fix explanation of in-place option for MPI_Allgather(v)

[jedbrown]

These man pages (for MPI_Allgather and MPI_Allgatherv when using MPI_IN_PLACE) were wrong (intended to call MPI_Gather(v)). The example is a bit confusing because recvbuf is modified in a loop of repeat calls, but I think the loop is still correct because the subset of recvbuf that acts as send area when using MPI_IN_PLACE should be invariant. I don't know if you want to keep this explanation; it certainly can't be used to explain sibling collectives that support MPI_IN_PLACE.

(Apologies for the incorrect explanation prior to edit.)

[github-actions]

Hello! The Git Commit Checker CI bot found a few problems with this PR:

f40711c: WIP: fix explanation of in-place option

• check_signed_off: does not contain a valid Signed-off-by line

Please fix these problems and, if necessary, force-push new commits back up to the PR branch. Thanks!

[github-actions]

Hello! The Git Commit Checker CI bot found a few problems with this PR:

c80a18d: doc: fix explanation of in-place option for MPI_Al...

• check_signed_off: does not contain a valid Signed-off-by line

Please fix these problems and, if necessary, force-push new commits back up to the PR branch. Thanks!

[github-actions]

Hello! The Git Commit Checker CI bot found a few problems with this PR:

c80a18d: doc: fix explanation of in-place option for MPI_Al...

• check_signed_off: does not contain a valid Signed-off-by line

Please fix these problems and, if necessary, force-push new commits back up to the PR branch. Thanks!

[jsquyres]

Woof. I pressed all the wrong buttons on this PR.

@bosilca Can you review?