Feed for discussion Advanced Graphics in project Graphics.

post122721: Graphics support for Beaglebone black(AM335x)

2026-05-14T08:03:08Z

I am using QNX6.6 with Graphics patch 3875 on Beaglebone black & LCD/TFT display interfaced through HDMI interface . I have been using a 7" inch display with a resolution of 1024X 600, which has been working fine. I got a new LCD/TFT display with a resolution of 720x1280. I am unable to configure it with the same combination. Graphics.conf doesn't accept 720x1280 but it accepts 1280x720. If I specify it, the screen comes up but with a small black band and mismatch of coordinates, hence unable to run any proper display application. No other video modes are working. How to make the this display work with BBB?

post122392: Qnx7.1 Screen splash screen

2023-10-16T07:52:31Z

Hi all, there is a quick way to load a splash screen in QNX7.1 ?

Or is a demo source available using screen?

Thanks,

Mario

post121656: Build OpenCV c/c++ application in QNX SDP 7.1

2021-11-11T12:42:32Z

I am newly started working on QNX environment. I want to build a sample c/c++ QNX OpenCV application to run on “aarch64le”. I am using Jacinto 7 board in which QNX is running on a72 core and QNX SDP Version 7.1. Please help me to build sample opencv application to run on QNX.

post121655: Sample OpenCV application in QNX

2021-11-11T12:40:59Z

Hi Team,

I am newly started working on QNX environment. I want to build a sample c/c++ QNX OpenCV application to run on “aarch64le”. I am using Jacinto 7 board in which QNX is running on a72 core and QNX SDP Version 7.1. Please help me to build sample opencv application to run on QNX.

Warm Regards,

Yatheesha S

________________________________

DISCLAIMER: This e-mail contains PRIVILEGED AND CONFIDENTIAL INFORMATION intended solely for the use of the addressee(s). If you are not the intended recipient, please notify the sender by e-mail and delete the original message. Opinions, conclusions and other information in this transmission that do not relate to the official business of QuEST Global and/or its subsidiaries, shall be understood as neither given nor endorsed by it. Any statements made herein that are tantamount to contractual obligations, promises, claims or commitments shall not be binding on the Company unless followed by written confirmation by an authorized signatory of the Company.

________________________________

post121630: QNX VM guest with Vmware graphics drivers on QEMU

2021-10-22T08:37:47Z

I am not sure if this is the right forum for this query. I am working on getting QNX neutrino 7.0 with vmware graphics working on QEMU. I did bit of reading and understood that QNX doesnt support this. Anyone here has any past experience with it. Any leads would be helpful. Thnx

post121574: Graphics driver for Intel Celeron Quad Core J1900 ?

2021-08-31T09:29:23Z

Hi

I've installed QNX 6.5 on Advantech UNO-2272G-J2AE with Intel Celeron Quad Core J1900 (Graphic Engine: Intel HD Graphic).

What graphics driver should I use? I've tried devg-intelhd.so and devg-i830.so but they don't work. Only vesabios and svga are working.

Best regards

Waldi

post121550: how to tell when screen mode change is applied

2021-08-02T16:15:39Z

I am using the SCREEN_PROPERTY_MODE property (screen_get_display_property_pv) to set/change the graphics mode. These mode changes have an impact on graphics output and display synchronization, so it is important to know when the mode change has been fully applied. However, the method for doing this is not clear.

Mode change time seems to vary drastically, taking 250ms one time and over 1.5 seconds the next. Mode change time can be seen from monitoring screen slog information as well as looking at the display.

I read the mode properly after setting (which flushes the command), but it returns the new mode even though that mode is not yet applied. I tried an explicit flush command, but that had no impact. I also tried "wait_vsync", to no effect.

post119804: libimg how to encode png to memory

2019-08-13T07:06:39Z

Hi

    In my case, I want to encode a pixel buffer into a piece of memory in png format.I just write the code:

BYTE* pbtBuffer = new BYTE[img.w * img.h];

s_writeBufferSize         = img.w * img.h;

io_stream_t* pOutStream = io_open( IO_MEM, IO_WRITE, img.w * img.h, pbtBuffer );

if ( !pOutStream )

{

CLogApp::info( String::format( L"io_open error: %d", errno ) );

return FALSE;

}

pOutStream->output.write_f = ioMemoryWrite;

But I cannot  do io_open( IO_MEM...) with IO_WRITE.  I search the history and find that qnx not support the write mode with io_mem.

How can I do this. Can someone help me? Thanks!!!!!!!!!!!!

post119693: QNX 7.0 screen configuration for Intel Skylake

2019-05-06T06:48:11Z

Hello,

I am using the generic Intel x86_64 BSP on a Dell computer with an Intel i3 Skylake processor with standard integrated graphics.

Will Screen run on such a machine and is there a Screen configuration available that I can try?

The BSP that I am using appears to provide a screen configuration for the Intel NUC only.

Thanks,

John

post119447: Re: calib-touch difficult to work and not accurate

2019-01-24T11:01:46Z

never mind,

use legacy driver "devi" instead of "hid" solve the problem.

Mike

post119434: Re: calib-touch difficult to work and not accurate

2019-01-22T13:35:29Z

another issue,

after the first unsuccessful "accept"ed calibration, I tried to do recalibration with

calib-touch -zorder=20 -config-file=/tmp/calib.localhost -recalibrate

this time stuck at Figure 1 of http://www.qnx.com/developers/docs/6.6.0_anm11_wf10/#com.qnx.doc.neutrino.utilities/topic/c/calib-touch.html

i.e., the target never moves no matter touch any place on the whole screen.

It seems, the previous calibrated data were still in use, and blocked the recalibration.

Please advise, thanks in advance.

Mike

post119431: calib-touch difficult to work and not accurate

2019-01-22T12:30:57Z

Hi QNX,

       qnx sdp 6.6, qnx sdk for apps and media 1.1 release,

       BSP_ti-am335x-beaglebone_br-660_be-660_SVN797070_JBN574.zip

       beaglebonedemo.ext.anm11.258.150325_2017

       touchscreen 800x480

I am working on touchscreen with beaglebone black on qnx 6.6.

Touchscreen starts with calib-touch 3-point calibration. However it stuck at the last step as described in Figure 4,

http://www.qnx.com/developers/docs/6.6.0_anm11_wf10/#com.qnx.doc.neutrino.utilities/topic/c/calib-touch.html

exactly same situation as previous post, "calib-touch program doesn't accept or recognize that."

http://community.qnx.com/sf/discussion/do/listPosts/projects.graphics/discussion.advanced_graphics.topc25633

After many times tries, calib-touch finally "accept" and save the data to calib.localhost. However, the calibration is not accurately at all. You have to estimate and try to figure out exact location, for example, in QT keyboard demo program from beaglebonedemo, push "2" maybe come out "8", very frustrating.

Please advise, thanks in advance

Mike

post119307: How to show YUV image on RGB display panel through QNX 6.6 Screen?

2018-11-27T01:32:54Z

Hi,

    I want to show YUV image on the RGB display panel through the APIs of Screen, but I can not work ok, could you please provide me some demos for it?

    The development environments are follow:

    1. QNX6.6.0

    2. i.MX6DL

    3.  Screen

    4. dual lvds display pannel(1920*720)

    5. NV12 images

Best regards,

Gary

post118575: Re: Gles2 app cannot find glesv2-sc-dlls

2018-02-14T09:23:19Z

Alright, thank you. I'll ignore those error messages then.

post118573: Re: Gles2 app cannot find glesv2-sc-dlls

2018-02-13T17:06:44Z

It's an artifact of the Vivante porting layer and can be ignored (as long as there is no file named 'glesv2-sc-dlls' in LD_LIBRARY_PATH).

post118564: Gles2 app cannot find glesv2-sc-dlls

2018-02-13T08:10:58Z

Dear community,

I have an i.MX6 quad based system running QNX6.6.0. While deploying Qt5, I noticed something a bit odd. When starting some of the example apps which came with the build, some of them would give me an error message (with LD_DEBUG=libs, the app does still start up despite the error):

libs: load_object: attempt load of glesv2-sc-dlls.so

libs: dlopen: Library cannot be found

libs: dlopen("/usr/lib/graphics/iMX6X/libGLSLC.so",1536)

Mode: LOCAL GROUP

I've attached a full dump. Someone in a different post suggested it might be of no relevance but I would like to be sure. I've also attached our graphics.conf - maybe I configured the glesv2-sc-dlls section wrong - the libSLSLC.so does exist, and is located in /usr/lib/graphics/iMX6X, executable. I get the error message for gles2-gears (although as stated above, it runs fine) and also for a few Qt example apps which use gles2.

Any thoughts on whether or not this error message is actually relevant?

post118257: Re: RE: Img library - io_open(IO_MEM, IO_WRITE, ...)

2017-12-01T15:40:41Z

Thank you for fast response.

I would like to do it myself.

In particular - how to properly initialize

struct _io_bstream {

io_stream_t stream;

unsigned flags;

_uint8 *buf;

unsigned bufsiz, buf_pos, buf_nleft;

unsigned pos, nleft;

io_close_f *close_f;

}

in my version of "io_open_f()"?

I have tried but it does not work properly (perhaps flags or sizes not set properly).

Is there doc or example?

post118256: RE: Img library - io_open(IO_MEM, IO_WRITE, ...)

2017-12-01T14:30:42Z

No, it is a read-only interface.  Only the IO_FD interface supports write functionality.

-----Original Message-----

From: Andrzej Polanski(deleted) [mailto:community-noreply@qnx.com]

Sent: Friday, December 1, 2017 5:29 AM

To: advanced-graphics <advanced-graphics@community.qnx.com>

Subject: Img library - io_open(IO_MEM, IO_WRITE, ...)

How to use io_open() for write into memory buffer?

Try like

   output = io_open(IO_MEM, IO_WRITE, (unsigned)BUF_SIZE, (void*)buf);

gives errno = ENOTSUP         48  /* Not supported (1003.1b-1993)   */

Is it possible?

Additional remark - documentation for image library is rather brief (e.g. not mentioned how to free memory allocated with img_load())

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post118255

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post118255: Img library - io_open(IO_MEM, IO_WRITE, ...)

2017-12-01T10:28:42Z

How to use io_open() for write into memory buffer?

Try like

output = io_open(IO_MEM, IO_WRITE, (unsigned)BUF_SIZE, (void*)buf);

gives errno = ENOTSUP 48 /* Not supported (1003.1b-1993) */

Is it possible?

Additional remark - documentation for image library is rather brief (e.g. not mentioned how to free memory allocated with img_load())

post117768: Re: RE: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-05-27T07:58:59Z

Hi Mr Yang：

I‘m from china too and have encountered the same problem,

Have you resolved this problem?

in case you have resolved, can you give me some suggestions on it?

Thank you very much!

post117358: SIGSEGV fault while calling vgCreateEGLImageTargetKHR API

2017-01-20T17:29:53Z

I'm creating a graphics application to run on i.MX6 Sabre SDB and I’m running into a SIGSEGV fault when I call ‘vgCreateEGLImageTargetKHR’ API.  Can someone please help me figure out the root cause for this?

I’m trying to use the EGL_KHR_image extension as a means to share images between the OpenGLES2 and OpenVG rendering APIs.

So, I wrote some code to create an EGLImageKHR type as shown below.  After successful creation of the EGLImageKHR, I was able to create an OpenGLES2 texture using the EGL image (glEGLImageTargetTexture2DOES ) and render it to the display.

But, I see a SIGSEGV fault, when I attempt to create an OpenVG texture using the same EGL Image.  The OpenVG extension function that I’m using is - vgCreateEGLImageTargetKHR

void * driver_CreateEglImageBuffer(EGL_DDB *pDdb)

{

    int             size[2] = {pDdb->width, pDdb->height};

    int             format = SCREEN_FORMAT_RGBA8888;

    void *          ret_buffer = NULL;

    int             result;

    /* Create a pixmap buffer.  This buffer will be used to hold the

    ** bitmap and eventually create an eglImage.

    */

    result = screen_create_pixmap((screen_pixmap_t *)&(pDdb->pixmap.pixmapId),

                                  egl_layer->win.screen_ctx);

    if (result == 0)

    {

        int usage = SCREEN_USAGE_OPENGL_ES2 | SCREEN_USAGE_OPENVG;

        screen_buffer_t screen_pbuf;

        screen_set_pixmap_property_iv(pDdb->pixmap.pixmapId, SCREEN_PROPERTY_USAGE, &usage);

        screen_set_pixmap_property_iv(pDdb->pixmap.pixmapId, SCREEN_PROPERTY_FORMAT, &format);

        screen_set_pixmap_property_iv(pDdb->pixmap.pixmapId, SCREEN_PROPERTY_BUFFER_SIZE, size);

        result = screen_create_pixmap_buffer(pDdb->pixmap.pixmapId);

        if (result != 0)

        {

            screen_destroy_pixmap(pDdb->pixmap.pixmapId);

            printf("Fail! screen_create_pixmap_buffer() returned %d\n", result);

        }

        else

        {

            screen_get_pixmap_property_pv(pDdb->pixmap.pixmapId, SCREEN_PROPERTY_RENDER_BUFFERS,

                                                   (void**)&screen_pbuf);



            /* Get the buffer pointer */

            screen_get_buffer_property_pv(screen_pbuf, SCREEN_PROPERTY_POINTER,

                                                   (void**)&pDdb->pixmap.pixmapBuffer);

            /* Get the buffer stride */

            screen_get_buffer_property_iv(screen_pbuf, SCREEN_PROPERTY_STRIDE,

                                                    (int *)&(pDdb->stride));

            /* Create EGL Image */

            pDdb->pixmap.eglImage = eglCreateImageKHR(egl_layer->win.eglDisplay,

                                                     EGL_NO_CONTEXT,

                                                     EGL_NATIVE_PIXMAP_KHR,

                                                     (void *)pDdb->pixmap.pixmapId,

                                                     NULL);

            if (pDdb->pixmap.eglImage == EGL_NO_IMAGE_KHR)

            {

                EGLint eglError = eglGetError();

                if (eglError != EGL_SUCCESS)

                {

                    printf("eglGetError() = %i\n", eglError);

                }

            }

            else

            {

                ret_buffer = pDdb->pixmap.pixmapBuffer;

            }

        }

    }

    return ret_buffer;

}

post117323: Re: RE: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-01-11T01:30:05Z

All right, thanks all the same.

> Yes, please ask your FAE to contact support directly on your behalf.

>

> Regards,

> Derek

>

> -----Original Message-----

> From: yang lei [mailto:community-noreply@qnx.com]

> Sent: Monday, January 9, 2017 3:29 AM

> To: advanced-graphics <advanced-graphics@community.qnx.com>

> Subject: Re: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-

> devices board

>

> Hi, Derek

>     Thanks for your reply, but my FAE told me that use the standard support

> plan. I new a case to request the sample source on 05/01/2017 1:34 AM(yes

> because I am in China), till now the case´s status is still ¨Unassigned¨.

>     I don´t know how to do next... I guess I should find my FAE again..

> > Hi,

> >

> > Your FAE/Sale Rep should be able to provide you with the latest source.

> >

> > Regards,

> > Derek

> >

> > -----Original Message-----

> > From: yang lei [mailto:community-noreply@qnx.com]

> > Sent: Thursday, January 5, 2017 6:01 AM

> > To: advanced-graphics <advanced-graphics@community.qnx.com>

> > Subject: How to get wfdcfg source code(LCD) for

> > I.MX6q-sabre-smart-devices board

> >

> > Hi, I'm new in writing video drivers for QNX 6.6, any suggestion will

> > be very helpful.

> >

> >     My target board is freescale I.MX6Q sabre smart device. I download

> > ¨ BSP_freescale-imx6q-sabresmart_br-660_be-660_PATCHES_JBN775¨ and I

> > read the sample source code(hdmi.c) for days, I can get something but

> > my target board´ s display device is a 12.3 inch LCD screen.

> > Unfortunately this hardware is not  on the BSP support list, so I need

> > a LCD driver sample source code, edit the timing data.

> >     How do I get the wfdcfg.c? My FAE told me that the DDK is out of

> > the date,  and let me go to here for help.

> >

> >

> >

> > _______________________________________________

> >

> > Advanced Graphics

> > http://community.qnx.com/sf/go/post117307

> > To cancel your subscription to this discussion, please e-mail

> > advanced- graphics-unsubscribe@community.qnx.com

>

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post117319

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

post117320: RE: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-01-09T13:16:56Z

Yes, please ask your FAE to contact support directly on your behalf.

Regards,

Derek

-----Original Message-----

From: yang lei [mailto:community-noreply@qnx.com]

Sent: Monday, January 9, 2017 3:29 AM

To: advanced-graphics <advanced-graphics@community.qnx.com>

Subject: Re: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

Hi, Derek

    Thanks for your reply, but my FAE told me that use the standard support plan. I new a case to request the sample source on 05/01/2017 1:34 AM(yes because I am in China), till now the case´s status is still ¨Unassigned¨.

    I don´t know how to do next... I guess I should find my FAE again..

> Hi,

>

> Your FAE/Sale Rep should be able to provide you with the latest source.

>

> Regards,

> Derek

>

> -----Original Message-----

> From: yang lei [mailto:community-noreply@qnx.com]

> Sent: Thursday, January 5, 2017 6:01 AM

> To: advanced-graphics <advanced-graphics@community.qnx.com>

> Subject: How to get wfdcfg source code(LCD) for

> I.MX6q-sabre-smart-devices board

>

> Hi, I'm new in writing video drivers for QNX 6.6, any suggestion will

> be very helpful.

>

>     My target board is freescale I.MX6Q sabre smart device. I download

> ¨ BSP_freescale-imx6q-sabresmart_br-660_be-660_PATCHES_JBN775¨ and I

> read the sample source code(hdmi.c) for days, I can get something but

> my target board´ s display device is a 12.3 inch LCD screen.

> Unfortunately this hardware is not  on the BSP support list, so I need

> a LCD driver sample source code, edit the timing data.

>     How do I get the wfdcfg.c? My FAE told me that the DDK is out of

> the date,  and let me go to here for help.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post117307

> To cancel your subscription to this discussion, please e-mail

> advanced- graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post117319

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post117319: Re: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-01-09T08:29:00Z

Hi, Derek

    Thanks for your reply, but my FAE told me that use the standard support plan. I new a case to request the sample source on 05/01/2017 1:34 AM(yes because I am in China), till now the case´s status is still ¨Unassigned¨.

    I don´t know how to do next... I guess I should find my FAE again..

> Hi,

>

> Your FAE/Sale Rep should be able to provide you with the latest source.

>

> Regards,

> Derek

>

> -----Original Message-----

> From: yang lei [mailto:community-noreply@qnx.com]

> Sent: Thursday, January 5, 2017 6:01 AM

> To: advanced-graphics <advanced-graphics@community.qnx.com>

> Subject: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices

> board

>

> Hi, I'm new in writing video drivers for QNX 6.6, any suggestion will be very

> helpful.

>

>     My target board is freescale I.MX6Q sabre smart device. I download ¨

> BSP_freescale-imx6q-sabresmart_br-660_be-660_PATCHES_JBN775¨ and I read the

> sample source code(hdmi.c) for days, I can get something but my target board´

> s display device is a 12.3 inch LCD screen. Unfortunately this hardware is not

>  on the BSP support list, so I need a LCD driver sample source code, edit the

> timing data.

>     How do I get the wfdcfg.c? My FAE told me that the DDK is out of the date,

>  and let me go to here for help.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post117307

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

post117310: RE: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-01-05T14:24:19Z

Hi,

Your FAE/Sale Rep should be able to provide you with the latest source.

Regards,

Derek

-----Original Message-----

From: yang lei [mailto:community-noreply@qnx.com]

Sent: Thursday, January 5, 2017 6:01 AM

To: advanced-graphics <advanced-graphics@community.qnx.com>

Subject: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

Hi, I'm new in writing video drivers for QNX 6.6, any suggestion will be very helpful.

My target board is freescale I.MX6Q sabre smart device. I download ¨BSP_freescale-imx6q-sabresmart_br-660_be-660_PATCHES_JBN775¨ and I read the sample source code(hdmi.c) for days, I can get something but my target board´s display device is a 12.3 inch LCD screen. Unfortunately this hardware is not on the BSP support list, so I need a LCD driver sample source code, edit the timing data.

How do I get the wfdcfg.c? My FAE told me that the DDK is out of the date, and let me go to here for help.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post117307

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post117307: How to get wfdcfg source code(LCD) for I.MX6q-sabre-smart-devices board

2017-01-05T11:00:35Z

Hi, I'm new in writing video drivers for QNX 6.6, any suggestion will be very helpful.

My target board is freescale I.MX6Q sabre smart device. I download ¨BSP_freescale-imx6q-sabresmart_br-660_be-660_PATCHES_JBN775¨ and I read the sample source code(hdmi.c) for days, I can get something but my target board´s display device is a 12.3 inch LCD screen. Unfortunately this hardware is not on the BSP support list, so I need a LCD driver sample source code, edit the timing data.

How do I get the wfdcfg.c? My FAE told me that the DDK is out of the date, and let me go to here for help.

post114472: Problem with graphics performance

2015-09-18T09:17:52Z

Hello,

I'm quite new to all the graphics stuff under QNX. I'm using QNX 6.4.1 on i.MX515 with 320x240 16 bit display and I have a problem with performance:

1. A simple loop drawing (gf_draw_rect()) one white rectangle on the whole screen at 30fps uses 26% of CPU.

2. vsync demo uses 66% of CPU

3. egl-gears uses 80% of CPU

My guess is that the software is not using hardware acceleration. I have been searching for more info about it, but can't find anything useful. What I keep finding is the sentence “(...) using hardware acceleration where possible”, but I have no idea how to make it possible.

The answer is probably simple enough to not put it in tutorials, but apparently I don't have enough knowledge in that topic and I don't even know where to look for it.

Could anyone point me in the right direction, please?

post114327: Re: libfont error: acessing a corrupted share lib on armv7,who can help me?

2015-08-23T19:30:25Z

Hi Kevin,

Sure that you use library for your CPU architecture (armle-v7). You should copy it from ${QNX_TARGET}/armle-v7/lib instead of /lib.

Respectfully,

Oleg

23 авг. 2015 г., в 7:33, kevin L <community-noreply@qnx.com> написал:

> when I use PfAttachLocalDll, there is an error: acessing a corrupted share lib.

> i find libfont in QNX installation /lib,and put in beagleboard /lib.

> i put ttf font file n /usr/photon/font-repository.

>

>

> I try it on vmware QNX, it's OK.

>

> hope for your help.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post114326

> To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post114326: libfont error: acessing a corrupted share lib on armv7,who can help me?

2015-08-23T04:33:36Z

when I use PfAttachLocalDll, there is an error: acessing a corrupted share lib.

i find libfont in QNX installation /lib,and put in beagleboard /lib.

i put ttf font file n /usr/photon/font-repository.

I try it on vmware QNX, it's OK.

hope for your help.

post114203: Video BIOS equivalent services for ARM

2015-08-04T02:49:32Z

Hi,

I am working on an application that would run on QNX RTOS 6.5. The application would run on ARM processor.

The application would have to access the Video memory, to check the contents of the display, in order to set and read any random pixel, also make sure the display is not frozen, etc.

I see that the services provided by Video BIOS services are what I need (e.g. vbios _ int(), vbios _ register(), etc), but these services are available only for x86 and not for ARM.

My question is, are there any equivalent services that I could use for ARM?

Thanks in advance.

Phil

post113932: Re: Why QNX6.6 after the close of the program is not completed drm-intel?

2015-06-08T15:42:16Z

What state is drm-intel after screen has terminated?

Thanks,

Dave

post113741: Why QNX6.6 after the close of the program is not completed drm-intel?

2015-04-22T22:59:09Z

Hello, in QNX6.6 for to run the Qt-application I use the next script:

-----------------------------------------------------------------------------------------------------------------------------------------------------

export LD_LIBRARY_PATH=/lib:/usr/lib:/lib/dll:/opt/qt5/lib:/opt/qnx6/x86/lib:/opt/qnx6/x86/usr:/opt/qnx6/x86/lib/dll

export QQNX_PHYSICAL_SCREEN_SIZE=200, 200

export QT_PLUGIN_PATH=/opt/qt5/plugins

export QML2_IMPORT_PATH=/opt/qt5/qml

export GRAPHICS_ROOT=/usr/lib/graphics/intel-drm

drm-intel

screen

/root/app/my_prog qnx:no-fullscreen

slay screen

slay drm-intel

-----------------------------------------------------------------------------------------------------------------------------------------------------

On the PC, after the close of my program, screen completed its work. But drm-intel no exited.

Because of what it could be?

On Vmware, when I use instead of drm-intel - vmware (though it is not explicitly called in a script), screen shuts down and I can continue to work with mqc.

Thanks

Vova

post113168: Re: screen_attach_window_buffers

2015-01-31T01:23:04Z

And the attachment

post113167: screen_attach_window_buffers

2015-01-31T01:17:34Z

Our video capture device is already using a linked list of buffers for transfer of video data via DMA.

I'd like to use these same buffers for screen and just post them when they are filled.  The attached code snippet outlines the basics of my approach, but it isn't working.

To illustrate and try to figure this out I'm using the sync tutorial.  You'll see the #ifdef clause that contains the code I would like to execute and have it work.

The result of the #ifdef block is a blank screen.  Of course, using the original code works.  Basically I would like to allocate the memory and assign it to a screen buffer and have screen use that memory instead of its own allocation.

Docs suggest that this is known and supported method, but apparently I'm missing some detail.

post113140: Re: Writing video driver for QNX 6.6

2015-01-29T14:42:00Z

OK, thank you very much!

post113138: Re: Writing video driver for QNX 6.6

2015-01-29T14:18:01Z

No, I can't provide you src.

You can use this doc as reference on start:

https://www.khronos.org/registry/wf/specs/OpenWF_Display_1_0_Specification.pdf

Hope it helps. Good luck.

post113137: Run screen on QNX6.5

2015-01-29T14:04:40Z

I'm new to QNX os, any suggested tutorial how to run screen on QNX6.5 will be very helpful.

I'm familiarized with:

a) Screen Graphics Subsystem Developer's Guide

b) OpenWF Display Configuration Developer's Guide

Since there is no WFD and Wfdcfg driver src available to me, I have no idea how to write my hardware specific WFD drivers. My processor is TI-AM3517, there is no available binary for me. Would anyone help me on how to write WFD driver? Thank you very much!

post113136: Re: Writing video driver for QNX 6.6

2015-01-29T13:53:45Z

Dear Lezhava,

I'm new inQNX os and I wanna run screen on QNX6.5. I have read the below articles:

a) Screen Graphics Subsystem Developer's Guide

b) OpenWF Display Configuration Developer's Guide

But I still have no idea how to write WFD driver. Would you please provide your WFD driver src as a sample to me? My email is wangwei_1720@126.com. Thank you very much!

post112996: Re: gf_surface_attach returns GF_ERR_OK but img_surf is a NULL pointer.

2015-01-19T16:05:56Z

As I understand you trying to show transparent image with gf_splash? To draw image with transparecy this application uses gf_draw_blit2() with configured chroma key. Does your driver supports this feature?

Best regards

Andrey

post112993: Re: gf_surface_attach returns GF_ERR_OK but img_surf is a NULL pointer.

2015-01-19T13:00:52Z

Its the gf_splash application - available in several bsp's. Running on an arm-v7 target. Looks like there might be a possibility for gf_surface_attach to return ok but not deliver a surface opaque pointer.

Any idea what (can) cause(s) this behaviour?

tia

George

post112992: Re: gf_surface_attach returns GF_ERR_OK but img_surf is a NULL pointer.

2015-01-19T12:06:56Z

Hello George,

What do you mean by "external memory"? Can you please provide a few examples of code?

Best regards

Andrey

post112985: gf_surface_attach returns GF_ERR_OK but img_surf is a NULL pointer.

2015-01-18T23:26:02Z

QNX 6.5.0 SP1

What would cause gf_surface_attach to return GF_ERR_OK, but not create a surface for the 'external' memory?

best regards

George Refseth

post112721: Writing video driver for QNX 6.6

2014-12-17T21:19:38Z

I'm new in writing video drivers for QNX 6.6, and any suggested tutorial how to do this will be very helpful.

