2.6.32 Status

************Issue list for 2.6.32 kernel port********************
Please update this doc with the results of your work or investigations.

Git for the 2.6.32 kernel: https://github.com/fat-tire/android_kernel_omap/

(Note it's forked so that [mbm] can push too)

Currently the two biggest issues are the MMC driver and the framebuffer driver.
It appears that once we get these two issues resolved we might at least have a bootable kernel.

UPDATE:  Now we can mount /system, though there are other issues.  See:

http://pastebin.com/raw.php?i=vMJXaTuY

Also, we need new 2.6.32 sfx modules in:

/system/bin/sfx

SOURCE FOR /system/bin/sgx:
http://software-dl.ti.com/dsps/dsps_public_sw/sdo_tii/TI_Android_DevKit/02_02_00/index_FDS.html

Also, we need 2.6.32 modules for /system/bin/sgx (GFX stuff):

MAKING GFX FROM SDK INSTRUCTIONS:
http://processors.wiki.ti.com/index.php/OMAP35x_Graphics_SDK_Getting_Started_Guide

GFX FILE: 
https://focus-webapps.ti.com/licreg/productdownload.tsp?toPerform=productDownload&downloadPage=true&location=http://software-dl.ti.com/dsps/dsps_public_sw/sdo_sb/targetcontent/gfxsdk/latest//exports/Graphics_SDK_setuplinux_4_03_00_01.bin

PSP FILE (required as described in instructinos)
https://myportal.ti.com/portal/dt?lt=myti&provider=TIPassLoginSingleContainer&lt=myti&URI=https%3A%2F%2Ffocus-webapps.ti.com%3A443%2Flicreg%2Fproductdownload.tsp%3FdownloadPage%3Dtrue%26location%3Dhttp%3A%2F%2Fsoftware-dl.ti.com%2Fdsps%2Fdsps_public_sw%2Fpsp%2FLinuxPSP%2FOMAP_03_00%2F03_00_01_06%2F%2F.%2Fexports%2FAM35x-OMAP35x-PSP-SDK-03.00.01.06.tgz%26toPerform%3DproductDownload

My attempt to build framebuffer kernel modules was a FAILURE:

http://pastebin.com/H6iJvMhd
http://pastebin.com/5p8LxAYg

Not sure which version to build exactly as you have a few choices.

The following is a kmsg from 2.6.32 with i2c debug: http://pastebin.com/zzF4QW0T
######################### Update after display fix: http://pastebin.com/UazQ9Vmh
############## Update 2 after few board file fixes: http://pastebin.com/zKwm9Pey
## Update 3 after changing init to reflect mmcblk1: http://pastebin.com/ZwURQhAg

For comparison here is a 2.6.29 kmsg: http://pastebin.com/b9H2aWNj

******Known issues listed in descending priority(as decided by Dalingrin)*******

-Issue #1 MMC driver. 
Currently little or no twl4030 devices are loading. The twl4030 multidevice controller houses the mmc controller as far as I know.
As defined in the board file (board-encore.c) "tps65921" is the name given to the mmc device and can be seen being registered on the i2c bus
on line 118 of the 2.6.32 kmsg. 
However, what cannot be seen in the log is that I have printk statements all over the mmc initialization functions and none of which get called. 
So, either the entire twl4030 chip or the mmc controller portion is failing somewhere between being registered and initialization.
###Update: Changing an i2c identifier for the twl4030 now gets MMC to at least try to load. In fact, it now recognizes the mmc but does not mount
           or see any partitions. http://pastebin.com/zKwm9Pey
###Update: MMC loads and partitions are recognized however, eMMC is set as mmcblk1 vs mmcblk0. Changing init to reflect this causes a reboot after 
           /system is mounted. http://pastebin.com/ZwURQhAg

############## FIXED ################ -Issue #2 Omapfb   
#As can be seen in the logs, omapfb is not getting started correctly. I believe this is the reason for the garbage on screen after the backlight 
#turns on during boot. I haven't looked into this at all yet so there could be an easy fix.
###Update: Somehow vram was set to 0. Updating this to 8mb results in no more garbage on screen, instead I get a beautiful tux logo. However,
#the display driver then KPs.
#####################################

-Issue #3 Sound
I have disabled the SOC sound codec because it causes a kernel panic. This can be revisted after we solve more prudent issues like those above.

-Issue #4 MUSB
I have disabled MUSB in the config because it also causes a kernel panic. Again, this can be revisted once we solve MMC and the framebuffer.
(Also, it appears that we have a bug fix for broken USB so check out again.)

**************************************************************