Commit 8c30e194 authored by Yunsheng Lin's avatar Yunsheng Lin Committed by David S. Miller

net: hns3: optimize the rx clean process

Currently HNS3_RING_RX_RING_FBDNUM_REG register is read to determine
how many rx desc can be cleaned. To avoid the register read operation
in the critical data path, use the valid bit in the rx desc to determine
if a specific rx desc can be cleaned.

The hns3 driver clear valid bit in the rx desc before notifying the
rx desc to the hw, and hw will only set the valid bit of the rx desc
after corresponding buffer is filled with packet data and other field
in the rx desc is set accordingly.

Add hns3_rx_ring_move_fw() function to clear the valid bit in the rx
desc before moving rx ring's next_to_clean forward to avoid double
cleaning a rx desc, also add a dma_rmb() barrier in hns3_handle_rx_bd()
to make sure valid bit is set before reading other field in the rx desc.
Signed-off-by: default avatarYunsheng Lin <linyunsheng@huawei.com>
Signed-off-by: default avatarHuazhong Tan <tanhuazhong@huawei.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
parent 20d06ca2
...@@ -2845,6 +2845,16 @@ static bool hns3_parse_vlan_tag(struct hns3_enet_ring *ring, ...@@ -2845,6 +2845,16 @@ static bool hns3_parse_vlan_tag(struct hns3_enet_ring *ring,
} }
} }
static void hns3_rx_ring_move_fw(struct hns3_enet_ring *ring)
{
ring->desc[ring->next_to_clean].rx.bd_base_info &=
cpu_to_le32(~BIT(HNS3_RXD_VLD_B));
ring->next_to_clean += 1;
if (unlikely(ring->next_to_clean == ring->desc_num))
ring->next_to_clean = 0;
}
static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
unsigned char *va) unsigned char *va)
{ {
...@@ -2880,7 +2890,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, ...@@ -2880,7 +2890,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
__page_frag_cache_drain(desc_cb->priv, __page_frag_cache_drain(desc_cb->priv,
desc_cb->pagecnt_bias); desc_cb->pagecnt_bias);
ring_ptr_move_fw(ring, next_to_clean); hns3_rx_ring_move_fw(ring);
return 0; return 0;
} }
u64_stats_update_begin(&ring->syncp); u64_stats_update_begin(&ring->syncp);
...@@ -2891,7 +2901,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length, ...@@ -2891,7 +2901,7 @@ static int hns3_alloc_skb(struct hns3_enet_ring *ring, unsigned int length,
__skb_put(skb, ring->pull_len); __skb_put(skb, ring->pull_len);
hns3_nic_reuse_page(skb, ring->frag_num++, ring, ring->pull_len, hns3_nic_reuse_page(skb, ring->frag_num++, ring, ring->pull_len,
desc_cb); desc_cb);
ring_ptr_move_fw(ring, next_to_clean); hns3_rx_ring_move_fw(ring);
return 0; return 0;
} }
...@@ -2946,7 +2956,7 @@ static int hns3_add_frag(struct hns3_enet_ring *ring) ...@@ -2946,7 +2956,7 @@ static int hns3_add_frag(struct hns3_enet_ring *ring)
hns3_nic_reuse_page(skb, ring->frag_num++, ring, 0, desc_cb); hns3_nic_reuse_page(skb, ring->frag_num++, ring, 0, desc_cb);
trace_hns3_rx_desc(ring); trace_hns3_rx_desc(ring);
ring_ptr_move_fw(ring, next_to_clean); hns3_rx_ring_move_fw(ring);
ring->pending_buf++; ring->pending_buf++;
} while (!(bd_base_info & BIT(HNS3_RXD_FE_B))); } while (!(bd_base_info & BIT(HNS3_RXD_FE_B)));
...@@ -3088,32 +3098,32 @@ static int hns3_handle_rx_bd(struct hns3_enet_ring *ring) ...@@ -3088,32 +3098,32 @@ static int hns3_handle_rx_bd(struct hns3_enet_ring *ring)
