New patches: [Fix #2759: add ability to serialize Rational 'Jose Pedro Magalhaes '**20081209125551] { hunk ./compiler/utils/Serialized.hs 89 -serializeConstr (FloatConstr d) = serializeWord8 3 . serializeDouble d +serializeConstr (FloatConstr r) = serializeWord8 3 . serializeRational r hunk ./compiler/utils/Serialized.hs 95 - 1 -> deserializeInt bytes $ \ix -> k (AlgConstr ix) - 2 -> deserializeInteger bytes $ \i -> k (IntConstr i) - 3 -> deserializeDouble bytes $ \d -> k (FloatConstr d) - 4 -> deserializeString bytes $ \s -> k (StringConstr s) + 1 -> deserializeInt bytes $ \ix -> k (AlgConstr ix) + 2 -> deserializeInteger bytes $ \i -> k (IntConstr i) + 3 -> deserializeRational bytes $ \r -> k (FloatConstr r) + 4 -> deserializeString bytes $ \s -> k (StringConstr s) hunk ./compiler/utils/Serialized.hs 143 -serializeDouble :: Double -> [Word8] -> [Word8] -serializeDouble = serializeString . show +serializeRational :: (Real a) => a -> [Word8] -> [Word8] +serializeRational = serializeString . show . toRational hunk ./compiler/utils/Serialized.hs 146 -deserializeDouble :: [Word8] -> (Double -> [Word8] -> a) -> a -deserializeDouble bytes k = deserializeString bytes (k . read) +deserializeRational :: (Fractional a) => [Word8] -> (a -> [Word8] -> b) -> b +deserializeRational bytes k = deserializeString bytes (k . fromRational . read) hunk ./compiler/utils/Serialized.hs 175 + + } Context: [Inject implicit bindings after CoreTidy, not before Simplify simonpj@microsoft.com**20081208173525 Originally I inject the "implicit bindings" (record selectors, class method selectors, data con wrappers...) after CoreTidy. However, in a misguided attempt to fix Trac #2070, I moved the injection point to before the Simplifier, so that record selectors would be optimised by the simplifier. This was misguided because record selectors (indeed all implicit bindings) are GlobalIds, whose IdInfo is meant to be frozen. But the Simplifier, and other Core-to-Core optimisations, merrily change the IdInfo. That ultimately made Trac #2844 happen, where a record selector got arity 2, but the GlobalId (which importing scopes re-construct from the class decl rather than reading from the interface file) has arity 1. So this patch moves the injection back to CoreTidy. Happily #2070 should still be OK because we now use CoreSubst.simpleOptExpr on the unfoldings for implict things, which gets rid of the most gratuitous infelicities. Still, there's a strong case for stoppping record selectors from being GlobalIds, and treating them much more like dict-funs. I'm thinking about that. Meanwhile, #2844 is ok now. ] [Add assertion for arity match (checks Trac #2844) simonpj@microsoft.com**20081208173241 The exported arity of a function must match the arity for the STG function. Trac #2844 was a pretty obscure manifestation of the failure of this invariant. This patch doesn't cure the bug; rather it adds an assertion to CoreToStg to check the invariant so we should get an earlier and less obscure warning if this fails in future. ] [Use CoreSubst.simpleOptExpr in place of the ad-hoc simpleSubst (reduces code too) simonpj@microsoft.com**20081208173018] [Move simpleOptExpr from CoreUnfold to CoreSubst simonpj@microsoft.com**20081208124840] [White space only simonpj@microsoft.com**20081208124310] [Comments only simonpj@microsoft.com**20081208124155] [Completely new treatment of INLINE pragmas (big patch) simonpj@microsoft.com**20081205165400 This is a major patch, which changes the way INLINE pragmas work. Although lots of files are touched, the net is only +21 lines of code -- and I bet that most of those are comments! HEADS UP: interface file format has changed, so you'll need to recompile everything. There is not much effect on overall performance for nofib, probably because those programs don't make heavy use of INLINE pragmas. Program Size Allocs Runtime Elapsed Min -11.3% -6.9% -9.2% -8.2% Max -0.1% +4.6% +7.5% +8.9% Geometric Mean -2.2% -0.2% -1.0% -0.8% (The +4.6% for on allocs is cichelli; see other patch relating to -fpass-case-bndr-to-join-points.) The old INLINE system ~~~~~~~~~~~~~~~~~~~~~ The old system worked like this. A function with an INLINE pragam got a right-hand side which looked like f = __inline_me__ (\xy. e) The __inline_me__ part was an InlineNote, and was treated specially in various ways. Notably, the simplifier didn't inline inside an __inline_me__ note. As a result, the code for f itself was pretty crappy. That matters if you say (map f xs), because then you execute the code for f, rather than inlining a copy at the call site. The new story: InlineRules ~~~~~~~~~~~~~~~~~~~~~~~~~~ The new system removes the InlineMe Note altogether. Instead there is a new constructor InlineRule in CoreSyn.Unfolding. This is a bit like a RULE, in that it remembers the template to be inlined inside the InlineRule. No simplification or inlining is done on an InlineRule, just like RULEs. An Id can