Related
>>> With UART access NookTab Secure BOOT Chain as been FULLY BROKEN, Custom Kernel and Custom Ramdisk have been succesfully run on NookTab, Look towards 2nd page or so for full info <<<
Hi,
Few days back I had got an idea to try and see if we can BYPASS the boot security chain by replacing the bootloader in memory, because NOOKTAB allows UART ACCESS to UBOOT.
My initial thought was to use a replacement UBOOT without Security checks. However on further thought, as UBOOT has memory access commands, I realised the simpler solution is to edit the UBOOT code directly in memory from UBOOT prompt itself.
In turn I had posted the concept and the commands to try and do the same on the below two threads, for people to try. However as no one seems to have tried it yet, I myself opened up the my NookTab and connected the UART signals and am continuing my experiments and the initial results are promising.
FINDING1: The MShield security logic doesn't mind if one modifies the UBOOT CODE. I was able to NOP the security check result logic check and the code continued to boot.
Next I have to try a modified RAMDisk and see it works fully.
My earlier posts on this can be got from these two threads
http://forum.xda-developers.com/showthread.php?t=1378886
http://forum.xda-developers.com/showthread.php?t=1418172
For someone interested in experimenting with this below are the commands to try on UART of NOOKTAB.
uboot Command summary
---------------------------------------
md.l address_in_Hex ---------- To cross check the memory content before overwriting (should match what I have mentioned as ORIG)
mw.l address_in_Hex 4ByteValueInHex -------------- To modify the given address location with new value
md.l address_in_Hex -------------- To cross check that the new value you have written has come properly.
Command sequence for Ramdisk check bypassing
-----------------------------------------------------------------------------
UBOOTPROMT> md.l 80e84808 ----- This should show 1a00000a
NOTE: I have verified that the 2nd possibility mentioned in my earlier post i.e 0x80e8.0000-0x120 is the load address to use to calculate the offsets.
next run
UBOOTPROMPT> mw.l 80e84808 e1a00000 ------------- This modify with NOP
Next run
UBOOTPROMPT> md.l 80e84808 ------ should show e1a00000
Next if you have updated the recovery.img with new ramdisk into /recovery partition RUN
UBOOTPROMPT> mmcinit 1; booti mmc1 recovery
HOWEVER instead if you have updated the flashing_boot.img file with new ramdisk in microSD then RUN
UBOOTPROMPT> mmcinit 0; fatload mmc 0:1 0x81000000 flashing_boot.img; booti 0x81000000
Now it should boot with out giving a signature error.
NOTE1: I have verified that changing the contents of UBOOT (i.e NOPing) in itself doesn't lock the ARM, next I have to try a updated ramdisk and see what happens. If you ask me It should work, fingers crossed, I will try and update.
NOTE2: In any android img file at offset 0x10 (i.e 16) the ramdisk size is stored as a 4 byte (long) value. Cross verify first that the original img and the ramdisk size at offset 0x10 in it matches the original ramdisk. Then update the 0x10 offset of new img file with new ramdisk's size.
NOTE3: kernel security check bypass address = '0x80e847a0'
[REPOSTING OLD, CONCEPT] BYPASS Kernel and Ramdisk check for People with UART ACCESS
**************************
>>> This was my original post to the other two threads on this concept, I have put this here for completeness. The load address confusion which I had is already resolved <<<
****************************
Hi,
NOTE: THis is based on a initial look at the source code and then the objdump of u-boot.bin. I haven't cross checked this yet, because for now I haven't opened up the nooktab for uart access yet. Also this assumes by default booti command is used for booting in BN uboot. If some one wants to use bootm, then a different location requires to be patched wrt the image loading security check.
If you are a lucky ;-) person working with opened up NookTab with UART access, then basically replacing the memory contents of these two offsets with NOP will 90% BYPASS the security check successfully and allow you to boot a MODIFIED KERNEL or RAMDISK as required.
All offsets specified Assuming u-boot is loaded at 0 (adjust for the actual address where u-boot.bin is loaded, haven't looked into that yet).
Check for Security check of Kernel image is at
[ORIG] 0x48c0 => bne 0x48d8 (0x1a00.0004)
Make this a NOP by overwriting using uboot memory write command to
[MODI] 0x48c0 => mov r0, r0 (0xe1a0.0000)
Check for Security check of RAMDisk image is at
[ORIG] 0x4928 => bne 0x4958 (1a00.000a)
Make this a NOP by overwriting with
[MODI] 0x4928 => mov r0, r0 (0xe1a0.0000)
Someone (Hi Adamoutler, maybe you) with opened up NookTab can try this and tell me if it worked or not.
NOTE: you have to add up the actual u-boot load address to the offsets specified.
UPDATE1: It appears the load address is either
Possibility 1) 0x80e8.0000 OR
Possibility 2) 0x80e8.0000-0x120 (More likely).
Have to dig thro bit more, but one of these two will potentially work.
So that means to NOP RAMDisk security check the offset is
Possibility 1 ==> 0x80e8.0000+0x4928
Possibility 2 ==> 0x80e8.0000-0x120+0x4928 (More likely)
Best is to cross check if the resultant address contains the BNE instruction bytes specified above.
Same concept applies for the Kernel security check Nopping offset.
NOTE: It appears there is a 0x120 size header before the actual u-boot.bin code starts and in turn, when I did the objdump, it included the 0x120 bytes of header also assumed as code. And inturn the full (including the header) u-boot.bin or for that matter the u-boot from emmc seems to load into 0x80e8.0000-0x120.
UPDATE 2:
Code around the locations to be noped to help identify the same in memory, in case my offset calculations are wrong
48b4: eb0030f1 bl 0x10c80
48b8: e59d3010 ldr r3, [sp, #16]
48bc: e3530000 cmp r3, #0
48c0: 1a000004 bne 0x48d8
48c4: e59f0104 ldr r0, [pc, #260] ; 0x49d0
48c8: e594100c ldr r1, [r4, #12]
48cc: e5942008 ldr r2, [r4, #8]
48d0: eb0015db bl 0xa044
............
491c: eb0030d7 bl 0x10c80
4920: e59d3010 ldr r3, [sp, #16]
4924: e3530000 cmp r3, #0
4928: 1a00000a bne 0x4958
492c: e59f00a4 ldr r0, [pc, #164] ; 0x49d8
4930: e5941014 ldr r1, [r4, #20]
4934: e5942010 ldr r2, [r4, #16]
4938: eb0015c1 bl 0xa044
UPDATE 3: ... for a rainy day in future ;-)
UPDATE 4: For maximum success, first try a changed RAMDisk rather than Changed Kernel. If Changed Ramdisk works then try Changed Kernel (THere is one more thing in Code, which I am not sure if it will impact a modified kernel or not yet, only way is to experiment).
UPDATE 5: I have cross verified on the target with UART access and the 2nd possibility mentioned above wrt load address is what is correct.
android img header structure for reference
from tools/mkbootimg/bootimg.h
#define BOOT_MAGIC "ANDROID!"
#define BOOT_MAGIC_SIZE 8
#define BOOT_NAME_SIZE 16
#define BOOT_ARGS_SIZE 512
struct boot_img_hdr
{
unsigned char magic[BOOT_MAGIC_SIZE];
unsigned kernel_size; /* size in bytes */
unsigned kernel_addr; /* physical load addr */
unsigned ramdisk_size; /* size in bytes */
unsigned ramdisk_addr; /* physical load addr */
unsigned second_size; /* size in bytes */
unsigned second_addr; /* physical load addr */
unsigned tags_addr; /* physical addr for kernel tags */
unsigned page_size; /* flash page size we assume */
unsigned unused[2]; /* future expansion: should be 0 */
unsigned char name[BOOT_NAME_SIZE]; /* asciiz product name */
unsigned char cmdline[BOOT_ARGS_SIZE];
unsigned id[8]; /* timestamp / checksum / sha1 / etc */
};
PARTIAL SUCCESS BYPASSING SEC CHECK using NOP
Hi
By BYPASSING both the Kernel and Ramdisk checks using NOPs, I am able to run the kernel (not modified, but repackaged, so bypassed Kernel sec check) and modified ramdisk.
However either
a) I seem to have done something wrong OR
b) Secure boot chain is doing something internally before passing control to uboot during kernel sec check, which is different between a successful call and a bad call.
Because the kernel crashes after control passes to it, almost immidiately.
NOTE: Have to try with only ramdisk change ...
The UART Dump of my run is given below.
OMAP44XX SDP # booti 0x81000000
[ERROR] [SEC_ENTRY] Call to Secure HAL failed!
kernel @ 80088000 (2689312)
[ERROR] [SEC_ENTRY] Call to Secure HAL failed!
ramdisk @ 81080000 (513429)
Initrd start : 81080000 , Initrd end : 810fd475Acclaim Board.
Starting kernel ...
undefined instruction
pc : [<800886e4>] lr : [<80e930c0>]
sp : 80e3fac4 ip : 00028f05 fp : 80eabe44
r10: 810fd475 r9 : 80eb1fb8 r8 : 80e3ffdc
r7 : 80088000 r6 : 00000000 r5 : 80e3ffb4 r4 : 80eb1fb8
r3 : 00000000 r2 : 80000100 r1 : 00000e18 r0 : 00000000
Flags: nZCv IRQs off FIQs on Mode SVC_32
Resetting CPU ...
NOTE: This requires UART access to NookTab.
UPDATE 1: I found one mistake in that the unpack tool was always using a fixed size 2048 for page size rather than 4096 in the BN recovery.img, I fixed it and repackaged the new set of files and now even thou success eludes me, I find that this time it didn't give a SEC ERROR for my modified ramdisk !?!?!? But it was slower with the checks this time.
OMAP44XX SDP # booti 0x81000000
kernel @ 80088000 (2687264)
[ERROR] [SEC_ENTRY] Call to Secure HAL failed!
ramdisk @ 81080000 (513416)
Initrd start : 81080000 , Initrd end : 810fd468Acclaim Board.
Starting kernel ...
SUCCESS SUCCESS SUCCESS with Modified Ramdisk
Hi All,
SHORT form for impatient people
-------------------------------------------------
OMAP44XX SDP # mmcinit 0; fatload mmc 0:1 0x81000000 new.recovery.img;
OMAP44XX SDP # md.l 80e84808 1; md.l 80e847a0 1; mw.l 80e84808 e1a00000; md.l 80e84808 1; md.l 80e847a0 1
OMAP44XX SDP # booti 0x81000000
LONG form for people who want bit more details
---------------------------------------------------------------------
I have been able to boot into a modified recovery image using my NOP based BYPASS logic for secure boot chain.
What I learnt in the process are
a) Secure boot chain logic doesn't bother if we change the UBoot / XYZ code space Key to any logic using/manipulating the memory of the NookTab from uboot.
b) The Android img images for BN NookTab contain
b.1) The standard 2K Android header (nothing special from BN in this).
However NOTE that pagesize is 4096 and a good base address (picked from recovery.img of factory.zip) is 0x80080000
b.2) The Kernel and the Ramdisk images with in the android img file in turn contain 0x120 Byte headers individually
b.3) The Secure Boot chain seems to be particular about these 0x120 byte headers
Even for my modified ramdisk, I had to use the original ramdisks' BN Header. Otherwise the security check seemed to take a hell lot of time most of the time and the end results were touchy (Have to debug this further ..., ALSO THERE IS THE OPTION OF AVOIDING THE SEC_ENTRY call in the FIRST PLACE ITSELF TO TRY AND BYPASS THIS, IF REQUIRED, I have to experiment this later).
So if one is using a tool which searchs for the GZIP MAGIC to decide where to split the img file into strictly two parts consisting of
dump_1) Android_Image_Header+Kernel_BNHeader+Kernel+Ramdisk_BNHeader and
dump_2) Ramdisk file
are fine.
However if one is using a program which uses the Android image header structure to dump the contents need to be careful to extract the BN header from the corresponding ramdisk file and then after manipulating/modifying the ramdisk file, RE PREPEND the BN header back to the ramdisk. Before clubing/joining all the files together.
Or tools which assume wrong pagesize (some I found used 2K page size instead of picking from android header) or which split the constituents into individual parts intelligently (which by the way will discard the BN Header potentially) will have to be MODIFIED before using.
I ended up writing my own c code to dump using Android header and inturn use shell script to extract the BN Header for safe keeping before merging everything back later. I will post the code and simple shell scripts in a day or two.
BELOW is the OUTPUT OF MY SUCESSFUL RUN with MODIFIED RAMDISK
--------------------------------------------------------------------------------------------------------------------
OMAP44XX SDP # mmcinit 0; fatload mmc 0:1 0x81000000 new.hdr.img;
3207168 bytes read
OMAP44XX SDP # md.l 80e84808 1; md.l 80e847a0 1; mw.l 80e84808 e1a00000; md.l 80e84808 1; md.l 80e847a0 1
80e84808: 1a00000a ....
