Shed some light on transparency?

2008-09-05 Thread Paul Stuart
Hi, Wondering if anyone has any insight into OSD transparency on the dm355. Specifically, I'm unclear on the role of OSD_OSDWINnMD_TEn and OSD_OSDWINnMD_BLEND_FACTOR with regards to the behavior of the attribute layer. If I put OSD1 in attribute layer mode and want it to control the blending

RE: How much memory needed in the 5MP jpeg encode on DM355?

2008-09-05 Thread Jon Povey
Re trying to get hold of large chunks of memory, have a look at alloc_bootmem_pages() and friends in linux/bootmem.h, or page 230 of O'Reilly "Linux Device Drivers" 3rd edition. I haven't used these but they sound like they might help. Apparently you have to compile into the kernel but you could w

Re: How much memory needed in the 5MP jpeg encode on DM355?

2008-09-05 Thread Stephen Berry
>From what I remember, the V4l2 driver needs 2 buffers (10MB), the JPEG encoder needs 19MB (at the very least, sometimes it asks for another big buffer but since the first call failed you never get the second request) and the IPIPE needs at least one source buffer (which can be the V4l2 buffer) and

gstreamer v4l2 and fbvideosink

2008-09-05 Thread ketan deshpande
Hello everybody, I am trying to run the gstreamer pipeline from v4l2 to fb using the following command gst-launch-0.10 v4l2src ! fbvideosink device=/dev/fb/3 but I keep getting the following error WARNING: from element /pipeline0/v4l2src0: Got unexpected frame size of 884736 instead o

How much memory needed in the 5MP jpeg encode on DM355?

2008-09-05 Thread pavelhan
My EVM board has 128M bytes DDR2 ram,and Now i am involved in a 5MP jpeg encode design on DM355,the sensor is MICRON MT9P031. I want to know the ccdc, ipie and jpegenc part on DM355 need how much memory to get the work more smoothly? My program hang at the ipipe initialization function,beacause I

Re: DaVinci ASoC DMA stalls

2008-09-05 Thread Caglar Akyuz
Mark Lokowich wrote: > I did have to add a getposition accessor to the DaVinci DMA module, > which is used by the pcm_pointer function. Other than that, I believe > the ASoC should build. My test is simply to run aplay or speaker-test, > then quit and run it again. The first run is audible, the