[dpdk-stable] [PATCH v2 01/10] net/atlantic: fix negative error codes

[dpdk-stable] [PATCH v2 01/10] net/atlantic: fix negative error codes

[Igor Russkikh]

These are just convention breakage on rte_errno,
no real harm from that.

Cc: stable@dpdk.org
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Fixes: 2b1472d7150c ("net/atlantic: implement Tx path")
---
 drivers/net/atlantic/atl_rxtx.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/atlantic/atl_rxtx.c b/drivers/net/atlantic/atl_rxtx.c
index 40c913795c79..aea58c0d5b83 100644
--- a/drivers/net/atlantic/atl_rxtx.c
+++ b/drivers/net/atlantic/atl_rxtx.c
@@ -812,12 +812,12 @@ atl_prep_pkts(__rte_unused void *tx_queue, struct rte_mbuf **tx_pkts,
 		ol_flags = m->ol_flags;
 
 		if (m->nb_segs > AQ_HW_MAX_SEGS_SIZE) {
-			rte_errno = -EINVAL;
+			rte_errno = EINVAL;
 			return i;
 		}
 
 		if (ol_flags & ATL_TX_OFFLOAD_NOTSUP_MASK) {
-			rte_errno = -ENOTSUP;
+			rte_errno = ENOTSUP;
 			return i;
 		}
 
-- 
2.17.1

[dpdk-stable] [PATCH v2 04/10] net/atlantic: fix buffer overflow

[Igor Russkikh]

From: Pavel Belous <Pavel.Belous@aquantia.com>

Found by Coverity scan. This is a real memory corruption.
There is no need in extra RTE_ALIGN macros since the
request/result structures are 4-byte aligned by definition.

Cc: stable@dpdk.org
Fixes: ce4e8d418097 ("net/atlantic: implement EEPROM get/set")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: Pavel Belous <Pavel.Belous@aquantia.com>
---
 drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c b/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
index 6841d9bce39c..f90ccfe9e010 100644
--- a/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
+++ b/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
@@ -501,7 +501,7 @@ static int aq_fw2x_get_eeprom(struct aq_hw_s *self, u32 *data, u32 len)
 	/* Write SMBUS request to cfg memory */
 	err = hw_atl_utils_fw_upload_dwords(self, self->rpc_addr,
 				(u32 *)(void *)&request,
-				RTE_ALIGN(sizeof(request), sizeof(u32)));
+				sizeof(request) / sizeof(u32));
 
 	if (err < 0)
 		return err;
@@ -523,7 +523,7 @@ static int aq_fw2x_get_eeprom(struct aq_hw_s *self, u32 *data, u32 len)
 
 	err = hw_atl_utils_fw_downld_dwords(self, self->rpc_addr + sizeof(u32),
 			&result,
-			RTE_ALIGN(sizeof(result), sizeof(u32)));
+			sizeof(result) / sizeof(u32));
 
 	if (err < 0)
 		return err;
@@ -558,7 +558,7 @@ static int aq_fw2x_set_eeprom(struct aq_hw_s *self, u32 *data, u32 len)
 	/* Write SMBUS request to cfg memory */
 	err = hw_atl_utils_fw_upload_dwords(self, self->rpc_addr,
 				(u32 *)(void *)&request,
-				RTE_ALIGN(sizeof(request), sizeof(u32)));
+				sizeof(request) / sizeof(u32));
 
 	if (err < 0)
 		return err;
@@ -589,7 +589,7 @@ static int aq_fw2x_set_eeprom(struct aq_hw_s *self, u32 *data, u32 len)
 	/* Read status of write operation */
 	err = hw_atl_utils_fw_downld_dwords(self, self->rpc_addr + sizeof(u32),
 				&result,
-				RTE_ALIGN(sizeof(result), sizeof(u32)));
+				sizeof(result) / sizeof(u32));
 
 	if (err < 0)
 		return err;
-- 
2.17.1

[dpdk-stable] [PATCH v2 06/10] net/atlantic: fix EEPROM get for small and uneven lengths

[Igor Russkikh]

