kieroxide wrote: @davemgreen vshll_n_u32(X, 32) produces the same output as before since llvm allowed 0..(elt*2)-1. We can see already existing tests for this exact case in test/CodeGen/AArch64/neon-misc.c. However, the IR to asm testing is lacking and I will add this to the PR.
https://github.com/llvm/llvm-project/pull/212459 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
