[GHC] #9221: (super!) linear slowdown of parallel builds on 40 core machine

[GHC]

#9221: (super!) linear slowdown of parallel builds on 40 core machine
-------------------------------------+-------------------------------------
              Reporter:  carter      |            Owner:
                  Type:  bug         |           Status:  new
              Priority:  high        |        Milestone:  7.10.1
             Component:  Compiler    |          Version:  7.8.2
            Resolution:              |         Keywords:
      Operating System:              |     Architecture:  Unknown/Multiple
  Unknown/Multiple                   |       Difficulty:  Unknown
       Type of failure:  Compile-    |       Blocked By:
  time performance bug               |  Related Tickets:  #910
             Test Case:              |
              Blocking:              |
Differential Revisions:              |
-------------------------------------+-------------------------------------

Comment (by gintas):

 While looking around I found #3758, and it appears that +RTS -qg (disable
 parallel GC) helps a lot with the superlinear overhead. For example, the
 benchmark above with jobs=24 & caps=24 without -qg took:

 real    1m3.596s
 user    6m31.072s
 sys     3m10.732s

 With -qg:

 real    0m47.747s
 user    1m33.352s
 sys     0m2.024s

 However, for smaller -j values -qg slightly increases walltime:
  2 jobs: 44s without -qg, 46s with -qg
  4 jobs: 37s vs 40s
  8 jobs: 37s vs 41s
  15 jobs: 42s vs 44s
  16 jobs: 49s vs 44s (walltime crossover point for this 16-core machine)

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/9221#comment:10>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler