Ticket #4335: fromRationalRatio.dpatch

File fromRationalRatio.dpatch, 50.6 KB (added by daniel.is.fischer, 4 years ago)
Line 
11 patch for repository /home/dafis/Haskell/Hacking/ghc/libraries/base:
2
3Tue Oct 19 03:01:09 CEST 2010  Daniel Fischer <daniel.is.fischer@web.de>
4  * FIX #4335
5  fromRational :: Rational -> Ratio a produced invalid results for fixed-width
6  types a. Reduce the fraction to avoid that.
7
8New patches:
9
10[FIX #4335
11Daniel Fischer <daniel.is.fischer@web.de>**20101019010109
12 Ignore-this: 3b8ad075637088df77937b923f623204
13 fromRational :: Rational -> Ratio a produced invalid results for fixed-width
14 types a. Reduce the fraction to avoid that.
15] {
16hunk ./GHC/Real.lhs 331
17     signum (x:%_)       =  signum x :% 1
18     fromInteger x       =  fromInteger x :% 1
19 
20+{-# RULES "fromRational/id" fromRational = id :: Rational -> Rational #-}
21 instance  (Integral a)  => Fractional (Ratio a)  where
22     {-# SPECIALIZE instance Fractional Rational #-}
23     (x:%y) / (x':%y')   =  (x*y') % (y*x')
24hunk ./GHC/Real.lhs 336
25     recip (x:%y)        =  y % x
26-    fromRational (x:%y) =  fromInteger x :% fromInteger y
27+    fromRational (x:%y) =  fromInteger x % fromInteger y
28 
29 instance  (Integral a)  => Real (Ratio a)  where
30     {-# SPECIALIZE instance Real Rational #-}
31}
32
33Context:
34
35[Define SpecConstrAnnotation in GHC.Exts, and import it from there
36simonpj@microsoft.com**20101018135857
37 Ignore-this: 8bf81cbc5787dbb5a3875b5622f67732
38   
39 Reason: avoid having to link the entire ghc package in modules
40 that use compile-time annotations:
41   
42        import GHC.Exts( SpecConstrAnnotation )
43        {-# ANN type T ForceSpecConstr #-}
44   
45 It's a kind of bug that the package exporting SpecConstrAnnotation
46 is linked even though it is only needed at compile time, but putting
47 the data type declaration in GHC.Exts is a simple way to sidestep
48 the problem
49   
50 See See Note [SpecConstrAnnotation] in SpecConstr
51]
52[throwTo: mention interruptible foreign calls
53Simon Marlow <marlowsd@gmail.com>**20101014084220
54 Ignore-this: dbc53d85f870cf649f87186c7185465a
55]
56[remove trailing whitespace
57Simon Marlow <marlowsd@gmail.com>**20101013101906
58 Ignore-this: b8b424540cacbbb3c6d934242e3af795
59]
60[FIX #4381
61Simon Marlow <marlowsd@gmail.com>**20101013101849
62 Ignore-this: f0daa4845eeb444231451b975b71d055
63 Fix scaleFloat by clamping the scaling parameter so that
64 exponent + scale doesn't overflow.
65 
66 Patch by: Daniel Fischer <daniel.is.fischer@web.de>
67]
68[Replaced some throws to throwIOs where the type is IO
69Bas van Dijk <v.dijk.bas@gmail.com>**20100924221340
70 Ignore-this: e74191e4527ae6f7551c95fd41063335
71]
72[Added initial .authorspellings
73Bas van Dijk <v.dijk.bas@gmail.com>**20101005072701
74 Ignore-this: 63628bcabfdd0b7beda4cd37daeccd89
75]
76[Lazier intersperse
77Daniel Fischer <daniel.is.fischer@web.de>**20101002231201
78 Ignore-this: a0fed65930cf19e68b4363381a5ab576
79 A lazier implementation of intersperse, and consequentially intercalate, to
80 avoid space leaks.
81]
82[FIX #4228 (atanh (-1) returns NaN instead of -Infinity)
83ghc@cainnorris.net**20100816213654
84 Ignore-this: dee89c24493e84a02bea711a1c83a73f
85]
86[Make intersectBy lazier
87Daniel Fischer <daniel.is.fischer@web.de>**20100930191731
88 Ignore-this: ef687bc75923434e85c14b57171576aa
89 Add shortcuts to intersectBy for empty list arguments.
90 In addition to being faster in that case, more inputs yield defined results.
91 Treats ticket #4323
92]
93[doc tweak for Directory file type: file names are '\0'-separated
94Simon Marlow <marlowsd@gmail.com>**20100922113811
95 Ignore-this: 96b7b004bd6e5bc3e958ad55bf238ba1
96]
97[documentation for IODeviceType (#4317, edited by me)
98Simon Marlow <marlowsd@gmail.com>**20100915131341
99 Ignore-this: 21c50ca7a189eebcf299523b6e942bae
100]
101[Allow Data.HashTable construction with user-supplied size
102**20100722210726
103 Ignore-this: bd54880bb16a106a992f03b040dc4164
104 
105 This avoids some resizing for users who know they will be inserting a
106 lot of data.
107 
108 http://hackage.haskell.org/trac/ghc/ticket/4193
109]
110[some fixes for hGetBufSome
111Simon Marlow <marlowsd@gmail.com>**20100916113732
112 Ignore-this: 3e596a606c180dc4859ea8f4c9132ca1
113  - fix one case where it was blocking when it shouldn't
114  - a couple of error-message tweaks
115]
116[Windows: map ERROR_NO_DATA to EPIPE, rather than EINVAL
117Simon Marlow <marlowsd@gmail.com>**20100915142618
118 Ignore-this: 9023e5f0542419f225aef26cb6b1d88d
119 WriteFile() returns ERROR_NO_DATA when writing to a pipe that is
120 "closing", however by default the write() wrapper in the CRT maps this
121 to EINVAL so we get confusing things like
122 
123   hPutChar: invalid argument (Invalid Argumnet)
124 
125 when piping the output of a Haskell program into something that closes
126 the pipe early.  This was happening in the testsuite in a few place.
127 
128 The solution is to map ERROR_NO_DATA to EPIPE correctly, as we
129 explicitly check for EPIPE on stdout (in GHC.TopHandler) so we can
130 exit without an error in this case.
131]
132[tighten up parsing of numbers (#1579)
133Simon Marlow <marlowsd@gmail.com>**20100913214733
134 Ignore-this: 3411bf3d2e98cfacb9e0afd11d79e722
135]
136[Add absentError. 
137simonpj@microsoft.com**20100914134639
138 Ignore-this: d0eef5a87e1def4cdbde92a55241c8c4
139 
140 This patch accompanies the HEAD patch:
141 
142   Tue Sep 14 12:38:27 BST 2010  simonpj@microsoft.com
143     * Make absent-arg wrappers work for unlifted types (fix Trac #4306)
144     
145     Previously we were simply passing arguments of unlifted
146     type to a wrapper, even if they were absent, which was
147     stupid.
148     
149     See Note [Absent error Id] in WwLib.
150]
151[Add missing import, fixes build on windows
152simonpj@microsoft.com**20100914122750
153 Ignore-this: 12ece15ef94982ddfbf5f9f7900619da
154]
155[Add a suitable Show instance for TextEncoding (#4273)
156Simon Marlow <marlowsd@gmail.com>**20100913154459
157 Ignore-this: 77f2235460895debd2827f34c42c3435
158]
159[don't fill a finalized handle with an error (see comment)
160Simon Marlow <marlowsd@gmail.com>**20100913153350
161 Ignore-this: c72cdb6898dffa88eca1d781171b2943
162]
163[deriving (Eq, Ord, Read, Show) for Newline and NewlineMode
164Simon Marlow <marlowsd@gmail.com>**20100913153031
165 Ignore-this: 9b9b29bfb7abf5550cfbfa7788f81bf
166]
167[fix warning on Windows
168Simon Marlow <marlowsd@gmail.com>**20100913111536
169 Ignore-this: dacc5448c452daad60ed37a1a5ed096e
170]
171[Put the state-token argument on fill, done, adjust on the RHS
172simonpj@microsoft.com**20100913101832
173 Ignore-this: d228b492de7d4635c026ed24cbc17e34
174 
175 This is so that the functions will inline when
176 applied to their normal (non-state-token) aguments.
