e_RTE_SCHED_SUBPORT_ARRAY_TOTAL,
};
-#ifdef RTE_SCHED_COLLECT_STATS
-
-static inline uint32_t
-rte_sched_port_queues_per_subport(struct rte_sched_port *port)
-{
- return RTE_SCHED_QUEUES_PER_PIPE * port->n_pipes_per_subport;
-}
-
-#endif
-
static inline uint32_t
rte_sched_subport_pipe_queues(struct rte_sched_subport *subport)
{
uint32_t queue)
{
return ((subport & (port->n_subports_per_port - 1)) <<
- (port->n_pipes_per_subport_log2 + 4)) |
- ((pipe & (port->n_pipes_per_subport - 1)) << 4) |
- ((rte_sched_port_pipe_queue(port, traffic_class) + queue) &
- (RTE_SCHED_QUEUES_PER_PIPE - 1));
+ (port->n_pipes_per_subport_log2 + 4)) |
+ ((pipe &
+ (port->subports[subport]->n_pipes_per_subport_enabled - 1)) << 4) |
+ ((rte_sched_port_pipe_queue(port, traffic_class) + queue) &
+ (RTE_SCHED_QUEUES_PER_PIPE - 1));
}
void
uint32_t queue_id = rte_mbuf_sched_queue_get(pkt);
*subport = queue_id >> (port->n_pipes_per_subport_log2 + 4);
- *pipe = (queue_id >> 4) & (port->n_pipes_per_subport - 1);
+ *pipe = (queue_id >> 4) &
+ (port->subports[*subport]->n_pipes_per_subport_enabled - 1);
*traffic_class = rte_sched_port_pipe_tc(port, queue_id);
*queue = rte_sched_port_tc_queue(port, queue_id);
}
return -EINVAL;
}
- s = port->subport + subport_id;
+ s = port->subports[subport_id];
/* Copy subport stats and clear */
memcpy(stats, &s->stats, sizeof(struct rte_sched_subport_stats));
#ifdef RTE_SCHED_COLLECT_STATS
static inline void
-rte_sched_port_update_subport_stats(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt)
+rte_sched_port_update_subport_stats(struct rte_sched_port *port,
+ struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt)
{
- struct rte_sched_subport *s = port->subport + (qindex / rte_sched_port_queues_per_subport(port));
uint32_t tc_index = rte_sched_port_pipe_tc(port, qindex);
uint32_t pkt_len = pkt->pkt_len;
- s->stats.n_pkts_tc[tc_index] += 1;
- s->stats.n_bytes_tc[tc_index] += pkt_len;
+ subport->stats.n_pkts_tc[tc_index] += 1;
+ subport->stats.n_bytes_tc[tc_index] += pkt_len;
}
#ifdef RTE_SCHED_RED
static inline void
rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port,
- uint32_t qindex,
- struct rte_mbuf *pkt, uint32_t red)
+ struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt,
+ uint32_t red)
#else
static inline void
rte_sched_port_update_subport_stats_on_drop(struct rte_sched_port *port,
- uint32_t qindex,
- struct rte_mbuf *pkt, __rte_unused uint32_t red)
+ struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt,
+ __rte_unused uint32_t red)
#endif
{
- struct rte_sched_subport *s = port->subport + (qindex / rte_sched_port_queues_per_subport(port));
uint32_t tc_index = rte_sched_port_pipe_tc(port, qindex);
uint32_t pkt_len = pkt->pkt_len;
- s->stats.n_pkts_tc_dropped[tc_index] += 1;
- s->stats.n_bytes_tc_dropped[tc_index] += pkt_len;
+ subport->stats.n_pkts_tc_dropped[tc_index] += 1;
+ subport->stats.n_bytes_tc_dropped[tc_index] += pkt_len;
#ifdef RTE_SCHED_RED
- s->stats.n_pkts_red_dropped[tc_index] += red;
+ subport->stats.n_pkts_red_dropped[tc_index] += red;
#endif
}
static inline void
-rte_sched_port_update_queue_stats(struct rte_sched_port *port, uint32_t qindex, struct rte_mbuf *pkt)
+rte_sched_port_update_queue_stats(struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt)
{
- struct rte_sched_queue_extra *qe = port->queue_extra + qindex;
