First half of G1 macro working for Round 1

block.go

@@ -37,21 +37,21 @@ func blocks(d *digest, p []uint8) {
 			j += 8
 		}
 
-		//// Round 1.
-		//v0 += m[0]
-		//v0 += v4
-		//v12 ^= v0
-		//v12 = v12<<(64-32) | v12>>32
-		//v8 += v12
-		//v4 ^= v8
-		//v4 = v4<<(64-24) | v4>>24
-		//v1 += m[2]
-		//v1 += v5
-		//v13 ^= v1
-		//v13 = v13<<(64-32) | v13>>32
-		//v9 += v13
-		//v5 ^= v9
-		//v5 = v5<<(64-24) | v5>>24
+		// Round 1.
+		v0 += m[0]
+		v0 += v4
+		v12 ^= v0
+		v12 = v12<<(64-32) | v12>>32
+		v8 += v12
+		v4 ^= v8
+		v4 = v4<<(64-24) | v4>>24
+		v1 += m[2]
+		v1 += v5
+		v13 ^= v1
+		v13 = v13<<(64-32) | v13>>32
+		v9 += v13
+		v5 ^= v9
+		v5 = v5<<(64-24) | v5>>24
 		//v2 += m[4]
 		//v2 += v6
 		//v14 ^= v2

compress_amd64.go

@@ -20,7 +20,7 @@
 package blake2b
 
 //go:noescape
-func compressSSE(p []uint8, in, iv, t, f, out []uint64)
+func compressSSE(p []uint8, in, iv  , t, f, shffle, out []uint64)
 
 func compress(d *digest, p []uint8) {
 	h0, h1, h2, h3, h4, h5, h6, h7 := d.h[0], d.h[1], d.h[2], d.h[3], d.h[4], d.h[5], d.h[6], d.h[7]
@@ -28,6 +28,11 @@ func compress(d *digest, p []uint8) {
 	in := make([]uint64, 8, 8)
 	out := make([]uint64, 8, 8)
 
+	shffle := make([]uint64, 2, 2)
+	// vector for PSHUFB instruction
+	shffle[0] = 0x0201000706050403
+	shffle[1] = 0x0a09080f0e0d0c0b
+
 	for len(p) >= BlockSize {
 		// Increment counter.
 		d.t[0] += BlockSize
@@ -37,7 +42,7 @@ func compress(d *digest, p []uint8) {
 
 		in[0], in[1], in[2], in[3], in[4], in[5], in[6], in[7] = h0, h1, h2, h3, h4, h5, h6, h7
 
-		compressSSE(p, in, iv[:], d.t[:], d.f[:], out)
+		compressSSE(p, in, iv[:], d.t[:], d.f[:], shffle, out)
 
 		h0, h1, h2, h3, h4, h5, h6, h7 = out[0], out[1], out[2], out[3], out[4], out[5], out[6], out[7]
 

compress_amd64.s

@@ -16,7 +16,7 @@
 // limitations under the License.
 //
 
-// func compressSSE(compressSSE(p []uint8, in, iv, t, f, out []uint64)
+// func compressSSE(compressSSE(p []uint8, in, iv, t, f, shffle, out []uint64)
 TEXT ·compressSSE(SB), 7, $0
 
     // Load digest
@@ -39,6 +39,32 @@ TEXT ·compressSSE(SB), 7, $0
     MOVOU   0(SI), X8         // X8 = f[0]+f[0]        /* row4h = _mm_xor_si128(                         LOAD( &S->f[0] )    */
     PXOR       X8, X7         // X7 = X7 ^ X8          /* row4h = _mm_xor_si128(                       ,                  ); */
 
+    ///////////////////////////////////////////////////////////////////////////
+    // R O U N D   1
+    ///////////////////////////////////////////////////////////////////////////
+
+    // LOAD_MSG_ ##r ##_1(b0, b1);
+    //   (X12 used as additional temp register)
+    MOVQ   message+0(FP), DX  // DX: &p (message)
+    MOVOU   0(DX), X12        // X12 = m[0]+m[1]
+    MOVOU  16(DX), X13        // X13 = m[2]+m[3]
+    MOVOU  32(DX), X14        // X14 = m[4]+m[5]
+    MOVOU  48(DX), X15        // X15 = m[6]+m[7]
+    BYTE $0xc4; BYTE $0x41; BYTE $0x19; BYTE $0x6c; BYTE $0xc5   // VPUNPCKLQDQ  XMM8, XMM12, XMM13  /* m[0], m[2] */
+
+    // Load shuffle value
+    MOVQ   shffle+120(FP), SI // SI: &shuffle
+    MOVOU  0(SI), X12         // X12 = 03040506 07000102 0b0c0d0e 0f08090a
+
+    // G1(row1l,row2l,row3l,row4l,row1h,row2h,row3h,row4h,b0,b1);
+    BYTE $0xc4; BYTE $0xc1; BYTE $0x79; BYTE $0xd4; BYTE $0xc0   // VPADDQ  XMM0,XMM0,XMM8   /* v0 += m[0], v1 += m[2] */
+    BYTE $0xc5; BYTE $0xf9; BYTE $0xd4; BYTE $0xc2               // VPADDQ  XMM0,XMM0,XMM2   /* v0 += v4, v1 += v5 */
+    BYTE $0xc5; BYTE $0xc9; BYTE $0xef; BYTE $0xf0               // VPXOR   XMM6,XMM6,XMM0   /* v12 ^= v0, v13 ^= v1 */
+    BYTE $0xc5; BYTE $0xf9; BYTE $0x70; BYTE $0xf6; BYTE $0xb1   // VPSHUFD XMM6,XMM6,0xb1   /* v12 = v12<<(64-32) | v12>>32, v13 = v13<<(64-32) | v13>>32 */
+    BYTE $0xc5; BYTE $0xd9; BYTE $0xd4; BYTE $0xe6               // VPADDQ  XMM4,XMM4,XMM6   /* v8 += v12, v9 += v13  */
+    BYTE $0xc5; BYTE $0xe9; BYTE $0xef; BYTE $0xd4               // VPXOR   XMM2,XMM2,XMM4   /* v4 ^= v8, v5 ^= v9 */
+    BYTE $0xc4; BYTE $0xc2; BYTE $0x69; BYTE $0x00; BYTE $0xd4   // VPSHUFB XMM2,XMM2,XMM12  /* v4 = v4<<(64-24) | v4>>24, v5 = v5<<(64-24) | v5>>24 */
+
     // Reload digest
     MOVQ   in+24(FP),  SI     // SI: &in
     MOVOU   0(SI), X12        // X12 = in[0]+in[1]      /* row1l = LOAD( &S->h[0] ); */
@@ -57,7 +83,7 @@ TEXT ·compressSSE(SB), 7, $0
     PXOR   X15, X3           // X3 = X3 ^ X15         /*  STORE( &S->h[6], _mm_xor_si128( LOAD( &S->h[6] ), row2h ) ); */
 
     // Store digest
-    MOVQ  out+120(FP), DX     // DX: &out
+    MOVQ  out+144(FP), DX     // DX: &out
     MOVOU  X0,  0(DX)         // out[0]+out[1] = X0
     MOVOU  X1, 16(DX)         // out[2]+out[3] = X1
     MOVOU  X2, 32(DX)         // out[4]+out[5] = X2