On Thu, 12 Mar 2020, Jakub Jelinek wrote:
> On Thu, Mar 12, 2020 at 09:20:08AM +0100, Richard Biener wrote:
> > On Thu, 12 Mar 2020, Jakub Jelinek wrote:
> >
> > > Hi!
> > >
> > > As the testcase shows, if DSE decides to head trim
> > > {mem{set,cpy,move},strncpy}
> > > and the call has lhs, it is incorrect to leave the lhs as is, because it
> > > will then point to the adjusted address (base + head_trim) instead of the
> > > original base.
> > > The following patch fixes that by dropping the lhs of the call and
> > > assigning
> > > lhs the original base in a following statement.
> > >
> > > Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk?
> >
> > OK, but I don't see where you need anything additional frmo gimplify.h?
>
> unshare_expr here:
> > > + gassign *newop = gimple_build_assign (lhs, unshare_expr (*where));
>
> While SSA_NAMEs obviously don't need to be unshared, if it is a gimple
> invariant, it might need to be.
Ah, ok.
Richard.
