[Buildroot] Statistics of build per architecture

[Thomas Petazzoni]

Hello,

Yann asked me on IRC what the statistics of build results per
architecture looked like. So here are the results, for all builds done
between April, 1st 2014 and now:

+--------------+---------+----------+-------+------------+------------+
| arch         | success | failures | total | pctsuccess | pctfailure |
+--------------+---------+----------+-------+------------+------------+
| bfin         |      12 |      162 |   174 |     6.8966 |    93.1034 |
| avr32        |      35 |       62 |    97 |    36.0825 |    63.9175 |
| aarch64      |      34 |       35 |    69 |    49.2754 |    50.7246 |
| microblazeel |      50 |       43 |    98 |    51.0204 |    43.8776 |
| xtensa       |      52 |       29 |    83 |    62.6506 |    34.9398 |
| arc          |      52 |       24 |    76 |    68.4211 |    31.5789 |
| arm          |     767 |      355 |  1126 |    68.1172 |    31.5275 |
| sh4          |      50 |       22 |    72 |    69.4444 |    30.5556 |
| nios2        |      64 |       26 |    90 |    71.1111 |    28.8889 |
| powerpc      |     394 |      154 |   549 |    71.7668 |    28.0510 |
| sh4a         |      64 |       19 |    83 |    77.1084 |    22.8916 |
| x86_64       |     271 |       80 |   357 |    75.9104 |    22.4090 |
| mips         |      61 |       16 |    77 |    79.2208 |    20.7792 |
| mips64el     |     167 |       28 |   200 |    83.5000 |    14.0000 |
| mipsel       |      94 |       15 |   109 |    86.2385 |    13.7615 |
| i686         |     974 |      138 |  1116 |    87.2760 |    12.3656 |
| i486         |      26 |        1 |    27 |    96.2963 |     3.7037 |
+--------------+---------+----------+-------+------------+------------+

For comparison, back in February, I gave a similar table, for the
results going from January 1st to February 13rd 2014:

+------------+---------+----------+-------+------------+------------+
| arch       | success | failures | total | pctsuccess | pctfailure |
+------------+---------+----------+-------+------------+------------+
| bfin       |       4 |      199 |   203 |     1.9704 |    98.0296 |
| xtensa     |      14 |       88 |   102 |    13.7255 |    86.2745 |
| microblaze |      24 |      102 |   126 |    19.0476 |    80.9524 |
| avr32      |      27 |       81 |   108 |    25.0000 |    75.0000 |
| arc        |      34 |       67 |   101 |    33.6634 |    66.3366 |
| aarch64    |      40 |       50 |    93 |    43.0108 |    53.7634 |
| x86_64     |     217 |      249 |   471 |    46.0722 |    52.8662 |
| mipsel     |      64 |       61 |   125 |    51.2000 |    48.8000 |
| mips64el   |     121 |      114 |   239 |    50.6276 |    47.6987 |
| nios2      |      59 |       58 |   126 |    46.8254 |    46.0317 |
| arm        |     579 |      436 |  1018 |    56.8762 |    42.8291 |
| mips       |      61 |       47 |   115 |    53.0435 |    40.8696 |
| sh4        |      57 |       39 |    97 |    58.7629 |    40.2062 |
| sh4a       |      63 |       47 |   121 |    52.0661 |    38.8430 |
| powerpc    |    1108 |      523 |  1640 |    67.5610 |    31.8902 |
| i686       |     938 |      308 |  1250 |    75.0400 |    24.6400 |
+------------+---------+----------+-------+------------+------------+

Note that in both cases, we're roughly comparing 1.5 months of builds,
with 1 month being the last month of the 'development' part of the
cycle, and 0.5 month being the beginning of the 'debugging' part of the
cycle. So I would say the results are fairly comparable.

And we can notice a pretty significant improvement in the results.
There's still a lot of work to be done, but it's pretty clear that
we're making progress. We've moved to over 60% of success for most of
the architectures.

Best regards,

Thomas
-- 
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com