Re: [PATCHv2 bpf-next 4/8] bpf: Take module reference on kprobe_multi link

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



On Mon, Oct 24, 2022 at 1:16 AM Jiri Olsa <olsajiri@xxxxxxxxx> wrote:
>
> On Fri, Oct 21, 2022 at 03:02:30PM -0700, Andrii Nakryiko wrote:
>
> SNIP
>
> > > +       if (err) {
> > > +               kprobe_multi_put_modules(args.mods, args.mods_cnt);
> > > +               kfree(args.mods);
> > > +               return err;
> > > +       }
> > > +
> > > +       /* or number of modules found if everything is ok. */
> > > +       *mods = args.mods;
> > > +       return args.mods_cnt;
> > > +}
> > > +
> > >  int bpf_kprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *prog)
> > >  {
> > >         struct bpf_kprobe_multi_link *link = NULL;
> > > @@ -2773,10 +2850,25 @@ int bpf_kprobe_multi_link_attach(const union bpf_attr *attr, struct bpf_prog *pr
> > >                        bpf_kprobe_multi_cookie_cmp,
> > >                        bpf_kprobe_multi_cookie_swap,
> > >                        link);
> > > +       } else {
> > > +               /*
> > > +                * We need to sort addrs array even if there are no cookies
> > > +                * provided, to allow bsearch in get_modules_for_addrs.
> > > +                */
> > > +               sort(addrs, cnt, sizeof(*addrs),
> > > +                      bpf_kprobe_multi_addrs_cmp, NULL);
> > > +       }
> > > +
> > > +       err = get_modules_for_addrs(&link->mods, addrs, cnt);
> > > +       if (err < 0) {
> > > +               bpf_link_cleanup(&link_primer);
> > > +               return err;
> > >         }
> > > +       link->mods_cnt = err;
> > >
> > >         err = register_fprobe_ips(&link->fp, addrs, cnt);
> > >         if (err) {
> > > +               kprobe_multi_put_modules(link->mods, link->mods_cnt);
> >
> > I don't think bpf_link_cleanup() will free link->mods, you have to do
> > it explicitly here
>
> hum, so bpf_link_cleanup sets link->prog to NULL so bpf_link_free
> won't call link->ops->release, but will call link->ops->dealloc,
> so it should be fine AFAICS

oh, I completely forgot that bpf_link_free() will be called eventually
due to fput(primer->file);

so yes, you are right!

Please add my ack for next version, this was the only (non-)issue I found.

Acked-by: Andrii Nakryiko <andrii@xxxxxxxxxx>

>
> jirka
>
> >
> > >                 bpf_link_cleanup(&link_primer);
> > >                 return err;
> > >         }
> > > --
> > > 2.37.3
> > >



[Index of Archives]     [Linux Samsung SoC]     [Linux Rockchip SoC]     [Linux Actions SoC]     [Linux for Synopsys ARC Processors]     [Linux NFS]     [Linux NILFS]     [Linux USB Devel]     [Video for Linux]     [Linux Audio Users]     [Yosemite News]     [Linux Kernel]     [Linux SCSI]


  Powered by Linux