[v4,2/2] net/ice: fix flow redirect failure
Checks
Commit Message
From: Dapeng Yu <dapengx.yu@intel.com>
It's possible that a switch rule can't be redirect successfully due
to kernel driver is busy to handle an ongoing VF reset, so the
redirect action need to be deferred into next redirect request which
is promised by kernel driver after VF reset done.
This patch uses the saved flow rule's data to replay switch rule
remove/add during next flow redirect.
Fixes: 397b4b3c5095 ("net/ice: enable flow redirect on switch")
Cc: stable@dpdk.org
Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
---
V2:
* Add more filter status and VSI number
V3:
* Use switch statement to make code clear
V4:
* Trim leading space
---
drivers/net/ice/ice_switch_filter.c | 108 ++++++++++++++++++++--------
1 file changed, 78 insertions(+), 30 deletions(-)
Comments
> -----Original Message-----
> From: Yu, DapengX <dapengx.yu@intel.com>
> Sent: Thursday, November 4, 2021 4:46 PM
> To: Yang, Qiming <qiming.yang@intel.com>; Zhang, Qi Z
> <qi.z.zhang@intel.com>
> Cc: dev@dpdk.org; Wang, Haiyue <haiyue.wang@intel.com>; Yu, DapengX
> <dapengx.yu@intel.com>; stable@dpdk.org
> Subject: [PATCH v4 2/2] net/ice: fix flow redirect failure
>
> From: Dapeng Yu <dapengx.yu@intel.com>
>
> It's possible that a switch rule can't be redirect successfully due to kernel
> driver is busy to handle an ongoing VF reset, so the redirect action need to be
> deferred into next redirect request which is promised by kernel driver after VF
> reset done.
>
> This patch uses the saved flow rule's data to replay switch rule remove/add
> during next flow redirect.
>
> Fixes: 397b4b3c5095 ("net/ice: enable flow redirect on switch")
> Cc: stable@dpdk.org
>
> Signed-off-by: Dapeng Yu <dapengx.yu@intel.com>
Acked-by: Qi Zhang <qi.z.zhang@intel.com>
Applied to dpdk-next-net-intel.
Thanks
Qi
@@ -1926,8 +1926,12 @@ ice_switch_redirect(struct ice_adapter *ad,
struct rte_flow *flow,
struct ice_flow_redirect *rd)
{
- struct ice_rule_query_data *rdata = flow->rule;
+ struct ice_rule_query_data *rdata;
+ struct ice_switch_filter_conf *filter_conf_ptr =
+ (struct ice_switch_filter_conf *)flow->rule;
+ struct ice_rule_query_data added_rdata = { 0 };
struct ice_adv_fltr_mgmt_list_entry *list_itr;
+ struct ice_adv_lkup_elem *lkups_ref = NULL;
struct ice_adv_lkup_elem *lkups_dp = NULL;
struct LIST_HEAD_TYPE *list_head;
struct ice_adv_rule_info rinfo;
@@ -1936,6 +1940,8 @@ ice_switch_redirect(struct ice_adapter *ad,
uint16_t lkups_cnt;
int ret;
+ rdata = &filter_conf_ptr->sw_query_data;
+
if (rdata->vsi_handle != rd->vsi_handle)
return 0;
@@ -1946,56 +1952,98 @@ ice_switch_redirect(struct ice_adapter *ad,
if (rd->type != ICE_FLOW_REDIRECT_VSI)
return -ENOTSUP;
- list_head = &sw->recp_list[rdata->rid].filt_rules;
- LIST_FOR_EACH_ENTRY(list_itr, list_head, ice_adv_fltr_mgmt_list_entry,
- list_entry) {
- rinfo = list_itr->rule_info;
- if ((rinfo.fltr_rule_id == rdata->rule_id &&
- rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI &&
- rinfo.sw_act.vsi_handle == rd->vsi_handle) ||
- (rinfo.fltr_rule_id == rdata->rule_id &&
- rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI_LIST)){
- lkups_cnt = list_itr->lkups_cnt;
- lkups_dp = (struct ice_adv_lkup_elem *)
- ice_memdup(hw, list_itr->lkups,
- sizeof(*list_itr->lkups) *
- lkups_cnt, ICE_NONDMA_TO_NONDMA);
-
- if (!lkups_dp) {
- PMD_DRV_LOG(ERR, "Failed to allocate memory.");
- return -EINVAL;
- }
+ switch (filter_conf_ptr->fltr_status) {
+ case ICE_SW_FLTR_ADDED:
+ list_head = &sw->recp_list[rdata->rid].filt_rules;
+ LIST_FOR_EACH_ENTRY(list_itr, list_head,
+ ice_adv_fltr_mgmt_list_entry,
+ list_entry) {
+ rinfo = list_itr->rule_info;
+ if ((rinfo.fltr_rule_id == rdata->rule_id &&
+ rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI &&
+ rinfo.sw_act.vsi_handle == rd->vsi_handle) ||
+ (rinfo.fltr_rule_id == rdata->rule_id &&
+ rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI_LIST)){
+ lkups_cnt = list_itr->lkups_cnt;
+
+ lkups_dp = (struct ice_adv_lkup_elem *)
+ ice_memdup(hw, list_itr->lkups,
+ sizeof(*list_itr->lkups) *
+ lkups_cnt,
+ ICE_NONDMA_TO_NONDMA);
+ if (!lkups_dp) {
+ PMD_DRV_LOG(ERR,
+ "Failed to allocate memory.");
+ return -EINVAL;
+ }
+ lkups_ref = lkups_dp;
- if (rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI_LIST) {
- rinfo.sw_act.vsi_handle = rd->vsi_handle;
- rinfo.sw_act.fltr_act = ICE_FWD_TO_VSI;
+ if (rinfo.sw_act.fltr_act ==
+ ICE_FWD_TO_VSI_LIST) {
+ rinfo.sw_act.vsi_handle =
+ rd->vsi_handle;
+ rinfo.sw_act.fltr_act = ICE_FWD_TO_VSI;
+ }
+ break;
}
- break;
}
- }
- if (!lkups_dp)
+ if (!lkups_ref)
+ return -EINVAL;
+
+ goto rmv_rule;
+ case ICE_SW_FLTR_RMV_FAILED_ON_RIDRECT:
+ /* Recover VSI context */
+ hw->vsi_ctx[rd->vsi_handle]->vsi_num = filter_conf_ptr->vsi_num;
+ rinfo = filter_conf_ptr->rule_info;
+ lkups_cnt = filter_conf_ptr->lkups_num;
+ lkups_ref = filter_conf_ptr->lkups;
+
+ if (rinfo.sw_act.fltr_act == ICE_FWD_TO_VSI_LIST) {
+ rinfo.sw_act.vsi_handle = rd->vsi_handle;
+ rinfo.sw_act.fltr_act = ICE_FWD_TO_VSI;
+ }
+
+ goto rmv_rule;
+ case ICE_SW_FLTR_ADD_FAILED_ON_RIDRECT:
+ rinfo = filter_conf_ptr->rule_info;
+ lkups_cnt = filter_conf_ptr->lkups_num;
+ lkups_ref = filter_conf_ptr->lkups;
+
+ goto add_rule;
+ default:
return -EINVAL;
+ }
+rmv_rule:
/* Remove the old rule */
- ret = ice_rem_adv_rule(hw, list_itr->lkups,
- lkups_cnt, &rinfo);
+ ret = ice_rem_adv_rule(hw, lkups_ref, lkups_cnt, &rinfo);
if (ret) {
PMD_DRV_LOG(ERR, "Failed to delete the old rule %d",
rdata->rule_id);
+ filter_conf_ptr->fltr_status =
+ ICE_SW_FLTR_RMV_FAILED_ON_RIDRECT;
ret = -EINVAL;
goto out;
}
+add_rule:
/* Update VSI context */
hw->vsi_ctx[rd->vsi_handle]->vsi_num = rd->new_vsi_num;
/* Replay the rule */
- ret = ice_add_adv_rule(hw, lkups_dp, lkups_cnt,
- &rinfo, rdata);
+ ret = ice_add_adv_rule(hw, lkups_ref, lkups_cnt,
+ &rinfo, &added_rdata);
if (ret) {
PMD_DRV_LOG(ERR, "Failed to replay the rule");
+ filter_conf_ptr->fltr_status =
+ ICE_SW_FLTR_ADD_FAILED_ON_RIDRECT;
ret = -EINVAL;
+ } else {
+ filter_conf_ptr->sw_query_data = added_rdata;
+ /* Save VSI number for failure recover */
+ filter_conf_ptr->vsi_num = rd->new_vsi_num;
+ filter_conf_ptr->fltr_status = ICE_SW_FLTR_ADDED;
}
out: