[PATCH net-next 00/12] code optimizations & bugfixes for HNS3 driver

[PATCH net-next 00/12] code optimizations & bugfixes for HNS3 driver

[Huazhong Tan]

This patchset includes bugfixes and code optimizations for the HNS3
ethernet controller driver

Huazhong Tan (1):
  net: hns3: fix a wrong checking in the hclge_tx_buffer_calc()

Jian Shen (1):
  net: hns3: don't allow user to change vlan filter state

Peng Li (2):
  net: hns3: fix a code style issue for hns3_update_new_int_gl()
  net: hns3: fix an issue for hns3_update_new_int_gl

Weihang Li (2):
  net: hns3: add hclge_cmd_check_retval() to parse comman's return value
  net: hns3: move some set_bit statement into hclge_prepare_mac_addr

Yonglong Liu (1):
  net: hns3: Modify parameter type from int to bool in set_gro_en

Yunsheng Lin (1):
  net: hns3: code optimization for hclge_rx_buffer_calc

liuzhongzhu (4):
  net: hns3: fix the problem that the supported port is empty
  net: hns3: optimize the maximum TC macro
  net: hns3: modify the upper limit judgment condition
  net: hns3: MAC table entry count function increases operation 0 value
    protection measures

 .../net/ethernet/hisilicon/hns3/hclge_mbx.h   |   4 +-
 drivers/net/ethernet/hisilicon/hns3/hnae3.h   |   2 +-
 .../net/ethernet/hisilicon/hns3/hns3_enet.c   |  33 ++--
 .../hisilicon/hns3/hns3pf/hclge_cmd.c         |  62 +++---
 .../hisilicon/hns3/hns3pf/hclge_cmd.h         |   7 +-
 .../hisilicon/hns3/hns3pf/hclge_debugfs.c     |   4 +-
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 184 ++++++++----------
 .../hisilicon/hns3/hns3pf/hclge_mbx.c         |  31 ++-
 .../hisilicon/hns3/hns3vf/hclgevf_main.c      |  34 +++-
 .../hisilicon/hns3/hns3vf/hclgevf_main.h      |   2 +
 .../hisilicon/hns3/hns3vf/hclgevf_mbx.c       |  12 ++
 11 files changed, 216 insertions(+), 159 deletions(-)

-- 
2.20.1

[PATCH net-next 01/12] net: hns3: fix a code style issue for hns3_update_new_int_gl()

[Huazhong Tan]

From: Peng Li <lipeng321@huawei.com>

Use the same code style for rx_group and tx_group in the
hns3_update_new_int_gl().

Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index c546b874d659..1d3f167ee1e4 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -2880,7 +2880,7 @@ static void hns3_update_new_int_gl(struct hns3_enet_tqp_vector *tqp_vector)
 	}
 
 	if (tx_group->coal.gl_adapt_enable) {
-		tx_update = hns3_get_new_int_gl(&tqp_vector->tx_group);
+		tx_update = hns3_get_new_int_gl(tx_group);
 		if (tx_update)
 			hns3_set_vector_coalesce_tx_gl(tqp_vector,
 						       tx_group->coal.int_gl);
-- 
2.20.1

[PATCH net-next 02/12] net: hns3: fix an issue for hns3_update_new_int_gl

[Huazhong Tan]

From: Peng Li <lipeng321@huawei.com>

HNS3 supports setting rx-usecs|tx-usecs as 0, but it will not
update dynamically when adaptive-tx or adaptive-rx is enable.
This patch removes the Redundant check.

Fixes: a95e1f8666e9 ("net: hns3: change the time interval of int_gl calculating")
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 1d3f167ee1e4..bc571b3d8ddc 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -2777,7 +2777,7 @@ static bool hns3_get_new_int_gl(struct hns3_enet_ring_group *ring_group)
 	u32 time_passed_ms;
 	u16 new_int_gl;
 
-	if (!ring_group->coal.int_gl || !tqp_vector->last_jiffies)
+	if (!tqp_vector->last_jiffies)
 		return false;
 
 	if (ring_group->total_packets == 0) {
-- 
2.20.1

[PATCH net-next 03/12] net: hns3: Modify parameter type from int to bool in set_gro_en

[Huazhong Tan]

From: Yonglong Liu <liuyonglong@huawei.com>

The second parameter to the hook function set_gro_en is always passed in
true/false, so modify it's type from int to bool.

Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hnae3.h   |  2 +-
 .../net/ethernet/hisilicon/hns3/hns3_enet.c   | 26 +++++++------------
 .../hisilicon/hns3/hns3pf/hclge_main.c        |  2 +-
 .../hisilicon/hns3/hns3vf/hclgevf_main.c      |  2 +-
 4 files changed, 12 insertions(+), 20 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.h b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
index e05b4926feb2..10f1cc2ba157 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.h
@@ -461,7 +461,7 @@ struct hnae3_ae_ops {
 	bool (*get_hw_reset_stat)(struct hnae3_handle *handle);
 	bool (*ae_dev_resetting)(struct hnae3_handle *handle);
 	unsigned long (*ae_dev_reset_cnt)(struct hnae3_handle *handle);
-	int (*set_gro_en)(struct hnae3_handle *handle, int enable);
+	int (*set_gro_en)(struct hnae3_handle *handle, bool enable);
 	u16 (*get_global_queue_id)(struct hnae3_handle *handle, u16 queue_id);
 	void (*set_timer_task)(struct hnae3_handle *handle, bool enable);
 	int (*mac_connect_phy)(struct hnae3_handle *handle);
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index bc571b3d8ddc..db06d36d408f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1349,6 +1349,7 @@ static int hns3_nic_set_features(struct net_device *netdev,
 	netdev_features_t changed = netdev->features ^ features;
 	struct hns3_nic_priv *priv = netdev_priv(netdev);
 	struct hnae3_handle *h = priv->ae_handle;
+	bool enable;
 	int ret;
 
