[libc-commits] [libc] [libc][math] Improve fmul performance by using double-double arithmetic. (PR #107517)

[via libc-commits]

================
@@ -5,16 +5,119 @@
 // SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
 //
 //===----------------------------------------------------------------------===//
-
 #include "src/math/fmul.h"
+#include "hdr/errno_macros.h"
+#include "hdr/fenv_macros.h"
+#include "src/__support/FPUtil/double_double.h"
 #include "src/__support/FPUtil/generic/mul.h"
 #include "src/__support/common.h"
 #include "src/__support/macros/config.h"
+#include <iostream>
 
 namespace LIBC_NAMESPACE_DECL {
 
 LLVM_LIBC_FUNCTION(float, fmul, (double x, double y)) {
+
+#ifndef LIBC_TARGET_CPU_HAS_FMA
   return fputil::generic::mul<float>(x, y);
-}
+#else
+  fputil::DoubleDouble prod = fputil::exact_mult(x, y);
+  float prod_hif = static_cast<float>(prod.hi);
+  fputil::FPBits<float> hif_bits(prod_hif);
+  using OutFPBits = fputil::FPBits<float>;
+  using OutStorageType = typename OutFPBits::StorageType;
+  using InFPBits = fputil::FPBits<double>;
+  using InStorageType = typename InFPBits::StorageType;
----------------
lntue wrote:

Remove these, move `DoubleBits` and `FloatBits` definition here, and use them for `x_bits`, `y_bits`.

https://github.com/llvm/llvm-project/pull/107517