octeontx2-pf: Cleanup flow rule management

Current MCAM allocation scheme allocates a single lot of
MCAM entries for ntuple filters, unicast filters and VF VLAN
rules. This patch attempts to cleanup this logic by segregating
MCAM rule allocation and management for Ntuple rules and unicast,
VF VLAN rules. This segregation will result in reusing most of
the logic for supporting ntuple filters for VF devices.

Also added debug messages for MCAM entry allocation failures.

Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/marvell/octeontx2/af/rvu_npc.c

@@ -2537,8 +2537,11 @@ int rvu_mbox_handler_npc_mcam_alloc_entry(struct rvu *rvu,
 	rsp->free_count = 0;
 
 	/* Check if ref_entry is within range */
-	if (req->priority && req->ref_entry >= mcam->bmap_entries)
+	if (req->priority && req->ref_entry >= mcam->bmap_entries) {
+		dev_err(rvu->dev, "%s: reference entry %d is out of range\n",
+			__func__, req->ref_entry);
 		return NPC_MCAM_INVALID_REQ;
+	}
 
 	/* ref_entry can't be '0' if requested priority is high.
 	 * Can't be last entry if requested priority is low.
@@ -2551,8 +2554,12 @@ int rvu_mbox_handler_npc_mcam_alloc_entry(struct rvu *rvu,
 	/* Since list of allocated indices needs to be sent to requester,
 	 * max number of non-contiguous entries per mbox msg is limited.
 	 */
-	if (!req->contig && req->count > NPC_MAX_NONCONTIG_ENTRIES)
+	if (!req->contig && req->count > NPC_MAX_NONCONTIG_ENTRIES) {
+		dev_err(rvu->dev,
+			"%s: %d Non-contiguous MCAM entries requested is morethan max (%d) allowed\n",
+			__func__, req->count, NPC_MAX_NONCONTIG_ENTRIES);
 		return NPC_MCAM_INVALID_REQ;
+	}
 
 	/* Alloc request from PFFUNC with no NIXLF attached should be denied */
 	if (!is_nixlf_attached(rvu, pcifunc))

drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h

@@ -267,24 +267,26 @@ struct otx2_mac_table {
 
 struct otx2_flow_config {
 	u16			entry[NPC_MAX_NONCONTIG_ENTRIES];
-	u32			nr_flows;
-#define OTX2_MAX_NTUPLE_FLOWS	32
-#define OTX2_MAX_UNICAST_FLOWS	8
-#define OTX2_MAX_VLAN_FLOWS	1
-#define OTX2_MAX_TC_FLOWS	OTX2_MAX_NTUPLE_FLOWS
-#define OTX2_MCAM_COUNT		(OTX2_MAX_NTUPLE_FLOWS + \
+	u16			*flow_ent;
+	u16			*def_ent;
+	u16			nr_flows;
+#define OTX2_DEFAULT_FLOWCOUNT		16
+#define OTX2_MAX_UNICAST_FLOWS		8
+#define OTX2_MAX_VLAN_FLOWS		1
+#define OTX2_MAX_TC_FLOWS	OTX2_DEFAULT_FLOWCOUNT
+#define OTX2_MCAM_COUNT		(OTX2_DEFAULT_FLOWCOUNT + \
 				 OTX2_MAX_UNICAST_FLOWS + \
 				 OTX2_MAX_VLAN_FLOWS)
-	u32			ntuple_offset;
-	u32			unicast_offset;
-	u32			rx_vlan_offset;
-	u32			vf_vlan_offset;
-#define OTX2_PER_VF_VLAN_FLOWS	2 /* rx+tx per VF */
+	u16			ntuple_offset;
+	u16			unicast_offset;
+	u16			rx_vlan_offset;
+	u16			vf_vlan_offset;
+#define OTX2_PER_VF_VLAN_FLOWS	2 /* Rx + Tx per VF */
 #define OTX2_VF_VLAN_RX_INDEX	0
 #define OTX2_VF_VLAN_TX_INDEX	1
-	u32			tc_flower_offset;
-	u32                     ntuple_max_flows;
-	u32			tc_max_flows;
+	u16			tc_flower_offset;
+	u16                     ntuple_max_flows;
+	u16			tc_max_flows;
 	struct list_head	flow_list;
 };
 

drivers/net/ethernet/marvell/octeontx2/nic/otx2_flows.c

@@ -20,13 +20,125 @@ struct otx2_flow {
 	int vf;
 };
 