177 
178 I forget why I did this, but it seems like the right thing anyway.
179]
180[avoid Foreign.unsafePerformIO
181Ross Paterson <ross@soi.city.ac.uk>**20100909125521
182 Ignore-this: b698101119ffd1bc6311cce0736f745d
183]
184[Remove debugging code accidentally left in
185Simon Marlow <marlowsd@gmail.com>**20100909113331
186 Ignore-this: 906a14176dd37030b8203782a687936b
187]
188[Fix Windows build; patches frmo ezyang
189Ian Lynagh <igloo@earth.li>**20100908123037
190 Ignore-this: 2f02986087edd7da8382221012c27cd0
191]
192[More accurate isatty test for MinGW.
193Edward Z. Yang <ezyang@mit.edu>**20100907154144
194 Ignore-this: 93bdc2b2a8e65a7c4c7d3906bdda01db
195]
196[Fix the build when HAVE_KQUEUE but not HAVE_KEVENT64
197Ian Lynagh <igloo@earth.li>**20100904223703]
198[Fix warnings
199benl@ouroborus.net**20100830044741
200 Ignore-this: 8397aaec7c36046c9ace403e65f32d32
201]
202[fix cache variable name used by FP_SEARCH_LIBS_PROTO
203Ross Paterson <ross@soi.city.ac.uk>**20100819204858
204 Ignore-this: b8113cb3c6f0e03c507297c99d3d82b7
205]
206[Add a missing castPtr (only shows up in -DDEBUG)
207simonpj@microsoft.com**20100815145127
208 Ignore-this: 30b9c42cd3ce7837bdabd254fe66078d
209]
210[Fixed a rounding error in threadDelay
211Johan Tibell <johan.tibell@gmail.com>**20100813124043
212 Ignore-this: 1cb77d0852233ffffb144b134064ee3c
213]
214[export allocaBytesAligned; make allocaArray use the correct alignment (#2917)
215Simon Marlow <marlowsd@gmail.com>**20100812105524
216 Ignore-this: deb6495f7b7b84deaf02b88927a5ba8c
217]
218[deprecate unGetChan and isEmptyChan (see #4154)
219Simon Marlow <marlowsd@gmail.com>**20100705125952
220 Ignore-this: b4e769959f131b2d0001eb7202bc1b92
221]
222[Add type signatures to cope with lack of local generalisation
223simonpj@microsoft.com**20100728124847
224 Ignore-this: d3af9a47c2821c6081bde05a135a92fb
225]
226[Add type signature in local where
227simonpj@microsoft.com**20100727151532
228 Ignore-this: 1c57063ad32d13e0d1ec8daf968bf055
229]
230[Integrated new I/O manager
231Simon Marlow <marlowsd@gmail.com>**20100810082248
232 Ignore-this: ed70a9066ac9b676a446fe99978fef7a
233 (patch originally by Johan Tibell <johan.tibell@gmail.com>, minor merging by me)
234]
235[Add mfilter to Control.Monad
236jon.fairbairn@cl.cam.ac.uk**20090917145616
237 Ignore-this: de4240b60684f3065b29378df3ea98f2
238 
239 Straightforward MonadPlus version of List.filter. I would
240 prefer to call it filter, but the current naming scheme for
241 Control.Monad implies mfilter.
242 
243]
244[move Monad and MonadFix instances for Either from mtl (proposal #4159)
245Ross Paterson <ross@soi.city.ac.uk>**20100729122449
246 Ignore-this: b0f8cd8643679948d1da43bd7c08c5aa
247 
248 The Monad and MonadFix instances for Either (formerly in the mtl
249 package) are moved to Control.Monad.Instances and Control.Monad.Fix
250 respectively.  The Monad instance is still an orphan, to retain Haskell
251 98 compatibility, but the MonadFix instance is together with its class.
252 The Error constraint is removed from both instances, and the default
253 definition of fail is used.
254]
255[Remove egregious ghc-ish from Foreign.Marshal
256Malcolm.Wallace@me.com**20100722075449]
257[add numSparks :: IO Int (#4167)
258Simon Marlow <marlowsd@gmail.com>**20100720153858
259 Ignore-this: 4543f57a7f137f8cae1c3efc5c023a9b
260]
261[add unsafeLocalState from Haskell 2010, and docs
262Simon Marlow <marlowsd@gmail.com>**20100720082819
263 Ignore-this: dcd79fb546ebe29ddff4df279ec2f38
264]
265[docs: mention that Foreign.unsafePerformIO is deprecated
266Simon Marlow <marlowsd@gmail.com>**20100720082804
267 Ignore-this: 4cfebb8f2a1cddc7d15e94e31b2befa4
268 We can't actually deprecate it without introducing a name clash
269 between Foreign.unsafePerformIO and System.IO.Unsafe.unsafePerformIO
270]
271[doc formatting fix
272Simon Marlow <marlowsd@gmail.com>**20100714151347
273 Ignore-this: 255edef607dcd290e198015240b5d125
274]
275[add module intro from Haskell 2010
276Simon Marlow <marlowsd@gmail.com>**20100714115853
277 Ignore-this: 59b5a07507a059ccccdff2dfb6490a27
278]
279[document exception-overriding behaviour in withFile
280Simon Marlow <marlowsd@gmail.com>**20100714104107
281 Ignore-this: f99e641ea2f46d872cb7420a62fa50dc
282]
283[doc: use "finalizer" consistently
284Simon Marlow <marlowsd@gmail.com>**20100714103649
285 Ignore-this: bdfea40f31dc5045fdbc6e12266dda93
286]
287[clarify meaning of bit
288Simon Marlow <marlowsd@gmail.com>**20100714103310
289 Ignore-this: 521b031f1e83ef34ca03d9aa9273df8a
290]
291[note shortcutting behaviour of any/all/elem
292Simon Marlow <marlowsd@gmail.com>**20100714103304
293 Ignore-this: 1605f362ba0712ad1cea1309636f3ea1
294]
295[add cast{C,U}CharToChar and castCharTo{C,U}Char, from Haskell 2010
296Simon Marlow <marlowsd@gmail.com>**20100713132515
297 Ignore-this: 9b1da827016c7b08668078b45964e9de
298]
299[mention that IntPtr and WordPtr can be marshalled to/from intptr_t and uintptr_t
300Simon Marlow <marlowsd@gmail.com>**20100713132403
301 Ignore-this: dcc112a72746ba117a84fa29e71b6800
302]
303[Partial fix for Trac #4136
304simonpj@microsoft.com**20100707135725
305 Ignore-this: 9548eeb3187d9779d4e5c858a0f35354
306 
307 In 'choose' (which is a library function designed specifically
308 to support derived instances of Read), we must match Symbol
309 as well as Ident, for nullary constructors that (wierdly) are
310 symbols.
311]
312[Fix typo in documentation
313Simon Hengel <simon.hengel@wiktory.org>**20100711141648
314 Ignore-this: c052dd8a681832ef598a323ad55eae3a
315]
316[Remove duplicated word in documentation
317Simon Hengel <simon.hengel@wiktory.org>**20100711072703
318 Ignore-this: fb3732dc57be55f14168792f923433
319]
320[Allow nhc98 to cope with recent changes to Control.Exception.
321Malcolm.Wallace@me.com**20100710170940]
322[ New asynchronous exception control API (base parts)
323Simon Marlow <marlowsd@gmail.com>**20100708152735
324 Ignore-this: 71a4811804f04259f1fe739f8863beaf
325   
326 As discussed on the libraries/haskell-cafe mailing lists
327   http://www.haskell.org/pipermail/libraries/2010-April/013420.html
328 
329 This is a replacement for block/unblock in the asychronous exceptions
330 API to fix a problem whereby a function could unblock asynchronous
331 exceptions even if called within a blocked context.
332 
333 The new terminology is "mask" rather than "block" (to avoid confusion
334 due to overloaded meanings of the latter).
335 
336 The following is the new API; the old API is deprecated but still
337 available for the time being.
