Re: [dpdk-dev] [PATCH v3 1/2] telemetry: support array values in data objects

[Thomas Monjalon]

02/07/2020 12:19, Ciara Power:
> +/**
> + * Add a container to a dictionary. A container is an existing telemetry data
> + * array. The dict the container is to be added to must have been started by
> + * rte_tel_data_start_dict(). The container must be an array of type
> + * uint64_t/int/string.
> + *
> + * @param d
> + *   The data structure passed to the callback
> + * @param val
> + *   The pointer to the container to be stored in the dict.
> + * @param keep
> + *   Flag to indicate that the container memory should not be automatically
> + *   freed by the telemetry library once it has finished with the data.
> + *   1 = keep, 0 = free.
> + * @return
> + *   0 on success, negative errno on error
> + */
> +__rte_experimental
> +int
> +rte_tel_data_add_dict_container(struct rte_tel_data *d, const char *name,
> + struct rte_tel_data *val, int keep);

Please fix this miss:

rte_telemetry.h:233: warning: The following parameter of 
rte_tel_data_add_dict_container(struct rte_tel_data *d, const char *name, 
struct rte_tel_data *val, int keep) is not documented:
+  parameter 'name'

Re: [dpdk-dev] [PATCH] rawdev: fix to remove EXPERIMENTAL from comment

[Thomas Monjalon]

07/07/2020 08:53, Hemant Agrawal:
> The experimental tags were removed, but the comment
> is still having API classification as EXPERIMENTAL
> 
> Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Hemant Agrawal 
> ---
>  lib/librte_rawdev/rte_rawdev.h | 3 ---
>  lib/librte_rawdev/rte_rawdev_pmd.h | 3 ---
>  2 files changed, 6 deletions(-)

Also in MAINTAINERS:
Raw device API - EXPERIMENTAL

Re: [dpdk-dev] [PATCH v6 2/2] ethdev: fix VLAN offloads set if no relative capabilities

[Jeff Guo]

On 7/6/2020 3:06 PM, Wei Hu (Xavier) wrote:

Currently, there is a potential problem that calling the API function
rte_eth_dev_set_vlan_offload to start VLAN hardware offloads which the
driver does not support. If the PMD driver does not support certain VLAN
hardware offloads and does not check for it, the hardware setting will
not change, but the VLAN offloads in dev->data->dev_conf.rxmode.offloads
will be turned on.

It is supposed to check the hardware capabilities to decide whether the
relative callback needs to be called just like the behavior in the API
function named rte_eth_dev_configure. And it is also needed to cleanup
duplicated checks which are done in some PMDs. Also, note that it is
behaviour change for some PMDs which simply ignore (with error/warning log
message) unsupported VLAN offloads, but now it will fail.

Fixes: a4996bd89c42 ("ethdev: new Rx/Tx offloads API")
Fixes: 0ebce6129bc6 ("net/dpaa2: support new ethdev offload APIs")
Fixes: f9416bbafd98 ("net/enic: remove VLAN filter handler")
Fixes: 4f7d9e383e5c ("fm10k: update vlan offload features")
Fixes: fdba3bf15c7b ("net/hinic: add VLAN filter and offload")
Fixes: b96fb2f0d22b ("net/i40e: handle QinQ strip")
Fixes: d4a27a3b092a ("nfp: add basic features")
Fixes: 56139e85abec ("net/octeontx: support VLAN filter offload")
Fixes: ba1b3b081edf ("net/octeontx2: support VLAN offloads")
Fixes: d87246a43759 ("net/qede: enable and disable VLAN filtering")
Cc: sta...@dpdk.org

Signed-off-by: Chengchang Tang 
Signed-off-by: Wei Hu (Xavier) 
Acked-by: Andrew Rybchenko 
Acked-by: Hyong Youb Kim 
Acked-by: Sachin Saxena 
Acked-by: Xiaoyun wang 
Acked-by: Harman Kalra 



Acked-by: Jeff Guo 



---
v5 -> v6: add the related history patch into the Fixes commit log.
v4 -> v5: no change.
v3 -> v4: Delete "next_mask" label and modify the function that when the
   offload is not supported the function fail.
v2 -> v3: Add __rte_unused to avoid unused parameter 'dev' and 'mask'
   warning.
v1 -> v2: Cleanup duplicated checks which are done in some PMDs.
---
  drivers/net/dpaa2/dpaa2_ethdev.c   | 12 +++-
  drivers/net/enic/enic_ethdev.c | 12 
  drivers/net/fm10k/fm10k_ethdev.c   | 23 ++-
  drivers/net/hinic/hinic_pmd_ethdev.c   |  6 --
  drivers/net/i40e/i40e_ethdev.c |  5 -
  drivers/net/nfp/nfp_net.c  |  5 -
  drivers/net/octeontx/octeontx_ethdev_ops.c | 10 --
  drivers/net/octeontx2/otx2_vlan.c  |  5 -
  drivers/net/qede/qede_ethdev.c |  3 ---
  lib/librte_ethdev/rte_ethdev.c | 21 +
  10 files changed, 26 insertions(+), 76 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index a1f1919..489d744 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -145,7 +145,7 @@ dpaa2_vlan_offload_set(struct rte_eth_dev *dev, int mask)
  {
struct dpaa2_dev_priv *priv = dev->data->dev_private;
struct fsl_mc_io *dpni = dev->process_private;
-   int ret;
+   int ret = 0;
  
  	PMD_INIT_FUNC_TRACE();
  
@@ -153,7 +153,7 @@ dpaa2_vlan_offload_set(struct rte_eth_dev *dev, int mask)

/* VLAN Filter not avaialble */
if (!priv->max_vlan_filters) {
DPAA2_PMD_INFO("VLAN filter not available");
-   goto next_mask;
+   return -ENOTSUP;
}
  
  		if (dev->data->dev_conf.rxmode.offloads &

@@ -166,14 +166,8 @@ dpaa2_vlan_offload_set(struct rte_eth_dev *dev, int mask)
if (ret < 0)
DPAA2_PMD_INFO("Unable to set vlan filter = %d", ret);
}
-next_mask:
-   if (mask & ETH_VLAN_EXTEND_MASK) {
-   if (dev->data->dev_conf.rxmode.offloads &
-   DEV_RX_OFFLOAD_VLAN_EXTEND)
-   DPAA2_PMD_INFO("VLAN extend offload not supported");
-   }
  
-	return 0;

+   return ret;
  }
  
  static int

diff --git a/drivers/net/enic/enic_ethdev.c b/drivers/net/enic/enic_ethdev.c
index 6a3580f..30a599d 100644
--- a/drivers/net/enic/enic_ethdev.c
+++ b/drivers/net/enic/enic_ethdev.c
@@ -367,18 +367,6 @@ static int enicpmd_vlan_offload_set(struct rte_eth_dev 
*eth_dev, int mask)
enic->ig_vlan_strip_en = 0;
}
  
-	if ((mask & ETH_VLAN_FILTER_MASK) &&

-   (offloads & DEV_RX_OFFLOAD_VLAN_FILTER)) {
-   dev_warning(enic,
-   "Configuration of VLAN filter is not supported\n");
-   }
-
-   if ((mask & ETH_VLAN_EXTEND_MASK) &&
-   (offloads & DEV_RX_OFFLOAD_VLAN_EXTEND)) {
-   dev_warning(enic,
-   "Configuration of extended VLAN is not supported\n");
-   }
-
return enic_set_vlan_strip(enic);
  }
  
diff --git a/drivers/net/fm10k/fm10k_ethdev.c b/drivers/net/fm10k/fm10k_ethdev.c

Re: [dpdk-dev] [PATCH 1/3] ring: remove experimental tag for ring reset API

[Kinsella, Ray]

On 07/07/2020 04:19, Feifei Wang wrote:
> 
> 
>> -Original Message-
>> From: Kinsella, Ray 
>> Sent: 2020年7月6日 14:23
>> To: Honnappa Nagarahalli ; Feifei Wang
>> ; Konstantin Ananyev
>> ; Neil Horman 
>> Cc: dev@dpdk.org; nd 
>> Subject: Re: [PATCH 1/3] ring: remove experimental tag for ring reset API
>>
>>
>>
>> On 03/07/2020 19:46, Honnappa Nagarahalli wrote:
>>> 
>>>

 On 03/07/2020 11:26, Feifei Wang wrote:
> Remove the experimental tag for rte_ring_reset API that have been
> around for 4 releases.
>
> Signed-off-by: Feifei Wang 
> Reviewed-by: Honnappa Nagarahalli 
> Reviewed-by: Ruifeng Wang 
> ---
>  lib/librte_ring/rte_ring.h   | 3 ---
>  lib/librte_ring/rte_ring_version.map | 4 +---
>  2 files changed, 1 insertion(+), 6 deletions(-)
>
> diff --git a/lib/librte_ring/rte_ring.h b/lib/librte_ring/rte_ring.h
> index f67141482..7181c33b4 100644
> --- a/lib/librte_ring/rte_ring.h
> +++ b/lib/librte_ring/rte_ring.h
> @@ -663,15 +663,12 @@ rte_ring_dequeue(struct rte_ring *r, void
>> **obj_p)
>   *
>   * This function flush all the elements in a ring
>   *
> - * @b EXPERIMENTAL: this API may change without prior notice
> - *
>   * @warning
>   * Make sure the ring is not in use while calling this function.
>   *
>   * @param r
>   *   A pointer to the ring structure.
>   */
> -__rte_experimental
>  void
>  rte_ring_reset(struct rte_ring *r);
>
> diff --git a/lib/librte_ring/rte_ring_version.map
> b/lib/librte_ring/rte_ring_version.map
> index e88c143cf..aec6f3820 100644
> --- a/lib/librte_ring/rte_ring_version.map
> +++ b/lib/librte_ring/rte_ring_version.map
> @@ -8,6 +8,7 @@ DPDK_20.0 {
>   rte_ring_init;
>   rte_ring_list_dump;
>   rte_ring_lookup;
> + rte_ring_reset;
>
>   local: *;
>  };
> @@ -15,9 +16,6 @@ DPDK_20.0 {
>  EXPERIMENTAL {
>   global:
>
> - # added in 19.08
> - rte_ring_reset;
> -
>   # added in 20.02
>   rte_ring_create_elem;
>   rte_ring_get_memsize_elem;

 So strictly speaking, rte_ring_reset is part of the DPDK_21 ABI, not
 the v20.0 ABI.
>>> Thanks Ray for clarifying this.
>>>
> Thanks very much for pointing this.

 The way to solve is to add it the DPDK_21 ABI in the map file.
 And then use the VERSION_SYMBOL_EXPERIMENTAL to alias to
>> experimental
 if necessary.
>>> Is using VERSION_SYMBOL_EXPERIMENTAL a must?
>>
>> Purely at the discretion of the contributor and maintainer.
>> If it has been around for a while, applications are using it and changing the
>> symbol will break them.
>>
>> You may choose to provide the alias or not.
> Ok, in the new patch version, I will add it into the DPDK_21 ABI but the 
> VERSION_SYMBOL_EXPERIMENTAL will not be added, because if it is added in this
> version, it is still needed to be removed in the near future.
> 
> Thanks very much for your review.

That is 100%

>>
>>> The documentation also seems to be vague. It says " The macro is used
>> when a symbol matures to become part of the stable ABI, to provide an alias
>> to experimental for some time". What does 'some time' mean?
>>
>> "Some time" is a bit vague alright, should be "until the next major ABI
>> version" - I will fix.
>>
>>>

 https://doc.dpdk.org/guides/contributing/abi_versioning.html#versioni
 ng-
 macros

Re: [dpdk-dev] [PATCH v3 1/4] eal: disable function versioning on Windows

[Kinsella, Ray]

On 05/07/2020 21:00, Thomas Monjalon wrote:
> 05/07/2020 11:00, Thomas Monjalon:
>> 05/07/2020 10:51, Fady Bader:
>>> From: Thomas Monjalon 
 05/07/2020 09:00, Fady Bader:
>> On 30/06/2020 11:49, Kinsella, Ray wrote:
>>> On 22/06/2020 12:55, Fady Bader wrote:
 --- a/lib/librte_eal/include/rte_function_versioning.h
 +++ b/lib/librte_eal/include/rte_function_versioning.h
 @@ -11,6 +11,10 @@
  #error Use of function versioning disabled, is
 "use_function_versioning=true"
>> in meson.build?
>>> Do we need a meson error/rule that catches if people explicitly
>>> try to enable
>> function_versioning.
>>> We don't want to just silently fail under such circumstances?
>>>
>>> Would something in lib/meson.build would work better?
>>>
>>> if use_function_versioning and target_machine.system == "Windows"
>>> // complain loudly ...
>>> endif
>>>
>
> I think this would be a better approach.
> But instead of checking if the use of function versioning is enabled
> under Windows and then complain, I think we should simply disable it
> from Windows in the meson.build file. Something like this:
>
> if target_machine.system == "Windows"
>   use_function_versioning = false
> endif

 Did you try?

 If you disable function versioning, compilation will fail with the #error 
 message
 above, right?

>>>
>>> Yes, the compilation fails, we can also change the rte_function_versioning.h
>>> code in order not to fail under Windows.
>>> What do you think ?
>>
>> Function versioning CANNOT be supported currently on Windows.
>> Function versioning macros are MANDATORY in some libraries.
>> That's why I think the only path is to have a specific
>> implementation of the function versioning macros for Windows.
> 
> After a closer look with Fady, the empty macros already exist
> for the static case.
> If disabling function versioning in shared case, the DLL is built
> with the object files compiled for the static library.
> Conclusion: disabling function versioning on Windows works
> (after disabling the error message).
> 

perfect, good work.
 

Re: [dpdk-dev] [PATCH 1/2] eal/windows: Add needed calls to detect vdev PMD

[Tal Shnaiderman]

Thank you for the comments, you're right, either master has changed since I 
sent those patches or I missed out on some exports/includes, anyhow I'll fix 
and resend v2.

The rte_panic you're seeing is unrelated to the new code (BTW, you should build 
a debug version if you want to see rte_panic's backtrace), it is a result of 
your setup not being configured with the memory management requirement of "Lock 
pages" (see [1]).

Dmitry, It looks like we got to this stage since hugepage_claim_privilege() 
cannot actually detect that "Lock pages" isn't granted to the current user, as 
a result we fail on the first usage of a memory management call [in this case 
rte_calloc()] without indication to the reason.

Is it possible to add an actual check that the current user is in the list of 
grantees?

Alternatively, It would be great to have this privilege added programmatically, 
I tried the MSDN example in [2] but it didn't work for me while testing, maybe 
Microsoft team can check if there is a way to do it?

[1] https://doc.dpdk.org/guides/windows_gsg/run_apps.html
[2] 
https://docs.microsoft.com/en-us/windows/win32/memory/creating-a-file-mapping-using-large-pages?redirectedfrom=MSDN

> -Original Message-
> From: Narcisa Ana Maria Vasile 
> Subject: Re: [PATCH 1/2] eal/windows: Add needed calls to detect vdev PMD
> 
> On Wed, Jun 24, 2020 at 05:56:20PM +0300, tal...@mellanox.com wrote:
> > From: Tal Shnaiderman 
> >
> > Add needed function calls in rte_eal_init to detect vdev PMD.
> >
> > eal_option_device_parse()
> > rte_service_init()
> > rte_bus_probe()
> >
> > Signed-off-by: Tal Shnaiderman 
> > ---
> >  lib/librte_eal/common/meson.build |  1 +
> >  lib/librte_eal/windows/eal.c  | 19 +++
> >  2 files changed, 20 insertions(+)
> >
> > diff --git a/lib/librte_eal/common/meson.build
> > b/lib/librte_eal/common/meson.build
> 
> Implicit declaration warning:
> ../lib/librte_eal/windows/eal.c:371:6: warning: implicit declaration of 
> function
> 'rte_service_init' is invalid in C99 [-Wimplicit-function-declaration]
> if (rte_service_init()) {
> ^
> 1 warning generated.
> 
> Adding "#include " in windows/eal.c should
> solve it.
> 
> Getting some linker error:
>Creating library drivers\librte_bus_vdev.dll.a and object
> drivers\librte_bus_vdev.dll.exp bus_vdev_vdev.c.obj : error LNK2019:
> unresolved external symbol rte_log_register_type_and_pick_level
> referenced in function __vdev_logtype_bus drivers\librte_bus_vdev-
> 0.200.3.dll : fatal error LNK1120: 1 unresolved externals
> 
> Probably just a missing export for "rte_log_register_type_and_pick_level" in
> rte_eal_exports.def?
> 
> After adding the missing include and export, compilation and linking are
> successful.
> I see some errors when running the app though:
> 
> EAL: error allocating rte services array
> EAL: FATAL: rte_service_init() failed
> EAL: rte_service_init() failed
> PANIC in main():
> Cannot init EAL
> 6: [ (RtlUserThreadStart+0x21)[0x7FFACF283460]]
> 5: [ (BaseThreadInitThunk+0x14)[0x7FFACE7A6DF0]]
> 4: []
> 3: []
> 2: []
> 1: []

Re: [dpdk-dev] [PATCH v5 15/51] net/bnxt: add HCAPI interface support

[Ferruh Yigit]

On 7/3/2020 10:01 PM, Ajit Khaparde wrote:
> From: Peter Spreadborough 
> 
> Add new hardware shim APIs to support multiple
> device generations
> 
> Signed-off-by: Peter Spreadborough 
> Signed-off-by: Venkat Duvvuru 
> Reviewed-by: Randy Schacher 
> Reviewed-by: Ajit Khaparde 
> ---
>  drivers/net/bnxt/hcapi/Makefile   |  10 +
>  drivers/net/bnxt/hcapi/hcapi_cfa.h| 271 +
>  drivers/net/bnxt/hcapi/hcapi_cfa_common.c |  92 +++
>  drivers/net/bnxt/hcapi/hcapi_cfa_defs.h   | 672 ++
>  drivers/net/bnxt/hcapi/hcapi_cfa_p4.c | 399 +
>  drivers/net/bnxt/hcapi/hcapi_cfa_p4.h | 451 +++
>  drivers/net/bnxt/meson.build  |   2 +
>  drivers/net/bnxt/tf_core/tf_em.c  |  28 +-
>  drivers/net/bnxt/tf_core/tf_tbl.c |  94 +--
>  drivers/net/bnxt/tf_core/tf_tbl.h |  24 +-
>  10 files changed, 1970 insertions(+), 73 deletions(-)
>  create mode 100644 drivers/net/bnxt/hcapi/Makefile
>  create mode 100644 drivers/net/bnxt/hcapi/hcapi_cfa.h
>  create mode 100644 drivers/net/bnxt/hcapi/hcapi_cfa_common.c
>  create mode 100644 drivers/net/bnxt/hcapi/hcapi_cfa_defs.h
>  create mode 100644 drivers/net/bnxt/hcapi/hcapi_cfa_p4.c
>  create mode 100644 drivers/net/bnxt/hcapi/hcapi_cfa_p4.h
> 
> diff --git a/drivers/net/bnxt/hcapi/Makefile b/drivers/net/bnxt/hcapi/Makefile
> new file mode 100644
> index 0..65cddd789
> --- /dev/null
> +++ b/drivers/net/bnxt/hcapi/Makefile
> @@ -0,0 +1,10 @@
> +# SPDX-License-Identifier: BSD-3-Clause
> +# Copyright(c) 2019-2020 Broadcom Limited.
> +# All rights reserved.
> +
> +SRCS-$(CONFIG_RTE_LIBRTE_BNXT_PMD) += hcapi/hcapi_cfa_p4.c
> +
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += hcapi/hcapi_cfa.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += hcapi/hcapi_cfa_defs.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += hcapi/hcapi_cfa_p4.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += hcapi/cfa_p40_hw.h

Why PMD header files added to install?
They should not accessed by the application, so they should not installed.

I will try to remove them.

Re: [dpdk-dev] [PATCH v5 16/51] net/bnxt: add core changes for EM and EEM lookups

[Ferruh Yigit]

On 7/3/2020 10:01 PM, Ajit Khaparde wrote:
> From: Randy Schacher 
> 
> - Move External Exact and Exact Match to device module using HCAPI
>   to add and delete entries
> - Make EM active through the device interface.
> 
> Signed-off-by: Randy Schacher 
> Signed-off-by: Venkat Duvvuru 
> Reviewed-by: Shahaji Bhosle 
> Reviewed-by: Ajit Khaparde 

<...>

> diff --git a/drivers/net/bnxt/tf_core/Makefile 
> b/drivers/net/bnxt/tf_core/Makefile
> index 2c02e29e7..5ed32f12a 100644
> --- a/drivers/net/bnxt/tf_core/Makefile
> +++ b/drivers/net/bnxt/tf_core/Makefile
> @@ -24,3 +24,11 @@ SRCS-$(CONFIG_RTE_LIBRTE_BNXT_PMD) += tf_core/tf_tbl_type.c
>  SRCS-$(CONFIG_RTE_LIBRTE_BNXT_PMD) += tf_core/tf_tcam.c
>  SRCS-$(CONFIG_RTE_LIBRTE_BNXT_PMD) += tf_core/tf_util.c
>  SRCS-$(CONFIG_RTE_LIBRTE_BNXT_PMD) += tf_core/tf_rm_new.c
> +
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_core.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_project.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_device.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_identifier.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_tbl.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/stack.h
> +SYMLINK-$(CONFIG_RTE_LIBRTE_BNXT_PMD)-include += tf_core/tf_tcam.h

Same comment here, these header files should keep PMD local and shouldn't 
installed.
Installing a header file as generic as "stack.h" is breaking some sample apps
but anyway I assume these are added by mistake and I will drop them.

Re: [dpdk-dev] [PATCH v3] net/i40e: support cloud filter with L4 port

[Xing, Beilei]

> -Original Message-
> From: Sun, GuinanX 
> Sent: Tuesday, July 7, 2020 2:34 PM
> To: dev@dpdk.org
> Cc: Xing, Beilei ; Guo, Jia ; Sun,
> GuinanX 
> Subject: [PATCH v3] net/i40e: support cloud filter with L4 port
> 
> This patch enables the filter that supports to create following two rules for
> the same packet type:
> One is to select source port only as input set and the other is for 
> destination
> port only.
> And the supported packet types are UDP/TCP/SCTP.
> 
> Signed-off-by: Guinan Sun 
> ---
> v3:
> * Add limitation to i40e.rst
> * Rename variables.
> * Fixed the variable assignment.
> v2:
> * Fixed code style and variable naming
> ---
>  doc/guides/nics/i40e.rst   |   9 +
>  doc/guides/rel_notes/release_20_08.rst |   8 +
>  drivers/net/i40e/i40e_ethdev.c | 200 +-
>  drivers/net/i40e/i40e_ethdev.h |  18 ++
>  drivers/net/i40e/i40e_flow.c   | 223 +
>  5 files changed, 457 insertions(+), 1 deletion(-)
> 
> diff --git a/doc/guides/nics/i40e.rst b/doc/guides/nics/i40e.rst index
> 00c3042d5..98a5537d7 100644
> --- a/doc/guides/nics/i40e.rst
> +++ b/doc/guides/nics/i40e.rst
> @@ -736,6 +736,15 @@ with DPDK, then the configuration will also impact
> port B in the NIC with  kernel driver, which don't want to use the TPID.
>  So PMD reports warning to clarify what is changed by writing global register.
> 
> +L4 port filter

Please don't use 'L4 port filter', there's no such filter type in i40e pipeline.

> +~~
> +
> +L4 port filter is a new type of cloud filter.
> +It is possible to generate two filters - one to select L4 DST port and
> +second to select L4 SRC port.
> +If the L4 port filter is used, the cloud filter using inner Vlan and
> +Tunnel Key will be invalid.
> +
>  High Performance of Small Packets on 40GbE NIC
>  --
> 
> diff --git a/doc/guides/rel_notes/release_20_08.rst
> b/doc/guides/rel_notes/release_20_08.rst
> index f4b858727..d7d4c4e3b 100644
> --- a/doc/guides/rel_notes/release_20_08.rst
> +++ b/doc/guides/rel_notes/release_20_08.rst
> @@ -108,6 +108,14 @@ New Features
>* Dump ``rte_flow`` memory consumption.
>* Measure packet per second forwarding.
> 
> +* **Updated Intel i40e driver.**
> +
> +  Updated i40e PMD with new features and improvements, including:
> +
> +  * Added a new type of cloud filter to support the coexistence of the
> +following two rules. One selects L4 destination port as input set and
> +the other one selects L4 source port. We call it L4 port filter.
We didn't call it L4 port filter.

> +But there are limitations on using the L4 port filter.
I think no need to mention the limitation here, since you have added in i40e.rst

Re: [dpdk-dev] [PATCH v6 1/2] vhost: introduce async enqueue registration API

[Xia, Chenbo]

> -Original Message-
> From: Fu, Patrick 
> Sent: Tuesday, July 7, 2020 1:07 PM
> To: dev@dpdk.org; maxime.coque...@redhat.com; Xia, Chenbo
> ; Wang, Zhihong 
> Cc: Fu, Patrick ; Wang, Yinan ;
> Jiang, Cheng1 ; Liang, Cunming
> 
> Subject: [PATCH v6 1/2] vhost: introduce async enqueue registration API
> 
> From: Patrick Fu 
> 
> Performing large memory copies usually takes up a major part of CPU cycles and
> becomes the hot spot in vhost-user enqueue operation. To offload the large
> copies from CPU to the DMA devices, asynchronous APIs are introduced, with
> which the CPU just submits copy jobs to the DMA but without waiting for its
> copy completion. Thus, there is no CPU intervention during data transfer. We
> can save precious CPU cycles and improve the overall throughput for vhost-user
> based applications. This patch introduces registration/un-registration APIs 
> for
> vhost async data enqueue operation. Together with the registration APIs
> implementations, data structures and the prototype of the async callback
> functions required for async enqueue data path are also defined.
> 
> Signed-off-by: Patrick Fu 
> ---
>  lib/librte_vhost/Makefile  |   2 +-
>  lib/librte_vhost/meson.build   |   2 +-
>  lib/librte_vhost/rte_vhost.h   |   1 +
>  lib/librte_vhost/rte_vhost_async.h | 136 +
>  lib/librte_vhost/rte_vhost_version.map |   4 +
>  lib/librte_vhost/socket.c  |  27 +
>  lib/librte_vhost/vhost.c   | 127 ++-
>  lib/librte_vhost/vhost.h   |  30 +-
>  lib/librte_vhost/vhost_user.c  |  23 -
>  9 files changed, 345 insertions(+), 7 deletions(-)  create mode 100644
> lib/librte_vhost/rte_vhost_async.h
> 
> diff --git a/lib/librte_vhost/Makefile b/lib/librte_vhost/Makefile index
> b7ff7dc4b..4f2f3e47d 100644
> --- a/lib/librte_vhost/Makefile
> +++ b/lib/librte_vhost/Makefile
> @@ -42,7 +42,7 @@ SRCS-$(CONFIG_RTE_LIBRTE_VHOST) := fd_man.c iotlb.c
> socket.c vhost.c \
> 
>  # install includes
>  SYMLINK-$(CONFIG_RTE_LIBRTE_VHOST)-include += rte_vhost.h rte_vdpa.h \
> - rte_vdpa_dev.h
> + rte_vdpa_dev.h
> rte_vhost_async.h
> 
>  # only compile vhost crypto when cryptodev is enabled  ifeq
> ($(CONFIG_RTE_LIBRTE_CRYPTODEV),y)
> diff --git a/lib/librte_vhost/meson.build b/lib/librte_vhost/meson.build index
> 882a0eaf4..cc9aa65c6 100644
> --- a/lib/librte_vhost/meson.build
> +++ b/lib/librte_vhost/meson.build
> @@ -22,5 +22,5 @@ sources = files('fd_man.c', 'iotlb.c', 'socket.c', 'vdpa.c',
>   'vhost.c', 'vhost_user.c',
>   'virtio_net.c', 'vhost_crypto.c')
>  headers = files('rte_vhost.h', 'rte_vdpa.h', 'rte_vdpa_dev.h',
> - 'rte_vhost_crypto.h')
> + 'rte_vhost_crypto.h', 'rte_vhost_async.h')
>  deps += ['ethdev', 'cryptodev', 'hash', 'pci'] diff --git
> a/lib/librte_vhost/rte_vhost.h b/lib/librte_vhost/rte_vhost.h index
> 8a5c332c8..f93f9595a 100644
> --- a/lib/librte_vhost/rte_vhost.h
> +++ b/lib/librte_vhost/rte_vhost.h
> @@ -35,6 +35,7 @@ extern "C" {
>  #define RTE_VHOST_USER_EXTBUF_SUPPORT(1ULL << 5)
>  /* support only linear buffers (no chained mbufs) */
>  #define RTE_VHOST_USER_LINEARBUF_SUPPORT (1ULL << 6)
> +#define RTE_VHOST_USER_ASYNC_COPY(1ULL << 7)
> 
>  /* Features. */
>  #ifndef VIRTIO_NET_F_GUEST_ANNOUNCE
> diff --git a/lib/librte_vhost/rte_vhost_async.h
> b/lib/librte_vhost/rte_vhost_async.h
> new file mode 100644
> index 0..d5a59279a
> --- /dev/null
> +++ b/lib/librte_vhost/rte_vhost_async.h
> @@ -0,0 +1,136 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2020 Intel Corporation
> + */
> +
> +#ifndef _RTE_VHOST_ASYNC_H_
> +#define _RTE_VHOST_ASYNC_H_
> +
> +#include "rte_vhost.h"
> +
> +/**
> + * iovec iterator
> + */
> +struct rte_vhost_iov_iter {
> + /** offset to the first byte of interesting data */
> + size_t offset;
> + /** total bytes of data in this iterator */
> + size_t count;
> + /** pointer to the iovec array */
> + struct iovec *iov;
> + /** number of iovec in this iterator */
> + unsigned long nr_segs;
> +};
> +
> +/**
> + * dma transfer descriptor pair
> + */
> +struct rte_vhost_async_desc {
> + /** source memory iov_iter */
> + struct rte_vhost_iov_iter *src;
> + /** destination memory iov_iter */
> + struct rte_vhost_iov_iter *dst;
> +};
> +
> +/**
> + * dma transfer status
> + */
> +struct rte_vhost_async_status {
> + /** An array of application specific data for source memory */
> + uintptr_t *src_opaque_data;
> + /** An array of application specific data for destination memory */
> + uintptr_t *dst_opaque_data;
> +};
> +
> +/**
> + * dma operation callbacks to be implemented by applications  */ struct
> +rte_vhost_async_channel_ops {
> + /**
> +  * instruct async engines to perform copies 

Re: [dpdk-dev] [PATCH v6 2/2] vhost: introduce async enqueue for split ring

[Xia, Chenbo]

> -Original Message-
> From: Fu, Patrick 
> Sent: Tuesday, July 7, 2020 1:07 PM
> To: dev@dpdk.org; maxime.coque...@redhat.com; Xia, Chenbo
> ; Wang, Zhihong 
> Cc: Fu, Patrick ; Wang, Yinan ;
> Jiang, Cheng1 ; Liang, Cunming
> 
> Subject: [PATCH v6 2/2] vhost: introduce async enqueue for split ring
> 
> From: Patrick Fu 
> 
> This patch implements async enqueue data path for split ring. 2 new async data
> path APIs are defined, by which applications can submit and poll packets 
> to/from
> async engines. The async engine is either a physical DMA device or it could 
> also
> be a software emulated backend.
> The async enqueue data path leverages callback functions registered by
> applications to work with the async engine.
> 
> Signed-off-by: Patrick Fu 
> ---
>  lib/librte_vhost/rte_vhost_async.h |  40 +++
>  lib/librte_vhost/virtio_net.c  | 551 -
>  2 files changed, 589 insertions(+), 2 deletions(-)
> 
> diff --git a/lib/librte_vhost/rte_vhost_async.h
> b/lib/librte_vhost/rte_vhost_async.h
> index d5a59279a..c8ad8dbc7 100644
> --- a/lib/librte_vhost/rte_vhost_async.h
> +++ b/lib/librte_vhost/rte_vhost_async.h
> @@ -133,4 +133,44 @@ int rte_vhost_async_channel_register(int vid, uint16_t
> queue_id,  __rte_experimental  int rte_vhost_async_channel_unregister(int vid,
> uint16_t queue_id);
> 
> +/**
> + * This function submit enqueue data to async engine. This function has
> + * no guranttee to the transfer completion upon return. Applications
> + * should poll transfer status by rte_vhost_poll_enqueue_completed()
> + *
> + * @param vid
> + *  id of vhost device to enqueue data
> + * @param queue_id
> + *  queue id to enqueue data
> + * @param pkts
> + *  array of packets to be enqueued
> + * @param count
> + *  packets num to be enqueued
> + * @return
> + *  num of packets enqueued
> + */
> +__rte_experimental
> +uint16_t rte_vhost_submit_enqueue_burst(int vid, uint16_t queue_id,
> + struct rte_mbuf **pkts, uint16_t count);
> +
> +/**
> + * This function check async completion status for a specific vhost
> + * device queue. Packets which finish copying (enqueue) operation
> + * will be returned in an array.
> + *
> + * @param vid
> + *  id of vhost device to enqueue data
> + * @param queue_id
> + *  queue id to enqueue data
> + * @param pkts
> + *  blank array to get return packet pointer
> + * @param count
> + *  size of the packet array
> + * @return
> + *  num of packets returned
> + */
> +__rte_experimental
> +uint16_t rte_vhost_poll_enqueue_completed(int vid, uint16_t queue_id,
> + struct rte_mbuf **pkts, uint16_t count);
> +
>  #endif /* _RTE_VHOST_ASYNC_H_ */
> diff --git a/lib/librte_vhost/virtio_net.c b/lib/librte_vhost/virtio_net.c 
> index
> 751c1f373..236498f71 100644
> --- a/lib/librte_vhost/virtio_net.c
> +++ b/lib/librte_vhost/virtio_net.c
> @@ -17,14 +17,15 @@
>  #include 
>  #include 
>  #include 
> +#include 
> 
>  #include "iotlb.h"
>  #include "vhost.h"
> 
> -#define MAX_PKT_BURST 32
> -
>  #define MAX_BATCH_LEN 256
> 
> +#define VHOST_ASYNC_BATCH_THRESHOLD 32
> +
>  static  __rte_always_inline bool
>  rxvq_is_mergeable(struct virtio_net *dev)  { @@ -116,6 +117,31 @@
> flush_shadow_used_ring_split(struct virtio_net *dev, struct vhost_virtqueue 
> *vq)
>   sizeof(vq->used->idx));
>  }
> 
> +static __rte_always_inline void
> +async_flush_shadow_used_ring_split(struct virtio_net *dev,
> + struct vhost_virtqueue *vq)
> +{
> + uint16_t used_idx = vq->last_used_idx & (vq->size - 1);
> +
> + if (used_idx + vq->shadow_used_idx <= vq->size) {
> + do_flush_shadow_used_ring_split(dev, vq, used_idx, 0,
> +   vq->shadow_used_idx);
> + } else {
> + uint16_t size;
> +
> + /* update used ring interval [used_idx, vq->size] */
> + size = vq->size - used_idx;
> + do_flush_shadow_used_ring_split(dev, vq, used_idx, 0, size);
> +
> + /* update the left half used ring interval [0, left_size] */
> + do_flush_shadow_used_ring_split(dev, vq, 0, size,
> +   vq->shadow_used_idx - size);
> + }
> +
> + vq->last_used_idx += vq->shadow_used_idx;
> + vq->shadow_used_idx = 0;
> +}
> +
>  static __rte_always_inline void
>  update_shadow_used_ring_split(struct vhost_virtqueue *vq,
>uint16_t desc_idx, uint32_t len)
> @@ -905,6 +931,200 @@ copy_mbuf_to_desc(struct virtio_net *dev, struct
> vhost_virtqueue *vq,
>   return error;
>  }
> 
> +static __rte_always_inline void
> +async_fill_vec(struct iovec *v, void *base, size_t len) {
> + v->iov_base = base;
> + v->iov_len = len;
> +}
> +
> +static __rte_always_inline void
> +async_fill_iter(struct rte_vhost_iov_iter *it, size_t count,
> + struct iovec *vec, unsigned long nr_seg) {
> + it->offset = 0;
> + it->count = count;
> +
> + if (count) {
> + it->iov = 

Re: [dpdk-dev] [PATCH v4 2/3] doc: update armv8-a IO barrier changes

[David Marchand]

On Tue, Jul 7, 2020 at 1:44 AM Honnappa Nagarahalli
 wrote:
>
> Updated the use of DMB instruction in rte_*mb APIs for Armv8-a.
>
> Signed-off-by: Honnappa Nagarahalli 
> Acked-by: Jerin Jacob 
> ---
>  doc/guides/rel_notes/release_20_08.rst | 7 +++
>  1 file changed, 7 insertions(+)
>
> diff --git a/doc/guides/rel_notes/release_20_08.rst 
> b/doc/guides/rel_notes/release_20_08.rst
> index 5cbc4ce14..567ae6b2a 100644
> --- a/doc/guides/rel_notes/release_20_08.rst
> +++ b/doc/guides/rel_notes/release_20_08.rst
> @@ -56,6 +56,13 @@ New Features
>   Also, make sure to start the actual text at the margin.
>   =
>

This release note update will be squashed with the change itself in patch 1.

> +* **rte_*mb APIs are updated to use DMB instruction for Armv8-a.**

We use "ARMv8" in the release notes, any objection if I update this
when applying?


> +
> +  Armv8-a memory model has been strengthened to require other-multi-copy
> +  atomicity. This allows for using DMB instruction instead of DSB for IO
> +  barriers. rte_*mb APIs, for Armv8-a platforms, are changed to use DMB
> +  instruction to reflect this.
> +
>  * **Updated PCAP driver.**
>
>Updated PCAP driver with new features and improvements, including:
> --
> 2.17.1
>


-- 
David Marchand

Re: [dpdk-dev] [PATCH v4 3/3] doc: update deprecation of CIO barrier APIs

[David Marchand]

On Tue, Jul 7, 2020 at 1:44 AM Honnappa Nagarahalli
 wrote:
>
> rte_cio_*mb APIs will be deprecated in 20.11 release.
>
> Signed-off-by: Honnappa Nagarahalli 
> Acked-by: Jerin Jacob 
> ---
>  doc/guides/rel_notes/deprecation.rst | 6 ++
>  1 file changed, 6 insertions(+)
>
> diff --git a/doc/guides/rel_notes/deprecation.rst 
> b/doc/guides/rel_notes/deprecation.rst
> index d1034f60f..59656da3d 100644
> --- a/doc/guides/rel_notes/deprecation.rst
> +++ b/doc/guides/rel_notes/deprecation.rst
> @@ -40,6 +40,12 @@ Deprecation Notices
>These wrappers must be used for patches that need to be merged in 20.08
>onwards. This change will not introduce any performance degradation.
>
> +* rte_cio_*mb: Since the IO barriers for ArmV8-a platforms are relaxed from 
> DSB
> +  to DMB, rte_cio_*mb APIs provide the same functionality as rte_io_*mb
> +  APIs(taking all platforms into consideration). rte_io_*mb APIs should be 
> used

Nit: missing space.


> +  in the place of rte_cio_*mb APIs. The rte_cio_*mb APIs will be deprecated 
> in
> +  20.11 release.
> +
>  * igb_uio: In the view of reducing the kernel dependency from the main tree,
>as a first step, the Technical Board decided to move ``igb_uio``
>kernel module to the dpdk-kmods repository in the /linux/igb_uio/ directory
> --
> 2.17.1
>

LGTM.

We need 3 acks (ideally from different vendors/companies) for a
deprecation notice.
Please maintainers?


Thanks.

-- 
David Marchand

Re: [dpdk-dev] [PATCH 1/2] eal/windows: Add needed calls to detect vdev PMD

[Dmitry Kozlyuk]

On Tue, 7 Jul 2020 08:04:00 +, Tal Shnaiderman wrote:
> Dmitry, It looks like we got to this stage since hugepage_claim_privilege() 
> cannot actually detect that "Lock pages" isn't granted to the current user, 
> as a result we fail on the first usage of a memory management call [in this 
> case rte_calloc()] without indication to the reason.
> 
> Is it possible to add an actual check that the current user is in the list of 
> grantees?

Thanks, I'll look into it.
 
> Alternatively, It would be great to have this privilege added 
> programmatically, I tried the MSDN example in [2] but it didn't work for me 
> while testing, maybe Microsoft team can check if there is a way to do it?

I don't think it's a good idea from security perspective if an application
grants its user new privileges implicitly. Process with SeLockMemory
privilege can affect overall system performance and stability.

-- 
Dmitry Kozlyuk

[dpdk-dev] [PATCH v2 2/2] bus/vdev: build on Windows

[talshn]

From: Tal Shnaiderman 

current support will build vdev with empty MP functions
currently unsupported for Windows.

Signed-off-by: Tal Shnaiderman 
---
 drivers/bus/vdev/meson.build   | 6 --
 lib/librte_eal/rte_eal_exports.def | 5 +
 2 files changed, 5 insertions(+), 6 deletions(-)

diff --git a/drivers/bus/vdev/meson.build b/drivers/bus/vdev/meson.build
index 3a662b1075..967d54e4f8 100644
--- a/drivers/bus/vdev/meson.build
+++ b/drivers/bus/vdev/meson.build
@@ -1,12 +1,6 @@
 # SPDX-License-Identifier: BSD-3-Clause
 # Copyright(c) 2017 Intel Corporation
 
-if is_windows
-   build = false
-   reason = 'not supported on Windows'
-   subdir_done()
-endif
-
 sources = files('vdev.c',
'vdev_params.c')
 install_headers('rte_bus_vdev.h')
diff --git a/lib/librte_eal/rte_eal_exports.def 
b/lib/librte_eal/rte_eal_exports.def
index 69204a92c6..f54ed74a58 100644
--- a/lib/librte_eal/rte_eal_exports.def
+++ b/lib/librte_eal/rte_eal_exports.def
@@ -11,6 +11,7 @@ EXPORTS
rte_calloc_socket
rte_bus_register
rte_dev_is_probed
+   rte_devargs_insert
rte_devargs_next
rte_devargs_remove
rte_eal_get_configuration
@@ -29,6 +30,9 @@ EXPORTS
rte_eal_using_phys_addrs
rte_free
rte_hexdump
+   rte_log_register
+   rte_log_register_type_and_pick_level
+   rte_log_set_level
rte_malloc
rte_malloc_dump_stats
rte_malloc_get_socket_stats
@@ -62,6 +66,7 @@ EXPORTS
rte_memzone_walk
rte_strerror
rte_strsplit
+   rte_sys_gettid
rte_vfio_container_dma_map
rte_vfio_container_dma_unmap
rte_vlog
-- 
2.16.1.windows.4

[dpdk-dev] [PATCH v2 1/2] eal/windows: add needed calls to detect vdev PMD

[talshn]

From: Tal Shnaiderman 

Add needed function calls in rte_eal_init to detect vdev PMD.

eal_option_device_parse()
rte_service_init()
rte_bus_probe()

Signed-off-by: Tal Shnaiderman 
---
 lib/librte_eal/common/meson.build |  1 +
 lib/librte_eal/windows/eal.c  | 20 
 2 files changed, 21 insertions(+)

diff --git a/lib/librte_eal/common/meson.build 
b/lib/librte_eal/common/meson.build
index c1d9f21488..9f32262450 100644
--- a/lib/librte_eal/common/meson.build
+++ b/lib/librte_eal/common/meson.build
@@ -31,6 +31,7 @@ if is_windows
'malloc_heap.c',
'rte_malloc.c',
'eal_common_timer.c',
+   'rte_service.c',
)
subdir_done()
 endif
diff --git a/lib/librte_eal/windows/eal.c b/lib/librte_eal/windows/eal.c
index eb10b4ef96..cbbf67e253 100644
--- a/lib/librte_eal/windows/eal.c
+++ b/lib/librte_eal/windows/eal.c
@@ -17,6 +17,7 @@
 #include 
 #include 
 #include 
+#include 
 #include 
 #include 
 
@@ -273,6 +274,11 @@ rte_eal_init(int argc, char **argv)
if (fctret < 0)
exit(1);
 
+   if (eal_option_device_parse()) {
+   rte_errno = ENODEV;
+   return -1;
+   }
+
/* Prevent creation of shared memory files. */
if (internal_conf->in_memory == 0) {
RTE_LOG(WARNING, EAL, "Multi-process support is requested, "
@@ -362,6 +368,20 @@ rte_eal_init(int argc, char **argv)
rte_panic("Cannot create thread\n");
}
 
+   /* initialize services so vdevs register service during bus_probe. */
+   if (rte_service_init()) {
+   rte_eal_init_alert("rte_service_init() failed");
+   rte_errno = ENOEXEC;
+   return -1;
+   }
+
+   /* Probe all the buses and devices/drivers on them */
+   if (rte_bus_probe()) {
+   rte_eal_init_alert("Cannot probe devices");
+   rte_errno = ENOTSUP;
+   return -1;
+   }
+
/*
 * Launch a dummy function on all slave lcores, so that master lcore
 * knows they are all ready when this function returns.
-- 
2.16.1.windows.4

[dpdk-dev] [PATCH v2 0/2] Windows bus/vdev support

[talshn]

From: Tal Shnaiderman 

This patchset implements the EAL functions needed for probing PMDs using vdev 
on Windows.

---
v2: Added missing exports and includes (NarcisaV)
---

Tal Shnaiderman (2):
  eal/windows: add needed calls to detect vdev PMD
  bus/vdev: build on Windows

 drivers/bus/vdev/meson.build   |  6 --
 lib/librte_eal/common/meson.build  |  1 +
 lib/librte_eal/rte_eal_exports.def |  5 +
 lib/librte_eal/windows/eal.c   | 20 
 4 files changed, 26 insertions(+), 6 deletions(-)

-- 
2.16.1.windows.4

[dpdk-dev] [PATCH] devtools: add checkpatch spelling dictionary builder

[Thomas Monjalon]

The script checkpatch.pl (used in checkpatches.sh) can use a dictionary
from the codespell project to check spelling.
There are multiple dictionaries to be used.

The script build-dict.sh concatenate multiple dictionaries and remove
some annoying false positives.

The dictionary built by this script must be saved in a file which
is referenced with the environment variable DPDK_CHECKPATCH_CODESPELL.
The easiest is to export this variable in ~/.config/dpdk/devel.config.

Signed-off-by: Thomas Monjalon 
---
 devtools/build-dict.sh | 22 ++
 1 file changed, 22 insertions(+)
 create mode 100755 devtools/build-dict.sh

diff --git a/devtools/build-dict.sh b/devtools/build-dict.sh
new file mode 100755
index 00..a8cac49029
--- /dev/null
+++ b/devtools/build-dict.sh
@@ -0,0 +1,22 @@
+#! /bin/sh -e
+# SPDX-License-Identifier: BSD-3-Clause
+# Copyright 2020 Mellanox Technologies, Ltd
+
+# Build a spelling dictionary suitable for DPDK_CHECKPATCH_CODESPELL
+
+# path to local clone of https://github.com/codespell-project/codespell.git
+codespell_path=$1
+
+# concatenate codespell dictionaries, except GB/US one
+for suffix in .txt _code.txt _informal.txt _names.txt _rare.txt _usage.txt ; do
+   cat $codespell_path/codespell_lib/data/dictionary$suffix
+done |
+
+# remove too short or wrong checks
+sed '/^..->/d' |
+sed '/^uint->/d' |
+sed "/^doesn'->/d" |
+sed '/^wasn->/d' |
+
+# print to stdout
+cat
-- 
2.27.0

[dpdk-dev] [PATCH v2] rawdev: fix to remove EXPERIMENTAL from comment

[Hemant Agrawal]

The experimental tags were removed, but the comment
is still having API classification as EXPERIMENTAL

Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
Cc: sta...@dpdk.org

Signed-off-by: Hemant Agrawal 
---
 MAINTAINERS| 2 +-
 lib/librte_rawdev/rte_rawdev.h | 3 ---
 lib/librte_rawdev/rte_rawdev_pmd.h | 3 ---
 3 files changed, 1 insertion(+), 7 deletions(-)

diff --git a/MAINTAINERS b/MAINTAINERS
index 27fcd6d648..6ba02d73aa 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -490,7 +490,7 @@ F: lib/librte_eventdev/*crypto_adapter*
 F: app/test/test_event_crypto_adapter.c
 F: doc/guides/prog_guide/event_crypto_adapter.rst
 
-Raw device API - EXPERIMENTAL
+Raw device API
 M: Nipun Gupta 
 M: Hemant Agrawal 
 F: lib/librte_rawdev/
diff --git a/lib/librte_rawdev/rte_rawdev.h b/lib/librte_rawdev/rte_rawdev.h
index ed011ca228..7a1f3213e9 100644
--- a/lib/librte_rawdev/rte_rawdev.h
+++ b/lib/librte_rawdev/rte_rawdev.h
@@ -12,9 +12,6 @@
  *
  * This API allow applications to configure and use generic devices having
  * no specific type already available in DPDK.
- *
- * @warning
- * @b EXPERIMENTAL: this API may change without prior notice
  */
 
 #ifdef __cplusplus
diff --git a/lib/librte_rawdev/rte_rawdev_pmd.h 
b/lib/librte_rawdev/rte_rawdev_pmd.h
index cb3555ab50..4395a2182d 100644
--- a/lib/librte_rawdev/rte_rawdev_pmd.h
+++ b/lib/librte_rawdev/rte_rawdev_pmd.h
@@ -11,9 +11,6 @@
  * @note
  * Driver facing APIs for a raw device. These are not to be called directly by
  * any application.
- *
- * @warning
- * @b EXPERIMENTAL: this API may change without prior notice
  */
 
 #ifdef __cplusplus
-- 
2.17.1

Re: [dpdk-dev] [PATCH v2] rawdev: fix to remove EXPERIMENTAL from comment

[Thomas Monjalon]

07/07/2020 10:54, Hemant Agrawal:
> The experimental tags were removed, but the comment
> is still having API classification as EXPERIMENTAL
> 
> Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
> Cc: sta...@dpdk.org
> 
> Signed-off-by: Hemant Agrawal 
> ---
>  MAINTAINERS| 2 +-
>  lib/librte_rawdev/rte_rawdev.h | 3 ---
>  lib/librte_rawdev/rte_rawdev_pmd.h | 3 ---
>  3 files changed, 1 insertion(+), 7 deletions(-)

Acked-by: Thomas Monjalon 

[dpdk-dev] [PATCH] lib/librte_timer:fix corruption with reset

[Sarosh Arif]

If the user tries to reset/stop some other timer in it's callback
function, which is also about to expire, using 
rte_timer_reset_sync/rte_timer_stop_sync the application goes into
an infinite loop. This happens because 
rte_timer_reset_sync/rte_timer_stop_sync loop until the timer 
resets/stops and there is check inside timer_set_config_state which
prevents a running timer from being reset/stopped by not it's own 
timer_cb. Therefore timer_set_config_state returns -1 due to which 
rte_timer_reset returns -1 and rte_timer_reset_sync goes into an 
infinite loop. 

The soloution to this problem is to return -1 from 
rte_timer_reset_sync/rte_timer_stop_sync in case the user tries to 
reset/stop some other timer in it's callback function.

Bugzilla ID: 491
Fixes: 20d159f20543 ("timer: fix corruption with reset")
Cc: h.mikit...@gmail.com
Signed-off-by: Sarosh Arif 
---
 lib/librte_timer/rte_timer.c | 25 +++--
 lib/librte_timer/rte_timer.h |  4 ++--
 2 files changed, 25 insertions(+), 4 deletions(-)

diff --git a/lib/librte_timer/rte_timer.c b/lib/librte_timer/rte_timer.c
index 6d19ce469..ed48bcbfd 100644
--- a/lib/librte_timer/rte_timer.c
+++ b/lib/librte_timer/rte_timer.c
@@ -576,14 +576,24 @@ rte_timer_alt_reset(uint32_t timer_data_id, struct 
rte_timer *tim,
 }
 
 /* loop until rte_timer_reset() succeed */
-void
+int
 rte_timer_reset_sync(struct rte_timer *tim, uint64_t ticks,
 enum rte_timer_type type, unsigned tim_lcore,
 rte_timer_cb_t fct, void *arg)
 {
+   struct rte_timer_data *timer_data;
+   TIMER_DATA_VALID_GET_OR_ERR_RET(default_data_id, timer_data, -EINVAL);
+
+   if (tim->status.state == RTE_TIMER_RUNNING &&   \
+   (tim->status.owner != (uint16_t)tim_lcore ||\
+   tim != timer_data->priv_timer[tim_lcore].running_tim))
+   return -1;
+
while (rte_timer_reset(tim, ticks, type, tim_lcore,
   fct, arg) != 0)
rte_pause();
+
+   return 0;
 }
 
 static int
@@ -642,11 +652,22 @@ rte_timer_alt_stop(uint32_t timer_data_id, struct 
rte_timer *tim)
 }
 
 /* loop until rte_timer_stop() succeed */
-void
+int
 rte_timer_stop_sync(struct rte_timer *tim)
 {
+   struct rte_timer_data *timer_data;
+   TIMER_DATA_VALID_GET_OR_ERR_RET(default_data_id, timer_data, -EINVAL);
+   unsigned int lcore_id = rte_lcore_id();
+
+   if (tim->status.state == RTE_TIMER_RUNNING &&   \
+   (tim->status.owner != (uint16_t)lcore_id || \
+   tim != timer_data->priv_timer[lcore_id].running_tim))
+   return -1;
+
while (rte_timer_stop(tim) != 0)
rte_pause();
+
+   return 0;
 }
 
 /* Test the PENDING status of the timer handle tim */
diff --git a/lib/librte_timer/rte_timer.h b/lib/librte_timer/rte_timer.h
index c6b3d450d..392ca423d 100644
--- a/lib/librte_timer/rte_timer.h
+++ b/lib/librte_timer/rte_timer.h
@@ -275,7 +275,7 @@ int rte_timer_reset(struct rte_timer *tim, uint64_t ticks,
  * @param arg
  *   The user argument of the callback function.
  */
-void
+int
 rte_timer_reset_sync(struct rte_timer *tim, uint64_t ticks,
 enum rte_timer_type type, unsigned tim_lcore,
 rte_timer_cb_t fct, void *arg);
@@ -314,7 +314,7 @@ int rte_timer_stop(struct rte_timer *tim);
  * @param tim
  *   The timer handle.
  */
-void rte_timer_stop_sync(struct rte_timer *tim);
+int rte_timer_stop_sync(struct rte_timer *tim);
 
 /**
  * Test if a timer is pending.
-- 
2.17.1

Re: [dpdk-dev] [PATCH v2] rawdev: fix to remove EXPERIMENTAL from comment

[Ferruh Yigit]

On 7/7/2020 10:02 AM, Thomas Monjalon wrote:
> 07/07/2020 10:54, Hemant Agrawal:
>> The experimental tags were removed, but the comment
>> is still having API classification as EXPERIMENTAL
>>
>> Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
>> Cc: sta...@dpdk.org
>>
>> Signed-off-by: Hemant Agrawal 
>> ---
>>  MAINTAINERS| 2 +-
>>  lib/librte_rawdev/rte_rawdev.h | 3 ---
>>  lib/librte_rawdev/rte_rawdev_pmd.h | 3 ---
>>  3 files changed, 1 insertion(+), 7 deletions(-)
> 
> Acked-by: Thomas Monjalon 
> 

It seems the APIs are already part of stable API, +1 to update the documentation
to match it.

Acked-by: Ferruh Yigit 

Re: [dpdk-dev] [PATCH v2] rawdev: fix to remove EXPERIMENTAL from comment

[David Marchand]

On Tue, Jul 7, 2020 at 10:58 AM Hemant Agrawal  wrote:
>
> The experimental tags were removed, but the comment
> is still having API classification as EXPERIMENTAL
>
> Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Hemant Agrawal 
> ---
>  MAINTAINERS| 2 +-
>  lib/librte_rawdev/rte_rawdev.h | 3 ---
>  lib/librte_rawdev/rte_rawdev_pmd.h | 3 ---
>  3 files changed, 1 insertion(+), 7 deletions(-)
>
> diff --git a/MAINTAINERS b/MAINTAINERS
> index 27fcd6d648..6ba02d73aa 100644
> --- a/MAINTAINERS
> +++ b/MAINTAINERS
> @@ -490,7 +490,7 @@ F: lib/librte_eventdev/*crypto_adapter*
>  F: app/test/test_event_crypto_adapter.c
>  F: doc/guides/prog_guide/event_crypto_adapter.rst
>
> -Raw device API - EXPERIMENTAL
> +Raw device API
>  M: Nipun Gupta 
>  M: Hemant Agrawal 
>  F: lib/librte_rawdev/
> diff --git a/lib/librte_rawdev/rte_rawdev.h b/lib/librte_rawdev/rte_rawdev.h
> index ed011ca228..7a1f3213e9 100644
> --- a/lib/librte_rawdev/rte_rawdev.h
> +++ b/lib/librte_rawdev/rte_rawdev.h
> @@ -12,9 +12,6 @@
>   *
>   * This API allow applications to configure and use generic devices having
>   * no specific type already available in DPDK.
> - *
> - * @warning
> - * @b EXPERIMENTAL: this API may change without prior notice
>   */
>
>  #ifdef __cplusplus
> diff --git a/lib/librte_rawdev/rte_rawdev_pmd.h 
> b/lib/librte_rawdev/rte_rawdev_pmd.h
> index cb3555ab50..4395a2182d 100644
> --- a/lib/librte_rawdev/rte_rawdev_pmd.h
> +++ b/lib/librte_rawdev/rte_rawdev_pmd.h
> @@ -11,9 +11,6 @@
>   * @note
>   * Driver facing APIs for a raw device. These are not to be called directly 
> by
>   * any application.
> - *
> - * @warning
> - * @b EXPERIMENTAL: this API may change without prior notice
>   */
>
>  #ifdef __cplusplus
> --
> 2.17.1
>

Acked-by: David Marchand 


-- 
David Marchand

Re: [dpdk-dev] [dpdk-dev v5 0/3] enable new hash flow for VF

[Zhang, Qi Z]

> -Original Message-
> From: Guo, Jia 
> Sent: Tuesday, July 7, 2020 1:14 PM
> To: Zhang, Qi Z ; Wu, Jingjing
> ; Xing, Beilei 
> Cc: dev@dpdk.org; Guo, Junfeng ; Yang, Qiming
> ; Su, Simei ; Guo, Jia
> 
> Subject: [dpdk-dev v5 0/3] enable new hash flow for VF
> 
> Some new hash flow will be supported to expend the flow hash capability,
> the input set are the 5 tuple for regular ip pattern and also GTPU inner ip
> pattern, and the session id for NAT-T ESP protocol, the l3 src/dst and the 
> teid
> for GTPU_IP protocol.
> 
> v5->v4:
> refine the lookup table for eth/vlan
> fix issue after splitting patch
> 
> v4->v3:
> split patch set
> 
> Jeff Guo (3):
>   net/iavf: refactor for hash flow
>   net/iavf: enable 5 tuple rss hash
>   net/iavf: enable some new hash flow
> 
>  drivers/net/iavf/iavf_hash.c | 2075 +-
>  1 file changed, 1560 insertions(+), 515 deletions(-)
> 
> --
> 2.20.1

Acked-by: Qi Zhang 

Applied to dpdk-next-net-intel.

Thanks
Qi

[dpdk-dev] [PATCH v2 03/29] net/dpaa2: enable timestamp for Rx offload case as well

[Hemant Agrawal]

From: Gagandeep Singh 

This patch enables the packet timestamping
conditionally when Rx offload is enabled for timestamp.

Signed-off-by: Gagandeep Singh 
---
 drivers/net/dpaa2/dpaa2_ethdev.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index a1f19194d..8edd4b3cd 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -524,8 +524,10 @@ dpaa2_eth_dev_configure(struct rte_eth_dev *dev)
return ret;
}
 
+#if !defined(RTE_LIBRTE_IEEE1588)
if (rx_offloads & DEV_RX_OFFLOAD_TIMESTAMP)
-   dpaa2_enable_ts = true;
+#endif
+   dpaa2_enable_ts = true;
 
if (tx_offloads & DEV_TX_OFFLOAD_IPV4_CKSUM)
tx_l3_csum_offload = true;
-- 
2.17.1

[dpdk-dev] [PATCH v2 00/29] NXP DPAAx enhancements

[Hemant Agrawal]

v2: dropping the fmlib changes - we will send them separately

This patch-set mainly address following enhancements

1. Supporting the non-EAL thread based I/O processing
2. Reducing the thread local storage
3. DPAA2 flow support
4. other minor fixes and enhancements

Gagandeep Singh (3):
  net/dpaa2: enable timestamp for Rx offload case as well
  bus/fslmc: combine thread specific variables
  net/dpaa: enable Tx queue taildrop

Hemant Agrawal (1):
  bus/fslmc: support handle portal alloc failure

Jun Yang (14):
  net/dpaa2: support dynamic flow control
  net/dpaa2: support key extracts of flow API
  net/dpaa2: add sanity check for flow extracts
  net/dpaa2: free flow rule memory
  net/dpaa2: support QoS or FS table entry indexing
  net/dpaa2: define the size of table entry
  net/dpaa2: add logging of flow extracts and rules
  net/dpaa2: support iscrimination between IPv4 and IPv6
  net/dpaa2: support distribution size set on multiple TCs
  net/dpaa2: support ndex of queue action for flow
  net/dpaa2: add flow data sanity check
  net/dpaa2: modify flow API QoS setup to follow FS setup
  net/dpaa2: support flow API FS miss action configuration
  net/dpaa2: configure per class distribution size

Nipun Gupta (7):
  bus/fslmc: fix getting the FD error
  net/dpaa: fix fd offset data type
  bus/fslmc: rework portal allocation to a per thread basis
  bus/fslmc: support portal migration
  bus/fslmc: rename the cinh read functions used for ls1088
  net/dpaa: update process specific device info
  net/dpaa2: support raw flow classification

Rohit Raj (3):
  drivers: optimize thread local storage for dpaa
  bus/dpaa: enable link state interrupt
  bus/dpaa: enable set link status

Sachin Saxena (1):
  net/dpaa: add 2.5G support

 doc/guides/nics/features/dpaa.ini |3 +-
 doc/guides/nics/features/dpaa2.ini|1 +
 doc/guides/rel_notes/release_20_08.rst|   13 +
 drivers/bus/dpaa/base/fman/fman.c |   10 +-
 drivers/bus/dpaa/base/fman/netcfg_layer.c |3 +-
 drivers/bus/dpaa/base/qbman/process.c |   99 +-
 drivers/bus/dpaa/base/qbman/qman.c|   43 +
 drivers/bus/dpaa/dpaa_bus.c   |   52 +-
 drivers/bus/dpaa/include/fman.h   |3 +
 drivers/bus/dpaa/include/fsl_qman.h   |   17 +
 drivers/bus/dpaa/include/process.h|   31 +
 drivers/bus/dpaa/rte_bus_dpaa_version.map |7 +-
 drivers/bus/dpaa/rte_dpaa_bus.h   |   48 +-
 drivers/bus/fslmc/Makefile|1 +
 drivers/bus/fslmc/fslmc_bus.c |2 -
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.c  |  284 +-
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.h  |   10 +-
 drivers/bus/fslmc/portal/dpaa2_hw_pvt.h   |   10 +-
 .../bus/fslmc/qbman/include/fsl_qbman_debug.h |1 +
 .../fslmc/qbman/include/fsl_qbman_portal.h|8 +-
 drivers/bus/fslmc/qbman/qbman_portal.c|  580 +-
 drivers/bus/fslmc/qbman/qbman_portal.h|   19 +-
 drivers/bus/fslmc/qbman/qbman_sys.h   |  135 +-
 drivers/bus/fslmc/rte_bus_fslmc_version.map   |1 -
 drivers/bus/fslmc/rte_fslmc.h |   18 -
 drivers/common/dpaax/compat.h |5 +-
 drivers/crypto/dpaa_sec/dpaa_sec.c|   11 +-
 drivers/event/dpaa/dpaa_eventdev.c|4 +-
 drivers/mempool/dpaa/dpaa_mempool.c   |6 +-
 drivers/net/dpaa/dpaa_ethdev.c|  431 +-
 drivers/net/dpaa/dpaa_ethdev.h|2 +-
 drivers/net/dpaa/dpaa_rxtx.c  |   77 +-
 drivers/net/dpaa/dpaa_rxtx.h  |3 +
 drivers/net/dpaa2/base/dpaa2_hw_dpni.c|   50 +-
 drivers/net/dpaa2/dpaa2_ethdev.c  |   95 +-
 drivers/net/dpaa2/dpaa2_ethdev.h  |   49 +-
 drivers/net/dpaa2/dpaa2_flow.c| 4767 -
 37 files changed, 5141 insertions(+), 1758 deletions(-)

-- 
2.17.1

[dpdk-dev] [PATCH v2 01/29] bus/fslmc: fix getting the FD error

[Hemant Agrawal]

From: Nipun Gupta 

Fix the incorrect register for getting error

Fixes: 03e36408b9fb ("bus/fslmc: add macros required by QDMA for FLE and FD")
Cc: sta...@dpdk.org

Signed-off-by: Nipun Gupta 
Acked-by: Akhil Goyal 
---
 drivers/bus/fslmc/portal/dpaa2_hw_pvt.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h 
b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
index 4682a5299..f1c70251a 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
@@ -286,7 +286,7 @@ enum qbman_fd_format {
 #define DPAA2_GET_FD_FRC(fd)   ((fd)->simple.frc)
 #define DPAA2_GET_FD_FLC(fd) \
(((uint64_t)((fd)->simple.flc_hi) << 32) + (fd)->simple.flc_lo)
-#define DPAA2_GET_FD_ERR(fd)   ((fd)->simple.bpid_offset & 0x00FF)
+#define DPAA2_GET_FD_ERR(fd)   ((fd)->simple.ctrl & 0x00FF)
 #define DPAA2_GET_FLE_OFFSET(fle) (((fle)->fin_bpid_offset & 0x0FFF) >> 16)
 #define DPAA2_SET_FLE_SG_EXT(fle) ((fle)->fin_bpid_offset |= (uint64_t)1 << 29)
 #define DPAA2_IS_SET_FLE_SG_EXT(fle)   \
-- 
2.17.1

[dpdk-dev] [PATCH v2 02/29] net/dpaa: fix fd offset data type

[Hemant Agrawal]

From: Nipun Gupta 

On DPAA fd offset is 9 bits, but we are using uint8_t in the
SG case. This patch fixes the same.
Fixes: 8cffdcbe85aa ("net/dpaa: support scattered Rx")
Cc: sta...@dpdk.org

Signed-off-by: Nipun Gupta 
Acked-by: Akhil Goyal 
---
 drivers/net/dpaa/dpaa_rxtx.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/dpaa/dpaa_rxtx.c b/drivers/net/dpaa/dpaa_rxtx.c
index 5dba1db8b..3aeecb7d2 100644
--- a/drivers/net/dpaa/dpaa_rxtx.c
+++ b/drivers/net/dpaa/dpaa_rxtx.c
@@ -305,7 +305,7 @@ dpaa_eth_sg_to_mbuf(const struct qm_fd *fd, uint32_t ifid)
struct qm_sg_entry *sgt, *sg_temp;
void *vaddr, *sg_vaddr;
int i = 0;
-   uint8_t fd_offset = fd->offset;
+   uint16_t fd_offset = fd->offset;
 
vaddr = DPAA_MEMPOOL_PTOV(bp_info, qm_fd_addr(fd));
if (!vaddr) {
-- 
2.17.1

[dpdk-dev] [PATCH v2 06/29] bus/fslmc: support handle portal alloc failure

[Hemant Agrawal]

Add the error handling on failure.

Signed-off-by: Nipun Gupta 
Signed-off-by: Hemant Agrawal 
---
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.c | 9 ++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c 
b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
index 47ae72749..5a12ff35d 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
@@ -284,8 +284,10 @@ static struct dpaa2_dpio_dev *dpaa2_get_qbman_swp(void)
if (dpio_dev && rte_atomic16_test_and_set(&dpio_dev->ref_count))
break;
}
-   if (!dpio_dev)
+   if (!dpio_dev) {
+   DPAA2_BUS_ERR("No software portal resource left");
return NULL;
+   }
 
DPAA2_BUS_DEBUG("New Portal %p (%d) affined thread - %lu",
dpio_dev, dpio_dev->index, syscall(SYS_gettid));
@@ -293,6 +295,7 @@ static struct dpaa2_dpio_dev *dpaa2_get_qbman_swp(void)
ret = dpaa2_configure_stashing(dpio_dev);
if (ret) {
DPAA2_BUS_ERR("dpaa2_configure_stashing failed");
+   rte_atomic16_clear(&dpio_dev->ref_count);
return NULL;
}
 
@@ -316,7 +319,7 @@ dpaa2_affine_qbman_swp(void)
if (!RTE_PER_LCORE(_dpaa2_io).dpio_dev) {
dpio_dev = dpaa2_get_qbman_swp();
if (!dpio_dev) {
-   DPAA2_BUS_ERR("No software portal resource left");
+   DPAA2_BUS_ERR("Error in software portal allocation");
return -1;
}
RTE_PER_LCORE(_dpaa2_io).dpio_dev = dpio_dev;
@@ -338,7 +341,7 @@ dpaa2_affine_qbman_ethrx_swp(void)
if (!RTE_PER_LCORE(_dpaa2_io).ethrx_dpio_dev) {
dpio_dev = dpaa2_get_qbman_swp();
if (!dpio_dev) {
-   DPAA2_BUS_ERR("No software portal resource left");
+   DPAA2_BUS_ERR("Error in software portal allocation");
return -1;
}
RTE_PER_LCORE(_dpaa2_io).ethrx_dpio_dev = dpio_dev;
-- 
2.17.1

[dpdk-dev] [PATCH v2 05/29] bus/fslmc: rework portal allocation to a per thread basis

[Hemant Agrawal]

From: Nipun Gupta 

The patch reworks the portal allocation which was previously
being done on per lcore basis to a per thread basis.
Now user can also create its own threads and use DPAA2 portals
for packet I/O.