I'm familiarized with:

a) Screen Graphics Subsystem Developer's Guide

b) OpenWF Display Configuration Developer's Guide

Currently I've written my WFD.so driver with all functions implemented as stubs, set it as "wfd-dlls" in graphics.conf and after running Screen I received SIGSEGV error:

"Process 106518 (screen) terminated SIGSEGV code=1 fltno=11 ip=0804f538(/sbin/screen@main+0x50f) ref=0000045c".

My graphics.conf:

begin khronos

  begin egl display 1

  end egl display

  begin wfd device 1

    wfd-dlls = WFD.so

  end wfd devide

end khronos

begin winmgr

  begin globals

  end globals

  begin display hdmi

  end display

  begin class framebuffer

  end class

  begin mtouch

  end mtouch

end winmg

Thanks.

post112356: Re: QNX 6.6 touch screen setup

2014-11-17T00:38:00Z

I've changed touch screen controller to HID compliant device. After figuring out what's going on, QNX support guy said that something is different between their egalax touch screen controller and mine. And if I want to go further, I have to pay 10000$ at least for priority support(bronze). Even though it works with QNX 6.5.0.

Anyway, now I can touch and move the cursor on the screen but coordinates are not fit.

So I've tried to calibrate but result was same as you. The cursor moves, but calib-touch program doesn't accept or recognize that.

I'm using a screen with 1024x600. But when I run 'devi-hid' from command line, some suspected output displayed.

# devi-hid -vvvvvvv mouse touch

...

Did not find calib info for [639x479]. Attempting to use calculated transformation.

...

I've set the resolution to 1024x600 through the same way as you and I don't know why that message displayed.

As your result is same with me, I wonder what's the output from devi-hid from command line with -vvvvvvv?

post112242: Re: QNX 6.6 touch screen setup

2014-11-04T10:56:02Z

Hi,

I am also trying to use touch screen with QNX 6.6 on TI Jacinto 6 (DRA74x) EVM (Rev E1) with 7" display.

I have used the screen patch and apps & media and run the driver for touch as per the QNX SDP user guide that came along with the BSP.

When I run the calib-touch utility, I get the calibration display but my touch inputs don't get registered.

I am not able to calibrate.

I have added the resolution of LCD I am using (800x480) to /etc/system/config/scaling.conf i.e.

800x480:mode=direct

Here is the sloginfo output when I run screen:

Jan 01 00:00:15    5     8   300 screen: starting up...

Jan 01 00:00:15    5     8   300 screen: connected to powman...

Jan 01 00:00:15    5     8   200 wfd: 1 wfdCreateDevice - Tiler is not available. Video buffer will be allocated in normal memory

Jan 01 00:00:15    5     8   200 wfd: 1 configure_sys_clock - System clock is set to: 20000 KHz

Jan 01 00:00:15    5     8   200 wfd: 1 configure_dss_func_clock - DSS func clock is set to: 192000 KHz

Jan 01 00:00:15    2     8   300 screen: invalid display id:

>>>   begin display 1

Jan 01 00:00:15    5     8   300 screen: loading EGL display 1...

Jan 01 00:00:15    2     8   100 SGX - Starting SGX release driver (version: 1.9@2291151)

Jan 01 00:00:15    2     8   100 SGX - Power management turn ON

Jan 01 00:00:15    2     8   100 SGX - TKT26818_OPTIMIZATION_SUBSET enabled

Jan 01 00:00:15    2     8   100 SGX - SKIP_ISR_PULSE_WHEN_GPU_OFF enabled

Jan 01 00:00:15    5     8   300 screen: loading blit module pvr2d...

Jan 01 00:00:15    5     8   300 screen: loading composition module pvr2d...

Jan 01 00:00:15    5     8   300 screen: loading mtouch module devi...

Jan 01 00:00:15    5     9   200 devi[INFO]:  Connected to TP on /dev/devi/touch0

Jan 01 00:00:15    5     9   100 libinputevents[INFO]: Attaching mtouch driver

Jan 01 00:00:15    5     9   100 libinputevents[INFO]: Opening mtouch device #0

Jan 01 00:00:15    5     8   300 screen: attached to mtouch0

So it doesn't look like there is any problem with the driver.

Can anyone suggest what could be going wrong?

Thanks in advance.

I am attaching the graphics.conf that I am using.

Regards,

Abhishek

post112123: Re: QNX 6.6 touch screen setup

2014-10-23T07:29:44Z

No, I'm still stuck on it.

In my case, touchscreen controller is not HID-compliant so I can't use "io-hid -dusb" instead of "io-hid -degalax".

I'm using ETP-4500UG-B touchscreen controller.

I've been asking to QNX support from two weeks ago but no luck yet.

post112098: Re: QNX 6.6 touch screen setup

2014-10-22T08:04:59Z

I solved my problem:

I didn`t use egalax driver, but used usb driver:

# io-usb -dehci -dohci -duhci -dxhci

# io-hid -dusb

# hidview

HIDD v1.00, v1.00 DDK

Device Address       : 0

Vendor               : 0x0eef (eGalax Inc.)

Product              : 0x0001 (USB TouchController)

Version              : r1.10

Usage                : Pointer

# devi-hid touch

# screen

# calib-touch -recalibrate

graphics.conf:

begin khronos

  begin egl display 1

    egl-dlls = libEGL-mesa.so

    glesv1-dlls = libglapi-mesa.so libGLESv1_CM-mesa.so

    glesv2-dlls = libglapi-mesa.so libGLESv2-mesa.so

  end egl display

  begin wfd device 1

    wfd-dlls = libwfdcfg-sample.so libWFDintel-drm.so

    pipeline1-display = 1

#    pipeline2-display = 2

    pipeline10-display = 1

#    pipeline11-display = 2

  end wfd device

end khronos

begin winmgr

  begin globals

    stack-size = 2048  # in units of 1024 bytes

    blit-config = gles2blt

  end globals

  begin display 1

    video-mode = 1280 x 1024 @ 60

    stack-size = 2048  # in units of 1024 bytes

  end display

#  begin display 2

#    video-mode = 800 x 600 @ 60

#    stack-size = 2048  # in units of 1024 bytes

#  end display

  begin mtouch

        driver = devi

        options = height=1024,width=1280

        display = 1

  end mtouch

end winmgr

scaling.conf:

# Scaling based on calibration data

1280x1024:mode=calib,disp_x=[102:922:922:102],disp_y=[60:60:540:540],mtouch_x=[172:1108:1108:172],mtouch_y=[110:110:658:658]

# Scaling based on physical dimensions

1280x1024:mode=dim,width_mm=154,height_mm=90,border_left_mm=5,border_right_mm=5,border_top_mm=5,border_bottom_mm=5

# Simple scaling to client resolution

1280x1024:mode=scale

# No scaling, report coordinates exactly as the driver hands them to the client

1280x1024:mode=direct

Good luck to everyone....

post112097: Re: QNX 6.6 touch screen setup

2014-10-22T06:06:30Z

May be uses egalax driver in section (as the documentation):

begin mtouch

    driver = egalax

    options = height=600,width=1024

    display = 1

  end mtouch

But then:

libs: load_object: attempt load of libmtouch-egalax.so

libs: dlopen: Library cannot be found

Where to get the library "libmtouch-egalax.so"?

post112085: Re: QNX 6.6 - i965 - screen startup failed

2014-10-21T16:32:50Z

Fixed after a update provided by QNX Support Team.

post112081: Re: QNX 6.6 touch screen setup

2014-10-21T14:23:17Z

Hi,

did you have any progress? I am having the same problem.

Regards

post112071: Re: Issue with QNX 6.6 drm_intel_hd

2014-10-20T12:41:58Z

Did you apply lastest patch?

http://www.qnx.com/developers/articles/rel_5948_25.html

Run qconfig on Windows prompt to check.

DRM now runs with drm-intel instead of drm_intel_hd.

post112056: Issue with QNX 6.6 drm_intel_hd

2014-10-16T07:02:39Z

Hi,

We have a target system with INTEL Graphics Chipset .

We were able to port QNX on to it and then tried running drm_intel_hd followed by Screen. But, on starting drm_intel_hd i am getting errors as Invalid Checksum and getting errors with Screen also. However, errors in screen is due to lack to drm_intel_hd being open for screen.

I am attaching you the errors , if any one could help me with solution, that would be great.

-Thank You.

post111993: Re: Intel DRM I7 4770K screen issues

2014-10-08T22:46:00Z

And the other attachment....

post111992: Intel DRM I7 4770K screen issues

2014-10-08T22:44:20Z

Couple of questions:

1.  How can I configure graphics.conf to have two displays mirror each other?  Attached is my graphcis.conf file

2.  Attached as well is output from slogger.  I'm having some weirdness surrounding "potential gpu lockup"  and "invlid max DP link"

Finally, when I am able to run gears on display 7, it runs really, really slow, about 0.995 frames per second.

Thanks for your help :)

Kevin

post111624: Re: QNX 6.6 touch screen setup

2014-09-04T01:40:52Z

If I run devi-hid in command line, output is different.

# devi-hid -vvvvvvv touch

Looking up touch in module table

Calibration filename: /etc/system/config/calib.localhost

Resetting module touch

Resetting module abs

Error: No such file or directory - cannot open calibration file for absolute filter module(/etc/system/config/calib.localhost)

Did not find calib info for [639x479]. Attempting to use calculated transformation.

Calibration Matrix An: 0, Bn: 0, Cn: 0, Dn: 0, En: 0, Fn: 0, Divider: 0

If /etc/system/config/calib.localhost exist, just "Error:No such ..." message removed and no difference.

post111623: Re: QNX 6.6 touch screen setup

2014-09-04T00:32:01Z

About the error of calibration file missing,

if I put calibration file from qnx650 to /etc/system/config then the error message has removed but still not working.

post111622: Re: QNX 6.6 touch screen setup

2014-09-04T00:20:51Z

Thanks for reply.

I've changed devi-hid command of rc.local to:

devi-hid -vvvvvvv touch

then sloginfo output after booting:

--------------------------------------------

# sloginfo

Time             Sev Major Minor Args

Sep 04 00:53:51    2    19   1500 devb-ahci cam ver 100 sim 1.00A (Feb 22 2014 18:20:52)

Sep 04 00:53:51    2     5     0 libcam.so (Feb 22 2014 18:20:44) bver 6060109

Sep 04 00:53:51    2    19   1500 eide_identify_devices: AHCI (generic) vid 0x8086, did 0x1e03, class 0x10601, rev 0x4, busno 0x0, dfunc 0xfa, base 0xf7d36000, irq 0xf, xlat 0x0

Sep 04 00:53:51    2    19   1500 eide_display_identify: mdl SAMSUNG MZMPC064HBDR-00000 sn S0VRNYAC200381 fw CXM1201Q port 2, max udma 6, cur udma 5, max mdma 2, cur mdma 0, pio 4, mblk 16

Sep 04 00:53:51    2     5   100 cam-disk.so (Feb 22 2014 18:20:50)

Sep 04 00:53:51    5     3   1000 fs-qnx6: trim was requested, and is usable

Sep 04 00:53:51    5     6     0 random: Selecting timer as entropy source

Sep 04 00:53:51    5     6     0 random: All ready, registering path names

Sep 04 00:53:51    5     6     0 random: Daemonizing the process

Sep 04 00:53:51    2     9     0 Error: No such file or directory - cannot open calibration file for absolute filter module(/etc/system/config/calib.localhost)

Sep 04 00:53:51    5    14     0 tcpip starting

Sep 04 00:53:51    3    14     0 Using pseudo random generator.  See "random" option

Sep 04 00:53:51    5    14     0 initializing IPsec... done

Sep 04 00:53:51    5    14     0 IPsec: Initialized Security Association Processing.

Sep 04 00:53:51    5    14     0 wm0

Sep 04 00:53:51    5    14     0 wm1

Sep 04 00:53:51    5    14     0 wm2

Sep 04 00:53:51    5    14     0 wm3

Sep 04 00:53:51    5    14     0 wm4

Sep 04 00:53:52    5    14     0 wm5

Sep 04 00:53:52    5     8   300 screen: starting up...

Sep 04 00:53:52    5     8   300 screen: connected to powman...

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    1     7     0 driver missing hw device index callback, ducking will not work properly between devices

Sep 04 00:53:52    3     8   915 [ihold:265]"warning"

Sep 04 00:53:52    5     8   300 screen: loading EGL display 1...

Sep 04 00:53:52    5     8   300 screen: loading blit module gles2blt...

Sep 04 00:53:53    3    12     0 devh-egalax.so - Failed to send check active (loopback) command to the controller: 5

Sep 04 00:53:53    5     8   300 screen: display_mode_set: selecting preferred mode 5 for display 1

Sep 04 00:53:53    5     8   300 screen: loading composition module gles2blt...

Sep 04 00:53:53    5     8   300 screen: win_cursor_init: No cursor specific in graphics.conf

Sep 04 00:53:53    5     8   300 screen: win_cursor_shape_init: using default cursor (9x16)

Sep 04 00:53:53    5     8   300 screen: loading mtouch module devi...

Sep 04 00:53:53    5     9   200 devi[INFO]:  Connected to TP on /dev/devi/touch0

Sep 04 00:53:53    5     9   100 libinputevents[INFO]: Attaching mtouch driver

Sep 04 00:53:53    5     9   100 libinputevents[INFO]: Changing mtouch coordinate mapping

Sep 04 00:53:53    5     9   100 libinputevents[INFO]: Opening mtouch device #0

Sep 04 00:53:53    5     8   300 screen: attached to mtouch0

Sep 04 00:53:53    5     8   300 screen: connected to HID v1.0 DDK 1.0

Sep 04 00:53:53    5     8   300 screen: HID device inserted:

Sep 04 00:53:53    5     8   300 screen: Device address: 0

Sep 04 00:53:53    5     8   300 screen: Vendor ID: 0x0eef (Unknown manufacturer)

Sep 04 00:53:53    5     8   300 screen: Product ID: 0x0001 (Unknown product)

Sep 04 00:53:53    5     8   300 screen: Version: r1.10

Sep 04 00:53:53    1    12     0 devh-egalax.so - Controller set to Calibrated point

Sep 04 00:53:53    5     8   300 screen: hid_insertion: attaching to report usage_page=0x1 usage=0x1

Sep 04 00:53:53    5     8   300 screen: HID device inserted:

Sep 04 00:53:53    5     8   300 screen: Device address: 1

Sep 04 00:53:53    5     8   300 screen: Vendor ID: 0x045e (Microsoft)

Sep 04 00:53:53    5     8   300 screen: Product ID: 0x0752 (Wired Keyboard 400)

Sep 04 00:53:53    5     8   300 screen: Version: r1.10

Sep 04 00:53:53    2     8   300 screen: Unable to load keyboard mapping file en_US_101 [2]

Sep 04 00:53:53    2     8   300 screen: win_input_create: failed to allocate a keymap for an new input device

Sep 04 00:53:53    5     8   300 screen: hid_insertion: attaching to report usage_page=0x1 usage=0x6

--------------------------------------------

I can guess below message tells me it's not communicating.

Sep 04 00:53:53    3    12     0 devh-egalax.so - Failed to send check active (loopback) command to the controller: 5

In sloginfo, touch pad recognized as unknown manufacturer and product but output of hidview:

# hidview

HIDD v1.00, v1.00 DDK

Device Address       : 0

Vendor               : 0x0eef (eGalax Inc.)

Product              : 0x0001

Version              : r1.10

Usage                : Pointer

Device Address       : 1

Vendor               : 0x045e (Microsoft)

Product              : 0x0752 (Wired Keyboard 400)

Version              : r1.10

Usage                : Keyboard

I don't know what to do next. Any suggestion?

post111618: Re: QNX 6.6 - i965 - screen startup failed

2014-09-03T15:31:38Z

Below sloginfo.

Runing LD_DEBUG=libs i checked i965 driver being loaded

# show_vesa

VESA Info Block

VESA Signature : VESA

VESA Version   : 0300

OEM String Ptr : 0080:0100

OEM String     : Intel(R)Valleyview Graphics Chipset Accelerated VGA BIOS

# drm-probe-displays

# drm-probe-displays

count_displays : 5

count_pipelines: 2

display 1: embedded displayport, connected

Mode: "1024x768" 1024x768 60

Mode: "1024x768" 1024x768 50

display 2: VGA, connected

Mode: "1024x768" 1024x768 60

Mode: "800x600" 800x600 60

Mode: "800x600" 800x600 56

Mode: "848x480" 848x480 60

Mode: "640x480" 640x480 60

display 3: HDMI-A, disconnected

display 4: HDMI-A, disconnected

display 5: displayport, disconnected

pipeline 1

pipeline 2

# sloginfo:

Sep 03 09:59:57    5     8   300 screen: starting up...

Sep 03 09:59:57    5     8   300 screen: connected to powman...

Sep 03 09:59:57    3     8   915 [ihold:265]"warning"

Sep 03 09:59:57    2     8   200 drm_intel_bufmgr_gem_init failed

Sep 03 09:59:57    2     8   300 screen: win_enumerate_displays: could not create display device 1

Sep 03 09:59:57    2     8   300 screen: invalid display id:

        >>>   begin display 1

Sep 03 09:59:57    2     8   300 screen: invalid display id:

        >>>   begin display 2

Sep 03 09:59:57    5     8   300 screen: loading EGL display 1...

Sep 03 09:59:57    5     8   300 screen: loading blit module gles2blt...

Sep 03 09:59:57    3     8   100 egl: DRI2: failed to create dri screen

Sep 03 09:59:57    2     8   300 screen-gles2blt: eglInitialize: EGL is not initialized, or could not be initialized, for the specified display

Sep 03 09:59:57    2     8   300 screen: win_blit_module_init: could not create a blitting context

post111617: Re: QNX 6.6 - i965 - screen startup failed

2014-09-03T15:21:56Z

Hi, are you running drm & screen inside vmware?

from the error message, it looks like the function drm_intel_bufmgr_gem_init failed, normally it should work fine, it will fail if the graphics device id is not supported.

so can you run sloginfo and post the output ? also it will be helpful to run command pci and post the output.

Thanks

post111616: Re: QNX 6.6 touch screen setup

2014-09-03T11:48:34Z

Can you post the output from devi-hid in verbose mode after you touch the screen:

Devi-hid -vvvvvvv touch

Leave out the kid module as its already handled by screen.

post111614: QNX 6.6 - i965 - screen startup failed

2014-09-03T11:19:32Z

Hi,

while i was working with VMWare my graphics configuration were ok, but after boot QNX660 in a x86 PC with Intel Pineview video board my screen don't run anymore.

Same error happens in Intel Valleyview video board.

There is any configuration or initialization that i have to do before start drm-intel?

# drm-intel

# screen

[intel_init_bufmgr:946] Error initializing buffer manager.

libEGL warning: DRI2: failed to create dri screen

Screen startup failed

# graphics.conf

begin khronos

  begin egl display 1

    egl-dlls = libEGL-mesa.so

    glesv1-dlls = libglapi-mesa.so libGLESv1_CM-mesa.so

    glesv2-dlls = libglapi-mesa.so libGLESv2-mesa.so

  end egl display

  begin wfd device 1

    wfd-dlls = libwfdcfg-sample.so libWFDintel-drm.so

    pipeline1-display = 1

    pipeline2-display = 2

  end wfd device

end khronos

begin winmgr

begin globals

    blit-config = gles2blt

    input = keyboard

    stack-size = 2048

  end globals

  begin display 1

    video-mode = 1024 x 768 @ 60

    stack-size = 2048

  end display

  begin display 2

    video-mode = 1024 x 768 @ 60

    stack-size = 2048

  end display

  begin class framebuffer

    display = 1

    pipeline = 1

    format = rgba8888

    usage = gles2blt

  end class



  begin class framebuffer

    display = 2

    pipeline = 2

    format = rgba8888

    usage = gles2blt

  end class

end winmgr

post111612: QNX 6.6 touch screen setup

2014-09-03T07:53:00Z

Hello.

I'm trying to use touch screen at QNX 6.6.

I've searched and followed answers and then I'm seeing calib-touch screen now.

But I can't move the cursor or click.

After launching below programs in rc.local, calib-touch program could be run.

But touch doesn't work and cannot calibrate.

With QNX 6.5 touch screen works fine.

Does anyone have any idea what's wrong?

Thanks.

Launch at rc.local:

io-usb -duhci -dohci -dehci

io-hid -dusb -degalax

devi-hid kbd touch

drm-intel &

screen &

Here is hidview result:

# hidview

Device Address       : 0

Vendor               : 0x0eef (eGalax Inc.)

Product              : 0x0001

Version              : r1.10

Usage                : Pointer

Device Address       : 1

Vendor               : 0x045e (Microsoft)

Product              : 0x0752 (Wired Keyboard 400)

Version              : r1.10

Usage                : Keyboard

And this is my env variables:

# env

_=/usr/bin/env

SSH_CONNECTION=192.168.100.118 64713 192.168.100.177 22

PATH=/sbin:/usr/sbin:/usr/bin:/bin:/usr/sbin:/sbin:/usr/local/bin:/opt/bin:/opt/sbin:/bin:/sbin:/usr/bin:/usr/sbin

SHELL=/bin/sh

TMPDIR=/tmp

HOSTNAME=localhost

USER=root

PROCESSOR=x86

MAIL=/var/spool/mail/root

LD_LIBRARY_PATH=/prob/boot:/lib:/usr/lib:/lib/dll:/opt/qt5/lib:/opt/qnx6/x86/lib:/opt/qnx6/x86/usr/lib:/opt/qnx6/x86/lib

/dll:/usr/lib/graphics/intel-drm:/opt/qt5/plugins/platforms

QT_DEBUG_PLUGINS=1

HOME=/root

SSH_CLIENT=192.168.100.118 64713 22

QNX_HOST=/

QT_QPA_PLATFORM=qnx:no-fullscreen

QNX_TARGET=/

GRAPHICS_ROOT=/usr/lib/graphics/intel-drm

QQNX_PHYSICAL_SCREEN_SIZE=200,117

TERM=xterm

QML_IMPORT_PATH=/opt/qt5/imports

QML2_IMPORT_PATH=/opt/qt5/qml

QT_PLUGIN_PATH=/opt/qt5/plugins

FONTCONFIG_FILE=/etc/fontconfig/local.conf

QNX_CONFIGURATION=/etc/qnx

SSH_TTY=/dev/ttyp0

SYSNAME=nto

LOGNAME=root

This is my /usr/lib/graphics/intel-drm/graphics.conf:

begin khronos

  begin egl display 1

    egl-dlls = libEGL-mesa.so

    glesv1-dlls = libglapi-mesa.so libGLESv1_CM-mesa.so

    glesv2-dlls = libglapi-mesa.so libGLESv2-mesa.so

  end egl display

  begin wfd device 1

    wfd-dlls = libwfdcfg-sample.so libWFDintel-drm.so

    # Run "drm-probe-displays" to list the available displays and pipelines,

    # and "use $GRAPHICS_ROOT/libWFDintel-drm.so" for more information on

    # these driver-specific settings.

    # Pipeline IDs 1 to 9 are used for DRM CRTCs.  One should be assigned

    # to each display that will be used.

    pipeline1-display = 1

    # Pipeline IDs 10 and above are used for DRM planes.  A plane can only

    # be used on a display with an active CRTC.

    pipeline10-display = 1

  end wfd device

end khronos

begin winmgr

  begin globals

    input = keyboard mouse

    # Adjust the stack size of Screen's resmgr threads.  The default size

    # is insufficient for blitters/compositors using Mesa (e.g., gles2blt).

    blit-config = gles2blt

  end globals

  begin display 1

    video-mode = 1024 x 600 @ 75

    # Adjust the stack size of Screen's composition thread; required when the

    # display's framebuffer uses Mesa (e.g. "usage = gles2"), as noted above.

    cursor = on

    stack-size = 4096  # in units of 1024 bytes

  end display

  begin class framebuffer

    # Be sure to configure the "pipeline%d-display" options (above) so the

    # specified display and pipeline are bindable.  Normally a framebuffer

    # will use a CRTC pipeline.

    display = 1

    pipeline = 1

    format = rgba8888

    usage = gles2blt

  end class

  begin mtouch

    driver = devi

    options = height=600,width=1024

    display = 1

  end mtouch

end winmgr

post111553: RE: RE: Address of frame buffer

2014-08-26T14:34:28Z

I'd suggest referring to the documentation for the product you're using to get more acquainted.

http://www.qnx.com/developers/docs/

If you then have more specific questions, we could then probably more helpful.

-Joel

________________________________________

From: Radu Stancu [community-noreply@qnx.com]

Sent: Tuesday, August 26, 2014 10:16 AM

To: advanced-graphics

Subject: Re: RE: Address of frame buffer

I'm working on an embedded system for automotive that has omap 5 and qnx on it. My application is written in Qt 5.2  (I render some objects, like cubes and such on 2 screens). The main thing I need to do is somehow capture a frame on one screen and then proceed as I mentioned in the original post. The framebuffer approach was the first that came to mind, as I need something raw for dce encode. I don't really need a solution in Qt, a separate program written in C would work as well, . But I'm kinda lost, because I'm not that familiar with qnx.

> Perhaps you could mention which framework you're using? Derek's likely

> assuming GF because you're posting in the advance graphics forum, but you

> don't actually provide any hints of the framework.

>

> -Joel

>

> ________________________________________

> From: Derek Leach [community-noreply@qnx.com]

> Sent: Tuesday, August 26, 2014 8:29 AM

> To: 'advanced-graphics@community.qnx.com'

> Subject: RE: Address of frame buffer

>

> http://www.qnx.com/developers/docs/6.4.0/gf/dev_guide/api/gf_surface_get_info.

> html

>

> vaddr in the info structure is the mapped in framebuffer of the provided

> surface.

>

> -----Original Message-----

> From: Radu Stancu [mailto:community-noreply@qnx.com]

> Sent: August-26-14 4:49 AM

> To: advanced-graphics

> Subject: Address of frame buffer

>

> Hello,

>

> Is there any way to find the address and size of the system frame buffer? I

> have an application that must save the current content of the screen, use

> hardware encoding to turn it to jpeg and send it via ethernet to another

> device where it is decoded and shown on screen.

>

> From my research, in order to do the above, I would need somehow to access the

>  frame buffer, copy it into a local buffer, apply some DCE functions and

> obtain the jpeg in a file I can then send. But I haven't found any reference

> to the actual address where the buffer is stored (if I know this, I can easily

>  use memcpy to access it).

>

> As a final note, I don't want to make a screen shot (this was already done in

> a previous version of my app, one without hardware encoding), but actually

> work with the raw frame buffer. Actual code or functions  would be of use, I'm

>  trying not to use other apps that may slow my device down (I'm working under

> strict cpu limitations).

>

> Any info would be appreciated.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111545

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111547

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post111551

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post111552: RE: RE: Address of frame buffer

2014-08-26T14:32:55Z

Posting to open Qt forum might provide a quick turnaround.  I would think they would have a Qt way of doing this.

-----Original Message-----

From: Radu Stancu [mailto:community-noreply@qnx.com]

Sent: August-26-14 10:17 AM

To: advanced-graphics

Subject: Re: RE: Address of frame buffer

I'm working on an embedded system for automotive that has omap 5 and qnx on it. My application is written in Qt 5.2  (I render some objects, like cubes and such on 2 screens). The main thing I need to do is somehow capture a frame on one screen and then proceed as I mentioned in the original post. The framebuffer approach was the first that came to mind, as I need something raw for dce encode. I don't really need a solution in Qt, a separate program written in C would work as well, . But I'm kinda lost, because I'm not that familiar with qnx.

> Perhaps you could mention which framework you're using? Derek's likely

> assuming GF because you're posting in the advance graphics forum, but

> you don't actually provide any hints of the framework.

>

> -Joel

>

> ________________________________________

> From: Derek Leach [community-noreply@qnx.com]

> Sent: Tuesday, August 26, 2014 8:29 AM

> To: 'advanced-graphics@community.qnx.com'

> Subject: RE: Address of frame buffer

>

> http://www.qnx.com/developers/docs/6.4.0/gf/dev_guide/api/gf_surface_get_info.

> html

>

> vaddr in the info structure is the mapped in framebuffer of the

> provided surface.

>

> -----Original Message-----

> From: Radu Stancu [mailto:community-noreply@qnx.com]

> Sent: August-26-14 4:49 AM

> To: advanced-graphics

> Subject: Address of frame buffer

>

> Hello,

>

> Is there any way to find the address and size of the system frame

> buffer? I have an application that must save the current content of

> the screen, use hardware encoding to turn it to jpeg and send it via

> ethernet to another device where it is decoded and shown on screen.

