idpf: configure resources for RX queues

Similar to the TX, RX also supports both single and split queue models.
In single queue model, the same descriptor queue is used by SW to post
buffer descriptors to HW and by HW to post completed descriptors
to SW. In split queue model, "RX buffer queues" are used to pass
descriptor buffers from SW to HW whereas "RX queues" are used to
post the descriptor completions i.e. descriptors that point to
completed buffers, from HW to SW. "RX queue group" is a set of
RX queues grouped together and will be serviced by a "RX buffer queue
group". IDPF supports 2 buffer queues i.e. large buffer (4KB) queue
and small buffer (2KB) queue per buffer queue group. HW uses large
buffers for 'hardware gro' feature and also if the packet size is
more than 2KB, if not 2KB buffers are used.

Add all the resources required for the RX queues initialization.
Allocate memory for the RX queue and RX buffer queue groups. Initialize
the software maintained refill queues for buffer management algorithm.

Same like the TX queues, initialize the queue parameters for the RX
queues and send the config RX queue virtchnl message to the device
Control Plane.

Signed-off-by: Alan Brady <[email protected]>
Co-developed-by: Alice Michael <[email protected]>
Signed-off-by: Alice Michael <[email protected]>
Co-developed-by: Joshua Hay <[email protected]>
Signed-off-by: Joshua Hay <[email protected]>
Co-developed-by: Madhu Chittim <[email protected]>
Signed-off-by: Madhu Chittim <[email protected]>
Co-developed-by: Phani Burra <[email protected]>
Signed-off-by: Phani Burra <[email protected]>
Reviewed-by: Sridhar Samudrala <[email protected]>
Reviewed-by: Willem de Bruijn <[email protected]>
Co-developed-by: Pavan Kumar Linga <[email protected]>
Signed-off-by: Pavan Kumar Linga <[email protected]>
Signed-off-by: Tony Nguyen <[email protected]>

Author: atbrady-intel

drivers/net/ethernet/intel/idpf/Makefile

@@ -11,6 +11,7 @@ idpf-y := \
 	idpf_dev.o		\
 	idpf_lib.o		\
 	idpf_main.o		\
+	idpf_singleq_txrx.o	\
 	idpf_txrx.o		\
 	idpf_virtchnl.o 	\
 	idpf_vf_dev.o

drivers/net/ethernet/intel/idpf/idpf.h

@@ -208,10 +208,20 @@ struct idpf_dev_ops {
 	STATE(IDPF_VC_DESTROY_VPORT_ERR)	\
 	STATE(IDPF_VC_CONFIG_TXQ)		\
 	STATE(IDPF_VC_CONFIG_TXQ_ERR)		\
+	STATE(IDPF_VC_CONFIG_RXQ)		\
+	STATE(IDPF_VC_CONFIG_RXQ_ERR)		\
 	STATE(IDPF_VC_ALLOC_VECTORS)		\
 	STATE(IDPF_VC_ALLOC_VECTORS_ERR)	\
 	STATE(IDPF_VC_DEALLOC_VECTORS)		\
 	STATE(IDPF_VC_DEALLOC_VECTORS_ERR)	\
+	STATE(IDPF_VC_GET_RSS_LUT)		\
+	STATE(IDPF_VC_GET_RSS_LUT_ERR)		\
+	STATE(IDPF_VC_SET_RSS_LUT)		\
+	STATE(IDPF_VC_SET_RSS_LUT_ERR)		\
+	STATE(IDPF_VC_GET_RSS_KEY)		\
+	STATE(IDPF_VC_GET_RSS_KEY_ERR)		\
+	STATE(IDPF_VC_SET_RSS_KEY)		\
+	STATE(IDPF_VC_SET_RSS_KEY_ERR)		\
 	STATE(IDPF_VC_ADD_MAC_ADDR)		\
 	STATE(IDPF_VC_ADD_MAC_ADDR_ERR)		\
 	STATE(IDPF_VC_DEL_MAC_ADDR)		\
@@ -248,6 +258,8 @@ extern const char * const idpf_vport_vc_state_str[];
  * @bufq_desc_count: Buffer queue descriptor count
  * @bufq_size: Size of buffers in ring (e.g. 2K, 4K, etc)
  * @num_rxq_grp: Number of RX queues in a group
+ * @rxq_grps: Total number of RX groups. Number of groups * number of RX per
+ *	      group will yield total number of RX queues.
  * @rxq_model: Splitq queue or single queue queuing model
  * @rx_ptype_lkup: Lookup table for ptypes on RX
  * @adapter: back pointer to associated adapter
@@ -284,6 +296,7 @@ struct idpf_vport {
 	u32 bufq_desc_count[IDPF_MAX_BUFQS_PER_RXQ_GRP];
 	u32 bufq_size[IDPF_MAX_BUFQS_PER_RXQ_GRP];
 	u16 num_rxq_grp;
+	struct idpf_rxq_group *rxq_grps;
 	u32 rxq_model;
 	struct idpf_rx_ptype_decoded rx_ptype_lkup[IDPF_RX_MAX_PTYPE];
 
@@ -307,9 +320,26 @@ struct idpf_vport {
 	struct mutex vc_buf_lock;
 };
 
+/**
+ * struct idpf_rss_data - Associated RSS data
+ * @rss_key_size: Size of RSS hash key
+ * @rss_key: RSS hash key
+ * @rss_lut_size: Size of RSS lookup table
+ * @rss_lut: RSS lookup table
+ * @cached_lut: Used to restore previously init RSS lut
+ */
+struct idpf_rss_data {
+	u16 rss_key_size;
+	u8 *rss_key;
+	u16 rss_lut_size;
+	u32 *rss_lut;
+	u32 *cached_lut;
+};
+
 /**
  * struct idpf_vport_user_config_data - User defined configuration values for
  *					each vport.
+ * @rss_data: See struct idpf_rss_data
  * @num_req_tx_qs: Number of user requested TX queues through ethtool
  * @num_req_rx_qs: Number of user requested RX queues through ethtool
  * @num_req_txq_desc: Number of user requested TX queue descriptors through
@@ -321,6 +351,7 @@ struct idpf_vport {
  * Used to restore configuration after a reset as the vport will get wiped.
  */
 struct idpf_vport_user_config_data {
+	struct idpf_rss_data rss_data;
 	u16 num_req_tx_qs;
 	u16 num_req_rx_qs;
 	u32 num_req_txq_desc;
@@ -667,6 +698,19 @@ static inline struct idpf_vport *idpf_netdev_to_vport(struct net_device *netdev)
 	return np->vport;
 }
 
+/**
+ * idpf_is_feature_ena - Determine if a particular feature is enabled
+ * @vport: Vport to check
+ * @feature: Netdev flag to check
+ *
+ * Returns true or false if a particular feature is enabled.
+ */
+static inline bool idpf_is_feature_ena(const struct idpf_vport *vport,
+				       netdev_features_t feature)
+{
+	return vport->netdev->features & feature;
+}
+
 /**
  * idpf_vport_ctrl_lock - Acquire the vport control lock
  * @netdev: Network interface device structure
@@ -706,6 +750,8 @@ int idpf_intr_req(struct idpf_adapter *adapter);
 void idpf_intr_rel(struct idpf_adapter *adapter);
 int idpf_send_destroy_vport_msg(struct idpf_vport *vport);
 int idpf_send_get_rx_ptype_msg(struct idpf_vport *vport);
+int idpf_send_get_set_rss_key_msg(struct idpf_vport *vport, bool get);
+int idpf_send_get_set_rss_lut_msg(struct idpf_vport *vport, bool get);
 int idpf_send_dealloc_vectors_msg(struct idpf_adapter *adapter);
 int idpf_send_alloc_vectors_msg(struct idpf_adapter *adapter, u16 num_vectors);
 void idpf_deinit_task(struct idpf_adapter *adapter);
@@ -727,7 +773,7 @@ void idpf_vport_init(struct idpf_vport *vport, struct idpf_vport_max_q *max_q);
 u32 idpf_get_vport_id(struct idpf_vport *vport);
 int idpf_vport_queue_ids_init(struct idpf_vport *vport);
 int idpf_queue_reg_init(struct idpf_vport *vport);
-int idpf_send_config_tx_queues_msg(struct idpf_vport *vport);
+int idpf_send_config_queues_msg(struct idpf_vport *vport);
 int idpf_send_create_vport_msg(struct idpf_adapter *adapter,
 			       struct idpf_vport_max_q *max_q);
 int idpf_check_supported_desc_ids(struct idpf_vport *vport);

