Struct TraceEventQueueList 

pub struct TraceEventQueueList {
    pub queues: Vec<Arc<TraceEventQueue>>,
    /* private fields */
}

Fields

queues: Vec<Arc<TraceEventQueue>>

Implementations

impl TraceEventQueueList

pub fn from(kernel: &Kernel) -> Arc<Self>

pub fn is_enabled(&self) -> bool

pub fn buffer_size_kb(&self) -> usize

pub fn set_buffer_size_kb(&self, size_kb: usize) -> Result<usize, Errno>

Replaces every queue’s ring buffer with an empty one of at least size_kb KiB, rounded up to whole pages and to LocklessRingBuffer’s 3-page minimum, and returns the size used. The new buffers keep the current tracing_on state.

Unlike Linux ring_buffer_resize(), which adds or removes pages in place, this discards every unread event, including events from writers still holding the old buffer. The old buffer’s dropped-page count is discarded too, so DROPPED_PAGES misses it.

pub fn enable(&self) -> Result<(), Errno>

Enables every queue. A queue that is already enabled is left alone, as Linux rb_simple_write() does when tracing_on is set to its current state. Unlike Linux, which keeps the buffer readable after tracing_on is cleared, disable() discards each queue’s contents, so enabling a disabled queue starts from empty.

pub fn disable(&self) -> Result<(), Errno>