Ticket #4228: 4228.patch

File 4228.patch, 44.6 KB (added by norriscm, 4 years ago)

FIX #4228 (use alternate formula for atanh)

Line 
1Mon Aug 16 17:36:54 EDT 2010  ghc@cainnorris.net
2  * FIX #4228 (atanh (-1) returns NaN instead of -Infinity)
3
4We can get the desired behavior without explicitly handling the branch
5cuts by using a different, but mathematically equivalent, expression
6for atanh, namely
7 > atanh x = 0.5 * log ((1.0+x) / (1.0-x))
8
9This patch changes the Floating instances for Double, Float, and
10Complex a to use the implementation above.
11
12
13New patches:
14
15[FIX #4228 (atanh (-1) returns NaN instead of -Infinity)
16ghc@cainnorris.net**20100816213654
17 Ignore-this: dee89c24493e84a02bea711a1c83a73f
18] {
19hunk ./Data/Complex.hs 201
20 
21     asinh z        =  log (z + sqrt (1+z*z))
22     acosh z        =  log (z + (z+1) * sqrt ((z-1)/(z+1)))
23-    atanh z        =  log ((1+z) / sqrt (1-z*z))
24+    atanh z        =  0.5 * log ((1.0+z) / (1.0-z))
25hunk ./GHC/Float.lhs 252
26 
27     asinh x = log (x + sqrt (1.0+x*x))
28     acosh x = log (x + (x+1.0) * sqrt ((x-1.0)/(x+1.0)))
29-    atanh x = log ((x+1.0) / sqrt (1.0-x*x))
30+    atanh x = 0.5 * log ((1.0+x) / (1.0-x))
31 
32 instance  RealFloat Float  where
33     floatRadix _        =  FLT_RADIX        -- from float.h
34hunk ./GHC/Float.lhs 334
35 
36     asinh x = log (x + sqrt (1.0+x*x))
37     acosh x = log (x + (x+1.0) * sqrt ((x-1.0)/(x+1.0)))
38-    atanh x = log ((x+1.0) / sqrt (1.0-x*x))
39+    atanh x = 0.5 * log ((1.0+x) / (1.0-x))
40 
41 {-# RULES "truncate/Double->Int" truncate = double2Int #-}
42 instance  RealFrac Double  where
43}
44
45Context:
46
47[export allocaBytesAligned; make allocaArray use the correct alignment (#2917)
48Simon Marlow <marlowsd@gmail.com>**20100812105524
49 Ignore-this: deb6495f7b7b84deaf02b88927a5ba8c
50]
51[deprecate unGetChan and isEmptyChan (see #4154)
52Simon Marlow <marlowsd@gmail.com>**20100705125952
53 Ignore-this: b4e769959f131b2d0001eb7202bc1b92
54]
55[Add type signatures to cope with lack of local generalisation
56simonpj@microsoft.com**20100728124847
57 Ignore-this: d3af9a47c2821c6081bde05a135a92fb
58]
59[Add type signature in local where
60simonpj@microsoft.com**20100727151532
61 Ignore-this: 1c57063ad32d13e0d1ec8daf968bf055
62]
63[Integrated new I/O manager
64Simon Marlow <marlowsd@gmail.com>**20100810082248
65 Ignore-this: ed70a9066ac9b676a446fe99978fef7a
66 (patch originally by Johan Tibell <johan.tibell@gmail.com>, minor merging by me)
67]
68[Add mfilter to Control.Monad
69jon.fairbairn@cl.cam.ac.uk**20090917145616
70 Ignore-this: de4240b60684f3065b29378df3ea98f2
71 
72 Straightforward MonadPlus version of List.filter. I would
73 prefer to call it filter, but the current naming scheme for
74 Control.Monad implies mfilter.
75 
76]
77[move Monad and MonadFix instances for Either from mtl (proposal #4159)
78Ross Paterson <ross@soi.city.ac.uk>**20100729122449
79 Ignore-this: b0f8cd8643679948d1da43bd7c08c5aa
80 
81 The Monad and MonadFix instances for Either (formerly in the mtl
82 package) are moved to Control.Monad.Instances and Control.Monad.Fix
83 respectively.  The Monad instance is still an orphan, to retain Haskell
84 98 compatibility, but the MonadFix instance is together with its class.
85 The Error constraint is removed from both instances, and the default
86 definition of fail is used.
87]
88[Remove egregious ghc-ish from Foreign.Marshal
89Malcolm.Wallace@me.com**20100722075449]
90[add numSparks :: IO Int (#4167)
91Simon Marlow <marlowsd@gmail.com>**20100720153858
92 Ignore-this: 4543f57a7f137f8cae1c3efc5c023a9b
93]
94[add unsafeLocalState from Haskell 2010, and docs
95Simon Marlow <marlowsd@gmail.com>**20100720082819
96 Ignore-this: dcd79fb546ebe29ddff4df279ec2f38
97]
98[docs: mention that Foreign.unsafePerformIO is deprecated
99Simon Marlow <marlowsd@gmail.com>**20100720082804
100 Ignore-this: 4cfebb8f2a1cddc7d15e94e31b2befa4
101 We can't actually deprecate it without introducing a name clash
102 between Foreign.unsafePerformIO and System.IO.Unsafe.unsafePerformIO
103]
104[doc formatting fix
105Simon Marlow <marlowsd@gmail.com>**20100714151347
106 Ignore-this: 255edef607dcd290e198015240b5d125
107]
108[add module intro from Haskell 2010
109Simon Marlow <marlowsd@gmail.com>**20100714115853
110 Ignore-this: 59b5a07507a059ccccdff2dfb6490a27
111]
112[document exception-overriding behaviour in withFile
113Simon Marlow <marlowsd@gmail.com>**20100714104107
114 Ignore-this: f99e641ea2f46d872cb7420a62fa50dc
115]
116[doc: use "finalizer" consistently
117Simon Marlow <marlowsd@gmail.com>**20100714103649
118 Ignore-this: bdfea40f31dc5045fdbc6e12266dda93
119]
120[clarify meaning of bit
121Simon Marlow <marlowsd@gmail.com>**20100714103310
122 Ignore-this: 521b031f1e83ef34ca03d9aa9273df8a
123]
124[note shortcutting behaviour of any/all/elem
125Simon Marlow <marlowsd@gmail.com>**20100714103304
126 Ignore-this: 1605f362ba0712ad1cea1309636f3ea1
127]
128[add cast{C,U}CharToChar and castCharTo{C,U}Char, from Haskell 2010
129Simon Marlow <marlowsd@gmail.com>**20100713132515
130 Ignore-this: 9b1da827016c7b08668078b45964e9de
131]
132[mention that IntPtr and WordPtr can be marshalled to/from intptr_t and uintptr_t
133Simon Marlow <marlowsd@gmail.com>**20100713132403
134 Ignore-this: dcc112a72746ba117a84fa29e71b6800
135]
136[Partial fix for Trac #4136
137simonpj@microsoft.com**20100707135725
138 Ignore-this: 9548eeb3187d9779d4e5c858a0f35354
139 
140 In 'choose' (which is a library function designed specifically
141 to support derived instances of Read), we must match Symbol
142 as well as Ident, for nullary constructors that (wierdly) are
143 symbols.
