Merge bb8f86f40e ("net: export a helper for adding up queue stats") into android16-6.12-lts

Steps on the way to 6.12.29

Change-Id: I176bb12ce2ae5614c558f22d50d03f8588b0d28a
Signed-off-by: Greg Kroah-Hartman <gregkh@google.com>

drivers/net/dsa/b53/b53_common.c

@@ -373,15 +373,17 @@ static void b53_enable_vlan(struct b53_device *dev, int port, bool enable,
 		b53_read8(dev, B53_VLAN_PAGE, B53_VLAN_CTRL5, &vc5);
 	}
 
+	vc1 &= ~VC1_RX_MCST_FWD_EN;
+
 	if (enable) {
 		vc0 |= VC0_VLAN_EN | VC0_VID_CHK_EN | VC0_VID_HASH_VID;
-		vc1 |= VC1_RX_MCST_UNTAG_EN | VC1_RX_MCST_FWD_EN;
+		vc1 |= VC1_RX_MCST_UNTAG_EN;
 		vc4 &= ~VC4_ING_VID_CHECK_MASK;
 		if (enable_filtering) {
 			vc4 |= VC4_ING_VID_VIO_DROP << VC4_ING_VID_CHECK_S;
 			vc5 |= VC5_DROP_VTABLE_MISS;
 		} else {
-			vc4 |= VC4_ING_VID_VIO_FWD << VC4_ING_VID_CHECK_S;
+			vc4 |= VC4_NO_ING_VID_CHK << VC4_ING_VID_CHECK_S;
 			vc5 &= ~VC5_DROP_VTABLE_MISS;
 		}
 
@@ -393,7 +395,7 @@ static void b53_enable_vlan(struct b53_device *dev, int port, bool enable,
 
 	} else {
 		vc0 &= ~(VC0_VLAN_EN | VC0_VID_CHK_EN | VC0_VID_HASH_VID);
-		vc1 &= ~(VC1_RX_MCST_UNTAG_EN | VC1_RX_MCST_FWD_EN);
+		vc1 &= ~VC1_RX_MCST_UNTAG_EN;
 		vc4 &= ~VC4_ING_VID_CHECK_MASK;
 		vc5 &= ~VC5_DROP_VTABLE_MISS;
 
@@ -576,6 +578,18 @@ static void b53_eee_enable_set(struct dsa_switch *ds, int port, bool enable)
 	b53_write16(dev, B53_EEE_PAGE, B53_EEE_EN_CTRL, reg);
 }
 
+int b53_setup_port(struct dsa_switch *ds, int port)
+{
+	struct b53_device *dev = ds->priv;
+
+	b53_port_set_ucast_flood(dev, port, true);
+	b53_port_set_mcast_flood(dev, port, true);
+	b53_port_set_learning(dev, port, false);
+
+	return 0;
+}
+EXPORT_SYMBOL(b53_setup_port);
+
 int b53_enable_port(struct dsa_switch *ds, int port, struct phy_device *phy)
 {
 	struct b53_device *dev = ds->priv;
@@ -588,10 +602,6 @@ int b53_enable_port(struct dsa_switch *ds, int port, struct phy_device *phy)
 
 	cpu_port = dsa_to_port(ds, port)->cpu_dp->index;
 
-	b53_port_set_ucast_flood(dev, port, true);
-	b53_port_set_mcast_flood(dev, port, true);
-	b53_port_set_learning(dev, port, false);
-
 	if (dev->ops->irq_enable)
 		ret = dev->ops->irq_enable(dev, port);
 	if (ret)
@@ -722,10 +732,6 @@ static void b53_enable_cpu_port(struct b53_device *dev, int port)
 	b53_write8(dev, B53_CTRL_PAGE, B53_PORT_CTRL(port), port_ctrl);
 
 	b53_brcm_hdr_setup(dev->ds, port);
-
-	b53_port_set_ucast_flood(dev, port, true);
-	b53_port_set_mcast_flood(dev, port, true);
-	b53_port_set_learning(dev, port, false);
 }
 
 static void b53_enable_mib(struct b53_device *dev)
@@ -761,6 +767,22 @@ static bool b53_vlan_port_needs_forced_tagged(struct dsa_switch *ds, int port)
 	return dev->tag_protocol == DSA_TAG_PROTO_NONE && dsa_is_cpu_port(ds, port);
 }
 
+static bool b53_vlan_port_may_join_untagged(struct dsa_switch *ds, int port)
+{
+	struct b53_device *dev = ds->priv;
+	struct dsa_port *dp;
+
+	if (!dev->vlan_filtering)
+		return true;
+
+	dp = dsa_to_port(ds, port);
+
+	if (dsa_port_is_cpu(dp))
+		return true;
+
+	return dp->bridge == NULL;
+}
+
 int b53_configure_vlan(struct dsa_switch *ds)
 {
 	struct b53_device *dev = ds->priv;
@@ -779,7 +801,7 @@ int b53_configure_vlan(struct dsa_switch *ds)
 		b53_do_vlan_op(dev, VTA_CMD_CLEAR);
 	}
 
-	b53_enable_vlan(dev, -1, dev->vlan_enabled, ds->vlan_filtering);
+	b53_enable_vlan(dev, -1, dev->vlan_enabled, dev->vlan_filtering);
 
 	/* Create an untagged VLAN entry for the default PVID in case
 	 * CONFIG_VLAN_8021Q is disabled and there are no calls to
@@ -787,26 +809,39 @@ int b53_configure_vlan(struct dsa_switch *ds)
 	 * entry. Do this only when the tagging protocol is not
 	 * DSA_TAG_PROTO_NONE
 	 */
+	v = &dev->vlans[def_vid];
 	b53_for_each_port(dev, i) {
-		v = &dev->vlans[def_vid];
+		if (!b53_vlan_port_may_join_untagged(ds, i))
-		v->members |= BIT(i);
-		if (!b53_vlan_port_needs_forced_tagged(ds, i))
-			v->untag = v->members;
-		b53_write16(dev, B53_VLAN_PAGE,
-			    B53_VLAN_PORT_DEF_TAG(i), def_vid);
-	}
-
-	/* Upon initial call we have not set-up any VLANs, but upon
-	 * system resume, we need to restore all VLAN entries.
-	 */
-	for (vid = def_vid; vid < dev->num_vlans; vid++) {
-		v = &dev->vlans[vid];
-
-		if (!v->members)
 			continue;
 
