jobspec: replace minus operator to multiplication

[dongahn]

Make the implementation come to compliance with RFC14's JRC rule section.
Tested with resource-query and this one-liner seems enough.

[coveralls]

Coverage increased (+0.004%) to 78.614% when pulling 036c631 on dongahn:rfc_compliance into a2b1063 on flux-framework:master.

[codecov-io]

Codecov Report

Merging #1293 into master will decrease coverage by 0.01%.
The diff coverage is n/a.

@@            Coverage Diff             @@
##           master    #1293      +/-   ##
==========================================
- Coverage   78.02%   78.01%   -0.02%     
==========================================
  Files         154      154              
  Lines       29104    29104              
==========================================
- Hits        22709    22706       -3     
- Misses       6395     6398       +3

Impacted Files	Coverage Δ
src/common/libflux/keepalive.c	86.66% <0%> (-6.67%)	⬇️
src/broker/modservice.c	79.61% <0%> (-1.95%)	⬇️
src/modules/connector-local/local.c	76.11% <0%> (-1.31%)	⬇️
src/common/libflux/response.c	83.73% <0%> (-0.82%)	⬇️
src/common/libflux/handle.c	83.66% <0%> (-0.5%)	⬇️
src/common/libflux/future.c	88.31% <0%> (-0.47%)	⬇️
src/common/libflux/message.c	81.36% <0%> (+0.23%)	⬆️
src/common/libkvs/kvs.c	66.54% <0%> (+0.24%)	⬆️
src/modules/kvs/kvs.c	64.95% <0%> (+0.25%)	⬆️
src/common/libkvs/treeobj.c	85.54% <0%> (+0.4%)	⬆️
... and 3 more

[dongahn]

Hmmm. I don't understand why this one line change causes a valgrind-related error....

https://travis-ci.org/flux-framework/flux-core/jobs/306163007#L9029

[garlick]

Looks like an unrelated transient error in czmq shutdown:

==10045== Memcheck, a memory error detector
==10045== Copyright (C) 2002-2013, and GNU GPL'd, by Julian Seward et al.
==10045== Using Valgrind-3.10.1 and LibVEX; rerun with -h for copyright info
==10045== Command: /home/travis/build/flux-framework/flux-core/flux-core-0.8.0-465-g95b20d2f/_build/sub/src/broker/.libs/lt-flux-broker --shutdown-grace=4 /home/travis/build/flux-framework/flux-core/flux-core-0.8.0-465-g95b20d2f/t/valgrind/valgrind-workload.sh 10
==10045== 
FLUX_URI=local:///tmp/flux-5SBsSk
==10045== 
==10045== HEAP SUMMARY:
==10045==     in use at exit: 111,117 bytes in 102 blocks
==10045==   total heap usage: 912,236 allocs, 912,134 frees, 78,583,150 bytes allocated
==10045== 
==10045== 288 bytes in 1 blocks are possibly lost in loss record 27 of 39
==10045==    at 0x4C2CC70: calloc (in /usr/lib/valgrind/vgpreload_memcheck-amd64-linux.so)
==10045==    by 0x4012FE4: allocate_dtv (dl-tls.c:296)
==10045==    by 0x4012FE4: _dl_allocate_tls (dl-tls.c:460)
==10045==    by 0x5E3ED92: allocate_stack (allocatestack.c:589)
==10045==    by 0x5E3ED92: pthread_create@@GLIBC_2.2.5 (pthread_create.c:500)
==10045==    by 0x54CD8DA: zactor_new (zactor.c:122)
==10045==    by 0x4E4826A: flux_sec_comms_init (security.c:197)
==10045==    by 0x407361: main (broker.c:411)
==10045== 
{
   <insert_a_suppression_name_here>
   Memcheck:Leak
   match-leak-kinds: possible
   fun:calloc
   fun:allocate_dtv
   fun:_dl_allocate_tls
   fun:allocate_stack
   fun:pthread_create@@GLIBC_2.2.5
   fun:zactor_new
   fun:flux_sec_comms_init
   fun:main
}
==10045== LEAK SUMMARY:
==10045==    definitely lost: 0 bytes in 0 blocks
==10045==    indirectly lost: 0 bytes in 0 blocks
==10045==      possibly lost: 288 bytes in 1 blocks
==10045==    still reachable: 110,317 bytes in 100 blocks
==10045==         suppressed: 512 bytes in 1 blocks
==10045== Reachable blocks (those to which a pointer was found) are not shown.
==10045== To see them, rerun with: --leak-check=full --show-leak-kinds=all
==10045== 
==10045== For counts of detected and suppressed errors, rerun with: -v
==10045== ERROR SUMMARY: 1 errors from 1 contexts (suppressed: 1 from 1)
not ok 1 - valgrind reports no new errors on single broker run

[garlick]

@dongahn: it would be nice to have a more descriptive commit message, for those of us who have not been following the jobspec design in detail.

jobspec: replace minus operator to multiplication

Did you mean "replace minus operator with multiplcation"?

Make the implementation come to compliance with RFC14's JRC rule section.
Tested with resource-query and this one-liner seems enough.

What specifically is out of compliance? Perhaps a quote from the spec would be useful here.

[dongahn]

@garlick: OK, that makes sense. I forced a push. Hopefully this works.

[garlick]

Much better, thank you!

Looks like you may need to rebase on current master?

We probably need @morrone or @grondo to give an ACK before merging.

[coveralls]

Coverage decreased (-0.01%) to 78.597% when pulling ad747cf on dongahn:rfc_compliance into a2b1063 on flux-framework:master.

[garlick]

Looks like you may need to rebase on current master?

Strike that - github was saying it was out of date, but not now.

[grondo]

Nice explanation @dongahn, makes sense to me! (Was likely just a typo in the orignal)

My only suggestion is maybe we should add a test case to t/jobspec/invalid to ensure a jobspec input using the invalid operator - is rejected?

[dongahn]

ok. should be easy enough.