+static void otx2_clear_ntuple_flow_info(struct otx2_nic *pfvf, struct otx2_flow_config *flow_cfg)
+{
+	devm_kfree(pfvf->dev, flow_cfg->flow_ent);
+	flow_cfg->flow_ent = NULL;
+	flow_cfg->ntuple_max_flows = 0;
+	flow_cfg->tc_max_flows = 0;
+}
+
+static int otx2_free_ntuple_mcam_entries(struct otx2_nic *pfvf)
+{
+	struct otx2_flow_config *flow_cfg = pfvf->flow_cfg;
+	struct npc_mcam_free_entry_req *req;
+	int ent, err;
+
+	if (!flow_cfg->ntuple_max_flows)
+		return 0;
+
+	mutex_lock(&pfvf->mbox.lock);
+	for (ent = 0; ent < flow_cfg->ntuple_max_flows; ent++) {
+		req = otx2_mbox_alloc_msg_npc_mcam_free_entry(&pfvf->mbox);
+		if (!req)
+			break;
+
+		req->entry = flow_cfg->flow_ent[ent];
+
+		/* Send message to AF to free MCAM entries */
+		err = otx2_sync_mbox_msg(&pfvf->mbox);
+		if (err)
+			break;
+	}
+	mutex_unlock(&pfvf->mbox.lock);
+	otx2_clear_ntuple_flow_info(pfvf, flow_cfg);
+	return 0;
+}
+
+static int otx2_alloc_ntuple_mcam_entries(struct otx2_nic *pfvf, u16 count)
+{
+	struct otx2_flow_config *flow_cfg = pfvf->flow_cfg;
+	struct npc_mcam_alloc_entry_req *req;
+	struct npc_mcam_alloc_entry_rsp *rsp;
+	int ent, allocated = 0;
+
+	/* Free current ones and allocate new ones with requested count */
+	otx2_free_ntuple_mcam_entries(pfvf);
+
+	if (!count)
+		return 0;
+
+	flow_cfg->flow_ent = devm_kmalloc_array(pfvf->dev, count,
+						sizeof(u16), GFP_KERNEL);
+	if (!flow_cfg->flow_ent)
+		return -ENOMEM;
+
+	mutex_lock(&pfvf->mbox.lock);
+
+	/* In a single request a max of NPC_MAX_NONCONTIG_ENTRIES MCAM entries
+	 * can only be allocated.
+	 */
+	while (allocated < count) {
+		req = otx2_mbox_alloc_msg_npc_mcam_alloc_entry(&pfvf->mbox);
+		if (!req)
+			goto exit;
+
+		req->contig = false;
+		req->count = (count - allocated) > NPC_MAX_NONCONTIG_ENTRIES ?
+				NPC_MAX_NONCONTIG_ENTRIES : count - allocated;
+		req->priority = NPC_MCAM_HIGHER_PRIO;
+		req->ref_entry = flow_cfg->def_ent[0];
+
+		/* Send message to AF */
+		if (otx2_sync_mbox_msg(&pfvf->mbox))
+			goto exit;
+
+		rsp = (struct npc_mcam_alloc_entry_rsp *)otx2_mbox_get_rsp
+			(&pfvf->mbox.mbox, 0, &req->hdr);
+
+		for (ent = 0; ent < rsp->count; ent++)
+			flow_cfg->flow_ent[ent + allocated] = rsp->entry_list[ent];
+
+		allocated += rsp->count;
+
+		/* If this request is not fulfilled, no need to send
+		 * further requests.
+		 */
+		if (rsp->count != req->count)
+			break;
+	}
+
+exit:
+	mutex_unlock(&pfvf->mbox.lock);
+
+	flow_cfg->ntuple_offset = 0;
+	flow_cfg->ntuple_max_flows = allocated;
+	flow_cfg->tc_max_flows = allocated;
+
+	if (allocated != count)
+		netdev_info(pfvf->netdev,
+			    "Unable to allocate %d MCAM entries for ntuple, got %d\n",
+			    count, allocated);
+
+	return allocated;
+}
+
 int otx2_alloc_mcam_entries(struct otx2_nic *pfvf)
 {
 	struct otx2_flow_config *flow_cfg = pfvf->flow_cfg;
 	struct npc_mcam_alloc_entry_req *req;
 	struct npc_mcam_alloc_entry_rsp *rsp;
 	int vf_vlan_max_flows;
-	int i;
+	int ent, count;
+
+	vf_vlan_max_flows = pfvf->total_vfs * OTX2_PER_VF_VLAN_FLOWS;
+	count = OTX2_MAX_UNICAST_FLOWS +
+			OTX2_MAX_VLAN_FLOWS + vf_vlan_max_flows;
+
+	flow_cfg->def_ent = devm_kmalloc_array(pfvf->dev, count,
+					       sizeof(u16), GFP_KERNEL);
+	if (!flow_cfg->def_ent)
+		return -ENOMEM;
 
