Skip to content

Libbpf userspace function ring_buffer__new

0.0.9

Create a new ring buffer manager.

Definition

typedef int (*ring_buffer_sample_fn)(void *ctx, void *data, size_t size); 

struct ring_buffer * ring_buffer__new(int map_fd, ring_buffer_sample_fn sample_cb, void *ctx, const struct ring_buffer_opts *opts);

Parameters

  • map_fd: file descriptor of the BPF_MAP_TYPE_RINGBUF map of the first ring buffer that will be part of the ring buffer manager
  • sample_cb: callback function that will be called when a sample is ready for the initial ring buffer
  • ctx: context that will be passed to the callback function
  • opts: options for the ring buffer

Returns

A pointer to the newly created ring buffer manager, or NULL on error. errno is set to the error code on error.

struct ring_buffer_opts

struct ring_buffer_opts {
    size_t sz; /* size of this struct, for forward/backward compatibility */
};

Usage

A ring buffer in this context is a circular buffer where eBPF programs are producers and userspace is the consumer. Even though the returned value is struct ring_buffer *, it actually is a "manager" for multiple ring buffers (struct ring).

All ring buffers that are part of this ring buffer manager can be epoll-ed together, to get notified of pending data on any of the rings.

A ring buffer manager always contains at least one ring, the map_fd, sample_cb, and ctx parameters are used to create the first ring buffer. It internally calls ring_buffer__add to add the first ring buffer.

Example

Docs could be improved

This part of the docs is incomplete, contributions are very welcome