diff --git a/library/stdarch/crates/core_arch/src/aarch64/neon/generated.rs b/library/stdarch/crates/core_arch/src/aarch64/neon/generated.rs
index d417838b543f..4fed40aa0366 100644
--- a/library/stdarch/crates/core_arch/src/aarch64/neon/generated.rs
+++ b/library/stdarch/crates/core_arch/src/aarch64/neon/generated.rs
@@ -9,6 +9,32 @@ use super::*;
 #[cfg(test)]
 use stdarch_test::assert_instr;
 
+/// Absolute difference between the arguments of Floating
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(test, assert_instr(fabd))]
+pub unsafe fn vabd_f64(a: float64x1_t, b: float64x1_t) -> float64x1_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fabd.v1f64")]
+        fn vabd_f64_(a: float64x1_t, a: float64x1_t) -> float64x1_t;
+    }
+    vabd_f64_(a, b)
+}
+
+/// Absolute difference between the arguments of Floating
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(test, assert_instr(fabd))]
+pub unsafe fn vabdq_f64(a: float64x2_t, b: float64x2_t) -> float64x2_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fabd.v2f64")]
+        fn vabdq_f64_(a: float64x2_t, a: float64x2_t) -> float64x2_t;
+    }
+    vabdq_f64_(a, b)
+}
+
 /// Compare bitwise Equal (vector)
 #[inline]
 #[target_feature(enable = "neon")]
@@ -570,6 +596,24 @@ mod test {
     use std::mem::transmute;
     use stdarch_test::simd_test;
 
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_f64() {
+        let a: f64 = 1.0;
+        let b: f64 = 9.0;
+        let e: f64 = 8.0;
+        let r: f64 = transmute(vabd_f64(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_f64() {
+        let a: f64x2 = f64x2::new(1.0, 2.0);
+        let b: f64x2 = f64x2::new(9.0, 3.0);
+        let e: f64x2 = f64x2::new(8.0, 1.0);
+        let r: f64x2 = transmute(vabdq_f64(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
     #[simd_test(enable = "neon")]
     unsafe fn test_vceq_u64() {
         let a: u64x1 = u64x1::new(0);
diff --git a/library/stdarch/crates/core_arch/src/arm/neon/generated.rs b/library/stdarch/crates/core_arch/src/arm/neon/generated.rs
index 5373e343d0c8..36e8dc1f4e05 100644
--- a/library/stdarch/crates/core_arch/src/arm/neon/generated.rs
+++ b/library/stdarch/crates/core_arch/src/arm/neon/generated.rs
@@ -489,6 +489,230 @@ pub unsafe fn veorq_u64(a: uint64x2_t, b: uint64x2_t) -> uint64x2_t {
     simd_xor(a, b)
 }
 
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabd_s8(a: int8x8_t, b: int8x8_t) -> int8x8_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v8i8")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v8i8")]
+        fn vabd_s8_(a: int8x8_t, b: int8x8_t) -> int8x8_t;
+    }
+vabd_s8_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabdq_s8(a: int8x16_t, b: int8x16_t) -> int8x16_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v16i8")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v16i8")]
+        fn vabdq_s8_(a: int8x16_t, b: int8x16_t) -> int8x16_t;
+    }
+vabdq_s8_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabd_s16(a: int16x4_t, b: int16x4_t) -> int16x4_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v4i16")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v4i16")]
+        fn vabd_s16_(a: int16x4_t, b: int16x4_t) -> int16x4_t;
+    }
+vabd_s16_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabdq_s16(a: int16x8_t, b: int16x8_t) -> int16x8_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v8i16")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v8i16")]
+        fn vabdq_s16_(a: int16x8_t, b: int16x8_t) -> int16x8_t;
+    }
+vabdq_s16_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabd_s32(a: int32x2_t, b: int32x2_t) -> int32x2_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v2i32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v2i32")]
+        fn vabd_s32_(a: int32x2_t, b: int32x2_t) -> int32x2_t;
+    }
+vabd_s32_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.s32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(sabd))]
+pub unsafe fn vabdq_s32(a: int32x4_t, b: int32x4_t) -> int32x4_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v4i32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.sabd.v4i32")]
+        fn vabdq_s32_(a: int32x4_t, b: int32x4_t) -> int32x4_t;
+    }
+vabdq_s32_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabd_u8(a: uint8x8_t, b: uint8x8_t) -> uint8x8_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v8i8")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v8i8")]
+        fn vabd_u8_(a: uint8x8_t, b: uint8x8_t) -> uint8x8_t;
+    }
+vabd_u8_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabdq_u8(a: uint8x16_t, b: uint8x16_t) -> uint8x16_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v16i8")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v16i8")]
+        fn vabdq_u8_(a: uint8x16_t, b: uint8x16_t) -> uint8x16_t;
+    }
+vabdq_u8_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabd_u16(a: uint16x4_t, b: uint16x4_t) -> uint16x4_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v4i16")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v4i16")]
+        fn vabd_u16_(a: uint16x4_t, b: uint16x4_t) -> uint16x4_t;
+    }
+vabd_u16_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabdq_u16(a: uint16x8_t, b: uint16x8_t) -> uint16x8_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v8i16")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v8i16")]
+        fn vabdq_u16_(a: uint16x8_t, b: uint16x8_t) -> uint16x8_t;
+    }
+vabdq_u16_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabd_u32(a: uint32x2_t, b: uint32x2_t) -> uint32x2_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v2i32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v2i32")]
+        fn vabd_u32_(a: uint32x2_t, b: uint32x2_t) -> uint32x2_t;
+    }
+vabd_u32_(a, b)
+}
+
+/// Absolute difference between the arguments
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.u32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(uabd))]
+pub unsafe fn vabdq_u32(a: uint32x4_t, b: uint32x4_t) -> uint32x4_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabdu.v4i32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.uabd.v4i32")]
+        fn vabdq_u32_(a: uint32x4_t, b: uint32x4_t) -> uint32x4_t;
+    }
+vabdq_u32_(a, b)
+}
+
+/// Absolute difference between the arguments of Floating
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.f32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(fabd))]
+pub unsafe fn vabd_f32(a: float32x2_t, b: float32x2_t) -> float32x2_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v2f32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fabd.v2f32")]
+        fn vabd_f32_(a: float32x2_t, b: float32x2_t) -> float32x2_t;
+    }
+vabd_f32_(a, b)
+}
+
+/// Absolute difference between the arguments of Floating
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vabd.f32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr(fabd))]
+pub unsafe fn vabdq_f32(a: float32x4_t, b: float32x4_t) -> float32x4_t {
+    #[allow(improper_ctypes)]
+    extern "C" {
+        #[cfg_attr(target_arch = "arm", link_name = "llvm.arm.neon.vabds.v4f32")]
+        #[cfg_attr(target_arch = "aarch64", link_name = "llvm.aarch64.neon.fabd.v4f32")]
+        fn vabdq_f32_(a: float32x4_t, b: float32x4_t) -> float32x4_t;
+    }
+vabdq_f32_(a, b)
+}
+
 /// Compare bitwise Equal (vector)
 #[inline]
 #[target_feature(enable = "neon")]
@@ -3473,6 +3697,132 @@ mod test {
         assert_eq!(r, e);
     }
 
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_s8() {
+        let a: i8x8 = i8x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b: i8x8 = i8x8::new(16, 15, 14, 13, 12, 11, 10, 9);
+        let e: i8x8 = i8x8::new(15, 13, 11, 9, 7, 5, 3, 1);
+        let r: i8x8 = transmute(vabd_s8(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_s8() {
+        let a: i8x16 = i8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
+        let b: i8x16 = i8x16::new(16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1);
+        let e: i8x16 = i8x16::new(15, 13, 11, 9, 7, 5, 3, 1, 1, 3, 5, 7, 9, 11, 13, 15);
+        let r: i8x16 = transmute(vabdq_s8(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_s16() {
