RE: ghc/rts/parallel/HLComms.c

2001-02-19 Thread Simon Marlow 

 Albert Lai [EMAIL PROTECTED] writes:
 
  I try to build ghc 4.08.2 with parallelism turned on, i.e., 
 GhcLibsWays=mp.
  At ghc/rts/parallel/HLComms.c line 550:
  
GarbageCollect(GetRoots);
  
  it barfs because the new GarbageCollect() function expects a boolean
  parameter after the first parameter.  (See 
 ghc/rts/Storage.h for example.)
 
 For now, I will just steal the code from gum 4.06 and replace this
 (and all other occurences in HLComms.c) by
 
   GarbageCollect(GetRoots, rtsFalse);
 
 and hope it works.

We don't maintain the GUM code in GHC, the GPH folks maintain it
separately, and as far as I know it is still based on GHC 4.06 (although
last I heard there were plans to update it to the current version).

The main GPH site is:

http://www.cee.hw.ac.uk/~dsg/gph/

Cheers,
Simon

___
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

ghc/rts/parallel/HLComms.c

2001-02-16 Thread Albert Lai 

I try to build ghc 4.08.2 with parallelism turned on, i.e., GhcLibsWays=mp.
At ghc/rts/parallel/HLComms.c line 550:

  GarbageCollect(GetRoots);

it barfs because the new GarbageCollect() function expects a boolean
parameter after the first parameter.  (See ghc/rts/Storage.h for example.)

Is this why I should use gum 4.06 instead? :)

I would really like to have just one installation of ghc for every
way I use it, not one for sequential execution and one for parallel
execution.  One installation is already faring at 90MB; having two
such beasts on a shared disk at school is not nice...

___
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs

Re: ghc/rts/parallel/HLComms.c

2001-02-16 Thread Albert Lai 

Albert Lai [EMAIL PROTECTED] writes:

 I try to build ghc 4.08.2 with parallelism turned on, i.e., GhcLibsWays=mp.
 At ghc/rts/parallel/HLComms.c line 550:
 
   GarbageCollect(GetRoots);
 
 it barfs because the new GarbageCollect() function expects a boolean
 parameter after the first parameter.  (See ghc/rts/Storage.h for example.)

For now, I will just steal the code from gum 4.06 and replace this
(and all other occurences in HLComms.c) by

  GarbageCollect(GetRoots, rtsFalse);

and hope it works.

___
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs