* [PATCH net-next 1/6] virtchnl: Add support for new VLAN capabilities
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 2/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 negotiation Tony Nguyen
` (5 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
Currently VIRTCHNL only allows for VLAN filtering and offloads to happen
on a single 802.1Q VLAN. Add support to filter and offload on inner,
outer, and/or inner + outer VLANs.
This is done by introducing the new capability
VIRTCHNL_VF_OFFLOAD_VLAN_V2. The flow to negotiate this new capability
is shown below.
1. VF - sets the VIRTCHNL_VF_OFFLOAD_VLAN_V2 bit in the
virtchnl_vf_resource.vf_caps_flags during the
VIRTCHNL_OP_GET_VF_RESOURCES request message. The VF should also set
the VIRTCHNL_VF_OFFLOAD_VLAN bit in case the PF driver doesn't support
the new capability.
2. PF - sets the VLAN capability bit it supports in the
VIRTCHNL_OP_GET_VF_RESOURCES response message. This will either be
VIRTCHNL_VF_OFFLOAD_VLAN_V2, VIRTCHNL_VF_OFFLOAD_VLAN, or none.
3. VF - If the VIRTCHNL_VF_OFFLOAD_VLAN_V2 capability was ACK'd by the
PF, then the VF needs to request the VLAN capabilities of the
PF/Device by issuing a VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS request.
If the VIRTCHNL_VF_OFFLOAD_VLAN capability was ACK'd then the VF
knows only single 802.1Q VLAN filtering/offloads are supported. If no
VLAN capability is ACK'd then the PF/Device doesn't support hardware
VLAN filtering/offloads for this VF.
4. PF - Populates the virtchnl_vlan_caps structure based on what it
allows/supports for that VF and sends that response via
VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS.
After VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS is successfully negotiated
the VF driver needs to interpret the capabilities supported by the
underlying PF/Device. The VF will be allowed to filter/offload the
inner 802.1Q, outer (various ethertype), inner 802.1Q + outer
(various ethertypes), or none based on which fields are set.
The VF will also need to interpret where the VLAN tag should be inserted
and/or stripped based on the negotiated capabilities.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
include/linux/avf/virtchnl.h | 377 +++++++++++++++++++++++++++++++++++
1 file changed, 377 insertions(+)
diff --git a/include/linux/avf/virtchnl.h b/include/linux/avf/virtchnl.h
index b30a1bc74fc7..2ce27e8e4f19 100644
--- a/include/linux/avf/virtchnl.h
+++ b/include/linux/avf/virtchnl.h
@@ -141,6 +141,13 @@ enum virtchnl_ops {
VIRTCHNL_OP_DEL_RSS_CFG = 46,
VIRTCHNL_OP_ADD_FDIR_FILTER = 47,
VIRTCHNL_OP_DEL_FDIR_FILTER = 48,
+ VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS = 51,
+ VIRTCHNL_OP_ADD_VLAN_V2 = 52,
+ VIRTCHNL_OP_DEL_VLAN_V2 = 53,
+ VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 = 54,
+ VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2 = 55,
+ VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2 = 56,
+ VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2 = 57,
VIRTCHNL_OP_MAX,
};
@@ -246,6 +253,7 @@ VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vsi_resource);
#define VIRTCHNL_VF_OFFLOAD_REQ_QUEUES BIT(6)
/* used to negotiate communicating link speeds in Mbps */
#define VIRTCHNL_VF_CAP_ADV_LINK_SPEED BIT(7)
+#define VIRTCHNL_VF_OFFLOAD_VLAN_V2 BIT(15)
#define VIRTCHNL_VF_OFFLOAD_VLAN BIT(16)
#define VIRTCHNL_VF_OFFLOAD_RX_POLLING BIT(17)
#define VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 BIT(18)
@@ -475,6 +483,351 @@ struct virtchnl_vlan_filter_list {
VIRTCHNL_CHECK_STRUCT_LEN(6, virtchnl_vlan_filter_list);
+/* This enum is used for all of the VIRTCHNL_VF_OFFLOAD_VLAN_V2_CAPS related
+ * structures and opcodes.
+ *
+ * VIRTCHNL_VLAN_UNSUPPORTED - This field is not supported and if a VF driver
+ * populates it the PF should return VIRTCHNL_STATUS_ERR_NOT_SUPPORTED.
+ *
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 - This field supports 0x8100 ethertype.
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 - This field supports 0x88A8 ethertype.
+ * VIRTCHNL_VLAN_ETHERTYPE_9100 - This field supports 0x9100 ethertype.
+ *
+ * VIRTCHNL_VLAN_ETHERTYPE_AND - Used when multiple ethertypes can be supported
+ * by the PF concurrently. For example, if the PF can support
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 AND VIRTCHNL_VLAN_ETHERTYPE_88A8 filters it
+ * would OR the following bits:
+ *
+ * VIRTHCNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_AND;
+ *
+ * The VF would interpret this as VLAN filtering can be supported on both 0x8100
+ * and 0x88A8 VLAN ethertypes.
+ *
+ * VIRTCHNL_ETHERTYPE_XOR - Used when only a single ethertype can be supported
+ * by the PF concurrently. For example if the PF can support
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 XOR VIRTCHNL_VLAN_ETHERTYPE_88A8 stripping
+ * offload it would OR the following bits:
+ *
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_XOR;
+ *
+ * The VF would interpret this as VLAN stripping can be supported on either
+ * 0x8100 or 0x88a8 VLAN ethertypes. So when requesting VLAN stripping via
+ * VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 the specified ethertype will override
+ * the previously set value.
+ *
+ * VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1 - Used to tell the VF to insert and/or
+ * strip the VLAN tag using the L2TAG1 field of the Tx/Rx descriptors.
+ *
+ * VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2 - Used to tell the VF to insert hardware
+ * offloaded VLAN tags using the L2TAG2 field of the Tx descriptor.
+ *
+ * VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2 - Used to tell the VF to strip hardware
+ * offloaded VLAN tags using the L2TAG2_2 field of the Rx descriptor.
+ *
+ * VIRTCHNL_VLAN_PRIO - This field supports VLAN priority bits. This is used for
+ * VLAN filtering if the underlying PF supports it.
+ *
+ * VIRTCHNL_VLAN_TOGGLE_ALLOWED - This field is used to say whether a
+ * certain VLAN capability can be toggled. For example if the underlying PF/CP
+ * allows the VF to toggle VLAN filtering, stripping, and/or insertion it should
+ * set this bit along with the supported ethertypes.
+ */
+enum virtchnl_vlan_support {
+ VIRTCHNL_VLAN_UNSUPPORTED = 0,
+ VIRTCHNL_VLAN_ETHERTYPE_8100 = BIT(0),
+ VIRTCHNL_VLAN_ETHERTYPE_88A8 = BIT(1),
+ VIRTCHNL_VLAN_ETHERTYPE_9100 = BIT(2),
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1 = BIT(8),
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2 = BIT(9),
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2_2 = BIT(10),
+ VIRTCHNL_VLAN_PRIO = BIT(24),
+ VIRTCHNL_VLAN_FILTER_MASK = BIT(28),
+ VIRTCHNL_VLAN_ETHERTYPE_AND = BIT(29),
+ VIRTCHNL_VLAN_ETHERTYPE_XOR = BIT(30),
+ VIRTCHNL_VLAN_TOGGLE = BIT(31),
+};
+
+/* This structure is used as part of the VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS
+ * for filtering, insertion, and stripping capabilities.
+ *
+ * If only outer capabilities are supported (for filtering, insertion, and/or
+ * stripping) then this refers to the outer most or single VLAN from the VF's
+ * perspective.
+ *
+ * If only inner capabilities are supported (for filtering, insertion, and/or
+ * stripping) then this refers to the outer most or single VLAN from the VF's
+ * perspective. Functionally this is the same as if only outer capabilities are
+ * supported. The VF driver is just forced to use the inner fields when
+ * adding/deleting filters and enabling/disabling offloads (if supported).
+ *
+ * If both outer and inner capabilities are supported (for filtering, insertion,
+ * and/or stripping) then outer refers to the outer most or single VLAN and
+ * inner refers to the second VLAN, if it exists, in the packet.
+ *
+ * There is no support for tunneled VLAN offloads, so outer or inner are never
+ * referring to a tunneled packet from the VF's perspective.
+ */
+struct virtchnl_vlan_supported_caps {
+ u32 outer;
+ u32 inner;
+};
+
+/* The PF populates these fields based on the supported VLAN filtering. If a
+ * field is VIRTCHNL_VLAN_UNSUPPORTED then it's not supported and the PF will
+ * reject any VIRTCHNL_OP_ADD_VLAN_V2 or VIRTCHNL_OP_DEL_VLAN_V2 messages using
+ * the unsupported fields.
+ *
+ * Also, a VF is only allowed to toggle its VLAN filtering setting if the
+ * VIRTCHNL_VLAN_TOGGLE bit is set.
+ *
+ * The ethertype(s) specified in the ethertype_init field are the ethertypes
+ * enabled for VLAN filtering. VLAN filtering in this case refers to the outer
+ * most VLAN from the VF's perspective. If both inner and outer filtering are
+ * allowed then ethertype_init only refers to the outer most VLAN as only
+ * VLAN ethertype supported for inner VLAN filtering is
+ * VIRTCHNL_VLAN_ETHERTYPE_8100. By default, inner VLAN filtering is disabled
+ * when both inner and outer filtering are allowed.
+ *
+ * The max_filters field tells the VF how many VLAN filters it's allowed to have
+ * at any one time. If it exceeds this amount and tries to add another filter,
+ * then the request will be rejected by the PF. To prevent failures, the VF
+ * should keep track of how many VLAN filters it has added and not attempt to
+ * add more than max_filters.
+ */
+struct virtchnl_vlan_filtering_caps {
+ struct virtchnl_vlan_supported_caps filtering_support;
+ u32 ethertype_init;
+ u16 max_filters;
+ u8 pad[2];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vlan_filtering_caps);
+
+/* This enum is used for the virtchnl_vlan_offload_caps structure to specify
+ * if the PF supports a different ethertype for stripping and insertion.
+ *
+ * VIRTCHNL_ETHERTYPE_STRIPPING_MATCHES_INSERTION - The ethertype(s) specified
+ * for stripping affect the ethertype(s) specified for insertion and visa versa
+ * as well. If the VF tries to configure VLAN stripping via
+ * VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 with VIRTCHNL_VLAN_ETHERTYPE_8100 then
+ * that will be the ethertype for both stripping and insertion.
+ *
+ * VIRTCHNL_ETHERTYPE_MATCH_NOT_REQUIRED - The ethertype(s) specified for
+ * stripping do not affect the ethertype(s) specified for insertion and visa
+ * versa.
+ */
+enum virtchnl_vlan_ethertype_match {
+ VIRTCHNL_ETHERTYPE_STRIPPING_MATCHES_INSERTION = 0,
+ VIRTCHNL_ETHERTYPE_MATCH_NOT_REQUIRED = 1,
+};
+
+/* The PF populates these fields based on the supported VLAN offloads. If a
+ * field is VIRTCHNL_VLAN_UNSUPPORTED then it's not supported and the PF will
+ * reject any VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 or
+ * VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2 messages using the unsupported fields.
+ *
+ * Also, a VF is only allowed to toggle its VLAN offload setting if the
+ * VIRTCHNL_VLAN_TOGGLE_ALLOWED bit is set.
+ *
+ * The VF driver needs to be aware of how the tags are stripped by hardware and
+ * inserted by the VF driver based on the level of offload support. The PF will
+ * populate these fields based on where the VLAN tags are expected to be
+ * offloaded via the VIRTHCNL_VLAN_TAG_LOCATION_* bits. The VF will need to
+ * interpret these fields. See the definition of the
+ * VIRTCHNL_VLAN_TAG_LOCATION_* bits above the virtchnl_vlan_support
+ * enumeration.
+ */
+struct virtchnl_vlan_offload_caps {
+ struct virtchnl_vlan_supported_caps stripping_support;
+ struct virtchnl_vlan_supported_caps insertion_support;
+ u32 ethertype_init;
+ u8 ethertype_match;
+ u8 pad[3];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(24, virtchnl_vlan_offload_caps);
+
+/* VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS
+ * VF sends this message to determine its VLAN capabilities.
+ *
+ * PF will mark which capabilities it supports based on hardware support and
+ * current configuration. For example, if a port VLAN is configured the PF will
+ * not allow outer VLAN filtering, stripping, or insertion to be configured so
+ * it will block these features from the VF.
+ *
+ * The VF will need to cross reference its capabilities with the PFs
+ * capabilities in the response message from the PF to determine the VLAN
+ * support.
+ */
+struct virtchnl_vlan_caps {
+ struct virtchnl_vlan_filtering_caps filtering;
+ struct virtchnl_vlan_offload_caps offloads;
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_vlan_caps);
+
+struct virtchnl_vlan {
+ u16 tci; /* tci[15:13] = PCP and tci[11:0] = VID */
+ u16 tci_mask; /* only valid if VIRTCHNL_VLAN_FILTER_MASK set in
+ * filtering caps
+ */
+ u16 tpid; /* 0x8100, 0x88a8, etc. and only type(s) set in
+ * filtering caps. Note that tpid here does not refer to
+ * VIRTCHNL_VLAN_ETHERTYPE_*, but it refers to the
+ * actual 2-byte VLAN TPID
+ */
+ u8 pad[2];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(8, virtchnl_vlan);
+
+struct virtchnl_vlan_filter {
+ struct virtchnl_vlan inner;
+ struct virtchnl_vlan outer;
+ u8 pad[16];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(32, virtchnl_vlan_filter);
+
+/* VIRTCHNL_OP_ADD_VLAN_V2
+ * VIRTCHNL_OP_DEL_VLAN_V2
+ *
+ * VF sends these messages to add/del one or more VLAN tag filters for Rx
+ * traffic.
+ *
+ * The PF attempts to add the filters and returns status.
+ *
+ * The VF should only ever attempt to add/del virtchnl_vlan_filter(s) using the
+ * supported fields negotiated via VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS.
+ */
+struct virtchnl_vlan_filter_list_v2 {
+ u16 vport_id;
+ u16 num_elements;
+ u8 pad[4];
+ struct virtchnl_vlan_filter filters[1];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(40, virtchnl_vlan_filter_list_v2);
+
+/* VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2
+ * VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2
+ * VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2
+ * VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2
+ *
+ * VF sends this message to enable or disable VLAN stripping or insertion. It
+ * also needs to specify an ethertype. The VF knows which VLAN ethertypes are
+ * allowed and whether or not it's allowed to enable/disable the specific
+ * offload via the VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS message. The VF needs to
+ * parse the virtchnl_vlan_caps.offloads fields to determine which offload
+ * messages are allowed.
+ *
+ * For example, if the PF populates the virtchnl_vlan_caps.offloads in the
+ * following manner the VF will be allowed to enable and/or disable 0x8100 inner
+ * VLAN insertion and/or stripping via the opcodes listed above. Inner in this
+ * case means the outer most or single VLAN from the VF's perspective. This is
+ * because no outer offloads are supported. See the comments above the
+ * virtchnl_vlan_supported_caps structure for more details.
+ *
+ * virtchnl_vlan_caps.offloads.stripping_support.inner =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100;
+ *
+ * virtchnl_vlan_caps.offloads.insertion_support.inner =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100;
+ *
+ * In order to enable inner (again note that in this case inner is the outer
+ * most or single VLAN from the VF's perspective) VLAN stripping for 0x8100
+ * VLANs, the VF would populate the virtchnl_vlan_setting structure in the
+ * following manner and send the VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 message.
+ *
+ * virtchnl_vlan_setting.inner_ethertype_setting =
+ * VIRTCHNL_VLAN_ETHERTYPE_8100;
+ *
+ * virtchnl_vlan_setting.vport_id = vport_id or vsi_id assigned to the VF on
+ * initialization.
+ *
+ * The reason that VLAN TPID(s) are not being used for the
+ * outer_ethertype_setting and inner_ethertype_setting fields is because it's
+ * possible a device could support VLAN insertion and/or stripping offload on
+ * multiple ethertypes concurrently, so this method allows a VF to request
+ * multiple ethertypes in one message using the virtchnl_vlan_support
+ * enumeration.
+ *
+ * For example, if the PF populates the virtchnl_vlan_caps.offloads in the
+ * following manner the VF will be allowed to enable 0x8100 and 0x88a8 outer
+ * VLAN insertion and stripping simultaneously. The
+ * virtchnl_vlan_caps.offloads.ethertype_match field will also have to be
+ * populated based on what the PF can support.
+ *
+ * virtchnl_vlan_caps.offloads.stripping_support.outer =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_AND;
+ *
+ * virtchnl_vlan_caps.offloads.insertion_support.outer =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_AND;
+ *
+ * In order to enable outer VLAN stripping for 0x8100 and 0x88a8 VLANs, the VF
+ * would populate the virthcnl_vlan_offload_structure in the following manner
+ * and send the VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2 message.
+ *
+ * virtchnl_vlan_setting.outer_ethertype_setting =
+ * VIRTHCNL_VLAN_ETHERTYPE_8100 |
+ * VIRTHCNL_VLAN_ETHERTYPE_88A8;
+ *
+ * virtchnl_vlan_setting.vport_id = vport_id or vsi_id assigned to the VF on
+ * initialization.
+ *
+ * There is also the case where a PF and the underlying hardware can support
+ * VLAN offloads on multiple ethertypes, but not concurrently. For example, if
+ * the PF populates the virtchnl_vlan_caps.offloads in the following manner the
+ * VF will be allowed to enable and/or disable 0x8100 XOR 0x88a8 outer VLAN
+ * offloads. The ethertypes must match for stripping and insertion.
+ *
+ * virtchnl_vlan_caps.offloads.stripping_support.outer =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_XOR;
+ *
+ * virtchnl_vlan_caps.offloads.insertion_support.outer =
+ * VIRTCHNL_VLAN_TOGGLE |
+ * VIRTCHNL_VLAN_ETHERTYPE_8100 |
+ * VIRTCHNL_VLAN_ETHERTYPE_88A8 |
+ * VIRTCHNL_VLAN_ETHERTYPE_XOR;
+ *
+ * virtchnl_vlan_caps.offloads.ethertype_match =
+ * VIRTCHNL_ETHERTYPE_STRIPPING_MATCHES_INSERTION;
+ *
+ * In order to enable outer VLAN stripping for 0x88a8 VLANs, the VF would
+ * populate the virtchnl_vlan_setting structure in the following manner and send
+ * the VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2. Also, this will change the
+ * ethertype for VLAN insertion if it's enabled. So, for completeness, a
+ * VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2 with the same ethertype should be sent.
+ *
+ * virtchnl_vlan_setting.outer_ethertype_setting = VIRTHCNL_VLAN_ETHERTYPE_88A8;
+ *
+ * virtchnl_vlan_setting.vport_id = vport_id or vsi_id assigned to the VF on
+ * initialization.
+ */
+struct virtchnl_vlan_setting {
+ u32 outer_ethertype_setting;
+ u32 inner_ethertype_setting;
+ u16 vport_id;
+ u8 pad[6];
+};
+
+VIRTCHNL_CHECK_STRUCT_LEN(16, virtchnl_vlan_setting);
+
/* VIRTCHNL_OP_CONFIG_PROMISCUOUS_MODE
* VF sends VSI id and flags.
* PF returns status code in retval.
@@ -1156,6 +1509,30 @@ virtchnl_vc_validate_vf_msg(struct virtchnl_version_info *ver, u32 v_opcode,
case VIRTCHNL_OP_DEL_FDIR_FILTER:
valid_len = sizeof(struct virtchnl_fdir_del);
break;
+ case VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS:
+ break;
+ case VIRTCHNL_OP_ADD_VLAN_V2:
+ case VIRTCHNL_OP_DEL_VLAN_V2:
+ valid_len = sizeof(struct virtchnl_vlan_filter_list_v2);
+ if (msglen >= valid_len) {
+ struct virtchnl_vlan_filter_list_v2 *vfl =
+ (struct virtchnl_vlan_filter_list_v2 *)msg;
+
+ valid_len += (vfl->num_elements - 1) *
+ sizeof(struct virtchnl_vlan_filter);
+
+ if (vfl->num_elements == 0) {
+ err_msg_format = true;
+ break;
+ }
+ }
+ break;
+ case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2:
+ case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2:
+ case VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2:
+ case VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2:
+ valid_len = sizeof(struct virtchnl_vlan_setting);
+ break;
/* These are always errors coming from the VF. */
case VIRTCHNL_OP_EVENT:
case VIRTCHNL_OP_UNKNOWN:
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH net-next 2/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 negotiation
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 1/6] virtchnl: Add support for new VLAN capabilities Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 3/6] iavf: Add support VIRTCHNL_VF_OFFLOAD_VLAN_V2 during netdev config Tony Nguyen
` (4 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
In order to support the new VIRTCHNL_VF_OFFLOAD_VLAN_V2 capability the
VF driver needs to rework it's initialization state machine and reset
flow. This has to be done because successful negotiation of
VIRTCHNL_VF_OFFLOAD_VLAN_V2 requires the VF driver to perform a second
capability request via VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS before
configuring the adapter and its netdev.
Add the VIRTCHNL_VF_OFFLOAD_VLAN_V2 bit when sending the
VIRTHCNL_OP_GET_VF_RESOURECES message. The underlying PF will either
support VIRTCHNL_VF_OFFLOAD_VLAN or VIRTCHNL_VF_OFFLOAD_VLAN_V2 or
neither. Both of these offloads should never be supported together.
Based on this, add 2 new states to the initialization state machine:
__IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS
__IAVF_INIT_CONFIG_ADAPTER
The __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS state is used to request/store
the new VLAN capabilities if and only if VIRTCHNL_VLAN_OFFLOAD_VLAN_V2
was successfully negotiated in the __IAVF_INIT_GET_RESOURCES state.
The __IAVF_INIT_CONFIG_ADAPTER state is used to configure the
adapter/netdev after the resource requests have finished. The VF will
move into this state regardless of whether it successfully negotiated
VIRTCHNL_VF_OFFLOAD_VLAN or VIRTCHNL_VF_OFFLOAD_VLAN_V2.
Also, add a the new flag IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS and set
it during VF reset. If VIRTCHNL_VF_OFFLOAD_VLAN_V2 was successfully
negotiated then the VF will request its VLAN capabilities via
VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS during the reset. This is needed
because the PF may change/modify the VF's configuration during VF reset
(i.e. modifying the VF's port VLAN configuration).
This also, required the VF to call netdev_update_features() since its
VLAN features may change during VF reset. Make sure to call this under
rtnl_lock().
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/iavf/iavf.h | 9 +
drivers/net/ethernet/intel/iavf/iavf_main.c | 205 +++++++++++++-----
.../net/ethernet/intel/iavf/iavf_virtchnl.c | 78 ++++++-
3 files changed, 240 insertions(+), 52 deletions(-)
diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h
index b5728bdbcf33..edb139834437 100644
--- a/drivers/net/ethernet/intel/iavf/iavf.h
+++ b/drivers/net/ethernet/intel/iavf/iavf.h
@@ -181,6 +181,8 @@ enum iavf_state_t {
__IAVF_REMOVE, /* driver is being unloaded */
__IAVF_INIT_VERSION_CHECK, /* aq msg sent, awaiting reply */
__IAVF_INIT_GET_RESOURCES, /* aq msg sent, awaiting reply */
+ __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS,
+ __IAVF_INIT_CONFIG_ADAPTER,
__IAVF_INIT_SW, /* got resources, setting up structs */
__IAVF_INIT_FAILED, /* init failed, restarting procedure */
__IAVF_RESETTING, /* in reset */
@@ -310,6 +312,7 @@ struct iavf_adapter {
#define IAVF_FLAG_AQ_ADD_ADV_RSS_CFG BIT(27)
#define IAVF_FLAG_AQ_DEL_ADV_RSS_CFG BIT(28)
#define IAVF_FLAG_AQ_REQUEST_STATS BIT(29)
+#define IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS BIT(30)
/* OS defined structs */
struct net_device *netdev;
@@ -349,6 +352,8 @@ struct iavf_adapter {
VIRTCHNL_VF_OFFLOAD_RSS_PF)))
#define VLAN_ALLOWED(_a) ((_a)->vf_res->vf_cap_flags & \
VIRTCHNL_VF_OFFLOAD_VLAN)
+#define VLAN_V2_ALLOWED(_a) ((_a)->vf_res->vf_cap_flags & \
+ VIRTCHNL_VF_OFFLOAD_VLAN_V2)
#define ADV_LINK_SUPPORT(_a) ((_a)->vf_res->vf_cap_flags & \
VIRTCHNL_VF_CAP_ADV_LINK_SPEED)
#define FDIR_FLTR_SUPPORT(_a) ((_a)->vf_res->vf_cap_flags & \
@@ -360,6 +365,7 @@ struct iavf_adapter {
struct virtchnl_version_info pf_version;
#define PF_IS_V11(_a) (((_a)->pf_version.major == 1) && \
((_a)->pf_version.minor == 1))
+ struct virtchnl_vlan_caps vlan_v2_caps;
u16 msg_enable;
struct iavf_eth_stats current_stats;
struct iavf_vsi vsi;
@@ -448,6 +454,7 @@ static inline void iavf_change_state(struct iavf_adapter *adapter,
int iavf_up(struct iavf_adapter *adapter);
void iavf_down(struct iavf_adapter *adapter);
int iavf_process_config(struct iavf_adapter *adapter);
+int iavf_parse_vf_resource_msg(struct iavf_adapter *adapter);
void iavf_schedule_reset(struct iavf_adapter *adapter);
void iavf_schedule_request_stats(struct iavf_adapter *adapter);
void iavf_reset(struct iavf_adapter *adapter);
@@ -466,6 +473,8 @@ int iavf_send_api_ver(struct iavf_adapter *adapter);
int iavf_verify_api_ver(struct iavf_adapter *adapter);
int iavf_send_vf_config_msg(struct iavf_adapter *adapter);
int iavf_get_vf_config(struct iavf_adapter *adapter);
+int iavf_get_vf_vlan_v2_caps(struct iavf_adapter *adapter);
+int iavf_send_vf_offload_vlan_v2_msg(struct iavf_adapter *adapter);
void iavf_irq_enable(struct iavf_adapter *adapter, bool flush);
void iavf_configure_queues(struct iavf_adapter *adapter);
void iavf_deconfigure_queues(struct iavf_adapter *adapter);
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index aa5b8042b13d..b301da6c0a96 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -1584,6 +1584,8 @@ static int iavf_process_aq_command(struct iavf_adapter *adapter)
{
if (adapter->aq_required & IAVF_FLAG_AQ_GET_CONFIG)
return iavf_send_vf_config_msg(adapter);
+ if (adapter->aq_required & IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS)
+ return iavf_send_vf_offload_vlan_v2_msg(adapter);
if (adapter->aq_required & IAVF_FLAG_AQ_DISABLE_QUEUES) {
iavf_disable_queues(adapter);
return 0;
@@ -1826,6 +1828,59 @@ static void iavf_init_version_check(struct iavf_adapter *adapter)
iavf_change_state(adapter, __IAVF_INIT_FAILED);
}
+/**
+ * iavf_parse_vf_resource_msg - parse response from VIRTCHNL_OP_GET_VF_RESOURCES
+ * @adapter: board private structure
+ */
+int iavf_parse_vf_resource_msg(struct iavf_adapter *adapter)
+{
+ int i, num_req_queues = adapter->num_req_queues;
+ struct iavf_vsi *vsi = &adapter->vsi;
+
+ for (i = 0; i < adapter->vf_res->num_vsis; i++) {
+ if (adapter->vf_res->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV)
+ adapter->vsi_res = &adapter->vf_res->vsi_res[i];
+ }
+ if (!adapter->vsi_res) {
+ dev_err(&adapter->pdev->dev, "No LAN VSI found\n");
+ return -ENODEV;
+ }
+
+ if (num_req_queues &&
+ num_req_queues > adapter->vsi_res->num_queue_pairs) {
+ /* Problem. The PF gave us fewer queues than what we had
+ * negotiated in our request. Need a reset to see if we can't
+ * get back to a working state.
+ */
+ dev_err(&adapter->pdev->dev,
+ "Requested %d queues, but PF only gave us %d.\n",
+ num_req_queues,
+ adapter->vsi_res->num_queue_pairs);
+ adapter->flags |= IAVF_FLAG_REINIT_ITR_NEEDED;
+ adapter->num_req_queues = adapter->vsi_res->num_queue_pairs;
+ iavf_schedule_reset(adapter);
+
+ return -EAGAIN;
+ }
+ adapter->num_req_queues = 0;
+ adapter->vsi.id = adapter->vsi_res->vsi_id;
+
+ adapter->vsi.back = adapter;
+ adapter->vsi.base_vector = 1;
+ adapter->vsi.work_limit = IAVF_DEFAULT_IRQ_WORK;
+ vsi->netdev = adapter->netdev;
+ vsi->qs_handle = adapter->vsi_res->qset_handle;
+ if (adapter->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_RSS_PF) {
+ adapter->rss_key_size = adapter->vf_res->rss_key_size;
+ adapter->rss_lut_size = adapter->vf_res->rss_lut_size;
+ } else {
+ adapter->rss_key_size = IAVF_HKEY_ARRAY_SIZE;
+ adapter->rss_lut_size = IAVF_HLUT_ARRAY_SIZE;
+ }
+
+ return 0;
+}
+
/**
* iavf_init_get_resources - third step of driver startup
* @adapter: board private structure
@@ -1837,7 +1892,6 @@ static void iavf_init_version_check(struct iavf_adapter *adapter)
**/
static void iavf_init_get_resources(struct iavf_adapter *adapter)
{
- struct net_device *netdev = adapter->netdev;
struct pci_dev *pdev = adapter->pdev;
struct iavf_hw *hw = &adapter->hw;
int err;
@@ -1855,7 +1909,7 @@ static void iavf_init_get_resources(struct iavf_adapter *adapter)
err = iavf_get_vf_config(adapter);
if (err == IAVF_ERR_ADMIN_QUEUE_NO_WORK) {
err = iavf_send_vf_config_msg(adapter);
- goto err;
+ goto err_alloc;
} else if (err == IAVF_ERR_PARAM) {
/* We only get ERR_PARAM if the device is in a very bad
* state or if we've been disabled for previous bad
@@ -1870,9 +1924,83 @@ static void iavf_init_get_resources(struct iavf_adapter *adapter)
goto err_alloc;
}
- err = iavf_process_config(adapter);
+ err = iavf_parse_vf_resource_msg(adapter);
if (err)
goto err_alloc;
+
+ err = iavf_send_vf_offload_vlan_v2_msg(adapter);
+ if (err == -EOPNOTSUPP) {
+ /* underlying PF doesn't support VIRTCHNL_VF_OFFLOAD_VLAN_V2, so
+ * go directly to finishing initialization
+ */
+ iavf_change_state(adapter, __IAVF_INIT_CONFIG_ADAPTER);
+ return;
+ } else if (err) {
+ dev_err(&pdev->dev, "Unable to send offload vlan v2 request (%d)\n",
+ err);
+ goto err_alloc;
+ }
+
+ /* underlying PF supports VIRTCHNL_VF_OFFLOAD_VLAN_V2, so update the
+ * state accordingly
+ */
+ iavf_change_state(adapter, __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS);
+ return;
+
+err_alloc:
+ kfree(adapter->vf_res);
+ adapter->vf_res = NULL;
+err:
+ iavf_change_state(adapter, __IAVF_INIT_FAILED);
+}
+
+/**
+ * iavf_init_get_offload_vlan_v2_caps - part of driver startup
+ * @adapter: board private structure
+ *
+ * Function processes __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS driver state if the
+ * VF negotiates VIRTCHNL_VF_OFFLOAD_VLAN_V2. If VIRTCHNL_VF_OFFLOAD_VLAN_V2 is
+ * not negotiated, then this state will never be entered.
