| Age | Commit message (Collapse) | Author |
|---|
|
BZ: IMINAN-51481
Fixes: 38135184
Test: Netflix, play movies, youtube
Change-Id: I15c497aba636159c67909b6f81083e8dcadc87e4
Signed-off-by: Austin Hu <austin.hu@intel.com>
(cherry picked from commit 86df600763e90207c90b4c4e9a823ad044cc41bd)
|
|
For build-system CFLAGS clean-up, remove unused variables.
Bug: 18632512
Change-Id: I4da2cb255b4bc08c5dfc04d7d31c58e63d90948e
|
|
BZ: 209178
remove the warning in wrs_core
Signed-off-by: ywan171 <yi.a.wang@intel.com>
Change-Id: I5655dcc283325cf7da5e0b088887e63cd24fb056
Signed-off-by: ywan171 <yi.a.wang@intel.com>
Reviewed-on: https://android.intel.com/220500
Reviewed-by: Fourdan, Olivier <olivier.fourdan@intel.com>
Tested-by: Fourdan, Olivier <olivier.fourdan@intel.com>
|
|
BZ: 209178
Change-Id: I07232c12f5466b362530315d33084a4d7ecfabdc
Signed-off-by: Thierry Strudel <thierry.strudel@intel.com>
Reviewed-on: https://android.intel.com/218917
Reviewed-by: Imberton, Guilhem <guilhem.imberton@intel.com>
Tested-by: Imberton, Guilhem <guilhem.imberton@intel.com>
|
|
BZ: 134837
reset the stop flag if the decoder thread restarts from stop state to fix cts MediaCodec
bug.
Change-Id: Ic8f22cb6eee576e4afa61197cbc390834ee7eb1b
Signed-off-by: ywan171 <yi.a.wang@intel.com>
Reviewed-on: http://android.intel.com:8080/130318
Reviewed-by: Zhou, ChuX <chux.zhou@intel.com>
Reviewed-by: Feng, Wei <wei.feng@intel.com>
Reviewed-by: Liang, Dan <dan.liang@intel.com>
Reviewed-by: Chen, Tianmi <tianmi.chen@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: cactus <cactus@intel.com>
Tested-by: cactus <cactus@intel.com>
|
|
BZ: 105774
The hw codec's working thread is spawned from TimedEventQueue
thread and inherits the parent's name. So we cannot distinguish
them for analyzing purpose.
Change-Id: I893874309bc669a1e812085001b04b6fb88c490c
Signed-off-by: Dan Liang <dan.liang@intel.com>
Reviewed-on: http://android.intel.com:8080/106287
Reviewed-by: cactus <cactus@intel.com>
Reviewed-by: Feng, Wei <wei.feng@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: buildbot <buildbot@intel.com>
Tested-by: buildbot <buildbot@intel.com>
|
|
BZ: 91187
Change to use return value of dlsym() instead of dlerror() to check whether
the result of dlsym() is successful or not. It can workaround the Bionic issue
that might cause other issues.
Signed-off-by: Liu Bolun <bolunx.liu@intel.com>
Change-Id: I88dc45c523dabd77cb0598dbd0b66d38b2d4d0e4
Reviewed-on: http://android.intel.com:8080/95452
Reviewed-by: Liu, BolunX <bolunx.liu@intel.com>
Reviewed-by: Yuan, Shengquan <shengquan.yuan@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: buildbot <buildbot@intel.com>
Tested-by: buildbot <buildbot@intel.com>
|
|
video editor carsh
BZ: 75145
1. the dlerror will return the error caused by the main thread, this will lead the
ret equals -1, and this will make the loop not end in CModule destructor;
2. remove the redundant pthread_mutex_unlock;
3. add some error logs to expose some corner cases
Change-Id: Ifab1bb704ae1a65553be1326716558a4e013fada
Signed-off-by: Gu, Wangyi <wangyi.gu@intel.com>
Reviewed-on: http://android.intel.com:8080/90189
Reviewed-by: cactus <cactus@intel.com>
Reviewed-by: Wang, Yi A <yi.a.wang@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: Gozalvez Herrero, Juan AntonioX <juan.antoniox.gozalvez.herrero@intel.com>
Tested-by: Gozalvez Herrero, Juan AntonioX <juan.antoniox.gozalvez.herrero@intel.com>
|
|
BZ: 83848
fix klocwork issue in queue.c
Signed-off-by: ywan171 <yi.a.wang@intel.com>
Change-Id: I57d75c2e3b3c33a03aa7a2c8d922ac0ce178ca4f
Reviewed-on: http://android.intel.com:8080/89313
Reviewed-by: Wang, Elaine <elaine.wang@intel.com>
Reviewed-by: Wang, Yi A <yi.a.wang@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: buildbot <buildbot@intel.com>
Tested-by: buildbot <buildbot@intel.com>
|
|
BZ: 73428
fix Klocwork issue under wrs_omxil_core, check point before refereced
Signed-off-by: ywan171 <yi.a.wang@intel.com>
Change-Id: Ic5b721a5cda4dacb0b89dc28bdf9aaad605e1d5a
Reviewed-on: http://android.intel.com:8080/86803
Reviewed-by: Wang, Yi A <yi.a.wang@intel.com>
Reviewed-by: Feng, Wei <wei.feng@intel.com>
Reviewed-by: Shi, PingX <pingx.shi@intel.com>
Tested-by: Shi, PingX <pingx.shi@intel.com>
Reviewed-by: cactus <cactus@intel.com>
Tested-by: cactus <cactus@intel.com>
|
|
BZ: 69531
The non-initialized dlerr variable from the module_open function
was passed to the module_set_error function. Incorrect dlerr
caused SEGV in the strdup function called from the
module_set_error function.
