* [PATCH V2 net-next 05/10] net/mlx4: Add a check if there are too many reserved QPs
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Dotan Barak, Or Gerlitz
In-Reply-To: <1418288280-334-1-git-send-email-ogerlitz@mellanox.com>
From: Dotan Barak <dotanb@dev.mellanox.co.il>
The number of reserved QPs is affected both from the firmware and
from the driver's requirements. This patch adds a check that
validates that this number is indeed feasable.
Signed-off-by: Dotan Barak <dotanb@dev.mellanox.co.il>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/qp.c | 8 +++++++-
1 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c
index 40e82ed..8720428 100644
--- a/drivers/net/ethernet/mellanox/mlx4/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx4/qp.c
@@ -478,6 +478,7 @@ int mlx4_init_qp_table(struct mlx4_dev *dev)
struct mlx4_qp_table *qp_table = &mlx4_priv(dev)->qp_table;
int err;
int reserved_from_top = 0;
+ int reserved_from_bot;
int k;
spin_lock_init(&qp_table->lock);
@@ -534,9 +535,14 @@ int mlx4_init_qp_table(struct mlx4_dev *dev)
* b. All the proxy SQPs (8 per function)
* c. All the tunnel QPs (8 per function)
*/
+ reserved_from_bot = mlx4_num_reserved_sqps(dev);
+ if (reserved_from_bot + reserved_from_top > dev->caps.num_qps) {
+ mlx4_err(dev, "Number of reserved QPs is higher than number of QPs\n");
+ return -EINVAL;
+ }
err = mlx4_bitmap_init(&qp_table->bitmap, dev->caps.num_qps,
- (1 << 23) - 1, mlx4_num_reserved_sqps(dev),
+ (1 << 23) - 1, reserved_from_bot,
reserved_from_top);
if (err)
return err;
--
1.7.1
^ permalink raw reply related
* [PATCH V2 net-next 04/10] net/mlx4: Change QP allocation scheme
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Eugenia Emantayev, Or Gerlitz
In-Reply-To: <1418288280-334-1-git-send-email-ogerlitz@mellanox.com>
From: Eugenia Emantayev <eugenia@mellanox.co.il>
When using BF (Blue-Flame), the QPN overrides the VLAN, CV, and SV fields
in the WQE. Thus, BF may only be used for QPNs with bits 6,7 unset.
The current Ethernet driver code reserves a Tx QP range with 256b alignment.
This is wrong because if there are more than 64 Tx QPs in use,
QPNs >= base + 65 will have bits 6/7 set.
This problem is not specific for the Ethernet driver, any entity that
tries to reserve more than 64 BF-enabled QPs should fail. Also, using
ranges is not necessary here and is wasteful.
The new mechanism introduced here will support reservation for
"Eth QPs eligible for BF" for all drivers: bare-metal, multi-PF, and VFs
(when hypervisors support WC in VMs). The flow we use is:
1. In mlx4_en, allocate Tx QPs one by one instead of a range allocation,
and request "BF enabled QPs" if BF is supported for the function
2. In the ALLOC_RES FW command, change param1 to:
a. param1[23:0] - number of QPs
b. param1[31-24] - flags controlling QPs reservation
Bit 31 refers to Eth blueflame supported QPs. Those QPs must have
bits 6 and 7 unset in order to be used in Ethernet.
Bits 24-30 of the flags are currently reserved.
When a function tries to allocate a QP, it states the required attributes
for this QP. Those attributes are considered "best-effort". If an attribute,
such as Ethernet BF enabled QP, is a must-have attribute, the function has
to check that attribute is supported before trying to do the allocation.
In a lower layer of the code, mlx4_qp_reserve_range masks out the bits
which are unsupported. If SRIOV is used, the PF validates those attributes
and masks out unsupported attributes as well. In order to notify VFs which
attributes are supported, the VF uses QUERY_FUNC_CAP command. This command's
mailbox is filled by the PF, which notifies which QP allocation attributes
it supports.
Signed-off-by: Eugenia Emantayev <eugenia@mellanox.co.il>
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/infiniband/hw/mlx4/main.c | 2 +-
drivers/infiniband/hw/mlx4/qp.c | 11 +++--
drivers/net/ethernet/mellanox/mlx4/alloc.c | 43 +++++++++++++++++---
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 10 +----
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 4 +-
drivers/net/ethernet/mellanox/mlx4/en_tx.c | 14 +++++-
drivers/net/ethernet/mellanox/mlx4/fw.c | 20 +++++++++-
drivers/net/ethernet/mellanox/mlx4/fw.h | 1 +
drivers/net/ethernet/mellanox/mlx4/main.c | 11 +++++-
drivers/net/ethernet/mellanox/mlx4/mlx4.h | 5 +-
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 2 +-
drivers/net/ethernet/mellanox/mlx4/qp.c | 24 +++++++++--
.../net/ethernet/mellanox/mlx4/resource_tracker.c | 7 +++-
include/linux/mlx4/device.h | 21 +++++++++-
14 files changed, 137 insertions(+), 38 deletions(-)
diff --git a/drivers/infiniband/hw/mlx4/main.c b/drivers/infiniband/hw/mlx4/main.c
index 0c33755..57ecc5b 100644
--- a/drivers/infiniband/hw/mlx4/main.c
+++ b/drivers/infiniband/hw/mlx4/main.c
@@ -2227,7 +2227,7 @@ static void *mlx4_ib_add(struct mlx4_dev *dev)
ibdev->steer_qpn_count = MLX4_IB_UC_MAX_NUM_QPS;
err = mlx4_qp_reserve_range(dev, ibdev->steer_qpn_count,
MLX4_IB_UC_STEER_QPN_ALIGN,
- &ibdev->steer_qpn_base);
+ &ibdev->steer_qpn_base, 0);
if (err)
goto err_counter;
diff --git a/drivers/infiniband/hw/mlx4/qp.c b/drivers/infiniband/hw/mlx4/qp.c
index 9c5150c..506d1bd 100644
--- a/drivers/infiniband/hw/mlx4/qp.c
+++ b/drivers/infiniband/hw/mlx4/qp.c
@@ -802,16 +802,19 @@ static int create_qp_common(struct mlx4_ib_dev *dev, struct ib_pd *pd,
}
}
} else {
- /* Raw packet QPNs must be aligned to 8 bits. If not, the WQE
- * BlueFlame setup flow wrongly causes VLAN insertion. */
+ /* Raw packet QPNs may not have bits 6,7 set in their qp_num;
+ * otherwise, the WQE BlueFlame setup flow wrongly causes
+ * VLAN insertion. */
if (init_attr->qp_type == IB_QPT_RAW_PACKET)
- err = mlx4_qp_reserve_range(dev->dev, 1, 1 << 8, &qpn);
+ err = mlx4_qp_reserve_range(dev->dev, 1, 1, &qpn,
+ init_attr->cap.max_send_wr ?
+ MLX4_RESERVE_ETH_BF_QP : 0);
else
if (qp->flags & MLX4_IB_QP_NETIF)
err = mlx4_ib_steer_qp_alloc(dev, 1, &qpn);
else
err = mlx4_qp_reserve_range(dev->dev, 1, 1,
- &qpn);
+ &qpn, 0);
if (err)
goto err_proxy;
}
diff --git a/drivers/net/ethernet/mellanox/mlx4/alloc.c b/drivers/net/ethernet/mellanox/mlx4/alloc.c
index b0297da..91a8acc 100644
--- a/drivers/net/ethernet/mellanox/mlx4/alloc.c
+++ b/drivers/net/ethernet/mellanox/mlx4/alloc.c
@@ -76,22 +76,53 @@ void mlx4_bitmap_free(struct mlx4_bitmap *bitmap, u32 obj, int use_rr)
mlx4_bitmap_free_range(bitmap, obj, 1, use_rr);
}
-u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt, int align)
+static unsigned long find_aligned_range(unsigned long *bitmap,
+ u32 start, u32 nbits,
+ int len, int align, u32 skip_mask)
+{
+ unsigned long end, i;
+
+again:
+ start = ALIGN(start, align);
+
+ while ((start < nbits) && (test_bit(start, bitmap) ||
+ (start & skip_mask)))
+ start += align;
+
+ if (start >= nbits)
+ return -1;
+
+ end = start+len;
+ if (end > nbits)
+ return -1;
+
+ for (i = start + 1; i < end; i++) {
+ if (test_bit(i, bitmap) || ((u32)i & skip_mask)) {
+ start = i + 1;
+ goto again;
+ }
+ }
+
+ return start;
+}
+
+u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt,
+ int align, u32 skip_mask)
{
u32 obj;
- if (likely(cnt == 1 && align == 1))
+ if (likely(cnt == 1 && align == 1 && !skip_mask))
return mlx4_bitmap_alloc(bitmap);
spin_lock(&bitmap->lock);
- obj = bitmap_find_next_zero_area(bitmap->table, bitmap->max,
- bitmap->last, cnt, align - 1);
+ obj = find_aligned_range(bitmap->table, bitmap->last,
+ bitmap->max, cnt, align, skip_mask);
if (obj >= bitmap->max) {
bitmap->top = (bitmap->top + bitmap->max + bitmap->reserved_top)
& bitmap->mask;
- obj = bitmap_find_next_zero_area(bitmap->table, bitmap->max,
- 0, cnt, align - 1);
+ obj = find_aligned_range(bitmap->table, 0, bitmap->max,
+ cnt, align, skip_mask);
}
if (obj < bitmap->max) {
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index dccf0e1..c67effb 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -595,7 +595,7 @@ static int mlx4_en_get_qp(struct mlx4_en_priv *priv)
return 0;
}
- err = mlx4_qp_reserve_range(dev, 1, 1, qpn);
+ err = mlx4_qp_reserve_range(dev, 1, 1, qpn, 0);
en_dbg(DRV, priv, "Reserved qp %d\n", *qpn);
if (err) {
en_err(priv, "Failed to reserve qp for mac registration\n");
@@ -1974,15 +1974,8 @@ int mlx4_en_alloc_resources(struct mlx4_en_priv *priv)
{
struct mlx4_en_port_profile *prof = priv->prof;
int i;
- int err;
int node;
- err = mlx4_qp_reserve_range(priv->mdev->dev, priv->tx_ring_num, 256, &priv->base_tx_qpn);
- if (err) {
- en_err(priv, "failed reserving range for TX rings\n");
- return err;
- }
-
/* Create tx Rings */
for (i = 0; i < priv->tx_ring_num; i++) {
node = cpu_to_node(i % num_online_cpus());
@@ -1991,7 +1984,6 @@ int mlx4_en_alloc_resources(struct mlx4_en_priv *priv)
goto err;
if (mlx4_en_create_tx_ring(priv, &priv->tx_ring[i],
- priv->base_tx_qpn + i,
prof->tx_ring_size, TXBB_SIZE,
node, i))
goto err;
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 69a2e1b..a850f24 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -1131,7 +1131,7 @@ int mlx4_en_create_drop_qp(struct mlx4_en_priv *priv)
int err;
u32 qpn;
- err = mlx4_qp_reserve_range(priv->mdev->dev, 1, 1, &qpn);
+ err = mlx4_qp_reserve_range(priv->mdev->dev, 1, 1, &qpn, 0);
if (err) {
en_err(priv, "Failed reserving drop qpn\n");
return err;
@@ -1174,7 +1174,7 @@ int mlx4_en_config_rss_steer(struct mlx4_en_priv *priv)
en_dbg(DRV, priv, "Configuring rss steering\n");
err = mlx4_qp_reserve_range(mdev->dev, priv->rx_ring_num,
priv->rx_ring_num,
- &rss_map->base_qpn);
+ &rss_map->base_qpn, 0);
if (err) {
en_err(priv, "Failed reserving %d qps\n", priv->rx_ring_num);
return err;
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index d0cecbd..a308d41 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -46,7 +46,7 @@
#include "mlx4_en.h"
int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
- struct mlx4_en_tx_ring **pring, int qpn, u32 size,
+ struct mlx4_en_tx_ring **pring, u32 size,
u16 stride, int node, int queue_index)
{
struct mlx4_en_dev *mdev = priv->mdev;
@@ -112,11 +112,17 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
ring, ring->buf, ring->size, ring->buf_size,
(unsigned long long) ring->wqres.buf.direct.map);
- ring->qpn = qpn;
+ err = mlx4_qp_reserve_range(mdev->dev, 1, 1, &ring->qpn,
+ MLX4_RESERVE_ETH_BF_QP);
+ if (err) {
+ en_err(priv, "failed reserving qp for TX ring\n");
+ goto err_map;
+ }
+
err = mlx4_qp_alloc(mdev->dev, ring->qpn, &ring->qp, GFP_KERNEL);
if (err) {
en_err(priv, "Failed allocating qp %d\n", ring->qpn);
- goto err_map;
+ goto err_reserve;
}
ring->qp.event = mlx4_en_sqp_event;
@@ -143,6 +149,8 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
*pring = ring;
return 0;
+err_reserve:
