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

[Chris Jackson via llvm-commits]

chrisjbris wrote:

So the i64 fneg case is failing. However, i64 fabs and fneg+fabs are succeeding. Quick analysis shows a combine somewhere adding an extra fneg, creating a double negative, which cancels out. I'll try to identify the offending combine now.

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