[RFC,2/3] tracing/hist: Support POLLPRI event for poll on histogram

Commit Message

Masami Hiramatsu (Google) June 25, 2024, 3:17 p.m. UTC

From: Masami Hiramatsu (Google) <[e-mail]>

Since POLLIN will not be flashed until read the hist file, user needs
to repeat read() and poll() on hist for monitoring the event
continuously. But the read() is somewhat redundant only for monitoring
events.

This add POLLPRI poll event on hist, this event returns when a histogram
is updated after open(), poll() or read(). Thus it is possible to wait
next event without read().

Signed-off-by: Masami Hiramatsu (Google) <[e-mail]>
---
 kernel/trace/trace_events_hist.c |   29 +++++++++++++++++++++++++++--
 1 file changed, 27 insertions(+), 2 deletions(-)

Patch

diff --git a/kernel/trace/trace_events_hist.c b/kernel/trace/trace_events_hist.c
index 136d91139949..a22454ad36d8 100644
--- a/kernel/trace/trace_events_hist.c
+++ b/kernel/trace/trace_events_hist.c
@@ -5599,6 +5599,7 @@  static void hist_trigger_show(struct seq_file *m,
 struct hist_file_data {
 	struct file *file;
 	u64 last_read;
+	u64 last_act;
 };
 
 static u64 get_hist_hit_count(struct trace_event_file *event_file)
@@ -5636,6 +5637,11 @@  static int hist_show(struct seq_file *m, void *v)
 			hist_trigger_show(m, data, n++);
 	}
 	hist_file->last_read = get_hist_hit_count(event_file);
+	/*
+	 * Update last_act too so that poll()/POLLPRI can wait for the next
+	 * event after any syscall on hist file.
+	 */
+	hist_file->last_act = hist_file->last_read;
 
  out_unlock:
 	mutex_unlock(&event_mutex);
@@ -5649,6 +5655,7 @@  static __poll_t event_hist_poll(struct file *file, struct poll_table_struct *wai
 	struct seq_file *m = file->private_data;
 	struct hist_file_data *hist_file = m->private;
 	__poll_t ret = 0;
+	u64 cnt;
 
 	mutex_lock(&event_mutex);
 
@@ -5660,8 +5667,13 @@  static __poll_t event_hist_poll(struct file *file, struct poll_table_struct *wai
 
 	poll_wait(file, &event_file->hist_wq, wait);
 
-	if (hist_file->last_read != get_hist_hit_count(event_file))
-		ret = EPOLLIN | EPOLLRDNORM;
+	cnt = get_hist_hit_count(event_file);
+	if (hist_file->last_read != cnt)
+		ret |= EPOLLIN | EPOLLRDNORM;
+	if (hist_file->last_act != cnt) {
+		hist_file->last_act = cnt;
+		ret |= EPOLLPRI;
+	}
 
 out_unlock:
 	mutex_unlock(&event_mutex);
@@ -5680,6 +5692,7 @@  static int event_hist_release(struct inode *inode, struct file *file)
 
 static int event_hist_open(struct inode *inode, struct file *file)
 {
+	struct trace_event_file *event_file;
 	struct hist_file_data *hist_file;
 	int ret;
 
@@ -5690,13 +5703,25 @@  static int event_hist_open(struct inode *inode, struct file *file)
 	hist_file = kzalloc(sizeof(*hist_file), GFP_KERNEL);
 	if (!hist_file)
 		return -ENOMEM;
+
+	mutex_lock(&event_mutex);
+	event_file = event_file_data(file);
+	if (!event_file) {
+		ret = -ENODEV;
+		goto out_unlock;
+	}
+
 	hist_file->file = file;
+	hist_file->last_act = get_hist_hit_count(event_file);
 
 	/* Clear private_data to avoid warning in single_open() */
 	file->private_data = NULL;
 	ret = single_open(file, hist_show, hist_file);
+
+out_unlock:
 	if (ret)
 		kfree(hist_file);
+	mutex_unlock(&event_mutex);
 	return ret;
 }