Now fdinfos are collected independently from reg files and
sockets. During tihs collect we effectively create the mirroring
list of both by checking which type-IDs are added first.

Fix this by removing the fdinfo_desc and attaching fds directly
to collected reg files and sockets. Pipes and unix sockets will
be reworked in the same manner soon.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Same as prev 2 patches now for the unix sockets. They are still in per-pid image files, but
this is going to change soon (I hope).
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Same as previous patch -- pull inet sockets in and store in the hash for the same reasons.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Instread of re-reading this image again and again on every fd restore, pull the
reg-files.img in early and store the entries in a hash. This will simplify the
further fd restoring fixes and will allow for dump/restore via a stream (socket).
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

At moment no-deps-targets is empty thus changes
in headers do not cause correspond source files
to be rebuilt. Fix it.
Reported-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This object file will be handled by ld script so .text
sections will be squashed into one common section.
Thus no need to force these functions being always-inlined.

It's a rudiment from time when we were not using ld at
all but #include source code file itself and now this
attribute might confuse code readers since it's
unclear _why_ it's needed at all.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

There is no need for specific parasite or restorer
head sections, a general "head" one is enough.

Also .stack section is dropped since it's never used.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Otherwise I get

 | sockets.c: In function ‘dump_socket’:
 | sockets.c:467: error: ‘st.st_ino’ may be used uninitialized in this function
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Since kernel commit 9c501935a3 the sys/socket.h
should be included prior the linux/netlink.h.

Make it so.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It has been used when debuggin without PTRACE_SEIZE,
not needed anymore.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It's wrong to treat restore_thread_exec_start as arguments
area (I managed to overlook this problem in commit
01484182) it's rather
a function start address.

The thread arguments area allocated dynamically after the
restorer blob itself.

We didn't hit any problems earlier simply because there
were a few bytes owerwritten in function prologue.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
CC: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Otherwise I'm getting error
 |
 | files.c: In function ‘prepare_fds’:
 | files.c:329:23: error: ‘tmp’ may be used uninitialized in this function [-Werror=uninitialized]
 | files.c:309:6: note: ‘tmp’ was declared here
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

It's used only to check whether we should do file re-send and verify we've served
all the cliens. This can be replaced with proper list manipulations.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

Since now we have local copies of a remove FDs we can dump socket queus without entering
a parasite code. This makes the code MUCH simpler.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We have two checks for fd being a chrdev:

One to skip stdio-s that are termilans and
The other one for any fd being a /dev/null or other special device.

Clean this a little bit.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Move the preID generation from dump_one_fdinfo to make it look cleaner.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

A bugfix. Drained fds should be closed after we are done with them.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

The statfs is not required, we now check for fd being a socket with S_IFSOCK.
The 2nd stat is just not needed, the caller provides stat info.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

One can be read from fd_parms.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This structure is actualy an fd parameter, so put it there. This will also
help with future patching.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Try meant that it might not be a socket, but with local fds we do know
it for sure.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Since we're able to simply drain file descriptors
we're interested in, just do that with parasite
help.

This changes the calling sequence a bit

 1) Collect file descriptors before parasite
    get intected the dumpee and remember them in
    local copy.

 2) Ask parasite to drain collected descriptrs
    into our space.

 3) Operate with file descriptors directly via fcntl
    calls and such.

Overall idea is to prepare ground for fowners
dumping which will be addressed in further patches.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Just implemented but not yet used in dumping procedure,
this will be addressed in further patches.

Note the space for file descriptors is statically allocated
in 8K arguments area, not on stack.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need these helpers to transfer file
descriptors from dumpee to our space.

Also make send_fd/recv_fd to be a wrappers over
send_fds/revc_fds to not duplicate the code.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

To be able to use scm_fdset structure the two helpers
added scm_fdset_init_chunk and scm_fdset_init.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

This structure will serve for multiple fds transmission/receive.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need them in file descriptors transfer addressed in further patches.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

We will need it in parasite code where we can't use libc functions.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

It's unused and wrong.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

This is not good to update images while restoring.

Thus, read vma_entry-es once into a list, put opened (when required) fds
in there and make restorer walk the entries in mem, not those read from
the image file.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

These chunks of memry, which transit into restorer code gets unmapped one-by-one and thus each of them
should be page-aligned.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

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

There's no need in seeking file on core dump.
Fix explicit sizeof(u32)-s in restore seeks.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

IOW, show family, type, state and proto as strings, not bare numbers.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Now every inetsk fd dump results in a new entry in the fdinfo.img file. Sockets itself are
dumped into inetsk.img global image file. On restore the generic fdinfo redistribution algo
is used and inet sockets are opened only when required.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

Currently only type is taken into account, but further we will need to distinguish
types as well.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>