make sure we don't delete files when -K is specified

fixes #7008

we need to run the function for runpart in the same shell, otherwise we
will lose the global variables set in the split functions.

fix so abuild -A does not exit with error if gitbase fails

this was introduced with the set -e

- versioned dependencies
- dependencies on 'provides' tags or library names

We chose to use kernel style HOST* instead of the gcc's
*_FOR_HOST.

Just checking the build, host and target triplet is not enough
due to various different combinations used in the early bootstrap.
So use detect specifical bootstrap setting.

this fixes problem with abump

Allowing bootstrap.sh to inject extra dependencies for build/host/target,
simplifies things greatly in the bootstrap script.

we need to replace noarch with CARCH for now

needed for upcoming 'rootbld' support

The check() function is an APKBUILD overridable function which runs a testsuite.  The packages listed in
$checkdepends are packages which will be installed at build time only if check() will be run.

for patches like:
patchname.patch::http://github/.../commit/<md5hash>.patch

use strict filename instead of the whole line

use only the strongest hash. this simplifies apkbuilds a bit.

Currently, if aarch64 exists in config.guess, it is not updated.
This breaks spl, which has aarch64 entry, but not ppc64le.

update_config_guess should update config.guess if any of those
does not exists.

This is the initial patches that starts to add support for the ppc64le
platform.

required tool for re-signing packages

this is where it is built too

so we can crosscompile for different architectures

So far default_prepare() ended up in "$builddir" only if there were any
patches.  It's better to be more consistent and make such change always.

This script uses variable expansion / pattern replacement, which is not
defined in POSIX-sh. Also APKBUILDs are not required to be strictly
POSIX-sh compatible. If someone run abuild on system with e.g. dash as
/bin/sh, then it fails (and someone reported exactly that on IRC
today). Therefore abuild should explicitly use /bin/ash in shebang and
not /bin/sh.