[dpdk-stable] patch 'net/octeontx: fix stop clearing Rx/Tx functions' has been queued to LTS release 17.11.4

[dpdk-stable] patch 'net/octeontx: fix stop clearing Rx/Tx functions' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From f6c78aecd8f42f951eaba2263d24b20511202f76 Mon Sep 17 00:00:00 2001
From: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Date: Tue, 24 Jul 2018 16:13:50 +0530
Subject: [PATCH] net/octeontx: fix stop clearing Rx/Tx functions

[ upstream commit efb9dd148e234aecd01966cd1d2aaa4b75836966 ]

On dev_stop the Rx/Tx_burst functions are being set to NULL, this causes
a SEGFAULT in cases where control path calls stop and a paket is still
being processed by a worker.
Instead clear the fastpath functions in dev_close.

Fixes: da6c687471a3 ("net/octeontx: add start and stop support")

Signed-off-by: Pavan Nikhilesh <pbhagavatula@caviumnetworks.com>
Acked-by: Jerin Jacob <jerin.jacob@caviumnetworks.com>
---
 drivers/net/octeontx/octeontx_ethdev.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/net/octeontx/octeontx_ethdev.c b/drivers/net/octeontx/octeontx_ethdev.c
index 33c6e78e6..13dfbbc04 100644
--- a/drivers/net/octeontx/octeontx_ethdev.c
+++ b/drivers/net/octeontx/octeontx_ethdev.c
@@ -377,6 +377,9 @@ octeontx_dev_close(struct rte_eth_dev *dev)
 
 		rte_free(txq);
 	}
+
+	dev->tx_pkt_burst = NULL;
+	dev->rx_pkt_burst = NULL;
 }
 
 static int
@@ -470,9 +473,6 @@ octeontx_dev_stop(struct rte_eth_dev *dev)
 			     ret);
 		return;
 	}
-
-	dev->tx_pkt_burst = NULL;
-	dev->rx_pkt_burst = NULL;
 }
 
 static void
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: fix filter freeing' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From b0468ac19bd9b1c18be39f66621b2d058d6687c1 Mon Sep 17 00:00:00 2001
From: Somnath Kotur <somnath.kotur@broadcom.com>
Date: Wed, 25 Jul 2018 18:15:44 -0700
Subject: [PATCH] net/bnxt: fix filter freeing

[ upstream commit eae0a36249acb365c9c5952b0ed7f1d4b625ab42 ]

bnxt_clear_hwrm_vnic_filters() was removing the created filter from HW,
but not clearing the internal data structures by removing it from the
struct bnxt_vnic_info->filter list.

Fixes: 5ef3b79fdfe6 ("net/bnxt: support flow filter ops")

Signed-off-by: Somnath Kotur <somnath.kotur@broadcom.com>
Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index f94c27bd7..9aa39e190 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -1804,6 +1804,7 @@ int bnxt_clear_hwrm_vnic_filters(struct bnxt *bp, struct bnxt_vnic_info *vnic)
 			rc = bnxt_hwrm_clear_ntuple_filter(bp, filter);
 		else
 			rc = bnxt_hwrm_clear_l2_filter(bp, filter);
+		STAILQ_REMOVE(&vnic->filter, filter, bnxt_filter_info, next);
 		//if (rc)
 			//break;
 	}
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: fix memory leaks in NVM commands' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 6421d9662b12d3a9226b33a9ea3cf8f762771618 Mon Sep 17 00:00:00 2001
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Date: Wed, 25 Jul 2018 18:15:45 -0700
Subject: [PATCH] net/bnxt: fix memory leaks in NVM commands

[ upstream commit 6621ae146166adaba8dc4c120ad8c0fabab95830 ]

In some cases we may not be freeing up memory allocated for certain
NVM commands because the code might have bailed out before reaching
rte_free(). This patch moves some code around to ensure the allocated
memory is freed before exiting the function.

Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM")

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Somnath Kotur <somnath.kotur@broadcom.com>
Reviewed-by: Ray Jui <ray.jui@broadcom.com>
Reviewed-by: Michael Wildt <michael.wildt@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index 9aa39e190..ce66dc53e 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -3106,13 +3106,12 @@ int bnxt_get_nvram_directory(struct bnxt *bp, uint32_t len, uint8_t *data)
 	req.host_dest_addr = rte_cpu_to_le_64(dma_handle);
 	rc = bnxt_hwrm_send_message(bp, &req, sizeof(req));
 
-	HWRM_CHECK_RESULT();
-	HWRM_UNLOCK();
-
 	if (rc == 0)
 		memcpy(data, buf, len > buflen ? buflen : len);
 
 	rte_free(buf);
+	HWRM_CHECK_RESULT();
+	HWRM_UNLOCK();
 
 	return rc;
 }
@@ -3144,12 +3143,13 @@ int bnxt_hwrm_get_nvram_item(struct bnxt *bp, uint32_t index,
 	req.offset = rte_cpu_to_le_32(offset);
 	req.len = rte_cpu_to_le_32(length);
 	rc = bnxt_hwrm_send_message(bp, &req, sizeof(req));
-	HWRM_CHECK_RESULT();
-	HWRM_UNLOCK();
 	if (rc == 0)
 		memcpy(data, buf, length);
 
 	rte_free(buf);
+	HWRM_CHECK_RESULT();
+	HWRM_UNLOCK();
+
 	return rc;
 }
 
@@ -3204,10 +3204,10 @@ int bnxt_hwrm_flash_nvram(struct bnxt *bp, uint16_t dir_type,
 
 	rc = bnxt_hwrm_send_message(bp, &req, sizeof(req));
 
+	rte_free(buf);
 	HWRM_CHECK_RESULT();
 	HWRM_UNLOCK();
 
-	rte_free(buf);
 	return rc;
 }
 
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: fix lock release on NVM write failure' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 063f7f0b46368f58da1a2d6e08dddfb999dfb393 Mon Sep 17 00:00:00 2001
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Date: Wed, 25 Jul 2018 18:15:46 -0700
Subject: [PATCH] net/bnxt: fix lock release on NVM write failure

[ upstream commit 4623c0d4b572e9bea28d59e6e54ae6d5dab2412e ]

In bnxt_hwrm_flash_nvram, before attempting to allocate a buffer
we are grabbing the rte_spinlock. And if the allocation fails we
are returning before releasing the spinlock. We avoid the situation
by calling HWRM_PREP which grabs the lock after the buffer is
allocated successfully.

Fixes: 19e6af01bb36 ("net/bnxt: support get/set EEPROM")

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index ce66dc53e..fa8f1faca 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -3180,14 +3180,6 @@ int bnxt_hwrm_flash_nvram(struct bnxt *bp, uint16_t dir_type,
 	rte_iova_t dma_handle;
 	uint8_t *buf;
 
-	HWRM_PREP(req, NVM_WRITE);
-
-	req.dir_type = rte_cpu_to_le_16(dir_type);
-	req.dir_ordinal = rte_cpu_to_le_16(dir_ordinal);
-	req.dir_ext = rte_cpu_to_le_16(dir_ext);
-	req.dir_attr = rte_cpu_to_le_16(dir_attr);
-	req.dir_data_length = rte_cpu_to_le_32(data_len);
-
 	buf = rte_malloc("nvm_write", data_len, 0);
 	rte_mem_lock_page(buf);
 	if (!buf)
@@ -3200,6 +3192,14 @@ int bnxt_hwrm_flash_nvram(struct bnxt *bp, uint16_t dir_type,
 		return -ENOMEM;
 	}
 	memcpy(buf, data, data_len);
+
+	HWRM_PREP(req, NVM_WRITE);
+
+	req.dir_type = rte_cpu_to_le_16(dir_type);
+	req.dir_ordinal = rte_cpu_to_le_16(dir_ordinal);
+	req.dir_ext = rte_cpu_to_le_16(dir_ext);
+	req.dir_attr = rte_cpu_to_le_16(dir_attr);
+	req.dir_data_length = rte_cpu_to_le_32(data_len);
 	req.host_src_addr = rte_cpu_to_le_64(dma_handle);
 
 	rc = bnxt_hwrm_send_message(bp, &req, sizeof(req));
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: check access denied for HWRM commands' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From e6af6f16d45dca27115526bd6201a6c658bf6d2c Mon Sep 17 00:00:00 2001
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Date: Wed, 25 Jul 2018 18:15:47 -0700
Subject: [PATCH] net/bnxt: check access denied for HWRM commands

[ upstream commit 6dc875ab5ff76377376906aa6d05d95ffd34b9f9 ]

Firmware can restrict access to certain resources in the hardware
depending on various factors. In such cases when the PMD tries to
configure these resources the firmware will return
HWRM_ERR_CODE_RESOURCE_ACCESS_DENIED error. Parse this and return
the standard EACCES error to the applications.

