Message-ID: <CAP-5=fUDQ6NTLWVfA2B+3022D6fZjvAksKH5EJ9Agnd1Qzvobw@mail.gmail.com>
Date: Fri, 16 Jan 2026 21:25:19 -0800
From: Ian Rogers <irogers@...gle.com>
To: Thomas Gleixner <tglx@...nel.org>
Cc: Peter Zijlstra <peterz@...radead.org>, kernel test robot <lkp@...el.com>, oe-kbuild-all@...ts.linux.dev, 
	linux-kernel@...r.kernel.org, x86@...nel.org, sparse@...isli.org, 
	linux-sparse@...r.kernel.org, Marco Elver <elver@...gle.com>
Subject: Re: [PATCH] compiler: Use __typeof_unqual__() for __unqual_scalar_typeof()

On Fri, Jan 16, 2026 at 10:18 AM Thomas Gleixner <tglx@...nel.org> wrote:
>
> From: Peter Zijlstra <peterz@...radead.org>
>
> The recent changes to get_unaligned() resulted in a new sparse warning:
>
>    net/rds/ib_cm.c:96:35: sparse: sparse: incorrect type in argument 1 (different modifiers) @@     expected void * @@     got restricted __be64 const * @@
>    net/rds/ib_cm.c:96:35: sparse:     expected void *
>    net/rds/ib_cm.c:96:35: sparse:     got restricted __be64 const *
>
> The updated get_unaligned_t() uses __unqual_scalar_typeof() to get an
> unqualified type. This works correctly for the compilers, but fails for
> sparse when the data type is __be64 (or any other __beNN variant).
>
> On sparse runs (C=[12]) __beNN types are annotated with
> __attribute__((bitwise)).
>
> That annotation allows sparse to detect incompatible operations on __beNN
> variables, but it also prevents sparse from evaluating the _Generic() in
> __unqual_scalar_typeof() and map __beNN to a unqualified scalar type, so it
> ends up with the default, i.e. the original qualified type of a 'const
> __beNN' pointer. That then ends up as the first pointer argument to
> builtin_memcpy(), which obviously causes the above sparse warnings.
>
> The sparse git tree supports typeof_unqual() now, which allows to use it
> instead of the _Generic() based __unqual_scalar_typeof(). With that sparse
> correctly evaluates the unqualified type and keeps the __beNN logic intact.

Wow, that's painful. Congratulations on finding the root case.

Acked-by: Ian Rogers <irogers@...gle.com>

Thanks,
Ian

> The downside is that this requires a top of tree sparse build and an old
> sparse version will emit a metric ton of incomprehensible error messages
> before it dies with a segfault.
>
> Therefore implement a sanity check which validates that the checker is
> available and capable of handling typeof_unqual(). Emit a warning if not so
> the user can take informed action.
>
> [ tglx: Move the evaluation of USE_TYPEOF_UNQUAL to compiler_types.h so it is
>         set before use and implement the sanity checker ]
>
> Reported-by: kernel test robot <lkp@...el.com>
> Signed-off-by: Peter Zijlstra <peterz@...radead.org>
> Signed-off-by: Thomas Gleixner <tglx@...nel.org>
> Closes: https://lore.kernel.org/oe-kbuild-all/202601150001.sKSN644a-lkp@intel.com/
> ---
>  Makefile                       |    8 ++++++++
>  include/linux/compiler.h       |   10 ----------
>  include/linux/compiler_types.h |   11 +++++++++++
>  scripts/checker-valid.sh       |   19 +++++++++++++++++++
>  4 files changed, 38 insertions(+), 10 deletions(-)
>
> --- a/Makefile
> +++ b/Makefile
> @@ -1178,6 +1178,14 @@ ifdef CONFIG_CC_IS_CLANG
>  KBUILD_USERLDFLAGS += --ld-path=$(LD)
>  endif
>
> +# Validate the checker is available and functional
> +ifneq ($(KBUILD_CHECKSRC), 0)
> +  ifneq ($(shell $(srctree)/scripts/checker-valid.sh $(CHECK)), 1)
> +    $(warning C=$(KBUILD_CHECKSRC) specified, but $(CHECK) is not available or not up to date)
> +    KBUILD_CHECKSRC = 0
> +  endif
> +endif
> +
>  # make the checker run with the right architecture
>  CHECKFLAGS += --arch=$(ARCH)
>
> --- a/include/linux/compiler.h
> +++ b/include/linux/compiler.h
> @@ -231,16 +231,6 @@ void ftrace_likely_update(struct ftrace_
>                                 "must be non-C-string (not NUL-terminated)")
>
>  /*
> - * Use __typeof_unqual__() when available.
> - *
> - * XXX: Remove test for __CHECKER__ once
> - * sparse learns about __typeof_unqual__().
> - */
> -#if CC_HAS_TYPEOF_UNQUAL && !defined(__CHECKER__)
> -# define USE_TYPEOF_UNQUAL 1
> -#endif
> -
> -/*
>   * Define TYPEOF_UNQUAL() to use __typeof_unqual__() as typeof
>   * operator when available, to return an unqualified type of the exp.
>   */
> --- a/include/linux/compiler_types.h
> +++ b/include/linux/compiler_types.h
> @@ -562,6 +562,13 @@ struct ftrace_likely_data {
>  #define asm_inline asm
>  #endif
>
> +/*
> + * Use __typeof_unqual__() when available.
> + */
> +#if CC_HAS_TYPEOF_UNQUAL || defined(__CHECKER__)
> +# define USE_TYPEOF_UNQUAL 1
> +#endif
> +
>  /* Are two types/vars the same type (ignoring qualifiers)? */
>  #define __same_type(a, b) __builtin_types_compatible_p(typeof(a), typeof(b))
>
> @@ -569,6 +576,7 @@ struct ftrace_likely_data {
>   * __unqual_scalar_typeof(x) - Declare an unqualified scalar type, leaving
>   *                            non-scalar types unchanged.
>   */
> +#ifndef USE_TYPEOF_UNQUAL
>  /*
>   * Prefer C11 _Generic for better compile-times and simpler code. Note: 'char'
>   * is not type-compatible with 'signed char', and we define a separate case.
> @@ -586,6 +594,9 @@ struct ftrace_likely_data {
>                          __scalar_type_to_expr_cases(long),             \
>                          __scalar_type_to_expr_cases(long long),        \
>                          default: (x)))
> +#else
> +#define __unqual_scalar_typeof(x) __typeof_unqual__(x)
> +#endif
>
>  /* Is this type a native word size -- useful for atomic operations */
>  #define __native_word(t) \
> --- /dev/null
> +++ b/scripts/checker-valid.sh
> @@ -0,0 +1,19 @@
> +#!/bin/sh -eu
> +# SPDX-License-Identifier: GPL-2.0
> +
> +[ ! -x "$(command -v "$1")" ] && exit 1
> +
> +tmp_file=$(mktemp)
> +trap "rm -f $tmp_file" EXIT
> +
> +cat << EOF >$tmp_file
> +static inline int u(const int *q)
> +{
> +       __typeof_unqual__(*q) v = *q;
> +       return v;
> +}
> +EOF
> +
> +# sparse happily exits with 0 on error so validate
> +# there is none on stderr. Use awk as grep is a pain with sh -e
> +$1 $tmp_file 2>&1 | awk -v c=1 '/error/{c=0}END{print c}'

Powered by blists - more mailing lists