bnxt: Tx NAPI disabling resiliency improvements
Jakub Kicinski
maintainerv2 cover bnxt: Tx NAPI disabling resiliency improvements
A lockdep warning was triggered by netpoll because napi poll was taking the xmit lock. Fix that and a couple more issues noticed while reading the code.
Jakub Kicinski (4):
bnxt: don't lock the tx queue from napi poll
bnxt: disable napi before canceling DIM
bnxt: make sure xmit_more + errors does not miss doorbells
bnxt: count Tx drops
drivers/net/ethernet/broadcom/bnxt/bnxt.c | 60 ++++++++++++++--------- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + 2 files changed, 37 insertions(+), 24 deletions(-)
Signature · 1 line
2.31.1
Jakub Kicinski
maintainerPATCH v2 1/4 bnxt: don't lock the tx queue from napi poll
We can't take the tx lock from the napi poll routine, because netpoll can poll napi at any moment, including with the tx lock already held.
It seems that the tx lock is only protecting against the disable path, appropriate barriers are already in place to make sure cleanup can safely run concurrently with start_xmit. I don't see any other reason why 'stopped && avail > thresh' needs to be re-checked under the lock.
Remove the tx lock and use synchronize_net() to make sure closing the device does not race we restarting the queues. Annotate accesses to dev_state against data races.
--
v2: keep the unlikely in bnxt_tx_int() [Edwin]
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +9 -10
@@ -731,14 +731,9 @@ static void bnxt_tx_int(struct bnxt *bp, struct bnxt_napi *bnapi, int nr_pkts)smp_mb ();if (unlikely (netif_tx_queue_stopped (txq)) && - (bnxt_tx_avail (bp, txr) > bp->tx_wake_thresh)) { -__netif_tx_lock (txq,smp_processor_id ()); -if (netif_tx_queue_stopped (txq) && -bnxt_tx_avail (bp, txr) > bp->tx_wake_thresh && - txr->dev_state !=BNXT_DEV_STATE_CLOSING ) -netif_tx_wake_queue (txq); -__netif_tx_unlock (txq); - } +bnxt_tx_avail (bp, txr) > bp->tx_wake_thresh && +READ_ONCE (txr->dev_state) !=BNXT_DEV_STATE_CLOSING ) +netif_tx_wake_queue (txq); } static struct page *__bnxt_alloc_rx_page(struct bnxt *bp, dma_addr_t *mapping,
@@ -9264,9 +9259,11 @@ void bnxt_tx_disable(struct bnxt *bp)if (bp->tx_ring) {for (i =0 ; i < bp->tx_nr_rings; i++) { txr = &bp->tx_ring[i]; - txr->dev_state =BNXT_DEV_STATE_CLOSING ; +WRITE_ONCE (txr->dev_state,BNXT_DEV_STATE_CLOSING ); } } +/* Make sure napi polls see @dev_state change */ +synchronize_net ();/* Drop carrier first to prevent TX timeout */ netif_carrier_off (bp->dev);/* Stop all TX queues */
@@ -9280,8 +9277,10 @@ void bnxt_tx_enable(struct bnxt *bp)for (i =0 ; i < bp->tx_nr_rings; i++) { txr = &bp->tx_ring[i]; - txr->dev_state =0 ; +WRITE_ONCE (txr->dev_state,0 ); } +/* Make sure napi polls see @dev_state change */ +synchronize_net ();netif_tx_wake_all_queues (bp->dev);if (bp->link_info.link_up)netif_carrier_on (bp->dev);
Signature · 1 line
2.31.1
Jakub Kicinski
maintainerPATCH v2 2/4 bnxt: disable napi before canceling DIM
In reply to Jakub Kicinskinapi schedules DIM, napi has to be disabled first, then DIM canceled.
Noticed while reading the code.
