diff --git a/userspace/units/interface/static_analysis/static_analysis.c b/userspace/units/interface/static_analysis/static_analysis.c
index 1007707f4..003ed1b87 100644
--- a/userspace/units/interface/static_analysis/static_analysis.c
+++ b/userspace/units/interface/static_analysis/static_analysis.c
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020, NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (c) 2020-2021, NVIDIA CORPORATION.  All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -29,132 +29,335 @@
 
 #include "static_analysis.h"
 
+/**
+ * unsigned addition tests
+ *
+ * parameters:
+ *   type_name: type (u8, u32 etc.)
+ *   type_max : Maximum value of the type.
+ *   tmp_operand: random value in the set (1, Maximum value)
+ *
+ * Boundary values: (0, 1, max-1, max)
+ *
+ * Valid tests: Addition result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Addition result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_ADD_TESTS(type_name, type_max, tmp_operand) do {\
+	unit_assert(nvgpu_safe_add_##type_name(type_max, 0) == type_max, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(type_max - 1, 1) == type_max, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(type_max - tmp_operand, tmp_operand) == type_max, \
+							return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(1, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(tmp_operand, type_max - tmp_operand + 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max - 1, tmp_operand > 1 ? tmp_operand : 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
+/**
+ * signed addition tests
+ *
+ * parameters:
+ *   type_name: type (s32, s64 etc.)
+ *   type_min:  Minimum value of the type.
+ *   type_max : Maximum value of the type.
+ *   tmp_operand1: random positive value in the set (1, Maximum value / 2)
+ *   tmp_operand2: random negative value in the set (-1, Minimum value / 2)
+ *
+ * Boundary values: (min, min+1, -1, 0, 1, max-1, max)
+ *
+ * Valid tests: Addition result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Addition result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_SIGNED_ADD_TESTS(type_name, type_min, type_max, tmp_operand1, tmp_operand2) do {\
+	unit_assert(nvgpu_safe_add_##type_name(type_min, type_max) == -1, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(0, type_max) == type_max, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(-1, -1) == -2, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(-1, 1) == 0, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(1, 1) == 2, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(type_max - tmp_operand1, tmp_operand1) == type_max, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(type_min - tmp_operand2, tmp_operand2) == type_min, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(tmp_operand1, tmp_operand2) == tmp_operand1 + tmp_operand2, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(tmp_operand1, type_min + 1) == tmp_operand1 + type_min + 1, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_add_##type_name(type_max - 1, tmp_operand2) == type_max - 1 + tmp_operand2, \
+							return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max - tmp_operand1 + 1, tmp_operand1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max, tmp_operand1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_min - tmp_operand2 - 1, tmp_operand2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_min, tmp_operand2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_min, type_min)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_min + 1, type_min + 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max - 1, type_max - 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_max - 1, tmp_operand1 > 1 ? tmp_operand1 : 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_add_##type_name(type_min + 1, tmp_operand2 < -1 ? tmp_operand2 : -2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
+/**
+ * unsigned subtraction tests
+ *
+ * parameters:
+ *   type_name: type (u8, u32 etc.)
+ *   type_max : Maximum value of the type.
+ *   tmp_operand: random value in the set (1, Maximum value)
+ *
+ * Boundary values: (0, 1, max-1, max)
+ *
+ * Valid tests: Subtraction result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Subtraction result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_SUBTRACT_TESTS(type_name, type_max, tmp_operand) do {\
+	unit_assert(nvgpu_safe_sub_##type_name(0, 0) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(1, 0) == 1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max, tmp_operand) == type_max - tmp_operand, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(tmp_operand, 0) == tmp_operand, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max, type_max - 1) == 1, \
+							return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max - 1, 1) == type_max - 2, \
+							return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(0, 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(0, tmp_operand)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(0, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(tmp_operand, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_max - 1, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
+/**
+ * signed subtraction tests
+ *
+ * parameters:
+ *   type_name: type (s32, s64 etc.)
+ *   type_min:  Minimum value of the type.
+ *   type_max : Maximum value of the type.
+ *   tmp_operand1: random positive value in the set (1, Maximum value / 2)
+ *   tmp_operand2: random negative value in the set (-1, Minimum value / 2)
+ *
+ * Boundary values: (min, min+1, -1, 0, 1, max-1, max)
+ *
