[v2] Compiler Attributes: Add __counted_by macro

  In an effort to annotate all flexible array members with their run-time
size information, the "element_count" attribute is being introduced by
Clang[1] and GCC[2] in future releases. This annotation will provide
the CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE features the ability
to perform run-time bounds checking on otherwise unknown-size flexible
arrays.

Even though the attribute is under development, we can start the
annotation process in the kernel. This requires defining a macro for
it, even if we have to change the name of the actual attribute later.
Since it is likely that this attribute may change its name to "counted_by"
in the future (to better align with a future total bytes "sized_by"
attribute), name the wrapper macro "__counted_by", which also reads more
clearly (and concisely) in structure definitions.

[1] https://reviews.llvm.org/D148381
[2] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108896

Cc: Miguel Ojeda <ojeda@kernel.org>
Cc: Bill Wendling <morbo@google.com>
Cc: Qing Zhao <qing.zhao@oracle.com>
Cc: Gustavo A. R. Silva <gustavoars@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Tom Rix <trix@redhat.com>
Cc: llvm@lists.linux.dev
Signed-off-by: Kees Cook <keescook@chromium.org>
---
v2: - update "Optional" comments
v1: https://lore.kernel.org/all/20230504181636.never.222-kees@kernel.org/
---
 include/linux/compiler_attributes.h | 13 +++++++++++++
 1 file changed, 13 insertions(+)
  

On Wed, May 17, 2023 at 12:08:44PM -0700, Kees Cook wrote:
> In an effort to annotate all flexible array members with their run-time
> size information, the "element_count" attribute is being introduced by
> Clang[1] and GCC[2] in future releases. This annotation will provide
> the CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE features the ability
> to perform run-time bounds checking on otherwise unknown-size flexible
> arrays.
> 
> Even though the attribute is under development, we can start the
> annotation process in the kernel. This requires defining a macro for
> it, even if we have to change the name of the actual attribute later.
> Since it is likely that this attribute may change its name to "counted_by"
> in the future (to better align with a future total bytes "sized_by"
> attribute), name the wrapper macro "__counted_by", which also reads more
> clearly (and concisely) in structure definitions.
> 
> [1] https://reviews.llvm.org/D148381
> [2] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108896
> 
> Cc: Miguel Ojeda <ojeda@kernel.org>
> Cc: Bill Wendling <morbo@google.com>
> Cc: Qing Zhao <qing.zhao@oracle.com>
> Cc: Gustavo A. R. Silva <gustavoars@kernel.org>
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Nathan Chancellor <nathan@kernel.org>
> Cc: Tom Rix <trix@redhat.com>
> Cc: llvm@lists.linux.dev
> Signed-off-by: Kees Cook <keescook@chromium.org>

Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>

Thanks
--
Gustavo

> ---
> v2: - update "Optional" comments
> v1: https://lore.kernel.org/all/20230504181636.never.222-kees@kernel.org/
> ---
>  include/linux/compiler_attributes.h | 13 +++++++++++++
>  1 file changed, 13 insertions(+)
> 
> diff --git a/include/linux/compiler_attributes.h b/include/linux/compiler_attributes.h
> index e659cb6fded3..a92d8887e8f0 100644
> --- a/include/linux/compiler_attributes.h
> +++ b/include/linux/compiler_attributes.h
> @@ -123,6 +123,19 @@
>  # define __designated_init
>  #endif
>  
> +/*
> + * Optional: only supported since gcc >= 14
> + * Optional: only supported since clang >= 17
> + *
> + *   gcc: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108896
> + * clang: https://reviews.llvm.org/D148381
> + */
> +#if __has_attribute(__element_count__)
> +# define __counted_by(member)		__attribute__((__element_count__(member)))
> +#else
> +# define __counted_by(member)
> +#endif
> +
>  /*
>   * Optional: only supported since clang >= 14.0
>   *
> -- 
> 2.34.1
>
  

