2013-07-07 14:25:49 +00:00
|
|
|
/*
|
2015-04-02 14:07:29 +00:00
|
|
|
* Copyright (c) 2013-2015, Mellanox Technologies. All rights reserved.
|
2013-07-07 14:25:49 +00:00
|
|
|
*
|
|
|
|
|
* 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.
|
|
|
|
|
*/
|
|
|
|
|
|
|
|
|
|
#ifndef MLX5_CORE_CQ_H
|
|
|
|
|
#define MLX5_CORE_CQ_H
|
|
|
|
|
|
|
|
|
|
#include <linux/mlx5/driver.h>
|
2017-10-20 07:23:40 +00:00
|
|
|
#include <linux/refcount.h>
|
2013-07-07 14:25:49 +00:00
|
|
|
|
|
|
|
|
struct mlx5_core_cq {
|
|
|
|
|
u32 cqn;
|
|
|
|
|
int cqe_sz;
|
|
|
|
|
__be32 *set_ci_db;
|
|
|
|
|
__be32 *arm_db;
|
2017-10-20 07:23:40 +00:00
|
|
|
refcount_t refcount;
|
2013-07-07 14:25:49 +00:00
|
|
|
struct completion free;
|
|
|
|
|
unsigned vector;
|
2016-01-17 09:25:47 +00:00
|
|
|
unsigned int irqn;
|
2019-06-30 16:23:27 +00:00
|
|
|
void (*comp)(struct mlx5_core_cq *cq, struct mlx5_eqe *eqe);
|
2013-07-07 14:25:49 +00:00
|
|
|
void (*event) (struct mlx5_core_cq *, enum mlx5_event);
|
|
|
|
|
u32 cons_index;
|
|
|
|
|
unsigned arm_sn;
|
|
|
|
|
struct mlx5_rsc_debug *dbg;
|
|
|
|
|
int pid;
|
net/mlx5_core: Use tasklet for user-space CQ completion events
Previously, we've fired all our completion callbacks straight from
our ISR.
Some of those callbacks were lightweight (for example, mlx5 Ethernet
napi callbacks), but some of them did more work (for example,
the user-space RDMA stack uverbs' completion handler). Besides that,
doing more than the minimal work in ISR is generally considered wrong,
it could even lead to a hard lockup of the system. Since when a lot
of completion events are generated by the hardware, the loop over
those events could be so long, that we'll get into a hard lockup by
the system watchdog.
In order to avoid that, add a new way of invoking completion events
callbacks. In the interrupt itself, we add the CQs which receive
completion event to a per-EQ list and schedule a tasklet. In the
tasklet context we loop over all the CQs in the list and invoke the
user callback.
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2016-04-17 14:08:40 +00:00
|
|
|
struct {
|
|
|
|
|
struct list_head list;
|
2019-06-30 16:23:27 +00:00
|
|
|
void (*comp)(struct mlx5_core_cq *cq, struct mlx5_eqe *eqe);
|
net/mlx5_core: Use tasklet for user-space CQ completion events
Previously, we've fired all our completion callbacks straight from
our ISR.
Some of those callbacks were lightweight (for example, mlx5 Ethernet
napi callbacks), but some of them did more work (for example,
the user-space RDMA stack uverbs' completion handler). Besides that,
doing more than the minimal work in ISR is generally considered wrong,
it could even lead to a hard lockup of the system. Since when a lot
of completion events are generated by the hardware, the loop over
those events could be so long, that we'll get into a hard lockup by
the system watchdog.
In order to avoid that, add a new way of invoking completion events
callbacks. In the interrupt itself, we add the CQs which receive
completion event to a per-EQ list and schedule a tasklet. In the
tasklet context we loop over all the CQs in the list and invoke the
user callback.
Signed-off-by: Matan Barak <matanb@mellanox.com>
Signed-off-by: Doug Ledford <dledford@redhat.com>
2016-04-17 14:08:40 +00:00
|
|
|
void *priv;
|
|
|
|
|
} tasklet_ctx;
|
2016-06-17 12:01:38 +00:00
|
|
|
int reset_notify_added;
|
|
|
|
|
struct list_head reset_notify;
|
2018-11-19 18:52:39 +00:00
|
|
|
struct mlx5_eq_comp *eq;
|
2018-09-20 18:35:20 +00:00
|
|
|
u16 uid;
|
2013-07-07 14:25:49 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
enum {
|
|
|
|
|
MLX5_CQE_SYNDROME_LOCAL_LENGTH_ERR = 0x01,
|
|
|
|
|
MLX5_CQE_SYNDROME_LOCAL_QP_OP_ERR = 0x02,
|
|
|
|
|
MLX5_CQE_SYNDROME_LOCAL_PROT_ERR = 0x04,
|
|
|
|
|
MLX5_CQE_SYNDROME_WR_FLUSH_ERR = 0x05,
|
|
|
|
|
MLX5_CQE_SYNDROME_MW_BIND_ERR = 0x06,
|
|
|
|
|
MLX5_CQE_SYNDROME_BAD_RESP_ERR = 0x10,
|
|
|
|
|
MLX5_CQE_SYNDROME_LOCAL_ACCESS_ERR = 0x11,
|
|
|
|
|
MLX5_CQE_SYNDROME_REMOTE_INVAL_REQ_ERR = 0x12,
|
|
|
|
|
MLX5_CQE_SYNDROME_REMOTE_ACCESS_ERR = 0x13,
|
|
|
|
|
MLX5_CQE_SYNDROME_REMOTE_OP_ERR = 0x14,
|
|
|
|
|
MLX5_CQE_SYNDROME_TRANSPORT_RETRY_EXC_ERR = 0x15,
|
|
|
|
|
MLX5_CQE_SYNDROME_RNR_RETRY_EXC_ERR = 0x16,
|
|
|
|
|
MLX5_CQE_SYNDROME_REMOTE_ABORTED_ERR = 0x22,
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
enum {
|
|
|
|
|
MLX5_CQE_OWNER_MASK = 1,
|
|
|
|
|
MLX5_CQE_REQ = 0,
|
|
|
|
|
MLX5_CQE_RESP_WR_IMM = 1,
|
|
|
|
|
MLX5_CQE_RESP_SEND = 2,
|
|
|
|
|
MLX5_CQE_RESP_SEND_IMM = 3,
|
|
|
|
|
MLX5_CQE_RESP_SEND_INV = 4,
|
2014-01-14 15:45:18 +00:00
|
|
|
MLX5_CQE_RESIZE_CQ = 5,
|
2014-02-23 12:19:12 +00:00
|
|
|
MLX5_CQE_SIG_ERR = 12,
|
2013-07-07 14:25:49 +00:00
|
|
|
MLX5_CQE_REQ_ERR = 13,
|
|
|
|
|
MLX5_CQE_RESP_ERR = 14,
|
2014-01-14 15:45:18 +00:00
|
|
|
MLX5_CQE_INVALID = 15,
|
2013-07-07 14:25:49 +00:00
|
|
|
};
|
|
|
|
|
|
|
|
|
|
enum {
|
2024-04-19 08:04:44 +00:00
|
|
|
MLX5_CQ_MODIFY_PERIOD = BIT(0),
|
|
|
|
|
MLX5_CQ_MODIFY_COUNT = BIT(1),
|
|
|
|
|
MLX5_CQ_MODIFY_OVERRUN = BIT(2),
|
|
|
|
|
MLX5_CQ_MODIFY_PERIOD_MODE = BIT(4),
|
2013-07-07 14:25:49 +00:00
|
|
|
};
|
|
|
|
|
|
2014-01-14 15:45:18 +00:00
|
|
|
enum {
|
|
|
|
|
MLX5_CQ_OPMOD_RESIZE = 1,
|
|
|
|
|
MLX5_MODIFY_CQ_MASK_LOG_SIZE = 1 << 0,