338 
339 Control.Exception
340 -----------------
341 
342 mask  :: ((forall a. IO a -> IO a) -> IO b) -> IO b
343 mask_ :: IO a -> IO a
344 
345 uninterruptibleMask :: ((forall a. IO a -> IO a) -> IO b) -> IO b
346 uninterruptibleMask_ :: IO a -> IO
347 
348 getMaskingState :: IO MaskingState
349 
350 data MaskingState
351   = Unmasked
352   | MaskedInterruptible
353   | MaskedUninterruptible
354 
355 
356 Control.Concurrent
357 ------------------
358 
359 forkIOUnmasked :: IO () -> IO ThreadId
360]
361[Async-exception safety, and avoid space leaks
362Simon Marlow <marlowsd@gmail.com>**20100708145819
363 Ignore-this: dbfd0e61551e9e7b4fc1c6fe9b9a83de
364 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
365 Modified slightly by me to remove non-functional changes.
366]
367[Async-exception safety, and avoid space leaks
368Simon Marlow <marlowsd@gmail.com>**20100708103154
369 Ignore-this: 190c3ac8f6633231624da8cf1316588
370 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
371 Modified slightly by me to remove non-functional changes.
372]
373[Fix a few places where we forgot to close the text codecs (#4029)
374Simon Marlow <marlowsd@gmail.com>**20100702130210
375 Ignore-this: 2e81a4b4cb343181cef34b0f9e2ded47
376 Each time you invoke :load in GHCi it resets the CAFs, including
377 stdin/stdout/stderr, and each of these was allocating a new iconv_t.
378]
379[remove docs from Monad that belonged on the instance for MonadPlus IO
380Simon Marlow <marlowsd@gmail.com>**20100701154203
381 Ignore-this: 59df02542a7ac9421552a2155d848d27
382]
383[docs: unqualify Prelude.IO
384Simon Marlow <marlowsd@gmail.com>**20100701153817
385 Ignore-this: 73b0202876c827e7a5b4a5ce74e724c4
386]
387[unqualify Float and Double
388Simon Marlow <marlowsd@gmail.com>**20100701142727
389 Ignore-this: cbe89d31a00bf49996a33933324fca17
390]
391[extract information about Data.Time from docs for CTime
392Simon Marlow <marlowsd@gmail.com>**20100701142415
393 Ignore-this: c48c9609b8d36e43e033a7bea81d6f17
394]
395[doc typo
396Simon Marlow <marlowsd@gmail.com>**20100701142354
397 Ignore-this: 17a1fd703831c888975ff63fbfa3a9b2
398]
399[peekArray docs: remove mentions of "this version" and "previous version"
400Simon Marlow <marlowsd@gmail.com>**20100701125333
401 Ignore-this: 39a744874258670bd935ba9e38390939
402]
403[doc typo
404Simon Marlow <marlowsd@gmail.com>**20100701124154
405 Ignore-this: 98f5c286e38c2c34c96b05d5e8bc5ad9
406]
407[doc typo
408Simon Marlow <marlowsd@gmail.com>**20100701124128
409 Ignore-this: 10a4314ec7aed336701fc616fb574ebc
410]
411[doc typo
412Simon Marlow <marlowsd@gmail.com>**20100701123715
413 Ignore-this: c4909a7bf7163460ee5d32f58812041e
414]
415[doc wibble: Haskell 98 I/O Error -> 'IOError'
416Simon Marlow <marlowsd@gmail.com>**20100701123612
417 Ignore-this: bf373df781acbc575e4ffe3b7e6059ae
418]
419[doc typo
420Simon Marlow <marlowsd@gmail.com>**20100701123014
421 Ignore-this: 16aaccae48ef3101adf78ea5b0d5a8fd
422]
423[Haddock hacks to fix whitespace consistency
424Simon Marlow <marlowsd@gmail.com>**20100701121631
425 Ignore-this: 61c58dec52a31fd2d3f331a87d2f903f
426]
427[use '==' consistently rather than '->' in examples
428Simon Marlow <marlowsd@gmail.com>**20100701121616
429 Ignore-this: 472b0a05a85d34d9712186040e1636d9
430]
431[doc wibble: remove confusing mention of "Prelude"
432Simon Marlow <marlowsd@gmail.com>**20100701113308
433 Ignore-this: 232283d0096d01cd45e9b3c5c1e63a6d
434]
435[doc wibble: nonstrict -> non-strict
436Simon Marlow <marlowsd@gmail.com>**20100701113253
437 Ignore-this: 4264f0ab23a0835fc13c6e8601d6b743
438]
439[doc whitespace
440Simon Marlow <marlowsd@gmail.com>**20100701112242
441 Ignore-this: 777a95b1d1140c61d3ab95d5eb5809e7
442]
443[move the doc for 'Char' to its new home in ghc-prim:GHC.Types
444Simon Marlow <marlowsd@gmail.com>**20100629134150
445 Ignore-this: 7687db0077a29498349bfb4b44983985
446]
447[doc wibble
448Simon Marlow <marlowsd@gmail.com>**20100629122608
449 Ignore-this: 9a909e5d015332dc445bd9592e6e386d
450]
451[doc updates in System.IO
452Simon Marlow <marlowsd@gmail.com>**20100629122118
453 Ignore-this: 2257ec1cc4cdb8b7804cfa1f3cf32753
454]
455[doc wibble
456Simon Marlow <marlowsd@gmail.com>**20100625134858
457 Ignore-this: 64c50f29df6c389273b818918fe7033a
458]
459[doc wibbles
460Simon Marlow <marlowsd@gmail.com>**20100624154614
461 Ignore-this: b364aad53beea6e741fee2824459b6e8
462]
463[Fix haddock formatting
464Ian Lynagh <igloo@earth.li>**20100625222623]
465[Give nub's complexity in the haddock docs; fixes #4086
466Ian Lynagh <igloo@earth.li>**20100625222059]
467[correct docs for exitWith: only stdout/stderr are flushed, not all Handles
468Simon Marlow <marlowsd@gmail.com>**20100624130506
469 Ignore-this: 33a938dad8f0bc061572e2ec571cacc7
470]
471[fix docs for isSpace
472Simon Marlow <marlowsd@gmail.com>**20100624130444
473 Ignore-this: b35ff080dbb9833176f08e39dbd9ff6d
474]
475[make the hGetBuf/hPutBuf family work with non-FD Handles (#4144)
476Simon Marlow <marlowsd@gmail.com>**20100624130425
477 Ignore-this: 8200f0208a9b1b1cf4824f343d75819a
478]
479[nit in docs for accumArray
480Simon Marlow <marlowsd@gmail.com>**20100622121131
481 Ignore-this: c066a456c40907e767df10c3990f35ff
482]
483[add doc for the ExitCode type
484Simon Marlow <marlowsd@gmail.com>**20100622120930
485 Ignore-this: 99c34332be7f3565da844528b470054a
486]
487[remove extraneous info from docs for Array
488Simon Marlow <marlowsd@gmail.com>**20100622120921
489 Ignore-this: e2a3f5e84fc23eb7bae911f0680e805e
490]
491[add an INLINE to the list version of traverse, to enable fusion
492Simon Marlow <marlowsd@gmail.com>**20100608082531
493 Ignore-this: ea98cdc3308b406bb04c0f7a38c4424b
494]
495[Don't define the C localeEncoding on Windows
496Ian Lynagh <igloo@earth.li>**20100620202342
497 Ignore-this: c4992f6832a391b0cccc5a9b7d643976
498 (it causes warnings, and isn't used)
499]
500[add Applicative instance for Either (proposal #4095)
501Ross Paterson <ross@soi.city.ac.uk>**20100617225110
502 Ignore-this: 50262ec4700dc16efec5755be5b308c5
503 
504 This is not the only possible instance for Either, but this one is
505 compatible with the usual Monad instance.
506]
507[Use libcharset instead of nl_langinfo(CODESET) if possible.
508pho@cielonegro.org**20100519013112
509 Ignore-this: 4c1e278e022a3d276848afc1dcba4425
510 
511 nl_langinfo(CODESET) doesn't always return standardized variations of the encoding names. Use libcharset if possible, which is shipped together with GNU libiconv.
512]
513[Add a note about the interruptibility of throwTo.
