Ticket #4343: add_throwSTM_and_generalize_catchSTM.dpatch

File add_throwSTM_and_generalize_catchSTM.dpatch, 51.3 KB (added by basvandijk, 4 years ago)

Patch for base

Line 
12 patches for repository http://darcs.haskell.org/packages/base:
2
3Sun Sep 26 21:21:06 CEST 2010  Bas van Dijk <v.dijk.bas@gmail.com>
4  * Generalize catchSTM
5
6Sun Sep 26 21:21:44 CEST 2010  Bas van Dijk <v.dijk.bas@gmail.com>
7  * Add throwSTM :: Exception e => e -> STM a
8
9New patches:
10
11[Generalize catchSTM
12Bas van Dijk <v.dijk.bas@gmail.com>**20100926192106
13 Ignore-this: d2038494582d2cde2247293dd162671c
14] {
15hunk ./GHC/Conc.lhs 61
16         , atomically    -- :: STM a -> IO a
17         , retry         -- :: STM a
18         , orElse        -- :: STM a -> STM a -> STM a
19-        , catchSTM      -- :: STM a -> (Exception -> STM a) -> STM a
20+        , catchSTM      -- :: Exception e => STM a -> (e -> STM a) -> STM a
21         , alwaysSucceeds -- :: STM a -> STM ()
22         , always        -- :: STM Bool -> STM ()
23         , TVar(..)
24hunk ./GHC/Conc/Sync.lhs 55
25         , atomically    -- :: STM a -> IO a
26         , retry         -- :: STM a
27         , orElse        -- :: STM a -> STM a -> STM a
28-        , catchSTM      -- :: STM a -> (Exception -> STM a) -> STM a
29+        , catchSTM      -- :: Exception e => STM a -> (e -> STM a) -> STM a
30         , alwaysSucceeds -- :: STM a -> STM ()
31         , always        -- :: STM Bool -> STM ()
32         , TVar(..)
33hunk ./GHC/Conc/Sync.lhs 510
34 orElse (STM m) e = STM $ \s -> catchRetry# m (unSTM e) s
35 
36 -- |Exception handling within STM actions.
37-catchSTM :: STM a -> (SomeException -> STM a) -> STM a
38-catchSTM (STM m) k = STM $ \s -> catchSTM# m (\ex -> unSTM (k ex)) s
39+catchSTM :: Exception e => STM a -> (e -> STM a) -> STM a
40+catchSTM (STM m) handler = STM $ catchSTM# m handler'
41+    where
42+      handler' e = case fromException e of
43+                     Just e' -> unSTM (handler e')
44+                     Nothing -> raiseIO# e
45 
46 -- | Low-level primitive on which always and alwaysSucceeds are built.
47 -- checkInv differs form these in that (i) the invariant is not
48}
49[Add throwSTM :: Exception e => e -> STM a
50Bas van Dijk <v.dijk.bas@gmail.com>**20100926192144
51 Ignore-this: c6bfdae0eab9f0cf1360bc06d088bfd5
52] {
53hunk ./GHC/Conc.lhs 61
54         , atomically    -- :: STM a -> IO a
55         , retry         -- :: STM a
56         , orElse        -- :: STM a -> STM a -> STM a
57+        , throwSTM      -- :: Exception e => e -> STM a
58         , catchSTM      -- :: Exception e => STM a -> (e -> STM a) -> STM a
59         , alwaysSucceeds -- :: STM a -> STM ()
60         , always        -- :: STM Bool -> STM ()
61hunk ./GHC/Conc/Sync.lhs 55
62         , atomically    -- :: STM a -> IO a
63         , retry         -- :: STM a
64         , orElse        -- :: STM a -> STM a -> STM a
65+        , throwSTM      -- :: Exception e => e -> STM a
66         , catchSTM      -- :: Exception e => STM a -> (e -> STM a) -> STM a
67         , alwaysSucceeds -- :: STM a -> STM ()
68         , always        -- :: STM Bool -> STM ()
69hunk ./GHC/Conc/Sync.lhs 510
70 orElse :: STM a -> STM a -> STM a
71 orElse (STM m) e = STM $ \s -> catchRetry# m (unSTM e) s
72 
73+-- | A variant of 'throw' that can only be used within the 'STM' monad.
74+--
75+-- Throwing an exception in @STM@ aborts the transaction and propagates the
76+-- exception.
77+--
78+-- Although 'throwSTM' has a type that is an instance of the type of 'throw', the
79+-- two functions are subtly different:
80+--
81+-- > throw e    `seq` x  ===> throw e
82+-- > throwSTM e `seq` x  ===> x
83+--
84+-- The first example will cause the exception @e@ to be raised,
85+-- whereas the second one won\'t.  In fact, 'throwSTM' will only cause
86+-- an exception to be raised when it is used within the 'STM' monad.
87+-- The 'throwSTM' variant should be used in preference to 'throw' to
88+-- raise an exception within the 'STM' monad because it guarantees
89+-- ordering with respect to other 'STM' operations, whereas 'throw'
90+-- does not.
91+throwSTM :: Exception e => e -> STM a
92+throwSTM e = STM $ raiseIO# (toException e)
93+
94 -- |Exception handling within STM actions.
95 catchSTM :: Exception e => STM a -> (e -> STM a) -> STM a
96 catchSTM (STM m) handler = STM $ catchSTM# m handler'
97}
98
99Context:
100
101[doc tweak for Directory file type: file names are '\0'-separated
102Simon Marlow <marlowsd@gmail.com>**20100922113811
103 Ignore-this: 96b7b004bd6e5bc3e958ad55bf238ba1
104]
105[documentation for IODeviceType (#4317, edited by me)
106Simon Marlow <marlowsd@gmail.com>**20100915131341
107 Ignore-this: 21c50ca7a189eebcf299523b6e942bae
108]
109[Allow Data.HashTable construction with user-supplied size
110**20100722210726
111 Ignore-this: bd54880bb16a106a992f03b040dc4164
112 
113 This avoids some resizing for users who know they will be inserting a
114 lot of data.
115 
116 http://hackage.haskell.org/trac/ghc/ticket/4193
117]
118[some fixes for hGetBufSome
119Simon Marlow <marlowsd@gmail.com>**20100916113732
120 Ignore-this: 3e596a606c180dc4859ea8f4c9132ca1
121  - fix one case where it was blocking when it shouldn't
122  - a couple of error-message tweaks
123]
124[Windows: map ERROR_NO_DATA to EPIPE, rather than EINVAL
125Simon Marlow <marlowsd@gmail.com>**20100915142618
126 Ignore-this: 9023e5f0542419f225aef26cb6b1d88d
127 WriteFile() returns ERROR_NO_DATA when writing to a pipe that is
128 "closing", however by default the write() wrapper in the CRT maps this
129 to EINVAL so we get confusing things like
130 
131   hPutChar: invalid argument (Invalid Argumnet)
132 
133 when piping the output of a Haskell program into something that closes
134 the pipe early.  This was happening in the testsuite in a few place.
135 
136 The solution is to map ERROR_NO_DATA to EPIPE correctly, as we
137 explicitly check for EPIPE on stdout (in GHC.TopHandler) so we can
138 exit without an error in this case.
139]
140[tighten up parsing of numbers (#1579)
141Simon Marlow <marlowsd@gmail.com>**20100913214733
142 Ignore-this: 3411bf3d2e98cfacb9e0afd11d79e722
143]
144[Add absentError. 
145simonpj@microsoft.com**20100914134639
146 Ignore-this: d0eef5a87e1def4cdbde92a55241c8c4
147 
148 This patch accompanies the HEAD patch:
149 
150   Tue Sep 14 12:38:27 BST 2010  simonpj@microsoft.com
151     * Make absent-arg wrappers work for unlifted types (fix Trac #4306)
152     
153     Previously we were simply passing arguments of unlifted
154     type to a wrapper, even if they were absent, which was
155     stupid.