>

> From my research, in order to do the above, I would need somehow to

> access the  frame buffer, copy it into a local buffer, apply some DCE

> functions and obtain the jpeg in a file I can then send. But I haven't

> found any reference to the actual address where the buffer is stored

> (if I know this, I can easily  use memcpy to access it).

>

> As a final note, I don't want to make a screen shot (this was already

> done in a previous version of my app, one without hardware encoding),

> but actually work with the raw frame buffer. Actual code or functions

> would be of use, I'm  trying not to use other apps that may slow my

> device down (I'm working under strict cpu limitations).

>

> Any info would be appreciated.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111545

> To cancel your subscription to this discussion, please e-mail

> advanced- graphics-unsubscribe@community.qnx.com

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111547

> To cancel your subscription to this discussion, please e-mail

> advanced- graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post111551

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post111551: Re: RE: Address of frame buffer

2014-08-26T14:16:31Z

I'm working on an embedded system for automotive that has omap 5 and qnx on it. My application is written in Qt 5.2  (I render some objects, like cubes and such on 2 screens). The main thing I need to do is somehow capture a frame on one screen and then proceed as I mentioned in the original post. The framebuffer approach was the first that came to mind, as I need something raw for dce encode. I don't really need a solution in Qt, a separate program written in C would work as well, . But I'm kinda lost, because I'm not that familiar with qnx.

> Perhaps you could mention which framework you're using? Derek's likely

> assuming GF because you're posting in the advance graphics forum, but you

> don't actually provide any hints of the framework.

>

> -Joel

>

> ________________________________________

> From: Derek Leach [community-noreply@qnx.com]

> Sent: Tuesday, August 26, 2014 8:29 AM

> To: 'advanced-graphics@community.qnx.com'

> Subject: RE: Address of frame buffer

>

> http://www.qnx.com/developers/docs/6.4.0/gf/dev_guide/api/gf_surface_get_info.

> html

>

> vaddr in the info structure is the mapped in framebuffer of the provided

> surface.

>

> -----Original Message-----

> From: Radu Stancu [mailto:community-noreply@qnx.com]

> Sent: August-26-14 4:49 AM

> To: advanced-graphics

> Subject: Address of frame buffer

>

> Hello,

>

> Is there any way to find the address and size of the system frame buffer? I

> have an application that must save the current content of the screen, use

> hardware encoding to turn it to jpeg and send it via ethernet to another

> device where it is decoded and shown on screen.

>

> From my research, in order to do the above, I would need somehow to access the

>  frame buffer, copy it into a local buffer, apply some DCE functions and

> obtain the jpeg in a file I can then send. But I haven't found any reference

> to the actual address where the buffer is stored (if I know this, I can easily

>  use memcpy to access it).

>

> As a final note, I don't want to make a screen shot (this was already done in

> a previous version of my app, one without hardware encoding), but actually

> work with the raw frame buffer. Actual code or functions  would be of use, I'm

>  trying not to use other apps that may slow my device down (I'm working under

> strict cpu limitations).

>

> Any info would be appreciated.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111545

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post111547

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

post111550: RE: Address of frame buffer

2014-08-26T14:02:08Z

Perhaps you could mention which framework you're using? Derek's likely assuming GF because you're posting in the advance graphics forum, but you don't actually provide any hints of the framework.

-Joel

________________________________________

From: Derek Leach [community-noreply@qnx.com]

Sent: Tuesday, August 26, 2014 8:29 AM

To: 'advanced-graphics@community.qnx.com'

Subject: RE: Address of frame buffer

http://www.qnx.com/developers/docs/6.4.0/gf/dev_guide/api/gf_surface_get_info.html

vaddr in the info structure is the mapped in framebuffer of the provided surface.

-----Original Message-----

From: Radu Stancu [mailto:community-noreply@qnx.com]

Sent: August-26-14 4:49 AM

To: advanced-graphics

Subject: Address of frame buffer

Hello,

Is there any way to find the address and size of the system frame buffer? I have an application that must save the current content of the screen, use hardware encoding to turn it to jpeg and send it via ethernet to another device where it is decoded and shown on screen.

From my research, in order to do the above, I would need somehow to access the frame buffer, copy it into a local buffer, apply some DCE functions and obtain the jpeg in a file I can then send. But I haven't found any reference to the actual address where the buffer is stored (if I know this, I can easily use memcpy to access it).

As a final note, I don't want to make a screen shot (this was already done in a previous version of my app, one without hardware encoding), but actually work with the raw frame buffer. Actual code or functions would be of use, I'm trying not to use other apps that may slow my device down (I'm working under strict cpu limitations).

Any info would be appreciated.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post111545

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post111547

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post111547: RE: Address of frame buffer

2014-08-26T12:29:47Z

http://www.qnx.com/developers/docs/6.4.0/gf/dev_guide/api/gf_surface_get_info.html

vaddr in the info structure is the mapped in framebuffer of the provided surface.

-----Original Message-----

From: Radu Stancu [mailto:community-noreply@qnx.com]

Sent: August-26-14 4:49 AM

To: advanced-graphics

Subject: Address of frame buffer

Hello,

Is there any way to find the address and size of the system frame buffer? I have an application that must save the current content of the screen, use hardware encoding to turn it to jpeg and send it via ethernet to another device where it is decoded and shown on screen.

From my research, in order to do the above, I would need somehow to access the frame buffer, copy it into a local buffer, apply some DCE functions and obtain the jpeg in a file I can then send. But I haven't found any reference to the actual address where the buffer is stored (if I know this, I can easily use memcpy to access it).

As a final note, I don't want to make a screen shot (this was already done in a previous version of my app, one without hardware encoding), but actually work with the raw frame buffer. Actual code or functions would be of use, I'm trying not to use other apps that may slow my device down (I'm working under strict cpu limitations).

Any info would be appreciated.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post111545

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post111545: Address of frame buffer

2014-08-26T08:48:59Z

Hello,

Is there any way to find the address and size of the system frame buffer? I have an application that must save the current content of the screen, use hardware encoding to turn it to jpeg and send it via ethernet to another device where it is decoded and shown on screen.

From my research, in order to do the above, I would need somehow to access the frame buffer, copy it into a local buffer, apply some DCE functions and obtain the jpeg in a file I can then send. But I haven't found any reference to the actual address where the buffer is stored (if I know this, I can easily use memcpy to access it).

As a final note, I don't want to make a screen shot (this was already done in a previous version of my app, one without hardware encoding), but actually work with the raw frame buffer. Actual code or functions would be of use, I'm trying not to use other apps that may slow my device down (I'm working under strict cpu limitations).

Any info would be appreciated.

post110576: Re: How run Screen Application

2014-06-04T08:26:14Z

Thanks! It works!

post110574: Re: How run Screen Application

2014-06-04T08:12:30Z

Hi,

You can use shell script like the following one:

---

#!/bin/sh

export GRAPHICS_ROOT=/usr/lib/graphics/vmware

screen

gles1-gears

---

Respectfully,

Oleg

04 июня 2014 г., в 10:46:51, Bartek P <community-noreply@qnx.com> написал:

> Hi,

>

> OK, I add this variable and screen running properly. But, how I can starting example app in screen?

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post110572

> To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post110572: Re: How run Screen Application

2014-06-04T06:46:51Z

Hi,

OK, I add this variable and screen running properly. But, how I can starting example app in screen?

post110566: Re: How run Screen Application

2014-06-03T17:30:08Z

Hi Bartek,

Try to declare environment variable GRAPHICS_ROOT before launch screen server:

export GRAPHICS_ROOT=/usr/lib/graphics/vmware

Respectfully,

Oleg

03 июня 2014 г., в 17:25:14, Bartek P <community-noreply@qnx.com> написал:

> Hello,

>

> How can I run application screen application in QNX 6.6?

> I want to start example "display_image" from "/usr/bin", and when start this app i have a massage:

> "screen_context_create: No such file or directory".

> What is command "screen" in QNX? When I call this command I receive "Memory fault"

>

> My host target is vmware.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post110563

> To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post110563: How run Screen Application

2014-06-03T13:25:14Z

Hello,

How can I run application screen application in QNX 6.6?

I want to start example "display_image" from "/usr/bin", and when start this app i have a massage:

"screen_context_create: No such file or directory".

What is command "screen" in QNX? When I call this command I receive "Memory fault"

My host target is vmware.

post110149: load and display an image using the 'libimg' library and the "QNX Screen" API.

2014-05-05T17:07:56Z

Hi,

I am a beginner in leaning QNX. I just have an assignment to load and display an image using the image library "libimg" and the "QNX Screen" API.

I created my code based on the sample code given in http://support7.qnx.com/download/downlo ... _Guide.pdf, also the sample code given in the post https://github.com/blackberry/NDK-Sampl ... ure/main.c. My code is below (omit all print code for debugging):

static img_t img;

screen_context_t screen_context;

screen_window_t screen_window;

int main(int argc, char * const argv[]) {

screen_create_context(&screen_context, SCREEN_APPLICATION_CONTEXT);

screen_create_window(&screen_window, screen_context);

int count = 0;

screen_get_context_property_iv(screen_context,SCREEN_PROPERTY_DISPLAY_COUNT, &count);

screen_display_t *screen_disps = calloc(count, sizeof(screen_display_t));

screen_get_context_property_pv(screen_context, SCREEN_PROPERTY_DISPLAYS, (void **) screen_disps);

screen_display_t screen_disp = screen_disps[0];

free(screen_disps);

int vis = 1;

int dims[2] = { 512, 512 };

int rect[4] = { 0, 0, 1, 1 };

screen_set_window_property_iv(screen_window, SCREEN_PROPERTY_VISIBLE, &vis);

screen_set_window_property_iv(screen_window, SCREEN_PROPERTY_BUFFER_SIZE,rect + 2);

rev = screen_set_window_property_iv(screen_window,SCREEN_PROPERTY_SOURCE_SIZE, dims);

screen_buffer_t screen_buf;

screen_create_window_buffers(screen_window, 1);

screen_get_window_property_pv(screen_window, SCREEN_PROPERTY_RENDER_BUFFERS, (void **)&screen_buf);

screen_get_buffer_property_pv(screen_buf, SCREEN_PROPERTY_POINTER, (void **)&img.access.direct.data);

screen_get_buffer_property_iv(screen_buf, SCREEN_PROPERTY_STRIDE, (int *)&img.access.direct.stride);

img.flags |= IMG_DIRECT;

loadImg();

screen_post_window(screen_window, screen_buf, 1, rect, 0);

}

static void loadImg(void) {

img_lib_t ilib = NULL;

int rc1 = 0;

img_lib_attach(&ilib);

img.w = 512;

img.flags |= IMG_W;

img.h = 512;

img.flags |= IMG_H;

rc1 == img_load_file(ilib,filename , NULL, &img);

img_lib_detach(ilib);

}

The debug info shows that loading image is successful, however what I can see is a 512*512 white image on my screen. I don't know what is the problem, hope I can get some help here.

Thank you very much.

Lila

post107536: Re: FlashLite 4 / Actionscript 3

2013-12-13T14:59:00Z

The information on that link is incorrect, and we have already asked our IT department to correct it.

FlashLite4 / AS3 is not supported.

post107335: Re: FlashLite 4 / Actionscript 3

2013-12-10T09:38:20Z

> I was under the impression that QNX 6.5 HMI Technology provided the ability to

> execute FlashLite 4.

>

> I believe the following link advertises this: http://www.qnx.com/products/hmi/

> hmi.html

>

> I have developed some entertainment applications in AS2 that run in the 6.4

> environment and I'm considering developing in AS3 for 6.5.

>

> When can I expect 6.5 to support Flash AS3 based applications? Will support be

> released as a patch to 6.5 or will it be rolled out in a future version of

> QNX?

Now, Is that Release?

I can't find AS3 Information, and FlashLite4 also.

Answer please..

post106310: Re: Image searching within another image

2013-10-28T13:49:45Z

Sorry but I cannot. Not familiar with the NDK and I'd suggest you ask for BlackBerry specific information on the BlackBerry forums.

You will have to do your homework but I know that most people's gut reaction when faced with similar problems is to say "OpenCV!" I do not know if it has something that solves your particular problem out of the box.

There appears to be an "Officially sanctioned port of OpenCV to the Blackberry/QNX platform" at the following location: https://github.com/blackberry/OpenCV/

If you are after a bit by bit perfect match and don't care about the CPU use then it should be fastest to check how to access the frame via the image library and write your own matcher with for(x0 on big image) for(y0 on big image) for(x on small image) for(y on small image).

If you need approximate matching, need robustness for scale/rotation/other transformations or just need more efficiency then do your research on existing libraries.

--

Mate

On Oct 28, 2013, at 10:23 PM, Dinyo Haygarov <community-noreply@qnx.com<mailto:community-noreply@qnx.com>>

wrote:

Can you point me to an existing image recognition library in the NDK which I can use?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post106308

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post106308: Re: Image searching within another image

2013-10-28T13:23:58Z

Can you point me to an existing image recognition library in the NDK which I can use?

post106307: Re: Image searching within another image

2013-10-28T12:04:33Z

It definitely is possible but QNX does not provide any libraries for that.

Depending on countless factors you may either read in both images using existing API-s and do the search yourself along the lines of a 2D strstr or rely on an image recognition library.

If you go with the roll your own solution approach but you hit performance issues then you can subsample both images multiple times (using a factor of 2 can be implemented most efficiently) resulting in an image pyramid and do the search at higher resolutions only in areas that gave a hit in lower resolutions.

--

Mate

On Oct 28, 2013, at 8:23 PM, Dinyo Haygarov <community-noreply@qnx.com>

wrote:

> is it possible to search for a certain image within another bigger image and retrieve it's coordinates within the bigger image?

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post106298

> To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post106298: Image searching within another image

2013-10-28T11:23:58Z

is it possible to search for a certain image within another bigger image and retrieve it's coordinates within the bigger image?

post106140: RE: How can I use opencv under QNX?

2013-10-22T13:56:52Z

OpenCV has not been ported to QNX. You could check out the source to OpenCV and attempt to port it yourself.

-----Original Message-----

From: robort smith [mailto:community-noreply@qnx.com]

Sent: October-22-13 6:28 AM

To: advanced-graphics

Subject: How can I use opencv under QNX?

How can I use opencv under QNX?

Now I am trying to use opencv under QNX to trace some objects. While opencv does't support QNX . Is there any thing I can do for this?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post106134

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post106134: How can I use opencv under QNX?

2013-10-22T10:27:43Z

How can I use opencv under QNX?

Now I am trying to use opencv under QNX to trace some objects. While opencv does't support QNX . Is there any thing I

can do for this?

post105268: How to deal with MJPEG foramt video?

2013-09-20T08:47:48Z

Now I am writing USB camera driver.I have already get single frames successfully.At current,The problems are that I don't know how to deal with these frames,and how to show these frames on screen.

post104263: memory leak in PxLoadImage (when repeating)

2013-08-18T10:20:57Z

It isn't new topic, but in prevoius discussion I haven't found any real decision of my problem. So, I use PxLoadImage many times during my program session. If it is, for instance, the 300-th call of PxLoadImage I have memory fault. I understand that problem is connected with memory leaks. But I don't know how to free memory properly. How must I release memory?

My code:

PhImage_t *temp_image= PhCreateImage( 0, 256 , 256 , Pg_IMAGE_DIRECT_888, 0, 0, 1);

if ((temp_image=PxLoadImage(file_name,NULL))==NULL){

fprintf(stderr,"Error opening next path: %s\n",file_name);

PtExit(EXIT_FAILURE);

...

temp_image->flags=Ph_RELEASE_IMAGE|Ph_RELEASE_PALETTE| Ph_RELEASE_TRANSPARENCY_MASK|Ph_RELEASE_GHOST_BITMAP|Ph_RELEASE_IMAGE_ALL;

PhReleaseImage(temp_image);

free(temp_image);

post103731: Re: Problem for gf-calib

2013-07-30T03:12:10Z

please reference the QNX650\target\qnx6\etc\system\config\img.conf.This is helpful.

post103615: Re: Problem for gf-calib

2013-07-26T02:02:08Z

I met the same problem。

# gf-calib -D -f /fs/mmc/config/calib.localhost

Number of layers: 1

Resolution: 800x480, refresh = 60Hz

layer_idx=0

img_load(target1): errno = 3

load_image2surf(): failed

Do you solve this problem?

post101196: gf_dev_attach() failed with GF_ERR_IODISPLAY

2013-05-06T23:37:00Z

I am trying to get Qt and QNX running on the Sabre Lite board. I have a Qt project (just a single button) and the project builds and loads to the target using the Momentics debugger. However, the first line of code that executes:

QApplication a(argc, argv, global);

Causes this error:

QQnxScreen: gf_dev_attach() failed with error code 2

The error in the code says this:

#define GF_ERR_IODISPLAY 0x00000002 /* error accessing io-display */

My board does not have an io-display process, and there is no /dev/io-display directory. The QNX demo that is part of the QNX accelerator kit quick start runs correctly and no io-display process is running.

How do I get past this error? Do I need an io-display process running?

Thanks,

Dave

post101156: 答复: How to show a PNG image with transparent background

2013-05-06T02:53:20Z

Dear Sir.

I found my PNG file is a 32bit PNG, and then I change it to 8bit and 24bit,

but it still can not be transparented.

Thanks and Best Regards!

*******************************************

金狄

天元鼎正电子科技（北京）有限公司

北京市海淀区上地七街1号汇众大厦106

邮编：100085

电话：010-62970815

手机：18601122606

*******************************************

-----邮件原件-----

发件人: Derek Leach [mailto:community-noreply@qnx.com]

发送时间: 2013年5月3日 21:27

收件人: advanced-graphics

主题: RE: How to show a PNG image with transparent background

What kind of PNG, 32-bit?

-----Original Message-----

From: Di Jin [mailto:community-noreply@qnx.com]

Sent: May-03-13 4:47 AM

To: advanced-graphics

Subject: How to show a PNG image with transparent background

I have a PNG image. I modify the background of image to transparency in the

Photoshop. I write the code like this:

    if (img.flags & IMG_TRANSPARENCY) {

            /* we can handle transparency in GF using chroma */

            gf_chroma_t        chroma;

            memset(&chroma, 0, sizeof chroma);

            chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

            if (img.format & IMG_FMT_PALETTE) {

                    chroma.color0 = img.palette[img.transparency.index];

            } else if (IMG_FMT_BPP(img.format) < 24) {

                    chroma.color0 = img.transparency.rgb16;

            } else {

                    chroma.color0 = img.transparency.rgb32;

            }

            gf_context_set_chroma(*p_context, &chroma);

    }

    if (img.format & IMG_FMT_ALPHA) {

            gf_alpha_t        alpha;

            memset(&alpha, 0, sizeof alpha);

            alpha.mode = GF_ALPHA_M1_SRC_PIXEL_ALPHA | GF_BLEND_SRC_M1 |

                GF_BLEND_DST_1mM1;

            gf_context_set_alpha(*p_context, &alpha);

    }

    // blit the image onto a target surface

    gf_draw_blit2(*p_context, *p_surface, NULL, 0, 0, img.w - 1, img.h - 1,

0, 0);

I found   (img.flags & IMG_TRANSPARENCY) this branch always nagative. But my

Image has transparent background. How shold I do?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post101108

To cancel your subscription to this discussion, please e-mail

advanced-graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post101114

To cancel your subscription to this discussion, please e-mail

advanced-graphics-unsubscribe@community.qnx.com

post101133: multiple layers-hardware

2013-05-03T16:58:56Z

Hi there,

If I am trying to display using 2 layers (2 processes), both using gf, but I found my hardware only had one layer. My question will be: how can I get the hardware which supports multiple layer? If I am doing it under vmware, what is the proper driver to use?

thanks,

Lu

post101119: 答复: How to show a PNG image with transparent background

2013-05-03T14:17:48Z

Dear,Sir.

Please check the attaching file.

I generate it by Fireworks.

Thanks and Best Regards!

*******************************************

金狄

天元鼎正电子科技（北京）有限公司

北京市海淀区上地七街1号汇众大厦106

邮编：100085

电话：010-62970815

手机：18601122606

*******************************************

-----邮件原件-----

发件人: Derek Leach [mailto:community-noreply@qnx.com]

发送时间: 2013年5月3日 21:27

收件人: advanced-graphics

主题: RE: How to show a PNG image with transparent background

What kind of PNG, 32-bit?

-----Original Message-----

From: Di Jin [mailto:community-noreply@qnx.com]

Sent: May-03-13 4:47 AM

To: advanced-graphics

Subject: How to show a PNG image with transparent background

I have a PNG image. I modify the background of image to transparency in the

Photoshop. I write the code like this:

    if (img.flags & IMG_TRANSPARENCY) {

            /* we can handle transparency in GF using chroma */

            gf_chroma_t        chroma;

            memset(&chroma, 0, sizeof chroma);

            chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

            if (img.format & IMG_FMT_PALETTE) {

                    chroma.color0 = img.palette[img.transparency.index];

            } else if (IMG_FMT_BPP(img.format) < 24) {

                    chroma.color0 = img.transparency.rgb16;

            } else {

                    chroma.color0 = img.transparency.rgb32;

            }

            gf_context_set_chroma(*p_context, &chroma);

    }

    if (img.format & IMG_FMT_ALPHA) {

            gf_alpha_t        alpha;

            memset(&alpha, 0, sizeof alpha);

            alpha.mode = GF_ALPHA_M1_SRC_PIXEL_ALPHA | GF_BLEND_SRC_M1 |

                GF_BLEND_DST_1mM1;

            gf_context_set_alpha(*p_context, &alpha);

    }

    // blit the image onto a target surface

    gf_draw_blit2(*p_context, *p_surface, NULL, 0, 0, img.w - 1, img.h - 1,

0, 0);

I found   (img.flags & IMG_TRANSPARENCY) this branch always nagative. But my

Image has transparent background. How shold I do?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post101108

To cancel your subscription to this discussion, please e-mail

advanced-graphics-unsubscribe@community.qnx.com

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post101114

To cancel your subscription to this discussion, please e-mail

advanced-graphics-unsubscribe@community.qnx.com

post101114: RE: How to show a PNG image with transparent background

2013-05-03T13:27:01Z

What kind of PNG, 32-bit?

-----Original Message-----

From: Di Jin [mailto:community-noreply@qnx.com]

Sent: May-03-13 4:47 AM

To: advanced-graphics

Subject: How to show a PNG image with transparent background

I have a PNG image. I modify the background of image to transparency in the Photoshop. I write the code like this:

    if (img.flags & IMG_TRANSPARENCY) {

            /* we can handle transparency in GF using chroma */

            gf_chroma_t        chroma;

            memset(&chroma, 0, sizeof chroma);

            chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

            if (img.format & IMG_FMT_PALETTE) {

                    chroma.color0 = img.palette[img.transparency.index];

            } else if (IMG_FMT_BPP(img.format) < 24) {

                    chroma.color0 = img.transparency.rgb16;

            } else {

                    chroma.color0 = img.transparency.rgb32;

            }

            gf_context_set_chroma(*p_context, &chroma);

    }

    if (img.format & IMG_FMT_ALPHA) {

            gf_alpha_t        alpha;

            memset(&alpha, 0, sizeof alpha);

            alpha.mode = GF_ALPHA_M1_SRC_PIXEL_ALPHA | GF_BLEND_SRC_M1 |

                GF_BLEND_DST_1mM1;

            gf_context_set_alpha(*p_context, &alpha);

    }

    // blit the image onto a target surface

    gf_draw_blit2(*p_context, *p_surface, NULL, 0, 0, img.w - 1, img.h - 1, 0, 0);

I found   (img.flags & IMG_TRANSPARENCY) this branch always nagative. But my Image has transparent background. How shold I do?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post101108

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post101108: How to show a PNG image with transparent background

2013-05-03T08:47:29Z

I have a PNG image. I modify the background of image to transparency in the Photoshop. I write the code like this:

    if (img.flags & IMG_TRANSPARENCY) {

            /* we can handle transparency in GF using chroma */

            gf_chroma_t        chroma;

            memset(&chroma, 0, sizeof chroma);

            chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

            if (img.format & IMG_FMT_PALETTE) {

                    chroma.color0 = img.palette[img.transparency.index];

            } else if (IMG_FMT_BPP(img.format) < 24) {

                    chroma.color0 = img.transparency.rgb16;

            } else {

                    chroma.color0 = img.transparency.rgb32;

            }

            gf_context_set_chroma(*p_context, &chroma);

    }

    if (img.format & IMG_FMT_ALPHA) {

            gf_alpha_t        alpha;

            memset(&alpha, 0, sizeof alpha);

            alpha.mode = GF_ALPHA_M1_SRC_PIXEL_ALPHA | GF_BLEND_SRC_M1 |

                GF_BLEND_DST_1mM1;

            gf_context_set_alpha(*p_context, &alpha);

    }

    // blit the image onto a target surface

    gf_draw_blit2(*p_context, *p_surface, NULL, 0, 0, img.w - 1, img.h - 1, 0, 0);

I found   (img.flags & IMG_TRANSPARENCY) this branch always nagative. But my Image has transparent background. How shold I do?

post100015: IMX28: writing directly to a layer surface

2013-03-19T20:10:28Z

I created a 24 BPP surface using gf_surface_create_layer, specifying CPU_LINEARLY_ACCESSIBLE, and verified that the surface does indeed have this flag set in the surface_info_t.

However, when I try and write to the surface, the access does not appear to be linear. It seems as though there is a swizzle or address redirection occurring. Drawing to a 32 bit surface and then using gf_draw_blit to blit to the 24 BPP layer_surface has the same issue.

Is it possible to write linearly (with the CPU) to a displayable surface on this chipset?

post99705: Re: Problem for gf-calib

2013-03-06T14:08:15Z

I got the same problem:

# gf-calib -D -f /tmp/calib.conf

Number of layers: 1

Resolution: 800x480, refresh = 60Hz

layer_idx=0

img_load(target1): errno = 3

load_image2surf(): failed

switch_lcd_on

switch_lcd_off

could any guys help me

post99336: how to display mouse cursor under non gf frameworks (ex: screen / libscreen.so )

2013-02-19T07:36:59Z

Hi all,

I am not sure where should I post this question.

Is there any simliar config file (such as /etc/system/config/display.conf, cursor=hardware) for screen.so or screen process on QNX CAR 2 ROM to show the mouse cursor if I use USB mouse instead of touch panel?

Regards,

Chinlin

post99320: Re: RE: Line dashing

2013-02-15T19:58:20Z

On my hardware (devg-i830.so) dashed line can't be drawn if width of line is set to 1.

If width of line is bigger than 1 then everything works ok.

Also looks like that my hardware handles line_repeat qual to 32 only.

Is this limitation common to all devices supported via the same driver in this case devg-i830.so ?

> gf_point_t p[] = { { 50,250},{w-100,250}};

> gf_context_set_fgcolor(context,0x00ffff);

> gf_context_set_bgcolor(context,0xffffff);

> gf_context_set_linedash(context,0x000ffff, 0, 32,

> GF_CONTEXT_LINEDASH_BACKFILL);

> gf_context_set_penwidth( context,10);

> gf_draw_polyline(context,p,2,0);

>

> -----Original Message-----

> From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

> Sent: February-14-13 2:55 PM

> To: advanced-graphics

> Subject: Re: Line dashing

>

> I need to draw dotted and dashed line.

>

> Can anybody explain how to use the gf_context_set_linedash() for such purpose

> ?

> Is it possible to draw a dashed line on the patchy background ?

>

> > Hi,

> >

> > Example from documentation "Setting line attributes" draws a solid

> > line, even if I play with function parameters.

> >

> > What is a proper usage of the gf_context_set_linedash() function ?

> >

> > Thank's,

> > Jacek

>

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post99297

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

post99298: RE: Line dashing

2013-02-14T20:02:01Z

gf_point_t p[] = { { 50,250},{w-100,250}};

gf_context_set_fgcolor(context,0x00ffff);

gf_context_set_bgcolor(context,0xffffff);

gf_context_set_linedash(context,0x000ffff, 0, 32, GF_CONTEXT_LINEDASH_BACKFILL);

gf_context_set_penwidth( context,10);

gf_draw_polyline(context,p,2,0);

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: February-14-13 2:55 PM

To: advanced-graphics

Subject: Re: Line dashing

I need to draw dotted and dashed line.

Can anybody explain how to use the gf_context_set_linedash() for such purpose ?

Is it possible to draw a dashed line on the patchy background ?

> Hi,

>

> Example from documentation "Setting line attributes" draws a solid

> line, even if I play with function parameters.