+ mlx4_qp_release_range(mdev->dev, ring->qpn, 1);
err_map:
mlx4_en_unmap_buffer(&ring->wqres.buf);
err_hwq_res:
diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c
index 5089f76..1469b5b 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.c
@@ -266,10 +266,15 @@ int mlx4_QUERY_FUNC_CAP_wrapper(struct mlx4_dev *dev, int slave,
#define QUERY_FUNC_CAP_MTT_QUOTA_OFFSET 0x64
#define QUERY_FUNC_CAP_MCG_QUOTA_OFFSET 0x68
+#define QUERY_FUNC_CAP_EXTRA_FLAGS_OFFSET 0x6c
+
#define QUERY_FUNC_CAP_FMR_FLAG 0x80
#define QUERY_FUNC_CAP_FLAG_RDMA 0x40
#define QUERY_FUNC_CAP_FLAG_ETH 0x80
#define QUERY_FUNC_CAP_FLAG_QUOTAS 0x10
+#define QUERY_FUNC_CAP_FLAG_VALID_MAILBOX 0x04
+
+#define QUERY_FUNC_CAP_EXTRA_FLAGS_BF_QP_ALLOC_FLAG (1UL << 31)
/* when opcode modifier = 1 */
#define QUERY_FUNC_CAP_PHYS_PORT_OFFSET 0x3
@@ -339,7 +344,7 @@ int mlx4_QUERY_FUNC_CAP_wrapper(struct mlx4_dev *dev, int slave,
mlx4_get_active_ports(dev, slave);
/* enable rdma and ethernet interfaces, and new quota locations */
field = (QUERY_FUNC_CAP_FLAG_ETH | QUERY_FUNC_CAP_FLAG_RDMA |
- QUERY_FUNC_CAP_FLAG_QUOTAS);
+ QUERY_FUNC_CAP_FLAG_QUOTAS | QUERY_FUNC_CAP_FLAG_VALID_MAILBOX);
MLX4_PUT(outbox->buf, field, QUERY_FUNC_CAP_FLAGS_OFFSET);
field = min(
@@ -401,6 +406,8 @@ int mlx4_QUERY_FUNC_CAP_wrapper(struct mlx4_dev *dev, int slave,
MLX4_PUT(outbox->buf, size, QUERY_FUNC_CAP_MCG_QUOTA_OFFSET);
MLX4_PUT(outbox->buf, size, QUERY_FUNC_CAP_MCG_QUOTA_OFFSET_DEP);
+ size = QUERY_FUNC_CAP_EXTRA_FLAGS_BF_QP_ALLOC_FLAG;
+ MLX4_PUT(outbox->buf, size, QUERY_FUNC_CAP_EXTRA_FLAGS_OFFSET);
} else
err = -EINVAL;
@@ -493,6 +500,17 @@ int mlx4_QUERY_FUNC_CAP(struct mlx4_dev *dev, u8 gen_or_port,
MLX4_GET(size, outbox, QUERY_FUNC_CAP_RESERVED_EQ_OFFSET);
func_cap->reserved_eq = size & 0xFFFFFF;
+ func_cap->extra_flags = 0;
+
+ /* Mailbox data from 0x6c and onward should only be treated if
+ * QUERY_FUNC_CAP_FLAG_VALID_MAILBOX is set in func_cap->flags
+ */
+ if (func_cap->flags & QUERY_FUNC_CAP_FLAG_VALID_MAILBOX) {
+ MLX4_GET(size, outbox, QUERY_FUNC_CAP_EXTRA_FLAGS_OFFSET);
+ if (size & QUERY_FUNC_CAP_EXTRA_FLAGS_BF_QP_ALLOC_FLAG)
+ func_cap->extra_flags |= MLX4_QUERY_FUNC_FLAGS_BF_RES_QP;
+ }
+
goto out;
}
diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.h b/drivers/net/ethernet/mellanox/mlx4/fw.h
index 475215e..0e910a4 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.h
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.h
@@ -144,6 +144,7 @@ struct mlx4_func_cap {
u8 port_flags;
u8 flags1;
u64 phys_port_id;
+ u32 extra_flags;
};
struct mlx4_func {
diff --git a/drivers/net/ethernet/mellanox/mlx4/main.c b/drivers/net/ethernet/mellanox/mlx4/main.c
index 3044f9e..6a9a941 100644
--- a/drivers/net/ethernet/mellanox/mlx4/main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/main.c
@@ -466,8 +466,13 @@ static int mlx4_dev_cap(struct mlx4_dev *dev, struct mlx4_dev_cap *dev_cap)
mlx4_is_master(dev))
dev->caps.function_caps |= MLX4_FUNC_CAP_64B_EQE_CQE;
- if (!mlx4_is_slave(dev))
+ if (!mlx4_is_slave(dev)) {
mlx4_enable_cqe_eqe_stride(dev);
+ dev->caps.alloc_res_qp_mask =
+ (dev->caps.bf_reg_size ? MLX4_RESERVE_ETH_BF_QP : 0);
+ } else {
+ dev->caps.alloc_res_qp_mask = 0;
+ }
return 0;
}
@@ -817,6 +822,10 @@ static int mlx4_slave_cap(struct mlx4_dev *dev)
slave_adjust_steering_mode(dev, &dev_cap, &hca_param);
+ if (func_cap.extra_flags & MLX4_QUERY_FUNC_FLAGS_BF_RES_QP &&
+ dev->caps.bf_reg_size)
+ dev->caps.alloc_res_qp_mask |= MLX4_RESERVE_ETH_BF_QP;
+
return 0;
err_mem:
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4.h b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
index b67ef48..6834da6 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
@@ -884,7 +884,8 @@ extern struct workqueue_struct *mlx4_wq;
u32 mlx4_bitmap_alloc(struct mlx4_bitmap *bitmap);
void mlx4_bitmap_free(struct mlx4_bitmap *bitmap, u32 obj, int use_rr);
-u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt, int align);
+u32 mlx4_bitmap_alloc_range(struct mlx4_bitmap *bitmap, int cnt,
+ int align, u32 skip_mask);
void mlx4_bitmap_free_range(struct mlx4_bitmap *bitmap, u32 obj, int cnt,
int use_rr);
u32 mlx4_bitmap_avail(struct mlx4_bitmap *bitmap);
@@ -970,7 +971,7 @@ int mlx4_DMA_wrapper(struct mlx4_dev *dev, int slave,
struct mlx4_cmd_mailbox *outbox,
struct mlx4_cmd_info *cmd);
int __mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align,
- int *base);
+ int *base, u8 flags);
void __mlx4_qp_release_range(struct mlx4_dev *dev, int base_qpn, int cnt);
int __mlx4_register_mac(struct mlx4_dev *dev, u8 port, u64 mac);
void __mlx4_unregister_mac(struct mlx4_dev *dev, u8 port, u64 mac);
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index ac48a8d..944a112 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -778,7 +778,7 @@ netdev_tx_t mlx4_en_xmit(struct sk_buff *skb, struct net_device *dev);
int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring **pring,
- int qpn, u32 size, u16 stride,
+ u32 size, u16 stride,
int node, int queue_index);
void mlx4_en_destroy_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring **pring);
diff --git a/drivers/net/ethernet/mellanox/mlx4/qp.c b/drivers/net/ethernet/mellanox/mlx4/qp.c
index 2301365..40e82ed 100644
--- a/drivers/net/ethernet/mellanox/mlx4/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx4/qp.c
@@ -42,6 +42,10 @@
#include "mlx4.h"
#include "icm.h"
+/* QP to support BF should have bits 6,7 cleared */
+#define MLX4_BF_QP_SKIP_MASK 0xc0
+#define MLX4_MAX_BF_QP_RANGE 0x40
+
void mlx4_qp_event(struct mlx4_dev *dev, u32 qpn, int event_type)
{
struct mlx4_qp_table *qp_table = &mlx4_priv(dev)->qp_table;
@@ -207,26 +211,36 @@ int mlx4_qp_modify(struct mlx4_dev *dev, struct mlx4_mtt *mtt,
EXPORT_SYMBOL_GPL(mlx4_qp_modify);
int __mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align,
- int *base)
+ int *base, u8 flags)
{
+ int bf_qp = !!(flags & (u8)MLX4_RESERVE_ETH_BF_QP);
+
struct mlx4_priv *priv = mlx4_priv(dev);
struct mlx4_qp_table *qp_table = &priv->qp_table;
- *base = mlx4_bitmap_alloc_range(&qp_table->bitmap, cnt, align);
+ if (cnt > MLX4_MAX_BF_QP_RANGE && bf_qp)
+ return -ENOMEM;
+
+ *base = mlx4_bitmap_alloc_range(&qp_table->bitmap, cnt, align,
+ bf_qp ? MLX4_BF_QP_SKIP_MASK : 0);
if (*base == -1)
return -ENOMEM;
return 0;
}
-int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align, int *base)
+int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align,
+ int *base, u8 flags)
{
u64 in_param = 0;
u64 out_param;
int err;
+ /* Turn off all unsupported QP allocation flags */
+ flags &= dev->caps.alloc_res_qp_mask;
+
if (mlx4_is_mfunc(dev)) {
- set_param_l(&in_param, cnt);
+ set_param_l(&in_param, (((u32)flags) << 24) | (u32)cnt);
set_param_h(&in_param, align);
err = mlx4_cmd_imm(dev, in_param, &out_param,
RES_QP, RES_OP_RESERVE,
@@ -238,7 +252,7 @@ int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align, int *base)
*base = get_param_l(&out_param);
return 0;
}
- return __mlx4_qp_reserve_range(dev, cnt, align, base);
+ return __mlx4_qp_reserve_range(dev, cnt, align, base, flags);
}
EXPORT_SYMBOL_GPL(mlx4_qp_reserve_range);
diff --git a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
index 16f617b..4efbd1e 100644
--- a/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
+++ b/drivers/net/ethernet/mellanox/mlx4/resource_tracker.c
@@ -1543,16 +1543,21 @@ static int qp_alloc_res(struct mlx4_dev *dev, int slave, int op, int cmd,
int align;
int base;
int qpn;
+ u8 flags;
switch (op) {
case RES_OP_RESERVE:
count = get_param_l(&in_param) & 0xffffff;
+ /* Turn off all unsupported QP allocation flags that the
+ * slave tries to set.
+ */
+ flags = (get_param_l(&in_param) >> 24) & dev->caps.alloc_res_qp_mask;
align = get_param_h(&in_param);
err = mlx4_grant_resource(dev, slave, RES_QP, count, 0);
if (err)
return err;
- err = __mlx4_qp_reserve_range(dev, count, align, &base);
+ err = __mlx4_qp_reserve_range(dev, count, align, &base, flags);
if (err) {
mlx4_release_resource(dev, slave, RES_QP, count, 0);
return err;
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 3951b53..272aa25 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -195,6 +195,22 @@ enum {
};
enum {
+ MLX4_QUERY_FUNC_FLAGS_BF_RES_QP = 1LL << 0
+};
+
+/* bit enums for an 8-bit flags field indicating special use
+ * QPs which require special handling in qp_reserve_range.
+ * Currently, this only includes QPs used by the ETH interface,
+ * where we expect to use blueflame. These QPs must not have
+ * bits 6 and 7 set in their qp number.
+ *
+ * This enum may use only bits 0..7.
+ */
+enum {
+ MLX4_RESERVE_ETH_BF_QP = 1 << 7,
+};
+
+enum {
MLX4_DEV_CAP_64B_EQE_ENABLED = 1LL << 0,
MLX4_DEV_CAP_64B_CQE_ENABLED = 1LL << 1,
MLX4_DEV_CAP_CQE_STRIDE_ENABLED = 1LL << 2,
@@ -501,6 +517,7 @@ struct mlx4_caps {
u64 phys_port_id[MLX4_MAX_PORTS + 1];
int tunnel_offload_mode;
u8 rx_checksum_flags_port[MLX4_MAX_PORTS + 1];
+ u8 alloc_res_qp_mask;
};
struct mlx4_buf_list {
@@ -950,8 +967,8 @@ int mlx4_cq_alloc(struct mlx4_dev *dev, int nent, struct mlx4_mtt *mtt,
struct mlx4_uar *uar, u64 db_rec, struct mlx4_cq *cq,
unsigned vector, int collapsed, int timestamp_en);
void mlx4_cq_free(struct mlx4_dev *dev, struct mlx4_cq *cq);
-
-int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align, int *base);
+int mlx4_qp_reserve_range(struct mlx4_dev *dev, int cnt, int align,
+ int *base, u8 flags);
void mlx4_qp_release_range(struct mlx4_dev *dev, int base_qpn, int cnt);
int mlx4_qp_alloc(struct mlx4_dev *dev, int qpn, struct mlx4_qp *qp,
--
1.7.1
^ permalink raw reply related
* [PATCH V2 net-next 03/10] net/mlx4_core: Use tasklet for user-space CQ completion events
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Or Gerlitz
In-Reply-To: <1418288280-334-1-git-send-email-ogerlitz@mellanox.com>
From: Matan Barak <matanb@mellanox.com>
Previously, we've fired all our completion callbacks straight from our ISR.
Some of those callbacks were lightweight (for example, mlx4_en's and
IPoIB napi callbacks), but some of them did more work (for example,
the user-space RDMA stack uverbs' completion handler). Besides that,
doing more than the minimal work in ISR is generally considered wrong,
it could even lead to a hard lockup of the system. Since when a lot
of completion events are generated by the hardware, the loop over those
events could be so long, that we'll get into a hard lockup by the system
watchdog.
In order to avoid that, add a new way of invoking completion events
callbacks. In the interrupt itself, we add the CQs which receive completion
event to a per-EQ list and schedule a tasklet. In the tasklet context
we loop over all the CQs in the list and invoke the user callback.
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/infiniband/hw/mlx4/cq.c | 5 ++-
drivers/net/ethernet/mellanox/mlx4/cq.c | 50 +++++++++++++++++++++++++++++
drivers/net/ethernet/mellanox/mlx4/eq.c | 16 +++++++++-
drivers/net/ethernet/mellanox/mlx4/mlx4.h | 12 +++++++
include/linux/mlx4/device.h | 5 +++
5 files changed, 86 insertions(+), 2 deletions(-)
diff --git a/drivers/infiniband/hw/mlx4/cq.c b/drivers/infiniband/hw/mlx4/cq.c
index 1066eec..a3b70f6 100644
--- a/drivers/infiniband/hw/mlx4/cq.c
+++ b/drivers/infiniband/hw/mlx4/cq.c
@@ -233,7 +233,10 @@ struct ib_cq *mlx4_ib_create_cq(struct ib_device *ibdev, int entries, int vector
if (err)
goto err_dbmap;
- cq->mcq.comp = mlx4_ib_cq_comp;
+ if (context)
+ cq->mcq.tasklet_ctx.comp = mlx4_ib_cq_comp;
+ else
+ cq->mcq.comp = mlx4_ib_cq_comp;
cq->mcq.event = mlx4_ib_cq_event;
if (context)
diff --git a/drivers/net/ethernet/mellanox/mlx4/cq.c b/drivers/net/ethernet/mellanox/mlx4/cq.c
index 56022d6..e71f313 100644
--- a/drivers/net/ethernet/mellanox/mlx4/cq.c
+++ b/drivers/net/ethernet/mellanox/mlx4/cq.c
@@ -52,6 +52,51 @@
#define MLX4_CQ_STATE_ARMED_SOL ( 6 << 8)
#define MLX4_EQ_STATE_FIRED (10 << 8)
+#define TASKLET_MAX_TIME 2
+#define TASKLET_MAX_TIME_JIFFIES msecs_to_jiffies(TASKLET_MAX_TIME)
+
+void mlx4_cq_tasklet_cb(unsigned long data)
+{
+ unsigned long flags;
+ unsigned long end = jiffies + TASKLET_MAX_TIME_JIFFIES;
+ struct mlx4_eq_tasklet *ctx = (struct mlx4_eq_tasklet *)data;
+ struct mlx4_cq *mcq, *temp;
+
+ spin_lock_irqsave(&ctx->lock, flags);
+ list_splice_tail_init(&ctx->list, &ctx->process_list);
+ spin_unlock_irqrestore(&ctx->lock, flags);
+
+ list_for_each_entry_safe(mcq, temp, &ctx->process_list, tasklet_ctx.list) {
+ list_del_init(&mcq->tasklet_ctx.list);
+ mcq->tasklet_ctx.comp(mcq);
+ if (atomic_dec_and_test(&mcq->refcount))
+ complete(&mcq->free);
+ if (time_after(jiffies, end))
+ break;
+ }
+
+ if (!list_empty(&ctx->process_list))
+ tasklet_schedule(&ctx->task);
+}
+
+static void mlx4_add_cq_to_tasklet(struct mlx4_cq *cq)
+{
+ unsigned long flags;
+ struct mlx4_eq_tasklet *tasklet_ctx = cq->tasklet_ctx.priv;
+
+ spin_lock_irqsave(&tasklet_ctx->lock, flags);
+ /* When migrating CQs between EQs will be implemented, please note
+ * that you need to sync this point. It is possible that
+ * while migrating a CQ, completions on the old EQs could
+ * still arrive.
+ */
+ if (list_empty_careful(&cq->tasklet_ctx.list)) {
+ atomic_inc(&cq->refcount);
+ list_add_tail(&cq->tasklet_ctx.list, &tasklet_ctx->list);
+ }
+ spin_unlock_irqrestore(&tasklet_ctx->lock, flags);
+}
+
void mlx4_cq_completion(struct mlx4_dev *dev, u32 cqn)
{
struct mlx4_cq *cq;
@@ -292,6 +337,11 @@ int mlx4_cq_alloc(struct mlx4_dev *dev, int nent,
cq->uar = uar;
atomic_set(&cq->refcount, 1);
init_completion(&cq->free);
+ cq->comp = mlx4_add_cq_to_tasklet;
+ cq->tasklet_ctx.priv =
+ &priv->eq_table.eq[cq->vector].tasklet_ctx;
+ INIT_LIST_HEAD(&cq->tasklet_ctx.list);
+
cq->irq = priv->eq_table.eq[cq->vector].irq;
return 0;
diff --git a/drivers/net/ethernet/mellanox/mlx4/eq.c b/drivers/net/ethernet/mellanox/mlx4/eq.c
index d68b264..3d275fb 100644
--- a/drivers/net/ethernet/mellanox/mlx4/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx4/eq.c
@@ -450,7 +450,7 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
{
struct mlx4_priv *priv = mlx4_priv(dev);
struct mlx4_eqe *eqe;
- int cqn;
+ int cqn = -1;
int eqes_found = 0;
int set_ci = 0;
int port;
@@ -758,6 +758,13 @@ static int mlx4_eq_int(struct mlx4_dev *dev, struct mlx4_eq *eq)
eq_set_ci(eq, 1);
+ /* cqn is 24bit wide but is initialized such that its higher bits
+ * are ones too. Thus, if we got any event, cqn's high bits should be off
+ * and we need to schedule the tasklet.
+ */
+ if (!(cqn & ~0xffffff))
+ tasklet_schedule(&eq->tasklet_ctx.task);
+
return eqes_found;
}
@@ -971,6 +978,12 @@ static int mlx4_create_eq(struct mlx4_dev *dev, int nent,
eq->cons_index = 0;
+ INIT_LIST_HEAD(&eq->tasklet_ctx.list);
+ INIT_LIST_HEAD(&eq->tasklet_ctx.process_list);
+ spin_lock_init(&eq->tasklet_ctx.lock);
+ tasklet_init(&eq->tasklet_ctx.task, mlx4_cq_tasklet_cb,
+ (unsigned long)&eq->tasklet_ctx);
+
return err;
err_out_free_mtt:
@@ -1027,6 +1040,7 @@ static void mlx4_free_eq(struct mlx4_dev *dev,
}
}
synchronize_irq(eq->irq);
+ tasklet_disable(&eq->tasklet_ctx.task);
mlx4_mtt_cleanup(dev, &eq->mtt);
for (i = 0; i < npages; ++i)
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4.h b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
index f48e7c3..b67ef48 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
@@ -43,6 +43,8 @@
#include <linux/timer.h>
#include <linux/semaphore.h>
#include <linux/workqueue.h>
+#include <linux/interrupt.h>
+#include <linux/spinlock.h>
#include <linux/mlx4/device.h>
#include <linux/mlx4/driver.h>
@@ -373,6 +375,14 @@ struct mlx4_srq_context {
__be64 db_rec_addr;
};
+struct mlx4_eq_tasklet {
+ struct list_head list;
+ struct list_head process_list;
+ struct tasklet_struct task;
+ /* lock on completion tasklet list */
+ spinlock_t lock;
+};
+
struct mlx4_eq {
struct mlx4_dev *dev;
void __iomem *doorbell;
@@ -383,6 +393,7 @@ struct mlx4_eq {
int nent;
struct mlx4_buf_list *page_list;
struct mlx4_mtt mtt;
+ struct mlx4_eq_tasklet tasklet_ctx;
};
struct mlx4_slave_eqe {
@@ -1146,6 +1157,7 @@ void mlx4_cmd_use_polling(struct mlx4_dev *dev);
int mlx4_comm_cmd(struct mlx4_dev *dev, u8 cmd, u16 param,
unsigned long timeout);
+void mlx4_cq_tasklet_cb(unsigned long data);
void mlx4_cq_completion(struct mlx4_dev *dev, u32 cqn);
void mlx4_cq_event(struct mlx4_dev *dev, u32 cqn, int event_type);
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index cf09e65..3951b53 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -621,6 +621,11 @@ struct mlx4_cq {
atomic_t refcount;
struct completion free;
+ struct {
+ struct list_head list;
+ void (*comp)(struct mlx4_cq *);
+ void *priv;
+ } tasklet_ctx;
};
struct mlx4_qp {
--
1.7.1
^ permalink raw reply related
* [PATCH V2 net-next 00/10] mlx4 driver update
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Or Gerlitz
Hi Dave,
This series from Matan, Jenny, Dotan and myself is mostly about adding
support to a new performance optimized flow steering mode (patches 4-10).
The 1st two patches are small fixes (one for VXLAN and one for SRIOV),
and the third patch is a fix to avoid hard-lockup situation when many
(hunderds) processes holding user-space QPs/CQs get events.
Matan and Or.
changes from V1:
- in patch #3, follow the pratice used by __do_softirq and limit
the tasklet running time by 2ms instread of hard coded 1000 iterations
changes from V0:
- avoid using magic constancts in patch #2
- open-coded trivial helpers in patch #6
Dotan Barak (1):
net/mlx4: Add a check if there are too many reserved QPs
Eugenia Emantayev (1):
net/mlx4: Change QP allocation scheme
Matan Barak (6):
net/mlx4_core: Use tasklet for user-space CQ completion events
net/mlx4: Add mlx4_bitmap zone allocator
net/mlx4: Add A0 hybrid steering
net/mlx4_core: Add explicit error message when rule doesn't meet configuration
net/mlx4: Refactor QUERY_PORT
net/mlx4: Add support for A0 steering
Or Gerlitz (2):
net/mlx4_en: Set csum level for encapsulated packets
net/mlx4_core: Mask out host side virtualization features for guests
drivers/infiniband/hw/mlx4/cq.c | 5 +-
drivers/infiniband/hw/mlx4/main.c | 2 +-
drivers/infiniband/hw/mlx4/qp.c | 13 +-
drivers/net/ethernet/mellanox/mlx4/alloc.c | 425 +++++++++++++++++++-
drivers/net/ethernet/mellanox/mlx4/cq.c | 50 +++
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 13 +-
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 8 +-
drivers/net/ethernet/mellanox/mlx4/en_tx.c | 14 +-
drivers/net/ethernet/mellanox/mlx4/eq.c | 16 +-
drivers/net/ethernet/mellanox/mlx4/fw.c | 225 ++++++++---
drivers/net/ethernet/mellanox/mlx4/fw.h | 42 ++-
drivers/net/ethernet/mellanox/mlx4/main.c | 218 +++++++++--
drivers/net/ethernet/mellanox/mlx4/mcg.c | 21 +-
drivers/net/ethernet/mellanox/mlx4/mlx4.h | 97 +++++-
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 2 +-
drivers/net/ethernet/mellanox/mlx4/qp.c | 303 +++++++++++++-
.../net/ethernet/mellanox/mlx4/resource_tracker.c | 7 +-
include/linux/mlx4/device.h | 51 +++-
18 files changed, 1342 insertions(+), 170 deletions(-)
^ permalink raw reply
* [PATCH V2 net-next 02/10] net/mlx4_core: Mask out host side virtualization features for guests
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Or Gerlitz
In-Reply-To: <1418288280-334-1-git-send-email-ogerlitz@mellanox.com>
When VFs (guests in this context) issue the QUERY_DEV_CAP command, they
need not be told that host side virtualization features such as VST, FSM
(MAC anti-spoofing) and running > 80 VFs are supported by the device.
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/fw.c | 12 +++++++++++-
1 files changed, 11 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/fw.c b/drivers/net/ethernet/mellanox/mlx4/fw.c
index 4251f81..5089f76 100644
--- a/drivers/net/ethernet/mellanox/mlx4/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx4/fw.c
@@ -973,6 +973,10 @@ out:
return err;
}
+#define DEV_CAP_EXT_2_FLAG_VLAN_CONTROL (1 << 26)
+#define DEV_CAP_EXT_2_FLAG_80_VFS (1 << 21)
+#define DEV_CAP_EXT_2_FLAG_FSM (1 << 20)
+
int mlx4_QUERY_DEV_CAP_wrapper(struct mlx4_dev *dev, int slave,
struct mlx4_vhcr *vhcr,
struct mlx4_cmd_mailbox *inbox,
@@ -982,7 +986,7 @@ int mlx4_QUERY_DEV_CAP_wrapper(struct mlx4_dev *dev, int slave,
u64 flags;
int err = 0;
u8 field;
- u32 bmme_flags;
+ u32 bmme_flags, field32;
int real_port;
int slave_port;
int first_port;
@@ -1053,6 +1057,12 @@ int mlx4_QUERY_DEV_CAP_wrapper(struct mlx4_dev *dev, int slave,
field &= ~0x80;
MLX4_PUT(outbox->buf, field, QUERY_DEV_CAP_FLOW_STEERING_IPOIB_OFFSET);
+ /* turn off host side virt features (VST, FSM, etc) for guests */
+ MLX4_GET(field32, outbox->buf, QUERY_DEV_CAP_EXT_2_FLAGS_OFFSET);
+ field32 &= ~(DEV_CAP_EXT_2_FLAG_VLAN_CONTROL | DEV_CAP_EXT_2_FLAG_80_VFS |
+ DEV_CAP_EXT_2_FLAG_FSM);
+ MLX4_PUT(outbox->buf, field32, QUERY_DEV_CAP_EXT_2_FLAGS_OFFSET);
+
return 0;
}
--
1.7.1
^ permalink raw reply related
* [PATCH V2 net-next 01/10] net/mlx4_en: Set csum level for encapsulated packets
From: Or Gerlitz @ 2014-12-11 8:57 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Matan Barak, Amir Vadai, Tal Alon, Jack Morgenstein,
Or Gerlitz
In-Reply-To: <1418288280-334-1-git-send-email-ogerlitz@mellanox.com>
This was dropped by mistake for the napi_gro_frags flow, fix that.