drivers/net/ethernet/intel/idpf/idpf_lan_txrx.h

@@ -4,6 +4,58 @@
 #ifndef _IDPF_LAN_TXRX_H_
 #define _IDPF_LAN_TXRX_H_
 
+enum idpf_rss_hash {
+	IDPF_HASH_INVALID			= 0,
+	/* Values 1 - 28 are reserved for future use */
+	IDPF_HASH_NONF_UNICAST_IPV4_UDP		= 29,
+	IDPF_HASH_NONF_MULTICAST_IPV4_UDP,
+	IDPF_HASH_NONF_IPV4_UDP,
+	IDPF_HASH_NONF_IPV4_TCP_SYN_NO_ACK,
+	IDPF_HASH_NONF_IPV4_TCP,
+	IDPF_HASH_NONF_IPV4_SCTP,
+	IDPF_HASH_NONF_IPV4_OTHER,
+	IDPF_HASH_FRAG_IPV4,
+	/* Values 37-38 are reserved */
+	IDPF_HASH_NONF_UNICAST_IPV6_UDP		= 39,
+	IDPF_HASH_NONF_MULTICAST_IPV6_UDP,
+	IDPF_HASH_NONF_IPV6_UDP,
+	IDPF_HASH_NONF_IPV6_TCP_SYN_NO_ACK,
+	IDPF_HASH_NONF_IPV6_TCP,
+	IDPF_HASH_NONF_IPV6_SCTP,
+	IDPF_HASH_NONF_IPV6_OTHER,
+	IDPF_HASH_FRAG_IPV6,
+	IDPF_HASH_NONF_RSVD47,
+	IDPF_HASH_NONF_FCOE_OX,
+	IDPF_HASH_NONF_FCOE_RX,
+	IDPF_HASH_NONF_FCOE_OTHER,
+	/* Values 51-62 are reserved */
+	IDPF_HASH_L2_PAYLOAD			= 63,
+
+	IDPF_HASH_MAX
+};
+
+/* Supported RSS offloads */
+#define IDPF_DEFAULT_RSS_HASH			\
+	(BIT_ULL(IDPF_HASH_NONF_IPV4_UDP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV4_SCTP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV4_TCP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV4_OTHER) |	\
+	BIT_ULL(IDPF_HASH_FRAG_IPV4) |		\
+	BIT_ULL(IDPF_HASH_NONF_IPV6_UDP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV6_TCP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV6_SCTP) |	\
+	BIT_ULL(IDPF_HASH_NONF_IPV6_OTHER) |	\
+	BIT_ULL(IDPF_HASH_FRAG_IPV6) |		\
+	BIT_ULL(IDPF_HASH_L2_PAYLOAD))
+
+#define IDPF_DEFAULT_RSS_HASH_EXPANDED (IDPF_DEFAULT_RSS_HASH | \
+	BIT_ULL(IDPF_HASH_NONF_IPV4_TCP_SYN_NO_ACK) |		\
+	BIT_ULL(IDPF_HASH_NONF_UNICAST_IPV4_UDP) |		\
+	BIT_ULL(IDPF_HASH_NONF_MULTICAST_IPV4_UDP) |		\
+	BIT_ULL(IDPF_HASH_NONF_IPV6_TCP_SYN_NO_ACK) |		\
+	BIT_ULL(IDPF_HASH_NONF_UNICAST_IPV6_UDP) |		\
+	BIT_ULL(IDPF_HASH_NONF_MULTICAST_IPV6_UDP))
+
 /* Transmit descriptors  */
 /* splitq tx buf, singleq tx buf and singleq compl desc */
 struct idpf_base_tx_desc {

drivers/net/ethernet/intel/idpf/idpf_lib.c

@@ -668,11 +668,16 @@ static void idpf_vport_rel(struct idpf_vport *vport)
 {
 	struct idpf_adapter *adapter = vport->adapter;
 	struct idpf_vport_config *vport_config;
+	struct idpf_rss_data *rss_data;
 	struct idpf_vport_max_q max_q;
 	u16 idx = vport->idx;
 	int i;
 
 	vport_config = adapter->vport_config[vport->idx];
+	idpf_deinit_rss(vport);
+	rss_data = &vport_config->user_config.rss_data;
+	kfree(rss_data->rss_key);
+	rss_data->rss_key = NULL;
 
 	idpf_send_destroy_vport_msg(vport);
 
@@ -743,6 +748,7 @@ static void idpf_vport_dealloc(struct idpf_vport *vport)
 static struct idpf_vport *idpf_vport_alloc(struct idpf_adapter *adapter,
 					   struct idpf_vport_max_q *max_q)
 {
+	struct idpf_rss_data *rss_data;
 	u16 idx = adapter->next_vport;
 	struct idpf_vport *vport;
 
@@ -773,6 +779,21 @@ static struct idpf_vport *idpf_vport_alloc(struct idpf_adapter *adapter,
 
 	idpf_vport_init(vport, max_q);
 
