trace2: write discard message to sentinel files

Add a new "discard" event type for trace2 event destinations. When the
trace2 file count check creates a sentinel file, it will include the
normal trace2 output in the sentinel, along with this new discard
event.

Writing this message into the sentinel file is useful for tracking how
often the file count check triggers in practice.

Bump up the event format version since we've added a new event type.

Signed-off-by: Josh Steadmon <steadmon@google.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

5 files changed, 53 insertions, 30 deletions

diff --git a/trace2/tr2_dst.c b/trace2/tr2_dst.c
index a88c466100..ae052a07fe 100644
--- a/trace2/tr2_dst.c
+++ b/trace2/tr2_dst.c
@@ -48,15 +48,19 @@ void tr2_dst_trace_disable(struct tr2_dst *dst)
 /*
  * Check to make sure we're not overloading the target directory with too many
  * files. First get the threshold (if present) from the config or envvar. If
- * it's zero or unset, disable this check.  Next check for the presence of a
- * sentinel file, then check file count. If we are overloaded, create the
- * sentinel file if it doesn't already exist.
+ * it's zero or unset, disable this check. Next check for the presence of a
+ * sentinel file, then check file count.
+ *
+ * Returns 0 if tracing should proceed as normal. Returns 1 if the sentinel file
+ * already exists, which means tracing should be disabled. Returns -1 if there
+ * are too many files but there was no sentinel file, which means we have
+ * created and should write traces to the sentinel file.
  *
  * We expect that some trace processing system is gradually collecting files
  * from the target directory; after it removes the sentinel file we'll start
  * writing traces again.
  */
-static int tr2_dst_too_many_files(const char *tgt_prefix)
+static int tr2_dst_too_many_files(struct tr2_dst *dst, const char *tgt_prefix)
 {
 	int file_count = 0, max_files = 0, ret = 0;
 	const char *max_files_var;
@@ -95,8 +99,9 @@ static int tr2_dst_too_many_files(const char *tgt_prefix)
 		closedir(dirp);
 
 	if (file_count >= tr2env_max_files) {
-		creat(sentinel_path.buf, 0666);
-		ret = 1;
+		dst->too_many_files = 1;
+		dst->fd = open(sentinel_path.buf, O_WRONLY | O_CREAT | O_EXCL, 0666);
+		ret = -1;
 		goto cleanup;
 	}
 
@@ -108,7 +113,7 @@ cleanup:
 
 static int tr2_dst_try_auto_path(struct tr2_dst *dst, const char *tgt_prefix)
 {
-	int fd;
+	int too_many_files;
 	const char *last_slash, *sid = tr2_sid_get();
 	struct strbuf path = STRBUF_INIT;
 	size_t base_path_len;
@@ -124,7 +129,19 @@ static int tr2_dst_try_auto_path(struct tr2_dst *dst, const char *tgt_prefix)
 	strbuf_addstr(&path, sid);
 	base_path_len = path.len;
 
-	if (tr2_dst_too_many_files(tgt_prefix)) {
+	too_many_files = tr2_dst_too_many_files(dst, tgt_prefix);
+	if (!too_many_files) {
+		for (attempt_count = 0; attempt_count < MAX_AUTO_ATTEMPTS; attempt_count++) {
+			if (attempt_count > 0) {
+				strbuf_setlen(&path, base_path_len);
+				strbuf_addf(&path, ".%d", attempt_count);
+			}
+
+			dst->fd = open(path.buf, O_WRONLY | O_CREAT | O_EXCL, 0666);
+			if (dst->fd != -1)
+				break;
+		}
+	} else if (too_many_files == 1) {
 		strbuf_release(&path);
 		if (tr2_dst_want_warning())
 			warning("trace2: not opening %s trace file due to too "
@@ -134,18 +151,7 @@ static int tr2_dst_try_auto_path(struct tr2_dst *dst, const char *tgt_prefix)
 		return 0;
 	}
 
-	for (attempt_count = 0; attempt_count < MAX_AUTO_ATTEMPTS; attempt_count++) {
-		if (attempt_count > 0) {
-			strbuf_setlen(&path, base_path_len);
-			strbuf_addf(&path, ".%d", attempt_count);
-		}
-
-		fd = open(path.buf, O_WRONLY | O_CREAT | O_EXCL, 0666);
-		if (fd != -1)
-			break;
-	}
-
-	if (fd == -1) {
+	if (dst->fd == -1) {
 		if (tr2_dst_want_warning())
 			warning("trace2: could not open '%.*s' for '%s' tracing: %s",
 				(int) base_path_len, path.buf,
@@ -159,7 +165,6 @@ static int tr2_dst_try_auto_path(struct tr2_dst *dst, const char *tgt_prefix)
 
 	strbuf_release(&path);
 
