Bad compare codegen

[TheIronBorn]

Just me or is this bad codegen?

extern crate packed_simd;
use std::arch::x86_64::*;
use packed_simd::*;

pub fn le_i8x16(x: i8x16, y: i8x16) -> bool {
    x.le(y).all()
}

	.section	__TEXT,__text,regular,pure_instructions
	.globl	__ZN9temp_test8le_i8x1617h13e72ae756b29df4E
	.p2align	4, 0x90
__ZN9temp_test8le_i8x1617h13e72ae756b29df4E:
	.cfi_startproc
	pushq	%rbp
	.cfi_def_cfa_offset 16
	.cfi_offset %rbp, -16
	movq	%rsp, %rbp
	.cfi_def_cfa_register %rbp
	vmovdqa	(%rdi), %xmm0
	vpcmpgtb	(%rsi), %xmm0, %xmm0
	vpcmpeqd	%xmm1, %xmm1, %xmm1
	vpxor	%xmm1, %xmm0, %xmm0
	vptest	%xmm1, %xmm0
	setb	%al
	popq	%rbp
	retq
	.cfi_endproc

vs

pub unsafe fn int_u8x16(x: __m128i, y: __m128i) -> bool {
    let mask = _mm_cmpgt_epi8(x, y);
    1 == _mm_test_all_zeros(mask, mask)
}

	.globl	__ZN9temp_test9int_u8x1617h00120f998e226c4eE
	.p2align	4, 0x90
__ZN9temp_test9int_u8x1617h00120f998e226c4eE:
	.cfi_startproc
	pushq	%rbp
	.cfi_def_cfa_offset 16
	.cfi_offset %rbp, -16
	movq	%rsp, %rbp
	.cfi_def_cfa_register %rbp
	vmovdqa	(%rdi), %xmm0
	vpcmpgtb	(%rsi), %xmm0, %xmm0
	vptest	%xmm0, %xmm0
	sete	%al
	popq	%rbp
	retq
	.cfi_endproc

[TheIronBorn]

with target-cpu=sandybridge

[TheIronBorn]

x.gt(y).none() appears to match the intrinsics

[gnzlbg]

this bad codegen

I think we'd need to fill this an an LLVM bug. LLVM has many of these.

[gnzlbg]

This is now: https://bugs.llvm.org/show_bug.cgi?id=38522

[gnzlbg]

@TheIronBorn we should re-check this once #156 is merged. Also I've opened #157 for exploring other things that we could do here codegen wise.