[dpdk-dev] [PATCH v2 38/44] event/octeontx2: add event timer arm timeout burst

[<pbhagavatula@marvell.com>]

From: Pavan Nikhilesh <pbhagavatula@marvell.com>

Add event timer arm timeout burst function.
All the timers requested to be armed have the same timeout.

Signed-off-by: Pavan Nikhilesh <pbhagavatula@marvell.com>
---
 drivers/event/octeontx2/otx2_tim_evdev.c  |  9 +++
 drivers/event/octeontx2/otx2_tim_evdev.h  | 16 ++++
 drivers/event/octeontx2/otx2_tim_worker.c | 53 ++++++++++++
 drivers/event/octeontx2/otx2_tim_worker.h | 98 +++++++++++++++++++++++
 4 files changed, 176 insertions(+)

diff --git a/drivers/event/octeontx2/otx2_tim_evdev.c b/drivers/event/octeontx2/otx2_tim_evdev.c
index f4651c281..fabcd3d0a 100644
--- a/drivers/event/octeontx2/otx2_tim_evdev.c
+++ b/drivers/event/octeontx2/otx2_tim_evdev.c
@@ -42,8 +42,17 @@ TIM_ARM_FASTPATH_MODES
 #undef FP
 	};
 
+	const rte_event_timer_arm_tmo_tick_burst_t arm_tmo_burst[2][2] = {
+#define FP(_name, _f2, _f1, flags) \
+		[_f2][_f1] = otx2_tim_arm_tmo_tick_burst_ ## _name,
+TIM_ARM_TMO_FASTPATH_MODES
+#undef FP
+	};
+
 	otx2_tim_ops.arm_burst = arm_burst[tim_ring->optimized]
 				[tim_ring->ena_dfb][prod_flag];
+	otx2_tim_ops.arm_tmo_tick_burst = arm_tmo_burst[tim_ring->optimized]
+				[tim_ring->ena_dfb];
 }
 
 static void
diff --git a/drivers/event/octeontx2/otx2_tim_evdev.h b/drivers/event/octeontx2/otx2_tim_evdev.h
index 01b271507..751659719 100644
--- a/drivers/event/octeontx2/otx2_tim_evdev.h
+++ b/drivers/event/octeontx2/otx2_tim_evdev.h
@@ -66,6 +66,8 @@
 #define OTX2_TIM_MAX_BUCKETS		(0xFFFFF)
 #define OTX2_TIM_RING_DEF_CHUNK_SZ	(4096)
 #define OTX2_TIM_CHUNK_ALIGNMENT	(16)
+#define OTX2_TIM_MAX_BURST		(RTE_CACHE_LINE_SIZE / \
+						OTX2_TIM_CHUNK_ALIGNMENT)
 #define OTX2_TIM_NB_CHUNK_SLOTS(sz)	(((sz) / OTX2_TIM_CHUNK_ALIGNMENT) - 1)
 #define OTX2_TIM_MIN_CHUNK_SLOTS	(0x1)
 #define OTX2_TIM_MAX_CHUNK_SLOTS	(0x1FFE)
@@ -175,6 +177,20 @@ uint16_t otx2_tim_arm_burst_ ## _name(					  \
 TIM_ARM_FASTPATH_MODES
 #undef FP
 
+#define TIM_ARM_TMO_FASTPATH_MODES				\
+FP(mod,       0, 0, OTX2_TIM_BKT_MOD | OTX2_TIM_ENA_DFB)	\
+FP(mod_fb,    0, 1, OTX2_TIM_BKT_MOD | OTX2_TIM_ENA_FB)		\
+FP(and,       1, 0, OTX2_TIM_BKT_AND | OTX2_TIM_ENA_DFB)	\
+FP(and_fb,    1, 1, OTX2_TIM_BKT_AND | OTX2_TIM_ENA_FB)		\
+
+#define FP(_name, _f2, _f1, flags)					\
+uint16_t otx2_tim_arm_tmo_tick_burst_ ## _name(				\
+		const struct rte_event_timer_adapter *adptr,		\
+		struct rte_event_timer **tim,				\
+		const uint64_t timeout_tick, const uint16_t nb_timers);
+TIM_ARM_TMO_FASTPATH_MODES
+#undef FP
+
 int otx2_tim_caps_get(const struct rte_eventdev *dev, uint64_t flags,
 		      uint32_t *caps,
 		      const struct rte_event_timer_adapter_ops **ops);
diff --git a/drivers/event/octeontx2/otx2_tim_worker.c b/drivers/event/octeontx2/otx2_tim_worker.c
index 409575ec4..737b167d1 100644
--- a/drivers/event/octeontx2/otx2_tim_worker.c
+++ b/drivers/event/octeontx2/otx2_tim_worker.c
@@ -72,6 +72,45 @@ tim_timer_arm_burst(const struct rte_event_timer_adapter *adptr,
 	return index;
 }
 
