Merge branch 'bnxt_en-support-header-page-pool-in-queue-api'

David Wei says:

====================
bnxt_en: support header page pool in queue API

Commit 7ed816b ("eth: bnxt: use page pool for head frags") added a
separate page pool for header frags. Now, frags are allocated from this
header page pool e.g. rxr->tpa_info.data.

The queue API did not properly handle rxr->tpa_info and so using the
queue API to i.e. reset any queues will result in pages being returned
to the incorrect page pool, causing inflight != 0 warnings.

Fix this bug by properly allocating/freeing tpa_info and copying/freeing
head_pool in the queue API implementation.

The 1st patch is a prep patch that refactors helpers out to be used by
the implementation patch later.

The 2nd patch is a drive-by refactor. Happy to take it out and re-send
to net-next if there are any objections.

The 3rd patch is the implementation patch that will properly alloc/free
rxr->tpa_info.
====================

Link: https://patch.msgid.link/20241204041022.56512-1-dw@davidwei.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Author: kuba-moo

drivers/net/ethernet/broadcom/bnxt/bnxt.c

@@ -3421,15 +3421,11 @@ static void bnxt_free_one_rx_agg_ring(struct bnxt *bp, struct bnxt_rx_ring_info
 	}
 }
 
-static void bnxt_free_one_rx_ring_skbs(struct bnxt *bp, int ring_nr)
+static void bnxt_free_one_tpa_info_data(struct bnxt *bp,
+					struct bnxt_rx_ring_info *rxr)
 {
-	struct bnxt_rx_ring_info *rxr = &bp->rx_ring[ring_nr];
-	struct bnxt_tpa_idx_map *map;
 	int i;
 
-	if (!rxr->rx_tpa)
-		goto skip_rx_tpa_free;
-
 	for (i = 0; i < bp->max_tpa; i++) {
 		struct bnxt_tpa_info *tpa_info = &rxr->rx_tpa[i];
 		u8 *data = tpa_info->data;
@@ -3440,6 +3436,17 @@ static void bnxt_free_one_rx_ring_skbs(struct bnxt *bp, int ring_nr)
 		tpa_info->data = NULL;
 		page_pool_free_va(rxr->head_pool, data, false);
 	}
+}
+
+static void bnxt_free_one_rx_ring_skbs(struct bnxt *bp,
+				       struct bnxt_rx_ring_info *rxr)
+{
+	struct bnxt_tpa_idx_map *map;
+
+	if (!rxr->rx_tpa)
+		goto skip_rx_tpa_free;
+
+	bnxt_free_one_tpa_info_data(bp, rxr);
 
 skip_rx_tpa_free:
 	if (!rxr->rx_buf_ring)
@@ -3467,7 +3474,7 @@ static void bnxt_free_rx_skbs(struct bnxt *bp)
 		return;
 
 	for (i = 0; i < bp->rx_nr_rings; i++)
-		bnxt_free_one_rx_ring_skbs(bp, i);
+		bnxt_free_one_rx_ring_skbs(bp, &bp->rx_ring[i]);
 }
 
 static void bnxt_free_skbs(struct bnxt *bp)
@@ -3608,29 +3615,64 @@ static int bnxt_alloc_ring(struct bnxt *bp, struct bnxt_ring_mem_info *rmem)
 	return 0;
 }
 
+static void bnxt_free_one_tpa_info(struct bnxt *bp,
+				   struct bnxt_rx_ring_info *rxr)
+{
+	int i;
+
+	kfree(rxr->rx_tpa_idx_map);
+	rxr->rx_tpa_idx_map = NULL;
+	if (rxr->rx_tpa) {
+		for (i = 0; i < bp->max_tpa; i++) {
+			kfree(rxr->rx_tpa[i].agg_arr);
+			rxr->rx_tpa[i].agg_arr = NULL;
+		}
+	}
+	kfree(rxr->rx_tpa);
+	rxr->rx_tpa = NULL;
+}
+
 static void bnxt_free_tpa_info(struct bnxt *bp)
 {
-	int i, j;
+	int i;
 
 	for (i = 0; i < bp->rx_nr_rings; i++) {
 		struct bnxt_rx_ring_info *rxr = &bp->rx_ring[i];
 
