| Versions |
  |
|
| Author |
 Topic  |
|
g_alin
USA
817 Posts |
 Posted - 04 déc. 2010 : 00:05:14
|
GelullaBarr,
I did :) now you can start DiscRW and download the SMFlash (in case you need it later ... :) ). |
 |
|
|
ablbd
342 Posts |
Posted - 04 déc. 2010 : 08:44:11
|
Let me explain the iboot issue.
Mio developpers had included an emergency "back-door" in every Moov bootloader (I don't know if C devices have it!!) which, if present on SD, will be executed first. The original iboot.exe is a very powerfull tool which has a similar effect as the "blue screen" and is able to replace not only the OS but nboot and uboot too, so be very carefull what files you put on SD together with iboot.exe!! It has no menu at all and will rewrite without any questioning winceimg, nboot and/or uboot if present. Obviously, in order to work, uboot must be functional.
Consequently, in all cases when the device stops responding and hangs in the boot screen, OS can be replaced by adding aside winceimg's the original iboot.exe or, as GellulaBarr explained, the file name "iboot.exe" can be used to run TC and correct something (an xml or ini which made the system to hang). |
Edited by - ablbd on 04 déc. 2010 08:57:56 |
|
|
|
GelullaBarr
25 Posts |
Posted - 04 déc. 2010 : 10:10:02
|
DiscRW it's not working on wince6, maybe another program.
Hardware Type is SEEKER2_600_EEU for those who bought from east european...I think it's important to know. This is another way to unlock this device, puting mioautorun in this folder and rename it auto.exe, for example
ablbd, iboot was just a way to regain control over mio S600. it was the only way open software. I used CECMD.exe(renamed iBoot.exe), but it's working very well mio pocket 4 R62
|
Edited by - GelullaBarr on 04 déc. 2010 11:43:53 |
|
|
|
ventzel
Romania
11 Posts |
Posted - 04 déc. 2010 : 18:37:30
|
quote:
Originally posted by ablbd
A Moov330/N179 is still laying on my desk as I couldn't find nboot/uboot for it. After some digging and testing I found a program (s3c2443_nr.exe, http://www.megaupload.com/?d=S5B4KQLP ) which is able to copy physical nand blocks. Should be executed from SD card and because bootloaders are not longer than 1,5MB, it can be stopped at 1/5 of the total size by simply reseting the device. If everything is OK, before pressing the "save" button you'll see the full nand ID of K9LAG08(EC D5 55 25 68). On SD you'll find 2 .hex files: pages.hex and spare.hex; it would be great to have both but actually I need pages.hex.
I have same problem with a Moov 500R2(S3C2450+K9GAG08). I fing another device with diferent nand(S3C2450+K9G4G08) and make 2 backups, first with H-jtag and second with S3C2443_NR.exe. Both are the same. I try to upload it into bricked device but it won't start.
I think bootloader have some parameters how to handle flash geometry in first block?
http://www.4shared.com/file/LIIuGCTb/T510r2_v1_00_0004.html
Can you tell me more about patching H-jtag driver to gain more speed? |
|
|
|
ablbd
342 Posts |
Posted - 04 déc. 2010 : 19:31:56
|
Well, can't tell you but can point out some issues:
a)The two nands are different in an essential way: K9G4G08 has 2K page size, K9GAG08 has 4K...
b) Have you loaded the image in one or you have separated nboot from uboot and loaded them separately?
c)How (using what tool??) did you loded the image into bricked device? Is this tool able to handle 4K page size, 128 pages/block?
d) have you choosen SLC mode for loading the image (h-jtag and s3c2443_nr are copying the pages as SLC) or you programmed it/them in two-plain mode?