Change-Id: I7e7b0b517a56968e356460dcc7544860ce84fa91
Signed-off-by: dvshkurk <Dmitry.V.Shkurko@intel.com>
Reviewed-on: http://android.intel.com:8080/80843
Reviewed-by: Shkurko, Dmitry V <dmitry.v.shkurko@intel.com>
Reviewed-by: Ding, Haitao <haitao.ding@intel.com>
Tested-by: Ding, Haitao <haitao.ding@intel.com>
Reviewed-by: cactus <cactus@intel.com>
Tested-by: cactus <cactus@intel.com>
|
|
string to workaround Bionic issue
BZ: 66993
Change to use return value of dlopen() instead of dlerror() to check whether the call to dlopen()
is successful or not, to workaround the synchronization issue of dl_err_str in Bionic.
Signed-off-by: Liu Bolun <bolunx.liu@intel.com>
Change-Id: I200fe18c8ac5841f7aeb29f44ba6d5e6dc2cf100
Reviewed-on: http://android.intel.com:8080/76392
Reviewed-by: Liu, BolunX <bolunx.liu@intel.com>
Reviewed-by: Feng, Wei <wei.feng@intel.com>
Tested-by: Tong, BoX <box.tong@intel.com>
Reviewed-by: Tong, BoX <box.tong@intel.com>
Reviewed-by: cactus <cactus@intel.com>
Tested-by: cactus <cactus@intel.com>
|
|
BZ: 18528
This is the fix for BZ: 18006
Protect access to a status variable in workqueue by a lock,
otherwise may cause endless wait.
Change-Id: Iccb2a19664852af29cd939f2591be44e7ce6eb3c
Signed-off-by: Chang Ying <ying.chang@intel.com>
Reviewed-on: http://android.intel.com:8080/30052
Reviewed-by: Ji, GuoliangX <guoliangx.ji@intel.com>
Reviewed-by: Yuan, Shengquan <shengquan.yuan@intel.com>
Reviewed-by: Ding, Haitao <haitao.ding@intel.com>
Tested-by: Ding, Haitao <haitao.ding@intel.com>
Reviewed-by: buildbot <buildbot@intel.com>
Tested-by: buildbot <buildbot@intel.com>
|
|
Change-Id: Ifb4bf6160fbc3469ede7b258dd72ef8c77c84d62
Signed-off-by: Liu, Shuo <shuo.liu@intel.com>
|
|
This reverts commit f92bee3efc605b340ff8930f630105ce0ffe827e.
|
|
Change-Id: Ib07001c2ae13eb38923c804eb86d0ea615b2bc56
Signed-off-by: Jack Ren <jack.ren@intel.com>
|
|
Change-Id: Ife66ebd3d89657d4fdffc057159a48f992c6fcde
Signed-off-by: $ <bruce.j.beare@intel.com>
|
|
Change-Id: Ia33c9e1c3e21e9f68c867a1b5f8616c0afcd056e
|
|
|
|
h264_parser uses some codes of mixvbp and it causes license issue.
recent OpenCore sets port parameters properly during port negotiation, so
we don't need to parse the sps/pps to get width and height for reconfiguring
ports and it (not parsing codec data) wouldn't cause any problem for now.
|
|
Signed-off-by: sahara <keun-o.park@windriver.com>
|
|
Signed-off-by: sahara <keun-o.park@windriver.com>
|
|
Signed-off-by: sahara <keun-o.park@windriver.com>
|
|
Signed-off-by: sahara <keun-o.park@windriver.com>
|
|
|
|
|
|
|
|
|
|
|
|
It's hard-coded for version 1, layer 3
|
|
|
|
we have following workqueue control functions
int StartWork(bool executing);
executing == true : start workqueue thread with executing state
!= true : start workqueue thread with paused state
void StopWork(void);
join workqueue thread and exit
void PauseWork(void);
pause workqueue thread, it guarantees any workqueue callback's not
running when it returns.
void ResumeWork(void);
resume paused workqueue thread
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
this patch prevent Start() from creating new thread without Join() and
join() from joining a thread not created yet.
|
|
|
|
this function discards qeueud works in works list to cancel queued job;
|
|
it had a race condition against list manipulation while processing Work()
this patch prevents race condition while processing Work()
|
|
|
|
WorkQueu doesn't start work thread implicitly. you should call
Start/StopWork() where you want at
|
|
|
|
it has list, queue, module and etc
|