 	mutex_lock(&pfvf->mbox.lock);
 
@@ -36,9 +148,8 @@ int otx2_alloc_mcam_entries(struct otx2_nic *pfvf)
 		return -ENOMEM;
 	}
 
-	vf_vlan_max_flows = pfvf->total_vfs * OTX2_PER_VF_VLAN_FLOWS;
 	req->contig = false;
-	req->count = OTX2_MCAM_COUNT + vf_vlan_max_flows;
+	req->count = count;
 
 	/* Send message to AF */
 	if (otx2_sync_mbox_msg(&pfvf->mbox)) {
@@ -51,37 +162,36 @@ int otx2_alloc_mcam_entries(struct otx2_nic *pfvf)
 
 	if (rsp->count != req->count) {
 		netdev_info(pfvf->netdev,
-			    "Unable to allocate %d MCAM entries, got %d\n",
-			    req->count, rsp->count);
-		/* support only ntuples here */
-		flow_cfg->ntuple_max_flows = rsp->count;
-		flow_cfg->ntuple_offset = 0;
-		pfvf->flags |= OTX2_FLAG_NTUPLE_SUPPORT;
-		flow_cfg->tc_max_flows = flow_cfg->ntuple_max_flows;
-		pfvf->flags |= OTX2_FLAG_TC_FLOWER_SUPPORT;
-	} else {
-		flow_cfg->vf_vlan_offset = 0;
-		flow_cfg->ntuple_offset = flow_cfg->vf_vlan_offset +
-						vf_vlan_max_flows;
-		flow_cfg->tc_flower_offset = flow_cfg->ntuple_offset;
-		flow_cfg->unicast_offset = flow_cfg->ntuple_offset +
-						OTX2_MAX_NTUPLE_FLOWS;
-		flow_cfg->rx_vlan_offset = flow_cfg->unicast_offset +
-						OTX2_MAX_UNICAST_FLOWS;
-		pfvf->flags |= OTX2_FLAG_NTUPLE_SUPPORT;
-		pfvf->flags |= OTX2_FLAG_UCAST_FLTR_SUPPORT;
-		pfvf->flags |= OTX2_FLAG_RX_VLAN_SUPPORT;
-		pfvf->flags |= OTX2_FLAG_VF_VLAN_SUPPORT;
-		pfvf->flags |= OTX2_FLAG_TC_FLOWER_SUPPORT;
-	}
-
-	for (i = 0; i < rsp->count; i++)
-		flow_cfg->entry[i] = rsp->entry_list[i];
+			    "Unable to allocate MCAM entries for ucast, vlan and vf_vlan\n");
+		mutex_unlock(&pfvf->mbox.lock);
+		devm_kfree(pfvf->dev, flow_cfg->def_ent);
+		return 0;
+	}
 
-	pfvf->flags |= OTX2_FLAG_MCAM_ENTRIES_ALLOC;
+	for (ent = 0; ent < rsp->count; ent++)
+		flow_cfg->def_ent[ent] = rsp->entry_list[ent];
 
