V4L2, encoding seems not work.

[jasonhsu]

Hi Community,

We have a application of video conference system, that it will get camera stream and encode to H263 format and send it to another receiver through internet. After runing the system over 12 hours or more for aging test, the encoding seems not work, and my board always prints below log:

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

I/FslCameraHAL( 3210): no buffer in v4l2, continue

............

Could you give me some advices about what happen and how to debug next step?

[igorpadykov]

Hi jianhua

this seems as memory leak bug, please try latest BSP, it has memory management fixes

L3.10.17_1.0.2_iMX6QDLS_BUNDLE : i.MX 6Quad, i.MX 6Dual, i.MX 6DualLite,

i.MX 6Solo Linux Binary Demo Files and Linux BSP Documentation.

Size (K): 717480 Format: gz Rev #: L3.10.17_1.0.2 Modified: 10/29/2014

i.MX6Q|i.MX 6Quad Processors|Quad Core|Freescale

Best regards

igor

[jasonhsu]

Hi igor,

Thx for your reply.

We development the project since last year with using IMX6_JB43_110_ANDROID_SOURCE_BSP. and many features are added to it. it is difficult to use new BSP in short time.

Is there any other way to porting the memory management fixes to the version we are using, ex. any patch?

[igorpadykov · Accepted solution]

Hi jianhua

refer to attached file

i.MX6Q|i.MX 6Quad Processors|Quad Core|Freescale

~igor

[jasonhsu]

Hi igor,

I have tried to apply the patch, many files applied failed.

Last year, we based on this version(IMX6_JB43_110_ANDROID_SOURCE_BSP) to init project, see below picture.

and I download new patch according to your suggest(IMX6_JB43_111_ANDROID_SOURCE_BSP), see below picture:

If apply IMX6_JB43_111_ANDROID_SOURCE_BSP to (IMX6_JB43_110_ANDROID_SOURCE_BSP), it will generate many patch failure in kernel_imx. and cannot fix manually, because some data structs do not exist at all.

About memory leak issues, is there patch based on  (IMX6_JB43_110_ANDROID_SOURCE_BSP) ?

[igorpadykov]

Hi jianhua

yes latest patches are very important and should be applied

to all old BSPs.

For procedure of applying patches I would suggest

to create new Community thread.

Best regards

igor

[jasonhsu]

Hi Igor

thx for your reply.

I will create a new thread. thx for your help.:smileylaugh:

[jasonhsu]

Hi igor,

thx for your reply.

I have tried to apply the patchs. it seems some mistakes happen.

the two patchs in device/fsl-proprietary: 0002xxx.patch and 0003.patch.

patch file is .so type, when I apply the patch, it will show below log:

$ patch -p1 < 0002-ENGR00327608-separate-GPU-low-memory-killer-and-rese.patch

   patch: **** Only garbage was found in the patch input.

I think it is failed to apply, do you have those file after patch applied:

gpu-viv/lib/egl/libEGL_VIVANTE.so       | Bin 50868 -> 50868 bytes

gpu-viv/lib/egl/libGLESv1_CM_VIVANTE.so | Bin 246320 -> 246312 bytes

gpu-viv/lib/egl/libGLESv2_VIVANTE.so    | Bin 363344 -> 363344 bytes

gpu-viv/lib/hw/gralloc_viv.imx6.so      | Bin 18040 -> 18040 bytes

gpu-viv/lib/hw/hwcomposer_viv.imx6.so   | Bin 30012 -> 30012 bytes

gpu-viv/lib/libGAL.so                   | Bin 630116 -> 638396 bytes

gpu-viv/lib/libGLSLC.so                 | Bin 349036 -> 349036 bytes

gpu-viv/lib/libg2d.so                   | Bin 17536 -> 17536 bytes

If you have, maybe you can share with me directly.:smileylaugh:

it will cost much time if I create android-4.3.1.1.1 from begining( get code form google -> apply all patch, getting code from google is very slow through my internet).