daemontools: remove native BBCLASSEXTEND

* it was used only to provide chkshsgr which is now replaced with no-op call
  since
  commit 50d526d06a742fa69ff698d7c2eefffb56e13afa
  Author: Khem Raj <raj.khem@gmail.com>
  Date:   Tue Jan 28 11:28:52 2020 -0800

    daemontools: Disable the chkshsgr tests

    Running the chkhsgr test during cross compile fails

    ./chkshsgr || ( cat warn-shsgr; exit 1 )
    Oops. Your getgroups() returned 0, and setgroups() failed; this means
    that I can't reliably do my shsgr test. Please either ``make'' as root
    or ``make'' while you're in one or more supplementary groups.

    All OE based targets have working getgroups()/setgroups()
    implementation, so its a safe assumption and therefore make the test
    to be a dummy

* the native chkshsgr from daemontools-native was actually being
  called only because of this chunk of cross-compile.patch:
  -       ./chkshsgr || ( cat warn-shsgr; exit 1 )
  +       chkshsgr || ( cat warn-shsgr; exit 1 )
  but all chkshsgr does is:
    short x[4];
    x[0] = x[1] = 0;
    if (getgroups(1,x) == 0) if (setgroups(1,x) == -1) _exit(1);
    _exit(0);
  which running on host system, doesn't say anything useful about the
  cross compile target, so it's easier to just remove the call in
  cross-compile.patch and simplify all this nonsense

* I came across this because daemontools-native was failing for me
  in "bitbake world" with zeus, which might be the same case as what
  Khem was seeing - just the final commit message doesn't reflect that

* daemontools-native fails to build without the above commit in
  zeus as well, when building inside docker container where my
  build user is in fewer groups (just 1) so the
  getgroups(1,x) call doesn't fail, but on more average OS the
  user will be in more than 4 groups and
  getgroups(1,x) would fail with errno 22 EINVAL
    - so setgroups isn't even called to return 1 error when chkshsgr is called
  http://man7.org/linux/man-pages/man2/setgroups.2.html
    If the calling process is a member of more than size
    supplementary groups, then an error results.
  if I increase the size of x enough for x to hold all groups, then setgroups
  will fail with errno 1 EPERM, which is the same error as shown
  under docker container where getgroups doesn't fail, because
  in both cases I'm using unprivileged user for builds

Signed-off-by: Martin Jansa <Martin.Jansa@gmail.com>
Signed-off-by: Khem Raj <raj.khem@gmail.com>

0 files changed, 0 insertions, 0 deletions