Memory Mapped Files

This article is intended to extend the idea of Inter Process Communication with a Memory Mapped File object by using a class that handles that memory object (or any memory) as a memory pool, a "heap". This class, CAllocationHandler, creates a layout of that memory consisting of a header, an allocation table, and a memory area where the allocated memory blocks resides. Those memory blocks may be moved by the class during the allocation/deallocation calls, so the memory blocks are mainly handled using "memory handles".

To directly access a memory block, it must be locked through this handle. The memory handle can also be transferred between processes that uses the same Memory Mapped File object and used to access the data.

The archive AllocationHandler_src contains the class files
  • AllocationHandler.h
  • AllocationHandler.cpp
  • AllocationHandler.inl
and the article (a plain text file)
  • AllocationHandler.txt
The archive AllocationHandler_demo contains a simple Console application that demonstrates the use of CAllocationHandler by simulating two producer threads of data allocated on the "heap", then saving the memory handles in a CList that is periodically checked by a consumer thread.

Downloads

Download demo project - 7 Kb
Download source - 15 Kb


Comments

  • Deadlocks!

    Posted by Legacy on 02/16/2004 12:00am

    Originally posted by: Raymond Lee

    The code seems to deadlock all the time on my machine (Win2K Pro). I haven't tried to debug it. Looks like a good piece of code otherwise!

    Reply
Leave a Comment
  • Your email address will not be published. All fields are required.

Top White Papers and Webcasts

  • Entire organizations suffer when their networks can't keep up and new opportunities are put on hold. Waiting on service providers isn't good business. In these examples, learn how to simplify network management so that your organization can better manage costs, adapt quickly to business demands, and seize market opportunities when they arise.

  • Today's enterprise datacenter can be one of the most complex business environments with dozens (for smaller business), hundreds (for larger and midsize business), even thousands (for hyperscale businesses) of servers that must be managed and monitored. At this level, just managing the cords can be challenging — let alone keeping up with the growing need for more agility and scalability within the datacenter. Simply put, companies are aggressively looking for less complexity and more agility from their …

Most Popular Programming Stories

More for Developers

RSS Feeds

Thanks for your registration, follow us on our social networks to keep up-to-date