have an InlineRule *or* a CoreUnfolding (since these are two constructors from Unfolding). The simplifier treats them differently: - An InlineRule is has the substitution applied (like RULES) but is otherwise left undisturbed. - A CoreUnfolding is updated with the new RHS of the definition, on each iteration of the simplifier. An InlineRule fires regardless of size, but *only* when the function is applied to enough arguments. The "arity" of the rule is specified (by the programmer) as the number of args on the LHS of the "=". So it makes a difference whether you say {-# INLINE f #-} f x = \y -> e or f x y = e This is one of the big new features that InlineRule gives us, and it is one that Roman really wanted. In contrast, a CoreUnfolding can fire when it is applied to fewer args than than the function has lambdas, provided the result is small enough. Consequential stuff ~~~~~~~~~~~~~~~~~~~ * A 'wrapper' no longer has a WrapperInfo in the IdInfo. Instead, the InlineRule has a field identifying wrappers. * Of course, IfaceSyn and interface serialisation changes appropriately. * Making implication constraints inline nicely was a bit fiddly. In the end I added a var_inline field to HsBInd.VarBind, which is why this patch affects the type checker slightly * I made some changes to the way in which eta expansion happens in CorePrep, mainly to ensure that *arguments* that become let-bound are also eta-expanded. I'm still not too happy with the clarity and robustness fo the result. * We now complain if the programmer gives an INLINE pragma for a recursive function (prevsiously we just ignored it). Reason for change: we don't want an InlineRule on a LoopBreaker, because then we'd have to check for loop-breaker-hood at occurrence sites (which isn't currenlty done). Some tests need changing as a result. This patch has been in my tree for quite a while, so there are probably some other minor changes. ] [Add -fpass-case-bndr-to-join-points simonpj@microsoft.com**20081205105159 See Note [Passing the case binder to join points] in Simplify.lhs The default now is *not* to pass the case binder. There are some nofib results with the above note; the effect is almost always negligible. I don't expect this flag to be used by users (hence no docs). It's just there to let me try the performance effects of switching on and off. ] [Add static flag -fsimple-list-literals simonpj@microsoft.com**20081205105002 The new static flag -fsimple-list-literals makes ExplicitList literals be desugared in the straightforward way, rather than using 'build' as now. See SLPJ comments with Note [Desugaring explicit lists]. I don't expect this flag to be used by users (hence no docs). It's just there to let me try the performance effects of switching on and off. ] [Comments only in OccurAnal simonpj@microsoft.com**20081205103252] [Comments only simonpj@microsoft.com**20081205102437] [Layout only simonpj@microsoft.com**20081205102252] [Comments only (Note [Entering error thunks]) simonpj@microsoft.com**20081205102149] [Make CoreToStg a little more robust to eta expansion simonpj@microsoft.com**20081205101932] [Add no-op case for addIdSpecialisations (very minor optimisation) simonpj@microsoft.com**20081205101022] [Trim redundant import simonpj@microsoft.com**20081205101006] [Make CoreTidy retain deadness info (better -ddump-simpl) simonpj@microsoft.com**20081205100518 GHC now retains more robust information about dead variables; but CoreTidy was throwing it away. This patch makes CoreTidy retain it, which gives better output for -ddump-simpl. New opportunity: shrink interface files by using wildcards for dead variables. ] [Remove INLINE pragmas on recursive functions simonpj@microsoft.com**20081205100353 INLINE pragmas on recursive functions are ignored; and this is checked in my upcoming patch for inlinings. ] [Comments only (on Activation) simonpj@microsoft.com**20081205100139] [We need to tell cabal-bin which version of Cabal to use Ian Lynagh **20081203123208 Otherwise, if the bootstrapping compiler has a newer version, we get a mismatch between the version used to compile ghc-prim's Setup.hs and the version that installPackage uses. ] [Document 'loadModule'. Thomas Schilling **20081202154800] [Add 'needsTemplateHaskell' utility function and document why one might Thomas Schilling **20081202152358 want to use it. ] [Documentation only. Thomas Schilling **20081202150158] [Export 'succeeded' and 'failed' helper functions. Thomas Schilling **20081202144451] [Put full ImportDecls in ModSummary instead of just ModuleNames Simon Marlow **20081202133736 ... and use it to make ghc -M generate correct cross-package dependencies when using package-qualified imports (needed for the new build system). Since we're already parsing the ImportDecl from the source file, there seems no good reason not to keep it in the ModSummary, it might be useful for other things too. ] [ghc -M: need to add a dep on Prelude unless -fno-implicit-prelude is on Simon Marlow **20081128165707] [make -include-pkg-deps work (not sure when this got lost) Simon Marlow **20081128135746] [Fix more problems caused by padding in the Capability structure Simon Marlow **20081202120735 Fixes crashes on Windows and Sparc ] [add missing case to Ord GlobalReg (EagerBlackhole == EagerBlackhole) Simon Marlow **20081128130106] [Better error message for fundep conflict simonpj@microsoft.com**20081201162845] [Fix typo in quasi-quote documentation's sample. shelarcy **20081129024344] [Remove the v_Split_info global variable and use a field of dflags instead Ian Lynagh **20081130152403] [Document the --machine-readable RTS flag Ian Lynagh **20081130152311] [Let 'loadModule' generate proper code depending on the 'hscTarget'. Thomas Schilling **20081128164412 With this change it should be possible to perform something similar to 'load' by traversing the module graph in dependency order and calling '{parse,typecheck,load}Module' on each. Of course, if you want smart recompilation checking you should still use 'load'. ] [Expose a separate 'hscBackend' phase for 'HsCompiler' and change Thomas Schilling **20081128163746 parameter to 'InteractiveStatus' to a 'Maybe'. ] [Whoops, *don't* reset the complete session in 'withLocalCallbacks'. Thomas Schilling **20081128150727] [Use a record instead of a typeclass for 'HsCompiler'. This is mostly Thomas Schilling **20081128121947 equivalent to a typeclass implementation that uses a functional dependency from the target mode to the result type. ] [Remove dead code Ian Lynagh **20081128193831] [Update docs not to talk about deprecated -optdep-* flags; fixes trac #2773 Ian Lynagh **20081128193633] [Use relative URLs in the GHC API haddock docs; fixes #2755 Ian Lynagh **20081128184511] [Teach runghc about --help; fixes trac #2757 Ian Lynagh **20081128191706] [Use a per-session data structure for callbacks. Make 'WarnErrLogger' Thomas Schilling **20081128103628 part of it. Part of the GHC API essentially represents a compilation framework. The difference of a *framework* as opposed to a *library* is that the overall structure of the functionality is pre-defined but certain details can be customised via callbacks. (Also known as the Hollywood Principle: "Don't call us, we'll call you.") This patch introduces a per-session data structure that contains all the callbacks instead of adding lots of small function arguments whenever we want to give the user more control over certain parts of the API. This should also help with future changes: Adding a new callback doesn't break old code since code that doesn't know about the new callback will use the (hopefully sane) default implementation. Overall, however, we should try and keep the number of callbacks small and well-defined (and provide useful defaults) and use simple library routines for the rest. ] [Improve error message for #2739 (but no fix). Thomas Schilling **20081127135725 This patch changes 'loadModule' to define a fake linkable. The previous implementation of providing no linkable at all violated a pre-condition in the ByteCode linker. This doesn't fix #2739, but it improves the error message a bit. ] [Remove the packing I added recently to the Capability structure Simon Marlow **20081128105046 The problem is that the packing caused some unaligned loads, which lead to bus errors on Sparc (and reduced performance elsewhere, presumably). ] [don't emit CmmComments for now Simon Marlow **20081127090145 - if the string contains */, we need to fix it (demonstrated by building Cabal with -fvia-C) - the strings can get quite large, so we probably only want to inject comments when some debugging option is on. ] [Collect instead of print warnings in 'warnUnnecessarySourceImports'. Thomas Schilling **20081127102534] [Force recompilation of BCOs when they were compiled in HscNothing mode. Thomas Schilling **20081126183402 Previously, loading a set of modules in HscNothing mode and then switching to HscInterpreted could lead to crashes since modules compiled with HscNothing were thought to be valid bytecode objects. This patch forces recompilation in these cases, hence switching between HscNothing and HscInterpreted should be safe now. ] [Documentation only: Add module description for HscMain. Thomas Schilling **20081126134344] [Include GHCi files in ctags/etags. Thomas Schilling **20081126122801] [drop some debugging traces and use only one flag for new codegen dias@eecs.harvard.edu**20081126180808] [one more missing patch from new codegen path dias@eecs.harvard.edu**20081126165742] [Fix Trac #2817 (TH syntax -> HsSyn conversion) simonpj@microsoft.com**20081126154022] [Fix Trac #2756: CorePrep strictness bug simonpj@microsoft.com**20081126143448] [Format output for :t more nicely simonpj@microsoft.com**20081126134814] [Fix Trac #2766: printing operator type variables simonpj@microsoft.com**20081126132202] [Fix build following codegen