Open source GPL/LGPL release

userspace/units/posix/utils/Makefile

@@ -0,0 +1,26 @@
+# Copyright (c) 2020, 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"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+# DEALINGS IN THE SOFTWARE.
+
+.SUFFIXES:
+
+OBJS   = posix-utils.o
+MODULE = posix-utils
+
+include ../../Makefile.units

userspace/units/posix/utils/Makefile.interface.tmk

@@ -0,0 +1,35 @@
+################################### tell Emacs this is a -*- makefile-gmake -*-
+#
+# Copyright (c) 2020, 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"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+# DEALINGS IN THE SOFTWARE.
+#
+# tmake for SW Mobile component makefile
+#
+###############################################################################
+
+NVGPU_UNIT_NAME=posix-utils
+
+include $(NV_COMPONENT_DIR)/../../Makefile.units.common.interface.tmk
+
+# Local Variables:
+# indent-tabs-mode: t
+# tab-width: 8
+# End:
+# vi: set tabstop=8 noexpandtab:

userspace/units/posix/utils/Makefile.tmk

@@ -0,0 +1,35 @@
+################################### tell Emacs this is a -*- makefile-gmake -*-
+#
+# Copyright (c) 2020, 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"),
+# to deal in the Software without restriction, including without limitation
+# the rights to use, copy, modify, merge, publish, distribute, sublicense,
+# and/or sell copies of the Software, and to permit persons to whom the
+# Software is furnished to do so, subject to the following conditions:
+#
+# The above copyright notice and this permission notice shall be included in
+# all copies or substantial portions of the Software.
+#
+# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+# IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+# FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+# THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+# LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+# FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+# DEALINGS IN THE SOFTWARE.
+#
+# tmake for SW Mobile component makefile
+#
+###############################################################################
+
+NVGPU_UNIT_NAME=posix-utils
+
+include $(NV_COMPONENT_DIR)/../../Makefile.units.common.tmk
+
+# Local Variables:
+# indent-tabs-mode: t
+# tab-width: 8
+# End:
+# vi: set tabstop=8 noexpandtab:

userspace/units/posix/utils/posix-utils.c

@@ -0,0 +1,628 @@
+/*
+ * 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"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+#include <stdlib.h>
+#include <stdint.h>
+
+#include <unit/io.h>
+#include <unit/unit.h>
+
+#include <nvgpu/utils.h>
+#include <nvgpu/barrier.h>
+
+#include "posix-utils.h"
+
+#define KHZ (1000U)
+#define MHZ (1000000U)
+
+#define ARRAY1_SIZE	4
+#define ARRAY2_SIZE	10
+
+#define PAGE_ALIGN_TEST_VALUE	0x3fffffff
+#define ALIGN_TEST_VALUE	0xffff
+#define ALIGN_WITH_VALUE	0x10
+#define ALIGN_WITH_MASK		0x3
+
+#define TO_ROUND_VALUE		11U
+#define ROUND_BY_VALUE		4U
+#define ROUND_UP_RESULT		12U
+#define ROUND_DOWN_RESULT	8U
+
+struct test_container {
+	uint32_t var1;
+	uint32_t var2;
+};
+
+struct test_container cont = {20, 30};
+
+/*
+ * Test to ensure the EXPECT_BUG construct works as intended by making sure it
+ * behaves properly when BUG is called or not.
+ * In the event that EXPECT_BUG is completely broken, the call to BUG() would
+ * cause the unit to crash and report a failure correctly.
