---
name: "io_uring_prep_sendmsg"
section: "3"
platform: "linux"
package: "liburing"
version: "2.15-1"
release: "Arch Linux 2026-11-01"
platforms: ["linux"]
url: "https://new.carta.tech/io_uring_prep_sendmsg"
---

# io\_uring\_prep\_sendmsg(3)

prepare a sendmsg request

## NAME

io\_uring\_prep\_sendmsg - prepare a sendmsg request

## SYNOPSIS

```
#include <sys/types.h>
#include <sys/socket.h>
#include <liburing.h>
```

**void io\_uring\_prep\_sendmsg(struct io\_uring\_sqe \****sqe***,**                            **int** *fd***,**                            **const struct msghdr \****msg***,**                            **unsigned** *flags***);**

**void io\_uring\_prep\_sendmsg\_zc(struct io\_uring\_sqe \****sqe***,**                               **int** *fd***,**                               **const struct msghdr \****msg***,**                               **unsigned** *flags***);**

## DESCRIPTION

The [**io\_uring\_prep\_sendmsg**(3)](/io_uring_prep_sendmsg/3.md) function prepares a sendmsg request. The submission queue entry *sqe* is setup to use the file descriptor *fd* to start sending the data indicated by *msg* with the [**sendmsg**(2)](/sendmsg/2.md) defined flags in the *flags* argument.

The [**io\_uring\_prep\_sendmsg\_zc**(3)](/io_uring_prep_sendmsg_zc/3.md) accepts the same parameters as  [**io\_uring\_prep\_sendmsg**(3)](/io_uring_prep_sendmsg/3.md) but prepares a zerocopy sendmsg request.

See [**io\_uring\_prep\_send**(3)](/io_uring_prep_send/3.md) for a description of flags that can be set in the SQE *ioprio* field. In addition to those, the zero-copy send also supports setting **IORING\_SEND\_ZC\_REPORT\_USAGE .** If set, the notification CQE *res* field will report the number of bytes that were copied rather than sent with zero copy. A value of **0** indicates success. If the value is **IORING\_NOTIF\_USAGE\_ZC\_COPIED ,** then data was copied.

As opposed to non-zerocopy send requests, a zerocopy send will usually generate two CQEs. The first CQE holds the result of the send operation itself, and if that CQE has **IORING\_CQE\_F\_MORE** set in the CQE *flags* field, then a second notification CQE will be posted for the operation. This second notification tells the application that the memory associated with the send is safe to get reused. The second CQE will have **IORING\_CQE\_F\_NOTIF** set in the CQE *flags* field. Also see the [**io\_uring\_enter**(2)](/io_uring_enter/2.md) man page for a fuller description of the notification CQE.

Note that using **IOSQE\_IO\_LINK** with this request type requires the setting of **MSG\_WAITALL** in the *flags* argument, as a short send isn't considered an error condition without that being set.

This function prepares an async [**sendmsg**(2)](/sendmsg/2.md) request. See that man page for details.

## RETURN VALUE

None

## ERRORS

The CQE *res* field will contain the result of the operation. See the related man page for details on possible values. Note that where synchronous system calls will return **-1** on failure and set *errno* to the actual error value, io\_uring never uses *errno*. Instead it returns the negated *errno* directly in the CQE *res* field. Some common error cases are:

- **-ENOMEM**

  The [**ulimit**(1)](/ulimit/1.md) -l setting is too low to support the size of the attempted zero copy send. Increasing the limit may help

- **-ENOMEM**

  The kernel ran out of memory.

## NOTES

As with any request that passes in data in a struct, that data must remain valid until the request has been successfully submitted. It need not remain valid until completion. Once a request has been submitted, the in-kernel state is stable. Very early kernels (5.4 and earlier) required state to be stable until the completion occurred. Applications can test for this behavior by inspecting the **IORING\_FEAT\_SUBMIT\_STABLE** flag passed back from [**io\_uring\_queue\_init\_params**(3)](/io_uring_queue_init_params/3.md).

Despite accepting an array of iovec's with a size\_t number of bytes each, these functions can transfer at most INT\_MAX bytes per call (the maximum for the underlying syscall interface).

## SEE ALSO

[**io\_uring\_get\_sqe**(3)](/io_uring_get_sqe/3.md), [**io\_uring\_submit**(3)](/io_uring_submit/3.md), [**io\_uring\_buf\_ring\_init**(3)](/io_uring_buf_ring_init/3.md), [**io\_uring\_buf\_ring\_add**(3)](/io_uring_buf_ring_add/3.md), [**sendmsg**(2)](/sendmsg/2.md)