Ticket #2159: bitmap.dpatch

File bitmap.dpatch, 220.5 KB (added by igloo, 19 months ago)
Line 
1
2New patches:
3
4[Make a bitmap for DynFlag's
5Ian Lynagh <igloo@earth.li>**20101021172536
6 Ignore-this: 63cae70b2759460bc79f7b1c4c23738a
7] {
8hunk ./compiler/main/DynFlags.hs 104
9+import Data.Bits
10hunk ./compiler/main/DynFlags.hs 109
11+import Data.Word
12hunk ./compiler/main/DynFlags.hs 116
13+data BitMapDynFlag = BitMapDynFlag !Word64 !Word64 !Word64
14+
15+-- BEGIN GEN
16+
17+bitmapDynFlagTest :: BitMapDynFlag -> DynFlag -> Bool
18+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cmm = testBit w1 0
19+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cmmz = testBit w1 1
20+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cmmz_pretty = testBit w1 2
21+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cps_cmm = testBit w1 3
22+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cvt_cmm = testBit w1 4
23+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm = testBit w1 5
24+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_native = testBit w1 6
25+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_liveness = testBit w1 7
26+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_coalesce = testBit w1 8
27+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_regalloc = testBit w1 9
28+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_regalloc_stages = testBit w1 10
29+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_conflicts = testBit w1 11
30+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_stats = testBit w1 12
31+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_asm_expanded = testBit w1 13
32+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_llvm = testBit w1 14
33+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cpranal = testBit w1 15
34+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_deriv = testBit w1 16
35+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_ds = testBit w1 17
36+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_flatC = testBit w1 18
37+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_foreign = testBit w1 19
38+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_inlinings = testBit w1 20
39+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rule_firings = testBit w1 21
40+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_occur_anal = testBit w1 22
41+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_parsed = testBit w1 23
42+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rn = testBit w1 24
43+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_simpl = testBit w1 25
44+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_simpl_iterations = testBit w1 26
45+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_simpl_phases = testBit w1 27
46+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_spec = testBit w1 28
47+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_prep = testBit w1 29
48+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_stg = testBit w1 30
49+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_stranal = testBit w1 31
50+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_tc = testBit w1 32
51+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_types = testBit w1 33
52+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rules = testBit w1 34
53+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_cse = testBit w1 35
54+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_worker_wrapper = testBit w1 36
55+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rn_trace = testBit w1 37
56+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rn_stats = testBit w1 38
57+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_opt_cmm = testBit w1 39
58+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_simpl_stats = testBit w1 40
59+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_tc_trace = testBit w1 41
60+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_if_trace = testBit w1 42
61+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_splices = testBit w1 43
62+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_BCOs = testBit w1 44
63+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_vect = testBit w1 45
64+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_hpc = testBit w1 46
65+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_rtti = testBit w1 47
66+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_source_stats = testBit w1 48
67+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_verbose_core2core = testBit w1 49
68+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_verbose_stg2stg = testBit w1 50
69+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_hi = testBit w1 51
70+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_hi_diffs = testBit w1 52
71+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_minimal_imports = testBit w1 53
72+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_mod_cycles = testBit w1 54
73+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_dump_view_pattern_commoning = testBit w1 55
74+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_faststring_stats = testBit w1 56
75+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_DumpToFile = testBit w1 57
76+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_D_no_debug_output = testBit w1 58
77+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_DoCoreLinting = testBit w1 59
78+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_DoStgLinting = testBit w1 60
79+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_DoCmmLinting = testBit w1 61
80+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_DoAsmLinting = testBit w1 62
81+bitmapDynFlagTest (BitMapDynFlag w1 _ _) Opt_WarnIsError = testBit w1 63
82+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnDuplicateExports = testBit w2 0
83+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnHiShadows = testBit w2 1
84+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnImplicitPrelude = testBit w2 2
85+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnIncompletePatterns = testBit w2 3
86+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnIncompletePatternsRecUpd = testBit w2 4
87+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMissingFields = testBit w2 5
88+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMissingImportList = testBit w2 6
89+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMissingMethods = testBit w2 7
90+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMissingSigs = testBit w2 8
91+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMissingLocalSigs = testBit w2 9
92+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnNameShadowing = testBit w2 10
93+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnOverlappingPatterns = testBit w2 11
94+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnTypeDefaults = testBit w2 12
95+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnMonomorphism = testBit w2 13
96+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnUnusedBinds = testBit w2 14
97+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnUnusedImports = testBit w2 15
98+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnUnusedMatches = testBit w2 16
99+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnWarningsDeprecations = testBit w2 17
100+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnDeprecatedFlags = testBit w2 18
101+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnDodgyExports = testBit w2 19
102+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnDodgyImports = testBit w2 20
103+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnOrphans = testBit w2 21
104+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnAutoOrphans = testBit w2 22
105+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnTabs = testBit w2 23
106+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnUnrecognisedPragmas = testBit w2 24
107+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnDodgyForeignImports = testBit w2 25
108+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnLazyUnliftedBindings = testBit w2 26
109+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnUnusedDoBind = testBit w2 27
110+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnWrongDoBind = testBit w2 28
111+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_WarnAlternativeLayoutRuleTransitional = testBit w2 29
112+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_PrintExplicitForalls = testBit w2 30
113+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_Strictness = testBit w2 31
114+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_FullLaziness = testBit w2 32
115+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_FloatIn = testBit w2 33
116+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_Specialise = testBit w2 34
117+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_StaticArgumentTransformation = testBit w2 35
118+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_CSE = testBit w2 36
119+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_LiberateCase = testBit w2 37
120+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_SpecConstr = testBit w2 38
121+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_DoLambdaEtaExpansion = testBit w2 39
122+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_IgnoreAsserts = testBit w2 40
123+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_DoEtaReduction = testBit w2 41
124+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_CaseMerge = testBit w2 42
125+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_UnboxStrictFields = testBit w2 43
126+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_MethodSharing = testBit w2 44
127+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_DictsCheap = testBit w2 45
128+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_EnableRewriteRules = testBit w2 46
129+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_Vectorise = testBit w2 47
130+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_RegsGraph = testBit w2 48
131+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_RegsIterative = testBit w2 49
132+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_IgnoreInterfacePragmas = testBit w2 50
133+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_OmitInterfacePragmas = testBit w2 51
134+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_ExposeAllUnfoldings = testBit w2 52
135+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_AutoSccsOnAllToplevs = testBit w2 53
136+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_AutoSccsOnExportedToplevs = testBit w2 54
137+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_AutoSccsOnIndividualCafs = testBit w2 55
138+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_Pp = testBit w2 56
139+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_ForceRecomp = testBit w2 57
140+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_DryRun = testBit w2 58
141+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_DoAsmMangling = testBit w2 59
142+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_ExcessPrecision = testBit w2 60
143+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_EagerBlackHoling = testBit w2 61
144+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_ReadUserPackageConf = testBit w2 62
145+bitmapDynFlagTest (BitMapDynFlag _ w2 _) Opt_NoHsMain = testBit w2 63
146+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_SplitObjs = testBit w3 0
147+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_StgStats = testBit w3 1
148+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_HideAllPackages = testBit w3 2
149+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_PrintBindResult = testBit w3 3
150+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_Haddock = testBit w3 4
151+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_HaddockOptions = testBit w3 5
152+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_Hpc_No_Auto = testBit w3 6
153+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_BreakOnException = testBit w3 7
154+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_BreakOnError = testBit w3 8
155+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_PrintEvldWithShow = testBit w3 9
156+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_PrintBindContents = testBit w3 10
157+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_GenManifest = testBit w3 11
158+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_EmbedManifest = testBit w3 12
159+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_EmitExternalCore = testBit w3 13
160+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_SharedImplib = testBit w3 14
161+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_BuildingCabalPackage = testBit w3 15
162+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_SSE2 = testBit w3 16
163+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_GhciSandbox = testBit w3 17
164+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_RunCPS = testBit w3 18
165+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_RunCPSZ = testBit w3 19
166+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_ConvertToZipCfgAndBack = testBit w3 20
167+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_AutoLinkPackages = testBit w3 21
168+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_ImplicitImportQualified = testBit w3 22
169+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_TryNewCodeGen = testBit w3 23
170+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepHiDiffs = testBit w3 24
171+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepHcFiles = testBit w3 25
172+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepSFiles = testBit w3 26
173+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepRawSFiles = testBit w3 27
174+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepTmpFiles = testBit w3 28
175+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepRawTokenStream = testBit w3 29
176+bitmapDynFlagTest (BitMapDynFlag _ _ w3) Opt_KeepLlvmFiles = testBit w3 30
177+
178+bitmapDynFlagSet :: BitMapDynFlag -> DynFlag -> BitMapDynFlag
179+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmm = BitMapDynFlag (setBit w1 0) w2 w3
180+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmmz = BitMapDynFlag (setBit w1 1) w2 w3
181+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmmz_pretty = BitMapDynFlag (setBit w1 2) w2 w3
182+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cps_cmm = BitMapDynFlag (setBit w1 3) w2 w3
183+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cvt_cmm = BitMapDynFlag (setBit w1 4) w2 w3
184+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm = BitMapDynFlag (setBit w1 5) w2 w3
185+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_native = BitMapDynFlag (setBit w1 6) w2 w3
186+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_liveness = BitMapDynFlag (setBit w1 7) w2 w3
187+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_coalesce = BitMapDynFlag (setBit w1 8) w2 w3
188+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_regalloc = BitMapDynFlag (setBit w1 9) w2 w3
189+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_regalloc_stages = BitMapDynFlag (setBit w1 10) w2 w3
190+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_conflicts = BitMapDynFlag (setBit w1 11) w2 w3
191+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_stats = BitMapDynFlag (setBit w1 12) w2 w3
192+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_expanded = BitMapDynFlag (setBit w1 13) w2 w3
193+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_llvm = BitMapDynFlag (setBit w1 14) w2 w3
194+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cpranal = BitMapDynFlag (setBit w1 15) w2 w3
195+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_deriv = BitMapDynFlag (setBit w1 16) w2 w3
196+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_ds = BitMapDynFlag (setBit w1 17) w2 w3
197+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_flatC = BitMapDynFlag (setBit w1 18) w2 w3
198+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_foreign = BitMapDynFlag (setBit w1 19) w2 w3
199+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_inlinings = BitMapDynFlag (setBit w1 20) w2 w3
200+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rule_firings = BitMapDynFlag (setBit w1 21) w2 w3
201+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_occur_anal = BitMapDynFlag (setBit w1 22) w2 w3
202+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_parsed = BitMapDynFlag (setBit w1 23) w2 w3
203+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn = BitMapDynFlag (setBit w1 24) w2 w3
204+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl = BitMapDynFlag (setBit w1 25) w2 w3
205+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_iterations = BitMapDynFlag (setBit w1 26) w2 w3
206+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_phases = BitMapDynFlag (setBit w1 27) w2 w3
207+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_spec = BitMapDynFlag (setBit w1 28) w2 w3
208+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_prep = BitMapDynFlag (setBit w1 29) w2 w3
209+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_stg = BitMapDynFlag (setBit w1 30) w2 w3
210+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_stranal = BitMapDynFlag (setBit w1 31) w2 w3
211+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_tc = BitMapDynFlag (setBit w1 32) w2 w3
212+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_types = BitMapDynFlag (setBit w1 33) w2 w3
213+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rules = BitMapDynFlag (setBit w1 34) w2 w3
214+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_cse = BitMapDynFlag (setBit w1 35) w2 w3
215+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_worker_wrapper = BitMapDynFlag (setBit w1 36) w2 w3
216+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn_trace = BitMapDynFlag (setBit w1 37) w2 w3
217+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn_stats = BitMapDynFlag (setBit w1 38) w2 w3
218+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_opt_cmm = BitMapDynFlag (setBit w1 39) w2 w3
219+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_stats = BitMapDynFlag (setBit w1 40) w2 w3
220+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_tc_trace = BitMapDynFlag (setBit w1 41) w2 w3
221+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_if_trace = BitMapDynFlag (setBit w1 42) w2 w3
222+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_splices = BitMapDynFlag (setBit w1 43) w2 w3
223+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_BCOs = BitMapDynFlag (setBit w1 44) w2 w3
224+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_vect = BitMapDynFlag (setBit w1 45) w2 w3
225+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_hpc = BitMapDynFlag (setBit w1 46) w2 w3
226+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_rtti = BitMapDynFlag (setBit w1 47) w2 w3
227+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_source_stats = BitMapDynFlag (setBit w1 48) w2 w3
228+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_verbose_core2core = BitMapDynFlag (setBit w1 49) w2 w3
229+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_verbose_stg2stg = BitMapDynFlag (setBit w1 50) w2 w3
230+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_hi = BitMapDynFlag (setBit w1 51) w2 w3
231+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_hi_diffs = BitMapDynFlag (setBit w1 52) w2 w3
232+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_minimal_imports = BitMapDynFlag (setBit w1 53) w2 w3
233+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_mod_cycles = BitMapDynFlag (setBit w1 54) w2 w3
234+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_dump_view_pattern_commoning = BitMapDynFlag (setBit w1 55) w2 w3
235+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_faststring_stats = BitMapDynFlag (setBit w1 56) w2 w3
236+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DumpToFile = BitMapDynFlag (setBit w1 57) w2 w3
237+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_D_no_debug_output = BitMapDynFlag (setBit w1 58) w2 w3
238+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoCoreLinting = BitMapDynFlag (setBit w1 59) w2 w3
239+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoStgLinting = BitMapDynFlag (setBit w1 60) w2 w3
240+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoCmmLinting = BitMapDynFlag (setBit w1 61) w2 w3
241+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoAsmLinting = BitMapDynFlag (setBit w1 62) w2 w3
242+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnIsError = BitMapDynFlag (setBit w1 63) w2 w3
243+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnDuplicateExports = BitMapDynFlag w1 (setBit w2 0) w3
244+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnHiShadows = BitMapDynFlag w1 (setBit w2 1) w3
245+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnImplicitPrelude = BitMapDynFlag w1 (setBit w2 2) w3
246+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnIncompletePatterns = BitMapDynFlag w1 (setBit w2 3) w3
247+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnIncompletePatternsRecUpd = BitMapDynFlag w1 (setBit w2 4) w3
248+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMissingFields = BitMapDynFlag w1 (setBit w2 5) w3
249+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMissingImportList = BitMapDynFlag w1 (setBit w2 6) w3
250+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMissingMethods = BitMapDynFlag w1 (setBit w2 7) w3
251+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMissingSigs = BitMapDynFlag w1 (setBit w2 8) w3
252+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMissingLocalSigs = BitMapDynFlag w1 (setBit w2 9) w3
253+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnNameShadowing = BitMapDynFlag w1 (setBit w2 10) w3
254+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnOverlappingPatterns = BitMapDynFlag w1 (setBit w2 11) w3
255+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnTypeDefaults = BitMapDynFlag w1 (setBit w2 12) w3
256+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnMonomorphism = BitMapDynFlag w1 (setBit w2 13) w3
257+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedBinds = BitMapDynFlag w1 (setBit w2 14) w3
258+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedImports = BitMapDynFlag w1 (setBit w2 15) w3
259+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedMatches = BitMapDynFlag w1 (setBit w2 16) w3
260+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnWarningsDeprecations = BitMapDynFlag w1 (setBit w2 17) w3
261+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnDeprecatedFlags = BitMapDynFlag w1 (setBit w2 18) w3
262+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyExports = BitMapDynFlag w1 (setBit w2 19) w3
263+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyImports = BitMapDynFlag w1 (setBit w2 20) w3
264+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnOrphans = BitMapDynFlag w1 (setBit w2 21) w3
265+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnAutoOrphans = BitMapDynFlag w1 (setBit w2 22) w3
266+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnTabs = BitMapDynFlag w1 (setBit w2 23) w3
267+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnUnrecognisedPragmas = BitMapDynFlag w1 (setBit w2 24) w3
268+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyForeignImports = BitMapDynFlag w1 (setBit w2 25) w3
269+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnLazyUnliftedBindings = BitMapDynFlag w1 (setBit w2 26) w3
270+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedDoBind = BitMapDynFlag w1 (setBit w2 27) w3
271+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnWrongDoBind = BitMapDynFlag w1 (setBit w2 28) w3
272+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_WarnAlternativeLayoutRuleTransitional = BitMapDynFlag w1 (setBit w2 29) w3
273+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_PrintExplicitForalls = BitMapDynFlag w1 (setBit w2 30) w3
274+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Strictness = BitMapDynFlag w1 (setBit w2 31) w3
275+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_FullLaziness = BitMapDynFlag w1 (setBit w2 32) w3
276+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_FloatIn = BitMapDynFlag w1 (setBit w2 33) w3
277+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Specialise = BitMapDynFlag w1 (setBit w2 34) w3
278+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_StaticArgumentTransformation = BitMapDynFlag w1 (setBit w2 35) w3
279+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_CSE = BitMapDynFlag w1 (setBit w2 36) w3
280+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_LiberateCase = BitMapDynFlag w1 (setBit w2 37) w3
281+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_SpecConstr = BitMapDynFlag w1 (setBit w2 38) w3
282+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoLambdaEtaExpansion = BitMapDynFlag w1 (setBit w2 39) w3
283+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_IgnoreAsserts = BitMapDynFlag w1 (setBit w2 40) w3
284+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoEtaReduction = BitMapDynFlag w1 (setBit w2 41) w3
285+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_CaseMerge = BitMapDynFlag w1 (setBit w2 42) w3
286+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_UnboxStrictFields = BitMapDynFlag w1 (setBit w2 43) w3
287+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_MethodSharing = BitMapDynFlag w1 (setBit w2 44) w3
288+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DictsCheap = BitMapDynFlag w1 (setBit w2 45) w3
289+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_EnableRewriteRules = BitMapDynFlag w1 (setBit w2 46) w3
290+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Vectorise = BitMapDynFlag w1 (setBit w2 47) w3
291+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_RegsGraph = BitMapDynFlag w1 (setBit w2 48) w3
292+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_RegsIterative = BitMapDynFlag w1 (setBit w2 49) w3
293+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_IgnoreInterfacePragmas = BitMapDynFlag w1 (setBit w2 50) w3
294+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_OmitInterfacePragmas = BitMapDynFlag w1 (setBit w2 51) w3
295+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ExposeAllUnfoldings = BitMapDynFlag w1 (setBit w2 52) w3
296+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnAllToplevs = BitMapDynFlag w1 (setBit w2 53) w3
297+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnExportedToplevs = BitMapDynFlag w1 (setBit w2 54) w3
298+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnIndividualCafs = BitMapDynFlag w1 (setBit w2 55) w3
299+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Pp = BitMapDynFlag w1 (setBit w2 56) w3
300+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ForceRecomp = BitMapDynFlag w1 (setBit w2 57) w3
301+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DryRun = BitMapDynFlag w1 (setBit w2 58) w3
302+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_DoAsmMangling = BitMapDynFlag w1 (setBit w2 59) w3
303+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ExcessPrecision = BitMapDynFlag w1 (setBit w2 60) w3
304+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_EagerBlackHoling = BitMapDynFlag w1 (setBit w2 61) w3
305+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ReadUserPackageConf = BitMapDynFlag w1 (setBit w2 62) w3
306+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_NoHsMain = BitMapDynFlag w1 (setBit w2 63) w3
307+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_SplitObjs = BitMapDynFlag w1 w2 (setBit w3 0)
308+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_StgStats = BitMapDynFlag w1 w2 (setBit w3 1)
309+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_HideAllPackages = BitMapDynFlag w1 w2 (setBit w3 2)
310+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_PrintBindResult = BitMapDynFlag w1 w2 (setBit w3 3)
311+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Haddock = BitMapDynFlag w1 w2 (setBit w3 4)
312+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_HaddockOptions = BitMapDynFlag w1 w2 (setBit w3 5)
313+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_Hpc_No_Auto = BitMapDynFlag w1 w2 (setBit w3 6)
314+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_BreakOnException = BitMapDynFlag w1 w2 (setBit w3 7)
315+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_BreakOnError = BitMapDynFlag w1 w2 (setBit w3 8)
316+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_PrintEvldWithShow = BitMapDynFlag w1 w2 (setBit w3 9)
317+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_PrintBindContents = BitMapDynFlag w1 w2 (setBit w3 10)
318+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_GenManifest = BitMapDynFlag w1 w2 (setBit w3 11)
319+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_EmbedManifest = BitMapDynFlag w1 w2 (setBit w3 12)
320+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_EmitExternalCore = BitMapDynFlag w1 w2 (setBit w3 13)
321+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_SharedImplib = BitMapDynFlag w1 w2 (setBit w3 14)
322+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_BuildingCabalPackage = BitMapDynFlag w1 w2 (setBit w3 15)
323+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_SSE2 = BitMapDynFlag w1 w2 (setBit w3 16)
324+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_GhciSandbox = BitMapDynFlag w1 w2 (setBit w3 17)
325+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_RunCPS = BitMapDynFlag w1 w2 (setBit w3 18)
326+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_RunCPSZ = BitMapDynFlag w1 w2 (setBit w3 19)
327+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ConvertToZipCfgAndBack = BitMapDynFlag w1 w2 (setBit w3 20)
328+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_AutoLinkPackages = BitMapDynFlag w1 w2 (setBit w3 21)
329+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_ImplicitImportQualified = BitMapDynFlag w1 w2 (setBit w3 22)
330+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_TryNewCodeGen = BitMapDynFlag w1 w2 (setBit w3 23)
331+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepHiDiffs = BitMapDynFlag w1 w2 (setBit w3 24)
332+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepHcFiles = BitMapDynFlag w1 w2 (setBit w3 25)
333+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepSFiles = BitMapDynFlag w1 w2 (setBit w3 26)
334+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepRawSFiles = BitMapDynFlag w1 w2 (setBit w3 27)
335+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepTmpFiles = BitMapDynFlag w1 w2 (setBit w3 28)
336+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepRawTokenStream = BitMapDynFlag w1 w2 (setBit w3 29)
337+bitmapDynFlagSet (BitMapDynFlag w1 w2 w3) Opt_KeepLlvmFiles = BitMapDynFlag w1 w2 (setBit w3 30)
338+
339+bitmapDynFlagClear :: BitMapDynFlag -> DynFlag -> BitMapDynFlag
340+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmm = BitMapDynFlag (clearBit w1 0) w2 w3
341+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmmz = BitMapDynFlag (clearBit w1 1) w2 w3
342+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cmmz_pretty = BitMapDynFlag (clearBit w1 2) w2 w3
343+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cps_cmm = BitMapDynFlag (clearBit w1 3) w2 w3
344+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cvt_cmm = BitMapDynFlag (clearBit w1 4) w2 w3
345+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm = BitMapDynFlag (clearBit w1 5) w2 w3
346+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_native = BitMapDynFlag (clearBit w1 6) w2 w3
347+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_liveness = BitMapDynFlag (clearBit w1 7) w2 w3
348+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_coalesce = BitMapDynFlag (clearBit w1 8) w2 w3
349+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_regalloc = BitMapDynFlag (clearBit w1 9) w2 w3
350+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_regalloc_stages = BitMapDynFlag (clearBit w1 10) w2 w3
351+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_conflicts = BitMapDynFlag (clearBit w1 11) w2 w3
352+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_stats = BitMapDynFlag (clearBit w1 12) w2 w3
353+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_asm_expanded = BitMapDynFlag (clearBit w1 13) w2 w3
354+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_llvm = BitMapDynFlag (clearBit w1 14) w2 w3
355+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cpranal = BitMapDynFlag (clearBit w1 15) w2 w3
356+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_deriv = BitMapDynFlag (clearBit w1 16) w2 w3
357+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_ds = BitMapDynFlag (clearBit w1 17) w2 w3
358+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_flatC = BitMapDynFlag (clearBit w1 18) w2 w3
359+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_foreign = BitMapDynFlag (clearBit w1 19) w2 w3
360+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_inlinings = BitMapDynFlag (clearBit w1 20) w2 w3
361+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rule_firings = BitMapDynFlag (clearBit w1 21) w2 w3
362+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_occur_anal = BitMapDynFlag (clearBit w1 22) w2 w3
363+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_parsed = BitMapDynFlag (clearBit w1 23) w2 w3
364+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn = BitMapDynFlag (clearBit w1 24) w2 w3
365+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl = BitMapDynFlag (clearBit w1 25) w2 w3
366+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_iterations = BitMapDynFlag (clearBit w1 26) w2 w3
367+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_phases = BitMapDynFlag (clearBit w1 27) w2 w3
368+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_spec = BitMapDynFlag (clearBit w1 28) w2 w3
369+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_prep = BitMapDynFlag (clearBit w1 29) w2 w3
370+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_stg = BitMapDynFlag (clearBit w1 30) w2 w3
371+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_stranal = BitMapDynFlag (clearBit w1 31) w2 w3
372+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_tc = BitMapDynFlag (clearBit w1 32) w2 w3
373+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_types = BitMapDynFlag (clearBit w1 33) w2 w3
374+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rules = BitMapDynFlag (clearBit w1 34) w2 w3
375+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_cse = BitMapDynFlag (clearBit w1 35) w2 w3
376+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_worker_wrapper = BitMapDynFlag (clearBit w1 36) w2 w3
377+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn_trace = BitMapDynFlag (clearBit w1 37) w2 w3
378+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rn_stats = BitMapDynFlag (clearBit w1 38) w2 w3
379+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_opt_cmm = BitMapDynFlag (clearBit w1 39) w2 w3
380+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_simpl_stats = BitMapDynFlag (clearBit w1 40) w2 w3
381+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_tc_trace = BitMapDynFlag (clearBit w1 41) w2 w3
382+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_if_trace = BitMapDynFlag (clearBit w1 42) w2 w3
383+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_splices = BitMapDynFlag (clearBit w1 43) w2 w3
384+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_BCOs = BitMapDynFlag (clearBit w1 44) w2 w3
385+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_vect = BitMapDynFlag (clearBit w1 45) w2 w3
386+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_hpc = BitMapDynFlag (clearBit w1 46) w2 w3
387+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_rtti = BitMapDynFlag (clearBit w1 47) w2 w3
388+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_source_stats = BitMapDynFlag (clearBit w1 48) w2 w3
389+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_verbose_core2core = BitMapDynFlag (clearBit w1 49) w2 w3
390+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_verbose_stg2stg = BitMapDynFlag (clearBit w1 50) w2 w3
391+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_hi = BitMapDynFlag (clearBit w1 51) w2 w3
392+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_hi_diffs = BitMapDynFlag (clearBit w1 52) w2 w3
393+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_minimal_imports = BitMapDynFlag (clearBit w1 53) w2 w3
394+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_mod_cycles = BitMapDynFlag (clearBit w1 54) w2 w3
395+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_dump_view_pattern_commoning = BitMapDynFlag (clearBit w1 55) w2 w3
396+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_faststring_stats = BitMapDynFlag (clearBit w1 56) w2 w3
397+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DumpToFile = BitMapDynFlag (clearBit w1 57) w2 w3
398+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_D_no_debug_output = BitMapDynFlag (clearBit w1 58) w2 w3
399+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoCoreLinting = BitMapDynFlag (clearBit w1 59) w2 w3
400+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoStgLinting = BitMapDynFlag (clearBit w1 60) w2 w3
401+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoCmmLinting = BitMapDynFlag (clearBit w1 61) w2 w3
402+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoAsmLinting = BitMapDynFlag (clearBit w1 62) w2 w3
403+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnIsError = BitMapDynFlag (clearBit w1 63) w2 w3
404+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnDuplicateExports = BitMapDynFlag w1 (clearBit w2 0) w3
405+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnHiShadows = BitMapDynFlag w1 (clearBit w2 1) w3
406+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnImplicitPrelude = BitMapDynFlag w1 (clearBit w2 2) w3
407+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnIncompletePatterns = BitMapDynFlag w1 (clearBit w2 3) w3
408+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnIncompletePatternsRecUpd = BitMapDynFlag w1 (clearBit w2 4) w3
409+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMissingFields = BitMapDynFlag w1 (clearBit w2 5) w3
410+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMissingImportList = BitMapDynFlag w1 (clearBit w2 6) w3
411+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMissingMethods = BitMapDynFlag w1 (clearBit w2 7) w3
412+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMissingSigs = BitMapDynFlag w1 (clearBit w2 8) w3
413+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMissingLocalSigs = BitMapDynFlag w1 (clearBit w2 9) w3
414+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnNameShadowing = BitMapDynFlag w1 (clearBit w2 10) w3
415+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnOverlappingPatterns = BitMapDynFlag w1 (clearBit w2 11) w3
416+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnTypeDefaults = BitMapDynFlag w1 (clearBit w2 12) w3
417+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnMonomorphism = BitMapDynFlag w1 (clearBit w2 13) w3
418+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedBinds = BitMapDynFlag w1 (clearBit w2 14) w3
419+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedImports = BitMapDynFlag w1 (clearBit w2 15) w3
420+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedMatches = BitMapDynFlag w1 (clearBit w2 16) w3
421+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnWarningsDeprecations = BitMapDynFlag w1 (clearBit w2 17) w3
422+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnDeprecatedFlags = BitMapDynFlag w1 (clearBit w2 18) w3
423+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyExports = BitMapDynFlag w1 (clearBit w2 19) w3
424+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyImports = BitMapDynFlag w1 (clearBit w2 20) w3
425+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnOrphans = BitMapDynFlag w1 (clearBit w2 21) w3
426+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnAutoOrphans = BitMapDynFlag w1 (clearBit w2 22) w3
427+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnTabs = BitMapDynFlag w1 (clearBit w2 23) w3
428+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnUnrecognisedPragmas = BitMapDynFlag w1 (clearBit w2 24) w3
429+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnDodgyForeignImports = BitMapDynFlag w1 (clearBit w2 25) w3
430+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnLazyUnliftedBindings = BitMapDynFlag w1 (clearBit w2 26) w3
431+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnUnusedDoBind = BitMapDynFlag w1 (clearBit w2 27) w3
432+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnWrongDoBind = BitMapDynFlag w1 (clearBit w2 28) w3
433+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_WarnAlternativeLayoutRuleTransitional = BitMapDynFlag w1 (clearBit w2 29) w3
434+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_PrintExplicitForalls = BitMapDynFlag w1 (clearBit w2 30) w3
435+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Strictness = BitMapDynFlag w1 (clearBit w2 31) w3
436+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_FullLaziness = BitMapDynFlag w1 (clearBit w2 32) w3
437+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_FloatIn = BitMapDynFlag w1 (clearBit w2 33) w3
438+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Specialise = BitMapDynFlag w1 (clearBit w2 34) w3
439+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_StaticArgumentTransformation = BitMapDynFlag w1 (clearBit w2 35) w3
440+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_CSE = BitMapDynFlag w1 (clearBit w2 36) w3
441+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_LiberateCase = BitMapDynFlag w1 (clearBit w2 37) w3
442+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_SpecConstr = BitMapDynFlag w1 (clearBit w2 38) w3
443+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoLambdaEtaExpansion = BitMapDynFlag w1 (clearBit w2 39) w3
444+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_IgnoreAsserts = BitMapDynFlag w1 (clearBit w2 40) w3
445+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoEtaReduction = BitMapDynFlag w1 (clearBit w2 41) w3
446+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_CaseMerge = BitMapDynFlag w1 (clearBit w2 42) w3
447+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_UnboxStrictFields = BitMapDynFlag w1 (clearBit w2 43) w3
448+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_MethodSharing = BitMapDynFlag w1 (clearBit w2 44) w3
449+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DictsCheap = BitMapDynFlag w1 (clearBit w2 45) w3
450+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_EnableRewriteRules = BitMapDynFlag w1 (clearBit w2 46) w3
451+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Vectorise = BitMapDynFlag w1 (clearBit w2 47) w3
452+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_RegsGraph = BitMapDynFlag w1 (clearBit w2 48) w3
453+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_RegsIterative = BitMapDynFlag w1 (clearBit w2 49) w3
454+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_IgnoreInterfacePragmas = BitMapDynFlag w1 (clearBit w2 50) w3
455+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_OmitInterfacePragmas = BitMapDynFlag w1 (clearBit w2 51) w3
456+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ExposeAllUnfoldings = BitMapDynFlag w1 (clearBit w2 52) w3
457+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnAllToplevs = BitMapDynFlag w1 (clearBit w2 53) w3
458+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnExportedToplevs = BitMapDynFlag w1 (clearBit w2 54) w3
459+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_AutoSccsOnIndividualCafs = BitMapDynFlag w1 (clearBit w2 55) w3
460+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Pp = BitMapDynFlag w1 (clearBit w2 56) w3
461+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ForceRecomp = BitMapDynFlag w1 (clearBit w2 57) w3
462+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DryRun = BitMapDynFlag w1 (clearBit w2 58) w3
463+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_DoAsmMangling = BitMapDynFlag w1 (clearBit w2 59) w3
464+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ExcessPrecision = BitMapDynFlag w1 (clearBit w2 60) w3
465+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_EagerBlackHoling = BitMapDynFlag w1 (clearBit w2 61) w3
466+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ReadUserPackageConf = BitMapDynFlag w1 (clearBit w2 62) w3
467+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_NoHsMain = BitMapDynFlag w1 (clearBit w2 63) w3
468+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_SplitObjs = BitMapDynFlag w1 w2 (clearBit w3 0)
469+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_StgStats = BitMapDynFlag w1 w2 (clearBit w3 1)
470+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_HideAllPackages = BitMapDynFlag w1 w2 (clearBit w3 2)
471+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_PrintBindResult = BitMapDynFlag w1 w2 (clearBit w3 3)
472+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Haddock = BitMapDynFlag w1 w2 (clearBit w3 4)
473+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_HaddockOptions = BitMapDynFlag w1 w2 (clearBit w3 5)
474+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_Hpc_No_Auto = BitMapDynFlag w1 w2 (clearBit w3 6)
475+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_BreakOnException = BitMapDynFlag w1 w2 (clearBit w3 7)
476+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_BreakOnError = BitMapDynFlag w1 w2 (clearBit w3 8)
477+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_PrintEvldWithShow = BitMapDynFlag w1 w2 (clearBit w3 9)
478+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_PrintBindContents = BitMapDynFlag w1 w2 (clearBit w3 10)
479+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_GenManifest = BitMapDynFlag w1 w2 (clearBit w3 11)
480+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_EmbedManifest = BitMapDynFlag w1 w2 (clearBit w3 12)
481+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_EmitExternalCore = BitMapDynFlag w1 w2 (clearBit w3 13)
482+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_SharedImplib = BitMapDynFlag w1 w2 (clearBit w3 14)
483+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_BuildingCabalPackage = BitMapDynFlag w1 w2 (clearBit w3 15)
484+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_SSE2 = BitMapDynFlag w1 w2 (clearBit w3 16)
485+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_GhciSandbox = BitMapDynFlag w1 w2 (clearBit w3 17)
486+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_RunCPS = BitMapDynFlag w1 w2 (clearBit w3 18)
487+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_RunCPSZ = BitMapDynFlag w1 w2 (clearBit w3 19)
488+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ConvertToZipCfgAndBack = BitMapDynFlag w1 w2 (clearBit w3 20)
489+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_AutoLinkPackages = BitMapDynFlag w1 w2 (clearBit w3 21)
490+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_ImplicitImportQualified = BitMapDynFlag w1 w2 (clearBit w3 22)
491+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_TryNewCodeGen = BitMapDynFlag w1 w2 (clearBit w3 23)
492+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepHiDiffs = BitMapDynFlag w1 w2 (clearBit w3 24)
493+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepHcFiles = BitMapDynFlag w1 w2 (clearBit w3 25)
494+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepSFiles = BitMapDynFlag w1 w2 (clearBit w3 26)
495+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepRawSFiles = BitMapDynFlag w1 w2 (clearBit w3 27)
496+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepTmpFiles = BitMapDynFlag w1 w2 (clearBit w3 28)
497+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepRawTokenStream = BitMapDynFlag w1 w2 (clearBit w3 29)
498+bitmapDynFlagClear (BitMapDynFlag w1 w2 w3) Opt_KeepLlvmFiles = BitMapDynFlag w1 w2 (clearBit w3 30)
499+
500+-- END GEN
501+
502hunk ./compiler/main/DynFlags.hs 993
503-  flags                 :: [DynFlag],
504+  flags                 :: BitMapDynFlag,
505hunk ./compiler/main/DynFlags.hs 1232
506-        flags = defaultFlags,
507+        flags = defaultFlagBitmap,
508hunk ./compiler/main/DynFlags.hs 1318
509-dopt f dflags  = f `elem` (flags dflags)
510+dopt f dflags  = bitmapDynFlagTest (flags dflags) f
511hunk ./compiler/main/DynFlags.hs 1322
512-dopt_set dfs f = dfs{ flags = f : flags dfs }
513+dopt_set dfs f = dfs{ flags = bitmapDynFlagSet (flags dfs) f }
514hunk ./compiler/main/DynFlags.hs 1326
515-dopt_unset dfs f = dfs{ flags = filter (/= f) (flags dfs) }
516+dopt_unset dfs f = dfs{ flags = bitmapDynFlagClear (flags dfs) f }
517hunk ./compiler/main/DynFlags.hs 2127
518+defaultFlagBitmap :: BitMapDynFlag
519+defaultFlagBitmap = foldl' bitmapDynFlagSet zeroBitmap defaultFlags
520+    where zeroBitmap = BitMapDynFlag 0 0 0
521+
522}
523
524Context:
525
526[Switch more uniqFromSupply+splitUniqSupply's to takeUniqFromSupply
527Ian Lynagh <igloo@earth.li>**20101021142824
528 Ignore-this: 27991ada81de2d3ef7f4757ac6cc8d11
529]
530[Whitespace only
531Ian Lynagh <igloo@earth.li>**20101021135551]
532[Whitespace only
533Ian Lynagh <igloo@earth.li>**20101021134404]
534[Whitespace only
535Ian Lynagh <igloo@earth.li>**20101021134133]
536[Windows installer improvements from Claus
537Ian Lynagh <igloo@earth.li>**20101021122942
538 - add link to inno setup docs, so readers don't have to guess what
539   ghc.iss might be
540 
541 - add Task section, and associated most Registry actions with separate
542   (sub)tasks, so that file associations, default action, and PATH
543   setting can be optional
544 
545 - copy license file into doc directory
546 
547 - install icon file again, so that DefaultIcon is no longer a dangling
548   pointer (#4352)
549 
550 - only delete ghc_haskell key if empty (in case there were other tools
551   using it)
552 
553 - add versioned GHCi to right-click menu (to allow for multiple tool installs)
554]
555[Use takeUniqFromSupply in ByteCodeGen
556Ian Lynagh <igloo@earth.li>**20101021121454]
557[Fix some whitespace
558Ian Lynagh <igloo@earth.li>**20101021121059]
559[Use takeUniqFromSupply in emitProcWithConvention
560Ian Lynagh <igloo@earth.li>**20101021120853
561 We were using the supply's unique, and then passing the same supply to
562 initUs_, which sounds like a bug waiting to happen.
