[llvm] [AMDGPU][SDAG] Support source modifiers on select integer operands (PR #147325)

[Chris Jackson via llvm-commits]

chrisjbris wrote:

I'm currently attempting move a simplified version of the selectcombine implemented into the TI selectcombine, to take advantage of the combiner splitting the i64 op for us. This will require careful use of  shouldFoldSelectWithIdentityConstant() from TargetLowering.h.

https://github.com/llvm/llvm-project/pull/147325