On Wed, May 17, 2023 at 12:08:44PM -0700, Kees Cook wrote:
> In an effort to annotate all flexible array members with their run-time
> size information, the "element_count" attribute is being introduced by
> Clang[1] and GCC[2] in future releases. This annotation will provide
> the CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE features the ability
> to perform run-time bounds checking on otherwise unknown-size flexible
> arrays.
> 
> Even though the attribute is under development, we can start the
> annotation process in the kernel. This requires defining a macro for
> it, even if we have to change the name of the actual attribute later.
> Since it is likely that this attribute may change its name to "counted_by"
> in the future (to better align with a future total bytes "sized_by"
> attribute), name the wrapper macro "__counted_by", which also reads more
> clearly (and concisely) in structure definitions.
> 
> [1] https://reviews.llvm.org/D148381
> [2] https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108896
> 
> Cc: Miguel Ojeda <ojeda@kernel.org>
> Cc: Bill Wendling <morbo@google.com>
> Cc: Qing Zhao <qing.zhao@oracle.com>
> Cc: Gustavo A. R. Silva <gustavoars@kernel.org>
> Cc: Nick Desaulniers <ndesaulniers@google.com>
> Cc: Nathan Chancellor <nathan@kernel.org>
> Cc: Tom Rix <trix@redhat.com>
> Cc: llvm@lists.linux.dev
> Signed-off-by: Kees Cook <keescook@chromium.org>

Reviewed-by: Nathan Chancellor <nathan@kernel.org>

> ---
> v2: - update "Optional" comments
> v1: https://lore.kernel.org/all/20230504181636.never.222-kees@kernel.org/
> ---
>  include/linux/compiler_attributes.h | 13 +++++++++++++
>  1 file changed, 13 insertions(+)
> 
> diff --git a/include/linux/compiler_attributes.h b/include/linux/compiler_attributes.h
> index e659cb6fded3..a92d8887e8f0 100644
> --- a/include/linux/compiler_attributes.h
> +++ b/include/linux/compiler_attributes.h
> @@ -123,6 +123,19 @@
>  # define __designated_init
>  #endif
>  
> +/*
> + * Optional: only supported since gcc >= 14
> + * Optional: only supported since clang >= 17
> + *
> + *   gcc: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108896
> + * clang: https://reviews.llvm.org/D148381
> + */
> +#if __has_attribute(__element_count__)
> +# define __counted_by(member)		__attribute__((__element_count__(member)))
> +#else
> +# define __counted_by(member)
> +#endif
> +
>  /*
>   * Optional: only supported since clang >= 14.0
>   *
> -- 
> 2.34.1
>
  

On Wed, 17 May 2023 12:08:44 -0700, Kees Cook wrote:
> In an effort to annotate all flexible array members with their run-time
> size information, the "element_count" attribute is being introduced by
> Clang[1] and GCC[2] in future releases. This annotation will provide
> the CONFIG_UBSAN_BOUNDS and CONFIG_FORTIFY_SOURCE features the ability
> to perform run-time bounds checking on otherwise unknown-size flexible
> arrays.
> 
> [...]

FYI, applied to for-next/hardening:

[1/1] Compiler Attributes: Add __counted_by macro
      https://git.kernel.org/kees/c/86a76e91cbab
  

On Fri, May 26, 2023 at 7:16 PM Kees Cook <keescook@chromium.org> wrote:
>
> FYI, applied to for-next/hardening:
>
> [1/1] Compiler Attributes: Add __counted_by macro
>       https://git.kernel.org/kees/c/86a76e91cbab

Sorry, I was going to apply it soon -- in case you want it:

    Acked-by: Miguel Ojeda <ojeda@kernel.org>

And thanks Nathan for resubmitting the `Reviewed-by` from v1!

