Thanks for the patch, can u please rebase your patch to the latest on trunk
before we review it ? There has been core gstreamer component updates since your
checkout. We need to know that this update will not cause any problem on
I update my work copy with the 563 review. I attached the new patch with the
You may want to give yourself credit as the author of the gstticodecs_omapl138.c
Also, did you test any container formats such as AVI?
Patch looks good but the main question is: where do I get DVSDK for L138 ?
gst_ti adds support for platform only after its DVSDK is released to the
It's very nice that we have this patch and it can be rebased when we get the
official DVSDK support.
A DVSDK for L138 is certainly a requirement for checking-in this patch to trunk.
I'm working this issue to determine when it will be generally available.
Thanks and regards,
There is a new update in the patch related to the gsttidmaivideosink.c file,
also the path of the files are against the trunk path and not against the work
The name of the new patch is gstreamer-ti-omapl138.patch
DMAI support for L138 is now on trunk. Commit 423. I verify the latest patch
from Jorge against this version.
I've found some small issues with your patch and will post my finding soon, but
just wanted to know if you guys have tested v4l2src on L138 device ? If not,
then is there any plans to test and support it ?
Rebasing patch to svn rev 612.
I test the gstreamer-ti-omapl138-trunk-r612.patch and is working correctly. I
also rebased the patch to rev 650 and upload the file.
I also upload the patch that make some more modifications in TIAuddec1/TIViddec2
to run freondemo, and for rev650
Thanks for the patches, please find my initial review comments below:
* Instead of doing #if0, i would suggest adding a property in TIViddec2 and
TIAuddec1 to use Real-time thread vs normal thread. May be we can name this
property as "RTCodecThread" and its default should be TRUE and
for freondemo you can set it to FALSE. Any comments ?
* codec configuration and other .cfg files looks okay to me.
* As per dmaivideosink changes are concerned, do you know if dmaivideosink has
any additional value over fbdevsink ? Since both sinks are doing memcpy in
RGB565 plane hence i was thinking that we should recommend (and document) to use
fbdevsink on OMAPL138 instead of dmaivideosink. Any thought ?
* About adding the property "RTCodecThread", its a better and
nice idea to implement to choose between Real-time thread vs normal thread.
* About dmaivideosink, this changes were made to add a feature for the omapL138,
but at the end this element was never used at all for freondemo, so you can
consider that patch if you want or simply ignore it.