In TProofBench::DrawCPU, add possibility to extract of a couple of numbers supposed
   to give an idea of the computing specs of the cluster being benchmarked. These are
   the maximum rate for the standard CPU intensive task and the normalized, per worker,
   rate. Both are expressed in RNGPS (RaNdom Generation Per Second).

   A new static method TProofBench::GetPerfSpecs provides a simpler interface to this
   new options, optionally scanning a directory for proofbench outputs and giving the
   possibility to choose which file to use to extarct the numbers:

 root [0] TProofBench::GetPerfSpecs()
Several possible files found:
  0     std:    2011-12-07 15:03:10     <no decription> (file: ./proofbench-plitehp24.cern.ch-24w-20111207-1559.root)
  1     std:    2011-12-07 15:09:26     <no decription> (file: ./proofbench-lite-24w-20111207-1605.root)
  2     std:    2011-12-07 17:08:34     PROOF-Lite on pcphsft64, 4 workers (file: ./proofbench-pcphsft64-lite-4w-20111207-1806.root)
  3     std:    2011-12-07 17:57:26     PROOF at skaf.saske.sk, 60 workers (file: ./proofbench-prf000-iep-grid.saske.sk-60w-20111207-1851.root)
  4     std:    2011-12-08 11:17:15     PROOF at kiaf.sdfarm.kr, 96 workers (file: ./proofbench-kiaf.sdfarm.kr-96w-20111208-1133.root)
  5     std:    2011-12-13 18:22:48     PROOF at cernvm24, 24 workers (file: ./proofbench-cernvm24.cern.ch-24w-20111213-1621.root)
Make your choice [5] 3
 FCN=1610.99 FROM MIGRAD    STATUS=CONVERGED      37 CALLS          38 TOTAL
                     EDM=5.60469e-11    STRATEGY= 1      ERROR MATRIX ACCURATE 
  EXT PARAMETER                                   STEP         FIRST   
  NO.   NAME      VALUE            ERROR          SIZE      DERIVATIVE 
   1  offset       2.50025e+05   8.07163e+03   8.71576e+01  -1.26091e-09
   2  slope        3.85933e+05   4.17679e+02   7.02211e+00   1.82330e-09
* ************************************************************ *
* Cluster: PROOF at kiaf.sdfarm.kr, 96 workers                 *
* Performance measurement from scalability plot:               *
*    rate max:         37.680   megaRNGPS (@ 95 workers)       *
*    per-worker rate:  0.386    megaRNGPS                      *
* ************************************************************ *
root [1] 

   The patch also adds a TNamed in the proofbench output file with a string describing
   the setup used for the test (in the example above: 'PROOF at kiaf.sdfarm.kr, 96 workers').
   The name of the TNamed is 'PB_description'.

   A web page is being created with the specs from some clusters where the bench has been run.


git-svn-id: http://root.cern.ch/svn/root/trunk@42496 27541ba8-7e3a-0410-8455-c3a389f83636