> Introduced through customer's feedback in their attempt to address > some bugs this introduces a memory barrier before posting ctlq tail. > This makes sure memory writes have a chance to take place before HW > starts messing with the descriptors. > > Signed-off-by: Soumyadeep Hore <soumyadeep.h...@intel.com> > ---
>From the description, it seems that this may be a bugfix patch. Can you >confirm this and whether it should be backported or not. Also, provide a fixes >tag for this. Thanks, /Bruce This is a new feature supported in MEV TS 1.4 which is only compatible with dpdk 24.07. So previous dpdk versions are not affected, hence backporting is not required. Regards, Soumya > drivers/common/idpf/base/idpf_controlq.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/drivers/common/idpf/base/idpf_controlq.c > b/drivers/common/idpf/base/idpf_controlq.c > index 65e5599614..4f47759a4f 100644 > --- a/drivers/common/idpf/base/idpf_controlq.c > +++ b/drivers/common/idpf/base/idpf_controlq.c > @@ -604,6 +604,8 @@ int idpf_ctlq_post_rx_buffs(struct idpf_hw *hw, struct > idpf_ctlq_info *cq, > /* Wrap to end of end ring since current ntp is 0 */ > cq->next_to_post = cq->ring_size - 1; > > + idpf_wmb(); > + > wr32(hw, cq->reg.tail, cq->next_to_post); > } > > -- > 2.43.0 >
