On Fri 2023-01-13 11:55:25, Song Liu wrote:
> On Fri, Jan 13, 2023 at 1:18 AM Miroslav Benes <mbenes@xxxxxxx> wrote:
> >
> > Hi,
> >
> > On Fri, 6 Jan 2023, Song Liu wrote:
> >
> > > From: Miroslav Benes <mbenes@xxxxxxx>
> > >
> > > 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.
> > >
> > > On ppc64le, we have a similar issue:
> > >
> > > module_64: livepatch_nfsd: Expected nop after call, got e8410018 at e_show+0x60/0x548 [livepatch_nfsd]
> > > livepatch: failed to initialize patch 'livepatch_nfsd' for module 'nfsd' (-8)
> > > livepatch: patch 'livepatch_nfsd' failed for module 'nfsd', refusing to load module 'nfsd'
> > >
> > > 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.
> > >
> > > Signed-off-by: Miroslav Benes <mbenes@xxxxxxx>
> > > Signed-off-by: Song Liu <song@xxxxxxxxxx>
> > > Reported-by: Josh Poimboeuf <jpoimboe@xxxxxxxxxx>
> >
> > I would be fine if you just claimed the authorship (and include my
> > Originally-by: tag for example), because you have reworked it quite a lot
> > since my first attempts.
>
> I am ok with either way. Or maybe with
>
> Co-developed-by: Song Liu <song@xxxxxxxxxx>
>
> >
> > > +int klp_apply_section_relocs(struct module *pmod, Elf_Shdr *sechdrs,
> > > + const char *shstrtab, const char *strtab,
> > > + unsigned int symndx, unsigned int secndx,
> > > + const char *objname)
> > > +{
> > > + return klp_write_section_relocs(pmod, sechdrs, shstrtab, strtab, symndx,
> > > + secndx, objname, true);
> > > }
I think that I proposed this wrapper :-)
> > Is this redirection needed somewhere? You could just replace
> > klp_apply_section_relocs() with klp_write_section_relocs() in
> > include/linux/livepatch.h and kernel/module/main.c.
> >
> > It may be cleaned up later.
>
> It might be a good practice to keep _write_ static in this file, and
> only expose _apply_ (maybe also _clear_ in the future)?
And I think that this was the reason. Also it looks better in
kernel/module/main.c in apply_relocations() that calls few more
*_apply_*reloc*() functions.
The idea is that functions with the same naming pattern do
the same operation. Also it is supposed to hide the true/false
parameter and self-explain the meaning by the function name.
> I don't have a strong preference either way.
I would prefer to keep the wrapper. But I do not resist on it :-)
Best Regards,
Petr