+	/* This alloc is done separate from the LUT because it's not strictly
+	 * dependent on how many queues we have. If we change number of queues
+	 * and soft reset we'll need a new LUT but the key can remain the same
+	 * for as long as the vport exists.
+	 */
+	rss_data = &adapter->vport_config[idx]->user_config.rss_data;
+	rss_data->rss_key = kzalloc(rss_data->rss_key_size, GFP_KERNEL);
+	if (!rss_data->rss_key) {
+		kfree(vport);
+
+		return NULL;
+	}
+	/* Initialize default rss key */
+	netdev_rss_key_fill((void *)rss_data->rss_key, rss_data->rss_key_size);
+
 	/* fill vport slot in the adapter struct */
 	adapter->vports[idx] = vport;
 	adapter->vport_ids[idx] = idpf_get_vport_id(vport);
@@ -837,6 +858,7 @@ static int idpf_vport_open(struct idpf_vport *vport, bool alloc_res)
 {
 	struct idpf_netdev_priv *np = netdev_priv(vport->netdev);
 	struct idpf_adapter *adapter = vport->adapter;
+	struct idpf_vport_config *vport_config;
 	int err;
 
 	if (np->state != __IDPF_VPORT_DOWN)
@@ -865,20 +887,38 @@ static int idpf_vport_open(struct idpf_vport *vport, bool alloc_res)
 		goto intr_rel;
 	}
 
+	err = idpf_rx_bufs_init_all(vport);
+	if (err) {
+		dev_err(&adapter->pdev->dev, "Failed to initialize RX buffers for vport %u: %d\n",
+			vport->vport_id, err);
+		goto intr_rel;
+	}
+
 	err = idpf_queue_reg_init(vport);
 	if (err) {
 		dev_err(&adapter->pdev->dev, "Failed to initialize queue registers for vport %u: %d\n",
 			vport->vport_id, err);
 		goto intr_rel;
 	}
 
-	err = idpf_send_config_tx_queues_msg(vport);
+	err = idpf_send_config_queues_msg(vport);
 	if (err) {
 		dev_err(&adapter->pdev->dev, "Failed to configure queues for vport %u, %d\n",
 			vport->vport_id, err);
 		goto intr_rel;
 	}
 
+	vport_config = adapter->vport_config[vport->idx];
+	if (vport_config->user_config.rss_data.rss_lut)
+		err = idpf_config_rss(vport);
+	else
+		err = idpf_init_rss(vport);
+	if (err) {
+		dev_err(&adapter->pdev->dev, "Failed to initialize RSS for vport %u: %d\n",
+			vport->vport_id, err);
+		goto intr_rel;
+	}
+
 	return 0;
 
 intr_rel:

drivers/net/ethernet/intel/idpf/idpf_singleq_txrx.c

@@ -0,0 +1,57 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/* Copyright (C) 2023 Intel Corporation */
+
+#include "idpf.h"
+
+/**
+ * idpf_rx_singleq_buf_hw_alloc_all - Replace used receive buffers
+ * @rx_q: queue for which the hw buffers are allocated
+ * @cleaned_count: number of buffers to replace
+ *
+ * Returns false if all allocations were successful, true if any fail
+ */
+bool idpf_rx_singleq_buf_hw_alloc_all(struct idpf_queue *rx_q,
+				      u16 cleaned_count)
+{
+	struct virtchnl2_singleq_rx_buf_desc *desc;
+	u16 nta = rx_q->next_to_alloc;
+	struct idpf_rx_buf *buf;
+
+	if (!cleaned_count)
+		return false;
+
+	desc = IDPF_SINGLEQ_RX_BUF_DESC(rx_q, nta);
+	buf = &rx_q->rx_buf.buf[nta];
+
+	do {
+		dma_addr_t addr;
+
+		addr = idpf_alloc_page(rx_q->pp, buf, rx_q->rx_buf_size);
+		if (unlikely(addr == DMA_MAPPING_ERROR))
+			break;
+
+		/* Refresh the desc even if buffer_addrs didn't change
+		 * because each write-back erases this info.
+		 */
+		desc->pkt_addr = cpu_to_le64(addr);
+		desc->hdr_addr = 0;
+		desc++;
+
+		buf++;
+		nta++;
+		if (unlikely(nta == rx_q->desc_count)) {
+			desc = IDPF_SINGLEQ_RX_BUF_DESC(rx_q, 0);
+			buf = rx_q->rx_buf.buf;
+			nta = 0;
+		}
+
+		cleaned_count--;
+	} while (cleaned_count);
+
+	if (rx_q->next_to_alloc != nta) {
+		idpf_rx_buf_hw_update(rx_q, nta);
+		rx_q->next_to_alloc = nta;
+	}
+
+	return !!cleaned_count;
+}