prefetch(desc); prefetch(desc);
length = le16_to_cpu(desc->rx.size); if (!skb) {
bd_base_info = le32_to_cpu(desc->rx.bd_base_info); bd_base_info = le32_to_cpu(desc->rx.bd_base_info);
/* Check valid BD */ /* Check valid BD */
if (unlikely(!(bd_base_info & BIT(HNS3_RXD_VLD_B)))) if (unlikely(!(bd_base_info & BIT(HNS3_RXD_VLD_B))))
return -ENXIO; return -ENXIO;
if (!skb) { dma_rmb();
length = le16_to_cpu(desc->rx.size);
ring->va = desc_cb->buf + desc_cb->page_offset; ring->va = desc_cb->buf + desc_cb->page_offset;
dma_sync_single_for_cpu(ring_to_dev(ring), dma_sync_single_for_cpu(ring_to_dev(ring),
desc_cb->dma + desc_cb->page_offset, desc_cb->dma + desc_cb->page_offset,
hns3_buf_size(ring), hns3_buf_size(ring),
DMA_FROM_DEVICE); DMA_FROM_DEVICE);
}
/* Prefetch first cache line of first page /* Prefetch first cache line of first page.
* Idea is to cache few bytes of the header of the packet. Our L1 Cache * Idea is to cache few bytes of the header of the packet.
* line size is 64B so need to prefetch twice to make it 128B. But in * Our L1 Cache line size is 64B so need to prefetch twice to make
* actual we can have greater size of caches with 128B Level 1 cache * it 128B. But in actual we can have greater size of caches with
* lines. In such a case, single fetch would suffice to cache in the * 128B Level 1 cache lines. In such a case, single fetch would
* relevant part of the header. * suffice to cache in the relevant part of the header.
*/ */
net_prefetch(ring->va); net_prefetch(ring->va);
if (!skb) {
ret = hns3_alloc_skb(ring, length, ring->va); ret = hns3_alloc_skb(ring, length, ring->va);
skb = ring->skb; skb = ring->skb;
...@@ -3153,19 +3163,11 @@ int hns3_clean_rx_ring(struct hns3_enet_ring *ring, int budget, ...@@ -3153,19 +3163,11 @@ int hns3_clean_rx_ring(struct hns3_enet_ring *ring, int budget,
#define RCB_NOF_ALLOC_RX_BUFF_ONCE 16 #define RCB_NOF_ALLOC_RX_BUFF_ONCE 16
int unused_count = hns3_desc_unused(ring); int unused_count = hns3_desc_unused(ring);
int recv_pkts = 0; int recv_pkts = 0;
int recv_bds = 0; int err;
int err, num;
num = readl_relaxed(ring->tqp->io_base + HNS3_RING_RX_RING_FBDNUM_REG);
num -= unused_count;
unused_count -= ring->pending_buf; unused_count -= ring->pending_buf;
if (num <= 0) while (recv_pkts < budget) {
goto out;
rmb(); /* Make sure num taken effect before the other data is touched */
while (recv_pkts < budget && recv_bds < num) {
/* Reuse or realloc buffers */ /* Reuse or realloc buffers */
if (unused_count >= RCB_NOF_ALLOC_RX_BUFF_ONCE) { if (unused_count >= RCB_NOF_ALLOC_RX_BUFF_ONCE) {
hns3_nic_alloc_rx_buffers(ring, unused_count); hns3_nic_alloc_rx_buffers(ring, unused_count);
...@@ -3183,7 +3185,6 @@ int hns3_clean_rx_ring(struct hns3_enet_ring *ring, int budget, ...@@ -3183,7 +3185,6 @@ int hns3_clean_rx_ring(struct hns3_enet_ring *ring, int budget,
recv_pkts++; recv_pkts++;
} }
recv_bds += ring->pending_buf;
unused_count += ring->pending_buf; unused_count += ring->pending_buf;
ring->skb = NULL; ring->skb = NULL;
ring->pending_buf = 0; ring->pending_buf = 0;
......
Markdown is supported
0%
or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment