insertions(+), 1

[dpdk-dev] [PATCH v3 13/16] drivers: support function name in logs trace

Signed-off-by: Hemant Agrawal Acked-by: Shreyansh Jain --- drivers/bus/fslmc/fslmc_logs.h | 2 +- drivers/crypto/dpaa2_sec/dpaa2_sec_logs.h | 2 +- drivers/crypto/dpaa_sec/dpaa_sec_log.h | 2 +- drivers/event/dpaa2/dpaa2_eventdev_logs.h | 2 +- drivers/net/dpaa2/dpaa2_pmd_logs.

Re: [dpdk-dev] [PATCH v2 09/16] net/dpaa: support default queue mode

On Wednesday 04 July 2018 03:13 PM, Hemant Agrawal wrote: > In case DPAA FMAN configuration tool (FMC) is not available. > System can still work with default queue. (1 queue per port). The commit message needs to reflect the fact this patch is also fixing some debugging logs which has not relati

Re: [dpdk-dev] [PATCH v2 05/16] net/dpaa2: fix the prefetch Rx to honor nb pkts

On Wednesday 04 July 2018 03:13 PM, Hemant Agrawal wrote: > This patch fix the prefetch rx routine to ^ fixes > set the next prefetch request to the size of nb_pkts. > This will assume that next request will ideally will be of same size. Incorrect wording. Maybe: "It

Re: [dpdk-dev] [PATCH v2 15/20] net/mlx5: support inner RSS computation

On Wed, Jun 27, 2018 at 05:07:47PM +0200, Nelio Laranjeiro wrote: > Signed-off-by: Nelio Laranjeiro > --- > drivers/net/mlx5/mlx5_flow.c | 131 +-- > drivers/net/mlx5/mlx5_rxtx.h | 1 - > 2 files changed, 96 insertions(+), 36 deletions(-) > > diff --git a/driver

Re: [dpdk-dev] [PATCH v2 12/20] net/mlx5: add mark/flag flow action

On Thu, Jul 05, 2018 at 12:56:09PM -0700, Yongseok Koh wrote: >[...] > > > > + if (mark->id >= MLX5_FLOW_MARK_MAX) > > > > + return rte_flow_error_set(error, EINVAL, > > > > + > > > > RTE_FLOW_ERROR_TYPE_ACTION_CONF, > > > > +

Re: [dpdk-dev] [PATCH v3 3/4] compressdev: replace mbuf scatter gather flag

> -Original Message- > From: Verma, Shally [mailto:shally.ve...@cavium.com] > Sent: Thursday, July 5, 2018 12:59 PM > To: De Lara Guarch, Pablo ; Gupta, Ashish > ; Trahe, Fiona ; Daly, Lee > ; Sahu, Sunila > Cc: dev@dpdk.org > Subject: RE: [PATCH v3 3/4] compressdev: replace mbuf scatte

Re: [dpdk-dev] [PATCH v3 3/4] compressdev: replace mbuf scatter gather flag

>-Original Message- >From: De Lara Guarch, Pablo [mailto:pablo.de.lara.gua...@intel.com] >Sent: 06 July 2018 14:10 >To: Verma, Shally ; Gupta, Ashish >; Trahe, Fiona ; >Daly, Lee ; Sahu, Sunila >Cc: dev@dpdk.org >Subject: RE: [PATCH v3 3/4] compressdev: replace mbuf scatter gather flag

Re: [dpdk-dev] [PATCH v3 3/4] compressdev: replace mbuf scatter gather flag

> -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Verma, Shally > Sent: Friday, July 6, 2018 9:53 AM > To: De Lara Guarch, Pablo ; Gupta, Ashish > ; Trahe, Fiona ; Daly, Lee > ; Sahu, Sunila > Cc: dev@dpdk.org > Subject: Re: [dpdk-dev] [PATCH v3 3/4] compressde

Re: [dpdk-dev] [PATCH] mk: using initial-exec model for thread local variable

On Fri, Jul 06, 2018 at 02:22:14AM +, Liu, Yong wrote: > > > > -Original Message- > > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Sachin Saxena > > Sent: Thursday, July 05, 2018 10:46 PM > > To: Liu, Yong ; Yang, Zhiyong ; > > tho...@monjalon.net; dev@dpdk.org > > Subject: Re