+ * Valid tests: Subtraction result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Subtraction result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_SIGNED_SUBTRACT_TESTS(type_name, type_min, type_max, tmp_operand1, tmp_operand2) do {\
+	unit_assert(nvgpu_safe_sub_##type_name(tmp_operand2, tmp_operand2) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(tmp_operand2, tmp_operand1) == tmp_operand2 - tmp_operand1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(tmp_operand1, tmp_operand2) == tmp_operand1 - tmp_operand2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(0, 0) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(0, type_max) == 0 - type_max, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max, 0) == type_max, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(-1, -1) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(-1, 1) == -2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(1, -1) == 2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(1, 1) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_min + 1, type_min + 1) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_min, type_min) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max - 1, type_max - 1) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max, type_max) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_min + 1, tmp_operand2) == type_min + 1 - tmp_operand2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max - 1, tmp_operand1) == type_max - 1 - tmp_operand1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(tmp_operand2, tmp_operand2 - type_min) == type_min, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_sub_##type_name(type_max, type_max - tmp_operand1) == tmp_operand1, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_min, tmp_operand1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_max, tmp_operand2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_max - 1, type_min + 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_min + 1, type_max - 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_max - 1, tmp_operand2 < -1 ? tmp_operand2 : -2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_min + 1, tmp_operand1 > 1 ? tmp_operand1 : 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(0, type_min)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_min, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_sub_##type_name(type_max, type_min)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
+/**
+ * unsigned multiplication tests
+ *
+ * parameters:
+ *   type_name: type (u8, u32 etc.)
+ *   type_max : Maximum value of the type.
+ *   tmp_operand: random value in the set (1, Maximum value / 2)
+ *
+ * Boundary values: (0, 1, max-1, max)
+ *
+ * Valid tests: Multiplication result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Multiplication result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_MULT_TESTS(type_name, type_max, tmp_operand) do {\
+	unit_assert(nvgpu_safe_mult_##type_name(0, type_max) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(type_max - 1, 1) == type_max - 1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(tmp_operand, 2) == tmp_operand * 2, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max - 1, 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max - 1, tmp_operand > 1 ? tmp_operand : 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
+/**
+ * signed multiplication tests
+ *
+ * parameters:
+ *   type_name: type (s32, s64 etc.)
+ *   type_min:  Minimum value of the type.
+ *   type_max : Maximum value of the type.
+ *   tmp_operand1: random positive value in the set (1, Maximum value / 2)
+ *   tmp_operand2: random negative value in the set (-1, Minimum value / 2)
+ *
+ * Boundary values: (min, min+1, -1, 0, 1, max-1, max)
+ *
+ * Valid tests: Multiplication result within range for each boundary value and
+ *              random value.
+ * Invalid tests: Multiplication result out of range if possible for each
+ *                boundary and random value.
+ */
+#define GENERATE_ARITHMETIC_SIGNED_MULT_TESTS(type_name, type_min, type_max, tmp_operand1, tmp_operand2) do {\
+	unit_assert(nvgpu_safe_mult_##type_name(0, type_max) == 0, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(1, type_min) == type_min, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(-1, -1) == 1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(-1, 1) == -1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(1, 1) == 1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(tmp_operand1, 2) == tmp_operand1 * 2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(tmp_operand2, 2) == tmp_operand2 * 2, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(type_max, -1) == type_max * -1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(type_max - 1, -1) == (type_max - 1) * -1, return UNIT_FAIL); \
+	unit_assert(nvgpu_safe_mult_##type_name(type_min + 1, -1) == (type_min + 1) * -1, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max, 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max, -2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min, 2)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min, -1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min, type_min)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min, type_max)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min + 1, type_min + 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_max - 1, type_max - 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+	err = EXPECT_BUG((void)nvgpu_safe_mult_##type_name(type_min + 1, type_max - 1)); \
+	unit_assert(err != 0, return UNIT_FAIL); \
+} while (0)
+
 int test_arithmetic(struct unit_module *m, struct gk20a *g, void *args)
 {
+	s64 tmp_s64, tmp_s64_neg;
+	s32 tmp_s32, tmp_s32_neg;
+	u64 tmp_u64;
+	u32 tmp_u32;
+	u8 tmp_u8;
 	int err;
 