From: Pavel Belous <Pavel.Belous@aquantia.com>

Fixes: ce4e8d418097 ("net/atlantic: implement EEPROM get/set")
Cc: stable@dpdk.org
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: Pavel Belous <Pavel.Belous@aquantia.com>
---
 .../net/atlantic/hw_atl/hw_atl_utils_fw2x.c   | 28 ++++++++++++++++---
 1 file changed, 24 insertions(+), 4 deletions(-)

diff --git a/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c b/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
index 78ca1eb1faaa..dc619bfdd75e 100644
--- a/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
+++ b/drivers/net/atlantic/hw_atl/hw_atl_utils_fw2x.c
@@ -534,13 +534,33 @@ static int aq_fw2x_get_eeprom(struct aq_hw_s *self, int dev_addr,
 		return err;
 
 	if (result == 0) {
-		err = hw_atl_utils_fw_downld_dwords(self,
+		u32 num_dwords = len / sizeof(u32);
+		u32 bytes_remains = len % sizeof(u32);
+
+		if (num_dwords) {
+			err = hw_atl_utils_fw_downld_dwords(self,
 				self->rpc_addr + sizeof(u32) * 2,
 				data,
-				RTE_ALIGN(len, sizeof(u32)));
+				num_dwords);
 
-		if (err < 0)
-			return err;
+			if (err < 0)
+				return err;
+		}
+
+		if (bytes_remains) {
+			u32 val = 0;
+
+			err = hw_atl_utils_fw_downld_dwords(self,
+				self->rpc_addr + sizeof(u32) * 2 + num_dwords,
+				&val,
+				sizeof(u32));
+
+			if (err < 0)
+				return err;
+
+			rte_memcpy((u8 *)data + len - bytes_remains,
+				   &val, bytes_remains);
+		}
 	}
 
 	return 0;
-- 
2.17.1

[dpdk-stable] [PATCH v2 07/10] net/atlantic: fix link configuration

[Igor Russkikh]

In case link speed is re configured after port start, it does not
takes the requested speed value, but instead just sets full autoneg
mask.

Cc: stable@dpdk.org
Fixes: 7943ba05f67c ("net/atlantic: add link status and interrupt management")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
 drivers/net/atlantic/atl_ethdev.c | 44 ++++++++++++++-----------------
 1 file changed, 20 insertions(+), 24 deletions(-)

diff --git a/drivers/net/atlantic/atl_ethdev.c b/drivers/net/atlantic/atl_ethdev.c
index 925f9722ec66..c869ff321bf5 100644
--- a/drivers/net/atlantic/atl_ethdev.c
+++ b/drivers/net/atlantic/atl_ethdev.c
@@ -465,8 +465,6 @@ atl_dev_start(struct rte_eth_dev *dev)
 	struct rte_pci_device *pci_dev = RTE_ETH_DEV_TO_PCI(dev);
 	struct rte_intr_handle *intr_handle = &pci_dev->intr_handle;
 	uint32_t intr_vector = 0;
-	uint32_t *link_speeds;
-	uint32_t speed = 0;
 	int status;
 	int err;
 
@@ -543,6 +541,8 @@ atl_dev_start(struct rte_eth_dev *dev)
 		goto error;
 	}
 
+	err = atl_dev_set_link_up(dev);
+
 	err = hw->aq_fw_ops->update_link_status(hw);
 
 	if (err)
@@ -550,26 +550,6 @@ atl_dev_start(struct rte_eth_dev *dev)
 
 	dev->data->dev_link.link_status = hw->aq_link_status.mbps != 0;
 
-	link_speeds = &dev->data->dev_conf.link_speeds;
-
-	speed = 0x0;
-
-	if (*link_speeds == ETH_LINK_SPEED_AUTONEG) {
-		speed = hw->aq_nic_cfg->link_speed_msk;
-	} else {
-		if (*link_speeds & ETH_LINK_SPEED_10G)
-			speed |= AQ_NIC_RATE_10G;
-		if (*link_speeds & ETH_LINK_SPEED_5G)
-			speed |= AQ_NIC_RATE_5G;
-		if (*link_speeds & ETH_LINK_SPEED_1G)
-			speed |= AQ_NIC_RATE_1G;
-		if (*link_speeds & ETH_LINK_SPEED_2_5G)
-			speed |=  AQ_NIC_RATE_2G5;
-		if (*link_speeds & ETH_LINK_SPEED_100M)
-			speed |= AQ_NIC_RATE_100M;
-	}
-
-	err = hw->aq_fw_ops->set_link_speed(hw, speed);
 	if (err)
 		goto error;
 
