- 13 Aug, 2015 8 commits
-
-
Cyrill Gorcunov authored
Seems was a typo. Signed-off-by:
Cyrill Gorcunov <gorcunov@openvz.org> Signed-off-by:
Pavel Emelyanov <xemul@parallels.com>
-
Cyrill Gorcunov authored
Signed-off-by:
-
Cyrill Gorcunov authored
Also use is_root_mount() helper instead of opencoded strcmp("./", m->mountpoint) and -Ex error codes in ERR_PTR. Signed-off-by: -
Cyrill Gorcunov authored
Signed-off-by:
-
Cyrill Gorcunov authored
It's quite unclean while this structure lives in proc_parse.h, which only have to fill this structure on procfs read, but real handling is inside mount.c. Move it as appropriate. Same time ext_mount structure should be moved into a header as well with sane @list name used instead of @l. Signed-off-by:
-
Cyrill Gorcunov authored
- gather structs on top - then extern vars Signed-off-by:
-
Cyrill Gorcunov authored
This a way easier for reading. Signed-off-by:
-
vkonyashkin authored
It 's necessary to align date, because the operand of movaps command must be aligned on a 16-byte (128-bit version) boundary or a general-protection exception (#GP) will be generated. Signed-off-by:
vkonyashkin <vkonyashkin@parallels.com> Acked-by:
Andrew Vagin <avagin@odin.com> Signed-off-by:
-
- 12 Aug, 2015 2 commits
-
-
Andrew Vagin authored
Jenkins runs all tests in a new session without a control terminal. Signed-off-by:
Andrew Vagin <avagin@openvz.org> Signed-off-by:
-
Andrew Vagin authored
Fixes: 94928cf8 ("test/rpc: include unistd.h") Signed-off-by:
-
- 10 Aug, 2015 9 commits
-
-
Cyrill Gorcunov authored
Signed-off-by:
-
Cyrill Gorcunov authored
Signed-off-by:
-
Cyrill Gorcunov authored
Signed-off-by:
-
Cyrill Gorcunov authored
For example we hit a case where systemd carries journal file with 4M in size. https://jira.sw.ru/browse/PSBM-38571Signed-off-by:
-
Cyrill Gorcunov authored
Signed-off-by:
-
Andrey Vagin authored
It allows to freeze processes faster Signed-off-by:
-
Andrey Vagin authored
Without using a freezer cgroup, we need to do a few iterations to catch all tasks, because a new tasks can be born. If new tasks appear faster than criu collects them, criu fails. The freezer cgroup allows to solve this problem. We freeze the freezer group, then attaches to tasks with ptrace and thaw the freezer cgroup. We suppose that all tasks which are going to be dumped in a specified freezer group. v2: fix comments from Christopher Reviewed-by:
Christopher Covington <cov@codeaurora.org> v3: refactor task_seize v4: fix comments from Pavel Cc: Christopher Covington <cov@codeaurora.org> Signed-off-by:
-
Andrew Vagin authored
Before cr_errno was called from open_pid_proc(), but now it isn't called if a root task doesn't exist. Reported-by: Mr Jenkins Fixes: b9b0730c ("ptrace: split task_seize into seize_catch_task and seize_wait_task") Signed-off-by:
-
Andrew Vagin authored
gcc -c test.c -I ./ test.c: In function ‘recv_resp’: test.c:18:8: warning: implicit declaration of function ‘read’ [-Wimplicit-function-declaration] len = read(socket_fd, buf, MAX_MSG_SIZE); ^ test.c: In function ‘send_req’: test.c:45:6: warning: implicit declaration of function ‘write’ [-Wimplicit-function-declaration] if (write(socket_fd, buf, len) == -1) { ^ test.c: In function ‘main’: test.c:165:2: warning: implicit declaration of function ‘close’ [-Wimplicit-function-declaration] close(fd); Reported-by: Mr Jenkins Signed-off-by:
-
- 07 Aug, 2015 9 commits
-
-
Pavel Emelyanov authored
Signed-off-by: -
Andrey Vagin authored
Signed-off-by:
-
Andrey Vagin authored
It's required to dump uid-s and gid-s from this userns. Signed-off-by:
-
Andrey Vagin authored
It's required for dumping tmpfs, where we use tar to save content. If we need to execute tar from a proper userns to get right uid-s and gid-s for files. Signed-off-by:
-
Andrey Vagin authored
uid and gid are shown in the init userns. We are going to restore mounts in a target userns, so we need to set these options in the target userns. Signed-off-by:
-
Gabriel Guimaraes authored
