[libc] Fix final conversion warnings

[michaelrj]

Summary

This patch fixes the floating point conversion warnings found with
-Wconversion and -Wno-sign-conversion. These were the last warnings
I found, meaning that once this lands https://reviews.llvm.org/D156630
should be unblocked.

[mcgrathr]

libc/src/math/generic/tanhf.cpp

59

The literals here use the f suffix to have float type, but then the variable is double.
Should the variable just be float? And I wonder too why it's not constexpr.

[lntue]

libc/src/math/generic/tanhf.cpp

59

+1, this should be declared as constexpr float

[michaelrj]

update tanhf signs array

[mcgrathr]

lgtm

libc/src/math/generic/exp2f.cpp

if (x_abs <= 0x3d00'0000) {
           0x1.3b2ab6fc92f5dp-7, 0x1.5d897cfe27125p-10, 0x1.43090e61e6af1p-13};
       double xd = static_cast<double>(x);
       double xsq = xd * xd;
       double c0 = fputil::multiply_add(xd, COEFFS[1], COEFFS[0]);
       double c1 = fputil::multiply_add(xd, COEFFS[3], COEFFS[2]);
       double c2 = fputil::multiply_add(xd, COEFFS[5], COEFFS[4]);
       double p = fputil::polyeval(xsq, c0, c1, c2);
       double r = fputil::multiply_add(p, xd, 1.0);
-      return r;
+      return static_cast<float>(r);
     }
 
     // x >= 128
     if (!xbits.get_sign()) {
       // x is finite
       if (x_u < 0x7f80'0000U) {
         int rounding = fputil::quick_get_round();
         if (rounding == FE_DOWNWARD || rounding == FE_TOWARDZERO)

libc/src/math/generic/tanhf.cpp

if (x_abs <= 0x3da0'0000U) {
 
       return static_cast<float>(fputil::multiply_add(xdbl, pe, xdbl));
     }
 
     // |x| >= 15
     if (LIBC_UNLIKELY(xbits.is_nan()))
       return x + 1.0f; // sNaN to qNaN + signal
 
-    const double SIGNS[2][2] = {{1.0f, -0x1.0p-25f}, {-1.0f, 0x1.0p-25f}};
+    constexpr float SIGNS[2][2] = {{1.0f, -0x1.0p-25f}, {-1.0f, 0x1.0p-25f}};
 
     bool sign = xbits.get_sign();
     int idx = static_cast<int>(sign);
 
     if (LIBC_UNLIKELY(xbits.is_inf()))
       return SIGNS[idx][0];
 
     return SIGNS[idx][0] + SIGNS[idx][1];

libc/src/time/difftime.cpp

 //===-- Implementation of difftime function -------------------------------===//
 //
 // Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
 // See https://llvm.org/LICENSE.txt for license information.
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
 
 #include "src/time/difftime.h"
 #include "src/__support/common.h"
 
 namespace __llvm_libc {
 
 LLVM_LIBC_FUNCTION(double, difftime, (time_t end, time_t beginning)) {
-  return end - beginning;
+  return static_cast<double>(end - beginning);
 }
 
 } // namespace __llvm_libc