[libc][support][UInt] implement 128b math helpers

[nickdesaulniers]

Flush out the remaining UInt<128> support and add test coverage.

We could have used cpp::popcount in the implementation of UInt::has_single_bit, but
has_single_bit has a perhaps useful early return.

[llvmbot]

@llvm/pr-subscribers-libc

Author: Nick Desaulniers (nickdesaulniers)

Changes

We could have used this in the implementation of UInt::has_single_bit, but
has_single_bit has a perhaps useful early return.

---

Full diff: https://github.com/llvm/llvm-project/pull/86531.diff

2 Files Affected:

• (modified) libc/src/__support/UInt.h (+11)
• (modified) libc/test/src/__support/CPP/bit_test.cpp (+1-8)

diff --git a/libc/src/__support/UInt.h b/libc/src/__support/UInt.h
index df01e081e3c19e..b04e8d02ad4c16 100644
--- a/libc/src/__support/UInt.h
+++ b/libc/src/__support/UInt.h
@@ -1082,6 +1082,17 @@ bit_cast(const UInt<Bits> &from) {
   return cpp::bit_cast<To>(from.val);
 }
 
+// Specialization of cpp::popcount ('bit.h') for BigInt.
+template <typename T>
+[[nodiscard]] LIBC_INLINE constexpr cpp::enable_if_t<is_big_int_v<T>, int>
+popcount(T value) {
+  int bits = 0;
+  for (auto word : value.val)
+    if (word)
+      bits += popcount(word);
+  return bits;
+}
+
 // Specialization of cpp::has_single_bit ('bit.h') for BigInt.
 template <typename T>
 [[nodiscard]] LIBC_INLINE constexpr cpp::enable_if_t<is_big_int_v<T>, bool>
diff --git a/libc/test/src/__support/CPP/bit_test.cpp b/libc/test/src/__support/CPP/bit_test.cpp
index cee5b90c8f4bdb..875b47e6a1980e 100644
--- a/libc/test/src/__support/CPP/bit_test.cpp
+++ b/libc/test/src/__support/CPP/bit_test.cpp
@@ -15,13 +15,6 @@
 
 namespace LIBC_NAMESPACE::cpp {
 
-using UnsignedTypesNoBigInt = testing::TypeList<
-#if defined(LIBC_TYPES_HAS_INT128)
-    __uint128_t,
-#endif // LIBC_TYPES_HAS_INT128
-    unsigned char, unsigned short, unsigned int, unsigned long,
-    unsigned long long>;
-
 using UnsignedTypes = testing::TypeList<
 #if defined(LIBC_TYPES_HAS_INT128)
     __uint128_t,
@@ -228,7 +221,7 @@ TEST(LlvmLibcBitTest, Rotr) {
             rotr<uint64_t>(0x12345678deadbeefULL, -19));
 }
 
-TYPED_TEST(LlvmLibcBitTest, CountOnes, UnsignedTypesNoBigInt) {
+TYPED_TEST(LlvmLibcBitTest, CountOnes, UnsignedTypes) {
   EXPECT_EQ(popcount(T(0)), 0);
   for (int i = 0; i != cpp::numeric_limits<T>::digits; ++i)
     EXPECT_EQ(popcount<T>(cpp::numeric_limits<T>::max() >> i),

[github-actions]

✅ With the latest revision this PR passed the Python code formatter.

[github-actions]

✅ With the latest revision this PR passed the C/C++ code formatter.

[nickdesaulniers]

Ah, I should fix up the math_extras, too.

[nickdesaulniers]

Ah, I should fix up the math_extras, too.

fixed in 457a41d, PTAL

[nickdesaulniers]

hmm...I'm on the fence about removing this TODO.

llvm-project/libc/src/__support/CPP/bit.h

Line 264 in 5ef0954

// TODO: 128b specializations?

On one hand, we now have __builtin_popcountg in nascent versions of clang and GCC. We'd need a __has_builtin guard for these.

Actually, I'll add it to this PR.

[nickdesaulniers]

hmm...I'm on the fence about removing this TODO.

llvm-project/libc/src/__support/CPP/bit.h

Line 264 in 5ef0954

// TODO: 128b specializations?

On one hand, we now have __builtin_popcountg in nascent versions of clang and GCC. We'd need a __has_builtin guard for these.

Actually, I'll add it to this PR.

Done in b1e40ed PTAL

[nickdesaulniers]

cc @overmighty

[nickdesaulniers]

Ah, should this be using LIBC_HAS_BUILTIN?

[nickdesaulniers]

gcc-10+ supports __has_builtin. While wrapping the use of __has_builtin used to be necessary to support older compilers, this is no longer necessary since we only support gcc-12+ . https://libc.llvm.org/compiler_support.html I'll leave this out, and file a bug about removing it.

[nickdesaulniers]

Filed #86546