Announcement

Collapse
No announcement yet.

F2FS File-System Driver Preparing A Low-Memory Mode

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • F2FS File-System Driver Preparing A Low-Memory Mode

    Phoronix: F2FS File-System Driver Preparing A Low-Memory Mode

    Google engineers are working on the notion of "memory modes" for the Flash-Friendly File-System (F2FS) with the intent on introducing a "low memory" mode for storage devices that would alter its behavior. Presumably Google is working on this new F2FS feature for low-end Android devices...

    Phoronix, Linux Hardware Reviews, Linux hardware benchmarks, Linux server benchmarks, Linux benchmarking, Desktop Linux, Linux performance, Open Source graphics, Linux How To, Ubuntu benchmarks, Ubuntu hardware, Phoronix Test Suite

  • #2
    Man, before the non-volatile era, memory in handheld devices was scarce but the file systems back then were very efficient, with minimum block size being technically zero (which would not be possible today as flash memory works with sectors)...

    Comment


    • #3
      Originally posted by tildearrow View Post
      Man, before the non-volatile era, memory in handheld devices was scarce but the file systems back then were very efficient, with minimum block size being technically zero (which would not be possible today as flash memory works with sectors)...
      Intel/Micron 3D XPoint is non-volatile and byte-addressable so all the tricks needed for NAND aren't needed.
      Too bad the "good ones" were limited to enterprise customers, but even the consumer Optanes were quite nice for what they were.

      Comment


      • #4
        Originally posted by tildearrow View Post
        Man, before the non-volatile era, memory in handheld devices was scarce but the file systems back then were very efficient, with minimum block size being technically zero (which would not be possible today as flash memory works with sectors)...
        The lower common denominator being a block size is mainly a result of how many modern filesystem API's work, they are addressed by blocks. In a similar vein one of the reason why ZFS works the way it does is its both the filesystem + raid combined in one so it can bypass that treating everything as a block limitation.

        Comment

        Working...
        X