+ **/
+static void iavf_init_get_offload_vlan_v2_caps(struct iavf_adapter *adapter)
+{
+ int ret;
+
+ WARN_ON(adapter->state != __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS);
+
+ memset(&adapter->vlan_v2_caps, 0, sizeof(adapter->vlan_v2_caps));
+
+ ret = iavf_get_vf_vlan_v2_caps(adapter);
+ if (ret) {
+ if (ret == IAVF_ERR_ADMIN_QUEUE_NO_WORK)
+ iavf_send_vf_offload_vlan_v2_msg(adapter);
+ goto err;
+ }
+
+ iavf_change_state(adapter, __IAVF_INIT_CONFIG_ADAPTER);
+ return;
+err:
+ iavf_change_state(adapter, __IAVF_INIT_FAILED);
+}
+
+/**
+ * iavf_init_config_adapter - last part of driver startup
+ * @adapter: board private structure
+ *
+ * After all the supported capabilities are negotiated, then the
+ * __IAVF_INIT_CONFIG_ADAPTER state will finish driver initialization.
+ */
+static void iavf_init_config_adapter(struct iavf_adapter *adapter)
+{
+ struct net_device *netdev = adapter->netdev;
+ struct pci_dev *pdev = adapter->pdev;
+ int err;
+
+ WARN_ON(adapter->state != __IAVF_INIT_CONFIG_ADAPTER);
+
+ if (iavf_process_config(adapter))
+ goto err;
+
adapter->current_op = VIRTCHNL_OP_UNKNOWN;
adapter->flags |= IAVF_FLAG_RX_CSUM_ENABLED;
@@ -1962,9 +2090,6 @@ static void iavf_init_get_resources(struct iavf_adapter *adapter)
iavf_free_misc_irq(adapter);
err_sw_init:
iavf_reset_interrupt_capability(adapter);
-err_alloc:
- kfree(adapter->vf_res);
- adapter->vf_res = NULL;
err:
iavf_change_state(adapter, __IAVF_INIT_FAILED);
}
@@ -2013,6 +2138,18 @@ static void iavf_watchdog_task(struct work_struct *work)
queue_delayed_work(iavf_wq, &adapter->watchdog_task,
msecs_to_jiffies(1));
return;
+ case __IAVF_INIT_GET_OFFLOAD_VLAN_V2_CAPS:
+ iavf_init_get_offload_vlan_v2_caps(adapter);
+ mutex_unlock(&adapter->crit_lock);
+ queue_delayed_work(iavf_wq, &adapter->watchdog_task,
+ msecs_to_jiffies(1));
+ return;
+ case __IAVF_INIT_CONFIG_ADAPTER:
+ iavf_init_config_adapter(adapter);
+ mutex_unlock(&adapter->crit_lock);
+ queue_delayed_work(iavf_wq, &adapter->watchdog_task,
+ msecs_to_jiffies(1));
+ return;
case __IAVF_INIT_FAILED:
if (++adapter->aq_wait_count > IAVF_AQ_MAX_ERR) {
dev_err(&adapter->pdev->dev,
@@ -2066,10 +2203,13 @@ static void iavf_watchdog_task(struct work_struct *work)
iavf_send_api_ver(adapter);
}
} else {
+ int ret = iavf_process_aq_command(adapter);
+
/* An error will be returned if no commands were
* processed; use this opportunity to update stats
+ * if the error isn't -ENOTSUPP
*/
- if (iavf_process_aq_command(adapter) &&
+ if (ret && ret != -EOPNOTSUPP &&
adapter->state == __IAVF_RUNNING)
iavf_request_stats(adapter);
}
@@ -2308,6 +2448,13 @@ static void iavf_reset_task(struct work_struct *work)
}
adapter->aq_required |= IAVF_FLAG_AQ_GET_CONFIG;
+ /* always set since VIRTCHNL_OP_GET_VF_RESOURCES has not been
+ * sent/received yet, so VLAN_V2_ALLOWED() cannot is not reliable here,
+ * however the VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS won't be sent until
+ * VIRTCHNL_OP_GET_VF_RESOURCES and VIRTCHNL_VF_OFFLOAD_VLAN_V2 have
+ * been successfully sent and negotiated
+ */
+ adapter->aq_required |= IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS;
adapter->aq_required |= IAVF_FLAG_AQ_MAP_VECTORS;
spin_lock_bh(&adapter->mac_vlan_list_lock);
@@ -3603,39 +3750,10 @@ static int iavf_check_reset_complete(struct iavf_hw *hw)
int iavf_process_config(struct iavf_adapter *adapter)
{
struct virtchnl_vf_resource *vfres = adapter->vf_res;
- int i, num_req_queues = adapter->num_req_queues;
struct net_device *netdev = adapter->netdev;
- struct iavf_vsi *vsi = &adapter->vsi;
netdev_features_t hw_enc_features;
netdev_features_t hw_features;
- /* got VF config message back from PF, now we can parse it */
- for (i = 0; i < vfres->num_vsis; i++) {
- if (vfres->vsi_res[i].vsi_type == VIRTCHNL_VSI_SRIOV)
- adapter->vsi_res = &vfres->vsi_res[i];
- }
- if (!adapter->vsi_res) {
- dev_err(&adapter->pdev->dev, "No LAN VSI found\n");
- return -ENODEV;
- }
-
- if (num_req_queues &&
- num_req_queues > adapter->vsi_res->num_queue_pairs) {
- /* Problem. The PF gave us fewer queues than what we had
- * negotiated in our request. Need a reset to see if we can't
- * get back to a working state.
- */
- dev_err(&adapter->pdev->dev,
- "Requested %d queues, but PF only gave us %d.\n",
- num_req_queues,
- adapter->vsi_res->num_queue_pairs);
- adapter->flags |= IAVF_FLAG_REINIT_ITR_NEEDED;
- adapter->num_req_queues = adapter->vsi_res->num_queue_pairs;
- iavf_schedule_reset(adapter);
- return -ENODEV;
- }
- adapter->num_req_queues = 0;
-
hw_enc_features = NETIF_F_SG |
NETIF_F_IP_CSUM |
NETIF_F_IPV6_CSUM |
@@ -3716,21 +3834,6 @@ int iavf_process_config(struct iavf_adapter *adapter)
netdev->features &= ~NETIF_F_GSO;
}
- adapter->vsi.id = adapter->vsi_res->vsi_id;
-
- adapter->vsi.back = adapter;
- adapter->vsi.base_vector = 1;
- adapter->vsi.work_limit = IAVF_DEFAULT_IRQ_WORK;
- vsi->netdev = adapter->netdev;
- vsi->qs_handle = adapter->vsi_res->qset_handle;
- if (vfres->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_RSS_PF) {
- adapter->rss_key_size = vfres->rss_key_size;
- adapter->rss_lut_size = vfres->rss_lut_size;
- } else {
- adapter->rss_key_size = IAVF_HKEY_ARRAY_SIZE;
- adapter->rss_lut_size = IAVF_HLUT_ARRAY_SIZE;
- }
-
return 0;
}
diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
index cfa1f0e0e2fe..1ebff8dc38ba 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
@@ -137,6 +137,7 @@ int iavf_send_vf_config_msg(struct iavf_adapter *adapter)
VIRTCHNL_VF_OFFLOAD_WB_ON_ITR |
VIRTCHNL_VF_OFFLOAD_RSS_PCTYPE_V2 |
VIRTCHNL_VF_OFFLOAD_ENCAP |
+ VIRTCHNL_VF_OFFLOAD_VLAN_V2 |
VIRTCHNL_VF_OFFLOAD_ENCAP_CSUM |
VIRTCHNL_VF_OFFLOAD_REQ_QUEUES |
VIRTCHNL_VF_OFFLOAD_ADQ |
@@ -155,6 +156,19 @@ int iavf_send_vf_config_msg(struct iavf_adapter *adapter)
NULL, 0);
}
+int iavf_send_vf_offload_vlan_v2_msg(struct iavf_adapter *adapter)
+{
+ adapter->aq_required &= ~IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS;
+
+ if (!VLAN_V2_ALLOWED(adapter))
+ return -EOPNOTSUPP;
+
+ adapter->current_op = VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS;
+
+ return iavf_send_pf_msg(adapter, VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS,
+ NULL, 0);
+}
+
/**
* iavf_validate_num_queues
* @adapter: adapter structure
@@ -235,6 +249,45 @@ int iavf_get_vf_config(struct iavf_adapter *adapter)
return err;
}
+int iavf_get_vf_vlan_v2_caps(struct iavf_adapter *adapter)
+{
+ struct iavf_hw *hw = &adapter->hw;
+ struct iavf_arq_event_info event;
+ enum virtchnl_ops op;
+ enum iavf_status err;
+ u16 len;
+
+ len = sizeof(struct virtchnl_vlan_caps);
+ event.buf_len = len;
+ event.msg_buf = kzalloc(event.buf_len, GFP_KERNEL);
+ if (!event.msg_buf) {
+ err = -ENOMEM;
+ goto out;
+ }
+
+ while (1) {
+ /* When the AQ is empty, iavf_clean_arq_element will return
+ * nonzero and this loop will terminate.
+ */
+ err = iavf_clean_arq_element(hw, &event, NULL);
+ if (err)
+ goto out_alloc;
+ op = (enum virtchnl_ops)le32_to_cpu(event.desc.cookie_high);
+ if (op == VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS)
+ break;
+ }
+
+ err = (enum iavf_status)le32_to_cpu(event.desc.cookie_low);
+ if (err)
+ goto out_alloc;
+
+ memcpy(&adapter->vlan_v2_caps, event.msg_buf, min(event.msg_len, len));
+out_alloc:
+ kfree(event.msg_buf);
+out:
+ return err;
+}
+
/**
* iavf_configure_queues
* @adapter: adapter structure
@@ -1759,6 +1812,26 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter,
}
spin_unlock_bh(&adapter->mac_vlan_list_lock);
+
+ iavf_parse_vf_resource_msg(adapter);
+
+ /* negotiated VIRTCHNL_VF_OFFLOAD_VLAN_V2, so wait for the
+ * response to VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS to finish
+ * configuration
+ */
+ if (VLAN_V2_ALLOWED(adapter))
+ break;
+ /* fallthrough and finish config if VIRTCHNL_VF_OFFLOAD_VLAN_V2
+ * wasn't successfully negotiated with the PF
+ */
+ }
+ fallthrough;
+ case VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS: {
+ if (v_opcode == VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS)
+ memcpy(&adapter->vlan_v2_caps, msg,
+ min_t(u16, msglen,
+ sizeof(adapter->vlan_v2_caps)));
+
iavf_process_config(adapter);
/* unlock crit_lock before acquiring rtnl_lock as other
@@ -1766,8 +1839,11 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter,
* crit_lock
*/
mutex_unlock(&adapter->crit_lock);
+ /* VLAN capabilities can change during VFR, so make sure to
+ * update the netdev features with the new capabilities
+ */
rtnl_lock();
- netdev_update_features(adapter->netdev);
+ netdev_update_features(netdev);
rtnl_unlock();
if (iavf_lock_timeout(&adapter->crit_lock, 10000))
dev_warn(&adapter->pdev->dev, "failed to acquire crit_lock in %s\n",
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH net-next 3/6] iavf: Add support VIRTCHNL_VF_OFFLOAD_VLAN_V2 during netdev config
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 1/6] virtchnl: Add support for new VLAN capabilities Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 2/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 negotiation Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 4/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 hotpath Tony Nguyen
` (3 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
Based on VIRTCHNL_VF_OFFLOAD_VLAN_V2, the VF can now support more VLAN
capabilities (i.e. 802.1AD offloads and filtering). In order to
communicate these capabilities to the netdev layer, the VF needs to
parse its VLAN capabilities based on whether it was able to negotiation
VIRTCHNL_VF_OFFLOAD_VLAN or VIRTCHNL_VF_OFFLOAD_VLAN_V2 or neither of
these.
In order to support this, add the following functionality:
iavf_get_netdev_vlan_hw_features() - This is used to determine the VLAN
features that the underlying hardware supports and that can be toggled
off/on based on the negotiated capabiltiies. For example, if
VIRTCHNL_VF_OFFLOAD_VLAN_V2 was negotiated, then any capability marked
with VIRTCHNL_VLAN_TOGGLE can be toggled on/off by the VF. If
VIRTCHNL_VF_OFFLOAD_VLAN was negotiated, then only VLAN insertion and/or
stripping can be toggled on/off.
iavf_get_netdev_vlan_features() - This is used to determine the VLAN
features that the underlying hardware supports and that should be
enabled by default. For example, if VIRTHCNL_VF_OFFLOAD_VLAN_V2 was
negotiated, then any supported capability that has its ethertype_init
filed set should be enabled by default. If VIRTCHNL_VF_OFFLOAD_VLAN was
negotiated, then filtering, stripping, and insertion should be enabled
by default.
Also, refactor iavf_fix_features() to take into account the new
capabilities. To do this, query all the supported features (enabled by
default and toggleable) and make sure the requested change is supported.