+        let a: i16x4 = i16x4::new(1, 2, 3, 4);
+        let b: i16x4 = i16x4::new(16, 15, 14, 13);
+        let e: i16x4 = i16x4::new(15, 13, 11, 9);
+        let r: i16x4 = transmute(vabd_s16(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_s16() {
+        let a: i16x8 = i16x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b: i16x8 = i16x8::new(16, 15, 14, 13, 12, 11, 10, 9);
+        let e: i16x8 = i16x8::new(15, 13, 11, 9, 7, 5, 3, 1);
+        let r: i16x8 = transmute(vabdq_s16(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_s32() {
+        let a: i32x2 = i32x2::new(1, 2);
+        let b: i32x2 = i32x2::new(16, 15);
+        let e: i32x2 = i32x2::new(15, 13);
+        let r: i32x2 = transmute(vabd_s32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_s32() {
+        let a: i32x4 = i32x4::new(1, 2, 3, 4);
+        let b: i32x4 = i32x4::new(16, 15, 14, 13);
+        let e: i32x4 = i32x4::new(15, 13, 11, 9);
+        let r: i32x4 = transmute(vabdq_s32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_u8() {
+        let a: u8x8 = u8x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b: u8x8 = u8x8::new(16, 15, 14, 13, 12, 11, 10, 9);
+        let e: u8x8 = u8x8::new(15, 13, 11, 9, 7, 5, 3, 1);
+        let r: u8x8 = transmute(vabd_u8(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_u8() {
+        let a: u8x16 = u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16);
+        let b: u8x16 = u8x16::new(16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1);
+        let e: u8x16 = u8x16::new(15, 13, 11, 9, 7, 5, 3, 1, 1, 3, 5, 7, 9, 11, 13, 15);
+        let r: u8x16 = transmute(vabdq_u8(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_u16() {
+        let a: u16x4 = u16x4::new(1, 2, 3, 4);
+        let b: u16x4 = u16x4::new(16, 15, 14, 13);
+        let e: u16x4 = u16x4::new(15, 13, 11, 9);
+        let r: u16x4 = transmute(vabd_u16(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_u16() {
+        let a: u16x8 = u16x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b: u16x8 = u16x8::new(16, 15, 14, 13, 12, 11, 10, 9);
+        let e: u16x8 = u16x8::new(15, 13, 11, 9, 7, 5, 3, 1);
+        let r: u16x8 = transmute(vabdq_u16(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_u32() {
+        let a: u32x2 = u32x2::new(1, 2);
+        let b: u32x2 = u32x2::new(16, 15);
+        let e: u32x2 = u32x2::new(15, 13);
+        let r: u32x2 = transmute(vabd_u32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_u32() {
+        let a: u32x4 = u32x4::new(1, 2, 3, 4);
+        let b: u32x4 = u32x4::new(16, 15, 14, 13);
+        let e: u32x4 = u32x4::new(15, 13, 11, 9);
+        let r: u32x4 = transmute(vabdq_u32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabd_f32() {
+        let a: f32x2 = f32x2::new(1.0, 2.0);
+        let b: f32x2 = f32x2::new(9.0, 3.0);
+        let e: f32x2 = f32x2::new(8.0, 1.0);
+        let r: f32x2 = transmute(vabd_f32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabdq_f32() {
+        let a: f32x4 = f32x4::new(1.0, 2.0, 5.0, -4.0);
+        let b: f32x4 = f32x4::new(9.0, 3.0, 2.0, 8.0);
+        let e: f32x4 = f32x4::new(8.0, 1.0, 3.0, 12.0);
+        let r: f32x4 = transmute(vabdq_f32(transmute(a), transmute(b)));
+        assert_eq!(r, e);
+    }
+
     #[simd_test(enable = "neon")]
     unsafe fn test_vceq_u8() {
         let a: u8x8 = u8x8::new(0, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07);
diff --git a/library/stdarch/crates/core_arch/src/arm/neon/mod.rs b/library/stdarch/crates/core_arch/src/arm/neon/mod.rs
index 9751bdb5c49b..8b469fdd924e 100644
--- a/library/stdarch/crates/core_arch/src/arm/neon/mod.rs
+++ b/library/stdarch/crates/core_arch/src/arm/neon/mod.rs
@@ -1093,6 +1093,105 @@ pub unsafe fn vld1q_dup_f32(ptr: *const f32) -> float32x4_t {
     simd_shuffle4(x, x, [0, 0, 0, 0])
 }
 
+// signed absolute difference and accumulate (64-bit)
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vaba_s8(a: int8x8_t, b: int8x8_t, c: int8x8_t) -> int8x8_t {