patch simonpj@microsoft.com**20081126125526] [Removed warnings, made Haddock happy, added examples in documentation dias@eecs.harvard.edu**20081017170707 The interesting examples talk about our story with heap checks in case alternatives and our story with the case scrutinee as a Boolean. ] [Fixed linear regalloc bug, dropped some tracing code dias@eecs.harvard.edu**20081016104218 o The linear-scan register allocator sometimes allocated a block before allocating one of its predecessors, which could lead to inconsistent allocations. Now, we allocate a block only if a predecessor has set the "incoming" assignments for the block (or if it's the procedure's entry block). o Also commented out some tracing code on the new codegen path. ] [Keep update frames live even in functions that never return dias@eecs.harvard.edu**20081014165437 An unusual case, but without it: (a) we had an assertion failure (b) we can overwrite the caller's infotable, which might cause the garbage collector to collect live data. Better to keep the update frame live at all call sites, not just at returns. ] [Removed space and time inefficiency in procpoint splitting dias@eecs.harvard.edu**20081014160354 I was adding extra jumps to every procpoint, even when the split-off graph referred to only some of the procpoints. No effect on correctness, but a big effect on space/time efficiency when there are lots of procpoints... ] [Clarify the SRT building process dias@eecs.harvard.edu**20081014140202 Before: building a closure that would build an SRT given the top-level SRT. It was somewhat difficult to understand the control flow, and it may have had held onto some data structures long after they should be dead. Now, I just bundle the info we need about CAFs along with the procedure and directly call a new top-level function to build the SRTs later. ] [Don't adjust hp up when the case scrutinee won't allocate dias@eecs.harvard.edu**20081014112618 If the case scrutinee can't allocate, we don't need to do a heap check in the case alternatives. (A previous patch got that right.) In that case, we had better not adjust the heap pointer to recover unused stack space before evaluating the scrutinee -- because we aren't going to reallocate for the case alternative. ] [Floating infotables were reversed in C back end dias@eecs.harvard.edu**20081013152718] [forgot a few files dias@eecs.harvard.edu**20081013134251] [Big collection of patches for the new codegen branch. dias@eecs.harvard.edu**20081013132556 o Fixed bug that emitted the copy-in code for closure entry in the wrong place -- at the initialization of the closure. o Refactored some of the closure entry code. o Added code to check that no LocalRegs are live-in to a procedure -- trip up some buggy programs earlier o Fixed environment bindings for thunks -- we weren't (re)binding the free variables in a thunk o Fixed a bug in proc-point splitting that dropped some updates to the entry block in a procedure. o Fixed improper calls to code that generates CmmLit's for strings o New invariant on cg_loc in CgIdInfo: the expression is always tagged o Code to load free vars on entry to a thunk was (wrongly) placed before the heap check. o Some of the StgCmm code was redundantly passing around Id's along with CgIdInfo's; no more. o Initialize the LocalReg's that point to a closure before allocating and initializing the closure itself -- otherwise, we have problems with recursive closure bindings o BlockEnv and BlockSet types are now abstract. o Update frames: - push arguments in Old call area - keep track of the return sp in the FCode monad - keep the return sp in every call, tail call, and return (because it might be different at different call sites, e.g. tail calls to the gc after a heap check are performed before pushing the update frame) - set the sp appropriately on returns and tail calls o Reduce call, tail call, and return to a single LastCall node o Added slow entry code, using different calling conventions on entry and tail call o More fixes to the calling convention code. The tricky stuff is all about the closure environment: it must be passed in R1, but in non-closures, there is no such argument, so we can't treat all arguments the same way: the closure environment is special. Maybe the right step forward would be to define a different calling convention for closure arguments. o Let-no-escapes need to be emitted out-of-line -- otherwise, we drop code. o Respect RTS requirement of word alignment for pointers My stack allocation can pack sub-word values into a single word on the stack, but it wasn't requiring word-alignment for pointers. It does now, by word-aligning both pointer registers and call areas. o CmmLint was over-aggresively ruling out non-word-aligned memory references, which may be kosher now that we can spill small values into a single word. o Wrong label order on a conditional branch when compiling switches. o void args weren't