If VIRTCHNL_VF_OFFLOAD_VLAN_V2 is successfully negotiated, there is no
need to check VIRTCHNL_VLAN_TOGGLE here because the driver already told
the netdev layer which features can be toggled via netdev->hw_features
during iavf_process_config(), so only those features will be requested
to change.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/iavf/iavf.h | 17 +-
drivers/net/ethernet/intel/iavf/iavf_main.c | 279 ++++++++++++++++--
.../net/ethernet/intel/iavf/iavf_virtchnl.c | 251 +++++++++++-----
3 files changed, 453 insertions(+), 94 deletions(-)
diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h
index edb139834437..5fb6ebf9a760 100644
--- a/drivers/net/ethernet/intel/iavf/iavf.h
+++ b/drivers/net/ethernet/intel/iavf/iavf.h
@@ -55,7 +55,8 @@ enum iavf_vsi_state_t {
struct iavf_vsi {
struct iavf_adapter *back;
struct net_device *netdev;
- unsigned long active_vlans[BITS_TO_LONGS(VLAN_N_VID)];
+ unsigned long active_cvlans[BITS_TO_LONGS(VLAN_N_VID)];
+ unsigned long active_svlans[BITS_TO_LONGS(VLAN_N_VID)];
u16 seid;
u16 id;
DECLARE_BITMAP(state, __IAVF_VSI_STATE_SIZE__);
@@ -146,9 +147,15 @@ struct iavf_mac_filter {
};
};
+#define IAVF_VLAN(vid, tpid) ((struct iavf_vlan){ vid, tpid })
+struct iavf_vlan {
+ u16 vid;
+ u16 tpid;
+};
+
struct iavf_vlan_filter {
struct list_head list;
- u16 vlan;
+ struct iavf_vlan vlan;
bool remove; /* filter needs to be removed */
bool add; /* filter needs to be added */
};
@@ -354,6 +361,12 @@ struct iavf_adapter {
VIRTCHNL_VF_OFFLOAD_VLAN)
#define VLAN_V2_ALLOWED(_a) ((_a)->vf_res->vf_cap_flags & \
VIRTCHNL_VF_OFFLOAD_VLAN_V2)
+#define VLAN_V2_FILTERING_ALLOWED(_a) \
+ (VLAN_V2_ALLOWED((_a)) && \
+ ((_a)->vlan_v2_caps.filtering.filtering_support.outer || \
+ (_a)->vlan_v2_caps.filtering.filtering_support.inner))
+#define VLAN_FILTERING_ALLOWED(_a) \
+ (VLAN_ALLOWED((_a)) || VLAN_V2_FILTERING_ALLOWED((_a)))
#define ADV_LINK_SUPPORT(_a) ((_a)->vf_res->vf_cap_flags & \
VIRTCHNL_VF_CAP_ADV_LINK_SPEED)
#define FDIR_FLTR_SUPPORT(_a) ((_a)->vf_res->vf_cap_flags & \
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index b301da6c0a96..6b7f59718097 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -646,14 +646,17 @@ static void iavf_configure_rx(struct iavf_adapter *adapter)
* mac_vlan_list_lock.
**/
static struct
-iavf_vlan_filter *iavf_find_vlan(struct iavf_adapter *adapter, u16 vlan)
+iavf_vlan_filter *iavf_find_vlan(struct iavf_adapter *adapter,
+ struct iavf_vlan vlan)
{
struct iavf_vlan_filter *f;
list_for_each_entry(f, &adapter->vlan_filter_list, list) {
- if (vlan == f->vlan)
+ if (f->vlan.vid == vlan.vid &&
+ f->vlan.tpid == vlan.tpid)
return f;
}
+
return NULL;
}
@@ -665,7 +668,8 @@ iavf_vlan_filter *iavf_find_vlan(struct iavf_adapter *adapter, u16 vlan)
* Returns ptr to the filter object or NULL when no memory available.
**/
static struct
-iavf_vlan_filter *iavf_add_vlan(struct iavf_adapter *adapter, u16 vlan)
+iavf_vlan_filter *iavf_add_vlan(struct iavf_adapter *adapter,
+ struct iavf_vlan vlan)
{
struct iavf_vlan_filter *f = NULL;
@@ -694,7 +698,7 @@ iavf_vlan_filter *iavf_add_vlan(struct iavf_adapter *adapter, u16 vlan)
* @adapter: board private structure
* @vlan: VLAN tag
**/
-static void iavf_del_vlan(struct iavf_adapter *adapter, u16 vlan)
+static void iavf_del_vlan(struct iavf_adapter *adapter, struct iavf_vlan vlan)
{
struct iavf_vlan_filter *f;
@@ -720,8 +724,11 @@ static void iavf_restore_filters(struct iavf_adapter *adapter)
u16 vid;
/* re-add all VLAN filters */
- for_each_set_bit(vid, adapter->vsi.active_vlans, VLAN_N_VID)
- iavf_add_vlan(adapter, vid);
+ for_each_set_bit(vid, adapter->vsi.active_cvlans, VLAN_N_VID)
+ iavf_add_vlan(adapter, IAVF_VLAN(vid, ETH_P_8021Q));
+
+ for_each_set_bit(vid, adapter->vsi.active_svlans, VLAN_N_VID)
+ iavf_add_vlan(adapter, IAVF_VLAN(vid, ETH_P_8021AD));
}
/**
@@ -735,13 +742,17 @@ static int iavf_vlan_rx_add_vid(struct net_device *netdev,
{
struct iavf_adapter *adapter = netdev_priv(netdev);
- if (!VLAN_ALLOWED(adapter))
+ if (!VLAN_FILTERING_ALLOWED(adapter))
return -EIO;
- if (iavf_add_vlan(adapter, vid) == NULL)
+ if (!iavf_add_vlan(adapter, IAVF_VLAN(vid, be16_to_cpu(proto))))
return -ENOMEM;
- set_bit(vid, adapter->vsi.active_vlans);
+ if (proto == cpu_to_be16(ETH_P_8021Q))
+ set_bit(vid, adapter->vsi.active_cvlans);
+ else
+ set_bit(vid, adapter->vsi.active_svlans);
+
return 0;
}
@@ -756,8 +767,11 @@ static int iavf_vlan_rx_kill_vid(struct net_device *netdev,
{
struct iavf_adapter *adapter = netdev_priv(netdev);
- iavf_del_vlan(adapter, vid);
- clear_bit(vid, adapter->vsi.active_vlans);
+ iavf_del_vlan(adapter, IAVF_VLAN(vid, be16_to_cpu(proto)));
+ if (proto == cpu_to_be16(ETH_P_8021Q))
+ clear_bit(vid, adapter->vsi.active_cvlans);
+ else
+ clear_bit(vid, adapter->vsi.active_svlans);
return 0;
}
@@ -3680,6 +3694,228 @@ static netdev_features_t iavf_features_check(struct sk_buff *skb,
return features & ~(NETIF_F_CSUM_MASK | NETIF_F_GSO_MASK);
}
+/**
+ * iavf_get_netdev_vlan_hw_features - get NETDEV VLAN features that can toggle on/off
+ * @adapter: board private structure
+ *
+ * Depending on whether VIRTHCNL_VF_OFFLOAD_VLAN or VIRTCHNL_VF_OFFLOAD_VLAN_V2
+ * were negotiated determine the VLAN features that can be toggled on and off.
+ **/
+static netdev_features_t
+iavf_get_netdev_vlan_hw_features(struct iavf_adapter *adapter)
+{
+ netdev_features_t hw_features = 0;
+
+ if (!adapter->vf_res || !adapter->vf_res->vf_cap_flags)
+ return hw_features;
+
+ /* Enable VLAN features if supported */
+ if (VLAN_ALLOWED(adapter)) {
+ hw_features |= (NETIF_F_HW_VLAN_CTAG_TX |
+ NETIF_F_HW_VLAN_CTAG_RX);
+ } else if (VLAN_V2_ALLOWED(adapter)) {
+ struct virtchnl_vlan_caps *vlan_v2_caps =
+ &adapter->vlan_v2_caps;
+ struct virtchnl_vlan_supported_caps *stripping_support =
+ &vlan_v2_caps->offloads.stripping_support;
+ struct virtchnl_vlan_supported_caps *insertion_support =
+ &vlan_v2_caps->offloads.insertion_support;
+
+ if (stripping_support->outer != VIRTCHNL_VLAN_UNSUPPORTED &&
+ stripping_support->outer & VIRTCHNL_VLAN_TOGGLE) {
+ if (stripping_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_8100)
+ hw_features |= NETIF_F_HW_VLAN_CTAG_RX;
+ if (stripping_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ hw_features |= NETIF_F_HW_VLAN_STAG_RX;
+ } else if (stripping_support->inner !=
+ VIRTCHNL_VLAN_UNSUPPORTED &&
+ stripping_support->inner & VIRTCHNL_VLAN_TOGGLE) {
+ if (stripping_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_8100)
+ hw_features |= NETIF_F_HW_VLAN_CTAG_RX;
+ }
+
+ if (insertion_support->outer != VIRTCHNL_VLAN_UNSUPPORTED &&
+ insertion_support->outer & VIRTCHNL_VLAN_TOGGLE) {
+ if (insertion_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_8100)
+ hw_features |= NETIF_F_HW_VLAN_CTAG_TX;
+ if (insertion_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ hw_features |= NETIF_F_HW_VLAN_STAG_TX;
+ } else if (insertion_support->inner &&
+ insertion_support->inner & VIRTCHNL_VLAN_TOGGLE) {
+ if (insertion_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_8100)
+ hw_features |= NETIF_F_HW_VLAN_CTAG_TX;
+ }
+ }
+
+ return hw_features;
+}
+
+/**
+ * iavf_get_netdev_vlan_features - get the enabled NETDEV VLAN fetures
+ * @adapter: board private structure
+ *
+ * Depending on whether VIRTHCNL_VF_OFFLOAD_VLAN or VIRTCHNL_VF_OFFLOAD_VLAN_V2
+ * were negotiated determine the VLAN features that are enabled by default.
+ **/
+static netdev_features_t
+iavf_get_netdev_vlan_features(struct iavf_adapter *adapter)
+{
+ netdev_features_t features = 0;
+
+ if (!adapter->vf_res || !adapter->vf_res->vf_cap_flags)
+ return features;
+
+ if (VLAN_ALLOWED(adapter)) {
+ features |= NETIF_F_HW_VLAN_CTAG_FILTER |
+ NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX;
+ } else if (VLAN_V2_ALLOWED(adapter)) {
+ struct virtchnl_vlan_caps *vlan_v2_caps =
+ &adapter->vlan_v2_caps;
+ struct virtchnl_vlan_supported_caps *filtering_support =
+ &vlan_v2_caps->filtering.filtering_support;
+ struct virtchnl_vlan_supported_caps *stripping_support =
+ &vlan_v2_caps->offloads.stripping_support;
+ struct virtchnl_vlan_supported_caps *insertion_support =
+ &vlan_v2_caps->offloads.insertion_support;
+ u32 ethertype_init;
+
+ /* give priority to outer stripping and don't support both outer
+ * and inner stripping
+ */
+ ethertype_init = vlan_v2_caps->offloads.ethertype_init;
+ if (stripping_support->outer != VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (stripping_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_RX;
+ else if (stripping_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ features |= NETIF_F_HW_VLAN_STAG_RX;
+ } else if (stripping_support->inner !=
+ VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (stripping_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_RX;
+ }
+
+ /* give priority to outer insertion and don't support both outer
+ * and inner insertion
+ */
+ if (insertion_support->outer != VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (insertion_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_TX;
+ else if (insertion_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ features |= NETIF_F_HW_VLAN_STAG_TX;
+ } else if (insertion_support->inner !=
+ VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (insertion_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_TX;
+ }
+
+ /* give priority to outer filtering and don't bother if both
+ * outer and inner filtering are enabled
+ */
+ ethertype_init = vlan_v2_caps->filtering.ethertype_init;
+ if (filtering_support->outer != VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (filtering_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_FILTER;
+ if (filtering_support->outer &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ features |= NETIF_F_HW_VLAN_STAG_FILTER;
+ } else if (filtering_support->inner !=
+ VIRTCHNL_VLAN_UNSUPPORTED) {
+ if (filtering_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_8100 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_8100)
+ features |= NETIF_F_HW_VLAN_CTAG_FILTER;
+ if (filtering_support->inner &
+ VIRTCHNL_VLAN_ETHERTYPE_88A8 &&
+ ethertype_init & VIRTCHNL_VLAN_ETHERTYPE_88A8)
+ features |= NETIF_F_HW_VLAN_STAG_FILTER;
+ }
+ }
+
+ return features;
+}
+
+#define IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested, allowed, feature_bit) \
+ (!(((requested) & (feature_bit)) && \
+ !((allowed) & (feature_bit))))
+
+/**
+ * iavf_fix_netdev_vlan_features - fix NETDEV VLAN features based on support
+ * @adapter: board private structure
+ * @requested_features: stack requested NETDEV features
+ **/
+static netdev_features_t
+iavf_fix_netdev_vlan_features(struct iavf_adapter *adapter,
+ netdev_features_t requested_features)
+{
+ netdev_features_t allowed_features;
+
+ allowed_features = iavf_get_netdev_vlan_hw_features(adapter) |
+ iavf_get_netdev_vlan_features(adapter);
+
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_CTAG_TX))
+ requested_features &= ~NETIF_F_HW_VLAN_CTAG_TX;
+
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_CTAG_RX))
+ requested_features &= ~NETIF_F_HW_VLAN_CTAG_RX;
+
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_STAG_TX))
+ requested_features &= ~NETIF_F_HW_VLAN_STAG_TX;
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_STAG_RX))
+ requested_features &= ~NETIF_F_HW_VLAN_STAG_RX;
+
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_CTAG_FILTER))
+ requested_features &= ~NETIF_F_HW_VLAN_CTAG_FILTER;
+
+ if (!IAVF_NETDEV_VLAN_FEATURE_ALLOWED(requested_features,
+ allowed_features,
+ NETIF_F_HW_VLAN_STAG_FILTER))
+ requested_features &= ~NETIF_F_HW_VLAN_STAG_FILTER;
+
+ if ((requested_features &
+ (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX)) &&
+ (requested_features &
+ (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_STAG_TX)) &&
+ adapter->vlan_v2_caps.offloads.ethertype_match ==
+ VIRTCHNL_ETHERTYPE_STRIPPING_MATCHES_INSERTION) {
+ netdev_warn(adapter->netdev, "cannot support CTAG and STAG VLAN stripping and/or insertion simultaneously since CTAG and STAG offloads are mutually exclusive, clearing STAG offload settings\n");
+ requested_features &= ~(NETIF_F_HW_VLAN_STAG_RX |
