Commit 039fbc47 authored by Lorenzo Bianconi's avatar Lorenzo Bianconi Committed by Jakub Kicinski

net: mvneta: alloc skb_shared_info on the mvneta_rx_swbm stack

Build skb_shared_info on mvneta_rx_swbm stack and sync it to xdp_buff
skb_shared_info area only on the last fragment. Leftover cache miss in
mvneta_swbm_rx_frame will be addressed introducing mb bit in
xdp_buff/xdp_frame struct
Signed-off-by: default avatarLorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: default avatarJesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: default avatarJakub Kicinski <kuba@kernel.org>
parent eb33f118
...@@ -2278,9 +2278,9 @@ mvneta_swbm_add_rx_fragment(struct mvneta_port *pp, ...@@ -2278,9 +2278,9 @@ mvneta_swbm_add_rx_fragment(struct mvneta_port *pp,
struct mvneta_rx_desc *rx_desc, struct mvneta_rx_desc *rx_desc,
struct mvneta_rx_queue *rxq, struct mvneta_rx_queue *rxq,
struct xdp_buff *xdp, int *size, struct xdp_buff *xdp, int *size,
struct skb_shared_info *xdp_sinfo,
struct page *page) struct page *page)
{ {
struct skb_shared_info *sinfo = xdp_get_shared_info_from_buff(xdp);
struct net_device *dev = pp->dev; struct net_device *dev = pp->dev;
enum dma_data_direction dma_dir; enum dma_data_direction dma_dir;
int data_len, len; int data_len, len;
...@@ -2298,13 +2298,22 @@ mvneta_swbm_add_rx_fragment(struct mvneta_port *pp, ...@@ -2298,13 +2298,22 @@ mvneta_swbm_add_rx_fragment(struct mvneta_port *pp,
len, dma_dir); len, dma_dir);
rx_desc->buf_phys_addr = 0; rx_desc->buf_phys_addr = 0;
if (data_len > 0 && sinfo->nr_frags < MAX_SKB_FRAGS) { if (data_len > 0 && xdp_sinfo->nr_frags < MAX_SKB_FRAGS) {
skb_frag_t *frag = &sinfo->frags[sinfo->nr_frags]; skb_frag_t *frag = &xdp_sinfo->frags[xdp_sinfo->nr_frags++];
skb_frag_off_set(frag, pp->rx_offset_correction); skb_frag_off_set(frag, pp->rx_offset_correction);
skb_frag_size_set(frag, data_len); skb_frag_size_set(frag, data_len);
__skb_frag_set_page(frag, page); __skb_frag_set_page(frag, page);
sinfo->nr_frags++;
/* last fragment */
if (len == *size) {
struct skb_shared_info *sinfo;
sinfo = xdp_get_shared_info_from_buff(xdp);
sinfo->nr_frags = xdp_sinfo->nr_frags;
memcpy(sinfo->frags, xdp_sinfo->frags,
sinfo->nr_frags * sizeof(skb_frag_t));
}
} else { } else {
page_pool_put_full_page(rxq->page_pool, page, true); page_pool_put_full_page(rxq->page_pool, page, true);
} }
...@@ -2348,6 +2357,7 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2348,6 +2357,7 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
{ {
int rx_proc = 0, rx_todo, refill, size = 0; int rx_proc = 0, rx_todo, refill, size = 0;
struct net_device *dev = pp->dev; struct net_device *dev = pp->dev;
struct skb_shared_info sinfo;
struct mvneta_stats ps = {}; struct mvneta_stats ps = {};
struct bpf_prog *xdp_prog; struct bpf_prog *xdp_prog;
u32 desc_status, frame_sz; u32 desc_status, frame_sz;
...@@ -2357,6 +2367,8 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2357,6 +2367,8 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
xdp_buf.frame_sz = PAGE_SIZE; xdp_buf.frame_sz = PAGE_SIZE;
xdp_buf.rxq = &rxq->xdp_rxq; xdp_buf.rxq = &rxq->xdp_rxq;
sinfo.nr_frags = 0;
/* Get number of received packets */ /* Get number of received packets */
rx_todo = mvneta_rxq_busy_desc_num_get(pp, rxq); rx_todo = mvneta_rxq_busy_desc_num_get(pp, rxq);
...@@ -2395,11 +2407,11 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2395,11 +2407,11 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
rx_desc->buf_phys_addr = 0; rx_desc->buf_phys_addr = 0;
page_pool_put_full_page(rxq->page_pool, page, page_pool_put_full_page(rxq->page_pool, page,
true); true);
continue; goto next;
} }
mvneta_swbm_add_rx_fragment(pp, rx_desc, rxq, &xdp_buf, mvneta_swbm_add_rx_fragment(pp, rx_desc, rxq, &xdp_buf,
&size, page); &size, &sinfo, page);
} /* Middle or Last descriptor */ } /* Middle or Last descriptor */
if (!(rx_status & MVNETA_RXD_LAST_DESC)) if (!(rx_status & MVNETA_RXD_LAST_DESC))
...@@ -2407,10 +2419,7 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2407,10 +2419,7 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
continue; continue;
if (size) { if (size) {
struct skb_shared_info *sinfo; mvneta_xdp_put_buff(pp, rxq, &xdp_buf, &sinfo, -1);
sinfo = xdp_get_shared_info_from_buff(&xdp_buf);
mvneta_xdp_put_buff(pp, rxq, &xdp_buf, sinfo, -1);
goto next; goto next;
} }
...@@ -2421,10 +2430,8 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2421,10 +2430,8 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
skb = mvneta_swbm_build_skb(pp, rxq, &xdp_buf, desc_status); skb = mvneta_swbm_build_skb(pp, rxq, &xdp_buf, desc_status);
if (IS_ERR(skb)) { if (IS_ERR(skb)) {
struct mvneta_pcpu_stats *stats = this_cpu_ptr(pp->stats); struct mvneta_pcpu_stats *stats = this_cpu_ptr(pp->stats);
struct skb_shared_info *sinfo;
sinfo = xdp_get_shared_info_from_buff(&xdp_buf); mvneta_xdp_put_buff(pp, rxq, &xdp_buf, &sinfo, -1);
mvneta_xdp_put_buff(pp, rxq, &xdp_buf, sinfo, -1);
u64_stats_update_begin(&stats->syncp); u64_stats_update_begin(&stats->syncp);
stats->es.skb_alloc_error++; stats->es.skb_alloc_error++;
...@@ -2441,15 +2448,12 @@ static int mvneta_rx_swbm(struct napi_struct *napi, ...@@ -2441,15 +2448,12 @@ static int mvneta_rx_swbm(struct napi_struct *napi,
napi_gro_receive(napi, skb); napi_gro_receive(napi, skb);
next: next:
xdp_buf.data_hard_start = NULL; xdp_buf.data_hard_start = NULL;
sinfo.nr_frags = 0;
} }
rcu_read_unlock(); rcu_read_unlock();
if (xdp_buf.data_hard_start) { if (xdp_buf.data_hard_start)
struct skb_shared_info *sinfo; mvneta_xdp_put_buff(pp, rxq, &xdp_buf, &sinfo, -1);
sinfo = xdp_get_shared_info_from_buff(&xdp_buf);
mvneta_xdp_put_buff(pp, rxq, &xdp_buf, sinfo, -1);
}
if (ps.xdp_redirect) if (ps.xdp_redirect)
xdp_do_flush_map(); xdp_do_flush_map();
......
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