From 62a404d90ebd1b1d7a2bbfe65694326b76c55731 Mon Sep 17 00:00:00 2001
From: lukechampine <luke.champine@gmail.com>
Date: Fri, 6 Jul 2018 17:10:20 -0400
Subject: [PATCH] fix SSE segfault

---
 xor_amd64.s | 71 ++++++++++++++---------------------------------------
 xor_test.go | 16 ++++++++++--
 2 files changed, 33 insertions(+), 54 deletions(-)

diff --git a/xor_amd64.s b/xor_amd64.s
index c137b67..24a2e14 100644
--- a/xor_amd64.s
+++ b/xor_amd64.s
@@ -8,68 +8,34 @@
 #define N R12
 
 // func xorBytesSSE(dst, a, b []byte, n int)
-TEXT ·xorBytesSSE(SB), NOSPLIT ,$0
+TEXT ·xorBytesSSE(SB), NOSPLIT, $0
 	MOVQ dst_data+0(FP), Dst
 	MOVQ a_data+24(FP), A
 	MOVQ b_data+48(FP), B
 	MOVQ n+72(FP), N
 
-XOR_LOOP_128_SSE:
-	CMPQ   N, $128
-	JB     XOR_LOOP_64_SSE
-
-	MOVOU     (A), X0
-	MOVOU   16(A), X1
-	MOVOU   32(A), X2
-	MOVOU   48(A), X3
-	MOVOU   64(A), X4
-	MOVOU   80(A), X5
-	MOVOU   96(A), X6
-	MOVOU  112(A), X7
-
-	PXOR      (B), X0
-	PXOR    16(B), X1
-	PXOR    32(B), X2
-	PXOR    48(B), X3
-	PXOR    64(B), X4
-	PXOR    80(B), X5
-	PXOR    96(B), X6
-	PXOR   112(B), X7
-
-	MOVOU  X0,    (Dst)
-	MOVOU  X1,  16(Dst)
-	MOVOU  X2,  32(Dst)
-	MOVOU  X3,  48(Dst)
-	MOVOU  X4,  64(Dst)
-	MOVOU  X5,  80(Dst)
-	MOVOU  X6,  96(Dst)
-	MOVOU  X7, 112(Dst)
-
-	ADDQ   $128, A
-	ADDQ   $128, B
-	ADDQ   $128, Dst
-	SUBQ   $128, N
-	JNZ    XOR_LOOP_128_SSE
-	RET
-
 XOR_LOOP_64_SSE:
 	CMPQ   N, $64
 	JB     XOR_LOOP_16_SSE
 
-	MOVOU    (A), X0
-	MOVOU  16(A), X1
-	MOVOU  32(A), X2
-	MOVOU  48(A), X3
+	MOVOU  0*16(A), X0
+	MOVOU  1*16(A), X1
+	MOVOU  2*16(A), X2
+	MOVOU  3*16(A), X3
+	MOVOU  0*16(B), X4
+	MOVOU  1*16(B), X5
+	MOVOU  2*16(B), X6
+	MOVOU  3*16(B), X7
 
-	PXOR     (B), X0
-	PXOR   16(B), X1
-	PXOR   32(B), X2
-	PXOR   48(B), X3
+	PXOR   X4, X0
+	PXOR   X5, X1
+	PXOR   X6, X2
+	PXOR   X7, X3
 
-	MOVOU  X0,   (Dst)
-	MOVOU  X1, 16(Dst)
-	MOVOU  X2, 32(Dst)
-	MOVOU  X3, 48(Dst)
+	MOVOU  X0, 0*16(Dst)
+	MOVOU  X1, 1*16(Dst)
+	MOVOU  X2, 2*16(Dst)
+	MOVOU  X3, 3*16(Dst)
 
 	ADDQ   $64, A
 	ADDQ   $64, B
@@ -82,7 +48,8 @@ XOR_LOOP_16_SSE:
 	CMPQ   N, $16
 	JB     XOR_LOOP_FINAL_SSE
 	MOVOU  (A), X0
-	PXOR   (B), X0
+	MOVOU  (B), X1
+	PXOR   X1, X0
 	MOVOU  X0, (Dst)
 	ADDQ   $16, A
 	ADDQ   $16, B
diff --git a/xor_test.go b/xor_test.go
index 6fea00b..6cf2c0f 100644
--- a/xor_test.go
+++ b/xor_test.go
@@ -56,8 +56,10 @@ func refBlock(dst, a, b []byte) {
 
 func TestBytes(t *testing.T) {
 	err := quick.Check(func(a, b []byte) bool {
-		// double size to increase chances of reaching 64 bytes
+		// quadruple size to increase coverage
 		a = append(a, a...)
+		a = append(a, a...)
+		b = append(b, b...)
 		b = append(b, b...)
 		if len(a) < 8 {
 			return true
@@ -72,7 +74,17 @@ func TestBytes(t *testing.T) {
 		return bytes.Equal(dst1, dst2)
 	}, &quick.Config{MaxCount: 10000})
 	if err != nil {
-		t.Fatal(err)
+		t.Error(err)
+	}
+
+	// test a large slice
+	a := make([]byte, 512-1)
+	for i := range a {
+		a[i] = byte(i)
+	}
+	Bytes(a, a, a)
+	if !bytes.Equal(a, make([]byte, len(a))) {
+		t.Error("bad xor of large slice")
 	}
 }