summaryrefslogtreecommitdiff
path: root/vendor/github.com/opencontainers/runc
diff options
context:
space:
mode:
authorMatthew Heon <matthew.heon@gmail.com>2017-11-01 11:24:59 -0400
committerMatthew Heon <matthew.heon@gmail.com>2017-11-01 11:24:59 -0400
commita031b83a09a8628435317a03f199cdc18b78262f (patch)
treebc017a96769ce6de33745b8b0b1304ccf38e9df0 /vendor/github.com/opencontainers/runc
parent2b74391cd5281f6fdf391ff8ad50fd1490f6bf89 (diff)
downloadpodman-a031b83a09a8628435317a03f199cdc18b78262f.tar.gz
podman-a031b83a09a8628435317a03f199cdc18b78262f.tar.bz2
podman-a031b83a09a8628435317a03f199cdc18b78262f.zip
Initial checkin from CRI-O repo
Signed-off-by: Matthew Heon <matthew.heon@gmail.com>
Diffstat (limited to 'vendor/github.com/opencontainers/runc')
-rw-r--r--vendor/github.com/opencontainers/runc/LICENSE191
-rw-r--r--vendor/github.com/opencontainers/runc/NOTICE17
-rw-r--r--vendor/github.com/opencontainers/runc/README.md244
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/README.md328
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor.go39
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor_disabled.go20
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/capabilities_linux.go114
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups.go64
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups_unsupported.go3
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/apply_raw.go360
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/blkio.go237
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpu.go125
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuacct.go121
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuset.go163
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/devices.go80
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/freezer.go61
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/fs_unsupported.go3
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/hugetlb.go71
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/memory.go313
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/name.go40
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_cls.go43
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_prio.go41
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/perf_event.go35
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/pids.go73
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/utils.go78
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/rootless/rootless.go128
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/stats.go108
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_nosystemd.go55
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_systemd.go553
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/cgroups/utils.go462
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/compat_1.5_linux.go10
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/blkio_device.go61
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_linux.go122
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_unsupported.go6
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_windows.go6
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/config.go344
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/config_linux.go61
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/device.go57
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/device_defaults.go111
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/hugepage_limit.go9
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/interface_priority_map.go14
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/mount.go39
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces.go5
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_linux.go122
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall.go31
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall_unsupported.go13
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_unsupported.go8
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/network.go72
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/validate/rootless.go117
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/configs/validate/validator.go195
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/console.go17
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/console_freebsd.go13
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/console_linux.go152
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/console_solaris.go11
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/console_windows.go30
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/container.go166
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/container_linux.go1654
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/container_solaris.go20
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/container_windows.go20
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/criu_opts_linux.go37
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/criu_opts_windows.go6
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.pb.go1178
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.proto209
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/devices/devices_linux.go100
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/devices/devices_unsupported.go3
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/devices/number.go24
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/error.go70
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/factory.go44
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/factory_linux.go325
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/generic_error.go92
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/init_linux.go502
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/keys/keyctl.go50
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/message_linux.go87
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/network_linux.go259
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/notify_linux.go90
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/README.md44
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/namespace.h32
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go12
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go25
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go5
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c862
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/process.go106
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/process_linux.go493
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/restored_process.go122
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/rootfs_linux.go812
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/seccomp/config.go76
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_linux.go227
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_unsupported.go24
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/setgroups_linux.go11
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/setns_init_linux.go65
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stacktrace/capture.go27
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stacktrace/frame.go38
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stacktrace/stacktrace.go5
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/standard_init_linux.go188
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/state_linux.go248
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stats.go15
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stats_freebsd.go5
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stats_linux.go8
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stats_solaris.go7
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/stats_windows.go5
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/sync.go107
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/linux.go136
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/proc.go113
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_386.go25
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_64.go25
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_arm.go25
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig.go12
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig_notcgo.go15
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/unsupported.go9
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/system/xattrs_linux.go35
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/user/lookup.go95
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unix.go46
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unsupported.go38
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/user/user.go441
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.go95
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/utils/utils.go127
-rw-r--r--vendor/github.com/opencontainers/runc/libcontainer/utils/utils_unix.go44
-rw-r--r--vendor/github.com/opencontainers/runc/vendor.conf21
118 files changed, 15603 insertions, 0 deletions
diff --git a/vendor/github.com/opencontainers/runc/LICENSE b/vendor/github.com/opencontainers/runc/LICENSE
new file mode 100644
index 000000000..27448585a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/LICENSE
@@ -0,0 +1,191 @@
+
+ Apache License
+ Version 2.0, January 2004
+ http://www.apache.org/licenses/
+
+ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION
+
+ 1. Definitions.
+
+ "License" shall mean the terms and conditions for use, reproduction,
+ and distribution as defined by Sections 1 through 9 of this document.
+
+ "Licensor" shall mean the copyright owner or entity authorized by
+ the copyright owner that is granting the License.
+
+ "Legal Entity" shall mean the union of the acting entity and all
+ other entities that control, are controlled by, or are under common
+ control with that entity. For the purposes of this definition,
+ "control" means (i) the power, direct or indirect, to cause the
+ direction or management of such entity, whether by contract or
+ otherwise, or (ii) ownership of fifty percent (50%) or more of the
+ outstanding shares, or (iii) beneficial ownership of such entity.
+
+ "You" (or "Your") shall mean an individual or Legal Entity
+ exercising permissions granted by this License.
+
+ "Source" form shall mean the preferred form for making modifications,
+ including but not limited to software source code, documentation
+ source, and configuration files.
+
+ "Object" form shall mean any form resulting from mechanical
+ transformation or translation of a Source form, including but
+ not limited to compiled object code, generated documentation,
+ and conversions to other media types.
+
+ "Work" shall mean the work of authorship, whether in Source or
+ Object form, made available under the License, as indicated by a
+ copyright notice that is included in or attached to the work
+ (an example is provided in the Appendix below).
+
+ "Derivative Works" shall mean any work, whether in Source or Object
+ form, that is based on (or derived from) the Work and for which the
+ editorial revisions, annotations, elaborations, or other modifications
+ represent, as a whole, an original work of authorship. For the purposes
+ of this License, Derivative Works shall not include works that remain
+ separable from, or merely link (or bind by name) to the interfaces of,
+ the Work and Derivative Works thereof.
+
+ "Contribution" shall mean any work of authorship, including
+ the original version of the Work and any modifications or additions
+ to that Work or Derivative Works thereof, that is intentionally
+ submitted to Licensor for inclusion in the Work by the copyright owner
+ or by an individual or Legal Entity authorized to submit on behalf of
+ the copyright owner. For the purposes of this definition, "submitted"
+ means any form of electronic, verbal, or written communication sent
+ to the Licensor or its representatives, including but not limited to
+ communication on electronic mailing lists, source code control systems,
+ and issue tracking systems that are managed by, or on behalf of, the
+ Licensor for the purpose of discussing and improving the Work, but
+ excluding communication that is conspicuously marked or otherwise
+ designated in writing by the copyright owner as "Not a Contribution."
+
+ "Contributor" shall mean Licensor and any individual or Legal Entity
+ on behalf of whom a Contribution has been received by Licensor and
+ subsequently incorporated within the Work.
+
+ 2. Grant of Copyright License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ copyright license to reproduce, prepare Derivative Works of,
+ publicly display, publicly perform, sublicense, and distribute the
+ Work and such Derivative Works in Source or Object form.
+
+ 3. Grant of Patent License. Subject to the terms and conditions of
+ this License, each Contributor hereby grants to You a perpetual,
+ worldwide, non-exclusive, no-charge, royalty-free, irrevocable
+ (except as stated in this section) patent license to make, have made,
+ use, offer to sell, sell, import, and otherwise transfer the Work,
+ where such license applies only to those patent claims licensable
+ by such Contributor that are necessarily infringed by their
+ Contribution(s) alone or by combination of their Contribution(s)
+ with the Work to which such Contribution(s) was submitted. If You
+ institute patent litigation against any entity (including a
+ cross-claim or counterclaim in a lawsuit) alleging that the Work
+ or a Contribution incorporated within the Work constitutes direct
+ or contributory patent infringement, then any patent licenses
+ granted to You under this License for that Work shall terminate
+ as of the date such litigation is filed.
+
+ 4. Redistribution. You may reproduce and distribute copies of the
+ Work or Derivative Works thereof in any medium, with or without
+ modifications, and in Source or Object form, provided that You
+ meet the following conditions:
+
+ (a) You must give any other recipients of the Work or
+ Derivative Works a copy of this License; and
+
+ (b) You must cause any modified files to carry prominent notices
+ stating that You changed the files; and
+
+ (c) You must retain, in the Source form of any Derivative Works
+ that You distribute, all copyright, patent, trademark, and
+ attribution notices from the Source form of the Work,
+ excluding those notices that do not pertain to any part of
+ the Derivative Works; and
+
+ (d) If the Work includes a "NOTICE" text file as part of its
+ distribution, then any Derivative Works that You distribute must
+ include a readable copy of the attribution notices contained
+ within such NOTICE file, excluding those notices that do not
+ pertain to any part of the Derivative Works, in at least one
+ of the following places: within a NOTICE text file distributed
+ as part of the Derivative Works; within the Source form or
+ documentation, if provided along with the Derivative Works; or,
+ within a display generated by the Derivative Works, if and
+ wherever such third-party notices normally appear. The contents
+ of the NOTICE file are for informational purposes only and
+ do not modify the License. You may add Your own attribution
+ notices within Derivative Works that You distribute, alongside
+ or as an addendum to the NOTICE text from the Work, provided
+ that such additional attribution notices cannot be construed
+ as modifying the License.
+
+ You may add Your own copyright statement to Your modifications and
+ may provide additional or different license terms and conditions
+ for use, reproduction, or distribution of Your modifications, or
+ for any such Derivative Works as a whole, provided Your use,
+ reproduction, and distribution of the Work otherwise complies with
+ the conditions stated in this License.
+
+ 5. Submission of Contributions. Unless You explicitly state otherwise,
+ any Contribution intentionally submitted for inclusion in the Work
+ by You to the Licensor shall be under the terms and conditions of
+ this License, without any additional terms or conditions.
+ Notwithstanding the above, nothing herein shall supersede or modify
+ the terms of any separate license agreement you may have executed
+ with Licensor regarding such Contributions.
+
+ 6. Trademarks. This License does not grant permission to use the trade
+ names, trademarks, service marks, or product names of the Licensor,
+ except as required for reasonable and customary use in describing the
+ origin of the Work and reproducing the content of the NOTICE file.
+
+ 7. Disclaimer of Warranty. Unless required by applicable law or
+ agreed to in writing, Licensor provides the Work (and each
+ Contributor provides its Contributions) on an "AS IS" BASIS,
+ WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or
+ implied, including, without limitation, any warranties or conditions
+ of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A
+ PARTICULAR PURPOSE. You are solely responsible for determining the
+ appropriateness of using or redistributing the Work and assume any
+ risks associated with Your exercise of permissions under this License.
+
+ 8. Limitation of Liability. In no event and under no legal theory,
+ whether in tort (including negligence), contract, or otherwise,
+ unless required by applicable law (such as deliberate and grossly
+ negligent acts) or agreed to in writing, shall any Contributor be
+ liable to You for damages, including any direct, indirect, special,
+ incidental, or consequential damages of any character arising as a
+ result of this License or out of the use or inability to use the
+ Work (including but not limited to damages for loss of goodwill,
+ work stoppage, computer failure or malfunction, or any and all
+ other commercial damages or losses), even if such Contributor
+ has been advised of the possibility of such damages.
+
+ 9. Accepting Warranty or Additional Liability. While redistributing
+ the Work or Derivative Works thereof, You may choose to offer,
+ and charge a fee for, acceptance of support, warranty, indemnity,
+ or other liability obligations and/or rights consistent with this
+ License. However, in accepting such obligations, You may act only
+ on Your own behalf and on Your sole responsibility, not on behalf
+ of any other Contributor, and only if You agree to indemnify,
+ defend, and hold each Contributor harmless for any liability
+ incurred by, or claims asserted against, such Contributor by reason
+ of your accepting any such warranty or additional liability.
+
+ END OF TERMS AND CONDITIONS
+
+ Copyright 2014 Docker, Inc.
+
+ 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.
diff --git a/vendor/github.com/opencontainers/runc/NOTICE b/vendor/github.com/opencontainers/runc/NOTICE
new file mode 100644
index 000000000..5c97abce4
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/NOTICE
@@ -0,0 +1,17 @@
+runc
+
+Copyright 2012-2015 Docker, Inc.
+
+This product includes software developed at Docker, Inc. (http://www.docker.com).
+
+The following is courtesy of our legal counsel:
+
+
+Use and transfer of Docker may be subject to certain restrictions by the
+United States and other governments.
+It is your responsibility to ensure that your use and/or transfer does not
+violate applicable laws.
+
+For more information, please see http://www.bis.doc.gov
+
+See also http://www.apache.org/dev/crypto.html and/or seek legal counsel.
diff --git a/vendor/github.com/opencontainers/runc/README.md b/vendor/github.com/opencontainers/runc/README.md
new file mode 100644
index 000000000..eabfb982b
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/README.md
@@ -0,0 +1,244 @@
+# runc
+
+[![Build Status](https://travis-ci.org/opencontainers/runc.svg?branch=master)](https://travis-ci.org/opencontainers/runc)
+[![Go Report Card](https://goreportcard.com/badge/github.com/opencontainers/runc)](https://goreportcard.com/report/github.com/opencontainers/runc)
+[![GoDoc](https://godoc.org/github.com/opencontainers/runc?status.svg)](https://godoc.org/github.com/opencontainers/runc)
+
+## Introduction
+
+`runc` is a CLI tool for spawning and running containers according to the OCI specification.
+
+## Releases
+
+`runc` depends on and tracks the [runtime-spec](https://github.com/opencontainers/runtime-spec) repository.
+We will try to make sure that `runc` and the OCI specification major versions stay in lockstep.
+This means that `runc` 1.0.0 should implement the 1.0 version of the specification.
+
+You can find official releases of `runc` on the [release](https://github.com/opencontainers/runc/releases) page.
+
+### Security
+
+If you wish to report a security issue, please disclose the issue responsibly
+to security@opencontainers.org.
+
+## Building
+
+`runc` currently supports the Linux platform with various architecture support.
+It must be built with Go version 1.6 or higher in order for some features to function properly.
+
+In order to enable seccomp support you will need to install `libseccomp` on your platform.
+> e.g. `libseccomp-devel` for CentOS, or `libseccomp-dev` for Ubuntu
+
+Otherwise, if you do not want to build `runc` with seccomp support you can add `BUILDTAGS=""` when running make.
+
+```bash
+# create a 'github.com/opencontainers' in your GOPATH/src
+cd github.com/opencontainers
+git clone https://github.com/opencontainers/runc
+cd runc
+
+make
+sudo make install
+```
+
+`runc` will be installed to `/usr/local/sbin/runc` on your system.
+
+#### Build Tags
+
+`runc` supports optional build tags for compiling support of various features.
+To add build tags to the make option the `BUILDTAGS` variable must be set.
+
+```bash
+make BUILDTAGS='seccomp apparmor'
+```
+
+| Build Tag | Feature | Dependency |
+|-----------|------------------------------------|-------------|
+| seccomp | Syscall filtering | libseccomp |
+| selinux | selinux process and mount labeling | <none> |
+| apparmor | apparmor profile support | libapparmor |
+| ambient | ambient capability support | kernel 4.3 |
+
+
+### Running the test suite
+
+`runc` currently supports running its test suite via Docker.
+To run the suite just type `make test`.
+
+```bash
+make test
+```
+
+There are additional make targets for running the tests outside of a container but this is not recommended as the tests are written with the expectation that they can write and remove anywhere.
+
+You can run a specific test case by setting the `TESTFLAGS` variable.
+
+```bash
+# make test TESTFLAGS="-run=SomeTestFunction"
+```
+
+### Dependencies Management
+
+`runc` uses [vndr](https://github.com/LK4D4/vndr) for dependencies management.
+Please refer to [vndr](https://github.com/LK4D4/vndr) for how to add or update
+new dependencies.
+
+## Using runc
+
+### Creating an OCI Bundle
+
+In order to use runc you must have your container in the format of an OCI bundle.
+If you have Docker installed you can use its `export` method to acquire a root filesystem from an existing Docker container.
+
+```bash
+# create the top most bundle directory
+mkdir /mycontainer
+cd /mycontainer
+
+# create the rootfs directory
+mkdir rootfs
+
+# export busybox via Docker into the rootfs directory
+docker export $(docker create busybox) | tar -C rootfs -xvf -
+```
+
+After a root filesystem is populated you just generate a spec in the format of a `config.json` file inside your bundle.
+`runc` provides a `spec` command to generate a base template spec that you are then able to edit.
+To find features and documentation for fields in the spec please refer to the [specs](https://github.com/opencontainers/runtime-spec) repository.
+
+```bash
+runc spec
+```
+
+### Running Containers
+
+Assuming you have an OCI bundle from the previous step you can execute the container in two different ways.
+
+The first way is to use the convenience command `run` that will handle creating, starting, and deleting the container after it exits.
+
+```bash
+# run as root
+cd /mycontainer
+runc run mycontainerid
+```
+
+If you used the unmodified `runc spec` template this should give you a `sh` session inside the container.
+
+The second way to start a container is using the specs lifecycle operations.
+This gives you more power over how the container is created and managed while it is running.
+This will also launch the container in the background so you will have to edit the `config.json` to remove the `terminal` setting for the simple examples here.
+Your process field in the `config.json` should look like this below with `"terminal": false` and `"args": ["sleep", "5"]`.
+
+
+```json
+ "process": {
+ "terminal": false,
+ "user": {
+ "uid": 0,
+ "gid": 0
+ },
+ "args": [
+ "sleep", "5"
+ ],
+ "env": [
+ "PATH=/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin",
+ "TERM=xterm"
+ ],
+ "cwd": "/",
+ "capabilities": {
+ "bounding": [
+ "CAP_AUDIT_WRITE",
+ "CAP_KILL",
+ "CAP_NET_BIND_SERVICE"
+ ],
+ "effective": [
+ "CAP_AUDIT_WRITE",
+ "CAP_KILL",
+ "CAP_NET_BIND_SERVICE"
+ ],
+ "inheritable": [
+ "CAP_AUDIT_WRITE",
+ "CAP_KILL",
+ "CAP_NET_BIND_SERVICE"
+ ],
+ "permitted": [
+ "CAP_AUDIT_WRITE",
+ "CAP_KILL",
+ "CAP_NET_BIND_SERVICE"
+ ],
+ "ambient": [
+ "CAP_AUDIT_WRITE",
+ "CAP_KILL",
+ "CAP_NET_BIND_SERVICE"
+ ]
+ },
+ "rlimits": [
+ {
+ "type": "RLIMIT_NOFILE",
+ "hard": 1024,
+ "soft": 1024
+ }
+ ],
+ "noNewPrivileges": true
+ },
+```
+
+Now we can go through the lifecycle operations in your shell.
+
+
+```bash
+# run as root
+cd /mycontainer
+runc create mycontainerid
+
+# view the container is created and in the "created" state
+runc list
+
+# start the process inside the container
+runc start mycontainerid
+
+# after 5 seconds view that the container has exited and is now in the stopped state
+runc list
+
+# now delete the container
+runc delete mycontainerid
+```
+
+This adds more complexity but allows higher level systems to manage runc and provides points in the containers creation to setup various settings after the container has created and/or before it is deleted.
+This is commonly used to setup the container's network stack after `create` but before `start` where the user's defined process will be running.
+
+#### Rootless containers
+`runc` has the ability to run containers without root privileges. This is called `rootless`. You need to pass some parameters to `runc` in order to run rootless containers. See below and compare with the previous version. Run the following commands as an ordinary user:
+```bash
+# Same as the first example
+mkdir ~/mycontainer
+cd ~/mycontainer
+mkdir rootfs
+docker export $(docker create busybox) | tar -C rootfs -xvf -
+
+# The --rootless parameter instructs runc spec to generate a configuration for a rootless container, which will allow you to run the container as a non-root user.
+runc spec --rootless
+
+# The --root parameter tells runc where to store the container state. It must be writable by the user.
+runc --root /tmp/runc run mycontainerid
+```
+
+#### Supervisors
+
+`runc` can be used with process supervisors and init systems to ensure that containers are restarted when they exit.
+An example systemd unit file looks something like this.
+
+```systemd
+[Unit]
+Description=Start My Container
+
+[Service]
+Type=forking
+ExecStart=/usr/local/sbin/runc run -d --pid-file /run/mycontainerid.pid mycontainerid
+ExecStopPost=/usr/local/sbin/runc delete mycontainerid
+WorkingDirectory=/mycontainer
+PIDFile=/run/mycontainerid.pid
+
+[Install]
+WantedBy=multi-user.target
+```
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/README.md b/vendor/github.com/opencontainers/runc/libcontainer/README.md
new file mode 100644
index 000000000..42f3efe56
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/README.md
@@ -0,0 +1,328 @@
+# libcontainer
+
+[![GoDoc](https://godoc.org/github.com/opencontainers/runc/libcontainer?status.svg)](https://godoc.org/github.com/opencontainers/runc/libcontainer)
+
+Libcontainer provides a native Go implementation for creating containers
+with namespaces, cgroups, capabilities, and filesystem access controls.
+It allows you to manage the lifecycle of the container performing additional operations
+after the container is created.
+
+
+#### Container
+A container is a self contained execution environment that shares the kernel of the
+host system and which is (optionally) isolated from other containers in the system.
+
+#### Using libcontainer
+
+Because containers are spawned in a two step process you will need a binary that
+will be executed as the init process for the container. In libcontainer, we use
+the current binary (/proc/self/exe) to be executed as the init process, and use
+arg "init", we call the first step process "bootstrap", so you always need a "init"
+function as the entry of "bootstrap".
+
+In addition to the go init function the early stage bootstrap is handled by importing
+[nsenter](https://github.com/opencontainers/runc/blob/master/libcontainer/nsenter/README.md).
+
+```go
+import (
+ _ "github.com/opencontainers/runc/libcontainer/nsenter"
+)
+
+func init() {
+ if len(os.Args) > 1 && os.Args[1] == "init" {
+ runtime.GOMAXPROCS(1)
+ runtime.LockOSThread()
+ factory, _ := libcontainer.New("")
+ if err := factory.StartInitialization(); err != nil {
+ logrus.Fatal(err)
+ }
+ panic("--this line should have never been executed, congratulations--")
+ }
+}
+```
+
+Then to create a container you first have to initialize an instance of a factory
+that will handle the creation and initialization for a container.
+
+```go
+factory, err := libcontainer.New("/var/lib/container", libcontainer.Cgroupfs, libcontainer.InitArgs(os.Args[0], "init"))
+if err != nil {
+ logrus.Fatal(err)
+ return
+}
+```
+
+Once you have an instance of the factory created we can create a configuration
+struct describing how the container is to be created. A sample would look similar to this:
+
+```go
+defaultMountFlags := unix.MS_NOEXEC | unix.MS_NOSUID | unix.MS_NODEV
+config := &configs.Config{
+ Rootfs: "/your/path/to/rootfs",
+ Capabilities: &configs.Capabilities{
+ Bounding: []string{
+ "CAP_CHOWN",
+ "CAP_DAC_OVERRIDE",
+ "CAP_FSETID",
+ "CAP_FOWNER",
+ "CAP_MKNOD",
+ "CAP_NET_RAW",
+ "CAP_SETGID",
+ "CAP_SETUID",
+ "CAP_SETFCAP",
+ "CAP_SETPCAP",
+ "CAP_NET_BIND_SERVICE",
+ "CAP_SYS_CHROOT",
+ "CAP_KILL",
+ "CAP_AUDIT_WRITE",
+ },
+ Effective: []string{
+ "CAP_CHOWN",
+ "CAP_DAC_OVERRIDE",
+ "CAP_FSETID",
+ "CAP_FOWNER",
+ "CAP_MKNOD",
+ "CAP_NET_RAW",
+ "CAP_SETGID",
+ "CAP_SETUID",
+ "CAP_SETFCAP",
+ "CAP_SETPCAP",
+ "CAP_NET_BIND_SERVICE",
+ "CAP_SYS_CHROOT",
+ "CAP_KILL",
+ "CAP_AUDIT_WRITE",
+ },
+ Inheritable: []string{
+ "CAP_CHOWN",
+ "CAP_DAC_OVERRIDE",
+ "CAP_FSETID",
+ "CAP_FOWNER",
+ "CAP_MKNOD",
+ "CAP_NET_RAW",
+ "CAP_SETGID",
+ "CAP_SETUID",
+ "CAP_SETFCAP",
+ "CAP_SETPCAP",
+ "CAP_NET_BIND_SERVICE",
+ "CAP_SYS_CHROOT",
+ "CAP_KILL",
+ "CAP_AUDIT_WRITE",
+ },
+ Permitted: []string{
+ "CAP_CHOWN",
+ "CAP_DAC_OVERRIDE",
+ "CAP_FSETID",
+ "CAP_FOWNER",
+ "CAP_MKNOD",
+ "CAP_NET_RAW",
+ "CAP_SETGID",
+ "CAP_SETUID",
+ "CAP_SETFCAP",
+ "CAP_SETPCAP",
+ "CAP_NET_BIND_SERVICE",
+ "CAP_SYS_CHROOT",
+ "CAP_KILL",
+ "CAP_AUDIT_WRITE",
+ },
+ Ambient: []string{
+ "CAP_CHOWN",
+ "CAP_DAC_OVERRIDE",
+ "CAP_FSETID",
+ "CAP_FOWNER",
+ "CAP_MKNOD",
+ "CAP_NET_RAW",
+ "CAP_SETGID",
+ "CAP_SETUID",
+ "CAP_SETFCAP",
+ "CAP_SETPCAP",
+ "CAP_NET_BIND_SERVICE",
+ "CAP_SYS_CHROOT",
+ "CAP_KILL",
+ "CAP_AUDIT_WRITE",
+ },
+ },
+ Namespaces: configs.Namespaces([]configs.Namespace{
+ {Type: configs.NEWNS},
+ {Type: configs.NEWUTS},
+ {Type: configs.NEWIPC},
+ {Type: configs.NEWPID},
+ {Type: configs.NEWUSER},
+ {Type: configs.NEWNET},
+ }),
+ Cgroups: &configs.Cgroup{
+ Name: "test-container",
+ Parent: "system",
+ Resources: &configs.Resources{
+ MemorySwappiness: nil,
+ AllowAllDevices: nil,
+ AllowedDevices: configs.DefaultAllowedDevices,
+ },
+ },
+ MaskPaths: []string{
+ "/proc/kcore",
+ "/sys/firmware",
+ },
+ ReadonlyPaths: []string{
+ "/proc/sys", "/proc/sysrq-trigger", "/proc/irq", "/proc/bus",
+ },
+ Devices: configs.DefaultAutoCreatedDevices,
+ Hostname: "testing",
+ Mounts: []*configs.Mount{
+ {
+ Source: "proc",
+ Destination: "/proc",
+ Device: "proc",
+ Flags: defaultMountFlags,
+ },
+ {
+ Source: "tmpfs",
+ Destination: "/dev",
+ Device: "tmpfs",
+ Flags: unix.MS_NOSUID | unix.MS_STRICTATIME,
+ Data: "mode=755",
+ },
+ {
+ Source: "devpts",
+ Destination: "/dev/pts",
+ Device: "devpts",
+ Flags: unix.MS_NOSUID | unix.MS_NOEXEC,
+ Data: "newinstance,ptmxmode=0666,mode=0620,gid=5",
+ },
+ {
+ Device: "tmpfs",
+ Source: "shm",
+ Destination: "/dev/shm",
+ Data: "mode=1777,size=65536k",
+ Flags: defaultMountFlags,
+ },
+ {
+ Source: "mqueue",
+ Destination: "/dev/mqueue",
+ Device: "mqueue",
+ Flags: defaultMountFlags,
+ },
+ {
+ Source: "sysfs",
+ Destination: "/sys",
+ Device: "sysfs",
+ Flags: defaultMountFlags | unix.MS_RDONLY,
+ },
+ },
+ UidMappings: []configs.IDMap{
+ {
+ ContainerID: 0,
+ HostID: 1000,
+ Size: 65536,
+ },
+ },
+ GidMappings: []configs.IDMap{
+ {
+ ContainerID: 0,
+ HostID: 1000,
+ Size: 65536,
+ },
+ },
+ Networks: []*configs.Network{
+ {
+ Type: "loopback",
+ Address: "127.0.0.1/0",
+ Gateway: "localhost",
+ },
+ },
+ Rlimits: []configs.Rlimit{
+ {
+ Type: unix.RLIMIT_NOFILE,
+ Hard: uint64(1025),
+ Soft: uint64(1025),
+ },
+ },
+}
+```
+
+Once you have the configuration populated you can create a container:
+
+```go
+container, err := factory.Create("container-id", config)
+if err != nil {
+ logrus.Fatal(err)
+ return
+}
+```
+
+To spawn bash as the initial process inside the container and have the
+processes pid returned in order to wait, signal, or kill the process:
+
+```go
+process := &libcontainer.Process{
+ Args: []string{"/bin/bash"},
+ Env: []string{"PATH=/bin"},
+ User: "daemon",
+ Stdin: os.Stdin,
+ Stdout: os.Stdout,
+ Stderr: os.Stderr,
+}
+
+err := container.Run(process)
+if err != nil {
+ container.Destroy()
+ logrus.Fatal(err)
+ return
+}
+
+// wait for the process to finish.
+_, err := process.Wait()
+if err != nil {
+ logrus.Fatal(err)
+}
+
+// destroy the container.
+container.Destroy()
+```
+
+Additional ways to interact with a running container are:
+
+```go
+// return all the pids for all processes running inside the container.
+processes, err := container.Processes()
+
+// get detailed cpu, memory, io, and network statistics for the container and
+// it's processes.
+stats, err := container.Stats()
+
+// pause all processes inside the container.
+container.Pause()
+
+// resume all paused processes.
+container.Resume()
+
+// send signal to container's init process.
+container.Signal(signal)
+
+// update container resource constraints.
+container.Set(config)
+
+// get current status of the container.
+status, err := container.Status()
+
+// get current container's state information.
+state, err := container.State()
+```
+
+
+#### Checkpoint & Restore
+
+libcontainer now integrates [CRIU](http://criu.org/) for checkpointing and restoring containers.
+This let's you save the state of a process running inside a container to disk, and then restore
+that state into a new process, on the same machine or on another machine.
+
+`criu` version 1.5.2 or higher is required to use checkpoint and restore.
+If you don't already have `criu` installed, you can build it from source, following the
+[online instructions](http://criu.org/Installation). `criu` is also installed in the docker image
+generated when building libcontainer with docker.
+
+
+## Copyright and license
+
+Code and documentation copyright 2014 Docker, inc. Code released under the Apache 2.0 license.
+Docs released under Creative commons.
+
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor.go b/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor.go
new file mode 100644
index 000000000..82ed1a68a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor.go
@@ -0,0 +1,39 @@
+// +build apparmor,linux
+
+package apparmor
+
+// #cgo LDFLAGS: -lapparmor
+// #include <sys/apparmor.h>
+// #include <stdlib.h>
+import "C"
+import (
+ "fmt"
+ "io/ioutil"
+ "os"
+ "unsafe"
+)
+
+// IsEnabled returns true if apparmor is enabled for the host.
+func IsEnabled() bool {
+ if _, err := os.Stat("/sys/kernel/security/apparmor"); err == nil && os.Getenv("container") == "" {
+ if _, err = os.Stat("/sbin/apparmor_parser"); err == nil {
+ buf, err := ioutil.ReadFile("/sys/module/apparmor/parameters/enabled")
+ return err == nil && len(buf) > 1 && buf[0] == 'Y'
+ }
+ }
+ return false
+}
+
+// ApplyProfile will apply the profile with the specified name to the process after
+// the next exec.
+func ApplyProfile(name string) error {
+ if name == "" {
+ return nil
+ }
+ cName := C.CString(name)
+ defer C.free(unsafe.Pointer(cName))
+ if _, err := C.aa_change_onexec(cName); err != nil {
+ return fmt.Errorf("apparmor failed to apply profile: %s", err)
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor_disabled.go b/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor_disabled.go
new file mode 100644
index 000000000..d4110cf0b
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/apparmor/apparmor_disabled.go
@@ -0,0 +1,20 @@
+// +build !apparmor !linux
+
+package apparmor
+
+import (
+ "errors"
+)
+
+var ErrApparmorNotEnabled = errors.New("apparmor: config provided but apparmor not supported")
+
+func IsEnabled() bool {
+ return false
+}
+
+func ApplyProfile(name string) error {
+ if name != "" {
+ return ErrApparmorNotEnabled
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/capabilities_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/capabilities_linux.go
new file mode 100644
index 000000000..8981b2a2f
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/capabilities_linux.go
@@ -0,0 +1,114 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/syndtr/gocapability/capability"
+)
+
+const allCapabilityTypes = capability.CAPS | capability.BOUNDS | capability.AMBS
+
+var capabilityMap map[string]capability.Cap
+
+func init() {
+ capabilityMap = make(map[string]capability.Cap)
+ last := capability.CAP_LAST_CAP
+ // workaround for RHEL6 which has no /proc/sys/kernel/cap_last_cap
+ if last == capability.Cap(63) {
+ last = capability.CAP_BLOCK_SUSPEND
+ }
+ for _, cap := range capability.List() {
+ if cap > last {
+ continue
+ }
+ capKey := fmt.Sprintf("CAP_%s", strings.ToUpper(cap.String()))
+ capabilityMap[capKey] = cap
+ }
+}
+
+func newContainerCapList(capConfig *configs.Capabilities) (*containerCapabilities, error) {
+ bounding := []capability.Cap{}
+ for _, c := range capConfig.Bounding {
+ v, ok := capabilityMap[c]
+ if !ok {
+ return nil, fmt.Errorf("unknown capability %q", c)
+ }
+ bounding = append(bounding, v)
+ }
+ effective := []capability.Cap{}
+ for _, c := range capConfig.Effective {
+ v, ok := capabilityMap[c]
+ if !ok {
+ return nil, fmt.Errorf("unknown capability %q", c)
+ }
+ effective = append(effective, v)
+ }
+ inheritable := []capability.Cap{}
+ for _, c := range capConfig.Inheritable {
+ v, ok := capabilityMap[c]
+ if !ok {
+ return nil, fmt.Errorf("unknown capability %q", c)
+ }
+ inheritable = append(inheritable, v)
+ }
+ permitted := []capability.Cap{}
+ for _, c := range capConfig.Permitted {
+ v, ok := capabilityMap[c]
+ if !ok {
+ return nil, fmt.Errorf("unknown capability %q", c)
+ }
+ permitted = append(permitted, v)
+ }
+ ambient := []capability.Cap{}
+ for _, c := range capConfig.Ambient {
+ v, ok := capabilityMap[c]
+ if !ok {
+ return nil, fmt.Errorf("unknown capability %q", c)
+ }
+ ambient = append(ambient, v)
+ }
+ pid, err := capability.NewPid(os.Getpid())
+ if err != nil {
+ return nil, err
+ }
+ return &containerCapabilities{
+ bounding: bounding,
+ effective: effective,
+ inheritable: inheritable,
+ permitted: permitted,
+ ambient: ambient,
+ pid: pid,
+ }, nil
+}
+
+type containerCapabilities struct {
+ pid capability.Capabilities
+ bounding []capability.Cap
+ effective []capability.Cap
+ inheritable []capability.Cap
+ permitted []capability.Cap
+ ambient []capability.Cap
+}
+
+// ApplyBoundingSet sets the capability bounding set to those specified in the whitelist.
+func (c *containerCapabilities) ApplyBoundingSet() error {
+ c.pid.Clear(capability.BOUNDS)
+ c.pid.Set(capability.BOUNDS, c.bounding...)
+ return c.pid.Apply(capability.BOUNDS)
+}
+
+// Apply sets all the capabilities for the current process in the config.
+func (c *containerCapabilities) ApplyCaps() error {
+ c.pid.Clear(allCapabilityTypes)
+ c.pid.Set(capability.BOUNDS, c.bounding...)
+ c.pid.Set(capability.PERMITTED, c.permitted...)
+ c.pid.Set(capability.INHERITABLE, c.inheritable...)
+ c.pid.Set(capability.EFFECTIVE, c.effective...)
+ c.pid.Set(capability.AMBIENT, c.ambient...)
+ return c.pid.Apply(allCapabilityTypes)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups.go
new file mode 100644
index 000000000..25ff51589
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups.go
@@ -0,0 +1,64 @@
+// +build linux
+
+package cgroups
+
+import (
+ "fmt"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type Manager interface {
+ // Applies cgroup configuration to the process with the specified pid
+ Apply(pid int) error
+
+ // Returns the PIDs inside the cgroup set
+ GetPids() ([]int, error)
+
+ // Returns the PIDs inside the cgroup set & all sub-cgroups
+ GetAllPids() ([]int, error)
+
+ // Returns statistics for the cgroup set
+ GetStats() (*Stats, error)
+
+ // Toggles the freezer cgroup according with specified state
+ Freeze(state configs.FreezerState) error
+
+ // Destroys the cgroup set
+ Destroy() error
+
+ // The option func SystemdCgroups() and Cgroupfs() require following attributes:
+ // Paths map[string]string
+ // Cgroups *configs.Cgroup
+ // Paths maps cgroup subsystem to path at which it is mounted.
+ // Cgroups specifies specific cgroup settings for the various subsystems
+
+ // Returns cgroup paths to save in a state file and to be able to
+ // restore the object later.
+ GetPaths() map[string]string
+
+ // Sets the cgroup as configured.
+ Set(container *configs.Config) error
+}
+
+type NotFoundError struct {
+ Subsystem string
+}
+
+func (e *NotFoundError) Error() string {
+ return fmt.Sprintf("mountpoint for %s not found", e.Subsystem)
+}
+
+func NewNotFoundError(sub string) error {
+ return &NotFoundError{
+ Subsystem: sub,
+ }
+}
+
+func IsNotFound(err error) bool {
+ if err == nil {
+ return false
+ }
+ _, ok := err.(*NotFoundError)
+ return ok
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups_unsupported.go
new file mode 100644
index 000000000..278d507e2
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/cgroups_unsupported.go
@@ -0,0 +1,3 @@
+// +build !linux
+
+package cgroups
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/apply_raw.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/apply_raw.go
new file mode 100644
index 000000000..22d82acb4
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/apply_raw.go
@@ -0,0 +1,360 @@
+// +build linux
+
+package fs
+
+import (
+ "errors"
+ "fmt"
+ "io"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+ "sync"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ libcontainerUtils "github.com/opencontainers/runc/libcontainer/utils"
+)
+
+var (
+ subsystems = subsystemSet{
+ &CpusetGroup{},
+ &DevicesGroup{},
+ &MemoryGroup{},
+ &CpuGroup{},
+ &CpuacctGroup{},
+ &PidsGroup{},
+ &BlkioGroup{},
+ &HugetlbGroup{},
+ &NetClsGroup{},
+ &NetPrioGroup{},
+ &PerfEventGroup{},
+ &FreezerGroup{},
+ &NameGroup{GroupName: "name=systemd", Join: true},
+ }
+ HugePageSizes, _ = cgroups.GetHugePageSize()
+)
+
+var errSubsystemDoesNotExist = errors.New("cgroup: subsystem does not exist")
+
+type subsystemSet []subsystem
+
+func (s subsystemSet) Get(name string) (subsystem, error) {
+ for _, ss := range s {
+ if ss.Name() == name {
+ return ss, nil
+ }
+ }
+ return nil, errSubsystemDoesNotExist
+}
+
+type subsystem interface {
+ // Name returns the name of the subsystem.
+ Name() string
+ // Returns the stats, as 'stats', corresponding to the cgroup under 'path'.
+ GetStats(path string, stats *cgroups.Stats) error
+ // Removes the cgroup represented by 'cgroupData'.
+ Remove(*cgroupData) error
+ // Creates and joins the cgroup represented by 'cgroupData'.
+ Apply(*cgroupData) error
+ // Set the cgroup represented by cgroup.
+ Set(path string, cgroup *configs.Cgroup) error
+}
+
+type Manager struct {
+ mu sync.Mutex
+ Cgroups *configs.Cgroup
+ Paths map[string]string
+}
+
+// The absolute path to the root of the cgroup hierarchies.
+var cgroupRootLock sync.Mutex
+var cgroupRoot string
+
+// Gets the cgroupRoot.
+func getCgroupRoot() (string, error) {
+ cgroupRootLock.Lock()
+ defer cgroupRootLock.Unlock()
+
+ if cgroupRoot != "" {
+ return cgroupRoot, nil
+ }
+
+ root, err := cgroups.FindCgroupMountpointDir()
+ if err != nil {
+ return "", err
+ }
+
+ if _, err := os.Stat(root); err != nil {
+ return "", err
+ }
+
+ cgroupRoot = root
+ return cgroupRoot, nil
+}
+
+type cgroupData struct {
+ root string
+ innerPath string
+ config *configs.Cgroup
+ pid int
+}
+
+func (m *Manager) Apply(pid int) (err error) {
+ if m.Cgroups == nil {
+ return nil
+ }
+ m.mu.Lock()
+ defer m.mu.Unlock()
+
+ var c = m.Cgroups
+
+ d, err := getCgroupData(m.Cgroups, pid)
+ if err != nil {
+ return err
+ }
+
+ m.Paths = make(map[string]string)
+ if c.Paths != nil {
+ for name, path := range c.Paths {
+ _, err := d.path(name)
+ if err != nil {
+ if cgroups.IsNotFound(err) {
+ continue
+ }
+ return err
+ }
+ m.Paths[name] = path
+ }
+ return cgroups.EnterPid(m.Paths, pid)
+ }
+
+ for _, sys := range subsystems {
+ // TODO: Apply should, ideally, be reentrant or be broken up into a separate
+ // create and join phase so that the cgroup hierarchy for a container can be
+ // created then join consists of writing the process pids to cgroup.procs
+ p, err := d.path(sys.Name())
+ if err != nil {
+ // The non-presence of the devices subsystem is
+ // considered fatal for security reasons.
+ if cgroups.IsNotFound(err) && sys.Name() != "devices" {
+ continue
+ }
+ return err
+ }
+ m.Paths[sys.Name()] = p
+
+ if err := sys.Apply(d); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (m *Manager) Destroy() error {
+ if m.Cgroups.Paths != nil {
+ return nil
+ }
+ m.mu.Lock()
+ defer m.mu.Unlock()
+ if err := cgroups.RemovePaths(m.Paths); err != nil {
+ return err
+ }
+ m.Paths = make(map[string]string)
+ return nil
+}
+
+func (m *Manager) GetPaths() map[string]string {
+ m.mu.Lock()
+ paths := m.Paths
+ m.mu.Unlock()
+ return paths
+}
+
+func (m *Manager) GetStats() (*cgroups.Stats, error) {
+ m.mu.Lock()
+ defer m.mu.Unlock()
+ stats := cgroups.NewStats()
+ for name, path := range m.Paths {
+ sys, err := subsystems.Get(name)
+ if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
+ continue
+ }
+ if err := sys.GetStats(path, stats); err != nil {
+ return nil, err
+ }
+ }
+ return stats, nil
+}
+
+func (m *Manager) Set(container *configs.Config) error {
+ // If Paths are set, then we are just joining cgroups paths
+ // and there is no need to set any values.
+ if m.Cgroups.Paths != nil {
+ return nil
+ }
+
+ paths := m.GetPaths()
+ for _, sys := range subsystems {
+ path := paths[sys.Name()]
+ if err := sys.Set(path, container.Cgroups); err != nil {
+ return err
+ }
+ }
+
+ if m.Paths["cpu"] != "" {
+ if err := CheckCpushares(m.Paths["cpu"], container.Cgroups.Resources.CpuShares); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// Freeze toggles the container's freezer cgroup depending on the state
+// provided
+func (m *Manager) Freeze(state configs.FreezerState) error {
+ paths := m.GetPaths()
+ dir := paths["freezer"]
+ prevState := m.Cgroups.Resources.Freezer
+ m.Cgroups.Resources.Freezer = state
+ freezer, err := subsystems.Get("freezer")
+ if err != nil {
+ return err
+ }
+ err = freezer.Set(dir, m.Cgroups)
+ if err != nil {
+ m.Cgroups.Resources.Freezer = prevState
+ return err
+ }
+ return nil
+}
+
+func (m *Manager) GetPids() ([]int, error) {
+ paths := m.GetPaths()
+ return cgroups.GetPids(paths["devices"])
+}
+
+func (m *Manager) GetAllPids() ([]int, error) {
+ paths := m.GetPaths()
+ return cgroups.GetAllPids(paths["devices"])
+}
+
+func getCgroupData(c *configs.Cgroup, pid int) (*cgroupData, error) {
+ root, err := getCgroupRoot()
+ if err != nil {
+ return nil, err
+ }
+
+ if (c.Name != "" || c.Parent != "") && c.Path != "" {
+ return nil, fmt.Errorf("cgroup: either Path or Name and Parent should be used")
+ }
+
+ // XXX: Do not remove this code. Path safety is important! -- cyphar
+ cgPath := libcontainerUtils.CleanPath(c.Path)
+ cgParent := libcontainerUtils.CleanPath(c.Parent)
+ cgName := libcontainerUtils.CleanPath(c.Name)
+
+ innerPath := cgPath
+ if innerPath == "" {
+ innerPath = filepath.Join(cgParent, cgName)
+ }
+
+ return &cgroupData{
+ root: root,
+ innerPath: innerPath,
+ config: c,
+ pid: pid,
+ }, nil
+}
+
+func (raw *cgroupData) path(subsystem string) (string, error) {
+ mnt, err := cgroups.FindCgroupMountpoint(subsystem)
+ // If we didn't mount the subsystem, there is no point we make the path.
+ if err != nil {
+ return "", err
+ }
+
+ // If the cgroup name/path is absolute do not look relative to the cgroup of the init process.
+ if filepath.IsAbs(raw.innerPath) {
+ // Sometimes subsystems can be mounted together as 'cpu,cpuacct'.
+ return filepath.Join(raw.root, filepath.Base(mnt), raw.innerPath), nil
+ }
+
+ // Use GetOwnCgroupPath instead of GetInitCgroupPath, because the creating
+ // process could in container and shared pid namespace with host, and
+ // /proc/1/cgroup could point to whole other world of cgroups.
+ parentPath, err := cgroups.GetOwnCgroupPath(subsystem)
+ if err != nil {
+ return "", err
+ }
+
+ return filepath.Join(parentPath, raw.innerPath), nil
+}
+
+func (raw *cgroupData) join(subsystem string) (string, error) {
+ path, err := raw.path(subsystem)
+ if err != nil {
+ return "", err
+ }
+ if err := os.MkdirAll(path, 0755); err != nil {
+ return "", err
+ }
+ if err := cgroups.WriteCgroupProc(path, raw.pid); err != nil {
+ return "", err
+ }
+ return path, nil
+}
+
+func writeFile(dir, file, data string) error {
+ // Normally dir should not be empty, one case is that cgroup subsystem
+ // is not mounted, we will get empty dir, and we want it fail here.
+ if dir == "" {
+ return fmt.Errorf("no such directory for %s", file)
+ }
+ if err := ioutil.WriteFile(filepath.Join(dir, file), []byte(data), 0700); err != nil {
+ return fmt.Errorf("failed to write %v to %v: %v", data, file, err)
+ }
+ return nil
+}
+
+func readFile(dir, file string) (string, error) {
+ data, err := ioutil.ReadFile(filepath.Join(dir, file))
+ return string(data), err
+}
+
+func removePath(p string, err error) error {
+ if err != nil {
+ return err
+ }
+ if p != "" {
+ return os.RemoveAll(p)
+ }
+ return nil
+}
+
+func CheckCpushares(path string, c uint64) error {
+ var cpuShares uint64
+
+ if c == 0 {
+ return nil
+ }
+
+ fd, err := os.Open(filepath.Join(path, "cpu.shares"))
+ if err != nil {
+ return err
+ }
+ defer fd.Close()
+
+ _, err = fmt.Fscanf(fd, "%d", &cpuShares)
+ if err != nil && err != io.EOF {
+ return err
+ }
+
+ if c > cpuShares {
+ return fmt.Errorf("The maximum allowed cpu-shares is %d", cpuShares)
+ } else if c < cpuShares {
+ return fmt.Errorf("The minimum allowed cpu-shares is %d", cpuShares)
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/blkio.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/blkio.go
new file mode 100644
index 000000000..a142cb991
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/blkio.go
@@ -0,0 +1,237 @@
+// +build linux
+
+package fs
+
+import (
+ "bufio"
+ "fmt"
+ "os"
+ "path/filepath"
+ "strconv"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type BlkioGroup struct {
+}
+
+func (s *BlkioGroup) Name() string {
+ return "blkio"
+}
+
+func (s *BlkioGroup) Apply(d *cgroupData) error {
+ _, err := d.join("blkio")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *BlkioGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.BlkioWeight != 0 {
+ if err := writeFile(path, "blkio.weight", strconv.FormatUint(uint64(cgroup.Resources.BlkioWeight), 10)); err != nil {
+ return err
+ }
+ }
+
+ if cgroup.Resources.BlkioLeafWeight != 0 {
+ if err := writeFile(path, "blkio.leaf_weight", strconv.FormatUint(uint64(cgroup.Resources.BlkioLeafWeight), 10)); err != nil {
+ return err
+ }
+ }
+ for _, wd := range cgroup.Resources.BlkioWeightDevice {
+ if err := writeFile(path, "blkio.weight_device", wd.WeightString()); err != nil {
+ return err
+ }
+ if err := writeFile(path, "blkio.leaf_weight_device", wd.LeafWeightString()); err != nil {
+ return err
+ }
+ }
+ for _, td := range cgroup.Resources.BlkioThrottleReadBpsDevice {
+ if err := writeFile(path, "blkio.throttle.read_bps_device", td.String()); err != nil {
+ return err
+ }
+ }
+ for _, td := range cgroup.Resources.BlkioThrottleWriteBpsDevice {
+ if err := writeFile(path, "blkio.throttle.write_bps_device", td.String()); err != nil {
+ return err
+ }
+ }
+ for _, td := range cgroup.Resources.BlkioThrottleReadIOPSDevice {
+ if err := writeFile(path, "blkio.throttle.read_iops_device", td.String()); err != nil {
+ return err
+ }
+ }
+ for _, td := range cgroup.Resources.BlkioThrottleWriteIOPSDevice {
+ if err := writeFile(path, "blkio.throttle.write_iops_device", td.String()); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *BlkioGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("blkio"))
+}
+
+/*
+examples:
+
+ blkio.sectors
+ 8:0 6792
+
+ blkio.io_service_bytes
+ 8:0 Read 1282048
+ 8:0 Write 2195456
+ 8:0 Sync 2195456
+ 8:0 Async 1282048
+ 8:0 Total 3477504
+ Total 3477504
+
+ blkio.io_serviced
+ 8:0 Read 124
+ 8:0 Write 104
+ 8:0 Sync 104
+ 8:0 Async 124
+ 8:0 Total 228
+ Total 228
+
+ blkio.io_queued
+ 8:0 Read 0
+ 8:0 Write 0
+ 8:0 Sync 0
+ 8:0 Async 0
+ 8:0 Total 0
+ Total 0
+*/
+
+func splitBlkioStatLine(r rune) bool {
+ return r == ' ' || r == ':'
+}
+
+func getBlkioStat(path string) ([]cgroups.BlkioStatEntry, error) {
+ var blkioStats []cgroups.BlkioStatEntry
+ f, err := os.Open(path)
+ if err != nil {
+ if os.IsNotExist(err) {
+ return blkioStats, nil
+ }
+ return nil, err
+ }
+ defer f.Close()
+
+ sc := bufio.NewScanner(f)
+ for sc.Scan() {
+ // format: dev type amount
+ fields := strings.FieldsFunc(sc.Text(), splitBlkioStatLine)
+ if len(fields) < 3 {
+ if len(fields) == 2 && fields[0] == "Total" {
+ // skip total line
+ continue
+ } else {
+ return nil, fmt.Errorf("Invalid line found while parsing %s: %s", path, sc.Text())
+ }
+ }
+
+ v, err := strconv.ParseUint(fields[0], 10, 64)
+ if err != nil {
+ return nil, err
+ }
+ major := v
+
+ v, err = strconv.ParseUint(fields[1], 10, 64)
+ if err != nil {
+ return nil, err
+ }
+ minor := v
+
+ op := ""
+ valueField := 2
+ if len(fields) == 4 {
+ op = fields[2]
+ valueField = 3
+ }
+ v, err = strconv.ParseUint(fields[valueField], 10, 64)
+ if err != nil {
+ return nil, err
+ }
+ blkioStats = append(blkioStats, cgroups.BlkioStatEntry{Major: major, Minor: minor, Op: op, Value: v})
+ }
+
+ return blkioStats, nil
+}
+
+func (s *BlkioGroup) GetStats(path string, stats *cgroups.Stats) error {
+ // Try to read CFQ stats available on all CFQ enabled kernels first
+ if blkioStats, err := getBlkioStat(filepath.Join(path, "blkio.io_serviced_recursive")); err == nil && blkioStats != nil {
+ return getCFQStats(path, stats)
+ }
+ return getStats(path, stats) // Use generic stats as fallback
+}
+
+func getCFQStats(path string, stats *cgroups.Stats) error {
+ var blkioStats []cgroups.BlkioStatEntry
+ var err error
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.sectors_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.SectorsRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_service_bytes_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoServiceBytesRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_serviced_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoServicedRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_queued_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoQueuedRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_service_time_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoServiceTimeRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_wait_time_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoWaitTimeRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.io_merged_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoMergedRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.time_recursive")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoTimeRecursive = blkioStats
+
+ return nil
+}
+
+func getStats(path string, stats *cgroups.Stats) error {
+ var blkioStats []cgroups.BlkioStatEntry
+ var err error
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.throttle.io_service_bytes")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoServiceBytesRecursive = blkioStats
+
+ if blkioStats, err = getBlkioStat(filepath.Join(path, "blkio.throttle.io_serviced")); err != nil {
+ return err
+ }
+ stats.BlkioStats.IoServicedRecursive = blkioStats
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpu.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpu.go
new file mode 100644
index 000000000..b712bd0b1
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpu.go
@@ -0,0 +1,125 @@
+// +build linux
+
+package fs
+
+import (
+ "bufio"
+ "os"
+ "path/filepath"
+ "strconv"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type CpuGroup struct {
+}
+
+func (s *CpuGroup) Name() string {
+ return "cpu"
+}
+
+func (s *CpuGroup) Apply(d *cgroupData) error {
+ // We always want to join the cpu group, to allow fair cpu scheduling
+ // on a container basis
+ path, err := d.path("cpu")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return s.ApplyDir(path, d.config, d.pid)
+}
+
+func (s *CpuGroup) ApplyDir(path string, cgroup *configs.Cgroup, pid int) error {
+ // This might happen if we have no cpu cgroup mounted.
+ // Just do nothing and don't fail.
+ if path == "" {
+ return nil
+ }
+ if err := os.MkdirAll(path, 0755); err != nil {
+ return err
+ }
+ // We should set the real-Time group scheduling settings before moving
+ // in the process because if the process is already in SCHED_RR mode
+ // and no RT bandwidth is set, adding it will fail.
+ if err := s.SetRtSched(path, cgroup); err != nil {
+ return err
+ }
+ // because we are not using d.join we need to place the pid into the procs file
+ // unlike the other subsystems
+ if err := cgroups.WriteCgroupProc(path, pid); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func (s *CpuGroup) SetRtSched(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.CpuRtPeriod != 0 {
+ if err := writeFile(path, "cpu.rt_period_us", strconv.FormatUint(cgroup.Resources.CpuRtPeriod, 10)); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.CpuRtRuntime != 0 {
+ if err := writeFile(path, "cpu.rt_runtime_us", strconv.FormatInt(cgroup.Resources.CpuRtRuntime, 10)); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (s *CpuGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.CpuShares != 0 {
+ if err := writeFile(path, "cpu.shares", strconv.FormatUint(cgroup.Resources.CpuShares, 10)); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.CpuPeriod != 0 {
+ if err := writeFile(path, "cpu.cfs_period_us", strconv.FormatUint(cgroup.Resources.CpuPeriod, 10)); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.CpuQuota != 0 {
+ if err := writeFile(path, "cpu.cfs_quota_us", strconv.FormatInt(cgroup.Resources.CpuQuota, 10)); err != nil {
+ return err
+ }
+ }
+ if err := s.SetRtSched(path, cgroup); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func (s *CpuGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("cpu"))
+}
+
+func (s *CpuGroup) GetStats(path string, stats *cgroups.Stats) error {
+ f, err := os.Open(filepath.Join(path, "cpu.stat"))
+ if err != nil {
+ if os.IsNotExist(err) {
+ return nil
+ }
+ return err
+ }
+ defer f.Close()
+
+ sc := bufio.NewScanner(f)
+ for sc.Scan() {
+ t, v, err := getCgroupParamKeyValue(sc.Text())
+ if err != nil {
+ return err
+ }
+ switch t {
+ case "nr_periods":
+ stats.CpuStats.ThrottlingData.Periods = v
+
+ case "nr_throttled":
+ stats.CpuStats.ThrottlingData.ThrottledPeriods = v
+
+ case "throttled_time":
+ stats.CpuStats.ThrottlingData.ThrottledTime = v
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuacct.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuacct.go
new file mode 100644
index 000000000..53afbaddf
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuacct.go
@@ -0,0 +1,121 @@
+// +build linux
+
+package fs
+
+import (
+ "fmt"
+ "io/ioutil"
+ "path/filepath"
+ "strconv"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/system"
+)
+
+const (
+ cgroupCpuacctStat = "cpuacct.stat"
+ nanosecondsInSecond = 1000000000
+)
+
+var clockTicks = uint64(system.GetClockTicks())
+
+type CpuacctGroup struct {
+}
+
+func (s *CpuacctGroup) Name() string {
+ return "cpuacct"
+}
+
+func (s *CpuacctGroup) Apply(d *cgroupData) error {
+ // we just want to join this group even though we don't set anything
+ if _, err := d.join("cpuacct"); err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+
+ return nil
+}
+
+func (s *CpuacctGroup) Set(path string, cgroup *configs.Cgroup) error {
+ return nil
+}
+
+func (s *CpuacctGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("cpuacct"))
+}
+
+func (s *CpuacctGroup) GetStats(path string, stats *cgroups.Stats) error {
+ userModeUsage, kernelModeUsage, err := getCpuUsageBreakdown(path)
+ if err != nil {
+ return err
+ }
+
+ totalUsage, err := getCgroupParamUint(path, "cpuacct.usage")
+ if err != nil {
+ return err
+ }
+
+ percpuUsage, err := getPercpuUsage(path)
+ if err != nil {
+ return err
+ }
+
+ stats.CpuStats.CpuUsage.TotalUsage = totalUsage
+ stats.CpuStats.CpuUsage.PercpuUsage = percpuUsage
+ stats.CpuStats.CpuUsage.UsageInUsermode = userModeUsage
+ stats.CpuStats.CpuUsage.UsageInKernelmode = kernelModeUsage
+ return nil
+}
+
+// Returns user and kernel usage breakdown in nanoseconds.
+func getCpuUsageBreakdown(path string) (uint64, uint64, error) {
+ userModeUsage := uint64(0)
+ kernelModeUsage := uint64(0)
+ const (
+ userField = "user"
+ systemField = "system"
+ )
+
+ // Expected format:
+ // user <usage in ticks>
+ // system <usage in ticks>
+ data, err := ioutil.ReadFile(filepath.Join(path, cgroupCpuacctStat))
+ if err != nil {
+ return 0, 0, err
+ }
+ fields := strings.Fields(string(data))
+ if len(fields) != 4 {
+ return 0, 0, fmt.Errorf("failure - %s is expected to have 4 fields", filepath.Join(path, cgroupCpuacctStat))
+ }
+ if fields[0] != userField {
+ return 0, 0, fmt.Errorf("unexpected field %q in %q, expected %q", fields[0], cgroupCpuacctStat, userField)
+ }
+ if fields[2] != systemField {
+ return 0, 0, fmt.Errorf("unexpected field %q in %q, expected %q", fields[2], cgroupCpuacctStat, systemField)
+ }
+ if userModeUsage, err = strconv.ParseUint(fields[1], 10, 64); err != nil {
+ return 0, 0, err
+ }
+ if kernelModeUsage, err = strconv.ParseUint(fields[3], 10, 64); err != nil {
+ return 0, 0, err
+ }
+
+ return (userModeUsage * nanosecondsInSecond) / clockTicks, (kernelModeUsage * nanosecondsInSecond) / clockTicks, nil
+}
+
+func getPercpuUsage(path string) ([]uint64, error) {
+ percpuUsage := []uint64{}
+ data, err := ioutil.ReadFile(filepath.Join(path, "cpuacct.usage_percpu"))
+ if err != nil {
+ return percpuUsage, err
+ }
+ for _, value := range strings.Fields(string(data)) {
+ value, err := strconv.ParseUint(value, 10, 64)
+ if err != nil {
+ return percpuUsage, fmt.Errorf("Unable to convert param value to uint64: %s", err)
+ }
+ percpuUsage = append(percpuUsage, value)
+ }
+ return percpuUsage, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuset.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuset.go
new file mode 100644
index 000000000..20c9eafac
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/cpuset.go
@@ -0,0 +1,163 @@
+// +build linux
+
+package fs
+
+import (
+ "bytes"
+ "fmt"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ libcontainerUtils "github.com/opencontainers/runc/libcontainer/utils"
+)
+
+type CpusetGroup struct {
+}
+
+func (s *CpusetGroup) Name() string {
+ return "cpuset"
+}
+
+func (s *CpusetGroup) Apply(d *cgroupData) error {
+ dir, err := d.path("cpuset")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return s.ApplyDir(dir, d.config, d.pid)
+}
+
+func (s *CpusetGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.CpusetCpus != "" {
+ if err := writeFile(path, "cpuset.cpus", cgroup.Resources.CpusetCpus); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.CpusetMems != "" {
+ if err := writeFile(path, "cpuset.mems", cgroup.Resources.CpusetMems); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (s *CpusetGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("cpuset"))
+}
+
+func (s *CpusetGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
+
+func (s *CpusetGroup) ApplyDir(dir string, cgroup *configs.Cgroup, pid int) error {
+ // This might happen if we have no cpuset cgroup mounted.
+ // Just do nothing and don't fail.
+ if dir == "" {
+ return nil
+ }
+ mountInfo, err := ioutil.ReadFile("/proc/self/mountinfo")
+ if err != nil {
+ return err
+ }
+ root := filepath.Dir(cgroups.GetClosestMountpointAncestor(dir, string(mountInfo)))
+ // 'ensureParent' start with parent because we don't want to
+ // explicitly inherit from parent, it could conflict with
+ // 'cpuset.cpu_exclusive'.
+ if err := s.ensureParent(filepath.Dir(dir), root); err != nil {
+ return err
+ }
+ if err := os.MkdirAll(dir, 0755); err != nil {
+ return err
+ }
+ // We didn't inherit cpuset configs from parent, but we have
+ // to ensure cpuset configs are set before moving task into the
+ // cgroup.
+ // The logic is, if user specified cpuset configs, use these
+ // specified configs, otherwise, inherit from parent. This makes
+ // cpuset configs work correctly with 'cpuset.cpu_exclusive', and
+ // keep backward compatbility.
+ if err := s.ensureCpusAndMems(dir, cgroup); err != nil {
+ return err
+ }
+
+ // because we are not using d.join we need to place the pid into the procs file
+ // unlike the other subsystems
+ if err := cgroups.WriteCgroupProc(dir, pid); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func (s *CpusetGroup) getSubsystemSettings(parent string) (cpus []byte, mems []byte, err error) {
+ if cpus, err = ioutil.ReadFile(filepath.Join(parent, "cpuset.cpus")); err != nil {
+ return
+ }
+ if mems, err = ioutil.ReadFile(filepath.Join(parent, "cpuset.mems")); err != nil {
+ return
+ }
+ return cpus, mems, nil
+}
+
+// ensureParent makes sure that the parent directory of current is created
+// and populated with the proper cpus and mems files copied from
+// it's parent.
+func (s *CpusetGroup) ensureParent(current, root string) error {
+ parent := filepath.Dir(current)
+ if libcontainerUtils.CleanPath(parent) == root {
+ return nil
+ }
+ // Avoid infinite recursion.
+ if parent == current {
+ return fmt.Errorf("cpuset: cgroup parent path outside cgroup root")
+ }
+ if err := s.ensureParent(parent, root); err != nil {
+ return err
+ }
+ if err := os.MkdirAll(current, 0755); err != nil {
+ return err
+ }
+ return s.copyIfNeeded(current, parent)
+}
+
+// copyIfNeeded copies the cpuset.cpus and cpuset.mems from the parent
+// directory to the current directory if the file's contents are 0
+func (s *CpusetGroup) copyIfNeeded(current, parent string) error {
+ var (
+ err error
+ currentCpus, currentMems []byte
+ parentCpus, parentMems []byte
+ )
+
+ if currentCpus, currentMems, err = s.getSubsystemSettings(current); err != nil {
+ return err
+ }
+ if parentCpus, parentMems, err = s.getSubsystemSettings(parent); err != nil {
+ return err
+ }
+
+ if s.isEmpty(currentCpus) {
+ if err := writeFile(current, "cpuset.cpus", string(parentCpus)); err != nil {
+ return err
+ }
+ }
+ if s.isEmpty(currentMems) {
+ if err := writeFile(current, "cpuset.mems", string(parentMems)); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (s *CpusetGroup) isEmpty(b []byte) bool {
+ return len(bytes.Trim(b, "\n")) == 0
+}
+
+func (s *CpusetGroup) ensureCpusAndMems(path string, cgroup *configs.Cgroup) error {
+ if err := s.Set(path, cgroup); err != nil {
+ return err
+ }
+ return s.copyIfNeeded(path, filepath.Dir(path))
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/devices.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/devices.go
new file mode 100644
index 000000000..0ac5b4ed7
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/devices.go
@@ -0,0 +1,80 @@
+// +build linux
+
+package fs
+
+import (
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/system"
+)
+
+type DevicesGroup struct {
+}
+
+func (s *DevicesGroup) Name() string {
+ return "devices"
+}
+
+func (s *DevicesGroup) Apply(d *cgroupData) error {
+ _, err := d.join("devices")
+ if err != nil {
+ // We will return error even it's `not found` error, devices
+ // cgroup is hard requirement for container's security.
+ return err
+ }
+ return nil
+}
+
+func (s *DevicesGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if system.RunningInUserNS() {
+ return nil
+ }
+
+ devices := cgroup.Resources.Devices
+ if len(devices) > 0 {
+ for _, dev := range devices {
+ file := "devices.deny"
+ if dev.Allow {
+ file = "devices.allow"
+ }
+ if err := writeFile(path, file, dev.CgroupString()); err != nil {
+ return err
+ }
+ }
+ return nil
+ }
+ if cgroup.Resources.AllowAllDevices != nil {
+ if *cgroup.Resources.AllowAllDevices == false {
+ if err := writeFile(path, "devices.deny", "a"); err != nil {
+ return err
+ }
+
+ for _, dev := range cgroup.Resources.AllowedDevices {
+ if err := writeFile(path, "devices.allow", dev.CgroupString()); err != nil {
+ return err
+ }
+ }
+ return nil
+ }
+
+ if err := writeFile(path, "devices.allow", "a"); err != nil {
+ return err
+ }
+ }
+
+ for _, dev := range cgroup.Resources.DeniedDevices {
+ if err := writeFile(path, "devices.deny", dev.CgroupString()); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *DevicesGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("devices"))
+}
+
+func (s *DevicesGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/freezer.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/freezer.go
new file mode 100644
index 000000000..e70dfe3b9
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/freezer.go
@@ -0,0 +1,61 @@
+// +build linux
+
+package fs
+
+import (
+ "fmt"
+ "strings"
+ "time"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type FreezerGroup struct {
+}
+
+func (s *FreezerGroup) Name() string {
+ return "freezer"
+}
+
+func (s *FreezerGroup) Apply(d *cgroupData) error {
+ _, err := d.join("freezer")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *FreezerGroup) Set(path string, cgroup *configs.Cgroup) error {
+ switch cgroup.Resources.Freezer {
+ case configs.Frozen, configs.Thawed:
+ if err := writeFile(path, "freezer.state", string(cgroup.Resources.Freezer)); err != nil {
+ return err
+ }
+
+ for {
+ state, err := readFile(path, "freezer.state")
+ if err != nil {
+ return err
+ }
+ if strings.TrimSpace(state) == string(cgroup.Resources.Freezer) {
+ break
+ }
+ time.Sleep(1 * time.Millisecond)
+ }
+ case configs.Undefined:
+ return nil
+ default:
+ return fmt.Errorf("Invalid argument '%s' to freezer.state", string(cgroup.Resources.Freezer))
+ }
+
+ return nil
+}
+
+func (s *FreezerGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("freezer"))
+}
+
+func (s *FreezerGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/fs_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/fs_unsupported.go
new file mode 100644
index 000000000..3ef9e0315
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/fs_unsupported.go
@@ -0,0 +1,3 @@
+// +build !linux
+
+package fs
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/hugetlb.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/hugetlb.go
new file mode 100644
index 000000000..2f9727719
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/hugetlb.go
@@ -0,0 +1,71 @@
+// +build linux
+
+package fs
+
+import (
+ "fmt"
+ "strconv"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type HugetlbGroup struct {
+}
+
+func (s *HugetlbGroup) Name() string {
+ return "hugetlb"
+}
+
+func (s *HugetlbGroup) Apply(d *cgroupData) error {
+ _, err := d.join("hugetlb")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *HugetlbGroup) Set(path string, cgroup *configs.Cgroup) error {
+ for _, hugetlb := range cgroup.Resources.HugetlbLimit {
+ if err := writeFile(path, strings.Join([]string{"hugetlb", hugetlb.Pagesize, "limit_in_bytes"}, "."), strconv.FormatUint(hugetlb.Limit, 10)); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *HugetlbGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("hugetlb"))
+}
+
+func (s *HugetlbGroup) GetStats(path string, stats *cgroups.Stats) error {
+ hugetlbStats := cgroups.HugetlbStats{}
+ for _, pageSize := range HugePageSizes {
+ usage := strings.Join([]string{"hugetlb", pageSize, "usage_in_bytes"}, ".")
+ value, err := getCgroupParamUint(path, usage)
+ if err != nil {
+ return fmt.Errorf("failed to parse %s - %v", usage, err)
+ }
+ hugetlbStats.Usage = value
+
+ maxUsage := strings.Join([]string{"hugetlb", pageSize, "max_usage_in_bytes"}, ".")
+ value, err = getCgroupParamUint(path, maxUsage)
+ if err != nil {
+ return fmt.Errorf("failed to parse %s - %v", maxUsage, err)
+ }
+ hugetlbStats.MaxUsage = value
+
+ failcnt := strings.Join([]string{"hugetlb", pageSize, "failcnt"}, ".")
+ value, err = getCgroupParamUint(path, failcnt)
+ if err != nil {
+ return fmt.Errorf("failed to parse %s - %v", failcnt, err)
+ }
+ hugetlbStats.Failcnt = value
+
+ stats.HugetlbStats[pageSize] = hugetlbStats
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/memory.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/memory.go
new file mode 100644
index 000000000..ad395a5d6
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/memory.go
@@ -0,0 +1,313 @@
+// +build linux
+
+package fs
+
+import (
+ "bufio"
+ "fmt"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+ "strconv"
+ "strings"
+ "syscall" // only for Errno
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+
+ "golang.org/x/sys/unix"
+)
+
+const (
+ cgroupKernelMemoryLimit = "memory.kmem.limit_in_bytes"
+ cgroupMemorySwapLimit = "memory.memsw.limit_in_bytes"
+ cgroupMemoryLimit = "memory.limit_in_bytes"
+)
+
+type MemoryGroup struct {
+}
+
+func (s *MemoryGroup) Name() string {
+ return "memory"
+}
+
+func (s *MemoryGroup) Apply(d *cgroupData) (err error) {
+ path, err := d.path("memory")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ } else if path == "" {
+ return nil
+ }
+ if memoryAssigned(d.config) {
+ if _, err := os.Stat(path); os.IsNotExist(err) {
+ if err := os.MkdirAll(path, 0755); err != nil {
+ return err
+ }
+ // Only enable kernel memory accouting when this cgroup
+ // is created by libcontainer, otherwise we might get
+ // error when people use `cgroupsPath` to join an existed
+ // cgroup whose kernel memory is not initialized.
+ if err := EnableKernelMemoryAccounting(path); err != nil {
+ return err
+ }
+ }
+ }
+ defer func() {
+ if err != nil {
+ os.RemoveAll(path)
+ }
+ }()
+
+ // We need to join memory cgroup after set memory limits, because
+ // kmem.limit_in_bytes can only be set when the cgroup is empty.
+ _, err = d.join("memory")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func EnableKernelMemoryAccounting(path string) error {
+ // Check if kernel memory is enabled
+ // We have to limit the kernel memory here as it won't be accounted at all
+ // until a limit is set on the cgroup and limit cannot be set once the
+ // cgroup has children, or if there are already tasks in the cgroup.
+ for _, i := range []int64{1, -1} {
+ if err := setKernelMemory(path, i); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func setKernelMemory(path string, kernelMemoryLimit int64) error {
+ if path == "" {
+ return fmt.Errorf("no such directory for %s", cgroupKernelMemoryLimit)
+ }
+ if !cgroups.PathExists(filepath.Join(path, cgroupKernelMemoryLimit)) {
+ // kernel memory is not enabled on the system so we should do nothing
+ return nil
+ }
+ if err := ioutil.WriteFile(filepath.Join(path, cgroupKernelMemoryLimit), []byte(strconv.FormatInt(kernelMemoryLimit, 10)), 0700); err != nil {
+ // Check if the error number returned by the syscall is "EBUSY"
+ // The EBUSY signal is returned on attempts to write to the
+ // memory.kmem.limit_in_bytes file if the cgroup has children or
+ // once tasks have been attached to the cgroup
+ if pathErr, ok := err.(*os.PathError); ok {
+ if errNo, ok := pathErr.Err.(syscall.Errno); ok {
+ if errNo == unix.EBUSY {
+ return fmt.Errorf("failed to set %s, because either tasks have already joined this cgroup or it has children", cgroupKernelMemoryLimit)
+ }
+ }
+ }
+ return fmt.Errorf("failed to write %v to %v: %v", kernelMemoryLimit, cgroupKernelMemoryLimit, err)
+ }
+ return nil
+}
+
+func setMemoryAndSwap(path string, cgroup *configs.Cgroup) error {
+ // If the memory update is set to -1 we should also
+ // set swap to -1, it means unlimited memory.
+ if cgroup.Resources.Memory == -1 {
+ // Only set swap if it's enabled in kernel
+ if cgroups.PathExists(filepath.Join(path, cgroupMemorySwapLimit)) {
+ cgroup.Resources.MemorySwap = -1
+ }
+ }
+
+ // When memory and swap memory are both set, we need to handle the cases
+ // for updating container.
+ if cgroup.Resources.Memory != 0 && cgroup.Resources.MemorySwap != 0 {
+ memoryUsage, err := getMemoryData(path, "")
+ if err != nil {
+ return err
+ }
+
+ // When update memory limit, we should adapt the write sequence
+ // for memory and swap memory, so it won't fail because the new
+ // value and the old value don't fit kernel's validation.
+ if cgroup.Resources.MemorySwap == -1 || memoryUsage.Limit < uint64(cgroup.Resources.MemorySwap) {
+ if err := writeFile(path, cgroupMemorySwapLimit, strconv.FormatInt(cgroup.Resources.MemorySwap, 10)); err != nil {
+ return err
+ }
+ if err := writeFile(path, cgroupMemoryLimit, strconv.FormatInt(cgroup.Resources.Memory, 10)); err != nil {
+ return err
+ }
+ } else {
+ if err := writeFile(path, cgroupMemoryLimit, strconv.FormatInt(cgroup.Resources.Memory, 10)); err != nil {
+ return err
+ }
+ if err := writeFile(path, cgroupMemorySwapLimit, strconv.FormatInt(cgroup.Resources.MemorySwap, 10)); err != nil {
+ return err
+ }
+ }
+ } else {
+ if cgroup.Resources.Memory != 0 {
+ if err := writeFile(path, cgroupMemoryLimit, strconv.FormatInt(cgroup.Resources.Memory, 10)); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.MemorySwap != 0 {
+ if err := writeFile(path, cgroupMemorySwapLimit, strconv.FormatInt(cgroup.Resources.MemorySwap, 10)); err != nil {
+ return err
+ }
+ }
+ }
+
+ return nil
+}
+
+func (s *MemoryGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if err := setMemoryAndSwap(path, cgroup); err != nil {
+ return err
+ }
+
+ if cgroup.Resources.KernelMemory != 0 {
+ if err := setKernelMemory(path, cgroup.Resources.KernelMemory); err != nil {
+ return err
+ }
+ }
+
+ if cgroup.Resources.MemoryReservation != 0 {
+ if err := writeFile(path, "memory.soft_limit_in_bytes", strconv.FormatInt(cgroup.Resources.MemoryReservation, 10)); err != nil {
+ return err
+ }
+ }
+
+ if cgroup.Resources.KernelMemoryTCP != 0 {
+ if err := writeFile(path, "memory.kmem.tcp.limit_in_bytes", strconv.FormatInt(cgroup.Resources.KernelMemoryTCP, 10)); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.OomKillDisable {
+ if err := writeFile(path, "memory.oom_control", "1"); err != nil {
+ return err
+ }
+ }
+ if cgroup.Resources.MemorySwappiness == nil || int64(*cgroup.Resources.MemorySwappiness) == -1 {
+ return nil
+ } else if *cgroup.Resources.MemorySwappiness <= 100 {
+ if err := writeFile(path, "memory.swappiness", strconv.FormatUint(*cgroup.Resources.MemorySwappiness, 10)); err != nil {
+ return err
+ }
+ } else {
+ return fmt.Errorf("invalid value:%d. valid memory swappiness range is 0-100", *cgroup.Resources.MemorySwappiness)
+ }
+
+ return nil
+}
+
+func (s *MemoryGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("memory"))
+}
+
+func (s *MemoryGroup) GetStats(path string, stats *cgroups.Stats) error {
+ // Set stats from memory.stat.
+ statsFile, err := os.Open(filepath.Join(path, "memory.stat"))
+ if err != nil {
+ if os.IsNotExist(err) {
+ return nil
+ }
+ return err
+ }
+ defer statsFile.Close()
+
+ sc := bufio.NewScanner(statsFile)
+ for sc.Scan() {
+ t, v, err := getCgroupParamKeyValue(sc.Text())
+ if err != nil {
+ return fmt.Errorf("failed to parse memory.stat (%q) - %v", sc.Text(), err)
+ }
+ stats.MemoryStats.Stats[t] = v
+ }
+ stats.MemoryStats.Cache = stats.MemoryStats.Stats["cache"]
+
+ memoryUsage, err := getMemoryData(path, "")
+ if err != nil {
+ return err
+ }
+ stats.MemoryStats.Usage = memoryUsage
+ swapUsage, err := getMemoryData(path, "memsw")
+ if err != nil {
+ return err
+ }
+ stats.MemoryStats.SwapUsage = swapUsage
+ kernelUsage, err := getMemoryData(path, "kmem")
+ if err != nil {
+ return err
+ }
+ stats.MemoryStats.KernelUsage = kernelUsage
+ kernelTCPUsage, err := getMemoryData(path, "kmem.tcp")
+ if err != nil {
+ return err
+ }
+ stats.MemoryStats.KernelTCPUsage = kernelTCPUsage
+
+ useHierarchy := strings.Join([]string{"memory", "use_hierarchy"}, ".")
+ value, err := getCgroupParamUint(path, useHierarchy)
+ if err != nil {
+ return err
+ }
+ if value == 1 {
+ stats.MemoryStats.UseHierarchy = true
+ }
+ return nil
+}
+
+func memoryAssigned(cgroup *configs.Cgroup) bool {
+ return cgroup.Resources.Memory != 0 ||
+ cgroup.Resources.MemoryReservation != 0 ||
+ cgroup.Resources.MemorySwap > 0 ||
+ cgroup.Resources.KernelMemory > 0 ||
+ cgroup.Resources.KernelMemoryTCP > 0 ||
+ cgroup.Resources.OomKillDisable ||
+ (cgroup.Resources.MemorySwappiness != nil && int64(*cgroup.Resources.MemorySwappiness) != -1)
+}
+
+func getMemoryData(path, name string) (cgroups.MemoryData, error) {
+ memoryData := cgroups.MemoryData{}
+
+ moduleName := "memory"
+ if name != "" {
+ moduleName = strings.Join([]string{"memory", name}, ".")
+ }
+ usage := strings.Join([]string{moduleName, "usage_in_bytes"}, ".")
+ maxUsage := strings.Join([]string{moduleName, "max_usage_in_bytes"}, ".")
+ failcnt := strings.Join([]string{moduleName, "failcnt"}, ".")
+ limit := strings.Join([]string{moduleName, "limit_in_bytes"}, ".")
+
+ value, err := getCgroupParamUint(path, usage)
+ if err != nil {
+ if moduleName != "memory" && os.IsNotExist(err) {
+ return cgroups.MemoryData{}, nil
+ }
+ return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", usage, err)
+ }
+ memoryData.Usage = value
+ value, err = getCgroupParamUint(path, maxUsage)
+ if err != nil {
+ if moduleName != "memory" && os.IsNotExist(err) {
+ return cgroups.MemoryData{}, nil
+ }
+ return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", maxUsage, err)
+ }
+ memoryData.MaxUsage = value
+ value, err = getCgroupParamUint(path, failcnt)
+ if err != nil {
+ if moduleName != "memory" && os.IsNotExist(err) {
+ return cgroups.MemoryData{}, nil
+ }
+ return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", failcnt, err)
+ }
+ memoryData.Failcnt = value
+ value, err = getCgroupParamUint(path, limit)
+ if err != nil {
+ if moduleName != "memory" && os.IsNotExist(err) {
+ return cgroups.MemoryData{}, nil
+ }
+ return cgroups.MemoryData{}, fmt.Errorf("failed to parse %s - %v", limit, err)
+ }
+ memoryData.Limit = value
+
+ return memoryData, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/name.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/name.go
new file mode 100644
index 000000000..d8cf1d87c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/name.go
@@ -0,0 +1,40 @@
+// +build linux
+
+package fs
+
+import (
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type NameGroup struct {
+ GroupName string
+ Join bool
+}
+
+func (s *NameGroup) Name() string {
+ return s.GroupName
+}
+
+func (s *NameGroup) Apply(d *cgroupData) error {
+ if s.Join {
+ // ignore errors if the named cgroup does not exist
+ d.join(s.GroupName)
+ }
+ return nil
+}
+
+func (s *NameGroup) Set(path string, cgroup *configs.Cgroup) error {
+ return nil
+}
+
+func (s *NameGroup) Remove(d *cgroupData) error {
+ if s.Join {
+ removePath(d.path(s.GroupName))
+ }
+ return nil
+}
+
+func (s *NameGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_cls.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_cls.go
new file mode 100644
index 000000000..8e74b645e
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_cls.go
@@ -0,0 +1,43 @@
+// +build linux
+
+package fs
+
+import (
+ "strconv"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type NetClsGroup struct {
+}
+
+func (s *NetClsGroup) Name() string {
+ return "net_cls"
+}
+
+func (s *NetClsGroup) Apply(d *cgroupData) error {
+ _, err := d.join("net_cls")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *NetClsGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.NetClsClassid != 0 {
+ if err := writeFile(path, "net_cls.classid", strconv.FormatUint(uint64(cgroup.Resources.NetClsClassid), 10)); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *NetClsGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("net_cls"))
+}
+
+func (s *NetClsGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_prio.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_prio.go
new file mode 100644
index 000000000..d0ab2af89
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/net_prio.go
@@ -0,0 +1,41 @@
+// +build linux
+
+package fs
+
+import (
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type NetPrioGroup struct {
+}
+
+func (s *NetPrioGroup) Name() string {
+ return "net_prio"
+}
+
+func (s *NetPrioGroup) Apply(d *cgroupData) error {
+ _, err := d.join("net_prio")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *NetPrioGroup) Set(path string, cgroup *configs.Cgroup) error {
+ for _, prioMap := range cgroup.Resources.NetPrioIfpriomap {
+ if err := writeFile(path, "net_prio.ifpriomap", prioMap.CgroupString()); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *NetPrioGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("net_prio"))
+}
+
+func (s *NetPrioGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/perf_event.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/perf_event.go
new file mode 100644
index 000000000..5693676d3
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/perf_event.go
@@ -0,0 +1,35 @@
+// +build linux
+
+package fs
+
+import (
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type PerfEventGroup struct {
+}
+
+func (s *PerfEventGroup) Name() string {
+ return "perf_event"
+}
+
+func (s *PerfEventGroup) Apply(d *cgroupData) error {
+ // we just want to join this group even though we don't set anything
+ if _, err := d.join("perf_event"); err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *PerfEventGroup) Set(path string, cgroup *configs.Cgroup) error {
+ return nil
+}
+
+func (s *PerfEventGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("perf_event"))
+}
+
+func (s *PerfEventGroup) GetStats(path string, stats *cgroups.Stats) error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/pids.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/pids.go
new file mode 100644
index 000000000..f1e372055
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/pids.go
@@ -0,0 +1,73 @@
+// +build linux
+
+package fs
+
+import (
+ "fmt"
+ "path/filepath"
+ "strconv"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type PidsGroup struct {
+}
+
+func (s *PidsGroup) Name() string {
+ return "pids"
+}
+
+func (s *PidsGroup) Apply(d *cgroupData) error {
+ _, err := d.join("pids")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ return nil
+}
+
+func (s *PidsGroup) Set(path string, cgroup *configs.Cgroup) error {
+ if cgroup.Resources.PidsLimit != 0 {
+ // "max" is the fallback value.
+ limit := "max"
+
+ if cgroup.Resources.PidsLimit > 0 {
+ limit = strconv.FormatInt(cgroup.Resources.PidsLimit, 10)
+ }
+
+ if err := writeFile(path, "pids.max", limit); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func (s *PidsGroup) Remove(d *cgroupData) error {
+ return removePath(d.path("pids"))
+}
+
+func (s *PidsGroup) GetStats(path string, stats *cgroups.Stats) error {
+ current, err := getCgroupParamUint(path, "pids.current")
+ if err != nil {
+ return fmt.Errorf("failed to parse pids.current - %s", err)
+ }
+
+ maxString, err := getCgroupParamString(path, "pids.max")
+ if err != nil {
+ return fmt.Errorf("failed to parse pids.max - %s", err)
+ }
+
+ // Default if pids.max == "max" is 0 -- which represents "no limit".
+ var max uint64
+ if maxString != "max" {
+ max, err = parseUint(maxString, 10, 64)
+ if err != nil {
+ return fmt.Errorf("failed to parse pids.max - unable to parse %q as a uint from Cgroup file %q", maxString, filepath.Join(path, "pids.max"))
+ }
+ }
+
+ stats.PidsStats.Current = current
+ stats.PidsStats.Limit = max
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/utils.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/utils.go
new file mode 100644
index 000000000..5ff0a1615
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/fs/utils.go
@@ -0,0 +1,78 @@
+// +build linux
+
+package fs
+
+import (
+ "errors"
+ "fmt"
+ "io/ioutil"
+ "path/filepath"
+ "strconv"
+ "strings"
+)
+
+var (
+ ErrNotValidFormat = errors.New("line is not a valid key value format")
+)
+
+// Saturates negative values at zero and returns a uint64.
+// Due to kernel bugs, some of the memory cgroup stats can be negative.
+func parseUint(s string, base, bitSize int) (uint64, error) {
+ value, err := strconv.ParseUint(s, base, bitSize)
+ if err != nil {
+ intValue, intErr := strconv.ParseInt(s, base, bitSize)
+ // 1. Handle negative values greater than MinInt64 (and)
+ // 2. Handle negative values lesser than MinInt64
+ if intErr == nil && intValue < 0 {
+ return 0, nil
+ } else if intErr != nil && intErr.(*strconv.NumError).Err == strconv.ErrRange && intValue < 0 {
+ return 0, nil
+ }
+
+ return value, err
+ }
+
+ return value, nil
+}
+
+// Parses a cgroup param and returns as name, value
+// i.e. "io_service_bytes 1234" will return as io_service_bytes, 1234
+func getCgroupParamKeyValue(t string) (string, uint64, error) {
+ parts := strings.Fields(t)
+ switch len(parts) {
+ case 2:
+ value, err := parseUint(parts[1], 10, 64)
+ if err != nil {
+ return "", 0, fmt.Errorf("unable to convert param value (%q) to uint64: %v", parts[1], err)
+ }
+
+ return parts[0], value, nil
+ default:
+ return "", 0, ErrNotValidFormat
+ }
+}
+
+// Gets a single uint64 value from the specified cgroup file.
+func getCgroupParamUint(cgroupPath, cgroupFile string) (uint64, error) {
+ fileName := filepath.Join(cgroupPath, cgroupFile)
+ contents, err := ioutil.ReadFile(fileName)
+ if err != nil {
+ return 0, err
+ }
+
+ res, err := parseUint(strings.TrimSpace(string(contents)), 10, 64)
+ if err != nil {
+ return res, fmt.Errorf("unable to parse %q as a uint from Cgroup file %q", string(contents), fileName)
+ }
+ return res, nil
+}
+
+// Gets a string value from the specified cgroup file
+func getCgroupParamString(cgroupPath, cgroupFile string) (string, error) {
+ contents, err := ioutil.ReadFile(filepath.Join(cgroupPath, cgroupFile))
+ if err != nil {
+ return "", err
+ }
+
+ return strings.TrimSpace(string(contents)), nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/rootless/rootless.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/rootless/rootless.go
new file mode 100644
index 000000000..b1efbfd99
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/rootless/rootless.go
@@ -0,0 +1,128 @@
+// +build linux
+
+package rootless
+
+import (
+ "fmt"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/cgroups/fs"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/configs/validate"
+)
+
+// TODO: This is copied from libcontainer/cgroups/fs, which duplicates this code
+// needlessly. We should probably export this list.
+
+var subsystems = []subsystem{
+ &fs.CpusetGroup{},
+ &fs.DevicesGroup{},
+ &fs.MemoryGroup{},
+ &fs.CpuGroup{},
+ &fs.CpuacctGroup{},
+ &fs.PidsGroup{},
+ &fs.BlkioGroup{},
+ &fs.HugetlbGroup{},
+ &fs.NetClsGroup{},
+ &fs.NetPrioGroup{},
+ &fs.PerfEventGroup{},
+ &fs.FreezerGroup{},
+ &fs.NameGroup{GroupName: "name=systemd"},
+}
+
+type subsystem interface {
+ // Name returns the name of the subsystem.
+ Name() string
+
+ // Returns the stats, as 'stats', corresponding to the cgroup under 'path'.
+ GetStats(path string, stats *cgroups.Stats) error
+}
+
+// The noop cgroup manager is used for rootless containers, because we currently
+// cannot manage cgroups if we are in a rootless setup. This manager is chosen
+// by factory if we are in rootless mode. We error out if any cgroup options are
+// set in the config -- this may change in the future with upcoming kernel features
+// like the cgroup namespace.
+
+type Manager struct {
+ Cgroups *configs.Cgroup
+ Paths map[string]string
+}
+
+func (m *Manager) Apply(pid int) error {
+ // If there are no cgroup settings, there's nothing to do.
+ if m.Cgroups == nil {
+ return nil
+ }
+
+ // We can't set paths.
+ // TODO(cyphar): Implement the case where the runner of a rootless container
+ // owns their own cgroup, which would allow us to set up a
+ // cgroup for each path.
+ if m.Cgroups.Paths != nil {
+ return fmt.Errorf("cannot change cgroup path in rootless container")
+ }
+
+ // We load the paths into the manager.
+ paths := make(map[string]string)
+ for _, sys := range subsystems {
+ name := sys.Name()
+
+ path, err := cgroups.GetOwnCgroupPath(name)
+ if err != nil {
+ // Ignore paths we couldn't resolve.
+ continue
+ }
+
+ paths[name] = path
+ }
+
+ m.Paths = paths
+ return nil
+}
+
+func (m *Manager) GetPaths() map[string]string {
+ return m.Paths
+}
+
+func (m *Manager) Set(container *configs.Config) error {
+ // We have to re-do the validation here, since someone might decide to
+ // update a rootless container.
+ return validate.New().Validate(container)
+}
+
+func (m *Manager) GetPids() ([]int, error) {
+ dir, err := cgroups.GetOwnCgroupPath("devices")
+ if err != nil {
+ return nil, err
+ }
+ return cgroups.GetPids(dir)
+}
+
+func (m *Manager) GetAllPids() ([]int, error) {
+ dir, err := cgroups.GetOwnCgroupPath("devices")
+ if err != nil {
+ return nil, err
+ }
+ return cgroups.GetAllPids(dir)
+}
+
+func (m *Manager) GetStats() (*cgroups.Stats, error) {
+ // TODO(cyphar): We can make this work if we figure out a way to allow usage
+ // of cgroups with a rootless container. While this doesn't
+ // actually require write access to a cgroup directory, the
+ // statistics are not useful if they can be affected by
+ // non-container processes.
+ return nil, fmt.Errorf("cannot get cgroup stats in rootless container")
+}
+
+func (m *Manager) Freeze(state configs.FreezerState) error {
+ // TODO(cyphar): We can make this work if we figure out a way to allow usage
+ // of cgroups with a rootless container.
+ return fmt.Errorf("cannot use freezer cgroup in rootless container")
+}
+
+func (m *Manager) Destroy() error {
+ // We don't have to do anything here because we didn't do any setup.
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/stats.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/stats.go
new file mode 100644
index 000000000..8eeedc55b
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/stats.go
@@ -0,0 +1,108 @@
+// +build linux
+
+package cgroups
+
+type ThrottlingData struct {
+ // Number of periods with throttling active
+ Periods uint64 `json:"periods,omitempty"`
+ // Number of periods when the container hit its throttling limit.
+ ThrottledPeriods uint64 `json:"throttled_periods,omitempty"`
+ // Aggregate time the container was throttled for in nanoseconds.
+ ThrottledTime uint64 `json:"throttled_time,omitempty"`
+}
+
+// CpuUsage denotes the usage of a CPU.
+// All CPU stats are aggregate since container inception.
+type CpuUsage struct {
+ // Total CPU time consumed.
+ // Units: nanoseconds.
+ TotalUsage uint64 `json:"total_usage,omitempty"`
+ // Total CPU time consumed per core.
+ // Units: nanoseconds.
+ PercpuUsage []uint64 `json:"percpu_usage,omitempty"`
+ // Time spent by tasks of the cgroup in kernel mode.
+ // Units: nanoseconds.
+ UsageInKernelmode uint64 `json:"usage_in_kernelmode"`
+ // Time spent by tasks of the cgroup in user mode.
+ // Units: nanoseconds.
+ UsageInUsermode uint64 `json:"usage_in_usermode"`
+}
+
+type CpuStats struct {
+ CpuUsage CpuUsage `json:"cpu_usage,omitempty"`
+ ThrottlingData ThrottlingData `json:"throttling_data,omitempty"`
+}
+
+type MemoryData struct {
+ Usage uint64 `json:"usage,omitempty"`
+ MaxUsage uint64 `json:"max_usage,omitempty"`
+ Failcnt uint64 `json:"failcnt"`
+ Limit uint64 `json:"limit"`
+}
+
+type MemoryStats struct {
+ // memory used for cache
+ Cache uint64 `json:"cache,omitempty"`
+ // usage of memory
+ Usage MemoryData `json:"usage,omitempty"`
+ // usage of memory + swap
+ SwapUsage MemoryData `json:"swap_usage,omitempty"`
+ // usage of kernel memory
+ KernelUsage MemoryData `json:"kernel_usage,omitempty"`
+ // usage of kernel TCP memory
+ KernelTCPUsage MemoryData `json:"kernel_tcp_usage,omitempty"`
+ // if true, memory usage is accounted for throughout a hierarchy of cgroups.
+ UseHierarchy bool `json:"use_hierarchy"`
+
+ Stats map[string]uint64 `json:"stats,omitempty"`
+}
+
+type PidsStats struct {
+ // number of pids in the cgroup
+ Current uint64 `json:"current,omitempty"`
+ // active pids hard limit
+ Limit uint64 `json:"limit,omitempty"`
+}
+
+type BlkioStatEntry struct {
+ Major uint64 `json:"major,omitempty"`
+ Minor uint64 `json:"minor,omitempty"`
+ Op string `json:"op,omitempty"`
+ Value uint64 `json:"value,omitempty"`
+}
+
+type BlkioStats struct {
+ // number of bytes tranferred to and from the block device
+ IoServiceBytesRecursive []BlkioStatEntry `json:"io_service_bytes_recursive,omitempty"`
+ IoServicedRecursive []BlkioStatEntry `json:"io_serviced_recursive,omitempty"`
+ IoQueuedRecursive []BlkioStatEntry `json:"io_queue_recursive,omitempty"`
+ IoServiceTimeRecursive []BlkioStatEntry `json:"io_service_time_recursive,omitempty"`
+ IoWaitTimeRecursive []BlkioStatEntry `json:"io_wait_time_recursive,omitempty"`
+ IoMergedRecursive []BlkioStatEntry `json:"io_merged_recursive,omitempty"`
+ IoTimeRecursive []BlkioStatEntry `json:"io_time_recursive,omitempty"`
+ SectorsRecursive []BlkioStatEntry `json:"sectors_recursive,omitempty"`
+}
+
+type HugetlbStats struct {
+ // current res_counter usage for hugetlb
+ Usage uint64 `json:"usage,omitempty"`
+ // maximum usage ever recorded.
+ MaxUsage uint64 `json:"max_usage,omitempty"`
+ // number of times hugetlb usage allocation failure.
+ Failcnt uint64 `json:"failcnt"`
+}
+
+type Stats struct {
+ CpuStats CpuStats `json:"cpu_stats,omitempty"`
+ MemoryStats MemoryStats `json:"memory_stats,omitempty"`
+ PidsStats PidsStats `json:"pids_stats,omitempty"`
+ BlkioStats BlkioStats `json:"blkio_stats,omitempty"`
+ // the map is in the format "size of hugepage: stats of the hugepage"
+ HugetlbStats map[string]HugetlbStats `json:"hugetlb_stats,omitempty"`
+}
+
+func NewStats() *Stats {
+ memoryStats := MemoryStats{Stats: make(map[string]uint64)}
+ hugetlbStats := make(map[string]HugetlbStats)
+ return &Stats{MemoryStats: memoryStats, HugetlbStats: hugetlbStats}
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_nosystemd.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_nosystemd.go
new file mode 100644
index 000000000..7de9ae605
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_nosystemd.go
@@ -0,0 +1,55 @@
+// +build !linux
+
+package systemd
+
+import (
+ "fmt"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type Manager struct {
+ Cgroups *configs.Cgroup
+ Paths map[string]string
+}
+
+func UseSystemd() bool {
+ return false
+}
+
+func (m *Manager) Apply(pid int) error {
+ return fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) GetPids() ([]int, error) {
+ return nil, fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) GetAllPids() ([]int, error) {
+ return nil, fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) Destroy() error {
+ return fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) GetPaths() map[string]string {
+ return nil
+}
+
+func (m *Manager) GetStats() (*cgroups.Stats, error) {
+ return nil, fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) Set(container *configs.Config) error {
+ return nil, fmt.Errorf("Systemd not supported")
+}
+
+func (m *Manager) Freeze(state configs.FreezerState) error {
+ return fmt.Errorf("Systemd not supported")
+}
+
+func Freeze(c *configs.Cgroup, state configs.FreezerState) error {
+ return fmt.Errorf("Systemd not supported")
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_systemd.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_systemd.go
new file mode 100644
index 000000000..b010b4b32
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/systemd/apply_systemd.go
@@ -0,0 +1,553 @@
+// +build linux
+
+package systemd
+
+import (
+ "errors"
+ "fmt"
+ "os"
+ "path/filepath"
+ "strings"
+ "sync"
+ "time"
+
+ systemdDbus "github.com/coreos/go-systemd/dbus"
+ systemdUtil "github.com/coreos/go-systemd/util"
+ "github.com/godbus/dbus"
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/cgroups/fs"
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type Manager struct {
+ mu sync.Mutex
+ Cgroups *configs.Cgroup
+ Paths map[string]string
+}
+
+type subsystem interface {
+ // Name returns the name of the subsystem.
+ Name() string
+ // Returns the stats, as 'stats', corresponding to the cgroup under 'path'.
+ GetStats(path string, stats *cgroups.Stats) error
+ // Set the cgroup represented by cgroup.
+ Set(path string, cgroup *configs.Cgroup) error
+}
+
+var errSubsystemDoesNotExist = errors.New("cgroup: subsystem does not exist")
+
+type subsystemSet []subsystem
+
+func (s subsystemSet) Get(name string) (subsystem, error) {
+ for _, ss := range s {
+ if ss.Name() == name {
+ return ss, nil
+ }
+ }
+ return nil, errSubsystemDoesNotExist
+}
+
+var subsystems = subsystemSet{
+ &fs.CpusetGroup{},
+ &fs.DevicesGroup{},
+ &fs.MemoryGroup{},
+ &fs.CpuGroup{},
+ &fs.CpuacctGroup{},
+ &fs.PidsGroup{},
+ &fs.BlkioGroup{},
+ &fs.HugetlbGroup{},
+ &fs.PerfEventGroup{},
+ &fs.FreezerGroup{},
+ &fs.NetPrioGroup{},
+ &fs.NetClsGroup{},
+ &fs.NameGroup{GroupName: "name=systemd"},
+}
+
+const (
+ testScopeWait = 4
+ testSliceWait = 4
+)
+
+var (
+ connLock sync.Mutex
+ theConn *systemdDbus.Conn
+ hasStartTransientUnit bool
+ hasStartTransientSliceUnit bool
+ hasTransientDefaultDependencies bool
+ hasDelegate bool
+)
+
+func newProp(name string, units interface{}) systemdDbus.Property {
+ return systemdDbus.Property{
+ Name: name,
+ Value: dbus.MakeVariant(units),
+ }
+}
+
+func UseSystemd() bool {
+ if !systemdUtil.IsRunningSystemd() {
+ return false
+ }
+
+ connLock.Lock()
+ defer connLock.Unlock()
+
+ if theConn == nil {
+ var err error
+ theConn, err = systemdDbus.New()
+ if err != nil {
+ return false
+ }
+
+ // Assume we have StartTransientUnit
+ hasStartTransientUnit = true
+
+ // But if we get UnknownMethod error we don't
+ if _, err := theConn.StartTransientUnit("test.scope", "invalid", nil, nil); err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ if dbusError.Name == "org.freedesktop.DBus.Error.UnknownMethod" {
+ hasStartTransientUnit = false
+ return hasStartTransientUnit
+ }
+ }
+ }
+
+ // Ensure the scope name we use doesn't exist. Use the Pid to
+ // avoid collisions between multiple libcontainer users on a
+ // single host.
+ scope := fmt.Sprintf("libcontainer-%d-systemd-test-default-dependencies.scope", os.Getpid())
+ testScopeExists := true
+ for i := 0; i <= testScopeWait; i++ {
+ if _, err := theConn.StopUnit(scope, "replace", nil); err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ if strings.Contains(dbusError.Name, "org.freedesktop.systemd1.NoSuchUnit") {
+ testScopeExists = false
+ break
+ }
+ }
+ }
+ time.Sleep(time.Millisecond)
+ }
+
+ // Bail out if we can't kill this scope without testing for DefaultDependencies
+ if testScopeExists {
+ return hasStartTransientUnit
+ }
+
+ // Assume StartTransientUnit on a scope allows DefaultDependencies
+ hasTransientDefaultDependencies = true
+ ddf := newProp("DefaultDependencies", false)
+ if _, err := theConn.StartTransientUnit(scope, "replace", []systemdDbus.Property{ddf}, nil); err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ if strings.Contains(dbusError.Name, "org.freedesktop.DBus.Error.PropertyReadOnly") {
+ hasTransientDefaultDependencies = false
+ }
+ }
+ }
+
+ // Not critical because of the stop unit logic above.
+ theConn.StopUnit(scope, "replace", nil)
+
+ // Assume StartTransientUnit on a scope allows Delegate
+ hasDelegate = true
+ dl := newProp("Delegate", true)
+ if _, err := theConn.StartTransientUnit(scope, "replace", []systemdDbus.Property{dl}, nil); err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ if strings.Contains(dbusError.Name, "org.freedesktop.DBus.Error.PropertyReadOnly") {
+ hasDelegate = false
+ }
+ }
+ }
+
+ // Assume we have the ability to start a transient unit as a slice
+ // This was broken until systemd v229, but has been back-ported on RHEL environments >= 219
+ // For details, see: https://bugzilla.redhat.com/show_bug.cgi?id=1370299
+ hasStartTransientSliceUnit = true
+
+ // To ensure simple clean-up, we create a slice off the root with no hierarchy
+ slice := fmt.Sprintf("libcontainer_%d_systemd_test_default.slice", os.Getpid())
+ if _, err := theConn.StartTransientUnit(slice, "replace", nil, nil); err != nil {
+ if _, ok := err.(dbus.Error); ok {
+ hasStartTransientSliceUnit = false
+ }
+ }
+
+ for i := 0; i <= testSliceWait; i++ {
+ if _, err := theConn.StopUnit(slice, "replace", nil); err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ if strings.Contains(dbusError.Name, "org.freedesktop.systemd1.NoSuchUnit") {
+ hasStartTransientSliceUnit = false
+ break
+ }
+ }
+ } else {
+ break
+ }
+ time.Sleep(time.Millisecond)
+ }
+
+ // Not critical because of the stop unit logic above.
+ theConn.StopUnit(scope, "replace", nil)
+ theConn.StopUnit(slice, "replace", nil)
+ }
+ return hasStartTransientUnit
+}
+
+func (m *Manager) Apply(pid int) error {
+ var (
+ c = m.Cgroups
+ unitName = getUnitName(c)
+ slice = "system.slice"
+ properties []systemdDbus.Property
+ )
+
+ if c.Paths != nil {
+ paths := make(map[string]string)
+ for name, path := range c.Paths {
+ _, err := getSubsystemPath(m.Cgroups, name)
+ if err != nil {
+ // Don't fail if a cgroup hierarchy was not found, just skip this subsystem
+ if cgroups.IsNotFound(err) {
+ continue
+ }
+ return err
+ }
+ paths[name] = path
+ }
+ m.Paths = paths
+ return cgroups.EnterPid(m.Paths, pid)
+ }
+
+ if c.Parent != "" {
+ slice = c.Parent
+ }
+
+ properties = append(properties, systemdDbus.PropDescription("libcontainer container "+c.Name))
+
+ // if we create a slice, the parent is defined via a Wants=
+ if strings.HasSuffix(unitName, ".slice") {
+ // This was broken until systemd v229, but has been back-ported on RHEL environments >= 219
+ if !hasStartTransientSliceUnit {
+ return fmt.Errorf("systemd version does not support ability to start a slice as transient unit")
+ }
+ properties = append(properties, systemdDbus.PropWants(slice))
+ } else {
+ // otherwise, we use Slice=
+ properties = append(properties, systemdDbus.PropSlice(slice))
+ }
+
+ // only add pid if its valid, -1 is used w/ general slice creation.
+ if pid != -1 {
+ properties = append(properties, newProp("PIDs", []uint32{uint32(pid)}))
+ }
+
+ if hasDelegate {
+ // This is only supported on systemd versions 218 and above.
+ properties = append(properties, newProp("Delegate", true))
+ }
+
+ // Always enable accounting, this gets us the same behaviour as the fs implementation,
+ // plus the kernel has some problems with joining the memory cgroup at a later time.
+ properties = append(properties,
+ newProp("MemoryAccounting", true),
+ newProp("CPUAccounting", true),
+ newProp("BlockIOAccounting", true))
+
+ if hasTransientDefaultDependencies {
+ properties = append(properties,
+ newProp("DefaultDependencies", false))
+ }
+
+ if c.Resources.Memory != 0 {
+ properties = append(properties,
+ newProp("MemoryLimit", c.Resources.Memory))
+ }
+
+ if c.Resources.CpuShares != 0 {
+ properties = append(properties,
+ newProp("CPUShares", c.Resources.CpuShares))
+ }
+
+ // cpu.cfs_quota_us and cpu.cfs_period_us are controlled by systemd.
+ if c.Resources.CpuQuota != 0 && c.Resources.CpuPeriod != 0 {
+ cpuQuotaPerSecUSec := uint64(c.Resources.CpuQuota*1000000) / c.Resources.CpuPeriod
+ properties = append(properties,
+ newProp("CPUQuotaPerSecUSec", cpuQuotaPerSecUSec))
+ }
+
+ if c.Resources.BlkioWeight != 0 {
+ properties = append(properties,
+ newProp("BlockIOWeight", uint64(c.Resources.BlkioWeight)))
+ }
+
+ // We have to set kernel memory here, as we can't change it once
+ // processes have been attached to the cgroup.
+ if c.Resources.KernelMemory != 0 {
+ if err := setKernelMemory(c); err != nil {
+ return err
+ }
+ }
+
+ if _, err := theConn.StartTransientUnit(unitName, "replace", properties, nil); err != nil && !isUnitExists(err) {
+ return err
+ }
+
+ if err := joinCgroups(c, pid); err != nil {
+ return err
+ }
+
+ paths := make(map[string]string)
+ for _, s := range subsystems {
+ subsystemPath, err := getSubsystemPath(m.Cgroups, s.Name())
+ if err != nil {
+ // Don't fail if a cgroup hierarchy was not found, just skip this subsystem
+ if cgroups.IsNotFound(err) {
+ continue
+ }
+ return err
+ }
+ paths[s.Name()] = subsystemPath
+ }
+ m.Paths = paths
+ return nil
+}
+
+func (m *Manager) Destroy() error {
+ if m.Cgroups.Paths != nil {
+ return nil
+ }
+ m.mu.Lock()
+ defer m.mu.Unlock()
+ theConn.StopUnit(getUnitName(m.Cgroups), "replace", nil)
+ if err := cgroups.RemovePaths(m.Paths); err != nil {
+ return err
+ }
+ m.Paths = make(map[string]string)
+ return nil
+}
+
+func (m *Manager) GetPaths() map[string]string {
+ m.mu.Lock()
+ paths := m.Paths
+ m.mu.Unlock()
+ return paths
+}
+
+func join(c *configs.Cgroup, subsystem string, pid int) (string, error) {
+ path, err := getSubsystemPath(c, subsystem)
+ if err != nil {
+ return "", err
+ }
+ if err := os.MkdirAll(path, 0755); err != nil {
+ return "", err
+ }
+ if err := cgroups.WriteCgroupProc(path, pid); err != nil {
+ return "", err
+ }
+ return path, nil
+}
+
+func joinCgroups(c *configs.Cgroup, pid int) error {
+ for _, sys := range subsystems {
+ name := sys.Name()
+ switch name {
+ case "name=systemd":
+ // let systemd handle this
+ case "cpuset":
+ path, err := getSubsystemPath(c, name)
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+ s := &fs.CpusetGroup{}
+ if err := s.ApplyDir(path, c, pid); err != nil {
+ return err
+ }
+ default:
+ _, err := join(c, name, pid)
+ if err != nil {
+ // Even if it's `not found` error, we'll return err
+ // because devices cgroup is hard requirement for
+ // container security.
+ if name == "devices" {
+ return err
+ }
+ // For other subsystems, omit the `not found` error
+ // because they are optional.
+ if !cgroups.IsNotFound(err) {
+ return err
+ }
+ }
+ }
+ }
+
+ return nil
+}
+
+// systemd represents slice hierarchy using `-`, so we need to follow suit when
+// generating the path of slice. Essentially, test-a-b.slice becomes
+// test.slice/test-a.slice/test-a-b.slice.
+func ExpandSlice(slice string) (string, error) {
+ suffix := ".slice"
+ // Name has to end with ".slice", but can't be just ".slice".
+ if len(slice) < len(suffix) || !strings.HasSuffix(slice, suffix) {
+ return "", fmt.Errorf("invalid slice name: %s", slice)
+ }
+
+ // Path-separators are not allowed.
+ if strings.Contains(slice, "/") {
+ return "", fmt.Errorf("invalid slice name: %s", slice)
+ }
+
+ var path, prefix string
+ sliceName := strings.TrimSuffix(slice, suffix)
+ // if input was -.slice, we should just return root now
+ if sliceName == "-" {
+ return "/", nil
+ }
+ for _, component := range strings.Split(sliceName, "-") {
+ // test--a.slice isn't permitted, nor is -test.slice.
+ if component == "" {
+ return "", fmt.Errorf("invalid slice name: %s", slice)
+ }
+
+ // Append the component to the path and to the prefix.
+ path += prefix + component + suffix + "/"
+ prefix += component + "-"
+ }
+
+ return path, nil
+}
+
+func getSubsystemPath(c *configs.Cgroup, subsystem string) (string, error) {
+ mountpoint, err := cgroups.FindCgroupMountpoint(subsystem)
+ if err != nil {
+ return "", err
+ }
+
+ initPath, err := cgroups.GetInitCgroup(subsystem)
+ if err != nil {
+ return "", err
+ }
+ // if pid 1 is systemd 226 or later, it will be in init.scope, not the root
+ initPath = strings.TrimSuffix(filepath.Clean(initPath), "init.scope")
+
+ slice := "system.slice"
+ if c.Parent != "" {
+ slice = c.Parent
+ }
+
+ slice, err = ExpandSlice(slice)
+ if err != nil {
+ return "", err
+ }
+
+ return filepath.Join(mountpoint, initPath, slice, getUnitName(c)), nil
+}
+
+func (m *Manager) Freeze(state configs.FreezerState) error {
+ path, err := getSubsystemPath(m.Cgroups, "freezer")
+ if err != nil {
+ return err
+ }
+ prevState := m.Cgroups.Resources.Freezer
+ m.Cgroups.Resources.Freezer = state
+ freezer, err := subsystems.Get("freezer")
+ if err != nil {
+ return err
+ }
+ err = freezer.Set(path, m.Cgroups)
+ if err != nil {
+ m.Cgroups.Resources.Freezer = prevState
+ return err
+ }
+ return nil
+}
+
+func (m *Manager) GetPids() ([]int, error) {
+ path, err := getSubsystemPath(m.Cgroups, "devices")
+ if err != nil {
+ return nil, err
+ }
+ return cgroups.GetPids(path)
+}
+
+func (m *Manager) GetAllPids() ([]int, error) {
+ path, err := getSubsystemPath(m.Cgroups, "devices")
+ if err != nil {
+ return nil, err
+ }
+ return cgroups.GetAllPids(path)
+}
+
+func (m *Manager) GetStats() (*cgroups.Stats, error) {
+ m.mu.Lock()
+ defer m.mu.Unlock()
+ stats := cgroups.NewStats()
+ for name, path := range m.Paths {
+ sys, err := subsystems.Get(name)
+ if err == errSubsystemDoesNotExist || !cgroups.PathExists(path) {
+ continue
+ }
+ if err := sys.GetStats(path, stats); err != nil {
+ return nil, err
+ }
+ }
+
+ return stats, nil
+}
+
+func (m *Manager) Set(container *configs.Config) error {
+ // If Paths are set, then we are just joining cgroups paths
+ // and there is no need to set any values.
+ if m.Cgroups.Paths != nil {
+ return nil
+ }
+ for _, sys := range subsystems {
+ // Get the subsystem path, but don't error out for not found cgroups.
+ path, err := getSubsystemPath(container.Cgroups, sys.Name())
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+
+ if err := sys.Set(path, container.Cgroups); err != nil {
+ return err
+ }
+ }
+
+ if m.Paths["cpu"] != "" {
+ if err := fs.CheckCpushares(m.Paths["cpu"], container.Cgroups.Resources.CpuShares); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func getUnitName(c *configs.Cgroup) string {
+ // by default, we create a scope unless the user explicitly asks for a slice.
+ if !strings.HasSuffix(c.Name, ".slice") {
+ return fmt.Sprintf("%s-%s.scope", c.ScopePrefix, c.Name)
+ }
+ return c.Name
+}
+
+func setKernelMemory(c *configs.Cgroup) error {
+ path, err := getSubsystemPath(c, "memory")
+ if err != nil && !cgroups.IsNotFound(err) {
+ return err
+ }
+
+ if err := os.MkdirAll(path, 0755); err != nil {
+ return err
+ }
+ return fs.EnableKernelMemoryAccounting(path)
+}
+
+// isUnitExists returns true if the error is that a systemd unit already exists.
+func isUnitExists(err error) bool {
+ if err != nil {
+ if dbusError, ok := err.(dbus.Error); ok {
+ return strings.Contains(dbusError.Name, "org.freedesktop.systemd1.UnitExists")
+ }
+ }
+ return false
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/cgroups/utils.go b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/utils.go
new file mode 100644
index 000000000..7c995efee
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/cgroups/utils.go
@@ -0,0 +1,462 @@
+// +build linux
+
+package cgroups
+
+import (
+ "bufio"
+ "fmt"
+ "io"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+ "strconv"
+ "strings"
+ "time"
+
+ "github.com/docker/go-units"
+)
+
+const (
+ cgroupNamePrefix = "name="
+ CgroupProcesses = "cgroup.procs"
+)
+
+// https://www.kernel.org/doc/Documentation/cgroup-v1/cgroups.txt
+func FindCgroupMountpoint(subsystem string) (string, error) {
+ mnt, _, err := FindCgroupMountpointAndRoot(subsystem)
+ return mnt, err
+}
+
+func FindCgroupMountpointAndRoot(subsystem string) (string, string, error) {
+ // We are not using mount.GetMounts() because it's super-inefficient,
+ // parsing it directly sped up x10 times because of not using Sscanf.
+ // It was one of two major performance drawbacks in container start.
+ if !isSubsystemAvailable(subsystem) {
+ return "", "", NewNotFoundError(subsystem)
+ }
+ f, err := os.Open("/proc/self/mountinfo")
+ if err != nil {
+ return "", "", err
+ }
+ defer f.Close()
+
+ scanner := bufio.NewScanner(f)
+ for scanner.Scan() {
+ txt := scanner.Text()
+ fields := strings.Split(txt, " ")
+ for _, opt := range strings.Split(fields[len(fields)-1], ",") {
+ if opt == subsystem {
+ return fields[4], fields[3], nil
+ }
+ }
+ }
+ if err := scanner.Err(); err != nil {
+ return "", "", err
+ }
+
+ return "", "", NewNotFoundError(subsystem)
+}
+
+func isSubsystemAvailable(subsystem string) bool {
+ cgroups, err := ParseCgroupFile("/proc/self/cgroup")
+ if err != nil {
+ return false
+ }
+ _, avail := cgroups[subsystem]
+ return avail
+}
+
+func GetClosestMountpointAncestor(dir, mountinfo string) string {
+ deepestMountPoint := ""
+ for _, mountInfoEntry := range strings.Split(mountinfo, "\n") {
+ mountInfoParts := strings.Fields(mountInfoEntry)
+ if len(mountInfoParts) < 5 {
+ continue
+ }
+ mountPoint := mountInfoParts[4]
+ if strings.HasPrefix(mountPoint, deepestMountPoint) && strings.HasPrefix(dir, mountPoint) {
+ deepestMountPoint = mountPoint
+ }
+ }
+ return deepestMountPoint
+}
+
+func FindCgroupMountpointDir() (string, error) {
+ f, err := os.Open("/proc/self/mountinfo")
+ if err != nil {
+ return "", err
+ }
+ defer f.Close()
+
+ scanner := bufio.NewScanner(f)
+ for scanner.Scan() {
+ text := scanner.Text()
+ fields := strings.Split(text, " ")
+ // Safe as mountinfo encodes mountpoints with spaces as \040.
+ index := strings.Index(text, " - ")
+ postSeparatorFields := strings.Fields(text[index+3:])
+ numPostFields := len(postSeparatorFields)
+
+ // This is an error as we can't detect if the mount is for "cgroup"
+ if numPostFields == 0 {
+ return "", fmt.Errorf("Found no fields post '-' in %q", text)
+ }
+
+ if postSeparatorFields[0] == "cgroup" {
+ // Check that the mount is properly formated.
+ if numPostFields < 3 {
+ return "", fmt.Errorf("Error found less than 3 fields post '-' in %q", text)
+ }
+
+ return filepath.Dir(fields[4]), nil
+ }
+ }
+ if err := scanner.Err(); err != nil {
+ return "", err
+ }
+
+ return "", NewNotFoundError("cgroup")
+}
+
+type Mount struct {
+ Mountpoint string
+ Root string
+ Subsystems []string
+}
+
+func (m Mount) GetOwnCgroup(cgroups map[string]string) (string, error) {
+ if len(m.Subsystems) == 0 {
+ return "", fmt.Errorf("no subsystem for mount")
+ }
+
+ return getControllerPath(m.Subsystems[0], cgroups)
+}
+
+func getCgroupMountsHelper(ss map[string]bool, mi io.Reader, all bool) ([]Mount, error) {
+ res := make([]Mount, 0, len(ss))
+ scanner := bufio.NewScanner(mi)
+ numFound := 0
+ for scanner.Scan() && numFound < len(ss) {
+ txt := scanner.Text()
+ sepIdx := strings.Index(txt, " - ")
+ if sepIdx == -1 {
+ return nil, fmt.Errorf("invalid mountinfo format")
+ }
+ if txt[sepIdx+3:sepIdx+10] == "cgroup2" || txt[sepIdx+3:sepIdx+9] != "cgroup" {
+ continue
+ }
+ fields := strings.Split(txt, " ")
+ m := Mount{
+ Mountpoint: fields[4],
+ Root: fields[3],
+ }
+ for _, opt := range strings.Split(fields[len(fields)-1], ",") {
+ if !ss[opt] {
+ continue
+ }
+ if strings.HasPrefix(opt, cgroupNamePrefix) {
+ m.Subsystems = append(m.Subsystems, opt[len(cgroupNamePrefix):])
+ } else {
+ m.Subsystems = append(m.Subsystems, opt)
+ }
+ if !all {
+ numFound++
+ }
+ }
+ res = append(res, m)
+ }
+ if err := scanner.Err(); err != nil {
+ return nil, err
+ }
+ return res, nil
+}
+
+// GetCgroupMounts returns the mounts for the cgroup subsystems.
+// all indicates whether to return just the first instance or all the mounts.
+func GetCgroupMounts(all bool) ([]Mount, error) {
+ f, err := os.Open("/proc/self/mountinfo")
+ if err != nil {
+ return nil, err
+ }
+ defer f.Close()
+
+ allSubsystems, err := ParseCgroupFile("/proc/self/cgroup")
+ if err != nil {
+ return nil, err
+ }
+
+ allMap := make(map[string]bool)
+ for s := range allSubsystems {
+ allMap[s] = true
+ }
+ return getCgroupMountsHelper(allMap, f, all)
+}
+
+// GetAllSubsystems returns all the cgroup subsystems supported by the kernel
+func GetAllSubsystems() ([]string, error) {
+ f, err := os.Open("/proc/cgroups")
+ if err != nil {
+ return nil, err
+ }
+ defer f.Close()
+
+ subsystems := []string{}
+
+ s := bufio.NewScanner(f)
+ for s.Scan() {
+ text := s.Text()
+ if text[0] != '#' {
+ parts := strings.Fields(text)
+ if len(parts) >= 4 && parts[3] != "0" {
+ subsystems = append(subsystems, parts[0])
+ }
+ }
+ }
+ if err := s.Err(); err != nil {
+ return nil, err
+ }
+ return subsystems, nil
+}
+
+// GetOwnCgroup returns the relative path to the cgroup docker is running in.
+func GetOwnCgroup(subsystem string) (string, error) {
+ cgroups, err := ParseCgroupFile("/proc/self/cgroup")
+ if err != nil {
+ return "", err
+ }
+
+ return getControllerPath(subsystem, cgroups)
+}
+
+func GetOwnCgroupPath(subsystem string) (string, error) {
+ cgroup, err := GetOwnCgroup(subsystem)
+ if err != nil {
+ return "", err
+ }
+
+ return getCgroupPathHelper(subsystem, cgroup)
+}
+
+func GetInitCgroup(subsystem string) (string, error) {
+ cgroups, err := ParseCgroupFile("/proc/1/cgroup")
+ if err != nil {
+ return "", err
+ }
+
+ return getControllerPath(subsystem, cgroups)
+}
+
+func GetInitCgroupPath(subsystem string) (string, error) {
+ cgroup, err := GetInitCgroup(subsystem)
+ if err != nil {
+ return "", err
+ }
+
+ return getCgroupPathHelper(subsystem, cgroup)
+}
+
+func getCgroupPathHelper(subsystem, cgroup string) (string, error) {
+ mnt, root, err := FindCgroupMountpointAndRoot(subsystem)
+ if err != nil {
+ return "", err
+ }
+
+ // This is needed for nested containers, because in /proc/self/cgroup we
+ // see pathes from host, which don't exist in container.
+ relCgroup, err := filepath.Rel(root, cgroup)
+ if err != nil {
+ return "", err
+ }
+
+ return filepath.Join(mnt, relCgroup), nil
+}
+
+func readProcsFile(dir string) ([]int, error) {
+ f, err := os.Open(filepath.Join(dir, CgroupProcesses))
+ if err != nil {
+ return nil, err
+ }
+ defer f.Close()
+
+ var (
+ s = bufio.NewScanner(f)
+ out = []int{}
+ )
+
+ for s.Scan() {
+ if t := s.Text(); t != "" {
+ pid, err := strconv.Atoi(t)
+ if err != nil {
+ return nil, err
+ }
+ out = append(out, pid)
+ }
+ }
+ return out, nil
+}
+
+// ParseCgroupFile parses the given cgroup file, typically from
+// /proc/<pid>/cgroup, into a map of subgroups to cgroup names.
+func ParseCgroupFile(path string) (map[string]string, error) {
+ f, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+ defer f.Close()
+
+ return parseCgroupFromReader(f)
+}
+
+// helper function for ParseCgroupFile to make testing easier
+func parseCgroupFromReader(r io.Reader) (map[string]string, error) {
+ s := bufio.NewScanner(r)
+ cgroups := make(map[string]string)
+
+ for s.Scan() {
+ text := s.Text()
+ // from cgroups(7):
+ // /proc/[pid]/cgroup
+ // ...
+ // For each cgroup hierarchy ... there is one entry
+ // containing three colon-separated fields of the form:
+ // hierarchy-ID:subsystem-list:cgroup-path
+ parts := strings.SplitN(text, ":", 3)
+ if len(parts) < 3 {
+ return nil, fmt.Errorf("invalid cgroup entry: must contain at least two colons: %v", text)
+ }
+
+ for _, subs := range strings.Split(parts[1], ",") {
+ cgroups[subs] = parts[2]
+ }
+ }
+ if err := s.Err(); err != nil {
+ return nil, err
+ }
+
+ return cgroups, nil
+}
+
+func getControllerPath(subsystem string, cgroups map[string]string) (string, error) {
+
+ if p, ok := cgroups[subsystem]; ok {
+ return p, nil
+ }
+
+ if p, ok := cgroups[cgroupNamePrefix+subsystem]; ok {
+ return p, nil
+ }
+
+ return "", NewNotFoundError(subsystem)
+}
+
+func PathExists(path string) bool {
+ if _, err := os.Stat(path); err != nil {
+ return false
+ }
+ return true
+}
+
+func EnterPid(cgroupPaths map[string]string, pid int) error {
+ for _, path := range cgroupPaths {
+ if PathExists(path) {
+ if err := WriteCgroupProc(path, pid); err != nil {
+ return err
+ }
+ }
+ }
+ return nil
+}
+
+// RemovePaths iterates over the provided paths removing them.
+// We trying to remove all paths five times with increasing delay between tries.
+// If after all there are not removed cgroups - appropriate error will be
+// returned.
+func RemovePaths(paths map[string]string) (err error) {
+ delay := 10 * time.Millisecond
+ for i := 0; i < 5; i++ {
+ if i != 0 {
+ time.Sleep(delay)
+ delay *= 2
+ }
+ for s, p := range paths {
+ os.RemoveAll(p)
+ // TODO: here probably should be logging
+ _, err := os.Stat(p)
+ // We need this strange way of checking cgroups existence because
+ // RemoveAll almost always returns error, even on already removed
+ // cgroups
+ if os.IsNotExist(err) {
+ delete(paths, s)
+ }
+ }
+ if len(paths) == 0 {
+ return nil
+ }
+ }
+ return fmt.Errorf("Failed to remove paths: %v", paths)
+}
+
+func GetHugePageSize() ([]string, error) {
+ var pageSizes []string
+ sizeList := []string{"B", "kB", "MB", "GB", "TB", "PB"}
+ files, err := ioutil.ReadDir("/sys/kernel/mm/hugepages")
+ if err != nil {
+ return pageSizes, err
+ }
+ for _, st := range files {
+ nameArray := strings.Split(st.Name(), "-")
+ pageSize, err := units.RAMInBytes(nameArray[1])
+ if err != nil {
+ return []string{}, err
+ }
+ sizeString := units.CustomSize("%g%s", float64(pageSize), 1024.0, sizeList)
+ pageSizes = append(pageSizes, sizeString)
+ }
+
+ return pageSizes, nil
+}
+
+// GetPids returns all pids, that were added to cgroup at path.
+func GetPids(path string) ([]int, error) {
+ return readProcsFile(path)
+}
+
+// GetAllPids returns all pids, that were added to cgroup at path and to all its
+// subcgroups.
+func GetAllPids(path string) ([]int, error) {
+ var pids []int
+ // collect pids from all sub-cgroups
+ err := filepath.Walk(path, func(p string, info os.FileInfo, iErr error) error {
+ dir, file := filepath.Split(p)
+ if file != CgroupProcesses {
+ return nil
+ }
+ if iErr != nil {
+ return iErr
+ }
+ cPids, err := readProcsFile(dir)
+ if err != nil {
+ return err
+ }
+ pids = append(pids, cPids...)
+ return nil
+ })
+ return pids, err
+}
+
+// WriteCgroupProc writes the specified pid into the cgroup's cgroup.procs file
+func WriteCgroupProc(dir string, pid int) error {
+ // Normally dir should not be empty, one case is that cgroup subsystem
+ // is not mounted, we will get empty dir, and we want it fail here.
+ if dir == "" {
+ return fmt.Errorf("no such directory for %s", CgroupProcesses)
+ }
+
+ // Dont attach any pid to the cgroup if -1 is specified as a pid
+ if pid != -1 {
+ if err := ioutil.WriteFile(filepath.Join(dir, CgroupProcesses), []byte(strconv.Itoa(pid)), 0700); err != nil {
+ return fmt.Errorf("failed to write %v to %v: %v", pid, CgroupProcesses, err)
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/compat_1.5_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/compat_1.5_linux.go
new file mode 100644
index 000000000..c7bdf1f60
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/compat_1.5_linux.go
@@ -0,0 +1,10 @@
+// +build linux,!go1.5
+
+package libcontainer
+
+import "syscall"
+
+// GidMappingsEnableSetgroups was added in Go 1.5, so do nothing when building
+// with earlier versions
+func enableSetgroups(sys *syscall.SysProcAttr) {
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/blkio_device.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/blkio_device.go
new file mode 100644
index 000000000..e0f3ca165
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/blkio_device.go
@@ -0,0 +1,61 @@
+package configs
+
+import "fmt"
+
+// blockIODevice holds major:minor format supported in blkio cgroup
+type blockIODevice struct {
+ // Major is the device's major number
+ Major int64 `json:"major"`
+ // Minor is the device's minor number
+ Minor int64 `json:"minor"`
+}
+
+// WeightDevice struct holds a `major:minor weight`|`major:minor leaf_weight` pair
+type WeightDevice struct {
+ blockIODevice
+ // Weight is the bandwidth rate for the device, range is from 10 to 1000
+ Weight uint16 `json:"weight"`
+ // LeafWeight is the bandwidth rate for the device while competing with the cgroup's child cgroups, range is from 10 to 1000, cfq scheduler only
+ LeafWeight uint16 `json:"leafWeight"`
+}
+
+// NewWeightDevice returns a configured WeightDevice pointer
+func NewWeightDevice(major, minor int64, weight, leafWeight uint16) *WeightDevice {
+ wd := &WeightDevice{}
+ wd.Major = major
+ wd.Minor = minor
+ wd.Weight = weight
+ wd.LeafWeight = leafWeight
+ return wd
+}
+
+// WeightString formats the struct to be writable to the cgroup specific file
+func (wd *WeightDevice) WeightString() string {
+ return fmt.Sprintf("%d:%d %d", wd.Major, wd.Minor, wd.Weight)
+}
+
+// LeafWeightString formats the struct to be writable to the cgroup specific file
+func (wd *WeightDevice) LeafWeightString() string {
+ return fmt.Sprintf("%d:%d %d", wd.Major, wd.Minor, wd.LeafWeight)
+}
+
+// ThrottleDevice struct holds a `major:minor rate_per_second` pair
+type ThrottleDevice struct {
+ blockIODevice
+ // Rate is the IO rate limit per cgroup per device
+ Rate uint64 `json:"rate"`
+}
+
+// NewThrottleDevice returns a configured ThrottleDevice pointer
+func NewThrottleDevice(major, minor int64, rate uint64) *ThrottleDevice {
+ td := &ThrottleDevice{}
+ td.Major = major
+ td.Minor = minor
+ td.Rate = rate
+ return td
+}
+
+// String formats the struct to be writable to the cgroup specific file
+func (td *ThrottleDevice) String() string {
+ return fmt.Sprintf("%d:%d %d", td.Major, td.Minor, td.Rate)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_linux.go
new file mode 100644
index 000000000..e15a662f5
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_linux.go
@@ -0,0 +1,122 @@
+package configs
+
+type FreezerState string
+
+const (
+ Undefined FreezerState = ""
+ Frozen FreezerState = "FROZEN"
+ Thawed FreezerState = "THAWED"
+)
+
+type Cgroup struct {
+ // Deprecated, use Path instead
+ Name string `json:"name,omitempty"`
+
+ // name of parent of cgroup or slice
+ // Deprecated, use Path instead
+ Parent string `json:"parent,omitempty"`
+
+ // Path specifies the path to cgroups that are created and/or joined by the container.
+ // The path is assumed to be relative to the host system cgroup mountpoint.
+ Path string `json:"path"`
+
+ // ScopePrefix describes prefix for the scope name
+ ScopePrefix string `json:"scope_prefix"`
+
+ // Paths represent the absolute cgroups paths to join.
+ // This takes precedence over Path.
+ Paths map[string]string
+
+ // Resources contains various cgroups settings to apply
+ *Resources
+}
+
+type Resources struct {
+ // If this is true allow access to any kind of device within the container. If false, allow access only to devices explicitly listed in the allowed_devices list.
+ // Deprecated
+ AllowAllDevices *bool `json:"allow_all_devices,omitempty"`
+ // Deprecated
+ AllowedDevices []*Device `json:"allowed_devices,omitempty"`
+ // Deprecated
+ DeniedDevices []*Device `json:"denied_devices,omitempty"`
+
+ Devices []*Device `json:"devices"`
+
+ // Memory limit (in bytes)
+ Memory int64 `json:"memory"`
+
+ // Memory reservation or soft_limit (in bytes)
+ MemoryReservation int64 `json:"memory_reservation"`
+
+ // Total memory usage (memory + swap); set `-1` to enable unlimited swap
+ MemorySwap int64 `json:"memory_swap"`
+
+ // Kernel memory limit (in bytes)
+ KernelMemory int64 `json:"kernel_memory"`
+
+ // Kernel memory limit for TCP use (in bytes)
+ KernelMemoryTCP int64 `json:"kernel_memory_tcp"`
+
+ // CPU shares (relative weight vs. other containers)
+ CpuShares uint64 `json:"cpu_shares"`
+
+ // CPU hardcap limit (in usecs). Allowed cpu time in a given period.
+ CpuQuota int64 `json:"cpu_quota"`
+
+ // CPU period to be used for hardcapping (in usecs). 0 to use system default.
+ CpuPeriod uint64 `json:"cpu_period"`
+
+ // How many time CPU will use in realtime scheduling (in usecs).
+ CpuRtRuntime int64 `json:"cpu_rt_quota"`
+
+ // CPU period to be used for realtime scheduling (in usecs).
+ CpuRtPeriod uint64 `json:"cpu_rt_period"`
+
+ // CPU to use
+ CpusetCpus string `json:"cpuset_cpus"`
+
+ // MEM to use
+ CpusetMems string `json:"cpuset_mems"`
+
+ // Process limit; set <= `0' to disable limit.
+ PidsLimit int64 `json:"pids_limit"`
+
+ // Specifies per cgroup weight, range is from 10 to 1000.
+ BlkioWeight uint16 `json:"blkio_weight"`
+
+ // Specifies tasks' weight in the given cgroup while competing with the cgroup's child cgroups, range is from 10 to 1000, cfq scheduler only
+ BlkioLeafWeight uint16 `json:"blkio_leaf_weight"`
+
+ // Weight per cgroup per device, can override BlkioWeight.
+ BlkioWeightDevice []*WeightDevice `json:"blkio_weight_device"`
+
+ // IO read rate limit per cgroup per device, bytes per second.
+ BlkioThrottleReadBpsDevice []*ThrottleDevice `json:"blkio_throttle_read_bps_device"`
+
+ // IO write rate limit per cgroup per device, bytes per second.
+ BlkioThrottleWriteBpsDevice []*ThrottleDevice `json:"blkio_throttle_write_bps_device"`
+
+ // IO read rate limit per cgroup per device, IO per second.
+ BlkioThrottleReadIOPSDevice []*ThrottleDevice `json:"blkio_throttle_read_iops_device"`
+
+ // IO write rate limit per cgroup per device, IO per second.
+ BlkioThrottleWriteIOPSDevice []*ThrottleDevice `json:"blkio_throttle_write_iops_device"`
+
+ // set the freeze value for the process
+ Freezer FreezerState `json:"freezer"`
+
+ // Hugetlb limit (in bytes)
+ HugetlbLimit []*HugepageLimit `json:"hugetlb_limit"`
+
+ // Whether to disable OOM Killer
+ OomKillDisable bool `json:"oom_kill_disable"`
+
+ // Tuning swappiness behaviour per cgroup
+ MemorySwappiness *uint64 `json:"memory_swappiness"`
+
+ // Set priority of network traffic for container
+ NetPrioIfpriomap []*IfPrioMap `json:"net_prio_ifpriomap"`
+
+ // Set class identifier for container's network packets
+ NetClsClassid uint32 `json:"net_cls_classid_u"`
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_unsupported.go
new file mode 100644
index 000000000..95e2830a4
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_unsupported.go
@@ -0,0 +1,6 @@
+// +build !windows,!linux,!freebsd
+
+package configs
+
+type Cgroup struct {
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_windows.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_windows.go
new file mode 100644
index 000000000..d74847b0d
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/cgroup_windows.go
@@ -0,0 +1,6 @@
+package configs
+
+// TODO Windows: This can ultimately be entirely factored out on Windows as
+// cgroups are a Unix-specific construct.
+type Cgroup struct {
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/config.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/config.go
new file mode 100644
index 000000000..269fffff3
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/config.go
@@ -0,0 +1,344 @@
+package configs
+
+import (
+ "bytes"
+ "encoding/json"
+ "fmt"
+ "os/exec"
+ "time"
+
+ "github.com/opencontainers/runtime-spec/specs-go"
+
+ "github.com/sirupsen/logrus"
+)
+
+type Rlimit struct {
+ Type int `json:"type"`
+ Hard uint64 `json:"hard"`
+ Soft uint64 `json:"soft"`
+}
+
+// IDMap represents UID/GID Mappings for User Namespaces.
+type IDMap struct {
+ ContainerID int `json:"container_id"`
+ HostID int `json:"host_id"`
+ Size int `json:"size"`
+}
+
+// Seccomp represents syscall restrictions
+// By default, only the native architecture of the kernel is allowed to be used
+// for syscalls. Additional architectures can be added by specifying them in
+// Architectures.
+type Seccomp struct {
+ DefaultAction Action `json:"default_action"`
+ Architectures []string `json:"architectures"`
+ Syscalls []*Syscall `json:"syscalls"`
+}
+
+// Action is taken upon rule match in Seccomp
+type Action int
+
+const (
+ Kill Action = iota + 1
+ Errno
+ Trap
+ Allow
+ Trace
+)
+
+// Operator is a comparison operator to be used when matching syscall arguments in Seccomp
+type Operator int
+
+const (
+ EqualTo Operator = iota + 1
+ NotEqualTo
+ GreaterThan
+ GreaterThanOrEqualTo
+ LessThan
+ LessThanOrEqualTo
+ MaskEqualTo
+)
+
+// Arg is a rule to match a specific syscall argument in Seccomp
+type Arg struct {
+ Index uint `json:"index"`
+ Value uint64 `json:"value"`
+ ValueTwo uint64 `json:"value_two"`
+ Op Operator `json:"op"`
+}
+
+// Syscall is a rule to match a syscall in Seccomp
+type Syscall struct {
+ Name string `json:"name"`
+ Action Action `json:"action"`
+ Args []*Arg `json:"args"`
+}
+
+// TODO Windows. Many of these fields should be factored out into those parts
+// which are common across platforms, and those which are platform specific.
+
+// Config defines configuration options for executing a process inside a contained environment.
+type Config struct {
+ // NoPivotRoot will use MS_MOVE and a chroot to jail the process into the container's rootfs
+ // This is a common option when the container is running in ramdisk
+ NoPivotRoot bool `json:"no_pivot_root"`
+
+ // ParentDeathSignal specifies the signal that is sent to the container's process in the case
+ // that the parent process dies.
+ ParentDeathSignal int `json:"parent_death_signal"`
+
+ // Path to a directory containing the container's root filesystem.
+ Rootfs string `json:"rootfs"`
+
+ // Readonlyfs will remount the container's rootfs as readonly where only externally mounted
+ // bind mounts are writtable.
+ Readonlyfs bool `json:"readonlyfs"`
+
+ // Specifies the mount propagation flags to be applied to /.
+ RootPropagation int `json:"rootPropagation"`
+
+ // Mounts specify additional source and destination paths that will be mounted inside the container's
+ // rootfs and mount namespace if specified
+ Mounts []*Mount `json:"mounts"`
+
+ // The device nodes that should be automatically created within the container upon container start. Note, make sure that the node is marked as allowed in the cgroup as well!
+ Devices []*Device `json:"devices"`
+
+ MountLabel string `json:"mount_label"`
+
+ // Hostname optionally sets the container's hostname if provided
+ Hostname string `json:"hostname"`
+
+ // Namespaces specifies the container's namespaces that it should setup when cloning the init process
+ // If a namespace is not provided that namespace is shared from the container's parent process
+ Namespaces Namespaces `json:"namespaces"`
+
+ // Capabilities specify the capabilities to keep when executing the process inside the container
+ // All capabilities not specified will be dropped from the processes capability mask
+ Capabilities *Capabilities `json:"capabilities"`
+
+ // Networks specifies the container's network setup to be created
+ Networks []*Network `json:"networks"`
+
+ // Routes can be specified to create entries in the route table as the container is started
+ Routes []*Route `json:"routes"`
+
+ // Cgroups specifies specific cgroup settings for the various subsystems that the container is
+ // placed into to limit the resources the container has available
+ Cgroups *Cgroup `json:"cgroups"`
+
+ // AppArmorProfile specifies the profile to apply to the process running in the container and is
+ // change at the time the process is execed
+ AppArmorProfile string `json:"apparmor_profile,omitempty"`
+
+ // ProcessLabel specifies the label to apply to the process running in the container. It is
+ // commonly used by selinux
+ ProcessLabel string `json:"process_label,omitempty"`
+
+ // Rlimits specifies the resource limits, such as max open files, to set in the container
+ // If Rlimits are not set, the container will inherit rlimits from the parent process
+ Rlimits []Rlimit `json:"rlimits,omitempty"`
+
+ // OomScoreAdj specifies the adjustment to be made by the kernel when calculating oom scores
+ // for a process. Valid values are between the range [-1000, '1000'], where processes with
+ // higher scores are preferred for being killed.
+ // More information about kernel oom score calculation here: https://lwn.net/Articles/317814/
+ OomScoreAdj int `json:"oom_score_adj"`
+
+ // UidMappings is an array of User ID mappings for User Namespaces
+ UidMappings []IDMap `json:"uid_mappings"`
+
+ // GidMappings is an array of Group ID mappings for User Namespaces
+ GidMappings []IDMap `json:"gid_mappings"`
+
+ // MaskPaths specifies paths within the container's rootfs to mask over with a bind
+ // mount pointing to /dev/null as to prevent reads of the file.
+ MaskPaths []string `json:"mask_paths"`
+
+ // ReadonlyPaths specifies paths within the container's rootfs to remount as read-only
+ // so that these files prevent any writes.
+ ReadonlyPaths []string `json:"readonly_paths"`
+
+ // Sysctl is a map of properties and their values. It is the equivalent of using
+ // sysctl -w my.property.name value in Linux.
+ Sysctl map[string]string `json:"sysctl"`
+
+ // Seccomp allows actions to be taken whenever a syscall is made within the container.
+ // A number of rules are given, each having an action to be taken if a syscall matches it.
+ // A default action to be taken if no rules match is also given.
+ Seccomp *Seccomp `json:"seccomp"`
+
+ // NoNewPrivileges controls whether processes in the container can gain additional privileges.
+ NoNewPrivileges bool `json:"no_new_privileges,omitempty"`
+
+ // Hooks are a collection of actions to perform at various container lifecycle events.
+ // CommandHooks are serialized to JSON, but other hooks are not.
+ Hooks *Hooks
+
+ // Version is the version of opencontainer specification that is supported.
+ Version string `json:"version"`
+
+ // Labels are user defined metadata that is stored in the config and populated on the state
+ Labels []string `json:"labels"`
+
+ // NoNewKeyring will not allocated a new session keyring for the container. It will use the
+ // callers keyring in this case.
+ NoNewKeyring bool `json:"no_new_keyring"`
+
+ // Rootless specifies whether the container is a rootless container.
+ Rootless bool `json:"rootless"`
+}
+
+type Hooks struct {
+ // Prestart commands are executed after the container namespaces are created,
+ // but before the user supplied command is executed from init.
+ Prestart []Hook
+
+ // Poststart commands are executed after the container init process starts.
+ Poststart []Hook
+
+ // Poststop commands are executed after the container init process exits.
+ Poststop []Hook
+}
+
+type Capabilities struct {
+ // Bounding is the set of capabilities checked by the kernel.
+ Bounding []string
+ // Effective is the set of capabilities checked by the kernel.
+ Effective []string
+ // Inheritable is the capabilities preserved across execve.
+ Inheritable []string
+ // Permitted is the limiting superset for effective capabilities.
+ Permitted []string
+ // Ambient is the ambient set of capabilities that are kept.
+ Ambient []string
+}
+
+func (hooks *Hooks) UnmarshalJSON(b []byte) error {
+ var state struct {
+ Prestart []CommandHook
+ Poststart []CommandHook
+ Poststop []CommandHook
+ }
+
+ if err := json.Unmarshal(b, &state); err != nil {
+ return err
+ }
+
+ deserialize := func(shooks []CommandHook) (hooks []Hook) {
+ for _, shook := range shooks {
+ hooks = append(hooks, shook)
+ }
+
+ return hooks
+ }
+
+ hooks.Prestart = deserialize(state.Prestart)
+ hooks.Poststart = deserialize(state.Poststart)
+ hooks.Poststop = deserialize(state.Poststop)
+ return nil
+}
+
+func (hooks Hooks) MarshalJSON() ([]byte, error) {
+ serialize := func(hooks []Hook) (serializableHooks []CommandHook) {
+ for _, hook := range hooks {
+ switch chook := hook.(type) {
+ case CommandHook:
+ serializableHooks = append(serializableHooks, chook)
+ default:
+ logrus.Warnf("cannot serialize hook of type %T, skipping", hook)
+ }
+ }
+
+ return serializableHooks
+ }
+
+ return json.Marshal(map[string]interface{}{
+ "prestart": serialize(hooks.Prestart),
+ "poststart": serialize(hooks.Poststart),
+ "poststop": serialize(hooks.Poststop),
+ })
+}
+
+// HookState is the payload provided to a hook on execution.
+type HookState specs.State
+
+type Hook interface {
+ // Run executes the hook with the provided state.
+ Run(HookState) error
+}
+
+// NewFunctionHook will call the provided function when the hook is run.
+func NewFunctionHook(f func(HookState) error) FuncHook {
+ return FuncHook{
+ run: f,
+ }
+}
+
+type FuncHook struct {
+ run func(HookState) error
+}
+
+func (f FuncHook) Run(s HookState) error {
+ return f.run(s)
+}
+
+type Command struct {
+ Path string `json:"path"`
+ Args []string `json:"args"`
+ Env []string `json:"env"`
+ Dir string `json:"dir"`
+ Timeout *time.Duration `json:"timeout"`
+}
+
+// NewCommandHook will execute the provided command when the hook is run.
+func NewCommandHook(cmd Command) CommandHook {
+ return CommandHook{
+ Command: cmd,
+ }
+}
+
+type CommandHook struct {
+ Command
+}
+
+func (c Command) Run(s HookState) error {
+ b, err := json.Marshal(s)
+ if err != nil {
+ return err
+ }
+ var stdout, stderr bytes.Buffer
+ cmd := exec.Cmd{
+ Path: c.Path,
+ Args: c.Args,
+ Env: c.Env,
+ Stdin: bytes.NewReader(b),
+ Stdout: &stdout,
+ Stderr: &stderr,
+ }
+ if err := cmd.Start(); err != nil {
+ return err
+ }
+ errC := make(chan error, 1)
+ go func() {
+ err := cmd.Wait()
+ if err != nil {
+ err = fmt.Errorf("error running hook: %v, stdout: %s, stderr: %s", err, stdout.String(), stderr.String())
+ }
+ errC <- err
+ }()
+ var timerCh <-chan time.Time
+ if c.Timeout != nil {
+ timer := time.NewTimer(*c.Timeout)
+ defer timer.Stop()
+ timerCh = timer.C
+ }
+ select {
+ case err := <-errC:
+ return err
+ case <-timerCh:
+ cmd.Process.Kill()
+ cmd.Wait()
+ return fmt.Errorf("hook ran past specified timeout of %.1fs", c.Timeout.Seconds())
+ }
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/config_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/config_linux.go
new file mode 100644
index 000000000..07da10804
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/config_linux.go
@@ -0,0 +1,61 @@
+package configs
+
+import "fmt"
+
+// HostUID gets the translated uid for the process on host which could be
+// different when user namespaces are enabled.
+func (c Config) HostUID(containerId int) (int, error) {
+ if c.Namespaces.Contains(NEWUSER) {
+ if c.UidMappings == nil {
+ return -1, fmt.Errorf("User namespaces enabled, but no uid mappings found.")
+ }
+ id, found := c.hostIDFromMapping(containerId, c.UidMappings)
+ if !found {
+ return -1, fmt.Errorf("User namespaces enabled, but no user mapping found.")
+ }
+ return id, nil
+ }
+ // Return unchanged id.
+ return containerId, nil
+}
+
+// HostRootUID gets the root uid for the process on host which could be non-zero
+// when user namespaces are enabled.
+func (c Config) HostRootUID() (int, error) {
+ return c.HostUID(0)
+}
+
+// HostGID gets the translated gid for the process on host which could be
+// different when user namespaces are enabled.
+func (c Config) HostGID(containerId int) (int, error) {
+ if c.Namespaces.Contains(NEWUSER) {
+ if c.GidMappings == nil {
+ return -1, fmt.Errorf("User namespaces enabled, but no gid mappings found.")
+ }
+ id, found := c.hostIDFromMapping(containerId, c.GidMappings)
+ if !found {
+ return -1, fmt.Errorf("User namespaces enabled, but no group mapping found.")
+ }
+ return id, nil
+ }
+ // Return unchanged id.
+ return containerId, nil
+}
+
+// HostRootGID gets the root gid for the process on host which could be non-zero
+// when user namespaces are enabled.
+func (c Config) HostRootGID() (int, error) {
+ return c.HostGID(0)
+}
+
+// Utility function that gets a host ID for a container ID from user namespace map
+// if that ID is present in the map.
+func (c Config) hostIDFromMapping(containerID int, uMap []IDMap) (int, bool) {
+ for _, m := range uMap {
+ if (containerID >= m.ContainerID) && (containerID <= (m.ContainerID + m.Size - 1)) {
+ hostID := m.HostID + (containerID - m.ContainerID)
+ return hostID, true
+ }
+ }
+ return -1, false
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/device.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/device.go
new file mode 100644
index 000000000..8701bb212
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/device.go
@@ -0,0 +1,57 @@
+package configs
+
+import (
+ "fmt"
+ "os"
+)
+
+const (
+ Wildcard = -1
+)
+
+// TODO Windows: This can be factored out in the future
+
+type Device struct {
+ // Device type, block, char, etc.
+ Type rune `json:"type"`
+
+ // Path to the device.
+ Path string `json:"path"`
+
+ // Major is the device's major number.
+ Major int64 `json:"major"`
+
+ // Minor is the device's minor number.
+ Minor int64 `json:"minor"`
+
+ // Cgroup permissions format, rwm.
+ Permissions string `json:"permissions"`
+
+ // FileMode permission bits for the device.
+ FileMode os.FileMode `json:"file_mode"`
+
+ // Uid of the device.
+ Uid uint32 `json:"uid"`
+
+ // Gid of the device.
+ Gid uint32 `json:"gid"`
+
+ // Write the file to the allowed list
+ Allow bool `json:"allow"`
+}
+
+func (d *Device) CgroupString() string {
+ return fmt.Sprintf("%c %s:%s %s", d.Type, deviceNumberString(d.Major), deviceNumberString(d.Minor), d.Permissions)
+}
+
+func (d *Device) Mkdev() int {
+ return int((d.Major << 8) | (d.Minor & 0xff) | ((d.Minor & 0xfff00) << 12))
+}
+
+// deviceNumberString converts the device number to a string return result.
+func deviceNumberString(number int64) string {
+ if number == Wildcard {
+ return "*"
+ }
+ return fmt.Sprint(number)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/device_defaults.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/device_defaults.go
new file mode 100644
index 000000000..4d348d217
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/device_defaults.go
@@ -0,0 +1,111 @@
+// +build linux freebsd
+
+package configs
+
+var (
+ // DefaultSimpleDevices are devices that are to be both allowed and created.
+ DefaultSimpleDevices = []*Device{
+ // /dev/null and zero
+ {
+ Path: "/dev/null",
+ Type: 'c',
+ Major: 1,
+ Minor: 3,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+ {
+ Path: "/dev/zero",
+ Type: 'c',
+ Major: 1,
+ Minor: 5,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+
+ {
+ Path: "/dev/full",
+ Type: 'c',
+ Major: 1,
+ Minor: 7,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+
+ // consoles and ttys
+ {
+ Path: "/dev/tty",
+ Type: 'c',
+ Major: 5,
+ Minor: 0,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+
+ // /dev/urandom,/dev/random
+ {
+ Path: "/dev/urandom",
+ Type: 'c',
+ Major: 1,
+ Minor: 9,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+ {
+ Path: "/dev/random",
+ Type: 'c',
+ Major: 1,
+ Minor: 8,
+ Permissions: "rwm",
+ FileMode: 0666,
+ },
+ }
+ DefaultAllowedDevices = append([]*Device{
+ // allow mknod for any device
+ {
+ Type: 'c',
+ Major: Wildcard,
+ Minor: Wildcard,
+ Permissions: "m",
+ },
+ {
+ Type: 'b',
+ Major: Wildcard,
+ Minor: Wildcard,
+ Permissions: "m",
+ },
+
+ {
+ Path: "/dev/console",
+ Type: 'c',
+ Major: 5,
+ Minor: 1,
+ Permissions: "rwm",
+ },
+ // /dev/pts/ - pts namespaces are "coming soon"
+ {
+ Path: "",
+ Type: 'c',
+ Major: 136,
+ Minor: Wildcard,
+ Permissions: "rwm",
+ },
+ {
+ Path: "",
+ Type: 'c',
+ Major: 5,
+ Minor: 2,
+ Permissions: "rwm",
+ },
+
+ // tuntap
+ {
+ Path: "",
+ Type: 'c',
+ Major: 10,
+ Minor: 200,
+ Permissions: "rwm",
+ },
+ }, DefaultSimpleDevices...)
+ DefaultAutoCreatedDevices = append([]*Device{}, DefaultSimpleDevices...)
+)
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/hugepage_limit.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/hugepage_limit.go
new file mode 100644
index 000000000..d30216380
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/hugepage_limit.go
@@ -0,0 +1,9 @@
+package configs
+
+type HugepageLimit struct {
+ // which type of hugepage to limit.
+ Pagesize string `json:"page_size"`
+
+ // usage limit for hugepage.
+ Limit uint64 `json:"limit"`
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/interface_priority_map.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/interface_priority_map.go
new file mode 100644
index 000000000..9a0395eaf
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/interface_priority_map.go
@@ -0,0 +1,14 @@
+package configs
+
+import (
+ "fmt"
+)
+
+type IfPrioMap struct {
+ Interface string `json:"interface"`
+ Priority int64 `json:"priority"`
+}
+
+func (i *IfPrioMap) CgroupString() string {
+ return fmt.Sprintf("%s %d", i.Interface, i.Priority)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/mount.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/mount.go
new file mode 100644
index 000000000..670757ddb
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/mount.go
@@ -0,0 +1,39 @@
+package configs
+
+const (
+ // EXT_COPYUP is a directive to copy up the contents of a directory when
+ // a tmpfs is mounted over it.
+ EXT_COPYUP = 1 << iota
+)
+
+type Mount struct {
+ // Source path for the mount.
+ Source string `json:"source"`
+
+ // Destination path for the mount inside the container.
+ Destination string `json:"destination"`
+
+ // Device the mount is for.
+ Device string `json:"device"`
+
+ // Mount flags.
+ Flags int `json:"flags"`
+
+ // Propagation Flags
+ PropagationFlags []int `json:"propagation_flags"`
+
+ // Mount data applied to the mount.
+ Data string `json:"data"`
+
+ // Relabel source if set, "z" indicates shared, "Z" indicates unshared.
+ Relabel string `json:"relabel"`
+
+ // Extensions are additional flags that are specific to runc.
+ Extensions int `json:"extensions"`
+
+ // Optional Command to be run before Source is mounted.
+ PremountCmds []Command `json:"premount_cmds"`
+
+ // Optional Command to be run after Source is mounted.
+ PostmountCmds []Command `json:"postmount_cmds"`
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces.go
new file mode 100644
index 000000000..a3329a31a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces.go
@@ -0,0 +1,5 @@
+package configs
+
+type NamespaceType string
+
+type Namespaces []Namespace
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_linux.go
new file mode 100644
index 000000000..5fc171a57
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_linux.go
@@ -0,0 +1,122 @@
+package configs
+
+import (
+ "fmt"
+ "os"
+ "sync"
+)
+
+const (
+ NEWNET NamespaceType = "NEWNET"
+ NEWPID NamespaceType = "NEWPID"
+ NEWNS NamespaceType = "NEWNS"
+ NEWUTS NamespaceType = "NEWUTS"
+ NEWIPC NamespaceType = "NEWIPC"
+ NEWUSER NamespaceType = "NEWUSER"
+)
+
+var (
+ nsLock sync.Mutex
+ supportedNamespaces = make(map[NamespaceType]bool)
+)
+
+// NsName converts the namespace type to its filename
+func NsName(ns NamespaceType) string {
+ switch ns {
+ case NEWNET:
+ return "net"
+ case NEWNS:
+ return "mnt"
+ case NEWPID:
+ return "pid"
+ case NEWIPC:
+ return "ipc"
+ case NEWUSER:
+ return "user"
+ case NEWUTS:
+ return "uts"
+ }
+ return ""
+}
+
+// IsNamespaceSupported returns whether a namespace is available or
+// not
+func IsNamespaceSupported(ns NamespaceType) bool {
+ nsLock.Lock()
+ defer nsLock.Unlock()
+ supported, ok := supportedNamespaces[ns]
+ if ok {
+ return supported
+ }
+ nsFile := NsName(ns)
+ // if the namespace type is unknown, just return false
+ if nsFile == "" {
+ return false
+ }
+ _, err := os.Stat(fmt.Sprintf("/proc/self/ns/%s", nsFile))
+ // a namespace is supported if it exists and we have permissions to read it
+ supported = err == nil
+ supportedNamespaces[ns] = supported
+ return supported
+}
+
+func NamespaceTypes() []NamespaceType {
+ return []NamespaceType{
+ NEWUSER, // Keep user NS always first, don't move it.
+ NEWIPC,
+ NEWUTS,
+ NEWNET,
+ NEWPID,
+ NEWNS,
+ }
+}
+
+// Namespace defines configuration for each namespace. It specifies an
+// alternate path that is able to be joined via setns.
+type Namespace struct {
+ Type NamespaceType `json:"type"`
+ Path string `json:"path"`
+}
+
+func (n *Namespace) GetPath(pid int) string {
+ return fmt.Sprintf("/proc/%d/ns/%s", pid, NsName(n.Type))
+}
+
+func (n *Namespaces) Remove(t NamespaceType) bool {
+ i := n.index(t)
+ if i == -1 {
+ return false
+ }
+ *n = append((*n)[:i], (*n)[i+1:]...)
+ return true
+}
+
+func (n *Namespaces) Add(t NamespaceType, path string) {
+ i := n.index(t)
+ if i == -1 {
+ *n = append(*n, Namespace{Type: t, Path: path})
+ return
+ }
+ (*n)[i].Path = path
+}
+
+func (n *Namespaces) index(t NamespaceType) int {
+ for i, ns := range *n {
+ if ns.Type == t {
+ return i
+ }
+ }
+ return -1
+}
+
+func (n *Namespaces) Contains(t NamespaceType) bool {
+ return n.index(t) != -1
+}
+
+func (n *Namespaces) PathOf(t NamespaceType) string {
+ i := n.index(t)
+ if i == -1 {
+ return ""
+ }
+ return (*n)[i].Path
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall.go
new file mode 100644
index 000000000..4ce6813d2
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall.go
@@ -0,0 +1,31 @@
+// +build linux
+
+package configs
+
+import "golang.org/x/sys/unix"
+
+func (n *Namespace) Syscall() int {
+ return namespaceInfo[n.Type]
+}
+
+var namespaceInfo = map[NamespaceType]int{
+ NEWNET: unix.CLONE_NEWNET,
+ NEWNS: unix.CLONE_NEWNS,
+ NEWUSER: unix.CLONE_NEWUSER,
+ NEWIPC: unix.CLONE_NEWIPC,
+ NEWUTS: unix.CLONE_NEWUTS,
+ NEWPID: unix.CLONE_NEWPID,
+}
+
+// CloneFlags parses the container's Namespaces options to set the correct
+// flags on clone, unshare. This function returns flags only for new namespaces.
+func (n *Namespaces) CloneFlags() uintptr {
+ var flag int
+ for _, v := range *n {
+ if v.Path != "" {
+ continue
+ }
+ flag |= namespaceInfo[v.Type]
+ }
+ return uintptr(flag)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall_unsupported.go
new file mode 100644
index 000000000..5d9a5c81f
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_syscall_unsupported.go
@@ -0,0 +1,13 @@
+// +build !linux,!windows
+
+package configs
+
+func (n *Namespace) Syscall() int {
+ panic("No namespace syscall support")
+}
+
+// CloneFlags parses the container's Namespaces options to set the correct
+// flags on clone, unshare. This function returns flags only for new namespaces.
+func (n *Namespaces) CloneFlags() uintptr {
+ panic("No namespace syscall support")
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_unsupported.go
new file mode 100644
index 000000000..19bf713de
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/namespaces_unsupported.go
@@ -0,0 +1,8 @@
+// +build !linux
+
+package configs
+
+// Namespace defines configuration for each namespace. It specifies an
+// alternate path that is able to be joined via setns.
+type Namespace struct {
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/network.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/network.go
new file mode 100644
index 000000000..ccdb228e1
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/network.go
@@ -0,0 +1,72 @@
+package configs
+
+// Network defines configuration for a container's networking stack
+//
+// The network configuration can be omitted from a container causing the
+// container to be setup with the host's networking stack
+type Network struct {
+ // Type sets the networks type, commonly veth and loopback
+ Type string `json:"type"`
+
+ // Name of the network interface
+ Name string `json:"name"`
+
+ // The bridge to use.
+ Bridge string `json:"bridge"`
+
+ // MacAddress contains the MAC address to set on the network interface
+ MacAddress string `json:"mac_address"`
+
+ // Address contains the IPv4 and mask to set on the network interface
+ Address string `json:"address"`
+
+ // Gateway sets the gateway address that is used as the default for the interface
+ Gateway string `json:"gateway"`
+
+ // IPv6Address contains the IPv6 and mask to set on the network interface
+ IPv6Address string `json:"ipv6_address"`
+
+ // IPv6Gateway sets the ipv6 gateway address that is used as the default for the interface
+ IPv6Gateway string `json:"ipv6_gateway"`
+
+ // Mtu sets the mtu value for the interface and will be mirrored on both the host and
+ // container's interfaces if a pair is created, specifically in the case of type veth
+ // Note: This does not apply to loopback interfaces.
+ Mtu int `json:"mtu"`
+
+ // TxQueueLen sets the tx_queuelen value for the interface and will be mirrored on both the host and
+ // container's interfaces if a pair is created, specifically in the case of type veth
+ // Note: This does not apply to loopback interfaces.
+ TxQueueLen int `json:"txqueuelen"`
+
+ // HostInterfaceName is a unique name of a veth pair that resides on in the host interface of the
+ // container.
+ HostInterfaceName string `json:"host_interface_name"`
+
+ // HairpinMode specifies if hairpin NAT should be enabled on the virtual interface
+ // bridge port in the case of type veth
+ // Note: This is unsupported on some systems.
+ // Note: This does not apply to loopback interfaces.
+ HairpinMode bool `json:"hairpin_mode"`
+}
+
+// Routes can be specified to create entries in the route table as the container is started
+//
+// All of destination, source, and gateway should be either IPv4 or IPv6.
+// One of the three options must be present, and omitted entries will use their
+// IP family default for the route table. For IPv4 for example, setting the
+// gateway to 1.2.3.4 and the interface to eth0 will set up a standard
+// destination of 0.0.0.0(or *) when viewed in the route table.
+type Route struct {
+ // Sets the destination and mask, should be a CIDR. Accepts IPv4 and IPv6
+ Destination string `json:"destination"`
+
+ // Sets the source and mask, should be a CIDR. Accepts IPv4 and IPv6
+ Source string `json:"source"`
+
+ // Sets the gateway. Accepts IPv4 and IPv6
+ Gateway string `json:"gateway"`
+
+ // The device to set this route up for, for example: eth0
+ InterfaceName string `json:"interface_name"`
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/rootless.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/rootless.go
new file mode 100644
index 000000000..0cebfaf80
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/rootless.go
@@ -0,0 +1,117 @@
+package validate
+
+import (
+ "fmt"
+ "os"
+ "reflect"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+var (
+ geteuid = os.Geteuid
+ getegid = os.Getegid
+)
+
+func (v *ConfigValidator) rootless(config *configs.Config) error {
+ if err := rootlessMappings(config); err != nil {
+ return err
+ }
+ if err := rootlessMount(config); err != nil {
+ return err
+ }
+ // Currently, cgroups cannot effectively be used in rootless containers.
+ // The new cgroup namespace doesn't really help us either because it doesn't
+ // have nice interactions with the user namespace (we're working with upstream
+ // to fix this).
+ if err := rootlessCgroup(config); err != nil {
+ return err
+ }
+
+ // XXX: We currently can't verify the user config at all, because
+ // configs.Config doesn't store the user-related configs. So this
+ // has to be verified by setupUser() in init_linux.go.
+
+ return nil
+}
+
+func rootlessMappings(config *configs.Config) error {
+ rootuid, err := config.HostRootUID()
+ if err != nil {
+ return fmt.Errorf("failed to get root uid from uidMappings: %v", err)
+ }
+ if euid := geteuid(); euid != 0 {
+ if !config.Namespaces.Contains(configs.NEWUSER) {
+ return fmt.Errorf("rootless containers require user namespaces")
+ }
+ if rootuid != euid {
+ return fmt.Errorf("rootless containers cannot map container root to a different host user")
+ }
+ }
+
+ rootgid, err := config.HostRootGID()
+ if err != nil {
+ return fmt.Errorf("failed to get root gid from gidMappings: %v", err)
+ }
+
+ // Similar to the above test, we need to make sure that we aren't trying to
+ // map to a group ID that we don't have the right to be.
+ if rootgid != getegid() {
+ return fmt.Errorf("rootless containers cannot map container root to a different host group")
+ }
+
+ // We can only map one user and group inside a container (our own).
+ if len(config.UidMappings) != 1 || config.UidMappings[0].Size != 1 {
+ return fmt.Errorf("rootless containers cannot map more than one user")
+ }
+ if len(config.GidMappings) != 1 || config.GidMappings[0].Size != 1 {
+ return fmt.Errorf("rootless containers cannot map more than one group")
+ }
+
+ return nil
+}
+
+// cgroup verifies that the user isn't trying to set any cgroup limits or paths.
+func rootlessCgroup(config *configs.Config) error {
+ // Nothing set at all.
+ if config.Cgroups == nil || config.Cgroups.Resources == nil {
+ return nil
+ }
+
+ // Used for comparing to the zero value.
+ left := reflect.ValueOf(*config.Cgroups.Resources)
+ right := reflect.Zero(left.Type())
+
+ // This is all we need to do, since specconv won't add cgroup options in
+ // rootless mode.
+ if !reflect.DeepEqual(left.Interface(), right.Interface()) {
+ return fmt.Errorf("cannot specify resource limits in rootless container")
+ }
+
+ return nil
+}
+
+// mount verifies that the user isn't trying to set up any mounts they don't have
+// the rights to do. In addition, it makes sure that no mount has a `uid=` or
+// `gid=` option that doesn't resolve to root.
+func rootlessMount(config *configs.Config) error {
+ // XXX: We could whitelist allowed devices at this point, but I'm not
+ // convinced that's a good idea. The kernel is the best arbiter of
+ // access control.
+
+ for _, mount := range config.Mounts {
+ // Check that the options list doesn't contain any uid= or gid= entries
+ // that don't resolve to root.
+ for _, opt := range strings.Split(mount.Data, ",") {
+ if strings.HasPrefix(opt, "uid=") && opt != "uid=0" {
+ return fmt.Errorf("cannot specify uid= mount options in rootless containers where argument isn't 0")
+ }
+ if strings.HasPrefix(opt, "gid=") && opt != "gid=0" {
+ return fmt.Errorf("cannot specify gid= mount options in rootless containers where argument isn't 0")
+ }
+ }
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/validator.go b/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/validator.go
new file mode 100644
index 000000000..828434544
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/configs/validate/validator.go
@@ -0,0 +1,195 @@
+package validate
+
+import (
+ "fmt"
+ "os"
+ "path/filepath"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+ selinux "github.com/opencontainers/selinux/go-selinux"
+)
+
+type Validator interface {
+ Validate(*configs.Config) error
+}
+
+func New() Validator {
+ return &ConfigValidator{}
+}
+
+type ConfigValidator struct {
+}
+
+func (v *ConfigValidator) Validate(config *configs.Config) error {
+ if err := v.rootfs(config); err != nil {
+ return err
+ }
+ if err := v.network(config); err != nil {
+ return err
+ }
+ if err := v.hostname(config); err != nil {
+ return err
+ }
+ if err := v.security(config); err != nil {
+ return err
+ }
+ if err := v.usernamespace(config); err != nil {
+ return err
+ }
+ if err := v.sysctl(config); err != nil {
+ return err
+ }
+ if config.Rootless {
+ if err := v.rootless(config); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// rootfs validates if the rootfs is an absolute path and is not a symlink
+// to the container's root filesystem.
+func (v *ConfigValidator) rootfs(config *configs.Config) error {
+ if _, err := os.Stat(config.Rootfs); err != nil {
+ if os.IsNotExist(err) {
+ return fmt.Errorf("rootfs (%s) does not exist", config.Rootfs)
+ }
+ return err
+ }
+ cleaned, err := filepath.Abs(config.Rootfs)
+ if err != nil {
+ return err
+ }
+ if cleaned, err = filepath.EvalSymlinks(cleaned); err != nil {
+ return err
+ }
+ if filepath.Clean(config.Rootfs) != cleaned {
+ return fmt.Errorf("%s is not an absolute path or is a symlink", config.Rootfs)
+ }
+ return nil
+}
+
+func (v *ConfigValidator) network(config *configs.Config) error {
+ if !config.Namespaces.Contains(configs.NEWNET) {
+ if len(config.Networks) > 0 || len(config.Routes) > 0 {
+ return fmt.Errorf("unable to apply network settings without a private NET namespace")
+ }
+ }
+ return nil
+}
+
+func (v *ConfigValidator) hostname(config *configs.Config) error {
+ if config.Hostname != "" && !config.Namespaces.Contains(configs.NEWUTS) {
+ return fmt.Errorf("unable to set hostname without a private UTS namespace")
+ }
+ return nil
+}
+
+func (v *ConfigValidator) security(config *configs.Config) error {
+ // restrict sys without mount namespace
+ if (len(config.MaskPaths) > 0 || len(config.ReadonlyPaths) > 0) &&
+ !config.Namespaces.Contains(configs.NEWNS) {
+ return fmt.Errorf("unable to restrict sys entries without a private MNT namespace")
+ }
+ if config.ProcessLabel != "" && !selinux.GetEnabled() {
+ return fmt.Errorf("selinux label is specified in config, but selinux is disabled or not supported")
+ }
+
+ return nil
+}
+
+func (v *ConfigValidator) usernamespace(config *configs.Config) error {
+ if config.Namespaces.Contains(configs.NEWUSER) {
+ if _, err := os.Stat("/proc/self/ns/user"); os.IsNotExist(err) {
+ return fmt.Errorf("USER namespaces aren't enabled in the kernel")
+ }
+ } else {
+ if config.UidMappings != nil || config.GidMappings != nil {
+ return fmt.Errorf("User namespace mappings specified, but USER namespace isn't enabled in the config")
+ }
+ }
+ return nil
+}
+
+// sysctl validates that the specified sysctl keys are valid or not.
+// /proc/sys isn't completely namespaced and depending on which namespaces
+// are specified, a subset of sysctls are permitted.
+func (v *ConfigValidator) sysctl(config *configs.Config) error {
+ validSysctlMap := map[string]bool{
+ "kernel.msgmax": true,
+ "kernel.msgmnb": true,
+ "kernel.msgmni": true,
+ "kernel.sem": true,
+ "kernel.shmall": true,
+ "kernel.shmmax": true,
+ "kernel.shmmni": true,
+ "kernel.shm_rmid_forced": true,
+ }
+
+ for s := range config.Sysctl {
+ if validSysctlMap[s] || strings.HasPrefix(s, "fs.mqueue.") {
+ if config.Namespaces.Contains(configs.NEWIPC) {
+ continue
+ } else {
+ return fmt.Errorf("sysctl %q is not allowed in the hosts ipc namespace", s)
+ }
+ }
+ if strings.HasPrefix(s, "net.") {
+ if config.Namespaces.Contains(configs.NEWNET) {
+ if path := config.Namespaces.PathOf(configs.NEWNET); path != "" {
+ if err := checkHostNs(s, path); err != nil {
+ return err
+ }
+ }
+ continue
+ } else {
+ return fmt.Errorf("sysctl %q is not allowed in the hosts network namespace", s)
+ }
+ }
+ return fmt.Errorf("sysctl %q is not in a separate kernel namespace", s)
+ }
+
+ return nil
+}
+
+func isSymbolicLink(path string) (bool, error) {
+ fi, err := os.Lstat(path)
+ if err != nil {
+ return false, err
+ }
+
+ return fi.Mode()&os.ModeSymlink == os.ModeSymlink, nil
+}
+
+// checkHostNs checks whether network sysctl is used in host namespace.
+func checkHostNs(sysctlConfig string, path string) error {
+ var currentProcessNetns = "/proc/self/ns/net"
+ // readlink on the current processes network namespace
+ destOfCurrentProcess, err := os.Readlink(currentProcessNetns)
+ if err != nil {
+ return fmt.Errorf("read soft link %q error", currentProcessNetns)
+ }
+
+ // First check if the provided path is a symbolic link
+ symLink, err := isSymbolicLink(path)
+ if err != nil {
+ return fmt.Errorf("could not check that %q is a symlink: %v", path, err)
+ }
+
+ if symLink == false {
+ // The provided namespace is not a symbolic link,
+ // it is not the host namespace.
+ return nil
+ }
+
+ // readlink on the path provided in the struct
+ destOfContainer, err := os.Readlink(path)
+ if err != nil {
+ return fmt.Errorf("read soft link %q error", path)
+ }
+ if destOfContainer == destOfCurrentProcess {
+ return fmt.Errorf("sysctl %q is not allowed in the hosts network namespace", sysctlConfig)
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/console.go b/vendor/github.com/opencontainers/runc/libcontainer/console.go
new file mode 100644
index 000000000..917acc702
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/console.go
@@ -0,0 +1,17 @@
+package libcontainer
+
+import (
+ "io"
+ "os"
+)
+
+// Console represents a pseudo TTY.
+type Console interface {
+ io.ReadWriteCloser
+
+ // Path returns the filesystem path to the slave side of the pty.
+ Path() string
+
+ // Fd returns the fd for the master of the pty.
+ File() *os.File
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/console_freebsd.go b/vendor/github.com/opencontainers/runc/libcontainer/console_freebsd.go
new file mode 100644
index 000000000..b7166a31f
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/console_freebsd.go
@@ -0,0 +1,13 @@
+// +build freebsd
+
+package libcontainer
+
+import (
+ "errors"
+)
+
+// newConsole returns an initialized console that can be used within a container by copying bytes
+// from the master side to the slave that is attached as the tty for the container's init process.
+func newConsole() (Console, error) {
+ return nil, errors.New("libcontainer console is not supported on FreeBSD")
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/console_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/console_linux.go
new file mode 100644
index 000000000..f70de3848
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/console_linux.go
@@ -0,0 +1,152 @@
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+ "unsafe"
+
+ "golang.org/x/sys/unix"
+)
+
+func ConsoleFromFile(f *os.File) Console {
+ return &linuxConsole{
+ master: f,
+ }
+}
+
+// newConsole returns an initialized console that can be used within a container by copying bytes
+// from the master side to the slave that is attached as the tty for the container's init process.
+func newConsole() (Console, error) {
+ master, err := os.OpenFile("/dev/ptmx", unix.O_RDWR|unix.O_NOCTTY|unix.O_CLOEXEC, 0)
+ if err != nil {
+ return nil, err
+ }
+ console, err := ptsname(master)
+ if err != nil {
+ return nil, err
+ }
+ if err := unlockpt(master); err != nil {
+ return nil, err
+ }
+ return &linuxConsole{
+ slavePath: console,
+ master: master,
+ }, nil
+}
+
+// linuxConsole is a linux pseudo TTY for use within a container.
+type linuxConsole struct {
+ master *os.File
+ slavePath string
+}
+
+func (c *linuxConsole) File() *os.File {
+ return c.master
+}
+
+func (c *linuxConsole) Path() string {
+ return c.slavePath
+}
+
+func (c *linuxConsole) Read(b []byte) (int, error) {
+ return c.master.Read(b)
+}
+
+func (c *linuxConsole) Write(b []byte) (int, error) {
+ return c.master.Write(b)
+}
+
+func (c *linuxConsole) Close() error {
+ if m := c.master; m != nil {
+ return m.Close()
+ }
+ return nil
+}
+
+// mount initializes the console inside the rootfs mounting with the specified mount label
+// and applying the correct ownership of the console.
+func (c *linuxConsole) mount() error {
+ oldMask := unix.Umask(0000)
+ defer unix.Umask(oldMask)
+ f, err := os.Create("/dev/console")
+ if err != nil && !os.IsExist(err) {
+ return err
+ }
+ if f != nil {
+ f.Close()
+ }
+ return unix.Mount(c.slavePath, "/dev/console", "bind", unix.MS_BIND, "")
+}
+
+// dupStdio opens the slavePath for the console and dups the fds to the current
+// processes stdio, fd 0,1,2.
+func (c *linuxConsole) dupStdio() error {
+ slave, err := c.open(unix.O_RDWR)
+ if err != nil {
+ return err
+ }
+ fd := int(slave.Fd())
+ for _, i := range []int{0, 1, 2} {
+ if err := unix.Dup3(fd, i, 0); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// open is a clone of os.OpenFile without the O_CLOEXEC used to open the pty slave.
+func (c *linuxConsole) open(flag int) (*os.File, error) {
+ r, e := unix.Open(c.slavePath, flag, 0)
+ if e != nil {
+ return nil, &os.PathError{
+ Op: "open",
+ Path: c.slavePath,
+ Err: e,
+ }
+ }
+ return os.NewFile(uintptr(r), c.slavePath), nil
+}
+
+func ioctl(fd uintptr, flag, data uintptr) error {
+ if _, _, err := unix.Syscall(unix.SYS_IOCTL, fd, flag, data); err != 0 {
+ return err
+ }
+ return nil
+}
+
+// unlockpt unlocks the slave pseudoterminal device corresponding to the master pseudoterminal referred to by f.
+// unlockpt should be called before opening the slave side of a pty.
+func unlockpt(f *os.File) error {
+ var u int32
+ return ioctl(f.Fd(), unix.TIOCSPTLCK, uintptr(unsafe.Pointer(&u)))
+}
+
+// ptsname retrieves the name of the first available pts for the given master.
+func ptsname(f *os.File) (string, error) {
+ n, err := unix.IoctlGetInt(int(f.Fd()), unix.TIOCGPTN)
+ if err != nil {
+ return "", err
+ }
+ return fmt.Sprintf("/dev/pts/%d", n), nil
+}
+
+// SaneTerminal sets the necessary tty_ioctl(4)s to ensure that a pty pair
+// created by us acts normally. In particular, a not-very-well-known default of
+// Linux unix98 ptys is that they have +onlcr by default. While this isn't a
+// problem for terminal emulators, because we relay data from the terminal we
+// also relay that funky line discipline.
+func SaneTerminal(terminal *os.File) error {
+ termios, err := unix.IoctlGetTermios(int(terminal.Fd()), unix.TCGETS)
+ if err != nil {
+ return fmt.Errorf("ioctl(tty, tcgets): %s", err.Error())
+ }
+
+ // Set -onlcr so we don't have to deal with \r.
+ termios.Oflag &^= unix.ONLCR
+
+ if err := unix.IoctlSetTermios(int(terminal.Fd()), unix.TCSETS, termios); err != nil {
+ return fmt.Errorf("ioctl(tty, tcsets): %s", err.Error())
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/console_solaris.go b/vendor/github.com/opencontainers/runc/libcontainer/console_solaris.go
new file mode 100644
index 000000000..e5ca54599
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/console_solaris.go
@@ -0,0 +1,11 @@
+package libcontainer
+
+import (
+ "errors"
+)
+
+// newConsole returns an initialized console that can be used within a container by copying bytes
+// from the master side to the slave that is attached as the tty for the container's init process.
+func newConsole() (Console, error) {
+ return nil, errors.New("libcontainer console is not supported on Solaris")
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/console_windows.go b/vendor/github.com/opencontainers/runc/libcontainer/console_windows.go
new file mode 100644
index 000000000..c61e866a5
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/console_windows.go
@@ -0,0 +1,30 @@
+package libcontainer
+
+// newConsole returns an initialized console that can be used within a container
+func newConsole() (Console, error) {
+ return &windowsConsole{}, nil
+}
+
+// windowsConsole is a Windows pseudo TTY for use within a container.
+type windowsConsole struct {
+}
+
+func (c *windowsConsole) Fd() uintptr {
+ return 0
+}
+
+func (c *windowsConsole) Path() string {
+ return ""
+}
+
+func (c *windowsConsole) Read(b []byte) (int, error) {
+ return 0, nil
+}
+
+func (c *windowsConsole) Write(b []byte) (int, error) {
+ return 0, nil
+}
+
+func (c *windowsConsole) Close() error {
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/container.go b/vendor/github.com/opencontainers/runc/libcontainer/container.go
new file mode 100644
index 000000000..2e31b4d4f
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/container.go
@@ -0,0 +1,166 @@
+// Package libcontainer provides a native Go implementation for creating containers
+// with namespaces, cgroups, capabilities, and filesystem access controls.
+// It allows you to manage the lifecycle of the container performing additional operations
+// after the container is created.
+package libcontainer
+
+import (
+ "os"
+ "time"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+// Status is the status of a container.
+type Status int
+
+const (
+ // Created is the status that denotes the container exists but has not been run yet.
+ Created Status = iota
+ // Running is the status that denotes the container exists and is running.
+ Running
+ // Pausing is the status that denotes the container exists, it is in the process of being paused.
+ Pausing
+ // Paused is the status that denotes the container exists, but all its processes are paused.
+ Paused
+ // Stopped is the status that denotes the container does not have a created or running process.
+ Stopped
+)
+
+func (s Status) String() string {
+ switch s {
+ case Created:
+ return "created"
+ case Running:
+ return "running"
+ case Pausing:
+ return "pausing"
+ case Paused:
+ return "paused"
+ case Stopped:
+ return "stopped"
+ default:
+ return "unknown"
+ }
+}
+
+// BaseState represents the platform agnostic pieces relating to a
+// running container's state
+type BaseState struct {
+ // ID is the container ID.
+ ID string `json:"id"`
+
+ // InitProcessPid is the init process id in the parent namespace.
+ InitProcessPid int `json:"init_process_pid"`
+
+ // InitProcessStartTime is the init process start time in clock cycles since boot time.
+ InitProcessStartTime uint64 `json:"init_process_start"`
+
+ // Created is the unix timestamp for the creation time of the container in UTC
+ Created time.Time `json:"created"`
+
+ // Config is the container's configuration.
+ Config configs.Config `json:"config"`
+}
+
+// BaseContainer is a libcontainer container object.
+//
+// Each container is thread-safe within the same process. Since a container can
+// be destroyed by a separate process, any function may return that the container
+// was not found. BaseContainer includes methods that are platform agnostic.
+type BaseContainer interface {
+ // Returns the ID of the container
+ ID() string
+
+ // Returns the current status of the container.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // Systemerror - System error.
+ Status() (Status, error)
+
+ // State returns the current container's state information.
+ //
+ // errors:
+ // SystemError - System error.
+ State() (*State, error)
+
+ // Returns the current config of the container.
+ Config() configs.Config
+
+ // Returns the PIDs inside this container. The PIDs are in the namespace of the calling process.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // Systemerror - System error.
+ //
+ // Some of the returned PIDs may no longer refer to processes in the Container, unless
+ // the Container state is PAUSED in which case every PID in the slice is valid.
+ Processes() ([]int, error)
+
+ // Returns statistics for the container.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // Systemerror - System error.
+ Stats() (*Stats, error)
+
+ // Set resources of container as configured
+ //
+ // We can use this to change resources when containers are running.
+ //
+ // errors:
+ // SystemError - System error.
+ Set(config configs.Config) error
+
+ // Start a process inside the container. Returns error if process fails to
+ // start. You can track process lifecycle with passed Process structure.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // ConfigInvalid - config is invalid,
+ // ContainerPaused - Container is paused,
+ // SystemError - System error.
+ Start(process *Process) (err error)
+
+ // Run immediately starts the process inside the container. Returns error if process
+ // fails to start. It does not block waiting for the exec fifo after start returns but
+ // opens the fifo after start returns.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // ConfigInvalid - config is invalid,
+ // ContainerPaused - Container is paused,
+ // SystemError - System error.
+ Run(process *Process) (err error)
+
+ // Destroys the container, if its in a valid state, after killing any
+ // remaining running processes.
+ //
+ // Any event registrations are removed before the container is destroyed.
+ // No error is returned if the container is already destroyed.
+ //
+ // Running containers must first be stopped using Signal(..).
+ // Paused containers must first be resumed using Resume(..).
+ //
+ // errors:
+ // ContainerNotStopped - Container is still running,
+ // ContainerPaused - Container is paused,
+ // SystemError - System error.
+ Destroy() error
+
+ // Signal sends the provided signal code to the container's initial process.
+ //
+ // If all is specified the signal is sent to all processes in the container
+ // including the initial process.
+ //
+ // errors:
+ // SystemError - System error.
+ Signal(s os.Signal, all bool) error
+
+ // Exec signals the container to exec the users process at the end of the init.
+ //
+ // errors:
+ // SystemError - System error.
+ Exec() error
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/container_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/container_linux.go
new file mode 100644
index 000000000..d7e7516e5
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/container_linux.go
@@ -0,0 +1,1654 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "bytes"
+ "encoding/json"
+ "fmt"
+ "io"
+ "io/ioutil"
+ "net"
+ "os"
+ "os/exec"
+ "path/filepath"
+ "reflect"
+ "strings"
+ "sync"
+ "syscall" // only for SysProcAttr and Signal
+ "time"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/criurpc"
+ "github.com/opencontainers/runc/libcontainer/system"
+ "github.com/opencontainers/runc/libcontainer/utils"
+
+ "github.com/golang/protobuf/proto"
+ "github.com/sirupsen/logrus"
+ "github.com/syndtr/gocapability/capability"
+ "github.com/vishvananda/netlink/nl"
+ "golang.org/x/sys/unix"
+)
+
+const stdioFdCount = 3
+
+type linuxContainer struct {
+ id string
+ root string
+ config *configs.Config
+ cgroupManager cgroups.Manager
+ initArgs []string
+ initProcess parentProcess
+ initProcessStartTime uint64
+ criuPath string
+ m sync.Mutex
+ criuVersion int
+ state containerState
+ created time.Time
+}
+
+// State represents a running container's state
+type State struct {
+ BaseState
+
+ // Platform specific fields below here
+
+ // Specifies if the container was started under the rootless mode.
+ Rootless bool `json:"rootless"`
+
+ // Path to all the cgroups setup for a container. Key is cgroup subsystem name
+ // with the value as the path.
+ CgroupPaths map[string]string `json:"cgroup_paths"`
+
+ // NamespacePaths are filepaths to the container's namespaces. Key is the namespace type
+ // with the value as the path.
+ NamespacePaths map[configs.NamespaceType]string `json:"namespace_paths"`
+
+ // Container's standard descriptors (std{in,out,err}), needed for checkpoint and restore
+ ExternalDescriptors []string `json:"external_descriptors,omitempty"`
+}
+
+// Container is a libcontainer container object.
+//
+// Each container is thread-safe within the same process. Since a container can
+// be destroyed by a separate process, any function may return that the container
+// was not found.
+type Container interface {
+ BaseContainer
+
+ // Methods below here are platform specific
+
+ // Checkpoint checkpoints the running container's state to disk using the criu(8) utility.
+ //
+ // errors:
+ // Systemerror - System error.
+ Checkpoint(criuOpts *CriuOpts) error
+
+ // Restore restores the checkpointed container to a running state using the criu(8) utility.
+ //
+ // errors:
+ // Systemerror - System error.
+ Restore(process *Process, criuOpts *CriuOpts) error
+
+ // If the Container state is RUNNING or CREATED, sets the Container state to PAUSING and pauses
+ // the execution of any user processes. Asynchronously, when the container finished being paused the
+ // state is changed to PAUSED.
+ // If the Container state is PAUSED, do nothing.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // ContainerNotRunning - Container not running or created,
+ // Systemerror - System error.
+ Pause() error
+
+ // If the Container state is PAUSED, resumes the execution of any user processes in the
+ // Container before setting the Container state to RUNNING.
+ // If the Container state is RUNNING, do nothing.
+ //
+ // errors:
+ // ContainerNotExists - Container no longer exists,
+ // ContainerNotPaused - Container is not paused,
+ // Systemerror - System error.
+ Resume() error
+
+ // NotifyOOM returns a read-only channel signaling when the container receives an OOM notification.
+ //
+ // errors:
+ // Systemerror - System error.
+ NotifyOOM() (<-chan struct{}, error)
+
+ // NotifyMemoryPressure returns a read-only channel signaling when the container reaches a given pressure level
+ //
+ // errors:
+ // Systemerror - System error.
+ NotifyMemoryPressure(level PressureLevel) (<-chan struct{}, error)
+}
+
+// ID returns the container's unique ID
+func (c *linuxContainer) ID() string {
+ return c.id
+}
+
+// Config returns the container's configuration
+func (c *linuxContainer) Config() configs.Config {
+ return *c.config
+}
+
+func (c *linuxContainer) Status() (Status, error) {
+ c.m.Lock()
+ defer c.m.Unlock()
+ return c.currentStatus()
+}
+
+func (c *linuxContainer) State() (*State, error) {
+ c.m.Lock()
+ defer c.m.Unlock()
+ return c.currentState()
+}
+
+func (c *linuxContainer) Processes() ([]int, error) {
+ pids, err := c.cgroupManager.GetAllPids()
+ if err != nil {
+ return nil, newSystemErrorWithCause(err, "getting all container pids from cgroups")
+ }
+ return pids, nil
+}
+
+func (c *linuxContainer) Stats() (*Stats, error) {
+ var (
+ err error
+ stats = &Stats{}
+ )
+ if stats.CgroupStats, err = c.cgroupManager.GetStats(); err != nil {
+ return stats, newSystemErrorWithCause(err, "getting container stats from cgroups")
+ }
+ for _, iface := range c.config.Networks {
+ switch iface.Type {
+ case "veth":
+ istats, err := getNetworkInterfaceStats(iface.HostInterfaceName)
+ if err != nil {
+ return stats, newSystemErrorWithCausef(err, "getting network stats for interface %q", iface.HostInterfaceName)
+ }
+ stats.Interfaces = append(stats.Interfaces, istats)
+ }
+ }
+ return stats, nil
+}
+
+func (c *linuxContainer) Set(config configs.Config) error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ status, err := c.currentStatus()
+ if err != nil {
+ return err
+ }
+ if status == Stopped {
+ return newGenericError(fmt.Errorf("container not running"), ContainerNotRunning)
+ }
+ c.config = &config
+ return c.cgroupManager.Set(c.config)
+}
+
+func (c *linuxContainer) Start(process *Process) error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ status, err := c.currentStatus()
+ if err != nil {
+ return err
+ }
+ if status == Stopped {
+ if err := c.createExecFifo(); err != nil {
+ return err
+ }
+ }
+ if err := c.start(process, status == Stopped); err != nil {
+ if status == Stopped {
+ c.deleteExecFifo()
+ }
+ return err
+ }
+ return nil
+}
+
+func (c *linuxContainer) Run(process *Process) error {
+ c.m.Lock()
+ status, err := c.currentStatus()
+ if err != nil {
+ c.m.Unlock()
+ return err
+ }
+ c.m.Unlock()
+ if err := c.Start(process); err != nil {
+ return err
+ }
+ if status == Stopped {
+ return c.exec()
+ }
+ return nil
+}
+
+func (c *linuxContainer) Exec() error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ return c.exec()
+}
+
+func (c *linuxContainer) exec() error {
+ path := filepath.Join(c.root, execFifoFilename)
+ f, err := os.OpenFile(path, os.O_RDONLY, 0)
+ if err != nil {
+ return newSystemErrorWithCause(err, "open exec fifo for reading")
+ }
+ defer f.Close()
+ data, err := ioutil.ReadAll(f)
+ if err != nil {
+ return err
+ }
+ if len(data) > 0 {
+ os.Remove(path)
+ return nil
+ }
+ return fmt.Errorf("cannot start an already running container")
+}
+
+func (c *linuxContainer) start(process *Process, isInit bool) error {
+ parent, err := c.newParentProcess(process, isInit)
+ if err != nil {
+ return newSystemErrorWithCause(err, "creating new parent process")
+ }
+ if err := parent.start(); err != nil {
+ // terminate the process to ensure that it properly is reaped.
+ if err := parent.terminate(); err != nil {
+ logrus.Warn(err)
+ }
+ return newSystemErrorWithCause(err, "starting container process")
+ }
+ // generate a timestamp indicating when the container was started
+ c.created = time.Now().UTC()
+ if isInit {
+ c.state = &createdState{
+ c: c,
+ }
+ state, err := c.updateState(parent)
+ if err != nil {
+ return err
+ }
+ c.initProcessStartTime = state.InitProcessStartTime
+
+ if c.config.Hooks != nil {
+ s := configs.HookState{
+ Version: c.config.Version,
+ ID: c.id,
+ Pid: parent.pid(),
+ Bundle: utils.SearchLabels(c.config.Labels, "bundle"),
+ }
+ for i, hook := range c.config.Hooks.Poststart {
+ if err := hook.Run(s); err != nil {
+ if err := parent.terminate(); err != nil {
+ logrus.Warn(err)
+ }
+ return newSystemErrorWithCausef(err, "running poststart hook %d", i)
+ }
+ }
+ }
+ } else {
+ c.state = &runningState{
+ c: c,
+ }
+ }
+ return nil
+}
+
+func (c *linuxContainer) Signal(s os.Signal, all bool) error {
+ if all {
+ return signalAllProcesses(c.cgroupManager, s)
+ }
+ if err := c.initProcess.signal(s); err != nil {
+ return newSystemErrorWithCause(err, "signaling init process")
+ }
+ return nil
+}
+
+func (c *linuxContainer) createExecFifo() error {
+ rootuid, err := c.Config().HostRootUID()
+ if err != nil {
+ return err
+ }
+ rootgid, err := c.Config().HostRootGID()
+ if err != nil {
+ return err
+ }
+
+ fifoName := filepath.Join(c.root, execFifoFilename)
+ if _, err := os.Stat(fifoName); err == nil {
+ return fmt.Errorf("exec fifo %s already exists", fifoName)
+ }
+ oldMask := unix.Umask(0000)
+ if err := unix.Mkfifo(fifoName, 0622); err != nil {
+ unix.Umask(oldMask)
+ return err
+ }
+ unix.Umask(oldMask)
+ if err := os.Chown(fifoName, rootuid, rootgid); err != nil {
+ return err
+ }
+ return nil
+}
+
+func (c *linuxContainer) deleteExecFifo() {
+ fifoName := filepath.Join(c.root, execFifoFilename)
+ os.Remove(fifoName)
+}
+
+func (c *linuxContainer) newParentProcess(p *Process, doInit bool) (parentProcess, error) {
+ parentPipe, childPipe, err := utils.NewSockPair("init")
+ if err != nil {
+ return nil, newSystemErrorWithCause(err, "creating new init pipe")
+ }
+ cmd, err := c.commandTemplate(p, childPipe)
+ if err != nil {
+ return nil, newSystemErrorWithCause(err, "creating new command template")
+ }
+ if !doInit {
+ return c.newSetnsProcess(p, cmd, parentPipe, childPipe)
+ }
+
+ // We only set up rootDir if we're not doing a `runc exec`. The reason for
+ // this is to avoid cases where a racing, unprivileged process inside the
+ // container can get access to the statedir file descriptor (which would
+ // allow for container rootfs escape).
+ rootDir, err := os.Open(c.root)
+ if err != nil {
+ return nil, err
+ }
+ cmd.ExtraFiles = append(cmd.ExtraFiles, rootDir)
+ cmd.Env = append(cmd.Env,
+ fmt.Sprintf("_LIBCONTAINER_STATEDIR=%d", stdioFdCount+len(cmd.ExtraFiles)-1))
+ return c.newInitProcess(p, cmd, parentPipe, childPipe, rootDir)
+}
+
+func (c *linuxContainer) commandTemplate(p *Process, childPipe *os.File) (*exec.Cmd, error) {
+ cmd := exec.Command(c.initArgs[0], c.initArgs[1:]...)
+ cmd.Stdin = p.Stdin
+ cmd.Stdout = p.Stdout
+ cmd.Stderr = p.Stderr
+ cmd.Dir = c.config.Rootfs
+ if cmd.SysProcAttr == nil {
+ cmd.SysProcAttr = &syscall.SysProcAttr{}
+ }
+ cmd.ExtraFiles = append(cmd.ExtraFiles, p.ExtraFiles...)
+ if p.ConsoleSocket != nil {
+ cmd.ExtraFiles = append(cmd.ExtraFiles, p.ConsoleSocket)
+ cmd.Env = append(cmd.Env,
+ fmt.Sprintf("_LIBCONTAINER_CONSOLE=%d", stdioFdCount+len(cmd.ExtraFiles)-1),
+ )
+ }
+ cmd.ExtraFiles = append(cmd.ExtraFiles, childPipe)
+ cmd.Env = append(cmd.Env,
+ fmt.Sprintf("_LIBCONTAINER_INITPIPE=%d", stdioFdCount+len(cmd.ExtraFiles)-1),
+ )
+ // NOTE: when running a container with no PID namespace and the parent process spawning the container is
+ // PID1 the pdeathsig is being delivered to the container's init process by the kernel for some reason
+ // even with the parent still running.
+ if c.config.ParentDeathSignal > 0 {
+ cmd.SysProcAttr.Pdeathsig = syscall.Signal(c.config.ParentDeathSignal)
+ }
+ return cmd, nil
+}
+
+func (c *linuxContainer) newInitProcess(p *Process, cmd *exec.Cmd, parentPipe, childPipe, rootDir *os.File) (*initProcess, error) {
+ cmd.Env = append(cmd.Env, "_LIBCONTAINER_INITTYPE="+string(initStandard))
+ nsMaps := make(map[configs.NamespaceType]string)
+ for _, ns := range c.config.Namespaces {
+ if ns.Path != "" {
+ nsMaps[ns.Type] = ns.Path
+ }
+ }
+ _, sharePidns := nsMaps[configs.NEWPID]
+ data, err := c.bootstrapData(c.config.Namespaces.CloneFlags(), nsMaps)
+ if err != nil {
+ return nil, err
+ }
+ return &initProcess{
+ cmd: cmd,
+ childPipe: childPipe,
+ parentPipe: parentPipe,
+ manager: c.cgroupManager,
+ config: c.newInitConfig(p),
+ container: c,
+ process: p,
+ bootstrapData: data,
+ sharePidns: sharePidns,
+ rootDir: rootDir,
+ }, nil
+}
+
+func (c *linuxContainer) newSetnsProcess(p *Process, cmd *exec.Cmd, parentPipe, childPipe *os.File) (*setnsProcess, error) {
+ cmd.Env = append(cmd.Env, "_LIBCONTAINER_INITTYPE="+string(initSetns))
+ state, err := c.currentState()
+ if err != nil {
+ return nil, newSystemErrorWithCause(err, "getting container's current state")
+ }
+ // for setns process, we don't have to set cloneflags as the process namespaces
+ // will only be set via setns syscall
+ data, err := c.bootstrapData(0, state.NamespacePaths)
+ if err != nil {
+ return nil, err
+ }
+ return &setnsProcess{
+ cmd: cmd,
+ cgroupPaths: c.cgroupManager.GetPaths(),
+ childPipe: childPipe,
+ parentPipe: parentPipe,
+ config: c.newInitConfig(p),
+ process: p,
+ bootstrapData: data,
+ }, nil
+}
+
+func (c *linuxContainer) newInitConfig(process *Process) *initConfig {
+ cfg := &initConfig{
+ Config: c.config,
+ Args: process.Args,
+ Env: process.Env,
+ User: process.User,
+ AdditionalGroups: process.AdditionalGroups,
+ Cwd: process.Cwd,
+ Capabilities: process.Capabilities,
+ PassedFilesCount: len(process.ExtraFiles),
+ ContainerId: c.ID(),
+ NoNewPrivileges: c.config.NoNewPrivileges,
+ Rootless: c.config.Rootless,
+ AppArmorProfile: c.config.AppArmorProfile,
+ ProcessLabel: c.config.ProcessLabel,
+ Rlimits: c.config.Rlimits,
+ }
+ if process.NoNewPrivileges != nil {
+ cfg.NoNewPrivileges = *process.NoNewPrivileges
+ }
+ if process.AppArmorProfile != "" {
+ cfg.AppArmorProfile = process.AppArmorProfile
+ }
+ if process.Label != "" {
+ cfg.ProcessLabel = process.Label
+ }
+ if len(process.Rlimits) > 0 {
+ cfg.Rlimits = process.Rlimits
+ }
+ cfg.CreateConsole = process.ConsoleSocket != nil
+ return cfg
+}
+
+func (c *linuxContainer) Destroy() error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ return c.state.destroy()
+}
+
+func (c *linuxContainer) Pause() error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ status, err := c.currentStatus()
+ if err != nil {
+ return err
+ }
+ switch status {
+ case Running, Created:
+ if err := c.cgroupManager.Freeze(configs.Frozen); err != nil {
+ return err
+ }
+ return c.state.transition(&pausedState{
+ c: c,
+ })
+ }
+ return newGenericError(fmt.Errorf("container not running or created: %s", status), ContainerNotRunning)
+}
+
+func (c *linuxContainer) Resume() error {
+ c.m.Lock()
+ defer c.m.Unlock()
+ status, err := c.currentStatus()
+ if err != nil {
+ return err
+ }
+ if status != Paused {
+ return newGenericError(fmt.Errorf("container not paused"), ContainerNotPaused)
+ }
+ if err := c.cgroupManager.Freeze(configs.Thawed); err != nil {
+ return err
+ }
+ return c.state.transition(&runningState{
+ c: c,
+ })
+}
+
+func (c *linuxContainer) NotifyOOM() (<-chan struct{}, error) {
+ // XXX(cyphar): This requires cgroups.
+ if c.config.Rootless {
+ return nil, fmt.Errorf("cannot get OOM notifications from rootless container")
+ }
+ return notifyOnOOM(c.cgroupManager.GetPaths())
+}
+
+func (c *linuxContainer) NotifyMemoryPressure(level PressureLevel) (<-chan struct{}, error) {
+ // XXX(cyphar): This requires cgroups.
+ if c.config.Rootless {
+ return nil, fmt.Errorf("cannot get memory pressure notifications from rootless container")
+ }
+ return notifyMemoryPressure(c.cgroupManager.GetPaths(), level)
+}
+
+var criuFeatures *criurpc.CriuFeatures
+
+func (c *linuxContainer) checkCriuFeatures(criuOpts *CriuOpts, rpcOpts *criurpc.CriuOpts, criuFeat *criurpc.CriuFeatures) error {
+
+ var t criurpc.CriuReqType
+ t = criurpc.CriuReqType_FEATURE_CHECK
+
+ // criu 1.8 => 10800
+ if err := c.checkCriuVersion(10800); err != nil {
+ // Feature checking was introduced with CRIU 1.8.
+ // Ignore the feature check if an older CRIU version is used
+ // and just act as before.
+ // As all automated PR testing is done using CRIU 1.7 this
+ // code will not be tested by automated PR testing.
+ return nil
+ }
+
+ // make sure the features we are looking for are really not from
+ // some previous check
+ criuFeatures = nil
+
+ req := &criurpc.CriuReq{
+ Type: &t,
+ // Theoretically this should not be necessary but CRIU
+ // segfaults if Opts is empty.
+ // Fixed in CRIU 2.12
+ Opts: rpcOpts,
+ Features: criuFeat,
+ }
+
+ err := c.criuSwrk(nil, req, criuOpts, false)
+ if err != nil {
+ logrus.Debugf("%s", err)
+ return fmt.Errorf("CRIU feature check failed")
+ }
+
+ logrus.Debugf("Feature check says: %s", criuFeatures)
+ missingFeatures := false
+
+ if *criuFeat.MemTrack && !*criuFeatures.MemTrack {
+ missingFeatures = true
+ logrus.Debugf("CRIU does not support MemTrack")
+ }
+
+ if missingFeatures {
+ return fmt.Errorf("CRIU is missing features")
+ }
+
+ return nil
+}
+
+func parseCriuVersion(path string) (int, error) {
+ var x, y, z int
+
+ out, err := exec.Command(path, "-V").Output()
+ if err != nil {
+ return 0, fmt.Errorf("Unable to execute CRIU command: %s", path)
+ }
+
+ x = 0
+ y = 0
+ z = 0
+ if ep := strings.Index(string(out), "-"); ep >= 0 {
+ // criu Git version format
+ var version string
+ if sp := strings.Index(string(out), "GitID"); sp > 0 {
+ version = string(out)[sp:ep]
+ } else {
+ return 0, fmt.Errorf("Unable to parse the CRIU version: %s", path)
+ }
+
+ n, err := fmt.Sscanf(string(version), "GitID: v%d.%d.%d", &x, &y, &z) // 1.5.2
+ if err != nil {
+ n, err = fmt.Sscanf(string(version), "GitID: v%d.%d", &x, &y) // 1.6
+ y++
+ } else {
+ z++
+ }
+ if n < 2 || err != nil {
+ return 0, fmt.Errorf("Unable to parse the CRIU version: %s %d %s", version, n, err)
+ }
+ } else {
+ // criu release version format
+ n, err := fmt.Sscanf(string(out), "Version: %d.%d.%d\n", &x, &y, &z) // 1.5.2
+ if err != nil {
+ n, err = fmt.Sscanf(string(out), "Version: %d.%d\n", &x, &y) // 1.6
+ }
+ if n < 2 || err != nil {
+ return 0, fmt.Errorf("Unable to parse the CRIU version: %s %d %s", out, n, err)
+ }
+ }
+
+ return x*10000 + y*100 + z, nil
+}
+
+func compareCriuVersion(criuVersion int, minVersion int) error {
+ // simple function to perform the actual version compare
+ if criuVersion < minVersion {
+ return fmt.Errorf("CRIU version %d must be %d or higher", criuVersion, minVersion)
+ }
+
+ return nil
+}
+
+// This is used to store the result of criu version RPC
+var criuVersionRPC *criurpc.CriuVersion
+
+// checkCriuVersion checks Criu version greater than or equal to minVersion
+func (c *linuxContainer) checkCriuVersion(minVersion int) error {
+
+ // If the version of criu has already been determined there is no need
+ // to ask criu for the version again. Use the value from c.criuVersion.
+ if c.criuVersion != 0 {
+ return compareCriuVersion(c.criuVersion, minVersion)
+ }
+
+ // First try if this version of CRIU support the version RPC.
+ // The CRIU version RPC was introduced with CRIU 3.0.
+
+ // First, reset the variable for the RPC answer to nil
+ criuVersionRPC = nil
+
+ var t criurpc.CriuReqType
+ t = criurpc.CriuReqType_VERSION
+ req := &criurpc.CriuReq{
+ Type: &t,
+ }
+
+ err := c.criuSwrk(nil, req, nil, false)
+ if err != nil {
+ return fmt.Errorf("CRIU version check failed: %s", err)
+ }
+
+ if criuVersionRPC != nil {
+ logrus.Debugf("CRIU version: %s", criuVersionRPC)
+ // major and minor are always set
+ c.criuVersion = int(*criuVersionRPC.Major) * 10000
+ c.criuVersion += int(*criuVersionRPC.Minor) * 100
+ if criuVersionRPC.Sublevel != nil {
+ c.criuVersion += int(*criuVersionRPC.Sublevel)
+ }
+ if criuVersionRPC.Gitid != nil {
+ // runc's convention is that a CRIU git release is
+ // always the same as increasing the minor by 1
+ c.criuVersion -= (c.criuVersion % 100)
+ c.criuVersion += 100
+ }
+ return compareCriuVersion(c.criuVersion, minVersion)
+ }
+
+ // This is CRIU without the version RPC and therefore
+ // older than 3.0. Parsing the output is required.
+
+ // This can be remove once runc does not work with criu older than 3.0
+
+ c.criuVersion, err = parseCriuVersion(c.criuPath)
+ if err != nil {
+ return err
+ }
+
+ return compareCriuVersion(c.criuVersion, minVersion)
+}
+
+const descriptorsFilename = "descriptors.json"
+
+func (c *linuxContainer) addCriuDumpMount(req *criurpc.CriuReq, m *configs.Mount) {
+ mountDest := m.Destination
+ if strings.HasPrefix(mountDest, c.config.Rootfs) {
+ mountDest = mountDest[len(c.config.Rootfs):]
+ }
+
+ extMnt := &criurpc.ExtMountMap{
+ Key: proto.String(mountDest),
+ Val: proto.String(mountDest),
+ }
+ req.Opts.ExtMnt = append(req.Opts.ExtMnt, extMnt)
+}
+
+func (c *linuxContainer) addMaskPaths(req *criurpc.CriuReq) error {
+ for _, path := range c.config.MaskPaths {
+ fi, err := os.Stat(fmt.Sprintf("/proc/%d/root/%s", c.initProcess.pid(), path))
+ if err != nil {
+ if os.IsNotExist(err) {
+ continue
+ }
+ return err
+ }
+ if fi.IsDir() {
+ continue
+ }
+
+ extMnt := &criurpc.ExtMountMap{
+ Key: proto.String(path),
+ Val: proto.String("/dev/null"),
+ }
+ req.Opts.ExtMnt = append(req.Opts.ExtMnt, extMnt)
+ }
+
+ return nil
+}
+
+func (c *linuxContainer) Checkpoint(criuOpts *CriuOpts) error {
+ c.m.Lock()
+ defer c.m.Unlock()
+
+ // TODO(avagin): Figure out how to make this work nicely. CRIU 2.0 has
+ // support for doing unprivileged dumps, but the setup of
+ // rootless containers might make this complicated.
+ if c.config.Rootless {
+ return fmt.Errorf("cannot checkpoint a rootless container")
+ }
+
+ // criu 1.5.2 => 10502
+ if err := c.checkCriuVersion(10502); err != nil {
+ return err
+ }
+
+ if criuOpts.ImagesDirectory == "" {
+ return fmt.Errorf("invalid directory to save checkpoint")
+ }
+
+ // Since a container can be C/R'ed multiple times,
+ // the checkpoint directory may already exist.
+ if err := os.Mkdir(criuOpts.ImagesDirectory, 0755); err != nil && !os.IsExist(err) {
+ return err
+ }
+
+ if criuOpts.WorkDirectory == "" {
+ criuOpts.WorkDirectory = filepath.Join(c.root, "criu.work")
+ }
+
+ if err := os.Mkdir(criuOpts.WorkDirectory, 0755); err != nil && !os.IsExist(err) {
+ return err
+ }
+
+ workDir, err := os.Open(criuOpts.WorkDirectory)
+ if err != nil {
+ return err
+ }
+ defer workDir.Close()
+
+ imageDir, err := os.Open(criuOpts.ImagesDirectory)
+ if err != nil {
+ return err
+ }
+ defer imageDir.Close()
+
+ rpcOpts := criurpc.CriuOpts{
+ ImagesDirFd: proto.Int32(int32(imageDir.Fd())),
+ WorkDirFd: proto.Int32(int32(workDir.Fd())),
+ LogLevel: proto.Int32(4),
+ LogFile: proto.String("dump.log"),
+ Root: proto.String(c.config.Rootfs),
+ ManageCgroups: proto.Bool(true),
+ NotifyScripts: proto.Bool(true),
+ Pid: proto.Int32(int32(c.initProcess.pid())),
+ ShellJob: proto.Bool(criuOpts.ShellJob),
+ LeaveRunning: proto.Bool(criuOpts.LeaveRunning),
+ TcpEstablished: proto.Bool(criuOpts.TcpEstablished),
+ ExtUnixSk: proto.Bool(criuOpts.ExternalUnixConnections),
+ FileLocks: proto.Bool(criuOpts.FileLocks),
+ EmptyNs: proto.Uint32(criuOpts.EmptyNs),
+ OrphanPtsMaster: proto.Bool(true),
+ }
+
+ fcg := c.cgroupManager.GetPaths()["freezer"]
+ if fcg != "" {
+ rpcOpts.FreezeCgroup = proto.String(fcg)
+ }
+
+ // append optional criu opts, e.g., page-server and port
+ if criuOpts.PageServer.Address != "" && criuOpts.PageServer.Port != 0 {
+ rpcOpts.Ps = &criurpc.CriuPageServerInfo{
+ Address: proto.String(criuOpts.PageServer.Address),
+ Port: proto.Int32(criuOpts.PageServer.Port),
+ }
+ }
+
+ //pre-dump may need parentImage param to complete iterative migration
+ if criuOpts.ParentImage != "" {
+ rpcOpts.ParentImg = proto.String(criuOpts.ParentImage)
+ rpcOpts.TrackMem = proto.Bool(true)
+ }
+
+ // append optional manage cgroups mode
+ if criuOpts.ManageCgroupsMode != 0 {
+ // criu 1.7 => 10700
+ if err := c.checkCriuVersion(10700); err != nil {
+ return err
+ }
+ mode := criurpc.CriuCgMode(criuOpts.ManageCgroupsMode)
+ rpcOpts.ManageCgroupsMode = &mode
+ }
+
+ var t criurpc.CriuReqType
+ if criuOpts.PreDump {
+ feat := criurpc.CriuFeatures{
+ MemTrack: proto.Bool(true),
+ }
+
+ if err := c.checkCriuFeatures(criuOpts, &rpcOpts, &feat); err != nil {
+ return err
+ }
+
+ t = criurpc.CriuReqType_PRE_DUMP
+ } else {
+ t = criurpc.CriuReqType_DUMP
+ }
+ req := &criurpc.CriuReq{
+ Type: &t,
+ Opts: &rpcOpts,
+ }
+
+ //no need to dump these information in pre-dump
+ if !criuOpts.PreDump {
+ for _, m := range c.config.Mounts {
+ switch m.Device {
+ case "bind":
+ c.addCriuDumpMount(req, m)
+ case "cgroup":
+ binds, err := getCgroupMounts(m)
+ if err != nil {
+ return err
+ }
+ for _, b := range binds {
+ c.addCriuDumpMount(req, b)
+ }
+ }
+ }
+
+ if err := c.addMaskPaths(req); err != nil {
+ return err
+ }
+
+ for _, node := range c.config.Devices {
+ m := &configs.Mount{Destination: node.Path, Source: node.Path}
+ c.addCriuDumpMount(req, m)
+ }
+
+ // Write the FD info to a file in the image directory
+ fdsJSON, err := json.Marshal(c.initProcess.externalDescriptors())
+ if err != nil {
+ return err
+ }
+
+ err = ioutil.WriteFile(filepath.Join(criuOpts.ImagesDirectory, descriptorsFilename), fdsJSON, 0655)
+ if err != nil {
+ return err
+ }
+ }
+
+ err = c.criuSwrk(nil, req, criuOpts, false)
+ if err != nil {
+ return err
+ }
+ return nil
+}
+
+func (c *linuxContainer) addCriuRestoreMount(req *criurpc.CriuReq, m *configs.Mount) {
+ mountDest := m.Destination
+ if strings.HasPrefix(mountDest, c.config.Rootfs) {
+ mountDest = mountDest[len(c.config.Rootfs):]
+ }
+
+ extMnt := &criurpc.ExtMountMap{
+ Key: proto.String(mountDest),
+ Val: proto.String(m.Source),
+ }
+ req.Opts.ExtMnt = append(req.Opts.ExtMnt, extMnt)
+}
+
+func (c *linuxContainer) restoreNetwork(req *criurpc.CriuReq, criuOpts *CriuOpts) {
+ for _, iface := range c.config.Networks {
+ switch iface.Type {
+ case "veth":
+ veth := new(criurpc.CriuVethPair)
+ veth.IfOut = proto.String(iface.HostInterfaceName)
+ veth.IfIn = proto.String(iface.Name)
+ req.Opts.Veths = append(req.Opts.Veths, veth)
+ case "loopback":
+ // Do nothing
+ }
+ }
+ for _, i := range criuOpts.VethPairs {
+ veth := new(criurpc.CriuVethPair)
+ veth.IfOut = proto.String(i.HostInterfaceName)
+ veth.IfIn = proto.String(i.ContainerInterfaceName)
+ req.Opts.Veths = append(req.Opts.Veths, veth)
+ }
+}
+
+func (c *linuxContainer) Restore(process *Process, criuOpts *CriuOpts) error {
+ c.m.Lock()
+ defer c.m.Unlock()
+
+ // TODO(avagin): Figure out how to make this work nicely. CRIU doesn't have
+ // support for unprivileged restore at the moment.
+ if c.config.Rootless {
+ return fmt.Errorf("cannot restore a rootless container")
+ }
+
+ // criu 1.5.2 => 10502
+ if err := c.checkCriuVersion(10502); err != nil {
+ return err
+ }
+ if criuOpts.WorkDirectory == "" {
+ criuOpts.WorkDirectory = filepath.Join(c.root, "criu.work")
+ }
+ // Since a container can be C/R'ed multiple times,
+ // the work directory may already exist.
+ if err := os.Mkdir(criuOpts.WorkDirectory, 0655); err != nil && !os.IsExist(err) {
+ return err
+ }
+ workDir, err := os.Open(criuOpts.WorkDirectory)
+ if err != nil {
+ return err
+ }
+ defer workDir.Close()
+ if criuOpts.ImagesDirectory == "" {
+ return fmt.Errorf("invalid directory to restore checkpoint")
+ }
+ imageDir, err := os.Open(criuOpts.ImagesDirectory)
+ if err != nil {
+ return err
+ }
+ defer imageDir.Close()
+ // CRIU has a few requirements for a root directory:
+ // * it must be a mount point
+ // * its parent must not be overmounted
+ // c.config.Rootfs is bind-mounted to a temporary directory
+ // to satisfy these requirements.
+ root := filepath.Join(c.root, "criu-root")
+ if err := os.Mkdir(root, 0755); err != nil {
+ return err
+ }
+ defer os.Remove(root)
+ root, err = filepath.EvalSymlinks(root)
+ if err != nil {
+ return err
+ }
+ err = unix.Mount(c.config.Rootfs, root, "", unix.MS_BIND|unix.MS_REC, "")
+ if err != nil {
+ return err
+ }
+ defer unix.Unmount(root, unix.MNT_DETACH)
+ t := criurpc.CriuReqType_RESTORE
+ req := &criurpc.CriuReq{
+ Type: &t,
+ Opts: &criurpc.CriuOpts{
+ ImagesDirFd: proto.Int32(int32(imageDir.Fd())),
+ WorkDirFd: proto.Int32(int32(workDir.Fd())),
+ EvasiveDevices: proto.Bool(true),
+ LogLevel: proto.Int32(4),
+ LogFile: proto.String("restore.log"),
+ RstSibling: proto.Bool(true),
+ Root: proto.String(root),
+ ManageCgroups: proto.Bool(true),
+ NotifyScripts: proto.Bool(true),
+ ShellJob: proto.Bool(criuOpts.ShellJob),
+ ExtUnixSk: proto.Bool(criuOpts.ExternalUnixConnections),
+ TcpEstablished: proto.Bool(criuOpts.TcpEstablished),
+ FileLocks: proto.Bool(criuOpts.FileLocks),
+ EmptyNs: proto.Uint32(criuOpts.EmptyNs),
+ OrphanPtsMaster: proto.Bool(true),
+ },
+ }
+
+ for _, m := range c.config.Mounts {
+ switch m.Device {
+ case "bind":
+ c.addCriuRestoreMount(req, m)
+ case "cgroup":
+ binds, err := getCgroupMounts(m)
+ if err != nil {
+ return err
+ }
+ for _, b := range binds {
+ c.addCriuRestoreMount(req, b)
+ }
+ }
+ }
+
+ if len(c.config.MaskPaths) > 0 {
+ m := &configs.Mount{Destination: "/dev/null", Source: "/dev/null"}
+ c.addCriuRestoreMount(req, m)
+ }
+
+ for _, node := range c.config.Devices {
+ m := &configs.Mount{Destination: node.Path, Source: node.Path}
+ c.addCriuRestoreMount(req, m)
+ }
+
+ if criuOpts.EmptyNs&unix.CLONE_NEWNET == 0 {
+ c.restoreNetwork(req, criuOpts)
+ }
+
+ // append optional manage cgroups mode
+ if criuOpts.ManageCgroupsMode != 0 {
+ // criu 1.7 => 10700
+ if err := c.checkCriuVersion(10700); err != nil {
+ return err
+ }
+ mode := criurpc.CriuCgMode(criuOpts.ManageCgroupsMode)
+ req.Opts.ManageCgroupsMode = &mode
+ }
+
+ var (
+ fds []string
+ fdJSON []byte
+ )
+ if fdJSON, err = ioutil.ReadFile(filepath.Join(criuOpts.ImagesDirectory, descriptorsFilename)); err != nil {
+ return err
+ }
+
+ if err := json.Unmarshal(fdJSON, &fds); err != nil {
+ return err
+ }
+ for i := range fds {
+ if s := fds[i]; strings.Contains(s, "pipe:") {
+ inheritFd := new(criurpc.InheritFd)
+ inheritFd.Key = proto.String(s)
+ inheritFd.Fd = proto.Int32(int32(i))
+ req.Opts.InheritFd = append(req.Opts.InheritFd, inheritFd)
+ }
+ }
+ return c.criuSwrk(process, req, criuOpts, true)
+}
+
+func (c *linuxContainer) criuApplyCgroups(pid int, req *criurpc.CriuReq) error {
+ // XXX: Do we need to deal with this case? AFAIK criu still requires root.
+ if err := c.cgroupManager.Apply(pid); err != nil {
+ return err
+ }
+
+ if err := c.cgroupManager.Set(c.config); err != nil {
+ return newSystemError(err)
+ }
+
+ path := fmt.Sprintf("/proc/%d/cgroup", pid)
+ cgroupsPaths, err := cgroups.ParseCgroupFile(path)
+ if err != nil {
+ return err
+ }
+
+ for c, p := range cgroupsPaths {
+ cgroupRoot := &criurpc.CgroupRoot{
+ Ctrl: proto.String(c),
+ Path: proto.String(p),
+ }
+ req.Opts.CgRoot = append(req.Opts.CgRoot, cgroupRoot)
+ }
+
+ return nil
+}
+
+func (c *linuxContainer) criuSwrk(process *Process, req *criurpc.CriuReq, opts *CriuOpts, applyCgroups bool) error {
+ fds, err := unix.Socketpair(unix.AF_LOCAL, unix.SOCK_SEQPACKET|unix.SOCK_CLOEXEC, 0)
+ if err != nil {
+ return err
+ }
+
+ var logPath string
+ if opts != nil {
+ logPath = filepath.Join(opts.WorkDirectory, req.GetOpts().GetLogFile())
+ } else {
+ // For the VERSION RPC 'opts' is set to 'nil' and therefore
+ // opts.WorkDirectory does not exist. Set logPath to "".
+ logPath = ""
+ }
+ criuClient := os.NewFile(uintptr(fds[0]), "criu-transport-client")
+ criuClientFileCon, err := net.FileConn(criuClient)
+ criuClient.Close()
+ if err != nil {
+ return err
+ }
+
+ criuClientCon := criuClientFileCon.(*net.UnixConn)
+ defer criuClientCon.Close()
+
+ criuServer := os.NewFile(uintptr(fds[1]), "criu-transport-server")
+ defer criuServer.Close()
+
+ args := []string{"swrk", "3"}
+ if c.criuVersion != 0 {
+ // If the CRIU Version is still '0' then this is probably
+ // the initial CRIU run to detect the version. Skip it.
+ logrus.Debugf("Using CRIU %d at: %s", c.criuVersion, c.criuPath)
+ }
+ logrus.Debugf("Using CRIU with following args: %s", args)
+ cmd := exec.Command(c.criuPath, args...)
+ if process != nil {
+ cmd.Stdin = process.Stdin
+ cmd.Stdout = process.Stdout
+ cmd.Stderr = process.Stderr
+ }
+ cmd.ExtraFiles = append(cmd.ExtraFiles, criuServer)
+
+ if err := cmd.Start(); err != nil {
+ return err
+ }
+ criuServer.Close()
+
+ defer func() {
+ criuClientCon.Close()
+ _, err := cmd.Process.Wait()
+ if err != nil {
+ return
+ }
+ }()
+
+ if applyCgroups {
+ err := c.criuApplyCgroups(cmd.Process.Pid, req)
+ if err != nil {
+ return err
+ }
+ }
+
+ var extFds []string
+ if process != nil {
+ extFds, err = getPipeFds(cmd.Process.Pid)
+ if err != nil {
+ return err
+ }
+ }
+
+ logrus.Debugf("Using CRIU in %s mode", req.GetType().String())
+ // In the case of criurpc.CriuReqType_FEATURE_CHECK req.GetOpts()
+ // should be empty. For older CRIU versions it still will be
+ // available but empty. criurpc.CriuReqType_VERSION actually
+ // has no req.GetOpts().
+ if !(req.GetType() == criurpc.CriuReqType_FEATURE_CHECK ||
+ req.GetType() == criurpc.CriuReqType_VERSION) {
+
+ val := reflect.ValueOf(req.GetOpts())
+ v := reflect.Indirect(val)
+ for i := 0; i < v.NumField(); i++ {
+ st := v.Type()
+ name := st.Field(i).Name
+ if strings.HasPrefix(name, "XXX_") {
+ continue
+ }
+ value := val.MethodByName("Get" + name).Call([]reflect.Value{})
+ logrus.Debugf("CRIU option %s with value %v", name, value[0])
+ }
+ }
+ data, err := proto.Marshal(req)
+ if err != nil {
+ return err
+ }
+ _, err = criuClientCon.Write(data)
+ if err != nil {
+ return err
+ }
+
+ buf := make([]byte, 10*4096)
+ oob := make([]byte, 4096)
+ for true {
+ n, oobn, _, _, err := criuClientCon.ReadMsgUnix(buf, oob)
+ if err != nil {
+ return err
+ }
+ if n == 0 {
+ return fmt.Errorf("unexpected EOF")
+ }
+ if n == len(buf) {
+ return fmt.Errorf("buffer is too small")
+ }
+
+ resp := new(criurpc.CriuResp)
+ err = proto.Unmarshal(buf[:n], resp)
+ if err != nil {
+ return err
+ }
+ if !resp.GetSuccess() {
+ typeString := req.GetType().String()
+ if typeString == "VERSION" {
+ // If the VERSION RPC fails this probably means that the CRIU
+ // version is too old for this RPC. Just return 'nil'.
+ return nil
+ }
+ return fmt.Errorf("criu failed: type %s errno %d\nlog file: %s", typeString, resp.GetCrErrno(), logPath)
+ }
+
+ t := resp.GetType()
+ switch {
+ case t == criurpc.CriuReqType_VERSION:
+ logrus.Debugf("CRIU version: %s", resp)
+ criuVersionRPC = resp.GetVersion()
+ break
+ case t == criurpc.CriuReqType_FEATURE_CHECK:
+ logrus.Debugf("Feature check says: %s", resp)
+ criuFeatures = resp.GetFeatures()
+ case t == criurpc.CriuReqType_NOTIFY:
+ if err := c.criuNotifications(resp, process, opts, extFds, oob[:oobn]); err != nil {
+ return err
+ }
+ t = criurpc.CriuReqType_NOTIFY
+ req = &criurpc.CriuReq{
+ Type: &t,
+ NotifySuccess: proto.Bool(true),
+ }
+ data, err = proto.Marshal(req)
+ if err != nil {
+ return err
+ }
+ _, err = criuClientCon.Write(data)
+ if err != nil {
+ return err
+ }
+ continue
+ case t == criurpc.CriuReqType_RESTORE:
+ case t == criurpc.CriuReqType_DUMP:
+ case t == criurpc.CriuReqType_PRE_DUMP:
+ default:
+ return fmt.Errorf("unable to parse the response %s", resp.String())
+ }
+
+ break
+ }
+
+ criuClientCon.CloseWrite()
+ // cmd.Wait() waits cmd.goroutines which are used for proxying file descriptors.
+ // Here we want to wait only the CRIU process.
+ st, err := cmd.Process.Wait()
+ if err != nil {
+ return err
+ }
+
+ // In pre-dump mode CRIU is in a loop and waits for
+ // the final DUMP command.
+ // The current runc pre-dump approach, however, is
+ // start criu in PRE_DUMP once for a single pre-dump
+ // and not the whole series of pre-dump, pre-dump, ...m, dump
+ // If we got the message CriuReqType_PRE_DUMP it means
+ // CRIU was successful and we need to forcefully stop CRIU
+ if !st.Success() && *req.Type != criurpc.CriuReqType_PRE_DUMP {
+ return fmt.Errorf("criu failed: %s\nlog file: %s", st.String(), logPath)
+ }
+ return nil
+}
+
+// block any external network activity
+func lockNetwork(config *configs.Config) error {
+ for _, config := range config.Networks {
+ strategy, err := getStrategy(config.Type)
+ if err != nil {
+ return err
+ }
+
+ if err := strategy.detach(config); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func unlockNetwork(config *configs.Config) error {
+ for _, config := range config.Networks {
+ strategy, err := getStrategy(config.Type)
+ if err != nil {
+ return err
+ }
+ if err = strategy.attach(config); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (c *linuxContainer) criuNotifications(resp *criurpc.CriuResp, process *Process, opts *CriuOpts, fds []string, oob []byte) error {
+ notify := resp.GetNotify()
+ if notify == nil {
+ return fmt.Errorf("invalid response: %s", resp.String())
+ }
+ logrus.Debugf("notify: %s\n", notify.GetScript())
+ switch {
+ case notify.GetScript() == "post-dump":
+ f, err := os.Create(filepath.Join(c.root, "checkpoint"))
+ if err != nil {
+ return err
+ }
+ f.Close()
+ case notify.GetScript() == "network-unlock":
+ if err := unlockNetwork(c.config); err != nil {
+ return err
+ }
+ case notify.GetScript() == "network-lock":
+ if err := lockNetwork(c.config); err != nil {
+ return err
+ }
+ case notify.GetScript() == "setup-namespaces":
+ if c.config.Hooks != nil {
+ s := configs.HookState{
+ Version: c.config.Version,
+ ID: c.id,
+ Pid: int(notify.GetPid()),
+ Bundle: utils.SearchLabels(c.config.Labels, "bundle"),
+ }
+ for i, hook := range c.config.Hooks.Prestart {
+ if err := hook.Run(s); err != nil {
+ return newSystemErrorWithCausef(err, "running prestart hook %d", i)
+ }
+ }
+ }
+ case notify.GetScript() == "post-restore":
+ pid := notify.GetPid()
+ r, err := newRestoredProcess(int(pid), fds)
+ if err != nil {
+ return err
+ }
+ process.ops = r
+ if err := c.state.transition(&restoredState{
+ imageDir: opts.ImagesDirectory,
+ c: c,
+ }); err != nil {
+ return err
+ }
+ // create a timestamp indicating when the restored checkpoint was started
+ c.created = time.Now().UTC()
+ if _, err := c.updateState(r); err != nil {
+ return err
+ }
+ if err := os.Remove(filepath.Join(c.root, "checkpoint")); err != nil {
+ if !os.IsNotExist(err) {
+ logrus.Error(err)
+ }
+ }
+ case notify.GetScript() == "orphan-pts-master":
+ scm, err := unix.ParseSocketControlMessage(oob)
+ if err != nil {
+ return err
+ }
+ fds, err := unix.ParseUnixRights(&scm[0])
+ if err != nil {
+ return err
+ }
+
+ master := os.NewFile(uintptr(fds[0]), "orphan-pts-master")
+ defer master.Close()
+
+ // While we can access console.master, using the API is a good idea.
+ if err := utils.SendFd(process.ConsoleSocket, master); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func (c *linuxContainer) updateState(process parentProcess) (*State, error) {
+ c.initProcess = process
+ state, err := c.currentState()
+ if err != nil {
+ return nil, err
+ }
+ err = c.saveState(state)
+ if err != nil {
+ return nil, err
+ }
+ return state, nil
+}
+
+func (c *linuxContainer) saveState(s *State) error {
+ f, err := os.Create(filepath.Join(c.root, stateFilename))
+ if err != nil {
+ return err
+ }
+ defer f.Close()
+ return utils.WriteJSON(f, s)
+}
+
+func (c *linuxContainer) deleteState() error {
+ return os.Remove(filepath.Join(c.root, stateFilename))
+}
+
+func (c *linuxContainer) currentStatus() (Status, error) {
+ if err := c.refreshState(); err != nil {
+ return -1, err
+ }
+ return c.state.status(), nil
+}
+
+// refreshState needs to be called to verify that the current state on the
+// container is what is true. Because consumers of libcontainer can use it
+// out of process we need to verify the container's status based on runtime
+// information and not rely on our in process info.
+func (c *linuxContainer) refreshState() error {
+ paused, err := c.isPaused()
+ if err != nil {
+ return err
+ }
+ if paused {
+ return c.state.transition(&pausedState{c: c})
+ }
+ t, err := c.runType()
+ if err != nil {
+ return err
+ }
+ switch t {
+ case Created:
+ return c.state.transition(&createdState{c: c})
+ case Running:
+ return c.state.transition(&runningState{c: c})
+ }
+ return c.state.transition(&stoppedState{c: c})
+}
+
+func (c *linuxContainer) runType() (Status, error) {
+ if c.initProcess == nil {
+ return Stopped, nil
+ }
+ pid := c.initProcess.pid()
+ stat, err := system.Stat(pid)
+ if err != nil {
+ return Stopped, nil
+ }
+ if stat.StartTime != c.initProcessStartTime || stat.State == system.Zombie || stat.State == system.Dead {
+ return Stopped, nil
+ }
+ // We'll create exec fifo and blocking on it after container is created,
+ // and delete it after start container.
+ if _, err := os.Stat(filepath.Join(c.root, execFifoFilename)); err == nil {
+ return Created, nil
+ }
+ return Running, nil
+}
+
+func (c *linuxContainer) isPaused() (bool, error) {
+ fcg := c.cgroupManager.GetPaths()["freezer"]
+ if fcg == "" {
+ // A container doesn't have a freezer cgroup
+ return false, nil
+ }
+ data, err := ioutil.ReadFile(filepath.Join(fcg, "freezer.state"))
+ if err != nil {
+ // If freezer cgroup is not mounted, the container would just be not paused.
+ if os.IsNotExist(err) {
+ return false, nil
+ }
+ return false, newSystemErrorWithCause(err, "checking if container is paused")
+ }
+ return bytes.Equal(bytes.TrimSpace(data), []byte("FROZEN")), nil
+}
+
+func (c *linuxContainer) currentState() (*State, error) {
+ var (
+ startTime uint64
+ externalDescriptors []string
+ pid = -1
+ )
+ if c.initProcess != nil {
+ pid = c.initProcess.pid()
+ startTime, _ = c.initProcess.startTime()
+ externalDescriptors = c.initProcess.externalDescriptors()
+ }
+ state := &State{
+ BaseState: BaseState{
+ ID: c.ID(),
+ Config: *c.config,
+ InitProcessPid: pid,
+ InitProcessStartTime: startTime,
+ Created: c.created,
+ },
+ Rootless: c.config.Rootless,
+ CgroupPaths: c.cgroupManager.GetPaths(),
+ NamespacePaths: make(map[configs.NamespaceType]string),
+ ExternalDescriptors: externalDescriptors,
+ }
+ if pid > 0 {
+ for _, ns := range c.config.Namespaces {
+ state.NamespacePaths[ns.Type] = ns.GetPath(pid)
+ }
+ for _, nsType := range configs.NamespaceTypes() {
+ if !configs.IsNamespaceSupported(nsType) {
+ continue
+ }
+ if _, ok := state.NamespacePaths[nsType]; !ok {
+ ns := configs.Namespace{Type: nsType}
+ state.NamespacePaths[ns.Type] = ns.GetPath(pid)
+ }
+ }
+ }
+ return state, nil
+}
+
+// orderNamespacePaths sorts namespace paths into a list of paths that we
+// can setns in order.
+func (c *linuxContainer) orderNamespacePaths(namespaces map[configs.NamespaceType]string) ([]string, error) {
+ paths := []string{}
+
+ for _, ns := range configs.NamespaceTypes() {
+
+ // Remove namespaces that we don't need to join.
+ if !c.config.Namespaces.Contains(ns) {
+ continue
+ }
+
+ if p, ok := namespaces[ns]; ok && p != "" {
+ // check if the requested namespace is supported
+ if !configs.IsNamespaceSupported(ns) {
+ return nil, newSystemError(fmt.Errorf("namespace %s is not supported", ns))
+ }
+ // only set to join this namespace if it exists
+ if _, err := os.Lstat(p); err != nil {
+ return nil, newSystemErrorWithCausef(err, "running lstat on namespace path %q", p)
+ }
+ // do not allow namespace path with comma as we use it to separate
+ // the namespace paths
+ if strings.ContainsRune(p, ',') {
+ return nil, newSystemError(fmt.Errorf("invalid path %s", p))
+ }
+ paths = append(paths, fmt.Sprintf("%s:%s", configs.NsName(ns), p))
+ }
+
+ }
+
+ return paths, nil
+}
+
+func encodeIDMapping(idMap []configs.IDMap) ([]byte, error) {
+ data := bytes.NewBuffer(nil)
+ for _, im := range idMap {
+ line := fmt.Sprintf("%d %d %d\n", im.ContainerID, im.HostID, im.Size)
+ if _, err := data.WriteString(line); err != nil {
+ return nil, err
+ }
+ }
+ return data.Bytes(), nil
+}
+
+// bootstrapData encodes the necessary data in netlink binary format
+// as a io.Reader.
+// Consumer can write the data to a bootstrap program
+// such as one that uses nsenter package to bootstrap the container's
+// init process correctly, i.e. with correct namespaces, uid/gid
+// mapping etc.
+func (c *linuxContainer) bootstrapData(cloneFlags uintptr, nsMaps map[configs.NamespaceType]string) (io.Reader, error) {
+ // create the netlink message
+ r := nl.NewNetlinkRequest(int(InitMsg), 0)
+
+ // write cloneFlags
+ r.AddData(&Int32msg{
+ Type: CloneFlagsAttr,
+ Value: uint32(cloneFlags),
+ })
+
+ // write custom namespace paths
+ if len(nsMaps) > 0 {
+ nsPaths, err := c.orderNamespacePaths(nsMaps)
+ if err != nil {
+ return nil, err
+ }
+ r.AddData(&Bytemsg{
+ Type: NsPathsAttr,
+ Value: []byte(strings.Join(nsPaths, ",")),
+ })
+ }
+
+ // write namespace paths only when we are not joining an existing user ns
+ _, joinExistingUser := nsMaps[configs.NEWUSER]
+ if !joinExistingUser {
+ // write uid mappings
+ if len(c.config.UidMappings) > 0 {
+ b, err := encodeIDMapping(c.config.UidMappings)
+ if err != nil {
+ return nil, err
+ }
+ r.AddData(&Bytemsg{
+ Type: UidmapAttr,
+ Value: b,
+ })
+ }
+
+ // write gid mappings
+ if len(c.config.GidMappings) > 0 {
+ b, err := encodeIDMapping(c.config.GidMappings)
+ if err != nil {
+ return nil, err
+ }
+ r.AddData(&Bytemsg{
+ Type: GidmapAttr,
+ Value: b,
+ })
+ // The following only applies if we are root.
+ if !c.config.Rootless {
+ // check if we have CAP_SETGID to setgroup properly
+ pid, err := capability.NewPid(os.Getpid())
+ if err != nil {
+ return nil, err
+ }
+ if !pid.Get(capability.EFFECTIVE, capability.CAP_SETGID) {
+ r.AddData(&Boolmsg{
+ Type: SetgroupAttr,
+ Value: true,
+ })
+ }
+ }
+ }
+ }
+
+ // write oom_score_adj
+ r.AddData(&Bytemsg{
+ Type: OomScoreAdjAttr,
+ Value: []byte(fmt.Sprintf("%d", c.config.OomScoreAdj)),
+ })
+
+ // write rootless
+ r.AddData(&Boolmsg{
+ Type: RootlessAttr,
+ Value: c.config.Rootless,
+ })
+
+ return bytes.NewReader(r.Serialize()), nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/container_solaris.go b/vendor/github.com/opencontainers/runc/libcontainer/container_solaris.go
new file mode 100644
index 000000000..bb84ff740
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/container_solaris.go
@@ -0,0 +1,20 @@
+package libcontainer
+
+// State represents a running container's state
+type State struct {
+ BaseState
+
+ // Platform specific fields below here
+}
+
+// A libcontainer container object.
+//
+// Each container is thread-safe within the same process. Since a container can
+// be destroyed by a separate process, any function may return that the container
+// was not found.
+type Container interface {
+ BaseContainer
+
+ // Methods below here are platform specific
+
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/container_windows.go b/vendor/github.com/opencontainers/runc/libcontainer/container_windows.go
new file mode 100644
index 000000000..bb84ff740
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/container_windows.go
@@ -0,0 +1,20 @@
+package libcontainer
+
+// State represents a running container's state
+type State struct {
+ BaseState
+
+ // Platform specific fields below here
+}
+
+// A libcontainer container object.
+//
+// Each container is thread-safe within the same process. Since a container can
+// be destroyed by a separate process, any function may return that the container
+// was not found.
+type Container interface {
+ BaseContainer
+
+ // Methods below here are platform specific
+
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_linux.go
new file mode 100644
index 000000000..9423d2464
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_linux.go
@@ -0,0 +1,37 @@
+package libcontainer
+
+// cgroup restoring strategy provided by criu
+type cgMode uint32
+
+const (
+ CRIU_CG_MODE_SOFT cgMode = 3 + iota // restore cgroup properties if only dir created by criu
+ CRIU_CG_MODE_FULL // always restore all cgroups and their properties
+ CRIU_CG_MODE_STRICT // restore all, requiring them to not present in the system
+ CRIU_CG_MODE_DEFAULT // the same as CRIU_CG_MODE_SOFT
+)
+
+type CriuPageServerInfo struct {
+ Address string // IP address of CRIU page server
+ Port int32 // port number of CRIU page server
+}
+
+type VethPairName struct {
+ ContainerInterfaceName string
+ HostInterfaceName string
+}
+
+type CriuOpts struct {
+ ImagesDirectory string // directory for storing image files
+ WorkDirectory string // directory to cd and write logs/pidfiles/stats to
+ ParentImage string // direcotry for storing parent image files in pre-dump and dump
+ LeaveRunning bool // leave container in running state after checkpoint
+ TcpEstablished bool // checkpoint/restore established TCP connections
+ ExternalUnixConnections bool // allow external unix connections
+ ShellJob bool // allow to dump and restore shell jobs
+ FileLocks bool // handle file locks, for safety
+ PreDump bool // call criu predump to perform iterative checkpoint
+ PageServer CriuPageServerInfo // allow to dump to criu page server
+ VethPairs []VethPairName // pass the veth to criu when restore
+ ManageCgroupsMode cgMode // dump or restore cgroup mode
+ EmptyNs uint32 // don't c/r properties for namespace from this mask
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_windows.go b/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_windows.go
new file mode 100644
index 000000000..bc9207703
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/criu_opts_windows.go
@@ -0,0 +1,6 @@
+package libcontainer
+
+// TODO Windows: This can ultimately be entirely factored out as criu is
+// a Unix concept not relevant on Windows.
+type CriuOpts struct {
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.pb.go b/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.pb.go
new file mode 100644
index 000000000..21af9db97
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.pb.go
@@ -0,0 +1,1178 @@
+// Code generated by protoc-gen-go.
+// source: criurpc.proto
+// DO NOT EDIT!
+
+/*
+Package criurpc is a generated protocol buffer package.
+
+It is generated from these files:
+ criurpc.proto
+
+It has these top-level messages:
+ CriuPageServerInfo
+ CriuVethPair
+ ExtMountMap
+ JoinNamespace
+ InheritFd
+ CgroupRoot
+ UnixSk
+ CriuOpts
+ CriuDumpResp
+ CriuRestoreResp
+ CriuNotify
+ CriuFeatures
+ CriuReq
+ CriuResp
+ CriuVersion
+*/
+package criurpc
+
+import proto "github.com/golang/protobuf/proto"
+import fmt "fmt"
+import math "math"
+
+// Reference imports to suppress errors if they are not otherwise used.
+var _ = proto.Marshal
+var _ = fmt.Errorf
+var _ = math.Inf
+
+// This is a compile-time assertion to ensure that this generated file
+// is compatible with the proto package it is being compiled against.
+// A compilation error at this line likely means your copy of the
+// proto package needs to be updated.
