From a8d82a63c78db131beab4ac0004062bc686e33b3 Mon Sep 17 00:00:00 2001
From: serge-sans-paille <sergesanspaille@free.fr>
Date: Mon, 22 Sep 2025 00:00:00 +0200
Subject: [PATCH] Support immediate version of btitwise shift on arm

---
 include/xsimd/arch/xsimd_neon.hpp | 98 +++++++++++++++++++++++++++++++
 1 file changed, 98 insertions(+)

diff --git a/include/xsimd/arch/xsimd_neon.hpp b/include/xsimd/arch/xsimd_neon.hpp
index 5bd7d9206..5c8900d8f 100644
--- a/include/xsimd/arch/xsimd_neon.hpp
+++ b/include/xsimd/arch/xsimd_neon.hpp
@@ -2314,6 +2314,55 @@ namespace xsimd
             return vshlq_s64(lhs, rhs);
         }
 
+        // immediate variant
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 1> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_u8(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 1> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_s8(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 2> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_u16(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 2> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_s16(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 4> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_u32(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 4> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_s32(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 8> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_u64(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 8> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_lshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshlq_n_s64(x, shift);
+        }
+
         /******************
          * bitwise_rshift *
          ******************/
@@ -2489,6 +2538,55 @@ namespace xsimd
             return vshlq_s32(lhs, vnegq_s32(rhs));
         }
 
+        // immediate variant
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 1> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_u8(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 1> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_s8(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 2> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_u16(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 2> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_s16(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 4> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_u32(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 4> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_s32(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_unsigned_t<T, 8> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_u64(x, shift);
+        }
+
+        template <size_t shift, class A, class T, detail::enable_sized_signed_t<T, 8> = 0>
+        XSIMD_INLINE batch<T, A> bitwise_rshift(batch<T, A> const& x, requires_arch<neon>) noexcept
+        {
+            return vshrq_n_s64(x, shift);
+        }
+
         // first
         template <class A>
         XSIMD_INLINE float first(batch<float, A> const& self, requires_arch<neon>) noexcept