* [net-next PATCH] dql: add a burst attribute
From: Eric Dumazet @ 2014-09-30 12:51 UTC (permalink / raw)
To: Jesper Dangaard Brouer
Cc: netdev, David S. Miller, Tom Herbert, Hannes Frederic Sowa,
Florian Westphal, Daniel Borkmann, Jamal Hadi Salim,
Alexander Duyck, John Fastabend, Dave Taht, toke
In-Reply-To: <1412080482.30721.68.camel@edumazet-glaptop2.roam.corp.google.com>
From: Eric Dumazet <edumazet@google.com>
Add a new dql attribute, to control how much packets we are allowed to
burst from qdisc to device.
Signed-off-by: Eric Dumazet <edumazet@google.com>
---
Jesper, please consider using this instead of 7 or 8 values you hard coded.
include/linux/dynamic_queue_limits.h | 1
lib/dynamic_queue_limits.c | 1
net/core/net-sysfs.c | 31 +++++++++++++++++++++++++
3 files changed, 33 insertions(+)
diff --git a/include/linux/dynamic_queue_limits.h b/include/linux/dynamic_queue_limits.h
index a4be70398ce1..8509053eb330 100644
--- a/include/linux/dynamic_queue_limits.h
+++ b/include/linux/dynamic_queue_limits.h
@@ -42,6 +42,7 @@ struct dql {
unsigned int num_queued; /* Total ever queued */
unsigned int adj_limit; /* limit + num_completed */
unsigned int last_obj_cnt; /* Count at last queuing */
+ unsigned int burst;
/* Fields accessed only by completion path (dql_completed) */
diff --git a/lib/dynamic_queue_limits.c b/lib/dynamic_queue_limits.c
index 0777c5a45fa0..acc0d6105191 100644
--- a/lib/dynamic_queue_limits.c
+++ b/lib/dynamic_queue_limits.c
@@ -132,6 +132,7 @@ int dql_init(struct dql *dql, unsigned hold_time)
dql->max_limit = DQL_MAX_LIMIT;
dql->min_limit = 0;
dql->slack_hold_time = hold_time;
+ dql->burst = 7;
dql_reset(dql);
return 0;
}
diff --git a/net/core/net-sysfs.c b/net/core/net-sysfs.c
index 9dd06699b09c..80cf6f166706 100644
--- a/net/core/net-sysfs.c
+++ b/net/core/net-sysfs.c
@@ -978,6 +978,36 @@ static struct netdev_queue_attribute bql_hold_time_attribute =
__ATTR(hold_time, S_IRUGO | S_IWUSR, bql_show_hold_time,
bql_set_hold_time);
+static ssize_t bql_show_burst(struct netdev_queue *queue,
+ struct netdev_queue_attribute *attr,
+ char *buf)
+{
+ const struct dql *dql = &queue->dql;
+
+ return sprintf(buf, "%u\n", dql->burst);
+}
+
+static ssize_t bql_set_burst(struct netdev_queue *queue,
+ struct netdev_queue_attribute *attribute,
+ const char *buf, size_t len)
+{
+ struct dql *dql = &queue->dql;
+ unsigned int value;
+ int err;
+
+ err = kstrtouint(buf, 10, &value);
+ if (err < 0)
+ return err;
+
+ dql->burst = value;
+
+ return len;
+}
+
+static struct netdev_queue_attribute bql_burst_attribute =
+ __ATTR(burst, S_IRUGO | S_IWUSR, bql_show_burst,
+ bql_set_burst);
+
static ssize_t bql_show_inflight(struct netdev_queue *queue,
struct netdev_queue_attribute *attr,
char *buf)
@@ -1019,6 +1049,7 @@ static struct attribute *dql_attrs[] = {
&bql_limit_min_attribute.attr,
&bql_hold_time_attribute.attr,
&bql_inflight_attribute.attr,
+ &bql_burst_attribute.attr,
NULL
};
^ permalink raw reply related
* How exactly does CHECKSUM_COMPLETE works?
From: Yuval Mintz @ 2014-09-30 12:51 UTC (permalink / raw)
To: Tom Herbert (Partner - google), netdev
Hi,
I was pondering what it would take for a network device support
CHECKSUM_COMPLETE on its Rx path. And I simply failed to understand the
specifics when using the napi_gro_receive() interface.
Scenario 1 - TCP/IPv4 packet. This will eventually reach tcp4_gro_receive(), and from
there into skb_gro_checksum_validate().
Since this SKB is marked as COMPLETE, the pseudo TCP checksum will be computed
and skb_gro_checksum_validate_complete() be called.
Now, it seems to me as if the only way by which the flow won't actually try to
re-calculate the checksum over TCP header and payload is if NAPI_GRO_CB(skb)->csum
is actually the one's complement to the TCP pseudo header. But that sounds.... odd,
since that's the csum that should have been listed on the SKB - and it doesn't protect
any of the data, only the "metadata".
Scenario 2 - IP-GRE packet, with CSUM option on the GRE header.
This scenario will also eventually reach skb_gro_checksum_validate_complete(),
this time via gre_gro_receive().
This one is even odder, since in order for the flow not to re-calculate the checksum
The driver should have set the SKB's csum to 0xffffffff [as it's using the
null_compute_pseudo() which will return a pseudo_csum of 0].
>From the comment in skbuff.h regarding CHECKSUM_COMPLETE it looks as if all
the HW needs to do is calculate the checksum over the entire data, starting at the
network header.
So, to summarize my questions -
1. What should a driver set as the SKBs csum value when passing CHECKSUM_COMPLETE?
2. Does the GRO flow for non-encapsulated packets work for packets marked as
CHECKSUM_COMPLETE? And by work, I obviously mean 'doesn't re-calculate checksums'.
3. Same as question 2, but for encapsulated packets.
Thanks,
Yuval
________________________________
This message and any attached documents contain information from QLogic Corporation or its wholly-owned subsidiaries that may be confidential. If you are not the intended recipient, you may not read, copy, distribute, or use this information. If you have received this transmission in error, please notify the sender immediately by reply e-mail and then delete this message.
^ permalink raw reply
* Re: [PATCH] net: ethernet : stmicro: fixed power suspend and resume failure in stmmac driver
From: Giuseppe CAVALLARO @ 2014-09-30 13:00 UTC (permalink / raw)
To: hliang1025; +Cc: netdev, linux-kernel, adi-linux-docs
In-Reply-To: <1412056534-7995-1-git-send-email-hliang1025@gmail.com>
Hello Hao Liang
On 9/30/2014 7:55 AM, hliang1025@gmail.com wrote:
> From: Hao Liang <hliang1025@gmail.com>
>
> This is the fix for a power management issue caused by suspend and resume function in stmmac_main.c.
> After enable CONFIG_DEBUG_ATOMIC_SLEEP which enable sleep-inside atomic section checking, power
> managemet can not work normally. Board couldn't wakeup successfully after suspend. Command
> "echo mem > /sys/power/state" suspend the board.
>
> In suspend and resume function of stmmac driver, there are some sleep-inside function in atomic section
> created by spin lock. These functions will causes system warnings and wakeup issue when enable
> CONFIG_DEBUG_ATOMIC_SLEEP.
>
> This bug was fixed by:
> * replace some sleep function with non-sleep function
> clk_disable_unprepare -> clk_disable ...
> * decrease the atomic area created by spin lock function. The original atomic area in resume function is
> too large.
I had done something in this patch:
[PATCH (net.git)] stmmac: fix and review whole driver locking
Indeed I was not able to review it after some advice but I will do it!
So we can review your patch and then I will re-base my work on top of
yours.
I just wonder if the spinlock can be safely removed instead of
surrounding just the code to clear the descriptors that should
not be claimed by others. Maybe the only part shared is the
stmmac_hw_setup that can be invoked by open method.
peppe
> Signed-off-by: Hao Liang <hliang1025@gmail.com>
> ---
> drivers/net/ethernet/stmicro/stmmac/stmmac_main.c | 13 +++++++------
> 1 file changed, 7 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> index 6e6ee22..2effbfe 100644
> --- a/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> +++ b/drivers/net/ethernet/stmicro/stmmac/stmmac_main.c
> @@ -2924,9 +2924,8 @@ int stmmac_suspend(struct net_device *ndev)
> if (priv->phydev)
> phy_stop(priv->phydev);
>
> - spin_lock_irqsave(&priv->lock, flags);
> -
> netif_device_detach(ndev);
> +
> netif_stop_queue(ndev);
>
> napi_disable(&priv->napi);
> @@ -2935,6 +2934,8 @@ int stmmac_suspend(struct net_device *ndev)
> priv->hw->dma->stop_tx(priv->ioaddr);
> priv->hw->dma->stop_rx(priv->ioaddr);
>
> + spin_lock_irqsave(&priv->lock, flags);
> +
> stmmac_clear_descriptors(priv);
>
> /* Enable Power down mode by programming the PMT regs */
> @@ -2945,7 +2946,7 @@ int stmmac_suspend(struct net_device *ndev)
> stmmac_set_mac(priv->ioaddr, false);
> pinctrl_pm_select_sleep_state(priv->device);
> /* Disable clock in case of PWM is off */
> - clk_disable_unprepare(priv->stmmac_clk);
> + clk_disable(priv->stmmac_clk);
> }
> spin_unlock_irqrestore(&priv->lock, flags);
>
> @@ -2977,12 +2978,14 @@ int stmmac_resume(struct net_device *ndev)
> } else {
> pinctrl_pm_select_default_state(priv->device);
> /* enable the clk prevously disabled */
> - clk_prepare_enable(priv->stmmac_clk);
> + clk_enable(priv->stmmac_clk);
> /* reset the phy so that it's ready */
> if (priv->mii)
> stmmac_mdio_reset(priv->mii);
> }
>
> + spin_unlock_irqrestore(&priv->lock, flags);
> +
> netif_device_attach(ndev);
>
> stmmac_hw_setup(ndev);
> @@ -2991,8 +2994,6 @@ int stmmac_resume(struct net_device *ndev)
>
> netif_start_queue(ndev);
>
> - spin_unlock_irqrestore(&priv->lock, flags);
> -
> if (priv->phydev)
> phy_start(priv->phydev);
>
>
^ permalink raw reply
* [PATCH net-next 0/4] mlx5 update for 3.18
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
Hi Dave,
This series integrates a new mechanism for populating and extracting field values
used in the driver/firmware interaction around command mailboxes.
Eli.
Eli Cohen (4):
net/mlx5_core: Update device capabilities handling
net/mlx5_core: Use hardware registers description header file
net/mlx5_core: use set/get macros in device caps
net/mlx5_core: Identify resources by their type
drivers/infiniband/hw/mlx5/cq.c | 8 +-
drivers/infiniband/hw/mlx5/mad.c | 2 +-
drivers/infiniband/hw/mlx5/main.c | 83 +-
drivers/infiniband/hw/mlx5/qp.c | 72 +-
drivers/infiniband/hw/mlx5/srq.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 77 +-
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 14 +-
drivers/net/ethernet/mellanox/mlx5/core/fw.c | 81 +-
drivers/net/ethernet/mellanox/mlx5/core/main.c | 229 +-
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 60 +-
drivers/net/ethernet/mellanox/mlx5/core/uar.c | 4 +-
include/linux/mlx5/device.h | 182 +-
include/linux/mlx5/driver.h | 118 +-
include/linux/mlx5/mlx5_ifc.h | 4464 ++++++++++++++++++++++++
include/linux/mlx5/qp.h | 3 +-
15 files changed, 4948 insertions(+), 455 deletions(-)
create mode 100644 include/linux/mlx5/mlx5_ifc.h
--
2.1.1
^ permalink raw reply
* [PATCH net-next 0/4] mlx5 update for 3.18
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412082135-30229-1-git-send-email-eli@mellanox.com>
Hi Dave,
This series integrates a new mechanism for populating and extracting field values
used in the driver/firmware interaction around command mailboxes.
Eli.
Eli Cohen (4):
net/mlx5_core: Update device capabilities handling
net/mlx5_core: Use hardware registers description header file
net/mlx5_core: use set/get macros in device caps
net/mlx5_core: Identify resources by their type
drivers/infiniband/hw/mlx5/cq.c | 8 +-
drivers/infiniband/hw/mlx5/mad.c | 2 +-
drivers/infiniband/hw/mlx5/main.c | 83 +-
drivers/infiniband/hw/mlx5/qp.c | 72 +-
drivers/infiniband/hw/mlx5/srq.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 77 +-
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 14 +-
drivers/net/ethernet/mellanox/mlx5/core/fw.c | 81 +-
drivers/net/ethernet/mellanox/mlx5/core/main.c | 229 +-
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 60 +-
drivers/net/ethernet/mellanox/mlx5/core/uar.c | 4 +-
include/linux/mlx5/device.h | 182 +-
include/linux/mlx5/driver.h | 118 +-
include/linux/mlx5/mlx5_ifc.h | 4464 ++++++++++++++++++++++++
include/linux/mlx5/qp.h | 3 +-
15 files changed, 4948 insertions(+), 455 deletions(-)
create mode 100644 include/linux/mlx5/mlx5_ifc.h
--
2.1.1
^ permalink raw reply
* [PATCH net-next 1/4] net/mlx5_core: Update device capabilities handling
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412082135-30229-1-git-send-email-eli@mellanox.com>
Rearrange struct mlx5_caps so it has a "gen" field to represent the current
capabilities configured for the device. Max capabilities can also be queried
from the device. Also update capabilities struct to contain more fields as per
the latest revision if firmware specification.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/infiniband/hw/mlx5/cq.c | 8 +-
drivers/infiniband/hw/mlx5/mad.c | 2 +-
drivers/infiniband/hw/mlx5/main.c | 83 ++++++----
drivers/infiniband/hw/mlx5/qp.c | 72 +++++---
drivers/infiniband/hw/mlx5/srq.c | 6 +-
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 24 ++-
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 2 +-
drivers/net/ethernet/mellanox/mlx5/core/fw.c | 81 +--------
drivers/net/ethernet/mellanox/mlx5/core/main.c | 219 +++++++++++++++++++------
drivers/net/ethernet/mellanox/mlx5/core/uar.c | 4 +-
include/linux/mlx5/device.h | 24 ++-
include/linux/mlx5/driver.h | 28 +++-
12 files changed, 331 insertions(+), 222 deletions(-)
diff --git a/drivers/infiniband/hw/mlx5/cq.c b/drivers/infiniband/hw/mlx5/cq.c
index e4056279166d..10cfce5119a9 100644
--- a/drivers/infiniband/hw/mlx5/cq.c
+++ b/drivers/infiniband/hw/mlx5/cq.c
@@ -752,7 +752,7 @@ struct ib_cq *mlx5_ib_create_cq(struct ib_device *ibdev, int entries,
return ERR_PTR(-EINVAL);
entries = roundup_pow_of_two(entries + 1);
- if (entries > dev->mdev->caps.max_cqes)
+ if (entries > dev->mdev->caps.gen.max_cqes)
return ERR_PTR(-EINVAL);
cq = kzalloc(sizeof(*cq), GFP_KERNEL);
@@ -919,7 +919,7 @@ int mlx5_ib_modify_cq(struct ib_cq *cq, u16 cq_count, u16 cq_period)
int err;
u32 fsel;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
+ if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_CQ_MODER))
return -ENOSYS;
in = kzalloc(sizeof(*in), GFP_KERNEL);
@@ -1074,7 +1074,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
int uninitialized_var(cqe_size);
unsigned long flags;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
+ if (!(dev->mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_RESIZE_CQ)) {
pr_info("Firmware does not support resize CQ\n");
return -ENOSYS;
}
@@ -1083,7 +1083,7 @@ int mlx5_ib_resize_cq(struct ib_cq *ibcq, int entries, struct ib_udata *udata)
return -EINVAL;
entries = roundup_pow_of_two(entries + 1);
- if (entries > dev->mdev->caps.max_cqes + 1)
+ if (entries > dev->mdev->caps.gen.max_cqes + 1)
return -EINVAL;
if (entries == ibcq->cqe + 1)
diff --git a/drivers/infiniband/hw/mlx5/mad.c b/drivers/infiniband/hw/mlx5/mad.c
index b514bbb5610f..657af9a1167c 100644
--- a/drivers/infiniband/hw/mlx5/mad.c
+++ b/drivers/infiniband/hw/mlx5/mad.c
@@ -129,7 +129,7 @@ int mlx5_query_ext_port_caps(struct mlx5_ib_dev *dev, u8 port)
packet_error = be16_to_cpu(out_mad->status);
- dev->mdev->caps.ext_port_cap[port - 1] = (!err && !packet_error) ?
+ dev->mdev->caps.gen.ext_port_cap[port - 1] = (!err && !packet_error) ?
MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO : 0;
out:
diff --git a/drivers/infiniband/hw/mlx5/main.c b/drivers/infiniband/hw/mlx5/main.c
index d8907b20522a..f3114d1132fb 100644
--- a/drivers/infiniband/hw/mlx5/main.c
+++ b/drivers/infiniband/hw/mlx5/main.c
@@ -157,11 +157,13 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
struct mlx5_ib_dev *dev = to_mdev(ibdev);
struct ib_smp *in_mad = NULL;
struct ib_smp *out_mad = NULL;
+ struct mlx5_general_caps *gen;
int err = -ENOMEM;
int max_rq_sg;
int max_sq_sg;
u64 flags;
+ gen = &dev->mdev->caps.gen;
in_mad = kzalloc(sizeof(*in_mad), GFP_KERNEL);
out_mad = kmalloc(sizeof(*out_mad), GFP_KERNEL);
if (!in_mad || !out_mad)
@@ -183,7 +185,7 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
IB_DEVICE_PORT_ACTIVE_EVENT |
IB_DEVICE_SYS_IMAGE_GUID |
IB_DEVICE_RC_RNR_NAK_GEN;
- flags = dev->mdev->caps.flags;
+ flags = gen->flags;
if (flags & MLX5_DEV_CAP_FLAG_BAD_PKEY_CNTR)
props->device_cap_flags |= IB_DEVICE_BAD_PKEY_CNTR;
if (flags & MLX5_DEV_CAP_FLAG_BAD_QKEY_CNTR)
@@ -213,30 +215,31 @@ static int mlx5_ib_query_device(struct ib_device *ibdev,
memcpy(&props->sys_image_guid, out_mad->data + 4, 8);
props->max_mr_size = ~0ull;
- props->page_size_cap = dev->mdev->caps.min_page_sz;
- props->max_qp = 1 << dev->mdev->caps.log_max_qp;
- props->max_qp_wr = dev->mdev->caps.max_wqes;
- max_rq_sg = dev->mdev->caps.max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
- max_sq_sg = (dev->mdev->caps.max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
+ props->page_size_cap = gen->min_page_sz;
+ props->max_qp = 1 << gen->log_max_qp;
+ props->max_qp_wr = gen->max_wqes;
+ max_rq_sg = gen->max_rq_desc_sz / sizeof(struct mlx5_wqe_data_seg);
+ max_sq_sg = (gen->max_sq_desc_sz - sizeof(struct mlx5_wqe_ctrl_seg)) /
sizeof(struct mlx5_wqe_data_seg);
props->max_sge = min(max_rq_sg, max_sq_sg);
- props->max_cq = 1 << dev->mdev->caps.log_max_cq;
- props->max_cqe = dev->mdev->caps.max_cqes - 1;
- props->max_mr = 1 << dev->mdev->caps.log_max_mkey;
- props->max_pd = 1 << dev->mdev->caps.log_max_pd;
- props->max_qp_rd_atom = dev->mdev->caps.max_ra_req_qp;
- props->max_qp_init_rd_atom = dev->mdev->caps.max_ra_res_qp;
+ props->max_cq = 1 << gen->log_max_cq;
+ props->max_cqe = gen->max_cqes - 1;
+ props->max_mr = 1 << gen->log_max_mkey;
+ props->max_pd = 1 << gen->log_max_pd;
+ props->max_qp_rd_atom = 1 << gen->log_max_ra_req_qp;
+ props->max_qp_init_rd_atom = 1 << gen->log_max_ra_res_qp;
+ props->max_srq = 1 << gen->log_max_srq;
+ props->max_srq_wr = gen->max_srq_wqes - 1;
+ props->local_ca_ack_delay = gen->local_ca_ack_delay;
props->max_res_rd_atom = props->max_qp_rd_atom * props->max_qp;
- props->max_srq = 1 << dev->mdev->caps.log_max_srq;
- props->max_srq_wr = dev->mdev->caps.max_srq_wqes - 1;
props->max_srq_sge = max_rq_sg - 1;
props->max_fast_reg_page_list_len = (unsigned int)-1;
- props->local_ca_ack_delay = dev->mdev->caps.local_ca_ack_delay;
+ props->local_ca_ack_delay = gen->local_ca_ack_delay;
props->atomic_cap = IB_ATOMIC_NONE;
props->masked_atomic_cap = IB_ATOMIC_NONE;
props->max_pkeys = be16_to_cpup((__be16 *)(out_mad->data + 28));
- props->max_mcast_grp = 1 << dev->mdev->caps.log_max_mcg;
- props->max_mcast_qp_attach = dev->mdev->caps.max_qp_mcg;
+ props->max_mcast_grp = 1 << gen->log_max_mcg;
+ props->max_mcast_qp_attach = gen->max_qp_mcg;
props->max_total_mcast_qp_attach = props->max_mcast_qp_attach *
props->max_mcast_grp;
props->max_map_per_fmr = INT_MAX; /* no limit in ConnectIB */
@@ -254,10 +257,12 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
struct mlx5_ib_dev *dev = to_mdev(ibdev);
struct ib_smp *in_mad = NULL;
struct ib_smp *out_mad = NULL;
+ struct mlx5_general_caps *gen;
int ext_active_speed;
int err = -ENOMEM;
- if (port < 1 || port > dev->mdev->caps.num_ports) {
+ gen = &dev->mdev->caps.gen;
+ if (port < 1 || port > gen->num_ports) {
mlx5_ib_warn(dev, "invalid port number %d\n", port);
return -EINVAL;
}
@@ -288,8 +293,8 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
props->phys_state = out_mad->data[33] >> 4;
props->port_cap_flags = be32_to_cpup((__be32 *)(out_mad->data + 20));
props->gid_tbl_len = out_mad->data[50];
- props->max_msg_sz = 1 << to_mdev(ibdev)->mdev->caps.log_max_msg;
- props->pkey_tbl_len = to_mdev(ibdev)->mdev->caps.port[port - 1].pkey_table_len;
+ props->max_msg_sz = 1 << gen->log_max_msg;
+ props->pkey_tbl_len = gen->port[port - 1].pkey_table_len;
props->bad_pkey_cntr = be16_to_cpup((__be16 *)(out_mad->data + 46));
props->qkey_viol_cntr = be16_to_cpup((__be16 *)(out_mad->data + 48));
props->active_width = out_mad->data[31] & 0xf;
@@ -316,7 +321,7 @@ int mlx5_ib_query_port(struct ib_device *ibdev, u8 port,
/* If reported active speed is QDR, check if is FDR-10 */
if (props->active_speed == 4) {
- if (dev->mdev->caps.ext_port_cap[port - 1] &
+ if (gen->ext_port_cap[port - 1] &
MLX_EXT_PORT_CAP_FLAG_EXTENDED_PORT_INFO) {
init_query_mad(in_mad);
in_mad->attr_id = MLX5_ATTR_EXTENDED_PORT_INFO;
@@ -470,6 +475,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
struct mlx5_ib_alloc_ucontext_req_v2 req;
struct mlx5_ib_alloc_ucontext_resp resp;
struct mlx5_ib_ucontext *context;
+ struct mlx5_general_caps *gen;
struct mlx5_uuar_info *uuari;
struct mlx5_uar *uars;
int gross_uuars;
@@ -480,6 +486,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
int i;
size_t reqlen;
+ gen = &dev->mdev->caps.gen;