80e847a0: 1a000004 ....
80e84808: e1a00000 ....
80e847a0: 1a000004 ....
OMAP44XX SDP # booti 0x81000000
kernel @ 80088000 (2682952)
[ERROR] [SEC_ENTRY] Call to Secure HAL failed!
ramdisk @ 81080000 (513707)
Initrd start : 81080000 , Initrd end : 810fd58bAcclaim Board.
Starting kernel ...
Linux version 2.6.35.7 ([email protected]) (gcc version 4.4.1 (Sourcery G++ Lite 2010q1-202) ) #1 SMP PREEMPT Fri Nov 11 12:35:42 PST 1
CPU: ARMv7 Processor [411fc093] revision 3 (ARMv7), cr=10c53c7f
CPU: VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
Machine: OMAP4430 ACCLAIM
Memory policy: ECC disabled, Data cache writealloc
...........
That all looks very good and sounds extremely promising
So in the realms of being able to boot modified roms where does this put us?
Uses of this method and its merits/demerits or needs/no needs
Hi Celtic/All,
*** The only places I see a meaningful use for this is mentioned towards the end, otherwise, I am working on this mainly for the fun of exploration. In most other cases where one thinks one needs this, I can tell you It can be achieved with out this, except for some exotic things which don't affect majority of users ***
My thoughts on ROMs with different complexities
------------------------------------------------------------------------
If a modified ROM is not using (doesn't need) custom Kernel or custom Ramdisk, then my 2ndihkvc or any other working 2nd-init method is the simple and straight forward way of doing a custom rom.
If it however requires a custom Ramdisk, then this NOP based BYPASS method will allow one to achieve the same. However I don't see any need for anyone to use a custom ramdisk. If someone is using a custom ramdisk, It can be 98% modified to use the generic 2nd-init method and in case of NookTab my simple 2ndihkvc method (As documented in my other thread, the default 2nd-init logic fails on NookTab as it uses ONE too many PTRACE calls).
If it requires a custom Kernel then with bit more work on this I don't see why a Custom kernel cann't be booted.
However if you ask me, we don't gain much with a custom kernel or ramdisk, which cann't be achieved using root access and or module loading support in default kernel, and inturn REMEMBER that both of these can be done on NookTab today (i.e 1. Root access and 2. Module loading).
Also NOTE that this requires UART access.
*** ONE PLACE WHERE THIS CAN HELP *** is, with BN 1.4.1 firmware, which has blocked the current rooting method If I am not wrong (Unless someone has found a way to break it recently, which I have missed). For 1.4.1, with this, we can boot into a specific custom recovery image and modify the /system partition, such that we put su and SuperUser back into it under /system/bin (with proper chmod settings) and /system/app, so that we can gain Root access again, on rebooting into the NookTAB normally after this change.
*** Another place *** is when the device is very old and the new kernel can bring in some feature missing badly in a very old device. Again in many of these cases, if one puts sufficient effort the feature may be back portable and or injectable into a older kernel using the module route.
REMEMBER IN LINUX - KERNEL MODULE IS SAME AS KERNEL as far as PRIVILAGES are concerned, as it stands TODAY, all LIMITS IF ANY are ARTIFICIAL.
HOPE THAT HELPS
I love reading these threads even though I don't fully understand everything going on in the code parts.
I'm interested in custom kernels because as far as I know there's no way to get ICS running on the Tab without one.
Nexus S 4G 4.0.3
I honestly dont know much about creating custom ROM's but I have been wondering why every thinks that we have to have the bootloader unlocked before we can get any type of custom ROM. I have a Moto X2. The bootloader is not and never will be unlocked but I am running a really sweet custom ROM on it. I know from other android phones that a ROM is possible with a locked bootloader.
My point is...I am glad to see someone working around this and taking the next step. I was wondering if DEV's have almost given up on the NT. Thank you for your work!
Rooting 1.4.1
hkvc said:
... BN 1.4.1 firmware, which has blocked the current rooting method If I am not wrong (Unless someone has found a way to break it recently, which I have missed).
Click to expand...
Click to collapse
See my method at http://forum.xda-developers.com/showthread.php?t=1413734 (since Dec 27)
Note that my method starts with either a rooted or (preferably) unrooted copy of 1.4.0, roots it if necessary, modifies it slightly, updates to 1.4.1, and then regains root. Requires ADB/USB access.
POTENTIAL SUCCESS with CUSTOM KERNEL (INDIRECT METHOD)
Hi All,
In SHORT for impatient
---------------------------------
OMAP44XX SDP # mmcinit 0; fatload mmc 0:1 0x81000000 new.hdr.img;
OMAP44XX SDP # md.l 80e84794 1; md.l 80e847fc 1; mw.l 80e84794 e1a00000; mw.l 80e847fc e1a00000; md.l 80e84794 1; md.l 80e847fc 1
OMAP44XX SDP # md.l 80e84808 1; md.l 80e847a0 1; mw.l 80e84808 e1a00000; mw.l 80e847a0 e1a00000; md.l 80e84808 1; md.l 80e847a0 1
OMAP44XX SDP # booti 0x81000000
(c) HKVC, GPL ;-)
The sufficient minimal Details
-------------------------------------
I have verified that NOT CALLING SEC_ENTRY calls, with in uboot, related to kernel and ramdisk check keeps things smooth. That should mean the FLOOD GATES are POTENTIALLY OPEN for CUSTOM KERNELs with UART ACCESS.
This requires few additional NOPs compared to what I had originally specified (My original set of NOPs had some issue with Kernel booting, which I have to debug later, however this work around seems to resolve it - I don't want to delve more into this than what I have already specified here, unless Secure Bootloader people get any ideas ;-).
UART Boot Dump/log
-------------------------
OMAP44XX SDP # mmcinit 0; fatload mmc 0:1 0x81000000 new.hdr.img;
3207168 bytes read
OMAP44XX SDP # md.l 80e84794 1; md.l 80e847fc 1; mw.l 80e84794 e1a00000; mw.l 80e847fc e1a00000; md.l 80e84794 1; md.l 80e847fc 1
80e84794: eb0030f1 .0..
80e847fc: eb0030d7 .0..
80e84794: e1a00000 ....
80e847fc: e1a00000 ....
OMAP44XX SDP # md.l 80e84808 1; md.l 80e847a0 1; mw.l 80e84808 e1a00000; mw.l 80e847a0 e1a00000; md.l 80e84808 1; md.l 80e847a0 1
80e84808: 1a00000a ....
80e847a0: 1a000004 ....
80e84808: e1a00000 ....
80e847a0: e1a00000 ....
OMAP44XX SDP # booti 0x81000000
kernel @ 80088000 (2682952)
ramdisk @ 81080000 (513707)
Initrd start : 81080000 , Initrd end : 810fd58bAcclaim Board.
Starting kernel ...
Linux version 2.6.35.7 ([email protected]) (gcc version 4.4.1 (Sourcery G++ Lite 2010q1-202) ) #1 SMP PREEMPT Fri Nov 11 12:35:42 PST 2011
CPU: ARMv7 Processor [411fc093] revision 3 (ARMv7), cr=10c53c7f
CPU: VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
Machine: OMAP4430 ACCLAIM
Memory policy: ECC disabled, Data cache writealloc
On node 0 totalpages: 245760
And if the merge bootloader of the nook color and from nook tablet, compare it and try to create one substitution?
DeanGibson said:
See my method at http://forum.xda-developers.com/showthread.php?t=1413734 (since Dec 27)
Note that my method starts with either a rooted or (preferably) unrooted copy of 1.4.0, roots it if necessary, modifies it slightly, updates to 1.4.1, and then regains root. Requires ADB/USB access.
Click to expand...
Click to collapse
Hi DeanGibson,
Thanks for your efforts on that. It should help people who get bumped into 1.41 by BN.
HOWEVER Do note that if the uSD based MLO and u-boot.bin gets loaded first before the ones in eMMC by the internal boot rom of the Omap (Should be the case based on what Pokey had mentioned sometime back, I haven't cross checked myself yet, as I have been busy with these stuff which I am looking into). Then what ever (except for one cavet - which I wont mention here) BN may do in a future update, with the UART based u-boot method which I have mentioned in this thread, one will always be able to get root access to the device.
OMG hkvc, between you, DG, and AO how can the NT win?
You guys are monsters! (in a good way)
Ok, little explaining before questioning is I'm not a tech guy. But from the all post of hkvc in this thread, the understanding that we can access more space in 16GB internal storage and custom ROM/kernel is on the way is correct?
camapghe said:
Ok, little explaining before questioning is I'm not a tech guy. But from the all post of hkvc in this thread, the understanding that we can access more space in 16GB internal storage and custom ROM/kernel is on the way is correct?
Click to expand...
Click to collapse
This requires repartitioning the drive, which we are not at this time confident, that it will not brick your nook by doing so. (This double negative actually is making a positive: Repartitioning might brick your nook.)
This hardware modification has nothing to do with accessing more of the space as that is entirely a software remedy. We're just not confident about how hard the Nook looks at the primary partition table.
SUCCESS SUCCESS SUCCESS with CUSTOM Kernel+ CUSTOM Ramdisk, UART NOP BYPASS
Hi All,
As I had mentioned yesterday/today early morning, By bypassing the SEC_ENTRY check I was able to run stock kernel with out any problem. And as I had mentioned then even thou it is a indirect way of verifying possibility of custom kernels, it should still open the flood gate for custom kernels (with UART access for NoW ;-).
Now I have actually verified by RUNNING a CUSTOM Kernel which I compiled along with a CUSTOM Ramdisk (with adb enabled - look at last few lines), which you can know from
a) the kernel version line while booting, which contains the machine used for compiling (Obviously I have redacted part of my name ;-),
b) as well as the size of the kernel and ramdisk images which is different from the stock img files, because this contains both a custom kernel as well as custom ramdisk from me.
SO IT IS SUCCESS with CUSTOM KERNELS+ CUSTOM RAMDISKS, using the uboot commands which I had mentioned in my older post.
UART DUMP including UBoot commands
--------------------------------------------------
OMAP44XX SDP # mmcinit 0; fatload mmc 0:1 0x81000000 new.kr.img;
4157440 bytes read
OMAP44XX SDP # md.l 80e84794 1; md.l 80e847fc 1; mw.l 80e84794 e1a00000; mw.l 80e847fc e1a00000; md.l 80e84794 1; md.l 80e847fc 1
80e84794: eb0030f1 .0..
80e847fc: eb0030d7 .0..
80e84794: e1a00000 ....
80e847fc: e1a00000 ....
OMAP44XX SDP # md.l 80e84808 1; md.l 80e847a0 1; mw.l 80e84808 e1a00000; mw.l 80e847a0 e1a00000; md.l 80e84808 1; md.l 80e847a0 1
80e84808: 1a00000a ....
80e847a0: 1a000004 ....
80e84808: e1a00000 ....
80e847a0: e1a00000 ....
OMAP44XX SDP # booti 0x81000000
kernel @ 80088000 (2693828)
ramdisk @ 81080000 (1455055)
Initrd start : 81080000 , Initrd end : 811e32afAcclaim Board.
Starting kernel ...
Linux version 2.6.35.7 ([email protected]) (gcc version 4.5.4 (Ubuntu/Linaro 4.5.3-9ubuntu1) ) #1 SMP PREEMPT Wed Jan 4 02:43:18 IST 2012
CPU: ARMv7 Processor [411fc093] revision 3 (ARMv7), cr=10c53c7f
CPU: VIPT nonaliasing data cache, VIPT nonaliasing instruction cache
Machine: OMAP4430 ACCLAIM
Memory policy: ECC disabled, Data cache writealloc
On node 0 totalpages: 245760
..... Chopped ...............
omapfb omapfb: Unknown ioctl 0x40044620
init: Unable to open persistent property directory /data/property errno: 2
enabling adb
adb_open
android_usb gadget: high speed config #1: android
SO ALL OF YOU out there ITHCING to experiment with Custom Kernels and What not, Go ahead and enjoy the freedom to do so on NOOK TABLET (with UART access for NoW ;-)
My Android Img file manipulation scripts including few older ones by others.
Hi,
I am attaching the simple C program and the scripts which I use for extracting BN Android Imgs consisting of
a) Dumping the individual sections of Android img
b) Allow seperating the header from the actual Kernel or Ramdisk
c) Allow concatenating (This is kind of dummy, but required to take care of u-boot logic of loading) the old header with new Kernel or Ramdisk
d) Pass proper arguments to recreate the Android IMG file.
Also I have attached some of the other open source tools which I started with originally, but due to few things here and there and also to get maximum flexibility I moved to my own set of scripts and program.
recovery img with simple ramdisk with ADB and Root shell on Adb and console
Hi,
Attached is a recovery.img file with the standard Kernel from NookTab and a modified Ramdisk which has support for
a) ADB shell
b) Root shell access (Both ADB and Console)
c) Console is enabled in UART.
Note that the sh on the ramdisk is renamed busybox with a symbolic link called busybox pointing to this sh.
go into /system/bin and run
busybox --install /system/bin
So that you have the standard commands available on the recovery shell.
Also remember to run
export PATH=/system/bin:$PATH
I am following this thread, Congrats on your findings and thanks for your time you spent on it.
I Have decided that this thread has served it's purpose and will now be closed to future posts. Please direct and 'non' SHV-E160L post's to
Brixfix V2
Please can all Ongoing jobs/works migrate to the above thread.
-----------Final Notes--------------
It has been mentioned many times that i should go back and correct the information below, i started to correct a few post's then realized i was removing the flavour in change of colour and size, parts of this thread documents my mistakes, assumptions and general lack of understanding of how we NOOBS post on XDA, It's with that in mind that i have decided to leave the mistakes in, so you can see in writing what i gained from the support of other Devs here.
Now, if you are NOOB in anyway or have a few questions please click HELP
If you are bricked and need help, read this thread first, there is NO one CLICK solution for anything, even this mentioned device.
-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
So you Brixed/bricced/BOD/QDL/EDLOAD/QHS-USB/05c6:9008/05c6:9025/ your device? Need a Oil and brush , Need help, follow this
One, Rules
Two, Understanding
--------------------------------------------------------------------------
Tip From the Author,
Some of you may have noticed that i did not start the original thread with a question, I did something my mentor taught me at around 9 years old but didn't put into good use until much later in life.
The tip is write things down as a question for yourself, in the writing process you get to pass the information past the part of your brain that interprets information, virtual sounding board, before posting as a question for others.
--------------------------------------------------------------------------
New Tools for debricking, goto
Brixfix V2
---------------------------Further Info Info -----------------------------
** I have Since Fixed the device and developed soultions for non shv-e160l devices. Prior posts are undergoing edit's for corrections.
** if you want the glory shot, sorry you will just have to read through.
** If you are selling this as a solution, dont. I know who you are.
---------------------------Original Post-----------------------------
Hi All
As i mentioned on this thread http://forum.xda-developers.com/showthread.php?p=32231827#post32231827 i will be attempting to come up with a home grown debrick solution for a SHV-E160L samsung note from korea.
I will use the forum to document what i am doing, i am very new to this so correct me please if i am wrong. I have never done Android dev work at any time but i have a very good understanding of the logic behind it all. `
Things i Have :-
Phone ( SHV-E160L)
bus pirate v3 with jtag firmware
openocd compiled on ubuntu and centos 6
smd jtag adapter and relay wire ( magnetic wire)
things i still need :-
openocd target config file for MSM8660 Snapdragon cpu (and a better understanding of eMMC access, how to load boot loaders either into ram or eMMC or trigger fail over boot to sc-card, USB via software or X0M/Boot pins)
assembled jtag (it's the smallest soldering i've ever seen)
.PIT file for 32GB model (if someone could pull the .PIT file from a working unit I would be happy, specify your radio/kernel versions when uploading)
micro fine solder iron tip and 20w iron (i've got 60w but too high for this type of work)
Does anyone have a idea of the SD-CARD partition layout, files for snapdragon devices, google has given me much for other devices but not a snapdragon .
Another question, I've used the USB jig to trigger 301K mode USB-Factory and seen no activity in dmesg for usb devices, i've yet to try windows, does windows/linux behave in a different way when it comes to usb , as in windows see's the qualcom usb mode but not linux ? does the usb client device always start the comms?
using the 615K usb jig i get nothing too, no pbl message from samsung (hence i am led to think is's the pbl/sbl thats damaged)
My understanding up boot is as follows
iROM code
This loads basic settings to boot the PBL (iROM is in rom) the PBL is loaded into radio(modem) cpu and then loads the SBL(s)
PBL/SBL stored in eMMC at address ????? (need to document the address for the masked access to eMMC and jtag/openocd access unmasked access)
Once the SBL is loaded you with have the ODIN mode (USB/UART)
from what i can see of commercial JTAG boxes is the access the radio cpu via jtag, write a new PBL/SBL to the eMMC then halt/reset cpu which now loads the new bootloaders, (resurrect dead body)
The openocd TAP id for the cpu should be 0x105310E1 but thats a number i got from a riff box log, not any actual testing ( still need to solder the fine pitch connector)
Here is a log from a riff box, not sure if the address's are usable accross to opencd
Taken from gsm-forums:-
Open serial port...OK
Connecting to the RIFF Box...OK
Firmware Version: 1.33, JTAG Manager Version: 1.44
Selected Resurrector: [Samsung E160K V1.0.4535.7001]
Connecting to the dead body...OK
Detected dead body ID: 0x105310E1 - IGNORED!
Set I/O Voltage reads as 1.79V, TCK Frequency is RTCK
Adaptive Clocking RTCK Sampling is: [Sample at MAX]
Resurrection sequence started.
Establish communication with the phone...OK
Initializing internal hardware configuration...OK
Uploading resurrector data into memory...OK
Starting communication with resurrector...OK
Detected an Initialized FLASH1 Chip, ID: 0x0015/0x0000 (KTS00M, 0x0003AB400000 Bytes = 14.68 GB)
Detected an Initialized FLASH2 Chip, ID: 0x0015/0x0000 (KTS00M, 0x000000200000 Bytes = 2.00 MB)
Flashing the dead body...OK
Resurrection complete!
Click to expand...
Click to collapse
I did notice one thing, the riff box opens the serial port, i wonder if they load PBL+SBL into memory, reset the cpu, then using the serial connection activate download mode ? (like on the captive)
I also dont know how the cpu (jtag TAP id? ) and flash variables translate accross to openocd as ive not found a target config file yet ( or my searching is wrong)
in the full stock Firmware I was able to extract the .tar file which contained,
Code:
amss.bin <-- application cpu boot files ?
boot.img <-- kernel/initrd ramdrive
mdm.bin <-- modem cpu boot files
recovery.img <--- recovery image
system.img.ext4 <---- rest of the system applications
so i think we have the two cpu firmware/boot loaders in the .bin files, these bin files are just fat32 images, to access in ubuntu use
Code:
mount -o loop mdm.bin /mnt/mdmmountlocation
My guess is my first approach is getting the right PBL/SBL into the system and getting some feed back via uart, i have the jtag pinouts and further reserach says there is a UART2 on the jtag header, so when soldering up my jtag adapter i will include all pins if i can and sniff for serial logic, i happen to have a Open source logic sniffer, great tool as i do a lot of hacking into serial devices like scales and till printers .
back to topic.
When i do get to the jtag part at a minimum i should have access to the modem radio, afaik jtag devices connect in chains and most of the IC's that have jtag on the phones board all should link to the master device (i am thinking it's the modem cpu, no application) and that the Two cpu's share the eMMC memory some how, or it could be one cpu loads it into the other (it is connected via jtag down the chain) .
hopefully someone could correct me there.
Most of this is theory and my guess work, correct me if you find a mistake. most of the research is only over a few days too so i am far from finished there, does not help that most of the users speak a language that google translate just does not have a flair for.
Most of the info seems to suggest the modem cpu is the first inline so i decided to look further into the files there, notice the mdm.bin file is 23Mb, thats large, when mounted i notice the is a folder called 'image' ( amms.bin has folder called IMAGE , note the case difference, dont yet know whay)
in image folder we have :-
Code:
1.3M Sep 30 13:07 AMSS.MBN
35K Sep 30 13:07 DBL.MBN
2.2M Sep 30 13:07 DSP1.MBN
19M Sep 30 13:07 DSP2.MBN
40 Sep 30 13:07 EFS1.MBN
40 Sep 30 13:07 EFS2.MBN
40 Sep 30 13:07 EFS3.MBN
295K Sep 30 13:07 OSBL.MBN
Ah, i see amss.mbm , that must be the boot loader for the application cpu, DBL.MBM seems to be the PBL , OSBL.MBM could be the SBL
then there is the DSP/EFS files, I did do the command strings on all the files,
DBL.MBM does not have any text in the file that points to being able to do UART on boot, all text seems internal like pointers and references to the original build files e.g
Code:
D:\Q1LGT_MDM\MDM9600\modem_proc\core\boot\secboot2\dbl\target\mdm9x00\src\dbl_ddr.c
9x00B-SCAQSVZM-31613102
D:\Q1LGT_MDM\MDM9600\modem_proc\core\boot\secboot2\dbl\target\mdm9x00\src\dbl_sahara.c
but it also does contain data like this
Code:
auth_image
@[email protected]
@configure_hw
@flash_init
l0:eek:SBL
load_osbl_img
@DBL, Start
hw_init
so it looks more likley that dbl is first in the chain, it refers to loading osbl and configure hardware, i wonder if it means USB/UART at this stage or setting up ram and other GPIO's
in OSBL.MBM we have more interesting text
Code:
MbP?
Unable to attached to ChipInfo DAL
SAMSUNG
TOSHIBA
Flash: Failed to do initialization for probe!
ONFIx
0:ALL
Flash: Multi 2X page read not supported!
Flash: Multi 2X page write not supported!
boot_qdsps
OSBL
hw_init
hw_init_secondary
OSBL, Start
create_vector_table
ram_init
retrieve_shared
clobber_add_protection
mmu_flush_cache
OSBL, End
OSBL, Delta
osbl_sahara_load_amss
osbl_sahara_load_dsp1
osbl_sahara_load_dsp2
osbl_sahara_load_ramfs1
osbl_sahara_load_ramfs2
osbl_sahara_load_ramfs3
smem_boot_init
so it is looking more and more like DBL then SBL which then loads all of the other parts , also if you notice EFS1/2/3 are all tiny 40byte files, now i see why, they are loaded as ram-drives, so i assume those file set out the basic EFS file system in the ram.
again from research the boot stages are often counted as 3, i am assuming the real first part is in rom of the cpu (is this what triggers the qualcom download mode ) that loads DBL from eMMC and chain loads SBL
Now looking around the riff forums i see the list the info in a different way
Code:
Partition 0
SBL1
SBL2
Partition 1
RPM
SBL3
eMMC APPSBoot
TZ
.PIT
Click to expand...
Click to collapse
TZ i think is Trusted Zone
RPM - Power manager ?
now how this translates to file name from full flash and to mmcblk0p1 partitions i have yet to find out, i still dont have a .PIT file from a 32gb model
More updates to come,
regards
DarkSpr1te
CPU Boot order updates
So my digging has taken me back round to some of me early searching which i forgot about , hardware level seems to support the qualcom usb mode, but it can be disabled by manufacturer, so even if you find a resistor to the BOOT_CONFIG GPIO and ground it , it still may not work, and you could toast your board. once the qfuse is gone for that track, the maker can now use the gpio for anything else, it no longer controls the iROM branch choice ( CPU:do i start usb first or last?), it my thinking that on the first board sent out by the designers for a final production run ( those first public devices) they keep the option open to print off DEV models by changing the resistors/value of while the hardware stays same, not to be confused with dev board, that is pin/track simlar but is used to design the software mainly, sometimes hardware debug but as you change the hardware between the dev platform and production this is less helpful, google new.intrinsyc.com and apq8060, they produce a dev board that is the same as the device we hold, but everything is broken out for testing so don't expect to see this left in a bar for you to e-bay.
EDIT:
Above I refer to a dev phone and dev board, these are SURF and FFA, FFA is form factor accurate and SURF is Subscriber Unit Reference.
Here is the link, http://forum.xda-developers.com/showthread.php?t=1856327
Now from what i see, it's the same(edit:simlar) X0M pin setup as other phones, ground the right pin, reverse boot order, but this maybe two pins in the snapdragon,
[copied from other link]
Simplified table:
Code:
------------------------------------------------------------------
BC[5:0] Mapping
------------------------------------------------------------------
0b00000 Emergency Boot from SDC3 (SD) followed by USB-HS
0b00001 SDC3 followed by SDC1 (eMMC)
0b00010 SDC3 followed by SDC2 (if used)
0b00011 SDC1 (eMMC)
Click to expand...
Click to collapse
So if 0b00000 is EM boot and the docs say the the two gpio's that control this (if qfuse not blown) are taken high then it's 0b00011, so grounding those two resistors should give us 0b00000 or EM boot, the cpu docs also say they are internally grounded, the schematic says the voltage goes throught a 10k resistor, so grounding that side of the resistor that 'goes' to the cpu should change the boot order, but before trying this out, remember if you get the live side of the resistor the is no resistor between your probe and ground, that full current, short, blown, no more johnny 5.
Have you managed to unbrick the E160L?
darkspr1te said:
So my digging has taken me back round to some of me early searching which i forgot about , hardware level seems to support the qualcom usb mode, but it can be disabled by manufacturer, so even if you find a resistor to the BOOT_CONFIG GPIO and ground it , it still may not work, and you could toast your board. once the qfuse is gone for that track, the maker can now use the gpio for anything else, it no longer controls the iROM branch choice ( CPU:do i start usb first or last?), it my thinking that on the first board sent out by the designers for a final production run ( those first public devices) they keep the option open to print off DEV models by changing the resistors/value of while the hardware stays same, not to be confused with dev board, that is pin/track simlar but is used to design the software mainly, sometimes hardware debug but as you change the hardware between the dev platform and production this is less helpful, google new.intrinsyc.com and apq8060, they produce a dev board that is the same as the device we hold, but everything is broken out for testing so don't expect to see this left in a bar for you to e-bay.
Here is the link, http://forum.xda-developers.com/showthread.php?t=1856327
Now from what i see, it's the same(edit:simlar) X0M pin setup as other phones, ground the right pin, reverse boot order, but this maybe two pins in the snapdragon,
[copied from other link]
Simplified table:
Code:
------------------------------------------------------------------
BC[5:0] Mapping
------------------------------------------------------------------
0b00000 Emergency Boot from SDC3 (SD) followed by USB-HS
0b00001 SDC3 followed by SDC1 (eMMC)
0b00010 SDC3 followed by SDC2 (if used)
0b00011 SDC1 (eMMC)
So if 0b00000 is EM boot and the docs say the the two gpio's that control this (if qfuse not blown) are taken high then it's 0b00011, so grounding those two resistors should give us 0b00000 or EM boot, the cpu docs also say they are internally grounded, the schematic says the voltage goes throught a 10k resistor, so grounding that side of the resistor that 'goes' to the cpu should change the boot order, but before trying this out, remember if you get the live side of the resistor the is no resistor between your probe and ground, that full current, short, blown, no more johnny 5.
Click to expand...
Click to collapse
I think my E160L got a real brick today after I tried to flash a modified Rom downloaded from a Chinese forum. It can not be powered on after rebooting (installed successfully). I desperately need advice now on how to deal with it.
Jeff_GTA said:
I think my E160L got a real brick today after I tried to flash a modified Rom downloaded from a Chinese forum. It can not be powered on after rebooting (installed successfully). I desperately need advice now on how to deal with it.
Click to expand...
Click to collapse
Do you have any backups like nandroid ? does the 3 button boot still work ?
Regards
Have you looked into using ort-jtag. It's only about $150 (USD).
I've been looking into this myself for low-level debugging/bootloader development on SGH-T959V and SGH-I717.
All three of these devices are supported by ort-jtag and have header connectors for the jtag pins.
So I'm also getting some of these from digi-key, and making a small receptacle, much like in AdamOutler's captivate bootloader development thread. (search for k-ww)
Again, ort-jtag does support the SHV-E160L. (search that link for SHV-E160L)
PBL Dump - I think
So ive been doing some tests.
I think i managed to dump the PBL
i dumped memory and a strings search return this
Code:
pbl_error_handler.c
pbl_flash_nand.c
pbl_flash.c
dload.c
pbl_flash_nand.c
pbl_flash_onenand.c
pbl_auth\secboot_rsa_math.c
pbl_error_handler.c
pbl_auth.c
pbl_auth.c
pbl_auth.c
pbl_auth.c
pbl_auth.c
pbl_mc.c
pbl_mc.c
pbl_error_handler.c
and
Code:
qhsusb\src\dci\qhsusb_dci.c
}^PBL_DloadVER1.0
!8}^
}]^}^
Q`omm
z8}]
DEBUG
SW_ID
OEM_ID
pbl_flash_onfi.c
pbl_flash_nand.c
pbl_flash_sflashc.c
pbl_loader.c
pbl_flash_sdcc.c
pbl_auth.c
pbl_auth\secboot.c
pbl_auth\secboot_x509.c
QUALCOMM COPYRIGHT 2009BOOT ROM VERSION: 1.4QHSUSB VERSION: 00.00.08
BOOT ROM AUTHOR: DHAVAL PATEL
07 0000 SHA1
does any one want the dump that can reverse it ?
Dumps & execute address
I also need the help of other SHV-E160? owners, i need dumps from working phones, i managed to create a 8660_msimage.mbn and flashed it, but i was using i717 bootloaders and i dont think they will work, i need working dumps from working phones, starting with partition table layout, sbl1.mbn and sbl2.mbn
Does anyone know if the is is correct
SBL1 exec address 0x2A000000
SBL2 exec address 0x2E000000
as i can upload the sbl to 0x2a000000 but not the sbl2 to 0x2e000000
i can also upload the tz.mbn to 0x2a020000
i am trying to use sec boot 3 based call stack but am unsure of the real exec values
Ive seen in another post these values
"
It looks like ours deviates slightly from this.
If the headers are to be believed,
TZ is loaded at 0x2A000000
SBL3 is loaded at 0x8FF00000
APPSBL/aboot is loaded at 0x88E00000
"
the post is
http://forum.xda-developers.com/showpost.php?p=30057296&postcount=243
it does explain why i cant load into 0x2e000000
Progress
So today i made real progress, I have been able to flash a basic program to allow me to access the EMMC, i have taken a full backup and now i need to start scanning the dump for need information,
I still need help from other users so please if you are will to provide me dumps of your working device that would help me a great deal
So Part One is a sucess, I have been able to flash my own code and power on the galaxy note. next step is rebuilding the emmc partition tables, testdisk can find the partitions but is not alowing me to write a non standard partition table (which emmc seems to be formatted with)
Thanks
darkspr1te
help QPST Software Download
Hi,
I'm stuck with the same problem can you tell me what image you use to the phone. I stuck here. I' m really don't know what to do?
Thank you for your help.
tyllerdurdent said:
Hi,
I'm stuck with the same problem can you tell me what image you use to the phone. I stuck here. I' m really don't know what to do?
Thank you for your help.
Click to expand...
Click to collapse
First thing i must say is dont flash your phone just yet!! walking blindly into this could render your phone useless due to certain data being lost for good.
if you still wish to continue i will upload a basic guide and files. My method is still in development, it has many bugs ( i flashed the phone with i717 roms, working, SHV-E120 roms, working, N7000 rom complete fail)
But first some questions,
Which model phone is it?
what happened to get you to the point of needing the flash ? ( i ask so i can trace why the bricks are happening and hopefully fix it)
thank you for your help, I will be waiting your method and your files.
Thank you so much for your help.
My phone is a samsung galaxy note SHV-E160L korean version.
what happen was:
I tried to upgrade the firmware with kies and suddenly the program crash. My phone enter in an error issue with the firmware and said use emergency recovery mode.
I tried the recovery several times (uninstalling kies and install it again but that never work).
So, I download odin and this files to restore the original firmware:
CSC - GT-N7000-MULTI-CSC-OZSLPF.tar.md5
Phone - MODEM_N7000XXLR1_REV_05_CL1144476.tar.md5
Bootloader- N7000_APBOOT_N7000ZSLPF_CL558430_REV02_user_low_sh ip.tar.md5
PDA - N7000_CODE_N7000ZSLPF_CL558430_REV02_user_low_ship .tar.md5
Pit for 16GB - Q1_20110914_16GB.pit
I connect my phone and try to install the firmware again, but odin fail and my samsung became a nice brick.
The phone currently does not turn on, the phone is in download mode and I install QPST and the program recognize the system in download mode.
I want to try your method because other information I collected said that I have to send it to guarantee.
Can I install i717 rom in the E160L?
I will be waiting for your post because sincerely I don't know how to repair it.
Thank you so much.
Hello darkspr1te
First of all, nice work there (though I didn't understood most of the things there, but seems there is some good work going on on our SHV-E160's
On your comment;
( i flashed the phone with i717 roms, working, SHV-E120 roms, working, N7000 rom complete fail)
Does that mean that i717 roms can work on the SHV-E160 devices? Please share if that is the case.
The geeky bits
tyllerdurdent said:
Thank you so much for your help.
My phone is a samsung galaxy note SHV-E160L korean version.
what happen was:
I tried to upgrade the firmware with kies and suddenly the program crash. My phone enter in an error issue with the firmware and said use emergency recovery mode.
I tried the recovery several times (uninstalling kies and install it again but that never work).
So, I download odin and this files to restore the original firmware:
CSC - GT-N7000-MULTI-CSC-OZSLPF.tar.md5
Phone - MODEM_N7000XXLR1_REV_05_CL1144476.tar.md5
Bootloader- N7000_APBOOT_N7000ZSLPF_CL558430_REV02_user_low_sh ip.tar.md5
PDA - N7000_CODE_N7000ZSLPF_CL558430_REV02_user_low_ship .tar.md5
Pit for 16GB - Q1_20110914_16GB.pit
I connect my phone and try to install the firmware again, but odin fail and my samsung became a nice brick.
The phone currently does not turn on, the phone is in download mode and I install QPST and the program recognize the system in download mode.
I want to try your method because other information I collected said that I have to send it to guarantee.
Can I install i717 rom in the E160L?
I will be waiting for your post because sincerely I don't know how to repair it.
Thank you so much.
Click to expand...
Click to collapse
Ok, as i said it's still a work in progress at the moment.
I used the i717 bootloaders (thats why we have a brick as it's not getting to the aboot loader or little kernel as some other refer to it) and E160 modem and application cpu as my first target is getting odin mode back.
I was able to also use the E120 bootloaders (screen was messed up though )
I've just got home from a very long shift so i will do a full and clear write up ( STILL a work in progress ) tomorrow (20th)
but i will explain the basic now as you do need to download large files before we continue.
First you need to download the same firmware as you were originally on before the brick, The reason is because between versions i suspect there is minor changes in partition tables (that why the n7000 roms brick )
If you dont have the latest QPST (2.7.3xx or higher ) please google for it now, there are many sites that offer it. (links will folllow tomorrow)
also down load :-
ABOOT_SGH-I717M_I717MUGLA2_user_CL875155_REV00.tar (or tar.md5 )
i717-GB-Modem.tar (or .md5)
now my initital work was based off a chinese link for the A820L
http://blog.csdn.net/su_ky/article/details/7773273
To save you the time of many hours of translation and cross reference here is the quick run down
When the phone is in QDLoad mode its because the PBL (Stored in ROM , read only memory) could not start SBL1 or SBL2 , it stores the error in IRAM location 0x3FF18 and then goes to QDLoad fail mode. At this point it has tried uart, sd-card before hand and those failed too.
IRAM is the small built in memory of the MSM8660 CPU, it has not initiated the main SYSTEM ram yet so our memory space ro running code is 87k and 256k (refer to document 8960_boot_architecture.pdf found the unlock bootloaders section.
Now because our partition table and or our bootloaders are damaged (or we have emmc brick bug) we have to rewrite that data again to revive our bricks.
This is where it gets hard, and where my warnings now come into play.
right now you must think of the EMMC chip (its the name for the internal SD-CARD we boot from and store our normal data, imei and all the other data of the system, it is just a sc-card with better security for our purpose)
This emmc chip holds all of you settings for phone function and we must not loose that,
But...
we have to write data to the chip to boot again, I am not fully aware of all the memory locations so this is assumptions on my part.
we are going to write a basic bootloader that turns the whole phone into a sd-card, then write new bootloaders
using QPST we upload 8660_msimage.mbn (its a out of the box emmc factory image) this file is ment for setting up of dev versions of the phone, it made up of the following parts
sector 0 partition table or (partition0.bin AFTER patching with info from patch0.xml) I do not have a real copy of the original of this, it can be pulled from a working SVH-E160x using the code at the end.
after the MBR (which is the first part of the partiton make up, EBR follows, we can have 3 primary partitions and the fourth is a extended which is just another partiton table pointing to the next EBR and so on, upto 29 parititons i think)
anyway, after the MBR is SBL1, which chainloads SBL2 then that side loads RPM, gets a go signal then loads SBL3, when SBL3 is done most of the device hardware has been mapped into the cpu's memory table, SDRAM is now ready for larger code,
aboot now loads
some of the above loading functions occur at the same time and some wait on go signals from other code in other CPU's and some fail due corruption and or security check fails( JTAG users can watch the memory as it changes and halt, change data and continue which is why JTAGers's have more power , we dont have loader outputting data yet so no feed back, hence the brick)
when aboot is loaded we now have access to odin, so thats the goal, get aboot loaded for now who cares about the rest of the funtions.
we do need to care about those function later so thats why we will backup the entire system, i dont know if this will really work when restored and bring back all of our settings, thats later,
So onto the writing and possibly overwriting of important information, WARNING, i dont know yet if we are overwriting imei or simalr data yet so proceed at your own risk.
