Wip/ancestor merging



See merge request !355

simuvex_c: call pkg-config for libs configuration string of glibc



See merge request !381

Add a new SimProcedure for mmap in libc.



See merge request !380

Implement x86g_calculate_RCR.



See merge request !379

raise simprocedure error on IO_FILE operations when the arch is missing



See merge request !378

Add support for arch_prctl syscall

Fast unicorn memory mapping

This MR implements batch unicorn memory mapping.

The old way of mapping memory page by page from SimuVEX to Unicorn causes Unicorn (QEMU in fact) to be extremely slow when there are too many pages mapped (like, over 200 pages). Instead of touching QEMU code to fix the underlying issue, this MR greatly mitigates the issue by trying to map 256 pages each time. Though it is not a fundamental fix, it should be "good enough" for now.

Preliminary performance tests show that the following program runs within 2.2 seconds with this patch, and takes over 19 seconds to run without.

```
#include <stdio.h>
#define SIZE 1000000

int buf[SIZE];

void perf_0()
{
    int i = 0;
    for (i = 0; i < SIZE; ++i)
    {
        buf[i] = i % 0x1337;
    }
}

int main()
{
    perf_0();
}
```

See merge request !376