Re: [dpdk-dev] [PATCH v2 1/6] hash: make duplicated code into functions

Hi Yipeng, > -Original Message- > From: Wang, Yipeng1 > Sent: Friday, June 29, 2018 1:25 PM > To: De Lara Guarch, Pablo > Cc: dev@dpdk.org; Wang, Yipeng1 ; Richardson, > Bruce ; honnappa.nagaraha...@arm.com; > vgu...@caviumnetworks.com; brijesh.s.si...@gmail.com > Subject: [PATCH v2 1/6]

Re: [dpdk-dev] [PATCH v2] librte_lpm: Improve performance of the delete and add functions

On Mon, Jul 02, 2018 at 07:42:11PM +0300, Alex Kiselev wrote: > There are two major problems with the library: > first, there is no need to rebuild the whole LPM tree > when a rule is deleted and second, due to the current > rules algorithm with complexity O(n) it's almost > impossible to deal with

Re: [dpdk-dev] [PATCH v2] librte_lpm: Improve performance of the delete and add functions

On Mon, Jul 02, 2018 at 07:42:11PM +0300, Alex Kiselev wrote: > There are two major problems with the library: > first, there is no need to rebuild the whole LPM tree > when a rule is deleted and second, due to the current > rules algorithm with complexity O(n) it's almost > impossible to deal with

Re: [dpdk-dev] [PATCH v2] librte_lpm: Improve performance of the delete and add functions

On Fri, Jul 06, 2018 at 11:13:53AM +0100, Bruce Richardson wrote: > On Mon, Jul 02, 2018 at 07:42:11PM +0300, Alex Kiselev wrote: > > There are two major problems with the library: > > first, there is no need to rebuild the whole LPM tree > > when a rule is deleted and second, due to the current >

Re: [dpdk-dev] [PATCH v4 00/23] net/softnic: refactoring

> -Original Message- > From: Singh, Jasvinder > Sent: Thursday, July 5, 2018 4:48 PM > To: dev@dpdk.org > Cc: Dumitrescu, Cristian > Subject: [PATCH v4 00/23] net/softnic: refactoring > > This patch set modifies the Soft NIC device driver to use the Packet > Framework, which makes it mu

Re: [dpdk-dev] [PATCH v2] librte_lpm: Improve performance of the delete and add functions

On Mon, Jul 02, 2018 at 07:42:11PM +0300, Alex Kiselev wrote: > There are two major problems with the library: > first, there is no need to rebuild the whole LPM tree > when a rule is deleted and second, due to the current > rules algorithm with complexity O(n) it's almost > impossible to deal with

[dpdk-dev] [PATCH v4 4/4] compressdev: add huffman encoding flags

Added Huffman fixed and dynamic encoding feature flags, so an application can query if a device supports these two types, when performing DEFLATE compression. Signed-off-by: Pablo de Lara Acked-by: Fiona Trahe --- v4/v3: - No change v2: - Fixed typo drivers/compress/isal/isal_compress_pmd_op

[dpdk-dev] [PATCH v4 2/4] doc: rename compress feature flag

Renamed feature "Bypass" to "Pass-through", as it is a more explicit name, meaning that the PMD is capable of passing the mbufs through it, without making any modifications (i.e.. NULL algorithm). Signed-off-by: Pablo de Lara Acked-by: Fiona Trahe --- v4: - Rephrased pass-through feature commen

[dpdk-dev] [PATCH v4 1/4] doc: cleanup ISA-L PMD feature matrix

In PMD feature matrices (.ini files), it is not required to have the list of features that are not supported, just the ones that are. Signed-off-by: Pablo de Lara Acked-by: Lee Daly --- v4: - No change doc/guides/compressdevs/features/isal.ini | 8 1 file changed, 8 deletions(-) dif

[dpdk-dev] [PATCH v4 3/4] compressdev: replace mbuf scatter gather flag

The current mbuf scatter gather feature flag is too ambiguous, as it is not clear if input and/or output buffers can be scatter gather mbufs or not. Therefore, three new flags will replace this flag: - RTE_COMP_FF_OOP_SGL_IN_SGL_OUT - RTE_COMP_FF_OOP_SGL_IN_FB_OUT - RTE_COMP_FF_OOP_LB_IN_SGL_OUT

Re: [dpdk-dev] [PATCH v2] librte_lpm: Improve performance of the delete and add functions

Hi Bruce. It's the test #1 which is giving you the error message. And I don't see anything wrong here. The test is trying to create LMP with the name "LPM1" which is in use by the already created LPM, so it writes error message to the LOG: LPM rules mempool allocation failed: File exists (17)

