Skip to content

Conversation

@ggouaillardet
Copy link
Contributor

@ggouaillardet ggouaillardet commented Apr 30, 2021

bot:notacherrypick

@ggouaillardet ggouaillardet added this to the v4.1.2 milestone Apr 30, 2021
@ggouaillardet ggouaillardet requested a review from bosilca April 30, 2021 04:00
@ggouaillardet ggouaillardet force-pushed the topic/v4.1.x/sendrecv_replace_big branch from bd2a35e to f1099d1 Compare April 30, 2021 04:06
@ggouaillardet
Copy link
Contributor Author

:bot:aws:retest

@ggouaillardet
Copy link
Contributor Author

@jsquyres could you please verify the bot?

this PR is made of two commits:

  • the first one is a backport
  • the second one is a straight cherry-pick

The first one contains the following messages

   (back-ported from commit 6a11873ef52191962b42ba8f56e3630e266222c7)
    bot:notacherrypick

the bot complained of a missing cherry pick message and did not notice the bot:notacherrypick directive (it even suggested me to add one !)

I would suggest to add the new bot:backport directive so we can make the distinction between backports (e.g. cherry-pick requiring manual action) and one-off commits for the release branches (because master do it differently).
A valid commit in master could even be checked when the bot:backport directive is used.

@bwbarrett
Copy link
Member

bwbarrett commented Apr 30, 2021

The bot:notacherrypick goes in the PR body, not in the git commit message. We are opinionated that bot commands don’t belong in git history.

I don’t get your back port comment. The bot doesn’t check that it was a clean cherry pick or that you changed things. Only that it is either a PR marked as not a cherry commit or there’s a cherry-commit line with a reference to a valid hash.

@ggouaillardet
Copy link
Contributor Author

Thanks @bwbarrett for the insight, I will update the PR accordingly.

Don't the bot:notacherrypick directive ends up in the merge commit (and hence git history)? (I have no intention to argue with this though)

My point is that, regardless the bot, I like to make the distinction between clean cherry pick and backports.
In the former case, the git message contains a cherry picked from commit ... line.
With the latter, after I resolve a conflict, I used to manually replace cherry picked with back-ported.

In the second case, I understand we have several options:

  • leave the cherry picked from commit message: bot is happy, I am not (because I like to make it clear this is a backport and not a straightforward cherry pick)
  • make a back-ported from commit message and a bot:notacherrypick message in the PR. This is stil suboptimal to me since
    • the bot does not check the back-port is from a valid commit from master
    • in the case of a PR containing multiple commits, the bot does not check the clean/straightforward commits were cherry-picked from a valid commit.
      FWIW, if this is a one off commit, I used to manually add a this is a one-off commit for the release branch
    • leave the cherry picked from commit message to keep the bot happy, and manually add a backport was required like message: in my highly subjective opinion, this does not look pretty.

Because MPI_Sendrecv_replace() uses PMPI_Sendrecv() with MPI_PACKED
under the hood, the data to be exchanged size = MPI_Type_size(datatype) * count
must fit in a signed integer.
Otherwise, PMPI_Sendrecv()
 - fails with an error message if (int)size < 0
 - silently truncate the data if (int)size >= 0

Refs. open-mpi#8862

Thanks Jakub Benda for reporting this issue and suggesting a fix.

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

(back-ported from commit 6a11873)
and fix a comment

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
(cherry picked from commit 0b38190)
@ggouaillardet ggouaillardet force-pushed the topic/v4.1.x/sendrecv_replace_big branch from f1099d1 to 19c30a0 Compare April 30, 2021 04:49
do ignore the status returned by opal_convertor_pack()
since it is *not* MPI_SUCCESS in case of success.

This fixes a regression introduced in open-mpi/ompi@0b38190

Refs. open-mpi#8907

Thanks Lisandro Dalcin for reporting this issue.

Signed-off-by: Gilles Gouaillardet <gilles@rist.or.jp>
(cherry picked from commit 34b764c)
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.

4 participants