d/numam-dpdk
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

net/sfc: make refill threshold check Rx datapath specific

Browse Source 
EFX_RXQ_LIMIT is libefx-specifics and it should not be used
for other Rx datapaths implementations (e.g. EF10 native).

EF10 native Rx datapath has its own understanding of the maximum
RxQ fill level imposed by EvQ clear strategy and space reserved
for Rx error and flush events.

Signed-off-by: Andrew Rybchenko <arybchenko@solarflare.com>
Reviewed-by: Andy Moreton <amoreton@solarflare.com>
This commit is contained in:
Andrew Rybchenko 2018-01-09 20:24:50 +00:00 committed by Ferruh Yigit
parent d6bde6b5ea
commit f7da270aea
3 changed files with 63 additions and 9 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

17
drivers/net/sfc/sfc_dp_rx.h
View File

@ -100,6 +100,22 @@ struct sfc_dp_rx_qcreate_info {
volatile void *mem_bar;
};
/**
* Get size of receive and event queue rings by the number of Rx
* descriptors.
*
* @param nb_rx_desc Number of Rx descriptors
* @param rxq_entries Location for number of Rx ring entries
* @param evq_entries Location for number of event ring entries
* @param rxq_max_fill_level Location for maximum Rx ring fill level
*
* @return 0 or positive errno.
*/
typedef int (sfc_dp_rx_qsize_up_rings_t)(uint16_t nb_rx_desc,
unsigned int *rxq_entries,
unsigned int *evq_entries,
unsigned int *rxq_max_fill_level);
/**
* Allocate and initialize datapath receive queue.
*
@ -168,6 +184,7 @@ struct sfc_dp_rx {
#define SFC_DP_RX_FEAT_SCATTER 0x1
#define SFC_DP_RX_FEAT_MULTI_PROCESS 0x2
#define SFC_DP_RX_FEAT_TUNNELS 0x4
sfc_dp_rx_qsize_up_rings_t *qsize_up_rings;
sfc_dp_rx_qcreate_t *qcreate;
sfc_dp_rx_qdestroy_t *qdestroy;
sfc_dp_rx_qstart_t *qstart;

15
drivers/net/sfc/sfc_ef10_rx.c
View File

@ -638,6 +638,20 @@ sfc_ef10_rx_qdesc_status(__rte_unused struct sfc_dp_rxq *dp_rxq,
}
static sfc_dp_rx_qsize_up_rings_t sfc_ef10_rx_qsize_up_rings;
static int
sfc_ef10_rx_qsize_up_rings(uint16_t nb_rx_desc,
unsigned int *rxq_entries,
unsigned int *evq_entries,
unsigned int *rxq_max_fill_level)
{
*rxq_entries = nb_rx_desc;
*evq_entries = nb_rx_desc;
*rxq_max_fill_level = SFC_EF10_RXQ_LIMIT(*rxq_entries);
return 0;
}
static uint64_t
sfc_ef10_mk_mbuf_rearm_data(uint16_t port_id, uint16_t prefix_size)
{
@ -794,6 +808,7 @@ struct sfc_dp_rx sfc_ef10_rx = {
},
.features = SFC_DP_RX_FEAT_MULTI_PROCESS |
SFC_DP_RX_FEAT_TUNNELS,
.qsize_up_rings = sfc_ef10_rx_qsize_up_rings,
.qcreate = sfc_ef10_rx_qcreate,
.qdestroy = sfc_ef10_rx_qdestroy,
.qstart = sfc_ef10_rx_qstart,

40
drivers/net/sfc/sfc_rx.c
View File

@ -413,6 +413,19 @@ sfc_rxq_by_dp_rxq(const struct sfc_dp_rxq *dp_rxq)
return rxq;
}
static sfc_dp_rx_qsize_up_rings_t sfc_efx_rx_qsize_up_rings;
static int
sfc_efx_rx_qsize_up_rings(uint16_t nb_rx_desc,
unsigned int *rxq_entries,
unsigned int *evq_entries,
unsigned int *rxq_max_fill_level)
{
*rxq_entries = nb_rx_desc;
*evq_entries = nb_rx_desc;
*rxq_max_fill_level = EFX_RXQ_LIMIT(*rxq_entries);
return 0;
}
static sfc_dp_rx_qcreate_t sfc_efx_rx_qcreate;
static int
sfc_efx_rx_qcreate(uint16_t port_id, uint16_t queue_id,
@ -535,6 +548,7 @@ struct sfc_dp_rx sfc_efx_rx = {
.hw_fw_caps = 0,
},
.features = SFC_DP_RX_FEAT_SCATTER,
.qsize_up_rings = sfc_efx_rx_qsize_up_rings,
.qcreate = sfc_efx_rx_qcreate,
.qdestroy = sfc_efx_rx_qdestroy,
.qstart = sfc_efx_rx_qstart,
@ -771,10 +785,9 @@ sfc_rx_qstop(struct sfc_adapter *sa, unsigned int sw_index)
}
static int
sfc_rx_qcheck_conf(struct sfc_adapter *sa, uint16_t nb_rx_desc,
sfc_rx_qcheck_conf(struct sfc_adapter *sa, unsigned int rxq_max_fill_level,
const struct rte_eth_rxconf *rx_conf)
{
const uint16_t rx_free_thresh_max = EFX_RXQ_LIMIT(nb_rx_desc);
int rc = 0;
if (rx_conf->rx_thresh.pthresh != 0 ||
@ -784,10 +797,10 @@ sfc_rx_qcheck_conf(struct sfc_adapter *sa, uint16_t nb_rx_desc,
"RxQ prefetch/host/writeback thresholds are not supported");
}
if (rx_conf->rx_free_thresh > rx_free_thresh_max) {
if (rx_conf->rx_free_thresh > rxq_max_fill_level) {
sfc_err(sa,
"RxQ free threshold too large: %u vs maximum %u",
rx_conf->rx_free_thresh, rx_free_thresh_max);
rx_conf->rx_free_thresh, rxq_max_fill_level);
rc = EINVAL;
}
@ -907,13 +920,21 @@ sfc_rx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
{
const efx_nic_cfg_t *encp = efx_nic_cfg_get(sa->nic);
int rc;
unsigned int rxq_entries;
unsigned int evq_entries;
unsigned int rxq_max_fill_level;
uint16_t buf_size;
struct sfc_rxq_info *rxq_info;
struct sfc_evq *evq;
struct sfc_rxq *rxq;
struct sfc_dp_rx_qcreate_info info;
rc = sfc_rx_qcheck_conf(sa, nb_rx_desc, rx_conf);
rc = sa->dp_rx->qsize_up_rings(nb_rx_desc, &rxq_entries, &evq_entries,
&rxq_max_fill_level);
if (rc != 0)
goto fail_size_up_rings;
rc = sfc_rx_qcheck_conf(sa, rxq_max_fill_level, rx_conf);
if (rc != 0)
goto fail_bad_conf;
@ -940,8 +961,8 @@ sfc_rx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
SFC_ASSERT(sw_index < sa->rxq_count);
rxq_info = &sa->rxq_info[sw_index];
SFC_ASSERT(nb_rx_desc <= rxq_info->max_entries);
rxq_info->entries = nb_rx_desc;
SFC_ASSERT(rxq_entries <= rxq_info->max_entries);
rxq_info->entries = rxq_entries;
rxq_info->type = EFX_RXQ_TYPE_DEFAULT;
rxq_info->type_flags =
sa->eth_dev->data->dev_conf.rxmode.enable_scatter ?
@ -952,7 +973,7 @@ sfc_rx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
rxq_info->type_flags |= EFX_RXQ_FLAG_INNER_CLASSES;
rc = sfc_ev_qinit(sa, SFC_EVQ_TYPE_RX, sw_index,
rxq_info->entries, socket_id, &evq);
evq_entries, socket_id, &evq);
if (rc != 0)
goto fail_ev_qinit;
@ -989,7 +1010,7 @@ sfc_rx_qinit(struct sfc_adapter *sa, unsigned int sw_index,
info.rxq_entries = rxq_info->entries;
info.rxq_hw_ring = rxq->mem.esm_base;
info.evq_entries = rxq_info->entries;
info.evq_entries = evq_entries;
info.evq_hw_ring = evq->mem.esm_base;
info.hw_index = rxq->hw_index;
info.mem_bar = sa->mem_bar.esb_base;
@ -1022,6 +1043,7 @@ fail_ev_qinit:
rxq_info->entries = 0;
fail_bad_conf:
fail_size_up_rings:
sfc_log_init(sa, "failed %d", rc);
return rc;
}