linux-kernel - uNNw: wrapping integer annotations

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [day] [month] [year] [list]

Date:   Thu, 30 Aug 2018 16:05:34 +0300
From:   Alexey Dobriyan <adobriyan@...il.com>
To:     keescook@...omium.org
Cc:     linux-kernel@...r.kernel.org
Subject: uNNw: wrapping integer annotations

LWN wrote:
> Marking places where unsigned overflow is expected is needed; it would be
> nice to get those annotations put into the kernel, Cook said.

Let's discuss the most important part -- naming :^)

In my opinion, signed universe should _not_ get anything before specific
examples are presented where the value is in fact can be negative and
can overflow.

As for size_t/uintptr_t/..., they can be added as usize_w Rust-style
or left until someone presents a usecase

--- a/include/linux/types.h
+++ b/include/linux/types.h
@@ -119,6 +119,18 @@ typedef s64			int64_t;
 #define aligned_be64		__aligned_be64
 #define aligned_le64		__aligned_le64
 
+#ifdef __CHECKER__	// change to sanitizer define
+#define __wrap __attribute__((whatever))
+#else
+#define __wrap
+#endif
+/* wrapping is OK */
+typedef u8 __wrap u8w;
+typedef u16 __wrap u16w;
+typedef u32 __wrap u32w;
+typedef u64 __wrap u64w;
+typedef unsigned long __wrap unsigned_long_w;	// name sucks but consistently so
+
 /**
  * The type used for indexing onto a disc or disc partition.
  *