Re: [PATCH] elf: Try not pointing empty PT_LOAD segment's offset past EOF

[Bálint Réczey <balint at balintreczey dot hu>]

Hi Alan,

Alan Modra <amodra@gmail.com> ezt írta (időpont: 2019. dec. 13., P, 12:36):
>
> On Fri, Dec 13, 2019 at 10:32:19AM +0100, Bálint Réczey wrote:
> > Hi Alan,
> >
> > Bálint Réczey <balint@balintreczey.hu> ezt írta (időpont: 2019. dec.
> > 11., Sze, 12:44):
> > >
> > > Alan Modra <amodra@gmail.com> ezt írta (időpont: 2019. dec. 11., Sze, 12:02):
> > > >
> > ...
> > > > "p->p_offset = off % maxpagesize" might be OK, I think.
> > >
> > > That's perfect, thanks. Attaching the updated patch.
> >
> > I have attached the updated patch to to bug, too.
> > Would you like me to perform any specific additional testing?
>
> I've run the binutils testsuite over multiple targets without seeing
> any problems so will commit the following variant.
>
>         PR 25237
>         * elf.c (assign_file_positions_for_load_sections): Attempt to
>         keep meaningless p_offset for PT_LOAD segments without file
>         contents within file size.
>
> diff --git a/bfd/elf.c b/bfd/elf.c
> index 1aa2603ee8..fd447fdb28 100644
> --- a/bfd/elf.c
> +++ b/bfd/elf.c
> @@ -5752,7 +5752,15 @@ assign_file_positions_for_load_sections (bfd *abfd,
>           || (p->p_type == PT_NOTE && bfd_get_format (abfd) == bfd_core))
>         {
>           if (!m->includes_filehdr && !m->includes_phdrs)
> -           p->p_offset = off;
> +           {
> +             p->p_offset = off;
> +             if (no_contents)
> +               /* Put meaningless p_offset for PT_LOAD segments
> +                  without file contents somewhere within the first
> +                  page, in an attempt to not point past EOF.  */
> +               p->p_offset = off % (p->p_align > maxpagesize
> +                                    ? p->p_align : maxpagesize);
> +           }
>           else
>             {
>               file_ptr adjust;

Thank you!

Cheers,
Balint