[net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

[net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

[Suman Ghosh]

Added support for adaptive IRQ coalescing. It uses net_dim
algorithm to find the suitable delay/IRQ count based on the
current packet rate.

Signed-off-by: Suman Ghosh <sumang@marvell.com>
Reviewed-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
---
 .../net/ethernet/marvell/octeontx2/Kconfig    |  1 +
 .../marvell/octeontx2/nic/otx2_common.c       |  5 ---
 .../marvell/octeontx2/nic/otx2_common.h       | 10 ++++++
 .../marvell/octeontx2/nic/otx2_ethtool.c      | 36 +++++++++++++++++--
 .../ethernet/marvell/octeontx2/nic/otx2_pf.c  | 22 ++++++++++++
 .../marvell/octeontx2/nic/otx2_txrx.c         | 25 ++++++++++++-
 .../marvell/octeontx2/nic/otx2_txrx.h         |  1 +
 7 files changed, 92 insertions(+), 8 deletions(-)

diff --git a/drivers/net/ethernet/marvell/octeontx2/Kconfig b/drivers/net/ethernet/marvell/octeontx2/Kconfig
index 639893d87055..e1036b0eb6b1 100644
--- a/drivers/net/ethernet/marvell/octeontx2/Kconfig
+++ b/drivers/net/ethernet/marvell/octeontx2/Kconfig
@@ -33,6 +33,7 @@ config OCTEONTX2_PF
 	select OCTEONTX2_MBOX
 	select NET_DEVLINK
 	depends on (64BIT && COMPILE_TEST) || ARM64
+	select DIMLIB
 	depends on PCI
 	depends on PTP_1588_CLOCK_OPTIONAL
 	help
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c
index b9d7601138ca..fb8db5888d2f 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.c
@@ -97,11 +97,6 @@ void otx2_get_dev_stats(struct otx2_nic *pfvf)
 {
 	struct otx2_dev_stats *dev_stats = &pfvf->hw.dev_stats;
 
-#define OTX2_GET_RX_STATS(reg) \
-	 otx2_read64(pfvf, NIX_LF_RX_STATX(reg))
-#define OTX2_GET_TX_STATS(reg) \
-	 otx2_read64(pfvf, NIX_LF_TX_STATX(reg))
-
 	dev_stats->rx_bytes = OTX2_GET_RX_STATS(RX_OCTS);
 	dev_stats->rx_drops = OTX2_GET_RX_STATS(RX_DROP);
 	dev_stats->rx_bcast_frames = OTX2_GET_RX_STATS(RX_BCAST);
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h
index c587c14ac2a3..ce2766317c0b 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_common.h
@@ -18,6 +18,7 @@
 #include <net/pkt_cls.h>
 #include <net/devlink.h>
 #include <linux/time64.h>
+#include <linux/dim.h>
 
 #include <mbox.h>
 #include <npc.h>
@@ -54,6 +55,11 @@ enum arua_mapped_qtypes {
 /* Send skid of 2000 packets required for CQ size of 4K CQEs. */
 #define SEND_CQ_SKID	2000
 
+#define OTX2_GET_RX_STATS(reg) \
+	otx2_read64(pfvf, NIX_LF_RX_STATX(reg))
+#define OTX2_GET_TX_STATS(reg) \
+	otx2_read64(pfvf, NIX_LF_TX_STATX(reg))
+
 struct otx2_lmt_info {
 	u64 lmt_addr;
 	u16 lmt_id;
@@ -351,6 +357,7 @@ struct otx2_nic {
 #define OTX2_FLAG_TC_MATCHALL_EGRESS_ENABLED	BIT_ULL(12)
 #define OTX2_FLAG_TC_MATCHALL_INGRESS_ENABLED	BIT_ULL(13)
 #define OTX2_FLAG_DMACFLTR_SUPPORT		BIT_ULL(14)
+#define OTX2_FLAG_ADPTV_INT_COAL_ENABLED BIT_ULL(16)
 	u64			flags;
 	u64			*cq_op_addr;
 
