[libc][math][C23] Cospif function should only return unsigned zeros (#98037)

Patches the cospif function, which should always return unsigned zeros
as specified in the standard. Also updates unit tests.

Author: HendrikHuebner

libc/src/math/generic/cospif.cpp

@@ -21,7 +21,6 @@ LLVM_LIBC_FUNCTION(float, cospif, (float x)) {
   using FPBits = typename fputil::FPBits<float>;
 
   FPBits xbits(x);
-  Sign xsign = xbits.sign();
   xbits.set_sign(Sign::POS);
 
   uint32_t x_abs = xbits.uintval();
@@ -86,7 +85,7 @@ LLVM_LIBC_FUNCTION(float, cospif, (float x)) {
   sincospif_eval(xd, sin_k, cos_k, sin_y, cosm1_y);
 
   if (LIBC_UNLIKELY(sin_y == 0 && cos_k == 0)) {
-    return FPBits::zero(xsign).get_val();
+    return 0.0f;
   }
 
   return static_cast<float>(fputil::multiply_add(

libc/test/src/math/cospif_test.cpp

@@ -38,8 +38,10 @@ TEST_F(LlvmLibcCospifTest, SpecialNumbers) {
 }
 
 TEST_F(LlvmLibcCospifTest, SpecificBitPatterns) {
-  constexpr int N = 36;
+  constexpr int N = 38;
   constexpr uint32_t INPUTS[N] = {
+      0x3f00'0000U, // x = 0.5
+      0x461d'd600U, // x = 10101.5
       0x3f06'0a92U, // x = pi/6
       0x3f3a'dc51U, // x = 0x1.75b8a2p-1f
       0x3f49'0fdbU, // x = pi/4
@@ -108,13 +110,12 @@ TEST_F(LlvmLibcCospifTest, SDCOMP_26094) {
   }
 }
 
-// sinpi(-n) = -0.0
-// sinpi(+n) = +0.0
+// sinpi(+n + 0.5) = sinpi(-n + 0.5) = +0.0
 TEST_F(LlvmLibcCospifTest, SignedZeros) {
   EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(100.5f));
-  EXPECT_FP_EQ(-0.0, LIBC_NAMESPACE::cospif(-100.5f));
+  EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(-100.5f));
   EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(45678.5f));
-  EXPECT_FP_EQ(-0.0, LIBC_NAMESPACE::cospif(-45678.5f));
+  EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(-45678.5f));
   EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(8000000.5f));
-  EXPECT_FP_EQ(-0.0, LIBC_NAMESPACE::cospif(-8000000.5f));
+  EXPECT_FP_EQ(0.0, LIBC_NAMESPACE::cospif(-8000000.5f));
 }