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +1 -2
@@ -9222,10 +9222,9 @@ static void bnxt_disable_napi(struct bnxt *bp)for (i =0 ; i < bp->cp_nr_rings; i++) {struct bnxt_cp_ring_info *cpr = &bp->bnapi[i]->cp_ring; +napi_disable (&bp->bnapi[i]->napi);if (bp->bnapi[i]->rx_ring)cancel_work_sync (&cpr->dim.work); - -napi_disable (&bp->bnapi[i]->napi); } }
Signature · 1 line
2.31.1
Jakub Kicinski
maintainerPATCH v2 3/4 bnxt: make sure xmit_more + errors does not miss doorbells
In reply to Jakub Kicinskiskbs are freed on error and not put on the ring. We may, however, be in a situation where we're freeing the last skb of a batch, and there is a doorbell ring pending because of xmit_more() being true earlier. Make sure we ring the door bell in such situations.
Since errors are rare don't pay attention to xmit_more() and just always flush the pending frames.
The busy case should be safe to be left alone because it can only happen if start_xmit races with completions and they both enable the queue. In that case the kick can't be pending.
Noticed while reading the code.
---
v2: - netdev_warn() -> netif_warn() [Edwin]
- use correct prod value [Michael]
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 36 +++++++++++++++-------- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + 2 files changed, 25 insertions(+), 12 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +24 -12
@@ -72,7 +72,8 @@#include "bnxt_debugfs.h" #define BNXT_TX_TIMEOUT (5 * HZ) -#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW) +#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW | \ + NETIF_MSG_TX_ERR)MODULE_LICENSE ("GPL" );MODULE_DESCRIPTION ("Broadcom BCM573xx network driver" );
@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);
@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );return NETDEV_TX_BUSY ; }
@@ -516,21 +526,16 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) normal_tx:if (length <BNXT_MIN_PKT_SIZE ) { pad =BNXT_MIN_PKT_SIZE - length; -if (skb_pad (skb, pad)) { +if (skb_pad (skb, pad))/* SKB already freed. */ - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } + goto tx_kick_pending; length =BNXT_MIN_PKT_SIZE ; } mapping =dma_map_single (&pdev->dev, skb->data, len,DMA_TO_DEVICE ); -if (unlikely (dma_mapping_error (&pdev->dev, mapping))) { -dev_kfree_skb_any (skb); - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } +if (unlikely (dma_mapping_error (&pdev->dev, mapping))) + goto tx_free;dma_unmap_addr_set (tx_buf, mapping, mapping); flags = (len << TX_BD_LEN_SHIFT) | TX_BD_TYPE_LONG_TX_BD |
@@ -617,13 +622,15 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) txr->tx_prod = prod;if (!netdev_xmit_more () ||netif_xmit_stopped (txq)) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod); +else + txr->kick_pending =1 ; tx_done:if (unlikely (bnxt_tx_avail (bp, txr) <=MAX_SKB_FRAGS +1 )) {if (netdev_xmit_more () && !tx_buf->is_push) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod);netif_tx_stop_queue (txq);
@@ -661,7 +668,12 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) PCI_DMA_TODEVICE); } +tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ; +if (txr->kick_pending) +bnxt_txr_db_kick (bp, txr, txr->tx_prod);return NETDEV_TX_OK ; }
drivers/net/ethernet/broadcom/bnxt/bnxt.h +1 -0
@@ -799,6 +799,7 @@ struct bnxt_tx_ring_info {u16 tx_prod;u16 tx_cons;u16 txq_index; +u8 kick_pending;struct bnxt_db_info tx_db;struct tx_bd *tx_desc_ring[MAX_TX_PAGES ];
Signature · 1 line
2.31.1
Michael Chan
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
Quoted 37 lines · Jakub Kicinski
---
v2: -netdev_warn()->netif_warn()[Edwin]
- use correct prod value [Michael]--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 36 +++++++++++++++-------- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + 2 files changed, 25 insertions(+), 12 deletions(-)drivers/net/ethernet/broadcom/bnxt/bnxt.c +9
-1@@ -72,7 +72,8 @@#include "bnxt_debugfs.h" #define BNXT_TX_TIMEOUT (5 * HZ) -#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW) +#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW | \ + NETIF_MSG_TX_ERR)MODULE_LICENSE ("GPL" );MODULE_DESCRIPTION ("Broadcom BCM573xx network driver" );@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 5 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );
You forgot to remove this.