563]
564[Use takeUniqFromSupply in IfaceEnv
565Ian Lynagh <igloo@earth.li>**20101021120304
566 This is a little nicer than having to explicitly split supplies and
567 throw half of them away.
568]
569[Define takeUniqFromSupply
570Ian Lynagh <igloo@earth.li>**20101021115822]
571[Remove some extraneous whitespace
572Ian Lynagh <igloo@earth.li>**20101021115600]
573[Add a comment on why some seq's are done
574Ian Lynagh <igloo@earth.li>**20101021113404
575 Ignore-this: 6c8e59eeb2f8733f7ff7686a28463728
576]
577[Tidy up RuntimeUnkSkols a bit more
578simonpj@microsoft.com**20101021100640
579 Ignore-this: eb3f5617eff99512a2e72d748c7ad99
580]
581[Fix haddock markup
582simonpj@microsoft.com**20101021085420
583 Ignore-this: a2705c5aaea64648770eb06ffbcf1e8f
584]
585[Improve rule checking, to fix panic Trac #4398
586simonpj@microsoft.com**20101021085402
587 Ignore-this: 7f1f8496f336b4e52374fc99110db4a8
588 
589 Lots of comments with decomposeRuleLhs
590]
591[Improve the simple expression optimiser so it does simple beta reduction
592simonpj@microsoft.com**20101021085030
593 Ignore-this: 7a8f99b59d01389ffada547dd29005f9
594]
595[Template Haskell: add view patterns (Trac #2399)
596Reiner Pope <reiner.pope@gmail.com>**20101010131720
597 Ignore-this: 710855eb7e9cdef96478f2f4da9cfbf4
598]
599[Tidy-up sweep, following the Great Skolemisation Simplification
600simonpj@microsoft.com**20101021070837
601 Ignore-this: ec821a200d54fb7c9d1135e2c4ce581
602]
603[Fix haddock markup
604Ian Lynagh <igloo@earth.li>**20101020151952
605 Ignore-this: 3d7bdf6a84a6264ae4cea215c415d821
606]
607[Tweak the haddock rules; no functional change
608Ian Lynagh <igloo@earth.li>**20101020151636
609 Ignore-this: bee8a936b3ebd33832fdd58161ba725
610]
611[Comments and layout only
612simonpj@microsoft.com**20101020132438
613 Ignore-this: ae58e1d0dbb2d7d9ce02fe765cd3591f
614]
615[Don't seq unfoldings
616Ian Lynagh <igloo@earth.li>**20101020143710
617 Ignore-this: 972ae43afdca1b7c6713c6ee5af14da4
618 We generate intermediate unfoldings which are just thrown away, so
619 evaluating them is a waste of time.
620]
621[fix markup
622Simon Marlow <marlowsd@gmail.com>**20101020123116
623 Ignore-this: 51372bccc13905f45610b106cacce31c
624]
625[remove xref to hasktags
626Simon Marlow <marlowsd@gmail.com>**20101020123106
627 Ignore-this: 1211f44dca795532adc729d08fde30b8
628]
629[Update the documentation on using DLL's from Windows, fixing several errors, in particular those relating to bug 3605
630Neil Mitchell **20101010100709
631 Ignore-this: 13b44b074b123c21f1b2257c9b1d0585
632]
633[(1) More lenient kind checking, (2) Fixed orientation problems and avoiding double unifications, (3) Comments
634dimitris@microsoft.com**20101020115526]
635[Midstream changes to deal with spontaneous solving and flatten skolem equivalence classes
636dimitris@microsoft.com**20101019171514]
637[hasktags was dropped in GHC 6.12, remove it from the docs
638Simon Marlow <marlowsd@gmail.com>**20101018084401
639 Ignore-this: e17307814d9c91f1197c07e0ea421f88
640]
641[Refactor, plus fix Trac #4418
642simonpj@microsoft.com**20101020090946
643 Ignore-this: 401774da6fbd69d567a27a58be3dd9b8
644 
645 We weren't doing fundeps for derived superclasses
646]
647[Add a comment, connecting the seq to the test (#4367) that shows its usefulness
648simonpj@microsoft.com**20101020075941
649 Ignore-this: 2dae7e226168a25e5753467c08e32414
650]
651[Avoid hanging on to old unfoldings; fixes #4367 (compiler space regression)
652Ian Lynagh <igloo@earth.li>**20101020131539]
653[Look for sources in Cabal's autogen directory too
654Ian Lynagh <igloo@earth.li>**20101020104759
655 Ignore-this: 709b8b491ed89c9d979b93dbaa6e7069
656]
657[Follow Cabal change: Use usedExtensions rather than extensions
658Ian Lynagh <igloo@earth.li>**20101020104739
659 Ignore-this: 18f7fe6729e573c0b75746b5f891bb71
660]
661[Define setIdUnfoldingLazily, and use it in Vectorise
662Ian Lynagh <igloo@earth.li>**20101019201537
663 Fixes a loop in the compiler, when running the dph tests
664]
665[Evaluate the results in coreToStgApp
666Ian Lynagh <igloo@earth.li>**20101019170532
667 Ignore-this: 3b3439d35ac21e664b5d24a070d6e8ca
668]
669[seq the unfolding in setUnfoldingInfo
670Ian Lynagh <igloo@earth.li>**20101019154552
671 Ignore-this: a18677dfdc1f1dddc82e89b3267d2bb1
672 Contrary to the comment, for the module in #4367 at least, it is a big
673 improvement. Without it we get a huge spike of drag.
674]
675[Comments only
676simonpj@microsoft.com**20101019153531
677 Ignore-this: 159849457d44f7036f22004075ac1377
678]
679[Fix debugger
680simonpj@microsoft.com**20101019153522
681 Ignore-this: ea1fb8253a1ac45bdc58d4f1ea9aa0f9
682 
683 A bit yukky; see Note [Runtime skolems] in TcErrors.
684 But it works, and the debugger just is yukky in places.
685]
686[Fix IPRun by fixing the inferred quantification mechanism
687simonpj@microsoft.com**20101019090220
688 Ignore-this: e7dc8c35979136237fa9a572b70f298e
689]
690[Recover after an error in an implication constraint
691simonpj@microsoft.com**20101019090202
692 Ignore-this: fbbbdc2c3fc0bac0662d6b40e5d57d47
693]
694[Reject programs with equality superclasses for now
695simonpj@microsoft.com**20101019090140
696 Ignore-this: 3e3eb8c876fa95290f0278e800fce5bc
697]
698[Layout and tiny refactoring only
699simonpj@microsoft.com**20101019090124
700 Ignore-this: ee21565fde8cd45514d83d0ca9af3253
701]
702[Clean up the debugger code
703simonpj@microsoft.com**20101019090100
704 Ignore-this: 8677a9084d787f45542419a58e5d8866
705 
706 In particular there is much less fiddly skolemisation now
707 Things are not *quite* right (break001 and 006 still fail),
708 but they are *much* better than before.
709]
710[Add new VarEnv functions minusVarEnv, intersectsVarEnv, unionInScope
711simonpj@microsoft.com**20101019085609
712 Ignore-this: a3a8e23fe522079e3eb9c34ea8032602
713]
714[Major pass through type checker:(1) prioritizing equalities, (2) improved Derived mechanism, (3) bugfixes
715dimitris@microsoft.com**20101018151510]
716[(1) Caching FD improvements for efficiency, (2) preventing cascading deriveds from entering the inert, (3) Fixing bugs in the creation of FlexiTcS variables
717dimitris@microsoft.com**20101015164421]
718[Midstream changes for performance improvement related to superclasses and functional dependencies.
719dimitris@microsoft.com**20101014143811]
720[Minor
721dimitris@microsoft.com**20101012080951]
722[Commentary changes
723dimitris@microsoft.com**20101011142235]
724[Kind checking bugfix (#4356) and preventing wanteds from rewriting wanteds
725dimitris@microsoft.com**20101008173700]
726[Retab CoreToStg, and remove trailing whitespace
727Ian Lynagh <igloo@earth.li>**20101019155530]
728[Fix a retainer profiling segfault
729Ian Lynagh <igloo@earth.li>**20101019132727
730 Ignore-this: 19f9e2a6dc8a64b30f3769f4101f82db
731 The bitmap type wasn't big enough to hold large bitmaps on 64 bit
732 platforms. Profiling GHC was segfaulting when retainStack was handling a
733 size 33 bitmap.
734]
735[Define SpecConstrAnnotation in GHC.Exts, and import it from there
736simonpj@microsoft.com**20101018135746
737 Ignore-this: a6ade815584a1d03be55a6c417de0ab0
738 
739 Reason: avoid having to link the entire ghc package in modules
740 that use compile-time annotations:
741 
742      import GHC.Exts( SpecConstrAnnotation )
743      {-# ANN type T ForceSpecConstr #-}
744 
745 It's a kind of bug that the package exporting SpecConstrAnnotation
746 is linked even though it is only needed at compile time, but putting
747 the data type declaration in GHC.Exts is a simple way to sidestep
748 the problem
749 
750 See See Note [SpecConstrAnnotation] in SpecConstr
751]
752[Fix warnings in AsmCodeGen
753David Terei <davidterei@gmail.com>**20101007143559
754 Ignore-this: 1c34b3a39968a5ca123862726f36a8ee
755]
756[LLVM: Fix compilation of writebarrier, #4308
757David Terei <davidterei@gmail.com>**20101004153843
758 Ignore-this: df256e78b90e3fc263a88d4f0af6ea96
759]
760[Fix -auto-all: Add SCCs to IDs which have a monotype too
761Ian Lynagh <igloo@earth.li>**20101018153957]
762[Change how the OS X installer's create-links finds the versin number
763Ian Lynagh <igloo@earth.li>**20101017122352
764 Ignore-this: 543d3f40c2ba5daf8b2b4ea22eb4bfca
765 It now gets created by configure, rather than trying to work out the
766 version number at runtime.
767]
768[Add more quoting to distrib/MacOS/installer-scripts/create-links
769Ian Lynagh <igloo@earth.li>**20101017112648
770 Ignore-this: 5119ea2efc925d568d96c616af550c7c
771]
772[change os x installer to allow multiple installed versions
773Evan Laforge <qdunkan@gmail.com>**20100929234538
774 Ignore-this: 4f70551293f53a52c39c5962d8a5bd5b
775 This puts the ghc version into the package name so they are considered separate
776 packages.
777]
778[Only put the boot packages in the haddock contents/index
779Ian Lynagh <igloo@earth.li>**20101016180031
780 Ignore-this: 3dc276dc734fec207def24a1d10eb369
781 We don't install dph etc, so don't put them in the doc index.
782]
783[Correct the regexp used to search for extra packages
784Ian Lynagh <igloo@earth.li>**20101016123421
785 Ignore-this: 98e3f8fda1c5678730a00ec1ac25fde3
786 We weren't ignoring comment lines
787]
788[New member "archiveMemberName" for struct _ObjectCode
789pho@cielonegro.org**20100927224145
790 Ignore-this: 628bc605e7dc4f0c4856c6f7ad23d9ee
791 
792 struct _ObjectCode should be able to retain the name of archive members.
793 Though currently the only use of those names are for debugging outputs.
794]
795[Add a -fghci-sandbox flag so that we can en/disable the ghci sandbox
796Ian Lynagh <igloo@earth.li>**20101015172746
797 Ignore-this: 52bc5729437a93a925d3586d9803623c
798 It's on by default (which matches the previous behaviour).
799 
800 Motivation:
801 GLUT on OS X needs to run on the main thread. If you
802 try to use it from another thread then you just get a
803 white rectangle rendered. For this, or anything else
804 with such restrictions, you can turn the GHCi sandbox off
805 and things will be run in the main thread.
806]
807[Fix boot; it was failing if darcs-all or validate were missing
808Ian Lynagh <igloo@earth.li>**20101015164549
809 Ignore-this: 14b3e98836647aff345e35bca2102f93
810 (which is the case in sdists)
811]
812[Comments and layout
813simonpj@microsoft.com**20101015131924
814 Ignore-this: 126fbdb629a08c1380c7a1f5cd967d97
815]
816[Make (Located a) an instance of Eq, Ord
817simonpj@microsoft.com**20101015131857
818 Ignore-this: 5da50f8dab06fcbc23ce149cd5080062
819 
820 Fulfils Trac #4369
821]
822[Give user-defined rules precedence over built-in rules
823simonpj@microsoft.com**20101015131814
824 Ignore-this: 76fbf36cb1a09e31c10c68c06b98937e
825 
826 This fixes Trac #4397.  See comments with 'isMoreSpecific'.
827]
828[Fix Trac #4401: meta-tyvars allocated by the constraint solver are always touchable
829simonpj@microsoft.com**20101015130818
830 Ignore-this: 7fcdaee825d9cadcf69c8c9b8967a446
831 
832   See Note [Touchable meta type variables] in TcSMonad
833]
834[Remove GHC.extendGlobalRdrScope, GHC.extendGlobalTypeScope
835simonpj@microsoft.com**20101013091107
836 Ignore-this: df65a43d261353ad53811b25507e913e
837 
838 These functions were added by
839 
840    Tue Apr 18 03:36:06 BST 2006  Lemmih <lemmih@gmail.com>
841    * Make the initial rdr and type scope available in the ghc-api
842 
843 The are extremely dubious, because they extend the Rdr and Type
844 env for every compilation.  The right thing to do is to use
845 the InteractiveContext for temporary extensions.
846 
847 So far as we know, no one uses them.  And if they are being used
848 it's probably a mistake.  So we're backing them out.
849]
850[InlinePrag needs an arity only for INLINE, not INLINABLE
851Simon Marlow <marlowsd@gmail.com>**20101015094925
852 Ignore-this: 3b338f0b2b49fa714b195067e0026ef7
853 This doesn't fix anything (we think), but it's morally correct.
854]
855[Fix #4346 (INLINABLE pragma not behaving consistently)
856Simon Marlow <marlowsd@gmail.com>**20101015094836
857 Ignore-this: 9a9d8ad42cfdf7f619adfac284bae021
858 Debugged thanks to lots of help from Simon PJ: we weren't updating the
859 UnfoldingGuidance when the unfolding changed.
860 Also, a bit of refactoring and additinoal comments.
861]
862[Have boot check that we have the dph packages when validating
863Ian Lynagh <igloo@earth.li>**20101014140556
864 Ignore-this: 61e365127933f4dbfb62be66115455bd
865]
866[Add more documentation for interruptible foreign calls
867Simon Marlow <marlowsd@gmail.com>**20101014084253
868 Ignore-this: 28ec0ddd958926a08ab816b6975da344
869]
870[minor refactoring
871Simon Marlow <marlowsd@gmail.com>**20100926105819
872 Ignore-this: 70d38e0a31a096c94dad5f346b0d91f6
873]
874[Fix for interruptible FFI handling
875Simon Marlow <marlowsd@gmail.com>**20100925193442
876 Ignore-this: a63ba3c60f577002a1d32b30bb45090c
877 Set tso->why_blocked before calling maybePerformBlockedException(), so
878 that throwToSingleThreaded() doesn't try to unblock the current thread
879 (it is already unblocked).
880]
881[interruptible FFI: more robust handling of the exception case in the interpreter
882Simon Marlow <marlowsd@gmail.com>**20100925193317
883 Ignore-this: 7f9e67835a3bd096154db2dcf533ec66
884]
885[Don't interrupt when task blocks exceptions, don't immediately start exception.
886Edward Z. Yang <ezyang@mit.edu>**20100925033026
887 Ignore-this: 6c0669995a2b66abf29d68b3711cb78e
888]
889[Interruptible FFI calls with pthread_kill and CancelSynchronousIO. v4
890Edward Z. Yang <ezyang@mit.edu>**20100919002905
891 Ignore-this: 43c260f90eeb9c03413f6e749e23101d
892 
893 This is patch that adds support for interruptible FFI calls in the form
894 of a new foreign import keyword 'interruptible', which can be used
895 instead of 'safe' or 'unsafe'.  Interruptible FFI calls act like safe
896 FFI calls, except that the worker thread they run on may be interrupted.
897 
898 Internally, it replaces BlockedOnCCall_NoUnblockEx with
899 BlockedOnCCall_Interruptible, and changes the behavior of the RTS
900 to not modify the TSO_ flags on the event of an FFI call from
901 a thread that was interruptible.  It also modifies the bytecode
902 format for foreign call, adding an extra Word16 to indicate
903 interruptibility.
904 
905 The semantics of interruption vary from platform to platform, but the
906 intent is that any blocking system calls are aborted with an error code.
907 This is most useful for making function calls to system library
908 functions that support interrupting.  There is no support for pre-Vista
909 Windows.
910 
911 There is a partner testsuite patch which adds several tests for this
912 functionality.
913]
914[Remove ghc-pkg's dependency on haskell98
915Ian Lynagh <igloo@earth.li>**20101013194356
916 Ignore-this: 273fdede3f21e682faa4dc6e61f4e7aa
917]
918[Build haskell98 and haskell2010 with stage2
919Ian Lynagh <igloo@earth.li>**20101013182759
920 Stops us accidentally depending on them
921]
922[Fix warning: Remove unused import
923Ian Lynagh <igloo@earth.li>**20101013141224
924 Ignore-this: ae8891440811f4676d9bb8e721ad22ca
925]
926[Fix warnings
927benl@ouroborus.net**20101013040335
928 Ignore-this: 4e5875827d2840de863000cdb35afb0e
929]
930[RegAlloc: Track slot liveness over jumps in spill cleaner
931benl@ouroborus.net**20101013015414
932 Ignore-this: ccd4a148908b7fbdc6ea76acf527c16b
933]
934[Bump Cabal dep
935Ian Lynagh <igloo@earth.li>**20101012154528
936 Ignore-this: da8539b957b74c7da91efff8ac25872
937]
938[Remove __HASKELL1__, __HASKELL98__, __CONCURRENT_HASKELL__
939Ian Lynagh <igloo@earth.li>**20101012134700
940 Ignore-this: d0db531be58537c52802de1c62694eb1
941 We used to define these CPP symbols, but nothing on hackage uses them
942 and the first 2 are no longer correct (as we support multiple Haskell
943 versions).
944]
945[Follow Cabal changes: Cabal no longer has a docbook userguide
946Ian Lynagh <igloo@earth.li>**20101012130538
947 Ignore-this: bf0b30e465c13f82725d72aec145e939
948 For now we don't build the Cabal userguide, but we should add markdown
949 support so that we can do so.
950]
951[Fix build on Windows: ghc-pkg/Main.hs needs ForeignFunctionInterface
952Ian Lynagh <igloo@earth.li>**20101012112111]
953[Remove unnecessary import
954Ian Lynagh <igloo@earth.li>**20101010222231
955 Ignore-this: 6c7d5cee72837e2af43c2807e10ad602
956]
957[Make "./validate --slow" run the full testsuite
958Ian Lynagh <igloo@earth.li>**20101007004327
959 Ignore-this: 2dee8262c19fd5d5034a186203e20d0f
960]
961[Fix build following haskell98 and -fglasgow-exts changes
962Ian Lynagh <igloo@earth.li>**20101006160656
963 Ignore-this: c56bdb0f01da897b3de75bcab1e2887c
964]
965[Don't automatically link the haskell98 package
966Ian Lynagh <igloo@earth.li>**20101006130235
967 The default language is now Haskell2010, so this was a little odd.
968 Also, --make is now on by default, so this was largely irrelevant.
969]
970[Deprecate -fglasgow-exts
971Ian Lynagh <igloo@earth.li>**20101006124413
972 Ignore-this: 70be67898a633ab31d76f6fca557d55
973]
974[Remove Opt_GADTs and Opt_TypeFamilies from -fglasgow-exts
975Ian Lynagh <igloo@earth.li>**20101006122000
976 Ignore-this: 7153ec880b2b5f9d332ae2f57b97afcc
977 This means most code doesn't get caught by monomorphic local bindings.
978]
979[Fix Trac #4360: omitted case in combineCtLoc
980simonpj@microsoft.com**20101008135747
981 Ignore-this: 834636a97af6469862a822809253db41
982]
983[Beautiful new approach to the skolem-escape check and untouchable
984simonpj@microsoft.com**20101008133751
985 Ignore-this: 33517a772cfdfbf4aa4678609f3dcd71
986 
987 Instead of keeping a *set* of untouchable variables in each
988 implication contraints, we keep a *range* of uniques for the
989 *touchable* variables of an implication.  This are precisely
990 the ones we would call the "existentials" if we were French.
991 
992 It turns out that the code is more efficient, and vastly easier
993 to get right, than the set-based approach.
994 
995 Fixes Trac #4355 among others
996]
997[Do less simplification when doing let-generalisation
998simonpj@microsoft.com**20101008133542
999 Ignore-this: 71366d0de37f10ffba2edc9f3927ddbe
1000 
1001 This fixes Trac #4361.  In a rather delicate way, but
1002 no more delicate than before.  A more remoseless typechecker
1003 would reject #4361 altogether.
1004 
1005 See Note [Avoid unecessary constraint simplification]
1006]
1007[Suppress ambiguity errors if any other errors occur
1008simonpj@microsoft.com**20101008111318
1009 Ignore-this: 40f014265c1ab15fe172baaf76c23c87
1010]
1011[Fix Trac #4361: be more discerning when inferring types
1012simonpj@microsoft.com**20101008111227
1013 Ignore-this: 33656f9a151f494b26b6318b5fcffef
1014 
1015 Note [Avoid unecessary constraint simplification] in TcSimplify
1016]
1017[Float out partial applications
1018Simon Marlow <marlowsd@gmail.com>**20101008092709
1019 Ignore-this: 2dc9d10597c19d0598ef2fc3cf74156d
1020 
1021 This fixes at least one case of performance regression in 7.0, and
1022 is nice win on nofib:
1023 
1024         Program           Size    Allocs   Runtime   Elapsed
1025             Min          +0.3%    -63.0%    -38.5%    -38.7%
1026             Max          +1.2%     +0.2%     +0.9%     +0.9%
1027  Geometric Mean          +0.6%     -3.0%     -6.4%     -6.6%
1028]
1029[Suppress knock-on typechecker errors
1030simonpj@microsoft.com**20101008094348
1031 Ignore-this: 8d125926286a7614fa1ce998e3b26d04
1032 
1033 The error cascade caused puzzling errors in T4093b, and
1034 suppressing some seems like a good plan.  Very few test
1035 outputs change.
1036]
1037[Some refactoring and simplification in TcInteract.occurCheck
1038simonpj@microsoft.com**20101007163500
1039 Ignore-this: d43d09370ab27b8796062e2e98ce7e9
1040]
1041[Comments only
1042simonpj@microsoft.com**20101007130301
1043 Ignore-this: ab46592edd3d24786bbce42c50feb4fd
1044]
1045[Implement auto-specialisation of imported Ids
1046simonpj@microsoft.com**20101007111051
1047 Ignore-this: 45257ff6e9597e4fa4de10b0657e27d6
1048 
1049 This big-ish patch arranges that if an Id 'f' is
1050   * Type-class overloaded
1051        f :: Ord a => [a] -> [a]
1052   * Defined with an INLINABLE pragma
1053        {-# INLINABLE f #-}
1054   * Exported from its defining module 'D'
1055 
1056 then in any module 'U' that imports D
1057 
1058 1. Any call of 'f' at a fixed type will generate
1059    (a) a specialised version of f in U
1060    (b) a RULE that rewrites unspecialised calls to the
1061        specialised on
1062 
1063   e.g. if the call is (f Int dOrdInt xs) then the
1064   specialiser will generate
1065      $sfInt :: [Int] -> [Int]
1066      $sfInt = <code for f, imported from D, specialised>
1067      {-# RULE forall d.  f Int d = $sfInt #-}
1068 
1069 2. In addition, you can give an explicit {-# SPECIALISE -#}
1070    pragma for the imported Id
1071      {-# SPECIALISE f :: [Bool] -> [Bool] #-}
1072    This too generates a local specialised definition,
1073    and the corresponding RULE
1074 
1075 The new RULES are exported from module 'U', so that any module
1076 importing U will see the specialised versions of 'f', and will
1077 not re-specialise them.
1078 
1079 There's a flag -fwarn-auto-orphan that warns you if the auto-generated
1080 RULES are orphan rules. It's not in -Wall, mainly to avoid lots of
1081 error messages with existing packages.
1082 
1083 Main implementation changes
1084 
1085  - A new flag on a CoreRule to say if it was auto-generated.
1086    This is persisted across interface files, so there's a small
1087    change in interface file format.
1088 
1089  - Quite a bit of fiddling with plumbing, to get the
1090    {-# SPECIALISE #-} pragmas for imported Ids.  In particular, a
1091    new field tgc_imp_specs in TcGblEnv, to keep the specialise
1092    pragmas for imported Ids between the typechecker and the desugarer.
1093 
1094  - Some new code (although surprisingly little) in Specialise,
1095    to deal with calls of imported Ids
1096]
1097[Make NameEnv back into type NameEnv a = UniqFM a
1098simonpj@microsoft.com**20101007104638
1099 Ignore-this: 24857c013461788be354520e84f4c286
1100 
1101 I don't think the type distinction of declaring NameEnv with a newtype
1102 (as it was) is really useful to us. Moreover, VarEnv is a UniqFM, and
1103 I do sometimes want to build an envt with Ids and look up with Names.
1104 
1105 This may not be the last word on the subject.
1106]
1107[Improve the rule-matcher
1108simonpj@microsoft.com**20101007103700
1109 Ignore-this: 9de96237dc4b73a43326bd568e34b53b
1110 
1111 Previously it was rejecting the match
1112 
1113   Template: forall s t. map s t
1114   Actual:   map Int t
1115 
1116 which should obviously be fine.  It turns out that this kind of match
1117 comes up when specialising.  By freshening that t we could avoid the
1118 difficulty, but morally the (forall t) binds t and the rule should
1119 be alpha-equivalent regardless of the forall'd variables.
1120 
1121 This patch makes it so, and incidentally makes matching a little
1122 more efficient.  See Note [Eta expansion] in VarEnv.
1123]
1124[Fix Trac #4345: simplifier bug
1125simonpj@microsoft.com**20101007102720
1126 Ignore-this: 261c1c9f094df344ce34de814f8b60c5
1127 
1128 This is another long-standing bug, in which there was a possibility
1129 that a loop-breaker could lose its loop-breaker-hood OccInfo,
1130 and then the simplifer re-simplified the expression. Result, either
1131 non-termination or, in the case of #4345, an unbound identifier.
1132 
1133 The fix is very simple, in Id.transferPolyIdInfo.
1134 See Note [transferPolyIdInfo].
1135]
1136[Avoid redundant simplification
1137simonpj@microsoft.com**20101007095935
1138 Ignore-this: 61bd1a2c508260f558866e6a88c29fa3
1139 
1140 When adding specialisation for imported Ids, I noticed that the
1141 Glorious Simplifier was repeatedly (and fruitlessly) simplifying the
1142 same term.  It turned out to be easy to fix this, because I already
1143 had a flag in the ApplyTo and Select constructors of SimplUtils.SimplCont.
1144 
1145 See Note [Avoid redundant simplification]
1146]
1147[Make the occurrence analyser deal correctly with RULES for imported Ids
1148simonpj@microsoft.com**20101007094100
1149 Ignore-this: 335b1cad013524e42b31e88c0a7a00f6
1150 
1151 This patch fixes a long-standing lurking bug, but it surfaced when I
1152 was adding specialisation for imported Ids.
1153 
1154 See Note [ImpRuleUsage], which explains the issue.   The solution
1155 seems more complicated than the problem really deserves, but I
1156 could not think of a simpler way, so I just bit the bullet and
1157 wrote the code.  Improvements welcome.
