[llvm] [AMDGPU][True16][CodeGen] uaddsat/usubsat true16 selection in gisel (PR #128233)

[Brox Chen via llvm-commits]

================
@@ -1800,36 +1806,37 @@ AMDGPULegalizerInfo::AMDGPULegalizerInfo(const GCNSubtarget &ST_,
                   IdxTy.getSizeInBits() == 32 &&
                   isLegalVecType;
         })
-      .bitcastIf(all(sizeIsMultipleOf32(VecTypeIdx), scalarOrEltNarrowerThan(VecTypeIdx, 32)),
----------------
broxigarchen wrote:

some of the changes here are pure clang-format change

https://github.com/llvm/llvm-project/pull/128233