Signed-off-by: Nipun Gupta 
---
 drivers/bus/fslmc/Makefile   |   1 +
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.c | 220 +--
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.h |   3 -
 3 files changed, 124 insertions(+), 100 deletions(-)

diff --git a/drivers/bus/fslmc/Makefile b/drivers/bus/fslmc/Makefile
index c70e359c8..b98d758ee 100644
--- a/drivers/bus/fslmc/Makefile
+++ b/drivers/bus/fslmc/Makefile
@@ -17,6 +17,7 @@ CFLAGS += -I$(RTE_SDK)/drivers/bus/fslmc/mc
 CFLAGS += -I$(RTE_SDK)/drivers/bus/fslmc/qbman/include
 CFLAGS += -I$(RTE_SDK)/drivers/common/dpaax
 CFLAGS += -I$(RTE_SDK)/lib/librte_eal/common
+LDLIBS += -lpthread
 LDLIBS += -lrte_eal -lrte_mbuf -lrte_mempool -lrte_ring
 LDLIBS += -lrte_ethdev
 LDLIBS += -lrte_common_dpaax
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c 
b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
index 21c535f2f..47ae72749 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
@@ -62,6 +62,9 @@ uint8_t dpaa2_dqrr_size;
 /* Variable to store DPAA2 EQCR size */
 uint8_t dpaa2_eqcr_size;
 
+/* Variable to hold the portal_key, once created.*/
+static pthread_key_t dpaa2_portal_key;
+
 /*Stashing Macros default for LS208x*/
 static int dpaa2_core_cluster_base = 0x04;
 static int dpaa2_cluster_sz = 2;
@@ -87,6 +90,32 @@ static int dpaa2_cluster_sz = 2;
  * Cluster 4 (ID = x07) : CPU14, CPU15;
  */
 
+static int
+dpaa2_get_core_id(void)
+{
+   rte_cpuset_t cpuset;
+   int i, ret, cpu_id = -1;
+
+   ret = pthread_getaffinity_np(pthread_self(), sizeof(cpu_set_t),
+   &cpuset);
+   if (ret) {
+   DPAA2_BUS_ERR("pthread_getaffinity_np() failed");
+   return ret;
+   }
+
+   for (i = 0; i < RTE_MAX_LCORE; i++) {
+   if (CPU_ISSET(i, &cpuset)) {
+   if (cpu_id == -1)
+   cpu_id = i;
+   else
+   /* Multiple cpus are affined */
+   return -1;
+   }
+   }
+
+   return cpu_id;
+}
+
 static int
 dpaa2_core_cluster_sdest(int cpu_id)
 {
@@ -97,7 +126,7 @@ dpaa2_core_cluster_sdest(int cpu_id)
 
 #ifdef RTE_LIBRTE_PMD_DPAA2_EVENTDEV
 static void
-dpaa2_affine_dpio_intr_to_respective_core(int32_t dpio_id, int lcoreid)
+dpaa2_affine_dpio_intr_to_respective_core(int32_t dpio_id, int cpu_id)
 {
 #define STRING_LEN 28
 #define COMMAND_LEN50
@@ -130,7 +159,7 @@ dpaa2_affine_dpio_intr_to_respective_core(int32_t dpio_id, 
int lcoreid)
return;
}
 
-   cpu_mask = cpu_mask << dpaa2_cpu[lcoreid];
+   cpu_mask = cpu_mask << dpaa2_cpu[cpu_id];
snprintf(command, COMMAND_LEN, "echo %X > /proc/irq/%s/smp_affinity",
 cpu_mask, token);
ret = system(command);
@@ -144,7 +173,7 @@ dpaa2_affine_dpio_intr_to_respective_core(int32_t dpio_id, 
int lcoreid)
fclose(file);
 }
 
-static int dpaa2_dpio_intr_init(struct dpaa2_dpio_dev *dpio_dev, int lcoreid)
+static int dpaa2_dpio_intr_init(struct dpaa2_dpio_dev *dpio_dev, int cpu_id)
 {
struct epoll_event epoll_ev;
int eventfd, dpio_epoll_fd, ret;
@@ -181,36 +210,42 @@ static int dpaa2_dpio_intr_init(struct dpaa2_dpio_dev 
*dpio_dev, int lcoreid)
}
dpio_dev->epoll_fd = dpio_epoll_fd;
 
-   dpaa2_affine_dpio_intr_to_respective_core(dpio_dev->hw_id, lcoreid);
+   dpaa2_affine_dpio_intr_to_respective_core(dpio_dev->hw_id, cpu_id);
 
return 0;
 }
+
+static void dpaa2_dpio_intr_deinit(struct dpaa2_dpio_dev *dpio_dev)
+{
+   int ret;
+
+   ret = rte_dpaa2_intr_disable(&dpio_dev->intr_handle, 0);
+   if (ret)
+   DPAA2_BUS_ERR("DPIO interrupt disable failed");
+
+   close(dpio_dev->epoll_fd);
+}
 #endif
 
 static int
-dpaa2_configure_stashing(struct dpaa2_dpio_dev *dpio_dev, int lcoreid)
+dpaa2_configure_stashing(struct dpaa2_dpio_dev *dpio_dev)
 {
int sdest, ret;
int cpu_id;
 
/* Set the Stashing Destination */
-   if (lcoreid < 0) {
-   lcoreid = rte_get_master_lcore();
-   if (lcoreid < 0) {
-   DPAA2_BUS_ERR("Getting CPU Index failed");
-   return -1;
-   }
+   cpu_id = dpaa2_get_core_id();
+   if (cpu_id < 0) {
+   DPAA2_BUS_ERR("Thread not affined to a single core");
+   return -1;
}
 
-   cpu_id = dpaa2_cpu[lcoreid];
-
/* Set the STASH Destination depending on Current CPU ID.
 * Valid values of SDEST are 4,5,6,7. Where,
 */
-
sdest = dpaa2_core_cluster_sdest(cpu_id);
-   DPAA2_BUS_DEBUG("Portal= %d  CPU= %u lcore id =%u SDEST= %d",
-

[dpdk-dev] [PATCH v2 04/29] bus/fslmc: combine thread specific variables

[Hemant Agrawal]

From: Gagandeep Singh 

This is to reduce the thread local storage

Signed-off-by: Gagandeep Singh 
---
 drivers/bus/fslmc/fslmc_bus.c   |  2 --
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.h|  7 +++
 drivers/bus/fslmc/portal/dpaa2_hw_pvt.h |  8 
 drivers/bus/fslmc/rte_bus_fslmc_version.map |  1 -
 drivers/bus/fslmc/rte_fslmc.h   | 18 --
 5 files changed, 15 insertions(+), 21 deletions(-)

diff --git a/drivers/bus/fslmc/fslmc_bus.c b/drivers/bus/fslmc/fslmc_bus.c
index 25d364e81..beb3dd008 100644
--- a/drivers/bus/fslmc/fslmc_bus.c
+++ b/drivers/bus/fslmc/fslmc_bus.c
@@ -35,8 +35,6 @@ rte_fslmc_get_device_count(enum rte_dpaa2_dev_type 
device_type)
return rte_fslmc_bus.device_count[device_type];
 }
 
-RTE_DEFINE_PER_LCORE(struct dpaa2_portal_dqrr, dpaa2_held_bufs);
-
 static void
 cleanup_fslmc_device_list(void)
 {
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.h 
b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.h
index 7c5966241..f6436f2e5 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.h
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.h
@@ -28,6 +28,13 @@ RTE_DECLARE_PER_LCORE(struct dpaa2_io_portal_t, _dpaa2_io);
 #define DPAA2_PER_LCORE_ETHRX_DPIO RTE_PER_LCORE(_dpaa2_io).ethrx_dpio_dev
 #define DPAA2_PER_LCORE_ETHRX_PORTAL DPAA2_PER_LCORE_ETHRX_DPIO->sw_portal
 
+#define DPAA2_PER_LCORE_DQRR_SIZE \
+   RTE_PER_LCORE(_dpaa2_io).dpio_dev->dpaa2_held_bufs.dqrr_size
+#define DPAA2_PER_LCORE_DQRR_HELD \
+   RTE_PER_LCORE(_dpaa2_io).dpio_dev->dpaa2_held_bufs.dqrr_held
+#define DPAA2_PER_LCORE_DQRR_MBUF(i) \
+   RTE_PER_LCORE(_dpaa2_io).dpio_dev->dpaa2_held_bufs.mbuf[i]
+
 /* Variable to store DPAA2 DQRR size */
 extern uint8_t dpaa2_dqrr_size;
 /* Variable to store DPAA2 EQCR size */
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h 
b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
index f1c70251a..be48462dd 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_pvt.h
@@ -87,6 +87,13 @@ struct eqresp_metadata {
struct rte_mempool *mp;
 };
 
+#define DPAA2_PORTAL_DEQUEUE_DEPTH 32
+struct dpaa2_portal_dqrr {
+   struct rte_mbuf *mbuf[DPAA2_PORTAL_DEQUEUE_DEPTH];
+   uint64_t dqrr_held;
+   uint8_t dqrr_size;
+};
+
 struct dpaa2_dpio_dev {
TAILQ_ENTRY(dpaa2_dpio_dev) next;
/**< Pointer to Next device instance */
@@ -112,6 +119,7 @@ struct dpaa2_dpio_dev {
struct rte_intr_handle intr_handle; /* Interrupt related info */
int32_t epoll_fd; /**< File descriptor created for interrupt polling */
int32_t hw_id; /**< An unique ID of this DPIO device instance */
+   struct dpaa2_portal_dqrr dpaa2_held_bufs;
 };
 
 struct dpaa2_dpbp_dev {
diff --git a/drivers/bus/fslmc/rte_bus_fslmc_version.map 
b/drivers/bus/fslmc/rte_bus_fslmc_version.map
index 69e7dc6ad..2a79f4518 100644
--- a/drivers/bus/fslmc/rte_bus_fslmc_version.map
+++ b/drivers/bus/fslmc/rte_bus_fslmc_version.map
@@ -57,7 +57,6 @@ INTERNAL {
mc_get_version;
mc_send_command;
per_lcore__dpaa2_io;
-   per_lcore_dpaa2_held_bufs;
qbman_check_command_complete;
qbman_check_new_result;
qbman_eq_desc_clear;
diff --git a/drivers/bus/fslmc/rte_fslmc.h b/drivers/bus/fslmc/rte_fslmc.h
index 5078b48ee..80873fffc 100644
--- a/drivers/bus/fslmc/rte_fslmc.h
+++ b/drivers/bus/fslmc/rte_fslmc.h
@@ -137,24 +137,6 @@ struct rte_fslmc_bus {
/**< Count of all devices scanned */
 };
 
-#define DPAA2_PORTAL_DEQUEUE_DEPTH 32
-
-/* Create storage for dqrr entries per lcore */
-struct dpaa2_portal_dqrr {
-   struct rte_mbuf *mbuf[DPAA2_PORTAL_DEQUEUE_DEPTH];
-   uint64_t dqrr_held;
-   uint8_t dqrr_size;
-};
-
-RTE_DECLARE_PER_LCORE(struct dpaa2_portal_dqrr, dpaa2_held_bufs);
-
-#define DPAA2_PER_LCORE_DQRR_SIZE \
-   RTE_PER_LCORE(dpaa2_held_bufs).dqrr_size
-#define DPAA2_PER_LCORE_DQRR_HELD \
-   RTE_PER_LCORE(dpaa2_held_bufs).dqrr_held
-#define DPAA2_PER_LCORE_DQRR_MBUF(i) \
-   RTE_PER_LCORE(dpaa2_held_bufs).mbuf[i]
-
 /**
  * Register a DPAA2 driver.
  *
-- 
2.17.1

[dpdk-dev] [PATCH v2 08/29] bus/fslmc: rename the cinh read functions used for ls1088

[Hemant Agrawal]

From: Nipun Gupta 

This patch changes the qbman I/O function names as they are
only reading from cinh register, but writing to cena registers.

This gives way to add functions which purely work in cinh mode

Signed-off-by: Nipun Gupta 
---
 drivers/bus/fslmc/qbman/qbman_portal.c | 250 +++--
 1 file changed, 233 insertions(+), 17 deletions(-)

diff --git a/drivers/bus/fslmc/qbman/qbman_portal.c 
b/drivers/bus/fslmc/qbman/qbman_portal.c
index 57f50b0d8..0a2af7be4 100644
--- a/drivers/bus/fslmc/qbman/qbman_portal.c
+++ b/drivers/bus/fslmc/qbman/qbman_portal.c
@@ -78,7 +78,7 @@ qbman_swp_enqueue_ring_mode_direct(struct qbman_swp *s,
const struct qbman_eq_desc *d,
const struct qbman_fd *fd);
 static int
-qbman_swp_enqueue_ring_mode_cinh_direct(struct qbman_swp *s,
+qbman_swp_enqueue_ring_mode_cinh_read_direct(struct qbman_swp *s,
const struct qbman_eq_desc *d,
const struct qbman_fd *fd);
 static int
@@ -97,7 +97,7 @@ qbman_swp_enqueue_multiple_direct(struct qbman_swp *s,
uint32_t *flags,
int num_frames);
 static int
-qbman_swp_enqueue_multiple_cinh_direct(struct qbman_swp *s,
+qbman_swp_enqueue_multiple_cinh_read_direct(struct qbman_swp *s,
const struct qbman_eq_desc *d,
const struct qbman_fd *fd,
uint32_t *flags,
@@ -122,7 +122,7 @@ qbman_swp_enqueue_multiple_fd_direct(struct qbman_swp *s,
uint32_t *flags,
int num_frames);
 static int
-qbman_swp_enqueue_multiple_fd_cinh_direct(struct qbman_swp *s,
+qbman_swp_enqueue_multiple_fd_cinh_read_direct(struct qbman_swp *s,
const struct qbman_eq_desc *d,
struct qbman_fd **fd,
uint32_t *flags,
@@ -146,7 +146,7 @@ qbman_swp_enqueue_multiple_desc_direct(struct qbman_swp *s,
const struct qbman_fd *fd,
int num_frames);
 static int
-qbman_swp_enqueue_multiple_desc_cinh_direct(struct qbman_swp *s,
+qbman_swp_enqueue_multiple_desc_cinh_read_direct(struct qbman_swp *s,
const struct qbman_eq_desc *d,
const struct qbman_fd *fd,
int num_frames);
@@ -309,15 +309,15 @@ struct qbman_swp *qbman_swp_init(const struct 
qbman_swp_desc *d)
&& (d->cena_access_mode == qman_cena_fastest_access)) {
p->eqcr.pi_ring_size = 32;
qbman_swp_enqueue_array_mode_ptr =
-   qbman_swp_enqueue_array_mode_mem_back;
+   qbman_swp_enqueue_array_mode_mem_back;
qbman_swp_enqueue_ring_mode_ptr =
-   qbman_swp_enqueue_ring_mode_mem_back;
+   qbman_swp_enqueue_ring_mode_mem_back;
qbman_swp_enqueue_multiple_ptr =
-   qbman_swp_enqueue_multiple_mem_back;
+   qbman_swp_enqueue_multiple_mem_back;
qbman_swp_enqueue_multiple_fd_ptr =
-   qbman_swp_enqueue_multiple_fd_mem_back;
+   qbman_swp_enqueue_multiple_fd_mem_back;
qbman_swp_enqueue_multiple_desc_ptr =
-   qbman_swp_enqueue_multiple_desc_mem_back;
+   qbman_swp_enqueue_multiple_desc_mem_back;
qbman_swp_pull_ptr = qbman_swp_pull_mem_back;
qbman_swp_dqrr_next_ptr = qbman_swp_dqrr_next_mem_back;
qbman_swp_release_ptr = qbman_swp_release_mem_back;
@@ -325,13 +325,13 @@ struct qbman_swp *qbman_swp_init(const struct 
qbman_swp_desc *d)
 
if (dpaa2_svr_family == SVR_LS1080A) {
qbman_swp_enqueue_ring_mode_ptr =
-   qbman_swp_enqueue_ring_mode_cinh_direct;
+   qbman_swp_enqueue_ring_mode_cinh_read_direct;
qbman_swp_enqueue_multiple_ptr =
-   qbman_swp_enqueue_multiple_cinh_direct;
+   qbman_swp_enqueue_multiple_cinh_read_direct;
qbman_swp_enqueue_multiple_fd_ptr =
-   qbman_swp_enqueue_multiple_fd_cinh_direct;
+   qbman_swp_enqueue_multiple_fd_cinh_read_direct;
qbman_swp_enqueue_multiple_desc_ptr =
-   qbman_swp_enqueue_multiple_desc_cinh_direct;
+   qbman_swp_enqueue_multiple_desc_cinh_read_direct;
}
 
for (mask_size = p->eqcr.pi_ring_size; mask_size > 0; mask_size >>= 1)
@@ -835,7 +835,7 @@ static int qbman_swp_enqueue_ring_mode_direct(struct 
qbman_swp *s,
return 0;
 }
 
-static int qbman_swp_enqueue_ring_mode_cinh_direct(
+static int qbman_swp_enqueue_ring_mode_cinh_read_direct(
struct qbman_swp *s,
const struct qbman_eq_desc *d,
const struct qbman_fd *fd)
@@ -873,6 +873,44 @@ static int qbman_swp_enqueue_ring_mode_cinh_

[dpdk-dev] [PATCH v2 07/29] bus/fslmc: support portal migration

[Hemant Agrawal]

From: Nipun Gupta 

The patch adds support for portal migration by disabling stashing
for the portals which is used in the non-affined threads, or on
threads affined to multiple cores

Signed-off-by: Nipun Gupta 
---
 doc/guides/rel_notes/release_20_08.rst|   5 +
 drivers/bus/fslmc/portal/dpaa2_hw_dpio.c  |  83 +
 .../bus/fslmc/qbman/include/fsl_qbman_debug.h |   1 +
 .../fslmc/qbman/include/fsl_qbman_portal.h|   8 +-
 drivers/bus/fslmc/qbman/qbman_portal.c| 340 +-
 drivers/bus/fslmc/qbman/qbman_portal.h|  19 +-
 drivers/bus/fslmc/qbman/qbman_sys.h   | 135 ++-
 7 files changed, 508 insertions(+), 83 deletions(-)

diff --git a/doc/guides/rel_notes/release_20_08.rst 
b/doc/guides/rel_notes/release_20_08.rst
index ffae463f4..d915fce12 100644
--- a/doc/guides/rel_notes/release_20_08.rst
+++ b/doc/guides/rel_notes/release_20_08.rst
@@ -119,6 +119,11 @@ New Features
   See the :doc:`../sample_app_ug/l2_forward_real_virtual` for more
   details of this parameter usage.
 
+* **Updated NXP dpaa2 ethdev PMD.**
+
+  Updated the NXP dpaa2 ethdev  with new features and improvements, including:
+
+  * Added support to use datapath APIs from non-EAL pthread
 
 Removed Items
 -
diff --git a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c 
b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
index 5a12ff35d..97be76116 100644
--- a/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
+++ b/drivers/bus/fslmc/portal/dpaa2_hw_dpio.c
@@ -53,10 +53,6 @@ static uint32_t io_space_count;
 /* Variable to store DPAA2 platform type */
 uint32_t dpaa2_svr_family;
 
-/* Physical core id for lcores running on dpaa2. */
-/* DPAA2 only support 1 lcore to 1 phy cpu mapping */
-static unsigned int dpaa2_cpu[RTE_MAX_LCORE];
-
 /* Variable to store DPAA2 DQRR size */
 uint8_t dpaa2_dqrr_size;
 /* Variable to store DPAA2 EQCR size */
@@ -159,7 +155,7 @@ dpaa2_affine_dpio_intr_to_respective_core(int32_t dpio_id, 
int cpu_id)
return;
}
 
-   cpu_mask = cpu_mask << dpaa2_cpu[cpu_id];
+   cpu_mask = cpu_mask << cpu_id;
snprintf(command, COMMAND_LEN, "echo %X > /proc/irq/%s/smp_affinity",
 cpu_mask, token);
ret = system(command);
@@ -228,17 +224,9 @@ static void dpaa2_dpio_intr_deinit(struct dpaa2_dpio_dev 
*dpio_dev)
 #endif
 
 static int
-dpaa2_configure_stashing(struct dpaa2_dpio_dev *dpio_dev)
+dpaa2_configure_stashing(struct dpaa2_dpio_dev *dpio_dev, int cpu_id)
 {
int sdest, ret;
-   int cpu_id;
-
-   /* Set the Stashing Destination */
-   cpu_id = dpaa2_get_core_id();
-   if (cpu_id < 0) {
-   DPAA2_BUS_ERR("Thread not affined to a single core");
-   return -1;
-   }
 
/* Set the STASH Destination depending on Current CPU ID.
 * Valid values of SDEST are 4,5,6,7. Where,
@@ -277,6 +265,7 @@ static void dpaa2_put_qbman_swp(struct dpaa2_dpio_dev 
*dpio_dev)
 static struct dpaa2_dpio_dev *dpaa2_get_qbman_swp(void)
 {
struct dpaa2_dpio_dev *dpio_dev = NULL;
+   int cpu_id;
int ret;
 
/* Get DPIO dev handle from list using index */
@@ -292,11 +281,19 @@ static struct dpaa2_dpio_dev *dpaa2_get_qbman_swp(void)
DPAA2_BUS_DEBUG("New Portal %p (%d) affined thread - %lu",
dpio_dev, dpio_dev->index, syscall(SYS_gettid));
 
-   ret = dpaa2_configure_stashing(dpio_dev);
-   if (ret) {
-   DPAA2_BUS_ERR("dpaa2_configure_stashing failed");
-   rte_atomic16_clear(&dpio_dev->ref_count);
-   return NULL;
+   /* Set the Stashing Destination */
+   cpu_id = dpaa2_get_core_id();
+   if (cpu_id < 0) {
+   DPAA2_BUS_WARN("Thread not affined to a single core");
+   if (dpaa2_svr_family != SVR_LX2160A)
+   qbman_swp_update(dpio_dev->sw_portal, 1);
+   } else {
+   ret = dpaa2_configure_stashing(dpio_dev, cpu_id);
+   if (ret) {
+   DPAA2_BUS_ERR("dpaa2_configure_stashing failed");
+   rte_atomic16_clear(&dpio_dev->ref_count);
+   return NULL;
+   }
}
 
ret = pthread_setspecific(dpaa2_portal_key, (void *)dpio_dev);
@@ -363,46 +360,6 @@ static void dpaa2_portal_finish(void *arg)
pthread_setspecific(dpaa2_portal_key, NULL);
 }
 
-/*
- * This checks for not supported lcore mappings as well as get the physical
- * cpuid for the lcore.
- * one lcore can only map to 1 cpu i.e. 1@10-14 not supported.
- * one cpu can be mapped to more than one lcores.
- */
-static int
-dpaa2_check_lcore_cpuset(void)
-{
-   unsigned int lcore_id, i;
-   int ret = 0;
-
-   for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++)
-   dpaa2_cpu[lcore_id] = 0x;
-
-   for (lcore_id = 0; lcore_id < RTE_MAX_LCORE; lcore_id++) {
-   rte_cpuset_t cpuset = rte_lcore_cpuset(lcore_id);
-
-   fo

[dpdk-dev] [PATCH v2 09/29] net/dpaa: enable Tx queue taildrop

[Hemant Agrawal]

From: Gagandeep Singh 

Enable congestion handling/tail drop for TX queues.

Signed-off-by: Gagandeep Singh 
---
 drivers/bus/dpaa/base/qbman/qman.c|  43 +
 drivers/bus/dpaa/include/fsl_qman.h   |  17 
 drivers/bus/dpaa/rte_bus_dpaa_version.map |   2 +
 drivers/net/dpaa/dpaa_ethdev.c| 111 --
 drivers/net/dpaa/dpaa_ethdev.h|   1 +
 drivers/net/dpaa/dpaa_rxtx.c  |  71 ++
 drivers/net/dpaa/dpaa_rxtx.h  |   3 +
 7 files changed, 242 insertions(+), 6 deletions(-)

diff --git a/drivers/bus/dpaa/base/qbman/qman.c 
b/drivers/bus/dpaa/base/qbman/qman.c
index b596e79c2..447c09177 100644
--- a/drivers/bus/dpaa/base/qbman/qman.c
+++ b/drivers/bus/dpaa/base/qbman/qman.c
@@ -40,6 +40,8 @@
spin_unlock(&__fq478->fqlock); \
} while (0)
 
+static qman_cb_free_mbuf qman_free_mbuf_cb;
+
 static inline void fq_set(struct qman_fq *fq, u32 mask)
 {
dpaa_set_bits(mask, &fq->flags);
@@ -790,6 +792,47 @@ static inline void fq_state_change(struct qman_portal *p, 
struct qman_fq *fq,
FQUNLOCK(fq);
 }
 
+void
+qman_ern_register_cb(qman_cb_free_mbuf cb)
+{
+   qman_free_mbuf_cb = cb;
+}
+
+
+void
+qman_ern_poll_free(void)
+{
+   struct qman_portal *p = get_affine_portal();
+   u8 verb, num = 0;
+   const struct qm_mr_entry *msg;
+   const struct qm_fd *fd;
+   struct qm_mr_entry swapped_msg;
+
+   qm_mr_pvb_update(&p->p);
+   msg = qm_mr_current(&p->p);
+
+   while (msg != NULL) {
+   swapped_msg = *msg;
+   hw_fd_to_cpu(&swapped_msg.ern.fd);
+   verb = msg->ern.verb & QM_MR_VERB_TYPE_MASK;
+   fd = &swapped_msg.ern.fd;
+
+   if (unlikely(verb & 0x20)) {
+   printf("HW ERN notification, Nothing to do\n");
+   } else {
+   if ((fd->bpid & 0xff) != 0xff)
+   qman_free_mbuf_cb(fd);
+   }
+
+   num++;
+   qm_mr_next(&p->p);
+   qm_mr_pvb_update(&p->p);
+   msg = qm_mr_current(&p->p);
+   }
+
+   qm_mr_cci_consume(&p->p, num);
+}
+
 static u32 __poll_portal_slow(struct qman_portal *p, u32 is)
 {
const struct qm_mr_entry *msg;
diff --git a/drivers/bus/dpaa/include/fsl_qman.h 
b/drivers/bus/dpaa/include/fsl_qman.h
index 78b698f39..0d9cfc339 100644
--- a/drivers/bus/dpaa/include/fsl_qman.h
+++ b/drivers/bus/dpaa/include/fsl_qman.h
@@ -1158,6 +1158,10 @@ typedef void (*qman_cb_mr)(struct qman_portal *qm, 
struct qman_fq *fq,
 /* This callback type is used when handling DCP ERNs */
 typedef void (*qman_cb_dc_ern)(struct qman_portal *qm,
const struct qm_mr_entry *msg);
+
+/* This callback function will be used to free mbufs of ERN */
+typedef uint16_t (*qman_cb_free_mbuf)(const struct qm_fd *fd);
+
 /*
  * s/w-visible states. Ie. tentatively scheduled + truly scheduled + active +
  * held-active + held-suspended are just "sched". Things like "retired" will 
not
@@ -1808,6 +1812,19 @@ __rte_internal
 int qman_enqueue_multi(struct qman_fq *fq, const struct qm_fd *fd, u32 *flags,
   int frames_to_send);
 
+/**
+ * qman_ern_poll_free - Polling on MR and calling a callback function to free
+ * mbufs when SW ERNs received.
+ */
+__rte_internal
+void qman_ern_poll_free(void);
+
+/**
+ * qman_ern_register_cb - Register a callback function to free buffers.
+ */
+__rte_internal
+void qman_ern_register_cb(qman_cb_free_mbuf cb);
+
 /**
  * qman_enqueue_multi_fq - Enqueue multiple frames to their respective frame
  * queues.
diff --git a/drivers/bus/dpaa/rte_bus_dpaa_version.map 
b/drivers/bus/dpaa/rte_bus_dpaa_version.map
index 46d42f7d6..8069b05af 100644
--- a/drivers/bus/dpaa/rte_bus_dpaa_version.map
+++ b/drivers/bus/dpaa/rte_bus_dpaa_version.map
@@ -61,6 +61,8 @@ INTERNAL {
qman_enqueue;
qman_enqueue_multi;
qman_enqueue_multi_fq;
+   qman_ern_poll_free;
+   qman_ern_register_cb;
qman_fq_fqid;
qman_fq_portal_irqsource_add;
qman_fq_portal_irqsource_remove;
diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c
index f1c9a7151..fd2c0c681 100644
--- a/drivers/net/dpaa/dpaa_ethdev.c
+++ b/drivers/net/dpaa/dpaa_ethdev.c
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: BSD-3-Clause
  *
  *   Copyright 2016 Freescale Semiconductor, Inc. All rights reserved.
- *   Copyright 2017-2019 NXP
+ *   Copyright 2017-2020 NXP
  *
  */
 /* System headers */
@@ -86,9 +86,12 @@ static int dpaa_push_mode_max_queue = 
DPAA_DEFAULT_PUSH_MODE_QUEUE;
 static int dpaa_push_queue_idx; /* Queue index which are in push mode*/
 
 
-/* Per FQ Taildrop in frame count */
+/* Per RX FQ Taildrop in frame count */
 static unsigned int td_threshold = CGR_RX_PERFQ_THRESH;
 
+/* Per TX FQ Taildrop in frame count, disabled by default */
+static unsigned int td_tx_threshold;

[dpdk-dev] [PATCH v2 10/29] net/dpaa: add 2.5G support

[Hemant Agrawal]

From: Sachin Saxena 

Handle 2.5Gbps ethernet ports as well.

Signed-off-by: Sachin Saxena 
Signed-off-by: Gagandeep Singh 
---
 doc/guides/nics/features/dpaa.ini | 2 +-
 drivers/bus/dpaa/base/fman/fman.c | 6 --
 drivers/bus/dpaa/base/fman/netcfg_layer.c | 3 ++-
 drivers/bus/dpaa/include/fman.h   | 1 +
 drivers/net/dpaa/dpaa_ethdev.c| 9 -
 5 files changed, 16 insertions(+), 5 deletions(-)

diff --git a/doc/guides/nics/features/dpaa.ini 
b/doc/guides/nics/features/dpaa.ini
index 24cfd8566..b00f46a97 100644
--- a/doc/guides/nics/features/dpaa.ini
+++ b/doc/guides/nics/features/dpaa.ini
@@ -4,7 +4,7 @@
 ; Refer to default.ini for the full list of available PMD features.
 ;
 [Features]
-Speed capabilities   = P
+Speed capabilities   = Y
 Link status  = Y
 Jumbo frame  = Y
 MTU update   = Y
diff --git a/drivers/bus/dpaa/base/fman/fman.c 
b/drivers/bus/dpaa/base/fman/fman.c
index 6d77a7e39..ae26041ca 100644
--- a/drivers/bus/dpaa/base/fman/fman.c
+++ b/drivers/bus/dpaa/base/fman/fman.c
@@ -263,7 +263,7 @@ fman_if_init(const struct device_node *dpa_node)
fman_dealloc_bufs_mask_hi = 0;
fman_dealloc_bufs_mask_lo = 0;
}
-   /* Is the MAC node 1G, 10G? */
+   /* Is the MAC node 1G, 2.5G, 10G? */
__if->__if.is_memac = 0;
 
if (of_device_is_compatible(mac_node, "fsl,fman-1g-mac"))
@@ -279,7 +279,9 @@ fman_if_init(const struct device_node *dpa_node)
/* Right now forcing memac to 1g in case of error*/
__if->__if.mac_type = fman_mac_1g;
} else {
-   if (strstr(char_prop, "sgmii"))
+   if (strstr(char_prop, "sgmii-2500"))
+   __if->__if.mac_type = fman_mac_2_5g;
+   else if (strstr(char_prop, "sgmii"))
__if->__if.mac_type = fman_mac_1g;
else if (strstr(char_prop, "rgmii")) {
__if->__if.mac_type = fman_mac_1g;
diff --git a/drivers/bus/dpaa/base/fman/netcfg_layer.c 
b/drivers/bus/dpaa/base/fman/netcfg_layer.c
index 36eca88cd..b7009f229 100644
--- a/drivers/bus/dpaa/base/fman/netcfg_layer.c
+++ b/drivers/bus/dpaa/base/fman/netcfg_layer.c
@@ -44,7 +44,8 @@ dump_netcfg(struct netcfg_info *cfg_ptr)
 
printf("\n+ Fman %d, MAC %d (%s);\n",
   __if->fman_idx, __if->mac_idx,
-  (__if->mac_type == fman_mac_1g) ? "1G" : "10G");
+  (__if->mac_type == fman_mac_1g) ? "1G" :
+  (__if->mac_type == fman_mac_2_5g) ? "2.5G" : "10G");
 
printf("\tmac_addr: %02x:%02x:%02x:%02x:%02x:%02x\n",
   (&__if->mac_addr)->addr_bytes[0],
diff --git a/drivers/bus/dpaa/include/fman.h b/drivers/bus/dpaa/include/fman.h
index c02d32d22..b6293b61c 100644
--- a/drivers/bus/dpaa/include/fman.h
+++ b/drivers/bus/dpaa/include/fman.h
@@ -72,6 +72,7 @@ enum fman_mac_type {
fman_offline = 0,
fman_mac_1g,
fman_mac_10g,
+   fman_mac_2_5g,
 };
 
 struct mac_addr {
diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c
index fd2c0c681..c0ded9086 100644
--- a/drivers/net/dpaa/dpaa_ethdev.c
+++ b/drivers/net/dpaa/dpaa_ethdev.c
@@ -356,8 +356,13 @@ static int dpaa_eth_dev_info(struct rte_eth_dev *dev,
 
if (dpaa_intf->fif->mac_type == fman_mac_1g) {
dev_info->speed_capa = ETH_LINK_SPEED_1G;
+   } else if (dpaa_intf->fif->mac_type == fman_mac_2_5g) {
+   dev_info->speed_capa = ETH_LINK_SPEED_1G
+   | ETH_LINK_SPEED_2_5G;
} else if (dpaa_intf->fif->mac_type == fman_mac_10g) {
-   dev_info->speed_capa = (ETH_LINK_SPEED_1G | ETH_LINK_SPEED_10G);
+   dev_info->speed_capa = ETH_LINK_SPEED_1G
+   | ETH_LINK_SPEED_2_5G
+   | ETH_LINK_SPEED_10G;
} else {
DPAA_PMD_ERR("invalid link_speed: %s, %d",
 dpaa_intf->name, dpaa_intf->fif->mac_type);
@@ -388,6 +393,8 @@ static int dpaa_eth_link_update(struct rte_eth_dev *dev,
 
if (dpaa_intf->fif->mac_type == fman_mac_1g)
link->link_speed = ETH_SPEED_NUM_1G;
+   else if (dpaa_intf->fif->mac_type == fman_mac_2_5g)
+   link->link_speed = ETH_SPEED_NUM_2_5G;
else if (dpaa_intf->fif->mac_type == fman_mac_10g)
link->link_speed = ETH_SPEED_NUM_10G;
else
-- 
2.17.1

[dpdk-dev] [PATCH v2 12/29] drivers: optimize thread local storage for dpaa

[Hemant Agrawal]

From: Rohit Raj 

Minimize the number of different thread variables

Add all the thread specific variables in dpaa_portal
structure to optimize TLS Usage.

Signed-off-by: Rohit Raj 
Acked-by: Akhil Goyal 
---
 doc/guides/rel_notes/release_20_08.rst|  6 
 drivers/bus/dpaa/dpaa_bus.c   | 24 ++---
 drivers/bus/dpaa/rte_bus_dpaa_version.map |  1 -
 drivers/bus/dpaa/rte_dpaa_bus.h   | 42 ++-
 drivers/crypto/dpaa_sec/dpaa_sec.c| 11 +++---
 drivers/event/dpaa/dpaa_eventdev.c|  4 +--
 drivers/mempool/dpaa/dpaa_mempool.c   |  6 ++--
 drivers/net/dpaa/dpaa_ethdev.c|  2 +-
 drivers/net/dpaa/dpaa_rxtx.c  |  4 +--
 9 files changed, 54 insertions(+), 46 deletions(-)

diff --git a/doc/guides/rel_notes/release_20_08.rst 
b/doc/guides/rel_notes/release_20_08.rst
index d915fce12..b1e039d03 100644
--- a/doc/guides/rel_notes/release_20_08.rst
+++ b/doc/guides/rel_notes/release_20_08.rst
@@ -119,6 +119,12 @@ New Features
   See the :doc:`../sample_app_ug/l2_forward_real_virtual` for more
   details of this parameter usage.
 
+* **Updated NXP dpaa ethdev PMD.**
+
+  Updated the NXP dpaa ethdev  with new features and improvements, including:
+
+  * Added support to use datapath APIs from non-EAL pthread
+
 * **Updated NXP dpaa2 ethdev PMD.**
 
   Updated the NXP dpaa2 ethdev  with new features and improvements, including:
diff --git a/drivers/bus/dpaa/dpaa_bus.c b/drivers/bus/dpaa/dpaa_bus.c
index 6770fbc52..aa906c34e 100644
--- a/drivers/bus/dpaa/dpaa_bus.c
+++ b/drivers/bus/dpaa/dpaa_bus.c
@@ -52,8 +52,7 @@ unsigned int dpaa_svr_family;
 
 #define FSL_DPAA_BUS_NAME  dpaa_bus
 
-RTE_DEFINE_PER_LCORE(bool, dpaa_io);
-RTE_DEFINE_PER_LCORE(struct dpaa_portal_dqrr, held_bufs);
+RTE_DEFINE_PER_LCORE(struct dpaa_portal *, dpaa_io);
 
 struct fm_eth_port_cfg *
 dpaa_get_eth_port_cfg(int dev_id)
@@ -253,7 +252,6 @@ int rte_dpaa_portal_init(void *arg)
 {
unsigned int cpu, lcore = rte_lcore_id();
int ret;
-   struct dpaa_portal *dpaa_io_portal;
 
BUS_INIT_FUNC_TRACE();
 
@@ -288,20 +286,21 @@ int rte_dpaa_portal_init(void *arg)
DPAA_BUS_LOG(DEBUG, "QMAN thread initialized - CPU=%d lcore=%d",
 cpu, lcore);
 
-   dpaa_io_portal = rte_malloc(NULL, sizeof(struct dpaa_portal),
+   DPAA_PER_LCORE_PORTAL = rte_malloc(NULL, sizeof(struct dpaa_portal),
RTE_CACHE_LINE_SIZE);
-   if (!dpaa_io_portal) {
+   if (!DPAA_PER_LCORE_PORTAL) {
DPAA_BUS_LOG(ERR, "Unable to allocate memory");
bman_thread_finish();
qman_thread_finish();
return -ENOMEM;
}
 
-   dpaa_io_portal->qman_idx = qman_get_portal_index();
-   dpaa_io_portal->bman_idx = bman_get_portal_index();
-   dpaa_io_portal->tid = syscall(SYS_gettid);
+   DPAA_PER_LCORE_PORTAL->qman_idx = qman_get_portal_index();
+   DPAA_PER_LCORE_PORTAL->bman_idx = bman_get_portal_index();
+   DPAA_PER_LCORE_PORTAL->tid = syscall(SYS_gettid);
 
-   ret = pthread_setspecific(dpaa_portal_key, (void *)dpaa_io_portal);
+   ret = pthread_setspecific(dpaa_portal_key,
+ (void *)DPAA_PER_LCORE_PORTAL);
if (ret) {
DPAA_BUS_LOG(ERR, "pthread_setspecific failed on core %u"
 " (lcore=%u) with ret: %d", cpu, lcore, ret);
@@ -310,8 +309,6 @@ int rte_dpaa_portal_init(void *arg)
return ret;
}
 
-   RTE_PER_LCORE(dpaa_io) = true;
-
DPAA_BUS_LOG(DEBUG, "QMAN thread initialized");
 
return 0;
@@ -324,7 +321,7 @@ rte_dpaa_portal_fq_init(void *arg, struct qman_fq *fq)
u32 sdqcr;
int ret;
 
-   if (unlikely(!RTE_PER_LCORE(dpaa_io))) {
+   if (unlikely(!DPAA_PER_LCORE_PORTAL)) {
ret = rte_dpaa_portal_init(arg);
if (ret < 0) {
DPAA_BUS_LOG(ERR, "portal initialization failure");
@@ -367,8 +364,7 @@ dpaa_portal_finish(void *arg)
 
rte_free(dpaa_io_portal);
dpaa_io_portal = NULL;
-
-   RTE_PER_LCORE(dpaa_io) = false;
+   DPAA_PER_LCORE_PORTAL = NULL;
 }
 
 static int
diff --git a/drivers/bus/dpaa/rte_bus_dpaa_version.map 
b/drivers/bus/dpaa/rte_bus_dpaa_version.map
index 8069b05af..2defa7992 100644
--- a/drivers/bus/dpaa/rte_bus_dpaa_version.map
+++ b/drivers/bus/dpaa/rte_bus_dpaa_version.map
@@ -48,7 +48,6 @@ INTERNAL {
netcfg_acquire;
netcfg_release;
per_lcore_dpaa_io;
-   per_lcore_held_bufs;
qman_alloc_cgrid_range;
qman_alloc_pool_range;
qman_clear_irq;
diff --git a/drivers/bus/dpaa/rte_dpaa_bus.h b/drivers/bus/dpaa/rte_dpaa_bus.h
index 2a186d83f..25aff2d30 100644
--- a/drivers/bus/dpaa/rte_dpaa_bus.h
+++ b/drivers/bus/dpaa/rte_dpaa_bus.h
@@ -35,8 +35,6 @@
 
 extern unsigned int dpaa_svr_family;
 
-extern RTE_DEFINE_PER_LCORE(bool, dpaa_io);
-
 st

[dpdk-dev] [PATCH v2 11/29] net/dpaa: update process specific device info

[Hemant Agrawal]

From: Nipun Gupta 

For DPAA devices the memory maps stored in the FMAN interface
information is per process. Store them in the device process specific
area.
This is required to support multi-process apps.

