During resolution of `..` component the last slash position won't be updated.

number of matched component should be decremented only when
component that was counted as matched gets removed.

for instance when `/tmp/ext/..` gets resolved to `/tmp` we should not
decrement the matched counter cause we didn't match ext before.

when resolving path we don't want always
to resolve also the last components if it is a link. Cause in some cases
we want exactly the path of that link and not the one it points to.

lstat and readlink are some example of functions that don't want to
resolve the last link.

relativize path now accepts a file descriptor.

If fd is AT_FDCWD relative path need to be considered with respect to
the current working directory. Otherwise the path is relative to the
directory with the given file descriptor

when the resolved path is the root dir we were returning an empty string

when only PSP are given as input path '//////', we got segfault

Instead of using the more complex adafs_pread function now is possible
to use adafs_read when it is necessary to read starting from the current
file position

Fix typo

when size is 0 the buffer to host the path of CWD must be allocated by
us

Always store complete path as CWD instead of internal and relative ones

`tar` and other gnu coreutils are using __openat{64,}_2 variant.
We need to intercept also that apart of the main `openat` function.

StackOverflow related question [1]

[1]: https://stackoverflow.com/questions/9161116/intercepting-the-openat-system-call-for-gnu-tar

add support for path resolution, this will be used to get canonical path.

More info at `man path_resolution`

During interception library startup the CWD of intercepted process is
stored in the PreloadContext object