Comment fixup

Author: alsepkow

tools/clang/unittests/HLSLExec/LongVectors.cpp

@@ -888,7 +888,8 @@ template <typename T, OpType OP> struct TrigonometricValidation {
   ValidationConfig ValidationConfig = ValidationConfig::Epsilon(0.0008f);
 };
 
-// Half precision trig functions have a larger tolerance. Note that the D3D spec
+// Half precision trig functions have a larger tolerance due to their lower
+// precision. Note that the D3D spec
 // does not mention half precision trig functions.
 template <OpType OP> struct TrigonometricValidation<HLSLHalf_t, OP> {
   ValidationConfig ValidationConfig = ValidationConfig::Epsilon(0.003f);