 	if (changed & (NETIF_F_TSO | NETIF_F_TSO6)) {
@@ -1359,38 +1360,29 @@ static int hns3_nic_set_features(struct net_device *netdev,
 	}
 
 	if (changed & (NETIF_F_GRO_HW) && h->ae_algo->ops->set_gro_en) {
-		if (features & NETIF_F_GRO_HW)
-			ret = h->ae_algo->ops->set_gro_en(h, true);
-		else
-			ret = h->ae_algo->ops->set_gro_en(h, false);
+		enable = !!(features & NETIF_F_GRO_HW);
+		ret = h->ae_algo->ops->set_gro_en(h, enable);
 		if (ret)
 			return ret;
 	}
 
 	if ((changed & NETIF_F_HW_VLAN_CTAG_FILTER) &&
 	    h->ae_algo->ops->enable_vlan_filter) {
-		if (features & NETIF_F_HW_VLAN_CTAG_FILTER)
-			h->ae_algo->ops->enable_vlan_filter(h, true);
-		else
-			h->ae_algo->ops->enable_vlan_filter(h, false);
+		enable = !!(features & NETIF_F_HW_VLAN_CTAG_FILTER);
+		h->ae_algo->ops->enable_vlan_filter(h, enable);
 	}
 
 	if ((changed & NETIF_F_HW_VLAN_CTAG_RX) &&
 	    h->ae_algo->ops->enable_hw_strip_rxvtag) {
-		if (features & NETIF_F_HW_VLAN_CTAG_RX)
-			ret = h->ae_algo->ops->enable_hw_strip_rxvtag(h, true);
-		else
-			ret = h->ae_algo->ops->enable_hw_strip_rxvtag(h, false);
-
+		enable = !!(features & NETIF_F_HW_VLAN_CTAG_RX);
+		ret = h->ae_algo->ops->enable_hw_strip_rxvtag(h, enable);
 		if (ret)
 			return ret;
 	}
 
 	if ((changed & NETIF_F_NTUPLE) && h->ae_algo->ops->enable_fd) {
-		if (features & NETIF_F_NTUPLE)
-			h->ae_algo->ops->enable_fd(h, true);
-		else
-			h->ae_algo->ops->enable_fd(h, false);
+		enable = !!(features & NETIF_F_NTUPLE);
+		h->ae_algo->ops->enable_fd(h, enable);
 	}
 
 	netdev->features = features;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 52fc57284be8..9630bbce5932 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -7989,7 +7989,7 @@ static void hclge_get_link_mode(struct hnae3_handle *handle,
 	}
 }
 
-static int hclge_gro_en(struct hnae3_handle *handle, int enable)
+static int hclge_gro_en(struct hnae3_handle *handle, bool enable)
 {
 	struct hclge_vport *vport = hclge_get_vport(handle);
 	struct hclge_dev *hdev = vport->back;
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index b9cdbd5dd6cb..bc2f29e67280 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -2550,7 +2550,7 @@ void hclgevf_update_speed_duplex(struct hclgevf_dev *hdev, u32 speed,
 	hdev->hw.mac.duplex = duplex;
 }
 
-static int hclgevf_gro_en(struct hnae3_handle *handle, int enable)
+static int hclgevf_gro_en(struct hnae3_handle *handle, bool enable)
 {
 	struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
 
-- 
2.20.1

[PATCH net-next 04/12] net: hns3: code optimization for hclge_rx_buffer_calc

[Huazhong Tan]

From: Yunsheng Lin <linyunsheng@huawei.com>

There are four steps to calcuate the rx private buffer, each step
can be done in a function to avoid code duplication and aid code
readability.

This patch adds three separate functions do the job. Also, the
function name more or less make the comment redundant, so remove
some obvious comment.

Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 140 ++++++++----------
 1 file changed, 63 insertions(+), 77 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 9630bbce5932..be7d82502f5f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1520,66 +1520,15 @@ static int hclge_tx_buffer_calc(struct hclge_dev *hdev,
 	return 0;
 }
 
-/* hclge_rx_buffer_calc: calculate the rx private buffer size for all TCs
- * @hdev: pointer to struct hclge_dev
- * @buf_alloc: pointer to buffer calculation data
- * @return: 0: calculate sucessful, negative: fail
- */
-static int hclge_rx_buffer_calc(struct hclge_dev *hdev,
-				struct hclge_pkt_buf_alloc *buf_alloc)
+static bool hclge_rx_buf_calc_all(struct hclge_dev *hdev, bool max,
+				  struct hclge_pkt_buf_alloc *buf_alloc)
 {
-	u32 rx_all = hdev->pkt_buf_size, aligned_mps;
-	int no_pfc_priv_num, pfc_priv_num;
-	struct hclge_priv_buf *priv;
+	u32 rx_all = hdev->pkt_buf_size - hclge_get_tx_buff_alloced(buf_alloc);
+	u32 aligned_mps = round_up(hdev->mps, HCLGE_BUF_SIZE_UNIT);
 	int i;
 
