- 22 Jan, 2012 13 commits
-
-
Pavel Emelyanov authored
Test that 2 different exit()-ed zombies and 2 killed with different signals are handled properly. Signed-off-by:
Pavel Emelyanov <xemul@parallels.com> Signed-off-by:
Cyrill Gorcunov <gorcunov@openvz.org>
-
Pavel Emelyanov authored
Dump the core-pid.img file only. On restore select the way of killing task based on his exit_code -- exit or kill with a signal. Before dying unblock all the handlers and set SIG_DFL to it (to make the dead signal other than KILL be deliverable). Signed-off-by:
-
Pavel Emelyanov authored
The absent image file on shared resources preparation now means -- no resources for this pid (zombies will not have these files). This is not the most elegant solution, but I don't have anything better in mind. Need to think over, all the more so we're most likely about to reimplement the way image is stored some day in the future. Signed-off-by:
-
Pavel Emelyanov authored
Don't dump tasks in any state but T (stopped). Will extend it with Z soon. Signed-off-by:
-
Pavel Emelyanov authored
Introduce 3 states we will have to work with: * alive for tasks sleeping or running * dead for zombies * stopped for stopped tasks. We cannot distinguish tasks in this state now, but with freezer cgroup this will become possible Signed-off-by:
-
Pavel Emelyanov authored
Requires patch #14 (for kernel). Also check for number of entries read be at least required, not exactly equal for forward compatibility. Signed-off-by:
-
Pavel Emelyanov authored
Move common code for tasks and threads dumping into routine. It will be used for zombies as well. Signed-off-by:
-
Pavel Emelyanov authored
Kill all the macros for reading/writing image parts. New API looks like * write_img_buf/write_img Write an object into an image. Reports 0 for OK, -1 for error. The _buf version accepts object size as an argument, the other one uses sizeof() * read_img_buf/read_img Reads an object from image. Reports 0 for OK, -1 for error or EOF. * read_img_buf_eof/read_img Reads an object from image. Reports 1 for OK, 0 for EOF and -1 for error. This is not symmetrical with the previous one, but it was done deliberately to make it possible to write code like ret = read_img_bug_eof(); if (ret <= 0) return ret; /* 0 means OK, all is done, -1 means error was met */. ... /* 1 means object was read, can proceed */ Signed-off-by:
-
Pavel Emelyanov authored
Keep task arch-independent fields in one struct (will be extended) in the beginning of the image and make pads be located separately. Signed-off-by:
-
Pavel Emelyanov authored
Rename prep_cr_fdset_for_dump into cr_fdset_open and make it reentable, i.e. every next enter will open more files in the same fdset. Required for zombies and makes the code cleaner. Signed-off-by:
-
Pavel Emelyanov authored
This is to make sure we dump zombies w/o threads. Signed-off-by:
-
Pavel Emelyanov authored
Move the /proc/pid/stat parsing before anything in dump_one_task (prepare for zombies dumping) and cleanup the rest of the code accordingly. Signed-off-by:
-
Pavel Emelyanov authored
These are checking in proc parse routine anyway. Signed-off-by:
-
- 20 Jan, 2012 5 commits
-
-
Stanislav Kinsbursky authored
Such targets as "tags", "cscope" and "clean" doesn't require blob headers to be rebuilt. Thi patch remove this dependence. Signed-off-by:
Stanislav Kinsbursky <skinsbursky@parallels.com> Signed-off-by:
-
Stanislav Kinsbursky authored
Signed-off-by:
-
Stanislav Kinsbursky authored
We just don't need them Signed-off-by:
-
Cyrill Gorcunov authored
There was API change in linux kernel 3.2 -> 3.3 transition | commit c8991362a0d3cf317dfbfb6cb946607870654e6d Reported-by:
-
Cyrill Gorcunov authored
It's not needed anymore, it was handing cases where no fork-with-pid functionality were in kernel, but now it's simply unneeded. Also drop redundant getpid() calls. Passes all tests (except fork test which known to fail). Signed-off-by:
Pavel Emelyanov <xemul@parllels.com>
-
- 19 Jan, 2012 13 commits
-
-
Pavel Emelyanov authored
I haven't found anything similar in there. It currently just fails most of the launches because seize doesn't work with zombies, but I'm working on it... However, when it manages to miss the zombie it triggers the problem with freezer (fixed by previous patch). Signed-off-by:
-
Pavel Emelyanov authored
Currently we read the children tree of running processes. This is deadly wrong. Stop task once we open it's proc dir before descending to its kids. Signed-off-by:
-
Pavel Emelyanov authored
I'd appreciate if we get rid of these misleading macros at all eventually... Signed-off-by:
-
Pavel Emelyanov authored
Signed-off-by:
-
Pavel Emelyanov authored
After Andrey's work with making restorer a regular .o file we can do it (the pthread00 test doesn't fail on it). Signed-off-by:
-
Pavel Emelyanov authored
This eats response skb mem, but we don't need it. Signed-off-by:
-
Pavel Emelyanov authored
We can do it since we may not want to dump the missed sockets. But later, when we will start dumping containers, we'll better fail here, rather than in the dump stage. Signed-off-by:
-
Cyrill Gorcunov authored
Signed-off-by: -
Andrey Vagin authored
Handle SIGCHLD and if someone failed, nr_in_progress is set to -1. If crtools notices that nr_in_progress is negative, it kills all tasks. v2: * Use named constants for task_entries->start in restorer.c * Use SA_NOCLDWAIT when setting sigchild handler, this makes sigchild handler simpler. Signed-off-by:Andrey Vagin <avagin@openvz.org> Acked-by:
-
Andrey Vagin authored
Signed-off-by:
-
Andrey Vagin authored
Signed-off-by:
-
Andrey Vagin authored
Now we have only one mutex nr_in_progress, it says how many tasks are not restored yet. A negative value signs that someone failed. Signed-off-by:
-
Andrey Vagin authored
to wait while a futex value bigger than defined value. v2: s/bigger/greater Signed-off-by:
-
- 18 Jan, 2012 9 commits
-
-
Andrey Vagin authored
It's really like in VZ. Signed-off-by:
-
Andrey Vagin authored
./crtools CMD [options] CMD may be show, restore, dump Signed-off-by:
-
Cyrill Gorcunov authored
Just drop a couple of spaces and add a comment. Signed-off-by: -
Pavel Emelyanov authored
Signed-off-by:
-
Pavel Emelyanov authored
Signed-off-by:
-
Pavel Emelyanov authored
There's no need in img fd pointer in opening functions. Signed-off-by:
-
Cyrill Gorcunov authored
This reverts commit 46dc368f. There will be another one. Instead of hard rebase I decide to do a revert instead. Signed-off-by:
-
Andrey Vagin authored
Signed-off-by:
-
Stanislav Kinsbursky authored
This test is based on socket_aio. It's actually the same, except that socket binding and listening is done before checkpointing and connection to socket is done after migration. Signed-off-by:
-
