netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH net] net: hns: fix for unmapping problem when SMMU is on
@ 2018-09-25  9:21 Salil Mehta
  2018-09-25 17:43 ` David Miller
  0 siblings, 1 reply; 2+ messages in thread
From: Salil Mehta @ 2018-09-25  9:21 UTC (permalink / raw)
  To: davem
  Cc: salil.mehta, yisen.zhuang, lipeng321, mehta.salil, netdev,
	linux-kernel, linuxarm, Yunsheng Lin

From: Yunsheng Lin <linyunsheng@huawei.com>

If SMMU is on, there is more likely that skb_shinfo(skb)->frags[i]
can not send by a single BD. when this happen, the
hns_nic_net_xmit_hw function map the whole data in a frags using
skb_frag_dma_map, but unmap each BD' data individually when tx is
done, which causes problem when SMMU is on.

This patch fixes this problem by ummapping the whole data in a
frags when tx is done.

Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: Peng Li <lipeng321@huawei.com>
Reviewed-by: Yisen Zhuang <yisen.zhuang@huawei.com>
Signed-off-by: Salil Mehta <salil.mehta@huawei.com>
---
 drivers/net/ethernet/hisilicon/hns/hnae.c     |  2 +-
 drivers/net/ethernet/hisilicon/hns/hns_enet.c | 30 +++++++++++++++++----------
 2 files changed, 20 insertions(+), 12 deletions(-)

diff --git a/drivers/net/ethernet/hisilicon/hns/hnae.c b/drivers/net/ethernet/hisilicon/hns/hnae.c
index a051e58..79d03f8 100644
--- a/drivers/net/ethernet/hisilicon/hns/hnae.c
+++ b/drivers/net/ethernet/hisilicon/hns/hnae.c
@@ -84,7 +84,7 @@ static void hnae_unmap_buffer(struct hnae_ring *ring, struct hnae_desc_cb *cb)
 	if (cb->type == DESC_TYPE_SKB)
 		dma_unmap_single(ring_to_dev(ring), cb->dma, cb->length,
 				 ring_to_dma_dir(ring));
-	else
+	else if (cb->length)
 		dma_unmap_page(ring_to_dev(ring), cb->dma, cb->length,
 			       ring_to_dma_dir(ring));
 }
diff --git a/drivers/net/ethernet/hisilicon/hns/hns_enet.c b/drivers/net/ethernet/hisilicon/hns/hns_enet.c
index f56855e..5ce23d4 100644
--- a/drivers/net/ethernet/hisilicon/hns/hns_enet.c
+++ b/drivers/net/ethernet/hisilicon/hns/hns_enet.c
@@ -40,9 +40,9 @@
 #define SKB_TMP_LEN(SKB) \
 	(((SKB)->transport_header - (SKB)->mac_header) + tcp_hdrlen(SKB))
 
-static void fill_v2_desc(struct hnae_ring *ring, void *priv,
-			 int size, dma_addr_t dma, int frag_end,
-			 int buf_num, enum hns_desc_type type, int mtu)
+static void fill_v2_desc_hw(struct hnae_ring *ring, void *priv, int size,
+			    int send_sz, dma_addr_t dma, int frag_end,
+			    int buf_num, enum hns_desc_type type, int mtu)
 {
 	struct hnae_desc *desc = &ring->desc[ring->next_to_use];
 	struct hnae_desc_cb *desc_cb = &ring->desc_cb[ring->next_to_use];
@@ -64,7 +64,7 @@ static void fill_v2_desc(struct hnae_ring *ring, void *priv,
 	desc_cb->type = type;
 
 	desc->addr = cpu_to_le64(dma);
-	desc->tx.send_size = cpu_to_le16((u16)size);
+	desc->tx.send_size = cpu_to_le16((u16)send_sz);
 
 	/* config bd buffer end */
 	hnae_set_bit(rrcfv, HNSV2_TXD_VLD_B, 1);
@@ -133,6 +133,14 @@ static void fill_v2_desc(struct hnae_ring *ring, void *priv,
 	ring_ptr_move_fw(ring, next_to_use);
 }
 
+static void fill_v2_desc(struct hnae_ring *ring, void *priv,
+			 int size, dma_addr_t dma, int frag_end,
+			 int buf_num, enum hns_desc_type type, int mtu)
+{
+	fill_v2_desc_hw(ring, priv, size, size, dma, frag_end,
+			buf_num, type, mtu);
+}
+
 static const struct acpi_device_id hns_enet_acpi_match[] = {
 	{ "HISI00C1", 0 },
 	{ "HISI00C2", 0 },
@@ -289,15 +297,15 @@ static void fill_tso_desc(struct hnae_ring *ring, void *priv,
 
 	/* when the frag size is bigger than hardware, split this frag */
 	for (k = 0; k < frag_buf_num; k++)
-		fill_v2_desc(ring, priv,
-			     (k == frag_buf_num - 1) ?
+		fill_v2_desc_hw(ring, priv, k == 0 ? size : 0,
+				(k == frag_buf_num - 1) ?
 					sizeoflast : BD_MAX_SEND_SIZE,
-			     dma + BD_MAX_SEND_SIZE * k,
-			     frag_end && (k == frag_buf_num - 1) ? 1 : 0,
-			     buf_num,
-			     (type == DESC_TYPE_SKB && !k) ?
+				dma + BD_MAX_SEND_SIZE * k,
+				frag_end && (k == frag_buf_num - 1) ? 1 : 0,
+				buf_num,
+				(type == DESC_TYPE_SKB && !k) ?
 					DESC_TYPE_SKB : DESC_TYPE_PAGE,
-			     mtu);
+				mtu);
 }
 
 netdev_tx_t hns_nic_net_xmit_hw(struct net_device *ndev,
-- 
2.7.4

^ permalink raw reply related	[flat|nested] 2+ messages in thread

* Re: [PATCH net] net: hns: fix for unmapping problem when SMMU is on
  2018-09-25  9:21 [PATCH net] net: hns: fix for unmapping problem when SMMU is on Salil Mehta
@ 2018-09-25 17:43 ` David Miller
  0 siblings, 0 replies; 2+ messages in thread
From: David Miller @ 2018-09-25 17:43 UTC (permalink / raw)
  To: salil.mehta
  Cc: yisen.zhuang, lipeng321, mehta.salil, netdev, linux-kernel,
	linuxarm, linyunsheng

From: Salil Mehta <salil.mehta@huawei.com>
Date: Tue, 25 Sep 2018 10:21:55 +0100

> From: Yunsheng Lin <linyunsheng@huawei.com>
> 
> If SMMU is on, there is more likely that skb_shinfo(skb)->frags[i]
> can not send by a single BD. when this happen, the
> hns_nic_net_xmit_hw function map the whole data in a frags using
> skb_frag_dma_map, but unmap each BD' data individually when tx is
> done, which causes problem when SMMU is on.
> 
> This patch fixes this problem by ummapping the whole data in a
> frags when tx is done.
> 
> Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
> Signed-off-by: Peng Li <lipeng321@huawei.com>
> Reviewed-by: Yisen Zhuang <yisen.zhuang@huawei.com>
> Signed-off-by: Salil Mehta <salil.mehta@huawei.com>

Applied, thank you.

^ permalink raw reply	[flat|nested] 2+ messages in thread

end of thread, other threads:[~2018-09-25 17:43 UTC | newest]

Thread overview: 2+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2018-09-25  9:21 [PATCH net] net: hns: fix for unmapping problem when SMMU is on Salil Mehta
2018-09-25 17:43 ` David Miller

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).