[GHC] #7254: Parallel processing and the -Nn flag

GHC cvs-ghc at haskell.org
Wed Sep 19 19:43:17 CEST 2012


#7254: Parallel processing and the -Nn flag
------------------------------+---------------------------------------------
 Reporter:  mcandre           |          Owner:                  
     Type:  feature request   |         Status:  new             
 Priority:  normal            |      Component:  Runtime System  
  Version:  7.0.3             |       Keywords:                  
       Os:  Unknown/Multiple  |   Architecture:  Unknown/Multiple
  Failure:  Runtime crash     |       Testcase:                  
Blockedby:                    |       Blocking:                  
  Related:                    |  
------------------------------+---------------------------------------------
 When I follow the Parallel Processing with Haskell tutorial
 ([http://www.yellosoft.us/parallel-processing-with-haskell]), it instructs
 users to run Haskell parallel programs with syntax like:

 {{{
 $ ./maths +RTS -N2
 }}}

 Where the -Nn flag informs the Haskell system that the computer has n
 cores. Modern video games are able to automatically determine the number
 of cores on a computer by some method of extracting such details; there's
 no reason Haskell can't do the same, so that the user doesn't have to
 specify it manually.

 This is important because specifying the wrong number of cores either
 leads the program to run slowly, or even crash. If Haskell detected the
 number of cores and the user still specified a manual value, it could
 correct the user's mistake in much the same way that the Haskell type
 system does inference and corrects the user's mistake.

 Specs at time of this behavior:

  * GHC 7.0.3
  * Haskell Platform 2011.2.0.1
  * Windows 7 Professional x86 and x64
  * Ubuntu 12.04 x86 and x64
  * Mac OS X 10.8.1 Mountain Lion (triple booting with rEFIt)
  * !MacBook Pro 2009

-- 
Ticket URL: <http://hackage.haskell.org/trac/ghc/ticket/7254>
GHC <http://www.haskell.org/ghc/>
The Glasgow Haskell Compiler



More information about the Glasgow-haskell-bugs mailing list