-		b53_set_vlan_entry(dev, vid, v);
+		vl.members |= BIT(i);
-		b53_fast_age_vlan(dev, vid);
+		if (!b53_vlan_port_needs_forced_tagged(ds, i))
+			vl.untag = vl.members;
+		b53_write16(dev, B53_VLAN_PAGE, B53_VLAN_PORT_DEF_TAG(i),
+			    def_vid);
+	}
+	b53_set_vlan_entry(dev, def_vid, &vl);
+
+	if (dev->vlan_filtering) {
+		/* Upon initial call we have not set-up any VLANs, but upon
+		 * system resume, we need to restore all VLAN entries.
+		 */
+		for (vid = def_vid + 1; vid < dev->num_vlans; vid++) {
+			v = &dev->vlans[vid];
+
+			if (!v->members)
+				continue;
+
+			b53_set_vlan_entry(dev, vid, v);
+			b53_fast_age_vlan(dev, vid);
+		}
+
+		b53_for_each_port(dev, i) {
+			if (!dsa_is_cpu_port(ds, i))
+				b53_write16(dev, B53_VLAN_PAGE,
+					    B53_VLAN_PORT_DEF_TAG(i),
+					    dev->ports[i].pvid);
+		}
 	}
 
 	return 0;
@@ -1126,7 +1161,9 @@ EXPORT_SYMBOL(b53_setup_devlink_resources);
 static int b53_setup(struct dsa_switch *ds)
 {
 	struct b53_device *dev = ds->priv;
+	struct b53_vlan *vl;
 	unsigned int port;
+	u16 pvid;
 	int ret;
 
 	/* Request bridge PVID untagged when DSA_TAG_PROTO_NONE is set
@@ -1134,12 +1171,26 @@ static int b53_setup(struct dsa_switch *ds)
 	 */
 	ds->untag_bridge_pvid = dev->tag_protocol == DSA_TAG_PROTO_NONE;
 
+	/* The switch does not tell us the original VLAN for untagged
+	 * packets, so keep the CPU port always tagged.
+	 */
+	ds->untag_vlan_aware_bridge_pvid = true;
+
 	ret = b53_reset_switch(dev);
 	if (ret) {
 		dev_err(ds->dev, "failed to reset switch\n");
 		return ret;
 	}
 
+	/* setup default vlan for filtering mode */
+	pvid = b53_default_pvid(dev);
+	vl = &dev->vlans[pvid];
+	b53_for_each_port(dev, port) {
+		vl->members |= BIT(port);
+		if (!b53_vlan_port_needs_forced_tagged(ds, port))
+			vl->untag |= BIT(port);
+	}
+
 	b53_reset_mib(dev);
 
 	ret = b53_apply_config(dev);
@@ -1493,7 +1544,10 @@ int b53_vlan_filtering(struct dsa_switch *ds, int port, bool vlan_filtering,
 {
 	struct b53_device *dev = ds->priv;
 
-	b53_enable_vlan(dev, port, dev->vlan_enabled, vlan_filtering);
+	if (dev->vlan_filtering != vlan_filtering) {
+		dev->vlan_filtering = vlan_filtering;
+		b53_apply_config(dev);
+	}
 
 	return 0;
 }
@@ -1518,7 +1572,7 @@ static int b53_vlan_prepare(struct dsa_switch *ds, int port,
 	if (vlan->vid >= dev->num_vlans)
 		return -ERANGE;
 
-	b53_enable_vlan(dev, port, true, ds->vlan_filtering);
+	b53_enable_vlan(dev, port, true, dev->vlan_filtering);
 
 	return 0;
 }
@@ -1531,18 +1585,29 @@ int b53_vlan_add(struct dsa_switch *ds, int port,
 	bool untagged = vlan->flags & BRIDGE_VLAN_INFO_UNTAGGED;
 	bool pvid = vlan->flags & BRIDGE_VLAN_INFO_PVID;
 	struct b53_vlan *vl;
+	u16 old_pvid, new_pvid;
 	int err;
 
 	err = b53_vlan_prepare(ds, port, vlan);
 	if (err)
 		return err;
 
+	if (vlan->vid == 0)
+		return 0;
+
+	old_pvid = dev->ports[port].pvid;
+	if (pvid)
+		new_pvid = vlan->vid;
+	else if (!pvid && vlan->vid == old_pvid)
+		new_pvid = b53_default_pvid(dev);
+	else
+		new_pvid = old_pvid;
+	dev->ports[port].pvid = new_pvid;
+
 	vl = &dev->vlans[vlan->vid];
 
-	b53_get_vlan_entry(dev, vlan->vid, vl);
+	if (dsa_is_cpu_port(ds, port))
-
+		untagged = false;
-	if (vlan->vid == 0 && vlan->vid == b53_default_pvid(dev))
-		untagged = true;
 
 	vl->members |= BIT(port);
 	if (untagged && !b53_vlan_port_needs_forced_tagged(ds, port))
@@ -1550,13 +1615,16 @@ int b53_vlan_add(struct dsa_switch *ds, int port,
 	else
 		vl->untag &= ~BIT(port);
 
+	if (!dev->vlan_filtering)
+		return 0;
+
 	b53_set_vlan_entry(dev, vlan->vid, vl);
 	b53_fast_age_vlan(dev, vlan->vid);
 
-	if (pvid && !dsa_is_cpu_port(ds, port)) {
+	if (!dsa_is_cpu_port(ds, port) && new_pvid != old_pvid) {
 		b53_write16(dev, B53_VLAN_PAGE, B53_VLAN_PORT_DEF_TAG(port),
-			    vlan->vid);
+			    new_pvid);
-		b53_fast_age_vlan(dev, vlan->vid);
+		b53_fast_age_vlan(dev, old_pvid);
 	}
 
 	return 0;
@@ -1571,20 +1639,25 @@ int b53_vlan_del(struct dsa_switch *ds, int port,
 	struct b53_vlan *vl;
 	u16 pvid;
 
-	b53_read16(dev, B53_VLAN_PAGE, B53_VLAN_PORT_DEF_TAG(port), &pvid);
+	if (vlan->vid == 0)
+		return 0;
+
+	pvid = dev->ports[port].pvid;
 
 	vl = &dev->vlans[vlan->vid];
 
-	b53_get_vlan_entry(dev, vlan->vid, vl);
-
 	vl->members &= ~BIT(port);
 
 	if (pvid == vlan->vid)
 		pvid = b53_default_pvid(dev);
+	dev->ports[port].pvid = pvid;
 
 	if (untagged && !b53_vlan_port_needs_forced_tagged(ds, port))
 		vl->untag &= ~(BIT(port));
 
+	if (!dev->vlan_filtering)
+		return 0;
+
 	b53_set_vlan_entry(dev, vlan->vid, vl);
 	b53_fast_age_vlan(dev, vlan->vid);
 
@@ -1917,8 +1990,9 @@ int b53_br_join(struct dsa_switch *ds, int port, struct dsa_bridge bridge,
 		bool *tx_fwd_offload, struct netlink_ext_ack *extack)
 {
 	struct b53_device *dev = ds->priv;
+	struct b53_vlan *vl;
 	s8 cpu_port = dsa_to_port(ds, port)->cpu_dp->index;
-	u16 pvlan, reg;
+	u16 pvlan, reg, pvid;
 	unsigned int i;
 
 	/* On 7278, port 7 which connects to the ASP should only receive
@@ -1927,15 +2001,29 @@ int b53_br_join(struct dsa_switch *ds, int port, struct dsa_bridge bridge,
 	if (dev->chip_id == BCM7278_DEVICE_ID && port == 7)
 		return -EINVAL;
 
-	/* Make this port leave the all VLANs join since we will have proper
+	pvid = b53_default_pvid(dev);
-	 * VLAN entries from now on
+	vl = &dev->vlans[pvid];
-	 */
+
-	if (is58xx(dev)) {
+	if (dev->vlan_filtering) {
-		b53_read16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, &reg);
+		/* Make this port leave the all VLANs join since we will have
-		reg &= ~BIT(port);
+		 * proper VLAN entries from now on
-		if ((reg & BIT(cpu_port)) == BIT(cpu_port))
+		 */
-			reg &= ~BIT(cpu_port);
+		if (is58xx(dev)) {
-		b53_write16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, reg);
+			b53_read16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN,
+				   &reg);
+			reg &= ~BIT(port);
+			if ((reg & BIT(cpu_port)) == BIT(cpu_port))
+				reg &= ~BIT(cpu_port);
+			b53_write16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN,
+				    reg);
+		}
+
+		b53_get_vlan_entry(dev, pvid, vl);
+		vl->members &= ~BIT(port);
+		if (vl->members == BIT(cpu_port))
+			vl->members &= ~BIT(cpu_port);
+		vl->untag = vl->members;
+		b53_set_vlan_entry(dev, pvid, vl);
 	}
 
 	b53_read16(dev, B53_PVLAN_PAGE, B53_PVLAN_PORT_MASK(port), &pvlan);
@@ -1968,7 +2056,7 @@ EXPORT_SYMBOL(b53_br_join);
 void b53_br_leave(struct dsa_switch *ds, int port, struct dsa_bridge bridge)
 {
 	struct b53_device *dev = ds->priv;
-	struct b53_vlan *vl = &dev->vlans[0];
+	struct b53_vlan *vl;
 	s8 cpu_port = dsa_to_port(ds, port)->cpu_dp->index;
 	unsigned int i;
 	u16 pvlan, reg, pvid;
@@ -1994,15 +2082,18 @@ void b53_br_leave(struct dsa_switch *ds, int port, struct dsa_bridge bridge)
 	dev->ports[port].vlan_ctl_mask = pvlan;
 
 	pvid = b53_default_pvid(dev);
+	vl = &dev->vlans[pvid];
+
+	if (dev->vlan_filtering) {
+		/* Make this port join all VLANs without VLAN entries */
+		if (is58xx(dev)) {
+			b53_read16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, &reg);
+			reg |= BIT(port);
+			if (!(reg & BIT(cpu_port)))
+				reg |= BIT(cpu_port);
+			b53_write16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, reg);
+		}
 
-	/* Make this port join all VLANs without VLAN entries */
-	if (is58xx(dev)) {
-		b53_read16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, &reg);
-		reg |= BIT(port);
-		if (!(reg & BIT(cpu_port)))
-			reg |= BIT(cpu_port);
-		b53_write16(dev, B53_VLAN_PAGE, B53_JOIN_ALL_VLAN_EN, reg);
-	} else {
 		b53_get_vlan_entry(dev, pvid, vl);
 		vl->members |= BIT(port) | BIT(cpu_port);
 		vl->untag |= BIT(port) | BIT(cpu_port);
@@ -2307,6 +2398,7 @@ static const struct dsa_switch_ops b53_switch_ops = {
 	.phy_read		= b53_phy_read16,
 	.phy_write		= b53_phy_write16,
 	.phylink_get_caps	= b53_phylink_get_caps,
+	.port_setup		= b53_setup_port,
 	.port_enable		= b53_enable_port,
 	.port_disable		= b53_disable_port,
 	.get_mac_eee		= b53_get_mac_eee,
@@ -2751,6 +2843,7 @@ struct b53_device *b53_switch_alloc(struct device *base,
 	ds->ops = &b53_switch_ops;
 	ds->phylink_mac_ops = &b53_phylink_mac_ops;
 	dev->vlan_enabled = true;
+	dev->vlan_filtering = false;
 	/* Let DSA handle the case were multiple bridges span the same switch
 	 * device and different VLAN awareness settings are requested, which
 	 * would be breaking filtering semantics for any of the other bridge

drivers/net/dsa/b53/b53_priv.h

@@ -95,6 +95,7 @@ struct b53_pcs {
 
 struct b53_port {
 	u16		vlan_ctl_mask;
+	u16		pvid;
 	struct ethtool_keee eee;
 };
 
