[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 (4):
  net: hns3: simplify hclgevf_cmd_csq_clean
  net: hns3: check resetting status in hns3_get_stats()
  net: hns3: prevent change MTU when resetting
  net: hns3: modify HNS3_NIC_STATE_INITED flag in
    hns3_reset_notify_uninit_enet

Jian Shen (2):
  net: hns3: add protect when handling mac addr list
  net: hns3: split function hnae3_match_n_instantiate()

Peng Li (2):
  net: hns3: check 1000M half for hns3_ethtool_ops.set_link_ksettings
  net: hns3: return 0 and print warning when hit duplicate MAC

Yonglong Liu (1):
  net: hns3: reduce resources use in kdump kernel

Yunsheng Lin (2):
  net: hns3: minor optimization for ring_space
  net: hns3: minor optimization for datapath

liuzhongzhu (1):
  net: hns3: modify the VF network port media type acquisition method

 drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h    |  1 +
 drivers/net/ethernet/hisilicon/hns3/hnae3.c        | 40 +++++++++++----------
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.c    | 41 +++++++++++++++++-----
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.h    | 15 +++-----
 drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c |  9 +++++
 .../ethernet/hisilicon/hns3/hns3pf/hclge_main.c    | 30 ++++++++++++++--
 .../net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c | 23 +++++++++---
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c   | 35 ++++++++++++------
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c  | 23 ++++++++++++
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c   |  1 -
 10 files changed, 163 insertions(+), 55 deletions(-)

-- 
2.7.4

[PATCH net-next 01/12] net: hns3: check 1000M half for hns3_ethtool_ops.set_link_ksettings

[Huazhong Tan]

From: Peng Li <lipeng321@huawei.com>

Hip08 SOC does not support 1000M half, this patch adds 1000M half
check for hns3_ethtool_ops.set_link_ksettings, so the user can not
set 1000M half by ethtool.

Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 359d473..0c05807 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -648,6 +648,10 @@ static int hns3_get_link_ksettings(struct net_device *netdev,
 static int hns3_set_link_ksettings(struct net_device *netdev,
 				   const struct ethtool_link_ksettings *cmd)
 {
+	/* Chip don't support this mode. */
+	if (cmd->base.speed == SPEED_1000 && cmd->base.duplex == DUPLEX_HALF)
+		return -EINVAL;
+
 	/* Only support ksettings_set for netdev with phy attached for now */
 	if (netdev->phydev)
 		return phy_ethtool_ksettings_set(netdev->phydev, cmd);
-- 
2.7.4

Re: [PATCH net-next 01/12] net: hns3: check 1000M half for hns3_ethtool_ops.set_link_ksettings

[Sergei Shtylyov]

Hello!

On 03.04.2019 6:07, Huazhong Tan wrote:

> From: Peng Li <lipeng321@huawei.com>
> 
> Hip08 SOC does not support 1000M half, this patch adds 1000M half
> check for hns3_ethtool_ops.set_link_ksettings, so the user can not
> set 1000M half by ethtool.
> 
> Signed-off-by: Peng Li <lipeng321@huawei.com>
> Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
> ---
>   drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 4 ++++
>   1 file changed, 4 insertions(+)
> 
> diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
> index 359d473..0c05807 100644
> --- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
> @@ -648,6 +648,10 @@ static int hns3_get_link_ksettings(struct net_device *netdev,
>   static int hns3_set_link_ksettings(struct net_device *netdev,
>   				   const struct ethtool_link_ksettings *cmd)
>   {
> +	/* Chip don't support this mode. */

    Doesn't.

[...]

MBR, Sergei

Re: [PATCH net-next 01/12] net: hns3: check 1000M half for hns3_ethtool_ops.set_link_ksettings

[tanhuazhong]

On 2019/4/3 17:12, Sergei Shtylyov wrote:
> Hello!
> 
> On 03.04.2019 6:07, Huazhong Tan wrote:
> 
>> From: Peng Li <lipeng321@huawei.com>
>>
>> Hip08 SOC does not support 1000M half, this patch adds 1000M half
>> check for hns3_ethtool_ops.set_link_ksettings, so the user can not
>> set 1000M half by ethtool.
>>
>> Signed-off-by: Peng Li <lipeng321@huawei.com>
>> Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
>> ---
>>   drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 4 ++++
>>   1 file changed, 4 insertions(+)
>>
>> diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c 
>> b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
>> index 359d473..0c05807 100644
>> --- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
>> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
>> @@ -648,6 +648,10 @@ static int hns3_get_link_ksettings(struct 
>> net_device *netdev,
>>   static int hns3_set_link_ksettings(struct net_device *netdev,
>>                      const struct ethtool_link_ksettings *cmd)
>>   {
>> +    /* Chip don't support this mode. */
> 
>     Doesn't.
> 
> [...]
> 
> MBR, Sergei
> 

ok, thanks.

