Re: [PATCH v2] module: harden ELF info handling

[Frank van der Linden]

On Wed, Jan 13, 2021 at 03:48:42PM +0100, Jessica Yu wrote:
> 
> 
[...]
> > +                      err = validate_section_offset(info, shdr);
> > +                      if (err < 0) {
> > +                              pr_err("Invalid ELF section in module (section %u type %u)\n",
> > +                                      i, shdr->sh_type);
> > +                              return err;
> > +                      }
> > +
> > +                      if (shdr->sh_flags & SHF_ALLOC) {
> > +                              if (shdr->sh_name >= strhdr->sh_size) {
> > +                                      pr_err("Invalid ELF section name in module (section num %u type %u)\n",
> 
> Small nit: Maybe remove "num", to be consistent with the other pr_err() above.

Sure, will do.

[...]
> > diff --git a/kernel/module_signing.c b/kernel/module_signing.c
> > index 9d9fc678c91d..9a057c5d1d4d 100644
> > --- a/kernel/module_signing.c
> > +++ b/kernel/module_signing.c
> > @@ -30,7 +30,7 @@ int mod_verify_sig(const void *mod, struct load_info *info)
> > 
> >       memcpy(&ms, mod + (modlen - sizeof(ms)), sizeof(ms));
> > 
> > -      ret = mod_check_sig(&ms, modlen, info->name);
> > +      ret = mod_check_sig(&ms, modlen, info->name ?: "module");
> 
> Since info->name is not expected to be valid anymore, as we're back to
> calling mod_sig_check() first thing, perhaps just stick with
> "module"? Or was there another reason for checking info->name here?

It's mainly "what if mod_check_sig() is called differently in the future?",
but you are of course right: info->name will be NULL here. I'll change it
to just "module".

Thanks,

- Frank