Quote:
Originally Posted by xms
My DX is US (b004) - yes, white DX, the very 1st DX produced.
It could be great if you do this images!
as i understand with somth like:
"dd if=/dev/mmcblk0 of=/mnt/us/mmcblk0_7688.dmp bs=512 count=7688"
"dd if=/dev/mmcblk0p1 of=/mnt/us/mmcblk0p1.dmp"
then as i undersatnd i need to
"3. Load MMC0 over USB storage"
it will show me 3 parttions sdx1=400mb, sdx2=24mb, sdx3=3Gb
and
"mkfs.ext3 /dev/sdX2"
"dd if=/path/to/mmcblk0p1.dmp of=/dev/sdX1 bs=1MiB"
and reboot
|
No.
Not if your DX is running Linux.
You have to pay attention to the device names and naming conventions, those are not arbitrary, they mean things.
sd<anything> Scsi Device - /dev/sd<whatever> is a device holding a file system.
mmc<whatever> - this is a storage device, almost always flash, that has a "high level", embedded, controller that accepts "disk drive" commands.
Disk drives do: "Read,Write,Seek, ..."
MTD - Memory Technology Devices (here: flash) do: "Read, Erase, Re-program, ..."
"Flash" :: eeprom (electrically erasable programmable read only memory)
Add an embedded device controller to a MTD storage part == a MMC storage part != file system
With that $0.05 tour of naming conventions, you should now be able to read the above "dd" command and explain how and why you are about to trash the system storage if you use it as written.
Quote:
Originally Posted by xms
maybe i need to format sdx1 also to verify bad blocks?
is it ext3?
|
No.
The controller of the underlying device manages the bad erase blocks (the underlaying device has "erase block" not "storage sectors" - the mmc controller is faking that).
It is using EEC to detect/correct and/or detect and retire the erase blocks in use at the storage media level.
The file system level should never see "bad blocks".
Quote:
Originally Posted by xms
p.s. 2knc1: i start to learn all this only after DX crash..
|
So how did you test what you are intending to post?