TEXT ·_dpavx_int32clang(SB),4, $0-32

        MOVQ a+0(FP), DI
        MOVQ b+8(FP), SI
        MOVQ gN+16(FP), DX
        MOVQ res+24(FP), CX

        MOVQ (DX),R8         // mov    r8d, dword [rdx]
        CMPQ R8,$0			 // test    r8d, r8d
        JLE  LBB0_8 // LBB0_1
        CMPQ R8,$15         // cmp    r8d, 15
        JA   LBB0_4
        XORQ R9, R9		// xor    r9d, r9d
        XORQ AX, AX		// xor    eax, eax
        JMP  LBB0_8 //LBB0_7

LBB0_1:
        XORQ AX, AX		 // xor    eax, eax
        JMP  LBB0_8

LBB0_4:
    	MOVQ R8,R9	 					// mov    r9d, r8d
        ANDQ $-16,R9        			// and    r9d, -16      
        VPXOR X0,X0,X0   			 // vpxor    xmm0, xmm0, xmm0   
        XORQ AX, AX             		// xor    eax, eax
        VPXOR X1, X1, X1    		// vpxor    xmm1, xmm1, xmm1
        VPXOR X2, X2, X2         // vpxor    xmm2, xmm2, xmm2
        VPXOR X3, X3, X3         // vpxor    xmm3, xmm3, xmm3
        
LBB0_5:
        VMOVDQU (SI)(AX*4), X4             // vmovdqu    xmm4, oword [rsi + 4*rax]
        VMOVDQU 16(SI)(AX*4), X5              // vmovdqu    xmm5, oword [rsi + 4*rax + 16]
        VMOVDQU 32(SI)(AX*4), X6              // vmovdqu    xmm6, oword [rsi + 4*rax + 32]
        VMOVDQU 48(SI)(AX*4), X7              // vmovdqu    xmm7, oword [rsi + 4*rax + 48]
        
        VPMULLD (DI)(AX*4), X4, X4  		// vpmulld    xmm4, xmm4, oword [rdi + 4*rax]
        VPADDD    X0, X4, X0                // vpaddd    xmm0, xmm4, xmm0
        
        VPMULLD  (DI)(AX*4),X5, X4 			// vpmulld    xmm4, xmm5, oword [rdi + 4*rax + 16]
        VPADDD    X1, X4, X1                           // vpaddd    xmm1, xmm4, xmm1
        VPMULLD (DI)(AX*4), X6, X4 			// vpmulld    xmm4, xmm6, oword [rdi + 4*rax + 32]
        VPMULLD (DI)(AX*4), X7, X5			 // vpmulld    xmm5, xmm7, oword [rdi + 4*rax + 48]
        VPADDD    X2, X4, X2   // vpaddd    xmm2, xmm4, xmm2
        VPADDD    X3, X5, X3   // vpaddd    xmm3, xmm5, xmm3
        ADDQ    $16,AX               // add    rax, 16
        CMPQ AX,R9                   // cmp    r9, rax
        JNE  LBB0_5
        VPADDD    X0, X1, X0    // vpaddd    xmm0, xmm1, xmm0
        VPADDD    X0, X2, X0    // vpaddd    xmm0, xmm2, xmm0
        VPADDD    X0, X3, X0    // vpaddd    xmm0, xmm3, xmm0
        
        VPSHUFD    $78 , X0,X1    // vpshufd    xmm1, xmm0, 78
        VPADDD    X1, X0, X0    // vpaddd    xmm0, xmm0, xmm1
        VPSHUFD    $229, X0, X1   // vpshufd    xmm1, xmm0, 229
        VPADDD    X1, X0, X0    // vpaddd    xmm0, xmm0, xmm1
        
        VMOVD X0,AX                // vmovd    eax, xmm0
        CMPQ R8,R9                   // cmp    r9, r8
        JE   LBB0_8

LBB0_7:
        MOVL (SI)(R9*4),DX   // mov    edx, dword [rsi + 4*r9]
        IMULL (DI)(R9*4),DX  // imul    edx, dword [rdi + 4*r9]
        ADDQ DX,AX           // add    eax, edx
        ADDQ $1,R9           // add    r9, 1
        CMPQ R9,R8  	     // cmp    r8, r9
        JNE  LBB0_7

LBB0_8:
        MOVQ (CX),AX		 // mov    dword [rcx], eax
        VZEROUPPER // JMB
        RET