This test checks that the OverlayFS fixup for CRIU works correctly by dumping and restoring a process that has an open file in an OverlayFS mounted directory. This should work no matter if the Kernel has the OverlayFS bug or not. Signed-off-by:
Gabriel Guimaraes <gabriellimaguimaraes@gmail.com> Signed-off-by:
-
Gabriel Guimaraes authored
This is here only to support the Linux Kernel between versions 3.18 and 4.2. After that, this workaround is not needed anymore, but it will work properly on both a kernel with and without the bug. The bug is that when a process has a file open in an OverlayFS directory, the information in /proc/<pid>/fd/<fd> and /proc/<pid>/fdinfo/<fd> is wrong, so we grab that information from the mountinfo table instead. This is done every time fill_fdlink is called. We first check to see if the mnt_id and st_dev numbers currently match some entry in the mountinfo table. If so, we already have the correct mnt_id and no fixup is needed. Then we proceed to see if there are any overlayFS mounted directories in the mountinfo table. If so, we concatenate the mountpoint with the name of the file, and stat the resulting path to check if we found the correct device id and node number. If that is the case, we update the mount id and link variables with the correct values. Signed-off-by:
-
Andrey Vagin authored
It's preparation to use a freezer cgroup for freezing tasks. Signed-off-by:
-
Ruslan Kuprieiev authored
Signed-off-by:
Ruslan Kuprieiev <rkuprieiev@cloudlinux.com> Signed-off-by:
-
- 06 Aug, 2015 3 commits
-
-
Pavel Emelyanov authored
Signed-off-by: -
Andrey Vagin authored
Signed-off-by:
-
Hui Kang authored
Signed-off-by:
Hui Kang <hkang.sunysb@gmail.com> Looks-good-to: Ruslan Kuprieiev <kupruser@gmail.com> Signed-off-by:
-
- 05 Aug, 2015 7 commits
-
-
Andrey Vagin authored
Currently we skip all userns tests if "criu check" fails. This code was written when the "freature" options of criu check didn't exist. Currently we add userns tests only if "criu check --feature userns" passes. Signed-off-by:
-
Cyrill Gorcunov authored
Dynamically allocated @name doesn't release if error happened. | ** CID 129898: (RESOURCE_LEAK) | /sk-unix.c: 505 in unix_process_name() | /sk-unix.c: 509 in unix_process_name() | /sk-unix.c: 519 in unix_process_name() Signed-off-by:
-
Andrey Vagin authored
They should do mostly the same, but the first one checks nothing after c/r. v2: remove sockets03 from Makefile Signed-off-by:
-
Andrey Vagin authored
because we are going to restore data of peer. Anyway this is wrong, because we need to restore a message with a sender address. Signed-off-by:
Tycho Andersen <tycho.andersen@canonical.com> Signed-off-by:
-
Andrey Vagin authored
Signed-off-by:
-
Andrey Vagin authored
Signed-off-by:
-
Andrey Vagin authored
This test fails currently, because these data are not restored. v2: remove the test from the execution list. Signed-off-by:
-
- 03 Aug, 2015 2 commits
-
-
Christopher Covington authored
If we want one CRIU binary to work across all AArch64 kernel configurations, a single task size value cannot be hard coded. While trivial applications successfully checkpoint and restore on AArch64 kernels with CONFIG_ARM64_64K_PAGES=y without this patch, replacing the remaining use of the hard-coded value seems like the best way to guard against failures that more complex process trees and future uses may expose. Signed-off-by:
-
Christopher Covington authored
If we want one CRIU binary to work across all AArch64 kernel configurations, a single task size value cannot be hard coded. Since vma_area_is_private() is used by both restorer blob code and non restorer blob code, which must use different variables for recording the task size, make task_size a function argument and modify the call sites accordingly. This fixes the following error on AArch64 kernels with CONFIG_ARM64_64K_PAGES=y. pie: Error (pie/restorer.c:929): Can't restore 0x3ffb7e70000 mapping w> pie: ith 0xfffffffffffffff7 Signed-off-by:
-
