On Sat, Dec 11, 2021 at 3:02 AM Maciej Fijalkowski
<maciej.fijalkowski@xxxxxxxxx> wrote:
>
> Follow mostly the logic from commit 9610bd988df9 ("ice: optimize XDP_TX
> workloads") that has been done in order to address the massive tx_busy
> statistic bump and improve the performance as well.
>
> One difference from 'xdpdrv' XDP_TX is when ring has less than
> ICE_TX_THRESH free entries, the cleaning routine will not stop after
> cleaning a single ICE_TX_THRESH amount of descs but rather will forward
> the next_dd pointer and check the DD bit and for this bit being set the
> cleaning will be repeated. IOW clean until there are descs that can be
> cleaned.
>
> Single instance of txonly scenario from xdpsock is improved by 20%. It
> takes four instances to achieve the line rate, which was not possible to
> achieve previously.
>
> Signed-off-by: Maciej Fijalkowski <maciej.fijalkowski@xxxxxxxxx>
> ---
> drivers/net/ethernet/intel/ice/ice_txrx.c | 2 +-
> drivers/net/ethernet/intel/ice/ice_txrx.h | 2 +-
> drivers/net/ethernet/intel/ice/ice_xsk.c | 131 ++++++++++++----------
> drivers/net/ethernet/intel/ice/ice_xsk.h | 5 +-
> 4 files changed, 73 insertions(+), 67 deletions(-)
>
> diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.c b/drivers/net/ethernet/intel/ice/ice_txrx.c
> index 227513b687b9..4b0ddb6df0c7 100644
> --- a/drivers/net/ethernet/intel/ice/ice_txrx.c
> +++ b/drivers/net/ethernet/intel/ice/ice_txrx.c
> @@ -1452,7 +1452,7 @@ int ice_napi_poll(struct napi_struct *napi, int budget)
> bool wd;
>
> if (tx_ring->xsk_pool)
> - wd = ice_clean_tx_irq_zc(tx_ring, budget);
> + wd = ice_clean_tx_irq_zc(tx_ring);
> else if (ice_ring_is_xdp(tx_ring))
> wd = true;
> else
> diff --git a/drivers/net/ethernet/intel/ice/ice_txrx.h b/drivers/net/ethernet/intel/ice/ice_txrx.h
> index b7b3bd4816f0..f2ebbe2158e7 100644
> --- a/drivers/net/ethernet/intel/ice/ice_txrx.h
> +++ b/drivers/net/ethernet/intel/ice/ice_txrx.h
> @@ -321,9 +321,9 @@ struct ice_tx_ring {
> u16 count; /* Number of descriptors */
> u16 q_index; /* Queue number of ring */
> /* stats structs */
> + struct ice_txq_stats tx_stats;
> struct ice_q_stats stats;
> struct u64_stats_sync syncp;
> - struct ice_txq_stats tx_stats;
Why this move?
> /* CL3 - 3rd cacheline starts here */
> struct rcu_head rcu; /* to avoid race on free */
> diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.c b/drivers/net/ethernet/intel/ice/ice_xsk.c
> index 925326c70701..a7f866b3fcd7 100644
> --- a/drivers/net/ethernet/intel/ice/ice_xsk.c
> +++ b/drivers/net/ethernet/intel/ice/ice_xsk.c
> @@ -613,55 +613,68 @@ int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget)
> /**
> * ice_xmit_zc - Completes AF_XDP entries, and cleans XDP entries
> * @xdp_ring: XDP Tx ring
> - * @budget: max number of frames to xmit
> *
> * Returns true if cleanup/transmission is done.
> */
> -static bool ice_xmit_zc(struct ice_tx_ring *xdp_ring, int budget)
> +static bool ice_xmit_zc(struct ice_tx_ring *xdp_ring)
> {
> + int total_packets = 0, total_bytes = 0;
Why not an unsigned type like u32?