if (!dev->ib_active)
return ERR_PTR(-EAGAIN);
@@ -512,14 +519,14 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
num_uars = req.total_num_uuars / MLX5_NON_FP_BF_REGS_PER_PAGE;
gross_uuars = num_uars * MLX5_BF_REGS_PER_PAGE;
- resp.qp_tab_size = 1 << dev->mdev->caps.log_max_qp;
- resp.bf_reg_size = dev->mdev->caps.bf_reg_size;
+ resp.qp_tab_size = 1 << gen->log_max_qp;
+ resp.bf_reg_size = gen->bf_reg_size;
resp.cache_line_size = L1_CACHE_BYTES;
- resp.max_sq_desc_sz = dev->mdev->caps.max_sq_desc_sz;
- resp.max_rq_desc_sz = dev->mdev->caps.max_rq_desc_sz;
- resp.max_send_wqebb = dev->mdev->caps.max_wqes;
- resp.max_recv_wr = dev->mdev->caps.max_wqes;
- resp.max_srq_recv_wr = dev->mdev->caps.max_srq_wqes;
+ resp.max_sq_desc_sz = gen->max_sq_desc_sz;
+ resp.max_rq_desc_sz = gen->max_rq_desc_sz;
+ resp.max_send_wqebb = gen->max_wqes;
+ resp.max_recv_wr = gen->max_wqes;
+ resp.max_srq_recv_wr = gen->max_srq_wqes;
context = kzalloc(sizeof(*context), GFP_KERNEL);
if (!context)
@@ -565,7 +572,7 @@ static struct ib_ucontext *mlx5_ib_alloc_ucontext(struct ib_device *ibdev,
mutex_init(&context->db_page_mutex);
resp.tot_uuars = req.total_num_uuars;
- resp.num_ports = dev->mdev->caps.num_ports;
+ resp.num_ports = gen->num_ports;
err = ib_copy_to_udata(udata, &resp,
sizeof(resp) - sizeof(resp.reserved));
if (err)
@@ -967,9 +974,11 @@ static void mlx5_ib_event(struct mlx5_core_dev *dev, void *context,
static void get_ext_port_caps(struct mlx5_ib_dev *dev)
{
+ struct mlx5_general_caps *gen;
int port;
- for (port = 1; port <= dev->mdev->caps.num_ports; port++)
+ gen = &dev->mdev->caps.gen;
+ for (port = 1; port <= gen->num_ports; port++)
mlx5_query_ext_port_caps(dev, port);
}
@@ -977,9 +986,11 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
{
struct ib_device_attr *dprops = NULL;
struct ib_port_attr *pprops = NULL;
+ struct mlx5_general_caps *gen;
int err = 0;
int port;
+ gen = &dev->mdev->caps.gen;
pprops = kmalloc(sizeof(*pprops), GFP_KERNEL);
if (!pprops)
goto out;
@@ -994,14 +1005,14 @@ static int get_port_caps(struct mlx5_ib_dev *dev)
goto out;
}
- for (port = 1; port <= dev->mdev->caps.num_ports; port++) {
+ for (port = 1; port <= gen->num_ports; port++) {
err = mlx5_ib_query_port(&dev->ib_dev, port, pprops);
if (err) {
mlx5_ib_warn(dev, "query_port %d failed %d\n", port, err);
break;
}
- dev->mdev->caps.port[port - 1].pkey_table_len = dprops->max_pkeys;
- dev->mdev->caps.port[port - 1].gid_table_len = pprops->gid_tbl_len;
+ gen->port[port - 1].pkey_table_len = dprops->max_pkeys;
+ gen->port[port - 1].gid_table_len = pprops->gid_tbl_len;
mlx5_ib_dbg(dev, "pkey_table_len %d, gid_table_len %d\n",
dprops->max_pkeys, pprops->gid_tbl_len);
}
@@ -1279,8 +1290,8 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
strlcpy(dev->ib_dev.name, "mlx5_%d", IB_DEVICE_NAME_MAX);
dev->ib_dev.owner = THIS_MODULE;
dev->ib_dev.node_type = RDMA_NODE_IB_CA;
- dev->ib_dev.local_dma_lkey = mdev->caps.reserved_lkey;
- dev->num_ports = mdev->caps.num_ports;
+ dev->ib_dev.local_dma_lkey = mdev->caps.gen.reserved_lkey;
+ dev->num_ports = mdev->caps.gen.num_ports;
dev->ib_dev.phys_port_cnt = dev->num_ports;
dev->ib_dev.num_comp_vectors = dev->num_comp_vectors;
dev->ib_dev.dma_device = &mdev->pdev->dev;
@@ -1355,7 +1366,7 @@ static void *mlx5_ib_add(struct mlx5_core_dev *mdev)
dev->ib_dev.free_fast_reg_page_list = mlx5_ib_free_fast_reg_page_list;
dev->ib_dev.check_mr_status = mlx5_ib_check_mr_status;
- if (mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC) {
+ if (mdev->caps.gen.flags & MLX5_DEV_CAP_FLAG_XRC) {
dev->ib_dev.alloc_xrcd = mlx5_ib_alloc_xrcd;
dev->ib_dev.dealloc_xrcd = mlx5_ib_dealloc_xrcd;
dev->ib_dev.uverbs_cmd_mask |=
diff --git a/drivers/infiniband/hw/mlx5/qp.c b/drivers/infiniband/hw/mlx5/qp.c
index 8c574b63d77b..dbfe498870c1 100644
--- a/drivers/infiniband/hw/mlx5/qp.c
+++ b/drivers/infiniband/hw/mlx5/qp.c
@@ -158,11 +158,13 @@ static void mlx5_ib_qp_event(struct mlx5_core_qp *qp, int type)
static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
int has_rq, struct mlx5_ib_qp *qp, struct mlx5_ib_create_qp *ucmd)
{
+ struct mlx5_general_caps *gen;
int wqe_size;
int wq_size;
+ gen = &dev->mdev->caps.gen;
/* Sanity check RQ size before proceeding */
- if (cap->max_recv_wr > dev->mdev->caps.max_wqes)
+ if (cap->max_recv_wr > gen->max_wqes)
return -EINVAL;
if (!has_rq) {
@@ -182,10 +184,10 @@ static int set_rq_size(struct mlx5_ib_dev *dev, struct ib_qp_cap *cap,
wq_size = roundup_pow_of_two(cap->max_recv_wr) * wqe_size;
wq_size = max_t(int, wq_size, MLX5_SEND_WQE_BB);
qp->rq.wqe_cnt = wq_size / wqe_size;
- if (wqe_size > dev->mdev->caps.max_rq_desc_sz) {
+ if (wqe_size > gen->max_rq_desc_sz) {
mlx5_ib_dbg(dev, "wqe_size %d, max %d\n",
wqe_size,
- dev->mdev->caps.max_rq_desc_sz);
+ gen->max_rq_desc_sz);
return -EINVAL;
}
qp->rq.wqe_shift = ilog2(wqe_size);
@@ -266,9 +268,11 @@ static int calc_send_wqe(struct ib_qp_init_attr *attr)
static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
struct mlx5_ib_qp *qp)
{
+ struct mlx5_general_caps *gen;
int wqe_size;
int wq_size;
+ gen = &dev->mdev->caps.gen;
if (!attr->cap.max_send_wr)
return 0;
@@ -277,9 +281,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
if (wqe_size < 0)
return wqe_size;
- if (wqe_size > dev->mdev->caps.max_sq_desc_sz) {
+ if (wqe_size > gen->max_sq_desc_sz) {
mlx5_ib_dbg(dev, "wqe_size(%d) > max_sq_desc_sz(%d)\n",
- wqe_size, dev->mdev->caps.max_sq_desc_sz);
+ wqe_size, gen->max_sq_desc_sz);
return -EINVAL;
}
@@ -292,9 +296,9 @@ static int calc_sq_size(struct mlx5_ib_dev *dev, struct ib_qp_init_attr *attr,
wq_size = roundup_pow_of_two(attr->cap.max_send_wr * wqe_size);
qp->sq.wqe_cnt = wq_size / MLX5_SEND_WQE_BB;
- if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+ if (qp->sq.wqe_cnt > gen->max_wqes) {
mlx5_ib_dbg(dev, "wqe count(%d) exceeds limits(%d)\n",
- qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+ qp->sq.wqe_cnt, gen->max_wqes);
return -ENOMEM;
}
qp->sq.wqe_shift = ilog2(MLX5_SEND_WQE_BB);
@@ -309,11 +313,13 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
struct mlx5_ib_qp *qp,
struct mlx5_ib_create_qp *ucmd)
{
+ struct mlx5_general_caps *gen;
int desc_sz = 1 << qp->sq.wqe_shift;
- if (desc_sz > dev->mdev->caps.max_sq_desc_sz) {
+ gen = &dev->mdev->caps.gen;
+ if (desc_sz > gen->max_sq_desc_sz) {
mlx5_ib_warn(dev, "desc_sz %d, max_sq_desc_sz %d\n",
- desc_sz, dev->mdev->caps.max_sq_desc_sz);
+ desc_sz, gen->max_sq_desc_sz);
return -EINVAL;
}
@@ -325,9 +331,9 @@ static int set_user_buf_size(struct mlx5_ib_dev *dev,
qp->sq.wqe_cnt = ucmd->sq_wqe_count;
- if (qp->sq.wqe_cnt > dev->mdev->caps.max_wqes) {
+ if (qp->sq.wqe_cnt > gen->max_wqes) {
mlx5_ib_warn(dev, "wqe_cnt %d, max_wqes %d\n",
- qp->sq.wqe_cnt, dev->mdev->caps.max_wqes);
+ qp->sq.wqe_cnt, gen->max_wqes);
return -EINVAL;
}
@@ -803,16 +809,18 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
struct mlx5_ib_resources *devr = &dev->devr;
struct mlx5_ib_create_qp_resp resp;
struct mlx5_create_qp_mbox_in *in;
+ struct mlx5_general_caps *gen;
struct mlx5_ib_create_qp ucmd;
int inlen = sizeof(*in);
int err;
+ gen = &dev->mdev->caps.gen;
mutex_init(&qp->mutex);
spin_lock_init(&qp->sq.lock);
spin_lock_init(&qp->rq.lock);
if (init_attr->create_flags & IB_QP_CREATE_BLOCK_MULTICAST_LOOPBACK) {
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_BLOCK_MCAST)) {
mlx5_ib_dbg(dev, "block multicast loopback isn't supported\n");
return -EINVAL;
} else {
@@ -851,9 +859,9 @@ static int create_qp_common(struct mlx5_ib_dev *dev, struct ib_pd *pd,
mlx5_ib_dbg(dev, "invalid rq params\n");
return -EINVAL;
}
- if (ucmd.sq_wqe_count > dev->mdev->caps.max_wqes) {
+ if (ucmd.sq_wqe_count > gen->max_wqes) {
mlx5_ib_dbg(dev, "requested sq_wqe_count (%d) > max allowed (%d)\n",
- ucmd.sq_wqe_count, dev->mdev->caps.max_wqes);
+ ucmd.sq_wqe_count, gen->max_wqes);
return -EINVAL;
}
err = create_user_qp(dev, pd, qp, udata, &in, &resp, &inlen);
@@ -1144,6 +1152,7 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
struct ib_qp_init_attr *init_attr,
struct ib_udata *udata)
{
+ struct mlx5_general_caps *gen;
struct mlx5_ib_dev *dev;
struct mlx5_ib_qp *qp;
u16 xrcdn = 0;
@@ -1161,11 +1170,12 @@ struct ib_qp *mlx5_ib_create_qp(struct ib_pd *pd,
}
dev = to_mdev(to_mxrcd(init_attr->xrcd)->ibxrcd.device);
}
+ gen = &dev->mdev->caps.gen;
switch (init_attr->qp_type) {
case IB_QPT_XRC_TGT:
case IB_QPT_XRC_INI:
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC)) {
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC)) {
mlx5_ib_dbg(dev, "XRC not supported\n");
return ERR_PTR(-ENOSYS);
}
@@ -1272,6 +1282,9 @@ enum {
static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
{
+ struct mlx5_general_caps *gen;
+
+ gen = &dev->mdev->caps.gen;
if (rate == IB_RATE_PORT_CURRENT) {
return 0;
} else if (rate < IB_RATE_2_5_GBPS || rate > IB_RATE_300_GBPS) {
@@ -1279,7 +1292,7 @@ static int ib_rate_to_mlx5(struct mlx5_ib_dev *dev, u8 rate)
} else {
while (rate != IB_RATE_2_5_GBPS &&
!(1 << (rate + MLX5_STAT_RATE_OFFSET) &
- dev->mdev->caps.stat_rate_support))
+ gen->stat_rate_support))
--rate;
}
@@ -1290,8 +1303,10 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
struct mlx5_qp_path *path, u8 port, int attr_mask,
u32 path_flags, const struct ib_qp_attr *attr)
{
+ struct mlx5_general_caps *gen;
int err;
+ gen = &dev->mdev->caps.gen;
path->fl = (path_flags & MLX5_PATH_FLAG_FL) ? 0x80 : 0;
path->free_ar = (path_flags & MLX5_PATH_FLAG_FREE_AR) ? 0x80 : 0;
@@ -1318,9 +1333,9 @@ static int mlx5_set_path(struct mlx5_ib_dev *dev, const struct ib_ah_attr *ah,
path->port = port;
if (ah->ah_flags & IB_AH_GRH) {
- if (ah->grh.sgid_index >= dev->mdev->caps.port[port - 1].gid_table_len) {
+ if (ah->grh.sgid_index >= gen->port[port - 1].gid_table_len) {
pr_err(KERN_ERR "sgid_index (%u) too large. max is %d\n",
- ah->grh.sgid_index, dev->mdev->caps.port[port - 1].gid_table_len);
+ ah->grh.sgid_index, gen->port[port - 1].gid_table_len);
return -EINVAL;
}
@@ -1492,6 +1507,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
struct mlx5_ib_qp *qp = to_mqp(ibqp);
struct mlx5_ib_cq *send_cq, *recv_cq;
struct mlx5_qp_context *context;
+ struct mlx5_general_caps *gen;
struct mlx5_modify_qp_mbox_in *in;
struct mlx5_ib_pd *pd;
enum mlx5_qp_state mlx5_cur, mlx5_new;
@@ -1500,6 +1516,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
int mlx5_st;
int err;
+ gen = &dev->mdev->caps.gen;
in = kzalloc(sizeof(*in), GFP_KERNEL);
if (!in)
return -ENOMEM;
@@ -1539,7 +1556,7 @@ static int __mlx5_ib_modify_qp(struct ib_qp *ibqp,
err = -EINVAL;
goto out;
}
- context->mtu_msgmax = (attr->path_mtu << 5) | dev->mdev->caps.log_max_msg;
+ context->mtu_msgmax = (attr->path_mtu << 5) | gen->log_max_msg;
}
if (attr_mask & IB_QP_DEST_QPN)
@@ -1685,9 +1702,11 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
struct mlx5_ib_dev *dev = to_mdev(ibqp->device);
struct mlx5_ib_qp *qp = to_mqp(ibqp);
enum ib_qp_state cur_state, new_state;
+ struct mlx5_general_caps *gen;
int err = -EINVAL;
int port;
+ gen = &dev->mdev->caps.gen;
mutex_lock(&qp->mutex);
cur_state = attr_mask & IB_QP_CUR_STATE ? attr->cur_qp_state : qp->state;
@@ -1699,21 +1718,21 @@ int mlx5_ib_modify_qp(struct ib_qp *ibqp, struct ib_qp_attr *attr,
goto out;
if ((attr_mask & IB_QP_PORT) &&
- (attr->port_num == 0 || attr->port_num > dev->mdev->caps.num_ports))
+ (attr->port_num == 0 || attr->port_num > gen->num_ports))
goto out;
if (attr_mask & IB_QP_PKEY_INDEX) {
port = attr_mask & IB_QP_PORT ? attr->port_num : qp->port;
- if (attr->pkey_index >= dev->mdev->caps.port[port - 1].pkey_table_len)
+ if (attr->pkey_index >= gen->port[port - 1].pkey_table_len)
goto out;
}
if (attr_mask & IB_QP_MAX_QP_RD_ATOMIC &&
- attr->max_rd_atomic > dev->mdev->caps.max_ra_res_qp)
+ attr->max_rd_atomic > (1 << gen->log_max_ra_res_qp))
goto out;
if (attr_mask & IB_QP_MAX_DEST_RD_ATOMIC &&
- attr->max_dest_rd_atomic > dev->mdev->caps.max_ra_req_qp)
+ attr->max_dest_rd_atomic > (1 << gen->log_max_ra_req_qp))
goto out;
if (cur_state == new_state && cur_state == IB_QPS_RESET) {
@@ -2893,7 +2912,8 @@ static void to_ib_ah_attr(struct mlx5_ib_dev *ibdev, struct ib_ah_attr *ib_ah_at
memset(ib_ah_attr, 0, sizeof(*ib_ah_attr));
ib_ah_attr->port_num = path->port;
- if (ib_ah_attr->port_num == 0 || ib_ah_attr->port_num > dev->caps.num_ports)
+ if (ib_ah_attr->port_num == 0 ||
+ ib_ah_attr->port_num > dev->caps.gen.num_ports)
return;
ib_ah_attr->sl = path->sl & 0xf;
@@ -3011,10 +3031,12 @@ struct ib_xrcd *mlx5_ib_alloc_xrcd(struct ib_device *ibdev,
struct ib_udata *udata)
{
struct mlx5_ib_dev *dev = to_mdev(ibdev);
+ struct mlx5_general_caps *gen;
struct mlx5_ib_xrcd *xrcd;
int err;
- if (!(dev->mdev->caps.flags & MLX5_DEV_CAP_FLAG_XRC))
+ gen = &dev->mdev->caps.gen;
+ if (!(gen->flags & MLX5_DEV_CAP_FLAG_XRC))
return ERR_PTR(-ENOSYS);
xrcd = kmalloc(sizeof(*xrcd), GFP_KERNEL);
diff --git a/drivers/infiniband/hw/mlx5/srq.c b/drivers/infiniband/hw/mlx5/srq.c
index 70bd131ba646..97cc1baaa8e3 100644
--- a/drivers/infiniband/hw/mlx5/srq.c
+++ b/drivers/infiniband/hw/mlx5/srq.c
@@ -238,6 +238,7 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
struct ib_udata *udata)
{
struct mlx5_ib_dev *dev = to_mdev(pd->device);
+ struct mlx5_general_caps *gen;
struct mlx5_ib_srq *srq;
int desc_size;
int buf_size;
@@ -247,11 +248,12 @@ struct ib_srq *mlx5_ib_create_srq(struct ib_pd *pd,
int is_xrc;
u32 flgs, xrcdn;
+ gen = &dev->mdev->caps.gen;
/* Sanity check SRQ size before proceeding */
- if (init_attr->attr.max_wr >= dev->mdev->caps.max_srq_wqes) {
+ if (init_attr->attr.max_wr >= gen->max_srq_wqes) {
mlx5_ib_dbg(dev, "max_wr %d, cap %d\n",
init_attr->attr.max_wr,
- dev->mdev->caps.max_srq_wqes);
+ gen->max_srq_wqes);
return ERR_PTR(-EINVAL);
}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 65a7da69e2ac..6eb0f85cf872 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1538,16 +1538,9 @@ static const char *cmd_status_str(u8 status)
}
}
-int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
+static int cmd_status_to_err(u8 status)
{
- if (!hdr->status)
- return 0;
-
- pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
- cmd_status_str(hdr->status), hdr->status,
- be32_to_cpu(hdr->syndrome));
-
- switch (hdr->status) {
+ switch (status) {
case MLX5_CMD_STAT_OK: return 0;
case MLX5_CMD_STAT_INT_ERR: return -EIO;
case MLX5_CMD_STAT_BAD_OP_ERR: return -EINVAL;
@@ -1567,3 +1560,16 @@ int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
default: return -EIO;
}
}
+
+/* this will be available till all the commands use set/get macros */
+int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
+{
+ if (!hdr->status)
+ return 0;
+
+ pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
+ cmd_status_str(hdr->status), hdr->status,
+ be32_to_cpu(hdr->syndrome));
+
+ return cmd_status_to_err(hdr->status);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 4e8bd0b34bb0..11b9b840ad4d 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -468,7 +468,7 @@ int mlx5_start_eqs(struct mlx5_core_dev *dev)
err = mlx5_create_map_eq(dev, &table->pages_eq,
MLX5_EQ_VEC_PAGES,
- dev->caps.max_vf + 1,
+ dev->caps.gen.max_vf + 1,
1 << MLX5_EVENT_TYPE_PAGE_REQUEST, "mlx5_pages_eq",
&dev->priv.uuari.uars[0]);
if (err) {
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/fw.c b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
index f012658b6a92..087c4c797deb 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/fw.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/fw.c
@@ -64,86 +64,9 @@ out_out:
return err;
}
-int mlx5_cmd_query_hca_cap(struct mlx5_core_dev *dev,
- struct mlx5_caps *caps)
+int mlx5_cmd_query_hca_cap(struct mlx5_core_dev *dev, struct mlx5_caps *caps)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *out;
- struct mlx5_cmd_query_hca_cap_mbox_in in;
- struct mlx5_query_special_ctxs_mbox_out ctx_out;
- struct mlx5_query_special_ctxs_mbox_in ctx_in;
- int err;
- u16 t16;
-
- out = kzalloc(sizeof(*out), GFP_KERNEL);
- if (!out)
- return -ENOMEM;
-
- memset(&in, 0, sizeof(in));
- in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- in.hdr.opmod = cpu_to_be16(0x1);
- err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
- if (err)
- goto out_out;
-
- if (out->hdr.status) {
- err = mlx5_cmd_status_to_err(&out->hdr);
- goto out_out;
- }
-
-
- caps->log_max_eq = out->hca_cap.log_max_eq & 0xf;
- caps->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
- caps->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
- caps->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
- caps->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
- caps->flags = be64_to_cpu(out->hca_cap.flags);
- caps->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
- caps->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
- caps->num_ports = out->hca_cap.num_ports & 0xf;
- caps->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
- if (caps->num_ports > MLX5_MAX_PORTS) {
- mlx5_core_err(dev, "device has %d ports while the driver supports max %d ports\n",
- caps->num_ports, MLX5_MAX_PORTS);
- err = -EINVAL;
- goto out_out;
- }
- caps->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
- caps->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
- caps->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
- caps->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
- caps->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
- caps->log_max_mcg = out->hca_cap.log_max_mcg;
- caps->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
- caps->max_ra_res_qp = 1 << (out->hca_cap.log_max_ra_res_qp & 0x3f);
- caps->max_ra_req_qp = 1 << (out->hca_cap.log_max_ra_req_qp & 0x3f);
- caps->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
- t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
- if (t16 & 0x8000) {
- caps->bf_reg_size = 1 << (t16 & 0x1f);
- caps->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
- } else {
- caps->bf_reg_size = 0;
- caps->bf_regs_per_page = 0;
- }
- caps->min_page_sz = ~(u32)((1 << out->hca_cap.log_pg_sz) - 1);
-
- memset(&ctx_in, 0, sizeof(ctx_in));
- memset(&ctx_out, 0, sizeof(ctx_out));
- ctx_in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS);
- err = mlx5_cmd_exec(dev, &ctx_in, sizeof(ctx_in),
- &ctx_out, sizeof(ctx_out));
- if (err)
- goto out_out;
-
- if (ctx_out.hdr.status)
- err = mlx5_cmd_status_to_err(&ctx_out.hdr);
-
- caps->reserved_lkey = be32_to_cpu(ctx_out.reserved_lkey);
-
-out_out:
- kfree(out);
-
- return err;
+ return mlx5_core_get_caps(dev, caps, HCA_CAP_OPMOD_GET_CUR);
}
int mlx5_cmd_init_hca(struct mlx5_core_dev *dev)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index f2716cc1f51d..d9f74618befa 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -207,11 +207,11 @@ static void release_bar(struct pci_dev *pdev)
static int mlx5_enable_msix(struct mlx5_core_dev *dev)
{
struct mlx5_eq_table *table = &dev->priv.eq_table;
- int num_eqs = 1 << dev->caps.log_max_eq;
+ int num_eqs = 1 << dev->caps.gen.log_max_eq;
int nvec;
int i;
- nvec = dev->caps.num_ports * num_online_cpus() + MLX5_EQ_VEC_COMP_BASE;
+ nvec = dev->caps.gen.num_ports * num_online_cpus() + MLX5_EQ_VEC_COMP_BASE;
nvec = min_t(int, nvec, num_eqs);
if (nvec <= MLX5_EQ_VEC_COMP_BASE)
return -ENOMEM;
@@ -250,13 +250,34 @@ struct mlx5_reg_host_endianess {
#define CAP_MASK(pos, size) ((u64)((1 << (size)) - 1) << (pos))
enum {
- MLX5_CAP_BITS_RW_MASK = CAP_MASK(MLX5_CAP_OFF_CMDIF_CSUM, 2) |
- CAP_MASK(MLX5_CAP_OFF_DCT, 1),
+ MLX5_CAP_BITS_RW_MASK = CAP_MASK(MLX5_CAP_OFF_CMDIF_CSUM, 2) |
+ MLX5_DEV_CAP_FLAG_DCT,
};
+static u16 to_fw_pkey_sz(u32 size)
+{
+ switch (size) {
+ case 128:
+ return 0;
+ case 256:
+ return 1;
+ case 512:
+ return 2;
+ case 1024:
+ return 3;
+ case 2048:
+ return 4;
+ case 4096:
+ return 5;
+ default:
+ pr_warn("invalid pkey table size %d\n", size);
+ return 0;
+ }
+}
+
/* selectively copy writable fields clearing any reserved area
*/
-static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_hca_cap *from)
+static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_general_caps *from)