+const _ = proto.ProtoPackageIsVersion2 // please upgrade the proto package
+
+type CriuCgMode int32
+
+const (
+ CriuCgMode_IGNORE CriuCgMode = 0
+ CriuCgMode_CG_NONE CriuCgMode = 1
+ CriuCgMode_PROPS CriuCgMode = 2
+ CriuCgMode_SOFT CriuCgMode = 3
+ CriuCgMode_FULL CriuCgMode = 4
+ CriuCgMode_STRICT CriuCgMode = 5
+ CriuCgMode_DEFAULT CriuCgMode = 6
+)
+
+var CriuCgMode_name = map[int32]string{
+ 0: "IGNORE",
+ 1: "CG_NONE",
+ 2: "PROPS",
+ 3: "SOFT",
+ 4: "FULL",
+ 5: "STRICT",
+ 6: "DEFAULT",
+}
+var CriuCgMode_value = map[string]int32{
+ "IGNORE": 0,
+ "CG_NONE": 1,
+ "PROPS": 2,
+ "SOFT": 3,
+ "FULL": 4,
+ "STRICT": 5,
+ "DEFAULT": 6,
+}
+
+func (x CriuCgMode) Enum() *CriuCgMode {
+ p := new(CriuCgMode)
+ *p = x
+ return p
+}
+func (x CriuCgMode) String() string {
+ return proto.EnumName(CriuCgMode_name, int32(x))
+}
+func (x *CriuCgMode) UnmarshalJSON(data []byte) error {
+ value, err := proto.UnmarshalJSONEnum(CriuCgMode_value, data, "CriuCgMode")
+ if err != nil {
+ return err
+ }
+ *x = CriuCgMode(value)
+ return nil
+}
+func (CriuCgMode) EnumDescriptor() ([]byte, []int) { return fileDescriptor0, []int{0} }
+
+type CriuReqType int32
+
+const (
+ CriuReqType_EMPTY CriuReqType = 0
+ CriuReqType_DUMP CriuReqType = 1
+ CriuReqType_RESTORE CriuReqType = 2
+ CriuReqType_CHECK CriuReqType = 3
+ CriuReqType_PRE_DUMP CriuReqType = 4
+ CriuReqType_PAGE_SERVER CriuReqType = 5
+ CriuReqType_NOTIFY CriuReqType = 6
+ CriuReqType_CPUINFO_DUMP CriuReqType = 7
+ CriuReqType_CPUINFO_CHECK CriuReqType = 8
+ CriuReqType_FEATURE_CHECK CriuReqType = 9
+ CriuReqType_VERSION CriuReqType = 10
+)
+
+var CriuReqType_name = map[int32]string{
+ 0: "EMPTY",
+ 1: "DUMP",
+ 2: "RESTORE",
+ 3: "CHECK",
+ 4: "PRE_DUMP",
+ 5: "PAGE_SERVER",
+ 6: "NOTIFY",
+ 7: "CPUINFO_DUMP",
+ 8: "CPUINFO_CHECK",
+ 9: "FEATURE_CHECK",
+ 10: "VERSION",
+}
+var CriuReqType_value = map[string]int32{
+ "EMPTY": 0,
+ "DUMP": 1,
+ "RESTORE": 2,
+ "CHECK": 3,
+ "PRE_DUMP": 4,
+ "PAGE_SERVER": 5,
+ "NOTIFY": 6,
+ "CPUINFO_DUMP": 7,
+ "CPUINFO_CHECK": 8,
+ "FEATURE_CHECK": 9,
+ "VERSION": 10,
+}
+
+func (x CriuReqType) Enum() *CriuReqType {
+ p := new(CriuReqType)
+ *p = x
+ return p
+}
+func (x CriuReqType) String() string {
+ return proto.EnumName(CriuReqType_name, int32(x))
+}
+func (x *CriuReqType) UnmarshalJSON(data []byte) error {
+ value, err := proto.UnmarshalJSONEnum(CriuReqType_value, data, "CriuReqType")
+ if err != nil {
+ return err
+ }
+ *x = CriuReqType(value)
+ return nil
+}
+func (CriuReqType) EnumDescriptor() ([]byte, []int) { return fileDescriptor0, []int{1} }
+
+type CriuPageServerInfo struct {
+ Address *string `protobuf:"bytes,1,opt,name=address" json:"address,omitempty"`
+ Port *int32 `protobuf:"varint,2,opt,name=port" json:"port,omitempty"`
+ Pid *int32 `protobuf:"varint,3,opt,name=pid" json:"pid,omitempty"`
+ Fd *int32 `protobuf:"varint,4,opt,name=fd" json:"fd,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuPageServerInfo) Reset() { *m = CriuPageServerInfo{} }
+func (m *CriuPageServerInfo) String() string { return proto.CompactTextString(m) }
+func (*CriuPageServerInfo) ProtoMessage() {}
+func (*CriuPageServerInfo) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{0} }
+
+func (m *CriuPageServerInfo) GetAddress() string {
+ if m != nil && m.Address != nil {
+ return *m.Address
+ }
+ return ""
+}
+
+func (m *CriuPageServerInfo) GetPort() int32 {
+ if m != nil && m.Port != nil {
+ return *m.Port
+ }
+ return 0
+}
+
+func (m *CriuPageServerInfo) GetPid() int32 {
+ if m != nil && m.Pid != nil {
+ return *m.Pid
+ }
+ return 0
+}
+
+func (m *CriuPageServerInfo) GetFd() int32 {
+ if m != nil && m.Fd != nil {
+ return *m.Fd
+ }
+ return 0
+}
+
+type CriuVethPair struct {
+ IfIn *string `protobuf:"bytes,1,req,name=if_in,json=ifIn" json:"if_in,omitempty"`
+ IfOut *string `protobuf:"bytes,2,req,name=if_out,json=ifOut" json:"if_out,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuVethPair) Reset() { *m = CriuVethPair{} }
+func (m *CriuVethPair) String() string { return proto.CompactTextString(m) }
+func (*CriuVethPair) ProtoMessage() {}
+func (*CriuVethPair) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{1} }
+
+func (m *CriuVethPair) GetIfIn() string {
+ if m != nil && m.IfIn != nil {
+ return *m.IfIn
+ }
+ return ""
+}
+
+func (m *CriuVethPair) GetIfOut() string {
+ if m != nil && m.IfOut != nil {
+ return *m.IfOut
+ }
+ return ""
+}
+
+type ExtMountMap struct {
+ Key *string `protobuf:"bytes,1,req,name=key" json:"key,omitempty"`
+ Val *string `protobuf:"bytes,2,req,name=val" json:"val,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *ExtMountMap) Reset() { *m = ExtMountMap{} }
+func (m *ExtMountMap) String() string { return proto.CompactTextString(m) }
+func (*ExtMountMap) ProtoMessage() {}
+func (*ExtMountMap) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{2} }
+
+func (m *ExtMountMap) GetKey() string {
+ if m != nil && m.Key != nil {
+ return *m.Key
+ }
+ return ""
+}
+
+func (m *ExtMountMap) GetVal() string {
+ if m != nil && m.Val != nil {
+ return *m.Val
+ }
+ return ""
+}
+
+type JoinNamespace struct {
+ Ns *string `protobuf:"bytes,1,req,name=ns" json:"ns,omitempty"`
+ NsFile *string `protobuf:"bytes,2,req,name=ns_file,json=nsFile" json:"ns_file,omitempty"`
+ ExtraOpt *string `protobuf:"bytes,3,opt,name=extra_opt,json=extraOpt" json:"extra_opt,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *JoinNamespace) Reset() { *m = JoinNamespace{} }
+func (m *JoinNamespace) String() string { return proto.CompactTextString(m) }
+func (*JoinNamespace) ProtoMessage() {}
+func (*JoinNamespace) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{3} }
+
+func (m *JoinNamespace) GetNs() string {
+ if m != nil && m.Ns != nil {
+ return *m.Ns
+ }
+ return ""
+}
+
+func (m *JoinNamespace) GetNsFile() string {
+ if m != nil && m.NsFile != nil {
+ return *m.NsFile
+ }
+ return ""
+}
+
+func (m *JoinNamespace) GetExtraOpt() string {
+ if m != nil && m.ExtraOpt != nil {
+ return *m.ExtraOpt
+ }
+ return ""
+}
+
+type InheritFd struct {
+ Key *string `protobuf:"bytes,1,req,name=key" json:"key,omitempty"`
+ Fd *int32 `protobuf:"varint,2,req,name=fd" json:"fd,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *InheritFd) Reset() { *m = InheritFd{} }
+func (m *InheritFd) String() string { return proto.CompactTextString(m) }
+func (*InheritFd) ProtoMessage() {}
+func (*InheritFd) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{4} }
+
+func (m *InheritFd) GetKey() string {
+ if m != nil && m.Key != nil {
+ return *m.Key
+ }
+ return ""
+}
+
+func (m *InheritFd) GetFd() int32 {
+ if m != nil && m.Fd != nil {
+ return *m.Fd
+ }
+ return 0
+}
+
+type CgroupRoot struct {
+ Ctrl *string `protobuf:"bytes,1,opt,name=ctrl" json:"ctrl,omitempty"`
+ Path *string `protobuf:"bytes,2,req,name=path" json:"path,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CgroupRoot) Reset() { *m = CgroupRoot{} }
+func (m *CgroupRoot) String() string { return proto.CompactTextString(m) }
+func (*CgroupRoot) ProtoMessage() {}
+func (*CgroupRoot) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{5} }
+
+func (m *CgroupRoot) GetCtrl() string {
+ if m != nil && m.Ctrl != nil {
+ return *m.Ctrl
+ }
+ return ""
+}
+
+func (m *CgroupRoot) GetPath() string {
+ if m != nil && m.Path != nil {
+ return *m.Path
+ }
+ return ""
+}
+
+type UnixSk struct {
+ Inode *uint32 `protobuf:"varint,1,req,name=inode" json:"inode,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *UnixSk) Reset() { *m = UnixSk{} }
+func (m *UnixSk) String() string { return proto.CompactTextString(m) }
+func (*UnixSk) ProtoMessage() {}
+func (*UnixSk) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{6} }
+
+func (m *UnixSk) GetInode() uint32 {
+ if m != nil && m.Inode != nil {
+ return *m.Inode
+ }
+ return 0
+}
+
+type CriuOpts struct {
+ ImagesDirFd *int32 `protobuf:"varint,1,req,name=images_dir_fd,json=imagesDirFd" json:"images_dir_fd,omitempty"`
+ Pid *int32 `protobuf:"varint,2,opt,name=pid" json:"pid,omitempty"`
+ LeaveRunning *bool `protobuf:"varint,3,opt,name=leave_running,json=leaveRunning" json:"leave_running,omitempty"`
+ ExtUnixSk *bool `protobuf:"varint,4,opt,name=ext_unix_sk,json=extUnixSk" json:"ext_unix_sk,omitempty"`
+ TcpEstablished *bool `protobuf:"varint,5,opt,name=tcp_established,json=tcpEstablished" json:"tcp_established,omitempty"`
+ EvasiveDevices *bool `protobuf:"varint,6,opt,name=evasive_devices,json=evasiveDevices" json:"evasive_devices,omitempty"`
+ ShellJob *bool `protobuf:"varint,7,opt,name=shell_job,json=shellJob" json:"shell_job,omitempty"`
+ FileLocks *bool `protobuf:"varint,8,opt,name=file_locks,json=fileLocks" json:"file_locks,omitempty"`
+ LogLevel *int32 `protobuf:"varint,9,opt,name=log_level,json=logLevel,def=2" json:"log_level,omitempty"`
+ LogFile *string `protobuf:"bytes,10,opt,name=log_file,json=logFile" json:"log_file,omitempty"`
+ Ps *CriuPageServerInfo `protobuf:"bytes,11,opt,name=ps" json:"ps,omitempty"`
+ NotifyScripts *bool `protobuf:"varint,12,opt,name=notify_scripts,json=notifyScripts" json:"notify_scripts,omitempty"`
+ Root *string `protobuf:"bytes,13,opt,name=root" json:"root,omitempty"`
+ ParentImg *string `protobuf:"bytes,14,opt,name=parent_img,json=parentImg" json:"parent_img,omitempty"`
+ TrackMem *bool `protobuf:"varint,15,opt,name=track_mem,json=trackMem" json:"track_mem,omitempty"`
+ AutoDedup *bool `protobuf:"varint,16,opt,name=auto_dedup,json=autoDedup" json:"auto_dedup,omitempty"`
+ WorkDirFd *int32 `protobuf:"varint,17,opt,name=work_dir_fd,json=workDirFd" json:"work_dir_fd,omitempty"`
+ LinkRemap *bool `protobuf:"varint,18,opt,name=link_remap,json=linkRemap" json:"link_remap,omitempty"`
+ Veths []*CriuVethPair `protobuf:"bytes,19,rep,name=veths" json:"veths,omitempty"`
+ CpuCap *uint32 `protobuf:"varint,20,opt,name=cpu_cap,json=cpuCap,def=4294967295" json:"cpu_cap,omitempty"`
+ ForceIrmap *bool `protobuf:"varint,21,opt,name=force_irmap,json=forceIrmap" json:"force_irmap,omitempty"`
+ ExecCmd []string `protobuf:"bytes,22,rep,name=exec_cmd,json=execCmd" json:"exec_cmd,omitempty"`
+ ExtMnt []*ExtMountMap `protobuf:"bytes,23,rep,name=ext_mnt,json=extMnt" json:"ext_mnt,omitempty"`
+ ManageCgroups *bool `protobuf:"varint,24,opt,name=manage_cgroups,json=manageCgroups" json:"manage_cgroups,omitempty"`
+ CgRoot []*CgroupRoot `protobuf:"bytes,25,rep,name=cg_root,json=cgRoot" json:"cg_root,omitempty"`
+ RstSibling *bool `protobuf:"varint,26,opt,name=rst_sibling,json=rstSibling" json:"rst_sibling,omitempty"`
+ InheritFd []*InheritFd `protobuf:"bytes,27,rep,name=inherit_fd,json=inheritFd" json:"inherit_fd,omitempty"`
+ AutoExtMnt *bool `protobuf:"varint,28,opt,name=auto_ext_mnt,json=autoExtMnt" json:"auto_ext_mnt,omitempty"`
+ ExtSharing *bool `protobuf:"varint,29,opt,name=ext_sharing,json=extSharing" json:"ext_sharing,omitempty"`
+ ExtMasters *bool `protobuf:"varint,30,opt,name=ext_masters,json=extMasters" json:"ext_masters,omitempty"`
+ SkipMnt []string `protobuf:"bytes,31,rep,name=skip_mnt,json=skipMnt" json:"skip_mnt,omitempty"`
+ EnableFs []string `protobuf:"bytes,32,rep,name=enable_fs,json=enableFs" json:"enable_fs,omitempty"`
+ UnixSkIno []*UnixSk `protobuf:"bytes,33,rep,name=unix_sk_ino,json=unixSkIno" json:"unix_sk_ino,omitempty"`
+ ManageCgroupsMode *CriuCgMode `protobuf:"varint,34,opt,name=manage_cgroups_mode,json=manageCgroupsMode,enum=CriuCgMode" json:"manage_cgroups_mode,omitempty"`
+ GhostLimit *uint32 `protobuf:"varint,35,opt,name=ghost_limit,json=ghostLimit,def=1048576" json:"ghost_limit,omitempty"`
+ IrmapScanPaths []string `protobuf:"bytes,36,rep,name=irmap_scan_paths,json=irmapScanPaths" json:"irmap_scan_paths,omitempty"`
+ External []string `protobuf:"bytes,37,rep,name=external" json:"external,omitempty"`
+ EmptyNs *uint32 `protobuf:"varint,38,opt,name=empty_ns,json=emptyNs" json:"empty_ns,omitempty"`
+ JoinNs []*JoinNamespace `protobuf:"bytes,39,rep,name=join_ns,json=joinNs" json:"join_ns,omitempty"`
+ CgroupProps *string `protobuf:"bytes,41,opt,name=cgroup_props,json=cgroupProps" json:"cgroup_props,omitempty"`
+ CgroupPropsFile *string `protobuf:"bytes,42,opt,name=cgroup_props_file,json=cgroupPropsFile" json:"cgroup_props_file,omitempty"`
+ CgroupDumpController []string `protobuf:"bytes,43,rep,name=cgroup_dump_controller,json=cgroupDumpController" json:"cgroup_dump_controller,omitempty"`
+ FreezeCgroup *string `protobuf:"bytes,44,opt,name=freeze_cgroup,json=freezeCgroup" json:"freeze_cgroup,omitempty"`
+ Timeout *uint32 `protobuf:"varint,45,opt,name=timeout" json:"timeout,omitempty"`
+ TcpSkipInFlight *bool `protobuf:"varint,46,opt,name=tcp_skip_in_flight,json=tcpSkipInFlight" json:"tcp_skip_in_flight,omitempty"`
+ WeakSysctls *bool `protobuf:"varint,47,opt,name=weak_sysctls,json=weakSysctls" json:"weak_sysctls,omitempty"`
+ LazyPages *bool `protobuf:"varint,48,opt,name=lazy_pages,json=lazyPages" json:"lazy_pages,omitempty"`
+ StatusFd *int32 `protobuf:"varint,49,opt,name=status_fd,json=statusFd" json:"status_fd,omitempty"`
+ OrphanPtsMaster *bool `protobuf:"varint,50,opt,name=orphan_pts_master,json=orphanPtsMaster" json:"orphan_pts_master,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuOpts) Reset() { *m = CriuOpts{} }
+func (m *CriuOpts) String() string { return proto.CompactTextString(m) }
+func (*CriuOpts) ProtoMessage() {}
+func (*CriuOpts) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{7} }
+
+const Default_CriuOpts_LogLevel int32 = 2
+const Default_CriuOpts_CpuCap uint32 = 4294967295
+const Default_CriuOpts_GhostLimit uint32 = 1048576
+
+func (m *CriuOpts) GetImagesDirFd() int32 {
+ if m != nil && m.ImagesDirFd != nil {
+ return *m.ImagesDirFd
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetPid() int32 {
+ if m != nil && m.Pid != nil {
+ return *m.Pid
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetLeaveRunning() bool {
+ if m != nil && m.LeaveRunning != nil {
+ return *m.LeaveRunning
+ }
+ return false
+}
+
+func (m *CriuOpts) GetExtUnixSk() bool {
+ if m != nil && m.ExtUnixSk != nil {
+ return *m.ExtUnixSk
+ }
+ return false
+}
+
+func (m *CriuOpts) GetTcpEstablished() bool {
+ if m != nil && m.TcpEstablished != nil {
+ return *m.TcpEstablished
+ }
+ return false
+}
+
+func (m *CriuOpts) GetEvasiveDevices() bool {
+ if m != nil && m.EvasiveDevices != nil {
+ return *m.EvasiveDevices
+ }
+ return false
+}
+
+func (m *CriuOpts) GetShellJob() bool {
+ if m != nil && m.ShellJob != nil {
+ return *m.ShellJob
+ }
+ return false
+}
+
+func (m *CriuOpts) GetFileLocks() bool {
+ if m != nil && m.FileLocks != nil {
+ return *m.FileLocks
+ }
+ return false
+}
+
+func (m *CriuOpts) GetLogLevel() int32 {
+ if m != nil && m.LogLevel != nil {
+ return *m.LogLevel
+ }
+ return Default_CriuOpts_LogLevel
+}
+
+func (m *CriuOpts) GetLogFile() string {
+ if m != nil && m.LogFile != nil {
+ return *m.LogFile
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetPs() *CriuPageServerInfo {
+ if m != nil {
+ return m.Ps
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetNotifyScripts() bool {
+ if m != nil && m.NotifyScripts != nil {
+ return *m.NotifyScripts
+ }
+ return false
+}
+
+func (m *CriuOpts) GetRoot() string {
+ if m != nil && m.Root != nil {
+ return *m.Root
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetParentImg() string {
+ if m != nil && m.ParentImg != nil {
+ return *m.ParentImg
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetTrackMem() bool {
+ if m != nil && m.TrackMem != nil {
+ return *m.TrackMem
+ }
+ return false
+}
+
+func (m *CriuOpts) GetAutoDedup() bool {
+ if m != nil && m.AutoDedup != nil {
+ return *m.AutoDedup
+ }
+ return false
+}
+
+func (m *CriuOpts) GetWorkDirFd() int32 {
+ if m != nil && m.WorkDirFd != nil {
+ return *m.WorkDirFd
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetLinkRemap() bool {
+ if m != nil && m.LinkRemap != nil {
+ return *m.LinkRemap
+ }
+ return false
+}
+
+func (m *CriuOpts) GetVeths() []*CriuVethPair {
+ if m != nil {
+ return m.Veths
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetCpuCap() uint32 {
+ if m != nil && m.CpuCap != nil {
+ return *m.CpuCap
+ }
+ return Default_CriuOpts_CpuCap
+}
+
+func (m *CriuOpts) GetForceIrmap() bool {
+ if m != nil && m.ForceIrmap != nil {
+ return *m.ForceIrmap
+ }
+ return false
+}
+
+func (m *CriuOpts) GetExecCmd() []string {
+ if m != nil {
+ return m.ExecCmd
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetExtMnt() []*ExtMountMap {
+ if m != nil {
+ return m.ExtMnt
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetManageCgroups() bool {
+ if m != nil && m.ManageCgroups != nil {
+ return *m.ManageCgroups
+ }
+ return false
+}
+
+func (m *CriuOpts) GetCgRoot() []*CgroupRoot {
+ if m != nil {
+ return m.CgRoot
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetRstSibling() bool {
+ if m != nil && m.RstSibling != nil {
+ return *m.RstSibling
+ }
+ return false
+}
+
+func (m *CriuOpts) GetInheritFd() []*InheritFd {
+ if m != nil {
+ return m.InheritFd
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetAutoExtMnt() bool {
+ if m != nil && m.AutoExtMnt != nil {
+ return *m.AutoExtMnt
+ }
+ return false
+}
+
+func (m *CriuOpts) GetExtSharing() bool {
+ if m != nil && m.ExtSharing != nil {
+ return *m.ExtSharing
+ }
+ return false
+}
+
+func (m *CriuOpts) GetExtMasters() bool {
+ if m != nil && m.ExtMasters != nil {
+ return *m.ExtMasters
+ }
+ return false
+}
+
+func (m *CriuOpts) GetSkipMnt() []string {
+ if m != nil {
+ return m.SkipMnt
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetEnableFs() []string {
+ if m != nil {
+ return m.EnableFs
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetUnixSkIno() []*UnixSk {
+ if m != nil {
+ return m.UnixSkIno
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetManageCgroupsMode() CriuCgMode {
+ if m != nil && m.ManageCgroupsMode != nil {
+ return *m.ManageCgroupsMode
+ }
+ return CriuCgMode_IGNORE
+}
+
+func (m *CriuOpts) GetGhostLimit() uint32 {
+ if m != nil && m.GhostLimit != nil {
+ return *m.GhostLimit
+ }
+ return Default_CriuOpts_GhostLimit
+}
+
+func (m *CriuOpts) GetIrmapScanPaths() []string {
+ if m != nil {
+ return m.IrmapScanPaths
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetExternal() []string {
+ if m != nil {
+ return m.External
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetEmptyNs() uint32 {
+ if m != nil && m.EmptyNs != nil {
+ return *m.EmptyNs
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetJoinNs() []*JoinNamespace {
+ if m != nil {
+ return m.JoinNs
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetCgroupProps() string {
+ if m != nil && m.CgroupProps != nil {
+ return *m.CgroupProps
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetCgroupPropsFile() string {
+ if m != nil && m.CgroupPropsFile != nil {
+ return *m.CgroupPropsFile
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetCgroupDumpController() []string {
+ if m != nil {
+ return m.CgroupDumpController
+ }
+ return nil
+}
+
+func (m *CriuOpts) GetFreezeCgroup() string {
+ if m != nil && m.FreezeCgroup != nil {
+ return *m.FreezeCgroup
+ }
+ return ""
+}
+
+func (m *CriuOpts) GetTimeout() uint32 {
+ if m != nil && m.Timeout != nil {
+ return *m.Timeout
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetTcpSkipInFlight() bool {
+ if m != nil && m.TcpSkipInFlight != nil {
+ return *m.TcpSkipInFlight
+ }
+ return false
+}
+
+func (m *CriuOpts) GetWeakSysctls() bool {
+ if m != nil && m.WeakSysctls != nil {
+ return *m.WeakSysctls
+ }
+ return false
+}
+
+func (m *CriuOpts) GetLazyPages() bool {
+ if m != nil && m.LazyPages != nil {
+ return *m.LazyPages
+ }
+ return false
+}
+
+func (m *CriuOpts) GetStatusFd() int32 {
+ if m != nil && m.StatusFd != nil {
+ return *m.StatusFd
+ }
+ return 0
+}
+
+func (m *CriuOpts) GetOrphanPtsMaster() bool {
+ if m != nil && m.OrphanPtsMaster != nil {
+ return *m.OrphanPtsMaster
+ }
+ return false
+}
+
+type CriuDumpResp struct {
+ Restored *bool `protobuf:"varint,1,opt,name=restored" json:"restored,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuDumpResp) Reset() { *m = CriuDumpResp{} }
+func (m *CriuDumpResp) String() string { return proto.CompactTextString(m) }
+func (*CriuDumpResp) ProtoMessage() {}
+func (*CriuDumpResp) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{8} }
+
+func (m *CriuDumpResp) GetRestored() bool {
+ if m != nil && m.Restored != nil {
+ return *m.Restored
+ }
+ return false
+}
+
+type CriuRestoreResp struct {
+ Pid *int32 `protobuf:"varint,1,req,name=pid" json:"pid,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuRestoreResp) Reset() { *m = CriuRestoreResp{} }
+func (m *CriuRestoreResp) String() string { return proto.CompactTextString(m) }
+func (*CriuRestoreResp) ProtoMessage() {}
+func (*CriuRestoreResp) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{9} }
+
+func (m *CriuRestoreResp) GetPid() int32 {
+ if m != nil && m.Pid != nil {
+ return *m.Pid
+ }
+ return 0
+}
+
+type CriuNotify struct {
+ Script *string `protobuf:"bytes,1,opt,name=script" json:"script,omitempty"`
+ Pid *int32 `protobuf:"varint,2,opt,name=pid" json:"pid,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuNotify) Reset() { *m = CriuNotify{} }
+func (m *CriuNotify) String() string { return proto.CompactTextString(m) }
+func (*CriuNotify) ProtoMessage() {}
+func (*CriuNotify) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{10} }
+
+func (m *CriuNotify) GetScript() string {
+ if m != nil && m.Script != nil {
+ return *m.Script
+ }
+ return ""
+}
+
+func (m *CriuNotify) GetPid() int32 {
+ if m != nil && m.Pid != nil {
+ return *m.Pid
+ }
+ return 0
+}
+
+//
+// List of features which can queried via
+// CRIU_REQ_TYPE__FEATURE_CHECK
+type CriuFeatures struct {
+ MemTrack *bool `protobuf:"varint,1,opt,name=mem_track,json=memTrack" json:"mem_track,omitempty"`
+ LazyPages *bool `protobuf:"varint,2,opt,name=lazy_pages,json=lazyPages" json:"lazy_pages,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuFeatures) Reset() { *m = CriuFeatures{} }
+func (m *CriuFeatures) String() string { return proto.CompactTextString(m) }
+func (*CriuFeatures) ProtoMessage() {}
+func (*CriuFeatures) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{11} }
+
+func (m *CriuFeatures) GetMemTrack() bool {
+ if m != nil && m.MemTrack != nil {
+ return *m.MemTrack
+ }
+ return false
+}
+
+func (m *CriuFeatures) GetLazyPages() bool {
+ if m != nil && m.LazyPages != nil {
+ return *m.LazyPages
+ }
+ return false
+}
+
+type CriuReq struct {
+ Type *CriuReqType `protobuf:"varint,1,req,name=type,enum=CriuReqType" json:"type,omitempty"`
+ Opts *CriuOpts `protobuf:"bytes,2,opt,name=opts" json:"opts,omitempty"`
+ NotifySuccess *bool `protobuf:"varint,3,opt,name=notify_success,json=notifySuccess" json:"notify_success,omitempty"`
+ //
+ // When set service won't close the connection but
+ // will wait for more req-s to appear. Works not
+ // for all request types.
+ KeepOpen *bool `protobuf:"varint,4,opt,name=keep_open,json=keepOpen" json:"keep_open,omitempty"`
+ //
+ // 'features' can be used to query which features
+ // are supported by the installed criu/kernel
+ // via RPC.
+ Features *CriuFeatures `protobuf:"bytes,5,opt,name=features" json:"features,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuReq) Reset() { *m = CriuReq{} }
+func (m *CriuReq) String() string { return proto.CompactTextString(m) }
+func (*CriuReq) ProtoMessage() {}
+func (*CriuReq) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{12} }
+
+func (m *CriuReq) GetType() CriuReqType {
+ if m != nil && m.Type != nil {
+ return *m.Type
+ }
+ return CriuReqType_EMPTY
+}
+
+func (m *CriuReq) GetOpts() *CriuOpts {
+ if m != nil {
+ return m.Opts
+ }
+ return nil
+}
+
+func (m *CriuReq) GetNotifySuccess() bool {
+ if m != nil && m.NotifySuccess != nil {
+ return *m.NotifySuccess
+ }
+ return false
+}
+
+func (m *CriuReq) GetKeepOpen() bool {
+ if m != nil && m.KeepOpen != nil {
+ return *m.KeepOpen
+ }
+ return false
+}
+
+func (m *CriuReq) GetFeatures() *CriuFeatures {
+ if m != nil {
+ return m.Features
+ }
+ return nil
+}
+
+type CriuResp struct {
+ Type *CriuReqType `protobuf:"varint,1,req,name=type,enum=CriuReqType" json:"type,omitempty"`
+ Success *bool `protobuf:"varint,2,req,name=success" json:"success,omitempty"`
+ Dump *CriuDumpResp `protobuf:"bytes,3,opt,name=dump" json:"dump,omitempty"`
+ Restore *CriuRestoreResp `protobuf:"bytes,4,opt,name=restore" json:"restore,omitempty"`
+ Notify *CriuNotify `protobuf:"bytes,5,opt,name=notify" json:"notify,omitempty"`
+ Ps *CriuPageServerInfo `protobuf:"bytes,6,opt,name=ps" json:"ps,omitempty"`
+ CrErrno *int32 `protobuf:"varint,7,opt,name=cr_errno,json=crErrno" json:"cr_errno,omitempty"`
+ Features *CriuFeatures `protobuf:"bytes,8,opt,name=features" json:"features,omitempty"`
+ CrErrmsg *string `protobuf:"bytes,9,opt,name=cr_errmsg,json=crErrmsg" json:"cr_errmsg,omitempty"`
+ Version *CriuVersion `protobuf:"bytes,10,opt,name=version" json:"version,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuResp) Reset() { *m = CriuResp{} }
+func (m *CriuResp) String() string { return proto.CompactTextString(m) }
+func (*CriuResp) ProtoMessage() {}
+func (*CriuResp) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{13} }
+
+func (m *CriuResp) GetType() CriuReqType {
+ if m != nil && m.Type != nil {
+ return *m.Type
+ }
+ return CriuReqType_EMPTY
+}
+
+func (m *CriuResp) GetSuccess() bool {
+ if m != nil && m.Success != nil {
+ return *m.Success
+ }
+ return false
+}
+
+func (m *CriuResp) GetDump() *CriuDumpResp {
+ if m != nil {
+ return m.Dump
+ }
+ return nil
+}
+
+func (m *CriuResp) GetRestore() *CriuRestoreResp {
+ if m != nil {
+ return m.Restore
+ }
+ return nil
+}
+
+func (m *CriuResp) GetNotify() *CriuNotify {
+ if m != nil {
+ return m.Notify
+ }
+ return nil
+}
+
+func (m *CriuResp) GetPs() *CriuPageServerInfo {
+ if m != nil {
+ return m.Ps
+ }
+ return nil
+}
+
+func (m *CriuResp) GetCrErrno() int32 {
+ if m != nil && m.CrErrno != nil {
+ return *m.CrErrno
+ }
+ return 0
+}
+
+func (m *CriuResp) GetFeatures() *CriuFeatures {
+ if m != nil {
+ return m.Features
+ }
+ return nil
+}
+
+func (m *CriuResp) GetCrErrmsg() string {
+ if m != nil && m.CrErrmsg != nil {
+ return *m.CrErrmsg
+ }
+ return ""
+}
+
+func (m *CriuResp) GetVersion() *CriuVersion {
+ if m != nil {
+ return m.Version
+ }
+ return nil
+}
+
+// Answer for criu_req_type.VERSION requests
+type CriuVersion struct {
+ Major *int32 `protobuf:"varint,1,req,name=major" json:"major,omitempty"`
+ Minor *int32 `protobuf:"varint,2,req,name=minor" json:"minor,omitempty"`
+ Gitid *string `protobuf:"bytes,3,opt,name=gitid" json:"gitid,omitempty"`
+ Sublevel *int32 `protobuf:"varint,4,opt,name=sublevel" json:"sublevel,omitempty"`
+ Extra *int32 `protobuf:"varint,5,opt,name=extra" json:"extra,omitempty"`
+ Name *string `protobuf:"bytes,6,opt,name=name" json:"name,omitempty"`
+ XXX_unrecognized []byte `json:"-"`
+}
+
+func (m *CriuVersion) Reset() { *m = CriuVersion{} }
+func (m *CriuVersion) String() string { return proto.CompactTextString(m) }
+func (*CriuVersion) ProtoMessage() {}
+func (*CriuVersion) Descriptor() ([]byte, []int) { return fileDescriptor0, []int{14} }
+
+func (m *CriuVersion) GetMajor() int32 {
+ if m != nil && m.Major != nil {
+ return *m.Major
+ }
+ return 0
+}
+
+func (m *CriuVersion) GetMinor() int32 {
+ if m != nil && m.Minor != nil {
+ return *m.Minor
+ }
+ return 0
+}
+
+func (m *CriuVersion) GetGitid() string {
+ if m != nil && m.Gitid != nil {
+ return *m.Gitid
+ }
+ return ""
+}
+
+func (m *CriuVersion) GetSublevel() int32 {
+ if m != nil && m.Sublevel != nil {
+ return *m.Sublevel
+ }
+ return 0
+}
+
+func (m *CriuVersion) GetExtra() int32 {
+ if m != nil && m.Extra != nil {
+ return *m.Extra
+ }
+ return 0
+}
+
+func (m *CriuVersion) GetName() string {
+ if m != nil && m.Name != nil {
+ return *m.Name
+ }
+ return ""
+}
+
+func init() {
+ proto.RegisterType((*CriuPageServerInfo)(nil), "criu_page_server_info")
+ proto.RegisterType((*CriuVethPair)(nil), "criu_veth_pair")
+ proto.RegisterType((*ExtMountMap)(nil), "ext_mount_map")
+ proto.RegisterType((*JoinNamespace)(nil), "join_namespace")
+ proto.RegisterType((*InheritFd)(nil), "inherit_fd")
+ proto.RegisterType((*CgroupRoot)(nil), "cgroup_root")
+ proto.RegisterType((*UnixSk)(nil), "unix_sk")
+ proto.RegisterType((*CriuOpts)(nil), "criu_opts")
+ proto.RegisterType((*CriuDumpResp)(nil), "criu_dump_resp")
+ proto.RegisterType((*CriuRestoreResp)(nil), "criu_restore_resp")
+ proto.RegisterType((*CriuNotify)(nil), "criu_notify")
+ proto.RegisterType((*CriuFeatures)(nil), "criu_features")
+ proto.RegisterType((*CriuReq)(nil), "criu_req")
+ proto.RegisterType((*CriuResp)(nil), "criu_resp")
+ proto.RegisterType((*CriuVersion)(nil), "criu_version")
+ proto.RegisterEnum("CriuCgMode", CriuCgMode_name, CriuCgMode_value)
+ proto.RegisterEnum("CriuReqType", CriuReqType_name, CriuReqType_value)
+}
+
+func init() { proto.RegisterFile("criurpc.proto", fileDescriptor0) }
+
+var fileDescriptor0 = []byte{
+ // 1781 bytes of a gzipped FileDescriptorProto
+ 0x1f, 0x8b, 0x08, 0x00, 0x00, 0x00, 0x00, 0x00, 0x02, 0xff, 0x8c, 0x56, 0xdd, 0x72, 0x5b, 0xb7,
+ 0x11, 0x0e, 0x29, 0xfe, 0x1c, 0x82, 0x3f, 0xa6, 0x10, 0xdb, 0x81, 0x93, 0xda, 0x62, 0xe8, 0x28,
+ 0x51, 0x15, 0x97, 0x4d, 0x58, 0x3b, 0xae, 0x33, 0xed, 0x85, 0x47, 0x22, 0x5d, 0x36, 0x92, 0xc8,
+ 0x01, 0x25, 0xcf, 0xe4, 0x0a, 0x73, 0x74, 0x0e, 0x48, 0xc1, 0x3c, 0x7f, 0x05, 0x40, 0x45, 0xf2,
+ 0x83, 0xf4, 0x29, 0xfa, 0x0c, 0x7d, 0x84, 0xbe, 0x4e, 0x6f, 0x3b, 0xbb, 0x00, 0x65, 0x29, 0xc9,
+ 0xb4, 0xbd, 0xc3, 0x7e, 0x58, 0x00, 0xbb, 0xfb, 0xed, 0x0f, 0x48, 0x3b, 0xd2, 0x6a, 0xad, 0x8b,
+ 0x68, 0x50, 0xe8, 0xdc, 0xe6, 0xfd, 0x25, 0x79, 0x00, 0x80, 0x28, 0xc2, 0xa5, 0x14, 0x46, 0xea,
+ 0x4b, 0xa9, 0x85, 0xca, 0x16, 0x39, 0x65, 0xa4, 0x1e, 0xc6, 0xb1, 0x96, 0xc6, 0xb0, 0x52, 0xaf,
+ 0xb4, 0xd7, 0xe0, 0x1b, 0x91, 0x52, 0x52, 0x29, 0x72, 0x6d, 0x59, 0xb9, 0x57, 0xda, 0xab, 0x72,
+ 0x5c, 0xd3, 0x2e, 0xd9, 0x2a, 0x54, 0xcc, 0xb6, 0x10, 0x82, 0x25, 0xed, 0x90, 0xf2, 0x22, 0x66,
+ 0x15, 0x04, 0xca, 0x8b, 0xb8, 0xff, 0x27, 0xd2, 0xc1, 0x87, 0x2e, 0xa5, 0xbd, 0x10, 0x45, 0xa8,
+ 0x34, 0xfd, 0x98, 0x54, 0xd5, 0x42, 0xa8, 0x8c, 0x95, 0x7a, 0xe5, 0xbd, 0x06, 0xaf, 0xa8, 0xc5,
+ 0x24, 0xa3, 0x0f, 0x48, 0x4d, 0x2d, 0x44, 0xbe, 0x86, 0xeb, 0x01, 0xad, 0xaa, 0xc5, 0x74, 0x6d,
+ 0xfb, 0x7f, 0x20, 0x6d, 0x79, 0x65, 0x45, 0x9a, 0xaf, 0x33, 0x2b, 0xd2, 0xb0, 0x80, 0x07, 0x57,
+ 0xf2, 0xda, 0x1f, 0x85, 0x25, 0x20, 0x97, 0x61, 0xe2, 0x8f, 0xc1, 0xb2, 0xff, 0x96, 0x74, 0xde,
+ 0xe5, 0x2a, 0x13, 0x59, 0x98, 0x4a, 0x53, 0x84, 0x91, 0x04, 0xa3, 0x32, 0xe3, 0x0f, 0x95, 0x33,
+ 0x43, 0x3f, 0x21, 0xf5, 0xcc, 0x88, 0x85, 0x4a, 0xa4, 0x3f, 0x57, 0xcb, 0xcc, 0x58, 0x25, 0x92,
+ 0x7e, 0x46, 0x1a, 0xf2, 0xca, 0xea, 0x50, 0xe4, 0x85, 0x45, 0xaf, 0x1a, 0x3c, 0x40, 0x60, 0x5a,
+ 0xd8, 0xfe, 0x80, 0x10, 0x95, 0x5d, 0x48, 0xad, 0xac, 0x58, 0xc4, 0xbf, 0x62, 0x89, 0x73, 0x1d,
+ 0x2e, 0x74, 0xae, 0xbf, 0x20, 0xcd, 0x68, 0xa9, 0xf3, 0x75, 0x21, 0x74, 0x9e, 0x5b, 0x88, 0x5f,
+ 0x64, 0x75, 0xe2, 0xc3, 0x8a, 0x6b, 0x8c, 0x69, 0x68, 0x2f, 0xbc, 0x15, 0xb8, 0xee, 0xef, 0x90,
+ 0xfa, 0x3a, 0x53, 0x57, 0xc2, 0xac, 0xe8, 0x7d, 0x52, 0x55, 0x59, 0x1e, 0x4b, 0x7c, 0xa5, 0xcd,
+ 0x9d, 0xd0, 0xff, 0x57, 0x9b, 0x34, 0x30, 0xa6, 0x79, 0x61, 0x0d, 0xed, 0x93, 0xb6, 0x4a, 0xc3,
+ 0xa5, 0x34, 0x22, 0x56, 0x5a, 0x2c, 0x62, 0xd4, 0xad, 0xf2, 0xa6, 0x03, 0x0f, 0x95, 0x1e, 0xc7,
+ 0x1b, 0x9a, 0xca, 0x1f, 0x68, 0x7a, 0x4a, 0xda, 0x89, 0x0c, 0x2f, 0xa5, 0xd0, 0xeb, 0x2c, 0x53,
+ 0xd9, 0x12, 0x9d, 0x0d, 0x78, 0x0b, 0x41, 0xee, 0x30, 0xfa, 0x84, 0x34, 0x21, 0xfa, 0xde, 0x1a,
+ 0x24, 0x35, 0xe0, 0x10, 0xa0, 0xb3, 0x4c, 0x5d, 0xcd, 0x57, 0xf4, 0x2b, 0x72, 0xcf, 0x46, 0x85,
+ 0x90, 0xc6, 0x86, 0xe7, 0x89, 0x32, 0x17, 0x32, 0x66, 0x55, 0xd4, 0xe9, 0xd8, 0xa8, 0x18, 0x7d,
+ 0x40, 0x41, 0x51, 0x5e, 0x86, 0x46, 0x5d, 0x4a, 0x11, 0xcb, 0x4b, 0x15, 0x49, 0xc3, 0x6a, 0x4e,
+ 0xd1, 0xc3, 0x87, 0x0e, 0x85, 0xf8, 0x9b, 0x0b, 0x99, 0x24, 0xe2, 0x5d, 0x7e, 0xce, 0xea, 0xa8,
+ 0x12, 0x20, 0xf0, 0xd7, 0xfc, 0x9c, 0x3e, 0x26, 0x04, 0x28, 0x13, 0x49, 0x1e, 0xad, 0x0c, 0x0b,
+ 0x9c, 0x35, 0x80, 0x1c, 0x01, 0x40, 0x9f, 0x90, 0x46, 0x92, 0x2f, 0x45, 0x22, 0x2f, 0x65, 0xc2,
+ 0x1a, 0xe0, 0xea, 0xf7, 0xa5, 0x21, 0x0f, 0x92, 0x7c, 0x79, 0x04, 0x10, 0x7d, 0x44, 0x60, 0xed,
+ 0x58, 0x27, 0x2e, 0xb5, 0x93, 0x7c, 0x89, 0xb4, 0x7f, 0x49, 0xca, 0x85, 0x61, 0xcd, 0x5e, 0x69,
+ 0xaf, 0x39, 0x7c, 0x38, 0xf8, 0xd5, 0xc2, 0xe0, 0xe5, 0xc2, 0xd0, 0x5d, 0xd2, 0xc9, 0x72, 0xab,
+ 0x16, 0xd7, 0xc2, 0x44, 0x5a, 0x15, 0xd6, 0xb0, 0x16, 0x5a, 0xd1, 0x76, 0xe8, 0xdc, 0x81, 0xc0,
+ 0x2a, 0x30, 0xce, 0xda, 0x8e, 0x69, 0x64, 0xff, 0x31, 0x21, 0x45, 0xa8, 0x65, 0x66, 0x85, 0x4a,
+ 0x97, 0xac, 0x83, 0x3b, 0x0d, 0x87, 0x4c, 0xd2, 0x25, 0x38, 0x6e, 0x75, 0x18, 0xad, 0x44, 0x2a,
+ 0x53, 0x76, 0xcf, 0x39, 0x8e, 0xc0, 0xb1, 0x4c, 0xe1, 0x6c, 0xb8, 0xb6, 0xb9, 0x88, 0x65, 0xbc,
+ 0x2e, 0x58, 0xd7, 0x39, 0x0e, 0xc8, 0x21, 0x00, 0x40, 0xd3, 0x4f, 0xb9, 0x5e, 0x6d, 0xf8, 0xdf,
+ 0x46, 0x96, 0x1b, 0x00, 0x39, 0xf6, 0x1f, 0x13, 0x92, 0xa8, 0x6c, 0x25, 0xb4, 0x4c, 0xc3, 0x82,
+ 0x51, 0x77, 0x1c, 0x10, 0x0e, 0x00, 0xdd, 0x25, 0x55, 0x28, 0x4e, 0xc3, 0x3e, 0xee, 0x6d, 0xed,
+ 0x35, 0x87, 0xf7, 0x06, 0x77, 0xeb, 0x95, 0xbb, 0x5d, 0xfa, 0x94, 0xd4, 0xa3, 0x62, 0x2d, 0xa2,
+ 0xb0, 0x60, 0xf7, 0x7b, 0xa5, 0xbd, 0xf6, 0xf7, 0xe4, 0xf9, 0xf0, 0xd5, 0xf3, 0x57, 0xdf, 0xbd,
+ 0x1c, 0xbe, 0x7a, 0xc1, 0x6b, 0x51, 0xb1, 0x3e, 0x08, 0x0b, 0xba, 0x43, 0x9a, 0x8b, 0x5c, 0x47,
+ 0x52, 0x28, 0x0d, 0x6f, 0x3d, 0xc0, 0xb7, 0x08, 0x42, 0x13, 0x40, 0x80, 0x04, 0x79, 0x25, 0x23,
+ 0x11, 0xa5, 0x31, 0x7b, 0xd8, 0xdb, 0x02, 0x12, 0x40, 0x3e, 0x48, 0x21, 0x49, 0xea, 0x58, 0xeb,
+ 0x99, 0x65, 0x9f, 0xa0, 0x25, 0x9d, 0xc1, 0x9d, 0xda, 0xe7, 0x35, 0x79, 0x65, 0x8f, 0x33, 0x0b,
+ 0x2c, 0xa4, 0x61, 0x06, 0xfc, 0xb8, 0xf2, 0x32, 0x8c, 0x39, 0x16, 0x1c, 0x7a, 0xe0, 0x40, 0xba,
+ 0x4b, 0xea, 0xd1, 0x12, 0x4b, 0x8f, 0x3d, 0xc2, 0xfb, 0x5a, 0x83, 0x5b, 0xe5, 0xc8, 0x6b, 0xd1,
+ 0x92, 0x03, 0x31, 0x3b, 0xa4, 0xa9, 0x8d, 0x15, 0x46, 0x9d, 0x27, 0x50, 0x07, 0x9f, 0x3a, 0x93,
+ 0xb5, 0xb1, 0x73, 0x87, 0xd0, 0xfd, 0xdb, 0x65, 0xcf, 0x3e, 0xc3, 0xab, 0x9a, 0x83, 0x0f, 0x10,
+ 0x6f, 0xf8, 0xf5, 0x38, 0xa6, 0x3d, 0xd2, 0x42, 0xa6, 0x36, 0x8e, 0xfc, 0xc6, 0xdd, 0x06, 0xd8,
+ 0xc8, 0x19, 0xbf, 0xe3, 0x6a, 0xca, 0x5c, 0x84, 0x1a, 0x9e, 0x7b, 0xec, 0x14, 0xe4, 0x95, 0x9d,
+ 0x3b, 0x64, 0xa3, 0x90, 0x86, 0xc6, 0x4a, 0x6d, 0xd8, 0x93, 0x1b, 0x85, 0x63, 0x87, 0x40, 0x08,
+ 0xcd, 0x4a, 0x15, 0x78, 0xff, 0x8e, 0x0b, 0x21, 0xc8, 0x70, 0x39, 0xb4, 0xaf, 0x2c, 0x3c, 0x4f,
+ 0xa4, 0x58, 0x18, 0xd6, 0xc3, 0xbd, 0xc0, 0x01, 0x63, 0x43, 0xf7, 0x48, 0xd3, 0x57, 0xb2, 0x50,
+ 0x59, 0xce, 0x3e, 0x47, 0x47, 0x82, 0x81, 0xc7, 0x78, 0x63, 0x8d, 0x45, 0x3d, 0xc9, 0x72, 0xfa,
+ 0x67, 0xf2, 0xf1, 0xdd, 0x00, 0x8b, 0x14, 0x9a, 0x50, 0xbf, 0x57, 0xda, 0xeb, 0x0c, 0xdb, 0x2e,
+ 0x3f, 0xa2, 0x25, 0x82, 0x7c, 0xfb, 0x4e, 0xd0, 0x8f, 0xf3, 0x58, 0xc2, 0x43, 0xcb, 0x8b, 0xdc,
+ 0x58, 0x91, 0xa8, 0x54, 0x59, 0xf6, 0x14, 0xb3, 0xa5, 0xfe, 0xed, 0x37, 0xcf, 0xff, 0xf8, 0xe2,
+ 0xe5, 0x77, 0x9c, 0xe0, 0xde, 0x11, 0x6c, 0xd1, 0x3d, 0xd2, 0xc5, 0x44, 0x11, 0x26, 0x0a, 0x33,
+ 0x01, 0xdd, 0xcf, 0xb0, 0x2f, 0xd0, 0xec, 0x0e, 0xe2, 0xf3, 0x28, 0xcc, 0x66, 0x80, 0xd2, 0x4f,
+ 0x21, 0x6f, 0xac, 0xd4, 0x59, 0x98, 0xb0, 0x5d, 0xef, 0x98, 0x97, 0x31, 0xa7, 0xd2, 0xc2, 0x5e,
+ 0x8b, 0xcc, 0xb0, 0x2f, 0xe1, 0x31, 0x5e, 0x47, 0xf9, 0x04, 0x7c, 0xae, 0xbb, 0x51, 0x60, 0xd8,
+ 0x57, 0x3e, 0xbb, 0xef, 0x8e, 0x06, 0x5e, 0x03, 0xf9, 0xc4, 0xd0, 0xcf, 0x49, 0xcb, 0x67, 0x47,
+ 0xa1, 0xf3, 0xc2, 0xb0, 0xdf, 0x62, 0x85, 0xfa, 0x06, 0x3e, 0x03, 0x88, 0xee, 0x93, 0xed, 0xdb,
+ 0x2a, 0xae, 0x93, 0xec, 0xa3, 0xde, 0xbd, 0x5b, 0x7a, 0xd8, 0x51, 0x9e, 0x93, 0x87, 0x5e, 0x37,
+ 0x5e, 0xa7, 0x85, 0x88, 0xf2, 0xcc, 0xea, 0x3c, 0x49, 0xa4, 0x66, 0x5f, 0xa3, 0xf5, 0xf7, 0xdd,
+ 0xee, 0xe1, 0x3a, 0x2d, 0x0e, 0x6e, 0xf6, 0xa0, 0x2b, 0x2f, 0xb4, 0x94, 0xef, 0x37, 0x81, 0x67,
+ 0xcf, 0xf0, 0xf6, 0x96, 0x03, 0x5d, 0x8c, 0x61, 0x42, 0x5b, 0x95, 0x4a, 0x98, 0x95, 0xbf, 0x73,
+ 0xde, 0x7a, 0x91, 0x7e, 0x4d, 0x28, 0xf4, 0x63, 0xcc, 0x0e, 0x95, 0x89, 0x45, 0xa2, 0x96, 0x17,
+ 0x96, 0x0d, 0x30, 0x83, 0xa0, 0x53, 0xcf, 0x57, 0xaa, 0x98, 0x64, 0x63, 0x84, 0xc1, 0xe1, 0x9f,
+ 0x64, 0xb8, 0x12, 0xe6, 0xda, 0x44, 0x36, 0x31, 0xec, 0xf7, 0xa8, 0xd6, 0x04, 0x6c, 0xee, 0x20,
+ 0x6c, 0x1c, 0xe1, 0xfb, 0x6b, 0xec, 0x85, 0x86, 0x7d, 0xe3, 0x1b, 0x47, 0xf8, 0xfe, 0x7a, 0x06,
+ 0x00, 0x36, 0x6b, 0x1b, 0xda, 0xb5, 0x81, 0xba, 0xf8, 0x16, 0xbb, 0x4e, 0xe0, 0x80, 0x71, 0x0c,
+ 0xc1, 0xca, 0x75, 0x71, 0x01, 0xb4, 0x5a, 0xe3, 0xb3, 0x99, 0x0d, 0x9d, 0x29, 0x6e, 0x63, 0x66,
+ 0x8d, 0x4b, 0xe9, 0xfe, 0x33, 0xff, 0x47, 0xc0, 0x50, 0x69, 0x69, 0x0a, 0xa0, 0x5b, 0x4b, 0x63,
+ 0x73, 0x2d, 0x63, 0x9c, 0x97, 0x01, 0xbf, 0x91, 0xfb, 0xbb, 0x64, 0x1b, 0xb5, 0x3d, 0xe0, 0x0e,
+ 0xf8, 0x09, 0xe7, 0x66, 0x1f, 0x2c, 0xfb, 0x2f, 0x49, 0x13, 0xd5, 0x5c, 0x6b, 0xa6, 0x0f, 0x49,
+ 0xcd, 0xf5, 0x6c, 0x3f, 0x7f, 0xbd, 0xf4, 0xcb, 0xd1, 0xd8, 0xff, 0xc1, 0xfd, 0x95, 0xc4, 0x42,
+ 0x86, 0x76, 0xad, 0x9d, 0x9f, 0xa9, 0x4c, 0x05, 0xb6, 0xe3, 0x8d, 0x35, 0xa9, 0x4c, 0x4f, 0x41,
+ 0xfe, 0x59, 0x8c, 0xca, 0x3f, 0x8b, 0x51, 0xff, 0x9f, 0x25, 0x12, 0x78, 0x6b, 0xff, 0x46, 0xfb,
+ 0xa4, 0x62, 0xaf, 0x0b, 0x37, 0xcd, 0x3b, 0xc3, 0xce, 0x60, 0xb3, 0x21, 0x00, 0xe5, 0xb8, 0x47,
+ 0x9f, 0x90, 0x0a, 0x8c, 0x75, 0xbc, 0xa9, 0x39, 0x24, 0x83, 0x9b, 0x41, 0xcf, 0x11, 0xbf, 0x3d,
+ 0x82, 0xd6, 0x51, 0x04, 0xdf, 0xb4, 0xad, 0x3b, 0x23, 0xc8, 0x81, 0x60, 0xf3, 0x4a, 0xca, 0x42,
+ 0xe4, 0x85, 0xcc, 0xfc, 0xe0, 0x0e, 0x00, 0x98, 0x16, 0x32, 0xa3, 0xfb, 0x24, 0xd8, 0x38, 0x87,
+ 0x03, 0xbb, 0xb9, 0xb1, 0x65, 0x83, 0xf2, 0x9b, 0xfd, 0xfe, 0xbf, 0xcb, 0xfe, 0xb3, 0x81, 0x61,
+ 0xfe, 0x7f, 0x3c, 0x60, 0xa4, 0xbe, 0x31, 0x0d, 0xbe, 0x35, 0x01, 0xdf, 0x88, 0xf4, 0x29, 0xa9,
+ 0x00, 0xc5, 0x68, 0xf1, 0xcd, 0xa0, 0xb9, 0x21, 0x9d, 0xe3, 0x26, 0x7d, 0x46, 0xea, 0x9e, 0x59,
+ 0xb4, 0xbb, 0x39, 0xa4, 0x83, 0x5f, 0xd0, 0xcd, 0x37, 0x2a, 0xf4, 0x0b, 0x52, 0x73, 0x8e, 0x7b,
+ 0x47, 0x5a, 0x83, 0x5b, 0xa4, 0x73, 0xbf, 0xe7, 0xe7, 0x7b, 0xed, 0x7f, 0xce, 0xf7, 0x47, 0x40,
+ 0x96, 0x90, 0x5a, 0x67, 0x39, 0xfe, 0x3e, 0xaa, 0xbc, 0x1e, 0xe9, 0x11, 0x88, 0x77, 0x62, 0x16,
+ 0xfc, 0xf7, 0x98, 0x41, 0xf0, 0xdd, 0x35, 0xa9, 0x59, 0xe2, 0x4f, 0xa4, 0xc1, 0x03, 0xbc, 0x27,
+ 0x35, 0x4b, 0x18, 0x73, 0x97, 0x52, 0x1b, 0x95, 0x67, 0xf8, 0x0b, 0x69, 0x6e, 0x1a, 0xaa, 0x07,
+ 0xf9, 0x66, 0xb7, 0xff, 0xf7, 0x12, 0x69, 0xdd, 0xde, 0x81, 0xdf, 0x60, 0x1a, 0xbe, 0xcb, 0xb5,
+ 0xcf, 0x72, 0x27, 0x20, 0xaa, 0xb2, 0x5c, 0xfb, 0x8f, 0xa7, 0x13, 0x00, 0x5d, 0x2a, 0xeb, 0xbf,
+ 0xe6, 0x0d, 0xee, 0x04, 0x28, 0x2b, 0xb3, 0x3e, 0x77, 0x3f, 0xa4, 0x8a, 0x2f, 0x58, 0x2f, 0xc3,
+ 0x09, 0xfc, 0xe9, 0x62, 0x20, 0xab, 0xdc, 0x09, 0xf0, 0x95, 0x81, 0x5e, 0x89, 0xb1, 0x6b, 0x70,
+ 0x5c, 0xef, 0x0b, 0x6f, 0x97, 0x1f, 0x01, 0x94, 0x90, 0xda, 0xe4, 0xcd, 0xc9, 0x94, 0x8f, 0xba,
+ 0x1f, 0xd1, 0x26, 0xa9, 0x1f, 0xbc, 0x11, 0x27, 0xd3, 0x93, 0x51, 0xb7, 0x44, 0x1b, 0xa4, 0x3a,
+ 0xe3, 0xd3, 0xd9, 0xbc, 0x5b, 0xa6, 0x01, 0xa9, 0xcc, 0xa7, 0xe3, 0xd3, 0xee, 0x16, 0xac, 0xc6,
+ 0x67, 0x47, 0x47, 0xdd, 0x0a, 0x9c, 0x9b, 0x9f, 0xf2, 0xc9, 0xc1, 0x69, 0xb7, 0x0a, 0xe7, 0x0e,
+ 0x47, 0xe3, 0xd7, 0x67, 0x47, 0xa7, 0xdd, 0xda, 0xfe, 0x3f, 0x4a, 0xbe, 0x04, 0x37, 0x99, 0x05,
+ 0x37, 0x8d, 0x8e, 0x67, 0xa7, 0x3f, 0x76, 0x3f, 0x82, 0xf3, 0x87, 0x67, 0xc7, 0xb3, 0x6e, 0x09,
+ 0xce, 0xf0, 0xd1, 0xfc, 0x14, 0x1e, 0x2e, 0x83, 0xc6, 0xc1, 0x5f, 0x46, 0x07, 0x3f, 0x74, 0xb7,
+ 0x68, 0x8b, 0x04, 0x33, 0x3e, 0x12, 0xa8, 0x55, 0xa1, 0xf7, 0x48, 0x73, 0xf6, 0xfa, 0xcd, 0x48,
+ 0xcc, 0x47, 0xfc, 0xed, 0x88, 0x77, 0xab, 0xf0, 0xec, 0xc9, 0xf4, 0x74, 0x32, 0xfe, 0xb1, 0x5b,
+ 0xa3, 0x5d, 0xd2, 0x3a, 0x98, 0x9d, 0x4d, 0x4e, 0xc6, 0x53, 0xa7, 0x5e, 0xa7, 0xdb, 0xa4, 0xbd,
+ 0x41, 0xdc, 0x7d, 0x01, 0x40, 0xe3, 0xd1, 0xeb, 0xd3, 0x33, 0x3e, 0xf2, 0x50, 0x03, 0x9e, 0x7e,
+ 0x3b, 0xe2, 0xf3, 0xc9, 0xf4, 0xa4, 0x4b, 0xfe, 0x13, 0x00, 0x00, 0xff, 0xff, 0x5f, 0x2a, 0xaf,
+ 0x49, 0x5b, 0x0d, 0x00, 0x00,
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.proto b/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.proto
new file mode 100644
index 000000000..48e42e26e
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/criurpc/criurpc.proto
@@ -0,0 +1,209 @@
+syntax = "proto2";
+
+message criu_page_server_info {
+ optional string address = 1;
+ optional int32 port = 2;
+ optional int32 pid = 3;
+ optional int32 fd = 4;
+}
+
+message criu_veth_pair {
+ required string if_in = 1;
+ required string if_out = 2;
+};
+
+message ext_mount_map {
+ required string key = 1;
+ required string val = 2;
+};
+
+message join_namespace {
+ required string ns = 1;
+ required string ns_file = 2;
+ optional string extra_opt = 3;
+}
+
+message inherit_fd {
+ required string key = 1;
+ required int32 fd = 2;
+};
+
+message cgroup_root {
+ optional string ctrl = 1;
+ required string path = 2;
+};
+
+message unix_sk {
+ required uint32 inode = 1;
+};
+
+enum criu_cg_mode {
+ IGNORE = 0;
+ CG_NONE = 1;
+ PROPS = 2;
+ SOFT = 3;
+ FULL = 4;
+ STRICT = 5;
+ DEFAULT = 6;
+};
+
+message criu_opts {
+ required int32 images_dir_fd = 1;
+ optional int32 pid = 2; /* if not set on dump, will dump requesting process */
+
+ optional bool leave_running = 3;
+ optional bool ext_unix_sk = 4;
+ optional bool tcp_established = 5;
+ optional bool evasive_devices = 6;
+ optional bool shell_job = 7;
+ optional bool file_locks = 8;
+ optional int32 log_level = 9 [default = 2];
+ optional string log_file = 10; /* No subdirs are allowed. Consider using work-dir */
+
+ optional criu_page_server_info ps = 11;
+
+ optional bool notify_scripts = 12;
+
+ optional string root = 13;
+ optional string parent_img = 14;
+ optional bool track_mem = 15;
+ optional bool auto_dedup = 16;
+
+ optional int32 work_dir_fd = 17;
+ optional bool link_remap = 18;
+ repeated criu_veth_pair veths = 19; /* DEPRECATED, use external instead */
+
+ optional uint32 cpu_cap = 20 [default = 0xffffffff];
+ optional bool force_irmap = 21;
+ repeated string exec_cmd = 22;
+
+ repeated ext_mount_map ext_mnt = 23; /* DEPRECATED, use external instead */
+ optional bool manage_cgroups = 24; /* backward compatibility */
+ repeated cgroup_root cg_root = 25;
+
+ optional bool rst_sibling = 26; /* swrk only */
+ repeated inherit_fd inherit_fd = 27; /* swrk only */
+
+ optional bool auto_ext_mnt = 28;
+ optional bool ext_sharing = 29;
+ optional bool ext_masters = 30;
+
+ repeated string skip_mnt = 31;
+ repeated string enable_fs = 32;
+
+ repeated unix_sk unix_sk_ino = 33; /* DEPRECATED, use external instead */
+
+ optional criu_cg_mode manage_cgroups_mode = 34;
+ optional uint32 ghost_limit = 35 [default = 0x100000];
+ repeated string irmap_scan_paths = 36;
+ repeated string external = 37;
+ optional uint32 empty_ns = 38;
+ repeated join_namespace join_ns = 39;
+
+ optional string cgroup_props = 41;
+ optional string cgroup_props_file = 42;
+ repeated string cgroup_dump_controller = 43;
+
+ optional string freeze_cgroup = 44;
+ optional uint32 timeout = 45;
+ optional bool tcp_skip_in_flight = 46;
+ optional bool weak_sysctls = 47;
+ optional bool lazy_pages = 48;
+ optional int32 status_fd = 49;
+ optional bool orphan_pts_master = 50;
+}
+
+message criu_dump_resp {
+ optional bool restored = 1;
+}
+
+message criu_restore_resp {
+ required int32 pid = 1;
+}
+
+message criu_notify {
+ optional string script = 1;
+ optional int32 pid = 2;
+}
+
+enum criu_req_type {
+ EMPTY = 0;
+ DUMP = 1;
+ RESTORE = 2;
+ CHECK = 3;
+ PRE_DUMP = 4;
+ PAGE_SERVER = 5;
+
+ NOTIFY = 6;
+
+ CPUINFO_DUMP = 7;
+ CPUINFO_CHECK = 8;
+
+ FEATURE_CHECK = 9;
+
+ VERSION = 10;
+}
+
+/*
+ * List of features which can queried via
+ * CRIU_REQ_TYPE__FEATURE_CHECK
+ */
+message criu_features {
+ optional bool mem_track = 1;
+ optional bool lazy_pages = 2;
+}
+
+/*
+ * Request -- each type corresponds to must-be-there
+ * request arguments of respective type
+ */
+
+message criu_req {
+ required criu_req_type type = 1;
+
+ optional criu_opts opts = 2;
+ optional bool notify_success = 3;
+
+ /*
+ * When set service won't close the connection but
+ * will wait for more req-s to appear. Works not
+ * for all request types.
+ */
+ optional bool keep_open = 4;
+ /*
+ * 'features' can be used to query which features
+ * are supported by the installed criu/kernel
+ * via RPC.
+ */
+ optional criu_features features = 5;
+}
+
+/*
+ * Response -- it states whether the request was served
+ * and additional request-specific information
+ */
+
+message criu_resp {
+ required criu_req_type type = 1;
+ required bool success = 2;
+
+ optional criu_dump_resp dump = 3;
+ optional criu_restore_resp restore = 4;
+ optional criu_notify notify = 5;
+ optional criu_page_server_info ps = 6;
+
+ optional int32 cr_errno = 7;
+ optional criu_features features = 8;
+ optional string cr_errmsg = 9;
+ optional criu_version version = 10;
+}
+
+/* Answer for criu_req_type.VERSION requests */
+message criu_version {
+ required int32 major = 1;
+ required int32 minor = 2;
+ optional string gitid = 3;
+ optional int32 sublevel = 4;
+ optional int32 extra = 5;
+ optional string name = 6;
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_linux.go
new file mode 100644
index 000000000..461dc097c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_linux.go
@@ -0,0 +1,100 @@
+package devices
+
+import (
+ "errors"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+
+ "golang.org/x/sys/unix"
+)
+
+var (
+ ErrNotADevice = errors.New("not a device node")
+)
+
+// Testing dependencies
+var (
+ unixLstat = unix.Lstat
+ ioutilReadDir = ioutil.ReadDir
+)
+
+// Given the path to a device and its cgroup_permissions(which cannot be easily queried) look up the information about a linux device and return that information as a Device struct.
+func DeviceFromPath(path, permissions string) (*configs.Device, error) {
+ var stat unix.Stat_t
+ err := unixLstat(path, &stat)
+ if err != nil {
+ return nil, err
+ }
+ var (
+ devType rune
+ mode = stat.Mode
+ )
+ switch {
+ case mode&unix.S_IFBLK == unix.S_IFBLK:
+ devType = 'b'
+ case mode&unix.S_IFCHR == unix.S_IFCHR:
+ devType = 'c'
+ default:
+ return nil, ErrNotADevice
+ }
+ devNumber := int(stat.Rdev)
+ uid := stat.Uid
+ gid := stat.Gid
+ return &configs.Device{
+ Type: devType,
+ Path: path,
+ Major: Major(devNumber),
+ Minor: Minor(devNumber),
+ Permissions: permissions,
+ FileMode: os.FileMode(mode),
+ Uid: uid,
+ Gid: gid,
+ }, nil
+}
+
+func HostDevices() ([]*configs.Device, error) {
+ return getDevices("/dev")
+}
+
+func getDevices(path string) ([]*configs.Device, error) {
+ files, err := ioutilReadDir(path)
+ if err != nil {
+ return nil, err
+ }
+ out := []*configs.Device{}
+ for _, f := range files {
+ switch {
+ case f.IsDir():
+ switch f.Name() {
+ // ".lxc" & ".lxd-mounts" added to address https://github.com/lxc/lxd/issues/2825
+ case "pts", "shm", "fd", "mqueue", ".lxc", ".lxd-mounts":
+ continue
+ default:
+ sub, err := getDevices(filepath.Join(path, f.Name()))
+ if err != nil {
+ return nil, err
+ }
+
+ out = append(out, sub...)
+ continue
+ }
+ case f.Name() == "console":
+ continue
+ }
+ device, err := DeviceFromPath(filepath.Join(path, f.Name()), "rwm")
+ if err != nil {
+ if err == ErrNotADevice {
+ continue
+ }
+ if os.IsNotExist(err) {
+ continue
+ }
+ return nil, err
+ }
+ out = append(out, device)
+ }
+ return out, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_unsupported.go
new file mode 100644
index 000000000..6649b9f2d
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/devices/devices_unsupported.go
@@ -0,0 +1,3 @@
+// +build !linux
+
+package devices
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/devices/number.go b/vendor/github.com/opencontainers/runc/libcontainer/devices/number.go
new file mode 100644
index 000000000..885b6e5dd
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/devices/number.go
@@ -0,0 +1,24 @@
+// +build linux freebsd
+
+package devices
+
+/*
+
+This code provides support for manipulating linux device numbers. It should be replaced by normal syscall functions once http://code.google.com/p/go/issues/detail?id=8106 is solved.
+
+You can read what they are here:
+
+ - http://www.makelinux.net/ldd3/chp-3-sect-2
+ - http://www.linux-tutorial.info/modules.php?name=MContent&pageid=94
+
+Note! These are NOT the same as the MAJOR(dev_t device);, MINOR(dev_t device); and MKDEV(int major, int minor); functions as defined in <linux/kdev_t.h> as the representation of device numbers used by go is different than the one used internally to the kernel! - https://github.com/torvalds/linux/blob/master/include/linux/kdev_t.h#L9
+
+*/
+
+func Major(devNumber int) int64 {
+ return int64((devNumber >> 8) & 0xfff)
+}
+
+func Minor(devNumber int) int64 {
+ return int64((devNumber & 0xff) | ((devNumber >> 12) & 0xfff00))
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/error.go b/vendor/github.com/opencontainers/runc/libcontainer/error.go
new file mode 100644
index 000000000..21a3789ba
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/error.go
@@ -0,0 +1,70 @@
+package libcontainer
+
+import "io"
+
+// ErrorCode is the API error code type.
+type ErrorCode int
+
+// API error codes.
+const (
+ // Factory errors
+ IdInUse ErrorCode = iota
+ InvalidIdFormat
+
+ // Container errors
+ ContainerNotExists
+ ContainerPaused
+ ContainerNotStopped
+ ContainerNotRunning
+ ContainerNotPaused
+
+ // Process errors
+ NoProcessOps
+
+ // Common errors
+ ConfigInvalid
+ ConsoleExists
+ SystemError
+)
+
+func (c ErrorCode) String() string {
+ switch c {
+ case IdInUse:
+ return "Id already in use"
+ case InvalidIdFormat:
+ return "Invalid format"
+ case ContainerPaused:
+ return "Container paused"
+ case ConfigInvalid:
+ return "Invalid configuration"
+ case SystemError:
+ return "System error"
+ case ContainerNotExists:
+ return "Container does not exist"
+ case ContainerNotStopped:
+ return "Container is not stopped"
+ case ContainerNotRunning:
+ return "Container is not running"
+ case ConsoleExists:
+ return "Console exists for process"
+ case ContainerNotPaused:
+ return "Container is not paused"
+ case NoProcessOps:
+ return "No process operations"
+ default:
+ return "Unknown error"
+ }
+}
+
+// Error is the API error type.
+type Error interface {
+ error
+
+ // Returns an error if it failed to write the detail of the Error to w.
+ // The detail of the Error may include the error message and a
+ // representation of the stack trace.
+ Detail(w io.Writer) error
+
+ // Returns the error code for this error.
+ Code() ErrorCode
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/factory.go b/vendor/github.com/opencontainers/runc/libcontainer/factory.go
new file mode 100644
index 000000000..0986cd77e
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/factory.go
@@ -0,0 +1,44 @@
+package libcontainer
+
+import (
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type Factory interface {
+ // Creates a new container with the given id and starts the initial process inside it.
+ // id must be a string containing only letters, digits and underscores and must contain
+ // between 1 and 1024 characters, inclusive.
+ //
+ // The id must not already be in use by an existing container. Containers created using
+ // a factory with the same path (and filesystem) must have distinct ids.
+ //
+ // Returns the new container with a running process.
+ //
+ // errors:
+ // IdInUse - id is already in use by a container
+ // InvalidIdFormat - id has incorrect format
+ // ConfigInvalid - config is invalid
+ // Systemerror - System error
+ //
+ // On error, any partially created container parts are cleaned up (the operation is atomic).
+ Create(id string, config *configs.Config) (Container, error)
+
+ // Load takes an ID for an existing container and returns the container information
+ // from the state. This presents a read only view of the container.
+ //
+ // errors:
+ // Path does not exist
+ // System error
+ Load(id string) (Container, error)
+
+ // StartInitialization is an internal API to libcontainer used during the reexec of the
+ // container.
+ //
+ // Errors:
+ // Pipe connection error
+ // System error
+ StartInitialization() error
+
+ // Type returns info string about factory type (e.g. lxc, libcontainer...)
+ Type() string
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/factory_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/factory_linux.go
new file mode 100644
index 000000000..42b6f5a05
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/factory_linux.go
@@ -0,0 +1,325 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "encoding/json"
+ "fmt"
+ "os"
+ "path/filepath"
+ "regexp"
+ "runtime/debug"
+ "strconv"
+
+ "github.com/docker/docker/pkg/mount"
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/cgroups/fs"
+ "github.com/opencontainers/runc/libcontainer/cgroups/rootless"
+ "github.com/opencontainers/runc/libcontainer/cgroups/systemd"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/configs/validate"
+ "github.com/opencontainers/runc/libcontainer/utils"
+
+ "golang.org/x/sys/unix"
+)
+
+const (
+ stateFilename = "state.json"
+ execFifoFilename = "exec.fifo"
+)
+
+var idRegex = regexp.MustCompile(`^[\w+-\.]+$`)
+
+// InitArgs returns an options func to configure a LinuxFactory with the
+// provided init binary path and arguments.
+func InitArgs(args ...string) func(*LinuxFactory) error {
+ return func(l *LinuxFactory) (err error) {
+ if len(args) > 0 {
+ // Resolve relative paths to ensure that its available
+ // after directory changes.
+ if args[0], err = filepath.Abs(args[0]); err != nil {
+ return newGenericError(err, ConfigInvalid)
+ }
+ }
+
+ l.InitArgs = args
+ return nil
+ }
+}
+
+// SystemdCgroups is an options func to configure a LinuxFactory to return
+// containers that use systemd to create and manage cgroups.
+func SystemdCgroups(l *LinuxFactory) error {
+ l.NewCgroupsManager = func(config *configs.Cgroup, paths map[string]string) cgroups.Manager {
+ return &systemd.Manager{
+ Cgroups: config,
+ Paths: paths,
+ }
+ }
+ return nil
+}
+
+// Cgroupfs is an options func to configure a LinuxFactory to return
+// containers that use the native cgroups filesystem implementation to
+// create and manage cgroups.
+func Cgroupfs(l *LinuxFactory) error {
+ l.NewCgroupsManager = func(config *configs.Cgroup, paths map[string]string) cgroups.Manager {
+ return &fs.Manager{
+ Cgroups: config,
+ Paths: paths,
+ }
+ }
+ return nil
+}
+
+// RootlessCgroups is an options func to configure a LinuxFactory to
+// return containers that use the "rootless" cgroup manager, which will
+// fail to do any operations not possible to do with an unprivileged user.
+// It should only be used in conjunction with rootless containers.
+func RootlessCgroups(l *LinuxFactory) error {
+ l.NewCgroupsManager = func(config *configs.Cgroup, paths map[string]string) cgroups.Manager {
+ return &rootless.Manager{
+ Cgroups: config,
+ Paths: paths,
+ }
+ }
+ return nil
+}
+
+// TmpfsRoot is an option func to mount LinuxFactory.Root to tmpfs.
+func TmpfsRoot(l *LinuxFactory) error {
+ mounted, err := mount.Mounted(l.Root)
+ if err != nil {
+ return err
+ }
+ if !mounted {
+ if err := unix.Mount("tmpfs", l.Root, "tmpfs", 0, ""); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// CriuPath returns an option func to configure a LinuxFactory with the
+// provided criupath
+func CriuPath(criupath string) func(*LinuxFactory) error {
+ return func(l *LinuxFactory) error {
+ l.CriuPath = criupath
+ return nil
+ }
+}
+
+// New returns a linux based container factory based in the root directory and
+// configures the factory with the provided option funcs.
+func New(root string, options ...func(*LinuxFactory) error) (Factory, error) {
+ if root != "" {
+ if err := os.MkdirAll(root, 0700); err != nil {
+ return nil, newGenericError(err, SystemError)
+ }
+ }
+ l := &LinuxFactory{
+ Root: root,
+ InitArgs: []string{"/proc/self/exe", "init"},
+ Validator: validate.New(),
+ CriuPath: "criu",
+ }
+ Cgroupfs(l)
+ for _, opt := range options {
+ if err := opt(l); err != nil {
+ return nil, err
+ }
+ }
+ return l, nil
+}
+
+// LinuxFactory implements the default factory interface for linux based systems.
+type LinuxFactory struct {
+ // Root directory for the factory to store state.
+ Root string
+
+ // InitArgs are arguments for calling the init responsibilities for spawning
+ // a container.
+ InitArgs []string
+
+ // CriuPath is the path to the criu binary used for checkpoint and restore of
+ // containers.
+ CriuPath string
+
+ // Validator provides validation to container configurations.
+ Validator validate.Validator
+
+ // NewCgroupsManager returns an initialized cgroups manager for a single container.
