Talk:UBIFS initial experiments: Difference between revisions

From OLPC
Jump to navigation Jump to search
m (New page: Hi, I'd like to comment this: * UBI is taking an extremely long time (~50s) to attach to the MTD device on the XO I am currently using for this testing. UBI attach time does scale li...)
 
No edit summary
Line 1: Line 1:
Hi, I'd like to comment this:
Hi, I'd like to comment this:
* UBI is taking an extremely long time (~50s) to attach to the MTD device on the XO I am currently using for this testing. UBI attach time does scale linearly w.r.t flash size, however the 50s seems wrong. According to this, we should only take about 2 seconds. At attach time I see a warning "UBI warning: ubi_eba_init_scan: cannot reserve enough PEBs for bad PEB handling, reserved 74, need 79" b/c my system partition has 5 bad PEBs in it and this may be related. TODO: Test on another XO.
* UBI is taking an extremely long time (~50s) to attach to the MTD device on the XO I am currently using for this testing.
UBI attach time does scale linearly w.r.t flash size, however the 50s seems wrong. According to this, we should only
take about 2 seconds.


I'm sure what you do is you erase the partition, then attach it to UBI. In this case UBI first read your partition and makes sure it contains all 0xFF bytes, which means it is empty, then it starts formatting this partition - it erases all eraseblocks again and writes the Erase counter and VID headers. This takes long time. If you attach it for the second time, it'll be fast, because it won't format it.
I'm sure what you do is you erase the partition, then attach it to UBI. In this case UBI first read your partition and makes sure it contains all 0xFF bytes, which means it is empty, then it starts formatting this partition - it erases all eraseblocks again and writes the Erase counter and VID headers. This takes long time. If you attach it for the second time, it'll be fast, because it won't format it.

Revision as of 07:45, 13 October 2008

Hi, I'd like to comment this:

  *  UBI is taking an extremely long time (~50s) to attach to the MTD device on the XO I am currently using for this testing.
     UBI attach time does scale linearly w.r.t flash size, however the 50s seems wrong. According to this, we should only
     take about 2 seconds.

I'm sure what you do is you erase the partition, then attach it to UBI. In this case UBI first read your partition and makes sure it contains all 0xFF bytes, which means it is empty, then it starts formatting this partition - it erases all eraseblocks again and writes the Erase counter and VID headers. This takes long time. If you attach it for the second time, it'll be fast, because it won't format it. Please, glance here: http://www.linux-mtd.infradead.org/faq/ubi.html#L_format_mtd Please, also glance here: http://www.linux-mtd.infradead.org/faq/ubi.html#L_ubierase

Also, make sure you do not have UBI debugging enabled. It slows down things a lot. The same is true for UBIFS.

HTH, Artem Bityutskiy.