Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

v3.1.x: Add "make check" support #1205

Merged
merged 11 commits into from
Apr 15, 2019
Merged

v3.1.x: Add "make check" support #1205

merged 11 commits into from
Apr 15, 2019

Conversation

rhc54
Copy link
Contributor

@rhc54 rhc54 commented Apr 12, 2019

No description provided.

rhc54 and others added 6 commits April 11, 2019 20:06
Add a script that runs the Jenkins tests and returns a non-zero status
if any of them fail.

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 2d07a6c)
Scan the output of the Jenkins run to get the list of tests

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 1f58246)
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 2bce385)
Required for "make check" to succeed

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 2122ec2)
Per chat with @jsquyres, use autogen.pl to assemble a file containing
the path names of all components in the code base, then let configure
AC_SUBST that string into the run_tests.pl for us. We then let Perl push
the result into the environment as an MCA parameter telling the system
where to find all the component libraries.

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 5b3ae33)
When pmix_mca_base_component_path is processed, it is not an error
(i.e., we should not abort in the middle) if one of the directories
fails to add.  Instead, we should warn the user that the directory
failed to add, but then keep processing all other directories listed
in the path.

Signed-off-by: Jeff Squyres <jsquyres@cisco.com>
(cherry picked from commit 1019712)
@rhc54 rhc54 added this to the v3.1.3 milestone Apr 12, 2019
@rhc54 rhc54 requested a review from jjhursey April 12, 2019 03:20
@rhc54 rhc54 self-assigned this Apr 12, 2019
bwbarrett and others added 5 commits April 12, 2019 17:26
"make check" can only succeed if the plugins have been installed or
"--disable-dlopen" has been passed to configure.  The first can't
happen in "make distcheck", so force the second for all invocations
of "make distcheck" by setting AM_DISTCHECK_CONFIGURE_FLAGS in
Makefile.am.

Clean up the handling of AM_DISTCHECK_CONFIGURE_FLAGS and
DISTCHECK_CONFIGURE_FLAGS in make_dist_tarball.  First, with the
above fix, passing --disable-dlopen is no longer required.  Second,
users (and scripts) should set DISTCHECK_CONFIGURE_FLAGS, not
AM_DISTCHECK_CONFIGURE_FLAGS.  Automake will combine the two options
when running "make distcheck", so the right things will happen with
this configuration.

Signed-off-by: Brian Barrett <bbarrett@amazon.com>
(cherry picked from commit e1b4611)
Since we are no longer modifying AM_DISTCHECK_CONFIGURE_FLAGS, we don't
want to add config_flags by itself to the distcheck_flags as those are
being passed to "make" and not "configure"

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 84d5189)
Two errors were found. First, the return code when data isn't found can
be either PMIX_ERR_NOT_FOUND or PMIX_ERR_PROC_ENTRY_NOT_FOUND - there is
no guarantee which might be returned. Only guarantee is that it will not
be PMIX_SUCCESS. Thus, the test has to be a little more flexible when
checking for error codes.

Second, when multiple servers are trying to mimic a multi-node setup,
each server is required to provide a proc map of the entire job - not
just the clients local to it.

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 96c8765)
If the system or user default component repository path doesn't exist,
then don't bark about it - there is no requirement that these exist.

Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit 4035bd4)
Signed-off-by: Ralph Castain <rhc@pmix.org>
(cherry picked from commit e49b100)
@rhc54 rhc54 merged commit 3160c22 into openpmix:v3.1 Apr 15, 2019
@rhc54 rhc54 deleted the cmr31/test branch April 22, 2019 09:55
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

4 participants