RE: [PATCH v2 1/4] net/i40e: fix buffer leak on Rx reconfiguration

[Zhang, Qi Z]

> -Original Message-
> From: Richardson, Bruce 
> Sent: Thursday, August 31, 2023 8:34 PM
> To: dev@dpdk.org
> Cc: Richardson, Bruce ; Zhang, Qi Z
> ; sta...@dpdk.org
> Subject: [PATCH v2 1/4] net/i40e: fix buffer leak on Rx reconfiguration
> 
> When reconfiguring a single queue on a device, the mbuf initializer value was
> not getting set, and remained at zero. This lead to mbuf leaks as the refcount
> was incorrect (0), so on free it wrapped around to UINT16_MAX. When
> setting up the mbuf initializer, also ensure that the queue is explicitly 
> marked
> as using a vector function by setting the "rx_using_sse" flag.
> 
> Fixes: a3c83a2527e1 ("net/i40e: enable runtime queue setup")
> Cc: qi.z.zh...@intel.com
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Bruce Richardson 
> ---
>  drivers/net/i40e/i40e_rxtx.c| 6 ++
>  drivers/net/i40e/i40e_rxtx_vec_common.h | 1 +
>  2 files changed, 7 insertions(+)
> 
> diff --git a/drivers/net/i40e/i40e_rxtx.c b/drivers/net/i40e/i40e_rxtx.c index
> b4f65b58fa..d96677 100644
> --- a/drivers/net/i40e/i40e_rxtx.c
> +++ b/drivers/net/i40e/i40e_rxtx.c
> @@ -1918,6 +1918,12 @@ i40e_dev_rx_queue_setup_runtime(struct
> rte_eth_dev *dev,
>   if (use_def_burst_func)
>   ad->rx_bulk_alloc_allowed = false;
>   i40e_set_rx_function(dev);
> +
> + if (ad->rx_vec_allowed && i40e_rxq_vec_setup(rxq)) {
> + PMD_DRV_LOG(ERR, "Failed vector rx setup.");
> + return -EINVAL;
> + }
> +
>   return 0;
>   } else if (ad->rx_vec_allowed && !rte_is_power_of_2(rxq-
> >nb_rx_desc)) {
>   PMD_DRV_LOG(ERR, "Vector mode is allowed, but
> descriptor"
> diff --git a/drivers/net/i40e/i40e_rxtx_vec_common.h
> b/drivers/net/i40e/i40e_rxtx_vec_common.h
> index fe1a6ec75e..8b745630e4 100644
> --- a/drivers/net/i40e/i40e_rxtx_vec_common.h
> +++ b/drivers/net/i40e/i40e_rxtx_vec_common.h
> @@ -201,6 +201,7 @@ i40e_rxq_vec_setup_default(struct i40e_rx_queue
> *rxq)
>   rte_compiler_barrier();
>   p = (uintptr_t)&mb_def.rearm_data;
>   rxq->mbuf_initializer = *(uint64_t *)p;
> + rxq->rx_using_sse = 1;

I saw some further code clean can be done by leveraging this change.

>   return 0;
>  }
> 
> --
> 2.39.2

Acked-by: Qi Zhang 

Applied to dpdk-next-net-intel.

Thanks
Qi

Re: [EXT] Re: [PATCH v5 1/1] devtools: add tracepoint check in checkpatch

[Thomas Monjalon]

Let's decide in a techboard meeting whether traces are mandatory or not.


01/09/2023 04:32, Jerin Jacob:
> On Thu, Aug 31, 2023 at 12:08 AM Morten Brørup  
> wrote:
> >
> > > From: Thomas Monjalon [mailto:tho...@monjalon.net]
> > > Sent: Wednesday, 30 August 2023 18.24
> > >
> > > 21/08/2023 16:46, Morten Brørup:
> > > > > From: Ankur Dwivedi [mailto:adwiv...@marvell.com]
> > > > > Sent: Monday, 21 August 2023 15.54
> 
> > >
> > > In general, I wonder how much the check is useful compared to the
> > > complexity.
> > >
> > >
> > > > The bigger question is: Do we really want to change tracepoints in
> > > functions from opt-in to opt-out?
> > >
> > > Yes that's the question: should traces be mandatory in some libs?
> > >
> > > > In my opinion, opt-in for trace is more appropriate.
> > >
> > > There was some work to add traces everywhere in few libs, so why not
> > > maintaining this state?
> >
> > I still think it's a silly and burdening requirement, but I'm not against 
> > it for the libs where it is already a de facto standard.
> >
> > 
> > I can imagine similar requirements regarding logging, telemetry and dump 
> > functions being imposed on select libs.
> >
> > Perhaps we should also require that new libs implement all four types of 
> > instrumentation, to ensure that only high quality (i.e. fully instrumented) 
> > libs are accepted into DPDK.
> 
> IMO, There is a lot of effort to put trace points to existing
> libraries, without these check, soon the disparity shows up when
> someone adds new APIs to library and forget
> to add trace points.
> 
> It is pretty easy to add trace point and there are a lot of
> references, so I don't think there is burden  for a developer
> comparing to the effort of adding a new API.
> 
> Most of the time, contributors forgets to add trace point because
> there is no automatic way to find it.
> Also, additional git commits are needs if some decide to add it later.
> 
> No strong opinion, If not find not useful, we could mark this patch is
> not appliable. Keeping the patch is limbo state is the issue. It is
> already reached to v5.
> So please decide one way or another.
> 
> 
> 
> 
> > 
> >
> > > I don't really like adding a skip list as one more burden for future
> > > authors.
> >
> > If the warning omitted from checkpatches refers to the skip list and its 
> > location, it is relatively easy for developers to manage.
> >
> > And reviewers will notice if new functions have been added to the skip 
> > list, indicating that trace has been omitted. So there are also advantages 
> > to the skip list.

Re: [PATCH v3 0/3] Release ethdev shared memory on port cleanup

[David Marchand]

On Thu, Aug 31, 2023 at 6:14 PM Thomas Monjalon  wrote:
>
> 21/08/2023 10:58, David Marchand:
> > This series was triggered after investigating why the
> > eal_flags_file_prefix_autotest unit test was failing in the case of
> > statically built binaries [1]).
> >
> > For now, I went with a simple (naive) approach and put all accesses to the
> > shared data under a single lock: ethdev maintainers, it is your turn to
> > shine and give me reasons why we should keep the locks the way they
> > were ;-).
>
> I think the reasons are:
> - we wanted to call rte_spinlock_init()
> - we didn't want to allocate an ethdev lock in EAL memory config

Hum, it is a choice of implementation, not a list of locking requirements.

As to why we would not want ethdev lock in EAL, I can understand the concern.
This could be enhanced with a new service provided by EAL to register
some space in the shared configuration for use by libraries.
However, seeing how the mempool and timer libraries already had one
lock stored there, I assumed it was ok to do the same.


> How eliminating a lock is making the last patch easier exactly?

Let's say that my goal is to cleanup resources once a DPDK application
exits (no hugepage files left).
Here, it is a first step in that direction, with releasing ethdev
shared mem data.
It is not a complete solution, as other device classes probably have
the same issues of leaving some shared data behind.


The ethdev shared memory hosts ethdev ports, and the ownership notion.
The current code implicitly relies on the assumption that the shared
memory will never go away.
So I had to revisit and protect places accessing this shared memory,
and having one shared lock for protecting access was necessary.
But doing so, the ownership lock would be nested in this global lock
while doing nothing more that the ethdev shared data lock.

I will enhance the patches description.


-- 
David Marchand

[PATCH] bus/vdev: fix devargs memory leak

[Mingjin Ye]

When a device is created by a secondary process, an empty devargs is
temporarily generated and bound to it. This causes the device to not
be associated with the correct devargs, and the empty devargs are not
released when the resource is freed.

This patch fixes the issue by matching the devargs when inserting a
device in secondary process.

Fixes: dda987315ca2 ("vdev: make virtual bus use its device struct")
Fixes: a16040453968 ("eal: extract vdev infra")
Cc: sta...@dpdk.org

Signed-off-by: Mingjin Ye 
---
 drivers/bus/vdev/vdev.c | 22 +-
 1 file changed, 21 insertions(+), 1 deletion(-)

diff --git a/drivers/bus/vdev/vdev.c b/drivers/bus/vdev/vdev.c
index 7974b27295..fe39a98a9c 100644
--- a/drivers/bus/vdev/vdev.c
+++ b/drivers/bus/vdev/vdev.c
@@ -259,6 +259,22 @@ alloc_devargs(const char *name, const char *args)
return devargs;
 }
 
+static struct rte_devargs *
+vdev_devargs_lookup(const char *name)
+{
+   struct rte_devargs *devargs;
+   char dev_name[32];
+
+   RTE_EAL_DEVARGS_FOREACH("vdev", devargs) {
+   devargs->bus->parse(devargs->name, &dev_name);
+   if (strcmp(dev_name, name) == 0) {
+   VDEV_LOG(INFO, "**Devargs matched %s", dev_name);
+   return devargs;
+   }
+   }
+   return NULL;
+}
+
 static int
 insert_vdev(const char *name, const char *args,
struct rte_vdev_device **p_dev,
@@ -271,7 +287,11 @@ insert_vdev(const char *name, const char *args,
if (name == NULL)
return -EINVAL;
 
-   devargs = alloc_devargs(name, args);
+   if (rte_eal_process_type() == RTE_PROC_PRIMARY)
+   devargs = alloc_devargs(name, args);
+   else
+   devargs = vdev_devargs_lookup(name);
+
if (!devargs)
return -ENOMEM;
 
-- 
2.25.1

RE: [PATCH 3/4] net/cpfl: introduce CPF common library

[Qiao, Wenjing]

Although it appears that 'hw' is being used, it is not actually being used. 
Look at the following macros which use 'hw':
#define idpf_calloc(h, c, s)rte_zmalloc(NULL, (c) * (s), 0)
#define idpf_free(h, m) rte_free(m)

If I remove '__rte_unused', it will show -Wunused-parameter warning when 
compiling.

-Original Message-
From: Xing, Beilei  
Sent: Thursday, August 24, 2023 5:20 PM
To: Qiao, Wenjing ; Zhang, Yuying 

Cc: dev@dpdk.org; Liu, Mingxia 
Subject: RE: [PATCH 3/4] net/cpfl: introduce CPF common library



> -Original Message-
> From: Qiao, Wenjing 
> Sent: Friday, August 11, 2023 5:31 PM
> To: Zhang, Yuying ; Xing, Beilei 
> 
> Cc: dev@dpdk.org; Liu, Mingxia ; Qiao, Wenjing 
> 
> Subject: [PATCH 3/4] net/cpfl: introduce CPF common library
> 
> Add common library support for CPFL rte_flow to create/delete rules.
> 
> Signed-off-by: Wenjing Qiao 
> ---


> +int
> +cpfl_ctlq_alloc_ring_res(struct idpf_hw *hw __rte_unused, struct

hw is used, so remove __rte_unused.
Please check other functions.

> idpf_ctlq_info *cq,
> +  struct cpfl_ctlq_create_info *qinfo) {

[PATCH v4 01/32] net/sssnic: add build and doc infrastructure

[wanry]

From: Renyong Wan 

Adding minimum PMD code, doc and build infrastructure for sssnic.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed 'Title underline too short' in doc/guides/nics/sssnic.rst.
* Removed error.h from including files.
---
 .mailmap|  5 +-
 MAINTAINERS |  8 
 doc/guides/nics/features/sssnic.ini |  9 
 doc/guides/nics/index.rst   |  1 +
 doc/guides/nics/sssnic.rst  | 73 +
 drivers/net/meson.build |  1 +
 drivers/net/sssnic/meson.build  | 18 +++
 drivers/net/sssnic/sssnic_ethdev.c  | 28 +++
 8 files changed, 140 insertions(+), 3 deletions(-)
 create mode 100644 doc/guides/nics/features/sssnic.ini
 create mode 100644 doc/guides/nics/sssnic.rst
 create mode 100644 drivers/net/sssnic/meson.build
 create mode 100644 drivers/net/sssnic/sssnic_ethdev.c

diff --git a/.mailmap b/.mailmap
index 864d33ee46..6fa73d3b79 100644
--- a/.mailmap
+++ b/.mailmap
@@ -151,7 +151,6 @@ Bao-Long Tran 
 Bar Neuman 
 Barak Enat 
 Barry Cao 
-Bartosz Staszewski 
 Baruch Siach 
 Bassam Zaid AlKilani 
 Beilei Xing 
@@ -496,7 +495,6 @@ Helin Zhang 
 Hemant Agrawal  
 Heng Ding 
 Hengjian Zhang 
-Heng Jiang 
 Heng Wang 
 Henning Schild 
 Henry Cai 
@@ -630,7 +628,6 @@ Jie Liu 
 Jie Pan 
 Jie Wang 
 Jie Zhou  
-Jieqiang Wang 
 Jijiang Liu 
 Jilei Chen 
 Jim Harris 
@@ -1156,6 +1153,7 @@ Rebecca Troy 
 Remi Pommarel 
 Remy Horton 
 Renata Saiakhova 
+Renyong Wan 
 Reshma Pattan  
 Ricardo Roldan 
 Ricardo Salveti 
@@ -1329,6 +1327,7 @@ Stephen Hurd  

 Steve Capper 
 Steven Lariau 
 Steven Luong 
+Steven Song 
 Steven Webster 
 Steven Zou 
 Steve Rempe 
diff --git a/MAINTAINERS b/MAINTAINERS
index a926155f26..1e57d29aa3 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -633,6 +633,13 @@ F: drivers/net/af_xdp/
 F: doc/guides/nics/af_xdp.rst
 F: doc/guides/nics/features/af_xdp.ini
 
+3SNIC sssnic
+M: Renyong Wan 
+M: Steven Song 
+F: driver/net/sssnic/
+F: doc/guides/nics/sssnic.rst
+F: doc/guides/nics/features/sssnic.ini
+
 Amazon ENA
 M: Michal Krawczyk 
 M: Shai Brandes 
@@ -1793,6 +1800,7 @@ F: doc/guides/tools/img/eventdev_*
 F: app/test/test_event_ring.c
 
 Procinfo tool
+M: Maryam Tahhan 
 M: Reshma Pattan 
 F: app/proc-info/
 F: doc/guides/tools/proc_info.rst
diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
new file mode 100644
index 00..6d9786db7e
--- /dev/null
+++ b/doc/guides/nics/features/sssnic.ini
@@ -0,0 +1,9 @@
+;
+; Supported features of the 'sssnic' network poll mode driver.
+;
+; Refer to default.ini for the full list of available PMD features.
+;
+[Features]
+Linux= Y
+ARMv8= Y
+x86-64   = Y
diff --git a/doc/guides/nics/index.rst b/doc/guides/nics/index.rst
index 7bfcac880f..9d2b29383b 100644
--- a/doc/guides/nics/index.rst
+++ b/doc/guides/nics/index.rst
@@ -61,6 +61,7 @@ Network Interface Controller Drivers
 qede
 sfc_efx
 softnic
+sssnic
 tap
 thunderx
 txgbe
diff --git a/doc/guides/nics/sssnic.rst b/doc/guides/nics/sssnic.rst
new file mode 100644
index 00..fe0180c2e6
--- /dev/null
+++ b/doc/guides/nics/sssnic.rst
@@ -0,0 +1,73 @@
+..  SPDX-License-Identifier: BSD-3-Clause
+Copyright(c) 2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+
+SSSNIC Poll Mode Driver
+===
+
+The sssnic PMD (**librte_pmd_sssnic**) provides poll mode driver support
+for 3SNIC 9x0 serials family of Ethernet adapters.
+
+
+Supported NICs
+--
+
+- 3S910 Dual Port SFP28 10/25GbE Ethernet adapter
+- 3S920 Quad Port SFP28 10/25GbE Ethernet adapter
+- 3S920 Quad Port QSFP28 100GbE Ethernet adapter
+
+
+Features
+
+
+Features of sssnic PMD are:
+
+- Link status
+- Link status event
+- Queue start/stop
+- Rx interrupt
+- Scattered Rx
+- TSO
+- LRO
+- Promiscuous mode
+- Allmulticast mode
+- Unicast MAC filter
+- Multicast MAC filte
+- RSS hash
+- RSS key update
+- RSS reta update
+- Inner RSS
+- VLAN filter
+- VLAN offload
+- L3 checksum offload
+- L4 checksum offload
+- Inner L3 checksum
+- Inner L4 checksum
+- Basic stats
+- Extended stats
+- Stats per queue
+- Flow control
+- FW version
+- Generic flow API
+
+
+Prerequisites
+-
+
+- Learning about 3SNIC Ethernet NICs using
+  ``_.
+
+- Follow the DPDK :ref:`Getting Started Guide for Linux ` to setup 
the basic DPDK environment.
+
+
+Driver compilation and testing
+--
+
+Refer to the document :ref:`compiling and testing a PMD for a NIC 
`
+for details.
+
+
+Limitations or Known issues
+---
+
+Build with ICC is not supported yet.
+Power8, ARMv7 and BSD are not supported yet.
diff --git a/drivers/net/meson.build b/drivers/net/meson.build
index bd38b533c5..224eab99a7 100644
--- a/drivers/net/meson.build
+++ b/drivers/net/meson.build
@@ -54,6 +54,7 @@ drivers = [

[PATCH v4 02/32] net/sssnic: add log type and log macros

[wanry]

From: Renyong Wan 

Adding log macros to print runtime messages and trace functions.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/sssnic_ethdev.c | 13 
 drivers/net/sssnic/sssnic_log.h| 51 ++
 2 files changed, 64 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_log.h

diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index dcda01eeb8..0f1017af9d 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -5,11 +5,14 @@
 #include 
 #include 
 
+#include "sssnic_log.h"
+
 static int
 sssnic_pci_probe(struct rte_pci_driver *pci_drv, struct rte_pci_device 
*pci_dev)
 {
RTE_SET_USED(pci_drv);
RTE_SET_USED(pci_dev);
+   PMD_INIT_FUNC_TRACE();
return -EINVAL;
 }
 
@@ -17,6 +20,7 @@ static int
 sssnic_pci_remove(struct rte_pci_device *pci_dev)
 {
RTE_SET_USED(pci_dev);
+   PMD_INIT_FUNC_TRACE();
return -EINVAL;
 }
 
@@ -26,3 +30,12 @@ static struct rte_pci_driver sssnic_pmd = {
 };
 
 RTE_PMD_REGISTER_PCI(net_sssnic, sssnic_pmd);
+
+RTE_LOG_REGISTER_SUFFIX(sssnic_logtype_driver, driver, INFO);
+RTE_LOG_REGISTER_SUFFIX(sssnic_logtype_init, init, NOTICE);
+#ifdef RTE_ETHDEV_DEBUG_RX
+RTE_LOG_REGISTER_SUFFIX(sssnic_logtype_rx, rx, DEBUG);
+#endif /*RTE_ETHDEV_DEBUG_RX*/
+#ifdef RTE_ETHDEV_DEBUG_TX
+RTE_LOG_REGISTER_SUFFIX(sssnic_logtype_tx, tx, DEBUG);
+#endif /*RTE_ETHDEV_DEBUG_TX*/
diff --git a/drivers/net/sssnic/sssnic_log.h b/drivers/net/sssnic/sssnic_log.h
new file mode 100644
index 00..629e12100c
--- /dev/null
+++ b/drivers/net/sssnic/sssnic_log.h
@@ -0,0 +1,51 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#ifndef _SSSNIC_LOG_H_
+#define _SSSNIC_LOG_H_
+
+#include 
+
+extern int sssnic_logtype_driver;
+extern int sssnic_logtype_init;
+
+#define SSSNIC_LOG_NAME "sssnic"
+#define PMD_DRV_LOG(level, fmt, args...)   
\
+   rte_log(RTE_LOG_##level, sssnic_logtype_driver,\
+   SSSNIC_LOG_NAME ": " fmt "\n", ##args)
+#define PMD_INIT_LOG(level, fmt, args...)  
\
+   rte_log(RTE_LOG_##level, sssnic_logtype_init, "%s(): " fmt "\n",   \
+   __func__, ##args)
+
+#define SSSNIC_DEBUG(fmt, args...) 
\
+   PMD_DRV_LOG(DEBUG, "[%s():%d] " fmt, __func__, __LINE__, ##args)
+
+/*
+ * Trace driver init and uninit.
+ */
+#define PMD_INIT_FUNC_TRACE() PMD_INIT_LOG(DEBUG, " >>")
+
+#ifdef RTE_ETHDEV_DEBUG_RX
+extern int sssnic_logtype_rx;
+#define SSSNIC_RX_LOG(level, fmt, args...) 
\
+   rte_log(RTE_LOG_##level, sssnic_logtype_rx,\
+   "sssnic_rx: [%s():%d] " fmt "\n", __func__, __LINE__, ##args)
+#else
+#define SSSNIC_RX_LOG(level, fmt, args...) 
\
+   do {   \
+   } while (0)
+#endif /*RTE_ETHDEV_DEBUG_RX*/
+
+#ifdef RTE_ETHDEV_DEBUG_TX
+extern int sssnic_logtype_tx;
+#define SSSNIC_TX_LOG(level, fmt, args...) 
\
+   rte_log(RTE_LOG_##level, sssnic_logtype_rx,\
+   "sssnic_tx: [%s():%d] " fmt "\n", __func__, __LINE__, ##args)
+#else
+#define SSSNIC_TX_LOG(level, fmt, args...) 
\
+   do {   \
+   } while (0)
+#endif /*RTE_ETHDEV_DEBUG_TX*/
+
+#endif /*_SSSNIC_LOG_H_*/
-- 
2.27.0

[PATCH v4 04/32] net/sssnic: initialize hardware base

[wanry]

From: Renyong Wan 

Initializing hardware base make hardware ready to be access.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/meson.build  |  13 ++
 drivers/net/sssnic/base/sssnic_hw.c  | 207 +++
 drivers/net/sssnic/base/sssnic_hw.h  |  49 +++
 drivers/net/sssnic/base/sssnic_reg.h | 169 ++
 drivers/net/sssnic/meson.build   |   3 +
 drivers/net/sssnic/sssnic_ethdev.c   |  46 +-
 drivers/net/sssnic/sssnic_ethdev.h   |  18 +++
 7 files changed, 501 insertions(+), 4 deletions(-)
 create mode 100644 drivers/net/sssnic/base/meson.build
 create mode 100644 drivers/net/sssnic/base/sssnic_hw.c
 create mode 100644 drivers/net/sssnic/base/sssnic_reg.h
 create mode 100644 drivers/net/sssnic/sssnic_ethdev.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
new file mode 100644
index 00..3e64112c72
--- /dev/null
+++ b/drivers/net/sssnic/base/meson.build
@@ -0,0 +1,13 @@
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+
+sources = [
+'sssnic_hw.c',
+]
+
+c_args = cflags
+base_lib = static_library('sssnic_base', sources,
+   dependencies: [static_rte_eal, static_rte_ethdev, static_rte_bus_pci, 
static_rte_net],
+   c_args: c_args)
+
+base_objs = base_lib.extract_all_objects()
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
new file mode 100644
index 00..8b7bba7644
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -0,0 +1,207 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_hw.h"
+#include "sssnic_reg.h"
+
+static int
+wait_for_sssnic_hw_ready(struct sssnic_hw *hw)
+{
+   struct sssnic_attr1_reg reg;
+   uint32_t timeout_ms = 10;
+
+   do {
+   reg.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR1_REG);
+   if (reg.u32 != 0x && reg.mgmt_init_status != 0)
+   return 0;
+   rte_delay_ms(1);
+   } while (--timeout_ms);
+
+   return -EBUSY;
+}
+
+static int
+wait_for_sssnic_db_enabled(struct sssnic_hw *hw)
+{
+   struct sssnic_attr4_reg r4;
+   struct sssnic_attr5_reg r5;
+   uint32_t timeout_ms = 6;
+
+   do {
+   r4.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR4_REG);
+   r5.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR5_REG);
+   if (r4.db_ctrl == SSSNIC_DB_CTRL_ENABLE &&
+   r5.outbound_ctrl == SSSNIC_DB_CTRL_ENABLE)
+   return 0;
+   rte_delay_ms(1);
+   } while (--timeout_ms);
+
+   return -EBUSY;
+}
+
+static void
+sssnic_attr_setup(struct sssnic_hw *hw)
+{
+   struct sssnic_attr0_reg attr0;
+   struct sssnic_attr1_reg attr1;
+   struct sssnic_attr2_reg attr2;
+   struct sssnic_attr3_reg attr3;
+   struct sssnic_hw_attr *attr = &hw->attr;
+
+   attr0.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR0_REG);
+   attr1.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR1_REG);
+   attr2.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR2_REG);
+   attr3.u32 = sssnic_cfg_reg_read(hw, SSSNIC_ATTR3_REG);
+
+   attr->func_idx = attr0.func_idx;
+   attr->pf_idx = attr0.pf_idx;
+   attr->pci_idx = attr0.pci_idx;
+   attr->vf_off = attr0.vf_off;
+   attr->func_type = attr0.func_type;
+   attr->af_idx = attr1.af_idx;
+   attr->num_aeq = RTE_BIT32(attr1.num_aeq);
+   attr->num_ceq = attr2.num_ceq;
+   attr->num_irq = attr2.num_irq;
+   attr->global_vf_off = attr3.global_vf_off;
+
+   PMD_DRV_LOG(DEBUG, "attr0=0x%x, attr1=0x%x, attr2=0x%x, attr3=0x%x",
+   attr0.u32, attr1.u32, attr2.u32, attr3.u32);
+}
+
+/* AF and MF election */
+static void
+sssnic_af_setup(struct sssnic_hw *hw)
+{
+   struct sssnic_af_election_reg reg0;
+   struct sssnic_mf_election_reg reg1;
+
+   /* AF election */
+   reg0.u32 = sssnic_mgmt_reg_read(hw, SSSNIC_AF_ELECTION_REG);
+   reg0.func_idx = hw->attr.func_idx;
+   sssnic_mgmt_reg_write(hw, SSSNIC_AF_ELECTION_REG, reg0.u32);
+   reg0.u32 = sssnic_mgmt_reg_read(hw, SSSNIC_AF_ELECTION_REG);
+   hw->attr.af_idx = reg0.func_idx;
+   if (hw->attr.af_idx == hw->attr.func_idx) {
+   hw->attr.func_type = SSSNIC_FUNC_TYPE_AF;
+   PMD_DRV_LOG(INFO, "Elected PF %d as AF", hw->attr.func_idx);
+
+   /* MF election */
+   reg1.u32 = sssnic_mgmt_reg_read(hw, SSSNIC_MF_ELECTION_REG);
+   reg1.func_idx = hw->attr.func_idx;
+   sssnic_mgmt_reg_write(hw, SSSNIC_MF_ELECTION_REG, reg1.u32);
+   reg1.u32 = sssnic_mgmt_reg_read(hw, SSSNIC_MF_ELECTION_REG);
+   hw->attr.mf_idx = reg1.func_idx;
+   if (hw->at

[PATCH v4 03/32] net/sssnic: support probe and remove

[wanry]

From: Renyong Wan 

Register device ID for 3SNIC ethernet adapter to support
PCI ethdev probe and remove.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/sssnic_hw.h | 11 +
 drivers/net/sssnic/sssnic_ethdev.c  | 37 +
 2 files changed, 44 insertions(+), 4 deletions(-)
 create mode 100644 drivers/net/sssnic/base/sssnic_hw.h

diff --git a/drivers/net/sssnic/base/sssnic_hw.h 
b/drivers/net/sssnic/base/sssnic_hw.h
new file mode 100644
index 00..db916b1977
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_hw.h
@@ -0,0 +1,11 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#ifndef _SSSNIC_HW_H_
+#define _SSSNIC_HW_H_
+
+#define SSSNIC_PCI_VENDOR_ID 0x1F3F
+#define SSSNIC_DEVICE_ID_STD 0x9020
+
+#endif /* _SSSNIC_HW_H_ */
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 0f1017af9d..4f8b5c2684 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -6,25 +6,54 @@
 #include 
 
 #include "sssnic_log.h"
+#include "base/sssnic_hw.h"
+
+static int
+sssnic_ethdev_init(struct rte_eth_dev *ethdev)
+{
+   RTE_SET_USED(ethdev);
+   PMD_INIT_FUNC_TRACE();
+
+   return -EINVAL;
+}
+
+static int
+sssnic_ethdev_uninit(struct rte_eth_dev *ethdev)
+{
+   RTE_SET_USED(ethdev);
+   PMD_INIT_FUNC_TRACE();
+
+   if (rte_eal_process_type() != RTE_PROC_PRIMARY)
+   return 0;
+
+   return -EINVAL;
+}
 
 static int
 sssnic_pci_probe(struct rte_pci_driver *pci_drv, struct rte_pci_device 
*pci_dev)
 {
RTE_SET_USED(pci_drv);
-   RTE_SET_USED(pci_dev);
PMD_INIT_FUNC_TRACE();
-   return -EINVAL;
+
+   return rte_eth_dev_pci_generic_probe(pci_dev, 0, sssnic_ethdev_init);
 }
 
 static int
 sssnic_pci_remove(struct rte_pci_device *pci_dev)
 {
-   RTE_SET_USED(pci_dev);
PMD_INIT_FUNC_TRACE();
-   return -EINVAL;
+
+   return rte_eth_dev_pci_generic_remove(pci_dev, sssnic_ethdev_uninit);
 }
 
+static const struct rte_pci_id sssnic_pci_id_map[] = {
+   { RTE_PCI_DEVICE(SSSNIC_PCI_VENDOR_ID, SSSNIC_DEVICE_ID_STD) },
+   { .vendor_id = 0 },
+};
+
 static struct rte_pci_driver sssnic_pmd = {
+   .id_table = sssnic_pci_id_map,
+   .drv_flags = RTE_PCI_DRV_NEED_MAPPING,
.probe = sssnic_pci_probe,
.remove = sssnic_pci_remove,
 };
-- 
2.27.0

[PATCH v4 06/32] net/sssnic/base: add message definition and utility

[wanry]

From: Renyong Wan 

sssnic message is used to encapsulate sssnic command for transmission
between driver and firmware. sssnic message is sent by driver via mail
box and is received by driver via event queue.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 drivers/net/sssnic/base/meson.build |   3 +-
 drivers/net/sssnic/base/sssnic_eventq.c |  29 +++
 drivers/net/sssnic/base/sssnic_hw.c |  15 +-
 drivers/net/sssnic/base/sssnic_hw.h |   2 +
 drivers/net/sssnic/base/sssnic_msg.c| 254 
 drivers/net/sssnic/base/sssnic_msg.h| 166 
 6 files changed, 467 insertions(+), 2 deletions(-)
 create mode 100644 drivers/net/sssnic/base/sssnic_msg.c
 create mode 100644 drivers/net/sssnic/base/sssnic_msg.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index 7758faa482..dd4dd08fc1 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -3,7 +3,8 @@
 
 sources = [
 'sssnic_hw.c',
-'sssnic_eventq.c'
+'sssnic_eventq.c',
+'sssnic_msg.c',
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_eventq.c 
b/drivers/net/sssnic/base/sssnic_eventq.c
index a74b74f756..e90d24bb6b 100644
--- a/drivers/net/sssnic/base/sssnic_eventq.c
+++ b/drivers/net/sssnic/base/sssnic_eventq.c
@@ -14,6 +14,7 @@
 #include "../sssnic_log.h"
 #include "sssnic_hw.h"
 #include "sssnic_reg.h"
+#include "sssnic_msg.h"
 #include "sssnic_eventq.h"
 
 #define SSSNIC_EVENTQ_DEF_DEPTH 64
@@ -184,6 +185,32 @@ sssnic_eventq_ci_update(struct sssnic_eventq *eq, int 
informed)
sssnic_eventq_reg_write(eq, SSSNIC_EVENTQ_CI_CTRL_REG, reg.u32);
 }
 
+static int
+sssnic_event_default_handler_func(struct sssnic_eventq *eq,
+   struct sssnic_event *ev, __rte_unused void *data)
+{
+   struct sssnic_hw *hw;
+   int ret;
+
+   hw = eq->hw;
+   ret = sssnic_msg_rx_handle(hw, (struct sssnic_msg_hdr *)(ev->data));
+   if (ret != SSSNIC_MSG_DONE)
+   return -1;
+
+   return SSSNIC_EVENT_DONE;
+}
+
+static void
+sssnic_eventq_handlers_init(struct sssnic_eventq *eq)
+{
+   int i;
+
+   for (i = SSSNIC_EVENT_CODE_MIN; i <= SSSNIC_EVENT_CODE_MAX; i++) {
+   eq->handlers[i].func = sssnic_event_default_handler_func;
+   eq->handlers[i].data = NULL;
+   }
+}
+
 static int
 sssnic_eventq_init(struct sssnic_hw *hw, struct sssnic_eventq *eq, uint16_t 
qid)
 {
@@ -230,6 +257,8 @@ sssnic_eventq_init(struct sssnic_hw *hw, struct 
sssnic_eventq *eq, uint16_t qid)
PMD_DRV_LOG(ERR, "Failed to setup eventq pages!");
return ret;
}
+
+   sssnic_eventq_handlers_init(eq);
sssnic_eventq_ctrl_setup(eq);
sssnic_eventq_ci_update(eq, 1);
if (qid == 0)
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
index 44e04486a5..387c823c7e 100644
--- a/drivers/net/sssnic/base/sssnic_hw.c
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -10,6 +10,7 @@
 #include "sssnic_hw.h"
 #include "sssnic_reg.h"
 #include "sssnic_eventq.h"
+#include "sssnic_msg.h"
 
 static int
 wait_for_sssnic_hw_ready(struct sssnic_hw *hw)
@@ -197,18 +198,30 @@ sssnic_hw_init(struct sssnic_hw *hw)
return ret;
}
 
+   ret = sssnic_msg_inbox_init(hw);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Failed to initialize message inbox.");
+   return ret;
+   }
+
ret = sssnic_eventq_all_init(hw);
if (ret != 0) {
PMD_DRV_LOG(ERR, "Failed to initialize event queues");
-   return ret;
+   goto eventq_init_fail;
}
 
return -EINVAL;
+
+eventq_init_fail:
+   sssnic_msg_inbox_shutdown(hw);
+   return ret;
 }
 
 void
 sssnic_hw_shutdown(struct sssnic_hw *hw)
 {
PMD_INIT_FUNC_TRACE();
+
sssnic_eventq_all_shutdown(hw);
+   sssnic_msg_inbox_shutdown(hw);
 }
diff --git a/drivers/net/sssnic/base/sssnic_hw.h 
b/drivers/net/sssnic/base/sssnic_hw.h
index 6caf3a6d66..38fb9ac1ac 100644
--- a/drivers/net/sssnic/base/sssnic_hw.h
+++ b/drivers/net/sssnic/base/sssnic_hw.h
@@ -52,11 +52,13 @@ struct sssnic_hw {
uint8_t *db_mem_len;
struct sssnic_hw_attr attr;
struct sssnic_eventq *eventqs;
+   struct sssnic_msg_inbox *msg_inbox;
uint8_t num_eventqs;
uint16_t eth_port_id;
 };
 
 #define SSSNIC_ETH_PORT_ID(hw) ((hw)->eth_port_id)
+#define SSSNIC_MPU_FUNC_IDX 0x1fff
 
 int sssnic_hw_init(struct sssnic_hw *hw);
 void sssnic_hw_shutdown(struct sssnic_hw *hw);
diff --git a/drivers/net/sssnic/base/sssnic_msg.c 
b/drivers/net/sssnic/base/sssnic_msg.c
new file mode 100644
index 00..4b98fee75b
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_msg.c
@@ -0,0 +1,254 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Techn

[PATCH v4 05/32] net/sssnic: add event queue

[wanry]

