Ticket #2557: 2557.patch

File 2557.patch, 53.0 KB (added by judahj, 6 years ago)
Line 
1Sat Aug 30 14:14:29 PDT 2008  Judah Jacobson <judah.jacobson@gmail.com>
2  * Don't build ghc in stage1 with editline (since we already don't build ghci).  Fixes #2557.
3
4New patches:
5
6[Don't build ghc in stage1 with editline (since we already don't build ghci).  Fixes #2557.
7Judah Jacobson <judah.jacobson@gmail.com>**20080830211429] {
8hunk ./compiler/Makefile 119
9 CONFIGURE_FLAGS_STAGE2 += --ghc-option=-DDEBUGGER
10 endif
11 
12-# Enable editline if either:
13-#   - we're building stage 1 and $(GhcHasEditline)="YES"
14+# Enable editline if:
15 #   - we're building stage 2/3, and we have built the editline package
16 #
17 # But we don't enable editline on Windows, as Windows terminals have
18hunk ./compiler/Makefile 129
19 CONFIGURE_FLAGS_STAGE1 += --flags=-editline
20 CONFIGURE_FLAGS_STAGE2 += --flags=-editline
21 else
22-ifeq "$(GhcHasEditline)" "YES"
23-CONFIGURE_FLAGS_STAGE1 += --flags=editline
24-else
25 CONFIGURE_FLAGS_STAGE1 += --flags=-editline
26hunk ./compiler/Makefile 130
27-endif
28 ifeq "$(wildcard $(FPTOOLS_TOP_ABS)/libraries/editline/unbuildable)" ""
29 CONFIGURE_FLAGS_STAGE2 += --flags=editline
30 else
31}
32
33Context:
34
35[Actually, that's use a separate haddock repo for now
36Ian Lynagh <igloo@earth.li>**20080830123032
37 Everything will be changing in a couple of weeks anyway, so let's
38 just keep it simple for the time being.
39]
40[Always build docs/{ext-core,storage-mgt}, even when making a bindist
41Ian Lynagh <igloo@earth.li>**20080829152141]
42[Fix linkage on OpenBSD.
43kili@outback.escape.de**20080828150331]
44[push-all needs to skip http:// repos, as we don't know how to push to them
45Ian Lynagh <igloo@earth.li>**20080829130019
46 Alternatives would be to have a GHC haddock repo (which I believe David
47 is keen to avoid), or to put the SSH URL in the packages file too (which
48 would probably mean moving to a more structured file format).
49]
50[Remove install-inplace directories when cleaning
51Ian Lynagh <igloo@earth.li>**20080828221701]
52[Use the right ghc-pkg when installing the rts package
53Ian Lynagh <igloo@earth.li>**20080828191512]
54[When installing, we can't rely on the ghc-pkg we just installed to work
55Ian Lynagh <igloo@earth.li>**20080828182447
56 If DESTDIR is defined, then on *nix machines ghc-pkg is a script that
57 doesn't point to the right path. Therefore we use the ghc-pkg binary
58 directly.
59]
60[Tweaks for installs into a DESTDIR
61Ian Lynagh <igloo@earth.li>**20080828182148]
62[Handle the html target in compiler/ and ghc/
63Ian Lynagh <igloo@earth.li>**20080828151429]
64[Turn on optimisation for stage2 when validating
65Ian Lynagh <igloo@earth.li>**20080828142320
66 This brings my validate down from 22mins to 16 mins.
67 Compiling stage2 takes longer, but we gain a faster haddock, faster
68 running of the tests, and faster building of the utils to be installed.
69]
70[haddock 2 doc fixes
71Ian Lynagh <igloo@earth.li>**20080828142210]
72[Remove an erroneous comma from the Makefile
73Ian Lynagh <igloo@earth.li>**20080828132316]
74[Disambiguate some names in haddock comments
75Ian Lynagh <igloo@earth.li>**20080828125018]
76[Update the build system to handle building and using haddock2
77Ian Lynagh <igloo@earth.li>**20080828124814
78 One side-effect of this is that we need to build the install-utils with
79 stage2 rather than stage1 as we need the ghc package.
80]
81[Fix a haddock2 failure
82Ian Lynagh <igloo@earth.li>**20080828124721]
83[Don't check that we have found haddock
84Ian Lynagh <igloo@earth.li>**20080828124608
85 We now use an in-tree haddock instead
86]
87[Don't fail if libraries/bootstrapping already exists
88Ian Lynagh <igloo@earth.li>**20080827185024]
89[When haddocking the libraries, tell Cabal where haddock is
90Ian Lynagh <igloo@earth.li>**20080827182731
91 We can't do it when configuring, because we haven't built haddock yet
92]
93[Build haddock with the stage1 compiler
94Ian Lynagh <igloo@earth.li>**20080814112221]
95[Teach darcs-all about absolute repos, and add haddock to packages
96Ian Lynagh <igloo@earth.li>**20080813220158]
97[Improve documentation of stolen syntax
98simonpj@microsoft.com**20080828125211
99 
100 This patch adds a section that summarises what syntax is stolen by
101 which flags.  The section is at the end of the "syntactic extensions"
102 section of the manual.
103 
104 
105]
106[Fix Trac #2529: derived read for prefix constructor operators
107simonpj@microsoft.com**20080828121006]
108[Fix Trac #745: improve error recoevery for type signatures
109simonpj@microsoft.com**20080827153322
110 
111 It turns out that fixing Trac #745 is easy using mapAndRecoverM,
112 and tidies up the code nicely in several places.  Hurrah.
113 
114]
115[Fix Trac #2538: better error messages when validating types
116simonpj@microsoft.com**20080827153051
117 
118 This fix solely concerns error messages, and uses a bit of contextual
119 information to suggest plausible flags. 
120 
121 It was rather more fiddly to implement than I expected.  Oh well.
122 
123]
124[Fix Trac #2520: duplicate symbols
125simonpj@microsoft.com**20080827152728
126 
127 The problem here was that were were quantifying over an *External* Name,
128 which causes no end of confusion.  See Note [Const rule dicts] in DsBinds.
129 
130 The fix is very easy, I'm happy to say.
131 
132 
133]
134[Only specialise on dictionaries that have some interesting structure
135simonpj@microsoft.com**20080827152540
136 
137 I discovered by accident that we were generating utterly useless
138 specialisations.  See Note [Interesting dictionary arguments] in Specialise.
139 
140 This patch used SimplUtils.interestingArg to restrict specialisation to
141 cases where the dictionary acutally has some information to give us.
142 
143]
144[Better documentation for -XLiberalTypeSynonyms, and steal forall keyword
145simonpj@microsoft.com**20080827151926
146 
147 In my travels through the front end I discoverd that -XLiberalTypeSynonyms is
148 rather thinly described.  Furthermore, it alleges that you can write a
149 forall on the RHS of a type synonym decl, so that means it should enable
150 the forall keyword.
151 
152 
153]
154[re-fix of #1205, fix #2542
155Simon Marlow <marlowsd@gmail.com>**20080827102414
156 New form of :load in GHCi:
157 
158 > :load *A
159 
160 forces A to be loaded as byte-code.  See the manual for details.  The
161 previous behaviour for specifying filenames vs. module names on the
162 command line and in :load has been restored.
163 
164 The Target datatype has a new Bool field, which is True if the target
165 is allowed to be loaded from compiled code, or False otherwise, so
166 this functionality is available via the GHC API.  guessTarget
167 understands the *-prefix form for specifying targets.
168]
169[Improve documentation of MagicHash and primitive types generally (Trac #2547)
170simonpj@microsoft.com**20080827090544]
171[Give locations of flag warnings/errors
172Ian Lynagh <igloo@earth.li>**20080826185641]
173[Remove a now-redundant comment
174Ian Lynagh <igloo@earth.li>**20080826182629]
175[Separate the static flag parser from the static global variables
176Ian Lynagh <igloo@earth.li>**20080826155612
177 This allows us to avoid a module import loop:
178 CmdLineParser -> SrcLoc -> Outputable -> StaticFlags -> CmdLineParser
179]
180[Move pprFastFilePath from SrcLoc to Outputable
181Ian Lynagh <igloo@earth.li>**20080826144452]
182[Fix flaggery for RULES (cf Trac #2497)
183simonpj@microsoft.com**20080826122121
184 
185 This patch executes the plan described in the discussion in Trac #2497.
186 Specficially:
187 
188     * Inside a RULE, switch on the forall-as-keyword in the lexer,
189       unconditionally. (Actually this is done by an earlier patch.)
