[RFT] mem-ref2 branch ready to merge

[Richard Guenther]

I am starting a review of the changes I made on the mem-ref2 branch
in preparation for a merge to trunk.  This is a request for testing.

I have personally bootstrapped and tested on x86_64-unknown-linux
and powerpc64-linux (both with 32bit multilibs included and for
all languages).  I also regularly build SPEC CPU 2006 on x86_64,
current results are

                           SPEC(R) CINT2006 Summary
                                  Estimated                       
Estimated
                Base     Base       Base        Peak     Peak       Peak
Benchmarks      Ref.   Run Time     Ratio       Ref.   Run Time     Ratio
==============================================================================
400.perlbench    9770        325       30.1 *    9770        327       
29.8 *
401.bzip2        9650        464       20.8 *    9650        465       
20.8 *
403.gcc          8050        290       27.8 *    8050        290       
27.7 *
429.mcf          9120        306       29.8 *    9120        307       
29.7 *
445.gobmk       10490        421       24.9 *   10490        421       
24.9 *
456.hmmer        9330        369       25.3 *    9330        369       
25.3 *
458.sjeng       12100        451       26.8 *   12100        447       
27.1 *
462.libquantum  20720        332       62.5 *   20720        327       
63.4 *
464.h264ref     22130        566       39.1 *   22130        563       
39.3 *
471.omnetpp      6250        297       21.0 *    6250        289       
21.6 *
473.astar        7020        382       18.4 *    7020        377       
18.6 *
483.xalancbmk    6900        212       32.6 *    6900        220       
31.3 *
 Est. SPECint(R)_base2006              28.4
 Est. SPECint2006                                                      
28.4

                           SPEC(R) CFP2006 Summary
                                  Estimated                       
Estimated
                Base     Base       Base        Peak     Peak       Peak
Benchmarks      Ref.   Run Time     Ratio       Ref.   Run Time     Ratio
410.bwaves      13590        336       40.5 *   13590    335           
40.6 *
416.gamess      19580        772       25.4 *   19580    770           
25.4 *
433.milc         9180        387       23.7 *    9180    388           
23.7 *
434.zeusmp       9100        357       25.5 *    9100    358           
25.5 *
435.gromacs      7140        341       20.9 *    7140    341           
21.0 *
436.cactusADM   11950        435       27.5 *   11950    436           
27.4 *
437.leslie3d     9400        316       29.8 *    9400    315           
29.8 *
444.namd         8020        412       19.4 *    8020    413           
19.4 *
447.dealII      11440        312       36.6 *   11440    314           
36.4 *
450.soplex       8340        218       38.2 *    8340    217           
38.4 *
453.povray       5320        192       27.6 *    5320    186           
28.5 *
454.calculix     8250        385       21.4 *    8250    384           
21.5 *
459.GemsFDTD    10610        357       29.8 *   10610    356           
29.8 *
465.tonto        9840        390       25.3 *    9840    388           
25.4 *
470.lbm         13740        258       53.4 *   13740    258           
53.2 *
481.wrf         11170        367       30.4 *   11170    368           
30.4 *
482.sphinx3     19490        478       40.8 *   19490    478           
40.8 *
 Est. SPECfp(R)_base2006               29.3
 Est. SPECfp2006                                                       
29.3

Measured on an iCore7 with -O3 -ffast-math -funroll-loops -march=core2.
No surprises here.

As said I am going over the changes, will prepare a changelog and will
post interesting pieces for re-review.  The branch merge will happen
not earlier than next week.

Thanks,
Richard.