514Simon Marlow <marlowsd@gmail.com>**20100615112720
515 Ignore-this: ae9fabe95310d7c364e95f7784793485
516]
517[docs: note that hGetBufNonBlocking isn't non-blocking on Windows
518Simon Marlow <marlowsd@gmail.com>**20100615112547
519 Ignore-this: 4f3e5213e142149affe08c5123d6efea
520]
521[don't depend on Prelude (#4122)
522Simon Marlow <marlowsd@gmail.com>**20100615105631
523 Ignore-this: 1a3fd49b103fe31cbb453f302c18767f
524]
525[Don't depend on Prelude (#4123)
526Simon Marlow <marlowsd@gmail.com>**20100615105401
527 Ignore-this: cc7616d85a1637bc7621b4f2bc181c0e
528]
529[bump version to 4.3.0.0, added instance MonadPlus STM
530Simon Marlow <marlowsd@gmail.com>**20100601144831
531 Ignore-this: 7c3cf7574499c4267372493f2636dc0
532]
533[Moved MonadPlus instance for STM from Control.Monad.STM to GHC.Conc to avoid an orphaned instance
534Bas van Dijk <v.dijk.bas@gmail.com>**20100516160651
535 Ignore-this: 651b852942b2fae2b93f996e39239b8f
536]
537[Added Applicative and Alternative instances for STM
538Bas van Dijk <v.dijk.bas@gmail.com>**20100516171756
539 Ignore-this: 567003bc4040bc97105cda4d31ebf04a
540]
541[expand Foldable instance for Array
542Ross Paterson <ross@soi.city.ac.uk>**20100602212154
543 Ignore-this: 9bd9e9666a9400431eb92352244fe7e7
544]
545[doc comment illustrating Foldable(foldr)
546Ross Paterson <ross@soi.city.ac.uk>**20100527150833
547 Ignore-this: 8f27d889379803f3ba86d6e928428f3c
548]
549[fix syntax in doc comments
550Ross Paterson <ross@soi.city.ac.uk>**20100527150757
551 Ignore-this: cb78da51d60ff6863dc395f1a892c103
552]
553[export hGetBufSome (#4046)
554Simon Marlow <marlowsd@gmail.com>**20100520093538
555 Ignore-this: f467fad9722e27edfad6b3dd75290e7b
556]
557[hWaitForInput: don't try to read from the device (#4078)
558Simon Marlow <marlowsd@gmail.com>**20100517133741
559 Ignore-this: 55ec33b03397380259b91e4ca62207a6
560 readTextDeviceNonBlocking is not non-blocking on Windows
561]
562[hSetEncoding: change the encoding on both read and write sides (#4066)
563Simon Marlow <marlowsd@gmail.com>**20100514124628
564 Ignore-this: 5b9e9caef06356d0296c584159709ebb
565]
566[Correct haddock formatting.
567Adam Vogt <vogt.adam@gmail.com>**20100423022103
568 Ignore-this: d2622339302048fda48080f7d5ce4a2f
569]
570[Fix for hGetBufSome
571Simon Marlow <marlowsd@gmail.com>**20100505135637
572 Ignore-this: 2019680f8fb223956cacfcf0d046f133
573]
574[improve the documentation for throwTo and killThread (#3884)
575Simon Marlow <marlowsd@gmail.com>**20100505135600
576 Ignore-this: ce881d96ddb729acb6ca09c779975e7d
577]
578[elaborate the docs for unsafePerformIO a bit
579Simon Marlow <marlowsd@gmail.com>**20100505101249
580 Ignore-this: 1cec3f67560b672c64c5a0dcf9a79eb7
581]
582[add Typeable instance
583Simon Marlow <marlowsd@gmail.com>**20100504152815
584 Ignore-this: 6d9cf9d62f0ef17fa459bf213a04098
585]
586[Add hGetBufSome, like hGetBuf but can return short reads
587Simon Marlow <marlowsd@gmail.com>**20100504152759
588 Ignore-this: 195c905b43f8d9505029364e2c5b18e
589]
590[Add swap (#3298)
591Simon Marlow <marlowsd@gmail.com>**20100504095339
592 Ignore-this: 13b007dc4594ce252997ec6fa0bbd976
593]
594[inline allocaArray0, to fix withCString benchmark
595Simon Marlow <marlowsd@gmail.com>**20100423124729
596 Ignore-this: 35c96816acc2f3aaf9dd29f7995fa6f0
597]
598[raise asynchronous exceptions asynchronously (#3997)
599Simon Marlow <marlowsd@gmail.com>**20100421094932
600 Ignore-this: 6d987d93d382c0f69c68c326312abd6b
601]
602[add NOINLINE pragmas for stdin/stdout/stderr
603Simon Marlow <marlowsd@gmail.com>**20100421082041
604 Ignore-this: 3fc130268ec786f28d945858d6690986
605]
606[INLINE alloca and malloc
607Simon Marlow <marlowsd@gmail.com>**20100419135333
608 Ignore-this: b218bd611f18721b1505a8c0b9e6a16a
609 See discussion on glasgow-haskell-users:
610   http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018740.html
611]
612[Move comment closer to the offending line
613Matthias Kilian <kili@outback.escape.de>**20100419155421
614 Ignore-this: b34a1d7affd66f67d210df2377b585d9
615]
616[Ignore the return code of c_fcntl_write again
617Matthias Kilian <kili@outback.escape.de>**20100415140452
618 Ignore-this: 266d8ba02cc3cb79c85629b3528261c9
619 
620 The return code has been ignored in the past on purpose, because
621 O_NONBLOCK will fail on BSDs for some special files. This fixes the
622 problem mentioned in
623 http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018698.html
624 
625]
626[Fix bitrot in IO debugging code
627Ian Lynagh <igloo@earth.li>**20100413134339
628 Also switched to using Haskell Bools (rather than CPP) to en/disable it,
629 so it shouldn't break again in the future.
630]
631[Tiny code tidy-up
632Ian Lynagh <igloo@earth.li>**20100413011147]
633[remove old/wrong comment
634Simon Marlow <marlowsd@gmail.com>**20100325161403
635 Ignore-this: e6e377d44af48c4162d17d55bdf3f821
636]
637[withThread: block asynchronous exceptions before installing exception handler.
638Bas van Dijk <v.dijk.bas@gmail.com>**20100329131624
639 Ignore-this: be5aeb47dbd73807b5f94df11afbb81c
640 Note that I don't unblock the given io computation. Because AFAICS
641 withThread is only called with 'waitFd' which only performs an FFI
642 call which can't receive asynchronous exceptions anyway.
643]
644[runInUnboundThread: block asynchronous exceptions before installing exception handler
645Bas van Dijk <v.dijk.bas@gmail.com>**20100329131549
646 Ignore-this: a00c5e32fe3981ff87bedd367a69051e
647]
648[fix the deprecation message (GHC.IO.Handle.Base -> GHC.IO.Handle)
649Simon Marlow <marlowsd@gmail.com>**20100330121137
650 Ignore-this: 4ca8500a01ac93454507aa8f9dd001f9
651]
652[Make SampleVar an abstract newtype
653Bas van Dijk <v.dijk.bas@gmail.com>**20100318200349
654 Ignore-this: 27939e2a064b75e71cb146117346be30
655]
656[Fix bugs regarding asynchronous exceptions and laziness in Control.Concurrent.SampleVar
657Bas van Dijk <v.dijk.bas@gmail.com>**20100318200104
658 Ignore-this: 7376b2a3afe155daf233a8f1ddc0a7a
659  - Block asynchronous exceptions at the right places
660  - Force thunks before putting them in a MVar
661]
662[Write the thunk 'next' to the MVar
663Bas van Dijk <v.dijk.bas@gmail.com>**20100319125951
664 Ignore-this: dd25636cf220131385ff2fd32493d456
665]
666[change to use STM, fixing 4 things
667Simon Marlow <marlowsd@gmail.com>**20100318104436
668 Ignore-this: 551d30280a7941c08f5c3b14576bdd70
669   1. there was no async exception protection
670   2. there was a space leak (now new value is strict)
671   3. using atomicModifyIORef would be slightly quicker, but can
672      suffer from adverse scheduling issues (see #3838)
673   4. also, the STM version is faster.