+ struct rte_sched_queue_extra *qe = subport->queue_extra + qindex;
uint32_t pkt_len = pkt->pkt_len;
qe->stats.n_pkts += 1;
#ifdef RTE_SCHED_RED
static inline void
-rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port,
- uint32_t qindex,
- struct rte_mbuf *pkt, uint32_t red)
+rte_sched_port_update_queue_stats_on_drop(struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt,
+ uint32_t red)
#else
static inline void
-rte_sched_port_update_queue_stats_on_drop(struct rte_sched_port *port,
- uint32_t qindex,
- struct rte_mbuf *pkt, __rte_unused uint32_t red)
+rte_sched_port_update_queue_stats_on_drop(struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf *pkt,
+ __rte_unused uint32_t red)
#endif
{
- struct rte_sched_queue_extra *qe = port->queue_extra + qindex;
+ struct rte_sched_queue_extra *qe = subport->queue_extra + qindex;
uint32_t pkt_len = pkt->pkt_len;
qe->stats.n_pkts_dropped += 1;
#ifdef RTE_SCHED_RED
static inline int
-rte_sched_port_red_drop(struct rte_sched_port *port, struct rte_mbuf *pkt, uint32_t qindex, uint16_t qlen)
+rte_sched_port_red_drop(struct rte_sched_port *port,
+ struct rte_sched_subport *subport,
+ struct rte_mbuf *pkt,
+ uint32_t qindex,
+ uint16_t qlen)
{
struct rte_sched_queue_extra *qe;
struct rte_red_config *red_cfg;
tc_index = rte_sched_port_pipe_tc(port, qindex);
color = rte_sched_port_pkt_read_color(pkt);
- red_cfg = &port->red_config[tc_index][color];
+ red_cfg = &subport->red_config[tc_index][color];
if ((red_cfg->min_th | red_cfg->max_th) == 0)
return 0;
- qe = port->queue_extra + qindex;
+ qe = subport->queue_extra + qindex;
red = &qe->red;
return rte_red_enqueue(red_cfg, red, qlen, port->time);
#else
-#define rte_sched_port_red_drop(port, pkt, qindex, qlen) 0
+static inline int rte_sched_port_red_drop(struct rte_sched_port *port __rte_unused,
+ struct rte_sched_subport *subport __rte_unused,
+ struct rte_mbuf *pkt __rte_unused,
+ uint32_t qindex __rte_unused,
+ uint16_t qlen __rte_unused)
+{
+ return 0;
+}
#define rte_sched_port_set_queue_empty_timestamp(port, qindex)
#endif /* RTE_SCHED_DEBUG */
+static inline struct rte_sched_subport *
+rte_sched_port_subport(struct rte_sched_port *port,
+ struct rte_mbuf *pkt)
+{
+ uint32_t queue_id = rte_mbuf_sched_queue_get(pkt);
+ uint32_t subport_id = queue_id >> (port->n_pipes_per_subport_log2 + 4);
+
+ return port->subports[subport_id];
+}
+
static inline uint32_t
-rte_sched_port_enqueue_qptrs_prefetch0(struct rte_sched_port *port,
- struct rte_mbuf *pkt)
+rte_sched_port_enqueue_qptrs_prefetch0(struct rte_sched_subport *subport,
+ struct rte_mbuf *pkt, uint32_t subport_qmask)
{
struct rte_sched_queue *q;
#ifdef RTE_SCHED_COLLECT_STATS
struct rte_sched_queue_extra *qe;
#endif
uint32_t qindex = rte_mbuf_sched_queue_get(pkt);
+ uint32_t subport_queue_id = subport_qmask & qindex;
- q = port->queue + qindex;
+ q = subport->queue + subport_queue_id;
rte_prefetch0(q);
#ifdef RTE_SCHED_COLLECT_STATS
- qe = port->queue_extra + qindex;
+ qe = subport->queue_extra + subport_queue_id;
rte_prefetch0(qe);
#endif
- return qindex;
+ return subport_queue_id;
}
static inline void
rte_sched_port_enqueue_qwa_prefetch0(struct rte_sched_port *port,
- uint32_t qindex, struct rte_mbuf **qbase)
+ struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf **qbase)
{
struct rte_sched_queue *q;
