On Thu, Feb 19, 2009 at 10:52:28AM +0000, Simon Marlow wrote:
> Ian Lynagh wrote:
> >Wed Feb 18 15:56:20 PST 2009 Ian Lynagh <[email protected]>
> > * Fix an off-by-one; fixes the second bug in trac #3001
> >
> > M ./rts/LdvProfile.h -1 +1
>
> Nice catch - though the off-by-one is actually at the call site (I think I
> broke this in the parallel GC code). Not a big deal though.
This function seems really odd to me, actually. Why do we call
LDV_FILL_SLOP(to + size_to_reserve, (int)(size_to_reserve - size_to_copy))
and then fill backwards from
to + size_to_reserve - 1
to
to + size_to_copy
inclusive, rather than calling
LDV_FILL_SLOP(to + size_to_copy, (int)(size_to_reserve - size_to_copy))
and then filling forwards from
to + size_to_copy
to
to + size_to_reserve - 1
?
Thanks
Ian
_______________________________________________
Cvs-ghc mailing list
[email protected]
http://www.haskell.org/mailman/listinfo/cvs-ghc