674]
675[Tweak docs
676Ian Lynagh <igloo@earth.li>**20100312214129]
677[Fixed dead links in documentation of forkIO
678Bas van Dijk <v.dijk.bas@gmail.com>**20100308222415
679 Ignore-this: 7deb8fd064c867fbede2a6b2e9da4f15
680]
681[Documentation fixes in Control.Exception
682Bas van Dijk <v.dijk.bas@gmail.com>**20100301220442
683 Ignore-this: 761fcba401cbd1f47276ddfc9b5b80f2
684]
685[Plug two race conditions that could lead to deadlocks in the IO manager
686Simon Marlow <marlowsd@gmail.com>**20100225120255
687 Ignore-this: e6983d6b953104d370278ab3e4617e8b
688]
689[FIX #3866: improve documentation of Data.Data.Constr
690jpm@cs.uu.nl**20100224125506
691 Ignore-this: 3818c5d8fee012a3cf322fb455b6e5dc
692]
693[UNDO: Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
694Simon Marlow <marlowsd@gmail.com>**20100223101603
695 Ignore-this: 78becb2d39b3cd9a1a473a5811ca7d92
696]
697[Put the complexity in the length docs. Fixes trac #3680
698Ian Lynagh <igloo@earth.li>**20100221191425]
699[nhc98 should build Data.Functor.
700Malcolm.Wallace@cs.york.ac.uk**20100221163218]
701[Update the exitWith docs
702Ian Lynagh <igloo@earth.li>**20100213140004
703 Error pointed out by Volker Wysk <vw@volker-wysk.de>
704]
705[Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
706Simon Marlow <marlowsd@gmail.com>**20100211101955
707 Ignore-this: 261415363303efca265e80290eac5f28
708]
709[For nhc98, import unsafeInterleaveIO rather than defining it here.
710Malcolm.Wallace@cs.york.ac.uk**20100204171021]
711[Stifle warning about unused return value
712benl@cse.unsw.edu.au**20100203025537]
713[fix #3832: use the locale encoding in openTempFile
714Simon Marlow <marlowsd@gmail.com>**20100120211830
715 Ignore-this: df4f778cc5fefb32290c798db722632c
716 Also while I was here fix an XXX: the Handle contained an
717 uninformative string like <fd: 4> for error messages rather than the
718 real file path.
719]
720[Fix the build: export void, so it doesn't give an unused binding warning
721Ian Lynagh <igloo@earth.li>**20100116174451]
722[hIsEOF: don't do any decoding (#3808)
723Simon Marlow <marlowsd@gmail.com>**20100112230317
724 Ignore-this: 6a384dd2d547ffe3ad3762920e5c1671
725]
726[Control.Monad: +void :: f a -> f ()
727gwern0@gmail.com**20100108214455
728 Ignore-this: 4dc07452315f2d1b4941903ff42fc45f
729 See http://hackage.haskell.org/trac/ghc/ticket/3292
730 Turns m a -> m (). Lets one call functions for their side-effects without
731 having to get rid of their return values with '>> return ()'. Very useful
732 in many contexts (parsing, IO etc.); particularly good for 'forkIO' and 'forM_',
733 as they demand return types of 'IO ()' though most interesting IO functions
734 return non-().
735]
736[Replace the implementation of mergesort with a 2x faster one.
737Malcolm.Wallace@cs.york.ac.uk**20091224152014
738 See ticket http://hackage.haskell.org/trac/ghc/ticket/2143.
739]
740[Restore previous Data.Typeable.typeOf*Default implementations for non-ghc.
741Malcolm.Wallace@cs.york.ac.uk**20091223142625
742 Not all compilers have ScopedTypeVariables.
743]
744[Add comments about double bounds-checking, and fast paths for rectangular arrays
745simonpj@microsoft.com**20091218165655
746 Ignore-this: ea0849419dc00927aba4bd410b1cc58d
747 
748 See Note [Double bounds-checking of index values] for the details.
749 
750 The fast paths omit the doubled checks for cases we know about
751]
752[Fix Trac #3245: memoising typeOf
753simonpj@microsoft.com**20091218155117
754 Ignore-this: 5a178a7f2222293c5ee0c3c43bd1b625
755 
756 The performance bug in #3245 was caused by computing the typeRep
757 once for each call of typeOf, rather than once for each dictionary
758 contruction.  (Computing TypeReps is reasonably expensive, because
759 of the hash-consing machinery.)
760 
761 This is readily fixed by putting the TypeRep construction outside
762 the lambda.  (Arguably GHC might have worked that out itself,
763 but it involves floating something between a type lambda and a
764 value lambda, which GHC doesn't currently do. If it happens a lot
765 we could fix that.)
766]
767[Mark 'index' as INLINE in GHC.Arr
768simonpj@microsoft.com**20091216170441
769 Ignore-this: a4df9d8acf496c8e0e9ce5a520509a2a
770 
771 This makes indexing much faster. See Trac #1216
772]
773[Comment the remaining orphan instance modules
774Ian Lynagh <igloo@earth.li>**20091206125021]
775[De-orphan Eq/Ord Float/Double
776Ian Lynagh <igloo@earth.li>**20091205181238]
777[Add comments to "OPTIONS_GHC -fno-warn-orphans" pragmas
778Ian Lynagh <igloo@earth.li>**20091205165854]
779[Data.Either.partitionEithers was insufficiently lazy.
780Malcolm.Wallace@cs.york.ac.uk**20091202032807
781 Ignore-this: 77e1b3288f66608c71458d8a91bcbe12
782 Spotted by Daniel Fischer.
783]
784[fix the docs regarding finalizer guarantees
785Simon Marlow <marlowsd@gmail.com>**20091130144409
786 Ignore-this: d1ab9532c74a002b8075ff60febcbe2d
787]
788[x86_64 requires more stack
789Malcolm.Wallace@cs.york.ac.uk**20091201033745]
790[check for size < 0 in mallocForeignPtrBytes and friends (#3514)
791Simon Marlow <marlowsd@gmail.com>**20091125143822
792 Ignore-this: 91077d01da2bbe1dfed5155e8b40da9
793]
794[hGetContents: close the handle properly on error
795Simon Marlow <marlowsd@gmail.com>**20091125123435
796 Ignore-this: bc37ff678acc6e547dc390285e056eb9
797 
798 When hGetContents caught an error it was closing the handle and then
799 throwing the exception, without updating the handle with the new
800 closed state.  This lead to a double-closed, which was the cause of
801 
802 *** glibc detected *** ./Setup: double free or corruption
803 
804 when iconv_close was called twice on the decoder.
805 
806 See http://hackage.haskell.org/trac/hackage/ticket/609
807]
808[Fix arities of mapFB and zipFB
809Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091126232219
810 Ignore-this: c4e14cd0a92622549c86e67237a40865
811]
812[Remove an unnecessary -fno-warn-orphans flag
813Ian Lynagh <igloo@earth.li>**20091126123404]
814[Tweak layout to work with alternative layout rule
815Ian Lynagh <igloo@earth.li>**20091125232349]
816[Tweak layout to be accepted by the alternative layout rul
817Ian Lynagh <igloo@earth.li>**20091125194147]
818[Make sure zipWithFB has arity 2
819Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091125010003
820 Ignore-this: 4cf60c55666f03d22a9f5a6e07f52d36
821 
822 It gets 2 arguments in the "zipWith" rule but its arity was higher and the new
823 inliner didn't inline it sometimes, for instance here:
824 
825 mpp ::  [Double] -> [Double] -> [Double] -> [Double] -> [Double]
826 mpp as bs cs ds = zipWith (*) (zipWith (+) as bs) (zipWith (+) cs ds)
827 
828 This was a regression vs. 6.10.
829]
830[Remove an old comment
831Ian Lynagh <igloo@earth.li>**20091124134647]
832[De-orphan the Eq/Ord Integer instances
833Ian Lynagh <igloo@earth.li>**20091124133639]
834[Whitespace only
835Ian Lynagh <igloo@earth.li>**20091124133421]
836[Derive some more instances, rather than writing them by hand
837Ian Lynagh <igloo@earth.li>**20091124011747]
838[We can now derive Ord ()
839Ian Lynagh <igloo@earth.li>**20091124011416]
840[De-orphan tuple Eq/Ord instances
841Ian Lynagh <igloo@earth.li>**20091123233343]
842[Control.Exception.Base no longer has any orphans
843Ian Lynagh <igloo@earth.li>**20091123224905]
844[De-orphan the MonadFix ST instance for GHC
845Ian Lynagh <igloo@earth.li>**20091123223544]
846[Rearrange the contents of Control.Monad.ST; no functionality changes
847Ian Lynagh <igloo@earth.li>**20091123222702]
848[De-orphan the Eq/Ord [a] instances
849Ian Lynagh <igloo@earth.li>**20091123215635]
850[De-orphan the Eq/Ord Char instances
851Ian Lynagh <igloo@earth.li>**20091123202253]
852[De-orphan the Eq/Ord Bool instances
853Ian Lynagh <igloo@earth.li>**20091123201817]
854[Move Eq/Ord Ordering instances to de-orphan them
855Ian Lynagh <igloo@earth.li>**20091123194310]
856[Remove ffi warnings for nhc98.
857Malcolm.Wallace@cs.york.ac.uk**20091123063743]
858[Second attempt to fix #1185 (forkProcess and -threaded)
859Simon Marlow <marlowsd@gmail.com>**20091111151915
860 Ignore-this: fa5f5d5e4e080d4b612a37244f937f9c
861 
862 Patch 2/2: first patch is to ghc
863 
864 This time without dynamic linker hacks, instead I've expanded the
865 existing rts/Globals.c to cache more CAFs, specifically those in
866 GHC.Conc.  We were already using this trick for signal handlers, I
867 should have realised before.
868 
869 It's still quite unsavoury, but we can do away with rts/Globals.c in
870 the future when we switch to a dynamically-linked GHCi.
871]
872[Rollback #1185 fix
873Simon Marlow <marlowsd@gmail.com>**20091106140629
874 Ignore-this: cd5667e8474e37e01ba26a1984274811
875 
876 rolling back:
877 
878 Tue Nov  3 16:05:40 GMT 2009  Simon Marlow <marlowsd@gmail.com>
879   * Fix #1185: restart the IO manager after fork()
880   
881   This is the libraries/base part of the patch; there is a corresponding
882   patch to GHC itself.
883   
884   The main change is that we now keep track of the IO manager's ThreadId
885   in a top-level MVar, and ensureIOManagerIsRunning checks whether a
886   previous IO manager thread is alive before starting one.  In the child
887   of fork(), we can hence call ensureIOManagerIsRunning to restart the
888   IO manager.
889 
890     M ./GHC/Conc.lhs -46 +44
891 
892 Wed Nov  4 17:49:45 GMT 2009  Ian Lynagh <igloo@earth.li>
893   * Fix the build on Windows
894 
895     M ./GHC/Conc.lhs -6 +4
896]
897[Fix the build on Windows
898Ian Lynagh <igloo@earth.li>**20091104174945]
899[Fix #1185: restart the IO manager after fork()
900Simon Marlow <marlowsd@gmail.com>**20091103160540
901 Ignore-this: 6dc05464f1500104554637f4759738cc
902 
903 This is the libraries/base part of the patch; there is a corresponding
904 patch to GHC itself.
905 
906 The main change is that we now keep track of the IO manager's ThreadId
907 in a top-level MVar, and ensureIOManagerIsRunning checks whether a
908 previous IO manager thread is alive before starting one.  In the child
909 of fork(), we can hence call ensureIOManagerIsRunning to restart the
910 IO manager.
911]
912[improve the documentation for throwErrnoIfRetry
913Simon Marlow <marlowsd@gmail.com>**20091016112404
914 Ignore-this: b77275cacf730e15757946027168f63e
915]
916[Don't inline unpackFoldrCString ever
917simonpj@microsoft.com**20091029135350
918 Ignore-this: 85d672649b1b776efc7e97500b05d4f9
919]
920[Inline more default methods
921simonpj@microsoft.com**20091029135330
922 Ignore-this: 289c44b0afd6d5631c2a4e0664275ca9
923 
924 Namely Monad: (>>)
925        Eq:    (==), (/=)
926        Num:   (-), negate
927        Real:  quot, rem, div, mod, recip, (/), truncate
928        Float: (**), logBase, sqrt, tan, tanh
929]
930[Move error messages out of INLINEd default methods
931simonpj@microsoft.com**20091029135118
932 Ignore-this: 9e35dc947f94827a3529eb53a41575fd
933 
934 No need to duplicate the error generation!
935]
936[Exploit now-working default-method INLINE pragmas for Data.Bits
937simonpj@microsoft.com**20091029135041
938 Ignore-this: 8adf225f31ca7a3181ee087e9e4fe535
939 
940 * Add INLINE pragmas to default methods for class Bits
941 
942 * Remove redundant instance methods elsewhere, now that
943   the default method will do the job
944]
945[Tidy up and comment imports
946simonpj@microsoft.com**20091029134414
947 Ignore-this: bf2be31035de975d8995e988933cc940
948]
949[Inline foldr and (.) when applied to two arguments not three
950simonpj@microsoft.com**20091029134335
951 Ignore-this: fccb6f3e90e15f44cb465814be85ede2
952 
953 The new INLINE story is (by design) arity-sensitive, so we must
954 put fewer argument on the LHS for foldr and (.)
955]
956[dirUtils.c no longer available
957Malcolm.Wallace@cs.york.ac.uk**20091013093833]
958[Make hGetContents throw an exception if an error is encountered
959Simon Marlow <marlowsd@gmail.com>**20091012152955
960 Ignore-this: 9f7a7176193eab25c9daaacd9261f2de
961 
962 Strictly speaking this breaks Haskell 98 compatibility, which requires
963 hGetContents to just end the lazy stream silently if an error is
964 encountered.  However, for a few reasons we think it will make
965 everyone's life a bit easier if we make this change
966 
967  1. Errors will be a lot more common in GHC 6.12.1, in the form
968     of Unicode decoding errors.
969 
970  2. When Haskell 98 was designed, we didn't know how to throw
971     exceptions from inside lazy I/O, but now we do.
972 
973  3. If anyone is actually relying on the previous behaviour, their
974     code is arguably broken.
975]
976[Re-instate System.Console.Getopt for nhc98 builds.
977Malcolm.Wallace@cs.york.ac.uk**20091013092843
978 Although it was split out of base a while back, that change was
979 reverted for ghc soon afterwards, but nhc98 never noticed.
980]
981[Roll back "Another instance of nhc98's strange import semantics."
982Ian Lynagh <igloo@earth.li>**20091009185618
983 Fri Oct  9 14:50:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
984 GHC (correctly) warns about the unused import, which breaks the validate
985 build.
986]
987[Roll back "Cope with nhc98's (occasionally-strange) import semantics"
988Ian Lynagh <igloo@earth.li>**20091009184704
989 Fri Oct  9 14:43:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
990 GHC (correctly) warns about the unused import, which breaks the validate
991 build.
992]
993[It seems that nhc98 needs defaulting in Data.Fixed.
994Malcolm.Wallace@cs.york.ac.uk**20091009135242]
995[Another instance of nhc98's strange import semantics.
996Malcolm.Wallace@cs.york.ac.uk**20091009135051]
997[Make Data.Functor compatible with non-GHC compilers.
998Malcolm.Wallace@cs.york.ac.uk**20091009134821]
999[Cope with nhc98's (occasionally-strange) import semantics.
1000Malcolm.Wallace@cs.york.ac.uk**20091009134351]
1001[Fix gratuitous breakage of nhc98 in System.IO.
1002Malcolm.Wallace@cs.york.ac.uk**20091009134001]
1003[Fix gratuitous breakage of nhc98 in Control.Exception.Base.
1004Malcolm.Wallace@cs.york.ac.uk**20091009133615]
1005[Fix gratuitous breakage of non-GHC in Data.Fixed.
1006Malcolm.Wallace@cs.york.ac.uk**20091009133330]
1007[Fix gratuitous breakage for non-GHC in Data.Bits.
1008Malcolm.Wallace@cs.york.ac.uk**20091009133257]
1009[Use UTF-32LE instead of UTF32LE
1010Simon Marlow <marlowsd@gmail.com>**20091006100207
1011 Ignore-this: 7f881e36543d250ef848c9f60d67655a
1012 The latter is not recognised by some iconv implementations.
1013]
1014[Strip any Byte Order Mark (BOM) from the front of decoded streams.
1015Ben.Lippmeier@anu.edu.au*-20090930084229
1016 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
1017 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
1018 of the stream, but Linux iconv doesn't.
1019]
1020[use UTF32BE/UTF32LE instead of UCS-4/UCS-4LE
1021Simon Marlow <marlowsd@gmail.com>**20091005101554
1022 Ignore-this: 2aef5e9bec421e714953b7aa1bdfc1b3
1023]
1024[Strip any Byte Order Mark (BOM) from the front of decoded streams.
1025Ben.Lippmeier@anu.edu.au**20090930084229
1026 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
1027 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
1028 of the stream, but Linux iconv doesn't.