+	flow_cfg->vf_vlan_offset = 0;
+	flow_cfg->unicast_offset = vf_vlan_max_flows;
+	flow_cfg->rx_vlan_offset = flow_cfg->unicast_offset +
+					OTX2_MAX_UNICAST_FLOWS;
+	pfvf->flags |= OTX2_FLAG_UCAST_FLTR_SUPPORT;
+	pfvf->flags |= OTX2_FLAG_RX_VLAN_SUPPORT;
+	pfvf->flags |= OTX2_FLAG_VF_VLAN_SUPPORT;
+
+	pfvf->flags |= OTX2_FLAG_MCAM_ENTRIES_ALLOC;
 	mutex_unlock(&pfvf->mbox.lock);
 
+	/* Allocate entries for Ntuple filters */
+	count = otx2_alloc_ntuple_mcam_entries(pfvf, OTX2_DEFAULT_FLOWCOUNT);
+	if (count <= 0) {
+		otx2_clear_ntuple_flow_info(pfvf, flow_cfg);
+		return 0;
+	}
+
+	pfvf->flags |= OTX2_FLAG_NTUPLE_SUPPORT;
+	pfvf->flags |= OTX2_FLAG_TC_FLOWER_SUPPORT;
+
 	return 0;
 }
 
@@ -96,13 +206,14 @@ int otx2_mcam_flow_init(struct otx2_nic *pf)
 
 	INIT_LIST_HEAD(&pf->flow_cfg->flow_list);
 
-	pf->flow_cfg->ntuple_max_flows = OTX2_MAX_NTUPLE_FLOWS;
-	pf->flow_cfg->tc_max_flows = pf->flow_cfg->ntuple_max_flows;
-
 	err = otx2_alloc_mcam_entries(pf);
 	if (err)
 		return err;
 
+	/* Check if MCAM entries are allocate or not */
+	if (!(pf->flags & OTX2_FLAG_UCAST_FLTR_SUPPORT))
+		return 0;
+
 	pf->mac_table = devm_kzalloc(pf->dev, sizeof(struct otx2_mac_table)
 					* OTX2_MAX_UNICAST_FLOWS, GFP_KERNEL);
 	if (!pf->mac_table)
@@ -146,7 +257,7 @@ static int otx2_do_add_macfilter(struct otx2_nic *pf, const u8 *mac)
 		ether_addr_copy(pf->mac_table[i].addr, mac);
 		pf->mac_table[i].inuse = true;
 		pf->mac_table[i].mcam_entry =
-			flow_cfg->entry[i + flow_cfg->unicast_offset];
+			flow_cfg->def_ent[i + flow_cfg->unicast_offset];
 		req->entry =  pf->mac_table[i].mcam_entry;
 		break;
 	}
@@ -732,8 +843,7 @@ int otx2_add_flow(struct otx2_nic *pfvf, struct ethtool_rxnfc *nfc)
 		if (!flow)
 			return -ENOMEM;
 		flow->location = fsp->location;
-		flow->entry = flow_cfg->entry[flow_cfg->ntuple_offset +
-						flow->location];
+		flow->entry = flow_cfg->flow_ent[flow->location];
 		new = true;
 	}
 	/* struct copy */
@@ -837,9 +947,8 @@ int otx2_destroy_ntuple_flows(struct otx2_nic *pfvf)
 		return -ENOMEM;
 	}
 
-	req->start = flow_cfg->entry[flow_cfg->ntuple_offset];
-	req->end   = flow_cfg->entry[flow_cfg->ntuple_offset +
-				      flow_cfg->ntuple_max_flows - 1];
+	req->start = flow_cfg->flow_ent[0];
+	req->end   = flow_cfg->flow_ent[flow_cfg->ntuple_max_flows - 1];
 	err = otx2_sync_mbox_msg(&pfvf->mbox);
 	mutex_unlock(&pfvf->mbox.lock);
 
@@ -906,7 +1015,7 @@ int otx2_install_rxvlan_offload_flow(struct otx2_nic *pfvf)
 		return -ENOMEM;
 	}
 