Signed-off-by: Nipun Gupta 
---
 drivers/net/dpaa/dpaa_ethdev.c | 207 -
 drivers/net/dpaa/dpaa_ethdev.h |   1 -
 2 files changed, 102 insertions(+), 106 deletions(-)

diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c
index c0ded9086..6c94fd396 100644
--- a/drivers/net/dpaa/dpaa_ethdev.c
+++ b/drivers/net/dpaa/dpaa_ethdev.c
@@ -149,7 +149,6 @@ dpaa_poll_queue_default_config(struct qm_mcc_initfq *opts)
 static int
 dpaa_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
 {
-   struct dpaa_if *dpaa_intf = dev->data->dev_private;
uint32_t frame_size = mtu + RTE_ETHER_HDR_LEN + RTE_ETHER_CRC_LEN
+ VLAN_TAG_SIZE;
uint32_t buffsz = dev->data->min_rx_buf_size - RTE_PKTMBUF_HEADROOM;
@@ -185,7 +184,7 @@ dpaa_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
 
dev->data->dev_conf.rxmode.max_rx_pkt_len = frame_size;
 
-   fman_if_set_maxfrm(dpaa_intf->fif, frame_size);
+   fman_if_set_maxfrm(dev->process_private, frame_size);
 
return 0;
 }
@@ -193,7 +192,6 @@ dpaa_mtu_set(struct rte_eth_dev *dev, uint16_t mtu)
 static int
 dpaa_eth_dev_configure(struct rte_eth_dev *dev)
 {
-   struct dpaa_if *dpaa_intf = dev->data->dev_private;
struct rte_eth_conf *eth_conf = &dev->data->dev_conf;
uint64_t rx_offloads = eth_conf->rxmode.offloads;
uint64_t tx_offloads = eth_conf->txmode.offloads;
@@ -232,14 +230,14 @@ dpaa_eth_dev_configure(struct rte_eth_dev *dev)
max_len = DPAA_MAX_RX_PKT_LEN;
}
 
-   fman_if_set_maxfrm(dpaa_intf->fif, max_len);
+   fman_if_set_maxfrm(dev->process_private, max_len);
dev->data->mtu = max_len
- RTE_ETHER_HDR_LEN - RTE_ETHER_CRC_LEN - VLAN_TAG_SIZE;
}
 
if (rx_offloads & DEV_RX_OFFLOAD_SCATTER) {
DPAA_PMD_DEBUG("enabling scatter mode");
-   fman_if_set_sg(dpaa_intf->fif, 1);
+   fman_if_set_sg(dev->process_private, 1);
dev->data->scattered_rx = 1;
}
 
@@ -283,18 +281,18 @@ static int dpaa_eth_dev_start(struct rte_eth_dev *dev)
else
dev->tx_pkt_burst = dpaa_eth_queue_tx;
 
-   fman_if_enable_rx(dpaa_intf->fif);
+   fman_if_enable_rx(dev->process_private);
 
return 0;
 }
 
 static void dpaa_eth_dev_stop(struct rte_eth_dev *dev)
 {
-   struct dpaa_if *dpaa_intf = dev->data->dev_private;
+   struct fman_if *fif = dev->process_private;
 
PMD_INIT_FUNC_TRACE();
 
-   fman_if_disable_rx(dpaa_intf->fif);
+   fman_if_disable_rx(fif);
dev->tx_pkt_burst = dpaa_eth_tx_drop_all;
 }
 
@@ -342,6 +340,7 @@ static int dpaa_eth_dev_info(struct rte_eth_dev *dev,
 struct rte_eth_dev_info *dev_info)
 {
struct dpaa_if *dpaa_intf = dev->data->dev_private;
+   struct fman_if *fif = dev->process_private;
 
DPAA_PMD_DEBUG(": %s", dpaa_intf->name);
 
@@ -354,18 +353,18 @@ static int dpaa_eth_dev_info(struct rte_eth_dev *dev,
dev_info->max_vmdq_pools = ETH_16_POOLS;
dev_info->flow_type_rss_offloads = DPAA_RSS_OFFLOAD_ALL;
 
-   if (dpaa_intf->fif->mac_type == fman_mac_1g) {
+   if (fif->mac_type == fman_mac_1g) {
dev_info->speed_capa = ETH_LINK_SPEED_1G;
-   } else if (dpaa_intf->fif->mac_type == fman_mac_2_5g) {
+   } else if (fif->mac_type == fman_mac_2_5g) {
dev_info->speed_capa = ETH_LINK_SPEED_1G
| ETH_LINK_SPEED_2_5G;
-   } else if (dpaa_intf->fif->mac_type == fman_mac_10g) {
+   } else if (fif->mac_type == fman_mac_10g) {
dev_info->speed_capa = ETH_LINK_SPEED_1G
| ETH_LINK_SPEED_2_5G
| ETH_LINK_SPEED_10G;
} else {
DPAA_PMD_ERR("invalid link_speed: %s, %d",
-dpaa_intf->name, dpaa_intf->fif->mac_type);
+dpaa_intf->name, fif->mac_type);
return -EINVAL;
}
 
@@ -388,18 +387,19 @@ static int dpaa_eth_link_update(struct rte_eth_dev *dev,
 {
struct dpaa_if *dpaa_intf = dev->data->dev_private;
struct rte_eth_link *link = &dev->data->dev_link;
+   struct fman_if *fif = dev->process_private;
 
PMD_INIT_FUNC_TRACE();
 
-   if (dpaa_intf->fif->mac_type == fman_mac_1g)
+   if (fif->mac_type == fman_mac_1g)
link->link_speed = ETH_SPEED_NUM_1G;
-   else if (dpaa_intf->fif->mac_type == fman_mac_2_5g)
+   else if (fif->mac_type == fman_mac_2_5g)
link->link_speed = ETH_SPEED_NUM_2_5G;
-   else if (dpaa

[dpdk-dev] [PATCH v2 15/29] net/dpaa2: support dynamic flow control

[Hemant Agrawal]

From: Jun Yang 

Dynamic flow used instead of layout defined.

The actual key/mask size depends on protocols and(or) fields
of patterns specified.
Also, the key and mask should start from the beginning of IOVA.

Signed-off-by: Jun Yang 
---
 doc/guides/nics/features/dpaa2.ini |   1 +
 doc/guides/rel_notes/release_20_08.rst |   1 +
 drivers/net/dpaa2/dpaa2_flow.c | 146 ++---
 3 files changed, 36 insertions(+), 112 deletions(-)

diff --git a/doc/guides/nics/features/dpaa2.ini 
b/doc/guides/nics/features/dpaa2.ini
index c2214fbd5..3685e2e02 100644
--- a/doc/guides/nics/features/dpaa2.ini
+++ b/doc/guides/nics/features/dpaa2.ini
@@ -16,6 +16,7 @@ Unicast MAC filter   = Y
 RSS hash = Y
 VLAN filter  = Y
 Flow control = Y
+Flow API = Y
 VLAN offload = Y
 L3 checksum offload  = Y
 L4 checksum offload  = Y
diff --git a/doc/guides/rel_notes/release_20_08.rst 
b/doc/guides/rel_notes/release_20_08.rst
index e5bc5cfd8..97267f7b7 100644
--- a/doc/guides/rel_notes/release_20_08.rst
+++ b/doc/guides/rel_notes/release_20_08.rst
@@ -131,6 +131,7 @@ New Features
   Updated the NXP dpaa2 ethdev  with new features and improvements, including:
 
   * Added support to use datapath APIs from non-EAL pthread
+  * Added support for dynamic flow management
 
 Removed Items
 -
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 8aa65db30..05d115c78 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -33,29 +33,6 @@ struct rte_flow {
uint16_t flow_id;
 };
 
-/* Layout for rule compositions for supported patterns */
-/* TODO: Current design only supports Ethernet + IPv4 based classification. */
-/* So corresponding offset macros are valid only. Rest are placeholder for */
-/* now. Once support for other netwrok headers will be added then */
-/* corresponding macros will be updated with correct values*/
-#define DPAA2_CLS_RULE_OFFSET_ETH  0   /*Start of buffer*/
-#define DPAA2_CLS_RULE_OFFSET_VLAN 14  /* DPAA2_CLS_RULE_OFFSET_ETH */
-   /*  + Sizeof Eth fields  */
-#define DPAA2_CLS_RULE_OFFSET_IPV4 14  /* DPAA2_CLS_RULE_OFFSET_VLAN */
-   /*  + Sizeof VLAN fields */
-#define DPAA2_CLS_RULE_OFFSET_IPV6 25  /* DPAA2_CLS_RULE_OFFSET_IPV4 */
-   /*  + Sizeof IPV4 fields */
-#define DPAA2_CLS_RULE_OFFSET_ICMP 58  /* DPAA2_CLS_RULE_OFFSET_IPV6 */
-   /*  + Sizeof IPV6 fields */
-#define DPAA2_CLS_RULE_OFFSET_UDP  60  /* DPAA2_CLS_RULE_OFFSET_ICMP */
-   /*  + Sizeof ICMP fields */
-#define DPAA2_CLS_RULE_OFFSET_TCP  64  /* DPAA2_CLS_RULE_OFFSET_UDP  */
-   /*  + Sizeof UDP fields  */
-#define DPAA2_CLS_RULE_OFFSET_SCTP 68  /* DPAA2_CLS_RULE_OFFSET_TCP  */
-   /*  + Sizeof TCP fields  */
-#define DPAA2_CLS_RULE_OFFSET_GRE  72  /* DPAA2_CLS_RULE_OFFSET_SCTP */
-   /*  + Sizeof SCTP fields */
-
 static const
 enum rte_flow_item_type dpaa2_supported_pattern_type[] = {
RTE_FLOW_ITEM_TYPE_END,
@@ -212,7 +189,7 @@ dpaa2_configure_flow_eth(struct rte_flow *flow,
(pattern->mask ? pattern->mask : default_mask);
 
/* Key rule */
-   key_iova = flow->rule.key_iova + DPAA2_CLS_RULE_OFFSET_ETH;
+   key_iova = flow->rule.key_iova + flow->key_size;
memcpy((void *)key_iova, (const void *)(spec->src.addr_bytes),
sizeof(struct rte_ether_addr));
key_iova += sizeof(struct rte_ether_addr);
@@ -223,7 +200,7 @@ dpaa2_configure_flow_eth(struct rte_flow *flow,
sizeof(rte_be16_t));
 
/* Key mask */
-   mask_iova = flow->rule.mask_iova + DPAA2_CLS_RULE_OFFSET_ETH;
+   mask_iova = flow->rule.mask_iova + flow->key_size;
memcpy((void *)mask_iova, (const void *)(mask->src.addr_bytes),
sizeof(struct rte_ether_addr));
mask_iova += sizeof(struct rte_ether_addr);
@@ -233,9 +210,9 @@ dpaa2_configure_flow_eth(struct rte_flow *flow,
memcpy((void *)mask_iova, (const void *)(&mask->type),
sizeof(rte_be16_t));
 
-   flow->rule.key_size = (DPAA2_CLS_RULE_OFFSET_ETH +
-   ((2  * sizeof(struct rte_ether_addr)) +
-   sizeof(rte_be16_t)));
+   flow->key_size += ((2  * sizeof(struct rte_ether_addr)) +
+   sizeof(rte_be16_t));
+
return device_configured;
 }
 
@@ -335,15 +312,15 @@ dpaa

[dpdk-dev] [PATCH v2 14/29] bus/dpaa: enable set link status

[Hemant Agrawal]

From: Rohit Raj 

Enabled set link status API to start/stop phy
device from application.

Signed-off-by: Rohit Raj 
---
 drivers/bus/dpaa/base/qbman/process.c | 27 +
 drivers/bus/dpaa/include/process.h| 11 +++
 drivers/bus/dpaa/rte_bus_dpaa_version.map |  1 +
 drivers/net/dpaa/dpaa_ethdev.c| 35 ---
 4 files changed, 63 insertions(+), 11 deletions(-)

diff --git a/drivers/bus/dpaa/base/qbman/process.c 
b/drivers/bus/dpaa/base/qbman/process.c
index 68b7af243..6f7e37957 100644
--- a/drivers/bus/dpaa/base/qbman/process.c
+++ b/drivers/bus/dpaa/base/qbman/process.c
@@ -366,3 +366,30 @@ int dpaa_get_link_status(char *if_name)
 
return args.link_status;
 }
+
+#define DPAA_IOCTL_UPDATE_LINK_STATUS \
+   _IOW(DPAA_IOCTL_MAGIC, 0x11, struct 
usdpaa_ioctl_update_link_status_args)
+
+int dpaa_update_link_status(char *if_name, int link_status)
+{
+   struct usdpaa_ioctl_update_link_status_args args;
+   int ret;
+
+   ret = check_fd();
+   if (ret)
+   return ret;
+
+   strcpy(args.if_name, if_name);
+   args.link_status = link_status;
+
+   ret = ioctl(fd, DPAA_IOCTL_UPDATE_LINK_STATUS, &args);
+   if (ret) {
+   if (errno == EINVAL)
+   printf("Failed to set link status: Not Supported\n");
+   else
+   printf("Failed to set link status");
+   return ret;
+   }
+
+   return 0;
+}
diff --git a/drivers/bus/dpaa/include/process.h 
b/drivers/bus/dpaa/include/process.h
index 7305762c2..f52ea1635 100644
--- a/drivers/bus/dpaa/include/process.h
+++ b/drivers/bus/dpaa/include/process.h
@@ -91,7 +91,18 @@ struct usdpaa_ioctl_link_status_args {
charif_name[IF_NAME_MAX_LEN];
int link_status;
 };
+
+struct usdpaa_ioctl_update_link_status_args {
+   /* network device node name */
+   charif_name[IF_NAME_MAX_LEN];
+   /* link status(ETH_LINK_UP/DOWN) */
+   int link_status;
+};
+
 __rte_internal
 int dpaa_get_link_status(char *if_name);
 
+__rte_internal
+int dpaa_update_link_status(char *if_name, int link_status);
+
 #endif /*  __PROCESS_H */
diff --git a/drivers/bus/dpaa/rte_bus_dpaa_version.map 
b/drivers/bus/dpaa/rte_bus_dpaa_version.map
index 96662d7be..5dec8d9e5 100644
--- a/drivers/bus/dpaa/rte_bus_dpaa_version.map
+++ b/drivers/bus/dpaa/rte_bus_dpaa_version.map
@@ -19,6 +19,7 @@ INTERNAL {
dpaa_intr_disable;
dpaa_intr_enable;
dpaa_svr_family;
+   dpaa_update_link_status;
fman_dealloc_bufs_mask_hi;
fman_dealloc_bufs_mask_lo;
fman_if_add_mac_addr;
diff --git a/drivers/net/dpaa/dpaa_ethdev.c b/drivers/net/dpaa/dpaa_ethdev.c
index 3f805b2b0..3a5b319d4 100644
--- a/drivers/net/dpaa/dpaa_ethdev.c
+++ b/drivers/net/dpaa/dpaa_ethdev.c
@@ -478,18 +478,15 @@ static int dpaa_eth_link_update(struct rte_eth_dev *dev,
DPAA_PMD_ERR("invalid link_speed: %s, %d",
 dpaa_intf->name, fif->mac_type);
 
-   ret = dpaa_get_link_status(__fif->node_name);
-   if (ret < 0) {
-   if (ret == -EINVAL) {
-   DPAA_PMD_DEBUG("Using default link status-No Support");
-   ret = 1;
-   } else {
-   DPAA_PMD_ERR("rte_dpaa_get_link_status %d", ret);
+   if (dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC) {
+   ret = dpaa_get_link_status(__fif->node_name);
+   if (ret < 0)
return ret;
-   }
+   link->link_status = ret;
+   } else {
+   link->link_status = dpaa_intf->valid;
}
 
-   link->link_status = ret;
link->link_duplex = ETH_LINK_FULL_DUPLEX;
link->link_autoneg = ETH_LINK_AUTONEG;
 
@@ -985,17 +982,33 @@ dpaa_dev_rx_queue_count(struct rte_eth_dev *dev, uint16_t 
rx_queue_id)
 
 static int dpaa_link_down(struct rte_eth_dev *dev)
 {
+   struct fman_if *fif = dev->process_private;
+   struct __fman_if *__fif;
+
PMD_INIT_FUNC_TRACE();
 
-   dpaa_eth_dev_stop(dev);
+   __fif = container_of(fif, struct __fman_if, __if);
+
+   if (dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC)
+   dpaa_update_link_status(__fif->node_name, ETH_LINK_DOWN);
+   else
+   dpaa_eth_dev_stop(dev);
return 0;
 }
 
 static int dpaa_link_up(struct rte_eth_dev *dev)
 {
+   struct fman_if *fif = dev->process_private;
+   struct __fman_if *__fif;
+
PMD_INIT_FUNC_TRACE();
 
-   dpaa_eth_dev_start(dev);
+   __fif = container_of(fif, struct __fman_if, __if);
+
+   if (dev->data->dev_flags & RTE_ETH_DEV_INTR_LSC)
+   dpaa_update_link_status(__fif->node_name, ETH_LINK_UP);
+   else
+   dpaa_eth_dev_start(dev);
return 0;
 }
 
-- 
2.17.1

[dpdk-dev] [PATCH v2 17/29] net/dpaa2: add sanity check for flow extracts

[Hemant Agrawal]

From: Jun Yang 

Define extracts support for each protocol and check the fields of each
pattern before building extracts of QoS/FS table.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_ethdev.c |   7 +-
 drivers/net/dpaa2/dpaa2_flow.c   | 250 +--
 2 files changed, 204 insertions(+), 53 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index 492b65840..fd3097c7d 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -2610,11 +2610,8 @@ dpaa2_dev_uninit(struct rte_eth_dev *eth_dev)
eth_dev->process_private = NULL;
rte_free(dpni);
 
-   for (i = 0; i < MAX_TCS; i++) {
-   if (priv->extract.tc_extract_param[i])
-   rte_free((void *)
-   (size_t)priv->extract.tc_extract_param[i]);
-   }
+   for (i = 0; i < MAX_TCS; i++)
+   rte_free((void *)(size_t)priv->extract.tc_extract_param[i]);
 
if (priv->extract.qos_extract_param)
rte_free((void *)(size_t)priv->extract.qos_extract_param);
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 779cb64ab..507a5d0e3 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -87,7 +87,68 @@ enum rte_flow_action_type dpaa2_supported_action_type[] = {
 #define DPAA2_FLOW_ITEM_TYPE_GENERIC_IP (RTE_FLOW_ITEM_TYPE_META + 1)
 
 enum rte_filter_type dpaa2_filter_type = RTE_ETH_FILTER_NONE;
-static const void *default_mask;
+
+#ifndef __cplusplus
+static const struct rte_flow_item_eth dpaa2_flow_item_eth_mask = {
+   .dst.addr_bytes = "\xff\xff\xff\xff\xff\xff",
+   .src.addr_bytes = "\xff\xff\xff\xff\xff\xff",
+   .type = RTE_BE16(0x),
+};
+
+static const struct rte_flow_item_vlan dpaa2_flow_item_vlan_mask = {
+   .tci = RTE_BE16(0x),
+};
+
+static const struct rte_flow_item_ipv4 dpaa2_flow_item_ipv4_mask = {
+   .hdr.src_addr = RTE_BE32(0x),
+   .hdr.dst_addr = RTE_BE32(0x),
+   .hdr.next_proto_id = 0xff,
+};
+
+static const struct rte_flow_item_ipv6 dpaa2_flow_item_ipv6_mask = {
+   .hdr = {
+   .src_addr =
+   "\xff\xff\xff\xff\xff\xff\xff\xff"
+   "\xff\xff\xff\xff\xff\xff\xff\xff",
+   .dst_addr =
+   "\xff\xff\xff\xff\xff\xff\xff\xff"
+   "\xff\xff\xff\xff\xff\xff\xff\xff",
+   .proto = 0xff
+   },
+};
+
+static const struct rte_flow_item_icmp dpaa2_flow_item_icmp_mask = {
+   .hdr.icmp_type = 0xff,
+   .hdr.icmp_code = 0xff,
+};
+
+static const struct rte_flow_item_udp dpaa2_flow_item_udp_mask = {
+   .hdr = {
+   .src_port = RTE_BE16(0x),
+   .dst_port = RTE_BE16(0x),
+   },
+};
+
+static const struct rte_flow_item_tcp dpaa2_flow_item_tcp_mask = {
+   .hdr = {
+   .src_port = RTE_BE16(0x),
+   .dst_port = RTE_BE16(0x),
+   },
+};
+
+static const struct rte_flow_item_sctp dpaa2_flow_item_sctp_mask = {
+   .hdr = {
+   .src_port = RTE_BE16(0x),
+   .dst_port = RTE_BE16(0x),
+   },
+};
+
+static const struct rte_flow_item_gre dpaa2_flow_item_gre_mask = {
+   .protocol = RTE_BE16(0x),
+};
+
+#endif
+
 
 static inline void dpaa2_flow_extract_key_set(
struct dpaa2_key_info *key_info, int index, uint8_t size)
@@ -555,6 +616,67 @@ dpaa2_flow_rule_move_ipaddr_tail(
return 0;
 }
 
+static int
+dpaa2_flow_extract_support(
+   const uint8_t *mask_src,
+   enum rte_flow_item_type type)
+{
+   char mask[64];
+   int i, size = 0;
+   const char *mask_support = 0;
+
+   switch (type) {
+   case RTE_FLOW_ITEM_TYPE_ETH:
+   mask_support = (const char *)&dpaa2_flow_item_eth_mask;
+   size = sizeof(struct rte_flow_item_eth);
+   break;
+   case RTE_FLOW_ITEM_TYPE_VLAN:
+   mask_support = (const char *)&dpaa2_flow_item_vlan_mask;
+   size = sizeof(struct rte_flow_item_vlan);
+   break;
+   case RTE_FLOW_ITEM_TYPE_IPV4:
+   mask_support = (const char *)&dpaa2_flow_item_ipv4_mask;
+   size = sizeof(struct rte_flow_item_ipv4);
+   break;
+   case RTE_FLOW_ITEM_TYPE_IPV6:
+   mask_support = (const char *)&dpaa2_flow_item_ipv6_mask;
+   size = sizeof(struct rte_flow_item_ipv6);
+   break;
+   case RTE_FLOW_ITEM_TYPE_ICMP:
+   mask_support = (const char *)&dpaa2_flow_item_icmp_mask;
+   size = sizeof(struct rte_flow_item_icmp);
+   break;
+   case RTE_FLOW_ITEM_TYPE_UDP:
+   mask_support = (const char *)&dpaa2_flow_item_udp_mask;
+   size = sizeof(struct rte_flow_item_udp);
+   break;
+   case RTE_FLOW_ITEM_TYPE_TCP:
+   

[dpdk-dev] [PATCH v2 13/29] bus/dpaa: enable link state interrupt

[Hemant Agrawal]

From: Rohit Raj 

Enable/disable link state interrupt and get link state api is
defined using IOCTL calls from kernel driver

Signed-off-by: Rohit Raj 
---
 doc/guides/nics/features/dpaa.ini |  1 +
 doc/guides/rel_notes/release_20_08.rst|  1 +
 drivers/bus/dpaa/base/fman/fman.c |  4 +-
 drivers/bus/dpaa/base/qbman/process.c | 72 -
 drivers/bus/dpaa/dpaa_bus.c   | 28 ++-
 drivers/bus/dpaa/include/fman.h   |  2 +
 drivers/bus/dpaa/include/process.h| 20 +
 drivers/bus/dpaa/rte_bus_dpaa_version.map |  3 +
 drivers/bus/dpaa/rte_dpaa_bus.h   |  6 +-
 drivers/common/dpaax/compat.h |  5 +-
 drivers/net/dpaa/dpaa_ethdev.c| 97 ++-
 11 files changed, 233 insertions(+), 6 deletions(-)

diff --git a/doc/guides/nics/features/dpaa.ini 
b/doc/guides/nics/features/dpaa.ini
index b00f46a97..816a6e08e 100644
--- a/doc/guides/nics/features/dpaa.ini
+++ b/doc/guides/nics/features/dpaa.ini
@@ -6,6 +6,7 @@
 [Features]
 Speed capabilities   = Y
 Link status  = Y
+Link status event= Y
 Jumbo frame  = Y
 MTU update   = Y
 Scattered Rx = Y
diff --git a/doc/guides/rel_notes/release_20_08.rst 
b/doc/guides/rel_notes/release_20_08.rst
index b1e039d03..e5bc5cfd8 100644
--- a/doc/guides/rel_notes/release_20_08.rst
+++ b/doc/guides/rel_notes/release_20_08.rst
@@ -123,6 +123,7 @@ New Features
 
   Updated the NXP dpaa ethdev  with new features and improvements, including:
 
+  * Added support for link status and interrupt
   * Added support to use datapath APIs from non-EAL pthread
 
 * **Updated NXP dpaa2 ethdev PMD.**
diff --git a/drivers/bus/dpaa/base/fman/fman.c 
b/drivers/bus/dpaa/base/fman/fman.c
index ae26041ca..33be9e5d7 100644
--- a/drivers/bus/dpaa/base/fman/fman.c
+++ b/drivers/bus/dpaa/base/fman/fman.c
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: (BSD-3-Clause OR GPL-2.0)
  *
  * Copyright 2010-2016 Freescale Semiconductor Inc.
- * Copyright 2017-2019 NXP
+ * Copyright 2017-2020 NXP
  *
  */
 
@@ -185,6 +185,8 @@ fman_if_init(const struct device_node *dpa_node)
}
memset(__if, 0, sizeof(*__if));
INIT_LIST_HEAD(&__if->__if.bpool_list);
+   strlcpy(__if->node_name, dpa_node->name, IF_NAME_MAX_LEN - 1);
+   __if->node_name[IF_NAME_MAX_LEN - 1] = '\0';
strlcpy(__if->node_path, dpa_node->full_name, PATH_MAX - 1);
__if->node_path[PATH_MAX - 1] = '\0';
 
diff --git a/drivers/bus/dpaa/base/qbman/process.c 
b/drivers/bus/dpaa/base/qbman/process.c
index 2c23c98df..68b7af243 100644
--- a/drivers/bus/dpaa/base/qbman/process.c
+++ b/drivers/bus/dpaa/base/qbman/process.c
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: (BSD-3-Clause OR GPL-2.0)
  *
  * Copyright 2011-2016 Freescale Semiconductor Inc.
- * Copyright 2017 NXP
+ * Copyright 2017,2020 NXP
  *
  */
 #include 
@@ -296,3 +296,73 @@ int bman_free_raw_portal(struct dpaa_raw_portal *portal)
 
return process_portal_free(&input);
 }
+
+#define DPAA_IOCTL_ENABLE_LINK_STATUS_INTERRUPT \
+   _IOW(DPAA_IOCTL_MAGIC, 0x0E, struct usdpaa_ioctl_link_status)
+
+#define DPAA_IOCTL_DISABLE_LINK_STATUS_INTERRUPT \
+   _IOW(DPAA_IOCTL_MAGIC, 0x0F, char*)
+
+int dpaa_intr_enable(char *if_name, int efd)
+{
+   struct usdpaa_ioctl_link_status args;
+
+   int ret = check_fd();
+
+   if (ret)
+   return ret;
+
+   args.efd = (uint32_t)efd;
+   strcpy(args.if_name, if_name);
+
+   ret = ioctl(fd, DPAA_IOCTL_ENABLE_LINK_STATUS_INTERRUPT, &args);
+   if (ret)
+   return errno;
+
+   return 0;
+}
+
+int dpaa_intr_disable(char *if_name)
+{
+   int ret = check_fd();
+
+   if (ret)
+   return ret;
+
+   ret = ioctl(fd, DPAA_IOCTL_DISABLE_LINK_STATUS_INTERRUPT, &if_name);
+   if (ret) {
+   if (errno == EINVAL)
+   printf("Failed to disable interrupt: Not Supported\n");
+   else
+   printf("Failed to disable interrupt\n");
+   return ret;
+   }
+
+   return 0;
+}
+
+#define DPAA_IOCTL_GET_LINK_STATUS \
+   _IOWR(DPAA_IOCTL_MAGIC, 0x10, struct usdpaa_ioctl_link_status_args)
+
+int dpaa_get_link_status(char *if_name)
+{
+   int ret = check_fd();
+   struct usdpaa_ioctl_link_status_args args;
+
+   if (ret)
+   return ret;
+
+   strcpy(args.if_name, if_name);
+   args.link_status = 0;
+
+   ret = ioctl(fd, DPAA_IOCTL_GET_LINK_STATUS, &args);
+   if (ret) {
+   if (errno == EINVAL)
+   printf("Failed to get link status: Not Supported\n");
+   else
+   printf("Failed to get link status\n");
+   return ret;
+   }
+
+   return args.link_status;
+}
diff --git a/drivers/bus/dpaa/dpaa_bus.c b/drivers/bus/dpaa/dpaa_bus.c
index aa906c34e..32e872da5 100644
--- a/drivers/bus/dpaa/dpaa_bus.c
+++ b/drivers/bus/dpaa/dpaa

[dpdk-dev] [PATCH v2 20/29] net/dpaa2: define the size of table entry

[Hemant Agrawal]

From: Jun Yang 

If entry size is not bigger than 27, MC alloc one TCAM entry,
otherwise, alloc 2 TCAM entries.
Extracts size by HW must be not bigger than TCAM entry size(27 or 54).
So define the flow entry size as 54.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 90 ++
 1 file changed, 60 insertions(+), 30 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 760a8a793..bcbd5977a 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -29,6 +29,8 @@
  */
 int mc_l4_port_identification;
 
+#define FIXED_ENTRY_SIZE 54
+
 enum flow_rule_ipaddr_type {
FLOW_NONE_IPADDR,
FLOW_IPV4_ADDR,
@@ -47,7 +49,8 @@ struct rte_flow {
LIST_ENTRY(rte_flow) next; /**< Pointer to the next flow structure. */
struct dpni_rule_cfg qos_rule;
struct dpni_rule_cfg fs_rule;
-   uint8_t key_size;
+   uint8_t qos_real_key_size;
+   uint8_t fs_real_key_size;
uint8_t tc_id; /** Traffic Class ID. */
uint8_t tc_index; /** index within this Traffic Class. */
enum rte_flow_action_type action;
@@ -478,6 +481,7 @@ dpaa2_flow_rule_data_set(
prot, field);
return -1;
}
+
memcpy((void *)(size_t)(rule->key_iova + offset), key, size);
memcpy((void *)(size_t)(rule->mask_iova + offset), mask, size);
 
@@ -523,9 +527,11 @@ _dpaa2_flow_rule_move_ipaddr_tail(
len = NH_FLD_IPV6_ADDR_SIZE;
 
memcpy(tmp, (char *)key_src, len);
+   memset((char *)key_src, 0, len);
memcpy((char *)key_dst, tmp, len);
 
memcpy(tmp, (char *)mask_src, len);
+   memset((char *)mask_src, 0, len);
memcpy((char *)mask_dst, tmp, len);
 
return 0;
@@ -1251,8 +1257,7 @@ dpaa2_configure_flow_generic_ip(
 
return -1;
}
-   local_cfg |= (DPAA2_QOS_TABLE_RECONFIGURE |
-   DPAA2_QOS_TABLE_IPADDR_EXTRACT);
+   local_cfg |= DPAA2_QOS_TABLE_RECONFIGURE;
}
 
index = dpaa2_flow_extract_search(
@@ -1269,8 +1274,7 @@ dpaa2_configure_flow_generic_ip(
 
return -1;
}
-   local_cfg |= (DPAA2_FS_TABLE_RECONFIGURE |
-   DPAA2_FS_TABLE_IPADDR_EXTRACT);
+   local_cfg |= DPAA2_FS_TABLE_RECONFIGURE;
}
 
if (spec_ipv4)
@@ -1339,8 +1343,7 @@ dpaa2_configure_flow_generic_ip(
 
return -1;
}
-   local_cfg |= (DPAA2_QOS_TABLE_RECONFIGURE |
-   DPAA2_QOS_TABLE_IPADDR_EXTRACT);
+   local_cfg |= DPAA2_QOS_TABLE_RECONFIGURE;
}
 
index = dpaa2_flow_extract_search(
@@ -1361,8 +1364,7 @@ dpaa2_configure_flow_generic_ip(
 
return -1;
}
-   local_cfg |= (DPAA2_FS_TABLE_RECONFIGURE |
-   DPAA2_FS_TABLE_IPADDR_EXTRACT);
+   local_cfg |= DPAA2_FS_TABLE_RECONFIGURE;
}
 
if (spec_ipv4)
@@ -2641,7 +2643,7 @@ dpaa2_flow_entry_update(
char ipdst_key[NH_FLD_IPV6_ADDR_SIZE];
char ipsrc_mask[NH_FLD_IPV6_ADDR_SIZE];
char ipdst_mask[NH_FLD_IPV6_ADDR_SIZE];
-   int extend = -1, extend1, size;
+   int extend = -1, extend1, size = -1;
uint16_t qos_index;
 
while (curr) {
@@ -2696,6 +2698,9 @@ dpaa2_flow_entry_update(
else
extend = extend1;
 
+   RTE_ASSERT((size == NH_FLD_IPV4_ADDR_SIZE) ||
+   (size == NH_FLD_IPV6_ADDR_SIZE));
+
memcpy(ipsrc_key,
(char *)(size_t)curr->qos_rule.key_iova +
curr->ipaddr_rule.qos_ipsrc_offset,
@@ -2725,6 +2730,9 @@ dpaa2_flow_entry_update(
else
extend = extend1;
 
+   RTE_ASSERT((size == NH_FLD_IPV4_ADDR_SIZE) ||
+   (size == NH_FLD_IPV6_ADDR_SIZE));
+
memcpy(ipdst_key,
(char *)(size_t)curr->qos_rule.key_iova +
curr->ipaddr_rule.qos_ipdst_offset,
@@ -2745,6 +2753,8 @@ dpaa2_flow_entry_update(
}
 
if (curr->ipaddr_rule.qos_ipsrc_offset >= 0) {
+   RTE_ASSERT((size == NH_FLD_IPV4_ADDR_SIZE) ||
+   (size == NH_FLD_IPV6_ADDR_SIZE));
memcpy((char *)(size_t)curr->qos_rule.key_iova +
curr->ipaddr_rule.qos_ipsrc_offset,
   

[dpdk-dev] [PATCH v2 18/29] net/dpaa2: free flow rule memory

[Hemant Agrawal]

From: Jun Yang 

Free rule memory when the flow is destroyed.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 5 +
 1 file changed, 5 insertions(+)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 507a5d0e3..941d62b80 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -3594,6 +3594,7 @@ int dpaa2_flow_destroy(struct rte_eth_dev *dev,
"Error in entry addition in QoS table(%d)", ret);
goto error;
}
+   priv->qos_index[flow->qos_index] = 0;
break;
default:
DPAA2_PMD_ERR(
@@ -3603,6 +3604,10 @@ int dpaa2_flow_destroy(struct rte_eth_dev *dev,
}
 
LIST_REMOVE(flow, next);
+   rte_free((void *)(size_t)flow->qos_rule.key_iova);
+   rte_free((void *)(size_t)flow->qos_rule.mask_iova);
+   rte_free((void *)(size_t)flow->fs_rule.key_iova);
+   rte_free((void *)(size_t)flow->fs_rule.mask_iova);
/* Now free the flow */
rte_free(flow);
 
-- 
2.17.1

[dpdk-dev] [PATCH v2 19/29] net/dpaa2: support QoS or FS table entry indexing

[Hemant Agrawal]

From: Jun Yang 

Calculate QoS/FS entry index by group and priority of flow.

1)The less index of entry, the higher priority of flow.
2)Verify if the flow with same group and priority has been added before
  creating flow.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_ethdev.c |   4 +
 drivers/net/dpaa2/dpaa2_ethdev.h |   5 +-
 drivers/net/dpaa2/dpaa2_flow.c   | 127 +--
 3 files changed, 59 insertions(+), 77 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index fd3097c7d..008e1c570 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -2392,6 +2392,10 @@ dpaa2_dev_init(struct rte_eth_dev *eth_dev)
}
 
priv->num_rx_tc = attr.num_rx_tcs;
+   priv->qos_entries = attr.qos_entries;
+   priv->fs_entries = attr.fs_entries;
+   priv->dist_queues = attr.num_queues;
+
/* only if the custom CG is enabled */
if (attr.options & DPNI_OPT_CUSTOM_CG)
priv->max_cgs = attr.num_cgs;
diff --git a/drivers/net/dpaa2/dpaa2_ethdev.h b/drivers/net/dpaa2/dpaa2_ethdev.h
index 030c625e3..b49b88a2d 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.h
+++ b/drivers/net/dpaa2/dpaa2_ethdev.h
@@ -145,6 +145,9 @@ struct dpaa2_dev_priv {
uint8_t max_mac_filters;
uint8_t max_vlan_filters;
uint8_t num_rx_tc;
+   uint16_t qos_entries;
+   uint16_t fs_entries;
+   uint8_t dist_queues;
uint8_t flags; /*dpaa2 config flags */
uint8_t en_ordered;
uint8_t en_loose_ordered;
@@ -152,8 +155,6 @@ struct dpaa2_dev_priv {
uint8_t cgid_in_use[MAX_RX_QUEUES];
 
struct extract_s extract;
-   uint8_t *qos_index;
-   uint8_t *fs_index;
 
uint16_t ss_offset;
uint64_t ss_iova;
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 941d62b80..760a8a793 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -47,11 +47,8 @@ struct rte_flow {
LIST_ENTRY(rte_flow) next; /**< Pointer to the next flow structure. */
struct dpni_rule_cfg qos_rule;
struct dpni_rule_cfg fs_rule;
-   uint16_t qos_index;
-   uint16_t fs_index;
uint8_t key_size;
uint8_t tc_id; /** Traffic Class ID. */
-   uint8_t flow_type;
uint8_t tc_index; /** index within this Traffic Class. */
enum rte_flow_action_type action;
uint16_t flow_id;
@@ -2645,6 +2642,7 @@ dpaa2_flow_entry_update(
char ipsrc_mask[NH_FLD_IPV6_ADDR_SIZE];
char ipdst_mask[NH_FLD_IPV6_ADDR_SIZE];
int extend = -1, extend1, size;
+   uint16_t qos_index;
 
while (curr) {
if (curr->ipaddr_rule.ipaddr_type ==
@@ -2676,6 +2674,9 @@ dpaa2_flow_entry_update(
size = NH_FLD_IPV6_ADDR_SIZE;
}
 
+   qos_index = curr->tc_id * priv->fs_entries +
+   curr->tc_index;
+
ret = dpni_remove_qos_entry(dpni, CMD_PRI_LOW,
priv->token, &curr->qos_rule);
if (ret) {
@@ -2769,7 +2770,7 @@ dpaa2_flow_entry_update(
 
ret = dpni_add_qos_entry(dpni, CMD_PRI_LOW,
priv->token, &curr->qos_rule,
-   curr->tc_id, curr->qos_index,
+   curr->tc_id, qos_index,
0, 0);
if (ret) {
DPAA2_PMD_ERR("Qos entry update failed.");
@@ -2875,7 +2876,7 @@ dpaa2_flow_entry_update(
curr->fs_rule.key_size += extend;
 
ret = dpni_add_fs_entry(dpni, CMD_PRI_LOW,
-   priv->token, curr->tc_id, curr->fs_index,
+   priv->token, curr->tc_id, curr->tc_index,
&curr->fs_rule, &curr->action_cfg);
if (ret) {
DPAA2_PMD_ERR("FS entry update failed.");
@@ -2888,6 +2889,28 @@ dpaa2_flow_entry_update(
return 0;
 }
 
+static inline int
+dpaa2_flow_verify_attr(
+   struct dpaa2_dev_priv *priv,
+   const struct rte_flow_attr *attr)
+{
+   struct rte_flow *curr = LIST_FIRST(&priv->flows);
+
+   while (curr) {
+   if (curr->tc_id == attr->group &&
+   curr->tc_index == attr->priority) {
+   DPAA2_PMD_ERR(
+   "Flow with group %d and priority %d already 
exists.",
+   attr->group, attr->priority);
+
+   return -1;
+   }
+   curr = LIST_NEXT(curr, next);
+   }
+
+   return 0;
+}
+
 static int
 dpaa2_generic_flow_set(struct rte_flow *flow,
   struct rte_eth_dev *dev,
@@ -2898,10 +2921,8 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
 {
const struct rte_flow_action_queue *dest_queue;
 

[dpdk-dev] [PATCH v2 16/29] net/dpaa2: support key extracts of flow API

[Hemant Agrawal]

From: Jun Yang 

1) Support QoS extracts and TC extracts for multiple TCs.

