Results of the CPython benchmark suite on dict_version-8.patch.

IMHO regex_v8 can be ignored, this benchmark is unstable (issue #26275).

Original python: ../pep509/python
3.6.0a0 (default:3a9b47b062b9, Apr 19 2016, 16:23:15) 
[GCC 5.3.1 20151207 (Red Hat 5.3.1-2)]

Patched python: ../pep509/python
3.6.0a0 (default:96f61aab2c6e, Apr 19 2016, 15:21:15) 
[GCC 5.3.1 20151207 (Red Hat 5.3.1-2)]

$ python3 -u perf.py --rigorous -b all ../default.copy/python ../pep509/python

INFO:root:Automatically selected timer: perf_counter
(...)
Report on Linux smithers 4.4.4-301.fc23.x86_64 #1 SMP Fri Mar 4 17:42:42 UTC 2016 x86_64 x86_64
Total CPU cores: 8

### 2to3 ###
Min: 6.714927 -> 6.856992: 1.02x slower
Avg: 6.726773 -> 6.984359: 1.04x slower
Significant (t=-5.16)
Stddev: 0.01160 -> 0.11111: 9.5816x larger

### call_method ###
Min: 0.312567 -> 0.323260: 1.03x slower
Avg: 0.313399 -> 0.323821: 1.03x slower
Significant (t=-284.62)
Stddev: 0.00042 -> 0.00048: 1.1430x larger

### call_simple ###
Min: 0.242091 -> 0.228922: 1.06x faster
Avg: 0.243613 -> 0.229294: 1.06x faster
Significant (t=2134.23)
Stddev: 0.00010 -> 0.00005: 1.9124x smaller

### etree_generate ###
Min: 0.255097 -> 0.265558: 1.04x slower
Avg: 0.256947 -> 0.267478: 1.04x slower
Significant (t=-67.89)
Stddev: 0.00111 -> 0.00108: 1.0250x smaller

### etree_parse ###
Min: 0.281726 -> 0.290976: 1.03x slower
Avg: 0.283626 -> 0.292642: 1.03x slower
Significant (t=-62.59)
Stddev: 0.00109 -> 0.00094: 1.1540x smaller

### etree_process ###
Min: 0.217164 -> 0.229552: 1.06x slower
Avg: 0.219401 -> 0.231242: 1.05x slower
Significant (t=-75.05)
Stddev: 0.00112 -> 0.00111: 1.0053x smaller

### fannkuch ###
Min: 1.012875 -> 0.985196: 1.03x faster
Avg: 1.014760 -> 0.992104: 1.02x faster
Significant (t=46.44)
Stddev: 0.00287 -> 0.00395: 1.3778x larger

### float ###
Min: 0.259226 -> 0.251918: 1.03x faster
Avg: 0.266530 -> 0.260176: 1.02x faster
Significant (t=10.69)
Stddev: 0.00403 -> 0.00437: 1.0836x larger

### json_load ###
Min: 0.430602 -> 0.433593: 1.01x slower
Avg: 0.431278 -> 0.486924: 1.13x slower
Significant (t=-24.86)
Stddev: 0.00045 -> 0.02238: 49.8974x larger

### normal_startup ###
Min: 0.319092 -> 0.326082: 1.02x slower
Avg: 0.320093 -> 0.326842: 1.02x slower
Significant (t=-70.52)
Stddev: 0.00067 -> 0.00069: 1.0283x larger

### regex_effbot ###
Min: 0.048969 -> 0.048313: 1.01x faster
Avg: 0.050019 -> 0.048415: 1.03x faster
Significant (t=31.41)
Stddev: 0.00051 -> 0.00006: 8.8152x smaller

### regex_v8 ###
Min: 0.051040 -> 0.043412: 1.18x faster
Avg: 0.051325 -> 0.043577: 1.18x faster
Significant (t=45.60)
Stddev: 0.00120 -> 0.00120: 1.0013x larger

### richards ###
Min: 0.157625 -> 0.161126: 1.02x slower
Avg: 0.158952 -> 0.162519: 1.02x slower
Significant (t=-34.92)
Stddev: 0.00073 -> 0.00071: 1.0351x smaller

### silent_logging ###
Min: 0.070547 -> 0.069902: 1.01x faster
Avg: 0.072246 -> 0.069934: 1.03x faster
Significant (t=65.51)
Stddev: 0.00035 -> 0.00003: 11.9828x smaller

The following not significant results are hidden, use -v to show them:
call_method_slots, call_method_unknown, chameleon_v2, chaos, django_v3, etree_iterparse, fastpickle, fastunpickle, formatted_logging, go, hexiom2, json_dump_v2, mako_v2, meteor_contest, nbody, nqueens, pathlib, pickle_dict, pickle_list, pidigits, raytrace, regex_compile, simple_logging, spectral_norm, startup_nosite, telco, tornado_http, unpack_sequence, unpickle_list.