-	dst->fd = fd;
 	dst->need_close = 1;
 	dst->initialized = 1;
 
diff --git a/trace2/tr2_dst.h b/trace2/tr2_dst.h
index 3adf3bac13..b1a8c144e0 100644
--- a/trace2/tr2_dst.h
+++ b/trace2/tr2_dst.h
@@ -9,6 +9,7 @@ struct tr2_dst {
 	int fd;
 	unsigned int initialized : 1;
 	unsigned int need_close : 1;
+	unsigned int too_many_files : 1;
 };
 
 /*
diff --git a/trace2/tr2_tgt_event.c b/trace2/tr2_tgt_event.c
index 9bcac20d1b..6353e8ad91 100644
--- a/trace2/tr2_tgt_event.c
+++ b/trace2/tr2_tgt_event.c
@@ -10,16 +10,17 @@
 #include "trace2/tr2_tgt.h"
 #include "trace2/tr2_tls.h"
 
-static struct tr2_dst tr2dst_event = { TR2_SYSENV_EVENT, 0, 0, 0 };
+static struct tr2_dst tr2dst_event = { TR2_SYSENV_EVENT, 0, 0, 0, 0 };
 
 /*
- * The version number of the JSON data generated by the EVENT target
- * in this source file.  Update this if you make a significant change
- * to the JSON fields or message structure.  You probably do not need
- * to update this if you just add another call to one of the existing
- * TRACE2 API methods.
+ * The version number of the JSON data generated by the EVENT target in this
+ * source file. The version should be incremented if new event types are added,
+ * if existing fields are removed, or if there are significant changes in
+ * interpretation of existing events or fields. Smaller changes, such as adding
+ * a new field to an existing event, do not require an increment to the EVENT
+ * format version.
  */
-#define TR2_EVENT_VERSION "1"
+#define TR2_EVENT_VERSION "2"
 
 /*
  * Region nesting limit for messages written to the event target.
@@ -107,6 +108,19 @@ static void event_fmt_prepare(const char *event_name, const char *file,
 		jw_object_intmax(jw, "repo", repo->trace2_repo_id);
 }
 
+static void fn_too_many_files_fl(const char *file, int line)
+{
+	const char *event_name = "too_many_files";
+	struct json_writer jw = JSON_WRITER_INIT;
+
+	jw_object_begin(&jw, 0);
+	event_fmt_prepare(event_name, file, line, NULL, &jw);
+	jw_end(&jw);
+
+	tr2_dst_write_line(&tr2dst_event, &jw.json);
+	jw_release(&jw);
+}
+
 static void fn_version_fl(const char *file, int line)
 {
 	const char *event_name = "version";
@@ -120,6 +134,9 @@ static void fn_version_fl(const char *file, int line)
 
 	tr2_dst_write_line(&tr2dst_event, &jw.json);
 	jw_release(&jw);
+
+	if (tr2dst_event.too_many_files)
+		fn_too_many_files_fl(file, line);
 }
 
 static void fn_start_fl(const char *file, int line,
diff --git a/trace2/tr2_tgt_normal.c b/trace2/tr2_tgt_normal.c
index 438ed05a40..31b602c171 100644
--- a/trace2/tr2_tgt_normal.c
+++ b/trace2/tr2_tgt_normal.c
@@ -9,7 +9,7 @@
 #include "trace2/tr2_tgt.h"
 #include "trace2/tr2_tls.h"
 
-static struct tr2_dst tr2dst_normal = { TR2_SYSENV_NORMAL, 0, 0, 0 };
+static struct tr2_dst tr2dst_normal = { TR2_SYSENV_NORMAL, 0, 0, 0, 0 };
 
 /*
  * Use the TR2_SYSENV_NORMAL_BRIEF setting to omit the "<time> <file>:<line>"
diff --git a/trace2/tr2_tgt_perf.c b/trace2/tr2_tgt_perf.c
index fd979db4ad..ffac8029ad 100644
--- a/trace2/tr2_tgt_perf.c
+++ b/trace2/tr2_tgt_perf.c
@@ -11,7 +11,7 @@
 #include "trace2/tr2_tgt.h"
 #include "trace2/tr2_tls.h"
 
-static struct tr2_dst tr2dst_perf = { TR2_SYSENV_PERF, 0, 0, 0 };
+static struct tr2_dst tr2dst_perf = { TR2_SYSENV_PERF, 0, 0, 0, 0 };
 
 /*
  * Use TR2_SYSENV_PERF_BRIEF to omit the "<time> <file>:<line>"