Don't accept null pointer in tests when requesting aligned memory

See merge request redox-os/relibc!272

Minor fixes to posix_memalign()

See merge request redox-os/relibc!271

Gcc regressions

See merge request redox-os/relibc!270

as it seams you can do something like
        printf ("A%*s%s/\n", -5, "B", "CC");
and it will print the padding to the left

Scanf function requires look ahead to function properly, In case of
scanning from a buffer that will not be an issue, but in our case we are
reading from file, so lookaheads needs to be undone (via lseek) in our
case. The only problem here is that if we opened a file that doesn't
support lseek such as many of the file /dev/*

At least in relibc, each call to ungetc should decrement ftell() by one
also allowing negative ftell() this is not possible on relibc thus gcc
failing to compile (gcc compiles tools that is later used to compile gcc
itself and these tools are the ones that fail)

According to the standards, only one ungetc may be guaranteed however
glibc allows more than one of those, and to be glibc compatiable, one
needs to be able to do the same, allowing only 1 ungetc may trigger bug
while compiling gcc as ungetc is used there alot

Normally one shouldn't be using this datatype ever, but then someone
have to tell that to gcc folks :(

Binutils regressions

See merge request redox-os/relibc!269

There was a bug (also uncovered via binutils) where R_X86_64_DTPOFF64 is
set uncorrectly. This program is the minimal reproducer of the seg fault

  #include <errno.h>
  int main() {
        int oerrno = errno;
  }

But it works after the bug fix.

There was bug in printf where space paddings cause segfault,
the problem was that it was pulled from the stack twice while it should
be only done once.

In patch 1182d120, I mistakingly added
the size of the gap to the total size of the binary, which was not
accurate. As the size of the binary was calculate by subtracting the
upperbound from the lower bound, thus all gaps in the middle are taking
into account.

Ld.so regressions

See merge request redox-os/relibc!268

This fixes a regression in gdb where sometimes it decides to ignore the
first entry in the list.

I noticed that the implementation is noth precise although it worked.
and instead of using the sections to identify memory addresses of
DT_DEBUG. I used segements

Some ELFs can have gaps between the segments, this results in problems
when mprotecting or when converting (addr + size) into rust slice.
Motivated by this This patch will fill these gaps with mmaped addresses.
In theory no real memory should be allocated because mmap is lazy
allocator.

Wide strings Functions

See merge request !267

This patch implements wcscasecmp and wcsncasecmp. These two
function are required for binutils to link against relibc.

This patch creates wctype.h and impelementat two functions
that belong to that header file towupper and towlower. These
functions are building blocks for wcscasecmp and wcsncasecmp
which are utilized by binutils.

The implementation for towlower and towupper seams to be complex
so this implementation is mimicking that of musl libc

Implement #include<sys/user.h> and #include<sys/procfs.h>

See merge request !266

This patch implements sys/user.h file that works for both x86_64 as well
as aarch64. This include file is used by sys/procfs.h which is needed
dependency for binutils. There is bug in this patch in aarch64 implementation
which is the lack of f128 implementation in rust, thus we can't create cbinding
for long double.

Init fini ld.so

See merge request redox-os/relibc!265