190 
191     * Merge the -XScopedTypeVariables and -XPatternSignatures flags,
192       and deprecate the latter. Distinguishing them isn't senseless,
193       but it's jolly confusing.
194 
195     * Inside a RULE, switch on -XScopedTypeVariables unconditionally.
196 
197     * Change -frewrite-rules to -fenable-rewrite-rules; deprecate the former.
198       Internally the DynFlag is now Opt_EnableRewriteRules.
199 
200 There's a test in typecheck/should_compile/T2497.hs
201 
202]
203[always treat 'forall' and '.' as reserved keywords inside RULES pragmas
204Simon Marlow <marlowsd@gmail.com>**20080820132911]
205[Fix a nasty float-in bug
206simonpj@microsoft.com**20080822133427
207 
208 This is a long-standing bug in FloatIn, which I somehow managed to
209 tickle (it's actually surprisingly hard to provoke which is why
210 it has not shown up before).
211 
212 The problem was that we had a specialisation like this:
213 
214   let
215        f_spec = ...
216   in let
217        {-# RULE f Int = f_spec #-}
218        f = ...
219   in
220        <body>
221 
222 The 'f_spec' binding was being floated inside the binding for 'f',
223 which makes the RULE invalid becuase 'f_spec' isn't in scope.
224 
225 We just need to add the free variables of the RULE in the right
226 places...
227       
228]
229[Make rule printing wobble less
230simonpj@microsoft.com**20080821123100
231 
232 a) When generating specialisations, include the types in the name
233    of the rule, to avoid having rules with duplicate names.
234    (The rule name is used to put rules in canonical order for
235    fingerprinting.)
236 
237 b) In Specialise and SpecConstr use a new function Rules.pprRulesForUser
238    to print rules in canonical order.  This reduces unnecessary wobbling
239    in test output, notably in T2486
240 
241]
242[syb is now in its own package
243Ian Lynagh <igloo@earth.li>**20080825214124]
244[Fix "runghc foo" where the program is foo.hs or foo.lhs
245Ian Lynagh <igloo@earth.li>**20080825194352]
246[When making bindists, check that we know where we are
247Ian Lynagh <igloo@earth.li>**20080825190912]
248[Fix the ghci script; fixes trac #2485
249Ian Lynagh <igloo@earth.li>**20080825183247]
250[Update .darcs-boring for the new libraries, plus some other odds and ends
251Ian Lynagh <igloo@earth.li>**20080825154249]
252[Read the packages file in strictly in darcs-all and push-all
253Ian Lynagh <igloo@earth.li>**20080825145908
254 This fixes a problem where darcs tries to update it while perl has it
255 open. On Windows this fails. Spotted by Claus Reinke.
256]
257[Remove the InstallShield directory; it is no longer used
258Ian Lynagh <igloo@earth.li>**20080825143356]
259[Remove the now-unused WindowsInstaller directory
260Ian Lynagh <igloo@earth.li>**20080825143214]
261[Set datadir=libdir; fixes trac #2541
262Ian Lynagh <igloo@earth.li>**20080825141653
263 GHC needs package.conf and things like unlit to be in the same place
264]
265[Get everything building with base 4 in the HEAD
266Ian Lynagh <igloo@earth.li>**20080825120907
267 Some things were using the base3 compat library.
268]
269[Change references to the old BIN_DIST_TARBALL to the new BIN_DIST_TAR_BZ2
270Ian Lynagh <igloo@earth.li>**20080824233821]
271[Fix the "have we got all the bootlibs" check
272Ian Lynagh <igloo@earth.li>**20080824175106
273 The packages file has an extra component (for "darcs" or "git") so we need
274 to take that into account when grepping.
275]
276[concurrent, unique, timeout have now been split off from base
277Ian Lynagh <igloo@earth.li>**20080824135052]
278[Actually tar up Windows bindists
279Ian Lynagh <igloo@earth.li>**20080824135032]
280[getopt is now split off from base
281Ian Lynagh <igloo@earth.li>**20080824020156]
282[st is now split off from base
283Ian Lynagh <igloo@earth.li>**20080823222941]
284[Ignore install-docs in compiler/Makefile and ghc/Makefile
285Ian Lynagh <igloo@earth.li>**20080821231349]
286[Fix making bindists
287Ian Lynagh <igloo@earth.li>**20080821223537
288 We were looking for executable things, but not requiring that they were
289 files. So we were adding lots of directories (and thus their contents)
290 to the bindists, massively inflating their size.
291]
292[mk/cabal.mk now turns on -Wall
293Ian Lynagh <igloo@earth.li>**20080821173102]
294[Fix warnings in runghc
295Ian Lynagh <igloo@earth.li>**20080821173051]
296[Make some utils -Wall clean
297Ian Lynagh <igloo@earth.li>**20080821153914]
298[unix is now warning-free
299Ian Lynagh <igloo@earth.li>**20080821120128]
300[Fix references to exceptions from the RTS
301Ian Lynagh <igloo@earth.li>**20080821110620
302 We now need to make sure that they have been toException'd.
303 Also, the RTS doesn't know about the Deadlock exception any more.
304]
305[It doesn't look like HSprel.def is used any more
306Ian Lynagh <igloo@earth.li>**20080821105914]
307[base (and base3-compat) is no warning-free
308Ian Lynagh <igloo@earth.li>**20080821001608]
309[Don't put symlinks in bindists, but the files they point at instead
310Ian Lynagh <igloo@earth.li>**20080820134141]
311[Fix making it possible to build without the NCG
312Ian Lynagh <igloo@earth.li>**20080818202845
313 If
314     GhcWithNativeCodeGen = NO
315 then we don't build the NCG modules, and we define OMIT_NATIVE_CODEGEN
316]
317[Comment fix
318Ian Lynagh <igloo@earth.li>**20080817210851]
319[Remove GHC_TOP; we don't actually use it
320Ian Lynagh <igloo@earth.li>**20080817210226]
321[We need to give make some more clues
322Ian Lynagh <igloo@earth.li>**20080817140005
323 Otherwise it can't work out how to make in ghc/ any more.
324]
325[Change how we know whether or not we are validating
326Ian Lynagh <igloo@earth.li>**20080817123311
327 We now set Validating=YES in mk/are-validating.mk rather than on the
328 commandline. This means that if you build a tree with validate then
329 just running make in it will use the validate flags.
330 "make distclean" removes mk/are-validating.mk, putting us back in
331 standard build mode.
332]
333[Don't clean the multi-stage stuff unless CLEAN_ALL_STAGES is YES
334Ian Lynagh <igloo@earth.li>**20080817121000]
335[When making in compiler/, automatically make in ghc/ when we are done
336Ian Lynagh <igloo@earth.li>**20080817114202]
337[Don't do the stage1 re-linking hack if we have GHC >= 6.9
338Ian Lynagh <igloo@earth.li>**20080817114132]
339[Avoid scary "Package contains no library to register" message in installPackage
340Ian Lynagh <igloo@earth.li>**20080817113128
341 We only call register if we actually have a library to be registered
342]
343[Don't use the cc-options from packages when compiling .hc files
344Simon Marlow <marlowsd@gmail.com>**20080821155549
345 Now that we don't include any header files in .hc apart from our own,
346 the cc-options from packages are at best superfluous, so don't pass
347 them.
348 
349 We still pass them to .c compilations, although I've just made changes
350 to Cabal so that cc-options from a .cabal file are not copied into the
351 InstalledPackageInfo.  Most uses of cc-options in Cabal are clearly
352 intended to be local to the package, but they were being propagated
353 everywhere, almost certainly unintentionally.
354 
355 The way is left open for Cabal to allow packages to specify cc-options
356 that get propagated in the future, if we find a use case for this.
357]
358[Improve error message when 'forall' is not a keyword
359simonpj@microsoft.com**20080821123502]
360[move directory after Win32/unix
361Simon Marlow <marlowsd@gmail.com>**20080821100436]
362[Fix Haddock comments in TcGenDeriv
363simonpj@microsoft.com**20080821074924]
364[Fix compacting GC on 64-bit machines
365Simon Marlow <marlowsd@gmail.com>**20080820125446
366 Some old experimental change leaked in with the parallel GC patches,
367 it seems
368]
369[Fix Trac #2456: eliminate duplicate bindings when deriving
370simonpj@microsoft.com**20080820120751
371 
372 Condsider deriving two overlapping Data declarations for the same type
373        deriving instance Data (T A)
374        deriving instance Data (T B)
375 
376 We were getting duplicate bindings for the data-con and tycon auxiliary
377 bindings for T.  This patch fixes the problem by doing these two decls
378 the same way as we do con2tag etc. 
379 
380 (Why might you want such instances; see Trac #2456.)
381 
382]
383[Elaborate for-all message slightly
384simonpj@microsoft.com**20080820120728]
385[set $(BOOTSTRAPPING_PACKAGE_CONF_HC_OPTS) automatically based on $(HC)
386Simon Marlow <marlowsd@gmail.com>**20080819101437
387 This fixes nofib
388]
389[Fix Trac #2518: add hs-boot files as an infelicty
390simonpj@microsoft.com**20080819123016]
391[use System.FilePath's isSearchPathSeparator instead of our own
392Simon Marlow <marlowsd@gmail.com>**20080818113555]
393[FIX #2521: trailing colon in GHC_PACKAGE_PATH
394Simon Marlow <marlowsd@gmail.com>**20080818113345
395 This was broken in the System.FilePath switchover, since filepath's
396 splitSearchPath doesn't do what we want (it ignores empty
397 components on Windows, and treats them as "." on Unix).  So we use our
398 own hand-rolled version, just like GHC.
399]
400[Test for and reject duplicate dependencies (#2330)
401Simon Marlow <marlowsd@gmail.com>**20080818112434]
402[When doing :l, abandon all breakpoints before we unload everything
403Ian Lynagh <igloo@earth.li>**20080816152135
404 I'm not 100% sure if this is the right fix, but it seems sensible and
405 stops break008 segfaulting for me on amd64/Linux.
406]
407[Comment fixes; trac #2468
408Ian Lynagh <igloo@earth.li>**20080816130910]
409[Stop dph/dph-{par,seq} reconfiguring themselves all the time
410Ian Lynagh <igloo@earth.li>**20080814173013]
411[Follow changes in Cabal
412Ian Lynagh <igloo@earth.li>**20080813215844]
413[When distcleaning in utils/, distclean (rather then merely clean) the utils
414Ian Lynagh <igloo@earth.li>**20080815114939]
415[Only clean utils/pwd/pwd when distcleaning, not when cleaning
416Ian Lynagh <igloo@earth.li>**20080812232544
417 Otherwise "make clean" in utils leaves us without a pwd, and lots of bad
418 things happen.
419]
420[In cabal.mk, clean is no longer identical to distclean
421Ian Lynagh <igloo@earth.li>**20080812232446
422 distclean will now also remove files listed in EXTRA_DISTCLEAN
423]
424[Improve -fwarn-orphans documentation (thanks to Tim)
425simonpj@microsoft.com**20080814152648]
426[Fix #2441 (unregister/expose/hide packages in non-first package databases)
427Simon Marlow <marlowsd@gmail.com>**20080814125348]
428[add --no-user-package-conf
429Simon Marlow <marlowsd@gmail.com>**20080814125312]
430[move INPLACE_DATA_DIR into mk/config.mk and share it
431Simon Marlow <marlowsd@gmail.com>**20080813144707]
432[FIX #1963: use Cabal's writeFileAtomic to write the new package.conf
433Simon Marlow <marlowsd@gmail.com>**20080813143738
434 This depends on #2298 also being fixed, which I'll do shortly
435]
436[FIX #2492: ghc-pkg insists on having HOME environment variable set
437Simon Marlow <marlowsd@gmail.com>**20080813143436]
438[FIX #2491 (ghc-pkg unregister should complain about breaking dependent packages)
439Simon Marlow <marlowsd@gmail.com>**20080813142555]
440[put back -fwarn-depcrecations
441Simon Marlow <marlowsd@gmail.com>**20080812141606
442 It was replaced by -fwarn-warnings-deprecations, but I think we want
443 to keep it for backwards compatibility.  I'm not sure we want to
444 deprecate it either...
445]
446[add special globbing file for Windows to ghc-pkg.cabal
447dias@eecs.harvard.edu**20080813105306]
448[Improve error message for deprecated flags (Trac #2513)
449simonpj@microsoft.com**20080813094506]
450[Improve docs for orphan decls (thanks Yitzchak Gale)
451simonpj@microsoft.com**20080813082828]
452[fix cut-and-pasto
453Simon Marlow <marlowsd@gmail.com>**20080811144158]
454[build base3-compat
455Simon Marlow <marlowsd@gmail.com>**20080811143831]
456[Don't warn if 'import Prelude' doesn't import anything
457Simon Marlow <marlowsd@gmail.com>**20080805133702
458 ... even if Prelude doesn't come from the base package (it might come from
459 a old backwards-compatible version of base, for example).
460]
461[Add -XPackageImports, new syntax for package-qualified imports
462Simon Marlow <marlowsd@gmail.com>**20080805133544
463 
464 Now you can say
465   
466   import "network" Network.Socket
467 
468 and get Network.Socket from package "network", even if there are
469 multiple Network.Socket modules in scope from different packages
470 and/or the current package.
471 
472 This is not really intended for general use, it's mainly so that we
473 can build backwards-compatible versions of packages, where we need to
474 be able to do
475 
476 module GHC.Base (module New.GHC.Base) where
477 import "base" GHC.Base as New.GHC.Base
478]
479[Suppress uniques when (and only when) we have -dsuppress-uniques
480simonpj@microsoft.com**20080812144028]
481[get exception names from Control.Exception.Base instead of Control.Exception
482Ross Paterson <ross@soi.city.ac.uk>**20080812122048]
483[Fix Trac #2490: sections should be parenthesised
484simonpj@microsoft.com**20080812082332
485 
486 When I added bang patterns I had to slightly generalise where the
487 parser would recognise sections.  See Note [Parsing sections] in
488 parser.y.pp.
489 
490 I forgot to check that ordinary H98 sections obey the original
491 rules.  This patch adds the check.
492 
493]
494[Refactoring: define TcRnMonad.failWith and use it in the renamer
495simonpj@microsoft.com**20080812082255]
496[Layout only
497simonpj@microsoft.com**20080812081822]
498[Use dph-{seq|par} instead of dph_{seq|par}
499Roman Leshchinskiy <rl@cse.unsw.edu.au>**20080812021939]
500[Simplify how we build dph
501Ian Lynagh <igloo@earth.li>**20080809203803]
502[Generalise libraries/Makefile.local
503Ian Lynagh <igloo@earth.li>**20080809161838
504 It's now possible to build libraries in, e.g.,
505     libraries/foo/bar
506 rather than just
507     libraries/foo
508]
509[The dph packages still contain warnings
510Ian Lynagh <igloo@earth.li>**20080809161815]
511[Print tidy rules in user style, to avoid gratuitous uniques
512simonpj@microsoft.com**20080811144257
513 
514 The uniques that come out in dumpStyle make it harder to compare
515 output in the testsuite.  And the rules are tidied, so uniques
516 are not necessary.  If you want the uniques, use -dppr-debug.
517 
518]
519[Export Depth (needed for mkUserStyle); collapse identical PrintUnqualified, QueryQualifies
520simonpj@microsoft.com**20080811144208]
521[When suppressing uniques, don't print the separating underscore
522simonpj@microsoft.com**20080811144118]
523[Make -dsuppress-uniques apply regardless of -ppr-debug
524simonpj@microsoft.com**20080811144032]
525[Fix Trac #2367: selectors for GADTs
526simonpj@microsoft.com**20080811123526
527 
528 The generation of record selectors for GADTs and the like was
529 pretty screwed up.  This patch fixes it.
530 
531 Note that Unify.refineType is now used only in the generation of
532 record seletctors -- but it really does seem to be needed there.
533 
534 Thanks to Max for finding this bug.
535 
536]
537[Mostly fix Trac #2431: make empty case acceptable to (most of) GHC
538simonpj@microsoft.com**20080811123158
539 
540 See the comments with Trac #2431.  This patch makes an empty HsCase
541 acceptable to the renamer onwards.  If you want to accept empty case
542 in Haskell source there's a little more to do: the ticket lists the
543 remaining tasks.
544 
545 
546]
547[Fix Trac #2412: type synonyms and hs-boot recursion
548simonpj@microsoft.com**20080811122523
549 
550 Max Bolingbroke found this awkward bug, which relates to the way in
551 which hs-boot files are handled.
552 
553    --> HEADS UP: interface file format change: recompile everything!
554 
555 When we import a type synonym, we want to *refrain* from looking at its
556 RHS until we've "tied the knot" in the module being compiled.  (Reason:
557 the type synonym might ultimately loop back to the module being compiled.)
558 To achieve this goal we need to know the *kind* of the synonym without
559 looking at its RHS.  And to do that we need its kind recorded in the interface
560 file.
561 
562 I slightly refactored the way that the IfaceSyn data constructor
563 fields work, eliminating the previous tricky re-use of the same field
564 as either a type or a kind.
565 
566 See Note [Synonym kind loop] in TcIface
567 
568 
569]
570[Fix Trac #2506: infix assert
571simonpj@microsoft.com**20080811105011]
572[Use do-notation
573simonpj@microsoft.com**20080811104949]
574[Fix Trac #2494: tcSimplifyRuleLhs
575simonpj@microsoft.com**20080811104438
576 
577 tcSimplifyRuleLhs is a cut-down constraint simplifier, intended for
578 use in RULE left-hand-sides.  But it was written before implication
579 constraints, and the exmaple of this bug report shows that when higher
580 rank types are involved we need to be a bit cleverer.
581 
582 The whole business of simplifying constraints on rule LHSs is a bit
583 of a hack; but for a good reason.  See the comments with tcSimplifyRuleLhs.
584 This patch at least cures the crash.
585 
586 
587]
588[Make comparison on equalities work right (ie look at the types)
589simonpj@microsoft.com**20080811104011
590 
591 This patch makes
592        (s1~t1) == (s2~t2) iff s1==s2, t1==t2
593 
594 The comparison and equality instances for Insts are really only
595 used in making the AvailEnv in TcSimplify, which equalities are
596 never put into, which is why we've gotten away with this for so long.
597 
598 But I'm now using 'nub' in the error messages for equalities,
599 and in any case it makes sense to have the right equality!
600 
601]
602[Minor refactoring; no functionality change
603simonpj@microsoft.com**20080811103821]
604[Fix Trac #2486: restore the specialiser to a working state
605simonpj@microsoft.com**20080809163123
606 
607 In improving the specialiser's data reprsentaion, thus
608   Mon Apr 28 16:57:11 GMT Daylight Time 2008  simonpj@microsoft.com
609   * Fix Trac #1969: perfomance bug in the specialiser
610 I got the sense of a pair of filter functions back to front.
611 As a result, almost all specialisation opportunities were being
612 filtered out, and no specialisations were being generated.
613 
614 Fortunately, dolio notice and reported Trac #2486.  The fix is
615 simple: put the filters the right way roud.
616 
617 
618]
619[Remove a pointless use of $(HERE_ABS)/
620Ian Lynagh <igloo@earth.li>**20080810173237]
621[Don't give the -d flag to rm; it goes wrong on Windows
622Ian Lynagh <igloo@earth.li>**20080810173053]
623[Add mk/bindist.mk
624Ian Lynagh <igloo@earth.li>**20080810150737]
625[Remove references to the files that the RTS no longer knows about
626Ian Lynagh <igloo@earth.li>**20080810144740
627 Spotted by Simon
628]
629[We don't need to generate driver/package.conf any more; spotted by Simon
630Ian Lynagh <igloo@earth.li>**20080810144259]
631[Remove inplace-datadir when cleaning
632Ian Lynagh <igloo@earth.li>**20080810144118]
633[Remove driver/ghc; The Cabal package in ghc/ now makes a wrapper itself
634Ian Lynagh <igloo@earth.li>**20080810143655]
635[Make "make clean" in ghc/ only clean the stage we want to clean
636Ian Lynagh <igloo@earth.li>**20080810142409]
637[Make "make clean" in compiler/ only clean the stage we want to clean
638Ian Lynagh <igloo@earth.li>**20080810142026
639 If you "make clean" in the root then we still clean all stages
640]
641[Also clean pwd.hi and pwd.o
642Ian Lynagh <igloo@earth.li>**20080810140731]
643[bindist fixes
644Ian Lynagh <igloo@earth.li>**20080810133925]
645[bindists are now some way towards working
646Ian Lynagh <igloo@earth.li>**20080810005016]
647[Use Cabal to build pwd
648Ian Lynagh <igloo@earth.li>**20080810002807]
649[Move allM to MonadUtils
650Max Bolingbroke <batterseapower@hotmail.com>**20080807224853]
651[Remove redundant fromIntegral calls
652Max Bolingbroke <batterseapower@hotmail.com>**20080807224333]
653[Remove CoreSyn SOURCE imports
654Max Bolingbroke <batterseapower@hotmail.com>**20080807223718]
655[Fixed performance bug in ext-core preprocessor
656Tim Chevalier <chevalier@alum.wellesley.edu>**20080809002051
657 
658 The Core preprocessor was rebuilding the type and data constructor environments every time it called the typechecker, which was horribly inefficient. Fixed.
