[Bug c++/105353] New: __builtin_shufflevector with template parameter fails to compile on GCC 12 but compiles on clang

[john_platts at hotmail dot com via Gcc-bugs]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=105353

            Bug ID: 105353
           Summary: __builtin_shufflevector with template parameter fails
                    to compile on GCC 12 but compiles on clang
           Product: gcc
           Version: 12.0
            Status: UNCONFIRMED
          Severity: normal
          Priority: P3
         Component: c++
          Assignee: unassigned at gcc dot gnu.org
          Reporter: john_platts at hotmail dot com
  Target Milestone: ---

The following code fails to compile with GCC 12 but compiles successfully on
clang (with the -std=c++17 flag):
#include <cstdint>

typedef std::uint8_t Simd128U8VectT __attribute__((__vector_size__(16)));

template<int ShuffleIndex>
static inline Simd128U8VectT ShufFunc(Simd128U8VectT vect) noexcept {
    if constexpr(unsigned(ShuffleIndex) >= 16)
        return Simd128U8VectT { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
    else if constexpr(ShuffleIndex == 0)
        return vect;
    else
        return __builtin_shufflevector(vect, vect, ShuffleIndex, ShuffleIndex +
1,
            ShuffleIndex + 2, ShuffleIndex + 3, ShuffleIndex + 4, ShuffleIndex
+ 5,
            ShuffleIndex + 6, ShuffleIndex + 7, ShuffleIndex + 8, ShuffleIndex
+ 9,
            ShuffleIndex + 10, ShuffleIndex + 11, ShuffleIndex + 12,
ShuffleIndex + 13,
            ShuffleIndex + 14, ShuffleIndex + 15);
}

auto func1(Simd128U8VectT vect) noexcept {
    return ShufFunc<5>(vect);
}

Here is the assembly output when the above code is compiled on clang 14.0.0
with the -O2 -std=c++17 flags:
func1(unsigned char __vector(16)):                         # @func1(unsigned
char __vector(16))
        movdqa  xmm1, xmm0
        psrldq  xmm1, 5                         # xmm1 =
xmm1[5,6,7,8,9,10,11,12,13,14,15],zero,zero,zero,zero,zero
        pslldq  xmm0, 11                        # xmm0 =
zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,zero,xmm0[0,1,2,3,4]
        por     xmm0, xmm1
        ret

The below code compiles successfully with GCC 11 and GCC 12 (which uses
__builtin_shuffle instead of __builtin_shufflevector):
#include <cstdint>

typedef std::uint8_t Simd128U8VectT __attribute__((__vector_size__(16)));

template<int ShuffleIndex>
static inline Simd128U8VectT ShufFunc(Simd128U8VectT vect) noexcept {
    if constexpr(unsigned(ShuffleIndex) >= 16)
        return Simd128U8VectT { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
};
    else if constexpr(ShuffleIndex == 0)
        return vect;
    else
        return __builtin_shuffle(vect, vect, (Simd128U8VectT){
            ShuffleIndex, ShuffleIndex + 1,
            ShuffleIndex + 2, ShuffleIndex + 3, ShuffleIndex + 4, ShuffleIndex
+ 5,
            ShuffleIndex + 6, ShuffleIndex + 7, ShuffleIndex + 8, ShuffleIndex
+ 9,
            ShuffleIndex + 10, ShuffleIndex + 11, ShuffleIndex + 12,
ShuffleIndex + 13,
            ShuffleIndex + 14, ShuffleIndex + 15 });
}

auto func1(Simd128U8VectT vect) noexcept {
    return ShufFunc<5>(vect);
}