1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
|
// Copyright 2018 The gVisor Authors.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package linux
import (
"gvisor.dev/gvisor/pkg/abi/linux"
"gvisor.dev/gvisor/pkg/bpf"
"gvisor.dev/gvisor/pkg/errors/linuxerr"
"gvisor.dev/gvisor/pkg/hostarch"
"gvisor.dev/gvisor/pkg/sentry/arch"
"gvisor.dev/gvisor/pkg/sentry/kernel"
)
// userSockFprog is equivalent to Linux's struct sock_fprog on amd64.
//
// +marshal
type userSockFprog struct {
// Len is the length of the filter in BPF instructions.
Len uint16
_ [6]byte // padding for alignment
// Filter is a user pointer to the struct sock_filter array that makes up
// the filter program. Filter is a uint64 rather than a hostarch.Addr
// because hostarch.Addr is actually uintptr, which is not a fixed-size
// type.
Filter uint64
}
// seccomp applies a seccomp policy to the current task.
func seccomp(t *kernel.Task, mode, flags uint64, addr hostarch.Addr) error {
// We only support SECCOMP_SET_MODE_FILTER at the moment.
if mode != linux.SECCOMP_SET_MODE_FILTER {
// Unsupported mode.
return linuxerr.EINVAL
}
tsync := flags&linux.SECCOMP_FILTER_FLAG_TSYNC != 0
// The only flag we support now is SECCOMP_FILTER_FLAG_TSYNC.
if flags&^linux.SECCOMP_FILTER_FLAG_TSYNC != 0 {
// Unsupported flag.
return linuxerr.EINVAL
}
var fprog userSockFprog
if _, err := fprog.CopyIn(t, addr); err != nil {
return err
}
if fprog.Len == 0 || fprog.Len > bpf.MaxInstructions {
// If the filter is already over the maximum number of instructions,
// do not go further and attempt to optimize the bytecode to make it
// smaller.
return linuxerr.EINVAL
}
filter := make([]linux.BPFInstruction, int(fprog.Len))
if _, err := linux.CopyBPFInstructionSliceIn(t, hostarch.Addr(fprog.Filter), filter); err != nil {
return err
}
bpfFilter := make([]bpf.Instruction, len(filter))
for i, ins := range filter {
bpfFilter[i] = bpf.Instruction(ins)
}
compiledFilter, err := bpf.Compile(bpfFilter, true /* optimize */)
if err != nil {
t.Debugf("Invalid seccomp-bpf filter: %v", err)
return linuxerr.EINVAL
}
return t.AppendSyscallFilter(compiledFilter, tsync)
}
// Seccomp implements linux syscall seccomp(2).
func Seccomp(t *kernel.Task, sysno uintptr, args arch.SyscallArguments) (uintptr, *kernel.SyscallControl, error) {
return 0, nil, seccomp(t, args[0].Uint64(), args[1].Uint64(), args[2].Pointer())
}
|
