On 21.09.23 15:19, Wen Gu wrote:
> This patch implements DMB registration, unregistration and data move
> operations of SMC-D loopback.
>
> Signed-off-by: Wen Gu <[email protected]>
> ---
> net/smc/smc_cdc.c | 6 +++
> net/smc/smc_cdc.h | 1 +
> net/smc/smc_loopback.c | 128 +++++++++++++++++++++++++++++++++++++++++++++++--
> net/smc/smc_loopback.h | 13 +++++
> 4 files changed, 145 insertions(+), 3 deletions(-)
>
> diff --git a/net/smc/smc_cdc.c b/net/smc/smc_cdc.c
> index 89105e9..2641800 100644
> --- a/net/smc/smc_cdc.c
> +++ b/net/smc/smc_cdc.c
> @@ -411,6 +411,12 @@ static void smc_cdc_msg_recv(struct smc_sock *smc, struct smc_cdc_msg *cdc)
> static void smcd_cdc_rx_tsklet(struct tasklet_struct *t)
> {
> struct smc_connection *conn = from_tasklet(conn, t, rx_tsklet);
> +
> + smcd_cdc_rx_handler(conn);
> +}
> +
> +void smcd_cdc_rx_handler(struct smc_connection *conn)
> +{
> struct smcd_cdc_msg *data_cdc;
> struct smcd_cdc_msg cdc;
> struct smc_sock *smc;
> diff --git a/net/smc/smc_cdc.h b/net/smc/smc_cdc.h
> index 696cc11..11559d4 100644
> --- a/net/smc/smc_cdc.h
> +++ b/net/smc/smc_cdc.h
> @@ -301,5 +301,6 @@ int smcr_cdc_msg_send_validation(struct smc_connection *conn,
> struct smc_wr_buf *wr_buf);
> int smc_cdc_init(void) __init;
> void smcd_cdc_rx_init(struct smc_connection *conn);
> +void smcd_cdc_rx_handler(struct smc_connection *conn);
>
> #endif /* SMC_CDC_H */
> diff --git a/net/smc/smc_loopback.c b/net/smc/smc_loopback.c
> index 9034ebd..cfbcabf 100644
> --- a/net/smc/smc_loopback.c
> +++ b/net/smc/smc_loopback.c
> @@ -16,6 +16,7 @@
> #include <linux/smc.h>
> #include <net/smc.h>
>
> +#include "smc_cdc.h"
> #include "smc_ism.h"
> #include "smc_loopback.h"
>
> @@ -74,6 +75,93 @@ static int smc_lo_query_rgid(struct smcd_dev *smcd, struct smcd_gid *rgid,
> return 0;
> }
>
> +static int smc_lo_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb,
> + void *client_priv)
> +{
> + struct smc_lo_dmb_node *dmb_node, *tmp_node;
> + struct smc_lo_dev *ldev = smcd->priv;
> + int sba_idx, rc;
> +
> + /* check space for new dmb */
> + for_each_clear_bit(sba_idx, ldev->sba_idx_mask, SMC_LODEV_MAX_DMBS) {
> + if (!test_and_set_bit(sba_idx, ldev->sba_idx_mask))
> + break;
> + }
> + if (sba_idx == SMC_LODEV_MAX_DMBS)
> + return -ENOSPC;
> +
> + dmb_node = kzalloc(sizeof(*dmb_node), GFP_KERNEL);
> + if (!dmb_node) {
> + rc = -ENOMEM;
> + goto err_bit;
> + }
> +
> + dmb_node->sba_idx = sba_idx;
> + dmb_node->cpu_addr = kzalloc(dmb->dmb_len, GFP_KERNEL |
> + __GFP_NOWARN | __GFP_NORETRY |
> + __GFP_NOMEMALLOC);
kzalloc()/kmalloc() allocates physically contigueous memory. Are you
sure it is suitable for allocating the dmb?
> + if (!dmb_node->cpu_addr) {
> + rc = -ENOMEM;
> + goto err_node;
> + }
> + dmb_node->len = dmb->dmb_len;
> + dmb_node->dma_addr = (dma_addr_t)dmb_node->cpu_addr;
> +
> +again:
> + /* add new dmb into hash table */
> + get_random_bytes(&dmb_node->token, sizeof(dmb_node->token));
> + write_lock(&ldev->dmb_ht_lock);
> + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb_node->token) {
> + if (tmp_node->token == dmb_node->token) {
> + write_unlock(&ldev->dmb_ht_lock);
> + goto again;
> + }
> + }
> + hash_add(ldev->dmb_ht, &dmb_node->list, dmb_node->token);
> + write_unlock(&ldev->dmb_ht_lock);
> +
> + dmb->sba_idx = dmb_node->sba_idx;
> + dmb->dmb_tok = dmb_node->token;
> + dmb->cpu_addr = dmb_node->cpu_addr;
> + dmb->dma_addr = dmb_node->dma_addr;
> + dmb->dmb_len = dmb_node->len;
> +
> + return 0;
> +
> +err_node:
> + kfree(dmb_node);
> +err_bit:
> + clear_bit(sba_idx, ldev->sba_idx_mask);
> + return rc;
> +}
> +
> +static int smc_lo_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb)
> +{
> + struct smc_lo_dmb_node *dmb_node = NULL, *tmp_node;
> + struct smc_lo_dev *ldev = smcd->priv;
> +
> + /* remove dmb from hash table */
> + write_lock(&ldev->dmb_ht_lock);
> + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb->dmb_tok) {
> + if (tmp_node->token == dmb->dmb_tok) {
> + dmb_node = tmp_node;
> + break;
> + }
> + }
> + if (!dmb_node) {
> + write_unlock(&ldev->dmb_ht_lock);
> + return -EINVAL;
> + }
> + hash_del(&dmb_node->list);
> + write_unlock(&ldev->dmb_ht_lock);
> +
> + clear_bit(dmb_node->sba_idx, ldev->sba_idx_mask);
> + kfree(dmb_node->cpu_addr);
> + kfree(dmb_node);
> +
> + return 0;
> +}
> +
> static int smc_lo_add_vlan_id(struct smcd_dev *smcd, u64 vlan_id)
> {
> return -EOPNOTSUPP;
> @@ -100,6 +188,38 @@ static int smc_lo_signal_event(struct smcd_dev *dev, struct smcd_gid *rgid,
> return 0;
> }
>
> +static int smc_lo_move_data(struct smcd_dev *smcd, u64 dmb_tok, unsigned int idx,
> + bool sf, unsigned int offset, void *data,
> + unsigned int size)
> +{
> + struct smc_lo_dmb_node *rmb_node = NULL, *tmp_node;
> + struct smc_lo_dev *ldev = smcd->priv;
> +
> + read_lock(&ldev->dmb_ht_lock);
> + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb_tok) {
> + if (tmp_node->token == dmb_tok) {
> + rmb_node = tmp_node;
> + break;
> + }
> + }
> + if (!rmb_node) {
> + read_unlock(&ldev->dmb_ht_lock);
> + return -EINVAL;
> + }
> + read_unlock(&ldev->dmb_ht_lock);
> +
> + memcpy((char *)rmb_node->cpu_addr + offset, data, size);
> +
> + if (sf) {
> + struct smc_connection *conn =
> + smcd->conn[rmb_node->sba_idx];
> +
> + if (conn && !conn->killed)
> + smcd_cdc_rx_handler(conn);
> + }
> + return 0;
> +}
> +
> static int smc_lo_supports_v2(void)
> {
> return SMC_LO_SUPPORTS_V2;
> @@ -131,14 +251,14 @@ static struct device *smc_lo_get_dev(struct smcd_dev *smcd)
>
> static const struct smcd_ops lo_ops = {
> .query_remote_gid = smc_lo_query_rgid,
> - .register_dmb = NULL,
> - .unregister_dmb = NULL,
> + .register_dmb = smc_lo_register_dmb,
> + .unregister_dmb = smc_lo_unregister_dmb,
> .add_vlan_id = smc_lo_add_vlan_id,
> .del_vlan_id = smc_lo_del_vlan_id,
> .set_vlan_required = smc_lo_set_vlan_required,
> .reset_vlan_required = smc_lo_reset_vlan_required,
> .signal_event = smc_lo_signal_event,
> - .move_data = NULL,
> + .move_data = smc_lo_move_data,
> .supports_v2 = smc_lo_supports_v2,
> .get_system_eid = smc_lo_get_system_eid,
> .get_local_gid = smc_lo_get_local_gid,
> @@ -212,6 +332,8 @@ static void smc_lo_dev_release(struct device *dev)
> static int smc_lo_dev_init(struct smc_lo_dev *ldev)
> {
> smc_lo_generate_id(ldev);
> + rwlock_init(&ldev->dmb_ht_lock);
> + hash_init(ldev->dmb_ht);
>
> return smcd_lo_register_dev(ldev);
> }
> diff --git a/net/smc/smc_loopback.h b/net/smc/smc_loopback.h
> index 2156f22..943424f 100644
> --- a/net/smc/smc_loopback.h
> +++ b/net/smc/smc_loopback.h
> @@ -20,12 +20,25 @@
>
> #define SMC_LO_CHID 0xFFFF
> #define SMC_LODEV_MAX_DMBS 5000
> +#define SMC_LODEV_DMBS_HASH_BITS 12
> +
> +struct smc_lo_dmb_node {
> + struct hlist_node list;
> + u64 token;
> + u32 len;
> + u32 sba_idx;
> + void *cpu_addr;
> + dma_addr_t dma_addr;
> +};
>
> struct smc_lo_dev {
> struct smcd_dev *smcd;
> struct device dev;
> u16 chid;
> struct smcd_gid local_gid;
> + DECLARE_BITMAP(sba_idx_mask, SMC_LODEV_MAX_DMBS);
> + rwlock_t dmb_ht_lock;
> + DECLARE_HASHTABLE(dmb_ht, SMC_LODEV_DMBS_HASH_BITS);
> };
>
> int smc_loopback_init(void);
On 2023/9/22 07:31, Wenjia Zhang wrote:
>
>
<...>
>> +static int smc_lo_register_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb,
>> + void *client_priv)
>> +{
>> + struct smc_lo_dmb_node *dmb_node, *tmp_node;
>> + struct smc_lo_dev *ldev = smcd->priv;
>> + int sba_idx, rc;
>> +
>> + /* check space for new dmb */
>> + for_each_clear_bit(sba_idx, ldev->sba_idx_mask, SMC_LODEV_MAX_DMBS) {
>> + if (!test_and_set_bit(sba_idx, ldev->sba_idx_mask))
>> + break;
>> + }
>> + if (sba_idx == SMC_LODEV_MAX_DMBS)
>> + return -ENOSPC;
>> +
>> + dmb_node = kzalloc(sizeof(*dmb_node), GFP_KERNEL);
>> + if (!dmb_node) {
>> + rc = -ENOMEM;
>> + goto err_bit;
>> + }
>> +
>> + dmb_node->sba_idx = sba_idx;
>> + dmb_node->cpu_addr = kzalloc(dmb->dmb_len, GFP_KERNEL |
>> + __GFP_NOWARN | __GFP_NORETRY |
>> + __GFP_NOMEMALLOC);
> kzalloc()/kmalloc() allocates physically contigueous memory. Are you sure it is suitable for allocating the dmb?
>
Yes, physically contigueous memory is little expensive here. I initially wanted to see the best performance.
I tried using vzalloc here, and the performance dropped a bit (2%~8%) compared to kzalloc. I think it is acceptable.
- ipc-benchmark
kzalloc vzalloc
Message
rate (msg/s) 152076 145753(-4.16%)
- sockperf
kzalloc vzalloc
Bandwidth(MBps) 8491.638 8002.380(-5.76%)
Latency(us) 3.222 3.508(+8.88%)
- nginx/wrk
kzalloc vzalloc
Requests/s 272519.36 256490.94(-5.88%)
- redis-benchmark
kzalloc vzalloc
GET(Requests/s) 123304.56 120084.05(-2.61%)
SET(Requests/s) 122062.87 118800.12(-2.67%)
>> + if (!dmb_node->cpu_addr) {
>> + rc = -ENOMEM;
>> + goto err_node;
>> + }
>> + dmb_node->len = dmb->dmb_len;
>> + dmb_node->dma_addr = (dma_addr_t)dmb_node->cpu_addr;
>> +
>> +again:
>> + /* add new dmb into hash table */
>> + get_random_bytes(&dmb_node->token, sizeof(dmb_node->token));
>> + write_lock(&ldev->dmb_ht_lock);
>> + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb_node->token) {
>> + if (tmp_node->token == dmb_node->token) {
>> + write_unlock(&ldev->dmb_ht_lock);
>> + goto again;
>> + }
>> + }
>> + hash_add(ldev->dmb_ht, &dmb_node->list, dmb_node->token);
>> + write_unlock(&ldev->dmb_ht_lock);
>> +
>> + dmb->sba_idx = dmb_node->sba_idx;
>> + dmb->dmb_tok = dmb_node->token;
>> + dmb->cpu_addr = dmb_node->cpu_addr;
>> + dmb->dma_addr = dmb_node->dma_addr;
>> + dmb->dmb_len = dmb_node->len;
>> +
>> + return 0;
>> +
>> +err_node:
>> + kfree(dmb_node);
>> +err_bit:
>> + clear_bit(sba_idx, ldev->sba_idx_mask);
>> + return rc;
>> +}
>> +
>> +static int smc_lo_unregister_dmb(struct smcd_dev *smcd, struct smcd_dmb *dmb)
>> +{
>> + struct smc_lo_dmb_node *dmb_node = NULL, *tmp_node;
>> + struct smc_lo_dev *ldev = smcd->priv;
>> +
>> + /* remove dmb from hash table */
>> + write_lock(&ldev->dmb_ht_lock);
>> + hash_for_each_possible(ldev->dmb_ht, tmp_node, list, dmb->dmb_tok) {
>> + if (tmp_node->token == dmb->dmb_tok) {
>> + dmb_node = tmp_node;
>> + break;
>> + }
>> + }
>> + if (!dmb_node) {
>> + write_unlock(&ldev->dmb_ht_lock);
>> + return -EINVAL;
>> + }
>> + hash_del(&dmb_node->list);
>> + write_unlock(&ldev->dmb_ht_lock);
>> +
>> + clear_bit(dmb_node->sba_idx, ldev->sba_idx_mask);
>> + kfree(dmb_node->cpu_addr);
>> + kfree(dmb_node);
>> +
>> + return 0;
>> +}
>> +