[Gta04-owner] MLO/boot woes?

Boudewijn wankelwankel at yahoo.com
Sun Oct 14 21:56:11 CEST 2012


Hi List,

Once every so often, my GTA04 has this while it won't boot using GTA-init. 

TL;DR: how sensitive/intelligent is MLO? Once MLO is on SD, will overwriting 
it result in another physical location? 

TS;TM (too short, tell more?):
I have the init system on SD card, so it will boot to NAND when there's no SD 
in the phone. Once it happens, I keep partitioning/formatting/writing/trying 
out SD cards (one 32GB, the other 8GB) till one of them "catches on" again. 
Once it does, I keep using that SD until it stops booting, and then repeat. 

It's doable, but I don't really feel in control and it's rather time consuming 
without giving too much of a result.

To my horror, I just found out that my current test SD has switched its 
head/sector count back to its defaults, rather than the values I set last 
night. My SD's seem to have difficulties keeping to the 63-sector-rule anyway, 
giving me a first suspect of the cause.

Current SD config:
Disk /dev/sdb: 31.9 GB, 31914983424 bytes
64 heads, 32 sectors/track, 30436 cylinders, total 62333952 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *        2048      133120       65536+   c  W95 FAT32 (LBA)
/dev/sdb2          133121     1181697      524288+  82  Linux swap / Solaris
/dev/sdb3         1181698    62333951    30576127    5  Extended
/dev/sdb5         1183746     4194304     1505279+  83  Linux  (aurora)
/dev/sdb6         4196353     8390657     2097152+  83  Linux  (qtmoko)
/dev/sdb7         8392706    12587010     2097152+  83  Linux  (shr)
/dev/sdb8        12589059    16783363     2097152+  83  Linux  (hwval)
/dev/sdb9        16785412    62333951    22774270   83  Linux  (data)


My config last night, following 
https://code.google.com/p/beagleboard/wiki/LinuxBootDiskFormat:

Disk /dev/sdb: 31.9 GB, 31914983424 bytes
255 heads, 63 sectors/track, 3880 cylinders, total 62333952 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

   Device Boot      Start         End      Blocks   Id  System

The thing I noticed after partitioning, was that the earliest option for the 
start of a partition was at 2048. A month ago or so I noticed that as well, 
but that was after the booting stopped; I have no reason to suspect it being 
otherwise before that (that time I prepared the SD with the makesd-script).

The other card, in use till last Thursday (also prepared via makesd):
root at tp:/# fdisk -l /dev/sdb

Disk /dev/sdb: 8270 MB, 8270118912 bytes
255 heads, 62 sectors/track, 1021 cylinders, total 16152576 sectors
Units = sectors of 1 * 512 = 512 bytes
Sector size (logical/physical): 512 bytes / 512 bytes
I/O size (minimum/optimal): 512 bytes / 512 bytes
Disk identifier: 0x00000000

   Device Boot      Start         End      Blocks   Id  System
/dev/sdb1   *         128      131071       65472    c  W95 FAT32 (LBA)
/dev/sdb2          131072     4325375     2097152   83  Linux   (qtmoko)
/dev/sdb3         4325376     8519679     2097152   83  Linux   (shr)
/dev/sdb4         8519680    16152575     3816448    5  Extended
/dev/sdb5         8521728    12716031     2097152   83  Linux   (hwval)
/dev/sdb6        12718080    13766655      524288   82  Linux swap / Solaris
/dev/sdb7        13768704    16152575     1191936    b  W95 FAT32 (data)
root at tp:/# 

With the board being so picky about where to find MLO, I'm sure there is cause 
for problems in at least one of the partition schemas. The thing is, though: 
the boot menu shows up in any case. It is just that no interaction seems 
possible with it. 

I have fsck'd the partitions on both SDs. I have also tried to skip the boot 
menu by preparing it with echo "sd" > gta04-init/bootdev according to
http://lists.goldelico.com/pipermail/gta04-owner/2012-March/002048.html 

I actually get half a boot while setting init to a particular partition as per 
https://github.com/radekp/gta04-init (echo "/dev/mmcblk0p3 /" > gta04-
init/bootdev) does boot on the given partition, but boot does not always 
finish.

I have a couple of 'next steps' in mind, but I hope to have some advice on 
which seems most likely to bring me further or whether there's something else 
to try/check first:
- debug via serial port
- init in NAND?

I have close to no experience with serial ports in general, so now would 
probably be a good moment to start making a connector on the RS232 cable end. 
The manual seems quite clear on how to solder the connector (page 35/36 in 
GTA04a3, page 47/48 in the latest GTA04a4 manual). I'm not so clear on my 
soldering skills. How clean do the joints have to be to get enough of a signal 
for communication? I will probably see it doesn't work if its too bad, but 
since I used not to get any connection while playing with serial devices in 
the past, I'll never know whether it was because of my soldering skills or 
because of my serial skills.

As for init in NAND: is that possible/useful at all? 

Another thing comes to mind: could the init system have problems with 
extended/logical partitions?

Thanks in advance!

Best regards,

Boudewijn

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20121014/432b874d/attachment-0001.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 490 bytes
Desc: This is a digitally signed message part.
URL: <http://lists.goldelico.com/pipermail/gta04-owner/attachments/20121014/432b874d/attachment-0001.bin>


More information about the Gta04-owner mailing list