aboutsummaryrefslogtreecommitdiffstats
path: root/kernel/bpf/syscall.c
diff options
context:
space:
mode:
authorArthur Fabre <afabre@cloudflare.com>2018-07-31 18:17:22 +0100
committerDaniel Borkmann <daniel@iogearbox.net>2018-07-31 22:09:33 +0200
commitfbeb1603bf4e9baa82da8f794de42949d0fe5e25 (patch)
tree18cda999382d4e19ab89c30221ffeafed8b26baa /kernel/bpf/syscall.c
parentselftests/bpf: Test for get_socket_cookie (diff)
downloadlinux-dev-fbeb1603bf4e9baa82da8f794de42949d0fe5e25.tar.xz
linux-dev-fbeb1603bf4e9baa82da8f794de42949d0fe5e25.zip
bpf: verifier: MOV64 don't mark dst reg unbounded
When check_alu_op() handles a BPF_MOV64 between two registers, it calls check_reg_arg(DST_OP) on the dst register, marking it as unbounded. If the src and dst register are the same, this marks the src as unbounded, which can lead to unexpected errors for further checks that rely on bounds info. For example: BPF_MOV64_IMM(BPF_REG_2, 0), BPF_MOV64_REG(BPF_REG_2, BPF_REG_2), BPF_ALU64_REG(BPF_ADD, BPF_REG_1, BPF_REG_2), BPF_MOV64_IMM(BPF_REG_0, 0), BPF_EXIT_INSN(), Results in: "math between ctx pointer and register with unbounded min value is not allowed" check_alu_op() now uses check_reg_arg(DST_OP_NO_MARK), and MOVs that need to mark the dst register (MOVIMM, MOV32) do so. Added a test case for MOV64 dst == src, and dst != src. Signed-off-by: Arthur Fabre <afabre@cloudflare.com> Acked-by: Edward Cree <ecree@solarflare.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Diffstat (limited to 'kernel/bpf/syscall.c')
0 files changed, 0 insertions, 0 deletions