Re: [PATCH v8 40/43] dyndbg: add epilogue to dynamic_debug/control file

Thank you for your contribution! Sashiko AI review found 1 potential issue(s) to consider:
- [Medium] ddebug_proc_start() fails to return EPILOGUE_TOKEN when resuming sequence iteration at the epilogue index, which violates seq_file semantics and causes the epilogue to be silently omitted if a read boundary falls exactly before it.
--

commit c254572f4a41eba87d7288c5cb6c6ef71571d0a3
Author: Jim Cromie [off-list ref]

dyndbg: add epilogue to dynamic_debug/control file

This commit adds an epilogue to the control-file to allow display of
statistics without disturbing the header. It introduces an EPILOGUE_TOKEN
yielded at the end of seq_file iteration.

[ ... ]

 static void *ddebug_proc_next(struct seq_file *m, void *p, loff_t *pos)
 {
 	struct ddebug_iter *iter = m->private;
 	struct _ddebug *dp;
 
+	(*pos)++;
+
+	if (p == EPILOGUE_TOKEN)
+		return NULL;
+
 	if (p == SEQ_START_TOKEN)
 		dp = ddebug_iter_first(iter);
 	else
 		dp = ddebug_iter_next(iter);
-	++*pos;
-	return dp;
+
+	if (dp)
+		return dp;
+
+	return EPILOGUE_TOKEN;

[Severity: Medium]
Since ddebug_proc_next() now yields EPILOGUE_TOKEN as an additional sequence
element, does ddebug_proc_start() also need to be updated to handle resuming
iteration at the epilogue index?

If a read boundary falls exactly before the epilogue (where pos equals the
total number of items plus one), ddebug_proc_start() will iterate until dp
is NULL, and then return NULL instead of EPILOGUE_TOKEN:

lib/dynamic_debug.c:ddebug_proc_start() {
	...
	dp = ddebug_iter_first(iter);
	while (dp != NULL && --n > 0)
		dp = ddebug_iter_next(iter);
	return dp;
}

Will this cause the epilogue to be silently omitted in that scenario?

-- 
Sashiko AI review · https://sashiko.dev/#/patchset/20260905-dd-cmap-part2-clean-v8-0-a4cc0674f6fd@gmail.com?part=40