Re: [dpdk-dev] [PATCH v4 2/4] doc: rename compress feature flag

>-Original Message- >From: Pablo de Lara [mailto:pablo.de.lara.gua...@intel.com] >Sent: 06 July 2018 08:24 >To: Verma, Shally ; Gupta, Ashish >; fiona.tr...@intel.com; >lee.d...@intel.com >Cc: dev@dpdk.org; Pablo de Lara >Subject: [PATCH v4 2/4] doc: rename compress feature flag > >Ext

Re: [dpdk-dev] [PATCH v3 03/16] bus/dpaa: fix the buffer offset setting in FMAN

On Friday 06 July 2018 01:40 PM, Hemant Agrawal wrote: The buffer offset was incorrectly being set at 64, thus not honoring the packet headroom. Fixes: 6d6b4f49a155 (bus/dpaa: add FMAN hardware operations") Cc: sta...@dpdk.org Signed-off-by: Hemant Agrawal --- Acked-by: Shreyansh Jain

Re: [dpdk-dev] [PATCH v4 3/4] compressdev: replace mbuf scatter gather flag

Hi Pablo Looks fine. Just minor comments: >-Original Message- >From: Pablo de Lara [mailto:pablo.de.lara.gua...@intel.com] >Sent: 06 July 2018 08:24 >To: Verma, Shally ; Gupta, Ashish >; fiona.tr...@intel.com; >lee.d...@intel.com >Cc: dev@dpdk.org; Pablo de Lara >Subject: [PATCH v4 3/4]

[dpdk-dev] [RFC 00/11] Support externally allocated memory in DPDK

This is a proposal to enable using externally allocated memory in DPDK. In a nutshell, here is what is being done here: - Index malloc heaps by NUMA node index, rather than NUMA node itself - Add identifier string to malloc heap, to uniquely identify it - Allow creating named heaps and add/remove

[dpdk-dev] [RFC 02/11] eal: add function to rerieve socket index by socket ID

We are preparing to switch to index heap based on heap indexes rather than by NUMA nodes. First few indexes will be equal to NUMA node ID indexes. For example, currently on a machine with NUMA nodes [0, 8], heaps 0 and 8 will be active, while we want to make it so that heaps 0 and 1 are active. How

[dpdk-dev] [RFC 09/11] malloc: allow removing memory from named heaps

Add an API to remove memory from specified heaps. This will first check if all elements within the region are free, and that the region is the original region that was added to the heap (by comparing its length to length of memory addressed by the underlying memseg list). Signed-off-by: Anatoly Bu

[dpdk-dev] [RFC 01/11] mem: allow memseg lists to be marked as external

When we allocate and use DPDK memory, we need to be able to differentiate between DPDK hugepage segments and segments that were made part of DPDK but are externally allocated. Add such a property to memseg lists. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/eal_common_memory.c |

[dpdk-dev] [RFC 06/11] malloc: enable allocating from named heaps

Add new malloc API to allocate memory from heap referenced to by specified name. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/include/rte_malloc.h | 25 ++ lib/librte_eal/common/malloc_heap.c| 2 +- lib/librte_eal/common/malloc_heap.h| 6 ++

[dpdk-dev] [RFC 08/11] malloc: allow adding memory to named heaps

Add an API to add externally allocated memory to malloc heap. The memory will be stored in memseg lists like regular DPDK memory. Multiple segments are allowed within a heap. If IOVA table is not provided, IOVA addresses are filled in with RTE_BAD_IOVA. Signed-off-by: Anatoly Burakov --- lib/lib

[dpdk-dev] [RFC 07/11] malloc: enable creating new malloc heaps

Add API to allow creating new malloc heaps. They will be created with indexes higher than heaps reserved for NUMA sockets, and up to RTE_MAX_HEAPS. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/include/rte_malloc.h | 21 ++ lib/librte_eal/common/malloc_heap.c| 16 +

[dpdk-dev] [RFC 04/11] malloc: add name to malloc heaps

We will need to refer to external heaps in some way. While we use heap ID's internally, for external API use it has to be something more user-friendly. So, we will be using a string to uniquely identify a heap. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/include/rte_malloc_heap.h |

[dpdk-dev] [RFC 10/11] malloc: allow destroying heaps

Add an API to destroy specified heap. Any memory regions still contained within the heap will be removed first. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/include/rte_malloc.h | 21 lib/librte_eal/common/malloc_heap.c| 29 ++ lib/librte_

[dpdk-dev] [RFC 05/11] malloc: enable retrieving statistics from named heaps

Add internal functions to look up heap by name, and enable dumping statistics for a specified named heap. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/include/rte_malloc.h | 19 +++-- lib/librte_eal/common/malloc_heap.c| 31 ++ lib/librte_eal/commo

