[GHC] #8224: Excessive system time -- new IO manager problem?

GHC ghc-devs at haskell.org
Sat Nov 12 22:14:14 UTC 2016


#8224: Excessive system time -- new IO manager problem?
-------------------------------------+-------------------------------------
        Reporter:  rrnewton          |                Owner:
            Type:  bug               |               Status:  new
        Priority:  high              |            Milestone:  8.2.1
       Component:  Runtime System    |              Version:  7.7
      Resolution:                    |             Keywords:  IO Manager,
                                     |  System Time
Operating System:  Linux             |         Architecture:  x86_64
 Type of failure:  Runtime           |  (amd64)
  performance bug                    |            Test Case:
      Blocked By:                    |             Blocking:
 Related Tickets:  #9221             |  Differential Rev(s):
       Wiki Page:                    |
-------------------------------------+-------------------------------------

Comment (by bgamari):

 Replying to [comment:34 alkar]:
 > A real-life example of a program where I'd want a long single thread
 computation with `-N` turned on is a faster analog of `fgrep -f
 patterns.txt input1 ... inputN`:
 >
 > 1. Read all lines from `patterns.txt` into a set.
 > 2. In parallel for each input file, `filter (\str -> member str set) .
 lines`

 I'm afraid I don't understand; this would be a multi-threaded computation,
 would it not?

 In general there are very real costs to synchronization, especially across
 cores on large machines. The performance hit that you report above sounds
 a bit high and it would be nice to bring it down. Perhaps you could do
 some digging to see where this time is going?

--
Ticket URL: <http://ghc.haskell.org/trac/ghc/ticket/8224#comment:35>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler


More information about the ghc-tickets mailing list