[Haskell-cafe] Ridiculously slow FFI, or cairo binding?
Felipe Almeida Lessa
felipe.lessa at gmail.com
Wed Nov 2 11:23:07 CET 2011
On Wed, Nov 2, 2011 at 8:15 AM, Eugene Kirpichov <ekirpichov at gmail.com> wrote:
> Any idea how to debug why all the GMP calls?
> I'm looking at even the auto-generated source for cairo bindings, but I
> don't see anything at all that could lead to *thousands* of them.
Found them. Look at the Types module and you'll see
cFloatConv :: (RealFloat a, RealFloat b) => a -> b
cFloatConv = realToFrac
This function (or its cousins peekFloatConv, withFloatConv...) are
used *everywhere*.
Looking at this module with ghc-core we see that GHC compiled a
generic version of cFloatConv:
Graphics.Rendering.Cairo.Types.$wcFloatConv
:: forall a_a3TN b_a3TO.
(RealFloat a_a3TN, RealFrac b_a3TO) =>
a_a3TN -> b_a3TO
[GblId,
Arity=3,
Unf=Unf{Src=<vanilla>, TopLvl=True, Arity=3, Value=True,
ConLike=True, Cheap=True, Expandable=True,
Guidance=IF_ARGS [3 3 0] 12 0}]
Graphics.Rendering.Cairo.Types.$wcFloatConv =
\ (@ a_a3TN)
(@ b_a3TO)
(w_s5zg :: RealFloat a_a3TN)
(ww_s5zj :: RealFrac b_a3TO)
(w1_s5zA :: a_a3TN) ->
fromRational
@ b_a3TO
($p2RealFrac @ b_a3TO ww_s5zj)
(toRational
@ a_a3TN
($p1RealFrac
@ a_a3TN ($p1RealFloat @ a_a3TN w_s5zg))
w1_s5zA)
Note that this is basically cFloatConv = fromRational . toRational.
*However*, GHC also compiled a Double -> Double specialization:
Graphics.Rendering.Cairo.Types.cFloatConv1
:: Double -> Double
[GblId,
Arity=1,
Unf=Unf{Src=InlineStable, TopLvl=True, Arity=1, Value=True,
ConLike=True, Cheap=True, Expandable=True,
Guidance=ALWAYS_IF(unsat_ok=True,boring_ok=False)
Tmpl= \ (eta_B1 [Occ=Once!] :: Double) ->
case eta_B1 of _ { D# ww_a5v3 [Occ=Once] ->
case $w$ctoRational ww_a5v3
of _ { (# ww2_a5v8 [Occ=Once], ww3_a5v9 [Occ=Once] #) ->
$wfromRat ww2_a5v8 ww3_a5v9
}
}}]
Graphics.Rendering.Cairo.Types.cFloatConv1 =
\ (eta_B1 :: Double) ->
case eta_B1 of _ { D# ww_a5v3 ->
case $w$ctoRational ww_a5v3
of _ { (# ww2_a5v8, ww3_a5v9 #) ->
$wfromRat ww2_a5v8 ww3_a5v9
}
}
...which is also equivalent to fromRational . toRational however with
the type class inlined! Oh, god...
Cheers,
--
Felipe.
More information about the Haskell-Cafe
mailing list