1158]
1159[Make warning-free
1160simonpj@microsoft.com**20101007092007
1161 Ignore-this: 4bae0c470a8a1f96d21990d1f3cf1f93
1162]
1163[This is just white-space and layout
1164simonpj@microsoft.com**20101007091618
1165 Ignore-this: 759c0335df70fce32558e967f140803a
1166 
1167 (At least, I don't think there is anything else.)
1168]
1169[Fix an ASSERT failure in FamInstEnv
1170simonpj@microsoft.com**20101007091327
1171 Ignore-this: a8c08ccb7ec2bc65864a674b5441539
1172 
1173 I added a lot of comments too, to explain the preconditions;
1174 esp Note [FamInstEnv]
1175]
1176[Fix a looping bug in the new occur-check code
1177simonpj@microsoft.com**20101007084104
1178 Ignore-this: a02a2deafb9ec986ef1565f4596049ed
1179]
1180[Fix test T4235 with -O
1181simonpj@microsoft.com**20101006155223
1182 Ignore-this: f0fa0fe2f0c493e362d528d71f7a64e1
1183 
1184 The tag2Enum rule wasn't doing the right thing for
1185 enumerations with a phantom type parameter, like
1186    data T a = A | B
1187]
1188[Make warning-free
1189simonpj@microsoft.com**20101006155033
1190 Ignore-this: 221a3c95a6079c6ecc0468996a38b048
1191]
1192[Major bugfixing pass through the type checker
1193dimitris@microsoft.com**20101006152854]
1194[Typechecker performance fixes and flatten skolem bugfixing
1195dimitris@microsoft.com**20101004130200
1196 Ignore-this: 86721ba3f09479c146a0710796b43459
1197]
1198[Performance bug fixes
1199dimitris@microsoft.com**20100923143918]
1200[Fix Trac #4371: matching of view patterns
1201simonpj@microsoft.com**20101006115316
1202 Ignore-this: 494b28b91f1e6392b2f1521cda0e83b1
1203]
1204[Remove unused NoMatchContext construtor
1205simonpj@microsoft.com**20101006115251
1206 Ignore-this: 8985ff1dac51fb652bd65657a630a792
1207]
1208[Refactoring: mainly rename ic_env_tvs to ic_untch
1209simonpj@microsoft.com**20101006102830
1210 Ignore-this: 32999403a3f447e14b59cec7896027ff
1211 
1212 Plus remember to zonk the free_tvs in TcUnify.newImplication
1213]
1214[remove unnecessary/broken definition of mask_
1215Simon Marlow <marlowsd@gmail.com>**20101002195118
1216 Ignore-this: 4cdc9c95d40e01cdfe2d0ac411476603
1217]
1218[-fwarn-tabs: add "Warning" to the message
1219Simon Marlow <marlowsd@gmail.com>**20101002195100
1220 Ignore-this: 589a36daa3426ab51f2fb140e38df6c
1221]
1222[give a better error message in the non-threaded RTS for out-of-range FDs
1223Simon Marlow <marlowsd@gmail.com>**20100929212916
1224 Ignore-this: e94c9f390b8f79d24895a80f9d16c8d9
1225 
1226 # ./aw
1227 aw: file descriptor 1027 out of range for select (0--1024).
1228 Recompile with -threaded to work around this.
1229]
1230[Fix a very rare crash in GHCi
1231Simon Marlow <marlowsd@gmail.com>**20101005144735
1232 Ignore-this: dad1cd08934bae2ba47e72c0c000acfa
1233 When a BCO with a zero-length bitmap was right at the edge of
1234 allocated memory, we were reading a word of non-existent memory.
1235 
1236 This showed up as a segfault in T789(ghci) for me, but the crash was
1237 extremely sensitive and went away with most changes.
1238 
1239 Also, optimised scavenge_large_bitmap a bit while I was in there.
1240]
1241[Using 'stdcall' when it is not supported is only a warning now (#3336)
1242Simon Marlow <marlowsd@gmail.com>**20100924152445
1243 Ignore-this: 66c5903a600a47485a7583535bb38455
1244]
1245[remove unnecessary stg_noForceIO (#3508)
1246Simon Marlow <marlowsd@gmail.com>**20100924150202
1247 Ignore-this: dec52de9cd9da7dcedae12b20691aba9
1248]
1249[Replace an outputStr with putStrLn calls; fixes #4332
1250Ian Lynagh <igloo@earth.li>**20101003125707
1251 Ignore-this: eb4b5f60d9d9d3f7dc203869927b28ba
1252]
1253[make test and fulltest targets in the main Makefile; fixes #4297
1254Ian Lynagh <igloo@earth.li>**20100930224741
1255 You can now run "make test" in the root, and the fast testsuite will be
1256 run with cleaning enabled. It will also put the summary in
1257 testsuite_summary.txt.
1258]
1259[Don't show the loaded packages in ":show packages"; fixes #4300
1260Ian Lynagh <igloo@earth.li>**20100930210128
1261 It's never worked properly, and the information is in ":show linker".
1262]
1263[Handle EXTRA_LIBRARIES when building programs
1264Ian Lynagh <igloo@earth.li>**20100930192552
1265 Ignore-this: 401a26e18d25dcaee010b13eaed8f011
1266 And set hp2ps's EXTRA_LIBRARIES. Based on a patch from Sergei Trofimovich.
1267]
1268[Fix the doc directory on Windows
1269Ian Lynagh <igloo@earth.li>**20100929133328]
1270[Remove an unused import on Windows
1271Ian Lynagh <igloo@earth.li>**20100929000024
1272 Ignore-this: 2899e0e5a47122e637fb5c8aa0df52ab
1273]
1274[Use showCommandForUser when showing tracing commands
1275Ian Lynagh <igloo@earth.li>**20100928235844
1276 Ignore-this: 8a4a9c9f8a8029e708c4297b096b6ef1
1277]
1278[Fix hsc2hs docs: 'gcc' is now the default compiler, not 'ghc'; fixes #4341
1279Ian Lynagh <igloo@earth.li>**20100928201938]
1280[Use an empty signal handler for SIGPIPE instead of SIG_IGN
1281Simon Marlow <marlowsd@gmail.com>**20100925193548
1282 Ignore-this: b4dc5de32740a7c5fd8fe4b3bfb1300f
1283 
1284 This is so that the SIGPIPE handler gets reset to the default
1285 automatically on exec().
1286]
1287[Fix the TH deps
1288Ian Lynagh <igloo@earth.li>**20100925210029
1289 Ignore-this: 32b832301a3625d4ba70f84c5c4f94d2
1290]
1291[Check inplace doesn't exist before we try to create it
1292Ian Lynagh <igloo@earth.li>**20100924191858
1293 This fixes rerunning configure in a tree which already has an inplace
1294 directory. Edward Z Yang ran into this; I guess whether it actually
1295 fails depends on details of your installation, or we'd have run into
1296 it sooner.
1297]
1298[Fix an egregious bug: INLINE pragmas on monomorphic Ids were being ignored
1299simonpj@microsoft.com**20100924155815
1300 Ignore-this: 38c6eec6710a92df7662a55fc5132c15
1301 
1302 I had do to some refactoring to make this work nicely
1303 but now it does. I can't think how this escaped our
1304 attention for so long!
1305]
1306[Eta expand only lambdas that bind a non-dictionary Id
1307simonpj@microsoft.com**20100924155707
1308 Ignore-this: 7cc265eaf6c0bb3fa12eb311d92594ac
1309 
1310 See Note [When to eta expand]. The idea is that dictionary
1311 lambdas are invisible to the user, so we shouldn't eta
1312 expand them.
1313]
1314[Add a comment
1315simonpj@microsoft.com**20100924155620
1316 Ignore-this: de210a1afdd40328824803e1d77b4d7f
1317]
1318[Add a debug print
1319simonpj@microsoft.com**20100924155614
1320 Ignore-this: 1a58b6d297fc77d6ded8eec7ea9f895d
1321]
1322[Just moving comments around
1323simonpj@microsoft.com**20100924155600
1324 Ignore-this: 96635b8e8c9d88b50d82938568152ef8
1325]
1326[use putStrLn instead of Haskeline's outputStrLn
1327Simon Marlow <marlowsd@gmail.com>**20100924133154
1328 Ignore-this: 7581ae11714a9a52e78ba098c3c216b3
1329 use of the latter caused problems for Claus Reinke's macros that
1330 redirect stdout.
1331]
1332[Change "OPTIONS" to "OPTIONS_GHC" in error messages; fixes #4327
1333Ian Lynagh <igloo@earth.li>**20100924120423
1334 Ignore-this: 1697c83a5c346db640c0a2e22c69ff55
1335]
1336[Add deps for TH uses in vector
1337Ian Lynagh <igloo@earth.li>**20100923220244
1338 Ignore-this: 54c3386b1c268821fcdd34b84bc8c6a4
1339]
1340[Bump Cabal dep
1341Ian Lynagh <igloo@earth.li>**20100923143241]
1342[Update Cabal's version number
1343Ian Lynagh <igloo@earth.li>**20100923141719]
1344[Build primitive with stage2
1345Ian Lynagh <igloo@earth.li>**20100923140525
1346 Ignore-this: 110a819b78a57629a7edf1d4facdc191
1347]
1348[Fix the Windows __chkstk build error (missing Linker symbol)
1349Simon Marlow <marlowsd@gmail.com>**20100924113837
1350 Ignore-this: 48f0907bb1bd5eaa0730b94a6bd94ea
1351]
1352[emit a helpful error message for missing DPH packages
1353Simon Marlow <marlowsd@gmail.com>**20100923141957
1354 Ignore-this: 55ff2ee90c94524e023e014243bfe5df
1355]
1356[Fix computation of installed packages
1357simonpj@microsoft.com**20100924084737
1358 Ignore-this: a597d2fa8be5135ba8ead6d2624b3d71
1359 
1360 This is a follow-on to Simon's patch yesterday, developed
1361 with him.  It cleans up the computation of how packages
1362 are installed, and installs the right ones.
1363]
1364[Fix braino in WwLib/Literal patch
1365simonpj@microsoft.com**20100924070914
1366 Ignore-this: f6eb3a42e10f8aa7920de541cdfe76d8
1367]
1368[For now, switch off incomplete-pattern warnings in containers
1369simonpj@microsoft.com**20100923130117
1370 Ignore-this: 7ffa58567f7a33aafe256492999da325
1371 
1372 Put it back on when my patch is applied to the containers repo.
1373 (the one that removes two refuable lambdas)
1374]
1375[Make -funfolding-dict-threshold work properly
1376simonpj@microsoft.com**20100923130032
1377 Ignore-this: 417788f5b09d1d624f6b6371852c80c7
1378 
1379 and increase its default value. This makes overloaded functions
1380 a bit keener to inline.  Which fixes Trac #4321
1381]
1382[Impredicative types is no longer deprecated
1383simonpj@microsoft.com**20100923125910
1384 Ignore-this: 2bbaeb38b5e8424551677c0add627683
1385]
1386[Do not make FunctionalDependencies force MonoLocalBinds
1387simonpj@microsoft.com**20100923125900
1388 Ignore-this: f4ae1fd07c87ec14f60bdfe3863ba7a9
1389]
1390[move CHECKED settings to the right place
1391Simon Marlow <marlowsd@gmail.com>**20100923123558
1392 Ignore-this: e00a0eb5855463cc9b953670b3bbf211
1393]
1394[turn off -Werror for primitive and vector
1395Simon Marlow <marlowsd@gmail.com>**20100923122055
1396 Ignore-this: 54d7b80f3f893385e1c3ef431e2a8a7b
1397]
1398[Add primitive and vector packages for DPH support
1399Simon Marlow <marlowsd@gmail.com>**20100923104542
1400 Ignore-this: c070d015385b0a0797394132dcbb7670
1401 DPH is now using the public vector package instead of its internal
1402 version.
1403 
1404 vector and primitive are not "boot" packages; they aren't required to
1405 build GHC, but they are required to validate (because we include DPH
1406 when validating).
1407 
1408 If you say './darcs-all get --no-dph' then you don't get DPH, vector,
1409 or primitive.
1410]
1411[Refactoring and tidy up in the build system
1412Simon Marlow <marlowsd@gmail.com>**20100923095642
1413 Ignore-this: f7bf3a1fd160149d89b26f464b064fb1
1414 
1415 Instead of the ghc-stage and ghc-stage2-package files in a package, we
1416 now have a list of these in ghc.mk.  There are other similar lists (of
1417 boot-packages and non-installable packages), so this is not too bad,
1418 and is simpler.
1419 
1420 While poking around in the top-level ghc.mk file I spotted various
1421 opportunities to clean up and re-order some of the cruft that has
1422 accumulated over time.
1423]
1424[Allow absent State# RealWorld arguments
1425simonpj@microsoft.com**20100923111356
1426 Ignore-this: c2d57633dec0293ebe6723ea3a4bb5df
1427]
1428[Add notSCCNote, and use it
1429simonpj@microsoft.com**20100923105949
1430 Ignore-this: c8cc758656558a7f366bf784d75f0304
1431 
1432 The point here is that SCCs get in the way of eta
1433 expansion and we must treat them uniformly.
1434]
1435[Remove use of lambda with a refutable pattern
1436simonpj@microsoft.com**20100923105901
1437 Ignore-this: d7d48b94e5744717a838591a1cc79cf0
1438]
1439[Avoid ASSERT black hole
1440simonpj@microsoft.com**20100923105820
1441 Ignore-this: 5419d450871be22c8781ac3f0f40d76a
1442 
1443 When this ASSERT tripped in CoreToStg it tried to print out
1444 too much, which tripped the asssertion again.  Result: an
1445 infinite loop with no output at all.  Hard to debug!
1446]
1447[Rejig the absent-arg stuff for unlifted types
1448simonpj@microsoft.com**20100923105732
1449 Ignore-this: 69daa35816b948b0c4d259c73a5e928e
1450 
1451 This is what was giving the "absent entered" messages
1452 See Note [Absent errors] in WwLib.  We now return a
1453 suitable literal for absent values of unlifted type.
1454]
1455[Remove -fwarn-simple-patterns, and make -fwarn-incomplete-patterns include lambdas
1456simonpj@microsoft.com**20100922133934
1457 Ignore-this: e851a2fb0377e10c28c506f0bf14cc85
1458 
1459 This makes
1460      \(x:xs) -> e
1461 want when you have -fwarn-incomplete-patterns, which is consistent.
1462]
1463[Get rid of non-exhaustive lambda
1464simonpj@microsoft.com**20100922133801
1465 Ignore-this: 748b2d5b43b02b6591b81abe7c105cd6
1466]
1467[Fix an ASSERT failure with profiling
1468simonpj@microsoft.com**20100922133741
1469 Ignore-this: 170b2e94d6ee8cc7444cc4bb515328a0
1470 
1471 The problem arose with this kind of thing
1472 
1473    x = (,) (scc "blah" Nothing)
1474 
1475 Then 'x' is marked NoCafRefs by CoreTidy, becuase it has
1476 arity 1, and doesn't mention any caffy things.
1477 
1478 That in turns means that CorePrep must not float out the
1479 sat binding to give
1480 
1481   sat = scc "blah" Nothing
1482   x = (,) sat
1483 
1484 Rather we must generate
1485 
1486   x = \eta. let sat = scc "blah" Nothing
1487             in (,) sat eta
1488 
1489 URGH! This Caf stuff is such a mess.
1490]
1491[Remove an out of date paragraph from the user guide; fixes #4331
1492Ian Lynagh <igloo@earth.li>**20100922225239]
1493[Fix bindisttest when GhcProfiled = YES
1494Ian Lynagh <igloo@earth.li>**20100921222634
1495 Ignore-this: 47c620fd6bec745e3eb699d9f53441d8
1496]
1497[Fixes for when HADDOCK_DOCS=NO
1498Ian Lynagh <igloo@earth.li>**20100921213916
1499 Ignore-this: e0e069555c6db9d01a8ac70ba4dde591
1500]
1501[Bump version to 7.1
1502Ian Lynagh <igloo@earth.li>**20100921195935
1503 Ignore-this: 4563987e6885d5ef55995ec0fa0d5ae8
1504]
1505[Don't use -march=i686 on powerpc-apple-darwin
1506Ian Lynagh <igloo@earth.li>**20100921193721
1507 Thorikil ran into this when doing a PPC OS X build. We now also don't
1508 use -m32 on PPC/OSX, but I don't think it should be necessary. We can
1509 add it back if it does turn out to be.
1510]
1511[add a simple trace facility to the build system
1512Simon Marlow <marlowsd@gmail.com>**20100921134729
1513 Ignore-this: d23ea2d62a648d0711b4b07d98e1b79f
1514 
1515 saying
1516 
1517   make TRACE=1
1518 
1519 prints most of the macro calls and their arguments.  It's easy to
1520 trace new macros; see rules/trace.mk.
1521]
1522[fix building with extra packages (packages were added to BUILD_DIRS twice)
1523Simon Marlow <marlowsd@gmail.com>**20100921100153
1524 Ignore-this: 4b425dff9777871ad5ba3e05e1d14483
1525 Also add some comments about what extra-packages is doing
1526]
1527[add extra packages to $(EXTRA_PACKAGES), so we avoid installing them by default
1528Simon Marlow <marlowsd@gmail.com>**20100920144307
1529 Ignore-this: 3395825d911a8bf7ba8385518d8b517b
1530]
1531[Fix indexing error in archive loader
1532Ian Lynagh <igloo@earth.li>**20100921121642]
1533[Add some -Dl belches
1534Ian Lynagh <igloo@earth.li>**20100921121624]
1535[Add casts to fix warnings
1536Ian Lynagh <igloo@earth.li>**20100921121714]
1537[Add support for BSD-variant large filenames in .a archives
1538Ian Lynagh <igloo@earth.li>**20100921000451]
1539[Tell Cabal that we're not building GHCi libs if UseArchivesForGhci=YES
1540Ian Lynagh <igloo@earth.li>**20100920230449]
1541["UseArchivesForGhci = YES" on darwin
1542Ian Lynagh <igloo@earth.li>**20100920211538]
1543[Add a dependency that my OS X build has recently started tripping up over
1544Ian Lynagh <igloo@earth.li>**20100920210239]
1545[Add "Use archives for ghci" to --info output
1546Ian Lynagh <igloo@earth.li>**20100920210523]
1547[Implement archive loading for ghci
1548Ian Lynagh <igloo@earth.li>**20100920201620]
1549[Tweak gen_contents_index now dph may not be there
1550Ian Lynagh <igloo@earth.li>**20100920201513]
1551[Filter out the FFI library when loading package in ghci
1552Ian Lynagh <igloo@earth.li>**20100920181032
1553 The FFI GHCi import lib isn't needed as
1554 compiler/ghci/Linker.lhs + rts/Linker.c link the
1555 interpreted references to FFI to the compiled FFI.
1556 We therefore filter it out so that we don't get
1557 duplicate symbol errors.
1558]
1559[Loosen the conditions for -XUndecidableInstances; fixes Trac #4200
1560simonpj@microsoft.com**20100919162623
1561 Ignore-this: 2f4323e278b1ce9250549727ffd0aa1b
1562]
1563[Further improvements in error messages
1564simonpj@microsoft.com**20100919153355
1565 Ignore-this: b6fa0b11ae893df1a3ca68f78e427fa
1566]
1567[Add a flag -fwarn-missing-local-sigs, and improve -fwarn-mising-signatures
1568simonpj@microsoft.com**20100919153327
1569 Ignore-this: fda8dfca450054ea692be0ee30b01885
1570 
1571 The new flag prints out a warning if you have a local,
1572 polymorphic binding that lacks a type signature. It's meant
1573 to help with the transition to the new typechecker, which
1574 discourages local let-generalisation.
1575 
1576 At the same time I moved the missing-signature code to TcHsSyn,
1577 where it takes place as part of zonking.  That way the
1578 types are reported after all typechecking is complete,
1579 thereby fixing Trac #3696.  (It's even more important for
1580 local bindings, which is why I made the change.)
1581]
1582[Include the "stupid theta" in the type of $con2tag
1583simonpj@microsoft.com**20100919152201
1584 Ignore-this: d95fae78a0e66f48bbd5862573a11f4d
1585]
1586[Add a release note about the typechecker
1587Ian Lynagh <igloo@earth.li>**20100919132927]
1588[Enable shared libs on OpenBSD
1589Matthias Kilian <kili@outback.escape.de>**20100918205040
1590 Ignore-this: 729dd7ac0bba5d758f43bc31b541896
1591]
1592[Add separate functions for querying DynFlag and ExtensionFlag options
1593Ian Lynagh <igloo@earth.li>**20100918163815
1594 and remove the temporary DOpt class workaround.
1595]
1596[Fix mkUserGuidePart deps
1597Ian Lynagh <igloo@earth.li>**20100918145904
1598 We need to directly depend on the stage1 libs. The stage1 compiler lib
1599 doesn't depend on them.
1600]
1601[Fix build on cygwin: Normalise slashes in .depend files to be /
1602Ian Lynagh <igloo@earth.li>**20100918132328
1603 Ignore-this: 664f5ef4a41a4461eb34fe2ca7f2729a
1604]
1605[extra packages info is now read from packages file
1606Ian Lynagh <igloo@earth.li>**20100917224409
1607 rather than being repeated in the build system
1608]
1609[Tweak darcs-all
1610Ian Lynagh <igloo@earth.li>**20100917194435]
1611[Bump dependencies
1612Ian Lynagh <igloo@earth.li>**20100917183609]
1613[Library release notes for 7.0.1
1614Ian Lynagh <igloo@earth.li>**20100917174850]
1615[Fix overriding of implicit parameters in the solver
1616simonpj@microsoft.com**20100917140403
1617 Ignore-this: af76732309c7e2ca6b04f49327e9c14b
1618]
1619[Minor type printing amomaly
1620simonpj@microsoft.com**20100917140204
1621 Ignore-this: c90cb2e51421b4543a827e096051772e
1622]
1623[Spaces only
1624simonpj@microsoft.com**20100917140156
1625 Ignore-this: 7e34479502f7cb87d762355e40cbd012
1626]
1627[Minor refactoring
1628simonpj@microsoft.com**20100917140150
1629 Ignore-this: 6c0648b949b91b7e2f23c136b124faf2
1630]
1631[Add types of implicit parameters as untouchable
1632simonpj@microsoft.com**20100917140138
1633 Ignore-this: ba80740a557a9ba062dc7756e2320d17
1634 
1635 This is a tricky point:
1636    see Note [Implicit parameter untouchables]
1637]
1638[Better pretty printing of implicit parameters
1639simonpj@microsoft.com**20100917140054
1640 Ignore-this: 867dd67818a5bd687b2b6a1b59e15775
1641]
1642[Yet more error message improvement
1643simonpj@microsoft.com**20100917121206
1644 Ignore-this: 647fe8129d1d39d81e8249debd8df94e
1645]
1646[More error message wibbles
1647simonpj@microsoft.com**20100917094721
1648 Ignore-this: 8ec2f150b96b26af2e9ab7ac2b371fc7
1649]
1650[More error refactoring
1651simonpj@microsoft.com**20100917092834
1652 Ignore-this: 2d570ac0b9cc11305ddd33d093d11324
1653]
1654[Refactor type errors a bit
1655simonpj@microsoft.com**20100917080726
1656 Ignore-this: 33da4549373f585064e2ee22b50ad6ac
1657 
1658 Improves kind error messages in paticular
1659]
1660[Fix a very subtle shadowing bug in optCoercion
1661simonpj@microsoft.com**20100916170452
1662 Ignore-this: 9041cfb3c93e27a5e644e57815313aae
1663 
1664 See Note [Subtle shadowing in coercions]
1665 
1666 This is what was going wrong in Trac 4160.
1667]
1668[Fix bad error in tyVarsOfType
1669simonpj@microsoft.com**20100916170348
1670 Ignore-this: 67c8ce96a668cf6e3a38b82c893bcd81
1671 
1672 We weren't gathering the type variables free in the kind
1673 of a coercion binder!
1674]
1675[More assertions
1676simonpj@microsoft.com**20100916170310
1677 Ignore-this: 7fdcb53c99d791621a3d7e01ef454404
1678]
1679[Add more location info in CoreLint
1680simonpj@microsoft.com**20100916170229
1681 Ignore-this: 6558bab544b4f30189e0430668db87c3
1682]
1683[Print coercion variables as such (debugging change only)
1684simonpj@microsoft.com**20100916165944
1685 Ignore-this: c6d2001c1d8279a2288cb63bc339577d
1686]
1687[Remove pprTrace
1688simonpj@microsoft.com**20100915225935
1689 Ignore-this: 28185bbfa9732386f3c0f3eb4781a637
1690]
1691[Remove dead code dealing with type refinement
1692simonpj@microsoft.com**20100915223230
1693 Ignore-this: 62824b5c2ec1077c7642163352559621
1694]
1695[Use mkAppTy
1696simonpj@microsoft.com**20100915223205
1697 Ignore-this: e79e087b6a49219e9088846a1253a153
1698 
1699 Using AppTy in CoreLint was giving a bogus Lint failure
1700]
1701[Comments only
1702simonpj@microsoft.com**20100915221253
1703 Ignore-this: 3a45ea614188ccbb4a462de5cac96eda
1704]
1705[Extend eta reduction to work with casted arguments
1706simonpj@microsoft.com**20100915221229
1707 Ignore-this: 24b103dcdf70331211507af929789f86
1708 
1709 See Trac #4201, and
1710 Note [Eta reduction with casted arguments]
1711 
1712 Thanks to Louis Wasserman for suggesting this, and
1713 implementing an early version of the patch
1714]
1715[Allow "INLINEABLE" as a synonym
1716simonpj@microsoft.com**20100915154249
1717 Ignore-this: f41f80cb769e9acd5b463b170df698d0
1718]
1719[Documentation for INLINABLE
1720simonpj@microsoft.com**20100915154235
1721 Ignore-this: f942c02bcadc0d2d2f05b9369f93e280
1722]
1723[Implement TH reification of instances (Trac #1835)
1724simonpj@microsoft.com**20100915151242
1725 Ignore-this: 97dfa83db7da8f6cbd1b96801a57f8c5
1726 
1727 Accompanying patch for template-haskell package is reqd
1728]
1729[errno corresponding to ERROR_NO_DATA should be EPIPE (non-threaded RTS)
1730Simon Marlow <marlowsd@gmail.com>**20100915141809
1731 Ignore-this: 709c7280fbaa762e7071fb8796e8c01e
1732]
1733[Windows: use a thread-local variable for myTask()
1734Simon Marlow <marlowsd@gmail.com>**20100915120627
1735 Ignore-this: 13ffa4f19ebd319fe672af53af8d0b9a
1736 Which entailed fixing an incorrect #ifdef in Task.c
1737]
1738[Fix typo
1739Ian Lynagh <igloo@earth.li>**20100915140814]
1740[Add quotes in error message
1741simonpj@microsoft.com**20100915144724
1742 Ignore-this: c5158047c0aa41947a79e4c8edbe54f4
1743]
1744[Fix isDefaultInlinePragma
1745simonpj@microsoft.com**20100915144710
1746 Ignore-this: c9addf6bf811b23dc12603cf8521aa6c
1747]
1748[Implement INLINABLE pragma
1749simonpj@microsoft.com**20100915124442
1750 Ignore-this: 80a4ab2c2d65b27868dc9b2e954d6c6f
1751 
1752 Implements Trac #4299.  Documentation to come.
1753]
1754[Less voluminous error when derived code doesn't typecheck
1755simonpj@microsoft.com**20100915072301
1756 Ignore-this: eca7871dcc50c1070a0b530711adea27
1757]
1758[Improve pretty-printing of family instances
1759simonpj@microsoft.com**20100915123219
1760 Ignore-this: 25ec6bcc7e8a7f7c303b38ca201db90e
1761 
1762 Fixed Trac #4246
1763]
1764[Fix Trac #4240: -ddump-minimal-imports
1765simonpj@microsoft.com**20100915121937
1766 Ignore-this: ab85057cb829a42ea44a92f7b4af24a3
1767 
1768 See Note [Partial export] for the details.
1769 I also fixed one egregious bug that was just
1770 waiting to bite: we were using loadSysInterface
1771 instead of loadSrcInterface.
1772]
1773[Comments only
1774simonpj@microsoft.com**20100915105707
1775 Ignore-this: ab3a5f16f8260b7b8570e748bf97998a
1776]
1777[implement setThreadAffinity on Windows (#1741)
1778Simon Marlow <marlowsd@gmail.com>**20100914155844
1779 Ignore-this: a14c7b4ef812007042342d0a25478f0b
1780]
1781[COFF: cope with new debug sections in gcc 4.x (fixes ghciprog004)
1782Simon Marlow <marlowsd@gmail.com>**20100914153026
1783 Ignore-this: f340e40a2b0390836bc61bba144a04ed
1784 Also updated the object file parser to properly handle the overflow
1785 case for section names longer than 8 chars.
1786]
1787[eliminate clutter from make output
1788Simon Marlow <marlowsd@gmail.com>**20100915105712
1789 Ignore-this: bfa4480dd239dda2a02ac391b6a9219c
1790]
1791[rts_isProfiled should be a visible API (fixes T2615(dyn))
1792Simon Marlow <marlowsd@gmail.com>**20100915083941
1793 Ignore-this: b8ac09bb9d1a929bf45c6122f8485561
1794]
1795[Fix the "lost due to fragmentation" calculation
1796Simon Marlow <marlowsd@gmail.com>**20100914145945
1797 Ignore-this: cdffcc9f3061c3a33da5171be111fc43
1798 It was counting the space used by block descriptors as "lost"
1799]
1800[fix +RTS -S output: use peak_mblocks_allocated, now that mblocks can be freed
1801Simon Marlow <marlowsd@gmail.com>**20100914135030
1802 Ignore-this: 65d21e5f86d3ab6ab4d6c255180b6968
1803]
1804[Fix egregious bug in deeplyInstantiate
1805simonpj@microsoft.com**20100915070325
1806 Ignore-this: 22ede973038877af2673339aaf5de6cf
1807 
1808 This resulted in an infinite loop in applyTypeToArgs, in syb
1809]
1810[Improve HsSyn pretty printing
1811simonpj@microsoft.com**20100915070255
1812 Ignore-this: 7c8e2d86a482453c7e69e22bc31cb03f
1813]
1814[Remove (most of) the FiniteMap wrapper
1815Ian Lynagh <igloo@earth.li>**20100914201703
1816 We still have
1817     insertList, insertListWith, deleteList
1818 which aren't in Data.Map, and
1819     foldRightWithKey
1820 which works around the fold(r)WithKey addition and deprecation.
1821]
1822[Improve ASSERT
1823simonpj@microsoft.com**20100914113900
1824 Ignore-this: dbc0363be5924f543316e77f7d18dd77
1825]
1826[Comment on what an "enumeration" type is
1827simonpj@microsoft.com**20100914113850
1828 Ignore-this: c09c8591e3140f305d55fbf945adbf95
1829]
1830[Make absent-arg wrappers work for unlifted types (fix Trac #4306)
1831simonpj@microsoft.com**20100914113827
1832 Ignore-this: 1945e56779329e8b79780403710aba98
1833 
1834 Previously we were simply passing arguments of unlifted
1835 type to a wrapper, even if they were absent, which was
1836 stupid.
1837 
1838 See Note [Absent error Id] in WwLib.
1839]
1840[Comments only
1841simonpj@microsoft.com**20100914113641
1842 Ignore-this: 3191ce856c9b5d9700cedc9b149b8097
1843]
1844[Move error-ids to MkCore (from PrelRules)
1845simonpj@microsoft.com**20100914113635
1846 Ignore-this: c3d820db62ba6139dd7c96bf97e51bb5
1847 
1848 and adjust imports accordingly
1849]
1850[More wibbles to deriving error messages
1851simonpj@microsoft.com**20100914113523
1852 Ignore-this: bd2df662644961138fa209aec843a2aa
1853]
1854[Fix getThreadCPUTime()
1855Simon Marlow <marlowsd@gmail.com>**20100913153838
1856 Ignore-this: 950e048a5724086534b74c609c7d5ed
1857 ever since the patch "Check with sysconf _POSIX_THREAD_CPUTIME", it
1858 has been returning incorrect results, because the sysconf variable to
1859 check should have been _SC_THREAD_CPUTIME, not _POSIX_THREAD_CPUTIME.
1860]
1861[filter out the gcc-lib directory from the rts package's library-dirs
1862Simon Marlow <marlowsd@gmail.com>**20100913101259
1863 Ignore-this: 46dc1dccbfee8a65f9243e125eee117f
1864 fixes problems when building with GHC 6.10 on Windows
1865]
1866[Don't include GC time in heap profiles (#4225)
1867Simon Marlow <marlowsd@gmail.com>**20100913133852
1868 Ignore-this: 68ac48b004b311384b5996c6b33ba5cc
1869]
1870[Use clock_gettime (if available) to measure the process CPU time
1871Simon Marlow <marlowsd@gmail.com>**20100913133818
1872 Ignore-this: 8c9300df9b929bfc1db4713c9b6065b3
1873 This is much more accurate than getrusage, which was giving misleading
1874 results when trying to time very quick operations like a minor GC.
1875]
1876[make stg_arg_bitmaps public, and available via the GHCi linker (#3672)
1877Simon Marlow <marlowsd@gmail.com>**20100913105235
1878 Ignore-this: e18efd0bd77c521e5530fb59e93b5a42
1879]
1880[fix typo
1881Simon Marlow <marlowsd@gmail.com>**20100913105100
1882 Ignore-this: 6049eea21208864203b2d79db2edd143
1883]
1884[Update release notes and docs with LLVM info.
1885David Terei <davidterei@gmail.com>**20100914072135
1886 Ignore-this: 5b3d0e5c9d5da98ed6ae9c2e8e1f6f30
1887]
1888[Remove defaultExtensionFlags
1889Ian Lynagh <igloo@earth.li>**20100913165949
1890 The default should do into languageExtensions instead
1891]
1892[Improve crash message
1893simonpj@microsoft.com**20100913170407
1894 Ignore-this: 5c26a9979f18be8cd12cea823c9f4b5a
1895]
1896[Fix Trac #4302, plus a little refactoring
1897simonpj@microsoft.com**20100913170355
1898 Ignore-this: cf6886b587aa0e8d723362183625d946
1899]
1900[Fix build with 6.10
1901Ian Lynagh <igloo@earth.li>**20100913160048]
1902[Haddock fixes
1903simonpj@microsoft.com**20100913120510
1904 Ignore-this: f3157d6969f10d4cbd593000a477138b
1905]
1906[Remove two old junk files
1907simonpj@microsoft.com**20100913103426
1908 Ignore-this: ed7af5ef1b9592178909a8d876345302
1909]
1910[Super-monster patch implementing the new typechecker -- at last
1911simonpj@microsoft.com**20100913095048
1912 Ignore-this: 14d14a1e4d7a414f5ae8d9d89d1c6a4b
1913 
1914 This major patch implements the new OutsideIn constraint solving
1915 algorithm in the typecheker, following our JFP paper "Modular type
1916 inference with local assumptions". 
1917 
1918 Done with major help from Dimitrios Vytiniotis and Brent Yorgey.
1919 
1920]
1921[Fix simplifier statistics
1922simonpj@microsoft.com**20100909085441
1923 Ignore-this: 48e383655aafc912dea15c4d94382863
1924]
1925[Trace output
1926simonpj@microsoft.com**20100908170056
1927 Ignore-this: 4b67fa4b310fbf0a16b852686d2d3294
1928]
1929[Better debug output
1930simonpj@microsoft.com**20100908170047
1931 Ignore-this: 410cef00616dda7c0c162f65216e8ca3
1932]
1933[Add Outputable instance for OccEncl
1934simonpj@microsoft.com**20100908150510
1935 Ignore-this: 6362ef9028287d84f070eaf8963c1bfc
1936]
1937[Better simplifier counting
1938simonpj@microsoft.com**20100907214840
1939 Ignore-this: 9d4722703f8f47447e86a28c8c50e0ea
1940]
1941[Put liftStringName into the known-key names
1942simonpj@microsoft.com**20100906112415
1943 Ignore-this: 287064d14ff484da1a6dea6924bc9235
1944]
1945[Deprecate NoRelaxedPolyRec
1946simonpj@microsoft.com**20100903234519
1947 Ignore-this: 607217e77f6bc1b91bf57dfd8dd2b967
1948]
1949[Buglet in Core Lint
1950simonpj@microsoft.com**20100903234457
1951 Ignore-this: 277535d51b396d3b4b0265a0939c2d4
1952]
1953[Give seqId the right type
1954simonpj@microsoft.com**20100903093556
1955 Ignore-this: d1fc7a73dea160614a8d4ddc930f99cd
1956]
1957[Remove dead code
1958simonpj@microsoft.com**20100903093548
1959 Ignore-this: 92cc3f7651445aa349ee7f114d3ec758
1960]
1961[Comments and layout
1962simonpj@microsoft.com**20100903093502
1963 Ignore-this: 9987d1409e654992c1cb1be35cb87728
1964]
1965[Remove checkFreeness; no longer needed
1966simonpj@microsoft.com**20100902233227
1967 Ignore-this: c96a12ac9794290aa30402317d88c095
1968]
1969[Assert
1970simonpj@microsoft.com**20100902073642
1971 Ignore-this: 4be1ab2f6096665ae5ec7fdd1f025a67
1972]
1973[Add aserts
1974simonpj@microsoft.com**20100902073211
1975 Ignore-this: e1409441217fd070c5a7f9ee4cca99ab
1976]
1977[Wibbles
1978simonpj@microsoft.com**20100831113540
1979 Ignore-this: 903811ab493a7b560a62eb86fcf3ee25
1980]
1981[Wibble to allow phantom types in Enum
1982simonpj@microsoft.com**20100825112711
1983 Ignore-this: fdef1c50d92b4a3d46bbe4cbfd8a83ea
1984]
1985[Add HsCoreTy to HsType
1986simonpj@microsoft.com**20100824141845
1987 Ignore-this: 4ca742b099f9cc90af3167f1012dbba6
1988 
1989 The main thing here is to allow us to provide type
1990 signatures for 'deriving' bindings without pain.
1991]
1992[Comments
1993simonpj@microsoft.com**20100823223654
1994 Ignore-this: dd412a55839430c436902d8699d6900b
1995]
1996[Wibbles to error message
1997simonpj@microsoft.com**20100823163308
1998 Ignore-this: 4d6cd8e613762dca8135c2e3b09264ec
1999]
2000[Correct type signatures
2001simonpj@microsoft.com**20100823153045
2002 Ignore-this: 42942309221a443258246098f9c0a13b
2003]
2004[Add missing signatures
2005simonpj@microsoft.com**20100823112413
2006 Ignore-this: 8ee1ce40456306de469938c02df4fed5
2007]
2008[Add type signatures in "deriving" bindings
2009simonpj@microsoft.com**20100820234230
2010 Ignore-this: 4726b28968cf65ec16cb65b7e0e7303e
2011]
2012[Minor
2013dimitris@microsoft.com**20100820131021]
2014[Be a bit less aggressive in mark-many inside a cast
2015simonpj@microsoft.com**20100819104804
2016 Ignore-this: 3fd48fe7647ec7a58c2032cd86ca4d4f
2017]
2018[Wibble
2019simonpj@microsoft.com**20100818185738
2020 Ignore-this: d5c939311377c0d0c7244aa339193315
2021]
2022[Pretty printing change
2023simonpj@microsoft.com**20100818065436
2024 Ignore-this: 4f7e70976dbe52f95effb3e634dfef5d
2025]
2026[Remember to zonk FlatSkols!
2027simonpj@microsoft.com**20100811143555
2028 Ignore-this: 84f7f9dbda97f561a918c69308ddef9a
2029]
2030[De-polymorphise
2031simonpj@microsoft.com**20100730151217
2032 Ignore-this: a9304487b983e517a9083fd697f77576
2033]
2034[Work around missing type signature in Happy
2035simonpj@microsoft.com**20100730122405
2036 Ignore-this: 7f241a655d93c5ad7763a7ffe8db0c7a
2037 
2038 Happy generates
2039 
2040       notHappyAtAll = error "Blah"
2041 
2042 without a type signature, and currently the new
2043 typechecker doesn't generalise it.  This patch
2044 says "no monomorphism restriction" which makes it
2045 generalise again.
2046 
2047 Better would be to add a type sig to Happy's template
2048]
2049[Add two local type signatures
2050simonpj@microsoft.com**20100729152611
2051 Ignore-this: afa99bcc515469aa0990d44d8c18a9e6
2052]
2053[Second test from Simon's laptop
2054simonpj@microsoft.com**20100729091703
2055 Ignore-this: 4dc64cadae314a5a1b05cc5326918a83
2056]
2057[Test commit from Simon's laptop
2058simonpj@microsoft.com**20100729091344
2059 Ignore-this: 109eff835cc19e9f93799d12f09b0ba7
2060]
2061[Add OutsideIn flag
2062simonpj@microsoft.com**20100728075525
2063 Ignore-this: 69c2f5c3a15fa653f6da80598aa8d74d
2064]
2065[Layout only
2066simonpj@microsoft.com**20100727141539
2067 Ignore-this: 1a58a8fe80ba8bced18ae81a2efb9495
2068]
2069[Improvement to SimplUtils.mkLam
2070simonpj@microsoft.com**20100727131659
2071 Ignore-this: 739beaefa79baa7e0ebeb5b2b6d1ea91
2072]
2073[Give the correct kind to unsafeCoerce#
2074simonpj@microsoft.com**20100727131538
2075 Ignore-this: 6b787de3b398c6d7a61fa04fccd15fd6
2076]
2077[Suppress warnings about recursive INLINE in output of desugarer
2078simonpj@microsoft.com**20100727094549
2079 Ignore-this: a361f7238c0fcba526d46326722c42e
2080]
2081[Rename CorePrep.tryEtaReduce to tryEtaReducePrep
2082simonpj@microsoft.com**20100726231253
2083 Ignore-this: 4375ddace205745244ba224ae012252
2084 
2085 This avoids the name clash with the similar but
2086 not identical CoreUtils.tryEtaReduce
2087]
2088[Add a trace message
2089simonpj@microsoft.com**20100719211111
2090 Ignore-this: b5daebe46e50c8cf28cc693f84bbf099
2091]
2092[Don't use RelaxedPolyRec in the compiler; it's built in now
2093simonpj@microsoft.com**20100719170441
2094 Ignore-this: a2e4489cdf63478e46282a421ee7aec3
2095]
2096[Remove duplicated #defines for FreeBSD
2097Matthias Kilian <kili@outback.escape.de>**20100912181518
2098 Ignore-this: d16214fef8635c7c9ef4edec4e8e7896
2099]
2100[Don't fail with absolute silence
2101Matthias Kilian <kili@outback.escape.de>**20100912150506
2102 Ignore-this: 479e2321f39b263fa2d9f80491e5e9f7
2103]
2104[Add a release note: "-dynload wrapper" removed
2105Ian Lynagh <igloo@earth.li>**20100911195809]
2106[put back the conversion of warn-lazy-unlifted-bindings into an error until 7.2
2107Ian Lynagh <igloo@earth.li>**20100911193434
2108 I think we'll currently still have too many people with old versions of
2109 alex/happy to want to make this an error now.
2110]
2111[6.14 -> 7.0
2112Ian Lynagh <igloo@earth.li>**20100911192837]
2113[Add a couple more release notes
2114Ian Lynagh <igloo@earth.li>**20100911162059]
2115[Document -dsuppress-module-prefixes
2116Ian Lynagh <igloo@earth.li>**20100911162005]
2117[Enable -fregs-graph with -O2; fixes #2790
2118Ian Lynagh <igloo@earth.li>**20100910191301]
2119[Remove unused code
2120Ian Lynagh <igloo@earth.li>**20100909170207]
2121[Fix warnings
2122Ian Lynagh <igloo@earth.li>**20100909154348]
2123[Fix warnings
2124Ian Lynagh <igloo@earth.li>**20100909150957]
2125[Remove context completion
2126lykahb@gmail.com**20100901160153
2127 Ignore-this: dc61b259dcb7063f0c76f56788b5d2af
2128 Now completion suggests to remove only modules added to context before.
2129]
2130[avoid Foreign.unsafePerformIO
2131Ross Paterson <ross@soi.city.ac.uk>**20100909125531
2132 Ignore-this: 5cabeae4cffec8fc17ef7c0cabbea22a
2133]
2134[updates to the release notes
2135Simon Marlow <marlowsd@gmail.com>**20100909111450
2136 Ignore-this: a4d25ad8815c305b7e0f21fd4f6ee37b
2137]
2138[newAlignedPinnedByteArray#: avoid allocating an extra word sometimes
2139Simon Marlow <marlowsd@gmail.com>**20100909110805
2140 Ignore-this: 996a3c0460068ab2835b4920905b3e75
2141]
2142[Finish breaking up vectoriser utils
2143benl@ouroborus.net**20100909061311
2144 Ignore-this: 217fe1d58a3e8bb13200bcb81353a416
2145]
2146[Move VectType module to Vectorise tree
2147benl@ouroborus.net**20100909042451
2148 Ignore-this: 5af8cf394d4835911259ca3ffb6774c5
2149]
2150[Sort all the PADict/PData/PRDict/PRepr stuff into their own modules
2151benl@ouroborus.net**20100909035147
2152 Ignore-this: 53436329773347cad793adbd83e90a9e
2153]
2154[Break out Repr and PADict stuff for vectorisation of ADTs to their own modules
2155benl@ouroborus.net**20100909025759
2156 Ignore-this: d2b7d2f79332eda13416449742f7cf1c
2157]
2158[Break out conversion functions to own module
2159benl@ouroborus.net**20100909023332
2160 Ignore-this: 613f2666b6ca7f2f8876fcc1e4a59593
2161]
2162[Comments and formatting only
2163benl@ouroborus.net**20100909022117
2164 Ignore-this: c8e30139d730669e5db44f0ef491a588
2165]
2166[Remove "-dynload wrapper"; fixes trac #4275
2167Ian Lynagh <igloo@earth.li>**20100908213251]
2168[Don't set visibility on Windows
2169Ian Lynagh <igloo@earth.li>**20100905122442
2170 With gcc 4.5.0-1, using visibility hidden gives:
2171     error: visibility attribute not supported in this configuration; ignored
2172]
2173[Fix warnings on Windows
2174Ian Lynagh <igloo@earth.li>**20100905111201
2175 Ignore-this: c5cce63bb1e0c7a27271bed78d25fbc5
2176]
2177[Fix gcc wrapper for new mingw binaries
2178Ian Lynagh <igloo@earth.li>**20100905001807
2179 Ignore-this: f6acc8c911055ffce632bac138ccc939
2180]
2181[Don't pass our gcc options to stage0 ghc's gcc; they may not be suitable
2182Ian Lynagh <igloo@earth.li>**20100905103129]
2183[Update intree-mingw creation
2184Ian Lynagh <igloo@earth.li>**20100904225559]
2185[Update commands to build in-tree mingw
2186Ian Lynagh <igloo@earth.li>**20100904215112]
2187[Break out hoisting utils into their own module
2188benl@ouroborus.net**20100908074102
2189 Ignore-this: e3ba4ed0252a2def1ed88a9e14c58fea
2190]
2191[Break out closure utils into own module
2192benl@ouroborus.net**20100908072040
2193 Ignore-this: 216172b046ff101cf31a1753667a5383
2194]
2195[Move VectVar module to Vectorise tree
2196benl@ouroborus.net**20100908065904
2197 Ignore-this: 1fba5333d29927dba4275381e1a7f315
2198]
2199[Break out vectorisation of expressions into own module
2200benl@ouroborus.net**20100908065128
2201 Ignore-this: 6a952b80fb024b5291f166477eb1976
2202]
2203[Break out TyCon classifier into own module
2204benl@ouroborus.net**20100908063111
2205 Ignore-this: da754c4ef6960b4e152ea1bf8c04ab6f
2206]
2207[Break out vectorisation of TyConDecls into own module
2208benl@ouroborus.net**20100908052004
2209 Ignore-this: c0ab4fb2a05ca396efe348b384db1ebf
2210]
2211[Break out type vectorisation into own module
2212benl@ouroborus.net**20100907110311
2213 Ignore-this: 67bd70a21d16468daf68dd3ec1ff7d62
2214]
2215[Tidy up the ArchHasAdjustorSupport definition
2216Ian Lynagh <igloo@earth.li>**20100904144234]
2217[ppc: switch handling of 'foreign import wrapper' (FIW) to libffi
2218Sergei Trofimovich <slyfox@community.haskell.org>**20100829192859
2219 Ignore-this: 662ea926681ebea0759e2a04a38e82b7
2220 
2221 Joseph Jezak reported darcs-2.4.4 SIGSEGV in interactive mode in ghc-6.12.3.
2222 So I've concluded ppc also has rotten native adjustor. I don't have hardware
2223 to verify the patch (ticket #3516 should help to test it), but I think it will
2224 help (as similar patch helped for ia64 and ppc64).
2225]
2226[Binary no longer has unusable UNPACK pragmas, so no need to turn of -Werror
2227Ian Lynagh <igloo@earth.li>**20100904133339]
2228[Don't haddock packages that we aren't going to install
2229Ian Lynagh <igloo@earth.li>**20100903231921]
2230[Give haddock per-package source entity paths; fixes #3810
2231Ian Lynagh <igloo@earth.li>**20100903221335]
2232[update for containers-0.4
2233Simon Marlow <marlowsd@gmail.com>**20100903105131
2234 Ignore-this: 556eac0e4926c9b8af6b66d7b069302c
2235]
2236[Fix for nursery resizing: the first block's back pointer should be NULL
2237Simon Marlow <marlowsd@gmail.com>**20100827102818
2238 Ignore-this: fb68938e3f1e291e3c9e5e8047f9dcd2
2239 I'm not sure if this could lead to a crash or not, but it upsets +RTS -DS
2240 Might be related to #4265
2241]
2242[Add some -no-user-package-conf flags
2243Ian Lynagh <igloo@earth.li>**20100902224726
2244 Stops user-installed packages breaking the build
2245]
2246[Fix warnings: Remove unused imports
2247Ian Lynagh <igloo@earth.li>**20100902204342]
2248[Finish breaking up VectBuiltIn and VectMonad, and add comments
2249benl@ouroborus.net**20100831100724
2250 Ignore-this: 65604f3d22d03433abc12f10be40050d
2251]
2252[Fix warnings
2253benl@ouroborus.net**20100830083746
2254 Ignore-this: 2a0e000985f694582a6f9a9261ff2739
2255]
2256[Break up vectoriser builtins module
2257benl@ouroborus.net**20100830070900
2258 Ignore-this: b86bd36a7875abdcf16763902ba2e637
2259]
2260[Move VectCore to Vectorise tree
2261benl@ouroborus.net**20100830053415
2262 Ignore-this: d5763ca6424285b39a58c7792f4a84a1
2263]
2264[Split out vectoriser environments into own module
2265benl@ouroborus.net**20100830050252
2266 Ignore-this: 5319111c74831394d2c29b9aedf5a766
2267]
2268[Comments and formatting to vectoriser, and split out varish stuff into own module
2269benl@ouroborus.net**20100830042722
2270 Ignore-this: d3f0c98ed8124dd0fca9a2ccea3e15fd
2271]
2272[Fix warnings
2273benl@ouroborus.net**20100830040340
2274 Ignore-this: d6cfad803ad4617e7fdaa62e4a895282
2275]
2276[Fix warning about multiply exported name
2277benl@ouroborus.net**20100830035243
2278 Ignore-this: 27ce2c1d22d9f99929d16a426343044e
2279]
2280[Vectorisation of method types
2281benl@ouroborus.net**20100830032941
2282 Ignore-this: 75614571d5c246a4906edb3b39ab1e0b
2283]
2284[Comments and formatting to vectoriser
2285benl@ouroborus.net**20100830032516
2286 Ignore-this: fe665b77108501c7960d858be3290761
2287]
2288[Implement -dsuppress-module-prefixes
2289benl@ouroborus.net**20100830032428
2290 Ignore-this: 2bb8bad9c60ef9044132bba118010687
2291]
2292[Whitespace only
2293benl@ouroborus.net**20100527045629
2294 Ignore-this: 4c160dfa77727e659817b6af9c84684a
2295]
2296[Disambiguate a function name
2297Ian Lynagh <igloo@earth.li>**20100828225827]
2298[users_guide.xml is now generated
2299Ian Lynagh <igloo@earth.li>**20100828225751]
2300[Pass more -pgm flags in the ghc wrapper; fixes #3863
2301Ian Lynagh <igloo@earth.li>**20100827204537]
2302[Add a new-IO manager release note
2303Ian Lynagh <igloo@earth.li>**20100827171616]
2304[Merge a duplicate release note
2305Ian Lynagh <igloo@earth.li>**20100827171511]
2306[Typo, spotted by Johan Tibell
2307Ian Lynagh <igloo@earth.li>**20100827153914]
2308[First pass at 6.14.1 release notes
2309Ian Lynagh <igloo@earth.li>**20100826220811]
2310[Fix typo
2311Ian Lynagh <igloo@earth.li>**20100824201330]
2312[FIX BUILD: add rts_isProfiled to the symbol table
2313Simon Marlow <marlowsd@gmail.com>**20100826094319
2314 Ignore-this: 9536ddb0a94721c8dec03a2a981cfa83
2315]
2316[Fix the DPH package cleaning/profiled mess even more (the build was broken)
2317Simon Marlow <marlowsd@gmail.com>**20100826084436
2318 Ignore-this: 49d7e4db2fb53b856c213c74c8969d82
2319]
2320[Remove the debugging memory allocator - valgrind does a better job
2321Simon Marlow <marlowsd@gmail.com>**20100824113537
2322 Ignore-this: a3731a83dc18b0fd0de49452e695a7ca
2323 
2324 I got fed up with the constant bogus output from the debugging memory
2325 allocator in RtsUtils.c.  One problem is that we allocate memory in
2326 constructors that then isn't tracked, because the debugging allocator
2327 hasn't been initialised yet.
2328 
2329 The bigger problem is that for a given piece of leaking memory it's
2330 impossible to find out where it was allocated; however Valgrind gives
2331 output like this:
2332 
2333 ==6967== 8 bytes in 1 blocks are still reachable in loss record 1 of 7
2334 ==6967==    at 0x4C284A8: malloc (vg_replace_malloc.c:236)
2335 ==6967==    by 0x4C28522: realloc (vg_replace_malloc.c:525)
2336 ==6967==    by 0x6745E9: stgReallocBytes (RtsUtils.c:213)
2337 ==6967==    by 0x68D812: setHeapAlloced (MBlock.c:91)
2338 ==6967==    by 0x68D8E2: markHeapAlloced (MBlock.c:116)
2339 ==6967==    by 0x68DB56: getMBlocks (MBlock.c:240)
2340 ==6967==    by 0x684F55: alloc_mega_group (BlockAlloc.c:305)
2341 ==6967==    by 0x6850C8: allocGroup (BlockAlloc.c:358)
2342 ==6967==    by 0x69484F: allocNursery (Storage.c:390)
2343 ==6967==    by 0x694ABD: allocNurseries (Storage.c:436)
2344 ==6967==    by 0x6944F2: initStorage (Storage.c:217)
2345 ==6967==    by 0x673E3C: hs_init (RtsStartup.c:160)
2346 
2347 which tells us exactly what the leaking bit of memory is.  So I don't
2348 think we need our own debugging allocator.
2349]
2350[free the entries in the thread label table on exit
2351Simon Marlow <marlowsd@gmail.com>**20100824112606
2352 Ignore-this: c9d577c06548cda80791e590e40d35b3
2353]
2354[Panic in the right way
2355simonpj@microsoft.com**20100825091614
2356 Ignore-this: e6ea4f6dfd2aea088828ea7a945ddd5f
2357]
2358[Fix the DPH/profiled make thing (again)
2359simonpj@microsoft.com**20100825091602
2360 Ignore-this: bc58fa48034ac40cf7be4170958ea29e
2361]
2362[Don't test for gcc flags before we've found gcc
2363Ian Lynagh <igloo@earth.li>**20100824131401]
2364[Change how the dblatex/lndir problem is worked around
2365Ian Lynagh <igloo@earth.li>**20100824130938
2366 Hack: dblatex normalises the name of the main input file using
2367 os.path.realpath, which means that if we're in a linked build tree,
2368 it find the real source files rather than the symlinks in our link
2369 tree. This is fine for the static sources, but it means it can't
2370 find the generated sources.
2371 
2372 We therefore also generate the main input file, so that it really
2373 is in the link tree, and thus dblatex can find everything.
2374]
2375[Clean the generated userguide sources
2376Ian Lynagh <igloo@earth.li>**20100824105827
2377 Ignore-this: 39b4f9702c688c053ed3273b20969597
2378]
2379[DPH should not even be built if GhcProfiled
2380simonpj@microsoft.com**20100823133439
2381 Ignore-this: 62acbf83de5b70ff6d27ab38ce9218ae
2382 
2383 It's not just when cleaning!
2384]
2385[The templateHaskellOk check should only run in stage2
2386simonpj@microsoft.com**20100823133353
2387 Ignore-this: f6dc9292923a1ca201953c5f58c0af3c
2388 
2389 Because rtsIsProfiled is only available in stage2
2390]
2391[Add a couple of missing tests for EAGER_BLACKHOLE
2392Simon Marlow <marlowsd@gmail.com>**20100823104654
2393 Ignore-this: 70c981b86370b0c7564b29b057650897
2394 This was leading to looping and excessive allocation, when the
2395 computation should have just blocked on the black hole. 
2396 
2397 Reported by Christian Höner zu Siederdissen <choener@tbi.univie.ac.at>
2398 on glasgow-haskell-users.
2399]
2400[Don't check for swept blocks in -DS.
2401Marco Túlio Gontijo e Silva <marcot@marcot.eti.br>**20100718225526
2402 Ignore-this: ad5dcf3c247bc19fbef5122c1142f3b2
2403 
2404 The checkHeap function assumed the allocated part of the block contained only
2405 alive objects and slops.  This was not true for blocks that are collected using
2406 mark sweep.  The code in this patch skip the test for this kind of blocks.
2407]
2408[Fix "darcs get"
2409Ian Lynagh <igloo@earth.li>**20100822183542]
2410[Add "darcs-all upstreampull"
2411Ian Lynagh <igloo@earth.li>**20100822163419
2412 This pulls from the upstream repos, for those packages which have
2413 upstreams
2414]
2415[Generate the bit in the user guide where we say what -fglasgow-exts does
2416Ian Lynagh <igloo@earth.li>**20100822155514
2417 Stops the docs going out of sync with the code.
2418]
2419[Factor out the packages file parsing in darcs-all
2420Ian Lynagh <igloo@earth.li>**20100822154813]
2421[Document --supported-extensions
2422Ian Lynagh <igloo@earth.li>**20100822134530]
2423[fix extraction of command stack of arguments of arrow "forms" (fixes #4236)
2424Ross Paterson <ross@soi.city.ac.uk>**20100822090022
2425 Ignore-this: a93db04ec4f20540642a19cdc67d1666
2426 
2427 The command stack was being extracted (by unscramble) with the outermost
2428 type first, contrary to the comment on the function.
2429]
2430[minor fix to comment
2431Ross Paterson <ross@soi.city.ac.uk>**20100822085838
2432 Ignore-this: 8d203ba2600eaf4cf21b043dcfa96cdc
2433]
2434[Add the RTS library path to the library search path
2435Ian Lynagh <igloo@earth.li>**20100820155523
2436 In case the RTS is being explicitly linked in. For #3807.
2437]
2438[Remove some duplication of C flags
2439Ian Lynagh <igloo@earth.li>**20100819233743
2440 We now use the CONF_CC_OPTS_STAGEn C flags in machdepCCOpts, rather than
2441 repeating them there.
2442]
2443[Set -fno-stack-protector in CONF_CC_OPTS_STAGE* rathre than extra-gcc-opts
2444Ian Lynagh <igloo@earth.li>**20100819233031
2445 The latter is only used when compiling .hc files, whereas we need it for
2446 .c files too.
2447]
2448[Give clearer errors for bad input in the packages file; suggested by pejo
2449Ian Lynagh <igloo@earth.li>**20100819232420]
2450[Set -march=i686 on OS X x86 in the configure variables
2451Ian Lynagh <igloo@earth.li>**20100819230939
2452 We used to set it only in machdepCCOpts, so this is more consistent
2453]
2454[Give each stage its own Config.hs
2455Ian Lynagh <igloo@earth.li>**20100819224709
2456 This also means the file is generated in a dist directory, not a
2457 source directory.
2458]
2459[Fix cleaning when GhcProfiled = YES
2460Ian Lynagh <igloo@earth.li>**20100819131346
2461 We need to include the DPH cleaning rules, even though we don't build DPH
2462 when GhcProfiled = YES.
2463]
2464[stgReallocBytes(DEBUG): don't fail when the ptr passed in is NULL
2465Simon Marlow <marlowsd@gmail.com>**20100817150836
2466 Ignore-this: 4b5063e65e01399f64a33f0d0555ff38
2467]
2468[Use make-command in rules/bindist.mk
2469Ian Lynagh <igloo@earth.li>**20100818191243
2470 Rather than it having its own specialised version
2471]
2472[Use make-command when installing packages
2473Ian Lynagh <igloo@earth.li>**20100818190600]
2474[Add _DATA_FILES to package-data.mk files
2475Ian Lynagh <igloo@earth.li>**20100818185801]
2476[Add a "make-command" utility Makefile function
2477Ian Lynagh <igloo@earth.li>**20100818183055]
2478[LLVM: Nicer format for lack of shared lib warning
2479David Terei <davidterei@gmail.com>**20100817145207
2480 Ignore-this: 753d45762601d87761614937a1bb6716
2481]
2482[fix FP_CHECK_ALIGNMENT for autoconf 2.66 (fixes #4252)
2483Ross Paterson <ross@soi.city.ac.uk>**20100816142442
2484 Ignore-this: cd784b8888d32b3b2cc2cc0969ec40f
2485 
2486 Recent versions of AS_LITERAL_IF don't like *'s.  Fix from
2487 
2488 http://blog.gmane.org/gmane.comp.sysutils.autoconf.general/month=20100701
2489]
2490[Refactor the command-line argument parsing (again)
2491simonpj@microsoft.com**20100816074453
2492 Ignore-this: 26dc9e37a88660a887a2e316ed7a9803
2493 
2494 This change allows the client of CmdLineParser a bit more flexibility,
2495 by giving him an arbitrary computation (not just a deprecation
2496 message) for each flag. 
2497 
2498 There are several clients, so there are lots of boilerplate changes.
2499 
2500 Immediate motivation: if RTS is not profiled, we want to make
2501 Template Haskell illegal.  That wasn't with the old setup.
2502]
2503[Add upstream repo to the packages file
2504Ian Lynagh <igloo@earth.li>**20100815154741]
2505[Make the "tag" column of the packages file always present
2506Ian Lynagh <igloo@earth.li>**20100815151657
2507 It makes the parsing simpler if we always have the same number of columns
2508]
2509[Disable object splitting on OSX; works around #4013
2510Ian Lynagh <igloo@earth.li>**20100815134759]
2511[Return memory to the OS; trac #698
2512Ian Lynagh <igloo@earth.li>**20100813170402]
2513[Reduce the xargs -s value we use on Windows
2514Ian Lynagh <igloo@earth.li>**20100812223721
2515 With 30000 I was getting:
2516     xargs: value for -s option should be < 28153
2517]
2518[LLVM: Enable shared lib support on Linux x64
2519David Terei <davidterei@gmail.com>**20100813191534
2520 Ignore-this: 642ed37af38e5f17d419bf4f09332671
2521]
2522[Re-do the arity calculation mechanism again (fix Trac #3959)
2523simonpj@microsoft.com**20100813161151
2524 Ignore-this: d4a2aa48150b503b20c25351a79decfb
2525 
2526 After rumination, yet again, on the subject of arity calculation,
2527 I have redone what an ArityType is (it's purely internal to the
2528 CoreArity module), and documented it better.  The result should
2529 fix #3959, and I hope the related #3961, #3983.
2530 
2531 There is lots of new documentation: in particular
2532  * Note [ArityType] 
2533    describes the new datatype for arity info
2534 
2535  * Note [State hack and bottoming functions]
2536    says how bottoming functions are dealt with, particularly
2537    covering catch# and Trac #3959
2538 
2539 I also found I had to be careful not to eta-expand single-method
2540 class constructors; see Note [Newtype classes and eta expansion].
2541 I think this part could be done better, but it works ok.
2542]
2543[Comments only
2544simonpj@microsoft.com**20100813161019
2545 Ignore-this: baf68300d8bc630bf0b7ab27647b33a0
2546]
2547[Modify FloatOut to fix Trac #4237
2548simonpj@microsoft.com**20100813163120
2549 Ignore-this: ffc8d00d4b7f0a8a785fcef312900413
2550 
2551 The problem was that a strict binding was getting floated
2552 out into a letrec. This only happened when profiling was
2553 on.  It exposed a fragility in the floating strategy.  This
2554 patch makes it more robust.  See
2555       Note [Avoiding unnecessary floating]
2556]
2557[Fix egregious bug in SetLevels.notWorthFloating
2558simonpj@microsoft.com**20100813161429
2559 Ignore-this: d22865f48d417e6a6b732de3dfba378f
2560 
2561 This bug just led to stupid code, which would
2562 later be optimised away, but better not to generate
2563 stupid code in the first place.
2564]
2565[Delete GhcLibProfiled
2566simonpj@microsoft.com**20100813140152
2567 Ignore-this: 2e1a3f677308be726bd022f45e2fd856
2568 
2569 Simon M and I looked at this, and we think GhcLibProfiled is
2570 (a) not needed (b) confusing.
2571 
2572 Ian should review.
2573 
2574 Really, if GhcProfiled is on we should also
2575 check that 'p' is in the GhcLibWays
2576]
2577[Do not build DPH when GhcProfiled (fixes #4172)
2578simonpj@microsoft.com**20100813140021
2579 Ignore-this: 9e20181643b456e841f845ae0cab0a9a
2580 
2581 Reason: DPH uses Template Haskell and TH doesn't work
2582 in a profiled compiler
2583]
2584[Fix Trac #4220
2585simonpj@microsoft.com**20100812131319
2586 Ignore-this: 33141cfd81627592150a9e5973411ff8
2587 
2588 For deriving Functor, Foldable, Traversable with empty
2589 data cons I just generate a null equation
2590    f _ = error "urk"
2591 
2592 There are probably more lurking (eg Enum) but this will do for now.
2593]
2594[Improve the Specialiser, fixing Trac #4203
2595simonpj@microsoft.com**20100812131133
2596 Ignore-this: 482afbf75165e24a80527a6e52080c07
2597 
2598 Simply fixing #4203 is a tiny fix: in case alterantives we should
2599 do dumpUDs *including* the case binder. 
2600 
2601 But I realised that we can do better and wasted far too much time
2602 implementing the idea.  It's described in
2603     Note [Floating dictionaries out of cases]
2604]
2605[Comments
2606simonpj@microsoft.com**20100812101456
2607 Ignore-this: 6362fe887d25688c12ef2c3cf5554ce4
2608]
2609[Comments only
2610simonpj@microsoft.com**20100812101439
2611 Ignore-this: 7ed2f5fc08811cbe9958c2309a9ed1fa
2612]
2613[Fix bug in linting of shadowed case-alternative binders
2614simonpj@microsoft.com**20100812101413
2615 Ignore-this: 9212a5e2c03421749f5935b3944ecf53
2616]
2617[Comments and spacing only
2618simonpj@microsoft.com**20100812101347
2619 Ignore-this: ed59a7dae7decb24470709dc1c118dbb
2620]
2621[Add more info to more parse error messages (#3811)
2622Ian Lynagh <igloo@earth.li>**20100809233108]
2623[Run finalizers *after* updating the stable pointer table (#4221)
2624Simon Marlow <marlowsd@gmail.com>**20100810133739
2625 Ignore-this: b0462f80dd64eac71e599d8a9f6dd665
2626 Silly bug really, we were running the C finalizers while the StablePtr
2627 table was still in a partially-updated state during GC, but finalizers
2628 are allowed to call freeStablePtr() (via hs_free_fun_ptr(), for
2629 example), and chaos ensues.
2630]
2631[Do the dependency-omitting for 'make 1' in a slightly different way
2632Simon Marlow <marlowsd@gmail.com>**20100810093446
2633 Ignore-this: af15edd3a1492cbd93111316b57e02e4
2634 
2635 I encountered a couple of things that broke after Ian's previous
2636 patch: one was my nightly build scripts that use 'make stage=2' at the
2637 top level, and the other is 'make fast' in libraries/base, which uses
2638 'stage=0' to avoid building any compilers.
2639 
2640 So my version of this patch is more direct: it just turns off the
2641 appropriate dependencies using a variable set by 'make 1', 'make 2',
2642 etc.
2643]
2644[Integrate new I/O manager, with signal support
2645Johan Tibell <johan.tibell@gmail.com>**20100724102355
2646 Ignore-this: eb092857a2a1b0ca966649caffe7ac2b
2647]
2648[Add DoAndIfThenElse support
2649Ian Lynagh <igloo@earth.li>**20100808194625]
2650[Make another parse error more informative
2651Ian Lynagh <igloo@earth.li>**20100808193340]
2652[Make a parse error say what it is failing to parse; part of #3811
2653Ian Lynagh <igloo@earth.li>**20100808155732]
2654[Send ghc progress output to stdout; fixes #3636
2655Ian Lynagh <igloo@earth.li>**20100808142542]
2656[Fix the HsColour test in the build system
2657Ian Lynagh <igloo@earth.li>**20100805155319
2658 Ignore-this: ba2752b04801a253e891b31e1914485d
2659]
2660[Fix the -lm configure test; fixes #4155
2661Ian Lynagh <igloo@earth.li>**20100805142508
2662 Ignore-this: 358b8b1074d2d22fb8d362ea6d8b80d6
2663]
2664[Don't restrict filenames in line pragmas to printable characters; fixes #4207
2665Ian Lynagh <igloo@earth.li>**20100805135011
2666 Ignore-this: e3d32312127165e40e6eaa919193d60b
2667 "printable" is ASCII-only, whereas in other locales we can get things like
2668 # 1 "<línea-de-orden>"
2669]
2670[Ensure extension flags are flattened in the Cmm phase
2671Ian Lynagh <igloo@earth.li>**20100805133614
2672 If we start with a .cmmcpp file then they don't get flattened in
2673 the CmmCpp phase, as we don't run that phase.
2674]
2675[Add "cmmcpp" as a Haskellish source suffix
2676Ian Lynagh <igloo@earth.li>**20100805132555]
2677[On amd64/OSX we don't need to be given memory in the first 31bits
2678Ian Lynagh <igloo@earth.li>**20100805120600
2679 Ignore-this: 42eb64e25ad4b66ae022884305e0297b
2680 as PIC is always on
2681]
2682[NCG: Don't worry about trying to re-freeze missing coalescences
2683benl@ouroborus.net**20100702053319
2684 Ignore-this: ea05cbee19b6c5c410db41292cbb64b0
2685]
2686[Make -rtsopts more flexible
2687Ian Lynagh <igloo@earth.li>**20100805011137
2688 The default is a new "some" state, which allows only known-safe flags
2689 that we want on by default. Currently this is only "--info".
2690]
2691[Test for (fd < 0) before trying to FD_SET it
2692Ian Lynagh <igloo@earth.li>**20100804173636]
2693[Remove "On by default" comments in DynFlags
2694Ian Lynagh <igloo@earth.li>**20100802110803
2695 Ignore-this: 2a51055277b5ce9f0e98e1438b212027
2696 These make less sense now we support multiple languges. The
2697 "languageExtensions" function gives the defaults.
2698]
2699[Fix build: Add newline to end of ghc-pkg/Main.hs
2700Ian Lynagh <igloo@earth.li>**20100801183206]
2701[Add a versions haddock binary for Windows
2702Ian Lynagh <igloo@earth.li>**20100801180917]
2703[ghc-pkg: don't fail, if a file is already removed
2704ich@christoph-bauer.net**20100725162606
2705 Ignore-this: 5501d6812c31f4da525c7fb24f6dcaed
2706]
2707[Remove push-all from file list in boot script (push-all no longer exists)
2708Ian Lynagh <igloo@earth.li>**20100801121841
2709 Ignore-this: eec130f06610d8728a57626682860a1a
2710]
2711[Add error checking to boot-pkgs script
2712Ian Lynagh <igloo@earth.li>**20100801121432
2713 Ignore-this: 8afd6663db443c774bad45d75bbfe950
2714]
2715[Add more error checking to the boot script
2716Ian Lynagh <igloo@earth.li>**20100801113628]
2717[Remove libHSrtsmain.a before creating it
2718Ian Lynagh <igloo@earth.li>**20100801005432
2719 Otherwise it isn't updated properly if rts/Main.c changes
2720]
2721[Expose the functions haddock needs even when haddock is disabled; #3558
2722Ian Lynagh <igloo@earth.li>**20100731115506]
2723[Always haddock by default
2724Ian Lynagh <igloo@earth.li>**20100730235001
2725 Revert this patch:
2726     Matthias Kilian <kili@outback.escape.de>**20090920181319
2727     Don't build haddock if HADDOC_DOCS = NO, and disable HADDOC_DOCS
2728         if GhcWithInterpreter = NO
2729     Haddock uses TcRnDriver.tcRnGetInfo, which is only available if
2730     GHCI is built. Set HADDOC_DOCS to NO if GhcWithInterpreter is NO,
2731     and disable the haddock build if HADDOC_DOCS = NO.
2732]
2733[Add a debugTrace for the phases that we run
2734Ian Lynagh <igloo@earth.li>**20100729201503]
2735[* Add StringPrimL as a constructor for Template Haskell (Trac #4168)
2736simonpj@microsoft.com**20100730131922
2737 Ignore-this: 520d0a0a14b499b299e8b2be8d148ff0
2738   
2739 There are already constructors for IntPrim, FloatPrim etc,
2740 so this makes it more uniform.
2741   
2742 There's a corresponding patch for the TH library
2743]
2744[Add thread affinity support for FreeBSD
2745Gabor Pali <pgj@FreeBSD.org>**20100720001409
2746 Ignore-this: 6c117b8219bfb45445089e82ee470410
2747 - Implement missing functions for setting thread affinity and getting real
2748   number of processors.
2749 - It is available starting from 7.1-RELEASE, which includes a native support
2750   for managing CPU sets.
2751 - Add __BSD_VISIBLE, since it is required for certain types to be visible in
2752   addition to POSIX & C99.
2753]
2754[Disable symbol visibility pragmas for FreeBSD
2755Ian Lynagh <igloo@earth.li>**20100729012507
2756 Do not use GCC pragmas for controlling visibility, because it causes
2757 "undefined reference" errors at link-time.  The true reasons are
2758 unknown, however FreeBSD 8.x includes GCC 4.2.1 in the base system,
2759 which might be buggy.
2760]
2761[Fix numeric escape sequences parsing
2762Anton Nikishaev <anton.nik@gmail.com>**20100721194208
2763 Ignore-this: dd71935b1866b5624f7975c45ad519a1
2764 This fixes trac bug #1344
2765]
2766[Explicitly give the right path to perl when making the OS X installer; #4183
2767Ian Lynagh <igloo@earth.li>**20100728163030]
2768[Set -fno-stack-protector in extra-gcc-opts; fixes #4206
2769Ian Lynagh <igloo@earth.li>**20100728161957
2770 We were using it only when building the RTS, and only on certain
2771 platforms. However, some versions of OS X need the flag, while others
2772 don't support it, so we now test for it properly.
2773]
2774[Make PersistentLinkerState fields strict; fixes #4208
2775Ian Lynagh <igloo@earth.li>**20100727201911
2776 Ignore-this: fc5cfba48cd16624f6bb15a7a03a3b4
2777 We modify fields a lot, so we retain the old value if they aren't forced.
2778]
2779[Don't rebuild dependency files unnecessarily when doing "make 1" etc
2780Ian Lynagh <igloo@earth.li>**20100726211512
2781 Ignore-this: d91a729e5113aa964cc67768e92e57ef
2782]
2783[LLVM: If user specifies optlo, don't use '-O' levels
2784David Terei <davidterei@gmail.com>**20100726105650
2785 Ignore-this: e05e103b09d1de937540ffad7983f88e
2786]
2787[Flatten flags for ghci's :show
2788Ian Lynagh <igloo@earth.li>**20100725135320]
2789[Add support for Haskell98 and Haskell2010 "languages"
2790Ian Lynagh <igloo@earth.li>**20100724230121]
2791[Rename "language" varibles etc to "extension", and add --supported-extensions
2792Ian Lynagh <igloo@earth.li>**20100724223624]
2793[Separate language option handling into 2 phases
2794Ian Lynagh <igloo@earth.li>**20100724212013
2795 We now first collect the option instructions (from the commandline,
2796 from pragmas in source files, etc), and then later flatten them into
2797 the list of enabled options. This will enable us to use different
2798 standards (H98, H2010, etc) as a base upon which to apply the
2799 instructions, when we don't know what the base will be when we start
2800 collecting instructions.
2801]
2802[Separate the language flags from the other DynFlag's
2803Ian Lynagh <igloo@earth.li>**20100724133103
2804 Ignore-this: 47bb8d42e621e47016b66c7472bd6cb5
2805]
2806[Set stage-specific CC/LD opts in the bindist configure.ac
2807Ian Lynagh <igloo@earth.li>**20100724113748
2808 Ignore-this: f06926d185a35ddd05490ca4a257e992
2809]
2810[Use different CC/LD options for different stages
2811Ian Lynagh <igloo@earth.li>**20100723223059]
2812[Add some error belchs to the linker, when we find bad magic numbers
2813Ian Lynagh <igloo@earth.li>**20100723200822]
2814[Add some more linker debugging prints
2815Ian Lynagh <igloo@earth.li>**20100723180237]
2816[When (un)loading an object fails, say which object in teh panic
2817Ian Lynagh <igloo@earth.li>**20100723162649]
2818[Add a release note: GHCi import syntax
2819Ian Lynagh <igloo@earth.li>**20100721193647]
2820[Deprecate NewQualifiedOperators extension (rejected by H')
2821Ian Lynagh <igloo@earth.li>**20100719150909
2822 Ignore-this: 6e7e3bedc5360c5975f73497b3e6cba5
2823]
2824[LLVM: Allow optlc and optlo to override default params for these systools
2825David Terei <davidterei@gmail.com>**20100722181631
2826 Ignore-this: e60af7941996f7170fb3bfb02a002082
2827]
2828[LLVM: Code and speed improvement to dominateAllocs pass.
2829David Terei <davidterei@gmail.com>**20100721143654
2830 Ignore-this: 9fb7058c8a2afc005521298c7b8d0036
2831]
2832[Comments only
2833simonpj@microsoft.com**20100721144257
2834 Ignore-this: b3091ddcd1df271eb85fe90978ab7adc
2835]
2836[Fix inlining for default methods
2837simonpj@microsoft.com**20100721144248
2838 Ignore-this: 61a11a8f741f775000c6318aae4b3191
2839 
2840 This was discombobulated by a patch a week ago;
2841 now fixed, quite straightforwardly.  See
2842 Note [Default methods and instances]
2843]
2844[Allow reification of existentials and GADTs
2845simonpj@microsoft.com**20100721090437
2846 Ignore-this: 20f1ccd336cc25aff4d4d67a9ac2211a
2847 
2848 It turns out that TH.Syntax is rich enough to express even GADTs,
2849 provided we express them in equality-predicate form.  So for
2850 example
2851 
2852   data T a where
2853      MkT1 :: a -> T [a]
2854      MkT2 :: T Int
2855 
2856 will appear in TH syntax like this
2857 
2858   data T a = forall b. (a ~ [b]) => MkT1 b
2859            | (a ~ Int) => MkT2
2860 
2861 While I was at it I also improved the reification of types,
2862 so that we use TH.TupleT and TH.ListT when we can.
2863]
2864[add numSparks# primop (#4167)
2865Simon Marlow <marlowsd@gmail.com>**20100720153746
2866 Ignore-this: f3f925e7de28f3f895213aefbdbe0b0f
2867]
2868[LLVM: Decrease max opt level used under OSX to avoid bug
2869David Terei <davidterei@gmail.com>**20100720160938
2870 Ignore-this: 34b0b3550f00b27b00ad92f8232745e5
2871 
2872 Currently, many programs compiled with GHC at -O2 and LLVM
2873 set to -O3 will segfault (only under OSX). Until this issue
2874 is fixed I have simply 'solved' the segfault by lowering
2875 the max opt level for LLVM used to -O2 under OSX.
2876 
2877 All these recent changes to OSX should mean its finally as
2878 stable as Linux and Windows.
2879]
2880[LLVM: Fix OSX to work again with TNTC disabled.
2881David Terei <davidterei@gmail.com>**20100720160845
2882 Ignore-this: 8dc98139cfa536b2a64aa364d040b581
2883]
2884[LLVM: Fix printing of local vars so LLVM works with -fnew-codegen
2885David Terei <davidterei@gmail.com>**20100720160302
2886 Ignore-this: d883c433dfaed67921a8c5360e1f9f6a
2887]
2888[Use a separate mutex to protect all_tasks, avoiding a lock-order-reversal
2889Simon Marlow <marlowsd@gmail.com>**20100716150832
2890 Ignore-this: ffbdb4ee502e0f724d57acb9bfbe9d92
2891 In GHC 6.12.x I found a rare deadlock caused by this
2892 lock-order-reversal:
2893 
2894 AQ cap->lock
2895   startWorkerTask
2896     newTask
2897       AQ sched_mutex
2898 
2899 scheduleCheckBlackHoles
2900   AQ sched_mutex
2901    unblockOne_
2902     wakeupThreadOnCapabilty
2903       AQ cap->lock
2904 
2905 so sched_mutex and cap->lock are taken in a different order in two
2906 places.
2907 
2908 This doesn't happen in the HEAD because we don't have
2909 scheduleCheckBlackHoles, but I thought it would be prudent to make
2910 this less likely to happen in the future by using a different mutex in
2911 newTask.  We can clearly see that the all_tasks mutex cannot be
2912 involved in a deadlock, becasue we never call anything else while
2913 holding it.
2914]
2915['make fast' in a package does not build any compilers
2916Simon Marlow <marlowsd@gmail.com>**20100715125904
2917 Ignore-this: f27e70faf3944831dad16e89a4e273da
2918]
2919[LLVM: Fix up botched last commit
2920David Terei <davidterei@gmail.com>**20100719104823
2921 Ignore-this: a32e0f6a38cb9e02527eb8ca69b3eb59
2922]
2923[LLVM: Fix warning introduce in last commit.
2924David Terei <davidterei@gmail.com>**20100719103411
2925 Ignore-this: e9c92a9402aff50d60ab26e6ad441bfc
2926]
2927[LLVM: Use mangler to fix up stack alignment issues on OSX
2928David Terei <davidterei@gmail.com>**20100718231000
2929 Ignore-this: 9f6e8cb855269cb3a5ac1a23480d0e71
2930]
2931[Fix #4195 (isGadtSyntaxTyCon returns opposite result)
2932illissius@gmail.com**20100715134134
2933 Ignore-this: a90403f893030432b5c15d743647f350
2934]
2935[Update to time 1.2.0.3
2936Ian Lynagh <igloo@earth.li>**20100717181810
2937 Ignore-this: 1ccb4801a73f399e6718ce556543ede1
2938]
2939[Reorder RTS --info output
2940Ian Lynagh <igloo@earth.li>**20100717162356]
2941[Fix unreg prof build: Define CCS_SYSTEM in stg/MiscClosures.h
2942Ian Lynagh <igloo@earth.li>**20100717142832
2943 Ignore-this: 9675f3f51b6dac40483155344e7f45b6
2944]
2945[Make mkDerivedConstants as a stage 1 program
2946Ian Lynagh <igloo@earth.li>**20100717000827
2947 Ignore-this: 5357403461b209b8606f1d33defb51cf
2948 This way it gets the defines for the right platform when cross-compiling
2949]
2950[Don't generate Haskell dependencies if we don't have any Haskell sources
2951Ian Lynagh <igloo@earth.li>**20100717000800
2952 Ignore-this: 454abd0358f535b7e789327125c9206c
2953]
2954[Link programs that have no Haskell objects with gcc rather than ghc
2955Ian Lynagh <igloo@earth.li>**20100716235303
2956 Ignore-this: f65588b69675edea616cc434e769b0a4
2957]
2958[Use gcc to build C programs for stages >= 1
2959Ian Lynagh <igloo@earth.li>**20100716223703
2960 Ignore-this: 9f843a4e17285cda582117504707f9e7
2961]
2962[Add platform info to "ghc --info" output
2963Ian Lynagh <igloo@earth.li>**20100716141953]
2964[Tidy up Config.hs generation
2965Ian Lynagh <igloo@earth.li>**20100716140630]
2966[Fix HC porting test in makefiles
2967Ian Lynagh <igloo@earth.li>**20100716010808
2968 Ignore-this: 6052c1dd022a6108ab2236a299ee1d84
2969 Now that we are trying to support cross compilation, we can't use
2970     "$(TARGETPLATFORM)" != "$(HOSTPLATFORM)"
2971 as a test for HC-porting.
2972]
2973[Change a BUILD var to a HOST var
2974Ian Lynagh <igloo@earth.li>**20100716002558]
2975[Remove an unnecessary #include
2976Ian Lynagh <igloo@earth.li>**20100715233930
2977 Ignore-this: dcede249de6be7e3c9305c9279c2ca07
2978]
2979[Split up some make commands, so that errors aren't overlooked
2980Ian Lynagh <igloo@earth.li>**20100715152237
2981 Ignore-this: fb69b0a25d9ca71dae5e75d38db675cd
2982 When we ask make to run "a | b", if a fails then the pipeline might
2983 still exit successfuly.
2984]
2985[Remove an unnecessary #include
2986Ian Lynagh <igloo@earth.li>**20100715143000
2987 Ignore-this: 4e098cac5dda2dd595ca0a0f5121853c
2988]
2989[Simplify some more CPP __GLASGOW_HASKELL__ tests
2990Ian Lynagh <igloo@earth.li>**20100715142500]
2991[Remove some code only used with GHC 6.11.*
2992Ian Lynagh <igloo@earth.li>**20100715141720]
2993[__GLASGOW_HASKELL__ >= 609 is now always true
2994Ian Lynagh <igloo@earth.li>**20100715141544]
2995[Correct the values in ghc_boot_platform.h
2996Ian Lynagh <igloo@earth.li>**20100714223717
2997 Ignore-this: 4c99116f7ac73fadbd6d16807f57a693
2998]
2999[Change some TARGET checks to HOST checks
3000Ian Lynagh <igloo@earth.li>**20100714184715]
3001[LLVM: Add inline assembly to binding.
3002David Terei <davidterei@gmail.com>**20100714152530
3003 Ignore-this: 72a7b5460c128ed511e8901e5889fe2b
3004]
3005[LLVM: Fix mistype in last commit which broke TNTC under win/linux.
3006David Terei <davidterei@gmail.com>**20100714153339
3007 Ignore-this: 302d7957e3dded80368ebade5312ab35
3008]
3009[Remove unnecessary #include
3010Ian Lynagh <igloo@earth.li>**20100713153704
3011 Ignore-this: c37d3127b1dc68f59270c07173994c28
3012]
3013[Change some TARGET tests to HOST tests in the RTS
3014Ian Lynagh <igloo@earth.li>**20100713141034
3015 Which was being used seemed to be random
3016]
3017[LLVM: Add in new LLVM mangler for implementing TNTC on OSX
3018David Terei <davidterei@gmail.com>**20100713183243
3019 Ignore-this: 394fb74d7f9657d8b454bd0148d24bf7
3020]
3021[Refactor where an error message is generated
3022simonpj@microsoft.com**20100713115733
3023 Ignore-this: f94467856238586fcbbe48537141cf78
3024]
3025[Comments only
3026simonpj@microsoft.com**20100713115703
3027 Ignore-this: 5815442c4e69b9ec331b34242a596253
3028]
3029[Comments on data type families
3030simonpj@microsoft.com**20100713115640
3031 Ignore-this: 90a333bb7f7d64a49fb7dd180d893f6b
3032]
3033[Fix Trac #T4136: take care with nullary symbol constructors
3034simonpj@microsoft.com**20100707135945
3035 Ignore-this: 2a717a24fefcd593ea41c23dad351db0
3036 
3037 When a nullary constructor is a symbol eg (:=:) we need
3038 to take care.  Annoying.
3039]
3040[Fix Trac #4127 (and hence #4173)
3041simonpj@microsoft.com**20100707123125
3042 Ignore-this: 98bb6d0f7182b59f8c93596c61f9785d
3043 
3044 The change involves a little refactoring, so that the default
3045 method Ids are brought into scope earlier, before the value
3046 declarations are compiled.  (Since a value decl may contain
3047 an instance decl in a quote.)
3048 
3049 See Note [Default method Ids and Template Haskell] in
3050 TcTyClsDcls.
3051]
3052[Fix second bug in Trac #4127
3053simonpj@microsoft.com**20100701140124
3054 Ignore-this: c8d1cc27364fe9ee5a52acb1ecb5cdd9
3055 
3056 This bug concerned the awkward shadowing we do for
3057 Template Haskell declaration brackets.  Lots of
3058 comments in
3059 
3060   Note [Top-level Names in Template Haskell decl quotes]
3061]
3062[ia64: switch handling of 'foreign import wrapper' (FIW) to libffi
3063Sergei Trofimovich <slyfox@community.haskell.org>**20100709213922
3064 Ignore-this: fd07687e0089aebabf62de85d2be693
3065 
3066 I tried to build darcs-2.4.4 with ghc-6.12.3 and got coredumps when darcs is used
3067 in interactive mode. I tried test from ticket #3516 and found out FIW code is broken.
3068 Instead of fixing it I just switched to libffi. Result built successfully, passed
3069 'foreign import wrapper' test from ticket #3516 and builds working darcs.
3070]
3071[* storage manager: preserve upper address bits on 64bit machines (thanks to zygoloid)
3072Sergei Trofimovich <slyfox@community.haskell.org>**20100709115917
3073 Ignore-this: 9f1958a19992091ddc2761c389ade940
3074 
3075 Patch does not touch amd64 as it's address lengts is 48 bits at most, so amd64 is unaffected.
3076 
3077 the issue: during ia64 ghc bootstrap (both 6.10.4 and 6.12.3) I
3078 got the failure on stage2 phase:
3079     "inplace/bin/ghc-stage2"   -H32m -O -H64m -O0 -w ...
3080     ghc-stage2: internal error: evacuate: strange closure type 15
3081         (GHC version 6.12.3 for ia64_unknown_linux)
3082         Please report this as a GHC bug:  http://www.haskell.org/ghc/reportabug
3083     make[1]: *** [libraries/dph/dph-base/dist-install/build/Data/Array/Parallel/Base/Hyperstrict.o] Aborted
3084 
3085 gdb backtrace (break on 'barf'):
3086 Breakpoint 1 at 0x400000000469ec31: file rts/RtsMessages.c, line 39.
3087 (gdb) run -B/var/tmp/portage/dev-lang/ghc-6.12.3/work/ghc-6.12.3/inplace/bin --info
3088 Starting program: /var/tmp/portage/dev-lang/ghc-6.12.3/work/ghc-6.12.3/inplace/lib/ghc-stage2 -B/var/tmp/portage/dev-lang/ghc-6.12.3/work/ghc-6.12.3/inplace/bin --info
3089 [Thread debugging using libthread_db enabled]
3090 
3091 Breakpoint 1, barf (s=0x40000000047915b0 "evacuate: strange closure type %d") at rts/RtsMessages.c:39
3092 39        va_start(ap,s);
3093 (gdb) bt
3094 #0  barf (s=0x40000000047915b0 "evacuate: strange closure type %d") at rts/RtsMessages.c:39
3095 #1  0x400000000474a1e0 in evacuate (p=0x6000000000147958) at rts/sm/Evac.c:756
3096 #2  0x40000000046d68c0 in scavenge_srt (srt=0x6000000000147958, srt_bitmap=7) at rts/sm/Scav.c:348
3097 ...
3098 
3099 > 16:52:53 < zygoloid> slyfox: i'm no ghc expert but it looks like HEAP_ALLOCED_GC(q)
3100 >                      is returning true for a FUN_STATIC closure
3101 > 17:18:43 < zygoloid> try: p HEAP_ALLOCED_miss((unsigned long)(*p) >> 20, *p)
3102 > 17:19:12 < slyfox> (gdb) p HEAP_ALLOCED_miss((unsigned long)(*p) >> 20, *p)
3103 > 17:19:12 < slyfox> $1 = 0
3104 > 17:19:40 < zygoloid> i /think/ that means the mblock_cache is broken
3105 > 17:22:45 < zygoloid> i can't help further. however i am suspicious that you seem to have pointers with similar-looking low 33
3106 >                      bits and different high 4 bits, and it looks like such pointers get put into the same bucket in
3107 >                      mblock_cache.
3108 ...
3109 > 17:36:16 < zygoloid> slyfox: try changing the definition of MbcCacheLine to StgWord64, see if that helps
3110 > 17:36:31 < zygoloid> that's in includes/rts/storage/MBlock.h
3111 And it helped!
3112]
3113[Fixing link failure of compiler on ia64 ('-Wl,' prefixed value passed directly to ld)
3114Sergei Trofimovich <slyfox@community.haskell.org>**20100708180943
3115 Ignore-this: ced99785e1f870ee97e5bec658e2504f
3116 
3117     /usr/bin/ld -Wl,--relax -r -o dist-stage1/build/HSghc-6.10.4.o \
3118                                   dist-stage1/build/BasicTypes.o dist-stage1/build/DataCon.o ...
3119     /usr/bin/ld: unrecognized option '-Wl,--relax'
3120 
3121 If we just drop '-Wl,' part it will not help as '-r' and '--relax' are incompatible.
3122 
3123 Looks like '-Wl,--relax' was skipped by earlier binutils' ld as unknown option.
3124 Removing ia64 specific path.
3125]
3126[LLVM: Allow getelementptr to use LlvmVar for indexes.
3127David Terei <davidterei@gmail.com>**20100712152529
3128 Ignore-this: 9e158d9b89a86bca8abf11d082328278
3129]
3130[Move all the warning workarounds to one place
3131Ian Lynagh <igloo@earth.li>**20100710161723]
3132[xhtml is now warning-free
3133Ian Lynagh <igloo@earth.li>**20100710144635]
3134[Move a bit of build system code
3135Ian Lynagh <igloo@earth.li>**20100709224534]
3136[adapt to the new async exceptions API
3137Simon Marlow <marlowsd@gmail.com>**20100709125238
3138 Ignore-this: 55d845e40b9daed3575c1479d8dda1d5
3139]
3140[quiet some new spewage
3141Simon Marlow <marlowsd@gmail.com>**20100709091521
3142 Ignore-this: de7f91976bbc9789e6fd7091f05c25c0
3143]
3144[New asynchronous exception control API (ghc parts)
3145Simon Marlow <marlowsd@gmail.com>**20100708144851
3146 Ignore-this: 56320c5fc61ae3602d586609387aae22
3147 
3148 As discussed on the libraries/haskell-cafe mailing lists
3149   http://www.haskell.org/pipermail/libraries/2010-April/013420.html
3150 
3151 This is a replacement for block/unblock in the asychronous exceptions
3152 API to fix a problem whereby a function could unblock asynchronous
3153 exceptions even if called within a blocked context.
3154 
3155 The new terminology is "mask" rather than "block" (to avoid confusion
3156 due to overloaded meanings of the latter).
3157 
3158 In GHC, we changed the names of some primops:
3159 
3160   blockAsyncExceptions#   -> maskAsyncExceptions#
3161   unblockAsyncExceptions# -> unmaskAsyncExceptions#
3162   asyncExceptionsBlocked# -> getMaskingState#
3163 
3164 and added one new primop:
3165 
3166   maskUninterruptible#
3167 
3168 See the accompanying patch to libraries/base for the API changes.
3169]
3170[remove outdated comment
3171Simon Marlow <marlowsd@gmail.com>**20100708100840
3172 Ignore-this: afb2e9f6fe1f1acda51b0cbdf2637176
3173]
3174[remove 'mode: xml' emacs settings (#2208)
3175Simon Marlow <marlowsd@gmail.com>**20100708100817
3176 Ignore-this: 3a8d997fb90e01ca88dc47fb95feeba0
3177]
3178[typo in comment
3179Simon Marlow <marlowsd@gmail.com>**20100616111359
3180 Ignore-this: d3ef9288d6d6b9ab3bacbe09e0d9801c
3181]
3182[Win32 getProcessElapsedTime: use a higher-resolution time source
3183Simon Marlow <marlowsd@gmail.com>**20100708093223
3184 Ignore-this: 821989d4ff7ff2bff40cee71a881521c
3185 QueryPerformanceCounter() on Windows gives much better resolution than
3186 GetSystemTimeAsFileTime().
3187]
3188[alpha: switch handling of 'foreign import wrapper' (FIW) to libffi
3189Sergei Trofimovich <slyfox@community.haskell.org>**20100708065318
3190 Ignore-this: ddee15876737a6aa7f6dabc8ff79ce0d
3191 
3192 I tried to build ghc-6.12.3 and found out FIW part of code
3193 does not compile anymore. It uses absent functions under #ifdef.
3194 Instead of fixing it I just switched to libffi. Result built successfully
3195 and passed 'foreign import wrapper' test I wrote for trac ticket #3516.
3196 
3197 I didn't try to build -HEAD yet, but this patch only removes code, so
3198 it should not make -HEAD worse.
3199]
3200[Reorder the CPP flags so -optP can override the platform defines
3201Ian Lynagh <igloo@earth.li>**20100708203523]
3202[Add docs for DatatypeContexts extension
3203Ian Lynagh <igloo@earth.li>**20100707230907
3204 Ignore-this: 8158f03b35a2d7442a75fe85d6f1b1c7
3205]
3206[Make datatype contexts an extension (on by default) (DatatypeContexts)
3207Ian Lynagh <igloo@earth.li>**20100707212529
3208 Ignore-this: 6885ff510a0060610eeeba65122caef5
3209]
3210[LLVM: Fix various typos in comments
3211David Terei <davidterei@gmail.com>**20100707220448
3212 Ignore-this: 1ba3e722f150492da2f9d485c5795e80
3213]
3214[Handle haddock headers when looking for LANGUAGE/OPTIONS_GHC pragmas
3215Ian Lynagh <igloo@earth.li>**20100707120423
3216 Ignore-this: a75aa67690284a6cee3e62c943d4fd01
3217]
3218[Make pragState call mkPState, rather than duplicating everything
3219Ian Lynagh <igloo@earth.li>**20100706173007
3220 Ignore-this: 61fe24b99dbe7a42efff1a9dd703a75c
3221 This also means that extsBitmap gets set, whereas is was just being set
3222 to 0 before.
3223]
3224[LLVM: Add alias type defenitions to LlvmModule.
3225David Terei <davidterei@gmail.com>**20100707142053
3226 Ignore-this: eee6ad5385563ccf08e664d2634a03f2
3227]
3228[LLVM: Use packed structure type instead of structure type
3229David Terei <davidterei@gmail.com>**20100707120320
3230 Ignore-this: a06e0359d182291b81cae56993ca385e
3231 
3232 The regular structure type adds padding to conform to the platform ABI,
3233 which causes problems with structures storing doubles under windows since
3234 we don't conform to the platform ABI there. So we use packed structures
3235 instead now that don't do any padding.
3236]
3237[Make mkPState and pragState take their arguments in the same order
3238Ian Lynagh <igloo@earth.li>**20100706172611]
3239[Remove an out-of-date comment
3240Ian Lynagh <igloo@earth.li>**20100706172217
3241 Ignore-this: 710ebd7d2dc01c1b0f1e58a5b6f85701
3242]
3243[LLVM: Stop llvm saving stg caller-save regs across C calls
3244David Terei <davidterei@gmail.com>**20100705162629
3245 Ignore-this: 28b4877b31b9358e682e38fc54b88658
3246 
3247 This is already handled by the Cmm code generator so LLVM is simply
3248 duplicating work. LLVM also doesn't know which ones are actually live
3249 so saves them all which causes a fair performance overhead for C calls
3250 on x64. We stop llvm saving them across the call by storing undef to
3251 them just before the call.
3252]
3253[LLVM: Add in literal undefined value to binding
3254David Terei <davidterei@gmail.com>**20100705161544
3255 Ignore-this: 95d8361b11584aaeec44c30e76916470
3256]
3257[LLVM: Add a literal NULL value to binding
3258David Terei <davidterei@gmail.com>**20100705161308
3259 Ignore-this: 9507b4b12c1157498704a9d1e5860f12
3260 
3261 Patch from Erik de Castro Lopo <erikd@mega-nerd.com>.
3262]
3263[refactor import declaration support (#2362)
3264Simon Marlow <marlowsd@gmail.com>**20100705104557
3265 Ignore-this: ee034ac377078a7e92bfada1907c86a0
3266]
3267[Disable dynamic linking optimisations on OS X
3268Simon Marlow <marlowsd@gmail.com>**20100705103014
3269 Ignore-this: b04420d3705c51112797758d17b2e40c
3270 To improve performance of the RTS when dynamically linked on x86, I
3271 previously disabled -fPIC for certain critical modules (the GC, and a
3272 few others).  However, build reports suggest that the dynamic linker
3273 on OS X doesn't like this, so I'm disabling this optimsation on that
3274 platform.
3275]
3276[trac #2362 (full import syntax in ghci)
3277amsay@amsay.net**20100625032632
3278 Ignore-this: a9d0859d84956beb74e27b797431bf9c
3279 'import' syntax is seperate from ':module' syntax
3280]
3281[Simplify ghc-pkg's Cabal dependencies
3282Ian Lynagh <igloo@earth.li>**20100704184155
3283 We no longer support building with a compiler that doesn't come with
3284 base 4.
3285]
3286[Use Cabal to configure the dist-install ghc-pkg; fixes trac #4156
3287Ian Lynagh <igloo@earth.li>**20100704132612]
3288[Remove dead code (standalone deriving flag no longer used in parser)
3289Ian Lynagh <igloo@earth.li>**20100701162058]
3290[LLVM: Use the inbounds keyword for getelementptr instructions.
3291David Terei <davidterei@gmail.com>**20100702160511
3292 Ignore-this: 3708e658a4c82b78b1402393f4405541
3293]
3294[threadPaused: fix pointer arithmetic
3295Simon Marlow <marlowsd@gmail.com>**20100701085046
3296 Ignore-this: b78210e5d978f18ffd235f1c78a55a23
3297 Noticed by Henrique Ferreiro <hferreiro@udc.es>, thanks!
3298]
3299[LLVM: Change more operations to use getelementptr
3300David Terei <davidterei@gmail.com>**20100701161856
3301 Ignore-this: fb24eb124e203f50680c6fec3ff9fe7d
3302]
3303[Add the haskell2010 package
3304Simon Marlow <marlowsd@gmail.com>**20100630125532
3305 Ignore-this: e9b011313f283a8ff2fcda7d029a01f
3306]
3307[LLVM: Use getelementptr instruction for a lot of situations
3308David Terei <davidterei@gmail.com>**20100630181157
3309 Ignore-this: 34d314dd8dffad9bdcffdc525261a49d
3310 
3311 LLVM supports creating pointers in two ways, firstly through
3312 pointer arithmetic (by casting between pointers and ints)
3313 and secondly using the getelementptr instruction. The second way
3314 is preferable as it gives LLVM more information to work with.
3315 
3316 This patch changes a lot of pointer related code from the first
3317 method to the getelementptr method.
3318]
3319[remove out of date comments; point to the wiki
3320Simon Marlow <marlowsd@gmail.com>**20100625100313
3321 Ignore-this: 95f363a373534b9471b1818102ec592d
3322]
3323[NCG: allocatableRegs is only giving us 8 SSE regs to allocate to
3324benl@ouroborus.net**20100629054321
3325 Ignore-this: b3e0fa0b4ce988a0258dc12261989ee0
3326]
3327[LLVM: Use intrinsic functions for pow, sqrt, sin, cos
3328David Terei <davidterei@gmail.com>**20100628182949
3329 Ignore-this: 98a0befaca3fe2b36d710d8ff9f062c4
3330 
3331 Instead of calling the C library for these Cmm functions
3332 we use intrinsic functions provided by llvm. LLVM will
3333 then either create a compile time constant if possible, or
3334 use a cpu instruction or as a last resort call the C
3335 library.
3336]
3337[LLVM: Fix test '2047' under linux-x64
3338David Terei <davidterei@gmail.com>**20100628165256
3339 Ignore-this: 41735d4f431a430db636621650ccd71e
3340]
3341[LLVM: Fix test 'ffi005' under linux-x64
3342David Terei <davidterei@gmail.com>**20100628155355
3343 Ignore-this: 841f3142c63cc898ac4c3f89698a837e
3344]
3345[LLVM: Update to use new fp ops introduced in 2.7
3346David Terei <davidterei@gmail.com>**20100628144037
3347 Ignore-this: 5dd2e5964e3c039d297ed586841e706b
3348]
3349[Add noalias and nocapture attributes to pointer stg registers
3350David Terei <davidterei@gmail.com>**20100628115120
3351 Ignore-this: 492a1e723cb3a62498d240d7de92dd7
3352 
3353 At the moment this gives a very slight performance boost of around 1 - 2%.
3354 Future changes to the generated code though so that pointers are kept as
3355 pointers more often instead of being cast to integer types straight away
3356 should hopefully improve the benefit this brings.
3357 
3358]
3359[during shutdown, only free the heap if we waited for foreign calls to exit
3360Simon Marlow <marlowsd@gmail.com>**20100628090536
3361 Ignore-this: d545384a4f641d701455d08ef1217479
3362]
3363[Fix typo in -ddump-pass's document.
3364shelarcy <shelarcy@gmail.com>**20100620070759
3365 Ignore-this: f4f1ddb53f147949e948147d89190c37
3366]
3367[Add #undefs for posix source symbols when including papi.h
3368dmp@rice.edu**20100624163514
3369 Ignore-this: 8a1cba21b880d12a75a75f7e96882053
3370 
3371 Validation fails when validating with PAPI support (i.e. GhcRtsWithPapi  = YES
3372 in validate.mk).  The problem is that the posix symbols are defined by a header
3373 included from papi.h. Compilation then fails because these symbols are
3374 redefined in PosixSource.h.
3375 
3376 This patch adds an undefine for the posix symbols after including papi.h and
3377 before including PosixSource.h. The #undefines are localized to Papi.c since
3378 that is the only case where they are getting defined twice.
3379]
3380[Use machdepCCOpts in runPhase_MoveBinary; fixes trac #3952
3381Ian Lynagh <igloo@earth.li>**20100625220953]
3382[LLVM: Fix bug with calling tail with empty list
3383David Terei <davidterei@gmail.com>**20100625115729
3384 Ignore-this: 46b4b32c8d92372a2d49794a96fe1613
3385]
3386[Fix warnings
3387benl@ouroborus.net**20100624091339
3388 Ignore-this: 5ba4bbd6abb9c9d1fb8c5d21ab73f218
3389]
3390[NCG: Comments and formatting only
3391benl@ouroborus.net**20100624083121
3392 Ignore-this: 86002e72c30d06bcc876d8c49f4caa5a
3393]
3394[NCG: Do the actual reversing of SCCs
3395benl@ouroborus.net**20100624082717
3396 Ignore-this: 12d2027ea118e751fbb48b27126150ef
3397]
3398[NCG: Fix dumping of graphs in regalloc stats for graph allocator
3399benl@ouroborus.net**20100624082625
3400 Ignore-this: 2b971bc9e0318099a9afb0e0db135730
3401]
3402[NCG: Reverse SCCs after each round in the graph allocator
3403benl@ouroborus.net**20100624082437
3404 Ignore-this: f0152e4039d6f16f7b5a99b286538116
3405]
3406[NCG: Don't actually complain on unreachable code blocks
3407benl@ouroborus.net**20100624081445
3408 Ignore-this: e7335ae6120917cb858c38c7c6da8e24
3409]
3410[NCG: Do explicit check for precondition of computeLiveness
3411benl@ouroborus.net**20100624080747
3412 Ignore-this: e7053c4e5e4c3c746b5ebf016913424a
3413 
3414  computeLiveness requires the SCCs of blocks to be in reverse dependent
3415  order, and if they're not it was silently giving bad liveness info,
3416  yielding a bad allocation.
3417 
3418  Now it complains, loudly.
3419]
3420[NCG: Fix off-by-one error in realRegSqueeze
3421benl@ouroborus.net**20100623095813
3422 Ignore-this: ab0698686d4c250da8e207f734f8252d
3423]
3424[NCG: Handle stripping of liveness info from procs with no blocks (like stg_split_marker)
3425benl@ouroborus.net**20100623091209
3426 Ignore-this: c0319b6cc62ec713afe4eb03790406e3
3427]
3428[NCG: Emit a warning on unreachable code block instead of panicing
3429benl@ouroborus.net**20100623085002
3430 Ignore-this: d20314b79e3c31e764ed4cd97290c696
3431]
3432[NCG: Remember to keep the entry block first when erasing liveness info
3433Ben.Lippmeier@anu.edu.au**20090917104429
3434 Ignore-this: 1b0c1df19d622858d50ffb6a01f2cef0
3435]
3436[NCG: Refactor representation of code with liveness info
3437Ben.Lippmeier@anu.edu.au**20090917090730
3438 Ignore-this: 2aebb3b02ebd92e547c5abad9feb0f0d
3439 
3440  * I've pushed the SPILL and RELOAD instrs down into the
3441    LiveInstr type to make them easier to work with.
3442 
3443  * When the graph allocator does a spill cycle it now just
3444    re-annotates the LiveCmmTops instead of converting them
3445    to NatCmmTops and back.
3446 
3447  * This saves working out the SCCS again, and avoids rewriting
3448    the SPILL and RELOAD meta instructions into real machine
3449    instructions.
3450]
3451[NCG: Add sanity checking to linear allocator
3452Ben.Lippmeier@anu.edu.au**20090917090335
3453 Ignore-this: 5a442be8b5087d04bc8b58dffa9ea080
3454 If there are are unreachable basic blocks in the native code then the
3455 linear allocator might loop. Detect this case and panic instead.
3456]
3457[NCG: Refactor LiveCmmTop to hold a list of SCCs instead of abusing ListGraph
3458Ben.Lippmeier@anu.edu.au**20090917060332
3459 Ignore-this: 3fec8d69ed0f760e53a202f873d5d9cb
3460]
3461[NCG: Allow the liveness map in a LiveInfo to be Nothing
3462Ben.Lippmeier@anu.edu.au**20090917043937
3463 Ignore-this: 5f82422d54d1b0ffc0589eb7e82fb7a4
3464]
3465[NCG: Also show the result of applying coalesings with -ddump-asm-regalloc-stages
3466Ben.Lippmeier.anu.edu.au**20090917034427
3467 Ignore-this: 76bd6d5ca43adb2167cb25832cbaa80b
3468]
3469[Fix panic when running "ghc -H"; trac #3364
3470Ian Lynagh <igloo@earth.li>**20100624234011
3471 The problem is that showing SDoc's looks at the static flags global
3472 variables, but those are panics while we are parsing the static flags.
3473 We work around this by explicitly using a fixed prettyprinter style.
3474]
3475[Allow for stg registers to have pointer type in llvm BE.
3476David Terei <davidterei@gmail.com>**20100621175839
3477 Ignore-this: fc09b1a8314aef0bde945c77af1124fb
3478 
3479 Before all the stg registers were simply a bit type or
3480 floating point type but now they can be declared to have
3481 a pointer type to one of these. This will allow various
3482 optimisations in the future in llvm since the type is
3483 more accurate.
3484]
3485[Add support for parameter attributes to the llvm BE binding
3486David Terei <davidterei@gmail.com>**20100624111744
3487 Ignore-this: 77f3c0c7bf8f81c4a154dc835ae7bcba
3488 
3489 These allow annotations of the code produced by the backend
3490 which should bring some perforamnce gains. At the moment
3491 the attributes aren't being used though.
3492]
3493[Cast some more nats to StgWord to be on the safe side
3494Simon Marlow <marlowsd@gmail.com>**20100624105700
3495 Ignore-this: e6176683856f9872fdeb2358bb065bb8
3496 And add a comment about the dangers of int overflow
3497]
3498[comments only
3499Simon Marlow <marlowsd@gmail.com>**20100624105105
3500 Ignore-this: fc8f762f4c3a5ffca2f8da2bc63ac2a4
3501]
3502[Fix an arithmetic overflow bug causing crashes with multi-GB heaps
3503Simon Marlow <marlowsd@gmail.com>**20100624104654
3504 Ignore-this: 67210755aa098740ff5230347be0fd5d
3505]
3506[Add support for collecting PAPI native events
3507dmp@rice.edu**20100622195953
3508 Ignore-this: 7269f9c4dfb2912a024eb632200fcd1
3509 
3510 This patch extends the PAPI support in the RTS to allow collection of native
3511 events. PAPI can collect data for native events that are exposed by the
3512 hardware beyond the PAPI present events. The native events supported on your
3513 hardware can found by using the papi_native_avail tool.
3514 
3515 The RTS already allows users to specify PAPI preset events from the command
3516 line. This patch extends that support to allow users to specify native events.
3517 The changes needed are:
3518 
3519 1) New option (#) for the RTS PAPI flag for native events. For example, to
3520    collect the native event 0x40000000, use ./a.out +RTS -a#0x40000000 -sstderr
3521 
3522 2) Update the PAPI_FLAGS struct to store whether the user specified event is a
3523    papi preset or a native event
3524 
3525 3) Update init_countable_events function to add the native events after parsing
3526    the event code and decoding the name using PAPI_event_code_to_name
3527 
3528]
3529[Don't warn about unused bindings with parents in .hs-boot files; trac #3449
3530Ian Lynagh <igloo@earth.li>**20100624110351]
3531[fix the home_imps filter to allow for 'import "this" <module>'
3532Simon Marlow <marlowsd@gmail.com>**20100621125535
3533 Ignore-this: da4e605b0513afc32a4e7caa921a2c76
3534 In the PackageImports extension, import "this" means "import from the
3535 current package".
3536]
3537[Use the standard C wrapper code for the ghc-$version.exe wrapper
3538Ian Lynagh <igloo@earth.li>**20100622202859
3539 Ignore-this: 60cd3e6db3afb63e6ba9e2db3b033580
3540]
3541[Don't rely on "-packagefoo" working; use "-package foo" instead
3542Ian Lynagh <igloo@earth.li>**20100622202547]
3543[Remove unnecessary C #includes
3544Ian Lynagh <igloo@earth.li>**20100622172919]
3545[Make the ghci.exe wrapper call the right ghc.exe
3546Ian Lynagh <igloo@earth.li>**20100622172247]
3547[More updates to datalayout description in llvm BE
3548David Terei <davidterei@gmail.com>**20100622165339
3549 Ignore-this: b0c604fe7673b0aa7c7064694d574437
3550]
3551[Remove LlvmAs phase as the llvm opt tool now handles this phase
3552David Terei <davidterei@gmail.com>**20100622144044
3553 Ignore-this: b9fd8f959702b6af014e2fa654bede3
3554 
3555 This phase originally invoked the llvm-as tool that turns a textual
3556 llvm assembly file into a bit code file for the rest of llvm to deal
3557 with. Now the llvm opt tool can do this itself, so we don't need to
3558 use llvm-as anymore.
3559]
3560[Update datalayout info in llvm BE
3561David Terei <davidterei@gmail.com>**20100622123457
3562 Ignore-this: 89b043d211225dcd819f30549afe1840
3563]
3564[Fix handling of float literals in llvm BE
3565David Terei <davidterei@gmail.com>**20100622121642
3566 Ignore-this: a3b5f382ad4b5a426ad4b581664506fa
3567]
3568[Declare some top level globals to be constant when appropriate
3569David Terei <davidterei@gmail.com>**20100621174954
3570 Ignore-this: 44832f65550d4f995d11c01cc1affef5
3571 
3572 This involved removing the old constant handling mechanism
3573 which was fairly hard to use. Now being constant or not is
3574 simply a property of a global variable instead of a separate
3575 type.
3576]
3577[Reduce the number of passes over the cmm in llvm BE
3578David Terei <davidterei@gmail.com>**20100621125220
3579 Ignore-this: cb2f4e54e8d0f982d5087fbeee35c73c
3580]
3581[Fix negate op not working for -0 in llvm backend
3582David Terei <davidterei@gmail.com>**20100621123606
3583 Ignore-this: c5d76e5cffa781fed074137851b1347f
3584]
3585[ROLLBACK: picCCOpts: -dynamic should not entail -optc-fPIC
3586Simon Marlow <marlowsd@gmail.com>**20100621100409
3587 Ignore-this: f2fac7df33d3919199befc59bd455414
3588 and add a comment to explain why it was wrong.  This fixes the dyn
3589 test failures that sprang up recently.
3590]
3591[Check files are really created in libffi
3592Ian Lynagh <igloo@earth.li>**20100620163724
3593 when we think that the libffi build creates them, so they just depend
3594 on the libffi build stamp.
3595]
3596[Improve the missing-import-list warning
3597Ian Lynagh <igloo@earth.li>**20100620124320
3598 Ignore-this: 551e5fdf2dfb56b49d249e0cebaa6115
3599]
3600[Tweak missing-import-list warning
3601Ian Lynagh <igloo@earth.li>**20100620122622
3602 Ignore-this: 360cdf59ae13d66ded181129325506c4
3603]
3604[trac #1789 (warnings for missing import lists)
3605amsay@amsay.net**20100618150649
3606 Ignore-this: b0b0b1e048fbca0817c1e6fade1153fa
3607]
3608[Refix docs for sizeofByteArray#/sizeofMutableByteArray# (#3800)
3609Ian Lynagh <igloo@earth.li>**20100620103749]
3610[Remove some old commented out code
3611Ian Lynagh <igloo@earth.li>**20100620000459]
3612[SET_ARR_HDR's last argument is now a number of bytes, rather than words
3613Ian Lynagh <igloo@earth.li>**20100619235214
3614 This avoids unnecessary work and potential loss of information
3615]
3616[Replace an (incorrect) bytes-to-words calculation with ROUNDUP_BYTES_TO_WDS
3617Ian Lynagh <igloo@earth.li>**20100619234310]
3618[FIX #38000 Store StgArrWords payload size in bytes
3619Antoine Latter <aslatter@gmail.com>**20100101183346
3620 Ignore-this: 7bf3ab4fc080c46311fc10b179361bb6
3621]
3622[Add win32 datalayout support to llvm backend
3623David Terei <davidterei@gmail.com>**20100618131733
3624 Ignore-this: 4b7bffaa8ef38c628ab852c1a6c1c009
3625]
3626[Remove unused 'ddump-opt-llvm' flag
3627David Terei <davidterei@gmail.com>**20100618101237
3628 Ignore-this: f78467496d986897e49d82646ee2907e
3629]
3630[generate "movl lbl(%reg1), %reg2" instructions, better codegen for -fPIC
3631Simon Marlow <marlowsd@gmail.com>**20100618082258
3632 Ignore-this: a25567ebff9f575303ddc8f2deafebbf
3633]
3634[joinToTargets: fix a case of panic "handleComponent cyclic"
3635Simon Marlow <marlowsd@gmail.com>**20100618082147
3636 Ignore-this: 765baeefbb5a41724004acd92405cecc
3637]
3638[comment typo
3639Simon Marlow <marlowsd@gmail.com>**20100618082102
3640 Ignore-this: e495610b7dd5ec30b02938638b56cb7
3641]
3642[Add support of TNTC to llvm backend
3643David Terei <davidterei@gmail.com>**20100618093205
3644 Ignore-this: 2c27d21668374a5b0d5e844882c69439
3645 
3646 We do this through a gnu as feature called subsections,
3647 where you can put data/code into a numbered subsection
3648 and those subsections will be joined together in descending
3649 order by gas at compile time.
3650]
3651[Don't automatically insert a -fvia-C flag in an unregisterised compiler
3652Ian Lynagh <igloo@earth.li>**20100617190901
3653 Ignore-this: eb25a9a338fade9e17c153da7c5f27e9
3654 The default object mode is already HscC, so it's unnecessary, and
3655 -fvia-C generates a deprecated flag warning now.
3656]
3657[In PosixSource.h, conditionally define things based on platform
3658Ian Lynagh <igloo@earth.li>**20100617174912
3659 This may not be ideal, but it should get GHC building on all platforms
3660 again.
3661]
3662[disable -fPIC for the GC for performance reasons
3663Simon Marlow <marlowsd@gmail.com>**20100617140025
3664 Ignore-this: c7c152bbff71ef7891eaee8ff39fc281
3665 see comment for details
3666]
3667[picCCOpts: -dynamic should not entail -optc-fPIC
3668Simon Marlow <marlowsd@gmail.com>**20100617115259
3669 Ignore-this: d71e42bd56e4bd107d2c431b801855e5
3670]
3671[Make getAllocations() visible
3672Simon Marlow <marlowsd@gmail.com>**20100617113259
3673 Ignore-this: 1b7fb38a01358c0acbe8987df07d23f2
3674]
3675[Fix the symbol visibility pragmas
3676Simon Marlow <marlowsd@gmail.com>**20100617105758
3677 Ignore-this: 76552500865473a1dbebbc1cc2def9f0
3678]
3679[pick up changes to $(GhcStage1HcOpts) without re-configuring the ghc package
3680Simon Marlow <marlowsd@gmail.com>**20100616124718
3681 Ignore-this: afb56d5560c813051285607fefb15493
3682]
3683[Fix bindisttest Makefile
3684Ian Lynagh <igloo@earth.li>**20100616205611
3685 Ignore-this: 39cd352152422f378572fc3859c5a377
3686]
3687[Remove some more unused make variables
3688Ian Lynagh <igloo@earth.li>**20100616180519]
3689[Convert some more variable names to FOO_CMD, for consistency
3690Ian Lynagh <igloo@earth.li>**20100616175916]
3691[Rename some variables from FOO to FOO_CMD
3692Ian Lynagh <igloo@earth.li>**20100616161108
3693 This fixes a problem with commands like gzip, where if $GZIP is exported
3694 in the environment, then when make runs a command it'll put the Makefile
3695 variable's value in the environment. But gzip treats $GZIP as arguments
3696 for itself, so when we run gzip it thinks we're giving it "gzip" as an
3697 argument.
3698]
3699[Make the "show" target work anywhere in the build tree
3700Ian Lynagh <igloo@earth.li>**20100616122910
3701 Ignore-this: 299d40cbe16112accd9f14e56fa12158
3702]
3703[Change ghc-pwd's license to a string Cabal recognises
3704Ian Lynagh <igloo@earth.li>**20100615204015
3705 Ignore-this: c935b6ad7f605aab0168997a90b40fc6
3706]
3707[fix warning
3708Simon Marlow <marlowsd@gmail.com>**20100604205933
3709 Ignore-this: 2aaa4ed6a8b9ae1e39adc4696aaf14a3
3710]
3711[--install-signal-handles=no does not affect the timer signal (#1908)
3712Simon Marlow <marlowsd@gmail.com>**20100527214627
3713 Ignore-this: b0c51f1abdb159dc360662485095a11a
3714]
3715[Small optimisation: allocate nursery blocks contiguously
3716Simon Marlow <marlowsd@gmail.com>**20100509194928
3717 Ignore-this: e650e99e9ea9493d2efb245d565beef4
3718 This lets automatic prefetching work better, for a tiny performance boost
3719]
3720[fix -fforce-recomp setting: module is PrimOp, not PrimOps
3721Simon Marlow <marlowsd@gmail.com>**20100507084507
3722 Ignore-this: f76e0d9b643682ec0e8fb7d91afdea68
3723]
3724[it should be an error to use relative directories (#4134)
3725Simon Marlow <marlowsd@gmail.com>**20100615151740
3726 Ignore-this: 2068021701832e018ca41b22877921d5
3727]
3728[missing include-dirs or library-dirs is only a warning now (#4104)
3729Simon Marlow <marlowsd@gmail.com>**20100615151702
3730 Ignore-this: e3114123cef147bbd28ccb64581a1afb
3731]
3732[fix #3822: desugaring case command in arrow notation
3733Ross Paterson <ross@soi.city.ac.uk>**20100615225110
3734 Ignore-this: 477d6c460b4174b94b4cd113fa5b9d19
3735 
3736 Get the set of free variables from the generated case expression:
3737 includes variables in the guards and decls that were missed before,
3738 and is also a bit simpler.
3739]
3740[Deprecate the -fvia-C flag; trac #3232
3741Ian Lynagh <igloo@earth.li>**20100615151836
3742 Ignore-this: c2452b2648bf7e44546465c1b964fce
3743]
3744[Avoid using the new ~~ perl operator in the mangler
3745Ian Lynagh <igloo@earth.li>**20100615151236
3746 Ignore-this: 709a7ba4e514b1596841b3ba7e5c6cc
3747]
3748[stmAddInvariantToCheck: add missing init of invariant->lock (#4057)
3749Simon Marlow <marlowsd@gmail.com>**20100615123643
3750 Ignore-this: 3b132547fa934cecf71a846db2a5f70e
3751]
3752[Add new LLVM code generator to GHC. (Version 2)
3753David Terei <davidterei@gmail.com>**20100615094714
3754 Ignore-this: 4dd2fe5854b64a3f0339d484fd5c238
3755 
3756 This was done as part of an honours thesis at UNSW, the paper describing the
3757 work and results can be found at:
3758 
3759 http://www.cse.unsw.edu.au/~pls/thesis/davidt-thesis.pdf
3760 
3761 A Homepage for the backend can be found at:
3762 
3763 http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/Backends/LLVM
3764 
3765 Quick summary of performance is that for the 'nofib' benchmark suite, runtimes
3766 are within 5% slower than the NCG and generally better than the C code
3767 generator.  For some code though, such as the DPH projects benchmark, the LLVM
3768 code generator outperforms the NCG and C code generator by about a 25%
3769 reduction in run times.
3770 
3771]
3772[Fix Trac #4127: build GlobalRdrEnv in GHCi correctly
3773simonpj@microsoft.com**20100615070626
3774 Ignore-this: d907e3bfa7882878cea0af172aaf6e84
3775 
3776 GHCi was building its GlobalRdrEnv wrongly, so that the
3777 gre_par field was bogus.  That in turn fooled the renamer.
3778 The fix is easy: use the right function!  Namely, call
3779 RnNames.gresFromAvail rather than availsToNameSet.
3780]
3781[Comments, and improvement to pretty-printing of HsGroup
3782simonpj@microsoft.com**20100615070409
3783 Ignore-this: ec8358f2485370b20226a97ec84e9024
3784]
3785[Don't reverse bindings in rnMethodBinds (fix Trac #4126)
3786simonpj@microsoft.com**20100614163935
3787 Ignore-this: a6ffbb5af6f51b142ed0aeae8ee5e3a9
3788]
3789[Fix Trac #4120: generate a proper coercion when unifying forall types
3790simonpj@microsoft.com**20100614134311
3791 Ignore-this: 601592bb505305f1954cbe730f168da4
3792 
3793 This was just a blatant omission, which hasn't come up before.
3794 Easily fixed, happily.
3795]
3796[Use mkFunTy to ensure that invariants are respected
3797simonpj@microsoft.com**20100614134159
3798 Ignore-this: 67dcada7a4e8d9927581cd77af71b6f
3799]
3800[Remove redundant debug code
3801simonpj@microsoft.com**20100601154151
3802 Ignore-this: e6ff11c04c631cf6aac73788cbcf02b5
3803]
3804[Fix Trac #4099: better error message for type functions
3805simonpj@microsoft.com**20100531140413
3806 Ignore-this: 3f53ca98cf770577818b9c0937482577
3807 
3808 Now we only want about "T is a type function and might not be
3809 injective" when matchin (T x) against (T y), which is the case
3810 that is really confusing.
3811]
3812[Gruesome fix in CorePrep to fix embarassing Trac #4121
3813simonpj@microsoft.com**20100614132726
3814 Ignore-this: fe82d15474afaac3e6133adfd7a7e055
3815 
3816 This is a long-lurking bug that has been flushed into
3817 the open by other arity-related changes.  There's a
3818 long comment
3819 
3820      Note [CafInfo and floating]
3821 
3822 to explain. 
3823 
3824 I really hate the contortions we have to do through to keep correct
3825 CafRef information on top-level binders.  The Right Thing, I believe,
3826 is to compute CAF and arity information later, and merge it into the
3827 interface-file information when the latter is generated.
3828 
3829 But for now, this hackily fixes the problem.
3830]
3831[Fix a bug in CorePrep that meant output invariants not satisfied
3832simonpj@microsoft.com**20100531150013
3833 Ignore-this: d34eb36d8877d3caf1cf2b20de426abd
3834 
3835 In cpePair I did things in the wrong order so that something that
3836 should have been a CprRhs wasn't.  Result: a crash in CoreToStg.
3837 Fix is easy, and I added more informative type signatures too.
3838]
3839[Robustify the treatement of DFunUnfolding
3840simonpj@microsoft.com**20100531145332
3841 Ignore-this: 8f5506ada4d89f6ab8ad1e8c3ffb09ba
3842 
3843 See Note [DFun unfoldings] in CoreSyn.  The issue here is that
3844 you can't tell how many dictionary arguments a DFun needs just
3845 from looking at the Arity of the DFun Id: if the dictionary is
3846 represented by a newtype the arity might include the dictionary
3847 and value arguments of the (single) method.
3848 
3849 So we need to record the number of arguments need by the DFun
3850 in the DFunUnfolding itself.  Details in
3851    Note [DFun unfoldings] in CoreSyn
3852]
3853[Fix spelling in comment
3854simonpj@microsoft.com**20100614132259
3855 Ignore-this: bbf0d55f2e5f10ef9c74592c12f9201c
3856]
3857[Update docs on view patterns
3858simonpj@microsoft.com**20100614074801
3859 Ignore-this: 8617b9078800d4942d71f142a5b6c831
3860]
3861[Fix printing of splices; part of #4124
3862Ian Lynagh <igloo@earth.li>**20100613154838
3863 Just putting parens around non-atomic expressions isn't sufficient
3864 for splices, as only the $x and $(e) forms are valid input.
3865]
3866[In ghci, catch IO exceptions when calling canonicalizePath
3867Ian Lynagh <igloo@earth.li>**20100613134627
3868 We now get an exception if the path doesn't exist
3869]
3870[Whitespace only
3871Ian Lynagh <igloo@earth.li>**20100612213119]
3872[Whitespace only
3873Ian Lynagh <igloo@earth.li>**20100612165450]
3874[Update ghci example output in user guide; patch from YitzGale in #4111
3875Ian Lynagh <igloo@earth.li>**20100612162250]
3876[Fix #4131 missing UNTAG_CLOSURE in messageBlackHole()
3877benl@ouroborus.net**20100611044614]
3878[messageBlackHole: fix deadlock bug caused by a missing 'volatile'
3879Simon Marlow <marlowsd@gmail.com>**20100610080636
3880 Ignore-this: 3cda3054bb45408aa9bd2d794b69c938
3881]
3882[Pass --no-tmp-comp-dir to Haddock (see comment)
3883Simon Marlow <marlowsd@gmail.com>**20100604083214
3884 Ignore-this: bfa4d74038637bd149f4d878b4eb8a87
3885]
3886[Track changes to DPH libs
3887Roman Leshchinskiy <rl@cse.unsw.edu.au>**20100607052903
3888 Ignore-this: 4dbc3f8418af3e74b3fc4f9a9dfe7764
3889]
3890[Track changes to DPH libs
3891Roman Leshchinskiy <rl@cse.unsw.edu.au>**20100607012642
3892 Ignore-this: 5d4e498171a3c57ab02621bfaea82cff
3893]
3894[In ghc-pkg, send warnings to stderr
3895Ian Lynagh <igloo@earth.li>**20100606161726
3896 Ignore-this: 56927d13b5e1c1ce2752734f0f9b665b
3897]
3898[Re-add newlines to enable layout for multi-line input.