+	srand(time(0));
+
+	/* random value in the set (1, U8_MAX - 1) */
+	tmp_u8 = rand() % U8_MAX ?: 1;
+
+	/* random value in the set ((u32)1, (u32)(U32_MAX - 1)) */
+	tmp_u32 = (u32)((rand() % U32_MAX) ?: 1);
+
+	/* random value in the set ((s32)1, (s32)(U32_MAX - 1) / 2) */
+	tmp_s32 = (s32)((rand() % U32_MAX) / 2 ?: 1);
+
+	/* random value in the set ((s32)-1, (s32)-((U32_MAX - 1) / 2)) */
+	tmp_s32_neg = (s32)(0 - (rand() % U32_MAX) / 2) ?: -1;
+
+	/* random value in the set ((u64)1, (u64)(U64_MAX - 1)) */
+	tmp_u64 = (u64)((rand() % U64_MAX) ?: 1);
+
+	/* random value in the set ((s64)1, (s64)(U64_MAX - 1) / 2) */
+	tmp_s64 = (s64)((rand() % U64_MAX) / 2 ?: 1);
+
+	/* random value in the set ((s64)-1, (s64)-((U64_MAX - 1) / 2)) */
+	tmp_s64_neg = (s64)(0 - (rand() % U64_MAX) / 2) ?: -1;
+
+	unit_info(m, "random operands\nu8: %u, u32: %u, s32: %d, s32_neg: %d\nu64: %llu, s64: %lld, s64_neg: %lld\n",
+		  tmp_u8, tmp_u32, tmp_s32, tmp_s32_neg, tmp_u64, tmp_s64, tmp_s64_neg);
+
 	/* U8 sub */
-	unit_assert(nvgpu_safe_sub_u8(1, 1) == 0, return UNIT_FAIL);
-	unit_assert(nvgpu_safe_sub_u8(U8_MAX, U8_MAX) == 0,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_u8(0, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SUBTRACT_TESTS(u8, U8_MAX, tmp_u8);
 
 	/* U32 add */
-	unit_assert(nvgpu_safe_add_u32(U32_MAX-1, 1) == U32_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_u32(U32_MAX, 0) == U32_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_u32(U32_MAX, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_ADD_TESTS(u32, U32_MAX, tmp_u32);
 
 	/* S32 add */
-	unit_assert(nvgpu_safe_add_s32(INT_MAX-1, 1) == INT_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_s32(INT_MAX, 0) == INT_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_s32(INT_MIN+1, -1) == INT_MIN,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_s32(INT_MAX, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_s32(INT_MIN, -1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SIGNED_ADD_TESTS(s32, INT_MIN, INT_MAX, tmp_s32, tmp_s32_neg);
 
 	/* U32 sub */
-	unit_assert(nvgpu_safe_sub_u32(1, 1) == 0, return UNIT_FAIL);
-	unit_assert(nvgpu_safe_sub_u32(U32_MAX, U32_MAX) == 0,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_u32(0, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SUBTRACT_TESTS(u32, U32_MAX, tmp_u32);
 