@@ -408,6 +415,9 @@ struct otx2_nic {
 	u8			pfc_en;
 	u8			*queue_to_pfc_map;
 #endif
+
+	/* napi event count. It is needed for adaptive irq coalescing. */
+	u32 napi_events;
 };
 
 static inline bool is_otx2_lbkvf(struct pci_dev *pdev)
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
index fc328de5345e..6d51e2921cf3 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_ethtool.c
@@ -455,6 +455,14 @@ static int otx2_get_coalesce(struct net_device *netdev,
 	cmd->rx_max_coalesced_frames = hw->cq_ecount_wait;
 	cmd->tx_coalesce_usecs = hw->cq_time_wait;
 	cmd->tx_max_coalesced_frames = hw->cq_ecount_wait;
+	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
+			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
+		cmd->use_adaptive_rx_coalesce = 1;
+		cmd->use_adaptive_tx_coalesce = 1;
+	} else {
+		cmd->use_adaptive_rx_coalesce = 0;
+		cmd->use_adaptive_tx_coalesce = 0;
+	}
 
 	return 0;
 }
@@ -466,11 +474,24 @@ static int otx2_set_coalesce(struct net_device *netdev,
 {
 	struct otx2_nic *pfvf = netdev_priv(netdev);
 	struct otx2_hw *hw = &pfvf->hw;
+	u8 priv_coalesce_status;
 	int qidx;
 
 	if (!ec->rx_max_coalesced_frames || !ec->tx_max_coalesced_frames)
 		return 0;
 
+	/* Check and update coalesce status */
+	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
+			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
+		priv_coalesce_status = 1;
+		if (!ec->use_adaptive_rx_coalesce || !ec->use_adaptive_tx_coalesce)
+			pfvf->flags &= ~OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
+	} else {
+		priv_coalesce_status = 0;
+		if (ec->use_adaptive_rx_coalesce || ec->use_adaptive_tx_coalesce)
+			pfvf->flags |= OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
+	}
+
 	/* 'cq_time_wait' is 8bit and is in multiple of 100ns,
 	 * so clamp the user given value to the range of 1 to 25usec.
 	 */
@@ -494,9 +515,9 @@ static int otx2_set_coalesce(struct net_device *netdev,
 	 * so clamp the user given value to the range of 1 to 64k.
 	 */
 	ec->rx_max_coalesced_frames = clamp_t(u32, ec->rx_max_coalesced_frames,
-					      1, U16_MAX);
+					      1, NAPI_POLL_WEIGHT);
 	ec->tx_max_coalesced_frames = clamp_t(u32, ec->tx_max_coalesced_frames,
-					      1, U16_MAX);
+					      1, NAPI_POLL_WEIGHT);
 
 	/* Rx and Tx are mapped to same CQ, check which one
 	 * is changed, if both then choose the min.
@@ -509,6 +530,17 @@ static int otx2_set_coalesce(struct net_device *netdev,
 		hw->cq_ecount_wait = min_t(u16, ec->rx_max_coalesced_frames,
 					   ec->tx_max_coalesced_frames);
 
+	/* Reset 'cq_time_wait' and 'cq_ecount_wait' to
+	 * default values if coalesce status changed from
+	 * 'on' to 'off'.
+	 */
+	if (priv_coalesce_status &&
+	    ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) !=
+	     OTX2_FLAG_ADPTV_INT_COAL_ENABLED)) {
+		hw->cq_time_wait = CQ_TIMER_THRESH_DEFAULT;
+		hw->cq_ecount_wait = CQ_CQE_THRESH_DEFAULT;
+	}
+
 	if (netif_running(netdev)) {
 		for (qidx = 0; qidx < pfvf->hw.cint_cnt; qidx++)
 			otx2_config_irq_coalescing(pfvf, qidx);
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c
index 53b2706d65a1..a7e919b81a2e 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_pf.c
@@ -1254,6 +1254,7 @@ static irqreturn_t otx2_cq_intr_handler(int irq, void *cq_irq)
 	otx2_write64(pf, NIX_LF_CINTX_ENA_W1C(qidx), BIT_ULL(0));
 
 	/* Schedule NAPI */
+	pf->napi_events++;
 	napi_schedule_irqoff(&cq_poll->napi);
 
 	return IRQ_HANDLED;
@@ -1267,6 +1268,7 @@ static void otx2_disable_napi(struct otx2_nic *pf)
 
 	for (qidx = 0; qidx < pf->hw.cint_cnt; qidx++) {
 		cq_poll = &qset->napi[qidx];
+		cancel_work_sync(&cq_poll->dim.work);
 		napi_disable(&cq_poll->napi);
 		netif_napi_del(&cq_poll->napi);
 	}
@@ -1546,6 +1548,24 @@ static void otx2_do_set_rx_mode(struct otx2_nic *pf)
 	mutex_unlock(&pf->mbox.lock);
 }
 