>

> What is a proper usage of the gf_context_set_linedash() function ?

>

> Thank's,

> Jacek

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post99297

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post99297: Re: Line dashing

2013-02-14T19:54:57Z

I need to draw dotted and dashed line.

Can anybody explain how to use the gf_context_set_linedash() for such purpose ?

Is it possible to draw a dashed line on the patchy background ?

> Hi,

>

> Example from documentation "Setting line attributes" draws a solid line,

> even if I play with function parameters.

>

> What is a proper usage of the gf_context_set_linedash() function ?

>

> Thank's,

> Jacek

post99210: Line dashing

2013-02-11T20:16:01Z

Hi,

Example from documentation "Setting line attributes" draws a solid line,

even if I play with function parameters.

What is a proper usage of the gf_context_set_linedash() function ?

Thank's,

Jacek

post99209: RE: Drawing a cicle

2013-02-11T20:15:44Z

You are correct, there is no 'draw circle' in GF. The original reason was that hardware did not generally implement such a request, and GF only generally mirrors hardware functionality. Photon supports drawing of circles, but at the lowest level, it is all broken down to many span draws.

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: February-11-13 3:12 PM

To: advanced-graphics

Subject: Re: Drawing a cicle

The GF library doesn't include function for drawing a cricle.

Looks like that I have to implement such routine itself.

> How to draw a circle ?

> Is it possible ?

>

> Jacek

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post99208

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post99208: Re: Drawing a cicle

2013-02-11T20:11:45Z

The GF library doesn't include function for drawing a cricle.

Looks like that I have to implement such routine itself.

> How to draw a circle ?

> Is it possible ?

>

> Jacek

post99206: Re: RE: RE: chroma operation increasing CPU hog

2013-02-11T20:05:57Z

Ok, so I have to look for the another solution...

> > I set chroma as follows:

> > chroma.mode = GF_CHROMA_OP_DST_MATCH | GF_CHROMA_OP_NO_DRAW;

>

> Sorry, but it is unaccelerated. Only SRC_MATCH | NO_DRAW is accelereted.

post99196: Re: RE: RE: chroma operation increasing CPU hog

2013-02-11T07:00:49Z

> I set chroma as follows:

> chroma.mode = GF_CHROMA_OP_DST_MATCH | GF_CHROMA_OP_NO_DRAW;

Sorry, but it is unaccelerated. Only SRC_MATCH | NO_DRAW is accelereted.

post99092: Drawing a cicle

2013-02-05T19:02:18Z

How to draw a circle ?

Is it possible ?

Jacek

post99049: RE: Layers - code example

2013-02-04T21:48:10Z

Are the pixel formats of the source and destination surfaces different?

Are planemask or rops enabled during this operation?

Are either of the surfaces allocated for CPU fast access?

Any of these will force chroma blit into CPU software mode ...

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: February-04-13 4:35 PM

To: advanced-graphics

Subject: Layers - code example

Hi,

Is there any example which shows how to properly use layers ?

Thank's

Jacek

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post99048

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post99048: Layers - code example

2013-02-04T21:34:51Z

Hi,

Is there any example which shows how to properly use layers ?

Thank's

Jacek

post99043: Re: RE: RE: chroma operation increasing CPU hog

2013-02-04T16:45:28Z

I set chroma as follows:

chroma.mode = GF_CHROMA_OP_DST_MATCH | GF_CHROMA_OP_NO_DRAW;

> Ok, there is a hw accel for certain chroma operations, but not all, what

> chroma flags are you setting?

>

> -----Original Message-----

> From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

> Sent: February-04-13 11:38 AM

> To: advanced-graphics

> Subject: Re: RE: chroma operation increasing CPU hog

>

> devg-i830

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post99041

> To cancel your subscription to this discussion, please e-mail advanced-

> graphics-unsubscribe@community.qnx.com

post99042: RE: RE: chroma operation increasing CPU hog

2013-02-04T16:40:58Z

Ok, there is a hw accel for certain chroma operations, but not all, what chroma flags are you setting?

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: February-04-13 11:38 AM

To: advanced-graphics

Subject: Re: RE: chroma operation increasing CPU hog

devg-i830

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post99041

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post99041: Re: RE: chroma operation increasing CPU hog

2013-02-04T16:37:56Z

devg-i830

post99032: RE: chroma operation increasing CPU hog

2013-02-04T13:50:19Z

What devg driver are you using?

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: February-03-13 2:54 PM

To: advanced-graphics

Subject: chroma operation increasing CPU hog

Hi,

In application respnsible for displaying video stream the gf_draw_blit2() is called very often.

When I set chroma via gf_context_set_chroma() then the following behavior is observed:

- increasing CPU hog,

- corrupted picture,

- picture is not displayed in real time.

I need to set chroma keying in order to not display video data on particular color.

Am I doing something wrong or my hardware is too slow for such usage ?

Application is tested with devg-i830.so driver.

Regards,

Jacek

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post99026

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post99026: chroma operation increasing CPU hog

2013-02-03T19:53:33Z

Hi,

In application respnsible for displaying video stream the gf_draw_blit2() is called very often.

When I set chroma via gf_context_set_chroma() then the following behavior is observed:

- increasing CPU hog,

- corrupted picture,

- picture is not displayed in real time.

I need to set chroma keying in order to not display video data on particular color.

Am I doing something wrong or my hardware is too slow for such usage ?

Application is tested with devg-i830.so driver.

Regards,

Jacek

post99006: Two layers - GF & Photon

2013-02-01T19:17:58Z

Hello,

I have application (hybrid of GF and Photon) which is responsible for displaying video stream.

Everything works fine and data is displayed in layer 0 which is shared with Photon.

I need to draw some lines and text on the top of video stream.

Idea is as follows:

Layer0: Photon application (lines, text, some other components, widgets, ...),

Layer1: GF+Photon (video stream).

I don't know if this is a right direction.

Regards,

Jacek

post98343: Re: Problem for gf-calib

2013-01-02T16:43:51Z

Try running your application with "LD_DEBUG=1 {your-app}". You should see which library your are missing.

post98313: Re: Problem for gf-calib

2012-12-24T06:29:35Z

Afetr including img.conf and img_codec so's i am getting error as

Number of layers: 3

Resolution: 720x400, refresh = 60Hz

layer_idx=0

img_load(target1): errno = 3

load_image2surf(): failed

Please help.

post97623: RE: RE: RE: RE: gf_layer_attach() failed !!!

2012-11-29T15:20:02Z

Please contact tech support for this level of support.

-----Original Message-----

From: Ashwini Dubey [mailto:community-noreply@qnx.com]

Sent: November-29-12 10:20 AM

To: advanced-graphics

Subject: Re: RE: RE: RE: gf_layer_attach() failed !!!

Can you please provide example of working GF drawing code.

Do let me also know how to toggle between GUI and text mode.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post97622

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post97622: Re: RE: RE: RE: gf_layer_attach() failed !!!

2012-11-29T15:19:43Z

Can you please provide example of working GF drawing code.

Do let me also know how to toggle between GUI and text mode.

post97615: RE: RE: RE: gf_layer_attach() failed !!!

2012-11-29T14:01:59Z

If you want to use Photon, ypu must use Pg photon drawing lines, please see helpviewer in Photon.

If you want to use GF application, then you must exit Photon to text mode, then to make it easy, telnet into the QNX box, then run the GF application.

-----Original Message-----

From: Ashwini Dubey [mailto:community-noreply@qnx.com]

Sent: November-29-12 3:16 AM

To: advanced-graphics

Subject: Re: RE: RE: gf_layer_attach() failed !!!

What else is running on the machine?

I have a default installation and basic GUI is running.

…......

Is Photon running?

Yes.

176152   1 /photon/bin/Photon  10r RECEIVE     1

…......

Number of displays: 1

Display 0: 1024X768, refresh = 60Hz

Number of layers: 1

…......

I am very new to QNX, but I am experienced in linux application and kernel development.

Currently I am using intel HD Graphics, PCI device, Vendor ID:8086, Device ID:0152.

….

please suggest what I need to draw a rectangle or circle or line.

Below is list of all the process running. Result of pidin command.

     pid tid name               prio STATE       Blocked

       1   1 boot/procnto-instr   0f READY

       1   2 boot/procnto-instr 255r RECEIVE     1

       1   3 boot/procnto-instr  10r RECEIVE     1

       1   4 boot/procnto-instr  10r RUNNING

       1   5 boot/procnto-instr 255r RECEIVE     1

       1   7 boot/procnto-instr  10r RECEIVE     1

       1   8 boot/procnto-instr  10r RECEIVE     1

       1  10 boot/procnto-instr  10r RECEIVE     1

       1  11 boot/procnto-instr  10r RECEIVE     1

       1  12 boot/procnto-instr  10r RECEIVE     1

       1  15 boot/procnto-instr  10r RECEIVE     1

       1  16 boot/procnto-instr  10r RECEIVE     1

       2   1 sbin/tinit          10o REPLY       1

    4099   1 proc/boot/pci-bios  10o RECEIVE     1

    4100   1 proc/boot/slogger   10o RECEIVE     1

    4101   1 proc/boot/io-usb    10o SIGWAITINFO

    4101   2 proc/boot/io-usb    21r RECEIVE     4

    4101   3 proc/boot/io-usb    21r RECEIVE     7

    4101   4 proc/boot/io-usb    21r RECEIVE     1

    4101   5 proc/boot/io-usb    10o RECEIVE     10

    4101   6 proc/boot/io-usb    10r NANOSLEEP

    4101   7 proc/boot/io-usb    10o RECEIVE     10

    4101   8 proc/boot/io-usb    10o RECEIVE     10

    4102   1 proc/boot/io-hid    10o SIGWAITINFO

    4102   2 proc/boot/io-hid    21r RECEIVE     1

    4102   4 proc/boot/io-hid    10o RECEIVE     4

    4102   5 proc/boot/io-hid    10r REPLY       4101

    4102   6 proc/boot/io-hid    10r REPLY       4101

    4102   7 proc/boot/io-hid     9r RECEIVE     12

    4102   8 proc/boot/io-hid     9r RECEIVE     16

    4102   9 proc/boot/io-hid    10o RECEIVE     4

    4103   1 /boot/devc-con-hid  20o RECEIVE     1

    4103   2 /boot/devc-con-hid  10o REPLY       4102

    8200   1 roc/boot/devb-eide  10o SIGWAITINFO

    8200   2 roc/boot/devb-eide  21r RECEIVE     1

    8200   3 roc/boot/devb-eide  21o RECEIVE     7

    8200   4 roc/boot/devb-eide  10o RECEIVE     4

    8200   5 roc/boot/devb-eide  10o RECEIVE     4

    8200   6 roc/boot/devb-eide  10o RECEIVE     4

    8200   7 roc/boot/devb-eide  10o RECEIVE     4

    8200  10 roc/boot/devb-eide  10o RECEIVE     4

    8200  13 roc/boot/devb-eide  10o RECEIVE     4

   20489   1 sbin/pipe           10o SIGWAITINFO

   20489   2 sbin/pipe           10o RECEIVE     1

   20489   3 sbin/pipe           10o RECEIVE     1

   20489   4 sbin/pipe           10o RECEIVE     1

   20489   5 sbin/pipe           10o RECEIVE     1

   24586   1 sbin/mqueue         10o RECEIVE     1

   53259   1 usr/sbin/mcd        10o RECEIVE     1

   53259   2 usr/sbin/mcd        10o NANOSLEEP

   53259   3 usr/sbin/mcd        10o SIGWAITINFO

   53259   4 usr/sbin/mcd         9o SIGWAITINFO

   53259   5 usr/sbin/mcd        10o SIGWAITINFO

   53259   6 usr/sbin/mcd        10o SIGWAITINFO

   53259   7 usr/sbin/mcd        10o SIGWAITINFO

   57356   1 usr/sbin/random     10o SIGWAITINFO

   57356   2 usr/sbin/random     10o RECEIVE     1

   57356   3 usr/sbin/random     10o NANOSLEEP

   61453   1 sbin/enum-devices   10o REPLY       20489

   77840   1 sbin/enum-usb       10o SIGWAITINFO

   77840   2 sbin/enum-usb       10r REPLY       4101

   94222   1 sbin/io-display     10o SIGWAITINFO

   94222   2 sbin/io-display     10o RECEIVE     1

   94222   3 sbin/io-display     12o RECEIVE     1

   94222   4 sbin/io-display     10o RECEIVE     3

   94222   5 sbin/io-display     10o RECEIVE     1

  118799   1 sbin/io-pkt-v4-hc   21o SIGWAITINFO

  118799   2 sbin/io-pkt-v4-hc   21o RECEIVE     1

  118801   1 sbin/devc-pty       10o RECEIVE     1

  131090   1 usr/sbin/dumper     10o RECEIVE     1

  147475   1 bin/login           10o REPLY       4103

  147476   1 bin/login           10o REPLY       4103

  147477   1 bin/login           10o REPLY       4103

  176152   1 /photon/bin/Photon  10r RECEIVE     1

  196633   1 on/bin/io-graphics  12r CONDVAR     (0x805dc40)

  196633   2 on/bin/io-graphics  10r RECEIVE     1

  196633   3 on/bin/io-graphics  12r REPLY       176152

  213020   1 hoton/bin/devi-hid  10o RECEIVE     1

  213020   2 hoton/bin/devi-hid  10o REPLY       4102

  213020   3 hoton/bin/devi-hid  12o SIGWAITINFO

  213020   4 hoton/bin/devi-hid  10o RECEIVE     1

  213020   5 hoton/bin/devi-hid  10o RECEIVE     1

  225302   1 bin/login           10o REPLY       4103

  786465   1 sbin/io-pkt-v4-hc   21r SIGWAITINFO

  786465   2 sbin/io-pkt-v4-hc   10r RECEIVE     1

  811044   1 r/sbin/dhcp.client  10r NANOSLEEP

3182629   1 usr/sbin/sshd       10r SIGWAITINFO

4292635   1 usr/photon/bin/pwm  10r RECEIVE     1

4321309   1 r/photon/bin/shelf  10r CONDVAR     (0x8076f88)

4321309   2 r/photon/bin/shelf  10r RECEIVE     1

4341790   1 photon/bin/bkgdmgr  10r RECEIVE     1

4341791   1 hoton/bin/wmswitch  10r RECEIVE     2

4341792   1 r/photon/bin/saver  10r RECEIVE     1

4689961   1 usr/sbin/inetd      10r SIGWAITINFO

4771874   1 usr/bin/phrelay     10r REPLY       176152

4771874   2 usr/bin/phrelay     10r REPLY       786465

4771874   3 usr/bin/phrelay     10r CONDVAR     (0x805be30)

4792343   1 r/photon/bin/pterm  10r RECEIVE     1

4792346   1 bin/sh              10r SIGSUSPEND

4792355   1 bin/pidin           10r REPLY       1

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post97598

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post97598: Re: RE: RE: gf_layer_attach() failed !!!

2012-11-29T08:16:13Z

What else is running on the machine?

I have a default installation and basic GUI is running.

…......

Is Photon running?

Yes.

176152   1 /photon/bin/Photon  10r RECEIVE     1

…......

Number of displays: 1

Display 0: 1024X768, refresh = 60Hz

Number of layers: 1

…......

I am very new to QNX, but I am experienced in linux application and kernel development.

Currently I am using intel HD Graphics, PCI device, Vendor ID:8086, Device ID:0152.

….

please suggest what I need to draw a rectangle or circle or line.

Below is list of all the process running. Result of pidin command.

     pid tid name               prio STATE       Blocked

       1   1 boot/procnto-instr   0f READY

       1   2 boot/procnto-instr 255r RECEIVE     1

       1   3 boot/procnto-instr  10r RECEIVE     1

       1   4 boot/procnto-instr  10r RUNNING

       1   5 boot/procnto-instr 255r RECEIVE     1

       1   7 boot/procnto-instr  10r RECEIVE     1

       1   8 boot/procnto-instr  10r RECEIVE     1

       1  10 boot/procnto-instr  10r RECEIVE     1

       1  11 boot/procnto-instr  10r RECEIVE     1

       1  12 boot/procnto-instr  10r RECEIVE     1

       1  15 boot/procnto-instr  10r RECEIVE     1

       1  16 boot/procnto-instr  10r RECEIVE     1

       2   1 sbin/tinit          10o REPLY       1

    4099   1 proc/boot/pci-bios  10o RECEIVE     1

    4100   1 proc/boot/slogger   10o RECEIVE     1

    4101   1 proc/boot/io-usb    10o SIGWAITINFO

    4101   2 proc/boot/io-usb    21r RECEIVE     4

    4101   3 proc/boot/io-usb    21r RECEIVE     7

    4101   4 proc/boot/io-usb    21r RECEIVE     1

    4101   5 proc/boot/io-usb    10o RECEIVE     10

    4101   6 proc/boot/io-usb    10r NANOSLEEP

    4101   7 proc/boot/io-usb    10o RECEIVE     10

    4101   8 proc/boot/io-usb    10o RECEIVE     10

    4102   1 proc/boot/io-hid    10o SIGWAITINFO

    4102   2 proc/boot/io-hid    21r RECEIVE     1

    4102   4 proc/boot/io-hid    10o RECEIVE     4

    4102   5 proc/boot/io-hid    10r REPLY       4101

    4102   6 proc/boot/io-hid    10r REPLY       4101

    4102   7 proc/boot/io-hid     9r RECEIVE     12

    4102   8 proc/boot/io-hid     9r RECEIVE     16

    4102   9 proc/boot/io-hid    10o RECEIVE     4

    4103   1 /boot/devc-con-hid  20o RECEIVE     1

    4103   2 /boot/devc-con-hid  10o REPLY       4102

    8200   1 roc/boot/devb-eide  10o SIGWAITINFO

    8200   2 roc/boot/devb-eide  21r RECEIVE     1

    8200   3 roc/boot/devb-eide  21o RECEIVE     7

    8200   4 roc/boot/devb-eide  10o RECEIVE     4

    8200   5 roc/boot/devb-eide  10o RECEIVE     4

    8200   6 roc/boot/devb-eide  10o RECEIVE     4

    8200   7 roc/boot/devb-eide  10o RECEIVE     4

    8200  10 roc/boot/devb-eide  10o RECEIVE     4

    8200  13 roc/boot/devb-eide  10o RECEIVE     4

   20489   1 sbin/pipe           10o SIGWAITINFO

   20489   2 sbin/pipe           10o RECEIVE     1

   20489   3 sbin/pipe           10o RECEIVE     1

   20489   4 sbin/pipe           10o RECEIVE     1

   20489   5 sbin/pipe           10o RECEIVE     1

   24586   1 sbin/mqueue         10o RECEIVE     1

   53259   1 usr/sbin/mcd        10o RECEIVE     1

   53259   2 usr/sbin/mcd        10o NANOSLEEP

   53259   3 usr/sbin/mcd        10o SIGWAITINFO

   53259   4 usr/sbin/mcd         9o SIGWAITINFO

   53259   5 usr/sbin/mcd        10o SIGWAITINFO

   53259   6 usr/sbin/mcd        10o SIGWAITINFO

   53259   7 usr/sbin/mcd        10o SIGWAITINFO

   57356   1 usr/sbin/random     10o SIGWAITINFO

   57356   2 usr/sbin/random     10o RECEIVE     1

   57356   3 usr/sbin/random     10o NANOSLEEP

   61453   1 sbin/enum-devices   10o REPLY       20489

   77840   1 sbin/enum-usb       10o SIGWAITINFO

   77840   2 sbin/enum-usb       10r REPLY       4101

   94222   1 sbin/io-display     10o SIGWAITINFO

   94222   2 sbin/io-display     10o RECEIVE     1

   94222   3 sbin/io-display     12o RECEIVE     1

   94222   4 sbin/io-display     10o RECEIVE     3

   94222   5 sbin/io-display     10o RECEIVE     1

  118799   1 sbin/io-pkt-v4-hc   21o SIGWAITINFO

  118799   2 sbin/io-pkt-v4-hc   21o RECEIVE     1

  118801   1 sbin/devc-pty       10o RECEIVE     1

  131090   1 usr/sbin/dumper     10o RECEIVE     1

  147475   1 bin/login           10o REPLY       4103

  147476   1 bin/login           10o REPLY       4103

  147477   1 bin/login           10o REPLY       4103

  176152   1 /photon/bin/Photon  10r RECEIVE     1

  196633   1 on/bin/io-graphics  12r CONDVAR     (0x805dc40)

  196633   2 on/bin/io-graphics  10r RECEIVE     1

  196633   3 on/bin/io-graphics  12r REPLY       176152

  213020   1 hoton/bin/devi-hid  10o RECEIVE     1

  213020   2 hoton/bin/devi-hid  10o REPLY       4102

  213020   3 hoton/bin/devi-hid  12o SIGWAITINFO

  213020   4 hoton/bin/devi-hid  10o RECEIVE     1

  213020   5 hoton/bin/devi-hid  10o RECEIVE     1

  225302   1 bin/login           10o REPLY       4103

  786465   1 sbin/io-pkt-v4-hc   21r SIGWAITINFO

  786465   2 sbin/io-pkt-v4-hc   10r RECEIVE     1

  811044   1 r/sbin/dhcp.client  10r NANOSLEEP

3182629   1 usr/sbin/sshd       10r SIGWAITINFO

4292635   1 usr/photon/bin/pwm  10r RECEIVE     1

4321309   1 r/photon/bin/shelf  10r CONDVAR     (0x8076f88)

4321309   2 r/photon/bin/shelf  10r RECEIVE     1

4341790   1 photon/bin/bkgdmgr  10r RECEIVE     1

4341791   1 hoton/bin/wmswitch  10r RECEIVE     2

4341792   1 r/photon/bin/saver  10r RECEIVE     1

4689961   1 usr/sbin/inetd      10r SIGWAITINFO

4771874   1 usr/bin/phrelay     10r REPLY       176152

4771874   2 usr/bin/phrelay     10r REPLY       786465

4771874   3 usr/bin/phrelay     10r CONDVAR     (0x805be30)

4792343   1 r/photon/bin/pterm  10r RECEIVE     1

4792346   1 bin/sh              10r SIGSUSPEND

4792355   1 bin/pidin           10r REPLY       1

post97562: RE: RE: gf_layer_attach() failed !!!

2012-11-28T15:09:05Z

What else is running on the machine? Is Photon running? You only have one layer on your machine, so no other graphics environment or application can be running.

-----Original Message-----

From: Ashwini Dubey [mailto:community-noreply@qnx.com]

Sent: November-28-12 3:45 AM

To: advanced-graphics

Subject: Re: RE: gf_layer_attach() failed !!!

Return Value = 0x8

which is GF_ERR_INUSE..

I have installed qnx on i5 intel machine....and i can connect to it and run basic hello world...

i can use web browser too.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post97541

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post97543: gf_surface_create_layer failed, returned 1, memory allocation failed

2012-11-28T10:48:05Z

hi all,

i am new to QNX graphics and i am facing problem regarding memory allocation failed in

gf_surface_create_layer();

return value = 1, GF_ERR_MEM

code is given below..

#include <stdlib.h>

#include <stdio.h>

#include<gf/gf.h>

int main(int argc, char *argv[]) {

gf_dev_t            gdev;

gf_dev_info_t       gdev_info;

gf_display_t        display;

gf_display_info_t   display_info;

gf_layer_t layer;

gf_surface_t surface;

int i=0;

if (gf_dev_attach(&gdev, GF_DEVICE_INDEX(0), &gdev_info) != GF_ERR_OK) {

    printf("gf_dev_attach() failed\n");

    return (-1);

}

printf("Number of displays: %d\n",gdev_info.ndisplays);

for (i = 0; i < gdev_info.ndisplays; i++ )

{

    printf("Display %d: ", i);

    if (gf_display_attach(&display, gdev, i, &display_info) == GF_ERR_OK)

    {

        printf("%dX%d, refresh = %dHz\n", display_info.xres,

            display_info.yres, display_info.refresh);

        printf("Number of layers: %d\n", display_info.nlayers);

        if (gf_layer_attach(&layer, display,0, 1) == GF_ERR_OK )

        {

         printf("\n\tLayer attached Successfully.");

         //printf("\n\t%d", gf_surface_create_layer(&surface, &layer, 1,0, display_info.xres, display_info.yres, display_info.format , NULL, 0 ));

         if(gf_surface_create_layer(&surface, &layer, 0,0, display_info.xres, display_info.yres, GF_FORMAT_PACK_RGB565 , NULL, 0 ) == GF_ERR_OK )

         {

         printf("\n\tSurface created successfully.");

         }

         else

         printf("\n\tSurface failed.");

        }

        else

         printf("\n\tLayer Failed.");

    }

    else

        printf("gf_display_attach() failed\n");

}

printf("\n\tWelcome to the QNX Momentics IDE\n");

if(gdev)

{

gf_dev_detach(gdev);

}

return EXIT_SUCCESS;

}

post97541: Re: RE: gf_layer_attach() failed !!!

2012-11-28T08:45:08Z

Return Value = 0x8

which is GF_ERR_INUSE..

I have installed qnx on i5 intel machine....and i can connect to it and run basic hello world...

i can use web browser too.

post97491: RE: gf_layer_attach() failed !!!

2012-11-27T14:15:55Z

What is the return value from gf_layer_attach()?

-----Original Message-----

From: Ashwini Dubey [mailto:community-noreply@qnx.com]

Sent: November-27-12 2:02 AM

To: advanced-graphics

Subject: gf_layer_attach() failed !!!

Dear All,

Dear All,

I am new to qnx GF, I have read “Advanced Graphics Developers Guide.pdf” and trying to draw a rectangle on GF. I have x86 target machine connected to QNX Momentics IDE 4.7.

gf_layer_attach() is failed... can some help...

find output and code below.

Output:

Number of displays: 1

Display 0: 1024X768, refresh = 60Hz

Number of layers: 1

code snippet:

#include <stdio.h>

#include <stdlib.h>

#include <gf/gf.h>

int main() {

//device & display

gf_dev_t            gdev;

gf_dev_info_t       gdev_info;

gf_display_t        display;

gf_display_info_t   display_info;

int i;

//layer

gf_layer_t  layer;

//surface

gf_surface_t    surface;

unsigned height;

unsigned width;

gf_layer_info_t layer_info;

//context

gf_context_t context;

if (gf_dev_attach(&gdev, GF_DEVICE_INDEX(0), &gdev_info) != GF_ERR_OK) {

    printf("gf_dev_attach() failed\n");

    return (-1);

}

printf("Number of displays: %d\n",gdev_info.ndisplays);

for (i = 0; i < gdev_info.ndisplays; i++) {

    printf("Display %d: ", i);

    if (gf_display_attach(&display, gdev, i, &display_info) == GF_ERR_OK) {

        printf("%dX%d, refresh = %dHz\n", display_info.xres,

            display_info.yres, display_info.refresh);

        printf("Number of layers: %d\n", display_info.nlayers);

    } else {

        printf("gf_display_attach() failed\n");

    }

}

if (display != NULL) {

    if (gf_layer_attach(&layer,display,display_info.main_layer_index, 0) == GF_ERR_OK) {

    //if(gf_layer_attach(&layer,display,0, 0) == GF_ERR_OK) {

printf("Attaching to Layer : 2\n ");

    }

    else {

        printf("gf_layer_attach() failed\n");

    }

}

if((gf_layer_query(layer,0,&layer_info))==GF_ERR_OK){

printf("Layer is of Format %d\n",layer_info.format); } else{

printf(" gf_layer_query() failed");

}

width = display_info.xres;

height = display_info.yres;

if (gf_surface_create_layer(&surface, &layer, 1, 0, width, height, layer_info.format, NULL, 0) == GF_ERR_OK) {

printf("Succesfully created surface on top of layer\n");

        gf_layer_set_surfaces(layer, &surface, 1);

        /* draw stuff here */

    } else {

        printf("gf_surface_create_layer() failed\n");

    }

if (gf_context_create(&context) != GF_ERR_OK) {

        fprintf(stderr, "gf_context_create failed\n");

        return -1;

    }

else{

printf(" Created the context succefully\n"); }

if (gf_context_set_surface(context,surface)!= GF_ERR_OK) {

            fprintf(stderr, "gf_context_set_surface failed\n");

            return -1;

        }

else{

        printf("Context Surface is Set Succesfully\n");

        }

gf_draw_begin(context);

gf_context_set_fgcolor(context,0xFFFFFF);

gf_draw_rect(context,10,10,50,50);

gf_context_set_fgcolor(context,0x000000);

gf_draw_rect(context,20,20,40,40);

gf_draw_end(context);

gf_dev_detach(gdev);

return 0;

}

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post97473

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post97473: gf_layer_attach() failed !!!

