All of this code was already correct, this just tidies it a little

1) use __builtin_cpu_supports("avx512f") only for gcc ver >= 5

2) Introduced two new macro's:
HAVE_ATTRIBUTE_TARGET_@ISA@_WITH_INTRINSICS for ensuring compiler can
compile functions that use intrinsics and are compiled with avx2/avx512f
attributes

This commit implements vectorized single precision exponential and
natural log using AVX2 and AVX512.

Accuracy:

| Function | Max ULP Error | Max Relative Error |
|----------|---------------|--------------------|
| np.exp   | 2.52          | 2.1E-07            |
| np.log   | 3.83          | 2.4E-07            |

Performance:

(1) Micro-benchmarks: measured execution time of np.exp and np.log using
timeit package in python. Each function is executed 1000 times and this
is repeated 100 times. The standard deviation for all the runs was less
than 2% of their mean value and hence not included in the data. The
vectorized implementation was upto 7.6x faster than the scalar version.

| Function | NumPy1.16 | AVX2   | AVX512 | AVX2 speedup | AVX512 speedup |
| -------- | --------- | ------ | ------ | ------------ | -------------- |
| np.exp   | 0.395s    | 0.112s | 0.055s | 3.56x        | 7.25x          |
| np.log   | 0.456s    | 0.147s | 0.059s | 3.10x        | 7.64x          |

(2) Logistic regression: exp and log are heavily used in training neural
networks (as part of sigmoid activation function and loss function
respectively). This patch significantly speeds up training a logistic
regression model. As an example, we measured how much time it takes to
train a model with 15 features using 1000 training data points. We
observed a 2x speed up to train the model to achieve a loss function
error < 10E-04.

| Function       | NumPy1.16  | AVX2   | AVX512 | AVX2 speedup | AVX512 speedup |
| -------------- | ---------- | ------ | ------ | ------------ | -------------- |
| logistic.train | 121.0s     | 75.02s | 60.60s | 1.61x        | 2.02x          |

 ENH: Generalized ufunc signature expansion for frozen and flexible dimensions

On distributions with /sys/kernel/mm/transparent_hugepage/enabled set to
madvise we do not automatically get our memory backed by huge pages.
Explicitly madvise these regions so they can use larger pages.

This improves performance of the common large array allocation/free
cycle in numpy as page fault overhead is reduced.

    asv continuous -E virtualenv:3.6 --bench bench_io.Copy HEAD^ HEAD

      2.64±0.3ms      1.99±0.06ms     0.75  bench_io.Copy.time_memcpy_large_out_of_place('float32')
      5.51±0.2ms       3.74±0.3ms     0.68  bench_io.Copy.time_memcpy_large_out_of_place('float64')
     4.76±0.02μs      3.02±0.04μs     0.63  bench_io.Copy.time_cont_assign('float32')
      5.55±0.1ms       3.31±0.1ms     0.60  bench_io.Copy.time_memcpy_large_out_of_place('complex64')
      11.3±0.2ms       6.26±0.2ms     0.56  bench_io.Copy.time_memcpy_large_out_of_place('complex128')

Closes gh-11919