From: Renyong Wan 

Event queue is intended for receiving event from hardware as well
as mailbox response message.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v4:
* Fixed dereferencing type-punned pointer.
* Fixed coding style issue of COMPLEX_MACRO.
---
 drivers/net/sssnic/base/meson.build |   1 +
 drivers/net/sssnic/base/sssnic_eventq.c | 432 
 drivers/net/sssnic/base/sssnic_eventq.h |  84 +
 drivers/net/sssnic/base/sssnic_hw.c |   9 +-
 drivers/net/sssnic/base/sssnic_hw.h |   5 +
 drivers/net/sssnic/base/sssnic_reg.h|  51 +++
 drivers/net/sssnic/sssnic_ethdev.c  |   1 +
 7 files changed, 582 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/sssnic/base/sssnic_eventq.c
 create mode 100644 drivers/net/sssnic/base/sssnic_eventq.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index 3e64112c72..7758faa482 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -3,6 +3,7 @@
 
 sources = [
 'sssnic_hw.c',
+'sssnic_eventq.c'
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_eventq.c 
b/drivers/net/sssnic/base/sssnic_eventq.c
new file mode 100644
index 00..a74b74f756
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_eventq.c
@@ -0,0 +1,432 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_hw.h"
+#include "sssnic_reg.h"
+#include "sssnic_eventq.h"
+
+#define SSSNIC_EVENTQ_DEF_DEPTH 64
+#define SSSNIC_EVENTQ_NUM_PAGES 4
+#define SSSNIC_EVENTQ_MAX_PAGE_SZ 0x40
+#define SSSNIC_EVENTQ_MIN_PAGE_SZ 0x1000
+
+#define SSSNIC_EVENT_ADDR(base_addr, event_sz, idx)
\
+   ((struct sssnic_event *)(((uint8_t *)(base_addr)) + ((idx) * 
(event_sz
+
+static inline struct sssnic_event *
+sssnic_eventq_peek(struct sssnic_eventq *eq)
+{
+   uint16_t page = eq->ci / eq->page_len;
+   uint16_t idx = eq->ci % eq->page_len;
+
+   return SSSNIC_EVENT_ADDR(eq->pages[page]->addr, eq->entry_size, idx);
+}
+
+static inline void
+sssnic_eventq_reg_write(struct sssnic_eventq *eq, uint32_t reg, uint32_t val)
+{
+   sssnic_cfg_reg_write(eq->hw, reg, val);
+}
+
+static inline uint32_t
+sssnic_eventq_reg_read(struct sssnic_eventq *eq, uint32_t reg)
+{
+   return sssnic_cfg_reg_read(eq->hw, reg);
+}
+
+static inline void
+sssnic_eventq_reg_write64(struct sssnic_eventq *eq, uint32_t reg, uint64_t val)
+{
+   sssnic_cfg_reg_write(eq->hw, reg, (uint32_t)((val >> 16) >> 16));
+   sssnic_cfg_reg_write(eq->hw, reg + sizeof(uint32_t), (uint32_t)val);
+}
+
+/* all eventq registers that to be access must be selected first */
+static inline void
+sssnic_eventq_reg_select(struct sssnic_eventq *eq)
+{
+   sssnic_eventq_reg_write(eq, SSSNIC_EVENTQ_IDX_SEL_REG, eq->qid);
+}
+
+static const struct rte_memzone *
+sssnic_eventq_page_alloc(struct sssnic_eventq *eq, int page_idx)
+{
+   const struct rte_memzone *mz = NULL;
+   char mz_name[RTE_MEMZONE_NAMESIZE];
+
+   snprintf(mz_name, sizeof(mz_name), "sssnic%u_eq%d_page%d",
+   SSSNIC_ETH_PORT_ID(eq->hw), eq->qid, page_idx);
+   mz = rte_memzone_reserve_aligned(mz_name, eq->page_size, SOCKET_ID_ANY,
+   RTE_MEMZONE_IOVA_CONTIG, eq->page_size);
+   return mz;
+}
+
+static uint32_t
+sssnic_eventq_page_size_calc(uint32_t depth, uint32_t entry_size)
+{
+   uint32_t pages = SSSNIC_EVENTQ_NUM_PAGES;
+   uint32_t size;
+
+   size = RTE_ALIGN(depth * entry_size, SSSNIC_EVENTQ_MIN_PAGE_SZ);
+   if (size <= pages * SSSNIC_EVENTQ_MIN_PAGE_SZ) {
+   /* use minimum page size */
+   return SSSNIC_EVENTQ_MIN_PAGE_SZ;
+   }
+
+   /* Calculate how many pages of minimum size page the big size page 
covers */
+   size = RTE_ALIGN(size / pages, SSSNIC_EVENTQ_MIN_PAGE_SZ);
+   pages = rte_fls_u32(size / SSSNIC_EVENTQ_MIN_PAGE_SZ);
+
+   return SSSNIC_EVENTQ_MIN_PAGE_SZ * pages;
+}
+
+static int
+sssnic_eventq_pages_setup(struct sssnic_eventq *eq)
+{
+   const struct rte_memzone *mz;
+   struct sssnic_event *ev;
+   int i, j;
+
+   eq->pages = rte_zmalloc(NULL,
+   eq->num_pages * sizeof(struct rte_memzone *), 1);
+   if (eq->pages == NULL) {
+   PMD_DRV_LOG(ERR, "Could not alloc memory for pages");
+   return -ENOMEM;
+   }
+
+   for (i = 0; i < eq->num_pages; i++) {
+   mz = sssnic_eventq_page_alloc(eq, i);
+   if (mz == NULL) {
+   PMD_DRV_LOG(ERR,
+   "Could not alloc DMA memory for eventq page %d",
+   i);
+   goto alloc_dma_fail;
+   }
+   /* init even

[PATCH v4 08/32] net/sssnic/base: add work queue

[wanry]

From: Renyong Wan 

Work queue is used to maintain hardware queue information by
driver, it is usually used in control queue, rx queue
and tx queue.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 drivers/net/sssnic/base/meson.build|   1 +
 drivers/net/sssnic/base/sssnic_workq.c | 141 +
 drivers/net/sssnic/base/sssnic_workq.h | 108 +++
 3 files changed, 250 insertions(+)
 create mode 100644 drivers/net/sssnic/base/sssnic_workq.c
 create mode 100644 drivers/net/sssnic/base/sssnic_workq.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index 4abd1a0daf..7c23a82ff3 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -6,6 +6,7 @@ sources = [
 'sssnic_eventq.c',
 'sssnic_msg.c',
 'sssnic_mbox.c',
+'sssnic_workq.c',
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_workq.c 
b/drivers/net/sssnic/base/sssnic_workq.c
new file mode 100644
index 00..25b7585246
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_workq.c
@@ -0,0 +1,141 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_workq.h"
+
+/* Consume num_entries and increase CI
+ * Return the first entry address of previous CI position
+ */
+void *
+sssnic_workq_consume(struct sssnic_workq *workq, uint16_t num_entries,
+   uint16_t *ci)
+{
+   void *e;
+   uint16_t current_ci;
+
+   if (workq->idle_entries + num_entries > workq->num_entries)
+   return NULL;
+
+   current_ci = sssnic_workq_ci_get(workq);
+   e = (void *)sssnic_workq_entry_get(workq, current_ci);
+   workq->idle_entries += num_entries;
+   workq->ci += num_entries;
+   if (ci != NULL)
+   *ci = current_ci;
+
+   return e;
+}
+
+/* Produce num_entries and increase pi.
+ * Return the first entry address of previous PI position
+ */
+void *
+sssnic_workq_produce(struct sssnic_workq *workq, uint16_t num_entries,
+   uint16_t *pi)
+{
+   void *e;
+   uint16_t current_pi;
+
+   if (workq->idle_entries < num_entries)
+   return NULL;
+
+   current_pi = sssnic_workq_pi_get(workq);
+   e = (void *)sssnic_workq_entry_get(workq, current_pi);
+   workq->idle_entries -= num_entries;
+   workq->pi += num_entries;
+   if (pi != NULL)
+   *pi = current_pi;
+
+   return e;
+}
+
+static int
+sssnic_workq_init(struct sssnic_workq *workq, const char *name, int socket_id,
+   uint32_t entry_size, uint32_t depth)
+{
+   char zname[RTE_MEMZONE_NAMESIZE];
+
+   if (!rte_is_power_of_2(entry_size)) {
+   PMD_DRV_LOG(ERR,
+   "The entry size(%u) of workq(%s) is not power of 2",
+   entry_size, name);
+   return -EINVAL;
+   }
+
+   if (!rte_is_power_of_2(depth)) {
+   PMD_DRV_LOG(ERR, "The depth(%u) of workq(%s) is not power of 2",
+   depth, name);
+   return -EINVAL;
+   }
+
+   workq->buf_size = entry_size * depth;
+   workq->entry_size = entry_size;
+   workq->entry_shift = rte_log2_u32(entry_size);
+   workq->num_entries = depth;
+   workq->idle_entries = depth;
+   workq->index_mask = depth - 1;
+
+   snprintf(zname, sizeof(zname), "%s_mz", name);
+   workq->buf_mz = rte_memzone_reserve_aligned(zname, workq->buf_size,
+   socket_id, RTE_MEMZONE_IOVA_CONTIG, RTE_PGSIZE_256K);
+   if (workq->buf_mz == NULL) {
+   PMD_DRV_LOG(ERR, "Failed to alloc DMA memory for %s", name);
+   return -ENOMEM;
+   }
+   workq->buf_addr = workq->buf_mz->addr;
+   workq->buf_phyaddr = workq->buf_mz->iova;
+
+   return 0;
+}
+
+static void
+sssnic_workq_cleanup(struct sssnic_workq *workq)
+{
+   if (workq != NULL && workq->buf_mz != NULL)
+   rte_memzone_free(workq->buf_mz);
+}
+
+/* Cleanup a work queue and free it*/
+void
+sssnic_workq_destroy(struct sssnic_workq *workq)
+{
+   if (workq != NULL) {
+   sssnic_workq_cleanup(workq);
+   rte_free(workq);
+   }
+}
+
+/*Create a work queue and initialize*/
+struct sssnic_workq *
+sssnic_workq_new(const char *name, int socket_id, uint32_t entry_size,
+   uint32_t depth)
+{
+   int ret;
+   struct sssnic_workq *workq;
+
+   if (name == NULL) {
+   PMD_DRV_LOG(ERR, "Bad parameter, workq name is NULL");
+   return NULL;
+   }
+
+   workq = rte_zmalloc(name, sizeof(struct sssnic_workq), 0);
+   if (workq == NULL) {
+   PMD_DRV_LOG(ERR, "Could not alloc memory for  %s", name);
+   return NULL;
+   }
+   ret = sssnic_workq_

[PATCH v4 07/32] net/sssnic/base: add mailbox support

[wanry]

From: Renyong Wan 

Mailbox is a message channel used to communicate between PF and
VF as well as driver and hardware functions.
Mailbox messages are received by driver through event queue, and
sent by driver through registers of mailbox.
There are two transfer modes for sending mailbox message, one is
DMA mode used to send message to PF, another is inline mode used
to send message to VF.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v3:
* Fixed dereferencing type-punned pointer.
---
 drivers/net/sssnic/base/meson.build   |   1 +
 drivers/net/sssnic/base/sssnic_hw.c   |  10 +
 drivers/net/sssnic/base/sssnic_hw.h   |   4 +
 drivers/net/sssnic/base/sssnic_mbox.c | 615 ++
 drivers/net/sssnic/base/sssnic_mbox.h |  45 ++
 drivers/net/sssnic/base/sssnic_misc.h |  11 +
 drivers/net/sssnic/base/sssnic_reg.h  |  47 ++
 7 files changed, 733 insertions(+)
 create mode 100644 drivers/net/sssnic/base/sssnic_mbox.c
 create mode 100644 drivers/net/sssnic/base/sssnic_mbox.h
 create mode 100644 drivers/net/sssnic/base/sssnic_misc.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index dd4dd08fc1..4abd1a0daf 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -5,6 +5,7 @@ sources = [
 'sssnic_hw.c',
 'sssnic_eventq.c',
 'sssnic_msg.c',
+'sssnic_mbox.c',
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
index 387c823c7e..ff527b2c7f 100644
--- a/drivers/net/sssnic/base/sssnic_hw.c
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -11,6 +11,7 @@
 #include "sssnic_reg.h"
 #include "sssnic_eventq.h"
 #include "sssnic_msg.h"
+#include "sssnic_mbox.h"
 
 static int
 wait_for_sssnic_hw_ready(struct sssnic_hw *hw)
@@ -210,8 +211,16 @@ sssnic_hw_init(struct sssnic_hw *hw)
goto eventq_init_fail;
}
 
+   ret = sssnic_mbox_init(hw);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Failed to initialize mailbox");
+   goto mbox_init_fail;
+   }
+
return -EINVAL;
 
+mbox_init_fail:
+   sssnic_eventq_all_shutdown(hw);
 eventq_init_fail:
sssnic_msg_inbox_shutdown(hw);
return ret;
@@ -222,6 +231,7 @@ sssnic_hw_shutdown(struct sssnic_hw *hw)
 {
PMD_INIT_FUNC_TRACE();
 
+   sssnic_mbox_shutdown(hw);
sssnic_eventq_all_shutdown(hw);
sssnic_msg_inbox_shutdown(hw);
 }
diff --git a/drivers/net/sssnic/base/sssnic_hw.h 
b/drivers/net/sssnic/base/sssnic_hw.h
index 38fb9ac1ac..41e65f5880 100644
--- a/drivers/net/sssnic/base/sssnic_hw.h
+++ b/drivers/net/sssnic/base/sssnic_hw.h
@@ -53,12 +53,16 @@ struct sssnic_hw {
struct sssnic_hw_attr attr;
struct sssnic_eventq *eventqs;
struct sssnic_msg_inbox *msg_inbox;
+   struct sssnic_mbox *mbox;
uint8_t num_eventqs;
uint16_t eth_port_id;
 };
 
+#define SSSNIC_FUNC_IDX(hw) ((hw)->attr.func_idx)
 #define SSSNIC_ETH_PORT_ID(hw) ((hw)->eth_port_id)
 #define SSSNIC_MPU_FUNC_IDX 0x1fff
+#define SSSNIC_FUNC_TYPE(hw) ((hw)->attr.func_type)
+#define SSSNIC_AF_FUNC_IDX(hw) ((hw)->attr.af_idx)
 
 int sssnic_hw_init(struct sssnic_hw *hw);
 void sssnic_hw_shutdown(struct sssnic_hw *hw);
diff --git a/drivers/net/sssnic/base/sssnic_mbox.c 
b/drivers/net/sssnic/base/sssnic_mbox.c
new file mode 100644
index 00..02957137ea
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_mbox.c
@@ -0,0 +1,615 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_hw.h"
+#include "sssnic_reg.h"
+#include "sssnic_misc.h"
+#include "sssnic_eventq.h"
+#include "sssnic_mbox.h"
+
+#define SSSNIC_MBOX_SEND_RESULT_SIZE 16
+#define SSSNIC_MBOX_SEND_BUF_SIZE 2048UL
+#define SSSNIC_MBOX_RESP_MSG_EVENTQ 1
+#define SSSNIC_MBOX_SEND_DONE_TIMEOUT 50 /* uint is 10us */
+#define SSSNIC_MBOX_DEF_REQ_TIMEOUT 4000 /* millisecond */
+#define SSSNIC_MBOX_REQ_ID_MASK 0xf /* request id only 4 bits*/
+
+struct sssnic_sendbox {
+   struct sssnic_mbox *mbox;
+   /* Send data memory */
+   uint8_t *data;
+   /* Send result DMA memory zone */
+   const struct rte_memzone *result_mz;
+   /* Send result DMA virtual address */
+   volatile uint64_t *result_addr;
+   /* DMA buffer mz */
+   const struct rte_memzone *buf_mz;
+   /* DMA buffer virtual address */
+   uint8_t *buf_addr;
+   pthread_mutex_t lock;
+};
+
+struct sssnic_mbox_msg_dma_desc {
+   /* 32bit xor checksum for DMA data */
+   uint32_t checksum;
+   /* dword of high DMA address */
+   uint32_t dma_addr_hi;
+   /* dword of low DMA address */
+   uint32_t dma_addr_lo;
+   /* DMA data length */
+   uint32_t len;
+   uint32_t resvd[2];
+};
+#define SSSNIC_MBOX_MSG_

[PATCH v4 10/32] net/sssnic: add dev configure and infos get

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/sssnic_api.c | 33 
 drivers/net/sssnic/base/sssnic_api.h |  8 +++
 drivers/net/sssnic/base/sssnic_cmd.h | 14 +
 drivers/net/sssnic/base/sssnic_hw.c  | 33 +++-
 drivers/net/sssnic/base/sssnic_hw.h  |  6 +++
 drivers/net/sssnic/sssnic_ethdev.c   | 76 
 drivers/net/sssnic/sssnic_ethdev.h   | 53 +++
 7 files changed, 222 insertions(+), 1 deletion(-)

diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 51a59f0f25..bf0859cd63 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -100,3 +100,36 @@ sssnic_msix_attr_set(struct sssnic_hw *hw, uint16_t 
msix_idx,
 
return 0;
 }
+
+int
+sssnic_capability_get(struct sssnic_hw *hw, struct sssnic_capability *capa)
+{
+   struct sssnic_capability_get_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   int ret;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_GET_CAPABILITY_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_CFG_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_CAPABILITY_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   capa->phy_port = cmd.phy_port;
+   capa->max_num_rxq = cmd.rxq_max_id + 1;
+   capa->max_num_txq = cmd.txq_max_id + 1;
+
+   return 0;
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index 3d54eb826a..8011cc8b0f 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -15,9 +15,17 @@ struct sssnic_msix_attr {
uint8_t resend_timer;
 };
 
+struct sssnic_capability {
+   uint16_t max_num_txq;
+   uint16_t max_num_rxq;
+   uint8_t phy_port;
+   uint8_t cos;
+};
+
 int sssnic_msix_attr_get(struct sssnic_hw *hw, uint16_t msix_idx,
struct sssnic_msix_attr *attr);
 int sssnic_msix_attr_set(struct sssnic_hw *hw, uint16_t msix_idx,
struct sssnic_msix_attr *attr);
+int sssnic_capability_get(struct sssnic_hw *hw, struct sssnic_capability 
*capa);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index ee9f536ac2..79192affbc 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -18,6 +18,8 @@ enum sssnic_mgmt_cmd_id {
SSSNIC_GET_FW_VERSION_CMD = 60,
 };
 
+#define SSSNIC_GET_CAPABILITY_CMD 0
+
 struct sssnic_cmd_common {
uint8_t status;
uint8_t version;
@@ -111,4 +113,16 @@ struct sssnic_msix_ctrl_cmd {
uint8_t resvd1[5];
 };
 
+struct sssnic_capability_get_cmd {
+   struct sssnic_cmd_common common;
+   uint16_t function;
+   uint16_t resvd0;
+   uint8_t resvd1[3];
+   uint8_t phy_port;
+   uint32_t resvd2[16];
+   uint16_t txq_max_id;
+   uint16_t rxq_max_id;
+   uint32_t resvd3[63];
+};
+
 #endif /* _SSSNIC_CMD_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
index 4ca75208af..8f5f556bde 100644
--- a/drivers/net/sssnic/base/sssnic_hw.c
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -253,6 +253,29 @@ sssnic_msix_attr_init(struct sssnic_hw *hw)
return 0;
 }
 
+static int
+sssnic_capability_init(struct sssnic_hw *hw)
+{
+   struct sssnic_capability cap;
+   int ret;
+
+   ret = sssnic_capability_get(hw, &cap);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to get sssnic capability");
+   return ret;
+   }
+
+   PMD_DRV_LOG(INFO,
+   "Initialized capability, physic port:%u, max %u txqs, max %u 
rxqs",
+   cap.phy_port, cap.max_num_txq, cap.max_num_rxq);
+
+   hw->phy_port = cap.phy_port;
+   hw->max_num_rxq = cap.max_num_rxq;
+   hw->max_num_txq = cap.max_num_txq;
+
+   return 0;
+}
+
 static int
 sssnic_base_init(struct sssnic_hw *hw)
 {
@@ -360,10 +383,18 @@ sssnic_hw_init(struct sssnic_hw *hw)
goto ctrlq_init_fail;
}
 
+   ret = sssnic_capability_init(hw);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to initialize capability");
+   goto capbility_init_fail;
+   }
+
sssnic_pf_status_set(hw, SSSNIC_PF_STATUS_ACTIVE);
 
-   return -EINVAL;
+   return 0;
 
+capbility_init_fail:
+   sssnic_ctrlq_s

[PATCH v4 09/32] net/sssnic/base: add control queue

[wanry]

From: Renyong Wan 

Control queue is used for communication between driver and datapath code
of firmware.

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed variable 'cmd_len' is uninitialized when used.
---
 drivers/net/sssnic/base/meson.build|   2 +
 drivers/net/sssnic/base/sssnic_api.c   | 102 +
 drivers/net/sssnic/base/sssnic_api.h   |  23 ++
 drivers/net/sssnic/base/sssnic_cmd.h   | 114 ++
 drivers/net/sssnic/base/sssnic_ctrlq.c | 521 +
 drivers/net/sssnic/base/sssnic_ctrlq.h |  58 +++
 drivers/net/sssnic/base/sssnic_hw.c| 149 +++
 drivers/net/sssnic/base/sssnic_hw.h|   8 +
 8 files changed, 977 insertions(+)
 create mode 100644 drivers/net/sssnic/base/sssnic_api.c
 create mode 100644 drivers/net/sssnic/base/sssnic_api.h
 create mode 100644 drivers/net/sssnic/base/sssnic_cmd.h
 create mode 100644 drivers/net/sssnic/base/sssnic_ctrlq.c
 create mode 100644 drivers/net/sssnic/base/sssnic_ctrlq.h

diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index 7c23a82ff3..e93ca7b24b 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -7,6 +7,8 @@ sources = [
 'sssnic_msg.c',
 'sssnic_mbox.c',
 'sssnic_workq.c',
+'sssnic_ctrlq.c',
+'sssnic_api.c',
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
new file mode 100644
index 00..51a59f0f25
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -0,0 +1,102 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_hw.h"
+#include "sssnic_cmd.h"
+#include "sssnic_mbox.h"
+#include "sssnic_api.h"
+
+int
+sssnic_msix_attr_get(struct sssnic_hw *hw, uint16_t msix_idx,
+   struct sssnic_msix_attr *attr)
+{
+   int ret;
+   struct sssnic_msg msg;
+   struct sssnic_msix_ctrl_cmd cmd;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.func_id = SSSNIC_FUNC_IDX(hw);
+   cmd.opcode = SSSNIC_CMD_OPCODE_GET;
+   cmd.idx = msix_idx;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_MSIX_CTRL_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_COMM_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to MSIX_CTRL_CMD, len=%u, status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+   attr->lli_credit = cmd.lli_credit;
+   attr->lli_timer = cmd.lli_timer;
+   attr->pending_limit = cmd.pending_count;
+   attr->coalescing_timer = cmd.coalescing_timer;
+   attr->resend_timer = cmd.resend_timer;
+
+   return 0;
+}
+
+int
+sssnic_msix_attr_set(struct sssnic_hw *hw, uint16_t msix_idx,
+   struct sssnic_msix_attr *attr)
+{
+   int ret;
+   struct sssnic_msg msg;
+   struct sssnic_msix_ctrl_cmd cmd;
+   struct sssnic_msix_attr tmp;
+   uint32_t cmd_len;
+
+   ret = sssnic_msix_attr_get(hw, msix_idx, &tmp);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to get interrupt configuration");
+   return ret;
+   }
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.func_id = SSSNIC_FUNC_IDX(hw);
+   cmd.opcode = SSSNIC_CMD_OPCODE_SET;
+   cmd.idx = msix_idx;
+   cmd.lli_credit = tmp.lli_credit;
+   cmd.lli_timer = tmp.lli_timer;
+   cmd.pending_count = tmp.pending_limit;
+   cmd.coalescing_timer = tmp.coalescing_timer;
+   cmd.resend_timer = tmp.resend_timer;
+   if (attr->lli_set != 0) {
+   cmd.lli_credit = attr->lli_credit;
+   cmd.lli_timer = attr->lli_timer;
+   }
+   if (attr->coalescing_set != 0) {
+   cmd.pending_count = attr->pending_limit;
+   cmd.coalescing_timer = attr->coalescing_timer;
+   cmd.resend_timer = attr->resend_timer;
+   }
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_MSIX_CTRL_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_COMM_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to MSIX_CTRL_CMD, len=%u, status=%u",
+   cmd_len, cmd.common.statu

[PATCH v4 15/32] net/sssnic: support Tx queue setup and release

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.

v4:
* Fixed coding style issue of REPEATED_WORD.
---
 drivers/net/sssnic/meson.build|   1 +
 drivers/net/sssnic/sssnic_ethdev.c|   4 +
 drivers/net/sssnic/sssnic_ethdev.h|   1 +
 drivers/net/sssnic/sssnic_ethdev_tx.c | 354 ++
 drivers/net/sssnic/sssnic_ethdev_tx.h |  27 ++
 5 files changed, 387 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_tx.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_tx.h

diff --git a/drivers/net/sssnic/meson.build b/drivers/net/sssnic/meson.build
index 7c3516a279..0c6e21310d 100644
--- a/drivers/net/sssnic/meson.build
+++ b/drivers/net/sssnic/meson.build
@@ -20,4 +20,5 @@ sources = files(
 'sssnic_ethdev.c',
 'sssnic_ethdev_link.c',
 'sssnic_ethdev_rx.c',
+'sssnic_ethdev_tx.c',
 )
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index f98510a55d..732fddfcf7 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -11,6 +11,7 @@
 #include "sssnic_ethdev.h"
 #include "sssnic_ethdev_link.h"
 #include "sssnic_ethdev_rx.h"
+#include "sssnic_ethdev_tx.h"
 
 static int
 sssnic_ethdev_infos_get(struct rte_eth_dev *ethdev,
@@ -336,6 +337,7 @@ sssnic_ethdev_release(struct rte_eth_dev *ethdev)
struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
 
sssnic_ethdev_link_intr_disable(ethdev);
+   sssnic_ethdev_tx_queue_all_release(ethdev);
sssnic_ethdev_rx_queue_all_release(ethdev);
sssnic_ethdev_mac_addrs_clean(ethdev);
sssnic_hw_shutdown(hw);
@@ -354,6 +356,8 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.set_mc_addr_list = sssnic_ethdev_set_mc_addr_list,
.rx_queue_setup = sssnic_ethdev_rx_queue_setup,
.rx_queue_release = sssnic_ethdev_rx_queue_release,
+   .tx_queue_setup = sssnic_ethdev_tx_queue_setup,
+   .tx_queue_release = sssnic_ethdev_tx_queue_release,
 };
 
 static int
diff --git a/drivers/net/sssnic/sssnic_ethdev.h 
b/drivers/net/sssnic/sssnic_ethdev.h
index 51740413c6..ab832d179f 100644
--- a/drivers/net/sssnic/sssnic_ethdev.h
+++ b/drivers/net/sssnic/sssnic_ethdev.h
@@ -57,6 +57,7 @@
 #define SSSNIC_ETHDEV_MAX_NUM_MC_MAC 2048
 
 #define SSSNIC_ETHDEV_DEF_RX_FREE_THRESH 32
+#define SSSNIC_ETHDEV_DEF_TX_FREE_THRESH 32
 
 struct sssnic_netdev {
void *hw;
diff --git a/drivers/net/sssnic/sssnic_ethdev_tx.c 
b/drivers/net/sssnic/sssnic_ethdev_tx.c
new file mode 100644
index 00..e80bf8c396
--- /dev/null
+++ b/drivers/net/sssnic/sssnic_ethdev_tx.c
@@ -0,0 +1,354 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+
+#include "sssnic_log.h"
+#include "sssnic_ethdev.h"
+#include "sssnic_ethdev_tx.h"
+#include "base/sssnic_hw.h"
+#include "base/sssnic_workq.h"
+#include "base/sssnic_api.h"
+#include "base/sssnic_misc.h"
+
+/* Hardware format of tx desc */
+struct sssnic_ethdev_tx_desc {
+   union {
+   uint32_t dw0;
+   struct {
+   /* length of the first tx seg data */
+   uint32_t data_len : 18;
+   uint32_t dw0_resvd0 : 1;
+   /* number of tx segments in tx entry */
+   uint32_t num_segs : 8;
+   /* offload desc enable */
+   uint32_t offload_en : 1;
+   /* data format, use SGL if 0 else inline */
+   uint32_t data_fmt : 1;
+   /* DN, always set 0  */
+   uint32_t dw0_resvd1 : 1;
+   /* refer sssnic_ethdev_txq_entry_type */
+   uint32_t entry_type : 1;
+   uint32_t owner : 1;
+   };
+   };
+   union {
+   uint32_t dw1;
+   struct {
+   uint32_t pkt_type : 2;
+   uint32_t payload_off : 8;
+   /* UFO, not used, always set 0 */
+   uint32_t dw1_resvd0 : 1;
+   uint32_t tso_en : 1;
+   /* TCP/UDP checksum offload enable flag */
+   uint32_t csum_en : 1;
+   uint32_t mss : 14;
+   uint32_t sctp_crc_en : 1;
+   /* set 1 if entry type is not compact else set 0 */
+   uint32_t uc : 1;
+   /* PRI, not used, always set 0  */
+   uint32_t dw1_resvd1 : 3;
+   };
+   };
+   union {
+   uint32_t dw2;
+   /* high 32bit of  DMA address of the first tx seg data */
+   uint32_t data_addr_hi;
+   };
+   union {
+   uint32_t dw3;
+ 

[PATCH v4 12/32] net/sssnic: support dev link status

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 doc/guides/nics/features/sssnic.ini |   1 +
 drivers/net/sssnic/base/sssnic_api.c| 127 
 drivers/net/sssnic/base/sssnic_api.h|  14 +++
 drivers/net/sssnic/base/sssnic_cmd.h|  37 +++
 drivers/net/sssnic/meson.build  |   1 +
 drivers/net/sssnic/sssnic_ethdev.c  |   4 +
 drivers/net/sssnic/sssnic_ethdev_link.c | 111 +
 drivers/net/sssnic/sssnic_ethdev_link.h |  12 +++
 8 files changed, 307 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_link.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_link.h

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 602839d301..a0688e70ef 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -4,6 +4,7 @@
 ; Refer to default.ini for the full list of available PMD features.
 ;
 [Features]
+Link status  = Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
 Linux= Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 76266dfbae..3bb31009c5 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -307,3 +307,130 @@ sssnic_mac_addr_del(struct sssnic_hw *hw, uint8_t *addr)
 
return 0;
 }
+
+int
+sssnic_netif_link_status_get(struct sssnic_hw *hw, uint8_t *status)
+{
+   int ret;
+   struct sssnic_netif_link_status_get_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   uint16_t func;
+
+   if (hw == NULL || status == NULL)
+   return -EINVAL;
+
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   func = SSSNIC_PF_FUNC_IDX(hw);
+   else
+   func = SSSNIC_MPU_FUNC_IDX;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.port = SSSNIC_PHY_PORT(hw);
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_GET_NETIF_LINK_STATUS_CMD, func, SSSNIC_NETIF_MODULE,
+   SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_NETIF_LINK_STATUS_CMD, 
len=%u, status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   *status = cmd.status;
+
+   return 0;
+}
+
+int
+sssnic_netif_link_info_get(struct sssnic_hw *hw,
+   struct sssnic_netif_link_info *info)
+{
+   int ret;
+   struct sssnic_netif_link_info_get_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   if (hw == NULL || info == NULL)
+   return -EINVAL;
+
+   ret = sssnic_netif_link_status_get(hw, &info->status);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to get netif link state!");
+   return ret;
+   }
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.port = SSSNIC_PHY_PORT(hw);
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_GET_NETIF_LINK_INFO_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_NETIF_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_NETIF_LINK_INFO_CMD, 
len=%u, status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   info->speed = cmd.speed;
+   info->duplex = cmd.duplex;
+   info->fec = cmd.fec;
+   info->type = cmd.type;
+   info->autoneg_capa = cmd.autoneg_capa;
+   info->autoneg = cmd.autoneg;
+
+   return 0;
+}
+
+int
+sssnic_netif_enable_set(struct sssnic_hw *hw, uint8_t state)
+{
+   int ret;
+   struct sssnic_netif_enable_set_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   if (hw == NULL)
+   return -EINVAL;
+
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   return 0;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   if (state != 0)
+   cmd.flag = SSSNIC_SET_NETIF_ENABLE_CMD_FLAG_RX_EN |
+  SSSNIC_SET_NETIF_ENABLE_CMD_FLAG_TX_EN;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_SET_NETIF_ENABLE_CMD, SSSNIC_MPU_FUNC_ID

[PATCH v4 11/32] net/sssnic: add dev MAC ops

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini  |   2 +
 drivers/net/sssnic/base/sssnic_api.c | 174 +
 drivers/net/sssnic/base/sssnic_api.h |   4 +
 drivers/net/sssnic/base/sssnic_cmd.h |  25 +++
 drivers/net/sssnic/base/sssnic_hw.h  |   1 +
 drivers/net/sssnic/sssnic_ethdev.c   | 270 ++-
 drivers/net/sssnic/sssnic_ethdev.h   |   2 +
 7 files changed, 477 insertions(+), 1 deletion(-)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 6d9786db7e..602839d301 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -4,6 +4,8 @@
 ; Refer to default.ini for the full list of available PMD features.
 ;
 [Features]
+Unicast MAC filter   = Y
+Multicast MAC filter = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index bf0859cd63..76266dfbae 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -133,3 +133,177 @@ sssnic_capability_get(struct sssnic_hw *hw, struct 
sssnic_capability *capa)
 
return 0;
 }
+
+int
+sssnic_mac_addr_get(struct sssnic_hw *hw, uint8_t *addr)
+{
+   int ret;
+   struct sssnic_mac_addr_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   uint16_t func;
+
+   if (hw == NULL || addr == NULL)
+   return -EINVAL;
+
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   func = SSSNIC_PF_FUNC_IDX(hw);
+   else
+   func = SSSNIC_MPU_FUNC_IDX;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_GET_MAC_ADDR_CMD,
+   func, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_DEF_MAC_ADDR_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   rte_memcpy(addr, cmd.addr, 6);
+
+   return 0;
+}
+
+int
+sssnic_mac_addr_update(struct sssnic_hw *hw, uint8_t *new, uint8_t *old)
+{
+   int ret;
+   struct sssnic_mac_addr_update_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   uint16_t func;
+
+   if (hw == NULL || new == NULL || old == NULL)
+   return -EINVAL;
+
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   func = SSSNIC_PF_FUNC_IDX(hw);
+   else
+   func = SSSNIC_MPU_FUNC_IDX;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   rte_memcpy(cmd.new_addr, new, 6);
+   rte_memcpy(cmd.old_addr, old, 6);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_UPDATE_MAC_ADDR_CMD, func, SSSNIC_LAN_MODULE,
+   SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   if (cmd.common.status == SSSNIC_MAC_ADDR_CMD_STATUS_IGNORED) {
+   PMD_DRV_LOG(WARNING,
+   "MAC address operation is ignored");
+   return 0;
+   }
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_UPDATE_MAC_ADDR_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_mac_addr_add(struct sssnic_hw *hw, uint8_t *addr)
+{
+   int ret;
+   struct sssnic_mac_addr_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   uint16_t func;
+
+   if (hw == NULL || addr == NULL)
+   return -EINVAL;
+
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   func = SSSNIC_PF_FUNC_IDX(hw);
+   else
+   func = SSSNIC_MPU_FUNC_IDX;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   rte_memcpy(cmd.addr, addr, 6);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_ADD_MAC_ADDR_CMD,
+   func, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbo

[PATCH v4 16/32] net/sssnic: support Rx queue start and stop

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/sssnic_api.c  |  63 +
 drivers/net/sssnic/base/sssnic_api.h  |   2 +
 drivers/net/sssnic/base/sssnic_cmd.h  |  49 
 drivers/net/sssnic/sssnic_ethdev.c|   2 +
 drivers/net/sssnic/sssnic_ethdev.h|   2 +
 drivers/net/sssnic/sssnic_ethdev_rx.c | 332 ++
 drivers/net/sssnic/sssnic_ethdev_rx.h |   4 +
 7 files changed, 454 insertions(+)

diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 3bb31009c5..3050d573bf 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -11,6 +11,7 @@
 #include "sssnic_hw.h"
 #include "sssnic_cmd.h"
 #include "sssnic_mbox.h"
+#include "sssnic_ctrlq.h"
 #include "sssnic_api.h"
 
 int
@@ -434,3 +435,65 @@ sssnic_netif_enable_set(struct sssnic_hw *hw, uint8_t 
state)
 
return 0;
 }
+
+int
+sssnic_port_enable_set(struct sssnic_hw *hw, bool state)
+{
+   int ret;
+   struct sssnic_port_enable_set_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   if (hw == NULL)
+   return -EINVAL;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.state = state ? 1 : 0;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_SET_PORT_ENABLE_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_SET_PORT_ENABLE_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_rxq_flush(struct sssnic_hw *hw, uint16_t qid)
+{
+   struct sssnic_ctrlq_cmd cmd;
+   struct sssnic_rxq_flush_cmd data;
+   int ret;
+
+   data.u32 = 0;
+   data.qid = qid;
+   data.u32 = rte_cpu_to_be_32(data.u32);
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.data = &data;
+   cmd.module = SSSNIC_LAN_MODULE;
+   cmd.data_len = sizeof(data);
+   cmd.cmd = SSSNIC_FLUSH_RXQ_CMD;
+
+   ret = sssnic_ctrlq_cmd_exec(hw, &cmd, 0);
+   if (ret != 0 || cmd.result != 0) {
+   PMD_DRV_LOG(ERR,
+   "Failed to execulte ctrlq command %s, ret=%d, result=%" 
PRIu64,
+   "SSSNIC_FLUSH_RXQ_CMD", ret, cmd.result);
+   return -EIO;
+   }
+
+   return 0;
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index 168aa152b9..29962aabf8 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -45,5 +45,7 @@ int sssnic_netif_link_status_get(struct sssnic_hw *hw, 
uint8_t *status);
 int sssnic_netif_link_info_get(struct sssnic_hw *hw,
struct sssnic_netif_link_info *info);
 int sssnic_netif_enable_set(struct sssnic_hw *hw, uint8_t state);
+int sssnic_port_enable_set(struct sssnic_hw *hw, bool state);
+int sssnic_rxq_flush(struct sssnic_hw *hw, uint16_t qid);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index 6957b742fc..6364058d36 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -35,6 +35,37 @@ enum sssnic_netif_cmd_id {
SSSNIC_GET_NETIF_LINK_INFO_CMD = 153,
 };
 