Fixes: 804e746c7b73 ("net/bnxt: add hardware resource manager init code")

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
---
 drivers/net/bnxt/bnxt_hwrm.c | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/drivers/net/bnxt/bnxt_hwrm.c b/drivers/net/bnxt/bnxt_hwrm.c
index fa8f1faca..db3222f4e 100644
--- a/drivers/net/bnxt/bnxt_hwrm.c
+++ b/drivers/net/bnxt/bnxt_hwrm.c
@@ -197,6 +197,10 @@ err_ret:
 		RTE_LOG(ERR, PMD, "%s failed rc:%d\n", \
 			__func__, rc); \
 		rte_spinlock_unlock(&bp->hwrm_lock); \
+		if (rc == HWRM_ERR_CODE_RESOURCE_ACCESS_DENIED) \
+			rc = -EACCES; \
+		else if (rc > 0) \
+			rc = -EINVAL; \
 		return rc; \
 	} \
 	if (resp->error_code) { \
@@ -218,6 +222,10 @@ err_ret:
 				"%s error %d\n", __func__, rc); \
 		} \
 		rte_spinlock_unlock(&bp->hwrm_lock); \
+		if (rc == HWRM_ERR_CODE_RESOURCE_ACCESS_DENIED) \
+			rc = -EACCES; \
+		else if (rc > 0) \
+			rc = -EINVAL; \
 		return rc; \
 	} \
 } while (0)
-- 
2.11.0

[dpdk-stable] patch 'net/bnxt: fix RETA size' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 2658600aca85843080ca6ae6f0afb4f08e065c50 Mon Sep 17 00:00:00 2001
From: Ajit Khaparde <ajit.khaparde@broadcom.com>
Date: Wed, 25 Jul 2018 18:15:48 -0700
Subject: [PATCH] net/bnxt: fix RETA size

[ upstream commit 7f675c27d9c80bc6ff3793140fa3215671b295d6 ]

The reta_size being indicated in the bnxt_dev_info_get_op was incorrect.
Set it to the value supported by the hardware.

Fixes: 0a6d2a720078 ("net/bnxt: get device infos")

Signed-off-by: Ajit Khaparde <ajit.khaparde@broadcom.com>
Reviewed-by: Scott Branden <scott.branden@broadcom.com>
Reviewed-by: Randy Schacher <stuart.schacher@broadcom.com>
Tested-by: Randy Schacher <stuart.schacher@broadcom.com>
---
 drivers/net/bnxt/bnxt_ethdev.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/net/bnxt/bnxt_ethdev.c b/drivers/net/bnxt/bnxt_ethdev.c
index 5b0bf80e3..7466a642e 100644
--- a/drivers/net/bnxt/bnxt_ethdev.c
+++ b/drivers/net/bnxt/bnxt_ethdev.c
@@ -440,7 +440,7 @@ static void bnxt_dev_info_get_op(struct rte_eth_dev *eth_dev,
 	/* For the sake of symmetry, max_rx_queues = max_tx_queues */
 	dev_info->max_rx_queues = max_rx_rings;
 	dev_info->max_tx_queues = max_rx_rings;
-	dev_info->reta_size = bp->max_rsscos_ctx;
+	dev_info->reta_size = HW_HASH_INDEX_SIZE;
 	dev_info->hash_key_size = 40;
 	max_vnics = bp->max_vnics;
 
-- 
2.11.0

[dpdk-stable] patch 'net/qede: fix interrupt handler unregister' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 56cadff31267ebc8a05b38c3227a0c06ef967f4b Mon Sep 17 00:00:00 2001
From: Shahed Shaikh <shahed.shaikh@cavium.com>
Date: Tue, 24 Jul 2018 15:18:41 -0700
Subject: [PATCH] net/qede: fix interrupt handler unregister

[ upstream commit 4eee1bbf25859b72723ec570d77bf5fb501bedcf ]

Commit 023d7a0449f11 ("net/qede: fix legacy interrupt mode")
added a handler for legacy interrupt mode but forgot to
unregister same handler in rte_intr_callback_unregister()

Fixes: 245aec289338 ("net/qede: fix legacy interrupt mode")

Signed-off-by: Shahed Shaikh <shahed.shaikh@cavium.com>
---
 drivers/net/qede/qede_ethdev.c | 16 ++++++++++++++--
 1 file changed, 14 insertions(+), 2 deletions(-)

diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index 89f6ce05a..3be45d3ee 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -1609,8 +1609,20 @@ static void qede_dev_close(struct rte_eth_dev *eth_dev)
 	qdev->ops->common->slowpath_stop(edev);
 	qdev->ops->common->remove(edev);
 	rte_intr_disable(&pci_dev->intr_handle);
-	rte_intr_callback_unregister(&pci_dev->intr_handle,
-				     qede_interrupt_handler, (void *)eth_dev);
+
+	switch (pci_dev->intr_handle.type) {
+	case RTE_INTR_HANDLE_UIO_INTX:
+	case RTE_INTR_HANDLE_VFIO_LEGACY:
+		rte_intr_callback_unregister(&pci_dev->intr_handle,
+					     qede_interrupt_handler_intx,
+					     (void *)eth_dev);
+		break;
+	default:
+		rte_intr_callback_unregister(&pci_dev->intr_handle,
+					   qede_interrupt_handler,
+					   (void *)eth_dev);
+	}
+
 	if (ECORE_IS_CMT(edev))
 		rte_eal_alarm_cancel(qede_poll_sp_sb_cb, (void *)eth_dev);
 }
-- 
2.11.0

[dpdk-stable] patch 'net/qede/base: fix GRC attention callback' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From f4dd0374363e29506d9b17fbd5dd680b19bbf1c6 Mon Sep 17 00:00:00 2001
From: Rasesh Mody <rasesh.mody@cavium.com>
Date: Tue, 24 Jul 2018 15:18:43 -0700
Subject: [PATCH] net/qede/base: fix GRC attention callback

[ upstream commit 538bcb378af5d1750a6831efeea4d9565a3267f0 ]

Treat any attention which is not for timeout event as invalid
and return status accordingly. The HW error handler logs and clears
the HW attention. Without this fix we can see flood of GRC attentions.

Fixes: e6051bd6b07d ("qede: add interrupt handling support")

Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
---
 drivers/net/qede/base/ecore_int.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/net/qede/base/ecore_int.c b/drivers/net/qede/base/ecore_int.c
index e6cef85bc..61e36a430 100644
--- a/drivers/net/qede/base/ecore_int.c
+++ b/drivers/net/qede/base/ecore_int.c
@@ -231,15 +231,19 @@ static const char *grc_timeout_attn_master_to_str(u8 master)
 
 static enum _ecore_status_t ecore_grc_attn_cb(struct ecore_hwfn *p_hwfn)
 {
+	enum _ecore_status_t rc = ECORE_SUCCESS;
 	u32 tmp, tmp2;
 
 	/* We've already cleared the timeout interrupt register, so we learn
-	 * of interrupts via the validity register
+	 * of interrupts via the validity register.
+	 * Any attention which is not for a timeout event is treated as fatal.
 	 */
 	tmp = ecore_rd(p_hwfn, p_hwfn->p_dpc_ptt,
 		       GRC_REG_TIMEOUT_ATTN_ACCESS_VALID);
-	if (!(tmp & ECORE_GRC_ATTENTION_VALID_BIT))
+	if (!(tmp & ECORE_GRC_ATTENTION_VALID_BIT)) {
+		rc = ECORE_INVAL;
 		goto out;
+	}
 
 	/* Read the GRC timeout information */
 	tmp = ecore_rd(p_hwfn, p_hwfn->p_dpc_ptt,
@@ -263,11 +267,11 @@ static enum _ecore_status_t ecore_grc_attn_cb(struct ecore_hwfn *p_hwfn)
 		  (tmp2 & ECORE_GRC_ATTENTION_VF_MASK) >>
 		  ECORE_GRC_ATTENTION_VF_SHIFT);
 
-out:
-	/* Regardles of anything else, clean the validity bit */
+	/* Clean the validity bit */
 	ecore_wr(p_hwfn, p_hwfn->p_dpc_ptt,
 		 GRC_REG_TIMEOUT_ATTN_ACCESS_VALID, 0);
-	return ECORE_SUCCESS;
+out:
+	return rc;
 }
 
 #define ECORE_PGLUE_ATTENTION_VALID (1 << 29)
-- 
2.11.0

[dpdk-stable] patch 'net/qede: fix MAC address removal failure message' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 9ff59831711b1c24ef25fa7e3602490d7da57d95 Mon Sep 17 00:00:00 2001
From: Shahed Shaikh <shahed.shaikh@cavium.com>
Date: Tue, 24 Jul 2018 15:18:44 -0700
Subject: [PATCH] net/qede: fix MAC address removal failure message

[ upstream commit 7e3060a3704e21ab63e9f1b4338be2fed9a345d1 ]

Don't treat MAC address removal failure as a fatal error
and print in logs.

