* [PATCH V2 1/7] net/mlx4_en: Force user priority by QP attribute
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 9:07 ` [PATCH V2 2/7] net/mlx4_core: set port QoS attributes Amir Vadai
` (5 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Instead of relying on HW to change schedule queue by UP, schedule
queue is fixed for a tx_ring, and UP in WQE is ignored in this aspect. This
resolves two issues with untagged traffic:
1. untagged traffic has no UP in packet which is needed for QoS. The change
above allows setting the schedule queue (and by that the UP) of such a stream.
2. BlueFlame uses the same field used by vlan tag. So forcing UP from QPC
allows using BF for untagged but prioritized traffic.
In old firmware that force UP is not supported, untagged traffic will not subject to
QoS.
Because UP is set by QP, need to always have a tx ring per UP, even if pfcrx
module paramter is false.
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/en_main.c | 2 +-
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 3 ++-
drivers/net/ethernet/mellanox/mlx4/en_resources.c | 6 +++++-
drivers/net/ethernet/mellanox/mlx4/en_rx.c | 4 ++--
drivers/net/ethernet/mellanox/mlx4/en_tx.c | 12 ++++--------
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 6 +++---
include/linux/mlx4/qp.h | 3 ++-
7 files changed, 19 insertions(+), 17 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_main.c b/drivers/net/ethernet/mellanox/mlx4/en_main.c
index 2097a7d..346fdb2 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_main.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_main.c
@@ -114,7 +114,7 @@ static int mlx4_en_get_profile(struct mlx4_en_dev *mdev)
params->prof[i].tx_ring_size = MLX4_EN_DEF_TX_RING_SIZE;
params->prof[i].rx_ring_size = MLX4_EN_DEF_RX_RING_SIZE;
params->prof[i].tx_ring_num = MLX4_EN_NUM_TX_RINGS +
- (!!pfcrx) * MLX4_EN_NUM_PPP_RINGS;
+ MLX4_EN_NUM_PPP_RINGS;
params->prof[i].rss_rings = 0;
}
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 31b455a..2322622 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -650,7 +650,8 @@ int mlx4_en_start_port(struct net_device *dev)
/* Configure ring */
tx_ring = &priv->tx_ring[i];
- err = mlx4_en_activate_tx_ring(priv, tx_ring, cq->mcq.cqn);
+ err = mlx4_en_activate_tx_ring(priv, tx_ring, cq->mcq.cqn,
+ max(0, i - MLX4_EN_NUM_TX_RINGS));
if (err) {
en_err(priv, "Failed allocating Tx ring\n");
mlx4_en_deactivate_cq(priv, cq);
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_resources.c b/drivers/net/ethernet/mellanox/mlx4/en_resources.c
index bcbc54c..10c24c7 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_resources.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_resources.c
@@ -39,7 +39,7 @@
void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride,
int is_tx, int rss, int qpn, int cqn,
- struct mlx4_qp_context *context)
+ int user_prio, struct mlx4_qp_context *context)
{
struct mlx4_en_dev *mdev = priv->mdev;
@@ -57,6 +57,10 @@ void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride,
context->local_qpn = cpu_to_be32(qpn);
context->pri_path.ackto = 1 & 0x07;
context->pri_path.sched_queue = 0x83 | (priv->port - 1) << 6;
+ if (user_prio >= 0) {
+ context->pri_path.sched_queue |= user_prio << 3;
+ context->pri_path.feup = 1 << 6;
+ }
context->pri_path.counter_index = 0xff;
context->cqn_send = cpu_to_be32(cqn);
context->cqn_recv = cpu_to_be32(cqn);
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_rx.c b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
index 9adbd53..d49a7ac 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_rx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_rx.c
@@ -823,7 +823,7 @@ static int mlx4_en_config_rss_qp(struct mlx4_en_priv *priv, int qpn,
memset(context, 0, sizeof *context);
mlx4_en_fill_qp_context(priv, ring->actual_size, ring->stride, 0, 0,
- qpn, ring->cqn, context);
+ qpn, ring->cqn, -1, context);
context->db_rec_addr = cpu_to_be64(ring->wqres.db.dma);
/* Cancel FCS removal if FW allows */
@@ -890,7 +890,7 @@ int mlx4_en_config_rss_steer(struct mlx4_en_priv *priv)
}
rss_map->indir_qp.event = mlx4_en_sqp_event;
mlx4_en_fill_qp_context(priv, 0, 0, 0, 1, priv->base_qpn,
- priv->rx_ring[0].cqn, &context);
+ priv->rx_ring[0].cqn, -1, &context);
if (!priv->prof->rss_rings || priv->prof->rss_rings > priv->rx_ring_num)
rss_rings = priv->rx_ring_num;
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index 1796824..94a605a 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -156,7 +156,7 @@ void mlx4_en_destroy_tx_ring(struct mlx4_en_priv *priv,
int mlx4_en_activate_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring *ring,
- int cq)
+ int cq, int user_prio)
{
struct mlx4_en_dev *mdev = priv->mdev;
int err;
@@ -174,7 +174,7 @@ int mlx4_en_activate_tx_ring(struct mlx4_en_priv *priv,
ring->doorbell_qpn = ring->qp.qpn << 8;
mlx4_en_fill_qp_context(priv, ring->size, ring->stride, 1, 0, ring->qpn,
- ring->cqn, &ring->context);
+ ring->cqn, user_prio, &ring->context);
if (ring->bf_enabled)
ring->context.usr_page = cpu_to_be32(ring->bf.uar->index);
