[PATCH net-next 0/8] be2net patch-set

[PATCH net-next 0/8] be2net patch-set

[Sriharsha Basavapatna]

Hi David,

The following patch set contains a few non-critical bug fixes. Please
consider applying this to the net-next tree. Thanks.

Patch-1 fixes be_set_phys_id() ethtool function to return an error code.
Patch-2 fixes a warning when some commands fail for VFs.
Patch-3 fixes be_vlan_rem_vid() to verify vlan being removed is in the list.
Patch-4 improves SRIOV queue distribution logic.
Patch-5 avoids running self test on VFs.
Patch-6 fixes error recovery in Lancer to clean up after moving to ready state.
Patch-7 adds retry logic to error recovery in case of recovery failures
Patch-8 fixes time interval used in eq delay computation routine

********

Padmanabh Ratnakar (3):
  benet: Fix Lancer error recovery
  benet: Add retry in case of error recovery failure
  benet: Fix interval calculation in interrupt moderation

Somnath Kotur (1):
  be2net: Don't run ethtool self-tests for VFs

Sriharsha Basavapatna (2):
  be2net: Fix be_vlan_rem_vid() to check vlan id being removed
  be2net: SRIOV Queue distribution should factor in EQ-count of VFs

Suresh Reddy (2):
  be2net: return error status from be_set_phys_id()
  be2net: check for INSUFFICIENT_PRIVILEGES error

 drivers/net/ethernet/emulex/benet/be.h         |  9 +++
 drivers/net/ethernet/emulex/benet/be_cmds.c    | 32 ++++++++-
 drivers/net/ethernet/emulex/benet/be_cmds.h    |  3 +-
 drivers/net/ethernet/emulex/benet/be_ethtool.c | 23 ++++---
 drivers/net/ethernet/emulex/benet/be_main.c    | 93 +++++++++++++++++---------
 5 files changed, 114 insertions(+), 46 deletions(-)

-- 
2.3.0.rc2

[PATCH net-next 1/8] be2net: return error status from be_set_phys_id()

[Sriharsha Basavapatna]

From: Suresh Reddy <suresh.reddy@broadcom.com>

be_set_phys_id() returns 0 to ethtool when the command fails in the FW.

This patch fixes the set_phys_id() to return -EIO in case the FW cmd fails.

Signed-off-by: Suresh Reddy <suresh.reddy@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_ethtool.c | 23 +++++++++++++----------
 1 file changed, 13 insertions(+), 10 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be_ethtool.c b/drivers/net/ethernet/emulex/benet/be_ethtool.c
index a19ac44..2ff6916 100644
--- a/drivers/net/ethernet/emulex/benet/be_ethtool.c
+++ b/drivers/net/ethernet/emulex/benet/be_ethtool.c
@@ -720,29 +720,32 @@ static int be_set_phys_id(struct net_device *netdev,
 			  enum ethtool_phys_id_state state)
 {
 	struct be_adapter *adapter = netdev_priv(netdev);
+	int status = 0;
 
 	switch (state) {
 	case ETHTOOL_ID_ACTIVE:
-		be_cmd_get_beacon_state(adapter, adapter->hba_port_num,
-					&adapter->beacon_state);
-		return 1;	/* cycle on/off once per second */
+		status = be_cmd_get_beacon_state(adapter, adapter->hba_port_num,
+						 &adapter->beacon_state);
+		if (status)
+			return be_cmd_status(status);
+		return 1;       /* cycle on/off once per second */
 
 	case ETHTOOL_ID_ON:
-		be_cmd_set_beacon_state(adapter, adapter->hba_port_num, 0, 0,
-					BEACON_STATE_ENABLED);
+		status = be_cmd_set_beacon_state(adapter, adapter->hba_port_num,
+						 0, 0, BEACON_STATE_ENABLED);
 		break;
 
 	case ETHTOOL_ID_OFF:
-		be_cmd_set_beacon_state(adapter, adapter->hba_port_num, 0, 0,
-					BEACON_STATE_DISABLED);
+		status = be_cmd_set_beacon_state(adapter, adapter->hba_port_num,
+						 0, 0, BEACON_STATE_DISABLED);
 		break;
 
 	case ETHTOOL_ID_INACTIVE:
-		be_cmd_set_beacon_state(adapter, adapter->hba_port_num, 0, 0,
-					adapter->beacon_state);
+		status = be_cmd_set_beacon_state(adapter, adapter->hba_port_num,
+						 0, 0, adapter->beacon_state);
 	}
 