144]
145[Fix typo in documentation
146Simon Hengel <simon.hengel@wiktory.org>**20100711141648
147 Ignore-this: c052dd8a681832ef598a323ad55eae3a
148]
149[Remove duplicated word in documentation
150Simon Hengel <simon.hengel@wiktory.org>**20100711072703
151 Ignore-this: fb3732dc57be55f14168792f923433
152]
153[Allow nhc98 to cope with recent changes to Control.Exception.
154Malcolm.Wallace@me.com**20100710170940]
155[ New asynchronous exception control API (base parts)
156Simon Marlow <marlowsd@gmail.com>**20100708152735
157 Ignore-this: 71a4811804f04259f1fe739f8863beaf
158   
159 As discussed on the libraries/haskell-cafe mailing lists
160   http://www.haskell.org/pipermail/libraries/2010-April/013420.html
161 
162 This is a replacement for block/unblock in the asychronous exceptions
163 API to fix a problem whereby a function could unblock asynchronous
164 exceptions even if called within a blocked context.
165 
166 The new terminology is "mask" rather than "block" (to avoid confusion
167 due to overloaded meanings of the latter).
168 
169 The following is the new API; the old API is deprecated but still
170 available for the time being.
171 
172 Control.Exception
173 -----------------
174 
175 mask  :: ((forall a. IO a -> IO a) -> IO b) -> IO b
176 mask_ :: IO a -> IO a
177 
178 uninterruptibleMask :: ((forall a. IO a -> IO a) -> IO b) -> IO b
179 uninterruptibleMask_ :: IO a -> IO
180 
181 getMaskingState :: IO MaskingState
182 
183 data MaskingState
184   = Unmasked
185   | MaskedInterruptible
186   | MaskedUninterruptible
187 
188 
189 Control.Concurrent
190 ------------------
191 
192 forkIOUnmasked :: IO () -> IO ThreadId
193]
194[Async-exception safety, and avoid space leaks
195Simon Marlow <marlowsd@gmail.com>**20100708145819
196 Ignore-this: dbfd0e61551e9e7b4fc1c6fe9b9a83de
197 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
198 Modified slightly by me to remove non-functional changes.
199]
200[Async-exception safety, and avoid space leaks
201Simon Marlow <marlowsd@gmail.com>**20100708103154
202 Ignore-this: 190c3ac8f6633231624da8cf1316588
203 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
204 Modified slightly by me to remove non-functional changes.
205]
206[Fix a few places where we forgot to close the text codecs (#4029)
207Simon Marlow <marlowsd@gmail.com>**20100702130210
208 Ignore-this: 2e81a4b4cb343181cef34b0f9e2ded47
209 Each time you invoke :load in GHCi it resets the CAFs, including
210 stdin/stdout/stderr, and each of these was allocating a new iconv_t.
211]
212[remove docs from Monad that belonged on the instance for MonadPlus IO
213Simon Marlow <marlowsd@gmail.com>**20100701154203
214 Ignore-this: 59df02542a7ac9421552a2155d848d27
215]
216[docs: unqualify Prelude.IO
217Simon Marlow <marlowsd@gmail.com>**20100701153817
218 Ignore-this: 73b0202876c827e7a5b4a5ce74e724c4
219]
220[unqualify Float and Double
221Simon Marlow <marlowsd@gmail.com>**20100701142727
222 Ignore-this: cbe89d31a00bf49996a33933324fca17
223]
224[extract information about Data.Time from docs for CTime
225Simon Marlow <marlowsd@gmail.com>**20100701142415
226 Ignore-this: c48c9609b8d36e43e033a7bea81d6f17
227]
228[doc typo
229Simon Marlow <marlowsd@gmail.com>**20100701142354
230 Ignore-this: 17a1fd703831c888975ff63fbfa3a9b2
231]
232[peekArray docs: remove mentions of "this version" and "previous version"
233Simon Marlow <marlowsd@gmail.com>**20100701125333
234 Ignore-this: 39a744874258670bd935ba9e38390939
235]
236[doc typo
237Simon Marlow <marlowsd@gmail.com>**20100701124154
238 Ignore-this: 98f5c286e38c2c34c96b05d5e8bc5ad9
239]
240[doc typo
241Simon Marlow <marlowsd@gmail.com>**20100701124128
242 Ignore-this: 10a4314ec7aed336701fc616fb574ebc
243]
244[doc typo
245Simon Marlow <marlowsd@gmail.com>**20100701123715
246 Ignore-this: c4909a7bf7163460ee5d32f58812041e
247]
248[doc wibble: Haskell 98 I/O Error -> 'IOError'
249Simon Marlow <marlowsd@gmail.com>**20100701123612
250 Ignore-this: bf373df781acbc575e4ffe3b7e6059ae
251]
252[doc typo
253Simon Marlow <marlowsd@gmail.com>**20100701123014
254 Ignore-this: 16aaccae48ef3101adf78ea5b0d5a8fd
255]
256[Haddock hacks to fix whitespace consistency
257Simon Marlow <marlowsd@gmail.com>**20100701121631
258 Ignore-this: 61c58dec52a31fd2d3f331a87d2f903f
259]
260[use '==' consistently rather than '->' in examples
261Simon Marlow <marlowsd@gmail.com>**20100701121616
262 Ignore-this: 472b0a05a85d34d9712186040e1636d9
263]
264[doc wibble: remove confusing mention of "Prelude"
265Simon Marlow <marlowsd@gmail.com>**20100701113308
266 Ignore-this: 232283d0096d01cd45e9b3c5c1e63a6d
267]
268[doc wibble: nonstrict -> non-strict
269Simon Marlow <marlowsd@gmail.com>**20100701113253
270 Ignore-this: 4264f0ab23a0835fc13c6e8601d6b743
271]
272[doc whitespace
273Simon Marlow <marlowsd@gmail.com>**20100701112242
274 Ignore-this: 777a95b1d1140c61d3ab95d5eb5809e7
275]
276[move the doc for 'Char' to its new home in ghc-prim:GHC.Types
277Simon Marlow <marlowsd@gmail.com>**20100629134150
278 Ignore-this: 7687db0077a29498349bfb4b44983985
279]
280[doc wibble
281Simon Marlow <marlowsd@gmail.com>**20100629122608
282 Ignore-this: 9a909e5d015332dc445bd9592e6e386d
283]
284[doc updates in System.IO
285Simon Marlow <marlowsd@gmail.com>**20100629122118
286 Ignore-this: 2257ec1cc4cdb8b7804cfa1f3cf32753
287]
288[doc wibble
289Simon Marlow <marlowsd@gmail.com>**20100625134858
290 Ignore-this: 64c50f29df6c389273b818918fe7033a
291]
292[doc wibbles
293Simon Marlow <marlowsd@gmail.com>**20100624154614
294 Ignore-this: b364aad53beea6e741fee2824459b6e8
295]
296[Fix haddock formatting
297Ian Lynagh <igloo@earth.li>**20100625222623]
298[Give nub's complexity in the haddock docs; fixes #4086
299Ian Lynagh <igloo@earth.li>**20100625222059]
300[correct docs for exitWith: only stdout/stderr are flushed, not all Handles
301Simon Marlow <marlowsd@gmail.com>**20100624130506
302 Ignore-this: 33a938dad8f0bc061572e2ec571cacc7
303]
304[fix docs for isSpace
305Simon Marlow <marlowsd@gmail.com>**20100624130444
306 Ignore-this: b35ff080dbb9833176f08e39dbd9ff6d
307]
308[make the hGetBuf/hPutBuf family work with non-FD Handles (#4144)
309Simon Marlow <marlowsd@gmail.com>**20100624130425
310 Ignore-this: 8200f0208a9b1b1cf4824f343d75819a
311]
312[nit in docs for accumArray
313Simon Marlow <marlowsd@gmail.com>**20100622121131
314 Ignore-this: c066a456c40907e767df10c3990f35ff
315]
316[add doc for the ExitCode type
317Simon Marlow <marlowsd@gmail.com>**20100622120930
318 Ignore-this: 99c34332be7f3565da844528b470054a
319]
320[remove extraneous info from docs for Array
321Simon Marlow <marlowsd@gmail.com>**20100622120921
322 Ignore-this: e2a3f5e84fc23eb7bae911f0680e805e
323]
324[add an INLINE to the list version of traverse, to enable fusion
325Simon Marlow <marlowsd@gmail.com>**20100608082531
326 Ignore-this: ea98cdc3308b406bb04c0f7a38c4424b
327]
328[Don't define the C localeEncoding on Windows
329Ian Lynagh <igloo@earth.li>**20100620202342
330 Ignore-this: c4992f6832a391b0cccc5a9b7d643976
331 (it causes warnings, and isn't used)
332]
333[add Applicative instance for Either (proposal #4095)
334Ross Paterson <ross@soi.city.ac.uk>**20100617225110
335 Ignore-this: 50262ec4700dc16efec5755be5b308c5
336 
337 This is not the only possible instance for Either, but this one is
338 compatible with the usual Monad instance.