@@ -570,18 +570,14 @@ static void build_inline_wqe(struct mlx4_en_tx_desc *tx_desc, struct sk_buff *sk
u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb)
{
- struct mlx4_en_priv *priv = netdev_priv(dev);
u16 vlan_tag = 0;
- /* If we support per priority flow control and the packet contains
- * a vlan tag, send the packet to the TX ring assigned to that priority
- */
- if (priv->prof->rx_ppp && vlan_tx_tag_present(skb)) {
+ if (vlan_tx_tag_present(skb)) {
vlan_tag = vlan_tx_tag_get(skb);
return MLX4_EN_NUM_TX_RINGS + (vlan_tag >> 13);
}
- return skb_tx_hash(dev, skb);
+ return __skb_tx_hash(dev, skb, MLX4_EN_NUM_TX_RINGS);
}
static void mlx4_bf_copy(void __iomem *dst, unsigned long *src, unsigned bytecnt)
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index 9e2b911..5bd7c2a 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -521,7 +521,7 @@ int mlx4_en_create_tx_ring(struct mlx4_en_priv *priv, struct mlx4_en_tx_ring *ri
void mlx4_en_destroy_tx_ring(struct mlx4_en_priv *priv, struct mlx4_en_tx_ring *ring);
int mlx4_en_activate_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring *ring,
- int cq);
+ int cq, int user_prio);
void mlx4_en_deactivate_tx_ring(struct mlx4_en_priv *priv,
struct mlx4_en_tx_ring *ring);
@@ -539,8 +539,8 @@ int mlx4_en_process_rx_cq(struct net_device *dev,
int budget);
int mlx4_en_poll_rx_cq(struct napi_struct *napi, int budget);
void mlx4_en_fill_qp_context(struct mlx4_en_priv *priv, int size, int stride,
- int is_tx, int rss, int qpn, int cqn,
- struct mlx4_qp_context *context);
+ int is_tx, int rss, int qpn, int cqn, int user_prio,
+ struct mlx4_qp_context *context);
void mlx4_en_sqp_event(struct mlx4_qp *qp, enum mlx4_event event);
int mlx4_en_map_buffer(struct mlx4_buf *buf);
void mlx4_en_unmap_buffer(struct mlx4_buf *buf);
diff --git a/include/linux/mlx4/qp.h b/include/linux/mlx4/qp.h
index 091f9e7..96005d7 100644
--- a/include/linux/mlx4/qp.h
+++ b/include/linux/mlx4/qp.h
@@ -139,7 +139,8 @@ struct mlx4_qp_path {
u8 rgid[16];
u8 sched_queue;
u8 vlan_index;
- u8 reserved3[2];
+ u8 feup;
+ u8 reserved3;
u8 reserved4[2];
u8 dmac[6];
};
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH V2 2/7] net/mlx4_core: set port QoS attributes
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
2012-03-26 9:07 ` [PATCH V2 1/7] net/mlx4_en: Force user priority by QP attribute Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 9:07 ` [PATCH V2 3/7] net/mlx4_en: DCB QoS support Amir Vadai
` (4 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Adding QoS firmware commands:
- mlx4_en_SET_PORT_PRIO2TC - set UP <=> TC
- mlx4_en_SET_PORT_SCHEDULER - set promised BW, max BW and PG number
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/en_port.h | 2 +
drivers/net/ethernet/mellanox/mlx4/mlx4.h | 21 ++++++++
drivers/net/ethernet/mellanox/mlx4/port.c | 64 ++++++++++++++++++++++++++
include/linux/mlx4/cmd.h | 4 ++
include/linux/mlx4/device.h | 3 +
5 files changed, 94 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_port.h b/drivers/net/ethernet/mellanox/mlx4/en_port.h
index 6934fd7..745090b 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_port.h
+++ b/drivers/net/ethernet/mellanox/mlx4/en_port.h
@@ -39,6 +39,8 @@
#define SET_PORT_PROMISC_SHIFT 31
#define SET_PORT_MC_PROMISC_SHIFT 30
+#define MLX4_EN_NUM_TC 8
+
#define VLAN_FLTR_SIZE 128
struct mlx4_set_vlan_fltr_mbox {
__be32 entry[VLAN_FLTR_SIZE];
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4.h b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
index 2a0ff2c..92fea8f 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4.h
@@ -53,6 +53,27 @@
#define DRV_VERSION "1.1"
#define DRV_RELDATE "Dec, 2011"
+#define MLX4_NUM_UP 8
+#define MLX4_NUM_TC 8
+#define MLX4_RATELIMIT_UNITS 3 /* 100 Mbps */
+#define MLX4_RATELIMIT_MB_TO_HW(_mb) ((_mb) / 100)
+#define MLX4_RATELIMIT_DEFAULT 0xffff
+
+struct mlx4_set_port_prio2tc_context {
+ u8 prio2tc[4];
+};
+
+struct mlx4_port_scheduler_tc_cfg_be {
+ __be16 pg;
+ __be16 bw_precentage;
+ __be16 max_bw_units; /* 3-100Mbps, 4-1Gbps, other values - reserved */
+ __be16 max_bw_value;
+};
+
+struct mlx4_set_port_scheduler_context {
+ struct mlx4_port_scheduler_tc_cfg_be tc[MLX4_NUM_TC];
+};
+
enum {
MLX4_HCR_BASE = 0x80680,
MLX4_HCR_SIZE = 0x0001c,
diff --git a/drivers/net/ethernet/mellanox/mlx4/port.c b/drivers/net/ethernet/mellanox/mlx4/port.c
index 77535ff..0f66cde 100644
--- a/drivers/net/ethernet/mellanox/mlx4/port.c
+++ b/drivers/net/ethernet/mellanox/mlx4/port.c
@@ -834,6 +834,70 @@ int mlx4_SET_PORT_qpn_calc(struct mlx4_dev *dev, u8 port, u32 base_qpn,
}
EXPORT_SYMBOL(mlx4_SET_PORT_qpn_calc);
+int mlx4_SET_PORT_PRIO2TC(struct mlx4_dev *dev, u8 port, u8 *prio2tc)
+{
+ struct mlx4_cmd_mailbox *mailbox;
+ struct mlx4_set_port_prio2tc_context *context;
+ int err;
+ u32 in_mod;
+ int i;
+
+ mailbox = mlx4_alloc_cmd_mailbox(dev);
+ if (IS_ERR(mailbox))
+ return PTR_ERR(mailbox);
+ context = mailbox->buf;
+ memset(context, 0, sizeof *context);
+
+ for (i = 0; i < MLX4_NUM_UP; i += 2)
+ context->prio2tc[i >> 1] = prio2tc[i] << 4 | prio2tc[i + 1];
+
+ in_mod = MLX4_SET_PORT_PRIO2TC << 8 | port;