-	return 0;
+	return be_cmd_status(status);
 }
 
 static int be_set_dump(struct net_device *netdev, struct ethtool_dump *dump)
-- 
2.3.0.rc2

[PATCH net-next 2/8] be2net: check for INSUFFICIENT_PRIVILEGES error

[Sriharsha Basavapatna]

From: Suresh Reddy <suresh.reddy@broadcom.com>

The driver currently logs the message "VF is not privileged to issue
opcode" by checking only the base_status field for UNAUTHORIZED_REQUEST.
Add check to look for INSUFFICIENT_PRIVILEGES in the additional status
field also as not all cmds fail with that base status.

Signed-off-by: Suresh Reddy <suresh.reddy@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_cmds.c | 3 ++-
 drivers/net/ethernet/emulex/benet/be_cmds.h | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index b63d8ad..3b665f1 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -236,7 +236,8 @@ static int be_mcc_compl_process(struct be_adapter *adapter,
 
 	if (base_status != MCC_STATUS_SUCCESS &&
 	    !be_skip_err_log(opcode, base_status, addl_status)) {
-		if (base_status == MCC_STATUS_UNAUTHORIZED_REQUEST) {
+		if (base_status == MCC_STATUS_UNAUTHORIZED_REQUEST ||
+		    addl_status == MCC_ADDL_STATUS_INSUFFICIENT_PRIVILEGES) {
 			dev_warn(&adapter->pdev->dev,
 				 "VF is not privileged to issue opcode %d-%d\n",
 				 opcode, subsystem);
diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.h b/drivers/net/ethernet/emulex/benet/be_cmds.h
index 241819b..f260ef3 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.h
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.h
@@ -68,7 +68,8 @@ enum mcc_addl_status {
 	MCC_ADDL_STATUS_TOO_MANY_INTERFACES = 0x4a,
 	MCC_ADDL_STATUS_INSUFFICIENT_VLANS = 0xab,
 	MCC_ADDL_STATUS_INVALID_SIGNATURE = 0x56,
-	MCC_ADDL_STATUS_MISSING_SIGNATURE = 0x57
+	MCC_ADDL_STATUS_MISSING_SIGNATURE = 0x57,
+	MCC_ADDL_STATUS_INSUFFICIENT_PRIVILEGES = 0x60
 };
 
 #define CQE_BASE_STATUS_MASK		0xFFFF
-- 
2.3.0.rc2

[PATCH net-next 3/8] be2net: Fix be_vlan_rem_vid() to check vlan id being removed

[Sriharsha Basavapatna]

The driver decrements its vlan count without checking if it is really
present in its list. This results in an invalid vlan count and impacts
subsequent vlan add/rem ops. The function be_vlan_rem_vid() should be
updated to fix this.

Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_main.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index f99de36..09e6f2c 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -1463,6 +1463,9 @@ static int be_vlan_rem_vid(struct net_device *netdev, __be16 proto, u16 vid)
 	if (lancer_chip(adapter) && vid == 0)
 		return 0;
 
+	if (!test_bit(vid, adapter->vids))
+		return 0;
+
 	clear_bit(vid, adapter->vids);
 	adapter->vlans_added--;
 
-- 
2.3.0.rc2

[PATCH net-next 4/8] be2net: SRIOV Queue distribution should factor in EQ-count of VFs

[Sriharsha Basavapatna]

The SRIOV resource distribution logic for RX/TX queue counts is not optimal
when a small number of VFs are enabled. It does not take into account the
VF's EQ count while computing the queue counts. Because of this, the VF
gets a large number of queues, though it doesn't have sufficient EQs,
resulting in wasted queue resources. And the PF gets a smaller share of
queues though it has more EQs. Fix this by capping the VF queue count at
its EQ count.

Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be.h      |  4 ++++
 drivers/net/ethernet/emulex/benet/be_main.c | 15 ++++++---------
 2 files changed, 10 insertions(+), 9 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index cf83783..f33d73c 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -89,6 +89,10 @@
 #define BE3_MAX_TX_QS		16
 #define BE3_MAX_EVT_QS		16
 #define BE3_SRIOV_MAX_EVT_QS	8
+#define	SH_VF_MAX_NIC_EQS	3	/* Skyhawk VFs can have a max of 4 EQs
+					 * and at least 1 is granted to either
+					 * SURF/DPDK
+					 */
 
 #define MAX_RSS_IFACES		15
 #define MAX_RX_QS		32
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 09e6f2c..62f6fbb 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -3792,18 +3792,15 @@ static u16 be_calculate_vf_qs(struct be_adapter *adapter, u16 num_vfs)
 	struct be_resources res = adapter->pool_res;
 	u16 num_vf_qs = 1;
 
-	/* Distribute the queue resources equally among the PF and it's VFs
+	/* Distribute the queue resources among the PF and it's VFs
 	 * Do not distribute queue resources in multi-channel configuration.
 	 */
 	if (num_vfs && !be_is_mc(adapter)) {
-		/* If number of VFs requested is 8 less than max supported,
-		 * assign 8 queue pairs to the PF and divide the remaining
-		 * resources evenly among the VFs
-		 */
-		if (num_vfs < (be_max_vfs(adapter) - 8))
-			num_vf_qs = (res.max_rss_qs - 8) / num_vfs;
-		else
-			num_vf_qs = res.max_rss_qs / num_vfs;
+		 /* Divide the qpairs evenly among the VFs and the PF, capped
+		  * at VF-EQ-count. Any remainder qpairs belong to the PF.
+		  */
+		num_vf_qs = min(SH_VF_MAX_NIC_EQS,
+				res.max_rss_qs / (num_vfs + 1));
 
 		/* Skyhawk-R chip supports only MAX_RSS_IFACES RSS capable
 		 * interfaces per port. Provide RSS on VFs, only if number
-- 
2.3.0.rc2

Re: [PATCH net-next 4/8] be2net: SRIOV Queue distribution should factor in EQ-count of VFs

[Sergei Shtylyov]

Hello.

On 2/2/2016 1:20 PM, Sriharsha Basavapatna wrote:

> The SRIOV resource distribution logic for RX/TX queue counts is not optimal
> when a small number of VFs are enabled. It does not take into account the
> VF's EQ count while computing the queue counts. Because of this, the VF
> gets a large number of queues, though it doesn't have sufficient EQs,
> resulting in wasted queue resources. And the PF gets a smaller share of
> queues though it has more EQs. Fix this by capping the VF queue count at
> its EQ count.
>
> Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
> ---
>   drivers/net/ethernet/emulex/benet/be.h      |  4 ++++
>   drivers/net/ethernet/emulex/benet/be_main.c | 15 ++++++---------
>   2 files changed, 10 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
> index cf83783..f33d73c 100644
> --- a/drivers/net/ethernet/emulex/benet/be.h
> +++ b/drivers/net/ethernet/emulex/benet/be.h
> @@ -89,6 +89,10 @@
>   #define BE3_MAX_TX_QS		16
>   #define BE3_MAX_EVT_QS		16
>   #define BE3_SRIOV_MAX_EVT_QS	8
> +#define	SH_VF_MAX_NIC_EQS	3	/* Skyhawk VFs can have a max of 4 EQs

    Other #define's seem to have a space after them, not tab...

> +					 * and at least 1 is granted to either
> +					 * SURF/DPDK
> +					 */
>
>   #define MAX_RSS_IFACES		15
>   #define MAX_RX_QS		32
[...]