|
|
|
|
|
MLX5_MODIFY_CQ_MASK_PG_OFFSET = 1 << 1,
|
|
|
|
|
MLX5_MODIFY_CQ_MASK_PG_SIZE = 1 << 2,
|
|
|
|
|
};
|
|
|
|
|
|
2013-07-07 14:25:49 +00:00
|
|
|
struct mlx5_cq_modify_params {
|
|
|
|
|
int type;
|
|
|
|
|
union {
|
|
|
|
|
struct {
|
|
|
|
|
u32 page_offset;
|
|
|
|
|
u8 log_cq_size;
|
|
|
|
|
} resize;
|
|
|
|
|
|
|
|
|
|
struct {
|
|
|
|
|
} moder;
|
|
|
|
|
|
|
|
|
|
struct {
|
|
|
|
|
} mapping;
|
|
|
|
|
} params;
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
enum {
|
2018-11-05 22:05:37 +00:00
|
|
|
CQE_STRIDE_64 = 0,
|
|
|
|
|
CQE_STRIDE_128 = 1,
|
|
|
|
|
CQE_STRIDE_128_PAD = 2,
|
2013-07-07 14:25:49 +00:00
|
|
|
};
|
|
|
|
|
|
2017-11-13 08:51:15 +00:00
|
|
|
#define MLX5_MAX_CQ_PERIOD (BIT(__mlx5_bit_sz(cqc, cq_period)) - 1)
|
|
|
|
|
#define MLX5_MAX_CQ_COUNT (BIT(__mlx5_bit_sz(cqc, cq_max_count)) - 1)
|
|
|
|
|
|
2017-10-19 05:25:53 +00:00
|
|
|
static inline int cqe_sz_to_mlx_sz(u8 size, int padding_128_en)
|
2013-07-07 14:25:49 +00:00
|
|
|
{
|
2018-11-05 22:05:37 +00:00
|
|
|
return padding_128_en ? CQE_STRIDE_128_PAD :
|
|
|
|
|
size == 64 ? CQE_STRIDE_64 : CQE_STRIDE_128;
|
2013-07-07 14:25:49 +00:00
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline void mlx5_cq_set_ci(struct mlx5_core_cq *cq)
|
|
|
|
|
{
|
|
|
|
|
*cq->set_ci_db = cpu_to_be32(cq->cons_index & 0xffffff);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
enum {
|
|
|
|
|
MLX5_CQ_DB_REQ_NOT_SOL = 1 << 24,
|
|
|
|
|
MLX5_CQ_DB_REQ_NOT = 0 << 24
|
|
|
|
|
};
|
|
|
|
|
|
|
|
|
|
static inline void mlx5_cq_arm(struct mlx5_core_cq *cq, u32 cmd,
|
|
|
|
|
void __iomem *uar_page,
|
2015-04-02 14:07:33 +00:00
|
|
|
u32 cons_index)
|
2013-07-07 14:25:49 +00:00
|
|
|
{
|
|
|
|
|
__be32 doorbell[2];
|
|
|
|
|
u32 sn;
|
|
|
|
|
u32 ci;
|
|
|
|
|
|
|
|
|
|
sn = cq->arm_sn & 3;
|
2015-04-02 14:07:33 +00:00
|
|
|
ci = cons_index & 0xffffff;
|
2013-07-07 14:25:49 +00:00
|
|
|
|
|
|
|
|
*cq->arm_db = cpu_to_be32(sn << 28 | cmd | ci);
|
|
|
|
|
|
|
|
|
|
/* Make sure that the doorbell record in host memory is
|
|
|
|
|
* written before ringing the doorbell via PCI MMIO.
|
|
|
|
|
*/
|
|
|
|
|
wmb();
|
|
|
|
|
|
|
|
|
|
doorbell[0] = cpu_to_be32(sn << 28 | cmd | ci);
|
|
|
|
|
doorbell[1] = cpu_to_be32(cq->cqn);
|
|
|
|
|
|
2019-03-29 22:37:52 +00:00
|
|
|
mlx5_write64(doorbell, uar_page + MLX5_CQ_DOORBELL);
|
2013-07-07 14:25:49 +00:00
|
|
|
}
|
|
|
|
|
|
2018-02-01 11:32:00 +00:00
|
|
|
static inline void mlx5_cq_hold(struct mlx5_core_cq *cq)
|
|
|
|
|
{
|
|
|
|
|
refcount_inc(&cq->refcount);
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
static inline void mlx5_cq_put(struct mlx5_core_cq *cq)
|
|
|
|
|
{
|
|
|
|
|
if (refcount_dec_and_test(&cq->refcount))
|
|
|
|
|
complete(&cq->free);
|
|
|
|
|
}
|
|
|
|
|
|
2025-11-09 09:49:03 +00:00
|
|
|
void mlx5_add_cq_to_tasklet(struct mlx5_core_cq *cq, struct mlx5_eqe *eqe);
|
net/mlx5: Use mlx5_cmd_do() in core create_{cq,dct}
mlx5_core_create_{cq/dct} functions are non-trivial mlx5 commands
functions. They check command execution status themselves and hide
valuable FW failure information.
For mlx5_core/eth kernel user this is what we actually want, but for a
devx/rdma user the hidden information is essential and should be propagated
up to the caller, thus we convert these commands to use mlx5_cmd_do
to return the FW/driver and command outbox status as is, and let the caller
decide what to do with it.
For kernel callers of mlx5_core_create_{cq/dct} or those who only care about
the binary status (FAIL/SUCCESS) they must check status themselves via
mlx5_cmd_check() to restore the current behavior.
err = mlx5_create_cq(in, out)
err = mlx5_cmd_check(err, in, out)
if (err)
// handle err
For DEVX users and those who care about full visibility, They will just
propagate the error to user space, and app can check if err == -EREMOTEIO,
then outbox.{status,syndrome} are valid.
API Note:
mlx5_cmd_check() must be used by kernel users since it allows the driver
to intercept the command execution status and return a driver simulated
status in case of driver induced error handling or reset/recovery flows.
Signed-off-by: Saeed Mahameed <saeedm@mellanox.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2020-03-31 04:12:58 +00:00
|
|
|
int mlx5_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
|
|
|
|
|
u32 *in, int inlen, u32 *out, int outlen);
|
2013-07-07 14:25:49 +00:00
|
|
|
int mlx5_core_create_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
|
2019-06-30 16:23:25 +00:00
|
|
|
u32 *in, int inlen, u32 *out, int outlen);
|
2013-07-07 14:25:49 +00:00
|
|
|
int mlx5_core_destroy_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq);
|
|
|
|
|
int mlx5_core_query_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
|
2020-04-09 08:39:14 +00:00
|
|
|
u32 *out);
|
2013-07-07 14:25:49 +00:00
|
|
|
int mlx5_core_modify_cq(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq,
|
2016-07-15 23:33:22 +00:00
|
|
|
u32 *in, int inlen);
|
2015-05-28 19:28:44 +00:00
|
|
|
int mlx5_core_modify_cq_moderation(struct mlx5_core_dev *dev,
|
|
|
|
|
struct mlx5_core_cq *cq, u16 cq_period,
|
|
|
|
|
u16 cq_max_count);
|
2017-12-31 10:55:26 +00:00
|
|
|
static inline void mlx5_dump_err_cqe(struct mlx5_core_dev *dev,
|
|
|
|
|
struct mlx5_err_cqe *err_cqe)
|
|
|
|
|
{
|
|
|
|
|
print_hex_dump(KERN_WARNING, "", DUMP_PREFIX_OFFSET, 16, 1, err_cqe,
|
|
|
|
|
sizeof(*err_cqe), false);
|
|
|
|
|
}
|
2013-07-07 14:25:49 +00:00
|
|
|
int mlx5_debug_cq_add(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq);
|
|
|
|
|
void mlx5_debug_cq_remove(struct mlx5_core_dev *dev, struct mlx5_core_cq *cq);
|
|
|
|
|
|
|
|
|
|
#endif /* MLX5_CORE_CQ_H */
|