[llvm] [VectorCombine] Use TCK_CodeSize for size-optimized functions (PR #202207)

[via llvm-commits]

================
@@ -0,0 +1,48 @@
+; NOTE: Assertions have been autogenerated by utils/update_test_checks.py UTC_ARGS: --version 4
+; RUN: opt < %s -passes=vector-combine -S -mtriple=x86_64-- -mattr=+avx2 | FileCheck %s
----------------
Michael-Chen-NJU wrote:

Thanks! I added coverage for x86-64 through x86-64-v4. They currently produce identical checked IR, so the RUN lines share the same check prefix.

https://github.com/llvm/llvm-project/pull/202207