MBR, Sergei

Re: [PATCH net-next 4/8] be2net: SRIOV Queue distribution should factor in EQ-count of VFs

[Sriharsha Basavapatna]

On Tue, Feb 2, 2016 at 8:00 PM, Sergei Shtylyov
<sergei.shtylyov@cogentembedded.com> wrote:
> Hello.
>
> On 2/2/2016 1:20 PM, Sriharsha Basavapatna wrote:
>
>> The SRIOV resource distribution logic for RX/TX queue counts is not
>> optimal
>> when a small number of VFs are enabled. It does not take into account the
>> VF's EQ count while computing the queue counts. Because of this, the VF
>> gets a large number of queues, though it doesn't have sufficient EQs,
>> resulting in wasted queue resources. And the PF gets a smaller share of
>> queues though it has more EQs. Fix this by capping the VF queue count at
>> its EQ count.
>>
>> Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
>> ---
>>   drivers/net/ethernet/emulex/benet/be.h      |  4 ++++
>>   drivers/net/ethernet/emulex/benet/be_main.c | 15 ++++++---------
>>   2 files changed, 10 insertions(+), 9 deletions(-)
>>
>> diff --git a/drivers/net/ethernet/emulex/benet/be.h
>> b/drivers/net/ethernet/emulex/benet/be.h
>> index cf83783..f33d73c 100644
>> --- a/drivers/net/ethernet/emulex/benet/be.h
>> +++ b/drivers/net/ethernet/emulex/benet/be.h
>> @@ -89,6 +89,10 @@
>>   #define BE3_MAX_TX_QS         16
>>   #define BE3_MAX_EVT_QS                16
>>   #define BE3_SRIOV_MAX_EVT_QS  8
>> +#define        SH_VF_MAX_NIC_EQS       3       /* Skyhawk VFs can have a
>> max of 4 EQs
>
>
>    Other #define's seem to have a space after them, not tab...

Thanks for catching it,  I'll fix it in a subsequent patch ?

Thanks,
-Harsha
>
>> +                                        * and at least 1 is granted to
>> either
>> +                                        * SURF/DPDK
>> +                                        */
>>
>>   #define MAX_RSS_IFACES                15
>>   #define MAX_RX_QS             32
>
> [...]
>
> MBR, Sergei
>

Re: [PATCH net-next 4/8] be2net: SRIOV Queue distribution should factor in EQ-count of VFs

[Sergei Shtylyov]

On 02/02/2016 06:41 PM, Sriharsha Basavapatna wrote:

>>> The SRIOV resource distribution logic for RX/TX queue counts is not
>>> optimal
>>> when a small number of VFs are enabled. It does not take into account the
>>> VF's EQ count while computing the queue counts. Because of this, the VF
>>> gets a large number of queues, though it doesn't have sufficient EQs,
>>> resulting in wasted queue resources. And the PF gets a smaller share of
>>> queues though it has more EQs. Fix this by capping the VF queue count at
>>> its EQ count.
>>>
>>> Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
>>> ---
>>>    drivers/net/ethernet/emulex/benet/be.h      |  4 ++++
>>>    drivers/net/ethernet/emulex/benet/be_main.c | 15 ++++++---------
>>>    2 files changed, 10 insertions(+), 9 deletions(-)
>>>
>>> diff --git a/drivers/net/ethernet/emulex/benet/be.h
>>> b/drivers/net/ethernet/emulex/benet/be.h
>>> index cf83783..f33d73c 100644
>>> --- a/drivers/net/ethernet/emulex/benet/be.h
>>> +++ b/drivers/net/ethernet/emulex/benet/be.h
>>> @@ -89,6 +89,10 @@
>>>    #define BE3_MAX_TX_QS         16
>>>    #define BE3_MAX_EVT_QS                16
>>>    #define BE3_SRIOV_MAX_EVT_QS  8
>>> +#define        SH_VF_MAX_NIC_EQS       3       /* Skyhawk VFs can have a
>>> max of 4 EQs
>>
>>
>>     Other #define's seem to have a space after them, not tab...
>
> Thanks for catching it,  I'll fix it in a subsequent patch ?

    Please just don't use tab here, no need for new patch.

> Thanks,
> -Harsha

[...]