> .
> 

[PATCH net-next 02/12] net: hns3: reduce resources use in kdump kernel

[Huazhong Tan]

From: Yonglong Liu <liuyonglong@huawei.com>

When the kdump kernel started, the HNS3 driver fail to register:
[14.753340] hns3 0000:7d:00.0: Alloc umv space failed, want 512, get 0
[14.795034] hns3 0000:7d:00.0: add uc mac address fail, ret =-22.

By default, the HNS3 driver will use about 512M memory, but
usually the reserved memory of kdump kernel is 576M, so the HNS3
driver fail to register. This patch reduces the memory use in
kdump kernel to about 16M.

And when the kdump kernel starts, we must clear ucast mac address
first to avoid add fail.

Signed-off-by: Yonglong Liu <liuyonglong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
---
 .../ethernet/hisilicon/hns3/hns3pf/hclge_main.c    | 23 +++++++++++++++++++++-
 1 file changed, 22 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
index deda606..e742a05 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -12,6 +12,7 @@
 #include <linux/pci.h>
 #include <linux/platform_device.h>
 #include <linux/if_vlan.h>
+#include <linux/crash_dump.h>
 #include <net/rtnetlink.h>
 #include "hclge_cmd.h"
 #include "hclge_dcb.h"
@@ -1015,6 +1016,23 @@ static int hclge_get_cap(struct hclge_dev *hdev)
 	return ret;
 }
 
+static void hclge_init_kdump_kernel_config(struct hclge_dev *hdev)
+{
+#define HCLGE_MIN_TX_DESC	64
+#define HCLGE_MIN_RX_DESC	64
+
+	if (!is_kdump_kernel())
+		return;
+
+	dev_info(&hdev->pdev->dev,
+		 "Running kdump kernel. Using minimal resources\n");
+
+	/* minimal queue pairs equals to the number of vports */
+	hdev->num_tqps = hdev->num_vmdq_vport + hdev->num_req_vfs + 1;
+	hdev->num_tx_desc = HCLGE_MIN_TX_DESC;
+	hdev->num_rx_desc = HCLGE_MIN_RX_DESC;
+}
+
 static int hclge_configure(struct hclge_dev *hdev)
 {
 	struct hclge_cfg cfg;
@@ -1074,6 +1092,8 @@ static int hclge_configure(struct hclge_dev *hdev)
 
 	hdev->tx_sch_mode = HCLGE_FLAG_TC_BASE_SCH_MODE;
 
+	hclge_init_kdump_kernel_config(hdev);
+
 	return ret;
 }
 
@@ -6293,7 +6313,8 @@ static int hclge_set_mac_addr(struct hnae3_handle *handle, void *p,
 		return -EINVAL;
 	}
 
-	if (!is_first && hclge_rm_uc_addr(handle, hdev->hw.mac.mac_addr))
+	if ((!is_first || is_kdump_kernel()) &&
+	    hclge_rm_uc_addr(handle, hdev->hw.mac.mac_addr))
 		dev_warn(&hdev->pdev->dev,
 			 "remove old uc mac address fail.\n");
 
-- 
2.7.4

[PATCH net-next 03/12] net: hns3: modify the VF network port media type acquisition method

[Huazhong Tan]

From: liuzhongzhu <liuzhongzhu@huawei.com>

Method for obtaining the media type of the VF network port periodically,
regular tasks will not run until the network port UP. When the network
port is DOWN, the network port cannot obtain the media type.
Modifies the madie type obtained when initializing the VF network port.

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/hclge_mbx.h    |  1 +
 .../net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c | 23 ++++++++++++++++++----
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c  | 23 ++++++++++++++++++++++
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c   |  1 -
 4 files changed, 43 insertions(+), 5 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h b/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
index 299b277..9d3d45f 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hclge_mbx.h
@@ -43,6 +43,7 @@ enum HCLGE_MBX_OPCODE {
 	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 */
+	HCLGE_MBX_GET_MEDIA_TYPE,       /* (VF -> PF) get media type */
 
 	HCLGE_MBX_GET_VF_FLR_STATUS = 200, /* (M7 -> PF) get vf reset status */
 };
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
index 306a23e..9aa9c64 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_mbx.c
@@ -385,24 +385,32 @@ static int hclge_get_vf_queue_depth(struct hclge_vport *vport,
 				    HCLGE_TQPS_DEPTH_INFO_LEN);
 }
 
+static int hclge_get_vf_media_type(struct hclge_vport *vport,
+				   struct hclge_mbx_vf_to_pf_cmd *mbx_req)
+{
+	struct hclge_dev *hdev = vport->back;
+	u8 resp_data;
+
+	resp_data = hdev->hw.mac.media_type;
+	return hclge_gen_resp_to_vf(vport, mbx_req, 0, &resp_data,
+				    sizeof(resp_data));
+}
+
 static int hclge_get_link_info(struct hclge_vport *vport,
 			       struct hclge_mbx_vf_to_pf_cmd *mbx_req)
 {
 	struct hclge_dev *hdev = vport->back;
 	u16 link_status;
-	u8 msg_data[10];
-	u16 media_type;
+	u8 msg_data[8];
 	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 */
@@ -662,6 +670,13 @@ void hclge_mbx_handler(struct hclge_dev *hdev)
 			hclge_rm_vport_all_vlan_table(vport, true);
 			mutex_unlock(&hdev->vport_cfg_mutex);
 			break;
+		case HCLGE_MBX_GET_MEDIA_TYPE:
+			ret = hclge_get_vf_media_type(vport, req);
+			if (ret)
+				dev_err(&hdev->pdev->dev,
+					"PF fail(%d) to media type for VF\n",
+					ret);
+			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 65bdc68..509ff93 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_main.c
@@ -307,6 +307,25 @@ static u16 hclgevf_get_qid_global(struct hnae3_handle *handle, u16 queue_id)
 	return qid_in_pf;
 }
 
+static int hclgevf_get_pf_media_type(struct hclgevf_dev *hdev)
+{
+	u8 resp_msg;
+	int ret;
+
+	ret = hclgevf_send_mbx_msg(hdev, HCLGE_MBX_GET_MEDIA_TYPE, 0, NULL, 0,
+				   true, &resp_msg, sizeof(resp_msg));
+	if (ret) {
+		dev_err(&hdev->pdev->dev,
+			"VF request to get the pf port media type failed %d",
+			ret);
+		return ret;
+	}
+
+	hdev->hw.mac.media_type = resp_msg;
+
+	return 0;
+}
+
 static int hclgevf_alloc_tqps(struct hclgevf_dev *hdev)
 {
 	struct hclgevf_tqp *tqp;
@@ -1824,6 +1843,10 @@ static int hclgevf_configure(struct hclgevf_dev *hdev)
 	if (ret)
 		return ret;
 
+	ret = hclgevf_get_pf_media_type(hdev);
+	if (ret)
+		return ret;
+
 	/* get tc configuration from PF */
 	return hclgevf_get_tc_info(hdev);
 }
diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
index 7dc3c9f..3bcf49b 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_mbx.c
@@ -272,7 +272,6 @@ 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);
-- 
2.7.4

[PATCH net-next 04/12] net: hns3: return 0 and print warning when hit duplicate MAC

[Huazhong Tan]

From: Peng Li <lipeng321@huawei.com>

When set 2 same MAC to different function of one port, IMP
will return error as the later one may modify the origin one.
This will cause bond fail for 2 VFs of one port.

Driver just print warning and return 0 with this patch, so
if set same MAC address, it will return 0 but do not really
configure to HW.

