[PATCH v2 0/2] vsprintf: Add test for restricted kernel pointers

[PATCH v2 0/2] vsprintf: Add test for restricted kernel pointers

[Thomas Weißschuh]

Fill out the tests for restricted kernel pointers, using the %pK format.

---
Changes in v2:
- Restrict the visibility of no_hash_pointers.
- Handle printf_kunit being a module.
- Add note about KUnit capabilities for kptr_restrict==1.
- Link to v1: https://lore.kernel.org/r/20260114-restricted-pointers-kunit-test-v1-1-a9d8a49c6b6f@linutronix.de

To: Andrew Morton <akpm@linux-foundation.org>
To: Petr Mladek <pmladek@suse.com>
To: Steven Rostedt <rostedt@goodmis.org>
To: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
To: Rasmus Villemoes <linux@rasmusvillemoes.dk>
To: Sergey Senozhatsky <senozhatsky@chromium.org>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>

---
Thomas Weißschuh (2):
      vsprintf: Only export no_hash_pointers to test module
      vsprintf: Add test for restricted kernel pointers

 lib/tests/printf_kunit.c | 22 +++++++++++++++++++++-
 lib/vsprintf.c           |  7 ++++++-
 2 files changed, 27 insertions(+), 2 deletions(-)
---
base-commit: 254f49634ee16a731174d2ae34bc50bd5f45e731
change-id: 20260109-restricted-pointers-kunit-test-ff4ec887bcd8

Best regards,
--  
Thomas Weißschuh <thomas.weissschuh@linutronix.de>

[PATCH v2 1/2] vsprintf: Only export no_hash_pointers to test module

[Thomas Weißschuh]

Aside from the printf test module, no module should ever use this symbol.

Suggested-by: Petr Mladek <pmladek@suse.com>
Link: https://lore.kernel.org/all/aWpwMyFEfpCNN297@pathway.suse.cz/
Suggested-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Link: https://lore.kernel.org/all/aW3em-KplLVofU5z@smile.fi.intel.com/
Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>
---
 lib/vsprintf.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/lib/vsprintf.c b/lib/vsprintf.c
index 9f359b31c8d1..f70d240a394c 100644
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -59,7 +59,9 @@
 
 /* Disable pointer hashing if requested */
 bool no_hash_pointers __ro_after_init;
-EXPORT_SYMBOL_GPL(no_hash_pointers);
+#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
+EXPORT_SYMBOL_FOR_MODULES(no_hash_pointers, "printf_kunit");
+#endif
 
 /*
  * Hashed pointers policy selected by "hash_pointers=..." boot param

-- 
2.53.0

Re: [PATCH v2 1/2] vsprintf: Only export no_hash_pointers to test module

[Andy Shevchenko]

On Mon, May 04, 2026 at 12:43:40PM +0200, Thomas Weißschuh wrote:
> Aside from the printf test module, no module should ever use this symbol.

...

> -EXPORT_SYMBOL_GPL(no_hash_pointers);

> +#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
> +EXPORT_SYMBOL_FOR_MODULES(no_hash_pointers, "printf_kunit");
> +#endif

But do we need that ugly ifdeffery? the infrastructure should handle that.
(Also, if I build a module separately after the kernel, it won't work.)


-- 
With Best Regards,
Andy Shevchenko

Re: [PATCH v2 1/2] vsprintf: Only export no_hash_pointers to test module

[Thomas Weißschuh]

On Mon, May 04, 2026 at 01:57:31PM +0300, Andy Shevchenko wrote:
> On Mon, May 04, 2026 at 12:43:40PM +0200, Thomas Weißschuh wrote:
> > Aside from the printf test module, no module should ever use this symbol.
> 
> ...
> 
> > -EXPORT_SYMBOL_GPL(no_hash_pointers);
> 
> > +#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
> > +EXPORT_SYMBOL_FOR_MODULES(no_hash_pointers, "printf_kunit");
> > +#endif
> 
> But do we need that ugly ifdeffery? the infrastructure should handle that.

Where does the generic infrastructure take the kconfig option into account?

> (Also, if I build a module separately after the kernel, it won't work.)

Only if it was not enabled when the kernel was built.

But if people are unhappy with the ifdeffery, I'm fine with dropping it.
It is not a hard requirement at all.