@@ -657,9 +637,25 @@ static int
 atl_dev_set_link_up(struct rte_eth_dev *dev)
 {
 	struct aq_hw_s *hw = ATL_DEV_PRIVATE_TO_HW(dev->data->dev_private);
+	uint32_t link_speeds = dev->data->dev_conf.link_speeds;
+	uint32_t speed_mask = 0;
+
+	if (link_speeds == ETH_LINK_SPEED_AUTONEG) {
+		speed_mask = hw->aq_nic_cfg->link_speed_msk;
+	} else {
+		if (link_speeds & ETH_LINK_SPEED_10G)
+			speed_mask |= AQ_NIC_RATE_10G;
+		if (link_speeds & ETH_LINK_SPEED_5G)
+			speed_mask |= AQ_NIC_RATE_5G;
+		if (link_speeds & ETH_LINK_SPEED_1G)
+			speed_mask |= AQ_NIC_RATE_1G;
+		if (link_speeds & ETH_LINK_SPEED_2_5G)
+			speed_mask |=  AQ_NIC_RATE_2G5;
+		if (link_speeds & ETH_LINK_SPEED_100M)
+			speed_mask |= AQ_NIC_RATE_100M;
+	}
 
-	return hw->aq_fw_ops->set_link_speed(hw,
-			hw->aq_nic_cfg->link_speed_msk);
+	return hw->aq_fw_ops->set_link_speed(hw, speed_mask);
 }
 
 /*
-- 
2.17.1

[dpdk-stable] [PATCH v2 09/10] net/atlantic: fix missing VLAN filter offload

[Igor Russkikh]

Original vlan offload code declared callbacks, but did not
enable the feature offload bit

Cc: stable@dpdk.org
Fixes: f7c2c2c8c558 ("net/atlantic: implement VLAN filters and offloads")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
 drivers/net/atlantic/atl_ethdev.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/net/atlantic/atl_ethdev.c b/drivers/net/atlantic/atl_ethdev.c
index c869ff321bf5..f2674f0cafe9 100644
--- a/drivers/net/atlantic/atl_ethdev.c
+++ b/drivers/net/atlantic/atl_ethdev.c
@@ -165,7 +165,8 @@ static struct rte_pci_driver rte_atl_pmd = {
 			| DEV_RX_OFFLOAD_IPV4_CKSUM \
 			| DEV_RX_OFFLOAD_UDP_CKSUM \
 			| DEV_RX_OFFLOAD_TCP_CKSUM \
-			| DEV_RX_OFFLOAD_JUMBO_FRAME)
+			| DEV_RX_OFFLOAD_JUMBO_FRAME \
+			| DEV_RX_OFFLOAD_VLAN_FILTER)
 
 #define ATL_TX_OFFLOADS (DEV_TX_OFFLOAD_VLAN_INSERT \
 			| DEV_TX_OFFLOAD_IPV4_CKSUM \
-- 
2.17.1

[dpdk-stable] [PATCH v2 10/10] net/atlantic: fix xstats to return correct number of items

[Igor Russkikh]

Max number of xstats items was returned instead of actual number
of filled in records.

Cc: stable@dpdk.org
Fixes: fbe059e87209 ("net/atlantic: implement device statistics")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
---
 drivers/net/atlantic/atl_ethdev.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/drivers/net/atlantic/atl_ethdev.c b/drivers/net/atlantic/atl_ethdev.c
index f2674f0cafe9..0f6edf878fe8 100644
--- a/drivers/net/atlantic/atl_ethdev.c
+++ b/drivers/net/atlantic/atl_ethdev.c
@@ -758,7 +758,7 @@ atl_dev_xstats_get_names(struct rte_eth_dev *dev __rte_unused,
 		snprintf(xstats_names[i].name, RTE_ETH_XSTATS_NAME_SIZE, "%s",
 			atl_xstats_tbl[i].name);
 
-	return size;
+	return i;
 }
 
 static int
@@ -778,7 +778,7 @@ atl_dev_xstats_get(struct rte_eth_dev *dev, struct rte_eth_xstat *stats,
 					atl_xstats_tbl[i].offset);
 	}
 
-	return n;
+	return i;
 }
 
 static int
-- 
2.17.1

Re: [dpdk-stable] [dpdk-dev] [PATCH v2 04/10] net/atlantic: fix buffer overflow

[Stephen Hemminger]

On Sat, 9 Mar 2019 14:03:30 +0000
Igor Russkikh <Igor.Russkikh@aquantia.com> wrote:

> From: Pavel Belous <Pavel.Belous@aquantia.com>
> 
> Found by Coverity scan. This is a real memory corruption.
> There is no need in extra RTE_ALIGN macros since the
> request/result structures are 4-byte aligned by definition.

When fixing bugs found by Coverity it is best to mark these with
the Coverity number. The convention (in Linux kernel) is shown
by these examples:

    Addresses-Coverity-ID: 1476095 ("Bad bit shift operation")
    Detected by CoverityScan, CID#1476031 ("Dereference before null check")

Should this go in the style guide somewhere?