@@ -146,6 +147,7 @@ struct b53_device {
 	unsigned int num_vlans;
 	struct b53_vlan *vlans;
 	bool vlan_enabled;
+	bool vlan_filtering;
 	unsigned int num_ports;
 	struct b53_port *ports;
 
@@ -380,6 +382,7 @@ enum dsa_tag_protocol b53_get_tag_protocol(struct dsa_switch *ds, int port,
 					   enum dsa_tag_protocol mprot);
 void b53_mirror_del(struct dsa_switch *ds, int port,
 		    struct dsa_mall_mirror_tc_entry *mirror);
+int b53_setup_port(struct dsa_switch *ds, int port);
 int b53_enable_port(struct dsa_switch *ds, int port, struct phy_device *phy);
 void b53_disable_port(struct dsa_switch *ds, int port);
 void b53_brcm_hdr_setup(struct dsa_switch *ds, int port);

drivers/net/dsa/bcm_sf2.c

@@ -1230,6 +1230,7 @@ static const struct dsa_switch_ops bcm_sf2_ops = {
 	.resume			= bcm_sf2_sw_resume,
 	.get_wol		= bcm_sf2_sw_get_wol,
 	.set_wol		= bcm_sf2_sw_set_wol,
+	.port_setup		= b53_setup_port,
 	.port_enable		= bcm_sf2_port_setup,
 	.port_disable		= bcm_sf2_port_disable,
 	.get_mac_eee		= b53_get_mac_eee,

drivers/net/ethernet/meta/fbnic/fbnic_csr.h

@@ -700,8 +700,10 @@ enum {
 /* PUL User Registers */
 #define FBNIC_CSR_START_PUL_USER	0x31000	/* CSR section delimiter */
 #define FBNIC_PUL_OB_TLP_HDR_AW_CFG	0x3103d		/* 0xc40f4 */
+#define FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH	CSR_BIT(19)
 #define FBNIC_PUL_OB_TLP_HDR_AW_CFG_BME		CSR_BIT(18)
 #define FBNIC_PUL_OB_TLP_HDR_AR_CFG	0x3103e		/* 0xc40f8 */
+#define FBNIC_PUL_OB_TLP_HDR_AR_CFG_FLUSH	CSR_BIT(19)
 #define FBNIC_PUL_OB_TLP_HDR_AR_CFG_BME		CSR_BIT(18)
 #define FBNIC_CSR_END_PUL_USER	0x31080	/* CSR section delimiter */
 

drivers/net/ethernet/meta/fbnic/fbnic_fw.c

@@ -17,11 +17,29 @@ static void __fbnic_mbx_wr_desc(struct fbnic_dev *fbd, int mbx_idx,
 {
 	u32 desc_offset = FBNIC_IPC_MBX(mbx_idx, desc_idx);
 
+	/* Write the upper 32b and then the lower 32b. Doing this the
+	 * FW can then read lower, upper, lower to verify that the state
+	 * of the descriptor wasn't changed mid-transaction.
+	 */
 	fw_wr32(fbd, desc_offset + 1, upper_32_bits(desc));
 	fw_wrfl(fbd);
 	fw_wr32(fbd, desc_offset, lower_32_bits(desc));
 }
 
+static void __fbnic_mbx_invalidate_desc(struct fbnic_dev *fbd, int mbx_idx,
+					int desc_idx, u32 desc)
+{
+	u32 desc_offset = FBNIC_IPC_MBX(mbx_idx, desc_idx);
+
+	/* For initialization we write the lower 32b of the descriptor first.
+	 * This way we can set the state to mark it invalid before we clear the
+	 * upper 32b.
+	 */
+	fw_wr32(fbd, desc_offset, desc);
+	fw_wrfl(fbd);
+	fw_wr32(fbd, desc_offset + 1, 0);
+}
+
 static u64 __fbnic_mbx_rd_desc(struct fbnic_dev *fbd, int mbx_idx, int desc_idx)
 {
 	u32 desc_offset = FBNIC_IPC_MBX(mbx_idx, desc_idx);
@@ -33,29 +51,41 @@ static u64 __fbnic_mbx_rd_desc(struct fbnic_dev *fbd, int mbx_idx, int desc_idx)
 	return desc;
 }
 
-static void fbnic_mbx_init_desc_ring(struct fbnic_dev *fbd, int mbx_idx)
+static void fbnic_mbx_reset_desc_ring(struct fbnic_dev *fbd, int mbx_idx)
 {
 	int desc_idx;
 
+	/* Disable DMA transactions from the device,
+	 * and flush any transactions triggered during cleaning
+	 */
+	switch (mbx_idx) {
+	case FBNIC_IPC_MBX_RX_IDX:
+		wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AW_CFG,
+		     FBNIC_PUL_OB_TLP_HDR_AW_CFG_FLUSH);
+		break;
+	case FBNIC_IPC_MBX_TX_IDX:
+		wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AR_CFG,
+		     FBNIC_PUL_OB_TLP_HDR_AR_CFG_FLUSH);
+		break;
+	}
+
+	wrfl(fbd);
+
 	/* Initialize first descriptor to all 0s. Doing this gives us a
 	 * solid stop for the firmware to hit when it is done looping
 	 * through the ring.
 	 */
-	__fbnic_mbx_wr_desc(fbd, mbx_idx, 0, 0);
+	__fbnic_mbx_invalidate_desc(fbd, mbx_idx, 0, 0);
-
-	fw_wrfl(fbd);
 