+static void otx2_dim_work(struct work_struct *w)
+{
+	struct dim_cq_moder cur_moder;
+	struct otx2_cq_poll *cq_poll;
+	struct otx2_nic *pfvf;
+	struct dim *dim;
+
+	dim = container_of(w, struct dim, work);
+	cur_moder = net_dim_get_rx_moderation(dim->mode, dim->profile_ix);
+	cq_poll = container_of(dim, struct otx2_cq_poll, dim);
+	pfvf = (struct otx2_nic *)cq_poll->dev;
+	pfvf->hw.cq_time_wait = (cur_moder.usec > CQ_TIMER_THRESH_MAX) ?
+		CQ_TIMER_THRESH_MAX : cur_moder.usec;
+	pfvf->hw.cq_ecount_wait = (cur_moder.pkts > NAPI_POLL_WEIGHT) ?
+		NAPI_POLL_WEIGHT : cur_moder.pkts;
+	dim->state = DIM_START_MEASURE;
+}
+
 int otx2_open(struct net_device *netdev)
 {
 	struct otx2_nic *pf = netdev_priv(netdev);
@@ -1612,6 +1632,8 @@ int otx2_open(struct net_device *netdev)
 			cq_poll->cq_ids[CQ_XDP] = CINT_INVALID_CQ;
 
 		cq_poll->dev = (void *)pf;
+		cq_poll->dim.mode = DIM_CQ_PERIOD_MODE_START_FROM_CQE;
+		INIT_WORK(&cq_poll->dim.work, otx2_dim_work);
 		netif_napi_add(netdev, &cq_poll->napi,
 			       otx2_napi_handler, NAPI_POLL_WEIGHT);
 		napi_enable(&cq_poll->napi);
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
index c26de15b2ac3..3cd35a4dfa3c 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.c
@@ -484,6 +484,18 @@ static int otx2_tx_napi_handler(struct otx2_nic *pfvf,
 	return 0;
 }
 
+static void otx2_adjust_adaptive_coalese(struct otx2_nic *pfvf, struct otx2_cq_poll *cq_poll)
+{
+	struct dim_sample dim_sample;
+	u64 rx_frames, rx_bytes;
+
+	rx_frames = OTX2_GET_RX_STATS(RX_BCAST) + OTX2_GET_RX_STATS(RX_MCAST) +
+		OTX2_GET_RX_STATS(RX_UCAST);
+	rx_bytes = OTX2_GET_RX_STATS(RX_OCTS);
+	dim_update_sample(pfvf->napi_events, rx_frames, rx_bytes, &dim_sample);
+	net_dim(&cq_poll->dim, dim_sample);
+}
+
 int otx2_napi_handler(struct napi_struct *napi, int budget)
 {
 	struct otx2_cq_queue *rx_cq = NULL;
@@ -521,9 +533,20 @@ int otx2_napi_handler(struct napi_struct *napi, int budget)
 		if (pfvf->flags & OTX2_FLAG_INTF_DOWN)
 			return workdone;
 
+		/* Check for adaptive interrupt coalesce */
+		if (workdone != 0 &&
+		    ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
+		     OTX2_FLAG_ADPTV_INT_COAL_ENABLED)) {
+			/* Adjust irq coalese using net_dim */
+			otx2_adjust_adaptive_coalese(pfvf, cq_poll);
+			/* Update irq coalescing */
+			for (i = 0; i < pfvf->hw.cint_cnt; i++)
+				otx2_config_irq_coalescing(pfvf, i);
+		}
+
 		/* Re-enable interrupts */
 		otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx),
-			     BIT_ULL(0));
+				BIT_ULL(0));
 	}
 	return workdone;
 }
diff --git a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h
index f1a04cf9210c..c88e8a436029 100644
--- a/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h
+++ b/drivers/net/ethernet/marvell/octeontx2/nic/otx2_txrx.h
@@ -109,6 +109,7 @@ struct otx2_cq_poll {
 #define CINT_INVALID_CQ		255
 	u8			cint_idx;
 	u8			cq_ids[CQS_PER_CINT];
+	struct dim		dim;
 	struct napi_struct	napi;
 };
 
-- 
2.25.1

Re: [net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

[Jakub Kicinski]

On Thu, 12 May 2022 12:49:12 +0530 Suman Ghosh wrote:
> Added support for adaptive IRQ coalescing. It uses net_dim
> algorithm to find the suitable delay/IRQ count based on the
> current packet rate.
> 
> Signed-off-by: Suman Ghosh <sumang@marvell.com>
> Reviewed-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>

You still claim this is tested? How can it work if you're not changing
.supported_coalesce_params? Do we have a bug in the core?

> +	/* Check and update coalesce status */
> +	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
> +			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
> +		priv_coalesce_status = 1;
> +		if (!ec->use_adaptive_rx_coalesce || !ec->use_adaptive_tx_coalesce)
> +			pfvf->flags &= ~OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
> +	} else {
> +		priv_coalesce_status = 0;
> +		if (ec->use_adaptive_rx_coalesce || ec->use_adaptive_tx_coalesce)
> +			pfvf->flags |= OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
> +	}

Won't this flip-flop the flag? You should probably reject 
adaptive_rx != adaptive_tx.

>  		/* Re-enable interrupts */
>  		otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx),
> -			     BIT_ULL(0));
> +				BIT_ULL(0));

Unrelated and possibly misalinged?

RE: [EXT] Re: [net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

[Suman Ghosh]

>You still claim this is tested? How can it work if you're not changing
>.supported_coalesce_params? Do we have a bug in the core?
[Suman] This feature is changing " cq_time_wait" and " cq_time_wait" dynamically based on net_dim. Which other .supported_coalesce_params do you mean?
>
>> +	/* Check and update coalesce status */
>> +	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
>> +			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
>> +		priv_coalesce_status = 1;
>> +		if (!ec->use_adaptive_rx_coalesce || !ec-
>>use_adaptive_tx_coalesce)
>> +			pfvf->flags &= ~OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
>> +	} else {
>> +		priv_coalesce_status = 0;
>> +		if (ec->use_adaptive_rx_coalesce || ec-
>>use_adaptive_tx_coalesce)
>> +			pfvf->flags |= OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
>> +	}
>
>Won't this flip-flop the flag? You should probably reject adaptive_rx !=
>adaptive_tx.
[Suman] The idea here is to enable/disable coalesce if either rx or tx status is enable/disable.
If we enable only adaptive_rx then adaptive_tx will also get enabled as for this hardware both RX and TX 
are mapped to the same queue. We do not support separate setting for rx/tx coalesce for a net-device.
>
>>  		/* Re-enable interrupts */
>>  		otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx),
>> -			     BIT_ULL(0));
>> +				BIT_ULL(0));
>
>Unrelated and possibly misalinged?
[Suman] Yes, will fix in V2.

RE: [EXT] Re: [net-next PATCH] octeontx2-pf: Add support for adaptive interrupt coalescing

[Suman Ghosh]

>>You still claim this is tested? How can it work if you're not changing
>>.supported_coalesce_params? Do we have a bug in the core?
>[Suman] Added the flag.
>>
>>> +	/* Check and update coalesce status */
>>> +	if ((pfvf->flags & OTX2_FLAG_ADPTV_INT_COAL_ENABLED) ==
>>> +			OTX2_FLAG_ADPTV_INT_COAL_ENABLED) {
>>> +		priv_coalesce_status = 1;
>>> +		if (!ec->use_adaptive_rx_coalesce || !ec-
>>>use_adaptive_tx_coalesce)
>>> +			pfvf->flags &= ~OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
>>> +	} else {
>>> +		priv_coalesce_status = 0;
>>> +		if (ec->use_adaptive_rx_coalesce || ec-
>>>use_adaptive_tx_coalesce)
>>> +			pfvf->flags |= OTX2_FLAG_ADPTV_INT_COAL_ENABLED;
>>> +	}
>>
>>Won't this flip-flop the flag? You should probably reject adaptive_rx
>>!= adaptive_tx.
>[Suman] The idea here is to enable/disable coalesce if either rx or tx
>status is enable/disable.
>If we enable only adaptive_rx then adaptive_tx will also get enabled as
>for this hardware both RX and TX are mapped to the same queue. We do not
>support separate setting for rx/tx coalesce for a net-device.
>>
>>>  		/* Re-enable interrupts */
>>>  		otx2_write64(pfvf, NIX_LF_CINTX_ENA_W1S(cq_poll->cint_idx),
>>> -			     BIT_ULL(0));
>>> +				BIT_ULL(0));
>>
>>Unrelated and possibly misalinged?
>[Suman] Yes, will fix in V2.