[Git][ghc/ghc][master] Add missing int64/word64-to-double/float rules (#23907)
Marge Bot (@marge-bot)
gitlab at gitlab.haskell.org
Fri Sep 15 15:18:22 UTC 2023
Marge Bot pushed to branch master at Glasgow Haskell Compiler / GHC
Commits:
5126a2fe by Sylvain Henry at 2023-09-15T11:18:02-04:00
Add missing int64/word64-to-double/float rules (#23907)
CLC proposal: https://github.com/haskell/core-libraries-committee/issues/203
- - - - -
5 changed files:
- libraries/base/GHC/Float.hs
- libraries/base/changelog.md
- + testsuite/tests/numeric/should_compile/T23907.hs
- + testsuite/tests/numeric/should_compile/T23907.stderr
- testsuite/tests/numeric/should_compile/all.T
Changes:
=====================================
libraries/base/GHC/Float.hs
=====================================
@@ -1810,3 +1810,22 @@ foreign import prim "stg_doubleToWord64zh"
"Word# -> Natural -> Double#"
forall x. naturalToDouble# (NS x) = word2Double# x #-}
+
+-- We don't have word64ToFloat/word64ToDouble primops (#23908), only
+-- word2Float/word2Double, so we can only perform these transformations when
+-- word-size is 64-bit.
+#if WORD_SIZE_IN_BITS == 64
+{-# RULES
+
+"Int64# -> Integer -> Float#"
+ forall x. integerToFloat# (integerFromInt64# x) = int2Float# (int64ToInt# x)
+
+"Int64# -> Integer -> Double#"
+ forall x. integerToDouble# (integerFromInt64# x) = int2Double# (int64ToInt# x)
+
+"Word64# -> Integer -> Float#"
+ forall x. integerToFloat# (integerFromWord64# x) = word2Float# (word64ToWord# x)
+
+"Word64# -> Integer -> Double#"
+ forall x. integerToDouble# (integerFromWord64# x) = word2Double# (word64ToWord# x) #-}
+#endif
=====================================
libraries/base/changelog.md
=====================================
@@ -4,6 +4,7 @@
* Export `foldl'` from `Prelude` ([CLC proposal #167](https://github.com/haskell/core-libraries-committee/issues/167))
* Add a `RULE` to `Prelude.lookup`, allowing it to participate in list fusion ([CLC proposal #174](https://github.com/haskell/core-libraries-committee/issues/175))
* The `Enum Int64` and `Enum Word64` instances now use native operations on 32-bit platforms, increasing performance by up to 1.5x on i386 and up to 5.6x with the JavaScript backend. ([CLC proposal #187](https://github.com/haskell/core-libraries-committee/issues/187))
+ * Add rewrite rules for conversion between Int64/Word64 and Float/Double on 64-bit architectures ([CLC proposal #203](https://github.com/haskell/core-libraries-committee/issues/203)).
## 4.19.0.0 *TBA*
* Add `{-# WARNING in "x-partial" #-}` to `Data.List.{head,tail}`.
=====================================
testsuite/tests/numeric/should_compile/T23907.hs
=====================================
@@ -0,0 +1,67 @@
+module T23907 (loop) where
+
+import Data.Word
+import Data.Bits
+
+{-# NOINLINE loop #-}
+loop :: Int -> Double -> SMGen -> (Double, SMGen)
+loop 0 !a !s = (a, s)
+loop n !a !s = loop (n - 1) (a + b) t where (b, t) = nextDouble s
+
+mix64 :: Word64 -> Word64
+mix64 z0 =
+ -- MurmurHash3Mixer
+ let z1 = shiftXorMultiply 33 0xff51afd7ed558ccd z0
+ z2 = shiftXorMultiply 33 0xc4ceb9fe1a85ec53 z1
+ z3 = shiftXor 33 z2
+ in z3
+
+shiftXor :: Int -> Word64 -> Word64
+shiftXor n w = w `xor` (w `shiftR` n)
+
+shiftXorMultiply :: Int -> Word64 -> Word64 -> Word64
+shiftXorMultiply n k w = shiftXor n w * k
+
+nextWord64 :: SMGen -> (Word64, SMGen)
+nextWord64 (SMGen seed gamma) = (mix64 seed', SMGen seed' gamma)
+ where
+ seed' = seed + gamma
+
+nextDouble :: SMGen -> (Double, SMGen)