2) Protocol type of L2 extract is used to parse L3.
   Next protocol of L3 extract is used to parse L4.

3) generic IP key extracts instead of IPv4 and IPv6 respectively.

4) Special for IP address extracts:
   Put IP(v4/v6) address extract(s)/rule(s) at the end of extracts array
   to make rest fields at fixed poisition.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_ethdev.c |   35 +-
 drivers/net/dpaa2/dpaa2_ethdev.h |   43 +-
 drivers/net/dpaa2/dpaa2_flow.c   | 3628 +-
 3 files changed, 2665 insertions(+), 1041 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index 8edd4b3cd..492b65840 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -1,7 +1,7 @@
 /* * SPDX-License-Identifier: BSD-3-Clause
  *
  *   Copyright (c) 2016 Freescale Semiconductor, Inc. All rights reserved.
- *   Copyright 2016 NXP
+ *   Copyright 2016-2020 NXP
  *
  */
 
@@ -2501,23 +2501,41 @@ dpaa2_dev_init(struct rte_eth_dev *eth_dev)
eth_dev->tx_pkt_burst = dpaa2_dev_tx;
 
/*Init fields w.r.t. classficaition*/
-   memset(&priv->extract.qos_key_cfg, 0, sizeof(struct dpkg_profile_cfg));
+   memset(&priv->extract.qos_key_extract, 0,
+   sizeof(struct dpaa2_key_extract));
priv->extract.qos_extract_param = (size_t)rte_malloc(NULL, 256, 64);
if (!priv->extract.qos_extract_param) {
DPAA2_PMD_ERR(" Error(%d) in allocation resources for flow "
" classificaiton ", ret);
goto init_err;
}
+   priv->extract.qos_key_extract.key_info.ipv4_src_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.qos_key_extract.key_info.ipv4_dst_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.qos_key_extract.key_info.ipv6_src_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.qos_key_extract.key_info.ipv6_dst_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+
for (i = 0; i < MAX_TCS; i++) {
-   memset(&priv->extract.fs_key_cfg[i], 0,
-   sizeof(struct dpkg_profile_cfg));
-   priv->extract.fs_extract_param[i] =
+   memset(&priv->extract.tc_key_extract[i], 0,
+   sizeof(struct dpaa2_key_extract));
+   priv->extract.tc_extract_param[i] =
(size_t)rte_malloc(NULL, 256, 64);
-   if (!priv->extract.fs_extract_param[i]) {
+   if (!priv->extract.tc_extract_param[i]) {
DPAA2_PMD_ERR(" Error(%d) in allocation resources for 
flow classificaiton",
 ret);
goto init_err;
}
+   priv->extract.tc_key_extract[i].key_info.ipv4_src_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.tc_key_extract[i].key_info.ipv4_dst_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.tc_key_extract[i].key_info.ipv6_src_offset =
+   IP_ADDRESS_OFFSET_INVALID;
+   priv->extract.tc_key_extract[i].key_info.ipv6_dst_offset =
+   IP_ADDRESS_OFFSET_INVALID;
}
 
ret = dpni_set_max_frame_length(dpni_dev, CMD_PRI_LOW, priv->token,
@@ -2593,8 +2611,9 @@ dpaa2_dev_uninit(struct rte_eth_dev *eth_dev)
rte_free(dpni);
 
for (i = 0; i < MAX_TCS; i++) {
-   if (priv->extract.fs_extract_param[i])
-   rte_free((void 
*)(size_t)priv->extract.fs_extract_param[i]);
+   if (priv->extract.tc_extract_param[i])
+   rte_free((void *)
+   (size_t)priv->extract.tc_extract_param[i]);
}
 
if (priv->extract.qos_extract_param)
diff --git a/drivers/net/dpaa2/dpaa2_ethdev.h b/drivers/net/dpaa2/dpaa2_ethdev.h
index c7fb6539f..030c625e3 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.h
+++ b/drivers/net/dpaa2/dpaa2_ethdev.h
@@ -96,10 +96,39 @@ extern enum pmd_dpaa2_ts dpaa2_enable_ts;
 #define DPAA2_QOS_TABLE_RECONFIGURE1
 #define DPAA2_FS_TABLE_RECONFIGURE 2
 
+#define DPAA2_QOS_TABLE_IPADDR_EXTRACT 4
+#define DPAA2_FS_TABLE_IPADDR_EXTRACT 8
+
+
 /*Externaly defined*/
 extern const struct rte_flow_ops dpaa2_flow_ops;
 extern enum rte_filter_type dpaa2_filter_type;
 
+#define IP_ADDRESS_OFFSET_INVALID (-1)
+
+struct dpaa2_key_info {
+   uint8_t key_offset[DPKG_MAX_NUM_OF_EXTRACTS];
+   uint8_t key_size[DPKG_MAX_NUM_OF_EXTRACTS];
+   /* Special for IP address. */
+   int ipv4_src_offset;
+   int ipv4_dst_offset;
+   int ipv6_src_offset;
+   int ipv6_dst_offset;
+   uint8_t key_total_size;
+};
+
+struct dpaa2_key_extract {
+   struct dpkg_profile_cfg dpkg;
+   struct dpaa2_key_info key_info;
+};
+
+

[dpdk-dev] [PATCH v2 24/29] net/dpaa2: support ndex of queue action for flow

[Hemant Agrawal]

From: Jun Yang 

Make more sense to use RXQ index for queue distribution

instead of flow ID.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 27 +--
 1 file changed, 17 insertions(+), 10 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 6f3139f86..76f68b903 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -56,7 +56,6 @@ struct rte_flow {
uint8_t tc_id; /** Traffic Class ID. */
uint8_t tc_index; /** index within this Traffic Class. */
enum rte_flow_action_type action;
-   uint16_t flow_id;
/* Special for IP address to specify the offset
 * in key/mask.
 */
@@ -3141,6 +3140,7 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
struct dpni_qos_tbl_cfg qos_cfg;
struct dpni_fs_action_cfg action;
struct dpaa2_dev_priv *priv = dev->data->dev_private;
+   struct dpaa2_queue *rxq;
struct fsl_mc_io *dpni = (struct fsl_mc_io *)priv->hw;
size_t param;
struct rte_flow *curr = LIST_FIRST(&priv->flows);
@@ -3244,10 +3244,10 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
case RTE_FLOW_ACTION_TYPE_QUEUE:
dest_queue =
(const struct rte_flow_action_queue 
*)(actions[j].conf);
-   flow->flow_id = dest_queue->index;
+   rxq = priv->rx_vq[dest_queue->index];
flow->action = RTE_FLOW_ACTION_TYPE_QUEUE;
memset(&action, 0, sizeof(struct dpni_fs_action_cfg));
-   action.flow_id = flow->flow_id;
+   action.flow_id = rxq->flow_id;
if (is_keycfg_configured & DPAA2_QOS_TABLE_RECONFIGURE) 
{
dpaa2_flow_qos_table_extracts_log(priv);
if (dpkg_prepare_key_cfg(
@@ -3303,8 +3303,6 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
}
/* Configure QoS table first */
 
-   action.flow_id = action.flow_id % priv->num_rx_tc;
-
qos_index = flow->tc_id * priv->fs_entries +
flow->tc_index;
 
@@ -3407,13 +3405,22 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
break;
case RTE_FLOW_ACTION_TYPE_RSS:
rss_conf = (const struct rte_flow_action_rss 
*)(actions[j].conf);
+   if (rss_conf->queue_num > priv->dist_queues) {
+   DPAA2_PMD_ERR(
+   "RSS number exceeds the distrbution 
size");
+   return -ENOTSUP;
+   }
+
for (i = 0; i < (int)rss_conf->queue_num; i++) {
-   if (rss_conf->queue[i] <
-   (attr->group * priv->dist_queues) ||
-   rss_conf->queue[i] >=
-   ((attr->group + 1) * 
priv->dist_queues)) {
+   if (rss_conf->queue[i] >= priv->nb_rx_queues) {
+   DPAA2_PMD_ERR(
+   "RSS RXQ number exceeds the 
total number");
+   return -ENOTSUP;
+   }
+   rxq = priv->rx_vq[rss_conf->queue[i]];
+   if (rxq->tc_index != attr->group) {
DPAA2_PMD_ERR(
-   "Queue/Group combination are not 
supported\n");
+   "RSS RXQ distributed is not in 
current group");
return -ENOTSUP;
}
}
-- 
2.17.1

[dpdk-dev] [PATCH v2 21/29] net/dpaa2: add logging of flow extracts and rules

[Hemant Agrawal]

From: Jun Yang 

This patch add support for logging the flow rules.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 213 -
 1 file changed, 209 insertions(+), 4 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index bcbd5977a..95756bf7b 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -29,6 +29,8 @@
  */
 int mc_l4_port_identification;
 
+static char *dpaa2_flow_control_log;
+
 #define FIXED_ENTRY_SIZE 54
 
 enum flow_rule_ipaddr_type {
@@ -149,6 +151,189 @@ static const struct rte_flow_item_gre 
dpaa2_flow_item_gre_mask = {
 
 #endif
 
+static inline void dpaa2_prot_field_string(
+   enum net_prot prot, uint32_t field,
+   char *string)
+{
+   if (!dpaa2_flow_control_log)
+   return;
+
+   if (prot == NET_PROT_ETH) {
+   strcpy(string, "eth");
+   if (field == NH_FLD_ETH_DA)
+   strcat(string, ".dst");
+   else if (field == NH_FLD_ETH_SA)
+   strcat(string, ".src");
+   else if (field == NH_FLD_ETH_TYPE)
+   strcat(string, ".type");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_VLAN) {
+   strcpy(string, "vlan");
+   if (field == NH_FLD_VLAN_TCI)
+   strcat(string, ".tci");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_IP) {
+   strcpy(string, "ip");
+   if (field == NH_FLD_IP_SRC)
+   strcat(string, ".src");
+   else if (field == NH_FLD_IP_DST)
+   strcat(string, ".dst");
+   else if (field == NH_FLD_IP_PROTO)
+   strcat(string, ".proto");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_TCP) {
+   strcpy(string, "tcp");
+   if (field == NH_FLD_TCP_PORT_SRC)
+   strcat(string, ".src");
+   else if (field == NH_FLD_TCP_PORT_DST)
+   strcat(string, ".dst");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_UDP) {
+   strcpy(string, "udp");
+   if (field == NH_FLD_UDP_PORT_SRC)
+   strcat(string, ".src");
+   else if (field == NH_FLD_UDP_PORT_DST)
+   strcat(string, ".dst");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_ICMP) {
+   strcpy(string, "icmp");
+   if (field == NH_FLD_ICMP_TYPE)
+   strcat(string, ".type");
+   else if (field == NH_FLD_ICMP_CODE)
+   strcat(string, ".code");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_SCTP) {
+   strcpy(string, "sctp");
+   if (field == NH_FLD_SCTP_PORT_SRC)
+   strcat(string, ".src");
+   else if (field == NH_FLD_SCTP_PORT_DST)
+   strcat(string, ".dst");
+   else
+   strcat(string, ".unknown field");
+   } else if (prot == NET_PROT_GRE) {
+   strcpy(string, "gre");
+   if (field == NH_FLD_GRE_TYPE)
+   strcat(string, ".type");
+   else
+   strcat(string, ".unknown field");
+   } else {
+   strcpy(string, "unknown protocol");
+   }
+}
+
+static inline void dpaa2_flow_qos_table_extracts_log(
+   const struct dpaa2_dev_priv *priv)
+{
+   int idx;
+   char string[32];
+
+   if (!dpaa2_flow_control_log)
+   return;
+
+   printf("Setup QoS table: number of extracts: %d\r\n",
+   priv->extract.qos_key_extract.dpkg.num_extracts);
+   for (idx = 0; idx < priv->extract.qos_key_extract.dpkg.num_extracts;
+   idx++) {
+   dpaa2_prot_field_string(priv->extract.qos_key_extract.dpkg
+   .extracts[idx].extract.from_hdr.prot,
+   priv->extract.qos_key_extract.dpkg.extracts[idx]
+   .extract.from_hdr.field,
+   string);
+   printf("%s", string);
+   if ((idx + 1) < priv->extract.qos_key_extract.dpkg.num_extracts)
+   printf(" / ");
+   }
+   printf("\r\n");
+}
+
+static inline void dpaa2_flow_fs_table_extracts_log(
+   const struct dpaa2_dev_priv *priv, int tc_id)
+{
+   int idx;
+   char string[32];
+
+   if (!dpaa2_flow_control_log)
+   return;
+
+   printf("Setup FS table: number of extracts of TC[%d]: %

[dpdk-dev] [PATCH v2 22/29] net/dpaa2: support iscrimination between IPv4 and IPv6

[Hemant Agrawal]

From: Jun Yang 

Discriminate between IPv4 and IPv6 in generic IP flow setup.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 153 +
 1 file changed, 80 insertions(+), 73 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 95756bf7b..6f3139f86 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -1284,6 +1284,70 @@ dpaa2_configure_flow_vlan(struct rte_flow *flow,
return 0;
 }
 
+static int
+dpaa2_configure_flow_ip_discrimation(
+   struct dpaa2_dev_priv *priv, struct rte_flow *flow,
+   const struct rte_flow_item *pattern,
+   int *local_cfg, int *device_configured,
+   uint32_t group)
+{
+   int index, ret;
+   struct proto_discrimination proto;
+
+   index = dpaa2_flow_extract_search(
+   &priv->extract.qos_key_extract.dpkg,
+   NET_PROT_ETH, NH_FLD_ETH_TYPE);
+   if (index < 0) {
+   ret = dpaa2_flow_proto_discrimination_extract(
+   &priv->extract.qos_key_extract,
+   RTE_FLOW_ITEM_TYPE_ETH);
+   if (ret) {
+   DPAA2_PMD_ERR(
+   "QoS Extract ETH_TYPE to discriminate IP failed.");
+   return -1;
+   }
+   (*local_cfg) |= DPAA2_QOS_TABLE_RECONFIGURE;
+   }
+
+   index = dpaa2_flow_extract_search(
+   &priv->extract.tc_key_extract[group].dpkg,
+   NET_PROT_ETH, NH_FLD_ETH_TYPE);
+   if (index < 0) {
+   ret = dpaa2_flow_proto_discrimination_extract(
+   &priv->extract.tc_key_extract[group],
+   RTE_FLOW_ITEM_TYPE_ETH);
+   if (ret) {
+   DPAA2_PMD_ERR(
+   "FS Extract ETH_TYPE to discriminate IP failed.");
+   return -1;
+   }
+   (*local_cfg) |= DPAA2_FS_TABLE_RECONFIGURE;
+   }
+
+   ret = dpaa2_flow_rule_move_ipaddr_tail(flow, priv, group);
+   if (ret) {
+   DPAA2_PMD_ERR(
+   "Move ipaddr before IP discrimination set failed");
+   return -1;
+   }
+
+   proto.type = RTE_FLOW_ITEM_TYPE_ETH;
+   if (pattern->type == RTE_FLOW_ITEM_TYPE_IPV4)
+   proto.eth_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4);
+   else
+   proto.eth_type = rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV6);
+   ret = dpaa2_flow_proto_discrimination_rule(priv, flow, proto, group);
+   if (ret) {
+   DPAA2_PMD_ERR("IP discrimination rule set failed");
+   return -1;
+   }
+
+   (*device_configured) |= (*local_cfg);
+
+   return 0;
+}
+
+
 static int
 dpaa2_configure_flow_generic_ip(
struct rte_flow *flow,
@@ -1327,73 +1391,16 @@ dpaa2_configure_flow_generic_ip(
flow->tc_id = group;
flow->tc_index = attr->priority;
 
-   if (!spec_ipv4 && !spec_ipv6) {
-   /* Don't care any field of IP header,
-* only care IP protocol.
-* Example: flow create 0 ingress pattern ipv6 /
-*/
-   /* Eth type is actually used for IP identification.
-*/
-   /* TODO: Current design only supports Eth + IP,
-*  Eth + vLan + IP needs to add.
-*/
-   struct proto_discrimination proto;
-
-   index = dpaa2_flow_extract_search(
-   &priv->extract.qos_key_extract.dpkg,
-   NET_PROT_ETH, NH_FLD_ETH_TYPE);
-   if (index < 0) {
-   ret = dpaa2_flow_proto_discrimination_extract(
-   &priv->extract.qos_key_extract,
-   RTE_FLOW_ITEM_TYPE_ETH);
-   if (ret) {
-   DPAA2_PMD_ERR(
-   "QoS Ext ETH_TYPE to discriminate IP failed.");
-
-   return -1;
-   }
-   local_cfg |= DPAA2_QOS_TABLE_RECONFIGURE;
-   }
-
-   index = dpaa2_flow_extract_search(
-   &priv->extract.tc_key_extract[group].dpkg,
-   NET_PROT_ETH, NH_FLD_ETH_TYPE);
-   if (index < 0) {
-   ret = dpaa2_flow_proto_discrimination_extract(
-   &priv->extract.tc_key_extract[group],
-   RTE_FLOW_ITEM_TYPE_ETH);
-   if (ret) {
-   DPAA2_PMD_ERR(
-   "FS Ext ETH_TYPE to discriminate IP failed");
-
-   return -1;
-   }
-   

[dpdk-dev] [PATCH v2 23/29] net/dpaa2: support distribution size set on multiple TCs

[Hemant Agrawal]

From: Jun Yang 

Default distribution size of TC is 1, which is limited by MC. We have to
set the distribution size for each TC to support multiple RXQs per TC.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/base/dpaa2_hw_dpni.c |  6 +--
 drivers/net/dpaa2/dpaa2_ethdev.c   | 51 --
 drivers/net/dpaa2/dpaa2_ethdev.h   |  2 +-
 3 files changed, 36 insertions(+), 23 deletions(-)

diff --git a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c 
b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
index 34de0d1f7..9f0dad6e7 100644
--- a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
+++ b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
@@ -81,14 +81,14 @@ rte_pmd_dpaa2_set_custom_hash(uint16_t port_id,
 
 int
 dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
- uint64_t req_dist_set)
+   uint64_t req_dist_set, int tc_index)
 {
struct dpaa2_dev_priv *priv = eth_dev->data->dev_private;
struct fsl_mc_io *dpni = priv->hw;
struct dpni_rx_tc_dist_cfg tc_cfg;
struct dpkg_profile_cfg kg_cfg;
void *p_params;
-   int ret, tc_index = 0;
+   int ret;
 
p_params = rte_malloc(
NULL, DIST_PARAM_IOVA_SIZE, RTE_CACHE_LINE_SIZE);
@@ -107,7 +107,7 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
return ret;
}
tc_cfg.key_cfg_iova = (uint64_t)(DPAA2_VADDR_TO_IOVA(p_params));
-   tc_cfg.dist_size = eth_dev->data->nb_rx_queues;
+   tc_cfg.dist_size = priv->dist_queues;
tc_cfg.dist_mode = DPNI_DIST_MODE_HASH;
 
ret = dpkg_prepare_key_cfg(&kg_cfg, p_params);
diff --git a/drivers/net/dpaa2/dpaa2_ethdev.c b/drivers/net/dpaa2/dpaa2_ethdev.c
index 008e1c570..020af4b03 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.c
+++ b/drivers/net/dpaa2/dpaa2_ethdev.c
@@ -453,7 +453,7 @@ dpaa2_eth_dev_configure(struct rte_eth_dev *dev)
int rx_l4_csum_offload = false;
int tx_l3_csum_offload = false;
int tx_l4_csum_offload = false;
-   int ret;
+   int ret, tc_index;
 
PMD_INIT_FUNC_TRACE();
 
@@ -493,12 +493,16 @@ dpaa2_eth_dev_configure(struct rte_eth_dev *dev)
}
 
if (eth_conf->rxmode.mq_mode == ETH_MQ_RX_RSS) {
-   ret = dpaa2_setup_flow_dist(dev,
-   eth_conf->rx_adv_conf.rss_conf.rss_hf);
-   if (ret) {
-   DPAA2_PMD_ERR("Unable to set flow distribution."
- "Check queue config");
-   return ret;
+   for (tc_index = 0; tc_index < priv->num_rx_tc; tc_index++) {
+   ret = dpaa2_setup_flow_dist(dev,
+   eth_conf->rx_adv_conf.rss_conf.rss_hf,
+   tc_index);
+   if (ret) {
+   DPAA2_PMD_ERR(
+   "Unable to set flow distribution on 
tc%d."
+   "Check queue config", tc_index);
+   return ret;
+   }
}
}
 
@@ -755,11 +759,11 @@ dpaa2_dev_tx_queue_setup(struct rte_eth_dev *dev,
flow_id = 0;
 
ret = dpni_set_queue(dpni, CMD_PRI_LOW, priv->token, DPNI_QUEUE_TX,
-tc_id, flow_id, options, &tx_flow_cfg);
+   tc_id, flow_id, options, &tx_flow_cfg);
if (ret) {
DPAA2_PMD_ERR("Error in setting the tx flow: "
- "tc_id=%d, flow=%d err=%d",
- tc_id, flow_id, ret);
+   "tc_id=%d, flow=%d err=%d",
+   tc_id, flow_id, ret);
return -1;
}
 
@@ -1984,22 +1988,31 @@ dpaa2_dev_rss_hash_update(struct rte_eth_dev *dev,
  struct rte_eth_rss_conf *rss_conf)
 {
struct rte_eth_dev_data *data = dev->data;
+   struct dpaa2_dev_priv *priv = data->dev_private;
struct rte_eth_conf *eth_conf = &data->dev_conf;
-   int ret;
+   int ret, tc_index;
 
PMD_INIT_FUNC_TRACE();
 
if (rss_conf->rss_hf) {
-   ret = dpaa2_setup_flow_dist(dev, rss_conf->rss_hf);
-   if (ret) {
-   DPAA2_PMD_ERR("Unable to set flow dist");
-   return ret;
+   for (tc_index = 0; tc_index < priv->num_rx_tc; tc_index++) {
+   ret = dpaa2_setup_flow_dist(dev, rss_conf->rss_hf,
+   tc_index);
+   if (ret) {
+   DPAA2_PMD_ERR("Unable to set flow dist on tc%d",
+   tc_index);
+   return ret;
+   }
}
} else {
-   ret = dpaa2_remove_flow_dist(dev, 0);
-   if (ret) {
-   DPAA2_PMD_ERR("Unable to remove flow dist");
-   

[dpdk-dev] [PATCH v2 28/29] net/dpaa2: configure per class distribution size

[Hemant Agrawal]

From: Jun Yang 

TC distribution size is set with dist_queues or

nb_rx_queues % dist_queues in order of TC priority.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/base/dpaa2_hw_dpni.c | 18 --
 1 file changed, 16 insertions(+), 2 deletions(-)

diff --git a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c 
b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
index d69156bcc..25b1d2bb6 100644
--- a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
+++ b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
@@ -88,7 +88,21 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
struct dpni_rx_dist_cfg tc_cfg;
struct dpkg_profile_cfg kg_cfg;
void *p_params;
-   int ret;
+   int ret, tc_dist_queues;
+
+   /*TC distribution size is set with dist_queues or
+* nb_rx_queues % dist_queues in order of TC priority index.
+* Calculating dist size for this tc_index:-
+*/
+   tc_dist_queues = eth_dev->data->nb_rx_queues -
+   tc_index * priv->dist_queues;
+   if (tc_dist_queues <= 0) {
+   DPAA2_PMD_INFO("No distribution on TC%d", tc_index);
+   return 0;
+   }
+
+   if (tc_dist_queues > priv->dist_queues)
+   tc_dist_queues = priv->dist_queues;
 
p_params = rte_malloc(
NULL, DIST_PARAM_IOVA_SIZE, RTE_CACHE_LINE_SIZE);
@@ -109,7 +123,7 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
}
 
tc_cfg.key_cfg_iova = (uint64_t)(DPAA2_VADDR_TO_IOVA(p_params));
-   tc_cfg.dist_size = priv->dist_queues;
+   tc_cfg.dist_size = tc_dist_queues;
tc_cfg.enable = true;
tc_cfg.tc = tc_index;
 
-- 
2.17.1

[dpdk-dev] [PATCH v2 27/29] net/dpaa2: support flow API FS miss action configuration

[Hemant Agrawal]

From: Jun Yang 

1) dpni_set_rx_hash_dist and dpni_set_rx_fs_dist used for TC configuration
instead of dpni_set_rx_tc_dist. Otherwise, re-configuration of
default TC of QoS fails.

2) Default miss action is to drop.
"export DPAA2_FLOW_CONTROL_MISS_FLOW=flow_id" is used receive
 the missed packets from flow with flow ID specified.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/base/dpaa2_hw_dpni.c | 30 +++--
 drivers/net/dpaa2/dpaa2_flow.c | 62 ++
 2 files changed, 60 insertions(+), 32 deletions(-)

diff --git a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c 
b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
index 9f0dad6e7..d69156bcc 100644
--- a/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
+++ b/drivers/net/dpaa2/base/dpaa2_hw_dpni.c
@@ -85,7 +85,7 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
 {
struct dpaa2_dev_priv *priv = eth_dev->data->dev_private;
struct fsl_mc_io *dpni = priv->hw;
-   struct dpni_rx_tc_dist_cfg tc_cfg;
+   struct dpni_rx_dist_cfg tc_cfg;
struct dpkg_profile_cfg kg_cfg;
void *p_params;
int ret;
@@ -96,8 +96,9 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
DPAA2_PMD_ERR("Unable to allocate flow-dist parameters");
return -ENOMEM;
}
+
memset(p_params, 0, DIST_PARAM_IOVA_SIZE);
-   memset(&tc_cfg, 0, sizeof(struct dpni_rx_tc_dist_cfg));
+   memset(&tc_cfg, 0, sizeof(struct dpni_rx_dist_cfg));
 
ret = dpaa2_distset_to_dpkg_profile_cfg(req_dist_set, &kg_cfg);
if (ret) {
@@ -106,9 +107,11 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
rte_free(p_params);
return ret;
}
+
tc_cfg.key_cfg_iova = (uint64_t)(DPAA2_VADDR_TO_IOVA(p_params));
tc_cfg.dist_size = priv->dist_queues;
-   tc_cfg.dist_mode = DPNI_DIST_MODE_HASH;
+   tc_cfg.enable = true;
+   tc_cfg.tc = tc_index;
 
ret = dpkg_prepare_key_cfg(&kg_cfg, p_params);
if (ret) {
@@ -117,8 +120,7 @@ dpaa2_setup_flow_dist(struct rte_eth_dev *eth_dev,
return ret;
}
 
-   ret = dpni_set_rx_tc_dist(dpni, CMD_PRI_LOW, priv->token, tc_index,
- &tc_cfg);
+   ret = dpni_set_rx_hash_dist(dpni, CMD_PRI_LOW, priv->token, &tc_cfg);
rte_free(p_params);
if (ret) {
DPAA2_PMD_ERR(
@@ -136,7 +138,7 @@ int dpaa2_remove_flow_dist(
 {
struct dpaa2_dev_priv *priv = eth_dev->data->dev_private;
struct fsl_mc_io *dpni = priv->hw;
-   struct dpni_rx_tc_dist_cfg tc_cfg;
+   struct dpni_rx_dist_cfg tc_cfg;
struct dpkg_profile_cfg kg_cfg;
void *p_params;
int ret;
@@ -147,13 +149,15 @@ int dpaa2_remove_flow_dist(
DPAA2_PMD_ERR("Unable to allocate flow-dist parameters");
return -ENOMEM;
}
-   memset(p_params, 0, DIST_PARAM_IOVA_SIZE);
-   memset(&tc_cfg, 0, sizeof(struct dpni_rx_tc_dist_cfg));
-   kg_cfg.num_extracts = 0;
-   tc_cfg.key_cfg_iova = (uint64_t)(DPAA2_VADDR_TO_IOVA(p_params));
+
+   memset(&tc_cfg, 0, sizeof(struct dpni_rx_dist_cfg));
tc_cfg.dist_size = 0;
-   tc_cfg.dist_mode = DPNI_DIST_MODE_NONE;
+   tc_cfg.key_cfg_iova = (uint64_t)(DPAA2_VADDR_TO_IOVA(p_params));
+   tc_cfg.enable = true;
+   tc_cfg.tc = tc_index;
 
+   memset(p_params, 0, DIST_PARAM_IOVA_SIZE);
+   kg_cfg.num_extracts = 0;
ret = dpkg_prepare_key_cfg(&kg_cfg, p_params);
if (ret) {
DPAA2_PMD_ERR("Unable to prepare extract parameters");
@@ -161,8 +165,8 @@ int dpaa2_remove_flow_dist(
return ret;
}
 
-   ret = dpni_set_rx_tc_dist(dpni, CMD_PRI_LOW, priv->token, tc_index,
- &tc_cfg);
+   ret = dpni_set_rx_hash_dist(dpni, CMD_PRI_LOW, priv->token,
+   &tc_cfg);
rte_free(p_params);
if (ret)
DPAA2_PMD_ERR(
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 9239fa459..cc789346a 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -30,6 +30,8 @@
 int mc_l4_port_identification;
 
 static char *dpaa2_flow_control_log;
+static int dpaa2_flow_miss_flow_id =
+   DPNI_FS_MISS_DROP;
 
 #define FIXED_ENTRY_SIZE 54
 
@@ -3201,7 +3203,7 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
const struct rte_flow_action_rss *rss_conf;
int is_keycfg_configured = 0, end_of_list = 0;
int ret = 0, i = 0, j = 0;
-   struct dpni_rx_tc_dist_cfg tc_cfg;
+   struct dpni_rx_dist_cfg tc_cfg;
struct dpni_qos_tbl_cfg qos_cfg;
struct dpni_fs_action_cfg action;
struct dpaa2_dev_priv *priv = dev->data->dev_private;
@@ -3330,20 +3332,30 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
return -1;
}
 
- 

[dpdk-dev] [PATCH v2 29/29] net/dpaa2: support raw flow classification

[Hemant Agrawal]

From: Nipun Gupta 

Add support for raw flow, which can be used for any
protocol rules.

Signed-off-by: Nipun Gupta 
---
 drivers/net/dpaa2/dpaa2_ethdev.h |   3 +-
 drivers/net/dpaa2/dpaa2_flow.c   | 135 +++
 2 files changed, 137 insertions(+), 1 deletion(-)

diff --git a/drivers/net/dpaa2/dpaa2_ethdev.h b/drivers/net/dpaa2/dpaa2_ethdev.h
index 52faeeefe..2bc0f3f5a 100644
--- a/drivers/net/dpaa2/dpaa2_ethdev.h
+++ b/drivers/net/dpaa2/dpaa2_ethdev.h
@@ -1,7 +1,7 @@
 /* SPDX-License-Identifier: BSD-3-Clause
  *
  *   Copyright (c) 2015-2016 Freescale Semiconductor, Inc. All rights reserved.
- *   Copyright 2016-2019 NXP
+ *   Copyright 2016-2020 NXP
  *
  */
 
@@ -99,6 +99,7 @@ extern enum pmd_dpaa2_ts dpaa2_enable_ts;
 #define DPAA2_QOS_TABLE_IPADDR_EXTRACT 4
 #define DPAA2_FS_TABLE_IPADDR_EXTRACT 8
 
+#define DPAA2_FLOW_MAX_KEY_SIZE16
 
 /*Externaly defined*/
 extern const struct rte_flow_ops dpaa2_flow_ops;
diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index cc789346a..136bdd5fa 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -493,6 +493,42 @@ static int dpaa2_flow_extract_add(
return 0;
 }
 
+static int dpaa2_flow_extract_add_raw(struct dpaa2_key_extract *key_extract,
+ int size)
+{
+   struct dpkg_profile_cfg *dpkg = &key_extract->dpkg;
+   struct dpaa2_key_info *key_info = &key_extract->key_info;
+   int last_extract_size, index;
+
+   if (dpkg->num_extracts != 0 && dpkg->extracts[0].type !=
+   DPKG_EXTRACT_FROM_DATA) {
+   DPAA2_PMD_WARN("RAW extract cannot be combined with others");
+   return -1;
+   }
+
+   last_extract_size = (size % DPAA2_FLOW_MAX_KEY_SIZE);
+   dpkg->num_extracts = (size / DPAA2_FLOW_MAX_KEY_SIZE);
+   if (last_extract_size)
+   dpkg->num_extracts++;
+   else
+   last_extract_size = DPAA2_FLOW_MAX_KEY_SIZE;
+
+   for (index = 0; index < dpkg->num_extracts; index++) {
+   dpkg->extracts[index].type = DPKG_EXTRACT_FROM_DATA;
+   if (index == dpkg->num_extracts - 1)
+   dpkg->extracts[index].extract.from_data.size =
+   last_extract_size;
+   else
+   dpkg->extracts[index].extract.from_data.size =
+   DPAA2_FLOW_MAX_KEY_SIZE;
+   dpkg->extracts[index].extract.from_data.offset =
+   DPAA2_FLOW_MAX_KEY_SIZE * index;
+   }
+
+   key_info->key_total_size = size;
+   return 0;
+}
+
 /* Protocol discrimination.
  * Discriminate IPv4/IPv6/vLan by Eth type.
  * Discriminate UDP/TCP/ICMP by next proto of IP.
@@ -674,6 +710,18 @@ dpaa2_flow_rule_data_set(
return 0;
 }
 
+static inline int
+dpaa2_flow_rule_data_set_raw(struct dpni_rule_cfg *rule,
+const void *key, const void *mask, int size)
+{
+   int offset = 0;
+
+   memcpy((void *)(size_t)(rule->key_iova + offset), key, size);
+   memcpy((void *)(size_t)(rule->mask_iova + offset), mask, size);
+
+   return 0;
+}
+
 static inline int
 _dpaa2_flow_rule_move_ipaddr_tail(
struct dpaa2_key_extract *key_extract,
@@ -2814,6 +2862,83 @@ dpaa2_configure_flow_gre(struct rte_flow *flow,
return 0;
 }
 
+static int
+dpaa2_configure_flow_raw(struct rte_flow *flow,
+struct rte_eth_dev *dev,
+const struct rte_flow_attr *attr,
+const struct rte_flow_item *pattern,
+const struct rte_flow_action actions[] __rte_unused,
+struct rte_flow_error *error __rte_unused,
+int *device_configured)
+{
+   struct dpaa2_dev_priv *priv = dev->data->dev_private;
+   const struct rte_flow_item_raw *spec = pattern->spec;
+   const struct rte_flow_item_raw *mask = pattern->mask;
+   int prev_key_size =
+   priv->extract.qos_key_extract.key_info.key_total_size;
+   int local_cfg = 0, ret;
+   uint32_t group;
+
+   /* Need both spec and mask */
+   if (!spec || !mask) {
+   DPAA2_PMD_ERR("spec or mask not present.");
+   return -EINVAL;
+   }
+   /* Only supports non-relative with offset 0 */
+   if (spec->relative || spec->offset != 0 ||
+   spec->search || spec->limit) {
+   DPAA2_PMD_ERR("relative and non zero offset not supported.");
+   return -EINVAL;
+   }
+   /* Spec len and mask len should be same */
+   if (spec->length != mask->length) {
+   DPAA2_PMD_ERR("Spec len and mask len mismatch.");
+   return -EINVAL;
+   }
+
+   /* Get traffic class index and flow id to be configured */
+   group = attr->group;
+   flow->tc_id = group;
+   flow->tc_index = attr->priority;
+
+  

[dpdk-dev] [PATCH v2 25/29] net/dpaa2: add flow data sanity check

[Hemant Agrawal]

From: Jun Yang 

Check flow attributions and actions before creating flow.
Otherwise, the QoS table and FS table need to re-build
if checking fails.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 84 ++
 1 file changed, 65 insertions(+), 19 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 76f68b903..3601829c9 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -3124,6 +3124,67 @@ dpaa2_flow_verify_attr(
return 0;
 }
 
+static inline int
+dpaa2_flow_verify_action(
+   struct dpaa2_dev_priv *priv,
+   const struct rte_flow_attr *attr,
+   const struct rte_flow_action actions[])
+{
+   int end_of_list = 0, i, j = 0;
+   const struct rte_flow_action_queue *dest_queue;
+   const struct rte_flow_action_rss *rss_conf;
+   struct dpaa2_queue *rxq;
+
+   while (!end_of_list) {
+   switch (actions[j].type) {
+   case RTE_FLOW_ACTION_TYPE_QUEUE:
+   dest_queue = (const struct rte_flow_action_queue *)
+   (actions[j].conf);
+   rxq = priv->rx_vq[dest_queue->index];
+   if (attr->group != rxq->tc_index) {
+   DPAA2_PMD_ERR(
+   "RXQ[%d] does not belong to the group 
%d",
+   dest_queue->index, attr->group);
+
+   return -1;
+   }
+   break;
+   case RTE_FLOW_ACTION_TYPE_RSS:
+   rss_conf = (const struct rte_flow_action_rss *)
+   (actions[j].conf);
+   if (rss_conf->queue_num > priv->dist_queues) {
+   DPAA2_PMD_ERR(
+   "RSS number exceeds the distrbution 
size");
+   return -ENOTSUP;
+   }
+   for (i = 0; i < (int)rss_conf->queue_num; i++) {
+   if (rss_conf->queue[i] >= priv->nb_rx_queues) {
+   DPAA2_PMD_ERR(
+   "RSS queue index exceeds the 
number of RXQs");
+   return -ENOTSUP;
+   }
+   rxq = priv->rx_vq[rss_conf->queue[i]];
+   if (rxq->tc_index != attr->group) {
+   DPAA2_PMD_ERR(
+   "Queue/Group combination are 
not supported\n");
+   return -ENOTSUP;
+   }
+   }
+
+   break;
+   case RTE_FLOW_ACTION_TYPE_END:
+   end_of_list = 1;
+   break;
+   default:
+   DPAA2_PMD_ERR("Invalid action type");
+   return -ENOTSUP;
+   }
+   j++;
+   }
+
+   return 0;
+}
+
 static int
 dpaa2_generic_flow_set(struct rte_flow *flow,
   struct rte_eth_dev *dev,
@@ -3150,6 +3211,10 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
if (ret)
return ret;
 
+   ret = dpaa2_flow_verify_action(priv, attr, actions);
+   if (ret)
+   return ret;
+
/* Parse pattern list to get the matching parameters */
while (!end_of_list) {
switch (pattern[i].type) {
@@ -3405,25 +3470,6 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
break;
case RTE_FLOW_ACTION_TYPE_RSS:
rss_conf = (const struct rte_flow_action_rss 
*)(actions[j].conf);
-   if (rss_conf->queue_num > priv->dist_queues) {
-   DPAA2_PMD_ERR(
-   "RSS number exceeds the distrbution 
size");
-   return -ENOTSUP;
-   }
-
-   for (i = 0; i < (int)rss_conf->queue_num; i++) {
-   if (rss_conf->queue[i] >= priv->nb_rx_queues) {
-   DPAA2_PMD_ERR(
-   "RSS RXQ number exceeds the 
total number");
-   return -ENOTSUP;
-   }
-   rxq = priv->rx_vq[rss_conf->queue[i]];
-   if (rxq->tc_index != attr->group) {
-   DPAA2_PMD_ERR(
-   "RSS RXQ distributed is not in 
current group");
-   return -ENOTSUP;
-   }
-  

[dpdk-dev] [PATCH v2 26/29] net/dpaa2: modify flow API QoS setup to follow FS setup

[Hemant Agrawal]

From: Jun Yang 

In HW/MC logical, QoS setup should follow FS setup.
In addition, Skip QoS setup if MAX TC number of DPNI is set 1.

Signed-off-by: Jun Yang 
---
 drivers/net/dpaa2/dpaa2_flow.c | 151 ++---
 1 file changed, 84 insertions(+), 67 deletions(-)

diff --git a/drivers/net/dpaa2/dpaa2_flow.c b/drivers/net/dpaa2/dpaa2_flow.c
index 3601829c9..9239fa459 100644
--- a/drivers/net/dpaa2/dpaa2_flow.c
+++ b/drivers/net/dpaa2/dpaa2_flow.c
@@ -2872,11 +2872,13 @@ dpaa2_flow_entry_update(
 
dpaa2_flow_qos_entry_log("Before update", curr, qos_index);
 
-   ret = dpni_remove_qos_entry(dpni, CMD_PRI_LOW,
-   priv->token, &curr->qos_rule);
-   if (ret) {
-   DPAA2_PMD_ERR("Qos entry remove failed.");
-   return -1;
+   if (priv->num_rx_tc > 1) {
+   ret = dpni_remove_qos_entry(dpni, CMD_PRI_LOW,
+   priv->token, &curr->qos_rule);
+   if (ret) {
+   DPAA2_PMD_ERR("Qos entry remove failed.");
+   return -1;
+   }
}
 
extend = -1;
@@ -2977,13 +2979,15 @@ dpaa2_flow_entry_update(
 
dpaa2_flow_qos_entry_log("Start update", curr, qos_index);
 
-   ret = dpni_add_qos_entry(dpni, CMD_PRI_LOW,
-   priv->token, &curr->qos_rule,
-   curr->tc_id, qos_index,
-   0, 0);
-   if (ret) {
-   DPAA2_PMD_ERR("Qos entry update failed.");
-   return -1;
+   if (priv->num_rx_tc > 1) {
+   ret = dpni_add_qos_entry(dpni, CMD_PRI_LOW,
+   priv->token, &curr->qos_rule,
+   curr->tc_id, qos_index,
+   0, 0);
+   if (ret) {
+   DPAA2_PMD_ERR("Qos entry update failed.");
+   return -1;
+   }
}
 
if (curr->action != RTE_FLOW_ACTION_TYPE_QUEUE) {
@@ -3313,31 +3317,8 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
flow->action = RTE_FLOW_ACTION_TYPE_QUEUE;
memset(&action, 0, sizeof(struct dpni_fs_action_cfg));
action.flow_id = rxq->flow_id;
-   if (is_keycfg_configured & DPAA2_QOS_TABLE_RECONFIGURE) 
{
-   dpaa2_flow_qos_table_extracts_log(priv);
-   if (dpkg_prepare_key_cfg(
-   &priv->extract.qos_key_extract.dpkg,
-   (uint8_t 
*)(size_t)priv->extract.qos_extract_param)
-   < 0) {
-   DPAA2_PMD_ERR(
-   "Unable to prepare extract parameters");
-   return -1;
-   }
 
-   memset(&qos_cfg, 0, sizeof(struct 
dpni_qos_tbl_cfg));
-   qos_cfg.discard_on_miss = true;
-   qos_cfg.keep_entries = true;
-   qos_cfg.key_cfg_iova =
-   (size_t)priv->extract.qos_extract_param;
-   ret = dpni_set_qos_table(dpni, CMD_PRI_LOW,
-   priv->token, &qos_cfg);
-   if (ret < 0) {
-   DPAA2_PMD_ERR(
-   "Distribution cannot be configured.(%d)"
-   , ret);
-   return -1;
-   }
-   }
+   /* Configure FS table first*/
if (is_keycfg_configured & DPAA2_FS_TABLE_RECONFIGURE) {
dpaa2_flow_fs_table_extracts_log(priv, 
flow->tc_id);
if (dpkg_prepare_key_cfg(
@@ -3366,17 +3347,39 @@ dpaa2_generic_flow_set(struct rte_flow *flow,
return -1;
}
}
-   /* Configure QoS table first */
 
-   qos_index = flow->tc_id * priv->fs_entries +
-   flow->tc_index;
+   /* Configure QoS table then.*/
+   if (is_keycfg_configured & DPAA2_QOS_TABLE_RECONFIGURE) 
{
+   dpaa2_flow_qos_table_extracts_log(priv);
+   if (dpkg_prepare_key_cfg(
+ 

Re: [dpdk-dev] [PATCH v2 5/5] app/testpmd: enable empty polls in burst stats

[Ferruh Yigit]

On 7/6/2020 8:11 PM, Honnappa Nagarahalli wrote:
> Hi Ferruh,
>   Thanks for the review comments
> 
> 
> 
>> Subject: Re: [PATCH v2 5/5] app/testpmd: enable empty polls in burst stats
>>
>> On 6/26/2020 11:09 PM, Honnappa Nagarahalli wrote:
>>> The number of empty polls provides information about available CPU
>>> head room in the presence of continuous polling.
>>
>> +1 to have it, it can be useful.
>>
>>>
>>> Signed-off-by: Honnappa Nagarahalli 
>>> Reviewed-by: Phil Yang 
>>> Reviewed-by: Ruifeng Wang 
>>> Tested-by: Phil Yang 
>>> Tested-by: Ali Alnubani 
>>
>> <...>
>>
>>> /*
>>>  * First compute the total number of packet bursts and the
>>>  * two highest numbers of bursts of the same number of packets.
>>>  */
>>> total_burst = 0;
>>> -   burst_stats[0] = burst_stats[1] = burst_stats[2] = 0;
>>> -   pktnb_stats[0] = pktnb_stats[1] = pktnb_stats[2] = 0;
>>> +   memset(&burst_stats, 0x0, sizeof(burst_stats));
>>> +   memset(&pktnb_stats, 0x0, sizeof(pktnb_stats));
>>> +
>>> for (nb_pkt = 0; nb_pkt < MAX_PKT_BURST; nb_pkt++) {
>>> nb_burst = pbs->pkt_burst_spread[nb_pkt];
>>> -   if (nb_burst == 0)
>>> -   continue;
>>> total_burst += nb_burst;
>>> -   if (nb_burst > burst_stats[0]) {
>>> -   burst_stats[1] = burst_stats[0];
>>> -   pktnb_stats[1] = pktnb_stats[0];
>>> +
>>> +   /* Show empty poll stats always */
>>> +   if (nb_pkt == 0) {
>>> burst_stats[0] = nb_burst;
>>> pktnb_stats[0] = nb_pkt;
>>
>> just a minor issue but this assignment is not needed.
> The empty poll stats are being shown always (even if there were no empty 
> polls). The check 'nb_pkts == 0' indicates that the burst stats are for empty 
> polls. Hence this assignment is required. But, this can be moved out of the 
> loop to provide clarity. I will do that.

It is not required because of 'memset' above, we know 'nb_pkt == 0' and we know
'pktnb_stats[0] == 0', so "pktnb_stats[0] = nb_pkt;" is redundant.

> 
>>
>>> -   } else if (nb_burst > burst_stats[1]) {
>>> +   continue;
>>> +   }
>>> +
>>> +   if (nb_burst == 0)
>>> +   continue;
>>
>> again another minor issue, can have this check above 'nb_pkt == 0', to 
>> prevent
>> unnecssary "burst_stats[0] = nb_burst;" assignment if "nb_burst == 0".
> The empty polls are always shown, even if there were 0% empty polls.

I got that, again because of memset, "burst_stats[0] == 0", you can skip
"burst_stats[0] = nb_burst;" in case "nb_burst == 0", that is why you can move
it up.

Anyway, both are trivial issues ...

> 
>>
>> <...>
>>
>> Reviewed-by: Ferruh Yigit 

Re: [dpdk-dev] [dpdk-dev v5 0/4] enable new hash flow for pf

[Zhang, Qi Z]

> -Original Message-
> From: Guo, Jia 
> Sent: Tuesday, July 7, 2020 1:26 PM
> To: Zhang, Qi Z ; Yang, Qiming
> 
> Cc: dev@dpdk.org; Wu, Jingjing ; Guo, Junfeng
> ; Xing, Beilei ; Su, Simei
> ; Guo, Jia 
> Subject: [dpdk-dev v5 0/4] enable new hash flow for pf
> 
> PF could add or delete a RSS rule base on the PF's hash capability.
> Some new rss input set will be supported, the protocols as below:
> eth/vlan/l2tpv3/esp/ah/pfcp/gtpu downlink/gtpu uplink.
> 
> v5->v4:
> fix checkpatch issue
> refine for eth/vlan hash
> 
> v4->v3:
> split patch set
> 
> Jeff Guo (4):
>   net/ice: refactor for pf hash flow
>   net/ice: support hash for new GTPU protocols
>   net/ice: enable new input set for rss hash
>   app/testpmd: add GTPU to RSS hash commands
> 
>  app/test-pmd/cmdline.c |   8 +-
>  app/test-pmd/config.c  |   4 +-
>  drivers/net/ice/ice_hash.c | 773 ++---
>  3 files changed, 562 insertions(+), 223 deletions(-)
> 
> --
> 2.20.1

Acked-by: Qi Zhang 

Applied to dpdk-next-net-intel.

Thanks
Qi

Re: [dpdk-dev] [PATCH v3 4/8] fib: introduce AVX512 lookup

[Bruce Richardson]

On Tue, May 19, 2020 at 01:12:59PM +0100, Vladimir Medvedkin wrote:
> Add new lookup implementation for DIR24_8 algorithm using
> AVX512 instruction set
> 
> Signed-off-by: Vladimir Medvedkin 
> ---
>  lib/librte_fib/Makefile |  14 
>  lib/librte_fib/dir24_8.c|  24 ++
>  lib/librte_fib/dir24_8_avx512.c | 165 
> 
>  lib/librte_fib/dir24_8_avx512.h |  24 ++
>  lib/librte_fib/meson.build  |  11 +++
>  lib/librte_fib/rte_fib.h|   3 +-
>  6 files changed, 240 insertions(+), 1 deletion(-)
>  create mode 100644 lib/librte_fib/dir24_8_avx512.c
>  create mode 100644 lib/librte_fib/dir24_8_avx512.h
>
 
> diff --git a/lib/librte_fib/meson.build b/lib/librte_fib/meson.build
> index 771828f..0963f3c 100644
> --- a/lib/librte_fib/meson.build
> +++ b/lib/librte_fib/meson.build
> @@ -5,3 +5,14 @@
>  sources = files('rte_fib.c', 'rte_fib6.c', 'dir24_8.c', 'trie.c')
>  headers = files('rte_fib.h', 'rte_fib6.h')
>  deps += ['rib']
> +
> +if dpdk_conf.has('RTE_ARCH_X86') and cc.has_argument('-mavx512f')
> + if cc.has_argument('-mavx512dq')
> + dir24_8_avx512_tmp = static_library('dir24_8_avx512_tmp',
> + 'dir24_8_avx512.c',
> + dependencies: static_rte_eal,
> + c_args: cflags + ['-mavx512f'] + ['-mavx512dq'])
> + objs += dir24_8_avx512_tmp.extract_objects('dir24_8_avx512.c')
> + cflags += '-DCC_DIR24_8_AVX512_SUPPORT'
> + endif
> +endif

This block looks wrong to me, especially comparing it with the equivalent
block in drivers/net/i40e. Firstly, the two if conditions are unnecessary
and can be merged. However, secondly, I think you should restructure it so
that you first check for AVX-512 already being enabled in the build, and
only if it is not do you need to see about checking compiler support and
using the static lib workaround to get just the one file compiled with
AVX-512. As Thomas suggested, a comment explaining this would also help -
again copying what is in the i40e/meson.build file would probably be a good
start.

/Bruce

Re: [dpdk-dev] [PATCH v2 1/3] ethdev: add new RSS offload types

[Ferruh Yigit]

On 7/3/2020 2:50 PM, Zhang, Qi Z wrote:
> 
> 
>> -Original Message-
>> From: Su, Simei 
>> Sent: Thursday, July 2, 2020 1:12 PM
>> To: Zhang, Qi Z ; Xing, Beilei 
>> Cc: dev@dpdk.org; Guo, Jia ; Guo, Junfeng
>> ; Su, Simei 
>> Subject: [PATCH v2 1/3] ethdev: add new RSS offload types
>>
>> This patch defines new RSS offload types for PPPoE. Typically, session id
>> would be the RSS input set for a PPPoE packet, but as a hint, each driver may
>> have different default behaviors.
>>
>> Signed-off-by: Simei Su 
>> ---
>>  lib/librte_ethdev/rte_ethdev.h | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
>> index a49242b..631b146 100644
>> --- a/lib/librte_ethdev/rte_ethdev.h
>> +++ b/lib/librte_ethdev/rte_ethdev.h
>> @@ -521,7 +521,7 @@ struct rte_eth_rss_conf {
>>  #define ETH_RSS_AH (1ULL << 28)
>>  #define ETH_RSS_L2TPV3 (1ULL << 29)
>>  #define ETH_RSS_PFCP   (1ULL << 30)
>> -
>> +#define ETH_RSS_PPPOE  (1ULL << 31)
>>
>>  /*
>>   * We use the following macros to combine with above ETH_RSS_* for
>> --
>> 1.8.3.1
> 
> Reviewed-by: Qi Zhang 
> 

Reviewed-by: Ferruh Yigit 

[dpdk-dev] [PATCH v6 0/4] generic rte atomic APIs deprecate proposal

[Phil Yang]

DPDK provides generic rte_atomic APIs to do several atomic operations.
These APIs are using the deprecated __sync built-ins and enforce full
memory barriers on aarch64. However, full barriers are not necessary
in many use cases. In order to address such use cases, C language offers
C11 atomic APIs. The C11 atomic APIs provide finer memory barrier control
by making use of the memory ordering parameter provided by the user.
Various patches submitted in the past [2] and the patches in this series
indicate significant performance gains on multiple aarch64 CPUs and no
performance loss on x86.

But the existing rte_atomic API implementations cannot be changed as the
APIs do not take the memory ordering parameter. The only choice available
is replacing the usage of the rte_atomic APIs with C11 atomic APIs. In
order to make this change, the following steps are proposed:

[1] deprecate rte_atomic APIs so that future patches do not use rte_atomic
APIs (a script is added to flag the usages).
[2] refactor the code that uses rte_atomic APIs to use c11 atomic APIs.

This patchset contains:
1) changes to programmer guide describing writing efficient code for aarch64.
2) the checkpatch script changes to flag rte_atomicNN_xxx API usage in patches.
3) wraps up __atomic_thread_fence with explicit memory ordering parameter.

v6:
Add check for rte_smp barriers APIs in the new code.

v5:
1. Wraps up __atomic_thread_fence to support optimized code for
__ATOMIC_SEQ_CST memory order.
2. Flag __atomic_thread_fence with __ATOMIC_SEQ_CST in new patches.
3. Fix email address typo in patch 2/4.

v4:
1. add reader-writer concurrency case describing.
2. claim maintainership of c11 atomics code for each platforms.
3. flag rte_atomicNN_xxx in new patches for modules that have been converted to
c11 style.
4. flag __sync_xxx built-ins in new patches.
5. wraps up compiler atomic built-ins
6. move the changes of libraries which make use of c11 atomic APIs out of this
patchset.

v3:
add libatomic dependency for 32-bit clang

v2:
1. fix Clang '-Wincompatible-pointer-types' WARNING.
2. fix typos.

Phil Yang (4):
  doc: add generic atomic deprecation section
  maintainers: claim maintainers of C11 atomics
  devtools: prevent use of rte atomic APIs in future patches
  eal/atomic: add wrapper for C11 atomic thread fence

 MAINTAINERS  |   4 +
 devtools/checkpatches.sh |  40 +++
 doc/guides/prog_guide/writing_efficient_code.rst | 139 ++-
 lib/librte_eal/arm/include/rte_atomic_32.h   |   6 +
 lib/librte_eal/arm/include/rte_atomic_64.h   |   6 +
 lib/librte_eal/include/generic/rte_atomic.h  |   6 +
 lib/librte_eal/ppc/include/rte_atomic.h  |   6 +
 lib/librte_eal/x86/include/rte_atomic.h  |  17 +++
 8 files changed, 223 insertions(+), 1 deletion(-)

-- 
2.7.4

[dpdk-dev] [PATCH v6 3/4] devtools: prevent use of rte atomic APIs in future patches

[Phil Yang]

In order to deprecate the rte_atomic and rte_smp barrier APIs, prevent
the patches from using these APIs in the converted modules and compilers
__sync built-ins in all modules.

The converted modules:
lib/librte_distributor
lib/librte_hash
lib/librte_kni
lib/librte_lpm
lib/librte_rcu
lib/librte_ring
lib/librte_stack
lib/librte_vhost
lib/librte_timer
lib/librte_ipsec
drivers/event/octeontx
drivers/event/octeontx2
drivers/event/opdl
drivers/net/bnx2x
drivers/net/hinic
drivers/net/hns3
drivers/net/memif
drivers/net/thunderx
drivers/net/virtio
examples/l2fwd-event

On x86 the __atomic_thread_fence(__ATOMIC_SEQ_CST) is quite expensive
for SMP case. Flag the new code which use __atomic_thread_fence API.

Signed-off-by: Phil Yang 
Reviewed-by: Ruifeng Wang 
---
 devtools/checkpatches.sh | 40 
 1 file changed, 40 insertions(+)

diff --git a/devtools/checkpatches.sh b/devtools/checkpatches.sh
index 58021aa..68dc598 100755
--- a/devtools/checkpatches.sh
+++ b/devtools/checkpatches.sh
@@ -51,6 +51,13 @@ print_usage () {
 
 check_forbidden_additions() { # 
res=0
+   c11_atomics_dir="lib/librte_distributor lib/librte_hash lib/librte_kni
+lib/librte_lpm lib/librte_rcu lib/librte_ring
+lib/librte_stack lib/librte_vhost
+drivers/event/octeontx drivers/event/octeontx2
+drivers/event/opdl drivers/net/bnx2x drivers/net/hinic
+drivers/net/hns3 drivers/net/memif drivers/net/thunderx
+drivers/net/virtio examples/l2fwd-event"
 
# refrain from new additions of rte_panic() and rte_exit()
# multiple folders and expressions are separated by spaces
@@ -74,6 +81,39 @@ check_forbidden_additions() { # 
-v 
EXPRESSIONS='for[[:space:]]*\\((char|u?int|unsigned|s?size_t)' \
-v RET_ON_FAIL=1 \
-v MESSAGE='Declaring a variable inside for()' \
+
+   # refrain from new additions of 16/32/64 bits rte_atomic_xxx()
+   # multiple folders and expressions are separated by spaces
+   awk -v FOLDERS="$c11_atomics_dir" \
+   -v EXPRESSIONS="rte_atomic[0-9][0-9]_.*\\\(" \
+   -v RET_ON_FAIL=1 \
+   -v MESSAGE='Use of rte_atomicNN_xxx APIs not allowed, use 
__atomic_xxx built-ins' \
+   -f $(dirname $(readlink -f $0))/check-forbidden-tokens.awk \
+   "$1" || res=1
+
+   # refrain from new additions of rte_smp_XXmb()
+   # multiple folders and expressions are separated by spaces
+   awk -v FOLDERS="$c11_atomics_dir" \
+   -v EXPRESSIONS="rte_smp_(r|w)?mb\\\(" \
+   -v RET_ON_FAIL=1 \
+   -v MESSAGE='Use of rte_smp_r/wmb not allowed, use __atomic_xxx 
built-ins' \
+   -f $(dirname $(readlink -f $0))/check-forbidden-tokens.awk \
+   "$1" || res=1
+
+   # refrain from using compiler __sync built-ins
+   awk -v FOLDERS="lib drivers app examples" \
+   -v EXPRESSIONS="__sync_.*\\\(" \
+   -v RET_ON_FAIL=1 \
+   -v MESSAGE='Use of __sync_xxx built-ins not allowed, use 
__atomic_xxx built-ins' \
+   -f $(dirname $(readlink -f $0))/check-forbidden-tokens.awk \
+   "$1" || res=1
+
+   # refrain from using compiler __atomic_thread_fence()
+   # It should be avoided on x86 for SMP case.
+   awk -v FOLDERS="lib drivers app examples" \
+   -v EXPRESSIONS="__atomic_thread_fence\\\(" \
+   -v RET_ON_FAIL=1 \
+   -v MESSAGE='Use of __atomic_thread_fence is not allowed, use 
rte_atomic_thread_fence' \
-f $(dirname $(readlink -f $0))/check-forbidden-tokens.awk \
"$1" || res=1
 
-- 
2.7.4

[dpdk-dev] [PATCH v6 1/4] doc: add generic atomic deprecation section

[Phil Yang]

Add deprecating the generic rte_atomic_xx APIs to c11 atomic built-ins
guide and examples.

Signed-off-by: Phil Yang 
Signed-off-by: Honnappa Nagarahalli 
---
 doc/guides/prog_guide/writing_efficient_code.rst | 139 ++-
 1 file changed, 138 insertions(+), 1 deletion(-)

diff --git a/doc/guides/prog_guide/writing_efficient_code.rst 
b/doc/guides/prog_guide/writing_efficient_code.rst
index 849f63e..3bd2601 100644
--- a/doc/guides/prog_guide/writing_efficient_code.rst
+++ b/doc/guides/prog_guide/writing_efficient_code.rst
@@ -167,7 +167,13 @@ but with the added cost of lower throughput.
 Locks and Atomic Operations
 ---
 
-Atomic operations imply a lock prefix before the instruction,
+This section describes some key considerations when using locks and atomic
+operations in the DPDK environment.
+
+Locks
+~
+
+On x86, atomic operations imply a lock prefix before the instruction,
 causing the processor's LOCK# signal to be asserted during execution of the 
following instruction.
 This has a big impact on performance in a multicore environment.
 
@@ -176,6 +182,137 @@ It can often be replaced by other solutions like 
per-lcore variables.
 Also, some locking techniques are more efficient than others.
 For instance, the Read-Copy-Update (RCU) algorithm can frequently replace 
simple rwlocks.
 
+Atomic Operations: Use C11 Atomic Built-ins
+~~~
+
+DPDK `generic rte_atomic 
`_
 operations are
+implemented by `__sync built-ins 
`_.
+These __sync built-ins result in full barriers on aarch64, which are 
unnecessary
+in many use cases. They can be replaced by `__atomic built-ins 
`_ that
+conform to the C11 memory model and provide finer memory order control.
+
+So replacing the rte_atomic operations with __atomic built-ins might improve
+performance for aarch64 machines. `More details 
`_.
+
+Some typical optimization cases are listed below:
+
+Atomicity
+^
+
+Some use cases require atomicity alone, the ordering of the memory operations
+does not matter. For example the packets statistics in the `vhost 
`_ example 
application.
+
+It just updates the number of transmitted packets, no subsequent logic depends
+on these counters. So the RELAXED memory ordering is sufficient:
+
+.. code-block:: c
+
+static __rte_always_inline void
+virtio_xmit(struct vhost_dev *dst_vdev, struct vhost_dev *src_vdev,
+struct rte_mbuf *m)
+{
+...
+...
+if (enable_stats) {
+__atomic_add_fetch(&dst_vdev->stats.rx_total_atomic, 1, 
__ATOMIC_RELAXED);
+__atomic_add_fetch(&dst_vdev->stats.rx_atomic, ret, 
__ATOMIC_RELAXED);
+...
+}
+}
+
+One-way Barrier
+^^^
+
+Some use cases allow for memory reordering in one way while requiring memory
+ordering in the other direction.
+
+For example, the memory operations before the `lock 
`_
 can move to the
+critical section, but the memory operations in the critical section cannot move
+above the lock. In this case, the full memory barrier in the CAS operation can
+be replaced to ACQUIRE. On the other hand, the memory operations after the
+`unlock 
`_
 can move to the critical section, but the memory operations in the
+critical section cannot move below the unlock. So the full barrier in the STORE
+operation can be replaced with RELEASE.
+
+Reader-Writer Concurrency
+^
+Lock-free reader-writer concurrency is one of the common use cases in DPDK.
+
+The payload or the data that the writer wants to communicate to the reader,
+can be written with RELAXED memory order. However, the guard variable should
+be written with RELEASE memory order. This ensures that the store to guard
+variable is observable only after the store to payload is observable.
+Refer to `rte_hash insert 
`_
 for an example.
+
+.. code-block:: c
+
+static inline int32_t
+rte_hash_cuckoo_insert_mw(const struct rte_hash *h,
+...
+int32_t *ret_val)
+{
+...
+...
+
+/* Insert new entry if there is room in the primary
+ * bucket.
+ */
+for (i = 0; i < RTE_HASH_BUCKET_ENTRIES; i++) {
+/* Check if slot is available */
+if (likely(prim_bkt->key_idx[i] == EMPTY_SLOT)) {
+ 

[dpdk-dev] [PATCH v6 4/4] eal/atomic: add wrapper for C11 atomic thread fence

[Phil Yang]

Provide a wrapper for __atomic_thread_fence built-in to support
optimized code for __ATOMIC_SEQ_CST memory order for x86 platforms.

Suggested-by: Honnappa Nagarahalli 
Signed-off-by: Phil Yang 
Reviewed-by: Ola Liljedahl 
Acked-by: Konstantin Ananyev 
---
 lib/librte_eal/arm/include/rte_atomic_32.h  |  6 ++
 lib/librte_eal/arm/include/rte_atomic_64.h  |  6 ++
 lib/librte_eal/include/generic/rte_atomic.h |  6 ++
 lib/librte_eal/ppc/include/rte_atomic.h |  6 ++
 lib/librte_eal/x86/include/rte_atomic.h | 17 +
 5 files changed, 41 insertions(+)

diff --git a/lib/librte_eal/arm/include/rte_atomic_32.h 
b/lib/librte_eal/arm/include/rte_atomic_32.h
index 7dc0d06..dbe7cc6 100644
--- a/lib/librte_eal/arm/include/rte_atomic_32.h
+++ b/lib/librte_eal/arm/include/rte_atomic_32.h
@@ -37,6 +37,12 @@ extern "C" {
 
 #define rte_cio_rmb() rte_rmb()
 
+static __rte_always_inline void
+rte_atomic_thread_fence(int mo)
+{
+   __atomic_thread_fence(mo);
+}
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/lib/librte_eal/arm/include/rte_atomic_64.h 
b/lib/librte_eal/arm/include/rte_atomic_64.h
index 7b7099c..22ff8ec 100644
--- a/lib/librte_eal/arm/include/rte_atomic_64.h
+++ b/lib/librte_eal/arm/include/rte_atomic_64.h
@@ -41,6 +41,12 @@ extern "C" {
 
 #define rte_cio_rmb() asm volatile("dmb oshld" : : : "memory")
 
+static __rte_always_inline void
+rte_atomic_thread_fence(int mo)
+{
+   __atomic_thread_fence(mo);
+}
+
 /* 128 bit atomic operations 
-*/
 
 #if defined(__ARM_FEATURE_ATOMICS) || defined(RTE_ARM_FEATURE_ATOMICS)
diff --git a/lib/librte_eal/include/generic/rte_atomic.h 
b/lib/librte_eal/include/generic/rte_atomic.h
index e6ab15a..5b941db 100644
--- a/lib/librte_eal/include/generic/rte_atomic.h
+++ b/lib/librte_eal/include/generic/rte_atomic.h
@@ -158,6 +158,12 @@ static inline void rte_cio_rmb(void);
asm volatile ("" : : : "memory");   \
 } while(0)
 
+/**
+ * Synchronization fence between threads based on the specified
+ * memory order.
+ */
+static inline void rte_atomic_thread_fence(int mo);
+
 /*- 16 bit atomic operations 
-*/
 
 /**
diff --git a/lib/librte_eal/ppc/include/rte_atomic.h 
b/lib/librte_eal/ppc/include/rte_atomic.h
index 7e3e131..91c5f30 100644
--- a/lib/librte_eal/ppc/include/rte_atomic.h
+++ b/lib/librte_eal/ppc/include/rte_atomic.h
@@ -40,6 +40,12 @@ extern "C" {
 
 #define rte_cio_rmb() rte_rmb()
 
+static __rte_always_inline void
+rte_atomic_thread_fence(int mo)
+{
+   __atomic_thread_fence(mo);
+}
+
 /*- 16 bit atomic operations 
-*/
 /* To be compatible with Power7, use GCC built-in functions for 16 bit
  * operations */
diff --git a/lib/librte_eal/x86/include/rte_atomic.h 
b/lib/librte_eal/x86/include/rte_atomic.h
index b9dcd30..bd256e7 100644
--- a/lib/librte_eal/x86/include/rte_atomic.h
+++ b/lib/librte_eal/x86/include/rte_atomic.h
@@ -83,6 +83,23 @@ rte_smp_mb(void)
 
 #define rte_cio_rmb() rte_compiler_barrier()
 
+/**
+ * Synchronization fence between threads based on the specified
+ * memory order.
+ *
+ * On x86 the __atomic_thread_fence(__ATOMIC_SEQ_CST) generates
+ * full 'mfence' which is quite expensive. The optimized
+ * implementation of rte_smp_mb is used instead.
+ */
+static __rte_always_inline void
+rte_atomic_thread_fence(int mo)
+{
+   if (mo == __ATOMIC_SEQ_CST)
+   rte_smp_mb();
+   else
+   __atomic_thread_fence(mo);
+}
+
 /*- 16 bit atomic operations 
-*/
 
 #ifndef RTE_FORCE_INTRINSICS
-- 
2.7.4

[dpdk-dev] [PATCH v6 2/4] maintainers: claim maintainers of C11 atomics

[Phil Yang]

Add the maintainership of C11 atomics code. Different maintainers focus
on different platforms C11 atomics.

Signed-off-by: Phil Yang 
Reviewed-by: Honnappa Nagarahalli 
---
 MAINTAINERS | 4 
 1 file changed, 4 insertions(+)

diff --git a/MAINTAINERS b/MAINTAINERS
index 27fcd6d..3682d2a 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -276,6 +276,10 @@ F: lib/librte_eal/include/rte_random.h
 F: lib/librte_eal/common/rte_random.c
 F: app/test/test_rand_perf.c
 
+C11 Atomics Code Maintainer
+M: Honnappa Nagarahalli 
+M: David Christensen 
+
 ARM v7
 M: Jan Viktorin 
 M: Ruifeng Wang 
-- 
2.7.4

[dpdk-dev] [PATCH 0/2] update base code batch 4

[Wei Zhao]

The patchset include 2 bug fixes for base code.

Wei Zhao (2):
  ice-shared: Fix tunnel type check for PPPoE
  ice-shared: Fix add special words for tunnel rule

 drivers/net/ice/base/ice_switch.c | 52 +--
 1 file changed, 14 insertions(+), 38 deletions(-)

-- 
2.19.1

[dpdk-dev] [PATCH 1/2] net/ice/base: fix tunnel type check for PPPoE

[Wei Zhao]

This patch fix tunnel type check for PPPoE type of recipe
when search for recipe which has been created before, and
also fix a bug of typo.

Fixes: 418d2563d10b ("net/ice/base: get tunnel type for recipe")

Signed-off-by: Wei Zhao 
---
 drivers/net/ice/base/ice_switch.c | 14 +-
 1 file changed, 5 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ice/base/ice_switch.c 
b/drivers/net/ice/base/ice_switch.c
index 0e220580a..caa080da2 100644
--- a/drivers/net/ice/base/ice_switch.c
+++ b/drivers/net/ice/base/ice_switch.c
@@ -1086,15 +1086,11 @@ static enum ice_sw_tunnel_type 
ice_get_tun_type_for_recipe(u8 rid)
tun_type = ICE_SW_TUN_PPPOE;
else if (!non_tun_valid && gtp_valid)
tun_type = ICE_SW_TUN_GTP;
-   else if ((non_tun_valid && vxlan_valid) ||
-(non_tun_valid && gre_valid) ||
-(non_tun_valid && gtp_valid) ||
-(non_tun_valid && pppoe_valid))
+   else if (non_tun_valid &&
+(vxlan_valid || gre_valid || gtp_valid || pppoe_valid))
tun_type = ICE_SW_TUN_AND_NON_TUN;
-   else if ((non_tun_valid && !vxlan_valid) ||
-(non_tun_valid && !gre_valid) ||
-(non_tun_valid && !gtp_valid) ||
-(non_tun_valid && !pppoe_valid))
+   else if (non_tun_valid && !vxlan_valid && !gre_valid && !gtp_valid &&
+!pppoe_valid)
tun_type = ICE_NON_TUN;
 
if (profile_num > 1 && tun_type == ICE_SW_TUN_PPPOE) {
@@ -1108,7 +1104,7 @@ static enum ice_sw_tunnel_type 
ice_get_tun_type_for_recipe(u8 rid)
tun_type = ICE_SW_TUN_PPPOE_IPV6;
}
 
-   if (profile_num == 1 && (flag_valid || non_tun_valid)) {
+   if (profile_num == 1 && (flag_valid || non_tun_valid || pppoe_valid)) {
for (j = 0; j < ICE_MAX_NUM_PROFILES; j++) {
if (ice_is_bit_set(recipe_to_profile[rid], j)) {
switch (j) {
-- 
2.19.1

[dpdk-dev] [PATCH 2/2] net/ice/base: fix add special words for tunnel rule

[Wei Zhao]

if we download a NVGRE rule like "eth / ipv4 / nvgre / eth / ipv4 src
is 192.168.1.2 dst is 192.168.1.3 / end actions queue index 3 / end"
the special word will not be added because of rm->n_grp_count = 1,
so we have to change the ice_add_special_words() function before
ice_create_recipe_group(), then the special words will be add
into rm->rg_list, btw the flag match_tun_mask is useless now
and no need to add special words in function
ice_add_sw_recipe() duplicatly.

