Am Mittwoch 28 April 2010 23:32:10 schrieb Philip Scott:
Hi all,
I kind of stumbled into this while doing something else - but I thought it was worth posting since I have never actually managed to crash ghci before :)
Prelude> let d = [1..1000000000000] :: [Int] Prelude> sum d 0
0? I mean, I might have expected an integer overflow, but 0?
I can deduce that you have a 32-bit system (as I do). Because: Prelude> 1000000000000 :: Int -727379968 So ([1 .. 1000000000000] :: [Int]) == [] and sum [] == 0 isn't surprising at all.
And now the really odd part... take away one zero
Prelude> let d = [1..100000000000] :: [Int]
Prelude> 100000000000 :: Int 1215752192 , which is a pretty large number. Since you gave the list a name, it stays in memory. For each Int in the list, at least 3 machine words are needed (one for the Int, pointer from cons-cell to value, pointer to next cons- cell; actually, I think the overhead is larger), so if you have less than 14GB of RAM, your ghci can't do it.
Prelude> sum d < ghci crashes and quits >
A slightly more reasonable number..
Prelude> let d = [1..10000000] :: [Int] Prelude> sum d *** Exception: stack overflow
At least I can appreciate what is going on in this one :)
That works with Prelude Data.List> foldl' (+) 0 [1 .. 10000000] :: Int -2004260032 Prelude Data.List> :set +s Prelude Data.List> foldl' (+) 0 [1 .. 100000000] :: Int 987459712 (3.47 secs, 4820224616 bytes) Prelude Data.List> foldl' (+) 0 [1 .. 100000000000] :: Int 2103145472 (44.75 secs, 58584335012 bytes)
I'm aware that this is a silly way to sum from 1 to n, but I am at a loss to explain the above behavior (Except the stack overflow, that is fair enough). I'm also aware that Int is a machine Int, not a clever infinite haskell Int.
Any ideas?
All the best,
Phil