#8313 NORM 9.1.0: Can OFW save-nand be made to go faster?
Zarro Boogs per Child
bugtracker at laptop.org
Fri Sep 5 17:41:46 EDT 2008
#8313: Can OFW save-nand be made to go faster?
-----------------------------------+----------------------------------------
Reporter: thomaswamm | Owner: wmb at firmworks.com
Type: enhancement | Status: new
Priority: normal | Milestone: 9.1.0
Component: ofw - open firmware | Version: Development build as of this date
Resolution: | Keywords:
Next_action: communicate | Verified: 0
Blockedby: | Blocking:
-----------------------------------+----------------------------------------
Comment(by wmb at firmworks.com):
Hmmm, here's another idea. I wonder if the USB devices that exhibit slow
write performance have been reformatted with filesystem parameters that
don't work well with the underlying page size of the NAND FLASH chip
inside? NAND FLASH is tricky to handle well. USB FLASH sticks have an
internal microprocessor that translates between the conventional disk
block paradigm and the NAND FLASH chip characteristics. That translation
is more efficient if the filesystem cluster size is a good match for the
NAND page and erase block size.
Such devices are factory-formatted with a suitable filesystem layout.
It's common practice in the FOSS community to reformat hard disks using
Linux tools like FDISK and mkfs. That works well with hard disks, but is
dodgy for FLASH storage, since you don't know what optimizations the
internal microprocessor is doing based on assumptions of cluster sizes,
FAT table locations, etc.
My best results with USB keys have always been with devices where I left
the factory filesystem format intact.
--
Ticket URL: <http://dev.laptop.org/ticket/8313#comment:5>
One Laptop Per Child <http://laptop.org/>
OLPC bug tracking system
More information about the Bugs
mailing list