[libcxx-commits] [libcxx] 586952f - Optimize std::midpoint for integers

[Jorg Brown via libcxx-commits]

Author: Jorg Brown
Date: 2019-11-04T19:00:23-08:00
New Revision: 586952f4cefd809b7becd16c6d1e751ea923adfd

URL: https://github.com/llvm/llvm-project/commit/586952f4cefd809b7becd16c6d1e751ea923adfd
DIFF: https://github.com/llvm/llvm-project/commit/586952f4cefd809b7becd16c6d1e751ea923adfd.diff

LOG: Optimize std::midpoint for integers

Same idea as the current algorithm, that is, add (half of the difference between a and b) to a.

But we use a different technique for computing the difference: we compute b - a into a pair of integers that are named "sign_bit" and "diff". We have to use a pair because subtracting two 32-bit integers produces a 33-bit result.

Computing half of that is a simple matter of shifting diff right by 1, and adding sign_bit shifted left by 31. llvm knows how to do that with one instruction: shld.

The only tricky part is that if the difference is odd and negative, then shifting it by one isn't the same as dividing it by two - shifting a negative one produces a negative one, for example. So there's one more adjustment: if the sign bit and the low bit of diff are one, we add one.

For a demonstration of the codegen difference, see https://godbolt.org/z/7ar3K9 , which also has a built-in test.

Differential Revision: https://reviews.llvm.org/D69459

Added: 
    

Modified: 
    libcxx/include/numeric

Removed: 
    


################################################################################
diff  --git a/libcxx/include/numeric b/libcxx/include/numeric
index 854bbfad65ac..11a5641bb3b0 100644
--- a/libcxx/include/numeric
+++ b/libcxx/include/numeric
@@ -532,17 +532,14 @@ midpoint(_Tp __a, _Tp __b) noexcept
 _LIBCPP_DISABLE_UBSAN_UNSIGNED_INTEGER_CHECK
 {
     using _Up = std::make_unsigned_t<_Tp>;
+    constexpr _Up __bitshift = std::numeric_limits<_Up>::digits - 1;
 
-    int __sign = 1;
-    _Up __m = __a;
-    _Up __M = __b;
-    if (__a > __b)
-    {
-        __sign = -1;
-        __m = __b;
-        __M = __a;
-    }
-     return __a + __sign * _Tp(_Up(__M-__m) >> 1);
+    _Up __
diff  = _Up(__b) - _Up(__a);
+    _Up __sign_bit = __b < __a;
+
+    _Up __half_
diff  = (__
diff  / 2) + (__sign_bit << __bitshift) + (__sign_bit & __
diff );
+
+    return __a + __half_
diff ;
 }