Quoted 48 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +8
-11return NETDEV_TX_BUSY ; }@@ -516,21 +526,16 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) normal_tx:if (length <BNXT_MIN_PKT_SIZE ) { pad =BNXT_MIN_PKT_SIZE - length; -if (skb_pad (skb, pad)) { +if (skb_pad (skb, pad))/* SKB already freed. */ - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } + goto tx_kick_pending; length =BNXT_MIN_PKT_SIZE ; } mapping =dma_map_single (&pdev->dev, skb->data, len,DMA_TO_DEVICE ); -if (unlikely (dma_mapping_error (&pdev->dev, mapping))) { -dev_kfree_skb_any (skb); - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } +if (unlikely (dma_mapping_error (&pdev->dev, mapping))) + goto tx_free;dma_unmap_addr_set (tx_buf, mapping, mapping); flags = (len << TX_BD_LEN_SHIFT) | TX_BD_TYPE_LONG_TX_BD |@@ -617,13 +622,15 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) txr->tx_prod = prod;if (!netdev_xmit_more () ||netif_xmit_stopped (txq)) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod); +else + txr->kick_pending =1 ; tx_done:if (unlikely (bnxt_tx_avail (bp, txr) <=MAX_SKB_FRAGS +1 )) {if (netdev_xmit_more () && !tx_buf->is_push) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod);netif_tx_stop_queue (txq);
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 674
Quoted 7 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +3
-0@@ -661,7 +668,12 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) PCI_DMA_TODEVICE); } +tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ;
I think we should remove the setting of tx_buf->skb to NULL in the tx_dma_error path since we are setting it here now.
Quoted 5 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0+if (txr->kick_pending) +bnxt_txr_db_kick (bp, txr, txr->tx_prod);return NETDEV_TX_OK ; }
Jakub Kicinski
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
Quoted 38 lines · On Wed, 11 Aug 2021 15:36:34 -0700 Michael Chan wrote:
Quoted 36 lines · Jakub Kicinski
---
v2: -netdev_warn()->netif_warn()[Edwin]
- use correct prod value [Michael]--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 36 +++++++++++++++-------- drivers/net/ethernet/broadcom/bnxt/bnxt.h | 1 + 2 files changed, 25 insertions(+), 12 deletions(-)drivers/net/ethernet/broadcom/bnxt/bnxt.c +9
-1@@ -72,7 +72,8 @@#include "bnxt_debugfs.h" #define BNXT_TX_TIMEOUT (5 * HZ) -#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW) +#define BNXT_DEF_MSG_ENABLE (NETIF_MSG_DRV | NETIF_MSG_HW | \ + NETIF_MSG_TX_ERR)MODULE_LICENSE ("GPL" );MODULE_DESCRIPTION ("Broadcom BCM573xx network driver" );@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);
Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 6 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );You forgot to remove this.
I changed my mind. I added the && txr->kick_pending to the condition, if there is a race and napi starts the queue unnecessarily the kick can't be pending.