Fixes: dd65beac48a5 ('net/mlx4_en: Extend usage of napi_gro_frags')
Signed-off-by: Or Gerlitz <ogerlitz@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 3 ++-
1 files changed, 2 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 4ca396e..69a2e1b 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -888,7 +888,8 @@ int mlx4_en_process_rx_cq(struct net_device *dev, struct mlx4_en_cq *cq, int bud
gro_skb->ip_summed = ip_summed;
if (l2_tunnel && ip_summed == CHECKSUM_UNNECESSARY)
- gro_skb->encapsulation = 1;
+ gro_skb->csum_level = 1;
+
if ((cqe->vlan_my_qpn &
cpu_to_be32(MLX4_CQE_VLAN_PRESENT_MASK)) &&
(dev->features & NETIF_F_HW_VLAN_CTAG_RX)) {
--
1.7.1
^ permalink raw reply related
* [PATCH v2 net] be2net: Export tunnel offloads only when a VxLAN tunnel is created
From: Sathya Perla @ 2014-12-11 8:24 UTC (permalink / raw)
To: netdev
From: Sriharsha Basavapatna <sriharsha.basavapatna@emulex.com>
The encapsulated offload flags shouldn't be unconditionally exported
to the stack. The stack expects offloading to work across all tunnel
types when those flags are set. This would break other tunnels (like
GRE) since be2net currently supports tunnel offload for VxLAN only.
Also, with VxLANs Skyhawk-R can offload only 1 UDP dport. If more
than 1 UDP port is added, we should disable offloads in that case too.
Signed-off-by: Sriharsha Basavapatna <sriharsha.basavapatna@emulex.com>
Signed-off-by: Sathya Perla <sathya.perla@emulex.com>
---
v2 changes: fix a bad indentation pointed out by Dave M.
drivers/net/ethernet/emulex/benet/be.h | 1 +
drivers/net/ethernet/emulex/benet/be_main.c | 42 ++++++++++++++++++++++-------
2 files changed, 33 insertions(+), 10 deletions(-)
diff --git a/drivers/net/ethernet/emulex/benet/be.h b/drivers/net/ethernet/emulex/benet/be.h
index 9a2d752..712e7f8 100644
--- a/drivers/net/ethernet/emulex/benet/be.h
+++ b/drivers/net/ethernet/emulex/benet/be.h
@@ -522,6 +522,7 @@ struct be_adapter {
u8 hba_port_num;
u16 pvid;
__be16 vxlan_port;
+ int vxlan_port_count;
struct phy_info phy;
u8 wol_cap;
bool wol_en;
diff --git a/drivers/net/ethernet/emulex/benet/be_main.c b/drivers/net/ethernet/emulex/benet/be_main.c
index 597c463..593943e 100644
--- a/drivers/net/ethernet/emulex/benet/be_main.c
+++ b/drivers/net/ethernet/emulex/benet/be_main.c
@@ -3123,6 +3123,8 @@ static void be_mac_clear(struct be_adapter *adapter)
#ifdef CONFIG_BE2NET_VXLAN
static void be_disable_vxlan_offloads(struct be_adapter *adapter)
{
+ struct net_device *netdev = adapter->netdev;
+
if (adapter->flags & BE_FLAGS_VXLAN_OFFLOADS)
be_cmd_manage_iface(adapter, adapter->if_handle,
OP_CONVERT_TUNNEL_TO_NORMAL);
@@ -3132,6 +3134,9 @@ static void be_disable_vxlan_offloads(struct be_adapter *adapter)
adapter->flags &= ~BE_FLAGS_VXLAN_OFFLOADS;
adapter->vxlan_port = 0;
+
+ netdev->hw_enc_features = 0;
+ netdev->hw_features &= ~(NETIF_F_GSO_UDP_TUNNEL);
}
#endif
@@ -4369,6 +4374,19 @@ static int be_ndo_bridge_getlink(struct sk_buff *skb, u32 pid, u32 seq,
}
#ifdef CONFIG_BE2NET_VXLAN
+/* VxLAN offload Notes:
+ *
+ * The stack defines tunnel offload flags (hw_enc_features) for IP and doesn't
+ * distinguish various types of transports (VxLAN, GRE, NVGRE ..). So, offload
+ * is expected to work across all types of IP tunnels once exported. Skyhawk
+ * supports offloads for either VxLAN or NVGRE, exclusively. So we export VxLAN
+ * offloads in hw_enc_features only when a VxLAN port is added. Note this only
+ * ensures that other tunnels work fine while VxLAN offloads are not enabled.
+ *
+ * Skyhawk supports VxLAN offloads only for one UDP dport. So, if the stack
+ * adds more than one port, disable offloads and don't re-enable them again
+ * until after all the tunnels are removed.
+ */
static void be_add_vxlan_port(struct net_device *netdev, sa_family_t sa_family,
__be16 port)
{
@@ -4380,13 +4398,16 @@ static void be_add_vxlan_port(struct net_device *netdev, sa_family_t sa_family,
return;
if (adapter->flags & BE_FLAGS_VXLAN_OFFLOADS) {
- dev_warn(dev, "Cannot add UDP port %d for VxLAN offloads\n",
- be16_to_cpu(port));
dev_info(dev,
"Only one UDP port supported for VxLAN offloads\n");
- return;
+ dev_info(dev, "Disabling VxLAN offloads\n");
+ adapter->vxlan_port_count++;
+ goto err;
}
+ if (adapter->vxlan_port_count++ >= 1)
+ return;
+
status = be_cmd_manage_iface(adapter, adapter->if_handle,
OP_CONVERT_NORMAL_TO_TUNNEL);
if (status) {
@@ -4402,6 +4423,11 @@ static void be_add_vxlan_port(struct net_device *netdev, sa_family_t sa_family,
adapter->flags |= BE_FLAGS_VXLAN_OFFLOADS;
adapter->vxlan_port = port;
+ netdev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
+ NETIF_F_TSO | NETIF_F_TSO6 |
+ NETIF_F_GSO_UDP_TUNNEL;
+ netdev->hw_features |= NETIF_F_GSO_UDP_TUNNEL;
+
dev_info(dev, "Enabled VxLAN offloads for UDP port %d\n",
be16_to_cpu(port));
return;
@@ -4418,13 +4444,15 @@ static void be_del_vxlan_port(struct net_device *netdev, sa_family_t sa_family,
return;
if (adapter->vxlan_port != port)
- return;
+ goto done;
be_disable_vxlan_offloads(adapter);
dev_info(&adapter->pdev->dev,
"Disabled VxLAN offloads for UDP port %d\n",
be16_to_cpu(port));
+done:
+ adapter->vxlan_port_count--;
}
static bool be_gso_check(struct sk_buff *skb, struct net_device *dev)
@@ -4468,12 +4496,6 @@ static void be_netdev_init(struct net_device *netdev)
{
struct be_adapter *adapter = netdev_priv(netdev);
- if (skyhawk_chip(adapter)) {
- netdev->hw_enc_features |= NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM |
- NETIF_F_TSO | NETIF_F_TSO6 |
- NETIF_F_GSO_UDP_TUNNEL;
- netdev->hw_features |= NETIF_F_GSO_UDP_TUNNEL;
- }
netdev->hw_features |= NETIF_F_SG | NETIF_F_TSO | NETIF_F_TSO6 |
NETIF_F_IP_CSUM | NETIF_F_IPV6_CSUM | NETIF_F_RXCSUM |
NETIF_F_HW_VLAN_CTAG_TX;
--
2.2.0
^ permalink raw reply related
* [PATCH iproute2] ip lib: Added shorter timestamp option
From: Vadim Kochan @ 2014-12-11 8:12 UTC (permalink / raw)
To: netdev; +Cc: Vadim Kochan
From: Vadim Kochan <vadim4j@gmail.com>
Added another timestamp format to look like more logging info:
[Dec 01 01:46:20.675589] 2: enp0s25: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default
link/ether 3c:97:0e:a3:86:2e brd ff:ff:ff:ff:ff:ff
Signed-off-by: Vadim Kochan <vadim4j@gmail.com>
---
include/utils.h | 1 +
ip/ip.c | 5 ++++-
lib/utils.c | 15 ++++++++++++---
man/man8/ip-monitor.8 | 13 +++++++++++++
4 files changed, 30 insertions(+), 4 deletions(-)
diff --git a/include/utils.h b/include/utils.h
index eef9c42..eecbc39 100644
--- a/include/utils.h
+++ b/include/utils.h
@@ -19,6 +19,7 @@ extern int show_raw;
extern int resolve_hosts;
extern int oneline;
extern int timestamp;
+extern int timestamp_short;
extern char * _SL_;
extern int max_flush_loops;
extern int batch_mode;
diff --git a/ip/ip.c b/ip/ip.c
index 5f759d5..9b90707 100644
--- a/ip/ip.c
+++ b/ip/ip.c
@@ -53,7 +53,7 @@ static void usage(void)
" -f[amily] { inet | inet6 | ipx | dnet | bridge | link } |\n"
" -4 | -6 | -I | -D | -B | -0 |\n"
" -l[oops] { maximum-addr-flush-attempts } |\n"
-" -o[neline] | -t[imestamp] | -b[atch] [filename] |\n"
+" -o[neline] | -t[imestamp] | -t[short] | -b[atch] [filename] |\n"
" -rc[vbuf] [size]}\n");
exit(-1);
}
@@ -232,6 +232,9 @@ int main(int argc, char **argv)
++oneline;
} else if (matches(opt, "-timestamp") == 0) {
++timestamp;
+ } else if (matches(opt, "-tshort") == 0) {
+ ++timestamp;
+ ++timestamp_short;
#if 0
} else if (matches(opt, "-numeric") == 0) {
rtnl_names_numeric++;
diff --git a/lib/utils.c b/lib/utils.c
index 987377b..3102920 100644
--- a/lib/utils.c
+++ b/lib/utils.c
@@ -32,6 +32,8 @@
#include "utils.h"
+int timestamp_short = 0;
+
int get_integer(int *val, const char *arg, int base)
{
long res;
@@ -773,13 +775,20 @@ int print_timestamp(FILE *fp)
{
struct timeval tv;
char *tstr;
+ char tshort[40] = {};
memset(&tv, 0, sizeof(tv));
gettimeofday(&tv, NULL);
- tstr = asctime(localtime(&tv.tv_sec));
- tstr[strlen(tstr)-1] = 0;
- fprintf(fp, "Timestamp: %s %ld usec\n", tstr, (long)tv.tv_usec);
+ if (timestamp_short) {
+ strftime(tshort, sizeof(tshort), "%b %d %H:%M:%S", localtime(&tv.tv_sec));
+ fprintf(fp, "[%s.%-6ld] ", tshort, (long)tv.tv_usec);
+ } else {
+ tstr = asctime(localtime(&tv.tv_sec));
+ tstr[strlen(tstr)-1] = 0;
+ fprintf(fp, "Timestamp: %s %ld usec\n", tstr, (long)tv.tv_usec);
+ }
+
return 0;
}
diff --git a/man/man8/ip-monitor.8 b/man/man8/ip-monitor.8
index 68e83f1..c131ae7 100644
--- a/man/man8/ip-monitor.8
+++ b/man/man8/ip-monitor.8
@@ -16,6 +16,19 @@ ip-monitor, rtmon \- state monitoring
]
.sp
+.SH OPTIONS
+
+.TP
+.BR "\-t" , " \-timestamp"
+Prints timestamp before the event message on the separated line in format:
+ Timestamp: <Day> <Month> <day> <HH:MM:SS> <YYYY> <usecs> usec
+ EVENT
+
+.TP
+.BR "\-ts" , " \-tshort"
+Prints short timestamp before the event message on the same line in format:
+ [<Month> <day> <HH:MM:SS>.usecs ] EVENT
+
.SH DESCRIPTION
The
.B ip
--
2.1.3
^ permalink raw reply related
* [net PATCH] fib_trie: Fix trie balancing issue if new node pushes down existing node
From: Alexander Duyck @ 2014-12-11 5:49 UTC (permalink / raw)
To: netdev; +Cc: davem
This patch addresses an issue with the level compression of the fib_trie.
Specifically in the case of adding a new leaf that triggers a new node to
be added that takes the place of the old node. The result is a trie where
the 1 child tnode is on one side and one leaf is on the other which gives
you a very deep trie. Below is the script I used to generate a trie on
dummy0 with a 10.X.X.X family of addresses.
ip link add type dummy
ipval=184549374
bit=2
for i in `seq 1 23`
do
ifconfig dummy0:$bit $ipval/8
ipval=`expr $ipval - $bit`
bit=`expr $bit \* 2`
done
cat /proc/net/fib_triestat
Running the script before the patch:
Local:
Aver depth: 10.82
Max depth: 23
Leaves: 29
Prefixes: 30
Internal nodes: 27
1: 26 2: 1
Pointers: 56
Null ptrs: 1
Total size: 5 kB
After applying the patch and repeating:
Local:
Aver depth: 4.72
Max depth: 9
Leaves: 29
Prefixes: 30
Internal nodes: 12
1: 3 2: 2 3: 7
Pointers: 70
Null ptrs: 30
Total size: 4 kB
What this fix does is start the rebalance at the newly created tnode
instead of at the parent tnode. This way if there is a gap between the
parent and the new node it doesn't prevent the new tnode from being
coalesced with any pre-existing nodes that may have been pushed into one
of the new nodes child branches.
Signed-off-by: Alexander Duyck <alexander.h.duyck@redhat.com>
---
net/ipv4/fib_trie.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)
diff --git a/net/ipv4/fib_trie.c b/net/ipv4/fib_trie.c
index e9cb258..18bcaf2 100644
--- a/net/ipv4/fib_trie.c
+++ b/net/ipv4/fib_trie.c
@@ -1143,8 +1143,9 @@ static struct list_head *fib_insert_node(struct trie *t, u32 key, int plen)
put_child(tp, cindex, (struct rt_trie_node *)tn);
} else {
rcu_assign_pointer(t->trie, (struct rt_trie_node *)tn);
- tp = tn;
}
+
+ tp = tn;
}
if (tp && tp->pos + tp->bits > 32)
^ permalink raw reply related
* Re: [PATCH net-next RESEND] net: Do not call ndo_dflt_fdb_dump if ndo_fdb_dump is defined.
From: Roopa Prabhu @ 2014-12-11 7:31 UTC (permalink / raw)
To: Hubert Sokolowski; +Cc: netdev, Jamal Hadi Salim
In-Reply-To: <c6756bd161048ac4b4407e308045fe73.squirrel@poczta.wsisiz.edu.pl>
On 12/10/14, 11:37 AM, Hubert Sokolowski wrote:
> This change restores the semantic that was present
> before 5e6d243587990a588143b9da3974833649595587
> "bridge: netlink dump interface at par with brctl"
> on how ndo_dflt_fdb_dump is called.
> This semantic is still used for add and del operations
> so let's keep it consistent.
> Driver can still call ndo_dflt_fdb_dump from inside
> its own fdb_dump routine when needed.
>
> Signed-off-by: Hubert Sokolowski <h.sokolowski@wit.edu.pl>
> ---
> net/core/rtnetlink.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index eaa057f..a9e5c37 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -2692,10 +2692,11 @@ static int rtnl_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb)
> idx);
> }
>
> - idx = ndo_dflt_fdb_dump(skb, cb, dev, NULL, idx);
> if (dev->netdev_ops->ndo_fdb_dump)
> idx = dev->netdev_ops->ndo_fdb_dump(skb, cb, bdev, dev,
> idx);
> + else
> + idx = ndo_dflt_fdb_dump(skb, cb, dev, NULL, idx);
>
> cops = NULL;
> }
I think commit
"5e6d243587990a588143b9da3974833649595587 "bridge: netlink dump
interface at par with brctl" tried to make sure even the dflt entries
(ie dev->uc and dev->mc) were also printed in the below case. ie the
'self' entries in the below output.
# ./bridge fdb show brport eth1
02:00:00:12:01:02 vlan 0 master br0 permanent
00:17:42:8a:b4:05 vlan 0 master br0 permanent
00:17:42:8a:b4:07 self permanent
33:33:00:00:00:01 self permanent
Am guessing reverting the patch is going to make the 'self' entries in
the above output to go away.
Can you please confirm ?.
Also, if i hear your concern correctly, for bridge ports that implement
ndo_fdb_dump, with commit 5e6d243587990a588143b9da3974833649595587, we
will get two entries for each 'self' entry above.
Can you also paste sample output for that ?.
Thanks,
Roopa
^ permalink raw reply
* RE: [PATCH net] be2net: Export tunnel offloads only when a VxLAN tunnel is created
From: Sathya Perla @ 2014-12-11 7:24 UTC (permalink / raw)
To: David Miller; +Cc: netdev@vger.kernel.org
In-Reply-To: <20141210.145338.683039710260795306.davem@davemloft.net>
> -----Original Message-----
> From: David Miller [mailto:davem@davemloft.net]
>
> From: Sathya Perla <sathya.perla@emulex.com>
> Date: Wed, 10 Dec 2014 04:56:04 -0500
>
> > + netdev->hw_enc_features |= (NETIF_F_IP_CSUM |
> NETIF_F_IPV6_CSUM |
> > + NETIF_F_TSO | NETIF_F_TSO6 | NETIF_F_GSO_UDP_TUNNEL);
>
> Please indent this properly:
>
> netdev->hw_enc_features |= (NETIF_F_IP_CSUM |
> NETIF_F_IPV6_CSUM |
> NETIF_F_TSO | NETIF_F_TSO6 |
> NETIF_F_GSO_UDP_TUNNEL);
Oops, checkpatch didn't seem to catch this...will fix it up and send out a v2 right away...
thanks!
^ permalink raw reply
* [net] gre: fix the inner mac header in nbma gre tunnels xmit path
From: Timo Teräs @ 2014-12-11 7:14 UTC (permalink / raw)
To: Linux Netdev List; +Cc: Timo Teräs, Tom Herbert, Alexander Duyck
In-Reply-To: <CA+mtBx-WsgMRAYDUaPiVaBt43cRx7tb-0hRJ+6Eao8vjd3j8aQ@mail.gmail.com>
The NBMA GRE tunnels temporarily push GRE header that contain the
per-packet NBMA destination on the skb via header ops early in xmit
path. It is the later pulled before the real GRE header is constructed.
The inner mac was thus set differently in nbma case: the GRE header
has been pushed by neighbor layer, and mac header points to beginning
of the temporary gre header (set by dev_queue_xmit).
Now that the offloads expect mac header to point to the gre payload,
fix the xmit patch to:
- pull first the temporary gre header away
- and reset mac header to point to gre payload
This fixes tso to work again with nbma tunnels.
Fixes: 14051f0452a2 ("gre: Use inner mac length when computing tunnel length"
Signed-off-by: Timo Teräs <timo.teras@iki.fi>
Cc: Tom Herbert <therbert@google.com>
Cc: Alexander Duyck <alexander.h.duyck@redhat.com>
---
Though, normally mac header does point to the begging of the hardware header.
E.g. in ethernet case it's pointing to the ethernet header. But now in gre
case it's instead pointing to the payload which seems counter-intuitive to me.
But I guess tunnels are a bit of special case, and there's valid reasons to
have it point to tunnel payload too.
Applying this patch on top of the Tom's previous fix of 14051f0452a2 seems to
now make my dmvpn scenario work again. So this should go to -stable too
(atleast 3.14).
net/ipv4/ip_gre.c | 9 +++++----
1 file changed, 5 insertions(+), 4 deletions(-)
diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
index 94213c8..6d2f84d 100644
--- a/net/ipv4/ip_gre.c
+++ b/net/ipv4/ip_gre.c
@@ -250,10 +250,6 @@ static netdev_tx_t ipgre_xmit(struct sk_buff *skb,
struct ip_tunnel *tunnel = netdev_priv(dev);
const struct iphdr *tnl_params;
- skb = gre_handle_offloads(skb, !!(tunnel->parms.o_flags&TUNNEL_CSUM));
- if (IS_ERR(skb))
- goto out;
-
if (dev->header_ops) {
/* Need space for new headers */
if (skb_cow_head(skb, dev->needed_headroom -
@@ -266,6 +262,7 @@ static netdev_tx_t ipgre_xmit(struct sk_buff *skb,
* to gre header.
*/
skb_pull(skb, tunnel->hlen + sizeof(struct iphdr));
+ skb_reset_mac_header(mac);
} else {
if (skb_cow_head(skb, dev->needed_headroom))
goto free_skb;
@@ -273,6 +270,10 @@ static netdev_tx_t ipgre_xmit(struct sk_buff *skb,
tnl_params = &tunnel->parms.iph;
}
+ skb = gre_handle_offloads(skb, !!(tunnel->parms.o_flags&TUNNEL_CSUM));
+ if (IS_ERR(skb))
+ goto out;
+
__gre_xmit(skb, dev, tnl_params, skb->protocol);
return NETDEV_TX_OK;
--
2.2.0
^ permalink raw reply related
* [PATCH net v2] gianfar: Fix dma check map error when DMA_API_DEBUG is enabled
From: Kevin Hao @ 2014-12-11 6:08 UTC (permalink / raw)
To: netdev; +Cc: Claudiu Manoil, David Miller
In-Reply-To: <20141211020646.GB18701@pek-khao-d1.corp.ad.wrs.com>
We need to use dma_mapping_error() to check the dma address returned
by dma_map_single/page(). Otherwise we would get warning like this:
WARNING: at lib/dma-debug.c:1140
Modules linked in:
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 3.18.0-rc2-next-20141029 #196
task: c0834300 ti: effe6000 task.ti: c0874000
NIP: c02b2c98 LR: c02b2c98 CTR: c030abc4
REGS: effe7d70 TRAP: 0700 Not tainted (3.18.0-rc2-next-20141029)
MSR: 00021000 <CE,ME> CR: 22044022 XER: 20000000
GPR00: c02b2c98 effe7e20 c0834300 00000098 00021000 00000000 c030b898 00000003
GPR08: 00000001 00000000 00000001 749eec9d 22044022 1001abe0 00000020 ef278678
GPR16: ef278670 ef278668 ef278660 070a8040 c087f99c c08cdc60 00029000 c0840d44
GPR24: c08be6e8 c0840000 effe7e78 ef041340 00000600 ef114e10 00000000 c08be6e0
NIP [c02b2c98] check_unmap+0x51c/0x9e4
LR [c02b2c98] check_unmap+0x51c/0x9e4
Call Trace:
[effe7e20] [c02b2c98] check_unmap+0x51c/0x9e4 (unreliable)
[effe7e70] [c02b31d8] debug_dma_unmap_page+0x78/0x8c
[effe7ed0] [c03d1640] gfar_clean_rx_ring+0x208/0x488
[effe7f40] [c03d1a9c] gfar_poll_rx_sq+0x3c/0xa8
[effe7f60] [c04f8714] net_rx_action+0xc0/0x178
[effe7f90] [c00435a0] __do_softirq+0x100/0x1fc
[effe7fe0] [c0043958] irq_exit+0xa4/0xc8
[effe7ff0] [c000d14c] call_do_irq+0x24/0x3c
[c0875e90] [c00048a0] do_IRQ+0x8c/0xf8
[c0875eb0] [c000ed10] ret_from_except+0x0/0x18
For TX, we need to unmap the pages which has already been mapped and
free the skb before return.
For RX, move the dma mapping and error check to gfar_new_skb(). We
would reuse the original skb in the rx ring when either allocating
skb failure or dma mapping error.
Signed-off-by: Kevin Hao <haokexin@gmail.com>
Signed-off-by: Claudiu Manoil <claudiu.manoil@freescale.com>
---
v2: Just update the RX path to reuse the original skb when dma mapping error
occurs as suggested by David.
drivers/net/ethernet/freescale/gianfar.c | 84 +++++++++++++++++++++-----------
1 file changed, 56 insertions(+), 28 deletions(-)
diff --git a/drivers/net/ethernet/freescale/gianfar.c b/drivers/net/ethernet/freescale/gianfar.c
index 4fdf0aa16978..73e1144f2a87 100644
--- a/drivers/net/ethernet/freescale/gianfar.c
+++ b/drivers/net/ethernet/freescale/gianfar.c
@@ -116,9 +116,7 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev);
static void gfar_reset_task(struct work_struct *work);
static void gfar_timeout(struct net_device *dev);
static int gfar_close(struct net_device *dev);
-struct sk_buff *gfar_new_skb(struct net_device *dev);
-static void gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
- struct sk_buff *skb);
+struct sk_buff *gfar_new_skb(struct net_device *dev, dma_addr_t *bufaddr);
static int gfar_set_mac_address(struct net_device *dev);
static int gfar_change_mtu(struct net_device *dev, int new_mtu);
static irqreturn_t gfar_error(int irq, void *dev_id);
@@ -178,6 +176,7 @@ static int gfar_init_bds(struct net_device *ndev)
struct txbd8 *txbdp;
struct rxbd8 *rxbdp;
int i, j;
+ dma_addr_t bufaddr;
for (i = 0; i < priv->num_tx_queues; i++) {
tx_queue = priv->tx_queue[i];
@@ -211,19 +210,17 @@ static int gfar_init_bds(struct net_device *ndev)
struct sk_buff *skb = rx_queue->rx_skbuff[j];
if (skb) {
- gfar_init_rxbdp(rx_queue, rxbdp,
- rxbdp->bufPtr);
+ bufaddr = rxbdp->bufPtr;
} else {
- skb = gfar_new_skb(ndev);
+ skb = gfar_new_skb(ndev, &bufaddr);
if (!skb) {
netdev_err(ndev, "Can't allocate RX buffers\n");
return -ENOMEM;
}
rx_queue->rx_skbuff[j] = skb;
-
- gfar_new_rxbdp(rx_queue, rxbdp, skb);
}
+ gfar_init_rxbdp(rx_queue, rxbdp, bufaddr);
rxbdp++;
}
@@ -2290,6 +2287,8 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev)
0,
frag_len,
DMA_TO_DEVICE);
+ if (unlikely(dma_mapping_error(priv->dev, bufaddr)))
+ goto dma_map_err;
/* set the TxBD length and buffer pointer */
txbdp->bufPtr = bufaddr;
@@ -2339,8 +2338,12 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev)
fcb->ptp = 1;
}
- txbdp_start->bufPtr = dma_map_single(priv->dev, skb->data,
- skb_headlen(skb), DMA_TO_DEVICE);
+ bufaddr = dma_map_single(priv->dev, skb->data, skb_headlen(skb),
+ DMA_TO_DEVICE);
+ if (unlikely(dma_mapping_error(priv->dev, bufaddr)))
+ goto dma_map_err;
+
+ txbdp_start->bufPtr = bufaddr;
/* If time stamping is requested one additional TxBD must be set up. The
* first TxBD points to the FCB and must have a data length of
@@ -2406,6 +2409,25 @@ static int gfar_start_xmit(struct sk_buff *skb, struct net_device *dev)
spin_unlock_irqrestore(&tx_queue->txlock, flags);
return NETDEV_TX_OK;
+
+dma_map_err:
+ txbdp = next_txbd(txbdp_start, base, tx_queue->tx_ring_size);
+ if (do_tstamp)
+ txbdp = next_txbd(txbdp, base, tx_queue->tx_ring_size);
+ for (i = 0; i < nr_frags; i++) {
+ lstatus = txbdp->lstatus;
+ if (!(lstatus & BD_LFLAG(TXBD_READY)))
+ break;
+
+ txbdp->lstatus = lstatus & ~BD_LFLAG(TXBD_READY);
+ bufaddr = txbdp->bufPtr;
+ dma_unmap_page(priv->dev, bufaddr, txbdp->length,
+ DMA_TO_DEVICE);
+ txbdp = next_txbd(txbdp, base, tx_queue->tx_ring_size);
+ }
+ gfar_wmb();
+ dev_kfree_skb_any(skb);
+ return NETDEV_TX_OK;
}
/* Stops the kernel queue, and halts the controller */
@@ -2606,18 +2628,6 @@ static void gfar_clean_tx_ring(struct gfar_priv_tx_q *tx_queue)
netdev_tx_completed_queue(txq, howmany, bytes_sent);
}
-static void gfar_new_rxbdp(struct gfar_priv_rx_q *rx_queue, struct rxbd8 *bdp,
- struct sk_buff *skb)
-{
- struct net_device *dev = rx_queue->dev;
- struct gfar_private *priv = netdev_priv(dev);
- dma_addr_t buf;
-
- buf = dma_map_single(priv->dev, skb->data,
- priv->rx_buffer_size, DMA_FROM_DEVICE);
- gfar_init_rxbdp(rx_queue, bdp, buf);
-}
-
static struct sk_buff *gfar_alloc_skb(struct net_device *dev)
{
struct gfar_private *priv = netdev_priv(dev);
@@ -2632,9 +2642,25 @@ static struct sk_buff *gfar_alloc_skb(struct net_device *dev)
return skb;
}
-struct sk_buff *gfar_new_skb(struct net_device *dev)
+struct sk_buff *gfar_new_skb(struct net_device *dev, dma_addr_t *bufaddr)
{
- return gfar_alloc_skb(dev);
+ struct gfar_private *priv = netdev_priv(dev);
+ struct sk_buff *skb;
+ dma_addr_t addr;
+
+ skb = gfar_alloc_skb(dev);
+ if (!skb)
+ return NULL;
+
+ addr = dma_map_single(priv->dev, skb->data,
+ priv->rx_buffer_size, DMA_FROM_DEVICE);
+ if (unlikely(dma_mapping_error(priv->dev, addr))) {
+ dev_kfree_skb_any(skb);
+ return NULL;
+ }
+
+ *bufaddr = addr;
+ return skb;
}
static inline void count_errors(unsigned short status, struct net_device *dev)
@@ -2805,11 +2831,12 @@ int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue, int rx_work_limit)
while (!((bdp->status & RXBD_EMPTY) || (--rx_work_limit < 0))) {
struct sk_buff *newskb;
+ dma_addr_t bufaddr;
rmb();
/* Add another skb for the future */
- newskb = gfar_new_skb(dev);
+ newskb = gfar_new_skb(dev, &bufaddr);
skb = rx_queue->rx_skbuff[rx_queue->skb_currx];
@@ -2825,9 +2852,10 @@ int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue, int rx_work_limit)
bdp->status & RXBD_ERR)) {
count_errors(bdp->status, dev);
- if (unlikely(!newskb))
+ if (unlikely(!newskb)) {
newskb = skb;
- else if (skb)
+ bufaddr = bdp->bufPtr;
+ } else if (skb)
dev_kfree_skb(skb);
} else {
/* Increment the number of packets */
@@ -2854,7 +2882,7 @@ int gfar_clean_rx_ring(struct gfar_priv_rx_q *rx_queue, int rx_work_limit)
rx_queue->rx_skbuff[rx_queue->skb_currx] = newskb;
/* Setup the new bdp */
- gfar_new_rxbdp(rx_queue, bdp, newskb);
+ gfar_init_rxbdp(rx_queue, bdp, bufaddr);
/* Update to the next pointer */
bdp = next_bd(bdp, base, rx_queue->rx_ring_size);
--
1.9.3
^ permalink raw reply related
* [PATCH] Driver: Vmxnet3: Make Rx ring 2 size configurable
From: Shrikrishna Khare @ 2014-12-11 5:37 UTC (permalink / raw)
To: sbhatewara, pv-drivers, netdev, linux-kernel
Cc: Shrikrishna Khare, Ramya Bolla
Rx ring 2 size can be configured by adjusting rx-jumbo parameter
of ethtool -G.
Signed-off-by: Ramya Bolla <bollar@vmware.com>
Signed-off-by: Shreyas Bhatewara <sbhatewara@vmware.com>
Signed-off-by: Shrikrishna Khare <skhare@vmware.com>
---
drivers/net/vmxnet3/vmxnet3_defs.h | 1 +
drivers/net/vmxnet3/vmxnet3_drv.c | 6 +++++-
drivers/net/vmxnet3/vmxnet3_ethtool.c | 27 ++++++++++++++++++++-------
drivers/net/vmxnet3/vmxnet3_int.h | 6 ++++--
4 files changed, 30 insertions(+), 10 deletions(-)
diff --git a/drivers/net/vmxnet3/vmxnet3_defs.h b/drivers/net/vmxnet3/vmxnet3_defs.h
index 4d84912..25b6fa4 100644
--- a/drivers/net/vmxnet3/vmxnet3_defs.h
+++ b/drivers/net/vmxnet3/vmxnet3_defs.h
@@ -342,6 +342,7 @@ union Vmxnet3_GenericDesc {
#define VMXNET3_TX_RING_MAX_SIZE 4096
#define VMXNET3_TC_RING_MAX_SIZE 4096
#define VMXNET3_RX_RING_MAX_SIZE 4096
+#define VMXNET3_RX_RING2_MAX_SIZE 2048
#define VMXNET3_RC_RING_MAX_SIZE 8192
/* a list of reasons for queue stop */
diff --git a/drivers/net/vmxnet3/vmxnet3_drv.c b/drivers/net/vmxnet3/vmxnet3_drv.c
index afd2953..7af1f5c 100644
--- a/drivers/net/vmxnet3/vmxnet3_drv.c
+++ b/drivers/net/vmxnet3/vmxnet3_drv.c
@@ -2505,6 +2505,9 @@ vmxnet3_adjust_rx_ring_size(struct vmxnet3_adapter *adapter)
ring0_size = min_t(u32, ring0_size, VMXNET3_RX_RING_MAX_SIZE /
sz * sz);
ring1_size = adapter->rx_queue[0].rx_ring[1].size;
+ ring1_size = (ring1_size + sz - 1) / sz * sz;
+ ring1_size = min_t(u32, ring1_size, VMXNET3_RX_RING2_MAX_SIZE /
+ sz * sz);
comp_size = ring0_size + ring1_size;
for (i = 0; i < adapter->num_rx_queues; i++) {
@@ -2585,7 +2588,7 @@ vmxnet3_open(struct net_device *netdev)
err = vmxnet3_create_queues(adapter, adapter->tx_ring_size,
adapter->rx_ring_size,
- VMXNET3_DEF_RX_RING_SIZE);
+ adapter->rx_ring2_size);
if (err)
goto queue_err;
@@ -2964,6 +2967,7 @@ vmxnet3_probe_device(struct pci_dev *pdev,
adapter->tx_ring_size = VMXNET3_DEF_TX_RING_SIZE;
adapter->rx_ring_size = VMXNET3_DEF_RX_RING_SIZE;
+ adapter->rx_ring2_size = VMXNET3_DEF_RX_RING2_SIZE;
spin_lock_init(&adapter->cmd_lock);
adapter->adapter_pa = dma_map_single(&adapter->pdev->dev, adapter,
diff --git a/drivers/net/vmxnet3/vmxnet3_ethtool.c b/drivers/net/vmxnet3/vmxnet3_ethtool.c
index b725fd9..70fb410 100644
--- a/drivers/net/vmxnet3/vmxnet3_ethtool.c
+++ b/drivers/net/vmxnet3/vmxnet3_ethtool.c
@@ -447,12 +447,12 @@ vmxnet3_get_ringparam(struct net_device *netdev,
param->rx_max_pending = VMXNET3_RX_RING_MAX_SIZE;
param->tx_max_pending = VMXNET3_TX_RING_MAX_SIZE;
param->rx_mini_max_pending = 0;
- param->rx_jumbo_max_pending = 0;
+ param->rx_jumbo_max_pending = VMXNET3_RX_RING2_MAX_SIZE;
param->rx_pending = adapter->rx_ring_size;
param->tx_pending = adapter->tx_ring_size;
param->rx_mini_pending = 0;
- param->rx_jumbo_pending = 0;
+ param->rx_jumbo_pending = adapter->rx_ring2_size;
}
@@ -461,7 +461,7 @@ vmxnet3_set_ringparam(struct net_device *netdev,
struct ethtool_ringparam *param)
{
struct vmxnet3_adapter *adapter = netdev_priv(netdev);
- u32 new_tx_ring_size, new_rx_ring_size;
+ u32 new_tx_ring_size, new_rx_ring_size, new_rx_ring2_size;
u32 sz;
int err = 0;
@@ -473,6 +473,10 @@ vmxnet3_set_ringparam(struct net_device *netdev,
VMXNET3_RX_RING_MAX_SIZE)
return -EINVAL;
+ if (param->rx_jumbo_pending == 0 ||
+ param->rx_jumbo_pending > VMXNET3_RX_RING2_MAX_SIZE)
+ return -EINVAL;
+
/* if adapter not yet initialized, do nothing */
if (adapter->rx_buf_per_pkt == 0) {
netdev_err(netdev, "adapter not completely initialized, "
@@ -500,8 +504,15 @@ vmxnet3_set_ringparam(struct net_device *netdev,
sz) != 0)
return -EINVAL;
- if (new_tx_ring_size == adapter->tx_queue[0].tx_ring.size &&
- new_rx_ring_size == adapter->rx_queue[0].rx_ring[0].size) {
+ /* ring2 has to be a multiple of VMXNET3_RING_SIZE_ALIGN */
+ new_rx_ring2_size = (param->rx_jumbo_pending + VMXNET3_RING_SIZE_MASK) &
+ ~VMXNET3_RING_SIZE_MASK;
+ new_rx_ring2_size = min_t(u32, new_rx_ring2_size,
+ VMXNET3_RX_RING2_MAX_SIZE);
+
+ if (new_tx_ring_size == adapter->tx_ring_size &&
+ new_rx_ring_size == adapter->rx_ring_size &&
+ new_rx_ring2_size == adapter->rx_ring2_size) {
return 0;
}
@@ -522,7 +533,7 @@ vmxnet3_set_ringparam(struct net_device *netdev,
vmxnet3_rq_destroy_all(adapter);
err = vmxnet3_create_queues(adapter, new_tx_ring_size,
- new_rx_ring_size, VMXNET3_DEF_RX_RING_SIZE);
+ new_rx_ring_size, new_rx_ring2_size);
if (err) {
/* failed, most likely because of OOM, try default
@@ -530,11 +541,12 @@ vmxnet3_set_ringparam(struct net_device *netdev,
netdev_err(netdev, "failed to apply new sizes, "
"try the default ones\n");
new_rx_ring_size = VMXNET3_DEF_RX_RING_SIZE;
+ new_rx_ring2_size = VMXNET3_DEF_RX_RING2_SIZE;
new_tx_ring_size = VMXNET3_DEF_TX_RING_SIZE;
err = vmxnet3_create_queues(adapter,
new_tx_ring_size,
new_rx_ring_size,
- VMXNET3_DEF_RX_RING_SIZE);
+ new_rx_ring2_size);
if (err) {
netdev_err(netdev, "failed to create queues "
"with default sizes. Closing it\n");
@@ -549,6 +561,7 @@ vmxnet3_set_ringparam(struct net_device *netdev,
}
adapter->tx_ring_size = new_tx_ring_size;
adapter->rx_ring_size = new_rx_ring_size;
+ adapter->rx_ring2_size = new_rx_ring2_size;
out:
clear_bit(VMXNET3_STATE_BIT_RESETTING, &adapter->state);
diff --git a/drivers/net/vmxnet3/vmxnet3_int.h b/drivers/net/vmxnet3/vmxnet3_int.h
index 5f0199f..048f020 100644
--- a/drivers/net/vmxnet3/vmxnet3_int.h
+++ b/drivers/net/vmxnet3/vmxnet3_int.h
@@ -69,10 +69,10 @@
/*
* Version numbers
*/
-#define VMXNET3_DRIVER_VERSION_STRING "1.2.1.0-k"
+#define VMXNET3_DRIVER_VERSION_STRING "1.3.1.0-k"
/* a 32-bit int, each byte encode a verion number in VMXNET3_DRIVER_VERSION */
-#define VMXNET3_DRIVER_VERSION_NUM 0x01020100
+#define VMXNET3_DRIVER_VERSION_NUM 0x01030100
#if defined(CONFIG_PCI_MSI)
/* RSS only makes sense if MSI-X is supported. */
@@ -352,6 +352,7 @@ struct vmxnet3_adapter {
/* Ring sizes */
u32 tx_ring_size;
u32 rx_ring_size;
+ u32 rx_ring2_size;
struct work_struct work;
@@ -384,6 +385,7 @@ struct vmxnet3_adapter {
/* must be a multiple of VMXNET3_RING_SIZE_ALIGN */
#define VMXNET3_DEF_TX_RING_SIZE 512
#define VMXNET3_DEF_RX_RING_SIZE 256
+#define VMXNET3_DEF_RX_RING2_SIZE 128
#define VMXNET3_MAX_ETH_HDR_SIZE 22
#define VMXNET3_MAX_SKB_BUF_SIZE (3*1024)
--
2.1.0
^ permalink raw reply related
* [PATCH] cxgb4/csiostor: Don't use MASTER_MUST for fw_hello call
From: Hariprasad Shenai @ 2014-12-11 5:41 UTC (permalink / raw)
To: netdev, linux-scsi
Cc: davem, JBottomley, hch, leedom, anish, nirranjan, praveenm,
Hariprasad Shenai
Remove use of calls into t4_fw_hello() with MASTER_MUST, which results in
FW_HELLO_CMD_MASTERFORCE being set. The firmware doesn't support this and of
course any existing PF Drivers will totally go for a toss.
