[dpdk-dev] [PATCH] net/ixgbe: fix too many interrupts
Zhang, Qi Z
qi.z.zhang at intel.com
Thu May 17 15:47:01 CEST 2018
> -----Original Message-----
> From: dev [mailto:dev-bounces at dpdk.org] On Behalf Of Wenzhuo Lu
> Sent: Thursday, May 17, 2018 1:34 PM
> To: dev at dpdk.org
> Cc: Lu, Wenzhuo <wenzhuo.lu at intel.com>; Luo, Michael
> <michael.luo at intel.com>
> Subject: [dpdk-dev] [PATCH] net/ixgbe: fix too many interrupts
>
> To support kernel VF, PBA bit is always set. But it may cause the too many
> interrupts issue on specific Linux kernel, like 4.4.0-116.
> PF host should set the atuo clean, mask and throttling as we always set the
> register for kernel VF.
>
> Fixes: 6b75183ac4d0 ("net/ixgbe: fix wrong PBA setting")
>
> Signed-off-by: Michael Luo <michael.luo at intel.com>
> Signed-off-by: Wenzhuo Lu <wenzhuo.lu at intel.com>
Reviewed-by: Qi Zhang <qi.z.zhang at intel.com>
> ---
> drivers/net/ixgbe/ixgbe_ethdev.c | 49
> +++++++++++++++++++++++-----------------
> 1 file changed, 28 insertions(+), 21 deletions(-)
>
> diff --git a/drivers/net/ixgbe/ixgbe_ethdev.c
> b/drivers/net/ixgbe/ixgbe_ethdev.c
> index f5006bc..7219f02 100644
> --- a/drivers/net/ixgbe/ixgbe_ethdev.c
> +++ b/drivers/net/ixgbe/ixgbe_ethdev.c
> @@ -5800,8 +5800,12 @@ static void ixgbevf_set_vfta_all(struct rte_eth_dev
> *dev, bool on)
>
> /* won't configure msix register if no mapping is done
> * between intr vector and event fd
> + * but if misx has been enabled already, need to configure
> + * auto clean, auto mask and throttling.
> */
> - if (!rte_intr_dp_is_en(intr_handle))
> + gpie = IXGBE_READ_REG(hw, IXGBE_GPIE);
> + if (!rte_intr_dp_is_en(intr_handle) &&
> + !(gpie & (IXGBE_GPIE_MSIX_MODE | IXGBE_GPIE_PBA_SUPPORT)))
> return;
>
> if (rte_intr_allow_others(intr_handle))
> @@ -5825,27 +5829,30 @@ static void ixgbevf_set_vfta_all(struct rte_eth_dev
> *dev, bool on)
> /* Populate the IVAR table and set the ITR values to the
> * corresponding register.
> */
> - for (queue_id = 0; queue_id < dev->data->nb_rx_queues;
> - queue_id++) {
> - /* by default, 1:1 mapping */
> - ixgbe_set_ivar_map(hw, 0, queue_id, vec);
> - intr_handle->intr_vec[queue_id] = vec;
> - if (vec < base + intr_handle->nb_efd - 1)
> - vec++;
> - }
> + if (rte_intr_dp_is_en(intr_handle)) {
> + for (queue_id = 0; queue_id < dev->data->nb_rx_queues;
> + queue_id++) {
> + /* by default, 1:1 mapping */
> + ixgbe_set_ivar_map(hw, 0, queue_id, vec);
> + intr_handle->intr_vec[queue_id] = vec;
> + if (vec < base + intr_handle->nb_efd - 1)
> + vec++;
> + }
>
> - switch (hw->mac.type) {
> - case ixgbe_mac_82598EB:
> - ixgbe_set_ivar_map(hw, -1, IXGBE_IVAR_OTHER_CAUSES_INDEX,
> - IXGBE_MISC_VEC_ID);
> - break;
> - case ixgbe_mac_82599EB:
> - case ixgbe_mac_X540:
> - case ixgbe_mac_X550:
> - ixgbe_set_ivar_map(hw, -1, 1, IXGBE_MISC_VEC_ID);
> - break;
> - default:
> - break;
> + switch (hw->mac.type) {
> + case ixgbe_mac_82598EB:
> + ixgbe_set_ivar_map(hw, -1,
> + IXGBE_IVAR_OTHER_CAUSES_INDEX,
> + IXGBE_MISC_VEC_ID);
> + break;
> + case ixgbe_mac_82599EB:
> + case ixgbe_mac_X540:
> + case ixgbe_mac_X550:
> + ixgbe_set_ivar_map(hw, -1, 1, IXGBE_MISC_VEC_ID);
> + break;
> + default:
> + break;
> + }
> }
> IXGBE_WRITE_REG(hw, IXGBE_EITR(IXGBE_MISC_VEC_ID),
>
> IXGBE_EITR_INTERVAL_US(IXGBE_QUEUE_ITR_INTERVAL_DEFAULT)
> --
> 1.9.3
More information about the dev
mailing list