[v2,14/62] common/sfc_efx/base: add action set spec init/fini APIs
Checks
Commit Message
From: Ivan Malov <ivan.malov@oktetlabs.ru>
The engine is only able to carry out chosen actions on matching packets in
a strict order. No MCDI exists to identify supported actions and the order.
Still, the definition of the latter is available from the FW documentation.
The general idea is to define an action specification structure and supply
a client driver with APIs for adding actions individually, order-dependent.
A client driver is supposed to invoke an API on every action passed by the
application, and if an out-of-order action follows, the API will reject it.
Add an action set specification stub and supply initialise / finalise APIs.
Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@xilinx.com>
---
drivers/common/sfc_efx/base/efx.h | 20 ++++++++++
drivers/common/sfc_efx/base/efx_impl.h | 3 ++
drivers/common/sfc_efx/base/efx_mae.c | 39 +++++++++++++++++++
.../sfc_efx/rte_common_sfc_efx_version.map | 3 ++
4 files changed, 65 insertions(+)
Comments
Hi,
Sent my report to v1 the first time by mistake.
> -----Original Message-----
> From: dev <dev-bounces@dpdk.org> On Behalf Of Andrew Rybchenko
> Sent: Tuesday, October 20, 2020 12:13 PM
> To: dev@dpdk.org
> Cc: Ivan Malov <ivan.malov@oktetlabs.ru>
> Subject: [dpdk-dev] [PATCH v2 14/62] common/sfc_efx/base: add action set
> spec init/fini APIs
>
> From: Ivan Malov <ivan.malov@oktetlabs.ru>
>
> The engine is only able to carry out chosen actions on matching packets in a
> strict order. No MCDI exists to identify supported actions and the order.
> Still, the definition of the latter is available from the FW documentation.
>
> The general idea is to define an action specification structure and supply a
> client driver with APIs for adding actions individually, order-dependent.
> A client driver is supposed to invoke an API on every action passed by the
> application, and if an out-of-order action follows, the API will reject it.
>
> Add an action set specification stub and supply initialise / finalise APIs.
>
> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
> ---
This patch is causing the following build failure on CentOS 7 with clang 3.4.2:
"""
In file included from drivers/common/sfc_efx/base/efx_crc32.c:8:
drivers/common/sfc_efx/base/efx_impl.h:1703:3: error: redefinition of typedef 'efx_mae_actions_t' is a C11 feature [-Werror,-Wtypedef-redefinition] } efx_mae_actions_t;
drivers/common/sfc_efx/base/efx.h:4101:34: note: previous definition is here typedef struct efx_mae_actions_s efx_mae_actions_t;
"""
Regards,
Ali
On 10/27/2020 9:13 AM, Ali Alnubani wrote:
> Hi,
> Sent my report to v1 the first time by mistake.
>
>> -----Original Message-----
>> From: dev <dev-bounces@dpdk.org> On Behalf Of Andrew Rybchenko
>> Sent: Tuesday, October 20, 2020 12:13 PM
>> To: dev@dpdk.org
>> Cc: Ivan Malov <ivan.malov@oktetlabs.ru>
>> Subject: [dpdk-dev] [PATCH v2 14/62] common/sfc_efx/base: add action set
>> spec init/fini APIs
>>
>> From: Ivan Malov <ivan.malov@oktetlabs.ru>
>>
>> The engine is only able to carry out chosen actions on matching packets in a
>> strict order. No MCDI exists to identify supported actions and the order.
>> Still, the definition of the latter is available from the FW documentation.
>>
>> The general idea is to define an action specification structure and supply a
>> client driver with APIs for adding actions individually, order-dependent.
>> A client driver is supposed to invoke an API on every action passed by the
>> application, and if an out-of-order action follows, the API will reject it.
>>
>> Add an action set specification stub and supply initialise / finalise APIs.
>>
>> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
>> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
>> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
>> ---
>
> This patch is causing the following build failure on CentOS 7 with clang 3.4.2:
>
> """
> In file included from drivers/common/sfc_efx/base/efx_crc32.c:8:
> drivers/common/sfc_efx/base/efx_impl.h:1703:3: error: redefinition of typedef 'efx_mae_actions_t' is a C11 feature [-Werror,-Wtypedef-redefinition] } efx_mae_actions_t;
> drivers/common/sfc_efx/base/efx.h:4101:34: note: previous definition is here typedef struct efx_mae_actions_s efx_mae_actions_t;
> """
>
Hi Andrew,
Are you working on this?
On 10/27/20 2:39 PM, Ferruh Yigit wrote:
> On 10/27/2020 9:13 AM, Ali Alnubani wrote:
>> Hi,
>> Sent my report to v1 the first time by mistake.
>>
>>> -----Original Message-----
>>> From: dev <dev-bounces@dpdk.org> On Behalf Of Andrew Rybchenko
>>> Sent: Tuesday, October 20, 2020 12:13 PM
>>> To: dev@dpdk.org
>>> Cc: Ivan Malov <ivan.malov@oktetlabs.ru>
>>> Subject: [dpdk-dev] [PATCH v2 14/62] common/sfc_efx/base: add action
>>> set
>>> spec init/fini APIs
>>>
>>> From: Ivan Malov <ivan.malov@oktetlabs.ru>
>>>
>>> The engine is only able to carry out chosen actions on matching
>>> packets in a
>>> strict order. No MCDI exists to identify supported actions and the
>>> order.
>>> Still, the definition of the latter is available from the FW
>>> documentation.
>>>
>>> The general idea is to define an action specification structure and
>>> supply a
>>> client driver with APIs for adding actions individually,
>>> order-dependent.
>>> A client driver is supposed to invoke an API on every action passed
>>> by the
>>> application, and if an out-of-order action follows, the API will
>>> reject it.
>>>
>>> Add an action set specification stub and supply initialise /
>>> finalise APIs.
>>>
>>> Signed-off-by: Ivan Malov <ivan.malov@oktetlabs.ru>
>>> Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
>>> Reviewed-by: Andy Moreton <amoreton@xilinx.com>
>>> ---
>>
>> This patch is causing the following build failure on CentOS 7 with
>> clang 3.4.2:
>>
>> """
>> In file included from drivers/common/sfc_efx/base/efx_crc32.c:8:
>> drivers/common/sfc_efx/base/efx_impl.h:1703:3: error: redefinition of
>> typedef 'efx_mae_actions_t' is a C11 feature
>> [-Werror,-Wtypedef-redefinition] } efx_mae_actions_t;
>> drivers/common/sfc_efx/base/efx.h:4101:34: note: previous definition
>> is here typedef struct efx_mae_actions_s efx_mae_actions_t;
>> """
>>
>
> Hi Andrew,
>
> Are you working on this?
Hi Ferruh,
Yes, I have a patch. I will send it shortly.
Andrew.
@@ -4098,6 +4098,26 @@ efx_mae_match_spec_is_valid(
__in efx_nic_t *enp,
__in const efx_mae_match_spec_t *spec);
+typedef struct efx_mae_actions_s efx_mae_actions_t;
+
+LIBEFX_API
+extern __checkReturn efx_rc_t
+efx_mae_action_set_spec_init(
+ __in efx_nic_t *enp,
+ __out efx_mae_actions_t **specp);
+
+LIBEFX_API
+extern void
+efx_mae_action_set_spec_fini(
+ __in efx_nic_t *enp,
+ __in efx_mae_actions_t *spec);
+
+LIBEFX_API
+extern __checkReturn boolean_t
+efx_mae_action_set_specs_equal(
+ __in const efx_mae_actions_t *left,
+ __in const efx_mae_actions_t *right);
+
/*
* Conduct a comparison to check whether two match specifications
* of equal rule type (action / outer) and priority would map to
@@ -1699,6 +1699,9 @@ struct efx_mae_match_spec_s {
} emms_mask_value_pairs;
};
+typedef struct efx_mae_actions_s {
+} efx_mae_actions_t;
+
#endif /* EFSYS_OPT_MAE */
#ifdef __cplusplus
@@ -434,6 +434,45 @@ efx_mae_match_spec_is_valid(
return (is_valid);
}
+ __checkReturn efx_rc_t
+efx_mae_action_set_spec_init(
+ __in efx_nic_t *enp,
+ __out efx_mae_actions_t **specp)
+{
+ efx_mae_actions_t *spec;
+ efx_rc_t rc;
+
+ EFSYS_KMEM_ALLOC(enp->en_esip, sizeof (*spec), spec);
+ if (spec == NULL) {
+ rc = ENOMEM;
+ goto fail1;
+ }
+
+ *specp = spec;
+
+ return (0);
+
+fail1:
+ EFSYS_PROBE1(fail1, efx_rc_t, rc);
+ return (rc);
+}
+
+ void
+efx_mae_action_set_spec_fini(
+ __in efx_nic_t *enp,
+ __in efx_mae_actions_t *spec)
+{
+ EFSYS_KMEM_FREE(enp->en_esip, sizeof (*spec), spec);
+}
+
+ __checkReturn boolean_t
+efx_mae_action_set_specs_equal(
+ __in const efx_mae_actions_t *left,
+ __in const efx_mae_actions_t *right)
+{
+ return ((memcmp(left, right, sizeof (*left)) == 0) ? B_TRUE : B_FALSE);
+}
+
__checkReturn efx_rc_t
efx_mae_match_specs_class_cmp(
__in efx_nic_t *enp,
@@ -85,6 +85,9 @@ INTERNAL {
efx_mac_stats_upload;
efx_mac_up;
+ efx_mae_action_set_spec_fini;
+ efx_mae_action_set_spec_init;
+ efx_mae_action_set_specs_equal;
efx_mae_fini;
efx_mae_get_limits;
efx_mae_init;