1029]
1030[Add traceEvent :: String -> IO ()
1031Simon Marlow <marlowsd@gmail.com>**20090925141257
1032 Ignore-this: 8b1888bbf9682ffba13f815b6000e4b1
1033 For emitting an event via the RTS tracing framework
1034]
1035[Fix the error message when flushing the read buffer of a non-seekable Handle
1036Simon Marlow <marlowsd@gmail.com>**20090923090536
1037 Ignore-this: 4342026df93759d99480f4e13f80a492
1038]
1039[Fix #3534: No need to flush the byte buffer when setting binary mode
1040Simon Marlow <marlowsd@gmail.com>**20090923090445
1041 Ignore-this: 625817ed7ae2c12291eb993a99dc640a
1042]
1043[Use let !y = x in .. x .. instead of seq in $! and evaluate (#2273)
1044Simon Marlow <marlowsd@gmail.com>**20090916140454]
1045[make some Applicative functions into methods, and split off Data.Functor (proposal #3335)
1046Ross Paterson <ross@soi.city.ac.uk>**20090915173109
1047 Ignore-this: a0cff4de6dfdbcbd56a66101bc4855a9
1048 
1049 The following functions
1050 
1051     (<$) :: Functor f => a -> f b -> f a
1052     (*>) :: Applicative f => f a -> f b -> f b
1053     (<*) :: Applicative f => f a -> f b -> f a
1054     some :: Alternative f => f a -> f [a]
1055     many :: Alternative f => f a -> f [a]
1056 
1057 are moved into the corresponding classes, with the existing implementations
1058 as default definitions.  This gives people creating instances the option of
1059 defining specialized implementations of these functions, though they should
1060 be equivalent to the default definitions.
1061 
1062 Although (<$) is now a method of the Functor class, it is hidden in the
1063 re-export by the Prelude, Control.Monad and Monad.  The new module
1064 Data.Functor exposes the full class, plus the function (<$>).  These are
1065 also re-exported by Control.Applicative.
1066]
1067[On Windows, use the console code page for text file encoding/decoding.
1068Judah Jacobson <judah.jacobson@gmail.com>**20090913022126
1069 Ignore-this: 86c2f2db8ef92b751599795d3195187b
1070 
1071 We keep all of the code page tables in the module
1072 GHC.IO.Encoding.CodePage.Table.  That file was generated automatically
1073 by running codepages/MakeTable.hs; more details are in the comments at the
1074 start of that script.
1075 
1076 Storing the lookup tables adds about 40KB to each statically linked executable;
1077 this only increases the size of a "hello world" program by about 7%.
1078 
1079 Currently we do not support double-byte encodings (Chinese/Japanese/Korean), since
1080 including those codepages would increase the table size to 400KB.  It will be
1081 straightforward to implement them once the work on library DLLs is finished.
1082]
1083[Fix "init" docs: the input list need not be finite. Fixes trac #3465
1084Ian Lynagh <igloo@earth.li>**20090911210437]
1085[Bump base version to 4.2.0.0
1086Ian Lynagh <igloo@earth.li>**20090911153913]
1087[Address #3310
1088Simon Marlow <marlowsd@gmail.com>**20090830152850
1089 Ignore-this: 40c7f7c171ee299a83092fd360a952b7
1090 
1091  - Rename BlockedOnDeadMVar   -> BlockedIndefinitelyOnMVar
1092  - Rename BlockedIndefinitely -> BlockedIndefinitelyOnSTM
1093  - instance Show BlockedIndefinitelyOnMVar is now
1094      "blocked indefinitely in an MVar operation"
1095  - instance Show BlockedIndefinitelyOnSTM is now
1096      "blocked indefinitely in an STM transaction"
1097 
1098 clients using Control.OldException will be unaffected (the new
1099 exceptions are mapped to the old names).  However, for base4-compat
1100 we'll need to make a version of catch/try that does a similar
1101 mapping.
1102]
1103[Fix unicode conversion for MSB architectures
1104Ben.Lippmeier@anu.edu.au**20090830130028
1105 This fixes the SPARC/Solaris build.
1106]
1107[Fix #3441: detect errors in partial sequences
1108Simon Marlow <marlowsd@gmail.com>**20090830075909
1109 Ignore-this: d12a75d95e0cae5eb1555266810ec281
1110]
1111[Fix hWaitForInput
1112Simon Marlow <marlowsd@gmail.com>**20090827152116
1113 Ignore-this: 2550e911f1a4d4357a5aa8d1764238ce
1114 It was erroneously waiting when there were bytes to decode waiting in
1115 the byte buffer.
1116]
1117[fix debugging code
1118Simon Marlow <marlowsd@gmail.com>**20090827150628
1119 Ignore-this: e1c82fdc19a22e247cd69ff6fa11921d
1120]
1121[Allow for configurable iconv include and library locations.
1122Matthias Kilian <kili@outback.escape.de>**20090826154406
1123 Ignore-this: be95fab611a5534cf184b508964ed498
1124 This should help to fix the build on OpenBSD.
1125]
1126[typo in comment
1127Simon Marlow <marlowsd@gmail.com>**20090826085252
1128 Ignore-this: 1903ee0f354157a6ed3871c100f6b1b9
1129]
1130[un-hide some modules from the Haddock docs
1131Simon Marlow <marlowsd@gmail.com>**20090825152457
1132 Ignore-this: dce6606f93cf977fb24ebe99082dfa62
1133]
1134[Apply fix for #1548, from squadette@gmail.com
1135Simon Marlow <marlowsd@gmail.com>**20090819120700
1136 Ignore-this: 31c237c46a6445f588ed4b8c51bb6231
1137]
1138[improvements to Data.Fixed: instances for Typeable and Data, more predefined types
1139Ashley Yakeley <ashley@semantic.org>**20090812055058
1140 Ignore-this: feeece36d5632f02a05d137d2a39ab78
1141]
1142[Fix "Cabal check" warnings
1143Ian Lynagh <igloo@earth.li>**20090811215856]
1144[Add a GHC.Constants module; fixes trac #3094
1145Ian Lynagh <igloo@earth.li>**20090809183252]
1146[Apply proposal #3393
1147Ian Lynagh <igloo@earth.li>**20090809134717
1148 Add openTempFileWithDefaultPermissions and
1149 openBinaryTempFileWithDefaultPermissions.
1150]
1151[Add some more C wrappers; patch from Krister Walfridsson
1152Ian Lynagh <igloo@earth.li>**20090807200631
1153 Fixes 21 testsuite errors on NetBSD 5.99.
1154]
1155[Fixing configure for autoconf 2.64
1156Alexander Dunlap <alexander.dunlap@gmail.com>**20090805060748
1157 Ignore-this: 992ab91ae3d68c12dbb265776e33e243
1158]
1159[add INLINE toList
1160Ross Paterson <ross@soi.city.ac.uk>**20090806142853
1161 Ignore-this: aba16aabb17d5dca44f15d188945680e
1162 
1163 In anticipation of the fixing of #2353.
1164]
1165[fix a copyright
1166Simon Marlow <marlowsd@gmail.com>**20090805134045
1167 Ignore-this: b0ffbdd38fbba121e8bcba37c4082a60
1168]
1169[Tweak the BufferedIO class to enable a memory-mapped file implementation
1170Simon Marlow <marlowsd@gmail.com>**20090805134036
1171 Ignore-this: ec67d7a0a6d977438deaa342503f77e0
1172 We have to eliminate the assumption that an empty write buffer can be
1173 constructed by setting the buffer pointers to zero: this isn't
1174 necessarily the case when the buffer corresponds to a memory-mapped
1175 file, or other in-memory device implementation.
