On Mon, Aug 24, 2026 at 9:02 AM <[email protected]> wrote:
>
> From: Kyrylo Tkachov <[email protected]>
>
> A VECTOR_CST can encode a stepped series without storing every lane. The
> fold-const negate_expr_p checks only the encoded elements. It can therefore
> approve a signed vector whose implicit final lane is INT_MIN.
>
> typedef int v4si __attribute__ ((vector_size (16)));
>
> const v4si c = { 2147483645, 2147483646, 2147483647,
> (-2147483647 - 1) };
>
> v4si f (v4si x) { return (-x) - c; }
> v4si g (v4si x) { return -(x + c); }
> v4si h () { return -c; }
>
> The predicate lets f and g negate c. It represents -INT_MIN as INT_MIN,
> which adds a false signed overflow for defined inputs. fold_negate_expr_1
> can also negate the encoded elements of c directly. This removes the
> required overflow diagnostic from h when overflow is instrumented.
>
> aarch64 -O2 -fsanitize=signed-integer-overflow
> -fsanitize-trap=signed-integer-overflow before:
>
> h:
> adrp x0, .LANCHOR0
> ldr q0, [x0, #:lo12:.LANCHOR0]
> ret
>
> After:
>
> h:
> brk #1000
>
> Reject a non-wrapping integral stepped vector in negate_expr_p. Also keep a
> direct stepped integral negation when sanitizer instrumentation makes
> overflow observable. Guard each TYPE_OVERFLOW_WRAPS query with its accepted
> integral type domain. VECTOR_CST can also represent fixed-point vectors,
> which the wrapping predicate does not accept. TYPE_OVERFLOW_SANITIZED
> accepts integral vectors.
>
> The two stepped checks have different conditions. negate_expr_p is a proof
> for moving a negation. Moving an implicit INT_MIN negation can introduce
> undefined overflow into f or g even when no overflow option is enabled, so
> that predicate rejects every non-wrapping integral stepped constant.
> fold_negate_expr_1 handles a direct source negation whose INT_MIN lane already
> overflows. GCC normally folds that constant overflow, and -ftrapv does not
> trap vector constant negation. Only sanitizer instrumentation makes the
> missing operation observable, so the direct-fold check is conditional on
> TYPE_OVERFLOW_SANITIZED.
>
> The sanitizer test checks that f and g do not report overflow for defined
> inputs. It also checks that h reports its real INT_MIN negation. The Arm
> test checks all six fixed-point vector modes in ordinary and saturating forms.
> Without the type-domain guards, a checking compiler fails. A release compiler
> also changes the results of the six saturating cases under -fwrapv.
>
> Bootstrapped and tested on aarch64-none-linux-gnu.
> Tested on arm-linux-gnueabihf with QEMU.
> Tested on x86_64-pc-linux-gnu.
> Ok for trunk?
> Thanks,
> Kyrill
>
> gcc/ChangeLog:
>
> * fold-const.cc (negate_expr_p): Reject non-wrapping stepped vector
> constants.
> (fold_negate_expr_1): Preserve sanitized stepped integral negations.
> * match.pd (negate_expr_p): Guard the wrapping query for vector
> constants.
>
> gcc/testsuite/ChangeLog:
>
> * g++.dg/ubsan/fold-negate-vector-1.C: New test.
> * gcc.target/arm/fixed-point-vector-negate-1.c: Likewise.
