idpf: improve when to set RE bit logic

jira KERNEL-168
commit-author Joshua Hay <joshua.a.hay@intel.com>
commit f2d18e1
upstream-diff |
	adjusted context in struct idpf_tx_queue because the order
	of the fields is different due to missing
	- 5a816aa ("idpf: strictly assert cachelines of queue and queue vector structures")

Track the gap between next_to_use and the last RE index. Set RE again
if the gap is large enough to ensure RE bit is set frequently. This is
critical before removing the stashing mechanisms because the
opportunistic descriptor ring cleaning from the out-of-order completions
will go away. Previously the descriptors would be "cleaned" by both the
descriptor (RE) completion and the out-of-order completions. Without the
latter, we must ensure the RE bit is set more frequently. Otherwise,
it's theoretically possible for the descriptor ring next_to_clean to
never advance.  The previous implementation was dependent on the start
of a packet falling on a 64th index in the descriptor ring, which is not
guaranteed with large packets.

	Signed-off-by: Luigi Rizzo <lrizzo@google.com>
	Signed-off-by: Brian Vazquez <brianvv@google.com>
	Signed-off-by: Joshua Hay <joshua.a.hay@intel.com>
	Reviewed-by: Madhu Chittim <madhu.chittim@intel.com>
	Tested-by: Samuel Salin <Samuel.salin@intel.com>
	Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
(cherry picked from commit f2d18e1)
	Signed-off-by: Roxana Nicolescu <rnicolescu@ciq.com>

Author: roxanan1996

drivers/net/ethernet/intel/idpf/idpf_txrx.c

@@ -293,6 +293,8 @@ static int idpf_tx_desc_alloc(const struct idpf_vport *vport,
 	 */
 	idpf_queue_change(GEN_CHK, refillq);
 
+	tx_q->last_re = tx_q->desc_count - IDPF_TX_SPLITQ_RE_MIN_GAP;
+
 	return 0;
 
 err_alloc:
@@ -2791,6 +2793,21 @@ netdev_tx_t idpf_tx_drop_skb(struct idpf_tx_queue *tx_q, struct sk_buff *skb)
 	return NETDEV_TX_OK;
 }
 
+/**
+ * idpf_tx_splitq_need_re - check whether RE bit needs to be set
+ * @tx_q: pointer to Tx queue
+ *
+ * Return: true if RE bit needs to be set, false otherwise
+ */
+static bool idpf_tx_splitq_need_re(struct idpf_tx_queue *tx_q)
+{
+	int gap = tx_q->next_to_use - tx_q->last_re;
+
+	gap += (gap < 0) ? tx_q->desc_count : 0;
+
+	return gap >= IDPF_TX_SPLITQ_RE_MIN_GAP;
+}
+
 /**
  * idpf_tx_splitq_frame - Sends buffer on Tx ring using flex descriptors
  * @skb: send buffer
@@ -2871,9 +2888,10 @@ static netdev_tx_t idpf_tx_splitq_frame(struct sk_buff *skb,
 		 * MIN_RING size to ensure it will be set at least once each
 		 * time around the ring.
 		 */
-		if (!(tx_q->next_to_use % IDPF_TX_SPLITQ_RE_MIN_GAP)) {
+		if (idpf_tx_splitq_need_re(tx_q)) {
 			tx_params.eop_cmd |= IDPF_TXD_FLEX_FLOW_CMD_RE;
 			tx_q->txq_grp->num_completions_pending++;
+			tx_q->last_re = tx_q->next_to_use;
 		}
 
 		if (skb->ip_summed == CHECKSUM_PARTIAL)

drivers/net/ethernet/intel/idpf/idpf_txrx.h

@@ -672,6 +672,8 @@ struct idpf_rx_queue {
  * @desc_count: Number of descriptors
  * @next_to_use: Next descriptor to use
  * @next_to_clean: Next descriptor to clean
+ * @last_re: last descriptor index that RE bit was set
+ * @tx_max_bufs: Max buffers that can be transmitted with scatter-gather
  * @netdev: &net_device corresponding to this queue
  * @cleaned_bytes: Splitq only, TXQ only: When a TX completion is received on
  *		   the TX completion queue, it can be for any TXQ associated
@@ -683,7 +685,6 @@ struct idpf_rx_queue {
  *		   only once at the end of the cleaning routine.
  * @clean_budget: singleq only, queue cleaning budget
  * @cleaned_pkts: Number of packets cleaned for the above said case
- * @tx_max_bufs: Max buffers that can be transmitted with scatter-gather
  * @tx_min_pkt_len: Min supported packet length
  * @refillq: Pointer to refill queue
  * @compl_tag_bufid_m: Completion tag buffer id mask
@@ -736,6 +737,8 @@ struct idpf_tx_queue {
 	u16 desc_count;
 	u16 next_to_use;
 	u16 next_to_clean;
+	u16 last_re;
+	u16 tx_max_bufs;
 
 	struct net_device *netdev;
 
@@ -745,7 +748,6 @@ struct idpf_tx_queue {
 	};
 	u16 cleaned_pkts;
 
-	u16 tx_max_bufs;
 	u16 tx_min_pkt_len;
 	struct idpf_sw_queue *refillq;