WAF 2.x upgrade status

[Alexander Bokovoy]

On to, 05 heinä 2018, Alexander Bokovoy via samba-technical wrote:
> On to, 05 heinä 2018, Amitay Isaacs wrote:
> > >> > I'm looking at some ways to mitigate this by setting ONNODE before
> > >> > running the tests but that doesn't help with the other failure.
> > >>
> > >> Attached patch fixes the tests.
> > >>
> > >> You can drop the patch that tries to use the installed run_tests.sh
> > >> (ctdb/wscript: use generated ctdb_run_tests).
> > > Thanks! Do you think we should keep bin in ctdb/?
> > >
> > 
> > I'm not sure what the correct thing to do is.  Did you change the bin
> > directory explicitly? Or did that change happen automatically with
> > waf-2.0?
> It is result of unification. waf 2.0 has now 'top' and 'out' to point to
> source and binary build directories. We also use 'top' to find out where
> buildtools are, so end result is that out is defined as $top/bin.
> 
> It seems to play well with lib/* libraries.
> 
> > If it's not too much trouble, then it's easier to do sub-directory
> > builds (e.g. do tdb or talloc build for testing) and discard them
> > without having to touch the top-level build.
> I'll look what could be done to preserve this use case.
I've got a green run!
https://gitlab.com/samba-team/devel/samba/pipelines/25198995

This patchset does not need additional patches for ctdb to find test
scripts as I restored previous defaults of build dir being 'bin'
everywhere.

However, the patchset still fails full autobuild on Fedora 28. The
failure I have is that ctdb test fails on F28:

*FAILED* tests/eventscripts/10.interface.multi.001.sh

while it passes in Gitlab runners (and on sn-devel, I'm sure). This
looks like a genuine issue Martin was going to look into.

I also found a bug in pidl tests which was a show-stopper for running
autobuild on F28: a number of pidl tests use ndr_push_init_ctx(NULL,
NULL) while we have removed second argument in 2010. Not sure why it is
not reproducible on Ubuntu, though, but I have samba-devel rpm package
with ndr.h headers installed on my system and this is what triggers gcc
to complain about mismatch of the signatures (and --picky-developer to
fail the test runs).

Tomorrow I'll work on fixing up commit messages and will submit the
patchset for reviews.

-- 
/ Alexander Bokovoy