+enum sssnic_port_cmd_id {
+   SSSNIC_REGISTER_VF_PORT_CMD = 0,
+   SSSNIC_SET_PORT_RXTX_SIZE_CMD = 5,
+   SSSNIC_SET_PORT_ENABLE_CMD = 6,
+   SSSNIC_SET_PORT_RX_MODE_CMD = 7,
+   SSSNIC_SET_PORT_TX_CI_ATTR_CMD = 8,
+   SSSNIC_GET_PORT_STATS_CMD = 9,
+   SSSNIC_CLEAR_PORT_STATS_CMD = 10,
+
+   SSSNIC_CLEAN_PORT_RES_CMD = 11,
+
+   SSSNIC_PORT_LRO_CFG_CMD = 13,
+   SSSNIC_PORT_LRO_TIMER_CMD = 14,
+   SSSNIC_PORT_FEATURE_CMD = 15,
+
+   SSSNIC_SET_PORT_VLAN_FILTER_CMD = 25,
+   SSSNIC_ENABLE_PORT_VLAN_FILTER_CMD = 26,
+   SSSNIC_ENABLE_PORT_VLAN_STRIP_CMD = 27,
+
+   SSSNIC_PORT_FLOW_CTRL_CMD = 101,
+};
+
+enum sssnic_ctrlq_cmd_id {
+   SSSNIC_SET_RXTXQ_CTX_CMD = 0,
+   SSSNIC_RESET_OFFLOAD_CTX_CMD = 1,
+   SSSNIC_SET_RSS_INDIR_TABLE_CMD = 4,
+   SSSNIC_SET_RSS_KEY_CTRLQ_CMD = 5,
+   SSSNIC_GET_RSS_INDIR_TABLE_CMD = 6,
+   SSSNIC_FLUSH_RXQ_CMD = 10,
+};
+
 struct sssnic_cmd_common {
uint8_t status;
uint8_t version;
@@ -187,4 +218,22 @@ struct sssnic_netif_enable_set_cmd {
uint8_t resvd1[3];
 };
 
+struct sssnic_port_enable_set_cmd {
+   struct sssnic_cmd_common common;
+   uint16_t function;
+   uint16_t resvd0;
+

[PATCH v4 13/32] net/sssnic: support link status event

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed 'EINVAL' undeclared.
---
 doc/guides/nics/features/sssnic.ini|   1 +
 drivers/net/sssnic/base/meson.build|   1 +
 drivers/net/sssnic/base/sssnic_exception.c | 116 +
 drivers/net/sssnic/base/sssnic_exception.h |  10 ++
 drivers/net/sssnic/base/sssnic_hw.c|  46 
 drivers/net/sssnic/base/sssnic_hw.h|  22 
 drivers/net/sssnic/sssnic_ethdev.c |   6 +-
 drivers/net/sssnic/sssnic_ethdev_link.c|  98 +
 drivers/net/sssnic/sssnic_ethdev_link.h|   2 +
 9 files changed, 301 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/sssnic/base/sssnic_exception.c
 create mode 100644 drivers/net/sssnic/base/sssnic_exception.h

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index a0688e70ef..82b527ba26 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -5,6 +5,7 @@
 ;
 [Features]
 Link status  = Y
+Link status event= Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
 Linux= Y
diff --git a/drivers/net/sssnic/base/meson.build 
b/drivers/net/sssnic/base/meson.build
index e93ca7b24b..cf0c177cab 100644
--- a/drivers/net/sssnic/base/meson.build
+++ b/drivers/net/sssnic/base/meson.build
@@ -9,6 +9,7 @@ sources = [
 'sssnic_workq.c',
 'sssnic_ctrlq.c',
 'sssnic_api.c',
+   'sssnic_exception.c',
 ]
 
 c_args = cflags
diff --git a/drivers/net/sssnic/base/sssnic_exception.c 
b/drivers/net/sssnic/base/sssnic_exception.c
new file mode 100644
index 00..738ca9c6cd
--- /dev/null
+++ b/drivers/net/sssnic/base/sssnic_exception.c
@@ -0,0 +1,116 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+
+#include "../sssnic_log.h"
+#include "sssnic_hw.h"
+#include "sssnic_reg.h"
+#include "sssnic_cmd.h"
+#include "sssnic_msg.h"
+#include "sssnic_eventq.h"
+#include "sssnic_mbox.h"
+#include "sssnic_exception.h"
+
+static void
+sssnic_link_event_msg_handle(struct sssnic_hw *hw, struct sssnic_msg *msg,
+   __rte_unused enum sssnic_msg_chann_id chan_id)
+{
+   struct sssnic_netif_link_status_get_cmd *cmd;
+   sssnic_link_event_cb_t *cb;
+   enum sssnic_link_status status;
+   void *priv;
+
+   cb = hw->link_event_handler.cb;
+   priv = hw->link_event_handler.priv;
+
+   cmd = (struct sssnic_netif_link_status_get_cmd *)msg->data_buf;
+   if (cb != NULL) {
+   if (cmd->status)
+   status = SSSNIC_LINK_STATUS_UP;
+   else
+   status = SSSNIC_LINK_STATUS_DOWN;
+
+   PMD_DRV_LOG(DEBUG, "Received sssnic%u link %s event",
+   SSSNIC_ETH_PORT_ID(hw), status ? "up" : "down");
+
+   return cb(cmd->port, status, priv);
+   }
+
+   PMD_DRV_LOG(WARNING, "Link event was not processed, port=%u, status=%u",
+   cmd->port, cmd->status);
+}
+
+static void
+sssnic_netif_vf_link_status_msg_handle(struct sssnic_hw *hw,
+   struct sssnic_msg *msg)
+{
+   int ret;
+
+   if (msg->ack != 0) {
+   msg->ack = 0;
+   msg->type = SSSNIC_MSG_TYPE_RESP;
+   msg->data_len = 1; /* indicate no data */
+   ret = sssnic_mbox_send(hw, msg, NULL, 0, 0);
+   if (ret != 0)
+   PMD_DRV_LOG(ERR,
+   "Failed to send VF link status response, 
ret=%d",
+   ret);
+   }
+}
+
+static void
+sssnic_netif_exception_msg_handle(struct sssnic_hw *hw, struct sssnic_msg *msg,
+   enum sssnic_msg_chann_id chan_id)
+{
+   if (msg->command == SSSNIC_GET_NETIF_LINK_STATUS_CMD) {
+   if (SSSNIC_FUNC_TYPE(hw) == SSSNIC_FUNC_TYPE_VF)
+   sssnic_netif_vf_link_status_msg_handle(hw, msg);
+
+   sssnic_link_event_msg_handle(hw, msg, chan_id);
+   return;
+   }
+
+   PMD_DRV_LOG(WARNING,
+   "Netif exception message was not processed, cmd=%u",
+   msg->command);
+}
+
+static int
+sssnic_exception_msg_process(struct sssnic_msg *msg,
+   enum sssnic_msg_chann_id chan_id, void *priv)
+{
+   struct sssnic_hw *hw = (struct sssnic_hw *)priv;
+
+   SSSNIC_DEBUG("command=%u, func=%u module=%u, type=%u, ack=%u, seq=%u, "
+"status=%u, id=%u data_buf=%p, data_len=%u",
+   msg->command, msg->func, msg->module, msg->type, msg->ack,
+   msg->seg, msg->status, msg->id, msg->data_buf, msg->data_len);
+
+   if (msg->module == SSSNIC_NETIF_MODULE) {
+   sssnic_netif_exception_msg_handle(hw, msg, chan_id);
+   return SSSNIC_MSG_DONE;
+   }
+
+   PMD_DRV_LOG(WARNING, "Exception message was not proce

[PATCH v4 17/32] net/sssnic: support Tx queue start and stop

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini   |   1 +
 drivers/net/sssnic/sssnic_ethdev.c|   2 +
 drivers/net/sssnic/sssnic_ethdev_tx.c | 155 ++
 drivers/net/sssnic/sssnic_ethdev_tx.h |   4 +
 4 files changed, 162 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 82b527ba26..b75c68cb33 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -6,6 +6,7 @@
 [Features]
 Link status  = Y
 Link status event= Y
+Queue start/stop = Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
 Linux= Y
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 208f0db402..8a18f25889 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -360,6 +360,8 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.tx_queue_release = sssnic_ethdev_tx_queue_release,
.rx_queue_start = sssnic_ethdev_rx_queue_start,
.rx_queue_stop = sssnic_ethdev_rx_queue_stop,
+   .tx_queue_start = sssnic_ethdev_tx_queue_start,
+   .tx_queue_stop = sssnic_ethdev_tx_queue_stop,
 };
 
 static int
diff --git a/drivers/net/sssnic/sssnic_ethdev_tx.c 
b/drivers/net/sssnic/sssnic_ethdev_tx.c
index e80bf8c396..47d7e3f343 100644
--- a/drivers/net/sssnic/sssnic_ethdev_tx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_tx.c
@@ -191,6 +191,18 @@ sssnic_ethdev_txq_ci_get(struct sssnic_ethdev_txq *txq)
return sssnic_workq_ci_get(txq->workq);
 }
 
+static inline int
+sssnic_ethdev_txq_pi_get(struct sssnic_ethdev_txq *txq)
+{
+   return sssnic_workq_pi_get(txq->workq);
+}
+
+static inline uint16_t
+sssnic_ethdev_txq_hw_ci_get(struct sssnic_ethdev_txq *txq)
+{
+   return *txq->hw_ci_addr & txq->idx_mask;
+}
+
 static inline void
 sssnic_ethdev_txq_consume(struct sssnic_ethdev_txq *txq, uint16_t num_entries)
 {
@@ -352,3 +364,146 @@ sssnic_ethdev_tx_queue_all_release(struct rte_eth_dev 
*ethdev)
for (qid = 0; qid < ethdev->data->nb_tx_queues; qid++)
sssnic_ethdev_tx_queue_release(ethdev, qid);
 }
+
+#define SSSNIC_ETHDEV_TX_FREE_BULK 64
+static inline int
+sssnic_ethdev_txq_pktmbufs_cleanup(struct sssnic_ethdev_txq *txq)
+{
+   struct sssnic_ethdev_tx_entry *txe;
+   struct rte_mbuf *free_pkts[SSSNIC_ETHDEV_TX_FREE_BULK];
+   uint16_t num_free_pkts = 0;
+   uint16_t hw_ci, ci, id_mask;
+   uint16_t count = 0;
+   int num_entries;
+
+   ci = sssnic_ethdev_txq_ci_get(txq);
+   hw_ci = sssnic_ethdev_txq_hw_ci_get(txq);
+   id_mask = txq->idx_mask;
+   num_entries = sssnic_ethdev_txq_num_used_entries(txq);
+
+   while (num_entries > 0) {
+   txe = &txq->txe[ci];
+
+   /* HW has not consumed enough entries of current packet */
+   if (((hw_ci - ci) & id_mask) < txe->num_workq_entries)
+   break;
+
+   num_entries -= txe->num_workq_entries;
+   count += txe->num_workq_entries;
+   ci = (ci + txe->num_workq_entries) & id_mask;
+
+   if (likely(txe->pktmbuf->nb_segs == 1)) {
+   struct rte_mbuf *pkt =
+   rte_pktmbuf_prefree_seg(txe->pktmbuf);
+   txe->pktmbuf = NULL;
+
+   if (unlikely(pkt == NULL))
+   continue;
+
+   free_pkts[num_free_pkts++] = pkt;
+   if (unlikely(pkt->pool != free_pkts[0]->pool ||
+num_free_pkts >=
+SSSNIC_ETHDEV_TX_FREE_BULK)) {
+   rte_mempool_put_bulk(free_pkts[0]->pool,
+   (void **)free_pkts, num_free_pkts - 1);
+   num_free_pkts = 0;
+   free_pkts[num_free_pkts++] = pkt;
+   }
+   } else {
+   rte_pktmbuf_free(txe->pktmbuf);
+   txe->pktmbuf = NULL;
+   }
+   }
+
+   if (num_free_pkts > 0)
+   rte_mempool_put_bulk(free_pkts[0]->pool, (void **)free_pkts,
+   num_free_pkts);
+
+   sssnic_ethdev_txq_consume(txq, count);
+
+   return count;
+}
+
+#define SSSNIC_ETHDEV_TXQ_FUSH_TIMEOUT 3000 /* 3 seconds */
+static int
+sssnic_ethdev_txq_flush(struct sssnic_ethdev_txq *txq)
+{
+   uint64_t timeout;
+   uint16_t used_entries;
+
+   timeout = rte_get_timer_cycles() +
+ rte_get_timer_hz() * SSSNIC_ETHDEV_TXQ_FUSH_TIMEOUT / 1000;
+
+   do {
+   sssnic_ethdev_txq_pktmbufs_cleanup(txq);
+   used_entries = sssnic_ethdev_txq_num_used_entries(txq);
+   if (used_entries == 0)
+   return 0;
+
+

[PATCH v4 14/32] net/sssnic: support Rx queue setup and release

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 drivers/net/sssnic/meson.build|   1 +
 drivers/net/sssnic/sssnic_ethdev.c|   4 +
 drivers/net/sssnic/sssnic_ethdev.h|   2 +
 drivers/net/sssnic/sssnic_ethdev_rx.c | 420 ++
 drivers/net/sssnic/sssnic_ethdev_rx.h |  24 ++
 5 files changed, 451 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_rx.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_rx.h

diff --git a/drivers/net/sssnic/meson.build b/drivers/net/sssnic/meson.build
index 12e967722a..7c3516a279 100644
--- a/drivers/net/sssnic/meson.build
+++ b/drivers/net/sssnic/meson.build
@@ -19,4 +19,5 @@ objs = [base_objs]
 sources = files(
 'sssnic_ethdev.c',
 'sssnic_ethdev_link.c',
+'sssnic_ethdev_rx.c',
 )
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 1e2b5c0450..f98510a55d 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -10,6 +10,7 @@
 #include "base/sssnic_api.h"
 #include "sssnic_ethdev.h"
 #include "sssnic_ethdev_link.h"
+#include "sssnic_ethdev_rx.h"
 
 static int
 sssnic_ethdev_infos_get(struct rte_eth_dev *ethdev,
@@ -335,6 +336,7 @@ sssnic_ethdev_release(struct rte_eth_dev *ethdev)
struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
 
sssnic_ethdev_link_intr_disable(ethdev);
+   sssnic_ethdev_rx_queue_all_release(ethdev);
sssnic_ethdev_mac_addrs_clean(ethdev);
sssnic_hw_shutdown(hw);
rte_free(hw);
@@ -350,6 +352,8 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.mac_addr_remove = sssnic_ethdev_mac_addr_remove,
.mac_addr_add = sssnic_ethdev_mac_addr_add,
.set_mc_addr_list = sssnic_ethdev_set_mc_addr_list,
+   .rx_queue_setup = sssnic_ethdev_rx_queue_setup,
+   .rx_queue_release = sssnic_ethdev_rx_queue_release,
 };
 
 static int
diff --git a/drivers/net/sssnic/sssnic_ethdev.h 
b/drivers/net/sssnic/sssnic_ethdev.h
index 636b3fd04c..51740413c6 100644
--- a/drivers/net/sssnic/sssnic_ethdev.h
+++ b/drivers/net/sssnic/sssnic_ethdev.h
@@ -56,6 +56,8 @@
 #define SSSNIC_ETHDEV_MAX_NUM_UC_MAC 128
 #define SSSNIC_ETHDEV_MAX_NUM_MC_MAC 2048
 
+#define SSSNIC_ETHDEV_DEF_RX_FREE_THRESH 32
+
 struct sssnic_netdev {
void *hw;
struct rte_ether_addr *mcast_addrs;
diff --git a/drivers/net/sssnic/sssnic_ethdev_rx.c 
b/drivers/net/sssnic/sssnic_ethdev_rx.c
new file mode 100644
index 00..0f489cf82f
--- /dev/null
+++ b/drivers/net/sssnic/sssnic_ethdev_rx.c
@@ -0,0 +1,420 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2018-2022 Shenzhen 3SNIC Information Technology Co., Ltd.
+ */
+
+#include 
+#include 
+#include 
+#include 
+
+#include "sssnic_log.h"
+#include "sssnic_ethdev.h"
+#include "sssnic_ethdev_rx.h"
+#include "base/sssnic_hw.h"
+#include "base/sssnic_workq.h"
+#include "base/sssnic_api.h"
+#include "base/sssnic_misc.h"
+
+/* hardware format of rx descriptor*/
+struct sssnic_ethdev_rx_desc {
+   /* status field */
+   union {
+   uint32_t dword0;
+   struct {
+   uint32_t ip_csum_err : 1;
+   uint32_t tcp_csum_err : 1;
+   uint32_t udp_csum_err : 1;
+   uint32_t igmp_csum_err : 1;
+   uint32_t icmpv4_csum_err : 1;
+   uint32_t icmpv6_csum_err : 1;
+   uint32_t sctp_crc_err : 1;
+   uint32_t hw_crc_err : 1;
+   uint32_t other_err : 1;
+   uint32_t err_resvd0 : 7;
+   uint32_t lro_num : 8;
+   uint32_t resvd1 : 1;
+   uint32_t lro_push : 1;
+   uint32_t lro_enter : 1;
+   uint32_t lro_intr : 1;
+   uint32_t flush : 1;
+   uint32_t decry : 1;
+   uint32_t bp_en : 1;
+   uint32_t done : 1;
+   };
+   struct {
+   uint32_t status_err : 16;
+   uint32_t status_rest : 16;
+   };
+   };
+
+   /* VLAN and length field */
+   union {
+   uint32_t dword1;
+   struct {
+   uint32_t vlan : 16;
+   uint32_t len : 16;
+   };
+   };
+
+   /* offload field */
+   union {
+   uint32_t dword2;
+   struct {
+   uint32_t pkt_type : 12;
+   uint32_t dword2_resvd0 : 9;
+   uint32_t vlan_en : 1;
+   uint32_t dword2_resvd1 : 2;
+   uint32_t rss_type : 8;
+   };
+   };
+
+   /* rss hash field */
+   union {
+   uint32_t dword3;
+  

[PATCH v4 18/32] net/sssnic: add Rx interrupt support

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini   |   1 +
 drivers/net/sssnic/base/sssnic_hw.c   |  14 +++
 drivers/net/sssnic/base/sssnic_hw.h   |   2 +
 drivers/net/sssnic/sssnic_ethdev.c|   2 +
 drivers/net/sssnic/sssnic_ethdev_rx.c | 135 ++
 drivers/net/sssnic/sssnic_ethdev_rx.h |   6 ++
 6 files changed, 160 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index b75c68cb33..e3b8166629 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -7,6 +7,7 @@
 Link status  = Y
 Link status event= Y
 Queue start/stop = Y
+Rx interrupt = Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
 Linux= Y
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
index 82eb4ea295..651a0aa7ef 100644
--- a/drivers/net/sssnic/base/sssnic_hw.c
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -145,6 +145,20 @@ sssnic_msix_resend_disable(struct sssnic_hw *hw, uint16_t 
msix_id)
sssnic_cfg_reg_write(hw, SSSNIC_MSIX_CTRL_REG, reg.u32);
 }
 
+void
+sssnic_msix_auto_mask_set(struct sssnic_hw *hw, uint16_t msix_id, int state)
+{
+   struct sssnic_msix_ctrl_reg reg;
+
+   reg.u32 = 0;
+   if (state == SSSNIC_MSIX_ENABLE)
+   reg.auto_msk_set = 1;
+   else
+   reg.auto_msk_clr = 1;
+   reg.msxi_idx = msix_id;
+   sssnic_cfg_reg_write(hw, SSSNIC_MSIX_CTRL_REG, reg.u32);
+}
+
 static void
 sssnic_pf_status_set(struct sssnic_hw *hw, enum sssnic_pf_status status)
 {
diff --git a/drivers/net/sssnic/base/sssnic_hw.h 
b/drivers/net/sssnic/base/sssnic_hw.h
index e25f5595e6..4820212543 100644
--- a/drivers/net/sssnic/base/sssnic_hw.h
+++ b/drivers/net/sssnic/base/sssnic_hw.h
@@ -100,6 +100,8 @@ int sssnic_hw_init(struct sssnic_hw *hw);
 void sssnic_hw_shutdown(struct sssnic_hw *hw);
 void sssnic_msix_state_set(struct sssnic_hw *hw, uint16_t msix_id, int state);
 void sssnic_msix_resend_disable(struct sssnic_hw *hw, uint16_t msix_id);
+void sssnic_msix_auto_mask_set(struct sssnic_hw *hw, uint16_t msix_id,
+   int state);
 int sssnic_link_event_callback_register(struct sssnic_hw *hw,
sssnic_link_event_cb_t *cb, void *priv);
 void sssnic_link_event_callback_unregister(struct sssnic_hw *hw);
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 8a18f25889..35bb26a0b1 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -362,6 +362,8 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.rx_queue_stop = sssnic_ethdev_rx_queue_stop,
.tx_queue_start = sssnic_ethdev_tx_queue_start,
.tx_queue_stop = sssnic_ethdev_tx_queue_stop,
+   .rx_queue_intr_enable = sssnic_ethdev_rx_queue_intr_enable,
+   .rx_queue_intr_disable = sssnic_ethdev_rx_queue_intr_disable,
 };
 
 static int
diff --git a/drivers/net/sssnic/sssnic_ethdev_rx.c 
b/drivers/net/sssnic/sssnic_ethdev_rx.c
index d8429e734d..9c1b2f20d1 100644
--- a/drivers/net/sssnic/sssnic_ethdev_rx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_rx.c
@@ -136,6 +136,12 @@ static const uint16_t sssnic_ethdev_rx_buf_size_tbl[] = { 
32, 64, 96, 128, 192,
 /* Doorbell offset 8192 */
 #define SSSNIC_ETHDEV_RXQ_DB_OFFSET 0x2000
 
+#define SSSNIC_ETHDEV_RX_MSIX_ID_START 1
+#define SSSNIC_ETHDEV_RX_MSIX_ID_INVAL 0
+#define SSSNIC_ETHDEV_RX_MSIX_PENDING_LIMIT 2
+#define SSSNIC_ETHDEV_RX_MSIX_COALESCING_TIMER 2
+#define SSSNIC_ETHDEV_RX_MSIX_RESNEDING_TIMER 7
+
 struct sssnic_ethdev_rxq_doorbell {
union {
uint64_t u64;
@@ -750,3 +756,132 @@ sssnic_ethdev_rx_queue_all_stop(struct rte_eth_dev 
*ethdev)
 
return 0;
 }
+
+static int
+sssinc_ethdev_rxq_intr_attr_init(struct sssnic_ethdev_rxq *rxq)
+{
+   int ret;
+   struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(rxq->ethdev);
+   struct sssnic_msix_attr attr;
+
+   attr.lli_set = 0;
+   attr.coalescing_set = 1;
+   attr.pending_limit = SSSNIC_ETHDEV_RX_MSIX_PENDING_LIMIT;
+   attr.coalescing_timer = SSSNIC_ETHDEV_RX_MSIX_COALESCING_TIMER;
+   attr.resend_timer = SSSNIC_ETHDEV_RX_MSIX_RESNEDING_TIMER;
+
+   ret = sssnic_msix_attr_set(hw, rxq->intr.msix_id, &attr);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to set msxi attributes");
+   return ret;
+   }
+
+   return 0;
+}
+
+int
+sssnic_ethdev_rx_queue_intr_enable(struct rte_eth_dev *ethdev, uint16_t qid)
+{
+   struct sssnic_ethdev_rxq *rxq = ethdev->data->rx_queues[qid];
+   struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
+
+   if (rxq->intr.enable)
+   return 0;
+
+   sssnic_msix_auto_mask_set(hw, rxq->intr.msix_id, SSSNIC_MSIX_ENABLE);
+   sssnic_msix_state_set(hw, rxq->intr.msix_id, SSSNIC_MSIX_ENABLE);
+   rxq->intr.enable = 1;
+
+   return 0;
+}
+

[PATCH v4 20/32] net/sssnic: support dev close and reset

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/sssnic_ethdev.c | 32 ++
 1 file changed, 32 insertions(+)

diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 8201a1e3c4..b59c4fd3ad 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -13,6 +13,8 @@
 #include "sssnic_ethdev_rx.h"
 #include "sssnic_ethdev_tx.h"
 
+static int sssnic_ethdev_init(struct rte_eth_dev *ethdev);
+
 static int
 sssnic_ethdev_infos_get(struct rte_eth_dev *ethdev,
struct rte_eth_dev_info *devinfo)
@@ -622,9 +624,39 @@ sssnic_ethdev_stop(struct rte_eth_dev *ethdev)
return 0;
 }
 
+static int
+sssnic_ethdev_close(struct rte_eth_dev *ethdev)
+{
+   sssnic_ethdev_release(ethdev);
+
+   PMD_DRV_LOG(INFO, "Port %u is closed", ethdev->data->port_id);
+
+   return 0;
+}
+
+static int
+sssnic_ethdev_reset(struct rte_eth_dev *ethdev)
+{
+   int ret;
+
+   sssnic_ethdev_release(ethdev);
+
+   ret = sssnic_ethdev_init(ethdev);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to initialize sssnic ethdev");
+   return ret;
+   }
+
+   PMD_DRV_LOG(INFO, "Port %u is reset", ethdev->data->port_id);
+
+   return 0;
+}
+
 static const struct eth_dev_ops sssnic_ethdev_ops = {
.dev_start = sssnic_ethdev_start,
.dev_stop = sssnic_ethdev_stop,
+   .dev_close = sssnic_ethdev_close,
+   .dev_reset = sssnic_ethdev_reset,
.dev_set_link_up = sssnic_ethdev_set_link_up,
.dev_set_link_down = sssnic_ethdev_set_link_down,
.link_update = sssnic_ethdev_link_update,
-- 
2.27.0

[PATCH v4 19/32] net/sssnic: support dev start and stop

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/sssnic_api.c  | 508 ++
 drivers/net/sssnic/base/sssnic_api.h  | 257 +
 drivers/net/sssnic/base/sssnic_cmd.h  | 100 +
 drivers/net/sssnic/base/sssnic_misc.h |  34 ++
 drivers/net/sssnic/sssnic_ethdev.c| 284 ++
 drivers/net/sssnic/sssnic_ethdev.h|  21 ++
 drivers/net/sssnic/sssnic_ethdev_rx.c | 270 +-
 drivers/net/sssnic/sssnic_ethdev_rx.h |   8 +
 drivers/net/sssnic/sssnic_ethdev_tx.c | 163 +
 drivers/net/sssnic/sssnic_ethdev_tx.h |   6 +
 10 files changed, 1650 insertions(+), 1 deletion(-)

diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 3050d573bf..81020387bd 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -13,6 +13,7 @@
 #include "sssnic_mbox.h"
 #include "sssnic_ctrlq.h"
 #include "sssnic_api.h"
+#include "sssnic_misc.h"
 
 int
 sssnic_msix_attr_get(struct sssnic_hw *hw, uint16_t msix_idx,
@@ -497,3 +498,510 @@ sssnic_rxq_flush(struct sssnic_hw *hw, uint16_t qid)
 
return 0;
 }
+
+static int
+sssnic_rxtx_size_set(struct sssnic_hw *hw, uint16_t rx_size, uint16_t tx_size,
+   uint32_t flags)
+{
+   int ret;
+   struct sssnic_rxtx_size_set_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   if (hw == NULL)
+   return -EINVAL;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.rx_size = rx_size;
+   cmd.tx_size = tx_size;
+   cmd.flags = flags;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_SET_PORT_RXTX_SIZE_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_SET_PORT_RXTX_SIZE_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_rxtx_max_size_init(struct sssnic_hw *hw, uint16_t rx_size,
+   uint16_t tx_size)
+{
+   return sssnic_rxtx_size_set(hw, rx_size, tx_size,
+   SSSNIC_CMD_INIT_RXTX_SIZE_FLAG | SSSNIC_CMD_SET_RX_SIZE_FLAG |
+   SSSNIC_CMD_SET_TX_SIZE_FLAG);
+}
+
+int
+sssnic_tx_max_size_set(struct sssnic_hw *hw, uint16_t tx_size)
+{
+   return sssnic_rxtx_size_set(hw, 0, tx_size,
+   SSSNIC_CMD_SET_TX_SIZE_FLAG);
+}
+
+int
+sssnic_port_features_get(struct sssnic_hw *hw, uint64_t *features)
+{
+   int ret;
+   struct sssnic_port_feature_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.opcode = SSSNIC_CMD_OPCODE_GET;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_PORT_FEATURE_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_PORT_FEATURE_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   *features = cmd.features;
+
+   return 0;
+}
+
+int
+sssnic_port_features_set(struct sssnic_hw *hw, uint64_t features)
+{
+   int ret;
+   struct sssnic_port_feature_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.features = features;
+   cmd.opcode = SSSNIC_CMD_OPCODE_SET;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_PORT_FEATURE_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_PORT_FEATURE_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+#define SSSNIC_MAX_NUM_RXTXQ_CTX_SET_IN_BULK   

[PATCH v4 23/32] net/sssnic: support Rx packet burst

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed wrong format of printing uint64_t.
---
 doc/guides/nics/features/sssnic.ini   |   2 +
 drivers/net/sssnic/sssnic_ethdev.c|   2 +
 drivers/net/sssnic/sssnic_ethdev_rx.c | 167 ++
 drivers/net/sssnic/sssnic_ethdev_rx.h |   2 +
 4 files changed, 173 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index aba0b78c95..320ac4533d 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -8,6 +8,8 @@ Link status  = Y
 Link status event= Y
 Queue start/stop = Y
 Rx interrupt = Y
+Scattered Rx = Y
+LRO  = Y
 Promiscuous mode = Y
 Allmulticast mode= Y
 Unicast MAC filter   = Y
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 99e6d6152a..021fabcbe5 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -769,6 +769,8 @@ sssnic_ethdev_init(struct rte_eth_dev *ethdev)
if (rte_eal_process_type() != RTE_PROC_PRIMARY)
return 0;
 
+   ethdev->rx_pkt_burst = sssnic_ethdev_rx_pkt_burst;
+
netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
pci_dev = RTE_ETH_DEV_TO_PCI(ethdev);
hw = rte_zmalloc("sssnic_hw", sizeof(struct sssnic_hw), 0);
diff --git a/drivers/net/sssnic/sssnic_ethdev_rx.c 
b/drivers/net/sssnic/sssnic_ethdev_rx.c
index 66045f7a98..82e65f2482 100644
--- a/drivers/net/sssnic/sssnic_ethdev_rx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_rx.c
@@ -1183,3 +1183,170 @@ sssnic_ethdev_rx_queue_stats_clear(struct rte_eth_dev 
*ethdev, uint16_t qid)
memset(&rxq->stats, 0, sizeof(rxq->stats));
}
 };
+
+static inline void
+sssnic_ethdev_rx_csum_offload(struct sssnic_ethdev_rxq *rxq,
+   struct rte_mbuf *rxm, volatile struct sssnic_ethdev_rx_desc *rxd)
+{
+   /* no errors */
+   if (likely(rxd->status_err == 0)) {
+   rxm->ol_flags |= RTE_MBUF_F_RX_IP_CKSUM_GOOD |
+RTE_MBUF_F_RX_L4_CKSUM_GOOD;
+   return;
+   }
+
+   /* bypass hw crc error*/
+   if (unlikely(rxd->hw_crc_err)) {
+   rxm->ol_flags |= RTE_MBUF_F_RX_IP_CKSUM_UNKNOWN;
+   return;
+   }
+
+   if (rxd->ip_csum_err) {
+   rxm->ol_flags |= RTE_MBUF_F_RX_IP_CKSUM_BAD;
+   rxq->stats.csum_errors++;
+   }
+
+   if (rxd->tcp_csum_err || rxd->udp_csum_err || rxd->sctp_crc_err) {
+   rxm->ol_flags |= RTE_MBUF_F_RX_L4_CKSUM_BAD;
+   rxq->stats.csum_errors++;
+   }
+
+   if (unlikely(rxd->other_err))
+   rxq->stats.other_errors++;
+}
+
+static inline void
+sssnic_ethdev_rx_vlan_offload(struct rte_mbuf *rxm,
+   volatile struct sssnic_ethdev_rx_desc *rxd)
+{
+   if (rxd->vlan_en == 0 || rxd->vlan == 0) {
+   rxm->vlan_tci = 0;
+   return;
+   }
+
+   rxm->vlan_tci = rxd->vlan;
+   rxm->ol_flags |= RTE_MBUF_F_RX_VLAN | RTE_MBUF_F_RX_VLAN_STRIPPED;
+}
+
+static inline void
+sssnic_ethdev_rx_segments(struct sssnic_ethdev_rxq *rxq, struct rte_mbuf *head,
+   uint32_t remain_size)
+{
+   struct sssnic_ethdev_rx_entry *rxe;
+   struct rte_mbuf *curr, *prev = head;
+   uint16_t rx_buf_size = rxq->rx_buf_size;
+   uint16_t ci;
+   uint32_t rx_size;
+
+   while (remain_size > 0) {
+   ci = sssnic_ethdev_rxq_ci_get(rxq);
+   rxe = &rxq->rxe[ci];
+   curr = rxe->pktmbuf;
+
+   sssnic_ethdev_rxq_consume(rxq, 1);
+
+   rx_size = RTE_MIN(remain_size, rx_buf_size);
+   remain_size -= rx_size;
+
+   curr->data_len = rx_size;
+   curr->next = NULL;
+   prev->next = curr;
+   prev = curr;
+   head->nb_segs++;
+   }
+}
+
+uint16_t
+sssnic_ethdev_rx_pkt_burst(void *rx_queue, struct rte_mbuf **rx_pkts,
+   uint16_t nb_pkts)
+{
+   struct sssnic_ethdev_rxq *rxq = (struct sssnic_ethdev_rxq *)rx_queue;
+   struct sssnic_ethdev_rx_entry *rxe;
+   struct rte_mbuf *rxm;
+   struct sssnic_ethdev_rx_desc *rxd, rx_desc;
+   uint16_t ci, idle_entries;
+   uint16_t rx_buf_size;
+   uint32_t rx_size;
+   uint64_t nb_rx = 0;
+   uint64_t rx_bytes = 0;
+
+   ci = sssnic_ethdev_rxq_ci_get(rxq);
+   rx_buf_size = rxq->rx_buf_size;
+   rxd = &rx_desc;
+
+   while (nb_rx < nb_pkts) {
+   rxd->dword0 = __atomic_load_n(&rxq->desc[ci].dword0,
+   __ATOMIC_ACQUIRE);
+   /* check rx done */
+   if (!rxd->done)
+   break;
+
+   rxd->dword1 = rxq->desc[ci].dword1;
+   rxd->dword2 = rxq->desc[ci].dword2;
+   rxd->dword3 = rxq->desc[ci].dword3;
+
+   /* reset rx d

[PATCH v4 22/32] net/sssnic: add basic and extended stats ops

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 doc/guides/nics/features/sssnic.ini  |   3 +
 drivers/net/sssnic/base/sssnic_api.c | 154 +
 drivers/net/sssnic/base/sssnic_api.h | 116 +++
 drivers/net/sssnic/base/sssnic_cmd.h |  12 +
 drivers/net/sssnic/meson.build   |   1 +
 drivers/net/sssnic/sssnic_ethdev.c   |   6 +
 drivers/net/sssnic/sssnic_ethdev_rx.c|  30 ++
 drivers/net/sssnic/sssnic_ethdev_rx.h|   4 +
 drivers/net/sssnic/sssnic_ethdev_stats.c | 391 +++
 drivers/net/sssnic/sssnic_ethdev_stats.h |  18 ++
 drivers/net/sssnic/sssnic_ethdev_tx.c|  36 +++
 drivers/net/sssnic/sssnic_ethdev_tx.h|   4 +
 12 files changed, 775 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_stats.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_stats.h

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 359834ce4c..aba0b78c95 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -12,6 +12,9 @@ Promiscuous mode = Y
 Allmulticast mode= Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
+Basic stats  = Y
+Extended stats   = Y
+Stats per queue  = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 81020387bd..9f063112f2 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1005,3 +1005,157 @@ sssnic_port_resource_clean(struct sssnic_hw *hw)
 
return 0;
 }
+
+int
+sssnic_port_stats_get(struct sssnic_hw *hw, struct sssnic_port_stats *stats)
+{
+   int ret;
+   struct sssnic_port_stats_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len, resp_len;
+   struct {
+   struct sssnic_cmd_common common;
+   uint32_t size;
+   uint32_t resvd0;
+   struct sssnic_port_stats stats;
+   uint64_t rsvd1[6];
+   } resp;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_GET_PORT_STATS_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   memset(&resp, 0, sizeof(resp));
+   resp_len = sizeof(resp);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&resp, &resp_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (resp_len == 0 || resp.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_PORT_STATS_CMD, len=%u, 
status=%u",
+   resp_len, resp.common.status);
+   return -EIO;
+   }
+
+   memcpy(stats, &resp.stats, sizeof(resp.stats));
+
+   return 0;
+}
+
+int
+sssnic_port_stats_clear(struct sssnic_hw *hw)
+{
+   int ret;
+   struct sssnic_port_stats_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_CLEAR_PORT_STATS_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_CLEAN_PORT_RES_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_mac_stats_get(struct sssnic_hw *hw, struct sssnic_mac_stats *stats)
+{
+   int ret;
+   struct sssnic_msg msg;
+   uint32_t cmd_len, resp_len;
+   struct sssnic_mac_stats_cmd cmd;
+   struct {
+   struct sssnic_cmd_common common;
+   struct sssnic_mac_stats stats;
+   uint64_t resvd[15];
+   } *resp;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.port = SSSNIC_PHY_PORT(hw);
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_GET_NETIF_MAC_STATS_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_NETIF_MODULE, SSSNIC_MSG_TYPE_REQ);
+
+   resp_len = sizeof(*resp);
+   resp = rte_zmalloc(NULL, resp_len, 0);
+   if (resp == NULL) {
+   PMD_DRV_LOG(ERR,
+   "Failed to alloc memory for mac stats response cmd");
+   return -ENOMEM;