+    simd_add(a, vabd_s8(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vaba_s16(a: int16x4_t, b: int16x4_t, c: int16x4_t) -> int16x4_t {
+    simd_add(a, vabd_s16(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vaba_s32(a: int32x2_t, b: int32x2_t, c: int32x2_t) -> int32x2_t {
+    simd_add(a, vabd_s32(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vaba_u8(a: uint8x8_t, b: uint8x8_t, c: uint8x8_t) -> uint8x8_t {
+    simd_add(a, vabd_u8(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vaba_u16(a: uint16x4_t, b: uint16x4_t, c: uint16x4_t) -> uint16x4_t {
+    simd_add(a, vabd_u16(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vaba_u32(a: uint32x2_t, b: uint32x2_t, c: uint32x2_t) -> uint32x2_t {
+    simd_add(a, vabd_u32(b, c))
+}
+// signed absolute difference and accumulate (128-bit)
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vabaq_s8(a: int8x16_t, b: int8x16_t, c: int8x16_t) -> int8x16_t {
+    simd_add(a, vabdq_s8(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vabaq_s16(a: int16x8_t, b: int16x8_t, c: int16x8_t) -> int16x8_t {
+    simd_add(a, vabdq_s16(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.s32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("saba"))]
+pub unsafe fn vabaq_s32(a: int32x4_t, b: int32x4_t, c: int32x4_t) -> int32x4_t {
+    simd_add(a, vabdq_s32(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u8"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vabaq_u8(a: uint8x16_t, b: uint8x16_t, c: uint8x16_t) -> uint8x16_t {
+    simd_add(a, vabdq_u8(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u16"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vabaq_u16(a: uint16x8_t, b: uint16x8_t, c: uint16x8_t) -> uint16x8_t {
+    simd_add(a, vabdq_u16(b, c))
+}
+#[inline]
+#[target_feature(enable = "neon")]
+#[cfg_attr(target_arch = "arm", target_feature(enable = "v7"))]
+#[cfg_attr(all(test, target_arch = "arm"), assert_instr("vaba.u32"))]
+#[cfg_attr(all(test, target_arch = "aarch64"), assert_instr("uaba"))]
+pub unsafe fn vabaq_u32(a: uint32x4_t, b: uint32x4_t, c: uint32x4_t) -> uint32x4_t {
+    simd_add(a, vabdq_u32(b, c))
+}
+
 /// Absolute value (wrapping).
 #[inline]
 #[target_feature(enable = "neon")]
@@ -8564,6 +8663,120 @@ mod tests {
         let e = i32x4::new(i32::MIN, i32::MAX, 0, 1);
         assert_eq!(r, e);
     }
+
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_s8() {
+        let a = i8x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b = i8x8::new(1, 1, 1, 1, 1, 1, 1, 1);
+        let c = i8x8::new(10, 9, 8, 7, 6, 5, 4, 3);
+        let r: i8x8 = transmute(vaba_s8(transmute(a), transmute(b), transmute(c)));
+        let e = i8x8::new(10, 10, 10, 10, 10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_s16() {
+        let a = i16x4::new(1, 2, 3, 4);
+        let b = i16x4::new(1, 1, 1, 1);
+        let c = i16x4::new(10, 9, 8, 7);
+        let r: i16x4 = transmute(vaba_s16(transmute(a), transmute(b), transmute(c)));
+        let e = i16x4::new(10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_s32() {
+        let a = i32x2::new(1, 2);
+        let b = i32x2::new(1, 1);
+        let c = i32x2::new(10, 9);
+        let r: i32x2 = transmute(vaba_s32(transmute(a), transmute(b), transmute(c)));
+        let e = i32x2::new(10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_u8() {
+        let a = u8x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b = u8x8::new(1, 1, 1, 1, 1, 1, 1, 1);
+        let c = u8x8::new(10, 9, 8, 7, 6, 5, 4, 3);
+        let r: u8x8 = transmute(vaba_u8(transmute(a), transmute(b), transmute(c)));
+        let e = u8x8::new(10, 10, 10, 10, 10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_u16() {
+        let a = u16x4::new(1, 2, 3, 4);
+        let b = u16x4::new(1, 1, 1, 1);
+        let c = u16x4::new(10, 9, 8, 7);
+        let r: u16x4 = transmute(vaba_u16(transmute(a), transmute(b), transmute(c)));
+        let e = u16x4::new(10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vaba_u32() {
+        let a = u32x2::new(1, 2);
+        let b = u32x2::new(1, 1);
+        let c = u32x2::new(10, 9);
