fix: avoid macro redefinition by namespacing helpers
Rename is_aligned and align_up to libtl_* variants and move header to
private include path. Prevents clashes with external projects that
define similar macros (e.g., TF-A).
Change-Id: I7e13efa55fe120dccc68d2a941dea7cdd1f005a6
Signed-off-by: Harrison Mutai <harrison.mutai@arm.com>
diff --git a/include/private/math_utils.h b/include/private/math_utils.h
new file mode 100644
index 0000000..ba1bb9c
--- /dev/null
+++ b/include/private/math_utils.h
@@ -0,0 +1,86 @@
+/*
+ * Copyright The Transfer List Library Contributors
+ *
+ * SPDX-License-Identifier: MIT OR GPL-2.0-or-later
+ */
+
+#ifndef MATH_UTILS_H
+#define MATH_UTILS_H
+
+#include <stdint.h>
+
+/**
+ * @brief Rounds up a given value to the nearest multiple of a boundary.
+ *
+ * @param value The value to round up.
+ * @param boundary The alignment boundary (must be a power of two).
+ * @return The smallest multiple of `boundary` that is greater than or equal to `value`.
+ */
+static inline uintptr_t libtl_align_up(uintptr_t value, uintptr_t boundary)
+{
+ return (value + boundary - 1) & ~(boundary - 1);
+}
+
+/**
+ * @brief Checks whether `value` is aligned to the specified `boundary`.
+ *
+ * @param value The value to check.
+ * @param boundary The alignment boundary (should be a power of two for efficiency).
+ * @return `1` if `value` is aligned, `0` otherwise.
+ */
+static inline int libtl_is_aligned(uintptr_t value, uintptr_t boundary)
+{
+ return (value % boundary) == 0;
+}
+
+/**
+ * @brief Safely adds `a` and `b`, detecting overflow.
+ *
+ * @param a First operand.
+ * @param b Second operand.
+ * @param res Pointer to store the result if no overflow occurs.
+ * @return `1` if overflow occurs, `0` otherwise.
+ */
+#define libtl_add_overflow(a, b, res) __builtin_add_overflow((a), (b), (res))
+
+/**
+ * @brief Rounds up `v` to the nearest multiple of `size`, detecting overflow.
+ *
+ * @param v The value to round up.
+ * @param size The alignment boundary (must be a power of two).
+ * @param res Pointer to store the rounded-up result.
+ * @return `1` if an overflow occurs, `0` otherwise.
+ */
+#define libtl_round_up_overflow(v, size, res) \
+ (__extension__({ \
+ typeof(res) __res = res; \
+ typeof(*(__res)) __roundup_tmp = 0; \
+ typeof(v) __roundup_mask = (typeof(v))(size) - 1; \
+ \
+ libtl_add_overflow((v), __roundup_mask, &__roundup_tmp) ? \
+ 1 : \
+ (void)(*(__res) = __roundup_tmp & ~__roundup_mask), \
+ 0; \
+ }))
+
+/**
+ * @brief Adds `a` and `b`, then rounds up the result to the nearest multiple
+ * of `size`, detecting overflow.
+ *
+ * @param a First operand for addition.
+ * @param b Second operand for addition.
+ * @param size The alignment boundary (must be positive).
+ * @param res Pointer to store the final rounded result.
+ * @return `1` if an overflow occurs during addition or rounding, `0` otherwise.
+ */
+#define libtl_add_with_round_up_overflow(a, b, size, res) \
+ (__extension__({ \
+ typeof(a) __a = (a); \
+ typeof(__a) __add_res = 0; \
+ \
+ libtl_add_overflow((__a), (b), &__add_res) ? 1 : \
+ libtl_round_up_overflow(__add_res, (size), (res)) ? 1 : \
+ 0; \
+ }))
+
+#endif /* MATH_UTILS_H */