>
> Signed-off-by: Kyrylo Tkachov <[email protected]>
> ---
> gcc/fold-const.cc | 12 ++-
> gcc/match.pd | 3 +-
> .../g++.dg/ubsan/fold-negate-vector-1.C | 46 +++++++++++
> .../arm/fixed-point-vector-negate-1.c | 76 +++++++++++++++++++
> 4 files changed, 134 insertions(+), 3 deletions(-)
> create mode 100644 gcc/testsuite/g++.dg/ubsan/fold-negate-vector-1.C
> create mode 100644 gcc/testsuite/gcc.target/arm/fixed-point-vector-negate-1.c
>
> diff --git a/gcc/fold-const.cc b/gcc/fold-const.cc
> index 420e3185a2a..dc324c9cfdb 100644
> --- a/gcc/fold-const.cc
> +++ b/gcc/fold-const.cc
> @@ -406,10 +406,14 @@ negate_expr_p (tree t)
>
> case VECTOR_CST:
> {
> - if (FLOAT_TYPE_P (TREE_TYPE (type)) || TYPE_OVERFLOW_WRAPS (type))
> + if (FLOAT_TYPE_P (TREE_TYPE (type))
> + || (ANY_INTEGRAL_TYPE_P (type) && TYPE_OVERFLOW_WRAPS (type)))
> return true;
>
> - /* Steps don't prevent negation. */
> + /* An implicit element of a stepped vector can be the minimum
> + value. */
> + if (VECTOR_CST_STEPPED_P (t))
So if for example we encode { 1, +, 2 } then INT_MIN cannot be reached
without overflow. Do we allow overflows in the stepped evaluation? Likewise
if it is { 0, +, -3 } then INT_MIN is not reached, but -2147483647 and then
it would overflow.
So, can we improve this?
> + return false;
> unsigned int count = vector_cst_encoded_nelts (t);
> for (unsigned int i = 0; i < count; ++i)
> if (!negate_expr_p (VECTOR_CST_ENCODED_ELT (t, i)))
> @@ -566,6 +570,10 @@ fold_negate_expr_1 (location_t loc, tree t)
>
> case VECTOR_CST:
> {
> + if (VECTOR_CST_STEPPED_P (t)
> + && TYPE_OVERFLOW_SANITIZED (type))
> + return NULL_TREE;
> +
> tree_vector_builder elts;
> elts.new_unary_operation (type, t, true);
> unsigned int count = elts.encoded_nelts ();
> diff --git a/gcc/match.pd b/gcc/match.pd
> index bcfc7f0e8ce..a969aa20417 100644
> --- a/gcc/match.pd
> +++ b/gcc/match.pd
> @@ -2430,7 +2430,8 @@ DEFINE_INT_AND_FLOAT_ROUND_FN (RINT)
> ways. */
> (match negate_expr_p
> VECTOR_CST
> - (if (FLOAT_TYPE_P (TREE_TYPE (type)) || TYPE_OVERFLOW_WRAPS (type))))
> + (if (FLOAT_TYPE_P (TREE_TYPE (type))
> + || (ANY_INTEGRAL_TYPE_P (type) && TYPE_OVERFLOW_WRAPS (type)))))
> (match negate_expr_p
> (minus @0 @1)
> (if ((ANY_INTEGRAL_TYPE_P (type) && TYPE_OVERFLOW_WRAPS (type))
> diff --git a/gcc/testsuite/g++.dg/ubsan/fold-negate-vector-1.C
> b/gcc/testsuite/g++.dg/ubsan/fold-negate-vector-1.C
> new file mode 100644
> index 00000000000..cda0fd9817f
> --- /dev/null
> +++ b/gcc/testsuite/g++.dg/ubsan/fold-negate-vector-1.C
> @@ -0,0 +1,46 @@
> +// { dg-do run { target int32 } }
> +// { dg-options "-O2 -Wno-psabi -fsanitize=signed-integer-overflow" }
> +
> +#define INT_MAX __INT_MAX__
> +#define INT_MIN (-INT_MAX - 1)
> +
> +typedef int v4si __attribute__ ((vector_size (16)));
> +
> +const v4si c = { INT_MAX - 2, INT_MAX - 1, INT_MAX, INT_MIN };
> +
> +v4si __attribute__ ((noipa))
> +f (v4si x)
> +{
> + return (-x) - c;
> +}
> +
> +v4si __attribute__ ((noipa))
> +g (v4si x)
> +{
> + return -(x + c);
> +}
> +
> +v4si __attribute__ ((noipa))
> +h ()
> +{
> + return -c;
> +}
> +
> +int
> +main ()
> +{
> + v4si x = { 0, 0, 0, 1 };
