cxl_internal_send_cmd(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

okay so bestie, welcome to cxl_internal_send_cmd(9) πŸ’– this page lives in the Kernel API Manual and she is literally the backstage pass for mailbox commands β€” no cap, she's the kernel-internal way to slide a command into the mailbox πŸŽ€


πŸŽ€ NAME πŸŽ€

cxl_internal_send_cmd β€” kernel internal interface to send a mailbox command ✨

🌸 SYNOPSIS 🌸

int cxl_internal_send_cmd (struct cxl_mailbox *cxl_mbox , struct cxl_mbox_cmd *mbox_cmd );

πŸ¦‹ ARGUMENTS πŸ¦‹

  • cxl_mbox β€” the CXL mailbox context, your main character 🫢
  • mbox_cmd β€” the initialized command you wanna execute, already dressed and ready πŸ’…

✨ CONTEXT ✨

Any context. like, genuinely any. she's flexible like that 🌈

πŸ”₯ RETURN πŸ”₯

  • %>=0 β€” number of bytes returned in out πŸ’–
  • -E2BIG β€” payload is too large for the hardware 😭
  • -EBUSY β€” couldn't acquire exclusive mailbox access, someone else is in the group chat πŸ™ˆ
  • -EFAULT β€” a hardware error occurred πŸ’€
  • -ENXIO β€” command completed, but the device reported an error 😬
  • -EIO β€” unexpected output size 🫠

πŸ‘‘ DESCRIPTION πŸ‘‘

okay this is the part that matters, so listen up πŸ“

Mailbox commands may execute successfully yet the device itself reported an error. two different things!! and while this distinction can be useful for commands from userspace, the kernel will only be able to use results when both are successful. both. not one. both. periodt 🧚

🌈 SEE ALSO 🌈


and that's the whole vibe check πŸ¦‹ she lives over in ./drivers/cxl/core, dated October 2026, and she answers to cxl_internal_send_cmd(9) β€” go read her friends, iykyk πŸ’œ

πŸ“¦ where this page comes from