Android P development 2

[Spookcity]

A continuation of #26

[skulko]

I patched the file with what I was able to link to the new source, but no improvement. Maybe some less errors/warning relating to codecs, but not sure.

Maybe we could need an additional shim for graphicsbuffer like here , because this looks really bad in log when trying to record: ExynosCamera: [CAM_ID(0)][]-ERR(m_handlePreviewFrame[4377]): Recording buffer is not available!! Recording Frames are Skipping(500 frames) (bufIndex=-2)

[Spookcity]

Interesting. Hopefully I'll be able to find some time this weekend and I'll connect to adb and try and record a video and see what it spits out. It's been a long time since I've done this. I know how to create shims for missing symbols, but would have to do some researching on creating other types of shims.

[skulko]

that would be awesome... because I don't know what exactly I'm doing sometimes :) I tried adding the libexynoscamera_shim, also found here, if you'd like to take a look.

[Spookcity]

Trust me,you are not alone in that. I just constantly search Google/Github/Gerrit :) I will try to look into this stuff over the weekend as long as I have the time. Time is just something I've really been lacking lately sadly.

[skulko]

You're probably right... :-/ I guess there is too much different hardware and abbreviations around. I wonder, how any single person can keep track..

For the camera, it might all boil down to this:

"Metadata in video buffer" mode is removed from Camera2Client and only "buffer queue" mode is supported.

No idea if it would work after reverting so much code...

[skulko]

...oh support for metadata in video buffer was restored I think. So no problem there. this is missing, however:

To include the Camera app in your device's system image, specify it in the PRODUCT_PACKAGES variable in your device's device///device.mk makefile: PRODUCT_PACKAGES := \ Gallery2 \ ... (https://source.android.com/devices/camera

[Spookcity]

Ahh,nice. I was hoping to spend some time on this over the weekend, as well as build an update for LOS 16 since the April security patch is now merged, but once again time was not on my side. Hopefully this week. I will try a build with this and see how it works. Thanks again for looking into all of this. I appreciate it and really hope I'm able to find the time soon to help out some.

[skulko]

The problem is most likely related with video buffer or something similiar in hardware. Someone would probably have to implement this overlay for hardware/samsung which has some stuff for exynos that looks useful. I don't see any other alternatives...

[Spookcity]

Yep,I think you are probably right about that. I'm not really sure how to go about implementing something similar for our device. I'd have to do some research on that.

[skulko]

I recently looked over your OC commits in kernel and in line 147 in /drivers/gpu/arm/mali400/mali/platform/exynos3470/exynos4_pmm.c I might have found a mistake: is 'cchar' correct there or should it be 'char'? Because the original file says 'char'

[Spookcity]

@skulko Hah,interesting. I haven't messed with this in forever,but I am sure you're correct and that's a typo on my part. Thanks for pointing it out!

[skulko]

Hello Spook

May I ask, how did you find the value for TARGET_ADDITIONAL_GRALLOC_10_USAGE_BITS := 0x4000000 (commit bea98ef37c67eeaafbb8d57e92c6a9933787aeaa)?

I'm asking because I might have found a solution for this frequent error W MapperHal: buffer descriptor with invalid usage bits 0x400 here: https://github.com/LineageOS/android_device_samsung_jf-common/commit/7e2dd5cd9c18133f1648f196338130eb51f73e12

Best,

[Spookcity]

Hey @skulko , Interesting. I think I looked into this commit awhile ago,but can't say whether I actually tried it or not.Plus,at this point you've made big improvements in things. So even if I did try and it didn't work then. It may now.

It has been awhile,but I got the specific value simply by connecting to adb shell and looking at the error/warning while running logcat Good luck!