2012-11-27T07:01:57Z

Dear All,

Dear All,

I am new to qnx GF, I have read “Advanced Graphics Developers Guide.pdf” and trying to draw a rectangle on GF. I have x86 target machine connected to QNX Momentics IDE 4.7.

gf_layer_attach() is failed... can some help...

find output and code below.

Output:

Number of displays: 1

Display 0: 1024X768, refresh = 60Hz

Number of layers: 1

code snippet:

#include <stdio.h>

#include <stdlib.h>

#include <gf/gf.h>

int main() {

//device & display

gf_dev_t            gdev;

gf_dev_info_t       gdev_info;

gf_display_t        display;

gf_display_info_t   display_info;

int i;

//layer

gf_layer_t  layer;

//surface

gf_surface_t    surface;

unsigned height;

unsigned width;

gf_layer_info_t layer_info;

//context

gf_context_t context;

if (gf_dev_attach(&gdev, GF_DEVICE_INDEX(0), &gdev_info) != GF_ERR_OK) {

    printf("gf_dev_attach() failed\n");

    return (-1);

}

printf("Number of displays: %d\n",gdev_info.ndisplays);

for (i = 0; i < gdev_info.ndisplays; i++) {

    printf("Display %d: ", i);

    if (gf_display_attach(&display, gdev, i, &display_info) == GF_ERR_OK) {

        printf("%dX%d, refresh = %dHz\n", display_info.xres,

            display_info.yres, display_info.refresh);

        printf("Number of layers: %d\n", display_info.nlayers);

    } else {

        printf("gf_display_attach() failed\n");

    }

}

if (display != NULL) {

    if (gf_layer_attach(&layer,display,display_info.main_layer_index, 0) == GF_ERR_OK) {

    //if(gf_layer_attach(&layer,display,0, 0) == GF_ERR_OK) {

printf("Attaching to Layer : 2\n ");

    }

    else {

        printf("gf_layer_attach() failed\n");

    }

}

if((gf_layer_query(layer,0,&layer_info))==GF_ERR_OK){

printf("Layer is of Format %d\n",layer_info.format);

}

else{

printf(" gf_layer_query() failed");

}

width = display_info.xres;

height = display_info.yres;

if (gf_surface_create_layer(&surface, &layer, 1, 0,

width, height, layer_info.format, NULL, 0) == GF_ERR_OK) {

printf("Succesfully created surface on top of layer\n");

        gf_layer_set_surfaces(layer, &surface, 1);

        /* draw stuff here */

    } else {

        printf("gf_surface_create_layer() failed\n");

    }

if (gf_context_create(&context) != GF_ERR_OK) {

        fprintf(stderr, "gf_context_create failed\n");

        return -1;

    }

else{

printf(" Created the context succefully\n");

}

if (gf_context_set_surface(context,surface)!= GF_ERR_OK) {

            fprintf(stderr, "gf_context_set_surface failed\n");

            return -1;

        }

else{

        printf("Context Surface is Set Succesfully\n");

        }

gf_draw_begin(context);

gf_context_set_fgcolor(context,0xFFFFFF);

gf_draw_rect(context,10,10,50,50);

gf_context_set_fgcolor(context,0x000000);

gf_draw_rect(context,20,20,40,40);

gf_draw_end(context);

gf_dev_detach(gdev);

return 0;

}

post96507: RE: Unable to launch io-graphics

2012-10-22T12:25:21Z

Good :)

-----Original Message-----

From: KB [mailto:community-noreply@qnx.com]

Sent: October-19-12 2:08 AM

To: advanced-graphics

Subject: Re: Unable to launch io-graphics

Never mind this issue is solved.

For further information check on this topic : http://www.openqnx.com/phpbbforum/viewtopic.php?f=8&t=13205

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post96441

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post96505: RE: Unable to launch io-graphics

2012-10-22T12:08:29Z

Do you have a /usr/photon/font_repository on your target?

-----Original Message-----

From: KB [mailto:community-noreply@qnx.com]

Sent: October-15-12 9:35 AM

To: advanced-graphics

Subject: Unable to launch io-graphics

Hi,

I'm trying to run a small photon application on my embedded system (for instance phcalc).

Here's my issue :

I am able to run io-display, see my display.conf :

# My PCI

device {

drivername=lx800

vid=0x1022

did=0x2081

deviceindex=0

display {

xres=1024

yres=768

refresh=60

pixel_format=argb8888

}

}

I am able to run vsync.

I can't launch demo-alpha > sloginfo :

When I run io-graphics I've got the following error :

Oct 15 15:07:47    6     8     0 Primary active: 0 9 7

Oct 15 15:07:58    6     8     0 VGA primary : bus 0x0 dev/func 0x9

Oct 15 15:07:58    6     8     0 Found 1 PCI/AGP display devices

Oct 15 15:07:58    6     8     0 pci_init: found PCI device 1022:2081

Oct 15 15:07:58    6     8     0 No config file specified

Oct 15 15:07:59    6     8     0 Primary active: 0 9 7

Il I launch Photon binary and then phcalc (for example), I have :

"unable to attach to font server, pf_attach() : No such file or directory"

"Ap: Unable to locate Photon (2)"

If I run phfont I have : "PfAttachLocalDllArgv() : Invalid argument "

and the following slog :

Oct 15 15:07:59    6     8     0 Primary active: 0 9 7

Oct 15 15:12:42    1     8     0 phfont: init...

Oct 15 15:12:42    1     8     0 phfont: unable to load font extension rules

If I run ph script I have : "Unable to start graphics driver(s)"

and the followin slog :

Oct 15 15:16:46    1     8     0 phfont.so: init...

Oct 15 15:16:46    1     8     0 phfont.so: unable to load font extension rules

Oct 15 15:16:46    2     8     0 io-graphics:  FAIL PfAttachLocalDll(): Invalid argument, please verify font server configuration.

Oct 15 15:16:46    2     8     0 fontserver_setup()

My script is attached. My environment variables are defined hereafter :

HOME=/

PATH=/proc/boot:/bin:/usr/bin:/usr/sbin:/sbin:/APD:/usr/photon/bin

LD_LIBRARY_PATH=/proc/boot:/lib:/usr/lib:/lib/dll

PHOTON_PATH=/usr/photon

PHOTON2_PATH=/usr/photon

PHOTON=/dev/photon

PHFONT=/dev/phfont

setconf _CS_LIBPATH :/proc/boot:/lib/dll:/dev/shmem:/lib:/usr/lib:/lib/dll:/lib/dll/font:/usr/photon/lib

I'm running QNX 6.5.0. My board is a Rhodeus Geode (AMD 500Mghz) : http://www.diamondsystems.com/products/rhodeus

My starting point was this official BSP : http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/wiki/AmdGeodeLxdb800

But I had to change things (surcvh as realtek driver)...

I use devg-lx800.so graphic library.

Any help would be appreciated.

Thanks.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post96328

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post96441: Re: Unable to launch io-graphics

2012-10-19T06:08:24Z

Never mind this issue is solved.

For further information check on this topic : http://www.openqnx.com/phpbbforum/viewtopic.php?f=8&t=13205

post96328: Unable to launch io-graphics

2012-10-15T13:35:13Z

Hi,

I'm trying to run a small photon application on my embedded system (for instance phcalc).

Here's my issue :

I am able to run io-display, see my display.conf :

# My PCI

device {

drivername=lx800

vid=0x1022

did=0x2081

deviceindex=0

display {

xres=1024

yres=768

refresh=60

pixel_format=argb8888

}

}

I am able to run vsync.

I can't launch demo-alpha > sloginfo :

When I run io-graphics I've got the following error :

Oct 15 15:07:47    6     8     0 Primary active: 0 9 7

Oct 15 15:07:58    6     8     0 VGA primary : bus 0x0 dev/func 0x9

Oct 15 15:07:58    6     8     0 Found 1 PCI/AGP display devices

Oct 15 15:07:58    6     8     0 pci_init: found PCI device 1022:2081

Oct 15 15:07:58    6     8     0 No config file specified

Oct 15 15:07:59    6     8     0 Primary active: 0 9 7

Il I launch Photon binary and then phcalc (for example), I have :

"unable to attach to font server, pf_attach() : No such file or directory"

"Ap: Unable to locate Photon (2)"

If I run phfont I have : "PfAttachLocalDllArgv() : Invalid argument "

and the following slog :

Oct 15 15:07:59    6     8     0 Primary active: 0 9 7

Oct 15 15:12:42    1     8     0 phfont: init...

Oct 15 15:12:42    1     8     0 phfont: unable to load font extension rules

If I run ph script I have : "Unable to start graphics driver(s)"

and the followin slog :

Oct 15 15:16:46    1     8     0 phfont.so: init...

Oct 15 15:16:46    1     8     0 phfont.so: unable to load font extension rules

Oct 15 15:16:46    2     8     0 io-graphics:  FAIL PfAttachLocalDll(): Invalid argument, please verify font server configuration.

Oct 15 15:16:46    2     8     0 fontserver_setup()

My script is attached. My environment variables are defined hereafter :

HOME=/

PATH=/proc/boot:/bin:/usr/bin:/usr/sbin:/sbin:/APD:/usr/photon/bin

LD_LIBRARY_PATH=/proc/boot:/lib:/usr/lib:/lib/dll

PHOTON_PATH=/usr/photon

PHOTON2_PATH=/usr/photon

PHOTON=/dev/photon

PHFONT=/dev/phfont

setconf _CS_LIBPATH :/proc/boot:/lib/dll:/dev/shmem:/lib:/usr/lib:/lib/dll:/lib/dll/font:/usr/photon/lib

I'm running QNX 6.5.0. My board is a Rhodeus Geode (AMD 500Mghz) : http://www.diamondsystems.com/products/rhodeus

My starting point was this official BSP : http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/wiki/AmdGeodeLxdb800

But I had to change things (surcvh as realtek driver)...

I use devg-lx800.so graphic library.

Any help would be appreciated.

Thanks.

post95938: Re: RE: devi-hid for armle-v7

2012-10-01T03:24:56Z

devi-hid for armle-v7 is available from QNX 6.5.0 SP1

post95905: Re: RE: BITMAPV5HEADER support

2012-09-28T03:43:48Z

I thought BITMAPV5HEADER has been around for a while since it was supported from Windows 98 onwards. Thanks Derek!

post95876: RE: BITMAPV5HEADER support

2012-09-27T12:59:06Z

650 supported v4, which was the latest at the initial 650 time of release.

-----Original Message-----

From: Philip Endaya [mailto:community-noreply@qnx.com]

Sent: September-27-12 2:51 AM

To: advanced-graphics

Subject: BITMAPV5HEADER support

Hi,

After some bmp codec soure code digging (bmp.h), it seems that QNX 6.5.0 only supports up to BITMAPV4HEADER but according to QNX documentation it provides full Microsoft BMP support for all known variants (http://www.qnx.com/developers/docs/6.5.0/index.jsp?topic=/com.qnx.doc.gf_dev_guide/images.html). Can someone confirm this?

BR,

Philip

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post95870

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post95870: BITMAPV5HEADER support

2012-09-27T06:50:31Z

Hi,

After some bmp codec soure code digging (bmp.h), it seems that QNX 6.5.0 only supports up to BITMAPV4HEADER but according to QNX documentation it provides full Microsoft BMP support for all known variants (http://www.qnx.com/developers/docs/6.5.0/index.jsp?topic=/com.qnx.doc.gf_dev_guide/images.html). Can someone confirm this?

BR,

Philip

post95434: RE: libffb.so Issues on Poulsbo Board

2012-09-07T12:39:58Z

Ok, glad you got it going.

-----Original Message-----

From: Joe Thompson [mailto:community-noreply@qnx.com]

Sent: September-06-12 3:29 PM

To: advanced-graphics

Subject: Re: libffb.so Issues on Poulsbo Board

Must be a driver conflict

I stripped everything I didn't need from the OS and finally got graphics started using the poulsbo driver.

Unfortunately I didn't do this systematically so I have no idea which driver or file was causing the problem

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post95421

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post95433: RE: RE: gf_draw_blit2() - problem with bigger surface size

2012-09-07T12:37:54Z

True, but is there any way that another code path could have modified the context?

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: September-06-12 5:09 PM

To: advanced-graphics

Subject: Re: RE: gf_draw_blit2() - problem with bigger surface size

I believe that the gf_context_t is inited properly.

Here is a part of code responsible for drawing:

   gf_draw_begin (context);

   gf_draw_blit2 (context, img_surf, NULL, 0, 0, 639, 479, 0, 0);

   gf_draw_flush (context);

   gf_draw_end (context);

nothing specific but fails.

When I do a blit2() operation more often but on smaller area then everything is ok:

  gf_draw_begin (context);

  gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2);

  gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2+1);

  gf_draw_flush (context);

  gf_draw_end (context);

BUFLEN is defined as 320.

In both examples img_suf is connected with the same 640x480 target area.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post95423

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post95426: Re: RE: gf_draw_blit2() - problem with bigger surface size

2012-09-07T08:15:36Z

> I believe that the gf_context_t is inited properly.

>

> Here is a part of code responsible for drawing:

>    gf_draw_begin (context);

>    gf_draw_blit2 (context, img_surf, NULL, 0, 0, 639, 479, 0, 0);

>    gf_draw_flush (context);

>    gf_draw_end (context);

> nothing specific but fails.>

> When I do a blit2() operation more often but on smaller area then everything

> is ok:

>   gf_draw_begin (context);

>   gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2);

>   gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2+1);

>

>   gf_draw_flush (context);

>   gf_draw_end (context);

> BUFLEN is defined as 320.

>

> In both examples img_suf is connected with the same 640x480 target area.

Hi,

What will happen in case of using static picture in img_surf, like previous captured frame?

Thanks.

post95423: Re: RE: gf_draw_blit2() - problem with bigger surface size

2012-09-06T21:09:23Z

I believe that the gf_context_t is inited properly.

Here is a part of code responsible for drawing:

   gf_draw_begin (context);

   gf_draw_blit2 (context, img_surf, NULL, 0, 0, 639, 479, 0, 0);

   gf_draw_flush (context);

   gf_draw_end (context);

nothing specific but fails.

When I do a blit2() operation more often but on smaller area then everything is ok:

  gf_draw_begin (context);

  gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2);

  gf_draw_blit2 (context, img_surf, NULL, 0, 0, BUFLEN*2-1, 0, 0, index*2+1);

  gf_draw_flush (context);

  gf_draw_end (context);

BUFLEN is defined as 320.

In both examples img_suf is connected with the same 640x480 target area.

post95421: Re: libffb.so Issues on Poulsbo Board

2012-09-06T19:28:32Z

Must be a driver conflict

I stripped everything I didn't need from the OS and finally got graphics started using the poulsbo driver.

Unfortunately I didn't do this systematically so I have no idea which driver or file was causing the problem

post95386: RE: gf_draw_blit2() - problem with bigger surface size

2012-09-05T12:40:41Z

Someone here fixed something similar in another unrelated piece of software, where the gf_context_t was not inited properly, and we were getting similar behaviour.

-----Original Message-----

From: Mike Gorchak [mailto:community-noreply@qnx.com]

Sent: September-05-12 1:48 AM

To: advanced-graphics

Subject: Re: gf_draw_blit2() - problem with bigger surface size

> Does mean that hardware is too slow ?

Maybe, do you using plain blit2 or with additional context flags enabled, like chroma, etc?

> Or maybe I'm doing something wrong ?

I've seen similar behavior when forgot to use flush at the end of picture rendering in case if you have mixed environment, like software+hardware operations on the same surface. Two or more consecutive blit2()s without flush() can be optimized by hardware, you will get the last blit2() result.

There are few possible issues:

1) You are not using flush() for blit2() operation. And blit2() is the only function which is called. Hardware will take video memory content for blitting in asynchronous way. For example, 1/1000 of sec is a delay between command to graphics controller has been issued and real blit2() execution in hardware. But in next 1/100000 sec you will get 3-4 optimized blit2() operations, while content of video memory was changed partially.

2) In the same time your application can use some graphics operations which are implemented in software. For example, output of some text and picture to the same surface. This is a race condition between software text rendering and hardware blit2(), so some strange visible artifacts could be observed. It depends on a GPU caching policy. So correct operations sequence will be blit2() -> flush() -> put_text("FPS: %d").

3) You are updating your video frame content asynchronously comparing to blit2() operation. For example, there is possible such situation: two tasks are not synchronized, one is filling data which it obtains from camera and second one is doing blit2() + flush(). Due to asynchronous nature of GPU, it can execute blit2() in the same time when changing of video frame content is in process. In such case double buffering for managing video stream will be a solution.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post95374

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post95374: Re: gf_draw_blit2() - problem with bigger surface size

2012-09-05T05:47:36Z

> Does mean that hardware is too slow ?

Maybe, do you using plain blit2 or with additional context flags enabled, like chroma, etc?

> Or maybe I'm doing something wrong ?

I've seen similar behavior when forgot to use flush at the end of picture rendering in case if you have mixed environment, like software+hardware operations on the same surface. Two or more consecutive blit2()s without flush() can be optimized by hardware, you will get the last blit2() result.

There are few possible issues:

1) You are not using flush() for blit2() operation. And blit2() is the only function which is called. Hardware will take video memory content for blitting in asynchronous way. For example, 1/1000 of sec is a delay between command to graphics controller has been issued and real blit2() execution in hardware. But in next 1/100000 sec you will get 3-4 optimized blit2() operations, while content of video memory was changed partially.

2) In the same time your application can use some graphics operations which are implemented in software. For example, output of some text and picture to the same surface. This is a race condition between software text rendering and hardware blit2(), so some strange visible artifacts could be observed. It depends on a GPU caching policy. So correct operations sequence will be blit2() -> flush() -> put_text("FPS: %d").

3) You are updating your video frame content asynchronously comparing to blit2() operation. For example, there is possible such situation: two tasks are not synchronized, one is filling data which it obtains from camera and second one is doing blit2() + flush(). Due to asynchronous nature of GPU, it can execute blit2() in the same time when changing of video frame content is in process. In such case double buffering for managing video stream will be a solution.

post95372: libffb.so Issues on Poulsbo Board

2012-09-04T21:43:34Z

Hardware: Congatec Conga-CA COM Express module (Intel Atom Z510/Intel US15W)

Trying to bring graphics up on this board using the generic X86 bsp. Started off with the OEM supplied bsp but now am running into the same issue.

As soon as I add the libffb.so driver to the OS, the board will no longer boot and I am left looking at a display of shifting colors. The crash is before anything is running on the system so I can't get any log information or debug information.

Have tried using io-display with both the Poulsbo graphics driver and the generic vesabios, the problem appears to be in the common libffb though so I'm stuck at a dead-end... Will be embarrassing to have to resort to a text-based gui!

Has anybody managed to get graphics support working on an Atom-based board?

post95236: Sharing interrupts

2012-08-28T20:48:26Z

In the our system a video stream is received from Ethernet device.

And then when the all data is already collected picture is displayed.

Can a shared interrupt of the ethernet and graphics controller

have an impact for application performance:

   io-pkt-v4-hc                 --> irq5,

   io-graphics / io-display --> irq5

?

Or better idea is to separate interrupts for both devices i.e. :

   io-pkt-v4-hc                 --> irq5,

   io-graphics / io-display --> irq10

?

post94943: Re: gf_draw_blit2() - problem with bigger surface size

2012-08-19T17:27:15Z

I have attached picture that shows a black area.

Rest from image lines are displayed properly.

Has anybody idea whay these black lines are visible and how can I get rid of them ?

post94942: gf_draw_blit2() - problem with bigger surface size

2012-08-19T12:52:09Z

In my application I had to call the gf_draw_blit2() very often.

When size of the image is 320x240:

gf_draw_blit2 (context, img_surf, NULL, 0, 0, 319, 239, 0, 0);

then everything works fine.

When size of the image is 640x480

gf_draw_blit2 (context, img_surf, NULL, 0, 0, 639, 479, 0, 0);

several consecutive lines in the window are black (not rendered at all).

Does mean that hardware is too slow ?

Or maybe I'm doing something wrong ?

I'm testing against the devg-i830.so driver (Device Id: 0x2582).

post94734: Re: Sample of image loading in libimg

2012-08-09T19:54:52Z

I'm also having troubles with libimg. After adding libimg the app crash when I execute

if (BPS_SUCCESS != screen_request_events(screen_cxt))

Any idea? I added the -limg on linking stage, its ok?

Thanks

post94680: RE: Combining GF and Photon together

2012-08-07T16:57:38Z

Yes, in the Foundry27 Graphics project, there is examples showing this.

-----Original Message-----

From: Jacek Rudnicki [mailto:community-noreply@qnx.com]

Sent: August-07-12 12:56 PM

To: advanced-graphics

Subject: Combining GF and Photon together

There is native application written under Photon and similar program that uses Graphics Framework API.

How to mix them together ?

Is there any documentation which describes such process ?

Jacek

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post94679

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post94679: Combining GF and Photon together

2012-08-07T16:55:56Z

There is native application written under Photon and similar program that uses Graphics Framework API.

How to mix them together ?

Is there any documentation which describes such process ?

Jacek

post94678: Re: Migration from Photon to GF

2012-08-07T16:51:52Z

I have successfully generated 8-bits gray palette as follows:

gf_palette_t palette;

gf_color_t g8pal[256];

...

for (i = 0; i < 256; i++) {

g8pal[i] = (0xff << 24) | ((i & 0xff) << 16) | ((i & 0xff) << 8) | (i & 0xff);

}

palette.ncolors = 256;

palette.colors = (gf_color_t*)g8pal;

It works fine with rgb565 format for io-display.

post94659: Re: Migration from Photon to GF

2012-08-06T20:00:03Z

I have changed pixel_format in the display.conf file from "rgb565" onto "pal8" value.

Now the gf_draw_blit2() works while with "rgb565" screen is dark all the time.

Picture is visible but colors are not correct.

I wonder how to generate 8-bits gray palette under GF ?

post94656: Re: Migration from Photon to GF

2012-08-06T09:40:28Z

I am having the same problem, I can't find 'g8pal'. It looks like it should be a pointer to a palette of type 'img_color_t' which is just an integer. So what should the g8pal palette look like. I am new to QNX and GF graphics and am finding it hard to get things working as a lot of info is missing from QNX examples.

post94650: Migration from Photon to GF

2012-08-05T20:28:59Z

Hi,

Application is responsible for displaying data (8-bits/pixel graymap). Data is received from Ethernet interface and displayed when image is complete (all frames were already gathered).

Under Photon as solution I use PhCreateImage() and  PgDrawPhImage() functions.

BTW Is there any other way in order to achive a better performane in Photon ?

It's time to move to GF.

In the "Working with Images" chapter there is a following code:

    palette.ncolors = 256;

    palette.colors = (img_color_t*)g8pal;

but compiler says that g8pal is undefined. How can I use this ?

Should I generate array of colors itself (in similar way that was done under Photon,  PgSetPalette()) ?

What function should I use under GF as equivalent ?

I found the gf_draw_image() but this function doesn't support GF_FORMAT_PAL8...

Regards,

Jacek

post94376: Re: RE: RE: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-20T21:55:39Z

Ok, thanks Derek. I've given the test code we're using to support, hopefully they can come up with some examples of how to make this work.

Thanks again,

Don

post94366: RE: RE: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-20T13:00:31Z

Ok, it looks like the imx51/53 driver supports this, please send your test case through support.

-----Original Message-----

From: Don Meyer [mailto:community-noreply@qnx.com]

Sent: July-19-12 6:53 PM

To: advanced-graphics

Subject: Re: RE: RE: Displaying layer 0 gf graphics with Photon in layer 1?

This is on an iMX53 eval board, the graphics are those integrated into the Cortex chip. It does support two layers in hardware, and we have Photon running on layer 1 (versus the default layer 0).

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post94353

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post94353: Re: RE: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-19T22:52:59Z

This is on an iMX53 eval board, the graphics are those integrated into the Cortex chip. It does support two layers in hardware, and we have Photon running on layer 1 (versus the default layer 0).

post94349: RE: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-19T19:35:35Z

You are attaching to the same layer as Photon?  What graphics chipset are you using?

-----Original Message-----

From: Don Meyer [mailto:community-noreply@qnx.com]

Sent: July-19-12 2:53 PM

To: advanced-graphics

Subject: Re: RE: Displaying layer 0 gf graphics with Photon in layer 1?

Thanks for the reply.  I did post essentially the same question in the Photon forum, and then posted it here today as I hadn't gotten any responses there yet.  Not certian which form is most appropriate for this question - it seems to span both.

I've gotten a similar suggestion from tech support (although with less detail, more of a general direction).  I've been going this way, so good to know that I'm on the right path.

Not having much luck however.

I'm attaching to the layer that Photon is using with a  passive attach.  Then, I set the chroma:

gf_chroma_t chroma;

chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

chroma.color0 = 0xff0000;

chroma.color1 = 0; // unused

gf_layer_set_chroma( layer, &chroma );

gf_layer_update( layer, 0 );

After that is done, we create our window and a PtOSContainer, whose background is set to full red.

That red background is still drawing however.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post94348

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post94348: Re: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-19T18:53:07Z

Thanks for the reply.  I did post essentially the same question in the Photon forum, and then posted it here today as I hadn't gotten any responses there yet.  Not certian which form is most appropriate for this question - it seems to span both.

I've gotten a similar suggestion from tech support (although with less detail, more of a general direction).  I've been going this way, so good to know that I'm on the right path.

Not having much luck however.

I'm attaching to the layer that Photon is using with a  passive attach.  Then, I set the chroma:

gf_chroma_t chroma;

chroma.mode = GF_CHROMA_OP_SRC_MATCH | GF_CHROMA_OP_NO_DRAW;

chroma.color0 = 0xff0000;

chroma.color1 = 0; // unused

gf_layer_set_chroma( layer, &chroma );

gf_layer_update( layer, 0 );

After that is done, we create our window and a PtOSContainer, whose background is set to full red.

That red background is still drawing however.

post94346: RE: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-19T17:50:08Z

I just answer a similar question in the Photon forum:

I have not done this approach lately, but try these suggestions. You could set the fill colour to Pg_TRANSPARENT,

unless you are using a PtOSContainer.

Now, I think this is the better approach here:

If you are using a PtOSContainer, set the fill colour to something not common,

then set the chroma on the layer to that value, and no draw.

-----Original Message-----

From: Don Meyer [mailto:community-noreply@qnx.com]

Sent: July-19-12 1:33 PM

To: advanced-graphics

Subject: Displaying layer 0 gf graphics with Photon in layer 1?

Is there any way to have Photon's background either not be drawn, or drawn as "transparent"?

I'm trying to render some graphics on layer 0 using the gf library, and have some Photon objects overlay them, with Photon on layer 1.

The problem is, Photon draws a background which totally obscures what's rendered on layer 0.

Thanks!

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post94343

To cancel your subscription to this discussion, please e-mail advanced-graphics-unsubscribe@community.qnx.com

post94343: Displaying layer 0 gf graphics with Photon in layer 1?

2012-07-19T17:32:30Z

Is there any way to have Photon's background either not be drawn, or drawn as "transparent"?

I'm trying to render some graphics on layer 0 using the gf library, and have some Photon objects overlay them, with Photon on layer 1.

The problem is, Photon draws a background which totally obscures what's rendered on layer 0.

Thanks!

post93716: Re: Sample of image loading in libimg

2012-06-18T09:05:17Z

Hi,

I successfully make a sample application of image loading using libimg. Now i want to is it possible to do image editing feature using Limbing. If yes then any documentation or source for this.

post93684: Sample of image loading in libimg

2012-06-15T05:37:57Z

Hi,

I am new to QNX development. Currently i am developing Native application for BB10 in which i come to work with QNX native. I want to do image editing in native can anyone help me how can i achieve this?

I tried to find libimg library but not get that much help. I make a sample application using libimg where i load the image but gives many error.

I attached my source code for the same to load the image please help me on this.

post93239: Re: Problem for gf-calib

2012-05-22T10:28:00Z

> I am using TI beagle-board. When I use gf-calib for calibration of my touch

> screen, the following message displays:

> # gf-calib

> Number of layers: 3

> Resolution: 720x400, refresh = 60Hz

> layer_idx=0

> gf_cursor_set(): error = 7

>

> I have no idea how to solve it. Any help?