156     
157     See Note [Absent error Id] in WwLib.
158]
159[Add missing import, fixes build on windows
160simonpj@microsoft.com**20100914122750
161 Ignore-this: 12ece15ef94982ddfbf5f9f7900619da
162]
163[Add a suitable Show instance for TextEncoding (#4273)
164Simon Marlow <marlowsd@gmail.com>**20100913154459
165 Ignore-this: 77f2235460895debd2827f34c42c3435
166]
167[don't fill a finalized handle with an error (see comment)
168Simon Marlow <marlowsd@gmail.com>**20100913153350
169 Ignore-this: c72cdb6898dffa88eca1d781171b2943
170]
171[deriving (Eq, Ord, Read, Show) for Newline and NewlineMode
172Simon Marlow <marlowsd@gmail.com>**20100913153031
173 Ignore-this: 9b9b29bfb7abf5550cfbfa7788f81bf
174]
175[fix warning on Windows
176Simon Marlow <marlowsd@gmail.com>**20100913111536
177 Ignore-this: dacc5448c452daad60ed37a1a5ed096e
178]
179[Put the state-token argument on fill, done, adjust on the RHS
180simonpj@microsoft.com**20100913101832
181 Ignore-this: d228b492de7d4635c026ed24cbc17e34
182 
183 This is so that the functions will inline when
184 applied to their normal (non-state-token) aguments.
185 
186 I forget why I did this, but it seems like the right thing anyway.
187]
188[avoid Foreign.unsafePerformIO
189Ross Paterson <ross@soi.city.ac.uk>**20100909125521
190 Ignore-this: b698101119ffd1bc6311cce0736f745d
191]
192[Remove debugging code accidentally left in
193Simon Marlow <marlowsd@gmail.com>**20100909113331
194 Ignore-this: 906a14176dd37030b8203782a687936b
195]
196[Fix Windows build; patches frmo ezyang
197Ian Lynagh <igloo@earth.li>**20100908123037
198 Ignore-this: 2f02986087edd7da8382221012c27cd0
199]
200[More accurate isatty test for MinGW.
201Edward Z. Yang <ezyang@mit.edu>**20100907154144
202 Ignore-this: 93bdc2b2a8e65a7c4c7d3906bdda01db
203]
204[Fix the build when HAVE_KQUEUE but not HAVE_KEVENT64
205Ian Lynagh <igloo@earth.li>**20100904223703]
206[Fix warnings
207benl@ouroborus.net**20100830044741
208 Ignore-this: 8397aaec7c36046c9ace403e65f32d32
209]
210[fix cache variable name used by FP_SEARCH_LIBS_PROTO
211Ross Paterson <ross@soi.city.ac.uk>**20100819204858
212 Ignore-this: b8113cb3c6f0e03c507297c99d3d82b7
213]
214[Add a missing castPtr (only shows up in -DDEBUG)
215simonpj@microsoft.com**20100815145127
216 Ignore-this: 30b9c42cd3ce7837bdabd254fe66078d
217]
218[Fixed a rounding error in threadDelay
219Johan Tibell <johan.tibell@gmail.com>**20100813124043
220 Ignore-this: 1cb77d0852233ffffb144b134064ee3c
221]
222[export allocaBytesAligned; make allocaArray use the correct alignment (#2917)
223Simon Marlow <marlowsd@gmail.com>**20100812105524
224 Ignore-this: deb6495f7b7b84deaf02b88927a5ba8c
225]
226[deprecate unGetChan and isEmptyChan (see #4154)
227Simon Marlow <marlowsd@gmail.com>**20100705125952
228 Ignore-this: b4e769959f131b2d0001eb7202bc1b92
229]
230[Add type signatures to cope with lack of local generalisation
231simonpj@microsoft.com**20100728124847
232 Ignore-this: d3af9a47c2821c6081bde05a135a92fb
233]
234[Add type signature in local where
235simonpj@microsoft.com**20100727151532
236 Ignore-this: 1c57063ad32d13e0d1ec8daf968bf055
237]
238[Integrated new I/O manager
239Simon Marlow <marlowsd@gmail.com>**20100810082248
240 Ignore-this: ed70a9066ac9b676a446fe99978fef7a
241 (patch originally by Johan Tibell <johan.tibell@gmail.com>, minor merging by me)
242]
243[Add mfilter to Control.Monad
244jon.fairbairn@cl.cam.ac.uk**20090917145616
245 Ignore-this: de4240b60684f3065b29378df3ea98f2
246 
247 Straightforward MonadPlus version of List.filter. I would
248 prefer to call it filter, but the current naming scheme for
249 Control.Monad implies mfilter.
250 
251]
252[move Monad and MonadFix instances for Either from mtl (proposal #4159)
253Ross Paterson <ross@soi.city.ac.uk>**20100729122449
254 Ignore-this: b0f8cd8643679948d1da43bd7c08c5aa
255 
256 The Monad and MonadFix instances for Either (formerly in the mtl
257 package) are moved to Control.Monad.Instances and Control.Monad.Fix
258 respectively.  The Monad instance is still an orphan, to retain Haskell
259 98 compatibility, but the MonadFix instance is together with its class.
260 The Error constraint is removed from both instances, and the default
261 definition of fail is used.
262]
263[Remove egregious ghc-ish from Foreign.Marshal
264Malcolm.Wallace@me.com**20100722075449]
265[add numSparks :: IO Int (#4167)
266Simon Marlow <marlowsd@gmail.com>**20100720153858
267 Ignore-this: 4543f57a7f137f8cae1c3efc5c023a9b
268]
269[add unsafeLocalState from Haskell 2010, and docs
270Simon Marlow <marlowsd@gmail.com>**20100720082819
271 Ignore-this: dcd79fb546ebe29ddff4df279ec2f38
272]
273[docs: mention that Foreign.unsafePerformIO is deprecated
274Simon Marlow <marlowsd@gmail.com>**20100720082804
275 Ignore-this: 4cfebb8f2a1cddc7d15e94e31b2befa4
276 We can't actually deprecate it without introducing a name clash
277 between Foreign.unsafePerformIO and System.IO.Unsafe.unsafePerformIO
278]
279[doc formatting fix
280Simon Marlow <marlowsd@gmail.com>**20100714151347
281 Ignore-this: 255edef607dcd290e198015240b5d125
282]
283[add module intro from Haskell 2010
284Simon Marlow <marlowsd@gmail.com>**20100714115853
285 Ignore-this: 59b5a07507a059ccccdff2dfb6490a27
286]
287[document exception-overriding behaviour in withFile
288Simon Marlow <marlowsd@gmail.com>**20100714104107
289 Ignore-this: f99e641ea2f46d872cb7420a62fa50dc
290]
291[doc: use "finalizer" consistently
292Simon Marlow <marlowsd@gmail.com>**20100714103649
293 Ignore-this: bdfea40f31dc5045fdbc6e12266dda93
294]
295[clarify meaning of bit
296Simon Marlow <marlowsd@gmail.com>**20100714103310
297 Ignore-this: 521b031f1e83ef34ca03d9aa9273df8a
298]
299[note shortcutting behaviour of any/all/elem
300Simon Marlow <marlowsd@gmail.com>**20100714103304
301 Ignore-this: 1605f362ba0712ad1cea1309636f3ea1
302]
303[add cast{C,U}CharToChar and castCharTo{C,U}Char, from Haskell 2010
304Simon Marlow <marlowsd@gmail.com>**20100713132515
305 Ignore-this: 9b1da827016c7b08668078b45964e9de
306]
307[mention that IntPtr and WordPtr can be marshalled to/from intptr_t and uintptr_t
308Simon Marlow <marlowsd@gmail.com>**20100713132403
309 Ignore-this: dcc112a72746ba117a84fa29e71b6800
310]
311[Partial fix for Trac #4136
312simonpj@microsoft.com**20100707135725
313 Ignore-this: 9548eeb3187d9779d4e5c858a0f35354
314 
315 In 'choose' (which is a library function designed specifically
316 to support derived instances of Read), we must match Symbol
317 as well as Ident, for nullary constructors that (wierdly) are
318 symbols.
319]
320[Fix typo in documentation
321Simon Hengel <simon.hengel@wiktory.org>**20100711141648
322 Ignore-this: c052dd8a681832ef598a323ad55eae3a
323]
324[Remove duplicated word in documentation
325Simon Hengel <simon.hengel@wiktory.org>**20100711072703
326 Ignore-this: fb3732dc57be55f14168792f923433
327]
328[Allow nhc98 to cope with recent changes to Control.Exception.
329Malcolm.Wallace@me.com**20100710170940]
330[ New asynchronous exception control API (base parts)
331Simon Marlow <marlowsd@gmail.com>**20100708152735
332 Ignore-this: 71a4811804f04259f1fe739f8863beaf
333   
334 As discussed on the libraries/haskell-cafe mailing lists
335   http://www.haskell.org/pipermail/libraries/2010-April/013420.html
336 
337 This is a replacement for block/unblock in the asychronous exceptions
338 API to fix a problem whereby a function could unblock asynchronous
339 exceptions even if called within a blocked context.
340 
341 The new terminology is "mask" rather than "block" (to avoid confusion
342 due to overloaded meanings of the latter).
343 
344 The following is the new API; the old API is deprecated but still
345 available for the time being.
346 
347 Control.Exception
348 -----------------
349 
350 mask  :: ((forall a. IO a -> IO a) -> IO b) -> IO b
351 mask_ :: IO a -> IO a
352 
353 uninterruptibleMask :: ((forall a. IO a -> IO a) -> IO b) -> IO b
354 uninterruptibleMask_ :: IO a -> IO
355 
356 getMaskingState :: IO MaskingState
357 
358 data MaskingState
359   = Unmasked
360   | MaskedInterruptible
361   | MaskedUninterruptible
362 
363 
364 Control.Concurrent
365 ------------------
366 
367 forkIOUnmasked :: IO () -> IO ThreadId
368]
369[Async-exception safety, and avoid space leaks
370Simon Marlow <marlowsd@gmail.com>**20100708145819
371 Ignore-this: dbfd0e61551e9e7b4fc1c6fe9b9a83de
372 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
373 Modified slightly by me to remove non-functional changes.
374]
375[Async-exception safety, and avoid space leaks
376Simon Marlow <marlowsd@gmail.com>**20100708103154
377 Ignore-this: 190c3ac8f6633231624da8cf1316588
378 Patch submitted by: Bas van Dijk <v.dijk.bas@gmail.com>
379 Modified slightly by me to remove non-functional changes.
380]
381[Fix a few places where we forgot to close the text codecs (#4029)
382Simon Marlow <marlowsd@gmail.com>**20100702130210
383 Ignore-this: 2e81a4b4cb343181cef34b0f9e2ded47
384 Each time you invoke :load in GHCi it resets the CAFs, including
385 stdin/stdout/stderr, and each of these was allocating a new iconv_t.
386]
387[remove docs from Monad that belonged on the instance for MonadPlus IO
388Simon Marlow <marlowsd@gmail.com>**20100701154203
389 Ignore-this: 59df02542a7ac9421552a2155d848d27
390]
391[docs: unqualify Prelude.IO
392Simon Marlow <marlowsd@gmail.com>**20100701153817
393 Ignore-this: 73b0202876c827e7a5b4a5ce74e724c4
394]
395[unqualify Float and Double
396Simon Marlow <marlowsd@gmail.com>**20100701142727
397 Ignore-this: cbe89d31a00bf49996a33933324fca17
398]
399[extract information about Data.Time from docs for CTime
400Simon Marlow <marlowsd@gmail.com>**20100701142415
401 Ignore-this: c48c9609b8d36e43e033a7bea81d6f17
402]
403[doc typo
404Simon Marlow <marlowsd@gmail.com>**20100701142354
405 Ignore-this: 17a1fd703831c888975ff63fbfa3a9b2
406]
407[peekArray docs: remove mentions of "this version" and "previous version"
408Simon Marlow <marlowsd@gmail.com>**20100701125333
409 Ignore-this: 39a744874258670bd935ba9e38390939
410]
411[doc typo
412Simon Marlow <marlowsd@gmail.com>**20100701124154
413 Ignore-this: 98f5c286e38c2c34c96b05d5e8bc5ad9
414]
415[doc typo
416Simon Marlow <marlowsd@gmail.com>**20100701124128
417 Ignore-this: 10a4314ec7aed336701fc616fb574ebc
418]
419[doc typo
420Simon Marlow <marlowsd@gmail.com>**20100701123715
421 Ignore-this: c4909a7bf7163460ee5d32f58812041e
422]
423[doc wibble: Haskell 98 I/O Error -> 'IOError'
424Simon Marlow <marlowsd@gmail.com>**20100701123612
425 Ignore-this: bf373df781acbc575e4ffe3b7e6059ae
426]
427[doc typo
428Simon Marlow <marlowsd@gmail.com>**20100701123014
429 Ignore-this: 16aaccae48ef3101adf78ea5b0d5a8fd
430]
431[Haddock hacks to fix whitespace consistency
432Simon Marlow <marlowsd@gmail.com>**20100701121631
433 Ignore-this: 61c58dec52a31fd2d3f331a87d2f903f
434]
435[use '==' consistently rather than '->' in examples
436Simon Marlow <marlowsd@gmail.com>**20100701121616
437 Ignore-this: 472b0a05a85d34d9712186040e1636d9
438]
439[doc wibble: remove confusing mention of "Prelude"
440Simon Marlow <marlowsd@gmail.com>**20100701113308
441 Ignore-this: 232283d0096d01cd45e9b3c5c1e63a6d
442]
443[doc wibble: nonstrict -> non-strict
444Simon Marlow <marlowsd@gmail.com>**20100701113253
445 Ignore-this: 4264f0ab23a0835fc13c6e8601d6b743
446]
447[doc whitespace
448Simon Marlow <marlowsd@gmail.com>**20100701112242
449 Ignore-this: 777a95b1d1140c61d3ab95d5eb5809e7
450]
451[move the doc for 'Char' to its new home in ghc-prim:GHC.Types
452Simon Marlow <marlowsd@gmail.com>**20100629134150
453 Ignore-this: 7687db0077a29498349bfb4b44983985
454]
455[doc wibble
456Simon Marlow <marlowsd@gmail.com>**20100629122608
457 Ignore-this: 9a909e5d015332dc445bd9592e6e386d
458]
459[doc updates in System.IO
460Simon Marlow <marlowsd@gmail.com>**20100629122118
461 Ignore-this: 2257ec1cc4cdb8b7804cfa1f3cf32753
462]
463[doc wibble
464Simon Marlow <marlowsd@gmail.com>**20100625134858
465 Ignore-this: 64c50f29df6c389273b818918fe7033a
466]
467[doc wibbles
468Simon Marlow <marlowsd@gmail.com>**20100624154614
469 Ignore-this: b364aad53beea6e741fee2824459b6e8
470]
471[Fix haddock formatting
472Ian Lynagh <igloo@earth.li>**20100625222623]
473[Give nub's complexity in the haddock docs; fixes #4086
474Ian Lynagh <igloo@earth.li>**20100625222059]
475[correct docs for exitWith: only stdout/stderr are flushed, not all Handles
476Simon Marlow <marlowsd@gmail.com>**20100624130506
477 Ignore-this: 33a938dad8f0bc061572e2ec571cacc7
478]
479[fix docs for isSpace
480Simon Marlow <marlowsd@gmail.com>**20100624130444
481 Ignore-this: b35ff080dbb9833176f08e39dbd9ff6d
482]
483[make the hGetBuf/hPutBuf family work with non-FD Handles (#4144)
484Simon Marlow <marlowsd@gmail.com>**20100624130425
485 Ignore-this: 8200f0208a9b1b1cf4824f343d75819a
486]
487[nit in docs for accumArray
488Simon Marlow <marlowsd@gmail.com>**20100622121131
489 Ignore-this: c066a456c40907e767df10c3990f35ff
490]
491[add doc for the ExitCode type
492Simon Marlow <marlowsd@gmail.com>**20100622120930
493 Ignore-this: 99c34332be7f3565da844528b470054a
494]
495[remove extraneous info from docs for Array
496Simon Marlow <marlowsd@gmail.com>**20100622120921
497 Ignore-this: e2a3f5e84fc23eb7bae911f0680e805e
498]
499[add an INLINE to the list version of traverse, to enable fusion
500Simon Marlow <marlowsd@gmail.com>**20100608082531
501 Ignore-this: ea98cdc3308b406bb04c0f7a38c4424b
502]
503[Don't define the C localeEncoding on Windows
504Ian Lynagh <igloo@earth.li>**20100620202342
505 Ignore-this: c4992f6832a391b0cccc5a9b7d643976
506 (it causes warnings, and isn't used)
507]
508[add Applicative instance for Either (proposal #4095)
509Ross Paterson <ross@soi.city.ac.uk>**20100617225110
510 Ignore-this: 50262ec4700dc16efec5755be5b308c5
511 
512 This is not the only possible instance for Either, but this one is
513 compatible with the usual Monad instance.
514]
515[Use libcharset instead of nl_langinfo(CODESET) if possible.
516pho@cielonegro.org**20100519013112
517 Ignore-this: 4c1e278e022a3d276848afc1dcba4425
518 
519 nl_langinfo(CODESET) doesn't always return standardized variations of the encoding names. Use libcharset if possible, which is shipped together with GNU libiconv.
520]
521[Add a note about the interruptibility of throwTo.
522Simon Marlow <marlowsd@gmail.com>**20100615112720
523 Ignore-this: ae9fabe95310d7c364e95f7784793485
524]
525[docs: note that hGetBufNonBlocking isn't non-blocking on Windows
526Simon Marlow <marlowsd@gmail.com>**20100615112547
527 Ignore-this: 4f3e5213e142149affe08c5123d6efea
528]
529[don't depend on Prelude (#4122)
530Simon Marlow <marlowsd@gmail.com>**20100615105631
531 Ignore-this: 1a3fd49b103fe31cbb453f302c18767f
532]
533[Don't depend on Prelude (#4123)
534Simon Marlow <marlowsd@gmail.com>**20100615105401
535 Ignore-this: cc7616d85a1637bc7621b4f2bc181c0e
536]
537[bump version to 4.3.0.0, added instance MonadPlus STM
538Simon Marlow <marlowsd@gmail.com>**20100601144831
539 Ignore-this: 7c3cf7574499c4267372493f2636dc0
540]
541[Moved MonadPlus instance for STM from Control.Monad.STM to GHC.Conc to avoid an orphaned instance
542Bas van Dijk <v.dijk.bas@gmail.com>**20100516160651
543 Ignore-this: 651b852942b2fae2b93f996e39239b8f
544]
545[expand Foldable instance for Array
546Ross Paterson <ross@soi.city.ac.uk>**20100602212154
547 Ignore-this: 9bd9e9666a9400431eb92352244fe7e7
548]
549[doc comment illustrating Foldable(foldr)
550Ross Paterson <ross@soi.city.ac.uk>**20100527150833
551 Ignore-this: 8f27d889379803f3ba86d6e928428f3c
552]
553[fix syntax in doc comments
554Ross Paterson <ross@soi.city.ac.uk>**20100527150757
555 Ignore-this: cb78da51d60ff6863dc395f1a892c103
556]
557[export hGetBufSome (#4046)
558Simon Marlow <marlowsd@gmail.com>**20100520093538
559 Ignore-this: f467fad9722e27edfad6b3dd75290e7b
560]
561[hWaitForInput: don't try to read from the device (#4078)
562Simon Marlow <marlowsd@gmail.com>**20100517133741
563 Ignore-this: 55ec33b03397380259b91e4ca62207a6
564 readTextDeviceNonBlocking is not non-blocking on Windows
565]
566[Added Applicative and Alternative instances for STM
567Bas van Dijk <v.dijk.bas@gmail.com>**20100516171756
568 Ignore-this: 567003bc4040bc97105cda4d31ebf04a
569]
570[hSetEncoding: change the encoding on both read and write sides (#4066)
571Simon Marlow <marlowsd@gmail.com>**20100514124628
572 Ignore-this: 5b9e9caef06356d0296c584159709ebb
573]
574[Correct haddock formatting.
575Adam Vogt <vogt.adam@gmail.com>**20100423022103
576 Ignore-this: d2622339302048fda48080f7d5ce4a2f
577]
578[Fix for hGetBufSome
579Simon Marlow <marlowsd@gmail.com>**20100505135637
580 Ignore-this: 2019680f8fb223956cacfcf0d046f133
581]
582[improve the documentation for throwTo and killThread (#3884)
583Simon Marlow <marlowsd@gmail.com>**20100505135600
584 Ignore-this: ce881d96ddb729acb6ca09c779975e7d
585]
586[elaborate the docs for unsafePerformIO a bit
587Simon Marlow <marlowsd@gmail.com>**20100505101249
588 Ignore-this: 1cec3f67560b672c64c5a0dcf9a79eb7
589]
590[add Typeable instance
591Simon Marlow <marlowsd@gmail.com>**20100504152815
592 Ignore-this: 6d9cf9d62f0ef17fa459bf213a04098
593]
594[Add hGetBufSome, like hGetBuf but can return short reads
595Simon Marlow <marlowsd@gmail.com>**20100504152759
596 Ignore-this: 195c905b43f8d9505029364e2c5b18e
597]
598[Add swap (#3298)
599Simon Marlow <marlowsd@gmail.com>**20100504095339
600 Ignore-this: 13b007dc4594ce252997ec6fa0bbd976
601]
602[inline allocaArray0, to fix withCString benchmark
603Simon Marlow <marlowsd@gmail.com>**20100423124729
604 Ignore-this: 35c96816acc2f3aaf9dd29f7995fa6f0
605]
606[raise asynchronous exceptions asynchronously (#3997)
607Simon Marlow <marlowsd@gmail.com>**20100421094932
608 Ignore-this: 6d987d93d382c0f69c68c326312abd6b
609]
610[add NOINLINE pragmas for stdin/stdout/stderr
611Simon Marlow <marlowsd@gmail.com>**20100421082041
612 Ignore-this: 3fc130268ec786f28d945858d6690986
613]
614[INLINE alloca and malloc
615Simon Marlow <marlowsd@gmail.com>**20100419135333
616 Ignore-this: b218bd611f18721b1505a8c0b9e6a16a
617 See discussion on glasgow-haskell-users:
618   http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018740.html
619]
620[Move comment closer to the offending line
621Matthias Kilian <kili@outback.escape.de>**20100419155421
622 Ignore-this: b34a1d7affd66f67d210df2377b585d9
623]
624[Ignore the return code of c_fcntl_write again
625Matthias Kilian <kili@outback.escape.de>**20100415140452
626 Ignore-this: 266d8ba02cc3cb79c85629b3528261c9
627 
628 The return code has been ignored in the past on purpose, because
629 O_NONBLOCK will fail on BSDs for some special files. This fixes the
630 problem mentioned in
631 http://www.haskell.org/pipermail/glasgow-haskell-users/2010-April/018698.html
632 
633]
634[Fix bitrot in IO debugging code
635Ian Lynagh <igloo@earth.li>**20100413134339
636 Also switched to using Haskell Bools (rather than CPP) to en/disable it,
637 so it shouldn't break again in the future.
638]
639[Tiny code tidy-up
640Ian Lynagh <igloo@earth.li>**20100413011147]
641[remove old/wrong comment
642Simon Marlow <marlowsd@gmail.com>**20100325161403
643 Ignore-this: e6e377d44af48c4162d17d55bdf3f821
644]
645[withThread: block asynchronous exceptions before installing exception handler.
646Bas van Dijk <v.dijk.bas@gmail.com>**20100329131624
647 Ignore-this: be5aeb47dbd73807b5f94df11afbb81c
648 Note that I don't unblock the given io computation. Because AFAICS
649 withThread is only called with 'waitFd' which only performs an FFI
650 call which can't receive asynchronous exceptions anyway.
651]
652[runInUnboundThread: block asynchronous exceptions before installing exception handler
653Bas van Dijk <v.dijk.bas@gmail.com>**20100329131549
654 Ignore-this: a00c5e32fe3981ff87bedd367a69051e
655]
656[fix the deprecation message (GHC.IO.Handle.Base -> GHC.IO.Handle)
657Simon Marlow <marlowsd@gmail.com>**20100330121137
658 Ignore-this: 4ca8500a01ac93454507aa8f9dd001f9
659]
660[Make SampleVar an abstract newtype
661Bas van Dijk <v.dijk.bas@gmail.com>**20100318200349
662 Ignore-this: 27939e2a064b75e71cb146117346be30
663]
664[Fix bugs regarding asynchronous exceptions and laziness in Control.Concurrent.SampleVar
665Bas van Dijk <v.dijk.bas@gmail.com>**20100318200104
666 Ignore-this: 7376b2a3afe155daf233a8f1ddc0a7a
667  - Block asynchronous exceptions at the right places
668  - Force thunks before putting them in a MVar
669]
670[Write the thunk 'next' to the MVar
671Bas van Dijk <v.dijk.bas@gmail.com>**20100319125951
672 Ignore-this: dd25636cf220131385ff2fd32493d456
673]
674[change to use STM, fixing 4 things
675Simon Marlow <marlowsd@gmail.com>**20100318104436
676 Ignore-this: 551d30280a7941c08f5c3b14576bdd70
677   1. there was no async exception protection
678   2. there was a space leak (now new value is strict)
679   3. using atomicModifyIORef would be slightly quicker, but can
680      suffer from adverse scheduling issues (see #3838)
681   4. also, the STM version is faster.
682]
683[Tweak docs
684Ian Lynagh <igloo@earth.li>**20100312214129]
685[Fixed dead links in documentation of forkIO
686Bas van Dijk <v.dijk.bas@gmail.com>**20100308222415
687 Ignore-this: 7deb8fd064c867fbede2a6b2e9da4f15
688]
689[Documentation fixes in Control.Exception
690Bas van Dijk <v.dijk.bas@gmail.com>**20100301220442
691 Ignore-this: 761fcba401cbd1f47276ddfc9b5b80f2
692]
693[Plug two race conditions that could lead to deadlocks in the IO manager
694Simon Marlow <marlowsd@gmail.com>**20100225120255
695 Ignore-this: e6983d6b953104d370278ab3e4617e8b
696]
697[FIX #3866: improve documentation of Data.Data.Constr
698jpm@cs.uu.nl**20100224125506
699 Ignore-this: 3818c5d8fee012a3cf322fb455b6e5dc
700]
701[UNDO: Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
702Simon Marlow <marlowsd@gmail.com>**20100223101603
703 Ignore-this: 78becb2d39b3cd9a1a473a5811ca7d92
704]
705[Put the complexity in the length docs. Fixes trac #3680
706Ian Lynagh <igloo@earth.li>**20100221191425]
707[nhc98 should build Data.Functor.
708Malcolm.Wallace@cs.york.ac.uk**20100221163218]
709[Update the exitWith docs
710Ian Lynagh <igloo@earth.li>**20100213140004
711 Error pointed out by Volker Wysk <vw@volker-wysk.de>
712]
713[Handle NaN, -Infinity and Infinity in the toRational for Float/Double (#3676)
714Simon Marlow <marlowsd@gmail.com>**20100211101955
715 Ignore-this: 261415363303efca265e80290eac5f28
716]
717[For nhc98, import unsafeInterleaveIO rather than defining it here.
718Malcolm.Wallace@cs.york.ac.uk**20100204171021]
719[Stifle warning about unused return value
720benl@cse.unsw.edu.au**20100203025537]
721[fix #3832: use the locale encoding in openTempFile
722Simon Marlow <marlowsd@gmail.com>**20100120211830
723 Ignore-this: df4f778cc5fefb32290c798db722632c
724 Also while I was here fix an XXX: the Handle contained an
725 uninformative string like <fd: 4> for error messages rather than the
726 real file path.
727]
728[Fix the build: export void, so it doesn't give an unused binding warning
729Ian Lynagh <igloo@earth.li>**20100116174451]
730[hIsEOF: don't do any decoding (#3808)
731Simon Marlow <marlowsd@gmail.com>**20100112230317
732 Ignore-this: 6a384dd2d547ffe3ad3762920e5c1671
733]
734[Control.Monad: +void :: f a -> f ()
735gwern0@gmail.com**20100108214455
736 Ignore-this: 4dc07452315f2d1b4941903ff42fc45f
737 See http://hackage.haskell.org/trac/ghc/ticket/3292
738 Turns m a -> m (). Lets one call functions for their side-effects without
739 having to get rid of their return values with '>> return ()'. Very useful
740 in many contexts (parsing, IO etc.); particularly good for 'forkIO' and 'forM_',
741 as they demand return types of 'IO ()' though most interesting IO functions
742 return non-().
743]
744[Replace the implementation of mergesort with a 2x faster one.
745Malcolm.Wallace@cs.york.ac.uk**20091224152014
746 See ticket http://hackage.haskell.org/trac/ghc/ticket/2143.
747]
748[Restore previous Data.Typeable.typeOf*Default implementations for non-ghc.
749Malcolm.Wallace@cs.york.ac.uk**20091223142625
750 Not all compilers have ScopedTypeVariables.
751]
752[Add comments about double bounds-checking, and fast paths for rectangular arrays
753simonpj@microsoft.com**20091218165655
754 Ignore-this: ea0849419dc00927aba4bd410b1cc58d
755 
756 See Note [Double bounds-checking of index values] for the details.
757 
758 The fast paths omit the doubled checks for cases we know about
759]
760[Fix Trac #3245: memoising typeOf
761simonpj@microsoft.com**20091218155117
762 Ignore-this: 5a178a7f2222293c5ee0c3c43bd1b625
763 
764 The performance bug in #3245 was caused by computing the typeRep
765 once for each call of typeOf, rather than once for each dictionary
766 contruction.  (Computing TypeReps is reasonably expensive, because
767 of the hash-consing machinery.)
768 
769 This is readily fixed by putting the TypeRep construction outside
770 the lambda.  (Arguably GHC might have worked that out itself,
771 but it involves floating something between a type lambda and a
772 value lambda, which GHC doesn't currently do. If it happens a lot
773 we could fix that.)
774]
775[Mark 'index' as INLINE in GHC.Arr
776simonpj@microsoft.com**20091216170441
777 Ignore-this: a4df9d8acf496c8e0e9ce5a520509a2a
778 
779 This makes indexing much faster. See Trac #1216
780]
781[Comment the remaining orphan instance modules
782Ian Lynagh <igloo@earth.li>**20091206125021]
783[De-orphan Eq/Ord Float/Double
784Ian Lynagh <igloo@earth.li>**20091205181238]
785[Add comments to "OPTIONS_GHC -fno-warn-orphans" pragmas
786Ian Lynagh <igloo@earth.li>**20091205165854]
787[Data.Either.partitionEithers was insufficiently lazy.
788Malcolm.Wallace@cs.york.ac.uk**20091202032807
789 Ignore-this: 77e1b3288f66608c71458d8a91bcbe12
790 Spotted by Daniel Fischer.
791]
792[fix the docs regarding finalizer guarantees
793Simon Marlow <marlowsd@gmail.com>**20091130144409
794 Ignore-this: d1ab9532c74a002b8075ff60febcbe2d
795]
796[x86_64 requires more stack
797Malcolm.Wallace@cs.york.ac.uk**20091201033745]
798[check for size < 0 in mallocForeignPtrBytes and friends (#3514)
799Simon Marlow <marlowsd@gmail.com>**20091125143822
800 Ignore-this: 91077d01da2bbe1dfed5155e8b40da9
801]
802[hGetContents: close the handle properly on error
803Simon Marlow <marlowsd@gmail.com>**20091125123435
804 Ignore-this: bc37ff678acc6e547dc390285e056eb9
805 
806 When hGetContents caught an error it was closing the handle and then
807 throwing the exception, without updating the handle with the new
808 closed state.  This lead to a double-closed, which was the cause of
809 
810 *** glibc detected *** ./Setup: double free or corruption
811 
812 when iconv_close was called twice on the decoder.
813 
814 See http://hackage.haskell.org/trac/hackage/ticket/609
815]
816[Fix arities of mapFB and zipFB
817Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091126232219
818 Ignore-this: c4e14cd0a92622549c86e67237a40865
819]
820[Remove an unnecessary -fno-warn-orphans flag
821Ian Lynagh <igloo@earth.li>**20091126123404]
822[Tweak layout to work with alternative layout rule
823Ian Lynagh <igloo@earth.li>**20091125232349]
824[Tweak layout to be accepted by the alternative layout rul
825Ian Lynagh <igloo@earth.li>**20091125194147]
826[Make sure zipWithFB has arity 2
827Roman Leshchinskiy <rl@cse.unsw.edu.au>**20091125010003
828 Ignore-this: 4cf60c55666f03d22a9f5a6e07f52d36
829 
830 It gets 2 arguments in the "zipWith" rule but its arity was higher and the new
831 inliner didn't inline it sometimes, for instance here:
832 
833 mpp ::  [Double] -> [Double] -> [Double] -> [Double] -> [Double]
834 mpp as bs cs ds = zipWith (*) (zipWith (+) as bs) (zipWith (+) cs ds)
835 
836 This was a regression vs. 6.10.
837]
838[Remove an old comment
839Ian Lynagh <igloo@earth.li>**20091124134647]
840[De-orphan the Eq/Ord Integer instances
841Ian Lynagh <igloo@earth.li>**20091124133639]
842[Whitespace only
843Ian Lynagh <igloo@earth.li>**20091124133421]
844[Derive some more instances, rather than writing them by hand
845Ian Lynagh <igloo@earth.li>**20091124011747]
846[We can now derive Ord ()
847Ian Lynagh <igloo@earth.li>**20091124011416]
848[De-orphan tuple Eq/Ord instances
849Ian Lynagh <igloo@earth.li>**20091123233343]
850[Control.Exception.Base no longer has any orphans
851Ian Lynagh <igloo@earth.li>**20091123224905]
852[De-orphan the MonadFix ST instance for GHC
853Ian Lynagh <igloo@earth.li>**20091123223544]
854[Rearrange the contents of Control.Monad.ST; no functionality changes
855Ian Lynagh <igloo@earth.li>**20091123222702]
856[De-orphan the Eq/Ord [a] instances
857Ian Lynagh <igloo@earth.li>**20091123215635]
858[De-orphan the Eq/Ord Char instances
859Ian Lynagh <igloo@earth.li>**20091123202253]
860[De-orphan the Eq/Ord Bool instances
861Ian Lynagh <igloo@earth.li>**20091123201817]
862[Move Eq/Ord Ordering instances to de-orphan them
863Ian Lynagh <igloo@earth.li>**20091123194310]
864[Remove ffi warnings for nhc98.
865Malcolm.Wallace@cs.york.ac.uk**20091123063743]
866[Second attempt to fix #1185 (forkProcess and -threaded)
867Simon Marlow <marlowsd@gmail.com>**20091111151915
868 Ignore-this: fa5f5d5e4e080d4b612a37244f937f9c
869 
870 Patch 2/2: first patch is to ghc
871 
872 This time without dynamic linker hacks, instead I've expanded the
873 existing rts/Globals.c to cache more CAFs, specifically those in
874 GHC.Conc.  We were already using this trick for signal handlers, I
875 should have realised before.
876 
877 It's still quite unsavoury, but we can do away with rts/Globals.c in
878 the future when we switch to a dynamically-linked GHCi.
879]
880[Rollback #1185 fix
881Simon Marlow <marlowsd@gmail.com>**20091106140629
882 Ignore-this: cd5667e8474e37e01ba26a1984274811
883 
884 rolling back:
885 
886 Tue Nov  3 16:05:40 GMT 2009  Simon Marlow <marlowsd@gmail.com>
887   * Fix #1185: restart the IO manager after fork()
888   
889   This is the libraries/base part of the patch; there is a corresponding
890   patch to GHC itself.
891   
892   The main change is that we now keep track of the IO manager's ThreadId
893   in a top-level MVar, and ensureIOManagerIsRunning checks whether a
894   previous IO manager thread is alive before starting one.  In the child
895   of fork(), we can hence call ensureIOManagerIsRunning to restart the
896   IO manager.
897 
898     M ./GHC/Conc.lhs -46 +44
899 
900 Wed Nov  4 17:49:45 GMT 2009  Ian Lynagh <igloo@earth.li>
901   * Fix the build on Windows
902 
903     M ./GHC/Conc.lhs -6 +4
904]
905[Fix the build on Windows
906Ian Lynagh <igloo@earth.li>**20091104174945]
907[Fix #1185: restart the IO manager after fork()
908Simon Marlow <marlowsd@gmail.com>**20091103160540
909 Ignore-this: 6dc05464f1500104554637f4759738cc
910 
911 This is the libraries/base part of the patch; there is a corresponding
912 patch to GHC itself.
913 
914 The main change is that we now keep track of the IO manager's ThreadId
915 in a top-level MVar, and ensureIOManagerIsRunning checks whether a
916 previous IO manager thread is alive before starting one.  In the child
917 of fork(), we can hence call ensureIOManagerIsRunning to restart the
918 IO manager.
919]
920[improve the documentation for throwErrnoIfRetry
921Simon Marlow <marlowsd@gmail.com>**20091016112404
922 Ignore-this: b77275cacf730e15757946027168f63e
923]
924[Don't inline unpackFoldrCString ever
925simonpj@microsoft.com**20091029135350
926 Ignore-this: 85d672649b1b776efc7e97500b05d4f9
927]
928[Inline more default methods
929simonpj@microsoft.com**20091029135330
930 Ignore-this: 289c44b0afd6d5631c2a4e0664275ca9
931 
932 Namely Monad: (>>)
933        Eq:    (==), (/=)
934        Num:   (-), negate
935        Real:  quot, rem, div, mod, recip, (/), truncate
936        Float: (**), logBase, sqrt, tan, tanh
937]
938[Move error messages out of INLINEd default methods
939simonpj@microsoft.com**20091029135118
940 Ignore-this: 9e35dc947f94827a3529eb53a41575fd
941 
942 No need to duplicate the error generation!
943]
944[Exploit now-working default-method INLINE pragmas for Data.Bits
945simonpj@microsoft.com**20091029135041
946 Ignore-this: 8adf225f31ca7a3181ee087e9e4fe535
947 
948 * Add INLINE pragmas to default methods for class Bits
949 
950 * Remove redundant instance methods elsewhere, now that
951   the default method will do the job
952]
953[Tidy up and comment imports
954simonpj@microsoft.com**20091029134414
955 Ignore-this: bf2be31035de975d8995e988933cc940
956]
957[Inline foldr and (.) when applied to two arguments not three
958simonpj@microsoft.com**20091029134335
959 Ignore-this: fccb6f3e90e15f44cb465814be85ede2
960 
961 The new INLINE story is (by design) arity-sensitive, so we must
962 put fewer argument on the LHS for foldr and (.)
963]
964[dirUtils.c no longer available
965Malcolm.Wallace@cs.york.ac.uk**20091013093833]
966[Make hGetContents throw an exception if an error is encountered
967Simon Marlow <marlowsd@gmail.com>**20091012152955
968 Ignore-this: 9f7a7176193eab25c9daaacd9261f2de
969 
970 Strictly speaking this breaks Haskell 98 compatibility, which requires
971 hGetContents to just end the lazy stream silently if an error is
972 encountered.  However, for a few reasons we think it will make
973 everyone's life a bit easier if we make this change
974 
975  1. Errors will be a lot more common in GHC 6.12.1, in the form
976     of Unicode decoding errors.
977 
978  2. When Haskell 98 was designed, we didn't know how to throw
979     exceptions from inside lazy I/O, but now we do.
980 
981  3. If anyone is actually relying on the previous behaviour, their
982     code is arguably broken.
983]
984[Re-instate System.Console.Getopt for nhc98 builds.
985Malcolm.Wallace@cs.york.ac.uk**20091013092843
986 Although it was split out of base a while back, that change was
987 reverted for ghc soon afterwards, but nhc98 never noticed.
988]
989[Roll back "Another instance of nhc98's strange import semantics."
990Ian Lynagh <igloo@earth.li>**20091009185618
991 Fri Oct  9 14:50:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
992 GHC (correctly) warns about the unused import, which breaks the validate
993 build.
994]
995[Roll back "Cope with nhc98's (occasionally-strange) import semantics"
996Ian Lynagh <igloo@earth.li>**20091009184704
997 Fri Oct  9 14:43:51 BST 2009  Malcolm.Wallace@cs.york.ac.uk
998 GHC (correctly) warns about the unused import, which breaks the validate
999 build.
1000]
1001[It seems that nhc98 needs defaulting in Data.Fixed.
1002Malcolm.Wallace@cs.york.ac.uk**20091009135242]
1003[Another instance of nhc98's strange import semantics.
1004Malcolm.Wallace@cs.york.ac.uk**20091009135051]
1005[Make Data.Functor compatible with non-GHC compilers.
1006Malcolm.Wallace@cs.york.ac.uk**20091009134821]
1007[Cope with nhc98's (occasionally-strange) import semantics.
1008Malcolm.Wallace@cs.york.ac.uk**20091009134351]
1009[Fix gratuitous breakage of nhc98 in System.IO.
1010Malcolm.Wallace@cs.york.ac.uk**20091009134001]
1011[Fix gratuitous breakage of nhc98 in Control.Exception.Base.
1012Malcolm.Wallace@cs.york.ac.uk**20091009133615]
1013[Fix gratuitous breakage of non-GHC in Data.Fixed.
1014Malcolm.Wallace@cs.york.ac.uk**20091009133330]
1015[Fix gratuitous breakage for non-GHC in Data.Bits.
1016Malcolm.Wallace@cs.york.ac.uk**20091009133257]
1017[Use UTF-32LE instead of UTF32LE
1018Simon Marlow <marlowsd@gmail.com>**20091006100207
1019 Ignore-this: 7f881e36543d250ef848c9f60d67655a
1020 The latter is not recognised by some iconv implementations.
1021]
1022[Strip any Byte Order Mark (BOM) from the front of decoded streams.
1023Ben.Lippmeier@anu.edu.au*-20090930084229
1024 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
1025 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
1026 of the stream, but Linux iconv doesn't.
1027]
1028[use UTF32BE/UTF32LE instead of UCS-4/UCS-4LE
1029Simon Marlow <marlowsd@gmail.com>**20091005101554
1030 Ignore-this: 2aef5e9bec421e714953b7aa1bdfc1b3
1031]
1032[Strip any Byte Order Mark (BOM) from the front of decoded streams.
1033Ben.Lippmeier@anu.edu.au**20090930084229
1034 Ignore-this: d0d0c3ae87b31d71ef1627c8e1786445
1035 When decoding to UTF-32, Solaris iconv inserts a BOM at the front
1036 of the stream, but Linux iconv doesn't.
1037]
1038[Add traceEvent :: String -> IO ()
1039Simon Marlow <marlowsd@gmail.com>**20090925141257
1040 Ignore-this: 8b1888bbf9682ffba13f815b6000e4b1
1041 For emitting an event via the RTS tracing framework
1042]
1043[Fix the error message when flushing the read buffer of a non-seekable Handle
1044Simon Marlow <marlowsd@gmail.com>**20090923090536
1045 Ignore-this: 4342026df93759d99480f4e13f80a492
1046]
1047[Fix #3534: No need to flush the byte buffer when setting binary mode
1048Simon Marlow <marlowsd@gmail.com>**20090923090445
1049 Ignore-this: 625817ed7ae2c12291eb993a99dc640a
1050]
1051[Use let !y = x in .. x .. instead of seq in $! and evaluate (#2273)
1052Simon Marlow <marlowsd@gmail.com>**20090916140454]
1053[make some Applicative functions into methods, and split off Data.Functor (proposal #3335)
1054Ross Paterson <ross@soi.city.ac.uk>**20090915173109
1055 Ignore-this: a0cff4de6dfdbcbd56a66101bc4855a9
1056 
1057 The following functions
1058 
1059     (<$) :: Functor f => a -> f b -> f a
1060     (*>) :: Applicative f => f a -> f b -> f b
1061     (<*) :: Applicative f => f a -> f b -> f a
1062     some :: Alternative f => f a -> f [a]
1063     many :: Alternative f => f a -> f [a]
1064 
1065 are moved into the corresponding classes, with the existing implementations
1066 as default definitions.  This gives people creating instances the option of
1067 defining specialized implementations of these functions, though they should
1068 be equivalent to the default definitions.
1069 
1070 Although (<$) is now a method of the Functor class, it is hidden in the
1071 re-export by the Prelude, Control.Monad and Monad.  The new module
1072 Data.Functor exposes the full class, plus the function (<$>).  These are
1073 also re-exported by Control.Applicative.
1074]
1075[On Windows, use the console code page for text file encoding/decoding.
1076Judah Jacobson <judah.jacobson@gmail.com>**20090913022126
1077 Ignore-this: 86c2f2db8ef92b751599795d3195187b
1078 
1079 We keep all of the code page tables in the module
1080 GHC.IO.Encoding.CodePage.Table.  That file was generated automatically
1081 by running codepages/MakeTable.hs; more details are in the comments at the
1082 start of that script.
1083 
1084 Storing the lookup tables adds about 40KB to each statically linked executable;
1085 this only increases the size of a "hello world" program by about 7%.
1086 
1087 Currently we do not support double-byte encodings (Chinese/Japanese/Korean), since
1088 including those codepages would increase the table size to 400KB.  It will be
1089 straightforward to implement them once the work on library DLLs is finished.
1090]
1091[Fix "init" docs: the input list need not be finite. Fixes trac #3465
1092Ian Lynagh <igloo@earth.li>**20090911210437]
1093[Bump base version to 4.2.0.0
1094Ian Lynagh <igloo@earth.li>**20090911153913]
1095[Address #3310
1096Simon Marlow <marlowsd@gmail.com>**20090830152850
1097 Ignore-this: 40c7f7c171ee299a83092fd360a952b7
1098 
1099  - Rename BlockedOnDeadMVar   -> BlockedIndefinitelyOnMVar
1100  - Rename BlockedIndefinitely -> BlockedIndefinitelyOnSTM
1101  - instance Show BlockedIndefinitelyOnMVar is now
1102      "blocked indefinitely in an MVar operation"
1103  - instance Show BlockedIndefinitelyOnSTM is now
1104      "blocked indefinitely in an STM transaction"
1105 
1106 clients using Control.OldException will be unaffected (the new
1107 exceptions are mapped to the old names).  However, for base4-compat
1108 we'll need to make a version of catch/try that does a similar
1109 mapping.
1110]
1111[Fix unicode conversion for MSB architectures
1112Ben.Lippmeier@anu.edu.au**20090830130028
1113 This fixes the SPARC/Solaris build.
1114]
1115[Fix #3441: detect errors in partial sequences
1116Simon Marlow <marlowsd@gmail.com>**20090830075909
1117 Ignore-this: d12a75d95e0cae5eb1555266810ec281
1118]
1119[Fix hWaitForInput
1120Simon Marlow <marlowsd@gmail.com>**20090827152116
1121 Ignore-this: 2550e911f1a4d4357a5aa8d1764238ce
1122 It was erroneously waiting when there were bytes to decode waiting in
1123 the byte buffer.
1124]
1125[fix debugging code
1126Simon Marlow <marlowsd@gmail.com>**20090827150628
1127 Ignore-this: e1c82fdc19a22e247cd69ff6fa11921d
1128]
1129[Allow for configurable iconv include and library locations.
1130Matthias Kilian <kili@outback.escape.de>**20090826154406
1131 Ignore-this: be95fab611a5534cf184b508964ed498
1132 This should help to fix the build on OpenBSD.
1133]
1134[typo in comment
1135Simon Marlow <marlowsd@gmail.com>**20090826085252
1136 Ignore-this: 1903ee0f354157a6ed3871c100f6b1b9
1137]
1138[un-hide some modules from the Haddock docs
1139Simon Marlow <marlowsd@gmail.com>**20090825152457
1140 Ignore-this: dce6606f93cf977fb24ebe99082dfa62
1141]
1142[Apply fix for #1548, from squadette@gmail.com
1143Simon Marlow <marlowsd@gmail.com>**20090819120700
1144 Ignore-this: 31c237c46a6445f588ed4b8c51bb6231
1145]
1146[improvements to Data.Fixed: instances for Typeable and Data, more predefined types
1147Ashley Yakeley <ashley@semantic.org>**20090812055058
1148 Ignore-this: feeece36d5632f02a05d137d2a39ab78
1149]
1150[Fix "Cabal check" warnings
1151Ian Lynagh <igloo@earth.li>**20090811215856]
1152[Add a GHC.Constants module; fixes trac #3094
1153Ian Lynagh <igloo@earth.li>**20090809183252]
1154[Apply proposal #3393
1155Ian Lynagh <igloo@earth.li>**20090809134717
1156 Add openTempFileWithDefaultPermissions and
1157 openBinaryTempFileWithDefaultPermissions.
1158]
1159[Add some more C wrappers; patch from Krister Walfridsson
1160Ian Lynagh <igloo@earth.li>**20090807200631
1161 Fixes 21 testsuite errors on NetBSD 5.99.
1162]
1163[Fixing configure for autoconf 2.64
1164Alexander Dunlap <alexander.dunlap@gmail.com>**20090805060748
1165 Ignore-this: 992ab91ae3d68c12dbb265776e33e243
1166]
1167[add INLINE toList
1168Ross Paterson <ross@soi.city.ac.uk>**20090806142853
1169 Ignore-this: aba16aabb17d5dca44f15d188945680e
1170 
1171 In anticipation of the fixing of #2353.
1172]
1173[fix a copyright
1174Simon Marlow <marlowsd@gmail.com>**20090805134045
1175 Ignore-this: b0ffbdd38fbba121e8bcba37c4082a60
1176]
1177[Tweak the BufferedIO class to enable a memory-mapped file implementation
1178Simon Marlow <marlowsd@gmail.com>**20090805134036
1179 Ignore-this: ec67d7a0a6d977438deaa342503f77e0
1180 We have to eliminate the assumption that an empty write buffer can be
1181 constructed by setting the buffer pointers to zero: this isn't
1182 necessarily the case when the buffer corresponds to a memory-mapped
1183 file, or other in-memory device implementation.