+ err = mlx4_cmd(dev, mailbox->dma, in_mod, 1, MLX4_CMD_SET_PORT,
+ MLX4_CMD_TIME_CLASS_B, MLX4_CMD_NATIVE);
+
+ mlx4_free_cmd_mailbox(dev, mailbox);
+ return err;
+}
+EXPORT_SYMBOL(mlx4_SET_PORT_PRIO2TC);
+
+int mlx4_SET_PORT_SCHEDULER(struct mlx4_dev *dev, u8 port, u8 *tc_tx_bw,
+ u8 *pg, u16 *ratelimit)
+{
+ struct mlx4_cmd_mailbox *mailbox;
+ struct mlx4_set_port_scheduler_context *context;
+ int err;
+ u32 in_mod;
+ int i;
+
+ mailbox = mlx4_alloc_cmd_mailbox(dev);
+ if (IS_ERR(mailbox))
+ return PTR_ERR(mailbox);
+ context = mailbox->buf;
+ memset(context, 0, sizeof *context);
+
+ for (i = 0; i < MLX4_NUM_TC; i++) {
+ struct mlx4_port_scheduler_tc_cfg_be *tc = &context->tc[i];
+ u16 r = MLX4_RATELIMIT_MB_TO_HW(ratelimit[i]);
+
+ if (!r)
+ r = MLX4_RATELIMIT_DEFAULT;
+
+ tc->pg = htons(pg[i]);
+ tc->bw_precentage = htons(tc_tx_bw[i]);
+
+ tc->max_bw_units = htons(MLX4_RATELIMIT_UNITS);
+ tc->max_bw_value = htons(r);
+ }
+
+ in_mod = MLX4_SET_PORT_SCHEDULER << 8 | port;
+ err = mlx4_cmd(dev, mailbox->dma, in_mod, 1, MLX4_CMD_SET_PORT,
+ MLX4_CMD_TIME_CLASS_B, MLX4_CMD_NATIVE);
+
+ mlx4_free_cmd_mailbox(dev, mailbox);
+ return err;
+}
+EXPORT_SYMBOL(mlx4_SET_PORT_SCHEDULER);
+
int mlx4_SET_MCAST_FLTR_wrapper(struct mlx4_dev *dev, int slave,
struct mlx4_vhcr *vhcr,
struct mlx4_cmd_mailbox *inbox,
diff --git a/include/linux/mlx4/cmd.h b/include/linux/mlx4/cmd.h
index 9958ff2..1f3860a 100644
--- a/include/linux/mlx4/cmd.h
+++ b/include/linux/mlx4/cmd.h
@@ -150,6 +150,10 @@ enum {
/* statistics commands */
MLX4_CMD_QUERY_IF_STAT = 0X54,
MLX4_CMD_SET_IF_STAT = 0X55,
+
+ /* set port opcode modifiers */
+ MLX4_SET_PORT_PRIO2TC = 0x8,
+ MLX4_SET_PORT_SCHEDULER = 0x9,
};
enum {
diff --git a/include/linux/mlx4/device.h b/include/linux/mlx4/device.h
index 834c96c..6d02824 100644
--- a/include/linux/mlx4/device.h
+++ b/include/linux/mlx4/device.h
@@ -628,6 +628,9 @@ int mlx4_SET_PORT_general(struct mlx4_dev *dev, u8 port, int mtu,
u8 pptx, u8 pfctx, u8 pprx, u8 pfcrx);
int mlx4_SET_PORT_qpn_calc(struct mlx4_dev *dev, u8 port, u32 base_qpn,
u8 promisc);
+int mlx4_SET_PORT_PRIO2TC(struct mlx4_dev *dev, u8 port, u8 *prio2tc);
+int mlx4_SET_PORT_SCHEDULER(struct mlx4_dev *dev, u8 port, u8 *tc_tx_bw,
+ u8 *pg, u16 *ratelimit);
int mlx4_find_cached_vlan(struct mlx4_dev *dev, u8 port, u16 vid, int *idx);
int mlx4_register_vlan(struct mlx4_dev *dev, u8 port, u16 vlan, int *index);
void mlx4_unregister_vlan(struct mlx4_dev *dev, u8 port, int index);
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH V2 3/7] net/mlx4_en: DCB QoS support
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
2012-03-26 9:07 ` [PATCH V2 1/7] net/mlx4_en: Force user priority by QP attribute Amir Vadai
2012-03-26 9:07 ` [PATCH V2 2/7] net/mlx4_core: set port QoS attributes Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 9:07 ` [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC Amir Vadai
` (3 subsequent siblings)
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Set TSA, promised BW and PFC using IEEE 802.1qaz netlink commands.
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/Kconfig | 12 ++
drivers/net/ethernet/mellanox/mlx4/Makefile | 1 +
drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c | 212 ++++++++++++++++++++++++
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 16 ++
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 23 +++
5 files changed, 264 insertions(+), 0 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
diff --git a/drivers/net/ethernet/mellanox/mlx4/Kconfig b/drivers/net/ethernet/mellanox/mlx4/Kconfig
index 1bb9353..5f027f9 100644
--- a/drivers/net/ethernet/mellanox/mlx4/Kconfig
+++ b/drivers/net/ethernet/mellanox/mlx4/Kconfig
@@ -11,6 +11,18 @@ config MLX4_EN
This driver supports Mellanox Technologies ConnectX Ethernet
devices.
+config MLX4_EN_DCB
+ bool "Data Center Bridging (DCB) Support"
+ default y
+ depends on MLX4_EN && DCB
+ ---help---
+ Say Y here if you want to use Data Center Bridging (DCB) in the
+ driver.
+ If set to N, will not be able to configure QoS and ratelimit attributes.
+ This flag is depended on the kernel's DCB support.
+
+ If unsure, set to Y
+
config MLX4_CORE
tristate
depends on PCI
diff --git a/drivers/net/ethernet/mellanox/mlx4/Makefile b/drivers/net/ethernet/mellanox/mlx4/Makefile
index 4a40ab9..293127d 100644
--- a/drivers/net/ethernet/mellanox/mlx4/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx4/Makefile
@@ -7,3 +7,4 @@ obj-$(CONFIG_MLX4_EN) += mlx4_en.o
mlx4_en-y := en_main.o en_tx.o en_rx.o en_ethtool.o en_port.o en_cq.o \
en_resources.o en_netdev.o en_selftest.o
+mlx4_en-$(CONFIG_MLX4_EN_DCB) += en_dcb_nl.o
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
new file mode 100644
index 0000000..abd4181
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx4/en_dcb_nl.c
@@ -0,0 +1,212 @@
+/*
+ * Copyright (c) 2011 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/dcbnl.h>
+
+#include "mlx4_en.h"
+
+static int mlx4_en_dcbnl_ieee_getets(struct net_device *dev,
+ struct ieee_ets *ets)
+{
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+ struct ieee_ets *my_ets = priv->mlx4_en_ieee_ets;
