From: James Hogan <james.hogan@imgtec.com>
To: <linux-kernel@vger.kernel.org>
Cc: <linux-arch@vger.kernel.org>, <linux-mips@linux-mips.org>,
James Hogan <james.hogan@imgtec.com>,
Kees Cook <keescook@chromium.org>,
Andrew Morton <akpm@linux-foundation.org>
Subject: [PATCH 6/7] test_user_copy: Check user string accessors
Date: Wed, 5 Aug 2015 16:48:54 +0100 [thread overview]
Message-ID: <1438789735-4643-7-git-send-email-james.hogan@imgtec.com> (raw)
In-Reply-To: <1438789735-4643-1-git-send-email-james.hogan@imgtec.com>
Add basic success/failure checking of the user string functions which
copy or find the length of userland strings.
The following cases are checked:
- strncpy_from_user() with legitimate user to kernel addresses, illegal
all-kernel and reversed addresses, and legitimate all-kernel
addresses.
- strnlen_user()/strlen_user() with a legitimate user address, an
illegal kernel address, and a legitimate kernel address.
This caught a bug in the MIPS Enhanced Virtual Memory (EVA)
implementation:
test_user_copy: illegal strlen_user passed
Due to strlen_user() accidentally always using normal kernel loads, and
the EVA user/kernel address ranges overlapping.
New tests:
- legitimate strncpy_from_user
- legitimate strnlen_user
- legitimate strlen_user
- illegal all-kernel strncpy_from_user
- illegal reversed strncpy_from_user
- illegal strnlen_user
- illegal strlen_user
- legitimate all-kernel strncpy_from_user
- legitimate kernel strnlen_user
- legitimate kernel strlen_user
Signed-off-by: James Hogan <james.hogan@imgtec.com>
Cc: Kees Cook <keescook@chromium.org>
Cc: Andrew Morton <akpm@linux-foundation.org>
---
lib/test_user_copy.c | 23 +++++++++++++++++++++++
1 file changed, 23 insertions(+)
diff --git a/lib/test_user_copy.c b/lib/test_user_copy.c
index 6cbdb0a15ca2..6d05ec5f6cfa 100644
--- a/lib/test_user_copy.c
+++ b/lib/test_user_copy.c
@@ -72,6 +72,12 @@ static int __init test_user_copy_init(void)
"legitimate put_user failed");
ret |= test(clear_user(usermem, PAGE_SIZE) != 0,
"legitimate clear_user passed");
+ ret |= test(strncpy_from_user(kmem, usermem, PAGE_SIZE) < 0,
+ "legitimate strncpy_from_user failed");
+ ret |= test(strnlen_user(usermem, PAGE_SIZE) == 0,
+ "legitimate strnlen_user failed");
+ ret |= test(strlen_user(usermem) == 0,
+ "legitimate strlen_user failed");
ret |= test(!access_ok(VERIFY_READ, usermem, PAGE_SIZE * 2),
"legitimate access_ok VERIFY_READ failed");
@@ -122,6 +128,16 @@ static int __init test_user_copy_init(void)
"illegal put_user passed");
ret |= test(clear_user((char __user *)kmem, PAGE_SIZE) != PAGE_SIZE,
"illegal kernel clear_user passed");
+ ret |= test(strncpy_from_user(kmem, (char __user *)(kmem + PAGE_SIZE),
+ PAGE_SIZE) >= 0,
+ "illegal all-kernel strncpy_from_user passed");
+ ret |= test(strncpy_from_user(bad_usermem, (char __user *)kmem,
+ PAGE_SIZE) >= 0,
+ "illegal reversed strncpy_from_user passed");
+ ret |= test(strnlen_user((char __user *)kmem, PAGE_SIZE) != 0,
+ "illegal strnlen_user passed");
+ ret |= test(strlen_user((char __user *)kmem) != 0,
+ "illegal strlen_user passed");
/*
* If unchecked user accesses (__*) on this architecture cannot access
@@ -201,6 +217,13 @@ static int __init test_user_copy_init(void)
"legitimate kernel put_user failed");
ret |= test(clear_user((char __user *)kmem, PAGE_SIZE) != 0,
"legitimate kernel clear_user failed");
+ ret |= test(strncpy_from_user(kmem, (char __user *)(kmem + PAGE_SIZE),
+ PAGE_SIZE) < 0,
+ "legitimate all-kernel strncpy_from_user failed");
+ ret |= test(strnlen_user((char __user *)kmem, PAGE_SIZE) == 0,
+ "legitimate kernel strnlen_user failed");
+ ret |= test(strlen_user((char __user *)kmem) == 0,
+ "legitimate kernel strlen_user failed");
ret |= test(!access_ok(VERIFY_READ, (char __user *)kmem, PAGE_SIZE * 2),
"legitimate kernel access_ok VERIFY_READ failed");
--
2.3.6
next prev parent reply other threads:[~2015-08-05 15:50 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-08-05 15:48 [PATCH 0/7] test_user_copy improvements James Hogan
2015-08-05 15:48 ` [PATCH 1/7] test_user_copy: Check legit kernel accesses James Hogan
2015-08-05 15:48 ` [PATCH 2/7] test_user_copy: Check unchecked accessors James Hogan
2015-08-05 15:48 ` [PATCH 3/7] test_user_copy: Check __clear_user()/clear_user() James Hogan
2015-08-05 15:48 ` [PATCH 4/7] test_user_copy: Check __copy_in_user()/copy_in_user() James Hogan
2015-08-05 15:48 ` [PATCH 5/7] test_user_copy: Check __copy_{to,from}_user_inatomic() James Hogan
2015-08-05 15:48 ` James Hogan [this message]
2015-08-05 15:48 ` [PATCH 7/7] test_user_copy: Check user checksum functions James Hogan
2015-08-05 20:26 ` [PATCH 0/7] test_user_copy improvements Kees Cook
2015-08-06 16:28 ` James Hogan
2015-08-06 9:50 ` Guenter Roeck
2015-08-06 10:01 ` James Hogan
2015-08-06 15:02 ` James Hogan
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1438789735-4643-7-git-send-email-james.hogan@imgtec.com \
--to=james.hogan@imgtec.com \
--cc=akpm@linux-foundation.org \
--cc=keescook@chromium.org \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mips@linux-mips.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox
all inboxes | Powered by JetHome®