xprt_rdma_send_request(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’– xprt_rdma_send_request(9) πŸ’–

okay so bestie, this little icon is straight outta the Kernel API Manual, and we're about to make her make sense. group-chat energy: on. 🫢✨


πŸŽ€ NAME πŸŽ€

xprt_rdma_send_request β€” marshal and send an RPC request. That's the whole name-section serve, and she ate. πŸ’…


🌸 SYNOPSIS 🌸

int xprt_rdma_send_request (struct rpc_rqst *rqst );

πŸ¦‹ ARGUMENTS πŸ¦‹

  • rqst β€” the RPC message hangin' out in rq_snd_buf, ready for its moment. πŸ’–

πŸ”₯ DESCRIPTION πŸ”₯

Real talk, queen: the caller holds the transport's write lock. That's the vibe check. πŸ”’βœ¨


πŸ’… RETURN πŸ’…

Here's the group-chat rundown of return values, and every single one matters:

  • 0 β€” the RPC message has been sent. slay πŸ’…
  • -ENOTCONN β€” the caller should reconnect and call again.
  • -EAGAIN β€” the caller should call again.
  • -ENOBUFS β€” the caller should call again after a delay.
  • -EMSGSIZE β€” encoding ran out of buffer space. The request was not sent. Do not try to send this message again.
  • -EIO β€” an I/O error occurred. The request was not sent. Do not try to send this message again.
😭 babe, those last two are a hard no-retry zone. Not me stressing over it. Do not try to send this message again, fr fr.

πŸ‘‘ SEE ALSO πŸ‘‘


source/version: ./net/sunrpc/xprtrdma Β· date: October 2026 β€” okay that's the whole lore, byeee bestie, mwah πŸ’‹πŸ«ΆπŸŽ€

πŸ“¦ where this page comes from