Hi All,

As some of the LayerManagerControl commands do not work when same surface
is attached to two layers.

Following command hangs and do not come out as expected. I tried to check
where it is hanging using GDB. Is this gives some hit on issue and
resolution? Please suggest if you have any pointers.

LayerManagerControl get scene

root@linux1:~# gdb -p 916
warning: Unable to find libthread_db matching inferior's thread library,
thread debugging will not be available.
warning: Unable to find libthread_db matching inferior's thread library,
thread debugging will not be available.

0x4ddf68f4 in wl_list_length () from /usr/lib/libwayland-client.so.0
(gdb) bt
#0  0x4ddf68f4 in wl_list_length () from /usr/lib/libwayland-client.so.0
#1  0x4deb45e4 in ilm_getSurfaceIDsOnLayer ()
   from /usr/lib/libilmControl.so.1.11.0
#2  0x00033f80 in printLayerProperties(unsigned int, char const*) ()
#3  0x00035a40 in printScene() ()
#4  0x0001d4e8 in func_2(Expression*) ()
#5  0x000330ac in
ExpressionInterpreter::interpretCommand(std::__cxx11::basic_string<char,
std::char_traits<char>, std::allocator<char> >) ()
#6  0x00018d50 in main ()

Regards,
Vikash

On Tue, Apr 17, 2018 at 10:51 AM, Vikas Patil <[email protected]> wrote:

> Hi Emre,
>
> Could you please suggest on this blocking behavior of LayerManagerControl
> with multi screen/layer?
>
> Thank You.
>
> Best Regards,
> Vikash
>
> On Wed, Apr 11, 2018 at 11:35 AM, Vikas Patil <[email protected]>
> wrote:
>
>> Hi Emre Ucan,
>>
>> Thanks a lot for your quick response. I am able to show same surface on
>> two layers now. I have taken following two commit to weston 1.11.0.
>> Attached here same as  patch to weston 1.11.0.
>>
>> "5e8d55da698e58"
>> "67bd21232fa549"
>>
>> However if I use any of the below commands to analyze then it is not
>> exiting and I need to prress "CTRl+C" to come out from command. Do you know
>> if this is the normal behavior or some fix is available for this ?
>>
>> root@linux-a1 :~# LayerManagerControl analyze surface 10
>> ^C
>>
>> root@ linux-a1 :~# LayerManagerControl get scene
>> screen 0 (0x0)
>> ---------------------------------------
>> - resolution:           x=800, y=480
>> - hardware layer count: 0
>> - layer render order:   1000(0x3e8), 2000(0x7d0),
>>
>>     layer 1000 (0x3e8)
>>     ---------------------------------------
>>     - created by pid:       0
>>     - original size:        x=400, y=480
>>     - destination region:   x=0, y=0, w=400, h=480
>>     - source region:        x=0, y=0, w=400, h=480
>>     - orientation:          0 (up is top)
>>     - opacity:              1
>>     - visibility:           1
>>     - type:                 0 (unknown)
>> ^C
>>
>> root@linux-a1:~# LayerManagerControl get layer 1000
>> layer 1000 (0x3e8)
>> ---------------------------------------
>> - created by pid:       0
>> - original size:        x=400, y=480
>> - destination region:   x=0, y=0, w=400, h=480
>> - source region:        x=0, y=0, w=400, h=480
>> - orientation:          0 (up is top)
>> - opacity:              1
>> - visibility:           1
>> - type:                 0 (unknown)
>> ^C
>>
>> root@orinoco-9939-a1:~# LayerManagerControl get surface 10
>> surface 10 (0xa)
>> ---------------------------------------
>> - created by pid:       821
>> - original size:      x=800, y=480
>> - destination region: x=0, y=0, w=400, h=480
>> - source region:      x=0, y=0, w=800, h=480
>> - orientation:        0 (up is top)
>> - opacity:            1
>> - visibility:         1
>> - pixel format:       0 (R-8)
>> - native surface:     0
>> - counters:           frame=0, draw=0, update=0
>> ^C
>>
>>
>> Also following commands worked successfully.
>>
>>
>> LayerManagerControl get screen 0
>> LayerManagerControl get layer 2000
>> LayerManagerControl get layers
>> LayerManagerControl get surfaces
>>
>> I used following commands to setup  and test
>>
>> export XDG_RUNTIME_DIR=/var/run/root/1000
>>
>> LayerManagerControl create layer 1000 400 480
>> LayerManagerControl set layer 1000 visibility 1
>> LayerManagerControl set layer 1000 destination region 0 0 400 480
>>
>> LayerManagerControl create layer 2000 400 480
>> LayerManagerControl set layer 2000 visibility 1
>> LayerManagerControl set layer 2000 destination region 400 0 400 480
>>
>> LayerManagerControl set screen 0 render order 1000,2000
>>
>> EGLWLMockNavigation &
>> LayerManagerControl add surface 10 to layer 1000
>> LayerManagerControl add surface 10 to layer 2000
>> LayerManagerControl set surface 10 visibility 1
>> LayerManagerControl set surface 10 source region 0 0 800 480
>> LayerManagerControl set surface 10 destination region 0 0 400 480
>>
>> Best Regards,
>> Vikash
>>
>> On Tue, Apr 10, 2018 at 7:43 PM, Ucan, Emre (ADITG/ESB) <
>> [email protected]> wrote:
>>
>>> Hi Vikas,
>>>
>>>
>>>
>>> This patch “5e8d55da698e58”  enabled the feature. It is part of weston
>>> 1.12 release.
>>>
>>>
>>>
>>> Best regards
>>>
>>> *Emre Ucan*
>>> Engineering Software Base (ADITG/ESB)
>>>
>>> Tel. +49 5121 49 6937
>>>
>>> *From:* wayland-devel [mailto:wayland-devel-bounces@
>>> lists.freedesktop.org] *On Behalf Of *Vikas Patil
>>> *Sent:* Dienstag, 10. April 2018 14:58
>>> *To:* [email protected]; Mizuno, Wataru
>>> (ADITJ/SWG); wayland mailing list
>>> *Subject:* Same ilm surface on multiple layer support
>>>
>>>
>>>
>>> +Subject
>>>
>>> Dear All,
>>>
>>> We are facing issue when we are trying to add same surface to multiple
>>> layers. When we try to attach surface to another layer, it is getting
>>> detached from the earlier layer.
>>>
>>> We are using wayland/weston/wayland-ivi-extension 1.11.0 with
>>> drm-backend on TI's Soc.
>>>
>>> Could anyone know if this is the limitation of ILM 1.11.0 ? Is this
>>> fixed in newer version and can it be ported to 1.11.0 ? or Is there any
>>> other way to show same surface on multiple layers?
>>>
>>> I see it was the limitation with wayland-ivi-extesnion 1.9.0 as below
>>> [1].
>>>
>>>
>>>
>>> *"Currently 1 layer can be only on 1 screen, and 1 surface can be only on 1 
>>> layer, we are planning to relax this limitation And allow 1 surface to be 
>>> on many layers but we would need to break the ABI and change the  
>>> ivi-controller protocol."*
>>>
>>> [1] https://lists.genivi.org/pipermail/genivi-ivi-layer-manageme
>>> nt/2016-October/005416.html
>>>
>>>
>>>
>>> Thanking you in advance.
>>>
>>>
>>>
>>> Best Regards,
>>>
>>> Vikash
>>>
>>
>>
>
_______________________________________________
wayland-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/wayland-devel

Reply via email to