MIME-Version: 1.0
References: <20230121004945.697003-1-song@kernel.org> <20230121004945.697003-2-song@kernel.org>
 <Y8/N7zMLUnMh259N@alley> <20230124172313.extovg6ig7dimpgb@treble>
In-Reply-To: <20230124172313.extovg6ig7dimpgb@treble>
From:   Song Liu <song@kernel.org>
Date:   Tue, 24 Jan 2023 09:30:41 -0800
Message-ID: <CAPhsuW7Wo1amPn2fdZHB9MAZB=iUMTQW13gu0Pj=0-N9EkRVvQ@mail.gmail.com>
Subject: Re: [PATCH v10 2/2] livepatch,x86: Clear relocation targets on a
 module removal
To:     Josh Poimboeuf <jpoimboe@kernel.org>
Cc:     Petr Mladek <pmladek@suse.com>, linux-kernel@vger.kernel.org,
        linux-modules@vger.kernel.org, live-patching@vger.kernel.org,
        x86@kernel.org, Josh Poimboeuf <jpoimboe@redhat.com>,
        Miroslav Benes <mbenes@suse.cz>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Jan 24, 2023 at 9:23 AM Josh Poimboeuf <jpoimboe@kernel.org> wrote:
>
> On Tue, Jan 24, 2023 at 01:24:15PM +0100, Petr Mladek wrote:
> > On Fri 2023-01-20 16:49:45, Song Liu wrote:
> > > Josh reported a bug:
> > >
> > >   When the object to be patched is a module, and that module is
> > >   rmmod'ed and reloaded, it fails to load with:
> > >
> > >   module: x86/modules: Skipping invalid relocation target, existing value is nonzero for type 2, loc 00000000ba0302e9, val ffffffffa03e293c
> > >   livepatch: failed to initialize patch 'livepatch_nfsd' for module 'nfsd' (-8)
> > >   livepatch: patch 'livepatch_nfsd' failed for module 'nfsd', refusing to load module 'nfsd'
> > >
> > >   The livepatch module has a relocation which references a symbol
> > >   in the _previous_ loading of nfsd. When apply_relocate_add()
> > >   tries to replace the old relocation with a new one, it sees that
> > >   the previous one is nonzero and it errors out.
> > >
> > > He also proposed three different solutions. We could remove the error
> > > check in apply_relocate_add() introduced by commit eda9cec4c9a1
> > > ("x86/module: Detect and skip invalid relocations"). However the check
> > > is useful for detecting corrupted modules.
> > >
> > > We could also deny the patched modules to be removed. If it proved to be
> > > a major drawback for users, we could still implement a different
> > > approach. The solution would also complicate the existing code a lot.
> > >
> > > We thus decided to reverse the relocation patching (clear all relocation
> > > targets on x86_64). The solution is not
> > > universal and is too much arch-specific, but it may prove to be simpler
> > > in the end.
> > >
> > > Reported-by: Josh Poimboeuf <jpoimboe@redhat.com>
> > > Originally-by: Miroslav Benes <mbenes@suse.cz>
> > > Signed-off-by: Song Liu <song@kernel.org>
> > > Acked-by: Miroslav Benes <mbenes@suse.cz>
> > >
> > > --- a/arch/x86/kernel/module.c
> > > +++ b/arch/x86/kernel/module.c
> > > @@ -129,22 +129,27 @@ int apply_relocate(Elf32_Shdr *sechdrs,
> > >     return 0;
> > >  }
> > >  #else /*X86_64*/
> > > -static int __apply_relocate_add(Elf64_Shdr *sechdrs,
> > > +static int __write_relocate_add(Elf64_Shdr *sechdrs,
> > >                const char *strtab,
> > >                unsigned int symindex,
> > >                unsigned int relsec,
> > >                struct module *me,
> > > -              void *(*write)(void *dest, const void *src, size_t len))
> > > +              void *(*write)(void *dest, const void *src, size_t len),
> > > +              bool apply)
> > >  {
> > >     unsigned int i;
> > >     Elf64_Rela *rel = (void *)sechdrs[relsec].sh_addr;
> > >     Elf64_Sym *sym;
> > >     void *loc;
> > >     u64 val;
> > > +   u64 zero = 0ULL;
> > >
> > > -   DEBUGP("Applying relocate section %u to %u\n",
> > > +   DEBUGP("%s relocate section %u to %u\n",
> > > +          apply ? "Applying" : "Clearing",
> > >            relsec, sechdrs[relsec].sh_info);
> > >     for (i = 0; i < sechdrs[relsec].sh_size / sizeof(*rel); i++) {
> > > +           int size = 0;
> >
> > The value 0 should never be used. It is better to do not initialize
> > it at all so that the compiler would warn when the variable might be
> > used uninitialized.
>
> Yes.  Also it can be unsigned, i.e. size_t.

Will fix this in the next version.

I guess we still need an Acked-by from x86 maintainers.

Thanks,
Song