diff options
author | Grygorii Strashko <grygorii.strashko@ti.com> | 2020-10-30 23:07:05 +0300 |
---|---|---|
committer | Jakub Kicinski <kuba@kernel.org> | 2020-11-03 03:41:07 +0300 |
commit | a9e60cf0b46044b5c40ac3cfd081eee43a647cdb (patch) | |
tree | 512a09075f861a3cee0d6cae916db63d4c78e122 /drivers/net/ethernet/ti | |
parent | 97067aaf127487788a297267dede0008cd75bb7b (diff) | |
download | linux-a9e60cf0b46044b5c40ac3cfd081eee43a647cdb.tar.xz |
net: ethernet: ti: am65-cpsw: prepare xmit/rx path for multi-port devices in mac-only mode
This patch adds multi-port support to TI AM65x CPSW driver xmit/rx path in
preparation for adding support for multi-port devices, like Main CPSW0 on
K3 J721E SoC or future CPSW3g on K3 AM64x SoC.
Hence DMA channels are common/shared for all ext Ports and the RX/TX NAPI
and DMA processing going to be assigned to first available netdev this patch:
- ensures all RX descriptors fields are initialized;
- adds synchronization for TX DMA push/pop operation (locking) as
Networking core locks are not enough any more;
- updates TX bql processing for every packet in
am65_cpsw_nuss_tx_compl_packets() as every completed TX skb can have
different ndev assigned (come from different netdevs).
To avoid performance issues for existing one-port CPSW2g devices the above
changes are done only for multi-port devices by splitting xmit path for
one-port and multi-port devices.
Signed-off-by: Grygorii Strashko <grygorii.strashko@ti.com>
Reviewed-by: Jesse Brandeburg <jesse.brandeburg@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Diffstat (limited to 'drivers/net/ethernet/ti')
-rw-r--r-- | drivers/net/ethernet/ti/am65-cpsw-nuss.c | 141 | ||||
-rw-r--r-- | drivers/net/ethernet/ti/am65-cpsw-nuss.h | 1 |
2 files changed, 108 insertions, 34 deletions
diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.c b/drivers/net/ethernet/ti/am65-cpsw-nuss.c index 2aa0c2acd059..86bfd253e295 100644 --- a/drivers/net/ethernet/ti/am65-cpsw-nuss.c +++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.c @@ -375,7 +375,7 @@ static int am65_cpsw_nuss_rx_push(struct am65_cpsw_common *common, cppi5_hdesc_init(desc_rx, CPPI5_INFO0_HDESC_EPIB_PRESENT, AM65_CPSW_NAV_PS_DATA_SIZE); - cppi5_hdesc_attach_buf(desc_rx, 0, 0, buf_dma, skb_tailroom(skb)); + cppi5_hdesc_attach_buf(desc_rx, buf_dma, skb_tailroom(skb), buf_dma, skb_tailroom(skb)); swdata = cppi5_hdesc_get_swdata(desc_rx); *((void **)swdata) = skb; @@ -911,10 +911,57 @@ static void am65_cpsw_nuss_tx_cleanup(void *data, dma_addr_t desc_dma) dev_kfree_skb_any(skb); } +static struct sk_buff * +am65_cpsw_nuss_tx_compl_packet(struct am65_cpsw_tx_chn *tx_chn, + dma_addr_t desc_dma) +{ + struct am65_cpsw_ndev_priv *ndev_priv; + struct am65_cpsw_ndev_stats *stats; + struct cppi5_host_desc_t *desc_tx; + struct net_device *ndev; + struct sk_buff *skb; + void **swdata; + + desc_tx = k3_cppi_desc_pool_dma2virt(tx_chn->desc_pool, + desc_dma); + swdata = cppi5_hdesc_get_swdata(desc_tx); + skb = *(swdata); + am65_cpsw_nuss_xmit_free(tx_chn, tx_chn->common->dev, desc_tx); + + ndev = skb->dev; + + am65_cpts_tx_timestamp(tx_chn->common->cpts, skb); + + ndev_priv = netdev_priv(ndev); + stats = this_cpu_ptr(ndev_priv->stats); + u64_stats_update_begin(&stats->syncp); + stats->tx_packets++; + stats->tx_bytes += skb->len; + u64_stats_update_end(&stats->syncp); + + return skb; +} + +static void am65_cpsw_nuss_tx_wake(struct am65_cpsw_tx_chn *tx_chn, struct net_device *ndev, + struct netdev_queue *netif_txq) +{ + if (netif_tx_queue_stopped(netif_txq)) { + /* Check whether the queue is stopped due to stalled + * tx dma, if the queue is stopped then wake the queue + * as we have free desc for tx + */ + __netif_tx_lock(netif_txq, smp_processor_id()); + if (netif_running(ndev) && + (k3_cppi_desc_pool_avail(tx_chn->desc_pool) >= MAX_SKB_FRAGS)) + netif_tx_wake_queue(netif_txq); + + __netif_tx_unlock(netif_txq); + } +} + static int am65_cpsw_nuss_tx_compl_packets(struct am65_cpsw_common *common, int chn, unsigned int budget) { - struct cppi5_host_desc_t *desc_tx; struct device *dev = common->dev; struct am65_cpsw_tx_chn *tx_chn; struct netdev_queue *netif_txq; @@ -923,15 +970,13 @@ static int am65_cpsw_nuss_tx_compl_packets(struct am65_cpsw_common *common, struct sk_buff *skb; dma_addr_t desc_dma; int res, num_tx = 0; - void **swdata; tx_chn = &common->tx_chns[chn]; while (true) { - struct am65_cpsw_ndev_priv *ndev_priv; - struct am65_cpsw_ndev_stats *stats; - + spin_lock(&tx_chn->lock); res = k3_udma_glue_pop_tx_chn(tx_chn->tx_chn, &desc_dma); + spin_unlock(&tx_chn->lock); if (res == -ENODATA) break; @@ -941,23 +986,52 @@ static int am65_cpsw_nuss_tx_compl_packets(struct am65_cpsw_common *common, break; } - desc_tx = k3_cppi_desc_pool_dma2virt(tx_chn->desc_pool, - desc_dma); - swdata = cppi5_hdesc_get_swdata(desc_tx); - skb = *(swdata); - am65_cpsw_nuss_xmit_free(tx_chn, dev, desc_tx); - + skb = am65_cpsw_nuss_tx_compl_packet(tx_chn, desc_dma); + total_bytes = skb->len; ndev = skb->dev; + napi_consume_skb(skb, budget); + num_tx++; - am65_cpts_tx_timestamp(common->cpts, skb); + netif_txq = netdev_get_tx_queue(ndev, chn); - ndev_priv = netdev_priv(ndev); - stats = this_cpu_ptr(ndev_priv->stats); - u64_stats_update_begin(&stats->syncp); - stats->tx_packets++; - stats->tx_bytes += skb->len; - u64_stats_update_end(&stats->syncp); + netdev_tx_completed_queue(netif_txq, num_tx, total_bytes); + + am65_cpsw_nuss_tx_wake(tx_chn, ndev, netif_txq); + } + dev_dbg(dev, "%s:%u pkt:%d\n", __func__, chn, num_tx); + + return num_tx; +} + +static int am65_cpsw_nuss_tx_compl_packets_2g(struct am65_cpsw_common *common, + int chn, unsigned int budget) +{ + struct device *dev = common->dev; + struct am65_cpsw_tx_chn *tx_chn; + struct netdev_queue *netif_txq; + unsigned int total_bytes = 0; + struct net_device *ndev; + struct sk_buff *skb; + dma_addr_t desc_dma; + int res, num_tx = 0; + + tx_chn = &common->tx_chns[chn]; + + while (true) { + res = k3_udma_glue_pop_tx_chn(tx_chn->tx_chn, &desc_dma); + if (res == -ENODATA) + break; + + if (cppi5_desc_is_tdcm(desc_dma)) { + if (atomic_dec_and_test(&common->tdown_cnt)) + complete(&common->tdown_complete); + break; + } + + skb = am65_cpsw_nuss_tx_compl_packet(tx_chn, desc_dma); + + ndev = skb->dev; total_bytes += skb->len; napi_consume_skb(skb, budget); num_tx++; @@ -970,19 +1044,8 @@ static int am65_cpsw_nuss_tx_compl_packets(struct am65_cpsw_common *common, netdev_tx_completed_queue(netif_txq, num_tx, total_bytes); - if (netif_tx_queue_stopped(netif_txq)) { - /* Check whether the queue is stopped due to stalled tx dma, - * if the queue is stopped then wake the queue as - * we have free desc for tx - */ - __netif_tx_lock(netif_txq, smp_processor_id()); - if (netif_running(ndev) && - (k3_cppi_desc_pool_avail(tx_chn->desc_pool) >= - MAX_SKB_FRAGS)) - netif_tx_wake_queue(netif_txq); + am65_cpsw_nuss_tx_wake(tx_chn, ndev, netif_txq); - __netif_tx_unlock(netif_txq); - } dev_dbg(dev, "%s:%u pkt:%d\n", __func__, chn, num_tx); return num_tx; @@ -993,8 +1056,11 @@ static int am65_cpsw_nuss_tx_poll(struct napi_struct *napi_tx, int budget) struct am65_cpsw_tx_chn *tx_chn = am65_cpsw_napi_to_tx_chn(napi_tx); int num_tx; - num_tx = am65_cpsw_nuss_tx_compl_packets(tx_chn->common, tx_chn->id, - budget); + if (AM65_CPSW_IS_CPSW2G(tx_chn->common)) + num_tx = am65_cpsw_nuss_tx_compl_packets_2g(tx_chn->common, tx_chn->id, budget); + else + num_tx = am65_cpsw_nuss_tx_compl_packets(tx_chn->common, tx_chn->id, budget); + num_tx = min(num_tx, budget); if (num_tx < budget) { napi_complete(napi_tx); @@ -1139,7 +1205,13 @@ done_tx: cppi5_hdesc_set_pktlen(first_desc, pkt_len); desc_dma = k3_cppi_desc_pool_virt2dma(tx_chn->desc_pool, first_desc); - ret = k3_udma_glue_push_tx_chn(tx_chn->tx_chn, first_desc, desc_dma); + if (AM65_CPSW_IS_CPSW2G(common)) { + ret = k3_udma_glue_push_tx_chn(tx_chn->tx_chn, first_desc, desc_dma); + } else { + spin_lock_bh(&tx_chn->lock); + ret = k3_udma_glue_push_tx_chn(tx_chn->tx_chn, first_desc, desc_dma); + spin_unlock_bh(&tx_chn->lock); + } if (ret) { dev_err(dev, "can't push desc %d\n", ret); /* inform bql */ @@ -1470,6 +1542,7 @@ static int am65_cpsw_nuss_init_tx_chns(struct am65_cpsw_common *common) snprintf(tx_chn->tx_chn_name, sizeof(tx_chn->tx_chn_name), "tx%d", i); + spin_lock_init(&tx_chn->lock); tx_chn->common = common; tx_chn->id = i; tx_chn->descs_num = max_desc_num; diff --git a/drivers/net/ethernet/ti/am65-cpsw-nuss.h b/drivers/net/ethernet/ti/am65-cpsw-nuss.h index b6f228ddc3a0..8e0dc5728253 100644 --- a/drivers/net/ethernet/ti/am65-cpsw-nuss.h +++ b/drivers/net/ethernet/ti/am65-cpsw-nuss.h @@ -60,6 +60,7 @@ struct am65_cpsw_tx_chn { struct am65_cpsw_common *common; struct k3_cppi_desc_pool *desc_pool; struct k3_udma_glue_tx_channel *tx_chn; + spinlock_t lock; /* protect TX rings in multi-port mode */ int irq; u32 id; u32 descs_num; |