linux-kernel - Re: [PATCH 7/8] locking/atomics, kcsan: Add KCSAN instrumentation

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

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

Message-ID: <20191016111847.GB44246@lakrids.cambridge.arm.com>
Date:   Wed, 16 Oct 2019 12:18:47 +0100
From:   Mark Rutland <mark.rutland@....com>
To:     Marco Elver <elver@...gle.com>
Cc:     akiyks@...il.com, stern@...land.harvard.edu, glider@...gle.com,
        parri.andrea@...il.com, andreyknvl@...gle.com, luto@...nel.org,
        ard.biesheuvel@...aro.org, arnd@...db.de, boqun.feng@...il.com,
        bp@...en8.de, dja@...ens.net, dlustig@...dia.com,
        dave.hansen@...ux.intel.com, dhowells@...hat.com,
        dvyukov@...gle.com, hpa@...or.com, mingo@...hat.com,
        j.alglave@....ac.uk, joel@...lfernandes.org, corbet@....net,
        jpoimboe@...hat.com, luc.maranget@...ia.fr, npiggin@...il.com,
        paulmck@...ux.ibm.com, peterz@...radead.org, tglx@...utronix.de,
        will@...nel.org, kasan-dev@...glegroups.com,
        linux-arch@...r.kernel.org, linux-doc@...r.kernel.org,
        linux-efi@...r.kernel.org, linux-kbuild@...r.kernel.org,
        linux-kernel@...r.kernel.org, linux-mm@...ck.org, x86@...nel.org
Subject: Re: [PATCH 7/8] locking/atomics, kcsan: Add KCSAN instrumentation

Hi Marco,

On Wed, Oct 16, 2019 at 10:39:58AM +0200, Marco Elver wrote:
> This adds KCSAN instrumentation to atomic-instrumented.h.
> 
> Signed-off-by: Marco Elver <elver@...gle.com>
> ---
>  include/asm-generic/atomic-instrumented.h | 192 +++++++++++++++++++++-
>  scripts/atomic/gen-atomic-instrumented.sh |   9 +-
>  2 files changed, 199 insertions(+), 2 deletions(-)
> 
> diff --git a/include/asm-generic/atomic-instrumented.h b/include/asm-generic/atomic-instrumented.h
> index e8730c6b9fe2..9e487febc610 100644
> --- a/include/asm-generic/atomic-instrumented.h
> +++ b/include/asm-generic/atomic-instrumented.h
> @@ -19,11 +19,13 @@
>  
>  #include <linux/build_bug.h>
>  #include <linux/kasan-checks.h>
> +#include <linux/kcsan-checks.h>
>  
>  static inline int
>  atomic_read(const atomic_t *v)
>  {
>  	kasan_check_read(v, sizeof(*v));
> +	kcsan_check_atomic(v, sizeof(*v), false);

For legibility and consistency with kasan, it would be nicer to avoid
the bool argument here and have kcsan_check_atomic_{read,write}()
helpers...

> diff --git a/scripts/atomic/gen-atomic-instrumented.sh b/scripts/atomic/gen-atomic-instrumented.sh
> index e09812372b17..c0553743a6f4 100755
> --- a/scripts/atomic/gen-atomic-instrumented.sh
> +++ b/scripts/atomic/gen-atomic-instrumented.sh
> @@ -12,15 +12,20 @@ gen_param_check()
>  	local type="${arg%%:*}"
>  	local name="$(gen_param_name "${arg}")"
>  	local rw="write"
> +	local is_write="true"
>  
>  	case "${type#c}" in
>  	i) return;;
>  	esac
>  
>  	# We don't write to constant parameters
> -	[ ${type#c} != ${type} ] && rw="read"
> +	if [ ${type#c} != ${type} ]; then
> +		rw="read"
> +		is_write="false"
> +	fi
>  
>  	printf "\tkasan_check_${rw}(${name}, sizeof(*${name}));\n"
> +	printf "\tkcsan_check_atomic(${name}, sizeof(*${name}), ${is_write});\n"

... which would also simplify this.

Though as below, we might want to wrap both in a helper local to
atomic-instrumented.h.

>  }
>  
>  #gen_param_check(arg...)
> @@ -108,6 +113,7 @@ cat <<EOF
>  ({									\\
>  	typeof(ptr) __ai_ptr = (ptr);					\\
>  	kasan_check_write(__ai_ptr, ${mult}sizeof(*__ai_ptr));		\\
> +	kcsan_check_atomic(__ai_ptr, ${mult}sizeof(*__ai_ptr), true);	\\
>  	arch_${xchg}(__ai_ptr, __VA_ARGS__);				\\
>  })
>  EOF
> @@ -148,6 +154,7 @@ cat << EOF
>  
>  #include <linux/build_bug.h>
>  #include <linux/kasan-checks.h>
> +#include <linux/kcsan-checks.h>

We could add the following to this preamble:

static inline void __atomic_check_read(const volatile void *v, size_t size)
{
	kasan_check_read(v, sizeof(*v));
	kcsan_check_atomic(v, sizeof(*v), false);
}

static inline void __atomic_check_write(const volatile void *v, size_t size)
{
	kasan_check_write(v, sizeof(*v));
	kcsan_check_atomic(v, sizeof(*v), true);
}

... and only have the one call in each atomic wrapper.

Otherwise, this looks good to me.

Thanks,
Mark.