339]
340[Use libcharset instead of nl_langinfo(CODESET) if possible.
341pho@cielonegro.org**20100519013112
342 Ignore-this: 4c1e278e022a3d276848afc1dcba4425
343 
344 nl_langinfo(CODESET) doesn't always return standardized variations of the encoding names. Use libcharset if possible, which is shipped together with GNU libiconv.
345]
346[Add a note about the interruptibility of throwTo.
347Simon Marlow <marlowsd@gmail.com>**20100615112720
348 Ignore-this: ae9fabe95310d7c364e95f7784793485
349]
350[docs: note that hGetBufNonBlocking isn't non-blocking on Windows
351Simon Marlow <marlowsd@gmail.com>**20100615112547
352 Ignore-this: 4f3e5213e142149affe08c5123d6efea
353]
354[don't depend on Prelude (#4122)
355Simon Marlow <marlowsd@gmail.com>**20100615105631
356 Ignore-this: 1a3fd49b103fe31cbb453f302c18767f
357]
358[Don't depend on Prelude (#4123)
359Simon Marlow <marlowsd@gmail.com>**20100615105401
360 Ignore-this: cc7616d85a1637bc7621b4f2bc181c0e
361]
362[bump version to 4.3.0.0, added instance MonadPlus STM
363Simon Marlow <marlowsd@gmail.com>**20100601144831
364 Ignore-this: 7c3cf7574499c4267372493f2636dc0
365]
366[Moved MonadPlus instance for STM from Control.Monad.STM to GHC.Conc to avoid an orphaned instance
367Bas van Dijk <v.dijk.bas@gmail.com>**20100516160651
368 Ignore-this: 651b852942b2fae2b93f996e39239b8f
369]
370[Added Applicative and Alternative instances for STM
371Bas van Dijk <v.dijk.bas@gmail.com>**20100516171756
372 Ignore-this: 567003bc4040bc97105cda4d31ebf04a
373]
374[expand Foldable instance for Array
375Ross Paterson <ross@soi.city.ac.uk>**20100602212154
376 Ignore-this: 9bd9e9666a9400431eb92352244fe7e7
377]
378[doc comment illustrating Foldable(foldr)
379Ross Paterson <ross@soi.city.ac.uk>**20100527150833
380 Ignore-this: 8f27d889379803f3ba86d6e928428f3c
381]
382[fix syntax in doc comments
383Ross Paterson <ross@soi.city.ac.uk>**20100527150757
384 Ignore-this: cb78da51d60ff6863dc395f1a892c103
385]
386[export hGetBufSome (#4046)
387Simon Marlow <marlowsd@gmail.com>**20100520093538
388 Ignore-this: f467fad9722e27edfad6b3dd75290e7b
389]
390[hWaitForInput: don't try to read from the device (#4078)
391Simon Marlow <marlowsd@gmail.com>**20100517133741
392 Ignore-this: 55ec33b03397380259b91e4ca62207a6
393 readTextDeviceNonBlocking is not non-blocking on Windows
394]
395[hSetEncoding: change the encoding on both read and write sides (#4066)
396Simon Marlow <marlowsd@gmail.com>**20100514124628
397 Ignore-this: 5b9e9caef06356d0296c584159709ebb
398]
399[Correct haddock formatting.
400Adam Vogt <vogt.adam@gmail.com>**20100423022103
401 Ignore-this: d2622339302048fda48080f7d5ce4a2f
402]
403[Fix for hGetBufSome
404Simon Marlow <marlowsd@gmail.com>**20100505135637
405 Ignore-this: 2019680f8fb223956cacfcf0d046f133
406]
407[improve the documentation for throwTo and killThread (#3884)
408Simon Marlow <marlowsd@gmail.com>**20100505135600
409 Ignore-this: ce881d96ddb729acb6ca09c779975e7d
410]
411[elaborate the docs for unsafePerformIO a bit
412Simon Marlow <marlowsd@gmail.com>**20100505101249
413 Ignore-this: 1cec3f67560b672c64c5a0dcf9a79eb7
414]
415[add Typeable instance
416Simon Marlow <marlowsd@gmail.com>**20100504152815
417 Ignore-this: 6d9cf9d62f0ef17fa459bf213a04098
418]
419[Add hGetBufSome, like hGetBuf but can return short reads
420Simon Marlow <marlowsd@gmail.com>**20100504152759
421 Ignore-this: 195c905b43f8d9505029364e2c5b18e
422]
423[Add swap (#3298)
424Simon Marlow <marlowsd@gmail.com>**20100504095339
425 Ignore-this: 13b007dc4594ce252997ec6fa0bbd976
426]
427[inline allocaArray0, to fix withCString benchmark
428Simon Marlow <marlowsd@gmail.com>**20100423124729
429 Ignore-this: 35c96816acc2f3aaf9dd29f7995fa6f0
430]
431[raise asynchronous exceptions asynchronously (#3997)
432Simon Marlow <marlowsd@gmail.com>**20100421094932
433 Ignore-this: 6d987d93d382c0f69c68c326312abd6b
434]
435[add NOINLINE pragmas for stdin/stdout/stderr
436Simon Marlow <marlowsd@gmail.com>**20100421082041
437 Ignore-this: 3fc130268ec786f28d945858d6690986
438]
439[INLINE alloca and malloc
440Simon Marlow <marlowsd@gmail.com>**20100419135333
441 Ignore-this: b218bd611f18721b1505a8c0b9e6a16a
442 See discussion on glasgow-haskell-users:
443   http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018740.html
444]
445[Move comment closer to the offending line
446Matthias Kilian <kili@outback.escape.de>**20100419155421
447 Ignore-this: b34a1d7affd66f67d210df2377b585d9
448]
449[Ignore the return code of c_fcntl_write again
450Matthias Kilian <kili@outback.escape.de>**20100415140452
451 Ignore-this: 266d8ba02cc3cb79c85629b3528261c9
452 
453 The return code has been ignored in the past on purpose, because
454 O_NONBLOCK will fail on BSDs for some special files. This fixes the
455 problem mentioned in
456 http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018698.html
457 
458]
459[Fix bitrot in IO debugging code
460Ian Lynagh <igloo@earth.li>**20100413134339
461 Also switched to using Haskell Bools (rather than CPP) to en/disable it,
462 so it shouldn't break again in the future.