{
u64 v64;
@@ -265,76 +286,172 @@ static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_hca_cap *from)
to->log_max_ra_res_dc = from->log_max_ra_res_dc & 0x3f;
to->log_max_ra_req_qp = from->log_max_ra_req_qp & 0x3f;
to->log_max_ra_res_qp = from->log_max_ra_res_qp & 0x3f;
- to->log_max_atomic_size_qp = from->log_max_atomic_size_qp;
- to->log_max_atomic_size_dc = from->log_max_atomic_size_dc;
- v64 = be64_to_cpu(from->flags) & MLX5_CAP_BITS_RW_MASK;
+ to->pkey_table_size = cpu_to_be16(to_fw_pkey_sz(from->pkey_table_size));
+ v64 = from->flags & MLX5_CAP_BITS_RW_MASK;
to->flags = cpu_to_be64(v64);
}
-enum {
- HCA_CAP_OPMOD_GET_MAX = 0,
- HCA_CAP_OPMOD_GET_CUR = 1,
-};
+static u16 get_pkey_table_size(int pkey)
+{
+ if (pkey > MLX5_MAX_LOG_PKEY_TABLE)
+ return 0;
-static int handle_hca_cap(struct mlx5_core_dev *dev)
+ return MLX5_MIN_PKEY_TABLE_SIZE << pkey;
+}
+
+static void fw2drv_caps(struct mlx5_caps *caps,
+ struct mlx5_cmd_query_hca_cap_mbox_out *out)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *query_out = NULL;
- struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
- struct mlx5_cmd_query_hca_cap_mbox_in query_ctx;
- struct mlx5_cmd_set_hca_cap_mbox_out set_out;
- u64 flags;
+ struct mlx5_general_caps *gen = &caps->gen;
+ u16 t16;
+
+ gen->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
+ gen->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
+ gen->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
+ gen->log_max_strq = out->hca_cap.log_max_strq_sz;
+ gen->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
+ gen->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
+ gen->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
+ gen->max_eqes = out->hca_cap.log_max_eq_sz;
+ gen->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
+ gen->log_max_eq = out->hca_cap.log_max_eq & 0xf;
+ gen->max_indirection = out->hca_cap.max_indirection;
+ gen->log_max_mrw_sz = out->hca_cap.log_max_mrw_sz;
+ gen->log_max_bsf_list_size = 0;
+ gen->log_max_klm_list_size = 0;
+ gen->log_max_ra_req_dc = out->hca_cap.log_max_ra_req_dc;
+ gen->log_max_ra_res_dc = out->hca_cap.log_max_ra_res_dc;
+ gen->log_max_ra_req_qp = out->hca_cap.log_max_ra_req_qp;
+ gen->log_max_ra_res_qp = out->hca_cap.log_max_ra_res_qp;
+ gen->max_qp_counters = be16_to_cpu(out->hca_cap.max_qp_count);
+ gen->pkey_table_size = get_pkey_table_size(be16_to_cpu(out->hca_cap.pkey_table_size));
+ gen->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
+ gen->num_ports = out->hca_cap.num_ports & 0xf;
+ gen->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
+ gen->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
+ gen->flags = be64_to_cpu(out->hca_cap.flags);
+ pr_debug("flags = 0x%llx\n", gen->flags);
+ gen->uar_sz = out->hca_cap.uar_sz;
+ gen->min_log_pg_sz = out->hca_cap.log_pg_sz;
+
+ t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
+ if (t16 & 0x8000) {
+ gen->bf_reg_size = 1 << (t16 & 0x1f);
+ gen->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
+ } else {
+ gen->bf_reg_size = 0;
+ gen->bf_regs_per_page = 0;
+ }
+ gen->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
+ gen->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
+ gen->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
+ gen->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
+ gen->log_max_xrcd = out->hca_cap.log_max_xrcd;
+ gen->log_uar_page_sz = be16_to_cpu(out->hca_cap.log_uar_page_sz);
+}
+
+static const char *caps_opmod_str(u16 opmod)
+{
+ switch (opmod) {
+ case HCA_CAP_OPMOD_GET_MAX:
+ return "GET_MAX";
+ case HCA_CAP_OPMOD_GET_CUR:
+ return "GET_CUR";
+ default:
+ return "Invalid";
+ }
+}
+
+int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
+ u16 opmod)
+{
+ struct mlx5_cmd_query_hca_cap_mbox_out *out;
+ struct mlx5_cmd_query_hca_cap_mbox_in in;
int err;
- memset(&query_ctx, 0, sizeof(query_ctx));
- query_out = kzalloc(sizeof(*query_out), GFP_KERNEL);
- if (!query_out)
+ memset(&in, 0, sizeof(in));
+ out = kzalloc(sizeof(*out), GFP_KERNEL);
+ if (!out)
return -ENOMEM;
- set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
- if (!set_ctx) {
- err = -ENOMEM;
+ in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
+ in.hdr.opmod = cpu_to_be16(opmod);
+ err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
+
+ err = mlx5_cmd_status_to_err(&out->hdr);
+ if (err) {
+ mlx5_core_warn(dev, "query max hca cap failed, %d\n", err);
goto query_ex;
}
+ mlx5_core_dbg(dev, "%s\n", caps_opmod_str(opmod));
+ fw2drv_caps(caps, out);
- query_ctx.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- query_ctx.hdr.opmod = cpu_to_be16(HCA_CAP_OPMOD_GET_CUR);
- err = mlx5_cmd_exec(dev, &query_ctx, sizeof(query_ctx),
- query_out, sizeof(*query_out));
+query_ex:
+ kfree(out);
+ return err;
+}
+
+static int set_caps(struct mlx5_core_dev *dev,
+ struct mlx5_cmd_set_hca_cap_mbox_in *in)
+{
+ struct mlx5_cmd_set_hca_cap_mbox_out out;
+ int err;
+
+ memset(&out, 0, sizeof(out));
+
+ in->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
+ err = mlx5_cmd_exec(dev, in, sizeof(*in), &out, sizeof(out));
if (err)
- goto query_ex;
+ return err;
- err = mlx5_cmd_status_to_err(&query_out->hdr);
- if (err) {
- mlx5_core_warn(dev, "query hca cap failed, %d\n", err);
+ err = mlx5_cmd_status_to_err(&out.hdr);
+
+ return err;
+}
+
+static int handle_hca_cap(struct mlx5_core_dev *dev)
+{
+ struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
+ struct mlx5_profile *prof = dev->profile;
+ struct mlx5_caps *cur_caps = NULL;
+ struct mlx5_caps *max_caps = NULL;
+ int err = -ENOMEM;
+
+ set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
+ if (!set_ctx)
goto query_ex;
- }
- copy_rw_fields(&set_ctx->hca_cap, &query_out->hca_cap);
+ max_caps = kzalloc(sizeof(*max_caps), GFP_KERNEL);
+ if (!max_caps)
+ goto query_ex;
- if (dev->profile && dev->profile->mask & MLX5_PROF_MASK_QP_SIZE)
- set_ctx->hca_cap.log_max_qp = dev->profile->log_max_qp;
+ cur_caps = kzalloc(sizeof(*cur_caps), GFP_KERNEL);
+ if (!cur_caps)
+ goto query_ex;
- flags = be64_to_cpu(query_out->hca_cap.flags);
- /* disable checksum */
- flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
-
- set_ctx->hca_cap.flags = cpu_to_be64(flags);
- memset(&set_out, 0, sizeof(set_out));
- set_ctx->hca_cap.log_uar_page_sz = cpu_to_be16(PAGE_SHIFT - 12);
- set_ctx->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
- err = mlx5_cmd_exec(dev, set_ctx, sizeof(*set_ctx),
- &set_out, sizeof(set_out));
- if (err) {
- mlx5_core_warn(dev, "set hca cap failed, %d\n", err);
+ err = mlx5_core_get_caps(dev, max_caps, HCA_CAP_OPMOD_GET_MAX);
+ if (err)
goto query_ex;
- }
- err = mlx5_cmd_status_to_err(&set_out.hdr);
+ err = mlx5_core_get_caps(dev, cur_caps, HCA_CAP_OPMOD_GET_CUR);
if (err)
goto query_ex;
+ /* we limit the size of the pkey table to 128 entries for now */
+ cur_caps->gen.pkey_table_size = 128;
+
+ if (prof->mask & MLX5_PROF_MASK_QP_SIZE)
+ cur_caps->gen.log_max_qp = prof->log_max_qp;
+
+ /* disable checksum */
+ cur_caps->gen.flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
+
+ copy_rw_fields(&set_ctx->hca_cap, &cur_caps->gen);
+ err = set_caps(dev, set_ctx);
+
query_ex:
- kfree(query_out);
+ kfree(cur_caps);
+ kfree(max_caps);
kfree(set_ctx);
return err;
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/uar.c b/drivers/net/ethernet/mellanox/mlx5/core/uar.c
index 68f5d9c77c7b..0a6348cefc01 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/uar.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/uar.c
@@ -174,11 +174,11 @@ int mlx5_alloc_uuars(struct mlx5_core_dev *dev, struct mlx5_uuar_info *uuari)
for (i = 0; i < tot_uuars; i++) {
bf = &uuari->bfs[i];
- bf->buf_size = dev->caps.bf_reg_size / 2;
+ bf->buf_size = dev->caps.gen.bf_reg_size / 2;
bf->uar = &uuari->uars[i / MLX5_BF_REGS_PER_PAGE];
bf->regreg = uuari->uars[i / MLX5_BF_REGS_PER_PAGE].map;
bf->reg = NULL; /* Add WC support */
- bf->offset = (i % MLX5_BF_REGS_PER_PAGE) * dev->caps.bf_reg_size +
+ bf->offset = (i % MLX5_BF_REGS_PER_PAGE) * dev->caps.gen.bf_reg_size +
MLX5_BF_OFFSET;
bf->need_lock = need_uuar_lock(i);
spin_lock_init(&bf->lock);
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 334947151dfc..dce01fd854a8 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -71,6 +71,11 @@ enum {
};
enum {
+ MLX5_MIN_PKEY_TABLE_SIZE = 128,
+ MLX5_MAX_LOG_PKEY_TABLE = 5,
+};
+
+enum {
MLX5_PERM_LOCAL_READ = 1 << 2,
MLX5_PERM_LOCAL_WRITE = 1 << 3,
MLX5_PERM_REMOTE_READ = 1 << 4,
@@ -184,10 +189,10 @@ enum {
MLX5_DEV_CAP_FLAG_CQ_MODER = 1LL << 29,
MLX5_DEV_CAP_FLAG_RESIZE_CQ = 1LL << 30,
MLX5_DEV_CAP_FLAG_RESIZE_SRQ = 1LL << 32,
+ MLX5_DEV_CAP_FLAG_DCT = 1LL << 37,
MLX5_DEV_CAP_FLAG_REMOTE_FENCE = 1LL << 38,
MLX5_DEV_CAP_FLAG_TLP_HINTS = 1LL << 39,
MLX5_DEV_CAP_FLAG_SIG_HAND_OVER = 1LL << 40,
- MLX5_DEV_CAP_FLAG_DCT = 1LL << 41,
MLX5_DEV_CAP_FLAG_CMDIF_CSUM = 3LL << 46,
};
@@ -243,10 +248,14 @@ enum {
};
enum {
- MLX5_CAP_OFF_DCT = 41,
MLX5_CAP_OFF_CMDIF_CSUM = 46,
};
+enum {
+ HCA_CAP_OPMOD_GET_MAX = 0,
+ HCA_CAP_OPMOD_GET_CUR = 1,
+};
+
struct mlx5_inbox_hdr {
__be16 opcode;
u8 rsvd[4];
@@ -303,9 +312,10 @@ struct mlx5_hca_cap {
u8 log_max_ra_req_qp;
u8 rsvd10;
u8 log_max_ra_res_qp;
- u8 rsvd11[4];
+ u8 pad_cap;
+ u8 rsvd11[3];
__be16 max_qp_count;
- __be16 rsvd12;
+ __be16 pkey_table_size;
u8 rsvd13;
u8 local_ca_ack_delay;
u8 rsvd14;
@@ -335,11 +345,7 @@ struct mlx5_hca_cap {
u8 log_max_xrcd;
u8 rsvd25[42];
__be16 log_uar_page_sz;
- u8 rsvd26[28];
- u8 log_max_atomic_size_qp;
- u8 rsvd27[2];
- u8 log_max_atomic_size_dc;
- u8 rsvd28[76];
+ u8 rsvd26[108];
};
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index b88e9b46d957..45a2add747e0 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -335,23 +335,30 @@ struct mlx5_port_caps {
int pkey_table_len;
};
-struct mlx5_caps {
+struct mlx5_general_caps {
u8 log_max_eq;
u8 log_max_cq;
u8 log_max_qp;
u8 log_max_mkey;
u8 log_max_pd;
u8 log_max_srq;
+ u8 log_max_strq;
+ u8 log_max_mrw_sz;
+ u8 log_max_bsf_list_size;
+ u8 log_max_klm_list_size;
u32 max_cqes;
int max_wqes;
+ u32 max_eqes;
+ u32 max_indirection;
int max_sq_desc_sz;
int max_rq_desc_sz;
+ int max_dc_sq_desc_sz;
u64 flags;
u16 stat_rate_support;
int log_max_msg;
int num_ports;
- int max_ra_res_qp;
- int max_ra_req_qp;
+ u8 log_max_ra_res_qp;
+ u8 log_max_ra_req_qp;
int max_srq_wqes;
int bf_reg_size;
int bf_regs_per_page;
@@ -363,6 +370,19 @@ struct mlx5_caps {
u8 log_max_mcg;
u32 max_qp_mcg;
int min_page_sz;
+ int pd_cap;
+ u32 max_qp_counters;
+ u32 pkey_table_size;
+ u8 log_max_ra_req_dc;
+ u8 log_max_ra_res_dc;
+ u32 uar_sz;
+ u8 min_log_pg_sz;
+ u8 log_max_xrcd;
+ u16 log_uar_page_sz;
+};
+
+struct mlx5_caps {
+ struct mlx5_general_caps gen;
};
struct mlx5_cmd_mailbox {
@@ -695,6 +715,8 @@ void mlx5_cmd_cleanup(struct mlx5_core_dev *dev);
void mlx5_cmd_use_events(struct mlx5_core_dev *dev);
void mlx5_cmd_use_polling(struct mlx5_core_dev *dev);
int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr);
+int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
+ u16 opmod);
int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
int out_size);
int mlx5_cmd_exec_cb(struct mlx5_core_dev *dev, void *in, int in_size,
--
2.1.1
^ permalink raw reply related
* [PATCH net-next 2/4] net/mlx5_core: Use hardware registers description header file
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412082135-30229-1-git-send-email-eli@mellanox.com>
Add an auto generated header file that describes hardware registers along with
set of macros that set/get values. The macros do static checks to avoid
overflow, handle endianess, and overall provide a clean way to code commands.
A few commands were removed since they are not supported currently and will be
added when support is available.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 36 -
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 3 -
include/linux/mlx5/device.h | 74 +
include/linux/mlx5/driver.h | 76 +-
include/linux/mlx5/mlx5_ifc.h | 4464 +++++++++++++++++++++++++
5 files changed, 4539 insertions(+), 114 deletions(-)
create mode 100644 include/linux/mlx5/mlx5_ifc.h
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 6eb0f85cf872..3ecef1310bae 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -357,60 +357,24 @@ const char *mlx5_command_str(int command)
case MLX5_CMD_OP_2ERR_QP:
return "2ERR_QP";
- case MLX5_CMD_OP_RTS2SQD_QP:
- return "RTS2SQD_QP";
-
- case MLX5_CMD_OP_SQD2RTS_QP:
- return "SQD2RTS_QP";
-
case MLX5_CMD_OP_2RST_QP:
return "2RST_QP";
case MLX5_CMD_OP_QUERY_QP:
return "QUERY_QP";
- case MLX5_CMD_OP_CONF_SQP:
- return "CONF_SQP";
-
case MLX5_CMD_OP_MAD_IFC:
return "MAD_IFC";
case MLX5_CMD_OP_INIT2INIT_QP:
return "INIT2INIT_QP";
- case MLX5_CMD_OP_SUSPEND_QP:
- return "SUSPEND_QP";
-
- case MLX5_CMD_OP_UNSUSPEND_QP:
- return "UNSUSPEND_QP";
-
- case MLX5_CMD_OP_SQD2SQD_QP:
- return "SQD2SQD_QP";
-
- case MLX5_CMD_OP_ALLOC_QP_COUNTER_SET:
- return "ALLOC_QP_COUNTER_SET";
-
- case MLX5_CMD_OP_DEALLOC_QP_COUNTER_SET:
- return "DEALLOC_QP_COUNTER_SET";
-
- case MLX5_CMD_OP_QUERY_QP_COUNTER_SET:
- return "QUERY_QP_COUNTER_SET";
-
case MLX5_CMD_OP_CREATE_PSV:
return "CREATE_PSV";
case MLX5_CMD_OP_DESTROY_PSV:
return "DESTROY_PSV";
- case MLX5_CMD_OP_QUERY_PSV:
- return "QUERY_PSV";
-
- case MLX5_CMD_OP_QUERY_SIG_RULE_TABLE:
- return "QUERY_SIG_RULE_TABLE";
-
- case MLX5_CMD_OP_QUERY_BLOCK_SIZE_TABLE:
- return "QUERY_BLOCK_SIZE_TABLE";
-
case MLX5_CMD_OP_CREATE_SRQ:
return "CREATE_SRQ";
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/qp.c b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
index 8145b4668229..415b67ce379e 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
@@ -184,13 +184,10 @@ int mlx5_core_qp_modify(struct mlx5_core_dev *dev, enum mlx5_qp_state cur_state,
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
[MLX5_QP_STATE_ERR] = MLX5_CMD_OP_2ERR_QP,
[MLX5_QP_STATE_RTS] = MLX5_CMD_OP_RTS2RTS_QP,
- [MLX5_QP_STATE_SQD] = MLX5_CMD_OP_RTS2SQD_QP,
},
[MLX5_QP_STATE_SQD] = {
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
[MLX5_QP_STATE_ERR] = MLX5_CMD_OP_2ERR_QP,
- [MLX5_QP_STATE_RTS] = MLX5_CMD_OP_SQD2RTS_QP,
- [MLX5_QP_STATE_SQD] = MLX5_CMD_OP_SQD2SQD_QP,
},
[MLX5_QP_STATE_SQER] = {
[MLX5_QP_STATE_RST] = MLX5_CMD_OP_2RST_QP,
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index dce01fd854a8..3db2d4d82ad3 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -44,6 +44,80 @@
#error Host endianness not defined
#endif
+/* helper macros */
+#define __mlx5_nullp(typ) ((struct mlx5_ifc_##typ##_bits *)0)
+#define __mlx5_bit_sz(typ, fld) sizeof(__mlx5_nullp(typ)->fld)
+#define __mlx5_bit_off(typ, fld) ((unsigned)(unsigned long)(&(__mlx5_nullp(typ)->fld)))
+#define __mlx5_dw_off(typ, fld) (__mlx5_bit_off(typ, fld) / 32)
+#define __mlx5_64_off(typ, fld) (__mlx5_bit_off(typ, fld) / 64)
+#define __mlx5_dw_bit_off(typ, fld) (32 - __mlx5_bit_sz(typ, fld) - (__mlx5_bit_off(typ, fld) & 0x1f))
+#define __mlx5_mask(typ, fld) ((u32)((1ull << __mlx5_bit_sz(typ, fld)) - 1))
+#define __mlx5_dw_mask(typ, fld) (__mlx5_mask(typ, fld) << __mlx5_dw_bit_off(typ, fld))
+#define __mlx5_st_sz_bits(typ) sizeof(struct mlx5_ifc_##typ##_bits)
+
+#define MLX5_FLD_SZ_BYTES(typ, fld) (__mlx5_bit_sz(typ, fld) / 8)
+#define MLX5_ST_SZ_BYTES(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 8)
+#define MLX5_ST_SZ_DW(typ) (sizeof(struct mlx5_ifc_##typ##_bits) / 32)
+#define MLX5_BYTE_OFF(typ, fld) (__mlx5_bit_off(typ, fld) / 8)
+#define MLX5_ADDR_OF(typ, p, fld) ((char *)(p) + MLX5_BYTE_OFF(typ, fld))
+
+static inline void non_existent_function(void)
+{
+ pr_info("%s\n", __func__);
+}
+
+static inline void non_existent_function_sz_align32(void)
+{
+ pr_info("%s\n", __func__);
+}
+
+static inline void non_existent_function_const_overflow(void)
+{
+ pr_info("%s\n", __func__);
+}
+
+static inline void memcpy_cpu_to_be32(void *dst, void *src, int len)
+{
+ u32 *dst_u32 = (u32 *)dst;
+ int i;
+
+ memcpy(dst, src, len);
+
+ for (i = 0; i < (len >> 2); i++)
+ dst_u32[i] = cpu_to_be32(dst_u32[i]);
+}
+
+/* insert a value to a struct */
+#define MLX5_SET(typ, p, fld, v) do { \
+ if (__mlx5_st_sz_bits(typ) % 32) \
+ non_existent_function_sz_align32(); \
+ *((__be32 *)(p) + __mlx5_dw_off(typ, fld)) = \
+ cpu_to_be32((be32_to_cpu(*((__be32 *)(p) + __mlx5_dw_off(typ, fld))) & \
+ (~__mlx5_dw_mask(typ, fld))) | (((v) & __mlx5_mask(typ, fld)) \
+ << __mlx5_dw_bit_off(typ, fld))); \
+} while (0)
+
+#define MLX5_GET(typ, p, fld) ((be32_to_cpu(*((__be32 *)(p) +\
+__mlx5_dw_off(typ, fld))) >> __mlx5_dw_bit_off(typ, fld)) & \
+__mlx5_mask(typ, fld))
+
+#define MLX5_GET_PR(typ, p, fld) ({ \
+ u32 ___t = MLX5_GET(typ, p, fld); \
+ pr_debug(#fld " = 0x%x\n", ___t); \
+ ___t; \
+})
+
+#define MLX5_SET64(typ, p, fld, v) do { \
+ if (__mlx5_bit_sz(typ, fld) != 64) \
+ non_existent_function(); \
+ else if (__mlx5_bit_off(typ, fld) % 64) \
+ non_existent_function(); \
+ else \
+ *((__be64 *)(p) + __mlx5_64_off(typ, fld)) = cpu_to_be64(v); \
+} while (0)
+
+#define MLX5_GET64(typ, p, fld) be64_to_cpu(*((__be64 *)(p) + __mlx5_64_off(typ, fld)))
+
enum {
MLX5_MAX_COMMANDS = 32,
MLX5_CMD_DATA_BLOCK_SIZE = 512,
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 45a2add747e0..6f48dc793b9f 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -44,6 +44,7 @@
#include <linux/mlx5/device.h>
#include <linux/mlx5/doorbell.h>
+#include <linux/mlx5/mlx5_ifc.h>
enum {
MLX5_BOARD_ID_LEN = 64,
@@ -99,81 +100,6 @@ enum {
};
enum {
- MLX5_CMD_OP_QUERY_HCA_CAP = 0x100,
- MLX5_CMD_OP_QUERY_ADAPTER = 0x101,
- MLX5_CMD_OP_INIT_HCA = 0x102,
- MLX5_CMD_OP_TEARDOWN_HCA = 0x103,
- MLX5_CMD_OP_ENABLE_HCA = 0x104,
- MLX5_CMD_OP_DISABLE_HCA = 0x105,
- MLX5_CMD_OP_QUERY_PAGES = 0x107,
- MLX5_CMD_OP_MANAGE_PAGES = 0x108,
- MLX5_CMD_OP_SET_HCA_CAP = 0x109,
-
- MLX5_CMD_OP_CREATE_MKEY = 0x200,
- MLX5_CMD_OP_QUERY_MKEY = 0x201,
- MLX5_CMD_OP_DESTROY_MKEY = 0x202,
- MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS = 0x203,
-
- MLX5_CMD_OP_CREATE_EQ = 0x301,
- MLX5_CMD_OP_DESTROY_EQ = 0x302,
- MLX5_CMD_OP_QUERY_EQ = 0x303,
-
- MLX5_CMD_OP_CREATE_CQ = 0x400,
- MLX5_CMD_OP_DESTROY_CQ = 0x401,
- MLX5_CMD_OP_QUERY_CQ = 0x402,
- MLX5_CMD_OP_MODIFY_CQ = 0x403,
-
- MLX5_CMD_OP_CREATE_QP = 0x500,
- MLX5_CMD_OP_DESTROY_QP = 0x501,
- MLX5_CMD_OP_RST2INIT_QP = 0x502,
- MLX5_CMD_OP_INIT2RTR_QP = 0x503,
- MLX5_CMD_OP_RTR2RTS_QP = 0x504,
- MLX5_CMD_OP_RTS2RTS_QP = 0x505,
- MLX5_CMD_OP_SQERR2RTS_QP = 0x506,
- MLX5_CMD_OP_2ERR_QP = 0x507,
- MLX5_CMD_OP_RTS2SQD_QP = 0x508,
- MLX5_CMD_OP_SQD2RTS_QP = 0x509,
- MLX5_CMD_OP_2RST_QP = 0x50a,
- MLX5_CMD_OP_QUERY_QP = 0x50b,
- MLX5_CMD_OP_CONF_SQP = 0x50c,
- MLX5_CMD_OP_MAD_IFC = 0x50d,
- MLX5_CMD_OP_INIT2INIT_QP = 0x50e,
- MLX5_CMD_OP_SUSPEND_QP = 0x50f,
- MLX5_CMD_OP_UNSUSPEND_QP = 0x510,
- MLX5_CMD_OP_SQD2SQD_QP = 0x511,
- MLX5_CMD_OP_ALLOC_QP_COUNTER_SET = 0x512,
- MLX5_CMD_OP_DEALLOC_QP_COUNTER_SET = 0x513,
- MLX5_CMD_OP_QUERY_QP_COUNTER_SET = 0x514,
-
- MLX5_CMD_OP_CREATE_PSV = 0x600,
- MLX5_CMD_OP_DESTROY_PSV = 0x601,
- MLX5_CMD_OP_QUERY_PSV = 0x602,
- MLX5_CMD_OP_QUERY_SIG_RULE_TABLE = 0x603,
- MLX5_CMD_OP_QUERY_BLOCK_SIZE_TABLE = 0x604,
-
- MLX5_CMD_OP_CREATE_SRQ = 0x700,
- MLX5_CMD_OP_DESTROY_SRQ = 0x701,
- MLX5_CMD_OP_QUERY_SRQ = 0x702,
- MLX5_CMD_OP_ARM_RQ = 0x703,
- MLX5_CMD_OP_RESIZE_SRQ = 0x704,
-
- MLX5_CMD_OP_ALLOC_PD = 0x800,
- MLX5_CMD_OP_DEALLOC_PD = 0x801,
- MLX5_CMD_OP_ALLOC_UAR = 0x802,
- MLX5_CMD_OP_DEALLOC_UAR = 0x803,
-
- MLX5_CMD_OP_ATTACH_TO_MCG = 0x806,
- MLX5_CMD_OP_DETACH_FROM_MCG = 0x807,
-
-
- MLX5_CMD_OP_ALLOC_XRCD = 0x80e,
- MLX5_CMD_OP_DEALLOC_XRCD = 0x80f,
-
- MLX5_CMD_OP_ACCESS_REG = 0x805,
- MLX5_CMD_OP_MAX = 0x810,
-};
-
-enum {
MLX5_REG_PCAP = 0x5001,
MLX5_REG_PMTU = 0x5003,
MLX5_REG_PTYS = 0x5004,
diff --git a/include/linux/mlx5/mlx5_ifc.h b/include/linux/mlx5/mlx5_ifc.h
new file mode 100644
index 000000000000..763ab31f54f6
--- /dev/null
+++ b/include/linux/mlx5/mlx5_ifc.h
@@ -0,0 +1,4464 @@
+/*
+ * Copyright (c) 2014, Mellanox Technologies inc. 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.
+ */
+
+/*
+
+ Autogenerated file.
+ Date: 2014-09-30 06:30
+ Source Document Name: Mellanox <Doc Name>
+ Source Document Version: 0.20
+ Generated by adb_to_c.py (EAT.ME Version: 1.0.62)
+*/
+#ifndef MLX5_IFC_H
+#define MLX5_IFC_H
+
+enum {
+ MLX5_EVENT_TYPE_CODING_COMPLETION_EVENTS = 0x0,
+ MLX5_EVENT_TYPE_CODING_PATH_MIGRATED_SUCCEEDED = 0x1,
+ MLX5_EVENT_TYPE_CODING_COMMUNICATION_ESTABLISHED = 0x2,
+ MLX5_EVENT_TYPE_CODING_SEND_QUEUE_DRAINED = 0x3,
+ MLX5_EVENT_TYPE_CODING_SRQ_LAST_WQE_REACHED = 0x13,
+ MLX5_EVENT_TYPE_CODING_SRQ_LIMIT = 0x14,
+ MLX5_EVENT_TYPE_CODING_DCT_ALL_CONNECTIONS_CLOSED = 0x1c,
+ MLX5_EVENT_TYPE_CODING_DCT_ACCESS_KEY_VIOLATION = 0x1d,
+ MLX5_EVENT_TYPE_CODING_CQ_ERROR = 0x4,
+ MLX5_EVENT_TYPE_CODING_LOCAL_WQ_CATASTROPHIC_ERROR = 0x5,
+ MLX5_EVENT_TYPE_CODING_PATH_MIGRATION_FAILED = 0x7,
+ MLX5_EVENT_TYPE_CODING_PAGE_FAULT_EVENT = 0xc,
+ MLX5_EVENT_TYPE_CODING_INVALID_REQUEST_LOCAL_WQ_ERROR = 0x10,
+ MLX5_EVENT_TYPE_CODING_LOCAL_ACCESS_VIOLATION_WQ_ERROR = 0x11,
+ MLX5_EVENT_TYPE_CODING_LOCAL_SRQ_CATASTROPHIC_ERROR = 0x12,
+ MLX5_EVENT_TYPE_CODING_INTERNAL_ERROR = 0x8,
+ MLX5_EVENT_TYPE_CODING_PORT_STATE_CHANGE = 0x9,
+ MLX5_EVENT_TYPE_CODING_GPIO_EVENT = 0x15,
+ MLX5_EVENT_TYPE_CODING_REMOTE_CONFIGURATION_PROTOCOL_EVENT = 0x19,
+ MLX5_EVENT_TYPE_CODING_DOORBELL_BLUEFLAME_CONGESTION_EVENT = 0x1a,
+ MLX5_EVENT_TYPE_CODING_STALL_VL_EVENT = 0x1b,
+ MLX5_EVENT_TYPE_CODING_DROPPED_PACKET_LOGGED_EVENT = 0x1f,
+ MLX5_EVENT_TYPE_CODING_COMMAND_INTERFACE_COMPLETION = 0xa,
+ MLX5_EVENT_TYPE_CODING_PAGE_REQUEST = 0xb
+};
+
+enum {
+ MLX5_CMD_OP_QUERY_HCA_CAP = 0x100,
+ MLX5_CMD_OP_QUERY_ADAPTER = 0x101,
+ MLX5_CMD_OP_INIT_HCA = 0x102,
+ MLX5_CMD_OP_TEARDOWN_HCA = 0x103,
+ MLX5_CMD_OP_ENABLE_HCA = 0x104,
+ MLX5_CMD_OP_DISABLE_HCA = 0x105,
+ MLX5_CMD_OP_QUERY_PAGES = 0x107,
+ MLX5_CMD_OP_MANAGE_PAGES = 0x108,
+ MLX5_CMD_OP_SET_HCA_CAP = 0x109,
+ MLX5_CMD_OP_CREATE_MKEY = 0x200,
+ MLX5_CMD_OP_QUERY_MKEY = 0x201,
+ MLX5_CMD_OP_DESTROY_MKEY = 0x202,
+ MLX5_CMD_OP_QUERY_SPECIAL_CONTEXTS = 0x203,
+ MLX5_CMD_OP_PAGE_FAULT_RESUME = 0x204,
+ MLX5_CMD_OP_CREATE_EQ = 0x301,
+ MLX5_CMD_OP_DESTROY_EQ = 0x302,
+ MLX5_CMD_OP_QUERY_EQ = 0x303,
+ MLX5_CMD_OP_GEN_EQE = 0x304,
+ MLX5_CMD_OP_CREATE_CQ = 0x400,
+ MLX5_CMD_OP_DESTROY_CQ = 0x401,
+ MLX5_CMD_OP_QUERY_CQ = 0x402,
+ MLX5_CMD_OP_MODIFY_CQ = 0x403,
+ MLX5_CMD_OP_CREATE_QP = 0x500,
+ MLX5_CMD_OP_DESTROY_QP = 0x501,
+ MLX5_CMD_OP_RST2INIT_QP = 0x502,
+ MLX5_CMD_OP_INIT2RTR_QP = 0x503,
+ MLX5_CMD_OP_RTR2RTS_QP = 0x504,
+ MLX5_CMD_OP_RTS2RTS_QP = 0x505,
+ MLX5_CMD_OP_SQERR2RTS_QP = 0x506,
+ MLX5_CMD_OP_2ERR_QP = 0x507,
+ MLX5_CMD_OP_2RST_QP = 0x50a,
+ MLX5_CMD_OP_QUERY_QP = 0x50b,
+ MLX5_CMD_OP_INIT2INIT_QP = 0x50e,
+ MLX5_CMD_OP_CREATE_PSV = 0x600,
+ MLX5_CMD_OP_DESTROY_PSV = 0x601,
+ MLX5_CMD_OP_CREATE_SRQ = 0x700,
+ MLX5_CMD_OP_DESTROY_SRQ = 0x701,
+ MLX5_CMD_OP_QUERY_SRQ = 0x702,
+ MLX5_CMD_OP_ARM_RQ = 0x703,
+ MLX5_CMD_OP_RESIZE_SRQ = 0x704,
+ MLX5_CMD_OP_CREATE_DCT = 0x710,
+ MLX5_CMD_OP_DESTROY_DCT = 0x711,
+ MLX5_CMD_OP_DRAIN_DCT = 0x712,
+ MLX5_CMD_OP_QUERY_DCT = 0x713,
+ MLX5_CMD_OP_ARM_DCT_FOR_KEY_VIOLATION = 0x714,
+ MLX5_CMD_OP_QUERY_VPORT_STATE = 0x750,
+ MLX5_CMD_OP_MODIFY_VPORT_STATE = 0x751,
+ MLX5_CMD_OP_QUERY_ESW_VPORT_CONTEXT = 0x752,
+ MLX5_CMD_OP_MODIFY_ESW_VPORT_CONTEXT = 0x753,
+ MLX5_CMD_OP_QUERY_NIC_VPORT_CONTEXT = 0x754,
+ MLX5_CMD_OP_MODIFY_NIC_VPORT_CONTEXT = 0x755,
+ MLX5_CMD_OP_QUERY_RCOE_ADDRESS = 0x760,
+ MLX5_CMD_OP_SET_ROCE_ADDRESS = 0x761,
+ MLX5_CMD_OP_QUERY_VPORT_COUNTER = 0x770,
+ MLX5_CMD_OP_ALLOC_Q_COUNTER = 0x771,
+ MLX5_CMD_OP_DEALLOC_Q_COUNTER = 0x772,
+ MLX5_CMD_OP_QUERY_Q_COUNTER = 0x773,
+ MLX5_CMD_OP_ALLOC_PD = 0x800,
+ MLX5_CMD_OP_DEALLOC_PD = 0x801,
+ MLX5_CMD_OP_ALLOC_UAR = 0x802,
+ MLX5_CMD_OP_DEALLOC_UAR = 0x803,
+ MLX5_CMD_OP_CONFIG_INT_MODERATION = 0x804,
+ MLX5_CMD_OP_ACCESS_REG = 0x805,
+ MLX5_CMD_OP_ATTACH_TO_MCG = 0x806,
+ MLX5_CMD_OP_DETACH_FROM_MCG = 0x807,
+ MLX5_CMD_OP_GET_DROPPED_PACKET_LOG = 0x80a,
+ MLX5_CMD_OP_MAD_IFC = 0x50d,
+ MLX5_CMD_OP_QUERY_MAD_DEMUX = 0x80b,
+ MLX5_CMD_OP_SET_MAD_DEMUX = 0x80c,
+ MLX5_CMD_OP_NOP = 0x80d,
+ MLX5_CMD_OP_ALLOC_XRCD = 0x80e,
+ MLX5_CMD_OP_DEALLOC_XRCD = 0x80f,
+ MLX5_CMD_OP_SET_BURST_SIZE = 0x812,
+ MLX5_CMD_OP_QUERY_BURST_SZIE = 0x813,
+ MLX5_CMD_OP_ACTIVATE_TRACER = 0x814,
+ MLX5_CMD_OP_DEACTIVATE_TRACER = 0x815,
+ MLX5_CMD_OP_CREATE_SNIFFER_RULE = 0x820,
+ MLX5_CMD_OP_DESTROY_SNIFFER_RULE = 0x821,
+ MLX5_CMD_OP_QUERY_CONG_PARAMS = 0x822,
+ MLX5_CMD_OP_MODIFY_CONG_PARAMS = 0x823,
+ MLX5_CMD_OP_QUERY_CONG_STATISTICS = 0x824,
+ MLX5_CMD_OP_CREATE_TIR = 0x900,
+ MLX5_CMD_OP_MODIFY_TIR = 0x901,
+ MLX5_CMD_OP_DESTROY_TIR = 0x902,
+ MLX5_CMD_OP_QUERY_TIR = 0x903,
+ MLX5_CMD_OP_CREATE_TIS = 0x912,
+ MLX5_CMD_OP_MODIFY_TIS = 0x913,
+ MLX5_CMD_OP_DESTROY_TIS = 0x914,
+ MLX5_CMD_OP_QUERY_TIS = 0x915,
+ MLX5_CMD_OP_CREATE_SQ = 0x904,
+ MLX5_CMD_OP_MODIFY_SQ = 0x905,
+ MLX5_CMD_OP_DESTROY_SQ = 0x906,
+ MLX5_CMD_OP_QUERY_SQ = 0x907,
+ MLX5_CMD_OP_CREATE_RQ = 0x908,
+ MLX5_CMD_OP_MODIFY_RQ = 0x909,
+ MLX5_CMD_OP_DESTROY_RQ = 0x90a,
+ MLX5_CMD_OP_QUERY_RQ = 0x90b,
+ MLX5_CMD_OP_CREATE_RMP = 0x90c,
+ MLX5_CMD_OP_MODIFY_RMP = 0x90d,
+ MLX5_CMD_OP_DESTROY_RMP = 0x90e,
+ MLX5_CMD_OP_QUERY_RMP = 0x90f,
+ MLX5_CMD_OP_SET_FLOW_TABLE_ENTRY = 0x910,
+ MLX5_CMD_OP_QUERY_FLOW_TABLE_ENTRY = 0x911,
+ MLX5_CMD_OP_MAX = 0x911
+};
+
+struct mlx5_ifc_cmd_pas_bits {
+ u8 pa_h[0x20];
+
+ u8 pa_l[0x17];
+ u8 reserved_0[0x9];
+};
+
+struct mlx5_ifc_uint64_bits {
+ u8 hi[0x20];
+
+ u8 lo[0x20];
+};
+
+struct mlx5_ifc_ads_bits {
+ u8 fl[0x1];
+ u8 free_ar[0x1];
+ u8 reserved_0[0xe];
+ u8 pkey_index[0x10];
+
+ u8 reserved_1[0x8];
+ u8 grh[0x1];
+ u8 mlid[0x7];
+ u8 rlid[0x10];
+
+ u8 ack_timeout[0x5];
+ u8 reserved_2[0x3];
+ u8 src_addr_index[0x8];
+ u8 reserved_3[0x4];
+ u8 stat_rate[0x4];
+ u8 hop_limit[0x8];
+
+ u8 reserved_4[0x4];
+ u8 tclass[0x8];
+ u8 flow_label[0x14];
+
+ u8 rgid_rip[16][0x8];
+
+ u8 reserved_5[0x4];
+ u8 f_dscp[0x1];
+ u8 f_ecn[0x1];
+ u8 reserved_6[0x1];
+ u8 f_eth_prio[0x1];
+ u8 ecn[0x2];
+ u8 dscp[0x6];
+ u8 udp_sport[0x10];
+
+ u8 dei_cfi[0x1];
+ u8 eth_prio[0x3];
+ u8 sl[0x4];
+ u8 port[0x8];
+ u8 rmac[6][0x8];
+};
+
+struct mlx5_ifc_rq_num_bits {
+ u8 reserved_0[0x8];
+ u8 rq_num[0x18];
+};
+
+enum {
+ MLX5_RSS_HASH_SELECTOR_IP_HASH_SIP = 0x1,
+ MLX5_RSS_HASH_SELECTOR_IP_HASH_DIP = 0x2,
+};
+
+enum {
+ MLX5_RSS_HASH_SELECTOR_TCP_IP_HASH_SIP = 0x1,
+ MLX5_RSS_HASH_SELECTOR_TCP_IP_HASH_DIP = 0x2,
+ MLX5_RSS_HASH_SELECTOR_TCP_IP_HASH_L4_SPORT = 0x8,
+ MLX5_RSS_HASH_SELECTOR_TCP_IP_HASH_L4_DPORT = 0x10,
+};
+
+enum {
+ MLX5_RSS_HASH_SELECTOR_UDP_IP_HASH_SIP = 0x1,
+ MLX5_RSS_HASH_SELECTOR_UDP_IP_HASH_DIP = 0x2,
+ MLX5_RSS_HASH_SELECTOR_UDP_IP_HASH_L4_SPORT = 0x8,
+ MLX5_RSS_HASH_SELECTOR_UDP_IP_HASH_L4_DPORT = 0x10,
+};
+
+enum {
+ MLX5_RSS_HASH_SELECTOR_IPSEC_HASH_SIP = 0x1,
+ MLX5_RSS_HASH_SELECTOR_IPSEC_HASH_DIP = 0x2,
+ MLX5_RSS_HASH_SELECTOR_IPSEC_HASH_IPSEC = 0x4,
+};
+
+struct mlx5_ifc_rss_hash_selector_bits {
+ u8 ip[0x10];
+ u8 tcp_ip[0x10];
+
+ u8 udp_ip[0x10];
+ u8 ipsec[0x10];
+};
+
+enum {
+ MLX5_WQ_TYPE_LINKED_LIST = 0x0,
+ MLX5_WQ_TYPE_CYCLIC = 0x1,
+ MLX5_WQ_TYPE_STRQ = 0x2,
+};
+
+enum {
+ MLX5_WQ_END_PAD_MODE_NONE = 0x0,
+ MLX5_WQ_END_PAD_MODE_ALIGN = 0x1,
+};
+
+struct mlx5_ifc_wq_bits {
+ u8 wq_type[0x4];
+ u8 wq_signature[0x1];
+ u8 end_padding_mode[0x2];
+ u8 cd_slave[0x1];
+ u8 reserved_0[0x3];
+ u8 log_wq_pg_sz[0x5];
+ u8 reserved_1[0x4];
+ u8 log_wq_stride[0x4];
+ u8 reserved_2[0x3];
+ u8 log_wq_size[0x5];
+
+ u8 hds_skip_first_sge[0x1];
+ u8 log2_hds_buf_size[0x3];
+ u8 reserved_3[0x7];
+ u8 page_offset[0x5];
+ u8 lwm[0x10];
+
+ u8 reserved_4[0x8];
+ u8 pd[0x18];
+
+ u8 reserved_5[0x8];
+ u8 uar_page[0x18];
+
+ u8 dbr_addr[8][0x8];
+
+ u8 hw_counter[0x20];
+
+ u8 sw_counter[0x20];
+
+ u8 reserved_6[0x80];
+
+ struct mlx5_ifc_cmd_pas_bits pas[0];
+};
+
+struct mlx5_ifc_allowed_address_context_bits {
+ u8 reserved_0[0x4];
+ u8 vlan[0xc];
+ u8 mac_addr[6][0x8];
+
+ u8 reserved_1[0x4];
+ u8 vlan_mask[0xc];
+ u8 mac_addr_mask[6][0x8];
+};
+
+struct mlx5_ifc_qpc_bits {
+ u8 state[0x4];
+ u8 reserved_0[0x4];
+ u8 st[0x8];
+ u8 reserved_1[0x3];
+ u8 pm_state[0x2];
+ u8 reserved_2[0x7];
+ u8 end_padding_mode[0x2];
+ u8 reserved_3[0x2];
+
+ u8 wq_signature[0x1];
+ u8 block_lb_mc[0x1];
+ u8 atomic_like_write_en[0x1];
+ u8 latency_sensitive[0x1];
+ u8 reserved_4[0x4];
+ u8 pd[0x18];
+
+ u8 mtu[0x3];
+ u8 log_msg_max[0x5];
+ u8 reserved_5[0x1];
+ u8 log_rq_size[0x4];
+ u8 log_rq_stride[0x3];
+ u8 no_sq[0x1];
+ u8 log_sq_size[0x4];
+ u8 reserved_6[0x1];
+ u8 log_strq_size[0x5];
+ u8 rlky[0x1];
+ u8 reserved_7[0x4];
+
+ u8 counter_set_id[0x8];
+ u8 uar_page[0x18];
+
+ u8 reserved_8[0x20];
+
+ u8 reserved_9[0x3];
+ u8 log_page_size[0x5];
+ u8 remote_qpn[0x18];
+
+ struct mlx5_ifc_ads_bits primary_address_path;
+
+ struct mlx5_ifc_ads_bits secondary_address_path;
+
+ u8 log_ack_req_freq[0x4];
+ u8 reserved_10[0x4];
+ u8 log_sra_max[0x3];
+ u8 reserved_11[0x2];
+ u8 retry_count[0x3];
+ u8 rnr_retry[0x3];
+ u8 reserved_12[0x1];
+ u8 fre[0x1];
+ u8 cur_rnr_retry[0x3];
+ u8 cur_retry_count[0x3];
+ u8 reserved_13[0x5];
+
+ u8 reserved_14[0x20];
+
+ u8 reserved_15[0x8];
+ u8 next_send_psn[0x18];
+
+ u8 reserved_16[0x8];
+ u8 cqn_snd[0x18];
+
+ u8 reserved_17[0x40];
+
+ u8 reserved_18[0x8];
+ u8 last_acked_psn[0x18];
+
+ u8 reserved_19[0x8];
+ u8 ssn[0x18];
+
+ u8 reserved_20[0x8];
+ u8 log_rra_max[0x3];
+ u8 reserved_21[0x1];
+ u8 atomic_mode[0x4];
+ u8 rre[0x1];
+ u8 rwe[0x1];
+ u8 rae[0x1];
+ u8 reserved_22[0x1];
+ u8 page_offset[0x6];
+ u8 reserved_23[0x3];
+ u8 cd_slave_receive[0x1];
+ u8 cd_slave_send[0x1];
+ u8 cd_master[0x1];
+
+ u8 reserved_24[0x3];
+ u8 min_rnr_nak[0x5];
+ u8 next_rcv_psn[0x18];
+
+ u8 reserved_25[0x8];
+ u8 xrcd[0x18];
+
+ u8 reserved_26[0x8];
+ u8 cqn_rcv[0x18];
+
+ u8 dbr_adr[0x40];
+
+ u8 q_key[0x20];
+
+ u8 reserved_27[0x5];
+ u8 rq_type[0x3];
+ u8 srqn[0x18];
+
+ u8 reserved_28[0x8];
+ u8 rmsn[0x18];
+
+ u8 hw_sq_wqebb_counter[0x10];
+ u8 sw_sq_wqebb_counter[0x10];
+
+ u8 hw_rcyclic_byte_counter[2][0x8];
+ u8 hw_rq_counter_hw_rcyclic_byte_counter[2][0x8];
+
+ u8 sw_rcyclic_byte_counter[2][0x8];
+ u8 sw_rq_counter_sw_rcyclic_byte_counter[2][0x8];
+
+ u8 reserved_29[0x20];
+
+ u8 reserved_30[0xf];
+ u8 cgs[0x1];
+ u8 cs_req[0x8];
+ u8 cs_res[0x8];
+
+ u8 dc_access_key[0x40];
+
+ u8 reserved_31[0xc0];
+};
+
+struct mlx5_ifc_roce_addr_layout_bits {
+ u8 source_l3_address[16][0x8];
+
+ u8 reserved_0[0x3];
+ u8 vlan_valid[0x1];
+ u8 vlan_id[0xc];
+ u8 source_mac[6][0x8];
+
+ u8 reserved_1[0x14];
+ u8 roce_l3_type[0x4];
+ u8 roce_version[0x8];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_cmd_hca_cap_bits {
+ u8 reserved_0[0x80];
+
+ u8 log_max_srq_sz[0x8];
+ u8 log_max_qp_sz[0x8];
+ u8 reserved_1[0xb];
+ u8 log_max_qp[0x5];
+
+ u8 log_max_strq_sz[0x8];
+ u8 reserved_2[0x3];
+ u8 log_max_srqs[0x5];
+ u8 reserved_3[0x10];
+
+ u8 reserved_4[0x8];
+ u8 log_max_cq_sz[0x8];
+ u8 reserved_5[0xb];
+ u8 log_max_cq[0x5];
+
+ u8 log_max_eq_sz[0x8];
+ u8 reserved_6[0x2];
+ u8 log_max_mkey[0x6];
+ u8 reserved_7[0xc];
+ u8 log_max_eq[0x4];
+
+ u8 max_indirection[0x8];
+ u8 reserved_8[0x1];
+ u8 log_max_mrw_sz[0x7];
+ u8 reserved_9[0x2];
+ u8 log_max_bsf_list_size[0x6];
+ u8 reserved_10[0x2];
+ u8 log_max_klm_list_size[0x6];
+
+ u8 reserved_11[0xa];
+ u8 log_max_ra_req_dc[0x6];
+ u8 reserved_12[0xa];
+ u8 log_max_ra_res_dc[0x6];
+
+ u8 reserved_13[0xa];
+ u8 log_max_ra_req_qp[0x6];
+ u8 reserved_14[0xa];
+ u8 log_max_ra_res_qp[0x6];
+
+ u8 pad_cap[0x1];
+ u8 cc_query_allowed[0x1];
+ u8 cc_modify_allowed[0x1];
+ u8 reserved_15[0x1d];
+
+ u8 reserved_16[0x6];
+ u8 max_qp_cnt[0xa];
+ u8 pkey_table_size[0x10];
+
+ u8 eswitch_owner[0x1];
+ u8 reserved_17[0xa];
+ u8 local_ca_ack_delay[0x5];
+ u8 reserved_18[0x8];
+ u8 num_ports[0x8];
+
+ u8 reserved_19[0x3];
+ u8 log_max_msg[0x5];
+ u8 reserved_20[0x18];
+
+ u8 stat_rate_support[0x10];
+ u8 reserved_21[0x10];
+
+ u8 reserved_22[0x10];
+ u8 cmdif_checksum[0x2];
+ u8 sigerr_cqe[0x1];
+ u8 reserved_23[0x1];
+ u8 wq_signature[0x1];
+ u8 sctr_data_cqe[0x1];
+ u8 reserved_24[0x1];
+ u8 sho[0x1];
+ u8 tph[0x1];
+ u8 rf[0x1];
+ u8 dc[0x1];
+ u8 reserved_25[0x2];
+ u8 roce[0x1];
+ u8 atomic[0x1];
+ u8 rsz_srq[0x1];
+
+ u8 cq_oi[0x1];
+ u8 cq_resize[0x1];
+ u8 cq_moderation[0x1];
+ u8 sniffer_rule_flow[0x1];
+ u8 sniffer_rule_vport[0x1];
+ u8 sniffer_rule_phy[0x1];
+ u8 reserved_26[0x1];
+ u8 pg[0x1];
+ u8 block_lb_mc[0x1];
+ u8 reserved_27[0x3];
+ u8 cd[0x1];
+ u8 reserved_28[0x1];
+ u8 apm[0x1];
+ u8 reserved_29[0x7];
+ u8 qkv[0x1];
+ u8 pkv[0x1];
+ u8 reserved_30[0x4];
+ u8 xrc[0x1];
+ u8 ud[0x1];
+ u8 uc[0x1];
+ u8 rc[0x1];
+
+ u8 reserved_31[0xa];
+ u8 uar_sz[0x6];
+ u8 reserved_32[0x8];
+ u8 log_pg_sz[0x8];
+
+ u8 bf[0x1];
+ u8 reserved_33[0xa];
+ u8 log_bf_reg_size[0x5];
+ u8 reserved_34[0x10];
+
+ u8 reserved_35[0x10];
+ u8 max_wqe_sz_sq[0x10];
+
+ u8 reserved_36[0x10];
+ u8 max_wqe_sz_rq[0x10];
+
+ u8 reserved_37[0x10];
+ u8 max_wqe_sz_sq_dc[0x10];
+
+ u8 reserved_38[0x7];
+ u8 max_qp_mcg[0x19];
+
+ u8 reserved_39[0x18];
+ u8 log_max_mcg[0x8];
+
+ u8 reserved_40[0xb];
+ u8 log_max_pd[0x5];
+ u8 reserved_41[0xb];
+ u8 log_max_xrcd[0x5];
+
+ u8 reserved_42[0x20];
+
+ u8 reserved_43[0x3];
+ u8 log_max_rq[0x5];
+ u8 reserved_44[0x3];
+ u8 log_max_sq[0x5];
+ u8 reserved_45[0x3];
+ u8 log_max_tir[0x5];
+ u8 reserved_46[0x3];
+ u8 log_max_tis[0x5];
+
+ u8 reserved_47[0x13];
+ u8 log_max_rq_per_tir[0x5];
+ u8 reserved_48[0x3];
+ u8 log_max_tis_per_sq[0x5];
+
+ u8 reserved_49[0xe0];
+
+ u8 reserved_50[0x10];
+ u8 log_uar_page_sz[0x10];
+
+ u8 reserved_51[0x100];
+
+ u8 reserved_52[0x1f];
+ u8 cqe_zip[0x1];
+
+ u8 cqe_zip_timeout[0x10];
+ u8 cqe_zip_max_num[0x10];
+
+ u8 reserved_53[0x220];
+};
+
+enum {
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_OUTER_SEC_PRIO_PRESENT = 0x1b,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_OUTER_SEC_CFI_PRESENT = 0x1a,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_OUTER_SEC_VID_PRESENT = 0x19,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_INNER_SEC_PRIO_PRESENT = 0x18,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_INNER_SEC_CFI_PRESENT = 0x17,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_INNER_SEC_VID_PRESENT = 0x16,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_OUTER_SEC_VLAN_PRESENT = 0x15,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_INNER_SEC_VLAN_PRESENT = 0x14,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_GRE_PROTO_PRESENT = 0x10,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_GRE_KEY_H_PRESENT = 0xf,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_GRE_KEY_L_PRESENT = 0xe,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_VXLAN_VNI_PRESENT = 0xd,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_OUTER_IPV6_FLOW_LABEL_PRESENT = 0x8,
+ MLX5_FLOW_TABLE_ENTRY_SET_MISC_FIELDS_VALID_BITS_INNER_IPV6_FLOW_LABEL_PRESENT = 0x7,
+};
+
+struct mlx5_ifc_flow_table_entry_set_misc_fields_bits {
+ u8 valid_bits[0x20];
+
+ u8 reserved_0[0x20];
+
+ u8 outer_second_prio[0x3];
+ u8 outer_second_cfi[0x1];
+ u8 outer_second_vid[0xc];
+ u8 inner_second_prio[0x3];
+ u8 inner_second_cfi[0x1];
+ u8 inner_second_vid[0xc];
+
+ u8 outer_second_vlan_tag[0x1];
+ u8 inner_second_vlan_tag[0x1];
+ u8 reserved_1[0xe];
+ u8 gre_protcol[0x10];
+
+ u8 gre_key_h[0x18];
+ u8 gre_key_l[0x8];
+
+ u8 vxlan_vni[0x18];
+ u8 reserved_2[0x8];
+
+ u8 reserved_3[0x20];
+
+ u8 reserved_4[0xc];
+ u8 outer_ipv6_flow_label[0x14];
+
+ u8 reserved_5[0xc];
+ u8 inner_ipv6_flow_label[0x14];
+
+ u8 reserved_6[0xe0];
+};
+
+enum {
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_SMAC_PRESENT = 0x18,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_ETHER_TYPE_PRESENT = 0x17,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_DMAC_PRESENT = 0x16,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_FIRST_PRIO_PRESENT = 0x15,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_FIRST_CFI_PRESENT = 0x14,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_FIRST_VID_PRESENT = 0x13,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_IP_PROROTOCOL_PRESENT = 0x12,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCLASS_PRESENT = 0x11,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_VLAN_TAG_PRESENT = 0x10,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_MCAST_PRESENT = 0xf,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_FRAG_PRESENT = 0xe,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_8_PRESENT = 0xc,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_7_PRESENT = 0xb,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_6_PRESENT = 0xa,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_5_PRESENT = 0x9,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_4_PRESENT = 0x8,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_3_PRESENT = 0x7,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_2_PRESENT = 0x6,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_1_PRESENT = 0x5,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_TCP_FLAGS_0_PRESENT = 0x4,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_L4_SPORT_PRESENT = 0x3,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_L4_DPORT_PRESENT = 0x2,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_SRC_IP_PRESENT = 0x1,
+ MLX5_FLOW_TABLE_ENTRY_SET_BASE_LAYER2_4_VALID_BITS_DST_IP_PRESENT = 0x0,
+};
+
+struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits {
+ u8 valid_bits[0x20];
+
+ u8 reserved_0[0x20];
+
+ u8 smac[6][0x8];
+
+ u8 ether_type[0x10];
+
+ u8 dmac[6][0x8];
+
+ u8 first_prio[0x3];
+ u8 first_cfi[0x1];
+ u8 first_vid[0xc];
+
+ u8 ip_protocol[0x8];
+ u8 tclass[0x8];
+ u8 vlan_tag[0x1];
+ u8 mcast[0x1];
+ u8 frag[0x1];
+ u8 reserved_1[0x4];
+ u8 tcp_flags[0x9];
+
+ u8 l4_sport[0x10];
+ u8 l4_dport[0x10];
+
+ u8 src_ip[16][0x8];
+
+ u8 dst_ip[16][0x8];
+};
+
+struct mlx5_ifc_traffic_counter_bits {
+ u8 packets[8][0x8];
+
+ u8 octets[8][0x8];
+};
+
+enum {
+ MLX5_TISC_STATE_RDY = 0x1,
+ MLX5_TISC_STATE_TSD = 0x2,
+};
+
+struct mlx5_ifc_tisc_bits {
+ u8 reserved_0[0x8];
+ u8 state[0x4];
+ u8 prio[0x4];
+ u8 reserved_1[0x10];
+
+ u8 reserved_2[0x4e0];
+};
+
+enum {
+ MLX5_TIRC_DISP_TYPE_TIR_RCV_DISP_ALL = 0x0,
+ MLX5_TIRC_DISP_TYPE_TIR_RCV_DISP_SINGLE = 0x1,
+};
+
+enum {
+ MLX5_TIRC_LRO_ENABLE_MASK_IPV4_LRO = 0x1,
+ MLX5_TIRC_LRO_ENABLE_MASK_IPV6_LRO = 0x2,
+};
+
+enum {
+ MLX5_TIRC_RX_HASH_FN_HASH_NONE = 0x0,
+ MLX5_TIRC_RX_HASH_FN_HASH_XOR = 0x1,
+ MLX5_TIRC_RX_HASH_FN_HASH_TOEPLITZ = 0x2,
+};
+
+struct mlx5_ifc_tirc_bits {
+ u8 reserved_0[0x20];
+
+ u8 disp_type[0x4];
+ u8 reserved_1[0x1c];
+
+ u8 reserved_2[0x40];
+
+ u8 lro_timeout_period_usecs[0x4];
+ u8 lro_enable_mask[0x4];
+ u8 reserved_3[0x10];
+ u8 lro_max_ip_payload_size[0x8];
+
+ u8 reserved_4[0x20];
+
+ u8 reserved_5[0x10];
+ u8 rq_lst_sz[0x10];
+
+ u8 reserved_6[0x20];
+
+ u8 rx_hash_symmetric[0x1];
+ u8 rx_hash_inner[0x1];
+ u8 reserved_7[0x6];
+ u8 rx_hash_default_rqn[0x18];
+
+ u8 rx_hash_fn[0x4];
+ u8 reserved_8[0x1c];
+
+ u8 rx_hash_toeplitz_key[40][0x8];
+
+ struct mlx5_ifc_rss_hash_selector_bits rx_hash_field_selector_ipv4;
+
+ struct mlx5_ifc_rss_hash_selector_bits rx_hash_field_selector_ipv6;
+
+ u8 reserved_9[0x480];
+
+ struct mlx5_ifc_rq_num_bits rq_num[0];
+};
+
+struct mlx5_ifc_srqc_bits {
+ u8 state[0x4];
+ u8 log_srq_size[0x4];
+ u8 reserved_0[0x18];
+
+ u8 wq_signature[0x1];
+ u8 cont_srq[0x1];
+ u8 reserved_1[0x1];
+ u8 rlky[0x1];
+ u8 reserved_2[0x1];
+ u8 log_rq_stride[0x3];
+ u8 xrcd[0x18];
+
+ u8 page_offset[0x6];
+ u8 reserved_3[0x2];
+ u8 cqn[0x18];
+
+ u8 reserved_4[0x20];
+
+ u8 reserved_5[0x2];
+ u8 log_page_size[0x6];
+ u8 reserved_6[0x18];
+
+ u8 reserved_7[0x20];
+
+ u8 reserved_8[0x8];
+ u8 pd[0x18];
+
+ u8 lwm[0x10];
+ u8 wqe_cnt[0x10];
+
+ u8 reserved_9[0x40];
+
+ u8 db_record_addr_h[0x20];
+
+ u8 db_record_addr_l[0x1e];
+ u8 reserved_10[0x2];
+
+ u8 reserved_11[0x80];
+};
+
+enum {
+ MLX5_SQC_STATE_RDY = 0x1,
+ MLX5_SQC_STATE_ERR = 0x3,
+};
+
+struct mlx5_ifc_sqc_bits {
+ u8 rlky[0x1];
+ u8 cd_master[0x1];
+ u8 fre[0x1];
+ u8 reserved_0[0x5];
+ u8 state[0x4];
+ u8 reserved_1[0x14];
+
+ u8 reserved_2[0x8];
+ u8 sq_index[0x18];
+
+ u8 reserved_3[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_4[0xa0];
+
+ u8 tis_lst_sz[0x10];
+ u8 reserved_5[0x10];
+
+ u8 reserved_6[0x40];
+
+ u8 reserved_7[0x8];
+ u8 tis_num_0[0x18];
+
+ struct mlx5_ifc_wq_bits wq;
+};
+
+enum {
+ MLX5_RQC_CS_CS_DISABLE = 0x0,
+ MLX5_RQC_CS_CS_32B = 0x1,
+ MLX5_RQC_CS_CS_64B = 0x2,
+};
+
+enum {
+ MLX5_RQ_TYPE_DIRECT = 0x0,
+ MLX5_RQ_TYPE_RMP = 0x1,
+ MLX5_RQ_TYPE_OVERLAY = 0x2,
+};
+
+enum {
+ MLX5_RQC_STATE_RDY = 0x1,
+ MLX5_RQC_STATE_ERR = 0x3,
+};
+
+struct mlx5_ifc_rqc_bits {
+ u8 rlky[0x1];
+ u8 cs[0x2];
+ u8 vsd[0x1];
+ u8 rq_type[0x4];
+ u8 state[0x4];
+ u8 reserved_0[0x14];
+
+ u8 reserved_1[0x8];
+ u8 rq_index[0x18];
+
+ u8 reserved_2[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 reserved_4[0x8];
+ u8 rmpn[0x18];
+
+ u8 reserved_5[0xe0];
+
+ struct mlx5_ifc_wq_bits wq;
+};
+
+enum {
+ MLX5_RMPC_STATE_RDY = 0x1,
+ MLX5_RMPC_STATE_ERR = 0x3,
+};
+
+struct mlx5_ifc_rmpc_bits {
+ u8 reserved_0[0x8];
+ u8 state[0x4];
+ u8 reserved_1[0x14];
+
+ u8 reserved_2[0xe0];
+
+ struct mlx5_ifc_wq_bits wq;
+};
+
+struct mlx5_ifc_nic_vport_context_bits {
+ u8 permanent_addresses_perm[0x1];
+ u8 reserved_0[0x1e];
+ u8 roce_en[0x1];
+
+ u8 reserved_1[0x760];
+
+ u8 reserved_2[0x14];
+ u8 num_of_permanent_addresses[0xc];
+
+ u8 reserved_3[0x60];
+
+ u8 permanent_address[0][0x40];
+};
+
+struct mlx5_ifc_mkc_bits {
+ u8 reserved_0[0x1];
+ u8 free[0x1];
+ u8 reserved_1[0xd];
+ u8 small_fence_on_rdma_read_response[0x1];
+ u8 umr_en[0x1];
+ u8 a[0x1];
+ u8 rw[0x1];
+ u8 rr[0x1];
+ u8 lw[0x1];
+ u8 lr[0x1];
+ u8 access_mode[0x2];
+ u8 reserved_2[0x8];
+
+ u8 qpn[0x18];
+ u8 mkey[0x8];
+
+ u8 reserved_3[0x20];
+
+ u8 length64[0x1];
+ u8 bsf_en[0x1];
+ u8 sync_umr[0x1];
+ u8 reserved_4[0x2];
+ u8 expected_sigerr_count[0x1];
+ u8 reserved_5[0x1];
+ u8 en_rinval[0x1];
+ u8 pd[0x18];
+
+ u8 start_addr[0x40];
+
+ u8 len[0x40];
+
+ u8 bsf_octword_size[0x20];
+
+ u8 reserved_6[0x80];
+
+ u8 translations_octword_size[0x20];
+
+ u8 reserved_7[0x1b];
+ u8 log_page_size[0x5];
+
+ u8 reserved_8[0x20];
+};
+
+struct mlx5_ifc_esw_vport_context_bits {
+ u8 reserved_0[0x780];
+
+ u8 reserved_1[0x14];
+ u8 num_of_allowed_forwarding_addresses[0xc];
+
+ u8 reserved_2[0x60];
+
+ struct mlx5_ifc_allowed_address_context_bits allowed_forwarding_address[0];
+};
+
+struct mlx5_ifc_eqc_bits {
+ u8 status[0x4];
+ u8 reserved_0[0x9];
+ u8 ec[0x1];
+ u8 oi[0x1];
+ u8 reserved_1[0x5];
+ u8 st[0x4];
+ u8 reserved_2[0x8];
+
+ u8 reserved_3[0x20];
+
+ u8 reserved_4[0x14];
+ u8 page_offset[0x6];
+ u8 reserved_5[0x6];
+
+ u8 reserved_6[0x3];
+ u8 log_eq_size[0x5];
+ u8 uar[0x18];
+
+ u8 reserved_7[0x20];
+
+ u8 reserved_8[0x18];
+ u8 intr[0x8];
+
+ u8 reserved_9[0x3];
+ u8 log_page_size[0x5];
+ u8 reserved_10[0x18];
+
+ u8 reserved_11[0x60];
+
+ u8 reserved_12[0x8];
+ u8 consumer_counter[0x18];
+
+ u8 reserved_13[0x8];
+ u8 producer_counter[0x18];
+
+ u8 reserved_14[0x80];
+};
+
+struct mlx5_ifc_dctc_bits {
+ u8 reserved_0[0x4];
+ u8 state[0x4];
+ u8 reserved_1[0x18];
+
+ u8 reserved_2[0x8];
+ u8 sw_context[0x18];
+
+ u8 reserved_3[0x8];
+ u8 cqn[0x18];
+
+ u8 counter_set_id[0x8];
+ u8 atomic_mode[0x4];
+ u8 rre[0x1];
+ u8 rwe[0x1];
+ u8 rae[0x1];
+ u8 atomic_like_write_en[0x1];
+ u8 latency_sensitive[0x1];
+ u8 rlky[0x1];
+ u8 free_ar[0x1];
+ u8 reserved_4[0xd];
+
+ u8 reserved_5[0x8];
+ u8 cs_res[0x8];
+ u8 reserved_6[0x3];
+ u8 min_rnr_nak[0x5];
+ u8 reserved_7[0x8];
+
+ u8 reserved_8[0x8];
+ u8 srqn[0x18];
+
+ u8 reserved_9[0x8];
+ u8 pd[0x18];
+
+ u8 tclass[0x8];
+ u8 reserved_10[0x4];
+ u8 flow_label[0x14];
+
+ u8 dc_access_key[0x40];
+
+ u8 reserved_11[0x5];
+ u8 mtu[0x3];
+ u8 port[0x8];
+ u8 pkey_index[0x10];
+
+ u8 reserved_12[0x8];
+ u8 my_addr_index[0x8];
+ u8 reserved_13[0x8];
+ u8 hop_limit[0x8];
+
+ u8 dc_access_key_violation_count[0x20];
+
+ u8 reserved_14[0x14];
+ u8 dei_cfi[0x1];
+ u8 eth_prio[0x3];
+ u8 ecn[0x2];
+ u8 dscp[0x6];
+
+ u8 reserved_15[0x40];
+};
+
+struct mlx5_ifc_cqc_bits {
+ u8 status[0x4];
+ u8 reserved_0[0x4];
+ u8 cqe_sz[0x3];
+ u8 cc[0x1];
+ u8 reserved_1[0x2];
+ u8 oi[0x1];
+ u8 reserved_2[0x2];
+ u8 cqe_zip_en[0x1];
+ u8 mini_cqe_res_format[0x2];
+ u8 st[0x4];
+ u8 reserved_3[0x8];
+
+ u8 reserved_4[0x20];
+
+ u8 reserved_5[0x14];
+ u8 page_offset[0x6];
+ u8 reserved_6[0x6];
+
+ u8 reserved_7[0x3];
+ u8 log_cq_size[0x5];
+ u8 uar_page[0x18];
+
+ u8 reserved_8[0x4];
+ u8 cq_period[0xc];
+ u8 cq_max_count[0x10];
+
+ u8 reserved_9[0x18];
+ u8 c_eqn[0x8];
+
+ u8 reserved_10[0x3];
+ u8 log_page_size[0x5];
+ u8 reserved_11[0x18];
+
+ u8 reserved_12[0x20];
+
+ u8 reserved_13[0x8];
+ u8 last_notified_index[0x18];
+
+ u8 reserved_14[0x8];
+ u8 last_solicit_index[0x18];
+
+ u8 reserved_15[0x8];
+ u8 consumer_counter[0x18];
+
+ u8 reserved_16[0x8];
+ u8 producer_counter[0x18];
+
+ u8 reserved_17[0x40];
+
+ u8 dbr_addr[0x40];
+};
+
+struct mlx5_ifc_query_adapter_param_block_bits {
+ u8 reserved_0[0x80];
+
+ u8 intapin[0x8];
+ u8 reserved_1[0x18];
+
+ u8 reserved_2[0x40];
+
+ u8 reserved_3[0x10];
+ u8 vsd_vendor_id[0x10];
+
+ u8 vsd[208][0x8];
+
+ u8 vsd_contd_psid[16][0x8];
+};
+
+struct mlx5_ifc_flow_table_field_bitmask_bits {
+ u8 outer_dmac[0x1];
+ u8 outer_smac[0x1];
+ u8 outer_ether_type[0x1];
+ u8 reserved_0[0x1];
+ u8 outer_first_prio[0x1];
+ u8 outer_first_cfi[0x1];
+ u8 outer_first_vid[0x1];
+ u8 reserved_1[0x1];
+ u8 outer_second_prio[0x1];
+ u8 outer_second_cfi[0x1];
+ u8 outer_second_vid[0x1];
+ u8 reserved_2[0x1];
+ u8 outer_sip[0x1];
+ u8 outer_dip[0x1];
+ u8 outer_frag[0x1];
+ u8 outer_ip_protocol[0x1];
+ u8 reserved_3[0x4];
+ u8 outer_l4_sport[0x1];
+ u8 outer_l4_dport[0x1];
+ u8 outer_tcp_flags[0x1];
+ u8 outer_gre_protcol[0x1];
+ u8 outer_gre_key[0x1];
+ u8 outer_vxlan_vni[0x1];
+ u8 reserved_4[0x6];
+
+ u8 inner_dmac[0x1];
+ u8 inner_smac[0x1];
+ u8 inner_ether_type[0x1];
+ u8 reserved_5[0x1];
+ u8 inner_first_prio[0x1];
+ u8 inner_first_cfi[0x1];
+ u8 inner_first_vid[0x1];
+ u8 reserved_6[0x1];
+ u8 inner_second_prio[0x1];
+ u8 inner_second_cfi[0x1];
+ u8 inner_second_vid[0x1];
+ u8 reserved_7[0x1];
+ u8 inner_sip[0x1];
+ u8 inner_dip[0x1];
+ u8 inner_frag[0x1];
+ u8 inner_ip_protocol[0x1];
+ u8 reserved_8[0x4];
+ u8 inner_l4_sport[0x1];
+ u8 inner_l4_dport[0x1];
+ u8 inner_tcp_flags[0x1];
+ u8 reserved_9[0x9];
+
+ u8 reserved_10[0x40];
+};
+
+struct mlx5_ifc_odp_per_transport_service_cap_bits {
+ u8 send[0x1];
+ u8 receive[0x1];
+ u8 write[0x1];
+ u8 reserved_0[0x1d];
+};
+
+struct mlx5_ifc_register_loopback_control_bits {
+ u8 no_lb[0x1];
+ u8 reserved_0[0x7];
+ u8 port[0x8];
+ u8 reserved_1[0x10];
+
+ u8 reserved_2[0x60];
+};
+
+struct mlx5_ifc_teardown_hca_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+enum {
+ MLX5_TEARDOWN_HCA_IN_OPCODE_GRACEFUL_CLOSE = 0x0,
+ MLX5_TEARDOWN_HCA_IN_OPCODE_PANIC_CLOSE = 0x1,
+};
+
+struct mlx5_ifc_teardown_hca_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x10];
+ u8 profile[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_sqerr2rts_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_sqerr2rts_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_set_roce_address_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_set_roce_address_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 roce_address_index[0x10];
+ u8 reserved_2[0x10];
+
+ u8 reserved_3[0x20];
+
+ struct mlx5_ifc_roce_addr_layout_bits roce_address;
+};
+
+struct mlx5_ifc_set_mad_demux_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+enum {
+ MLX5_SET_MAD_DEMUX_IN_DEMUX_MODE_PASS_ALL = 0x0,
+ MLX5_SET_MAD_DEMUX_IN_DEMUX_MODE_SELECTIVE = 0x2,
+};
+
+struct mlx5_ifc_set_mad_demux_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x20];
+
+ u8 reserved_3[0x6];
+ u8 demux_mode[0x2];
+ u8 reserved_4[0x18];
+};
+
+struct mlx5_ifc_set_hca_cap_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_set_hca_cap_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_cmd_hca_cap_bits hca_capability_struct;
+};
+
+struct mlx5_ifc_set_flow_table_entry_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+enum {
+ MLX5_SET_FLOW_TABLE_ENTRY_IN_DESTINATION_TYPE_TIR = 0x1,
+};
+
+struct mlx5_ifc_set_flow_table_entry_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x20];
+
+ u8 flow_index[0x20];
+
+ u8 destination_type[0x8];
+ u8 destination_id[0x18];
+
+ u8 reserved_3[0x8];
+ u8 table_id[0x18];
+
+ u8 valid[0x1];
+ u8 reserved_4[0x19];
+ u8 miscellaneous_headers_masks_valid[0x1];
+ u8 inner_packet_headers_masks_valid[0x1];
+ u8 outer_packet_headers_masks_valid[0x1];
+ u8 miscellaneous_headers_valid[0x1];
+ u8 inner_packet_headers_valid[0x1];
+ u8 outer_packet_headers_valid[0x1];
+
+ u8 reserved_5[0x20];
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits outer_packet_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits inner_packet_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_misc_fields_bits miscellaneous_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits outer_packet_headers_masks;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits inner_packet_headers_masks;
+
+ struct mlx5_ifc_flow_table_entry_set_misc_fields_bits miscellaneous_headers_masks;
+
+ u8 reserved_6[0x800];
+};
+
+struct mlx5_ifc_set_burst_size_out_bits {
+ u8 status[0x10];
+ u8 reserved_0[0x10];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_set_burst_size_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x20];
+
+ u8 reserved_3[0x9];
+ u8 device_burst_size[0x17];
+};
+
+struct mlx5_ifc_rts2rts_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_rts2rts_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_rtr2rts_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_rtr2rts_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_rst2init_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_rst2init_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_roce_cap_bits {
+ u8 reserved_0[0x80];
+
+ u8 reserved_1[0xc];
+ u8 l3_type[0x4];
+ u8 reserved_2[0x8];
+ u8 roce_version[0x8];
+
+ u8 reserved_3[0x10];
+ u8 r_roce_dest_udp_port[0x10];
+
+ u8 r_roce_max_src_udp_port[0x10];
+ u8 r_roce_min_src_udp_port[0x10];
+
+ u8 reserved_4[0x10];
+ u8 roce_address_table_size[0x10];
+
+ u8 reserved_5[0x700];
+};
+
+struct mlx5_ifc_resize_field_select_bits {
+ u8 resize_field_select[0x20];
+};
+
+enum {
+ MLX5_QUERY_VPORT_STATE_OUT_STATE_DOWN = 0x0,
+ MLX5_QUERY_VPORT_STATE_OUT_STATE_UP = 0x1,
+};
+
+struct mlx5_ifc_query_vport_state_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x20];
+
+ u8 reserved_2[0x1c];
+ u8 state[0x4];
+};
+
+enum {
+ MLX5_QUERY_VPORT_STATE_IN_OP_MOD_VNIC_VPORT = 0x0,
+ MLX5_QUERY_VPORT_STATE_IN_OP_MOD_ESW_VPORT = 0x1,
+ MLX5_QUERY_VPORT_STATE_IN_OP_MOD_UPLINK = 0x2,
+};
+
+struct mlx5_ifc_query_vport_state_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_vport_counter_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_traffic_counter_bits received_errors;
+
+ struct mlx5_ifc_traffic_counter_bits transmit_errors;
+
+ struct mlx5_ifc_traffic_counter_bits received_unicast;
+
+ struct mlx5_ifc_traffic_counter_bits transmitted_unicast;
+
+ struct mlx5_ifc_traffic_counter_bits received_multicast;
+
+ struct mlx5_ifc_traffic_counter_bits transmitted_multicast;
+
+ struct mlx5_ifc_traffic_counter_bits received_broadcast;
+
+ struct mlx5_ifc_traffic_counter_bits transmitted_broadcast;
+
+ u8 reserved_2[0x400];
+};
+
+struct mlx5_ifc_query_vport_counter_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 reserved_3[0x60];
+
+ u8 clear[0x1];
+ u8 reserved_4[0x1f];
+
+ u8 reserved_5[0x20];
+};
+
+struct mlx5_ifc_query_tis_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_tisc_bits tis_context;
+};
+
+struct mlx5_ifc_query_tis_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 tis_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_tir_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0xc0];
+
+ struct mlx5_ifc_tirc_bits tir_context;
+};
+
+struct mlx5_ifc_query_tir_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 tir_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_srq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_srqc_bits srq_context_entry;
+
+ u8 reserved_2[0x600];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_query_srq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 srqn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_sq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_sqc_bits sq_context;
+};
+
+struct mlx5_ifc_query_sq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 sq_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_special_contexts_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x20];
+
+ u8 resd_lkey[0x20];
+};
+
+struct mlx5_ifc_query_special_contexts_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_rq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_rqc_bits rq_context;
+};
+
+struct mlx5_ifc_query_rq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 rq_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_roce_address_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_roce_addr_layout_bits roce_address;
+};
+
+struct mlx5_ifc_query_roce_address_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 roce_address_index[0x10];
+ u8 reserved_2[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_rmp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_rmpc_bits rmp_context;
+};
+
+struct mlx5_ifc_query_rmp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 rmp_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_2[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_3[0x80];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_query_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_q_counter_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 rx_write_requests[0x20];
+
+ u8 reserved_2[0x20];
+
+ u8 rx_read_requests[0x20];
+
+ u8 reserved_3[0x20];
+
+ u8 rx_atomic_requests[0x20];
+
+ u8 reserved_4[0x20];
+
+ u8 rx_dct_connect[0x20];
+
+ u8 reserved_5[0x20];
+
+ u8 out_of_buffer[0x20];
+
+ u8 reserved_6[0x660];
+};
+
+struct mlx5_ifc_query_q_counter_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x80];
+
+ u8 clear[0x1];
+ u8 reserved_3[0x1f];
+
+ u8 reserved_4[0x18];
+ u8 counter_set_id[0x8];
+};
+
+struct mlx5_ifc_query_pages_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x10];
+ u8 function_id[0x10];
+
+ u8 num_pages[0x20];
+};
+
+enum {
+ MLX5_QUERY_PAGES_IN_OP_MOD_BOOT_PAGES = 0x1,
+ MLX5_QUERY_PAGES_IN_OP_MOD_INIT_PAGES = 0x2,
+ MLX5_QUERY_PAGES_IN_OP_MOD_REGULAR_PAGES = 0x3,
+};
+
+struct mlx5_ifc_query_pages_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_nic_vport_context_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_nic_vport_context_bits nic_vport_context;
+};
+
+struct mlx5_ifc_query_nic_vport_context_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_mkey_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_mkc_bits memory_key_mkey_entry;
+
+ u8 reserved_2[0x600];
+
+ u8 bsf0_klm0_pas_mtt0_1[16][0x8];
+
+ u8 bsf1_klm1_pas_mtt2_3[16][0x8];
+};
+
+struct mlx5_ifc_query_mkey_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 mkey_index[0x18];
+
+ u8 pg_access[0x1];
+ u8 reserved_3[0x1f];
+};
+
+struct mlx5_ifc_query_mad_demux_out_bits {
+ u8 status[0x10];
+ u8 reserved_0[0x10];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 mad_dumux_parameters_block[0x20];
+};
+
+struct mlx5_ifc_query_mad_demux_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_hca_cap_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 capability_struct[256][0x8];
+};
+
+struct mlx5_ifc_query_hca_cap_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_flow_table_entry_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x80];
+
+ u8 valid[0x1];
+ u8 reserved_2[0x19];
+ u8 miscellaneous_headers_masks_valid[0x1];
+ u8 inner_packet_headers_masks_valid[0x1];
+ u8 outer_packet_headers_masks_valid[0x1];
+ u8 miscellaneous_headers_valid[0x1];
+ u8 inner_packet_headers_valid[0x1];
+ u8 outer_packet_headers_valid[0x1];
+
+ u8 reserved_3[0x20];
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits outer_packet_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits inner_packet_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_misc_fields_bits miscellaneous_headers;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits outer_packet_headers_masks;
+
+ struct mlx5_ifc_flow_table_entry_set_base_layer2_4_bits inner_packet_headers_masks;
+
+ struct mlx5_ifc_flow_table_entry_set_misc_fields_bits miscellaneous_headers_masks;
+
+ u8 reserved_4[0x800];
+};
+
+struct mlx5_ifc_query_flow_table_entry_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x20];
+
+ u8 flow_index[0x20];
+
+ u8 reserved_3[0x20];
+
+ u8 reserved_4[0x8];
+ u8 table_id[0x18];
+
+ u8 reserved_5[0x40];
+};
+
+struct mlx5_ifc_query_esw_vport_context_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_esw_vport_context_bits esw_vport_context;
+};
+
+struct mlx5_ifc_query_esw_vport_context_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_eq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_eqc_bits eq_context_entry;
+
+ u8 reserved_2[0x40];
+
+ u8 event_bitmask[0x40];
+
+ u8 reserved_3[0x580];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_query_eq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x18];
+ u8 eq_number[0x8];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_dct_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_dctc_bits dct_context_entry;
+
+ u8 reserved_2[0x180];
+};
+
+struct mlx5_ifc_query_dct_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 dct_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_cq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_cqc_bits cq_context;
+
+ u8 reserved_2[0x600];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_query_cq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_cong_statistics_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 cur_flows[0x20];
+
+ u8 sum_flows[0x20];
+
+ u8 cnp_ignored_high[0x20];
+
+ u8 cnp_ignored_low[0x20];
+
+ u8 cnp_handled_high[0x20];
+
+ u8 cnp_handled_low[0x20];
+
+ u8 cur_total_rate_high[0x20];
+
+ u8 cur_total_rate_low[0x20];
+
+ u8 sum_total_rate_high[0x20];
+
+ u8 sum_total_rate_low[0x20];
+
+ u8 min_total_rate_high[0x20];
+
+ u8 min_total_rate_low[0x20];
+
+ u8 max_total_rate_high[0x20];
+
+ u8 max_total_rate_low[0x20];
+
+ u8 time_stamp_high[0x20];
+
+ u8 time_stamp_low[0x20];
+
+ u8 accumulators_period[0x20];
+
+ u8 ecn_marked_roce_packets_high[0x20];
+
+ u8 ecn_marked_roce_packets_low[0x20];
+
+ u8 cnps_sent_high[0x20];
+
+ u8 cnps_sent_low[0x20];
+
+ u8 reserved_2[0x560];
+};
+
+struct mlx5_ifc_query_cong_statistics_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 clear[0x1];
+ u8 reserved_2[0x1f];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_cong_params_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 congestion_parameters802_1qau_rp_r_roce_ecn_rp_r_roce_ecn_np[256][0x8];
+};
+
+struct mlx5_ifc_query_cong_params_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x1c];
+ u8 cong_protocol[0x4];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_query_burst_size_out_bits {
+ u8 status[0x10];
+ u8 reserved_0[0x10];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x20];
+
+ u8 reserved_2[0x9];
+ u8 device_burst_size[0x17];
+};
+
+struct mlx5_ifc_query_burst_size_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_query_adapter_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ struct mlx5_ifc_query_adapter_param_block_bits query_adapter_struct;
+};
+
+struct mlx5_ifc_query_adapter_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_per_protocol_networking_offload_caps_bits {
+ u8 csum_cap[0x1];
+ u8 vlan_cap[0x1];
+ u8 lro_cap[0x1];
+ u8 lro_psh_flag[0x1];
+ u8 lro_time_stamp[0x1];
+ u8 reserved_0[0x6];
+ u8 max_lso_cap[0x5];
+ u8 reserved_1[0x4];
+ u8 rss_ind_tbl_cap[0x4];
+ u8 reserved_2[0x8];
+
+ u8 reserved_3[0x2];
+ u8 log_flow_table_sz[0x6];
+ u8 reserved_4[0x18];
+
+ u8 reserved_5[0x10];
+ u8 lro_min_mss_size[0x10];
+
+ u8 reserved_6[0x20];
+
+ struct mlx5_ifc_flow_table_field_bitmask_bits flow_table_field_support;
+
+ struct mlx5_ifc_flow_table_field_bitmask_bits flow_table_field_bitmask_support;
+
+ u8 lro_timer_supported_periods[4][0x20];
+
+ u8 reserved_7[0x80];
+};
+
+struct mlx5_ifc_page_fault_resume_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_page_fault_resume_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 error[0x1];
+ u8 reserved_2[0x4];
+ u8 rdma[0x1];
+ u8 read_write[0x1];
+ u8 requestor[0x1];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_odp_cap_bits {
+ u8 reserved_0[0x40];
+
+ u8 sig[0x1];
+ u8 reserved_1[0x1f];
+
+ u8 reserved_2[0x20];
+
+ struct mlx5_ifc_odp_per_transport_service_cap_bits rc_odp_caps;
+
+ struct mlx5_ifc_odp_per_transport_service_cap_bits uc_odp_caps;
+
+ struct mlx5_ifc_odp_per_transport_service_cap_bits ud_odp_caps;
+
+ u8 reserved_3[0x720];
+};
+
+struct mlx5_ifc_modify_vport_state_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_vport_state_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 reserved_3[0x1c];
+ u8 state[0x4];
+};
+
+struct mlx5_ifc_modify_tis_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_tis_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 tis_state[0x4];
+ u8 reserved_2[0x4];
+ u8 tis_number[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 modify_bitmask[8][0x8];
+
+ u8 reserved_4[0x40];
+
+ struct mlx5_ifc_tisc_bits ctx;
+};
+
+struct mlx5_ifc_modify_tir_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_tir_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 tir_number[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 modify_bitmask[8][0x8];
+
+ u8 reserved_4[0x40];
+
+ struct mlx5_ifc_tirc_bits ctx;
+};
+
+struct mlx5_ifc_modify_sq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_sq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 sq_state[0x4];
+ u8 reserved_2[0x4];
+ u8 sq_number[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 modify_bitmask[8][0x8];
+
+ u8 reserved_4[0x40];
+
+ struct mlx5_ifc_sqc_bits ctx;
+};
+
+struct mlx5_ifc_modify_rq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_rq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 rq_state[0x4];
+ u8 reserved_2[0x4];
+ u8 rq_number[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 modify_bitmask[8][0x8];
+
+ u8 reserved_4[0x40];
+
+ struct mlx5_ifc_rqc_bits ctx;
+};
+
+struct mlx5_ifc_modify_rmp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_rmp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 rmp_state[0x4];