+ NETIF_F_HW_VLAN_STAG_TX);
+ }
+
+ return requested_features;
+}
+
/**
* iavf_fix_features - fix up the netdev feature bits
* @netdev: our net device
@@ -3692,13 +3928,7 @@ static netdev_features_t iavf_fix_features(struct net_device *netdev,
{
struct iavf_adapter *adapter = netdev_priv(netdev);
- if (adapter->vf_res &&
- !(adapter->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_VLAN))
- features &= ~(NETIF_F_HW_VLAN_CTAG_TX |
- NETIF_F_HW_VLAN_CTAG_RX |
- NETIF_F_HW_VLAN_CTAG_FILTER);
-
- return features;
+ return iavf_fix_netdev_vlan_features(adapter, features);
}
static const struct net_device_ops iavf_netdev_ops = {
@@ -3750,6 +3980,7 @@ static int iavf_check_reset_complete(struct iavf_hw *hw)
int iavf_process_config(struct iavf_adapter *adapter)
{
struct virtchnl_vf_resource *vfres = adapter->vf_res;
+ netdev_features_t hw_vlan_features, vlan_features;
struct net_device *netdev = adapter->netdev;
netdev_features_t hw_enc_features;
netdev_features_t hw_features;
@@ -3797,19 +4028,19 @@ int iavf_process_config(struct iavf_adapter *adapter)
*/
hw_features = hw_enc_features;
- /* Enable VLAN features if supported */
- if (vfres->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_VLAN)
- hw_features |= (NETIF_F_HW_VLAN_CTAG_TX |
- NETIF_F_HW_VLAN_CTAG_RX);
+ /* get HW VLAN features that can be toggled */
+ hw_vlan_features = iavf_get_netdev_vlan_hw_features(adapter);
+
/* Enable cloud filter if ADQ is supported */
if (vfres->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_ADQ)
hw_features |= NETIF_F_HW_TC;
if (vfres->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_USO)
hw_features |= NETIF_F_GSO_UDP_L4;
- netdev->hw_features |= hw_features;
+ netdev->hw_features |= hw_features | hw_vlan_features;
+ vlan_features = iavf_get_netdev_vlan_features(adapter);
- netdev->features |= hw_features;
+ netdev->features |= hw_features | vlan_features;
if (vfres->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_VLAN)
netdev->features |= NETIF_F_HW_VLAN_CTAG_FILTER;
diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
index 1ebff8dc38ba..783d829874bc 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
@@ -642,7 +642,6 @@ static void iavf_mac_add_reject(struct iavf_adapter *adapter)
**/
void iavf_add_vlans(struct iavf_adapter *adapter)
{
- struct virtchnl_vlan_filter_list *vvfl;
int len, i = 0, count = 0;
struct iavf_vlan_filter *f;
bool more = false;
@@ -660,48 +659,105 @@ void iavf_add_vlans(struct iavf_adapter *adapter)
if (f->add)
count++;
}
- if (!count || !VLAN_ALLOWED(adapter)) {
+ if (!count || !VLAN_FILTERING_ALLOWED(adapter)) {
adapter->aq_required &= ~IAVF_FLAG_AQ_ADD_VLAN_FILTER;
spin_unlock_bh(&adapter->mac_vlan_list_lock);
return;
}
- adapter->current_op = VIRTCHNL_OP_ADD_VLAN;
- len = sizeof(struct virtchnl_vlan_filter_list) +
- (count * sizeof(u16));
- if (len > IAVF_MAX_AQ_BUF_SIZE) {
- dev_warn(&adapter->pdev->dev, "Too many add VLAN changes in one request\n");
- count = (IAVF_MAX_AQ_BUF_SIZE -
- sizeof(struct virtchnl_vlan_filter_list)) /
- sizeof(u16);
- len = sizeof(struct virtchnl_vlan_filter_list) +
- (count * sizeof(u16));
- more = true;
- }
- vvfl = kzalloc(len, GFP_ATOMIC);
- if (!vvfl) {
+ if (VLAN_ALLOWED(adapter)) {
+ struct virtchnl_vlan_filter_list *vvfl;
+
+ adapter->current_op = VIRTCHNL_OP_ADD_VLAN;
+
+ len = sizeof(*vvfl) + (count * sizeof(u16));
+ if (len > IAVF_MAX_AQ_BUF_SIZE) {
+ dev_warn(&adapter->pdev->dev, "Too many add VLAN changes in one request\n");
+ count = (IAVF_MAX_AQ_BUF_SIZE - sizeof(*vvfl)) /
+ sizeof(u16);
+ len = sizeof(*vvfl) + (count * sizeof(u16));
+ more = true;
+ }
+ vvfl = kzalloc(len, GFP_ATOMIC);
+ if (!vvfl) {
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ return;
+ }
+
+ vvfl->vsi_id = adapter->vsi_res->vsi_id;
+ vvfl->num_elements = count;
+ list_for_each_entry(f, &adapter->vlan_filter_list, list) {
+ if (f->add) {
+ vvfl->vlan_id[i] = f->vlan.vid;
+ i++;
+ f->add = false;
+ if (i == count)
+ break;
+ }
+ }
+ if (!more)
+ adapter->aq_required &= ~IAVF_FLAG_AQ_ADD_VLAN_FILTER;
+
spin_unlock_bh(&adapter->mac_vlan_list_lock);
- return;
- }
- vvfl->vsi_id = adapter->vsi_res->vsi_id;
- vvfl->num_elements = count;
- list_for_each_entry(f, &adapter->vlan_filter_list, list) {
- if (f->add) {
- vvfl->vlan_id[i] = f->vlan;
- i++;
- f->add = false;
- if (i == count)
- break;
+ iavf_send_pf_msg(adapter, VIRTCHNL_OP_ADD_VLAN, (u8 *)vvfl, len);
+ kfree(vvfl);
+ } else {
+ struct virtchnl_vlan_filter_list_v2 *vvfl_v2;
+
+ adapter->current_op = VIRTCHNL_OP_ADD_VLAN_V2;
+
+ len = sizeof(*vvfl_v2) + ((count - 1) *
+ sizeof(struct virtchnl_vlan_filter));
+ if (len > IAVF_MAX_AQ_BUF_SIZE) {
+ dev_warn(&adapter->pdev->dev, "Too many add VLAN changes in one request\n");
+ count = (IAVF_MAX_AQ_BUF_SIZE - sizeof(*vvfl_v2)) /
+ sizeof(struct virtchnl_vlan_filter);
+ len = sizeof(*vvfl_v2) +
+ ((count - 1) *
+ sizeof(struct virtchnl_vlan_filter));
+ more = true;
}
- }
- if (!more)
- adapter->aq_required &= ~IAVF_FLAG_AQ_ADD_VLAN_FILTER;
- spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ vvfl_v2 = kzalloc(len, GFP_ATOMIC);
+ if (!vvfl_v2) {
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ return;
+ }
- iavf_send_pf_msg(adapter, VIRTCHNL_OP_ADD_VLAN, (u8 *)vvfl, len);
- kfree(vvfl);
+ vvfl_v2->vport_id = adapter->vsi_res->vsi_id;
+ vvfl_v2->num_elements = count;
+ list_for_each_entry(f, &adapter->vlan_filter_list, list) {
+ if (f->add) {
+ struct virtchnl_vlan_supported_caps *filtering_support =
+ &adapter->vlan_v2_caps.filtering.filtering_support;
+ struct virtchnl_vlan *vlan;
+
+ /* give priority over outer if it's enabled */
+ if (filtering_support->outer)
+ vlan = &vvfl_v2->filters[i].outer;
+ else
+ vlan = &vvfl_v2->filters[i].inner;
+
+ vlan->tci = f->vlan.vid;
+ vlan->tpid = f->vlan.tpid;
+
+ i++;
+ f->add = false;
+ if (i == count)
+ break;
+ }
+ }
+
+ if (!more)
+ adapter->aq_required &= ~IAVF_FLAG_AQ_ADD_VLAN_FILTER;
+
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+
+ iavf_send_pf_msg(adapter, VIRTCHNL_OP_ADD_VLAN_V2,
+ (u8 *)vvfl_v2, len);
+ kfree(vvfl_v2);
+ }
}
/**
@@ -712,7 +768,6 @@ void iavf_add_vlans(struct iavf_adapter *adapter)
**/
void iavf_del_vlans(struct iavf_adapter *adapter)
{
- struct virtchnl_vlan_filter_list *vvfl;
struct iavf_vlan_filter *f, *ftmp;
int len, i = 0, count = 0;
bool more = false;
@@ -733,56 +788,116 @@ void iavf_del_vlans(struct iavf_adapter *adapter)
* filters marked for removal to enable bailing out before
* sending a virtchnl message
*/
- if (f->remove && !VLAN_ALLOWED(adapter)) {
+ if (f->remove && !VLAN_FILTERING_ALLOWED(adapter)) {
list_del(&f->list);
kfree(f);
} else if (f->remove) {
count++;
}
}
- if (!count) {
+ if (!count || !VLAN_FILTERING_ALLOWED(adapter)) {
adapter->aq_required &= ~IAVF_FLAG_AQ_DEL_VLAN_FILTER;
spin_unlock_bh(&adapter->mac_vlan_list_lock);
return;
}
- adapter->current_op = VIRTCHNL_OP_DEL_VLAN;
- len = sizeof(struct virtchnl_vlan_filter_list) +
- (count * sizeof(u16));
- if (len > IAVF_MAX_AQ_BUF_SIZE) {
- dev_warn(&adapter->pdev->dev, "Too many delete VLAN changes in one request\n");
- count = (IAVF_MAX_AQ_BUF_SIZE -
- sizeof(struct virtchnl_vlan_filter_list)) /
- sizeof(u16);
- len = sizeof(struct virtchnl_vlan_filter_list) +
- (count * sizeof(u16));
- more = true;
- }
- vvfl = kzalloc(len, GFP_ATOMIC);
- if (!vvfl) {
+ if (VLAN_ALLOWED(adapter)) {
+ struct virtchnl_vlan_filter_list *vvfl;
+
+ adapter->current_op = VIRTCHNL_OP_DEL_VLAN;
+
+ len = sizeof(*vvfl) + (count * sizeof(u16));
+ if (len > IAVF_MAX_AQ_BUF_SIZE) {
+ dev_warn(&adapter->pdev->dev, "Too many delete VLAN changes in one request\n");
+ count = (IAVF_MAX_AQ_BUF_SIZE - sizeof(*vvfl)) /
+ sizeof(u16);
+ len = sizeof(*vvfl) + (count * sizeof(u16));
+ more = true;
+ }
+ vvfl = kzalloc(len, GFP_ATOMIC);
+ if (!vvfl) {
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ return;
+ }
+
+ vvfl->vsi_id = adapter->vsi_res->vsi_id;
+ vvfl->num_elements = count;
+ list_for_each_entry_safe(f, ftmp, &adapter->vlan_filter_list, list) {
+ if (f->remove) {
+ vvfl->vlan_id[i] = f->vlan.vid;
+ i++;
+ list_del(&f->list);
+ kfree(f);
+ if (i == count)
+ break;
+ }
+ }
+
+ if (!more)
+ adapter->aq_required &= ~IAVF_FLAG_AQ_DEL_VLAN_FILTER;
+
spin_unlock_bh(&adapter->mac_vlan_list_lock);
- return;
- }
- vvfl->vsi_id = adapter->vsi_res->vsi_id;
- vvfl->num_elements = count;
- list_for_each_entry_safe(f, ftmp, &adapter->vlan_filter_list, list) {
- if (f->remove) {
- vvfl->vlan_id[i] = f->vlan;
- i++;
- list_del(&f->list);
- kfree(f);
- if (i == count)
- break;
+ iavf_send_pf_msg(adapter, VIRTCHNL_OP_DEL_VLAN, (u8 *)vvfl, len);
+ kfree(vvfl);
+ } else {
+ struct virtchnl_vlan_filter_list_v2 *vvfl_v2;
+
+ adapter->current_op = VIRTCHNL_OP_DEL_VLAN_V2;
+
+ len = sizeof(*vvfl_v2) +
+ ((count - 1) * sizeof(struct virtchnl_vlan_filter));
+ if (len > IAVF_MAX_AQ_BUF_SIZE) {
+ dev_warn(&adapter->pdev->dev, "Too many add VLAN changes in one request\n");
+ count = (IAVF_MAX_AQ_BUF_SIZE -
+ sizeof(*vvfl_v2)) /
+ sizeof(struct virtchnl_vlan_filter);
+ len = sizeof(*vvfl_v2) +
+ ((count - 1) *
+ sizeof(struct virtchnl_vlan_filter));
+ more = true;
}
- }
- if (!more)
- adapter->aq_required &= ~IAVF_FLAG_AQ_DEL_VLAN_FILTER;
- spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ vvfl_v2 = kzalloc(len, GFP_ATOMIC);
+ if (!vvfl_v2) {
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+ return;
+ }
+
+ vvfl_v2->vport_id = adapter->vsi_res->vsi_id;
+ vvfl_v2->num_elements = count;
+ list_for_each_entry_safe(f, ftmp, &adapter->vlan_filter_list, list) {
+ if (f->remove) {
+ struct virtchnl_vlan_supported_caps *filtering_support =
+ &adapter->vlan_v2_caps.filtering.filtering_support;
+ struct virtchnl_vlan *vlan;
+
+ /* give priority over outer if it's enabled */
+ if (filtering_support->outer)
+ vlan = &vvfl_v2->filters[i].outer;
+ else
+ vlan = &vvfl_v2->filters[i].inner;
+
+ vlan->tci = f->vlan.vid;
+ vlan->tpid = f->vlan.tpid;
+
+ list_del(&f->list);
+ kfree(f);
+ i++;
+ if (i == count)
+ break;
+ }
+ }
- iavf_send_pf_msg(adapter, VIRTCHNL_OP_DEL_VLAN, (u8 *)vvfl, len);
- kfree(vvfl);
+ if (!more)
+ adapter->aq_required &= ~IAVF_FLAG_AQ_DEL_VLAN_FILTER;
+
+ spin_unlock_bh(&adapter->mac_vlan_list_lock);
+
+ iavf_send_pf_msg(adapter, VIRTCHNL_OP_DEL_VLAN_V2,
+ (u8 *)vvfl_v2, len);
+ kfree(vvfl_v2);
+ }
}
/**
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH net-next 4/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 hotpath
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
` (2 preceding siblings ...)
2021-12-17 22:06 ` [PATCH net-next 3/6] iavf: Add support VIRTCHNL_VF_OFFLOAD_VLAN_V2 during netdev config Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 5/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 offload enable/disable Tony Nguyen
` (2 subsequent siblings)
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
The new VIRTCHNL_VF_OFFLOAD_VLAN_V2 capability added support that allows
the PF to set the location of the Tx and Rx VLAN tag for insertion and
stripping offloads. In order to support this functionality a few changes
are needed.
1. Add a new method to cache the VLAN tag location based on negotiated
capabilities for the Tx and Rx ring flags. This needs to be called in
the initialization and reset paths.
2. Refactor the transmit hotpath to account for the new Tx ring flags.
When IAVF_TXR_FLAGS_VLAN_LOC_L2TAG2 is set, then the driver needs to
insert the VLAN tag in the L2TAG2 field of the transmit descriptor.
When the IAVF_TXRX_FLAGS_VLAN_LOC_L2TAG1 is set, then the driver needs
to use the l2tag1 field of the data descriptor (same behavior as
before).
3. Refactor the iavf_tx_prepare_vlan_flags() function to simplify
transmit hardware VLAN offload functionality by only depending on the
skb_vlan_tag_present() function. This can be done because the OS
won't request transmit offload for a VLAN unless the driver told the
OS it's supported and enabled.
4. Refactor the receive hotpath to account for the new Rx ring flags and
VLAN ethertypes. This requires checking the Rx ring flags and
descriptor status bits to determine the location of the VLAN tag.