+        let r: u32x2 = transmute(vaba_u32(transmute(a), transmute(b), transmute(c)));
+        let e = u32x2::new(10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_s8() {
+        let a = i8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 9, 8, 7, 6, 5, 4, 3, 2);
+        let b = i8x16::new(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1);
+        let c = i8x16::new(10, 9, 8, 7, 6, 5, 4, 3, 12, 13, 14, 15, 16, 17, 18, 19);
+        let r: i8x16 = transmute(vabaq_s8(transmute(a), transmute(b), transmute(c)));
+        let e = i8x16::new(
+            10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20,
+        );
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_s16() {
+        let a = i16x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b = i16x8::new(1, 1, 1, 1, 1, 1, 1, 1);
+        let c = i16x8::new(10, 9, 8, 7, 6, 5, 4, 3);
+        let r: i16x8 = transmute(vabaq_s16(transmute(a), transmute(b), transmute(c)));
+        let e = i16x8::new(10, 10, 10, 10, 10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_s32() {
+        let a = i32x4::new(1, 2, 3, 4);
+        let b = i32x4::new(1, 1, 1, 1);
+        let c = i32x4::new(10, 9, 8, 7);
+        let r: i32x4 = transmute(vabaq_s32(transmute(a), transmute(b), transmute(c)));
+        let e = i32x4::new(10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_u8() {
+        let a = u8x16::new(1, 2, 3, 4, 5, 6, 7, 8, 9, 8, 7, 6, 5, 4, 3, 2);
+        let b = u8x16::new(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1);
+        let c = u8x16::new(10, 9, 8, 7, 6, 5, 4, 3, 12, 13, 14, 15, 16, 17, 18, 19);
+        let r: u8x16 = transmute(vabaq_u8(transmute(a), transmute(b), transmute(c)));
+        let e = u8x16::new(
+            10, 10, 10, 10, 10, 10, 10, 10, 20, 20, 20, 20, 20, 20, 20, 20,
+        );
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_u16() {
+        let a = u16x8::new(1, 2, 3, 4, 5, 6, 7, 8);
+        let b = u16x8::new(1, 1, 1, 1, 1, 1, 1, 1);
+        let c = u16x8::new(10, 9, 8, 7, 6, 5, 4, 3);
+        let r: u16x8 = transmute(vabaq_u16(transmute(a), transmute(b), transmute(c)));
+        let e = u16x8::new(10, 10, 10, 10, 10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+    #[simd_test(enable = "neon")]
+    unsafe fn test_vabaq_u32() {
+        let a = u32x4::new(1, 2, 3, 4);
+        let b = u32x4::new(1, 1, 1, 1);
+        let c = u32x4::new(10, 9, 8, 7);
+        let r: u32x4 = transmute(vabaq_u32(transmute(a), transmute(b), transmute(c)));
+        let e = u32x4::new(10, 10, 10, 10);
+        assert_eq!(r, e);
+    }
+
     #[simd_test(enable = "neon")]
     unsafe fn test_vpadd_s16() {
         let a = i16x4::new(1, 2, 3, 4);
diff --git a/library/stdarch/crates/stdarch-gen/neon.spec b/library/stdarch/crates/stdarch-gen/neon.spec
index 28d25e314a0f..1d3749abfe81 100644
--- a/library/stdarch/crates/stdarch-gen/neon.spec
+++ b/library/stdarch/crates/stdarch-gen/neon.spec
@@ -102,6 +102,44 @@ b = 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00
 validate 0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0A, 0x0B, 0x0C, 0x0D, 0x0E, 0x0F
 generate int*_t, uint*_t, int64x*_t, uint64x*_t
 
+////////////////////
+// Absolute difference between the arguments
+////////////////////
+
+/// Absolute difference between the arguments
+name = vabd
+a = 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16
+b = 16, 15, 14, 13, 12, 11, 10, 9, 8, 7, 6, 5, 4, 3, 2, 1
+validate 15, 13, 11, 9, 7, 5, 3, 1, 1, 3, 5, 7, 9, 11, 13, 15
+
+arm = vabd.s
+aarch64 = sabd
+link-arm = vabds._EXT_
+link-aarch64 = sabd._EXT_
+generate int*_t
+
+arm = vabd.s
+aarch64 = uabd
+link-arm = vabdu._EXT_
+link-aarch64 = uabd._EXT_
+generate uint*_t
+
+/// Absolute difference between the arguments of Floating
+name = vabd
+a = 1.0, 2.0, 5.0, -4.0
+b = 9.0, 3.0, 2.0, 8.0
+validate 8.0, 1.0, 3.0, 12.0
+
+aarch64 = fabd
+link-aarch64 = fabd._EXT_
+generate float64x*_t
+
+arm = vabd.s
+aarch64 = fabd
+link-arm = vabds._EXT_
+link-aarch64 = fabd._EXT_
+generate float*_t
+
 ////////////////////
 // equality
 ////////////////////