Quoted 51 lines
Quoted 51 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +8
-11return NETDEV_TX_BUSY ; }@@ -516,21 +526,16 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) normal_tx:if (length <BNXT_MIN_PKT_SIZE ) { pad =BNXT_MIN_PKT_SIZE - length; -if (skb_pad (skb, pad)) { +if (skb_pad (skb, pad))/* SKB already freed. */ - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } + goto tx_kick_pending; length =BNXT_MIN_PKT_SIZE ; } mapping =dma_map_single (&pdev->dev, skb->data, len,DMA_TO_DEVICE ); -if (unlikely (dma_mapping_error (&pdev->dev, mapping))) { -dev_kfree_skb_any (skb); - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } +if (unlikely (dma_mapping_error (&pdev->dev, mapping))) + goto tx_free;dma_unmap_addr_set (tx_buf, mapping, mapping); flags = (len << TX_BD_LEN_SHIFT) | TX_BD_TYPE_LONG_TX_BD |@@ -617,13 +622,15 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) txr->tx_prod = prod;if (!netdev_xmit_more () ||netif_xmit_stopped (txq)) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod); +else + txr->kick_pending =1 ; tx_done:if (unlikely (bnxt_tx_avail (bp, txr) <=MAX_SKB_FRAGS +1 )) {if (netdev_xmit_more () && !tx_buf->is_push) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod);netif_tx_stop_queue (txq);@@ -661,7 +668,12 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) PCI_DMA_TODEVICE); }
Quoted 7 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 674
Quoted 4 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +3
-0+tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ;I think we should remove the setting of
tx_buf->skbto NULL in thetx_dma_errorpath since we are setting it here now.
But tx_buf gets moved IIRC - if we hit tx_dma_error tx_buf will be one of the fragment bufs at this point. It should be legal to clear the skb pointer on those AFAICT.
Are you suggesting to do something along the lines of:
txr->tx_buf_ring[txr->tx_prod].skb = NULL;
?
Quoted 5 lines
Quoted 5 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0+if (txr->kick_pending) +bnxt_txr_db_kick (bp, txr, txr->tx_prod);return NETDEV_TX_OK ; }
Michael Chan
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
Quoted 26 lines · Jakub Kicinski
Quoted 15 lines · On Wed, 11 Aug 2021 15:36:34 -0700 Michael Chan wrote:
Quoted 14 lines · Jakub Kicinski
drivers/net/ethernet/broadcom/bnxt/bnxt.c +7
-0@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 6 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );You forgot to remove this.
Quoted 3 lines
I changed my mind. I added the &&
txr->kick_pendingto the condition, if there is a race and napi starts the queue unnecessarily the kick can't be pending.
I don't understand. The queue should be stopped if we have <= MAX_SKB_FRAGS + 1 descriptors left. If there is a race and the queue is awake, the first TX packet may slip through if skb_shinfo(skb)->nr_frags is small and we have enough descriptors for it. Let's say xmit_more is set for this packet and so kick is pending. The next packet may not fit anymore and it will hit this check here.
Quoted 60 lines
Quoted 52 lines
Quoted 52 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +8
-11return NETDEV_TX_BUSY ; }@@ -516,21 +526,16 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) normal_tx:if (length <BNXT_MIN_PKT_SIZE ) { pad =BNXT_MIN_PKT_SIZE - length; -if (skb_pad (skb, pad)) { +if (skb_pad (skb, pad))/* SKB already freed. */ - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } + goto tx_kick_pending; length =BNXT_MIN_PKT_SIZE ; } mapping =dma_map_single (&pdev->dev, skb->data, len,DMA_TO_DEVICE ); -if (unlikely (dma_mapping_error (&pdev->dev, mapping))) { -dev_kfree_skb_any (skb); - tx_buf->skb =NULL ; -return NETDEV_TX_OK ; - } +if (unlikely (dma_mapping_error (&pdev->dev, mapping))) + goto tx_free;dma_unmap_addr_set (tx_buf, mapping, mapping); flags = (len << TX_BD_LEN_SHIFT) | TX_BD_TYPE_LONG_TX_BD |@@ -617,13 +622,15 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) txr->tx_prod = prod;if (!netdev_xmit_more () ||netif_xmit_stopped (txq)) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod); +else + txr->kick_pending =1 ; tx_done:if (unlikely (bnxt_tx_avail (bp, txr) <=MAX_SKB_FRAGS +1 )) {if (netdev_xmit_more () && !tx_buf->is_push) -bnxt_db_write (bp, &txr->tx_db, prod); +bnxt_txr_db_kick (bp, txr, prod);netif_tx_stop_queue (txq);@@ -661,7 +668,12 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) PCI_DMA_TODEVICE); }Quoted 7 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 674
Quoted 4 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +3
-0+tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ;I think we should remove the setting of
tx_buf->skbto NULL in thetx_dma_errorpath since we are setting it here now.
