Skip to content

Conversation

@ggouaillardet
Copy link
Contributor

No description provided.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(cherry picked from commit open-mpi/ompi@ef10d3f)
PMIx_server_register_nspace() is an asynchronous operation, so
the pmix glue wait for it completes before returning.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(back-ported from commit open-mpi/ompi@dccb189)
PMIx_server_register_nspace() is an asynchronous operation, so
the pmix glue wait for it completes before returning.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>

(back-ported from commit open-mpi/ompi@dccb189)
@ggouaillardet ggouaillardet requested a review from rhc54 January 31, 2017 01:20
@ggouaillardet ggouaillardet added this to the v2.1.0 milestone Jan 31, 2017
@ggouaillardet ggouaillardet changed the title v2.x: fix PMIx_server_register_nspace() usage v2.x: pmix: fix PMIx_server_register_nspace() usage Jan 31, 2017
@ggouaillardet
Copy link
Contributor Author

:bot:lanl:retest

@hppritcha
Copy link
Member

We have a regression with --disable-open on v2.x

@ggouaillardet
Copy link
Contributor Author

@hppritcha is it a really a regression or a jenkins issue ?
(see a previous comment of mine)
basically

  • i cannot reproduce the issue (i tried OpenSUSE Leap 42.2)
  • at the beginning of ./autogen.pl
./ompi/include/mpif-handles.h unchanged; not written
./ompi/include/mpif-constants.h unchanged; not written
./ompi/include/mpif-io-handles.h unchanged; not written
./ompi/include/mpif-io-constants.h unchanged; not written
./ompi/mpi/fortran/use-mpi-f08/constants.h unchanged; not written

that suggests the working directory is not empty when Jenkins starts building a PR

since i cannot make any sense of this error, i am thinking the root cause could be some leftover in Jenkins.

@hppritcha
Copy link
Member

bot:lanl:retest

@hppritcha hppritcha merged commit 18a0219 into open-mpi:v2.x Feb 5, 2017
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
initialized it leads to early data release and memory corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
initialized it leads to early release of the callback data and memory
corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
activated it leads to early release of the callback data and memory
corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
activated it leads to early release of the callback data and memory
corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
activated it leads to early release of the callback data and memory
corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
karasevb added a commit to karasevb/ompi that referenced this pull request Feb 15, 2017
Corresponds to PR open-mpi#2981
This bug introduced in PR open-mpi#2885. The status variable of waiting was not
activated it leads to early release of the callback data and memory
corruption.

Signed-off-by: Boris Karasev <karasev.b@gmail.com>
(cherry picked from commit 179abe4)
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.

3 participants