BEGINBLOCK intrinsic instruction

[chfast]

Instead of checking if an instruction starts a basic block, an "internal" BEGINBLOCK instruction is inserted that performs basic block checks. This simplifies the dispatch loop implementation and removes a branch there.

This gives 2% - 8% execution speed improvement.

Comparing bin/evmone-bench-master to bin/evmone-bench
Benchmark                            Time             CPU      Time Old      Time New       CPU Old       CPU New
-----------------------------------------------------------------------------------------------------------------
sha1_shifts/empty                 -0.0881         -0.0881            63            58            63            58
sha1_shifts/1351                  -0.0715         -0.0715          1205          1119          1205          1119
sha1_shifts/2737                  -0.0845         -0.0845          2375          2175          2375          2175
sha1_shifts/5311                  -0.0803         -0.0803          4614          4244          4614          4244
sha1_shifts/65536                 -0.0722         -0.0722         56199         52142         56198         52142
stop                              +0.0156         +0.0156             0             0             0             0
blake2b_huff/empty                -0.0804         -0.0804            81            74            81            74
blake2b_huff/abc                  -0.0786         -0.0786            80            74            80            74
blake2b_huff/2805nulls            -0.0414         -0.0413           602           577           602           577
blake2b_huff/2805aa               -0.0317         -0.0317           598           579           598           579
blake2b_huff/5610nulls            -0.0335         -0.0335          1124          1086          1124          1086
blake2b_huff/8415nulls            -0.0332         -0.0332          1627          1573          1627          1573
blake2b_huff/65536nulls           -0.0250         -0.0250         12128         11824         12128         11824
sha1_divs/empty                   -0.0163         -0.0162           105           103           105           103
sha1_divs/1351                    -0.0260         -0.0260          2112          2057          2112          2057
sha1_divs/2737                    -0.0277         -0.0276          4135          4020          4135          4020
sha1_divs/5311                    -0.0192         -0.0191          8029          7875          8029          7875
sha1_divs/65536                   -0.0214         -0.0213         97719         95631         97714         95632
blake2b_shifts/empty              +0.0000         +0.0000             0             0             0             0
blake2b_shifts/2805nulls          -0.0319         -0.0319          7585          7343          7585          7343
blake2b_shifts/5610nulls          -0.0277         -0.0276         15070         14653         15069         14653
blake2b_shifts/8415nulls          -0.0326         -0.0326         22651         21911         22650         21911
blake2b_shifts/65536nulls         -0.0338         -0.0338        177036        171051        177030        171051

[codecov-io]

Codecov Report

Merging #74 into master will decrease coverage by 0.16%.
The diff coverage is 100%.

@@            Coverage Diff             @@
##           master      #74      +/-   ##
==========================================
- Coverage   99.43%   99.27%   -0.17%     
==========================================
  Files          16       16              
  Lines        1778     1791      +13     
  Branches      161      160       -1     
==========================================
+ Hits         1768     1778      +10     
- Misses          1        4       +3     
  Partials        9        9