Searched refs:reduced_max_exponent (Results 1 – 2 of 2) sorted by relevance
120 const uint32_t reduced_max_exponent = in EmitReducePrecisionFloat() local130 llvm::ConstantInt::get(int_type, reduced_max_exponent << 23)); in EmitReducePrecisionFloat()
2587 const uint32_t reduced_max_exponent = in HandleReducePrecision()2594 const bool x_overflows = x_exponent > (reduced_max_exponent << 23); in HandleReducePrecision()