Bug in work-stealing dequeu (possible memory write access reordering in pushWSDeque)

[David Peixotto]

If you run

$ ./sync-all --extra get

It will grab the parallel, deepseq, and stm packages for you.

On Jul 12, 2011, at 11:14 AM, Karel Gardas wrote:

> On 07/11/11 01:10 PM, Simon Marlow wrote:
>> On 11/07/2011 11:36, Karel Gardas wrote:
>> 
>>>> Have you tried running some heavy -N2 tests? e.g. try the benchmarks in
>>>> nofib/parallel.
>>> 
>>> If this is part of normal nofib's `make' then this runs already.
>> 
>> No, it's not part of a normal nofib run. You want something like this:
>> 
>> $ cd nofib/parallel
>> $ make boot
> 
> Unfortunately this fails with:
> 
> ------------------------------------------------------------------------
> == make boot ;
> in /home/kgardas/src/github.com-ghc-arm/ghc-haskell.org/nofib/parallel/parfib
> ------------------------------------------------------------------------
> /home/kgardas/src/github.com-ghc-arm/ghc-haskell.org/inplace/bin/ghc-stage2   -M -dep-makefile .depend  -osuf o     -H32m -O -O -Rghc-timing -package array -H32m -hisuf hi -package parallel -rtsopts Main.hs
> <command line>: cannot satisfy -package parallel
>    (use -v for more information)
> <<ghc: 6797280 bytes, 2 GCs, 69312/69312 avg/max bytes residency (1 samples), 31M in use, 0.00 INIT (0.01 elapsed), 0.03 MUT (0.07 elapsed), 0.03 GC (0.07 elapsed) :ghc>>
> make[1]: *** [depend] Error 1
> Failed making boot in parfib: 1
> make: *** [boot] Error 1
> 
> 
> And inplace package list looks like:
> /home/kgardas/src/github.com-ghc-arm/ghc-haskell.org/inplace/lib/package.conf.d:
>    Cabal-1.11.2
>    array-0.3.0.3
>    base-4.4.0.0
>    bin-package-db-0.0.0.0
>    binary-0.5.0.2
>    bytestring-0.9.2.0
>    containers-0.4.0.0
>    directory-1.1.0.1
>    dph-base-0.5
>    (dph-par-0.5)
>    (dph-prim-interface-0.5)
>    (dph-prim-par-0.5)
>    (dph-prim-seq-0.5)
>    (dph-seq-0.5)
>    extensible-exceptions-0.1.1.3
>    ffi-1.0
>    filepath-1.2.0.1
>    (ghc-7.3.20110711)
>    ghc-prim-0.2.0.0
>    haskeline-0.6.4.0
>    (haskell2010-1.1.0.0)
>    (haskell98-2.0.0.0)
>    hoopl-3.8.7.0
>    hpc-0.5.1.0
>    integer-gmp-0.3.0.0
>    mtl-1.1.1.1
>    old-locale-1.0.0.3
>    old-time-1.0.0.7
>    pretty-1.1.0.0
>    primitive-0.3.1
>    process-1.1.0.0
>    random-1.0.0.4
>    rts-1.0
>    template-haskell-2.6.0.0
>    terminfo-0.3.1.3
>    time-1.2.0.5
>    unix-2.5.0.0
>    utf8-string-0.3.7
>    vector-0.7.2
>    xhtml-3000.2.0.1
> 
> 
> Do you have any idea how to fix this?
> 
> Thanks,
> Karel
> 
> _______________________________________________
> Cvs-ghc mailing list
> Cvs-ghc at haskell.org
> http://www.haskell.org/mailman/listinfo/cvs-ghc
>