+ NewCgroupsManager func(config *configs.Cgroup, paths map[string]string) cgroups.Manager
+}
+
+func (l *LinuxFactory) Create(id string, config *configs.Config) (Container, error) {
+ if l.Root == "" {
+ return nil, newGenericError(fmt.Errorf("invalid root"), ConfigInvalid)
+ }
+ if err := l.validateID(id); err != nil {
+ return nil, err
+ }
+ if err := l.Validator.Validate(config); err != nil {
+ return nil, newGenericError(err, ConfigInvalid)
+ }
+ containerRoot := filepath.Join(l.Root, id)
+ if _, err := os.Stat(containerRoot); err == nil {
+ return nil, newGenericError(fmt.Errorf("container with id exists: %v", id), IdInUse)
+ } else if !os.IsNotExist(err) {
+ return nil, newGenericError(err, SystemError)
+ }
+ if err := os.MkdirAll(containerRoot, 0711); err != nil {
+ return nil, newGenericError(err, SystemError)
+ }
+ if err := os.Chown(containerRoot, unix.Geteuid(), unix.Getegid()); err != nil {
+ return nil, newGenericError(err, SystemError)
+ }
+ if config.Rootless {
+ RootlessCgroups(l)
+ }
+ c := &linuxContainer{
+ id: id,
+ root: containerRoot,
+ config: config,
+ initArgs: l.InitArgs,
+ criuPath: l.CriuPath,
+ cgroupManager: l.NewCgroupsManager(config.Cgroups, nil),
+ }
+ c.state = &stoppedState{c: c}
+ return c, nil
+}
+
+func (l *LinuxFactory) Load(id string) (Container, error) {
+ if l.Root == "" {
+ return nil, newGenericError(fmt.Errorf("invalid root"), ConfigInvalid)
+ }
+ containerRoot := filepath.Join(l.Root, id)
+ state, err := l.loadState(containerRoot, id)
+ if err != nil {
+ return nil, err
+ }
+ r := &nonChildProcess{
+ processPid: state.InitProcessPid,
+ processStartTime: state.InitProcessStartTime,
+ fds: state.ExternalDescriptors,
+ }
+ // We have to use the RootlessManager.
+ if state.Rootless {
+ RootlessCgroups(l)
+ }
+ c := &linuxContainer{
+ initProcess: r,
+ initProcessStartTime: state.InitProcessStartTime,
+ id: id,
+ config: &state.Config,
+ initArgs: l.InitArgs,
+ criuPath: l.CriuPath,
+ cgroupManager: l.NewCgroupsManager(state.Config.Cgroups, state.CgroupPaths),
+ root: containerRoot,
+ created: state.Created,
+ }
+ c.state = &loadedState{c: c}
+ if err := c.refreshState(); err != nil {
+ return nil, err
+ }
+ return c, nil
+}
+
+func (l *LinuxFactory) Type() string {
+ return "libcontainer"
+}
+
+// StartInitialization loads a container by opening the pipe fd from the parent to read the configuration and state
+// This is a low level implementation detail of the reexec and should not be consumed externally
+func (l *LinuxFactory) StartInitialization() (err error) {
+ var (
+ pipefd, rootfd int
+ consoleSocket *os.File
+ envInitPipe = os.Getenv("_LIBCONTAINER_INITPIPE")
+ envStateDir = os.Getenv("_LIBCONTAINER_STATEDIR")
+ envConsole = os.Getenv("_LIBCONTAINER_CONSOLE")
+ )
+
+ // Get the INITPIPE.
+ pipefd, err = strconv.Atoi(envInitPipe)
+ if err != nil {
+ return fmt.Errorf("unable to convert _LIBCONTAINER_INITPIPE=%s to int: %s", envInitPipe, err)
+ }
+
+ var (
+ pipe = os.NewFile(uintptr(pipefd), "pipe")
+ it = initType(os.Getenv("_LIBCONTAINER_INITTYPE"))
+ )
+ defer pipe.Close()
+
+ // Only init processes have STATEDIR.
+ rootfd = -1
+ if it == initStandard {
+ if rootfd, err = strconv.Atoi(envStateDir); err != nil {
+ return fmt.Errorf("unable to convert _LIBCONTAINER_STATEDIR=%s to int: %s", envStateDir, err)
+ }
+ }
+
+ if envConsole != "" {
+ console, err := strconv.Atoi(envConsole)
+ if err != nil {
+ return fmt.Errorf("unable to convert _LIBCONTAINER_CONSOLE=%s to int: %s", envConsole, err)
+ }
+ consoleSocket = os.NewFile(uintptr(console), "console-socket")
+ defer consoleSocket.Close()
+ }
+
+ // clear the current process's environment to clean any libcontainer
+ // specific env vars.
+ os.Clearenv()
+
+ defer func() {
+ // We have an error during the initialization of the container's init,
+ // send it back to the parent process in the form of an initError.
+ if werr := utils.WriteJSON(pipe, syncT{procError}); werr != nil {
+ fmt.Fprintln(os.Stderr, err)
+ return
+ }
+ if werr := utils.WriteJSON(pipe, newSystemError(err)); werr != nil {
+ fmt.Fprintln(os.Stderr, err)
+ return
+ }
+ }()
+ defer func() {
+ if e := recover(); e != nil {
+ err = fmt.Errorf("panic from initialization: %v, %v", e, string(debug.Stack()))
+ }
+ }()
+
+ i, err := newContainerInit(it, pipe, consoleSocket, rootfd)
+ if err != nil {
+ return err
+ }
+
+ // If Init succeeds, syscall.Exec will not return, hence none of the defers will be called.
+ return i.Init()
+}
+
+func (l *LinuxFactory) loadState(root, id string) (*State, error) {
+ f, err := os.Open(filepath.Join(root, stateFilename))
+ if err != nil {
+ if os.IsNotExist(err) {
+ return nil, newGenericError(fmt.Errorf("container %q does not exist", id), ContainerNotExists)
+ }
+ return nil, newGenericError(err, SystemError)
+ }
+ defer f.Close()
+ var state *State
+ if err := json.NewDecoder(f).Decode(&state); err != nil {
+ return nil, newGenericError(err, SystemError)
+ }
+ return state, nil
+}
+
+func (l *LinuxFactory) validateID(id string) error {
+ if !idRegex.MatchString(id) {
+ return newGenericError(fmt.Errorf("invalid id format: %v", id), InvalidIdFormat)
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/generic_error.go b/vendor/github.com/opencontainers/runc/libcontainer/generic_error.go
new file mode 100644
index 000000000..6e7de2fe7
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/generic_error.go
@@ -0,0 +1,92 @@
+package libcontainer
+
+import (
+ "fmt"
+ "io"
+ "text/template"
+ "time"
+
+ "github.com/opencontainers/runc/libcontainer/stacktrace"
+)
+
+var errorTemplate = template.Must(template.New("error").Parse(`Timestamp: {{.Timestamp}}
+Code: {{.ECode}}
+{{if .Message }}
+Message: {{.Message}}
+{{end}}
+Frames:{{range $i, $frame := .Stack.Frames}}
+---
+{{$i}}: {{$frame.Function}}
+Package: {{$frame.Package}}
+File: {{$frame.File}}@{{$frame.Line}}{{end}}
+`))
+
+func newGenericError(err error, c ErrorCode) Error {
+ if le, ok := err.(Error); ok {
+ return le
+ }
+ gerr := &genericError{
+ Timestamp: time.Now(),
+ Err: err,
+ ECode: c,
+ Stack: stacktrace.Capture(1),
+ }
+ if err != nil {
+ gerr.Message = err.Error()
+ }
+ return gerr
+}
+
+func newSystemError(err error) Error {
+ return createSystemError(err, "")
+}
+
+func newSystemErrorWithCausef(err error, cause string, v ...interface{}) Error {
+ return createSystemError(err, fmt.Sprintf(cause, v...))
+}
+
+func newSystemErrorWithCause(err error, cause string) Error {
+ return createSystemError(err, cause)
+}
+
+// createSystemError creates the specified error with the correct number of
+// stack frames skipped. This is only to be called by the other functions for
+// formatting the error.
+func createSystemError(err error, cause string) Error {
+ gerr := &genericError{
+ Timestamp: time.Now(),
+ Err: err,
+ ECode: SystemError,
+ Cause: cause,
+ Stack: stacktrace.Capture(2),
+ }
+ if err != nil {
+ gerr.Message = err.Error()
+ }
+ return gerr
+}
+
+type genericError struct {
+ Timestamp time.Time
+ ECode ErrorCode
+ Err error `json:"-"`
+ Cause string
+ Message string
+ Stack stacktrace.Stacktrace
+}
+
+func (e *genericError) Error() string {
+ if e.Cause == "" {
+ return e.Message
+ }
+ frame := e.Stack.Frames[0]
+ return fmt.Sprintf("%s:%d: %s caused %q", frame.File, frame.Line, e.Cause, e.Message)
+}
+
+func (e *genericError) Code() ErrorCode {
+ return e.ECode
+}
+
+func (e *genericError) Detail(w io.Writer) error {
+ return errorTemplate.Execute(w, e)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/init_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/init_linux.go
new file mode 100644
index 000000000..63afd28eb
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/init_linux.go
@@ -0,0 +1,502 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "encoding/json"
+ "fmt"
+ "io"
+ "net"
+ "os"
+ "strings"
+ "syscall" // only for Errno
+ "unsafe"
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/system"
+ "github.com/opencontainers/runc/libcontainer/user"
+ "github.com/opencontainers/runc/libcontainer/utils"
+
+ "github.com/sirupsen/logrus"
+ "github.com/vishvananda/netlink"
+ "golang.org/x/sys/unix"
+)
+
+type initType string
+
+const (
+ initSetns initType = "setns"
+ initStandard initType = "standard"
+)
+
+type pid struct {
+ Pid int `json:"pid"`
+}
+
+// network is an internal struct used to setup container networks.
+type network struct {
+ configs.Network
+
+ // TempVethPeerName is a unique temporary veth peer name that was placed into
+ // the container's namespace.
+ TempVethPeerName string `json:"temp_veth_peer_name"`
+}
+
+// initConfig is used for transferring parameters from Exec() to Init()
+type initConfig struct {
+ Args []string `json:"args"`
+ Env []string `json:"env"`
+ Cwd string `json:"cwd"`
+ Capabilities *configs.Capabilities `json:"capabilities"`
+ ProcessLabel string `json:"process_label"`
+ AppArmorProfile string `json:"apparmor_profile"`
+ NoNewPrivileges bool `json:"no_new_privileges"`
+ User string `json:"user"`
+ AdditionalGroups []string `json:"additional_groups"`
+ Config *configs.Config `json:"config"`
+ Networks []*network `json:"network"`
+ PassedFilesCount int `json:"passed_files_count"`
+ ContainerId string `json:"containerid"`
+ Rlimits []configs.Rlimit `json:"rlimits"`
+ CreateConsole bool `json:"create_console"`
+ Rootless bool `json:"rootless"`
+}
+
+type initer interface {
+ Init() error
+}
+
+func newContainerInit(t initType, pipe *os.File, consoleSocket *os.File, stateDirFD int) (initer, error) {
+ var config *initConfig
+ if err := json.NewDecoder(pipe).Decode(&config); err != nil {
+ return nil, err
+ }
+ if err := populateProcessEnvironment(config.Env); err != nil {
+ return nil, err
+ }
+ switch t {
+ case initSetns:
+ return &linuxSetnsInit{
+ pipe: pipe,
+ consoleSocket: consoleSocket,
+ config: config,
+ }, nil
+ case initStandard:
+ return &linuxStandardInit{
+ pipe: pipe,
+ consoleSocket: consoleSocket,
+ parentPid: unix.Getppid(),
+ config: config,
+ stateDirFD: stateDirFD,
+ }, nil
+ }
+ return nil, fmt.Errorf("unknown init type %q", t)
+}
+
+// populateProcessEnvironment loads the provided environment variables into the
+// current processes's environment.
+func populateProcessEnvironment(env []string) error {
+ for _, pair := range env {
+ p := strings.SplitN(pair, "=", 2)
+ if len(p) < 2 {
+ return fmt.Errorf("invalid environment '%v'", pair)
+ }
+ if err := os.Setenv(p[0], p[1]); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// finalizeNamespace drops the caps, sets the correct user
+// and working dir, and closes any leaked file descriptors
+// before executing the command inside the namespace
+func finalizeNamespace(config *initConfig) error {
+ // Ensure that all unwanted fds we may have accidentally
+ // inherited are marked close-on-exec so they stay out of the
+ // container
+ if err := utils.CloseExecFrom(config.PassedFilesCount + 3); err != nil {
+ return err
+ }
+
+ capabilities := &configs.Capabilities{}
+ if config.Capabilities != nil {
+ capabilities = config.Capabilities
+ } else if config.Config.Capabilities != nil {
+ capabilities = config.Config.Capabilities
+ }
+ w, err := newContainerCapList(capabilities)
+ if err != nil {
+ return err
+ }
+ // drop capabilities in bounding set before changing user
+ if err := w.ApplyBoundingSet(); err != nil {
+ return err
+ }
+ // preserve existing capabilities while we change users
+ if err := system.SetKeepCaps(); err != nil {
+ return err
+ }
+ if err := setupUser(config); err != nil {
+ return err
+ }
+ if err := system.ClearKeepCaps(); err != nil {
+ return err
+ }
+ if err := w.ApplyCaps(); err != nil {
+ return err
+ }
+ if config.Cwd != "" {
+ if err := unix.Chdir(config.Cwd); err != nil {
+ return fmt.Errorf("chdir to cwd (%q) set in config.json failed: %v", config.Cwd, err)
+ }
+ }
+ return nil
+}
+
+// setupConsole sets up the console from inside the container, and sends the
+// master pty fd to the config.Pipe (using cmsg). This is done to ensure that
+// consoles are scoped to a container properly (see runc#814 and the many
+// issues related to that). This has to be run *after* we've pivoted to the new
+// rootfs (and the users' configuration is entirely set up).
+func setupConsole(socket *os.File, config *initConfig, mount bool) error {
+ defer socket.Close()
+ // At this point, /dev/ptmx points to something that we would expect. We
+ // used to change the owner of the slave path, but since the /dev/pts mount
+ // can have gid=X set (at the users' option). So touching the owner of the
+ // slave PTY is not necessary, as the kernel will handle that for us. Note
+ // however, that setupUser (specifically fixStdioPermissions) *will* change
+ // the UID owner of the console to be the user the process will run as (so
+ // they can actually control their console).
+ console, err := newConsole()
+ if err != nil {
+ return err
+ }
+ // After we return from here, we don't need the console anymore.
+ defer console.Close()
+
+ linuxConsole, ok := console.(*linuxConsole)
+ if !ok {
+ return fmt.Errorf("failed to cast console to *linuxConsole")
+ }
+ // Mount the console inside our rootfs.
+ if mount {
+ if err := linuxConsole.mount(); err != nil {
+ return err
+ }
+ }
+ // While we can access console.master, using the API is a good idea.
+ if err := utils.SendFd(socket, linuxConsole.File()); err != nil {
+ return err
+ }
+ // Now, dup over all the things.
+ return linuxConsole.dupStdio()
+}
+
+// syncParentReady sends to the given pipe a JSON payload which indicates that
+// the init is ready to Exec the child process. It then waits for the parent to
+// indicate that it is cleared to Exec.
+func syncParentReady(pipe io.ReadWriter) error {
+ // Tell parent.
+ if err := writeSync(pipe, procReady); err != nil {
+ return err
+ }
+
+ // Wait for parent to give the all-clear.
+ if err := readSync(pipe, procRun); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+// syncParentHooks sends to the given pipe a JSON payload which indicates that
+// the parent should execute pre-start hooks. It then waits for the parent to
+// indicate that it is cleared to resume.
+func syncParentHooks(pipe io.ReadWriter) error {
+ // Tell parent.
+ if err := writeSync(pipe, procHooks); err != nil {
+ return err
+ }
+
+ // Wait for parent to give the all-clear.
+ if err := readSync(pipe, procResume); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+// setupUser changes the groups, gid, and uid for the user inside the container
+func setupUser(config *initConfig) error {
+ // Set up defaults.
+ defaultExecUser := user.ExecUser{
+ Uid: 0,
+ Gid: 0,
+ Home: "/",
+ }
+
+ passwdPath, err := user.GetPasswdPath()
+ if err != nil {
+ return err
+ }
+
+ groupPath, err := user.GetGroupPath()
+ if err != nil {
+ return err
+ }
+
+ execUser, err := user.GetExecUserPath(config.User, &defaultExecUser, passwdPath, groupPath)
+ if err != nil {
+ return err
+ }
+
+ var addGroups []int
+ if len(config.AdditionalGroups) > 0 {
+ addGroups, err = user.GetAdditionalGroupsPath(config.AdditionalGroups, groupPath)
+ if err != nil {
+ return err
+ }
+ }
+
+ if config.Rootless {
+ if execUser.Uid != 0 {
+ return fmt.Errorf("cannot run as a non-root user in a rootless container")
+ }
+
+ if execUser.Gid != 0 {
+ return fmt.Errorf("cannot run as a non-root group in a rootless container")
+ }
+
+ // We cannot set any additional groups in a rootless container and thus we
+ // bail if the user asked us to do so. TODO: We currently can't do this
+ // earlier, but if libcontainer.Process.User was typesafe this might work.
+ if len(addGroups) > 0 {
+ return fmt.Errorf("cannot set any additional groups in a rootless container")
+ }
+ }
+
+ // before we change to the container's user make sure that the processes STDIO
+ // is correctly owned by the user that we are switching to.
+ if err := fixStdioPermissions(config, execUser); err != nil {
+ return err
+ }
+
+ // This isn't allowed in an unprivileged user namespace since Linux 3.19.
+ // There's nothing we can do about /etc/group entries, so we silently
+ // ignore setting groups here (since the user didn't explicitly ask us to
+ // set the group).
+ if !config.Rootless {
+ suppGroups := append(execUser.Sgids, addGroups...)
+ if err := unix.Setgroups(suppGroups); err != nil {
+ return err
+ }
+ }
+
+ if err := system.Setgid(execUser.Gid); err != nil {
+ return err
+ }
+
+ if err := system.Setuid(execUser.Uid); err != nil {
+ return err
+ }
+
+ // if we didn't get HOME already, set it based on the user's HOME
+ if envHome := os.Getenv("HOME"); envHome == "" {
+ if err := os.Setenv("HOME", execUser.Home); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// fixStdioPermissions fixes the permissions of PID 1's STDIO within the container to the specified user.
+// The ownership needs to match because it is created outside of the container and needs to be
+// localized.
+func fixStdioPermissions(config *initConfig, u *user.ExecUser) error {
+ var null unix.Stat_t
+ if err := unix.Stat("/dev/null", &null); err != nil {
+ return err
+ }
+ for _, fd := range []uintptr{
+ os.Stdin.Fd(),
+ os.Stderr.Fd(),
+ os.Stdout.Fd(),
+ } {
+ var s unix.Stat_t
+ if err := unix.Fstat(int(fd), &s); err != nil {
+ return err
+ }
+
+ // Skip chown of /dev/null if it was used as one of the STDIO fds.
+ if s.Rdev == null.Rdev {
+ continue
+ }
+
+ // Skip chown if s.Gid is actually an unmapped gid in the host. While
+ // this is a bit dodgy if it just so happens that the console _is_
+ // owned by overflow_gid, there's no way for us to disambiguate this as
+ // a userspace program.
+ if _, err := config.Config.HostGID(int(s.Gid)); err != nil {
+ continue
+ }
+
+ // We only change the uid owner (as it is possible for the mount to
+ // prefer a different gid, and there's no reason for us to change it).
+ // The reason why we don't just leave the default uid=X mount setup is
+ // that users expect to be able to actually use their console. Without
+ // this code, you couldn't effectively run as a non-root user inside a
+ // container and also have a console set up.
+ if err := unix.Fchown(int(fd), u.Uid, int(s.Gid)); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+// setupNetwork sets up and initializes any network interface inside the container.
+func setupNetwork(config *initConfig) error {
+ for _, config := range config.Networks {
+ strategy, err := getStrategy(config.Type)
+ if err != nil {
+ return err
+ }
+ if err := strategy.initialize(config); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func setupRoute(config *configs.Config) error {
+ for _, config := range config.Routes {
+ _, dst, err := net.ParseCIDR(config.Destination)
+ if err != nil {
+ return err
+ }
+ src := net.ParseIP(config.Source)
+ if src == nil {
+ return fmt.Errorf("Invalid source for route: %s", config.Source)
+ }
+ gw := net.ParseIP(config.Gateway)
+ if gw == nil {
+ return fmt.Errorf("Invalid gateway for route: %s", config.Gateway)
+ }
+ l, err := netlink.LinkByName(config.InterfaceName)
+ if err != nil {
+ return err
+ }
+ route := &netlink.Route{
+ Scope: netlink.SCOPE_UNIVERSE,
+ Dst: dst,
+ Src: src,
+ Gw: gw,
+ LinkIndex: l.Attrs().Index,
+ }
+ if err := netlink.RouteAdd(route); err != nil {
+ return err
+ }
+ }
+ return nil
+}
+
+func setupRlimits(limits []configs.Rlimit, pid int) error {
+ for _, rlimit := range limits {
+ if err := system.Prlimit(pid, rlimit.Type, unix.Rlimit{Max: rlimit.Hard, Cur: rlimit.Soft}); err != nil {
+ return fmt.Errorf("error setting rlimit type %v: %v", rlimit.Type, err)
+ }
+ }
+ return nil
+}
+
+const _P_PID = 1
+
+type siginfo struct {
+ si_signo int32
+ si_errno int32
+ si_code int32
+ // below here is a union; si_pid is the only field we use
+ si_pid int32
+ // Pad to 128 bytes as detailed in blockUntilWaitable
+ pad [96]byte
+}
+
+// isWaitable returns true if the process has exited false otherwise.
+// Its based off blockUntilWaitable in src/os/wait_waitid.go
+func isWaitable(pid int) (bool, error) {
+ si := &siginfo{}
+ _, _, e := unix.Syscall6(unix.SYS_WAITID, _P_PID, uintptr(pid), uintptr(unsafe.Pointer(si)), unix.WEXITED|unix.WNOWAIT|unix.WNOHANG, 0, 0)
+ if e != 0 {
+ return false, os.NewSyscallError("waitid", e)
+ }
+
+ return si.si_pid != 0, nil
+}
+
+// isNoChildren returns true if err represents a unix.ECHILD (formerly syscall.ECHILD) false otherwise
+func isNoChildren(err error) bool {
+ switch err := err.(type) {
+ case syscall.Errno:
+ if err == unix.ECHILD {
+ return true
+ }
+ case *os.SyscallError:
+ if err.Err == unix.ECHILD {
+ return true
+ }
+ }
+ return false
+}
+
+// signalAllProcesses freezes then iterates over all the processes inside the
+// manager's cgroups sending the signal s to them.
+// If s is SIGKILL then it will wait for each process to exit.
+// For all other signals it will check if the process is ready to report its
+// exit status and only if it is will a wait be performed.
+func signalAllProcesses(m cgroups.Manager, s os.Signal) error {
+ var procs []*os.Process
+ if err := m.Freeze(configs.Frozen); err != nil {
+ logrus.Warn(err)
+ }
+ pids, err := m.GetAllPids()
+ if err != nil {
+ m.Freeze(configs.Thawed)
+ return err
+ }
+ for _, pid := range pids {
+ p, err := os.FindProcess(pid)
+ if err != nil {
+ logrus.Warn(err)
+ continue
+ }
+ procs = append(procs, p)
+ if err := p.Signal(s); err != nil {
+ logrus.Warn(err)
+ }
+ }
+ if err := m.Freeze(configs.Thawed); err != nil {
+ logrus.Warn(err)
+ }
+
+ for _, p := range procs {
+ if s != unix.SIGKILL {
+ if ok, err := isWaitable(p.Pid); err != nil {
+ if !isNoChildren(err) {
+ logrus.Warn("signalAllProcesses: ", p.Pid, err)
+ }
+ continue
+ } else if !ok {
+ // Not ready to report so don't wait
+ continue
+ }
+ }
+
+ if _, err := p.Wait(); err != nil {
+ if !isNoChildren(err) {
+ logrus.Warn("wait: ", err)
+ }
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/keys/keyctl.go b/vendor/github.com/opencontainers/runc/libcontainer/keys/keyctl.go
new file mode 100644
index 000000000..82ffa7a88
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/keys/keyctl.go
@@ -0,0 +1,50 @@
+// +build linux
+
+package keys
+
+import (
+ "fmt"
+ "strconv"
+ "strings"
+
+ "golang.org/x/sys/unix"
+)
+
+type KeySerial uint32
+
+func JoinSessionKeyring(name string) (KeySerial, error) {
+ sessKeyId, err := unix.KeyctlJoinSessionKeyring(name)
+ if err != nil {
+ return 0, fmt.Errorf("could not create session key: %v", err)
+ }
+ return KeySerial(sessKeyId), nil
+}
+
+// ModKeyringPerm modifies permissions on a keyring by reading the current permissions,
+// anding the bits with the given mask (clearing permissions) and setting
+// additional permission bits
+func ModKeyringPerm(ringId KeySerial, mask, setbits uint32) error {
+ dest, err := unix.KeyctlString(unix.KEYCTL_DESCRIBE, int(ringId))
+ if err != nil {
+ return err
+ }
+
+ res := strings.Split(string(dest), ";")
+ if len(res) < 5 {
+ return fmt.Errorf("Destination buffer for key description is too small")
+ }
+
+ // parse permissions
+ perm64, err := strconv.ParseUint(res[3], 16, 32)
+ if err != nil {
+ return err
+ }
+
+ perm := (uint32(perm64) & mask) | setbits
+
+ if err := unix.KeyctlSetperm(int(ringId), perm); err != nil {
+ return err
+ }
+
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/message_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/message_linux.go
new file mode 100644
index 000000000..8829b71ad
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/message_linux.go
@@ -0,0 +1,87 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "github.com/vishvananda/netlink/nl"
+ "golang.org/x/sys/unix"
+)
+
+// list of known message types we want to send to bootstrap program
+// The number is randomly chosen to not conflict with known netlink types
+const (
+ InitMsg uint16 = 62000
+ CloneFlagsAttr uint16 = 27281
+ NsPathsAttr uint16 = 27282
+ UidmapAttr uint16 = 27283
+ GidmapAttr uint16 = 27284
+ SetgroupAttr uint16 = 27285
+ OomScoreAdjAttr uint16 = 27286
+ RootlessAttr uint16 = 27287
+)
+
+type Int32msg struct {
+ Type uint16
+ Value uint32
+}
+
+// Serialize serializes the message.
+// Int32msg has the following representation
+// | nlattr len | nlattr type |
+// | uint32 value |
+func (msg *Int32msg) Serialize() []byte {
+ buf := make([]byte, msg.Len())
+ native := nl.NativeEndian()
+ native.PutUint16(buf[0:2], uint16(msg.Len()))
+ native.PutUint16(buf[2:4], msg.Type)
+ native.PutUint32(buf[4:8], msg.Value)
+ return buf
+}
+
+func (msg *Int32msg) Len() int {
+ return unix.NLA_HDRLEN + 4
+}
+
+// Bytemsg has the following representation
+// | nlattr len | nlattr type |
+// | value | pad |
+type Bytemsg struct {
+ Type uint16
+ Value []byte
+}
+
+func (msg *Bytemsg) Serialize() []byte {
+ l := msg.Len()
+ buf := make([]byte, (l+unix.NLA_ALIGNTO-1) & ^(unix.NLA_ALIGNTO-1))
+ native := nl.NativeEndian()
+ native.PutUint16(buf[0:2], uint16(l))
+ native.PutUint16(buf[2:4], msg.Type)
+ copy(buf[4:], msg.Value)
+ return buf
+}
+
+func (msg *Bytemsg) Len() int {
+ return unix.NLA_HDRLEN + len(msg.Value) + 1 // null-terminated
+}
+
+type Boolmsg struct {
+ Type uint16
+ Value bool
+}
+
+func (msg *Boolmsg) Serialize() []byte {
+ buf := make([]byte, msg.Len())
+ native := nl.NativeEndian()
+ native.PutUint16(buf[0:2], uint16(msg.Len()))
+ native.PutUint16(buf[2:4], msg.Type)
+ if msg.Value {
+ buf[4] = 1
+ } else {
+ buf[4] = 0
+ }
+ return buf
+}
+
+func (msg *Boolmsg) Len() int {
+ return unix.NLA_HDRLEN + 1
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/network_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/network_linux.go
new file mode 100644
index 000000000..5075bee4d
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/network_linux.go
@@ -0,0 +1,259 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "io/ioutil"
+ "net"
+ "path/filepath"
+ "strconv"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/utils"
+ "github.com/vishvananda/netlink"
+)
+
+var strategies = map[string]networkStrategy{
+ "veth": &veth{},
+ "loopback": &loopback{},
+}
+
+// networkStrategy represents a specific network configuration for
+// a container's networking stack
+type networkStrategy interface {
+ create(*network, int) error
+ initialize(*network) error
+ detach(*configs.Network) error
+ attach(*configs.Network) error
+}
+
+// getStrategy returns the specific network strategy for the
+// provided type.
+func getStrategy(tpe string) (networkStrategy, error) {
+ s, exists := strategies[tpe]
+ if !exists {
+ return nil, fmt.Errorf("unknown strategy type %q", tpe)
+ }
+ return s, nil
+}
+
+// Returns the network statistics for the network interfaces represented by the NetworkRuntimeInfo.
+func getNetworkInterfaceStats(interfaceName string) (*NetworkInterface, error) {
+ out := &NetworkInterface{Name: interfaceName}
+ // This can happen if the network runtime information is missing - possible if the
+ // container was created by an old version of libcontainer.
+ if interfaceName == "" {
+ return out, nil
+ }
+ type netStatsPair struct {
+ // Where to write the output.
+ Out *uint64
+ // The network stats file to read.
+ File string
+ }
+ // Ingress for host veth is from the container. Hence tx_bytes stat on the host veth is actually number of bytes received by the container.
+ netStats := []netStatsPair{
+ {Out: &out.RxBytes, File: "tx_bytes"},
+ {Out: &out.RxPackets, File: "tx_packets"},
+ {Out: &out.RxErrors, File: "tx_errors"},
+ {Out: &out.RxDropped, File: "tx_dropped"},
+
+ {Out: &out.TxBytes, File: "rx_bytes"},
+ {Out: &out.TxPackets, File: "rx_packets"},
+ {Out: &out.TxErrors, File: "rx_errors"},
+ {Out: &out.TxDropped, File: "rx_dropped"},
+ }
+ for _, netStat := range netStats {
+ data, err := readSysfsNetworkStats(interfaceName, netStat.File)
+ if err != nil {
+ return nil, err
+ }
+ *(netStat.Out) = data
+ }
+ return out, nil
+}
+
+// Reads the specified statistics available under /sys/class/net/<EthInterface>/statistics
+func readSysfsNetworkStats(ethInterface, statsFile string) (uint64, error) {
+ data, err := ioutil.ReadFile(filepath.Join("/sys/class/net", ethInterface, "statistics", statsFile))
+ if err != nil {
+ return 0, err
+ }
+ return strconv.ParseUint(strings.TrimSpace(string(data)), 10, 64)
+}
+
+// loopback is a network strategy that provides a basic loopback device
+type loopback struct {
+}
+
+func (l *loopback) create(n *network, nspid int) error {
+ return nil
+}
+
+func (l *loopback) initialize(config *network) error {
+ return netlink.LinkSetUp(&netlink.Device{LinkAttrs: netlink.LinkAttrs{Name: "lo"}})
+}
+
+func (l *loopback) attach(n *configs.Network) (err error) {
+ return nil
+}
+
+func (l *loopback) detach(n *configs.Network) (err error) {
+ return nil
+}
+
+// veth is a network strategy that uses a bridge and creates
+// a veth pair, one that is attached to the bridge on the host and the other
+// is placed inside the container's namespace
+type veth struct {
+}
+
+func (v *veth) detach(n *configs.Network) (err error) {
+ return netlink.LinkSetMaster(&netlink.Device{LinkAttrs: netlink.LinkAttrs{Name: n.HostInterfaceName}}, nil)
+}
+
+// attach a container network interface to an external network
+func (v *veth) attach(n *configs.Network) (err error) {
+ brl, err := netlink.LinkByName(n.Bridge)
+ if err != nil {
+ return err
+ }
+ br, ok := brl.(*netlink.Bridge)
+ if !ok {
+ return fmt.Errorf("Wrong device type %T", brl)
+ }
+ host, err := netlink.LinkByName(n.HostInterfaceName)
+ if err != nil {
+ return err
+ }
+
+ if err := netlink.LinkSetMaster(host, br); err != nil {
+ return err
+ }
+ if err := netlink.LinkSetMTU(host, n.Mtu); err != nil {
+ return err
+ }
+ if n.HairpinMode {
+ if err := netlink.LinkSetHairpin(host, true); err != nil {
+ return err
+ }
+ }
+ if err := netlink.LinkSetUp(host); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func (v *veth) create(n *network, nspid int) (err error) {
+ tmpName, err := v.generateTempPeerName()
+ if err != nil {
+ return err
+ }
+ n.TempVethPeerName = tmpName
+ if n.Bridge == "" {
+ return fmt.Errorf("bridge is not specified")
+ }
+ veth := &netlink.Veth{
+ LinkAttrs: netlink.LinkAttrs{
+ Name: n.HostInterfaceName,
+ TxQLen: n.TxQueueLen,
+ },
+ PeerName: n.TempVethPeerName,
+ }
+ if err := netlink.LinkAdd(veth); err != nil {
+ return err
+ }
+ defer func() {
+ if err != nil {
+ netlink.LinkDel(veth)
+ }
+ }()
+ if err := v.attach(&n.Network); err != nil {
+ return err
+ }
+ child, err := netlink.LinkByName(n.TempVethPeerName)
+ if err != nil {
+ return err
+ }
+ return netlink.LinkSetNsPid(child, nspid)
+}
+
+func (v *veth) generateTempPeerName() (string, error) {
+ return utils.GenerateRandomName("veth", 7)
+}
+
+func (v *veth) initialize(config *network) error {
+ peer := config.TempVethPeerName
+ if peer == "" {
+ return fmt.Errorf("peer is not specified")
+ }
+ child, err := netlink.LinkByName(peer)
+ if err != nil {
+ return err
+ }
+ if err := netlink.LinkSetDown(child); err != nil {
+ return err
+ }
+ if err := netlink.LinkSetName(child, config.Name); err != nil {
+ return err
+ }
+ // get the interface again after we changed the name as the index also changes.
+ if child, err = netlink.LinkByName(config.Name); err != nil {
+ return err
+ }
+ if config.MacAddress != "" {
+ mac, err := net.ParseMAC(config.MacAddress)
+ if err != nil {
+ return err
+ }
+ if err := netlink.LinkSetHardwareAddr(child, mac); err != nil {
+ return err
+ }
+ }
+ ip, err := netlink.ParseAddr(config.Address)
+ if err != nil {
+ return err
+ }
+ if err := netlink.AddrAdd(child, ip); err != nil {
+ return err
+ }
+ if config.IPv6Address != "" {
+ ip6, err := netlink.ParseAddr(config.IPv6Address)
+ if err != nil {
+ return err
+ }
+ if err := netlink.AddrAdd(child, ip6); err != nil {
+ return err
+ }
+ }
+ if err := netlink.LinkSetMTU(child, config.Mtu); err != nil {
+ return err
+ }
+ if err := netlink.LinkSetUp(child); err != nil {
+ return err
+ }
+ if config.Gateway != "" {
+ gw := net.ParseIP(config.Gateway)
+ if err := netlink.RouteAdd(&netlink.Route{
+ Scope: netlink.SCOPE_UNIVERSE,
+ LinkIndex: child.Attrs().Index,
+ Gw: gw,
+ }); err != nil {
+ return err
+ }
+ }
+ if config.IPv6Gateway != "" {
+ gw := net.ParseIP(config.IPv6Gateway)
+ if err := netlink.RouteAdd(&netlink.Route{
+ Scope: netlink.SCOPE_UNIVERSE,
+ LinkIndex: child.Attrs().Index,
+ Gw: gw,
+ }); err != nil {
+ return err
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/notify_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/notify_linux.go
new file mode 100644
index 000000000..81587b9b1
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/notify_linux.go
@@ -0,0 +1,90 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "io/ioutil"
+ "os"
+ "path/filepath"
+
+ "golang.org/x/sys/unix"
+)
+
+const oomCgroupName = "memory"
+
+type PressureLevel uint
+
+const (
+ LowPressure PressureLevel = iota
+ MediumPressure
+ CriticalPressure
+)
+
+func registerMemoryEvent(cgDir string, evName string, arg string) (<-chan struct{}, error) {
+ evFile, err := os.Open(filepath.Join(cgDir, evName))
+ if err != nil {
+ return nil, err
+ }
+ fd, err := unix.Eventfd(0, unix.EFD_CLOEXEC)
+ if err != nil {
+ evFile.Close()
+ return nil, err
+ }
+
+ eventfd := os.NewFile(uintptr(fd), "eventfd")
+
+ eventControlPath := filepath.Join(cgDir, "cgroup.event_control")
+ data := fmt.Sprintf("%d %d %s", eventfd.Fd(), evFile.Fd(), arg)
+ if err := ioutil.WriteFile(eventControlPath, []byte(data), 0700); err != nil {
+ eventfd.Close()
+ evFile.Close()
+ return nil, err
+ }
+ ch := make(chan struct{})
+ go func() {
+ defer func() {
+ close(ch)
+ eventfd.Close()
+ evFile.Close()
+ }()
+ buf := make([]byte, 8)
+ for {
+ if _, err := eventfd.Read(buf); err != nil {
+ return
+ }
+ // When a cgroup is destroyed, an event is sent to eventfd.
+ // So if the control path is gone, return instead of notifying.
+ if _, err := os.Lstat(eventControlPath); os.IsNotExist(err) {
+ return
+ }
+ ch <- struct{}{}
+ }
+ }()
+ return ch, nil
+}
+
+// notifyOnOOM returns channel on which you can expect event about OOM,
+// if process died without OOM this channel will be closed.
+func notifyOnOOM(paths map[string]string) (<-chan struct{}, error) {
+ dir := paths[oomCgroupName]
+ if dir == "" {
+ return nil, fmt.Errorf("path %q missing", oomCgroupName)
+ }
+
+ return registerMemoryEvent(dir, "memory.oom_control", "")
+}
+
+func notifyMemoryPressure(paths map[string]string, level PressureLevel) (<-chan struct{}, error) {
+ dir := paths[oomCgroupName]
+ if dir == "" {
+ return nil, fmt.Errorf("path %q missing", oomCgroupName)
+ }
+
+ if level > CriticalPressure {
+ return nil, fmt.Errorf("invalid pressure level %d", level)
+ }
+
+ levelStr := []string{"low", "medium", "critical"}[level]
+ return registerMemoryEvent(dir, "memory.pressure_level", levelStr)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/README.md b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/README.md
new file mode 100644
index 000000000..575701371
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/README.md
@@ -0,0 +1,44 @@
+## nsenter
+
+The `nsenter` package registers a special init constructor that is called before
+the Go runtime has a chance to boot. This provides us the ability to `setns` on
+existing namespaces and avoid the issues that the Go runtime has with multiple
+threads. This constructor will be called if this package is registered,
+imported, in your go application.
+
+The `nsenter` package will `import "C"` and it uses [cgo](https://golang.org/cmd/cgo/)
+package. In cgo, if the import of "C" is immediately preceded by a comment, that comment,
+called the preamble, is used as a header when compiling the C parts of the package.
+So every time we import package `nsenter`, the C code function `nsexec()` would be
+called. And package `nsenter` is now only imported in `main_unix.go`, so every time
+before we call `cmd.Start` on linux, that C code would run.
+
+Because `nsexec()` must be run before the Go runtime in order to use the
+Linux kernel namespace, you must `import` this library into a package if
+you plan to use `libcontainer` directly. Otherwise Go will not execute
+the `nsexec()` constructor, which means that the re-exec will not cause
+the namespaces to be joined. You can import it like this:
+
+```go
+import _ "github.com/opencontainers/runc/libcontainer/nsenter"
+```
+
+`nsexec()` will first get the file descriptor number for the init pipe
+from the environment variable `_LIBCONTAINER_INITPIPE` (which was opened
+by the parent and kept open across the fork-exec of the `nsexec()` init
+process). The init pipe is used to read bootstrap data (namespace paths,
+clone flags, uid and gid mappings, and the console path) from the parent
+process. `nsexec()` will then call `setns(2)` to join the namespaces
+provided in the bootstrap data (if available), `clone(2)` a child process
+with the provided clone flags, update the user and group ID mappings, do
+some further miscellaneous setup steps, and then send the PID of the
+child process to the parent of the `nsexec()` "caller". Finally,
+the parent `nsexec()` will exit and the child `nsexec()` process will
+return to allow the Go runtime take over.
+
+NOTE: We do both `setns(2)` and `clone(2)` even if we don't have any
+CLONE_NEW* clone flags because we must fork a new process in order to
+enter the PID namespace.
+
+
+
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/namespace.h b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/namespace.h
new file mode 100644
index 000000000..9e9bdca05
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/namespace.h
@@ -0,0 +1,32 @@
+#ifndef NSENTER_NAMESPACE_H
+#define NSENTER_NAMESPACE_H
+
+#ifndef _GNU_SOURCE
+# define _GNU_SOURCE
+#endif
+#include <sched.h>
+
+/* All of these are taken from include/uapi/linux/sched.h */
+#ifndef CLONE_NEWNS
+# define CLONE_NEWNS 0x00020000 /* New mount namespace group */
+#endif
+#ifndef CLONE_NEWCGROUP
+# define CLONE_NEWCGROUP 0x02000000 /* New cgroup namespace */
+#endif
+#ifndef CLONE_NEWUTS
+# define CLONE_NEWUTS 0x04000000 /* New utsname namespace */
+#endif
+#ifndef CLONE_NEWIPC
+# define CLONE_NEWIPC 0x08000000 /* New ipc namespace */
+#endif
+#ifndef CLONE_NEWUSER
+# define CLONE_NEWUSER 0x10000000 /* New user namespace */
+#endif
+#ifndef CLONE_NEWPID
+# define CLONE_NEWPID 0x20000000 /* New pid namespace */
+#endif
+#ifndef CLONE_NEWNET
+# define CLONE_NEWNET 0x40000000 /* New network namespace */
+#endif
+
+#endif /* NSENTER_NAMESPACE_H */
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go
new file mode 100644
index 000000000..07f4d63e4
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter.go
@@ -0,0 +1,12 @@
+// +build linux,!gccgo
+
+package nsenter
+
+/*
+#cgo CFLAGS: -Wall
+extern void nsexec();
+void __attribute__((constructor)) init(void) {
+ nsexec();
+}
+*/
+import "C"
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go
new file mode 100644
index 000000000..63c7a3ec2
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_gccgo.go
@@ -0,0 +1,25 @@
+// +build linux,gccgo
+
+package nsenter
+
+/*
+#cgo CFLAGS: -Wall
+extern void nsexec();
+void __attribute__((constructor)) init(void) {
+ nsexec();
+}
+*/
+import "C"
+
+// AlwaysFalse is here to stay false
+// (and be exported so the compiler doesn't optimize out its reference)
+var AlwaysFalse bool
+
+func init() {
+ if AlwaysFalse {
+ // by referencing this C init() in a noop test, it will ensure the compiler
+ // links in the C function.
+ // https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65134
+ C.init()
+ }
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go
new file mode 100644
index 000000000..ac701ca39
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsenter_unsupported.go
@@ -0,0 +1,5 @@
+// +build !linux !cgo
+
+package nsenter
+
+import "C"
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c
new file mode 100644
index 000000000..197e6d08e
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/nsenter/nsexec.c
@@ -0,0 +1,862 @@
+#define _GNU_SOURCE
+#include <endian.h>
+#include <errno.h>
+#include <fcntl.h>
+#include <grp.h>
+#include <sched.h>
+#include <setjmp.h>
+#include <signal.h>
+#include <stdarg.h>
+#include <stdbool.h>
+#include <stdint.h>
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdbool.h>
+#include <string.h>
+#include <unistd.h>
+
+#include <sys/ioctl.h>
+#include <sys/prctl.h>
+#include <sys/socket.h>
+#include <sys/types.h>
+
+#include <linux/limits.h>
+#include <linux/netlink.h>
+#include <linux/types.h>
+
+/* Get all of the CLONE_NEW* flags. */
+#include "namespace.h"
+
+/* Synchronisation values. */
+enum sync_t {
+ SYNC_USERMAP_PLS = 0x40, /* Request parent to map our users. */
+ SYNC_USERMAP_ACK = 0x41, /* Mapping finished by the parent. */
+ SYNC_RECVPID_PLS = 0x42, /* Tell parent we're sending the PID. */
+ SYNC_RECVPID_ACK = 0x43, /* PID was correctly received by parent. */
+ SYNC_GRANDCHILD = 0x44, /* The grandchild is ready to run. */
+ SYNC_CHILD_READY = 0x45, /* The child or grandchild is ready to return. */
+
+ /* XXX: This doesn't help with segfaults and other such issues. */
+ SYNC_ERR = 0xFF, /* Fatal error, no turning back. The error code follows. */
+};
+
+/* longjmp() arguments. */
+#define JUMP_PARENT 0x00
+#define JUMP_CHILD 0xA0
+#define JUMP_INIT 0xA1
+
+/* JSON buffer. */
+#define JSON_MAX 4096
+
+/* Assume the stack grows down, so arguments should be above it. */
+struct clone_t {
+ /*
+ * Reserve some space for clone() to locate arguments
+ * and retcode in this place
+ */
+ char stack[4096] __attribute__ ((aligned(16)));
+ char stack_ptr[0];
+
+ /* There's two children. This is used to execute the different code. */
+ jmp_buf *env;
+ int jmpval;
+};
+
+struct nlconfig_t {
+ char *data;
+ uint32_t cloneflags;
+ char *uidmap;
+ size_t uidmap_len;
+ char *gidmap;
+ size_t gidmap_len;
+ char *namespaces;
+ size_t namespaces_len;
+ uint8_t is_setgroup;
+ uint8_t is_rootless;
+ char *oom_score_adj;
+ size_t oom_score_adj_len;
+};
+
+/*
+ * List of netlink message types sent to us as part of bootstrapping the init.
+ * These constants are defined in libcontainer/message_linux.go.
+ */
+#define INIT_MSG 62000
+#define CLONE_FLAGS_ATTR 27281
+#define NS_PATHS_ATTR 27282
+#define UIDMAP_ATTR 27283
+#define GIDMAP_ATTR 27284
+#define SETGROUP_ATTR 27285
+#define OOM_SCORE_ADJ_ATTR 27286
+#define ROOTLESS_ATTR 27287
+
+/*
+ * Use the raw syscall for versions of glibc which don't include a function for
+ * it, namely (glibc 2.12).
+ */
+#if __GLIBC__ == 2 && __GLIBC_MINOR__ < 14
+# define _GNU_SOURCE
+# include "syscall.h"
+# if !defined(SYS_setns) && defined(__NR_setns)
+# define SYS_setns __NR_setns
+# endif
+
+#ifndef SYS_setns
+# error "setns(2) syscall not supported by glibc version"
+#endif
+
+int setns(int fd, int nstype)
+{
+ return syscall(SYS_setns, fd, nstype);
+}
+#endif
+
+/* XXX: This is ugly. */
+static int syncfd = -1;
+
+/* TODO(cyphar): Fix this so it correctly deals with syncT. */
+#define bail(fmt, ...) \
+ do { \
+ int ret = __COUNTER__ + 1; \
+ fprintf(stderr, "nsenter: " fmt ": %m\n", ##__VA_ARGS__); \
+ if (syncfd >= 0) { \
+ enum sync_t s = SYNC_ERR; \
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) \
+ fprintf(stderr, "nsenter: failed: write(s)"); \
+ if (write(syncfd, &ret, sizeof(ret)) != sizeof(ret)) \
+ fprintf(stderr, "nsenter: failed: write(ret)"); \
+ } \
+ exit(ret); \
+ } while(0)
+
+static int write_file(char *data, size_t data_len, char *pathfmt, ...)
+{
+ int fd, len, ret = 0;
+ char path[PATH_MAX];
+
+ va_list ap;
+ va_start(ap, pathfmt);
+ len = vsnprintf(path, PATH_MAX, pathfmt, ap);
+ va_end(ap);
+ if (len < 0)
+ return -1;
+
+ fd = open(path, O_RDWR);
+ if (fd < 0) {
+ return -1;
+ }
+
+ len = write(fd, data, data_len);
+ if (len != data_len) {
+ ret = -1;
+ goto out;
+ }
+
+out:
+ close(fd);
+ return ret;
+}
+
+enum policy_t {
+ SETGROUPS_DEFAULT = 0,
+ SETGROUPS_ALLOW,
+ SETGROUPS_DENY,
+};
+
+/* This *must* be called before we touch gid_map. */
+static void update_setgroups(int pid, enum policy_t setgroup)
+{
+ char *policy;
+
+ switch (setgroup) {
+ case SETGROUPS_ALLOW:
+ policy = "allow";
+ break;
+ case SETGROUPS_DENY:
+ policy = "deny";
+ break;
+ case SETGROUPS_DEFAULT:
+ default:
+ /* Nothing to do. */
+ return;
+ }
+
+ if (write_file(policy, strlen(policy), "/proc/%d/setgroups", pid) < 0) {
+ /*
+ * If the kernel is too old to support /proc/pid/setgroups,
+ * open(2) or write(2) will return ENOENT. This is fine.
+ */
+ if (errno != ENOENT)
+ bail("failed to write '%s' to /proc/%d/setgroups", policy, pid);
+ }
+}
+
+static void update_uidmap(int pid, char *map, size_t map_len)
+{
+ if (map == NULL || map_len <= 0)
+ return;
+
+ if (write_file(map, map_len, "/proc/%d/uid_map", pid) < 0)
+ bail("failed to update /proc/%d/uid_map", pid);
+}
+
+static void update_gidmap(int pid, char *map, size_t map_len)
+{
+ if (map == NULL || map_len <= 0)
+ return;
+
+ if (write_file(map, map_len, "/proc/%d/gid_map", pid) < 0)
+ bail("failed to update /proc/%d/gid_map", pid);
+}
+
+static void update_oom_score_adj(char *data, size_t len)
+{
+ if (data == NULL || len <= 0)
+ return;
+
+ if (write_file(data, len, "/proc/self/oom_score_adj") < 0)
+ bail("failed to update /proc/self/oom_score_adj");
+}
+
+/* A dummy function that just jumps to the given jumpval. */
+static int child_func(void *arg) __attribute__ ((noinline));
+static int child_func(void *arg)
+{
+ struct clone_t *ca = (struct clone_t *)arg;
+ longjmp(*ca->env, ca->jmpval);
+}
+
+static int clone_parent(jmp_buf *env, int jmpval) __attribute__ ((noinline));
+static int clone_parent(jmp_buf *env, int jmpval)
+{
+ struct clone_t ca = {
+ .env = env,
+ .jmpval = jmpval,
+ };
+
+ return clone(child_func, ca.stack_ptr, CLONE_PARENT | SIGCHLD, &ca);
+}
+
+/*
+ * Gets the init pipe fd from the environment, which is used to read the
+ * bootstrap data and tell the parent what the new pid is after we finish
+ * setting up the environment.
+ */
+static int initpipe(void)
+{
+ int pipenum;
+ char *initpipe, *endptr;
+
+ initpipe = getenv("_LIBCONTAINER_INITPIPE");
+ if (initpipe == NULL || *initpipe == '\0')
+ return -1;
+
+ pipenum = strtol(initpipe, &endptr, 10);
+ if (*endptr != '\0')
+ bail("unable to parse _LIBCONTAINER_INITPIPE");
+
+ return pipenum;
+}
+
+/* Returns the clone(2) flag for a namespace, given the name of a namespace. */
+static int nsflag(char *name)
+{
+ if (!strcmp(name, "cgroup"))
+ return CLONE_NEWCGROUP;
+ else if (!strcmp(name, "ipc"))
+ return CLONE_NEWIPC;
+ else if (!strcmp(name, "mnt"))
+ return CLONE_NEWNS;
+ else if (!strcmp(name, "net"))
+ return CLONE_NEWNET;
+ else if (!strcmp(name, "pid"))
+ return CLONE_NEWPID;
+ else if (!strcmp(name, "user"))
+ return CLONE_NEWUSER;
+ else if (!strcmp(name, "uts"))
+ return CLONE_NEWUTS;
+
+ /* If we don't recognise a name, fallback to 0. */
+ return 0;
+}
+
+static uint32_t readint32(char *buf)
+{
+ return *(uint32_t *) buf;
+}
+
+static uint8_t readint8(char *buf)
+{
+ return *(uint8_t *) buf;
+}
+
+static void nl_parse(int fd, struct nlconfig_t *config)
+{
+ size_t len, size;
+ struct nlmsghdr hdr;
+ char *data, *current;
+
+ /* Retrieve the netlink header. */
+ len = read(fd, &hdr, NLMSG_HDRLEN);
+ if (len != NLMSG_HDRLEN)
+ bail("invalid netlink header length %zu", len);
+
+ if (hdr.nlmsg_type == NLMSG_ERROR)
+ bail("failed to read netlink message");
+
+ if (hdr.nlmsg_type != INIT_MSG)
+ bail("unexpected msg type %d", hdr.nlmsg_type);
+
+ /* Retrieve data. */
+ size = NLMSG_PAYLOAD(&hdr, 0);
+ current = data = malloc(size);
+ if (!data)
+ bail("failed to allocate %zu bytes of memory for nl_payload", size);
+
+ len = read(fd, data, size);
+ if (len != size)
+ bail("failed to read netlink payload, %zu != %zu", len, size);
+
+ /* Parse the netlink payload. */
+ config->data = data;
+ while (current < data + size) {
+ struct nlattr *nlattr = (struct nlattr *)current;
+ size_t payload_len = nlattr->nla_len - NLA_HDRLEN;
+
+ /* Advance to payload. */
+ current += NLA_HDRLEN;
+
+ /* Handle payload. */
+ switch (nlattr->nla_type) {
+ case CLONE_FLAGS_ATTR:
+ config->cloneflags = readint32(current);
+ break;
+ case ROOTLESS_ATTR:
+ config->is_rootless = readint8(current);
+ break;
+ case OOM_SCORE_ADJ_ATTR:
+ config->oom_score_adj = current;
+ config->oom_score_adj_len = payload_len;
+ break;
+ case NS_PATHS_ATTR:
+ config->namespaces = current;
+ config->namespaces_len = payload_len;
+ break;
+ case UIDMAP_ATTR:
+ config->uidmap = current;
+ config->uidmap_len = payload_len;
+ break;
+ case GIDMAP_ATTR:
+ config->gidmap = current;
+ config->gidmap_len = payload_len;
+ break;
+ case SETGROUP_ATTR:
+ config->is_setgroup = readint8(current);
+ break;
+ default:
+ bail("unknown netlink message type %d", nlattr->nla_type);
+ }
+
+ current += NLA_ALIGN(payload_len);
+ }
+}
+
+void nl_free(struct nlconfig_t *config)
+{
+ free(config->data);
+}
+
+void join_namespaces(char *nslist)
+{
+ int num = 0, i;
+ char *saveptr = NULL;
+ char *namespace = strtok_r(nslist, ",", &saveptr);
+ struct namespace_t {
+ int fd;
+ int ns;
+ char type[PATH_MAX];
+ char path[PATH_MAX];
+ } *namespaces = NULL;
+
+ if (!namespace || !strlen(namespace) || !strlen(nslist))
+ bail("ns paths are empty");
+
+ /*
+ * We have to open the file descriptors first, since after
+ * we join the mnt namespace we might no longer be able to
+ * access the paths.
+ */
+ do {
+ int fd;
+ char *path;
+ struct namespace_t *ns;
+
+ /* Resize the namespace array. */
+ namespaces = realloc(namespaces, ++num * sizeof(struct namespace_t));
+ if (!namespaces)
+ bail("failed to reallocate namespace array");
+ ns = &namespaces[num - 1];
+
+ /* Split 'ns:path'. */
+ path = strstr(namespace, ":");
+ if (!path)
+ bail("failed to parse %s", namespace);
+ *path++ = '\0';
+
+ fd = open(path, O_RDONLY);
+ if (fd < 0)
+ bail("failed to open %s", path);
+
+ ns->fd = fd;
+ ns->ns = nsflag(namespace);
+ strncpy(ns->path, path, PATH_MAX);
+ } while ((namespace = strtok_r(NULL, ",", &saveptr)) != NULL);
+
+ /*
+ * The ordering in which we join namespaces is important. We should
+ * always join the user namespace *first*. This is all guaranteed
+ * from the container_linux.go side of this, so we're just going to
+ * follow the order given to us.
+ */
+
+ for (i = 0; i < num; i++) {
+ struct namespace_t ns = namespaces[i];
+
+ if (setns(ns.fd, ns.ns) < 0)
+ bail("failed to setns to %s", ns.path);
+
+ close(ns.fd);
+ }
+
+ free(namespaces);
+}
+
+void nsexec(void)
+{
+ int pipenum;
+ jmp_buf env;
+ int sync_child_pipe[2], sync_grandchild_pipe[2];
+ struct nlconfig_t config = {0};
+
+ /*
+ * If we don't have an init pipe, just return to the go routine.
+ * We'll only get an init pipe for start or exec.
+ */
+ pipenum = initpipe();
+ if (pipenum == -1)
+ return;
+
+ /* Parse all of the netlink configuration. */
+ nl_parse(pipenum, &config);
+
+ /* Set oom_score_adj. This has to be done before !dumpable because
+ * /proc/self/oom_score_adj is not writeable unless you're an privileged
+ * user (if !dumpable is set). All children inherit their parent's
+ * oom_score_adj value on fork(2) so this will always be propagated
+ * properly.
+ */
+ update_oom_score_adj(config.oom_score_adj, config.oom_score_adj_len);
+
+ /*
+ * Make the process non-dumpable, to avoid various race conditions that
+ * could cause processes in namespaces we're joining to access host
+ * resources (or potentially execute code).
+ *
+ * However, if the number of namespaces we are joining is 0, we are not
+ * going to be switching to a different security context. Thus setting
+ * ourselves to be non-dumpable only breaks things (like rootless
+ * containers), which is the recommendation from the kernel folks.
+ */
+ if (config.namespaces) {
+ if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
+ bail("failed to set process as non-dumpable");
+ }
+
+ /* Pipe so we can tell the child when we've finished setting up. */
+ if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_child_pipe) < 0)
+ bail("failed to setup sync pipe between parent and child");
+
+ /*
+ * We need a new socketpair to sync with grandchild so we don't have
+ * race condition with child.
+ */
+ if (socketpair(AF_LOCAL, SOCK_STREAM, 0, sync_grandchild_pipe) < 0)
+ bail("failed to setup sync pipe between parent and grandchild");
+
+ /* TODO: Currently we aren't dealing with child deaths properly. */
+
+ /*
+ * Okay, so this is quite annoying.
+ *
+ * In order for this unsharing code to be more extensible we need to split
+ * up unshare(CLONE_NEWUSER) and clone() in various ways. The ideal case
+ * would be if we did clone(CLONE_NEWUSER) and the other namespaces
+ * separately, but because of SELinux issues we cannot really do that. But
+ * we cannot just dump the namespace flags into clone(...) because several
+ * usecases (such as rootless containers) require more granularity around
+ * the namespace setup. In addition, some older kernels had issues where
+ * CLONE_NEWUSER wasn't handled before other namespaces (but we cannot
+ * handle this while also dealing with SELinux so we choose SELinux support
+ * over broken kernel support).
+ *
+ * However, if we unshare(2) the user namespace *before* we clone(2), then
+ * all hell breaks loose.
+ *
+ * The parent no longer has permissions to do many things (unshare(2) drops
+ * all capabilities in your old namespace), and the container cannot be set
+ * up to have more than one {uid,gid} mapping. This is obviously less than
+ * ideal. In order to fix this, we have to first clone(2) and then unshare.
+ *
+ * Unfortunately, it's not as simple as that. We have to fork to enter the
+ * PID namespace (the PID namespace only applies to children). Since we'll
+ * have to double-fork, this clone_parent() call won't be able to get the
+ * PID of the _actual_ init process (without doing more synchronisation than
+ * I can deal with at the moment). So we'll just get the parent to send it
+ * for us, the only job of this process is to update
+ * /proc/pid/{setgroups,uid_map,gid_map}.
+ *
+ * And as a result of the above, we also need to setns(2) in the first child
+ * because if we join a PID namespace in the topmost parent then our child
+ * will be in that namespace (and it will not be able to give us a PID value
+ * that makes sense without resorting to sending things with cmsg).
+ *
+ * This also deals with an older issue caused by dumping cloneflags into
+ * clone(2): On old kernels, CLONE_PARENT didn't work with CLONE_NEWPID, so
+ * we have to unshare(2) before clone(2) in order to do this. This was fixed
+ * in upstream commit 1f7f4dde5c945f41a7abc2285be43d918029ecc5, and was
+ * introduced by 40a0d32d1eaffe6aac7324ca92604b6b3977eb0e. As far as we're
+ * aware, the last mainline kernel which had this bug was Linux 3.12.
+ * However, we cannot comment on which kernels the broken patch was
+ * backported to.
+ *
+ * -- Aleksa "what has my life come to?" Sarai
+ */
+
+ switch (setjmp(env)) {
+ /*
+ * Stage 0: We're in the parent. Our job is just to create a new child
+ * (stage 1: JUMP_CHILD) process and write its uid_map and
+ * gid_map. That process will go on to create a new process, then
+ * it will send us its PID which we will send to the bootstrap
+ * process.
+ */
+ case JUMP_PARENT: {
+ int len;
+ pid_t child;
+ char buf[JSON_MAX];
+ bool ready = false;
+
+ /* For debugging. */
+ prctl(PR_SET_NAME, (unsigned long) "runc:[0:PARENT]", 0, 0, 0);
+
+ /* Start the process of getting a container. */
+ child = clone_parent(&env, JUMP_CHILD);
+ if (child < 0)
+ bail("unable to fork: child_func");
+
+ /*
+ * State machine for synchronisation with the children.
+ *
+ * Father only return when both child and grandchild are
+ * ready, so we can receive all possible error codes
+ * generated by children.
+ */
+ while (!ready) {
+ enum sync_t s;
+ int ret;
+
+ syncfd = sync_child_pipe[1];
+ close(sync_child_pipe[0]);
+
+ if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with child: next state");
+
+ switch (s) {
+ case SYNC_ERR:
+ /* We have to mirror the error code of the child. */
+ if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
+ bail("failed to sync with child: read(error code)");
+
+ exit(ret);
+ case SYNC_USERMAP_PLS:
+ /*
+ * Enable setgroups(2) if we've been asked to. But we also
+ * have to explicitly disable setgroups(2) if we're
+ * creating a rootless container (this is required since
+ * Linux 3.19).
+ */
+ if (config.is_rootless && config.is_setgroup) {
+ kill(child, SIGKILL);
+ bail("cannot allow setgroup in an unprivileged user namespace setup");
+ }
+
+ if (config.is_setgroup)
+ update_setgroups(child, SETGROUPS_ALLOW);
+ if (config.is_rootless)
+ update_setgroups(child, SETGROUPS_DENY);
+
+ /* Set up mappings. */
+ update_uidmap(child, config.uidmap, config.uidmap_len);
+ update_gidmap(child, config.gidmap, config.gidmap_len);
+
+ s = SYNC_USERMAP_ACK;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with child: write(SYNC_USERMAP_ACK)");
+ }
+ break;
+ case SYNC_RECVPID_PLS: {
+ pid_t old = child;
+
+ /* Get the init_func pid. */
+ if (read(syncfd, &child, sizeof(child)) != sizeof(child)) {
+ kill(old, SIGKILL);
+ bail("failed to sync with child: read(childpid)");
+ }
+
+ /* Send ACK. */
+ s = SYNC_RECVPID_ACK;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(old, SIGKILL);
+ kill(child, SIGKILL);
+ bail("failed to sync with child: write(SYNC_RECVPID_ACK)");
+ }
+ }
+ break;
+ case SYNC_CHILD_READY:
+ ready = true;
+ break;
+ default:
+ bail("unexpected sync value: %u", s);
+ }
+ }
+
+ /* Now sync with grandchild. */
+
+ ready = false;
+ while (!ready) {
+ enum sync_t s;
+ int ret;
+
+ syncfd = sync_grandchild_pipe[1];
+ close(sync_grandchild_pipe[0]);
+
+ s = SYNC_GRANDCHILD;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with child: write(SYNC_GRANDCHILD)");
+ }
+
+ if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with child: next state");
+
+ switch (s) {
+ case SYNC_ERR:
+ /* We have to mirror the error code of the child. */
+ if (read(syncfd, &ret, sizeof(ret)) != sizeof(ret))
+ bail("failed to sync with child: read(error code)");
+
+ exit(ret);
+ case SYNC_CHILD_READY:
+ ready = true;
+ break;
+ default:
+ bail("unexpected sync value: %u", s);
+ }
+ }
+
+ /* Send the init_func pid back to our parent. */
+ len = snprintf(buf, JSON_MAX, "{\"pid\": %d}\n", child);
+ if (len < 0) {
+ kill(child, SIGKILL);
+ bail("unable to generate JSON for child pid");
+ }
+ if (write(pipenum, buf, len) != len) {
+ kill(child, SIGKILL);
+ bail("unable to send child pid to bootstrapper");
+ }
+
+ exit(0);
+ }
+
+ /*
+ * Stage 1: We're in the first child process. Our job is to join any
+ * provided namespaces in the netlink payload and unshare all
+ * of the requested namespaces. If we've been asked to
+ * CLONE_NEWUSER, we will ask our parent (stage 0) to set up
+ * our user mappings for us. Then, we create a new child
+ * (stage 2: JUMP_INIT) for PID namespace. We then send the
+ * child's PID to our parent (stage 0).
+ */
+ case JUMP_CHILD: {
+ pid_t child;
+ enum sync_t s;
+
+ /* We're in a child and thus need to tell the parent if we die. */
+ syncfd = sync_child_pipe[0];
+ close(sync_child_pipe[1]);
+
+ /* For debugging. */
+ prctl(PR_SET_NAME, (unsigned long) "runc:[1:CHILD]", 0, 0, 0);
+
+ /*
+ * We need to setns first. We cannot do this earlier (in stage 0)
+ * because of the fact that we forked to get here (the PID of
+ * [stage 2: JUMP_INIT]) would be meaningless). We could send it
+ * using cmsg(3) but that's just annoying.
+ */
+ if (config.namespaces)
+ join_namespaces(config.namespaces);
+
+ /*
+ * Unshare all of the namespaces. Now, it should be noted that this
+ * ordering might break in the future (especially with rootless
+ * containers). But for now, it's not possible to split this into
+ * CLONE_NEWUSER + [the rest] because of some RHEL SELinux issues.
+ *
+ * Note that we don't merge this with clone() because there were
+ * some old kernel versions where clone(CLONE_PARENT | CLONE_NEWPID)
+ * was broken, so we'll just do it the long way anyway.
+ */
+ if (unshare(config.cloneflags) < 0)
+ bail("failed to unshare namespaces");
+
+ /*
+ * Deal with user namespaces first. They are quite special, as they
+ * affect our ability to unshare other namespaces and are used as
+ * context for privilege checks.
+ */
+ if (config.cloneflags & CLONE_NEWUSER) {
+ /*
+ * We don't have the privileges to do any mapping here (see the
+ * clone_parent rant). So signal our parent to hook us up.
+ */
+
+ /* Switching is only necessary if we joined namespaces. */
+ if (config.namespaces) {
+ if (prctl(PR_SET_DUMPABLE, 1, 0, 0, 0) < 0)
+ bail("failed to set process as dumpable");
+ }
+ s = SYNC_USERMAP_PLS;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with parent: write(SYNC_USERMAP_PLS)");
+
+ /* ... wait for mapping ... */
+
+ if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with parent: read(SYNC_USERMAP_ACK)");
+ if (s != SYNC_USERMAP_ACK)
+ bail("failed to sync with parent: SYNC_USERMAP_ACK: got %u", s);
+ /* Switching is only necessary if we joined namespaces. */
+ if (config.namespaces) {
+ if (prctl(PR_SET_DUMPABLE, 0, 0, 0, 0) < 0)
+ bail("failed to set process as dumpable");
+ }
+ }
+
+ /*
+ * TODO: What about non-namespace clone flags that we're dropping here?
+ *
+ * We fork again because of PID namespace, setns(2) or unshare(2) don't
+ * change the PID namespace of the calling process, because doing so
+ * would change the caller's idea of its own PID (as reported by getpid()),
+ * which would break many applications and libraries, so we must fork
+ * to actually enter the new PID namespace.
+ */
+ child = clone_parent(&env, JUMP_INIT);
+ if (child < 0)
+ bail("unable to fork: init_func");
+
+ /* Send the child to our parent, which knows what it's doing. */
+ s = SYNC_RECVPID_PLS;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with parent: write(SYNC_RECVPID_PLS)");
+ }
+ if (write(syncfd, &child, sizeof(child)) != sizeof(child)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with parent: write(childpid)");
+ }
+
+ /* ... wait for parent to get the pid ... */
+
+ if (read(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with parent: read(SYNC_RECVPID_ACK)");
+ }
+ if (s != SYNC_RECVPID_ACK) {
+ kill(child, SIGKILL);
+ bail("failed to sync with parent: SYNC_RECVPID_ACK: got %u", s);
+ }
+
+ s = SYNC_CHILD_READY;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s)) {
+ kill(child, SIGKILL);
+ bail("failed to sync with parent: write(SYNC_CHILD_READY)");
+ }
+
+ /* Our work is done. [Stage 2: JUMP_INIT] is doing the rest of the work. */
+ exit(0);
+ }
+
+ /*
+ * Stage 2: We're the final child process, and the only process that will
+ * actually return to the Go runtime. Our job is to just do the
+ * final cleanup steps and then return to the Go runtime to allow
+ * init_linux.go to run.
+ */
+ case JUMP_INIT: {
+ /*
+ * We're inside the child now, having jumped from the
+ * start_child() code after forking in the parent.
+ */
+ enum sync_t s;
+
+ /* We're in a child and thus need to tell the parent if we die. */
+ syncfd = sync_grandchild_pipe[0];
+ close(sync_grandchild_pipe[1]);
+ close(sync_child_pipe[0]);
+ close(sync_child_pipe[1]);
+
+ /* For debugging. */
+ prctl(PR_SET_NAME, (unsigned long) "runc:[2:INIT]", 0, 0, 0);
+
+ if (read(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with parent: read(SYNC_GRANDCHILD)");
+ if (s != SYNC_GRANDCHILD)
+ bail("failed to sync with parent: SYNC_GRANDCHILD: got %u", s);
+
+ if (setsid() < 0)
+ bail("setsid failed");
+
+ if (setuid(0) < 0)
+ bail("setuid failed");
+
+ if (setgid(0) < 0)
+ bail("setgid failed");
+
+ if (!config.is_rootless && config.is_setgroup) {
+ if (setgroups(0, NULL) < 0)
+ bail("setgroups failed");
+ }
+
+ s = SYNC_CHILD_READY;
+ if (write(syncfd, &s, sizeof(s)) != sizeof(s))
+ bail("failed to sync with patent: write(SYNC_CHILD_READY)");
+
+ /* Close sync pipes. */
+ close(sync_grandchild_pipe[0]);
+
+ /* Free netlink data. */
+ nl_free(&config);
+
+ /* Finish executing, let the Go runtime take over. */
+ return;
+ }
+ default:
+ bail("unexpected jump value");
+ }
+
+ /* Should never be reached. */
+ bail("should never be reached");
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/process.go b/vendor/github.com/opencontainers/runc/libcontainer/process.go
new file mode 100644
index 000000000..f1ad08149
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/process.go
@@ -0,0 +1,106 @@
+package libcontainer
+
+import (
+ "fmt"
+ "io"
+ "math"
+ "os"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+type processOperations interface {
+ wait() (*os.ProcessState, error)
+ signal(sig os.Signal) error
+ pid() int
+}
+
+// Process specifies the configuration and IO for a process inside
+// a container.
+type Process struct {
+ // The command to be run followed by any arguments.
+ Args []string
+
+ // Env specifies the environment variables for the process.
+ Env []string
+
+ // User will set the uid and gid of the executing process running inside the container
+ // local to the container's user and group configuration.
+ User string
+
+ // AdditionalGroups specifies the gids that should be added to supplementary groups
+ // in addition to those that the user belongs to.
+ AdditionalGroups []string
+
+ // Cwd will change the processes current working directory inside the container's rootfs.
+ Cwd string
+
+ // Stdin is a pointer to a reader which provides the standard input stream.
+ Stdin io.Reader
+
+ // Stdout is a pointer to a writer which receives the standard output stream.
+ Stdout io.Writer
+
+ // Stderr is a pointer to a writer which receives the standard error stream.
+ Stderr io.Writer
+
+ // ExtraFiles specifies additional open files to be inherited by the container
+ ExtraFiles []*os.File
+
+ // Capabilities specify the capabilities to keep when executing the process inside the container
+ // All capabilities not specified will be dropped from the processes capability mask
+ Capabilities *configs.Capabilities
+
+ // AppArmorProfile specifies the profile to apply to the process and is
+ // changed at the time the process is execed
+ AppArmorProfile string
+
+ // Label specifies the label to apply to the process. It is commonly used by selinux
+ Label string
+
+ // NoNewPrivileges controls whether processes can gain additional privileges.
+ NoNewPrivileges *bool
+
+ // Rlimits specifies the resource limits, such as max open files, to set in the container
+ // If Rlimits are not set, the container will inherit rlimits from the parent process
+ Rlimits []configs.Rlimit
+
+ // ConsoleSocket provides the masterfd console.
+ ConsoleSocket *os.File
+
+ ops processOperations
+}
+
+// Wait waits for the process to exit.
+// Wait releases any resources associated with the Process
+func (p Process) Wait() (*os.ProcessState, error) {
+ if p.ops == nil {
+ return nil, newGenericError(fmt.Errorf("invalid process"), NoProcessOps)
+ }
+ return p.ops.wait()
+}
+
+// Pid returns the process ID
+func (p Process) Pid() (int, error) {
+ // math.MinInt32 is returned here, because it's invalid value
+ // for the kill() system call.
+ if p.ops == nil {
+ return math.MinInt32, newGenericError(fmt.Errorf("invalid process"), NoProcessOps)
+ }
+ return p.ops.pid(), nil
+}
+
+// Signal sends a signal to the Process.
+func (p Process) Signal(sig os.Signal) error {
+ if p.ops == nil {
+ return newGenericError(fmt.Errorf("invalid process"), NoProcessOps)
+ }
+ return p.ops.signal(sig)
+}
+
+// IO holds the process's STDIO
+type IO struct {
+ Stdin io.WriteCloser
+ Stdout io.ReadCloser
+ Stderr io.ReadCloser
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/process_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/process_linux.go
new file mode 100644
index 000000000..171685ccd
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/process_linux.go
@@ -0,0 +1,493 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "encoding/json"
+ "errors"
+ "fmt"
+ "io"
+ "os"
+ "os/exec"
+ "path/filepath"
+ "strconv"
+ "syscall" // only for Signal
+
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/system"
+ "github.com/opencontainers/runc/libcontainer/utils"
+
+ "golang.org/x/sys/unix"
+)
+
+type parentProcess interface {
+ // pid returns the pid for the running process.
+ pid() int
+
+ // start starts the process execution.
+ start() error
+
+ // send a SIGKILL to the process and wait for the exit.
+ terminate() error
+
+ // wait waits on the process returning the process state.
+ wait() (*os.ProcessState, error)
+
+ // startTime returns the process start time.
+ startTime() (uint64, error)
+
+ signal(os.Signal) error
+
+ externalDescriptors() []string
+
+ setExternalDescriptors(fds []string)
+}
+
+type setnsProcess struct {
+ cmd *exec.Cmd
+ parentPipe *os.File
+ childPipe *os.File
+ cgroupPaths map[string]string
+ config *initConfig
+ fds []string
+ process *Process
+ bootstrapData io.Reader
+}
+
+func (p *setnsProcess) startTime() (uint64, error) {
+ stat, err := system.Stat(p.pid())
+ return stat.StartTime, err
+}
+
+func (p *setnsProcess) signal(sig os.Signal) error {
+ s, ok := sig.(syscall.Signal)
+ if !ok {
+ return errors.New("os: unsupported signal type")
+ }
+ return unix.Kill(p.pid(), s)
+}
+
+func (p *setnsProcess) start() (err error) {
+ defer p.parentPipe.Close()
+ err = p.cmd.Start()
+ p.childPipe.Close()
+ if err != nil {
+ return newSystemErrorWithCause(err, "starting setns process")
+ }
+ if p.bootstrapData != nil {
+ if _, err := io.Copy(p.parentPipe, p.bootstrapData); err != nil {
+ return newSystemErrorWithCause(err, "copying bootstrap data to pipe")
+ }
+ }
+ if err = p.execSetns(); err != nil {
+ return newSystemErrorWithCause(err, "executing setns process")
+ }
+ // We can't join cgroups if we're in a rootless container.
+ if !p.config.Rootless && len(p.cgroupPaths) > 0 {
+ if err := cgroups.EnterPid(p.cgroupPaths, p.pid()); err != nil {
+ return newSystemErrorWithCausef(err, "adding pid %d to cgroups", p.pid())
+ }
+ }
+ // set rlimits, this has to be done here because we lose permissions
+ // to raise the limits once we enter a user-namespace
+ if err := setupRlimits(p.config.Rlimits, p.pid()); err != nil {
+ return newSystemErrorWithCause(err, "setting rlimits for process")
+ }
+ if err := utils.WriteJSON(p.parentPipe, p.config); err != nil {
+ return newSystemErrorWithCause(err, "writing config to pipe")
+ }
+
+ ierr := parseSync(p.parentPipe, func(sync *syncT) error {
+ switch sync.Type {
+ case procReady:
+ // This shouldn't happen.
+ panic("unexpected procReady in setns")
+ case procHooks:
+ // This shouldn't happen.
+ panic("unexpected procHooks in setns")
+ default:
+ return newSystemError(fmt.Errorf("invalid JSON payload from child"))
+ }
+ })
+
+ if err := unix.Shutdown(int(p.parentPipe.Fd()), unix.SHUT_WR); err != nil {
+ return newSystemErrorWithCause(err, "calling shutdown on init pipe")
+ }
+ // Must be done after Shutdown so the child will exit and we can wait for it.
+ if ierr != nil {
+ p.wait()
+ return ierr
+ }
+ return nil
+}
+
+// execSetns runs the process that executes C code to perform the setns calls
+// because setns support requires the C process to fork off a child and perform the setns
+// before the go runtime boots, we wait on the process to die and receive the child's pid
+// over the provided pipe.
+func (p *setnsProcess) execSetns() error {
+ status, err := p.cmd.Process.Wait()
+ if err != nil {
+ p.cmd.Wait()
+ return newSystemErrorWithCause(err, "waiting on setns process to finish")
+ }
+ if !status.Success() {
+ p.cmd.Wait()
+ return newSystemError(&exec.ExitError{ProcessState: status})
+ }
+ var pid *pid
+ if err := json.NewDecoder(p.parentPipe).Decode(&pid); err != nil {
+ p.cmd.Wait()
+ return newSystemErrorWithCause(err, "reading pid from init pipe")
+ }
+ process, err := os.FindProcess(pid.Pid)
+ if err != nil {
+ return err
+ }
+ p.cmd.Process = process
+ p.process.ops = p
+ return nil
+}
+
+// terminate sends a SIGKILL to the forked process for the setns routine then waits to
+// avoid the process becoming a zombie.
+func (p *setnsProcess) terminate() error {
+ if p.cmd.Process == nil {
+ return nil
+ }
+ err := p.cmd.Process.Kill()
+ if _, werr := p.wait(); err == nil {
+ err = werr
+ }
+ return err
+}
+
+func (p *setnsProcess) wait() (*os.ProcessState, error) {
+ err := p.cmd.Wait()
+
+ // Return actual ProcessState even on Wait error
+ return p.cmd.ProcessState, err
+}
+
+func (p *setnsProcess) pid() int {
+ return p.cmd.Process.Pid
+}
+
+func (p *setnsProcess) externalDescriptors() []string {
+ return p.fds
+}
+
+func (p *setnsProcess) setExternalDescriptors(newFds []string) {
+ p.fds = newFds
+}
+
+type initProcess struct {
+ cmd *exec.Cmd
+ parentPipe *os.File
+ childPipe *os.File
+ config *initConfig
+ manager cgroups.Manager
+ container *linuxContainer
+ fds []string
+ process *Process
+ bootstrapData io.Reader
+ sharePidns bool
+ rootDir *os.File
+}
+
+func (p *initProcess) pid() int {
+ return p.cmd.Process.Pid
+}
+
+func (p *initProcess) externalDescriptors() []string {
+ return p.fds
+}
+
+// execSetns runs the process that executes C code to perform the setns calls
+// because setns support requires the C process to fork off a child and perform the setns
+// before the go runtime boots, we wait on the process to die and receive the child's pid
+// over the provided pipe.
+// This is called by initProcess.start function
+func (p *initProcess) execSetns() error {
+ status, err := p.cmd.Process.Wait()
+ if err != nil {
+ p.cmd.Wait()
+ return err
+ }
+ if !status.Success() {
+ p.cmd.Wait()
+ return &exec.ExitError{ProcessState: status}
+ }
+ var pid *pid
+ if err := json.NewDecoder(p.parentPipe).Decode(&pid); err != nil {
+ p.cmd.Wait()
+ return err
+ }
+ process, err := os.FindProcess(pid.Pid)
+ if err != nil {
+ return err
+ }
+ p.cmd.Process = process
+ p.process.ops = p
+ return nil
+}
+
+func (p *initProcess) start() error {
+ defer p.parentPipe.Close()
+ err := p.cmd.Start()
+ p.process.ops = p
+ p.childPipe.Close()
+ p.rootDir.Close()
+ if err != nil {
+ p.process.ops = nil
+ return newSystemErrorWithCause(err, "starting init process command")
+ }
+ if _, err := io.Copy(p.parentPipe, p.bootstrapData); err != nil {
+ return newSystemErrorWithCause(err, "copying bootstrap data to pipe")
+ }
+ if err := p.execSetns(); err != nil {
+ return newSystemErrorWithCause(err, "running exec setns process for init")
+ }
+ // Save the standard descriptor names before the container process
+ // can potentially move them (e.g., via dup2()). If we don't do this now,
+ // we won't know at checkpoint time which file descriptor to look up.
+ fds, err := getPipeFds(p.pid())
+ if err != nil {
+ return newSystemErrorWithCausef(err, "getting pipe fds for pid %d", p.pid())
+ }
+ p.setExternalDescriptors(fds)
+ // Do this before syncing with child so that no children can escape the
+ // cgroup. We don't need to worry about not doing this and not being root
+ // because we'd be using the rootless cgroup manager in that case.
+ if err := p.manager.Apply(p.pid()); err != nil {
+ return newSystemErrorWithCause(err, "applying cgroup configuration for process")
+ }
+ defer func() {
+ if err != nil {
+ // TODO: should not be the responsibility to call here
+ p.manager.Destroy()
+ }
+ }()
+ if err := p.createNetworkInterfaces(); err != nil {
+ return newSystemErrorWithCause(err, "creating network interfaces")
+ }
+ if err := p.sendConfig(); err != nil {
+ return newSystemErrorWithCause(err, "sending config to init process")
+ }
+ var (
+ sentRun bool
+ sentResume bool
+ )
+
+ ierr := parseSync(p.parentPipe, func(sync *syncT) error {
+ switch sync.Type {
+ case procReady:
+ // set rlimits, this has to be done here because we lose permissions
+ // to raise the limits once we enter a user-namespace
+ if err := setupRlimits(p.config.Rlimits, p.pid()); err != nil {
+ return newSystemErrorWithCause(err, "setting rlimits for ready process")
+ }
+ // call prestart hooks
+ if !p.config.Config.Namespaces.Contains(configs.NEWNS) {
+ // Setup cgroup before prestart hook, so that the prestart hook could apply cgroup permissions.
+ if err := p.manager.Set(p.config.Config); err != nil {
+ return newSystemErrorWithCause(err, "setting cgroup config for ready process")
+ }
+
+ if p.config.Config.Hooks != nil {
+ s := configs.HookState{
+ Version: p.container.config.Version,
+ ID: p.container.id,
+ Pid: p.pid(),
+ Bundle: utils.SearchLabels(p.config.Config.Labels, "bundle"),
+ }
+ for i, hook := range p.config.Config.Hooks.Prestart {
+ if err := hook.Run(s); err != nil {
+ return newSystemErrorWithCausef(err, "running prestart hook %d", i)
+ }
+ }
+ }
+ }
+ // Sync with child.
+ if err := writeSync(p.parentPipe, procRun); err != nil {
+ return newSystemErrorWithCause(err, "writing syncT 'run'")
+ }
+ sentRun = true
+ case procHooks:
+ // Setup cgroup before prestart hook, so that the prestart hook could apply cgroup permissions.
+ if err := p.manager.Set(p.config.Config); err != nil {
+ return newSystemErrorWithCause(err, "setting cgroup config for procHooks process")
+ }
+ if p.config.Config.Hooks != nil {
+ s := configs.HookState{
+ Version: p.container.config.Version,
+ ID: p.container.id,
+ Pid: p.pid(),
+ Bundle: utils.SearchLabels(p.config.Config.Labels, "bundle"),
+ }
+ for i, hook := range p.config.Config.Hooks.Prestart {
+ if err := hook.Run(s); err != nil {
+ return newSystemErrorWithCausef(err, "running prestart hook %d", i)
+ }
+ }
+ }
+ // Sync with child.
+ if err := writeSync(p.parentPipe, procResume); err != nil {
+ return newSystemErrorWithCause(err, "writing syncT 'resume'")
+ }
+ sentResume = true
+ default:
+ return newSystemError(fmt.Errorf("invalid JSON payload from child"))
+ }
+
+ return nil
+ })
+
+ if !sentRun {
+ return newSystemErrorWithCause(ierr, "container init")
+ }
+ if p.config.Config.Namespaces.Contains(configs.NEWNS) && !sentResume {
+ return newSystemError(fmt.Errorf("could not synchronise after executing prestart hooks with container process"))
+ }
+ if err := unix.Shutdown(int(p.parentPipe.Fd()), unix.SHUT_WR); err != nil {
+ return newSystemErrorWithCause(err, "shutting down init pipe")
+ }
+
+ // Must be done after Shutdown so the child will exit and we can wait for it.
+ if ierr != nil {
+ p.wait()
+ return ierr
+ }
+ return nil
+}
+
+func (p *initProcess) wait() (*os.ProcessState, error) {
+ err := p.cmd.Wait()
+ if err != nil {
+ return p.cmd.ProcessState, err
+ }
+ // we should kill all processes in cgroup when init is died if we use host PID namespace
+ if p.sharePidns {
+ signalAllProcesses(p.manager, unix.SIGKILL)
+ }
+ return p.cmd.ProcessState, nil
+}
+
+func (p *initProcess) terminate() error {
+ if p.cmd.Process == nil {
+ return nil
+ }
+ err := p.cmd.Process.Kill()
+ if _, werr := p.wait(); err == nil {
+ err = werr
+ }
+ return err
+}
+
+func (p *initProcess) startTime() (uint64, error) {
+ stat, err := system.Stat(p.pid())
+ return stat.StartTime, err
+}
+
+func (p *initProcess) sendConfig() error {
+ // send the config to the container's init process, we don't use JSON Encode
+ // here because there might be a problem in JSON decoder in some cases, see:
+ // https://github.com/docker/docker/issues/14203#issuecomment-174177790
+ return utils.WriteJSON(p.parentPipe, p.config)
+}
+
+func (p *initProcess) createNetworkInterfaces() error {
+ for _, config := range p.config.Config.Networks {
+ strategy, err := getStrategy(config.Type)
+ if err != nil {
+ return err
+ }
+ n := &network{
+ Network: *config,
+ }
+ if err := strategy.create(n, p.pid()); err != nil {
+ return err
+ }
+ p.config.Networks = append(p.config.Networks, n)
+ }
+ return nil
+}
+
+func (p *initProcess) signal(sig os.Signal) error {
+ s, ok := sig.(syscall.Signal)
+ if !ok {
+ return errors.New("os: unsupported signal type")
+ }
+ return unix.Kill(p.pid(), s)
+}
+
+func (p *initProcess) setExternalDescriptors(newFds []string) {
+ p.fds = newFds
+}
+
+func getPipeFds(pid int) ([]string, error) {
+ fds := make([]string, 3)
+
+ dirPath := filepath.Join("/proc", strconv.Itoa(pid), "/fd")
+ for i := 0; i < 3; i++ {
+ // XXX: This breaks if the path is not a valid symlink (which can
+ // happen in certain particularly unlucky mount namespace setups).
+ f := filepath.Join(dirPath, strconv.Itoa(i))
+ target, err := os.Readlink(f)
+ if err != nil {
+ // Ignore permission errors, for rootless containers and other
+ // non-dumpable processes. if we can't get the fd for a particular
+ // file, there's not much we can do.
+ if os.IsPermission(err) {
+ continue
+ }
+ return fds, err
+ }
+ fds[i] = target
+ }
+ return fds, nil
+}
+
+// InitializeIO creates pipes for use with the process's stdio and returns the
+// opposite side for each. Do not use this if you want to have a pseudoterminal
+// set up for you by libcontainer (TODO: fix that too).
+// TODO: This is mostly unnecessary, and should be handled by clients.
+func (p *Process) InitializeIO(rootuid, rootgid int) (i *IO, err error) {
+ var fds []uintptr
+ i = &IO{}
+ // cleanup in case of an error
+ defer func() {
+ if err != nil {
+ for _, fd := range fds {
+ unix.Close(int(fd))
+ }
+ }
+ }()
+ // STDIN
+ r, w, err := os.Pipe()
+ if err != nil {
+ return nil, err
+ }
+ fds = append(fds, r.Fd(), w.Fd())
+ p.Stdin, i.Stdin = r, w
+ // STDOUT
+ if r, w, err = os.Pipe(); err != nil {
+ return nil, err
+ }
+ fds = append(fds, r.Fd(), w.Fd())
+ p.Stdout, i.Stdout = w, r
+ // STDERR
+ if r, w, err = os.Pipe(); err != nil {
+ return nil, err
+ }
+ fds = append(fds, r.Fd(), w.Fd())
+ p.Stderr, i.Stderr = w, r
+ // change ownership of the pipes incase we are in a user namespace
+ for _, fd := range fds {
+ if err := unix.Fchown(int(fd), rootuid, rootgid); err != nil {
+ return nil, err
+ }
+ }
+ return i, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/restored_process.go b/vendor/github.com/opencontainers/runc/libcontainer/restored_process.go
new file mode 100644
index 000000000..408916ad9
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/restored_process.go
@@ -0,0 +1,122 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+
+ "github.com/opencontainers/runc/libcontainer/system"
+)
+
+func newRestoredProcess(pid int, fds []string) (*restoredProcess, error) {
+ var (
+ err error
+ )
+ proc, err := os.FindProcess(pid)
+ if err != nil {
+ return nil, err
+ }
+ stat, err := system.Stat(pid)
+ if err != nil {
+ return nil, err
+ }
+ return &restoredProcess{
+ proc: proc,
+ processStartTime: stat.StartTime,
+ fds: fds,
+ }, nil
+}
+
+type restoredProcess struct {
+ proc *os.Process
+ processStartTime uint64
+ fds []string
+}
+
+func (p *restoredProcess) start() error {
+ return newGenericError(fmt.Errorf("restored process cannot be started"), SystemError)
+}
+
+func (p *restoredProcess) pid() int {
+ return p.proc.Pid
+}
+
+func (p *restoredProcess) terminate() error {
+ err := p.proc.Kill()
+ if _, werr := p.wait(); err == nil {
+ err = werr
+ }
+ return err
+}
+
+func (p *restoredProcess) wait() (*os.ProcessState, error) {
+ // TODO: how do we wait on the actual process?
+ // maybe use --exec-cmd in criu
+ st, err := p.proc.Wait()
+ if err != nil {
+ return nil, err
+ }
+ return st, nil
+}
+
+func (p *restoredProcess) startTime() (uint64, error) {
+ return p.processStartTime, nil
+}
+
+func (p *restoredProcess) signal(s os.Signal) error {
+ return p.proc.Signal(s)
+}
+
+func (p *restoredProcess) externalDescriptors() []string {
+ return p.fds
+}
+
+func (p *restoredProcess) setExternalDescriptors(newFds []string) {
+ p.fds = newFds
+}
+
+// nonChildProcess represents a process where the calling process is not
+// the parent process. This process is created when a factory loads a container from
+// a persisted state.
+type nonChildProcess struct {
+ processPid int
+ processStartTime uint64
+ fds []string
+}
+
+func (p *nonChildProcess) start() error {
+ return newGenericError(fmt.Errorf("restored process cannot be started"), SystemError)
+}
+
+func (p *nonChildProcess) pid() int {
+ return p.processPid
+}
+
+func (p *nonChildProcess) terminate() error {
+ return newGenericError(fmt.Errorf("restored process cannot be terminated"), SystemError)
+}
+
+func (p *nonChildProcess) wait() (*os.ProcessState, error) {
+ return nil, newGenericError(fmt.Errorf("restored process cannot be waited on"), SystemError)
+}
+
+func (p *nonChildProcess) startTime() (uint64, error) {
+ return p.processStartTime, nil
+}
+
+func (p *nonChildProcess) signal(s os.Signal) error {
+ proc, err := os.FindProcess(p.processPid)
+ if err != nil {
+ return err
+ }
+ return proc.Signal(s)
+}
+
+func (p *nonChildProcess) externalDescriptors() []string {
+ return p.fds
+}
+
+func (p *nonChildProcess) setExternalDescriptors(newFds []string) {
+ p.fds = newFds
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/rootfs_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/rootfs_linux.go
new file mode 100644
index 000000000..e2e734a85
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/rootfs_linux.go
@@ -0,0 +1,812 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "io"
+ "io/ioutil"
+ "os"
+ "os/exec"
+ "path"
+ "path/filepath"
+ "strings"
+ "time"
+
+ "github.com/docker/docker/pkg/mount"
+ "github.com/docker/docker/pkg/symlink"
+ "github.com/mrunalp/fileutils"
+ "github.com/opencontainers/runc/libcontainer/cgroups"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/system"
+ libcontainerUtils "github.com/opencontainers/runc/libcontainer/utils"
+ "github.com/opencontainers/selinux/go-selinux/label"
+
+ "golang.org/x/sys/unix"
+)
+
+const defaultMountFlags = unix.MS_NOEXEC | unix.MS_NOSUID | unix.MS_NODEV
+
+// needsSetupDev returns true if /dev needs to be set up.
+func needsSetupDev(config *configs.Config) bool {
+ for _, m := range config.Mounts {
+ if m.Device == "bind" && libcontainerUtils.CleanPath(m.Destination) == "/dev" {
+ return false
+ }
+ }
+ return true
+}
+
+// prepareRootfs sets up the devices, mount points, and filesystems for use
+// inside a new mount namespace. It doesn't set anything as ro. You must call
+// finalizeRootfs after this function to finish setting up the rootfs.
+func prepareRootfs(pipe io.ReadWriter, config *configs.Config) (err error) {
+ if err := prepareRoot(config); err != nil {
+ return newSystemErrorWithCause(err, "preparing rootfs")
+ }
+
+ setupDev := needsSetupDev(config)
+ for _, m := range config.Mounts {
+ for _, precmd := range m.PremountCmds {
+ if err := mountCmd(precmd); err != nil {
+ return newSystemErrorWithCause(err, "running premount command")
+ }
+ }
+
+ if err := mountToRootfs(m, config.Rootfs, config.MountLabel); err != nil {
+ return newSystemErrorWithCausef(err, "mounting %q to rootfs %q at %q", m.Source, config.Rootfs, m.Destination)
+ }
+
+ for _, postcmd := range m.PostmountCmds {
+ if err := mountCmd(postcmd); err != nil {
+ return newSystemErrorWithCause(err, "running postmount command")
+ }
+ }
+ }
+
+ if setupDev {
+ if err := createDevices(config); err != nil {
+ return newSystemErrorWithCause(err, "creating device nodes")
+ }
+ if err := setupPtmx(config); err != nil {
+ return newSystemErrorWithCause(err, "setting up ptmx")
+ }
+ if err := setupDevSymlinks(config.Rootfs); err != nil {
+ return newSystemErrorWithCause(err, "setting up /dev symlinks")
+ }
+ }
+
+ // Signal the parent to run the pre-start hooks.
+ // The hooks are run after the mounts are setup, but before we switch to the new
+ // root, so that the old root is still available in the hooks for any mount
+ // manipulations.
+ if err := syncParentHooks(pipe); err != nil {
+ return err
+ }
+
+ // The reason these operations are done here rather than in finalizeRootfs
+ // is because the console-handling code gets quite sticky if we have to set
+ // up the console before doing the pivot_root(2). This is because the
+ // Console API has to also work with the ExecIn case, which means that the
+ // API must be able to deal with being inside as well as outside the
+ // container. It's just cleaner to do this here (at the expense of the
+ // operation not being perfectly split).
+
+ if err := unix.Chdir(config.Rootfs); err != nil {
+ return newSystemErrorWithCausef(err, "changing dir to %q", config.Rootfs)
+ }
+
+ if config.NoPivotRoot {
+ err = msMoveRoot(config.Rootfs)
+ } else {
+ err = pivotRoot(config.Rootfs)
+ }
+ if err != nil {
+ return newSystemErrorWithCause(err, "jailing process inside rootfs")
+ }
+
+ if setupDev {
+ if err := reOpenDevNull(); err != nil {
+ return newSystemErrorWithCause(err, "reopening /dev/null inside container")
+ }
+ }
+
+ return nil
+}
+
+// finalizeRootfs sets anything to ro if necessary. You must call
+// prepareRootfs first.
+func finalizeRootfs(config *configs.Config) (err error) {
+ // remount dev as ro if specified
+ for _, m := range config.Mounts {
+ if libcontainerUtils.CleanPath(m.Destination) == "/dev" {
+ if m.Flags&unix.MS_RDONLY == unix.MS_RDONLY {
+ if err := remountReadonly(m); err != nil {
+ return newSystemErrorWithCausef(err, "remounting %q as readonly", m.Destination)
+ }
+ }
+ break
+ }
+ }
+
+ // set rootfs ( / ) as readonly
+ if config.Readonlyfs {
+ if err := setReadonly(); err != nil {
+ return newSystemErrorWithCause(err, "setting rootfs as readonly")
+ }
+ }
+
+ unix.Umask(0022)
+ return nil
+}
+
+func mountCmd(cmd configs.Command) error {
+ command := exec.Command(cmd.Path, cmd.Args[:]...)
+ command.Env = cmd.Env
+ command.Dir = cmd.Dir
+ if out, err := command.CombinedOutput(); err != nil {
+ return fmt.Errorf("%#v failed: %s: %v", cmd, string(out), err)
+ }
+ return nil
+}
+
+func mountToRootfs(m *configs.Mount, rootfs, mountLabel string) error {
+ var (
+ dest = m.Destination
+ )
+ if !strings.HasPrefix(dest, rootfs) {
+ dest = filepath.Join(rootfs, dest)
+ }
+
+ switch m.Device {
+ case "proc", "sysfs":
+ if err := os.MkdirAll(dest, 0755); err != nil {
+ return err
+ }
+ // Selinux kernels do not support labeling of /proc or /sys
+ return mountPropagate(m, rootfs, "")
+ case "mqueue":
+ if err := os.MkdirAll(dest, 0755); err != nil {
+ return err
+ }
+ if err := mountPropagate(m, rootfs, mountLabel); err != nil {
+ // older kernels do not support labeling of /dev/mqueue
+ if err := mountPropagate(m, rootfs, ""); err != nil {
+ return err
+ }
+ return label.SetFileLabel(dest, mountLabel)
+ }
+ return nil
+ case "tmpfs":
+ copyUp := m.Extensions&configs.EXT_COPYUP == configs.EXT_COPYUP
+ tmpDir := ""
+ stat, err := os.Stat(dest)
+ if err != nil {
+ if err := os.MkdirAll(dest, 0755); err != nil {
+ return err
+ }
+ }
+ if copyUp {
+ tmpDir, err = ioutil.TempDir("/tmp", "runctmpdir")
+ if err != nil {
+ return newSystemErrorWithCause(err, "tmpcopyup: failed to create tmpdir")
+ }
+ defer os.RemoveAll(tmpDir)
+ m.Destination = tmpDir
+ }
+ if err := mountPropagate(m, rootfs, mountLabel); err != nil {
+ return err
+ }
+ if copyUp {
+ if err := fileutils.CopyDirectory(dest, tmpDir); err != nil {
+ errMsg := fmt.Errorf("tmpcopyup: failed to copy %s to %s: %v", dest, tmpDir, err)
+ if err1 := unix.Unmount(tmpDir, unix.MNT_DETACH); err1 != nil {
+ return newSystemErrorWithCausef(err1, "tmpcopyup: %v: failed to unmount", errMsg)
+ }
+ return errMsg
+ }
+ if err := unix.Mount(tmpDir, dest, "", unix.MS_MOVE, ""); err != nil {
+ errMsg := fmt.Errorf("tmpcopyup: failed to move mount %s to %s: %v", tmpDir, dest, err)
+ if err1 := unix.Unmount(tmpDir, unix.MNT_DETACH); err1 != nil {
+ return newSystemErrorWithCausef(err1, "tmpcopyup: %v: failed to unmount", errMsg)
+ }
+ return errMsg
+ }
+ }
+ if stat != nil {
+ if err = os.Chmod(dest, stat.Mode()); err != nil {
+ return err
+ }
+ }
+ return nil
+ case "bind":
+ stat, err := os.Stat(m.Source)
+ if err != nil {
+ // error out if the source of a bind mount does not exist as we will be
+ // unable to bind anything to it.
+ return err
+ }
+ // ensure that the destination of the bind mount is resolved of symlinks at mount time because
+ // any previous mounts can invalidate the next mount's destination.
+ // this can happen when a user specifies mounts within other mounts to cause breakouts or other
+ // evil stuff to try to escape the container's rootfs.
+ if dest, err = symlink.FollowSymlinkInScope(dest, rootfs); err != nil {
+ return err
+ }
+ if err := checkMountDestination(rootfs, dest); err != nil {
+ return err
+ }
+ // update the mount with the correct dest after symlinks are resolved.
+ m.Destination = dest
+ if err := createIfNotExists(dest, stat.IsDir()); err != nil {
+ return err
+ }
+ if err := mountPropagate(m, rootfs, mountLabel); err != nil {
+ return err
+ }
+ // bind mount won't change mount options, we need remount to make mount options effective.
+ // first check that we have non-default options required before attempting a remount
+ if m.Flags&^(unix.MS_REC|unix.MS_REMOUNT|unix.MS_BIND) != 0 {
+ // only remount if unique mount options are set
+ if err := remount(m, rootfs); err != nil {
+ return err
+ }
+ }
+
+ if m.Relabel != "" {
+ if err := label.Validate(m.Relabel); err != nil {
+ return err
+ }
+ shared := label.IsShared(m.Relabel)
+ if err := label.Relabel(m.Source, mountLabel, shared); err != nil {
+ return err
+ }
+ }
+ case "cgroup":
+ binds, err := getCgroupMounts(m)
+ if err != nil {
+ return err
+ }
+ var merged []string
+ for _, b := range binds {
+ ss := filepath.Base(b.Destination)
+ if strings.Contains(ss, ",") {
+ merged = append(merged, ss)
+ }
+ }
+ tmpfs := &configs.Mount{
+ Source: "tmpfs",
+ Device: "tmpfs",
+ Destination: m.Destination,
+ Flags: defaultMountFlags,
+ Data: "mode=755",
+ PropagationFlags: m.PropagationFlags,
+ }
+ if err := mountToRootfs(tmpfs, rootfs, mountLabel); err != nil {
+ return err
+ }
+ for _, b := range binds {
+ if err := mountToRootfs(b, rootfs, mountLabel); err != nil {
+ return err
+ }
+ }
+ for _, mc := range merged {
+ for _, ss := range strings.Split(mc, ",") {
+ // symlink(2) is very dumb, it will just shove the path into
+ // the link and doesn't do any checks or relative path
+ // conversion. Also, don't error out if the cgroup already exists.
+ if err := os.Symlink(mc, filepath.Join(rootfs, m.Destination, ss)); err != nil && !os.IsExist(err) {
+ return err
+ }
+ }
+ }
+ if m.Flags&unix.MS_RDONLY != 0 {
+ // remount cgroup root as readonly
+ mcgrouproot := &configs.Mount{
+ Source: m.Destination,
+ Device: "bind",
+ Destination: m.Destination,
+ Flags: defaultMountFlags | unix.MS_RDONLY | unix.MS_BIND,
+ }
+ if err := remount(mcgrouproot, rootfs); err != nil {
+ return err
+ }
+ }
+ default:
+ // ensure that the destination of the mount is resolved of symlinks at mount time because
+ // any previous mounts can invalidate the next mount's destination.
+ // this can happen when a user specifies mounts within other mounts to cause breakouts or other
+ // evil stuff to try to escape the container's rootfs.
+ var err error
+ if dest, err = symlink.FollowSymlinkInScope(dest, rootfs); err != nil {
+ return err
+ }
+ if err := checkMountDestination(rootfs, dest); err != nil {
+ return err
+ }
+ // update the mount with the correct dest after symlinks are resolved.
+ m.Destination = dest
+ if err := os.MkdirAll(dest, 0755); err != nil {
+ return err
+ }
+ return mountPropagate(m, rootfs, mountLabel)
+ }
+ return nil
+}
+
+func getCgroupMounts(m *configs.Mount) ([]*configs.Mount, error) {
+ mounts, err := cgroups.GetCgroupMounts(false)
+ if err != nil {
+ return nil, err
+ }
+
+ cgroupPaths, err := cgroups.ParseCgroupFile("/proc/self/cgroup")
+ if err != nil {
+ return nil, err
+ }
+
+ var binds []*configs.Mount
+
+ for _, mm := range mounts {
+ dir, err := mm.GetOwnCgroup(cgroupPaths)
+ if err != nil {
+ return nil, err
+ }
+ relDir, err := filepath.Rel(mm.Root, dir)
+ if err != nil {
+ return nil, err
+ }
+ binds = append(binds, &configs.Mount{
+ Device: "bind",
+ Source: filepath.Join(mm.Mountpoint, relDir),
+ Destination: filepath.Join(m.Destination, filepath.Base(mm.Mountpoint)),
+ Flags: unix.MS_BIND | unix.MS_REC | m.Flags,
+ PropagationFlags: m.PropagationFlags,
+ })
+ }
+
+ return binds, nil
+}
+
+// checkMountDestination checks to ensure that the mount destination is not over the top of /proc.
+// dest is required to be an abs path and have any symlinks resolved before calling this function.
+func checkMountDestination(rootfs, dest string) error {
+ invalidDestinations := []string{
+ "/proc",
+ }
+ // White list, it should be sub directories of invalid destinations
+ validDestinations := []string{
+ // These entries can be bind mounted by files emulated by fuse,
+ // so commands like top, free displays stats in container.
+ "/proc/cpuinfo",
+ "/proc/diskstats",
+ "/proc/meminfo",
+ "/proc/stat",
+ "/proc/swaps",
+ "/proc/uptime",
+ "/proc/net/dev",
+ }
+ for _, valid := range validDestinations {
+ path, err := filepath.Rel(filepath.Join(rootfs, valid), dest)
+ if err != nil {
+ return err
+ }
+ if path == "." {
+ return nil
+ }
+ }
+ for _, invalid := range invalidDestinations {
+ path, err := filepath.Rel(filepath.Join(rootfs, invalid), dest)
+ if err != nil {
+ return err
+ }
+ if path == "." || !strings.HasPrefix(path, "..") {
+ return fmt.Errorf("%q cannot be mounted because it is located inside %q", dest, invalid)
+ }
+ }
+ return nil
+}
+
+func setupDevSymlinks(rootfs string) error {
+ var links = [][2]string{
+ {"/proc/self/fd", "/dev/fd"},
+ {"/proc/self/fd/0", "/dev/stdin"},
+ {"/proc/self/fd/1", "/dev/stdout"},
+ {"/proc/self/fd/2", "/dev/stderr"},
+ }
+ // kcore support can be toggled with CONFIG_PROC_KCORE; only create a symlink
+ // in /dev if it exists in /proc.
+ if _, err := os.Stat("/proc/kcore"); err == nil {
+ links = append(links, [2]string{"/proc/kcore", "/dev/core"})
+ }
+ for _, link := range links {
+ var (
+ src = link[0]
+ dst = filepath.Join(rootfs, link[1])
+ )
+ if err := os.Symlink(src, dst); err != nil && !os.IsExist(err) {
+ return fmt.Errorf("symlink %s %s %s", src, dst, err)
+ }
+ }
+ return nil
+}
+
+// If stdin, stdout, and/or stderr are pointing to `/dev/null` in the parent's rootfs
+// this method will make them point to `/dev/null` in this container's rootfs. This
+// needs to be called after we chroot/pivot into the container's rootfs so that any
+// symlinks are resolved locally.
+func reOpenDevNull() error {
+ var stat, devNullStat unix.Stat_t
+ file, err := os.OpenFile("/dev/null", os.O_RDWR, 0)
+ if err != nil {
+ return fmt.Errorf("Failed to open /dev/null - %s", err)
+ }
+ defer file.Close()
+ if err := unix.Fstat(int(file.Fd()), &devNullStat); err != nil {
+ return err
+ }
+ for fd := 0; fd < 3; fd++ {
+ if err := unix.Fstat(fd, &stat); err != nil {
+ return err
+ }
+ if stat.Rdev == devNullStat.Rdev {
+ // Close and re-open the fd.
+ if err := unix.Dup3(int(file.Fd()), fd, 0); err != nil {
+ return err
+ }
+ }
+ }
+ return nil
+}
+
+// Create the device nodes in the container.
+func createDevices(config *configs.Config) error {
+ useBindMount := system.RunningInUserNS() || config.Namespaces.Contains(configs.NEWUSER)
+ oldMask := unix.Umask(0000)
+ for _, node := range config.Devices {
+ // containers running in a user namespace are not allowed to mknod
+ // devices so we can just bind mount it from the host.
+ if err := createDeviceNode(config.Rootfs, node, useBindMount); err != nil {
+ unix.Umask(oldMask)
+ return err
+ }
+ }
+ unix.Umask(oldMask)
+ return nil
+}
+
+func bindMountDeviceNode(dest string, node *configs.Device) error {
+ f, err := os.Create(dest)
+ if err != nil && !os.IsExist(err) {
+ return err
+ }
+ if f != nil {
+ f.Close()
+ }
+ return unix.Mount(node.Path, dest, "bind", unix.MS_BIND, "")
+}
+
+// Creates the device node in the rootfs of the container.
+func createDeviceNode(rootfs string, node *configs.Device, bind bool) error {
+ dest := filepath.Join(rootfs, node.Path)
+ if err := os.MkdirAll(filepath.Dir(dest), 0755); err != nil {
+ return err
+ }
+
+ if bind {
+ return bindMountDeviceNode(dest, node)
+ }
+ if err := mknodDevice(dest, node); err != nil {
+ if os.IsExist(err) {
+ return nil
+ } else if os.IsPermission(err) {
+ return bindMountDeviceNode(dest, node)
+ }
+ return err
+ }
+ return nil
+}
+
+func mknodDevice(dest string, node *configs.Device) error {
+ fileMode := node.FileMode
+ switch node.Type {
+ case 'c', 'u':
+ fileMode |= unix.S_IFCHR
+ case 'b':
+ fileMode |= unix.S_IFBLK
+ case 'p':
+ fileMode |= unix.S_IFIFO
+ default:
+ return fmt.Errorf("%c is not a valid device type for device %s", node.Type, node.Path)
+ }
+ if err := unix.Mknod(dest, uint32(fileMode), node.Mkdev()); err != nil {
+ return err
+ }
+ return unix.Chown(dest, int(node.Uid), int(node.Gid))
+}
+
+func getMountInfo(mountinfo []*mount.Info, dir string) *mount.Info {
+ for _, m := range mountinfo {
+ if m.Mountpoint == dir {
+ return m
+ }
+ }
+ return nil
+}
+
+// Get the parent mount point of directory passed in as argument. Also return
+// optional fields.
+func getParentMount(rootfs string) (string, string, error) {
+ var path string
+
+ mountinfos, err := mount.GetMounts()
+ if err != nil {
+ return "", "", err
+ }
+
+ mountinfo := getMountInfo(mountinfos, rootfs)
+ if mountinfo != nil {
+ return rootfs, mountinfo.Optional, nil
+ }
+
+ path = rootfs
+ for {
+ path = filepath.Dir(path)
+
+ mountinfo = getMountInfo(mountinfos, path)
+ if mountinfo != nil {
+ return path, mountinfo.Optional, nil
+ }
+
+ if path == "/" {
+ break
+ }
+ }
+
+ // If we are here, we did not find parent mount. Something is wrong.
+ return "", "", fmt.Errorf("Could not find parent mount of %s", rootfs)
+}
+
+// Make parent mount private if it was shared
+func rootfsParentMountPrivate(rootfs string) error {
+ sharedMount := false
+
+ parentMount, optionalOpts, err := getParentMount(rootfs)
+ if err != nil {
+ return err
+ }
+
+ optsSplit := strings.Split(optionalOpts, " ")
+ for _, opt := range optsSplit {
+ if strings.HasPrefix(opt, "shared:") {
+ sharedMount = true
+ break
+ }
+ }
+
+ // Make parent mount PRIVATE if it was shared. It is needed for two
+ // reasons. First of all pivot_root() will fail if parent mount is
+ // shared. Secondly when we bind mount rootfs it will propagate to
+ // parent namespace and we don't want that to happen.
+ if sharedMount {
+ return unix.Mount("", parentMount, "", unix.MS_PRIVATE, "")
+ }
+
+ return nil
+}
+
+func prepareRoot(config *configs.Config) error {
+ flag := unix.MS_SLAVE | unix.MS_REC
+ if config.RootPropagation != 0 {
+ flag = config.RootPropagation
+ }
+ if err := unix.Mount("", "/", "", uintptr(flag), ""); err != nil {
+ return err
+ }
+
+ // Make parent mount private to make sure following bind mount does
+ // not propagate in other namespaces. Also it will help with kernel
+ // check pass in pivot_root. (IS_SHARED(new_mnt->mnt_parent))
+ if err := rootfsParentMountPrivate(config.Rootfs); err != nil {
+ return err
+ }
+
+ return unix.Mount(config.Rootfs, config.Rootfs, "bind", unix.MS_BIND|unix.MS_REC, "")
+}
+
+func setReadonly() error {
+ return unix.Mount("/", "/", "bind", unix.MS_BIND|unix.MS_REMOUNT|unix.MS_RDONLY|unix.MS_REC, "")
+}
+
+func setupPtmx(config *configs.Config) error {
+ ptmx := filepath.Join(config.Rootfs, "dev/ptmx")
+ if err := os.Remove(ptmx); err != nil && !os.IsNotExist(err) {
+ return err
+ }
+ if err := os.Symlink("pts/ptmx", ptmx); err != nil {
+ return fmt.Errorf("symlink dev ptmx %s", err)
+ }
+ return nil
+}
+
+// pivotRoot will call pivot_root such that rootfs becomes the new root
+// filesystem, and everything else is cleaned up.
+func pivotRoot(rootfs string) error {
+ // While the documentation may claim otherwise, pivot_root(".", ".") is
+ // actually valid. What this results in is / being the new root but
+ // /proc/self/cwd being the old root. Since we can play around with the cwd
+ // with pivot_root this allows us to pivot without creating directories in
+ // the rootfs. Shout-outs to the LXC developers for giving us this idea.
+
+ oldroot, err := unix.Open("/", unix.O_DIRECTORY|unix.O_RDONLY, 0)
+ if err != nil {
+ return err
+ }
+ defer unix.Close(oldroot)
+
+ newroot, err := unix.Open(rootfs, unix.O_DIRECTORY|unix.O_RDONLY, 0)
+ if err != nil {
+ return err
+ }
+ defer unix.Close(newroot)
+
+ // Change to the new root so that the pivot_root actually acts on it.
+ if err := unix.Fchdir(newroot); err != nil {
+ return err
+ }
+
+ if err := unix.PivotRoot(".", "."); err != nil {
+ return fmt.Errorf("pivot_root %s", err)
+ }
+
+ // Currently our "." is oldroot (according to the current kernel code).
+ // However, purely for safety, we will fchdir(oldroot) since there isn't
+ // really any guarantee from the kernel what /proc/self/cwd will be after a
+ // pivot_root(2).
+
+ if err := unix.Fchdir(oldroot); err != nil {
+ return err
+ }
+
+ // Make oldroot rprivate to make sure our unmounts don't propagate to the
+ // host (and thus bork the machine).
+ if err := unix.Mount("", ".", "", unix.MS_PRIVATE|unix.MS_REC, ""); err != nil {
+ return err
+ }
+ // Preform the unmount. MNT_DETACH allows us to unmount /proc/self/cwd.
+ if err := unix.Unmount(".", unix.MNT_DETACH); err != nil {
+ return err
+ }
+
+ // Switch back to our shiny new root.
+ if err := unix.Chdir("/"); err != nil {
+ return fmt.Errorf("chdir / %s", err)
+ }
+ return nil
+}
+
+func msMoveRoot(rootfs string) error {
+ if err := unix.Mount(rootfs, "/", "", unix.MS_MOVE, ""); err != nil {
+ return err
+ }
+ if err := unix.Chroot("."); err != nil {
+ return err
+ }
+ return unix.Chdir("/")
+}
+
+// createIfNotExists creates a file or a directory only if it does not already exist.
+func createIfNotExists(path string, isDir bool) error {
+ if _, err := os.Stat(path); err != nil {
+ if os.IsNotExist(err) {
+ if isDir {
+ return os.MkdirAll(path, 0755)
+ }
+ if err := os.MkdirAll(filepath.Dir(path), 0755); err != nil {
+ return err
+ }
+ f, err := os.OpenFile(path, os.O_CREATE, 0755)
+ if err != nil {
+ return err
+ }
+ f.Close()
+ }
+ }
+ return nil
+}
+
+// readonlyPath will make a path read only.
+func readonlyPath(path string) error {
+ if err := unix.Mount(path, path, "", unix.MS_BIND|unix.MS_REC, ""); err != nil {
+ if os.IsNotExist(err) {
+ return nil
+ }
+ return err
+ }
+ return unix.Mount(path, path, "", unix.MS_BIND|unix.MS_REMOUNT|unix.MS_RDONLY|unix.MS_REC, "")
+}
+
+// remountReadonly will remount an existing mount point and ensure that it is read-only.
+func remountReadonly(m *configs.Mount) error {
+ var (
+ dest = m.Destination
+ flags = m.Flags
+ )
+ for i := 0; i < 5; i++ {
+ if err := unix.Mount("", dest, "", uintptr(flags|unix.MS_REMOUNT|unix.MS_RDONLY), ""); err != nil {
+ switch err {
+ case unix.EBUSY:
+ time.Sleep(100 * time.Millisecond)
+ continue
+ default:
+ return err
+ }
+ }
+ return nil
+ }
+ return fmt.Errorf("unable to mount %s as readonly max retries reached", dest)
+}
+
+// maskPath masks the top of the specified path inside a container to avoid
+// security issues from processes reading information from non-namespace aware
+// mounts ( proc/kcore ).
+// For files, maskPath bind mounts /dev/null over the top of the specified path.
+// For directories, maskPath mounts read-only tmpfs over the top of the specified path.
+func maskPath(path string) error {
+ if err := unix.Mount("/dev/null", path, "", unix.MS_BIND, ""); err != nil && !os.IsNotExist(err) {
+ if err == unix.ENOTDIR {
+ return unix.Mount("tmpfs", path, "tmpfs", unix.MS_RDONLY, "")
+ }
+ return err
+ }
+ return nil
+}
+
+// writeSystemProperty writes the value to a path under /proc/sys as determined from the key.
+// For e.g. net.ipv4.ip_forward translated to /proc/sys/net/ipv4/ip_forward.
+func writeSystemProperty(key, value string) error {
+ keyPath := strings.Replace(key, ".", "/", -1)
+ return ioutil.WriteFile(path.Join("/proc/sys", keyPath), []byte(value), 0644)
+}
+
+func remount(m *configs.Mount, rootfs string) error {
+ var (
+ dest = m.Destination
+ )
+ if !strings.HasPrefix(dest, rootfs) {
+ dest = filepath.Join(rootfs, dest)
+ }
+ if err := unix.Mount(m.Source, dest, m.Device, uintptr(m.Flags|unix.MS_REMOUNT), ""); err != nil {
+ return err
+ }
+ return nil
+}
+
+// Do the mount operation followed by additional mounts required to take care
+// of propagation flags.
+func mountPropagate(m *configs.Mount, rootfs string, mountLabel string) error {
+ var (
+ dest = m.Destination
+ data = label.FormatMountLabel(m.Data, mountLabel)
+ flags = m.Flags
+ )
+ if libcontainerUtils.CleanPath(dest) == "/dev" {
+ flags &= ^unix.MS_RDONLY
+ }
+
+ copyUp := m.Extensions&configs.EXT_COPYUP == configs.EXT_COPYUP
+ if !(copyUp || strings.HasPrefix(dest, rootfs)) {
+ dest = filepath.Join(rootfs, dest)
+ }
+
+ if err := unix.Mount(m.Source, dest, m.Device, uintptr(flags), data); err != nil {
+ return err
+ }
+
+ for _, pflag := range m.PropagationFlags {
+ if err := unix.Mount("", dest, "", uintptr(pflag), ""); err != nil {
+ return err
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/seccomp/config.go b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/config.go
new file mode 100644
index 000000000..ded5a6bbc
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/config.go
@@ -0,0 +1,76 @@
+package seccomp
+
+import (
+ "fmt"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+var operators = map[string]configs.Operator{
+ "SCMP_CMP_NE": configs.NotEqualTo,
+ "SCMP_CMP_LT": configs.LessThan,
+ "SCMP_CMP_LE": configs.LessThanOrEqualTo,
+ "SCMP_CMP_EQ": configs.EqualTo,
+ "SCMP_CMP_GE": configs.GreaterThanOrEqualTo,
+ "SCMP_CMP_GT": configs.GreaterThan,
+ "SCMP_CMP_MASKED_EQ": configs.MaskEqualTo,
+}
+
+var actions = map[string]configs.Action{
+ "SCMP_ACT_KILL": configs.Kill,
+ "SCMP_ACT_ERRNO": configs.Errno,
+ "SCMP_ACT_TRAP": configs.Trap,
+ "SCMP_ACT_ALLOW": configs.Allow,
+ "SCMP_ACT_TRACE": configs.Trace,
+}
+
+var archs = map[string]string{
+ "SCMP_ARCH_X86": "x86",
+ "SCMP_ARCH_X86_64": "amd64",
+ "SCMP_ARCH_X32": "x32",
+ "SCMP_ARCH_ARM": "arm",
+ "SCMP_ARCH_AARCH64": "arm64",
+ "SCMP_ARCH_MIPS": "mips",
+ "SCMP_ARCH_MIPS64": "mips64",
+ "SCMP_ARCH_MIPS64N32": "mips64n32",
+ "SCMP_ARCH_MIPSEL": "mipsel",
+ "SCMP_ARCH_MIPSEL64": "mipsel64",
+ "SCMP_ARCH_MIPSEL64N32": "mipsel64n32",
+ "SCMP_ARCH_PPC": "ppc",
+ "SCMP_ARCH_PPC64": "ppc64",
+ "SCMP_ARCH_PPC64LE": "ppc64le",
+ "SCMP_ARCH_S390": "s390",
+ "SCMP_ARCH_S390X": "s390x",
+}
+
+// ConvertStringToOperator converts a string into a Seccomp comparison operator.
+// Comparison operators use the names they are assigned by Libseccomp's header.
+// Attempting to convert a string that is not a valid operator results in an
+// error.
+func ConvertStringToOperator(in string) (configs.Operator, error) {
+ if op, ok := operators[in]; ok == true {
+ return op, nil
+ }
+ return 0, fmt.Errorf("string %s is not a valid operator for seccomp", in)
+}
+
+// ConvertStringToAction converts a string into a Seccomp rule match action.
+// Actions use the names they are assigned in Libseccomp's header, though some
+// (notable, SCMP_ACT_TRACE) are not available in this implementation and will
+// return errors.
+// Attempting to convert a string that is not a valid action results in an
+// error.
+func ConvertStringToAction(in string) (configs.Action, error) {
+ if act, ok := actions[in]; ok == true {
+ return act, nil
+ }
+ return 0, fmt.Errorf("string %s is not a valid action for seccomp", in)
+}
+
+// ConvertStringToArch converts a string into a Seccomp comparison arch.
+func ConvertStringToArch(in string) (string, error) {
+ if arch, ok := archs[in]; ok == true {
+ return arch, nil
+ }
+ return "", fmt.Errorf("string %s is not a valid arch for seccomp", in)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_linux.go
new file mode 100644
index 000000000..2523cbf99
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_linux.go
@@ -0,0 +1,227 @@
+// +build linux,cgo,seccomp
+
+package seccomp
+
+import (
+ "bufio"
+ "fmt"
+ "os"
+ "strings"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+ libseccomp "github.com/seccomp/libseccomp-golang"
+
+ "golang.org/x/sys/unix"
+)
+
+var (
+ actAllow = libseccomp.ActAllow
+ actTrap = libseccomp.ActTrap
+ actKill = libseccomp.ActKill
+ actTrace = libseccomp.ActTrace.SetReturnCode(int16(unix.EPERM))
+ actErrno = libseccomp.ActErrno.SetReturnCode(int16(unix.EPERM))
+)
+
+// Filters given syscalls in a container, preventing them from being used
+// Started in the container init process, and carried over to all child processes
+// Setns calls, however, require a separate invocation, as they are not children
+// of the init until they join the namespace
+func InitSeccomp(config *configs.Seccomp) error {
+ if config == nil {
+ return fmt.Errorf("cannot initialize Seccomp - nil config passed")
+ }
+
+ defaultAction, err := getAction(config.DefaultAction)
+ if err != nil {
+ return fmt.Errorf("error initializing seccomp - invalid default action")
+ }
+
+ filter, err := libseccomp.NewFilter(defaultAction)
+ if err != nil {
+ return fmt.Errorf("error creating filter: %s", err)
+ }
+
+ // Add extra architectures
+ for _, arch := range config.Architectures {
+ scmpArch, err := libseccomp.GetArchFromString(arch)
+ if err != nil {
+ return err
+ }
+
+ if err := filter.AddArch(scmpArch); err != nil {
+ return err
+ }
+ }
+
+ // Unset no new privs bit
+ if err := filter.SetNoNewPrivsBit(false); err != nil {
+ return fmt.Errorf("error setting no new privileges: %s", err)
+ }
+
+ // Add a rule for each syscall
+ for _, call := range config.Syscalls {
+ if call == nil {
+ return fmt.Errorf("encountered nil syscall while initializing Seccomp")
+ }
+
+ if err = matchCall(filter, call); err != nil {
+ return err
+ }
+ }
+
+ if err = filter.Load(); err != nil {
+ return fmt.Errorf("error loading seccomp filter into kernel: %s", err)
+ }
+
+ return nil
+}
+
+// IsEnabled returns if the kernel has been configured to support seccomp.
+func IsEnabled() bool {
+ // Try to read from /proc/self/status for kernels > 3.8
+ s, err := parseStatusFile("/proc/self/status")
+ if err != nil {
+ // Check if Seccomp is supported, via CONFIG_SECCOMP.
+ if err := unix.Prctl(unix.PR_GET_SECCOMP, 0, 0, 0, 0); err != unix.EINVAL {
+ // Make sure the kernel has CONFIG_SECCOMP_FILTER.
+ if err := unix.Prctl(unix.PR_SET_SECCOMP, unix.SECCOMP_MODE_FILTER, 0, 0, 0); err != unix.EINVAL {
+ return true
+ }
+ }
+ return false
+ }
+ _, ok := s["Seccomp"]
+ return ok
+}
+
+// Convert Libcontainer Action to Libseccomp ScmpAction
+func getAction(act configs.Action) (libseccomp.ScmpAction, error) {
+ switch act {
+ case configs.Kill:
+ return actKill, nil
+ case configs.Errno:
+ return actErrno, nil
+ case configs.Trap:
+ return actTrap, nil
+ case configs.Allow:
+ return actAllow, nil
+ case configs.Trace:
+ return actTrace, nil
+ default:
+ return libseccomp.ActInvalid, fmt.Errorf("invalid action, cannot use in rule")
+ }
+}
+
+// Convert Libcontainer Operator to Libseccomp ScmpCompareOp
+func getOperator(op configs.Operator) (libseccomp.ScmpCompareOp, error) {
+ switch op {
+ case configs.EqualTo:
+ return libseccomp.CompareEqual, nil
+ case configs.NotEqualTo:
+ return libseccomp.CompareNotEqual, nil
+ case configs.GreaterThan:
+ return libseccomp.CompareGreater, nil
+ case configs.GreaterThanOrEqualTo:
+ return libseccomp.CompareGreaterEqual, nil
+ case configs.LessThan:
+ return libseccomp.CompareLess, nil
+ case configs.LessThanOrEqualTo:
+ return libseccomp.CompareLessOrEqual, nil
+ case configs.MaskEqualTo:
+ return libseccomp.CompareMaskedEqual, nil
+ default:
+ return libseccomp.CompareInvalid, fmt.Errorf("invalid operator, cannot use in rule")
+ }
+}
+
+// Convert Libcontainer Arg to Libseccomp ScmpCondition
+func getCondition(arg *configs.Arg) (libseccomp.ScmpCondition, error) {
+ cond := libseccomp.ScmpCondition{}
+
+ if arg == nil {
+ return cond, fmt.Errorf("cannot convert nil to syscall condition")
+ }
+
+ op, err := getOperator(arg.Op)
+ if err != nil {
+ return cond, err
+ }
+
+ return libseccomp.MakeCondition(arg.Index, op, arg.Value, arg.ValueTwo)
+}
+
+// Add a rule to match a single syscall
+func matchCall(filter *libseccomp.ScmpFilter, call *configs.Syscall) error {
+ if call == nil || filter == nil {
+ return fmt.Errorf("cannot use nil as syscall to block")
+ }
+
+ if len(call.Name) == 0 {
+ return fmt.Errorf("empty string is not a valid syscall")
+ }
+
+ // If we can't resolve the syscall, assume it's not supported on this kernel
+ // Ignore it, don't error out
+ callNum, err := libseccomp.GetSyscallFromName(call.Name)
+ if err != nil {
+ return nil
+ }
+
+ // Convert the call's action to the libseccomp equivalent
+ callAct, err := getAction(call.Action)
+ if err != nil {
+ return err
+ }
+
+ // Unconditional match - just add the rule
+ if len(call.Args) == 0 {
+ if err = filter.AddRule(callNum, callAct); err != nil {
+ return err
+ }
+ } else {
+ // Conditional match - convert the per-arg rules into library format
+ conditions := []libseccomp.ScmpCondition{}
+
+ for _, cond := range call.Args {
+ newCond, err := getCondition(cond)
+ if err != nil {
+ return err
+ }
+
+ conditions = append(conditions, newCond)
+ }
+
+ if err = filter.AddRuleConditional(callNum, callAct, conditions); err != nil {
+ return err
+ }
+ }
+
+ return nil
+}
+
+func parseStatusFile(path string) (map[string]string, error) {
+ f, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+ defer f.Close()
+
+ s := bufio.NewScanner(f)
+ status := make(map[string]string)
+
+ for s.Scan() {
+ text := s.Text()
+ parts := strings.Split(text, ":")
+
+ if len(parts) <= 1 {
+ continue
+ }
+
+ status[parts[0]] = parts[1]
+ }
+ if err := s.Err(); err != nil {
+ return nil, err
+ }
+
+ return status, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_unsupported.go
new file mode 100644
index 000000000..44df1ad4c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/seccomp/seccomp_unsupported.go
@@ -0,0 +1,24 @@
+// +build !linux !cgo !seccomp
+
+package seccomp
+
+import (
+ "errors"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+)
+
+var ErrSeccompNotEnabled = errors.New("seccomp: config provided but seccomp not supported")
+
+// InitSeccomp does nothing because seccomp is not supported.
+func InitSeccomp(config *configs.Seccomp) error {
+ if config != nil {
+ return ErrSeccompNotEnabled
+ }
+ return nil
+}
+
+// IsEnabled returns false, because it is not supported.
+func IsEnabled() bool {
+ return false
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/setgroups_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/setgroups_linux.go
new file mode 100644
index 000000000..c7bdb605a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/setgroups_linux.go
@@ -0,0 +1,11 @@
+// +build linux,go1.5
+
+package libcontainer
+
+import "syscall"
+
+// Set the GidMappingsEnableSetgroups member to true, so the process's
+// setgroups proc entry wont be set to 'deny' if GidMappings are set
+func enableSetgroups(sys *syscall.SysProcAttr) {
+ sys.GidMappingsEnableSetgroups = true
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/setns_init_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/setns_init_linux.go
new file mode 100644
index 000000000..35b84219c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/setns_init_linux.go
@@ -0,0 +1,65 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+
+ "github.com/opencontainers/runc/libcontainer/apparmor"
+ "github.com/opencontainers/runc/libcontainer/keys"
+ "github.com/opencontainers/runc/libcontainer/seccomp"
+ "github.com/opencontainers/runc/libcontainer/system"
+ "github.com/opencontainers/selinux/go-selinux/label"
+
+ "golang.org/x/sys/unix"
+)
+
+// linuxSetnsInit performs the container's initialization for running a new process
+// inside an existing container.
+type linuxSetnsInit struct {
+ pipe *os.File
+ consoleSocket *os.File
+ config *initConfig
+}
+
+func (l *linuxSetnsInit) getSessionRingName() string {
+ return fmt.Sprintf("_ses.%s", l.config.ContainerId)
+}
+
+func (l *linuxSetnsInit) Init() error {
+ if !l.config.Config.NoNewKeyring {
+ // do not inherit the parent's session keyring
+ if _, err := keys.JoinSessionKeyring(l.getSessionRingName()); err != nil {
+ return err
+ }
+ }
+ if l.config.CreateConsole {
+ if err := setupConsole(l.consoleSocket, l.config, false); err != nil {
+ return err
+ }
+ if err := system.Setctty(); err != nil {
+ return err
+ }
+ }
+ if l.config.NoNewPrivileges {
+ if err := unix.Prctl(unix.PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); err != nil {
+ return err
+ }
+ }
+ if l.config.Config.Seccomp != nil {
+ if err := seccomp.InitSeccomp(l.config.Config.Seccomp); err != nil {
+ return err
+ }
+ }
+ if err := finalizeNamespace(l.config); err != nil {
+ return err
+ }
+ if err := apparmor.ApplyProfile(l.config.AppArmorProfile); err != nil {
+ return err
+ }
+ if err := label.SetProcessLabel(l.config.ProcessLabel); err != nil {
+ return err
+ }
+ return system.Execv(l.config.Args[0], l.config.Args[0:], os.Environ())
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/capture.go b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/capture.go
new file mode 100644
index 000000000..0bbe14950
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/capture.go
@@ -0,0 +1,27 @@
+package stacktrace
+
+import "runtime"
+
+// Capture captures a stacktrace for the current calling go program
+//
+// skip is the number of frames to skip
+func Capture(userSkip int) Stacktrace {
+ var (
+ skip = userSkip + 1 // add one for our own function
+ frames []Frame
+ prevPc uintptr
+ )
+ for i := skip; ; i++ {
+ pc, file, line, ok := runtime.Caller(i)
+ //detect if caller is repeated to avoid loop, gccgo
+ //currently runs into a loop without this check
+ if !ok || pc == prevPc {
+ break
+ }
+ frames = append(frames, NewFrame(pc, file, line))
+ prevPc = pc
+ }
+ return Stacktrace{
+ Frames: frames,
+ }
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/frame.go b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/frame.go
new file mode 100644
index 000000000..0d590d9a5
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/frame.go
@@ -0,0 +1,38 @@
+package stacktrace
+
+import (
+ "path/filepath"
+ "runtime"
+ "strings"
+)
+
+// NewFrame returns a new stack frame for the provided information
+func NewFrame(pc uintptr, file string, line int) Frame {
+ fn := runtime.FuncForPC(pc)
+ if fn == nil {
+ return Frame{}
+ }
+ pack, name := parseFunctionName(fn.Name())
+ return Frame{
+ Line: line,
+ File: filepath.Base(file),
+ Package: pack,
+ Function: name,
+ }
+}
+
+func parseFunctionName(name string) (string, string) {
+ i := strings.LastIndex(name, ".")
+ if i == -1 {
+ return "", name
+ }
+ return name[:i], name[i+1:]
+}
+
+// Frame contains all the information for a stack frame within a go program
+type Frame struct {
+ File string
+ Function string
+ Package string
+ Line int
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/stacktrace.go b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/stacktrace.go
new file mode 100644
index 000000000..5e8b58d2d
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stacktrace/stacktrace.go
@@ -0,0 +1,5 @@
+package stacktrace
+
+type Stacktrace struct {
+ Frames []Frame
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/standard_init_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/standard_init_linux.go
new file mode 100644
index 000000000..580b3fe45
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/standard_init_linux.go
@@ -0,0 +1,188 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+ "os/exec"
+ "syscall" //only for Exec
+
+ "github.com/opencontainers/runc/libcontainer/apparmor"
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/keys"
+ "github.com/opencontainers/runc/libcontainer/seccomp"
+ "github.com/opencontainers/runc/libcontainer/system"
+ "github.com/opencontainers/selinux/go-selinux/label"
+
+ "golang.org/x/sys/unix"
+)
+
+type linuxStandardInit struct {
+ pipe *os.File
+ consoleSocket *os.File
+ parentPid int
+ stateDirFD int
+ config *initConfig
+}
+
+func (l *linuxStandardInit) getSessionRingParams() (string, uint32, uint32) {
+ var newperms uint32
+
+ if l.config.Config.Namespaces.Contains(configs.NEWUSER) {
+ // with user ns we need 'other' search permissions
+ newperms = 0x8
+ } else {
+ // without user ns we need 'UID' search permissions
+ newperms = 0x80000
+ }
+
+ // create a unique per session container name that we can
+ // join in setns; however, other containers can also join it
+ return fmt.Sprintf("_ses.%s", l.config.ContainerId), 0xffffffff, newperms
+}
+
+func (l *linuxStandardInit) Init() error {
+ if !l.config.Config.NoNewKeyring {
+ ringname, keepperms, newperms := l.getSessionRingParams()
+
+ // do not inherit the parent's session keyring
+ sessKeyId, err := keys.JoinSessionKeyring(ringname)
+ if err != nil {
+ return err
+ }
+ // make session keyring searcheable
+ if err := keys.ModKeyringPerm(sessKeyId, keepperms, newperms); err != nil {
+ return err
+ }
+ }
+
+ if err := setupNetwork(l.config); err != nil {
+ return err
+ }
+ if err := setupRoute(l.config.Config); err != nil {
+ return err
+ }
+
+ label.Init()
+
+ // prepareRootfs() can be executed only for a new mount namespace.
+ if l.config.Config.Namespaces.Contains(configs.NEWNS) {
+ if err := prepareRootfs(l.pipe, l.config.Config); err != nil {
+ return err
+ }
+ }
+
+ // Set up the console. This has to be done *before* we finalize the rootfs,
+ // but *after* we've given the user the chance to set up all of the mounts
+ // they wanted.
+ if l.config.CreateConsole {
+ if err := setupConsole(l.consoleSocket, l.config, true); err != nil {
+ return err
+ }
+ if err := system.Setctty(); err != nil {
+ return err
+ }
+ }
+
+ // Finish the rootfs setup.
+ if l.config.Config.Namespaces.Contains(configs.NEWNS) {
+ if err := finalizeRootfs(l.config.Config); err != nil {
+ return err
+ }
+ }
+
+ if hostname := l.config.Config.Hostname; hostname != "" {
+ if err := unix.Sethostname([]byte(hostname)); err != nil {
+ return err
+ }
+ }
+ if err := apparmor.ApplyProfile(l.config.AppArmorProfile); err != nil {
+ return err
+ }
+ if err := label.SetProcessLabel(l.config.ProcessLabel); err != nil {
+ return err
+ }
+
+ for key, value := range l.config.Config.Sysctl {
+ if err := writeSystemProperty(key, value); err != nil {
+ return err
+ }
+ }
+ for _, path := range l.config.Config.ReadonlyPaths {
+ if err := readonlyPath(path); err != nil {
+ return err
+ }
+ }
+ for _, path := range l.config.Config.MaskPaths {
+ if err := maskPath(path); err != nil {
+ return err
+ }
+ }
+ pdeath, err := system.GetParentDeathSignal()
+ if err != nil {
+ return err
+ }
+ if l.config.NoNewPrivileges {
+ if err := unix.Prctl(unix.PR_SET_NO_NEW_PRIVS, 1, 0, 0, 0); err != nil {
+ return err
+ }
+ }
+ // Tell our parent that we're ready to Execv. This must be done before the
+ // Seccomp rules have been applied, because we need to be able to read and
+ // write to a socket.
+ if err := syncParentReady(l.pipe); err != nil {
+ return err
+ }
+ // Without NoNewPrivileges seccomp is a privileged operation, so we need to
+ // do this before dropping capabilities; otherwise do it as late as possible
+ // just before execve so as few syscalls take place after it as possible.
+ if l.config.Config.Seccomp != nil && !l.config.NoNewPrivileges {
+ if err := seccomp.InitSeccomp(l.config.Config.Seccomp); err != nil {
+ return err
+ }
+ }
+ if err := finalizeNamespace(l.config); err != nil {
+ return err
+ }
+ // finalizeNamespace can change user/group which clears the parent death
+ // signal, so we restore it here.
+ if err := pdeath.Restore(); err != nil {
+ return err
+ }
+ // compare the parent from the initial start of the init process and make sure that it did not change.
+ // if the parent changes that means it died and we were reparented to something else so we should
+ // just kill ourself and not cause problems for someone else.
+ if unix.Getppid() != l.parentPid {
+ return unix.Kill(unix.Getpid(), unix.SIGKILL)
+ }
+ // check for the arg before waiting to make sure it exists and it is returned
+ // as a create time error.
+ name, err := exec.LookPath(l.config.Args[0])
+ if err != nil {
+ return err
+ }
+ // close the pipe to signal that we have completed our init.
+ l.pipe.Close()
+ // wait for the fifo to be opened on the other side before
+ // exec'ing the users process.
+ fd, err := unix.Openat(l.stateDirFD, execFifoFilename, os.O_WRONLY|unix.O_CLOEXEC, 0)
+ if err != nil {
+ return newSystemErrorWithCause(err, "openat exec fifo")
+ }
+ if _, err := unix.Write(fd, []byte("0")); err != nil {
+ return newSystemErrorWithCause(err, "write 0 exec fifo")
+ }
+ if l.config.Config.Seccomp != nil && l.config.NoNewPrivileges {
+ if err := seccomp.InitSeccomp(l.config.Config.Seccomp); err != nil {
+ return newSystemErrorWithCause(err, "init seccomp")
+ }
+ }
+ // close the statedir fd before exec because the kernel resets dumpable in the wrong order
+ // https://github.com/torvalds/linux/blob/v4.9/fs/exec.c#L1290-L1318
+ unix.Close(l.stateDirFD)
+ if err := syscall.Exec(name, l.config.Args[0:], os.Environ()); err != nil {
+ return newSystemErrorWithCause(err, "exec user process")
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/state_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/state_linux.go
new file mode 100644
index 000000000..44fa6b43a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/state_linux.go
@@ -0,0 +1,248 @@
+// +build linux
+
+package libcontainer
+
+import (
+ "fmt"
+ "os"
+ "path/filepath"
+
+ "github.com/opencontainers/runc/libcontainer/configs"
+ "github.com/opencontainers/runc/libcontainer/utils"
+
+ "github.com/sirupsen/logrus"
+ "golang.org/x/sys/unix"
+)
+
+func newStateTransitionError(from, to containerState) error {
+ return &stateTransitionError{
+ From: from.status().String(),
+ To: to.status().String(),
+ }
+}
+
+// stateTransitionError is returned when an invalid state transition happens from one
+// state to another.
+type stateTransitionError struct {
+ From string
+ To string
+}
+
+func (s *stateTransitionError) Error() string {
+ return fmt.Sprintf("invalid state transition from %s to %s", s.From, s.To)
+}
+
+type containerState interface {
+ transition(containerState) error
+ destroy() error
+ status() Status
+}
+
+func destroy(c *linuxContainer) error {
+ if !c.config.Namespaces.Contains(configs.NEWPID) {
+ if err := signalAllProcesses(c.cgroupManager, unix.SIGKILL); err != nil {
+ logrus.Warn(err)
+ }
+ }
+ err := c.cgroupManager.Destroy()
+ if rerr := os.RemoveAll(c.root); err == nil {
+ err = rerr
+ }
+ c.initProcess = nil
+ if herr := runPoststopHooks(c); err == nil {
+ err = herr
+ }
+ c.state = &stoppedState{c: c}
+ return err
+}
+
+func runPoststopHooks(c *linuxContainer) error {
+ if c.config.Hooks != nil {
+ s := configs.HookState{
+ Version: c.config.Version,
+ ID: c.id,
+ Bundle: utils.SearchLabels(c.config.Labels, "bundle"),
+ }
+ for _, hook := range c.config.Hooks.Poststop {
+ if err := hook.Run(s); err != nil {
+ return err
+ }
+ }
+ }
+ return nil
+}
+
+// stoppedState represents a container is a stopped/destroyed state.
+type stoppedState struct {
+ c *linuxContainer
+}
+
+func (b *stoppedState) status() Status {
+ return Stopped
+}
+
+func (b *stoppedState) transition(s containerState) error {
+ switch s.(type) {
+ case *runningState, *restoredState:
+ b.c.state = s
+ return nil
+ case *stoppedState:
+ return nil
+ }
+ return newStateTransitionError(b, s)
+}
+
+func (b *stoppedState) destroy() error {
+ return destroy(b.c)
+}
+
+// runningState represents a container that is currently running.
+type runningState struct {
+ c *linuxContainer
+}
+
+func (r *runningState) status() Status {
+ return Running
+}
+
+func (r *runningState) transition(s containerState) error {
+ switch s.(type) {
+ case *stoppedState:
+ t, err := r.c.runType()
+ if err != nil {
+ return err
+ }
+ if t == Running {
+ return newGenericError(fmt.Errorf("container still running"), ContainerNotStopped)
+ }
+ r.c.state = s
+ return nil
+ case *pausedState:
+ r.c.state = s
+ return nil
+ case *runningState:
+ return nil
+ }
+ return newStateTransitionError(r, s)
+}
+
+func (r *runningState) destroy() error {
+ t, err := r.c.runType()
+ if err != nil {
+ return err
+ }
+ if t == Running {
+ return newGenericError(fmt.Errorf("container is not destroyed"), ContainerNotStopped)
+ }
+ return destroy(r.c)
+}
+
+type createdState struct {
+ c *linuxContainer
+}
+
+func (i *createdState) status() Status {
+ return Created
+}
+
+func (i *createdState) transition(s containerState) error {
+ switch s.(type) {
+ case *runningState, *pausedState, *stoppedState:
+ i.c.state = s
+ return nil
+ case *createdState:
+ return nil
+ }
+ return newStateTransitionError(i, s)
+}
+
+func (i *createdState) destroy() error {
+ i.c.initProcess.signal(unix.SIGKILL)
+ return destroy(i.c)
+}
+
+// pausedState represents a container that is currently pause. It cannot be destroyed in a
+// paused state and must transition back to running first.
+type pausedState struct {
+ c *linuxContainer
+}
+
+func (p *pausedState) status() Status {
+ return Paused
+}
+
+func (p *pausedState) transition(s containerState) error {
+ switch s.(type) {
+ case *runningState, *stoppedState:
+ p.c.state = s
+ return nil
+ case *pausedState:
+ return nil
+ }
+ return newStateTransitionError(p, s)
+}
+
+func (p *pausedState) destroy() error {
+ t, err := p.c.runType()
+ if err != nil {
+ return err
+ }
+ if t != Running && t != Created {
+ if err := p.c.cgroupManager.Freeze(configs.Thawed); err != nil {
+ return err
+ }
+ return destroy(p.c)
+ }
+ return newGenericError(fmt.Errorf("container is paused"), ContainerPaused)
+}
+
+// restoredState is the same as the running state but also has associated checkpoint
+// information that maybe need destroyed when the container is stopped and destroy is called.
+type restoredState struct {
+ imageDir string
+ c *linuxContainer
+}
+
+func (r *restoredState) status() Status {
+ return Running
+}
+
+func (r *restoredState) transition(s containerState) error {
+ switch s.(type) {
+ case *stoppedState, *runningState:
+ return nil
+ }
+ return newStateTransitionError(r, s)
+}
+
+func (r *restoredState) destroy() error {
+ if _, err := os.Stat(filepath.Join(r.c.root, "checkpoint")); err != nil {
+ if !os.IsNotExist(err) {
+ return err
+ }
+ }
+ return destroy(r.c)
+}
+
+// loadedState is used whenever a container is restored, loaded, or setting additional
+// processes inside and it should not be destroyed when it is exiting.
+type loadedState struct {
+ c *linuxContainer
+ s Status
+}
+
+func (n *loadedState) status() Status {
+ return n.s
+}
+
+func (n *loadedState) transition(s containerState) error {
+ n.c.state = s
+ return nil
+}
+
+func (n *loadedState) destroy() error {
+ if err := n.c.refreshState(); err != nil {
+ return err
+ }
+ return n.c.state.destroy()
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stats.go b/vendor/github.com/opencontainers/runc/libcontainer/stats.go
new file mode 100644
index 000000000..303e4b94c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stats.go
@@ -0,0 +1,15 @@
+package libcontainer
+
+type NetworkInterface struct {
+ // Name is the name of the network interface.
+ Name string
+
+ RxBytes uint64
+ RxPackets uint64
+ RxErrors uint64
+ RxDropped uint64
+ TxBytes uint64
+ TxPackets uint64
+ TxErrors uint64
+ TxDropped uint64
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stats_freebsd.go b/vendor/github.com/opencontainers/runc/libcontainer/stats_freebsd.go
new file mode 100644
index 000000000..f8d1d689c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stats_freebsd.go
@@ -0,0 +1,5 @@
+package libcontainer
+
+type Stats struct {
+ Interfaces []*NetworkInterface
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stats_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/stats_linux.go
new file mode 100644
index 000000000..c629dc67d
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stats_linux.go
@@ -0,0 +1,8 @@
+package libcontainer
+
+import "github.com/opencontainers/runc/libcontainer/cgroups"
+
+type Stats struct {
+ Interfaces []*NetworkInterface
+ CgroupStats *cgroups.Stats
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stats_solaris.go b/vendor/github.com/opencontainers/runc/libcontainer/stats_solaris.go
new file mode 100644
index 000000000..da78c1c2e
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stats_solaris.go
@@ -0,0 +1,7 @@
+package libcontainer
+
+// Solaris - TODO
+
+type Stats struct {
+ Interfaces []*NetworkInterface
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/stats_windows.go b/vendor/github.com/opencontainers/runc/libcontainer/stats_windows.go
new file mode 100644
index 000000000..f8d1d689c
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/stats_windows.go
@@ -0,0 +1,5 @@
+package libcontainer
+
+type Stats struct {
+ Interfaces []*NetworkInterface
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/sync.go b/vendor/github.com/opencontainers/runc/libcontainer/sync.go
new file mode 100644
index 000000000..cf7b45bc3
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/sync.go
@@ -0,0 +1,107 @@
+package libcontainer
+
+import (
+ "encoding/json"
+ "fmt"
+ "io"
+
+ "github.com/opencontainers/runc/libcontainer/utils"
+)
+
+type syncType string
+
+// Constants that are used for synchronisation between the parent and child
+// during container setup. They come in pairs (with procError being a generic
+// response which is followed by a &genericError).
+//
+// [ child ] <-> [ parent ]
+//
+// procHooks --> [run hooks]
+// <-- procResume
+//
+// procConsole -->
+// <-- procConsoleReq
+// [send(fd)] --> [recv(fd)]
+// <-- procConsoleAck
+//
+// procReady --> [final setup]
+// <-- procRun
+const (
+ procError syncType = "procError"
+ procReady syncType = "procReady"
+ procRun syncType = "procRun"
+ procHooks syncType = "procHooks"
+ procResume syncType = "procResume"
+)
+
+type syncT struct {
+ Type syncType `json:"type"`
+}
+
+// writeSync is used to write to a synchronisation pipe. An error is returned
+// if there was a problem writing the payload.
+func writeSync(pipe io.Writer, sync syncType) error {
+ if err := utils.WriteJSON(pipe, syncT{sync}); err != nil {
+ return err
+ }
+ return nil
+}
+
+// readSync is used to read from a synchronisation pipe. An error is returned
+// if we got a genericError, the pipe was closed, or we got an unexpected flag.
+func readSync(pipe io.Reader, expected syncType) error {
+ var procSync syncT
+ if err := json.NewDecoder(pipe).Decode(&procSync); err != nil {
+ if err == io.EOF {
+ return fmt.Errorf("parent closed synchronisation channel")
+ }
+
+ if procSync.Type == procError {
+ var ierr genericError
+
+ if err := json.NewDecoder(pipe).Decode(&ierr); err != nil {
+ return fmt.Errorf("failed reading error from parent: %v", err)
+ }
+
+ return &ierr
+ }
+
+ if procSync.Type != expected {
+ return fmt.Errorf("invalid synchronisation flag from parent")
+ }
+ }
+ return nil
+}
+
+// parseSync runs the given callback function on each syncT received from the
+// child. It will return once io.EOF is returned from the given pipe.
+func parseSync(pipe io.Reader, fn func(*syncT) error) error {
+ dec := json.NewDecoder(pipe)
+ for {
+ var sync syncT
+ if err := dec.Decode(&sync); err != nil {
+ if err == io.EOF {
+ break
+ }
+ return err
+ }
+
+ // We handle this case outside fn for cleanliness reasons.
+ var ierr *genericError
+ if sync.Type == procError {
+ if err := dec.Decode(&ierr); err != nil && err != io.EOF {
+ return newSystemErrorWithCause(err, "decoding proc error from init")
+ }
+ if ierr != nil {
+ return ierr
+ }
+ // Programmer error.
+ panic("No error following JSON procError payload.")
+ }
+
+ if err := fn(&sync); err != nil {
+ return err
+ }
+ }
+ return nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/linux.go b/vendor/github.com/opencontainers/runc/libcontainer/system/linux.go
new file mode 100644
index 000000000..4837085a7
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/linux.go
@@ -0,0 +1,136 @@
+// +build linux
+
+package system
+
+import (
+ "bufio"
+ "fmt"
+ "os"
+ "os/exec"
+ "syscall" // only for exec
+ "unsafe"
+
+ "golang.org/x/sys/unix"
+)
+
+// If arg2 is nonzero, set the "child subreaper" attribute of the
+// calling process; if arg2 is zero, unset the attribute. When a
+// process is marked as a child subreaper, all of the children
+// that it creates, and their descendants, will be marked as
+// having a subreaper. In effect, a subreaper fulfills the role
+// of init(1) for its descendant processes. Upon termination of
+// a process that is orphaned (i.e., its immediate parent has
+// already terminated) and marked as having a subreaper, the
+// nearest still living ancestor subreaper will receive a SIGCHLD
+// signal and be able to wait(2) on the process to discover its
+// termination status.
+const PR_SET_CHILD_SUBREAPER = 36
+
+type ParentDeathSignal int
+
+func (p ParentDeathSignal) Restore() error {
+ if p == 0 {
+ return nil
+ }
+ current, err := GetParentDeathSignal()
+ if err != nil {
+ return err
+ }
+ if p == current {
+ return nil
+ }
+ return p.Set()
+}
+
+func (p ParentDeathSignal) Set() error {
+ return SetParentDeathSignal(uintptr(p))
+}
+
+func Execv(cmd string, args []string, env []string) error {
+ name, err := exec.LookPath(cmd)
+ if err != nil {
+ return err
+ }
+
+ return syscall.Exec(name, args, env)
+}
+
+func Prlimit(pid, resource int, limit unix.Rlimit) error {
+ _, _, err := unix.RawSyscall6(unix.SYS_PRLIMIT64, uintptr(pid), uintptr(resource), uintptr(unsafe.Pointer(&limit)), uintptr(unsafe.Pointer(&limit)), 0, 0)
+ if err != 0 {
+ return err
+ }
+ return nil
+}
+
+func SetParentDeathSignal(sig uintptr) error {
+ if err := unix.Prctl(unix.PR_SET_PDEATHSIG, sig, 0, 0, 0); err != nil {
+ return err
+ }
+ return nil
+}
+
+func GetParentDeathSignal() (ParentDeathSignal, error) {
+ var sig int
+ if err := unix.Prctl(unix.PR_GET_PDEATHSIG, uintptr(unsafe.Pointer(&sig)), 0, 0, 0); err != nil {
+ return -1, err
+ }
+ return ParentDeathSignal(sig), nil
+}
+
+func SetKeepCaps() error {
+ if err := unix.Prctl(unix.PR_SET_KEEPCAPS, 1, 0, 0, 0); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func ClearKeepCaps() error {
+ if err := unix.Prctl(unix.PR_SET_KEEPCAPS, 0, 0, 0, 0); err != nil {
+ return err
+ }
+
+ return nil
+}
+
+func Setctty() error {
+ if err := unix.IoctlSetInt(0, unix.TIOCSCTTY, 0); err != nil {
+ return err
+ }
+ return nil
+}
+
+// RunningInUserNS detects whether we are currently running in a user namespace.
+// Copied from github.com/lxc/lxd/shared/util.go
+func RunningInUserNS() bool {
+ file, err := os.Open("/proc/self/uid_map")
+ if err != nil {
+ // This kernel-provided file only exists if user namespaces are supported
+ return false
+ }
+ defer file.Close()
+
+ buf := bufio.NewReader(file)
+ l, _, err := buf.ReadLine()
+ if err != nil {
+ return false
+ }
+
+ line := string(l)
+ var a, b, c int64
+ fmt.Sscanf(line, "%d %d %d", &a, &b, &c)
+ /*
+ * We assume we are in the initial user namespace if we have a full
+ * range - 4294967295 uids starting at uid 0.
+ */
+ if a == 0 && b == 0 && c == 4294967295 {
+ return false
+ }
+ return true
+}
+
+// SetSubreaper sets the value i as the subreaper setting for the calling process
+func SetSubreaper(i int) error {
+ return unix.Prctl(PR_SET_CHILD_SUBREAPER, uintptr(i), 0, 0, 0)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/proc.go b/vendor/github.com/opencontainers/runc/libcontainer/system/proc.go
new file mode 100644
index 000000000..79232a437
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/proc.go
@@ -0,0 +1,113 @@
+package system
+
+import (
+ "fmt"
+ "io/ioutil"
+ "path/filepath"
+ "strconv"
+ "strings"
+)
+
+// State is the status of a process.
+type State rune
+
+const ( // Only values for Linux 3.14 and later are listed here
+ Dead State = 'X'
+ DiskSleep State = 'D'
+ Running State = 'R'
+ Sleeping State = 'S'
+ Stopped State = 'T'
+ TracingStop State = 't'
+ Zombie State = 'Z'
+)
+
+// String forms of the state from proc(5)'s documentation for
+// /proc/[pid]/status' "State" field.
+func (s State) String() string {
+ switch s {
+ case Dead:
+ return "dead"
+ case DiskSleep:
+ return "disk sleep"
+ case Running:
+ return "running"
+ case Sleeping:
+ return "sleeping"
+ case Stopped:
+ return "stopped"
+ case TracingStop:
+ return "tracing stop"
+ case Zombie:
+ return "zombie"
+ default:
+ return fmt.Sprintf("unknown (%c)", s)
+ }
+}
+
+// Stat_t represents the information from /proc/[pid]/stat, as
+// described in proc(5) with names based on the /proc/[pid]/status
+// fields.
+type Stat_t struct {
+ // PID is the process ID.
+ PID uint
+
+ // Name is the command run by the process.
+ Name string
+
+ // State is the state of the process.
+ State State
+
+ // StartTime is the number of clock ticks after system boot (since
+ // Linux 2.6).
+ StartTime uint64
+}
+
+// Stat returns a Stat_t instance for the specified process.
+func Stat(pid int) (stat Stat_t, err error) {
+ bytes, err := ioutil.ReadFile(filepath.Join("/proc", strconv.Itoa(pid), "stat"))
+ if err != nil {
+ return stat, err
+ }
+ return parseStat(string(bytes))
+}
+
+// GetProcessStartTime is deprecated. Use Stat(pid) and
+// Stat_t.StartTime instead.
+func GetProcessStartTime(pid int) (string, error) {
+ stat, err := Stat(pid)
+ if err != nil {
+ return "", err
+ }
+ return fmt.Sprintf("%d", stat.StartTime), nil
+}
+
+func parseStat(data string) (stat Stat_t, err error) {
+ // From proc(5), field 2 could contain space and is inside `(` and `)`.
+ // The following is an example:
+ // 89653 (gunicorn: maste) S 89630 89653 89653 0 -1 4194560 29689 28896 0 3 146 32 76 19 20 0 1 0 2971844 52965376 3920 18446744073709551615 1 1 0 0 0 0 0 16781312 137447943 0 0 0 17 1 0 0 0 0 0 0 0 0 0 0 0 0 0
+ i := strings.LastIndex(data, ")")
+ if i <= 2 || i >= len(data)-1 {
+ return stat, fmt.Errorf("invalid stat data: %q", data)
+ }
+
+ parts := strings.SplitN(data[:i], "(", 2)
+ if len(parts) != 2 {
+ return stat, fmt.Errorf("invalid stat data: %q", data)
+ }
+
+ stat.Name = parts[1]
+ _, err = fmt.Sscanf(parts[0], "%d", &stat.PID)
+ if err != nil {
+ return stat, err
+ }
+
+ // parts indexes should be offset by 3 from the field number given
+ // proc(5), because parts is zero-indexed and we've removed fields
+ // one (PID) and two (Name) in the paren-split.
+ parts = strings.Split(data[i+2:], " ")
+ var state int
+ fmt.Sscanf(parts[3-3], "%c", &state)
+ stat.State = State(state)
+ fmt.Sscanf(parts[22-3], "%d", &stat.StartTime)
+ return stat, nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_386.go b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_386.go
new file mode 100644
index 000000000..3f7235ed1
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_386.go
@@ -0,0 +1,25 @@
+// +build linux,386
+
+package system
+
+import (
+ "golang.org/x/sys/unix"
+)
+
+// Setuid sets the uid of the calling thread to the specified uid.
+func Setuid(uid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETUID32, uintptr(uid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
+
+// Setgid sets the gid of the calling thread to the specified gid.
+func Setgid(gid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETGID32, uintptr(gid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_64.go b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_64.go
new file mode 100644
index 000000000..d7891a2ff
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_64.go
@@ -0,0 +1,25 @@
+// +build linux,arm64 linux,amd64 linux,ppc linux,ppc64 linux,ppc64le linux,s390x
+
+package system
+
+import (
+ "golang.org/x/sys/unix"
+)
+
+// Setuid sets the uid of the calling thread to the specified uid.
+func Setuid(uid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETUID, uintptr(uid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
+
+// Setgid sets the gid of the calling thread to the specified gid.
+func Setgid(gid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETGID, uintptr(gid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_arm.go b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_arm.go
new file mode 100644
index 000000000..31ff3deb1
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/syscall_linux_arm.go
@@ -0,0 +1,25 @@
+// +build linux,arm
+
+package system
+
+import (
+ "golang.org/x/sys/unix"
+)
+
+// Setuid sets the uid of the calling thread to the specified uid.
+func Setuid(uid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETUID32, uintptr(uid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
+
+// Setgid sets the gid of the calling thread to the specified gid.
+func Setgid(gid int) (err error) {
+ _, _, e1 := unix.RawSyscall(unix.SYS_SETGID32, uintptr(gid), 0, 0)
+ if e1 != 0 {
+ err = e1
+ }
+ return
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig.go b/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig.go
new file mode 100644
index 000000000..b3a07cba3
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig.go
@@ -0,0 +1,12 @@
+// +build cgo,linux cgo,freebsd
+
+package system
+
+/*
+#include <unistd.h>
+*/
+import "C"
+
+func GetClockTicks() int {
+ return int(C.sysconf(C._SC_CLK_TCK))
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig_notcgo.go b/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig_notcgo.go
new file mode 100644
index 000000000..d93b5d5fd
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/sysconfig_notcgo.go
@@ -0,0 +1,15 @@
+// +build !cgo windows
+
+package system
+
+func GetClockTicks() int {
+ // TODO figure out a better alternative for platforms where we're missing cgo
+ //
+ // TODO Windows. This could be implemented using Win32 QueryPerformanceFrequency().
+ // https://msdn.microsoft.com/en-us/library/windows/desktop/ms644905(v=vs.85).aspx
+ //
+ // An example of its usage can be found here.
+ // https://msdn.microsoft.com/en-us/library/windows/desktop/dn553408(v=vs.85).aspx
+
+ return 100
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/system/unsupported.go
new file mode 100644
index 000000000..e7cfd62b2
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/unsupported.go
@@ -0,0 +1,9 @@
+// +build !linux
+
+package system
+
+// RunningInUserNS is a stub for non-Linux systems
+// Always returns false
+func RunningInUserNS() bool {
+ return false
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/system/xattrs_linux.go b/vendor/github.com/opencontainers/runc/libcontainer/system/xattrs_linux.go
new file mode 100644
index 000000000..a6823fc99
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/system/xattrs_linux.go
@@ -0,0 +1,35 @@
+package system
+
+import "golang.org/x/sys/unix"
+
+// Returns a []byte slice if the xattr is set and nil otherwise
+// Requires path and its attribute as arguments
+func Lgetxattr(path string, attr string) ([]byte, error) {
+ var sz int
+ // Start with a 128 length byte array
+ dest := make([]byte, 128)
+ sz, errno := unix.Lgetxattr(path, attr, dest)
+
+ switch {
+ case errno == unix.ENODATA:
+ return nil, errno
+ case errno == unix.ENOTSUP:
+ return nil, errno
+ case errno == unix.ERANGE:
+ // 128 byte array might just not be good enough,
+ // A dummy buffer is used to get the real size
+ // of the xattrs on disk
+ sz, errno = unix.Lgetxattr(path, attr, []byte{})
+ if errno != nil {
+ return nil, errno
+ }
+ dest = make([]byte, sz)
+ sz, errno = unix.Lgetxattr(path, attr, dest)
+ if errno != nil {
+ return nil, errno
+ }
+ case errno != nil:
+ return nil, errno
+ }
+ return dest[:sz], nil
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/user/lookup.go b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup.go
new file mode 100644
index 000000000..95e9eebc0
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup.go
@@ -0,0 +1,95 @@
+package user
+
+import (
+ "errors"
+)
+
+var (
+ // The current operating system does not provide the required data for user lookups.
+ ErrUnsupported = errors.New("user lookup: operating system does not provide passwd-formatted data")
+ // No matching entries found in file.
+ ErrNoPasswdEntries = errors.New("no matching entries in passwd file")
+ ErrNoGroupEntries = errors.New("no matching entries in group file")
+)
+
+func lookupUser(filter func(u User) bool) (User, error) {
+ // Get operating system-specific passwd reader-closer.
+ passwd, err := GetPasswd()
+ if err != nil {
+ return User{}, err
+ }
+ defer passwd.Close()
+
+ // Get the users.
+ users, err := ParsePasswdFilter(passwd, filter)
+ if err != nil {
+ return User{}, err
+ }
+
+ // No user entries found.
+ if len(users) == 0 {
+ return User{}, ErrNoPasswdEntries
+ }
+
+ // Assume the first entry is the "correct" one.
+ return users[0], nil
+}
+
+// LookupUser looks up a user by their username in /etc/passwd. If the user
+// cannot be found (or there is no /etc/passwd file on the filesystem), then
+// LookupUser returns an error.
+func LookupUser(username string) (User, error) {
+ return lookupUser(func(u User) bool {
+ return u.Name == username
+ })
+}
+
+// LookupUid looks up a user by their user id in /etc/passwd. If the user cannot
+// be found (or there is no /etc/passwd file on the filesystem), then LookupId
+// returns an error.
+func LookupUid(uid int) (User, error) {
+ return lookupUser(func(u User) bool {
+ return u.Uid == uid
+ })
+}
+
+func lookupGroup(filter func(g Group) bool) (Group, error) {
+ // Get operating system-specific group reader-closer.
+ group, err := GetGroup()
+ if err != nil {
+ return Group{}, err
+ }
+ defer group.Close()
+
+ // Get the users.
+ groups, err := ParseGroupFilter(group, filter)
+ if err != nil {
+ return Group{}, err
+ }
+
+ // No user entries found.
+ if len(groups) == 0 {
+ return Group{}, ErrNoGroupEntries
+ }
+
+ // Assume the first entry is the "correct" one.
+ return groups[0], nil
+}
+
+// LookupGroup looks up a group by its name in /etc/group. If the group cannot
+// be found (or there is no /etc/group file on the filesystem), then LookupGroup
+// returns an error.
+func LookupGroup(groupname string) (Group, error) {
+ return lookupGroup(func(g Group) bool {
+ return g.Name == groupname
+ })
+}
+
+// LookupGid looks up a group by its group id in /etc/group. If the group cannot
+// be found (or there is no /etc/group file on the filesystem), then LookupGid
+// returns an error.
+func LookupGid(gid int) (Group, error) {
+ return lookupGroup(func(g Group) bool {
+ return g.Gid == gid
+ })
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unix.go b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unix.go
new file mode 100644
index 000000000..c2bb9ec90
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unix.go
@@ -0,0 +1,46 @@
+// +build darwin dragonfly freebsd linux netbsd openbsd solaris
+
+package user
+
+import (
+ "io"
+ "os"
+
+ "golang.org/x/sys/unix"
+)
+
+// Unix-specific path to the passwd and group formatted files.
+const (
+ unixPasswdPath = "/etc/passwd"
+ unixGroupPath = "/etc/group"
+)
+
+func GetPasswdPath() (string, error) {
+ return unixPasswdPath, nil
+}
+
+func GetPasswd() (io.ReadCloser, error) {
+ return os.Open(unixPasswdPath)
+}
+
+func GetGroupPath() (string, error) {
+ return unixGroupPath, nil
+}
+
+func GetGroup() (io.ReadCloser, error) {
+ return os.Open(unixGroupPath)
+}
+
+// CurrentUser looks up the current user by their user id in /etc/passwd. If the
+// user cannot be found (or there is no /etc/passwd file on the filesystem),
+// then CurrentUser returns an error.
+func CurrentUser() (User, error) {
+ return LookupUid(unix.Getuid())
+}
+
+// CurrentGroup looks up the current user's group by their primary group id's
+// entry in /etc/passwd. If the group cannot be found (or there is no
+// /etc/group file on the filesystem), then CurrentGroup returns an error.
+func CurrentGroup() (Group, error) {
+ return LookupGid(unix.Getgid())
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unsupported.go b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unsupported.go
new file mode 100644
index 000000000..4a8d00acb
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/user/lookup_unsupported.go
@@ -0,0 +1,38 @@
+// +build !darwin,!dragonfly,!freebsd,!linux,!netbsd,!openbsd,!solaris
+
+package user
+
+import (
+ "io"
+ "syscall"
+)
+
+func GetPasswdPath() (string, error) {
+ return "", ErrUnsupported
+}
+
+func GetPasswd() (io.ReadCloser, error) {
+ return nil, ErrUnsupported
+}
+
+func GetGroupPath() (string, error) {
+ return "", ErrUnsupported
+}
+
+func GetGroup() (io.ReadCloser, error) {
+ return nil, ErrUnsupported
+}
+
+// CurrentUser looks up the current user by their user id in /etc/passwd. If the
+// user cannot be found (or there is no /etc/passwd file on the filesystem),
+// then CurrentUser returns an error.
+func CurrentUser() (User, error) {
+ return LookupUid(syscall.Getuid())
+}
+
+// CurrentGroup looks up the current user's group by their primary group id's
+// entry in /etc/passwd. If the group cannot be found (or there is no
+// /etc/group file on the filesystem), then CurrentGroup returns an error.
+func CurrentGroup() (Group, error) {
+ return LookupGid(syscall.Getgid())
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/user/user.go b/vendor/github.com/opencontainers/runc/libcontainer/user/user.go
new file mode 100644
index 000000000..8962cab33
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/user/user.go
@@ -0,0 +1,441 @@
+package user
+
+import (
+ "bufio"
+ "fmt"
+ "io"
+ "os"
+ "strconv"
+ "strings"
+)
+
+const (
+ minId = 0
+ maxId = 1<<31 - 1 //for 32-bit systems compatibility
+)
+
+var (
+ ErrRange = fmt.Errorf("uids and gids must be in range %d-%d", minId, maxId)
+)
+
+type User struct {
+ Name string
+ Pass string
+ Uid int
+ Gid int
+ Gecos string
+ Home string
+ Shell string
+}
+
+type Group struct {
+ Name string
+ Pass string
+ Gid int
+ List []string
+}
+
+func parseLine(line string, v ...interface{}) {
+ if line == "" {
+ return
+ }
+
+ parts := strings.Split(line, ":")
+ for i, p := range parts {
+ // Ignore cases where we don't have enough fields to populate the arguments.
+ // Some configuration files like to misbehave.
+ if len(v) <= i {
+ break
+ }
+
+ // Use the type of the argument to figure out how to parse it, scanf() style.
+ // This is legit.
+ switch e := v[i].(type) {
+ case *string:
+ *e = p
+ case *int:
+ // "numbers", with conversion errors ignored because of some misbehaving configuration files.
+ *e, _ = strconv.Atoi(p)
+ case *[]string:
+ // Comma-separated lists.
+ if p != "" {
+ *e = strings.Split(p, ",")
+ } else {
+ *e = []string{}
+ }
+ default:
+ // Someone goof'd when writing code using this function. Scream so they can hear us.
+ panic(fmt.Sprintf("parseLine only accepts {*string, *int, *[]string} as arguments! %#v is not a pointer!", e))
+ }
+ }
+}
+
+func ParsePasswdFile(path string) ([]User, error) {
+ passwd, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+ defer passwd.Close()
+ return ParsePasswd(passwd)
+}
+
+func ParsePasswd(passwd io.Reader) ([]User, error) {
+ return ParsePasswdFilter(passwd, nil)
+}
+
+func ParsePasswdFileFilter(path string, filter func(User) bool) ([]User, error) {
+ passwd, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+ defer passwd.Close()
+ return ParsePasswdFilter(passwd, filter)
+}
+
+func ParsePasswdFilter(r io.Reader, filter func(User) bool) ([]User, error) {
+ if r == nil {
+ return nil, fmt.Errorf("nil source for passwd-formatted data")
+ }
+
+ var (
+ s = bufio.NewScanner(r)
+ out = []User{}
+ )
+
+ for s.Scan() {
+ if err := s.Err(); err != nil {
+ return nil, err
+ }
+
+ line := strings.TrimSpace(s.Text())
+ if line == "" {
+ continue
+ }
+
+ // see: man 5 passwd
+ // name:password:UID:GID:GECOS:directory:shell
+ // Name:Pass:Uid:Gid:Gecos:Home:Shell
+ // root:x:0:0:root:/root:/bin/bash
+ // adm:x:3:4:adm:/var/adm:/bin/false
+ p := User{}
+ parseLine(line, &p.Name, &p.Pass, &p.Uid, &p.Gid, &p.Gecos, &p.Home, &p.Shell)
+
+ if filter == nil || filter(p) {
+ out = append(out, p)
+ }
+ }
+
+ return out, nil
+}
+
+func ParseGroupFile(path string) ([]Group, error) {
+ group, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+
+ defer group.Close()
+ return ParseGroup(group)
+}
+
+func ParseGroup(group io.Reader) ([]Group, error) {
+ return ParseGroupFilter(group, nil)
+}
+
+func ParseGroupFileFilter(path string, filter func(Group) bool) ([]Group, error) {
+ group, err := os.Open(path)
+ if err != nil {
+ return nil, err
+ }
+ defer group.Close()
+ return ParseGroupFilter(group, filter)
+}
+
+func ParseGroupFilter(r io.Reader, filter func(Group) bool) ([]Group, error) {
+ if r == nil {
+ return nil, fmt.Errorf("nil source for group-formatted data")
+ }
+
+ var (
+ s = bufio.NewScanner(r)
+ out = []Group{}
+ )
+
+ for s.Scan() {
+ if err := s.Err(); err != nil {
+ return nil, err
+ }
+
+ text := s.Text()
+ if text == "" {
+ continue
+ }
+
+ // see: man 5 group
+ // group_name:password:GID:user_list
+ // Name:Pass:Gid:List
+ // root:x:0:root
+ // adm:x:4:root,adm,daemon
+ p := Group{}
+ parseLine(text, &p.Name, &p.Pass, &p.Gid, &p.List)
+
+ if filter == nil || filter(p) {
+ out = append(out, p)
+ }
+ }
+
+ return out, nil
+}
+
+type ExecUser struct {
+ Uid int
+ Gid int
+ Sgids []int
+ Home string
+}
+
+// GetExecUserPath is a wrapper for GetExecUser. It reads data from each of the
+// given file paths and uses that data as the arguments to GetExecUser. If the
+// files cannot be opened for any reason, the error is ignored and a nil
+// io.Reader is passed instead.
+func GetExecUserPath(userSpec string, defaults *ExecUser, passwdPath, groupPath string) (*ExecUser, error) {
+ var passwd, group io.Reader
+
+ if passwdFile, err := os.Open(passwdPath); err == nil {
+ passwd = passwdFile
+ defer passwdFile.Close()
+ }
+
+ if groupFile, err := os.Open(groupPath); err == nil {
+ group = groupFile
+ defer groupFile.Close()
+ }
+
+ return GetExecUser(userSpec, defaults, passwd, group)
+}
+
+// GetExecUser parses a user specification string (using the passwd and group
+// readers as sources for /etc/passwd and /etc/group data, respectively). In
+// the case of blank fields or missing data from the sources, the values in
+// defaults is used.
+//
+// GetExecUser will return an error if a user or group literal could not be
+// found in any entry in passwd and group respectively.
+//
+// Examples of valid user specifications are:
+// * ""
+// * "user"
+// * "uid"
+// * "user:group"
+// * "uid:gid
+// * "user:gid"
+// * "uid:group"
+//
+// It should be noted that if you specify a numeric user or group id, they will
+// not be evaluated as usernames (only the metadata will be filled). So attempting
+// to parse a user with user.Name = "1337" will produce the user with a UID of
+// 1337.
+func GetExecUser(userSpec string, defaults *ExecUser, passwd, group io.Reader) (*ExecUser, error) {
+ if defaults == nil {
+ defaults = new(ExecUser)
+ }
+
+ // Copy over defaults.
+ user := &ExecUser{
+ Uid: defaults.Uid,
+ Gid: defaults.Gid,
+ Sgids: defaults.Sgids,
+ Home: defaults.Home,
+ }
+
+ // Sgids slice *cannot* be nil.
+ if user.Sgids == nil {
+ user.Sgids = []int{}
+ }
+
+ // Allow for userArg to have either "user" syntax, or optionally "user:group" syntax
+ var userArg, groupArg string
+ parseLine(userSpec, &userArg, &groupArg)
+
+ // Convert userArg and groupArg to be numeric, so we don't have to execute
+ // Atoi *twice* for each iteration over lines.
+ uidArg, uidErr := strconv.Atoi(userArg)
+ gidArg, gidErr := strconv.Atoi(groupArg)
+
+ // Find the matching user.
+ users, err := ParsePasswdFilter(passwd, func(u User) bool {
+ if userArg == "" {
+ // Default to current state of the user.
+ return u.Uid == user.Uid
+ }
+
+ if uidErr == nil {
+ // If the userArg is numeric, always treat it as a UID.
+ return uidArg == u.Uid
+ }
+
+ return u.Name == userArg
+ })
+
+ // If we can't find the user, we have to bail.
+ if err != nil && passwd != nil {
+ if userArg == "" {
+ userArg = strconv.Itoa(user.Uid)
+ }
+ return nil, fmt.Errorf("unable to find user %s: %v", userArg, err)
+ }
+
+ var matchedUserName string
+ if len(users) > 0 {
+ // First match wins, even if there's more than one matching entry.
+ matchedUserName = users[0].Name
+ user.Uid = users[0].Uid
+ user.Gid = users[0].Gid
+ user.Home = users[0].Home
+ } else if userArg != "" {
+ // If we can't find a user with the given username, the only other valid
+ // option is if it's a numeric username with no associated entry in passwd.
+
+ if uidErr != nil {
+ // Not numeric.
+ return nil, fmt.Errorf("unable to find user %s: %v", userArg, ErrNoPasswdEntries)
+ }
+ user.Uid = uidArg
+
+ // Must be inside valid uid range.
+ if user.Uid < minId || user.Uid > maxId {
+ return nil, ErrRange
+ }
+
+ // Okay, so it's numeric. We can just roll with this.
+ }
+
+ // On to the groups. If we matched a username, we need to do this because of
+ // the supplementary group IDs.
+ if groupArg != "" || matchedUserName != "" {
+ groups, err := ParseGroupFilter(group, func(g Group) bool {
+ // If the group argument isn't explicit, we'll just search for it.
+ if groupArg == "" {
+ // Check if user is a member of this group.
+ for _, u := range g.List {
+ if u == matchedUserName {
+ return true
+ }
+ }
+ return false
+ }
+
+ if gidErr == nil {
+ // If the groupArg is numeric, always treat it as a GID.
+ return gidArg == g.Gid
+ }
+
+ return g.Name == groupArg
+ })
+ if err != nil && group != nil {
+ return nil, fmt.Errorf("unable to find groups for spec %v: %v", matchedUserName, err)
+ }
+
+ // Only start modifying user.Gid if it is in explicit form.
+ if groupArg != "" {
+ if len(groups) > 0 {
+ // First match wins, even if there's more than one matching entry.
+ user.Gid = groups[0].Gid
+ } else {
+ // If we can't find a group with the given name, the only other valid
+ // option is if it's a numeric group name with no associated entry in group.
+
+ if gidErr != nil {
+ // Not numeric.
+ return nil, fmt.Errorf("unable to find group %s: %v", groupArg, ErrNoGroupEntries)
+ }
+ user.Gid = gidArg
+
+ // Must be inside valid gid range.
+ if user.Gid < minId || user.Gid > maxId {
+ return nil, ErrRange
+ }
+
+ // Okay, so it's numeric. We can just roll with this.
+ }
+ } else if len(groups) > 0 {
+ // Supplementary group ids only make sense if in the implicit form.
+ user.Sgids = make([]int, len(groups))
+ for i, group := range groups {
+ user.Sgids[i] = group.Gid
+ }
+ }
+ }
+
+ return user, nil
+}
+
+// GetAdditionalGroups looks up a list of groups by name or group id
+// against the given /etc/group formatted data. If a group name cannot
+// be found, an error will be returned. If a group id cannot be found,
+// or the given group data is nil, the id will be returned as-is
+// provided it is in the legal range.
+func GetAdditionalGroups(additionalGroups []string, group io.Reader) ([]int, error) {
+ var groups = []Group{}
+ if group != nil {
+ var err error
+ groups, err = ParseGroupFilter(group, func(g Group) bool {
+ for _, ag := range additionalGroups {
+ if g.Name == ag || strconv.Itoa(g.Gid) == ag {
+ return true
+ }
+ }
+ return false
+ })
+ if err != nil {
+ return nil, fmt.Errorf("Unable to find additional groups %v: %v", additionalGroups, err)
+ }
+ }
+
+ gidMap := make(map[int]struct{})
+ for _, ag := range additionalGroups {
+ var found bool
+ for _, g := range groups {
+ // if we found a matched group either by name or gid, take the
+ // first matched as correct
+ if g.Name == ag || strconv.Itoa(g.Gid) == ag {
+ if _, ok := gidMap[g.Gid]; !ok {
+ gidMap[g.Gid] = struct{}{}
+ found = true
+ break
+ }
+ }
+ }
+ // we asked for a group but didn't find it. let's check to see
+ // if we wanted a numeric group
+ if !found {
+ gid, err := strconv.Atoi(ag)
+ if err != nil {
+ return nil, fmt.Errorf("Unable to find group %s", ag)
+ }
+ // Ensure gid is inside gid range.
+ if gid < minId || gid > maxId {
+ return nil, ErrRange
+ }
+ gidMap[gid] = struct{}{}
+ }
+ }
+ gids := []int{}
+ for gid := range gidMap {
+ gids = append(gids, gid)
+ }
+ return gids, nil
+}
+
+// GetAdditionalGroupsPath is a wrapper around GetAdditionalGroups
+// that opens the groupPath given and gives it as an argument to
+// GetAdditionalGroups.
+func GetAdditionalGroupsPath(additionalGroups []string, groupPath string) ([]int, error) {
+ var group io.Reader
+
+ if groupFile, err := os.Open(groupPath); err == nil {
+ group = groupFile
+ defer groupFile.Close()
+ }
+ return GetAdditionalGroups(additionalGroups, group)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.go b/vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.go
new file mode 100644
index 000000000..2cbb6491a
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/utils/cmsg.go
@@ -0,0 +1,95 @@
+// +build linux
+
+package utils
+
+/*
+ * Copyright 2016, 2017 SUSE LLC
+ *
+ * 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.
+ */
+
+import (
+ "fmt"
+ "os"
+
+ "golang.org/x/sys/unix"
+)
+
+// MaxSendfdLen is the maximum length of the name of a file descriptor being
+// sent using SendFd. The name of the file handle returned by RecvFd will never
+// be larger than this value.
+const MaxNameLen = 4096
+
+// oobSpace is the size of the oob slice required to store a single FD. Note
+// that unix.UnixRights appears to make the assumption that fd is always int32,
+// so sizeof(fd) = 4.
+var oobSpace = unix.CmsgSpace(4)
+
+// RecvFd waits for a file descriptor to be sent over the given AF_UNIX
+// socket. The file name of the remote file descriptor will be recreated
+// locally (it is sent as non-auxiliary data in the same payload).
+func RecvFd(socket *os.File) (*os.File, error) {
+ // For some reason, unix.Recvmsg uses the length rather than the capacity
+ // when passing the msg_controllen and other attributes to recvmsg. So we
+ // have to actually set the length.
+ name := make([]byte, MaxNameLen)
+ oob := make([]byte, oobSpace)
+
+ sockfd := socket.Fd()
+ n, oobn, _, _, err := unix.Recvmsg(int(sockfd), name, oob, 0)
+ if err != nil {
+ return nil, err
+ }
+
+ if n >= MaxNameLen || oobn != oobSpace {
+ return nil, fmt.Errorf("recvfd: incorrect number of bytes read (n=%d oobn=%d)", n, oobn)
+ }
+
+ // Truncate.
+ name = name[:n]
+ oob = oob[:oobn]
+
+ scms, err := unix.ParseSocketControlMessage(oob)
+ if err != nil {
+ return nil, err
+ }
+ if len(scms) != 1 {
+ return nil, fmt.Errorf("recvfd: number of SCMs is not 1: %d", len(scms))
+ }
+ scm := scms[0]
+
+ fds, err := unix.ParseUnixRights(&scm)
+ if err != nil {
+ return nil, err
+ }
+ if len(fds) != 1 {
+ return nil, fmt.Errorf("recvfd: number of fds is not 1: %d", len(fds))
+ }
+ fd := uintptr(fds[0])
+
+ return os.NewFile(fd, string(name)), nil
+}
+
+// SendFd sends a file descriptor over the given AF_UNIX socket. In
+// addition, the file.Name() of the given file will also be sent as
+// non-auxiliary data in the same payload (allowing to send contextual
+// information for a file descriptor).
+func SendFd(socket, file *os.File) error {
+ name := []byte(file.Name())
+ if len(name) >= MaxNameLen {
+ return fmt.Errorf("sendfd: filename too long: %s", file.Name())
+ }
+ oob := unix.UnixRights(int(file.Fd()))
+
+ return unix.Sendmsg(int(socket.Fd()), name, oob, nil, 0)
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/utils/utils.go b/vendor/github.com/opencontainers/runc/libcontainer/utils/utils.go
new file mode 100644
index 000000000..baa54c9ba
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/utils/utils.go
@@ -0,0 +1,127 @@
+package utils
+
+import (
+ "crypto/rand"
+ "encoding/hex"
+ "encoding/json"
+ "io"
+ "os"
+ "path/filepath"
+ "strings"
+ "unsafe"
+
+ "golang.org/x/sys/unix"
+)
+
+const (
+ exitSignalOffset = 128
+)
+
+// GenerateRandomName returns a new name joined with a prefix. This size
+// specified is used to truncate the randomly generated value
+func GenerateRandomName(prefix string, size int) (string, error) {
+ id := make([]byte, 32)
+ if _, err := io.ReadFull(rand.Reader, id); err != nil {
+ return "", err
+ }
+ if size > 64 {
+ size = 64
+ }
+ return prefix + hex.EncodeToString(id)[:size], nil
+}
+
+// ResolveRootfs ensures that the current working directory is
+// not a symlink and returns the absolute path to the rootfs
+func ResolveRootfs(uncleanRootfs string) (string, error) {
+ rootfs, err := filepath.Abs(uncleanRootfs)
+ if err != nil {
+ return "", err
+ }
+ return filepath.EvalSymlinks(rootfs)
+}
+
+// ExitStatus returns the correct exit status for a process based on if it
+// was signaled or exited cleanly
+func ExitStatus(status unix.WaitStatus) int {
+ if status.Signaled() {
+ return exitSignalOffset + int(status.Signal())
+ }
+ return status.ExitStatus()
+}
+
+// WriteJSON writes the provided struct v to w using standard json marshaling
+func WriteJSON(w io.Writer, v interface{}) error {
+ data, err := json.Marshal(v)
+ if err != nil {
+ return err
+ }
+ _, err = w.Write(data)
+ return err
+}
+
+// CleanPath makes a path safe for use with filepath.Join. This is done by not
+// only cleaning the path, but also (if the path is relative) adding a leading
+// '/' and cleaning it (then removing the leading '/'). This ensures that a
+// path resulting from prepending another path will always resolve to lexically
+// be a subdirectory of the prefixed path. This is all done lexically, so paths
+// that include symlinks won't be safe as a result of using CleanPath.
+func CleanPath(path string) string {
+ // Deal with empty strings nicely.
+ if path == "" {
+ return ""
+ }
+
+ // Ensure that all paths are cleaned (especially problematic ones like
+ // "/../../../../../" which can cause lots of issues).
+ path = filepath.Clean(path)
+
+ // If the path isn't absolute, we need to do more processing to fix paths
+ // such as "../../../../<etc>/some/path". We also shouldn't convert absolute
+ // paths to relative ones.
+ if !filepath.IsAbs(path) {
+ path = filepath.Clean(string(os.PathSeparator) + path)
+ // This can't fail, as (by definition) all paths are relative to root.
+ path, _ = filepath.Rel(string(os.PathSeparator), path)
+ }
+
+ // Clean the path again for good measure.
+ return filepath.Clean(path)
+}
+
+// SearchLabels searches a list of key-value pairs for the provided key and
+// returns the corresponding value. The pairs must be separated with '='.
+func SearchLabels(labels []string, query string) string {
+ for _, l := range labels {
+ parts := strings.SplitN(l, "=", 2)
+ if len(parts) < 2 {
+ continue
+ }
+ if parts[0] == query {
+ return parts[1]
+ }
+ }
+ return ""
+}
+
+// Annotations returns the bundle path and user defined annotations from the
+// libcontainer state. We need to remove the bundle because that is a label
+// added by libcontainer.
+func Annotations(labels []string) (bundle string, userAnnotations map[string]string) {
+ userAnnotations = make(map[string]string)
+ for _, l := range labels {
+ parts := strings.SplitN(l, "=", 2)
+ if len(parts) < 2 {
+ continue
+ }
+ if parts[0] == "bundle" {
+ bundle = parts[1]
+ } else {
+ userAnnotations[parts[0]] = parts[1]
+ }
+ }
+ return
+}
+
+func GetIntSize() int {
+ return int(unsafe.Sizeof(1))
+}
diff --git a/vendor/github.com/opencontainers/runc/libcontainer/utils/utils_unix.go b/vendor/github.com/opencontainers/runc/libcontainer/utils/utils_unix.go
new file mode 100644
index 000000000..c96088988
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/libcontainer/utils/utils_unix.go
@@ -0,0 +1,44 @@
+// +build !windows
+
+package utils
+
+import (
+ "io/ioutil"
+ "os"
+ "strconv"
+
+ "golang.org/x/sys/unix"
+)
+
+func CloseExecFrom(minFd int) error {
+ fdList, err := ioutil.ReadDir("/proc/self/fd")
+ if err != nil {
+ return err
+ }
+ for _, fi := range fdList {
+ fd, err := strconv.Atoi(fi.Name())
+ if err != nil {
+ // ignore non-numeric file names
+ continue
+ }
+
+ if fd < minFd {
+ // ignore descriptors lower than our specified minimum
+ continue
+ }
+
+ // intentionally ignore errors from unix.CloseOnExec
+ unix.CloseOnExec(fd)
+ // the cases where this might fail are basically file descriptors that have already been closed (including and especially the one that was created when ioutil.ReadDir did the "opendir" syscall)
+ }
+ return nil
+}
+
+// NewSockPair returns a new unix socket pair
+func NewSockPair(name string) (parent *os.File, child *os.File, err error) {
+ fds, err := unix.Socketpair(unix.AF_LOCAL, unix.SOCK_STREAM|unix.SOCK_CLOEXEC, 0)
+ if err != nil {
+ return nil, nil, err
+ }
+ return os.NewFile(uintptr(fds[1]), name+"-p"), os.NewFile(uintptr(fds[0]), name+"-c"), nil
+}
diff --git a/vendor/github.com/opencontainers/runc/vendor.conf b/vendor/github.com/opencontainers/runc/vendor.conf
new file mode 100644
index 000000000..9506b5c67
--- /dev/null
+++ b/vendor/github.com/opencontainers/runc/vendor.conf
@@ -0,0 +1,21 @@
+# OCI runtime-spec. When updating this, make sure you use a version tag rather
+# than a commit ID so it's much more obvious what version of the spec we are
+# using.
+github.com/opencontainers/runtime-spec v1.0.0
+# Core libcontainer functionality.
+github.com/mrunalp/fileutils ed869b029674c0e9ce4c0dfa781405c2d9946d08
+github.com/opencontainers/selinux v1.0.0-rc1
+github.com/seccomp/libseccomp-golang 32f571b70023028bd57d9288c20efbcb237f3ce0
+github.com/sirupsen/logrus a3f95b5c423586578a4e099b11a46c2479628cac
+github.com/syndtr/gocapability db04d3cc01c8b54962a58ec7e491717d06cfcc16
+github.com/vishvananda/netlink 1e2e08e8a2dcdacaae3f14ac44c5cfa31361f270
+# systemd integration.
+github.com/coreos/go-systemd v14
+github.com/coreos/pkg v3
+github.com/godbus/dbus v3
+github.com/golang/protobuf 18c9bb3261723cd5401db4d0c9fbc5c3b6c70fe8
+# Command-line interface.
+github.com/docker/docker 0f5c9d301b9b1cca66b3ea0f9dec3b5317d3686d
+github.com/docker/go-units v0.2.0
+github.com/urfave/cli d53eb991652b1d438abdd34ce4bfa3ef1539108e
+golang.org/x/sys 0e0164865330d5cf1c00247be08330bf96e2f87c https://github.com/golang/sys