Fixes: 77fac1b54fc9 ("net/qede: fix filtering code")

Signed-off-by: Shahed Shaikh <shahed.shaikh@cavium.com>
---
 drivers/net/qede/qede_ethdev.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/net/qede/qede_ethdev.c b/drivers/net/qede/qede_ethdev.c
index 3be45d3ee..4a5e48579 100644
--- a/drivers/net/qede/qede_ethdev.c
+++ b/drivers/net/qede/qede_ethdev.c
@@ -845,7 +845,10 @@ qede_mac_int_ops(struct rte_eth_dev *eth_dev, struct ecore_filter_ucast *ucast,
 	if (rc == 0)
 		rc = ecore_filter_ucast_cmd(edev, ucast,
 					    ECORE_SPQ_MODE_CB, NULL);
-	if (rc != ECORE_SUCCESS)
+	/* Indicate error only for add filter operation.
+	 * Delete filter operations are not severe.
+	 */
+	if ((rc != ECORE_SUCCESS) && add)
 		DP_ERR(edev, "MAC filter failed, rc = %d, op = %d\n",
 		       rc, add);
 
-- 
2.11.0

[dpdk-stable] patch 'net/bnx2x: fix FW command timeout during stop' has been queued to LTS release 17.11.4

[Yongseok Koh]

Hi,

FYI, your patch has been queued to LTS release 17.11.4

Note it hasn't been pushed to http://dpdk.org/browse/dpdk-stable yet.
It will be pushed if I get no objections before 08/15/18. So please
shout if anyone has objections.

Thanks.

Yongseok

---
>From 669b749935fad556b65f3a274183d88d42a08c55 Mon Sep 17 00:00:00 2001
From: Rasesh Mody <rasesh.mody@cavium.com>
Date: Wed, 1 Aug 2018 11:19:20 -0700
Subject: [PATCH] net/bnx2x: fix FW command timeout during stop

[ upstream commit 4a4607bb9cadcc49a062a46588dbdcc873dfe27a ]

This patch fixes firmware command timeout error seen during device stop
while stopping queues. It patially reverts an earlier preventive change
commit 91b7e432bcef ("net/bnx2x: disable fast path interrupts") to now
enable fast path interrupts.

The original issue of performance degradation is not observed anymore,
with or without the fix.

Fixes: 91b7e432bcef ("net/bnx2x: disable fast path interrupts")

Signed-off-by: Rasesh Mody <rasesh.mody@cavium.com>
---
 drivers/net/bnx2x/bnx2x.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/drivers/net/bnx2x/bnx2x.c b/drivers/net/bnx2x/bnx2x.c
index 98b08d11c..cc5f28a0f 100644
--- a/drivers/net/bnx2x/bnx2x.c
+++ b/drivers/net/bnx2x/bnx2x.c
@@ -4492,6 +4492,8 @@ static void bnx2x_handle_fp_tq(struct bnx2x_fastpath *fp, int scan_fp)
 	struct bnx2x_softc *sc = fp->sc;
 	uint8_t more_rx = FALSE;
 
+	PMD_DRV_LOG(DEBUG, "---> FP TASK QUEUE (%d) <--", fp->index);
+
 	/* update the fastpath index */
 	bnx2x_update_fp_sb_idx(fp);
 
@@ -4508,7 +4510,7 @@ static void bnx2x_handle_fp_tq(struct bnx2x_fastpath *fp, int scan_fp)
 	}
 
 	bnx2x_ack_sb(sc, fp->igu_sb_id, USTORM_ID,
-		   le16toh(fp->fp_hc_idx), IGU_INT_DISABLE, 1);
+		   le16toh(fp->fp_hc_idx), IGU_INT_ENABLE, 1);
 }
 
 /*
-- 
2.11.0