> struct ice_tx_desc *tx_desc = NULL;
> - bool work_done = true;
> + u16 ntu = xdp_ring->next_to_use;
> struct xdp_desc desc;
> dma_addr_t dma;
> + u16 budget = 0;
>
> - while (likely(budget-- > 0)) {
> - struct ice_tx_buf *tx_buf;
> -
> - if (unlikely(!ICE_DESC_UNUSED(xdp_ring))) {
> - xdp_ring->tx_stats.tx_busy++;
> - work_done = false;
> - break;
> - }
> -
> - tx_buf = &xdp_ring->tx_buf[xdp_ring->next_to_use];
> + budget = ICE_DESC_UNUSED(xdp_ring);
> + if (unlikely(!budget)) {
> + xdp_ring->tx_stats.tx_busy++;
> + return false;
> + }
>
> + while (budget-- > 0) {
> if (!xsk_tx_peek_desc(xdp_ring->xsk_pool, &desc))
> break;
>
> + total_packets++;
> dma = xsk_buff_raw_get_dma(xdp_ring->xsk_pool, desc.addr);
> xsk_buff_raw_dma_sync_for_device(xdp_ring->xsk_pool, dma,
> desc.len);
>
> - tx_buf->bytecount = desc.len;
> + total_bytes += desc.len;
>
> - tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_to_use);
> + tx_desc = ICE_TX_DESC(xdp_ring, ntu);
> tx_desc->buf_addr = cpu_to_le64(dma);
> tx_desc->cmd_type_offset_bsz =
> - ice_build_ctob(ICE_TXD_LAST_DESC_CMD, 0, desc.len, 0);
> -
> - xdp_ring->next_to_use++;
> - if (xdp_ring->next_to_use == xdp_ring->count)
> - xdp_ring->next_to_use = 0;
> + ice_build_ctob(ICE_TX_DESC_CMD_EOP, 0, desc.len, 0);
> +
> + ntu++;
> + if (ntu == xdp_ring->count) {
> + ntu = 0;
> + tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_rs);
> + tx_desc->cmd_type_offset_bsz |=
> + cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S);
> + xdp_ring->next_rs = ICE_TX_THRESH - 1;
> + }
> + if (ntu > xdp_ring->next_rs) {
> + tx_desc = ICE_TX_DESC(xdp_ring, xdp_ring->next_rs);
> + tx_desc->cmd_type_offset_bsz |=
> + cpu_to_le64(ICE_TX_DESC_CMD_RS << ICE_TXD_QW1_CMD_S);
> + xdp_ring->next_rs += ICE_TX_THRESH;
> + }
> }
>
> + xdp_ring->next_to_use = ntu;
> if (tx_desc) {
> ice_xdp_ring_update_tail(xdp_ring);
> xsk_tx_release(xdp_ring->xsk_pool);
> }
>
> - return budget > 0 && work_done;
> -}
> + if (xsk_uses_need_wakeup(xdp_ring->xsk_pool))
> + xsk_set_tx_need_wakeup(xdp_ring->xsk_pool);
> + ice_update_tx_ring_stats(xdp_ring, total_packets, total_bytes);
>
> + return budget > 0;
> +}
> /**
> * ice_clean_xdp_tx_buf - Free and unmap XDP Tx buffer
> * @xdp_ring: XDP Tx ring
> @@ -679,30 +692,31 @@ ice_clean_xdp_tx_buf(struct ice_tx_ring *xdp_ring, struct ice_tx_buf *tx_buf)
> /**
> * ice_clean_tx_irq_zc - Completes AF_XDP entries, and cleans XDP entries
> * @xdp_ring: XDP Tx ring
> - * @budget: NAPI budget
> *
> * Returns true if cleanup/tranmission is done.
Nit and not your fault, but it should be transmission with an "s". If
there is a v2, you might as well change it.
> */
> -bool ice_clean_tx_irq_zc(struct ice_tx_ring *xdp_ring, int budget)
> +bool ice_clean_tx_irq_zc(struct ice_tx_ring *xdp_ring)
> {
> - int total_packets = 0, total_bytes = 0;
> - s16 ntc = xdp_ring->next_to_clean;
> + u16 next_dd = xdp_ring->next_dd;
> + u16 desc_cnt = xdp_ring->count;
> struct ice_tx_desc *tx_desc;
> struct ice_tx_buf *tx_buf;
> - u32 xsk_frames = 0;
> - bool xmit_done;
> + u32 xsk_frames;
> + u16 ntc;
> + int i;
u32 since it can never be negative?
