[LLVMdev] new set of superoptimizer results

David Chisnall David.Chisnall at cl.cam.ac.uk
Thu Nov 27 08:19:16 PST 2014


On 25 Nov 2014, at 17:58, Reid Kleckner <rnk at google.com> wrote:

> Cool! Looks like we do lots of provably unnecessary alignment checks. :)

We've hit this with clang and a target that didn't support unaligned accesses.  Almost all loads and stores emitted by clang have align 1, rather than the alignment that the C language mandates.  There's even a PowerPC test that fails when you fix this in clang, as it depends on the alignment being 1.

David





More information about the llvm-dev mailing list