LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [PATCH] kasan: test: Bypass __alloc_size checks
@ 2021-10-06  3:55 Kees Cook
  2021-10-06 11:38 ` Mark Rutland
  0 siblings, 1 reply; 3+ messages in thread
From: Kees Cook @ 2021-10-06  3:55 UTC (permalink / raw)
  To: Andrew Morton
  Cc: Kees Cook, Andrey Ryabinin, Alexander Potapenko,
	Andrey Konovalov, Dmitry Vyukov, kasan-dev, linux-kernel,
	linux-hardening

Intentional overflows, as performed by the KASAN tests, are detected
at compile time[1] (instead of only at run-time) with the addition of
__alloc_size. Fix this by forcing the compiler into not being able to
trust the size used following the kmalloc()s.

[1] https://lore.kernel.org/lkml/20211005184717.65c6d8eb39350395e387b71f@linux-foundation.org

Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
Cc: Alexander Potapenko <glider@google.com>
Cc: Andrey Konovalov <andreyknvl@gmail.com>
Cc: Dmitry Vyukov <dvyukov@google.com>
Cc: kasan-dev@googlegroups.com
Signed-off-by: Kees Cook <keescook@chromium.org>
---
 lib/test_kasan.c        | 10 +++++-----
 lib/test_kasan_module.c |  2 +-
 2 files changed, 6 insertions(+), 6 deletions(-)

diff --git a/lib/test_kasan.c b/lib/test_kasan.c
index 8835e0784578..0e1f8d5281b4 100644
--- a/lib/test_kasan.c
+++ b/lib/test_kasan.c
@@ -435,7 +435,7 @@ static void kmalloc_uaf_16(struct kunit *test)
 static void kmalloc_oob_memset_2(struct kunit *test)
 {
 	char *ptr;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 
 	ptr = kmalloc(size, GFP_KERNEL);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
@@ -447,7 +447,7 @@ static void kmalloc_oob_memset_2(struct kunit *test)
 static void kmalloc_oob_memset_4(struct kunit *test)
 {
 	char *ptr;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 
 	ptr = kmalloc(size, GFP_KERNEL);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
@@ -459,7 +459,7 @@ static void kmalloc_oob_memset_4(struct kunit *test)
 static void kmalloc_oob_memset_8(struct kunit *test)
 {
 	char *ptr;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 
 	ptr = kmalloc(size, GFP_KERNEL);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
@@ -471,7 +471,7 @@ static void kmalloc_oob_memset_8(struct kunit *test)
 static void kmalloc_oob_memset_16(struct kunit *test)
 {
 	char *ptr;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 
 	ptr = kmalloc(size, GFP_KERNEL);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
@@ -483,7 +483,7 @@ static void kmalloc_oob_memset_16(struct kunit *test)
 static void kmalloc_oob_in_memset(struct kunit *test)
 {
 	char *ptr;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 
 	ptr = kmalloc(size, GFP_KERNEL);
 	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
diff --git a/lib/test_kasan_module.c b/lib/test_kasan_module.c
index 7ebf433edef3..c8cc77b1dcf3 100644
--- a/lib/test_kasan_module.c
+++ b/lib/test_kasan_module.c
@@ -19,7 +19,7 @@ static noinline void __init copy_user_test(void)
 {
 	char *kmem;
 	char __user *usermem;
-	size_t size = 128 - KASAN_GRANULE_SIZE;
+	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
 	int __maybe_unused unused;
 
 	kmem = kmalloc(size, GFP_KERNEL);
-- 
2.30.2


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] kasan: test: Bypass __alloc_size checks
  2021-10-06  3:55 [PATCH] kasan: test: Bypass __alloc_size checks Kees Cook
@ 2021-10-06 11:38 ` Mark Rutland
  2021-10-06 16:33   ` Kees Cook
  0 siblings, 1 reply; 3+ messages in thread
From: Mark Rutland @ 2021-10-06 11:38 UTC (permalink / raw)
  To: Kees Cook
  Cc: Andrew Morton, Andrey Ryabinin, Alexander Potapenko,
	Andrey Konovalov, Dmitry Vyukov, kasan-dev, linux-kernel,
	linux-hardening

Hi Kees,

On Tue, Oct 05, 2021 at 08:55:22PM -0700, Kees Cook wrote:
> Intentional overflows, as performed by the KASAN tests, are detected
> at compile time[1] (instead of only at run-time) with the addition of
> __alloc_size. Fix this by forcing the compiler into not being able to
> trust the size used following the kmalloc()s.

It might be better to use OPTIMIZER_HIDE_VAR(), since that's intended to
make the value opaque to the compiler, and volatile might not always do
that depending on how the compiler tracks the variable.

Thanks,
Mark.

