Werner Almesberger <[email protected]> writes: > Paul Fertser wrote: >> In this case, yes. With Qi in NAND that's not so, as far as i >> understand. Nevertheless, what will be if one block wears out in the >> kernel partition? The user won't be able to even mount his NAND rootfs >> after the next boot to save configuration/data and that is probably >> important for many users. > > You mean that the work-out block would change the offset calculation ? > That won't happen, because the work-out block gets marked only in the > BBT and Qi does't look there. (See my previous mail on the topic.)
Ok, i see it now. But then it's even more inconsistent and somewhat defeats the purpose of marking worn-out blocks ;) Moreover, i understand that the kernel marks worn-out blocks only in BBT, but does NOR u-boot mark worn-out blocks in both BBT and OOB? If yes, then a user risks to lose his rootfs by flashing the kernel or bootloader via DFU. >> That is, to be absolutely sure that the flashing will be done to the >> correct addresses, everytime before dfu-ing anything a user must issue >> a dynpart command in NOR u-boot? > > Naw, NOR u-boot takes care of that for you. See the script starting at > line 87 of http://svn.openmoko.org/trunk/src/host/devirginator/mknor Yes, i see. That's why i asked Andy whether NOR u-boot does dynpart on every boot or not. >> Basically, people tried to "nandwrite -pm /dev/mtd3 uImage-GTA02.bin" >> and nandwrite marked plenty of blocks bad. > > Hmm, did they flash_eraseall /dev/mtd3 first ? They didn't. > Actually, where do these flash instructions come from ? I thought we > used to have them right ... The wiki has only one place where nandwrite is mentioned, and that is http://wiki.openmoko.org/wiki/Flashing_the_Neo_FreeRunner#Alternative:_using_nandwrite . "flash_eraseall" is not mentioned at all. -- Be free, use free (http://www.gnu.org/philosophy/free-sw.html) software! mailto:[email protected]
