[PATCH v5] net/i40e: rework maximum frame size configuration

[David Marchand]

On Fri, Feb 3, 2023 at 9:39 AM Yang, Qiming <qiming.yang at intel.com> wrote:
>
>
>
> > -----Original Message-----
> > From: David Marchand <david.marchand at redhat.com>
> > Sent: Thursday, February 2, 2023 9:49 PM
> > To: Su, Simei <simei.su at intel.com>; Zhang, Yuying
> > <yuying.zhang at intel.com>; Xing, Beilei <beilei.xing at intel.com>; Zhang, Qi Z
> > <qi.z.zhang at intel.com>
> > Cc: dev at dpdk.org; Yang, Qiming <qiming.yang at intel.com>;
> > stable at dpdk.org
> > Subject: Re: [PATCH v5] net/i40e: rework maximum frame size configuration
> >
> > On Thu, Feb 2, 2023 at 2:24 PM David Marchand
> > <david.marchand at redhat.com> wrote:
> > >
> > > On Thu, Feb 2, 2023 at 1:37 PM Simei Su <simei.su at intel.com> wrote:
> > > > @@ -2467,8 +2466,16 @@ i40e_dev_start(struct rte_eth_dev *dev)
> > > >                             "please call hierarchy_commit() "
> > > >                             "before starting the port");
> > > >
> > > > -       max_frame_size = dev->data->mtu + I40E_ETH_OVERHEAD;
> > > > -       i40e_set_mac_max_frame(dev, max_frame_size);
> > > > +       max_frame_size = dev->data->mtu ?
> > > > +               dev->data->mtu + I40E_ETH_OVERHEAD :
> > > > +               I40E_FRAME_SIZE_MAX;
> > > > +
> > > > +       /* Set the max frame size to HW*/
> > > > +       ret = i40e_aq_set_mac_config(hw, max_frame_size, TRUE, false, 0,
> > NULL);
> > > > +       if (ret) {
> > > > +               PMD_DRV_LOG(ERR, "Fail to set mac config");
> > > > +               return ret;
> > > > +       }
> > >
> > > Reading this patch again.
> > >
> > > Returning here seems incorrect as we leave rx/tx queue in started state.
> > > Don't we need to jump to tx_err label on error?
> >
> > There is probably an issue with interrupt handler still being registered too.
> > Qi, i40e maintainers, please review this patch carefully, and ping me when it
> > is ready so that we can test it at RH.
> >
>
> This change will not break interrupt handler, the second parameter is waiting to complete. Just waiting more time to make sure adminq command execute completed. So that subsequent commands(MTU set) can be executed.
> And if you have other issues report at RH system, please report it.

If i40e_aq_set_mac_config() fails and we return directly, an interrupt
handler is left registered, don't you think it is an issue?

In any case, I will wait for the next revision before asking for tests
on my side.
Thanks.


-- 
David Marchand