| Message ID | 20211027120447.28075-1-mateusz.palczewski@intel.com |
|---|---|
| State | Changes Requested |
| Delegated to: | Anthony Nguyen |
| Headers | show |
| Series | [net,v1] i40e: Fix VF failed to init adminq: -53 | expand |
> -----Original Message----- > From: Intel-wired-lan <intel-wired-lan-bounces@osuosl.org> On Behalf Of > Mateusz Palczewski > Sent: środa, 27 października 2021 14:05 > To: intel-wired-lan@lists.osuosl.org > Cc: Palczewski, Mateusz <mateusz.palczewski@intel.com>; Zulinski, > NorbertX <norbertx.zulinski@intel.com> > Subject: [Intel-wired-lan] [PATCH net v1] i40e: Fix VF failed to init adminq: -53 > > From: Norbert Zulinski <norbertx.zulinski@intel.com> > > Fix the problem with init adminq in VF reset handler. > When the PF finished reinitialize VF resource set VFR_VFACTIVE bit in VF > Reset Status register. It is sign for VF driver among others to shut down and > reinitialize the admin queue. VF handle reset procedure is sampling this > register to check every 10ms. PF driver give up to 20ms to VF reset > procedure. For a single VF reset it is enoungh time to do it but in case request > VF reset twice, the first VF reset can be not completely finished when PF > requests it one more time. > Fixed by adding additional time for VF to finish reset procedure before > sending next VF reset request by PF. > > Fixes: 5c3c48ac6bf5 ("i40e: implement virtual device interface") > Signed-off-by: Grzegorz Szczurek <grzegorzx.szczurek@intel.com> > Signed-off-by: Norbert Zulinski <norbertx.zulinski@intel.com> > Signed-off-by: Mateusz Palczewski <mateusz.palczewski@intel.com> > --- > drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c | 4 ++++ > drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h | 2 ++ > 2 files changed, 6 insertions(+) > > diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c > b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c > index 5a488ce..ec0e52c 100644 > --- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c > +++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c index 5a488ce..ec0e52c 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c +++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.c @@ -172,6 +172,9 @@ void i40e_vc_notify_vf_reset(struct i40e_vf *vf) !test_bit(I40E_VF_STATE_ACTIVE, &vf->vf_states)) return; + if (ktime_get_ns() - vf->reset_timestamp < I40E_VF_RESET_TIME_MIN) + usleep_range(30000, 60000); + abs_vf_id = vf->vf_id + (int)vf->pf->hw.func_caps.vf_base_id; pfe.event = VIRTCHNL_EVENT_RESET_IMPENDING; @@ -1536,6 +1539,7 @@ bool i40e_reset_vf(struct i40e_vf *vf, bool flr) i40e_cleanup_reset_vf(vf); i40e_flush(hw); + vf->reset_timestamp = ktime_get_ns(); clear_bit(__I40E_VF_DISABLE, pf->state); return true; diff --git a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h index 49575a6..0e7d555 100644 --- a/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h +++ b/drivers/net/ethernet/intel/i40e/i40e_virtchnl_pf.h @@ -19,6 +19,7 @@ #define I40E_MAX_VF_PROMISC_FLAGS 3 #define I40E_VF_STATE_WAIT_COUNT 20 +#define I40E_VF_RESET_TIME_MIN 30000000 // time in nsec /* Various queue ctrls */ enum i40e_queue_ctrl { @@ -80,6 +81,7 @@ struct i40e_vf { u16 port_vlan_id; bool pf_set_mac; /* The VMM admin set the VF MAC address */ bool trusted; + u64 reset_timestamp; /* VSI indices - actual VSI pointers are maintained in the PF structure * When assigned, these will be non-zero, because VSI 0 is always