Thomas

Re: [PATCH v2 1/2] vsprintf: Only export no_hash_pointers to test module

[Andy Shevchenko]

On Mon, May 04, 2026 at 01:53:47PM +0200, Thomas Weißschuh wrote:
> On Mon, May 04, 2026 at 01:57:31PM +0300, Andy Shevchenko wrote:
> > On Mon, May 04, 2026 at 12:43:40PM +0200, Thomas Weißschuh wrote:

...

> > > -EXPORT_SYMBOL_GPL(no_hash_pointers);
> > 
> > > +#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
> > > +EXPORT_SYMBOL_FOR_MODULES(no_hash_pointers, "printf_kunit");
> > > +#endif
> > 
> > But do we need that ugly ifdeffery? the infrastructure should handle that.
> 
> Where does the generic infrastructure take the kconfig option into account?

Nowhere, but my point that it's taken care of at run-time.

> > (Also, if I build a module separately after the kernel, it won't work.)
> 
> Only if it was not enabled when the kernel was built.
> 
> But if people are unhappy with the ifdeffery, I'm fine with dropping it.
> It is not a hard requirement at all.

I think it's unnecessary to have. But I'm not a maintainer of this code.

-- 
With Best Regards,
Andy Shevchenko

[PATCH v2 2/2] vsprintf: Add test for restricted kernel pointers

[Thomas Weißschuh]

Fill out the tests for restricted kernel pointers, using the %pK format.

Signed-off-by: Thomas Weißschuh <thomas.weissschuh@linutronix.de>

---
Please note that changes to the kptr_restrict sysctl from the kernel
commandline are only applied *after* the boot-time KUnit tests run.
---
 lib/tests/printf_kunit.c | 22 +++++++++++++++++++++-
 lib/vsprintf.c           |  3 +++
 2 files changed, 24 insertions(+), 1 deletion(-)

diff --git a/lib/tests/printf_kunit.c b/lib/tests/printf_kunit.c
index bb70b9cddadd..58e639b01e83 100644
--- a/lib/tests/printf_kunit.c
+++ b/lib/tests/printf_kunit.c
@@ -319,7 +319,27 @@ symbol_ptr(struct kunit *kunittest)
 static void
 kernel_ptr(struct kunit *kunittest)
 {
-	/* We can't test this without access to kptr_restrict. */
+	switch (kptr_restrict) {
+	case 0:
+		if (no_hash_pointers) {
+			test(PTR_STR, "%pK", PTR);
+		} else {
+			char buf[PLAIN_BUF_SIZE];
+
+			plain_hash_to_buffer(kunittest, PTR, buf, PLAIN_BUF_SIZE);
+			/* %pK behaves the same as hashing */
+			test(buf, "%pK", PTR);
+		}
+		break;
+	case 1:
+		/* The KUnit kthread has all capabilities, including CAP_SYSLOG */
+		test(PTR_STR, "%pK", PTR);
+		break;
+	case 2:
+	default:
+		test(ZEROS "00000000", "%pK", PTR);
+		break;
+	}
 }
 
 static void
diff --git a/lib/vsprintf.c b/lib/vsprintf.c
index f70d240a394c..dc7e012a6ed7 100644
--- a/lib/vsprintf.c
+++ b/lib/vsprintf.c
@@ -859,6 +859,9 @@ static char *default_pointer(char *buf, char *end, const void *ptr,
 }
 
 int kptr_restrict __read_mostly;
+#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
+EXPORT_SYMBOL_FOR_MODULES(kptr_restrict, "printf_kunit");
+#endif
 
 static noinline_for_stack
 char *restricted_pointer(char *buf, char *end, const void *ptr,

-- 
2.53.0

Re: [PATCH v2 2/2] vsprintf: Add test for restricted kernel pointers

[Andy Shevchenko]

On Mon, May 04, 2026 at 12:43:41PM +0200, Thomas Weißschuh wrote:
> Fill out the tests for restricted kernel pointers, using the %pK format.

...

> +#if IS_MODULE(CONFIG_PRINTF_KUNIT_TEST)
> +EXPORT_SYMBOL_FOR_MODULES(kptr_restrict, "printf_kunit");
> +#endif

Same Q here.

-- 
With Best Regards,
Andy Shevchenko