 	/* S32 sub */
-	unit_assert(nvgpu_safe_sub_s32(INT_MIN+1, 1) == INT_MIN,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_sub_s32(INT_MAX-1, -1) == INT_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_s32(INT_MIN, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_s32(INT_MAX, -1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SIGNED_SUBTRACT_TESTS(s32, INT_MIN, INT_MAX, tmp_s32, tmp_s32_neg);
 
 	/* U32 Mult */
-	unit_assert(nvgpu_safe_mult_u32(0, 0) == 0,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_u32(U32_MAX, 0) == 0,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_u32(U32_MAX, 1) == U32_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_u32(U32_MAX, 2));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_MULT_TESTS(u32, U32_MAX, tmp_u32);
 
 	/* U64 add */
-	unit_assert(nvgpu_safe_add_u64(U64_MAX-1, 1) == U64_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_u64(U64_MAX, 0) == U64_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_u64(U64_MAX, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_ADD_TESTS(u64, U64_MAX, tmp_u64);
 
 	/* S64 add */
-	unit_assert(nvgpu_safe_add_s64(LONG_MAX-1, 1) == LONG_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_s64(LONG_MAX, 0) == LONG_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_add_s64(LONG_MIN+1, -1) == LONG_MIN,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_s64(LONG_MAX, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_add_s64(LONG_MIN, -1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SIGNED_ADD_TESTS(s64, LONG_MIN, LONG_MAX, tmp_s64, tmp_s64_neg);
 
 	/* U64 sub */
-	unit_assert(nvgpu_safe_sub_u64(1, 1) == 0, return UNIT_FAIL);
-	unit_assert(nvgpu_safe_sub_u64(U64_MAX, U64_MAX) == 0,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_u64(0, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SUBTRACT_TESTS(u64, U64_MAX, tmp_u64);
 
 	/* S64 sub */
-	unit_assert(nvgpu_safe_sub_s64(LONG_MIN+1, 1) == LONG_MIN,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_sub_s64(LONG_MAX-1, -1) == LONG_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_s64(LONG_MIN, 1));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_sub_s64(LONG_MAX, -1));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SIGNED_SUBTRACT_TESTS(s64, LONG_MIN, LONG_MAX, tmp_s64, tmp_s64_neg);
 
 	/* U64 Mult */
-	unit_assert(nvgpu_safe_mult_u64(0, 0) == 0,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_u64(U64_MAX, 0) == 0,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_u64(U64_MAX, 1) == U64_MAX,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_u64(U64_MAX, 2));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_MULT_TESTS(u64, U64_MAX, tmp_u64);
 
 	/* S64 Mult */
-	unit_assert(nvgpu_safe_mult_s64(LONG_MAX, 1) == LONG_MAX,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_s64(LONG_MAX, -1) == (-1 * LONG_MAX),
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_s64(-1, LONG_MAX) == (-1 * LONG_MAX),
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_s64(0, LONG_MAX) == 0,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_s64(-1, -1) == 1,
-							return UNIT_FAIL);
-	unit_assert(nvgpu_safe_mult_s64(0, -1) == 0,
-							return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_s64(LONG_MAX, 2));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_s64(LONG_MAX, -2));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_s64(-1*LONG_MAX, 2));
-	unit_assert(err != 0, return UNIT_FAIL);
-	err = EXPECT_BUG((void)nvgpu_safe_mult_s64(-1*LONG_MAX, -2));
-	unit_assert(err != 0, return UNIT_FAIL);
+	GENERATE_ARITHMETIC_SIGNED_MULT_TESTS(s64, LONG_MIN, LONG_MAX, tmp_s64, tmp_s64_neg);
 
 	return UNIT_SUCCESS;
 }
@@ -279,4 +482,4 @@ struct unit_module_test static_analysis_tests[] = {
 	UNIT_TEST(safety_checks,	test_safety_checks,	NULL, 0),
 };
 
