octeontx2-af: enable tx shaping feature for 96xx C0

Starting from 96xx C0 onwards all silicons support traffic shaping.
This patch enables that feature along with other changes
- When PIR/CIR shaping config is modified, toggle SW_XOFF
  for config to take effect
- Before SMQ flush, clear SW_XOFF at all parent schedulers
- Support to read current transmit scheduler configuration via mbox
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/marvell/octeontx2/af/mbox.h

@@ -87,7 +87,7 @@ struct mbox_msghdr {
 #define OTX2_MBOX_REQ_SIG (0xdead)
 #define OTX2_MBOX_RSP_SIG (0xbeef)
 	u16 sig;         /* Signature, for validating corrupted msgs */
-#define OTX2_MBOX_VERSION (0x0007)
+#define OTX2_MBOX_VERSION (0x0009)
 	u16 ver;         /* Version of msg's structure for this ID */
 	u16 next_msgoff; /* Offset of next msg within mailbox region */
 	int rc;          /* Msg process'ed response code */
@@ -243,7 +243,8 @@ M(NIX_HWCTX_DISABLE,	0x8003, nix_hwctx_disable,			\
 M(NIX_TXSCH_ALLOC,	0x8004, nix_txsch_alloc,			\
 				 nix_txsch_alloc_req, nix_txsch_alloc_rsp)   \
 M(NIX_TXSCH_FREE,	0x8005, nix_txsch_free, nix_txsch_free_req, msg_rsp) \
-M(NIX_TXSCHQ_CFG,	0x8006, nix_txschq_cfg, nix_txschq_config, msg_rsp)  \
+M(NIX_TXSCHQ_CFG,	0x8006, nix_txschq_cfg, nix_txschq_config,	\
+				nix_txschq_config)			\
 M(NIX_STATS_RST,	0x8007, nix_stats_rst, msg_req, msg_rsp)	\
 M(NIX_VTAG_CFG,		0x8008, nix_vtag_cfg, nix_vtag_config,		\
 				 nix_vtag_config_rsp)			\
@@ -370,16 +371,20 @@ struct msix_offset_rsp {
 	struct mbox_msghdr hdr;
 	u16  npa_msixoff;
 	u16  nix_msixoff;
-	u8   sso;
-	u8   ssow;
-	u8   timlfs;
-	u8   cptlfs;
+	u16  sso;
+	u16  ssow;
+	u16  timlfs;
+	u16  cptlfs;
 	u16  sso_msixoff[MAX_RVU_BLKLF_CNT];
 	u16  ssow_msixoff[MAX_RVU_BLKLF_CNT];
 	u16  timlf_msixoff[MAX_RVU_BLKLF_CNT];
 	u16  cptlf_msixoff[MAX_RVU_BLKLF_CNT];
-	u8   cpt1_lfs;
+	u16  cpt1_lfs;
+	u16  ree0_lfs;
+	u16  ree1_lfs;
 	u16  cpt1_lf_msixoff[MAX_RVU_BLKLF_CNT];
+	u16  ree0_lf_msixoff[MAX_RVU_BLKLF_CNT];
+	u16  ree1_lf_msixoff[MAX_RVU_BLKLF_CNT];
 };
 
 struct get_hw_cap_rsp {
@@ -836,6 +841,7 @@ struct nix_txsch_free_req {
 struct nix_txschq_config {
 	struct mbox_msghdr hdr;
 	u8 lvl;	/* SMQ/MDQ/TL4/TL3/TL2/TL1 */
+	u8 read;
 #define TXSCHQ_IDX_SHIFT	16
 #define TXSCHQ_IDX_MASK		(BIT_ULL(10) - 1)
 #define TXSCHQ_IDX(reg, shift)	(((reg) >> (shift)) & TXSCHQ_IDX_MASK)
@@ -843,6 +849,8 @@ struct nix_txschq_config {
 #define MAX_REGS_PER_MBOX_MSG	20
 	u64 reg[MAX_REGS_PER_MBOX_MSG];
 	u64 regval[MAX_REGS_PER_MBOX_MSG];
+	/* All 0's => overwrite with new value */
+	u64 regval_mask[MAX_REGS_PER_MBOX_MSG];
 };
 
 struct nix_vtag_config {

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

@@ -70,18 +70,21 @@ static void rvu_setup_hw_capabilities(struct rvu *rvu)
 	hw->cap.nix_shaping = true;
 	hw->cap.nix_tx_link_bp = true;
 	hw->cap.nix_rx_multicast = true;
+	hw->cap.nix_shaper_toggle_wait = false;
 	hw->rvu = rvu;
 
-	if (is_rvu_96xx_B0(rvu)) {
+	if (is_rvu_pre_96xx_C0(rvu)) {
 		hw->cap.nix_fixed_txschq_mapping = true;
 		hw->cap.nix_txsch_per_cgx_lmac = 4;
 		hw->cap.nix_txsch_per_lbk_lmac = 132;
 		hw->cap.nix_txsch_per_sdp_lmac = 76;
 		hw->cap.nix_shaping = false;
 		hw->cap.nix_tx_link_bp = false;
-		if (is_rvu_96xx_A0(rvu))
+		if (is_rvu_96xx_A0(rvu) || is_rvu_95xx_A0(rvu))
 			hw->cap.nix_rx_multicast = false;
 	}
+	if (!is_rvu_pre_96xx_C0(rvu))
+		hw->cap.nix_shaper_toggle_wait = true;
 
 	if (!is_rvu_otx2(rvu))
 		hw->cap.per_pf_mbox_regs = true;

drivers/net/ethernet/marvell/octeontx2/af/rvu.h

@@ -328,6 +328,7 @@ struct hw_cap {
 	u16	nix_txsch_per_sdp_lmac; /* Max Q's transmitting to SDP LMAC */
 	bool	nix_fixed_txschq_mapping; /* Schq mapping fixed or flexible */
 	bool	nix_shaping;		 /* Is shaping and coloring supported */
+	bool    nix_shaper_toggle_wait; /* Shaping toggle needs poll/wait */
 	bool	nix_tx_link_bp;		 /* Can link backpressure TL queues ? */
 	bool	nix_rx_multicast;	 /* Rx packet replication support */
 	bool	nix_common_dwrr_mtu;	 /* Common DWRR MTU for quantum config */
@@ -517,20 +518,34 @@ static inline u64 rvupf_read64(struct rvu *rvu, u64 offset)
 }
 
 /* Silicon revisions */
+static inline bool is_rvu_pre_96xx_C0(struct rvu *rvu)
+{
+	struct pci_dev *pdev = rvu->pdev;
+	/* 96XX A0/B0, 95XX A0/A1/B0 chips */
+	return ((pdev->revision == 0x00) || (pdev->revision == 0x01) ||
+		(pdev->revision == 0x10) || (pdev->revision == 0x11) ||
+		(pdev->revision == 0x14));
+}
+
 static inline bool is_rvu_96xx_A0(struct rvu *rvu)
 {
 	struct pci_dev *pdev = rvu->pdev;
 
-	return (pdev->revision == 0x00) &&
-		(pdev->subsystem_device == PCI_SUBSYS_DEVID_96XX);
+	return (pdev->revision == 0x00);
 }
 
 static inline bool is_rvu_96xx_B0(struct rvu *rvu)
 {
 	struct pci_dev *pdev = rvu->pdev;
 
-	return ((pdev->revision == 0x00) || (pdev->revision == 0x01)) &&
-		(pdev->subsystem_device == PCI_SUBSYS_DEVID_96XX);
+	return (pdev->revision == 0x00) || (pdev->revision == 0x01);
+}
+
+static inline bool is_rvu_95xx_A0(struct rvu *rvu)
+{
+	struct pci_dev *pdev = rvu->pdev;
+
+	return (pdev->revision == 0x10) || (pdev->revision == 0x11);
 }
 
 /* REVID for PCIe devices.

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

@@ -33,8 +33,8 @@ static struct hw_reg_map txsch_reg_map[NIX_TXSCH_LVL_CNT] = {
 	{NIX_TXSCH_LVL_SMQ, 2, 0xFFFF, {{0x0700, 0x0708}, {0x1400, 0x14C8} } },
 	{NIX_TXSCH_LVL_TL4, 3, 0xFFFF, {{0x0B00, 0x0B08}, {0x0B10, 0x0B18},
 			      {0x1200, 0x12E0} } },
-	{NIX_TXSCH_LVL_TL3, 3, 0xFFFF, {{0x1000, 0x10E0}, {0x1600, 0x1608},
-			      {0x1610, 0x1618} } },
+	{NIX_TXSCH_LVL_TL3, 4, 0xFFFF, {{0x1000, 0x10E0}, {0x1600, 0x1608},
+			      {0x1610, 0x1618}, {0x1700, 0x17B0} } },
 	{NIX_TXSCH_LVL_TL2, 2, 0xFFFF, {{0x0E00, 0x0EE0}, {0x1700, 0x17B0} } },
 	{NIX_TXSCH_LVL_TL1, 1, 0xFFFF, {{0x0C00, 0x0D98} } },
 };