Also, since only a single ethertype can be supported at a time, check
the enabled netdev features before specifying a VLAN ethertype in
__vlan_hwaccel_put_tag().
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/iavf/iavf.h | 1 +
drivers/net/ethernet/intel/iavf/iavf_main.c | 82 +++++++++++++++++++
drivers/net/ethernet/intel/iavf/iavf_txrx.c | 71 ++++++++--------
drivers/net/ethernet/intel/iavf/iavf_txrx.h | 30 ++++---
.../net/ethernet/intel/iavf/iavf_virtchnl.c | 2 +
5 files changed, 135 insertions(+), 51 deletions(-)
diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h
index 5fb6ebf9a760..2660d46da1b5 100644
--- a/drivers/net/ethernet/intel/iavf/iavf.h
+++ b/drivers/net/ethernet/intel/iavf/iavf.h
@@ -488,6 +488,7 @@ int iavf_send_vf_config_msg(struct iavf_adapter *adapter);
int iavf_get_vf_config(struct iavf_adapter *adapter);
int iavf_get_vf_vlan_v2_caps(struct iavf_adapter *adapter);
int iavf_send_vf_offload_vlan_v2_msg(struct iavf_adapter *adapter);
+void iavf_set_queue_vlan_tag_loc(struct iavf_adapter *adapter);
void iavf_irq_enable(struct iavf_adapter *adapter, bool flush);
void iavf_configure_queues(struct iavf_adapter *adapter);
void iavf_deconfigure_queues(struct iavf_adapter *adapter);
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index 6b7f59718097..cd9c7484ec99 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -1165,6 +1165,86 @@ static void iavf_free_queues(struct iavf_adapter *adapter)
adapter->rx_rings = NULL;
}
+/**
+ * iavf_set_queue_vlan_tag_loc - set location for VLAN tag offload
+ * @adapter: board private structure
+ *
+ * Based on negotiated capabilities, the VLAN tag needs to be inserted and/or
+ * stripped in certain descriptor fields. Instead of checking the offload
+ * capability bits in the hot path, cache the location the ring specific
+ * flags.
+ */
+void iavf_set_queue_vlan_tag_loc(struct iavf_adapter *adapter)
+{
+ int i;
+
+ for (i = 0; i < adapter->num_active_queues; i++) {
+ struct iavf_ring *tx_ring = &adapter->tx_rings[i];
+ struct iavf_ring *rx_ring = &adapter->rx_rings[i];
+
+ /* prevent multiple L2TAG bits being set after VFR */
+ tx_ring->flags &=
+ ~(IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1 |
+ IAVF_TXR_FLAGS_VLAN_TAG_LOC_L2TAG2);
+ rx_ring->flags &=
+ ~(IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1 |
+ IAVF_RXR_FLAGS_VLAN_TAG_LOC_L2TAG2_2);
+
+ if (VLAN_ALLOWED(adapter)) {
+ tx_ring->flags |= IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ rx_ring->flags |= IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ } else if (VLAN_V2_ALLOWED(adapter)) {
+ struct virtchnl_vlan_supported_caps *stripping_support;
+ struct virtchnl_vlan_supported_caps *insertion_support;
+
+ stripping_support =
+ &adapter->vlan_v2_caps.offloads.stripping_support;
+ insertion_support =
+ &adapter->vlan_v2_caps.offloads.insertion_support;
+
+ if (stripping_support->outer) {
+ if (stripping_support->outer &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1)
+ rx_ring->flags |=
+ IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ else if (stripping_support->outer &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2_2)
+ rx_ring->flags |=
+ IAVF_RXR_FLAGS_VLAN_TAG_LOC_L2TAG2_2;
+ } else if (stripping_support->inner) {
+ if (stripping_support->inner &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1)
+ rx_ring->flags |=
+ IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ else if (stripping_support->inner &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2_2)
+ rx_ring->flags |=
+ IAVF_RXR_FLAGS_VLAN_TAG_LOC_L2TAG2_2;
+ }
+
+ if (insertion_support->outer) {
+ if (insertion_support->outer &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1)
+ tx_ring->flags |=
+ IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ else if (insertion_support->outer &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2)
+ tx_ring->flags |=
+ IAVF_TXR_FLAGS_VLAN_TAG_LOC_L2TAG2;
+ } else if (insertion_support->inner) {
+ if (insertion_support->inner &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG1)
+ tx_ring->flags |=
+ IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1;
+ else if (insertion_support->inner &
+ VIRTCHNL_VLAN_TAG_LOCATION_L2TAG2)
+ tx_ring->flags |=
+ IAVF_TXR_FLAGS_VLAN_TAG_LOC_L2TAG2;
+ }
+ }
+ }
+}
+
/**
* iavf_alloc_queues - Allocate memory for all rings
* @adapter: board private structure to initialize
@@ -1226,6 +1306,8 @@ static int iavf_alloc_queues(struct iavf_adapter *adapter)
adapter->num_active_queues = num_active_queues;
+ iavf_set_queue_vlan_tag_loc(adapter);
+
return 0;
err_out:
diff --git a/drivers/net/ethernet/intel/iavf/iavf_txrx.c b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
index 42c9f9dc235c..a0b1c18a3273 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_txrx.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
@@ -865,6 +865,9 @@ static void iavf_receive_skb(struct iavf_ring *rx_ring,
if ((rx_ring->netdev->features & NETIF_F_HW_VLAN_CTAG_RX) &&
(vlan_tag & VLAN_VID_MASK))
__vlan_hwaccel_put_tag(skb, htons(ETH_P_8021Q), vlan_tag);
+ else if ((rx_ring->netdev->features & NETIF_F_HW_VLAN_STAG_RX) &&
+ vlan_tag & VLAN_VID_MASK)
+ __vlan_hwaccel_put_tag(skb, htons(ETH_P_8021AD), vlan_tag);
napi_gro_receive(&q_vector->napi, skb);
}
@@ -1468,7 +1471,7 @@ static int iavf_clean_rx_irq(struct iavf_ring *rx_ring, int budget)
struct iavf_rx_buffer *rx_buffer;
union iavf_rx_desc *rx_desc;
unsigned int size;
- u16 vlan_tag;
+ u16 vlan_tag = 0;
u8 rx_ptype;
u64 qword;
@@ -1551,9 +1554,13 @@ static int iavf_clean_rx_irq(struct iavf_ring *rx_ring, int budget)
/* populate checksum, VLAN, and protocol */
iavf_process_skb_fields(rx_ring, rx_desc, skb, rx_ptype);
-
- vlan_tag = (qword & BIT(IAVF_RX_DESC_STATUS_L2TAG1P_SHIFT)) ?
- le16_to_cpu(rx_desc->wb.qword0.lo_dword.l2tag1) : 0;
+ if (qword & BIT(IAVF_RX_DESC_STATUS_L2TAG1P_SHIFT) &&
+ rx_ring->flags & IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1)
+ vlan_tag = le16_to_cpu(rx_desc->wb.qword0.lo_dword.l2tag1);
+ if (rx_desc->wb.qword2.ext_status &
+ cpu_to_le16(BIT(IAVF_RX_DESC_EXT_STATUS_L2TAG2P_SHIFT)) &&
+ rx_ring->flags & IAVF_RXR_FLAGS_VLAN_TAG_LOC_L2TAG2_2)
+ vlan_tag = le16_to_cpu(rx_desc->wb.qword2.l2tag2_2);
iavf_trace(clean_rx_irq_rx, rx_ring, rx_desc, skb);
iavf_receive_skb(rx_ring, skb, vlan_tag);
@@ -1781,46 +1788,29 @@ int iavf_napi_poll(struct napi_struct *napi, int budget)
* Returns error code indicate the frame should be dropped upon error and the
* otherwise returns 0 to indicate the flags has been set properly.
**/
-static inline int iavf_tx_prepare_vlan_flags(struct sk_buff *skb,
- struct iavf_ring *tx_ring,
- u32 *flags)
+static void iavf_tx_prepare_vlan_flags(struct sk_buff *skb,
+ struct iavf_ring *tx_ring, u32 *flags)
{
- __be16 protocol = skb->protocol;
u32 tx_flags = 0;
- if (protocol == htons(ETH_P_8021Q) &&
- !(tx_ring->netdev->features & NETIF_F_HW_VLAN_CTAG_TX)) {
- /* When HW VLAN acceleration is turned off by the user the
- * stack sets the protocol to 8021q so that the driver
- * can take any steps required to support the SW only
- * VLAN handling. In our case the driver doesn't need
- * to take any further steps so just set the protocol
- * to the encapsulated ethertype.
- */
- skb->protocol = vlan_get_protocol(skb);
- goto out;
- }
- /* if we have a HW VLAN tag being added, default to the HW one */
- if (skb_vlan_tag_present(skb)) {
- tx_flags |= skb_vlan_tag_get(skb) << IAVF_TX_FLAGS_VLAN_SHIFT;
- tx_flags |= IAVF_TX_FLAGS_HW_VLAN;
- /* else if it is a SW VLAN, check the next protocol and store the tag */
- } else if (protocol == htons(ETH_P_8021Q)) {
- struct vlan_hdr *vhdr, _vhdr;
-
- vhdr = skb_header_pointer(skb, ETH_HLEN, sizeof(_vhdr), &_vhdr);
- if (!vhdr)
- return -EINVAL;
+ /* stack will only request hardware VLAN insertion offload for protocols
+ * that the driver supports and has enabled
+ */
+ if (!skb_vlan_tag_present(skb))
+ return;
- protocol = vhdr->h_vlan_encapsulated_proto;
- tx_flags |= ntohs(vhdr->h_vlan_TCI) << IAVF_TX_FLAGS_VLAN_SHIFT;
- tx_flags |= IAVF_TX_FLAGS_SW_VLAN;
+ tx_flags |= skb_vlan_tag_get(skb) << IAVF_TX_FLAGS_VLAN_SHIFT;
+ if (tx_ring->flags & IAVF_TXR_FLAGS_VLAN_TAG_LOC_L2TAG2) {
+ tx_flags |= IAVF_TX_FLAGS_HW_OUTER_SINGLE_VLAN;
+ } else if (tx_ring->flags & IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1) {
+ tx_flags |= IAVF_TX_FLAGS_HW_VLAN;
+ } else {
+ dev_dbg(tx_ring->dev, "Unsupported Tx VLAN tag location requested\n");
+ return;
}
-out:
*flags = tx_flags;
- return 0;
}
/**
@@ -2440,8 +2430,13 @@ static netdev_tx_t iavf_xmit_frame_ring(struct sk_buff *skb,
first->gso_segs = 1;
/* prepare the xmit flags */
- if (iavf_tx_prepare_vlan_flags(skb, tx_ring, &tx_flags))
- goto out_drop;
+ iavf_tx_prepare_vlan_flags(skb, tx_ring, &tx_flags);
+ if (tx_flags & IAVF_TX_FLAGS_HW_OUTER_SINGLE_VLAN) {
+ cd_type_cmd_tso_mss |= IAVF_TX_CTX_DESC_IL2TAG2 <<
+ IAVF_TXD_CTX_QW1_CMD_SHIFT;
+ cd_l2tag2 = (tx_flags & IAVF_TX_FLAGS_VLAN_MASK) >>
+ IAVF_TX_FLAGS_VLAN_SHIFT;
+ }
/* obtain protocol of skb */
protocol = vlan_get_protocol(skb);
diff --git a/drivers/net/ethernet/intel/iavf/iavf_txrx.h b/drivers/net/ethernet/intel/iavf/iavf_txrx.h
index e5b9ba42dd00..2624bf6d009e 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_txrx.h
+++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.h
@@ -243,19 +243,20 @@ static inline unsigned int iavf_txd_use_count(unsigned int size)
#define DESC_NEEDED (MAX_SKB_FRAGS + 6)
#define IAVF_MIN_DESC_PENDING 4
-#define IAVF_TX_FLAGS_HW_VLAN BIT(1)
-#define IAVF_TX_FLAGS_SW_VLAN BIT(2)
-#define IAVF_TX_FLAGS_TSO BIT(3)
-#define IAVF_TX_FLAGS_IPV4 BIT(4)
-#define IAVF_TX_FLAGS_IPV6 BIT(5)
-#define IAVF_TX_FLAGS_FCCRC BIT(6)
-#define IAVF_TX_FLAGS_FSO BIT(7)
-#define IAVF_TX_FLAGS_FD_SB BIT(9)
-#define IAVF_TX_FLAGS_VXLAN_TUNNEL BIT(10)
-#define IAVF_TX_FLAGS_VLAN_MASK 0xffff0000
-#define IAVF_TX_FLAGS_VLAN_PRIO_MASK 0xe0000000
-#define IAVF_TX_FLAGS_VLAN_PRIO_SHIFT 29
-#define IAVF_TX_FLAGS_VLAN_SHIFT 16
+#define IAVF_TX_FLAGS_HW_VLAN BIT(1)
+#define IAVF_TX_FLAGS_SW_VLAN BIT(2)
+#define IAVF_TX_FLAGS_TSO BIT(3)
+#define IAVF_TX_FLAGS_IPV4 BIT(4)
+#define IAVF_TX_FLAGS_IPV6 BIT(5)
+#define IAVF_TX_FLAGS_FCCRC BIT(6)
+#define IAVF_TX_FLAGS_FSO BIT(7)
+#define IAVF_TX_FLAGS_FD_SB BIT(9)
+#define IAVF_TX_FLAGS_VXLAN_TUNNEL BIT(10)
+#define IAVF_TX_FLAGS_HW_OUTER_SINGLE_VLAN BIT(11)
+#define IAVF_TX_FLAGS_VLAN_MASK 0xffff0000
+#define IAVF_TX_FLAGS_VLAN_PRIO_MASK 0xe0000000
+#define IAVF_TX_FLAGS_VLAN_PRIO_SHIFT 29
+#define IAVF_TX_FLAGS_VLAN_SHIFT 16
struct iavf_tx_buffer {
struct iavf_tx_desc *next_to_watch;
@@ -362,6 +363,9 @@ struct iavf_ring {
u16 flags;
#define IAVF_TXR_FLAGS_WB_ON_ITR BIT(0)
#define IAVF_RXR_FLAGS_BUILD_SKB_ENABLED BIT(1)
+#define IAVF_TXRX_FLAGS_VLAN_TAG_LOC_L2TAG1 BIT(3)
+#define IAVF_TXR_FLAGS_VLAN_TAG_LOC_L2TAG2 BIT(4)
+#define IAVF_RXR_FLAGS_VLAN_TAG_LOC_L2TAG2_2 BIT(5)
/* stats structs */
struct iavf_queue_stats stats;
diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
index 783d829874bc..3e1b95011146 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
@@ -1964,6 +1964,8 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter,
dev_warn(&adapter->pdev->dev, "failed to acquire crit_lock in %s\n",
__FUNCTION__);
+ iavf_set_queue_vlan_tag_loc(adapter);
+
}
break;
case VIRTCHNL_OP_ENABLE_QUEUES:
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH net-next 5/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 offload enable/disable
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
` (3 preceding siblings ...)
2021-12-17 22:06 ` [PATCH net-next 4/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 hotpath Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-17 22:06 ` [PATCH net-next 6/6] iavf: Restrict maximum VLAN filters for VIRTCHNL_VF_OFFLOAD_VLAN_V2 Tony Nguyen
2021-12-18 12:30 ` [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 patchwork-bot+netdevbpf
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
The new VIRTCHNL_VF_OFFLOAD_VLAN_V2 capability added support that allows
the VF to support 802.1Q and 802.1ad VLAN insertion and stripping if
successfully negotiated via VIRTCHNL_OP_GET_OFFLOAD_VLAN_V2_CAPS.
Multiple changes were needed to support this new functionality.
1. Added new aq_required flags to support any kind of VLAN stripping and
insertion offload requests via virtchnl.
2. Added the new method iavf_set_vlan_offload_features() that's
used during VF initialization, VF reset, and iavf_set_features() to
set the aq_required bits based on the current VLAN offload
configuration of the VF's netdev.
