Hi Ken,

It would be helpful to know more about your specific run-time scenario and DFB 
environment, since we were unable to reproduce the difficulty you have 
encountered.  For example, what app is running as the master process and what 
parameters were used, and are there any other slave processes running.  Also 
information about the directfbrc configuration used and the window manager in 
use.  A log from the master process would identify some of these details.

We are able to successfully run multiple instances of the multiAppTestTool on 
multiple display layers, as well as on the same display layer.  Here are three 
example valid usages we ran:

1.
    # Master process rendering on display layer UPP A with exclusive access
    ./multiAppTestTool -r=2 -l=A -t=20

    # Slave process rendering on UPP B with exclusive access (launched 
    # shortly after the master of course)
    ./multiAppTestTool -r=2 -l=B -t=10

2.
    # Master process rendering on UPP B with shared access in a window
    ./multiAppTestTool -r=2 -l=Bw -t=20

    # Slave process also rendering on UPP B with shared access in a window
    ./multiAppTestTool -r=2 -l=Bw -t=10

3.
    # Master process rendering on UPP A with shared access in a window
    ./multiAppTestTool -r=2 -l=Aw -t=30

    # Slave process also rendering on UPP A with shared access in a window
    ./multiAppTestTool -r=2 -l=Aw -t=10

    # Slave process rendering on UPP C with exclusive access
    ./multiAppTestTool -r=2 -l=C -t=10

    # Slave 2 process rendering on UPP D with exclusive access
    ./multiAppTestTool -r=2 -l=D -t=10

    # Slave 3 process rendering on UPP B with shared access in a window
    ./multiAppTestTool -r=2 -l=Bw -t=10

    # Slave 4 process also rendering on UPP B with shared access in a window
    # (all slaves must be launched in quick succession of course due to 
    # the timeouts specified on the master)
    ./multiAppTestTool -r=2 -l=Bw -t=10 

Note:  We tested with a more recent version of the DFB driver set than you are 
using, however these should work on the version you are working with.

BTW:  You probably figured this out already, but the multiAppTestTool sample is 
a custom tool included with the Intel® CE Media Processors SDK and tailored for 
our hardware (related to the display layers available in current DFB driver 
set).  As such, please feel free to contact me directly to continue to work 
this issue (benjamin.r.def...@intel.com)

Thanks,
Ben

-----Original Message-----
From: directfb-dev-boun...@directfb.org 
[mailto:directfb-dev-boun...@directfb.org] On Behalf Of Kenneth Rogers -X 
(kenroger - Apex Systems, Inc. at Cisco)
Sent: Wednesday, September 29, 2010 7:36 AM
To: Lionel Landwerlin
Cc: directfb-dev@directfb.org
Subject: Re: [directfb-dev] GetDisplayLayer Not Supported?

-r is which test to run.
-r=X,--run=X     The test to run. Square=0, Triangle=1, Circle=2, Daemon=3

Kenneth R.

-----Original Message-----
From: Lionel Landwerlin [mailto:llandwer...@gmail.com] 
Sent: Wednesday, September 29, 2010 10:26 AM
To: Kenneth Rogers -X (kenroger - Apex Systems, Inc. at Cisco)
Cc: directfb-dev@directfb.org
Subject: Re: [directfb-dev] GetDisplayLayer Not Supported?

I guess you will have to look at the Intel DirectFB driver sources to
find out why it says this operation is unsupported.

What's the meaning of the -r=2 option ?