Fixes: eda40e22b858 ("net/ice/base: improve switch chained recipe")
Cc: sta...@dpdk.org

Signed-off-by: Wei Zhao 
---
 drivers/net/ice/base/ice_switch.c | 38 ---
 1 file changed, 9 insertions(+), 29 deletions(-)

diff --git a/drivers/net/ice/base/ice_switch.c 
b/drivers/net/ice/base/ice_switch.c
index caa080da2..296aa5484 100644
--- a/drivers/net/ice/base/ice_switch.c
+++ b/drivers/net/ice/base/ice_switch.c
@@ -5993,12 +5993,11 @@ ice_find_free_recp_res_idx(struct ice_hw *hw, const 
ice_bitmap_t *profiles,
  * ice_add_sw_recipe - function to call AQ calls to create switch recipe
  * @hw: pointer to hardware structure
  * @rm: recipe management list entry
- * @match_tun_mask: tunnel mask that needs to be programmed
  * @profiles: bitmap of profiles that will be associated.
  */
 static enum ice_status
 ice_add_sw_recipe(struct ice_hw *hw, struct ice_sw_recipe *rm,
- u16 match_tun_mask, ice_bitmap_t *profiles)
+ ice_bitmap_t *profiles)
 {
ice_declare_bitmap(result_idx_bm, ICE_MAX_FV_WORDS);
struct ice_aqc_recipe_data_elem *tmp;
@@ -6213,15 +6212,6 @@ ice_add_sw_recipe(struct ice_hw *hw, struct 
ice_sw_recipe *rm,
}
buf[recps].content.act_ctrl_fwd_priority = rm->priority;
 
-   /* To differentiate among different UDP tunnels, a meta data ID
-* flag is used.
-*/
-   if (match_tun_mask) {
-   buf[recps].content.lkup_indx[i] = ICE_TUN_FLAG_FV_IND;
-   buf[recps].content.mask[i] =
-   CPU_TO_LE16(match_tun_mask);
-   }
-
recps++;
rm->root_rid = (u8)rid;
}
@@ -6596,8 +6586,6 @@ ice_add_adv_recipe(struct ice_hw *hw, struct 
ice_adv_lkup_elem *lkups,
struct ice_sw_fv_list_entry *tmp;
enum ice_status status = ICE_SUCCESS;
struct ice_sw_recipe *rm;
-   u16 match_tun_mask = 0;
-   u16 mask;
u8 i;
 
if (!ice_is_prof_rule(rinfo->tun_type) && !lkups_cnt)
@@ -6648,6 +6636,13 @@ ice_add_adv_recipe(struct ice_hw *hw, struct 
ice_adv_lkup_elem *lkups,
if (status)
goto err_unroll;
 
+   /* Create any special protocol/offset pairs, such as looking at tunnel
+* bits by extracting metadata
+*/
+   status = ice_add_special_words(rinfo, lkup_exts);
+   if (status)
+   goto err_free_lkup_exts;
+
/* Group match words into recipes using preferred recipe grouping
 * criteria.
 */
@@ -6655,14 +6650,6 @@ ice_add_adv_recipe(struct ice_hw *hw, struct 
ice_adv_lkup_elem *lkups,
if (status)
goto err_unroll;
 
-   /* For certain tunnel types it is necessary to use a metadata ID flag to
-* differentiate different tunnel types. A separate recipe needs to be
-* used for the metadata.
-*/
-   if (ice_tun_type_match_word(rinfo->tun_type, &mask) &&
-   rm->n_grp_count > 1)
-   match_tun_mask = mask;
-
/* set the recipe priority if specified */
rm->priority = (u8)rinfo->priority;
 
@@ -6701,13 +6688,6 @@ ice_add_adv_recipe(struct ice_hw *hw, struct 
ice_adv_lkup_elem *lkups,
ice_set_bit((u16)fvit->profile_id, profiles);
}
 
-   /* Create any special protocol/offset pairs, such as looking at tunnel
-* bits by extracting metadata
-*/
-   status = ice_add_special_words(rinfo, lkup_exts);
-   if (status)
-   goto err_free_lkup_exts;
-
/* Look for a recipe which matches our requested fv / mask list */
*rid = ice_find_recp(hw, lkup_exts, rinfo->tun_type);
if (*rid < ICE_MAX_NUM_RECIPES)
@@ -6716,7 +6696,7 @@ ice_add_adv_recipe(struct ice_hw *hw, struct 
ice_adv_lkup_elem *lkups,
 
rm->tun_type = rinfo->tun_type;
/* Recipe we need does not exist, add a recipe */
-   status = ice_add_sw_recipe(hw, rm, match_tun_mask, profiles);
+   status = ice_add_sw_recipe(hw, rm, profiles);
if (status)
goto err_unroll;
 
-- 
2.19.1

[dpdk-dev] [PATCH v2] mbuf: use C11 atomics for refcnt operations

[Phil Yang]

Use C11 atomics with explicit ordering instead of rte_atomic ops which
enforce unnecessary barriers on aarch64.

Signed-off-by: Phil Yang 
Reviewed-by: Ruifeng Wang 
---
v2:
Fix ABI issue: revert the rte_mbuf_ext_shared_info struct refcnt field
to refcnt_atomic.

 lib/librte_mbuf/rte_mbuf.c  |  1 -
 lib/librte_mbuf/rte_mbuf.h  | 19 ++-
 lib/librte_mbuf/rte_mbuf_core.h | 11 +++
 3 files changed, 13 insertions(+), 18 deletions(-)

diff --git a/lib/librte_mbuf/rte_mbuf.c b/lib/librte_mbuf/rte_mbuf.c
index ae91ae2..8a456e5 100644
--- a/lib/librte_mbuf/rte_mbuf.c
+++ b/lib/librte_mbuf/rte_mbuf.c
@@ -22,7 +22,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 #include 
diff --git a/lib/librte_mbuf/rte_mbuf.h b/lib/librte_mbuf/rte_mbuf.h
index f8e492e..4a7a98c 100644
--- a/lib/librte_mbuf/rte_mbuf.h
+++ b/lib/librte_mbuf/rte_mbuf.h
@@ -37,7 +37,6 @@
 #include 
 #include 
 #include 
-#include 
 #include 
 #include 
 #include 
@@ -365,7 +364,7 @@ rte_pktmbuf_priv_flags(struct rte_mempool *mp)
 static inline uint16_t
 rte_mbuf_refcnt_read(const struct rte_mbuf *m)
 {
-   return (uint16_t)(rte_atomic16_read(&m->refcnt_atomic));
+   return __atomic_load_n(&m->refcnt, __ATOMIC_RELAXED);
 }
 
 /**
@@ -378,14 +377,15 @@ rte_mbuf_refcnt_read(const struct rte_mbuf *m)
 static inline void
 rte_mbuf_refcnt_set(struct rte_mbuf *m, uint16_t new_value)
 {
-   rte_atomic16_set(&m->refcnt_atomic, (int16_t)new_value);
+   __atomic_store_n(&m->refcnt, new_value, __ATOMIC_RELAXED);
 }
 
 /* internal */
 static inline uint16_t
 __rte_mbuf_refcnt_update(struct rte_mbuf *m, int16_t value)
 {
-   return (uint16_t)(rte_atomic16_add_return(&m->refcnt_atomic, value));
+   return (uint16_t)(__atomic_add_fetch((int16_t *)&m->refcnt, value,
+   __ATOMIC_ACQ_REL));
 }
 
 /**
@@ -466,7 +466,7 @@ rte_mbuf_refcnt_set(struct rte_mbuf *m, uint16_t new_value)
 static inline uint16_t
 rte_mbuf_ext_refcnt_read(const struct rte_mbuf_ext_shared_info *shinfo)
 {
-   return (uint16_t)(rte_atomic16_read(&shinfo->refcnt_atomic));
+   return __atomic_load_n(&shinfo->refcnt_atomic, __ATOMIC_RELAXED);
 }
 
 /**
@@ -481,7 +481,7 @@ static inline void
 rte_mbuf_ext_refcnt_set(struct rte_mbuf_ext_shared_info *shinfo,
uint16_t new_value)
 {
-   rte_atomic16_set(&shinfo->refcnt_atomic, (int16_t)new_value);
+   __atomic_store_n(&shinfo->refcnt_atomic, new_value, __ATOMIC_RELAXED);
 }
 
 /**
@@ -505,7 +505,8 @@ rte_mbuf_ext_refcnt_update(struct rte_mbuf_ext_shared_info 
*shinfo,
return (uint16_t)value;
}
 
-   return (uint16_t)rte_atomic16_add_return(&shinfo->refcnt_atomic, value);
+   return (uint16_t)(__atomic_add_fetch((int16_t *)&shinfo->refcnt_atomic,
+   value, __ATOMIC_ACQ_REL));
 }
 
 /** Mbuf prefetch */
@@ -1304,8 +1305,8 @@ static inline int 
__rte_pktmbuf_pinned_extbuf_decref(struct rte_mbuf *m)
 * Direct usage of add primitive to avoid
 * duplication of comparing with one.
 */
-   if (likely(rte_atomic16_add_return
-   (&shinfo->refcnt_atomic, -1)))
+   if (likely(__atomic_add_fetch((int *)&shinfo->refcnt_atomic, -1,
+__ATOMIC_ACQ_REL)))
return 1;
 
/* Reinitialize counter before mbuf freeing. */
diff --git a/lib/librte_mbuf/rte_mbuf_core.h b/lib/librte_mbuf/rte_mbuf_core.h
index 16600f1..806313a 100644
--- a/lib/librte_mbuf/rte_mbuf_core.h
+++ b/lib/librte_mbuf/rte_mbuf_core.h
@@ -18,7 +18,6 @@
 
 #include 
 #include 
-#include 
 
 #ifdef __cplusplus
 extern "C" {
@@ -495,12 +494,8 @@ struct rte_mbuf {
 * or non-atomic) is controlled by the CONFIG_RTE_MBUF_REFCNT_ATOMIC
 * config option.
 */
-   RTE_STD_C11
-   union {
-   rte_atomic16_t refcnt_atomic; /**< Atomically accessed refcnt */
-   /** Non-atomically accessed refcnt */
-   uint16_t refcnt;
-   };
+   uint16_t refcnt;
+
uint16_t nb_segs; /**< Number of segments. */
 
/** Input port (16 bits to support more than 256 virtual ports).
@@ -679,7 +674,7 @@ typedef void (*rte_mbuf_extbuf_free_callback_t)(void *addr, 
void *opaque);
 struct rte_mbuf_ext_shared_info {
rte_mbuf_extbuf_free_callback_t free_cb; /**< Free callback function */
void *fcb_opaque;/**< Free callback argument */
-   rte_atomic16_t refcnt_atomic;/**< Atomically accessed refcnt */
+   uint16_t refcnt_atomic;  /**< Atomically accessed refcnt */
 };
 
 /**< Maximum number of nb_segs allowed. */
-- 
2.7.4

Re: [dpdk-dev] [PATCH v2 1/3] ethdev: add new RSS types for IPv6 prefix

[Ferruh Yigit]

On 6/16/2020 9:16 AM, Junfeng Guo wrote:
> This patch defines new RSS offload types for IPv6 prefix with 32, 48,
> 64 bits of both SRC and DST IPv6 address.
> 
> Signed-off-by: Junfeng Guo 

Reviewed-by: Ferruh Yigit 

Re: [dpdk-dev] [PATCH v3 1/7] ethdev: introduce sample action for rte flow

[Andrew Rybchenko]

On 7/6/20 8:51 PM, Jiawei Wang wrote:
> When using full offload, all traffic will be handled by the HW, and
> directed to the requested VF or wire, the control application loses
> visibility on the traffic.
> So there's a need for an action that will enable the control application
> some visibility.
> 
> The solution is introduced a new action that will sample the incoming
> traffic and send a duplicated traffic with the specified ratio to the
> application, while the original packet will continue to the target
> destination.
> 
> The packets sampled equals is '1/ratio', if the ratio value be set to 1,
> means that the packets would be completely mirrored. The sample packet
> can be assigned with different set of actions from the original packet.
> 
> In order to support the sample packet in rte_flow, new rte_flow action
> definition RTE_FLOW_ACTION_TYPE_SAMPLE and structure rte_flow_action_sample
> will be introduced.
> 
> Signed-off-by: Jiawei Wang 
> Acked-by: Ori Kam 
> Acked-by: Jerin Jacob 

Acked-by: Andrew Rybchenko 

[dpdk-dev] [PATCH v3 1/2] service: fix wrong lcore indices

[Andrew Rybchenko]

From: Igor Romanov 

The service core list is populated, but not used. Incorrect
lcore states are examined for a service.

Use the populated list to iterate over service cores.

Fixes: e30dd31847d2 ("service: add mechanism for quiescing")
Cc: sta...@dpdk.org

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
---
 lib/librte_eal/common/rte_service.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/lib/librte_eal/common/rte_service.c 
b/lib/librte_eal/common/rte_service.c
index 6123a2124d..e2795f857e 100644
--- a/lib/librte_eal/common/rte_service.c
+++ b/lib/librte_eal/common/rte_service.c
@@ -422,7 +422,7 @@ rte_service_may_be_active(uint32_t id)
return -EINVAL;
 
for (i = 0; i < lcore_count; i++) {
-   if (lcore_states[i].service_active_on_lcore[id])
+   if (lcore_states[ids[i]].service_active_on_lcore[id])
return 1;
}
 
-- 
2.17.1

[dpdk-dev] [PATCH v3 2/2] test/service: add test for service active on two lcores

[Andrew Rybchenko]

From: Igor Romanov 

The test checks that the service may be active API works
when there are two cores: a non-service lcore and a service one.

The API notes to take care when checking the status of a running
service, but the test setup allows for a safe usage in that case.

Signed-off-by: Igor Romanov 
Signed-off-by: Andrew Rybchenko 
---
v3:
 - skip the test if number of cores is insufficient

 app/test/test_service_cores.c | 54 +++
 1 file changed, 54 insertions(+)

diff --git a/app/test/test_service_cores.c b/app/test/test_service_cores.c
index 981e212130..4f003ed1af 100644
--- a/app/test/test_service_cores.c
+++ b/app/test/test_service_cores.c
@@ -910,6 +910,59 @@ service_may_be_active(void)
return unregister_all();
 }
 
+/* check service may be active when service is running on a second lcore */
+static int
+service_active_two_cores(void)
+{
+   if (!rte_lcore_is_enabled(0) || !rte_lcore_is_enabled(1) ||
+   !rte_lcore_is_enabled(2))
+   return TEST_SKIPPED;
+
+   const uint32_t sid = 0;
+   int i;
+
+   uint32_t lcore = rte_get_next_lcore(/* start core */ -1,
+   /* skip master */ 1,
+   /* wrap */ 0);
+   uint32_t slcore = rte_get_next_lcore(/* start core */ lcore,
+/* skip master */ 1,
+/* wrap */ 0);
+
+   /* start the service on the second available lcore */
+   TEST_ASSERT_EQUAL(0, rte_service_runstate_set(sid, 1),
+   "Starting valid service failed");
+   TEST_ASSERT_EQUAL(0, rte_service_lcore_add(slcore),
+   "Add service core failed when not in use before");
+   TEST_ASSERT_EQUAL(0, rte_service_map_lcore_set(sid, slcore, 1),
+   "Enabling valid service on valid core failed");
+   TEST_ASSERT_EQUAL(0, rte_service_lcore_start(slcore),
+   "Service core start after add failed");
+
+   /* ensures core really is running the service function */
+   TEST_ASSERT_EQUAL(1, service_lcore_running_check(),
+   "Service core expected to poll service but it didn't");
+
+   /* ensures that service may be active reports running state */
+   TEST_ASSERT_EQUAL(1, rte_service_may_be_active(sid),
+   "Service may be active did not report running state");
+
+   /* stop the service */
+   TEST_ASSERT_EQUAL(0, rte_service_runstate_set(sid, 0),
+   "Error: Service stop returned non-zero");
+
+   /* give the service 100ms to stop running */
+   for (i = 0; i < 100; i++) {
+   if (!rte_service_may_be_active(sid))
+   break;
+   rte_delay_ms(SERVICE_DELAY);
+   }
+
+   TEST_ASSERT_EQUAL(0, rte_service_may_be_active(sid),
+ "Error: Service not stopped after 100ms");
+
+   return unregister_all();
+}
+
 static struct unit_test_suite service_tests  = {
.suite_name = "service core test suite",
.setup = testsuite_setup,
@@ -931,6 +984,7 @@ static struct unit_test_suite service_tests  = {
TEST_CASE_ST(dummy_register, NULL, service_app_lcore_mt_safe),
TEST_CASE_ST(dummy_register, NULL, service_app_lcore_mt_unsafe),
TEST_CASE_ST(dummy_register, NULL, service_may_be_active),
+   TEST_CASE_ST(dummy_register, NULL, service_active_two_cores),
TEST_CASES_END() /**< NULL terminate unit test array */
}
 };
-- 
2.17.1

Re: [dpdk-dev] [dpdk-techboard] [PATCH v3 0/3] Experimental/internal libraries cleanup

[Thomas Monjalon]

26/06/2020 11:25, Bruce Richardson:
> On Fri, Jun 26, 2020 at 10:16:35AM +0200, David Marchand wrote:
> > Following discussions on the mailing list and the 05/20 TB meeting, here
> > is a series that drops the special versioning for non stable libraries.
> > 
[...]
> The build changes, and patchset as a whole, look ok to me and good to see
> the code simplified by this.
> 
> Series-acked-by: Bruce Richardson 

Applied, thanks

Re: [dpdk-dev] [dpdk-stable] [PATCH v2] rawdev: fix to remove EXPERIMENTAL from comment

[Thomas Monjalon]

07/07/2020 11:13, David Marchand:
> On Tue, Jul 7, 2020 at 10:58 AM Hemant Agrawal  wrote:
> >
> > The experimental tags were removed, but the comment
> > is still having API classification as EXPERIMENTAL
> >
> > Fixes: 931cc531aad2 ("rawdev: remove experimental tag")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Hemant Agrawal 
> 
> Acked-by: David Marchand 

Applied, thanks

Re: [dpdk-dev] [PATCH v2 1/3] ethdev: add new RSS types for IPv6 prefix

[Thomas Monjalon]

16/06/2020 10:16, Junfeng Guo:
> This patch defines new RSS offload types for IPv6 prefix with 32, 48,
> 64 bits of both SRC and DST IPv6 address.
> 
> Signed-off-by: Junfeng Guo 
> ---
>  lib/librte_ethdev/rte_ethdev.h | 51 ++
>  1 file changed, 51 insertions(+)
> 
> diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
> index 631b146bd..5a7ba36d8 100644
> --- a/lib/librte_ethdev/rte_ethdev.h
> +++ b/lib/librte_ethdev/rte_ethdev.h
> @@ -538,6 +538,9 @@ struct rte_eth_rss_conf {
>  #define ETH_RSS_L4_DST_ONLY(1ULL << 60)
>  #define ETH_RSS_L2_SRC_ONLY(1ULL << 59)
>  #define ETH_RSS_L2_DST_ONLY(1ULL << 58)
> +#define ETH_RSS_L3_PRE32   (1ULL << 57)
> +#define ETH_RSS_L3_PRE48   (1ULL << 56)
> +#define ETH_RSS_L3_PRE64   (1ULL << 55)

PRE32, 48 and 64 are not obvious.
Why is it needed?
At least, please add comments for the values of this API.

Do we want to continue with the RTE_ prefix missing?
Can't we add the prefix for the new values?

Note: ethdev maintainers were not Cc'ed.
This reason is enough to nack the patch.

Re: [dpdk-dev] [PATCH] log: remove unneeded log type declaration

[David Marchand]

On Mon, Jul 6, 2020 at 8:10 PM Eads, Gage  wrote:
>
> > Subject: [dpdk-dev] [PATCH] log: remove unneeded log type declaration
> >
> > From: Jerin Jacob 
> >
> > RTE_LOG_REGISTER macro already declares the the log type.
> > Remove the unneeded log type declaration.
> >
> > Fixes: 9c99878aa1b16de2 ("log: introduce logtype register macro")
> >
> > Reported-by: David Marchand 
> > Signed-off-by: Jerin Jacob 
>
> Acked-by: Gage Eads 

Applied, thanks.


-- 
David Marchand

[dpdk-dev] [PATCH v3 1/4] eventdev: fix race condition on timer list counter

[Phil Yang]

The n_poll_lcores counter and poll_lcore array are shared between lcores
and the update of these variables are out of the protection of spinlock
on each lcore timer list. The read-modify-write operations of the counter
are not atomic, so it has the potential of race condition between lcores.

Use c11 atomics with RELAXED ordering to prevent confliction.

Fixes: cc7b73ea9e3b ("eventdev: add new software timer adapter")
Cc: erik.g.carri...@intel.com
Cc: sta...@dpdk.org

Signed-off-by: Phil Yang 
Reviewed-by: Dharmik Thakkar 
Reviewed-by: Ruifeng Wang 
Acked-by: Erik Gabriel Carrillo 
---
v2:
Align the code. (Erik)

 lib/librte_eventdev/rte_event_timer_adapter.c | 16 
 1 file changed, 12 insertions(+), 4 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
b/lib/librte_eventdev/rte_event_timer_adapter.c
index 2321803..370ea40 100644
--- a/lib/librte_eventdev/rte_event_timer_adapter.c
+++ b/lib/librte_eventdev/rte_event_timer_adapter.c
@@ -583,6 +583,7 @@ swtim_callback(struct rte_timer *tim)
uint16_t nb_evs_invalid = 0;
uint64_t opaque;
int ret;
+   int n_lcores;
 
opaque = evtim->impl_opaque[1];
adapter = (struct rte_event_timer_adapter *)(uintptr_t)opaque;
@@ -605,8 +606,12 @@ swtim_callback(struct rte_timer *tim)
  "with immediate expiry value");
}
 
-   if (unlikely(rte_atomic16_test_and_set(&sw->in_use[lcore].v)))
-   sw->poll_lcores[sw->n_poll_lcores++] = lcore;
+   if (unlikely(rte_atomic16_test_and_set(&sw->in_use[lcore].v))) {
+   n_lcores = __atomic_fetch_add(&sw->n_poll_lcores, 1,
+__ATOMIC_RELAXED);
+   __atomic_store_n(&sw->poll_lcores[n_lcores], lcore,
+   __ATOMIC_RELAXED);
+   }
} else {
EVTIM_BUF_LOG_DBG("buffered an event timer expiry event");
 
@@ -1011,6 +1016,7 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
uint32_t lcore_id = rte_lcore_id();
struct rte_timer *tim, *tims[nb_evtims];
uint64_t cycles;
+   int n_lcores;
 
 #ifdef RTE_LIBRTE_EVENTDEV_DEBUG
/* Check that the service is running. */
@@ -1033,8 +1039,10 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
if (unlikely(rte_atomic16_test_and_set(&sw->in_use[lcore_id].v))) {
EVTIM_LOG_DBG("Adding lcore id = %u to list of lcores to poll",
  lcore_id);
-   sw->poll_lcores[sw->n_poll_lcores] = lcore_id;
-   ++sw->n_poll_lcores;
+   n_lcores = __atomic_fetch_add(&sw->n_poll_lcores, 1,
+__ATOMIC_RELAXED);
+   __atomic_store_n(&sw->poll_lcores[n_lcores], lcore_id,
+   __ATOMIC_RELAXED);
}
 
ret = rte_mempool_get_bulk(sw->tim_pool, (void **)tims,
-- 
2.7.4

[dpdk-dev] [PATCH v3 3/4] eventdev: remove redundant code

[Phil Yang]

There is no thread will access these impl_opaque data after timer
canceled. When new timer armed, it got refilled. So the cleanup
process is unnecessary.

Signed-off-by: Phil Yang 
Reviewed-by: Dharmik Thakkar 
---
 lib/librte_eventdev/rte_event_timer_adapter.c | 2 --
 1 file changed, 2 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
b/lib/librte_eventdev/rte_event_timer_adapter.c
index 6d01a34..d75415c 100644
--- a/lib/librte_eventdev/rte_event_timer_adapter.c
+++ b/lib/librte_eventdev/rte_event_timer_adapter.c
@@ -1167,8 +1167,6 @@ swtim_cancel_burst(const struct rte_event_timer_adapter 
*adapter,
rte_mempool_put(sw->tim_pool, (void **)timp);
 
evtims[i]->state = RTE_EVENT_TIMER_CANCELED;
-   evtims[i]->impl_opaque[0] = 0;
-   evtims[i]->impl_opaque[1] = 0;
 
rte_smp_wmb();
}
-- 
2.7.4

[dpdk-dev] [PATCH v3 2/4] eventdev: use C11 atomics for lcore timer armed flag

[Phil Yang]

The in_use flag is a per core variable which is not shared between
lcores in the normal case and the access of this variable should be
ordered on the same core. However, if non-EAL thread pick the highest
lcore to insert timers into, there is the possibility of conflicts
on this flag between threads. Then the atomic CAS operation is needed.

Use the C11 atomic CAS instead of the generic rte_atomic operations
to avoid the unnecessary barrier on aarch64.

Signed-off-by: Phil Yang 
Reviewed-by: Dharmik Thakkar 
Reviewed-by: Ruifeng Wang 
Acked-by: Erik Gabriel Carrillo 
---
v2:
1. Make the code comments more accurate. (Erik)
2. Define the in_use flag as an unsigned type. (Stephen)

 lib/librte_eventdev/rte_event_timer_adapter.c | 15 +++
 1 file changed, 11 insertions(+), 4 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
b/lib/librte_eventdev/rte_event_timer_adapter.c
index 370ea40..6d01a34 100644
--- a/lib/librte_eventdev/rte_event_timer_adapter.c
+++ b/lib/librte_eventdev/rte_event_timer_adapter.c
@@ -554,7 +554,7 @@ struct swtim {
uint32_t timer_data_id;
/* Track which cores have actually armed a timer */
struct {
-   rte_atomic16_t v;
+   uint16_t v;
} __rte_cache_aligned in_use[RTE_MAX_LCORE];
/* Track which cores' timer lists should be polled */
unsigned int poll_lcores[RTE_MAX_LCORE];
@@ -606,7 +606,8 @@ swtim_callback(struct rte_timer *tim)
  "with immediate expiry value");
}
 
-   if (unlikely(rte_atomic16_test_and_set(&sw->in_use[lcore].v))) {
+   if (unlikely(sw->in_use[lcore].v == 0)) {
+   sw->in_use[lcore].v = 1;
n_lcores = __atomic_fetch_add(&sw->n_poll_lcores, 1,
 __ATOMIC_RELAXED);
__atomic_store_n(&sw->poll_lcores[n_lcores], lcore,
@@ -834,7 +835,7 @@ swtim_init(struct rte_event_timer_adapter *adapter)
 
/* Initialize the variables that track in-use timer lists */
for (i = 0; i < RTE_MAX_LCORE; i++)
-   rte_atomic16_init(&sw->in_use[i].v);
+   sw->in_use[i].v = 0;
 
/* Initialize the timer subsystem and allocate timer data instance */
ret = rte_timer_subsystem_init();
@@ -1017,6 +1018,8 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
struct rte_timer *tim, *tims[nb_evtims];
uint64_t cycles;
int n_lcores;
+   /* Timer list for this lcore is not in use. */
+   uint16_t exp_state = 0;
 
 #ifdef RTE_LIBRTE_EVENTDEV_DEBUG
/* Check that the service is running. */
@@ -1035,8 +1038,12 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
/* If this is the first time we're arming an event timer on this lcore,
 * mark this lcore as "in use"; this will cause the service
 * function to process the timer list that corresponds to this lcore.
+* The atomic CAS operation can prevent the race condition on in_use
+* flag between multiple non-EAL threads.
 */
-   if (unlikely(rte_atomic16_test_and_set(&sw->in_use[lcore_id].v))) {
+   if (unlikely(__atomic_compare_exchange_n(&sw->in_use[lcore_id].v,
+   &exp_state, 1, 0,
+   __ATOMIC_RELAXED, __ATOMIC_RELAXED))) {
EVTIM_LOG_DBG("Adding lcore id = %u to list of lcores to poll",
  lcore_id);
n_lcores = __atomic_fetch_add(&sw->n_poll_lcores, 1,
-- 
2.7.4

[dpdk-dev] [PATCH v3 4/4] eventdev: relax smp barriers with C11 atomics

[Phil Yang]

The impl_opaque field is shared between the timer arm and cancel
operations. Meanwhile, the state flag acts as a guard variable to
make sure the update of impl_opaque is synchronized. The original
code uses rte_smp barriers to achieve that. This patch uses C11
atomics with an explicit one-way memory barrier instead of full
barriers rte_smp_w/rmb() to avoid the unnecessary barrier on aarch64.

Since compilers can generate the same instructions for volatile and
non-volatile variable in C11 __atomics built-ins, so remain the volatile
keyword in front of state enum to avoid the ABI break issue.

Signed-off-by: Phil Yang 
Reviewed-by: Dharmik Thakkar 
Reviewed-by: Ruifeng Wang 
Acked-by: Erik Gabriel Carrillo 
---
v3:
Fix ABI issue: revert to 'volatile enum rte_event_timer_state type state'.

v2:
1. Removed implementation-specific opaque data cleanup code.
2. Replaced thread fence with atomic ACQURE/RELEASE ordering on state access.

 lib/librte_eventdev/rte_event_timer_adapter.c | 55 ++-
 1 file changed, 37 insertions(+), 18 deletions(-)

diff --git a/lib/librte_eventdev/rte_event_timer_adapter.c 
b/lib/librte_eventdev/rte_event_timer_adapter.c
index d75415c..eb2c93a 100644
--- a/lib/librte_eventdev/rte_event_timer_adapter.c
+++ b/lib/librte_eventdev/rte_event_timer_adapter.c
@@ -629,7 +629,8 @@ swtim_callback(struct rte_timer *tim)
sw->expired_timers[sw->n_expired_timers++] = tim;
sw->stats.evtim_exp_count++;
 
-   evtim->state = RTE_EVENT_TIMER_NOT_ARMED;
+   __atomic_store_n(&evtim->state, RTE_EVENT_TIMER_NOT_ARMED,
+   __ATOMIC_RELEASE);
}
 
if (event_buffer_batch_ready(&sw->buffer)) {
@@ -1020,6 +1021,7 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
int n_lcores;
/* Timer list for this lcore is not in use. */
uint16_t exp_state = 0;
+   enum rte_event_timer_state n_state;
 
 #ifdef RTE_LIBRTE_EVENTDEV_DEBUG
/* Check that the service is running. */
@@ -1060,30 +1062,36 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
}
 
for (i = 0; i < nb_evtims; i++) {
-   /* Don't modify the event timer state in these cases */
-   if (evtims[i]->state == RTE_EVENT_TIMER_ARMED) {
+   n_state = __atomic_load_n(&evtims[i]->state, __ATOMIC_ACQUIRE);
+   if (n_state == RTE_EVENT_TIMER_ARMED) {
rte_errno = EALREADY;
break;
-   } else if (!(evtims[i]->state == RTE_EVENT_TIMER_NOT_ARMED ||
-evtims[i]->state == RTE_EVENT_TIMER_CANCELED)) {
+   } else if (!(n_state == RTE_EVENT_TIMER_NOT_ARMED ||
+n_state == RTE_EVENT_TIMER_CANCELED)) {
rte_errno = EINVAL;
break;
}
 
ret = check_timeout(evtims[i], adapter);
if (unlikely(ret == -1)) {
-   evtims[i]->state = RTE_EVENT_TIMER_ERROR_TOOLATE;
+   __atomic_store_n(&evtims[i]->state,
+   RTE_EVENT_TIMER_ERROR_TOOLATE,
+   __ATOMIC_RELAXED);
rte_errno = EINVAL;
break;
} else if (unlikely(ret == -2)) {
-   evtims[i]->state = RTE_EVENT_TIMER_ERROR_TOOEARLY;
+   __atomic_store_n(&evtims[i]->state,
+   RTE_EVENT_TIMER_ERROR_TOOEARLY,
+   __ATOMIC_RELAXED);
rte_errno = EINVAL;
break;
}
 
if (unlikely(check_destination_event_queue(evtims[i],
   adapter) < 0)) {
-   evtims[i]->state = RTE_EVENT_TIMER_ERROR;
+   __atomic_store_n(&evtims[i]->state,
+   RTE_EVENT_TIMER_ERROR,
+   __ATOMIC_RELAXED);
rte_errno = EINVAL;
break;
}
@@ -1099,13 +1107,18 @@ __swtim_arm_burst(const struct rte_event_timer_adapter 
*adapter,
  SINGLE, lcore_id, NULL, evtims[i]);
if (ret < 0) {
/* tim was in RUNNING or CONFIG state */
-   evtims[i]->state = RTE_EVENT_TIMER_ERROR;
+   __atomic_store_n(&evtims[i]->state,
+   RTE_EVENT_TIMER_ERROR,
+   __ATOMIC_RELEASE);
break;
}
 
-   rte_smp_wmb();
EVTIM_LOG_DBG("armed an event timer");
-   evtims[i]->state = RTE_EVENT_TIMER_ARMED;
+

Re: [dpdk-dev] [PATCH v2 1/2] eal/arm: generic counter based loop for CPU freq calculation

[David Marchand]

On Fri, Jun 26, 2020 at 10:35 PM Honnappa Nagarahalli
 wrote:
>
> get_tsc_freq uses 'nanosleep' system call to calculate the CPU
> frequency. However, 'nanosleep' results in the process getting
> un-scheduled. The kernel saves and restores the PMU state. This
> ensures that the PMU cycles are not counted towards a sleeping
> process. When RTE_ARM_EAL_RDTSC_USE_PMU is defined, this results
> in incorrect CPU frequency calculation. This logic is replaced
> with generic counter based loop.
>
> Bugzilla ID: 450
> Fixes: f91bcbb2d9a6 ("eal/arm: use high-resolution cycle counter")
> Cc: sta...@dpdk.org
>
> Signed-off-by: Honnappa Nagarahalli 
> Reviewed-by: Ruifeng Wang 
> Reviewed-by: Dharmik Thakkar 
> Reviewed-by: Phil Yang 
> Acked-by: Jerin Jacob 

Series applied, thanks.


-- 
David Marchand

Re: [dpdk-dev] [dpdk-stable] [PATCH] app/test: fix rpath for drivers in meson build

[David Marchand]

On Wed, Jul 1, 2020 at 7:26 PM Timothy Redaelli  wrote:
> This commit fixes the setting of relative rpath on dpdk-test for
> drivers ($libdir/dpdk/pmd-$abiver) to the correct absolute rpath
> ($prefix$libdir/dpdk/pmd-$abiver)
>
> Fixes: b5dc795a8a55 ("test: build app with meson as dpdk-test")
> Cc: sta...@dpdk.org
> Signed-off-by: Timothy Redaelli 

Acked-by: Bruce Richarson 

Applied, thanks.


-- 
David Marchand

Re: [dpdk-dev] [PATCH v4 01/10] eal: introduce macros for getting valuefor bit

[Parav Pandit]

> From: Morten Brørup 
> Sent: Monday, July 6, 2020 4:24 PM
> 
> > From: dev [mailto:dev-boun...@dpdk.org] On Behalf Of Parav Pandit
> > Sent: Friday, July 3, 2020 3:47 PM
> >
> > There are several drivers which duplicate bit generation macro.
> > Introduce a generic bit macros so that such drivers avoid redefining
> > same in multiple drivers.
> >
> > Signed-off-by: Parav Pandit 
> > Acked-by: Matan Azrad 
> > ---
> > Changelog:
> > v1->v2:
> >  - Addressed comments from Thomas and Gaten.
> >  - Avoided new file, added macro to rte_bitops.h
> > ---
> >  lib/librte_eal/include/rte_bitops.h | 2 ++
> >  1 file changed, 2 insertions(+)
> >
> > diff --git a/lib/librte_eal/include/rte_bitops.h
> > b/lib/librte_eal/include/rte_bitops.h
> > index 740927f3b..d72c7cd93 100644
> > --- a/lib/librte_eal/include/rte_bitops.h
> > +++ b/lib/librte_eal/include/rte_bitops.h
> > @@ -17,6 +17,8 @@
> >  #include 
> >  #include 
> >
> > +#define RTE_BIT(bit_num)   (1UL << (bit_num))
> 
> Is the return value 32 or 64 bit, or is intended to depend on the target
> architecture?
>
It should be 64-bit.
 
> Please be explicit by using UINT32_C(1) or UINT64_C(1) instead of 1UL, if you
> want a specific size.
> 
Will do UINT64_C(1).

> It could be a static inline __attribute__((__pure__)) function instead of a 
> macro,
> but it's not important for me.
> 
> The macro/function needs a description for the documentation.
> 
In this header file or outside?

> I'm also concerned about the name of the macro being too generic. But the
> effort of changing all the drivers where it is being used already could be 
> too big
> if the name changes too.
>
Right. Currently drivers have generic name as BIT(). Close to 3000 entries.
So doing at RTE_BIT to match other rte_ APIs.
Drivers can slowly migrate at their pace to this one.
 
> And the macro/function is new, so shouldn't it - in theory - be marked as
> experimental?

How to mark a macro as experimental?

Re: [dpdk-dev] [PATCH] eal: return -1 if rte_lcore_index called from non-DPDK thread

[David Marchand]

On Wed, Apr 8, 2020 at 10:24 PM Stephen Hemminger
 wrote:
>
> If rte_lcore_index() is asked to give the index of the
> current lcore (argument -1) and is called from a non-DPDK thread
> then it would invalid result. The result would come
> lcore_config[-1].core_index which is some other data in the
> per-thread area.
>
> The resolution is to return -1 which is what rte_lcore_index()
> returns if handed an invalid lcore.
>
> Same issue existed with rte_lcore_to_cpu_id().
>
> Bugzilla ID: 446
> Signed-off-by: Stephen Hemminger 

Acked-by: David Marchand 

Applied, thanks.


-- 
David Marchand

Re: [dpdk-dev] [PATCH 1/2] mbuf: introduce accurate packet Tx scheduling

[Olivier Matz]

Hi Slava,

Few question/comments below.

On Wed, Jul 01, 2020 at 03:36:26PM +, Viacheslav Ovsiienko wrote:
> There is the requirement on some networks for precise traffic timing
> management. The ability to send (and, generally speaking, receive)
> the packets at the very precisely specified moment of time provides
> the opportunity to support the connections with Time Division
> Multiplexing using the contemporary general purpose NIC without involving
> an auxiliary hardware. For example, the supporting of O-RAN Fronthaul
> interface is one of the promising features for potentially usage of the
> precise time management for the egress packets.
> 
> The main objective of this RFC is to specify the way how applications
> can provide the moment of time at what the packet transmission must be
> started and to describe in preliminary the supporting this feature from
> mlx5 PMD side.
> 
> The new dynamic timestamp field is proposed, it provides some timing
> information, the units and time references (initial phase) are not
> explicitly defined but are maintained always the same for a given port.
> Some devices allow to query rte_eth_read_clock() that will return
> the current device timestamp. The dynamic timestamp flag tells whether
> the field contains actual timestamp value. For the packets being sent
> this value can be used by PMD to schedule packet sending.
> 
> After PKT_RX_TIMESTAMP flag and fixed timestamp field deprecation
> and obsoleting, these dynamic flag and field will be used to manage
> the timestamps on receiving datapath as well.

