bnxt_en: Add support for 2nd firmware message channel.

Earlier, some of the firmware commands (ex: CFA_FLOW_*) which are processed
by KONG processor were sent to the CHIMP processor from the host. This
approach was taken as there was no direct message channel to KONG.
CHIMP in turn used to send them to KONG. Newer firmware supports a new
message channel which the host can send messages directly to the KONG
processor.

This patch adds support for required changes needed in the driver
to support direct KONG message channel.  This speeds up flow related
messages sent to the firmware for CLS_FLOWER offload.
Signed-off-by: Venkat Duvvuru <venkatkumar.duvvuru@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/broadcom/bnxt/bnxt.c

@@ -3279,6 +3279,27 @@ static void bnxt_free_hwrm_resources(struct bnxt *bp)
 				  bp->hwrm_cmd_resp_dma_addr);
 		bp->hwrm_cmd_resp_addr = NULL;
 	}
+
+	if (bp->hwrm_cmd_kong_resp_addr) {
+		dma_free_coherent(&pdev->dev, PAGE_SIZE,
+				  bp->hwrm_cmd_kong_resp_addr,
+				  bp->hwrm_cmd_kong_resp_dma_addr);
+		bp->hwrm_cmd_kong_resp_addr = NULL;
+	}
+}
+
+static int bnxt_alloc_kong_hwrm_resources(struct bnxt *bp)
+{
+	struct pci_dev *pdev = bp->pdev;
+
+	bp->hwrm_cmd_kong_resp_addr =
+		dma_alloc_coherent(&pdev->dev, PAGE_SIZE,
+				   &bp->hwrm_cmd_kong_resp_dma_addr,
+				   GFP_KERNEL);
+	if (!bp->hwrm_cmd_kong_resp_addr)
+		return -ENOMEM;
+
+	return 0;
 }
 
 static int bnxt_alloc_hwrm_resources(struct bnxt *bp)
@@ -3740,7 +3761,10 @@ void bnxt_hwrm_cmd_hdr_init(struct bnxt *bp, void *request, u16 req_type,
 	req->req_type = cpu_to_le16(req_type);
 	req->cmpl_ring = cpu_to_le16(cmpl_ring);
 	req->target_id = cpu_to_le16(target_id);
-	req->resp_addr = cpu_to_le64(bp->hwrm_cmd_resp_dma_addr);
+	if (bnxt_kong_hwrm_message(bp, req))
+		req->resp_addr = cpu_to_le64(bp->hwrm_cmd_kong_resp_dma_addr);
+	else
+		req->resp_addr = cpu_to_le64(bp->hwrm_cmd_resp_dma_addr);
 }
 
 static int bnxt_hwrm_do_send_msg(struct bnxt *bp, void *msg, u32 msg_len,
@@ -3758,11 +3782,7 @@ static int bnxt_hwrm_do_send_msg(struct bnxt *bp, void *msg, u32 msg_len,
 	u32 doorbell_offset = BNXT_GRCPF_REG_CHIMP_COMM_TRIGGER;
 	u8 *resp_addr = (u8 *)bp->hwrm_cmd_resp_addr;
 	u32 bar_offset = BNXT_GRCPF_REG_CHIMP_COMM;
-
-	req->seq_id = cpu_to_le16(bnxt_get_hwrm_seq_id(bp));
-	memset(resp, 0, PAGE_SIZE);
-	cp_ring_id = le16_to_cpu(req->cmpl_ring);
-	intr_process = (cp_ring_id == INVALID_HW_RING_ID) ? 0 : 1;
+	u16 dst = BNXT_HWRM_CHNL_CHIMP;
 
 	if (msg_len > BNXT_HWRM_MAX_REQ_LEN) {
 		if (msg_len > bp->hwrm_max_ext_req_len ||
@@ -3770,6 +3790,23 @@ static int bnxt_hwrm_do_send_msg(struct bnxt *bp, void *msg, u32 msg_len,
 			return -EINVAL;
 	}
 
