Announcement

Collapse
No announcement yet.

XFS Is Becoming Leaner While Btrfs & EXT4 Gain Weight

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

  • XFS Is Becoming Leaner While Btrfs & EXT4 Gain Weight

    Phoronix: XFS Is Becoming Leaner While Btrfs & EXT4 Gain Weight

    Red Hat's Eric Sandeen has written an interesting blog post concerning the size of popular Linux file-systems and their kernel modules. It turns out that the XFS file-system is losing lines of code, while maintaining the same feature-set and robustness, but the EXT4 and Btrfs file-systems continue to have a net increase in lines of code...

    http://www.phoronix.com/vr.php?view=OTU4OA

  • #2
    A quoto

    "Measuring software productivity by lines of code is like measuring progress on an airplane by how much it weighs."- Bill Gates

    May be we can find if there is any real improvement by a phoronix benchmark

    Comment


    • #3
      Originally posted by phoronix View Post
      It's not too surprising though that the Btrfs file-system size is on the rise as it's still under active development and new features are continuing to be added.
      That's the only sentence that is needed.

      Comment


      • #4
        It's also noteworthy that according to the graph on the original post ext4 is only approaching half of the LOCs of xfs...

        Comment


        • #5
          Lol

          Originally posted by ChrisXY View Post
          It's also noteworthy that according to the graph on the original post ext4 is only approaching half of the LOCs of xfs...
          That kinda is what actually matters....

          +1

          Comment


          • #6
            Originally posted by zeealpal View Post
            That kinda is what actually matters....

            +1
            Yeah, the fact that XFS code has been decreasing kind of proves the point. That it was complicated and not extremely well integrated with Linux, otherwise it shouldn't be possible to be consistently reducing the LOCs like that over an extended period of time. But at least it's getting better.

            Comment


            • #7
              XFS has always served me well. During the days EXT4 was still causing data corruption, XFS did its job.
              The only problem I had, was when I broke the hard drive's partition table. It took me a while to get a utility to recover the XFS partition. It worked, eventually.

              Comment


              • #8
                Given that XFS was originally ported directly from IRIX I wouldn't be surprised if there was a "compatability layer" in place to map any systemcalls that weren't present in Linux. Given that IRIX has been dead in water since 2006 the guys in SGI/XFS can over time reduce dependency on such a layer -- as they don't have to maintain codebase with IRIX XFS anymore.

                Comment


                • #9
                  It's not the first time forum is more informative than article.

                  Comment


                  • #10
                    Originally posted by kraftman View Post
                    It's not the first time forum is more informative than article.
                    I am not sure why you call it an article. This was submitted under "Latest Linux News".

                    Comment


                    • #11
                      Originally posted by ChrisXY View Post
                      It's also noteworthy that according to the graph on the original post ext4 is only approaching half of the LOCs of xfs...
                      Isn't there code that is shared across ext2/3/4? If yes, is that counted in?

                      Comment


                      • #12
                        Originally posted by not.sure View Post
                        Isn't there code that is shared across ext2/3/4? If yes, is that counted in?
                        IIRC, the common code is duplicated because people were wary of doing any kind of experimental development in the ext3 tree at that point. There is (or was last time I looked) a kernel config option to the effect of "use ext4 driver for ext2/3 filesystems", which doesn't make much sense unless the ext2/3 code is duplicated.

                        Comment


                        • #13
                          Originally posted by sabriah View Post
                          I am not sure why you call it an article. This was submitted under "Latest Linux News".
                          I have all the articles and news from Phoronix in the same place. I wouldn't expect something like that as a news.

                          Comment

                          Working...
                          X