bnxt_en: Fix TX timeout during netpoll.
Michael Chan
maintainerPATCH v2 1/1 bnxt_en: Fix TX timeout during netpoll.
The current netpoll implementation in the bnxt_en driver has problems that may miss TX completion events. bnxt_poll_work() in effect is only handling at most 1 TX packet before exiting. In addition, there may be in flight TX completions that ->poll() may miss even after we fix bnxt_poll_work() to handle all visible TX completions. netpoll may not call ->poll() again and HW may not generate IRQ because the driver does not ARM the IRQ when the budget (0 for netpoll) is reached.
We fix it by handling all TX completions and to always ARM the IRQ when we exit ->poll() with 0 budget.
Also, the logic to ACK the completion ring in case it is almost filled with TX completions need to be adjusted to take care of the 0 budget case, as discussed with Eric Dumazet <[email protected]>
--- drivers/net/ethernet/broadcom/bnxt/bnxt.c | 13 ++++++++++--- 1 file changed, 10 insertions(+), 3 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.c +10 -3
@@ -1884,8 +1884,11 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget)if (TX_CMP_TYPE (txcmp) ==CMP_TYPE_TX_L2_CMP ) { tx_pkts++;/* return full budget so NAPI will complete. */ -if (unlikely (tx_pkts > bp->tx_wake_thresh)) +if (unlikely (tx_pkts > bp->tx_wake_thresh)) { rx_pkts = budget; + raw_cons =NEXT_RAW_CMP (raw_cons); +break ; + } }else if ((TX_CMP_TYPE (txcmp) &0x30 ) ==0x10 ) {if (likely (budget)) rc =bnxt_rx_pkt (bp, bnapi, &raw_cons, &event);
@@ -1913,7 +1916,7 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget) } raw_cons =NEXT_RAW_CMP (raw_cons); -if (rx_pkts == budget) +if (rx_pkts && rx_pkts == budget)break ; }
@@ -2027,8 +2030,12 @@ static int bnxt_poll(struct napi_struct *napi, int budget)while (1 ) { work_done +=bnxt_poll_work (bp, bnapi, budget - work_done); -if (work_done >= budget) +if (work_done >= budget) { +if (!budget) +BNXT_CP_DB_REARM (cpr->cp_doorbell, + cpr->cp_raw_cons);break ; + }if (!bnxt_has_work (bp, cpr)) {if (napi_complete_done (napi, work_done))
Signature · 1 line
2.5.1
Quoted 16 lines · Michael Chan
The current netpoll implementation in the
bnxt_endriver has problems that may miss TX completion events.bnxt_poll_work()in effect is only handling at most 1 TX packet before exiting. In addition, there may be in flight TX completions that ->poll() may miss even after we fixbnxt_poll_work()to handle all visible TX completions. netpoll may not call ->poll() again and HW may not generate IRQ because the driver does not ARM the IRQ when the budget (0 for netpoll) is reached.We fix it by handling all TX completions and to always ARM the IRQ when we exit ->poll() with 0 budget.
Quoted 6 lines
Also, the logic to ACK the completion ring in case it is almost filled with TX completions need to be adjusted to take care of the 0 budget case, as discussed with Eric Dumazet <[email protected]>
Reported-by Song Liu <[email protected]>Signed-off-by Michael Chan <[email protected]>
Quoted 47 lines
---
drivers/net/ethernet/broadcom/bnxt/bnxt.c| 13 ++++++++++---1 file changed, 10 insertions(+), 3 deletions(-)drivers/net/ethernet/broadcom/bnxt/bnxt.c +10
-3@@ -1884,8 +1884,11 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget)if (TX_CMP_TYPE (txcmp) ==CMP_TYPE_TX_L2_CMP ) { tx_pkts++;/* return full budget so NAPI will complete. */ -if (unlikely (tx_pkts > bp->tx_wake_thresh)) +if (unlikely (tx_pkts > bp->tx_wake_thresh)) { rx_pkts = budget; + raw_cons =NEXT_RAW_CMP (raw_cons); +break ; + } }else if ((TX_CMP_TYPE (txcmp) &0x30 ) ==0x10 ) {if (likely (budget)) rc =bnxt_rx_pkt (bp, bnapi, &raw_cons, &event);@@ -1913,7 +1916,7 @@ static int bnxt_poll_work(struct bnxt *bp, struct bnxt_napi *bnapi, int budget) } raw_cons =NEXT_RAW_CMP (raw_cons); -if (rx_pkts == budget) +if (rx_pkts && rx_pkts == budget)break ; }@@ -2027,8 +2030,12 @@ static int bnxt_poll(struct napi_struct *napi, int budget)while (1 ) { work_done +=bnxt_poll_work (bp, bnapi, budget - work_done); -if (work_done >= budget) +if (work_done >= budget) { +if (!budget) +BNXT_CP_DB_REARM (cpr->cp_doorbell, + cpr->cp_raw_cons);break ; + }if (!bnxt_has_work (bp, cpr)) {if (napi_complete_done (napi, work_done))Signature · 1 line
2.5.1
David Miller
maintainerFrom: Michael Chan <[email protected]>
Date: Wed, 26 Sep 2018 00:41:04 -0400
Quoted 12 lines
The current netpoll implementation in the
bnxt_endriver has problems that may miss TX completion events.bnxt_poll_work()in effect is only handling at most 1 TX packet before exiting. In addition, there may be in flight TX completions that ->poll() may miss even after we fixbnxt_poll_work()to handle all visible TX completions. netpoll may not call ->poll() again and HW may not generate IRQ because the driver does not ARM the IRQ when the budget (0 for netpoll) is reached.We fix it by handling all TX completions and to always ARM the IRQ when we exit ->poll() with 0 budget.
Quoted 6 lines
Also, the logic to ACK the completion ring in case it is almost filled with TX completions need to be adjusted to take care of the 0 budget case, as discussed with Eric Dumazet <[email protected]>
Reported-by Song Liu <[email protected]>Signed-off-by Michael Chan <[email protected]>
Applied and queued up for -stable, thanks Michael.
Quoted 19 lines · David Miller
From: Michael Chan <[email protected]>
Date: Wed, 26 Sep 2018 00:41:04 -0400Quoted 12 lines
The current netpoll implementation in the
bnxt_endriver has problems that may miss TX completion events.bnxt_poll_work()in effect is only handling at most 1 TX packet before exiting. In addition, there may be in flight TX completions that ->poll() may miss even after we fixbnxt_poll_work()to handle all visible TX completions. netpoll may not call ->poll() again and HW may not generate IRQ because the driver does not ARM the IRQ when the budget (0 for netpoll) is reached.We fix it by handling all TX completions and to always ARM the IRQ when we exit ->poll() with 0 budget.
Quoted 8 lines
Quoted 6 lines
Also, the logic to ACK the completion ring in case it is almost filled with TX completions need to be adjusted to take care of the 0 budget case, as discussed with Eric Dumazet <[email protected]>
Reported-by Song Liu <[email protected]>Signed-off-by Michael Chan <[email protected]>Applied and queued up for -stable, thanks Michael.
Hi David,
We also need this patch from Eric:
https://marc.info/?l=linux-netdev&m=153780304905946
Thanks,
Song
Quoted 6 lines · Song Liu
We also need this patch from Eric:
I will submit this formally tomorrow, thanks.