http://gcc.gnu.org/bugzilla/show_bug.cgi?id=48377
--- Comment #15 from Ira Rosen <irar at il dot ibm.com> 2011-04-07 09:00:43 UTC --- (In reply to comment #14) > Sure, but for types with user alignment smaller than their size, loop peeling > may often never be able to align it. > > This regressed with http://gcc.gnu.org/viewcvs?root=gcc&view=rev&rev=161797 This patch only allowed peeling for loads. So the problem existed for stores before this revision. > > What I mean is that for say unsigned int type with 32-bit alignment loop > peeling for valid code should always end up with correct alignment, but for > unsigned int with 8-bit alignment it can't. And e.g. i?86 long long is of the > same category, it has 64-bit size, but 32-bit alignment, so if you have a > pointer > to long long array, the whole array might be just 32-bit aligned, but not > 64-bit, > then any kind of peeling will still result in misalignment. The vectorizer calls builtin_vector_alignment_reachable in case of unknown alignment to verify that the access is naturally aligned. (i?86 doesn't implement this builtin and uses the default implementation). There is PR 42652 for a similar problem. Thanks, Ira
