Re: [PATCH v4 bpf-next 09/13] bpf: introduce multibuff support to bpf_prog_test_run_xdp()

[Jesper Dangaard Brouer <brouer@redhat.com>]

On Thu, 8 Oct 2020 11:06:14 +0300
Shay Agroskin <shayagr@amazon.com> wrote:

> Lorenzo Bianconi <lorenzo@kernel.org> writes:
> 
> > Introduce the capability to allocate a xdp multi-buff in
> > bpf_prog_test_run_xdp routine. This is a preliminary patch to 
> > introduce
> > the selftests for new xdp multi-buff ebpf helpers
> >
> > Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
> > ---
> >  net/bpf/test_run.c | 51  ++++++++++++++++++++++++++++++++++++++--------
> >  1 file changed, 43 insertions(+), 8 deletions(-)
> >
> > diff --git a/net/bpf/test_run.c b/net/bpf/test_run.c
> > index bd291f5f539c..ec7286cd051b 100644
> > --- a/net/bpf/test_run.c
> > +++ b/net/bpf/test_run.c
> > @@ -617,44 +617,79 @@ int bpf_prog_test_run_xdp(struct bpf_prog 
> > *prog, const union bpf_attr *kattr,
> >  {
> >  	u32 tailroom = SKB_DATA_ALIGN(sizeof(struct 
> >  skb_shared_info));
> >  	u32 headroom = XDP_PACKET_HEADROOM;
> > -	u32 size = kattr->test.data_size_in;
> >  	u32 repeat = kattr->test.repeat;
> >  	struct netdev_rx_queue *rxqueue;
> > +	struct skb_shared_info *sinfo;
> >  	struct xdp_buff xdp = {};
> > +	u32 max_data_sz, size;
> >  	u32 retval, duration;
> > -	u32 max_data_sz;
> > +	int i, ret, data_len;
> >  	void *data;
> > -	int ret;
> >  
> >  	if (kattr->test.ctx_in || kattr->test.ctx_out)
> >  		return -EINVAL;
> >  
> > -	/* XDP have extra tailroom as (most) drivers use full page 
> > */
> >  	max_data_sz = 4096 - headroom - tailroom;  
> 
> For the sake of consistency, can this 4096 be changed to PAGE_SIZE 
> ?

The size 4096 is explicitly use, because the selftest xdp_adjust_tail
expect this, else it will fail on ARCHs with 64K PAGE_SIZE.  It also
seems excessive to create 64K packets for testing XDP.

See: tools/testing/selftests/bpf/prog_tests/xdp_adjust_tail.c

> Same as in
>      data_len = min_t(int, kattr->test.data_size_in - size, 
>      PAGE_SIZE);
> 
> expression below



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