 	/* We then fill the rest of the ring starting at the end and moving
 	 * back toward descriptor 0 with skip descriptors that have no
 	 * length nor address, and tell the firmware that they can skip
 	 * them and just move past them to the one we initialized to 0.
 	 */
-	for (desc_idx = FBNIC_IPC_MBX_DESC_LEN; --desc_idx;) {
+	for (desc_idx = FBNIC_IPC_MBX_DESC_LEN; --desc_idx;)
-		__fbnic_mbx_wr_desc(fbd, mbx_idx, desc_idx,
+		__fbnic_mbx_invalidate_desc(fbd, mbx_idx, desc_idx,
-				    FBNIC_IPC_MBX_DESC_FW_CMPL |
+					    FBNIC_IPC_MBX_DESC_FW_CMPL |
-				    FBNIC_IPC_MBX_DESC_HOST_CMPL);
+					    FBNIC_IPC_MBX_DESC_HOST_CMPL);
-		fw_wrfl(fbd);
-	}
 }
 
 void fbnic_mbx_init(struct fbnic_dev *fbd)
@@ -76,7 +106,7 @@ void fbnic_mbx_init(struct fbnic_dev *fbd)
 	wr32(fbd, FBNIC_INTR_CLEAR(0), 1u << FBNIC_FW_MSIX_ENTRY);
 
 	for (i = 0; i < FBNIC_IPC_MBX_INDICES; i++)
-		fbnic_mbx_init_desc_ring(fbd, i);
+		fbnic_mbx_reset_desc_ring(fbd, i);
 }
 
 static int fbnic_mbx_map_msg(struct fbnic_dev *fbd, int mbx_idx,
@@ -141,7 +171,7 @@ static void fbnic_mbx_clean_desc_ring(struct fbnic_dev *fbd, int mbx_idx)
 {
 	int i;
 
-	fbnic_mbx_init_desc_ring(fbd, mbx_idx);
+	fbnic_mbx_reset_desc_ring(fbd, mbx_idx);
 
 	for (i = FBNIC_IPC_MBX_DESC_LEN; i--;)
 		fbnic_mbx_unmap_and_free_msg(fbd, mbx_idx, i);
@@ -265,67 +295,41 @@ static int fbnic_fw_xmit_simple_msg(struct fbnic_dev *fbd, u32 msg_type)
 	return err;
 }
 
-/**
+static void fbnic_mbx_init_desc_ring(struct fbnic_dev *fbd, int mbx_idx)
- * fbnic_fw_xmit_cap_msg - Allocate and populate a FW capabilities message
- * @fbd: FBNIC device structure
- *
- * Return: NULL on failure to allocate, error pointer on error, or pointer
- * to new TLV test message.
- *
- * Sends a single TLV header indicating the host wants the firmware to
- * confirm the capabilities and version.
- **/
-static int fbnic_fw_xmit_cap_msg(struct fbnic_dev *fbd)
-{
-	int err = fbnic_fw_xmit_simple_msg(fbd, FBNIC_TLV_MSG_ID_HOST_CAP_REQ);
-
-	/* Return 0 if we are not calling this on ASIC */
-	return (err == -EOPNOTSUPP) ? 0 : err;
-}
-
-static void fbnic_mbx_postinit_desc_ring(struct fbnic_dev *fbd, int mbx_idx)
 {
 	struct fbnic_fw_mbx *mbx = &fbd->mbx[mbx_idx];
 
-	/* This is a one time init, so just exit if it is completed */
-	if (mbx->ready)
-		return;
-
 	mbx->ready = true;
 
 	switch (mbx_idx) {
 	case FBNIC_IPC_MBX_RX_IDX:
+		/* Enable DMA writes from the device */
+		wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AW_CFG,
+		     FBNIC_PUL_OB_TLP_HDR_AW_CFG_BME);
+
 		/* Make sure we have a page for the FW to write to */
 		fbnic_mbx_alloc_rx_msgs(fbd);
 		break;
 	case FBNIC_IPC_MBX_TX_IDX:
-		/* Force version to 1 if we successfully requested an update
+		/* Enable DMA reads from the device */
-		 * from the firmware. This should be overwritten once we get
+		wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AR_CFG,
-		 * the actual version from the firmware in the capabilities
+		     FBNIC_PUL_OB_TLP_HDR_AR_CFG_BME);
-		 * request message.
-		 */
-		if (!fbnic_fw_xmit_cap_msg(fbd) &&
-		    !fbd->fw_cap.running.mgmt.version)
-			fbd->fw_cap.running.mgmt.version = 1;
 		break;
 	}
 }
 
-static void fbnic_mbx_postinit(struct fbnic_dev *fbd)
+static bool fbnic_mbx_event(struct fbnic_dev *fbd)
 {
-	int i;
+	/* We only need to do this on the first interrupt following reset.
-
-	/* We only need to do this on the first interrupt following init.
 	 * this primes the mailbox so that we will have cleared all the
 	 * skip descriptors.
 	 */
 	if (!(rd32(fbd, FBNIC_INTR_STATUS(0)) & (1u << FBNIC_FW_MSIX_ENTRY)))
-		return;
+		return false;
 
 	wr32(fbd, FBNIC_INTR_CLEAR(0), 1u << FBNIC_FW_MSIX_ENTRY);
 
-	for (i = 0; i < FBNIC_IPC_MBX_INDICES; i++)
+	return true;
-		fbnic_mbx_postinit_desc_ring(fbd, i);
 }
 
 /**
@@ -726,7 +730,7 @@ next_page:
 
 void fbnic_mbx_poll(struct fbnic_dev *fbd)
 {
-	fbnic_mbx_postinit(fbd);
+	fbnic_mbx_event(fbd);
 
 	fbnic_mbx_process_tx_msgs(fbd);
 	fbnic_mbx_process_rx_msgs(fbd);
@@ -734,60 +738,80 @@ void fbnic_mbx_poll(struct fbnic_dev *fbd)
 
 int fbnic_mbx_poll_tx_ready(struct fbnic_dev *fbd)
 {
-	struct fbnic_fw_mbx *tx_mbx;
+	unsigned long timeout = jiffies + 10 * HZ + 1;
-	int attempts = 50;
+	int err, i;
 
