On Wed, Sep 16, 2020 at 2:21 PM Toke Høiland-Jørgensen <toke@xxxxxxxxxx> wrote:
>
> Andrii Nakryiko <andrii.nakryiko@xxxxxxxxx> writes:
>
> > On Wed, Sep 16, 2020 at 1:37 PM Andrii Nakryiko
> > <andrii.nakryiko@xxxxxxxxx> wrote:
> >>
> >> On Tue, Sep 15, 2020 at 5:50 PM Toke Høiland-Jørgensen <toke@xxxxxxxxxx> wrote:
> >> >
> >> > From: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
> >> >
> >> > This adds support for supplying a target btf ID for the bpf_link_create()
> >> > operation, and adds a new bpf_program__attach_freplace() high-level API for
> >> > attaching freplace functions with a target.
> >> >
> >> > Signed-off-by: Toke Høiland-Jørgensen <toke@xxxxxxxxxx>
> >> > ---
> >> > tools/lib/bpf/bpf.c | 1 +
> >> > tools/lib/bpf/bpf.h | 3 ++-
> >> > tools/lib/bpf/libbpf.c | 24 ++++++++++++++++++------
> >> > tools/lib/bpf/libbpf.h | 3 +++
> >> > tools/lib/bpf/libbpf.map | 1 +
> >> > 5 files changed, 25 insertions(+), 7 deletions(-)
> >> >
> >> > diff --git a/tools/lib/bpf/bpf.c b/tools/lib/bpf/bpf.c
> >> > index 82b983ff6569..e928456c0dd6 100644
> >> > --- a/tools/lib/bpf/bpf.c
> >> > +++ b/tools/lib/bpf/bpf.c
> >> > @@ -599,6 +599,7 @@ int bpf_link_create(int prog_fd, int target_fd,
> >> > attr.link_create.iter_info =
> >> > ptr_to_u64(OPTS_GET(opts, iter_info, (void *)0));
> >> > attr.link_create.iter_info_len = OPTS_GET(opts, iter_info_len, 0);
> >> > + attr.link_create.target_btf_id = OPTS_GET(opts, target_btf_id, 0);
> >> >
> >> > return sys_bpf(BPF_LINK_CREATE, &attr, sizeof(attr));
> >> > }
> >> > diff --git a/tools/lib/bpf/bpf.h b/tools/lib/bpf/bpf.h
> >> > index 015d13f25fcc..f8dbf666b62b 100644
> >> > --- a/tools/lib/bpf/bpf.h
> >> > +++ b/tools/lib/bpf/bpf.h
> >> > @@ -174,8 +174,9 @@ struct bpf_link_create_opts {
> >> > __u32 flags;
> >> > union bpf_iter_link_info *iter_info;
> >> > __u32 iter_info_len;
> >> > + __u32 target_btf_id;
> >> > };
> >> > -#define bpf_link_create_opts__last_field iter_info_len
> >> > +#define bpf_link_create_opts__last_field target_btf_id
> >> >
> >> > LIBBPF_API int bpf_link_create(int prog_fd, int target_fd,
> >> > enum bpf_attach_type attach_type,
> >> > diff --git a/tools/lib/bpf/libbpf.c b/tools/lib/bpf/libbpf.c
> >> > index 550950eb1860..165131c73f40 100644
> >> > --- a/tools/lib/bpf/libbpf.c
> >> > +++ b/tools/lib/bpf/libbpf.c
> >> > @@ -9322,12 +9322,14 @@ static struct bpf_link *attach_iter(const struct bpf_sec_def *sec,
> >> >
> >> > static struct bpf_link *
> >> > bpf_program__attach_fd(struct bpf_program *prog, int target_fd,
> >> > - const char *target_name)
> >> > + int target_btf_id, const char *target_name)
> >> > {
> >> > enum bpf_attach_type attach_type;
> >> > char errmsg[STRERR_BUFSIZE];
> >> > struct bpf_link *link;
> >> > int prog_fd, link_fd;
> >> > + DECLARE_LIBBPF_OPTS(bpf_link_create_opts, opts,
> >> > + .target_btf_id = target_btf_id);
> >> >
> >> > prog_fd = bpf_program__fd(prog);
> >> > if (prog_fd < 0) {
> >> > @@ -9340,8 +9342,12 @@ bpf_program__attach_fd(struct bpf_program *prog, int target_fd,
> >> > return ERR_PTR(-ENOMEM);
> >> > link->detach = &bpf_link__detach_fd;
> >> >
> >> > - attach_type = bpf_program__get_expected_attach_type(prog);
> >> > - link_fd = bpf_link_create(prog_fd, target_fd, attach_type, NULL);
> >> > + if (bpf_program__get_type(prog) == BPF_PROG_TYPE_EXT)
> >> > + attach_type = BPF_TRACE_FREPLACE;
> >>
> >> doing this unconditionally will break an old-style freplace without
> >> target_fd/btf_id on older kernels. Safe and simple way would be to
> >> continue using raw_tracepoint_open when there is no target_fd/btf_id,
> >> and use LINK_CREATE for newer stuff. Alternatively, you'd need to do
> >> feature detection, but it's still would be nice to handle old-style
> >> attach through raw_tracepoint_open for bpf_program__attach_freplace.
> >>
> >> so I suggest leaving bpf_program__attach_fd() as is and to create a
> >> custom bpf_program__attach_freplace() implementation.
>
> Sure, I'll take another pass at this. Not sure how useful feature
> detection in libbpf is; if the caller passes a target, libbpf can't
> really do much if the kernel doesn't support it...
I was thinking about bpf_program__attach_freplace(prog, 0, NULL) doing
bpf_raw_tracepoint_open(). It would be nice to support this, for API
uniformity, no?
>
> >> > + else
> >> > + attach_type = bpf_program__get_expected_attach_type(prog);
> >> > +
> >> > + link_fd = bpf_link_create(prog_fd, target_fd, attach_type, &opts);
> >> > if (link_fd < 0) {
> >> > link_fd = -errno;
> >> > free(link);
> >> > @@ -9357,19 +9363,25 @@ bpf_program__attach_fd(struct bpf_program *prog, int target_fd,
> >> > struct bpf_link *
> >> > bpf_program__attach_cgroup(struct bpf_program *prog, int cgroup_fd)
> >> > {
> >> > - return bpf_program__attach_fd(prog, cgroup_fd, "cgroup");
> >> > + return bpf_program__attach_fd(prog, cgroup_fd, 0, "cgroup");
> >> > }
> >> >
> >> > struct bpf_link *
> >> > bpf_program__attach_netns(struct bpf_program *prog, int netns_fd)
> >> > {
> >> > - return bpf_program__attach_fd(prog, netns_fd, "netns");
> >> > + return bpf_program__attach_fd(prog, netns_fd, 0, "netns");
> >> > }
> >> >
> >> > struct bpf_link *bpf_program__attach_xdp(struct bpf_program *prog, int ifindex)
> >> > {
> >> > /* target_fd/target_ifindex use the same field in LINK_CREATE */
> >> > - return bpf_program__attach_fd(prog, ifindex, "xdp");
> >> > + return bpf_program__attach_fd(prog, ifindex, 0, "xdp");
> >> > +}
> >> > +
> >> > +struct bpf_link *bpf_program__attach_freplace(struct bpf_program *prog,
> >> > + int target_fd, int target_btf_id)
> >> > +{
> >> > + return bpf_program__attach_fd(prog, target_fd, target_btf_id, "freplace");
> >> > }
> >> >
> >> > struct bpf_link *
> >> > diff --git a/tools/lib/bpf/libbpf.h b/tools/lib/bpf/libbpf.h
> >> > index a750f67a23f6..ce5add9b9203 100644
> >> > --- a/tools/lib/bpf/libbpf.h
> >> > +++ b/tools/lib/bpf/libbpf.h
> >> > @@ -261,6 +261,9 @@ LIBBPF_API struct bpf_link *
> >> > bpf_program__attach_netns(struct bpf_program *prog, int netns_fd);
> >> > LIBBPF_API struct bpf_link *
> >> > bpf_program__attach_xdp(struct bpf_program *prog, int ifindex);
> >> > +LIBBPF_API struct bpf_link *
> >> > +bpf_program__attach_freplace(struct bpf_program *prog,
> >> > + int target_fd, int target_btf_id);
> >>
> >> maybe a const char * function name instead of target_btf_id would be a
> >> nicer API? Users won't have to deal with fetching target prog's BTF,
> >> searching it, etc. That's all pretty straightforward for libbpf to do,
> >> leaving users with more natural and simpler API.
> >>
> >
> > bpf_program__set_attach_target() uses string name for target
> > functions, we should definitely be consistent here
>
> All right, fair enough :)
>
> -Toke
>
