Sung-jae Park [Mon, 20 Aug 2012 08:10:12 +0000 (17:10 +0900)]
Check the created viewer window state(visibility)
Change-Id: I608502ef08daf7b2e7179817705c3202b45ecf08
Sung-jae Park [Tue, 14 Aug 2012 07:21:58 +0000 (16:21 +0900)]
Fix the build error
Change-Id: If551d6df84736e8e7bd1a20d37603f97501492c4
Sung-jae Park [Tue, 14 Aug 2012 03:02:44 +0000 (12:02 +0900)]
Update buffer management code
Add PIXMAP (but not yet implemented)
Sung-jae Park [Sat, 11 Aug 2012 11:00:33 +0000 (20:00 +0900)]
Update OSP provider package name
Change-Id: I3378a814d098daf84fb3641dcaba76ea429534cf
Sung-jae Park [Sat, 11 Aug 2012 08:57:44 +0000 (17:57 +0900)]
Use the separated content folder for reader
Change-Id: I7e11c2731200ace9b233c2d56a9d60426fd3afb7
Sung-jae Park [Sat, 11 Aug 2012 08:23:10 +0000 (17:23 +0900)]
Initiate the DB loader
Change-Id: I2e42a9e2be3398838c7c275b4d0af7215b9d6494
Sung-jae Park [Sat, 11 Aug 2012 05:30:47 +0000 (14:30 +0900)]
Merge with old repository
Change-Id: I6128e8904d5eff2e9f590a8e8ae84826246b1ca3
Sung-jae Park [Mon, 6 Aug 2012 11:17:30 +0000 (20:17 +0900)]
Close the PD if the client is deactivated.
Change-Id: Ica04645071f6123e79ef0ec7f745fda1345a9f43
Sung-jae Park [Thu, 2 Aug 2012 12:08:54 +0000 (21:08 +0900)]
Add livebox type 700x172
Don't send pause/resume event if the slave is not activated
Change-Id: I944b9a018e91a7dc404cba3c3f6ee66d3746d88c
Sung-jae Park [Thu, 2 Aug 2012 08:52:23 +0000 (17:52 +0900)]
Implementing the free space calculation
Change-Id: I265acfe6d2e79f89d6841f41c976a6cae47227f9
Sung-jae Park [Thu, 2 Aug 2012 03:59:56 +0000 (12:59 +0900)]
Update livebox package manager parser plugin
Remove useless damage_set function
Change-Id: Ib5cf75ab49b45486b3cf571a6772b247f93b716d
Sung-jae Park [Wed, 1 Aug 2012 07:56:32 +0000 (16:56 +0900)]
Revise the event callback management code
Change-Id: I8655fae39050b9af5cb27f11690d01a3b3935f61
Sung-jae Park [Mon, 30 Jul 2012 09:58:56 +0000 (18:58 +0900)]
Try to create liveboxes in the system cluster.
Update acquire_buffer status value for web-slave
Fix the bug of acquire_buffer
Change-Id: I5cf6515d8bb970bc5866943f5e30b3ae594ba587
Sung-jae Park [Sat, 28 Jul 2012 15:47:43 +0000 (00:47 +0900)]
SEL Verification - bug fixed
S1-6489
S1-6520
S1-6337
S1-6494
S1-5750
P120725-0076
Change-Id: I32d7b8f3b0f050fe48478cb3d4e5266e4d1410cc
Sung-jae Park [Sat, 28 Jul 2012 06:24:13 +0000 (15:24 +0900)]
use the casecmp for comparing the cluster name
Change-Id: Ie9181711b045fddfabb69572e6b5b79e128a02d2
Sung-jae Park [Fri, 27 Jul 2012 08:35:26 +0000 (17:35 +0900)]
Updated for supporting the web-slave
Change-Id: I2db5d4e542eec15d5191ad6d397152a8cb2e52fb
Sung-jae Park [Fri, 27 Jul 2012 07:34:39 +0000 (16:34 +0900)]
Update the cluster/sub-cluster ID
Change-Id: Ia5ad6f8eadaa3ff6243e2becb0d4b0f70c8b4ce8
Sung-jae Park [Fri, 27 Jul 2012 06:20:51 +0000 (15:20 +0900)]
Send move event right before handling the up event
Change-Id: Ie8d8f615d77ce67fe700b834f1839ff9b0c0e51a
Sung-jae Park [Fri, 27 Jul 2012 04:37:11 +0000 (13:37 +0900)]
Add debugging message
Change-Id: I1b594a0fe099a569078f37defc713d7f4d511e1a
Sung-jae Park [Thu, 26 Jul 2012 09:33:13 +0000 (18:33 +0900)]
Code for launching the web-slave
Freeze/That the ping timer when in the idle state.