-	/* Immediate fail if BAR4 isn't there */
+	do {
-	if (!fbnic_fw_present(fbd))
+		if (!time_is_after_jiffies(timeout))
-		return -ENODEV;
+			return -ETIMEDOUT;
 
-	tx_mbx = &fbd->mbx[FBNIC_IPC_MBX_TX_IDX];
-	while (!tx_mbx->ready && --attempts) {
 		/* Force the firmware to trigger an interrupt response to
 		 * avoid the mailbox getting stuck closed if the interrupt
 		 * is reset.
 		 */
-		fbnic_mbx_init_desc_ring(fbd, FBNIC_IPC_MBX_TX_IDX);
+		fbnic_mbx_reset_desc_ring(fbd, FBNIC_IPC_MBX_TX_IDX);
 
-		msleep(200);
+		/* Immediate fail if BAR4 went away */
+		if (!fbnic_fw_present(fbd))
+			return -ENODEV;
 
-		fbnic_mbx_poll(fbd);
+		msleep(20);
-	}
+	} while (!fbnic_mbx_event(fbd));
 
-	return attempts ? 0 : -ETIMEDOUT;
+	/* FW has shown signs of life. Enable DMA and start Tx/Rx */
+	for (i = 0; i < FBNIC_IPC_MBX_INDICES; i++)
+		fbnic_mbx_init_desc_ring(fbd, i);
+
+	/* Request an update from the firmware. This should overwrite
+	 * mgmt.version once we get the actual version from the firmware
+	 * in the capabilities request message.
+	 */
+	err = fbnic_fw_xmit_simple_msg(fbd, FBNIC_TLV_MSG_ID_HOST_CAP_REQ);
+	if (err)
+		goto clean_mbx;
+
+	/* Use "1" to indicate we entered the state waiting for a response */
+	fbd->fw_cap.running.mgmt.version = 1;
+
+	return 0;
+clean_mbx:
+	/* Cleanup Rx buffers and disable mailbox */
+	fbnic_mbx_clean(fbd);
+	return err;
 }
 
 void fbnic_mbx_flush_tx(struct fbnic_dev *fbd)
 {
+	unsigned long timeout = jiffies + 10 * HZ + 1;
 	struct fbnic_fw_mbx *tx_mbx;
-	int attempts = 50;
+	u8 tail;
-	u8 count = 0;
-
-	/* Nothing to do if there is no mailbox */
-	if (!fbnic_fw_present(fbd))
-		return;
 
 	/* Record current Rx stats */
 	tx_mbx = &fbd->mbx[FBNIC_IPC_MBX_TX_IDX];
 
-	/* Nothing to do if mailbox never got to ready */
+	spin_lock_irq(&fbd->fw_tx_lock);
-	if (!tx_mbx->ready)
+
-		return;
+	/* Clear ready to prevent any further attempts to transmit */
+	tx_mbx->ready = false;
+
+	/* Read tail to determine the last tail state for the ring */
+	tail = tx_mbx->tail;
+
+	spin_unlock_irq(&fbd->fw_tx_lock);
 
 	/* Give firmware time to process packet,
-	 * we will wait up to 10 seconds which is 50 waits of 200ms.
+	 * we will wait up to 10 seconds which is 500 waits of 20ms.
 	 */
 	do {
 		u8 head = tx_mbx->head;
 
-		if (head == tx_mbx->tail)
+		/* Tx ring is empty once head == tail */
+		if (head == tail)
 			break;
 
-		msleep(200);
+		msleep(20);
 		fbnic_mbx_process_tx_msgs(fbd);
-
+	} while (time_is_after_jiffies(timeout));
-		count += (tx_mbx->head - head) % FBNIC_IPC_MBX_DESC_LEN;
-	} while (count < FBNIC_IPC_MBX_DESC_LEN && --attempts);
 }
 
 void fbnic_get_fw_ver_commit_str(struct fbnic_dev *fbd, char *fw_version,

drivers/net/ethernet/meta/fbnic/fbnic_mac.c

@@ -79,12 +79,6 @@ static void fbnic_mac_init_axi(struct fbnic_dev *fbd)
 	fbnic_init_readrq(fbd, FBNIC_QM_RNI_RBP_CTL, cls, readrq);
 	fbnic_init_mps(fbd, FBNIC_QM_RNI_RDE_CTL, cls, mps);
 	fbnic_init_mps(fbd, FBNIC_QM_RNI_RCM_CTL, cls, mps);
-
-	/* Enable XALI AR/AW outbound */
-	wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AW_CFG,
-	     FBNIC_PUL_OB_TLP_HDR_AW_CFG_BME);
-	wr32(fbd, FBNIC_PUL_OB_TLP_HDR_AR_CFG,
-	     FBNIC_PUL_OB_TLP_HDR_AR_CFG_BME);
 }
 
 static void fbnic_mac_init_qm(struct fbnic_dev *fbd)

fs/erofs/zdata.c

@@ -82,9 +82,6 @@ struct z_erofs_pcluster {
 	/* L: whether partial decompression or not */
 	bool partial;
 
-	/* L: indicate several pageofs_outs or not */
-	bool multibases;
-
 	/* L: whether extra buffer allocations are best-effort */
 	bool besteffort;
 