struct rte_mbuf **q_qw;
uint16_t qsize;
- q = port->queue + qindex;
- qsize = rte_sched_port_qsize(port, qindex);
+ q = subport->queue + qindex;
+ qsize = rte_sched_subport_pipe_qsize(port, subport, qindex);
q_qw = qbase + (q->qw & (qsize - 1));
rte_prefetch0(q_qw);
- rte_bitmap_prefetch0(port->bmp, qindex);
+ rte_bitmap_prefetch0(subport->bmp, qindex);
}
static inline int
-rte_sched_port_enqueue_qwa(struct rte_sched_port *port, uint32_t qindex,
- struct rte_mbuf **qbase, struct rte_mbuf *pkt)
+rte_sched_port_enqueue_qwa(struct rte_sched_port *port,
+ struct rte_sched_subport *subport,
+ uint32_t qindex,
+ struct rte_mbuf **qbase,
+ struct rte_mbuf *pkt)
{
struct rte_sched_queue *q;
uint16_t qsize;
uint16_t qlen;
- q = port->queue + qindex;
- qsize = rte_sched_port_qsize(port, qindex);
+ q = subport->queue + qindex;
+ qsize = rte_sched_subport_pipe_qsize(port, subport, qindex);
qlen = q->qw - q->qr;
/* Drop the packet (and update drop stats) when queue is full */
- if (unlikely(rte_sched_port_red_drop(port, pkt, qindex, qlen) ||
+ if (unlikely(rte_sched_port_red_drop(port, subport, pkt, qindex, qlen) ||
(qlen >= qsize))) {
rte_pktmbuf_free(pkt);
#ifdef RTE_SCHED_COLLECT_STATS
- rte_sched_port_update_subport_stats_on_drop(port, qindex, pkt,
- qlen < qsize);
- rte_sched_port_update_queue_stats_on_drop(port, qindex, pkt,
- qlen < qsize);
+ rte_sched_port_update_subport_stats_on_drop(port, subport,
+ qindex, pkt, qlen < qsize);
+ rte_sched_port_update_queue_stats_on_drop(subport, qindex, pkt,
+ qlen < qsize);
#endif
return 0;
}
qbase[q->qw & (qsize - 1)] = pkt;
q->qw++;
- /* Activate queue in the port bitmap */
- rte_bitmap_set(port->bmp, qindex);
+ /* Activate queue in the subport bitmap */
+ rte_bitmap_set(subport->bmp, qindex);
/* Statistics */
#ifdef RTE_SCHED_COLLECT_STATS
- rte_sched_port_update_subport_stats(port, qindex, pkt);
- rte_sched_port_update_queue_stats(port, qindex, pkt);
+ rte_sched_port_update_subport_stats(port, subport, qindex, pkt);
+ rte_sched_port_update_queue_stats(subport, qindex, pkt);
#endif
return 1;
*pkt30, *pkt31, *pkt_last;
struct rte_mbuf **q00_base, **q01_base, **q10_base, **q11_base,
**q20_base, **q21_base, **q30_base, **q31_base, **q_last_base;
+ struct rte_sched_subport *subport00, *subport01, *subport10, *subport11,
+ *subport20, *subport21, *subport30, *subport31, *subport_last;
uint32_t q00, q01, q10, q11, q20, q21, q30, q31, q_last;
uint32_t r00, r01, r10, r11, r20, r21, r30, r31, r_last;
+ uint32_t subport_qmask;
uint32_t result, i;
result = 0;
+ subport_qmask = (1 << (port->n_pipes_per_subport_log2 + 4)) - 1;
/*
* Less then 6 input packets available, which is not enough to
* feed the pipeline
*/
if (unlikely(n_pkts < 6)) {
+ struct rte_sched_subport *subports[5];
struct rte_mbuf **q_base[5];
uint32_t q[5];
for (i = 0; i < n_pkts; i++)
rte_prefetch0(pkts[i]);
+ /* Prefetch the subport structure for each packet */
+ for (i = 0; i < n_pkts; i++)
+ subports[i] = rte_sched_port_subport(port, pkts[i]);
+
/* Prefetch the queue structure for each queue */
for (i = 0; i < n_pkts; i++)
- q[i] = rte_sched_port_enqueue_qptrs_prefetch0(port,
- pkts[i]);
+ q[i] = rte_sched_port_enqueue_qptrs_prefetch0(subports[i],
+ pkts[i], subport_qmask);
/* Prefetch the write pointer location of each queue */
for (i = 0; i < n_pkts; i++) {
- q_base[i] = rte_sched_port_qbase(port, q[i]);
- rte_sched_port_enqueue_qwa_prefetch0(port, q[i],
- q_base[i]);
+ q_base[i] = rte_sched_subport_pipe_qbase(subports[i], q[i]);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subports[i],
+ q[i], q_base[i]);
}
/* Write each packet to its queue */
for (i = 0; i < n_pkts; i++)
- result += rte_sched_port_enqueue_qwa(port, q[i],
- q_base[i], pkts[i]);
+ result += rte_sched_port_enqueue_qwa(port, subports[i],
+ q[i], q_base[i], pkts[i]);
return result;
}
rte_prefetch0(pkt10);
rte_prefetch0(pkt11);
- q20 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt20);
- q21 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt21);
+ subport20 = rte_sched_port_subport(port, pkt20);
+ subport21 = rte_sched_port_subport(port, pkt21);
+ q20 = rte_sched_port_enqueue_qptrs_prefetch0(subport20,
+ pkt20, subport_qmask);
+ q21 = rte_sched_port_enqueue_qptrs_prefetch0(subport21,
+ pkt21, subport_qmask);
pkt00 = pkts[4];
pkt01 = pkts[5];
rte_prefetch0(pkt00);
rte_prefetch0(pkt01);
- q10 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt10);
- q11 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt11);
+ subport10 = rte_sched_port_subport(port, pkt10);
+ subport11 = rte_sched_port_subport(port, pkt11);
+ q10 = rte_sched_port_enqueue_qptrs_prefetch0(subport10,
+ pkt10, subport_qmask);
+ q11 = rte_sched_port_enqueue_qptrs_prefetch0(subport11,
+ pkt11, subport_qmask);
- q20_base = rte_sched_port_qbase(port, q20);
- q21_base = rte_sched_port_qbase(port, q21);
- rte_sched_port_enqueue_qwa_prefetch0(port, q20, q20_base);
- rte_sched_port_enqueue_qwa_prefetch0(port, q21, q21_base);
+ q20_base = rte_sched_subport_pipe_qbase(subport20, q20);
+ q21_base = rte_sched_subport_pipe_qbase(subport21, q21);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport20, q20, q20_base);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport21, q21, q21_base);
/* Run the pipeline */
for (i = 6; i < (n_pkts & (~1)); i += 2) {
q31 = q21;
q20 = q10;
q21 = q11;
+ subport30 = subport20;
+ subport31 = subport21;
+ subport20 = subport10;
+ subport21 = subport11;
q30_base = q20_base;
q31_base = q21_base;
rte_prefetch0(pkt00);
rte_prefetch0(pkt01);
- /* Stage 1: Prefetch queue structure storing queue pointers */
- q10 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt10);
- q11 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt11);
+ /* Stage 1: Prefetch subport and queue structure storing queue pointers */
+ subport10 = rte_sched_port_subport(port, pkt10);
+ subport11 = rte_sched_port_subport(port, pkt11);
+ q10 = rte_sched_port_enqueue_qptrs_prefetch0(subport10,
+ pkt10, subport_qmask);
+ q11 = rte_sched_port_enqueue_qptrs_prefetch0(subport11,
+ pkt11, subport_qmask);
/* Stage 2: Prefetch queue write location */
- q20_base = rte_sched_port_qbase(port, q20);
- q21_base = rte_sched_port_qbase(port, q21);
- rte_sched_port_enqueue_qwa_prefetch0(port, q20, q20_base);
- rte_sched_port_enqueue_qwa_prefetch0(port, q21, q21_base);
+ q20_base = rte_sched_subport_pipe_qbase(subport20, q20);
+ q21_base = rte_sched_subport_pipe_qbase(subport21, q21);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport20, q20, q20_base);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport21, q21, q21_base);
/* Stage 3: Write packet to queue and activate queue */
- r30 = rte_sched_port_enqueue_qwa(port, q30, q30_base, pkt30);
- r31 = rte_sched_port_enqueue_qwa(port, q31, q31_base, pkt31);
+ r30 = rte_sched_port_enqueue_qwa(port, subport30,
+ q30, q30_base, pkt30);
+ r31 = rte_sched_port_enqueue_qwa(port, subport31,
+ q31, q31_base, pkt31);
result += r30 + r31;
}
pkt_last = pkts[n_pkts - 1];
rte_prefetch0(pkt_last);
- q00 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt00);
- q01 = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt01);
-
- q10_base = rte_sched_port_qbase(port, q10);
- q11_base = rte_sched_port_qbase(port, q11);
- rte_sched_port_enqueue_qwa_prefetch0(port, q10, q10_base);
- rte_sched_port_enqueue_qwa_prefetch0(port, q11, q11_base);
-
- r20 = rte_sched_port_enqueue_qwa(port, q20, q20_base, pkt20);
- r21 = rte_sched_port_enqueue_qwa(port, q21, q21_base, pkt21);
+ subport00 = rte_sched_port_subport(port, pkt00);
+ subport01 = rte_sched_port_subport(port, pkt01);
+ q00 = rte_sched_port_enqueue_qptrs_prefetch0(subport00,
+ pkt00, subport_qmask);
+ q01 = rte_sched_port_enqueue_qptrs_prefetch0(subport01,
+ pkt01, subport_qmask);
+
+ q10_base = rte_sched_subport_pipe_qbase(subport10, q10);
+ q11_base = rte_sched_subport_pipe_qbase(subport11, q11);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport10, q10, q10_base);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport11, q11, q11_base);
+
+ r20 = rte_sched_port_enqueue_qwa(port, subport20,
+ q20, q20_base, pkt20);
+ r21 = rte_sched_port_enqueue_qwa(port, subport21,
+ q21, q21_base, pkt21);
result += r20 + r21;
- q_last = rte_sched_port_enqueue_qptrs_prefetch0(port, pkt_last);
+ subport_last = rte_sched_port_subport(port, pkt_last);
+ q_last = rte_sched_port_enqueue_qptrs_prefetch0(subport_last,
+ pkt_last, subport_qmask);
- q00_base = rte_sched_port_qbase(port, q00);
- q01_base = rte_sched_port_qbase(port, q01);
- rte_sched_port_enqueue_qwa_prefetch0(port, q00, q00_base);
- rte_sched_port_enqueue_qwa_prefetch0(port, q01, q01_base);
+ q00_base = rte_sched_subport_pipe_qbase(subport00, q00);
+ q01_base = rte_sched_subport_pipe_qbase(subport01, q01);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport00, q00, q00_base);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport01, q01, q01_base);
- r10 = rte_sched_port_enqueue_qwa(port, q10, q10_base, pkt10);
- r11 = rte_sched_port_enqueue_qwa(port, q11, q11_base, pkt11);
+ r10 = rte_sched_port_enqueue_qwa(port, subport10, q10,
+ q10_base, pkt10);
+ r11 = rte_sched_port_enqueue_qwa(port, subport11, q11,
+ q11_base, pkt11);
result += r10 + r11;
- q_last_base = rte_sched_port_qbase(port, q_last);
- rte_sched_port_enqueue_qwa_prefetch0(port, q_last, q_last_base);
+ q_last_base = rte_sched_subport_pipe_qbase(subport_last, q_last);
+ rte_sched_port_enqueue_qwa_prefetch0(port, subport_last,
+ q_last, q_last_base);
- r00 = rte_sched_port_enqueue_qwa(port, q00, q00_base, pkt00);
- r01 = rte_sched_port_enqueue_qwa(port, q01, q01_base, pkt01);
+ r00 = rte_sched_port_enqueue_qwa(port, subport00, q00,
+ q00_base, pkt00);
+ r01 = rte_sched_port_enqueue_qwa(port, subport01, q01,
+ q01_base, pkt01);
result += r00 + r01;
if (n_pkts & 1) {
- r_last = rte_sched_port_enqueue_qwa(port, q_last, q_last_base, pkt_last);
+ r_last = rte_sched_port_enqueue_qwa(port, subport_last,
+ q_last, q_last_base, pkt_last);
result += r_last;
}