On December 30, 2019 3:15:14 PM GMT+01:00, Richard Sandiford
wrote:
>This patch fixes an awkward corner case in which:
>
>(a) we apply if-conversion to a loop;
>
>(b) the original scalar loop doesn't have a vdef, and thus doesn't
>need a virtual phi;
>
>(c) the vectorised main loop does need
This patch fixes an awkward corner case in which:
(a) we apply if-conversion to a loop;
(b) the original scalar loop doesn't have a vdef, and thus doesn't
need a virtual phi;
(c) the vectorised main loop does need a vdef and a virtual phi (see below);
(d) we also vectorise the epilogue; and
