diff --git a/mask_amd64.s b/mask_amd64.s
index 73ae59b4..8464440b 100644
--- a/mask_amd64.s
+++ b/mask_amd64.s
@@ -117,10 +117,10 @@ less_than_4:
 
 less_than_2:
 	TESTQ $1, CX
-	JZ    done
+	JZ    end
 	XORB  SI, (AX)
 	ROLL  $24, SI
 
-done:
+end:
 	MOVL SI, ret+24(FP)
 	RET
diff --git a/mask_arm64.s b/mask_arm64.s
index 8fd49aa9..42a1211f 100644
--- a/mask_arm64.s
+++ b/mask_arm64.s
@@ -15,7 +15,7 @@ TEXT ·maskAsm(SB), NOSPLIT, $0-28
 	CMP   $64, R1
 	BLT   less_than_64
 
-// TODO: allign memory like amd64
+// TODO: align memory like amd64
 
 loop_64:
 	VLD1   (R0), [V1.B16, V2.B16, V3.B16, V4.B16]
@@ -29,41 +29,39 @@ loop_64:
 	BGE    loop_64
 
 less_than_64:
-	// quick end
-	CBZ    R1, end
-	TBZ    $5, R1, less_than32
+	TBZ    $5, R1, less_than_32
 	VLD1   (R0), [V1.B16, V2.B16]
 	VEOR   V1.B16, V0.B16, V1.B16
 	VEOR   V2.B16, V0.B16, V2.B16
 	VST1.P [V1.B16, V2.B16], 32(R0)
 
-less_than32:
-	TBZ   $4, R1, less_than16
+less_than_32:
+	TBZ   $4, R1, less_than_16
 	LDP   (R0), (R11, R12)
 	EOR   R11, R2, R11
 	EOR   R12, R2, R12
 	STP.P (R11, R12), 16(R0)
 
-less_than16:
-	TBZ    $3, R1, less_than8
+less_than_16:
+	TBZ    $3, R1, less_than_8
 	MOVD   (R0), R11
 	EOR    R2, R11, R11
 	MOVD.P R11, 8(R0)
 
-less_than8:
-	TBZ     $2, R1, less_than4
+less_than_8:
+	TBZ     $2, R1, less_than_4
 	MOVWU   (R0), R11
 	EORW    R2, R11, R11
 	MOVWU.P R11, 4(R0)
 
-less_than4:
-	TBZ     $1, R1, less_than2
+less_than_4:
+	TBZ     $1, R1, less_than_2
 	MOVHU   (R0), R11
 	EORW    R3, R11, R11
 	MOVHU.P R11, 2(R0)
 	RORW    $16, R3
 
-less_than2:
+less_than_2:
 	TBZ     $0, R1, end
 	MOVBU   (R0), R11
 	EORW    R3, R11, R11