On Mon, May 06, 2019 at 01:45:40PM +0000, Maxim Mikityanskiy wrote:

On 2019-05-04 20:25, Björn Töpel wrote:

quoted

On Tue, 30 Apr 2019 at 20:12, Maxim Mikityanskiy [off-list ref] wrote:

quoted

Make it possible for the application to determine whether the AF_XDP
socket is running in zero-copy mode. To achieve this, add a new
getsockopt option XDP_OPTIONS that returns flags. The only flag
supported for now is the zero-copy mode indicator.

Signed-off-by: Maxim Mikityanskiy <redacted>
Reviewed-by: Tariq Toukan <redacted>
Acked-by: Saeed Mahameed <redacted>
---
  include/uapi/linux/if_xdp.h       |  7 +++++++
  net/xdp/xsk.c                     | 22 ++++++++++++++++++++++
  tools/include/uapi/linux/if_xdp.h |  7 +++++++
  3 files changed, 36 insertions(+)

diff --git a/include/uapi/linux/if_xdp.h b/include/uapi/linux/if_xdp.h
index caed8b1614ff..9ae4b4e08b68 100644
--- a/include/uapi/linux/if_xdp.h
+++ b/include/uapi/linux/if_xdp.h

@@ -46,6 +46,7 @@ struct xdp_mmap_offsets {
  #define XDP_UMEM_FILL_RING             5
  #define XDP_UMEM_COMPLETION_RING       6
  #define XDP_STATISTICS                 7
+#define XDP_OPTIONS                    8

  struct xdp_umem_reg {
         __u64 addr; /* Start of packet data area */

@@ -60,6 +61,12 @@ struct xdp_statistics {
         __u64 tx_invalid_descs; /* Dropped due to invalid descriptor */
  };

+struct xdp_options {
+       __u32 flags;
+};
+
+#define XDP_OPTIONS_FLAG_ZEROCOPY (1 << 0)

Nit: The other flags doesn't use "FLAG" in its name, but that doesn't
really matter.

quoted

+
  /* Pgoff for mmaping the rings */
  #define XDP_PGOFF_RX_RING                        0
  #define XDP_PGOFF_TX_RING               0x80000000

diff --git a/net/xdp/xsk.c b/net/xdp/xsk.c
index b68a380f50b3..998199109d5c 100644
--- a/net/xdp/xsk.c
+++ b/net/xdp/xsk.c

@@ -650,6 +650,28 @@ static int xsk_getsockopt(struct socket *sock, int level, int optname,

                 return 0;
         }
+       case XDP_OPTIONS:
+       {
+               struct xdp_options opts;
+
+               if (len < sizeof(opts))
+                       return -EINVAL;
+
+               opts.flags = 0;

Maybe get rid of this, in favor of "opts = {}" if the structure grows?

I'm OK with any of these options. Should I respin the series, or can I 
follow up with the change in RCs if the series gets to 5.2?

Alexei, is it even possible to still make changes to this series? The 
window appears closed.

The series were not applied.
Please resubmit addressing all feedback when bpf-next reopens.
Likely in ~2 weeks.