We expect:

- all 9 scripts are called
- there's always images dir variable
- for 7 of those scripts there's root-pid variable
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
Acked-by: Andrew Vagin <avagin@virtuozzo.com>

On restore root_item may exist, but its real pid is not yet resolved
Fixes e5e8f791.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
Acked-by: Andrew Vagin <avagin@virtuozzo.com>

--record option allows us to keep track of files that are being
installed by writing them to specified file. We can than use that
file to do proper cleanup on uninstall.

v2, drop -r, as we shouldn't really care about dirs, because
    setup.py doesn't report them to us.
Signed-off-by: Ruslan Kuprieiev <kupruser@gmail.com>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

qemu-static from ubuntu/xenial contains a required fix
and now we can execute compilation for ppc64le in travis.

v2: rebase on criu-dev
v3: set QEMU_CPU=POWER8 to avoid segfaults of qemu-ppc64le-static

> > By default, both qemu and gcc5 produce code on ppc64el which will run on
> > POWER7 platform.  However, in debian gcc build instruction, gcc is explicitly
> > built to produce code for POWER8 platform, which, obviously has additional
> > instructions.

v4: use ubuntu:trusty. xenial and wily has a bug:
Processing triggers for libc-bin (2.21-0ubuntu4.1) ...
/sbin/ldconfig.real: 5: /sbin/ldconfig.real: Syntax error: ")" unexpected
/sbin/ldconfig.real: 5: /sbin/ldconfig.real: Syntax error: ")" unexpected

v5: arm, x86, aarch64 were removed by mistake
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

v2: Don't clear_breakpoints too.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

This is log cleanup patch.
Pr_perror prints new line by itself.
Signed-off-by: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

The IMAGES_DIR and INIT_PID are constants so it's
enough to set the only once per criu.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

We effectively have two modes of actions -- call shell
scripts and call RPC. So make ths code work such. As a
side effect -- we don't mess with environment variables
in case of RPC notifications.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

This lets us clean the action-script.c eventually.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

The same -- it's internal action-scripts.c business.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

It's used purely inside action-scripts.c, so move
it there not to pollute the global opts.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Fixes: 1dc2f87b33ce ("util: shutdown log in cr_system_userns if error fd is negative")
Cc: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

While most are handled from the scratch there is a significant
problem with python setup. So I added some preliminaty solution,
probably someohe with good knowledge of how setup.py works
improve it later.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Will need it for uninstall action.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Otherwise error in case of exec error won't be printed
The problem is that when err fd is negative, it's replaced by log fd.
Then err is moved to STDERR (that means, that log_fd is _closed_).
But log facility still consider log fd as valid and tries to use it to print
error message in case of exec failure.
Which is equal to writing to /dev/null, basically.
This patch shutdown log, if err fd was negative, thus forcing criu to output
exec error to STDERR (which was replaced by log fs, btw).
Signed-off-by: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

If several copies of binfmt_misc test are running, we may fail because of duplicate names.
Fix that by giving a custom name for registered /proc/sys/fs/binfmt_misc/xxx.

################## Test zdtm/static/binfmt_misc FAIL at start ##################
Test output: ================================
18:03:33.594: 4: string: :magic_file:M:9:\x73\x51\xff\x4a\xec\x29\xcd\xba\xab\xf2::/bin/interpreter:OCP
18:03:33.594: 4: FAIL: binfmt_misc.c:133: write magic_file (errno = 22 (Invalid argument))
18:03:33.594: 3: ERR: test.c:204: Test exited unexpectedly with code 1
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

When user namespace are stepping in we should wait for their preparation
to complete before start using userns daemon (internally the kernel
checks for uids and if uids are not set -EINVAL will be returned
when usersn calls for sendmsg()).

v2: change stages correctly
v3: Wait prepare_userns() before making any calls to usernsd()

Cc: Cyrill Gorcunov <gorcunov@gmail.com>
Acked-by: Cyrill Gorcunov <gorcunov@gmail.com>
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Fixes small bug:
__open_proc() macro is called from parse_pid_loginuid as:
>	fd = __open_proc(pid, (ignore_noent) ? ENOENT : 0,
>			O_RDONLY, "loginuid");

So, ier parameter is badly expanded with current version:
>	if (__fd < 0 && (errno != (ignore_noent) ? ENOENT : 0)

Which in result does not hide "No such file" error on feature
test (at least with arm-gcc). Not a big deal, tho.
Reported-by: alex vk <avankemp@gmail.com>
Signed-off-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Impact: little optimization
Signed-off-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

drop_unicast_in_l2_multicast and drop_gratuitous_arp appeared in MS:

commit 12b74dfadb5a ("ipv4: add option to drop unicast encapsulated in
L2 multicast")
commit 97daf3314550 ("ipv4: add option to drop gratuitous ARP packets")
Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

We found that linkat for "unsafe" files doesn't work in userns
if a file uid isn't equal to the currect fsuid. This issue was
fixed by changing fsuid before calling linkat. But in this
case we are not able to createa link if a target directory doesn't
have write premissions.

Starting with the 4.3 kernel, it's possible to create links of
"unsafe files":

f2ca379642d7 ("namei: permit linking with CAP_FOWNER in userns")

So we can try to call linkat() without changing fsuid and make one
more attempt with changing fsuid if the first one failed with EPERM.
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

When not set up it leaves empty making setup of
pycriu python library wrong. Define it as root.
Reported-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Cc: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Acked-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Do not fail, if task doesn't have open files. This fixes
zdtm/static/fd test, which is broken now:

(00.004411) Error (cr-dump.c:1312): Dump files (pid: 25) failed with -1
(00.004548) Error (cr-dump.c:1614): Dumping FAILED.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Acked-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

It might be nil on predump, so dereference only when present.
Signed-off-by: Cyrill Gorcunov <gorcunov@virtuozzo.com>
Acked-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

The only one (so far) "post-resume" script needs some pid to join its
Signed-off-by: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

TCP code doesn't need one any longer.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Don't mess with sysctl, just try sending queues in greedy mode
shrinking the max_chunk eventually.

This is needed for 2 reasons -- first, to get rig of reading the
max_rshare sysctl and to make libsoccr possible and simple.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

CID 159475 (#1 of 1): Unchecked return value from library (CHECKED_RETURN)
1. check_return: Calling lseek(img_raw_fd(img), pkt->entry->length, 1)
   without checking return value. This library function may fail and return an error code.
Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

The only one (so far) "post-resume" script needs some pid to join its
namespaces. Let it be containers root.

v4:
1) Enviroonment setup has been was moved to run_scripts
2) Environment variable NS_PID was renamed to CRTOOLS_INIT_PID
Signed-off-by: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

This hook is called at the very end, when everything is restored and processes
were resumed.
Can be used for some actions, which require operation container, like
restarting of systemd autofs services.

v3:
Call "post-resume" scripts before detach option check (to make sure, that
restored process(es) are still alive

v2:
Remove code chunk, escaped from debugging
Signed-off-by: Stanislav Kinsburskiy <skinsbursky@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Here's the first once-a-month release with whatever is there in
the master branch. We have quite a few new features, but a lot of
bugifxes :)
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>

Signed-off-by: Andrew Vagin <avagin@virtuozzo.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>