3899Ian Lynagh <igloo@earth.li>**20100602180737
3900 Patch from Adam Vogt <vogt.adam@gmail.com>
3901 Partial fix for #3984
3902]
3903[Don't use unnecessary parens when printing types (Fix Trac 4107)
3904simonpj@microsoft.com**20100604110143
3905 Ignore-this: a833714ab13013c4345b222f4e87db1d
3906 
3907    f :: Eq a => a -> a
3908 rather than
3909    f :: (Eq a) => a -> a
3910]
3911[Track DPH library changes
3912Roman Leshchinskiy <rl@cse.unsw.edu.au>**20100604005728
3913 Ignore-this: 32bc2fbea6ad975e89545d4c42fd7c30
3914]
3915[fix --source-entity option passed to Haddock: we needed to escape a #
3916Simon Marlow <marlowsd@gmail.com>**20100603125459
3917 Ignore-this: d52ae6188b510c482bcebb23f0e553ae
3918]
3919[__stg_EAGER_BLACKHOLE_INFO -> __stg_EAGER_BLACKHOLE_info (#4106)
3920Simon Marlow <marlowsd@gmail.com>**20100602091419
3921 Ignore-this: 293315ac8f86fd366b8d61992ecc7961
3922]
3923[Add xhtml package (a new dependency of Haddock; not installed/shipped)
3924Simon Marlow <marlowsd@gmail.com>**20100602090101
3925 Ignore-this: af0ac8b91abe98f7fdb624ea0a4dee20
3926]
3927[Use UserInterrupt rather than our own Interrupted exception (#4100)
3928Simon Marlow <marlowsd@gmail.com>**20100602082345
3929 Ignore-this: 1909acf2f452593138b9f85024711714
3930]
3931[Add the global package DB to ghc --info (#4103)
3932Simon Marlow <marlowsd@gmail.com>**20100602082233
3933 Ignore-this: fd5c0e207e70eb0f62606c45dc5b8124
3934]
3935[rts/sm/GC.c: resize_generations(): Remove unneeded check of number of generations.
3936Marco Túlio Gontijo e Silva <marcot@debian.org>**20100528115612
3937 Ignore-this: 6f1bea62917c01c7adac636146132c97
3938 
3939 This "if" is inside another "if" which checks for RtsFlags.GcFlags.generations
3940 > 1, so testing this again is redundant, assuming the number of generations
3941 won't change during program execution.
3942]
3943[rts/sm/BlockAlloc.c: Small comment correction.
3944Marco Túlio Gontijo e Silva <marcot@debian.org>**20100526205839
3945 Ignore-this: bd2fcd4597cc872d80b0e2eeb1c3998a
3946]
3947[rts/sm/GC.c: Annotate constants.
3948Marco Túlio Gontijo e Silva <marcot@debian.org>**20100526205707
3949 Ignore-this: f232edb89383564d759ed890a18f602f
3950]
3951[includes/rts/storage/GC.h: generation_: n_words: Improve comment.
3952Marco Túlio Gontijo e Silva <marcot@debian.org>**20100526204615
3953 Ignore-this: f5d5feefa8f7b552303978f1804fea23
3954]
3955[Add PPC_RELOC_LOCAL_SECTDIFF support; patch from PHO in #3654
3956Ian Lynagh <igloo@earth.li>**20100601204211
3957 Ignore-this: 51293b7041cdce3ce7619ef11cf7ceb
3958]
3959[powerpc-apple-darwin now supports shared libs
3960Ian Lynagh <igloo@earth.li>**20100601173325]
3961[PIC support for PowerPC
3962pho@cielonegro.org**20100508143900
3963 Ignore-this: 3673859a305398c4acae3f4d7c997615
3964 
3965 PPC.CodeGen.getRegister was not properly handling PicBaseReg.
3966 It seems working with this patch, but I'm not sure this change is correct.
3967]
3968[Vectoriser: only treat a function as scalar if it actually computes something
3969Roman Leshchinskiy <rl@cse.unsw.edu.au>**20100601045630
3970 Ignore-this: e5d99a6ddb62052e3520094a5af47552
3971]
3972[Add a release notes file for 6.14.1
3973Ian Lynagh <igloo@earth.li>**20100530171117
3974 Ignore-this: 1941e6d3d1f4051b69ca2f17a1cf84d6
3975]
3976[Check dblatex actually creates the files we tell it to
3977Ian Lynagh <igloo@earth.li>**20100530171043
3978 Ignore-this: ccc72caea2313be05cbac59bb54c0603
3979 If it fails, it still exits successfully.
3980]
3981[Add darwin to the list of OSes for which we use mmap
3982Ian Lynagh <igloo@earth.li>**20100529145016
3983 Ignore-this: a86d12a3334aaaafc86f7af9dbb0a7ae
3984 Patch from Barney Stratford
3985]
3986[Simplify the CPP logic in rts/Linker.c
3987Ian Lynagh <igloo@earth.li>**20100529144929
3988 Ignore-this: 1288f5b752cc1ab8b1c90cfd0ecfdf68
3989]
3990[Fix validate on OS X
3991Ian Lynagh <igloo@earth.li>**20100529154726]
3992[OS X x86_64 fix from Barney Stratford
3993Ian Lynagh <igloo@earth.li>**20100529122440]
3994[OS X 64 installer fixes from Barney Stratford
3995Ian Lynagh <igloo@earth.li>**20100528234935]
3996[fix warning
3997Simon Marlow <marlowsd@gmail.com>**20100525155812
3998 Ignore-this: f34eee3fe3d89579fd8d381c91ced750
3999]
4000[Fix doc bugs (#4071)
4001Simon Marlow <marlowsd@gmail.com>**20100525155728
4002 Ignore-this: aa25be196de567de360075022a1942f7
4003]
4004[Make sparks into weak pointers (#2185)
4005Simon Marlow <marlowsd@gmail.com>**20100525150435
4006 Ignore-this: feea0bb5006007b82c932bc3006124d7
4007 The new strategies library (parallel-2.0+, preferably 2.2+) is now
4008 required for parallel programming, otherwise parallelism will be lost.
4009]
4010[If you say 'make' or 'make stage=2' here, pretend we're in the ghc dir
4011Simon Marlow <marlowsd@gmail.com>**20100525085301
4012 Ignore-this: 78b740337aa460915c812cbbcdae5321
4013]
4014[Another attempt to get these #defines right
4015Simon Marlow <marlowsd@gmail.com>**20100525154313
4016 Ignore-this: 460ca0c47d81cd25eae6542114f67899
4017 Apparently on Solaris it is an error to omit _ISOC99_SOURCE when using
4018 _POSIX_C_SOURCE==200112L.
4019]
4020[Add configure flags for the location of GMP includes/library; fixes #4022
4021Ian Lynagh <igloo@earth.li>**20100525221616
4022 Ignore-this: fc3060caf995d07274ec975eeefbdf3e
4023]
4024[Refactor pretty printing of TyThings to fix Trac #4015
4025simonpj@microsoft.com**20100525153126
4026 Ignore-this: 8f15053b7554f62caa84201d2e4976d2
4027]
4028[When haddocking, we need the dependencies to have been built
4029Ian Lynagh <igloo@earth.li>**20100525145830
4030 as haddock loads the .hi files with the GHC API.
4031]
4032[Fix profiling output; spotted by jlouis
4033Ian Lynagh <igloo@earth.li>**20100525111217
4034 We were outputing the number of words allocated in a column titled "bytes".
4035]
4036[Improve printing of TyThings; fixes Trac #4087
4037simonpj@microsoft.com**20100525114045
4038 Ignore-this: da2a757a533454bba80b9b77cc5a771
4039]
4040[Spelling in comments
4041simonpj@microsoft.com**20100525114001
4042 Ignore-this: 270f3da655e526cf04e27db7a01e29c0
4043]
4044[Refactor (again) the handling of default methods
4045simonpj@microsoft.com**20100525113910
4046 Ignore-this: 6686f6cdb878d57abf6b49fec64fcbb1
4047 
4048 This patch fixes Trac #4056, by
4049 
4050  a) tidying up the treatment of default method names
4051  b) removing the 'module' argument to newTopSrcBinder
4052 
4053 The details aren't that interesting, but the result
4054 is much tidier. The original bug was a 'nameModule' panic,
4055 caused by trying to find the module of a top-level name.
4056 But TH quotes generate Internal top-level names that don't
4057 have a module, and that is generally a good thing. 
4058 
4059 Fixing that in turn led to the default-method refactoring,
4060 which also makes the Name for a default method be handled
4061 in the same way as other derived names, generated in BuildTyCl
4062 via a call newImplicitBinder.  Hurrah.
4063]
4064[Don't do SpecConstr on NOINLINE things (Trac #4064)
4065simonpj@microsoft.com**20100525112807
4066 Ignore-this: 452be0a2cef0042fb67275c2827b5f72
4067 
4068 Since the RULE from specialising gets the same Activation as
4069 the inlining for the Id itself there's no point in specialising
4070 a NOINLINE thing, because the rule will be permanently switched
4071 off.
4072 
4073 See Note [Transfer activation] in SpecConstr
4074 and Note [Auto-specialisation and RULES] in Specialise.
4075]
4076[Change our #defines to work on FreeBSD too
4077Simon Marlow <marlowsd@gmail.com>**20100524105828
4078 Ignore-this: b23ede46211e67859206c0ec57d6a86f
4079 With glibc, things like _POSIX_C_SOURCE and _ISOC99_SOURCE are
4080 additive, but on FreeBSD they are mutually exclusive.  However, it
4081 turns out we only need to define _POSIX_C_SOURCE and _XOPEN_SOURCE to
4082 get all the C99 stuff we need too, so there's no need for any #ifdefs.
4083 
4084 Submitted by: Gabor PALI <pgj@FreeBSD.org>
4085]
4086[Add a missing UNTAG_CLOSURE, causing bus errors on Sparc
4087Simon Marlow <marlowsd@gmail.com>**20100524105547
4088 Ignore-this: a590b5391d6f05d50c8c088456c3c166
4089 We just about got away with this on x86 which isn't
4090 alignment-sensitive.  The result of the memory load is compared
4091 against a few different values, but there is a fallback case that
4092 happened to be the right thing when the pointer was tagged.  A good
4093 bug to find, nonetheless.
4094]
4095[Add wiki links
4096Simon Marlow <marlowsd@gmail.com>**20100520095953
4097 Ignore-this: c22f126cde166e6207922b2eb51d29e3
4098]
4099[the 'stage=0' trick to disable all compiler builds stopped working; fix it
4100Simon Marlow <marlowsd@gmail.com>**20100520104455
4101 Ignore-this: bb6fae9056471612c8dbf06916188c33
4102]
4103[Comments and formatting only
4104benl@ouroborus.net**20100524014021
4105 Ignore-this: 64579c38154728b632e358bec751cc0b
4106]
4107[Core prettyprinter fixes. Patch from Tim Chevalier. Fixes #4085
4108Ian Lynagh <igloo@earth.li>**20100522225048]
4109[Correct install-name for dynamic Darwin rts
4110pho@cielonegro.org**20100508151155
4111 Ignore-this: 6d31716c8c113dcb46e9cb925c4201df
4112]
4113[Fix the RTS debug_p build
4114Ian Lynagh <igloo@earth.li>**20100522163127]
4115[Unset $CFLAGS for "GNU non-executable stack" configure test; fixes #3889
4116Ian Lynagh <igloo@earth.li>**20100521165005
4117 With gcc 4.4 we get
4118     Error: can't resolve `.note.GNU-stack' {.note.GNU-stack section} - `.Ltext0' {.text section}
4119 when running gcc with the -g flag. To work around this we unset
4120 CFLAGS when running the test.
4121]
4122[Don't run "set -o igncr" before configuring libffi
4123Ian Lynagh <igloo@earth.li>**20100520162918
4124 Ignore-this: 489fa94df23f2adf4ff63c8ede2c0794
4125 It used to make the build work on cygwin, but now it breaks it instead:
4126     config.status: creating include/Makefile
4127     gawk: ./confLqjohp/subs.awk:1: BEGIN {\r
4128     gawk: ./confLqjohp/subs.awk:1: ^ backslash not last character on line
4129     config.status: error: could not create include/Makefile
4130     make[2]: *** [libffi/stamp.ffi.configure-shared] Error 1
4131     make[1]: *** [all] Error 2
4132]
4133[Stop passing -Wl,-macosx_version_min to gcc
4134Ian Lynagh <igloo@earth.li>**20100520154003
4135 Fixes a build failure on OS X 10.6. When linking
4136     rts/dist/build/libHSrts-ghc6.13.20100519.dylib
4137 we got
4138     ld: symbol dyld_stub_binding_helper not defined (usually in crt1.o/dylib1.o/bundle1.o)
4139     collect2: ld returned 1 exit status
4140]
4141[Fix build on FreeBSD; patch from Gabor PALI
4142Ian Lynagh <igloo@earth.li>**20100519140552]
4143[Fix package shadowing order (#4072)
4144Simon Marlow <marlowsd@gmail.com>**20100519104617
4145 Ignore-this: 26ea5e4bb5dff18618b807a54c7d6ebb
4146 
4147 Later packages are supposed to shadow earlier ones in the stack,
4148 unless the ordering is overriden with -package-id flags.
4149 Unfortunately an earlier fix for something else had sorted the list of
4150 packages so that it was in lexicographic order by installedPackageId,
4151 and sadly our test (cabal/shadow) didn't pick this up because the
4152 lexicographic ordering happened to work for the test.  I've now fixed
4153 the test so it tries both orderings.
4154]
4155[Set more env variables when configuring libffi
4156Ian Lynagh <igloo@earth.li>**20100518185014
4157 We now tell it where to find ld, nm and ar
4158]
4159[Set the location of ar to be the in-tree ar on Windows
4160Ian Lynagh <igloo@earth.li>**20100518181556]
4161[Change another / to </> to avoid building paths containing \/
4162Ian Lynagh <igloo@earth.li>**20100518172015
4163 This will hopefully fix #2889.
4164]
4165[Fix #4074 (I hope).
4166Simon Marlow <marlowsd@gmail.com>**20100518113214
4167 Ignore-this: 73cd70f5bc6f5add5247b61985c03fc1
4168 
4169 1. allow multiple threads to call startTimer()/stopTimer() pairs
4170 2. disable the timer around fork() in forkProcess()
4171 
4172 A corresponding change to the process package is required.
4173]
4174[we don't have a gcc-lib in LIB_DIR any more
4175Simon Marlow <marlowsd@gmail.com>**20100401102351
4176 Ignore-this: f41acd2d8f8e6763aa8bd57a0b44a7e4
4177]
4178[In validate, use gmake if available; based on a patch from Gabor PALI
4179Ian Lynagh <igloo@earth.li>**20100517200654]
4180[Remove duplicate "./configure --help" output; fixes #4075
4181Ian Lynagh <igloo@earth.li>**20100516141206]
4182[Update various 'sh boot's to 'perl boot'
4183Ian Lynagh <igloo@earth.li>**20100516122609
4184 Spotted by Marco Túlio Gontijo e Silva
4185]
4186[add missing initialisation for eventBufMutex
4187Simon Marlow <marlowsd@gmail.com>**20100514094943
4188 Ignore-this: 7f75594a8cb54fbec5aebd46bb959f45
4189]
4190[Undo part of #4003 patch
4191Simon Marlow <marlowsd@gmail.com>**20100513142017
4192 Ignore-this: cb65db86a38a7e5ccee9f779e489d104
4193 We still need the workaround for when compiling HEAD with 6.12.2
4194 
4195]
4196[Fix makefile loop (#4050)
4197pho@cielonegro.org**20100507140707
4198 Ignore-this: 3a1cb13d0600977e74d17ac26cbef83d
4199 
4200 The libtool creates "libffi.dylib" and "libffi.5.dylib" but not "libffi.5.0.9.dylib". Having it in libffi_DYNAMIC_LIBS causes an infinite makefile loop.
4201]
4202[fix !TABLES_NEXT_TO_CODE
4203Simon Marlow <marlowsd@gmail.com>**20100510151934
4204 Ignore-this: fccb859b114bef1c3122c98e60af51
4205]
4206[looksLikeModuleName: allow apostrophe in module names (#4051)
4207Simon Marlow <marlowsd@gmail.com>**20100510094741
4208 Ignore-this: df9348f3ba90608bec57257b47672985
4209]
4210[add the proper library dependencies for GhcProfiled=YES
4211Simon Marlow <marlowsd@gmail.com>**20100506122118
4212 Ignore-this: 6236993aa308ab5b5e1e5ea5f65982a
4213]
4214[Fix Trac #4003: fix the knot-tying in checkHiBootIface
4215simonpj@microsoft.com**20100511075026
4216 Ignore-this: a9ce2a318386fdc8782848df84592002
4217 
4218 I had incorrectly "optimised" checkHiBootIface so that it forgot
4219 to update the "knot-tied" type environment.
4220 
4221 This patch fixes the HEAD
4222]
4223[Re-engineer the derived Ord instance generation code (fix Trac #4019)
4224simonpj@microsoft.com**20100510133333
4225 Ignore-this: 8fe46e4dad27fbee211a7928acf372c2
4226   
4227 As well as fixing #4019, I rejigged the way that Ord instances are
4228 generated, which should make them faster in general.  See the
4229 Note [Generating Ord instances].
4230 
4231 I tried to measure the performance difference from this change, but
4232 the #4019 fix only removes one conditional branch per iteration, and
4233 I couldn't measure a consistent improvement.  But still, tihs is
4234 better than before.
4235]
4236[Make arity of INLINE things consistent
4237simonpj@microsoft.com**20100510133005
4238 Ignore-this: 15e7abf803d1dcb3f4ca760d2d939d0d
4239 
4240 We eta-expand things with INLINE pragmas;
4241 see Note [Eta-expanding INLINE things].
4242 
4243 But I eta-expanded it the wrong amount when the function
4244 was overloaded.  Ooops.
4245]
4246[Compacting GC fix, we forgot to thread the new bq field of StgTSO.
4247Simon Marlow <marlowsd@gmail.com>**20100510082325
4248 Ignore-this: a079c8446e2ad53efff6fd95d0f3ac80
4249]
4250[Add version constraints for the boot packages; fixes trac #3852
4251Ian Lynagh <igloo@earth.li>**20100509175051
4252 When using the bootstrapping compiler, we now explicitly constrain
4253 the version of boot packages (Cabal, extensible-exceptions, etc) to the
4254 in-tree version, so that the build system is less fragile should the
4255 user have a newer version installed for the bootstrapping compiler.
4256]
4257[Don't include inter-package dependencies when compiling with stage 0; #4031
4258Ian Lynagh <igloo@earth.li>**20100509130511
4259 This fixes a problem when building with GHC 6.12 on Windows, where
4260 dependencies on stage 0 (bootstrapping compiler) packages have absolute
4261 paths c:/ghc/..., and make gets confused by the colon.
4262]
4263[Add a ghc.mk for bindisttest/
4264Ian Lynagh <igloo@earth.li>**20100508223911]
4265[Move some make variables around so they are available when cleaning
4266Ian Lynagh <igloo@earth.li>**20100508212405]
4267[Optimise checkremove a bit
4268Ian Lynagh <igloo@earth.li>**20100508202006]
4269[Improve the bindisttest Makefile
4270Ian Lynagh <igloo@earth.li>**20100508195450]
4271[Add tools to test that cleaning works properly
4272Ian Lynagh <igloo@earth.li>**20100508194105]
4273[Tweak the ghc-pkg finding code
4274Ian Lynagh <igloo@earth.li>**20100508125815
4275 It now understand the ghc-stage[123] names we use in-tree, and it won't
4276 go looking for any old ghc-pkg if it can't find the one that matches
4277 ghc.
4278]
4279[Add a way to show what cleaning would be done, without actually doing it
4280Ian Lynagh <igloo@earth.li>**20100508122438]
4281[Tidy up the "rm" flags in the build system
4282Ian Lynagh <igloo@earth.li>**20100508115745]
4283[Fix crash in nested callbacks (#4038)
4284Simon Marlow <marlowsd@gmail.com>**20100507093222
4285 Ignore-this: cade85e361534ce711865a4820276388
4286 Broken by "Split part of the Task struct into a separate struct
4287 InCall".
4288]
4289[Add $(GhcDynamic) knob, set to YES to get stage2 linked with -dynamic
4290Simon Marlow <marlowsd@gmail.com>**20100428205241
4291 Ignore-this: 1db8bccf92099785ecac39aebd27c92d
4292 Default currently NO.
4293 
4294 Validate passed with GhcDynamic=YES on x86/Linux here.
4295 
4296 The compiler is currently slower on x86 when linked -dynamic,
4297 because the GC inner loop has been adversely affected by -fPIC, I'm
4298 looking into how to fix it.
4299]
4300[omit "dyn" from the way appended to the __stginit label
4301Simon Marlow <marlowsd@gmail.com>**20100428204914
4302 Ignore-this: 14183f3defa9f2bde68fda6729b740bc
4303 When GHCi is linked dynamically, we still want to be able to load
4304 non-dynamic object files.
4305]
4306[improvements to findPtr(), a neat hack for browsing the heap in gdb
4307Simon Marlow <marlowsd@gmail.com>**20100506115427
4308 Ignore-this: ac57785bb3e13b97a5945f753f068738
4309]
4310[Fix +RTS -G1
4311Simon Marlow <marlowsd@gmail.com>**20100506110739
4312 Ignore-this: 86a5de39a94d3331a4ee1213f82be497
4313]
4314[Enable the "redundant specialise pragmas" warning; fixes trac #3855
4315Ian Lynagh <igloo@earth.li>**20100506175351]
4316[Find the correct external ids when there's a wrapper
4317simonpj@microsoft.com**20100506164135
4318 Ignore-this: 636266407b174b05b2b8646cc73062c0
4319 
4320 We were failing to externalise the wrapper id for a function
4321 that had one.
4322]
4323[Add a comment about pattern coercions
4324simonpj@microsoft.com**20100506164027
4325 Ignore-this: 17428089f3df439f65d892e23e8ed61a
4326]
4327[Comments only
4328simonpj@microsoft.com**20100506163829
4329 Ignore-this: 169167b6463873ab173cc5750c5be469
4330]
4331[Make a missing name in mkUsageInfo into a panic
4332simonpj@microsoft.com**20100506163813
4333 Ignore-this: b82ff1b8bf89f74f146db7cb5cc4c4d7
4334 
4335 We really want to know about this!
4336]
4337[Refactoring of hsXxxBinders
4338simonpj@microsoft.com**20100506163737
4339 Ignore-this: 97c6667625262b160f9746f7bea1c980
4340 
4341 This patch moves various functions that extract the binders
4342 from a HsTyClDecl, HsForeignDecl etc into HsUtils, and gives
4343 them consistent names.
4344]
4345[Fix Trac #3966: warn about useless UNPACK pragmas
4346simonpj@microsoft.com**20100506163337
4347 Ignore-this: 5beb24b686eda6113b614dfac8490df1
4348 
4349 Warning about useless UNPACK pragmas wasn't as easy as I thought.
4350 I did quite a bit of refactoring, which improved the code by refining
4351 the types somewhat.  In particular notice that in DataCon, we have
4352 
4353     dcStrictMarks   :: [HsBang]
4354     dcRepStrictness :: [StrictnessMarks]
4355 
4356 The former relates to the *source-code* annotation, the latter to
4357 GHC's representation choice.
4358]
4359[Make tcg_dus behave more sanely; fixes a mkUsageInfo panic
4360simonpj@microsoft.com**20100506162719
4361 Ignore-this: d000bca15b0e127e297378ded1bfb81b
4362 
4363 The tcg_dus field used to contain *uses* of type and class decls,
4364 but not *defs*.  That was inconsistent, and it really went wrong
4365 for Template Haskell bracket.  What happened was that
4366  foo = [d| data A = A
4367                   f :: A -> A
4368                   f x = x |]
4369 would find a "use" of A when processing the top level of the module,
4370 which in turn led to a mkUsageInfo panic in MkIface.  The cause was
4371 the fact that the tcg_dus for the nested quote didn't have defs for
4372 A.
4373]
4374[Add a HsExplicitFlag to SpliceDecl, to improve Trac #4042
4375simonpj@microsoft.com**20100506161523
4376 Ignore-this: e4e563bac2fd831cc9e94612f5b4fa9d
4377 
4378 The issue here is that
4379 
4380     g :: A -> A
4381     f
4382     data A = A
4383 
4384 is treated as if you'd written $(f); that is the call of
4385 f is a top-level Template Haskell splice.  This patch
4386 makes sure that we *first* check the -XTemplateHaskellFlag
4387 and bleat about a parse error if it's off.  Othewise we
4388 get strange seeing "A is out of scope" errors.
4389]
4390[Change an assert to a warn
4391simonpj@microsoft.com**20100506161111
4392 Ignore-this: 739a4fb4c7940376b0f2c8ad52a1966c
4393 
4394 This is in the constraint simplifier which I'm about
4395 to rewrite, so I'm hoping the assert isn't fatal!
4396]
4397[Tidy up debug print a little
4398simonpj@microsoft.com**20100506161027
4399 Ignore-this: bd5492878e06bee1cddcbb3fc4df66d8
4400]
4401[Remove useless UNPACK pragmas
4402simonpj@microsoft.com**20100506161012
4403 Ignore-this: 3e5ab1a7cf58107034412a798bc214e5
4404]
4405[Add WARNM2 macro, plus some refactoring
4406simonpj@microsoft.com**20100506160808
4407 Ignore-this: 2ab4f1f0b5d94be683036e77aec09255
4408]
4409[Use -Wwarn for the binary package, becuase it has redundant UNPACK pragmas
4410simonpj@microsoft.com**20100506160750
4411 Ignore-this: cf0d3a11473e28bfce9602e716e69a5f
4412]
4413[Fix Trac #3966: warn about unused UNPACK pragmas
4414simonpj@microsoft.com**20100409201812
4415 Ignore-this: c96412596b39c918b5fb9b3c39ce2119
4416]
4417[Fix Trac #3953: fail earlier when using a bogus quasiquoter
4418simonpj@microsoft.com**20100409201748
4419 Ignore-this: ef48e39aa932caed538643985234f043
4420]
4421[Fix Trac #3965: tighten conditions when deriving Data
4422simonpj@microsoft.com**20100409184420
4423 Ignore-this: 96f7d7d2da11565d26b465d7d0497ac9
4424 
4425 It's tricky to set up the context for a Data instance.  I got it wrong
4426 once, and fixed it -- hence the "extra_constraints" in
4427 TcDeriv.inferConstraints. 
4428 
4429 But it still wasn't right!  The tricky bit is that dataCast1 is only
4430 generated when T :: *->*, and dataCast2 when T :: *->*->*. (See
4431 the code in TcGenDeriv for dataCastX.
4432]
4433[Fix Trac #3964: view patterns in DsArrows
4434simonpj@microsoft.com**20100409165557
4435 Ignore-this: d823c182831d5e2e592e995b16180e2f
4436 
4437 Just a missing case; I've eliminated the catch-all so
4438 that we get a warning next time we extend HsPat
4439]
4440[Fix Trac #3955: renamer and type variables
4441simonpj@microsoft.com**20100409163710
4442 Ignore-this: bd5ec64d76c0f583bf5f224792bf294c
4443 
4444 The renamer wasn't computing the free variables of a type declaration
4445 properly.  This patch refactors a bit, and makes it more robust,
4446 fixing #3955 and several other closely-related bugs.  (We were
4447 omitting some free variables and that could just possibly lead to a
4448 usage-version tracking error.
4449]
4450[Layout only
4451simonpj@microsoft.com**20100409163506
4452 Ignore-this: 1f14990b5aa0b9821b84452fb34e9f41
4453]
4454[Give a better deprecated message for INCLUDE pragmas; fixes #3933
4455Ian Lynagh <igloo@earth.li>**20100506130910
4456 We now have a DeprecatedFullText constructor, so we can override the
4457 "-#include is deprecated: " part of the warning.
4458]
4459[De-haddock a comment that confuses haddock
4460Ian Lynagh <igloo@earth.li>**20100506123607]
4461[Fix comment to not confuse haddock
4462Ian Lynagh <igloo@earth.li>**20100506113642]
4463[Detect EOF when trying to parse a string in hp2ps
4464Ian Lynagh <igloo@earth.li>**20100506000830]
4465[Make the demand analyser sdd demands for strict constructors
4466simonpj@microsoft.com**20100505200936
4467 Ignore-this: eb32632adbc354eb7a5cf884c263e0d3
4468 
4469 This opportunity was spotted by Roman, and is documented in
4470 Note [Add demands for strict constructors] in DmdAnal.
4471]
4472[Fix interaction of exprIsCheap and the lone-variable inlining check
4473simonpj@microsoft.com**20100505200723
4474 Ignore-this: f3cb65085c5673a99153d5d7b6559ab1
4475 
4476 See Note [Interaction of exprIsCheap and lone variables] in CoreUnfold
4477 
4478 This buglet meant that a nullary definition with an INLINE pragma
4479 counter-intuitively didn't get inlined at all.  Roman identified
4480 the bug.
4481]
4482[Matching cases in SpecConstr and Rules
4483simonpj@microsoft.com**20100505200543
4484 Ignore-this: f5c28c780fbf8badce84c6fdc9aa1779
4485 
4486 This patch has zero effect.  It includes comments,
4487 a bit of refactoring, and a tiny bit of commment-out
4488 code go implement the "matching cases" idea below.
4489 
4490 In the end I've left it disabled because while I think
4491 it does no harm I don't think it'll do any good either.
4492 But I didn't want to lose the idea totally. There's
4493 a thread called "Storable and constant memory" on
4494 the libraries@haskell.org list (Apr 2010) about it.
4495 
4496 Note [Matching cases]
4497 ~~~~~~~~~~~~~~~~~~~~~
4498 {- NOTE: This idea is currently disabled.  It really only works if
4499          the primops involved are OkForSpeculation, and, since
4500         they have side effects readIntOfAddr and touch are not.
4501         Maybe we'll get back to this later .  -}
4502   
4503 Consider
4504    f (case readIntOffAddr# p# i# realWorld# of { (# s#, n# #) ->
4505       case touch# fp s# of { _ ->
4506       I# n# } } )
4507 This happened in a tight loop generated by stream fusion that
4508 Roman encountered.  We'd like to treat this just like the let
4509 case, because the primops concerned are ok-for-speculation.
4510 That is, we'd like to behave as if it had been
4511    case readIntOffAddr# p# i# realWorld# of { (# s#, n# #) ->
4512    case touch# fp s# of { _ ->
4513    f (I# n# } } )
4514]
4515[Comments only
4516simonpj@microsoft.com**20100504163629
4517 Ignore-this: 3be12df04714aa820bce706b5dc8a9cb
4518]
4519[Comments only
4520simonpj@microsoft.com**20100504163529
4521 Ignore-this: 791e2fd39c7d880ce1dc80ebdf3a5398
4522]
4523[Comments only
4524simonpj@microsoft.com**20100504163457
4525 Ignore-this: f19e9ffeb3d65770b1595bca5f97a59d
4526]
4527[Comments only (about type families)
4528simonpj@microsoft.com**20100417145032
4529 Ignore-this: dd39425ef2155d52dbf55a4d5fd97cb8
4530]
4531[Fix hp2ps when the .hp file has large string literals
4532Ian Lynagh <igloo@earth.li>**20100505191921]
4533[In build system, call package-config after including package data
4534Ian Lynagh <igloo@earth.li>**20100504225035
4535 Otherwise the $1_$2_HC_OPTS variable gets clobbered.
4536]
4537[runghc: flush stdout/stderr on an exception (#3890)
4538Simon Marlow <marlowsd@gmail.com>**20100505133848
4539 Ignore-this: 224c1898cec64cb1c94e0d7033e7590e
4540]
4541[Remove the Unicode alternative for ".." (#3894)
4542Simon Marlow <marlowsd@gmail.com>**20100505121202
4543 Ignore-this: 2452cd67281667106f9169747b6d784f
4544]
4545[tidyup; no functional changes
4546Simon Marlow <marlowsd@gmail.com>**20100505115015
4547 Ignore-this: d0787e5cdeef1dee628682fa0a46019
4548]
4549[Make the running_finalizers flag task-local
4550Simon Marlow <marlowsd@gmail.com>**20100505114947
4551 Ignore-this: 345925d00f1dca203941b3c5d84c90e1
4552 Fixes a bug reported by Lennart Augustsson, whereby we could get an
4553 incorrect error from the RTS about re-entry from a finalizer,
4554]
4555[add a MAYBE_GC() in killThread#, fixes throwto003(threaded2) looping
4556Simon Marlow <marlowsd@gmail.com>**20100505114746
4557 Ignore-this: efea04991d6feed04683a42232fc85da
4558]
4559[Allow filepath-1.2.*
4560Simon Marlow <marlowsd@gmail.com>**20100505101139
4561 Ignore-this: 1b5580cd9cd041ec48f40cd37603326a
4562]
4563[BlockedOnMsgThrowTo is possible in resurrectThreads (#4030)
4564Simon Marlow <marlowsd@gmail.com>**20100505094534
4565 Ignore-this: ac24a22f95ffeaf480187a1620fdddb2
4566]
4567[Don't raise a throwTo when the target is masking and BlockedOnBlackHole
4568Simon Marlow <marlowsd@gmail.com>**20100505094506
4569 Ignore-this: 302616931f61667030d77ddfbb02374e
4570]
4571[Fix build with GHC 6.10
4572Ian Lynagh <igloo@earth.li>**20100504180302
4573 In GHC 6.10, intersectionWith is (a -> b -> a) instead of (a -> b -> c),
4574 so we need to jump through some hoops to get the more general type.
