[clang] [CIR] [X86] Add support for vfmadd and vfmaddsub builtins. (PR #217326)

[Mohammed Zayer Alam via cfe-commits]

zayeralam wrote:

@andykaylor  Thank you for the detailed feedback and precise guidance. I deeply appreciate you taking the time to point out the build configuration ('-DCLANG_ENABLE_CIR_=ON') and the correct source pathing. I see where the error was regarding 'CIRGenBuiltin.cpp' vs target-specific handling in 'CIRGenBuiltinX86.cpp'. I'm updating the implementation to properly map the 'BI__builtin_ia32_vfmadd*' builtins  following 'TargetBuiltins/X86.cpp', and I'll move the tests to 'avx512f-builtins.c' using the proper intrinsic wrappers with CIR, LLVM IR and OGCG verification lines. I'll push the revised commit shortly. Thanks again for steering me in the right direction..!

https://github.com/llvm/llvm-project/pull/217326