3. Added virtchnl handling for VIRTCHNL_OP_ENABLE_STRIPPING_V2,
VIRTCHNL_OP_DISABLE_STRIPPING_V2, VIRTCHNL_OP_ENABLE_INSERTION_V2,
and VIRTCHNL_OP_ENABLE_INSERTION_V2.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/iavf/iavf.h | 80 ++++---
drivers/net/ethernet/intel/iavf/iavf_main.c | 151 +++++++++++--
.../net/ethernet/intel/iavf/iavf_virtchnl.c | 203 ++++++++++++++++++
3 files changed, 383 insertions(+), 51 deletions(-)
diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h
index 2660d46da1b5..59806d1f7e79 100644
--- a/drivers/net/ethernet/intel/iavf/iavf.h
+++ b/drivers/net/ethernet/intel/iavf/iavf.h
@@ -287,39 +287,47 @@ struct iavf_adapter {
/* duplicates for common code */
#define IAVF_FLAG_DCB_ENABLED 0
/* flags for admin queue service task */
- u32 aq_required;
-#define IAVF_FLAG_AQ_ENABLE_QUEUES BIT(0)
-#define IAVF_FLAG_AQ_DISABLE_QUEUES BIT(1)
-#define IAVF_FLAG_AQ_ADD_MAC_FILTER BIT(2)
-#define IAVF_FLAG_AQ_ADD_VLAN_FILTER BIT(3)
-#define IAVF_FLAG_AQ_DEL_MAC_FILTER BIT(4)
-#define IAVF_FLAG_AQ_DEL_VLAN_FILTER BIT(5)
-#define IAVF_FLAG_AQ_CONFIGURE_QUEUES BIT(6)
-#define IAVF_FLAG_AQ_MAP_VECTORS BIT(7)
-#define IAVF_FLAG_AQ_HANDLE_RESET BIT(8)
-#define IAVF_FLAG_AQ_CONFIGURE_RSS BIT(9) /* direct AQ config */
-#define IAVF_FLAG_AQ_GET_CONFIG BIT(10)
+ u64 aq_required;
+#define IAVF_FLAG_AQ_ENABLE_QUEUES BIT_ULL(0)
+#define IAVF_FLAG_AQ_DISABLE_QUEUES BIT_ULL(1)
+#define IAVF_FLAG_AQ_ADD_MAC_FILTER BIT_ULL(2)
+#define IAVF_FLAG_AQ_ADD_VLAN_FILTER BIT_ULL(3)
+#define IAVF_FLAG_AQ_DEL_MAC_FILTER BIT_ULL(4)
+#define IAVF_FLAG_AQ_DEL_VLAN_FILTER BIT_ULL(5)
+#define IAVF_FLAG_AQ_CONFIGURE_QUEUES BIT_ULL(6)
+#define IAVF_FLAG_AQ_MAP_VECTORS BIT_ULL(7)
+#define IAVF_FLAG_AQ_HANDLE_RESET BIT_ULL(8)
+#define IAVF_FLAG_AQ_CONFIGURE_RSS BIT_ULL(9) /* direct AQ config */
+#define IAVF_FLAG_AQ_GET_CONFIG BIT_ULL(10)
/* Newer style, RSS done by the PF so we can ignore hardware vagaries. */
-#define IAVF_FLAG_AQ_GET_HENA BIT(11)
-#define IAVF_FLAG_AQ_SET_HENA BIT(12)
-#define IAVF_FLAG_AQ_SET_RSS_KEY BIT(13)
-#define IAVF_FLAG_AQ_SET_RSS_LUT BIT(14)
-#define IAVF_FLAG_AQ_REQUEST_PROMISC BIT(15)
-#define IAVF_FLAG_AQ_RELEASE_PROMISC BIT(16)
-#define IAVF_FLAG_AQ_REQUEST_ALLMULTI BIT(17)
-#define IAVF_FLAG_AQ_RELEASE_ALLMULTI BIT(18)
-#define IAVF_FLAG_AQ_ENABLE_VLAN_STRIPPING BIT(19)
-#define IAVF_FLAG_AQ_DISABLE_VLAN_STRIPPING BIT(20)
-#define IAVF_FLAG_AQ_ENABLE_CHANNELS BIT(21)
-#define IAVF_FLAG_AQ_DISABLE_CHANNELS BIT(22)
-#define IAVF_FLAG_AQ_ADD_CLOUD_FILTER BIT(23)
-#define IAVF_FLAG_AQ_DEL_CLOUD_FILTER BIT(24)
-#define IAVF_FLAG_AQ_ADD_FDIR_FILTER BIT(25)
-#define IAVF_FLAG_AQ_DEL_FDIR_FILTER BIT(26)
-#define IAVF_FLAG_AQ_ADD_ADV_RSS_CFG BIT(27)
-#define IAVF_FLAG_AQ_DEL_ADV_RSS_CFG BIT(28)
-#define IAVF_FLAG_AQ_REQUEST_STATS BIT(29)
-#define IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS BIT(30)
+#define IAVF_FLAG_AQ_GET_HENA BIT_ULL(11)
+#define IAVF_FLAG_AQ_SET_HENA BIT_ULL(12)
+#define IAVF_FLAG_AQ_SET_RSS_KEY BIT_ULL(13)
+#define IAVF_FLAG_AQ_SET_RSS_LUT BIT_ULL(14)
+#define IAVF_FLAG_AQ_REQUEST_PROMISC BIT_ULL(15)
+#define IAVF_FLAG_AQ_RELEASE_PROMISC BIT_ULL(16)
+#define IAVF_FLAG_AQ_REQUEST_ALLMULTI BIT_ULL(17)
+#define IAVF_FLAG_AQ_RELEASE_ALLMULTI BIT_ULL(18)
+#define IAVF_FLAG_AQ_ENABLE_VLAN_STRIPPING BIT_ULL(19)
+#define IAVF_FLAG_AQ_DISABLE_VLAN_STRIPPING BIT_ULL(20)
+#define IAVF_FLAG_AQ_ENABLE_CHANNELS BIT_ULL(21)
+#define IAVF_FLAG_AQ_DISABLE_CHANNELS BIT_ULL(22)
+#define IAVF_FLAG_AQ_ADD_CLOUD_FILTER BIT_ULL(23)
+#define IAVF_FLAG_AQ_DEL_CLOUD_FILTER BIT_ULL(24)
+#define IAVF_FLAG_AQ_ADD_FDIR_FILTER BIT_ULL(25)
+#define IAVF_FLAG_AQ_DEL_FDIR_FILTER BIT_ULL(26)
+#define IAVF_FLAG_AQ_ADD_ADV_RSS_CFG BIT_ULL(27)
+#define IAVF_FLAG_AQ_DEL_ADV_RSS_CFG BIT_ULL(28)
+#define IAVF_FLAG_AQ_REQUEST_STATS BIT_ULL(29)
+#define IAVF_FLAG_AQ_GET_OFFLOAD_VLAN_V2_CAPS BIT_ULL(30)
+#define IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_STRIPPING BIT_ULL(31)
+#define IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_STRIPPING BIT_ULL(32)
+#define IAVF_FLAG_AQ_ENABLE_STAG_VLAN_STRIPPING BIT_ULL(33)
+#define IAVF_FLAG_AQ_DISABLE_STAG_VLAN_STRIPPING BIT_ULL(34)
+#define IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_INSERTION BIT_ULL(35)
+#define IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_INSERTION BIT_ULL(36)
+#define IAVF_FLAG_AQ_ENABLE_STAG_VLAN_INSERTION BIT_ULL(37)
+#define IAVF_FLAG_AQ_DISABLE_STAG_VLAN_INSERTION BIT_ULL(38)
/* OS defined structs */
struct net_device *netdev;
@@ -524,6 +532,14 @@ void iavf_enable_channels(struct iavf_adapter *adapter);
void iavf_disable_channels(struct iavf_adapter *adapter);
void iavf_add_cloud_filter(struct iavf_adapter *adapter);
void iavf_del_cloud_filter(struct iavf_adapter *adapter);
+void iavf_enable_vlan_stripping_v2(struct iavf_adapter *adapter, u16 tpid);
+void iavf_disable_vlan_stripping_v2(struct iavf_adapter *adapter, u16 tpid);
+void iavf_enable_vlan_insertion_v2(struct iavf_adapter *adapter, u16 tpid);
+void iavf_disable_vlan_insertion_v2(struct iavf_adapter *adapter, u16 tpid);
+void
+iavf_set_vlan_offload_features(struct iavf_adapter *adapter,
+ netdev_features_t prev_features,
+ netdev_features_t features);
void iavf_add_fdir_filter(struct iavf_adapter *adapter);
void iavf_del_fdir_filter(struct iavf_adapter *adapter);
void iavf_add_adv_rss_cfg(struct iavf_adapter *adapter);
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index cd9c7484ec99..ab59c9235d03 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -1815,6 +1815,39 @@ static int iavf_process_aq_command(struct iavf_adapter *adapter)
iavf_del_adv_rss_cfg(adapter);
return 0;
}
+ if (adapter->aq_required & IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_STRIPPING) {
+ iavf_disable_vlan_stripping_v2(adapter, ETH_P_8021Q);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_DISABLE_STAG_VLAN_STRIPPING) {
+ iavf_disable_vlan_stripping_v2(adapter, ETH_P_8021AD);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_STRIPPING) {
+ iavf_enable_vlan_stripping_v2(adapter, ETH_P_8021Q);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_ENABLE_STAG_VLAN_STRIPPING) {
+ iavf_enable_vlan_stripping_v2(adapter, ETH_P_8021AD);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_INSERTION) {
+ iavf_disable_vlan_insertion_v2(adapter, ETH_P_8021Q);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_DISABLE_STAG_VLAN_INSERTION) {
+ iavf_disable_vlan_insertion_v2(adapter, ETH_P_8021AD);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_INSERTION) {
+ iavf_enable_vlan_insertion_v2(adapter, ETH_P_8021Q);
+ return 0;
+ }
+ if (adapter->aq_required & IAVF_FLAG_AQ_ENABLE_STAG_VLAN_INSERTION) {
+ iavf_enable_vlan_insertion_v2(adapter, ETH_P_8021AD);
+ return 0;
+ }
+
if (adapter->aq_required & IAVF_FLAG_AQ_REQUEST_STATS) {
iavf_request_stats(adapter);
return 0;
@@ -1823,6 +1856,91 @@ static int iavf_process_aq_command(struct iavf_adapter *adapter)
return -EAGAIN;
}
+/**
+ * iavf_set_vlan_offload_features - set VLAN offload configuration
+ * @adapter: board private structure
+ * @prev_features: previous features used for comparison
+ * @features: updated features used for configuration
+ *
+ * Set the aq_required bit(s) based on the requested features passed in to
+ * configure VLAN stripping and/or VLAN insertion if supported. Also, schedule
+ * the watchdog if any changes are requested to expedite the request via
+ * virtchnl.
+ **/
+void
+iavf_set_vlan_offload_features(struct iavf_adapter *adapter,
+ netdev_features_t prev_features,
+ netdev_features_t features)
+{
+ bool enable_stripping = true, enable_insertion = true;
+ u16 vlan_ethertype = 0;
+ u64 aq_required = 0;
+
+ /* keep cases separate because one ethertype for offloads can be
+ * disabled at the same time as another is disabled, so check for an
+ * enabled ethertype first, then check for disabled. Default to
+ * ETH_P_8021Q so an ethertype is specified if disabling insertion and
+ * stripping.
+ */
+ if (features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_STAG_TX))
+ vlan_ethertype = ETH_P_8021AD;
+ else if (features & (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX))
+ vlan_ethertype = ETH_P_8021Q;
+ else if (prev_features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_STAG_TX))
+ vlan_ethertype = ETH_P_8021AD;
+ else if (prev_features & (NETIF_F_HW_VLAN_CTAG_RX | NETIF_F_HW_VLAN_CTAG_TX))
+ vlan_ethertype = ETH_P_8021Q;
+ else
+ vlan_ethertype = ETH_P_8021Q;
+
+ if (!(features & (NETIF_F_HW_VLAN_STAG_RX | NETIF_F_HW_VLAN_CTAG_RX)))
+ enable_stripping = false;
+ if (!(features & (NETIF_F_HW_VLAN_STAG_TX | NETIF_F_HW_VLAN_CTAG_TX)))
+ enable_insertion = false;
+
+ if (VLAN_ALLOWED(adapter)) {
+ /* VIRTCHNL_VF_OFFLOAD_VLAN only has support for toggling VLAN
+ * stripping via virtchnl. VLAN insertion can be toggled on the
+ * netdev, but it doesn't require a virtchnl message
+ */
+ if (enable_stripping)
+ aq_required |= IAVF_FLAG_AQ_ENABLE_VLAN_STRIPPING;
+ else
+ aq_required |= IAVF_FLAG_AQ_DISABLE_VLAN_STRIPPING;
+
+ } else if (VLAN_V2_ALLOWED(adapter)) {
+ switch (vlan_ethertype) {
+ case ETH_P_8021Q:
+ if (enable_stripping)
+ aq_required |= IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_STRIPPING;
+ else
+ aq_required |= IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_STRIPPING;
+
+ if (enable_insertion)
+ aq_required |= IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_INSERTION;
+ else
+ aq_required |= IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_INSERTION;
+ break;
+ case ETH_P_8021AD:
+ if (enable_stripping)
+ aq_required |= IAVF_FLAG_AQ_ENABLE_STAG_VLAN_STRIPPING;
+ else
+ aq_required |= IAVF_FLAG_AQ_DISABLE_STAG_VLAN_STRIPPING;
+
+ if (enable_insertion)
+ aq_required |= IAVF_FLAG_AQ_ENABLE_STAG_VLAN_INSERTION;
+ else
+ aq_required |= IAVF_FLAG_AQ_DISABLE_STAG_VLAN_INSERTION;
+ break;
+ }
+ }
+
+ if (aq_required) {
+ adapter->aq_required |= aq_required;
+ mod_delayed_work(iavf_wq, &adapter->watchdog_task, 0);
+ }
+}
+
/**
* iavf_startup - first step of driver startup
* @adapter: board private structure
@@ -2179,6 +2297,10 @@ static void iavf_init_config_adapter(struct iavf_adapter *adapter)
else
iavf_init_rss(adapter);
+ if (VLAN_V2_ALLOWED(adapter))
+ /* request initial VLAN offload settings */
+ iavf_set_vlan_offload_features(adapter, 0, netdev->features);
+
return;
err_mem:
iavf_free_rss(adapter);
@@ -3684,6 +3806,11 @@ static int iavf_change_mtu(struct net_device *netdev, int new_mtu)
return 0;
}
+#define NETIF_VLAN_OFFLOAD_FEATURES (NETIF_F_HW_VLAN_CTAG_RX | \
+ NETIF_F_HW_VLAN_CTAG_TX | \
+ NETIF_F_HW_VLAN_STAG_RX | \
+ NETIF_F_HW_VLAN_STAG_TX)
+
/**
* iavf_set_features - set the netdev feature flags
* @netdev: ptr to the netdev being adjusted
@@ -3695,25 +3822,11 @@ static int iavf_set_features(struct net_device *netdev,
{
struct iavf_adapter *adapter = netdev_priv(netdev);
- /* Don't allow enabling VLAN features when adapter is not capable
- * of VLAN offload/filtering
- */
- if (!VLAN_ALLOWED(adapter)) {
- netdev->hw_features &= ~(NETIF_F_HW_VLAN_CTAG_RX |
- NETIF_F_HW_VLAN_CTAG_TX |
- NETIF_F_HW_VLAN_CTAG_FILTER);
- if (features & (NETIF_F_HW_VLAN_CTAG_RX |
- NETIF_F_HW_VLAN_CTAG_TX |
- NETIF_F_HW_VLAN_CTAG_FILTER))
- return -EINVAL;
- } else if ((netdev->features ^ features) & NETIF_F_HW_VLAN_CTAG_RX) {
- if (features & NETIF_F_HW_VLAN_CTAG_RX)
- adapter->aq_required |=
- IAVF_FLAG_AQ_ENABLE_VLAN_STRIPPING;
- else
- adapter->aq_required |=
- IAVF_FLAG_AQ_DISABLE_VLAN_STRIPPING;
- }
+ /* trigger update on any VLAN feature change */
+ if ((netdev->features & NETIF_VLAN_OFFLOAD_FEATURES) ^
+ (features & NETIF_VLAN_OFFLOAD_FEATURES))
+ iavf_set_vlan_offload_features(adapter, netdev->features,
+ features);
return 0;
}
diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
index 3e1b95011146..5ee1d118fd30 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
@@ -1124,6 +1124,204 @@ void iavf_disable_vlan_stripping(struct iavf_adapter *adapter)
iavf_send_pf_msg(adapter, VIRTCHNL_OP_DISABLE_VLAN_STRIPPING, NULL, 0);
}
+/**
+ * iavf_tpid_to_vc_ethertype - transform from VLAN TPID to virtchnl ethertype
+ * @tpid: VLAN TPID (i.e. 0x8100, 0x88a8, etc.)