Try the -D option. Also, you need to have "img.conf" under /etc/system/config, plus the img_codec so's in your library path.

- Malte

post92952: Problem for gf-calib

2012-05-11T01:55:53Z

I am using TI beagle-board. When I use gf-calib for calibration of my touch screen, the following message displays:

# gf-calib

Number of layers: 3

Resolution: 720x400, refresh = 60Hz

layer_idx=0

gf_cursor_set(): error = 7

I have no idea how to solve it. Any help?

post91489: Re: Source code of vsync?

2012-02-10T19:54:08Z

bump

post91261: Re: ScreenShot in GF

2012-01-31T22:18:58Z

Hi John,

Did you ever find an answer to this? Looks like I'm following down the same road now.

Thanks,

Devan

post90739: How to Repair damage to proportional text in Graphics Framework ?

2011-12-22T11:11:49Z

Hi ,

When dealing with proportional fonts, sometimes the vectors of one glyph

run into the vectors of another. This is especially evident when using a

font such as Nuptial BT. We  need to take special care when repairing

damage to such fonts.

PfExtentTextCharPositions()

<http://www.qnx.com/developers/docs/6.4.1/photon/lib_ref/pf/pfextenttext

charpositions.html>  addresses this issue in Photon environment. This

position is where you should draw the next character.

If you use the PF_CHAR_DRAW_POSITIONS flag, the bearing x of the

following character isn't applied to the position, which is useful when

you're placing cursors.

What is the equivalent on PfExtentTextCharPositions()

<http://www.qnx.com/developers/docs/6.4.1/photon/lib_ref/pf/pfextenttext

charpositions.html>  in Graphics Framework ?

Please let me know regarding this.

Thanks,

Karthik Kannan M

General Electric - Transportation

post90709: Re: RE: Can't load driver dll

2011-12-21T07:56:15Z

Christophe,

You were right, the problem was with another library included in devg-flat.so that was not present in the library path.

Thanks a lot for your valuable help.

Regards,

Nicolas

post90694: RE: Can't load driver dll

2011-12-20T16:42:37Z

Hello,

Probably the problem was not with your library but with another one included with your library.

Try this:

# DL_DEBUG=1 io-display  -c /mnt/sdhc/gf/display.conf -d vid=0x10ee,did=0x6024

And check the missing library.

Christophe Hannoyer

-----Message d'origine-----

De : Nicolas Pinault [mailto:community-noreply@qnx.com]

Envoyé : mardi 20 décembre 2011 17:11

À : advanced-graphics

Objet : Can't load driver dll

Hello,

I'm trying to run graphics on my system (Freescale P4080 PowerPC). I'm using a self made driverl derived from devg-flat.

To run my driver, I do the folowing :

# io-display  -c /mnt/sdhc/gf/display.conf -d vid=0x10ee,did=0x6024 Could not initialize device "vid=0x10ee,did=0x6024": No such file or directory Check 'sloginfo' output for more details io-display failed to initialise

# sloginfo -c

Time             Sev Major Minor Args

Aug 02 01:41:36    2     8   200 devg-flat.so: gfpr_dlopen: Library cannot be found

Here is the display.conf file content :

device {

        drivername=flat

        vid=0x10ee

        did=0x6024

        deviceindex=0

        memopts=/mnt/sdhc/gf/flat.conf

        display {

                xres=1920

                yres=1080

                refresh=30

                pixel_format=argb8888

        }

}

The driver devg-flat.so is in /mnt/sdhc/lib which is added to LD_LIBRARY_PATH. It is also added to procnto LD_LIBRARY_PATH in build file.

I spent hours to find why the library is not found.

Can anybody help me ?

Regards,

Nicolas

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post90693

post90693: Can't load driver dll

2011-12-20T16:11:25Z

Hello,

I'm trying to run graphics on my system (Freescale P4080 PowerPC). I'm using a self made driverl derived from devg-flat.

To run my driver, I do the folowing :

# io-display  -c /mnt/sdhc/gf/display.conf -d vid=0x10ee,did=0x6024

Could not initialize device "vid=0x10ee,did=0x6024": No such file or directory

Check 'sloginfo' output for more details

io-display failed to initialise

# sloginfo -c

Time             Sev Major Minor Args

Aug 02 01:41:36    2     8   200 devg-flat.so: gfpr_dlopen: Library cannot be found

Here is the display.conf file content :

device {

        drivername=flat

        vid=0x10ee

        did=0x6024

        deviceindex=0

        memopts=/mnt/sdhc/gf/flat.conf

        display {

                xres=1920

                yres=1080

                refresh=30

                pixel_format=argb8888

        }

}

The driver devg-flat.so is in /mnt/sdhc/lib which is added to LD_LIBRARY_PATH. It is also added to procnto LD_LIBRARY_PATH in build file.

I spent hours to find why the library is not found.

Can anybody help me ?

Regards,

Nicolas

post90683: Re: RE: devi-hid for armle-v7

2011-12-19T15:04:55Z

Do you solve you problem ?

Have you got devi-hid for armle-v7 ?

thanks, regards

post90352: Bitmap copy from src device context to dst context

2011-11-29T22:23:54Z

Hello,

I need to copy an entire frame buffer (src) to the screen (dst). Would the function PgContextBlitArea() do the trick?

Any help is appreciated.

Thanks,

-Zahira

post89931: Video Display from memory buffer

2011-11-08T10:39:20Z

Hi,

I am trying to display a video from a memory buffer into my display.

I will be using the Beagle BSP for OMAP3730.

Is there any functionality on GF to display video from the memory buffer?

Thanks in advance!

post89653: RE: RE: mmap_peer and mmap64_peer - Undefined reference error

2011-10-26T16:53:13Z

Those functions were written for some custom work and were never included in a general product release. The technote was erroneously included in the documentation, but has since been removed.

Steve Reid (stever@qnx.com)

Technical Editor

QNX Software Systems

-----Original Message-----

From: Nikshep Patil [mailto:community-noreply@qnx.com]

Sent: Wednesday, October 26, 2011 10:29 AM

To: advanced-graphics

Subject: Re: RE: mmap_peer and mmap64_peer - Undefined reference error

Any update on this...?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post89648

post89648: Re: RE: mmap_peer and mmap64_peer - Undefined reference error

2011-10-26T14:29:12Z

Any update on this...?

post89527: Re: gf_context_set_pattern function

2011-10-21T12:36:47Z

If I understand you correctly you should try something like this:

-- disable all ROP operation

-- set fgcolor to your "bgcolor"

-- draw a solid rectangle (gf_draw_rect()) where you need it, let's say @ X,Y

-- set the pattern

-- set the ROP

-- set fgcolor to your "fgcolor"

-- draw a solid rectangle again @ X,Y

post89520: Re: gf_context_set_pattern function

2011-10-21T09:40:02Z

Hi Misha,

thanks for your help.

I have an another question, how can I use this 8x8 value combine with ROPs(gf_context_set_rop());

for example,

I want to draw an 8x8 rectangle and set all the needed values like this,

mypattern[8] = {0x00, 0x44, 0x00,0x11, 0x00,0x44, 0x00,0x11 };

fgcolor:    0xFFFF0000

bgcolor:   0xFF0000FF

rop:          GF_ROP_PDSnaon

what is the formula to get the result ?

post89492: Re: RE: mmap_peer and mmap64_peer - Undefined reference error

2011-10-20T14:44:32Z

Thanks Thomas.

Does that mean I cannot use the mmap_peer()/mmap64_peer() anymore...?

Is there an alternative to these functions in 6.5...?

I see this documentation on mmap_peer functions -

http://www.qnx.com/developers/docs/6.4.1/neutrino/technotes/mmap_and_munmap_peer.html

post89490: RE: mmap_peer and mmap64_peer - Undefined reference error

2011-10-20T14:38:07Z

I may be mistaken, but I believe they aren't part of libc any more in 6.5.

Regards,

Thomas

-----Original Message-----

From: Nikshep Patil [mailto:community-noreply@qnx.com]

Sent: Donnerstag, 20. Oktober 2011 16:27

To: advanced-graphics

Subject: mmap_peer and mmap64_peer - Undefined reference error

I am trying to use mmap64_peer() in my resource manager.

I get "Undefined reference to mmap64_peer()" error.

Same with mmap_peer().

Including "sys/mman.h" does not help.

What header file/library do I need to include to get mmap64_peer() to work...?

Thank you.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post89488

post89488: mmap_peer and mmap64_peer - Undefined reference error

2011-10-20T14:27:25Z

I am trying to use mmap64_peer() in my resource manager.

I get "Undefined reference to mmap64_peer()" error.

Same with mmap_peer().

Including "sys/mman.h" does not help.

What header file/library do I need to include to get mmap64_peer() to work...?

Thank you.

post89451: Re: gf_context_set_pattern function

2011-10-19T13:38:51Z

The pattern is a pointer to an array of characters defining the pattern bitmap.

This array should define an 8X8 single bit/pixel bitmap packing 8 pixels/byte, with no padding.

post89441: gf_context_set_pattern function

2011-10-19T08:41:37Z

Hi ,

from the help documents, it says the second parameter of this function is pattern which is A pointer to the bitmap containing the pattern.

my question is how can I define the bitmap ?

post89358: Source code of vsync?

2011-10-13T07:23:08Z

I'm trying to showing some picture by using GF, my application seems to be running OK but there is nothing displayed on the screen.

However I run the vsync sample app and it works fine. Is it possible to get the source code of it? Or anybody could tell me about the details of how vsync works.

Your help would be appreciated.

post88943: Re: RE: devi-hid for armle-v7

2011-09-19T22:11:54Z

Brian,

Did you ever get anywhere with this? I'm also looking for devi-hid drivers for armle-v7

Thanks!

post88789: Re: Problem with gf_surface_attach()

2011-09-12T19:03:20Z

I recommend posting your code -- so others can look at the code and may spot what the problem is.

post88787: resolution change

2011-09-12T18:15:34Z

Hi,

Can anyone tell me how to change the resolution or display preferences in photon in code? I know “phgrafx” does this. But this is in photon GUI. But I want to implement the same functionality in my code.

In short, I like to know about the ways to reconfigure and restart photon with a new config file, after once the display is already been up.

Thanks,

Arun Johnson

post88774: Write an Image with raw pixel Data

2011-09-12T09:35:38Z

Hello :)

i want to write a .bmp file from raw pixel data.

My Programm i have, listen on the ethernet for my Client and accept the Client when he connects. The Client send some data and there are raw pixel data.

With a function named "const unsigned int* GetImage()" i get a pointer to the first pixel. My Pictures are 320(W) x 240(H).

Now how can i display or save the raw pixel data?

i hope for hints :)

with best regards

post88667: Problem with gf_surface_attach()

2011-09-06T20:55:55Z

Hi,

I need to render a picture to the screen.

I am trying to create a surface from my picture buffer using gf_surface_attach().

But the call fails with return value 7.

The arguments to gf_surface_attach() seem correct.

Could you help me with this...?

Thanks

post88552: io-winmgr sendev

2011-08-31T20:10:38Z

Hi,

Is there a way to use io-winmgr to bring a window to the front?

win_globals.disp_data[0]->win_list = 0x80000003, 0x80000000, 0x80000002

win_globals.disp_data[0]->planes[0].buffered[0].win_list = (null)

win_globals.disp_data[0]->planes[0].buffered[1].win_list = (null)

win_globals.disp_data[0]->planes[1].buffered[0].win_list = 0x80000002, 0x80000003

win_globals.disp_data[0]->planes[1].buffered[1].win_list = (null)

I see from its dump the window IDs.

Can I send a msg to io-winmgr to bring window e.g. 0x80000002 or 3 to top? Using io-winmgr –sendev ???

-Hue

post88200: RE: devi-hid for armle-v7

2011-08-18T15:24:01Z

To be honest I'm not sure when 6.5.0 was released any platforms with

graphics were actually running armle-v7. A couple of customers

running armle-v7 needed custom drivers and wrote their own.

I checked whether the input DDK posted on the QNX website could be used

to build a driver and was told:

"The DDK won't work on 6.5.0, better to just get the imx25 BSP:

http://community.qnx.com/sf/wiki/do/viewPage/projects.bsp/wiki/Freescale

Imx253dsPdk

and grab the source archive from it, it contains the input libs with the

gf support and a sample driver."

Maybe this is something you can try?

Mike

-----Original Message-----

From: Brian Edmond [mailto:community-noreply@qnx.com]

Sent: August 17, 2011 1:52 PM

To: advanced-graphics

Subject: devi-hid for armle-v7

Probably not the correct place for this but the input drivers were

always kinda graphics. There are no devi drivers for armle-v7 targets.

How does one get a mouse or keyboard input in 6.5?

Brian

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88172

post88180: RE: Creating a PNG output file. [bcc][auto-ip]

2011-08-18T07:15:03Z

No idea how you can change the compression /quality :(,

Check if img_load_resize_file() is of any use....

Regards,

Lavin

-----Original Message-----

From: Ronald Burkey [mailto:community-noreply@qnx.com]

Sent: Wednesday, August 17, 2011 10:38 PM

To: advanced-graphics

Subject: Re: Creating a PNG output file. [bcc][auto-ip]

True. I don't lack space, so that doesn't actually concern me. But I

gather you don't know how it can be done, so the point is moot.

Thanks,

Ron

On 08/17/2011 12:02 PM, Lavin Pottekkat wrote:

> Though i have not tried changing the compression techniques, i have seen noticeable difference in the size of the images created with different formats.

>

> Regards,

> Lavin

> ________________________________________

> From: Ronald Burkey [community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 9:52 PM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file.

>

> One final (I hope) question: If JPGs were being output rather than

> PNGs, how can the quality/compression be set?

>

> -- Ron

>

>

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88170

post88172: devi-hid for armle-v7

2011-08-17T17:51:32Z

Probably not the correct place for this but the input drivers were always kinda graphics. There are no devi drivers for armle-v7 targets. How does one get a mouse or keyboard input in 6.5?

Brian

post88170: Re: Creating a PNG output file. [bcc][auto-ip]

2011-08-17T17:08:18Z

True. I don't lack space, so that doesn't actually concern me. But I

gather you don't know how it can be done, so the point is moot.

Thanks,

Ron

On 08/17/2011 12:02 PM, Lavin Pottekkat wrote:

> Though i have not tried changing the compression techniques, i have seen noticeable difference in the size of the images created with different formats.

>

> Regards,

> Lavin

> ________________________________________

> From: Ronald Burkey [community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 9:52 PM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file.

>

> One final (I hope) question: If JPGs were being output rather than

> PNGs, how can the quality/compression be set?

>

> -- Ron

>

>

post88169: RE: Creating a PNG output file.

2011-08-17T17:02:02Z

Though i have not tried changing the compression techniques, i have seen noticeable difference in the size of the images created with different formats.

Regards,

Lavin

________________________________________

From: Ronald Burkey [community-noreply@qnx.com]

Sent: Wednesday, August 17, 2011 9:52 PM

To: advanced-graphics

Subject: RE: Creating a PNG output file.

One final (I hope) question:  If JPGs were being output rather than

PNGs, how can the quality/compression be set?

-- Ron

> -----Original Message-----

> From: Lavin Pottekkat [mailto:community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 7:53 AM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file. [bcc]

>

> Yes, you can change the mime type for different file formats.

>

> Regards,

> Lavin

>

> -----Original Message-----

> From: Ronald Burkey [mailto:community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 6:10 PM

> To: advanced-graphics

> Subject: Re: Creating a PNG output file. [bcc]

>

> I had just made that change about 10 seconds before your message

> arrived. :-) I haven't gotten to the point yet where I can run the

> code.  I'll let you know what happens.

>

> I assume that if I want jpg or bmp, it's just a matter of changing the

> mime type appropriately?

>

> Thanks!

> Ron

>

>

> On 08/17/2011 07:34 AM, Lavin Pottekkat wrote:

> > Just now I noticed that you have RGBA8888, in this case I think you

> should modify below lines highlighted below using '-->'

> > Also pl make sure you configure the img.w/h appropriately.

> >

> > Cheers,

> > Lavin

> >

> > -----Original Message-----

> > From: Pottekkat, Lavin

> > Sent: Wednesday, August 17, 2011 5:43 PM

> > To: advanced-graphics

> > Subject: RE: Creating a PNG output file.

> >

> > Hi Ronald,

> >

> > Try this it should work :) .... search for "Lavin" to see my

comments

> inline....

> >

> >     uint32_t *img_buf = NULL;

> >     uint32_t image_size = Max size of the RGB data buffer;

> >     int32_t           result = EOK;

> >

> >     img_lib_t         ilib;

> >     img_codec_t       codec_buf;

> >     int               n_codecs;

> >     img_t             img;

> >     io_stream_t       *image_stream_fd;

> >     char              mime[64];

> >     uintptr_t         temp_encode_data = 0;

> >

> >     /* allocate (image_size/sizeof int-(4bytes)) elements of type

int */

> >     if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))),

> sizeof(uint32_t)))) {

> >        _error("%s: Memory allocation for %d bytes failed",

__FUNCTION__,

> image_size);

> >        return ENOMEM;

> >     }

> >

> >   /* Lavin : copy the image RGB values linearly to the above

allocated

> img_buf, from your rectangular array in memory */

> >

> >   /* Lavin : at this point I assume that you have already copied the

> image data to img_buf.*/

> >

> >   /* Lavin : NOTE: you can very well use your rectangular array

directly

> as image_buf as well :) */

> >

> >

> >        //Attaching to the image library

> >        memset(&img,0,sizeof(img_t));

> >

> >        /* get the resolution details */

> >        img.w = 1440; // 1440 or 960 , use accordingly

> >        img.h = 540; // 540, use accordingly

> >

> >        img.format =  IMG_FMT_RGBA8888;

> >

> >        img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

> >

> >        img.access.direct.data = img_buf;

> > -->       img.access.direct.stride = (img.w * 4 );

> >

> >        if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

> >           fprintf(stderr, "img_lib_attach() failed: %d\n", result);

> >           return result;

> >        }

> >

> >        memset(mime, '\0', 64 );

> >        sprintf(mime, "image/%s", "png");

> >

> >        if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

> >                               mime,

> >                               &codec_buf,

> >                               10 )))

> >        {

> >           _error("%s: img_codec_list() No Codecs found on target !",

> __FUNCTION__);

> >           return -1;

> >        }

> >

> >        //convert the image_fd to stream_fd required by image libs.

> >     /* Lavin : here file_fs should be your xxx.png file fd after

> open(...O_CREATE) */

> >        if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE,

file_fd

> )))

> >        {

> >           _error("%s: io_open() failed: %d", __FUNCTION__);

> >           return -1;

> >        }

> >

> >        //set the codec

> >

> >        //Prepare to encode a frame to a stream

> >        if (img_encode_begin(codec_buf,

> image_stream_fd,&temp_encode_data) != IMG_ERR_OK)

> >        {

> >           _error("%s: img_encode_begin() failed", __FUNCTION__);

> >           return -1;

> >        }

> >

> >        if ( (result = img_encode_frame(codec_buf, image_stream_fd,

> NULL,&img,&temp_encode_data)) != IMG_ERR_OK)

> >        {

> >           _error("%s: img_encode_frame() failed: %d", __FUNCTION__,

> result);

> >           return -1;

> >        }

> >

> >        (void)img_encode_finish(codec_buf,

> image_stream_fd,&temp_encode_data);

> >

> >        io_close( image_stream_fd );

> >

> >        /* done with img lib */

> >        img_lib_detach(ilib);

> >

> >     /* Lavin : Now your PNG fromat file is ready :) :) */

> >

> >

> > Regards,

> > Lavin,

> >

> > -----Original Message-----

> > From: Ronald Burkey [mailto:community-noreply@qnx.com]

> > Sent: Friday, August 12, 2011 9:52 PM

> > To: advanced-graphics

> > Subject: Creating a PNG output file.

> >

> > I have a rectangular array in memory of RGB values (i.e., it is

> IMG_FMT_RGBA8888 data) and would like to output a PNG file from that

data.

> How would this be done?

> >

> >

> >

> >

> >

> > _______________________________________________

> >

> > Advanced Graphics

> > http://community.qnx.com/sf/go/post88079

> >

> >

> >

> >

> > _______________________________________________

> >

> > Advanced Graphics

> > http://community.qnx.com/sf/go/post88149

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88150

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88152

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88168

post88168: RE: Creating a PNG output file.

2011-08-17T16:22:11Z

One final (I hope) question:  If JPGs were being output rather than

PNGs, how can the quality/compression be set?

-- Ron

> -----Original Message-----

> From: Lavin Pottekkat [mailto:community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 7:53 AM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file. [bcc]

>

> Yes, you can change the mime type for different file formats.

>

> Regards,

> Lavin

>

> -----Original Message-----

> From: Ronald Burkey [mailto:community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 6:10 PM

> To: advanced-graphics

> Subject: Re: Creating a PNG output file. [bcc]

>

> I had just made that change about 10 seconds before your message

> arrived. :-) I haven't gotten to the point yet where I can run the

> code.  I'll let you know what happens.

>

> I assume that if I want jpg or bmp, it's just a matter of changing the

> mime type appropriately?

>

> Thanks!

> Ron

>

>

> On 08/17/2011 07:34 AM, Lavin Pottekkat wrote:

> > Just now I noticed that you have RGBA8888, in this case I think you

> should modify below lines highlighted below using '-->'

> > Also pl make sure you configure the img.w/h appropriately.

> >

> > Cheers,

> > Lavin

> >

> > -----Original Message-----

> > From: Pottekkat, Lavin

> > Sent: Wednesday, August 17, 2011 5:43 PM

> > To: advanced-graphics

> > Subject: RE: Creating a PNG output file.

> >

> > Hi Ronald,

> >

> > Try this it should work :) .... search for "Lavin" to see my

comments

> inline....

> >

> >     uint32_t *img_buf = NULL;

> >     uint32_t image_size = Max size of the RGB data buffer;

> >     int32_t           result = EOK;

> >

> >     img_lib_t         ilib;

> >     img_codec_t       codec_buf;

> >     int               n_codecs;

> >     img_t             img;

> >     io_stream_t       *image_stream_fd;

> >     char              mime[64];

> >     uintptr_t         temp_encode_data = 0;

> >

> >     /* allocate (image_size/sizeof int-(4bytes)) elements of type

int */

> >     if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))),

> sizeof(uint32_t)))) {

> >        _error("%s: Memory allocation for %d bytes failed",

__FUNCTION__,

> image_size);

> >        return ENOMEM;

> >     }

> >

> >   /* Lavin : copy the image RGB values linearly to the above

allocated

> img_buf, from your rectangular array in memory */

> >

> >   /* Lavin : at this point I assume that you have already copied the

> image data to img_buf.*/

> >

> >   /* Lavin : NOTE: you can very well use your rectangular array

directly

> as image_buf as well :) */

> >

> >

> >        //Attaching to the image library

> >        memset(&img,0,sizeof(img_t));

> >

> >        /* get the resolution details */

> >        img.w = 1440; // 1440 or 960 , use accordingly

> >        img.h = 540; // 540, use accordingly

> >

> >        img.format =  IMG_FMT_RGBA8888;

> >

> >        img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

> >

> >        img.access.direct.data = img_buf;

> > -->       img.access.direct.stride = (img.w * 4 );

> >

> >        if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

> >           fprintf(stderr, "img_lib_attach() failed: %d\n", result);

> >           return result;

> >        }

> >

> >        memset(mime, '\0', 64 );

> >        sprintf(mime, "image/%s", "png");

> >

> >        if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

> >                               mime,

> >                               &codec_buf,

> >                               10 )))

> >        {

> >           _error("%s: img_codec_list() No Codecs found on target !",

> __FUNCTION__);

> >           return -1;

> >        }

> >

> >        //convert the image_fd to stream_fd required by image libs.

> > /* Lavin : here file_fs should be your xxx.png file fd after

> open(...O_CREATE) */

> >        if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE,

file_fd

> )))

> >        {

> >           _error("%s: io_open() failed: %d", __FUNCTION__);

> >           return -1;

> >        }

> >

> >        //set the codec

> >

> >        //Prepare to encode a frame to a stream

> >        if (img_encode_begin(codec_buf,

> image_stream_fd,&temp_encode_data) != IMG_ERR_OK)

> >        {

> >           _error("%s: img_encode_begin() failed", __FUNCTION__);

> >           return -1;

> >        }

> >

> >        if ( (result = img_encode_frame(codec_buf, image_stream_fd,

> NULL,&img,&temp_encode_data)) != IMG_ERR_OK)

> >        {

> >           _error("%s: img_encode_frame() failed: %d", __FUNCTION__,

> result);

> >           return -1;

> >        }

> >

> >        (void)img_encode_finish(codec_buf,

> image_stream_fd,&temp_encode_data);

> >

> >        io_close( image_stream_fd );

> >

> >        /* done with img lib */

> >        img_lib_detach(ilib);

> >

> > /* Lavin : Now your PNG fromat file is ready :) :) */

> >

> >

> > Regards,

> > Lavin,

> >

> > -----Original Message-----

> > From: Ronald Burkey [mailto:community-noreply@qnx.com]

> > Sent: Friday, August 12, 2011 9:52 PM

> > To: advanced-graphics

> > Subject: Creating a PNG output file.

> >

> > I have a rectangular array in memory of RGB values (i.e., it is

> IMG_FMT_RGBA8888 data) and would like to output a PNG file from that

data.

> How would this be done?

> >

> >

> >

> >

> >

> > _______________________________________________

> >

> > Advanced Graphics

> > http://community.qnx.com/sf/go/post88079

> >

> >

> >

> >

> > _______________________________________________

> >

> > Advanced Graphics

> > http://community.qnx.com/sf/go/post88149

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88150

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88152

post88158: Re: Creating a PNG output file. [bcc]

2011-08-17T14:24:37Z

Thanks, Lavin. It seems to work fine, after appropriate changes for my

environment.

On 08/17/2011 07:53 AM, Lavin Pottekkat wrote:

> Yes, you can change the mime type for different file formats.

>

> Regards,

> Lavin

>

> -----Original Message-----

> From: Ronald Burkey [mailto:community-noreply@qnx.com]

> Sent: Wednesday, August 17, 2011 6:10 PM

> To: advanced-graphics

> Subject: Re: Creating a PNG output file. [bcc]

>

> I had just made that change about 10 seconds before your message

> arrived. :-) I haven't gotten to the point yet where I can run the

> code. I'll let you know what happens.

>

> I assume that if I want jpg or bmp, it's just a matter of changing the

> mime type appropriately?

>

> Thanks!

> Ron

>

>

>

post88152: RE: Creating a PNG output file. [bcc]

2011-08-17T12:53:15Z

Yes, you can change the mime type for different file formats.

Regards,

Lavin

-----Original Message-----

From: Ronald Burkey [mailto:community-noreply@qnx.com]

Sent: Wednesday, August 17, 2011 6:10 PM

To: advanced-graphics

Subject: Re: Creating a PNG output file. [bcc]

I had just made that change about 10 seconds before your message

arrived. :-) I haven't gotten to the point yet where I can run the

code.  I'll let you know what happens.

I assume that if I want jpg or bmp, it's just a matter of changing the

mime type appropriately?

Thanks!

Ron

On 08/17/2011 07:34 AM, Lavin Pottekkat wrote:

> Just now I noticed that you have RGBA8888, in this case I think you should modify below lines highlighted below using '-->'

> Also pl make sure you configure the img.w/h appropriately.

>

> Cheers,

> Lavin

>

> -----Original Message-----

> From: Pottekkat, Lavin

> Sent: Wednesday, August 17, 2011 5:43 PM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file.

>

> Hi Ronald,

>

> Try this it should work :) .... search for "Lavin" to see my comments inline....