[PATCH v4 21/32] net/sssnic: add allmulticast and promiscuous ops

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini |  2 +
 drivers/net/sssnic/sssnic_ethdev.c  | 72 +
 2 files changed, 74 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index e3b8166629..359834ce4c 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -8,6 +8,8 @@ Link status  = Y
 Link status event= Y
 Queue start/stop = Y
 Rx interrupt = Y
+Promiscuous mode = Y
+Allmulticast mode= Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
 Linux= Y
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index b59c4fd3ad..e1c805aeea 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -652,6 +652,74 @@ sssnic_ethdev_reset(struct rte_eth_dev *ethdev)
return 0;
 }
 
+static int
+sssnic_ethdev_allmulticast_enable(struct rte_eth_dev *ethdev)
+{
+   struct sssnic_netdev *netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
+   uint32_t rx_mode;
+   int ret;
+
+   rx_mode = netdev->rx_mode | SSSNIC_ETHDEV_RX_ALL_MCAST;
+   ret = sssnic_ethdev_rx_mode_set(ethdev, rx_mode);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to set rx_mode: %x", rx_mode);
+   return ret;
+   }
+
+   return 0;
+}
+
+static int
+sssnic_ethdev_allmulticast_disable(struct rte_eth_dev *ethdev)
+{
+   struct sssnic_netdev *netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
+   uint32_t rx_mode;
+   int ret;
+
+   rx_mode = netdev->rx_mode & (~SSSNIC_ETHDEV_RX_ALL_MCAST);
+   ret = sssnic_ethdev_rx_mode_set(ethdev, rx_mode);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to set rx_mode: %x", rx_mode);
+   return ret;
+   }
+
+   return 0;
+}
+
+static int
+sssnic_ethdev_promiscuous_enable(struct rte_eth_dev *ethdev)
+{
+   struct sssnic_netdev *netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
+   uint32_t rx_mode;
+   int ret;
+
+   rx_mode = netdev->rx_mode | SSSNIC_ETHDEV_RX_PROMISC;
+   ret = sssnic_ethdev_rx_mode_set(ethdev, rx_mode);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to set rx_mode: %x", rx_mode);
+   return ret;
+   }
+
+   return 0;
+}
+
+static int
+sssnic_ethdev_promiscuous_disable(struct rte_eth_dev *ethdev)
+{
+   struct sssnic_netdev *netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
+   uint32_t rx_mode;
+   int ret;
+
+   rx_mode = netdev->rx_mode & (~SSSNIC_ETHDEV_RX_PROMISC);
+   ret = sssnic_ethdev_rx_mode_set(ethdev, rx_mode);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to set rx_mode: %x", rx_mode);
+   return ret;
+   }
+
+   return 0;
+}
+
 static const struct eth_dev_ops sssnic_ethdev_ops = {
.dev_start = sssnic_ethdev_start,
.dev_stop = sssnic_ethdev_stop,
@@ -676,6 +744,10 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.tx_queue_stop = sssnic_ethdev_tx_queue_stop,
.rx_queue_intr_enable = sssnic_ethdev_rx_queue_intr_enable,
.rx_queue_intr_disable = sssnic_ethdev_rx_queue_intr_disable,
+   .allmulticast_enable = sssnic_ethdev_allmulticast_enable,
+   .allmulticast_disable = sssnic_ethdev_allmulticast_disable,
+   .promiscuous_enable = sssnic_ethdev_promiscuous_enable,
+   .promiscuous_disable = sssnic_ethdev_promiscuous_disable,
 };
 
 static int
-- 
2.27.0

[PATCH v4 24/32] net/sssnic: support Tx packet burst

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed wrong format of printing uint64_t.
---
 doc/guides/nics/features/sssnic.ini   |   5 +
 drivers/net/sssnic/sssnic_ethdev.c|   1 +
 drivers/net/sssnic/sssnic_ethdev_tx.c | 404 ++
 drivers/net/sssnic/sssnic_ethdev_tx.h |   2 +
 4 files changed, 412 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 320ac4533d..7e6b70684a 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -9,11 +9,16 @@ Link status event= Y
 Queue start/stop = Y
 Rx interrupt = Y
 Scattered Rx = Y
+TSO  = Y
 LRO  = Y
 Promiscuous mode = Y
 Allmulticast mode= Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
+L3 checksum offload  = Y
+L4 checksum offload  = Y
+Inner L3 checksum= Y
+Inner L4 checksum= Y
 Basic stats  = Y
 Extended stats   = Y
 Stats per queue  = Y
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 021fabcbe5..328fb85d30 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -770,6 +770,7 @@ sssnic_ethdev_init(struct rte_eth_dev *ethdev)
return 0;
 
ethdev->rx_pkt_burst = sssnic_ethdev_rx_pkt_burst;
+   ethdev->tx_pkt_burst = sssnic_ethdev_tx_pkt_burst;
 
netdev = SSSNIC_ETHDEV_PRIVATE(ethdev);
pci_dev = RTE_ETH_DEV_TO_PCI(ethdev);
diff --git a/drivers/net/sssnic/sssnic_ethdev_tx.c 
b/drivers/net/sssnic/sssnic_ethdev_tx.c
index d167e3f307..533befb6ea 100644
--- a/drivers/net/sssnic/sssnic_ethdev_tx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_tx.c
@@ -171,6 +171,17 @@ enum sssnic_ethdev_txq_entry_type {
SSSNIC_ETHDEV_TXQ_ENTRY_EXTEND = 1,
 };
 
+struct sssnic_ethdev_tx_info {
+   /* offload enable flag */
+   uint16_t offload_en;
+   /*l4 payload offset*/
+   uint16_t payload_off;
+   /* number of txq entries */
+   uint16_t nb_entries;
+   /* number of tx segs */
+   uint16_t nb_segs;
+};
+
 #define SSSNIC_ETHDEV_TXQ_ENTRY_SZ_BITS 4
 #define SSSNIC_ETHDEV_TXQ_ENTRY_SZ (RTE_BIT32(SSSNIC_ETHDEV_TXQ_ENTRY_SZ_BITS))
 
@@ -182,12 +193,44 @@ enum sssnic_ethdev_txq_entry_type {
 #define SSSNIC_ETHDEV_TX_CI_DEF_COALESCING_TIME 16
 #define SSSNIC_ETHDEV_TX_CI_DEF_PENDING_TIME 4
 
+#define SSSNIC_ETHDEV_TX_CSUM_OFFLOAD_MASK 
\
+   (RTE_MBUF_F_TX_IP_CKSUM | RTE_MBUF_F_TX_TCP_CKSUM |\
+   RTE_MBUF_F_TX_UDP_CKSUM | RTE_MBUF_F_TX_SCTP_CKSUM |   \
+   RTE_MBUF_F_TX_OUTER_IP_CKSUM | RTE_MBUF_F_TX_TCP_SEG)
+
+#define SSSNIC_ETHDEV_TX_OFFLOAD_MASK  
\
+   (RTE_MBUF_F_TX_VLAN | SSSNIC_ETHDEV_TX_CSUM_OFFLOAD_MASK)
+
+#define SSSNIC_ETHDEV_TX_MAX_NUM_SEGS 38
+#define SSSNIC_ETHDEV_TX_MAX_SEG_SIZE 65535
+#define SSSNIC_ETHDEV_TX_MAX_PAYLOAD_OFF 221
+#define SSSNIC_ETHDEV_TX_DEF_MSS 0x3e00
+#define SSSNIC_ETHDEV_TX_MIN_MSS 0x50
+#define SSSNIC_ETHDEV_TX_COMPACT_SEG_MAX_SIZE 0x3fff
+
+#define SSSNIC_ETHDEV_TXQ_DESC_ENTRY(txq, idx) 
\
+   (SSSNIC_WORKQ_ENTRY_CAST((txq)->workq, idx,\
+   struct sssnic_ethdev_tx_desc))
+
+#define SSSNIC_ETHDEV_TXQ_OFFLOAD_ENTRY(txq, idx)  
\
+   SSSNIC_WORKQ_ENTRY_CAST((txq)->workq, idx, \
+   struct sssnic_ethdev_tx_offload)
+
+#define SSSNIC_ETHDEV_TXQ_SEG_ENTRY(txq, idx)  
\
+   SSSNIC_WORKQ_ENTRY_CAST((txq)->workq, idx, struct sssnic_ethdev_tx_seg)
+
 static inline uint16_t
 sssnic_ethdev_txq_num_used_entries(struct sssnic_ethdev_txq *txq)
 {
return sssnic_workq_num_used_entries(txq->workq);
 }
 
+static inline uint16_t
+sssnic_ethdev_txq_num_idle_entries(struct sssnic_ethdev_txq *txq)
+{
+   return sssnic_workq_num_idle_entries(txq->workq);
+}
+
 static inline uint16_t
 sssnic_ethdev_txq_ci_get(struct sssnic_ethdev_txq *txq)
 {
@@ -212,6 +255,12 @@ sssnic_ethdev_txq_consume(struct sssnic_ethdev_txq *txq, 
uint16_t num_entries)
sssnic_workq_consume_fast(txq->workq, num_entries);
 }
 
+static inline void
+sssnic_ethdev_txq_produce(struct sssnic_ethdev_txq *txq, uint16_t num_entries)
+{
+   sssnic_workq_produce_fast(txq->workq, num_entries);
+}
+
 int
 sssnic_ethdev_tx_queue_setup(struct rte_eth_dev *ethdev, uint16_t tx_queue_id,
uint16_t nb_tx_desc, unsigned int socket_id,
@@ -706,3 +755,358 @@ sssnic_ethdev_tx_queue_stats_clear(struct rte_eth_dev 
*ethdev, uint16_t qid)
*(stat++) = 0;
}
 }
+
+static inline uint16_t
+sssnic_ethdev_tx_payload_calc(struct rte_mbuf *tx_mbuf)
+{
+   if ((tx_mbuf->ol_flags & RTE_MBUF_F_TX_TUNNEL_MASK) != 0) {
+   uint64_t mask = RTE_MBU

[PATCH v4 26/32] net/sssnic: support dev MTU set

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/sssnic_ethdev.c | 7 +++
 1 file changed, 7 insertions(+)

diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index a00e96bebe..b086e91d2d 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -733,6 +733,12 @@ sssnic_ethdev_promiscuous_disable(struct rte_eth_dev 
*ethdev)
return 0;
 }
 
+static int
+sssnic_ethdev_mtu_set(struct rte_eth_dev *ethdev, uint16_t mtu)
+{
+   return sssnic_ethdev_tx_max_size_set(ethdev, mtu);
+}
+
 static const struct eth_dev_ops sssnic_ethdev_ops = {
.dev_start = sssnic_ethdev_start,
.dev_stop = sssnic_ethdev_stop,
@@ -770,6 +776,7 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.rss_hash_update = sssnic_ethdev_rss_hash_update,
.reta_update = sssnic_ethdev_rss_reta_update,
.reta_query = sssnic_ethdev_rss_reta_query,
+   .mtu_set = sssnic_ethdev_mtu_set,
 };
 
 static int
-- 
2.27.0

[PATCH v4 25/32] net/sssnic: add RSS support

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Removed error.h from including files.
---
 doc/guides/nics/features/sssnic.ini|   4 +
 drivers/net/sssnic/base/sssnic_api.c   | 338 ++
 drivers/net/sssnic/base/sssnic_api.h   |  36 +++
 drivers/net/sssnic/base/sssnic_cmd.h   |  58 
 drivers/net/sssnic/meson.build |   1 +
 drivers/net/sssnic/sssnic_ethdev.c |  16 ++
 drivers/net/sssnic/sssnic_ethdev.h |   2 +
 drivers/net/sssnic/sssnic_ethdev_rss.c | 377 +
 drivers/net/sssnic/sssnic_ethdev_rss.h |  20 ++
 drivers/net/sssnic/sssnic_ethdev_rx.c  |  13 +
 10 files changed, 865 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_rss.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_rss.h

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 7e6b70684a..020a9e7056 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -15,6 +15,10 @@ Promiscuous mode = Y
 Allmulticast mode= Y
 Unicast MAC filter   = Y
 Multicast MAC filter = Y
+RSS hash = Y
+RSS key update   = Y
+RSS reta update  = Y
+Inner RSS= Y
 L3 checksum offload  = Y
 L4 checksum offload  = Y
 Inner L3 checksum= Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 9f063112f2..32b24e841c 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1159,3 +1159,341 @@ sssnic_mac_stats_clear(struct sssnic_hw *hw)
 
return 0;
 }
+
+int
+sssnic_rss_enable_set(struct sssnic_hw *hw, bool state)
+{
+   int ret;
+   struct sssnic_rss_enable_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.state = state ? 1 : 0;
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_ENABLE_RSS_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_ENABLE_RSS_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+static int
+sssnic_rss_profile_config(struct sssnic_hw *hw, bool new)
+{
+   int ret;
+   struct sssnic_rss_profile_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.opcode = new ? SSSNIC_RSS_PROFILE_CMD_OP_NEW :
+SSSNIC_RSS_PROFILE_CMD_OP_DEL;
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_RSS_PROFILE_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_RSS_PROFILE_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_rss_profile_create(struct sssnic_hw *hw)
+{
+   return sssnic_rss_profile_config(hw, true);
+}
+
+int
+sssnic_rss_profile_destroy(struct sssnic_hw *hw)
+{
+   return sssnic_rss_profile_config(hw, false);
+}
+
+int
+sssnic_rss_hash_key_set(struct sssnic_hw *hw, uint8_t *key, uint16_t len)
+{
+   int ret;
+   struct sssnic_rss_hash_key_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   if (len > sizeof(cmd.key)) {
+   PMD_DRV_LOG(ERR, "Invalid rss hash key length: %u", len);
+   return -EINVAL;
+   }
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.opcode = SSSNIC_CMD_OPCODE_SET;
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   rte_memcpy(cmd.key, key, len);
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_RSS_HASH_KEY_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Ba

[PATCH v4 27/32] net/sssnic: support dev queue info get

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/sssnic_ethdev.c|  2 ++
 drivers/net/sssnic/sssnic_ethdev_rx.c | 13 +
 drivers/net/sssnic/sssnic_ethdev_rx.h |  2 ++
 drivers/net/sssnic/sssnic_ethdev_tx.c | 11 +++
 drivers/net/sssnic/sssnic_ethdev_tx.h |  2 ++
 5 files changed, 30 insertions(+)

diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index b086e91d2d..bde8d89ddc 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -777,6 +777,8 @@ static const struct eth_dev_ops sssnic_ethdev_ops = {
.reta_update = sssnic_ethdev_rss_reta_update,
.reta_query = sssnic_ethdev_rss_reta_query,
.mtu_set = sssnic_ethdev_mtu_set,
+   .rxq_info_get = sssnic_ethdev_rx_queue_info_get,
+   .txq_info_get = sssnic_ethdev_tx_queue_info_get,
 };
 
 static int
diff --git a/drivers/net/sssnic/sssnic_ethdev_rx.c 
b/drivers/net/sssnic/sssnic_ethdev_rx.c
index 2874a93a54..6c5f209262 100644
--- a/drivers/net/sssnic/sssnic_ethdev_rx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_rx.c
@@ -1363,3 +1363,16 @@ sssnic_ethdev_rx_pkt_burst(void *rx_queue, struct 
rte_mbuf **rx_pkts,
 
return nb_rx;
 }
+
+void
+sssnic_ethdev_rx_queue_info_get(struct rte_eth_dev *ethdev,
+   uint16_t rx_queue_id, struct rte_eth_rxq_info *qinfo)
+{
+   struct sssnic_ethdev_rxq *rxq = ethdev->data->rx_queues[rx_queue_id];
+
+   qinfo->rx_buf_size = rxq->rx_buf_size;
+   qinfo->nb_desc = rxq->depth;
+   qinfo->queue_state = ethdev->data->rx_queue_state[rx_queue_id];
+   qinfo->mp = rxq->mp;
+   qinfo->conf.rx_free_thresh = rxq->rx_free_thresh;
+}
diff --git a/drivers/net/sssnic/sssnic_ethdev_rx.h 
b/drivers/net/sssnic/sssnic_ethdev_rx.h
index b0b35dee73..20e4d1ac0e 100644
--- a/drivers/net/sssnic/sssnic_ethdev_rx.h
+++ b/drivers/net/sssnic/sssnic_ethdev_rx.h
@@ -44,5 +44,7 @@ void sssnic_ethdev_rx_queue_stats_clear(struct rte_eth_dev 
*ethdev,
uint16_t qid);
 uint16_t sssnic_ethdev_rx_pkt_burst(void *rx_queue, struct rte_mbuf **rx_pkts,
uint16_t nb_pkts);
+void sssnic_ethdev_rx_queue_info_get(struct rte_eth_dev *ethdev,
+   uint16_t rx_queue_id, struct rte_eth_rxq_info *qinfo);
 
 #endif
diff --git a/drivers/net/sssnic/sssnic_ethdev_tx.c 
b/drivers/net/sssnic/sssnic_ethdev_tx.c
index 533befb6ea..51931df645 100644
--- a/drivers/net/sssnic/sssnic_ethdev_tx.c
+++ b/drivers/net/sssnic/sssnic_ethdev_tx.c
@@ -1110,3 +1110,14 @@ sssnic_ethdev_tx_pkt_burst(void *tx_queue, struct 
rte_mbuf **tx_pkts,
 
return nb_tx;
 }
+
+void
+sssnic_ethdev_tx_queue_info_get(struct rte_eth_dev *ethdev,
+   uint16_t tx_queue_id, struct rte_eth_txq_info *qinfo)
+{
+   struct sssnic_ethdev_txq *txq = ethdev->data->tx_queues[tx_queue_id];
+
+   qinfo->nb_desc = txq->depth;
+   qinfo->queue_state = ethdev->data->tx_queue_state[tx_queue_id];
+   qinfo->conf.tx_free_thresh = txq->tx_free_thresh;
+}
diff --git a/drivers/net/sssnic/sssnic_ethdev_tx.h 
b/drivers/net/sssnic/sssnic_ethdev_tx.h
index 3a7cd47080..6130ade4d1 100644
--- a/drivers/net/sssnic/sssnic_ethdev_tx.h
+++ b/drivers/net/sssnic/sssnic_ethdev_tx.h
@@ -39,5 +39,7 @@ void sssnic_ethdev_tx_queue_stats_clear(struct rte_eth_dev 
*ethdev,
uint16_t qid);
 uint16_t sssnic_ethdev_tx_pkt_burst(void *tx_queue, struct rte_mbuf **tx_pkts,
uint16_t nb_pkts);
+void sssnic_ethdev_tx_queue_info_get(struct rte_eth_dev *ethdev,
+   uint16_t tx_queue_id, struct rte_eth_txq_info *qinfo);
 
 #endif /* _SSSNIC_ETHDEV_TX_H_ */
-- 
2.27.0

[PATCH v4 29/32] net/sssnic: add dev flow control ops

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini  |  1 +
 drivers/net/sssnic/base/sssnic_api.c | 68 
 drivers/net/sssnic/base/sssnic_api.h |  4 ++
 drivers/net/sssnic/base/sssnic_cmd.h | 11 +
 drivers/net/sssnic/sssnic_ethdev.c   | 65 ++
 5 files changed, 149 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index a40b509558..9bf05cb968 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -26,6 +26,7 @@ Inner L4 checksum= Y
 Basic stats  = Y
 Extended stats   = Y
 Stats per queue  = Y
+Flow control = P
 FW version   = Y
 Linux= Y
 ARMv8= Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 12aca16995..d91896cdd2 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1533,3 +1533,71 @@ sssnic_fw_version_get(struct sssnic_hw *hw, struct 
sssnic_fw_version *version)
 
return 0;
 }
+
+int
+sssnic_flow_ctrl_set(struct sssnic_hw *hw, bool autoneg, bool rx_en, bool 
tx_en)
+{
+   int ret;
+   struct sssnic_msg msg;
+   struct sssnic_flow_ctrl_cmd cmd;
+   uint32_t cmdlen = sizeof(cmd);
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.auto_neg = autoneg ? 1 : 0;
+   cmd.rx_en = rx_en ? 1 : 0;
+   cmd.tx_en = tx_en ? 1 : 0;
+   cmd.opcode = SSSNIC_CMD_OPCODE_SET;
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmdlen,
+   SSSNIC_PORT_FLOW_CTRL_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmdlen, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmdlen == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_PORT_FLOW_CTRL_CMD, len=%u, 
status=%u",
+   cmdlen, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_flow_ctrl_get(struct sssnic_hw *hw, bool *autoneg, bool *rx_en,
+   bool *tx_en)
+{
+   int ret;
+   struct sssnic_msg msg;
+   struct sssnic_flow_ctrl_cmd cmd;
+   uint32_t cmdlen = sizeof(cmd);
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.opcode = SSSNIC_CMD_OPCODE_GET;
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmdlen,
+   SSSNIC_PORT_FLOW_CTRL_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmdlen, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmdlen == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_PORT_FLOW_CTRL_CMD, len=%u, 
status=%u",
+   cmdlen, cmd.common.status);
+   return -EIO;
+   }
+
+   *autoneg = cmd.auto_neg;
+   *rx_en = cmd.rx_en;
+   *tx_en = cmd.tx_en;
+
+   return 0;
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index bd4f01d388..36544a5dc3 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -465,5 +465,9 @@ int sssnic_rss_indir_table_get(struct sssnic_hw *hw, 
uint16_t *entry,
uint32_t num_entries);
 int sssnic_fw_version_get(struct sssnic_hw *hw,
struct sssnic_fw_version *version);
+int sssnic_flow_ctrl_set(struct sssnic_hw *hw, bool autoneg, bool rx_en,
+   bool tx_en);
+int sssnic_flow_ctrl_get(struct sssnic_hw *hw, bool *autoneg, bool *rx_en,
+   bool *tx_en);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index 9da07770b1..d2054fad5a 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -414,4 +414,15 @@ struct sssnic_fw_version_get_cmd {
uint8_t time[20];
 };
 
+struct sssnic_flow_ctrl_cmd {
+   struct sssnic_cmd_common common;
+   uint8_t port;
+   uint8_t opcode;
+   uint16_t resvd0;
+   uint8_t auto_neg;
+   uint8_t rx_en;
+   uint8_t tx_en;
+   uint8_t resvd1[5];
+};
+
 #endif /* _SSSNIC_CMD_H_ */
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index cceaa5c8be..8999693027 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -758,6 +758,69 @@ sssnic_ethdev_fw_version_get(struct rte_eth_dev *ethdev, 
char *fw_version,
return 0;
 }
 
+static int
+sssnic_ethdev_flow_ctrl_set(struct rte_eth_dev *ethdev,
+   struct rte_eth_fc_c

[PATCH v4 28/32] net/sssnic: support dev firmware version get

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini  |  1 +
 drivers/net/sssnic/base/sssnic_api.c | 36 
 drivers/net/sssnic/base/sssnic_api.h |  9 +++
 drivers/net/sssnic/base/sssnic_cmd.h |  8 +++
 drivers/net/sssnic/sssnic_ethdev.c   | 20 
 5 files changed, 74 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 020a9e7056..a40b509558 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -26,6 +26,7 @@ Inner L4 checksum= Y
 Basic stats  = Y
 Extended stats   = Y
 Stats per queue  = Y
+FW version   = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 32b24e841c..12aca16995 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1497,3 +1497,39 @@ sssnic_rss_indir_table_get(struct sssnic_hw *hw, 
uint16_t *entry,
sssnic_ctrlq_cmd_destroy(hw, cmd);
return ret;
 }
+
+int
+sssnic_fw_version_get(struct sssnic_hw *hw, struct sssnic_fw_version *version)
+{
+   int ret;
+   struct sssnic_msg msg;
+   struct sssnic_fw_version_get_cmd cmd;
+   uint32_t cmdlen = sizeof(cmd);
+   int len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.type = 1; /* get MPU firmware version */
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmdlen,
+   SSSNIC_GET_FW_VERSION_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_COMM_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmdlen, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmdlen == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_GET_FW_VERSION_CMD, len=%u, 
status=%u",
+   cmdlen, cmd.common.status);
+   return -EIO;
+   }
+
+   len = RTE_MIN(sizeof(version->version), sizeof(cmd.version));
+   rte_memcpy(version->version, cmd.version, len);
+   len = RTE_MIN(sizeof(version->time), sizeof(cmd.time));
+   rte_memcpy(version->time, cmd.time, len);
+
+   return 0;
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index 1d80b93e38..bd4f01d388 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -402,6 +402,13 @@ enum sssnic_rss_hash_engine_type {
SSSNIC_RSS_HASH_ENGINE_COUNT,
 };
 
+#define SSSNIC_FW_VERSION_LEN 16
+#define SSSNIC_FW_TIME_LEN 20
+struct sssnic_fw_version {
+   char version[SSSNIC_FW_VERSION_LEN];
+   char time[SSSNIC_FW_VERSION_LEN];
+};
+
 int sssnic_msix_attr_get(struct sssnic_hw *hw, uint16_t msix_idx,
struct sssnic_msix_attr *attr);
 int sssnic_msix_attr_set(struct sssnic_hw *hw, uint16_t msix_idx,
@@ -456,5 +463,7 @@ int sssnic_rss_indir_table_set(struct sssnic_hw *hw, const 
uint16_t *entry,
uint32_t num_entries);
 int sssnic_rss_indir_table_get(struct sssnic_hw *hw, uint16_t *entry,
uint32_t num_entries);
+int sssnic_fw_version_get(struct sssnic_hw *hw,
+   struct sssnic_fw_version *version);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index 56818471b6..9da07770b1 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -406,4 +406,12 @@ struct sssnic_rss_indir_table_cmd {
uint16_t entry[256];
 };
 
+struct sssnic_fw_version_get_cmd {
+   struct sssnic_cmd_common common;
+   uint16_t type;
+   uint16_t resvd;
+   uint8_t version[16];
+   uint8_t time[20];
+};
+
 #endif /* _SSSNIC_CMD_H_ */
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index bde8d89ddc..cceaa5c8be 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -739,6 +739,25 @@ sssnic_ethdev_mtu_set(struct rte_eth_dev *ethdev, uint16_t 
mtu)
return sssnic_ethdev_tx_max_size_set(ethdev, mtu);
 }
 
+static int
+sssnic_ethdev_fw_version_get(struct rte_eth_dev *ethdev, char *fw_version,
+   size_t fw_size)
+{
+   struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
+   struct sssnic_fw_version version;
+   int ret;
+
+   ret = sssnic_fw_version_get(hw, &version);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Failed to get firmware version");
+   return ret;
+   }
+
+   snprintf(fw_version, fw_size, "%s", version.version);
+
+   return 0;
+}
+
 static const struct eth_dev_ops sssnic_ethdev_ops = {
.dev_start = sssnic_ethdev_start,
.dev_stop = sssnic_ethdev_stop,
@@ -779,6 +

[PATCH v4 30/32] net/sssnic: support VLAN offload and filter

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 doc/guides/nics/features/sssnic.ini  |  2 +
 drivers/net/sssnic/base/sssnic_api.c | 34 +++
 drivers/net/sssnic/base/sssnic_api.h |  1 +
 drivers/net/sssnic/base/sssnic_cmd.h |  9 +++
 drivers/net/sssnic/sssnic_ethdev.c   | 87 
 5 files changed, 133 insertions(+)

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index 9bf05cb968..f5738ac934 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -19,6 +19,8 @@ RSS hash = Y
 RSS key update   = Y
 RSS reta update  = Y
 Inner RSS= Y
+VLAN filter  = Y
+VLAN offload = Y
 L3 checksum offload  = Y
 L4 checksum offload  = Y
 Inner L3 checksum= Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index d91896cdd2..68c16c9c1e 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1601,3 +1601,37 @@ sssnic_flow_ctrl_get(struct sssnic_hw *hw, bool 
*autoneg, bool *rx_en,
 
return 0;
 }
+
+int
+sssnic_vlan_filter_set(struct sssnic_hw *hw, uint16_t vid, bool add)
+{
+   int ret;
+   struct sssnic_vlan_filter_set_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.add = add ? 1 : 0;
+   cmd.vid = vid;
+   cmd_len = sizeof(cmd);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_SET_PORT_VLAN_FILTER_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_SET_PORT_VLAN_FILTER_CMD, 
len=%u, status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index 36544a5dc3..28b235dda2 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -469,5 +469,6 @@ int sssnic_flow_ctrl_set(struct sssnic_hw *hw, bool 
autoneg, bool rx_en,
bool tx_en);
 int sssnic_flow_ctrl_get(struct sssnic_hw *hw, bool *autoneg, bool *rx_en,
bool *tx_en);
+int sssnic_vlan_filter_set(struct sssnic_hw *hw, uint16_t vid, bool add);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index d2054fad5a..3e70d0e223 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -425,4 +425,13 @@ struct sssnic_flow_ctrl_cmd {
uint8_t resvd1[5];
 };
 
+struct sssnic_vlan_filter_set_cmd {
+   struct sssnic_cmd_common common;
+   uint16_t function;
+   uint8_t add;
+   uint8_t resvd0;
+   uint16_t vid;
+   uint16_t resvd1;
+};
+
 #endif /* _SSSNIC_CMD_H_ */
diff --git a/drivers/net/sssnic/sssnic_ethdev.c 
b/drivers/net/sssnic/sssnic_ethdev.c
index 8999693027..8a1ccff70b 100644
--- a/drivers/net/sssnic/sssnic_ethdev.c
+++ b/drivers/net/sssnic/sssnic_ethdev.c
@@ -16,6 +16,7 @@
 #include "sssnic_ethdev_rss.h"
 
 static int sssnic_ethdev_init(struct rte_eth_dev *ethdev);
+static void sssnic_ethdev_vlan_filter_clean(struct rte_eth_dev *ethdev);
 
 static int
 sssnic_ethdev_infos_get(struct rte_eth_dev *ethdev,
@@ -340,6 +341,7 @@ sssnic_ethdev_release(struct rte_eth_dev *ethdev)
 {
struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
 
+   sssnic_ethdev_vlan_filter_clean(ethdev);
sssnic_ethdev_link_intr_disable(ethdev);
sssnic_ethdev_tx_queue_all_release(ethdev);
sssnic_ethdev_rx_queue_all_release(ethdev);
@@ -821,6 +823,89 @@ sssnic_ethdev_flow_ctrl_get(struct rte_eth_dev *ethdev,
return 0;
 }
 
+static int
+sssnic_ethdev_vlan_offload_set(struct rte_eth_dev *ethdev, int mask)
+{
+   struct sssnic_hw *hw = SSSNIC_ETHDEV_TO_HW(ethdev);
+   struct rte_eth_conf *dev_conf = ðdev->data->dev_conf;
+   uint8_t vlan_strip_en;
+   uint32_t vlan_filter_en;
+   int ret;
+
+   if (mask & RTE_ETH_VLAN_STRIP_MASK) {
+   if (dev_conf->rxmode.offloads & RTE_ETH_RX_OFFLOAD_VLAN_STRIP)
+   vlan_strip_en = 1;
+   else
+   vlan_strip_en = 0;
+
+   ret = sssnic_vlan_strip_enable_set(hw, vlan_strip_en);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to %s vlan strip offload",
+   vlan_strip_en ? "enable" : "disable");
+   return ret

[PATCH v4 31/32] net/sssnic: add generic flow ops

[wanry]

From: Renyong Wan 

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
v2:
* Fixed 'mask->hdr.src_addr' will always evaluate to 'true'.
* Removed error.h from including files.
---
 doc/guides/nics/features/sssnic.ini |   12 +
 drivers/net/sssnic/base/sssnic_api.c|  264 ++
 drivers/net/sssnic/base/sssnic_api.h|   22 +
 drivers/net/sssnic/base/sssnic_cmd.h|   71 ++
 drivers/net/sssnic/base/sssnic_hw.h |3 +
 drivers/net/sssnic/base/sssnic_misc.h   |7 +
 drivers/net/sssnic/meson.build  |2 +
 drivers/net/sssnic/sssnic_ethdev.c  |   12 +
 drivers/net/sssnic/sssnic_ethdev.h  |1 +
 drivers/net/sssnic/sssnic_ethdev_fdir.c | 1017 +++
 drivers/net/sssnic/sssnic_ethdev_fdir.h |  332 
 drivers/net/sssnic/sssnic_ethdev_flow.c |  981 ++
 drivers/net/sssnic/sssnic_ethdev_flow.h |   11 +
 drivers/net/sssnic/sssnic_ethdev_rx.c   |   18 +
 14 files changed, 2753 insertions(+)
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_fdir.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_fdir.h
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_flow.c
 create mode 100644 drivers/net/sssnic/sssnic_ethdev_flow.h

diff --git a/doc/guides/nics/features/sssnic.ini 
b/doc/guides/nics/features/sssnic.ini
index f5738ac934..57e7440d86 100644
--- a/doc/guides/nics/features/sssnic.ini
+++ b/doc/guides/nics/features/sssnic.ini
@@ -33,3 +33,15 @@ FW version   = Y
 Linux= Y
 ARMv8= Y
 x86-64   = Y
+
+[rte_flow items]
+any  = Y
+eth  = Y
+ipv4 = Y
+ipv6 = Y
+tcp  = Y
+udp  = Y
+vxlan= Y
+
+[rte_flow actions]
+queue= Y
diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 68c16c9c1e..0e965442fd 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1635,3 +1635,267 @@ sssnic_vlan_filter_set(struct sssnic_hw *hw, uint16_t 
vid, bool add)
 
return 0;
 }
+
+int
+sssnic_tcam_enable_set(struct sssnic_hw *hw, bool enabled)
+{
+   struct sssnic_tcam_enable_set_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   int ret;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+   cmd.enabled = enabled ? 1 : 0;
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_SET_TCAM_ENABLE_CMD, SSSNIC_MPU_FUNC_IDX,
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   if (cmd.common.status == SSSNIC_TCAM_CMD_STATUS_UNSUPPORTED)
+   PMD_DRV_LOG(WARNING,
+   "SSSNIC_SET_TCAM_ENABLED_CMD is unsupported");
+   else
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_SET_TCAM_ENABLE_CMD, 
len=%u, status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_tcam_flush(struct sssnic_hw *hw)
+{
+   struct sssnic_tcam_flush_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+   int ret;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd_len = sizeof(cmd);
+   cmd.function = SSSNIC_FUNC_IDX(hw);
+
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len, SSSNIC_FLUSH_TCAM_CMD,
+   SSSNIC_MPU_FUNC_IDX, SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   if (cmd.common.status == SSSNIC_TCAM_CMD_STATUS_UNSUPPORTED)
+   PMD_DRV_LOG(WARNING,
+   "SSSNIC_FLUSH_TCAM_CMD is unsupported");
+   else
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_FLUSH_TCAM_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+   return 0;
+}
+
+int
+sssnic_tcam_disable_and_flush(struct sssnic_hw *hw)
+{
+   int ret;
+
+   ret = sssnic_tcam_enable_set(hw, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Could not disable TCAM");
+   return ret;
+   }
+
+   ret = sssnic_tcam_flush(hw);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Could not flush TCAM");
+   return ret;
+   }
+

[PATCH v4 32/32] net/sssnic: add VF dev support

[wanry]

From: Renyong Wan 

In comparison to PF, VF PMD does not support the following features:

1. link up and link down set
2. promiscuous enable and disable
3. MAC stats in extend xstats

