Project

General

Profile

Bug #1790

+oneWthPerHost not implemented in standalone builds

Added by Sam White 5 months ago. Updated 4 months ago.

Status:
Merged
Priority:
Normal
Assignee:
Category:
-
Target version:
Start date:
02/01/2018
Due date:
% Done:

0%


Description

$ ./charmrun ./jacobi 2 2 2 40 +vp8 +balancer RotateLB +LBDebug 1 +oneWthPerHost
Running command: ./jacobi 2 2 2 40 +vp8 +balancer RotateLB +LBDebug 1 +oneWthPerHost

Charm++> No provisioning arguments specified. Running with a single PE.
         Use +auto-provision to fully subscribe resources or +p1 to silence this message.
Charm++: standalone mode (not using charmrun)
Charm++> Running in Multicore mode:  1 threads
Converse/Charm++ Commit ID: v6.8.0-423-g3aa5cfe08
CharmLB> Verbose level 1, load balancing period: 0.5 seconds
CharmLB> Load balancer assumes all CPUs are same.
Charm++> Running on 1 unique compute nodes (8-way SMP).
Charm++> cpu topology info is gathered in 0.001 seconds.
WARNING: +oneWthPerHost is a command line argument beginning with a '+' but was not parsed by the RTS.
If any of the above arguments were intended for the RTS you may need to recompile Charm++ with different options.
[0] RotateLB created
iter 1 time: 0.100631 maxerr: 2020.200000
iter 2 time: 0.102280 maxerr: 1696.968000
iter 3 time: 0.098243 maxerr: 1477.170240
iter 4 time: 0.098893 maxerr: 1319.433024
iter 5 time: 0.101593 maxerr: 1200.918072
iter 6 time: 0.101801 maxerr: 1108.425519
iter 7 time: 0.105620 maxerr: 1033.970839
iter 8 time: 0.107805 maxerr: 972.509242
iter 9 time: 0.109841 maxerr: 920.721889
iter 10 time: 0.107849 maxerr: 876.344030
iter 11 time: 0.106372 maxerr: 837.779089
iter 12 time: 0.110128 maxerr: 803.868831
iter 13 time: 0.112698 maxerr: 773.751705
iter 14 time: 0.104304 maxerr: 746.772667
iter 15 time: 0.103350 maxerr: 722.424056
iter 16 time: 0.116914 maxerr: 700.305763
iter 17 time: 0.109865 maxerr: 680.097726
iter 18 time: 0.107713 maxerr: 661.540528
iter 19 time: 0.101935 maxerr: 644.421422
iter 20 time: 0.101064 maxerr: 628.564089
[Partition 0][Node 0] End of program

All of the other +oneWthPer* options work.
Running with or without charmrun doesn't matter.

History

#1 Updated by Evan Ramos 5 months ago

I specifically did not add oneWthPerHost to standalone builds (which are all multicore has) because host count is not relevant to a single machine--it is always one.

#2 Updated by Sam White 5 months ago

Yes, but we should accept that argument since it does make sense, and because we should always parse and consume all of these options. If the options don't make sense, we can throw a warning or error, but we shouldn't let the argument through to the user's arguments.

#3 Updated by Evan Ramos 5 months ago

  • Status changed from New to Implemented
  • Subject changed from +oneWthPerHost not recognized by multicore-darwin-x86_64 to +oneWthPerHost not implemented in standalone builds

#4 Updated by Evan Ramos 4 months ago

  • Status changed from Implemented to Merged

Also available in: Atom PDF