MTL/OFI: Check threshold number of peers allowed per rank

[wckzhang]

When the provider does not support FI_REMOTE_CQ_DATA, the OFI tag does not have
sizeof(int) bits for the rank. Therefore, unexpected behavior will occur when
this limit is crossed.

Check the max allowed number of ranks during add_procs() and return if there is
danger of exceeding this threshold.

Signed-off-by: Aravind Gopalakrishnan Aravind.Gopalakrishnan@intel.com

[wckzhang]

ompi_mtl_ofi_add_procs returning failure would cause a segfault. Cherry-picked 5cf43de to add check for the segfault.

[wckzhang]

Cray seems broken, can anybody check it out?

[jsquyres]

Cray sometimes just times out. 😦

bot:ompi:retest

[bwbarrett]

@wckzhang this looks like a cherry pick from master? if so, you should add the -x argument to the cherry-pick command line, so that it adds the “cherry-pick from” message in the commit message.

[wckzhang]

@wckzhang this looks like a cherry pick from master? if so, you should add the -x argument to the cherry-pick command line, so that it adds the “cherry-pick from” message in the commit message.

Done, thanks.

[ibm-ompi]

The IBM CI (XL Compiler) build failed! Please review the log, linked below.

Gist: https://gist.github.com/ibm-ompi/88b480fb98dfd2bdac939f023e1f8fe6

[ibm-ompi]

The IBM CI (GNU Compiler) build failed! Please review the log, linked below.

Gist: https://gist.github.com/ibm-ompi/88b480fb98dfd2bdac939f023e1f8fe6

[wckzhang]

IBM CI failed, can someone check that?

[jjhursey]

bot:ibm:gnu:retest
Somehow the prrte PR mixed with this PR. I think I figured out what happened. Let's see if this works.

[jjhursey]

bot:ibm:retest

@wckzhang, recherry-picked with -x

[gpaulsen]

Thanks.