659]
660[Fix darcs-all get
661Ian Lynagh <igloo@earth.li>**20080807203258]
662[Fix errors with haddock 0.8
663Ian Lynagh <igloo@earth.li>**20080807131506]
664[Document Name and expand it's API
665Max Bolingbroke <batterseapower@hotmail.com>**20080807122901]
666[Follow OccName changes in Convert
667Max Bolingbroke <batterseapower@hotmail.com>**20080731120741]
668[Use DynFlags.getMainFun in TcRnDriver
669Max Bolingbroke <batterseapower@hotmail.com>**20080731055210]
670[Fix Vectorise after introduction of MonadThings
671Max Bolingbroke <batterseapower@hotmail.com>**20080731054813]
672[Follow introduction of MkCore in VectUtils
673Max Bolingbroke <batterseapower@hotmail.com>**20080731054722]
674[Fix VectMonad after introduction of MonadThings
675Max Bolingbroke <batterseapower@hotmail.com>**20080731054623]
676[Minor refactorings in TcEnv
677Max Bolingbroke <batterseapower@hotmail.com>**20080731054437]
678[Handle introduction of MkCore in DsMonad and expand API
679Max Bolingbroke <batterseapower@hotmail.com>**20080731054239]
680[Add MkCore, holding general Core construction facilities
681Max Bolingbroke <batterseapower@hotmail.com>**20080731054128]
682[Document HscTypes, refactor it somewhat, remove unused type and add MonadThings
683Max Bolingbroke <batterseapower@hotmail.com>**20080731054042]
684[Use new RdrName export in RdrHsSyn
685Max Bolingbroke <batterseapower@hotmail.com>**20080731053731]
686[Document CoreUtils
687Max Bolingbroke <batterseapower@hotmail.com>**20080731053542]
688[Document Packages and a minor refactoring
689Max Bolingbroke <batterseapower@hotmail.com>**20080731053054]
690[Document CoreSyn and expand its API
691Max Bolingbroke <batterseapower@hotmail.com>**20080731052941]
692[Document IdInfo
693Max Bolingbroke <batterseapower@hotmail.com>**20080731052819]
694[Document DynFlags and expand API
695Max Bolingbroke <batterseapower@hotmail.com>**20080731052653]
696[Document OccName and expand it's API
697Max Bolingbroke <batterseapower@hotmail.com>**20080731052238]
698[Follow Digraph changes in VectType
699Max Bolingbroke <batterseapower@hotmail.com>**20080731012356]
700[Follow OccName changes in VectBuiltIn
701Max Bolingbroke <batterseapower@hotmail.com>**20080731012355]
702[Document Util
703Max Bolingbroke <batterseapower@hotmail.com>**20080807115505]
704[Document UniqSet
705Max Bolingbroke <batterseapower@hotmail.com>**20080731012355]
706[Document UniqFM and add listToUFM_C
707Max Bolingbroke <batterseapower@hotmail.com>**20080731012355]
708[Follow FastString changes in Pretty
709Max Bolingbroke <batterseapower@hotmail.com>**20080731012354]
710[Document Outputable
711Max Bolingbroke <batterseapower@hotmail.com>**20080807114806]
712[Document FiniteMap
713Max Bolingbroke <batterseapower@hotmail.com>**20080731012353]
714[Document FastString and rename strLength to lengthLS
715Max Bolingbroke <batterseapower@hotmail.com>**20080731012353]
716[Add more functionality to Digraph and refactor it's interface somewhat, including adding a Graph ADT
717Max Bolingbroke <batterseapower@hotmail.com>**20080731012353]
718[Document TypeRep and follow OccName change
719Max Bolingbroke <batterseapower@hotmail.com>**20080731012352]
720[Document Type
721Max Bolingbroke <batterseapower@hotmail.com>**20080731012352]
722[Document TyCon
723Max Bolingbroke <batterseapower@hotmail.com>**20080731012352]
724[Document Coercion
725Max Bolingbroke <batterseapower@hotmail.com>**20080731012351]
726[Follow Digraph changes in TcTyDecls
727Max Bolingbroke <batterseapower@hotmail.com>**20080731012351]
728[Make IOEnv a MonadUnique
729Max Bolingbroke <batterseapower@hotmail.com>**20080731012351]
730[Follow TcEnv API addition in TcForeign
731Max Bolingbroke <batterseapower@hotmail.com>**20080731012350]
732[Follow Digraph changes in TcBinds
733Max Bolingbroke <batterseapower@hotmail.com>**20080731012350]
734[Follow OccName and MkCore changes in Inst
735Max Bolingbroke <batterseapower@hotmail.com>**20080731012349]
736[Document Rules
737Max Bolingbroke <batterseapower@hotmail.com>**20080731012349]
738[Follow Literal change in Simplify
739Max Bolingbroke <batterseapower@hotmail.com>**20080731012349]
740[Follow Digraph changes in OccurAnal
741Max Bolingbroke <batterseapower@hotmail.com>**20080731012348]
742[Follow Digraph changes in RnBinds and small refactoring
743Max Bolingbroke <batterseapower@hotmail.com>**20080731012348]
744[Document TysWiredIn and follow OccName changes
745Max Bolingbroke <batterseapower@hotmail.com>**20080731012347]
746[Follow OccName changes in TysPrim
747Max Bolingbroke <batterseapower@hotmail.com>**20080731012347]
748[Follow OccName changes in ParserCore
749Max Bolingbroke <batterseapower@hotmail.com>**20080731012347]
750[Follow Digraph changes in RegLiveness
751Max Bolingbroke <batterseapower@hotmail.com>**20080731012346]
752[Follow Digraph changes in RegAllocLinear.hs
753Max Bolingbroke <batterseapower@hotmail.com>**20080731012346]
754[Follow Digraph changes in AsmCodeGen
755Max Bolingbroke <batterseapower@hotmail.com>**20080731012346]
756[Document InteractiveEval and follow OccName change
757Max Bolingbroke <batterseapower@hotmail.com>**20080731012345]
758[Follow Digraph changes in GHC; patch from Max Bolingbroke
759Ian Lynagh <igloo@earth.li>**20080807112510]
760[Expand ErrUtils api with another plain variant
761Max Bolingbroke <batterseapower@hotmail.com>**20080731012344]
762[Follow OccName changes and minor refactorings in TcIface
763Max Bolingbroke <batterseapower@hotmail.com>**20080731012343]
764[Follow Digraph changes in MkIface
765Max Bolingbroke <batterseapower@hotmail.com>**20080731012343]
766[Minor refactoring in RtClosureInspect
767Max Bolingbroke <batterseapower@hotmail.com>**20080731012342]
768[Handle optSrcSpanFileName in InteractiveUI
769Max Bolingbroke <batterseapower@hotmail.com>**20080731012342]
770[Handle introduction of MkCore in MatchLit
771Max Bolingbroke <batterseapower@hotmail.com>**20080731012342]
772[Handle introduction of MkCore in MatchCon
773Max Bolingbroke <batterseapower@hotmail.com>**20080731012342]
774[Handle introduction of MkCore in Match
775Max Bolingbroke <batterseapower@hotmail.com>**20080731012341]
776[Handle introduction of MkCore in DsUtils
777Max Bolingbroke <batterseapower@hotmail.com>**20080731012341]
778[Handle introduction of MkCore in DsMeta
779Max Bolingbroke <batterseapower@hotmail.com>**20080731012340]
780[Handle introduction of MkCore in DsListComp
781Max Bolingbroke <batterseapower@hotmail.com>**20080731012340]
782[Handle introduction of MkCore in DsForeign
783Max Bolingbroke <batterseapower@hotmail.com>**20080731012340]
784[Handle introduction of MkCore in DsExpr
785Max Bolingbroke <batterseapower@hotmail.com>**20080731012340]
786[Handle introduction of MkCore in DsBinds
787Max Bolingbroke <batterseapower@hotmail.com>**20080731012339]
788[Handle introduction of MkCore in DsArrows
789Max Bolingbroke <batterseapower@hotmail.com>**20080731012339]
790[Handle optSrcSpanFileName rename in Coverage
791Max Bolingbroke <batterseapower@hotmail.com>**20080731012339]
792[Document CgUtils, follow Digraph changes
793Max Bolingbroke <batterseapower@hotmail.com>**20080731012337]
794[Document UniqSupply
795Max Bolingbroke <batterseapower@hotmail.com>**20080731012334]
796[Document SrcLoc
797Max Bolingbroke <batterseapower@hotmail.com>**20080731012334]
798[Document RdrName and expand its API
799Max Bolingbroke <batterseapower@hotmail.com>**20080731012334]
800[Make Module Uniquable
801Max Bolingbroke <batterseapower@hotmail.com>**20080731012332]
802[Tweak comments in MkId and follow mkStringLit renaming
803Max Bolingbroke <batterseapower@hotmail.com>**20080731012332]
804[Document Literal, expand it's API and rename mkStringLit to mkMachString
805Max Bolingbroke <batterseapower@hotmail.com>**20080731012331]
806[Document Id
807Max Bolingbroke <batterseapower@hotmail.com>**20080731012331]
808[Split the Id related functions out from Var into Id, document Var and some of Id
809Max Bolingbroke <batterseapower@hotmail.com>**20080731012330]
810[Eq and Ord have moved into GHC.Classes
811Ian Lynagh <igloo@earth.li>**20080806161815]
812[Move Int, Float and Double into ghc-prim:GHC.Types
813Ian Lynagh <igloo@earth.li>**20080806224359]
814[Move Int, Float and Double into ghc-prim:GHC.Types
815Ian Lynagh <igloo@earth.li>**20080806191458]
816[Prepare GHC for building with Git
817Max Bolingbroke <batterseapower@hotmail.com>**20080806210207
818 * New packages format lets you select source control system in use
819 * Packages file now includes root repo explicitly
820 * Scripts darcs-all and push-all updated for the new packages format only
821 * New sync-all script, intended for use after Git changeover and for buildbots right now
822 * Had to remove libraries/bootstrapping from tree since Git cannot track empty directories without a hack
823 * Determine checkout date with Git using Darcs fallback in aclocal.m4
824]
825[Add -fno-implicit-import-qualified (#2452)
826Simon Marlow <marlowsd@gmail.com>**20080805151730
827 The flag is off by default, but GHCi turns it on (in Main.hs).  For
828 GHCi it can be overriden on the command-line or using :set.
829]
830[in stage1, always rebuild Main.hs
831Simon Marlow <marlowsd@gmail.com>**20080805133823
832 GHC 6.9+ can properly handle cross-package recompilation checking, but
833 older versions of GHC couldn't, so in stage1 we always rebuild
834 Main.hs.  In other words, 'make' in ghc/ should do the right thing now.
835]
836[Ooops; lack of mk/confi.mk doesn't mean validate shouldn't run configure!