463]
464[Tiny code tidy-up
465Ian Lynagh <igloo@earth.li>**20100413011147]
466[remove old/wrong comment
467Simon Marlow <marlowsd@gmail.com>**20100325161403
468 Ignore-this: e6e377d44af48c4162d17d55bdf3f821
469]
470[withThread: block asynchronous exceptions before installing exception handler.
471Bas van Dijk <v.dijk.bas@gmail.com>**20100329131624
472 Ignore-this: be5aeb47dbd73807b5f94df11afbb81c
473 Note that I don't unblock the given io computation. Because AFAICS
474 withThread is only called with 'waitFd' which only performs an FFI
475 call which can't receive asynchronous exceptions anyway.
476]
477[runInUnboundThread: block asynchronous exceptions before installing exception handler
478Bas van Dijk <v.dijk.bas@gmail.com>**20100329131549
479 Ignore-this: a00c5e32fe3981ff87bedd367a69051e
480]
481[fix the deprecation message (GHC.IO.Handle.Base -> GHC.IO.Handle)
482Simon Marlow <marlowsd@gmail.com>**20100330121137
483 Ignore-this: 4ca8500a01ac93454507aa8f9dd001f9
484]
485[Make SampleVar an abstract newtype
486Bas van Dijk <v.dijk.bas@gmail.com>**20100318200349
487 Ignore-this: 27939e2a064b75e71cb146117346be30
488]
489[Fix bugs regarding asynchronous exceptions and laziness in Control.Concurrent.SampleVar
490Bas van Dijk <v.dijk.bas@gmail.com>**20100318200104
491 Ignore-this: 7376b2a3afe155daf233a8f1ddc0a7a
492  - Block asynchronous exceptions at the right places
493  - Force thunks before putting them in a MVar
494]
495[Write the thunk 'next' to the MVar
496Bas van Dijk <v.dijk.bas@gmail.com>**20100319125951
497 Ignore-this: dd25636cf220131385ff2fd32493d456
498]
499[change to use STM, fixing 4 things
500Simon Marlow <marlowsd@gmail.com>**20100318104436
501 Ignore-this: 551d30280a7941c08f5c3b14576bdd70
502   1. there was no async exception protection
503   2. there was a space leak (now new value is strict)
504   3. using atomicModifyIORef would be slightly quicker, but can
505      suffer from adverse scheduling issues (see #3838)
506   4. also, the STM version is faster.
507]
508[Tweak docs
509Ian Lynagh <igloo@earth.li>**20100312214129]
510[Fixed dead links in documentation of forkIO
511Bas van Dijk <v.dijk.bas@gmail.com>**20100308222415
512 Ignore-this: 7deb8fd064c867fbede2a6b2e9da4f15
513]
514[Documentation fixes in Control.Exception
515Bas van Dijk <v.dijk.bas@gmail.com>**20100301220442
516 Ignore-this: 761fcba401cbd1f47276ddfc9b5b80f2
517]
518[Plug two race conditions that could lead to deadlocks in the IO manager
519Simon Marlow <marlowsd@gmail.com>**20100225120255
520 Ignore-this: e6983d6b953104d370278ab3e4617e8b
521]
522[FIX #3866: improve documentation of Data.Data.Constr
523jpm@cs.uu.nl**20100224125506
524 Ignore-this: 3818c5d8fee012a3cf322fb455b6e5dc
525]
526[UNDO: Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
527Simon Marlow <marlowsd@gmail.com>**20100223101603
528 Ignore-this: 78becb2d39b3cd9a1a473a5811ca7d92
529]
530[Put the complexity in the length docs. Fixes trac #3680
531Ian Lynagh <igloo@earth.li>**20100221191425]
532[nhc98 should build Data.Functor.
533Malcolm.Wallace@cs.york.ac.uk**20100221163218]
534[Update the exitWith docs
535Ian Lynagh <igloo@earth.li>**20100213140004
536 Error pointed out by Volker Wysk <vw@volker-wysk.de>
537]
538[Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
539Simon Marlow <marlowsd@gmail.com>**20100211101955
540 Ignore-this: 261415363303efca265e80290eac5f28
541]
542[For nhc98, import unsafeInterleaveIO rather than defining it here.
543Malcolm.Wallace@cs.york.ac.uk**20100204171021]
544[Stifle warning about unused return value
545benl@cse.unsw.edu.au**20100203025537]
546[fix #3832: use the locale encoding in openTempFile
547Simon Marlow <marlowsd@gmail.com>**20100120211830
548 Ignore-this: df4f778cc5fefb32290c798db722632c
549 Also while I was here fix an XXX: the Handle contained an
550 uninformative string like <fd: 4> for error messages rather than the
551 real file path.
552]
553[Fix the build: export void, so it doesn't give an unused binding warning
554Ian Lynagh <igloo@earth.li>**20100116174451]
555[hIsEOF: don't do any decoding (#3808)
556Simon Marlow <marlowsd@gmail.com>**20100112230317
557 Ignore-this: 6a384dd2d547ffe3ad3762920e5c1671
558]
559[Control.Monad: +void :: f a -> f ()
560gwern0@gmail.com**20100108214455
561 Ignore-this: 4dc07452315f2d1b4941903ff42fc45f
562 See http://hackage.haskell.org/trac/ghc/ticket/3292
563 Turns m a -> m (). Lets one call functions for their side-effects without
564 having to get rid of their return values with '>> return ()'. Very useful
565 in many contexts (parsing, IO etc.); particularly good for 'forkIO' and 'forM_',
566 as they demand return types of 'IO ()' though most interesting IO functions
567 return non-().
568]
569[Replace the implementation of mergesort with a 2x faster one.
570Malcolm.Wallace@cs.york.ac.uk**20091224152014
571 See ticket http://hackage.haskell.org/trac/ghc/ticket/2143.
572]
573[Restore previous Data.Typeable.typeOf*Default implementations for non-ghc.
574Malcolm.Wallace@cs.york.ac.uk**20091223142625
575 Not all compilers have ScopedTypeVariables.
576]
577[Add comments about double bounds-checking, and fast paths for rectangular arrays
578simonpj@microsoft.com**20091218165655
579 Ignore-this: ea0849419dc00927aba4bd410b1cc58d
580 
581 See Note [Double bounds-checking of index values] for the details.
582 
583 The fast paths omit the doubled checks for cases we know about
584]
585[Fix Trac #3245: memoising typeOf
586simonpj@microsoft.com**20091218155117
587 Ignore-this: 5a178a7f2222293c5ee0c3c43bd1b625
588 
589 The performance bug in #3245 was caused by computing the typeRep
590 once for each call of typeOf, rather than once for each dictionary
591 contruction.  (Computing TypeReps is reasonably expensive, because
592 of the hash-consing machinery.)
593 
594 This is readily fixed by putting the TypeRep construction outside
595 the lambda.  (Arguably GHC might have worked that out itself,
596 but it involves floating something between a type lambda and a
597 value lambda, which GHC doesn't currently do. If it happens a lot
598 we could fix that.)
599]
600[Mark 'index' as INLINE in GHC.Arr
601simonpj@microsoft.com**20091216170441
602 Ignore-this: a4df9d8acf496c8e0e9ce5a520509a2a
603 
604 This makes indexing much faster. See Trac #1216
605]
606[Comment the remaining orphan instance modules
607Ian Lynagh <igloo@earth.li>**20091206125021]
608[De-orphan Eq/Ord Float/Double
609Ian Lynagh <igloo@earth.li>**20091205181238]
610[Add comments to "OPTIONS_GHC -fno-warn-orphans" pragmas
611Ian Lynagh <igloo@earth.li>**20091205165854]
612[Data.Either.partitionEithers was insufficiently lazy.
613Malcolm.Wallace@cs.york.ac.uk**20091202032807
614 Ignore-this: 77e1b3288f66608c71458d8a91bcbe12
615 Spotted by Daniel Fischer.
616]
617[fix the docs regarding finalizer guarantees
618Simon Marlow <marlowsd@gmail.com>**20091130144409
619 Ignore-this: d1ab9532c74a002b8075ff60febcbe2d
620]
621[x86_64 requires more stack
622Malcolm.Wallace@cs.york.ac.uk**20091201033745]
623[check for size < 0 in mallocForeignPtrBytes and friends (#3514)
624Simon Marlow <marlowsd@gmail.com>**20091125143822
625 Ignore-this: 91077d01da2bbe1dfed5155e8b40da9
626]
627[hGetContents: close the handle properly on error
628Simon Marlow <marlowsd@gmail.com>**20091125123435
629 Ignore-this: bc37ff678acc6e547dc390285e056eb9
630 
631 When hGetContents caught an error it was closing the handle and then
632 throwing the exception, without updating the handle with the new
633 closed state.  This lead to a double-closed, which was the cause of
634 
635 *** glibc detected *** ./Setup: double free or corruption
636 
637 when iconv_close was called twice on the decoder.
638 
639 See http://hackage.haskell.org/trac/hackage/ticket/609
640]
641[Fix arities of mapFB and zipFB
642Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091126232219
643 Ignore-this: c4e14cd0a92622549c86e67237a40865
644]
645[Remove an unnecessary -fno-warn-orphans flag
646Ian Lynagh <igloo@earth.li>**20091126123404]
647[Tweak layout to work with alternative layout rule
648Ian Lynagh <igloo@earth.li>**20091125232349]
649[Tweak layout to be accepted by the alternative layout rul
650Ian Lynagh <igloo@earth.li>**20091125194147]
651[Make sure zipWithFB has arity 2
652Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091125010003
653 Ignore-this: 4cf60c55666f03d22a9f5a6e07f52d36
654 
655 It gets 2 arguments in the "zipWith" rule but its arity was higher and the new
656 inliner didn't inline it sometimes, for instance here:
657 
658 mpp ::  [Double] -> [Double] -> [Double] -> [Double] -> [Double]
659 mpp as bs cs ds = zipWith (*) (zipWith (+) as bs) (zipWith (+) cs ds)
660 
661 This was a regression vs. 6.10.
662]
663[Remove an old comment
664Ian Lynagh <igloo@earth.li>**20091124134647]
665[De-orphan the Eq/Ord Integer instances
666Ian Lynagh <igloo@earth.li>**20091124133639]
667[Whitespace only
668Ian Lynagh <igloo@earth.li>**20091124133421]
669[Derive some more instances, rather than writing them by hand
670Ian Lynagh <igloo@earth.li>**20091124011747]
671[We can now derive Ord ()
672Ian Lynagh <igloo@earth.li>**20091124011416]
673[De-orphan tuple Eq/Ord instances
674Ian Lynagh <igloo@earth.li>**20091123233343]
675[Control.Exception.Base no longer has any orphans
676Ian Lynagh <igloo@earth.li>**20091123224905]
677[De-orphan the MonadFix ST instance for GHC
678Ian Lynagh <igloo@earth.li>**20091123223544]
679[Rearrange the contents of Control.Monad.ST; no functionality changes
680Ian Lynagh <igloo@earth.li>**20091123222702]
681[De-orphan the Eq/Ord [a] instances
682Ian Lynagh <igloo@earth.li>**20091123215635]
683[De-orphan the Eq/Ord Char instances
684Ian Lynagh <igloo@earth.li>**20091123202253]
685[De-orphan the Eq/Ord Bool instances
686Ian Lynagh <igloo@earth.li>**20091123201817]
687[Move Eq/Ord Ordering instances to de-orphan them
688Ian Lynagh <igloo@earth.li>**20091123194310]
689[Remove ffi warnings for nhc98.
690Malcolm.Wallace@cs.york.ac.uk**20091123063743]
691[Second attempt to fix #1185 (forkProcess and -threaded)
692Simon Marlow <marlowsd@gmail.com>**20091111151915
693 Ignore-this: fa5f5d5e4e080d4b612a37244f937f9c
694 
695 Patch 2/2: first patch is to ghc
696 
697 This time without dynamic linker hacks, instead I've expanded the
698 existing rts/Globals.c to cache more CAFs, specifically those in
699 GHC.Conc.  We were already using this trick for signal handlers, I
700 should have realised before.
701 
702 It's still quite unsavoury, but we can do away with rts/Globals.c in
703 the future when we switch to a dynamically-linked GHCi.
704]
705[Rollback #1185 fix
706Simon Marlow <marlowsd@gmail.com>**20091106140629
707 Ignore-this: cd5667e8474e37e01ba26a1984274811
708 
709 rolling back:
710 
711 Tue Nov  3 16:05:40 GMT 2009  Simon Marlow <marlowsd@gmail.com>
712   * Fix #1185: restart the IO manager after fork()
713   
714   This is the libraries/base part of the patch; there is a corresponding
715   patch to GHC itself.
716   
717   The main change is that we now keep track of the IO manager's ThreadId
718   in a top-level MVar, and ensureIOManagerIsRunning checks whether a
719   previous IO manager thread is alive before starting one.  In the child
720   of fork(), we can hence call ensureIOManagerIsRunning to restart the
721   IO manager.
722 
723     M ./GHC/Conc.lhs -46 +44
724 
725 Wed Nov  4 17:49:45 GMT 2009  Ian Lynagh <igloo@earth.li>
726   * Fix the build on Windows
727 
728     M ./GHC/Conc.lhs -6 +4
729]
730[Fix the build on Windows
731Ian Lynagh <igloo@earth.li>**20091104174945]
732[Fix #1185: restart the IO manager after fork()
733Simon Marlow <marlowsd@gmail.com>**20091103160540
734 Ignore-this: 6dc05464f1500104554637f4759738cc
735 
736 This is the libraries/base part of the patch; there is a corresponding
737 patch to GHC itself.