Change-Id: I09324cb9b084f627eff8a6d1034b4524b227b9a3
Sung-jae Park [Thu, 26 Jul 2012 02:04:37 +0000 (11:04 +0900)]
Use the 10 secs for timeout handling
Change-Id: I214ef34916cf6c7f99521fd035852ef7524efa67
Sung-jae Park [Wed, 25 Jul 2012 15:23:54 +0000 (00:23 +0900)]
Mouse Enter/Leave handler added
Change-Id: I48b64c00fc3d63e3802df28ec93f994e12a9fa9f
Sung-jae Park [Wed, 25 Jul 2012 14:43:49 +0000 (23:43 +0900)]
Update cluster/sub-cluster table
Change-Id: I36f0637919942a9d7c4623a13e691c993519dc09
Sung-jae Park [Wed, 25 Jul 2012 08:46:56 +0000 (17:46 +0900)]
Remove thread safety code for ctx engine.
ctx engine is not uses thread anymore.
Change-Id: Icbf5a8e89bd211deded6e3dad940ac67f3e66a05
Sung-jae Park [Wed, 25 Jul 2012 07:24:00 +0000 (16:24 +0900)]
Add log to figure out the event blocking bug
Change-Id: I09c57124639c837c77b832c315152bafc9ea05cc
Sung-jae Park [Wed, 25 Jul 2012 05:31:23 +0000 (14:31 +0900)]
Clear subscribed list of the client when unref it
Change-Id: I740e198a7a55c84a6987766ed3ce0d99913a9782
Sung-jae Park [Tue, 24 Jul 2012 11:05:17 +0000 (20:05 +0900)]
Flush the memory (for pause)
Change-Id: I1eb4b78770feba364c15418fb78071457e82263e
Sung-jae Park [Tue, 24 Jul 2012 10:37:41 +0000 (19:37 +0900)]
Update pkgmgr module
Change-Id: I506acf3da76c6bbef5cc2bd4b721db1733ea2a26
Sung-jae Park [Tue, 24 Jul 2012 04:22:57 +0000 (13:22 +0900)]
Update pkgmgr code
Apply commented code
Change-Id: Ie61b2e743106bee2912519b87c069930a4253c22
Sung-jae Park [Tue, 24 Jul 2012 01:50:13 +0000 (10:50 +0900)]
Add pkgmgr (dummy)
Change-Id: I43e9a549f910b1d457f7c95a30d6b59f7caa97f9
Sung-jae Park [Tue, 24 Jul 2012 01:49:12 +0000 (10:49 +0900)]
PM state for checking the pause/resume state
Change-Id: I117d52c94e1f2e4d11655e6cdfb8f5b02737cf1c
Sung-jae Park [Mon, 23 Jul 2012 15:12:44 +0000 (00:12 +0900)]
Even if the lockscreen is crashed(or disappeared),
Do not resume the slaves if the LCD is off.
Change-Id: I9ed010d5a0df4243c725f092f6fd9014c133ec57
Sung-jae Park [Mon, 23 Jul 2012 14:11:32 +0000 (23:11 +0900)]
While the instance is changing its state,
Don't try to recover it.
Change-Id: Id93d2964700811a3028c7e554a77cf1c11ec1fa7
Sung-jae Park [Mon, 23 Jul 2012 12:31:21 +0000 (21:31 +0900)]
Update provider buffer management code.
For the web-slave. (which are handling the canvas buffer directly)
Change-Id: I19391b12d5c094382c5c3a78aad3e6a88d178663
Sung-jae Park [Mon, 23 Jul 2012 08:10:33 +0000 (17:10 +0900)]
Add -g option for generating stripped binary
Add "timeout" parameter(default 0u)
Change-Id: I06f4f41ef8571de9fae87913b632991fde46644e
Sung-jae Park [Sat, 21 Jul 2012 08:32:37 +0000 (17:32 +0900)]
Remove scaling code
Use the fixed size for livebox developer.
Scaling factor of the livebox contents should be cared.
Change-Id: I46a423b90ec55141fec7bdf9f28da2eaac372a7c
Sung-jae Park [Fri, 20 Jul 2012 10:31:09 +0000 (19:31 +0900)]
Add the SIGTERM handler for handling the power off
To handle it more more safely.
Change-Id: I98d35556eee7d8a5e9651aad64151a83404583d7
Sung-jae Park [Fri, 20 Jul 2012 07:33:37 +0000 (16:33 +0900)]
The owner of an instance can see it
Change-Id: Ibfbf7d396924de27682c61a22f142c239db054c5
Sung-jae Park [Fri, 20 Jul 2012 01:00:14 +0000 (10:00 +0900)]
Strip the binary and add feature (refresh API)
Released binary includes needless section for launching.
This patch will optimize the size of binary.
This patch also implements some livebox refresh API
Change-Id: I4e46b5d214a16a5f6925cdd3bb92a2ead553b2f3
Sung-jae Park [Thu, 19 Jul 2012 02:23:09 +0000 (11:23 +0900)]
Use the "default" for default content
This should be changed.
Delete cluster/category feature is added but those are not fully operational.
Change-Id: Ia7644fe4717f6f4e73d08f6e962c4dfbc6b32722
Sung-jae Park [Tue, 17 Jul 2012 08:50:53 +0000 (17:50 +0900)]
Update TAG
Sung-jae Park [Mon, 16 Jul 2012 12:53:56 +0000 (21:53 +0900)]
Updated for subscribe/unsubscribe events.
Change-Id: Ic20faae5a02d8169c368d51b12880a7aa7c1d4cd
Sung-jae Park [Mon, 16 Jul 2012 05:20:15 +0000 (14:20 +0900)]
Handling the power off state for daemon.
And use the default size if the buffer type livebox has no size.
Change-Id: I97de2610f9c27c14b311d41c9fcb7144726cbce5
Sung-jae Park [Fri, 13 Jul 2012 15:45:28 +0000 (00:45 +0900)]
Pinup implemented
Change-Id: I7afbc219c52c48ad9c8bee284d04c5142fb2165a
Sung-jae Park [Fri, 13 Jul 2012 13:03:15 +0000 (22:03 +0900)]
Subscribe cluster/sub-cluster feature is added
Change-Id: I8f81f746622970fad9d41b181c6a8b0cc1340f2a
Sung-jae Park [Fri, 13 Jul 2012 12:06:44 +0000 (21:06 +0900)]
Find proper slave for specified ABI.
And add missing file (buffer_handler.c)
Now the slave_activate function will launch the
proper slave for given ABI.
if the "abi" is "HTML", the master will launch the web slave.
or if the "abi" is "OSP", the master will launch the osp slave.
if the "abi" is "c" or "cpp" the master will launch the native slave.
Change-Id: If87a905d1de71f45e4f8de8a78e320abdb358a4c
Sung-jae Park [Fri, 13 Jul 2012 08:36:21 +0000 (17:36 +0900)]
Add instance state check code in the render cb.
For fixing the crash (occured from libevas.so.1 XXX)
Change-Id: I9902d3ba89f92204e154cf476c9bf3667c639d70
Sung-jae Park [Wed, 11 Jul 2012 15:23:26 +0000 (00:23 +0900)]
Acquire buffer/Release buffer implemented.
Slave can acquire the buffer of livebox and PD only if its type is
buffer_livebox, buffer_pd or the slave couldn't get any buffer of livebox or PD.
To support this feature,
livebox should use the "buffer_livebox, buffer_pd" option in its conf
file.
then the master will prepare the buffer and share it with the slave.
Change-Id: Id08a86184f668af9de3842cd8edce655f1df6099
Sung-jae Park [Mon, 9 Jul 2012 11:01:16 +0000 (20:01 +0900)]
Fix the crash (Evas *e == NULL)
Change-Id: I091c7f52dd7283fe1af847a845fc13be152fe83d
Sung-jae Park [Sat, 7 Jul 2012 12:06:07 +0000 (21:06 +0900)]
Fix the duplicated "created" event sending code.
When a new client is connected, the master will send
created event for all instances.