Signed-off-by: Hariprasad Shenai <hariprasad@chelsio.com>
---
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 12 ++----------
drivers/scsi/csiostor/csio_hw.c | 6 +-----
drivers/scsi/csiostor/csio_hw.h | 1 -
3 files changed, 3 insertions(+), 16 deletions(-)
diff --git a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
index 279873c..30deaf1 100644
--- a/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
+++ b/drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c
@@ -5687,14 +5687,8 @@ static int adap_init0(struct adapter *adap)
struct fw_caps_config_cmd caps_cmd;
int reset = 1;
- /*
- * Contact FW, advertising Master capability (and potentially forcing
- * ourselves as the Master PF if our module parameter force_init is
- * set).
- */
- ret = t4_fw_hello(adap, adap->mbox, adap->fn,
- force_init ? MASTER_MUST : MASTER_MAY,
- &state);
+ /* Contact FW, advertising Master capability */
+ ret = t4_fw_hello(adap, adap->mbox, adap->mbox, MASTER_MAY, &state);
if (ret < 0) {
dev_err(adap->pdev_dev, "could not connect to FW, error %d\n",
ret);
@@ -5702,8 +5696,6 @@ static int adap_init0(struct adapter *adap)
}
if (ret == adap->mbox)
adap->flags |= MASTER_PF;
- if (force_init && state == DEV_STATE_INIT)
- state = DEV_STATE_UNINIT;
/*
* If we're the Master PF Driver and the device is uninitialized,
diff --git a/drivers/scsi/csiostor/csio_hw.c b/drivers/scsi/csiostor/csio_hw.c
index 0eaec47..5b909c0 100644
--- a/drivers/scsi/csiostor/csio_hw.c
+++ b/drivers/scsi/csiostor/csio_hw.c
@@ -47,7 +47,6 @@
#include "csio_lnode.h"
#include "csio_rnode.h"
-int csio_force_master;
int csio_dbg_level = 0xFEFF;
unsigned int csio_port_mask = 0xf;
@@ -889,7 +888,6 @@ csio_do_hello(struct csio_hw *hw, enum csio_dev_state *state)
{
struct csio_mb *mbp;
int rv = 0;
- enum csio_dev_master master;
enum fw_retval retval;
uint8_t mpfn;
char state_str[16];
@@ -904,11 +902,9 @@ csio_do_hello(struct csio_hw *hw, enum csio_dev_state *state)
goto out;
}
- master = csio_force_master ? CSIO_MASTER_MUST : CSIO_MASTER_MAY;
-
retry:
csio_mb_hello(hw, mbp, CSIO_MB_DEFAULT_TMO, hw->pfn,
- hw->pfn, master, NULL);
+ hw->pfn, CSIO_MASTER_MAY, NULL);
rv = csio_mb_issue(hw, mbp);
if (rv) {
diff --git a/drivers/scsi/csiostor/csio_hw.h b/drivers/scsi/csiostor/csio_hw.h
index 5db2d85..68248da 100644
--- a/drivers/scsi/csiostor/csio_hw.h
+++ b/drivers/scsi/csiostor/csio_hw.h
@@ -110,7 +110,6 @@ struct csio_scsi_cpu_info {
};
extern int csio_dbg_level;
-extern int csio_force_master;
extern unsigned int csio_port_mask;
extern int csio_msi;
--
1.7.1
^ permalink raw reply related
* Re: [PATCH v7 0/4] arch: Add lightweight memory barriers for coherent memory access
From: Alexander Duyck @ 2014-12-11 5:28 UTC (permalink / raw)
To: Alexander Duyck, linux-arch, netdev, linux-kernel, arnd, davem
Cc: mathieu.desnoyers, peterz, benh, heiko.carstens, mingo, mikey,
linux, donald.c.skidmore, matthew.vick, geert, jeffrey.t.kirsher,
romieu, paulmck, nic_swsd, will.deacon, michael, tony.luck,
torvalds, oleg, schwidefsky, fweisbec
In-Reply-To: <20141125203310.8240.27370.stgit@ahduyck-server>
On 11/25/2014 12:35 PM, Alexander Duyck wrote:
> These patches introduce two new primitives for synchronizing cache coherent
> memory writes and reads. These two new primitives are:
>
> dma_rmb()
> dma_wmb()
>
> The first patch cleans up some unnecessary overhead related to the
> definition of read_barrier_depends, smp_read_barrier_depends, and comments
> related to the barrier.
>
> The second patch adds the primitives for the applicable architectures and
> asm-generic.
>
> The third patch adds the barriers to r8169 which turns out to be a good
> example of where the new barriers might be useful as they have full
> rmb()/wmb() barriers ordering accesses to the descriptors and the DescOwn
> bit.
>
> The fourth patch adds support for coherent_rmb() to the Intel fm10k, igb,
> and ixgbe drivers. Testing with the ixgbe driver has shown a processing
> time reduction of at least 7ns per 64B frame on a Core i7-4930K.
>
> This patch series is essentially the v7 for:
> v4-6: Add lightweight memory barriers for coherent memory access
> v3: Add lightweight memory barriers fast_rmb() and fast_wmb()
> v2: Introduce load_acquire() and store_release()
> v1: Introduce read_acquire()
>
> The key changes in this patch series versus the earlier patches are:
> v7:
> - Dropped test/debug patch that was accidentally slipped in
> v6:
> - Replaced "memory based device I/O" with "consistent memory" in
> docs
> - Added reference to DMA-API.txt to explain consistent memory
> v5:
> - Renamed barriers dma_rmb and dma_wmb
> - Undid smp_wmb changes in x86 and PowerPC
> - Defined smp_rmb as __lwsync for SMP case on PowerPC
> v4:
> - Renamed barriers coherent_rmb and coherent_wmb
> - Added smp_lwsync for use in smp_load_acquire/smp_store_release
> v3:
> - Moved away from acquire()/store() and instead focused on barriers
> - Added cleanup of read_barrier_depends
> - Added change in r8169 to fix cur_tx/DescOwn ordering
> - Simplified changes to just replacing/moving barriers in r8169
> - Added update to documentation with code example
> v2:
> - Renamed read_acquire() to be consistent with smp_load_acquire()
> - Changed barrier used to be consistent with smp_load_acquire()
> - Updated PowerPC code to use __lwsync based on IBM article
> - Added store_release() as this is a viable use case for drivers
> - Added r8169 patch which is able to fully use primitives
> - Added fm10k/igb/ixgbe patch which is able to test performance
>
> ---
>
> Alexander Duyck (4):
> arch: Cleanup read_barrier_depends() and comments
> arch: Add lightweight memory barriers dma_rmb() and dma_wmb()
> r8169: Use dma_rmb() and dma_wmb() for DescOwn checks
> fm10k/igb/ixgbe: Use dma_rmb on Rx descriptor reads
>
>
> Documentation/memory-barriers.txt | 42 +++++++++++++++
> arch/alpha/include/asm/barrier.h | 51 ++++++++++++++++++
> arch/arm/include/asm/barrier.h | 4 +
> arch/arm64/include/asm/barrier.h | 3 +
> arch/blackfin/include/asm/barrier.h | 51 ++++++++++++++++++
> arch/ia64/include/asm/barrier.h | 25 ++++-----
> arch/metag/include/asm/barrier.h | 19 ++++---
> arch/mips/include/asm/barrier.h | 61 ++--------------------
> arch/powerpc/include/asm/barrier.h | 19 ++++---
> arch/s390/include/asm/barrier.h | 7 ++-
> arch/sparc/include/asm/barrier_64.h | 7 ++-
> arch/x86/include/asm/barrier.h | 70 ++++---------------------
> arch/x86/um/asm/barrier.h | 20 ++++---
> drivers/net/ethernet/intel/fm10k/fm10k_main.c | 6 +-
> drivers/net/ethernet/intel/igb/igb_main.c | 6 +-
> drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 9 +--
> drivers/net/ethernet/realtek/r8169.c | 29 ++++++++--
> include/asm-generic/barrier.h | 8 +++
> 18 files changed, 258 insertions(+), 179 deletions(-)
>
> --
It occurs to me that I never got a sign off from any of the maintainers
on getting this pulled in.
Since the merge window is open I was wondering which tree I should make
sure these patches apply to and who will be the one to pull them in?
Since I was modifying network drivers should I resubmit them for netdev,
or should I submit them for asm-generic or some other tree?
- Alex
^ permalink raw reply
* Re: [PATCH net-next 0/3] Kill arch_fast_hash
From: David Miller @ 2014-12-11 4:37 UTC (permalink / raw)
To: dborkman; +Cc: netdev, tgraf, hannes
In-Reply-To: <5488D3F5.8040701@redhat.com>
From: Daniel Borkmann <dborkman@redhat.com>
Date: Thu, 11 Dec 2014 00:15:01 +0100
> On 12/10/2014 09:18 PM, David Miller wrote:
> ...
>> Series applied, and I'll queue patch #1 to -stable, thanks.
>
> Sounds good, thanks!
>
> Btw, this will create a "silent" merge conflict with Linus' tree:
> As nios2 arch has been merged for 3.19, the 'generic-y += hash.h'
> from arch/nios2/include/asm/Kbuild needs to be removed there, too.
Thanks for letting me know.
^ permalink raw reply
* Re: [PATCH net-next v2 0/3] net: fec: driver code clean and bug fix
From: David Miller @ 2014-12-11 4:37 UTC (permalink / raw)
To: b38611; +Cc: netdev, bhutchings, stephen
In-Reply-To: <1418260833-896-1-git-send-email-b38611@freescale.com>
From: Fugang Duan <b38611@freescale.com>
Date: Thu, 11 Dec 2014 09:20:30 +0800
> The patch serial include code clean and bug fix:
> Patch#1: avoid dummy operation during suspend/resume test.
> Patch#2: bug fix for i.MX6SX SOC that clean all interrupt events during MAC initial process.
> Patch#3: before phy device link status is up, only enable MDIO bus interrupt.
>
> V2:
> - Modify the comment form from David's suggestion.
Series applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next] net: sock: fix access via invalid file descriptor
From: David Miller @ 2014-12-11 4:34 UTC (permalink / raw)
To: ast; +Cc: fengguang.wu, davej, dborkman, netdev, linux-kernel
In-Reply-To: <1418271295-5829-1-git-send-email-ast@plumgrid.com>
From: Alexei Starovoitov <ast@plumgrid.com>
Date: Wed, 10 Dec 2014 20:14:55 -0800
> 0day robot reported the following crash:
> [ 21.233581] BUG: unable to handle kernel NULL pointer dereference at 0000000000000007
> [ 21.234709] IP: [<ffffffff8156ebda>] sk_attach_bpf+0x39/0xc2
>
> It's due to bpf_prog_get() returning ERR_PTR.
> Check it properly.
>
> Reported-by: Fengguang Wu <fengguang.wu@intel.com>
> Fixes: 89aa075832b0 ("net: sock: allow eBPF programs to be attached to sockets")
> Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
> ---
> Silly mistake. I was sure I've checked this error path. Apparently not :(
Applied, thanks Alexei.
^ permalink raw reply
* Re: [PATCH net-next RESEND] net: Do not call ndo_dflt_fdb_dump if ndo_fdb_dump is defined.
From: David Miller @ 2014-12-11 4:32 UTC (permalink / raw)
To: h.sokolowski; +Cc: netdev, jhs
In-Reply-To: <c6756bd161048ac4b4407e308045fe73.squirrel@poczta.wsisiz.edu.pl>
From: "Hubert Sokolowski" <h.sokolowski@wit.edu.pl>
Date: Wed, 10 Dec 2014 19:37:01 -0000
> This change restores the semantic that was present
> before 5e6d243587990a588143b9da3974833649595587
> "bridge: netlink dump interface at par with brctl"
> on how ndo_dflt_fdb_dump is called.
> This semantic is still used for add and del operations
> so let's keep it consistent.
> Driver can still call ndo_dflt_fdb_dump from inside
> its own fdb_dump routine when needed.
>
> Signed-off-by: Hubert Sokolowski <h.sokolowski@wit.edu.pl>
Jamal, please review.
> ---
> net/core/rtnetlink.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c
> index eaa057f..a9e5c37 100644
> --- a/net/core/rtnetlink.c
> +++ b/net/core/rtnetlink.c
> @@ -2692,10 +2692,11 @@ static int rtnl_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb)
> idx);
> }
>
> - idx = ndo_dflt_fdb_dump(skb, cb, dev, NULL, idx);
> if (dev->netdev_ops->ndo_fdb_dump)
> idx = dev->netdev_ops->ndo_fdb_dump(skb, cb, bdev, dev,
> idx);
> + else
> + idx = ndo_dflt_fdb_dump(skb, cb, dev, NULL, idx);
>
> cops = NULL;
> }
> --
> 1.9.3
>
>
>
>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* Re: [PATCH net-next v2 1/4] netdev: introduce new NETIF_F_HW_NETFUNC_OFFLOAD feature flag for switch device offloads
From: Roopa Prabhu @ 2014-12-11 4:31 UTC (permalink / raw)
To: Jiri Pirko
Cc: sfeldma, jhs, bcrl, tgraf, john.fastabend, stephen, linville,
vyasevic, netdev, davem, shm, gospo
In-Reply-To: <20141210093150.GA1863@nanopsycho.orion>
On 12/10/14, 1:31 AM, Jiri Pirko wrote:
> Wed, Dec 10, 2014 at 10:05:17AM CET, roopa@cumulusnetworks.com wrote:
>> From: Roopa Prabhu <roopa@cumulusnetworks.com>
>>
>> This is a high level feature flag for all switch asic offloads
>>
>> switch drivers set this flag on switch ports. Logical devices like
>> bridge, bonds, vxlans can inherit this flag from their slaves/ports.