We will get the required from factory (qualcomm test or dev board not samsung factory in the box for consumer) from the MUI phone firmware
http://bigota.d.miui.com/QDN43/Mioneplus_QDN43_fastboot_Android_4.0_d3d83nmdk2.zip
from this zip we want 8660_msimage.mbn, patch0.xml, partition0.bin MPRG8660.hex ( this file is uploaded first, its a serial bootloader that is loaded at 0x2a000000 (start of PBL IRAM space 256k in size) and that setups a emmc to command access (we use revskill to upload the same file and dump memory , sadly ive not found a way of pulling the entire emmc to a backup, if we can figure that out we can pull the entire boot chain, fix it and send it back with what ever versions we desire, for now revskill is used to read the PBL error so we can at least see why we cant boot, not quite jtag but best we got ))
so now we have a phone running a basic bit of code that allows us to use code sent to serial port to write (possibly read) the emmc
we then use QPST to write the 8660_msimage.mbn as a one to one copy to the very start of the emmc , reboot phone and then when the phone restarts, it sets up the ram, some hardware (charging system, you will now notice your phone gets warmer that before when plugged in) and gives us direct access to the emmc as if it was a sd-card
at this point you could move the phone to any pc and it's just a sd-card branded qualcomm
BUT at this point the pc or any other computer you connect it too only see's the partition table contained in the 8660_msimage.mbn file , you other data is there so i advise the next step you MUST do.
connect the phone to a linux computer (use a live cd or live usb if you are not a normal linux user)
you will then run the following command
Code:
dd if=/dev/sd? of=/mount/location/shv-e160-full-emmc.bin bs=512
? is the letter of the drive , use dmesg and look for sdb or sdc , if you dont understand this part then i would suggest waiting for a possible script/one click solution. right now i am still booting only 1 in 20 boots and do not yet know why the boots fail and why some work.
of=/mount... this is where you will place the entire 16GB (32GB for 32gb models ) which should be a one to one copy of the system
the bs=512 is very important, it's block size, again, if you dont understand then maybe wait.
Thats enough for now, i am going to spend a hour or two working on some theories i came up with today.
user with working phones, please google how to backup parts of your phone, this may happen to you so it's best to backup asap !!!
from the blog.csd site a script to grab the partition table data, if a working usr could please run this and post the file, it does not contain user data only the partiton table and a direct 1 to 1 restore for any phone, i think it possible to write that direct back to a QDLoad mode phone, re write the bootloaders from linux and bingo working phone. i dont have backups as it's not my phone, it belongs to a client who knows i like to tinker with electronics.
anyway, once i have the partition file i can overlay it on my test phone (which i can activate QSLoad at any time, hence it's unbrick-able dev mode)
once the partition file is written to my phone, i can build a script to backup your important data, write known working bootloaders, and reboot the phone into a usable device.
here is the script in python (user linux live cd with a copy of adb, just google adb linux pack, there is a windows and linux allin one pack)
or you can get the original from the link above, i've not tested this as i dont have a device in adb mode but i've read through it and it looks sound but never tested by me.
Well i hope that enlightens you, am sorry i dont have a all in one solution for you, it's still a dev project and most of the information i have has only been collected over the past week, i only discovered it's QSDload after getting a msm8660 schematic and i still dont know what i am trully shorting out to trigger the QSDload when ever i want, even when it's booted
If any one from the unbrickable project(s) want to get in touch to share info i would be happy, i am also sure this is a usable solution for HTC phones as well
oh and one last thing
i read only a hour ago (via cell phone while in a car so not 100%) that once the phone is in QSDload and stays in QSDload on every power cycle then we can write the partition table to a SD-CARD and it will boot that, i have not tested that yet, i will try and see if the 8660_msimage.mbn file written to a sd-card works
I also suspect that some of my good boots have been when i've mixed up the sdcard with system.img.ext4 etc on it with the one with just update.zip on it. it's one my list of things to check , any suggestions are welcome as to how i correctly format the card (heads,cylinders, block size etc)
ok folks, hope this helps
COPY TEXT BELOW ONLY INTO A FILE AND RUN WITH PYTHON (linux is easier, may be possible to use a vm box, i am but linux is my main os and windows is the vm)
Code:
import os
from struct import *
def mbr():
global offset, partitions
os.popen("adb shell su -c 'dd if=/dev/block/mmcblk0 of=/cache/partition0.bin bs=512 count=1'").close()
os.popen("adb shell su -c 'cp /cache/partition0.bin /sdcard/partition0.bin'").close()
os.popen("adb pull /sdcard/partition0.bin .").close()
f = open("partition0.bin", 'rb')
data = f.read()
f.close()
partitions = [ ]
n=0
while True:
buf = data[446+(16*n):446+(16*(n+1))]
partition = dict(zip(('boot', 'id', 'start', 'size'), unpack('4I', buf)))
partition['type'] = "MBR"
n += 1
partition['no'] = n
partitions.append(partition)
if partition['id'] == 5:
offset = partition['start']
break
def ebr():
global offset, partitions
n = 0
while True:
a = 0
os.popen("adb shell su -c 'dd if=/dev/block/mmcblk0 of=/cache/ebr bs=512 count=1 skip=" + str(offset+n) + "\'").close()
n += 1
os.popen("adb shell su -c 'dd if=/cache/ebr of=/cache/partition0.bin bs=512 count=1 seek=" + str(n) + "'").close()
os.popen("adb shell su -c 'cp /cache/ebr /sdcard/partition0.bin'").close()
os.popen("adb pull /sdcard/partition0.bin .").close()
f = open("partition0.bin", 'rb')
data = f.read()
f.close()
while True:
buf = data[446+16*a:446+16*(a+1)]
partition = dict(zip(('boot', 'id', 'start', 'size'), unpack('4I', buf)))
if partition['id'] == 5:
break
if partition['id'] == 0:
return
partition['type'] = "EBR"
partition['no'] = n
partition['start'] += n-1+offset
partitions.append(partition)
a += 1
if __name__ == "__main__":
mbr()
ebr()
os.popen("adb shell su -c 'cp /cache/partition0.bin /sdcard/partition0.bin'").close()
os.popen("adb pull /sdcard/partition0.bin .").close()
for part in partitions:
print "%s %2i, Boot: 0x%02X, Id: 0x%02X, Start: 0x%08X (%8i), Size: 0x%08X (%8i, %8i KB)" % (part['type'], part['no'], part['boot'],part['id'], part['start'], part['start'], part['size'], part['size'], part['size']/2)
Click to expand...
Click to collapse
beginning
thank you for your help,
I currently have the qpst version 2.7 build 373. You think is enough of download the same version of Chinese post QPST.2.7.374.rar
I will begin to download the other files required and I will be commenting my progress.
Thank you so much for your help, i really appreciate that you share you r knowledge.
Requests
While i try some theories if othe users could possibly provide me with :-
Original partition table via script above and also via adb
use
adb and run
Code:
cat /proc/partitions > /sdcard/partitions.txt
fdisk -l /dev/block/mmcblk0 > /sdcard/fdisklist.txt
mount > /sdcard/mountlist.txt
Then on the pc side using ADB again do the following
Code:
adb pull /sdcard/partitions.txt
adb pull /sdcard/fdisklist.txt
adb pull /sdcard/mountlist.txt
and post those files.
there are many posts on it so wont repeat but later will add a link.
along with some spell checks :laugh:
if you can dump the boot loaders from a original e160x too as my data started currupt.
i also need to talk to someone who can assist me in writing a program to take the pit file and turn it into this
Code:
<?xml version="1.0" ?>
<data>
<!--NOTE: Sector size is 512bytes-->
<program file_sector_offset="0" filename="" label="SMD_HDR" num_partition_sectors="65536" physical_partition_number="0" size_in_KB="32768.0" start_sector="1"/>
<program file_sector_offset="0" filename="sbl1.mbn" label="SBL1" num_partition_sectors="1000" physical_partition_number="0" size_in_KB="500.0" start_sector="65537"/>
<program file_sector_offset="0" filename="sbl2.mbn" label="SBL2" num_partition_sectors="3000" physical_partition_number="0" size_in_KB="1500.0" start_sector="66537"/>
<program file_sector_offset="0" filename="rpm.mbn" label="RPM" num_partition_sectors="1000" physical_partition_number="0" size_in_KB="500.0" start_sector="69559"/>
<program file_sector_offset="0" filename="sbl3.mbn" label="SBL3" num_partition_sectors="4096" physical_partition_number="0" size_in_KB="2048.0" start_sector="70559"/>
<program file_sector_offset="0" filename="aboot.mbn" label="ABOOT" num_partition_sectors="5000" physical_partition_number="0" size_in_KB="2500.0" start_sector="74655"/>
<program file_sector_offset="0" filename="" label="BOOT" num_partition_sectors="20480" physical_partition_number="0" size_in_KB="10240.0" start_sector="79655"/>
<program file_sector_offset="0" filename="tz.mbn" label="TZ" num_partition_sectors="1000" physical_partition_number="0" size_in_KB="500.0" start_sector="100135"/>
<program file_sector_offset="0" filename="partition0.bin" label="MBR" num_partition_sectors="1" physical_partition_number="0" size_in_KB="0.5" start_sector="0"/>
<program file_sector_offset="1" filename="partition0.bin" label="EXT" num_partition_sectors="22" physical_partition_number="0" size_in_KB="11.0" start_sector="69537"/>
</data>
Click to expand...
Click to collapse
*edit
the partiton0.bin provided below is 8.5kb (.5kb MBR, 8kb EBR) and in raw_program0.xml bove it say 0.5kb and 11kb, making that file 11.5kb, i dont know if the A810 has larger or smaller EBR than us, it could be they pulled extra, in my reading of the dumps i've seen lots of padded 0's after files (between sbl2/ebr/rpm) anyway if you just copy paste it will throw a error, ive got it set at 0.5 and 8.
EDIT:- Do not use this file, ive uploaded newer files later on.
some of the questions i need to answer are :-
1. what is the first partition, it's dos, around 105mb and labled smd_hdr and is filled with smd_hdr.bin (or mbn)
2. what are the real sector locations of the files, above you will see the rawpartiton0.xml file, this tells QPST where in the emmc to put the data num_partiton_sectors does match data from the pit files, but i dont know the real offsets yet, (samsung or htc could put the rest of the partiton table in cpu qfuse data areas and not write it to the emmc to confuse us and write the real files to another location and use the pit file as a base+offset calculation)
start_sector is the real location on the emmc, where it starts writing the file.
at the end is partiton locations(its a generic file containing the first few byes of default partition table, patch0.xml then updates this data), i dont have our device specific figures yet, i also dont fully understand patch0.xml and the difference in figures used.
if we have a backup of each of the different version of android partitons we could just write that in replacement of partiton0.bin and we dont need patch0.xml, this file sole job to alter the generic files, oem's have the choice of changing this data.
Code:
<?xml version="1.0" ?>
<patches>
<!--NOTE: This is an ** Autogenerated file **-->
<!--NOTE: Patching is in little endian format, i.e. 0xAABBCCDD will look like DD CC BB AA in the file or on disk-->
<!--NOTE: This file is used by Trace32 - So make sure to add decimals, i.e. 0x10-10=0, *but* 0x10-10.=6.-->
<patch byte_offset="506" filename="partition0.bin" physical_partition_number="0" size_in_bytes="4" start_sector="0" value="NUM_DISK_SECTORS-208801." what="Update MBR with the length of the EXT Partition."/>
<patch byte_offset="506" filename="DISK" physical_partition_number="0" size_in_bytes="4" start_sector="0" value="NUM_DISK_SECTORS-208801." what="Update MBR with the length of the EXT Partition."/>
<patch byte_offset="458" filename="partition0.bin" physical_partition_number="0" size_in_bytes="4" start_sector="16" value="NUM_DISK_SECTORS-1695744." what="Update final partition with actual size."/>
<patch byte_offset="458" filename="DISK" physical_partition_number="0" size_in_bytes="4" start_sector="208816" value="NUM_DISK_SECTORS-1695744." what="Update final partition with actual size."/>
</patches>
Click to expand...
Click to collapse
please note that it's two lines of the same code except one is partition0.bin and the other is DISK,
Do we need both? i know if i dont add the partiton0 section used in raw_program.xml then the drive is blank in linux,
now it's my understanding that the ebr comes as the forth partiton and it point to the next one , above in patch0.xml it start at NUM_DISK_SECTORS-1695744
i am still trying to better understand these figures,
Well time to grab coffee, i guess it's a dev night in.
the file MPRG8660.HEX can be renamed EMMCBLD.HEX and it triggers QPST to always look for a QDLoad mode phone and not debug, you can place all the files you need in one folder, i advise you to keep the originals in one location and only extract what your need to your worrking folder, copy emmcswdowload.exe from the QPST folder there too, we might need to do command line work, ive read that you can pre-create images in emmcswdownload (the same way 8660_msimage.mbn was created ) that you could just drop onto a phone once it's in emmc sd-card mode, almost a one click.
More info, plus help offered
Your welcome tyllerdurdent,
I am going to be putting a few hours into the dev from now actually for if you want assistance then no problems,
I also advise the following, download ubuntu live cd, it has a lot of tools your going to need to extract data you require, if we go step by step we might be good, i did a lot of test writing before i got my first boot, and that again only happens one in 20, i dont know why.
the rawpartiton0.xml above is incorrect for our devices as it states the first partion is 32mb, (i think it's ment to be amss.mbn, or NON-HLOS.mbn , our pit file which i did extract from my emmc dump says it's 105mb. i am confused and to why rawpartiton0.xml says the first bootloader is at start_sector="65537" but fdisk shows it as start 204801, i think someone needs to show me how to convert from blocks to sectors,
in patch0.xml it says
Code:
<patch byte_offset="506" filename="partition0.bin" physical_partition_number="0" size_in_bytes="4" start_sector="0" value="NUM_DISK_SECTORS-208801." what="Update MBR with the length of the EXT Partition."/>
Click to expand...
Click to collapse
208801 is where we have our ebr start,
i also think the IROM based pbl, sbl etc use the partition types in some way, why else have so many types? can any one explain that
this is a fdisk view of what i think our partition table looks like
Code:
Device Boot Start End Blocks Id System
/dev/sdb1 1 204800 102400 c W95 FAT32 (LBA)
/dev/sdb2 * 204801 205800 500 4d QNX4.x
/dev/sdb3 205801 208800 1500 51 OnTrack DM6 Aux1
/dev/sdb4 208801 208801 0 5 Extended
/dev/sdb5 212992 213991 500 47 Unknown
/dev/sdb6 221184 225279 2048 45 Unknown
/dev/sdb7 229376 234375 2500 4c Unknown
/dev/sdb8 237568 258047 10240 48 Unknown
/dev/sdb9 262144 263143 500 46 Unknown
/dev/sdb10 270336 271335 500 5d Unknown
/dev/sdb11 278528 279527 500 91 Unknown
/dev/sdb12 286720 307199 10240 93 Amoeba
/dev/sdb13 311296 511999 100352 c W95 FAT32 (LBA)
/dev/sdb14 516096 522239 3072 4a Unknown
/dev/sdb15 524288 530431 3072 4b Unknown
/dev/sdb16 532480 538623 3072 58 Unknown
/dev/sdb17 540672 741375 100352 8f Unknown
/dev/sdb18 745472 751615 3072 59 Unknown
/dev/sdb19 753664 759807 3072 5a Unknown
/dev/sdb20 761856 29843455 14540800 5b Unknown
/dev/sdb21 770048 790527 10240 ab Darwin boot
/dev/sdb22 794624 815103 10240 60 Unknown
/dev/sdb23 819200 839679 10240 94 Amoeba BBT
/dev/sdb24 843776 3911679 1533952 a5 FreeBSD
/dev/sdb25 3915776 8114175 2099200 a6 OpenBSD
/dev/sdb26 8118272 8736767 309248 a8 Darwin UFS
/dev/sdb27 8740864 9005055 132096 a9 NetBSD
/dev/sdb28 9011200 10035199 512000 95 Unknown
/dev/sdb29 10035200 30777343 10371072 90 Unknown
Oh, download wxdhex or wimlar program, you going to need a hex editor that can load BIG files , 16gb worth
i717-GB-Modem.zip IS THE SAME AS TAR?
i717-GB-Modem.zip 21.35 MB 7 0 2012-06-30 08:45:11
I could not find the i717-gb as tar file but I find it as a zip file. but I'm not sure about thif the contents are correct. Could you check
http://d-h.st/1aP
i717-GB-Modem.zip contents
META-INF
COM
GOOGLE
ANDROID
update-binary
updater-script
TMP
amss.bin
mdm.bin
Blocks and sectors
This may explain it , the different figure in the xml files
Because sectors are logical on the drive (Logical Block Addressing = LBA) you need to convert between LBA and physical (file system) sectors. This is pretty easy to do:
First - get a table of the start and end sectors of the partition table:
Code:
[[email protected] ~]# fdisk -lu /dev/hda
Disk /dev/hda: 120.0 GB, 120034123776 bytes
255 heads, 63 sectors/track, 14593 cylinders, total 234441648 sectors
Units = sectors of 1 * 512 = 512 bytes
Device Boot Start End Blocks Id System
/dev/hda1 * 63 208844 104391 83 Linux
/dev/hda2 208845 4401809 2096482+ 83 Linux
/dev/hda3 4401810 8482319 2040255 82 Linux swap
/dev/hda4 8482320 234436544 112977112+ 5 Extended
/dev/hda5 8482383 29447144 10482381 83 Linux
/dev/hda6 29447208 50411969 10482381 83 Linux
/dev/hda7 50412033 52516484 1052226 83 Linux
/dev/hda8 52516548 234436544 90959998+ 83 Linux
Use this to determine what partition the bad sector is in. In this case 232962120 is inside the start and end values for /dev/hda5
NOTE: This is in partition 5 - ignore partition 4 as it is the extended partition. Any block from partitions 5 through 8 will also be in partition 4, but you want the real partition, not the extended partition.
Next, calculate the file system block using the formula:
b = (int)((L-S)*512/B)
where:
b = File System block number B = File system block size in bytes (almost always is 4096) L = LBA of bad sector S = Starting sector of partition as shown by fdisk -lu and (int) denotes the integer part.
For example:
The reported sector from the smart log above is 232962120, thus:
((14858312 - 8482383) * 512) / 4096 = 796991.125
^Bad Sec. ^Start Sec. ^Cha Ching! This is the sector!
(Use the block number from the smart test section, not from the smart error log section. They are using different methods of reporting file system vs. physical blocks.)
((BadBLock - StartPartition) * 512) / 4096
You can just paste this into Google as a template
Any fraction left indicates the problem sector is in the mid or latter part of the block (which contains a number of sectors). Ignore the fraction and just use the integer.
Next, use debugfs to locate the inode and then file associated with that sector:
Click to expand...
Click to collapse
[[email protected]]# debugfs
debugfs 1.35 (28-Feb-2004)
debugfs: open /dev/hda5
debugfs: icheck 796991
Block Inode number
796991 <block not found>
debugfs: quit
Ah! It didn't give the inode! It if did, you could have found the file with:
[[email protected]]# debugfs
debugfs 1.35 (28-Feb-2004)
debugfs: open /dev/hda5
debugfs: icheck 796991
Block Inode number
796991 41032
debugfs: ncheck 41032
Inode Pathname
41032 /S1/R/H/714197568-714203359/H-R-714202192-16.gwf
So what the heck? Why no inode? Well, remember how it said the sector might be bad?
Click to expand...
Click to collapse
the above copied from
http://timelordz.com/wiki/SMART_Rewriting_Bad_Sectors
i have a feeling we may need to shift our files (the basic files need to start odin are listed in rawpatch0 above, i dont know if that 100% true but it was the only files i wrote on by first sucess)
also
http://forum.xda-developers.com/showthread.php?p=31843525&postcount=13
in the above link they talk about the header of the qualcomm file
+------------+
|Dbl-preamble|
+------------+
|Dbl-header |
+------------+
|Dbl.bin |
+------------+
Click to expand...
Click to collapse
and
data_ptr = autodetectpage;
*data_ptr = sbl_header.codeword;
data_ptr++;
*data_ptr = sbl_header.magic;
data_ptr++;
*data_ptr = AUTODETECT_PAGE_SIZE_MAGIC_NUM;
Click to expand...
Click to collapse
now i used this in a way to find my bootloaders (i717 by this time, not shve-160l )
and to find the partitons
you will see in a hex editor at the start of each boot loader
something else to think about, my lack of success that last two days to produce a boot could be because my partitons are not clean , thats is to say if i write my sbl1 to 1000, and the trailing 0000 of the partition definition of my 99 block ebr/mbr ends at 999 , if i have dirt data between 999 and 1000 the cpu/pbl my interpret that as code(some of my boots is brick, some are into QDLoad, i have no pattern yet) , something i must test or confirm, or just worry about.
tyllerdurdent said:
i717-GB-Modem.zip 21.35 MB 7 0 2012-06-30 08:45:11
I could not find the i717-gb as tar file but I find it as a zip file. but I'm not sure about thif the contents are correct. Could you check
http://d-h.st/1aP
i717-GB-Modem.zip contents
META-INF
COM
GOOGLE
ANDROID
update-binary
updater-script
TMP
amss.bin
mdm.bin
Click to expand...
Click to collapse
Yes thats correct
updater script btw contains text, binary is the flashing exe i think,
Code:
run_program("/sbin/dd", "if=/tmp/mdm.bin", "of=/dev/block/mmcblk0p17");
run_program("/sbin/dd", "if=/tmp/amss.bin", "of=/dev/block/mmcblk0p13");
Click to expand...
Click to collapse
and a google of a simlar sansung product the skyrocket gives me a simlar pit layout
Device Name Size Part Name ODIN tar file Mount Point
mmcblk0boot0 512KB (empty) n/a (empty partition)
mmcblk0boot1 512KB (empty) n/a (empty partition)
mmcblk0p1 100MB SMD_HDR (partition info)
mmcblk0p2 500KB SBL1 sbl1.mbn
mmcblk0p3 1500KB SBL2 sbl2.mbn
mmcblk0p4 1KB (unnamed partition with '55 AA' MBR signature)
mmcblk0p5 500KB RPM rpm.mbn
mmcblk0p6 2MB SBL3 sbl3.mbn
mmcblk0p7 2500KB ABOOT aboot.mbn
mmcblk0p8 10MB BOOT boot.img
mmcblk0p9 500KB TZ tz.mbn
mmcblk0p10 500KB SSD n/a (empty partition)
mmcblk0p11 500KB PIT celox.pit
mmcblk0p12 10MB PARAM param.lfs
mmcblk0p13 98MB MODEM amss.bin /system/etc/firmware/misc
mmcblk0p14 3MB MSM_ST1 efs.img
mmcblk0p15 3MB MSM_ST2 n/a
mmcblk0p16 3MB MSM_FSG n/a
mmcblk0p17 98MB MDM mdm.bin /system/etc/firmware/misc_mdm
mmcblk0p18 3MB M9K_EFS1 efsclear1.bin
mmcblk0p19 3MB M9K_EFS2 efsclear2.bin
mmcblk0p20 3MB M9K_FSG n/a
mmcblk0p21 10MB DEVENC enc.img.ext4 /efs
mmcblk0p22 10MB RECOVERY recovery.img
mmcblk0p23 3MB FOTA n/a
mmcblk0p24 598MB SYSTEM system.img.ext4 /system
mmcblk0p25 2GB USERDATA userdata.img.ext4 /data
mmcblk0p26 302MB CACHE cache.img.ext4 /cache
mmcblk0p27 129MB TOMBSTONES tomb.img.ext4 /tombstones
mmcblk0p28 11.2GB UMS ums.rfs /mnt/sdcard
Click to expand...
Click to collapse
Other files
contents of the i717 boot loaders i used
ABOOT_SGH-I717M_I717MUGLA2_user_CL875155_REV00
Code:
527K Jan 6 2012 aboot.mbn
115K Jan 6 2012 rpm.mbn
72K Jan 6 2012 sbl1.mbn
111K Jan 6 2012 sbl2.mbn
601K Jan 6 2012 sbl3.mbn
117K Jan 6 2012 tz.mbn
other files pulled from
ABOOT_SGH-I717M_I717MUGLA2_user_CL875155_REV00 (no bootloader but all the other system files )
I am looking to compile a new kernel for this device. I have pulled off the boot.img file from /dev/block/nandc and used split_bootimg.pl to pull apart the kernel and the boot ramdisk. I have used extract-ikconfig to get the kernel config and built my new 3.4 kernel (with modules). So far so good...