[dpdk-dev] [RFC 03/11] malloc: index heaps using heap ID rather than NUMA node

Switch over all parts of EAL to use heap ID instead of NUMA node ID to identify heaps. Heap ID for DPDK-internal heaps is NUMA node's index within the detected NUMA node list. Signed-off-by: Anatoly Burakov --- config/common_base| 1 + lib/librte_eal/common/eal_commo

[dpdk-dev] [RFC 11/11] memzone: enable reserving memory from named heaps

Add ability to allocate memory for memzones from named heaps. The semantics are kept similar to regular allocations, and as much of the code as possible is shared. Signed-off-by: Anatoly Burakov --- lib/librte_eal/common/eal_common_memzone.c | 237 +++- lib/librte_eal/common/inc

[dpdk-dev] [PATCH v1] lib/metrics: add check for invalid metric keys

This patchset adds a check to rte_metrics_update_values() that prevents the updating of metrics when presented with an invalid metric key. Previously, doing the latter could result in a crash. Fixes: 349950ddb9c5 ("metrics: add information metrics library") Signed-off-by: Remy Horton --- lib/li

[dpdk-dev] [PATCH v5 1/4] doc: cleanup ISA-L PMD feature matrix

In PMD feature matrices (.ini files), it is not required to have the list of features that are not supported, just the ones that are. Signed-off-by: Pablo de Lara Acked-by: Lee Daly --- v5: - Removed "HW Accelerated" from isa-l feature list v4: - No change doc/guides/compressdevs/features/is

[dpdk-dev] [PATCH v5 4/4] compressdev: add huffman encoding flags

Added Huffman fixed and dynamic encoding feature flags, so an application can query if a device supports these two types, when performing DEFLATE compression. Signed-off-by: Pablo de Lara Acked-by: Fiona Trahe --- v5-v3: - No change v2: - Fixed typo drivers/compress/isal/isal_compress_pmd_op

[dpdk-dev] [PATCH v5 3/4] compressdev: replace mbuf scatter gather flag

The current mbuf scatter gather feature flag is too ambiguous, as it is not clear if input and/or output buffers can be scatter gather mbufs or not. Therefore, three new flags will replace this flag: - RTE_COMP_FF_OOP_SGL_IN_SGL_OUT - RTE_COMP_FF_OOP_SGL_IN_FB_OUT - RTE_COMP_FF_OOP_LB_IN_SGL_OUT

[dpdk-dev] [PATCH v5 2/4] doc: rename compress feature flag

Renamed feature "Bypass" to "Pass-through", as it is a more explicit name, meaning that the PMD is capable of passing the mbufs through it, without making any modifications (i.e.. NULL algorithm). Signed-off-by: Pablo de Lara Acked-by: Fiona Trahe Acked-by: Shally Verma --- v5: - No change v4

[dpdk-dev] [PATCH] cryptodev: rename experimental private data APIs

The name private_data is confusing in these APIs: rte_cryptodev_sym_session_set_private_data() rte_cryptodev_sym_session_get_private_data() It refers to data added at the end of the session hdr for use by the application. The session already contains sess_private_data[index] which is used to store

Re: [dpdk-dev] [PATCH v4 2/4] cryptodev: support asymmetric operations

Hi Shally, Umesh, > -Original Message- > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Shally Verma > Sent: Tuesday, July 3, 2018 4:24 PM > To: De Lara Guarch, Pablo > Cc: dev@dpdk.org; pathr...@caviumnetworks.com; nmur...@caviumnetworks.com; > Umesh Kartha > ; Sunila Sahu > ; As

Re: [dpdk-dev] [PATCH v4 2/4] cryptodev: support asymmetric operations

Hi Fiona >-Original Message- >From: Trahe, Fiona [mailto:fiona.tr...@intel.com] >Sent: 06 July 2018 19:11 >To: Verma, Shally ; De Lara Guarch, Pablo > >Cc: dev@dpdk.org; Athreya, Narayana Prasad >; Murthy, Nidadavolu >; Kartha, Umesh ; Sahu, >Sunila ; Gupta, >Ashish ; Trahe, Fiona >Sub

Re: [dpdk-dev] [PATCH v4 2/4] cryptodev: support asymmetric operations

Hi Shally > Ya . right now they were there for consistency. You prefer to remove them? Yes.

