Picked from patch "[PATCH RFC] namespaces: use CLONE_VFORK
with CLONE_VM when it is possible" by Andrew Vagin.
Currenly parent touches child's stack, as in moment of clone() call
its stack pointer is above the child's (we allocate char stack[128]
on parent's stack). This prevents to create CLONE_VM|CLONE_VFORK
processes, because the child uses stack addresses occupied by parent.
The patch changes clone_noasan() behaviour and allows to do that
with the same memory consumption. We give a child memory, which
is not used by parent clone(), so parent's and child's stacks
have no tntersection.
This allows to create CLONE_VM|CLONE_VFORK processes.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
(Was "user_ns: Block SIGCHLD during namespaces generation")
We don't want asynchronous signal handler during creation
of namespaces (for example, in create_user_ns_hierarhy())
as we do wait() synchronous. So we need to block the signal.
Do this once globally.
v2: Set initial ret = 0
v3: Block signal globally in root_item before its children
are created.
v4: Move block to prepare_namespace()
Suggested-by: Andrew Vagin <avagin@virtuozzo.com>
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
In next patches usernsd will need to create transport
socket in the same net_ns as other tasks do their
TRANSPORT_FD_OFF sockets.
Choose criu net_ns for that: this allows usernsd
to do not wait for creation of other net_ns, i.e.
to do not introduce new dependencies between tasks.
In case of (root_ns_mask & CLONE_NEWUSER) != 0
root_item's user_ns does not allow to restore criu net_ns,
so do prepare_net_namespaces() in sub-process to do not
lose criu net.
v3: Introduce __prepare_net_namespaces and execute it in cloned task.
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Since net ns is assigned after prepare_fds() and,
in common case, at the moment of open_ns_fd() call
task points to a net ns, which differs to its target
net ns, we can't get the ns from a task. So, get it
from fdstore. Also, support userns ns fds.
v2: Add comment
Signed-off-by: Kirill Tkhai <ktkhai@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
We shave a test case for external veth devices. This test case
checks veth devices which are living in two dumped network
namespaces.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
A network device, which is connected to a bridge, is restored
after the bridge. In this case we can set the master attribute and
the device will be connected to the bridge automatically.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
When we dump a veth device, the kernel reports where a peer device lives
and we use this information to restore this veth pair.
On restore we set a net ns id for a peer and it is created in the required
netns.
v2: add more comments
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
In each network namespace we can set an id for another network namespace
to be able to address it in netlink messages.
For example, we can say that a peer of a veth devices has to be created
in a network namespace with a specified id. If we request information about
a veth device, a kernel will report where a peer device lives.
An user are able to set this ID-s, so we have to dump and restore them.
v2: add more commetns
v3: make a union of nsfd_id and ns_fd, they are not used together
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Each network namespaces has a list of ID-s for other namespaces,
so if we request infomation about a veth device, we get an id
for a namespace of a peer device.
These ID-s can be set by users or by kernel when they are required.
CRIU has to restore these ID-s for network namespaces. We have to
remember that one netns can have different id-s in different network
namespaces.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
It is possible to assign id for network namespaces and
this id will be used by the kernel in some netlink messages.
If no id is assigned when the kernel needs it, it will be
automatically assigned by the kernel.
For example, this id is reported for peer veth devices.
v2: add a comment
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
This tests create a few processes which live in three network namespaces
and have a few sockets which are created in different network namespaces.
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Each sockets belongs to one network namespace and operates
in this network namespace.
socket_diag reports informations about sockets from
one network namespace, but it doesn't report sockets which
are not bound or connected to somewhere. So we need to have
a way to get network namespaces for such sockets.
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
This ioctl is called for a socket and returns a file descriptor
for network namespace where a socket has been created.
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Each socket has to be restored from a proper network namespaces
where it was created.
We set a specified network namespace before restoring a socket.
A task network namespace is set after restoring all files.
v2: don't set the root netns for transport sockets
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
We need this to avoid conflicts with file descriptors,
which has to be restored.
Currently open_proc_pid() doesn't used during restoring
file descriptors, but we are going to use it to restore
sockets in proper network namespaces.
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Restore all network namespaces from the root task and then set
a proper namespace for each task after restoring sockets, because
we need to switch network namespaces to restore sockets.
Each socket has to be created in a proper network namespace.
v2: fix a typo bug
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Each socket has to be restored in a proper namespaces where
it has been created.
Here is an issue about unconnected and unbound sockets,
they are not reported via socket-diag and we can't to
get their network namespaces.
v2: add a comment before get_socket_ns()
remove nsid from sk_packet_entry
Acked-by: Pavel Emelyanov <xemul@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
PID ussualy means processs ID, but prepare_net_ns works with namespaces.
travis-ci: success for Dump and restore nested network namespaces (rev4)
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
ns_id will be used to collect sockets and other per-netns
resources
travis-ci: success for Dump and restore nested network namespaces (rev4)
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
There are a number of global variables around this descriptor
(e.g. open_proc_fd) and their values are saved in memory which
are not shared between processes.
travis-ci: success for Dump and restore nested network namespaces (rev4)
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Reviewed-by: Dmitry Safonov <dsafonov@virtuozzo.com>
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
This time we shifted the update one month, due to the amount of
changes not being very big. And got pretty big update, including
a set of bugfixes and new functionality.
Signed-off-by: Pavel Emelyanov <xemul@virtuozzo.com>
It has two arguments "pos_l and "pos_h" instead of one "off". It is used
to handle 64-bit offsets on 32-bit kernels.
SYSCALL_DEFINE5(preadv, unsigned long, fd, const struct iovec __user *, vec,
unsigned long, vlen, unsigned long, pos_l, unsigned long, pos_h)
https://github.com/checkpoint-restore/criu/issues/424
Signed-off-by: Andrei Vagin <avagin@openvz.org>
Reviewed-by: Dmitry Safonov <0x7f454c46@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Due to way CRIU handles paths (as relative to workdir), there's a case,
where migration would fail. Simple example is a ghost file in filesystem
root (with root being cwd). For example, "/unlinked" becomes "unlinked".
And original code piece scans path for other slashes, which would be
missing in this case. But it's still a perfectly valid case, and there's
no need to fail. So if there's no parent dir - we just don't need to
create one and we can just return 0 here instead of failing.
Signed-off-by: Vitaly Ostrosablin <vostrosablin@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
The kernel doesn't have an interface to get a sent queue for udp
sockets, so currently we can't dump them and criu dump has to fail in
such cases.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Now we block all sockets with non-zero idiag_wqueue, but it doesn't mean
that a CORK option is enabled for a socket. A packet can be in a network
stack and it is accounted into idiag_wqueue.
https://github.com/checkpoint-restore/criu/issues/409
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Except for several false positives done by:
find -type f -name "*.c" -not -path "./test/*" -exec sed -i
's/\(\<pr_err.*[^\][^n]\)\("[,)]\)/\1\\n\2/g' {} \;
Signed-off-by: Pavel Tikhomirov <ptikhomirov@virtuozzo.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Leases can be set only on regular files. Thus, as optimization we can
skip attempts to find associated leases in 'correct_file_leases_type'
for other fd types.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
-- check childs' errors in file_leases03
-- test c/r of lease transfered to child process
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
CRIU creates dictinct lock record for each file descriptor on the same
OFD. The patch removes this duplicates. To do so, it adds new field into
struct file_lock, which stores pid of fd, on which lock was found.
'owner pid' is not actually helpful, because the original fd, on which
lock have been set, can be already closed.
Also it purges crutches doing the same stuff but only for file leases.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Test cases:
0. Basic non-breaking read/write leases.
1. Multiple read leases and OFDs with no lease for the same file.
2. Breaking leases.
3. Multiple fds (dup + inherited) for single lease (mutual OFD).
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Information about locks in /proc/<pid>/fdinfo is presented only since
kernel v4.1. This patch adds logic to *note_file_lock* to match leases
and OFDs.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
restore of breaking leases is executed in 2 steps:
1. restore the lease in a state it was before break
2. break it by opening associated file.
The patch fixes type of broken leases to 'target lease type',
because procfs always returns 'READ' in this case.
Also, it adds 'updated' field in lock structure. It's used to remove all
duplicated records for single lease from the image, which wasn't
corrected by 'correct_lease_type'.
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Leases in breaking state are not supported. In that case criu will
report an error during the dumping. Also lock info in
/proc/<pid>/fdinfo should be presented (since kernel 4.1).
Before taking out new lease it modifies process fsuid to match file uid
(see fcntl F_SETLEASE).
Signed-off-by: Pavel Begunkov <asml.silence@gmail.com>
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Now it's probably one valide use case, because there is no way to commit
a container when a container is being checkpointed.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
All static tests has to stop any activity before C/R.
./tempfs_subns --pidfile=tempfs_subns.pid --outfile=tempfs_subns.out --dirname=tempfs_subns.test
Run criu dump
Unable to kill 128: [Errno 3] No such process
Run criu restore
7: Old mounts lost: []
7: New mounts appeared: [('/rootfs/criu/test', '/'), ('/', '/proc'), ('/', '/dev/pts')]
:
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
Currently we use an additional pipe to steal data from a pipe, but we
don't check that we steal all data. And the additional pipe can have a
smaller size.
Signed-off-by: Andrei Vagin <avagin@virtuozzo.com>
