Merge tag 'kvm-riscv-6.1-1' of https://github.com/kvm-riscv/linux into HEAD

KVM/riscv changes for 6.1

- Improved instruction encoding infrastructure for
  instructions not yet supported by binutils
- Svinval support for both KVM Host and KVM Guest
- Zihintpause support for KVM Guest
- Zicbom support for KVM Guest
- Record number of signal exits as a VCPU stat
- Use generic guest entry infrastructure

8 files changed, 104 insertions, 45 deletions

diff --git a/drivers/net/ethernet/intel/i40e/i40e_main.c b/drivers/net/ethernet/intel/i40e/i40e_main.c
index 10c1e1ea83a1..e3d9804aeb25 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_main.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_main.c
@@ -5909,6 +5909,26 @@ static int i40e_get_link_speed(struct i40e_vsi *vsi)
 }
 
 /**
+ * i40e_bw_bytes_to_mbits - Convert max_tx_rate from bytes to mbits
+ * @vsi: Pointer to vsi structure
+ * @max_tx_rate: max TX rate in bytes to be converted into Mbits
+ *
+ * Helper function to convert units before send to set BW limit
+ **/
+static u64 i40e_bw_bytes_to_mbits(struct i40e_vsi *vsi, u64 max_tx_rate)
+{
+	if (max_tx_rate < I40E_BW_MBPS_DIVISOR) {
+		dev_warn(&vsi->back->pdev->dev,
+			 "Setting max tx rate to minimum usable value of 50Mbps.\n");
+		max_tx_rate = I40E_BW_CREDIT_DIVISOR;
+	} else {
+		do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR);
+	}
+
+	return max_tx_rate;
+}
+
+/**
  * i40e_set_bw_limit - setup BW limit for Tx traffic based on max_tx_rate
  * @vsi: VSI to be configured
  * @seid: seid of the channel/VSI
@@ -5930,10 +5950,10 @@ int i40e_set_bw_limit(struct i40e_vsi *vsi, u16 seid, u64 max_tx_rate)
 			max_tx_rate, seid);
 		return -EINVAL;
 	}
-	if (max_tx_rate && max_tx_rate < 50) {
+	if (max_tx_rate && max_tx_rate < I40E_BW_CREDIT_DIVISOR) {
 		dev_warn(&pf->pdev->dev,
 			 "Setting max tx rate to minimum usable value of 50Mbps.\n");
-		max_tx_rate = 50;
+		max_tx_rate = I40E_BW_CREDIT_DIVISOR;
 	}
 
 	/* Tx rate credits are in values of 50Mbps, 0 is disabled */
@@ -8224,9 +8244,9 @@ config_tc:
 
 	if (i40e_is_tc_mqprio_enabled(pf)) {
 		if (vsi->mqprio_qopt.max_rate[0]) {
-			u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0];
+			u64 max_tx_rate = i40e_bw_bytes_to_mbits(vsi,
+						  vsi->mqprio_qopt.max_rate[0]);
 
-			do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR);
 			ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate);
 			if (!ret) {
 				u64 credits = max_tx_rate;
@@ -10971,10 +10991,10 @@ static void i40e_rebuild(struct i40e_pf *pf, bool reinit, bool lock_acquired)
 	}
 
 	if (vsi->mqprio_qopt.max_rate[0]) {
-		u64 max_tx_rate = vsi->mqprio_qopt.max_rate[0];
+		u64 max_tx_rate = i40e_bw_bytes_to_mbits(vsi,
+						  vsi->mqprio_qopt.max_rate[0]);
 		u64 credits = 0;
 
-		do_div(max_tx_rate, I40E_BW_MBPS_DIVISOR);
 		ret = i40e_set_bw_limit(vsi, vsi->seid, max_tx_rate);
 		if (ret)
 			goto end_unlock;
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
index 4f184c50f6e8..7e9f6a69eb10 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c
@@ -2039,6 +2039,25 @@ static void i40e_del_qch(struct i40e_vf *vf)
 }
 
 /**
+ * i40e_vc_get_max_frame_size
+ * @vf: pointer to the VF
+ *
+ * Max frame size is determined based on the current port's max frame size and
+ * whether a port VLAN is configured on this VF. The VF is not aware whether
+ * it's in a port VLAN so the PF needs to account for this in max frame size
+ * checks and sending the max frame size to the VF.
+ **/
+static u16 i40e_vc_get_max_frame_size(struct i40e_vf *vf)
+{
+	u16 max_frame_size = vf->pf->hw.phy.link_info.max_frame_size;
+
+	if (vf->port_vlan_id)
+		max_frame_size -= VLAN_HLEN;
+
+	return max_frame_size;
+}
+
+/**
  * i40e_vc_get_vf_resources_msg
  * @vf: pointer to the VF info
  * @msg: pointer to the msg buffer
@@ -2139,6 +2158,7 @@ static int i40e_vc_get_vf_resources_msg(struct i40e_vf *vf, u8 *msg)
 	vfres->max_vectors = pf->hw.func_caps.num_msix_vectors_vf;
 	vfres->rss_key_size = I40E_HKEY_ARRAY_SIZE;
 	vfres->rss_lut_size = I40E_VF_HLUT_ARRAY_SIZE;
+	vfres->max_mtu = i40e_vc_get_max_frame_size(vf);
 
 	if (vf->lan_vsi_idx) {
 		vfres->vsi_res[0].vsi_id = vf->lan_vsi_id;
diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c
index 10aa99dfdcdb..0c89f16bf1e2 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_main.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_main.c
@@ -1077,7 +1077,6 @@ static int iavf_set_mac(struct net_device *netdev, void *p)
 {
 	struct iavf_adapter *adapter = netdev_priv(netdev);
 	struct sockaddr *addr = p;
-	bool handle_mac = iavf_is_mac_set_handled(netdev, addr->sa_data);
 	int ret;
 
 	if (!is_valid_ether_addr(addr->sa_data))
@@ -1094,10 +1093,9 @@ static int iavf_set_mac(struct net_device *netdev, void *p)
 		return 0;
 	}
 
-	if (handle_mac)
-		goto done;
-
-	ret = wait_event_interruptible_timeout(adapter->vc_waitqueue, false, msecs_to_jiffies(2500));
+	ret = wait_event_interruptible_timeout(adapter->vc_waitqueue,
+					       iavf_is_mac_set_handled(netdev, addr->sa_data),
+					       msecs_to_jiffies(2500));
 
 	/* If ret < 0 then it means wait was interrupted.
 	 * If ret == 0 then it means we got a timeout.
@@ -1111,7 +1109,6 @@ static int iavf_set_mac(struct net_device *netdev, void *p)
 	if (!ret)
 		return -EAGAIN;
 
-done:
 	if (!ether_addr_equal(netdev->dev_addr, addr->sa_data))
 		return -EACCES;
 
diff --git a/drivers/net/ethernet/intel/iavf/iavf_txrx.c b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
index 06d18797d25a..18b6a702a1d6 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_txrx.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_txrx.c
@@ -114,8 +114,11 @@ u32 iavf_get_tx_pending(struct iavf_ring *ring, bool in_sw)
 {
 	u32 head, tail;
 
+	/* underlying hardware might not allow access and/or always return
+	 * 0 for the head/tail registers so just use the cached values
+	 */
 	head = ring->next_to_clean;
-	tail = readl(ring->tail);
+	tail = ring->next_to_use;
 
 	if (head != tail)
 		return (head < tail) ?
@@ -1390,7 +1393,7 @@ static struct sk_buff *iavf_build_skb(struct iavf_ring *rx_ring,
 #endif
 	struct sk_buff *skb;
 