-UNIT_MODULE(static_analysis, static_analysis_tests, UNIT_PRIO_NVGPU_TEST);
\ No newline at end of file
+UNIT_MODULE(static_analysis, static_analysis_tests, UNIT_PRIO_NVGPU_TEST);
diff --git a/userspace/units/interface/static_analysis/static_analysis.h b/userspace/units/interface/static_analysis/static_analysis.h
index 66933316b..48d308520 100644
--- a/userspace/units/interface/static_analysis/static_analysis.h
+++ b/userspace/units/interface/static_analysis/static_analysis.h
@@ -1,5 +1,5 @@
 /*
- * Copyright (c) 2020, NVIDIA CORPORATION.  All rights reserved.
+ * Copyright (c) 2020-2021, NVIDIA CORPORATION.  All rights reserved.
  *
  * Permission is hereby granted, free of charge, to any person obtaining a
  * copy of this software and associated documentation files (the "Software"),
@@ -37,7 +37,7 @@ struct unit_module;
  *
  * Description: Verify functionality of static analysis safe arithmetic APIs.
  *
- * Test Type: Feature, Error guessing
+ * Test Type: Feature, Error guessing, Boundary Value
  *
  * Targets: nvgpu_safe_sub_u8, nvgpu_safe_add_u32, nvgpu_safe_add_s32,
  *          nvgpu_safe_sub_u32, nvgpu_safe_sub_s32, nvgpu_safe_mult_u32,
@@ -46,6 +46,66 @@ struct unit_module;
  *
  * Input: None
  *
+ * -# Unsigned addition tests:
+ *    Boundary values: {0, max}
+ *
+ *    Equivalence classes:
+ *    Variable: Addition result of two unsigned operands.
+ *    - Valid tests: Addition result within range for each boundary value and
+ *                   random value in the set.
+ *    - Invalid tests: Addition result out of range if possible for each
+ *                     boundary and random value.
+ *
+ * -# Signed addition tests:
+ *     Boundary values: {min, 0, max}
+ *
+ *     Equivalence classes:
+ *     Variable: Addition result of two signed operands.
+ *     - Valid tests: Addition result within range for each boundary value and
+ *                    random value.
+ *     - Invalid tests: Addition result out of range if possible for each
+ *                      boundary and random value.
+ *
+ * -# Unsigned subtraction tests:
+ *    Boundary values: {0, max}
+ *
+ *    Equivalence classes:
+ *    Variable: Subtraction result of two unsigned operands.
+ *    - Valid tests: Subtraction result within range for each boundary value and
+ *                   random value.
+ *    - Invalid tests: Subtraction result out of range if possible for each
+ *                     boundary and random value.
+ *
+ * -# Signed subtraction tests:
+ *    Boundary values: {min, 0, max}
+ *
+ *    Equivalence classes:
+ *    Variable: Subtraction result of two signed operands.
+ *    - Valid tests: Subtraction output within range for each boundary value and
+ *                   random value.
+ *    - Invalid tests: Subtraction output out of range if possible for each
+ *                     boundary and random value.
+ *
+ * -# Unsigned multiplication tests:
+ *    Boundary values: {0, max}
+ *
+ *    Equivalence classes:
+ *    Variable: Multiplication result of two unsigned operands.
+ *    - Valid tests: Multiplication result within range for each boundary value and
+ *                   random value.
+ *    - Invalid tests: Multiplication result out of range if possible for each
+ *                     boundary and random value.
+ *
+ * -# Signed multiplication tests:
+ *    Boundary values: {min, 0, max}
+ *
+ *    Equivalence classes:
+ *    Variable: Multiplication result of two signed operands.
+ *    - Valid tests: Multiplication result within range for each boundary value and
+ *                   random value.
+ *    - Invalid tests: Multiplication result out of range if possible for each
+ *                     boundary and random value.
+ *
  * Steps:
  * - Call the static analysis arithmetic APIs. Pass in valid values and verify
  *   correct return.