+
+ /* No IEEE PFC settings available */
+ if (!my_ets)
+ return -EINVAL;
+
+ ets->ets_cap = IEEE_8021QAZ_MAX_TCS;
+ ets->cbs = my_ets->cbs;
+ memcpy(ets->tc_tx_bw, my_ets->tc_tx_bw, sizeof(ets->tc_tx_bw));
+ memcpy(ets->tc_tsa, my_ets->tc_tsa, sizeof(ets->tc_tsa));
+ memcpy(ets->prio_tc, my_ets->prio_tc, sizeof(ets->prio_tc));
+
+ return 0;
+}
+
+static int mlx4_en_ets_validate(struct mlx4_en_priv *priv, struct ieee_ets *ets)
+{
+ int i;
+ int total_ets_bw = 0;
+ int has_ets_tc = 0;
+
+ for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) {
+ if (ets->prio_tc[i] > MLX4_EN_NUM_UP) {
+ en_err(priv, "Bad priority in UP <=> TC mapping. "
+ "TC: %d, UP: %d\n", i, ets->prio_tc[i]);
+ return -EINVAL;
+ }
+
+ switch (ets->tc_tsa[i]) {
+ case IEEE_8021QAZ_TSA_STRICT:
+ break;
+ case IEEE_8021QAZ_TSA_ETS:
+ has_ets_tc = 1;
+ total_ets_bw += ets->tc_tx_bw[i];
+ break;
+ default:
+ en_err(priv, "TC[%d]: Not supported TSA: %d\n",
+ i, ets->tc_tsa[i]);
+ return -ENOTSUPP;
+ }
+ }
+
+ if (has_ets_tc && total_ets_bw != MLX4_EN_BW_MAX) {
+ en_err(priv, "Bad ETS BW sum: %d. Should be exactly 100%%\n",
+ total_ets_bw);
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+int mlx4_en_config_port_scheduler(struct mlx4_en_priv *priv,
+ struct ieee_ets *ets, u16 *ratelimit)
+{
+ struct mlx4_en_dev *mdev = priv->mdev;
+ int num_strict = 0;
+ int i;
+ __u8 tc_tx_bw[IEEE_8021QAZ_MAX_TCS] = { 0 };
+ __u8 pg[IEEE_8021QAZ_MAX_TCS] = { 0 };
+
+ ets = ets ?: priv->mlx4_en_ieee_ets;
+ ratelimit = ratelimit ?: priv->ratelimit;
+
+ /* higher TC means higher priority => lower pg */
+ for (i = IEEE_8021QAZ_MAX_TCS - 1; i >= 0; i--) {
+ switch (ets->tc_tsa[i]) {
+ case IEEE_8021QAZ_TSA_STRICT:
+ pg[i] = num_strict++;
+ tc_tx_bw[i] = MLX4_EN_BW_MAX;
+ break;
+ case IEEE_8021QAZ_TSA_ETS:
+ pg[i] = MLX4_EN_TC_ETS;
+ tc_tx_bw[i] = ets->tc_tx_bw[i] ?: MLX4_EN_BW_MIN;
+ break;
+ }
+ }
+
+ return mlx4_SET_PORT_SCHEDULER(mdev->dev, priv->port, tc_tx_bw, pg,
+ ratelimit);
+}
+
+static int
+mlx4_en_dcbnl_ieee_setets(struct net_device *dev, struct ieee_ets *ets)
+{
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+ struct mlx4_en_dev *mdev = priv->mdev;
+ int err;
+
+ err = mlx4_en_ets_validate(priv, ets);
+ if (err)
+ return err;
+
+ err = mlx4_SET_PORT_PRIO2TC(mdev->dev, priv->port, ets->prio_tc);
+ if (err)
+ return err;
+
+ err = mlx4_en_config_port_scheduler(priv, ets, NULL);
+ if (err)
+ return err;
+
+ if (ets != priv->mlx4_en_ieee_ets)
+ memcpy(priv->mlx4_en_ieee_ets, ets,
+ sizeof(*priv->mlx4_en_ieee_ets));
+
+ return 0;
+}
+
+static int mlx4_en_dcbnl_ieee_getpfc(struct net_device *dev,
+ struct ieee_pfc *pfc)
+{
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+
+ pfc->pfc_cap = IEEE_8021QAZ_MAX_TCS;
+ pfc->pfc_en = priv->prof->tx_ppp;
+
+ return 0;
+}
+
+static int mlx4_en_dcbnl_ieee_setpfc(struct net_device *dev,
+ struct ieee_pfc *pfc)
+{
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+ struct mlx4_en_dev *mdev = priv->mdev;
+ int err;
+
+ en_dbg(DRV, priv, "cap: 0x%x en: 0x%x mbc: 0x%x delay: %d\n",
+ pfc->pfc_cap,
+ pfc->pfc_en,
+ pfc->mbc,
+ pfc->delay);
+
+ priv->prof->rx_pause = priv->prof->tx_pause = !!pfc->pfc_en;
+ priv->prof->rx_ppp = priv->prof->tx_ppp = pfc->pfc_en;
+
+ err = mlx4_SET_PORT_general(mdev->dev, priv->port,
+ priv->rx_skb_size + ETH_FCS_LEN,
+ priv->prof->tx_pause,
+ priv->prof->tx_ppp,
+ priv->prof->rx_pause,
+ priv->prof->rx_ppp);
+ if (err)
+ en_err(priv, "Failed setting pause params\n");
+
+ return err;
+}
+
+static u8 mlx4_en_dcbnl_getdcbx(struct net_device *dev)
+{
+ return DCB_CAP_DCBX_VER_IEEE;
+}
+
+static u8 mlx4_en_dcbnl_setdcbx(struct net_device *dev, u8 mode)
+{
+ if ((mode & DCB_CAP_DCBX_LLD_MANAGED) ||
+ (mode & DCB_CAP_DCBX_VER_CEE) ||
+ !(mode & DCB_CAP_DCBX_VER_IEEE) ||
+ !(mode & DCB_CAP_DCBX_HOST))
+ return 1;
+
+ return 0;
+}
+
+const struct dcbnl_rtnl_ops mlx4_en_dcbnl_ops = {
+ .ieee_getets = mlx4_en_dcbnl_ieee_getets,
+ .ieee_setets = mlx4_en_dcbnl_ieee_setets,
+ .ieee_getpfc = mlx4_en_dcbnl_ieee_getpfc,
+ .ieee_setpfc = mlx4_en_dcbnl_ieee_setpfc,
+
+ .getdcbx = mlx4_en_dcbnl_getdcbx,
+ .setdcbx = mlx4_en_dcbnl_setdcbx,
+};
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 2322622..9b456ae 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -967,6 +967,11 @@ void mlx4_en_destroy_netdev(struct net_device *dev)
mutex_unlock(&mdev->state_lock);
mlx4_en_free_resources(priv);
+
+#ifdef CONFIG_MLX4_EN_DCB
+ vfree(priv->mlx4_en_ieee_ets);
+#endif
+
free_netdev(dev);
}
@@ -1080,6 +1085,17 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
INIT_WORK(&priv->watchdog_task, mlx4_en_restart);
INIT_WORK(&priv->linkstate_task, mlx4_en_linkstate);
INIT_DELAYED_WORK(&priv->stats_task, mlx4_en_do_get_stats);
+#ifdef CONFIG_MLX4_EN_DCB
+ if (!mlx4_is_slave(priv->mdev->dev)) {
+ dev->dcbnl_ops = &mlx4_en_dcbnl_ops;
+
+ priv->mlx4_en_ieee_ets = vzalloc(sizeof(struct ieee_ets));
+ if (!priv->mlx4_en_ieee_ets) {
+ err = -ENOMEM;
+ goto out;
+ }
+ }
+#endif
/* Query for default mac and max mtu */
priv->max_mtu = mdev->dev->caps.eth_mtu_cap[priv->port];
