Fixes #2074: run_start property when loading SNS files will allow the right instrument definition to be loaded.

Add a GNUSetup file to the CMake setup scripts. This contains options that only apply to the gcc/g++ compilers. At the moment it just activates the usual wrarning sets that we were missing. Apologies for creating a far messier compile output but these really should be on. Re #1881

pooling isn't cooperating as well on the build servers as on my machine where everything ran fine./MantidPlot/ Re #2276

Created a pool for parametrized detector objects to avoid excessive calls to the new operator in multi-threaded loops. If a cached object already exists and is unique then it's memory is reused rather than reallocating another object. This alleviates some of the locking within threaded loops that was impeding the scaling with the number of cores. There is still something to done with instruments but an issue remains with plottable components. Re #2276

Added an executeSubAlg() method to PythonAlgorithm. It creates and runs a sub-alg with the provided parameters. Re #2270

Refs #2167: You can't put ClassName:: in front of your function declarations in a header file; I'm surprised this builds under windows.

Fixes #2324: Call to sortAll before CompressEvents speeds things up from 80 to 30 seconds in one example (95 million event list). Same thing in ConvertToMatrixWorkspace. Nicer reporting.

 Fixes #2277: A million tests for every possible permutation of multiply and divide workspace. Whew!