Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 2620:137:e000::3:7 as permitted sender) client-ip=2620:137:e000::3:7;
From:   Hengqi Chen <hengqi.chen@gmail.com>
To:     linux-kernel@vger.kernel.org
Cc:     keescook@chromium.org, luto@amacapital.net, wad@chromium.org,
        alexyonghe@tencent.com, hengqi.chen@gmail.com
Subject: [PATCH v3 0/3] seccomp: Make seccomp filter reusable
Date:   Wed, 29 Nov 2023 05:34:37 +0000
Message-Id: <20231129053440.41522-1-hengqi.chen@gmail.com>
MIME-Version: 1.0
Content-Transfer-Encoding: 8bit
Precedence: bulk

This patchset introduces a new operation and a new flag
to split the SECCOMP_SET_MODE_FILTER process into two steps:
load filter and attach filter. Thus we can reuse the filter.

The new SECCOMP_LOAD_FILTER loads the filter and returns a fd
which can be reused for different processes via UDS.
With this new operation, we can eliminate a hot path of JITing
BPF program (the filter) where we apply the same seccomp filter
to thousands of micro VMs on a bare metal instance.

The new flag SECCOMP_FILTER_FLAG_FILTER_FD is used to attach
a filter which is loaded via SECCOMP_LOAD_FILTER and represented
by a seccomp filter fd.

Change logs:
  v2 -> v3:
    - Stick to cBPF, don't interfere with bpffs

  v1 -> v2:
    - Add a flag to attach filter from fd
    - Update selftests

  RFC -> v1:
    - Addressed comments from Kees
    - Reuse filter copy/create code (patch 1)
    - Add a selftest (patch 4)

Hengqi Chen (3):
  seccomp: Introduce SECCOMP_LOAD_FILTER operation
  seccomp: Introduce new flag SECCOMP_FILTER_FLAG_FILTER_FD
  selftests/seccomp: Test seccomp filter load and attach

 include/linux/seccomp.h                       |  3 +-
 include/uapi/linux/seccomp.h                  |  3 +
 kernel/seccomp.c                              | 78 ++++++++++++++++++-
 tools/testing/selftests/seccomp/seccomp_bpf.c | 71 +++++++++++++++++
 4 files changed, 151 insertions(+), 4 deletions(-)

--
2.34.1