Cheers,
Miguel
  

On Fri, May 26, 2023 at 07:47:03PM +0200, Miguel Ojeda wrote:
> On Fri, May 26, 2023 at 7:16 PM Kees Cook <keescook@chromium.org> wrote:
> >
> > FYI, applied to for-next/hardening:
> >
> > [1/1] Compiler Attributes: Add __counted_by macro
> >       https://git.kernel.org/kees/c/86a76e91cbab
> 
> Sorry, I was going to apply it soon -- in case you want it:
> 
>     Acked-by: Miguel Ojeda <ojeda@kernel.org>

Thanks!

> And thanks Nathan for resubmitting the `Reviewed-by` from v1!

Yes, apologies for missing this in my v2 submission!
  

On Fri, May 26, 2023 at 11:56 AM Kees Cook <keescook@chromium.org> wrote:
>
> On Fri, May 26, 2023 at 07:47:03PM +0200, Miguel Ojeda wrote:
> > On Fri, May 26, 2023 at 7:16 PM Kees Cook <keescook@chromium.org> wrote:
> > >
> > > FYI, applied to for-next/hardening:
> > >
> > > [1/1] Compiler Attributes: Add __counted_by macro
> > >       https://git.kernel.org/kees/c/86a76e91cbab
> >
> > Sorry, I was going to apply it soon -- in case you want it:
> >
> >     Acked-by: Miguel Ojeda <ojeda@kernel.org>
>
> Thanks!
>
> > And thanks Nathan for resubmitting the `Reviewed-by` from v1!
>
> Yes, apologies for missing this in my v2 submission!
>
> --
> Kees Cook
>

https://discourse.llvm.org/t/rfc-enforcing-bounds-safety-in-c-fbounds-safety/70854
proposes a macro __counted_by as well.
This patch uses the same name:

> # define __counted_by(member)          __attribute__((__element_count__(member)))

I wonder whether the two use cases are compatible so that using the
same macro name will be fine.

#if defined(__has_feature) && __has_feature(bounds_safety)
#define __counted_by(T) __attribute__((__counted_by__(T)))
// ... other bounds annotations
#else
#define __counted_by(T) // defined as nothing // ... other bounds annotations
#endif
  

On Fri, May 26, 2023 at 12:48:26PM -0700, Fangrui Song wrote:
> On Fri, May 26, 2023 at 11:56 AM Kees Cook <keescook@chromium.org> wrote:
> >
> > On Fri, May 26, 2023 at 07:47:03PM +0200, Miguel Ojeda wrote:
> > > On Fri, May 26, 2023 at 7:16 PM Kees Cook <keescook@chromium.org> wrote:
> > > >
> > > > FYI, applied to for-next/hardening:
> > > >
> > > > [1/1] Compiler Attributes: Add __counted_by macro
> > > >       https://git.kernel.org/kees/c/86a76e91cbab
> > >
> > > Sorry, I was going to apply it soon -- in case you want it:
> > >
> > >     Acked-by: Miguel Ojeda <ojeda@kernel.org>
> >
> > Thanks!
> >
> > > And thanks Nathan for resubmitting the `Reviewed-by` from v1!
> >
> > Yes, apologies for missing this in my v2 submission!
> >
> > --
> > Kees Cook
> >
> 
> https://discourse.llvm.org/t/rfc-enforcing-bounds-safety-in-c-fbounds-safety/70854
> proposes a macro __counted_by as well.
> This patch uses the same name:
> 
> > # define __counted_by(member)          __attribute__((__element_count__(member)))
> 
> I wonder whether the two use cases are compatible so that using the
> same macro name will be fine.

Yeah, I have suggest the name change for the GCC proposal. However,
given that there is still no code to test for -fbounds-safety, I'm
sticking with __element_count for the moment, as there is code
implementing that name in both GCC and Clang today.

> #if defined(__has_feature) && __has_feature(bounds_safety)
> #define __counted_by(T) __attribute__((__counted_by__(T)))
> // ... other bounds annotations
> #else
> #define __counted_by(T) // defined as nothing // ... other bounds annotations
> #endif

Right. My main consideration for getting __counted_by defined by the
kernel at all is so that annotation can begin. We can adjust the
define's contents as needed. :)