[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <3830bef7b52414e3a0b874d3fd23d7e8bc4c1c2f.camel@redhat.com>
Date: Thu, 21 Dec 2023 08:23:39 +0100
From: Paolo Abeni <pabeni@...hat.com>
To: "Nelson, Shannon" <shannon.nelson@....com>, Larysa Zaremba
<larysa.zaremba@...el.com>
Cc: Tony Nguyen <anthony.l.nguyen@...el.com>, davem@...emloft.net,
kuba@...nel.org, edumazet@...gle.com, netdev@...r.kernel.org,
maciej.fijalkowski@...el.com, magnus.karlsson@...el.com, ast@...nel.org,
daniel@...earbox.net, hawk@...nel.org, john.fastabend@...il.com,
bpf@...r.kernel.org, Przemek Kitszel <przemyslaw.kitszel@...el.com>, Simon
Horman <horms@...nel.org>, Chandan Kumar Rout <chandanx.rout@...el.com>
Subject: Re: [PATCH net 3/3] ice: Fix PF with enabled XDP going no-carrier
after reset
On Wed, 2023-12-20 at 09:04 -0800, Nelson, Shannon wrote:
> On 12/20/2023 1:23 AM, Larysa Zaremba wrote:
> >
> > On Tue, Dec 19, 2023 at 04:09:09PM -0800, Nelson, Shannon wrote:
> > > On 12/18/2023 11:27 AM, Tony Nguyen wrote:
> > > > Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.
> > > >
> > > >
> > > > From: Larysa Zaremba <larysa.zaremba@...el.com>
> > > >
> > > > Commit 6624e780a577fc596788 ("ice: split ice_vsi_setup into smaller
> > > > functions") has refactored a bunch of code involved in PFR. In this
> > > > process, TC queue number adjustment for XDP was lost. Bring it back.
> > > >
> > > > Lack of such adjustment causes interface to go into no-carrier after a
> > > > reset, if XDP program is attached, with the following message:
> > > >
> > > > ice 0000:b1:00.0: Failed to set LAN Tx queue context, error: -22
> > > > ice 0000:b1:00.0 ens801f0np0: Failed to open VSI 0x0006 on switch 0x0001
> > > > ice 0000:b1:00.0: enable VSI failed, err -22, VSI index 0, type ICE_VSI_PF
> > > > ice 0000:b1:00.0: PF VSI rebuild failed: -22
> > > > ice 0000:b1:00.0: Rebuild failed, unload and reload driver
> > > >
> > > > Fixes: 6624e780a577 ("ice: split ice_vsi_setup into smaller functions")
> > > > Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@...el.com>
> > > > Signed-off-by: Larysa Zaremba <larysa.zaremba@...el.com>
> > > > Reviewed-by: Simon Horman <horms@...nel.org>
> > > > Tested-by: Chandan Kumar Rout <chandanx.rout@...el.com> (A Contingent Worker at Intel)
> > > > Signed-off-by: Tony Nguyen <anthony.l.nguyen@...el.com>
> > > > ---
> > > > drivers/net/ethernet/intel/ice/ice_lib.c | 3 +++
> > > > 1 file changed, 3 insertions(+)
> > > >
> > > > diff --git a/drivers/net/ethernet/intel/ice/ice_lib.c b/drivers/net/ethernet/intel/ice/ice_lib.c
> > > > index de7ba87af45d..1bad6e17f9be 100644
> > > > --- a/drivers/net/ethernet/intel/ice/ice_lib.c
> > > > +++ b/drivers/net/ethernet/intel/ice/ice_lib.c
> > > > @@ -2371,6 +2371,9 @@ static int ice_vsi_cfg_tc_lan(struct ice_pf *pf, struct ice_vsi *vsi)
> > > > } else {
> > > > max_txqs[i] = vsi->alloc_txq;
> > > > }
> > > > +
> > > > + if (vsi->type == ICE_VSI_PF)
> > > > + max_txqs[i] += vsi->num_xdp_txq;
> > >
> > > Since this new code is coming right after an existing
> > > if (vsi->type == ICE_VSI_CHNL)
> > > it looks like it would make sense to make it an 'else if' in that last
> > > block, e.g.:
> > >
> > > if (vsi->type == ICE_VSI_CHNL) {
> > > if (!vsi->alloc_txq && vsi->num_txq)
> > > max_txqs[i] = vsi->num_txq;
> > > else
> > > max_txqs[i] = pf->num_lan_tx;
> > > } else if (vsi->type == ICE_VSI_PF) {
> > > max_txqs[i] += vsi->num_xdp_txq;
> >
> > Would need to be
> > max_txqs[i] = vsi->alloc_txq + vsi->num_xdp_txq;
> >
> > > } else {
> > > max_txqs[i] = vsi->alloc_txq;
> > > }
> > >
> > > Of course this begins to verge on the switch/case/default format.
> > >
> > > sln
> > >
> >
> > I was going for logic: assign default values first, adjust based on enabled
> > features (well, a single feature) second. The thing that in my opinion would
> > make it more clear would be replacing 'vsi->type == ICE_VSI_PF' with
> > ice_is_xdp_ena_vsi(). Do you think this is worth doing?
>
> Hmm... I made a dumb error in a quick read of the code. This suggests
> that making the intent of the code more clear would be a good idea. I
> think that the ice_is_xdp_ena_vsi() would definitely make it more clear
> as opposed to the bare ICE_VCSI_PF.
I think that the current patch fits well for stable, and the issue
looks relevant enough that we should prefer have it fixed in this
cycle. Any refactoring/change would not allow such result due to the
timing.
I'll apply the series as-is, please follow-up on net-next as needed (no
rush).
Cheers,
Paolo
Powered by blists - more mailing lists