4575]
4576[The libffi patches are no longer needed
4577Ian Lynagh <igloo@earth.li>**20100504171603]
4578[Use the in-tree windres; fixes trac #4032
4579Ian Lynagh <igloo@earth.li>**20100504170941]
4580[Print unfoldings on lambda-bound variables
4581Simon PJ <simonpj@microsoft.com>**20100503181822
4582 Ignore-this: 2fd5a7502cc6273d96258e0914f0f8cd
4583 
4584 ...in the unusual case where they have one;
4585 see Note [Case binders and join points] in Simplify.lhs
4586]
4587[Replace FiniteMap and UniqFM with counterparts from containers.
4588Milan Straka <fox@ucw.cz>**20100503171315
4589 Ignore-this: a021972239163dbf728284b19928cebb
4590 
4591 The original interfaces are kept. There is small performance improvement:
4592 - when compiling for five nofib, we get following speedups:
4593     Average                -----           -2.5%
4594     Average                -----           -0.6%
4595     Average                -----           -0.5%
4596     Average                -----           -5.5%
4597     Average                -----          -10.3%
4598 - when compiling HPC ten times, we get:
4599     switches                          oldmaps   newmaps
4600     -O -fasm                          117.402s  116.081s (98.87%)
4601     -O -fasm -fregs-graph             119.993s  118.735s (98.95%)
4602     -O -fasm -fregs-iterative         120.191s  118.607s (98.68%)
4603]
4604[Make the demand analyser take account of lambda-bound unfoldings
4605Simon PJ <simonpj@microsoft.com>**20100503151630
4606 Ignore-this: 2ee8e27d4df2debfc79e6b8a17c32bc1
4607 
4608 This is a long-standing lurking bug. See Note [Lamba-bound unfoldings]
4609 in DmdAnal.
4610 
4611 I'm still not really happy with this lambda-bound-unfolding stuff.
4612]
4613[Fix dynamic libs on OS X, and enable them by default
4614Ian Lynagh <igloo@earth.li>**20100503150302]
4615[Switch back to using bytestring from the darcs repo; partially fixes #3855
4616Ian Lynagh <igloo@earth.li>**20100502113458]
4617[Fix some cpp warnings when building on FreeBSD; patch from Gabor PALI
4618Ian Lynagh <igloo@earth.li>**20100428150700]
4619[Fix "make 2"
4620Ian Lynagh <igloo@earth.li>**20100427162212
4621 The new Makefile logic was enabling the stage 1 rules when stage=2,
4622 so "make 2" was rebuilding stage 1.
4623]
4624[Inplace programs depend on their shell wrappers
4625Ian Lynagh <igloo@earth.li>**20100427160038]
4626[--make is now the default (#3515), and -fno-code works with --make (#3783)
4627Simon Marlow <marlowsd@gmail.com>**20100427122851
4628 Ignore-this: 33330474fa4703f32bf9997462b4bf3c
4629 If the command line contains any Haskell source files, then we behave
4630 as if --make had been given.
4631 
4632 The meaning of the -c flag has changed (back): -c now selects one-shot
4633 compilation, but stops before linking.  However, to retain backwards
4634 compatibility, -c is still allowed with --make, and means the same as
4635 --make -no-link.  The -no-link flag has been un-deprecated.
4636 
4637 -fno-code is now allowed with --make (#3783); the fact that it was
4638 disabled before was largely accidental, it seems.  We also had some
4639 regressions in this area: it seems that -fno-code was causing a .hc
4640 file to be emitted in certain cases.  I've tidied up the code, there
4641 was no need for -fno-code to be a "mode" flag, as far as I can tell.
4642 
4643 -fno-code does not emit interface files, nor does it do recompilation
4644 checking, as suggested in #3783.  This would make Haddock emit
4645 interface files, for example, and I'm fairly sure we don't want to do
4646 that.  Compiling with -fno-code is pretty quick anyway, perhaps we can
4647 get away without recompilation checking.
4648]
4649[remove duplicate docs for -e in --help output (#4010)
4650Simon Marlow <marlowsd@gmail.com>**20100426140642
4651 Ignore-this: 187ff893ba8ffa0ec127867a7590e38d
4652]
4653[workaround for #4003, fixes HEAD build with 6.12.2
4654Simon Marlow <marlowsd@gmail.com>**20100426103428
4655 Ignore-this: c4bc445dc8052d4e6efef3f1daf63562
4656]
4657[Make sure all the clean rules are always included
4658Ian Lynagh <igloo@earth.li>**20100424181823
4659 In particular, this fixes a problem where stage3 bits weren't being cleaned
4660]
4661[Correct the name of the amd64/FreeBSD platform in PlatformSupportsSharedLibs
4662Ian Lynagh <igloo@earth.li>**20100424132830
4663 We weren't getting sharedlibs on amd64/FreeBSD because of this
4664]
4665[Include DPH docs in bindists
4666Ian Lynagh <igloo@earth.li>**20100424123101]
4667[reinstate eta-expansion during SimplGently, to fix inlining of sequence_
4668Simon Marlow <marlowsd@gmail.com>**20100423124853
4669 Ignore-this: 4fa0fd5bafe0d6b58fc81076f50d5f8d
4670]
4671[fix 64-bit value for W_SHIFT, which thankfully appears to be not used
4672Simon Marlow <marlowsd@gmail.com>**20100422213605
4673 Ignore-this: 525c062d2456c224ec8d0e083edd3b55
4674]
4675[Add missing constant folding and optimisation for unsigned division
4676Simon Marlow <marlowsd@gmail.com>**20100422213443
4677 Ignore-this: fb10d1cda0852fab0cbcb47247498fb3
4678 Noticed by Denys Rtveliashvili <rtvd@mac.com>, see #4004
4679]
4680[Fix the GHC API link in the main doc index.html
4681Ian Lynagh <igloo@earth.li>**20100422213226]
4682[Give the right exit code in darcs-all
4683Ian Lynagh <igloo@earth.li>**20100421171339
4684 Our END block was calling system, which alters $?. So now we save and
4685 restore it.
4686]
4687[Use StgWord64 instead of ullong
4688Ian Lynagh <igloo@earth.li>**20100421162336
4689 This patch also fixes ullong_format_string (renamed to showStgWord64)
4690 so that it works with values outside the 32bit range (trac #3979), and
4691 simplifies the without-commas case.
4692]
4693[Implement try10Times in Makefile
4694Ian Lynagh <igloo@earth.li>**20100420165909
4695 Avoid using seq, as FreeBSD has jot instead.
4696]
4697[Fix crash in non-threaded RTS on Windows
4698Simon Marlow <marlowsd@gmail.com>**20100420122125
4699 Ignore-this: 28b0255a914a8955dce02d89a7dfaca
4700 The tso->block_info field is now overwritten by pushOnRunQueue(), but
4701 stg_block_async_info was assuming that it still held a pointer to the
4702 StgAsyncIOResult.  We must therefore save this value somewhere safe
4703 before putting the TSO on the run queue.
4704]
4705[Expand the scope of the event_buf_mutex to cover io_manager_event
4706Simon Marlow <marlowsd@gmail.com>**20100420122026
4707 Ignore-this: 185a6d84f7d4a35997f10803f6dacef1
4708 I once saw a failure that I think was due to a race on
4709 io_manager_event, this should fix it.
4710]
4711[Flags -auto and -auto-all operate only on functions not marked INLINE.
4712Milan Straka <fox@ucw.cz>**20100331191050
4713 Ignore-this: 3b63580cfcb3c33d62ad697c36d94d05
4714]
4715[Spelling correction for LANGUAGE pragmas
4716Max Bolingbroke <batterseapower@hotmail.com>**20100413192825
4717 Ignore-this: 311b51ba8d43f6c7fd32f48db9a88dee
4718]
4719[Update the user guide so it talks about the newer "do rec" notation everywhere
4720Ian Lynagh <igloo@earth.li>**20100416205416
4721 Some of the problems highlighted in trac #3968.
4722]
4723[Fix typo
4724Ian Lynagh <igloo@earth.li>**20100416205412]
4725[Fix Trac #3950: unifying types of different kinds
4726simonpj@microsoft.com**20100412151845
4727 Ignore-this: d145b9de5ced136ef2c39f3ea4a04f4a
4728 
4729 I was assuming that the unifer only unified types of the
4730 same kind, but now we can "defer" unsolved constraints that
4731 invariant no longer holds.  Or at least is's more complicated
4732 to ensure. 
4733 
4734 This patch takes the path of not assuming the invariant, which
4735 is simpler and more robust.  See
4736 Note [Mismatched type lists and application decomposition]
4737]
4738[Fix Trac #3943: incorrect unused-variable warning
4739simonpj@microsoft.com**20100412151630
4740 Ignore-this: 52459f2b8b02c3cb120abe674dc9a060
4741 
4742 In fixing this I did the usual little bit of refactoring
4743]
4744[Convert boot and boot-pkgs to perl
4745Ian Lynagh <igloo@earth.li>**20100415143919
4746 This stops us having to worry about sh/sed/... portability.
4747]
4748[Use $(MAKE), not make, when recursively calling make
4749Ian Lynagh <igloo@earth.li>**20100415121453]
4750[Remove the ghc_ge_609 makefile variables
4751Ian Lynagh <igloo@earth.li>**20100412235658
4752 They are now guaranteed to be YES
4753]
4754[Increase the minimum version number required to 6.10 in configure.ac
4755Ian Lynagh <igloo@earth.li>**20100412235313]
4756[The bootstrapping compiler is now required to be > 609
4757Ian Lynagh <igloo@earth.li>**20100409161046]
4758[Handle IND_STATIC in isRetainer
4759Ian Lynagh <igloo@earth.li>**20100409104207
4760 IND_STATIC used to be an error, but at the moment it can happen
4761 as isAlive doesn't look through IND_STATIC as it ignores static
4762 closures. See trac #3956 for a program that hit this error.
4763]
4764[Add Data and Typeable instances to HsSyn
4765David Waern <david.waern@gmail.com>**20100330011020
4766 Ignore-this: c3f2717207b15539fea267c36b686e6a
4767 
4768 The instances (and deriving declarations) have been taken from the ghc-syb
4769 package.
4770]
4771[Fix for derefing ThreadRelocated TSOs in MVar operations
4772Simon Marlow <marlowsd@gmail.com>**20100407092824
4773 Ignore-this: 94dd7c68a6094eda667e2375921a8b78
4774]
4775[sanity check fix
4776Simon Marlow <marlowsd@gmail.com>**20100407092746
4777 Ignore-this: 9c18cd5f5393e5049015ca52e62a1269
4778]
4779[get the reg liveness right in the putMVar# heap check
4780Simon Marlow <marlowsd@gmail.com>**20100407092724
4781 Ignore-this: b1ba07a59ecfae00e9a1f8391741abc
4782]
4783[initialise the headers of MSG_BLACKHOLE objects properly
4784Simon Marlow <marlowsd@gmail.com>**20100407081712
4785 Ignore-this: 183dcd0ca6a395d08db2be12b02bdd79
4786]
4787[initialise the headers of MVAR_TSO_QUEUE objects properly
4788Simon Marlow <marlowsd@gmail.com>**20100407081514
4789 Ignore-this: 4b4a2f30cf2fb69ca4128c41744687bb
4790]
4791[undo debugging code
4792Simon Marlow <marlowsd@gmail.com>**20100406142740
4793 Ignore-this: 323c2248f817b6717c19180482fc4b00
4794]
4795[putMVar#: fix reg liveness in the heap check
4796Simon Marlow <marlowsd@gmail.com>**20100406135832
4797 Ignore-this: cddd2c7807ac7612c9b2c4c0d384d284
4798]
4799[account for the new BLACKHOLEs in the GHCi debugger
4800Simon Marlow <marlowsd@gmail.com>**20100406133406
4801 Ignore-this: 4d4aeb4bbada3f50dc1fb0123f565e8f
4802]
4803[don't forget to deRefTSO() in tryWakeupThread()
4804Simon Marlow <marlowsd@gmail.com>**20100406130411
4805 Ignore-this: 171d57c4f8653835dec0b69f9be9881c
4806]
4807[Fix bug in popRunQueue
4808Simon Marlow <marlowsd@gmail.com>**20100406091453
4809 Ignore-this: 9d3cec8f18f5c5cbd51751797386eb6f
4810]
4811[fix bug in migrateThread()
4812Simon Marlow <marlowsd@gmail.com>**20100401105840
4813 Ignore-this: 299bcf0d1ea0f8865f3e845eb93d2ad3
4814]
4815[Remove the IND_OLDGEN and IND_OLDGEN_PERM closure types
4816Simon Marlow <marlowsd@gmail.com>**20100401093519
4817 Ignore-this: 95f2480c8a45139835eaf5610217780b
4818 These are no longer used: once upon a time they used to have different
4819 layout from IND and IND_PERM respectively, but that is no longer the
4820 case since we changed the remembered set to be an array of addresses
4821 instead of a linked list of closures.
4822]
4823[Change the representation of the MVar blocked queue
4824Simon Marlow <marlowsd@gmail.com>**20100401091605
4825 Ignore-this: 20a35bfabacef2674df362905d7834fa
4826 
4827 The list of threads blocked on an MVar is now represented as a list of
4828 separately allocated objects rather than being linked through the TSOs
4829 themselves.  This lets us remove a TSO from the list in O(1) time
4830 rather than O(n) time, by marking the list object.  Removing this
4831 linear component fixes some pathalogical performance cases where many
4832 threads were blocked on an MVar and became unreachable simultaneously
4833 (nofib/smp/threads007), or when sending an asynchronous exception to a
4834 TSO in a long list of thread blocked on an MVar.
4835 
4836 MVar performance has actually improved by a few percent as a result of
4837 this change, slightly to my surprise.
4838 
4839 This is the final cleanup in the sequence, which let me remove the old
4840 way of waking up threads (unblockOne(), MSG_WAKEUP) in favour of the
4841 new way (tryWakeupThread and MSG_TRY_WAKEUP, which is idempotent).  It
4842 is now the case that only the Capability that owns a TSO may modify
4843 its state (well, almost), and this simplifies various things.  More of
4844 the RTS is based on message-passing between Capabilities now.
4845]
4846[eliminate some duplication with a bit of CPP
4847Simon Marlow <marlowsd@gmail.com>**20100330154355
4848 Ignore-this: 838f7d341f096ca14c86ab9c81193e36
4849]
4850[Make ioManagerDie() idempotent
4851Simon Marlow <marlowsd@gmail.com>**20100401100705
4852 Ignore-this: a5996b43cdb2e2d72e6e971d7ea925fb
4853 Avoids screeds of "event buffer overflowed; event dropped" in
4854 conc059(threaded1).
4855]
4856[Move a thread to the front of the run queue when another thread blocks on it
4857Simon Marlow <marlowsd@gmail.com>**20100329144521
4858 Ignore-this: c518ff0d41154680edc811d891826a29
4859 This fixes #3838, and was made possible by the new BLACKHOLE
4860 infrastructure.  To allow reording of the run queue I had to make it
4861 doubly-linked, which entails some extra trickiness with regard to
4862 GC write barriers and suchlike.
4863]
4864[remove non-existent MUT_CONS symbols
4865Simon Marlow <marlowsd@gmail.com>**20100330152600
4866 Ignore-this: 885628257a9d03f2ece2a754d993014a
4867]
4868[change throwTo to use tryWakeupThread rather than unblockOne
4869Simon Marlow <marlowsd@gmail.com>**20100329144613
4870 Ignore-this: 10ad4965e6c940db71253f1c72218bbb
4871]
4872[tiny GC optimisation
4873Simon Marlow <marlowsd@gmail.com>**20100329144551
4874 Ignore-this: 9e095b9b73fff0aae726f9937846ba92
4875]
4876[New implementation of BLACKHOLEs
4877Simon Marlow <marlowsd@gmail.com>**20100329144456
4878 Ignore-this: 96cd26793b4e6ab9ddd0d59aae5c2f1d
4879 
4880 This replaces the global blackhole_queue with a clever scheme that
4881 enables us to queue up blocked threads on the closure that they are
4882 blocked on, while still avoiding atomic instructions in the common
4883 case.
4884 
4885 Advantages:
4886 
4887  - gets rid of a locked global data structure and some tricky GC code
4888    (replacing it with some per-thread data structures and different
4889    tricky GC code :)
4890 
4891  - wakeups are more prompt: parallel/concurrent performance should
4892    benefit.  I haven't seen anything dramatic in the parallel
4893    benchmarks so far, but a couple of threading benchmarks do improve
4894    a bit.
4895 
4896  - waking up a thread blocked on a blackhole is now O(1) (e.g. if
4897    it is the target of throwTo).
4898 
4899  - less sharing and better separation of Capabilities: communication
4900    is done with messages, the data structures are strictly owned by a
4901    Capability and cannot be modified except by sending messages.
4902 
4903  - this change will utlimately enable us to do more intelligent
4904    scheduling when threads block on each other.  This is what started
4905    off the whole thing, but it isn't done yet (#3838).
4906 
4907 I'll be documenting all this on the wiki in due course.
4908 
4909]
4910[Fix warnings (allow pushOnRunQueue() to not be inlined)
4911Simon Marlow <marlowsd@gmail.com>**20100401114559
4912 Ignore-this: f40bfbfad70a5165a946d11371605b7d
4913]
4914[remove out of date comment
4915Simon Marlow <marlowsd@gmail.com>**20100401105853
4916 Ignore-this: 26af88dd418ee0bcda7223b3b7e4e8d2
4917]
4918[tidy up spacing in stderr traces
4919Simon Marlow <marlowsd@gmail.com>**20100326163122
4920 Ignore-this: 16558b0433a274be217d4bf39aa4946
4921]
4922[Fix an assertion that was not safe when running in parallel
4923Simon Marlow <marlowsd@gmail.com>**20100325143656
4924 Ignore-this: cad08fb8900eb3a475547af0189fcc47
4925]
4926[Never jump directly to a thunk's entry code, even if it is single-entry
4927Simon Marlow <marlowsd@gmail.com>**20100325114847
4928 Ignore-this: 938da172c06a97762ef605c8fccfedf1
4929 I don't think this fixes any bugs as we don't have single-entry thunks
4930 at the moment, but it could cause problems for parallel execution if
4931 we ever did re-introduce update avoidance.
4932]
4933[Rename forgotten -dverbose-simpl to -dverbose-core2core in the docs.
4934Milan Straka <fox@ucw.cz>**20100331153626
4935 Ignore-this: 2da58477fb96e1cfb80f37dddd7c422c
4936]
4937[Add -pa and -V to the documentation of time profiling options.
4938Milan Straka <fox@ucw.cz>**20100329191121
4939 Ignore-this: be74d216481ec5a19e5f40f85e6e3d65
4940]
4941[Keep gcc 4.5 happy
4942Simon Marlow <marlowsd@gmail.com>**20100330120425
4943 Ignore-this: 7811878cc2bd1ce9cfbb5bf102fe3454
4944]
4945[Fix warning compiling Linker.c for PPC Mac
4946naur@post11.tele.dk**20100403182355
4947 Ignore-this: e2d2448770c9714ce17dd6cf3e297063
4948 The warning message eliminated is:
4949 > rts/Linker.c:4756:0:
4950 >      warning: nested extern declaration of 'symbolsWithoutUnderscore'
4951]
4952[Fix error compiling AsmCodeGen.lhs for PPC Mac (mkRtsCodeLabel)
4953naur@post11.tele.dk**20100403181656
4954 Ignore-this: deb7524ea7852a15a2ac0849c8c82f74
4955 The error messages eliminated are:
4956 > compiler/nativeGen/AsmCodeGen.lhs:875:31:
4957 >     Not in scope: `mkRtsCodeLabel'
4958 > compiler/nativeGen/AsmCodeGen.lhs:879:31:
4959 >     Not in scope: `mkRtsCodeLabel'
4960 > compiler/nativeGen/AsmCodeGen.lhs:883:31:
4961 >     Not in scope: `mkRtsCodeLabel'
4962]
4963[Fix error compiling AsmCodeGen.lhs for PPC Mac (DestBlockId)
4964naur@post11.tele.dk**20100403180643
4965 Ignore-this: 71e833e94ed8371b2ffabc2cf80bf585
4966 The error message eliminated is:
4967 > compiler/nativeGen/AsmCodeGen.lhs:637:16:
4968 >     Not in scope: data constructor `DestBlockId'
4969]
4970[Fix boot-pkgs's sed usage to work with Solaris's sed
4971Ian Lynagh <igloo@earth.li>**20100401153441]
4972[Pass "-i org.haskell.GHC" to packagemaker when building the OS X installer
4973Ian Lynagh <igloo@earth.li>**20100331144707
4974 This seems to fix this failure:
4975 [...]
4976 ** BUILD SUCCEEDED **
4977 rm -f -f GHC-system.pmdoc/*-contents.xml
4978 /Developer/usr/bin/packagemaker -v --doc GHC-system.pmdoc\
4979              -o /Users/ian/to_release/ghc-6.12.1.20100330/GHC-6.12.1.20100330-i386.pkg
4980 2010-03-31 15:08:15.695 packagemaker[13909:807] Setting to : 0 (null)
4981 2010-03-31 15:08:15.709 packagemaker[13909:807] Setting to : 0 org.haskell.glasgowHaskellCompiler.ghc.pkg
4982 2010-03-31 15:08:15.739 packagemaker[13909:807] relocate: (null) 0
4983 2010-03-31 15:08:15.740 packagemaker[13909:807] *** Terminating app due to uncaught exception 'NSInvalidArgumentException', reason: '*** -[NSXMLDocument initWithXMLString:options:error:]: nil argument'
4984 2010-03-31 15:08:15.741 packagemaker[13909:807] Stack: (
4985     2511962091,
4986     2447007291,
4987     2511961547,
4988     2511961610,
4989     2432803204,
4990     453371,
4991     447720,
4992     436209,
4993     435510,
4994     9986,
4995     9918
4996 )
4997 make[1]: *** [framework-pkg] Trace/BPT trap
4998 make: *** [framework-pkg] Error 2
4999]
5000[Use machdepCCOpts when compiling the file to toggle -(no-)rtsopts
5001Ian Lynagh <igloo@earth.li>**20100331161302
5002 Should fix toggling on OS X "Snow Leopard". Diagnosed by Roman Leshchinskiy.
5003]
5004[Avoid a non-portable use of tar reported by Roman Leshchinskiy
5005Ian Lynagh <igloo@earth.li>**20100330145802]
5006[Don't install EXTRA_PACKAGES by default
5007Simon Marlow <marlowsd@gmail.com>**20100330142714
5008 Ignore-this: d4cc8f87a6de8d9d1d6dc9b77130b3
5009]
5010[fix a non-portable printf format
5011Simon Marlow <marlowsd@gmail.com>**20100330134437
5012 Ignore-this: d41c23c54ec29654cb2049de1e588570
5013]
5014[avoid single quote in #error
5015Simon Marlow <marlowsd@gmail.com>**20100330120346
5016 Ignore-this: 663f39e7a27fead2f648fbf22d345bb4
5017]
5018[use FMT_Word64 instead of locally-defined version
5019Simon Marlow <marlowsd@gmail.com>**20100330114650
5020 Ignore-this: 82697b8095dffb3a8e196c687006ece0
5021]
5022[remove old/unused DotnetSupport and GhcLibsWithUnix
5023Simon Marlow <marlowsd@gmail.com>**20100330123732
5024 Ignore-this: c68814868b3671abdc369105bbeafe6c
5025]
5026[fix return type cast in f.i.wrapper when using libffi (#3516)
5027Simon Marlow <marlowsd@gmail.com>**20100329154220
5028 Ignore-this: f898eb8c9ae2ca2009e539735b92c438
5029 
5030 Original fix submitted by
5031   Sergei Trofimovich <slyfox@community.haskell.org>
5032 modified by me:
5033  - exclude 64-bit types
5034  - compare uniques, not strings
5035  - #include "ffi.h" is conditional
5036]
5037[libffi: install 'ffitarget.h' header as sole 'ffi.h' is unusable
5038Simon Marlow <marlowsd@gmail.com>**20100329135734
5039 Ignore-this: f9b555ea289d8df1aa22cb6faa219a39
5040 Submitted by: Sergei Trofimovich <slyfox@community.haskell.org>
5041 Re-recorded against HEAD.
5042]
5043[avoid a fork deadlock (see comments)
5044Simon Marlow <marlowsd@gmail.com>**20100329132329
5045 Ignore-this: 3377f88b83bb3b21e42d7fc5f0d866f
5046]
5047[tidy up the end of the all_tasks list after forking
5048Simon Marlow <marlowsd@gmail.com>**20100329132253
5049 Ignore-this: 819d679875be5f344e816210274d1c29
5050]
5051[Add a 'setKeepCAFs' external function (#3900)
5052Simon Marlow <marlowsd@gmail.com>**20100329110036
5053 Ignore-this: ec532a18cad4259a09847b0b9ae2e1d2
5054]
5055[Explicitly check whether ar supports the @file syntax
5056Ian Lynagh <igloo@earth.li>**20100329123325
5057 rather than assuming that all GNU ar's do.
5058 Apparently OpenBSD's older version doesn't.
5059]
5060[Fix the format specifier for Int64/Word64 on Windows
5061Ian Lynagh <igloo@earth.li>**20100327182126
5062 mingw doesn't understand %llu/%lld - it treats them as 32-bit rather
5063 than 64-bit. We use %I64u/%I64d instead.
5064]
5065[Fix the ghci startmenu item
5066Ian Lynagh <igloo@earth.li>**20100326235934
5067 I'm not sure what changed, but it now doesn't work for me without
5068 the "Start in" field being set.
5069]
5070[Fix paths to docs in "Start Menu" entries in Windows installer; fixes #3847
5071Ian Lynagh <igloo@earth.li>**20100326155917]
5072[Add a licence file for the Windows installer to use
5073Ian Lynagh <igloo@earth.li>**20100326155130]
5074[Add gcc-g++ to the inplace mingw installation; fixes #3893
5075Ian Lynagh <igloo@earth.li>**20100326154714]
5076[Add the licence file to the Windows installer. Fixes #3934
5077Ian Lynagh <igloo@earth.li>**20100326152449]
5078[Quote the paths to alex and happy in configure
5079Ian Lynagh <igloo@earth.li>**20100325143449
5080 Ignore-this: d6d6e1a250f88985bbeea760e63a79db
5081]
5082[Use </> rather than ++ "/"
5083Ian Lynagh <igloo@earth.li>**20100325133237
5084 This stops us generating paths like
5085     c:\foo\/ghc460_0/ghc460_0.o
5086 which windres doesn't understand.
5087]
5088[Append $(exeext) to utils/ghc-pkg_dist_PROG
5089Ian Lynagh <igloo@earth.li>**20100324233447
5090 Fixes bindist creation
5091]
5092[A sanity check
5093Simon Marlow <marlowsd@gmail.com>**20100325110500
5094 Ignore-this: 3b3b76d898c822456857e506b7531e65
5095]
5096[do_checks: do not set HpAlloc if the stack check fails
5097Simon Marlow <marlowsd@gmail.com>**20100325110328
5098 Ignore-this: 899ac8c29ca975d03952dbf4608d758
5099 
5100 This fixes a very rare heap corruption bug, whereby
5101 
5102  - a context switch is requested, which sets HpLim to zero
5103    (contextSwitchCapability(), called by the timer signal or
5104    another Capability).
5105 
5106  - simultaneously a stack check fails, in a code fragment that has
5107    both a stack and a heap check.
5108 
5109 The RTS then assumes that a heap-check failure has occurred and
5110 subtracts HpAlloc from Hp, although in fact it was a stack-check
5111 failure and retreating Hp will overwrite valid heap objects.  The bug
5112 is that HpAlloc should only be set when Hp has been incremented by the
5113 heap check.  See comments in rts/HeapStackCheck.cmm for more details.
5114 
5115 This bug is probably incredibly rare in practice, but I happened to be
5116 working on a test that triggers it reliably:
5117 concurrent/should_run/throwto001, compiled with -O -threaded, args 30
5118 300 +RTS -N2, run repeatedly in a loop.
5119]
5120[comments and formatting only
5121Simon Marlow <marlowsd@gmail.com>**20100325104617
5122 Ignore-this: c0a211e15b5953bb4a84771bcddd1d06
5123]
5124[Change how perl scripts get installed; partially fixes #3863
5125Ian Lynagh <igloo@earth.li>**20100324171422
5126 We now regenerate them when installing, which means the path for perl
5127 doesn't get baked in
5128]
5129[Pass the location of gcc in the ghc wrapper script; partially fixes #3863
5130Ian Lynagh <igloo@earth.li>**20100324171408
5131 This means we don't rely on baking a path to gcc into the executable
5132]
5133[Quote the ar path in configure
5134Ian Lynagh <igloo@earth.li>**20100324162043]
5135[Remove unused cUSER_WAY_NAMES cUSER_WAY_OPTS
5136Ian Lynagh <igloo@earth.li>**20100324145048]
5137[Remove unused cCONTEXT_DIFF
5138Ian Lynagh <igloo@earth.li>**20100324145013]
5139[Remove unused cEnableWin32DLLs
5140Ian Lynagh <igloo@earth.li>**20100324144841]
5141[Remove unused cGHC_CP
5142Ian Lynagh <igloo@earth.li>**20100324144656]
5143[Fix the build for non-GNU-ar
5144Ian Lynagh <igloo@earth.li>**20100324132907]
5145[Tweak the Makefile code for making .a libs; fixes trac #3642
5146Ian Lynagh <igloo@earth.li>**20100323221325
5147 The main change is that, rather than using "xargs ar" we now put
5148 all the filenames into a file, and do "ar @file". This means that
5149 ar adds all the files at once, which works around a problem where
5150 files with the same basename in a later invocation were overwriting
5151 the existing file in the .a archive.
5152]
5153[Enable shared libraries on Windows; fixes trac #3879
5154Ian Lynagh <igloo@earth.li>**20100320231414
5155 Ignore-this: c93b35ec5b7a7fa6ddb286d17a616216
5156]
5157[Add the external core PDF to the new build system
5158Ian Lynagh <igloo@earth.li>**20100321161909]
5159[Allow specifying $threads directly when validating
5160Ian Lynagh <igloo@earth.li>**20100321112835]
5161[Remove LazyUniqFM; fixes trac #3880
5162Ian Lynagh <igloo@earth.li>**20100320213837]
5163[UNDO: slight improvement to scavenging ...
5164Simon Marlow <marlowsd@gmail.com>**20100319153413
5165 Ignore-this: f0ab581c07361f7b57eae02dd6ec893c
5166 
5167 Accidnetally pushed this patch which, while it validates, isn't
5168 correct.
5169 
5170 rolling back:
5171 
5172 Fri Mar 19 11:21:27 GMT 2010  Simon Marlow <marlowsd@gmail.com>
5173   * slight improvement to scavenging of update frames when a collision has occurred
5174 
5175     M ./rts/sm/Scav.c -19 +15
5176]
5177[slight improvement to scavenging of update frames when a collision has occurred
5178Simon Marlow <marlowsd@gmail.com>**20100319112127
5179 Ignore-this: 6de2bb9614978975f17764a0f259d9bf
5180]
5181[Don't install the utf8-string package
5182Ian Lynagh <igloo@earth.li>**20100317212709]
5183[Don't use -Bsymbolic when linking the RTS
5184Ian Lynagh <igloo@earth.li>**20100316233357
5185 This makes the RTS hooks work when doing dynamic linking
5186]
5187[Fix Trac #3920: Template Haskell kinds
5188simonpj@microsoft.com**20100317123519
5189 Ignore-this: 426cac7920446e04f3cc30bd1d9f76e2
5190 
5191 Fix two places where we were doing foldl instead of foldr
5192 after decomposing a Kind.  Strange that the same bug appears
5193 in two quite different places!
5194]
5195[copy_tag_nolock(): fix write ordering and add a write_barrier()
5196Simon Marlow <marlowsd@gmail.com>**20100316143103
5197 Ignore-this: ab7ca42904f59a0381ca24f3eb38d314
5198 
5199 Fixes a rare crash in the parallel GC.
5200 
5201 If we copy a closure non-atomically during GC, as we do for all
5202 immutable values, then before writing the forwarding pointer we better
5203 make sure that the closure itself is visible to other threads that
5204 might follow the forwarding pointer.  I imagine this doesn't happen
5205 very often, but I just found one case of it: in scavenge_stack, the
5206 RET_FUN case, after evacuating ret_fun->fun we then follow it and look
5207 up the info pointer.
5208]
5209[Add sliceP mapping to vectoriser builtins
5210benl@ouroborus.net**20100316060517
5211 Ignore-this: 54c3cafff584006b6fbfd98124330aa3
5212]
5213[Comments only
5214benl@ouroborus.net**20100311064518
5215 Ignore-this: d7dc718cc437d62aa5b1b673059a9b22
5216]
5217[TAG 2010-03-16
5218Ian Lynagh <igloo@earth.li>**20100316005137
5219 Ignore-this: 234e3bc29e2f26cc59d7b03d780cc352
5220]
5221Patch bundle hash:
52220745eba3a27dc4f9d05ca9b6e31568a9ece47ca4