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index 5bd7c2a..cc30677 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -40,6 +40,9 @@
#include <linux/mutex.h>
#include <linux/netdevice.h>
#include <linux/if_vlan.h>
+#ifdef CONFIG_MLX4_EN_DCB
+#include <linux/dcbnl.h>
+#endif
#include <linux/mlx4/device.h>
#include <linux/mlx4/qp.h>
@@ -110,6 +113,7 @@ enum {
#define MLX4_EN_NUM_TX_RINGS 8
#define MLX4_EN_NUM_PPP_RINGS 8
#define MAX_TX_RINGS (MLX4_EN_NUM_TX_RINGS + MLX4_EN_NUM_PPP_RINGS)
+#define MLX4_EN_NUM_UP 8
#define MLX4_EN_DEF_TX_RING_SIZE 512
#define MLX4_EN_DEF_RX_RING_SIZE 1024
@@ -410,6 +414,15 @@ struct mlx4_en_frag_info {
};
+#ifdef CONFIG_MLX4_EN_DCB
+/* Minimal TC BW - setting to 0 will block traffic */
+#define MLX4_EN_BW_MIN 1
+#define MLX4_EN_BW_MAX 100 /* Utilize 100% of the line */
+
+#define MLX4_EN_TC_ETS 7
+
+#endif
+
struct mlx4_en_priv {
struct mlx4_en_dev *mdev;
struct mlx4_en_port_profile *prof;
@@ -483,6 +496,10 @@ struct mlx4_en_priv {
int vids[128];
bool wol;
struct device *ddev;
+
+#ifdef CONFIG_MLX4_EN_DCB
+ struct ieee_ets *mlx4_en_ieee_ets;
+#endif
};
enum mlx4_en_wol {
@@ -557,6 +574,12 @@ int mlx4_SET_VLAN_FLTR(struct mlx4_dev *dev, struct mlx4_en_priv *priv);
int mlx4_en_DUMP_ETH_STATS(struct mlx4_en_dev *mdev, u8 port, u8 reset);
int mlx4_en_QUERY_PORT(struct mlx4_en_dev *mdev, u8 port);
+#ifdef CONFIG_MLX4_EN_DCB
+extern const struct dcbnl_rtnl_ops mlx4_en_dcbnl_ops;
+int mlx4_en_config_port_scheduler(struct mlx4_en_priv *priv,
+ struct ieee_ets *ets, u16 *ratelimit);
+#endif
+
#define MLX4_EN_NUM_SELF_TEST 5
void mlx4_en_ex_selftest(struct net_device *dev, u32 *flags, u64 *buf);
u64 mlx4_en_mac_to_u64(u8 *addr);
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
` (2 preceding siblings ...)
2012-03-26 9:07 ` [PATCH V2 3/7] net/mlx4_en: DCB QoS support Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 14:46 ` Ben Hutchings
2012-03-26 19:49 ` Roland Dreier
2012-03-26 9:07 ` [PATCH V2 5/7] net/mlx4_en: sk_prio <=> UP for untagged traffic Amir Vadai
` (2 subsequent siblings)
6 siblings, 2 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Set max rate-limit using sysfs file /sys/class/net/<interface>/ratelimit/tc<n>
Values in Mbps units.
For example, to set ratelimit of 5G to TC3 issue:
echo 5000 > /sys/class/net/eth2/ratelimit/tc3
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
We used sysfs since max bw isn't part of the ETS / DCBX NL support, and we're
open to other suggestions to add generic support for max bw, e.g add call to
the DCBX NL API.
drivers/net/ethernet/mellanox/mlx4/Makefile | 2 +-
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 13 +++
drivers/net/ethernet/mellanox/mlx4/en_sysfs.c | 99 ++++++++++++++++++++++++
drivers/net/ethernet/mellanox/mlx4/mlx4_en.h | 11 +++
4 files changed, 124 insertions(+), 1 deletions(-)
create mode 100644 drivers/net/ethernet/mellanox/mlx4/en_sysfs.c
diff --git a/drivers/net/ethernet/mellanox/mlx4/Makefile b/drivers/net/ethernet/mellanox/mlx4/Makefile
index 293127d..8e5bd88 100644
--- a/drivers/net/ethernet/mellanox/mlx4/Makefile
+++ b/drivers/net/ethernet/mellanox/mlx4/Makefile
@@ -7,4 +7,4 @@ obj-$(CONFIG_MLX4_EN) += mlx4_en.o
mlx4_en-y := en_main.o en_tx.o en_rx.o en_ethtool.o en_port.o en_cq.o \
en_resources.o en_netdev.o en_selftest.o
-mlx4_en-$(CONFIG_MLX4_EN_DCB) += en_dcb_nl.o
+mlx4_en-$(CONFIG_MLX4_EN_DCB) += en_dcb_nl.o en_sysfs.o
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index 9b456ae..f45d544 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -810,6 +810,18 @@ static void mlx4_en_restart(struct work_struct *work)
mutex_unlock(&mdev->state_lock);
}
+static int mlx4_en_init(struct net_device *dev)
+{
+#ifdef CONFIG_MLX4_EN_DCB
+ struct mlx4_en_priv *priv = netdev_priv(dev);
+
+ if (!mlx4_is_slave(priv->mdev->dev))
+ mlx4_en_prepare_sysfs_group(priv);
+#endif
+
+ return 0;
+}
+
static void mlx4_en_clear_stats(struct net_device *dev)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
@@ -1026,6 +1038,7 @@ static int mlx4_en_set_features(struct net_device *netdev,
}
static const struct net_device_ops mlx4_netdev_ops = {
+ .ndo_init = mlx4_en_init,
.ndo_open = mlx4_en_open,
.ndo_stop = mlx4_en_close,
.ndo_start_xmit = mlx4_en_xmit,
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_sysfs.c b/drivers/net/ethernet/mellanox/mlx4/en_sysfs.c
new file mode 100644
index 0000000..f633f61
--- /dev/null
+++ b/drivers/net/ethernet/mellanox/mlx4/en_sysfs.c
@@ -0,0 +1,99 @@
+/*
+ * Copyright (c) 2011 Mellanox Technologies. All rights reserved.
+ *
+ * This software is available to you under a choice of one of two
+ * licenses. You may choose to be licensed under the terms of the GNU
+ * General Public License (GPL) Version 2, available from the file
+ * COPYING in the main directory of this source tree, or the
+ * OpenIB.org BSD license below:
+ *
+ * Redistribution and use in source and binary forms, with or
+ * without modification, are permitted provided that the following
+ * conditions are met:
+ *
+ * - Redistributions of source code must retain the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer.
