rpcrdma_send...unget_locked(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

πŸ’… rpcrdma_sendctx_unget_locked(9) β€” like, the kernel API manual, bestie ✨

okay so you found your way into the Kernel API Manual and the page you want is rpcrdma_send...unget_locked(9) β€” yes the header truncates it with dots, that's just how it vibes. anyway. sit down, hydrate, let's go 🫢

🌸 NAME 🌸

rpcrdma_sendctx_unget_locked β€” Release an unposted send context. that's it. that's the whole job. give it back, no sending happened πŸ’…

✨ SYNOPSIS ✨

void rpcrdma_sendctx_unget_locked (struct rpcrdma_xprt *r_xprt , struct
       rpcrdma_sendctx *sc );

it returns void β€” noθΏ”ε›žε€Ό, no receipts, it just does the thing and dips 😌

πŸŽ€ ARGUMENTS πŸŽ€

  • r_xprt β€” the controlling transport instance, aka the main character of this whole xprt situation πŸ‘‘
  • sc β€” the send context to release, aka the thing you're handing back. bye bestie πŸ¦‹

πŸ’– DESCRIPTION πŸ’–

Usage: this one gets called when no Send is posted for the sendctx that was most recently returned by rpcrdma_sendctx_get_locked. like, you grabbed it, you didn't use it, you're putting it back. no cap. that's the whole plot πŸ™ˆ

the caller serializes calls to this function and to rpcrdma_sendctx_get_locked (per transport). so like, one at a time, per transport, in an orderly little line β€” it's giving polite queue energy ✨

πŸ”₯ SEE ALSO πŸ”₯

kernel file ./net/sunrpc/xprtrdma/verbs.c β€” that's where the lore lives πŸ“–

like, the whole friend group fr fr β€” go read them too, iykyk πŸ’œ

lowkey a glow-up note: if you're calling this without having gotten the sendctx from rpcrdma_sendctx_get_locked first, that's not it. read the usage line again bestie 😭

and that's the page!! this little icon comes from ./net/sunrpc/xprtrdma, stamped October 2026, and it's still answering to the name rpcrdma_send...unget_locked(9) πŸŽ€ slay, go post some sends (or don't, apparently) πŸ’…βœ¨

πŸ“¦ where this page comes from