1. 17 3月, 2022 11 次提交
  2. 16 3月, 2022 2 次提交
    • D
      io_uring: make tracing format consistent · 052ebf1f
      Dylan Yudaken 提交于
      Make the tracing formatting for user_data and flags consistent.
      
      Having consistent formatting allows one for example to grep for a specific
      user_data/flags and be able to trace a single sqe through easily.
      
      Change user_data to 0x%llx and flags to 0x%x everywhere. The '0x' is
      useful to disambiguate for example "user_data 100".
      
      Additionally remove the '=' for flags in io_uring_req_failed, again for consistency.
      Signed-off-by: NDylan Yudaken <dylany@fb.com>
      Link: https://lore.kernel.org/r/20220316095204.2191498-1-dylany@fb.comSigned-off-by: NJens Axboe <axboe@kernel.dk>
      052ebf1f
    • J
      io_uring: recycle apoll_poll entries · 4d9237e3
      Jens Axboe 提交于
      Particularly for networked workloads, io_uring intensively uses its
      poll based backend to get a notification when data/space is available.
      Profiling workloads, we see 3-4% of alloc+free that is directly attributed
      to just the apoll allocation and free (and the rest being skb alloc+free).
      
      For the fast path, we have ctx->uring_lock held already for both issue
      and the inline completions, and we can utilize that to avoid any extra
      locking needed to have a basic recycling cache for the apoll entries on
      both the alloc and free side.
      
      Double poll still requires an allocation. But those are rare and not
      a fast path item.
      
      With the simple cache in place, we see a 3-4% reduction in overhead for
      the workload.
      Signed-off-by: NJens Axboe <axboe@kernel.dk>
      4d9237e3
  3. 12 3月, 2022 1 次提交
  4. 11 3月, 2022 7 次提交
  5. 10 3月, 2022 18 次提交
  6. 07 3月, 2022 1 次提交
新手
引导
客服 返回
顶部