Patching h-jtag driver was not made for speed but to adapt it for a certain nand(page size, pages/block). Anyway, "patching" is history now, I am recompiling the modified driver. |
Edited by - ablbd on 04 déc. 2010 19:39:39 |
|
|
|
ventzel
Romania
11 Posts |
|
|
ablbd
342 Posts |
Posted - 04 déc. 2010 : 21:10:39
|
I looked into source code for k9gag08 because I knew that the so-called "prototype" drivers had some errors (and those who have modified them were careless enough to disregard them), such as:
-according Samsung specifications, bad block marking is placed in the LAST page's spare area; this means that if really exists a factory bad block, the driver will not "see" the mark because it doesn't look in the right place and will consider such a block as good.
-the driver has some hard-coded parameters in nand_mark_badblock and nand_is_badblock routines (2048!!):
*SysAddr32(NFADDR) = (2048>>0) & 0xFF;
*SysAddr32(NFADDR) = (2048>>8) & 0xFF;
instead of
*SysAddr32(NFADDR) = (NAND_PAGE_SIZE>>0) & 0xFF;
*SysAddr32(NFADDR) = (NAND_PAGE_SIZE>>8) & 0xFF;
Just think at consequences: k9gag08 has 4096 bytes page size, so spare area starts at 4096 and not at 2048 !!!
Those guys made a very beneffic modification in the driver by setting the valid read-write area to cover full nand, which I used since long time ago. The problem here is that this trick works only if the nand was locked by software bit... |
Edited by - ablbd on 04 déc. 2010 21:27:24 |
|
|
|
ventzel
Romania
11 Posts |
Posted - 04 déc. 2010 : 21:29:06
|
| Thanks ablbd! I didn't notice that. I think i'll replace k9gag08 with K9K8G08 witch 2k/page and i'll try to flash it. |
|
|
|
ablbd
342 Posts |
Posted - 05 déc. 2010 : 08:35:01
|
@ventzel: don't do that, won't work!
I'll make the modification into source code and I will recompile the driver. Another (and much safer) solution would be a modified SJF, if you have a little patience I will include into sjf2450 the k9gag08 handling module.
Edit: here is the recompiled driver + source codes
http://www.4shared.com/file/ZzNqq-RD/DRIVER-1808_S3C2450K9GAG08.html
BTW, I'm not absolutely sure that the nand init bits are OK but try to write the nand, may work. First trial should be to erase some blocks (or entire nand) followed by check blank for the same selection. If selected area is blank indeed than you can write anything on it... |
Edited by - ablbd on 05 déc. 2010 09:33:21 |
|
|
|
ventzel
Romania
11 Posts |
Posted - 05 déc. 2010 : 19:16:23
|
@ablbd
I know how device is booting and it won't work only if i have same flash or something similar that have same page size and number of pages/block like the one i backup from device that works.
You are right about bad block handling. Its in last page spare area of every block for K9GAG08. Thanks again!
I need to wait until i find a similar mio moov 500 with K9gag08 flash chip and make a backup.
Init bits are not ok because for writing or erasing i need to pull WP of chip up, but page size and addrcycle in NFCONF of S3c2450 are ok. |
|
|
|
ablbd
342 Posts |
|
|
ventzel
Romania
11 Posts |
Posted - 05 déc. 2010 : 22:23:51
|
| Can you change ID of cpu please? I have this error ERROR: No CPU is detected(ID=0x07926f0f). |
|
|
|
ablbd
342 Posts |
|
|
marciport
Hungary
15 Posts |
Posted - 06 déc. 2010 : 13:18:46
|
@abldb
Hello i red back the forum and i just want to ask u maybe u found the solution for the mio N179 device what i explaind ago... if u have solution can u explain it and help me out how to fix it? thank u |
|
|
|
white_flag
22 Posts |
Posted - 06 déc. 2010 : 13:44:48
|
Hello, I made also an update to my firmware, Mio s505 with the file.fimware from the first page.. everything went well, but now, I have an problem with the screen, it is not calibrated, in the center is ok but on the edge "the touch" is in offset (I can not access close button an so on ..)
what can I do? |
|
|
|
Topic |
|
|
|
| This page was generated in 1,97 seconds. |
 |
|
- 
- 
How to upgrade/downgrade the firm. on Mio (Moov+C)
