I ported pigz from Unix to Windows(blog.kowalczyk.info)
83 pointsby speckx7 days ago11 comments
  • igrunert4 days ago
    I recently ported WebKit's libpas memory allocator[1] to Windows, which used pthreads on the Linux and Darwin ports. Depending on what pthreads features you're using it's not that much code to shim to Windows APIs. It's around ~200 LOC[2] for WebKit's usage, which a lot smaller than pthread-win32.

    [1] https://github.com/WebKit/WebKit/pull/41945 [2] https://github.com/WebKit/WebKit/blob/main/Source/bmalloc/li...

    • kjksf3 days ago
      At the time (11 years ago) I wanted this to run on Windows XP.

      The APIs you use there (e.g. SleepConditionVariableSRW()) were only added in Vista.

      I assume a big chunk of pthread emulation code at that time was implementing things like that.

    • malkia4 days ago
      These VirtualAlloc's may intermittently fail if the pagefile is growing...
      • igrunert4 days ago
        Ah yeah, I see Firefox ran into that and added retries:

        https://hacks.mozilla.org/2022/11/improving-firefox-stabilit...

        Seems like a worthwhile change, though I'm not sure when I'll get around to it.

        • account423 days ago
          This is something you also need to do for other Win32 APIs, e.g. file write access may be temporarily blocked by anti-virus programs or whatever and not handling that makes unhappy users.
    • adzm4 days ago
      Never knew about the destructor feature for fiber local allocations!
  • andy994 days ago
    I'm a big fan of pigz, I discovered it 6 years ago when I had some massive files I needed to zip and and 48 core server I was underutilizing. It was very satisfying to open htop and watch all the cores max out.

    Edit: found the screenshot https://imgur.com/a/w5fnXKS

  • haunter4 days ago
    Very old post, needs 2013 in the title

    https://web.archive.org/web/20130407195442/https://blog.kowa...

    • frainfreeze3 days ago
      Seems to be updated, no?
      • jwilk3 days ago
        Not much. The only non-cosmetic difference is:

          -Premake supports Visual Studio 2008 and 2010 (and 2012 supports 2010 project files via conversion).
  +Premake supports latest Visual Studio 2018 and 2022 project files via conversion).
  • kjksf7 days ago
    Worth mentioning that this is only of interest as technical info on porting process.

    The port itself is very old and therefore very outdated.

  • mid-kid3 days ago
    I'm not sure how willing I'd be to trust a pthread library fork from a single no-name github person. The mingw-w64 project provides libwinpthread, which you can download as source from their sourceforge, or as a binary+headers from a well-known repository like msys2.
  • account423 days ago
    > Porting pthreads code to Windows would be a nightmare.

    Porting one application using pthreads to use the Win32 API directly is however a lot more reasonable and provides you more opportunity to deal with impedance mismatches than a full API shim has. Same goes for dirent and other things as well as for the reverse direction. Some slightly higher level abstraction for the thnings your program actually needs is usually a better solution for cross-platform applications than using one OS API and emulating it on other systems.

  • themadsens4 days ago
    I wish premake could gain more traction. It is the comprehensible alternative to Cmake etc.
    • account423 days ago
      I'd rather everyone use CMake than have to deal with yet another build system. Wouldn't be so bad if build systems could at least agree on the user interface and package registry format.
    • beagle34 days ago
      Xmake[0] is as-simple-as-premake and does IIRC everything Premake does and a whole lot more.

      [0] https://xmake.io/

    • PeakKS4 days ago
      It's 2025, just use meson
      • nly4 days ago
        Completely useless in an airgapped environment
        • throwaway20463 days ago
          Could you elaborate on that?
          • carlmr3 days ago
            I'm guessing it needs internet for everything and can't work with local repositories.
            • account423 days ago
              Not really a fan of Meson but I doubt that that's the case as it is very popular in big OSS projects and distributions aren't throwing a fit.
        • PeakKS2 days ago
          No?
  • kristianp4 days ago
    Repository link: https://github.com/kjk/pigz
  • nialv73 days ago
    The best kind of porting - other people have already done most of the work for you!
  • anilakar3 days ago
    Pigz? Good old Pigzip? :)

    https://pc-freak.net/files/hackles.org/cgi-bin/archives.pl%3...

    • ObscureScience3 days ago
      I don't see any relation. Pigz is a multithreaded reimplenentation of gzip (drop in replacement)
  • jqpabc1237 days ago
    This is clearly aimed at faster results in a single user desktop environment.

    In a threaded server type app where available processor cores are already being utilized, I don't see much real advantage in this --- if any.

    • GuinansEyebrows4 days ago
      depends on the current load. i've worked places where we would create nightly postgres dumps via pg_dumpall, then pipe through pigz to compress. it's great if you run it when load is otherwise low and you want to squeeze every bit of performance out of the box during that quiet window.

      this predates the maturation of pg_dump/pg_restore concurrency features :)

      • ggm4 days ago
        Not to over state it, embedding the parallelism into the application drives to the logic "the application is where we know we can do it" but embedding the parallelism into a discrete lower layer and using pipes drives to "this is a generic UNIX model of how to process data"

        The thing with "and pipe to <thing>" is that you then reduce to a serial buffer delay decoding the pipe input. I do this, because often its both logically simple and the component of serial->parallel delay deblocking on a pipe is low.

        Which is where xargs and the prefork model comes in, because instead you segment/shard the process, and either don't have a re-unification burden or its a simple serialise over the outputs.

        When I know I can shard, and I don't know how to tell the appication to be parallel, this is my path out.

    • 4 days ago
      undefined