Hello skaller, Sunday, August 13, 2006, 4:34:14 AM, you wrote:
I know very little about Haskell, let alone GHC internals
me too. so it's better to wait for comments about your thoughts from GHC team than from me. but at least i can said that
But the state of the art is then two stages behind the requirement: Haskell still has to 'world stop' threads to do a major collection.
is not exactly true. look at "Non-stop Haskell" (http://www.haskell.org/~simonmar/papers/nonstop.pdf) i don't know why it is not included in 6.6 or previous version
So I'm bringing into question whether these nice 'optimisations' are actually worthwhile. They actually seem to *degrade* performance, not improve it, when we're running with a large number of CPUs. Stopping the world if you have 20,000 CPU's will happen so often, all the CPU's will be idle 99.99% of the time :)
btw, one GHC intern worked on multi-processor GC and i hope that it will be included in 6.6. so, the GC will also use all these 20k cpus :) or Intel/IBM/Sun will start make some FP chips. they already do this actually, just these chips still emulate x86/sparc/... ones :) -- Best regards, Bulat mailto:Bulat.Ziganshin@gmail.com