-	if (!rx_buffer)
+	if (!rx_buffer || !size)
 		return NULL;
 	/* prefetch first cache line of first page */
 	va = page_address(rx_buffer->page) + rx_buffer->page_offset;
@@ -1548,7 +1551,7 @@ static int iavf_clean_rx_irq(struct iavf_ring *rx_ring, int budget)
 		/* exit if we failed to retrieve a buffer */
 		if (!skb) {
 			rx_ring->rx_stats.alloc_buff_failed++;
-			if (rx_buffer)
+			if (rx_buffer && size)
 				rx_buffer->pagecnt_bias++;
 			break;
 		}
diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
index 15ee85dc33bd..5a9e6563923e 100644
--- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
+++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c
@@ -269,11 +269,14 @@ int iavf_get_vf_vlan_v2_caps(struct iavf_adapter *adapter)
 void iavf_configure_queues(struct iavf_adapter *adapter)
 {
 	struct virtchnl_vsi_queue_config_info *vqci;
-	struct virtchnl_queue_pair_info *vqpi;
+	int i, max_frame = adapter->vf_res->max_mtu;
 	int pairs = adapter->num_active_queues;
-	int i, max_frame = IAVF_MAX_RXBUFFER;
+	struct virtchnl_queue_pair_info *vqpi;
 	size_t len;
 
+	if (max_frame > IAVF_MAX_RXBUFFER || !max_frame)
+		max_frame = IAVF_MAX_RXBUFFER;
+
 	if (adapter->current_op != VIRTCHNL_OP_UNKNOWN) {
 		/* bail because we already have a command pending */
 		dev_err(&adapter->pdev->dev, "Cannot configure queues, command %d pending\n",
diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
index 0c4ec9264071..58d483e2f539 100644
--- a/drivers/net/ethernet/intel/ice/ice_lib.c
+++ b/drivers/net/ethernet/intel/ice/ice_lib.c
@@ -914,7 +914,7 @@ static void ice_set_dflt_vsi_ctx(struct ice_hw *hw, struct ice_vsi_ctx *ctxt)
  */
 static int ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 {
-	u16 offset = 0, qmap = 0, tx_count = 0, pow = 0;
+	u16 offset = 0, qmap = 0, tx_count = 0, rx_count = 0, pow = 0;
 	u16 num_txq_per_tc, num_rxq_per_tc;
 	u16 qcount_tx = vsi->alloc_txq;
 	u16 qcount_rx = vsi->alloc_rxq;
@@ -981,23 +981,25 @@ static int ice_vsi_setup_q_map(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt)
 	 * at least 1)
 	 */
 	if (offset)
-		vsi->num_rxq = offset;
+		rx_count = offset;
 	else
-		vsi->num_rxq = num_rxq_per_tc;
+		rx_count = num_rxq_per_tc;
 
-	if (vsi->num_rxq > vsi->alloc_rxq) {
+	if (rx_count > vsi->alloc_rxq) {
 		dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Rx queues (%u), than were allocated (%u)!\n",
-			vsi->num_rxq, vsi->alloc_rxq);
+			rx_count, vsi->alloc_rxq);
 		return -EINVAL;
 	}
 
-	vsi->num_txq = tx_count;
-	if (vsi->num_txq > vsi->alloc_txq) {
+	if (tx_count > vsi->alloc_txq) {
 		dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Tx queues (%u), than were allocated (%u)!\n",
-			vsi->num_txq, vsi->alloc_txq);
+			tx_count, vsi->alloc_txq);
 		return -EINVAL;
 	}
 
+	vsi->num_txq = tx_count;
+	vsi->num_rxq = rx_count;
+
 	if (vsi->type == ICE_VSI_VF && vsi->num_txq != vsi->num_rxq) {
 		dev_dbg(ice_pf_to_dev(vsi->back), "VF VSI should have same number of Tx and Rx queues. Hence making them equal\n");
 		/* since there is a chance that num_rxq could have been changed
@@ -3490,6 +3492,7 @@ ice_vsi_setup_q_map_mqprio(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt,
 	u16 pow, offset = 0, qcount_tx = 0, qcount_rx = 0, qmap;
 	u16 tc0_offset = vsi->mqprio_qopt.qopt.offset[0];
 	int tc0_qcount = vsi->mqprio_qopt.qopt.count[0];
+	u16 new_txq, new_rxq;
 	u8 netdev_tc = 0;
 	int i;
 
@@ -3530,21 +3533,24 @@ ice_vsi_setup_q_map_mqprio(struct ice_vsi *vsi, struct ice_vsi_ctx *ctxt,
 		}
 	}
 
-	/* Set actual Tx/Rx queue pairs */
-	vsi->num_txq = offset + qcount_tx;
-	if (vsi->num_txq > vsi->alloc_txq) {
+	new_txq = offset + qcount_tx;
+	if (new_txq > vsi->alloc_txq) {
 		dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Tx queues (%u), than were allocated (%u)!\n",
-			vsi->num_txq, vsi->alloc_txq);
+			new_txq, vsi->alloc_txq);
 		return -EINVAL;
 	}
 
-	vsi->num_rxq = offset + qcount_rx;
-	if (vsi->num_rxq > vsi->alloc_rxq) {
+	new_rxq = offset + qcount_rx;
+	if (new_rxq > vsi->alloc_rxq) {
 		dev_err(ice_pf_to_dev(vsi->back), "Trying to use more Rx queues (%u), than were allocated (%u)!\n",
-			vsi->num_rxq, vsi->alloc_rxq);
+			new_rxq, vsi->alloc_rxq);
 		return -EINVAL;
 	}
 
+	/* Set actual Tx/Rx queue pairs */
+	vsi->num_txq = new_txq;
+	vsi->num_rxq = new_rxq;
+
 	/* Setup queue TC[0].qmap for given VSI context */
 	ctxt->info.tc_mapping[0] = cpu_to_le16(qmap);
 	ctxt->info.q_mapping[0] = cpu_to_le16(vsi->rxq_map[0]);
@@ -3576,6 +3582,7 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc)
 {
 	u16 max_txqs[ICE_MAX_TRAFFIC_CLASS] = { 0 };
 	struct ice_pf *pf = vsi->back;
+	struct ice_tc_cfg old_tc_cfg;
 	struct ice_vsi_ctx *ctx;
 	struct device *dev;
 	int i, ret = 0;
@@ -3600,6 +3607,7 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc)
 			max_txqs[i] = vsi->num_txq;
 	}
 
+	memcpy(&old_tc_cfg, &vsi->tc_cfg, sizeof(old_tc_cfg));
 	vsi->tc_cfg.ena_tc = ena_tc;
 	vsi->tc_cfg.numtc = num_tc;
 
@@ -3616,8 +3624,10 @@ int ice_vsi_cfg_tc(struct ice_vsi *vsi, u8 ena_tc)
 	else
 		ret = ice_vsi_setup_q_map(vsi, ctx);
 
-	if (ret)
+	if (ret) {
+		memcpy(&vsi->tc_cfg, &old_tc_cfg, sizeof(vsi->tc_cfg));
 		goto out;
+	}
 
 	/* must to indicate which section of VSI context are being modified */
 	ctx->info.valid_sections = cpu_to_le16(ICE_AQ_VSI_PROP_RXQ_MAP_VALID);
diff --git a/drivers/net/ethernet/intel/ice/ice_main.c b/drivers/net/ethernet/intel/ice/ice_main.c
index 8c30eea61b6d..e109cb93886b 100644
--- a/drivers/net/ethernet/intel/ice/ice_main.c
+++ b/drivers/net/ethernet/intel/ice/ice_main.c
@@ -2399,8 +2399,6 @@ int ice_schedule_reset(struct ice_pf *pf, enum ice_reset_req reset)
 		return -EBUSY;
 	}
 
