Hello everyone,

I'm currently trying to get an Android port running on a custom
device. It is an Armadeus APF51 Board with a touchscreen attached,
basic details are here:

Processor: Freescale i.MX51 (Cortex-A8 @ 800MHz)
RAM: Mobile DDR 400. 64 to 512MB. 32 bits data bus. Default capacity
will be 256MB.
Flash: Mobile NAND. 512MB, 8 bits data bus.
Ethernet: onboard Physical (ready to use Ethernet 10/100Mbit link)
USB: High speed USB OTG (OnTheGo) with onboard Physical (ready to use
USB OTG link)
USB: 2 High speed Hosts with integrated PHY
FPGA: Xilinx Spartan 6 (LX9 or LX16). Default: LX9
Touchscreen controler (4/5 wires)

More details can be found on http://www.armadeus.com/wiki/index.php?title=APF51


What do I have already?

I have my aosp-android-2.6.38 kernel, where I had to apply some
patches specific to the board I use. (I have to use this kernel
version, because the board-developers supply patches for that kernel
version). This seems to be fine as the kernel is bootable and can give
a console.

However, my Android rootfs (tried boot via NFS and MMC) ends in a loop
while booting, and shows the A N D R O I D Ascii-like on the
touchscreen permanently.

According to the log, it seems to be a problem with dalvikvm, so
basically kernel and java-environment in my rootfs seems ok.

I'm using aosp-gingerbread and also tried binaries supplied from
Freescale here: 
http://www.freescale.com/webapp/sps/site/prod_summary.jsp?code=IMX51ANDROID&fpsp=1&tab=Design_Tools_Tab
(10.3)

Here is my full log: http://pastebin.com/iFY1emn9 (output from kernel
and logcat)
Please have a look especially at line ~370 and below.
The bottom part of the log repeats permanently in a loop.

"No vspace available" seems to be problematic, although the device has
enough memory imho.
Has anyone an idea what the problem could be?

I have access to the android shell.

-- 
unsubscribe: [email protected]
website: http://groups.google.com/group/android-porting

Reply via email to