-	aligned_mps = round_up(hdev->mps, HCLGE_BUF_SIZE_UNIT);
-	rx_all -= hclge_get_tx_buff_alloced(buf_alloc);
-
-	/* When DCB is not supported, rx private
-	 * buffer is not allocated.
-	 */
-	if (!hnae3_dev_dcb_supported(hdev)) {
-		if (!hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
-			return -ENOMEM;
-
-		return 0;
-	}
-
-	/* step 1, try to alloc private buffer for all enabled tc */
-	for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
-		priv = &buf_alloc->priv_buf[i];
-		if (hdev->hw_tc_map & BIT(i)) {
-			priv->enable = 1;
-			if (hdev->tm_info.hw_pfc_map & BIT(i)) {
-				priv->wl.low = aligned_mps;
-				priv->wl.high =
-					roundup(priv->wl.low + aligned_mps,
-						HCLGE_BUF_SIZE_UNIT);
-				priv->buf_size = priv->wl.high +
-					hdev->dv_buf_size;
-			} else {
-				priv->wl.low = 0;
-				priv->wl.high = 2 * aligned_mps;
-				priv->buf_size = priv->wl.high +
-						hdev->dv_buf_size;
-			}
-		} else {
-			priv->enable = 0;
-			priv->wl.low = 0;
-			priv->wl.high = 0;
-			priv->buf_size = 0;
-		}
-	}
-
-	if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
-		return 0;
-
-	/* step 2, try to decrease the buffer size of
-	 * no pfc TC's private buffer
-	 */
 	for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
-		priv = &buf_alloc->priv_buf[i];
+		struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
 
 		priv->enable = 0;
 		priv->wl.low = 0;
@@ -1592,28 +1541,30 @@ static int hclge_rx_buffer_calc(struct hclge_dev *hdev,
 		priv->enable = 1;
 
 		if (hdev->tm_info.hw_pfc_map & BIT(i)) {
-			priv->wl.low = 256;
-			priv->wl.high = priv->wl.low + aligned_mps;
-			priv->buf_size = priv->wl.high + hdev->dv_buf_size;
+			priv->wl.low = max ? aligned_mps : 256;
+			priv->wl.high = roundup(priv->wl.low + aligned_mps,
+						HCLGE_BUF_SIZE_UNIT);
 		} else {
 			priv->wl.low = 0;
-			priv->wl.high = aligned_mps;
-			priv->buf_size = priv->wl.high + hdev->dv_buf_size;
+			priv->wl.high = max ? (aligned_mps * 2) : aligned_mps;
 		}
+
+		priv->buf_size = priv->wl.high + hdev->dv_buf_size;
 	}
 
-	if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
-		return 0;
+	return hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all);
+}
 
-	/* step 3, try to reduce the number of pfc disabled TCs,
-	 * which have private buffer
-	 */
-	/* get the total no pfc enable TC number, which have private buffer */
-	no_pfc_priv_num = hclge_get_no_pfc_priv_num(hdev, buf_alloc);
+static bool hclge_drop_nopfc_buf_till_fit(struct hclge_dev *hdev,
+					  struct hclge_pkt_buf_alloc *buf_alloc)
+{
+	u32 rx_all = hdev->pkt_buf_size - hclge_get_tx_buff_alloced(buf_alloc);
+	int no_pfc_priv_num = hclge_get_no_pfc_priv_num(hdev, buf_alloc);
+	int i;
 
 	/* let the last to be cleared first */
 	for (i = HCLGE_MAX_TC_NUM - 1; i >= 0; i--) {
-		priv = &buf_alloc->priv_buf[i];
+		struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
 
 		if (hdev->hw_tc_map & BIT(i) &&
 		    !(hdev->tm_info.hw_pfc_map & BIT(i))) {
@@ -1630,17 +1581,19 @@ static int hclge_rx_buffer_calc(struct hclge_dev *hdev,
 			break;
 	}
 
-	if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
-		return 0;
+	return hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all);
+}
 
-	/* step 4, try to reduce the number of pfc enabled TCs
-	 * which have private buffer.
-	 */
-	pfc_priv_num = hclge_get_pfc_priv_num(hdev, buf_alloc);
+static bool hclge_drop_pfc_buf_till_fit(struct hclge_dev *hdev,
+					struct hclge_pkt_buf_alloc *buf_alloc)
+{
+	u32 rx_all = hdev->pkt_buf_size - hclge_get_tx_buff_alloced(buf_alloc);
+	int pfc_priv_num = hclge_get_pfc_priv_num(hdev, buf_alloc);
+	int i;
 
 	/* let the last to be cleared first */
 	for (i = HCLGE_MAX_TC_NUM - 1; i >= 0; i--) {
-		priv = &buf_alloc->priv_buf[i];
+		struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
 
 		if (hdev->hw_tc_map & BIT(i) &&
 		    hdev->tm_info.hw_pfc_map & BIT(i)) {
@@ -1656,7 +1609,40 @@ static int hclge_rx_buffer_calc(struct hclge_dev *hdev,
 		    pfc_priv_num == 0)
 			break;
 	}
-	if (hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
+
+	return hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all);
+}
+
+/* hclge_rx_buffer_calc: calculate the rx private buffer size for all TCs
+ * @hdev: pointer to struct hclge_dev
+ * @buf_alloc: pointer to buffer calculation data
+ * @return: 0: calculate sucessful, negative: fail
+ */
+static int hclge_rx_buffer_calc(struct hclge_dev *hdev,
+				struct hclge_pkt_buf_alloc *buf_alloc)
+{
+	/* When DCB is not supported, rx private buffer is not allocated. */
+	if (!hnae3_dev_dcb_supported(hdev)) {
+		u32 rx_all = hdev->pkt_buf_size;
+
+		rx_all -= hclge_get_tx_buff_alloced(buf_alloc);
+		if (!hclge_is_rx_buf_ok(hdev, buf_alloc, rx_all))
+			return -ENOMEM;
+
+		return 0;
+	}
+
+	if (hclge_rx_buf_calc_all(hdev, true, buf_alloc))
+		return 0;
+
+	/* try to decrease the buffer size */
+	if (hclge_rx_buf_calc_all(hdev, false, buf_alloc))
+		return 0;
+
+	if (hclge_drop_nopfc_buf_till_fit(hdev, buf_alloc))
+		return 0;
+
+	if (hclge_drop_pfc_buf_till_fit(hdev, buf_alloc))
 		return 0;
 
 	return -ENOMEM;
-- 
2.20.1

[PATCH net-next 05/12] net: hns3: add hclge_cmd_check_retval() to parse comman's return value

[Huazhong Tan]

From: Weihang Li <liweihang@hisilicon.com>

For simplifying the code, this patch adds hclge_cmd_check_retval() to
check the return value of the command.

Also, according the IMP's description, when there are several descriptors
in a command, then the IMP will save the return value on the last
description, so hclge_cmd_check_retval() just check the last one for this
case.

Signed-off-by: Weihang Li <liweihang@hisilicon.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 .../hisilicon/hns3/hns3pf/hclge_cmd.c         | 60 ++++++++++---------
 1 file changed, 33 insertions(+), 27 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
index 81dbe1b6abb0..57c50b947857 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
@@ -186,6 +186,38 @@ static bool hclge_is_special_opcode(u16 opcode)
 	return false;
 }
 
+static int hclge_cmd_check_retval(struct hclge_hw *hw, struct hclge_desc *desc,
+				  int num, int ntc)
+{
+	u16 opcode, desc_ret;
+	int handle;
+	int retval;
+
+	opcode = le16_to_cpu(desc[0].opcode);
+	for (handle = 0; handle < num; handle++) {
+		desc[handle] = hw->cmq.csq.desc[ntc];
+		ntc++;
+		if (ntc == hw->cmq.csq.desc_num)
+			ntc = 0;
+	}
+	if (likely(!hclge_is_special_opcode(opcode)))
+		desc_ret = le16_to_cpu(desc[num - 1].retval);
+	else
+		desc_ret = le16_to_cpu(desc[0].retval);
+
+	if (desc_ret == HCLGE_CMD_EXEC_SUCCESS)
+		retval = 0;
+	else if (desc_ret == HCLGE_CMD_NO_AUTH)
+		retval = -EPERM;
+	else if (desc_ret == HCLGE_CMD_NOT_SUPPORTED)
+		retval = -EOPNOTSUPP;
+	else
+		retval = -EIO;
+	hw->cmq.last_status = desc_ret;
+
+	return retval;
+}
+
 /**
  * hclge_cmd_send - send command to command queue
  * @hw: pointer to the hw struct
@@ -203,7 +235,6 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num)
 	u32 timeout = 0;
 	int handle = 0;
 	int retval = 0;
-	u16 opcode, desc_ret;
 	int ntc;
 
 	spin_lock_bh(&hw->cmq.csq.lock);
@@ -219,7 +250,6 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num)
 	 * which will be use for hardware to write back
 	 */
 	ntc = hw->cmq.csq.next_to_use;
-	opcode = le16_to_cpu(desc[0].opcode);
 	while (handle < num) {
 		desc_to_use = &hw->cmq.csq.desc[hw->cmq.csq.next_to_use];
 		*desc_to_use = desc[handle];
@@ -250,31 +280,7 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num)
 	if (!complete) {
 		retval = -EAGAIN;
 	} else {
-		handle = 0;
-		while (handle < num) {
-			/* Get the result of hardware write back */
-			desc_to_use = &hw->cmq.csq.desc[ntc];
-			desc[handle] = *desc_to_use;
-
-			if (likely(!hclge_is_special_opcode(opcode)))
-				desc_ret = le16_to_cpu(desc[handle].retval);
-			else
-				desc_ret = le16_to_cpu(desc[0].retval);
-
-			if (desc_ret == HCLGE_CMD_EXEC_SUCCESS)
-				retval = 0;
-			else if (desc_ret == HCLGE_CMD_NO_AUTH)
-				retval = -EPERM;
-			else if (desc_ret == HCLGE_CMD_NOT_SUPPORTED)
-				retval = -EOPNOTSUPP;
-			else
-				retval = -EIO;
-			hw->cmq.last_status = desc_ret;
-			ntc++;
-			handle++;
-			if (ntc == hw->cmq.csq.desc_num)
-				ntc = 0;
-		}
+		retval = hclge_cmd_check_retval(hw, desc, num, ntc);
 	}
 
 	/* Clean the command send queue */
-- 
2.20.1

[PATCH net-next 06/12] net: hns3: move some set_bit statement into hclge_prepare_mac_addr

[Huazhong Tan]

From: Weihang Li <liweihang@hisilicon.com>

This patch does not change the code logic. There are some same
set_bit statements called by add/rm_uc/mc_addr_common, and move
this statements into hclge_prepare_mac_addr to reduce duplicate
code.

Signed-off-by: Weihang Li <liweihang@hisilicon.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 .../hisilicon/hns3/hns3pf/hclge_main.c        | 24 +++++++++----------
 1 file changed, 11 insertions(+), 13 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index be7d82502f5f..a037dc5e36b7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -5586,13 +5586,19 @@ static bool hclge_is_all_function_id_zero(struct hclge_desc *desc)
 }
 
 static void hclge_prepare_mac_addr(struct hclge_mac_vlan_tbl_entry_cmd *new_req,
-				   const u8 *addr)
+				   const u8 *addr, bool is_mc)
 {
 	const unsigned char *mac_addr = addr;
 	u32 high_val = mac_addr[2] << 16 | (mac_addr[3] << 24) |
 		       (mac_addr[0]) | (mac_addr[1] << 8);
 	u32 low_val  = mac_addr[4] | (mac_addr[5] << 8);
 
+	hnae3_set_bit(new_req->flags, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
+	if (is_mc) {
+		hnae3_set_bit(new_req->entry_type, HCLGE_MAC_VLAN_BIT1_EN_B, 1);
+		hnae3_set_bit(new_req->mc_mac_en, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
+	}
+
 	new_req->mac_addr_hi32 = cpu_to_le32(high_val);
 	new_req->mac_addr_lo16 = cpu_to_le16(low_val & 0xffff);
 }
@@ -5863,14 +5869,13 @@ int hclge_add_uc_addr_common(struct hclge_vport *vport,
 	}
 
 	memset(&req, 0, sizeof(req));
-	hnae3_set_bit(req.flags, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
 
 	hnae3_set_field(egress_port, HCLGE_MAC_EPORT_VFID_M,
 			HCLGE_MAC_EPORT_VFID_S, vport->vport_id);
 
 	req.egress_port = cpu_to_le16(egress_port);
 