-	ice_unplug_aux_dev(pf);
-
 	switch (reset) {
 	case ICE_RESET_PFR:
 		set_bit(ICE_PFR_REQ, pf->state);
@@ -6651,7 +6649,7 @@ static void ice_napi_disable_all(struct ice_vsi *vsi)
  */
 int ice_down(struct ice_vsi *vsi)
 {
-	int i, tx_err, rx_err, link_err = 0, vlan_err = 0;
+	int i, tx_err, rx_err, vlan_err = 0;
 
 	WARN_ON(!test_bit(ICE_VSI_DOWN, vsi->state));
 
@@ -6685,20 +6683,13 @@ int ice_down(struct ice_vsi *vsi)
 
 	ice_napi_disable_all(vsi);
 
-	if (test_bit(ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA, vsi->back->flags)) {
-		link_err = ice_force_phys_link_state(vsi, false);
-		if (link_err)
-			netdev_err(vsi->netdev, "Failed to set physical link down, VSI %d error %d\n",
-				   vsi->vsi_num, link_err);
-	}
-
 	ice_for_each_txq(vsi, i)
 		ice_clean_tx_ring(vsi->tx_rings[i]);
 
 	ice_for_each_rxq(vsi, i)
 		ice_clean_rx_ring(vsi->rx_rings[i]);
 
-	if (tx_err || rx_err || link_err || vlan_err) {
+	if (tx_err || rx_err || vlan_err) {
 		netdev_err(vsi->netdev, "Failed to close VSI 0x%04X on switch 0x%04X\n",
 			   vsi->vsi_num, vsi->vsw->sw_id);
 		return -EIO;
@@ -6860,6 +6851,8 @@ int ice_vsi_open(struct ice_vsi *vsi)
 	if (err)
 		goto err_setup_rx;
 
+	ice_vsi_cfg_netdev_tc(vsi, vsi->tc_cfg.ena_tc);
+
 	if (vsi->type == ICE_VSI_PF) {
 		/* Notify the stack of the actual queue counts. */
 		err = netif_set_real_num_tx_queues(vsi->netdev, vsi->num_txq);
@@ -8892,6 +8885,16 @@ int ice_stop(struct net_device *netdev)
 		return -EBUSY;
 	}
 
+	if (test_bit(ICE_FLAG_LINK_DOWN_ON_CLOSE_ENA, vsi->back->flags)) {
+		int link_err = ice_force_phys_link_state(vsi, false);
+
+		if (link_err) {
+			netdev_err(vsi->netdev, "Failed to set physical link down, VSI %d error %d\n",
+				   vsi->vsi_num, link_err);
+			return -EIO;
+		}
+	}
+
 	ice_vsi_close(vsi);
 
 	return 0;
diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
index 836dce840712..97453d1dfafe 100644
--- a/drivers/net/ethernet/intel/ice/ice_txrx.c
+++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
@@ -610,7 +610,7 @@ ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 	if (test_bit(ICE_VSI_DOWN, vsi->state))
 		return -ENETDOWN;
 
-	if (!ice_is_xdp_ena_vsi(vsi) || queue_index >= vsi->num_xdp_txq)
+	if (!ice_is_xdp_ena_vsi(vsi))
 		return -ENXIO;
 
 	if (unlikely(flags & ~XDP_XMIT_FLAGS_MASK))
@@ -621,6 +621,9 @@ ice_xdp_xmit(struct net_device *dev, int n, struct xdp_frame **frames,
 		xdp_ring = vsi->xdp_rings[queue_index];
 		spin_lock(&xdp_ring->tx_lock);
 	} else {
+		/* Generally, should not happen */
+		if (unlikely(queue_index >= vsi->num_xdp_txq))
+			return -ENXIO;
 		xdp_ring = vsi->xdp_rings[queue_index];
 	}