Menu

#1 Dynamic allocation of memory

closed
None
5
2009-01-15
2005-06-05
No

Hi!

Andreas Tille <tillea_AT_rki.de> has created a patch
that implements dynamic allocation of memory on Mummer.
Could you check it and see if everything is OK?
Also, if it won't cause any problems, it might be a
good idea to apply it to upstream version of Mummer.

Thank you!
Nelson

Discussion

  • Nelson A. de Oliveira

    Patch to implement dynamic allocation of memory

     
  • Nelson A. de Oliveira

    New implementantion of dynamic allocation of memory

     
  • Nelson A. de Oliveira

    Logged In: YES
    user_id=1269234

    Hi!

    It was missing some informations with my last comment.

    First, a problem has raised when creating a Debian package
    of mummer. When trying to use ldd on the annotate binary,
    ldd exits with a "not a dynamic executable". This issue has
    been discussed on the debian-devel mailing list
    (http://lists.debian.org/debian-devel/2005/06/msg00037.html).
    The conclusion was that annotate was allocating a large size
    of memory statically. With machines with less than 512 MB of
    RAM, it was not possible to execute annotate (and because
    this, ldd was not possible to get the needed informations of
    the program). annotate gets killed if trying to run with
    less than 512 MB.

    A first solution to that problem was done by Andreas Tille
    (the patch that I have submitted), but it wasnt tested.
    With test, we saw that it was giving segmentation fault.

    Steffen Moeller <moeller_AT_pzr.uni-rostock.de> worked on a
    new patch (the one attached now), that also implements
    dynamic allocation of memory.
    It was tested and the results obtained with it applied were
    the same from the original source code. The program run
    perfectly.

    The only difference between the original and the modified
    version, is that the original is faster.

    The differences are (average of a test):

    Original version: 0.265s (with -O3 from g++) and 0.979s
    (without any -O)
    Patched version: 0.610s (with -O3) and 1.963s (without any -O)

    (Bigger time due to overhead of dynamic allocation?)

    If you can give some comments about that, I would be very
    grateful.

    Thank you.
    Nelson

     
  • Adam Phillippy

    Adam Phillippy - 2005-06-06
    • assigned_to: nobody --> aphillippy
     
  • Adam Phillippy

    Adam Phillippy - 2005-06-06

    Logged In: YES
    user_id=1231950

    Sure thing. I'll take a look at this when time permits.
    Thanks for bringing this to my attention.

    -Adam

     
  • Adam Phillippy

    Adam Phillippy - 2009-01-15
    • status: open --> closed
     
  • Nelson A. de Oliveira

    Hi.
    Is it fixed, please?

     

Log in to post a comment.