+ u8 reserved_2[0x4];
+ u8 rmp_number[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 modify_bitmask[8][0x8];
+
+ struct mlx5_ifc_rmpc_bits ctx;
+};
+
+struct mlx5_ifc_modify_nic_vport_context_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_nic_vport_context_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 field_select[0x20];
+
+ u8 reserved_3[0x780];
+
+ struct mlx5_ifc_nic_vport_context_bits nic_vport_context;
+};
+
+struct mlx5_ifc_modify_field_select_bits {
+ u8 modify_field_select[0x20];
+};
+
+struct mlx5_ifc_modify_esw_vport_context_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_esw_vport_context_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 vport_number[0x10];
+
+ u8 field_select[0x20];
+
+ u8 reserved_3[0x780];
+
+ struct mlx5_ifc_esw_vport_context_bits esw_vport_context;
+};
+
+struct mlx5_ifc_modify_cq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_cq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 cqn[0x18];
+
+ u8 modify_field_select_resize_field_select[0x20];
+
+ struct mlx5_ifc_cqc_bits cq_context;
+
+ u8 reserved_3[0x600];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_modify_cong_params_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_modify_cong_params_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x1c];
+ u8 cong_protocol[0x4];
+
+ u8 field_select_8021qaurp_field_select_r_roce_rp_field_select_r_r[0x20];
+
+ u8 reserved_3[0x80];
+
+ u8 congestion_parameters802_1qau_rp_r_roce_ecn_rp_r_roce_ecn_np[272][0x8];
+};
+
+struct mlx5_ifc_manage_pages_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 output_num_entries[0x20];
+
+ u8 reserved_1[0x20];
+
+ u8 pas[0][0x40];
+};
+
+enum {
+ MLX5_MANAGE_PAGES_IN_OP_MOD_ALLOCATION_FAIL = 0x0,
+ MLX5_MANAGE_PAGES_IN_OP_MOD_ALLOCATION_SUCCESS = 0x1,
+ MLX5_MANAGE_PAGES_IN_OP_MOD_HCA_RETURN_PAGES = 0x2,
+};
+
+struct mlx5_ifc_manage_pages_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x10];
+ u8 function_id[0x10];
+
+ u8 input_num_entries[0x20];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_mad_ifc_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 response_mad_packet[256][0x8];
+};
+
+struct mlx5_ifc_mad_ifc_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 remote_lid[0x10];
+ u8 reserved_2[0x8];
+ u8 port[0x8];
+
+ u8 reserved_3[0x20];
+
+ u8 request_mad_packet[256][0x8];
+};
+
+struct mlx5_ifc_init_hca_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_init_hca_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x10];
+ u8 profile[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_init2rtr_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_init2rtr_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_init2init_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_init2init_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_4[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_5[0x80];
+};
+
+struct mlx5_ifc_get_dropped_packet_log_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 packet_headers_log[128][0x8];
+
+ u8 packet_syndrome[64][0x8];
+};
+
+struct mlx5_ifc_gen_eqe_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x18];
+ u8 eq_number[0x8];
+
+ u8 reserved_3[0x20];
+
+ u8 eqe[64][0x8];
+};
+
+struct mlx5_ifc_gen_eq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_filed_select_r_roce_rp_bits {
+ u8 field_select_r_roce_rp[0x20];
+};
+
+struct mlx5_ifc_filed_select_r_roce_np_bits {
+ u8 field_select_r_roce_np[0x20];
+};
+
+struct mlx5_ifc_filed_select_802_1qau_rp_bits {
+ u8 field_select_8021qaurp[0x20];
+};
+
+struct mlx5_ifc_enable_hca_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x20];
+};
+
+struct mlx5_ifc_enable_hca_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_drain_dct_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_drain_dct_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 dct_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_disable_hca_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x20];
+};
+
+struct mlx5_ifc_disable_hca_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_detach_from_sniffer_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_detach_from_sniffer_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x10];
+ u8 sniffer_rule_id[0x10];
+
+ u8 reserved_3[0xa0];
+};
+
+struct mlx5_ifc_detach_from_mcg_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_detach_from_mcg_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 multicast_gid[16][0x8];
+};
+
+struct mlx5_ifc_destroy_tis_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_tis_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 tis_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_tir_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_tir_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 tir_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_srq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_srq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 srqn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_sq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_sq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 sq_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_rq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_rq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 rq_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_rmp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_rmp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 rmp_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_psv_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_psv_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 psv_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_mkey_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_mkey_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 mkey[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_eq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_eq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x18];
+ u8 eq_number[0x8];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_dct_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_dct_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 dct_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_destroy_cq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_destroy_cq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_dealloc_xrcd_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_dealloc_xrcd_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 xrcd[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_dealloc_uar_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_dealloc_uar_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 uar[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_dealloc_q_counter_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_dealloc_q_counter_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x18];
+ u8 counter_set_id[0x8];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_dealloc_pd_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_dealloc_pd_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 pd[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_deactivate_tracer_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_deactivate_tracer_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 mkey[0x20];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_tis_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 tis_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_tis_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0xc0];
+
+ struct mlx5_ifc_tisc_bits ctx;
+};
+
+struct mlx5_ifc_create_tir_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 tir_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_tir_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0xc0];
+
+ struct mlx5_ifc_tirc_bits ctx;
+};
+
+struct mlx5_ifc_create_srq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 srqn[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_srq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_srqc_bits srq_context_entry;
+
+ u8 reserved_3[0x600];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_create_sq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 sq_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_sq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0xc0];
+
+ struct mlx5_ifc_sqc_bits ctx;
+};
+
+struct mlx5_ifc_create_rq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 rq_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_rq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0xc0];
+
+ struct mlx5_ifc_rqc_bits ctx;
+};
+
+struct mlx5_ifc_create_rmp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 rmp_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_rmp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_rmpc_bits ctx;
+};
+
+struct mlx5_ifc_create_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ u8 opt_param_mask[0x20];
+
+ u8 reserved_3[0x20];
+
+ struct mlx5_ifc_qpc_bits qpc;
+
+ u8 reserved_4[0x80];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_create_psv_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 reserved_2[0x8];
+ u8 psv0_index[0x18];
+
+ u8 reserved_3[0x8];
+ u8 psv1_index[0x18];
+
+ u8 reserved_4[0x8];
+ u8 psv2_index[0x18];
+
+ u8 reserved_5[0x8];
+ u8 psv3_index[0x18];
+};
+
+struct mlx5_ifc_create_psv_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 num_psv[0x4];
+ u8 reserved_2[0x4];
+ u8 pd[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_create_mkey_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 mkey_index[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_mkey_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x20];
+
+ u8 pg_access[0x1];
+ u8 reserved_3[0x1f];
+
+ struct mlx5_ifc_mkc_bits memory_key_mkey_entry;
+
+ u8 reserved_4[0x80];
+
+ u8 translations_octword_actual_size[0x20];
+
+ u8 reserved_5[0x560];
+
+ u8 klm_pas_mtt[0][0x20];
+};
+
+struct mlx5_ifc_create_eq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x18];
+ u8 eq_number[0x8];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_eq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_eqc_bits eq_context_entry;
+
+ u8 reserved_3[0x40];
+
+ u8 event_bitmask[0x40];
+
+ u8 reserved_4[0x580];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_create_dct_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 dct_number[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_dct_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_dctc_bits dct_context_entry;
+
+ u8 reserved_3[0x180];
+};
+
+struct mlx5_ifc_create_cq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_create_cq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+
+ struct mlx5_ifc_cqc_bits cq_context;
+
+ u8 reserved_3[0x600];
+
+ u8 pas[0][0x40];
+};
+
+struct mlx5_ifc_cong_control_r_roce_ecn_rp_bits {
+ u8 reserved_0[0x60];
+
+ u8 cnp_receive_enable[0x1];
+ u8 reserved_1[0x3];
+ u8 clamp_tgt_rate[0x1];
+ u8 reserved_2[0x3];
+ u8 clamp_tgt_rate_after_time_inc[0x1];
+ u8 reserved_3[0x17];
+
+ u8 reserved_4[0x20];
+
+ u8 rpg_time_reset[0x20];
+
+ u8 rpg_byte_reset[0x20];
+
+ u8 rpg_threshold[0x20];
+
+ u8 rpg_max_rate[0x20];
+
+ u8 rpg_ai_rate[0x20];
+
+ u8 rpg_hai_rate[0x20];
+
+ u8 reserved_5[0x20];
+
+ u8 rpg_min_dec_fac[0x20];
+
+ u8 rpg_min_rate[0x20];
+
+ u8 reserved_6[0xe0];
+
+ u8 rate_to_set_on_first_cnp[0x20];
+
+ u8 dce_tcp_g[0x20];
+
+ u8 dce_tcp_rtt[0x20];
+
+ u8 rate_reduce_monitor_period[0x20];
+
+ u8 reserved_7[0x20];
+
+ u8 initial_alpha_value[0x20];
+
+ u8 reserved_8[0x4a0];
+};
+
+struct mlx5_ifc_cong_control_r_roce_ecn_np_bits {
+ u8 reserved_0[0x60];
+
+ u8 ecn_receive_enable[0x1];
+ u8 reserved_1[0x1f];
+
+ u8 reserved_2[0x20];
+
+ u8 min_time_between_cnps[0x20];
+
+ u8 reserved_3[0x12];
+ u8 cnp_dscp[0x6];
+ u8 reserved_4[0x5];
+ u8 cnp_802p_prio[0x3];
+
+ u8 reserved_5[0x720];
+};
+
+struct mlx5_ifc_cong_control_802_1qau_rp_bits {
+ u8 reserved_0[0x60];
+
+ u8 rpg_enable[0x1];
+ u8 tag_enable[0x1];
+ u8 reserved_1[0x1e];
+
+ u8 rppp_max_rps[0x20];
+
+ u8 rpg_time_reset[0x20];
+
+ u8 rpg_byte_reset[0x20];
+
+ u8 rpg_threshold[0x20];
+
+ u8 rpg_max_rate[0x20];
+
+ u8 rpg_ai_rate[0x20];
+
+ u8 rpg_hai_rate[0x20];
+
+ u8 rpg_gd[0x20];
+
+ u8 rpg_min_dec_fac[0x20];
+
+ u8 rpg_min_rate[0x20];
+
+ u8 reserved_2[0x640];
+};
+
+struct mlx5_ifc_config_int_moderation_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x4];
+ u8 min_delay[0xc];
+ u8 int_vector[0x10];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_config_int_moderation_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x4];
+ u8 min_delay[0xc];
+ u8 int_vector[0x10];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_attach_to_sniffer_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x10];
+ u8 sniffer_rule_id[0x10];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_attach_to_sniffer_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 other_vport[0x1];
+ u8 reserved_2[0xf];
+ u8 port_num[0x10];
+
+ u8 sx[0x1];
+ u8 rx[0x1];
+ u8 reserved_3[0x6];
+ u8 tirn[0x18];
+
+ u8 reserved_4[0x1480];
+};
+
+struct mlx5_ifc_attach_to_mcg_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_attach_to_mcg_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+
+ u8 multicast_gid[16][0x8];
+};
+
+struct mlx5_ifc_atomic_cap_bits {
+ u8 reserved_0[0x40];
+
+ u8 atomic_req_endianess[0x1];
+ u8 reserved_1[0x1f];
+
+ u8 reserved_2[0x20];
+
+ u8 reserved_3[0x10];
+ u8 atomic_operations[0x10];
+
+ u8 reserved_4[0x10];
+ u8 atomic_size_qp[0x10];
+
+ u8 reserved_5[0x10];
+ u8 atomic_size_dc[0x10];
+
+ u8 reserved_6[0x720];
+};
+
+struct mlx5_ifc_arm_rq_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_arm_rq_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 srq_number[0x18];
+
+ u8 reserved_3[0x10];
+ u8 lwm[0x10];
+};
+
+struct mlx5_ifc_arm_dct_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_arm_dct_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 dct_number[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_alloc_xrcd_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 xrcd[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_alloc_xrcd_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_alloc_uar_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 uar[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_alloc_uar_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_alloc_q_counter_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x18];
+ u8 counter_set_id[0x8];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_alloc_q_counter_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_alloc_pd_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x8];
+ u8 pd[0x18];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_alloc_pd_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x40];
+};
+
+struct mlx5_ifc_activate_tracer_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_activate_tracer_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 mkey[0x20];
+
+ u8 reserved_2[0x20];
+};
+
+struct mlx5_ifc_access_register_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+
+ u8 register_data[0][0x20];
+};
+
+struct mlx5_ifc_access_register_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x10];
+ u8 register_id[0x10];
+
+ u8 argument[0x20];
+
+ u8 register_data[0][0x20];
+};
+
+struct mlx5_ifc__2rst_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc__2rst_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc__2err_qp_out_bits {
+ u8 status[0x8];
+ u8 reserved_0[0x18];
+
+ u8 syndrome[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc__2err_qp_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 reserved_2[0x8];
+ u8 qpn[0x18];
+
+ u8 reserved_3[0x20];
+};
+
+struct mlx5_ifc_register_power_settings_bits {
+ u8 reserved_0[0x18];
+ u8 power_settings_level[0x8];
+
+ u8 reserved_1[0x60];
+};
+
+struct mlx5_ifc_register_host_endianess_bits {
+ u8 he[0x1];
+ u8 reserved_0[0x1f];
+
+ u8 reserved_1[0x60];
+};
+
+struct mlx5_ifc_pplr_bits {
+ u8 reserved_0[0x8];
+ u8 local_port[0x8];
+ u8 reserved_1[0x10];
+
+ u8 reserved_2[0x1d];
+ u8 exl[0x1];
+ u8 el[0x1];
+ u8 il[0x1];
+};
+
+struct mlx5_ifc_initial_seg_bits {
+ u8 fw_rev_minor[0x10];
+ u8 fw_rev_major[0x10];
+
+ u8 cmd_interface_rev[0x10];
+ u8 fw_rev_subminor[0x10];
+
+ u8 reserved_0[0x40];
+
+ u8 cmdq_phy_addr[4][0x8];
+
+ u8 cmdq_phy_addr_31_12[0x14];
+ u8 reserved_1[0x4];
+ u8 log_cmdq_size[0x4];
+ u8 log_cmdq_stride[0x4];
+
+ u8 command_doorbell_vector[0x20];
+
+ u8 reserved_2[0xf00];
+
+ u8 initializing[0x1];
+ u8 reserved_3[0x1f];
+
+ u8 health_buffer[64][0x8];
+
+ u8 reserved_4[0x6e80];
+
+ u8 health_syndrome[0x8];
+ u8 health_counter[0x18];
+
+ u8 reserved_5[0x7fc0];
+
+ u8 reserved_6[0x1c];
+ u8 clear_intx[0x4];
+
+ u8 reserved_7[0xffe0];
+};
+
+struct mlx5_ifc_wqe_associated_page_fault_event_bits {
+ u8 bytes_commited[0x20];
+
+ u8 reserved_0[0x10];
+ u8 wqe_index[0x10];
+
+ u8 reserved_1[0x10];
+ u8 packet_len[0x10];
+
+ u8 reserved_2[0x60];
+
+ u8 reserved_3[0x5];
+ u8 rdma[0x1];
+ u8 write[0x1];
+ u8 requestor[0x1];
+ u8 qp_number[0x18];
+};
+
+struct mlx5_ifc_ud_adrs_vector_bits {
+ u8 dc_key[8][0x8];
+
+ u8 _1[0x1];
+ u8 reserved_0[0x7];
+ u8 destination_qp_dct[0x18];
+
+ u8 static_rate[0x4];
+ u8 sl_eth_prio[0x4];
+ u8 fl[0x1];
+ u8 mlid[0x7];
+ u8 rlid_udp_sport[0x10];
+
+ u8 reserved_1[0x20];
+
+ u8 rmac[6][0x8];
+
+ u8 tclass[0x8];
+ u8 hop_limit[0x8];
+
+ u8 reserved_2[0x1];
+ u8 grh[0x1];
+ u8 reserved_3[0x2];
+ u8 src_addr_index[0x8];
+ u8 flow_label[0x14];
+
+ u8 rgid_rip[16][0x8];
+};
+
+struct mlx5_ifc_stall_vl_event_bits {
+ u8 reserved_0[0x18];
+ u8 port_num[0x1];
+ u8 reserved_1[0x3];
+ u8 vl[0x4];
+
+ u8 reserved_2[0xa0];
+};
+
+struct mlx5_ifc_res_mini_cqe_rx_hash_layout_bits {
+ u8 rx_hash_result[0x20];
+
+ u8 byte_count[0x20];
+};
+
+struct mlx5_ifc_res_mini_cqe_checksum_layout_bits {
+ u8 checksum[0x20];
+
+ u8 byte_count[0x20];
+};
+
+struct mlx5_ifc_req_mini_cqe_layout_bits {
+ u8 wqe_counter[0x10];
+ u8 s_wqe_opcode[0x8];
+ u8 reserved_0[0x8];
+
+ u8 byte_count[0x20];
+};
+
+struct mlx5_ifc_rdma_page_fault_event_bits {
+ u8 bytes_commited[0x20];
+
+ u8 r_key[0x20];
+
+ u8 reserved_0[0x10];
+ u8 packet_len[0x10];
+
+ u8 rdma_op_len[0x20];
+
+ u8 rdma_va[8][0x8];
+
+ u8 reserved_1[0x5];
+ u8 rdma[0x1];
+ u8 write[0x1];
+ u8 requestor[0x1];
+ u8 qp_number[0x18];
+};
+
+struct mlx5_ifc_qp_events_bits {
+ u8 reserved_0[0xc0];
+
+ u8 reserved_1[0x8];
+ u8 qp_number[0x18];
+};
+
+struct mlx5_ifc_port_state_change_event_bits {
+ u8 reserved_0[0x40];
+
+ u8 port[0x4];
+ u8 reserved_1[0x1c];
+
+ u8 reserved_2[0x80];
+};
+
+struct mlx5_ifc_pages_req_event_bits {
+ u8 reserved_0[0x10];
+ u8 function_id[0x10];
+
+ u8 num_pages[0x20];
+
+ u8 reserved_1[0xa0];
+};
+
+struct mlx5_ifc_gpio_event_bits {
+ u8 reserved_0[0x60];
+
+ u8 gpio_event_hi[0x20];
+
+ u8 gpio_event_lo[0x20];
+
+ u8 reserved_1[0x40];
+};
+
+struct mlx5_ifc_eqe_bits {
+ u8 reserved_0[0x8];
+ u8 event_type[0x8];
+ u8 reserved_1[0x8];
+ u8 event_sub_type[0x8];
+
+ u8 reserved_2[0xe0];
+
+ u8 event_data[28][0x8];
+
+ u8 reserved_3[0x10];
+ u8 signature[0x8];
+ u8 reserved_4[0x7];
+ u8 owner[0x1];
+};
+
+struct mlx5_ifc_dropped_packet_logged_bits {
+ u8 reserved_0[0xe0];
+};
+
+struct mlx5_ifc_dct_events_bits {
+ u8 reserved_0[0xc0];
+
+ u8 reserved_1[0x8];
+ u8 dct_number[0x18];
+};
+
+struct mlx5_ifc_db_bf_congestion_event_bits {
+ u8 event_subtype[0x8];
+ u8 reserved_0[0x8];
+ u8 congestion_level[0x8];
+ u8 reserved_1[0x8];
+
+ u8 reserved_2[0xa0];
+};
+
+struct mlx5_ifc_cq_error_bits {
+ u8 reserved_0[0x8];
+ u8 cqn[0x18];
+
+ u8 reserved_1[0x20];
+
+ u8 reserved_2[0x18];
+ u8 syndrome[0x8];
+
+ u8 reserved_3[0x80];
+};
+
+struct mlx5_ifc_comp_event_bits {
+ u8 reserved_0[0xc0];
+
+ u8 reserved_1[0x8];
+ u8 cq_number[0x18];
+};
+
+struct mlx5_ifc_cmd_queue_entry_bits {
+ u8 type_7[0x8];
+ u8 reserved_0[0x18];
+
+ u8 input_length[0x20];
+
+ u8 input_mailbox_pointer[4][0x8];
+
+ u8 input_mailbox_pointer_31_9[0x17];
+ u8 reserved_1[0x9];
+
+ u8 command_input_inline_data[16][0x8];
+
+ u8 command_output_inline_data[16][0x8];
+
+ u8 output_mailbox_pointer[4][0x8];
+
+ u8 output_mailbox_pointer_31_9[0x17];
+ u8 reserved_2[0x9];
+
+ u8 output_length[0x20];
+
+ u8 token[0x8];
+ u8 signature[0x8];
+ u8 reserved_3[0x8];
+ u8 status[0x7];
+ u8 ownership[0x1];
+};
+
+struct mlx5_ifc_cmd_out_bits {
+ u8 status[0x10];
+ u8 reserved_0[0x10];
+
+ u8 syndrome[0x20];
+
+ u8 command_output[0x20];
+};
+
+struct mlx5_ifc_cmd_inter_comp_event_bits {
+ u8 command_completion_vector[0x20];
+
+ u8 reserved_0[0xc0];
+};
+
+struct mlx5_ifc_cmd_in_bits {
+ u8 opcode[0x10];
+ u8 reserved_0[0x10];
+
+ u8 reserved_1[0x10];
+ u8 op_mod[0x10];
+
+ u8 command[0][0x20];
+};
+
+struct mlx5_ifc_cmd_if_box_bits {
+ u8 mailbox_data[512][0x8];
+
+ u8 reserved_0[0x180];
+
+ u8 next_pointer[4][0x8];
+
+ u8 next_pointer_31_10[0x16];
+ u8 reserved_1[0xa];
+
+ u8 block_number[0x20];
+
+ u8 reserved_2[0x8];
+ u8 token[0x8];
+ u8 ctrl_signature[0x8];
+ u8 signature[0x8];
+};
+
+struct mlx5_ifc_address_context_bits {
+ u8 reserved_0[0x10];
+ u8 mac_addr[6][0x8];
+};
+
+struct mlx5_ifc_mtt_bits {
+ u8 ptag[7][0x8];
+
+ u8 reserved_0[0x6];
+ u8 wr_en[0x1];
+ u8 rd_en[0x1];
+};
+
+union mlx5_ifc_initialization_and_teardown_document_bits {
+ struct mlx5_ifc_initial_seg_bits initial_seg;
+ u8 reserved_0[0x20060];