+ *
+ * - Redistributions in binary form must reproduce the above
+ * copyright notice, this list of conditions and the following
+ * disclaimer in the documentation and/or other materials
+ * provided with the distribution.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
+ * EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
+ * MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
+ * NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
+ * BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
+ * ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
+ * CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+ * SOFTWARE.
+ *
+ */
+
+#include <linux/device.h>
+#include <linux/netdevice.h>
+
+#include "mlx4_en.h"
+
+#define to_en_ratelimit_attr(x) container_of(x, struct en_ratelimit_attr, attr)
+static ssize_t mlx4_en_show_ratelimit(struct device *dev,
+ struct device_attribute *attr,
+ char *buf)
+{
+ struct en_ratelimit_attr *a = to_en_ratelimit_attr(attr);
+ struct mlx4_en_priv *priv = a->priv;
+ int tc = a->tc;
+
+ return snprintf(buf, PAGE_SIZE, "%d\n", priv->ratelimit[tc]);
+}
+
+static ssize_t mlx4_en_store_ratelimit(struct device *d,
+ struct device_attribute *attr,
+ const char *buf, size_t size)
+{
+ struct en_ratelimit_attr *a = to_en_ratelimit_attr(attr);
+ struct mlx4_en_priv *priv = a->priv;
+ int tc = a->tc;
+ u16 ratelimit[IEEE_8021QAZ_MAX_TCS];
+ int ret;
+
+ memcpy(ratelimit, priv->ratelimit, sizeof(ratelimit));
+
+ ret = kstrtou16(buf, 0, &ratelimit[tc]);
+ if (ret)
+ return ret;
+
+ ret = mlx4_en_config_port_scheduler(priv, NULL, ratelimit);
+ if (ret)
+ return ret;
+
+ priv->ratelimit[tc] = ratelimit[tc];
+
+ return size;
+}
+
+void mlx4_en_prepare_sysfs_group(struct mlx4_en_priv *priv)
+{
+ static char *dev_attr_tcs_name[] = { "tc0", "tc1", "tc2", "tc3", "tc4",
+ "tc5", "tc6", "tc7"};
+ int i;
+
+ priv->ratelimit_group.name = "ratelimit",
+ priv->ratelimit_group.attrs =
+ (struct attribute **)priv->ratelimit_attrs_p;
+ for (i = 0; i < IEEE_8021QAZ_MAX_TCS; i++) {
+ struct en_ratelimit_attr *dev_attr = &priv->ratelimit_attrs[i];
+
+ dev_attr->attr.attr.name = dev_attr_tcs_name[i];
+ dev_attr->attr.attr.mode = S_IRUGO | S_IWUSR;
+ dev_attr->attr.show = mlx4_en_show_ratelimit;
+ dev_attr->attr.store = mlx4_en_store_ratelimit;
+ dev_attr->priv = priv;
+ dev_attr->tc = i;
+
+ priv->ratelimit_attrs_p[i] = dev_attr;
+ }
+
+ priv->dev->sysfs_groups[0] = &priv->ratelimit_group;
+}
diff --git a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
index cc30677..7c32973 100644
--- a/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
+++ b/drivers/net/ethernet/mellanox/mlx4/mlx4_en.h
@@ -421,6 +421,12 @@ struct mlx4_en_frag_info {
#define MLX4_EN_TC_ETS 7
+struct en_ratelimit_attr {
+ struct device_attribute attr;
+ struct mlx4_en_priv *priv;
+ int tc;
+};
+
#endif
struct mlx4_en_priv {
@@ -499,6 +505,10 @@ struct mlx4_en_priv {
#ifdef CONFIG_MLX4_EN_DCB
struct ieee_ets *mlx4_en_ieee_ets;
+ u16 ratelimit[IEEE_8021QAZ_MAX_TCS];
+ struct attribute_group ratelimit_group;
+ struct en_ratelimit_attr ratelimit_attrs[IEEE_8021QAZ_MAX_TCS];
+ struct en_ratelimit_attr *ratelimit_attrs_p[IEEE_8021QAZ_MAX_TCS + 1];
#endif
};
@@ -578,6 +588,7 @@ int mlx4_en_QUERY_PORT(struct mlx4_en_dev *mdev, u8 port);
extern const struct dcbnl_rtnl_ops mlx4_en_dcbnl_ops;
int mlx4_en_config_port_scheduler(struct mlx4_en_priv *priv,
struct ieee_ets *ets, u16 *ratelimit);
+void mlx4_en_prepare_sysfs_group(struct mlx4_en_priv *priv);
#endif
#define MLX4_EN_NUM_SELF_TEST 5
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* Re: [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 9:07 ` [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC Amir Vadai
@ 2012-03-26 14:46 ` Ben Hutchings
2012-03-26 17:00 ` Or Gerlitz
2012-03-26 19:49 ` Roland Dreier
1 sibling, 1 reply; 13+ messages in thread
From: Ben Hutchings @ 2012-03-26 14:46 UTC (permalink / raw)
To: Amir Vadai
Cc: David S. Miller, netdev, Roland Dreier, Yevgeny Petrilin,
Oren Duer, Amir Vadai
On Mon, 2012-03-26 at 11:07 +0200, Amir Vadai wrote:
> Set max rate-limit using sysfs file /sys/class/net/<interface>/ratelimit/tc<n>
> Values in Mbps units.
>
> For example, to set ratelimit of 5G to TC3 issue:
> echo 5000 > /sys/class/net/eth2/ratelimit/tc3
This looks the wrong way round; wouldn't a tc be an object with a
ratelimit as attribute?
> Signed-off-by: Amir Vadai <amirv@mellanox.com>
> ---
>
> We used sysfs since max bw isn't part of the ETS / DCBX NL support, and we're
> open to other suggestions to add generic support for max bw, e.g add call to
> the DCBX NL API.
[...]
netlink interfaces are generally easily extensible and it doesn't make
sense to me to augment such an interface through sysfs. Perhaps you're
concerned that netlink extensions won't be supported in older kernel
versions running your OOT driver? That's unfortunate, but let's not
standardise an ugly interface based on a temporary problem like that.
Ben.
--
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 14:46 ` Ben Hutchings
@ 2012-03-26 17:00 ` Or Gerlitz
2012-03-26 18:55 ` John Fastabend
0 siblings, 1 reply; 13+ messages in thread
From: Or Gerlitz @ 2012-03-26 17:00 UTC (permalink / raw)
To: Ben Hutchings, Eric Dumazet, David Miller
Cc: Amir Vadai, netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer,
Amir Vadai
On Mon, Mar 26, 2012 at 4:46 PM, Ben Hutchings
<bhutchings@solarflare.com> wrote:
>> We used sysfs since max bw isn't part of the ETS / DCBX NL support, and we're
>> open to other suggestions to add generic support for max bw, e.g add call to
>> the DCBX NL API.