>
> Interesting thing is that you are talking about "switch" (mentioned 3
> times on 3 lines), yet the name of bit is "NETFUNC_OFFLOAD". I'm sorry,
> I might be missing some obvious facts but "netfunc" seems ambiguous
> to me.
scotts and your series and mine too is being introduced for switch asics
right now.
And we are all only targeting l2 today. But, Keeping future in mind, the
names we choose today
should be applicable for all other features and also other similar devices.
Even the 'switch' in the api names is making it not fit well in some cases.
I am not hung up on 'network function', if there are other suggestions I
will be happy to take.
Thanks,
Roopa
>
>
>> The patch also adds the flag to NETIF_F_ONE_FOR_ALL, so that it gets
>> propagated to the upperdevices (bridges and bonds).
>>
>> (In case you did not notice, I am trying a new name NETIF_F_HW_NETFUNC_OFFLOAD for the flag ;)
>>
>> Signed-off-by: Roopa Prabhu <roopa@cumulusnetworks.com>
>> ---
>> include/linux/netdev_features.h | 5 ++++-
>> 1 file changed, 4 insertions(+), 1 deletion(-)
>>
>> diff --git a/include/linux/netdev_features.h b/include/linux/netdev_features.h
>> index 8e30685..8ab175b 100644
>> --- a/include/linux/netdev_features.h
>> +++ b/include/linux/netdev_features.h
>> @@ -66,6 +66,7 @@ enum {
>> NETIF_F_HW_VLAN_STAG_FILTER_BIT,/* Receive filtering on VLAN STAGs */
>> NETIF_F_HW_L2FW_DOFFLOAD_BIT, /* Allow L2 Forwarding in Hardware */
>> NETIF_F_BUSY_POLL_BIT, /* Busy poll */
>> + NETIF_F_HW_NETFUNC_OFFLOAD_BIT, /* HW switch offload */
>>
>> /*
>> * Add your fresh new feature above and remember to update
>> @@ -124,6 +125,7 @@ enum {
>> #define NETIF_F_HW_VLAN_STAG_TX __NETIF_F(HW_VLAN_STAG_TX)
>> #define NETIF_F_HW_L2FW_DOFFLOAD __NETIF_F(HW_L2FW_DOFFLOAD)
>> #define NETIF_F_BUSY_POLL __NETIF_F(BUSY_POLL)
>> +#define NETIF_F_HW_NETFUNC_OFFLOAD __NETIF_F(HW_NETFUNC_OFFLOAD)
>>
>> /* Features valid for ethtool to change */
>> /* = all defined minus driver/device-class-related */
>> @@ -159,7 +161,8 @@ enum {
>> */
>> #define NETIF_F_ONE_FOR_ALL (NETIF_F_GSO_SOFTWARE | NETIF_F_GSO_ROBUST | \
>> NETIF_F_SG | NETIF_F_HIGHDMA | \
>> - NETIF_F_FRAGLIST | NETIF_F_VLAN_CHALLENGED)
>> + NETIF_F_FRAGLIST | NETIF_F_VLAN_CHALLENGED | \
>> + NETIF_F_HW_NETFUNC_OFFLOAD)
>> /*
>> * If one device doesn't support one of these features, then disable it
>> * for all in netdev_increment_features.
>> --
>> 1.7.10.4
>>
> --
> To unsubscribe from this list: send the line "unsubscribe netdev" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply
* [PATCH net-next] net: sock: fix access via invalid file descriptor
From: Alexei Starovoitov @ 2014-12-11 4:14 UTC (permalink / raw)
To: David S. Miller
Cc: Fengguang Wu, Dave Jones, Daniel Borkmann, netdev, linux-kernel
0day robot reported the following crash:
[ 21.233581] BUG: unable to handle kernel NULL pointer dereference at 0000000000000007
[ 21.234709] IP: [<ffffffff8156ebda>] sk_attach_bpf+0x39/0xc2
It's due to bpf_prog_get() returning ERR_PTR.
Check it properly.
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Fixes: 89aa075832b0 ("net: sock: allow eBPF programs to be attached to sockets")
Signed-off-by: Alexei Starovoitov <ast@plumgrid.com>
---
Silly mistake. I was sure I've checked this error path. Apparently not :(
net/core/filter.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/core/filter.c b/net/core/filter.c
index 8cc3c03078b3..ec9baea10c16 100644
--- a/net/core/filter.c
+++ b/net/core/filter.c
@@ -1103,8 +1103,8 @@ int sk_attach_bpf(u32 ufd, struct sock *sk)
return -EPERM;
prog = bpf_prog_get(ufd);
- if (!prog)
- return -EINVAL;
+ if (IS_ERR(prog))
+ return PTR_ERR(prog);
if (prog->aux->prog_type != BPF_PROG_TYPE_SOCKET_FILTER) {
/* valid fd, but invalid program type */
--
1.7.9.5
^ permalink raw reply related
* Re: [PATCH net v8 1/7] cxgb4i: fix tx immediate data credit check
From: David Miller @ 2014-12-11 3:54 UTC (permalink / raw)
To: kxie; +Cc: linux-scsi, netdev, hariprasad, anish, hch, James.Bottomley,
michaelc
In-Reply-To: <201412102111.sBALB6sI011656@localhost6.localdomain6>
From: Karen Xie <kxie@chelsio.com>
Date: Wed, 10 Dec 2014 13:11:06 -0800
> From: Karen Xie <kxie@chelsio.com>
>
> Only data skbs need the wr header added while control skbs do not. Make sure they are treated differently.
>
> Signed-off-by: Karen Xie <kxie@chelsio.com>
Please text format your commit messages to 80 characters or less per
line.
People are going to read your commits using text tools, not
necessarily GUI tools which will auto-format your commit message.
I know this seems like we are being difficult, but you are making many
fundamental mistakes in your efforts to merge in these changes.
^ permalink raw reply
* Re: [PATCH v3] net: introduce helper macro for_each_cmsghdr
From: David Miller @ 2014-12-11 3:42 UTC (permalink / raw)
To: guz.fnst; +Cc: netdev, linux-kernel, joe
In-Reply-To: <54890DDC.3040402@cn.fujitsu.com>
From: Gu Zheng <guz.fnst@cn.fujitsu.com>
Date: Thu, 11 Dec 2014 11:22:04 +0800
> Introduce helper macro for_each_cmsghdr as a wrapper of the enumerating
> cmsghdr from msghdr, just cleanup.
>
> Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
> ---
> v3:
> -drop the changes about user-land programs as David suggested.
> v2:
> -use the lower-case macro name as Joe suggested.
Applied, thanks.
^ permalink raw reply
* [PATCH v3] net: introduce helper macro for_each_cmsghdr
From: Gu Zheng @ 2014-12-11 3:22 UTC (permalink / raw)
To: David S. Miller; +Cc: netdev, linux-kernel, Joe Perches
Introduce helper macro for_each_cmsghdr as a wrapper of the enumerating
cmsghdr from msghdr, just cleanup.
Signed-off-by: Gu Zheng <guz.fnst@cn.fujitsu.com>
---
v3:
-drop the changes about user-land programs as David suggested.
v2:
-use the lower-case macro name as Joe suggested.
crypto/af_alg.c | 2 +-
include/linux/socket.h | 4 ++++
net/core/scm.c | 3 +--
net/dccp/proto.c | 5 ++---
net/ipv4/ip_sockglue.c | 2 +-
net/ipv6/datagram.c | 2 +-
net/iucv/af_iucv.c | 4 +---
net/rds/send.c | 4 ++--
net/rxrpc/ar-output.c | 2 +-
net/sctp/socket.c | 3 +--
10 files changed, 15 insertions(+), 16 deletions(-)
diff --git a/crypto/af_alg.c b/crypto/af_alg.c
index 6a3ad80..bc21f52 100644
--- a/crypto/af_alg.c
+++ b/crypto/af_alg.c
@@ -399,7 +399,7 @@ int af_alg_cmsg_send(struct msghdr *msg, struct af_alg_control *con)
{
struct cmsghdr *cmsg;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
if (cmsg->cmsg_level != SOL_ALG)
diff --git a/include/linux/socket.h b/include/linux/socket.h
index bb9b836..0e71278 100644
--- a/include/linux/socket.h
+++ b/include/linux/socket.h
@@ -94,6 +94,10 @@ struct cmsghdr {
(cmsg)->cmsg_len <= (unsigned long) \
((mhdr)->msg_controllen - \
((char *)(cmsg) - (char *)(mhdr)->msg_control)))
+#define for_each_cmsghdr(cmsg, msg) \
+ for (cmsg = CMSG_FIRSTHDR(msg); \
+ cmsg; \
+ cmsg = CMSG_NXTHDR(msg, cmsg))
/*
* Get the next cmsg header
diff --git a/net/core/scm.c b/net/core/scm.c
index b442e7e..3b6899b 100644
--- a/net/core/scm.c
+++ b/net/core/scm.c
@@ -129,8 +129,7 @@ int __scm_send(struct socket *sock, struct msghdr *msg, struct scm_cookie *p)
struct cmsghdr *cmsg;
int err;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg))
- {
+ for_each_cmsghdr(cmsg, msg) {
err = -EINVAL;
/* Verify that cmsg_len is at least sizeof(struct cmsghdr) */
diff --git a/net/dccp/proto.c b/net/dccp/proto.c
index 5ab6627..e7413a9 100644
--- a/net/dccp/proto.c
+++ b/net/dccp/proto.c
@@ -703,7 +703,7 @@ EXPORT_SYMBOL_GPL(compat_dccp_getsockopt);
static int dccp_msghdr_parse(struct msghdr *msg, struct sk_buff *skb)
{
- struct cmsghdr *cmsg = CMSG_FIRSTHDR(msg);
+ struct cmsghdr *cmsg;
/*
* Assign an (opaque) qpolicy priority value to skb->priority.
@@ -717,8 +717,7 @@ static int dccp_msghdr_parse(struct msghdr *msg, struct sk_buff *skb)
*/
skb->priority = 0;
- for (; cmsg != NULL; cmsg = CMSG_NXTHDR(msg, cmsg)) {
-
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
diff --git a/net/ipv4/ip_sockglue.c b/net/ipv4/ip_sockglue.c
index 9daf217..839db9d 100644
--- a/net/ipv4/ip_sockglue.c
+++ b/net/ipv4/ip_sockglue.c
@@ -192,7 +192,7 @@ int ip_cmsg_send(struct net *net, struct msghdr *msg, struct ipcm_cookie *ipc,
int err, val;
struct cmsghdr *cmsg;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
#if IS_ENABLED(CONFIG_IPV6)
diff --git a/net/ipv6/datagram.c b/net/ipv6/datagram.c
index 2cdc383..7252965 100644
--- a/net/ipv6/datagram.c
+++ b/net/ipv6/datagram.c
@@ -640,7 +640,7 @@ int ip6_datagram_send_ctl(struct net *net, struct sock *sk,
int len;
int err = 0;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
int addr_type;
if (!CMSG_OK(msg, cmsg)) {
diff --git a/net/iucv/af_iucv.c b/net/iucv/af_iucv.c
index a089b6b..b69d87e 100644
--- a/net/iucv/af_iucv.c
+++ b/net/iucv/af_iucv.c
@@ -1070,9 +1070,7 @@ static int iucv_sock_sendmsg(struct kiocb *iocb, struct socket *sock,
txmsg.class = 0;
/* iterate over control messages */
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg;
- cmsg = CMSG_NXTHDR(msg, cmsg)) {
-
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg)) {
err = -EINVAL;
goto out;
diff --git a/net/rds/send.c b/net/rds/send.c
index 0a64541..b23e7b8 100644
--- a/net/rds/send.c
+++ b/net/rds/send.c
@@ -826,7 +826,7 @@ static int rds_rm_size(struct msghdr *msg, int data_len)
int cmsg_groups = 0;
int retval;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
@@ -878,7 +878,7 @@ static int rds_cmsg_send(struct rds_sock *rs, struct rds_message *rm,
struct cmsghdr *cmsg;
int ret = 0;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
diff --git a/net/rxrpc/ar-output.c b/net/rxrpc/ar-output.c
index 0b4b9a7..83b4616 100644
--- a/net/rxrpc/ar-output.c
+++ b/net/rxrpc/ar-output.c
@@ -45,7 +45,7 @@ static int rxrpc_sendmsg_cmsg(struct rxrpc_sock *rx, struct msghdr *msg,
if (msg->msg_controllen == 0)
return -EINVAL;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg; cmsg = CMSG_NXTHDR(msg, cmsg)) {
+ for_each_cmsghdr(cmsg, msg) {
if (!CMSG_OK(msg, cmsg))
return -EINVAL;
diff --git a/net/sctp/socket.c b/net/sctp/socket.c
index 634a2ab..a3802ed 100644
--- a/net/sctp/socket.c
+++ b/net/sctp/socket.c
@@ -6592,8 +6592,7 @@ static int sctp_msghdr_parse(const struct msghdr *msg, sctp_cmsgs_t *cmsgs)
struct cmsghdr *cmsg;
struct msghdr *my_msg = (struct msghdr *)msg;
- for (cmsg = CMSG_FIRSTHDR(msg); cmsg != NULL;
- cmsg = CMSG_NXTHDR(my_msg, cmsg)) {
+ for_each_cmsghdr(cmsg, my_msg) {
if (!CMSG_OK(my_msg, cmsg))
return -EINVAL;
--
1.7.7
^ permalink raw reply related
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox