[debian-mysql] Bug#920854: Bug#920854: mariadb-10.1: FTBFS on s390x (test failure)

[Otto Kekäläinen]

> Justification: fails to build from source (but built successfully in the past)
> The mariadb-10.1 package on security.d.o doesn't build on s390x, one test fails:

Unfortunately there are no public build logs of this available at
https://buildd.debian.org/status/package.php?p=mariadb-10.1&suite=stretch
Last build log for Stretch official s390x at
https://buildd.debian.org/status/logs.php?pkg=mariadb-10.1&arch=s390x
is too old to be useful here, the test wasn't even attempted at in
10.1.23-9+deb9u1.

In https://launchpadlibrarian.net/417117661/buildlog_ubuntu-xenial-s390x.mariadb-10.1_10.1.38-0+deb9u1~ubuntu16.04.1~1553970696.2efba5470d_BUILDING.txt.gz
this is disabled:

unit.pcre_test                           [ disabled ]  MDEV-17865

It is one of the tests disabled by upstream, so it should not even run.

Upstream unstable tests have been skipped since
https://salsa.debian.org/mariadb-team/mariadb-10.1/commit/22e773ffa3eb23058543caa83b23cbc8596bb587
and this has been as is both in 10.1.26-0+deb9u1 and .38.

So it is quite strange why this test runs on s390x in stretch-security.

I will now limit the test suite to main as has been done in
mariadb-10.3 packaging, so it should take care of this issue as well.