OpenBSD Journal

g2k16 Hackathon Report: Otto Moerbeek on back to malloc, and more

Contributed by pitrh on from the make me malloc dept.

Next in the g2k16 series is the Otto Moerbeek's report. Otto writes,

It has been a while since I was at a hackathon, due to various reasons the last hackathon I attended was in 2008 where I did commit my implementation of malloc. During the years, various improvements to my malloc were done, both by others and myself.
But now, 8 years later I came back to to commit a important change: make malloc more friendly to multi-threaded programs. Most of the actual work had been done during the months before the hackathon, but in Cambridge I did the last checks, got my OK's and committed! Of course this is not the last commit in this areas, the current approach is simple (but effective), it can be improved after it has been proven to be stable.

Apart from that I worked on some changes to newfs(8), disklabel(8) and fsck_ffs(8) to make the code to find alternative superblocks (which is used if the primary superblock is corrupted) work better. Testing that code takes quite some time, but it was good to see my manually corrupted filesystems come back to clean state.

A hackathon is not complete without pub and restaurant visits and some tourist activities, we went punting on the Cam. All in all, I had a very good time, it was really nice to meet other developers in person (again). In the futue, I hope to be able to come more often!

Thanks for the report, Otto! And thanks for continuing to provide excellent code to our favorite operating system!

(Comments are closed)


Comments
  1. By Philipp (pb) on

    how about scan_ffs(8) for that arena of messed up labels?

    Comments
    1. By Otto Moerbeek (otto) on http://www.drijf.net

      > how about scan_ffs(8) for that arena of messed up labels?

      Still useful. My changes affect the case of a corrupted primary superblock while the label is still OK.

      scan_ffs is used to reconstruct a lost label.

      Comments
      1. By pb (84.184.179.216) on

        > Still useful. My changes affect the case of a corrupted primary superblock while the label is still OK.
        >
        > scan_ffs is used to reconstruct a lost label

        sure, I was more about the point if scan_ffs needs to be aware of additional superblocks or make use of it.

        and also a happy user of scan_ffs to save the day some years back where i messed the label by dd(1) the first megabyte of the disk.

Latest Articles

Credits

Copyright © - Daniel Hartmeier. All rights reserved. Articles and comments are copyright their respective authors, submission implies license to publish on this web site. Contents of the archive prior to as well as images and HTML templates were copied from the fabulous original deadly.org with Jose's and Jim's kind permission. This journal runs as CGI with httpd(8) on OpenBSD, the source code is BSD licensed. undeadly \Un*dead"ly\, a. Not subject to death; immortal. [Obs.]