Re: [PATCH bpf 0/3] xsk: require XDP_UMEM_TX_METADATA_LEN to actuate tx_metadata_len

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On 7/19/24 5:29 PM, Fijalkowski, Maciej wrote:
On 7/13/24 3:52 AM, Stanislav Fomichev wrote:
Julian reports that commit 341ac980eab9 ("xsk: Support tx_metadata_len")
can break existing use cases which don't zero-initialize xdp_umem_reg
padding. Fix it (while still breaking a minority of new users of tx
metadata), update the docs, update the selftest and sprinkle some
BUILD_BUG_ONs to hopefully catch similar issues in the future.

Thank you Julian for the report and for helping to chase it down!

Reported-by: Julian Schindel <mail@xxxxxxxxxxxxxxxx>
Cc: Magnus Karlsson <magnus.karlsson@xxxxxxxxx>

Stanislav Fomichev (3):
    xsk: require XDP_UMEM_TX_METADATA_LEN to actuate tx_metadata_len
    selftests/bpf: Add XDP_UMEM_TX_METADATA_LEN to XSK TX metadata test
    xsk: Try to make xdp_umem_reg extension a bit more future-proof

   Documentation/networking/xsk-tx-metadata.rst  | 16 ++++++++-----
   include/uapi/linux/if_xdp.h                   |  4 ++++
   net/xdp/xdp_umem.c                            |  9 +++++---
   net/xdp/xsk.c                                 | 23 ++++++++++---------
   tools/include/uapi/linux/if_xdp.h             |  4 ++++
   .../selftests/bpf/prog_tests/xdp_metadata.c   |  3 ++-
   6 files changed, 38 insertions(+), 21 deletions(-)


Magnus or Maciej, ptal when you get a chance.

I'll do so on Monday as I'll be back from vacation, Magnus will be out for
yet another week. Hope it works for you?

Sounds good, just making sure this doesn't fall off the radar. :)

Thanks Maciej!




[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux