[PATCH v3 2/2] net: nps_enet: bug fix - handle lost tx interrupts
From: Elad Kanfi <hidden>
Date: 2016-05-09 17:47:40
Also in:
lkml
Subsystem:
networking drivers, the rest · Maintainers:
Andrew Lunn, "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Linus Torvalds
From: Elad Kanfi <redacted> The tx interrupt is of edge type, and in case such interrupt is triggered while it is masked it will not be handled even after tx interrupts are re-enabled in the end of NAPI poll. This will cause tx network to stop in the following scenario: * Rx is being handled, hence interrupts are masked. * Tx interrupt is triggered after checking if there is some tx to handle and before re-enabling the interrupts. In this situation only rx transaction will release tx requests. In order to handle the tx that was missed( if there was one ), a NAPI reschdule was added after enabling the interrupts. Signed-off-by: Elad Kanfi <redacted> Acked-by: Noam Camus <redacted> Acked-by: Gilad Ben-Yossef <redacted> --- drivers/net/ethernet/ezchip/nps_enet.c | 15 +++++++++++++++ 1 files changed, 15 insertions(+), 0 deletions(-)
diff --git a/drivers/net/ethernet/ezchip/nps_enet.c b/drivers/net/ethernet/ezchip/nps_enet.c
index 25ac2de..085f912 100644
--- a/drivers/net/ethernet/ezchip/nps_enet.c
+++ b/drivers/net/ethernet/ezchip/nps_enet.c@@ -183,6 +183,9 @@ static int nps_enet_poll(struct napi_struct *napi, int budget) work_done = nps_enet_rx_handler(ndev); if (work_done < budget) { u32 buf_int_enable_value = 0; + u32 tx_ctrl_value = nps_enet_reg_get(priv, NPS_ENET_REG_TX_CTL); + u32 tx_ctrl_ct = + (tx_ctrl_value & TX_CTL_CT_MASK) >> TX_CTL_CT_SHIFT; napi_complete(napi);
@@ -192,6 +195,18 @@ static int nps_enet_poll(struct napi_struct *napi, int budget) nps_enet_reg_set(priv, NPS_ENET_REG_BUF_INT_ENABLE, buf_int_enable_value); + + /* in case we will get a tx interrupt while interrupts + * are masked, we will lose it since the tx is edge interrupt. + * specifically, while executing the code section above, + * between nps_enet_tx_handler and the interrupts enable, all + * tx requests will be stuck until we will get an rx interrupt. + * the two code lines below will solve this situation by + * re-adding ourselves to the poll list. + */ + + if (priv->tx_skb && !tx_ctrl_ct) + napi_reschedule(napi); } return work_done;
--
1.7.1