Re: [PATCH v2] minmax: clamp more efficiently by avoiding extra comparison

From: Kees Cook
Date: Fri Sep 23 2022 - 12:42:17 EST


On Fri, Sep 23, 2022 at 05:40:01PM +0200, Jason A. Donenfeld wrote:
> Currently the clamp algorithm does:
>
> if (val > hi)
> val = hi;
> if (val < lo)
> val = lo;
>
> But since hi > lo by definition, this can be made more efficient with:
>
> if (val > hi)
> val = hi;
> else if (val < lo)
> val = lo;
>
> So fix up the clamp and clamp_t functions to do this, adding the same
> argument checking as for min and min_t.
>
> Cc: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
> Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
> Cc: Kees Cook <keescook@xxxxxxxxxxxx>
> Signed-off-by: Jason A. Donenfeld <Jason@xxxxxxxxx>
> ---
> include/linux/minmax.h | 25 +++++++++++++++++++++++--
> 1 file changed, 23 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/minmax.h b/include/linux/minmax.h
> index 5433c08fcc68..30e2e2cd0f44 100644
> --- a/include/linux/minmax.h
> +++ b/include/linux/minmax.h
> @@ -37,6 +37,27 @@
> __cmp(x, y, op), \
> __cmp_once(x, y, __UNIQUE_ID(__x), __UNIQUE_ID(__y), op))
>
> +#define __clamp(val, lo, hi) \
> + ((val) >= (hi) ? (hi) : ((val) <= (lo) ? (lo) : (val)))
> +
> +#define __clamp_once(val, lo, hi, unique_val, unique_lo, unique_hi) ({ \
> + typeof(val) unique_val = (val); \
> + typeof(lo) unique_lo = (lo); \
> + typeof(hi) unique_hi = (hi); \
> + __clamp(unique_val, unique_lo, unique_hi); })
> +
> +#define __clamp_input_check(lo, hi) \
> + (BUILD_BUG_ON_ZERO(__builtin_choose_expr( \
> + __is_constexpr((lo) > (hi)), (lo) > (hi), false)))

Nice. :)

> +
> +#define __careful_clamp(val, lo, hi) ({ \
> + __clamp_input_check(lo, hi) + \
> + __builtin_choose_expr(__typecheck(val, lo) && __typecheck(val, hi) && \
> + __typecheck(hi, lo) && __is_constexpr(val) && \
> + __is_constexpr(lo) && __is_constexpr(hi), \

I really like it! I might have used:

__safe_cmp(val, lo) && __safe_cmp(val, hi)

instead of the "open coded" __typecheck()s and __is_constexpr()s, but
it's the same result.

> + __clamp(val, lo, hi), \
> + __clamp_once(val, lo, hi, __UNIQUE_ID(__val), __UNIQUE_ID(__lo), __UNIQUE_ID(__hi))); })

*complaint about line being >100 characters, but I don't really care* If
anyone is really bothered, this looks fine, too:

__clamp_once(val, lo, hi, \
__UNIQUE_ID(__val), __UNIQUE_ID(__lo), __UNIQUE_ID(__hi))); })

*shrug*

> +
> /**
> * min - return minimum of two values of the same or compatible types
> * @x: first value
> @@ -86,7 +107,7 @@
> * This macro does strict typechecking of @lo/@hi to make sure they are of the
> * same type as @val. See the unnecessary pointer comparisons.
> */
> -#define clamp(val, lo, hi) min((typeof(val))max(val, lo), hi)
> +#define clamp(val, lo, hi) __careful_clamp(val, lo, hi)
>
> /*
> * ..and if you can't take the strict
> @@ -121,7 +142,7 @@
> * This macro does no typechecking and uses temporary variables of type
> * @type to make all the comparisons.
> */
> -#define clamp_t(type, val, lo, hi) min_t(type, max_t(type, val, lo), hi)
> +#define clamp_t(type, val, lo, hi) __careful_clamp((type)(val), (type)(lo), (type)(hi))
>
> /**
> * clamp_val - return a value clamped to a given range using val's type
> --
> 2.37.3
>

Reviewed-by: Kees Cook <keescook@xxxxxxxxxxxx>

I can take this unless akpm wants it?

--
Kees Cook