From: Andrii Nakryiko <andrii.nakryiko@gmail.com>
To: Martin Lau <kafai@fb.com>
Cc: bpf <bpf@vger.kernel.org>, Alexei Starovoitov <ast@kernel.org>,
Daniel Borkmann <daniel@iogearbox.net>,
David Miller <davem@davemloft.net>,
Kernel Team <Kernel-team@fb.com>,
Networking <netdev@vger.kernel.org>
Subject: Re: [PATCH bpf-next v2 05/11] bpf: Introduce BPF_PROG_TYPE_STRUCT_OPS
Date: Mon, 23 Dec 2019 14:55:12 -0800 [thread overview]
Message-ID: <CAEf4BzaM7OGnocOc=58hXAAcLvM0qaYRWuwiqt1L2cPY1rWykA@mail.gmail.com> (raw)
In-Reply-To: <20191223222955.2d2hxboqzgp7662r@kafai-mbp>
On Mon, Dec 23, 2019 at 2:30 PM Martin Lau <kafai@fb.com> wrote:
>
> On Mon, Dec 23, 2019 at 12:29:37PM -0800, Andrii Nakryiko wrote:
> > On Fri, Dec 20, 2019 at 10:26 PM Martin KaFai Lau <kafai@fb.com> wrote:
> > >
> > > This patch allows the kernel's struct ops (i.e. func ptr) to be
> > > implemented in BPF. The first use case in this series is the
> > > "struct tcp_congestion_ops" which will be introduced in a
> > > latter patch.
> > >
> > > This patch introduces a new prog type BPF_PROG_TYPE_STRUCT_OPS.
> > > The BPF_PROG_TYPE_STRUCT_OPS prog is verified against a particular
> > > func ptr of a kernel struct. The attr->attach_btf_id is the btf id
> > > of a kernel struct. The attr->expected_attach_type is the member
> > > "index" of that kernel struct. The first member of a struct starts
> > > with member index 0. That will avoid ambiguity when a kernel struct
> > > has multiple func ptrs with the same func signature.
> > >
> > > For example, a BPF_PROG_TYPE_STRUCT_OPS prog is written
> > > to implement the "init" func ptr of the "struct tcp_congestion_ops".
> > > The attr->attach_btf_id is the btf id of the "struct tcp_congestion_ops"
> > > of the _running_ kernel. The attr->expected_attach_type is 3.
> > >
> > > The ctx of BPF_PROG_TYPE_STRUCT_OPS is an array of u64 args saved
> > > by arch_prepare_bpf_trampoline that will be done in the next
> > > patch when introducing BPF_MAP_TYPE_STRUCT_OPS.
> > >
> > > "struct bpf_struct_ops" is introduced as a common interface for the kernel
> > > struct that supports BPF_PROG_TYPE_STRUCT_OPS prog. The supporting kernel
> > > struct will need to implement an instance of the "struct bpf_struct_ops".
> > >
> > > The supporting kernel struct also needs to implement a bpf_verifier_ops.
> > > During BPF_PROG_LOAD, bpf_struct_ops_find() will find the right
> > > bpf_verifier_ops by searching the attr->attach_btf_id.
> > >
> > > A new "btf_struct_access" is also added to the bpf_verifier_ops such
> > > that the supporting kernel struct can optionally provide its own specific
> > > check on accessing the func arg (e.g. provide limited write access).
> > >
> > > After btf_vmlinux is parsed, the new bpf_struct_ops_init() is called
> > > to initialize some values (e.g. the btf id of the supporting kernel
> > > struct) and it can only be done once the btf_vmlinux is available.
> > >
> > > The R0 checks at BPF_EXIT is excluded for the BPF_PROG_TYPE_STRUCT_OPS prog
> > > if the return type of the prog->aux->attach_func_proto is "void".
> > >
> > > Signed-off-by: Martin KaFai Lau <kafai@fb.com>
> > > ---
> > > include/linux/bpf.h | 30 +++++++
> > > include/linux/bpf_types.h | 4 +
> > > include/linux/btf.h | 34 ++++++++
> > > include/uapi/linux/bpf.h | 1 +
> > > kernel/bpf/Makefile | 2 +-
> > > kernel/bpf/bpf_struct_ops.c | 122 +++++++++++++++++++++++++++
> > > kernel/bpf/bpf_struct_ops_types.h | 4 +
> > > kernel/bpf/btf.c | 88 ++++++++++++++------
> > > kernel/bpf/syscall.c | 17 ++--
> > > kernel/bpf/verifier.c | 134 +++++++++++++++++++++++-------
> > > 10 files changed, 372 insertions(+), 64 deletions(-)
> > > create mode 100644 kernel/bpf/bpf_struct_ops.c
> > > create mode 100644 kernel/bpf/bpf_struct_ops_types.h
> > >
> >
> > All looks good, apart from the concern with partially-initialized
> > bpf_struct_ops.
> >
> > [...]
> >
> > > +const struct bpf_prog_ops bpf_struct_ops_prog_ops = {
> > > +};
> > > +
> > > +void bpf_struct_ops_init(struct btf *_btf_vmlinux)
> >
> > this is always get passed vmlinux's btf, so why not call it short and
> > sweet "btf"? _btf_vmlinux is kind of ugly and verbose.
> >
> > > +{
> > > + const struct btf_member *member;
> > > + struct bpf_struct_ops *st_ops;
> > > + struct bpf_verifier_log log = {};
> > > + const struct btf_type *t;
> > > + const char *mname;
> > > + s32 type_id;
> > > + u32 i, j;
> > > +
> >
> > [...]
> >
> > > +static int check_struct_ops_btf_id(struct bpf_verifier_env *env)
> > > +{
> > > + const struct btf_type *t, *func_proto;
> > > + const struct bpf_struct_ops *st_ops;
> > > + const struct btf_member *member;
> > > + struct bpf_prog *prog = env->prog;
> > > + u32 btf_id, member_idx;
> > > + const char *mname;
> > > +
> > > + btf_id = prog->aux->attach_btf_id;
> > > + st_ops = bpf_struct_ops_find(btf_id);
> >
> > if struct_ops initialization fails, type will be NULL and type_id will
> > be 0, which we rely on here to not get partially-initialized
> > bpf_struct_ops, right? Small comment mentioning this would be helpful.
> >
> >
> > > + if (!st_ops) {
> > > + verbose(env, "attach_btf_id %u is not a supported struct\n",
> > > + btf_id);
> > > + return -ENOTSUPP;
> > > + }
> > > +
> >
> > [...]
> >
> > > static int check_attach_btf_id(struct bpf_verifier_env *env)
> > > {
> > > struct bpf_prog *prog = env->prog;
> > > @@ -9520,6 +9591,9 @@ static int check_attach_btf_id(struct bpf_verifier_env *env)
> > > long addr;
> > > u64 key;
> > >
> > > + if (prog->type == BPF_PROG_TYPE_STRUCT_OPS)
> > > + return check_struct_ops_btf_id(env);
> > > +
> >
> > There is a btf_id == 0 check below, you need to check that for
> > STRUCT_OPS as well, otherwise you can get partially-initialized
> > bpf_struct_ops struct in check_struct_ops_btf_id.
> This btf_id == 0 check is done at the beginning of bpf_struct_ops_find().
> Hence, bpf_struct_ops_find() won't try to search if btf_id is 0.
>
Ah right, I missed that check. Then yeah, it's not a concern. I still
don't like _btf_vmlinux name, but that's just a nit.
Acked-by: Andrii Nakryiko <andriin@fb.com>
> st_ops fields is only set when everything passed, so individual st_ops
> will not be partially initialized.
>
>
> >
> > > if (prog->type != BPF_PROG_TYPE_TRACING)
> > > return 0;
> > >
> > > --
> > > 2.17.1
> > >
next prev parent reply other threads:[~2019-12-23 22:55 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-12-21 6:25 [PATCH bpf-next v2 00/11] Introduce BPF STRUCT_OPS Martin KaFai Lau
2019-12-21 6:25 ` [PATCH bpf-next v2 01/11] bpf: Save PTR_TO_BTF_ID register state when spilling to stack Martin KaFai Lau
2019-12-21 6:25 ` [PATCH bpf-next v2 02/11] bpf: Avoid storing modifier to info->btf_id Martin KaFai Lau
2019-12-21 6:26 ` [PATCH bpf-next v2 03/11] bpf: Add enum support to btf_ctx_access() Martin KaFai Lau
2019-12-21 6:26 ` [PATCH bpf-next v2 04/11] bpf: Support bitfield read access in btf_struct_access Martin KaFai Lau
2019-12-23 7:49 ` Yonghong Song
2019-12-23 20:05 ` Andrii Nakryiko
2019-12-23 21:21 ` Yonghong Song
2019-12-21 6:26 ` [PATCH bpf-next v2 05/11] bpf: Introduce BPF_PROG_TYPE_STRUCT_OPS Martin KaFai Lau
2019-12-23 19:33 ` Yonghong Song
2019-12-23 20:29 ` Andrii Nakryiko
2019-12-23 22:29 ` Martin Lau
2019-12-23 22:55 ` Andrii Nakryiko [this message]
2019-12-24 11:46 ` kbuild test robot
2019-12-21 6:26 ` [PATCH bpf-next v2 06/11] bpf: Introduce BPF_MAP_TYPE_STRUCT_OPS Martin KaFai Lau
2019-12-23 19:57 ` Yonghong Song
2019-12-23 21:44 ` Andrii Nakryiko
2019-12-23 22:15 ` Martin Lau
2019-12-27 6:16 ` Martin Lau
2019-12-23 23:05 ` Andrii Nakryiko
2019-12-28 1:47 ` Martin Lau
2019-12-28 2:24 ` Andrii Nakryiko
2019-12-28 5:16 ` Martin Lau
2019-12-24 12:28 ` kbuild test robot
2019-12-21 6:26 ` [PATCH bpf-next v2 07/11] bpf: tcp: Support tcp_congestion_ops in bpf Martin KaFai Lau
2019-12-23 20:18 ` Yonghong Song
2019-12-23 23:20 ` Andrii Nakryiko
2019-12-24 7:16 ` kbuild test robot
2019-12-24 13:06 ` kbuild test robot
2019-12-21 6:26 ` [PATCH bpf-next v2 08/11] bpf: Add BPF_FUNC_tcp_send_ack helper Martin KaFai Lau
2019-12-21 6:26 ` [PATCH bpf-next v2 09/11] bpf: Synch uapi bpf.h to tools/ Martin KaFai Lau
2019-12-21 6:26 ` [PATCH bpf-next v2 10/11] bpf: libbpf: Add STRUCT_OPS support Martin KaFai Lau
2019-12-23 19:54 ` Andrii Nakryiko
2019-12-26 22:47 ` Martin Lau
2019-12-21 6:26 ` [PATCH bpf-next v2 11/11] bpf: Add bpf_dctcp example Martin KaFai Lau
2019-12-23 23:26 ` Andrii Nakryiko
2019-12-24 1:31 ` Martin Lau
2019-12-24 7:01 ` Andrii Nakryiko
2019-12-24 7:32 ` Martin Lau
2019-12-24 16:50 ` Martin Lau
2019-12-26 19:02 ` Andrii Nakryiko
2019-12-26 20:25 ` Martin Lau
2019-12-26 20:48 ` Andrii Nakryiko
2019-12-26 22:20 ` Martin Lau
2019-12-26 22:25 ` Andrii Nakryiko
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to='CAEf4BzaM7OGnocOc=58hXAAcLvM0qaYRWuwiqt1L2cPY1rWykA@mail.gmail.com' \
--to=andrii.nakryiko@gmail.com \
--cc=Kernel-team@fb.com \
--cc=ast@kernel.org \
--cc=bpf@vger.kernel.org \
--cc=daniel@iogearbox.net \
--cc=davem@davemloft.net \
--cc=kafai@fb.com \
--cc=netdev@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).