1184]
1185[Deprecate Control.OldException
1186Ian Lynagh <igloo@earth.li>**20090804143910]
1187[Windows build fix, following RTS tidyup
1188Simon Marlow <marlowsd@gmail.com>**20090803131121
1189 Ignore-this: ce862fb91c2b234211a8757f98690778
1190]
1191[Updates to follow the RTS tidyup
1192Simon Marlow <marlowsd@gmail.com>**20090801220743
1193 Ignore-this: 6e92412df93a66c12d75344053d5634
1194 C functions like isDoubleNaN moved here (primFloat.c)
1195]
1196[Add integer-simple as a build option
1197Ian Lynagh <igloo@earth.li>**20090722013151]
1198[Use shift[LR]Integer in the Bits Integer instance
1199Ian Lynagh <igloo@earth.li>**20090721222440]
1200[depend directly on integer-gmp, rather than indirecting through integer
1201Ian Lynagh <igloo@earth.li>**20090721185228]
1202[Move the instances of Functor and Monad IO to GHC.Base, to avoid orphans
1203Simon Marlow <marlowsd@gmail.com>**20090722102130
1204 Ignore-this: a7d85ac0025d559674249de0108dbcf4
1205]
1206[move "instance Exception Dynamic" so it isn't an orphan
1207Simon Marlow <marlowsd@gmail.com>**20090721093854
1208 Ignore-this: 5ede91ecfec2112c91b699d4de87cd02
1209]
1210[Improve the index checking for array accesses; fixes #2120 #2669
1211Ian Lynagh <igloo@earth.li>**20090719153228
1212 As well as checking that offset we are reading is actually inside the
1213 array, we now also check that it is "in range" as defined by the Ix
1214 instance. This fixes confusing behaviour (#2120) and improves some error
1215 messages (#2669).
1216]
1217[Make chr say what its argument was, if it's a bad argument
1218Ian Lynagh <igloo@earth.li>**20090718151049]
1219[remove unused warning
1220Simon Marlow <marlowsd@gmail.com>**20090715124416
1221 Ignore-this: 31f613654089d0f4a44363946087b41e
1222]
1223[warning fix: -fno-implicit-prelude -> -XNoImplicitPrelude
1224Simon Marlow <marlowsd@gmail.com>**20090715122839
1225 Ignore-this: dc8957249731d5bcb71c01899e5adf2b
1226]
1227[Add hGetEncoding :: Handle -> IO (Maybe TextEncoding)
1228Simon Marlow <marlowsd@gmail.com>**20090715122519
1229 Ignore-this: 14c3eff996db062da1199739781e4708
1230 as suggested during the discussion on the libraries list
1231]
1232[Add more documentation to mkTextEncoding
1233Simon Marlow <marlowsd@gmail.com>**20090715122414
1234 Ignore-this: 97253b2624267df3a246a18121e8ea81
1235 noting that "//IGNORE" and "//TRANSLIT" suffixes can be used with GNU
1236 iconv.
1237]
1238[Add the utf8_bom codec
1239Simon Marlow <marlowsd@gmail.com>**20090715122257
1240 Ignore-this: 1c9396cd805201fe873a39382ced79c7
1241 as suggested during the discussion on the libraries list.
1242]
1243[Export Unicode and newline functionality from System.IO; update Haddock docs
1244Simon Marlow <marlowsd@gmail.com>**20090713113104
1245 Ignore-this: c3f017a555335aa55d106253393f72e2
1246]
1247[add a comment about the non-workingness of CHARBUF_UTF16
1248Simon Marlow <marlowsd@gmail.com>**20090707124406
1249 Ignore-this: 98d00411b68d688b3b4cffc9507b1f35
1250]
1251[Fix build on Windows
1252Ian Lynagh <igloo@earth.li>**20090711004351]
1253[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1254Ian Lynagh <igloo@earth.li>**20090710204513]
1255[Use throwErrnoIfMinus1_ when calling getrusage
1256Ian Lynagh <igloo@earth.li>**20090710204221]
1257[Remove an unused import
1258Ian Lynagh <igloo@earth.li>**20090710153345]
1259[reportStackOverflow now returns IO ()
1260Ian Lynagh <igloo@earth.li>**20090710153257
1261 It used to do "return undefined" to return IO a.
1262]
1263[GHC.Conc.reportError now returns IO ()
1264Ian Lynagh <igloo@earth.li>**20090710152646
1265 It used to return IO a, by "return undefined".
1266]
1267[Fix some "warn-unused-do-bind" warnings where we want to ignore the value
1268Ian Lynagh <igloo@earth.li>**20090710152526]
1269[Minor SampleVar refactoring
1270Ian Lynagh <igloo@earth.li>**20090710151438]
1271[Fix "warn-unused-do-bind" warnings in GHC/IO/Handle/Text.hs
1272Ian Lynagh <igloo@earth.li>**20090710122905]
1273[Fix some "warn-unused-do-bind" warnings where we just want to ignore the result
1274Ian Lynagh <igloo@earth.li>**20090710005638]
1275[Use the result of writeCharBuf in GHC/IO/Encoding/Latin1.hs too
1276Ian Lynagh <igloo@earth.li>**20090710004032]
1277[Minor code tidyups in GHC.Conc
1278Ian Lynagh <igloo@earth.li>**20090710003801]
1279[Fix "warn-unused-do-bind" warning in GHC.Conc
1280Ian Lynagh <igloo@earth.li>**20090710003530
1281 If we fail to communicate with the IO manager then we print a warning
1282 using debugErrLn from the ghc-prim package.
1283]
1284[Fix "warn-unused-do-bind" warnings in System.Posix.Internals
1285Ian Lynagh <igloo@earth.li>**20090709164546]
1286[Fix "warn-unused-do-bind" warnings where we really do want to ignore the result
1287Ian Lynagh <igloo@earth.li>**20090709163912]
1288[Add back imports needed on Windows
1289Ian Lynagh <igloo@earth.li>**20090707181924]
1290[Remove unused imports
1291Ian Lynagh <igloo@earth.li>**20090707115810]
1292[Remove unused imports from base
1293simonpj@microsoft.com**20090706111842
1294 Ignore-this: f9b5f353e3bb820f787c56d615b28765
1295 
1296 These unused imports are detected by the new unused-import code
1297 
1298]
1299[Use the result of writeCharBuf
1300Simon Marlow <marlowsd@gmail.com>**20090706133303
1301 Ignore-this: 52288dd559bf4c4f313df6197091d935
1302   
1303 This only makes a difference when CHARBUF_UTF16 is in use, which it
1304 normally isn't.  I suspect CHARBUF_UTF16 doesn't currently work for
1305 other reasons (CHARBUF_UTF16 was an experiment before I wrote the
1306 GHC.IO.Encoding.UTF* codecs), but this patch at least makes it
1307 slightly closer to working.
1308]
1309[Remove some cruft from Data.HashTable
1310Ian Lynagh <igloo@earth.li>**20090706181630]
1311[Add 'eof' to Text.ParserCombinators.ReadP
1312simonpj@microsoft.com**20090706111801
1313 Ignore-this: 2aea7b848e00c894761bc4011adaa95d
1314 
1315 Add a ReadP parser that succeeds at the end of input. Very useful!
1316 
1317]
1318[Don't export CLDouble for GHC; fixes trac #2793
1319Ian Lynagh <igloo@earth.li>**20090705155120
1320 We never really supported CLDouble (it was a plain old double underneath),
1321 and pretending that we do does more harm than good.
1322]
1323[a byte between 0x80 and 0xBF is illegal immediately (#3341)
1324Simon Marlow <marlowsd@gmail.com>**20090702081415
1325 Ignore-this: dc19ef59a1a21118d5a7dd38aa2f611c
1326]
1327[avoid a warning
1328Simon Marlow <marlowsd@gmail.com>**20090630084134
1329 Ignore-this: c92a45ee216faf01327feae9fe06d6e2
1330]
1331[Add a wrapper for libiconv.
1332Matthias Kilian <kili@outback.escape.de>**20090629183634
1333 Ignore-this: 23c6047c0d71b745b495cc223574a47f
1334]
1335[#include <sys/times.h> if we have it (should fix build problems)
1336Simon Marlow <marlowsd@gmail.com>**20090629085351
1337 Ignore-this: a35e93b37ca9595c73460243180f4b9d
1338]
1339[set binary mode for existing FDs on Windows (fixes some GHCi test failures)
1340Simon Marlow <marlowsd@gmail.com>**20090626120522
1341 Ignore-this: 580cf636e9c77d8427aff6861d089481
1342]
1343[Move directory-related stuff to the unix package
1344Simon Marlow <marlowsd@gmail.com>**20090625120325
1345 Ignore-this: b997b3cbce0a46ca87ad825bbdc0a411
1346 now that it isn't used on Windows any more.
1347]
1348[TAG 2009-06-25
1349Ian Lynagh <igloo@earth.li>**20090625160056]
1350Patch bundle hash:
13513c3b0f34d7ef092b9a9caa21a997ec01cf09a90e