Quoted 3 lines
But
tx_bufgets moved IIRC - if we hittx_dma_errortx_bufwill be one of the fragment bufs at this point. It should be legal to clear the skb pointer on those AFAICT.
Ah, you're right.
Quoted 4 lines
Are you suggesting to do something along the lines of:
txr->tx_buf_ring[txr->tx_prod].skb = NULL;
Yeah, I like this the best.
Quoted 9 lines
?
Quoted 5 lines
Quoted 5 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0+if (txr->kick_pending) +bnxt_txr_db_kick (bp, txr, txr->tx_prod);return NETDEV_TX_OK ; }
Jakub Kicinski
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
Quoted 30 lines · On Wed, 11 Aug 2021 16:00:52 -0700 Michael Chan wrote:
Quoted 25 lines · Jakub Kicinski
Quoted 15 lines · On Wed, 11 Aug 2021 15:36:34 -0700 Michael Chan wrote:
Quoted 14 lines · Jakub Kicinski
drivers/net/ethernet/broadcom/bnxt/bnxt.c +7
-0@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 6 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );You forgot to remove this.
Quoted 3 lines
I changed my mind. I added the &&
txr->kick_pendingto the condition, if there is a race and napi starts the queue unnecessarily the kick can't be pending.
Quoted 7 lines
I don't understand. The queue should be stopped if we have <=
MAX_SKB_FRAGS+ 1 descriptors left. If there is a race and the queue is awake, the first TX packet may slip through if skb_shinfo(skb)->nr_frags is small and we have enough descriptors for it. Let's sayxmit_moreis set for this packet and so kick is pending. The next packet may not fit anymore and it will hit this check here.
But even if we slip past this check we can only do it once, the check at the end of start_xmit() will see we have fewer slots than MAX_FRAGS
+ 2, ring the doorbell and stop.
Quoted 12 lines
Quoted 12 lines
Quoted 11 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 674
Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +3
-0@@ -661,7 +668,12 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) PCI_DMA_TODEVICE); } +tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ;I think we should remove the setting of
tx_buf->skbto NULL in thetx_dma_errorpath since we are setting it here now.
Quoted 5 lines
Quoted 3 lines
Are you suggesting to do something along the lines of:
txr->tx_buf_ring[txr->tx_prod].skb = NULL;Yeah, I like this the best.
Roger that, I'll send v3 tomorrow, I run out of day.
Michael Chan
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
Quoted 40 lines · Jakub Kicinski
Quoted 30 lines · On Wed, 11 Aug 2021 16:00:52 -0700 Michael Chan wrote:
Quoted 25 lines · Jakub Kicinski
Quoted 15 lines · On Wed, 11 Aug 2021 15:36:34 -0700 Michael Chan wrote:
Quoted 14 lines · Jakub Kicinski
drivers/net/ethernet/broadcom/bnxt/bnxt.c +7
-0@@ -367,6 +368,13 @@ static u16 bnxt_xmit_get_cfa_action(struct sk_buff *skb)return md_dst->u.port_info.port_id; } +static void bnxt_txr_db_kick (struct bnxt *bp,struct bnxt_tx_ring_info *txr, +u16 prod) +{ +bnxt_db_write (bp, &txr->tx_db, prod); + txr->kick_pending =0 ; +} +static netdev_tx_t bnxt_start_xmit (struct sk_buff *skb,struct net_device *dev) {struct bnxt *bp =netdev_priv (dev);Quoted 8 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 6 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );You forgot to remove this.