-	hclge_prepare_mac_addr(&req, addr);
+	hclge_prepare_mac_addr(&req, addr, false);
 
 	/* Lookup the mac address in the mac_vlan table, and add
 	 * it if the entry is inexistent. Repeated unicast entry
@@ -5928,9 +5933,8 @@ int hclge_rm_uc_addr_common(struct hclge_vport *vport,
 	}
 
 	memset(&req, 0, sizeof(req));
-	hnae3_set_bit(req.flags, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
 	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT0_EN_B, 0);
-	hclge_prepare_mac_addr(&req, addr);
+	hclge_prepare_mac_addr(&req, addr, false);
 	ret = hclge_remove_mac_vlan_tbl(vport, &req);
 	if (!ret)
 		hclge_update_umv_space(vport, true);
@@ -5962,11 +5966,8 @@ int hclge_add_mc_addr_common(struct hclge_vport *vport,
 		return -EINVAL;
 	}
 	memset(&req, 0, sizeof(req));
-	hnae3_set_bit(req.flags, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
 	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT0_EN_B, 0);
-	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT1_EN_B, 1);
-	hnae3_set_bit(req.mc_mac_en, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
-	hclge_prepare_mac_addr(&req, addr);
+	hclge_prepare_mac_addr(&req, addr, true);
 	status = hclge_lookup_mac_vlan_tbl(vport, &req, desc, true);
 	if (!status) {
 		/* This mac addr exist, update VFID for it */
@@ -6012,11 +6013,8 @@ int hclge_rm_mc_addr_common(struct hclge_vport *vport,
 	}
 
 	memset(&req, 0, sizeof(req));
-	hnae3_set_bit(req.flags, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
 	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT0_EN_B, 0);
-	hnae3_set_bit(req.entry_type, HCLGE_MAC_VLAN_BIT1_EN_B, 1);
-	hnae3_set_bit(req.mc_mac_en, HCLGE_MAC_VLAN_BIT0_EN_B, 1);
-	hclge_prepare_mac_addr(&req, addr);
+	hclge_prepare_mac_addr(&req, addr, true);
 	status = hclge_lookup_mac_vlan_tbl(vport, &req, desc, true);
 	if (!status) {
 		/* This mac addr exist, remove this handle's VFID for it */
-- 
2.20.1

[PATCH net-next 07/12] net: hns3: fix a wrong checking in the hclge_tx_buffer_calc()

[Huazhong Tan]

Only the TC is enabled, we need to check whether the buffer is enough,
otherwise it may lead to a wrong -ENOMEM case.

Fixes: 9ffe79a9c2ee ("net: hns3: Support for dynamically assigning tx buffer to TC")
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index a037dc5e36b7..5f6fc23fab5d 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1506,13 +1506,14 @@ static int hclge_tx_buffer_calc(struct hclge_dev *hdev,
 	for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
 		struct hclge_priv_buf *priv = &buf_alloc->priv_buf[i];
 
-		if (total_size < hdev->tx_buf_size)
-			return -ENOMEM;
+		if (hdev->hw_tc_map & BIT(i)) {
+			if (total_size < hdev->tx_buf_size)
+				return -ENOMEM;
 
-		if (hdev->hw_tc_map & BIT(i))
 			priv->tx_buf_size = hdev->tx_buf_size;
-		else
+		} else {
 			priv->tx_buf_size = 0;
+		}
 
 		total_size -= priv->tx_buf_size;
 	}
-- 
2.20.1

[PATCH net-next 08/12] net: hns3: fix the problem that the supported port is empty

[Huazhong Tan]

From: liuzhongzhu <liuzhongzhu@huawei.com>

Run ethtool ethx when displaying device information in VF，
the supported port and link mode items will be empty.

This patch fixes it.

Fixes: e2cb1dec9779 ("net: hns3: Add HNS3 VF HCL(Hardware Compatibility Layer) Support")
Signed-off-by: liuzhongzhu <liuzhongzhu@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 .../net/ethernet/hisilicon/hns3/hclge_mbx.h   |  4 ++-
 .../hisilicon/hns3/hns3pf/hclge_mbx.c         | 31 +++++++++++++++++-
 .../hisilicon/hns3/hns3vf/hclgevf_main.c      | 32 +++++++++++++++++--
 .../hisilicon/hns3/hns3vf/hclgevf_main.h      |  2 ++
 .../hisilicon/hns3/hns3vf/hclgevf_mbx.c       | 12 +++++++
 5 files changed, 77 insertions(+), 4 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h b/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
index 691d12174902..53089cd980d9 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
@@ -40,6 +40,8 @@ enum HCLGE_MBX_OPCODE {
 	HCLGE_MBX_SET_ALIVE,		/* (VF -> PF) set alive state */
 	HCLGE_MBX_SET_MTU,		/* (VF -> PF) set mtu */
 	HCLGE_MBX_GET_QID_IN_PF,	/* (VF -> PF) get queue id in pf */
+	HCLGE_MBX_LINK_STAT_MODE,	/* (PF -> VF) link mode has changed */
+	HCLGE_MBX_GET_LINK_MODE,	/* (VF -> PF) get the link mode of pf */
 };
 
 /* below are per-VF mac-vlan subcodes */
@@ -60,7 +62,7 @@ enum hclge_mbx_vlan_cfg_subcode {
 };
 
 #define HCLGE_MBX_MAX_MSG_SIZE	16
-#define HCLGE_MBX_MAX_RESP_DATA_SIZE	8
+#define HCLGE_MBX_MAX_RESP_DATA_SIZE	16
 #define HCLGE_MBX_RING_MAP_BASIC_MSG_NUM	3
 #define HCLGE_MBX_RING_NODE_VARIABLE_NUM	3
 
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index 6afb0a4b73f7..7e4a104582d6 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -355,16 +355,19 @@ static int hclge_get_link_info(struct hclge_vport *vport,
 {
 	struct hclge_dev *hdev = vport->back;
 	u16 link_status;
-	u8 msg_data[8];
+	u8 msg_data[10];
+	u16 media_type;
 	u8 dest_vfid;
 	u16 duplex;
 
 	/* mac.link can only be 0 or 1 */
 	link_status = (u16)hdev->hw.mac.link;
 	duplex = hdev->hw.mac.duplex;
+	media_type = hdev->hw.mac.media_type;
 	memcpy(&msg_data[0], &link_status, sizeof(u16));
 	memcpy(&msg_data[2], &hdev->hw.mac.speed, sizeof(u32));
 	memcpy(&msg_data[6], &duplex, sizeof(u16));
+	memcpy(&msg_data[8], &media_type, sizeof(u16));
 	dest_vfid = mbx_req->mbx_src_vfid;
 
 	/* send this requested info to VF */
@@ -372,6 +375,29 @@ static int hclge_get_link_info(struct hclge_vport *vport,
 				  HCLGE_MBX_LINK_STAT_CHANGE, dest_vfid);
 }
 
+static void hclge_get_link_mode(struct hclge_vport *vport,
+				struct hclge_mbx_vf_to_pf_cmd *mbx_req)
+{
+#define HCLGE_SUPPORTED   1
+	struct hclge_dev *hdev = vport->back;
+	unsigned long advertising;
+	unsigned long supported;
+	unsigned long send_data;
+	u8 msg_data[10];
+	u8 dest_vfid;
+
+	advertising = hdev->hw.mac.advertising[0];
+	supported = hdev->hw.mac.supported[0];
+	dest_vfid = mbx_req->mbx_src_vfid;
+	msg_data[0] = mbx_req->msg[2];
+
+	send_data = msg_data[0] == HCLGE_SUPPORTED ? supported : advertising;
+
+	memcpy(&msg_data[2], &send_data, sizeof(unsigned long));
+	hclge_send_mbx_msg(vport, msg_data, sizeof(msg_data),
+			   HCLGE_MBX_LINK_STAT_MODE, dest_vfid);
+}
+
 static void hclge_mbx_reset_vf_queue(struct hclge_vport *vport,
 				     struct hclge_mbx_vf_to_pf_cmd *mbx_req)
 {
@@ -556,6 +582,9 @@ void hclge_mbx_handler(struct hclge_dev *hdev)
 					"PF failed(%d) to get qid for VF\n",
 					ret);
 			break;
+		case HCLGE_MBX_GET_LINK_MODE:
+			hclge_get_link_mode(vport, req);
+			break;
 		default:
 			dev_err(&hdev->pdev->dev,
 				"un-supported mailbox message, code = %d\n",
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
index bc2f29e67280..4a897cfb573b 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -381,6 +381,21 @@ void hclgevf_update_link_status(struct hclgevf_dev *hdev, int link_state)
 	}
 }
 
+void hclgevf_update_link_mode(struct hclgevf_dev *hdev)
+{
+#define HCLGEVF_ADVERTISING 0
+#define HCLGEVF_SUPPORTED   1
+	u8 send_msg;
+	u8 resp_msg;
+
+	send_msg = HCLGEVF_ADVERTISING;
+	hclgevf_send_mbx_msg(hdev, HCLGE_MBX_GET_LINK_MODE, 0, &send_msg,
+			     sizeof(u8), false, &resp_msg, sizeof(u8));
+	send_msg = HCLGEVF_SUPPORTED;
+	hclgevf_send_mbx_msg(hdev, HCLGE_MBX_GET_LINK_MODE, 0, &send_msg,
+			     sizeof(u8), false, &resp_msg, sizeof(u8));
+}
+
 static int hclgevf_set_handle_info(struct hclgevf_dev *hdev)
 {
 	struct hnae3_handle *nic = &hdev->nic;
@@ -1646,6 +1661,8 @@ static void hclgevf_service_task(struct work_struct *work)
 	 */
 	hclgevf_request_link_info(hdev);
 
+	hclgevf_update_link_mode(hdev);
+
 	hclgevf_deferred_task_schedule(hdev);
 
 	clear_bit(HCLGEVF_STATE_SERVICE_SCHED, &hdev->state);
@@ -1726,8 +1743,6 @@ static int hclgevf_configure(struct hclgevf_dev *hdev)
 {
 	int ret;
 
-	hdev->hw.mac.media_type = HNAE3_MEDIA_TYPE_NONE;
-
 	/* get queue configuration from PF */
 	ret = hclgevf_get_queue_info(hdev);
 	if (ret)
@@ -1880,6 +1895,8 @@ static int hclgevf_ae_start(struct hnae3_handle *handle)
 
 	hclgevf_request_link_info(hdev);
 
+	hclgevf_update_link_mode(hdev);
+
 	clear_bit(HCLGEVF_STATE_DOWN, &hdev->state);
 
 	return 0;
@@ -2586,6 +2603,16 @@ static unsigned long hclgevf_ae_dev_reset_cnt(struct hnae3_handle *handle)
 	return hdev->reset_count;
 }
 
