Make sure the CYGWIN environment variable is set whenever system tests
are run on Windows to prevent stop.pl from making incorrect assumptions
about the environment it is running in, which triggers e.g. false
reports about named instances crashing on shutdown when system tests are
run on Windows. This issue has not been caught earlier because the
CYGWIN environment variable was incidentally being set on a higher level
in our Windows test environments.
Error reporting for parallel system tests on Windows has been broken all
along: since all parallel.mk targets generated by parallel.sh pipe their
output through "tee", the return code from run.sh is lost and thus
running "make -f parallel.mk check" will not yield a non-zero return
code if some system tests fail. The same applies to runsequential.sh.
Yet, runall.sh on Windows only sets its return code to a non-zero value
if either "make -f parallel.mk check" or runsequential.sh returns a
non-zero return code. Fix by making runall.sh yield a non-zero return
code when testsummary.sh fails, which is the same approach as the one
used in the "test" target in bin/tests/system/Makefile.
this moves the creation of "parallel.mk" into a separate shell script
instead of bin/tests/system/Makefile. that shell script can now be
executed by runall.sh, allowing us to make use of the cygwin "make"
command, which supports parallel execution.
- add CHANGES note
- update copyrights and license headers
- add -j to the make commands in .gitlab-ci.yml to take
advantage of parallelization in the gitlab CI process
This enables the environment variables controlling run.sh behavior to be
permanently set in a working environment (e.g. to automatically force
colored output without using "-c" in each runall.sh invocation).
Relevant runall.sh command line arguments still have a higher priority.
As parallel.mk and runsequential.sh both pipe system test output through
"tee" (for the purpose of creating test.output), run.sh invoked from
these two files detects it is not writing to a terminal, which prevents
colored output from being generated. Allow forcing colored output using
a new command line argument for runall.sh, "-c", which sets an
environment variable (SYSTEMTEST_FORCE_COLOR) causing conf.sh to
unconditionally enable colored output.
The same environment variable can also be used directly to force colored
output when using "make test" instead of runall.sh.
Instead of exporting an environment variable containing a command line
argument (NOCLEAN="-n"), extend run.sh to handle a "boolean" environment
variable (SYSTEMTEST_NO_CLEAN) itself. The former method is buggy
because the value of NOCLEAN is set in parallel.mk when that file is
first created, but it is not subsequently updated upon each test run
(because make considers parallel.mk to be up to date).
To retain backward compatibility, the "-n" command line argument for
run.sh is still supported (and has a higher priority than the relevant
environment variable).
The SYSTEMTEST_NO_CLEAN environment variable can also be used directly
to prevent cleanup when using "make test" instead of runall.sh.
Apart from fixing a bug, this simplifies the way runall.sh controls
run.sh behavior due to the Makefile being bypassed. Direct processing
of environment variables in run.sh is more scalable in the long run,
given that the previously utilized technique, even with its
implementation fixed, would still require Makefile.in to be modified in
two places each time a new flag needed to be passed from runall.sh to
run.sh.
1. Track changes to conf.sh.in in conf.sh.win32
2. Modifications to prevent Windows "Configure" script replacing
the sed "@PORT@" substitution tokens in conf.sh.win32.
3. runall.sh now runs Windows tests sequentially
When running all the system tests, output from a test is sent to a
test.output file in the test directory. These are combined in to
systests.output when the run finishes.
Miscellaneous tidying up of run management. The most significant
change is that "runall.sh" now runs _all_ the tests, even the
ones that can run in parallel. runsequential.sh is the script
to run tests that have not been converted to parallel running.
Via an intermediate make file, tests that have been modified to be able
to run in parallel are assigned unique query and control port numbers
(other than 5300 and 9953 respectively). Tests that have not yet been
modified all use ports 5300 and 9953, so must be run sequentially.