-		kfree(rxr->rx_tpa_idx_map);
-		rxr->rx_tpa_idx_map = NULL;
-		if (rxr->rx_tpa) {
-			for (j = 0; j < bp->max_tpa; j++) {
-				kfree(rxr->rx_tpa[j].agg_arr);
-				rxr->rx_tpa[j].agg_arr = NULL;
-			}
-		}
-		kfree(rxr->rx_tpa);
-		rxr->rx_tpa = NULL;
+		bnxt_free_one_tpa_info(bp, rxr);
+	}
+}
+
+static int bnxt_alloc_one_tpa_info(struct bnxt *bp,
+				   struct bnxt_rx_ring_info *rxr)
+{
+	struct rx_agg_cmp *agg;
+	int i;
+
+	rxr->rx_tpa = kcalloc(bp->max_tpa, sizeof(struct bnxt_tpa_info),
+			      GFP_KERNEL);
+	if (!rxr->rx_tpa)
+		return -ENOMEM;
+
+	if (!(bp->flags & BNXT_FLAG_CHIP_P5_PLUS))
+		return 0;
+	for (i = 0; i < bp->max_tpa; i++) {
+		agg = kcalloc(MAX_SKB_FRAGS, sizeof(*agg), GFP_KERNEL);
+		if (!agg)
+			return -ENOMEM;
+		rxr->rx_tpa[i].agg_arr = agg;
 	}
+	rxr->rx_tpa_idx_map = kzalloc(sizeof(*rxr->rx_tpa_idx_map),
+				      GFP_KERNEL);
+	if (!rxr->rx_tpa_idx_map)
+		return -ENOMEM;
+
+	return 0;
 }
 
 static int bnxt_alloc_tpa_info(struct bnxt *bp)
 {
-	int i, j;
+	int i, rc;
 
 	bp->max_tpa = MAX_TPA;
 	if (bp->flags & BNXT_FLAG_CHIP_P5_PLUS) {
@@ -3641,25 +3683,10 @@ static int bnxt_alloc_tpa_info(struct bnxt *bp)
 
 	for (i = 0; i < bp->rx_nr_rings; i++) {
 		struct bnxt_rx_ring_info *rxr = &bp->rx_ring[i];
-		struct rx_agg_cmp *agg;
-
-		rxr->rx_tpa = kcalloc(bp->max_tpa, sizeof(struct bnxt_tpa_info),
-				      GFP_KERNEL);
-		if (!rxr->rx_tpa)
-			return -ENOMEM;
 
-		if (!(bp->flags & BNXT_FLAG_CHIP_P5_PLUS))
-			continue;
-		for (j = 0; j < bp->max_tpa; j++) {
-			agg = kcalloc(MAX_SKB_FRAGS, sizeof(*agg), GFP_KERNEL);
-			if (!agg)
-				return -ENOMEM;
-			rxr->rx_tpa[j].agg_arr = agg;
-		}
-		rxr->rx_tpa_idx_map = kzalloc(sizeof(*rxr->rx_tpa_idx_map),
-					      GFP_KERNEL);
-		if (!rxr->rx_tpa_idx_map)
-			return -ENOMEM;
+		rc = bnxt_alloc_one_tpa_info(bp, rxr);
+		if (rc)
+			return rc;
 	}
 	return 0;
 }
@@ -3683,7 +3710,7 @@ static void bnxt_free_rx_rings(struct bnxt *bp)
 			xdp_rxq_info_unreg(&rxr->xdp_rxq);
 
 		page_pool_destroy(rxr->page_pool);
-		if (rxr->page_pool != rxr->head_pool)
+		if (bnxt_separate_head_pool())
 			page_pool_destroy(rxr->head_pool);
 		rxr->page_pool = rxr->head_pool = NULL;
 
@@ -3737,6 +3764,19 @@ static int bnxt_alloc_rx_page_pool(struct bnxt *bp,
 	return PTR_ERR(pool);
 }
 
+static int bnxt_alloc_rx_agg_bmap(struct bnxt *bp, struct bnxt_rx_ring_info *rxr)
+{
+	u16 mem_size;
+
+	rxr->rx_agg_bmap_size = bp->rx_agg_ring_mask + 1;
+	mem_size = rxr->rx_agg_bmap_size / 8;
+	rxr->rx_agg_bmap = kzalloc(mem_size, GFP_KERNEL);
+	if (!rxr->rx_agg_bmap)
+		return -ENOMEM;
+
+	return 0;
+}
+
 static int bnxt_alloc_rx_rings(struct bnxt *bp)
 {
 	int numa_node = dev_to_node(&bp->pdev->dev);
@@ -3781,19 +3821,15 @@ static int bnxt_alloc_rx_rings(struct bnxt *bp)
 
 		ring->grp_idx = i;
 		if (agg_rings) {
-			u16 mem_size;
-
 			ring = &rxr->rx_agg_ring_struct;
 			rc = bnxt_alloc_ring(bp, &ring->ring_mem);
 			if (rc)
 				return rc;
 
 			ring->grp_idx = i;
-			rxr->rx_agg_bmap_size = bp->rx_agg_ring_mask + 1;
-			mem_size = rxr->rx_agg_bmap_size / 8;
-			rxr->rx_agg_bmap = kzalloc(mem_size, GFP_KERNEL);
-			if (!rxr->rx_agg_bmap)
-				return -ENOMEM;
+			rc = bnxt_alloc_rx_agg_bmap(bp, rxr);
+			if (rc)
+				return rc;
 		}
 	}
 	if (bp->flags & BNXT_FLAG_TPA)
@@ -4268,10 +4304,31 @@ static void bnxt_alloc_one_rx_ring_page(struct bnxt *bp,
 	rxr->rx_agg_prod = prod;
 }
 
+static int bnxt_alloc_one_tpa_info_data(struct bnxt *bp,
+					struct bnxt_rx_ring_info *rxr)
+{
+	dma_addr_t mapping;
+	u8 *data;
+	int i;
+
+	for (i = 0; i < bp->max_tpa; i++) {
+		data = __bnxt_alloc_rx_frag(bp, &mapping, rxr,
+					    GFP_KERNEL);
+		if (!data)
+			return -ENOMEM;
+
+		rxr->rx_tpa[i].data = data;
+		rxr->rx_tpa[i].data_ptr = data + bp->rx_offset;
+		rxr->rx_tpa[i].mapping = mapping;
+	}
+
+	return 0;
+}
+
 static int bnxt_alloc_one_rx_ring(struct bnxt *bp, int ring_nr)
 {
 	struct bnxt_rx_ring_info *rxr = &bp->rx_ring[ring_nr];
-	int i;
+	int rc;
 
 	bnxt_alloc_one_rx_ring_skb(bp, rxr, ring_nr);
 
@@ -4281,19 +4338,9 @@ static int bnxt_alloc_one_rx_ring(struct bnxt *bp, int ring_nr)
 	bnxt_alloc_one_rx_ring_page(bp, rxr, ring_nr);
 
 	if (rxr->rx_tpa) {
-		dma_addr_t mapping;
-		u8 *data;
-
-		for (i = 0; i < bp->max_tpa; i++) {
-			data = __bnxt_alloc_rx_frag(bp, &mapping, rxr,
-						    GFP_KERNEL);
-			if (!data)
-				return -ENOMEM;
-
-			rxr->rx_tpa[i].data = data;
-			rxr->rx_tpa[i].data_ptr = data + bp->rx_offset;
-			rxr->rx_tpa[i].mapping = mapping;
-		}
+		rc = bnxt_alloc_one_tpa_info_data(bp, rxr);
+		if (rc)
+			return rc;
 	}
 	return 0;
 }
@@ -13663,7 +13710,7 @@ static void bnxt_rx_ring_reset(struct bnxt *bp)
 			bnxt_reset_task(bp, true);
 			break;
 		}
-		bnxt_free_one_rx_ring_skbs(bp, i);
+		bnxt_free_one_rx_ring_skbs(bp, rxr);
 		rxr->rx_prod = 0;
 		rxr->rx_agg_prod = 0;
 		rxr->rx_sw_agg_prod = 0;
@@ -15293,19 +15340,6 @@ static const struct netdev_stat_ops bnxt_stat_ops = {
 	.get_base_stats		= bnxt_get_base_stats,
 };
 