dropped in many cases. To help prevent this kind of mistake, I defined a NonVoid wrapper, which I'm applying only to Id's for now, although there are probably other good candidates. o A little code refactoring: separate modules for procpoint analysis splitting, stack layout, and building infotables. o Stack limit check: insert along with the heap limit check, using a symbolic constant (a special CmmLit), then replace it when the stack layout is known. o Removed last node: MidAddToContext o Adding block id as a literal: means that the lowering of the calling conventions no longer has to produce labels early, which was inhibiting common-block elimination. Will also make it easier for the non-procpoint-splitting path. o Info tables: don't try to describe the update frame! o Over aggressive use of NonVoid!!!! Don't drop the non-void args before setting the type of the closure!!! o Sanity checking: Added a pass to stub dead dead slots on the stack (only ~10 lines with the dataflow framework) o More sanity checking: Check that incoming pointer arguments are non-stubbed. Note: these checks are still subject to dead-code removal, but they should still be quite helpful. o Better sanity checking: why stop at function arguments? Instead, in mkAssign, check that _any_ assignment to a pointer type is non-null -- the sooner the crash, the easier it is to debug. Still need to add the debugging flag to turn these checks on explicitly. o Fixed yet another calling convention bug. This time, the calls to the GC were wrong. I've added a new convention for GC calls and invoked it where appropriate. We should really straighten out the calling convention stuff: some of the code (and documentation) is spread across the compiler, and there's some magical use of the node register that should really be handled (not avoided) by calling conventions. o Switch bug: the arms in mkCmmLitSwitch weren't returning to a single join point. o Environment shadowing problem in Stg->Cmm: When a closure f is bound at the top-level, we should not bind f to the node register on entry to the closure. Why? Because if the body of f contains a let-bound closure g that refers to f, we want to make sure that it refers to the static closure for f. Normally, this would all be fine, because when we compile a closure, we rebind free variables in the environment. But f doesn't look like a free variable because it's a static value. So, the binding for f remains in the environment when we compile g, inconveniently referring to the wrong thing. Now, I bind the variable in the local environment only if the closure is not bound at the top level. It's still okay to make assumptions about the node holding the closure environment; we just won't find the binding in the environment, so code that names the closure will now directly get the label of the static closure, not the node register holding a pointer to the static closure. o Don't generate bogus Cmm code containing SRTs during the STG -> Cmm pass! The tables made reference to some labels that don't exist when we compute and generate the tables in the back end. o Safe foreign calls need some special treatment (at least until we have the integrated codegen). In particular: o they need info tables o they are not procpoints -- the successor had better be in the same procedure o we cannot (yet) implement the calling conventions early, which means we have to carry the calling-conv info all the way to the end o We weren't following the old convention when registering a module. Now, we use update frames to push any new modules that have to be registered and enter the youngest one on the stack. We also use the update frame machinery to specify that the return should pop the return address off the stack. o At each safe foreign call, an infotable must be at the bottom of the stack, and the TSO->sp must point to it. o More problems with void args in a direct call to a function: We were checking the args (minus voids) to check whether the call was saturated, which caused problems when the function really wasn't saturated because it took an extra void argument. o Forgot to distinguish integer != from floating != during Stg->Cmm o Updating slotEnv and areaMap to include safe foreign calls The dataflow analyses that produce the slotEnv and areaMap give results for each basic block, but we also need the results for a safe foreign call, which is a middle node. After running the dataflow analysis, we have another pass that updates the results to includ any safe foreign calls. o Added a static flag for the debugging technique that inserts instructions to stub dead slots on the stack and crashes when a stubbed value is loaded into a pointer-typed LocalReg. o C back end expects to see return continuations before their call sites. Sorted the flowgraphs appropriately after splitting. o PrimOp calling conventions are special -- unlimited registers, no stack Yet another calling convention... o