[llvm] [DirectX] Add support to lower LLVM intrinsics ceil, cos, fabs, floor and smax to DXIL Ops. (PR #78767)

[David Peixotto via llvm-commits]

================
@@ -0,0 +1,69 @@
+; Make sure dxil operation function calls for abs are appropriately strength reduced for int and int64_t.
+; RUN: opt -S -dxil-strength-reduce < %s | FileCheck %s -check-prefix=TEST_SR
+
+; Make sure output of strength reduction pass is lowered to DXIL code as expected.
+; RUN: opt -S -dxil-strength-reduce -dxil-op-lower < %s | FileCheck %s -check-prefix=TEST_SR_OP_LOWER
+
+
+target datalayout = "e-m:e-p:32:32-i1:32-i8:8-i16:16-i32:32-i64:64-f16:16-f32:32-f64:64-n8:16:32:64"
+target triple = "dxil-unknown-shadermodel6.5-compute"
+
+@"?a@@3HA" = local_unnamed_addr global i32 0, align 4
----------------
dmpots wrote:

I find using the mangled names makes the tests harder to read. They are used throughout these tests cases for both function names and variables.

Using simpler names makes it easier to read I think.

https://github.com/llvm/llvm-project/pull/78767