Signed-off-by: Steven Song 
Signed-off-by: Renyong Wan 
---
 drivers/net/sssnic/base/sssnic_api.c | 42 +
 drivers/net/sssnic/base/sssnic_api.h |  2 +
 drivers/net/sssnic/base/sssnic_cmd.h |  6 +++
 drivers/net/sssnic/base/sssnic_hw.c  | 19 ++--
 drivers/net/sssnic/base/sssnic_hw.h  |  4 ++
 drivers/net/sssnic/sssnic_ethdev.c   | 60 +++-
 drivers/net/sssnic/sssnic_ethdev_stats.c | 29 ++--
 7 files changed, 153 insertions(+), 9 deletions(-)

diff --git a/drivers/net/sssnic/base/sssnic_api.c 
b/drivers/net/sssnic/base/sssnic_api.c
index 0e965442fd..2d829bc884 100644
--- a/drivers/net/sssnic/base/sssnic_api.c
+++ b/drivers/net/sssnic/base/sssnic_api.c
@@ -1899,3 +1899,45 @@ sssnic_tcam_entry_del(struct sssnic_hw *hw, uint32_t 
entry_idx)
 
return 0;
 }
+
+static int
+sssnic_vf_port_register_op(struct sssnic_hw *hw, bool op)
+{
+   int ret;
+   struct sssnic_vf_port_register_cmd cmd;
+   struct sssnic_msg msg;
+   uint32_t cmd_len;
+
+   memset(&cmd, 0, sizeof(cmd));
+   cmd.op = op ? 1 : 0;
+   cmd_len = sizeof(cmd);
+   sssnic_msg_init(&msg, (uint8_t *)&cmd, cmd_len,
+   SSSNIC_REGISTER_VF_PORT_CMD, SSSNIC_PF_FUNC_IDX(hw),
+   SSSNIC_LAN_MODULE, SSSNIC_MSG_TYPE_REQ);
+   ret = sssnic_mbox_send(hw, &msg, (uint8_t *)&cmd, &cmd_len, 0);
+   if (ret != 0) {
+   PMD_DRV_LOG(ERR, "Failed to send mbox message, ret=%d", ret);
+   return ret;
+   }
+
+   if (cmd_len == 0 || cmd.common.status != 0) {
+   PMD_DRV_LOG(ERR,
+   "Bad response to SSSNIC_REGISTER_VF_PORT_CMD, len=%u, 
status=%u",
+   cmd_len, cmd.common.status);
+   return -EIO;
+   }
+
+   return 0;
+}
+
+int
+sssnic_vf_port_register(struct sssnic_hw *hw)
+{
+   return sssnic_vf_port_register_op(hw, true);
+}
+
+int
+sssnic_vf_port_unregister(struct sssnic_hw *hw)
+{
+   return sssnic_vf_port_register_op(hw, false);
+}
diff --git a/drivers/net/sssnic/base/sssnic_api.h 
b/drivers/net/sssnic/base/sssnic_api.h
index 7a02ec61ee..2506682821 100644
--- a/drivers/net/sssnic/base/sssnic_api.h
+++ b/drivers/net/sssnic/base/sssnic_api.h
@@ -492,5 +492,7 @@ int sssnic_tcam_packet_type_filter_set(struct sssnic_hw 
*hw, uint8_t ptype,
 int sssnic_tcam_entry_add(struct sssnic_hw *hw,
struct sssnic_tcam_entry *entry);
 int sssnic_tcam_entry_del(struct sssnic_hw *hw, uint32_t entry_idx);
+int sssnic_vf_port_register(struct sssnic_hw *hw);
+int sssnic_vf_port_unregister(struct sssnic_hw *hw);
 
 #endif /* _SSSNIC_API_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_cmd.h 
b/drivers/net/sssnic/base/sssnic_cmd.h
index c75cb0dad3..058ab298f3 100644
--- a/drivers/net/sssnic/base/sssnic_cmd.h
+++ b/drivers/net/sssnic/base/sssnic_cmd.h
@@ -505,4 +505,10 @@ struct sssnic_tcam_entry_del_cmd {
uint32_t num; /* number of entries to be deleted */
 };
 
+struct sssnic_vf_port_register_cmd {
+   struct sssnic_cmd_common common;
+   uint8_t op; /* 0: unregister, 1: register */
+   uint8_t resvd[39];
+};
+
 #endif /* _SSSNIC_CMD_H_ */
diff --git a/drivers/net/sssnic/base/sssnic_hw.c 
b/drivers/net/sssnic/base/sssnic_hw.c
index 651a0aa7ef..0edd5b9508 100644
--- a/drivers/net/sssnic/base/sssnic_hw.c
+++ b/drivers/net/sssnic/base/sssnic_hw.c
@@ -345,10 +345,16 @@ sssnic_base_init(struct sssnic_hw *hw)
pci_dev = hw->pci_dev;
 
/* get base addresses of hw registers */
-   hw->cfg_base_addr =
-   (uint8_t *)pci_dev->mem_resource[SSSNIC_PCI_BAR_CFG].addr;
-   hw->mgmt_base_addr =
-   (uint8_t *)pci_dev->mem_resource[SSSNIC_PCI_BAR_MGMT].addr;
+   if (pci_dev->id.device_id == SSSNIC_VF_DEVICE_ID) {
+   uint8_t *addr =
+   (uint8_t 
*)pci_dev->mem_resource[SSSNIC_VF_PCI_BAR_CFG].addr;
+   hw->cfg_base_addr = addr + SSSNIC_VF_CFG_ADDR_OFFSET;
+   } else {
+   hw->cfg_base_addr =
+   (uint8_t 
*)pci_dev->mem_resource[SSSNIC_PCI_BAR_CFG].addr;
+   hw->mgmt_base_addr =
+   (uint8_t 
*)pci_dev->mem_resource[SSSNIC_PCI_BAR_MGMT].addr;
+   }
hw->db_base_addr =
(uint8_t *)pci_dev->mem_resource[SSSNIC_PCI_BAR_DB].addr;
hw->db_mem_len =
@@ -365,7 +371,10 @@ sssnic_base_init(struct sssnic_hw *hw)
PMD_DRV_LOG(ERR, "Doorbell is not enabled!");
return -EBUSY;
}
-   sssnic_af_setup(hw);
+
+   if (SSSNIC_FUNC_TYPE(hw) != SSSNIC_FUNC_TYPE_VF)
+   sssnic_af_setup(hw);
+
sssnic_msix_all_disable(hw);
sssnic_pf_status_set(hw, SSSNIC_PF_STATUS_INIT);
 
diff --git a/drivers/net/sssnic/base

Re: DPDK community: RTE_FLOW support for P4-programmable devices

[Jerin Jacob]

On Fri, Sep 1, 2023 at 12:27 PM Ori Kam  wrote:
>
>
>
> > -Original Message-
> > From: Jerin Jacob 
> > Sent: Friday, September 1, 2023 5:07 AM
> >
> > On Thu, Aug 31, 2023 at 4:02 PM Ori Kam  wrote:
> > >
> > > Hi
> >
> > > > >>
> > > > >> 3. Everybody on the call agreed that the P4-programmable devices from
> > > > Intel,
> > > > >> AMD and others need to be fully supported by DPDK and that there are
> > > > some
> > > > >> gaps in RTE_FLOW to be fixed for supporting these devices.
> > > > >
> > > > > Personally, It makes sense to me to have normative DPDK API to send p4
> > > > > runtime message to the
> > > > > ethdev so that we have "vendor neutral + DPDK based" p4 runtime
> > backend.
> > > > >
> > > > > I prefer to have specialized ethdev ops for this due to the following
> > reasons.
> > > > >
> > > > > # If the ethdev has both real TCAM based HW(for existing rte_flow
> > > > > patterns and actions) and SW agent to receive P4 runtime message etc.
> > > > > Typically, it needs to take a different path in driver to talk. 
> > > > > Assume, if you
> > > > > have cascaded patterns/actions, One is targeted for TCAM and other for
> > > > > SW agent for p4, one
> > > > > need to have serious amount checking for dispatching.It complicates
> > > > > the driver and forbid to have
> > > > > driver optimization especially cases for templates etc. if user making
> > > > > rules for both category of HW.
> > > > >
> > > >
> > > > Indeed I am not against dedicated APIs for P4 runtime backend.
> > > >
> > > > But assuming there is a dedicated rte_flow item for P4, how it is
> > > > different than dedicated API in above scenario?
> > > > If driver detects P4 runtime specific rule, it can bail it out to SW 
> > > > agent.
> > > > Can you please elaborate the complexity it introduces?
> >
> > Assume, normal existing rte-flow programming include a bunch of
> > register writes and
> > p4 runtime backend is more of SW ring. If a template has both patterns
> > and actions
> > as cascaded, it will be difficult for driver to optimize the template.
> >
> >
> > > >
> > > > > # All we need "char buffer//string" based communication ethdev <-> 
> > > > > app.
> > > > >
> > > >
> > > > Yes, and both a dedicated API or dedicated rte_flow item can provide
> > > > medium for this communication.
> > > >
> > > > rte_flow one has flexibility & extensibility advantages, but maybe not
> > > > as straightforward as an API.
> > >
> > > I think not using the rte_flow will also require duplication of all the 
> > > rule
> > handling functions and table creations, for example aync rule create/destroy
> > query ..
> >
> > Yes. That is a fair point. I am OK with exposing as rte_flow.
> > As a driver implementation note, to get rid of the above problem,
> > driver can choose to have pseudo ethdev devices for p4 if needed(if
> > driver finds difficult to combine TCAM based on HW rules and p4
> > runtime rule).
> >
>
> What about the following concept:
> The p4 compiler can generate the translation to known PMD objects in rte_flow,
> then when a command is sent from the p4 agent to the offload using GRPC or 
> any other way, the DPDK will convert from
> p4 protocol to rte_flow commands (this should be very fast since the commands 
> are known and the mapping is already
> defined).
>
> To support the above idea we need to add two new functions to rte_flow (each 
> function will be implemented in PMD level)

If the implemention of rte_flow_p4_runtime((p4 command based on the p4 spec))
is defined in PMD level, The PMD driver to decide to use rte_flow or
something else.
I think it is good, this is actually going back to specialized API.
BTW, rte_flow prefix is not needed in that case.



> Rte_flow_register_p4(void *p4_info, void *p4_blob)
> {
> Creates the static structures/objects
> Internal register the p4 commands to PMD translation table.
> }
>
> Rte_flow_p4_runtime(p4 command based on the p4 spec)
> {
> Based on the registered mapping, translate the command to rte_flow 
> commands.
> Rte_flow_xxx() calls
> }
>
> As far as I see, the above code will also allow PMD to implement internal 
> logic if needed, while from DPDK API,
> we will only add two new functions.
>
> >
> > >

Re: [PATCH v2 1/3] eventdev: introduce event dispatcher

[Mattias Rönnblom]

On 2023-08-18 08:09, Jerin Jacob wrote:




2) Have new library lib/event_dispatcher



Should the library be named event_dispatcher, or simply dispatcher?

Underscore in library isn't exactly aesthetically pleasing, and shorter 
is better. Also, the rte_event_* namespace is avoided altogether.


On the other hand "dispatcher" is a little too generic, and somewhat 
grandiose name, for a relatively simple thing. "event_dispatcher" makes 
the relation to eventdev obvious.

Re: [PATCH v2 1/3] eventdev: introduce event dispatcher

[Jerin Jacob]

On Fri, Sep 1, 2023 at 4:23 PM Mattias Rönnblom  wrote:
>
> On 2023-08-18 08:09, Jerin Jacob wrote:
>
> 
>
> > 2) Have new library lib/event_dispatcher
> >
>
> Should the library be named event_dispatcher, or simply dispatcher?

Looks good to me.

>
> Underscore in library isn't exactly aesthetically pleasing, and shorter

> is better. Also, the rte_event_* namespace is avoided altogether.

+1

>
> On the other hand "dispatcher" is a little too generic, and somewhat
> grandiose name, for a relatively simple thing. "event_dispatcher" makes
> the relation to eventdev obvious.
>
> 

[PATCH v2 0/8] add rte flow support for cpfl

[Yuying Zhang]

This patchset add rte_flow support for cpfl driver.
It depends on the following patch set:
http://patchwork.dpdk.org/project/dpdk/cover/20230816150541.144532-1-beilei.x...@intel.com/

Wenjing Qiao (4):
  net/cpfl: parse flow parser file in devargs
  net/cpfl: add flow json parser
  net/cpfl: add FXP low level implementation
  net/cpfl: setup ctrl path

Yuying Zhang (4):
  net/cpfl: set up rte flow skeleton
  net/cpfl: add fxp rule module
  net/cpfl: add fxp flow engine
  net/cpfl: add flow support for representor

 doc/guides/nics/cpfl.rst|   45 +
 doc/guides/rel_notes/release_23_11.rst  |1 +
 drivers/net/cpfl/cpfl_actions.h |  858 +++
 drivers/net/cpfl/cpfl_controlq.c|  803 ++
 drivers/net/cpfl/cpfl_controlq.h|   75 +
 drivers/net/cpfl/cpfl_ethdev.c  |  390 -
 drivers/net/cpfl/cpfl_ethdev.h  |  109 ++
 drivers/net/cpfl/cpfl_flow.c|  339 +
 drivers/net/cpfl/cpfl_flow.h|   85 ++
 drivers/net/cpfl/cpfl_flow_engine_fxp.c |  603 
 drivers/net/cpfl/cpfl_flow_parser.c | 1769 +++
 drivers/net/cpfl/cpfl_flow_parser.h |  220 +++
 drivers/net/cpfl/cpfl_fxp_rule.c|  297 
 drivers/net/cpfl/cpfl_fxp_rule.h|   68 +
 drivers/net/cpfl/cpfl_representor.c |   29 +
 drivers/net/cpfl/cpfl_rules.c   |  126 ++
 drivers/net/cpfl/cpfl_rules.h   |  306 
 drivers/net/cpfl/cpfl_vchnl.c   |  144 ++
 drivers/net/cpfl/meson.build|   14 +
 19 files changed, 6280 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/cpfl/cpfl_actions.h
 create mode 100644 drivers/net/cpfl/cpfl_controlq.c
 create mode 100644 drivers/net/cpfl/cpfl_controlq.h
 create mode 100644 drivers/net/cpfl/cpfl_flow.c
 create mode 100644 drivers/net/cpfl/cpfl_flow.h
 create mode 100644 drivers/net/cpfl/cpfl_flow_engine_fxp.c
 create mode 100644 drivers/net/cpfl/cpfl_flow_parser.c
 create mode 100644 drivers/net/cpfl/cpfl_flow_parser.h
 create mode 100644 drivers/net/cpfl/cpfl_fxp_rule.c
 create mode 100644 drivers/net/cpfl/cpfl_fxp_rule.h
 create mode 100644 drivers/net/cpfl/cpfl_rules.c
 create mode 100644 drivers/net/cpfl/cpfl_rules.h

-- 
2.25.1

[PATCH v2 1/8] net/cpfl: parse flow parser file in devargs

[Yuying Zhang]

From: Wenjing Qiao 

Add devargs "flow_parser" for rte_flow json parser.

Signed-off-by: Wenjing Qiao 
---
 doc/guides/nics/cpfl.rst   | 32 
 drivers/net/cpfl/cpfl_ethdev.c | 38 +-
 drivers/net/cpfl/cpfl_ethdev.h |  3 +++
 drivers/net/cpfl/meson.build   |  6 ++
 4 files changed, 78 insertions(+), 1 deletion(-)

diff --git a/doc/guides/nics/cpfl.rst b/doc/guides/nics/cpfl.rst
index c20334230b..7032dd1a1a 100644
--- a/doc/guides/nics/cpfl.rst
+++ b/doc/guides/nics/cpfl.rst
@@ -128,12 +128,24 @@ Runtime Configuration
 
 -a BDF,representor=vf[0-3],representor=c1pf1
 
+- ``flow_parser`` (default ``not enabled``)
+
+  The PMD supports using a JSON file to parse rte_flow tokens into low level 
hardware
+  resources defined in a DDP package file.
+
+  The user can specify the path of json file, for example::
+
+-a ca:00.0,flow_parser="refpkg.json"
+
+  Then the PMD will load json file for device ``ca:00.0``.
+  The parameter is optional.
 
 Driver compilation and testing
 --
 
 Refer to the document :doc:`build_and_test` for details.
 
+Rte flow need to install json-c library.
 
 Features
 
@@ -164,3 +176,23 @@ Hairpin queue
 E2100 Series can loopback packets from RX port to TX port.
 This feature is called port-to-port or hairpin.
 Currently, the PMD only supports single port hairpin.
+
+Rte_flow
+~
+
+Rte_flow uses a json file to direct CPF PMD to parse rte_flow tokens into
+low level hardware resources defined in a DDP package file.
+
+#. install json-c library::
+
+   .. code-block:: console
+
+   git clone https://github.com/json-c/json-c.git
+   cd json-c
+   git checkout 777dd06be83ef7fac71c2218b565557cd068a714
+
+#. run testpmd with the json file::
+
+   .. code-block:: console
+
+   dpdk-testpmd -c 0x3 -n 4 -a 
:af:00.6,vport=[0],flow_parser="refpkg.json" -- -i
diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 88c1479f3a..3c4a6a4724 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -21,6 +21,10 @@
 #define CPFL_RX_SINGLE_Q   "rx_single"
 #define CPFL_VPORT "vport"
 
+#ifdef CPFL_FLOW_JSON_SUPPORT
+#define CPFL_FLOW_PARSER   "flow_parser"
+#endif
+
 rte_spinlock_t cpfl_adapter_lock;
 /* A list for all adapters, one adapter matches one PCI device */
 struct cpfl_adapter_list cpfl_adapter_list;
@@ -31,6 +35,9 @@ static const char * const cpfl_valid_args_first[] = {
CPFL_TX_SINGLE_Q,
CPFL_RX_SINGLE_Q,
CPFL_VPORT,
+#ifdef CPFL_FLOW_JSON_SUPPORT
+   CPFL_FLOW_PARSER,
+#endif
NULL
 };
 
@@ -1537,6 +1544,24 @@ parse_repr(const char *key __rte_unused, const char 
*value, void *args)
return 0;
 }
 
+#ifdef CPFL_FLOW_JSON_SUPPORT
+static int
+parse_file(const char *key, const char *value, void *args)
+{
+   char *name = args;
+
+   if (strlen(value) > CPFL_FLOW_FILE_LEN - 1) {
+   PMD_DRV_LOG(ERR, "file path(%s) is too long.", value);
+   return -1;
+   }
+
+   PMD_DRV_LOG(DEBUG, "value:\"%s\" for key:\"%s\"", value, key);
+   strlcpy(name, value, CPFL_FLOW_FILE_LEN);
+
+   return 0;
+}
+#endif
+
 static int
 cpfl_parse_devargs(struct rte_pci_device *pci_dev, struct cpfl_adapter_ext 
*adapter, bool first)
 {
@@ -1585,7 +1610,18 @@ cpfl_parse_devargs(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *adap
 &adapter->base.is_rx_singleq);
if (ret != 0)
goto fail;
-
+#ifdef CPFL_FLOW_JSON_SUPPORT
+   if (rte_kvargs_get(kvlist, CPFL_FLOW_PARSER)) {
+   ret = rte_kvargs_process(kvlist, CPFL_FLOW_PARSER,
+&parse_file, cpfl_args->flow_parser);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Failed to parser flow_parser, ret: 
%d", ret);
+   goto fail;
+   }
+   } else {
+   cpfl_args->flow_parser[0] = '\0';
+   }
+#endif
 fail:
rte_kvargs_free(kvlist);
return ret;
diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index a891bd8df9..ed730cc0e9 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -87,6 +87,8 @@
 #define ACC_LCE_ID 15
 #define IMC_MBX_EFD_ID 0
 
+#define CPFL_FLOW_FILE_LEN 100
+
 struct cpfl_vport_param {
struct cpfl_adapter_ext *adapter;
uint16_t devarg_id; /* arg id from user */
@@ -100,6 +102,7 @@ struct cpfl_devargs {
uint16_t req_vport_nb;
uint8_t repr_args_num;
struct rte_eth_devargs repr_args[CPFL_REPR_ARG_NUM_MAX];
+   char flow_parser[CPFL_FLOW_FILE_LEN];
 };
 
 struct p2p_queue_chunks_info {
diff --git a/drivers/net/cpfl/meson.build b/drivers/net/cpfl/meson.build
index fb075c6860..0be25512c3 100644
--- a/drivers/net/cpfl/meson.build
+++ b/drivers/net/cpfl/meson.build
@@ -38,3 +38,9 @@ if arch

[PATCH v2 2/8] net/cpfl: add flow json parser

[Yuying Zhang]

From: Wenjing Qiao 

A JSON file will be used to direct DPDK CPF PMD to
parse rte_flow tokens into low level hardware resources
defined in a DDP package file.

Signed-off-by: Wenjing Qiao 
---
 drivers/net/cpfl/cpfl_flow_parser.c | 1769 +++
 drivers/net/cpfl/cpfl_flow_parser.h |  220 
 drivers/net/cpfl/meson.build|3 +
 3 files changed, 1992 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_flow_parser.c
 create mode 100644 drivers/net/cpfl/cpfl_flow_parser.h

diff --git a/drivers/net/cpfl/cpfl_flow_parser.c 
b/drivers/net/cpfl/cpfl_flow_parser.c
new file mode 100644
index 00..f14ceefed6
--- /dev/null
+++ b/drivers/net/cpfl/cpfl_flow_parser.c
@@ -0,0 +1,1769 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2023 Intel Corporation
+ */
+
+#include 
+#include 
+#include 
+
+#include "cpfl_flow_parser.h"
+#include "cpfl_ethdev.h"
+#include "rte_malloc.h"
+
+static enum rte_flow_item_type
+cpfl_get_item_type_by_str(const char *type)
+{
+   if (strcmp(type, "eth") == 0)
+   return RTE_FLOW_ITEM_TYPE_ETH;
+   else if (strcmp(type, "ipv4") == 0)
+   return RTE_FLOW_ITEM_TYPE_IPV4;
+   else if (strcmp(type, "tcp") == 0)
+   return RTE_FLOW_ITEM_TYPE_TCP;
+   else if (strcmp(type, "udp") == 0)
+   return RTE_FLOW_ITEM_TYPE_UDP;
+   else if (strcmp(type, "vxlan") == 0)
+   return RTE_FLOW_ITEM_TYPE_VXLAN;
+   else if (strcmp(type, "icmp") == 0)
+   return RTE_FLOW_ITEM_TYPE_ICMP;
+   else if (strcmp(type, "vlan") == 0)
+   return RTE_FLOW_ITEM_TYPE_VLAN;
+
+   PMD_DRV_LOG(ERR, "Not support this type: %s.", type);
+   return RTE_FLOW_ITEM_TYPE_VOID;
+}
+
+static enum rte_flow_action_type
+cpfl_get_action_type_by_str(const char *type)
+{
+   if (strcmp(type, "vxlan_encap") == 0)
+   return RTE_FLOW_ACTION_TYPE_VXLAN_ENCAP;
+   else if (strcmp(type, "vxlan_decap") == 0)
+   return RTE_FLOW_ACTION_TYPE_VXLAN_DECAP;
+
+   PMD_DRV_LOG(ERR, "Not support this type: %s.", type);
+   return RTE_FLOW_ACTION_TYPE_VOID;
+}
+
+static const char *
+cpfl_json_object_to_string(json_object *object, const char *name)
+{
+   json_object *subobject;
+
+   if (!object) {
+   PMD_DRV_LOG(ERR, "object doesn't exist.");
+   return NULL;
+   }
+   subobject = json_object_object_get(object, name);
+   if (!subobject) {
+   PMD_DRV_LOG(ERR, "%s doesn't exist.", name);
+   return NULL;
+   }
+
+   return json_object_get_string(subobject);
+}
+
+static int
+cpfl_json_object_to_int(json_object *object, const char *name, int *value)
+{
+   json_object *subobject;
+
+   if (!object) {
+   PMD_DRV_LOG(ERR, "object doesn't exist.");
+   return -EINVAL;
+   }
+   subobject = json_object_object_get(object, name);
+   if (!subobject) {
+   PMD_DRV_LOG(ERR, "%s doesn't exist.", name);
+   return -EINVAL;
+   }
+   *value = json_object_get_int(subobject);
+
+   return 0;
+}
+
+static int
+cpfl_json_object_to_uint16(json_object *object, const char *name, uint16_t 
*value)
+{
+   json_object *subobject;
+
+   if (!object) {
+   PMD_DRV_LOG(ERR, "object doesn't exist.");
+   return -EINVAL;
+   }
+   subobject = json_object_object_get(object, name);
+   if (!subobject) {
+   PMD_DRV_LOG(ERR, "%s doesn't exist.", name);
+   return -EINVAL;
+   }
+   *value = json_object_get_int(subobject);
+
+   return 0;
+}
+
+static int
+cpfl_json_object_to_uint32(json_object *object, const char *name, uint32_t 
*value)
+{
+   json_object *subobject;
+
+   if (!object) {
+   PMD_DRV_LOG(ERR, "object doesn't exist.");
+   return -EINVAL;
+   }
+   subobject = json_object_object_get(object, name);
+   if (!subobject) {
+   PMD_DRV_LOG(ERR, "%s doesn't exist.", name);
+   return -EINVAL;
+   }
+   *value = json_object_get_int64(subobject);
+
+   return 0;
+}
+
+static int
+cpfl_flow_js_pattern_key_attr(json_object *cjson_pr_key_attr, struct 
cpfl_flow_js_pr *js_pr)
+{
+   int i, len;
+   struct cpfl_flow_js_pr_key_attr *attr;
+
+   len = json_object_array_length(cjson_pr_key_attr);
+   js_pr->key.attributes = rte_malloc(NULL, sizeof(struct 
cpfl_flow_js_pr_key_attr), 0);
+   if (!js_pr->key.attributes) {
+   PMD_DRV_LOG(ERR, "Failed to alloc memory.");
+   return -ENOMEM;
+   }
+   js_pr->key.attr_size = len;
+   attr = js_pr->key.attributes;
+   for (i = 0; i < len; i++) {
+   json_object *object;
+   const char *name;
+   uint16_t value = 0;
+   int ret;
+
+   object = json_object_array_get_idx(cjson_pr_key_attr, i);
+

[PATCH v2 3/8] net/cpfl: add FXP low level implementation

[Yuying Zhang]

From: Wenjing Qiao 

Add FXP low level implementation for CPFL rte_flow to
create/delete rules.

Signed-off-by: Wenjing Qiao 
---
 drivers/net/cpfl/cpfl_actions.h  | 858 +++
 drivers/net/cpfl/cpfl_controlq.c | 379 ++
 drivers/net/cpfl/cpfl_controlq.h |  51 ++
 drivers/net/cpfl/cpfl_rules.c| 126 +
 drivers/net/cpfl/cpfl_rules.h| 306 +++
 drivers/net/cpfl/meson.build |   2 +
 6 files changed, 1722 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_actions.h
 create mode 100644 drivers/net/cpfl/cpfl_controlq.c
 create mode 100644 drivers/net/cpfl/cpfl_controlq.h
 create mode 100644 drivers/net/cpfl/cpfl_rules.c
 create mode 100644 drivers/net/cpfl/cpfl_rules.h

diff --git a/drivers/net/cpfl/cpfl_actions.h b/drivers/net/cpfl/cpfl_actions.h
new file mode 100644
index 00..7b82119e39
--- /dev/null
+++ b/drivers/net/cpfl/cpfl_actions.h
@@ -0,0 +1,858 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2001-2023 Intel Corporation
+ */
+
+#ifndef _CPFL_ACTIONS_H_
+#define _CPFL_ACTIONS_H_
+
+#include "base/idpf_osdep.h"
+
+#pragma pack(1)
+
+union cpfl_action_set {
+   uint32_t data;
+
+   struct {
+   uint32_t val : 24;
+   uint32_t idx : 4;
+   uint32_t tag : 1;
+   uint32_t prec : 3;
+   } set_24b_a;
+
+   struct {
+   uint32_t val : 24;
+   uint32_t idx : 3;
+   uint32_t tag : 2;
+   uint32_t prec : 3;
+   } set_24b_b;
+
+   struct {
+   uint32_t val : 16;
+   uint32_t idx : 4;
+   uint32_t unused : 6;
+   uint32_t tag : 3;
+   uint32_t prec : 3;
+   } set_16b;
+
+   struct {
+   uint32_t val_a : 8;
+   uint32_t val_b : 8;
+   uint32_t idx_a : 4;
+   uint32_t idx_b : 4;
+   uint32_t tag : 5;
+   uint32_t prec : 3;
+   } set_8b;
+
+   struct {
+   uint32_t val : 10;
+   uint32_t ena : 10;
+   uint32_t idx : 4;
+   uint32_t tag : 5;
+   uint32_t prec : 3;
+   } set_1b;
+
+   struct {
+   uint32_t val : 24;
+   uint32_t tag : 5;
+   uint32_t prec : 3;
+   } nop;
+
+   struct {
+   uint32_t val : 24;
+   uint32_t tag : 5;
+   uint32_t prec : 3;
+   } chained_24b;
+
+   struct {
+   uint32_t val : 24;
+   uint32_t tag : 5;
+   uint32_t prec : 3;
+   } aux_flags;
+};
+
+struct cpfl_action_set_ext {
+#define CPFL_ACTION_SET_EXT_CNT 2
+   union cpfl_action_set acts[CPFL_ACTION_SET_EXT_CNT];
+};
+
+#pragma pack()
+
+/**
+ * cpfl_act_nop - Encode a NOP action
+ */
+static inline union cpfl_action_set
+cpfl_act_nop(void)
+{
+   union cpfl_action_set act;
+
+   act.data = 0;
+   return act;
+}
+
+/**
+ * cpfl_is_nop_action - Indicate if an action set is a NOP
+ */
+static inline bool
+cpfl_is_nop_action(union cpfl_action_set *act)
+{
+   return act->data == cpfl_act_nop().data;
+}
+
+#define CPFL_MAKE_MASK32(b, s) uint32_t)1 << (b)) - 1) << (s))
+
+#define CPFL_ACT_PREC_MAX  7
+#define CPFL_ACT_PREC_S29
+#define CPFL_ACT_PREC_MCPFL_MAKE_MASK32(3, CPFL_ACT_PREC_S)
+#define CPFL_ACT_PREC_SET(p)   \
+   (((uint32_t)(p) << CPFL_ACT_PREC_S) & CPFL_ACT_PREC_M)
+#define CPFL_ACT_PREC_CHECK(p) ((p) > 0 && (p) <= CPFL_ACT_PREC_MAX)
+
+#define CPFL_METADATA_ID_CNT   32  /* Max number of metadata IDs */
+#define CPFL_METADATA_STRUCT_MAX_SZ128 /* Max metadata size per ID */
+
+/***
+ * 1-Bit Actions
+ 
**/
+#define CPFL_ACT_1B_OP_S   24
+#define CPFL_ACT_1B_OP_M   CPFL_MAKE_MASK32(5, CPFL_ACT_1B_OP_S)
+#define CPFL_ACT_1B_OP ((uint32_t)(0x01) << CPFL_ACT_1B_OP_S)
+
+#define CPFL_ACT_1B_VAL_S  0
+#define CPFL_ACT_1B_VAL_M  CPFL_MAKE_MASK32(10, CPFL_ACT_1B_VAL_S)
+#define CPFL_ACT_1B_EN_S   10
+#define CPFL_ACT_1B_EN_M   CPFL_MAKE_MASK32(10, CPFL_ACT_1B_EN_S)
+#define CPFL_ACT_1B_INDEX_S20
+#define CPFL_ACT_1B_INDEX_MCPFL_MAKE_MASK32(4, CPFL_ACT_1B_INDEX_S)
+
+/* 1-bit actions currently uses only INDEX of 0 */
+#define CPFL_ACT_MAKE_1B(prec, en, val) \
+   ((CPFL_ACT_PREC_SET(prec)) | CPFL_ACT_1B_OP | \
+uint32_t)0) << CPFL_ACT_1B_INDEX_S) & CPFL_ACT_1B_INDEX_M) | \
+(((uint32_t)(en) << CPFL_ACT_1B_EN_S) & CPFL_ACT_1B_EN_M) | \
+(((uint32_t)(val) << CPFL_ACT_1B_VAL_S) & CPFL_ACT_1B_VAL_M))
+
+enum cpfl_act_1b_op {
+   CPFL_ACT_1B_OP_DROP = 0x01,
+   CPFL_ACT_1B_OP_HDR_SPLIT= 0x02,
+   CPFL_ACT_1B_OP_DIR_CHANGE   = 0x04,
+   CPFL_ACT_1B_OP_DEFER_DROP   

[PATCH v2 4/8] net/cpfl: setup ctrl path

[Yuying Zhang]

From: Wenjing Qiao 

Setup the control vport and control queue for flow offloading.

Signed-off-by: Yuying Zhang 
Signed-off-by: Beilei Xing 
Signed-off-by: Qi Zhang 
Signed-off-by: Wenjing Qiao 
---
 drivers/net/cpfl/cpfl_ethdev.c | 267 +
 drivers/net/cpfl/cpfl_ethdev.h |  14 ++
 drivers/net/cpfl/cpfl_vchnl.c  | 144 ++
 3 files changed, 425 insertions(+)

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 3c4a6a4724..22f3e72894 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -1657,6 +1657,10 @@ cpfl_handle_vchnl_event_msg(struct cpfl_adapter_ext 
*adapter, uint8_t *msg, uint
return;
}
 
+   /* ignore if it is ctrl vport */
+   if (adapter->ctrl_vport.base.vport_id == vc_event->vport_id)
+   return;
+
vport = cpfl_find_vport(adapter, vc_event->vport_id);
if (!vport) {
PMD_DRV_LOG(ERR, "Can't find vport.");
@@ -1852,6 +1856,260 @@ cpfl_dev_alarm_handler(void *param)
rte_eal_alarm_set(CPFL_ALARM_INTERVAL, cpfl_dev_alarm_handler, adapter);
 }
 
+static int
+cpfl_stop_cfgqs(struct cpfl_adapter_ext *adapter)
+{
+   int i, ret;
+
+   for (i = 0; i < CPFL_TX_CFGQ_NUM; i++) {
+   ret = idpf_vc_queue_switch(&adapter->ctrl_vport.base, i, false, 
false);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to disable Tx config queue.");
+   return ret;
+   }
+   }
+
+   for (i = 0; i < CPFL_RX_CFGQ_NUM; i++) {
+   ret = idpf_vc_queue_switch(&adapter->ctrl_vport.base, i, true, 
false);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to disable Rx config queue.");
+   return ret;
+   }
+   }
+
+   return 0;
+}
+
+static int
+cpfl_start_cfgqs(struct cpfl_adapter_ext *adapter)
+{
+   int i, ret;
+
+   ret = cpfl_config_ctlq_tx(adapter);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to configure Tx config queue.");
+   return ret;
+   }
+
+   ret = cpfl_config_ctlq_rx(adapter);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to configure Rx config queue.");
+   return ret;
+   }
+
+   for (i = 0; i < CPFL_TX_CFGQ_NUM; i++) {
+   ret = idpf_vc_queue_switch(&adapter->ctrl_vport.base, i, false, 
true);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to enable Tx config queue.");
+   return ret;
+   }
+   }
+
+   for (i = 0; i < CPFL_RX_CFGQ_NUM; i++) {
+   ret = idpf_vc_queue_switch(&adapter->ctrl_vport.base, i, true, 
true);
+   if (ret) {
+   PMD_DRV_LOG(ERR, "Fail to enable Rx config queue.");
+   return ret;
+   }
+   }
+
+   return 0;
+}
+
+static void
+cpfl_remove_cfgqs(struct cpfl_adapter_ext *adapter)
+{
+   struct idpf_hw *hw = (struct idpf_hw *)(&adapter->base.hw);
+   struct cpfl_ctlq_create_info *create_cfgq_info;
+   int i;
+
+   create_cfgq_info = adapter->cfgq_info;
+
+   for (i = 0; i < CPFL_CFGQ_NUM; i++) {
+   cpfl_vport_ctlq_remove(hw, adapter->ctlqp[i]);
+   if (create_cfgq_info[i].ring_mem.va)
+   idpf_free_dma_mem(&adapter->base.hw, 
&create_cfgq_info[i].ring_mem);
+   if (create_cfgq_info[i].buf_mem.va)
+   idpf_free_dma_mem(&adapter->base.hw, 
&create_cfgq_info[i].buf_mem);
+   }
+}
+
+static int
+cpfl_add_cfgqs(struct cpfl_adapter_ext *adapter)
+{
+   struct idpf_ctlq_info *cfg_cq;
+   int ret = 0;
+   int i = 0;
+
+   for (i = 0; i < CPFL_CFGQ_NUM; i++) {
+   ret = cpfl_vport_ctlq_add((struct idpf_hw *)(&adapter->base.hw),
+ &adapter->cfgq_info[i],
+ &cfg_cq);
+   if (ret || !cfg_cq) {
+   PMD_DRV_LOG(ERR, "ctlq add failed for queue id: %d",
+   adapter->cfgq_info[i].id);
+   cpfl_remove_cfgqs(adapter);
+   return ret;
+   }
+   PMD_DRV_LOG(INFO, "added cfgq to hw. queue id: %d",
+   adapter->cfgq_info[i].id);
+   adapter->ctlqp[i] = cfg_cq;
+   }
+
+   return ret;
+}
+
+#define CPFL_CFGQ_RING_LEN 512
+#define CPFL_CFGQ_DESCRIPTOR_SIZE  32
+#define CPFL_CFGQ_BUFFER_SIZE  256
+#define CPFL_CFGQ_RING_SIZE512
+
+static int
+cpfl_cfgq_setup(struct cpfl_adapter_ext *adapter)
+{
+   struct cpfl_ctlq_create_info *create_cfgq_info;
+   struct cpfl_vport *vport;
+   int i, err;
+   uint32_t ring_size = CPFL_CFGQ_RING_SIZE * sizeof(struct 
idpf_ctlq_desc);
+   uint32_t buf_size = CPFL_CFGQ_RING_SIZE * 

[PATCH v2 5/8] net/cpfl: set up rte flow skeleton

[Yuying Zhang]

Set up the rte_flow backend skeleton. Introduce the framework
to support different engines as rte_flow backend. Bridge rte_flow
driver API to flow engines.

