Merge branch 'dpaa2-eth-sw-TSO'

Ioana Ciornei says:

====================
dpaa2-eth: add support for software TSO

This series adds support for driver level TSO in the dpaa2-eth driver.

The first 5 patches lay the ground work for the actual feature:
rearrange some variable declaration, cleaning up the interraction with
the S/G Table buffer cache etc.

The 6th patch adds the actual driver level software TSO support by using
the usual tso_build_hdr()/tso_build_data() APIs and creates the S/G FDs.

With this patch set we can see the following improvement in a TCP flow
running on a single A72@2.2GHz of the LX2160A SoC:

before: 6.38Gbit/s
after:  8.48Gbit/s
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

drivers/net/ethernet/freescale/dpaa2/dpaa2-eth.h

@@ -122,6 +122,7 @@ enum dpaa2_eth_swa_type {
 	DPAA2_ETH_SWA_SINGLE,
 	DPAA2_ETH_SWA_SG,
 	DPAA2_ETH_SWA_XDP,
+	DPAA2_ETH_SWA_SW_TSO,
 };
 
 /* Must keep this struct smaller than DPAA2_ETH_SWA_SIZE */
@@ -142,6 +143,12 @@ struct dpaa2_eth_swa {
 			int dma_size;
 			struct xdp_frame *xdpf;
 		} xdp;
+		struct {
+			struct sk_buff *skb;
+			int num_sg;
+			int sgt_size;
+			int is_last_fd;
+		} tso;
 	};
 };
 
@@ -354,6 +361,8 @@ struct dpaa2_eth_drv_stats {
 	__u64	tx_conf_bytes;
 	__u64	tx_sg_frames;
 	__u64	tx_sg_bytes;
+	__u64	tx_tso_frames;
+	__u64	tx_tso_bytes;
 	__u64	rx_sg_frames;
 	__u64	rx_sg_bytes;
 	/* Linear skbs sent as a S/G FD due to insufficient headroom */
@@ -493,8 +502,15 @@ struct dpaa2_eth_trap_data {
 	struct dpaa2_eth_priv *priv;
 };
 
+#define DPAA2_ETH_SG_ENTRIES_MAX	(PAGE_SIZE / sizeof(struct scatterlist))
+
 #define DPAA2_ETH_DEFAULT_COPYBREAK	512
 
+#define DPAA2_ETH_ENQUEUE_MAX_FDS	200
+struct dpaa2_eth_fds {
+	struct dpaa2_fd array[DPAA2_ETH_ENQUEUE_MAX_FDS];
+};
+
 /* Driver private data */
 struct dpaa2_eth_priv {
 	struct net_device *net_dev;
@@ -577,6 +593,8 @@ struct dpaa2_eth_priv {
 	struct devlink_port devlink_port;
 
 	u32 rx_copybreak;
+
+	struct dpaa2_eth_fds __percpu *fd;
 };
 
 struct dpaa2_eth_devlink_priv {

drivers/net/ethernet/freescale/dpaa2/dpaa2-ethtool.c

@@ -44,6 +44,8 @@ static char dpaa2_ethtool_extras[][ETH_GSTRING_LEN] = {
 	"[drv] tx conf bytes",
 	"[drv] tx sg frames",
 	"[drv] tx sg bytes",
+	"[drv] tx tso frames",
+	"[drv] tx tso bytes",
 	"[drv] rx sg frames",
 	"[drv] rx sg bytes",
 	"[drv] tx converted sg frames",

drivers/soc/fsl/dpio/qbman-portal.c

@@ -743,8 +743,8 @@ int qbman_swp_enqueue_multiple_mem_back(struct qbman_swp *s,
 	full_mask = s->eqcr.pi_ci_mask;
 	if (!s->eqcr.available) {
 		eqcr_ci = s->eqcr.ci;
-		p = s->addr_cena + QBMAN_CENA_SWP_EQCR_CI_MEMBACK;
-		s->eqcr.ci = *p & full_mask;
+		s->eqcr.ci = qbman_read_register(s, QBMAN_CINH_SWP_EQCR_CI);
+		s->eqcr.ci &= full_mask;
 		s->eqcr.available = qm_cyc_diff(s->eqcr.pi_ring_size,
 					eqcr_ci, s->eqcr.ci);
 		if (!s->eqcr.available) {
@@ -887,8 +887,8 @@ int qbman_swp_enqueue_multiple_desc_mem_back(struct qbman_swp *s,
 	full_mask = s->eqcr.pi_ci_mask;
 	if (!s->eqcr.available) {
 		eqcr_ci = s->eqcr.ci;
-		p = s->addr_cena + QBMAN_CENA_SWP_EQCR_CI_MEMBACK;
-		s->eqcr.ci = *p & full_mask;
+		s->eqcr.ci = qbman_read_register(s, QBMAN_CINH_SWP_EQCR_CI);
+		s->eqcr.ci &= full_mask;
 		s->eqcr.available = qm_cyc_diff(s->eqcr.pi_ring_size,
 					eqcr_ci, s->eqcr.ci);
 		if (!s->eqcr.available)