linux/net/smc/smc_wr.h
Kai Shen 79a22238b4 net/smc: Use percpu ref for wr tx reference
The refcount wr_tx_refcnt may cause cache thrashing problems among
cores and we can use percpu ref to mitigate this issue here. We
gain some performance improvement with percpu ref here on our
customized smc-r verion. Applying cache alignment may also mitigate
this problem but it seem more reasonable to use percpu ref here.
We can also replace wr_reg_refcnt with one percpu reference like
wr_tx_refcnt.

redis-benchmark on smc-r with atomic wr_tx_refcnt:
SET: 525707.06 requests per second, p50=0.087 msec
GET: 554877.38 requests per second, p50=0.087 msec

redis-benchmark on the percpu_ref version:
SET: 540482.06 requests per second, p50=0.087 msec
GET: 570711.12 requests per second, p50=0.079 msec

Cases are like "redis-benchmark -h x.x.x.x -q -t set,get -P 1 -n
5000000 -c 50 -d 10 --threads 4".

Signed-off-by: Kai Shen <KaiShen@linux.alibaba.com>
Reviewed-by: Tony Lu <tonylu@linux.alibaba.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2023-03-17 08:59:01 +00:00

140 lines
4.0 KiB
C

/* SPDX-License-Identifier: GPL-2.0 */
/*
* Shared Memory Communications over RDMA (SMC-R) and RoCE
*
* Work Requests exploiting Infiniband API
*
* Copyright IBM Corp. 2016
*
* Author(s): Steffen Maier <maier@linux.vnet.ibm.com>
*/
#ifndef SMC_WR_H
#define SMC_WR_H
#include <linux/atomic.h>
#include <rdma/ib_verbs.h>
#include <asm/div64.h>
#include "smc.h"
#include "smc_core.h"
#define SMC_WR_BUF_CNT 16 /* # of ctrl buffers per link */
#define SMC_WR_TX_WAIT_FREE_SLOT_TIME (10 * HZ)
#define SMC_WR_TX_SIZE 44 /* actual size of wr_send data (<=SMC_WR_BUF_SIZE) */
#define SMC_WR_TX_PEND_PRIV_SIZE 32
struct smc_wr_tx_pend_priv {
u8 priv[SMC_WR_TX_PEND_PRIV_SIZE];
};
typedef void (*smc_wr_tx_handler)(struct smc_wr_tx_pend_priv *,
struct smc_link *,
enum ib_wc_status);
typedef bool (*smc_wr_tx_filter)(struct smc_wr_tx_pend_priv *,
unsigned long);
typedef void (*smc_wr_tx_dismisser)(struct smc_wr_tx_pend_priv *);
struct smc_wr_rx_handler {
struct hlist_node list; /* hash table collision resolution */
void (*handler)(struct ib_wc *, void *);
u8 type;
};
/* Only used by RDMA write WRs.
* All other WRs (CDC/LLC) use smc_wr_tx_send handling WR_ID implicitly
*/
static inline long smc_wr_tx_get_next_wr_id(struct smc_link *link)
{
return atomic_long_inc_return(&link->wr_tx_id);
}
static inline void smc_wr_tx_set_wr_id(atomic_long_t *wr_tx_id, long val)
{
atomic_long_set(wr_tx_id, val);
}
static inline bool smc_wr_tx_link_hold(struct smc_link *link)
{
if (!smc_link_sendable(link))
return false;
percpu_ref_get(&link->wr_tx_refs);
return true;
}
static inline void smc_wr_tx_link_put(struct smc_link *link)
{
percpu_ref_put(&link->wr_tx_refs);
}
static inline void smc_wr_drain_cq(struct smc_link *lnk)
{
wait_event(lnk->wr_rx_empty_wait, lnk->wr_rx_id_compl == lnk->wr_rx_id);
}
static inline void smc_wr_wakeup_tx_wait(struct smc_link *lnk)
{
wake_up_all(&lnk->wr_tx_wait);
}
static inline void smc_wr_wakeup_reg_wait(struct smc_link *lnk)
{
wake_up(&lnk->wr_reg_wait);
}
/* post a new receive work request to fill a completed old work request entry */
static inline int smc_wr_rx_post(struct smc_link *link)
{
int rc;
u64 wr_id, temp_wr_id;
u32 index;
wr_id = ++link->wr_rx_id; /* tasklet context, thus not atomic */
temp_wr_id = wr_id;
index = do_div(temp_wr_id, link->wr_rx_cnt);
link->wr_rx_ibs[index].wr_id = wr_id;
rc = ib_post_recv(link->roce_qp, &link->wr_rx_ibs[index], NULL);
return rc;
}
int smc_wr_create_link(struct smc_link *lnk);
int smc_wr_alloc_link_mem(struct smc_link *lnk);
int smc_wr_alloc_lgr_mem(struct smc_link_group *lgr);
void smc_wr_free_link(struct smc_link *lnk);
void smc_wr_free_link_mem(struct smc_link *lnk);
void smc_wr_free_lgr_mem(struct smc_link_group *lgr);
void smc_wr_remember_qp_attr(struct smc_link *lnk);
void smc_wr_remove_dev(struct smc_ib_device *smcibdev);
void smc_wr_add_dev(struct smc_ib_device *smcibdev);
int smc_wr_tx_get_free_slot(struct smc_link *link, smc_wr_tx_handler handler,
struct smc_wr_buf **wr_buf,
struct smc_rdma_wr **wrs,
struct smc_wr_tx_pend_priv **wr_pend_priv);
int smc_wr_tx_get_v2_slot(struct smc_link *link,
smc_wr_tx_handler handler,
struct smc_wr_v2_buf **wr_buf,
struct smc_wr_tx_pend_priv **wr_pend_priv);
int smc_wr_tx_put_slot(struct smc_link *link,
struct smc_wr_tx_pend_priv *wr_pend_priv);
int smc_wr_tx_send(struct smc_link *link,
struct smc_wr_tx_pend_priv *wr_pend_priv);
int smc_wr_tx_v2_send(struct smc_link *link,
struct smc_wr_tx_pend_priv *priv, int len);
int smc_wr_tx_send_wait(struct smc_link *link, struct smc_wr_tx_pend_priv *priv,
unsigned long timeout);
void smc_wr_tx_cq_handler(struct ib_cq *ib_cq, void *cq_context);
void smc_wr_tx_wait_no_pending_sends(struct smc_link *link);
int smc_wr_rx_register_handler(struct smc_wr_rx_handler *handler);
int smc_wr_rx_post_init(struct smc_link *link);
void smc_wr_rx_cq_handler(struct ib_cq *ib_cq, void *cq_context);
int smc_wr_reg_send(struct smc_link *link, struct ib_mr *mr);
#endif /* SMC_WR_H */