This is more correct, as if st_dev == phys_dev check fails
we have to treat phys_dev as kdev for path resolve device
comparison.

Howver, this is not the case for non-btrfs FSs, and for the
latter one doesn't change anything as it uses anon devices
which are equal for kdev and odev cases.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

When dumping a ghost file we put real device in its header,
not the (btrfs) virtual one. This is done since we put real
devices into fsnotify images (we get them from proc). That
said on fsnotify ghost restore we don't need to do path
resolution, just devices compare.

And one more thing. When dumping device for ghost file for
_non_ btrfs case we have to convert stat dev_t into kernel
dev_t as all the other places in criu manipulate the latter
ones.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It's used by phys_stat_resolve_dev (broken by c5d2386a)
Signed-off-by: Andrew Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

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

Now we have more robust and fs agnostic path-resolution
engine for resolving dev conflicts.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Instead of scanning btrfs subvolumes (which can be even unaccessbile
if mount point lays on directory instead of subvolume itself) we use
path resolving feature here -- once we need to figure out if some
device number need to be altered up to mount point (as we know stat()
called on subvolume returns st_dev for subvolume itself, but not
one that associated with a superblock and shown in /proc/self/mountinfo
output).

This as well implies that we need to check if device number for ghost
files are to be updated to match mountinfo, thus we use phys_stat_resolve_dev
helper here.

After this patch the previously merged btrfs engine is no longer needed
(at least it seems so) and can be dropped.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This routine is aimed to find a mount point on which
the path passed as argument is laying on. We walk over
all mount points and see which one is matching.

Once found (in worst case it will be a root mount point
so function is never failing) we're checking if this is
btrfs and then return subvolume0 device id.

See commit 921cf873
for details what the hell we're doing here.

v2: rewrite mount_resolve_path w/o recursion
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Andrew Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

For paths resolution we will need mount tree to be parsed
and built, but it's not that simple -- the current code
implies that once parsed the tree must not be re-parsed
again, so we pass @parse argument from a caller: if a task
we're restoring do not use mount namespace, we should parse
mount tree early, otherwise defer this action until mount
tree is read from the image.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This variable should be carrying root of a parsed
mount tree pointed by @mntinfo. Thus if @mntinfo
get destroyed the @mntinfo_root should be set to NULL.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

If mount point is not a subvolume we need more complex logic
to figure out which subvolumes it might have. It'll be addressed
later, at moment simply don't fail in such case: for worst
scenarion we will fail later in CRIU code when devs are not
matching, in best case if application is not using notify,
sockets or deleted files -- we should continue without errors.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Subvolume 0 has own device number as well, don't miss it
when looking up for match.
Reported-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Tested-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Signed-off-by: Ruslan Kuprieiev <kupruser@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

And make use of it in for btrfs.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>

We will need to parse btrfs stuff, but this one is not
in the supported list yet (as it's bound to hardware).
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Acked-by: Cyrill Gorcunov <gorcunov@openvz.org>

In case if ghost file is on BTRFS subvolume we should not fail
but test if there is a match.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Because socket migh lay on btrfs volume (thus the device
number reported by diag module won't be the same as obtained
from stat(2)) we need to do an additional test and try
to resolve device number with help of btrfs engine.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This helper serves to hide fs specifics (in particular
btrfs) thus the caller won't need the details.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This helper will cause BTRFS engine to collect all subvolumes.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It is been found that BTRFS uses per subvolume device
numbering scheme  which causes some of our tests to fail
if they are laying on this filesystem.

Here we introduce minimal engine which parses subvolumes
present on mount point.

The interface is two functions

 - btrfs_parse_mountinfo: this one parses subvolume and caches the results
 - is_btrfs_subvol: test if device id is laying on subvolume

In the code we assume that cpu we're running on
is little endian, thus if one day we run on big
ending machine the code must be updated accordingly.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It will hold auxiliary data associated with mount point. We
will use it for btrfs handling but in future it can be extended.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need it for btrfs handling. Also print out the
FS type for easier debug

 | (00.003545)     type unsupported (cgroup) source cgroup 1c / @ /sys/fs/cgroup/blkio flags 30000e options blkio,
 | (00.003558)     type unsupported (cgroup) source cgroup 1d / @ /sys/fs/cgroup/perf_event flags 30000e options perf_event,
 | (00.003571)     type unsupported (cgroup) source cgroup 1e / @ /sys/fs/cgroup/hugetlb flags 30000e options hugetlb,
 | (00.003584)     type unsupported (ext4) source /dev/sda2 800002 / @ / flags 300000 options data=ordered,
 | (00.003670)     type tmpfs (tmpfs) source tmpfs 20 / @ /tmp flags 100000 options
 | (00.003696)     type unsupported (mqueue) source mqueue d / @ /dev/mqueue flags 300000 options
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

@path is always nil here, we actually need @remap->path
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

For debug purpose.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We allocate mount_info with xzalloc, no need for
additional NULL assignment.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

If we fail in xmalloc the function occasionally return
0 meaning that everything is fine. Don't do that, wait
until routine complete.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

First thing people typically do with criu is run something
in shell and try to dump it. This is typicall pitfall, as
in that case we face external sid and tty problem. Print a
message on stdout about this and advice wiki page to read.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Currently we have a bug: if no leave_running is set in request, service won't send dump response. We must not send response only if it was a self-dump request and no leave_running option was set.
Signed-off-by: Ruslan Kuprieiev <kupruser@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Signed-off-by: Tikhomirov Pavel <snorcht@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need it for additional handling once parsing
of mount entry is complete (in particular btrfs requires
additional processing to figure out subvolumes names).
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need it for btrfs subvolumes handling.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It's designed specificly for that.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

The error message is rather confusing people. In worst case (if
it happened that we need this uncollected socket), criu will
print out real error message later.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Useful for bug hunting.

 | (00.005209) unix: Dropping path /mnt/disk1/new_subvol/criu/test/zdtm/live/static/sockets00.test
 | for unlinked bound sk 0x26.0x1d4e real 0x23.0x1d4e
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

The next version is ready. It looks like it's time to make it
be the first major release, so here it is -- the v1.0. It's as
functional as v0.8 but with a lot of bug fixes and improvements.
And it's still :) compatible wrt images and cli options with
older criu-s.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Currently only addresses are compared. It's obviously not enough.

* First of all the parent vma must be private.
* Both vma-s must have the identical set of MAP_GROWSDOWN and MAP_FILES
  flags.
* Both vma-s must be linked to the same file.

https://bugzilla.openvz.org/show_bug.cgi?id=2824Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

shmid contains a file id for file mappings. It's required to determine,
which VMA-s are cowed. The parent maps a VMA and saves premmaped
address. Then  child trys to determing, which VMA-s must be inhereted
from parent, for that it compares addresses, flags and file id.

We don't want to transfer vma_area-s in restorer, so when a VMA entry is
copied in restorer memory, the premmaped address is save in shmid.
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Signed-off-by: Ruslan Kuprieiev <kupruser@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

Kenrel API is stable enough not to ignore it.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>