I have tried to build the new boot.img file with no avail. My concern is with the original kernel as the new one should be build the same (I think). It looks like it is not compressed as I run the command
file boot.img-kernel
It returns
boot.img-kernel: data
So it is not a Linux kernel ARM boot executable zImage
Is this correct? Can Android boot with this or is the split_bootimg.pl tool not creating the file correctly (it does pull the ramdisk as I have uncompressed it fine! and the file does contain the kernel config)
Ok So assuming I can create a proper boot.img with my built zImage and the old ramdisk, what do I need to do with the modules? Do I update the boot.img on the tablet and copy the modules over to the existing /system/lib/modules? (this appears to not work)
I did not find an exact guide on android kernel and the modules....
Domenic
link to the old (working) boot.img hxxps://docs.google.com/file/d/0B0c1jpaziDZ-NU5Sd18wbUVjN1E/edit?usp=sharing
some progress
domenictroilo said:
I am looking to compile a new kernel for this device. I have pulled off the boot.img file from /dev/block/nandc and used split_bootimg.pl to pull apart the kernel and the boot ramdisk. I have used extract-ikconfig to get the kernel config and built my new 3.4 kernel (with modules). So far so good...
I have tried to build the new boot.img file with no avail. My concern is with the original kernel as the new one should be build the same (I think). It looks like it is not compressed as I run the command
file boot.img-kernel
It returns
boot.img-kernel: data
So it is not a Linux kernel ARM boot executable zImage
Is this correct? Can Android boot with this or is the split_bootimg.pl tool not creating the file correctly (it does pull the ramdisk as I have uncompressed it fine! and the file does contain the kernel config)
Ok So assuming I can create a proper boot.img with my built zImage and the old ramdisk, what do I need to do with the modules? Do I update the boot.img on the tablet and copy the modules over to the existing /system/lib/modules? (this appears to not work)
I did not find an exact guide on android kernel and the modules....
Domenic
link to the old (working) boot.img hxxps://docs.google.com/file/d/0B0c1jpaziDZ-NU5Sd18wbUVjN1E/edit?usp=sharing
Click to expand...
Click to collapse
Ok so the mkimage command can be used to pull the uImage out of the Kernel Part of the file like this:
mkimage -A arm -O linux -T kernel -C none -a 0x40008000 -e 0x40008000 -n "Linux 3.0.8" -d uImage boot.img-kernel2
The only question is where in the boot process can I find the "Load Address" and "Entry Point" values that are required to create my uImage. I would think that these are going to be needed as options when I build my new kernel, and I will need to change the boot process or add a boot.scr file with my options. I have mounted the /dev/block/nanda partition (uBoot partition) but am not certain on where I would find the current environment values.
I did find this little snippet on the net that referes to the base address for this device xxx.tabletroms.com/forums/all-other-android-tablets-general-discussion/5791-full-specs-whatever-else-polaroid-pmid10c-tablet-2.htmlx
mkbootimg --kernel boot.img-kernel --ramdisk boot.img-ramdisk.cpio.gz --cmdline 'console=ttyS0,115200 rw init=/init loglevel=8' --board sun4i --base 0x40000000 -o boot.img
But he never discusses where he found the -base address. Is this the same as the Load Address and Entry Point?
Domenic
domenictroilo said:
Ok so the mkimage command can be used to pull the uImage out of the Kernel Part of the file like this:
mkimage -A arm -O linux -T kernel -C none -a 0x40008000 -e 0x40008000 -n "Linux 3.0.8" -d uImage boot.img-kernel2
The only question is where in the boot process can I find the "Load Address" and "Entry Point" values that are required to create my uImage. I would think that these are going to be needed as options when I build my new kernel, and I will need to change the boot process or add a boot.scr file with my options. I have mounted the /dev/block/nanda partition (uBoot partition) but am not certain on where I would find the current environment values.
I did find this little snippet on the net that referes to the base address for this device xxx.tabletroms.com/forums/all-other-android-tablets-general-discussion/5791-full-specs-whatever-else-polaroid-pmid10c-tablet-2.htmlx
mkbootimg --kernel boot.img-kernel --ramdisk boot.img-ramdisk.cpio.gz --cmdline 'console=ttyS0,115200 rw init=/init loglevel=8' --board sun4i --base 0x40000000 -o boot.img
But he never discusses where he found the -base address. Is this the same as the Load Address and Entry Point?
Domenic
Click to expand...
Click to collapse
So the base address is found in script.bin (converted to text file with bin2fex tool). This is the information found in it:
[dram_para]
dram_baseaddr = 0x40000000
Still how do I find the "Load Address" and "Entry Point" for the current kernel...
unmkbootimg
domenictroilo said:
So the base address is found in script.bin (converted to text file with bin2fex tool). This is the information found in it:
[dram_para]
dram_baseaddr = 0x40000000
Still how do I find the "Load Address" and "Entry Point" for the current kernel...
Click to expand...
Click to collapse
Well more reading leads me to use unmkbootimg but on my linux 32bit system all I get is:
# strace ./unmkbootimg
execve("./unmkbootimg", ["./unmkbootimg"], [/* 20 vars */]) = 0
--- SIGILL (Illegal instruction) @ 0 (0) ---
+++ killed by SIGILL +++
Illegal instruction
Too bad I can see the souce code to find out what this program is doing or compile it for myself.....
split_bootimg.pl should give you the parameters i listed on tabletroms.
What source are you using? Polaroid never coughed any up and i last harassed them in February.
bigsupersquid said:
split_bootimg.pl should give you the parameters i listed on tabletroms.
What source are you using? Polaroid never coughed any up and i last harassed them in February.
Click to expand...
Click to collapse
Here is the split:
letourneau:~/tmp# ../split_bootimg.pl /root/boot.img
Page size: 2048 (0x00000800)
Kernel size: 8112836 (0x007bcac4)
Ramdisk size: 210343 (0x000335a7)
Second size: 0 (0x00000000)
Board name:
Command line: console=ttyS0,115200 rw init=/init loglevel=8
Writing boot.img-kernel ... complete.
Writing boot.img-ramdisk.gz ... complete.
I am using the linux--sunxi kernel for now with an additional patch for dmard06.ko (which I do not expect to work!) I emailed them for the kernel source but just received a form email saying they will get back to me..... oh well I will as again.
Just playing around for now to see if a newer kernel will tweek up the speed on this device as it is very laggy.
Domenic
I may be using an older version of split_bootimg.pl?
On another note, I was able to execute unmkbootimg on a different linux server. I was attempting to run it on a PIII 800 so it needs something > a pentium (I guess).
./unmkbootimg boot.img
unmkbootimg version 1.2 - Mikael Q Kuisma <[email protected]>
Kernel size 8112836
Kernel address 0x40008000
Ramdisk size 210343
Ramdisk address 0x41000000
Secondary size 0
Secondary address 0x40f00000
Kernel tags address 0x40000100
Flash page size 2048
Board name is ""
Command line "console=ttyS0,115200 rw init=/init loglevel=8"
This image is built using standard mkbootimg
Extracting kernel to file zImage ...
Extracting root filesystem to file initramfs.cpio.gz ...
All done.
---------------
To recompile this image, use:
mkbootimg --kernel zImage --ramdisk initramfs.cpio.gz --base 0x40000000 --cmdline 'console=ttyS0,115200 rw init=/init loglevel=8' -o new_boot.img
---------------
Domenic
think the parameters you're searching for are specific for building the new kernel without proper source code to extract them from?
maybe you could get them from dmesg while the tablet is running?
else think there might be a way to grep the variable names in the existing kernel binary or headers?
i also would like a fresh kernel.
maybe useful stuff here:
https://vilimpoc.org/research/sh3dev/kernel-hacking.html
http://forum.xda-developers.com/showpost.php?p=46668556&postcount=31
thanks jjchico
i haven't checked this out yet, but may be useful.
Well I have build a new kernel and copied the modules over, but I does not boot. It appears that it never mounts the ramdisk as it never gets to the second splash screen, but I am flying blind as there is no serial port to debug with. I suspect that I do not quite understand the boot process for this tablet.... here is what is on the nanda vfat partition: (good information here http://rhombus-tech.net/allwinner_a10/a10_boot_process/)
since this is in my linux.ini
cat linux.ini
[segment]
img_name = c:\linux\u-boot.bin
img_size = 0x80000
img_base = 0x4A000000
[script_info]
script_base = 0x43000000
script_size = 0x10000
[logo_info]
logo_name = c:\linux\linux.bmp
logo_show = 1
Not quite certain how to setup img_base on my compile I guess.
I have tried the linix.ini from the link above but the tablet booted to livesuite image... that is a fail.
Not being a kernel developer.... I am wondering if the option to setup SUNXI_NAND_COMPAT_DEV is creating /dev/block/nandX or /dev/nandX as this will definatly impact the init.sun4i.rc file as it is mounting /dev/block/nandX This could be my first issue.... but then does /init first display the initlogo.rle or start processing init.sun4i.rc. Well if the first thing it does in load initlogo.rle then I am not even getting this far....below is the description of SUNXI_NAND_COMPAT_DEV
Symbol: SUNXI_NAND_COMPAT_DEV [=y] │
│ Type : boolean │
│ Prompt: Create old nand device names (nanda-nandz) │
│ Defined at drivers/block/Kconfig:564 │
│ Depends on: BLK_DEV [=y] && SUNXI_NAND [=y] │
│ Location: │
│ -> Device Drivers │
│ -> Block devices (BLK_DEV [=y]) │
│ -> SUNXI Nandflash Driver (SUNXI_NAND [=y]) │
well a little more reading leads me to this page on the CarbonRom http://www.slatedroid.com/topic/64378-rom-carbonrom-422/ turns out that if you want to run a 3.4 kernel you MUST update u-boot (http://www.mediafire...4ro667xcfkd2dv7).... so I have flashed the CarbonRom and it does boot with the u-boot update. Not to worry the new u-boot will still boot the old kernel. The only issue I have with the CarbonRom kernel is the missing dmard06 driver so no touch screen..... I may get a new kernel working yet!
domenictroilo said:
well a little more reading leads me to this page on the CarbonRom http://www.slatedroid.com/topic/64378-rom-carbonrom-422/ turns out that if you want to run a 3.4 kernel you MUST update u-boot (http://www.mediafire...4ro667xcfkd2dv7).... so I have flashed the CarbonRom and it does boot with the u-boot update. Not to worry the new u-boot will still boot the old kernel. The only issue I have with the CarbonRom kernel is the missing dmard06 driver so no touch screen..... I may get a new kernel working yet!
Click to expand...
Click to collapse
https://github.com/fsebentley/linux_allwinner_drivers_input_touchscreen_mods
https://groups.google.com/forum/#!topic/linux-sunxi/D0n8dyaVA34
?
bigsupersquid said:
https://github.com/fsebentley/linux_allwinner_drivers_input_touchscreen_mods
https://groups.google.com/forum/#!topic/linux-sunxi/D0n8dyaVA34
?
Click to expand...
Click to collapse
the patch compiles,you need to comment out #include <linux/earlysuspend.h> as it is not supported on the newer kernel and not required by the driver anyway. I just need to get the kernel to actually load lol..... I have at least gotten past the boot screen now, but it hangs.... I will try a different arm tool chain and see what happens next.....
So close but yet so far..... well I have the new kernel installed and booted with adb shell available but it turns out the new mma7660.ko is not compatable with my device I get this error from dmesg over and over and over....
<4>[ 3254.934675] incomplete xfer (0x20)
<3>[ 3254.938093] Assertion failed! drivers/hwmon/mma7660.c,196,mma7660_read_xyz,result>=0
Will rename the driver and see if I can get past init further in the boot....
Edit
Well more issues with drivers the dmard06 is not compatible either, I had more success with the Carbon Kernel I will see if I can find the source git, should be around....
dmesg | busybox grep dmard
<4>[ 5.408528] dmard06: init
<3>[ 5.447835] dmard06gsensor_fetch_sysconfig_para: after: gsensor_twi_addr is 0x4c, dirty_addr_buf: 0x4c. dirty_addr_buf[1]: 0xfffe
<4>[ 5.488546] dmard06_init: after fetch_sysconfig_para: normal_i2c: 0x4c. normal_i2c[1]: 0xfffe
<6>[ 5.508612] gsensor_detect: Detected chip dmard06 at adapter 1, address 0x4c
<4>[ 5.601533] dmard06 gsensor I2C err = 0!
