HN Explorer

4 comments

[ 5.0 ms ] story [ 18.3 ms ] thread
nathanappere
Although it does nos focus on memory profiling, the best serie of post I know on memory allocation is Gustavo Duartes one: http://duartes.org/gustavo/blog/post/anatomy-of-a-program-in... http://duartes.org/gustavo/blog/post/how-the-kernel-manages-... http://duartes.org/gustavo/blog/post/page-cache-the-affair-b...
ajross
And any discussion about memory topics under linux has to include Ulrich Drepper's fantastic series of LWN.net articles: http://lwn.net/Articles/257209/

Very thick in places, but absolutely worth reading. Still the only good place to find an exposition about how SDRAM timings actually work.

acg
Also worth looking into tcmalloc: http://goog-perftools.sourceforge.net/doc/tcmalloc.html

Which comes with tools: http://goog-perftools.sourceforge.net/doc/heap_checker.html http://goog-perftools.sourceforge.net/doc/heap_profiler.html

srean
Also nedmalloc https://github.com/ned14/nedmalloc and hoard http://www.hoard.org/. There is some comparison between the three here http://www.nedprod.com/programs/portable/nedmalloc/#FAQ The following is a snippet from that FAQ

  Is tcmalloc better or worse than nedmalloc?... nedmalloc
  is about equal to tcmalloc for threadcache-only ops and
  substantially beats it for non-threadcache ops. nedmalloc
  is also written in C rather than C++ and v0.5 of tcmalloc
  only works on Unix systems and not win32. tcmalloc
  achieves its speed by never returning memory to the
  system - free space reclamation is one of the slowest
  parts of any allocator. Therefore tcmalloc should NOT be
  used outside long running server processes (and indeed
  its own docs say the same).
However I am not sure if the criticism about not returning the memory is true anymore for the latest version tcmalloc.