Re: [PATCH v8 bpf-next 00/14] mvneta: introduce XDP multi-buffer support

From: Jesper Dangaard Brouer <brouer@redhat.com>
To: Magnus Karlsson <magnus.karlsson@gmail.com>
Cc: Lorenzo Bianconi <lorenzo@kernel.org>, bpf <bpf@vger.kernel.org>,
	Network Development <netdev@vger.kernel.org>,
	"David S. Miller" <davem@davemloft.net>,
	Jakub Kicinski <kuba@kernel.org>,
	Alexei Starovoitov <ast@kernel.org>,
	Daniel Borkmann <daniel@iogearbox.net>,
	John Fastabend <john.fastabend@gmail.com>,
	David Ahern <dsahern@kernel.org>,
	Eelco Chaudron <echaudro@redhat.com>,
	Jason Wang <jasowang@redhat.com>,
	Alexander Duyck <alexander.duyck@gmail.com>,
	Saeed Mahameed <saeed@kernel.org>,
	"Fijalkowski, Maciej" <maciej.fijalkowski@intel.com>,
	Tirthendu <tirthendu.sarkar@intel.com>,
	brouer@redhat.com
Subject: Re: [PATCH v8 bpf-next 00/14] mvneta: introduce XDP multi-buffer support
Date: Thu, 22 Apr 2021 16:42:23 +0200	[thread overview]
Message-ID: <20210422164223.77870d28@carbon> (raw)
In-Reply-To: <CAJ8uoz2JpfdjvjJp-vjWuhw5z1=2D32jj-KktFnLN6Zd9ZVmAQ@mail.gmail.com>

On Thu, 22 Apr 2021 12:24:32 +0200
Magnus Karlsson <magnus.karlsson@gmail.com> wrote:

> On Wed, Apr 21, 2021 at 5:39 PM Jesper Dangaard Brouer
> <brouer@redhat.com> wrote:
> >
> > On Wed, 21 Apr 2021 16:12:32 +0200
> > Magnus Karlsson <magnus.karlsson@gmail.com> wrote:
> >  
[...]
> > > more than I get.  
> >
> > I clearly have a bug in the i40e driver.  As I wrote later, I don't see
> > any packets transmitted for XDP_TX.  Hmm, I using Mel Gorman's tree,
> > which contains the i40e/ice/ixgbe bug we fixed earlier.

Something is wrong with i40e, I changed git-tree to net-next (at
commit 5d869070569a) and XDP seems to have stopped working on i40e :-(

$ uname -a
Linux broadwell 5.12.0-rc7-net-next+ #600 SMP PREEMPT Thu Apr 22 15:13:15 CEST 2021 x86_64 x86_64 x86_64 GNU/Linux

When I load any XDP prog almost no packets are let through:

 [kernel-bpf-samples]$ sudo ./xdp1 i40e2
 libbpf: elf: skipping unrecognized data section(16) .eh_frame
 libbpf: elf: skipping relo section(17) .rel.eh_frame for section(16) .eh_frame
 proto 17:          1 pkt/s
 proto 0:          0 pkt/s
 proto 17:          0 pkt/s
 proto 0:          0 pkt/s
 proto 17:          1 pkt/s

On the same system my mlx5 NIC works fine:

 [kernel-bpf-samples]$ sudo ./xdp1 mlx5p1
 libbpf: elf: skipping unrecognized data section(16) .eh_frame
 libbpf: elf: skipping relo section(17) .rel.eh_frame for section(16) .eh_frame
 proto 17:   10984608 pkt/s
 proto 17:   24374656 pkt/s
 proto 17:   24339904 pkt/s

> > The call to xdp_convert_buff_to_frame() fails, but (see below) that
> > error is simply converted to I40E_XDP_CONSUMED.  Thus, not even the
> > 'trace_xdp_exception' will be able to troubleshoot this.  You/Intel
> > should consider making XDP_TX errors detectable (this will also happen
> > if TX ring don't have room).  
> 
> This is not good. Will submit a fix. Thanks for reporting Jesper.

Usually I use this command to troubleshoot:
  sudo ./xdp_monitor --stats 

But as driver i40e doesn't call the 'trace_xdp_exception' then I don't
capture any error this way.

-- 
Best regards,
  Jesper Dangaard Brouer
  MSc.CS, Principal Kernel Engineer at Red Hat
  LinkedIn: http://www.linkedin.com/in/brouer