But this, patch will not send "INIT" state instance's created event.
because it will be sent after activated.
Change-Id: I9f45f6b21eaaf0754d0fa2717df8a89a0d3f1b00
Sung-jae Park [Fri, 6 Jul 2012 11:37:55 +0000 (20:37 +0900)]
Fix the crash bug
Update revision
Change-Id: I381b46e2be391cb09a8331ad6236e34b5000adbb
Sung-jae Park [Fri, 6 Jul 2012 11:17:23 +0000 (20:17 +0900)]
Update the function prototype.
Service handler's packet should not be editable.
So I added "const" type qualifier.
Change-Id: Ic6825f1525cf16817744650fb46c13312dec0487
Sung-jae Park [Thu, 5 Jul 2012 09:49:40 +0000 (18:49 +0900)]
Update return value.
Request for faulted package returns EFAULT instead of EINVAL
Change-Id: I07ab2e8f1cbe1ace16a653152baa09ba472645ac
Sung-jae Park [Thu, 5 Jul 2012 08:48:33 +0000 (17:48 +0900)]
Update for heap monitor.
Just increase the PING/PONG time for slave to 240 secs.
Each slave has to ping to master every 120 secs period.
(ANR)
Change-Id: If939501d5dfcab602acf918d29c90ba5f5fb2c53
Sung-jae Park [Tue, 3 Jul 2012 02:38:54 +0000 (11:38 +0900)]
Add some flags
Change-Id: I59c0229a5a8a4437db408903f12cad0d59f45b44
Sung-jae Park [Sun, 1 Jul 2012 05:30:08 +0000 (14:30 +0900)]
Check memory leak
Please check this URL.
http://git.futuremobile.net/?p=heap-monitor&a=summary
The heap monitor is now available.
But it is not stablized yet.
(Not yet published, has no public git repository)
Differences between small(written by the SE team) tool
1. Faster than the "small(SE)".
This heap monitor is aimed to be used by the running process.
So this library should be optimized as possible as to reduce the
performance penalty.
But the small is aimed to be check the memory leak as a tool.
It is a big difference.
2. This is only usable by GLIBC and ELF(not EFL) based system.
This library crawling the ELF header(Program header) to find the
memory region of address location.
3. This library has false alaram.
So we cannot trust all statistics fully.
It depends on the value in the stack.
So if the stack contains some special values, the module detection
can go wrong.
Change-Id: I95c13bd2410420b055f4b603afa9475d4b91d7d6
Sung-jae Park [Wed, 27 Jun 2012 04:18:59 +0000 (13:18 +0900)]
Remove verbose log message
Add SHM type FB management code (slave is not preapred yet)
Replace libconnector with libcore-com
Fix some bug of instance management code.
Try to send a packet again to a slave if it fails but the slave is not faulted.
Fix the com-core handle manage code (Reset it if fails to communicate w/t other)
Change-Id: I43620520329f158aab620e5fe071fe52d9812ae3
Sung-jae Park [Thu, 21 Jun 2012 09:20:17 +0000 (18:20 +0900)]
Fix the pause/resume bug
Add log for pause/resume state check
Change-Id: I1635ff446876e0c20e0d26fbd6c2adb57cca490a
Sung-jae Park [Thu, 21 Jun 2012 04:59:59 +0000 (13:59 +0900)]
Optimize the instance state.
Remove deactivate, request_to_deactivate state.
Just right after the creating a new instance,
its state would be "INST_INIT" and then activate it.
so the created instance could not be hold the INST_INIT.
Because it will be change to INST_REQUEST_TO_ACTIVATED.
Instead of deactivating state,
Just provider will reset the instance state.
Then the state of instance would be "INST_INIT" again.
Change-Id: Ia3ca248d221f35e4c673ca491bc5a8f254bc7809
Sung-jae Park [Tue, 19 Jun 2012 09:30:35 +0000 (18:30 +0900)]
Fix the bug of util_basename
Change-Id: I801039bdf52b2b56811fa5e5d718c5f555536f5b
Sung-jae Park [Tue, 19 Jun 2012 02:40:25 +0000 (11:40 +0900)]
Remove verbose log
Change-Id: I503e5ffb06c67acc186ba54264ac3964aa3e20a8
Sung-jae Park [Mon, 18 Jun 2012 07:16:12 +0000 (16:16 +0900)]
Remove connector implementations.
It is replaced with the libconnector library
Change-Id: I977f996282417bd100a5f5fcb3364bee4860c132
Sung-jae Park [Mon, 18 Jun 2012 06:25:46 +0000 (15:25 +0900)]
Maximum restarting counter is added.
Only the data provider can be re-launched 5 times.
Change-Id: Ie92c9aa8a3a28e253833d7e9c69b037673566ee8
Sung-jae Park [Thu, 14 Jun 2012 14:20:14 +0000 (23:20 +0900)]
Replace dbus with socket (connector)
And there is more minor patche is applied.
------------------------------------------
Keep the last state of the Buffer
When the slave is deactivated,
all instance on that slave will be destroyed.
And reactivating normal instances again.
While executing this procedure,
master will deactivate the instance then its buffers are destroyed.
But after destroying all buffers, it doesn't re-create after
re-activated the instances.
This patch fix it.
Change-Id: I36367a8c6ce11ad597b899d0841b73e52f2385a2
Sung-jae Park [Thu, 14 Jun 2012 13:04:16 +0000 (22:04 +0900)]
Change object reloading code of resizing canvas
Change-Id: I31c5b57f674dec2b185a14655cde472e266dfcf1
Sung-jae Park [Thu, 14 Jun 2012 09:44:01 +0000 (18:44 +0900)]
Make sure always redner the canvas to flush it.
Change-Id: Ie758807fdf3978ae0d09fb7a5e4e21dd95849435
Sung-jae Park [Tue, 12 Jun 2012 23:23:21 +0000 (08:23 +0900)]
Update for thread.
context engine uses thread for handling the context change event.
So I should keep data in safe even if multiple threads are enabled.
This patch keeps he critical section in safe using pthread_mutex.
Change-Id: I95f0f93805fcae302a751d2fb6b03acd6b5ea94d
Sung-jae Park [Tue, 12 Jun 2012 10:14:33 +0000 (19:14 +0900)]
Revise the dbus code and fix some bugs.
Change-Id: I02c97d0e7c4968736c7738f29ec13bdbb759af2b
Sung-jae Park [Tue, 12 Jun 2012 01:34:36 +0000 (10:34 +0900)]
Reset the Proxy if it fails to keep connection.
Change-Id: Iff4f45bf3e099925d3d1fdec8502fb6b8c713bc9
Sung-jae Park [Mon, 11 Jun 2012 11:00:28 +0000 (20:00 +0900)]
Write comments for briefing the state of instances.
Change-Id: I24c0a07372d50cae5b7328e6c6f2f90d9ff738ec
Sung-jae Park [Mon, 11 Jun 2012 07:05:57 +0000 (16:05 +0900)]
Update the state management code for slave.
Pause/Resume state is updated
Resource leaks are resolved.
Change-Id: I02f13ac1c3ab8f076d21bbb5b0d0947053f5d124
Sung-jae Park [Mon, 11 Jun 2012 06:09:25 +0000 (15:09 +0900)]
SMACK Phase 1 - Update spec, manifest files
Change-Id: Ib3e675d63f76f158bddfd2395fcc51d7174c79ec
Sung-jae Park [Sun, 10 Jun 2012 08:38:27 +0000 (17:38 +0900)]
Revise the states of an instance.
Change-Id: Id592fded29566a811f5a19a1a1cbfc4d8d368d26
Sung-jae Park [Sat, 9 Jun 2012 12:18:45 +0000 (21:18 +0900)]
Fix the mouse down/move/up bug.
Before send the down event, send the move event first.
Change-Id: I3556739eab29b259f49f1095da39de53589b6607
Sung-jae Park [Sat, 9 Jun 2012 09:48:03 +0000 (18:48 +0900)]
Fix the buffer type LB's size hanling bug
When data master receives "updated" event,
it reloads livebox buffer again even if it has same size.
This patch will check the size of buffer
and if it is not changed, skip the buffer resizing procedures.
Change-Id: Id4119ce1d91168e40fc5c4eb11d158d3309a27f6
Sung-jae Park [Sat, 9 Jun 2012 07:26:50 +0000 (16:26 +0900)]
Send deleted evt to clients for destroyed livebox
Change-Id: I7247ee6ee0bf760ac1b37218ef086cc87eba2b07
Sung-jae Park [Fri, 8 Jun 2012 15:14:44 +0000 (00:14 +0900)]
Revise the instance state management code.