@@ -1073,8 +1070,6 @@ static int z_erofs_scan_folio(struct z_erofs_decompress_frontend *f,
 				break;
 
 			erofs_onlinefolio_split(folio);
-			if (f->pcl->pageofs_out != (map->m_la & ~PAGE_MASK))
-				f->pcl->multibases = true;
 			if (f->pcl->length < offset + end - map->m_la) {
 				f->pcl->length = offset + end - map->m_la;
 				f->pcl->pageofs_out = map->m_la & ~PAGE_MASK;
@@ -1120,7 +1115,6 @@ struct z_erofs_decompress_backend {
 	struct page *onstack_pages[Z_EROFS_ONSTACK_PAGES];
 	struct super_block *sb;
 	struct z_erofs_pcluster *pcl;
-
 	/* pages with the longest decompressed length for deduplication */
 	struct page **decompressed_pages;
 	/* pages to keep the compressed data */
@@ -1129,6 +1123,8 @@ struct z_erofs_decompress_backend {
 	struct list_head decompressed_secondary_bvecs;
 	struct page **pagepool;
 	unsigned int onstack_used, nr_pages;
+	/* indicate if temporary copies should be preserved for later use */
+	bool keepxcpy;
 };
 
 struct z_erofs_bvec_item {
@@ -1139,18 +1135,20 @@ struct z_erofs_bvec_item {
 static void z_erofs_do_decompressed_bvec(struct z_erofs_decompress_backend *be,
 					 struct z_erofs_bvec *bvec)
 {
+	int poff = bvec->offset + be->pcl->pageofs_out;
 	struct z_erofs_bvec_item *item;
-	unsigned int pgnr;
+	struct page **page;
 
-	if (!((bvec->offset + be->pcl->pageofs_out) & ~PAGE_MASK) &&
+	if (!(poff & ~PAGE_MASK) && (bvec->end == PAGE_SIZE ||
-	    (bvec->end == PAGE_SIZE ||
+			bvec->offset + bvec->end == be->pcl->length)) {
-	     bvec->offset + bvec->end == be->pcl->length)) {
+		DBG_BUGON((poff >> PAGE_SHIFT) >= be->nr_pages);
-		pgnr = (bvec->offset + be->pcl->pageofs_out) >> PAGE_SHIFT;
+		page = be->decompressed_pages + (poff >> PAGE_SHIFT);
-		DBG_BUGON(pgnr >= be->nr_pages);
+		if (!*page) {
-		if (!be->decompressed_pages[pgnr]) {
+			*page = bvec->page;
-			be->decompressed_pages[pgnr] = bvec->page;
 			return;
 		}
+	} else {
+		be->keepxcpy = true;
 	}
 
 	/* (cold path) one pcluster is requested multiple times */
@@ -1316,7 +1314,7 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
 					.alg = pcl->algorithmformat,
 					.inplace_io = overlapped,
 					.partial_decoding = pcl->partial,
-					.fillgaps = pcl->multibases,
+					.fillgaps = be->keepxcpy,
 					.gfp = pcl->besteffort ? GFP_KERNEL :
 						GFP_NOWAIT | __GFP_NORETRY
 				 }, be->pagepool);
@@ -1370,7 +1368,6 @@ static int z_erofs_decompress_pcluster(struct z_erofs_decompress_backend *be,
 
 	pcl->length = 0;
 	pcl->partial = true;
-	pcl->multibases = false;
 	pcl->besteffort = false;
 	pcl->bvset.nextpage = NULL;
 	pcl->vcnt = 0;

include/net/netdev_queues.h

@@ -92,6 +92,12 @@ struct netdev_stat_ops {
 			       struct netdev_queue_stats_tx *tx);
 };
 
+void netdev_stat_queue_sum(struct net_device *netdev,
+			   int rx_start, int rx_end,
+			   struct netdev_queue_stats_rx *rx_sum,
+			   int tx_start, int tx_end,
+			   struct netdev_queue_stats_tx *tx_sum);
+
 /**
  * struct netdev_queue_mgmt_ops - netdev ops for queue management
  *

net/core/filter.c

@@ -2534,6 +2534,7 @@ int skb_do_redirect(struct sk_buff *skb)
 			goto out_drop;
 		skb->dev = dev;
 		dev_sw_netstats_rx_add(dev, skb->len);
+		skb_scrub_packet(skb, false);
 		return -EAGAIN;
 	}
 	return flags & BPF_F_NEIGH ?

net/core/netdev-genl.c

@@ -616,25 +616,66 @@ netdev_nl_stats_by_queue(struct net_device *netdev, struct sk_buff *rsp,
 	return 0;
 }
 
+/**
+ * netdev_stat_queue_sum() - add up queue stats from range of queues
+ * @netdev:	net_device
+ * @rx_start:	index of the first Rx queue to query
+ * @rx_end:	index after the last Rx queue (first *not* to query)
+ * @rx_sum:	output Rx stats, should be already initialized
+ * @tx_start:	index of the first Tx queue to query
+ * @tx_end:	index after the last Tx queue (first *not* to query)
+ * @tx_sum:	output Tx stats, should be already initialized
+ *
+ * Add stats from [start, end) range of queue IDs to *x_sum structs.
+ * The sum structs must be already initialized. Usually this
+ * helper is invoked from the .get_base_stats callbacks of drivers
+ * to account for stats of disabled queues. In that case the ranges
+ * are usually [netdev->real_num_*x_queues, netdev->num_*x_queues).
+ */
+void netdev_stat_queue_sum(struct net_device *netdev,
+			   int rx_start, int rx_end,
+			   struct netdev_queue_stats_rx *rx_sum,
+			   int tx_start, int tx_end,
+			   struct netdev_queue_stats_tx *tx_sum)
+{
+	const struct netdev_stat_ops *ops;
+	struct netdev_queue_stats_rx rx;
+	struct netdev_queue_stats_tx tx;
+	int i;
+
+	ops = netdev->stat_ops;
+
+	for (i = rx_start; i < rx_end; i++) {
+		memset(&rx, 0xff, sizeof(rx));
+		if (ops->get_queue_stats_rx)
+			ops->get_queue_stats_rx(netdev, i, &rx);
+		netdev_nl_stats_add(rx_sum, &rx, sizeof(rx));
+	}
+	for (i = tx_start; i < tx_end; i++) {
+		memset(&tx, 0xff, sizeof(tx));
+		if (ops->get_queue_stats_tx)
+			ops->get_queue_stats_tx(netdev, i, &tx);
+		netdev_nl_stats_add(tx_sum, &tx, sizeof(tx));
+	}
+}
+EXPORT_SYMBOL(netdev_stat_queue_sum);
+
 static int
 netdev_nl_stats_by_netdev(struct net_device *netdev, struct sk_buff *rsp,
 			  const struct genl_info *info)
 {
-	struct netdev_queue_stats_rx rx_sum, rx;
+	struct netdev_queue_stats_rx rx_sum;
-	struct netdev_queue_stats_tx tx_sum, tx;
+	struct netdev_queue_stats_tx tx_sum;
-	const struct netdev_stat_ops *ops;
 	void *hdr;
-	int i;
 
-	ops = netdev->stat_ops;
 	/* Netdev can't guarantee any complete counters */
-	if (!ops->get_base_stats)
+	if (!netdev->stat_ops->get_base_stats)
 		return 0;
 
 	memset(&rx_sum, 0xff, sizeof(rx_sum));
 	memset(&tx_sum, 0xff, sizeof(tx_sum));
 
-	ops->get_base_stats(netdev, &rx_sum, &tx_sum);
+	netdev->stat_ops->get_base_stats(netdev, &rx_sum, &tx_sum);
 
 	/* The op was there, but nothing reported, don't bother */
 	if (!memchr_inv(&rx_sum, 0xff, sizeof(rx_sum)) &&
@@ -647,18 +688,8 @@ netdev_nl_stats_by_netdev(struct net_device *netdev, struct sk_buff *rsp,
 	if (nla_put_u32(rsp, NETDEV_A_QSTATS_IFINDEX, netdev->ifindex))
 		goto nla_put_failure;
 
-	for (i = 0; i < netdev->real_num_rx_queues; i++) {
+	netdev_stat_queue_sum(netdev, 0, netdev->real_num_rx_queues, &rx_sum,
-		memset(&rx, 0xff, sizeof(rx));
+			      0, netdev->real_num_tx_queues, &tx_sum);
-		if (ops->get_queue_stats_rx)
-			ops->get_queue_stats_rx(netdev, i, &rx);
-		netdev_nl_stats_add(&rx_sum, &rx, sizeof(rx));
-	}
-	for (i = 0; i < netdev->real_num_tx_queues; i++) {
-		memset(&tx, 0xff, sizeof(tx));
-		if (ops->get_queue_stats_tx)
-			ops->get_queue_stats_tx(netdev, i, &tx);
-		netdev_nl_stats_add(&tx_sum, &tx, sizeof(tx));
-	}
 
 	if (netdev_nl_stats_write_rx(rsp, &rx_sum) ||
 	    netdev_nl_stats_write_tx(rsp, &tx_sum))

net/netfilter/ipset/ip_set_hash_gen.h

@@ -64,7 +64,7 @@ struct hbucket {
 #define ahash_sizeof_regions(htable_bits)		\
 	(ahash_numof_locks(htable_bits) * sizeof(struct ip_set_region))
 #define ahash_region(n, htable_bits)		\
-	((n) % ahash_numof_locks(htable_bits))
+	((n) / jhash_size(HTABLE_REGION_BITS))
 #define ahash_bucket_start(h,  htable_bits)	\
 	((htable_bits) < HTABLE_REGION_BITS ? 0	\
 		: (h) * jhash_size(HTABLE_REGION_BITS))

net/netfilter/ipvs/ip_vs_xmit.c

@@ -119,13 +119,12 @@ __mtu_check_toobig_v6(const struct sk_buff *skb, u32 mtu)
 	return false;
 }
 
-/* Get route to daddr, update *saddr, optionally bind route to saddr */
+/* Get route to daddr, optionally bind route to saddr */
 static struct rtable *do_output_route4(struct net *net, __be32 daddr,
-				       int rt_mode, __be32 *saddr)
+				       int rt_mode, __be32 *ret_saddr)
 {
 	struct flowi4 fl4;
 	struct rtable *rt;
-	bool loop = false;
 
 	memset(&fl4, 0, sizeof(fl4));
 	fl4.daddr = daddr;
@@ -135,23 +134,17 @@ static struct rtable *do_output_route4(struct net *net, __be32 daddr,
 retry:
 	rt = ip_route_output_key(net, &fl4);
 	if (IS_ERR(rt)) {
-		/* Invalid saddr ? */
-		if (PTR_ERR(rt) == -EINVAL && *saddr &&
-		    rt_mode & IP_VS_RT_MODE_CONNECT && !loop) {
-			*saddr = 0;
-			flowi4_update_output(&fl4, 0, daddr, 0);
-			goto retry;
-		}
 		IP_VS_DBG_RL("ip_route_output error, dest: %pI4\n", &daddr);
 		return NULL;
-	} else if (!*saddr && rt_mode & IP_VS_RT_MODE_CONNECT && fl4.saddr) {
+	}
+	if (rt_mode & IP_VS_RT_MODE_CONNECT && fl4.saddr) {
 		ip_rt_put(rt);
-		*saddr = fl4.saddr;
 		flowi4_update_output(&fl4, 0, daddr, fl4.saddr);
-		loop = true;
+		rt_mode = 0;
 		goto retry;
 	}
-	*saddr = fl4.saddr;
+	if (ret_saddr)
+		*ret_saddr = fl4.saddr;
 	return rt;
 }
 
@@ -344,19 +337,15 @@ __ip_vs_get_out_rt(struct netns_ipvs *ipvs, int skb_af, struct sk_buff *skb,
 		if (ret_saddr)
 			*ret_saddr = dest_dst->dst_saddr.ip;
 	} else {
-		__be32 saddr = htonl(INADDR_ANY);
-
 		noref = 0;
 
 		/* For such unconfigured boxes avoid many route lookups
 		 * for performance reasons because we do not remember saddr
 		 */
 		rt_mode &= ~IP_VS_RT_MODE_CONNECT;
-		rt = do_output_route4(net, daddr, rt_mode, &saddr);
+		rt = do_output_route4(net, daddr, rt_mode, ret_saddr);
 		if (!rt)
 			goto err_unreach;
-		if (ret_saddr)
-			*ret_saddr = saddr;
 	}
 
 	local = (rt->rt_flags & RTCF_LOCAL) ? 1 : 0;