linux-kernel - Re: linux-next: Tree for May 16 (security/landlock/ruleset.c)

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 linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

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

Message-ID: <d645a0a2-4ffd-4dc6-b8a9-522ec3d27d7f@infradead.org>
Date: Mon, 19 May 2025 13:26:52 -0700
From: Randy Dunlap <rdunlap@...radead.org>
To: Kees Cook <kees@...nel.org>, Mickaël Salaün
 <mic@...ikod.net>
Cc: Steven Rostedt <rostedt@...dmis.org>,
 Stephen Rothwell <sfr@...b.auug.org.au>,
 Linux Next Mailing List <linux-next@...r.kernel.org>,
 Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
 linux-security-module@...r.kernel.org, Günther Noack
 <gnoack@...gle.com>
Subject: Re: linux-next: Tree for May 16 (security/landlock/ruleset.c)



> From 6fbf66fdfd0a7dac809b77faafdd72c60112bb8d Mon Sep 17 00:00:00 2001
> From: Kees Cook <kees@...nel.org>
> Date: Mon, 19 May 2025 11:52:06 -0700
> Subject: [PATCH] string.h: Provide basic sanity checks for fallback memcpy()
> MIME-Version: 1.0
> Content-Type: text/plain; charset=UTF-8
> Content-Transfer-Encoding: 8bit
> 
> Instead of defining memcpy() in terms of __builtin_memcpy() deep
> in arch/x86/include/asm/string_32.h, notice that it is needed up in
> the general string.h, as done with other common C String APIs. This
> allows us to add basic sanity checking for pathological "size"
> arguments to memcpy(). Besides the run-time checking benefit, this
> avoids GCC trying to be very smart about value range tracking[1] when
> CONFIG_PROFILE_ALL_BRANCHES=y but FORTIFY_SOURCE=n.
> 
> Link: https://lore.kernel.org/all/202505191117.C094A90F88@keescook/ [1]
> Reported-by: kernel test robot <lkp@...el.com>
> Closes: https://lore.kernel.org/all/202501040747.S3LYfvYq-lkp@intel.com/
> Reported-by: Randy Dunlap <rdunlap@...radead.org>
> Closes: https://lore.kernel.org/all/e3754f69-1dea-4542-8de0-a567a14fb95b@infradead.org/
> Signed-off-by: Kees Cook <kees@...nel.org>

Tested-by: Randy Dunlap <rdunlap@...radead.org>

Thanks.

> ---
> Cc: "Mickaël Salaün" <mic@...ikod.net>
> Cc: Thomas Gleixner <tglx@...utronix.de>
> Cc: Ingo Molnar <mingo@...hat.com>
> Cc: Borislav Petkov <bp@...en8.de>
> Cc: Dave Hansen <dave.hansen@...ux.intel.com>
> Cc: <x86@...nel.org>
> Cc: "H. Peter Anvin" <hpa@...or.com>
> Cc: Andy Shevchenko <andy@...nel.org>
> Cc: Uros Bizjak <ubizjak@...il.com>
> Cc: <linux-hardening@...r.kernel.org>
> ---
>  arch/x86/include/asm/string_32.h |  6 ------
>  include/linux/string.h           | 13 +++++++++++++
>  2 files changed, 13 insertions(+), 6 deletions(-)
> 
> diff --git a/arch/x86/include/asm/string_32.h b/arch/x86/include/asm/string_32.h
> index e9cce169bb4c..74397c95fa37 100644
> --- a/arch/x86/include/asm/string_32.h
> +++ b/arch/x86/include/asm/string_32.h
> @@ -145,12 +145,6 @@ static __always_inline void *__constant_memcpy(void *to, const void *from,
>  #define __HAVE_ARCH_MEMCPY
>  extern void *memcpy(void *, const void *, size_t);
>  
> -#ifndef CONFIG_FORTIFY_SOURCE
> -
> -#define memcpy(t, f, n) __builtin_memcpy(t, f, n)
> -
> -#endif /* !CONFIG_FORTIFY_SOURCE */
> -
>  #define __HAVE_ARCH_MEMMOVE
>  void *memmove(void *dest, const void *src, size_t n);
>  
> diff --git a/include/linux/string.h b/include/linux/string.h
> index 01621ad0f598..ffcee31a14f9 100644
> --- a/include/linux/string.h
> +++ b/include/linux/string.h
> @@ -3,6 +3,7 @@
>  #define _LINUX_STRING_H_
>  
>  #include <linux/args.h>
> +#include <linux/bug.h>
>  #include <linux/array_size.h>
>  #include <linux/cleanup.h>	/* for DEFINE_FREE() */
>  #include <linux/compiler.h>	/* for inline */
> @@ -390,7 +391,19 @@ static inline const char *kbasename(const char *path)
>  
>  #if !defined(__NO_FORTIFY) && defined(__OPTIMIZE__) && defined(CONFIG_FORTIFY_SOURCE)
>  #include <linux/fortify-string.h>
> +#else
> +/* Basic sanity checking even without FORTIFY_SOURCE */
> +# ifndef __HAVE_ARCH_MEMCPY
> +#  define memcpy(t, f, n)					\
> +	do {							\
> +		typeof(n) __n = (n);				\
> +		/* Skip impossible sizes. */			\
> +		if (!WARN_ON(__n < 0 || __n == SIZE_MAX))	\
> +			__builtin_memcpy(t, f, __n);		\
> +	} while (0)
> +# endif
>  #endif
> +
>  #ifndef unsafe_memcpy
>  #define unsafe_memcpy(dst, src, bytes, justification)		\
>  	memcpy(dst, src, bytes)

-- 
~Randy