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 e742a05..d45789e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
@@ -5962,8 +5962,11 @@ int hclge_add_uc_addr_common(struct hclge_vport *vport,
 	}
 
 	/* check if we just hit the duplicate */
-	if (!ret)
-		ret = -EINVAL;
+	if (!ret) {
+		dev_warn(&hdev->pdev->dev, "VF %d mac(%pM) has been existed\n",
+			 vport->vport_id, addr);
+		return 0;
+	}
 
 	dev_err(&hdev->pdev->dev,
 		"PF failed to add unicast entry(%pM) in the MAC table\n",
-- 
2.7.4

Re: [PATCH net-next 04/12] net: hns3: return 0 and print warning when hit duplicate MAC

[Sergei Shtylyov]

On 03.04.2019 6:07, Huazhong Tan wrote:

> From: Peng Li <lipeng321@huawei.com>
> 
> When set 2 same MAC to different function of one port, IMP
> will return error as the later one may modify the origin one.
> This will cause bond fail for 2 VFs of one port.
> 
> Driver just print warning and return 0 with this patch, so
> if set same MAC address, it will return 0 but do not really
> configure to HW.

    "To" not needed here.

> 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 e742a05..d45789e 100644
> --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
> @@ -5962,8 +5962,11 @@ int hclge_add_uc_addr_common(struct hclge_vport *vport,
>   	}
>   
>   	/* check if we just hit the duplicate */
> -	if (!ret)
> -		ret = -EINVAL;
> +	if (!ret) {
> +		dev_warn(&hdev->pdev->dev, "VF %d mac(%pM) has been existed\n",

    Maybe just "exists"? The grammar is certainly wrong as is. :-)

> +			 vport->vport_id, addr);
> +		return 0;
> +	}
>   
>   	dev_err(&hdev->pdev->dev,
>   		"PF failed to add unicast entry(%pM) in the MAC table\n",

MBR, Sergei

Re: [PATCH net-next 04/12] net: hns3: return 0 and print warning when hit duplicate MAC

[tanhuazhong]

On 2019/4/3 17:15, Sergei Shtylyov wrote:
> On 03.04.2019 6:07, Huazhong Tan wrote:
> 
>> From: Peng Li <lipeng321@huawei.com>
>>
>> When set 2 same MAC to different function of one port, IMP
>> will return error as the later one may modify the origin one.
>> This will cause bond fail for 2 VFs of one port.
>>
>> Driver just print warning and return 0 with this patch, so
>> if set same MAC address, it will return 0 but do not really
>> configure to HW.
> 
>     "To" not needed here.
> 

ok, thanks.

>> 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 e742a05..d45789e 100644
>> --- a/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
>> +++ b/drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c
>> @@ -5962,8 +5962,11 @@ int hclge_add_uc_addr_common(struct hclge_vport 
>> *vport,
>>       }
>>       /* check if we just hit the duplicate */
>> -    if (!ret)
>> -        ret = -EINVAL;
>> +    if (!ret) {
>> +        dev_warn(&hdev->pdev->dev, "VF %d mac(%pM) has been existed\n",
> 
>     Maybe just "exists"? The grammar is certainly wrong as is. :-)
> 
>> +             vport->vport_id, addr);
>> +        return 0;
>> +    }
>>       dev_err(&hdev->pdev->dev,
>>           "PF failed to add unicast entry(%pM) in the MAC table\n",
> 
> MBR, Sergei
> 

ok, thanks.

> .
> 

[PATCH net-next 05/12] net: hns3: minor optimization for ring_space

[Huazhong Tan]

From: Yunsheng Lin <linyunsheng@huawei.com>

This patch optimizes the ring_space by calculating the
ring space without calling ring_dist.

Also ring_dist is only used by ring_space, so this patch
removes it when it is no longer used.

Signed-off-by: Yunsheng Lin <linyunsheng@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.h | 15 +++++----------
 1 file changed, 5 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
index 75669cd..025d0f7 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.h
@@ -577,18 +577,13 @@ union l4_hdr_info {
 	unsigned char *hdr;
 };
 
-/* the distance between [begin, end) in a ring buffer
- * note: there is a unuse slot between the begin and the end
- */
-static inline int ring_dist(struct hns3_enet_ring *ring, int begin, int end)
-{
-	return (end - begin + ring->desc_num) % ring->desc_num;
-}
-
 static inline int ring_space(struct hns3_enet_ring *ring)
 {
-	return ring->desc_num -
-		ring_dist(ring, ring->next_to_clean, ring->next_to_use) - 1;
+	int begin = ring->next_to_clean;
+	int end = ring->next_to_use;
+
+	return ((end >= begin) ? (ring->desc_num - end + begin) :
+			(begin - end)) - 1;
 }
 
 static inline int is_ring_empty(struct hns3_enet_ring *ring)
-- 
2.7.4

[PATCH net-next 06/12] net: hns3: minor optimization for datapath

[Huazhong Tan]

From: Yunsheng Lin <linyunsheng@huawei.com>

This patch adds a likely case for hns3_fill_desc and
limits the local variables' scope as much as possible,
also avoid div operation when the tqp_vector->num_tqps
is one.

