[AArch64][GlobalISel] Improve codegen for G_VECREDUCE_{SMIN,SMAX,UMIN,UMAX} for odd-sized vectors#2

dc03-work wants to merge 1 commit intousers/spr/dc03-work/main.aarch64globalisel-improve-codegen-for-g_vecreduce_sminsmaxuminumax-for-odd-sized-vectorsdc03-work/llvm-project:users/spr/dc03-work/main.aarch64globalisel-improve-codegen-for-g_vecreduce_sminsmaxuminumax-for-odd-sized-vectorsfrom

users/spr/dc03-work/aarch64globalisel-improve-codegen-for-g_vecreduce_sminsmaxuminumax-for-odd-sized-vectorsdc03-work/llvm-project:users/spr/dc03-work/aarch64globalisel-improve-codegen-for-g_vecreduce_sminsmaxuminumax-for-odd-sized-vectorsCopy head branch name to clipboard