To my
Me> For, with -O and much in-lining, the interfaces may be very large.
Simon Marlow <[EMAIL PROTECTED]> replies
Ma> You really don't want to use -O with GHCi (in fact, at the moment
Ma> you can't).
Me> This may be serious. For example, putStr $ show (sum1 [1..n])
Me> will need a constant memory or O(n) memory depending on the -O
Me> possibility.
Ma> True. But compiling with optimsation takes 2-3 times longer than
Ma>interpreting - if you're prepared to wait that long, you might as well
Ma> compile instead. Compiling will give more of a performance boost in
Ma> general (about 10x) than turning on -O (perhaps 2x).
??
I do not understand. I meant -O only for ghc --make -O
(has -O sense in other situation?)
to compile _once_ the whole application library.
.o, .hi files are prepared once
(Why are you saying `to wait that long' ?).
Then, thousands of times ghci is run that uses the ready library
optimized code and loads smaller examples to interptret.
(I never tried -O to compile for ghci because so far we are
considering the bugs made by ghci using the code produced with
-Onot).
Typical example: the product of concrete matrices matrMul M N
is called for several M, N - in interpreted mode.
But matrMul refers to the application library code compiled (once)
with -O.
Is this possible?
Please, document these questions in the manual.
-----------------
Serge Mechveliani
[EMAIL PROTECTED]
_______________________________________________
Glasgow-haskell-bugs mailing list
[EMAIL PROTECTED]
http://www.haskell.org/mailman/listinfo/glasgow-haskell-bugs