Do you mean the same flag will be used for both Rx and Tx?  I wonder if
it's a good idea: if you enable the timestamp on Rx, the packets will be
flagged and it will impact Tx, except if the application explicitly
resets the flag in all mbufs. Wouldn't it be safer to have an Rx flag
and a Tx flag?

> When PMD sees the "rte_dynfield_timestamp" set on the packet being sent
> it tries to synchronize the time of packet appearing on the wire with
> the specified packet timestamp. It the specified one is in the past it
> should be ignored, if one is in the distant future it should be capped
> with some reasonable value (in range of seconds). These specific cases
> ("too late" and "distant future") can be optionally reported via
> device xstats to assist applications to detect the time-related
> problems.

I think what to do with packets to be send in the "past" could be
configurable through an ethdev API in the future (drop or send).

> There is no any packet reordering according timestamps is supposed,
> neither within packet burst, nor between packets, it is an entirely
> application responsibility to generate packets and its timestamps
> in desired order. The timestamps can be put only in the first packet
> in the burst providing the entire burst scheduling.

This constraint makes sense. At first glance, it looks it is imposed by
a PMD or hw limitation, but thinking more about it, I think it is the
correct behavior to have. Packets are ordered within a PMD queue, and
the ability to set the timestamp for one packet to delay the subsequent
ones looks useful.

Should this behavior be documented somewhere? Maybe in the API comment
documenting the dynamic flag?

> PMD reports the ability to synchronize packet sending on timestamp
> with new offload flag:
> 
> This is palliative and is going to be replaced with new eth_dev API
> about reporting/managing the supported dynamic flags and its related
> features. This API would break ABI compatibility and can't be introduced
> at the moment, so is postponed to 20.11.
> 
> For testing purposes it is proposed to update testpmd "txonly"
> forwarding mode routine. With this update testpmd application generates
> the packets and sets the dynamic timestamps according to specified time
> pattern if it sees the "rte_dynfield_timestamp" is registered.
> 
> The new testpmd command is proposed to configure sending pattern:
> 
> set tx_times ,
> 
>  - the delay between the packets within the burst
>   specified in the device clock units. The number
>   of packets in the burst is defined by txburst parameter
> 
>  - the delay between the bursts in the device clock units
> 
> As the result the bursts of packet will be transmitted with specific
> delays between the packets within the burst and specific delay between
> the bursts. The rte_eth_get_clock is supposed to be engaged to get the
> current device clock value and provide the reference for the timestamps.
> 
> Signed-off-by: Viacheslav Ovsiienko 
> ---
>  lib/librte_ethdev/rte_ethdev.c |  1 +
>  lib/librte_ethdev/rte_ethdev.h |  4 
>  lib/librte_mbuf/rte_mbuf_dyn.h | 16 
>  3 files changed, 21 insertions(+)
> 
> diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c
> index 8e10a6f..02157d5 100644
> --- a/lib/librte_ethdev/rte_ethdev.c
> +++ b/lib/librte_ethdev/rte_ethdev.c
> @@ -162,6 +162,7 @@ struct rte_eth_xstats_na

Re: [dpdk-dev] [dpdk-stable] [PATCH] vfio: remove unused variable

[David Marchand]

On Mon, Jun 29, 2020 at 11:44 AM David Marchand
 wrote:
>
> On Sat, May 16, 2020 at 9:59 AM wangyunjian  wrote:
> >
> > From: Yunjian Wang 
> >
> > The 'group_status' has never been used and can be removed.
> >
> > Fixes: 94c0776b1bad ("vfio: support hotplug")
> > Cc: sta...@dpdk.org
> >
> > Signed-off-by: Yunjian Wang 
> Reviewed-by: David Marchand 

Applied, thanks.


-- 
David Marchand

Re: [dpdk-dev] [PATCH v2] eal/vfio: reduce severity of startup messages

[David Marchand]

On Fri, Jun 12, 2020 at 11:07 AM Burakov, Anatoly
 wrote:
>
> On 12-Jun-20 1:10 AM, Stephen Hemminger wrote:
> > The startup of VFIO is too noisy. Logging is expensive on some
> > systems, and distracting to the user.
> >
> > It should not be logging at NOTICE level, reduce it to INFO level.
> > It really should be DEBUG here but that would hide it by default.
> >
> > Signed-off-by: Stephen Hemminger 
> Acked-by: Anatoly Burakov 

Applied, thanks.


-- 
David Marchand

Re: [dpdk-dev] [PATCH v2 2/3] app/testpmd: support extended RSS offload types

[Ferruh Yigit]

On 7/2/2020 6:11 AM, Simei Su wrote:
> This patch adds testpmd cmdline support for PPPoE.
> 
> Signed-off-by: Simei Su 

Reviewed-by: Ferruh Yigit 

Re: [dpdk-dev] [PATCH 09/27] event/dlb: add inline functions used in multiple files

[Bruce Richardson]

On Fri, Jun 26, 2020 at 11:37:33PM -0500, Tim McDaniel wrote:
> From: "McDaniel, Timothy" 
> 
> Signed-off-by: McDaniel, Timothy 
> ---
>  drivers/event/dlb/dlb_inline_fns.h |   80 
> 
>  1 file changed, 80 insertions(+)
>  create mode 100644 drivers/event/dlb/dlb_inline_fns.h
> 
> diff --git a/drivers/event/dlb/dlb_inline_fns.h 
> b/drivers/event/dlb/dlb_inline_fns.h
> new file mode 100644
> index 000..86f85aa
> --- /dev/null
> +++ b/drivers/event/dlb/dlb_inline_fns.h
> @@ -0,0 +1,80 @@
> +/* SPDX-License-Identifier: BSD-3-Clause
> + * Copyright(c) 2016-2020 Intel Corporation
> + */
> +
> +#include "rte_memcpy.h"
> +#include "rte_io.h"
> +
> +/* Inline functions required in more than one source file.
> + */
> +
> +static inline struct dlb_eventdev *
> +dlb_pmd_priv(const struct rte_eventdev *eventdev)
> +{
> + return eventdev->data->dev_private;
> +}
> +
> +static inline void
> +dlb_umonitor(volatile void *addr)
> +{
> + asm volatile(".byte 0xf3, 0x0f, 0xae, 0xf7\t\n"
> + :
> + : "D" (addr));
> +}
> +
> +static inline void
> +dlb_umwait(int state, uint64_t timeout)
> +{
> + uint32_t eax = timeout & UINT32_MAX;
> + uint32_t edx = timeout >> 32;
> +
> + asm volatile(".byte 0xf2, 0x0f, 0xae, 0xf7\t\n"
> + :
> + : "D" (state),  "a" (eax), "d" (edx));
> +}
> +
> +static inline void
> +dlb_movntdq(void *qe4, void *pp_addr)
> +{
> + /* Move entire 64B cache line of QEs, 128 bits (16B) at a time. */
> + long long *_qe  = (long long *)qe4;
> + __v2di src_data0 = (__v2di){_qe[0], _qe[1]};
> + __v2di src_data1 = (__v2di){_qe[2], _qe[3]};
> + __v2di src_data2 = (__v2di){_qe[4], _qe[5]};
> + __v2di src_data3 = (__v2di){_qe[6], _qe[7]};
> +
> + __builtin_ia32_movntdq((__v2di *)pp_addr + 0, (__v2di)src_data0);
> + rte_wmb();
> + __builtin_ia32_movntdq((__v2di *)pp_addr + 1, (__v2di)src_data1);
> + rte_wmb();
> + __builtin_ia32_movntdq((__v2di *)pp_addr + 2, (__v2di)src_data2);
> + rte_wmb();
> + __builtin_ia32_movntdq((__v2di *)pp_addr + 3, (__v2di)src_data3);
> + rte_wmb();
> +}
> +
> +static inline void
> +dlb_movntdq_single(void *qe4, void *pp_addr)
> +{
> + long long *_qe  = (long long *)qe4;
> + __v2di src_data0 = (__v2di){_qe[0], _qe[1]};
> +
> + __builtin_ia32_movntdq((__v2di *)pp_addr, (__v2di)src_data0);
> +}
> +
> +static inline void
> +dlb_cldemote(void *addr)
> +{
> + /* Load addr into RSI, then demote the cache line of the address
> +  * contained in that register.
> +  */
> + asm volatile(".byte 0x0f, 0x1c, 0x06" :: "S" (addr));
> +}
> +
> +static inline void
> +dlb_movdir64b(void *qe4, void *pp_addr)
> +{
> + asm volatile(".byte 0x66, 0x0f, 0x38, 0xf8, 0x02"
> +  :
> +  : "a" (pp_addr), "d" (qe4));
> +}
> -- 

Although these are mostly x86-specific, a number of these functions may be
needed in other places in DPDK also, so they should perhaps be put in a
common location.

Also, for the use of these, can you try and align the parameter names and
order as far as possible with those of the equivalent x86 intrinsics
documented at
https://software.intel.com/sites/landingpage/IntrinsicsGuide/. This will
make it easier in the future to switch to use the proper intrinsics rather
than our own versions. For example, _movdir64b intrinsic has parameters
"void *dst, const void *src", which is reversed to what you have here.

Thanks,
/Bruce

Re: [dpdk-dev] [PATCH v17 2/2] eal: support for VFIO-PCI VF token

[David Marchand]

On Fri, Jul 3, 2020 at 5:05 PM Haiyue Wang  wrote:
>
> The kernel module vfio-pci introduces the VF token to enable SR-IOV
> support since 5.7.
>
> The VF token can be set by a vfio-pci based PF driver and must be known
> by the vfio-pci based VF driver in order to gain access to the device.
>
> Since the vfio-pci module uses the VF token as internal data to provide
> the collaboration between SR-IOV PF and VFs, so DPDK can use the same
> VF token for all PF devices by specifying the related EAL option.
>
> Signed-off-by: Haiyue Wang 
> Acked-by: Anatoly Burakov 
> Acked-by: Andrew Rybchenko 
> Tested-by: Harman Kalra 

Series applied, thanks Haiyue.


-- 
David Marchand

Re: [dpdk-dev] [PATCH v4 01/10] eal: introduce macros for getting valuefor bit

[Thomas Monjalon]

07/07/2020 13:38, Parav Pandit:
> From: Morten Brørup 
> > From: Parav Pandit
> > > --- a/lib/librte_eal/include/rte_bitops.h
> > > +++ b/lib/librte_eal/include/rte_bitops.h
> > > @@ -17,6 +17,8 @@
> > >  #include 
> > >  #include 
> > >
> > > +#define RTE_BIT(bit_num) (1UL << (bit_num))
> > 
> > Is the return value 32 or 64 bit, or is intended to depend on the target
> > architecture?
> >
> It should be 64-bit.
>  
> > Please be explicit by using UINT32_C(1) or UINT64_C(1) instead of 1UL, if 
> > you
> > want a specific size.
> > 
> Will do UINT64_C(1).
> 
> > It could be a static inline __attribute__((__pure__)) function instead of a 
> > macro,
> > but it's not important for me.
> > 
> > The macro/function needs a description for the documentation.
> > 
> In this header file or outside?

It is asked to add a doxygen comment.


> > I'm also concerned about the name of the macro being too generic. But the
> > effort of changing all the drivers where it is being used already could be 
> > too big
> > if the name changes too.
> >
> Right. Currently drivers have generic name as BIT(). Close to 3000 entries.
> So doing at RTE_BIT to match other rte_ APIs.
> Drivers can slowly migrate at their pace to this one.
>  
> > And the macro/function is new, so shouldn't it - in theory - be marked as
> > experimental?
> 
> How to mark a macro as experimental?

A macro cannot be experimental.

Re: [dpdk-dev] [PATCH 1/3] rawdev: allow calling info function for unknown rawdevs

[Thomas Monjalon]

06/07/2020 12:31, Bruce Richardson:
> To call the rte_rawdev_info_get() function, the user currently has to know
> the underlying type of the device in order to pass an appropriate structure
> or buffer as the dev_private pointer in the info structure. By allowing a
> NULL value for this field, we can skip getting the device-specific info and
> just return the generic info - including the device name and driver, which
> can be used to determine the device type - to the user.
> 
> This ensures that basic info can be get for all rawdevs, without knowing
> the type, and even if the info driver API call has not been implemented for
> the device.
> 
> Cc: sta...@dpdk.org

This is a new feature. Why should it be backported?

Re: [dpdk-dev] [dpdk-stable] [PATCH 2/3] rawdev: return NUMA socket id to the user

[Thomas Monjalon]

06/07/2020 12:31, Bruce Richardson:
> The rawdev info struct has a socket_id field which was not filled in.
> 
> We can also omit the checks for the parameter struct being null, since
> that is previously checked in the function.
> 
> Cc: sta...@dpdk.org

It looks to be a fix.
Which commit introduced this bug? The original one?
Fixes: c88b3f2558ed ("rawdev: introduce raw device library")

Re: [dpdk-dev] [PATCH 1/3] rawdev: allow calling info function for unknown rawdevs

[Bruce Richardson]

On Tue, Jul 07, 2020 at 02:18:31PM +0200, Thomas Monjalon wrote:
> 06/07/2020 12:31, Bruce Richardson:
> > To call the rte_rawdev_info_get() function, the user currently has to
> > know the underlying type of the device in order to pass an appropriate
> > structure or buffer as the dev_private pointer in the info structure.
> > By allowing a NULL value for this field, we can skip getting the
> > device-specific info and just return the generic info - including the
> > device name and driver, which can be used to determine the device type
> > - to the user.
> > 
> > This ensures that basic info can be get for all rawdevs, without
> > knowing the type, and even if the info driver API call has not been
> > implemented for the device.
> > 
> > Cc: sta...@dpdk.org
> 
> This is a new feature. Why should it be backported?
> 

I'd largely leave it up to the stable maintainers, but IMHO it should be
backported as its not a big change and I don't see the existing rawdev APIs
as very usable without it. I think the current API leaves the user in a bit
of a catch 22, since info_get() is the API you'd expect to call to find out
the actual type of a rawdev, but unfortunately, to call the info_get API
you need to know the type to pass in the appropriate type-specific
structure parameter. Therefore I see this as fixing a usability bug. :-)

Regards,
/Bruce

Re: [dpdk-dev] [dpdk-stable] [PATCH 2/3] rawdev: return NUMA socket id to the user

[Bruce Richardson]

On Tue, Jul 07, 2020 at 02:23:04PM +0200, Thomas Monjalon wrote:
> 06/07/2020 12:31, Bruce Richardson:
> > The rawdev info struct has a socket_id field which was not filled in.
> > 
> > We can also omit the checks for the parameter struct being null, since
> > that is previously checked in the function.
> > 
> > Cc: sta...@dpdk.org
> 
> It looks to be a fix.
> Which commit introduced this bug? The original one?
> Fixes: c88b3f2558ed ("rawdev: introduce raw device library")
>

Yes, it can be viewed as a fix and that is the commit introducing the
issue.

/Bruce

Re: [dpdk-dev] [dpdk-dev v4 1/4] cryptodev: add symmetric crypto data-path APIs

[Zhang, Roy Fan]

Hi Akhil,

> -Original Message-
> From: Akhil Goyal 
> Sent: Monday, July 6, 2020 1:13 PM
> To: Zhang, Roy Fan ; dev@dpdk.org;
> ano...@marvell.com; asoma...@amd.com; ruifeng.w...@arm.com;
> Nagadheeraj Rottela ; Michael Shamis
> ; Ankur Dwivedi ; Jay
> Zhou ; De Lara Guarch, Pablo
> 
> Cc: Trahe, Fiona ; Bronowski, PiotrX
> ; Ananyev, Konstantin
> ; Thomas Monjalon
> 
> Subject: RE: [dpdk-dev v4 1/4] cryptodev: add symmetric crypto data-path
> APIs
> 
...
> >
> > As you may have seen the structure definition is hW centric with the
> > IOVA addresses all over. Also as you will from the patch series the
> operation is
> > Per operation basis instead of operating in a burst. The external 
> > application
> > may sooner know when a specific enqueue is failed.
> 
> You may also need to save a virtual address as well. As some hardware are
> able to
> Convert virtual to physical addresses on it's own giving a performance
> improvement.
> 
> I do not see an issue in using enqueue burst with burst size=1 , but since you
> are doing
> Optimizations, none of the hardware can perform well with burst = 1, I think
> it is always
> Greater than 1.

Shall I update the rte_crypto_sym_vec as the following - so the 2 problems can 
be
resolved?

struct rte_crypto_sym_vec {
/** array of SGL vectors */
struct rte_crypto_sgl *sgl;
union {
/* Supposed to be used with CPU crypto API call. */
struct {
/** array of pointers to IV */
void **iv;
/** array of pointers to AAD */
void **aad;
/** array of pointers to digest */
void **digest;
/**
 * array of statuses for each operation:
 *  - 0 on success
 *  - errno on error
 */
int32_t *status;
};

/* Supposed to be used with HW crypto API call. */
struct {
/** array of pointers to IV */
struct rte_crypto_vec *iv_hw;
/** array of pointers to AAD */
struct rte_crypto_vec *aad_hw;
/** array of pointers to Digest */
struct rte_crypto_vec *digest_hw;
};

};
/** number of operations to perform */
uint32_t num;
};

> >
> > > Now for this patchset, the requirement is
> > > - raw buffers
> > > - asynchronous APIs
> > >
> > > The data structure for raw buffers and crypto related offsets are already
> > > defined
> > > So they should be reused.
> > > And I believe with some changes in rte_crypto_op  and
> rte_crypto_sym_op,
> > > We can support raw buffers with the same APIs.
> > > Instead of m_src and m_dst, raw buffer data structures can be combined
> in a
> > > Union and some of the fields in the rte_crypto_op can be left NULL in
> case of
> > > raw buffers.
> > >
> >
> > This is a good point but we still face too many unnecessary fields to be
> NULL,
> > such as
> > digest pointers, I have given a lot thought to this structure. Hopefully it
> covers
> > all vendor's HW symmetric crypto needs and in the same time it well
> squeeze
> > the required HW addresses into 1 cacheline, instead of rte_crypto_op +
> > rte_crypto_sym_op 3 cacheline footprint. Another purpose of the
> structure
> > design
> > is the structure buffer can be taken from stack and can be used to fill all
> > jobs to the PMD HW.
> 
> Which fields you think are not useful and should be set as NULL?
> Digest pointers you are anyways setting in the new structure.
> Your new struct does not support session less as well as security sessions.
> It does not take care of asymmetric crypto.
> So whenever, a vendor need to support all these, we would end up getting
> the rte_crypto_op structure.
> IMO, you only need to make m_src and m_dst as union to a raw
> input/output
> buffers. Everything else will be relevant.
> 

Rte_crypto_op is designed to be allocated from mempool with HW address
info contained so it is possible to deduct IV and AAD physical address from
it. More importantly rte_crypto_op is designed to be taken from heap and 
being freed after dequeue. So they cannot be allocated from stack - for this
reason I think rte_crypot_sym_vec is a better fit for the patch, do you agree? 
(the Proposed change is at above).

> Have you done some profiling with using rte_crypto_op instead of this new
> struct?
> 
Yes, the code are actually upstreamed in VPP
https://gerrit.fd.io/r/c/vpp/+/18036, please try out. If you have a look at the
enqueue/dequeue functions you should see the struggle we had to translate
ops, and creating a second software ring to make sure we only dequeue a 
frame of data. Lucky VPP has space to store mbufs otherwise the perf will
be even wor

Re: [dpdk-dev] [PATCH v4 01/10] eal: introduce macros for getting valuefor bit

[Morten Brørup]

Adding Joyce Kong to this discussion as the rte_bitops maintainer.

> From: Thomas Monjalon [mailto:tho...@monjalon.net]
> Sent: Tuesday, July 7, 2020 2:13 PM
> 
> 07/07/2020 13:38, Parav Pandit:
> > From: Morten Brørup 
> > > From: Parav Pandit
> > > > --- a/lib/librte_eal/include/rte_bitops.h
> > > > +++ b/lib/librte_eal/include/rte_bitops.h
> > > > @@ -17,6 +17,8 @@
> > > >  #include 
> > > >  #include 
> > > >
> > > > +#define RTE_BIT(bit_num)   (1UL << (bit_num))
> > >
> > > Is the return value 32 or 64 bit, or is intended to depend on the
> target
> > > architecture?
> > >
> > It should be 64-bit.
> >
> > > Please be explicit by using UINT32_C(1) or UINT64_C(1) instead of
> 1UL, if you
> > > want a specific size.
> > >
> > Will do UINT64_C(1).
> >
> > > It could be a static inline __attribute__((__pure__)) function
> instead of a macro,
> > > but it's not important for me.
> > >
> > > The macro/function needs a description for the documentation.
> > >
> > In this header file or outside?
> 
> It is asked to add a doxygen comment.
> 
> 
> > > I'm also concerned about the name of the macro being too generic.
> But the
> > > effort of changing all the drivers where it is being used already
> could be too big
> > > if the name changes too.
> > >
> > Right. Currently drivers have generic name as BIT(). Close to 3000
> entries.
> > So doing at RTE_BIT to match other rte_ APIs.
> > Drivers can slowly migrate at their pace to this one.
> >
> > > And the macro/function is new, so shouldn't it - in theory - be
> marked as
> > > experimental?
> >
> > How to mark a macro as experimental?
> 
> A macro cannot be experimental.
> 

OK. If the macro is given a future proof name, I guess it should be accepted.

If we want boundary checks, I suggest a macro like:

#define RTE_BIT64(nr)   \
({  \
typeof(nr) n = nr;  \
RTE_BUILD_BUG_ON((n > 64) || (n < 0));  \
UINT64_C(1) << (n); \
})

Or a function:

__rte_experimental
static __rte_always_inline __attribute__((const)) uint64_t
rte_bit64(const unsigned int nr)
{
RTE_ASSERT(nr < 64);

return UINT64_C(1) << nr;
}

Re: [dpdk-dev] [PATCH] acl: Fix RTE_ACL_RULE_SZ macro definition

[Ananyev, Konstantin]

> From: Levend Sayar 

Could you provide some explanation:
What do you think is wrong with current version and why,
and what your fix does. 

> 
> Signed-off-by: Levend Sayar 
> ---
>  lib/librte_acl/rte_acl.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/lib/librte_acl/rte_acl.h b/lib/librte_acl/rte_acl.h
> index aa22e70c6..d34fdbc0e 100644
> --- a/lib/librte_acl/rte_acl.h
> +++ b/lib/librte_acl/rte_acl.h
> @@ -116,7 +116,7 @@ struct rte_acl_rule_data {
>  RTE_ACL_RULE_DEF(rte_acl_rule,);
> 
>  #define  RTE_ACL_RULE_SZ(fld_num)\
> - (sizeof(struct rte_acl_rule) + sizeof(struct rte_acl_field) * (fld_num))
> + (sizeof(struct rte_acl_rule_data) + sizeof(struct rte_acl_field) * 
> (fld_num))
> 
> 
>  /** Max number of characters in name.*/
> --
> 2.27.0

Re: [dpdk-dev] [PATCH 1/2] mbuf: introduce accurate packet Tx scheduling

[Slava Ovsiienko]

Hi, Olivier

Thanks a lot for the review.

> -Original Message-
> From: Olivier Matz 
> Sent: Tuesday, July 7, 2020 14:51
> To: Slava Ovsiienko 
> Cc: dev@dpdk.org; Matan Azrad ; Raslan
> Darawsheh ; bernard.iremon...@intel.com;
> tho...@mellanox.net
> Subject: Re: [PATCH 1/2] mbuf: introduce accurate packet Tx scheduling
> 
> Hi Slava,
> 
> Few question/comments below.
> 
> On Wed, Jul 01, 2020 at 03:36:26PM +, Viacheslav Ovsiienko wrote:
> > There is the requirement on some networks for precise traffic timing
> > management. The ability to send (and, generally speaking, receive) the
> > packets at the very precisely specified moment of time provides the
> > opportunity to support the connections with Time Division Multiplexing
> > using the contemporary general purpose NIC without involving an
> > auxiliary hardware. For example, the supporting of O-RAN Fronthaul
> > interface is one of the promising features for potentially usage of
> > the precise time management for the egress packets.
> >
> > The main objective of this RFC is to specify the way how applications
> > can provide the moment of time at what the packet transmission must be
> > started and to describe in preliminary the supporting this feature
> > from
> > mlx5 PMD side.
> >
> > The new dynamic timestamp field is proposed, it provides some timing
> > information, the units and time references (initial phase) are not
> > explicitly defined but are maintained always the same for a given port.
> > Some devices allow to query rte_eth_read_clock() that will return the
> > current device timestamp. The dynamic timestamp flag tells whether the
> > field contains actual timestamp value. For the packets being sent this
> > value can be used by PMD to schedule packet sending.
> >
> > After PKT_RX_TIMESTAMP flag and fixed timestamp field deprecation and
> > obsoleting, these dynamic flag and field will be used to manage the
> > timestamps on receiving datapath as well.
> 
> Do you mean the same flag will be used for both Rx and Tx?  I wonder if it's a
> good idea: if you enable the timestamp on Rx, the packets will be flagged
> and it will impact Tx, except if the application explicitly resets the flag 
> in all
> mbufs. Wouldn't it be safer to have an Rx flag and a Tx flag?

A little bit difficult to say ambiguously, I thought about and did not make 
strong decision.
We have the flag sharing for the Rx/Tx metadata and just follow the same 
approach.
OK, I will promote ones to:
RTE_MBUF_DYNFLAG_TX_TIMESTAMP_NAME
RTE_MBUF_DYNFIELD_TX_TIMESTAMP_NAME

And, possible, we should reconsider metadata dynamic flags.
> 
> > When PMD sees the "rte_dynfield_timestamp" set on the packet being
> > sent it tries to synchronize the time of packet appearing on the wire
> > with the specified packet timestamp. It the specified one is in the
> > past it should be ignored, if one is in the distant future it should
> > be capped with some reasonable value (in range of seconds). These
> > specific cases ("too late" and "distant future") can be optionally
> > reported via device xstats to assist applications to detect the
> > time-related problems.
> 
> I think what to do with packets to be send in the "past" could be configurable
> through an ethdev API in the future (drop or send).
Yes, currently there is no complete understanding how to handle packets out of 
the time slot.
In 20.11 we are going to introduce the time-based rte flow API to manage 
out-of-window packets.
 
> 
> > There is no any packet reordering according timestamps is supposed,
> > neither within packet burst, nor between packets, it is an entirely
> > application responsibility to generate packets and its timestamps in
> > desired order. The timestamps can be put only in the first packet in
> > the burst providing the entire burst scheduling.

"can" should be replaced with "might". Current mlx5 implementation
checks each packet in the burst for the timestamp presence.

> 
> This constraint makes sense. At first glance, it looks it is imposed by a PMD 
> or
> hw limitation, but thinking more about it, I think it is the correct behavior 
> to
> have. Packets are ordered within a PMD queue, and the ability to set the
> timestamp for one packet to delay the subsequent ones looks useful.
> 
> Should this behavior be documented somewhere? Maybe in the API
> comment documenting the dynamic flag?

It is documented in mlx5.rst (coming soon), do you think it should be
more common? OK, will update.

> 
> > PMD reports the ability to synchronize packet sending on timestamp
> > with new offload flag:
> >
> > This is palliative and is going to be replaced with new eth_dev API
> > about reporting/managing the supported dynamic flags and its related
> > features. This API would break ABI compatibility and can't be
> > introduced at the moment, so is postponed to 20.11.
> >
> > For testing purposes it is proposed to update testpmd "txonly"
> > forwarding mode routine. With this update testpmd application
> 

[dpdk-dev] [PATCH v2 2/2] app/testpmd: add send scheduling test capability

[Viacheslav Ovsiienko]

This commit adds testpmd capability to provide timestamps on the packets
being sent in the txonly mode. This includes:

 - SEND_ON_TIMESTAMP support
   new device Tx offload capability support added, example:

 testpmd> port config 0 tx_offload send_on_timestamp on

 - set txtimes, registers field and flag, example:

 testpmd> set txtimes 100,0

   This command enables the packet send scheduling on timestamps if
   the first parameter is not zero, generic format:

 testpmd> set txtimes (inter),(intra)

   where:

 inter - is the delay between the bursts in the device clock units.
 intra - is the delay between the packets within the burst specified
 in the device clock units

 As the result the bursts of packet will be transmitted with
 specific delay between the packets within the burst and specific
 delay between the bursts. The rte_eth_get_clock() is supposed to be
 engaged to get the current device clock value and provide
 the reference for the timestamps.

 - show txtimes, displays the timing settings
 - txonly burst time pattern

Signed-off-by: Viacheslav Ovsiienko 
---
 app/test-pmd/cmdline.c  | 59 --
 app/test-pmd/config.c   | 43 
 app/test-pmd/testpmd.c  |  3 ++
 app/test-pmd/testpmd.h  |  3 ++
 app/test-pmd/txonly.c   | 77 +++--
 doc/guides/testpmd_app_ug/testpmd_funcs.rst | 36 +-
 6 files changed, 212 insertions(+), 9 deletions(-)

diff --git a/app/test-pmd/cmdline.c b/app/test-pmd/cmdline.c
index 996a498..57e34a0 100644
--- a/app/test-pmd/cmdline.c
+++ b/app/test-pmd/cmdline.c
@@ -3926,6 +3926,52 @@ struct cmd_set_txsplit_result {
},
 };
 
+/* *** SET TIMES FOR TXONLY PACKETS SCHEDULING ON TIMESTAMPS *** */
+
+struct cmd_set_txtimes_result {
+   cmdline_fixed_string_t cmd_keyword;
+   cmdline_fixed_string_t txtimes;
+   cmdline_fixed_string_t tx_times;
+};
+
+static void
+cmd_set_txtimes_parsed(void *parsed_result,
+  __rte_unused struct cmdline *cl,
+  __rte_unused void *data)
+{
+   struct cmd_set_txtimes_result *res;
+   unsigned int tx_times[2] = {0, 0};
+   unsigned int n_times;
+
+   res = parsed_result;
+   n_times = parse_item_list(res->tx_times, "tx times",
+ 2, tx_times, 0);
+   if (n_times == 2)
+   set_tx_pkt_times(tx_times);
+}
+
+cmdline_parse_token_string_t cmd_set_txtimes_keyword =
+   TOKEN_STRING_INITIALIZER(struct cmd_set_txtimes_result,
+cmd_keyword, "set");
+cmdline_parse_token_string_t cmd_set_txtimes_name =
+   TOKEN_STRING_INITIALIZER(struct cmd_set_txtimes_result,
+txtimes, "txtimes");
+cmdline_parse_token_string_t cmd_set_txtimes_value =
+   TOKEN_STRING_INITIALIZER(struct cmd_set_txtimes_result,
+tx_times, NULL);
+
+cmdline_parse_inst_t cmd_set_txtimes = {
+   .f = cmd_set_txtimes_parsed,
+   .data = NULL,
+   .help_str = "set txtimes ,",
+   .tokens = {
+   (void *)&cmd_set_txtimes_keyword,
+   (void *)&cmd_set_txtimes_name,
+   (void *)&cmd_set_txtimes_value,
+   NULL,
+   },
+};
+
 /* *** ADD/REMOVE ALL VLAN IDENTIFIERS TO/FROM A PORT VLAN RX FILTER *** */
 struct cmd_rx_vlan_filter_all_result {
cmdline_fixed_string_t rx_vlan;
@@ -7414,6 +7460,8 @@ static void cmd_showcfg_parsed(void *parsed_result,
pkt_fwd_config_display(&cur_fwd_config);
else if (!strcmp(res->what, "txpkts"))
show_tx_pkt_segments();
+   else if (!strcmp(res->what, "txtimes"))
+   show_tx_pkt_times();
 }
 
 cmdline_parse_token_string_t cmd_showcfg_show =
@@ -7422,12 +7470,12 @@ static void cmd_showcfg_parsed(void *parsed_result,
TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, cfg, "config");
 cmdline_parse_token_string_t cmd_showcfg_what =
TOKEN_STRING_INITIALIZER(struct cmd_showcfg_result, what,
-"rxtx#cores#fwd#txpkts");
+"rxtx#cores#fwd#txpkts#txtimes");
 
 cmdline_parse_inst_t cmd_showcfg = {
.f = cmd_showcfg_parsed,
.data = NULL,
-   .help_str = "show config rxtx|cores|fwd|txpkts",
+   .help_str = "show config rxtx|cores|fwd|txpkts|txtimes",
.tokens = {
(void *)&cmd_showcfg_show,
(void *)&cmd_showcfg_port,
@@ -18662,7 +18710,8 @@ struct cmd_config_per_port_tx_offload_result {
  "sctp_cksum#tcp_tso#udp_tso#outer_ipv4_cksum#"
  "qinq_insert#vxlan_tnl_tso#gre_tnl_tso#"
  "ipip_tnl_tso#geneve_tnl_tso#macsec_insert#"
- "mt_lockfree#multi_segs#mbuf_fast_free#security");
+

[dpdk-dev] [PATCH v2 1/2] mbuf: introduce accurate packet Tx scheduling

[Viacheslav Ovsiienko]

There is the requirement on some networks for precise traffic timing
management. The ability to send (and, generally speaking, receive)
the packets at the very precisely specified moment of time provides
the opportunity to support the connections with Time Division
Multiplexing using the contemporary general purpose NIC without involving
an auxiliary hardware. For example, the supporting of O-RAN Fronthaul
interface is one of the promising features for potentially usage of the
precise time management for the egress packets.

The main objective of this RFC is to specify the way how applications
can provide the moment of time at what the packet transmission must be
started and to describe in preliminary the supporting this feature from
mlx5 PMD side.

The new dynamic timestamp field is proposed, it provides some timing
information, the units and time references (initial phase) are not
explicitly defined but are maintained always the same for a given port.
Some devices allow to query rte_eth_read_clock() that will return
the current device timestamp. The dynamic timestamp flag tells whether
the field contains actual timestamp value. For the packets being sent
this value can be used by PMD to schedule packet sending.

After PKT_RX_TIMESTAMP flag and fixed timestamp field deprecation
and obsoleting, these dynamic flag and field will be used to manage
the timestamps on receiving datapath as well.

When PMD sees the "rte_dynfield_timestamp" set on the packet being sent
it tries to synchronize the time of packet appearing on the wire with
the specified packet timestamp. If the specified one is in the past it
should be ignored, if one is in the distant future it should be capped
with some reasonable value (in range of seconds). These specific cases
("too late" and "distant future") can be optionally reported via
device xstats to assist applications to detect the time-related
problems.

There is no any packet reordering according timestamps is supposed,
neither within packet burst, nor between packets, it is an entirely
application responsibility to generate packets and its timestamps
in desired order. The timestamps can be put only in the first packet
in the burst providing the entire burst scheduling.

PMD reports the ability to synchronize packet sending on timestamp
with new offload flag:

This is palliative and is going to be replaced with new eth_dev API
about reporting/managing the supported dynamic flags and its related
features. This API would break ABI compatibility and can't be introduced
at the moment, so is postponed to 20.11.

For testing purposes it is proposed to update testpmd "txonly"
forwarding mode routine. With this update testpmd application generates
the packets and sets the dynamic timestamps according to specified time
pattern if it sees the "rte_dynfield_timestamp" is registered.

The new testpmd command is proposed to configure sending pattern:

set tx_times ,

 - the delay between the packets within the burst
  specified in the device clock units. The number
  of packets in the burst is defined by txburst parameter

 - the delay between the bursts in the device clock units

As the result the bursts of packet will be transmitted with specific
delays between the packets within the burst and specific delay between
the bursts. The rte_eth_get_clock is supposed to be engaged to get the
current device clock value and provide the reference for the timestamps.

Signed-off-by: Viacheslav Ovsiienko 
---
 lib/librte_ethdev/rte_ethdev.c |  1 +
 lib/librte_ethdev/rte_ethdev.h |  4 
 lib/librte_mbuf/rte_mbuf_dyn.h | 32 
 3 files changed, 37 insertions(+)

diff --git a/lib/librte_ethdev/rte_ethdev.c b/lib/librte_ethdev/rte_ethdev.c
index 8e10a6f..02157d5 100644
--- a/lib/librte_ethdev/rte_ethdev.c
+++ b/lib/librte_ethdev/rte_ethdev.c
@@ -162,6 +162,7 @@ struct rte_eth_xstats_name_off {
RTE_TX_OFFLOAD_BIT2STR(UDP_TNL_TSO),
RTE_TX_OFFLOAD_BIT2STR(IP_TNL_TSO),
RTE_TX_OFFLOAD_BIT2STR(OUTER_UDP_CKSUM),
+   RTE_TX_OFFLOAD_BIT2STR(SEND_ON_TIMESTAMP),
 };
 
 #undef RTE_TX_OFFLOAD_BIT2STR
diff --git a/lib/librte_ethdev/rte_ethdev.h b/lib/librte_ethdev/rte_ethdev.h
index a49242b..6f6454c 100644
--- a/lib/librte_ethdev/rte_ethdev.h
+++ b/lib/librte_ethdev/rte_ethdev.h
@@ -1178,6 +1178,10 @@ struct rte_eth_conf {
 /** Device supports outer UDP checksum */
 #define DEV_TX_OFFLOAD_OUTER_UDP_CKSUM  0x0010
 
+/** Device supports send on timestamp */
+#define DEV_TX_OFFLOAD_SEND_ON_TIMESTAMP 0x0020
+
+
 #define RTE_ETH_DEV_CAPA_RUNTIME_RX_QUEUE_SETUP 0x0001
 /**< Device supports Rx queue setup after device started*/
 #define RTE_ETH_DEV_CAPA_RUNTIME_TX_QUEUE_SETUP 0x0002
diff --git a/lib/librte_mbuf/rte_mbuf_dyn.h b/lib/librte_mbuf/rte_mbuf_dyn.h
index 96c3631..834acdd 100644
--- a/lib/librte_mbuf/rte_mbuf_dyn.h
+++ b/lib/librte_mbuf/rte_mbuf_dyn.h
@@ -250,4 +250,36 @@ int rte_mbuf_dynflag_lookup(const char *name,
 #d