-static int bnxt_alloc_rx_agg_bmap(struct bnxt *bp, struct bnxt_rx_ring_info *rxr)
-{
-	u16 mem_size;
-
-	rxr->rx_agg_bmap_size = bp->rx_agg_ring_mask + 1;
-	mem_size = rxr->rx_agg_bmap_size / 8;
-	rxr->rx_agg_bmap = kzalloc(mem_size, GFP_KERNEL);
-	if (!rxr->rx_agg_bmap)
-		return -ENOMEM;
-
-	return 0;
-}
-
 static int bnxt_queue_mem_alloc(struct net_device *dev, void *qmem, int idx)
 {
 	struct bnxt_rx_ring_info *rxr, *clone;
@@ -15354,25 +15388,37 @@ static int bnxt_queue_mem_alloc(struct net_device *dev, void *qmem, int idx)
 			goto err_free_rx_agg_ring;
 	}
 
+	if (bp->flags & BNXT_FLAG_TPA) {
+		rc = bnxt_alloc_one_tpa_info(bp, clone);
+		if (rc)
+			goto err_free_tpa_info;
+	}
+
 	bnxt_init_one_rx_ring_rxbd(bp, clone);
 	bnxt_init_one_rx_agg_ring_rxbd(bp, clone);
 
 	bnxt_alloc_one_rx_ring_skb(bp, clone, idx);
 	if (bp->flags & BNXT_FLAG_AGG_RINGS)
 		bnxt_alloc_one_rx_ring_page(bp, clone, idx);
+	if (bp->flags & BNXT_FLAG_TPA)
+		bnxt_alloc_one_tpa_info_data(bp, clone);
 
 	return 0;
 
+err_free_tpa_info:
+	bnxt_free_one_tpa_info(bp, clone);
 err_free_rx_agg_ring:
 	bnxt_free_ring(bp, &clone->rx_agg_ring_struct.ring_mem);
 err_free_rx_ring:
 	bnxt_free_ring(bp, &clone->rx_ring_struct.ring_mem);
 err_rxq_info_unreg:
 	xdp_rxq_info_unreg(&clone->xdp_rxq);
 err_page_pool_destroy:
-	clone->page_pool->p.napi = NULL;
 	page_pool_destroy(clone->page_pool);
+	if (bnxt_separate_head_pool())
+		page_pool_destroy(clone->head_pool);
 	clone->page_pool = NULL;
+	clone->head_pool = NULL;
 	return rc;
 }
 
@@ -15382,13 +15428,15 @@ static void bnxt_queue_mem_free(struct net_device *dev, void *qmem)
 	struct bnxt *bp = netdev_priv(dev);
 	struct bnxt_ring_struct *ring;
 
-	bnxt_free_one_rx_ring(bp, rxr);
-	bnxt_free_one_rx_agg_ring(bp, rxr);
+	bnxt_free_one_rx_ring_skbs(bp, rxr);
 
 	xdp_rxq_info_unreg(&rxr->xdp_rxq);
 
 	page_pool_destroy(rxr->page_pool);
+	if (bnxt_separate_head_pool())
+		page_pool_destroy(rxr->head_pool);
 	rxr->page_pool = NULL;
+	rxr->head_pool = NULL;
 
 	ring = &rxr->rx_ring_struct;
 	bnxt_free_ring(bp, &ring->ring_mem);
@@ -15470,7 +15518,10 @@ static int bnxt_queue_start(struct net_device *dev, void *qmem, int idx)
 	rxr->rx_agg_prod = clone->rx_agg_prod;
 	rxr->rx_sw_agg_prod = clone->rx_sw_agg_prod;
 	rxr->rx_next_cons = clone->rx_next_cons;
+	rxr->rx_tpa = clone->rx_tpa;
+	rxr->rx_tpa_idx_map = clone->rx_tpa_idx_map;
 	rxr->page_pool = clone->page_pool;
+	rxr->head_pool = clone->head_pool;
 	rxr->xdp_rxq = clone->xdp_rxq;
 
 	bnxt_copy_rx_ring(bp, rxr, clone);
@@ -15529,6 +15580,8 @@ static int bnxt_queue_stop(struct net_device *dev, void *qmem, int idx)
 	bnxt_hwrm_rx_agg_ring_free(bp, rxr, false);
 	rxr->rx_next_cons = 0;
 	page_pool_disable_direct_recycling(rxr->page_pool);
+	if (bnxt_separate_head_pool())
+		page_pool_disable_direct_recycling(rxr->head_pool);
 
 	memcpy(qmem, rxr, sizeof(*rxr));
 	bnxt_init_rx_ring_struct(bp, qmem);