>
> - tx_desc = ICE_TX_DESC(xdp_ring, ntc);
> - tx_buf = &xdp_ring->tx_buf[ntc];
> - ntc -= xdp_ring->count;
> + tx_desc = ICE_TX_DESC(xdp_ring, next_dd);
>
> - do {
> - if (!(tx_desc->cmd_type_offset_bsz &
> - cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE)))
> - break;
> + if (!(tx_desc->cmd_type_offset_bsz &
> + cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE)))
> + return ice_xmit_zc(xdp_ring);
>
> - total_bytes += tx_buf->bytecount;
> - total_packets++;
> +again:
> + xsk_frames = 0;
> + ntc = xdp_ring->next_to_clean;
> +
> + for (i = 0; i < ICE_TX_THRESH; i++) {
> + tx_buf = &xdp_ring->tx_buf[ntc];
>
> if (tx_buf->raw_buf) {
> ice_clean_xdp_tx_buf(xdp_ring, tx_buf);
> @@ -711,34 +725,27 @@ bool ice_clean_tx_irq_zc(struct ice_tx_ring *xdp_ring, int budget)
> xsk_frames++;
> }
>
> - tx_desc->cmd_type_offset_bsz = 0;
> - tx_buf++;
> - tx_desc++;
> ntc++;
> -
> - if (unlikely(!ntc)) {
> - ntc -= xdp_ring->count;
> - tx_buf = xdp_ring->tx_buf;
> - tx_desc = ICE_TX_DESC(xdp_ring, 0);
> - }
> -
> - prefetch(tx_desc);
Did the removal of the prefetch improve performance?
> -
> - } while (likely(--budget));
> -
> - ntc += xdp_ring->count;
> - xdp_ring->next_to_clean = ntc;
> -
> + if (ntc >= xdp_ring->count)
> + ntc = 0;
> + }
> + xdp_ring->next_to_clean += ICE_TX_THRESH;
> + if (xdp_ring->next_to_clean >= desc_cnt)
> + xdp_ring->next_to_clean -= desc_cnt;
> if (xsk_frames)
> xsk_tx_completed(xdp_ring->xsk_pool, xsk_frames);
> -
> - if (xsk_uses_need_wakeup(xdp_ring->xsk_pool))
> - xsk_set_tx_need_wakeup(xdp_ring->xsk_pool);
> -
> - ice_update_tx_ring_stats(xdp_ring, total_packets, total_bytes);
> - xmit_done = ice_xmit_zc(xdp_ring, ICE_DFLT_IRQ_WORK);
> -
> - return budget > 0 && xmit_done;
> + tx_desc->cmd_type_offset_bsz = 0;
> + next_dd += ICE_TX_THRESH;
> + if (next_dd > desc_cnt)
> + next_dd = ICE_TX_THRESH - 1;
> +
> + tx_desc = ICE_TX_DESC(xdp_ring, next_dd);
> + if ((tx_desc->cmd_type_offset_bsz &
> + cpu_to_le64(ICE_TX_DESC_DTYPE_DESC_DONE)))
> + goto again;
> + xdp_ring->next_dd = next_dd;
> +
> + return ice_xmit_zc(xdp_ring);
> }
>
> /**
> diff --git a/drivers/net/ethernet/intel/ice/ice_xsk.h b/drivers/net/ethernet/intel/ice/ice_xsk.h
> index 4c7bd8e9dfc4..1f98ad090f89 100644
> --- a/drivers/net/ethernet/intel/ice/ice_xsk.h
> +++ b/drivers/net/ethernet/intel/ice/ice_xsk.h
> @@ -12,7 +12,7 @@ struct ice_vsi;
> int ice_xsk_pool_setup(struct ice_vsi *vsi, struct xsk_buff_pool *pool,
> u16 qid);
> int ice_clean_rx_irq_zc(struct ice_rx_ring *rx_ring, int budget);
> -bool ice_clean_tx_irq_zc(struct ice_tx_ring *xdp_ring, int budget);
> +bool ice_clean_tx_irq_zc(struct ice_tx_ring *xdp_ring);
> int ice_xsk_wakeup(struct net_device *netdev, u32 queue_id, u32 flags);
> bool ice_alloc_rx_bufs_zc(struct ice_rx_ring *rx_ring, u16 count);
> bool ice_xsk_any_rx_ring_ena(struct ice_vsi *vsi);
> @@ -35,8 +35,7 @@ ice_clean_rx_irq_zc(struct ice_rx_ring __always_unused *rx_ring,
> }
>
> static inline bool
> -ice_clean_tx_irq_zc(struct ice_tx_ring __always_unused *xdp_ring,
> - int __always_unused budget)
> +ice_clean_tx_irq_zc(struct ice_tx_ring __always_unused *xdp_ring)
> {
> return false;
> }
> --
> 2.33.1
>