1176]
1177[Deprecate Control.OldException
1178Ian Lynagh <igloo@earth.li>**20090804143910]
1179[Windows build fix, following RTS tidyup
1180Simon Marlow <marlowsd@gmail.com>**20090803131121
1181 Ignore-this: ce862fb91c2b234211a8757f98690778
1182]
1183[Updates to follow the RTS tidyup
1184Simon Marlow <marlowsd@gmail.com>**20090801220743
1185 Ignore-this: 6e92412df93a66c12d75344053d5634
1186 C functions like isDoubleNaN moved here (primFloat.c)
1187]
1188[Add integer-simple as a build option
1189Ian Lynagh <igloo@earth.li>**20090722013151]
1190[Use shift[LR]Integer in the Bits Integer instance
1191Ian Lynagh <igloo@earth.li>**20090721222440]
1192[depend directly on integer-gmp, rather than indirecting through integer
1193Ian Lynagh <igloo@earth.li>**20090721185228]
1194[Move the instances of Functor and Monad IO to GHC.Base, to avoid orphans
1195Simon Marlow <marlowsd@gmail.com>**20090722102130
1196 Ignore-this: a7d85ac0025d559674249de0108dbcf4
1197]
1198[move "instance Exception Dynamic" so it isn't an orphan
1199Simon Marlow <marlowsd@gmail.com>**20090721093854
1200 Ignore-this: 5ede91ecfec2112c91b699d4de87cd02
1201]
1202[Improve the index checking for array accesses; fixes #2120 #2669
1203Ian Lynagh <igloo@earth.li>**20090719153228
1204 As well as checking that offset we are reading is actually inside the
1205 array, we now also check that it is "in range" as defined by the Ix
1206 instance. This fixes confusing behaviour (#2120) and improves some error
1207 messages (#2669).
1208]
1209[Make chr say what its argument was, if it's a bad argument
1210Ian Lynagh <igloo@earth.li>**20090718151049]
1211[remove unused warning
1212Simon Marlow <marlowsd@gmail.com>**20090715124416
1213 Ignore-this: 31f613654089d0f4a44363946087b41e
1214]
1215[warning fix: -fno-implicit-prelude -> -XNoImplicitPrelude
1216Simon Marlow <marlowsd@gmail.com>**20090715122839
1217 Ignore-this: dc8957249731d5bcb71c01899e5adf2b
1218]
1219[Add hGetEncoding :: Handle -> IO (Maybe TextEncoding)
1220Simon Marlow <marlowsd@gmail.com>**20090715122519
1221 Ignore-this: 14c3eff996db062da1199739781e4708
1222 as suggested during the discussion on the libraries list
1223]
1224[Add more documentation to mkTextEncoding
1225Simon Marlow <marlowsd@gmail.com>**20090715122414
1226 Ignore-this: 97253b2624267df3a246a18121e8ea81
1227 noting that "//IGNORE" and "//TRANSLIT" suffixes can be used with GNU
1228 iconv.
1229]
1230[Add the utf8_bom codec
1231Simon Marlow <marlowsd@gmail.com>**20090715122257
1232 Ignore-this: 1c9396cd805201fe873a39382ced79c7
1233 as suggested during the discussion on the libraries list.
1234]
1235[Export Unicode and newline functionality from System.IO; update Haddock docs
1236Simon Marlow <marlowsd@gmail.com>**20090713113104
1237 Ignore-this: c3f017a555335aa55d106253393f72e2
1238]
1239[add a comment about the non-workingness of CHARBUF_UTF16
1240Simon Marlow <marlowsd@gmail.com>**20090707124406
1241 Ignore-this: 98d00411b68d688b3b4cffc9507b1f35
1242]
1243[Fix build on Windows
1244Ian Lynagh <igloo@earth.li>**20090711004351]
1245[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1246Ian Lynagh <igloo@earth.li>**20090710204513]
1247[Use throwErrnoIfMinus1_ when calling getrusage
1248Ian Lynagh <igloo@earth.li>**20090710204221]
1249[Remove an unused import
1250Ian Lynagh <igloo@earth.li>**20090710153345]
1251[reportStackOverflow now returns IO ()
1252Ian Lynagh <igloo@earth.li>**20090710153257
1253 It used to do "return undefined" to return IO a.
1254]
1255[GHC.Conc.reportError now returns IO ()
1256Ian Lynagh <igloo@earth.li>**20090710152646
1257 It used to return IO a, by "return undefined".
1258]
1259[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1260Ian Lynagh <igloo@earth.li>**20090710152526]
1261[Minor SampleVar refactoring
1262Ian Lynagh <igloo@earth.li>**20090710151438]
1263[Fix "warn-unused-do-bind" warnings in GHC/IO/Handle/Text.hs
1264Ian Lynagh <igloo@earth.li>**20090710122905]
1265[Fix some "warn-unused-do-bind" warnings where we just want to ignore the result
1266Ian Lynagh <igloo@earth.li>**20090710005638]
1267[Use the result of writeCharBuf in GHC/IO/Encoding/Latin1.hs too
1268Ian Lynagh <igloo@earth.li>**20090710004032]
1269[Minor code tidyups in GHC.Conc
1270Ian Lynagh <igloo@earth.li>**20090710003801]
1271[Fix "warn-unused-do-bind" warning in GHC.Conc
1272Ian Lynagh <igloo@earth.li>**20090710003530
1273 If we fail to communicate with the IO manager then we print a warning
1274 using debugErrLn from the ghc-prim package.
1275]
1276[Fix "warn-unused-do-bind" warnings in System.Posix.Internals
1277Ian Lynagh <igloo@earth.li>**20090709164546]
1278[Fix "warn-unused-do-bind" warnings where we really do want to ignore the result
1279Ian Lynagh <igloo@earth.li>**20090709163912]
1280[Add back imports needed on Windows
1281Ian Lynagh <igloo@earth.li>**20090707181924]
1282[Remove unused imports
1283Ian Lynagh <igloo@earth.li>**20090707115810]
1284[Remove unused imports from base
1285simonpj@microsoft.com**20090706111842
1286 Ignore-this: f9b5f353e3bb820f787c56d615b28765
1287 
1288 These unused imports are detected by the new unused-import code
1289 
1290]
1291[Use the result of writeCharBuf
1292Simon Marlow <marlowsd@gmail.com>**20090706133303
1293 Ignore-this: 52288dd559bf4c4f313df6197091d935
1294   
1295 This only makes a difference when CHARBUF_UTF16 is in use, which it
1296 normally isn't.  I suspect CHARBUF_UTF16 doesn't currently work for
1297 other reasons (CHARBUF_UTF16 was an experiment before I wrote the
1298 GHC.IO.Encoding.UTF* codecs), but this patch at least makes it
1299 slightly closer to working.
1300]
1301[Remove some cruft from Data.HashTable
1302Ian Lynagh <igloo@earth.li>**20090706181630]
1303[Add 'eof' to Text.ParserCombinators.ReadP
1304simonpj@microsoft.com**20090706111801
1305 Ignore-this: 2aea7b848e00c894761bc4011adaa95d
1306 
1307 Add a ReadP parser that succeeds at the end of input. Very useful!
1308 
1309]
1310[Don't export CLDouble for GHC; fixes trac #2793
1311Ian Lynagh <igloo@earth.li>**20090705155120
1312 We never really supported CLDouble (it was a plain old double underneath),
1313 and pretending that we do does more harm than good.
1314]
1315[a byte between 0x80 and 0xBF is illegal immediately (#3341)
1316Simon Marlow <marlowsd@gmail.com>**20090702081415
1317 Ignore-this: dc19ef59a1a21118d5a7dd38aa2f611c
1318]
1319[avoid a warning
1320Simon Marlow <marlowsd@gmail.com>**20090630084134
1321 Ignore-this: c92a45ee216faf01327feae9fe06d6e2
1322]
1323[Add a wrapper for libiconv.
1324Matthias Kilian <kili@outback.escape.de>**20090629183634
1325 Ignore-this: 23c6047c0d71b745b495cc223574a47f
1326]
1327[#include <sys/times.h> if we have it (should fix build problems)
1328Simon Marlow <marlowsd@gmail.com>**20090629085351
1329 Ignore-this: a35e93b37ca9595c73460243180f4b9d
1330]
1331[set binary mode for existing FDs on Windows (fixes some GHCi test failures)
1332Simon Marlow <marlowsd@gmail.com>**20090626120522
1333 Ignore-this: 580cf636e9c77d8427aff6861d089481
1334]
1335[Move directory-related stuff to the unix package
1336Simon Marlow <marlowsd@gmail.com>**20090625120325
1337 Ignore-this: b997b3cbce0a46ca87ad825bbdc0a411
1338 now that it isn't used on Windows any more.
1339]
1340[TAG 2009-06-25
1341Ian Lynagh <igloo@earth.li>**20090625160056]
1342Patch bundle hash:
1343e977fa2e1cce48998eb714768415319ef4d90482