+	if (bnxt_hwrm_kong_chnl(bp, req)) {
+		dst = BNXT_HWRM_CHNL_KONG;
+		bar_offset = BNXT_GRCPF_REG_KONG_COMM;
+		doorbell_offset = BNXT_GRCPF_REG_KONG_COMM_TRIGGER;
+		resp = bp->hwrm_cmd_kong_resp_addr;
+		resp_addr = (u8 *)bp->hwrm_cmd_kong_resp_addr;
+	}
+
+	memset(resp, 0, PAGE_SIZE);
+	cp_ring_id = le16_to_cpu(req->cmpl_ring);
+	intr_process = (cp_ring_id == INVALID_HW_RING_ID) ? 0 : 1;
+
+	req->seq_id = cpu_to_le16(bnxt_get_hwrm_seq_id(bp, dst));
+	/* currently supports only one outstanding message */
+	if (intr_process)
+		bp->hwrm_intr_seq_id = le16_to_cpu(req->seq_id);
+
 	if ((bp->fw_cap & BNXT_FW_CAP_SHORT_CMD) ||
 	    msg_len > BNXT_HWRM_MAX_REQ_LEN) {
 		void *short_cmd_req = bp->hwrm_short_cmd_req_addr;
@@ -3808,10 +3845,6 @@ static int bnxt_hwrm_do_send_msg(struct bnxt *bp, void *msg, u32 msg_len,
 	for (i = msg_len; i < max_req_len; i += 4)
 		writel(0, bp->bar0 + bar_offset + i);
 
-	/* currently supports only one outstanding message */
-	if (intr_process)
-		bp->hwrm_intr_seq_id = le16_to_cpu(req->seq_id);
-
 	/* Ring channel doorbell */
 	writel(1, bp->bar0 + doorbell_offset);
 
@@ -6488,6 +6521,9 @@ static int bnxt_hwrm_ver_get(struct bnxt *bp)
 	    (dev_caps_cfg & VER_GET_RESP_DEV_CAPS_CFG_SHORT_CMD_REQUIRED))
 		bp->fw_cap |= BNXT_FW_CAP_SHORT_CMD;
 
+	if (dev_caps_cfg & VER_GET_RESP_DEV_CAPS_CFG_KONG_MB_CHNL_SUPPORTED)
+		bp->fw_cap |= BNXT_FW_CAP_KONG_MB_CHNL;
+
 hwrm_ver_get_exit:
 	mutex_unlock(&bp->hwrm_cmd_lock);
 	return rc;
@@ -10226,6 +10262,12 @@ static int bnxt_init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	if (rc)
 		goto init_err_pci_clean;
 
+	if (bp->fw_cap & BNXT_FW_CAP_KONG_MB_CHNL) {
+		rc = bnxt_alloc_kong_hwrm_resources(bp);
+		if (rc)
+			bp->fw_cap &= ~BNXT_FW_CAP_KONG_MB_CHNL;
+	}
+
 	if ((bp->fw_cap & BNXT_FW_CAP_SHORT_CMD) ||
 	    bp->hwrm_max_ext_req_len > BNXT_HWRM_MAX_REQ_LEN) {
 		rc = bnxt_alloc_hwrm_short_cmd_req(bp);

drivers/net/ethernet/broadcom/bnxt/bnxt.h

@@ -584,6 +584,9 @@ struct nqe_cn {
 
 #define HWRM_VALID_BIT_DELAY_USEC	20
 
+#define BNXT_HWRM_CHNL_CHIMP	0
+#define BNXT_HWRM_CHNL_KONG	1
+
 #define BNXT_RX_EVENT	1
 #define BNXT_AGG_EVENT	2
 #define BNXT_TX_EVENT	4
@@ -1118,6 +1121,9 @@ struct bnxt_test_info {
 #define BNXT_CAG_REG_LEGACY_INT_STATUS		0x4014
 #define BNXT_CAG_REG_BASE			0x300000
 
+#define BNXT_GRCPF_REG_KONG_COMM		0xA00
+#define BNXT_GRCPF_REG_KONG_COMM_TRIGGER	0xB00
+
 struct bnxt_tc_flow_stats {
 	u64		packets;
 	u64		bytes;
@@ -1458,20 +1464,24 @@ struct bnxt {
 	u32			msg_enable;
 
 	u32			fw_cap;
-	#define BNXT_FW_CAP_SHORT_CMD	0x00000001
-	#define BNXT_FW_CAP_LLDP_AGENT	0x00000002
-	#define BNXT_FW_CAP_DCBX_AGENT	0x00000004
-	#define BNXT_FW_CAP_NEW_RM	0x00000008
-	#define BNXT_FW_CAP_IF_CHANGE	0x00000010
+	#define BNXT_FW_CAP_SHORT_CMD			0x00000001
+	#define BNXT_FW_CAP_LLDP_AGENT			0x00000002
+	#define BNXT_FW_CAP_DCBX_AGENT			0x00000004
+	#define BNXT_FW_CAP_NEW_RM			0x00000008
+	#define BNXT_FW_CAP_IF_CHANGE			0x00000010
+	#define BNXT_FW_CAP_KONG_MB_CHNL		0x00000080
 