+ */
+static u32 iavf_tpid_to_vc_ethertype(u16 tpid)
+{
+ switch (tpid) {
+ case ETH_P_8021Q:
+ return VIRTCHNL_VLAN_ETHERTYPE_8100;
+ case ETH_P_8021AD:
+ return VIRTCHNL_VLAN_ETHERTYPE_88A8;
+ }
+
+ return 0;
+}
+
+/**
+ * iavf_set_vc_offload_ethertype - set virtchnl ethertype for offload message
+ * @adapter: adapter structure
+ * @msg: message structure used for updating offloads over virtchnl to update
+ * @tpid: VLAN TPID (i.e. 0x8100, 0x88a8, etc.)
+ * @offload_op: opcode used to determine which support structure to check
+ */
+static int
+iavf_set_vc_offload_ethertype(struct iavf_adapter *adapter,
+ struct virtchnl_vlan_setting *msg, u16 tpid,
+ enum virtchnl_ops offload_op)
+{
+ struct virtchnl_vlan_supported_caps *offload_support;
+ u16 vc_ethertype = iavf_tpid_to_vc_ethertype(tpid);
+
+ /* reference the correct offload support structure */
+ switch (offload_op) {
+ case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2:
+ case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2:
+ offload_support =
+ &adapter->vlan_v2_caps.offloads.stripping_support;
+ break;
+ case VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2:
+ case VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2:
+ offload_support =
+ &adapter->vlan_v2_caps.offloads.insertion_support;
+ break;
+ default:
+ dev_err(&adapter->pdev->dev, "Invalid opcode %d for setting virtchnl ethertype to enable/disable VLAN offloads\n",
+ offload_op);
+ return -EINVAL;
+ }
+
+ /* make sure ethertype is supported */
+ if (offload_support->outer & vc_ethertype &&
+ offload_support->outer & VIRTCHNL_VLAN_TOGGLE) {
+ msg->outer_ethertype_setting = vc_ethertype;
+ } else if (offload_support->inner & vc_ethertype &&
+ offload_support->inner & VIRTCHNL_VLAN_TOGGLE) {
+ msg->inner_ethertype_setting = vc_ethertype;
+ } else {
+ dev_dbg(&adapter->pdev->dev, "opcode %d unsupported for VLAN TPID 0x%04x\n",
+ offload_op, tpid);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+/**
+ * iavf_clear_offload_v2_aq_required - clear AQ required bit for offload request
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID
+ * @offload_op: opcode used to determine which AQ required bit to clear
+ */
+static void
+iavf_clear_offload_v2_aq_required(struct iavf_adapter *adapter, u16 tpid,
+ enum virtchnl_ops offload_op)
+{
+ switch (offload_op) {
+ case VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2:
+ if (tpid == ETH_P_8021Q)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_STRIPPING;
+ else if (tpid == ETH_P_8021AD)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_ENABLE_STAG_VLAN_STRIPPING;
+ break;
+ case VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2:
+ if (tpid == ETH_P_8021Q)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_STRIPPING;
+ else if (tpid == ETH_P_8021AD)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_DISABLE_STAG_VLAN_STRIPPING;
+ break;
+ case VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2:
+ if (tpid == ETH_P_8021Q)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_ENABLE_CTAG_VLAN_INSERTION;
+ else if (tpid == ETH_P_8021AD)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_ENABLE_STAG_VLAN_INSERTION;
+ break;
+ case VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2:
+ if (tpid == ETH_P_8021Q)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_DISABLE_CTAG_VLAN_INSERTION;
+ else if (tpid == ETH_P_8021AD)
+ adapter->aq_required &=
+ ~IAVF_FLAG_AQ_DISABLE_STAG_VLAN_INSERTION;
+ break;
+ default:
+ dev_err(&adapter->pdev->dev, "Unsupported opcode %d specified for clearing aq_required bits for VIRTCHNL_VF_OFFLOAD_VLAN_V2 offload request\n",
+ offload_op);
+ }
+}
+
+/**
+ * iavf_send_vlan_offload_v2 - send offload enable/disable over virtchnl
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID used for the command (i.e. 0x8100 or 0x88a8)
+ * @offload_op: offload_op used to make the request over virtchnl
+ */
+static void
+iavf_send_vlan_offload_v2(struct iavf_adapter *adapter, u16 tpid,
+ enum virtchnl_ops offload_op)
+{
+ struct virtchnl_vlan_setting *msg;
+ int len = sizeof(*msg);
+
+ if (adapter->current_op != VIRTCHNL_OP_UNKNOWN) {
+ /* bail because we already have a command pending */
+ dev_err(&adapter->pdev->dev, "Cannot send %d, command %d pending\n",
+ offload_op, adapter->current_op);
+ return;
+ }
+
+ adapter->current_op = offload_op;
+
+ msg = kzalloc(len, GFP_KERNEL);
+ if (!msg)
+ return;
+
+ msg->vport_id = adapter->vsi_res->vsi_id;
+
+ /* always clear to prevent unsupported and endless requests */
+ iavf_clear_offload_v2_aq_required(adapter, tpid, offload_op);
+
+ /* only send valid offload requests */
+ if (!iavf_set_vc_offload_ethertype(adapter, msg, tpid, offload_op))
+ iavf_send_pf_msg(adapter, offload_op, (u8 *)msg, len);
+ else
+ adapter->current_op = VIRTCHNL_OP_UNKNOWN;
+
+ kfree(msg);
+}
+
+/**
+ * iavf_enable_vlan_stripping_v2 - enable VLAN stripping
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID used to enable VLAN stripping
+ */
+void iavf_enable_vlan_stripping_v2(struct iavf_adapter *adapter, u16 tpid)
+{
+ iavf_send_vlan_offload_v2(adapter, tpid,
+ VIRTCHNL_OP_ENABLE_VLAN_STRIPPING_V2);
+}
+
+/**
+ * iavf_disable_vlan_stripping_v2 - disable VLAN stripping
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID used to disable VLAN stripping
+ */
+void iavf_disable_vlan_stripping_v2(struct iavf_adapter *adapter, u16 tpid)
+{
+ iavf_send_vlan_offload_v2(adapter, tpid,
+ VIRTCHNL_OP_DISABLE_VLAN_STRIPPING_V2);
+}
+
+/**
+ * iavf_enable_vlan_insertion_v2 - enable VLAN insertion
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID used to enable VLAN insertion
+ */
+void iavf_enable_vlan_insertion_v2(struct iavf_adapter *adapter, u16 tpid)
+{
+ iavf_send_vlan_offload_v2(adapter, tpid,
+ VIRTCHNL_OP_ENABLE_VLAN_INSERTION_V2);
+}
+
+/**
+ * iavf_disable_vlan_insertion_v2 - disable VLAN insertion
+ * @adapter: adapter structure
+ * @tpid: VLAN TPID used to disable VLAN insertion
+ */
+void iavf_disable_vlan_insertion_v2(struct iavf_adapter *adapter, u16 tpid)
+{
+ iavf_send_vlan_offload_v2(adapter, tpid,
+ VIRTCHNL_OP_DISABLE_VLAN_INSERTION_V2);
+}
+
#define IAVF_MAX_SPEED_STRLEN 13
/**
@@ -1964,6 +2162,11 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter,
dev_warn(&adapter->pdev->dev, "failed to acquire crit_lock in %s\n",
__FUNCTION__);
+ /* Request VLAN offload settings */
+ if (VLAN_V2_ALLOWED(adapter))
+ iavf_set_vlan_offload_features(adapter, 0,
+ netdev->features);
+
iavf_set_queue_vlan_tag_loc(adapter);
}
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* [PATCH net-next 6/6] iavf: Restrict maximum VLAN filters for VIRTCHNL_VF_OFFLOAD_VLAN_V2
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
` (4 preceding siblings ...)
2021-12-17 22:06 ` [PATCH net-next 5/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 offload enable/disable Tony Nguyen
@ 2021-12-17 22:06 ` Tony Nguyen
2021-12-18 12:30 ` [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 patchwork-bot+netdevbpf
6 siblings, 0 replies; 8+ messages in thread
From: Tony Nguyen @ 2021-12-17 22:06 UTC (permalink / raw)
To: davem, kuba
Cc: Brett Creeley, netdev, anthony.l.nguyen, sassmann,
Konrad Jankowski
From: Brett Creeley <brett.creeley@intel.com>
For VIRTCHNL_VF_OFFLOAD_VLAN, PF's would limit the number of VLAN
filters a VF was allowed to add. However, by the time the opcode failed,
the VLAN netdev had already been added. VIRTCHNL_VF_OFFLOAD_VLAN_V2
added the ability for a PF to tell the VF how many VLAN filters it's
allowed to add. Make changes to support that functionality.
Signed-off-by: Brett Creeley <brett.creeley@intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
---
drivers/net/ethernet/intel/iavf/iavf_main.c | 50 +++++++++++++++++++++
1 file changed, 50 insertions(+)
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index ab59c9235d03..0fbac51b7cca 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -731,6 +731,50 @@ static void iavf_restore_filters(struct iavf_adapter *adapter)
iavf_add_vlan(adapter, IAVF_VLAN(vid, ETH_P_8021AD));
}
+/**
+ * iavf_get_num_vlans_added - get number of VLANs added
+ * @adapter: board private structure
+ */
+static u16 iavf_get_num_vlans_added(struct iavf_adapter *adapter)
+{
+ return bitmap_weight(adapter->vsi.active_cvlans, VLAN_N_VID) +
+ bitmap_weight(adapter->vsi.active_svlans, VLAN_N_VID);
+}
+
+/**
+ * iavf_get_max_vlans_allowed - get maximum VLANs allowed for this VF
+ * @adapter: board private structure
+ *
+ * This depends on the negotiated VLAN capability. For VIRTCHNL_VF_OFFLOAD_VLAN,
+ * do not impose a limit as that maintains current behavior and for
+ * VIRTCHNL_VF_OFFLOAD_VLAN_V2, use the maximum allowed sent from the PF.
+ **/
+static u16 iavf_get_max_vlans_allowed(struct iavf_adapter *adapter)
+{
+ /* don't impose any limit for VIRTCHNL_VF_OFFLOAD_VLAN since there has
+ * never been a limit on the VF driver side
+ */
+ if (VLAN_ALLOWED(adapter))
+ return VLAN_N_VID;
+ else if (VLAN_V2_ALLOWED(adapter))
+ return adapter->vlan_v2_caps.filtering.max_filters;
+
+ return 0;
+}
+
+/**
+ * iavf_max_vlans_added - check if maximum VLANs allowed already exist
+ * @adapter: board private structure
+ **/
+static bool iavf_max_vlans_added(struct iavf_adapter *adapter)
+{
+ if (iavf_get_num_vlans_added(adapter) <
+ iavf_get_max_vlans_allowed(adapter))
+ return false;
+
+ return true;
+}
+
/**
* iavf_vlan_rx_add_vid - Add a VLAN filter to a device
* @netdev: network device struct
@@ -745,6 +789,12 @@ static int iavf_vlan_rx_add_vid(struct net_device *netdev,
if (!VLAN_FILTERING_ALLOWED(adapter))
return -EIO;
+ if (iavf_max_vlans_added(adapter)) {
+ netdev_err(netdev, "Max allowed VLAN filters %u. Remove existing VLANs or disable filtering via Ethtool if supported.\n",
+ iavf_get_max_vlans_allowed(adapter));
+ return -EIO;
+ }
+
if (!iavf_add_vlan(adapter, IAVF_VLAN(vid, be16_to_cpu(proto))))
return -ENOMEM;
--
2.31.1
^ permalink raw reply related [flat|nested] 8+ messages in thread* Re: [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17
2021-12-17 22:06 [PATCH net-next 0/6][pull request] 40GbE Intel Wired LAN Driver Updates 2021-12-17 Tony Nguyen
` (5 preceding siblings ...)
2021-12-17 22:06 ` [PATCH net-next 6/6] iavf: Restrict maximum VLAN filters for VIRTCHNL_VF_OFFLOAD_VLAN_V2 Tony Nguyen
@ 2021-12-18 12:30 ` patchwork-bot+netdevbpf
6 siblings, 0 replies; 8+ messages in thread
From: patchwork-bot+netdevbpf @ 2021-12-18 12:30 UTC (permalink / raw)
To: Tony Nguyen; +Cc: davem, kuba, netdev, sassmann
Hello:
This series was applied to netdev/net-next.git (master)
by Tony Nguyen <anthony.l.nguyen@intel.com>:
On Fri, 17 Dec 2021 14:06:41 -0800 you wrote:
> Brett Creeley says:
>
> This patch series adds support in the iavf driver for communicating and
> using VIRTCHNL_VF_OFFLOAD_VLAN_V2. The current VIRTCHNL_VF_OFFLOAD_VLAN
> is very limited and covers all 802.1Q VLAN offloads and filtering with
> no granularity.
>
> [...]
Here is the summary with links:
- [net-next,1/6] virtchnl: Add support for new VLAN capabilities
https://git.kernel.org/netdev/net-next/c/bd0b536dc2e1
- [net-next,2/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 negotiation
https://git.kernel.org/netdev/net-next/c/209f2f9c7181
- [net-next,3/6] iavf: Add support VIRTCHNL_VF_OFFLOAD_VLAN_V2 during netdev config
https://git.kernel.org/netdev/net-next/c/48ccc43ecf10
- [net-next,4/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 hotpath
https://git.kernel.org/netdev/net-next/c/ccd219d2ea13
- [net-next,5/6] iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 offload enable/disable
https://git.kernel.org/netdev/net-next/c/8afadd1cd8ba
- [net-next,6/6] iavf: Restrict maximum VLAN filters for VIRTCHNL_VF_OFFLOAD_VLAN_V2
https://git.kernel.org/netdev/net-next/c/92fc50859872
You are awesome, thank you!
--
Deet-doot-dot, I am a bot.
https://korg.docs.kernel.org/patchwork/pwbot.html
^ permalink raw reply [flat|nested] 8+ messages in thread