> 
> [1] https://lore.kernel.org/lkml/20211005184717.65c6d8eb39350395e387b71f@linux-foundation.org
> 
> Cc: Andrey Ryabinin <ryabinin.a.a@gmail.com>
> Cc: Alexander Potapenko <glider@google.com>
> Cc: Andrey Konovalov <andreyknvl@gmail.com>
> Cc: Dmitry Vyukov <dvyukov@google.com>
> Cc: kasan-dev@googlegroups.com
> Signed-off-by: Kees Cook <keescook@chromium.org>
> ---
>  lib/test_kasan.c        | 10 +++++-----
>  lib/test_kasan_module.c |  2 +-
>  2 files changed, 6 insertions(+), 6 deletions(-)
> 
> diff --git a/lib/test_kasan.c b/lib/test_kasan.c
> index 8835e0784578..0e1f8d5281b4 100644
> --- a/lib/test_kasan.c
> +++ b/lib/test_kasan.c
> @@ -435,7 +435,7 @@ static void kmalloc_uaf_16(struct kunit *test)
>  static void kmalloc_oob_memset_2(struct kunit *test)
>  {
>  	char *ptr;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  
>  	ptr = kmalloc(size, GFP_KERNEL);
>  	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
> @@ -447,7 +447,7 @@ static void kmalloc_oob_memset_2(struct kunit *test)
>  static void kmalloc_oob_memset_4(struct kunit *test)
>  {
>  	char *ptr;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  
>  	ptr = kmalloc(size, GFP_KERNEL);
>  	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
> @@ -459,7 +459,7 @@ static void kmalloc_oob_memset_4(struct kunit *test)
>  static void kmalloc_oob_memset_8(struct kunit *test)
>  {
>  	char *ptr;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  
>  	ptr = kmalloc(size, GFP_KERNEL);
>  	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
> @@ -471,7 +471,7 @@ static void kmalloc_oob_memset_8(struct kunit *test)
>  static void kmalloc_oob_memset_16(struct kunit *test)
>  {
>  	char *ptr;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  
>  	ptr = kmalloc(size, GFP_KERNEL);
>  	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
> @@ -483,7 +483,7 @@ static void kmalloc_oob_memset_16(struct kunit *test)
>  static void kmalloc_oob_in_memset(struct kunit *test)
>  {
>  	char *ptr;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  
>  	ptr = kmalloc(size, GFP_KERNEL);
>  	KUNIT_ASSERT_NOT_ERR_OR_NULL(test, ptr);
> diff --git a/lib/test_kasan_module.c b/lib/test_kasan_module.c
> index 7ebf433edef3..c8cc77b1dcf3 100644
> --- a/lib/test_kasan_module.c
> +++ b/lib/test_kasan_module.c
> @@ -19,7 +19,7 @@ static noinline void __init copy_user_test(void)
>  {
>  	char *kmem;
>  	char __user *usermem;
> -	size_t size = 128 - KASAN_GRANULE_SIZE;
> +	volatile size_t size = 128 - KASAN_GRANULE_SIZE;
>  	int __maybe_unused unused;
>  
>  	kmem = kmalloc(size, GFP_KERNEL);
> -- 
> 2.30.2
> 

^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH] kasan: test: Bypass __alloc_size checks
  2021-10-06 11:38 ` Mark Rutland
@ 2021-10-06 16:33   ` Kees Cook
  0 siblings, 0 replies; 3+ messages in thread
From: Kees Cook @ 2021-10-06 16:33 UTC (permalink / raw)
  To: Mark Rutland
  Cc: Andrew Morton, Andrey Ryabinin, Alexander Potapenko,
	Andrey Konovalov, Dmitry Vyukov, kasan-dev, linux-kernel,
	linux-hardening

On Wed, Oct 06, 2021 at 12:38:36PM +0100, Mark Rutland wrote:
> Hi Kees,
> 
> On Tue, Oct 05, 2021 at 08:55:22PM -0700, Kees Cook wrote:
> > Intentional overflows, as performed by the KASAN tests, are detected
> > at compile time[1] (instead of only at run-time) with the addition of
> > __alloc_size. Fix this by forcing the compiler into not being able to
> > trust the size used following the kmalloc()s.
> 
> It might be better to use OPTIMIZER_HIDE_VAR(), since that's intended to
> make the value opaque to the compiler, and volatile might not always do
> that depending on how the compiler tracks the variable.

Given both you and Jann[1] have suggested this, I'll send a v2 with that.
:) Thanks!

-Kees

[1] https://lore.kernel.org/lkml/CAG48ez19raco+s+UF8eiXqTvaDEoMAo6_qmW2KdO24QDpmZpFQ@mail.gmail.com/

-- 
Kees Cook

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2021-10-06 16:33 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2021-10-06  3:55 [PATCH] kasan: test: Bypass __alloc_size checks Kees Cook
2021-10-06 11:38 ` Mark Rutland
2021-10-06 16:33   ` Kees Cook

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).