MBR, Sergei

[PATCH net-next 5/8] be2net: Don't run ethtool self-tests for VFs

[Sriharsha Basavapatna]

From: Somnath Kotur <somnath.kotur@broadcom.com>

The CMD_SUBSYSTEM_LOWLEVEL cmds need DEV_CFG Privilege to run
which VFs don't have by default.
Self-tests need to be issued only for PFs.

Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_cmds.c | 29 ++++++++++++++++++++++++++++-
 1 file changed, 28 insertions(+), 1 deletion(-)

diff --git a/drivers/net/ethernet/emulex/benet/be_cmds.c b/drivers/net/ethernet/emulex/benet/be_cmds.c
index 3b665f1..7d51d47 100644
--- a/drivers/net/ethernet/emulex/benet/be_cmds.c
+++ b/drivers/net/ethernet/emulex/benet/be_cmds.c
@@ -65,7 +65,22 @@ static struct be_cmd_priv_map cmd_priv_map[] = {
 		CMD_SUBSYSTEM_COMMON,
 		BE_PRIV_LNKMGMT | BE_PRIV_VHADM |
 		BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
-	}
+	},
+	{
+		OPCODE_LOWLEVEL_HOST_DDR_DMA,
+		CMD_SUBSYSTEM_LOWLEVEL,
+		BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+	},
+	{
+		OPCODE_LOWLEVEL_LOOPBACK_TEST,
+		CMD_SUBSYSTEM_LOWLEVEL,
+		BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+	},
+	{
+		OPCODE_LOWLEVEL_SET_LOOPBACK_MODE,
+		CMD_SUBSYSTEM_LOWLEVEL,
+		BE_PRIV_DEVCFG | BE_PRIV_DEVSEC
+	},
 };
 
 static bool be_cmd_allowed(struct be_adapter *adapter, u8 opcode, u8 subsystem)
@@ -3169,6 +3184,10 @@ int be_cmd_set_loopback(struct be_adapter *adapter, u8 port_num,
 	struct be_cmd_req_set_lmode *req;
 	int status;
 
+	if (!be_cmd_allowed(adapter, OPCODE_LOWLEVEL_SET_LOOPBACK_MODE,
+			    CMD_SUBSYSTEM_LOWLEVEL))
+		return -EPERM;
+
 	spin_lock_bh(&adapter->mcc_lock);
 
 	wrb = wrb_from_mccq(adapter);
@@ -3214,6 +3233,10 @@ int be_cmd_loopback_test(struct be_adapter *adapter, u32 port_num,
 	struct be_cmd_resp_loopback_test *resp;
 	int status;
 
+	if (!be_cmd_allowed(adapter, OPCODE_LOWLEVEL_LOOPBACK_TEST,
+			    CMD_SUBSYSTEM_LOWLEVEL))
+		return -EPERM;
+
 	spin_lock_bh(&adapter->mcc_lock);
 
 	wrb = wrb_from_mccq(adapter);
@@ -3260,6 +3283,10 @@ int be_cmd_ddr_dma_test(struct be_adapter *adapter, u64 pattern,
 	int status;
 	int i, j = 0;
 
+	if (!be_cmd_allowed(adapter, OPCODE_LOWLEVEL_HOST_DDR_DMA,
+			    CMD_SUBSYSTEM_LOWLEVEL))
+		return -EPERM;
+
 	spin_lock_bh(&adapter->mcc_lock);
 
 	wrb = wrb_from_mccq(adapter);
-- 
2.3.0.rc2

[PATCH net-next 6/8] benet: Fix Lancer error recovery

[Sriharsha Basavapatna]

From: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>

After error is detected, wait for adapter to move to ready state
before destroying queues and cleanup of other resources. Also
skip performing any cleanup for non-Lancer chips and move debug
messages to correct routine.

Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_main.c | 51 +++++++++++++++++++----------
 1 file changed, 34 insertions(+), 17 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 62f6fbb..6eb3aba 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -4859,21 +4859,27 @@ static int be_resume(struct be_adapter *adapter)
 
 static int be_err_recover(struct be_adapter *adapter)
 {
-	struct device *dev = &adapter->pdev->dev;
 	int status;
 
+	/* Error recovery is supported only Lancer as of now */
+	if (!lancer_chip(adapter))
+		return -EIO;
+
+	/* Wait for adapter to reach quiescent state before
+	 * destroying queues
+	 */
+	status = be_fw_wait_ready(adapter);
+	if (status)
+		goto err;
+
+	be_cleanup(adapter);
+
 	status = be_resume(adapter);
 	if (status)
 		goto err;
 
-	dev_info(dev, "Adapter recovery successful\n");
 	return 0;
 err:
-	if (be_physfn(adapter))
-		dev_err(dev, "Adapter recovery failed\n");
-	else
-		dev_err(dev, "Re-trying adapter recovery\n");
-
 	return status;
 }
 
@@ -4882,21 +4888,32 @@ static void be_err_detection_task(struct work_struct *work)
 	struct be_adapter *adapter =
 				container_of(work, struct be_adapter,
 					     be_err_detection_work.work);
-	int status = 0;
+	struct device *dev = &adapter->pdev->dev;
+	int recovery_status;
 
 	be_detect_error(adapter);
 
-	if (be_check_error(adapter, BE_ERROR_HW)) {
-		be_cleanup(adapter);
-
-		/* As of now error recovery support is in Lancer only */
-		if (lancer_chip(adapter))
-			status = be_err_recover(adapter);
+	if (be_check_error(adapter, BE_ERROR_HW))
+		recovery_status = be_err_recover(adapter);
+	else
+		goto reschedule_task;
+
+	if (!recovery_status) {
+		dev_info(dev, "Adapter recovery successful\n");
+		goto reschedule_task;
+	} else if (be_virtfn(adapter)) {
+		/* For VFs, check if PF have allocated resources
+		 * every second.
+		 */
+		dev_err(dev, "Re-trying adapter recovery\n");
+		goto reschedule_task;
+	} else {
+		dev_err(dev, "Adapter recovery failed\n");
 	}
 
-	/* Always attempt recovery on VFs */
-	if (!status || be_virtfn(adapter))
-		be_schedule_err_detection(adapter);
+	return;
+reschedule_task:
+	be_schedule_err_detection(adapter);
 }
 
 static void be_log_sfp_info(struct be_adapter *adapter)
-- 
2.3.0.rc2

[PATCH net-next 7/8] benet: Add retry in case of error recovery failure

[Sriharsha Basavapatna]

From: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>

Retry error recovery MAX_ERR_RECOVERY_RETRY_COUNT times in case of
failure during error recovery.

Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be.h      |  5 +++++
 drivers/net/ethernet/emulex/benet/be_main.c | 23 +++++++++++++++++------
 2 files changed, 22 insertions(+), 6 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index f33d73c..5277974 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -397,6 +397,10 @@ enum vf_state {
 #define BE_UC_PMAC_COUNT			30
 #define BE_VF_UC_PMAC_COUNT			2
 
+#define MAX_ERR_RECOVERY_RETRY_COUNT		3
+#define ERR_DETECTION_DELAY			1000
+#define ERR_RECOVERY_RETRY_DELAY		30000
+
 /* Ethtool set_dump flags */
 #define LANCER_INITIATE_FW_DUMP			0x1
 #define LANCER_DELETE_FW_DUMP			0x2
@@ -534,6 +538,7 @@ struct be_adapter {
 	u16 work_counter;
 
 	struct delayed_work be_err_detection_work;
+	u8 recovery_retries;
 	u8 err_flags;
 	u32 flags;
 	u32 cmd_privileges;
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 6eb3aba..d5286d3 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -4265,10 +4265,10 @@ static void be_schedule_worker(struct be_adapter *adapter)
 	adapter->flags |= BE_FLAGS_WORKER_SCHEDULED;
 }
 
-static void be_schedule_err_detection(struct be_adapter *adapter)
+static void be_schedule_err_detection(struct be_adapter *adapter, u32 delay)
 {
 	schedule_delayed_work(&adapter->be_err_detection_work,
-			      msecs_to_jiffies(1000));
+			      msecs_to_jiffies(delay));
 	adapter->flags |= BE_FLAGS_ERR_DETECTION_SCHEDULED;
 }
 
@@ -4890,6 +4890,7 @@ static void be_err_detection_task(struct work_struct *work)
 					     be_err_detection_work.work);
 	struct device *dev = &adapter->pdev->dev;
 	int recovery_status;
+	int delay = ERR_DETECTION_DELAY;
 
 	be_detect_error(adapter);
 
@@ -4899,6 +4900,7 @@ static void be_err_detection_task(struct work_struct *work)
 		goto reschedule_task;
 
 	if (!recovery_status) {
+		adapter->recovery_retries = 0;
 		dev_info(dev, "Adapter recovery successful\n");
 		goto reschedule_task;
 	} else if (be_virtfn(adapter)) {
@@ -4907,13 +4909,22 @@ static void be_err_detection_task(struct work_struct *work)
 		 */
 		dev_err(dev, "Re-trying adapter recovery\n");
 		goto reschedule_task;
+	} else if (adapter->recovery_retries++ <
+		   MAX_ERR_RECOVERY_RETRY_COUNT) {
+		/* In case of another error during recovery, it takes 30 sec
+		 * for adapter to come out of error. Retry error recovery after
+		 * this time interval.
+		 */
+		dev_err(&adapter->pdev->dev, "Re-trying adapter recovery\n");
+		delay = ERR_RECOVERY_RETRY_DELAY;
+		goto reschedule_task;
 	} else {
 		dev_err(dev, "Adapter recovery failed\n");
 	}
 
 	return;
 reschedule_task:
-	be_schedule_err_detection(adapter);
+	be_schedule_err_detection(adapter, delay);
 }
 
 static void be_log_sfp_info(struct be_adapter *adapter)
@@ -5309,7 +5320,7 @@ static int be_probe(struct pci_dev *pdev, const struct pci_device_id *pdev_id)
 
 	be_roce_dev_add(adapter);
 
-	be_schedule_err_detection(adapter);
+	be_schedule_err_detection(adapter, ERR_DETECTION_DELAY);
 
 	/* On Die temperature not supported for VF. */
 	if (be_physfn(adapter) && IS_ENABLED(CONFIG_BE2NET_HWMON)) {
@@ -5376,7 +5387,7 @@ static int be_pci_resume(struct pci_dev *pdev)
 	if (status)
 		return status;
 
-	be_schedule_err_detection(adapter);
+	be_schedule_err_detection(adapter, ERR_DETECTION_DELAY);
 
 	if (adapter->wol_en)
 		be_setup_wol(adapter, false);
@@ -5476,7 +5487,7 @@ static void be_eeh_resume(struct pci_dev *pdev)
 	if (status)
 		goto err;
 
-	be_schedule_err_detection(adapter);
+	be_schedule_err_detection(adapter, ERR_DETECTION_DELAY);
 	return;
 err:
 	dev_err(&adapter->pdev->dev, "EEH resume failed\n");
-- 
2.3.0.rc2

[PATCH net-next 8/8] benet: Fix interval calculation in interrupt moderation

[Sriharsha Basavapatna]

From: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>

Interrupt moderation parameters need to be recalculated only
after a time interval of 1 ms. Interval calculation is wrong
when there is a rollover of jiffies. Using recommended way of interval
calculation using jiffies to fix this.

Signed-off-by: Padmanabh Ratnakar <padmanabh.ratnakar@broadcom.com>
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@broadcom.com>
---
 drivers/net/ethernet/emulex/benet/be_main.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index d5286d3..9c1fc9d 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -1917,8 +1917,7 @@ static u32 be_get_eq_delay_mult_enc(struct be_eq_obj *eqo)
 	if (!aic->enable)
 		return 0;
 
-	if (time_before_eq(now, aic->jiffies) ||
-	    jiffies_to_msecs(now - aic->jiffies) < 1)
+	if (jiffies_to_msecs(now - aic->jiffies) < 1)
 		eqd = aic->prev_eqd;
 	else
 		eqd = be_get_new_eqd(eqo);
-- 
2.3.0.rc2