 #define BNXT_NEW_RM(bp)		((bp)->fw_cap & BNXT_FW_CAP_NEW_RM)
 	u32			hwrm_spec_code;
 	u16			hwrm_cmd_seq;
+	u16                     hwrm_cmd_kong_seq;
 	u16			hwrm_intr_seq_id;
 	void			*hwrm_short_cmd_req_addr;
 	dma_addr_t		hwrm_short_cmd_req_dma_addr;
 	void			*hwrm_cmd_resp_addr;
 	dma_addr_t		hwrm_cmd_resp_dma_addr;
+	void			*hwrm_cmd_kong_resp_addr;
+	dma_addr_t		hwrm_cmd_kong_resp_dma_addr;
 
 	struct rtnl_link_stats64	net_stats_prev;
 	struct rx_port_stats	*hw_rx_port_stats;
@@ -1673,16 +1683,63 @@ static inline void bnxt_db_write(struct bnxt *bp, struct bnxt_db_info *db,
 	}
 }
 
+static inline bool bnxt_cfa_hwrm_message(u16 req_type)
+{
+	switch (req_type) {
+	case HWRM_CFA_ENCAP_RECORD_ALLOC:
+	case HWRM_CFA_ENCAP_RECORD_FREE:
+	case HWRM_CFA_DECAP_FILTER_ALLOC:
+	case HWRM_CFA_DECAP_FILTER_FREE:
+	case HWRM_CFA_NTUPLE_FILTER_ALLOC:
+	case HWRM_CFA_NTUPLE_FILTER_FREE:
+	case HWRM_CFA_NTUPLE_FILTER_CFG:
+	case HWRM_CFA_EM_FLOW_ALLOC:
+	case HWRM_CFA_EM_FLOW_FREE:
+	case HWRM_CFA_EM_FLOW_CFG:
+	case HWRM_CFA_FLOW_ALLOC:
+	case HWRM_CFA_FLOW_FREE:
+	case HWRM_CFA_FLOW_INFO:
+	case HWRM_CFA_FLOW_FLUSH:
+	case HWRM_CFA_FLOW_STATS:
+	case HWRM_CFA_METER_PROFILE_ALLOC:
+	case HWRM_CFA_METER_PROFILE_FREE:
+	case HWRM_CFA_METER_PROFILE_CFG:
+	case HWRM_CFA_METER_INSTANCE_ALLOC:
+	case HWRM_CFA_METER_INSTANCE_FREE:
+		return true;
+	default:
+		return false;
+	}
+}
+
+static inline bool bnxt_kong_hwrm_message(struct bnxt *bp, struct input *req)
+{
+	return (bp->fw_cap & BNXT_FW_CAP_KONG_MB_CHNL &&
+		bnxt_cfa_hwrm_message(le16_to_cpu(req->req_type)));
+}
+
+static inline bool bnxt_hwrm_kong_chnl(struct bnxt *bp, struct input *req)
+{
+	return (bp->fw_cap & BNXT_FW_CAP_KONG_MB_CHNL &&
+		req->resp_addr == cpu_to_le64(bp->hwrm_cmd_kong_resp_dma_addr));
+}
+
 static inline void *bnxt_get_hwrm_resp_addr(struct bnxt *bp, void *req)
 {
-	return bp->hwrm_cmd_resp_addr;
+	if (bnxt_hwrm_kong_chnl(bp, (struct input *)req))
+		return bp->hwrm_cmd_kong_resp_addr;
+	else
+		return bp->hwrm_cmd_resp_addr;
 }
 
-static inline u16 bnxt_get_hwrm_seq_id(struct bnxt *bp)
+static inline u16 bnxt_get_hwrm_seq_id(struct bnxt *bp, u16 dst)
 {
 	u16 seq_id;
 
-	seq_id = bp->hwrm_cmd_seq++;
+	if (dst == BNXT_HWRM_CHNL_CHIMP)
+		seq_id = bp->hwrm_cmd_seq++;
+	else
+		seq_id = bp->hwrm_cmd_kong_seq++;
 	return seq_id;
 }