New Mac OS X port for Macintosh users! Discuss Chaos Strikes Back for Windows and Linux, an unofficial port of Chaos Strikes Back to PC by Paul Stevens, as well as CSBuild, an associated dungeon editor. Dungeon Masters purchasing this module can use it to run the adventure with very little prep, other than reading through the adventure in advance. The contents of the story are indexed and linked, where appropriate. Combat encounters and boxed text are preloaded and ready to drop onto maps and into the chat window with a few clicks.
The Hoard Memory Allocator
Copyright (C) 1998-2014 by Emery Berger
The Hoard memory allocator is a fast, scalable, and memory-efficientmemory allocator that works on a range of platforms,including Linux, Solaris, Mac OS X, and Windows.
Hoard is a drop-in replacement for malloc that can dramaticallyimprove application performance, especially for multithreaded programsrunning on multiprocessors and multicore CPUs. No source code changesnecessary: just link it in or set one environment variable (seeBuilding Hoard, below).
Press
'If you'll be running on multiprocessor machines, ... use EmeryBerger's excellent Hoard multiprocessor memory management code. It's adrop-in replacement for the C and C++ memory routines and is very faston multiprocessor machines.'
'(To improve scalability), consider an open source alternative such asthe Hoard Memory Manager...'
'Hoard dramatically improves program performance through its moreefficient use of memory. Moreover, Hoard has provably bounded memoryblowup and low synchronization costs.'
Users
Companies using Hoard in their products and servers include AOL,British Telecom, Blue Vector, Business Objects (formerly CrystalDecisions), Cisco, Credit Suisse, Entrust, InfoVista, Kamakura,Novell, Oktal SE, OpenText, OpenWave Systems (for their Typhoon andTwister servers), Pervasive Software, Plath GmbH, Quest Software,Reuters, Royal Bank of Canada, SAP, Sonus Networks, TataCommunications, and Verite Group.
Open source projects using Hoard include the Asterisk Open SourceTelephony Project, Bayonne GNU telephony server, the Cilk parallelprogramming language, the GNU Common C++ system, the OpenFOAMcomputational fluid dynamics toolkit, and the SafeSquid web proxy.
Hoard is now a standard compiler option for the Standard PerformanceEvaluation Corporation's CPU2006 benchmark suite for the Intel andOpen64 compilers.
Licensing
Hoard is distributed under the GPL (v2.0), and can also be licensedfor commercial use.
Because of the restrictions imposed by the GPL license (you must makeyour code open-source), commercial users of Hoard can purchase non-GPLlicenses through the University of Texas at Austin. Please consult thecurrent Hoard pricinginformation(updated 2/5/2009), which lists a number of options for purchasinglicenses, as well as software license terms andconditions,and the software licenseagreement:note that the main UT-Austin licensingpage always contains themost up-to-date documents.
To obtain a license, please contact Jitendra Jain directly(jjain@otc.utexas.edu) and copy Emery Berger (emery@cs.umass.edu).
Jitendra Jain's full contact information follows:
Jitendra Jain
The University of Texas at Austin
Office of Technology Commercialization
MCC Building, Suite 1.9A
3925 West Braker Lane
Austin, Texas 78759
(512) 471-9055, (512) 475-6894 (fax)
Why Hoard?
There are a number of problems with existing memory allocators thatmake Hoard a better choice.
Contention
Multithreaded programs often do not scale because the heap is abottleneck. When multiple threads simultaneously allocate ordeallocate memory from the allocator, the allocator will serializethem. Programs making intensive use of the allocator actually slowdown as the number of processors increases. Your program may beallocation-intensive without you realizing it, for instance, if yourprogram makes many calls to the C++ Standard Template Library (STL). Hoard eliminates this bottleneck.
False Sharing
System-provided memory allocators can cause insidious problems for multithreaded code. They canlead to a phenomenon known as 'false sharing': threads on different CPUscan end up with memory in the same cache line, or chunk ofmemory. Accessing these falsely-shared cache lines is hundreds oftimes slower than accessing unshared cache lines. Hoard is designed to prevent false sharing.
Blowup
Multithreaded programs can also lead the allocator to blowup memoryconsumption. This effect can multiply the amount of memory needed torun your application by the number of CPUs on your machine: four CPUscould mean that you need four times as much memory. Hoard is guaranteed (provably!) to bound memory consumption.
Building Hoard (Unix/Mac)
NOTE: Make sure to invoke git as follows:
To build Hoard on non-Windows platforms, change into the src/directory and run make followed by the appropriate target. If youtype make, it will present a list of available targets. Theseinclude linux-gcc-x86, solaris-sunw-sparc, macos, windows, andmore.
You can then use Hoard by linking it with your executable, orby setting the LD_PRELOAD environment variable, as in
in Solaris:
(32-bit version)
(64-bit version)
Mac Os Versions
or, in Mac OS X:
Building Hoard (Windows)
(64-bit version)
Mac Os Versions
or, in Mac OS X:
Building Hoard (Windows)
Change into the src directory and build the Windows version:
To use Hoard, link your executable with sourceuselibhoard.cpp and libhoard.lib.You must use the /MD flag.
Example:
To run yourapp.exe, you will need to have libhoard.dll in your path.
Dungeon Hoard Mac Os Download
Benchmarks
Dungeon Hoard Mac Os Catalina
The directory benchmarks/ contains a number of benchmarks used toevaluate and tune Hoard.
Dungeon Hoard Mac Os 11
Technical Information
Hoard has changed quite a bit over the years, but for technical details of the first version of Hoard, read Hoard: AScalable Memory Allocator for Multithreaded Applications,by Emery D. Berger, Kathryn S. McKinley, Robert D. Blumofe, and PaulR. Wilson. The Ninth International Conference on Architectural Supportfor Programming Languages and Operating Systems(ASPLOS-IX). Cambridge, MA, November 2000.