+nextDouble g = case nextWord64 g of
+ (w64, g') -> (fromIntegral (w64 `shiftR` 11) * doubleUlp, g')
+
+data SMGen = SMGen !Word64 !Word64 -- seed and gamma; gamma is odd
+
+mkSMGen :: Word64 -> SMGen
+mkSMGen s = SMGen (mix64 s) (mixGamma (s + goldenGamma))
+
+goldenGamma :: Word64
+goldenGamma = 0x9e3779b97f4a7c15
+
+floatUlp :: Float
+floatUlp = 1.0 / fromIntegral (1 `shiftL` 24 :: Word32)
+
+doubleUlp :: Double
+doubleUlp = 1.0 / fromIntegral (1 `shiftL` 53 :: Word64)
+
+mix64variant13 :: Word64 -> Word64
+mix64variant13 z0 =
+ -- Better Bit Mixing - Improving on MurmurHash3's 64-bit Finalizer
+ -- http://zimbry.blogspot.fi/2011/09/better-bit-mixing-improving-on.html
+ --
+ -- Stafford's Mix13
+ let z1 = shiftXorMultiply 30 0xbf58476d1ce4e5b9 z0 -- MurmurHash3 mix constants
+ z2 = shiftXorMultiply 27 0x94d049bb133111eb z1
+ z3 = shiftXor 31 z2
+ in z3
+
+mixGamma :: Word64 -> Word64
+mixGamma z0 =
+ let z1 = mix64variant13 z0 .|. 1 -- force to be odd
+ n = popCount (z1 `xor` (z1 `shiftR` 1))
+ -- see: http://www.pcg-random.org/posts/bugs-in-splitmix.html
+ -- let's trust the text of the paper, not the code.
+ in if n >= 24
+ then z1
+ else z1 `xor` 0xaaaaaaaaaaaaaaaa
=====================================
testsuite/tests/numeric/should_compile/T23907.stderr
=====================================
@@ -0,0 +1,57 @@
+
+==================== Tidy Core ====================
+Result size of Tidy Core
+ = {terms: 90, types: 62, coercions: 0, joins: 0/3}
+
+$WSMGen
+ = \ conrep conrep1 ->
+ case conrep of { W64# unbx ->
+ case conrep1 of { W64# unbx1 -> SMGen unbx unbx1 }
+ }
+
+Rec {
+$wloop
+ = \ ww ww1 ww2 ww3 ->
+ case ww of ds {
+ __DEFAULT ->
+ let { seed' = plusWord64# ww2 ww3 } in
+ let {
+ x#
+ = timesWord64#
+ (xor64# seed' (uncheckedShiftRL64# seed' 33#))
+ 18397679294719823053#Word64 } in
+ let {
+ x#1
+ = timesWord64#
+ (xor64# x# (uncheckedShiftRL64# x# 33#))
+ 14181476777654086739#Word64 } in
+ $wloop
+ (-# ds 1#)
+ (+##
+ ww1
+ (*##
+ (word2Double#
+ (word64ToWord#
+ (uncheckedShiftRL64#
+ (xor64# x#1 (uncheckedShiftRL64# x#1 33#)) 11#)))
+ 1.1102230246251565e-16##))
+ seed'
+ ww3;
+ 0# -> (# ww1, ww2, ww3 #)
+ }
+end Rec }
+
+loop
+ = \ ds a s ->
+ case ds of { I# ww ->
+ case a of { D# ww1 ->
+ case s of { SMGen ww2 ww3 ->
+ case $wloop ww ww1 ww2 ww3 of { (# ww4, ww5, ww6 #) ->
+ (D# ww4, SMGen ww5 ww6)
+ }
+ }
+ }
+ }
+
+
+
=====================================
testsuite/tests/numeric/should_compile/all.T
=====================================
@@ -20,3 +20,4 @@ test('T20448', normal, compile, ['-ddump-simpl -O -dsuppress-all -dno-typeable-b
test('T19641', normal, compile, ['-ddump-simpl -O -dsuppress-all -dno-typeable-binds -dsuppress-uniques'])
test('T15547', normal, compile, ['-ddump-simpl -O -dsuppress-all -dno-typeable-binds -dsuppress-uniques'])
test('T23019', normal, compile, ['-O'])
+test('T23907', [ when(wordsize(32), expect_broken(23908))], compile, ['-ddump-simpl -O2 -dsuppress-all -dno-typeable-binds -dsuppress-uniques'])
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5126a2fef0385e206643b6af0543d10ff0c219d8
--
View it on GitLab: https://gitlab.haskell.org/ghc/ghc/-/commit/5126a2fef0385e206643b6af0543d10ff0c219d8
You're receiving this email because of your account on gitlab.haskell.org.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.haskell.org/pipermail/ghc-commits/attachments/20230915/e17813bf/attachment-0001.html>
More information about the ghc-commits
mailing list