Explanation of exclude periodic test is that we can't set overrun for
timer, so all checks will fail in case of big overrun before dump.
Signed-off-by: Pavel Tikhomirov <snorcht@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Add fake timers to inshure that then timer ids don't go one after
another c/r go well.
Signed-off-by: Pavel Tikhomirov <snorcht@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
This also add function to check if time represented by two numbers is
valid. I.e. for timespec(sec, nsec), sec and nsec must be > 0 and nsec
must be less when NSEC_PER_SEC.
Signed-off-by: Pavel Tikhomirov <snorcht@gmail.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Only last iteration stops processes, all other iterations make
snapshots. Processes are restored from the last snapshot.
bash -x test/zdtm.sh -s -i 3
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
This patch prevents page dump failure screening
by the PARASITE_CMD_MPROTECT_VMAS command success.
Signed-off-by: Alexander Kartashov <alekskartashov@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
PARASITE_CMD_DUMPPAGES is called many times and the parasite args
contains an array with vmas at this time, so VMAs can be unprotected in
error case
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Since *all* of them just call do_dump_gen_file with proper ops,
just call one directly. Compacts the code.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
This code was written, when all threads were daemonized.
Now only a thread leader is daemonized.
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
For some reasons a task can go out from a parasite code.
Signed-off-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
In commit 2f5df09da5
we included asm/type.h for bool definition. This
is fine in terms of CRIU but makes cpt2 converter
to carry the whole asm/type.h, sometimes causing
error due to confilicting definitions.
So lets be simplier and include <stdbool.h> instead.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Just to protect ourself from unpredicted changes.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
The -v option is declared as "optional argument" and thus
only -v=N or -vN are valid. Currently used -v N syntax no
longer works as expected, this will be treated as LOG_ERR
level (single -v) with dangling (unparsed) argument.
Sorry for that :( but it's better to this earlier.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Kernel has more and more links with rtnl-ops, which report
a string kind of the device, which is handy for debugging.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Sys_mmap returns 'unsigned long' value. Contrary to x64, ARM has 32bit size of
'unsigned long'. If sys_mmap returns 32bit value, higher 32bit will be filled
0xffffffff.
Signed-off-by: Chanho Park <chanho61.park@samsung.com>
Signed-off-by: Myungjoo Ham <myungjoo.ham@samsung.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Reminder: the '&<string>' arg turns out into an address in memory
containing <string>. The '@<size>' arg turns out into an address
to uninitilized memory of size <size>, which contents it printed
after syscall returns.
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
If we meet shared mount point without share master belonging
to us -- it means we might fail on restore, thus require both
master/slave mount peers to be collected on dump.
In other words, the output will be like
| (00.077025) Error (mount.c:421): Mount 49 (master_id: 2 shared_id: 0) has unreachable sharing
| (00.077123) Error (mount.c:472): Can't proceed 4237's mountinfo
| (00.077865) Error (namespaces.c:442): Namespaces dumping finished with error 65280
https://bugzilla.openvz.org/show_bug.cgi?id=2608
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
Allocate it with xzalloc instead of massive
NULL assignment. Moreover, don't forget to
initialize @siblings.
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
The jenkins framework might need some lifting up
(crtools.$commit changed to criu.$commit).
Signed-off-by: Cyrill Gorcunov <gorcunov@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
This patch adds the test static/arm-neon00 modelled
after the test static/fpu00: it initializes ARM NEON
registers before a checkpoint, carries out a simple
computation on ARM NEON after a restore and compares
the result against the result of the same computation
carried out before a checkpoint.
Signed-off-by: Alexander Kartashov <alekskartashov@parallels.com>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
A struct sigframe* instead of fpu_state_t should be passed
to the routine restore_fpu() since FPU registers are stored
in different fields of the sigframe in different architectures.
An architecture-specific implementation of the routine restore_fpu()
should know details of this layout instead of construct_sigframe().
This change makes it possible to move ARM FPU restoration
from sigreturn_prep_fpu_frame() (where it caused a segfault
since the pointer fpu_stat has become invalid in the dumper
address space) to restore_fpu()
Signed-off-by: Alexander Kartashov <alekskartashov@parallels.com>
Acked-by: Andrey Vagin <avagin@openvz.org>
Signed-off-by: Pavel Emelyanov <xemul@parallels.com>