+static void hclgevf_get_link_mode(struct hnae3_handle *handle,
+				  unsigned long *supported,
+				  unsigned long *advertising)
+{
+	struct hclgevf_dev *hdev = hclgevf_ae_get_hdev(handle);
+
+	*supported = hdev->hw.mac.supported;
+	*advertising = hdev->hw.mac.advertising;
+}
+
 #define MAX_SEPARATE_NUM	4
 #define SEPARATOR_VALUE		0xFFFFFFFF
 #define REG_NUM_PER_LINE	4
@@ -2704,6 +2731,7 @@ static const struct hnae3_ae_ops hclgevf_ops = {
 	.set_mtu = hclgevf_set_mtu,
 	.get_global_queue_id = hclgevf_get_qid_global,
 	.set_timer_task = hclgevf_set_timer_task,
+	.get_link_mode = hclgevf_get_link_mode,
 };
 
 static struct hnae3_ae_algo ae_algovf = {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
index 787bc06944e5..eba1118a766d 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.h
@@ -145,6 +145,8 @@ struct hclgevf_mac {
 	int link;
 	u8 duplex;
 	u32 speed;
+	u64 supported;
+	u64 advertising;
 };
 
 struct hclgevf_hw {
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
index 84653f58b2d1..7dc3c9f79169 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
@@ -197,6 +197,7 @@ void hclgevf_mbx_handler(struct hclgevf_dev *hdev)
 			break;
 		case HCLGE_MBX_LINK_STAT_CHANGE:
 		case HCLGE_MBX_ASSERTING_RESET:
+		case HCLGE_MBX_LINK_STAT_MODE:
 			/* set this mbx event as pending. This is required as we
 			 * might loose interrupt event when mbx task is busy
 			 * handling. This shall be cleared when mbx task just
@@ -247,6 +248,7 @@ void hclgevf_mbx_async_handler(struct hclgevf_dev *hdev)
 	u8 duplex;
 	u32 speed;
 	u32 tail;
+	u8 idx;
 
 	/* we can safely clear it now as we are at start of the async message
 	 * processing
@@ -270,12 +272,22 @@ void hclgevf_mbx_async_handler(struct hclgevf_dev *hdev)
 			link_status = le16_to_cpu(msg_q[1]);
 			memcpy(&speed, &msg_q[2], sizeof(speed));
 			duplex = (u8)le16_to_cpu(msg_q[4]);
+			hdev->hw.mac.media_type = (u8)le16_to_cpu(msg_q[5]);
 
 			/* update upper layer with new link link status */
 			hclgevf_update_link_status(hdev, link_status);
 			hclgevf_update_speed_duplex(hdev, speed, duplex);
 
 			break;
+		case HCLGE_MBX_LINK_STAT_MODE:
+			idx = (u8)le16_to_cpu(msg_q[1]);
+			if (idx)
+				memcpy(&hdev->hw.mac.supported, &msg_q[2],
+				       sizeof(unsigned long));
+			else
+				memcpy(&hdev->hw.mac.advertising, &msg_q[2],
+				       sizeof(unsigned long));
+			break;
 		case HCLGE_MBX_ASSERTING_RESET:
 			/* PF has asserted reset hence VF should go in pending
 			 * state and poll for the hardware reset status till it
-- 
2.20.1

[PATCH net-next 09/12] net: hns3: optimize the maximum TC macro

[Huazhong Tan]

From: liuzhongzhu <liuzhongzhu@huawei.com>

Multiple macros with the largest number of TCs in the system,
optimized to HCLGE_MAX_TC_NUM.

Signed-off-by: liuzhongzhu <liuzhongzhu@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h     | 7 +++----
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c | 4 ++--
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c    | 2 +-
 3 files changed, 6 insertions(+), 7 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
index e26a25128693..ac2cd219a9fb 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.h
@@ -312,16 +312,16 @@ struct hclge_ctrl_vector_chain_cmd {
 	u8 rsv;
 };
 
-#define HCLGE_TC_NUM		8
+#define HCLGE_MAX_TC_NUM		8
 #define HCLGE_TC0_PRI_BUF_EN_B	15 /* Bit 15 indicate enable or not */
 #define HCLGE_BUF_UNIT_S	7  /* Buf size is united by 128 bytes */
 struct hclge_tx_buff_alloc_cmd {
-	__le16 tx_pkt_buff[HCLGE_TC_NUM];
+	__le16 tx_pkt_buff[HCLGE_MAX_TC_NUM];
 	u8 tx_buff_rsv[8];
 };
 
 struct hclge_rx_priv_buff_cmd {
-	__le16 buf_num[HCLGE_TC_NUM];
+	__le16 buf_num[HCLGE_MAX_TC_NUM];
 	__le16 shared_buf;
 	u8 rsv[6];
 };
@@ -367,7 +367,6 @@ struct hclge_priv_buf {
 	u32 enable;	/* Enable TC private buffer or not */
 };
 
-#define HCLGE_MAX_TC_NUM	8
 struct hclge_shared_buf {
 	struct hclge_waterline self;
 	struct hclge_tc_thrd tc_thrd[HCLGE_MAX_TC_NUM];
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
index 26d80504c730..1192cf6f2321 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_debugfs.c
@@ -691,7 +691,7 @@ static void hclge_dbg_dump_qos_buf_cfg(struct hclge_dev *hdev)
 	dev_info(&hdev->pdev->dev, "dump qos buf cfg\n");
 
 	tx_buf_cmd = (struct hclge_tx_buff_alloc_cmd *)desc[0].data;
-	for (i = 0; i < HCLGE_TC_NUM; i++)
+	for (i = 0; i < HCLGE_MAX_TC_NUM; i++)
 		dev_info(&hdev->pdev->dev, "tx_packet_buf_tc_%d: 0x%x\n", i,
 			 tx_buf_cmd->tx_pkt_buff[i]);
 