Quoted 3 lines
I changed my mind. I added the &&
txr->kick_pendingto the condition, if there is a race and napi starts the queue unnecessarily the kick can't be pending.Quoted 7 lines
I don't understand. The queue should be stopped if we have <=
MAX_SKB_FRAGS+ 1 descriptors left. If there is a race and the queue is awake, the first TX packet may slip through if skb_shinfo(skb)->nr_frags is small and we have enough descriptors for it. Let's sayxmit_moreis set for this packet and so kick is pending. The next packet may not fit anymore and it will hit this check here.
Quoted 3 lines
But even if we slip past this check we can only do it once, the check at the end of
start_xmit()will see we have fewer slots thanMAX_FRAGS
+ 2, ring the doorbell and stop.
Yeah, I think you're right.
Michael Chan
maintainerRe: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
In reply to Jakub Kicinskidrivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 408
Quoted 6 lines · Jakub Kicinski
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2
-0@@ -396,6 +404,8 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) free_size =bnxt_tx_avail (bp, txr);if (unlikely (free_size <skb_shinfo (skb)->nr_frags +2 )) {netif_tx_stop_queue (txq); +if (net_ratelimit () && txr->kick_pending) +netif_warn (bp, tx_err, dev,"bnxt: ring busy!\n" );
I think there is one more problem here. Now that it is possible to get here, we can race with bnxt_tx_int() again here. We can call netif_tx_stop_queue() here after bnxt_tx_int() has already cleaned the entire TX ring. So I think we need to call bnxt_tx_wake_queue() here again if descriptors have become available.
Quoted 3 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +0
-0return NETDEV_TX_BUSY ; }
RE: [PATCH net v2 3/4] bnxt: make sure xmit_more + errors does not miss doorbells
In reply to Jakub KicinskiFrom: Jakub Kicinski
Quoted 10 lines
Sent: 11 August 2021 22:38
skbs are freed on error and not put on the ring. We may, however, be in a situation where we're freeing the last skb of a batch, and there is a doorbell ring pending because of
xmit_more()being true earlier. Make sure we ring the door bell in such situations.Since errors are rare don't pay attention to
xmit_more()and just always flush the pending frames.
...
drivers/net/ethernet/broadcom/bnxt/bnxt.c static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) · line 677
Quoted 7 lines
drivers/net/ethernet/broadcom/bnxt/bnxt.c +5
-0+tx_free:dev_kfree_skb_any (skb); +tx_kick_pending: + tx_buf->skb =NULL ; +if (txr->kick_pending) +bnxt_txr_db_kick (bp, txr, txr->tx_prod);return NETDEV_TX_OK ;
Is this case actually so unlikely that the 'kick' can be done unconditionally?
Then all the conditionals can be removed from the hot path.
David
-
Registered Address Lakeside, Bramley Road, Mount Farm, Milton Keynes, MK1 1PT, UK Registration No: 1397386 (Wales)
Jakub Kicinski
maintainerPATCH v2 4/4 bnxt: count Tx drops
In reply to Jakub KicinskiDrivers should count packets they are dropping.
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +2 -0
@@ -394,6 +394,7 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) i =skb_get_queue_mapping (skb);if (unlikely (i >= bp->tx_nr_rings)) {dev_kfree_skb_any (skb); +atomic_long_inc (&dev->tx_dropped);return NETDEV_TX_OK ; }
@@ -674,6 +675,7 @@ static netdev_tx_t bnxt_start_xmit(struct sk_buff *skb, struct net_device *dev) tx_buf->skb =NULL ;if (txr->kick_pending)bnxt_txr_db_kick (bp, txr, txr->tx_prod); +atomic_long_inc (&dev->tx_dropped);return NETDEV_TX_OK ; }
Signature · 1 line
2.31.1