[libc] add printf auto float conversion

This patch adds the final conversion to printf, %g. This is a floating
point conversion that selects automatically between the %e and %f
formats based on the precision requested and resulting exponent.
Additionally it trims trailing zeroes. With this done all that's left
for finishing printf is adding long double support to the decimal float
conversions.

Reviewed By: sivachandra

Differential Revision: https://reviews.llvm.org/D143006

libc/src/__support/float_to_string.h

@@ -89,6 +89,8 @@ LIBC_INLINE constexpr uint32_t length_for_num(const uint32_t idx,
 // Rewritten slightly we get:
 // floor(5^(-9i) * 2^(e + c_1 - 9i) + 1) % (10^9 * 2^c_1)
 
+// TODO: Fix long doubles (needs bigger table or alternate algorithm.)
+// Currently the table values are generated, which is very slow.
 template <size_t INT_SIZE>
 LIBC_INLINE constexpr cpp::UInt<MID_INT_SIZE>
 get_table_positive(int exponent, size_t i, const size_t constant) {

libc/src/stdio/printf_core/converter.cpp

@@ -52,9 +52,9 @@ int convert(Writer *writer, const FormatSection &to_conv) {
   case 'a':
   case 'A':
     return convert_float_hex_exp(writer, to_conv);
-    // case 'g':
-    // case 'G':
-    // return convert_float_mixed(writer, to_conv);
+  case 'g':
+  case 'G':
+    return convert_float_dec_auto(writer, to_conv);
 #endif // LLVM_LIBC_PRINTF_DISABLE_FLOAT
 #ifndef LLVM_LIBC_PRINTF_DISABLE_WRITE_INT
   case 'n':

libc/src/stdio/printf_core/converter_atlas.h

@@ -24,11 +24,11 @@
 
 #ifndef LLVM_LIBC_PRINTF_DISABLE_FLOAT
 // defines convert_float_decimal
-#include "src/stdio/printf_core/float_dec_converter.h"
 // defines convert_float_dec_exp
+// defines convert_float_dec_auto
+#include "src/stdio/printf_core/float_dec_converter.h"
 // defines convert_float_hex_exp
 #include "src/stdio/printf_core/float_hex_converter.h"
-// defines convert_float_mixed
 #endif // LLVM_LIBC_PRINTF_DISABLE_FLOAT
 
 #ifndef LLVM_LIBC_PRINTF_DISABLE_WRITE_INT