[dpdk-dev] [PATCH v9 00/19] enable hotplug on multi-process

v9: - Move hotplug IPC from rte_eth_dev_attach/rte_eth_dev_detach to eal_dev_hotplug_add and eal_dev_hotplug_remove, now all kinds of devices will be synced in multi-process. - Fix couple issue when a device is bound to vfio. 1) The device can't be detached clearly in a secondary process, whi

[dpdk-dev] [PATCH v9 09/19] net/af_packet: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/af_packet/rte_eth_af_packet.c | 11 +++

[dpdk-dev] [PATCH v9 02/19] bus/pci: fix PCI address compare

When use memcmp to compare two PCI address, sizeof(struct rte_pci_addr) is 4 bytes aligned, and it is 8. While only 7 byte of struct rte_pci_addr is valid. So compare the 8th byte will cause the unexpected result, which happens when repeatedly attach/detach a device. Fixes: c752998b5e2e ("pci: int

[dpdk-dev] [PATCH v9 08/19] net/ixgbe: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/ixgbe/ixgbe_ethdev.c | 3 +++ 1 file changed, 3

[dpdk-dev] [PATCH v9 04/19] vfio: remove uneccessary IPC for group fd clear

Clear vfio_group_fd is not necessary to involve any IPC. Also, current IPC implementation for SOCKET_CLR_GROUP is not correct. rte_vfio_clear_group on secondary will always fail, that prevent device be detached correctly on a secondary process. The patch simply removes all IPC related stuff in rte_

[dpdk-dev] [PATCH v9 07/19] net/i40e: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/i40e/i40e_ethdev.c | 2 ++ 1 file changed, 2 ins

[dpdk-dev] [PATCH v9 05/19] eal: enable hotplug on multi-process

We are going to introduce the solution to handle hotplug in multi-process, it includes the below scenario: 1. Attach a device from the primary 2. Detach a device from the primary 3. Attach a device from a secondary 4. Detach a device from a secondary In the primary-secondary process model, we ass

[dpdk-dev] [PATCH v9 03/19] bus/pci: enable vfio unmap resource for secondary

Subroutine to unmap VFIO resource is shared by secondary and primary, and it does not work on the secondary process. The patch adds a dedicate function to handle the situation when a device is unmapped on a secondary process. Signed-off-by: Qi Zhang --- drivers/bus/pci/linux/pci_vfio.c | 75

[dpdk-dev] [PATCH v9 01/19] ethdev: add function to release port in local process

Add driver API rte_eth_release_port_private to support the case when an ethdev need to be detached on a secondary process. Local state is set to unused and shared data will not be reset so the primary process can still use it. Signed-off-by: Qi Zhang Reviewed-by: Andrew Rybchenko Acked-by: Remy

[dpdk-dev] [PATCH v9 11/19] net/kni: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/kni/rte_eth_kni.c | 11 +++ 1 file chang

[dpdk-dev] [PATCH v9 06/19] eal: support attach or detach share device from secondary

This patch cover the multi-process hotplug case when a device attach/detach request be issued from a secondary process device attach on secondary: a) secondary send sync request to the primary. b) primary receive the request and attach the new device if failed goto i). c) primary forward attach

[dpdk-dev] [PATCH v9 12/19] net/null: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/null/rte_eth_null.c | 16 +++- 1 fil

[dpdk-dev] [PATCH v9 13/19] net/octeontx: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/octeontx/octeontx_ethdev.c | 16

[dpdk-dev] [PATCH v9 14/19] net/pcap: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/pcap/rte_eth_pcap.c | 15 ++- 1 file

[dpdk-dev] [PATCH v9 10/19] net/bonding: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/bonding/rte_eth_bond_pmd.c | 11 +++ 1 f

[dpdk-dev] [PATCH v9 15/19] net/softnic: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/softnic/rte_eth_softnic.c | 19 -

[dpdk-dev] [PATCH v9 17/19] net/vhost: enable port detach on secondary process

Previously, detach port on a secondary process will mess primary process and cause the same device can't be attached back again. A secondary process should use rte_eth_release_port_private to release a port. Signed-off-by: Qi Zhang --- drivers/net/vhost/rte_eth_vhost.c | 11 +++ 1 file c

[dpdk-dev] [PATCH v9 18/19] examples/multi_process: add hotplug sample

The sample code demonstrates device (ethdev only) management at a multi-process environment. The user can attach/detach a device on primary process and see it is synced on secondary process automatically. How to start? ./hotplug_mp --proc-type=auto Command Line Example: >help >list /* attach a

  1   2   >