linux-kernel - Re: [PATCH v6] livepatch: Clear relocation targets on a module removal

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <alpine.LSU.2.21.2212091348090.18933@pobox.suse.cz>
Date:   Fri, 9 Dec 2022 13:49:32 +0100 (CET)
From:   Miroslav Benes <mbenes@...e.cz>
To:     Song Liu <song@...nel.org>
cc:     Petr Mladek <pmladek@...e.com>, live-patching@...r.kernel.org,
        linux-kernel@...r.kernel.org, jpoimboe@...nel.org,
        jikos@...nel.org, x86@...nel.org, joe.lawrence@...hat.com,
        linuxppc-dev@...ts.ozlabs.org, Josh Poimboeuf <jpoimboe@...hat.com>
Subject: Re: [PATCH v6] livepatch: Clear relocation targets on a module
 removal

> > > --- a/kernel/livepatch/core.c
> > > +++ b/kernel/livepatch/core.c
> > > @@ -316,6 +316,45 @@ int klp_apply_section_relocs(struct module *pmod, Elf_Shdr *sechdrs,
> > >       return apply_relocate_add(sechdrs, strtab, symndx, secndx, pmod);
> > >  }
> > >
> > > +static void klp_clear_object_relocations(struct module *pmod,
> > > +                                     struct klp_object *obj)
> > > +{
> > > +     int i, cnt;
> > > +     const char *objname, *secname;
> > > +     char sec_objname[MODULE_NAME_LEN];
> > > +     Elf_Shdr *sec;
> > > +
> > > +     objname = klp_is_module(obj) ? obj->name : "vmlinux";
> > > +
> > > +     /* For each klp relocation section */
> > > +     for (i = 1; i < pmod->klp_info->hdr.e_shnum; i++) {
> > > +             sec = pmod->klp_info->sechdrs + i;
> > > +             secname = pmod->klp_info->secstrings + sec->sh_name;
> > > +             if (!(sec->sh_flags & SHF_RELA_LIVEPATCH))
> > > +                     continue;
> > > +
> > > +             /*
> > > +              * Format: .klp.rela.sec_objname.section_name
> > > +              * See comment in klp_resolve_symbols() for an explanation
> > > +              * of the selected field width value.
> > > +              */
> > > +             secname = pmod->klp_info->secstrings + sec->sh_name;
> > > +             cnt = sscanf(secname, ".klp.rela.%55[^.]", sec_objname);
> > > +             if (cnt != 1) {
> > > +                     pr_err("section %s has an incorrectly formatted name\n",
> > > +                            secname);
> > > +                     continue;
> > > +             }
> 
> Actually, I think we don't need the cnt check here. Once it is removed,
> there isn't much duplicated logic.

Because it must have passed the check once already during 
klp_apply_section_relocs()? Yes, perhaps.

Miroslav