A comparison between Minix and Linux through benchmarking - Minix

This is a discussion on A comparison between Minix and Linux through benchmarking - Minix ; Hi to all,first time to write(actually first time to ever come to a newsgroup,since I just learned what it is and how to use it-thanks Al) as part of my thesis(how we got from minix2 to minix3) I must do ...

+ Reply to Thread
Results 1 to 3 of 3

Thread: A comparison between Minix and Linux through benchmarking

  1. A comparison between Minix and Linux through benchmarking

    Hi to all,first time to write(actually first time to ever come to a
    newsgroup,since I just learned what it is and how to use it-thanks Al)
    as part of my thesis(how we got from minix2 to minix3) I must do some
    benchmarking between Minix and Linux and discuss how and why. I start
    thinking about doing one CPU-bound an one I/O-bound-no networking,
    problem was I didn't find any. I have heard that only Dhrystone runs
    with no problems on both OS (I was planning to use a simple gcc command
    on linux but a friend told me to use the make command-have never used
    it, must dig a bit to see what it does).So the main question, does
    anyone know any I/O benchmark? Info from anyone who has done anything
    similar (some sort of a guide/tutorial would be great) is more than
    welcome.


  2. Re: A comparison between Minix and Linux through benchmarking

    mike_k12003@yahoo.gr wrote:
    > Hi to all,first time to write(actually first time to ever come to a
    > newsgroup,since I just learned what it is and how to use it-thanks Al)
    > as part of my thesis(how we got from minix2 to minix3) I must do some
    > benchmarking between Minix and Linux and discuss how and why. I start
    > thinking about doing one CPU-bound an one I/O-bound-no networking,
    > problem was I didn't find any. I have heard that only Dhrystone runs
    > with no problems on both OS (I was planning to use a simple gcc command
    > on linux but a friend told me to use the make command-have never used
    > it, must dig a bit to see what it does).So the main question, does
    > anyone know any I/O benchmark? Info from anyone who has done anything
    > similar (some sort of a guide/tutorial would be great) is more than
    > welcome.


    Perhaps you can modify some of the existing Linux benchmarks to run on
    minix. This site may be helpful:
    http://math.nmu.edu/~benchmark/index.php?page=intro (try googling for
    linux kernel benchmark to get more)

    I also recommend doing some scaling benchmarks. For each test program
    that you have, you can run 1, 2, 4, 8, ... copies in parallel and see
    how that affects things.

    Can you post the results when you're done ?


  3. Re: A comparison between Minix and Linux through benchmarking

    > Hi to all,first time to write(actually first time to ever come to a
    > newsgroup,since I just learned what it is and how to use it-thanks Al)
    > as part of my thesis(how we got from minix2 to minix3) I must do some
    > benchmarking between Minix and Linux and discuss how and why. I start
    > thinking about doing one CPU-bound an one I/O-bound-no networking,
    > problem was I didn't find any. I have heard that only Dhrystone runs
    > with no problems on both OS (I was planning to use a simple gcc
    > command on linux but a friend told me to use the make command-have
    > never used it, must dig a bit to see what it does).So the main
    > question, does anyone know any I/O benchmark? Info from anyone who
    > has done anything similar (some sort of a guide/tutorial would be
    > great) is more than welcome.


    When benchmarking, keep in mind that Minix does not support floating
    point instructions. This means the result will be quite useless if you
    use a benchmark which performs floating point operations (Linux would
    win with a huge difference in this case). The same goes for MMX and SSE.




+ Reply to Thread