+};
+
+#endif /* MLX5_IFC_H */
--
2.1.1
^ permalink raw reply related
* [PATCH net-next 3/4] net/mlx5_core: use set/get macros in device caps
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412082135-30229-1-git-send-email-eli@mellanox.com>
Transform device capabilities related commands to use set/get macros to
manipulate command mailboxes.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/cmd.c | 17 +++
drivers/net/ethernet/mellanox/mlx5/core/main.c | 150 ++++++++++++-------------
include/linux/mlx5/device.h | 92 ---------------
include/linux/mlx5/driver.h | 1 +
4 files changed, 92 insertions(+), 168 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
index 3ecef1310bae..368c6c5ea014 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/cmd.c
@@ -1537,3 +1537,20 @@ int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr)
return cmd_status_to_err(hdr->status);
}
+
+int mlx5_cmd_status_to_err_v2(void *ptr)
+{
+ u32 syndrome;
+ u8 status;
+
+ status = be32_to_cpu(*(__be32 *)ptr) >> 24;
+ if (!status)
+ return 0;
+
+ syndrome = be32_to_cpu(*(__be32 *)(ptr + 4));
+
+ pr_warn("command failed, status %s(0x%x), syndrome 0x%x\n",
+ cmd_status_str(status), status, syndrome);
+
+ return cmd_status_to_err(status);
+}
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/main.c b/drivers/net/ethernet/mellanox/mlx5/core/main.c
index d9f74618befa..b9e3259e415f 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/main.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/main.c
@@ -43,6 +43,7 @@
#include <linux/mlx5/qp.h>
#include <linux/mlx5/srq.h>
#include <linux/debugfs.h>
+#include <linux/mlx5/mlx5_ifc.h>
#include "mlx5_core.h"
#define DRIVER_NAME "mlx5_core"
@@ -277,18 +278,20 @@ static u16 to_fw_pkey_sz(u32 size)
/* selectively copy writable fields clearing any reserved area
*/
-static void copy_rw_fields(struct mlx5_hca_cap *to, struct mlx5_general_caps *from)
+static void copy_rw_fields(void *to, struct mlx5_caps *from)
{
+ __be64 *flags_off = (__be64 *)MLX5_ADDR_OF(cmd_hca_cap, to, reserved_22);
u64 v64;
- to->log_max_qp = from->log_max_qp & 0x1f;
- to->log_max_ra_req_dc = from->log_max_ra_req_dc & 0x3f;
- to->log_max_ra_res_dc = from->log_max_ra_res_dc & 0x3f;
- to->log_max_ra_req_qp = from->log_max_ra_req_qp & 0x3f;
- to->log_max_ra_res_qp = from->log_max_ra_res_qp & 0x3f;
- to->pkey_table_size = cpu_to_be16(to_fw_pkey_sz(from->pkey_table_size));
- v64 = from->flags & MLX5_CAP_BITS_RW_MASK;
- to->flags = cpu_to_be64(v64);
+ MLX5_SET(cmd_hca_cap, to, log_max_qp, from->gen.log_max_qp);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_req_qp, from->gen.log_max_ra_req_qp);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_res_qp, from->gen.log_max_ra_res_qp);
+ MLX5_SET(cmd_hca_cap, to, pkey_table_size, from->gen.pkey_table_size);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_req_dc, from->gen.log_max_ra_req_dc);
+ MLX5_SET(cmd_hca_cap, to, log_max_ra_res_dc, from->gen.log_max_ra_res_dc);
+ MLX5_SET(cmd_hca_cap, to, pkey_table_size, to_fw_pkey_sz(from->gen.pkey_table_size));
+ v64 = from->gen.flags & MLX5_CAP_BITS_RW_MASK;
+ *flags_off = cpu_to_be64(v64);
}
static u16 get_pkey_table_size(int pkey)
@@ -299,55 +302,47 @@ static u16 get_pkey_table_size(int pkey)
return MLX5_MIN_PKEY_TABLE_SIZE << pkey;
}
-static void fw2drv_caps(struct mlx5_caps *caps,
- struct mlx5_cmd_query_hca_cap_mbox_out *out)
+static void fw2drv_caps(struct mlx5_caps *caps, void *out)
{
struct mlx5_general_caps *gen = &caps->gen;
- u16 t16;
-
- gen->max_srq_wqes = 1 << out->hca_cap.log_max_srq_sz;
- gen->max_wqes = 1 << out->hca_cap.log_max_qp_sz;
- gen->log_max_qp = out->hca_cap.log_max_qp & 0x1f;
- gen->log_max_strq = out->hca_cap.log_max_strq_sz;
- gen->log_max_srq = out->hca_cap.log_max_srqs & 0x1f;
- gen->max_cqes = 1 << out->hca_cap.log_max_cq_sz;
- gen->log_max_cq = out->hca_cap.log_max_cq & 0x1f;
- gen->max_eqes = out->hca_cap.log_max_eq_sz;
- gen->log_max_mkey = out->hca_cap.log_max_mkey & 0x3f;
- gen->log_max_eq = out->hca_cap.log_max_eq & 0xf;
- gen->max_indirection = out->hca_cap.max_indirection;
- gen->log_max_mrw_sz = out->hca_cap.log_max_mrw_sz;
- gen->log_max_bsf_list_size = 0;
- gen->log_max_klm_list_size = 0;
- gen->log_max_ra_req_dc = out->hca_cap.log_max_ra_req_dc;
- gen->log_max_ra_res_dc = out->hca_cap.log_max_ra_res_dc;
- gen->log_max_ra_req_qp = out->hca_cap.log_max_ra_req_qp;
- gen->log_max_ra_res_qp = out->hca_cap.log_max_ra_res_qp;
- gen->max_qp_counters = be16_to_cpu(out->hca_cap.max_qp_count);
- gen->pkey_table_size = get_pkey_table_size(be16_to_cpu(out->hca_cap.pkey_table_size));
- gen->local_ca_ack_delay = out->hca_cap.local_ca_ack_delay & 0x1f;
- gen->num_ports = out->hca_cap.num_ports & 0xf;
- gen->log_max_msg = out->hca_cap.log_max_msg & 0x1f;
- gen->stat_rate_support = be16_to_cpu(out->hca_cap.stat_rate_support);
- gen->flags = be64_to_cpu(out->hca_cap.flags);
- pr_debug("flags = 0x%llx\n", gen->flags);
- gen->uar_sz = out->hca_cap.uar_sz;
- gen->min_log_pg_sz = out->hca_cap.log_pg_sz;
- t16 = be16_to_cpu(out->hca_cap.bf_log_bf_reg_size);
- if (t16 & 0x8000) {
- gen->bf_reg_size = 1 << (t16 & 0x1f);
- gen->bf_regs_per_page = MLX5_BF_REGS_PER_PAGE;
- } else {
- gen->bf_reg_size = 0;
- gen->bf_regs_per_page = 0;
- }
- gen->max_sq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_sq);
- gen->max_rq_desc_sz = be16_to_cpu(out->hca_cap.max_desc_sz_rq);
- gen->max_qp_mcg = be32_to_cpu(out->hca_cap.max_qp_mcg) & 0xffffff;
- gen->log_max_pd = out->hca_cap.log_max_pd & 0x1f;
- gen->log_max_xrcd = out->hca_cap.log_max_xrcd;
- gen->log_uar_page_sz = be16_to_cpu(out->hca_cap.log_uar_page_sz);
+ gen->max_srq_wqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_srq_sz);
+ gen->max_wqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_qp_sz);
+ gen->log_max_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_qp);
+ gen->log_max_strq = MLX5_GET_PR(cmd_hca_cap, out, log_max_strq_sz);
+ gen->log_max_srq = MLX5_GET_PR(cmd_hca_cap, out, log_max_srqs);
+ gen->max_cqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_cq_sz);
+ gen->log_max_cq = MLX5_GET_PR(cmd_hca_cap, out, log_max_cq);
+ gen->max_eqes = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_max_eq_sz);
+ gen->log_max_mkey = MLX5_GET_PR(cmd_hca_cap, out, log_max_mkey);
+ gen->log_max_eq = MLX5_GET_PR(cmd_hca_cap, out, log_max_eq);
+ gen->max_indirection = MLX5_GET_PR(cmd_hca_cap, out, max_indirection);
+ gen->log_max_mrw_sz = MLX5_GET_PR(cmd_hca_cap, out, log_max_mrw_sz);
+ gen->log_max_bsf_list_size = MLX5_GET_PR(cmd_hca_cap, out, log_max_bsf_list_size);
+ gen->log_max_klm_list_size = MLX5_GET_PR(cmd_hca_cap, out, log_max_klm_list_size);
+ gen->log_max_ra_req_dc = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_req_dc);
+ gen->log_max_ra_res_dc = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_res_dc);
+ gen->log_max_ra_req_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_req_qp);
+ gen->log_max_ra_res_qp = MLX5_GET_PR(cmd_hca_cap, out, log_max_ra_res_qp);
+ gen->max_qp_counters = MLX5_GET_PR(cmd_hca_cap, out, max_qp_cnt);
+ gen->pkey_table_size = get_pkey_table_size(MLX5_GET_PR(cmd_hca_cap, out, pkey_table_size));
+ gen->local_ca_ack_delay = MLX5_GET_PR(cmd_hca_cap, out, local_ca_ack_delay);
+ gen->num_ports = MLX5_GET_PR(cmd_hca_cap, out, num_ports);
+ gen->log_max_msg = MLX5_GET_PR(cmd_hca_cap, out, log_max_msg);
+ gen->stat_rate_support = MLX5_GET_PR(cmd_hca_cap, out, stat_rate_support);
+ gen->flags = be64_to_cpu(*(__be64 *)MLX5_ADDR_OF(cmd_hca_cap, out, reserved_22));
+ pr_debug("flags = 0x%llx\n", gen->flags);
+ gen->uar_sz = MLX5_GET_PR(cmd_hca_cap, out, uar_sz);
+ gen->min_log_pg_sz = MLX5_GET_PR(cmd_hca_cap, out, log_pg_sz);
+ gen->bf_reg_size = MLX5_GET_PR(cmd_hca_cap, out, bf);
+ gen->bf_reg_size = 1 << MLX5_GET_PR(cmd_hca_cap, out, log_bf_reg_size);
+ gen->max_sq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_sq);
+ gen->max_rq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_rq);
+ gen->max_dc_sq_desc_sz = MLX5_GET_PR(cmd_hca_cap, out, max_wqe_sz_sq_dc);
+ gen->max_qp_mcg = MLX5_GET_PR(cmd_hca_cap, out, max_qp_mcg);
+ gen->log_max_pd = MLX5_GET_PR(cmd_hca_cap, out, log_max_pd);
+ gen->log_max_xrcd = MLX5_GET_PR(cmd_hca_cap, out, log_max_xrcd);
+ gen->log_uar_page_sz = MLX5_GET_PR(cmd_hca_cap, out, log_uar_page_sz);
}
static const char *caps_opmod_str(u16 opmod)
@@ -365,59 +360,61 @@ static const char *caps_opmod_str(u16 opmod)
int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
u16 opmod)
{
- struct mlx5_cmd_query_hca_cap_mbox_out *out;
- struct mlx5_cmd_query_hca_cap_mbox_in in;
+ u8 in[MLX5_ST_SZ_BYTES(query_hca_cap_in)];
+ int out_sz = MLX5_ST_SZ_BYTES(query_hca_cap_out);
+ void *out;
int err;
- memset(&in, 0, sizeof(in));
- out = kzalloc(sizeof(*out), GFP_KERNEL);
+ memset(in, 0, sizeof(in));
+ out = kzalloc(out_sz, GFP_KERNEL);
if (!out)
return -ENOMEM;
+ MLX5_SET(query_hca_cap_in, in, opcode, MLX5_CMD_OP_QUERY_HCA_CAP);
+ MLX5_SET(query_hca_cap_in, in, op_mod, opmod);
+ err = mlx5_cmd_exec(dev, in, sizeof(in), out, out_sz);
+ if (err)
+ goto query_ex;
- in.hdr.opcode = cpu_to_be16(MLX5_CMD_OP_QUERY_HCA_CAP);
- in.hdr.opmod = cpu_to_be16(opmod);
- err = mlx5_cmd_exec(dev, &in, sizeof(in), out, sizeof(*out));
-
- err = mlx5_cmd_status_to_err(&out->hdr);
+ err = mlx5_cmd_status_to_err_v2(out);
if (err) {
mlx5_core_warn(dev, "query max hca cap failed, %d\n", err);
goto query_ex;
}
mlx5_core_dbg(dev, "%s\n", caps_opmod_str(opmod));
- fw2drv_caps(caps, out);
+ fw2drv_caps(caps, MLX5_ADDR_OF(query_hca_cap_out, out, capability_struct));
query_ex:
kfree(out);
return err;
}
-static int set_caps(struct mlx5_core_dev *dev,
- struct mlx5_cmd_set_hca_cap_mbox_in *in)
+static int set_caps(struct mlx5_core_dev *dev, void *in, int in_sz)
{
- struct mlx5_cmd_set_hca_cap_mbox_out out;
+ u32 out[MLX5_ST_SZ_DW(set_hca_cap_out)];
int err;
- memset(&out, 0, sizeof(out));
+ memset(out, 0, sizeof(out));
- in->hdr.opcode = cpu_to_be16(MLX5_CMD_OP_SET_HCA_CAP);
- err = mlx5_cmd_exec(dev, in, sizeof(*in), &out, sizeof(out));
+ MLX5_SET(set_hca_cap_in, in, opcode, MLX5_CMD_OP_SET_HCA_CAP);
+ err = mlx5_cmd_exec(dev, in, in_sz, out, sizeof(out));
if (err)
return err;
- err = mlx5_cmd_status_to_err(&out.hdr);
+ err = mlx5_cmd_status_to_err_v2(out);
return err;
}
static int handle_hca_cap(struct mlx5_core_dev *dev)
{
- struct mlx5_cmd_set_hca_cap_mbox_in *set_ctx = NULL;
+ void *set_ctx = NULL;
struct mlx5_profile *prof = dev->profile;
struct mlx5_caps *cur_caps = NULL;
struct mlx5_caps *max_caps = NULL;
int err = -ENOMEM;
+ int set_sz = MLX5_ST_SZ_BYTES(set_hca_cap_in);
- set_ctx = kzalloc(sizeof(*set_ctx), GFP_KERNEL);
+ set_ctx = kzalloc(set_sz, GFP_KERNEL);
if (!set_ctx)
goto query_ex;
@@ -446,8 +443,9 @@ static int handle_hca_cap(struct mlx5_core_dev *dev)
/* disable checksum */
cur_caps->gen.flags &= ~MLX5_DEV_CAP_FLAG_CMDIF_CSUM;
- copy_rw_fields(&set_ctx->hca_cap, &cur_caps->gen);
- err = set_caps(dev, set_ctx);
+ copy_rw_fields(MLX5_ADDR_OF(set_hca_cap_in, set_ctx, hca_capability_struct),
+ cur_caps);
+ err = set_caps(dev, set_ctx, set_sz);
query_ex:
kfree(cur_caps);
diff --git a/include/linux/mlx5/device.h b/include/linux/mlx5/device.h
index 3db2d4d82ad3..08a963c894dc 100644
--- a/include/linux/mlx5/device.h
+++ b/include/linux/mlx5/device.h
@@ -357,98 +357,6 @@ struct mlx5_cmd_query_adapter_mbox_out {
u8 vsd_psid[16];
};
-struct mlx5_hca_cap {
- u8 rsvd1[16];
- u8 log_max_srq_sz;
- u8 log_max_qp_sz;
- u8 rsvd2;
- u8 log_max_qp;
- u8 log_max_strq_sz;
- u8 log_max_srqs;
- u8 rsvd4[2];
- u8 rsvd5;
- u8 log_max_cq_sz;
- u8 rsvd6;
- u8 log_max_cq;
- u8 log_max_eq_sz;
- u8 log_max_mkey;
- u8 rsvd7;
- u8 log_max_eq;
- u8 max_indirection;
- u8 log_max_mrw_sz;
- u8 log_max_bsf_list_sz;
- u8 log_max_klm_list_sz;
- u8 rsvd_8_0;
- u8 log_max_ra_req_dc;
- u8 rsvd_8_1;
- u8 log_max_ra_res_dc;
- u8 rsvd9;
- u8 log_max_ra_req_qp;
- u8 rsvd10;
- u8 log_max_ra_res_qp;
- u8 pad_cap;
- u8 rsvd11[3];
- __be16 max_qp_count;
- __be16 pkey_table_size;
- u8 rsvd13;
- u8 local_ca_ack_delay;
- u8 rsvd14;
- u8 num_ports;
- u8 log_max_msg;
- u8 rsvd15[3];
- __be16 stat_rate_support;
- u8 rsvd16[2];
- __be64 flags;
- u8 rsvd17;
- u8 uar_sz;
- u8 rsvd18;
- u8 log_pg_sz;
- __be16 bf_log_bf_reg_size;
- u8 rsvd19[4];
- __be16 max_desc_sz_sq;
- u8 rsvd20[2];
- __be16 max_desc_sz_rq;
- u8 rsvd21[2];
- __be16 max_desc_sz_sq_dc;
- __be32 max_qp_mcg;
- u8 rsvd22[3];
- u8 log_max_mcg;
- u8 rsvd23;
- u8 log_max_pd;
- u8 rsvd24;
- u8 log_max_xrcd;
- u8 rsvd25[42];
- __be16 log_uar_page_sz;
- u8 rsvd26[108];
-};
-
-
-struct mlx5_cmd_query_hca_cap_mbox_in {
- struct mlx5_inbox_hdr hdr;
- u8 rsvd[8];
-};
-
-
-struct mlx5_cmd_query_hca_cap_mbox_out {
- struct mlx5_outbox_hdr hdr;
- u8 rsvd0[8];
- struct mlx5_hca_cap hca_cap;
-};
-
-
-struct mlx5_cmd_set_hca_cap_mbox_in {
- struct mlx5_inbox_hdr hdr;
- u8 rsvd[8];
- struct mlx5_hca_cap hca_cap;
-};
-
-
-struct mlx5_cmd_set_hca_cap_mbox_out {
- struct mlx5_outbox_hdr hdr;
- u8 rsvd0[8];
-};
-
-
struct mlx5_cmd_init_hca_mbox_in {
struct mlx5_inbox_hdr hdr;
u8 rsvd0[2];
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index 6f48dc793b9f..c439f9c59b93 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -641,6 +641,7 @@ void mlx5_cmd_cleanup(struct mlx5_core_dev *dev);
void mlx5_cmd_use_events(struct mlx5_core_dev *dev);
void mlx5_cmd_use_polling(struct mlx5_core_dev *dev);
int mlx5_cmd_status_to_err(struct mlx5_outbox_hdr *hdr);
+int mlx5_cmd_status_to_err_v2(void *ptr);
int mlx5_core_get_caps(struct mlx5_core_dev *dev, struct mlx5_caps *caps,
u16 opmod);
int mlx5_cmd_exec(struct mlx5_core_dev *dev, void *in, int in_size, void *out,
--
2.1.1
^ permalink raw reply related
* [PATCH net-next 4/4] net/mlx5_core: Identify resources by their type
From: Eli Cohen @ 2014-09-30 13:02 UTC (permalink / raw)
To: davem; +Cc: netdev, ogerlitz, yevgenyp, Eli Cohen
In-Reply-To: <1412082135-30229-1-git-send-email-eli@mellanox.com>
This patch puts a common part as the first field of mlx5_core_qp. This field is
used to identify which resource generated an event. This is required since upcoming
new resource types such as DC targets are allocated for the same numerical space
as regular QPs and may generate the same events. By searching the resource in the
same table we can then look at the common field to identify the resource.
Signed-off-by: Eli Cohen <eli@mellanox.com>
---
drivers/net/ethernet/mellanox/mlx5/core/eq.c | 12 +++---
drivers/net/ethernet/mellanox/mlx5/core/qp.c | 57 ++++++++++++++++++++--------
include/linux/mlx5/driver.h | 13 ++++++-
include/linux/mlx5/qp.h | 3 +-
4 files changed, 60 insertions(+), 25 deletions(-)
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/eq.c b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
index 11b9b840ad4d..ed53291468f3 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/eq.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/eq.c
@@ -198,7 +198,7 @@ static int mlx5_eq_int(struct mlx5_core_dev *dev, struct mlx5_eq *eq)
int eqes_found = 0;
int set_ci = 0;
u32 cqn;
- u32 srqn;
+ u32 rsn;
u8 port;
while ((eqe = next_eqe_sw(eq))) {
@@ -224,18 +224,18 @@ static int mlx5_eq_int(struct mlx5_core_dev *dev, struct mlx5_eq *eq)
case MLX5_EVENT_TYPE_PATH_MIG_FAILED:
case MLX5_EVENT_TYPE_WQ_INVAL_REQ_ERROR:
case MLX5_EVENT_TYPE_WQ_ACCESS_ERROR:
+ rsn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
mlx5_core_dbg(dev, "event %s(%d) arrived\n",
eqe_type_str(eqe->type), eqe->type);
- mlx5_qp_event(dev, be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff,
- eqe->type);
+ mlx5_rsc_event(dev, rsn, eqe->type);
break;
case MLX5_EVENT_TYPE_SRQ_RQ_LIMIT:
case MLX5_EVENT_TYPE_SRQ_CATAS_ERROR:
- srqn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
+ rsn = be32_to_cpu(eqe->data.qp_srq.qp_srq_n) & 0xffffff;
mlx5_core_dbg(dev, "SRQ event %s(%d): srqn 0x%x\n",
- eqe_type_str(eqe->type), eqe->type, srqn);
- mlx5_srq_event(dev, srqn, eqe->type);
+ eqe_type_str(eqe->type), eqe->type, rsn);
+ mlx5_srq_event(dev, rsn, eqe->type);
break;
case MLX5_EVENT_TYPE_CMD:
diff --git a/drivers/net/ethernet/mellanox/mlx5/core/qp.c b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
index 415b67ce379e..5261a2b0da43 100644
--- a/drivers/net/ethernet/mellanox/mlx5/core/qp.c
+++ b/drivers/net/ethernet/mellanox/mlx5/core/qp.c
@@ -39,28 +39,53 @@
#include "mlx5_core.h"
-void mlx5_qp_event(struct mlx5_core_dev *dev, u32 qpn, int event_type)
+static struct mlx5_core_rsc_common *mlx5_get_rsc(struct mlx5_core_dev *dev,
+ u32 rsn)
{
struct mlx5_qp_table *table = &dev->priv.qp_table;
- struct mlx5_core_qp *qp;
+ struct mlx5_core_rsc_common *common;
spin_lock(&table->lock);
- qp = radix_tree_lookup(&table->tree, qpn);
- if (qp)
- atomic_inc(&qp->refcount);
+ common = radix_tree_lookup(&table->tree, rsn);
+ if (common)
+ atomic_inc(&common->refcount);
spin_unlock(&table->lock);
- if (!qp) {
- mlx5_core_warn(dev, "Async event for bogus QP 0x%x\n", qpn);
- return;
+ if (!common) {
+ mlx5_core_warn(dev, "Async event for bogus resource 0x%x\n",
+ rsn);
+ return NULL;
}
+ return common;
+}
- qp->event(qp, event_type);
+void mlx5_core_put_rsc(struct mlx5_core_rsc_common *common)
+{
+ if (atomic_dec_and_test(&common->refcount))
+ complete(&common->free);
+}
+
+void mlx5_rsc_event(struct mlx5_core_dev *dev, u32 rsn, int event_type)
+{
+ struct mlx5_core_rsc_common *common = mlx5_get_rsc(dev, rsn);
+ struct mlx5_core_qp *qp;
+
+ if (!common)
+ return;
+
+ switch (common->res) {
+ case MLX5_RES_QP:
+ qp = (struct mlx5_core_qp *)common;
+ qp->event(qp, event_type);
+ break;
+
+ default:
+ mlx5_core_warn(dev, "invalid resource type for 0x%x\n", rsn);
+ }
- if (atomic_dec_and_test(&qp->refcount))
- complete(&qp->free);
+ mlx5_core_put_rsc(common);
}
int mlx5_core_create_qp(struct mlx5_core_dev *dev,
@@ -92,6 +117,7 @@ int mlx5_core_create_qp(struct mlx5_core_dev *dev,
qp->qpn = be32_to_cpu(out.qpn) & 0xffffff;
mlx5_core_dbg(dev, "qpn = 0x%x\n", qp->qpn);
+ qp->common.res = MLX5_RES_QP;
spin_lock_irq(&table->lock);
err = radix_tree_insert(&table->tree, qp->qpn, qp);
spin_unlock_irq(&table->lock);
@@ -106,9 +132,9 @@ int mlx5_core_create_qp(struct mlx5_core_dev *dev,
qp->qpn);
qp->pid = current->pid;
- atomic_set(&qp->refcount, 1);
+ atomic_set(&qp->common.refcount, 1);
atomic_inc(&dev->num_qps);
- init_completion(&qp->free);
+ init_completion(&qp->common.free);
return 0;
@@ -138,9 +164,8 @@ int mlx5_core_destroy_qp(struct mlx5_core_dev *dev,
radix_tree_delete(&table->tree, qp->qpn);
spin_unlock_irqrestore(&table->lock, flags);
- if (atomic_dec_and_test(&qp->refcount))
- complete(&qp->free);
- wait_for_completion(&qp->free);
+ mlx5_core_put_rsc((struct mlx5_core_rsc_common *)qp);
+ wait_for_completion(&qp->common.free);
memset(&in, 0, sizeof(in));
memset(&out, 0, sizeof(out));
diff --git a/include/linux/mlx5/driver.h b/include/linux/mlx5/driver.h
index c439f9c59b93..246310dc8bef 100644
--- a/include/linux/mlx5/driver.h
+++ b/include/linux/mlx5/driver.h
@@ -375,6 +375,16 @@ struct mlx5_core_mr {
u32 pd;
};
+enum mlx5_res_type {
+ MLX5_RES_QP,
+};
+
+struct mlx5_core_rsc_common {
+ enum mlx5_res_type res;
+ atomic_t refcount;
+ struct completion free;
+};
+
struct mlx5_core_srq {
u32 srqn;
int max;
@@ -700,7 +710,7 @@ int mlx5_eq_init(struct mlx5_core_dev *dev);
void mlx5_eq_cleanup(struct mlx5_core_dev *dev);
void mlx5_fill_page_array(struct mlx5_buf *buf, __be64 *pas);
void mlx5_cq_completion(struct mlx5_core_dev *dev, u32 cqn);
-void mlx5_qp_event(struct mlx5_core_dev *dev, u32 qpn, int event_type);
+void mlx5_rsc_event(struct mlx5_core_dev *dev, u32 rsn, int event_type);
void mlx5_srq_event(struct mlx5_core_dev *dev, u32 srqn, int event_type);
struct mlx5_core_srq *mlx5_core_get_srq(struct mlx5_core_dev *dev, u32 srqn);
void mlx5_cmd_comp_handler(struct mlx5_core_dev *dev, unsigned long vector);
@@ -737,6 +747,7 @@ void mlx5_cmdif_debugfs_cleanup(struct mlx5_core_dev *dev);
int mlx5_core_create_psv(struct mlx5_core_dev *dev, u32 pdn,
int npsvs, u32 *sig_index);
int mlx5_core_destroy_psv(struct mlx5_core_dev *dev, int psv_num);
+void mlx5_core_put_rsc(struct mlx5_core_rsc_common *common);
static inline u32 mlx5_mkey_to_idx(u32 mkey)
{
diff --git a/include/linux/mlx5/qp.h b/include/linux/mlx5/qp.h
index 9709b30e2d69..7c4c0f1f5805 100644
--- a/include/linux/mlx5/qp.h
+++ b/include/linux/mlx5/qp.h
@@ -342,10 +342,9 @@ struct mlx5_stride_block_ctrl_seg {
};
struct mlx5_core_qp {
+ struct mlx5_core_rsc_common common; /* must be first */
void (*event) (struct mlx5_core_qp *, int);
int qpn;
- atomic_t refcount;
- struct completion free;
struct mlx5_rsc_debug *dbg;
int pid;
};
--
2.1.1
^ permalink raw reply related
* Re: [PATCH 2/7] netfilter: Convert print_tuple functions to return void
From: Joe Perches @ 2014-09-30 13:04 UTC (permalink / raw)
To: Petr Mladek
Cc: Steven Rostedt, Al Viro, Andrew Morton, Linus Torvalds,
Pablo Neira Ayuso, Patrick McHardy, Jozsef Kadlecsik,
David S. Miller, netfilter-devel, coreteam, netdev, linux-kernel
In-Reply-To: <20140930102221.GB2707@dhcp128.suse.cz>
On Tue, 2014-09-30 at 12:22 +0200, Petr Mladek wrote:
> On Mon 29-09-14 16:08:22, Joe Perches wrote:
> > Since adding a new function to seq_file (seq_is_full)
> > there isn't any value for functions called from seq_show to
> > return anything. Remove the int returns of the various
> > print_tuple/<foo>_print_tuple functions.