Change-Id: If5d7769d0d0504be861058dbc735fd67bb9f02f5
Sung-jae Park [Fri, 8 Jun 2012 13:55:55 +0000 (22:55 +0900)]
Cares the exceptional case.
Change-Id: Ia088c5f74663ba160c2de63bedf9564aac8768a2
Sung-jae Park [Fri, 8 Jun 2012 13:42:13 +0000 (22:42 +0900)]
Fix the bug of mismatched buffer size.
In case of the buffer type livebox is not handles the size of the
buffer correctly.
This patch will fix the mismatched buffer size problem.
Change-Id: I7c191434bd32d559c286cd37728530a947b3c5a8
Sung-jae Park [Fri, 8 Jun 2012 13:16:32 +0000 (22:16 +0900)]
Prevent from duplicated initialzation.
After successfully initialize the master,
Someone invokes aul_launch again.
Then the app_create will be called again,
in this case, we have not to do initialization again.
This patch will fix it.
Change-Id: Iec3d7dbe8ec5365363aee002414af3eba1c1860b
Sung-jae Park [Fri, 8 Jun 2012 12:35:30 +0000 (21:35 +0900)]
Fix the crash bug.
While the master tries to activate instances of liveboxes
The slave is crashed without dead callback invocation (via AUL).
(or timing issue that the provider receives connection failed event first)
In this case,
activate_cb function of instance will be called with "result = NULL".
Then the activate_cb function will try to delete it. (just decrease the
refcnt)
But if the instance is referenced by the other functions,
(such as other requests resize, clicked, etc ...)
The refcnt will not reached to the ZERO even if the activate_cb was
called, so the instance will be kept in alive.
Now the problem is raised.
After executing above routines,
The data provider will forcely terminates the slave to recover this unexpected problem.
(because of the timing issue (or AUL missed send dead event).
--> Data provider just think that the current slave is not terminated but it
couldn't managed request correctly so it will terminates the problem
slave, to recover this problem, it will terminates the slave forcely.
Then the slave deactivate callback will be called and tries to delete
the instances which are recorded as a fault package,
because of the instance refcnt is not reached to ZERO, slave deactivate
callback will destroy thes instances again even if it is already cared
by previous routines.
After processing this routines again, the refcnt will be tained.
So who really references this instance will not be guaranted about this instance.
So, to fix this, I just separate the case of switch statement in the
slave_deactivated_cb.
I prevent decreasing when the state of an instance is "INST_DESTROYED".
(INST_DESTROYED means already someone cares this instance,
so don't try to destroy again from any other stages)
Change-Id: I7b41bcb69ecf8f0f0a82536ab9af57faf0d3b42a
Sung-jae Park [Fri, 8 Jun 2012 11:23:28 +0000 (20:23 +0900)]
Build the cluster tree for every packages at the first time.
And remove the data provider launch script from
booting scripts
Change-Id: Iea738d1028d0396339d4dae2b82859019e527d64
Sung-jae Park [Fri, 8 Jun 2012 03:10:56 +0000 (12:10 +0900)]
Handling the exceptional case for aul_terminate_pid
Add "const" to simple functions (setter/getter)
Change-Id: If764ad5f89435850957d8741e705b24840a30431
Sung-jae Park [Tue, 5 Jun 2012 07:09:02 +0000 (16:09 +0900)]
Revise the codes.
Reduce the verbose debugging message.
I may as well separate implementation of package manager to
package.c and instance.c
So hard to understand the code that if the size of a
implementation file is too big.
And it also make hard to manage it.
Fix the bugs
Change-Id: Iccc11895837635b22a2f39c2d0305270ff4079ee
Sung-jae Park [Tue, 5 Jun 2012 04:51:38 +0000 (13:51 +0900)]
Make the delayed ctx connection
Change-Id: I81473c38fb78d3ce8dd67161b4c9ef64be914ecb
Sung-jae Park [Mon, 4 Jun 2012 09:29:05 +0000 (18:29 +0900)]
Pinup bug is fixed.
Every livebox is pinned up when it is created.
So I fix this bug.
