github.com/meulengracht/snapd@v0.0.0-20210719210640-8bde69bcc84e/cmd/snap-confine/snap-confine.rst (about) 1 ============== 2 snap-confine 3 ============== 4 5 ----------------------------------------------- 6 internal tool for confining snappy applications 7 ----------------------------------------------- 8 9 :Author: zygmunt.krynicki@canonical.com 10 :Date: 2017-09-18 11 :Copyright: Canonical Ltd. 12 :Version: 2.28 13 :Manual section: 8 14 :Manual group: snappy 15 16 SYNOPSIS 17 ======== 18 19 snap-confine [--classic] [--base BASE] SECURITY_TAG COMMAND [...ARGUMENTS] 20 21 DESCRIPTION 22 =========== 23 24 The `snap-confine` is a program used internally by `snapd` to construct the 25 execution environment for snap applications. 26 27 OPTIONS 28 ======= 29 30 The `snap-confine` program accepts two options: 31 32 `--classic` requests the so-called _classic_ _confinement_ in which 33 applications are not confined at all (like in classic systems, hence the 34 name). This disables the use of a dedicated, per-snap mount namespace. The 35 `snapd` service generates permissive apparmor and seccomp profiles that 36 allow everything. 37 38 `--base BASE` directs snap-confine to use the given base snap as the root 39 filesystem. If omitted it defaults to the `core` snap. This is derived from 40 snap meta-data by `snapd` when starting the application process. 41 42 FEATURES 43 ======== 44 45 Apparmor profiles 46 ----------------- 47 48 `snap-confine` switches to the apparmor profile `$SECURITY_TAG`. The profile is 49 **mandatory** and `snap-confine` will refuse to run without it. 50 51 The profile has to be loaded into the kernel prior to using `snap-confine`. 52 Typically this is arranged for by `snapd`. The profile contains rich 53 description of what the application process is allowed to do, this includes 54 system calls, file paths, access patterns, linux capabilities, etc. The 55 apparmor profile can also do extensive dbus mediation. Refer to apparmor 56 documentation for more details. 57 58 Seccomp profiles 59 ---------------- 60 61 `snap-confine` looks for the 62 `/var/lib/snapd/seccomp/bpf/$SECURITY_TAG.bin` file. This file is 63 **mandatory** and `snap-confine` will refuse to run without it. This 64 file contains the seccomp bpf binary program that is loaded into the 65 kernel by snap-confine. 66 67 The file is generated with the `/usr/lib/snapd/snap-seccomp` compiler from the 68 `$SECURITY_TAG.src` file that uses a custom syntax that describes the set of 69 allowed system calls and optionally their arguments. The profile is then used 70 to confine the started application. 71 72 As a security precaution disallowed system calls cause the started application 73 executable to be killed by the kernel. In the future this restriction may be 74 lifted to return `EPERM` instead. 75 76 Mount profiles 77 -------------- 78 79 `snap-confine` uses a helper process, `snap-update-ns`, to apply the mount 80 namespace profile to freshly constructed mount namespace. That tool looks for 81 the `/var/lib/snapd/mount/snap.$SNAP_NAME.fstab` file. If present it is read, 82 parsed and treated like a mostly-typical `fstab(5)` file. The mount directives 83 listed there are executed in order. All directives must succeed as any failure 84 will abort execution. 85 86 By default all mount entries start with the following flags: `bind`, `ro`, 87 `nodev`, `nosuid`. Some of those flags can be reversed by an appropriate 88 option (e.g. `rw` can cause the mount point to be writable). 89 90 Certain additional features are enabled and conveyed through the use of mount 91 options prefixed with `x-snapd-`. 92 93 As a security precaution only `bind` mounts are supported at this time. 94 95 Sharing of the mount namespace 96 ------------------------------ 97 98 As of version 1.0.41 all the applications from the same snap will share the 99 same mount namespace. Applications from different snaps continue to use 100 separate mount namespaces. 101 102 ENVIRONMENT 103 =========== 104 105 `snap-confine` responds to the following environment variables 106 107 `SNAP_CONFINE_DEBUG`: 108 When defined the program will print additional diagnostic information about 109 the actions being performed. All the output goes to stderr. 110 111 The following variables are only used when `snap-confine` is not setuid root. 112 This is only applicable when testing the program itself. 113 114 `SNAPPY_LAUNCHER_INSIDE_TESTS`: 115 Internal variable that should not be relied upon. 116 117 `SNAPPY_LAUNCHER_SECCOMP_PROFILE_DIR`: 118 Internal variable that should not be relied upon. 119 120 `SNAP_USER_DATA`: 121 Full path to the directory like /home/$LOGNAME/snap/$SNAP_NAME/$SNAP_REVISION. 122 123 This directory is created by snap-confine on startup. This is a temporary 124 feature that will be merged into snapd's snap-run command. The set of directories 125 that can be created is confined with apparmor. 126 127 FILES 128 ===== 129 130 `snap-confine` and `snap-update-ns` use the following files: 131 132 `/var/lib/snapd/mount/snap.*.fstab`: 133 134 Description of the mount profile. 135 136 `/var/lib/snapd/seccomp/bpf/*.src`: 137 138 Input for the /usr/lib/snapd/snap-seccomp profile compiler. 139 140 `/var/lib/snapd/seccomp/bpf/*.bin`: 141 142 Compiled seccomp bpf profile programs. 143 144 `/run/snapd/ns/`: 145 146 Directory used to keep shared mount namespaces. 147 148 `snap-confine` internally converts this directory to a private bind mount. 149 Semantically the behavior is identical to the following mount commands: 150 151 mount --bind /run/snapd/ns /run/snapd/ns 152 mount --make-private /run/snapd/ns 153 154 `/run/snapd/ns/.lock`: 155 156 A `flock(2)`-based lock file acquired to create and convert 157 `/run/snapd/ns/` to a private bind mount. 158 159 `/run/snapd/ns/$SNAP_NAME.lock`: 160 161 A `flock(2)`-based lock file acquired to create or join the mount namespace 162 represented as `/run/snaps/ns/$SNAP_NAME.mnt`. 163 164 `/run/snapd/ns/$SNAP_NAME.mnt`: 165 166 This file can be either: 167 168 - An empty file that may be seen before the mount namespace is preserved or 169 when the mount namespace is unmounted. 170 - A file belonging to the `nsfs` file system, representing a fully 171 populated mount namespace of a given snap. The file is bind mounted from 172 `/proc/self/ns/mnt` from the first process in any snap. 173 174 `/proc/self/mountinfo`: 175 176 This file is read to decide if `/run/snapd/ns/` needs to be created and 177 converted to a private bind mount, as described above. 178 179 Note that the apparmor profile is external to `snap-confine` and is loaded 180 directly into the kernel. The actual apparmor profile is managed by `snapd`. 181 182 BUGS 183 ==== 184 185 Please report all bugs with https://bugs.launchpad.net/snap-confine/+filebug