> netlink interfaces are generally easily extensible and it doesn't make
> sense to me to augment such an interface through sysfs. Perhaps you're
> concerned that netlink extensions won't be supported in older kernel
> versions running your OOT driver? That's unfortunate, but let's not
> standardise an ugly interface based on a temporary problem like that.
As written above, that was done since ratelimit isn't part of ETS, we can
that through netlink extensions that you mentioned, if this is the preffered
way to go, David? Eric? Ben - could you provide pointer to these extensions?
Or.
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 17:00 ` Or Gerlitz
@ 2012-03-26 18:55 ` John Fastabend
2012-03-27 12:03 ` Amir Vadai
0 siblings, 1 reply; 13+ messages in thread
From: John Fastabend @ 2012-03-26 18:55 UTC (permalink / raw)
To: Or Gerlitz
Cc: Ben Hutchings, Eric Dumazet, David Miller, Amir Vadai, netdev,
Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai
On 3/26/2012 10:00 AM, Or Gerlitz wrote:
> On Mon, Mar 26, 2012 at 4:46 PM, Ben Hutchings
> <bhutchings@solarflare.com> wrote:
>>> We used sysfs since max bw isn't part of the ETS / DCBX NL support, and we're
>>> open to other suggestions to add generic support for max bw, e.g add call to
>>> the DCBX NL API.
>
>> netlink interfaces are generally easily extensible and it doesn't make
>> sense to me to augment such an interface through sysfs. Perhaps you're
>> concerned that netlink extensions won't be supported in older kernel
>> versions running your OOT driver? That's unfortunate, but let's not
>> standardise an ugly interface based on a temporary problem like that.
>
> As written above, that was done since ratelimit isn't part of ETS, we can
> that through netlink extensions that you mentioned, if this is the preffered
> way to go, David? Eric? Ben - could you provide pointer to these extensions?
>
> Or.
> --
I think I original suggested it didn't belong in DCBNL because it
wasn't part of ETS (802.1Qaz). But it _is_ a traffic selection
algorithm and could fall into the vendor specific part of 802.1Q.
I would suggest either adding it as an option to mqprio to take
a max bandwidth. The advantage here is it would be tied in with
the usual QOS tooling 'tc'.
# tc qdisc add dev eth3 root mqprio help
Usage: ... mqprio [num_tc NUMBER] [map P0 P1 ...]
[queues count1@offset1 count2@offset2 ...] [hw 1|0]
[max_rate rate@tc ...]
Or extending DCBNL being careful not to break backwards
compatibility. I tend to think extending mqprio is cleaner but
a DCBNL extension could likely work as well. Would need a
'DCBNL_IEEE_SET_MAXRATE' and 'DCBNL_IEEE_GET_MAXRATE' for this
I expect.
.John
^ permalink raw reply [flat|nested] 13+ messages in thread* Re: [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 18:55 ` John Fastabend
@ 2012-03-27 12:03 ` Amir Vadai
0 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-27 12:03 UTC (permalink / raw)
To: John Fastabend
Cc: Or Gerlitz, Ben Hutchings, netdev, Yevgeny Petrilin, Oren Duer,
Amir Vadai, Liran Liss
On 03/26/2012 08:55 PM, John Fastabend wrote:
> On 3/26/2012 10:00 AM, Or Gerlitz wrote:
>> On Mon, Mar 26, 2012 at 4:46 PM, Ben Hutchings
>> <bhutchings@solarflare.com> wrote:
>>>> We used sysfs since max bw isn't part of the ETS / DCBX NL support, and we're
>>>> open to other suggestions to add generic support for max bw, e.g add call to
>>>> the DCBX NL API.
>>
>>> netlink interfaces are generally easily extensible and it doesn't make
>>> sense to me to augment such an interface through sysfs. Perhaps you're
>>> concerned that netlink extensions won't be supported in older kernel
>>> versions running your OOT driver? That's unfortunate, but let's not
>>> standardise an ugly interface based on a temporary problem like that.
>>
>> As written above, that was done since ratelimit isn't part of ETS, we can
>> that through netlink extensions that you mentioned, if this is the preffered
>> way to go, David? Eric? Ben - could you provide pointer to these extensions?
>>
>> Or.
>> --
>
> I think I original suggested it didn't belong in DCBNL because it
> wasn't part of ETS (802.1Qaz). But it _is_ a traffic selection
> algorithm and could fall into the vendor specific part of 802.1Q.
>
> I would suggest either adding it as an option to mqprio to take
> a max bandwidth. The advantage here is it would be tied in with
> the usual QOS tooling 'tc'.
>
> # tc qdisc add dev eth3 root mqprio help
> Usage: ... mqprio [num_tc NUMBER] [map P0 P1 ...]
> [queues count1@offset1 count2@offset2 ...] [hw 1|0]
> [max_rate rate@tc ...]
>
This could be elegant, but since tc here is a logical traffic class and
ratelimit in our context is an attribute of ETS TC, it could be
problematic.
>
> Or extending DCBNL being careful not to break backwards
> compatibility. I tend to think extending mqprio is cleaner but
> a DCBNL extension could likely work as well. Would need a
> 'DCBNL_IEEE_SET_MAXRATE' and 'DCBNL_IEEE_GET_MAXRATE' for this
> I expect.
I do prefer this option. I'm checking now if it is possible to add
infrastructure for a vendor specific netlink command. It could be
connected by lldpad to the vendor TLV in DCBX.
>
> .John
I will send a V3 of this patchset without the ratelimit patch, to make
sure the patchset will get into net-next on time.
Thanks,
Amir
^ permalink raw reply [flat|nested] 13+ messages in thread
* Re: [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC
2012-03-26 9:07 ` [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC Amir Vadai
2012-03-26 14:46 ` Ben Hutchings
@ 2012-03-26 19:49 ` Roland Dreier
1 sibling, 0 replies; 13+ messages in thread
From: Roland Dreier @ 2012-03-26 19:49 UTC (permalink / raw)
To: Amir Vadai
Cc: David S. Miller, netdev, Yevgeny Petrilin, Oren Duer, Amir Vadai
On Mon, Mar 26, 2012 at 2:07 AM, Amir Vadai <amirv@mellanox.com> wrote:
> Values in Mbps units.
Hopefully not nitpicking too much, but Mbps still seems to coarse to me.
Eg in wireless world I can imagine wanting to limit a tos to a small # of kbps.
If this is done through netlink, the ABI can use a small unit and then tc
can accept abbreviations for kbps / Mbps / Gbps?
- R.
^ permalink raw reply [flat|nested] 13+ messages in thread
* [PATCH V2 5/7] net/mlx4_en: sk_prio <=> UP for untagged traffic
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
` (3 preceding siblings ...)