+static __rte_always_inline uint16_t
+tim_timer_arm_tmo_brst(const struct rte_event_timer_adapter *adptr,
+		       struct rte_event_timer **tim,
+		       const uint64_t timeout_tick,
+		       const uint16_t nb_timers, const uint8_t flags)
+{
+	struct otx2_tim_ent entry[OTX2_TIM_MAX_BURST] __rte_cache_aligned;
+	struct otx2_tim_ring *tim_ring = adptr->data->adapter_priv;
+	uint16_t set_timers = 0;
+	uint16_t arr_idx = 0;
+	uint16_t idx;
+	int ret;
+
+	if (unlikely(!timeout_tick || timeout_tick >= tim_ring->nb_bkts)) {
+		const enum rte_event_timer_state state = timeout_tick ?
+			RTE_EVENT_TIMER_ERROR_TOOLATE :
+			RTE_EVENT_TIMER_ERROR_TOOEARLY;
+		for (idx = 0; idx < nb_timers; idx++)
+			tim[idx]->state = state;
+
+		rte_errno = EINVAL;
+		return 0;
+	}
+
+	while (arr_idx < nb_timers) {
+		for (idx = 0; idx < OTX2_TIM_MAX_BURST && (arr_idx < nb_timers);
+		     idx++, arr_idx++) {
+			tim_format_event(tim[arr_idx], &entry[idx]);
+		}
+		ret = tim_add_entry_brst(tim_ring, timeout_tick,
+					 &tim[set_timers], entry, idx, flags);
+		set_timers += ret;
+		if (ret != idx)
+			break;
+	}
+
+	return set_timers;
+}
+
 #define FP(_name, _f3, _f2, _f1, _flags)				  \
 uint16_t __rte_noinline							  \
 otx2_tim_arm_burst_ ## _name(const struct rte_event_timer_adapter *adptr, \
@@ -82,3 +121,17 @@ otx2_tim_arm_burst_ ## _name(const struct rte_event_timer_adapter *adptr, \
 }
 TIM_ARM_FASTPATH_MODES
 #undef FP
+
+#define FP(_name, _f2, _f1, _flags)					\
+uint16_t __rte_noinline							\
+otx2_tim_arm_tmo_tick_burst_ ## _name(					\
+			const struct rte_event_timer_adapter *adptr,	\
+				      struct rte_event_timer **tim,	\
+				      const uint64_t timeout_tick,	\
+				      const uint16_t nb_timers)		\
+{									\
+	return tim_timer_arm_tmo_brst(adptr, tim, timeout_tick,		\
+			nb_timers, _flags);				\
+}
+TIM_ARM_TMO_FASTPATH_MODES
+#undef FP
diff --git a/drivers/event/octeontx2/otx2_tim_worker.h b/drivers/event/octeontx2/otx2_tim_worker.h
index a5e0d56bc..da8c93ff2 100644
--- a/drivers/event/octeontx2/otx2_tim_worker.h
+++ b/drivers/event/octeontx2/otx2_tim_worker.h
@@ -312,4 +312,102 @@ tim_add_entry_mp(struct otx2_tim_ring * const tim_ring,
 	return 0;
 }
 
+static inline uint16_t
+tim_cpy_wrk(uint16_t index, uint16_t cpy_lmt,
+	    struct otx2_tim_ent *chunk,
+	    struct rte_event_timer ** const tim,
+	    const struct otx2_tim_ent * const ents,
+	    const struct otx2_tim_bkt * const bkt)
+{
+	for (; index < cpy_lmt; index++) {
+		*chunk = *(ents + index);
+		tim[index]->impl_opaque[0] = (uintptr_t)chunk++;
+		tim[index]->impl_opaque[1] = (uintptr_t)bkt;
+		tim[index]->state = RTE_EVENT_TIMER_ARMED;
+	}
+
+	return index;
+}
+
+/* Burst mode functions */
+static inline int
+tim_add_entry_brst(struct otx2_tim_ring * const tim_ring,
+		   const uint16_t rel_bkt,
+		   struct rte_event_timer ** const tim,
+		   const struct otx2_tim_ent *ents,
+		   const uint16_t nb_timers, const uint8_t flags)
+{
+	struct otx2_tim_ent *chunk;
+	struct otx2_tim_bkt *bkt;
+	uint16_t chunk_remainder;
+	uint16_t index = 0;
+	uint64_t lock_sema;
+	int16_t rem, crem;
+	uint8_t lock_cnt;
+
+__retry:
+	bkt = tim_get_target_bucket(tim_ring, rel_bkt, flags);
+
+	/* Only one thread beyond this. */
+	lock_sema = tim_bkt_inc_lock(bkt);
+	lock_cnt = (uint8_t)
+		((lock_sema >> TIM_BUCKET_W1_S_LOCK) & TIM_BUCKET_W1_M_LOCK);
+
+	if (lock_cnt) {
+		tim_bkt_dec_lock(bkt);
+		goto __retry;
+	}
+
+	/* Bucket related checks. */
+	if (unlikely(tim_bkt_get_hbt(lock_sema))) {
+		tim_bkt_dec_lock(bkt);
+		goto __retry;
+	}
+
+	chunk_remainder = tim_bkt_fetch_rem(lock_sema);
+	rem = chunk_remainder - nb_timers;
+	if (rem < 0) {
+		crem = tim_ring->nb_chunk_slots - chunk_remainder;
+		if (chunk_remainder && crem) {
+			chunk = ((struct otx2_tim_ent *)
+					(uintptr_t)bkt->current_chunk) + crem;
+
+			index = tim_cpy_wrk(index, chunk_remainder, chunk, tim,
+					    ents, bkt);
+			tim_bkt_sub_rem(bkt, chunk_remainder);
+			tim_bkt_add_nent(bkt, chunk_remainder);
+		}
+
+		if (flags & OTX2_TIM_ENA_FB)
+			chunk = tim_refill_chunk(bkt, tim_ring);
+		if (flags & OTX2_TIM_ENA_DFB)
+			chunk = tim_insert_chunk(bkt, tim_ring);
+
+		if (unlikely(chunk == NULL)) {
+			tim_bkt_dec_lock(bkt);
+			rte_errno = ENOMEM;
+			tim[index]->state = RTE_EVENT_TIMER_ERROR;
+			return crem;
+		}
+		*(uint64_t *)(chunk + tim_ring->nb_chunk_slots) = 0;
+		bkt->current_chunk = (uintptr_t)chunk;
+		tim_cpy_wrk(index, nb_timers, chunk, tim, ents, bkt);
+
+		rem = nb_timers - chunk_remainder;
+		tim_bkt_set_rem(bkt, tim_ring->nb_chunk_slots - rem);
+		tim_bkt_add_nent(bkt, rem);
+	} else {
+		chunk = (struct otx2_tim_ent *)(uintptr_t)bkt->current_chunk;
+		chunk += (tim_ring->nb_chunk_slots - chunk_remainder);
+
+		tim_cpy_wrk(index, nb_timers, chunk, tim, ents, bkt);
+		tim_bkt_sub_rem(bkt, nb_timers);
+		tim_bkt_add_nent(bkt, nb_timers);
+	}
+
+	tim_bkt_dec_lock(bkt);
+
+	return nb_timers;
+}
+
 #endif /* __OTX2_TIM_WORKER_H__ */
-- 
2.22.0