Signed-off-by: Yuying Zhang 
---
 drivers/net/cpfl/cpfl_ethdev.c |  54 ++
 drivers/net/cpfl/cpfl_ethdev.h |   5 +
 drivers/net/cpfl/cpfl_flow.c   | 339 +
 drivers/net/cpfl/cpfl_flow.h   |  85 +
 drivers/net/cpfl/meson.build   |   3 +-
 5 files changed, 485 insertions(+), 1 deletion(-)
 create mode 100644 drivers/net/cpfl/cpfl_flow.c
 create mode 100644 drivers/net/cpfl/cpfl_flow.h

diff --git a/drivers/net/cpfl/cpfl_ethdev.c b/drivers/net/cpfl/cpfl_ethdev.c
index 22f3e72894..618a6a0fe2 100644
--- a/drivers/net/cpfl/cpfl_ethdev.c
+++ b/drivers/net/cpfl/cpfl_ethdev.c
@@ -15,6 +15,7 @@
 #include "cpfl_ethdev.h"
 #include 
 #include "cpfl_rxtx.h"
+#include "cpfl_flow.h"
 
 #define CPFL_REPRESENTOR   "representor"
 #define CPFL_TX_SINGLE_Q   "tx_single"
@@ -1074,6 +1075,19 @@ cpfl_dev_stop(struct rte_eth_dev *dev)
return 0;
 }
 
+static void
+cpfl_flow_free(struct cpfl_vport *vport)
+{
+   struct rte_flow *p_flow;
+
+   while ((p_flow = TAILQ_FIRST(&vport->itf.flow_list))) {
+   TAILQ_REMOVE(&vport->itf.flow_list, p_flow, next);
+   if (p_flow->engine->free)
+   p_flow->engine->free(p_flow);
+   rte_free(p_flow);
+   }
+}
+
 static int
 cpfl_p2p_queue_grps_del(struct idpf_vport *vport)
 {
@@ -1105,6 +1119,7 @@ cpfl_dev_close(struct rte_eth_dev *dev)
if (!adapter->base.is_rx_singleq && !adapter->base.is_tx_singleq)
cpfl_p2p_queue_grps_del(vport);
 
+   cpfl_flow_free(cpfl_vport);
idpf_vport_deinit(vport);
rte_free(cpfl_vport->p2p_q_chunks_info);
 
@@ -1117,6 +1132,29 @@ cpfl_dev_close(struct rte_eth_dev *dev)
return 0;
 }
 
+static int
+cpfl_dev_flow_ops_get(struct rte_eth_dev *dev,
+ const struct rte_flow_ops **ops)
+{
+   struct cpfl_itf *itf;
+
+   if (!dev)
+   return -EINVAL;
+
+   itf = CPFL_DEV_TO_ITF(dev);
+
+   /* only vport support rte_flow */
+   if (itf->type != CPFL_ITF_TYPE_VPORT)
+   return -ENOTSUP;
+#ifdef CPFL_FLOW_JSON_SUPPORT
+   *ops = &cpfl_flow_ops;
+#else
+   *ops = NULL;
+   PMD_DRV_LOG(NOTICE, "not support rte_flow, please install json-c 
library.");
+#endif
+   return 0;
+}
+
 static int
 cpfl_hairpin_get_peer_ports(struct rte_eth_dev *dev, uint16_t *peer_ports,
size_t len, uint32_t tx)
@@ -1318,6 +1356,7 @@ static const struct eth_dev_ops cpfl_eth_dev_ops = {
.xstats_get = cpfl_dev_xstats_get,
.xstats_get_names   = cpfl_dev_xstats_get_names,
.xstats_reset   = cpfl_dev_xstats_reset,
+   .flow_ops_get   = cpfl_dev_flow_ops_get,
.hairpin_cap_get= cpfl_hairpin_cap_get,
.rx_hairpin_queue_setup = cpfl_rx_hairpin_queue_setup,
.tx_hairpin_queue_setup = cpfl_tx_hairpin_queue_setup,
@@ -2283,6 +2322,13 @@ cpfl_adapter_ext_init(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *a
goto err_create_ctrl_vport;
}
 
+#ifdef CPFL_FLOW_JSON_SUPPORT
+   ret = cpfl_flow_init(adapter);
+   if (ret) {
+   PMD_INIT_LOG(ERR, "Failed to init flow module");
+   goto err_flow_init;
+   }
+#endif
adapter->cur_vports = 0;
adapter->cur_vport_nb = 0;
 
@@ -2290,6 +2336,10 @@ cpfl_adapter_ext_init(struct rte_pci_device *pci_dev, 
struct cpfl_adapter_ext *a
 
return ret;
 
+#ifdef CPFL_FLOW_JSON_SUPPORT
+err_flow_init:
+   cpfl_ctrl_path_close(adapter);
+#endif
 err_create_ctrl_vport:
rte_free(adapter->vports);
 err_vports_alloc:
@@ -2446,6 +2496,7 @@ cpfl_dev_vport_init(struct rte_eth_dev *dev, void 
*init_params)
cpfl_vport->itf.type = CPFL_ITF_TYPE_VPORT;
cpfl_vport->itf.adapter = adapter;
cpfl_vport->itf.data = dev->data;
+   TAILQ_INIT(&cpfl_vport->itf.flow_list);
adapter->vports[param->idx] = cpfl_vport;
adapter->cur_vports |= RTE_BIT32(param->devarg_id);
adapter->cur_vport_nb++;
@@ -2526,6 +2577,9 @@ cpfl_find_adapter_ext(struct rte_pci_device *pci_dev)
 static void
 cpfl_adapter_ext_deinit(struct cpfl_adapter_ext *adapter)
 {
+#ifdef CPFL_FLOW_JSON_SUPPORT
+   cpfl_flow_uninit(adapter);
+#endif
cpfl_ctrl_path_close(adapter);
rte_eal_alarm_cancel(cpfl_dev_alarm_handler, adapter);
cpfl_vport_map_uninit(adapter);
diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index 90e71a6550..40a27f8b74 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -143,9 +143,12 @@ enum cpfl_itf_type {
CPFL_ITF_TYPE_REPRESENTOR
 };
 
+TAILQ_HEAD(cpfl_flow_list, rte_flow);
+
 struct cpfl_itf {
enum cp

[PATCH v2 6/8] net/cpfl: add fxp rule module

[Yuying Zhang]

Added low level fxp module for rule packing / creation / destroying.

Signed-off-by: Yuying Zhang 
---
 drivers/net/cpfl/cpfl_controlq.c | 424 +++
 drivers/net/cpfl/cpfl_controlq.h |  24 ++
 drivers/net/cpfl/cpfl_ethdev.c   |  31 +++
 drivers/net/cpfl/cpfl_ethdev.h   |   6 +
 drivers/net/cpfl/cpfl_fxp_rule.c | 297 ++
 drivers/net/cpfl/cpfl_fxp_rule.h |  68 +
 drivers/net/cpfl/meson.build |   1 +
 7 files changed, 851 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_fxp_rule.c
 create mode 100644 drivers/net/cpfl/cpfl_fxp_rule.h

diff --git a/drivers/net/cpfl/cpfl_controlq.c b/drivers/net/cpfl/cpfl_controlq.c
index 476c78f235..ed76282b0c 100644
--- a/drivers/net/cpfl/cpfl_controlq.c
+++ b/drivers/net/cpfl/cpfl_controlq.c
@@ -331,6 +331,402 @@ cpfl_ctlq_add(struct idpf_hw *hw, struct 
cpfl_ctlq_create_info *qinfo,
return status;
 }
 
+/**
+ * cpfl_ctlq_send - send command to Control Queue (CTQ)
+ * @hw: pointer to hw struct
+ * @cq: handle to control queue struct to send on
+ * @num_q_msg: number of messages to send on control queue
+ * @q_msg: pointer to array of queue messages to be sent
+ *
+ * The caller is expected to allocate DMAable buffers and pass them to the
+ * send routine via the q_msg struct / control queue specific data struct.
+ * The control queue will hold a reference to each send message until
+ * the completion for that message has been cleaned.
+ */
+int
+cpfl_ctlq_send(struct idpf_hw *hw, struct idpf_ctlq_info *cq,
+  uint16_t num_q_msg, struct idpf_ctlq_msg q_msg[])
+{
+   struct idpf_ctlq_desc *desc;
+   int num_desc_avail = 0;
+   int status = 0;
+   int i = 0;
+
+   if (!cq || !cq->ring_size)
+   return -ENOBUFS;
+
+   idpf_acquire_lock(&cq->cq_lock);
+
+   /* Ensure there are enough descriptors to send all messages */
+   num_desc_avail = IDPF_CTLQ_DESC_UNUSED(cq);
+   if (num_desc_avail == 0 || num_desc_avail < num_q_msg) {
+   status = -ENOSPC;
+   goto sq_send_command_out;
+   }
+
+   for (i = 0; i < num_q_msg; i++) {
+   struct idpf_ctlq_msg *msg = &q_msg[i];
+   uint64_t msg_cookie;
+
+   desc = IDPF_CTLQ_DESC(cq, cq->next_to_use);
+   desc->opcode = CPU_TO_LE16(msg->opcode);
+   desc->pfid_vfid = CPU_TO_LE16(msg->func_id);
+   msg_cookie = *(uint64_t *)&msg->cookie;
+   desc->cookie_high =
+   CPU_TO_LE32(IDPF_HI_DWORD(msg_cookie));
+   desc->cookie_low =
+   CPU_TO_LE32(IDPF_LO_DWORD(msg_cookie));
+   desc->flags = CPU_TO_LE16((msg->host_id & IDPF_HOST_ID_MASK) <<
+   IDPF_CTLQ_FLAG_HOST_ID_S);
+   if (msg->data_len) {
+   struct idpf_dma_mem *buff = msg->ctx.indirect.payload;
+
+   desc->datalen |= CPU_TO_LE16(msg->data_len);
+   desc->flags |= CPU_TO_LE16(IDPF_CTLQ_FLAG_BUF);
+   desc->flags |= CPU_TO_LE16(IDPF_CTLQ_FLAG_RD);
+   /* Update the address values in the desc with the pa
+* value for respective buffer
+*/
+   desc->params.indirect.addr_high =
+   CPU_TO_LE32(IDPF_HI_DWORD(buff->pa));
+   desc->params.indirect.addr_low =
+   CPU_TO_LE32(IDPF_LO_DWORD(buff->pa));
+   idpf_memcpy(&desc->params, msg->ctx.indirect.context,
+   IDPF_INDIRECT_CTX_SIZE, IDPF_NONDMA_TO_DMA);
+   } else {
+   idpf_memcpy(&desc->params, msg->ctx.direct,
+   IDPF_DIRECT_CTX_SIZE, IDPF_NONDMA_TO_DMA);
+   }
+
+   /* Store buffer info */
+   cq->bi.tx_msg[cq->next_to_use] = msg;
+   (cq->next_to_use)++;
+   if (cq->next_to_use == cq->ring_size)
+   cq->next_to_use = 0;
+   }
+
+   /* Force memory write to complete before letting hardware
+* know that there are new descriptors to fetch.
+*/
+   idpf_wmb();
+   wr32(hw, cq->reg.tail, cq->next_to_use);
+
+sq_send_command_out:
+   idpf_release_lock(&cq->cq_lock);
+
+   return status;
+}
+
+/**
+ * __cpfl_ctlq_clean_sq - helper function to reclaim descriptors on HW write
+ * back for the requested queue
+ * @cq: pointer to the specific Control queue
+ * @clean_count: (input|output) number of descriptors to clean as input, and
+ * number of descriptors actually cleaned as output
+ * @msg_status: (output) pointer to msg pointer array to be populated; needs
+ * to be allocated by caller
+ * @force: (input) clean descriptors which were not done yet. Use with caution
+ * in kernel mode only
+ *
+ * Returns an array of message pointers as

[PATCH v2 7/8] net/cpfl: add fxp flow engine

[Yuying Zhang]

Adapt fxp low level as a flow engine.

Signed-off-by: Yuying Zhang 
---
 drivers/net/cpfl/cpfl_ethdev.h  |  81 
 drivers/net/cpfl/cpfl_flow_engine_fxp.c | 583 
 drivers/net/cpfl/meson.build|   1 +
 3 files changed, 665 insertions(+)
 create mode 100644 drivers/net/cpfl/cpfl_flow_engine_fxp.c

diff --git a/drivers/net/cpfl/cpfl_ethdev.h b/drivers/net/cpfl/cpfl_ethdev.h
index c29da92e81..0dbf8fb21b 100644
--- a/drivers/net/cpfl/cpfl_ethdev.h
+++ b/drivers/net/cpfl/cpfl_ethdev.h
@@ -92,6 +92,8 @@
 
 #define CPFL_RX_CFGQ_NUM   4
 #define CPFL_TX_CFGQ_NUM   4
+#define CPFL_FPCP_CFGQ_TX   0
+#define CPFL_FPCP_CFGQ_RX   1
 #define CPFL_CFGQ_NUM  8
 
 struct cpfl_vport_param {
@@ -203,6 +205,8 @@ struct cpfl_adapter_ext {
struct rte_hash *repr_whitelist_hash;
 
struct cpfl_flow_js_parser *flow_parser;
+   struct rte_bitmap *mod_bm;
+   void *mod_bm_mem;
 
/* ctrl vport and ctrl queues. */
struct cpfl_vport ctrl_vport;
@@ -239,5 +243,82 @@ int cpfl_alloc_dma_mem_batch(struct idpf_dma_mem 
*orig_dma, struct idpf_dma_mem
((struct cpfl_repr *)((dev)->data->dev_private))
 #define CPFL_DEV_TO_ITF(dev)   \
((struct cpfl_itf *)((dev)->data->dev_private))
+#define CPFL_INVALID_HW_ID  UINT16_MAX
 
+static inline uint16_t
+cpfl_get_port_id(struct cpfl_itf *itf)
+{
+   if (!itf)
+   return CPFL_INVALID_HW_ID;
+
+   if (itf->type == CPFL_ITF_TYPE_VPORT) {
+   struct cpfl_vport *vport = (void *)itf;
+
+   return vport->base.devarg_id;
+   }
+
+   return CPFL_INVALID_HW_ID;
+}
+
+static inline uint16_t
+cpfl_get_vsi_id(struct cpfl_itf *itf)
+{
+   struct cpfl_adapter_ext *adapter = itf->adapter;
+   struct cpfl_vport_info *info;
+   uint32_t vport_id;
+   int ret;
+   struct cpfl_vport_id vport_identity;
+
+   if (!itf)
+   return CPFL_INVALID_HW_ID;
+
+   if (itf->type == CPFL_ITF_TYPE_REPRESENTOR) {
+   struct cpfl_repr *repr = (void *)itf;
+
+   return repr->vport_info->vport_info.vsi_id;
+   } else if (itf->type == CPFL_ITF_TYPE_VPORT) {
+   vport_id = ((struct cpfl_vport *)itf)->base.vport_id;
+   vport_identity.func_type = CPCHNL2_FUNC_TYPE_PF;
+   /* host: HOST0_CPF_ID, acc: ACC_CPF_ID */
+   vport_identity.pf_id = ACC_CPF_ID;
+   vport_identity.vf_id = 0;
+   vport_identity.vport_id = vport_id;
+
+   ret = rte_hash_lookup_data(adapter->vport_map_hash,
+  &vport_identity, (void **)&info);
+   if (ret < 0) {
+   PMD_DRV_LOG(ERR, "vport id not exist");
+   goto err;
+   }
+
+   return info->vport_info.vsi_id;
+   }
+
+err:
+   return CPFL_INVALID_HW_ID;
+}
+
+static inline struct cpfl_itf *
+cpfl_get_itf_by_port_id(uint16_t port_id)
+{
+   struct rte_eth_dev *dev;
+
+   if (port_id >= RTE_MAX_ETHPORTS) {
+   PMD_DRV_LOG(ERR, "port_id should be < %d.", RTE_MAX_ETHPORTS);
+   return NULL;
+   }
+
+   dev = &rte_eth_devices[port_id];
+   if (dev->state == RTE_ETH_DEV_UNUSED) {
+   PMD_DRV_LOG(ERR, "eth_dev[%d] is unused.", port_id);
+   return NULL;
+   }
+
+   if (!dev->data) {
+   PMD_DRV_LOG(ERR, "eth_dev[%d] data not be allocated.", port_id);
+   return NULL;
+   }
+
+   return CPFL_DEV_TO_ITF(dev);
+}
 #endif /* _CPFL_ETHDEV_H_ */
diff --git a/drivers/net/cpfl/cpfl_flow_engine_fxp.c 
b/drivers/net/cpfl/cpfl_flow_engine_fxp.c
new file mode 100644
index 00..e0c08a77c3
--- /dev/null
+++ b/drivers/net/cpfl/cpfl_flow_engine_fxp.c
@@ -0,0 +1,583 @@
+/* SPDX-License-Identifier: BSD-3-Clause
+ * Copyright(c) 2022 Intel Corporation
+ */
+
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include 
+#include "cpfl_rules.h"
+#include "cpfl_logs.h"
+#include "cpfl_ethdev.h"
+#include "cpfl_flow.h"
+#include "cpfl_fxp_rule.h"
+#include "cpfl_flow_parser.h"
+
+#define CPFL_COOKIE_DEF0x1000
+#define CPFL_MOD_COOKIE_DEF0x1237561
+#define CPFL_PREC_DEF  1
+#define CPFL_PREC_SET  5
+#define CPFL_TYPE_ID   3
+#define CPFL_OFFSET0x0a
+#define CPFL_HOST_ID_DEF   0
+#define CPFL_PF_NUM_DEF0
+#define CPFL_PORT_NUM_DEF  0
+#define CPFL_RESP_REQ_DEF  2
+#define CPFL_PIN_TO_CACHE_DEF  0
+#define CPFL_CLEAR_MIRROR_1ST_STATE_DEF0
+#define CPFL_FIXED_FETCH_DEF   0
+#define CPFL_PTI_DEF   0
+#define CPFL_MOD_OBJ_SIZE_DEF  0
+#define CPFL_PIN_MOD_CONTENT_DEF   0
+
+#define CPFL_MAX_MOD_CONTENT_INDEX 256
+#define

[PATCH v2 8/8] net/cpfl: add flow support for representor

[Yuying Zhang]

Add flow support for representor, so representor can
create, destroy, validate and flush rules.

Signed-off-by: Yuying Zhang 
---
 doc/guides/nics/cpfl.rst| 13 +++
 doc/guides/rel_notes/release_23_11.rst  |  1 +
 drivers/net/cpfl/cpfl_flow_engine_fxp.c | 24 ++--
 drivers/net/cpfl/cpfl_representor.c | 29 +
 4 files changed, 65 insertions(+), 2 deletions(-)

diff --git a/doc/guides/nics/cpfl.rst b/doc/guides/nics/cpfl.rst
index 7032dd1a1a..e2fe5430ed 100644
--- a/doc/guides/nics/cpfl.rst
+++ b/doc/guides/nics/cpfl.rst
@@ -196,3 +196,16 @@ low level hardware resources defined in a DDP package file.
.. code-block:: console
 
dpdk-testpmd -c 0x3 -n 4 -a 
:af:00.6,vport=[0],flow_parser="refpkg.json" -- -i
+
+#. Create one flow to forward ETH-IPV4-TCP from I/O port to a local(CPF's) 
vport::
+
+   .. code-block:: console
+
+   flow create 0 ingress group 1 pattern eth dst is 00:01:00:00:03:14 / ipv4 
src is 192.168.0.1 \
+   dst is 192.168.0.2 / tcp / end actions port_representor port_id 0 / end
+
+#. Send the packet, and it should be displayed on PMD::
+
+   .. code-block:: console
+
+   
sendp(Ether(dst='00:01:00:00:03:14')/IP(src='192.168.0.1',dst='192.168.0.2')/TCP(),iface="enp24s0f0")
diff --git a/doc/guides/rel_notes/release_23_11.rst 
b/doc/guides/rel_notes/release_23_11.rst
index 688bee4d6d..eded3ecc84 100644
--- a/doc/guides/rel_notes/release_23_11.rst
+++ b/doc/guides/rel_notes/release_23_11.rst
@@ -58,6 +58,7 @@ New Features
 * **Updated Intel cpfl driver.**
 
   * Added support for port representor.
+  * Added support for rte_flow.
 
 Removed Items
 -
diff --git a/drivers/net/cpfl/cpfl_flow_engine_fxp.c 
b/drivers/net/cpfl/cpfl_flow_engine_fxp.c
index e0c08a77c3..f40402a912 100644
--- a/drivers/net/cpfl/cpfl_flow_engine_fxp.c
+++ b/drivers/net/cpfl/cpfl_flow_engine_fxp.c
@@ -73,6 +73,7 @@ cpfl_fxp_create(struct rte_eth_dev *dev,
struct cpfl_adapter_ext *ad = itf->adapter;
struct cpfl_rule_info_meta *rim = meta;
struct cpfl_vport *vport;
+   struct cpfl_repr *repr;
 
if (!rim)
return ret;
@@ -83,6 +84,10 @@ cpfl_fxp_create(struct rte_eth_dev *dev,
 * Even index is tx queue and odd index is rx queue.
 */
cpq_id = vport->base.devarg_id * 2;
+   } else if (itf->type == CPFL_ITF_TYPE_REPRESENTOR) {
+   repr = (struct cpfl_repr *)itf;
+   cpq_id = ((repr->repr_id.pf_id  + repr->repr_id.vf_id) &
+ (CPFL_TX_CFGQ_NUM - 1)) * 2;
} else {
rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_HANDLE, 
NULL,
   "fail to find correct control queue");
@@ -122,6 +127,7 @@ cpfl_fxp_destroy(struct rte_eth_dev *dev,
struct cpfl_rule_info_meta *rim;
uint32_t i;
struct cpfl_vport *vport;
+   struct cpfl_repr *repr;
 
rim = flow->rule;
if (!rim) {
@@ -135,6 +141,10 @@ cpfl_fxp_destroy(struct rte_eth_dev *dev,
if (itf->type == CPFL_ITF_TYPE_VPORT) {
vport = (struct cpfl_vport *)itf;
cpq_id = vport->base.devarg_id * 2;
+   } else if (itf->type == CPFL_ITF_TYPE_REPRESENTOR) {
+   repr = (struct cpfl_repr *)itf;
+   cpq_id = ((repr->repr_id.pf_id  + repr->repr_id.vf_id) &
+ (CPFL_TX_CFGQ_NUM - 1)) * 2;
} else {
rte_flow_error_set(error, EINVAL, RTE_FLOW_ERROR_TYPE_HANDLE, 
NULL,
   "fail to find correct control queue");
@@ -257,6 +267,7 @@ cpfl_fxp_parse_action(struct cpfl_itf *itf,
int queue_id = -1;
bool fwd_vsi = false;
bool fwd_q = false;
+   bool is_vsi;
uint32_t i;
struct cpfl_rule_info *rinfo = &rim->rules[index];
union cpfl_action_set *act_set = (void *)rinfo->act_bytes;
@@ -267,6 +278,7 @@ cpfl_fxp_parse_action(struct cpfl_itf *itf,
action_type = action->type;
switch (action_type) {
case RTE_FLOW_ACTION_TYPE_PORT_REPRESENTOR:
+   case RTE_FLOW_ACTION_TYPE_REPRESENTED_PORT:
if (!fwd_vsi)
fwd_vsi = true;
else
@@ -285,12 +297,20 @@ cpfl_fxp_parse_action(struct cpfl_itf *itf,
queue_id = CPFL_INVALID_QUEUE_ID;
}
 
-   dev_id = cpfl_get_vsi_id(dst_itf);
+   is_vsi = (action_type == 
RTE_FLOW_ACTION_TYPE_PORT_REPRESENTOR ||
+ dst_itf->type == CPFL_ITF_TYPE_REPRESENTOR);
+   if (is_vsi)
+   dev_id = cpfl_get_vsi_id(dst_itf);
+   else
+   dev_id = cpfl_get_port_id(dst_itf);
 
if (dev_id == CPFL_INVALID_HW_ID)
   

Re: [PATCH v8 0/3] Split logging functionality out of EAL

[Thomas Monjalon]

11/08/2023 14:59, Bruce Richardson:
> On Fri, Aug 11, 2023 at 02:46:13PM +0200, David Marchand wrote:
> > On Wed, Aug 9, 2023 at 3:36 PM Bruce Richardson
> >  wrote:
> > >
> > > There is a general desire to reduce the size and scope of EAL. To this
> > > end, this patchset makes a (very) small step in that direction by taking
> > > the logging functionality out of EAL and putting it into its own library
> > > that can be built and maintained separately.
> > >
> > > As with the first RFC for this, the main obstacle is the "fnmatch"
> > > function which is needed by both EAL and the new log function when
> > > building on windows. While the function cannot stay in EAL - or we would
> > > have a circular dependency, moving it to a new library or just putting
> > > it in the log library have the disadvantages that it then "leaks" into
> > > the public namespace without an rte_prefix, which could cause issues.
> > > Since only a single function is involved, subsequent versions take a
> > > different approach to v1, and just moves the offending function to be a
> > > static function in a header file. This allows use by multiple libs
> > > without conflicting names or making it public.
> > >
> > > The other complication, as explained in v1 RFC was that of multiple
> > > implementations for different OS's. This is solved here in the same
> > > way as v1, by including the OS in the name and having meson pick the
> > > correct file for each build. Since only one file is involved, there
> > > seemed little need for replicating EAL's separate subdirectories
> > > per-OS.
> > 
> > Series applied, thanks Bruce for this first step.
> > 
> > As mentionned during the maintainers weekly call yesterday, this is
> > only a first "easy" step but, thinking of next steps, more splitting
> > may not be that easy.
> 
> I took a look after doing this patchset to try and find more easy to
> extract parts, but did not find many. The EAL is pretty intertwined now.
> As I look at it, there are really two ways to try and dis-entangle it - 
> bottoms-up or top-down. This patchset is an example of the former, taking a
> logical set of related APIs and pulling them out into a separate library
> that EAL can depend upon. There may be some other API-sets like this we can
> pull out, but in my search I didn't find any.

A small thing to easily separate is rte_bitmap.

> The other tops-down approach may be worth looking at too. We can try and
> take the top-level EAL init function and separate it out into a separate
> initialization library (which may be called "EAL" with the rest being
> called something else). I have not tried this approach to see how it goes,
> but pulling out the init may allow further dis-entangling of other parts of
> EAL.

I agree we should start with separating the init logic.
I would call it rte_init. We may need rte_opts for command line parsing.
I think the name EAL should be kept for the low-level functions,
abstracting differences between OS, CPU and toolchains.

Next steps would be to try to separate memory and CPU management
in 2 different libraries.
We'll need to decide whether rte_service is part of the CPU management.
Same question for multiprocess communication rte_mp and rte_keepalive.
I suppose we can keep IRQ and threading in EAL.
VFIO may be managed in a separate library perhaps.

Once we do that, the low-level libraries like log, kvargs and telemetry
can depend on EAL, being the very low-level common denominator.
The trace subsystem can probably become a separate library as well.

In a later step, we can think about bus and device management.

And the ideal would be to extract tailq, once all logic is out of EAL.

How does it sound?

Re: [PATCH v4 01/32] net/sssnic: add build and doc infrastructure

[David Marchand]

On Fri, Sep 1, 2023 at 11:35 AM  wrote:
>
> From: Renyong Wan 
>
> Adding minimum PMD code, doc and build infrastructure for sssnic.
>
> Signed-off-by: Steven Song 
> Signed-off-by: Renyong Wan 
> ---
> v2:
> * Fixed 'Title underline too short' in doc/guides/nics/sssnic.rst.
> * Removed error.h from including files.
> ---
>  .mailmap|  5 +-
>  MAINTAINERS |  8 
>  doc/guides/nics/features/sssnic.ini |  9 
>  doc/guides/nics/index.rst   |  1 +
>  doc/guides/nics/sssnic.rst  | 73 +
>  drivers/net/meson.build |  1 +
>  drivers/net/sssnic/meson.build  | 18 +++
>  drivers/net/sssnic/sssnic_ethdev.c  | 28 +++
>  8 files changed, 140 insertions(+), 3 deletions(-)
>  create mode 100644 doc/guides/nics/features/sssnic.ini
>  create mode 100644 doc/guides/nics/sssnic.rst
>  create mode 100644 drivers/net/sssnic/meson.build
>  create mode 100644 drivers/net/sssnic/sssnic_ethdev.c
>
> diff --git a/.mailmap b/.mailmap
> index 864d33ee46..6fa73d3b79 100644
> --- a/.mailmap
> +++ b/.mailmap
> @@ -151,7 +151,6 @@ Bao-Long Tran 
>  Bar Neuman 
>  Barak Enat 
>  Barry Cao 
> -Bartosz Staszewski 
>  Baruch Siach 
>  Bassam Zaid AlKilani 
>  Beilei Xing 

This looks like some rebase damage, or at least unwanted changes.
Please fix.

> @@ -496,7 +495,6 @@ Helin Zhang 
>  Hemant Agrawal  
>  Heng Ding 
>  Hengjian Zhang 
> -Heng Jiang 
>  Heng Wang 
>  Henning Schild 
>  Henry Cai 
> @@ -630,7 +628,6 @@ Jie Liu 
>  Jie Pan 
>  Jie Wang 
>  Jie Zhou  
> -Jieqiang Wang 
>  Jijiang Liu 
>  Jilei Chen 
>  Jim Harris 
> @@ -1156,6 +1153,7 @@ Rebecca Troy 
>  Remi Pommarel 
>  Remy Horton 
>  Renata Saiakhova 
> +Renyong Wan 
>  Reshma Pattan  
>  Ricardo Roldan 
>  Ricardo Salveti 
> @@ -1329,6 +1327,7 @@ Stephen Hurd  
> 
>  Steve Capper 
>  Steven Lariau 
>  Steven Luong 
> +Steven Song 
>  Steven Webster 
>  Steven Zou 
>  Steve Rempe 
> diff --git a/MAINTAINERS b/MAINTAINERS
> index a926155f26..1e57d29aa3 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -633,6 +633,13 @@ F: drivers/net/af_xdp/
>  F: doc/guides/nics/af_xdp.rst
>  F: doc/guides/nics/features/af_xdp.ini
>
> +3SNIC sssnic
> +M: Renyong Wan 
> +M: Steven Song 
> +F: driver/net/sssnic/

This path is incorrect.
*drivers*

> +F: doc/guides/nics/sssnic.rst
> +F: doc/guides/nics/features/sssnic.ini
> +
>  Amazon ENA
>  M: Michal Krawczyk 
>  M: Shai Brandes 
> @@ -1793,6 +1800,7 @@ F: doc/guides/tools/img/eventdev_*
>  F: app/test/test_event_ring.c
>
>  Procinfo tool
> +M: Maryam Tahhan 
>  M: Reshma Pattan 
>  F: app/proc-info/
>  F: doc/guides/tools/proc_info.rst


-- 
David Marchand

[PATCH] maintainers: assign meson.build for drivers with subtrees

[David Marchand]

When a new (for example, net) driver gets submitted, the CI scripts
can't determine which subtree (here, dpdk-next-net) this series is
associated to because the MAINTAINERS update for the new driver is
part of the series.

Put meson.build of the drivers classes under the responsibility of
subtree maintainers.

Signed-off-by: David Marchand 
---
 MAINTAINERS | 10 +++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index a926155f26..598edadc42 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -611,6 +611,7 @@ Networking Drivers
 --
 M: Ferruh Yigit 
 T: git://dpdk.org/next/dpdk-next-net
+F: drivers/net/meson.build
 F: doc/guides/nics/features/default.ini
 
 Link bonding
@@ -1057,6 +1058,7 @@ F: doc/guides/nics/features/memif.ini
 Crypto Drivers
 --
 T: git://dpdk.org/next/dpdk-next-crypto
+F: drivers/crypto/meson.build
 F: doc/guides/cryptodevs/features/default.ini
 
 AMD CCP Crypto
@@ -1193,6 +1195,7 @@ F: doc/guides/cryptodevs/features/virtio.ini
 Compression Drivers
 ---
 T: git://dpdk.org/next/dpdk-next-crypto
+F: drivers/compress/meson.build
 
 Cavium OCTEON TX zipvf
 M: Ashish Gupta 
@@ -1290,6 +1293,7 @@ F: doc/guides/mldevs/cnxk.rst
 vDPA Drivers
 
 T: git://dpdk.org/next/dpdk-next-virtio
+F: drivers/vdpa/meson.build
 
 Intel ifc
 M: Xiao Wang 
@@ -1315,6 +1319,7 @@ Eventdev Drivers
 
 M: Jerin Jacob 
 T: git://dpdk.org/next/dpdk-next-eventdev
+F: drivers/event/meson.build
 
 Cavium OCTEON TX ssovf
 M: Jerin Jacob 
@@ -1369,10 +1374,11 @@ F: doc/guides/eventdevs/opdl.rst
 
 Baseband Drivers
 
+T: git://dpdk.org/next/dpdk-next-baseband
+F: drivers/baseband/meson.build
 
 Intel baseband
 M: Nicolas Chautru 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/turbo_sw/
 F: doc/guides/bbdevs/turbo_sw.rst
 F: doc/guides/bbdevs/features/turbo_sw.ini
@@ -1391,7 +1397,6 @@ F: doc/guides/bbdevs/features/vrb1.ini
 
 Null baseband
 M: Nicolas Chautru 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/null/
 F: doc/guides/bbdevs/null.rst
 F: doc/guides/bbdevs/features/null.ini
@@ -1399,7 +1404,6 @@ F: doc/guides/bbdevs/features/null.ini
 NXP LA12xx
 M: Gagandeep Singh 
 M: Hemant Agrawal 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/la12xx/
 F: doc/guides/bbdevs/la12xx.rst
 F: doc/guides/bbdevs/features/la12xx.ini
-- 
2.41.0

Re: [PATCH v1] maintainers: update maintainer for DLB Driver

[Thomas Monjalon]

31/08/2023 09:51, Bruce Richardson:
> On Mon, Jun 26, 2023 at 03:04:11AM -0500, Abdullah Sevincer wrote:
> > Abdullah Sevincer is now maintainer for DLB Driver.
> > 
> > Signed-off-by: Abdullah Sevincer 
> > ---
> >  MAINTAINERS | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/MAINTAINERS b/MAINTAINERS
> > index fda0b55513..58ef06dd74 100644
> > --- a/MAINTAINERS
> > +++ b/MAINTAINERS
> > @@ -1334,7 +1334,7 @@ M: Pavan Nikhilesh 
> >  F: drivers/event/octeontx/timvf_*
> >  
> >  Intel DLB2
> > -M: Timothy McDaniel 
> > +M: Abdullah Sevincer 
> >  F: drivers/event/dlb2/
> >  F: doc/guides/eventdevs/dlb2.rst
> >  
> Acked-by: Bruce Richardson 

Timothy is not replying, let's assume he already abandoned.

Applied, thanks.

Re: [PATCH] app/testpmd: add support for MPLS TTL

[Singh, Aman Deep]

On 8/22/2023 7:47 PM, Michael Baum wrote:

Add support for matching "ttl" field in MPLS item.

Signed-off-by: Michael Baum 


Acked-by: Aman Singh 


---
  app/test-pmd/cmdline_flow.c | 9 +
  1 file changed, 9 insertions(+)

diff --git a/app/test-pmd/cmdline_flow.c b/app/test-pmd/cmdline_flow.c
index 94827bcc4a..95c0a19beb 100644
--- a/app/test-pmd/cmdline_flow.c
+++ b/app/test-pmd/cmdline_flow.c
@@ -365,6 +365,7 @@ enum index {
ITEM_MPLS_LABEL,
ITEM_MPLS_TC,
ITEM_MPLS_S,
+   ITEM_MPLS_TTL,
ITEM_GRE,
ITEM_GRE_PROTO,
ITEM_GRE_C_RSVD0_VER,
@@ -1712,6 +1713,7 @@ static const enum index item_mpls[] = {
ITEM_MPLS_LABEL,
ITEM_MPLS_TC,
ITEM_MPLS_S,
+   ITEM_MPLS_TTL,
ITEM_NEXT,
ZERO,
  };
@@ -4650,6 +4652,13 @@ static const struct token token_list[] = {
  label_tc_s,
  "\x00\x00\x01")),
},
+   [ITEM_MPLS_TTL] = {
+   .name = "ttl",
+   .help = "MPLS Time-to-Live",
+   .next = NEXT(item_mpls, NEXT_ENTRY(COMMON_UNSIGNED),
+item_param),
+   .args = ARGS(ARGS_ENTRY_HTON(struct rte_flow_item_mpls, ttl)),
+   },
[ITEM_GRE] = {
.name = "gre",
.help = "match GRE header",

RE: [PATCH v11 2/4] net/i40e: implement mbufs recycle mode

[Feifei Wang]

> -Original Message-
> From: Konstantin Ananyev 
> Sent: Friday, September 1, 2023 1:25 AM
> To: Feifei Wang ; Konstantin Ananyev
> 
> Cc: dev@dpdk.org; nd ; Honnappa Nagarahalli
> ; Ruifeng Wang
> ; Yuying Zhang ; Beilei
> Xing ; nd 
> Subject: RE: [PATCH v11 2/4] net/i40e: implement mbufs recycle mode
> 
> 
> 
> > >
> > > Define specific function implementation for i40e driver.
> > > Currently, mbufs recycle mode can support 128bit vector path and avx2
> path.
> > > And can be enabled both in fast free and no fast free mode.
> > >
> > > Suggested-by: Honnappa Nagarahalli 
> > > Signed-off-by: Feifei Wang 
> > > Reviewed-by: Ruifeng Wang 
> > > Reviewed-by: Honnappa Nagarahalli 
> > > ---
> > >  drivers/net/i40e/i40e_ethdev.c|   1 +
> > >  drivers/net/i40e/i40e_ethdev.h|   2 +
> > >  .../net/i40e/i40e_recycle_mbufs_vec_common.c  | 147
> > > ++
> > >  drivers/net/i40e/i40e_rxtx.c  |  32 
> > >  drivers/net/i40e/i40e_rxtx.h  |   4 +
> > >  drivers/net/i40e/meson.build  |   1 +
> > >  6 files changed, 187 insertions(+)
> > >  create mode 100644 drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > >
> > > diff --git a/drivers/net/i40e/i40e_ethdev.c
> > > b/drivers/net/i40e/i40e_ethdev.c index 8271bbb394..50ba9aac94 100644
> > > --- a/drivers/net/i40e/i40e_ethdev.c
> > > +++ b/drivers/net/i40e/i40e_ethdev.c
> > > @@ -496,6 +496,7 @@ static const struct eth_dev_ops i40e_eth_dev_ops
> = {
> > >   .flow_ops_get = i40e_dev_flow_ops_get,
> > >   .rxq_info_get = i40e_rxq_info_get,
> > >   .txq_info_get = i40e_txq_info_get,
> > > + .recycle_rxq_info_get = i40e_recycle_rxq_info_get,
> > >   .rx_burst_mode_get= i40e_rx_burst_mode_get,
> > >   .tx_burst_mode_get= i40e_tx_burst_mode_get,
> > >   .timesync_enable  = i40e_timesync_enable,
> > > diff --git a/drivers/net/i40e/i40e_ethdev.h
> > > b/drivers/net/i40e/i40e_ethdev.h index 6f65d5e0ac..af758798e1 100644
> > > --- a/drivers/net/i40e/i40e_ethdev.h
> > > +++ b/drivers/net/i40e/i40e_ethdev.h
> > > @@ -1355,6 +1355,8 @@ void i40e_rxq_info_get(struct rte_eth_dev
> > > *dev, uint16_t queue_id,
> > >   struct rte_eth_rxq_info *qinfo);
> > >  void i40e_txq_info_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > >   struct rte_eth_txq_info *qinfo);
> > > +void i40e_recycle_rxq_info_get(struct rte_eth_dev *dev, uint16_t
> queue_id,
> > > + struct rte_eth_recycle_rxq_info *recycle_rxq_info);
> > >  int i40e_rx_burst_mode_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > >  struct rte_eth_burst_mode *mode);  int
> > > i40e_tx_burst_mode_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > > diff -- git a/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > b/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > new file mode 100644
> > > index 00..5663ecccde
> > > --- /dev/null
> > > +++ b/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > @@ -0,0 +1,147 @@
> > > +/* SPDX-License-Identifier: BSD-3-Clause
> > > + * Copyright (c) 2023 Arm Limited.
> > > + */
> > > +
> > > +#include 
> > > +#include 
> > > +
> > > +#include "base/i40e_prototype.h"
> > > +#include "base/i40e_type.h"
> > > +#include "i40e_ethdev.h"
> > > +#include "i40e_rxtx.h"
> > > +
> > > +#pragma GCC diagnostic ignored "-Wcast-qual"
> > > +
> > > +void
> > > +i40e_recycle_rx_descriptors_refill_vec(void *rx_queue, uint16_t
> > > +nb_mbufs) {
> > > + struct i40e_rx_queue *rxq = rx_queue;
> > > + struct i40e_rx_entry *rxep;
> > > + volatile union i40e_rx_desc *rxdp;
> > > + uint16_t rx_id;
> > > + uint64_t paddr;
> > > + uint64_t dma_addr;
> > > + uint16_t i;
> > > +
> > > + rxdp = rxq->rx_ring + rxq->rxrearm_start;
> > > + rxep = &rxq->sw_ring[rxq->rxrearm_start];
> > > +
> > > + for (i = 0; i < nb_mbufs; i++) {
> > > + /* Initialize rxdp descs. */
> > > + paddr = (rxep[i].mbuf)->buf_iova +
> > > RTE_PKTMBUF_HEADROOM;
> > > + dma_addr = rte_cpu_to_le_64(paddr);
> > > + /* flush desc with pa dma_addr */
> > > + rxdp[i].read.hdr_addr = 0;
> > > + rxdp[i].read.pkt_addr = dma_addr;
> > > + }
> > > +
> > > + /* Update the descriptor initializer index */
> > > + rxq->rxrearm_start += nb_mbufs;
> > > + rx_id = rxq->rxrearm_start - 1;
> > > +
> > > + if (unlikely(rxq->rxrearm_start >= rxq->nb_rx_desc)) {
> > > + rxq->rxrearm_start = 0;
> > > + rx_id = rxq->nb_rx_desc - 1;
> > > + }
> > > +
> > > + rxq->rxrearm_nb -= nb_mbufs;
> > > +
> > > + rte_io_wmb();
> > > + /* Update the tail pointer on the NIC */
> > > + I40E_PCI_REG_WRITE_RELAXED(rxq->qrx_tail, rx_id); }
> > > +
> > > +uint16_t
> > > +i40e_recycle_tx_mbufs_reuse_vec(void *tx_queue,
> > > + struct rte_eth_recycle_rxq_info *recycle_rxq_info) {
> > > + struct i40e_tx_queue *txq = tx_queue;
> > > + struct i40e_tx_entry *txep;
> > > + struct rte_mbuf **rxep;
> > > + int 

Re: [RFC] cache guard

[Thomas Monjalon]

27/08/2023 10:34, Morten Brørup:
> +CC Honnappa and Konstantin, Ring lib maintainers
> +CC Mattias, PRNG lib maintainer
> 
> > From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> > Sent: Friday, 25 August 2023 11.24
> > 
> > On Fri, Aug 25, 2023 at 11:06:01AM +0200, Morten Brørup wrote:
> > > +CC mempool maintainers
> > >
> > > > From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> > > > Sent: Friday, 25 August 2023 10.23
> > > >
> > > > On Fri, Aug 25, 2023 at 08:45:12AM +0200, Morten Brørup wrote:
> > > > > Bruce,
> > > > >
> > > > > With this patch [1], it is noted that the ring producer and
> > consumer data
> > > > should not be on adjacent cache lines, for performance reasons.
> > > > >
> > > > > [1]:
> > > >
> > https://git.dpdk.org/dpdk/commit/lib/librte_ring/rte_ring.h?id=d9f0d3a1f
> > fd4b66
> > > > e75485cc8b63b9aedfbdfe8b0
> > > > >
> > > > > (It's obvious that they cannot share the same cache line, because
> > they are
> > > > accessed by two different threads.)
> > > > >
> > > > > Intuitively, I would think that having them on different cache
> > lines would
> > > > suffice. Why does having an empty cache line between them make a
> > difference?
> > > > >
> > > > > And does it need to be an empty cache line? Or does it suffice
> > having the
> > > > second structure start at two cache lines after the start of the
> > first
> > > > structure (e.g. if the size of the first structure is two cache
> > lines)?
> > > > >
> > > > > I'm asking because the same principle might apply to other code
> > too.
> > > > >
> > > > Hi Morten,
> > > >
> > > > this was something we discovered when working on the distributor
> > library.
> > > > If we have cachelines per core where there is heavy access, having
> > some
> > > > cachelines as a gap between the content cachelines can help
> > performance. We
> > > > believe this helps due to avoiding issues with the HW prefetchers
> > (e.g.
> > > > adjacent cacheline prefetcher) bringing in the second cacheline
> > > > speculatively when an operation is done on the first line.
> > >
> > > I guessed that it had something to do with speculative prefetching,
> > but wasn't sure. Good to get confirmation, and that it has a measureable
> > effect somewhere. Very interesting!
> > >
> > > NB: More comments in the ring lib about stuff like this would be nice.
> > >
> > > So, for the mempool lib, what do you think about applying the same
> > technique to the rte_mempool_debug_stats structure (which is an array
> > indexed per lcore)... Two adjacent lcores heavily accessing their local
> > mempool caches seems likely to me. But how heavy does the access need to
> > be for this technique to be relevant?
> > >
> > 
> > No idea how heavy the accesses need to be for this to have a noticable
> > effect. For things like debug stats, I wonder how worthwhile making such
> > a
> > change would be, but then again, any change would have very low impact
> > too
> > in that case.
> 
> I just tried adding padding to some of the hot structures in our own 
> application, and observed a significant performance improvement for those.
> 
> So I think this technique should have higher visibility in DPDK by adding a 
> new cache macro to rte_common.h:

+1 to make more visibility in doc and adding a macro, good idea!

RE: [PATCH v8 0/3] Split logging functionality out of EAL

[Morten Brørup]

> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Friday, 1 September 2023 13.25
> 
> 11/08/2023 14:59, Bruce Richardson:
> > On Fri, Aug 11, 2023 at 02:46:13PM +0200, David Marchand wrote:
> > > On Wed, Aug 9, 2023 at 3:36 PM Bruce Richardson
> > >  wrote:
> > > >
> > > > There is a general desire to reduce the size and scope of EAL. To this
> > > > end, this patchset makes a (very) small step in that direction by taking
> > > > the logging functionality out of EAL and putting it into its own library
> > > > that can be built and maintained separately.
> > > >
> > > > As with the first RFC for this, the main obstacle is the "fnmatch"
> > > > function which is needed by both EAL and the new log function when
> > > > building on windows. While the function cannot stay in EAL - or we would
> > > > have a circular dependency, moving it to a new library or just putting
> > > > it in the log library have the disadvantages that it then "leaks" into
> > > > the public namespace without an rte_prefix, which could cause issues.
> > > > Since only a single function is involved, subsequent versions take a
> > > > different approach to v1, and just moves the offending function to be a
> > > > static function in a header file. This allows use by multiple libs
> > > > without conflicting names or making it public.
> > > >
> > > > The other complication, as explained in v1 RFC was that of multiple
> > > > implementations for different OS's. This is solved here in the same
> > > > way as v1, by including the OS in the name and having meson pick the
> > > > correct file for each build. Since only one file is involved, there
> > > > seemed little need for replicating EAL's separate subdirectories
> > > > per-OS.
> > >
> > > Series applied, thanks Bruce for this first step.
> > >
> > > As mentionned during the maintainers weekly call yesterday, this is
> > > only a first "easy" step but, thinking of next steps, more splitting
> > > may not be that easy.
> >
> > I took a look after doing this patchset to try and find more easy to
> > extract parts, but did not find many. The EAL is pretty intertwined now.
> > As I look at it, there are really two ways to try and dis-entangle it -
> > bottoms-up or top-down. This patchset is an example of the former, taking a
> > logical set of related APIs and pulling them out into a separate library
> > that EAL can depend upon. There may be some other API-sets like this we can
> > pull out, but in my search I didn't find any.
> 
> A small thing to easily separate is rte_bitmap.
> 
> > The other tops-down approach may be worth looking at too. We can try and
> > take the top-level EAL init function and separate it out into a separate
> > initialization library (which may be called "EAL" with the rest being
> > called something else). I have not tried this approach to see how it goes,
> > but pulling out the init may allow further dis-entangling of other parts of
> > EAL.
> 
> I agree we should start with separating the init logic.
> I would call it rte_init. We may need rte_opts for command line parsing.
> I think the name EAL should be kept for the low-level functions,
> abstracting differences between OS, CPU and toolchains.
> 
> Next steps would be to try to separate memory and CPU management
> in 2 different libraries.
> We'll need to decide whether rte_service is part of the CPU management.
> Same question for multiprocess communication rte_mp and rte_keepalive.
> I suppose we can keep IRQ and threading in EAL.
> VFIO may be managed in a separate library perhaps.
> 
> Once we do that, the low-level libraries like log, kvargs and telemetry
> can depend on EAL, being the very low-level common denominator.
> The trace subsystem can probably become a separate library as well.
> 
> In a later step, we can think about bus and device management.
> 
> And the ideal would be to extract tailq, once all logic is out of EAL.

Yes, tailq is a good example of a utility library, which can stand on its own 
(i.e. it could be used for non-DPDK applications too).

> 
> How does it sound?
> 

Not really a reply to Thomas, just thinking out loud here...

From a high level perspective, thinking about two dimensions, layers and silos, 
(the boxes a marketing person would draw on a PowerPoint slide to describe what 
DPDK comprises of) might also help defining the borders.

For vertical separation:

There are the architecture specific definitions, e.g. header files created when 
selecting an architecture. There should be no functions in this box, only 
descriptions/definitions.

Then there are the utility libraries/functions, e.g. bit manipulation and log2 
functions, the command line parser, etc.. Some of these depend on the 
architecture definitions. But none of these should depend on the DPDK runtime 
itself - i.e. it should be possible to use this for writing a non-dataplane 
application (in theory).

Then there is the DPDK core itself. The core of this should be kept at an 
absolute minimum, mak

Re: [PATCH] maintainers: assign meson.build for drivers with subtrees

[Thomas Monjalon]

01/09/2023 14:03, David Marchand:
> When a new (for example, net) driver gets submitted, the CI scripts
> can't determine which subtree (here, dpdk-next-net) this series is
> associated to because the MAINTAINERS update for the new driver is
> part of the series.
> 
> Put meson.build of the drivers classes under the responsibility of
> subtree maintainers.
[...]
> @@ -1315,6 +1319,7 @@ Eventdev Drivers
>  
>  M: Jerin Jacob 
>  T: git://dpdk.org/next/dpdk-next-eventdev
> +F: drivers/event/meson.build

While at it, can we remove Jerin's name here?
He's already referenced for the whole git tree at the beginning of the file.
Other device classes don't add a name for drivers, it is redundant.

Re: [PATCH] maintainers: assign meson.build for drivers with subtrees

[David Marchand]

On Fri, Sep 1, 2023 at 2:28 PM Thomas Monjalon  wrote:
>
> 01/09/2023 14:03, David Marchand:
> > When a new (for example, net) driver gets submitted, the CI scripts
> > can't determine which subtree (here, dpdk-next-net) this series is
> > associated to because the MAINTAINERS update for the new driver is
> > part of the series.
> >
> > Put meson.build of the drivers classes under the responsibility of
> > subtree maintainers.
> [...]
> > @@ -1315,6 +1319,7 @@ Eventdev Drivers
> >  
> >  M: Jerin Jacob 
> >  T: git://dpdk.org/next/dpdk-next-eventdev
> > +F: drivers/event/meson.build
>
> While at it, can we remove Jerin's name here?
> He's already referenced for the whole git tree at the beginning of the file.
> Other device classes don't add a name for drivers, it is redundant.

The same is true for Ferruh too.
v2 on the way.


-- 
David Marchand

Re: [PATCH v1] maintainers: take maintainership of AMD axgbe driver

[Thomas Monjalon]

> Chandu is not working on dpdk axgbe driver anymore.
> 
> Signed-off-by: Selwin Sebastian 
> ---
>  AMD axgbe
> -M: Chandubabu Namburu 
> +M: Selwin Sebastian 
> 
> Acked-by: Chandubabu Namburu 

Applied, thanks.

[PATCH v2] maintainers: assign meson.build for drivers with subtrees

[David Marchand]

When a new (for example, net) driver gets submitted, the CI scripts
can't determine which subtree (here, dpdk-next-net) this series is
associated to because the MAINTAINERS update for the new driver is
part of the series.

Put meson.build of the drivers classes under the responsibility of
subtree maintainers.

While at it, cleanup redundant statements:
- Ferruh or Jerin are maintainers of next-net / next-event but this
  information is already present at the top of the MAINTAINERS file,
- the subtree for baseband drivers can be put at the start of the
  baseband section,

Signed-off-by: David Marchand 
---
Changes since v1:
- added some minor cleanup,

---
 MAINTAINERS | 12 +++-
 1 file changed, 7 insertions(+), 5 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index a926155f26..f3e3dda872 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -609,8 +609,8 @@ F: drivers/bus/vmbus/
 
 Networking Drivers
 --
-M: Ferruh Yigit 
 T: git://dpdk.org/next/dpdk-next-net
+F: drivers/net/meson.build
 F: doc/guides/nics/features/default.ini
 
 Link bonding
@@ -1057,6 +1057,7 @@ F: doc/guides/nics/features/memif.ini
 Crypto Drivers
 --
 T: git://dpdk.org/next/dpdk-next-crypto
+F: drivers/crypto/meson.build
 F: doc/guides/cryptodevs/features/default.ini
 
 AMD CCP Crypto
@@ -1193,6 +1194,7 @@ F: doc/guides/cryptodevs/features/virtio.ini
 Compression Drivers
 ---
 T: git://dpdk.org/next/dpdk-next-crypto
+F: drivers/compress/meson.build
 
 Cavium OCTEON TX zipvf
 M: Ashish Gupta 
@@ -1290,6 +1292,7 @@ F: doc/guides/mldevs/cnxk.rst
 vDPA Drivers
 
 T: git://dpdk.org/next/dpdk-next-virtio
+F: drivers/vdpa/meson.build
 
 Intel ifc
 M: Xiao Wang 
@@ -1313,8 +1316,8 @@ F: doc/guides/vdpadevs/features/sfc.ini
 
 Eventdev Drivers
 
-M: Jerin Jacob 
 T: git://dpdk.org/next/dpdk-next-eventdev
+F: drivers/event/meson.build
 
 Cavium OCTEON TX ssovf
 M: Jerin Jacob 
@@ -1369,10 +1372,11 @@ F: doc/guides/eventdevs/opdl.rst
 
 Baseband Drivers
 
+T: git://dpdk.org/next/dpdk-next-baseband
+F: drivers/baseband/meson.build
 
 Intel baseband
 M: Nicolas Chautru 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/turbo_sw/
 F: doc/guides/bbdevs/turbo_sw.rst
 F: doc/guides/bbdevs/features/turbo_sw.ini
@@ -1391,7 +1395,6 @@ F: doc/guides/bbdevs/features/vrb1.ini
 
 Null baseband
 M: Nicolas Chautru 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/null/
 F: doc/guides/bbdevs/null.rst
 F: doc/guides/bbdevs/features/null.ini
@@ -1399,7 +1402,6 @@ F: doc/guides/bbdevs/features/null.ini
 NXP LA12xx
 M: Gagandeep Singh 
 M: Hemant Agrawal 
-T: git://dpdk.org/next/dpdk-next-baseband
 F: drivers/baseband/la12xx/
 F: doc/guides/bbdevs/la12xx.rst
 F: doc/guides/bbdevs/features/la12xx.ini
-- 
2.41.0

Re: [PATCH v2] maintainers: assign meson.build for drivers with subtrees

[Thomas Monjalon]

01/09/2023 14:41, David Marchand:
> When a new (for example, net) driver gets submitted, the CI scripts
> can't determine which subtree (here, dpdk-next-net) this series is
> associated to because the MAINTAINERS update for the new driver is
> part of the series.
> 
> Put meson.build of the drivers classes under the responsibility of
> subtree maintainers.
> 
> While at it, cleanup redundant statements:
> - Ferruh or Jerin are maintainers of next-net / next-event but this
>   information is already present at the top of the MAINTAINERS file,
> - the subtree for baseband drivers can be put at the start of the
>   baseband section,
> 
> Signed-off-by: David Marchand 

Applied, thanks.

quick thread in DLB2

[Thomas Monjalon]

Hello Abdullah,

In the DLB2 code, I see a thread is created for a single operation:
In drivers/event/dlb2/pf/base/dlb2_resource.c
pthread_create(&pthread, NULL, &dlb2_pp_profile_func, &dlb2_thread_data[i]);
and just after:
pthread_join(pthread, NULL);

Can we avoid creating this thread?
I guess no, because it must spawn on a specific CPU.

RE: [PATCH v11 2/4] net/i40e: implement mbufs recycle mode

[Konstantin Ananyev]

> > > >
> > > > Define specific function implementation for i40e driver.
> > > > Currently, mbufs recycle mode can support 128bit vector path and avx2
> > path.
> > > > And can be enabled both in fast free and no fast free mode.
> > > >
> > > > Suggested-by: Honnappa Nagarahalli 
> > > > Signed-off-by: Feifei Wang 
> > > > Reviewed-by: Ruifeng Wang 
> > > > Reviewed-by: Honnappa Nagarahalli 
> > > > ---
> > > >  drivers/net/i40e/i40e_ethdev.c|   1 +
> > > >  drivers/net/i40e/i40e_ethdev.h|   2 +
> > > >  .../net/i40e/i40e_recycle_mbufs_vec_common.c  | 147
> > > > ++
> > > >  drivers/net/i40e/i40e_rxtx.c  |  32 
> > > >  drivers/net/i40e/i40e_rxtx.h  |   4 +
> > > >  drivers/net/i40e/meson.build  |   1 +
> > > >  6 files changed, 187 insertions(+)
> > > >  create mode 100644 drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > >
> > > > diff --git a/drivers/net/i40e/i40e_ethdev.c
> > > > b/drivers/net/i40e/i40e_ethdev.c index 8271bbb394..50ba9aac94 100644
> > > > --- a/drivers/net/i40e/i40e_ethdev.c
> > > > +++ b/drivers/net/i40e/i40e_ethdev.c
> > > > @@ -496,6 +496,7 @@ static const struct eth_dev_ops i40e_eth_dev_ops
> > = {
> > > > .flow_ops_get = i40e_dev_flow_ops_get,
> > > > .rxq_info_get = i40e_rxq_info_get,
> > > > .txq_info_get = i40e_txq_info_get,
> > > > +   .recycle_rxq_info_get = i40e_recycle_rxq_info_get,
> > > > .rx_burst_mode_get= i40e_rx_burst_mode_get,
> > > > .tx_burst_mode_get= i40e_tx_burst_mode_get,
> > > > .timesync_enable  = i40e_timesync_enable,
> > > > diff --git a/drivers/net/i40e/i40e_ethdev.h
> > > > b/drivers/net/i40e/i40e_ethdev.h index 6f65d5e0ac..af758798e1 100644
> > > > --- a/drivers/net/i40e/i40e_ethdev.h
> > > > +++ b/drivers/net/i40e/i40e_ethdev.h
> > > > @@ -1355,6 +1355,8 @@ void i40e_rxq_info_get(struct rte_eth_dev
> > > > *dev, uint16_t queue_id,
> > > > struct rte_eth_rxq_info *qinfo);
> > > >  void i40e_txq_info_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > > > struct rte_eth_txq_info *qinfo);
> > > > +void i40e_recycle_rxq_info_get(struct rte_eth_dev *dev, uint16_t
> > queue_id,
> > > > +   struct rte_eth_recycle_rxq_info *recycle_rxq_info);
> > > >  int i40e_rx_burst_mode_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > > >struct rte_eth_burst_mode *mode);  int
> > > > i40e_tx_burst_mode_get(struct rte_eth_dev *dev, uint16_t queue_id,
> > > > diff -- git a/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > > b/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > > new file mode 100644
> > > > index 00..5663ecccde
> > > > --- /dev/null
> > > > +++ b/drivers/net/i40e/i40e_recycle_mbufs_vec_common.c
> > > > @@ -0,0 +1,147 @@
> > > > +/* SPDX-License-Identifier: BSD-3-Clause
> > > > + * Copyright (c) 2023 Arm Limited.
> > > > + */
> > > > +
> > > > +#include 
> > > > +#include 
> > > > +
> > > > +#include "base/i40e_prototype.h"
> > > > +#include "base/i40e_type.h"
> > > > +#include "i40e_ethdev.h"
> > > > +#include "i40e_rxtx.h"
> > > > +
> > > > +#pragma GCC diagnostic ignored "-Wcast-qual"
> > > > +
> > > > +void
> > > > +i40e_recycle_rx_descriptors_refill_vec(void *rx_queue, uint16_t
> > > > +nb_mbufs) {
> > > > +   struct i40e_rx_queue *rxq = rx_queue;
> > > > +   struct i40e_rx_entry *rxep;
> > > > +   volatile union i40e_rx_desc *rxdp;
> > > > +   uint16_t rx_id;
> > > > +   uint64_t paddr;
> > > > +   uint64_t dma_addr;
> > > > +   uint16_t i;
> > > > +
> > > > +   rxdp = rxq->rx_ring + rxq->rxrearm_start;
> > > > +   rxep = &rxq->sw_ring[rxq->rxrearm_start];
> > > > +
> > > > +   for (i = 0; i < nb_mbufs; i++) {
> > > > +   /* Initialize rxdp descs. */
> > > > +   paddr = (rxep[i].mbuf)->buf_iova +
> > > > RTE_PKTMBUF_HEADROOM;
> > > > +   dma_addr = rte_cpu_to_le_64(paddr);
> > > > +   /* flush desc with pa dma_addr */
> > > > +   rxdp[i].read.hdr_addr = 0;
> > > > +   rxdp[i].read.pkt_addr = dma_addr;
> > > > +   }
> > > > +
> > > > +   /* Update the descriptor initializer index */
> > > > +   rxq->rxrearm_start += nb_mbufs;
> > > > +   rx_id = rxq->rxrearm_start - 1;
> > > > +
> > > > +   if (unlikely(rxq->rxrearm_start >= rxq->nb_rx_desc)) {
> > > > +   rxq->rxrearm_start = 0;
> > > > +   rx_id = rxq->nb_rx_desc - 1;
> > > > +   }
> > > > +
> > > > +   rxq->rxrearm_nb -= nb_mbufs;
> > > > +
> > > > +   rte_io_wmb();
> > > > +   /* Update the tail pointer on the NIC */
> > > > +   I40E_PCI_REG_WRITE_RELAXED(rxq->qrx_tail, rx_id); }
> > > > +
> > > > +uint16_t
> > > > +i40e_recycle_tx_mbufs_reuse_vec(void *tx_queue,
> > > > +   struct rte_eth_recycle_rxq_info *recy

[PATCH 1/2] build: fail if explicitly requested lib is unbuildable

[Bruce Richardson]

When the user passes a list of desired libraries to build via the
"enable_libs" option, the expectation is that those libraries should be
part of the build. However, if those libs have either external or
internal dependencies, they still may be silently disabled, for example:
running "meson setup -Denable_libs=security build" will successfully
run, but the security lib will not be configured as "cryptodev" is
missing.

We can fix this by setting a flag to indicate when the libraries are
specified via an enable_libs flag. If so, then we error out when a
library is unbuildable, giving a suitable error message. For the above
example case, the "meson setup" run fails with:

Message: Disabling security [lib/security]: missing internal dependency 
"cryptodev"

lib/meson.build:218:16: ERROR: Problem encountered: Cannot build explicitly 
requested lib "security".
Please add missing dependency "cryptodev" to "enable_libs" option

Signed-off-by: Bruce Richardson 
---
 lib/meson.build | 10 ++
 1 file changed, 10 insertions(+)

diff --git a/lib/meson.build b/lib/meson.build
index 099b0ed18a..cf4aa63630 100644
--- a/lib/meson.build
+++ b/lib/meson.build
@@ -131,7 +131,9 @@ endforeach
 disable_libs = run_command(list_dir_globs, get_option('disable_libs'), check: 
true).stdout().split()
 
 enable_libs = run_command(list_dir_globs, get_option('enable_libs'), check: 
true).stdout().split()
+require_libs = true
 if enable_libs.length() == 0
+require_libs = false
 enable_libs += optional_libs
 endif
 enable_libs += always_enable
@@ -189,6 +191,10 @@ foreach l:libraries
 
 if build
 subdir(l)
+if not build and require_libs
+error('Cannot build explicitly requested lib "@0@".\n'.format(name)
++'\tReason: ' + reason)
+endif
 endif
 if name != l
 warning('Library name, "@0@", and directory name, "@1@", do not 
match'.format(name, l))
@@ -208,6 +214,10 @@ foreach l:libraries
 endif
 message('Disabling @1@ [@2@]: missing internal dependency "@0@"'
 .format(d, name, 'lib/' + l))
+if require_libs
+error('Cannot build explicitly requested lib 
"@0@".\n'.format(name)
++ '\tPlease add missing dependency "@0@" to 
"enable_libs" option'.format(d))
+endif
 else
 shared_deps += [get_variable('shared_rte_' + d)]
 static_deps += [get_variable('static_rte_' + d)]
-- 
2.39.2

[PATCH 2/2] build: fail if explicitly requested driver is unbuildable

[Bruce Richardson]

When the user passes a list of desired drivers to build via the
"enable_drivers" option, the expectation is that those drivers should be
part of the build. However, if those drivers have either external or
internal dependencies, they still may be silently disabled, for example:
running "meson setup -Denable_drivers=net/iavf build" will successfully
run, but the iavf net driver will not be configured as "common/iavf" is
missing.

We can fix this by setting a flag to indicate when the drivers are
specified via an enable_drivers flag. However, unlike when erroring out
on missing libs, we don't error out if a driver in unbuildable, unless
the driver name explicitly appears in the "enable_drivers" list. This is
implemented this way to ensure that wildcarding still works. For
example: we still want to allow "meson setup -Denable_drivers=net/*" to
work, configuring only the buildable network drivers. While it's true
that this additional restriction may cause some builds to pass when they
should fail, e.g. if the wildcard refers only to a single driver,
implementing things this way avoids massive amounts of complexity, and
is still an improvement on the status-quo.

Suggested-by: Anatoly Burakov 
Signed-off-by: Bruce Richardson 
---
 drivers/meson.build | 14 ++
 1 file changed, 14 insertions(+)

diff --git a/drivers/meson.build b/drivers/meson.build
index 417b64b8fc..8c775bbe62 100644
--- a/drivers/meson.build
+++ b/drivers/meson.build
@@ -41,7 +41,9 @@ disable_drivers = run_command(list_dir_globs, 
disable_drivers, check: true).stdo
 # add cmdline enabled drivers and meson enabled drivers together
 enable_drivers = ',' + get_option('enable_drivers')
 enable_drivers = run_command(list_dir_globs, enable_drivers, check: 
true).stdout().split()
+require_drivers = true
 if enable_drivers.length() == 0
+require_drivers = false
 enable_drivers = run_command(list_dir_globs, '*/*', check: 
true).stdout().split()
 endif
 
@@ -155,6 +157,12 @@ foreach subpath:subdirs
 build = false
 reason = 'requires IOVA in mbuf (set enable_iova_as_pa option)'
 endif
+# error out if we can't build a driver and that driver was 
explicitly requested,
+# i.e. not via wildcard.
+if not build and require_drivers and 
get_option('enable_drivers').contains(drv_path)
+error('Cannot build explicitly requested driver 
"@0@".\n'.format(drv_path)
++'\tReason: ' + reason)
+endif
 
 # get dependency objs from strings
 shared_deps = ext_deps
@@ -171,6 +179,12 @@ foreach subpath:subdirs
 endif
 message('Disabling @1@ [@2@]: missing internal dependency 
"@0@"'
 .format(d, name, 'drivers/' + drv_path))
+# error out if we can't build a driver and that driver was 
explicitly
+# requested, i.e. not via wildcard.
+if require_drivers and 
get_option('enable_drivers').contains(drv_path)
+error('Cannot build explicitly requested driver 
"@0@".\n'.format(drv_path)
++'\tPlease enable missing dependency 
"@0@"'.format(d))
+endif
 else
 shared_deps += [get_variable('shared_rte_' + d)]
 static_deps += [get_variable('static_rte_' + d)]
-- 
2.39.2

Re: [PATCH 1/2] build: fail if explicitly requested lib is unbuildable

[David Marchand]

On Fri, Sep 1, 2023 at 4:23 PM Bruce Richardson
 wrote:
>
> When the user passes a list of desired libraries to build via the
> "enable_libs" option, the expectation is that those libraries should be
> part of the build. However, if those libs have either external or
> internal dependencies, they still may be silently disabled, for example:
> running "meson setup -Denable_libs=security build" will successfully
> run, but the security lib will not be configured as "cryptodev" is
> missing.
>
> We can fix this by setting a flag to indicate when the libraries are
> specified via an enable_libs flag. If so, then we error out when a
> library is unbuildable, giving a suitable error message. For the above
> example case, the "meson setup" run fails with:
>
> Message: Disabling security [lib/security]: missing internal dependency 
> "cryptodev"
>
> lib/meson.build:218:16: ERROR: Problem encountered: Cannot build explicitly 
> requested lib "security".
> Please add missing dependency "cryptodev" to "enable_libs" option
>
> Signed-off-by: Bruce Richardson 

I remember setting a enable_drivers=net/af_xdp on a system lacking
libbpf-devel and not getting warned about its absence.
So I suspect enable_drivers is affected by the same issue.


-- 
David Marchand

Re: [PATCH 1/2] build: fail if explicitly requested lib is unbuildable

[David Marchand]

On Fri, Sep 1, 2023 at 4:29 PM David Marchand  wrote:
>
> On Fri, Sep 1, 2023 at 4:23 PM Bruce Richardson
>  wrote:
> >
> > When the user passes a list of desired libraries to build via the
> > "enable_libs" option, the expectation is that those libraries should be
> > part of the build. However, if those libs have either external or
> > internal dependencies, they still may be silently disabled, for example:
> > running "meson setup -Denable_libs=security build" will successfully
> > run, but the security lib will not be configured as "cryptodev" is
> > missing.
> >
> > We can fix this by setting a flag to indicate when the libraries are
> > specified via an enable_libs flag. If so, then we error out when a
> > library is unbuildable, giving a suitable error message. For the above
> > example case, the "meson setup" run fails with:
> >
> > Message: Disabling security [lib/security]: missing internal dependency 
> > "cryptodev"
> >
> > lib/meson.build:218:16: ERROR: Problem encountered: Cannot build explicitly 
> > requested lib "security".
> > Please add missing dependency "cryptodev" to "enable_libs" option
> >
> > Signed-off-by: Bruce Richardson 
>
> I remember setting a enable_drivers=net/af_xdp on a system lacking
> libbpf-devel and not getting warned about its absence.
> So I suspect enable_drivers is affected by the same issue.

Arf sorry, did not see it was a two patch series :-).

-- 
David Marchand

Re: [PATCH 1/2] build: fail if explicitly requested lib is unbuildable

[Bruce Richardson]

On Fri, Sep 01, 2023 at 04:30:56PM +0200, David Marchand wrote:
> On Fri, Sep 1, 2023 at 4:29 PM David Marchand  
> wrote:
> >
> > On Fri, Sep 1, 2023 at 4:23 PM Bruce Richardson
> >  wrote:
> > >
> > > When the user passes a list of desired libraries to build via the
> > > "enable_libs" option, the expectation is that those libraries should be
> > > part of the build. However, if those libs have either external or
> > > internal dependencies, they still may be silently disabled, for example:
> > > running "meson setup -Denable_libs=security build" will successfully
> > > run, but the security lib will not be configured as "cryptodev" is
> > > missing.
> > >
> > > We can fix this by setting a flag to indicate when the libraries are
> > > specified via an enable_libs flag. If so, then we error out when a
> > > library is unbuildable, giving a suitable error message. For the above
> > > example case, the "meson setup" run fails with:
> > >
> > > Message: Disabling security [lib/security]: missing internal dependency 
> > > "cryptodev"
> > >
> > > lib/meson.build:218:16: ERROR: Problem encountered: Cannot build 
> > > explicitly requested lib "security".
> > > Please add missing dependency "cryptodev" to "enable_libs" option
> > >
> > > Signed-off-by: Bruce Richardson 
> >
> > I remember setting a enable_drivers=net/af_xdp on a system lacking
> > libbpf-devel and not getting warned about its absence.
> > So I suspect enable_drivers is affected by the same issue.
> 
> Arf sorry, did not see it was a two patch series :-).
> 
No problem.

At least your response confirms to me that this is indeed a common issue,
that could do with a solution. As I explain in patch 2, fixing for drivers
is a little trickier than for libs, as for drivers we really need to take
account of wildcards to avoid breaking people's existing builds. [Or, at
minimum, my own builds! :-)].

BTW: while we could look to handle wildcards for libs, I think their use is
probably much more limited there, and I decided to go with the simplest
possible solution instead.