[a bunch of quoted stuff]
Please remember to cut out from your replies the unnecessary old stuff.
It can take quite awhile to scan through looking for your comments.
> > diff --git a/net/netfilter/nf_conntrack_standalone.c b/net/netfilter/nf_conntrack_standalone.c
[]
> > @@ -202,9 +203,8 @@ static int ct_seq_show(struct seq_file *s, void *v)
> > if (l4proto->print_conntrack && l4proto->print_conntrack(s, ct))
> > goto release;
> >
> > - if (print_tuple(s, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple,
> > - l3proto, l4proto))
> > - goto release;
> > + print_tuple(s, &ct->tuplehash[IP_CT_DIR_ORIGINAL].tuple,
> > + l3proto, l4proto);
>
> To be precise, we should add:
>
> if (seq_overflow(s))
> goto release;
Precision isn't all that useful when checking seq_<output>.
There really isn't much value in checking each possible
overflow site. A periodic check prior to or in the middle
of a more costly/longish operation should be acceptable.
The entire block that precedes any seq buffer full test will
be redone when the buffer is expanded.
^ permalink raw reply
* RE: [PATCH v5 2/3] net: Add Keystone NetCP ethernet driver
From: David Laight @ 2014-09-30 13:09 UTC (permalink / raw)
To: 'David Miller', santosh.shilimkar@ti.com
Cc: netdev@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, robh+dt@kernel.org,
grant.likely@linaro.org, devicetree@vger.kernel.org,
sandeep_n@ti.com, joe@perches.com
In-Reply-To: <20140929.155200.2231132602322345812.davem@davemloft.net>
From: David Miller
> From: Santosh Shilimkar <santosh.shilimkar@ti.com>
> Date: Thu, 25 Sep 2014 13:48:36 -0400
>
> > +static inline int gbe_phy_link_status(struct gbe_slave *slave)
> > +{
> > + if (!slave->phy)
> > + return 1;
> > +
> > + if (slave->phy->link)
> > + return 1;
> > +
> > + return 0;
> > +}
>
> Please use 'bool' as the return type and return 'true' or 'false'.
That function body could also be just:
return !slave->phy && slave->phy->link;
which might be more readable if directly coded.
I also wonder if slave>phy can actually be NULL?
If it can be under unusual circumstances it might be worth
assigning the address of a dummy 'phy' structure so that the
tests can all be removed.
David
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Wolfram Sang @ 2014-09-30 13:26 UTC (permalink / raw)
To: Roger Quadros
Cc: wg, mkl, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar,
nm, sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <1410273070-22485-3-git-send-email-rogerq@ti.com>
[-- Attachment #1: Type: text/plain, Size: 1868 bytes --]
On Tue, Sep 09, 2014 at 05:31:09PM +0300, Roger Quadros wrote:
> Some TI SoCs like DRA7 have a RAMINIT register specification
> different from the other AMxx SoCs and as expected by the
> existing driver.
>
> To add more insanity, this register is shared with other
> IPs like DSS, PCIe and PWM.
>
> Provides a more generic mechanism to specify the RAMINIT
> register location and START/DONE bit position and use the
> syscon/regmap framework to access the register.
>
> Signed-off-by: Roger Quadros <rogerq@ti.com>
> ---
> .../devicetree/bindings/net/can/c_can.txt | 7 ++
> drivers/net/can/c_can/c_can.h | 11 ++-
> drivers/net/can/c_can/c_can_platform.c | 109 +++++++++++++++------
> 3 files changed, 95 insertions(+), 32 deletions(-)
>
> diff --git a/Documentation/devicetree/bindings/net/can/c_can.txt b/Documentation/devicetree/bindings/net/can/c_can.txt
> index 8f1ae81..e12d1a1 100644
> --- a/Documentation/devicetree/bindings/net/can/c_can.txt
> +++ b/Documentation/devicetree/bindings/net/can/c_can.txt
> @@ -13,6 +13,13 @@ Optional properties:
> - ti,hwmods : Must be "d_can<n>" or "c_can<n>", n being the
> instance number
>
> +- ti,raminit-syscon : Handle to system control region that contains the
> + RAMINIT register. If specified, the second memory resource
> + in the reg property must index into the RAMINIT
> + register within the syscon region
There seems to be a simple "syscon" property these days.
> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
This should not be encoded in DT! This is not describing hardware setup.
The driver should know where the bits are for the syscon phandle,
depending on which SoC it runs...
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH v5 2/3] net: Add Keystone NetCP ethernet driver
From: Geert Uytterhoeven @ 2014-09-30 13:28 UTC (permalink / raw)
To: David Laight
Cc: David Miller, santosh.shilimkar@ti.com, netdev@vger.kernel.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org, robh+dt@kernel.org,
grant.likely@linaro.org, devicetree@vger.kernel.org,
sandeep_n@ti.com, joe@perches.com
In-Reply-To: <063D6719AE5E284EB5DD2968C1650D6D174A1012@AcuExch.aculab.com>
On Tue, Sep 30, 2014 at 3:09 PM, David Laight <David.Laight@aculab.com> wrote:
>> > +static inline int gbe_phy_link_status(struct gbe_slave *slave)
>> > +{
>> > + if (!slave->phy)
>> > + return 1;
>> > +
>> > + if (slave->phy->link)
>> > + return 1;
>> > +
>> > + return 0;
>> > +}
>>
>> Please use 'bool' as the return type and return 'true' or 'false'.
>
> That function body could also be just:
> return !slave->phy && slave->phy->link;
> which might be more readable if directly coded.
return !slave->phy || slave->phy->link;
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Roger Quadros @ 2014-09-30 13:33 UTC (permalink / raw)
To: Wolfram Sang
Cc: wg, mkl, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar,
nm, sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <20140930132650.GN1325@katana>
On 09/30/2014 04:26 PM, Wolfram Sang wrote:
> On Tue, Sep 09, 2014 at 05:31:09PM +0300, Roger Quadros wrote:
>> Some TI SoCs like DRA7 have a RAMINIT register specification
>> different from the other AMxx SoCs and as expected by the
>> existing driver.
>>
>> To add more insanity, this register is shared with other
>> IPs like DSS, PCIe and PWM.
>>
>> Provides a more generic mechanism to specify the RAMINIT
>> register location and START/DONE bit position and use the
>> syscon/regmap framework to access the register.
>>
>> Signed-off-by: Roger Quadros <rogerq@ti.com>
>> ---
>> .../devicetree/bindings/net/can/c_can.txt | 7 ++
>> drivers/net/can/c_can/c_can.h | 11 ++-
>> drivers/net/can/c_can/c_can_platform.c | 109 +++++++++++++++------
>> 3 files changed, 95 insertions(+), 32 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/net/can/c_can.txt b/Documentation/devicetree/bindings/net/can/c_can.txt
>> index 8f1ae81..e12d1a1 100644
>> --- a/Documentation/devicetree/bindings/net/can/c_can.txt
>> +++ b/Documentation/devicetree/bindings/net/can/c_can.txt
>> @@ -13,6 +13,13 @@ Optional properties:
>> - ti,hwmods : Must be "d_can<n>" or "c_can<n>", n being the
>> instance number
>>
>> +- ti,raminit-syscon : Handle to system control region that contains the
>> + RAMINIT register. If specified, the second memory resource
>> + in the reg property must index into the RAMINIT
>> + register within the syscon region
>
> There seems to be a simple "syscon" property these days.
I had used plain "syscon" in the earlier revisions but was asked to make it a TI specific
property since only TI uses this mechanism.
>
>> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
>> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
>
> This should not be encoded in DT! This is not describing hardware setup.
> The driver should know where the bits are for the syscon phandle,
> depending on which SoC it runs...
>
OK. I'll think of matching the compatible ID with SOC specific data in the driver.
cheers,
-roger
^ permalink raw reply
* Re: [net-next PATCH] dql: add a burst attribute
From: Florian Westphal @ 2014-09-30 13:46 UTC (permalink / raw)
To: Eric Dumazet
Cc: Jesper Dangaard Brouer, netdev, David S. Miller, Tom Herbert,
Hannes Frederic Sowa, Florian Westphal, Daniel Borkmann,
Jamal Hadi Salim, Alexander Duyck, John Fastabend, Dave Taht,
toke
In-Reply-To: <1412081485.30721.75.camel@edumazet-glaptop2.roam.corp.google.com>
Eric Dumazet <eric.dumazet@gmail.com> wrote:
> From: Eric Dumazet <edumazet@google.com>
>
> Add a new dql attribute, to control how much packets we are allowed to
> burst from qdisc to device.
I understand the motivation for this, but I find it a bit out-of-place
to have a 'packet' type counter in bql context?
Would it perhaps make more sense to restrict bulk dequeues by an upper
(possibly changeable from userspace) byte counter limit?
Thanks,
Florian
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Marc Kleine-Budde @ 2014-09-30 13:45 UTC (permalink / raw)
To: Wolfram Sang, Roger Quadros
Cc: wg, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar, nm,
sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <20140930132650.GN1325@katana>
[-- Attachment #1: Type: text/plain, Size: 2390 bytes --]
On 09/30/2014 03:26 PM, Wolfram Sang wrote:
> On Tue, Sep 09, 2014 at 05:31:09PM +0300, Roger Quadros wrote:
>> Some TI SoCs like DRA7 have a RAMINIT register specification
>> different from the other AMxx SoCs and as expected by the
>> existing driver.
>>
>> To add more insanity, this register is shared with other
>> IPs like DSS, PCIe and PWM.
>>
>> Provides a more generic mechanism to specify the RAMINIT
>> register location and START/DONE bit position and use the
>> syscon/regmap framework to access the register.
>>
>> Signed-off-by: Roger Quadros <rogerq@ti.com>
>> ---
>> .../devicetree/bindings/net/can/c_can.txt | 7 ++
>> drivers/net/can/c_can/c_can.h | 11 ++-
>> drivers/net/can/c_can/c_can_platform.c | 109 +++++++++++++++------
>> 3 files changed, 95 insertions(+), 32 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/net/can/c_can.txt b/Documentation/devicetree/bindings/net/can/c_can.txt
>> index 8f1ae81..e12d1a1 100644
>> --- a/Documentation/devicetree/bindings/net/can/c_can.txt
>> +++ b/Documentation/devicetree/bindings/net/can/c_can.txt
>> @@ -13,6 +13,13 @@ Optional properties:
>> - ti,hwmods : Must be "d_can<n>" or "c_can<n>", n being the
>> instance number
>>
>> +- ti,raminit-syscon : Handle to system control region that contains the
>> + RAMINIT register. If specified, the second memory resource
>> + in the reg property must index into the RAMINIT
>> + register within the syscon region
>
> There seems to be a simple "syscon" property these days.
>
>> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
>> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
>
> This should not be encoded in DT! This is not describing hardware setup.
> The driver should know where the bits are for the syscon phandle,
> depending on which SoC it runs...
Is the register shared by more than one core? If so the information has
to go somewhere. Using an alias in the DT is probably the wrong approach.
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Wolfram Sang @ 2014-09-30 13:52 UTC (permalink / raw)
To: Roger Quadros
Cc: wg, mkl, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar,
nm, sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <542AB137.30507@ti.com>
[-- Attachment #1: Type: text/plain, Size: 1414 bytes --]
> >> +- ti,raminit-syscon : Handle to system control region that contains the
> >> + RAMINIT register. If specified, the second memory resource
> >> + in the reg property must index into the RAMINIT
> >> + register within the syscon region
> >
> > There seems to be a simple "syscon" property these days.
>
> I had used plain "syscon" in the earlier revisions but was asked to make it a TI specific
> property since only TI uses this mechanism.
I see, when was that? Currently, it looks messy :( Grepping through the
dts files in 3.17-rc7, I see that bcm7445 uses "syscon" and variants
with prefixes, STE uses it, too. Samsung uses "samsung,syscon-phandle",
st uses "st,syscon"...
The MACID readout patches for AM335x (just applied right now) also use
"syscon": https://patchwork.ozlabs.org/patch/394289/
I'd vote for a generic "syscon" to be OK, yet I guess the DT maintainers
will have the final word.
> >> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
> >> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
> >
> > This should not be encoded in DT! This is not describing hardware setup.
> > The driver should know where the bits are for the syscon phandle,
> > depending on which SoC it runs...
> >
> OK. I'll think of matching the compatible ID with SOC specific data in the driver.
Great, thanks!
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Roger Quadros @ 2014-09-30 13:58 UTC (permalink / raw)
To: Wolfram Sang
Cc: wg, mkl, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar,
nm, sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <20140930135226.GO1325@katana>
On 09/30/2014 04:52 PM, Wolfram Sang wrote:
>
>>>> +- ti,raminit-syscon : Handle to system control region that contains the
>>>> + RAMINIT register. If specified, the second memory resource
>>>> + in the reg property must index into the RAMINIT
>>>> + register within the syscon region
>>>
>>> There seems to be a simple "syscon" property these days.
>>
>> I had used plain "syscon" in the earlier revisions but was asked to make it a TI specific
>> property since only TI uses this mechanism.
>
> I see, when was that? Currently, it looks messy :( Grepping through the
> dts files in 3.17-rc7, I see that bcm7445 uses "syscon" and variants
> with prefixes, STE uses it, too. Samsung uses "samsung,syscon-phandle",
> st uses "st,syscon"...
>
> The MACID readout patches for AM335x (just applied right now) also use
> "syscon": https://patchwork.ozlabs.org/patch/394289/
>
> I'd vote for a generic "syscon" to be OK, yet I guess the DT maintainers
> will have the final word.
Now that I actually checked, it was never just "syscon" but just without the "ti" vendor prefix.
Sorry for the misinformation.
As just TI is using this out of band RAMINIT mechanism, should it be "ti,syscon" or just "syscon"?
cheers,
-roger
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Roger Quadros @ 2014-09-30 14:02 UTC (permalink / raw)
To: Marc Kleine-Budde, Wolfram Sang
Cc: wg, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar, nm,
sergei.shtylyov, linux-omap, linux-can, netdev@vger.kernel.org
In-Reply-To: <542AB400.6080608@pengutronix.de>
On 09/30/2014 04:45 PM, Marc Kleine-Budde wrote:
> On 09/30/2014 03:26 PM, Wolfram Sang wrote:
>> On Tue, Sep 09, 2014 at 05:31:09PM +0300, Roger Quadros wrote:
>>> Some TI SoCs like DRA7 have a RAMINIT register specification
>>> different from the other AMxx SoCs and as expected by the
>>> existing driver.
>>>
>>> To add more insanity, this register is shared with other
>>> IPs like DSS, PCIe and PWM.
>>>
>>> Provides a more generic mechanism to specify the RAMINIT
>>> register location and START/DONE bit position and use the
>>> syscon/regmap framework to access the register.
>>>
>>> Signed-off-by: Roger Quadros <rogerq@ti.com>
>>> ---
>>> .../devicetree/bindings/net/can/c_can.txt | 7 ++
>>> drivers/net/can/c_can/c_can.h | 11 ++-
>>> drivers/net/can/c_can/c_can_platform.c | 109 +++++++++++++++------
>>> 3 files changed, 95 insertions(+), 32 deletions(-)
>>>
>>> diff --git a/Documentation/devicetree/bindings/net/can/c_can.txt b/Documentation/devicetree/bindings/net/can/c_can.txt
>>> index 8f1ae81..e12d1a1 100644
>>> --- a/Documentation/devicetree/bindings/net/can/c_can.txt
>>> +++ b/Documentation/devicetree/bindings/net/can/c_can.txt
>>> @@ -13,6 +13,13 @@ Optional properties:
>>> - ti,hwmods : Must be "d_can<n>" or "c_can<n>", n being the
>>> instance number
>>>
>>> +- ti,raminit-syscon : Handle to system control region that contains the
>>> + RAMINIT register. If specified, the second memory resource
>>> + in the reg property must index into the RAMINIT
>>> + register within the syscon region
>>
>> There seems to be a simple "syscon" property these days.
>>
>>> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
>>> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
>>
>> This should not be encoded in DT! This is not describing hardware setup.
>> The driver should know where the bits are for the syscon phandle,
>> depending on which SoC it runs...
>
> Is the register shared by more than one core? If so the information has
> to go somewhere. Using an alias in the DT is probably the wrong approach.
In case of the DRA7xx Soc, this syscon register is shared by multiple cores. (CAN, Display, etc), sigh!!
For AM335x and AM43xx SoCs, the register is used only for DCAN. but shared by 2 DCAN instances.
Shouldn't the information for the CAN specific bits lie in the CAN driver?
cheers,
-roger
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Marc Kleine-Budde @ 2014-09-30 14:11 UTC (permalink / raw)
To: Roger Quadros, Wolfram Sang
Cc: wg, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar, nm,
sergei.shtylyov, linux-omap, linux-can, netdev@vger.kernel.org
In-Reply-To: <542AB801.6050805@ti.com>
[-- Attachment #1: Type: text/plain, Size: 3096 bytes --]
On 09/30/2014 04:02 PM, Roger Quadros wrote:
> On 09/30/2014 04:45 PM, Marc Kleine-Budde wrote:
>> On 09/30/2014 03:26 PM, Wolfram Sang wrote:
>>> On Tue, Sep 09, 2014 at 05:31:09PM +0300, Roger Quadros wrote:
>>>> Some TI SoCs like DRA7 have a RAMINIT register specification
>>>> different from the other AMxx SoCs and as expected by the
>>>> existing driver.
>>>>
>>>> To add more insanity, this register is shared with other
>>>> IPs like DSS, PCIe and PWM.
>>>>
>>>> Provides a more generic mechanism to specify the RAMINIT
>>>> register location and START/DONE bit position and use the
>>>> syscon/regmap framework to access the register.
>>>>
>>>> Signed-off-by: Roger Quadros <rogerq@ti.com>
>>>> ---
>>>> .../devicetree/bindings/net/can/c_can.txt | 7 ++
>>>> drivers/net/can/c_can/c_can.h | 11 ++-
>>>> drivers/net/can/c_can/c_can_platform.c | 109 +++++++++++++++------
>>>> 3 files changed, 95 insertions(+), 32 deletions(-)
>>>>
>>>> diff --git a/Documentation/devicetree/bindings/net/can/c_can.txt b/Documentation/devicetree/bindings/net/can/c_can.txt
>>>> index 8f1ae81..e12d1a1 100644
>>>> --- a/Documentation/devicetree/bindings/net/can/c_can.txt
>>>> +++ b/Documentation/devicetree/bindings/net/can/c_can.txt
>>>> @@ -13,6 +13,13 @@ Optional properties:
>>>> - ti,hwmods : Must be "d_can<n>" or "c_can<n>", n being the
>>>> instance number
>>>>
>>>> +- ti,raminit-syscon : Handle to system control region that contains the
>>>> + RAMINIT register. If specified, the second memory resource
>>>> + in the reg property must index into the RAMINIT
>>>> + register within the syscon region
>>>
>>> There seems to be a simple "syscon" property these days.
>>>
>>>> +- ti,raminit-start-bit : Bit posistion of START bit in the RAMINIT register
>>>> +- ti,raminit-done-bit : Bit position of DONE bit in the RAMINIT register
>>>
>>> This should not be encoded in DT! This is not describing hardware setup.
>>> The driver should know where the bits are for the syscon phandle,
>>> depending on which SoC it runs...
>>
>> Is the register shared by more than one core? If so the information has
>> to go somewhere. Using an alias in the DT is probably the wrong approach.
>
> In case of the DRA7xx Soc, this syscon register is shared by multiple cores. (CAN, Display, etc), sigh!!
> For AM335x and AM43xx SoCs, the register is used only for DCAN. but shared by 2 DCAN instances.
>
> Shouldn't the information for the CAN specific bits lie in the CAN driver?
As there are more than one DCAN instance on the same SOC the information
which DCAN instance uses which bits in the register has to go somewhere.
We might add these bits as the 3rd and 4th parameter to the syscon phandle.
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]
^ permalink raw reply
* Re: [net-next PATCH] dql: add a burst attribute
From: Eric Dumazet @ 2014-09-30 14:17 UTC (permalink / raw)
To: Florian Westphal
Cc: Jesper Dangaard Brouer, netdev, David S. Miller, Tom Herbert,
Hannes Frederic Sowa, Daniel Borkmann, Jamal Hadi Salim,
Alexander Duyck, John Fastabend, Dave Taht, toke
In-Reply-To: <20140930134655.GA24565@breakpoint.cc>
On Tue, 2014-09-30 at 15:46 +0200, Florian Westphal wrote:
> Eric Dumazet <eric.dumazet@gmail.com> wrote:
> > From: Eric Dumazet <edumazet@google.com>
> >
> > Add a new dql attribute, to control how much packets we are allowed to
> > burst from qdisc to device.
>
> I understand the motivation for this, but I find it a bit out-of-place
> to have a 'packet' type counter in bql context?
>
> Would it perhaps make more sense to restrict bulk dequeues by an upper
> (possibly changeable from userspace) byte counter limit?
The byte count is already provided : its the BQL limit.
We already have ways to tune it (limit_min & limit_max)
We do not think we need something else here.
That was the whole point of reusing BQL in the first place : Its already
powerful and implemented.
This new attribute is a packet burst.
When I suggested 'pburst' attribute for pktgen to Alexei, he sent a
patch with "burst". Nobody complained.
Now I am reusing what apparently was OK for pktgen, you seem to object.
If you feel not comfortable with "burst", rename it to whatever you
think is best.
But please, do not hard code magic 7 in your code.
Thanks
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Wolfram Sang @ 2014-09-30 14:19 UTC (permalink / raw)
To: Roger Quadros
Cc: wg, mkl, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar,
nm, sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <542AB6E9.9000907@ti.com>
[-- Attachment #1: Type: text/plain, Size: 578 bytes --]
> As just TI is using this out of band RAMINIT mechanism, should it be "ti,syscon" or just "syscon"?
Yes, only TI uses this out-of-band RAMINIT (currently, at least). So, we
need an (optional) way to describe that. However, accessing syscon
registers in general is not TI specific and a generic way to do this
should be used. Which looks to me like the "syscon" property to allow
access to the register. Still, we should ask DT maintainers about it,
maybe they prefer a more precise property like "syscon-raminit" to allow
for further syscon extensions later or something...
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 819 bytes --]
^ permalink raw reply
* Re: [PATCH v2 2/3] net: can: c_can: Add syscon/regmap RAMINIT mechanism
From: Marc Kleine-Budde @ 2014-09-30 14:22 UTC (permalink / raw)
To: Wolfram Sang, Roger Quadros
Cc: wg, tony, tglx, mugunthanvnm, george.cherian, balbi, nsekhar, nm,
sergei.shtylyov, linux-omap, linux-can, netdev
In-Reply-To: <20140930141909.GP1325@katana>
[-- Attachment #1: Type: text/plain, Size: 1047 bytes --]
On 09/30/2014 04:19 PM, Wolfram Sang wrote:
>
>> As just TI is using this out of band RAMINIT mechanism, should it be "ti,syscon" or just "syscon"?
>
> Yes, only TI uses this out-of-band RAMINIT (currently, at least). So, we
> need an (optional) way to describe that. However, accessing syscon
> registers in general is not TI specific and a generic way to do this
> should be used. Which looks to me like the "syscon" property to allow
> access to the register. Still, we should ask DT maintainers about it,
> maybe they prefer a more precise property like "syscon-raminit" to allow
> for further syscon extensions later or something...
What do you think about putting the bit information in the
syscon-raminit phandle as additional arguments?
Marc
--
Pengutronix e.K. | Marc Kleine-Budde |
Industrial Linux Solutions | Phone: +49-231-2826-924 |
Vertretung West/Dortmund | Fax: +49-5121-206917-5555 |
Amtsgericht Hildesheim, HRA 2686 | http://www.pengutronix.de |
[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 181 bytes --]
^ permalink raw reply
* Salaam Alaikum,
From: pastor james @ 2014-09-30 14:25 UTC (permalink / raw)
Salaam Alaikum,
This mail might come to you as a surprise and the temptation to
ignore it as unserious could come into your mind; but please, consider
it a divine wish and accept it with a deep sense of humility.
I work with a bank In West Africa, Togo Lome,
I have a business transaction for you. In my department we discovered
an abandoned sum of $10.5 Million US Dollars.(Ten Million Five
Hundread Thousand United States Dollars) In an account that belongs to
one of our foreign customer who died along with his entire family in a
Bomb blast in irag Since his supposed next of kin died
along side with him at the Bomb blast, there is nobody to claim the
left over balance in the account.
It is therefore upon this discovery that I and other officials in my
department decided to seek your assistance and present you to the bank
as his Next of kin. If you accept i would give you the guide lines of
how we can achieve this transfer of the balance (10.5Million Dollars)
to your account..and we will shear the money 50//50%.i will not fail
to bring to your notice that this transaction is 100% risk free that
you should not entertain any atom of fear as all the arrangements has
been made for the transfer. I expect that you will give me your
telephone and fax numbers for easy communication with you and for more
details through my private email ID Below;
(h_idris0o5@yahoo.com)
Best Regards
Hamidu.
^ permalink raw reply
* Re: [net-next PATCH] dql: add a burst attribute
From: Tom Herbert @ 2014-09-30 14:26 UTC (permalink / raw)
To: Eric Dumazet
Cc: Florian Westphal, Jesper Dangaard Brouer, Linux Netdev List,
David S. Miller, Hannes Frederic Sowa, Daniel Borkmann,
Jamal Hadi Salim, Alexander Duyck, John Fastabend, Dave Taht,
Toke Høiland-Jørgensen
In-Reply-To: <1412086646.30721.81.camel@edumazet-glaptop2.roam.corp.google.com>
On Tue, Sep 30, 2014 at 7:17 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Tue, 2014-09-30 at 15:46 +0200, Florian Westphal wrote:
>> Eric Dumazet <eric.dumazet@gmail.com> wrote:
>> > From: Eric Dumazet <edumazet@google.com>
>> >
>> > Add a new dql attribute, to control how much packets we are allowed to
>> > burst from qdisc to device.
>>
>> I understand the motivation for this, but I find it a bit out-of-place
>> to have a 'packet' type counter in bql context?
>>
>> Would it perhaps make more sense to restrict bulk dequeues by an upper
>> (possibly changeable from userspace) byte counter limit?
>
> The byte count is already provided : its the BQL limit.
> We already have ways to tune it (limit_min & limit_max)
> We do not think we need something else here.
> That was the whole point of reusing BQL in the first place : Its already
> powerful and implemented.
>
> This new attribute is a packet burst.
>
Yes, this is the wrong place to put it. DQL knows nothing about
packets or technically even bytes, it's a generic library that deal
with "objects" on a queue. I don't see how burst relates to this.
I would infer from the patch that what you want is to add an attribute
to the TX queue. Why not do this directly?
>
> When I suggested 'pburst' attribute for pktgen to Alexei, he sent a
> patch with "burst". Nobody complained.
>
> Now I am reusing what apparently was OK for pktgen, you seem to object.
>
> If you feel not comfortable with "burst", rename it to whatever you
> think is best.
>
> But please, do not hard code magic 7 in your code.
>
> Thanks
>
>
^ permalink raw reply
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