837Ian Lynagh <igloo@earth.li>**20080806113102]
838[C# has moved to ghc-prim:GHC.Types
839Ian Lynagh <igloo@earth.li>**20080805220155]
840[Move the Char datatype into ghc-prim
841Ian Lynagh <igloo@earth.li>**20080805195643]
842[Move the [] definion from base to ghc-prim
843Ian Lynagh <igloo@earth.li>**20080805180624]
844[Don't boot/configure if we are validating --no-clean
845Ian Lynagh <igloo@earth.li>**20080804214657]
846[Remove the cgi package from extralibs
847Ian Lynagh <igloo@earth.li>**20080805125551
848 It has some sort of Error Monad using the old Exception type.
849 I'm not familiar with it enough to know what the right thing to do
850 for it with extensible exceptions is.
851]
852[Follow the tuple datatype movements
853Ian Lynagh <igloo@earth.li>**20080804155402]
854[Don't make a "windows" flag in the Cabal file - os(windows) already exists!
855Ian Lynagh <igloo@earth.li>**20080804153430
856 Pointed out by Duncan Coutts
857]
858[Fix Trac #2449
859simonpj@microsoft.com**20080805105402
860 
861 Deriving isn't allowed in hs-boot files (says the user manual)
862 This patch checks properly instead of crashing!
863 
864]
865[2nd try: remove lochash, it isn't needed (now)
866Simon Marlow <marlowsd@gmail.com>**20080804125944]
867[FIX BUILD on Windows
868Simon Marlow <marlowsd@gmail.com>**20080804105740]
869[Fix Trac #2467: decent warnings for orphan instances
870simonpj@microsoft.com**20080804162129
871 
872 This patch makes
873   * Orphan instances and rules obey -Werror
874   * They look nicer when printed
875 
876]
877[Fix the bug part of Trac #1930
878simonpj@microsoft.com**20080804161039]
879[Fix Trac #2433 (deriving Typeable)
880simonpj@microsoft.com**20080804141503]
881[Fix Trac #2478
882simonpj@microsoft.com**20080801122223
883 
884 A minor glitch that shows up only when a data constructor has *both* a
885 "stupid theta" in the data type decl, *and* an existential type variable.
886 
887]
888[Improve docs for GADTs
889simonpj@microsoft.com**20080729145313]
890[Document -dsuppress-uniques
891simonpj@microsoft.com**20080729145247]
892[UNDO:  FIX #2375: remove oc->lochash completely, it apparently isn't used
893Simon Marlow <marlowsd@gmail.com>**20080804111801]
894[haddock fixes
895Ian Lynagh <igloo@earth.li>**20080803180330]
896[Follow the move of assertError from Control.Exception to GHC.IOBase
897Ian Lynagh <igloo@earth.li>**20080803141146]
898[Document PackageConfig
899Max Bolingbroke <batterseapower@hotmail.com>**20080731012345]
900[Document CoreSubst
901Max Bolingbroke <batterseapower@hotmail.com>**20080731012338]
902[Document CoreFVs
903Max Bolingbroke <batterseapower@hotmail.com>**20080731012337]
904[Document CmmZipUtil
905Max Bolingbroke <batterseapower@hotmail.com>**20080731012335]
906[Document VarSet
907Max Bolingbroke <batterseapower@hotmail.com>**20080731012335]
908[Document VarEnv
909Max Bolingbroke <batterseapower@hotmail.com>**20080731012335]
910[Document Unique
911Max Bolingbroke <batterseapower@hotmail.com>**20080731012334]
912[Document LazyUniqFM
913Max Bolingbroke <batterseapower@hotmail.com>**20080731012354]
914[Document FastTypes
915Max Bolingbroke <batterseapower@hotmail.com>**20080731012353]
916[Add some type signatures to RnNames
917Max Bolingbroke <batterseapower@hotmail.com>**20080731012348]
918[Comment only in IfaceENv
919Max Bolingbroke <batterseapower@hotmail.com>**20080731012343]
920[Document ZipCfg
921Max Bolingbroke <batterseapower@hotmail.com>**20080731012336]
922[Document MachOp
923Max Bolingbroke <batterseapower@hotmail.com>**20080731012336]
924[Document Dataflow
925Max Bolingbroke <batterseapower@hotmail.com>**20080731012336]
926[Document DFMonad
927Max Bolingbroke <batterseapower@hotmail.com>**20080731012336]
928[Document NameSet
929Max Bolingbroke <batterseapower@hotmail.com>**20080731012333]
930[Document NameEnv
931Max Bolingbroke <batterseapower@hotmail.com>**20080731012333]
932[Document Module
933Max Bolingbroke <batterseapower@hotmail.com>**20080731012332]
934[Document DataCon
935Max Bolingbroke <batterseapower@hotmail.com>**20080731012316]
936[Document BasicTypes
937Max Bolingbroke <batterseapower@hotmail.com>**20080731012306]
938[Rename maybeTyConSingleCon to tyConSingleDataCon_maybe
939Max Bolingbroke <batterseapower@hotmail.com>**20080731010537]
940[Ignore git files
941Max Bolingbroke <batterseapower@hotmail.com>**20080731010509]
942[Fix ifBuildable
943Ian Lynagh <igloo@earth.li>**20080801141731]
944[Update assertErrorName; assertError has moved to Control.Exception
945Ian Lynagh <igloo@earth.li>**20080801011028]
946[Fix the catching of "IOEnv failure" with extensible extensions
947Ian Lynagh <igloo@earth.li>**20080731194252]
948[Follow changes in the base library
949Ian Lynagh <igloo@earth.li>**20080731173354
950 TopHandler now uses the new extensible exceptions module, so we
951 need to interact with it using the new types.
952]
953[Add the Exception module to ghc.cabal
954Ian Lynagh <igloo@earth.li>**20080730213419]
955[Fix building with extensible exceptions
956Ian Lynagh <igloo@earth.li>**20080730194508]
957[Mark the ghc package as not exposed
958Ian Lynagh <igloo@earth.li>**20080730172124]
959[Follow extensible exception changes
960Ian Lynagh <igloo@earth.li>**20080730120134]
961[When raising NonTermination with the RTS, build the right value
962Ian Lynagh <igloo@earth.li>**20080621144528
963 We now use a nonTermination value in the base library to take take of
964 constructing the SomeException value, with the dictionaries etc, for us.
965 We'll probably need to do the same for some other exceptions too
966]
967[Fix the way we pass GMP_INCLUDE_DIRS to hsc2hs; spotted by Andres Loh
968Ian Lynagh <igloo@earth.li>**20080730114713
969 We were still building the flags in Haskell list syntax, but we now pass
970 the arguments directly rather than constructing a Haskell program with
971 them.