2012-03-26 9:07 ` [PATCH V2 4/7] net/mlx4_en: Set max rate-limit for a TC Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 9:07 ` [PATCH V2 6/7] net/route: export symbol ip_tos2prio Amir Vadai
2012-03-26 9:07 ` [PATCH V2 7/7] IB/rdma_cm: TOS <=> UP mapping for IBoE Amir Vadai
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Since vlan egress map is only good for tagged traffic, need to have other
mapping to be used by untagged traffic.
For that, the driver uses sch_mqprio mapping. This mapping could be set by
using tc tool from iproute2 package.
Mapped UP will be used by the HW for QoS purposes, but won't go out on the
wire.
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx4/en_netdev.c | 18 ++++++++++++++++++
drivers/net/ethernet/mellanox/mlx4/en_tx.c | 2 +-
2 files changed, 19 insertions(+), 1 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
index f45d544..680f4dc 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_netdev.c
@@ -45,6 +45,14 @@
#include "mlx4_en.h"
#include "en_port.h"
+static int mlx4_en_setup_tc(struct net_device *dev, u8 up)
+{
+ if (up != MLX4_EN_NUM_UP)
+ return -EINVAL;
+
+ return 0;
+}
+
static int mlx4_en_vlan_rx_add_vid(struct net_device *dev, unsigned short vid)
{
struct mlx4_en_priv *priv = netdev_priv(dev);
@@ -1055,6 +1063,7 @@ static const struct net_device_ops mlx4_netdev_ops = {
.ndo_poll_controller = mlx4_en_netpoll,
#endif
.ndo_set_features = mlx4_en_set_features,
+ .ndo_setup_tc = mlx4_en_setup_tc,
};
int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
@@ -1143,6 +1152,15 @@ int mlx4_en_init_netdev(struct mlx4_en_dev *mdev, int port,
netif_set_real_num_tx_queues(dev, priv->tx_ring_num);
netif_set_real_num_rx_queues(dev, priv->rx_ring_num);
+ netdev_set_num_tc(dev, MLX4_EN_NUM_UP);
+
+ /* First 9 rings are for UP 0 */
+ netdev_set_tc_queue(dev, 0, MLX4_EN_NUM_TX_RINGS + 1, 0);
+
+ /* Partition Tx queues evenly amongst UP's 1-7 */
+ for (i = 1; i < MLX4_EN_NUM_UP; i++)
+ netdev_set_tc_queue(dev, i, 1, MLX4_EN_NUM_TX_RINGS + i);
+
SET_ETHTOOL_OPS(dev, &mlx4_en_ethtool_ops);
/* Set defualt MAC */
diff --git a/drivers/net/ethernet/mellanox/mlx4/en_tx.c b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
index 94a605a..d9bab53 100644
--- a/drivers/net/ethernet/mellanox/mlx4/en_tx.c
+++ b/drivers/net/ethernet/mellanox/mlx4/en_tx.c
@@ -577,7 +577,7 @@ u16 mlx4_en_select_queue(struct net_device *dev, struct sk_buff *skb)
return MLX4_EN_NUM_TX_RINGS + (vlan_tag >> 13);
}
- return __skb_tx_hash(dev, skb, MLX4_EN_NUM_TX_RINGS);
+ return skb_tx_hash(dev, skb);
}
static void mlx4_bf_copy(void __iomem *dst, unsigned long *src, unsigned bytecnt)
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH V2 6/7] net/route: export symbol ip_tos2prio
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
` (4 preceding siblings ...)
2012-03-26 9:07 ` [PATCH V2 5/7] net/mlx4_en: sk_prio <=> UP for untagged traffic Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
2012-03-26 9:07 ` [PATCH V2 7/7] IB/rdma_cm: TOS <=> UP mapping for IBoE Amir Vadai
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai
Need to export this to enable drivers use rt_tos2priority()
Signed-off-by: Amir Vadai <amirv@mellanox.com>
---
net/ipv4/route.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/ipv4/route.c b/net/ipv4/route.c
index 12ccf88..69e0b86 100644
--- a/net/ipv4/route.c
+++ b/net/ipv4/route.c
@@ -230,7 +230,7 @@ const __u8 ip_tos2prio[16] = {
TC_PRIO_INTERACTIVE_BULK,
ECN_OR_COST(INTERACTIVE_BULK)
};
-
+EXPORT_SYMBOL(ip_tos2prio);
/*
* Route cache.
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread* [PATCH V2 7/7] IB/rdma_cm: TOS <=> UP mapping for IBoE
2012-03-26 9:07 [PATCH V2 0/7] net/mlx4_en: DCB QoS support Amir Vadai
` (5 preceding siblings ...)
2012-03-26 9:07 ` [PATCH V2 6/7] net/route: export symbol ip_tos2prio Amir Vadai
@ 2012-03-26 9:07 ` Amir Vadai
6 siblings, 0 replies; 13+ messages in thread
From: Amir Vadai @ 2012-03-26 9:07 UTC (permalink / raw)
To: David S. Miller
Cc: netdev, Roland Dreier, Yevgeny Petrilin, Oren Duer, Amir Vadai,
Amir Vadai, Sean Hefty
Both tagged traffic and untagged traffic use tc tool mapping.
Treat RDMA TOS same as IP TOS when mapping to SL
Signed-off-by: Amir Vadai <amirv@mellanox.com>
CC: Sean Hefty <sean.hefty@intel.com>
---
drivers/infiniband/core/cma.c | 6 +++++-
1 files changed, 5 insertions(+), 1 deletions(-)
diff --git a/drivers/infiniband/core/cma.c b/drivers/infiniband/core/cma.c
index e3e470f..59fbd70 100644
--- a/drivers/infiniband/core/cma.c
+++ b/drivers/infiniband/core/cma.c
@@ -42,6 +42,7 @@
#include <linux/inetdevice.h>
#include <linux/slab.h>
#include <linux/module.h>
+#include <net/route.h>
#include <net/tcp.h>
#include <net/ipv6.h>
@@ -1826,7 +1827,10 @@ static int cma_resolve_iboe_route(struct rdma_id_private *id_priv)
route->path_rec->reversible = 1;
route->path_rec->pkey = cpu_to_be16(0xffff);
route->path_rec->mtu_selector = IB_SA_EQ;
- route->path_rec->sl = id_priv->tos >> 5;
+ route->path_rec->sl = netdev_get_prio_tc_map(
+ ndev->priv_flags & IFF_802_1Q_VLAN ?
+ vlan_dev_real_dev(ndev) : ndev,
+ rt_tos2priority(id_priv->tos));
route->path_rec->mtu = iboe_get_mtu(ndev->mtu);
route->path_rec->rate_selector = IB_SA_EQ;
--
1.7.8.2
^ permalink raw reply related [flat|nested] 13+ messages in thread