On Wed, Sep 29, 2010 at 4:13 PM, Kenneth Rogers -X (kenroger - Apex
Systems, Inc. at Cisco) <kenro...@cisco.com> wrote:
> It actually wants the letters,
>
> multiAppTestTool -r=2 -l=1
> Unrecognized value:1: Expected a single letter A-D
>
> I guess I could hack the source to make it take a number, but using -l=C 
> actually works, and C is the default value, according to the -h.
> -l=X,--layer=X   The layer to use. Valid Values:A-D Default=C
>
> Kenneth R.
>
> -----Original Message-----
> From: Lionel Landwerlin [mailto:llandwer...@gmail.com]
> Sent: Wednesday, September 29, 2010 10:05 AM
> To: Kenneth Rogers -X (kenroger - Apex Systems, Inc. at Cisco)
> Cc: directfb-dev@directfb.org
> Subject: Re: [directfb-dev] GetDisplayLayer Not Supported?
>
> Hi Kenneth,
>
> The DirectFB layers are numeroted from 0 to nb_layer, not with a
> letter. B refers to the Intel's GDL interface.
> Could you try with layer 1 as showed by dfbinfo ?
>
> Regards,
>
> --
> Lionel Landwerlin
>
> On Wed, Sep 29, 2010 at 3:53 PM, Kenneth Rogers -X (kenroger - Apex
> Systems, Inc. at Cisco) <kenro...@cisco.com> wrote:
>> I'm trying to learn how to make two apps that use DirectFB display on
>> screen at the same time, so I was experimenting with the sample code for
>> multiAppTestTool, But if I start it with a layer that's not 'C' it
>> fails.  Here's the output:
>>
>>
>> # multiAppTestTool -r=2 -l=B
>> Starting. PID:3772
>>
>>   ~~~~~~~~~~~~~~~~~~~~~~~~~| DirectFB 1.4.3++
>> |~~~~~~~~~~~~~~~~~~~~~~~~~
>>        (c) 2007-2010  Intel Corporation
>>        (c) 2001-2009  The world wide DirectFB Open Source Community
>>        (c) 2000-2004  Convergence (integrated media) GmbH
>>      ----------------------------------------------------------------
>>
>> (*) DirectFB/Core: Multi Application Core. (2010-09-21 21:49)
>> (*) Fusion/SHM: Using MADV_REMOVE (2.6.28.0 >= 2.6.19.2)
>> (*) Direct/Thread: Started 'Fusion Dispatch' (3773) [MESSAGING
>> OTHER/OTHER 0/0] <8388608>...
>> (*) IntelCE/Systems: Available GDL driver video memory: 51MB of 70MB
>> (*) IntelCE/Systems: Default video memory caching policy: Uncached
>> (*) IntelCE/Systems: Slave Process
>> (*) DirectFB/Genefx: MMX detected and enabled
>> (*) IntelCE/GfxDriver: Using OpenGL ES 1.1 graphics library
>> (*) IntelCE/GfxDriver: Using SRB graphics library
>> (*) IntelCE/GfxDriver: Default graphics library: SRB
>> (*) DirectFB/Graphics: Intel Corporation Intel_CE Graphics Driver  2.0
>> (Intel Corporation)
>> (*) Direct/Interface: Loaded 'FT2' implementation of 'IDirectFBFont'.
>> (!) DirectFB/core/layers: Could not set primary region config!
>>    --> Not supported!
>> multiAppTestTool.c <376>:
>>        (#) DirectFBError [dfb->GetDisplayLayer(dfb, gDisplayLayerId,
>> &pLayer)]: Not supported!
>>  (!!!)  *** WARNING [Application exited without deinitialization of
>> DirectFB!] *** [core.c:874 in dfb_core_deinit_check()]
>> (!) Direct/Thread: Canceling 'Fusion Dispatch' (3773)!
>>
>>
>>
>> dfbinfo shows that B is a layer, so I'm not sure what's wrong.
>>     Layer (00) DLID_UPP_C                      (primary layer)
>>        Type:    GRAPHICS VIDEO STILL_PICTURE
>>        Caps:    SURFACE OPACITY ALPHACHANNEL SRC_COLORKEY PREMULTIPLIED
>> SCREEN_POSITION
>>
>>     Layer (01) DLID_UPP_B
>>        Type:    GRAPHICS VIDEO STILL_PICTURE
>>        Caps:    SURFACE OPACITY ALPHACHANNEL SRC_COLORKEY PREMULTIPLIED
>> SCREEN_POSITION
>>
>>
>> Maybe I'm going about this whole process the wrong way? What is the
>> recommended way to have two applications be able to share the screen?
>>
>> Thanks,
>> Ken Rogers
>> _______________________________________________
>> directfb-dev mailing list
>> directfb-dev@directfb.org
>> http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
>>
>
_______________________________________________
directfb-dev mailing list
directfb-dev@directfb.org
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
_______________________________________________
directfb-dev mailing list
directfb-dev@directfb.org
http://mail.directfb.org/cgi-bin/mailman/listinfo/directfb-dev
  • [directfb-dev] ... Kenneth Rogers -X (kenroger - Apex Systems, Inc. at Cisco)
    • Re: [direc... Lionel Landwerlin
      • Re: [d... Kenneth Rogers -X (kenroger - Apex Systems, Inc. at Cisco)
        • Re... Lionel Landwerlin
          • ... Kenneth Rogers -X (kenroger - Apex Systems, Inc. at Cisco)
            • ... Defnet, Benjamin R

Reply via email to