972]
973[oops, fix a small pessimisation made in previous refactoring
974Simon Marlow <marlowsd@gmail.com>**20080730105203]
975[FIX #2388: check that the operand fits before using the 'test' opcode
976Simon Marlow <marlowsd@gmail.com>**20080730105231]
977[workaround #2277: turn off the RTS timer when calling into editline
978Simon Marlow <marlowsd@gmail.com>**20080730135918]
979[FIX #2375: remove oc->lochash completely, it apparently isn't used
980Simon Marlow <marlowsd@gmail.com>**20080730101252]
981[Fix a typo in powerpc/Linux-only code; spotted by Jeroen Pulles
982Ian Lynagh <igloo@earth.li>**20080729214007]
983[Add the runghc wrapper script
984Ian Lynagh <igloo@earth.li>**20080729211852]
985[Make cabal-bin not do any building, even of Setup.hs, when it is asked to clean
986Ian Lynagh <igloo@earth.li>**20080729202410]
987[Update the test in Makefile that we have all the boot libs
988Ian Lynagh <igloo@earth.li>**20080729201640]
989[Update boot's test that we have all of the bootlibs
990Ian Lynagh <igloo@earth.li>**20080729201032]
991[Make the push-all script complain about bad lines
992Ian Lynagh <igloo@earth.li>**20080729200613]
993[Add some comments to packages/darcs-all
994Ian Lynagh <igloo@earth.li>**20080729151934]
995[Remove ndp from libraries/Makefile. We now use dph instead.
996Ian Lynagh <igloo@earth.li>**20080729141917]
997[Add dph to ./packages and darcs-all
998Ian Lynagh <igloo@earth.li>**20080729141850]
999[Remove cabal-install from ./packages; we've decided not to build it
1000Ian Lynagh <igloo@earth.li>**20080729141824]
1001[FIX #2327: a fault in the thunk-selector machinery (again)
1002Simon Marlow <marlowsd@gmail.com>**20080729150518
1003 This program contains an expression of the form
1004 
1005    let x = snd (_, snd (_, snd (_, x)))
1006 
1007 (probably not explicitly, but that's what appears in the heap at
1008 runtime).  Obviously the program should deadlock if it ever enters
1009 this thing, but apparently the test program in #2327 never does.
1010 
1011 The GC tries to evaluate the snd closures, and gets confused due to
1012 the loop.  In particular the earlier fix for #1038 was to blame.
1013]
1014[FIX #2332: avoid overflow on 64-bit machines in the memory allocator
1015Simon Marlow <marlowsd@gmail.com>**20080729150459]
1016[Change the calling conventions for unboxed tuples slightly
1017Simon Marlow <marlowsd@gmail.com>**20080728155621
1018 When returning an unboxed tuple with a single non-void component, we
1019 now use the same calling convention as for returning a value of the
1020 same type as that component.  This means that the return convention
1021 for IO now doesn't vary depending on the platform, which make some
1022 parts of the RTS simpler, and fixes a problem I was having with making
1023 the FFI work in unregisterised GHCi (the byte-code compiler makes
1024 some assumptions about calling conventions to keep things simple).
1025]
1026[don't strip the inplace GHC executables (for debugging)
1027Simon Marlow <marlowsd@gmail.com>**20080728134647]
1028[Complete the changes for #1205
1029Simon Marlow <marlowsd@gmail.com>**20080728105141
1030 Now ":load M" always searches for a module called "M", rather than
1031 using a file called "M.hs" if that exists.  To get the file semantics
1032 (i.e. not loading "M.o"), use ":load M.hs".
1033]
1034[update the comments about how we find $topdir
1035Simon Marlow <marlowsd@gmail.com>**20080725151406]
1036[understand absolute pathnames on Windows too
1037Simon Marlow <marlowsd@gmail.com>**20080728102243]
1038[change where we put gcc-lib/ld.exe to keep Cabal happy
1039Simon Marlow <marlowsd@gmail.com>**20080728100852]
1040[move an inline function to keep older versions of gcc happy
1041Simon Marlow <marlowsd@gmail.com>**20080725144708
1042 no idea why this only just showed up...
1043]
1044[try to fix the way we find $topdir
1045Simon Marlow <marlowsd@gmail.com>**20080725142828]
1046[for the installed versions, don't use dynamic-linking wrappers
1047Simon Marlow <marlowsd@gmail.com>**20080725134551]
1048[don't steal %ebx for the GC on x86: it's also used by PIC
1049Simon Marlow <marlowsd@gmail.com>**20080725122921]
1050[SRT labels don't need to be globally visible
1051Simon Marlow <marlowsd@gmail.com>**20080725080901
1052 Saves space in the symbol table and speeds up linking
1053]
1054[Don't prematurely link shared libraries against the RTS package
1055Simon Marlow <marlowsd@gmail.com>**20080724155001
1056 We want to be able to pick the RTS flavour (e.g. -threaded) when we
1057 link the final program.
1058]
1059[add --enable-shared to configure, and $(BuildSharedLibs) to the build system
1060Simon Marlow <marlowsd@gmail.com>**20080724154925]
1061[use RTLD_LAZY instead of RTLD_NOW
1062Simon Marlow <marlowsd@gmail.com>**20080724152727
1063 RTLD_NOW apparently causes some problems, according to previous
1064 mailing-list discussion
1065 
1066  http://www.haskell.org/pipermail/cvs-ghc/2007-September/038570.html
1067]
1068[debug output tweak
1069Simon Marlow <marlowsd@gmail.com>**20080724152636]
1070[small cleanup
1071Simon Marlow <marlowsd@gmail.com>**20080724151614]
1072[Fix building runghc on Windows
1073Ian Lynagh <igloo@earth.li>**20080724182831]
1074[Follow darcs-all changes in push-all
1075Ian Lynagh <igloo@earth.li>**20080724164153]
1076[Rejig how darcs-all works
1077Ian Lynagh <igloo@earth.li>**20080724164142
1078 It's now easier to add new repos anywhere in the source tree
1079]
1080[Remove the OpenGL family of libraries from extralibs
1081Ian Lynagh <igloo@earth.li>**20080724102736]
1082[compiler/package.conf.in is no longer used, so remove it
1083Ian Lynagh <igloo@earth.li>**20080724101610]
1084[allow EXTRA_HC_OPTS to be used from the command-line
1085Simon Marlow <marlowsd@gmail.com>**20080724081728]
1086[put the inplace GHC in stageN-inplace/ghc instead of stageN-inplace/bin/ghc
1087Simon Marlow <marlowsd@gmail.com>**20080724080951
1088 just saves a bit of typing
1089]
1090[add a "rebuild" target for convenience
1091Simon Marlow <marlowsd@gmail.com>**20080723143201]
1092[set PAPI_LIB_DIR="" when we don't have PAPI (clean up package.conf)
1093Simon Marlow <marlowsd@gmail.com>**20080722141327]
1094[remove -fvia-C that I apparrently accidentally added recently
1095Simon Marlow <marlowsd@gmail.com>**20080722141255]
1096[Fix a build error on powerpc/Linux; spotted by Jeroen Pulles
1097Ian Lynagh <igloo@earth.li>**20080723191948]
1098[If the extension is not .lhs, runghc now treats it as .hs; fixes trac #1232
1099Ian Lynagh <igloo@earth.li>**20080723182156]
1100[runghc now uses the compiler that it comes with; fixes trac #1281
1101Ian Lynagh <igloo@earth.li>**20080723181115
1102 rather than the first one that it finds on the PATH
1103]
1104[Use the upstream hsc2hs repo
1105Ian Lynagh <igloo@earth.li>**20080723155021]
1106[Remove some redundancy in darcs-all
1107Ian Lynagh <igloo@earth.li>**20080723143804]
1108[Tell Cabal where gcc is
1109Ian Lynagh <igloo@earth.li>**20080723001202]
1110[Undo fix for #2185: sparks really should be treated as roots
1111Simon Marlow <marlowsd@gmail.com>**20080723125205
1112 Unless sparks are roots, strategies don't work at all: all the sparks
1113 get GC'd.  We need to think about this some more.