> + v4si y = f (x);
> + if (y[3] != INT_MAX)
> + __builtin_abort ();
> +
> + y = g (x);
> + if (y[3] != INT_MAX)
> + __builtin_abort ();
> +
> + volatile v4si z = h ();
> + if (z[3] != INT_MIN)
> + __builtin_abort ();
> +}
> +
> +// { dg-output "negation of -2147483648 cannot be represented in type 'int';
> cast to an unsigned type to negate this value to itself" }
> diff --git a/gcc/testsuite/gcc.target/arm/fixed-point-vector-negate-1.c
> b/gcc/testsuite/gcc.target/arm/fixed-point-vector-negate-1.c
> new file mode 100644
> index 00000000000..e34f9633929
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/arm/fixed-point-vector-negate-1.c
> @@ -0,0 +1,76 @@
> +/* { dg-do run { target fixed_point } } */
> +/* { dg-require-effective-target arm_arch_v6_arm_multilib } */
> +/* { dg-options "-O2 -std=gnu99 -fwrapv" } */
> +/* { dg-add-options arm_arch_v6_arm } */
> +
> +#include <stdfix.h>
> +
> +/* Check all Arm fixed-point vector modes and their saturating forms. The
> + saturating cases make an invalid X - C to X + (-C) rewrite observable. */
> +
> +#define DEF2(NAME, TYPE, X, C, R) \
> + typedef TYPE NAME##_type __attribute__ ((vector_size (4))); \
> + static void __attribute__ ((noipa)) \
> + NAME (NAME##_type *out, const NAME##_type *in) \
> + { \
> + *out = *in - (NAME##_type) { C, C }; \
> + } \
> + static void \
> + check_##NAME (void) \
> + { \
> + NAME##_type in = { X, X }; \
> + NAME##_type out; \
> + NAME (&out, &in); \
> + if (out[0] != R || out[1] != R) \
> + __builtin_abort (); \
> + }
> +
> +#define DEF4(NAME, TYPE, X, C, R) \
> + typedef TYPE NAME##_type __attribute__ ((vector_size (4))); \
> + static void __attribute__ ((noipa)) \
> + NAME (NAME##_type *out, const NAME##_type *in) \
> + { \
> + *out = *in - (NAME##_type) { C, C, C, C }; \
> + } \
> + static void \
> + check_##NAME (void) \
> + { \
> + NAME##_type in = { X, X, X, X }; \
> + NAME##_type out; \
> + NAME (&out, &in); \
> + if (out[0] != R || out[1] != R || out[2] != R || out[3] != R) \
> + __builtin_abort (); \
> + }
> +
> +DEF4 (v4qq, short _Fract, 0.5hr, 0.25hr, 0.25hr)
> +DEF4 (v4uqq, unsigned short _Fract, 0.5uhr, 0.25uhr, 0.25uhr)
> +DEF4 (sat_v4qq, _Sat short _Fract, SFRACT_MIN, SFRACT_MIN, 0.0hr)
> +DEF4 (sat_v4uqq, _Sat unsigned short _Fract, 0.5uhr, 0.5uhr, 0.0uhr)
> +
> +DEF2 (v2hq, _Fract, 0.5r, 0.25r, 0.25r)
> +DEF2 (v2uhq, unsigned _Fract, 0.5ur, 0.25ur, 0.25ur)
> +DEF2 (sat_v2hq, _Sat _Fract, FRACT_MIN, FRACT_MIN, 0.0r)
> +DEF2 (sat_v2uhq, _Sat unsigned _Fract, 0.5ur, 0.5ur, 0.0ur)
> +
> +DEF2 (v2ha, short _Accum, 0.5hk, 0.25hk, 0.25hk)
> +DEF2 (v2uha, unsigned short _Accum, 0.5uhk, 0.25uhk, 0.25uhk)
> +DEF2 (sat_v2ha, _Sat short _Accum, SACCUM_MIN, SACCUM_MIN, 0.0hk)
> +DEF2 (sat_v2uha, _Sat unsigned short _Accum, 0.5uhk, 0.5uhk, 0.0uhk)
> +
> +int
> +main (void)
> +{
> + check_v4qq ();
> + check_v4uqq ();
> + check_sat_v4qq ();
> + check_sat_v4uqq ();
> + check_v2hq ();
> + check_v2uhq ();
> + check_sat_v2hq ();
> + check_sat_v2uhq ();
> + check_v2ha ();
> + check_v2uha ();
> + check_sat_v2ha ();
> + check_sat_v2uha ();
> + return 0;
> +}
> --
> 2.50.1 (Apple Git-155)
>