Signed-off-by: Yunsheng Lin <linyunsheng@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 | 24 +++++++++++++++++++++---
 1 file changed, 21 insertions(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 21085c4..5b34d2e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1012,7 +1012,6 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv,
 	struct hns3_desc_cb *desc_cb = &ring->desc_cb[ring->next_to_use];
 	struct hns3_desc *desc = &ring->desc[ring->next_to_use];
 	struct device *dev = ring_to_dev(ring);
-	u16 bdtp_fe_sc_vld_ra_ri = 0;
 	struct skb_frag_struct *frag;
 	unsigned int frag_buf_num;
 	int k, sizeoflast;
@@ -1080,12 +1079,30 @@ static int hns3_fill_desc(struct hns3_enet_ring *ring, void *priv,
 
 	desc_cb->length = size;
 
+	if (likely(size <= HNS3_MAX_BD_SIZE)) {
+		u16 bdtp_fe_sc_vld_ra_ri = 0;
+
+		desc_cb->priv = priv;
+		desc_cb->dma = dma;
+		desc_cb->type = type;
+		desc->addr = cpu_to_le64(dma);
+		desc->tx.send_size = cpu_to_le16(size);
+		hns3_set_txbd_baseinfo(&bdtp_fe_sc_vld_ra_ri, frag_end);
+		desc->tx.bdtp_fe_sc_vld_ra_ri =
+			cpu_to_le16(bdtp_fe_sc_vld_ra_ri);
+
+		ring_ptr_move_fw(ring, next_to_use);
+		return 0;
+	}
+
 	frag_buf_num = hns3_tx_bd_count(size);
 	sizeoflast = size & HNS3_TX_LAST_SIZE_M;
 	sizeoflast = sizeoflast ? sizeoflast : HNS3_MAX_BD_SIZE;
 
 	/* When frag size is bigger than hardware limit, split this frag */
 	for (k = 0; k < frag_buf_num; k++) {
+		u16 bdtp_fe_sc_vld_ra_ri = 0;
+
 		/* The txbd's baseinfo of DESC_TYPE_PAGE & DESC_TYPE_SKB */
 		desc_cb->priv = priv;
 		desc_cb->dma = dma + HNS3_MAX_BD_SIZE * k;
@@ -2891,7 +2908,7 @@ static int hns3_nic_common_poll(struct napi_struct *napi, int budget)
 	struct hns3_enet_tqp_vector *tqp_vector =
 		container_of(napi, struct hns3_enet_tqp_vector, napi);
 	bool clean_complete = true;
-	int rx_budget;
+	int rx_budget = budget;
 
 	if (unlikely(test_bit(HNS3_NIC_STATE_DOWN, &priv->state))) {
 		napi_complete(napi);
@@ -2905,7 +2922,8 @@ static int hns3_nic_common_poll(struct napi_struct *napi, int budget)
 		hns3_clean_tx_ring(ring);
 
 	/* make sure rx ring budget not smaller than 1 */
-	rx_budget = max(budget / tqp_vector->num_tqps, 1);
+	if (tqp_vector->num_tqps > 1)
+		rx_budget = max(budget / tqp_vector->num_tqps, 1);
 
 	hns3_for_each_ring(ring, tqp_vector->rx_group) {
 		int rx_cleaned = hns3_clean_rx_ring(ring, rx_budget,
-- 
2.7.4

[PATCH net-next 07/12] net: hns3: simplify hclgevf_cmd_csq_clean

[Huazhong Tan]

csq is used as a ring buffer, the value of the desc will be replaced
in next use. This patch removes the unnecessary memset, and just
updates the next_to_clean.

Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
---
 .../ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c   | 35 +++++++++++++++-------
 1 file changed, 24 insertions(+), 11 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
index 9441b45..930b175 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3vf/hclgevf_cmd.c
@@ -27,26 +27,39 @@ static int hclgevf_ring_space(struct hclgevf_cmq_ring *ring)
 	return ring->desc_num - used - 1;
 }
 
+static int hclgevf_is_valid_csq_clean_head(struct hclgevf_cmq_ring *ring,
+					   int head)
+{
+	int ntu = ring->next_to_use;
+	int ntc = ring->next_to_clean;
+
+	if (ntu > ntc)
+		return head >= ntc && head <= ntu;
+
+	return head >= ntc || head <= ntu;
+}
+
 static int hclgevf_cmd_csq_clean(struct hclgevf_hw *hw)
 {
+	struct hclgevf_dev *hdev = container_of(hw, struct hclgevf_dev, hw);
 	struct hclgevf_cmq_ring *csq = &hw->cmq.csq;
-	u16 ntc = csq->next_to_clean;
-	struct hclgevf_desc *desc;
 	int clean = 0;
 	u32 head;
 
-	desc = &csq->desc[ntc];
 	head = hclgevf_read_dev(hw, HCLGEVF_NIC_CSQ_HEAD_REG);
-	while (head != ntc) {
-		memset(desc, 0, sizeof(*desc));
-		ntc++;
-		if (ntc == csq->desc_num)
-			ntc = 0;
-		desc = &csq->desc[ntc];
-		clean++;
+	rmb(); /* Make sure head is ready before touch any data */
+
+	if (!hclgevf_is_valid_csq_clean_head(csq, head)) {
+		dev_warn(&hdev->pdev->dev, "wrong cmd head (%d, %d-%d)\n", head,
+			 csq->next_to_use, csq->next_to_clean);
+		dev_warn(&hdev->pdev->dev,
+			 "Disabling any further commands to IMP firmware\n");
+		set_bit(HCLGEVF_STATE_CMD_DISABLE, &hdev->state);
+		return -EIO;
 	}
-	csq->next_to_clean = ntc;
 
+	clean = (head - csq->next_to_clean + csq->desc_num) % csq->desc_num;
+	csq->next_to_clean = head;
 	return clean;
 }
 
-- 
2.7.4

[PATCH net-next 08/12] net: hns3: add protect when handling mac addr list

[Huazhong Tan]

From: Jian Shen <shenjian15@huawei.com>

It used netdev->uc and netdev->mc list in function
hns3_recover_hw_addr() and hns3_remove_hw_addr().
We should add protect for them.

Fixes: f05e21097121 ("net: hns3: Clear mac vlan table entries when unload driver or function reset")
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 | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 5b34d2e..476c23d 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -3791,12 +3791,13 @@ static int hns3_recover_hw_addr(struct net_device *ndev)
 	struct netdev_hw_addr *ha, *tmp;
 	int ret = 0;
 
+	netif_addr_lock_bh(ndev);
 	/* go through and sync uc_addr entries to the device */
 	list = &ndev->uc;
 	list_for_each_entry_safe(ha, tmp, &list->list, list) {
 		ret = hns3_nic_uc_sync(ndev, ha->addr);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
 	/* go through and sync mc_addr entries to the device */
@@ -3804,9 +3805,11 @@ static int hns3_recover_hw_addr(struct net_device *ndev)
 	list_for_each_entry_safe(ha, tmp, &list->list, list) {
 		ret = hns3_nic_mc_sync(ndev, ha->addr);
 		if (ret)
-			return ret;
+			goto out;
 	}
 
+out:
+	netif_addr_unlock_bh(ndev);
 	return ret;
 }
 
@@ -3817,6 +3820,7 @@ static void hns3_remove_hw_addr(struct net_device *netdev)
 
 	hns3_nic_uc_unsync(netdev, netdev->dev_addr);
 
+	netif_addr_lock_bh(netdev);
 	/* go through and unsync uc_addr entries to the device */
 	list = &netdev->uc;
 	list_for_each_entry_safe(ha, tmp, &list->list, list)
@@ -3827,6 +3831,8 @@ static void hns3_remove_hw_addr(struct net_device *netdev)
 	list_for_each_entry_safe(ha, tmp, &list->list, list)
 		if (ha->refcount > 1)
 			hns3_nic_mc_unsync(netdev, ha->addr);
+
+	netif_addr_unlock_bh(netdev);
 }
 
 static void hns3_clear_tx_ring(struct hns3_enet_ring *ring)
-- 
2.7.4

[PATCH net-next 09/12] net: hns3: check resetting status in hns3_get_stats()

[Huazhong Tan]

hns3_get_stats() should check the resetting status firstly,
since the device will be reinitialized when resetting. If the
reset has not completed, the hns3_get_stats() may access
invalid memory.

Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c | 5 +++++
 1 file changed, 5 insertions(+)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
index 0c05807..482523e 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_ethtool.c
@@ -483,6 +483,11 @@ static void hns3_get_stats(struct net_device *netdev,
 	struct hnae3_handle *h = hns3_get_handle(netdev);
 	u64 *p = data;
 
+	if (hns3_nic_resetting(netdev)) {
+		netdev_err(netdev, "dev resetting, could not get stats\n");
+		return;
+	}
+
 	if (!h->ae_algo->ops->get_stats || !h->ae_algo->ops->update_stats) {
 		netdev_err(netdev, "could not get any statistics\n");
 		return;
-- 
2.7.4

[PATCH net-next 10/12] net: hns3: prevent change MTU when resetting

[Huazhong Tan]

When resetting, the changing of MTU is not allowed, so this patch
adds checking reset status in hns3_nic_change_mtu() to do that.

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

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 476c23d..0e31f74 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -1591,6 +1591,9 @@ static int hns3_nic_change_mtu(struct net_device *netdev, int new_mtu)
 	struct hnae3_handle *h = hns3_get_handle(netdev);
 	int ret;
 
+	if (hns3_nic_resetting(netdev))
+		return -EBUSY;
+
 	if (!h->ae_algo->ops->set_mtu)
 		return -EOPNOTSUPP;
 
-- 
2.7.4

[PATCH net-next 11/12] net: hns3: modify HNS3_NIC_STATE_INITED flag in hns3_reset_notify_uninit_enet

[Huazhong Tan]

In the hns3_reset_notify_uninit_enet() HNS3_NIC_STATE_INITED flag
should be checked and cleared firstly.

Signed-off-by: Huazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns3/hns3_enet.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
index 0e31f74..0f389a4 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hns3_enet.c
@@ -4128,7 +4128,7 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
 	struct hns3_nic_priv *priv = netdev_priv(netdev);
 	int ret;
 
-	if (!test_bit(HNS3_NIC_STATE_INITED, &priv->state)) {
+	if (!test_and_clear_bit(HNS3_NIC_STATE_INITED, &priv->state)) {
 		netdev_warn(netdev, "already uninitialized\n");
 		return 0;
 	}
@@ -4150,8 +4150,6 @@ static int hns3_reset_notify_uninit_enet(struct hnae3_handle *handle)
 	hns3_put_ring_config(priv);
 	priv->ring_data = NULL;
 
-	clear_bit(HNS3_NIC_STATE_INITED, &priv->state);
-
 	return ret;
 }
 
-- 
2.7.4

[PATCH net-next 12/12] net: hns3: split function hnae3_match_n_instantiate()

[Huazhong Tan]

From: Jian Shen <shenjian15@huawei.com>

The function hnae3_match_n_instantiate() was called both by
initializing or uninitializing client. For uninitializing, the
return value was never used.

To make it more clear, this patch splits it to two functions,
hnae3_init_client_instance() and hnae3_uninit_client_instance().

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/hnae3.c | 40 ++++++++++++++++-------------
 1 file changed, 22 insertions(+), 18 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns3/hnae3.c b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
index 17ab4f4..fa8b850 100644
--- a/drivers/net/ethernet/hisilicon/hns3/hnae3.c
+++ b/drivers/net/ethernet/hisilicon/hns3/hnae3.c
@@ -76,8 +76,8 @@ static int hnae3_get_client_init_flag(struct hnae3_client *client,
 	return inited;
 }
 
-static int hnae3_match_n_instantiate(struct hnae3_client *client,
-				     struct hnae3_ae_dev *ae_dev, bool is_reg)
+static int hnae3_init_client_instance(struct hnae3_client *client,
+				      struct hnae3_ae_dev *ae_dev)
 {
 	int ret;
 
@@ -87,23 +87,27 @@ static int hnae3_match_n_instantiate(struct hnae3_client *client,
 		return 0;
 	}
 
-	/* now, (un-)instantiate client by calling lower layer */
-	if (is_reg) {
-		ret = ae_dev->ops->init_client_instance(client, ae_dev);
-		if (ret)
-			dev_err(&ae_dev->pdev->dev,
-				"fail to instantiate client, ret = %d\n", ret);
+	ret = ae_dev->ops->init_client_instance(client, ae_dev);
+	if (ret)
+		dev_err(&ae_dev->pdev->dev,
+			"fail to instantiate client, ret = %d\n", ret);
 
-		return ret;
-	}
+	return ret;
+}
+
+static void hnae3_uninit_client_instance(struct hnae3_client *client,
+					 struct hnae3_ae_dev *ae_dev)
+{
+	/* check if this client matches the type of ae_dev */
+	if (!(hnae3_client_match(client->type, ae_dev->dev_type) &&
+	      hnae3_get_bit(ae_dev->flag, HNAE3_DEV_INITED_B)))
+		return;
 
 	if (hnae3_get_client_init_flag(client, ae_dev)) {
 		ae_dev->ops->uninit_client_instance(client, ae_dev);
 
 		hnae3_set_client_init_flag(client, ae_dev, 0);
 	}
-
-	return 0;
 }
 
 int hnae3_register_client(struct hnae3_client *client)
@@ -129,7 +133,7 @@ int hnae3_register_client(struct hnae3_client *client)
 		/* if the client could not be initialized on current port, for
 		 * any error reasons, move on to next available port
 		 */
-		ret = hnae3_match_n_instantiate(client, ae_dev, true);
+		ret = hnae3_init_client_instance(client, ae_dev);
 		if (ret)
 			dev_err(&ae_dev->pdev->dev,
 				"match and instantiation failed for port, ret = %d\n",
@@ -153,7 +157,7 @@ void hnae3_unregister_client(struct hnae3_client *client)
 	mutex_lock(&hnae3_common_lock);
 	/* un-initialize the client on every matched port */
 	list_for_each_entry(ae_dev, &hnae3_ae_dev_list, node) {
-		hnae3_match_n_instantiate(client, ae_dev, false);
+		hnae3_uninit_client_instance(client, ae_dev);
 	}
 
 	list_del(&client->node);
@@ -205,7 +209,7 @@ void hnae3_register_ae_algo(struct hnae3_ae_algo *ae_algo)
 		 * initialize the figure out client instance
 		 */
 		list_for_each_entry(client, &hnae3_client_list, node) {
-			ret = hnae3_match_n_instantiate(client, ae_dev, true);
+			ret = hnae3_init_client_instance(client, ae_dev);
 			if (ret)
 				dev_err(&ae_dev->pdev->dev,
 					"match and instantiation failed, ret = %d\n",
@@ -243,7 +247,7 @@ void hnae3_unregister_ae_algo(struct hnae3_ae_algo *ae_algo)
 		 * un-initialize the figure out client instance
 		 */
 		list_for_each_entry(client, &hnae3_client_list, node)
-			hnae3_match_n_instantiate(client, ae_dev, false);
+			hnae3_uninit_client_instance(client, ae_dev);
 
 		ae_algo->ops->uninit_ae_dev(ae_dev);
 		hnae3_set_bit(ae_dev->flag, HNAE3_DEV_INITED_B, 0);
@@ -301,7 +305,7 @@ int hnae3_register_ae_dev(struct hnae3_ae_dev *ae_dev)
 	 * initialize the figure out client instance
 	 */
 	list_for_each_entry(client, &hnae3_client_list, node) {
-		ret = hnae3_match_n_instantiate(client, ae_dev, true);
+		ret = hnae3_init_client_instance(client, ae_dev);
 		if (ret)
 			dev_err(&ae_dev->pdev->dev,
 				"match and instantiation failed, ret = %d\n",
@@ -343,7 +347,7 @@ void hnae3_unregister_ae_dev(struct hnae3_ae_dev *ae_dev)
 			continue;
 
 		list_for_each_entry(client, &hnae3_client_list, node)
-			hnae3_match_n_instantiate(client, ae_dev, false);
+			hnae3_uninit_client_instance(client, ae_dev);
 
 		ae_algo->ops->uninit_ae_dev(ae_dev);
 		hnae3_set_bit(ae_dev->flag, HNAE3_DEV_INITED_B, 0);
-- 
2.7.4