1114]
1115[fix bug in sparkPoolSize (affects debug output only)
1116Simon Marlow <marlowsd@gmail.com>**20080723104322]
1117[debug message tweaks
1118Simon Marlow <marlowsd@gmail.com>**20080723090050]
1119[refactoring/tidyup: (not.is64BitInteger) -> is32BitInteger
1120Simon Marlow <marlowsd@gmail.com>**20080722092113]
1121[non-threaded RTS: don't assume deadlock if there are signal handlers to run
1122Simon Marlow <marlowsd@gmail.com>**20080715130316]
1123[update the text about header files and -#include
1124Simon Marlow <marlowsd@gmail.com>**20080715101119]
1125[add NetBSD to some of the #ifdefs (patch partly from 6.8 branch)
1126Simon Marlow <marlowsd@gmail.com>**20080714145040]
1127[Warn about unrecognised pragmas; these often mean we've typoed
1128Ian Lynagh <igloo@earth.li>**20080722235550]
1129[Sync hsc2hs's Main.hs with the Cabal repo
1130Ian Lynagh <igloo@earth.li>**20080722203646]
1131[We need to clean the utils on "distclean", as well as "clean"
1132Ian Lynagh <igloo@earth.li>**20080722170754]
1133[Clean stage 3
1134Ian Lynagh <igloo@earth.li>**20080722170542]
1135[Add replacements for the -optdep flags, and deprecate the old ones
1136Ian Lynagh <igloo@earth.li>**20080722163308]
1137[Fix the stage3 build
1138Ian Lynagh <igloo@earth.li>**20080722125743]
1139[Fixes for haddock 0.8
1140Ian Lynagh <igloo@earth.li>**20080721095256]
1141[haddock the stage2 compiler if HADDOCK_DOCS is YES
1142Ian Lynagh <igloo@earth.li>**20080720220622]
1143[First step for getting rid of the old -optdep flags
1144Ian Lynagh <igloo@earth.li>**20080720203239
1145 They are now handled by the main flag parser, rather than having their
1146 own praser that runs later.
1147 
1148 As an added bonus, 5 global variables are also gone.
1149]
1150[Fix Haddock errors.
1151Thomas Schilling <nominolo@googlemail.com>**20080720173151]
1152[Fix Haddock errors.
1153Thomas Schilling <nominolo@googlemail.com>**20080720173117]
1154[Fix Haddock errors.
1155Thomas Schilling <nominolo@googlemail.com>**20080720173105]
1156[Fix Haddock errors.
1157Thomas Schilling <nominolo@googlemail.com>**20080720173017]
1158[Fix Haddock errors.
1159Thomas Schilling <nominolo@googlemail.com>**20080720172614]
1160[Fix Haddock errors.
1161Thomas Schilling <nominolo@googlemail.com>**20080720172401]
1162[Fix Haddock errors.
1163Thomas Schilling <nominolo@googlemail.com>**20080720172242]
1164[Fix Haddock errors.
1165Thomas Schilling <nominolo@googlemail.com>**20080720172222]
1166[Fix Haddock errors.
1167Thomas Schilling <nominolo@googlemail.com>**20080720172139]
1168[Fix Haddock errors.
1169Thomas Schilling <nominolo@googlemail.com>**20080720172114]
1170[Fix Haddock errors.
1171Thomas Schilling <nominolo@googlemail.com>**20080720172054]
1172[Fix Haddock errors.
1173Thomas Schilling <nominolo@googlemail.com>**20080720172010]
1174[Fix Haddock errors.
1175Thomas Schilling <nominolo@googlemail.com>**20080720171723]
1176[Fix Haddock errors.
1177Thomas Schilling <nominolo@googlemail.com>**20080720171554]
1178[Fix Haddock errors.
1179Thomas Schilling <nominolo@googlemail.com>**20080720171529]
1180[Fix Haddock errors.
1181Thomas Schilling <nominolo@googlemail.com>**20080720171424]
1182[Fix Haddock errors.
1183Thomas Schilling <nominolo@googlemail.com>**20080720171113]
1184[Fix Haddock errors.
1185Thomas Schilling <nominolo@googlemail.com>**20080720170708]
1186[Fix Haddock errors.
1187Thomas Schilling <nominolo@googlemail.com>**20080720170601]
1188[Fix Haddock errors.
1189Thomas Schilling <nominolo@googlemail.com>**20080720170421]
1190[Fix Haddock errors.
1191Thomas Schilling <nominolo@googlemail.com>**20080720165845]
1192[Fix Haddock errors.
1193Thomas Schilling <nominolo@googlemail.com>**20080720165637]
1194[Fix Haddock errors.
1195Thomas Schilling <nominolo@googlemail.com>**20080720164133]
1196[Properly comment out unused pragmas
1197Ian Lynagh <igloo@earth.li>**20080720135604
1198 We now say
1199     -- {-# SPECIALIZE ...
1200 rather than
1201     {-# -- SPECIALIZE ...
1202]
1203[Add a WARNING pragma
1204Ian Lynagh <igloo@earth.li>**20080720120918]
1205[Put a #! line in ghc-pkg's shell wrapper
1206Ian Lynagh <igloo@earth.li>**20080719112544]
1207[Fix ghc-pkg inplace on Windows
1208Ian Lynagh <igloo@earth.li>**20080719002613]
1209[Some "install" and "clean" fixes
1210Ian Lynagh <igloo@earth.li>**20080718223656]
1211[Change how inplace detection works, so that it also works on Windows
1212Ian Lynagh <igloo@earth.li>**20080718210836]
1213[More dependency wibbling
1214Ian Lynagh <igloo@earth.li>**20080718193454]
1215[Build system tweaks
1216Ian Lynagh <igloo@earth.li>**20080718184706]
1217[We need to make Parser.y and Config.hs earlier
1218Ian Lynagh <igloo@earth.li>**20080718180441]
1219[Explicitly list HpcParser as a module in hpc-bin
1220Ian Lynagh <igloo@earth.li>**20080718174657
1221 Cabal doesn't preprocess the .y file otherwise.
1222]
1223[Disable building pwd and lndir for now
1224Ian Lynagh <igloo@earth.li>**20080718170329]
1225[Build hpc with Cabal
1226Ian Lynagh <igloo@earth.li>**20080718170047]
1227[Build runghc with Cabal
1228Ian Lynagh <igloo@earth.li>**20080718165317]
1229[Add a comment
1230Ian Lynagh <igloo@earth.li>**20080718154238]
1231[Tweak the build system for installPackage
1232Ian Lynagh <igloo@earth.li>**20080718153956]
1233[More build system changes; hasktags is now built with Cabal
1234Ian Lynagh <igloo@earth.li>**20080718153459]
1235[Remove a comment
1236Ian Lynagh <igloo@earth.li>**20080718115044]
1237[More build system changes; ghc-pkg is now built with Cabal
1238Ian Lynagh <igloo@earth.li>**20080718114753]
1239[Fix some argument names
1240Ian Lynagh <igloo@earth.li>**20080717223543]
1241[Tweak the hsc2hs wrapper script
1242Ian Lynagh <igloo@earth.li>**20080717194916]
1243[Fix the order in which things get built
1244Ian Lynagh <igloo@earth.li>**20080717192402]
1245[Split building the ghc package and binary into "boot" and "all" steps
1246Ian Lynagh <igloo@earth.li>**20080717150746
1247 In "boot" we configure, and in "all" we do the actual building.
1248]
1249[Install the compiler during make install
1250Ian Lynagh <igloo@earth.li>**20080717150453
1251 For now we always install stage 2
1252]
1253[Do the building and installing of hsc2hs with the stage1 compiler
1254Ian Lynagh <igloo@earth.li>**20080717150420]
1255[Remove some duplication
1256Ian Lynagh <igloo@earth.li>**20080717144906]
1257[Windows fixes
1258Ian Lynagh <igloo@earth.li>**20080716222719]
1259[Fix GHC finding extra-gcc-opts on Windows
1260Ian Lynagh <igloo@earth.li>**20080716222457]
1261[Fix the inplace compiler finding package.conf on Windows
1262Ian Lynagh <igloo@earth.li>**20080716215000]
1263[Fix the build with GHC 6.4.2
1264Ian Lynagh <igloo@earth.li>**20080716192836]
1265[Get building GHC itself with Cabal more-or-less working
1266Ian Lynagh <igloo@earth.li>**20080716150441
1267 Installing and bindist creation don't work, but they were already broken.
1268 Only tested validating with one setup.
1269]
1270[TAG Before cabalised-GHC
1271Ian Lynagh <igloo@earth.li>**20080719132217]
1272Patch bundle hash:
1273e46abf062924f12b250a6dab260cd52fec66d866