+ */
+int test_hamming_weight(struct unit_module *m,
+				struct gk20a *g, void *args)
+{
+	unsigned int result;
+	unsigned int i;
+	uint8_t hwt_8bit;
+	uint16_t hwt_16bit;
+	uint32_t hwt_32bit;
+	uint64_t hwt_64bit;
+
+	for (i = 0; i < 8; i++) {
+		hwt_8bit = (unsigned int) 1 << i;
+		result = nvgpu_posix_hweight8(hwt_8bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"8 bit hwt failed for %d\n", hwt_8bit);
+		}
+	}
+
+	for (i = 0; i < 16; i++) {
+		hwt_16bit = (unsigned int) 1 << i;
+		result = nvgpu_posix_hweight16(hwt_16bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"16 bit hwt failed for %d\n", hwt_16bit);
+		}
+	}
+
+	for (i = 0; i < 32; i++) {
+		hwt_32bit = (unsigned int) 1 << i;
+		result = nvgpu_posix_hweight32(hwt_32bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"32 bit hwt failed for %d\n", hwt_32bit);
+		}
+	}
+
+	for (i = 0; i < 32; i++) {
+		hwt_32bit = (unsigned int) 1 << i;
+		result = hweight32(hwt_32bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"hweight32 failed for %d\n", hwt_32bit);
+		}
+	}
+
+	for (i = 0; i < 64; i++) {
+		hwt_64bit = (unsigned long) 1 << i;
+		result = nvgpu_posix_hweight64(hwt_64bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"64 bit hwt failed for %lx\n", hwt_64bit);
+		}
+	}
+
+	for (i = 0; i < 64; i++) {
+		hwt_64bit = (unsigned long) 1 << i;
+		result = hweight_long(hwt_64bit);
+		if (result != 1) {
+			unit_return_fail(m,
+				"hweight_long failed for %lx\n", hwt_64bit);
+		}
+	}
+
+	hwt_8bit = 0x0;
+	result = nvgpu_posix_hweight8(hwt_8bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"8 bit hwt failed for %d\n", hwt_8bit);
+	}
+
+	hwt_8bit = 0xff;
+	result = nvgpu_posix_hweight8(hwt_8bit);
+	if (result != 8) {
+		unit_return_fail(m,
+			"8 bit hwt failed for %d\n", hwt_8bit);
+	}
+
+	hwt_16bit = 0x0;
+	result = nvgpu_posix_hweight16(hwt_16bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"16 bit hwt failed for %d\n", hwt_16bit);
+	}
+
+	hwt_16bit = 0xffff;
+	result = nvgpu_posix_hweight16(hwt_16bit);
+	if (result != 16) {
+		unit_return_fail(m,
+			"16 bit hwt failed for %d\n", hwt_16bit);
+	}
+
+	hwt_32bit = 0x0;
+	result = nvgpu_posix_hweight32(hwt_32bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"32 bit hwt failed for %d\n", hwt_32bit);
+	}
+
+	hwt_32bit = 0xffffffff;
+	result = nvgpu_posix_hweight32(hwt_32bit);
+	if (result != 32) {
+		unit_return_fail(m,
+			"32 bit hwt failed for %d\n", hwt_32bit);
+	}
+
+	hwt_32bit = 0x0;
+	result = hweight32(hwt_32bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"hweight32 failed for %d\n", hwt_32bit);
+	}
+
+	hwt_32bit = 0xffffffff;
+	result = hweight32(hwt_32bit);
+	if (result != 32) {
+		unit_return_fail(m,
+			"hweight32 failed for %d\n", hwt_32bit);
+	}
+
+	hwt_64bit = 0x0;
+	result = nvgpu_posix_hweight64(hwt_64bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"64 bit hwt failed for %ld\n", hwt_64bit);
+	}
+
+	hwt_64bit = 0xffffffffffffffff;
+	result = nvgpu_posix_hweight64(hwt_64bit);
+	if (result != 64) {
+		unit_return_fail(m,
+			"64 bit hwt failed for %ld\n", hwt_64bit);
+	}
+
+	hwt_64bit = 0x0;
+	result = hweight_long(hwt_64bit);
+	if (result != 0) {
+		unit_return_fail(m,
+			"hweight_long failed for %ld\n", hwt_64bit);
+	}
+
+	hwt_64bit = 0xffffffffffffffff;
+	result = hweight_long(hwt_64bit);
+	if (result != 64) {
+		unit_return_fail(m,
+			"hweight_long failed for %ld\n", hwt_64bit);
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_be32tocpu(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t pattern;
+	uint32_t result;
+	uint8_t *ptr;
+
+	pattern = 0xaabbccdd;
+
+	ptr = (uint8_t*)&pattern;
+
+	result = be32_to_cpu(pattern);
+
+	if (*ptr == 0xdd) {
+		if (result != 0xddccbbaa) {
+			unit_return_fail(m,
+				"be32tocpu failed for %x %x\n",
+				pattern, result);
+		}
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_minmax(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t i;
+	uint32_t a;
+	uint32_t b;
+	uint32_t c;
+	uint32_t result;
+
+	a = 10;
+	b = 20;
+	c = 30;
+	for (i = 0; i < 10; i++) {
+		result = min(a, b);
+		if (result != a) {
+			unit_return_fail(m, "min failure %d\n", result);
+		}
+
+		result = min(b, a);
+		if (result != a) {
+			unit_return_fail(m, "min failure %d\n", result);
+		}
+
+		a += 5;
+		b += 5;
+	}
+
+	a = 100;
+	b = 200;
+	c = 300;
+	for (i = 0; i < 10; i++) {
+		result = min3(a, b, c);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		result = min3(a, c, b);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		result = min3(b, a, c);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		result = min3(b, c, a);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		result = min3(c, a, b);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		result = min3(c, b, a);
+		if (result != a) {
+			unit_return_fail(m, "min3 failure %d\n", result);
+		}
+
+		a += 5;
+		b += 5;
+		c += 5;
+	}
+
+	b = 2000;
+	c = 3000;
+	for (i = 0; i < 10; i++) {
+		result = min_t(uint32_t, b, c);
+		if (result != b) {
+			unit_return_fail(m, "min_t failure %d\n", result);
+		}
+
+		result = min_t(uint32_t, c, b);
+		if (result != b) {
+			unit_return_fail(m, "min_t failure %d\n", result);
+		}
+
+		b += 100;
+		c += 100;
+	}
+
+	a = 1000;
+	b = 2000;
+	for (i = 0; i < 10; i++) {
+		result = max(a, b);
+		if (result != b) {
+			unit_return_fail(m, "max failure %d\n", result);
+		}
+
+		result = max(b, a);
+		if (result != b) {
+			unit_return_fail(m, "max failure %d\n", result);
+		}
+
+		a += 100;
+		b += 100;
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_arraysize(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result;
+	uint32_t array1[ARRAY1_SIZE] = {0};
+	uint64_t array2[ARRAY2_SIZE] = {0};
+
+	result = ARRAY_SIZE(array1);
+	if (result != ARRAY1_SIZE) {
+		unit_return_fail(m, "ARRAY SIZE failure %d\n", result);
+	}
+
+	result = ARRAY_SIZE(array2);
+	if (result != ARRAY2_SIZE) {
+		unit_return_fail(m, "ARRAY SIZE failure %d\n", result);
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_typecheck(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result;
+	unsigned int test1;
+	unsigned long  test2;
+	signed int test3;
+	signed long test4;
+
+	result = IS_UNSIGNED_TYPE(test1);
+	if (!result) {
+		unit_return_fail(m,
+			"IS_UNSIGNED_TYPE failure for uint %d\n", result);
+	}
+
+	result = IS_UNSIGNED_TYPE(test2);
+	if (!result) {
+		unit_return_fail(m,
+			"IS_UNSIGNED_TYPE failure for ulong %d\n", result);
+	}
+
+	result = IS_UNSIGNED_TYPE(test3);
+	if (result) {
+		unit_return_fail(m,
+			"IS_UNSIGNED_TYPE failure for int %d\n", result);
+	}
+
+	result = IS_UNSIGNED_LONG_TYPE(test2);
+	if (!result) {
+		unit_return_fail(m,
+			"IS_UNSIGNED_LONG_TYPE failure for ulong %d\n",
+			result);
+	}
+
+	result = IS_UNSIGNED_LONG_TYPE(test4);
+	if (result) {
+		unit_return_fail(m,
+			"IS_UNSIGNED_LONG_TYPE failure for long %d\n",
+			result);
+	}
+
+	result = IS_SIGNED_LONG_TYPE(test2);
+	if (result) {
+		unit_return_fail(m,
+			"IS_SIGNED_LONG_TYPE failure for ulong %d\n",
+			result);
+	}
+
+	result = IS_SIGNED_LONG_TYPE(test4);
+	if (!result) {
+		unit_return_fail(m,
+			"IS_SIGNED_LONG_TYPE failure for long %d\n",
+			result);
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_align_macros(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result;
+	unsigned int test1;
+
+	test1 = ALIGN_TEST_VALUE;
+	result = ALIGN_WITH_VALUE;
+	test1 = NVGPU_ALIGN(test1, result);
+	if (test1 & (ALIGN_WITH_VALUE - 1)) {
+		unit_return_fail(m,
+			"ALIGN failure %x\n", test1);
+	}
+
+	test1 = ALIGN_TEST_VALUE;
+	result = ALIGN_WITH_MASK;
+	test1 = ALIGN_MASK(test1, result);
+	if (test1 & ALIGN_WITH_MASK) {
+		unit_return_fail(m,
+			"ALIGN_MASK failure %x\n", test1);
+	}
+
+	test1 = PAGE_ALIGN_TEST_VALUE;
+	result = PAGE_ALIGN(test1);
+	if (result & (NVGPU_CPU_PAGE_SIZE - 1)) {
+		unit_return_fail(m,
+			"PAGE_ALIGN failure %x\n", result);
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_round_macros(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result, i, test1;
+
+	for (i = 1; i < 8; i++) {
+		result = 1U << i;
+		if (round_mask(test1, result) != (result - 1U)) {
+			unit_return_fail(m,
+				"round_mask failure %d\n", result);
+		}
+	}
+
+	result = ROUND_BY_VALUE;
+	for (i = 0; i < ROUND_BY_VALUE; i++) {
+		test1 = (ROUND_DOWN_RESULT + 1U) + i;
+		if (round_up(test1, result) != ROUND_UP_RESULT) {
+			unit_return_fail(m, "round_up failure %d %d\n", test1, i);
+		}
+	}
+
+	result = ROUND_BY_VALUE;
+	for (i = 0; i < ROUND_BY_VALUE; i++) {
+		test1 = (ROUND_UP_RESULT - 1U) - i;
+		if (round_down(test1, result) != ROUND_DOWN_RESULT) {
+			unit_return_fail(m, "round_down failure\n");
+		}
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_write_once(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result, i, test1;
+
+	test1 = 20;
+	for (i = 0 ; i < 10; i++) {
+		test1 += 1;
+		NV_WRITE_ONCE(result, test1);
+		if (result != test1) {
+			unit_return_fail(m,
+				"NV_WRITE_ONCE failure %d\n", result);
+		}
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_div_macros(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t result, test1;
+	uint64_t test2, test5;
+
+	test1 = 199/20;
+	test2 = DIV_ROUND_UP_U64(199, 20);
+	if (test2 != (test1 + 1)) {
+		unit_return_fail(m,
+			"DIV_ROUND_UP_U64 failure %ld\n", test2);
+	}
+
+	test1 = 239/40;
+	result = DIV_ROUND_UP((uint32_t) 239, (uint32_t) 40);
+	if (result != (test1 + 1)) {
+		unit_return_fail(m,
+			"DIV_ROUND_UP failure %d\n", result);
+	}
+
+	test1 = 640;
+	result = 100;
+	do_div(test1, result);
+	if (test1 != 6) {
+		unit_return_fail(m,
+			"do_div failure %d\n", test1);
+	}
+
+	test2 = 800;
+	test5 = 200;
+	result = div64_u64(test2, test5);
+	if (result != (test2/test5)) {
+		unit_return_fail(m,
+			"div64_u64 failure %d\n", result);
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_containerof(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	struct test_container *contptr;
+	struct test_container *contptr1;
+	struct test_container *contptr2;
+	uint32_t *varptr1;
+	uint32_t *varptr2;
+
+	contptr = &cont;
+	varptr1 = &cont.var1;
+	varptr2 = &cont.var2;
+
+	contptr1 = container_of(varptr1, struct test_container, var1);
+	contptr2 = container_of(varptr2, struct test_container, var2);
+
+	if ((contptr1 != contptr) || (contptr2 != contptr)) {
+		unit_return_fail(m, "container_of failure\n");
+	}
+
+	return UNIT_SUCCESS;
+}
+
+int test_hertzconversion(struct unit_module *m,
+			struct gk20a *g, void *args)
+{
+	uint32_t i, hz, khz, mhz;
+	uint64_t long_hz;
+
+	for (i = 1; i < 10; i++) {
+		hz = i * 1000U;
+		khz = HZ_TO_KHZ(hz);
+		if (khz != i) {
+			unit_return_fail(m, "HZ_TO_KHZ failure\n");
+		}
+
+		if (hz != KHZ_TO_HZ(i)) {
+			unit_return_fail(m, "KHZ_TO_HZ failure\n");
+		}
+
+		hz = i * 1000000U;
+		mhz = HZ_TO_MHZ(hz);
+		if (mhz != i) {
+			unit_return_fail(m, "HZ_TO_MHZ failure\n");
+		}
+
+		long_hz = i * 1000000U;
+		mhz = HZ_TO_MHZ_ULL(long_hz);
+		if (mhz != i) {
+			unit_return_fail(m, "HZ_TO_MHZ_ULL failure\n");
+		}
+	}
+
+	for (i = 0; i < 10; i++) {
+		khz = i * 1000U;
+		mhz = KHZ_TO_MHZ(khz);
+		if (mhz != i) {
+			unit_return_fail(m, "KHZ_TO_MHZ failure\n");
+		}
+
+		if (khz != MHZ_TO_KHZ(i)) {
+			unit_return_fail(m, "MHZ_TO_KHZ failure\n");
+		}
+	}
+
+	for (i = 0; i < 10; i++) {
+		hz = i * 1000000;
+		if (hz != MHZ_TO_HZ_ULL(i)) {
+			unit_return_fail(m, "MHZ_TO_HZ_ULL failure\n");
+		}
+	}
+
+	return UNIT_SUCCESS;
+}
+struct unit_module_test posix_utils_tests[] = {
+	UNIT_TEST(hweight_test, test_hamming_weight, NULL, 0),
+	UNIT_TEST(be32tocpu_test, test_be32tocpu, NULL, 0),
+	UNIT_TEST(minmax_test, test_minmax, NULL, 0),
+	UNIT_TEST(arraysize_test, test_arraysize, NULL, 0),
+	UNIT_TEST(typecheck_test, test_typecheck, NULL, 0),
+	UNIT_TEST(alignmacros_test, test_align_macros, NULL, 0),
+	UNIT_TEST(roundmacros_test, test_round_macros, NULL, 0),
+	UNIT_TEST(writeonce_test, test_write_once, NULL, 0),
+	UNIT_TEST(divmacros_test, test_div_macros, NULL, 0),
+	UNIT_TEST(containerof_test, test_containerof, NULL, 0),
+	UNIT_TEST(conversion_test, test_hertzconversion, NULL, 0),
+};
+
+UNIT_MODULE(posix_utils, posix_utils_tests, UNIT_PRIO_POSIX_TEST);

userspace/units/posix/utils/posix-utils.h

@@ -0,0 +1,326 @@
+/*
+ * Copyright (c) 2020, 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"),
+ * to deal in the Software without restriction, including without limitation
+ * the rights to use, copy, modify, merge, publish, distribute, sublicense,
+ * and/or sell copies of the Software, and to permit persons to whom the
+ * Software is furnished to do so, subject to the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
+ * THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
+ * DEALINGS IN THE SOFTWARE.
+ */
+
+/**
+ * @addtogroup SWUTS-posix-utils
+ * @{
+ *
+ * Software Unit Test Specification for posix-utils
+ */
+
+#ifndef __UNIT_POSIX_UTILS_H__
+#define __UNIT_POSIX_UTILS_H__
+
+/**
+ * Test specification for test_hamming_weight
+ *
+ * Description: Test the hamming weight implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: nvgpu_posix_hweight8, nvgpu_posix_hweight16,
+ *          nvgpu_posix_hweight32, nvgpu_posix_hweight64,
+ *          hweight32, hweight_long
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Call nvgpu_posix_hweight8 in loop with only the loop index bit position
+ *    set.
+ * 2) Return FAIL if the return value from nvgpu_posix_hweight8 is not equal
+ *    to 1 in any of the iterations.
+ * 3) Repeat steps 1 and 2 for nvgpu_posix_hweight16, nvgpu_posix_hweight32
+ *    nvgpu_posix_hweight64, hweight32 and hweight_long.
+ * 4) Call nvgpu_posix_hweight8 with input parameter set as 0.
+ * 5) Return FAIL if the return value from nvgpu_posix_hweight8 is not equal
+ *    to 0.
+ * 6) Call nvgpu_posix_hweight8 with input parameter set to maximum value.
+ * 7) Return FAIL if the return value from nvgpu_posix_hweight8 is not equal
+ *    to the number of bits in the input parameter.
+ * 8) Repeat steps 4,5,6 and 7 for nvgpu_posix_hweight16, nvgpu_posix_hweight32
+ *    nvgpu_posix_hweight64, hweight32 and hweight_long.
+ *
+ * Output:
+ * The test returns PASS if all the hamming weight function invocations return
+ * the expected value. Otherwise the test returns FAIL.
+ *
+ */
+int test_hamming_weight(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_be32tocpu
+ *
+ * Description: Test the endian conversion implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: be32_to_cpu
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke function be32_to_cpu with a fixed pattern as input.
+ * 2) Check if the machine is little endian.
+ * 3) If the machine is little endian, confirm that the return value from
+ *    be32_to_cpu is equal to the little endian order of the pattern, else
+ *    return FAIL.
+ * 3) EXPECT_BUG is also tested to make sure that BUG is not called where it is
+ *    not expected.
+ *
+ * Output:
+ * The test returns PASS if BUG is called as expected based on the parameters
+ * passed and EXPECT_BUG handles it accordingly.
+ * The test returns FAIL if either BUG is not called as expected or if
+ * EXPECT_BUG indicates that a BUG call was made which was not requested by
+ * the test.
+ *
+ */
+int test_be32tocpu(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_minmax
+ *
+ * Description: Test the min and max implementations.
+ *
+ * Test Type: Feature
+ *
+ * Targets: min_t, min, min3, max
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke function min in loop with different input parameter values.
+ * 2) Check if the return value is the minimum value among the parameters
+ *    passed. Else return FAIL.
+ * 3) Invoke function min3 in loop with different input parameter values.
+ * 4) Check if the return value is the minimum value among the parameters
+ *    passed. Else return FAIL.
+ * 5) Invoke function min_t in loop with type and values as input parameters.
+ * 6) Check if the return value is the minimum value among the parameters
+ *    passed for every iteration. Else return FAIL.
+ * 7) Invoke function max in loop.
+ * 8) Check if the return value is the maximum value among the parameters
+ *    passed for every iteration. Else return FAIL.
+ * 9) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of min and max implementations
+ * returns the expected value. Otherwise, test returns FAIL.
+ *
+ */
+int test_minmax(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_arraysize
+ *
+ * Description: Test ARRAY_SIZE macro implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: ARRAY_SIZE
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macro ARRAY_SIZE with multiple arrays and confirm
+ *    that the results are as expected. Otherwise, return FAIL.
+ * 4) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of ARRAY_SIZE macro returns
+ * the results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_arraysize(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_typecheck
+ *
+ * Description: Test type checking macros.
+ *
+ * Test Type: Feature
+ *
+ * Targets: IS_UNSIGNED_TYPE, IS_UNSIGNED_LONG_TYPE,
+ *          IS_SIGNED_LONG_TYPE
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macros IS_UNSIGNED_TYPE, IS_UNSIGNED_LONG_TYPE,
+ *    IS_SIGNED_LONG_TYPE ARRAY_SIZE with multiple data types and confirm
+ *    that the results are as expected. Otherwise, return FAIL.
+ * 2) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of type checking macros returns
+ * the results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_typecheck(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_align_macros
+ *
+ * Description: Test align macro implementations.
+ *
+ * Test Type: Feature
+ *
+ * Targets: ALIGN, ALIGN_MASK, PAGE_ALIGN
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macros ALIGN, ALIGN_MASK and PAGE_ALIGN and confirm that the
+ *    results are masked as expected. Otherwise, return FAIL.
+ * 2) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of various align macros
+ * returns the results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_align_macros(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_round_macros
+ *
+ * Description: Test rounding macro implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: round_mask, round_up, round_down
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macro round_mask in loop and confirm that the mask generated is
+ *    as expected. Otherwise, return FAIL.
+ * 2) Invoke macros round_up and round_up in loop for various input values and
+ *    confirm that the values are rounded off as expected. Otherwise, return
+ *    FAIL.
+ * 3) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of round macros returns the
+ * results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_round_macros(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_write_once
+ *
+ * Description: Test WRITE_ONCE macro implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: WRITE_ONCE
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macro WRITE_ONCE in loop and confirm that the value is written
+ *    into the variable as expected. Otherwise, return FAIL.
+ * 2) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of WRITE_ONCE macro writes the
+ * value into the variable. Otherwise, the test returns FAIL.
+ *
+ */
+int test_write_once(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_div_macros
+ *
+ * Description: Test various division macro implementations.
+ *
+ * Test Type: Feature
+ *
+ * Targets: DIV_ROUND_UP_U64, DIV_ROUND_UP, do_div, div64_u64
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke macros DIV_ROUND_UP_U64, DIV_ROUND_UP, do_div and div64_u64 and
+ *    confirm that the results are as expected. Otherwise, return FAIL.
+ * 2) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of various division macros
+ * returns the results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_div_macros(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_containerof
+ *
+ * Description: Test container_of implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: container_of
+ *
+ * Inputs: Global struct instance cont.
+ *
+ * Steps:
+ * 1) Invoke container_of with the first variable ptr in cont.
+ * 2) Invoke container_of with the second variable ptr in cont.
+ * 3) Confirm if both the invocation of container_of macro returns the address
+ *    of the global struct instance cont. Otherwise, return FAIL.
+ * 4) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if both the invocation of container_of returns the
+ * same address as that of the global struct instance cont. Otherwise, the test
+ * returns FAIL.
+ *
+ */
+int test_containerof(struct unit_module *m, struct gk20a *g, void *args);
+
+/**
+ * Test specification for test_hertzconversion
+ *
+ * Description: Test hertz conversion macro implementation.
+ *
+ * Test Type: Feature
+ *
+ * Targets: HZ_TO_KHZ, HZ_TO_MHZ, HZ_TO_MHZ_ULL,
+ *          KHZ_TO_HZ, MHZ_TO_KHZ, KHZ_TO_MHZ, MHZ_TO_HZ_ULL
+ *
+ * Inputs: None
+ *
+ * Steps:
+ * 1) Invoke various hertz conversion macros with different input values.
+ * 2) Check and confirm if the conversion macro results in expected value.
+ *    Otherwise, return FAIL.
+ * 4) Return PASS.
+ *
+ * Output:
+ * The test returns PASS if all the invocations of various hertz conversion
+ * functions returns the results as expected. Otherwise, the test returns FAIL.
+ *
+ */
+int test_hertzconversion(struct unit_module *m, struct gk20a *g, void *args);
+
+#endif /* __UNIT_POSIX_UTILS_H__ */