<4>[ 5.613430] dmard06 probe failed
<4>[ 5.624926] dmard06: probe of 1-004c failed with error -1
This all makes some sense as Polaroid has not get gotten me the kernel and the drivers they load (their kernel have very unique names:
Module Size Used by Not tainted
8192cu 575431 0
rtl8150 8910 0
mcs7830 6557 0
qf9700 8513 0
asix 22500 0
usbnet 23745 3 mcs7830,qf9700,asix
dmard06_gl 8202 2
mma7660 6541 0
mali 130177 14
ump 42178 19 mali
gt811_ts_828 19947 0
sun4i_csi0 27946 0
nt99250 14254 1
gc0308 12848 1
videobuf_dma_contig 5659 1 sun4i_csi0
videobuf_core 18067 2 sun4i_csi0,videobuf_dma_contig
Well more issues with drivers the dmard06 is not compatible either, I had more success with the Carbon Kernel I will see if I can find the kernel source git, should be around....
dmesg | busybox grep dmard
<4>[ 5.408528] dmard06: init
<3>[ 5.447835] dmard06gsensor_fetch_sysconfig_para: after: gsensor_twi_addr is 0x4c, dirty_addr_buf: 0x4c. dirty_addr_buf[1]: 0xfffe
<4>[ 5.488546] dmard06_init: after fetch_sysconfig_para: normal_i2c: 0x4c. normal_i2c[1]: 0xfffe
<6>[ 5.508612] gsensor_detect: Detected chip dmard06 at adapter 1, address 0x4c
<4>[ 5.601533] dmard06 gsensor I2C err = 0!
<4>[ 5.613430] dmard06 probe failed
<4>[ 5.624926] dmard06: probe of 1-004c failed with error -1
This all makes some sense as Polaroid has not provided me the kernel they used and the drivers they load have unique names see below from the original stock rom:
Module Size Used by Not tainted
8192cu 575431 0
rtl8150 8910 0
mcs7830 6557 0
qf9700 8513 0
asix 22500 0
usbnet 23745 3 mcs7830,qf9700,asix
dmard06_gl 8202 2
mma7660 6541 0
mali 130177 14
ump 42178 19 mali
gt811_ts_828 19947 0
sun4i_csi0 27946 0
nt99250 14254 1
gc0308 12848 1
videobuf_dma_contig 5659 1 sun4i_csi0
videobuf_core 18067 2 sun4i_csi0,videobuf_dma_contig
I suspect after all the trouble I had with LG source that there are device-specific bits in the drivers.
Polaroid will most likely have to cough up source to get it all working.
Have you considered trying punchmod to use the old drivers? I dunno if that'd work considering the stock kernel is a lower version... but it might be worth a try.
bigsupersquid said:
I suspect after all the trouble I had with LG source that there are device-specific bits in the drivers.
Polaroid will most likely have to cough up source to get it all working.
Have you considered trying punchmod to use the old drivers? I dunno if that'd work considering the stock kernel is a lower version... but it might be worth a try.
Click to expand...
Click to collapse
I have changed the vermagic on the old modules and tried with the newer kernel but no go.... the latest 3.0 sunxi kernel is at 3.0.96 so I will build this one and see if I can get the special 3.0.8 polaroid modules to load.
I have a ticket open with Polaroid lol I guess we will see if the cough up a kernel or not.... I would not put money on it...
domenictroilo said:
I have changed the vermagic on the old modules and tried with the newer kernel but no go.... the latest 3.0 sunxi kernel is at 3.0.96 so I will build this one and see if I can get the special 3.0.8 polaroid modules to load.
I have a ticket open with Polaroid lol I guess we will see if the cough up a kernel or not.... I would not put money on it...
Click to expand...
Click to collapse
I've had an open ticket since November from talking on the phone after email did nothing.
nothing but autoresponder.
bigsupersquid said:
I've had an open ticket since November from talking on the phone after email did nothing.
nothing but autoresponder.
Click to expand...
Click to collapse
Do you believe that something like this http://gpl-violations.org/faq/violation-faq.html could force them to provide the source code? I could try to get the group at "FreakTab" to make the same requests for the kernel sources. Maybe with enough requests the gpl-violations group could take it on?
(K I am getting off topic here sorry anyone out there......)
I have attempted to build the 3.0.96 kernel but it did not boot.... back to hacking at the .config to see what I have missed (I need to move my build environment off this old p3 800 lol)
History:
This was a virgin stock rom'd kindle that just didn't boot one day at the end of last November. The battery was showing signs of degradation, i.e. false/flaky battery life, and I think random restarts! My guess is that it was most likely updating when something went wrong.
I cannot be sure what firmware it was on or upgrading to, but at the same time that it happened, around the 22/10/2017, I could only manually download update-kindle-14.4.5.5_user_455004320.bin from amazon. But my daughters identical KFAPWI took the 4.5.5.3 update on the 17/10/2017.
So I'm going to take a stab, and say it most likely on 4.5.5.2.:fingers-crossed:
At one point I had the grey boot logo only, but after pressing too many combinations, getting drunk and falling asleep, it now does nothing:crying:. It does not connect to the pc at all, despite my persistence holding down the power button for far too long, and even using a factory cable - have got drivers previously installed from rooting/buggering/fixing another kindle, both standard and qbulk. Using a DVM meter on the battery yields the correct voltage for both cells too...
I'm better with the down and dirty hardware side of things, so I have taken the less-than-direct direct emmc access root by Ivan_nik in his post her on XDA: https://forum.xda-developers.com/kindle-fire-hdx/general/direct-emmc-access-t3249534 and have success in reading the EMMC.
I now have a full EMMC image, along with individual partition dumps done with DD in Hiren's partition magic Linux, hooray! So in my mind I can do whatever I like, and I should be able to restore from any mess that I make.
But WTF to do now?? Reading through the many posts here, there's talk of flashing partition images from the stock firmware, .mbn and .img etc. Some of them 'seem' easily mapped to partitions:
rpm.mbn > rpm
sbl1.mbn > sbl1
sdi.mbn > dbi
tz.mbn > tz
emmc_appsboot.mbn > aboot??
from this folder within the unpacked firmware:
Directory of C:\Android\update-kindle-14.4.5.5_user_455004320\images
30/05/2014 22:26 358,404 emmc_appsboot.mbn
30/05/2014 22:26 155,880 rpm.mbn
30/05/2014 22:26 277,076 sbl1.mbn
30/05/2014 22:26 18,004 sdi.mbn
30/05/2014 22:26 322,744 tz.mbn
5 File(s) 1,132,108 bytes
But if you unpack the fwupdate.tar, there seem to be duplicates:
Directory of C:\Android\update-kindle-14.4.5.5_user_455004320\images\fwupdate
27/08/2013 03:16 70,944 appsboot.mbn
12/10/2013 22:25 3,428,888 dsp1.mbn
12/10/2013 22:25 34,903,448 dsp2.mbn
27/08/2013 03:16 464 partition.mbn
27/08/2013 02:04 104,360 rpm.mbn
27/08/2013 02:04 115,308 sbl1.mbn
27/08/2013 02:04 368,684 sbl2.mbn
7 File(s) 38,992,096 bytes
And which boot image to dd to which boot partition?:
Directory of C:\Android\update-kindle-14.4.5.5_user_455004320
30/05/2014 22:26 7,766,016 boot.img
1 File(s) 7,766,016 bytes
Directory of C:\Android\update-kindle-14.4.5.5_user_455004320\images\fwupdate
27/08/2013 03:16 5,591,040 boot-oe-msm9615.img
1 File(s) 5,591,040 bytes
I found this reference in the install-recovery.sh, and one of the references below match the boot.img above:
"#!/system/bin/sh
if ! applypatch -c EMMC:/dev/block/platform/msm_sdcc.1/by-name/recovery:9525248:ba9ceca5992d27e1a24d630f54991376082375f0; then
log -t recovery "Installing new recovery image"
applypatch -b /system/etc/recovery-resource.dat EMMC:/dev/block/platform/msm_sdcc.1/by-name/boot:7766016:5cd16d9155f1e148b867dfbbbde5e4ce0faee8c9 EMMC:/dev/block/platform/msm_sdcc.1/by-name/recovery ba9ceca5992d27e1a24d630f54991376082375f0 9525248 5cd16d9155f1e148b867dfbbbde5e4ce0faee8c9:/system/recovery-from-boot.p
else
log -t recovery "Recovery image already installed"
fi"
Linux is not my strong point, Me being a mainly M$ based IT tech, but I'm a quick learner - if only I can stop my eyes from going curly.
Any tips, pointers or even a bit of humour at this point would be greatly appreciated.
Thanks for reading.
Ok, I have confirmed that at least in part, it was 4.5.5.3. Comparing what partitions I could discover here on XDA to the ones found in the stock image.bin using HxD, some of the following are identical in all but file size, due to trailing zeros in the backed up img files:
01.modem.img same as 4.5.5.3 NON-HLOS.bin
02.sbl1.img same as 4.5.5.3 sbl1.mbn
03.dbi.img same as 4.5.5.3 sdi.mbn
04.DDR.img
05.misc.img empty - all zeros
06.aboot.img same as 4.5.5.3 emmc_appsboot.mbn
07.rpm.img same as 4.5.5.3 rpm.mbn
08.boot.img same as 4.5.5.3 boot.img
09.tz.img same as 4.5.5.3 tz.mbn
10.pad.img empty - all zeros
11.modemst1.img
12.modemst2.img
13.fsg.img empty - all zeros
14.fsc.img empty - all zeros
15.ssd.img empty - all zeros
16.dkernel.img
17.dfs.img
18.recovery.img same as 4.5.5.3 boot.img patched with recovery-resource.dat
19.persist.img
20.persistbackup.img
21.cache.img
22.system.img
23.userdata.img
Partitions 17, 19, 20, 21, 22, & 23 are all ext4, and mount easily in parted magic. Build.prop in system states that it is 4.5.5.3.
That leaves very little partitions with info on them that I don't know how to compare/check, 4, 11, 12, & 16.
And are partitions 5, 10, 13, 14, & 15 normally empty?
Hello! help please help with Amazon Kindle Fire hdx 8.9 OUR CITY DOES NOT INCLUDED AND THERE ARE NO MASTERS WHO COULD MAKE IT HAVE A FRIENDLY MASTER HE HAS A PROGRAMMER SAYS IF THERE WOULD BE a dump then he would help fix other things but he says he could if someone gave a direction! Thank you
Hello, I recently got a TicWatch Pro 2020 model (1 GB RAM, no LTE), and wanted to install Magisk. I tried booting TWRP for the original TicWatch Pro (catfish) as well as the LTE model (catshark) to no avail. I decided to try and port the recovery myself. So I cloned the repository for the original model and tried just changing the codename. I had never done it before so I had no idea that I first needed to get the kernel for my watch (which requires TWRP to pull). Eventually I found a TWRP image that worked on my watch and I wanted to see if I could manage to compile one myself. I pulled boot.img from my watch and this is what binwalk shows on it:
Code:
DECIMAL HEXADECIMAL DESCRIPTION
--------------------------------------------------------------------------------
0 0x0 Android bootimg, kernel size: 8469500 bytes, kernel addr: 0x80008000, ramdisk size: 1869123 bytes, ramdisk addr: 0x81000000, product name: ""
2048 0x800 Linux kernel ARM boot executable zImage (little-endian)
18675 0x48F3 gzip compressed data, maximum compression, from Unix, last modified: 1970-01-01 00:00:00 (null date)
7879104 0x7839C0 Flattened device tree, size: 197476 bytes, version: 17
8053880 0x7AE478 Unix path: /dev/block/platform/soc/7824900.sdhci/by-name/vendor
8076580 0x7B3D24 Flattened device tree, size: 197492 bytes, version: 17
8251372 0x7DE7EC Unix path: /dev/block/platform/soc/7824900.sdhci/by-name/vendor
8274072 0x7E4098 Flattened device tree, size: 197476 bytes, version: 17
8448848 0x80EB50 Unix path: /dev/block/platform/soc/7824900.sdhci/by-name/vendor
8472576 0x814800 gzip compressed data, maximum compression, from Unix, last modified: 1970-01-01 00:00:00 (null date)
I used dd to cut off the first and last entry, to (hopefully) give me the zImage-dtb that I needed. Amazingly, it worked! The recovery compiled and I was able to boot it. However it's not perfect and I would like to ask for help with two things:
1: When I run:
Code:
mka recoveryimage
the process finishes and the recovery.img is built, but make returns an error: recovery.img too large. Luckily it doesn't affect me as for now I'm just using fastboot boot and not flashing anything. But if I wanted to flash the image I guess I would need to find a way to reduce the image size. Also the other recovery that I found online is much smaller (~14 kB vs ~21 kB). What can I do to make the final image smaller?
2: When running fastboot boot recovery.img, the screen changes to a random colorful noise for about 10 seconds before showing the TWRP logo and booting successfully. What could cause that and how can I fix it?
Here is the repository with all of the files: https://github.com/sasodoma/android_device_mobvoi_catfish_ext