Change-Id: I89a9e757c2abacd1d684859051d0572edc36cceb
Sung-jae Park [Mon, 4 Jun 2012 01:20:35 +0000 (10:20 +0900)]
Counting the fault
Replace "state" with "count".
Update fault management code.
Update the pause/resume check code
Check pause or resume status when activating a slave
Add some log for event handling of ctx engine
Change-Id: Id4a2dc7a8a59d6a8dbae412228b9265ca436ae18
Sung-jae Park [Sat, 2 Jun 2012 17:37:05 +0000 (02:37 +0900)]
Revise the slave_manager.c
Separate it to two files.
slave_life.c
slave_rpc.c
slave_life.c will care the life cycle of the slaves.
slave_rpc.c will care the RPC for slave.
pkgmgr, client_mgr also should be revised.
And more testing required.
Version is updated to 0.2.0
Change-Id: I00582e1f56ce0b5e8d8fb949fab91e6ff5c5f6b3
Sung-jae Park [Fri, 1 Jun 2012 10:12:49 +0000 (19:12 +0900)]
Update slave life cycle monitoring code
especially, for the crash
Change-Id: I367dd3c1bdfef5c6072ed2d9b2d33e6a5bf100c2
Sung-jae Park [Fri, 1 Jun 2012 06:10:01 +0000 (15:10 +0900)]
Check the lock state.
Change-Id: Ia3f9367a87feb6252ef6aa710b611b4c36011b66
Sung-jae Park [Fri, 1 Jun 2012 01:50:58 +0000 (10:50 +0900)]
Enable contex engine
Toggling it by the vconf.
Fix the bug of conf parser
Change-Id: I6b936624db2a708d854653a71dc427611a7f43d9
Sung-jae Park [Thu, 31 May 2012 09:36:12 +0000 (18:36 +0900)]
Recover util_unlink
Update slave process management code.
Change-Id: I4e3cce719676db5218cf45cff5ccbe0454301e03
Sung-jae Park [Wed, 30 May 2012 11:14:41 +0000 (20:14 +0900)]
Update the debugging message
And destroy a slave if it has no loaded packages
Change-Id: I86fb4780b78eb491a5780f04b0b721997ef68320
Sung-jae Park [Tue, 29 May 2012 11:10:12 +0000 (20:10 +0900)]
Fix the bug of lievbox_has_pd function.
master data provider always set the default PD even a livebox has
no information about the PD.
Because of this, the livebox_has_pd function always returns true
(exists)
But this patch will not set default value anymore.
If a livebox didn't set any path and group for PD,
The data provider will not make any buffer for it.
After applying this patch,
the viewer can decide whether a livebox has PD or not.
-
Delete unnecessary function (util_unlink).
Change-Id: I3efe06f06fa23c44e3694ec1d82fdc5ab81a164e
Sung-jae Park [Tue, 29 May 2012 06:19:46 +0000 (15:19 +0900)]
Add comments
Change-Id: I5556710b7fcb91661f0f8a4bd62f07ac5931f397
Sung-jae Park [Sun, 27 May 2012 09:01:17 +0000 (18:01 +0900)]
Review entire codes.
Add error message.
Change-Id: I7b4088a87d935fc5390fa25f5547933882056e1d
Sung-jae Park [Sat, 26 May 2012 02:48:41 +0000 (11:48 +0900)]
Define the minimum update period.
Check the period value.
if the period is in 0.0 < ~ < MINIMUM_PERIOD, set it to MINIMUM_PERIOD
MINIMUM_PERIOD is the minimum value of the period.
and it is defined as "1.0f" now.
Change-Id: I8573005f9167647ec88571380d5445a711c78f96
Sung-jae Park [Thu, 24 May 2012 08:10:13 +0000 (17:10 +0900)]
Supporting the OSP Livebox(CPP)
Change-Id: Ife376427757a2d89f59bf89025db97d3bdf1d632
Sung-jae Park [Wed, 23 May 2012 12:41:26 +0000 (21:41 +0900)]
content_event function is updated.
Give the current pointer location
to the livebox via livebox_content_info function.
Change-Id: I9c8e96a84ac66515f76b9fec7fc4d9c128ddf587
Sung-jae Park [Wed, 23 May 2012 06:08:11 +0000 (15:08 +0900)]
Create share folder
Change-Id: Iad91e8671e1b03ad104161e532f450131a049b55