@@ -703,7 +703,7 @@ static void hclge_dbg_dump_qos_buf_cfg(struct hclge_dev *hdev)
 
 	dev_info(&hdev->pdev->dev, "\n");
 	rx_buf_cmd = (struct hclge_rx_priv_buff_cmd *)desc[0].data;
-	for (i = 0; i < HCLGE_TC_NUM; i++)
+	for (i = 0; i < HCLGE_MAX_TC_NUM; i++)
 		dev_info(&hdev->pdev->dev, "rx_packet_buf_tc_%d: 0x%x\n", i,
 			 rx_buf_cmd->buf_num[i]);
 
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index 5f6fc23fab5d..f55835dee5c7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -1331,7 +1331,7 @@ static int  hclge_cmd_alloc_tx_buff(struct hclge_dev *hdev,
 	req = (struct hclge_tx_buff_alloc_cmd *)desc.data;
 
 	hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_TX_BUFF_ALLOC, 0);
-	for (i = 0; i < HCLGE_TC_NUM; i++) {
+	for (i = 0; i < HCLGE_MAX_TC_NUM; i++) {
 		u32 buf_size = buf_alloc->priv_buf[i].tx_buf_size;
 
 		req->tx_pkt_buff[i] =
-- 
2.20.1

[PATCH net-next 10/12] net: hns3: don't allow user to change vlan filter state

[Huazhong Tan]

From: Jian Shen <shenjian15@huawei.com>

When user disables vlan filter, and adds vlan device, it won't
notify the driver the update the vlan filter. In this case, when
user enables vlan filter again, the packets with new vlan tag
will be filtered by vlan filter.

Signed-off-by: Jian Shen <shenjian15@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index db06d36d408f..40b17223ee41 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1929,8 +1929,7 @@ static void hns3_set_default_feature(struct net_device *netdev)
 		NETIF_F_GSO_UDP_TUNNEL_CSUM | NETIF_F_SCTP_CRC;
 
 	if (pdev->revision >= 0x21) {
-		netdev->hw_features |= NETIF_F_HW_VLAN_CTAG_FILTER |
-			NETIF_F_GRO_HW;
+		netdev->hw_features |= NETIF_F_GRO_HW;
 		netdev->features |= NETIF_F_GRO_HW;
 
 		if (!(h->flags & HNAE3_SUPPORT_VF)) {
-- 
2.20.1

[PATCH net-next 11/12] net: hns3: modify the upper limit judgment condition

[Huazhong Tan]

From: liuzhongzhu <liuzhongzhu@huawei.com>

In order to prevent the variable anomaly from being larger than desc_num,
the upper limit judgment condition becomes >=.

Signed-off-by: liuzhongzhu <liuzhongzhu@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
index 57c50b947857..1dddfcba41a9 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_cmd.c
@@ -197,7 +197,7 @@ static int hclge_cmd_check_retval(struct hclge_hw *hw, struct hclge_desc *desc,
 	for (handle = 0; handle < num; handle++) {
 		desc[handle] = hw->cmq.csq.desc[ntc];
 		ntc++;
-		if (ntc == hw->cmq.csq.desc_num)
+		if (ntc >= hw->cmq.csq.desc_num)
 			ntc = 0;
 	}
 	if (likely(!hclge_is_special_opcode(opcode)))
@@ -254,7 +254,7 @@ int hclge_cmd_send(struct hclge_hw *hw, struct hclge_desc *desc, int num)
 		desc_to_use = &hw->cmq.csq.desc[hw->cmq.csq.next_to_use];
 		*desc_to_use = desc[handle];
 		(hw->cmq.csq.next_to_use)++;
-		if (hw->cmq.csq.next_to_use == hw->cmq.csq.desc_num)
+		if (hw->cmq.csq.next_to_use >= hw->cmq.csq.desc_num)
 			hw->cmq.csq.next_to_use = 0;
 		handle++;
 	}
-- 
2.20.1

[PATCH net-next 12/12] net: hns3: MAC table entry count function increases operation 0 value protection measures

[Huazhong Tan]

From: liuzhongzhu <liuzhongzhu@huawei.com>

When updating the available MAC VLAN table counts,
MAC VLAN table entry count function adds
operation 0 value protection measures.

Signed-off-by: liuzhongzhu <liuzhongzhu@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index f55835dee5c7..0a64c7f2dc6c 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -5830,9 +5830,12 @@ static void hclge_update_umv_space(struct hclge_vport *vport, bool is_free)
 	if (is_free) {
 		if (vport->used_umv_num > hdev->priv_umv_size)
 			hdev->share_umv_size++;
-		vport->used_umv_num--;
+
+		if (vport->used_umv_num > 0)
+			vport->used_umv_num--;
 	} else {
-		if (vport->used_umv_num >= hdev->priv_umv_size)
+		if (vport->used_umv_num >= hdev->priv_umv_size &&
+		    hdev->share_umv_size > 0)
 			hdev->share_umv_size--;
 		vport->used_umv_num++;
 	}
-- 
2.20.1

Re: [PATCH net-next 00/12] code optimizations & bugfixes for HNS3 driver

[David Miller]

From: Huazhong Tan <tanhuazhong@huawei.com>
Date: Sat, 2 Feb 2019 22:39:25 +0800

> This patchset includes bugfixes and code optimizations for the HNS3
> ethernet controller driver

Series applied.

Re: [PATCH net-next 10/12] net: hns3: don't allow user to change vlan filter state

[Sergei Shtylyov]

Hello!

On 02.02.2019 17:39, Huazhong Tan wrote:

> From: Jian Shen <shenjian15@huawei.com>
> 
> When user disables vlan filter, and adds vlan device, it won't
> notify the driver the update the vlan filter. In this case, when

    To update, perhaps?

> user enables vlan filter again, the packets with new vlan tag
> will be filtered by vlan filter.
> 
> Signed-off-by: Jian Shen <shenjian15@huawei.com>
> Signed-off-by: Peng Li <lipeng321@huawei.com>
> Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
[...]

MBR, Sergei