Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::1:20 as permitted sender) client-ip=2620:137:e000::1:20;
From:   Pu Lehui <pulehui@huaweicloud.com>
To:     Martin KaFai Lau <martin.lau@linux.dev>,
        Stanislav Fomichev <sdf@google.com>, bpf@vger.kernel.org,
        linux-kernel@vger.kernel.org
Cc:     Alexei Starovoitov <ast@kernel.org>,
        Daniel Borkmann <daniel@iogearbox.net>,
        Andrii Nakryiko <andrii@kernel.org>,
        Quentin Monnet <quentin@isovalent.com>,
        Song Liu <song@kernel.org>, Yonghong Song <yhs@fb.com>,
        John Fastabend <john.fastabend@gmail.com>,
        KP Singh <kpsingh@kernel.org>, Hao Luo <haoluo@google.com>,
        Jiri Olsa <jolsa@kernel.org>, Pu Lehui <pulehui@huawei.com>,
        Pu Lehui <pulehui@huaweicloud.com>
Subject: [PATCH bpf v5 0/3] Fix wrong cgroup attach flags being assigned to effective progs
Date:   Wed, 21 Sep 2022 10:46:01 +0000
Message-Id: <20220921104604.2340580-1-pulehui@huaweicloud.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit
Precedence: bulk

From: Pu Lehui <pulehui@huawei.com>

When root-cgroup attach multi progs and sub-cgroup attach a
override prog, bpftool will display incorrectly for the attach
flags of the sub-cgroup’s effective progs:

$ bpftool cgroup tree /sys/fs/cgroup effective
CgroupPath
ID       AttachType      AttachFlags     Name
/sys/fs/cgroup
6        cgroup_sysctl   multi           sysctl_tcp_mem
13       cgroup_sysctl   multi           sysctl_tcp_mem
/sys/fs/cgroup/cg1
20       cgroup_sysctl   override        sysctl_tcp_mem
6        cgroup_sysctl   override        sysctl_tcp_mem <- wrong
13       cgroup_sysctl   override        sysctl_tcp_mem <- wrong
/sys/fs/cgroup/cg1/cg2
20       cgroup_sysctl                   sysctl_tcp_mem
6        cgroup_sysctl                   sysctl_tcp_mem
13       cgroup_sysctl                   sysctl_tcp_mem

For cg1, obviously, the attach flags of prog6 and prog13 can not be
OVERRIDE. And for query with EFFECTIVE flags, exporting attach flags
does not make sense. So let's remove the AttachFlags field and the
associated logic. After these patches, the above effective cgroup
tree will show as bellow:

# bpftool cgroup tree /sys/fs/cgroup effective
CgroupPath
ID       AttachType      Name
/sys/fs/cgroup
6        cgroup_sysctl   sysctl_tcp_mem
13       cgroup_sysctl   sysctl_tcp_mem
/sys/fs/cgroup/cg1
20       cgroup_sysctl   sysctl_tcp_mem
6        cgroup_sysctl   sysctl_tcp_mem
13       cgroup_sysctl   sysctl_tcp_mem
/sys/fs/cgroup/cg1/cg2
20       cgroup_sysctl   sysctl_tcp_mem
6        cgroup_sysctl   sysctl_tcp_mem
13       cgroup_sysctl   sysctl_tcp_mem

v5:
- Adapt selftests for effective query uapi change.

v4:
https://lore.kernel.org/bpf/20220920154233.1494352-1-pulehui@huaweicloud.com
- Reject prog_attach_flags array when effective query. (Martin)
- Target to bpf tree. (Martin)

v3:
https://lore.kernel.org/bpf/20220914161742.3180731-1-pulehui@huaweicloud.com
- Don't show attach flags when effective query. (John, sdf, Martin)

v2:
https://lore.kernel.org/bpf/20220908145304.3436139-1-pulehui@huaweicloud.com
- Limit prog_cnt to avoid overflow. (John)
- Add more detail message.

v1:
https://lore.kernel.org/bpf/20220820120234.2121044-1-pulehui@huawei.com

Pu Lehui (3):
  bpf, cgroup: Reject prog_attach_flags array when effective query
  bpftool: Fix wrong cgroup attach flags being assigned to effective
    progs
  selftests/bpf: Adapt cgroup effective query uapi change

 include/uapi/linux/bpf.h                      |  7 ++-
 kernel/bpf/cgroup.c                           | 28 ++++++----
 tools/bpf/bpftool/cgroup.c                    | 54 +++++++++++++++++--
 tools/include/uapi/linux/bpf.h                |  7 ++-
 .../selftests/bpf/prog_tests/cgroup_link.c    | 11 ++--
 5 files changed, 81 insertions(+), 26 deletions(-)

-- 
2.25.1