Justin Bogner
4ac26f45fa
We've improved these error messages in #199217, update the tests for the DirectX backend.
11 lines
429 B
LLVM
11 lines
429 B
LLVM
; RUN: not opt -S -dxil-op-lower -mtriple=dxil-pc-shadermodel6.3-library %s 2>&1 | FileCheck %s
|
|
|
|
; DXIL operation firstbitshigh does not support double overload type
|
|
; CHECK: error: intrinsic argument 0 type (overload type 0) expected any integer or integer vector, but got double
|
|
|
|
define noundef double @firstbitshigh_double(double noundef %a) {
|
|
entry:
|
|
%1 = call double @llvm.dx.firstbitshigh.f64(double %a)
|
|
ret double %1
|
|
}
|