Hi, i'm trying to compile cm9, i got it to boot and show the start screen for cm9 but then it hangs, razor or any one can share the device/vendor files for the huawei s7 fat not he slim version, so i can compare the files and see what i'm doing wrong.
Any links for rom? Even 4 years(soon 5) passed many of us still use this tablet and want good rom for it,so if you need any help just say and pls upload rom somewhere :/
I built kernel 2.6.32 with vmsplit 3g, which is required for cm7, yesterday and rictec compiled a cm7 build with that kernel and it booted fine, at least once battery was removed.
Basically, its not in a usable state at all but the fact we got it booting is nice progress and we're working on fixing each issue at a time.
Issues:
No Radio, No Wifi, No Camera
No USB (Mounting / ADB broken)
Battery has to be removed if not blue screen occurs (doesn't happen to all testers, might be specific to some S7's Revisions)
rictec deserves most of the credit since he is the one who did the vendor config.
I am currently cleaning that vendor config up and progress is going fine.
Currently: Discussion is mostly done on irc but discussing on here is fine.
irc channel #androidtablets on freenode irc server.
Current Developers/Tester:
Razor950
Rictec
Joenilan
WillEY_
I will post a link to a extremely alpha release once we get usb working but if you would like to test, just drop by the irc
Ok, merged the init's and stuff, but noticed in the end the problems were down to one permission in ueventd.rc ....
/dev/kgsl-3d0 0666 root root
/dev/kgsl-2d0 0666 root root
gaah!
That error occurs from system/core/rootdir/ueventd.rc . The huawei s7's kernel uses the "old" way of kgsl-3d0 and 2d0 instead of just kgsl 0/1
anyway, andreno is now back to usable in 2.2.2 , no corruption etc. http://threader.zapto.org/experimental/boot-new-1.img Testing it now with dragon fly.
Now i just need to back everything the hell up and try 2.3.7 =) (bootloop currently)
.orig btw because i was sidetracked with trying to increase the /data partition with a 2gig ext4 partition and use userdata as swap
I also just discovered some malconfiguration for my 10x atleast, in BoardConfig.mk
BOARD_BOOTIMAGE_PARTITION_SIZE := 0x00500000
BOARD_RECOVERYIMAGE_PARTITION_SIZE := 0x00500000
BOARD_SYSTEMIMAGE_PARTITION_SIZE := 0x0fa00000
BOARD_USERDATAIMAGE_PARTITION_SIZE := 0x0a000000
BOARD_FLASH_BLOCK_SIZE := 131072
Isnt this still an option? http://gitorious.org/cm4l/vendor/bl...1a8af28f1/cyanogen/products/cyanogen_salsa.mk
#Build with 2G map (Acer used for their ROMs), or remove this line to build with 3G/1G (standard build that works with phhusson's Froyo kernel)
#PRODUCT_SPECIFIC_DEFINES += TARGET_PRELINKER_MAP=$(TOP)/vendor/cyanogen/prelink-linux-arm-liquid.map
Anyway, i fixed up the kernel some more, discovered a problem with the previous 3g split one using some flags that didnt work. Tested this one properly for a good few in linux before releasing.
To get some useful information.
fastboot -c 'androidboot.hardware=qsd8k_s7' boot boot.img
I dont know exactly what im doing wrong here, but no matter how much i fiddle with executable binaries, libraries, frameworks or whatever else, i still cannot get this to boot properly. If anyone's a rom wiz please give me a shout, experience on the qsd8k would help! There has to be something obvous im missing. It gets to the point where its launching zygote, then two processes crash, and this it can do all year long if i leave it.
txpxsx
I compiled that kernel some weeks ago for my Gnu/Linux project ( http://xdaforums.com/showthread.php?p=14417403#post14417403 ), making the sources compile required hours for fixing errors and ponds of coffee, who did the kernel work for you guys? I believe the andreno is used using a propriety library on android, isnt it? I will certainly try compiling 2.3.7 tho.
i have 2.3.7 compiled, i just need to figure out how to make a boot image the device will accept, compiled 2.3.7 tonight, seems the 3g split doesnt boot here, probably runs out of address apace?.
Okay, im at the gfx corruption bit now, thing is , it also happens with 2.2.2, so the problem is probably in the ramdisk, i've used this kernel with 2.2.2 before, but then with the old ramdisk