Commits · 64e7d2435a105fa6fce4a441ca5b72400aaa71d1 · zhul / criu

20 Aug, 2013 9 commits

collect: Reduce amount of args to collect_image call · 64e7d243
Pavel Emelyanov authored Aug 20, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
64e7d243
rst: Compact file-descs collects a bit · 9917c4fe
Pavel Emelyanov authored Aug 20, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
9917c4fe
rst: Rework remaps collection on generic collect routine · e03ab6b4
Pavel Emelyanov authored Aug 19, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
e03ab6b4
pb: Free image object of collection doesn't need it · 0543f49c
Pavel Emelyanov authored Aug 19, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
0543f49c

restore: do not kill processes if not-all of them have been created · ef3ca3a1

Andrey Vagin authored Aug 20, 2013

If processes are restored without pidns, criu knows pidtheir -s from images,
but part of those task may have not yet forked, and thus the pids can not
exist or (!) be used by other processes.

To address that we abort stages RESTORE_NS and FORKING without killing tasks,
but with task_entries->start futex by writing STATE_FAIL into it and making
the tasks to check that. Since during RESTORE_NS and FORKING stages tasks can
only block on the mentioned futes, we can safely do it.
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

ef3ca3a1

zdtm: add a new test, where both sides of a tcp connection is dumped · 0c5b606b

Andrey Vagin authored Aug 20, 2013

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

0c5b606b

zdtm: clean up test_fork · 43292877

Andrey Vagin authored Aug 20, 2013

Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

43292877

test: create env.sh for environment setting · 156b1b81

Yicheng Qin authored Aug 20, 2013

Currently, env.sh only contains the location of criu executable.
Signed-off-by: Yicheng Qin <yichengq@google.com>
Acked-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

156b1b81

test/tcp/run.sh: fix permission · f1684be8

Yicheng Qin authored Aug 20, 2013

Signed-off-by: Yicheng Qin <yichengq@google.com>
Acked-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

f1684be8

16 Aug, 2013 10 commits

atomic: Use atomic_read instead of atomic_get · 71f7f754

Cyrill Gorcunov authored Aug 16, 2013

To switch to kernel's style.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

71f7f754

arm: atomic -- Add atomic_read/atomic_set helpers · 56cb0f8b

Cyrill Gorcunov authored Aug 16, 2013

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

56cb0f8b

atomic -- Switch to linux kernel templates · aea8a605

Cyrill Gorcunov authored Aug 16, 2013

Use same code as provided in kernel. In first place
we used own prototypes in case of simplicity (they
all were based on "lock xadd" instruction. There is
no more need for that and we can switch to well known
kernel's api.

Because kernel uses plain int type to carry atomic
counters I had to add explicit u32 type for futexes,
as well as a couple of fixes for new api usage.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

aea8a605

arm: atomic -- Use signed types for atomic_add_return/atomic_sub_return · 19391c85

Cyrill Gorcunov authored Aug 16, 2013

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

19391c85

x86: asm -- Add cmpxchg helpers · 862e0fb2

Cyrill Gorcunov authored Aug 16, 2013

Will need it for atomic data conversion.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

862e0fb2

sockets: inet -- Fix typo in message · 6d97dba7

Cyrill Gorcunov authored Aug 16, 2013

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

6d97dba7

restore: remove restore_one_fake, which duplicates the code · f824d8c3
Andrey Vagin authored Aug 16, 2013
```
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
f824d8c3
zdtm: Simplify TEST_LIST formation · 33c70a24
Pavel Emelyanov authored Aug 16, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
33c70a24

zdtm: Remove explicit namespaces specification on restore · a9e10c60

Pavel Emelyanov authored Aug 16, 2013

Yet again -- namespaces are auto-detected now, no need in this.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

a9e10c60

zdtm: Remove per-namespace list · aa7e079d

Pavel Emelyanov authored Aug 16, 2013

We have automatic namespaces detection, no need in this.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

aa7e079d

12 Aug, 2013 5 commits

rst: Remove threads restore serialization · 01f113ec

Pavel Emelyanov authored Aug 12, 2013

This thing was introduced by 01f8f8f4 to help not mixing
per-thread error messages in log files. Now messages are
not mixed by other means, so this thing is useless.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

01f113ec

rst: Check core near the place read it from image · 0a6da43d
Pavel Emelyanov authored Aug 12, 2013
```
Doing it several steps further is less clean.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
0a6da43d
rst: Make clean error path for root task restore · e0b8f914
Pavel Emelyanov authored Aug 12, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
e0b8f914
rst: Switch-and-wait creds restore stage with existing helper · 44a85ee0
Pavel Emelyanov authored Aug 12, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
44a85ee0

rstore: Invert switch_restore_stage logic · e5534e8e

Pavel Emelyanov authored Aug 12, 2013

Currentl it waits for previous stage to complete and starts the
next one. Now it starts the next one and waits for it to complete.
The latter way fits better into both -- the code and the head.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

e5534e8e

11 Aug, 2013 12 commits
- stats: Account total time to restore · 9b45833b
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  9b45833b
- stats: Account for time to fork tasks on restore · 2df39a4b
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  2df39a4b
- stats: Add timing stats for restore · c65f0684
  Pavel Emelyanov authored Aug 11, 2013
```
This will only work if timiings are reported by a single
task. Collecting them from several tasks is to be done.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  c65f0684
- rst: Reuse vma_list.priv_size for premapped area size calculation · ebd826c9
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  ebd826c9
- rst: Close vmas image earlier · 4c8e1c99
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  4c8e1c99
- rst: Write comment about why we need two calls to restore_pgid · e93566c7
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  e93566c7
- rst: Minor formatting fix · 4c83c432
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  4c83c432
- restore: Move kids (and session) restore into a helper · ede980c0
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  ede980c0
- rst: Rename read_vmas into prepare_vmas · 26aa61a3
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  26aa61a3
- util: Make set_proc_mountpoint static · 8ff15e5c
  Pavel Emelyanov authored Aug 11, 2013
```
And rename it to better reflect what it does.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  8ff15e5c
- util: Use open_detahc_mount helper for proc service fd creation · e2a0258d
  Pavel Emelyanov authored Aug 11, 2013
```
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  e2a0258d
- util: Make set_proc_fd report success/error only · cdce5d21
  Pavel Emelyanov authored Aug 11, 2013
```
Returning the new proc fd value is useless.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
```
  cdce5d21
14 Aug, 2013 4 commits

zdtm: exit with the zero code if everything is ok · 3e6f1ae8

Andrey Vagin authored Aug 14, 2013

Cc: Yicheng Qin <yichengq@google.com>
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

3e6f1ae8

test/zdtm: add option -n for batch test · 05745cee

Yicheng Qin authored Aug 14, 2013

In batch-test mode, the script skips failed individual tests
instead of exiting. This ensures that all zdtm tests can be run
in one pass, which provides better overview of test result.
Signed-off-by: Yicheng Qin <yichengq@google.com>
Acked-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

05745cee

test: adjust the output for tests · a7a787a4

Yicheng Qin authored Aug 14, 2013

Adjust the format of information printed.
Print out more necessary debug information.
Display error message in STDERR and make it more readable.
Signed-off-by: Yicheng Qin <yichengq@google.com>
Acked-by: Andrew Vagin <avagin@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

a7a787a4

mm: do check on shared pages only if a vma is inherited (v2) · 1f9da830

Andrey Vagin authored Aug 14, 2013

v2: restore page contents directly into memory
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>

1f9da830