Re: [PATCH v2 03/18] mm/ksw: add /proc/kstackwatch interface

On Thu, Sep 04, 2025 at 10:51:41AM +0300, Mike Rapoport wrote:

On Thu, Sep 04, 2025 at 08:21:00AM +0800, Jinchao Wang wrote:

quoted

Provide the /proc/kstackwatch file to read or update the KSW configuration.
Writing a valid config string starts watching; empty input stops watching.
Invalid input stops watching and reports an error.

Allocate a ksw_config struct on module init and free it on exit.
Manage watching state with ksw_start_watching() and ksw_stop_watching().

Signed-off-by: Jinchao Wang <redacted>
---
 mm/kstackwatch/kernel.c | 140 ++++++++++++++++++++++++++++++++++++++++
 1 file changed, 140 insertions(+)

diff --git a/mm/kstackwatch/kernel.c b/mm/kstackwatch/kernel.c
index 4a6dc49449fe..95ade95abde1 100644
--- a/mm/kstackwatch/kernel.c
+++ b/mm/kstackwatch/kernel.c

@@ -1,7 +1,10 @@
 // SPDX-License-Identifier: GPL-2.0
 #include <linux/kstrtox.h>
 #include <linux/module.h>
+#include <linux/proc_fs.h>
+#include <linux/seq_file.h>
 #include <linux/string.h>
+#include <linux/uaccess.h>
 
 #include "kstackwatch.h"
 

@@ -9,6 +12,29 @@ MODULE_AUTHOR("Jinchao Wang");
 MODULE_DESCRIPTION("Kernel Stack Watch");
 MODULE_LICENSE("GPL");
 
+struct ksw_config *ksw_config;

Does a global ksw_config imply that this is a singleton?

Yes, it is.

What happens when there are several "echo .. > /proc/kstackwatch"?
And even more interesting what happens if they are parallel?

I did not consider this deeply enough.
I will address it properly in the next series.
Thank you for pointing it out.

quoted

+bool watching_active;
+
+bool panic_on_catch;
+module_param(panic_on_catch, bool, 0644);
+MODULE_PARM_DESC(panic_on_catch,
+		 "Trigger a kernel panic immediately on corruption catch");
+
+static int ksw_start_watching(void)
+{
+	watching_active = true;
+
+	pr_info("KSW: start watching %s\n", ksw_config->config_str);
+	return 0;
+}
+
+static void ksw_stop_watching(void)
+{
+	watching_active = false;
+
+	pr_info("KSW: stop watching %s\n", ksw_config->config_str);

This module is overly verbose. Do you really need all the printks?

You’re right — many of these printks are for early debugging and not
needed for production. I will remove or downgrade them to pr_debug() in
the next version.

For critical paths, I’ll ensure only actionable or WARN-level messages
remain.

Appreciate the push for cleaner logging.

quoted

+}

...

quoted

 static int __init kstackwatch_init(void)
 {
+	ksw_config = kmalloc(sizeof(*ksw_config), GFP_KERNEL);
+	if (!ksw_config)
+		return -ENOMEM;
+
+	/* Create proc interface */
+	if (!proc_create("kstackwatch", 0644, NULL, &kstackwatch_proc_ops)) {

ksw_config must be freed here

quoted

+		pr_err("KSW: create proc kstackwatch fail");
+		return -ENOMEM;
+	}
+
 	pr_info("KSW: module loaded\n");
 	pr_info("KSW: usage:\n");
 	pr_info("KSW: echo 'function+ip_offset[+depth] [local_var_offset:local_var_len]' > /proc/kstackwatch\n");

@@ -85,6 +217,14 @@ static int __init kstackwatch_init(void)
 
 static void __exit kstackwatch_exit(void)
 {
+	/* Cleanup active watching */
+	if (watching_active)
+		ksw_stop_watching();
+
+	/* Remove proc interface */
+	remove_proc_entry("kstackwatch", NULL);
+	kfree(ksw_config);
+
 	pr_info("KSW: Module unloaded\n");
 }
 

-- 
2.43.0

-- 
Sincerely yours,
Mike.