This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH][RFC] Fix PR26421, improve find_used_portions
On Wed, 22 Feb 2006, Daniel Berlin wrote:
> On Wed, 2006-02-22 at 22:47 +0100, Richard Guenther wrote:
> > On Wed, 22 Feb 2006, Richard Guenther wrote:
> >
> > > It seems this patch causes gcc.dg/tree-ssa/stdarg-5.c to ICE during
> > > cplxlower in verify_ssa (thus RFC):
> > >
> > > /abuild/rguenther/gcc/gcc/testsuite/gcc.dg/tree-ssa/stdarg-5.c: In
> > > function ?f7?:
> > > /abuild/rguenther/gcc/gcc/testsuite/gcc.dg/tree-ssa/stdarg-5.c:108:
> > > internal compiler error: tree check: expected ssa_name, have
> > > name_memory_tag in verify_ssa, at tree-ssa.c:735
> > >
> > > we don't like
> > >
> > > addr.44_7 = &va_arg_tmp.41;
> > > # VUSE <SFT.154_105>;
> > > D.2281_132 = REALPART_EXPR <*addr.44_7>;
> > > # VUSE <NMT.168>;
> > > D.2282_133 = IMAGPART_EXPR <*addr.44_7>;
> >
> > Actually we start from (before cplxlower)
> >
> > addr.44_7 = &va_arg_tmp.41;
> > # VUSE <SFT.154_105>;
> > D.2089_8 = *addr.44_7;
> >
> > and lowering introduces an explicit use of IMAGPART_EXPR. As we
> > don't treat INDIRECT_REF (var) as explicit use.
>
> Of course, we can't treat it as an explicit use, because that would
> require knowing what it points to, which we don't have for
> find_used_portions.
Hmm right.
> The statement
>
> > addr.44_7 = &va_arg_tmp.41;
>
> Should have caused it to mark the entire variable as used, no?
Yes, but only implicitly (though I don't remember what was the
difference later in SFT creation for implicit/explicit uses).
Richard.