-	req->entry = flow_cfg->entry[flow_cfg->rx_vlan_offset];
+	req->entry = flow_cfg->def_ent[flow_cfg->rx_vlan_offset];
 	req->intf = NIX_INTF_RX;
 	ether_addr_copy(req->packet.dmac, pfvf->netdev->dev_addr);
 	eth_broadcast_addr((u8 *)&req->mask.dmac);
@@ -935,7 +1044,7 @@ static int otx2_delete_rxvlan_offload_flow(struct otx2_nic *pfvf)
 		return -ENOMEM;
 	}
 
-	req->entry = flow_cfg->entry[flow_cfg->rx_vlan_offset];
+	req->entry = flow_cfg->def_ent[flow_cfg->rx_vlan_offset];
 	/* Send message to AF */
 	err = otx2_sync_mbox_msg(&pfvf->mbox);
 	mutex_unlock(&pfvf->mbox.lock);

drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c

@@ -2109,7 +2109,7 @@ static int otx2_do_set_vf_vlan(struct otx2_nic *pf, int vf, u16 vlan, u8 qos,
 		}
 		idx = ((vf * OTX2_PER_VF_VLAN_FLOWS) + OTX2_VF_VLAN_RX_INDEX);
 		del_req->entry =
-			flow_cfg->entry[flow_cfg->vf_vlan_offset + idx];
+			flow_cfg->def_ent[flow_cfg->vf_vlan_offset + idx];
 		err = otx2_sync_mbox_msg(&pf->mbox);
 		if (err)
 			goto out;
@@ -2122,7 +2122,7 @@ static int otx2_do_set_vf_vlan(struct otx2_nic *pf, int vf, u16 vlan, u8 qos,
 		}
 		idx = ((vf * OTX2_PER_VF_VLAN_FLOWS) + OTX2_VF_VLAN_TX_INDEX);
 		del_req->entry =
-			flow_cfg->entry[flow_cfg->vf_vlan_offset + idx];
+			flow_cfg->def_ent[flow_cfg->vf_vlan_offset + idx];
 		err = otx2_sync_mbox_msg(&pf->mbox);
 
 		goto out;
@@ -2136,7 +2136,7 @@ static int otx2_do_set_vf_vlan(struct otx2_nic *pf, int vf, u16 vlan, u8 qos,
 	}
 
 	idx = ((vf * OTX2_PER_VF_VLAN_FLOWS) + OTX2_VF_VLAN_RX_INDEX);
-	req->entry = flow_cfg->entry[flow_cfg->vf_vlan_offset + idx];
+	req->entry = flow_cfg->def_ent[flow_cfg->vf_vlan_offset + idx];
 	req->packet.vlan_tci = htons(vlan);
 	req->mask.vlan_tci = htons(VLAN_VID_MASK);
 	/* af fills the destination mac addr */
@@ -2187,7 +2187,7 @@ static int otx2_do_set_vf_vlan(struct otx2_nic *pf, int vf, u16 vlan, u8 qos,
 
 	eth_zero_addr((u8 *)&req->mask.dmac);
 	idx = ((vf * OTX2_PER_VF_VLAN_FLOWS) + OTX2_VF_VLAN_TX_INDEX);
-	req->entry = flow_cfg->entry[flow_cfg->vf_vlan_offset + idx];
+	req->entry = flow_cfg->def_ent[flow_cfg->vf_vlan_offset + idx];
 	req->features = BIT_ULL(NPC_DMAC);
 	req->channel = pf->hw.tx_chan_base;
 	req->intf = NIX_INTF_TX;

drivers/net/ethernet/marvell/octeontx2/nic/otx2_tc.c

@@ -570,8 +570,8 @@ static int otx2_tc_add_flow(struct otx2_nic *nic,
 	new_node->bitpos = find_first_zero_bit(tc_info->tc_entries_bitmap,
 					       nic->flow_cfg->tc_max_flows);
 	req->channel = nic->hw.rx_chan_base;
-	req->entry = nic->flow_cfg->entry[nic->flow_cfg->tc_flower_offset +
-					  nic->flow_cfg->tc_max_flows - new_node->bitpos];
+	req->entry = nic->flow_cfg->flow_ent[nic->flow_cfg->tc_flower_offset +
+				nic->flow_cfg->tc_max_flows - new_node->bitpos];
 	req->intf = NIX_INTF_RX;
 	req->set_cntr = 1;
 	new_node->entry = req->entry;