/Bruce

Re: [PATCH 1/3] vhost: fix build for powerpc

[Bruce Richardson]

+PPC maintainer

On Thu, Aug 31, 2023 at 01:10:56PM +0100, Bruce Richardson wrote:
> When building on Ubuntu using the packaged powerpc compiler[1], a
> warning is issued about the print format of the __u64 values.
> 
> ../../lib/vhost/vduse.c: In function ‘vduse_vring_setup’:
> ../../lib/vhost/vhost.h:676:17: error: format ‘%llx’ expects argument of
> type ‘long long unsigned int’, but argument 5 has type ‘__u64’ {aka
> ‘long unsigned int’} [-Werror=format=]
>   676 | "VHOST_CONFIG: (%s) " fmt, prefix, ##args)
>   | ^
> 
> Changing the format specifier to %lx, or to use PRIx64 breaks other
> builds, so the safest solution is to explicitly typecast the printed
> values to match the format string.
> 
> [1] powerpc64le-linux-gnu-gcc (Ubuntu 12.3.0-1ubuntu1~23.04) 12.3.0
> 
> Fixes: a9120db8b98b ("vhost: add VDUSE device startup")
> Cc: maxime.coque...@redhat.com
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Bruce Richardson 
> ---
>  lib/vhost/vduse.c | 9 ++---
>  1 file changed, 6 insertions(+), 3 deletions(-)
> 
> diff --git a/lib/vhost/vduse.c b/lib/vhost/vduse.c
> index 73ed424232..e2b6d35d37 100644
> --- a/lib/vhost/vduse.c
> +++ b/lib/vhost/vduse.c
> @@ -162,9 +162,12 @@ vduse_vring_setup(struct virtio_net *dev, unsigned int 
> index)
>  
>   VHOST_LOG_CONFIG(dev->ifname, INFO, "VQ %u info:\n", index);
>   VHOST_LOG_CONFIG(dev->ifname, INFO, "\tnum: %u\n", vq_info.num);
> - VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdesc_addr: %llx\n", 
> vq_info.desc_addr);
> - VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdriver_addr: %llx\n", 
> vq_info.driver_addr);
> - VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdevice_addr: %llx\n", 
> vq_info.device_addr);
> + VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdesc_addr: %llx\n",
> + (unsigned long long)vq_info.desc_addr);
> + VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdriver_addr: %llx\n",
> + (unsigned long long)vq_info.driver_addr);
> + VHOST_LOG_CONFIG(dev->ifname, INFO, "\tdevice_addr: %llx\n",
> + (unsigned long long)vq_info.device_addr);
>   VHOST_LOG_CONFIG(dev->ifname, INFO, "\tavail_idx: %u\n", 
> vq_info.split.avail_index);
>   VHOST_LOG_CONFIG(dev->ifname, INFO, "\tready: %u\n", vq_info.ready);
>  
> -- 
> 2.39.2
> 

RE: [PATCH 1/2] build: fail if explicitly requested lib is unbuildable

[Morten Brørup]

> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Friday, 1 September 2023 16.24
> 
> When the user passes a list of desired libraries to build via the
> "enable_libs" option, the expectation is that those libraries should be
> part of the build. However, if those libs have either external or
> internal dependencies, they still may be silently disabled, for example:
> running "meson setup -Denable_libs=security build" will successfully
> run, but the security lib will not be configured as "cryptodev" is
> missing.
> 
> We can fix this by setting a flag to indicate when the libraries are
> specified via an enable_libs flag. If so, then we error out when a
> library is unbuildable, giving a suitable error message. For the above
> example case, the "meson setup" run fails with:
> 
> Message: Disabling security [lib/security]: missing internal dependency
> "cryptodev"
> 
> lib/meson.build:218:16: ERROR: Problem encountered: Cannot build explicitly
> requested lib "security".
>   Please add missing dependency "cryptodev" to "enable_libs" option
> 
> Signed-off-by: Bruce Richardson 

Acked-by: Morten Brørup 

RE: [PATCH 2/2] build: fail if explicitly requested driver is unbuildable

[Morten Brørup]

> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> Sent: Friday, 1 September 2023 16.24
> 
> When the user passes a list of desired drivers to build via the
> "enable_drivers" option, the expectation is that those drivers should be
> part of the build. However, if those drivers have either external or
> internal dependencies, they still may be silently disabled, for example:
> running "meson setup -Denable_drivers=net/iavf build" will successfully
> run, but the iavf net driver will not be configured as "common/iavf" is
> missing.
> 
> We can fix this by setting a flag to indicate when the drivers are
> specified via an enable_drivers flag. However, unlike when erroring out
> on missing libs, we don't error out if a driver in unbuildable, unless
> the driver name explicitly appears in the "enable_drivers" list. This is
> implemented this way to ensure that wildcarding still works. For
> example: we still want to allow "meson setup -Denable_drivers=net/*" to
> work, configuring only the buildable network drivers. While it's true
> that this additional restriction may cause some builds to pass when they
> should fail, e.g. if the wildcard refers only to a single driver,
> implementing things this way avoids massive amounts of complexity, and
> is still an improvement on the status-quo.
> 
> Suggested-by: Anatoly Burakov 
> Signed-off-by: Bruce Richardson 
> ---

Acked-by: Morten Brørup 

Re: [PATCH v2 3/6] baseband/fpga_5gnr_fec: renaming for consistency

[Maxime Coquelin]

On 5/25/23 20:28, Hernan Vargas wrote:

Rename generic functions and constants using the FPGA 5GNR prefix naming
to prepare for code reuse for new FPGA implementation variant.
No functional impact.

Signed-off-by: Hernan Vargas 
---
  .../baseband/fpga_5gnr_fec/fpga_5gnr_fec.h| 117 +++--
  .../fpga_5gnr_fec/rte_fpga_5gnr_fec.c | 455 --
  .../fpga_5gnr_fec/rte_pmd_fpga_5gnr_fec.h |  17 +-
  3 files changed, 269 insertions(+), 320 deletions(-)



Reviewed-by: Maxime Coquelin 

Re: [PATCH v2 4/6] baseband/fpga_5gnr_fec: add Vista Creek variant

[Maxime Coquelin]

On 5/25/23 20:28, Hernan Vargas wrote:

Create a new file vc_5gnr_pmd.h to store structures and macros specific
to Vista Creek 5G FPGA implementation and rename functions specific to
the Vista Creek variant.

Signed-off-by: Hernan Vargas 
---
  .../baseband/fpga_5gnr_fec/fpga_5gnr_fec.h| 183 +-
  .../fpga_5gnr_fec/rte_fpga_5gnr_fec.c | 531 +-
  drivers/baseband/fpga_5gnr_fec/vc_5gnr_pmd.h  | 140 +
  3 files changed, 426 insertions(+), 428 deletions(-)
  create mode 100644 drivers/baseband/fpga_5gnr_fec/vc_5gnr_pmd.h

diff --git a/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h 
b/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
index 9300349a731b..c88d276cc48f 100644
--- a/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
+++ b/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
@@ -8,6 +8,8 @@
  #include 
  #include 
  
+#include "vc_5gnr_pmd.h"

+
  /* Helper macro for logging */
  #define rte_bbdev_log(level, fmt, ...) \
rte_log(RTE_LOG_ ## level, fpga_5gnr_fec_logtype, fmt "\n", \
@@ -25,32 +27,20 @@
  #define FPGA_5GNR_FEC_PF_DRIVER_NAME intel_fpga_5gnr_fec_pf
  #define FPGA_5GNR_FEC_VF_DRIVER_NAME intel_fpga_5gnr_fec_vf
  
-/* FPGA 5GNR FEC PCI vendor & device IDs */

-#define FPGA_5GNR_FEC_VENDOR_ID (0x8086)
-#define FPGA_5GNR_FEC_PF_DEVICE_ID (0x0D8F)
-#define FPGA_5GNR_FEC_VF_DEVICE_ID (0x0D90)
-
-/* Align DMA descriptors to 256 bytes - cache-aligned */
-#define FPGA_5GNR_RING_DESC_ENTRY_LENGTH (8)
-/* Ring size is in 256 bits (32 bytes) units */
-#define FPGA_RING_DESC_LEN_UNIT_BYTES (32)
-/* Maximum size of queue */
-#define FPGA_5GNR_RING_MAX_SIZE (1024)
-
-#define FPGA_NUM_UL_QUEUES (32)
-#define FPGA_NUM_DL_QUEUES (32)
-#define FPGA_TOTAL_NUM_QUEUES (FPGA_NUM_UL_QUEUES + FPGA_NUM_DL_QUEUES)
-#define FPGA_NUM_INTR_VEC (FPGA_TOTAL_NUM_QUEUES - RTE_INTR_VEC_RXTX_OFFSET)
-
  #define FPGA_5GNR_INVALID_HW_QUEUE_ID (0x)
-
  #define FPGA_5GNR_QUEUE_FLUSH_TIMEOUT_US (1000)
  #define FPGA_5GNR_HARQ_RDY_TIMEOUT (10)
  #define FPGA_5GNR_TIMEOUT_CHECK_INTERVAL (5)
  #define FPGA_5GNR_DDR_OVERFLOW (0x10)
-
  #define FPGA_5GNR_DDR_WR_DATA_LEN_IN_BYTES 8
  #define FPGA_5GNR_DDR_RD_DATA_LEN_IN_BYTES 8
+/* Align DMA descriptors to 256 bytes - cache-aligned. */
+#define FPGA_5GNR_RING_DESC_ENTRY_LENGTH (8)
+/* Maximum size of queue. */
+#define FPGA_5GNR_RING_MAX_SIZE (1024)
+
+#define VC_5GNR_FPGA_VARIANT   0
+#define AGX100_FPGA_VARIANT1
  
  /* Constants from K0 computation from 3GPP 38.212 Table 5.4.2.1-2 */

  #define N_ZC_1 66 /* N = 66 Zc for BG 1 */
@@ -62,32 +52,7 @@
  #define K0_3_1 56 /* K0 fraction numerator for rv 3 and BG 1 */
  #define K0_3_2 43 /* K0 fraction numerator for rv 3 and BG 2 */
  
-/* FPGA 5GNR FEC Register mapping on BAR0 */

-enum {
-   FPGA_5GNR_FEC_VERSION_ID = 0x, /* len: 4B */
-   FPGA_5GNR_FEC_CONFIGURATION = 0x0004, /* len: 2B */
-   FPGA_5GNR_FEC_QUEUE_PF_VF_MAP_DONE = 0x0008, /* len: 1B */
-   FPGA_5GNR_FEC_LOAD_BALANCE_FACTOR = 0x000a, /* len: 2B */
-   FPGA_5GNR_FEC_RING_DESC_LEN = 0x000c, /* len: 2B */
-   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_LW = 0x0018, /* len: 4B */
-   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_HI = 0x001c, /* len: 4B */
-   FPGA_5GNR_FEC_QUEUE_MAP = 0x0040, /* len: 256B */
-   FPGA_5GNR_FEC_RING_CTRL_REGS = 0x0200, /* len: 2048B */
-   FPGA_5GNR_FEC_DDR4_WR_ADDR_REGS = 0x0A00, /* len: 4B */
-   FPGA_5GNR_FEC_DDR4_WR_DATA_REGS = 0x0A08, /* len: 8B */
-   FPGA_5GNR_FEC_DDR4_WR_DONE_REGS = 0x0A10, /* len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_ADDR_REGS = 0x0A18, /* len: 4B */
-   FPGA_5GNR_FEC_DDR4_RD_DONE_REGS = 0x0A20, /* len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_RDY_REGS = 0x0A28, /* len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_DATA_REGS = 0x0A30, /* len: 8B */
-   FPGA_5GNR_FEC_DDR4_ADDR_RDY_REGS = 0x0A38, /* len: 1B */
-   FPGA_5GNR_FEC_HARQ_BUF_SIZE_RDY_REGS = 0x0A40, /* len: 1B */
-   FPGA_5GNR_FEC_HARQ_BUF_SIZE_REGS = 0x0A48, /* len: 4B */
-   FPGA_5GNR_FEC_MUTEX = 0x0A60, /* len: 4B */
-   FPGA_5GNR_FEC_MUTEX_RESET = 0x0A68  /* len: 4B */
-};
-
-/* FPGA 5GNR FEC Ring Control Registers */
+/* FPGA 5GNR Ring Control Registers. */
  enum {
FPGA_5GNR_FEC_RING_HEAD_ADDR = 0x0008,
FPGA_5GNR_FEC_RING_SIZE = 0x0010,
@@ -98,113 +63,27 @@ enum {
FPGA_5GNR_FEC_RING_HEAD_POINT = 0x001C
  };
  
-/* FPGA 5GNR FEC DESCRIPTOR ERROR */

+/* VC 5GNR and AGX100 common register mapping on BAR0. */
  enum {
-   DESC_ERR_NO_ERR = 0x0,
-   DESC_ERR_K_P_OUT_OF_RANGE = 0x1,
-   DESC_ERR_Z_C_NOT_LEGAL = 0x2,
-   DESC_ERR_DESC_OFFSET_ERR = 0x3,
-   DESC_ERR_DESC_READ_FAIL = 0x8,
-   DESC_ERR_DESC_READ_TIMEOUT = 0x9,
-   DESC_ERR_DESC_READ_TLP_POISONED = 0xA,
-   DESC_ERR_HARQ_INPUT_LEN = 0xB,
-   DESC_ERR_CB_READ_FAIL = 0xC,
-   DESC_ERR_CB_READ_TIMEOUT = 0xD,
-   DESC_ERR_CB_READ_TLP_POISONED = 0xE,
-   DE

Re: [RFC] cache guard

[Mattias Rönnblom]

On 2023-09-01 14:26, Thomas Monjalon wrote:

27/08/2023 10:34, Morten Brørup:

+CC Honnappa and Konstantin, Ring lib maintainers
+CC Mattias, PRNG lib maintainer


From: Bruce Richardson [mailto:bruce.richard...@intel.com]
Sent: Friday, 25 August 2023 11.24

On Fri, Aug 25, 2023 at 11:06:01AM +0200, Morten Brørup wrote:

+CC mempool maintainers


From: Bruce Richardson [mailto:bruce.richard...@intel.com]
Sent: Friday, 25 August 2023 10.23

On Fri, Aug 25, 2023 at 08:45:12AM +0200, Morten Brørup wrote:

Bruce,

With this patch [1], it is noted that the ring producer and

consumer data

should not be on adjacent cache lines, for performance reasons.


[1]:



https://git.dpdk.org/dpdk/commit/lib/librte_ring/rte_ring.h?id=d9f0d3a1f
fd4b66

e75485cc8b63b9aedfbdfe8b0


(It's obvious that they cannot share the same cache line, because

they are

accessed by two different threads.)


Intuitively, I would think that having them on different cache

lines would

suffice. Why does having an empty cache line between them make a

difference?


And does it need to be an empty cache line? Or does it suffice

having the

second structure start at two cache lines after the start of the

first

structure (e.g. if the size of the first structure is two cache

lines)?


I'm asking because the same principle might apply to other code

too.



Hi Morten,

this was something we discovered when working on the distributor

library.

If we have cachelines per core where there is heavy access, having

some

cachelines as a gap between the content cachelines can help

performance. We

believe this helps due to avoiding issues with the HW prefetchers

(e.g.

adjacent cacheline prefetcher) bringing in the second cacheline
speculatively when an operation is done on the first line.


I guessed that it had something to do with speculative prefetching,

but wasn't sure. Good to get confirmation, and that it has a measureable
effect somewhere. Very interesting!


NB: More comments in the ring lib about stuff like this would be nice.

So, for the mempool lib, what do you think about applying the same

technique to the rte_mempool_debug_stats structure (which is an array
indexed per lcore)... Two adjacent lcores heavily accessing their local
mempool caches seems likely to me. But how heavy does the access need to
be for this technique to be relevant?




No idea how heavy the accesses need to be for this to have a noticable
effect. For things like debug stats, I wonder how worthwhile making such
a
change would be, but then again, any change would have very low impact
too
in that case.


I just tried adding padding to some of the hot structures in our own 
application, and observed a significant performance improvement for those.

So I think this technique should have higher visibility in DPDK by adding a new 
cache macro to rte_common.h:


+1 to make more visibility in doc and adding a macro, good idea!





A worry I have is that for CPUs with large (in this context) N, you will 
end up with a lot of padding to avoid next-N-lines false sharing. That 
would be padding after, and in the general (non-array) case also before, 
the actual per-lcore data. A slight nuisance is also that those 
prefetched lines of padding, will never contain anything useful, and 
thus fetching them will always be a waste.


Padding/alignment may not be the only way to avoid HW-prefetcher-induced 
false sharing for per-lcore data structures.


What we are discussing here is organizing the statically allocated 
per-lcore structs of a particular module in an array with the 
appropriate padding/alignment. In this model, all data related to a 
particular module is close (memory address/page-wise), but not so close 
to cause false sharing.


/* rte_a.c */

struct rte_a_state
{
int x;
RTE_CACHE_GUARD;
} __rte_cache_aligned;

static struct rte_a_state a_states[RTE_MAX_LCORE];

/* rte_b.c */

struct rte_b_state
{
char y;
char z;
RTE_CACHE_GUARD;
} __rte_cache_aligned;


static struct rte_b_state b_states[RTE_MAX_LCORE];

What you would end up with in runtime when the linker has done its job 
is something that essentially looks like this (in memory):


struct {
struct rte_a_state a_states[RTE_MAX_LCORE];
struct rte_b_state b_states[RTE_MAX_LCORE];
};

You could consider turning it around, and keeping data (i.e., module 
structs) related to a particular lcore, for all modules, close. In other 
words, keeping a per-lcore arrays of variable-sized elements.


So, something that will end up looking like this (in memory, not in the 
source code):


struct rte_lcore_state
{
struct rte_a_state a_state;
struct rte_b_state b_state;
RTE_CACHE_GUARD;
};

struct rte_lcore_state lcore_states[RTE_LCORE_MAX];

In such a scenario, the per-lcore struct type for a module need not (and 
should not) be cache-line-aligned (but may still have some alignment 
requirements). Data will be more tightly packed, and the "nex

Re: [RFC PATCH v4 4/4] dts: format docstrigs to google format

[Jeremy Spewock]

On Thu, Aug 31, 2023 at 6:04 AM Juraj Linkeš 
wrote:

> WIP: only one module is reformatted to serve as a demonstration.
>
> The google format is documented here [0].
>
> [0]: https://google.github.io/styleguide/pyguide.html
>
> Signed-off-by: Juraj Linkeš 
> ---
>  dts/framework/testbed_model/node.py | 171 +++-
>  1 file changed, 118 insertions(+), 53 deletions(-)
>
> diff --git a/dts/framework/testbed_model/node.py
> b/dts/framework/testbed_model/node.py
> index 23efa79c50..619743ebe7 100644
> --- a/dts/framework/testbed_model/node.py
> +++ b/dts/framework/testbed_model/node.py
> @@ -3,8 +3,13 @@
>  # Copyright(c) 2022-2023 PANTHEON.tech s.r.o.
>  # Copyright(c) 2022-2023 University of New Hampshire
>
> -"""
> -A node is a generic host that DTS connects to and manages.
> +"""Common functionality for node management.
> +
> +There's a base class, Node, that's supposed to be extended by other
> classes
> +with functionality specific to that node type.
> +The only part that can be used standalone is the Node.skip_setup static
> method,
> +which is a decorator used to skip method execution if skip_setup is passed
> +by the user on the cmdline or in an env variable.
>  """
>
>  from abc import ABC
> @@ -35,10 +40,26 @@
>
>
>  class Node(ABC):
> -"""
> -Basic class for node management. This class implements methods that
> -manage a node, such as information gathering (of CPU/PCI/NIC) and
> -environment setup.
> +"""The base class for node management.
> +
> +It shouldn't be instantiated, but rather extended.
> +It implements common methods to manage any node:
> +
> +   * connection to the node
> +   * information gathering of CPU
> +   * hugepages setup
> +
> +Arguments:
>

My only comment would be we might want to make this Args instead of
arguments just to line up with the rest of the comments, but like you said
this is just a proof of concept really.

Acked-by: Jeremy Spweock 

Re: [PATCH v2 5/6] baseband/fpga_5gnr_fec: add AGX100 support

[Maxime Coquelin]

On 5/25/23 20:28, Hernan Vargas wrote:

Add support for new FPGA variant AGX100 (on Arrow Creek N6000).

Signed-off-by: Hernan Vargas 
---
  doc/guides/bbdevs/fpga_5gnr_fec.rst   |   72 +-
  drivers/baseband/fpga_5gnr_fec/agx100_pmd.h   |  273 
  .../baseband/fpga_5gnr_fec/fpga_5gnr_fec.h|6 +
  .../fpga_5gnr_fec/rte_fpga_5gnr_fec.c | 1197 +++--
  4 files changed, 1395 insertions(+), 153 deletions(-)
  create mode 100644 drivers/baseband/fpga_5gnr_fec/agx100_pmd.h

diff --git a/doc/guides/bbdevs/fpga_5gnr_fec.rst 
b/doc/guides/bbdevs/fpga_5gnr_fec.rst
index 9d71585e9e18..c27db695a834 100644
--- a/doc/guides/bbdevs/fpga_5gnr_fec.rst
+++ b/doc/guides/bbdevs/fpga_5gnr_fec.rst
@@ -6,12 +6,13 @@ Intel(R) FPGA 5GNR FEC Poll Mode Driver
  
  The BBDEV FPGA 5GNR FEC poll mode driver (PMD) supports an FPGA implementation of a VRAN

  LDPC Encode / Decode 5GNR wireless acceleration function, using Intel's PCI-e 
and FPGA
-based Vista Creek device.
+based Vista Creek (N3000, referred to as VC_5GNR in the code) as well as Arrow 
Creek (N6000,
+referred to as AGX100 in the code).
  
  Features

  
  
-FPGA 5GNR FEC PMD supports the following features:

+FPGA 5GNR FEC PMD supports the following BBDEV capabilities:
  
  - LDPC Encode in the DL

  - LDPC Decode in the UL
@@ -67,10 +68,18 @@ Initialization
  
  When the device first powers up, its PCI Physical Functions (PF) can be listed through this command:
  
+Vista Creek (N3000)

+
  .. code-block:: console
  
sudo lspci -vd8086:0d8f
  
+Arrow Creek (N6000)

+
+.. code-block:: console
+
+  sudo lspci -vd8086:5799
+
  The physical and virtual functions are compatible with Linux UIO drivers:
  ``vfio`` and ``igb_uio``. However, in order to work the FPGA 5GNR FEC device 
firstly needs
  to be bound to one of these linux drivers through DPDK.
@@ -85,24 +94,34 @@ Install the DPDK igb_uio driver, bind it with the PF PCI 
device ID and use
  The igb_uio driver may be bound to the PF PCI device using one of two methods:
  
  
-1. PCI functions (physical or virtual, depending on the use case) can be bound to

-the UIO driver by repeating this command for every function.
+1. PCI functions (physical or virtual, depending on the use case) can be bound 
to the UIO driver by repeating this command for every function.
  
-.. code-block:: console

+  .. code-block:: console
+
+insmod igb_uio.ko
+
+  Bind N3000 to igb_uio
+
+  .. code-block:: console
  
-  insmod igb_uio.ko

-  echo "8086 0d8f" > /sys/bus/pci/drivers/igb_uio/new_id
-  lspci -vd8086:0d8f
+echo "8086 0d8f" > /sys/bus/pci/drivers/igb_uio/new_id
+lspci -vd8086:0d8f
  
+  Bind N6000 to igb_uio

+
+  .. code-block:: console
+
+echo "8086 5799" > /sys/bus/pci/drivers/igb_uio/new_id
+lspci -vd8086:5799
  
  2. Another way to bind PF with DPDK UIO driver is by using the ``dpdk-devbind.py`` tool
  
-.. code-block:: console

+  .. code-block:: console
  
-  cd 

-  ./usertools/dpdk-devbind.py -b igb_uio :06:00.0
+cd 
+./usertools/dpdk-devbind.py -b igb_uio :06:00.0
  
-where the PCI device ID (example: :06:00.0) is obtained using lspci -vd8086:0d8f

+where the PCI device ID (example: :06:00.0) is obtained using lspci 
-vd8086:0d8f for N3000 or lspci -vd8086:5799 for N6000
  
  
  In the same way the FPGA 5GNR FEC PF can be bound with vfio, but vfio driver does not

@@ -165,7 +184,6 @@ parameters defined in ``rte_fpga_5gnr_fec_conf`` structure:
uint8_t dl_bandwidth;
uint8_t ul_load_balance;
uint8_t dl_load_balance;
-  uint16_t flr_time_out;
};
  
  - ``pf_mode_en``: identifies whether only PF is to be used, or the VFs. PF and

@@ -176,12 +194,12 @@ parameters defined in ``rte_fpga_5gnr_fec_conf`` 
structure:
  
  - ``vf_*l_queues_number``: defines the hardware queue mapping for every VF.
  
-- ``*l_bandwidth``: in case of congestion on PCIe interface. The device

-  allocates different bandwidth to UL and DL. The weight is configured by this
-  setting. The unit of weight is 3 code blocks. For example, if the code block
-  cbps (code block per second) ratio between UL and DL is 12:1, then the
-  configuration value should be set to 36:3. The schedule algorithm is based
-  on code block regardless the length of each block.
+- ``*l_bandwidth``: Only used for the Vista Creek schedule algorithm in case of
+  congestion on PCIe interface. The device allocates different bandwidth to UL
+  and DL. The weight is configured by this setting. The unit of weight is 3 
code
+  blocks. For example, if the code block cbps (code block per second) ratio 
between
+  UL and DL is 12:1, then the configuration value should be set to 36:3.
+  The schedule algorithm is based on code block regardless the length of each 
block.
  
  - ``*l_load_balance``: hardware queues are load-balanced in a round-robin

fashion. Queues get filled first-in first-out until they reach a pre-defined
@@ -191,10 +209,6 @@ parameters defined 

Re: [PATCH v2 6/6] baseband/fpga_5gnr_fec: cosmetic comment changes

[Maxime Coquelin]

On 5/25/23 20:28, Hernan Vargas wrote:

Cosmetic changes for comments.
No functional impact.

Signed-off-by: Hernan Vargas 
---
  .../baseband/fpga_5gnr_fec/fpga_5gnr_fec.h|  93 ++--
  .../fpga_5gnr_fec/rte_fpga_5gnr_fec.c | 398 +-
  .../fpga_5gnr_fec/rte_pmd_fpga_5gnr_fec.h |  16 +-
  3 files changed, 252 insertions(+), 255 deletions(-)



Some of the changes are fixing code introduced earlier in the series.
Fix that in these patches directly.


diff --git a/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h 
b/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
index d0d9ee64dbde..c2aa5af2af40 100644
--- a/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
+++ b/drivers/baseband/fpga_5gnr_fec/fpga_5gnr_fec.h
@@ -11,7 +11,7 @@
  #include "agx100_pmd.h"
  #include "vc_5gnr_pmd.h"
  
-/* Helper macro for logging */

+/* Helper macro for logging. */
  #define rte_bbdev_log(level, fmt, ...) \
rte_log(RTE_LOG_ ## level, fpga_5gnr_fec_logtype, fmt "\n", \
##__VA_ARGS__)
@@ -24,7 +24,7 @@
  #define rte_bbdev_log_debug(fmt, ...)
  #endif
  
-/* FPGA 5GNR FEC driver names */

+/* FPGA 5GNR FEC driver names. */
  #define FPGA_5GNR_FEC_PF_DRIVER_NAME intel_fpga_5gnr_fec_pf
  #define FPGA_5GNR_FEC_VF_DRIVER_NAME intel_fpga_5gnr_fec_vf
  
@@ -43,15 +43,15 @@

  #define VC_5GNR_FPGA_VARIANT  0
  #define AGX100_FPGA_VARIANT   1
  
-/* Constants from K0 computation from 3GPP 38.212 Table 5.4.2.1-2 */

-#define N_ZC_1 66 /* N = 66 Zc for BG 1 */
-#define N_ZC_2 50 /* N = 50 Zc for BG 2 */
-#define K0_1_1 17 /* K0 fraction numerator for rv 1 and BG 1 */
-#define K0_1_2 13 /* K0 fraction numerator for rv 1 and BG 2 */
-#define K0_2_1 33 /* K0 fraction numerator for rv 2 and BG 1 */
-#define K0_2_2 25 /* K0 fraction numerator for rv 2 and BG 2 */
-#define K0_3_1 56 /* K0 fraction numerator for rv 3 and BG 1 */
-#define K0_3_2 43 /* K0 fraction numerator for rv 3 and BG 2 */
+/* Constants from K0 computation from 3GPP 38.212 Table 5.4.2.1-2. */
+#define N_ZC_1 66 /**< N = 66 Zc for BG 1. */
+#define N_ZC_2 50 /**< N = 50 Zc for BG 2. */
+#define K0_1_1 17 /**< K0 fraction numerator for rv 1 and BG 1. */
+#define K0_1_2 13 /**< K0 fraction numerator for rv 1 and BG 2. */
+#define K0_2_1 33 /**< K0 fraction numerator for rv 2 and BG 1. */
+#define K0_2_2 25 /**< K0 fraction numerator for rv 2 and BG 2. */
+#define K0_3_1 56 /**< K0 fraction numerator for rv 3 and BG 1. */
+#define K0_3_2 43 /**< K0 fraction numerator for rv 3 and BG 2. */
  
  /* FPGA 5GNR Ring Control Registers. */

  enum {
@@ -66,25 +66,25 @@ enum {
  
  /* VC 5GNR and AGX100 common register mapping on BAR0. */

  enum {
-   FPGA_5GNR_FEC_VERSION_ID = 0x, /**< len: 4B */
-   FPGA_5GNR_FEC_QUEUE_PF_VF_MAP_DONE = 0x0008, /**< len: 1B */
-   FPGA_5GNR_FEC_LOAD_BALANCE_FACTOR = 0x000A, /**< len: 2B */
-   FPGA_5GNR_FEC_RING_DESC_LEN = 0x000C, /**< len: 2B */
-   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_LW = 0x0018, /**< len: 4B */
-   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_HI = 0x001C, /**< len: 4B */
-   FPGA_5GNR_FEC_RING_CTRL_REGS = 0x0200, /**< len: 2048B */
-   FPGA_5GNR_FEC_DDR4_WR_ADDR_REGS = 0x0A00, /**< len: 4B */
-   FPGA_5GNR_FEC_DDR4_WR_DATA_REGS = 0x0A08, /**< len: 8B */
-   FPGA_5GNR_FEC_DDR4_WR_DONE_REGS = 0x0A10, /**< len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_ADDR_REGS = 0x0A18, /**< len: 4B */
-   FPGA_5GNR_FEC_DDR4_RD_DONE_REGS = 0x0A20, /**< len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_RDY_REGS = 0x0A28, /**< len: 1B */
-   FPGA_5GNR_FEC_DDR4_RD_DATA_REGS = 0x0A30, /**< len: 8B */
-   FPGA_5GNR_FEC_DDR4_ADDR_RDY_REGS = 0x0A38, /**< len: 1B */
-   FPGA_5GNR_FEC_HARQ_BUF_SIZE_RDY_REGS = 0x0A40, /**< len: 1B */
-   FPGA_5GNR_FEC_HARQ_BUF_SIZE_REGS = 0x0A48, /**< len: 4B */
-   FPGA_5GNR_FEC_MUTEX = 0x0A60, /**< len: 4B */
-   FPGA_5GNR_FEC_MUTEX_RESET = 0x0A68  /**< len: 4B */
+   FPGA_5GNR_FEC_VERSION_ID = 0x, /**< len: 4B. */
+   FPGA_5GNR_FEC_QUEUE_PF_VF_MAP_DONE = 0x0008, /**< len: 1B. */
+   FPGA_5GNR_FEC_LOAD_BALANCE_FACTOR = 0x000A, /**< len: 2B. */
+   FPGA_5GNR_FEC_RING_DESC_LEN = 0x000C, /**< len: 2B. */
+   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_LW = 0x0018, /**< len: 4B. */
+   FPGA_5GNR_FEC_VFQ_FLUSH_STATUS_HI = 0x001C, /**< len: 4B. */
+   FPGA_5GNR_FEC_RING_CTRL_REGS = 0x0200, /**< len: 2048B. */
+   FPGA_5GNR_FEC_DDR4_WR_ADDR_REGS = 0x0A00, /**< len: 4B. */
+   FPGA_5GNR_FEC_DDR4_WR_DATA_REGS = 0x0A08, /**< len: 8B. */
+   FPGA_5GNR_FEC_DDR4_WR_DONE_REGS = 0x0A10, /**< len: 1B. */
+   FPGA_5GNR_FEC_DDR4_RD_ADDR_REGS = 0x0A18, /**< len: 4B. */
+   FPGA_5GNR_FEC_DDR4_RD_DONE_REGS = 0x0A20, /**< len: 1B. */
+   FPGA_5GNR_FEC_DDR4_RD_RDY_REGS = 0x0A28, /**< len: 1B. */
+   FPGA_5GNR_FEC_DDR4_RD_DATA_REGS = 0x0A30, /**< len: 8B. */
+   FPGA_5GNR_FEC_DDR4_AD

RE: [RFC] cache guard

[Morten Brørup]

> From: Mattias Rönnblom [mailto:hof...@lysator.liu.se]
> Sent: Friday, 1 September 2023 18.58
> 
> On 2023-09-01 14:26, Thomas Monjalon wrote:
> > 27/08/2023 10:34, Morten Brørup:
> >> +CC Honnappa and Konstantin, Ring lib maintainers
> >> +CC Mattias, PRNG lib maintainer
> >>
> >>> From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> >>> Sent: Friday, 25 August 2023 11.24
> >>>
> >>> On Fri, Aug 25, 2023 at 11:06:01AM +0200, Morten Brørup wrote:
>  +CC mempool maintainers
> 
> > From: Bruce Richardson [mailto:bruce.richard...@intel.com]
> > Sent: Friday, 25 August 2023 10.23
> >
> > On Fri, Aug 25, 2023 at 08:45:12AM +0200, Morten Brørup wrote:
> >> Bruce,
> >>
> >> With this patch [1], it is noted that the ring producer and
> >>> consumer data
> > should not be on adjacent cache lines, for performance reasons.
> >>
> >> [1]:
> >
> >>>
> https://git.dpdk.org/dpdk/commit/lib/librte_ring/rte_ring.h?id=d9f0d3a1f
> >>> fd4b66
> > e75485cc8b63b9aedfbdfe8b0
> >>
> >> (It's obvious that they cannot share the same cache line, because
> >>> they are
> > accessed by two different threads.)
> >>
> >> Intuitively, I would think that having them on different cache
> >>> lines would
> > suffice. Why does having an empty cache line between them make a
> >>> difference?
> >>
> >> And does it need to be an empty cache line? Or does it suffice
> >>> having the
> > second structure start at two cache lines after the start of the
> >>> first
> > structure (e.g. if the size of the first structure is two cache
> >>> lines)?
> >>
> >> I'm asking because the same principle might apply to other code
> >>> too.
> >>
> > Hi Morten,
> >
> > this was something we discovered when working on the distributor
> >>> library.
> > If we have cachelines per core where there is heavy access, having
> >>> some
> > cachelines as a gap between the content cachelines can help
> >>> performance. We
> > believe this helps due to avoiding issues with the HW prefetchers
> >>> (e.g.
> > adjacent cacheline prefetcher) bringing in the second cacheline
> > speculatively when an operation is done on the first line.
> 
>  I guessed that it had something to do with speculative prefetching,
> >>> but wasn't sure. Good to get confirmation, and that it has a
> measureable
> >>> effect somewhere. Very interesting!
> 
>  NB: More comments in the ring lib about stuff like this would be
> nice.
> 
>  So, for the mempool lib, what do you think about applying the same
> >>> technique to the rte_mempool_debug_stats structure (which is an
> array
> >>> indexed per lcore)... Two adjacent lcores heavily accessing their
> local
> >>> mempool caches seems likely to me. But how heavy does the access
> need to
> >>> be for this technique to be relevant?
> 
> >>>
> >>> No idea how heavy the accesses need to be for this to have a
> noticable
> >>> effect. For things like debug stats, I wonder how worthwhile making
> such
> >>> a
> >>> change would be, but then again, any change would have very low
> impact
> >>> too
> >>> in that case.
> >>
> >> I just tried adding padding to some of the hot structures in our own
> application, and observed a significant performance improvement for
> those.
> >>
> >> So I think this technique should have higher visibility in DPDK by
> adding a new cache macro to rte_common.h:
> >
> > +1 to make more visibility in doc and adding a macro, good idea!
> >
> >
> >
> 
> A worry I have is that for CPUs with large (in this context) N, you will
> end up with a lot of padding to avoid next-N-lines false sharing. That
> would be padding after, and in the general (non-array) case also before,
> the actual per-lcore data. A slight nuisance is also that those
> prefetched lines of padding, will never contain anything useful, and
> thus fetching them will always be a waste.

Out of curiosity, what is the largest N anyone here on the list is aware of?

> 
> Padding/alignment may not be the only way to avoid HW-prefetcher-induced
> false sharing for per-lcore data structures.
> 
> What we are discussing here is organizing the statically allocated
> per-lcore structs of a particular module in an array with the
> appropriate padding/alignment. In this model, all data related to a
> particular module is close (memory address/page-wise), but not so close
> to cause false sharing.
> 
> /* rte_a.c */
> 
> struct rte_a_state
> {
>   int x;
>  RTE_CACHE_GUARD;
> } __rte_cache_aligned;
> 
> static struct rte_a_state a_states[RTE_MAX_LCORE];
> 
> /* rte_b.c */
> 
> struct rte_b_state
> {
>   char y;
>  char z;
>  RTE_CACHE_GUARD;
> } __rte_cache_aligned;
> 
> 
> static struct rte_b_state b_states[RTE_MAX_LCORE];
> 
> What you would end up with in runtime when the linker has done its job
> is something that essentially looks like this (in memory):
> 
> struct {
>