>

>     uint32_t *img_buf = NULL;

>     uint32_t image_size = Max size of the RGB data buffer;

>     int32_t           result = EOK;

>

>     img_lib_t         ilib;

>     img_codec_t       codec_buf;

>     int               n_codecs;

>     img_t             img;

>     io_stream_t       *image_stream_fd;

>     char              mime[64];

>     uintptr_t         temp_encode_data = 0;

>

>     /* allocate (image_size/sizeof int-(4bytes)) elements of type int */

>     if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))), sizeof(uint32_t)))) {

>        _error("%s: Memory allocation for %d bytes failed", __FUNCTION__, image_size);

>        return ENOMEM;

>     }

>

>   /* Lavin : copy the image RGB values linearly to the above allocated img_buf, from your rectangular array in memory */

>

>   /* Lavin : at this point I assume that you have already copied the image data to img_buf.*/

>

>   /* Lavin : NOTE: you can very well use your rectangular array directly as image_buf as well :) */

>

>

>        //Attaching to the image library

>        memset(&img,0,sizeof(img_t));

>

>        /* get the resolution details */

>        img.w = 1440; // 1440 or 960 , use accordingly

>        img.h = 540; // 540, use accordingly

>

>        img.format =  IMG_FMT_RGBA8888;

>

>        img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

>

>        img.access.direct.data = img_buf;

> -->       img.access.direct.stride = (img.w * 4 );

>

>        if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

>           fprintf(stderr, "img_lib_attach() failed: %d\n", result);

>           return result;

>        }

>

>        memset(mime, '\0', 64 );

>        sprintf(mime, "image/%s", "png");

>

>        if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

>                               mime,

>                               &codec_buf,

>                               10 )))

>        {

>           _error("%s: img_codec_list() No Codecs found on target !", __FUNCTION__);

>           return -1;

>        }

>

>        //convert the image_fd to stream_fd required by image libs.

> /* Lavin : here file_fs should be your xxx.png file fd after open(...O_CREATE) */

>        if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE, file_fd )))

>        {

>           _error("%s: io_open() failed: %d", __FUNCTION__);

>           return -1;

>        }

>

>        //set the codec

>

>        //Prepare to encode a frame to a stream

>        if (img_encode_begin(codec_buf, image_stream_fd,&temp_encode_data) != IMG_ERR_OK)

>        {

>           _error("%s: img_encode_begin() failed", __FUNCTION__);

>           return -1;

>        }

>

>        if ( (result = img_encode_frame(codec_buf, image_stream_fd, NULL,&img,&temp_encode_data)) != IMG_ERR_OK)

>        {

>           _error("%s: img_encode_frame() failed: %d", __FUNCTION__, result);

>           return -1;

>        }

>

>        (void)img_encode_finish(codec_buf, image_stream_fd,&temp_encode_data);

>

>        io_close( image_stream_fd );

>

>        /* done with img lib */

>        img_lib_detach(ilib);

>

> /* Lavin : Now your PNG fromat file is ready :) :) */

>

>

> Regards,

> Lavin,

>

> -----Original Message-----

> From: Ronald Burkey [mailto:community-noreply@qnx.com]

> Sent: Friday, August 12, 2011 9:52 PM

> To: advanced-graphics

> Subject: Creating a PNG output file.

>

> I have a rectangular array in memory of RGB values (i.e., it is IMG_FMT_RGBA8888 data) and would like to output a PNG file from that data.  How would this be done?

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88079

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88149

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88150

post88150: Re: Creating a PNG output file. [bcc]

2011-08-17T12:40:07Z

I had just made that change about 10 seconds before your message

arrived. :-) I haven't gotten to the point yet where I can run the

code.  I'll let you know what happens.

I assume that if I want jpg or bmp, it's just a matter of changing the

mime type appropriately?

Thanks!

Ron

On 08/17/2011 07:34 AM, Lavin Pottekkat wrote:

> Just now I noticed that you have RGBA8888, in this case I think you should modify below lines highlighted below using '-->'

> Also pl make sure you configure the img.w/h appropriately.

>

> Cheers,

> Lavin

>

> -----Original Message-----

> From: Pottekkat, Lavin

> Sent: Wednesday, August 17, 2011 5:43 PM

> To: advanced-graphics

> Subject: RE: Creating a PNG output file.

>

> Hi Ronald,

>

> Try this it should work :) .... search for "Lavin" to see my comments inline....

>

>     uint32_t *img_buf = NULL;

>     uint32_t image_size = Max size of the RGB data buffer;

>     int32_t           result = EOK;

>

>     img_lib_t         ilib;

>     img_codec_t       codec_buf;

>     int               n_codecs;

>     img_t             img;

>     io_stream_t       *image_stream_fd;

>     char              mime[64];

>     uintptr_t         temp_encode_data = 0;

>

>     /* allocate (image_size/sizeof int-(4bytes)) elements of type int */

>     if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))), sizeof(uint32_t)))) {

>        _error("%s: Memory allocation for %d bytes failed", __FUNCTION__, image_size);

>        return ENOMEM;

>     }

>

>   /* Lavin : copy the image RGB values linearly to the above allocated img_buf, from your rectangular array in memory */

>

>   /* Lavin : at this point I assume that you have already copied the image data to img_buf.*/

>

>   /* Lavin : NOTE: you can very well use your rectangular array directly as image_buf as well :) */

>

>

>        //Attaching to the image library

>        memset(&img,0,sizeof(img_t));

>

>        /* get the resolution details */

>        img.w = 1440; // 1440 or 960 , use accordingly

>        img.h = 540; // 540, use accordingly

>

>        img.format =  IMG_FMT_RGBA8888;

>

>        img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

>

>        img.access.direct.data = img_buf;

> -->       img.access.direct.stride = (img.w * 4 );

>

>        if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

>           fprintf(stderr, "img_lib_attach() failed: %d\n", result);

>           return result;

>        }

>

>        memset(mime, '\0', 64 );

>        sprintf(mime, "image/%s", "png");

>

>        if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

>                               mime,

>                               &codec_buf,

>                               10 )))

>        {

>           _error("%s: img_codec_list() No Codecs found on target !", __FUNCTION__);

>           return -1;

>        }

>

>        //convert the image_fd to stream_fd required by image libs.

> /* Lavin : here file_fs should be your xxx.png file fd after open(...O_CREATE) */

>        if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE, file_fd )))

>        {

>           _error("%s: io_open() failed: %d", __FUNCTION__);

>           return -1;

>        }

>

>        //set the codec

>

>        //Prepare to encode a frame to a stream

>        if (img_encode_begin(codec_buf, image_stream_fd,&temp_encode_data) != IMG_ERR_OK)

>        {

>           _error("%s: img_encode_begin() failed", __FUNCTION__);

>           return -1;

>        }

>

>        if ( (result = img_encode_frame(codec_buf, image_stream_fd, NULL,&img,&temp_encode_data)) != IMG_ERR_OK)

>        {

>           _error("%s: img_encode_frame() failed: %d", __FUNCTION__, result);

>           return -1;

>        }

>

>        (void)img_encode_finish(codec_buf, image_stream_fd,&temp_encode_data);

>

>        io_close( image_stream_fd );

>

>        /* done with img lib */

>        img_lib_detach(ilib);

>

> /* Lavin : Now your PNG fromat file is ready :) :) */

>

>

> Regards,

> Lavin,

>

> -----Original Message-----

> From: Ronald Burkey [mailto:community-noreply@qnx.com]

> Sent: Friday, August 12, 2011 9:52 PM

> To: advanced-graphics

> Subject: Creating a PNG output file.

>

> I have a rectangular array in memory of RGB values (i.e., it is IMG_FMT_RGBA8888 data) and would like to output a PNG file from that data.  How would this be done?

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88079

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post88149

post88149: RE: Creating a PNG output file.

2011-08-17T12:34:51Z

Just now I noticed that you have RGBA8888, in this case I think you should modify below lines highlighted below using '-->'

Also pl make sure you configure the img.w/h appropriately.

Cheers,

Lavin

-----Original Message-----

From: Pottekkat, Lavin

Sent: Wednesday, August 17, 2011 5:43 PM

To: advanced-graphics

Subject: RE: Creating a PNG output file.

Hi Ronald,

Try this it should work :) .... search for "Lavin" to see my comments inline....

   uint32_t *img_buf = NULL;

   uint32_t image_size = Max size of the RGB data buffer;

   int32_t           result = EOK;

   img_lib_t         ilib;

   img_codec_t       codec_buf;

   int               n_codecs;

   img_t             img;

   io_stream_t       *image_stream_fd;

   char              mime[64];

   uintptr_t         temp_encode_data = 0;

   /* allocate (image_size/sizeof int-(4bytes)) elements of type int */

   if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))), sizeof(uint32_t)))) {

      _error("%s: Memory allocation for %d bytes failed", __FUNCTION__, image_size);

      return ENOMEM;

   }

/* Lavin : copy the image RGB values linearly to the above allocated img_buf, from your rectangular array in memory */

/* Lavin : at this point I assume that you have already copied the image data to img_buf.*/

/* Lavin : NOTE: you can very well use your rectangular array directly as image_buf as well :) */

      //Attaching to the image library

      memset(&img,0,sizeof(img_t));

      /* get the resolution details */

      img.w = 1440; // 1440 or 960 , use accordingly

      img.h = 540; // 540, use accordingly

      img.format =  IMG_FMT_RGBA8888;

      img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

      img.access.direct.data = img_buf;

-->      img.access.direct.stride = (img.w * 4 );

      if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

         fprintf(stderr, "img_lib_attach() failed: %d\n", result);

         return result;

      }

      memset(mime, '\0', 64 );

      sprintf(mime, "image/%s", "png");

      if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

                             mime,

                             &codec_buf,

                             10 )))

      {

         _error("%s: img_codec_list() No Codecs found on target !", __FUNCTION__);

         return -1;

      }

      //convert the image_fd to stream_fd required by image libs.

/* Lavin : here file_fs should be your xxx.png file fd after open(...O_CREATE) */

      if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE, file_fd )))

      {

         _error("%s: io_open() failed: %d", __FUNCTION__);

         return -1;

      }

      //set the codec

      //Prepare to encode a frame to a stream

      if (img_encode_begin(codec_buf, image_stream_fd, &temp_encode_data) != IMG_ERR_OK)

      {

         _error("%s: img_encode_begin() failed", __FUNCTION__);

         return -1;

      }

      if ( (result = img_encode_frame(codec_buf, image_stream_fd, NULL, &img, &temp_encode_data)) != IMG_ERR_OK)

      {

         _error("%s: img_encode_frame() failed: %d", __FUNCTION__, result);

         return -1;

      }

      (void)img_encode_finish(codec_buf, image_stream_fd,  &temp_encode_data);

      io_close( image_stream_fd );

      /* done with img lib */

      img_lib_detach(ilib);

/* Lavin : Now your PNG fromat file is ready :) :) */

Regards,

Lavin,

-----Original Message-----

From: Ronald Burkey [mailto:community-noreply@qnx.com]

Sent: Friday, August 12, 2011 9:52 PM

To: advanced-graphics

Subject: Creating a PNG output file.

I have a rectangular array in memory of RGB values (i.e., it is IMG_FMT_RGBA8888 data) and would like to output a PNG file from that data.  How would this be done?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88079

post88148: RE: Creating a PNG output file.

2011-08-17T12:14:51Z

Hi Ronald,

Try this it should work :) .... search for "Lavin" to see my comments inline....

   uint32_t *img_buf = NULL;

   uint32_t image_size = Max size of the RGB data buffer;

   int32_t           result = EOK;

   img_lib_t         ilib;

   img_codec_t       codec_buf;

   int               n_codecs;

   img_t             img;

   io_stream_t       *image_stream_fd;

   char              mime[64];

   uintptr_t         temp_encode_data = 0;

   /* allocate (image_size/sizeof int-(4bytes)) elements of type int */

   if (NULL == (img_buf = calloc((image_size / (sizeof(uint32_t))), sizeof(uint32_t)))) {

      _error("%s: Memory allocation for %d bytes failed", __FUNCTION__, image_size);

      return ENOMEM;

   }

/* Lavin : copy the image RGB values linearly to the above allocated img_buf, from your rectangular array in memory */

/* Lavin : at this point I assume that you have already copied the image data to img_buf.*/

/* Lavin : NOTE: you can very well use your rectangular array directly as image_buf as well :) */

      //Attaching to the image library

      memset(&img,0,sizeof(img_t));

      /* get the resolution details */

      img.w = 1440; // 1440 or 960 , use accordingly

      img.h = 540; // 540, use accordingly

      img.format =  IMG_FMT_RGBA8888;

      img.flags |= IMG_FORMAT | IMG_W | IMG_H | IMG_DIRECT;

      img.access.direct.data = img_buf;

      img.access.direct.stride = (img.w * 3 );

      if ((result = img_lib_attach(&ilib)) != IMG_ERR_OK) {

         fprintf(stderr, "img_lib_attach() failed: %d\n", result);

         return result;

      }

      memset(mime, '\0', 64 );

      sprintf(mime, "image/%s", "png");

      if(EOK == (n_codecs =  img_codec_list_bymime( ilib,

                             mime,

                             &codec_buf,

                             10 )))

      {

         _error("%s: img_codec_list() No Codecs found on target !", __FUNCTION__);

         return -1;

      }

      //convert the image_fd to stream_fd required by image libs.

/* Lavin : here file_fs should be your xxx.png file fd after open(...O_CREATE) */

      if(NULL == (image_stream_fd = io_open( IO_FD, IO_WRITE, file_fd )))

      {

         _error("%s: io_open() failed: %d", __FUNCTION__);

         return -1;

      }

      //set the codec

      //Prepare to encode a frame to a stream

      if (img_encode_begin(codec_buf, image_stream_fd, &temp_encode_data) != IMG_ERR_OK)

      {

         _error("%s: img_encode_begin() failed", __FUNCTION__);

         return -1;

      }

      if ( (result = img_encode_frame(codec_buf, image_stream_fd, NULL, &img, &temp_encode_data)) != IMG_ERR_OK)

      {

         _error("%s: img_encode_frame() failed: %d", __FUNCTION__, result);

         return -1;

      }

      (void)img_encode_finish(codec_buf, image_stream_fd,  &temp_encode_data);

      io_close( image_stream_fd );

      /* done with img lib */

      img_lib_detach(ilib);

/* Lavin : Now your PNG fromat file is ready :) :) */

Regards,

Lavin,

-----Original Message-----

From: Ronald Burkey [mailto:community-noreply@qnx.com]

Sent: Friday, August 12, 2011 9:52 PM

To: advanced-graphics

Subject: Creating a PNG output file.

I have a rectangular array in memory of RGB values (i.e., it is IMG_FMT_RGBA8888 data) and would like to output a PNG file from that data.  How would this be done?

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post88079

post88107: RE: gf_layer_attach() fails.

2011-08-15T20:56:48Z

Photon (and io-graphics) is running and there's only 1 layer on the

system so your app fails to attach to the layer.

-----Original Message-----

From: Hieronim Fet [mailto:community-noreply@qnx.com]

Sent: August-09-11 6:49 PM

To: advanced-graphics

Subject: gf_layer_attach() fails.

Hello everyone,

I`m having a problem with example program that draws rectangle. The code

is from topic

http://community.qnx.com/sf/discussion/do/listPosts/projects.graphics/di

scussion.advanced_graphics.topc8189?_pagenum=1,

but I enclose copy of source without long comments so it is easier to

look through.

As mentioned in topic of this message there is problem with

gf_layer_attach() function that fails. The output of the running program

is:

Number of displays: 1

Display 0: 1280X1024, refresh = 60Hz

Number of layers: 1

gf_layer_attach() failed

I`m sending sloginfo and pidin outputs and my display.conf.

I have no idea why gf_layer_attach fails, hope you will help me.

Generally I`m working in Momentics IDE, so if I run this program as QNX

C/C++ App will I see any "drawn rectangles" or should I run the program

directly under QNX? Maybe this question is silly but I`m just a

beginner.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post87952

post88079: Creating a PNG output file.

2011-08-12T16:22:11Z

I have a rectangular array in memory of RGB values (i.e., it is IMG_FMT_RGBA8888 data) and would like to output a PNG file from that data. How would this be done?

post87957: RE: gf_layer_attach() fails.

2011-08-10T07:05:29Z

Hi,

I don't immediately see a reason for failure, either. You should examine

the return value of gf_layer_attach() more closely to see what the error

was. Perhaps another instance of your application was already running?

In that case, you'd get an GF_ERR_INUSE.

You will not see any rectangles drawn on your Windows/Linux development

host; QNX executables simple will not run there. They can only execute

on systems running the QNX Neutrino OS.

To run your program from the Momentics IDE, you'll want to use a "C/C++

QNX QConn (IP)" launch configuration; this will automatically transfer

your program to the defined target and run it there. As well, you could

transfer the binary to your target and run it there directly.

Hope this helps,

Thomas

-----Original Message-----

From: Hieronim Fet [mailto:community-noreply@qnx.com]

Sent: Mittwoch, 10. August 2011 00:49

To: advanced-graphics

Subject: gf_layer_attach() fails.

Hello everyone,

I`m having a problem with example program that draws rectangle. The code

is from topic

http://community.qnx.com/sf/discussion/do/listPosts/projects.graphics/di

scussion.advanced_graphics.topc8189?_pagenum=1,

but I enclose copy of source without long comments so it is easier to

look through.

As mentioned in topic of this message there is problem with

gf_layer_attach() function that fails. The output of the running program

is:

Number of displays: 1

Display 0: 1280X1024, refresh = 60Hz

Number of layers: 1

gf_layer_attach() failed

I`m sending sloginfo and pidin outputs and my display.conf.

I have no idea why gf_layer_attach fails, hope you will help me.

Generally I`m working in Momentics IDE, so if I run this program as QNX

C/C++ App will I see any "drawn rectangles" or should I run the program

directly under QNX? Maybe this question is silly but I`m just a

beginner.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post87952

post87954: Re: gf_layer_attach() fails.

2011-08-09T22:54:28Z

Sorry for posting twice, my mistake.

post87953: gf_layer_attach() fails.

2011-08-09T22:53:00Z

Hello everyone,

I`m having a problem with example program that draws rectangle. The code is from topic http://community.qnx.com/sf/discussion/do/listPosts/projects.graphics/discussion.advanced_graphics.topc8189?_pagenum=1, but I enclose copy of source without long comments so it is easier to look through.

As mentioned in topic of this message there is problem with gf_layer_attach() function that fails. The output of the running program is:

Number of displays: 1

Display 0: 1280X1024, refresh = 60Hz

Number of layers: 1

gf_layer_attach() failed

I`m sending sloginfo and pidin outputs and my display.conf.

I have no idea why gf_layer_attach fails, hope you will help me.

Generally I`m working in Momentics IDE, so if I run this program as QNX C/C++ App will I see any "drawn rectangles" or should I run the program directly under QNX? Maybe this question is silly but I`m just a beginner.

post87952: gf_layer_attach() fails.

2011-08-09T22:49:26Z

Hello everyone,

I`m having a problem with example program that draws rectangle. The code is

from topic

http://community.qnx.com/sf/discussion/do/listPosts/projects.graphics/discussion.advanced_graphics.topc8189?_pagenum=1,

but I enclose copy of source without long comments so it is easier to look

through.

As mentioned in topic of this message there is problem with

gf_layer_attach() function that fails. The output of the running program is:

Number of displays: 1

Display 0: 1280X1024, refresh = 60Hz

Number of layers: 1

gf_layer_attach() failed

I`m sending sloginfo and pidin outputs and my display.conf.

I have no idea why gf_layer_attach fails, hope you will help me.

Generally I`m working in Momentics IDE, so if I run this program as QNX

C/C++ App will I see any "drawn rectangles" or should I run the program

directly under QNX? Maybe this question is silly but I`m just a beginner.

post87581: RE: Display problem for iMx51 - vsync and eglgears not working

2011-07-26T04:42:08Z

Am using LCD, i had a reference driver for LCD.

I was able to integrate and it is working fine.

Now am able to run few example applications.

Thank you.

Has anyone done a Photon GUI environment on the iMx51 or iMx53 platform.

Kishore.S

>

> What kind of display (DVI, VGA, LCD) are you using?

>

> Thanks

>

>

> -----Original Message-----

> From: kishore s [mailto:community-noreply@qnx.com]

> Sent: July 22, 2011 2:43 AM

> To: advanced-graphics

> Subject: Display problem for iMx51 - vsync and eglgears not working

>

> Hi,

>

> I am working on iMx51 reference board from Freescale.

>

> When i run the vsync or egl-gears application nothing is comming up on

> the

> display.

>

> When i run the egl application i get the below prints.

> egl-gears

> 185 frames in  5.027 seconds = 36.801 FPS

> 184 frames in  5.027 seconds = 36.602 FPS

> 218 frames in  5.946 seconds = 36.663 FPS

>

> I am not getting any errors also when i run sloginfo.

>

> Jan 01 00:00:00    2     5     0 libcam.so (Jul  9 2010 13:52:59) bver

> 6040207

> Jan 01 00:00:00    5     6     0 Watchdog Timer is disable now.

> Jan 01 00:00:00    2     5   100 cam-disk.so (Jul  9 2010 13:53:13)

> Jan 01 00:00:14    6     8     0 get_config_data(): xres=800 yres=480

> refresh=60 output_fmt=24 di_sel=1 display_type=1 yuv_layer=1

> Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): type=1,

> gpio3=0x0600102a DVI_RST LVDS_RST LCD_DET

> Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): set

> gpio3=0x0600103a DVI_RST LVDS_RST LCD_DET LCD_BKLIGHT

>

> For the display.conf you can refer the attached buildfile.

> for modopts in display.conf am using the imx51-1.conf.

> xres=800,yres=480,refresh=60,pixclk=32264000,hsw=124,hss=60,hew=32,vsw=2

> ,vss=33,vew=10,vpol=1,epol=1,dpol=0,cpol=0,hpol=1,ofmt=16,di=1,type=1,yu

> v=1

>

> In the release notes they have mentioned about display-detect but i dint

> find any binary for the same, so i have directly given the imx51-1.conf

> file.

>

>

> Can anyone help me on this issue.

>

>

> Kishore.S

>

>

>

>

>

>

>

> --

> This message has been scanned for viruses and

> dangerous content by MailScanner, and is

> believed to be clean.

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post87517

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post87560

>

>

> --

> This message has been scanned for viruses and

> dangerous content by MailScanner, and is

> believed to be clean.

>

>

--

This message has been scanned for viruses and

dangerous content by MailScanner, and is

believed to be clean.

post87560: RE: Display problem for iMx51 - vsync and eglgears not working

2011-07-25T12:51:51Z

What kind of display (DVI, VGA, LCD) are you using?

Thanks

-----Original Message-----

From: kishore s [mailto:community-noreply@qnx.com]

Sent: July 22, 2011 2:43 AM

To: advanced-graphics

Subject: Display problem for iMx51 - vsync and eglgears not working

Hi,

I am working on iMx51 reference board from Freescale.

When i run the vsync or egl-gears application nothing is comming up on

the

display.

When i run the egl application i get the below prints.

egl-gears

185 frames in  5.027 seconds = 36.801 FPS

184 frames in  5.027 seconds = 36.602 FPS

218 frames in  5.946 seconds = 36.663 FPS

I am not getting any errors also when i run sloginfo.

Jan 01 00:00:00    2     5     0 libcam.so (Jul  9 2010 13:52:59) bver

6040207

Jan 01 00:00:00    5     6     0 Watchdog Timer is disable now.

Jan 01 00:00:00    2     5   100 cam-disk.so (Jul  9 2010 13:53:13)

Jan 01 00:00:14    6     8     0 get_config_data(): xres=800 yres=480

refresh=60 output_fmt=24 di_sel=1 display_type=1 yuv_layer=1

Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): type=1,

gpio3=0x0600102a DVI_RST LVDS_RST LCD_DET

Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): set

gpio3=0x0600103a DVI_RST LVDS_RST LCD_DET LCD_BKLIGHT

For the display.conf you can refer the attached buildfile.

for modopts in display.conf am using the imx51-1.conf.

xres=800,yres=480,refresh=60,pixclk=32264000,hsw=124,hss=60,hew=32,vsw=2

,vss=33,vew=10,vpol=1,epol=1,dpol=0,cpol=0,hpol=1,ofmt=16,di=1,type=1,yu

v=1

In the release notes they have mentioned about display-detect but i dint

find any binary for the same, so i have directly given the imx51-1.conf

file.

Can anyone help me on this issue.

Kishore.S

--

This message has been scanned for viruses and

dangerous content by MailScanner, and is

believed to be clean.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post87517

post87517: Display problem for iMx51 - vsync and eglgears not working

2011-07-22T06:42:37Z

Hi,

I am working on iMx51 reference board from Freescale.

When i run the vsync or egl-gears application nothing is comming up on the

display.

When i run the egl application i get the below prints.

egl-gears

185 frames in  5.027 seconds = 36.801 FPS

184 frames in  5.027 seconds = 36.602 FPS

218 frames in  5.946 seconds = 36.663 FPS

I am not getting any errors also when i run sloginfo.

Jan 01 00:00:00    2     5     0 libcam.so (Jul  9 2010 13:52:59) bver

6040207

Jan 01 00:00:00    5     6     0 Watchdog Timer is disable now.

Jan 01 00:00:00    2     5   100 cam-disk.so (Jul  9 2010 13:53:13)

Jan 01 00:00:14    6     8     0 get_config_data(): xres=800 yres=480

refresh=60 output_fmt=24 di_sel=1 display_type=1 yuv_layer=1

Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): type=1,

gpio3=0x0600102a DVI_RST LVDS_RST LCD_DET

Jan 01 00:00:14    6     8     0 imx51_layer_update_end(): set

gpio3=0x0600103a DVI_RST LVDS_RST LCD_DET LCD_BKLIGHT

For the display.conf you can refer the attached buildfile.

for modopts in display.conf am using the imx51-1.conf.

xres=800,yres=480,refresh=60,pixclk=32264000,hsw=124,hss=60,hew=32,vsw=2,vss=33,vew=10,vpol=1,epol=1,dpol=0,cpol=0,hpol=1,ofmt=16,di=1,type=1,yuv=1

In the release notes they have mentioned about display-detect but i dint

find any binary for the same, so i have directly given the imx51-1.conf

file.

Can anyone help me on this issue.

Kishore.S

--

This message has been scanned for viruses and

dangerous content by MailScanner, and is

believed to be clean.

post86633: Re: Poulsbo and screen resiolution

2011-06-15T11:59:15Z

Check your video memory setting in BIOS! Had the same problem with the poulsbo driver on a iEi UIBX-200 machine. Changing the video memory from 4MB to 8MB it works fine with 1280x1024 16M colors.

post86430: IMG_W and IMG_H in img_t

2011-06-03T17:33:23Z

I am loading a 640X480 rgb png as follow:

img_t                   image1;

rc = img_load_file(ilib, "/tmp/black.png", NULL, &image1);

I am trying to walk through every pixel in this image and do per pixel calculation.  However, the loop below causes my system to crash!

for (i = 0; i <( image1.h * image1.w);  i++)

// walk through the image using a pointer

After debugging this, it seems if I use image size to be 640*361 everything works fine.  So, why is it that imag1.access.direct.data is pointing to a memory size that is (640*361) where the image size is 640X480?  My image is a ARGB, so it I should have 4 bytes/pixel.

According to the documentation in img_t, the img_w and ima_h fields are the number of pixels.

Can someone help me understand the discrepancy that I am seeing?

Thanks,

R.S.

post86372: img_t flags

2011-06-02T03:40:24Z

Hi,

I have a question in regards to the following img_t flags:

IMG_TRANSPARENCY;

IMG_FMT_ALPHA;

I have a file that is ARGB8888. When I load this file as below

rc = img_load_file(ilib, "/tmp/Day_24hr.png", NULL, &image1);

the IMG_TRANSPARENCY, IMG_FMT_ALPHA are not set! Aren't these flags suppose to be set by the img_load_file, if the file is ARGB8888 png file? I am trying to blit (blend) two ARGB files, but for some reason, I keep overwriting the destination instead of the blending. I also tried the following without luck

alpha.mode = GF_ALPHA_M1_SRC_PIXEL_ALPHA;

gf_context_set_dst_alpha(context, &alpha);

I would appreciate if someone can show me me some code on how to blend to two ARGB png files together.

I am trying to evaluate the QNX for our next company project but the documentation is not always very easy to follow (at least for a non graphic expert like me)

Thanks,

R.S.

post86359: Re: How to blit insttead of overwite using gf_draw_blit2?

2011-06-01T21:36:34Z

For some reason, when I do:

image2.flags |= IMG_FORMAT;

image2.format = GF_FORMAT_ARGB8888;

rc = img_load_file(ilib, "/tmp/Day_24hr.png", NULL, &image1);

rc = gf_surface_attach(&img_surf1, gdev, image1.w, image1.h, image1.access.direct.stride, img_fmt_to_gf(image1.format),

&palette, image1.access.direct.data, 0);

gf_context_create(&context);

Then my call to rc = gf_context_set_surface(context, img_surf1); hangs and never returns!!!

I am running out of ideas to try!!!

Thanks,

Reza.

post86358: Re: How to blit insttead of overwite using gf_draw_blit2?

2011-06-01T21:27:37Z

Correct, you'll need an alpha channel for source alpha blending.

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Wednesday, June 01, 2011 05:17 PM

To: advanced-graphics <post86357@community.qnx.com>

Subject: Re: How to blit insttead of overwite using gf_draw_blit2?

Thanks for the quick reply Joel. However, I guess my problem is that I can not figure out how to enable blending! I have looked at the document and I found the gf_context_set_alpha. However, my images are RGB888 and not ARGB888. Should I read the original image as ARGB8888 (by setting the img_t format field).

I would appreciate it if you can point me to small code snipet or site, as I have been trying to figure this out for past two days.

Regards,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post86357

post86357: Re: How to blit insttead of overwite using gf_draw_blit2?

2011-06-01T21:17:32Z

Thanks for the quick reply Joel. However, I guess my problem is that I can not figure out how to enable blending! I have looked at the document and I found the gf_context_set_alpha. However, my images are RGB888 and not ARGB888. Should I read the original image as ARGB8888 (by setting the img_t format field).

I would appreciate it if you can point me to small code snipet or site, as I have been trying to figure this out for past two days.

Regards,

Reza.

post86356: Re: How to blit insttead of overwite using gf_draw_blit2?

2011-06-01T20:56:56Z

You'll probably want to enable blending and use a source over blending mode.

-Joel

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Wednesday, June 01, 2011 04:51 PM

To: advanced-graphics <post86355@community.qnx.com>

Subject: How to blit insttead of overwite using gf_draw_blit2?

Hi All,

I am trying to blit two .png files.  I have attached the .png files to two seperate surfaces and am using the following code(only the important calls):

rc = gf_surface_attach(&img_surf1, gdev, image1.w, image1.h, image1.access.direct.stride, img_fmt_to_gf(image1.format),

        &palette, image1.access.direct.data, 0);

rc = gf_surface_attach(&img_surf2, gdev, image2.w, image2.h,   image2.access.direct.stride, img_fmt_to_gf(image2.format),

        &palette2, image2.access.direct.data, 0);

gf_context_create(&context);

gf_context_set_surface(context, img_surf1);

gf_draw_begin(context);

gf_draw_blit2( context,  img_surf2, NULL, 107, 4, 334, 222, 0, 0);

gf_draw_end(context);

rc = img_write_file(ilib, "/tmp/reza.png", NULL, &image1);

The two images are getting combined.  However, the target image is overwriting (covering) the target image.

Can someone please tell me what I need to do (probably to the context) to have the source image blited into the target image, instead of overwriting it?

Thanks in advance,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post86355

post86355: How to blit insttead of overwite using gf_draw_blit2?

2011-06-01T20:51:18Z

Hi All,

I am trying to blit two .png files.  I have attached the .png files to two seperate surfaces and am using the following code(only the important calls):

rc = gf_surface_attach(&img_surf1, gdev, image1.w, image1.h, image1.access.direct.stride, img_fmt_to_gf(image1.format),

        &palette, image1.access.direct.data, 0);

rc = gf_surface_attach(&img_surf2, gdev, image2.w, image2.h,   image2.access.direct.stride, img_fmt_to_gf(image2.format),

        &palette2, image2.access.direct.data, 0);

gf_context_create(&context);

gf_context_set_surface(context, img_surf1);

gf_draw_begin(context);

gf_draw_blit2( context,  img_surf2, NULL, 107, 4, 334, 222, 0, 0);

gf_draw_end(context);

rc = img_write_file(ilib, "/tmp/reza.png", NULL, &image1);

The two images are getting combined.  However, the target image is overwriting (covering) the target image.

Can someone please tell me what I need to do (probably to the context) to have the source image blited into the target image, instead of overwriting it?

Thanks in advance,

Reza.

post86287: Re: Write data into draw context memory slowly!

2011-05-31T12:51:49Z

1. What graphics driver are you using?

2. What version of the OS?

3. What is the output of 'pci -v' (relevant part only -- search for VGA)

4. What is the output of show_vesa ? (just the header -- not need for modes' descriptions).

post86285: Re: Write data into draw context memory slowly!

2011-05-31T10:59:57Z

This is the link which contains datasheet of ARK-6622H-18ZE:

-

http://www.alldatasheet.net/datasheet-pdf/pdf/327854/ADVANTECH/ARK-6622H-18ZE.html

and this is the link for the board:

-

http://buy.advantech.com/AIMB270G200A1E/AIMB270G200A1E/model-AIMB-270G2-00A1E.htm

2011/5/31 Misha Nefedov <community-noreply@qnx.com>

> The code is for a chassis. Please post a link to the board/graphics adapter

> details.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post86276

>

>

post86276: Re: Write data into draw context memory slowly!

2011-05-30T18:38:46Z

The code is for a chassis. Please post a link to the board/graphics adapter details.

post86275: Re: Write data into draw context memory slowly!

2011-05-30T17:38:35Z

I bought a computer from Advantech, its name is ARK-6622H-18ZE. You can see

details about graphic adapter on website of Advantech. Thank you.

2011/5/31 Thắng Bùi Quý <bqthangcndt@gmail.com>

> Dear you!

> I am using the graphics adapter Intel GFX. I used window OS to build

> programs with the context memory and they always run normal, there were not

> any problem. Do you have solutions for my problem? Can you help me. Please!

>

>

> 2011/5/31 Misha Nefedov <community-noreply@qnx.com>

>

>> This is most likely expected -- when you get a pointer to a video

>> (dedicated) memory. The memory access is facilitated by the graphics

>> controller and is usually slow for reading (comparing to a RAM memory

>> controller) and fast for writing.

>> What graphics adapter are you using?

>>

>>

>>

>> _______________________________________________

>>

>> Advanced Graphics

>> http://community.qnx.com/sf/go/post86272

>>

>>

>

post86274: Re: Write data into draw context memory slowly!

2011-05-30T17:32:36Z

Dear you!

I am using the graphics adapter Intel GFX. I used window OS to build

programs with the context memory and they always run normal, there were not

any problem. Do you have solutions for my problem? Can you help me. Please!

2011/5/31 Misha Nefedov <community-noreply@qnx.com>

> This is most likely expected -- when you get a pointer to a video

> (dedicated) memory. The memory access is facilitated by the graphics

> controller and is usually slow for reading (comparing to a RAM memory

> controller) and fast for writing.

> What graphics adapter are you using?

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post86272

>

>

post86272: Re: Write data into draw context memory slowly!

2011-05-30T17:10:33Z

This is most likely expected -- when you get a pointer to a video (dedicated) memory. The memory access is facilitated by the graphics controller and is usually slow for reading (comparing to a RAM memory controller) and fast for writing.

What graphics adapter are you using?

post86234: Write data into draw context memory slowly!

2011-05-28T17:55:44Z

I write a program using draw context memory by the function PdCreateOffscreenContext(). This function returns a pointer to a PdOffscreenContext_t. After that i use the function PdGetOffscreenContextPtr() to get a pointer to the shared memory object. When i have that pointer i can change data in the context to blit them on the screen by the function PgContextBlit(). However i had a problem with writing data back into the context, it was very slow. For example: ptr is the pointer to context, d is a variable.

for(int i=0; i < CONTEXT_HEIGHT; ++i)

    for(int j=0; j < CONTEXT_WITH; ++j) {

       d = *(ptr + i * CONTEXT_WIDTH * 4 + j);(1)

      *(ptr + i * CONTEXT_WIDTH * 4 + j) = 2 * d; (2)

}

if i set CONTEXT_HEIGHT and CONTEXT_WIDTH small (about smaller than 256) the program run normal.

if i set CONTEXT_HEIGHT and CONTEXT_WIDTH biggerl (about 1024) the program run very slow.

I tried to test with bypassing either the line (1) or the line (2) and i realized that the line (2) made the program very slow. Can you help me to solve it?

post86181: Re: Who is responsible for alloc mem for access.direct.data?

2011-05-26T22:02:30Z

I'd have to take a closer look. Perhaps someone else knows off hand.

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 26, 2011 05:59 PM

To: advanced-graphics <post86180@community.qnx.com>

Subject: Re: Who is responsible for alloc mem for access.direct.data?

Hi Joel,

Thanks for answering. I though the same. However, I wrote a small code and it seems like data is NULL before and after img_load_file() call! The function returns OK so the image does get loaded (somewhere), but the point is still data pointer is still NULL. So, when I try to use this pointer to attach it to a gf surface, the system hangs.

img.flags = 0;

img.flags |= IMG_DIRECT;

printf("data before load %p \n",img.access.direct.data);

if ((rc = img_load_file(ilib, "/tmp/n_day_r.png", NULL, &img)) != IMG_ERR_OK) {

fprintf(stderr, "img_load_file(%s) failed: %d\n", "/tmp/n_day_r.png", rc);

return -1;

}

else

{

fprintf(stderr, "img_load_file(%s) sucess : %d\n", "/tmp/n_day_r.png", rc);

fprintf(stdout, "img is %dx%dx%d\n", img.w, img.h, IMG_FMT_BPP(img.format));

}

printf("data after load %d \n",img.access.direct.data);

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post86180

post86180: Re: Who is responsible for alloc mem for access.direct.data?

2011-05-26T21:59:31Z

Hi Joel,

Thanks for answering. I though the same. However, I wrote a small code and it seems like data is NULL before and after img_load_file() call! The function returns OK so the image does get loaded (somewhere), but the point is still data pointer is still NULL. So, when I try to use this pointer to attach it to a gf surface, the system hangs.

img.flags = 0;

img.flags |= IMG_DIRECT;

printf("data before load %p \n",img.access.direct.data);

if ((rc = img_load_file(ilib, "/tmp/n_day_r.png", NULL, &img)) != IMG_ERR_OK) {

fprintf(stderr, "img_load_file(%s) failed: %d\n", "/tmp/n_day_r.png", rc);

return -1;

}

else

{

fprintf(stderr, "img_load_file(%s) sucess : %d\n", "/tmp/n_day_r.png", rc);

fprintf(stdout, "img is %dx%dx%d\n", img.w, img.h, IMG_FMT_BPP(img.format));

}

printf("data after load %d \n",img.access.direct.data);

post86178: Re: Who is responsible for alloc mem for access.direct.data?

2011-05-26T21:23:40Z

It been a long time, but i'm pretty sure the lib will allocate the memory, however I think you're responsible for freeing it.

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 26, 2011 05:21 PM

To: advanced-graphics <post86177@community.qnx.com>

Subject: Who is responsible for alloc mem for access.direct.data?

Hi All,

I am having a problem with setting img.flags |= IMG_DIRECT. Who is responsible for alloc memory for acess.direct.data? Is it the user or does the img_load_file() allocates memeory when the user sets the IMG_DIRECT flag. I have been spinning why wheels on this. I am trying to grab the acess.direct.data pointer after loading a png file, but my system hangs everytime I set the IMG_DIRECT flag.

Thanks,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post86177

post86177: Who is responsible for alloc mem for access.direct.data?

2011-05-26T21:21:19Z

Hi All,

I am having a problem with setting img.flags |= IMG_DIRECT. Who is responsible for alloc memory for acess.direct.data? Is it the user or does the img_load_file() allocates memeory when the user sets the IMG_DIRECT flag. I have been spinning why wheels on this. I am trying to grab the acess.direct.data pointer after loading a png file, but my system hangs everytime I set the IMG_DIRECT flag.

Thanks,

Reza.

post86166: How to update driver from 6.3 to 6.4/6.5?

2011-05-26T14:50:05Z

Hello,

For one of our embedded system we have a custom video driver working with QNX 6.3. Now I want to update it for QNX 6.4/6.5.

I'm currently reading this doc :

http://www.qnx.com/download/feature.html?programid=20957

Now I need to get the updated sources of the includes and sample drivers. But since the last changes in F27 I have not access to these sources.

How is it possible to get these access?

Thanks,

Christophe

post86102: Re: RE: cleaning up an img_t* in C++

2011-05-25T00:13:23Z

Hi Derek,

I have a question with regard to img_t but in regard to who is allocating memory for access.direct.data pointer when using setting the IMG_DIRECT flag? Will the img_load_file() allocate memory for it?

I am trying to read a .png file, make some modification to the file and then save it as new .png file. However, when I set the IMG_DIRECT flag, my system hangs. If I don't set the IMG_DIRECT flag, the file is loaded correctly.

I am using QNX 6.4.1 on the host and the same version running on a VM target.

I appreciate it if you can give me pointer on where the problem could be.

Thanks,

RS.

post86101: Re: RE: How to use devg-flat in gf_dev_attach?

2011-05-24T23:06:36Z

Hi Joel,

I tried to use the direct pointer approach, but my system hangs. Below is a code snipet where it hangs:

img_t img_day;

if ((rc = img_lib_attach(&ilib)) != IMG_ERR_OK) {

fprintf(stderr, "img_lib_attach() failed: %d\n", rc);

return rc;

}

img_day.flags = 0;

img_day.flags |= IMG_DIRECT;

// /tmp/stop.png is the location of the file on the target VM system

img_template_file_Day = "/tmp/stop.png";

// The code hangs after making the call below. If I don't set the IMG_DIRECT flag, it works.

if ((rc = img_load_file(ilib, img_template_file_Day, NULL, &img_day)) != IMG_ERR_OK) {

fprintf(stderr, "img_load_file(%s) failed on Day Template: %d\n", img_template_file_Day, rc);

// return rc;

}

post86097: Re: RE: How to use devg-flat in gf_dev_attach?

2011-05-24T20:14:43Z

Well, I think my problem is that I am not familiar enough with the graphic sub-systems of QNX to know what I want to do :). Basically, I have some graphic artists that are providing me base .png files. I need to draw an ellipse based on some dynamic data and save the result as another .png file. So, whatever is the fastest to accomplish this would work for me.

I appreciate it if you have any suggestion or code snippet to help me get started.

Thanks,

Reza

post86096: RE: How to use devg-flat in gf_dev_attach?

2011-05-24T19:38:35Z

Do you want to modify the img data using gf rendering or just via a pointer in software? If you don't plan on using gf rendering, then there's no need for a gf surface.

-Joel

-----Original Message-----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: May 24, 2011 3:02 PM

To: advanced-graphics

Subject: Re: How to use devg-flat in gf_dev_attach?

Thanks Joel.  But, my understanding was that in order to modify the image data, I will have to first attach it to a gf surface, modify it, and then save it.  Just so I understand, if I only need need to modify a .png file (not display it), I don't need to use gf at all?  In other words, I don't need to attach and great a gf surface at all?  Basically, below is what I am trying to do:

1.  Read a .png file from the file system (i.e. /tmp/test.png)

2.  Modify this file with some algorithm to draw an ellipse

3.  Same the result from #2 as /tmp/new_image.png

Thanks,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post86093

post86093: Re: How to use devg-flat in gf_dev_attach?

2011-05-24T19:02:28Z

Thanks Joel.  But, my understanding was that in order to modify the image data, I will have to first attach it to a gf surface, modify it, and then save it.  Just so I understand, if I only need need to modify a .png file (not display it), I don't need to use gf at all?  In other words, I don't need to attach and great a gf surface at all?  Basically, below is what I am trying to do:

1.  Read a .png file from the file system (i.e. /tmp/test.png)

2.  Modify this file with some algorithm to draw an ellipse

3.  Same the result from #2 as /tmp/new_image.png

Thanks,

Reza.

post86090: Re: How to use devg-flat in gf_dev_attach?

2011-05-24T18:30:37Z

If you just want to encode/decode image data you don't need GF. You can either pass or get a pointer to the img data by just using libimg.

post86089: How to use devg-flat in gf_dev_attach?

2011-05-24T18:27:12Z

I am trying to read a .png file, modify it and write a back to the file system. In order to do that, my understanding is, that I need to use both gf and img libraries. I don't want to use any specific HW to attach to as I don't need to display my .png file. All I want to do is to be able to create surface so that I can modify my .png file.

In order to create a surface, I need to first attach to a device. My understanding is that I can use devg-flat as a device. However I am not sure, how to install devg-flat driver and related files into my IDE and target. I also don't know how to use devg-flat in the call gf_dev_attach() call? I have already added libffb.so to extra libraries section of IDE.

I appreciate it if anyone can give me some pointers on this.

Regards,

Reza.

post86031: Re: Newbie Question - Help with GF

2011-05-21T11:43:26Z

Reza Salehi wrote:

> Hi Armin,

>

> Thanks for your comment.  Do you know if there is a site that tells you how to install sdl_gfx library/ies and write a hello world program to draw an ellipse on VM x86 target running Nuternia 6.5?

SDL  1.3:

http://community.qnx.com/sf/projects/qnx_community_sdl_project/   -->

have a look to the OpenGL ES support

SDL_Gfx: http://www.ferzkopp.net/joomla/content/view/19/14/  ... it

compiles out of the box.

--Armin

> Thanks in advance,

>

> Reza.

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post86028

>

>

post86028: Re: Newbie Question - Help with GF

2011-05-20T23:13:14Z

Hi Armin,

Thanks for your comment. Do you know if there is a site that tells you how to install sdl_gfx library/ies and write a hello world program to draw an ellipse on VM x86 target running Nuternia 6.5?

Thanks in advance,

Reza.

post86000: Re: Newbie Question - Help with GF

2011-05-20T08:53:18Z

Joel Pilon wrote:

> My mistake, my memory is a little rusty. You'll have to generate and tessellate your own ellipses in gf. Or you might be able to use photon and the pgdrawellise function.

... with SDL_Gfx you will get antialized ellipses :)

--Armin

http://www.steinhoff-automation.com

> ----- Original Message -----

> From: Reza Salehi [mailto:community-noreply@qnx.com]

> Sent: Thursday, May 19, 2011 08:42 PM

> To: advanced-graphics<post85987@community.qnx.com>

> Subject: Re: Newbie Question - Help with GF

>

> Thanks Joel. However, after checking the libgf, it seems like it doesn't support drawing of ellipses! I only see polyline and rectangular drawing functions.

>

> Cheers,

>

> Reza.

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post85987

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post85995

>

>

post85995: Re: Newbie Question - Help with GF

2011-05-20T02:52:36Z

My mistake, my memory is a little rusty. You'll have to generate and tessellate your own ellipses in gf. Or you might be able to use photon and the pgdrawellise function.

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 19, 2011 08:42 PM

To: advanced-graphics <post85987@community.qnx.com>

Subject: Re: Newbie Question - Help with GF

Thanks Joel. However, after checking the libgf, it seems like it doesn't support drawing of ellipses! I only see polyline and rectangular drawing functions.

Cheers,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post85987

post85987: Re: Newbie Question - Help with GF

2011-05-20T00:42:55Z

Thanks Joel. However, after checking the libgf, it seems like it doesn't support drawing of ellipses! I only see polyline and rectangular drawing functions.

Cheers,

Reza.

post85984: Re: Newbie Question - Help with GF

2011-05-19T23:49:01Z

Oops accidently hit send to soon. You can get a pointer from a gf surface using gf to either read or write image data. Libimg accepts a pointer for the src or dst for image encode/decode operations.

----- Original Message -----

From: Joel Pilon

Sent: Thursday, May 19, 2011 07:46 PM

To: 'post85982@community.qnx.com' <post85982@community.qnx.com>

Subject: Re: Newbie Question - Help with GF

I don't have a link handy, but you can get a pointer from a CPU accessible surface. You can pass the pointrr

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 19, 2011 07:44 PM

To: advanced-graphics <post85982@community.qnx.com>

Subject: Re: Newbie Question - Help with GF

Thank you very much Joel. At least I know where to start looking.

After looking into libimg, I found the img_load and img_load_file APIs. However, one of these APIs deal with streaming data and another with a filesystem file! I could not find and API that would directly read and load an image from GF surface. Can you please tell which libimg API does that?

Also, QNX usually has some very good code examples or tutorials on these types of things. Do you know a link or a location where I can find an example of how to do something like this instead of me asking stupid questions and wasting peoples time on the this Forum?

Thanks again,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post85982

post85983: Re: Newbie Question - Help with GF

2011-05-19T23:46:10Z

I don't have a link handy, but you can get a pointer from a CPU accessible surface. You can pass the pointrr

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 19, 2011 07:44 PM

To: advanced-graphics <post85982@community.qnx.com>

Subject: Re: Newbie Question - Help with GF

Thank you very much Joel. At least I know where to start looking.

After looking into libimg, I found the img_load and img_load_file APIs. However, one of these APIs deal with streaming data and another with a filesystem file! I could not find and API that would directly read and load an image from GF surface. Can you please tell which libimg API does that?

Also, QNX usually has some very good code examples or tutorials on these types of things. Do you know a link or a location where I can find an example of how to do something like this instead of me asking stupid questions and wasting peoples time on the this Forum?

Thanks again,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post85982

post85982: Re: Newbie Question - Help with GF

2011-05-19T23:44:14Z

Thank you very much Joel. At least I know where to start looking.

After looking into libimg, I found the img_load and img_load_file APIs. However, one of these APIs deal with streaming data and another with a filesystem file! I could not find and API that would directly read and load an image from GF surface. Can you please tell which libimg API does that?

Also, QNX usually has some very good code examples or tutorials on these types of things. Do you know a link or a location where I can find an example of how to do something like this instead of me asking stupid questions and wasting peoples time on the this Forum?

Thanks again,

Reza.

post85979: Re: Newbie Question - Help with GF

2011-05-19T22:38:51Z

I'd suggest taking a look at libimg. There also doc for the api as well. The gf api allows you to render things like ellipses and libimg would allow you to write data from gf surfaces as pngs. You could then load png files into gf surfaces or to render them as gf images.

----- Original Message -----

From: Reza Salehi [mailto:community-noreply@qnx.com]

Sent: Thursday, May 19, 2011 06:30 PM

To: advanced-graphics <post85978@community.qnx.com>

Subject: Re: Newbie Question - Help with GF

Sorry, I had a typo in the previous e-mail: I was trying to say that I am trying to draw ellipses and save them as a png file.

Thanks,

Reza.

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post85978

post85978: Re: Newbie Question - Help with GF

2011-05-19T22:30:37Z

Sorry, I had a typo in the previous e-mail: I was trying to say that I am trying to draw ellipses and save them as a png file.

Thanks,

Reza.

post85977: Newbie Question - Help with GF

2011-05-19T21:35:09Z

Hi All,

I am Newbie with QNX GF and QNX in general.  I am trying to use Momentics IDE to connect to a x86 VM target using qconn.  What I want to do is to be able to draw an based on some parameters and then save it in a PNG format on the target file system (VM running neutrino).  I would like to then be able to load the png file in a browser and display the image.

Can anyone someone tell me what is the best way to start?  Is GF the correct place to start.  With GF, I could not find anyway to render to a file rather that a graphic device!

I apologize if some my graphic terminology is not correct.  I am trying to evaluate the QNX platform for a future project and this would help me get started.

Thanks in advance.

Reza.

post85798: Re: What are the basics of font rendering? [bcc] [bcc][auto-ip]

2011-05-17T16:15:31Z

Thanks, Misha. It does seem to do something now.

On 05/17/2011 09:19 AM, Misha Nefedov wrote:

> Yes, I got this. The problem is that we asked to NO rendering buffer when doing th PfAttachCx(). Give it a number, for example:

>

> PfAttachCx( NULL, 32*1024 );

>

> The size of the buffer should depend on how large the strings and the font size are.

>

>

>

>

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post85785

post85785: Re: What are the basics of font rendering? [bcc] [bcc][auto-ip]

2011-05-17T14:19:24Z

Yes, I got this. The problem is that we asked to NO rendering buffer when doing th PfAttachCx(). Give it a number, for example:

PfAttachCx( NULL, 32*1024 );

The size of the buffer should depend on how large the strings and the font size are.

post85745: Re: What are the basics of font rendering? [bcc] [bcc][auto-ip]

2011-05-16T20:15:32Z

On 05/16/2011 02:55 PM, Misha Nefedov wrote:

> Hard to tell. Can you please post the entire source file -- for me to try?

Here's code that fails for me as described.  I compiled it with

"ntox86-gcc -o FontTest FontTest.c -lph" and ran the font server with

"phfont -d /usr/photon/font_repository -D /usr/photon/font_repository".

#include <stdio.h>

#include <errno.h>

#include <photon/Pf.h>

void

Callback (void *ctx, const pf_point_t *pos, const FontRender *fr)

{

   fprintf (stderr, "Callback: %d,%d\n", pos->x, pos->y);

   fprintf (stderr, "Callback: %d,%d %d,%d %d, %d, %d, %p\n",

       fr->size.x, fr->size.y, fr->offset.x, fr->offset.y,

       fr->width, fr->bpl, fr->bpp, fr->bmptr);

}

int

main (void)

{

   char *MyFontName = "Courier10 BT";

   int FontSize = 10;

   int i;

   struct _Pf_ctrl *FontServer = NULL;

   FontID *ID;

   FontName Description;

   pf_point_t pp = { 50, 50 };

   FontServer = PfAttachCx (NULL, 0);

   if (FontServer == NULL)

     {

       fprintf (stderr, "Cannot attach to font server.\n");

       return (1);

     }

   if (NULL == PfGenerateFontNameCx (FontServer, MyFontName, 0,

                                     FontSize, Description))

     {

       fprintf (stderr, "Cannot get font name %s size %d style 0x%X\n",

           MyFontName, FontSize, 0);

       return (1);

     }

   i = PfRenderCx (FontServer, FontServer, Description, 0L, 0L,

                   "Hello, world!", 0, 0, &pp, NULL, Callback);

   fprintf (stderr, "PfRenderCx() returned %d\n", i);

   if (i)

     fprintf (stderr, "%s\n", strerror (errno));

   return (0);

}

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post85742

post85743: RE: RE: gf_context_set_surface(): Problem with color format GF_FORMAT_PACK_YUV_UYVY

2011-05-16T19:55:11Z

GF does not support blitting to/from YUV surfaces so you will not be

able to set a context to this surface. The other formats are all RGB

so they work.

If you need to transfer YUV data between surfaces you'll have to

manually copy this. Alternately depending on your

application/use-case you could have multiple YUV surfaces and just

re-target the layer to point at the different surfaces.

(gf_layer_set_surfaces())

-----Original Message-----

From: Nikshep Patil [mailto:community-noreply@qnx.com]

Sent: May 16, 2011 3:35 PM

To: advanced-graphics

Subject: Re: RE: gf_context_set_surface(): Problem with color format

GF_FORMAT_PACK_YUV_UYVY

Hi Michael,

I am working with an internal driver.

Regards

_______________________________________________

Advanced Graphics

http://community.qnx.com/sf/go/post85736

post85742: Re: What are the basics of font rendering? [bcc] [bcc][auto-ip]

2011-05-16T19:54:59Z

Hard to tell. Can you please post the entire source file -- for me to try?

post85739: Re: gf_context_set_surface(): Problem with color format GF_FORMAT_PACK_YUV_UYVY

2011-05-16T19:39:34Z

I am having problems with gf_context_set_surface() of step 3 with the color format GF_FORMAT_PACK_YUV_UYVY.

post85737: Re: What are the basics of font rendering? [bcc] [bcc][auto-ip]

2011-05-16T19:35:18Z

On 05/16/2011 01:17 PM, Misha Nefedov wrote:

> The 'Description' should be replaced with 'FontName' when calling the PfFindFont(). The PfFindFont() is similar to PfGenerateFontName(), it only returns FontID. I am not sure why you need it.

Well, I was simply following the sample code in the PfRenderCx() API

doc, as you suggested.  The sample code calls PfFindFontCx().

Personally, I'm willing to do without it, even though I don't see why it

should be failing.

> The 'Description' you got from the Generate function is good to be used with the subsequent PfRenderCx().

Okay, I think you're suggesting something like this:

   FontName Description;

   char *FontName = "Courier10 BT";

   pf_point_t pp = { 50, 50 };

   int FontSize = 10;

   FontServer = PfAttachCx (NULL, 0);

   ... check for error ...

   if (NULL == PfGenerateFontNameCx (FontServer, FontName, 0, FontSize,

Description))

     {

       ... error ...

     }

   i = PfRenderCx (FontServer, FontServer, Description, 0L, 0L, "Hello,

world!", 0,

        0, &pp, NULL, Callback);

   ... check for error ...

This code fails when it reaches PfRenderCx().  The error message

displayed by "perror(strerror(errno))" is "Invalid argument".

>

>

> _______________________________________________

>

> Advanced Graphics

> http://community.qnx.com/sf/go/post85727