738 
739 The main change is that we now keep track of the IO manager's ThreadId
740 in a top-level MVar, and ensureIOManagerIsRunning checks whether a
741 previous IO manager thread is alive before starting one.  In the child
742 of fork(), we can hence call ensureIOManagerIsRunning to restart the
743 IO manager.
744]
745[improve the documentation for throwErrnoIfRetry
746Simon Marlow <marlowsd@gmail.com>**20091016112404
747 Ignore-this: b77275cacf730e15757946027168f63e
748]
749[Don't inline unpackFoldrCString ever
750simonpj@microsoft.com**20091029135350
751 Ignore-this: 85d672649b1b776efc7e97500b05d4f9
752]
753[Inline more default methods
754simonpj@microsoft.com**20091029135330
755 Ignore-this: 289c44b0afd6d5631c2a4e0664275ca9
756 
757 Namely Monad: (>>)
758        Eq:    (==), (/=)
759        Num:   (-), negate
760        Real:  quot, rem, div, mod, recip, (/), truncate
761        Float: (**), logBase, sqrt, tan, tanh
762]
763[Move error messages out of INLINEd default methods
764simonpj@microsoft.com**20091029135118
765 Ignore-this: 9e35dc947f94827a3529eb53a41575fd
766 
767 No need to duplicate the error generation!
768]
769[Exploit now-working default-method INLINE pragmas for Data.Bits
770simonpj@microsoft.com**20091029135041
771 Ignore-this: 8adf225f31ca7a3181ee087e9e4fe535
772 
773 * Add INLINE pragmas to default methods for class Bits
774 
775 * Remove redundant instance methods elsewhere, now that
776   the default method will do the job
777]
778[Tidy up and comment imports
779simonpj@microsoft.com**20091029134414
780 Ignore-this: bf2be31035de975d8995e988933cc940
781]
782[Inline foldr and (.) when applied to two arguments not three
783simonpj@microsoft.com**20091029134335
784 Ignore-this: fccb6f3e90e15f44cb465814be85ede2
785 
786 The new INLINE story is (by design) arity-sensitive, so we must
787 put fewer argument on the LHS for foldr and (.)
788]
789[dirUtils.c no longer available
790Malcolm.Wallace@cs.york.ac.uk**20091013093833]
791[Make hGetContents throw an exception if an error is encountered
792Simon Marlow <marlowsd@gmail.com>**20091012152955
793 Ignore-this: 9f7a7176193eab25c9daaacd9261f2de
794 
795 Strictly speaking this breaks Haskell 98 compatibility, which requires
796 hGetContents to just end the lazy stream silently if an error is
797 encountered.  However, for a few reasons we think it will make
798 everyone's life a bit easier if we make this change
799 
800  1. Errors will be a lot more common in GHC 6.12.1, in the form
801     of Unicode decoding errors.
802 
803  2. When Haskell 98 was designed, we didn't know how to throw
804     exceptions from inside lazy I/O, but now we do.
805 
806  3. If anyone is actually relying on the previous behaviour, their
807     code is arguably broken.
808]
809[Re-instate System.Console.Getopt for nhc98 builds.
810Malcolm.Wallace@cs.york.ac.uk**20091013092843
811 Although it was split out of base a while back, that change was
812 reverted for ghc soon afterwards, but nhc98 never noticed.
813]
814[Roll back "Another instance of nhc98's strange import semantics."
815Ian Lynagh <igloo@earth.li>**20091009185618
816 Fri Oct  9 14:50:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
817 GHC (correctly) warns about the unused import, which breaks the validate
818 build.
819]
820[Roll back "Cope with nhc98's (occasionally-strange) import semantics"
821Ian Lynagh <igloo@earth.li>**20091009184704
822 Fri Oct  9 14:43:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
823 GHC (correctly) warns about the unused import, which breaks the validate
824 build.
825]
826[It seems that nhc98 needs defaulting in Data.Fixed.
827Malcolm.Wallace@cs.york.ac.uk**20091009135242]
828[Another instance of nhc98's strange import semantics.
829Malcolm.Wallace@cs.york.ac.uk**20091009135051]
830[Make Data.Functor compatible with non-GHC compilers.
831Malcolm.Wallace@cs.york.ac.uk**20091009134821]
832[Cope with nhc98's (occasionally-strange) import semantics.
833Malcolm.Wallace@cs.york.ac.uk**20091009134351]
834[Fix gratuitous breakage of nhc98 in System.IO.
835Malcolm.Wallace@cs.york.ac.uk**20091009134001]
836[Fix gratuitous breakage of nhc98 in Control.Exception.Base.
837Malcolm.Wallace@cs.york.ac.uk**20091009133615]
838[Fix gratuitous breakage of non-GHC in Data.Fixed.
839Malcolm.Wallace@cs.york.ac.uk**20091009133330]
840[Fix gratuitous breakage for non-GHC in Data.Bits.
841Malcolm.Wallace@cs.york.ac.uk**20091009133257]
842[Use UTF-32LE instead of UTF32LE
843Simon Marlow <marlowsd@gmail.com>**20091006100207
844 Ignore-this: 7f881e36543d250ef848c9f60d67655a
845 The latter is not recognised by some iconv implementations.
846]
847[Strip any Byte Order Mark (BOM) from the front of decoded streams.
848Ben.Lippmeier@anu.edu.au*-20090930084229
849 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
850 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
851 of the stream, but Linux iconv doesn't.
852]
853[use UTF32BE/UTF32LE instead of UCS-4/UCS-4LE
854Simon Marlow <marlowsd@gmail.com>**20091005101554
855 Ignore-this: 2aef5e9bec421e714953b7aa1bdfc1b3
856]
857[Strip any Byte Order Mark (BOM) from the front of decoded streams.
858Ben.Lippmeier@anu.edu.au**20090930084229
859 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
860 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
861 of the stream, but Linux iconv doesn't.
862]
863[Add traceEvent :: String -> IO ()
864Simon Marlow <marlowsd@gmail.com>**20090925141257
865 Ignore-this: 8b1888bbf9682ffba13f815b6000e4b1
866 For emitting an event via the RTS tracing framework
867]
868[Fix the error message when flushing the read buffer of a non-seekable Handle
869Simon Marlow <marlowsd@gmail.com>**20090923090536
870 Ignore-this: 4342026df93759d99480f4e13f80a492
871]
872[Fix #3534: No need to flush the byte buffer when setting binary mode
873Simon Marlow <marlowsd@gmail.com>**20090923090445
874 Ignore-this: 625817ed7ae2c12291eb993a99dc640a
875]
876[Use let !y = x in .. x .. instead of seq in $! and evaluate (#2273)
877Simon Marlow <marlowsd@gmail.com>**20090916140454]
878[make some Applicative functions into methods, and split off Data.Functor (proposal #3335)
879Ross Paterson <ross@soi.city.ac.uk>**20090915173109
880 Ignore-this: a0cff4de6dfdbcbd56a66101bc4855a9
881 
882 The following functions
883 
884     (<$) :: Functor f => a -> f b -> f a
885     (*>) :: Applicative f => f a -> f b -> f b
886     (<*) :: Applicative f => f a -> f b -> f a
887     some :: Alternative f => f a -> f [a]
888     many :: Alternative f => f a -> f [a]
889 
890 are moved into the corresponding classes, with the existing implementations
891 as default definitions.  This gives people creating instances the option of
892 defining specialized implementations of these functions, though they should
893 be equivalent to the default definitions.
894 
895 Although (<$) is now a method of the Functor class, it is hidden in the
896 re-export by the Prelude, Control.Monad and Monad.  The new module
897 Data.Functor exposes the full class, plus the function (<$>).  These are
898 also re-exported by Control.Applicative.
899]
900[On Windows, use the console code page for text file encoding/decoding.
901Judah Jacobson <judah.jacobson@gmail.com>**20090913022126
902 Ignore-this: 86c2f2db8ef92b751599795d3195187b
903 
904 We keep all of the code page tables in the module
905 GHC.IO.Encoding.CodePage.Table.  That file was generated automatically
906 by running codepages/MakeTable.hs; more details are in the comments at the
907 start of that script.
908 
909 Storing the lookup tables adds about 40KB to each statically linked executable;
910 this only increases the size of a "hello world" program by about 7%.
911 
912 Currently we do not support double-byte encodings (Chinese/Japanese/Korean), since
913 including those codepages would increase the table size to 400KB.  It will be
914 straightforward to implement them once the work on library DLLs is finished.
915]
916[Fix "init" docs: the input list need not be finite. Fixes trac #3465
917Ian Lynagh <igloo@earth.li>**20090911210437]
918[Bump base version to 4.2.0.0
919Ian Lynagh <igloo@earth.li>**20090911153913]
920[Address #3310
921Simon Marlow <marlowsd@gmail.com>**20090830152850
922 Ignore-this: 40c7f7c171ee299a83092fd360a952b7
923 
924  - Rename BlockedOnDeadMVar   -> BlockedIndefinitelyOnMVar
925  - Rename BlockedIndefinitely -> BlockedIndefinitelyOnSTM
926  - instance Show BlockedIndefinitelyOnMVar is now
927      "blocked indefinitely in an MVar operation"
928  - instance Show BlockedIndefinitelyOnSTM is now
929      "blocked indefinitely in an STM transaction"
930 
931 clients using Control.OldException will be unaffected (the new
932 exceptions are mapped to the old names).  However, for base4-compat
933 we'll need to make a version of catch/try that does a similar
934 mapping.
935]
936[Fix unicode conversion for MSB architectures
937Ben.Lippmeier@anu.edu.au**20090830130028
938 This fixes the SPARC/Solaris build.
939]
940[Fix #3441: detect errors in partial sequences
941Simon Marlow <marlowsd@gmail.com>**20090830075909
942 Ignore-this: d12a75d95e0cae5eb1555266810ec281
943]
944[Fix hWaitForInput
945Simon Marlow <marlowsd@gmail.com>**20090827152116
946 Ignore-this: 2550e911f1a4d4357a5aa8d1764238ce
947 It was erroneously waiting when there were bytes to decode waiting in
948 the byte buffer.
949]
950[fix debugging code
951Simon Marlow <marlowsd@gmail.com>**20090827150628
952 Ignore-this: e1c82fdc19a22e247cd69ff6fa11921d
953]
954[Allow for configurable iconv include and library locations.
955Matthias Kilian <kili@outback.escape.de>**20090826154406
956 Ignore-this: be95fab611a5534cf184b508964ed498
957 This should help to fix the build on OpenBSD.
958]
959[typo in comment
960Simon Marlow <marlowsd@gmail.com>**20090826085252
961 Ignore-this: 1903ee0f354157a6ed3871c100f6b1b9
962]
963[un-hide some modules from the Haddock docs
964Simon Marlow <marlowsd@gmail.com>**20090825152457
965 Ignore-this: dce6606f93cf977fb24ebe99082dfa62
966]
967[Apply fix for #1548, from squadette@gmail.com
968Simon Marlow <marlowsd@gmail.com>**20090819120700
969 Ignore-this: 31c237c46a6445f588ed4b8c51bb6231
970]
971[improvements to Data.Fixed: instances for Typeable and Data, more predefined types
972Ashley Yakeley <ashley@semantic.org>**20090812055058
973 Ignore-this: feeece36d5632f02a05d137d2a39ab78
974]
975[Fix "Cabal check" warnings
976Ian Lynagh <igloo@earth.li>**20090811215856]
977[Add a GHC.Constants module; fixes trac #3094
978Ian Lynagh <igloo@earth.li>**20090809183252]
979[Apply proposal #3393
980Ian Lynagh <igloo@earth.li>**20090809134717
981 Add openTempFileWithDefaultPermissions and
982 openBinaryTempFileWithDefaultPermissions.
983]
984[Add some more C wrappers; patch from Krister Walfridsson
985Ian Lynagh <igloo@earth.li>**20090807200631
986 Fixes 21 testsuite errors on NetBSD 5.99.
987]
988[Fixing configure for autoconf 2.64
989Alexander Dunlap <alexander.dunlap@gmail.com>**20090805060748
990 Ignore-this: 992ab91ae3d68c12dbb265776e33e243
991]
992[add INLINE toList
993Ross Paterson <ross@soi.city.ac.uk>**20090806142853
994 Ignore-this: aba16aabb17d5dca44f15d188945680e
995 
996 In anticipation of the fixing of #2353.
997]
998[fix a copyright
999Simon Marlow <marlowsd@gmail.com>**20090805134045
1000 Ignore-this: b0ffbdd38fbba121e8bcba37c4082a60
1001]
1002[Tweak the BufferedIO class to enable a memory-mapped file implementation
1003Simon Marlow <marlowsd@gmail.com>**20090805134036
1004 Ignore-this: ec67d7a0a6d977438deaa342503f77e0
1005 We have to eliminate the assumption that an empty write buffer can be
1006 constructed by setting the buffer pointers to zero: this isn't
1007 necessarily the case when the buffer corresponds to a memory-mapped
1008 file, or other in-memory device implementation.
1009]
1010[Deprecate Control.OldException
1011Ian Lynagh <igloo@earth.li>**20090804143910]
1012[Windows build fix, following RTS tidyup
1013Simon Marlow <marlowsd@gmail.com>**20090803131121
1014 Ignore-this: ce862fb91c2b234211a8757f98690778
1015]
1016[Updates to follow the RTS tidyup
1017Simon Marlow <marlowsd@gmail.com>**20090801220743
1018 Ignore-this: 6e92412df93a66c12d75344053d5634
1019 C functions like isDoubleNaN moved here (primFloat.c)
1020]
1021[Add integer-simple as a build option
1022Ian Lynagh <igloo@earth.li>**20090722013151]
1023[Use shift[LR]Integer in the Bits Integer instance
1024Ian Lynagh <igloo@earth.li>**20090721222440]
1025[depend directly on integer-gmp, rather than indirecting through integer
1026Ian Lynagh <igloo@earth.li>**20090721185228]
1027[Move the instances of Functor and Monad IO to GHC.Base, to avoid orphans
1028Simon Marlow <marlowsd@gmail.com>**20090722102130
1029 Ignore-this: a7d85ac0025d559674249de0108dbcf4
1030]
1031[move "instance Exception Dynamic" so it isn't an orphan
1032Simon Marlow <marlowsd@gmail.com>**20090721093854
1033 Ignore-this: 5ede91ecfec2112c91b699d4de87cd02
1034]
1035[Improve the index checking for array accesses; fixes #2120 #2669
1036Ian Lynagh <igloo@earth.li>**20090719153228
1037 As well as checking that offset we are reading is actually inside the
1038 array, we now also check that it is "in range" as defined by the Ix
1039 instance. This fixes confusing behaviour (#2120) and improves some error
1040 messages (#2669).
1041]
1042[Make chr say what its argument was, if it's a bad argument
1043Ian Lynagh <igloo@earth.li>**20090718151049]
1044[remove unused warning
1045Simon Marlow <marlowsd@gmail.com>**20090715124416
1046 Ignore-this: 31f613654089d0f4a44363946087b41e
1047]
1048[warning fix: -fno-implicit-prelude -> -XNoImplicitPrelude
1049Simon Marlow <marlowsd@gmail.com>**20090715122839
1050 Ignore-this: dc8957249731d5bcb71c01899e5adf2b
1051]
1052[Add hGetEncoding :: Handle -> IO (Maybe TextEncoding)
1053Simon Marlow <marlowsd@gmail.com>**20090715122519
1054 Ignore-this: 14c3eff996db062da1199739781e4708
1055 as suggested during the discussion on the libraries list
1056]
1057[Add more documentation to mkTextEncoding
1058Simon Marlow <marlowsd@gmail.com>**20090715122414
1059 Ignore-this: 97253b2624267df3a246a18121e8ea81
1060 noting that "//IGNORE" and "//TRANSLIT" suffixes can be used with GNU
1061 iconv.
1062]
1063[Add the utf8_bom codec
1064Simon Marlow <marlowsd@gmail.com>**20090715122257
1065 Ignore-this: 1c9396cd805201fe873a39382ced79c7
1066 as suggested during the discussion on the libraries list.
1067]
1068[Export Unicode and newline functionality from System.IO; update Haddock docs
1069Simon Marlow <marlowsd@gmail.com>**20090713113104
1070 Ignore-this: c3f017a555335aa55d106253393f72e2
1071]
1072[add a comment about the non-workingness of CHARBUF_UTF16
1073Simon Marlow <marlowsd@gmail.com>**20090707124406
1074 Ignore-this: 98d00411b68d688b3b4cffc9507b1f35
1075]
1076[Fix build on Windows
1077Ian Lynagh <igloo@earth.li>**20090711004351]
1078[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1079Ian Lynagh <igloo@earth.li>**20090710204513]
1080[Use throwErrnoIfMinus1_ when calling getrusage
1081Ian Lynagh <igloo@earth.li>**20090710204221]
1082[Remove an unused import
1083Ian Lynagh <igloo@earth.li>**20090710153345]
1084[reportStackOverflow now returns IO ()
1085Ian Lynagh <igloo@earth.li>**20090710153257
1086 It used to do "return undefined" to return IO a.
1087]
1088[GHC.Conc.reportError now returns IO ()
1089Ian Lynagh <igloo@earth.li>**20090710152646
1090 It used to return IO a, by "return undefined".
1091]
1092[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1093Ian Lynagh <igloo@earth.li>**20090710152526]
1094[Minor SampleVar refactoring
1095Ian Lynagh <igloo@earth.li>**20090710151438]
1096[Fix "warn-unused-do-bind" warnings in GHC/IO/Handle/Text.hs
1097Ian Lynagh <igloo@earth.li>**20090710122905]
1098[Fix some "warn-unused-do-bind" warnings where we just want to ignore the result
1099Ian Lynagh <igloo@earth.li>**20090710005638]
1100[Use the result of writeCharBuf in GHC/IO/Encoding/Latin1.hs too
1101Ian Lynagh <igloo@earth.li>**20090710004032]
1102[Minor code tidyups in GHC.Conc
1103Ian Lynagh <igloo@earth.li>**20090710003801]
1104[Fix "warn-unused-do-bind" warning in GHC.Conc
1105Ian Lynagh <igloo@earth.li>**20090710003530
1106 If we fail to communicate with the IO manager then we print a warning
1107 using debugErrLn from the ghc-prim package.
1108]
1109[Fix "warn-unused-do-bind" warnings in System.Posix.Internals
1110Ian Lynagh <igloo@earth.li>**20090709164546]
1111[Fix "warn-unused-do-bind" warnings where we really do want to ignore the result
1112Ian Lynagh <igloo@earth.li>**20090709163912]
1113[Add back imports needed on Windows
1114Ian Lynagh <igloo@earth.li>**20090707181924]
1115[Remove unused imports
1116Ian Lynagh <igloo@earth.li>**20090707115810]
1117[Remove unused imports from base
1118simonpj@microsoft.com**20090706111842
1119 Ignore-this: f9b5f353e3bb820f787c56d615b28765
1120 
1121 These unused imports are detected by the new unused-import code
1122 
1123]
1124[Use the result of writeCharBuf
1125Simon Marlow <marlowsd@gmail.com>**20090706133303
1126 Ignore-this: 52288dd559bf4c4f313df6197091d935
1127   
1128 This only makes a difference when CHARBUF_UTF16 is in use, which it
1129 normally isn't.  I suspect CHARBUF_UTF16 doesn't currently work for
1130 other reasons (CHARBUF_UTF16 was an experiment before I wrote the
1131 GHC.IO.Encoding.UTF* codecs), but this patch at least makes it
1132 slightly closer to working.
1133]
1134[Remove some cruft from Data.HashTable
1135Ian Lynagh <igloo@earth.li>**20090706181630]
1136[Add 'eof' to Text.ParserCombinators.ReadP
1137simonpj@microsoft.com**20090706111801
1138 Ignore-this: 2aea7b848e00c894761bc4011adaa95d
1139 
1140 Add a ReadP parser that succeeds at the end of input. Very useful!
1141 
1142]
1143[Don't export CLDouble for GHC; fixes trac #2793
1144Ian Lynagh <igloo@earth.li>**20090705155120
1145 We never really supported CLDouble (it was a plain old double underneath),
1146 and pretending that we do does more harm than good.
1147]
1148[a byte between 0x80 and 0xBF is illegal immediately (#3341)
1149Simon Marlow <marlowsd@gmail.com>**20090702081415
1150 Ignore-this: dc19ef59a1a21118d5a7dd38aa2f611c
1151]
1152[avoid a warning
1153Simon Marlow <marlowsd@gmail.com>**20090630084134
1154 Ignore-this: c92a45ee216faf01327feae9fe06d6e2
1155]
1156[Add a wrapper for libiconv.
1157Matthias Kilian <kili@outback.escape.de>**20090629183634
1158 Ignore-this: 23c6047c0d71b745b495cc223574a47f
1159]
1160[#include <sys/times.h> if we have it (should fix build problems)
1161Simon Marlow <marlowsd@gmail.com>**20090629085351
1162 Ignore-this: a35e93b37ca9595c73460243180f4b9d
1163]
1164[set binary mode for existing FDs on Windows (fixes some GHCi test failures)
1165Simon Marlow <marlowsd@gmail.com>**20090626120522
1166 Ignore-this: 580cf636e9c77d8427aff6861d089481
1167]
1168[Move directory-related stuff to the unix package
1169Simon Marlow <marlowsd@gmail.com>**20090625120325
1170 Ignore-this: b997b3cbce0a46ca87ad825bbdc0a411
1171 now that it isn't used on Windows any more.
1172]
1173[TAG 2009-06-25
1174Ian Lynagh <igloo@earth.li>**20090625160056]
1175Patch bundle hash:
11764fdf834be1107c5b2c22fb9cd19603426a3b98d5