More void value problems: if the RHS of a case arm is a void-typed variable, don't try to return it. o When calling some primOp, they may allocate memory; if so, we need to do a heap check when we return from the call. ] [Merging in the new codegen branch dias@eecs.harvard.edu**20080814124027 This merge does not turn on the new codegen (which only compiles a select few programs at this point), but it does introduce some changes to the old code generator. The high bits: 1. The Rep Swamp patch is finally here. The highlight is that the representation of types at the machine level has changed. Consequently, this patch contains updates across several back ends. 2. The new Stg -> Cmm path is here, although it appears to have a fair number of bugs lurking. 3. Many improvements along the CmmCPSZ path, including: o stack layout o some code for infotables, half of which is right and half wrong o proc-point splitting ] [Major clean-up of HscMain. Thomas Schilling **20081125153201 This patch entails a major restructuring of HscMain and a small bugfix to MkIface (which required the restructuring in HscMain). In MkIface: - mkIface* no longer outputs orphan warnings directly and also no longer quits GHC when -Werror is set. Instead, errors are reported using the common IO (Messages, Maybe result) scheme. In HscMain: - Get rid of the 'Comp' monad. This monad was mostly GhcMonad + two reader arguments, a ModSummary for the currently compiled module and a possible old interface. The latter actually lead to a small space-leak since only its hash was needed (to check whether the newly-generated interface file was the same as the original one). Functions originally of type 'Comp' now only take the arguments that they actually need. This leads to slighly longer argument lists in some places, however, it is now much easier to see what is actually going on. - Get rid of 'myParseModule'. Rename 'parseFile' to 'hscParse'. - Join 'deSugarModule' and 'hscDesugar' (keeping the latter). - Rename 'typecheck{Rename}Module{'}' to 'hscTypecheck{Rename}'. One variant keeps the renamed syntax, the other doesn't. - Parameterise 'HscStatus', so that 'InteractiveStatus' is just a different parameterisation of 'HscStatus'. - 'hscCompile{OneShot,Batch,Nothing,Interactive}' are now implemented using a (local) typeclass called 'HsCompiler'. The idea is to make the common structure more obvious. Using this typeclass we now have two functions 'genericHscCompile' (original 'hscCompiler') and 'genericHscRecompile' (original 'genComp') describing the default pipeline. The methods of the typeclass describe a sort of "hook" interface (in OO-terms this would be called the "template method" pattern). One problem with this approach is that we parameterise over the /result/ type which, in fact, is not actually different for "nothing" and "batch" mode. To avoid functional dependencies or associated types, we use type tags to make them artificially different and parameterise the type class over the result type. A perhaps better approach might be to use records instead. - Drop some redundant 'HscEnv' arguments. These were likely different from what 'getSession' would return because during compilation we temporarily set the module's DynFlags as well as a few other fields. We now use the 'withTempSession' combinator to temporarily change the 'HscEnv' and automatically restore the original session after the enclosed action has returned (even in case of exceptions). - Rename 'hscCompile' to 'hscGenHardCode' (since that is what it does). Calls in 'GHC' and 'DriverPipeline' accordingly needed small adaptions. ] [Fix Trac #2799: TcType.isOverloadedTy simonpj@microsoft.com**20081125110540 A missing case (for equality predicates) in isOverloadedTy make bindInstsOfLocalFuns/Pats do the wrong thing. Core Lint nailed it. Merge to 6.10 branch. ] [Fix #2740: we were missing the free variables on some expressions Simon Marlow **20081125103113 Particularly boolean expresions: the conditional of an 'if', and guards, were missing their free variables. ] [Fix symbol macro names in Linker.c Thorkil Naur **20081121160149] [Add a --machine-readable RTS flag Ian Lynagh **20081123152127 Currently it only affects the -t flag output ] [Return errors instead of dying in myParseModule. Thomas Schilling **20081122154151] [Comments/Haddockification only. Thomas Schilling **20081122143018] [Report source span instead of just source location for unused names. Thomas Schilling **20081122142641] [Change 'handleFlagWarnings' to throw exceptions instead of dying. Thomas Schilling **20081122130658 It now uses the standard warning log and error reporting mechanism. ] [Document exported functions in main/HeaderInfo. Thomas Schilling **20081121145307] [Remove warning supression klugde in main/HeaderInfo Thomas Schilling **20081121144155] [Use mutator threads to do GC, instead of having a separate pool of GC threads Simon Marlow **20081121151233 Previously, the GC had its own pool of threads to use as workers when doing parallel GC. There was a "leader", which was the mutator thread that initiated the GC, and the other threads were taken from the pool. This was simple and worked fine for sequential programs, where we did most of the benchmarking for the parallel GC, but falls down for parallel programs. When we have N mutator threads and N cores, at GC time we would have to stop N-1 mutator threads and start up N-1 GC threads, and hope that the OS schedules them all onto separate cores. It practice it doesn't, as you might expect. Now we use the mutator threads to do GC. This works quite nicely, particularly for parallel programs, where each mutator thread scans its own spark pool, which is probably in its cache anyway. There are some flag changes: -g is removed (-g1 is still accepted for backwards compat). There's no way to have a different number of GC threads than mutator threads now. -q1 Use one OS thread for GC (turns off parallel GC) -qg Use parallel GC for generations >= (default: 1) Using parallel GC only for generations >=1 works well for sequential programs. Compiling an ordinary sequential program with -threaded and running it with -N2 or more should help if you do a lot of GC. I've found that adding -qg0 (do parallel GC for generation 0 too) speeds up some parallel programs, but slows down some sequential programs. Being conservative, I left the threshold at 1. ToDo: document the new options. ] [we shouldn't update topBound in discardSparks() Simon Marlow **20081121113539] [Throw SourceErrors instead of ProgramErrors in main/HeaderInfo. Thomas Schilling **20081121141339 Parse errors during dependency analysis or options parsing really shouldn't kill GHC; this is particularly annoying for GHC API clients. ] [fix the build when !USE_MMAP Simon Marlow **20081121085418] [round the size up to a page in mmapForLinker() instead of in the caller Simon Marlow **20081120154014] [Fix a race in the deadlock-detection code Simon Marlow **20081120112438 After a deadlock it was possible for the timer signal to remain off, which meant that the next deadlock would not be detected, and the system would hang. Spotted by conc047(threaded2). ] [error message wibble Simon Marlow **20081120084901] [Fix flag name -XDisambiguateRecordFields simonpj@microsoft.com**20081120123205] [Fix regTableToCapability() if gcc introduces padding Simon Marlow **20081119162910 Also avoid padding if possible using __attribute__((packed)) Fixes the Windows build ] [Fix 32-bit breakage Simon Marlow **20081119145429] [Small refactoring, and add comments Simon Marlow **20081119143702 I discovered a new invariant while experimenting (blackholing is not optional when using parallel GC), so documented it. ] [Fix some unsigned comparisions that should be signed Simon Marlow **20081119143205 Fixes crashes when using reclaimSpark() (not used currently, but may be in the future). ] [Remove incorrect assertions in steal() Simon Marlow **20081119143043] [don't run sparks if there are other threads on this Capability Simon Marlow **20081114121022] [Fix typo (HAVE_LIBGMP => HAVE_LIB_GMP); omit local gmp includes if HAVE_LIB_GMP Simon Marlow **20081119131056 If we're using the system's installed GMP, we don't want to be picking up the local gmp.h header file. Fixes 2469(ghci) for me, because it turns out the system's GMP is more up-to-date than GHC's version and has a fix for more recent versions of gcc. We also need to pull in a more recent GMP, but that's a separte issue. ] [Fix some more shutdown races Simon Marlow **20081119124848 There were races between workerTaskStop() and freeTaskManager(): we need to be sure that all Tasks have exited properly before we start tearing things down. This isn't completely straighforward, see comments for details. ] [Add help messages about --with-editline-(includes,libraries) to the ghc configure script. Judah Jacobson **20081114183334] [Add optional eager black-holing, with new flag -feager-blackholing Simon Marlow **20081118142442 Eager blackholing can improve parallel performance by reducing the chances that two threads perform the same computation. However, it has a cost: one extra memory write per thunk entry. To get the best results, any code which may be executed in parallel should be compiled with eager blackholing turned on. But since there's a cost for sequential code, we make it optional and turn it on for the parallel package only. It might be a good idea to compile applications (or modules) with parallel code in with -feager-blackholing. ToDo: document -feager-blackholing. ] [Fix #2783: detect black-hole loops properly Simon Marlow **20081117144515 At some point we regressed on detecting simple black-hole loops. This happened due to the introduction of duplicate-work detection for parallelism: a black-hole loop looks very much like duplicate work, except it's duplicate work being performed by the very same thread. So we have to detect and handle this case. ] [Fix warning on Windows (use deleteThread() not deleteThread_()) Simon Marlow **20081117143047] [fix compile breakage on Windows Simon Marlow **20081117142831] [Attempt to fix #2512 and #2063; add +RTS -xm
-RTS option Simon Marlow **20081117120556 On x86_64, the RTS needs to allocate memory in the low 2Gb of the address space. On Linux we can do this with MAP_32BIT, but sometimes this doesn't work (#2512) and other OSs don't support it at all (#2063). So to work around this: - Try MAP_32BIT first, if available. - Otherwise, try allocating memory from a fixed address (by default 1Gb) - We now provide an option to configure the address to allocate from. This allows a workaround on machines where the default breaks, and also provides a way for people to test workarounds that we can incorporate in future releases. ] [Another shutdown fix Simon Marlow **20081117094350 If we encounter a runnable thread during shutdown, just kill it. All the threads are supposed to be dead at this stage, but this catches threads that might have just returned from a foreign call, or were finalizers created by the GC. Fixes memo002(threaded1) ] [Correct an example in the users guide Ian Lynagh **20081116174938] [Fix gen_contents_index when not run inplace; trac #2764 Ian Lynagh **20081116174122 Based on a patch from juhpetersen. ] [close the temporary Handle before removing the file Simon Marlow **20081114130958] [refactor: move unlockClosure() into SMPClosureOps() where it should be Simon Marlow **20081114095817] [Omit definitions of cas() and xchg() in .hc code Simon Marlow **20081114095738 They cause compilation errors (correctly) with newer gccs Shows up compiling the RTS via C, which happens on Windows ] [Don't put stdin into non-blocking mode (#2778, #2777) Simon Marlow **20081114130546 This used to be necessary when our I/O library needed all FDs in O_NONBLOCK mode, and readline used to put stdin back into blocking mode. Nowadays the I/O library can cope with FDs in blocking mode, and #2778/#2777 show why this is important. ] [Rmoeve --enable-dotnet Simon Marlow **20081114124929] [#2751: disourage --enable-shared in ./configure --help Simon Marlow **20081114124748] [add a warning that --enable-shared is experimental Simon Marlow **20081114104034] [lookupSymbol: revert to looking up both with and without the @N suffix Simon Marlow **20081113122927] [#2768: fix compatibility problem with newer version of mingw Simon Marlow **20081113114626] [notice ^C exceptions when waiting for I/O Simon Marlow **20081113114342] [Fix a bug in the recompilation checking logic. Thomas Schilling **20081113162653 Previously, using target HscNothing resulted in unnessesary recompilation because 'upsweep_mod' treated HscInterface specially. This patch changes relaxes this. When running GHC with debug level 5, 'upsweep_mod' will now also be more verbose about what it is doing. There is (at least) one possible remaining problem, though: When using target 'HscNothing' we generate a fake linkable to signal that we have processed a module. When switching to 'HscInterpreted' this may cause objects to not be recompiled. Switching from HscNothing to HscInterpreted is therefore only safe if we unload everything first. ] [Fix another subtle shutdown deadlock Simon Marlow **20081113160005 The problem occurred when a thread tries to GC during shutdown. In order to GC it has to acquire all the Capabilities in the system, but during shutdown, some of the Capabilities have already been closed and can never be acquired. ] [Fix an extremely subtle deadlock bug on x86_64 Simon Marlow **20081113155730 The recent_activity flag was an unsigned int, but we sometimes do a 64-bit xchg() on it, which overwrites the next word in memory. This happened to contain the sched_state flag, which is used to control the orderly shutdown of the system. If the xchg() happened during shutdown, the scheduler would get confused and deadlock. Don't you just love C? ] [move an assertion Simon Marlow **20081113154542] [Always zap the trailing @N from symbols when looking up in a DLL Simon Marlow **20081112111518 Fixes win32002(ghci) Previously we only did this for references from object files, but we should do it for all symbols, including those that GHCi looks up due to FFI calls from bytecode. ] [Only allocate a mark stack if we're actually doing marking Simon Marlow **20081112112144 saves a bit of memory in major GCs ] [Fix parse error with older gccs (#2752) Simon Marlow **20081111135344] [Fix to i386_insert_ffrees (#2724, #1944) Simon Marlow **20081111125619 The i386 native code generator has to arrange that the FPU stack is clear on exit from any function that uses the FPU. Unfortunately it was getting this wrong (and has been ever since this code was written, I think): it was looking for basic blocks that used the FPU and adding the code to clear the FPU stack on any non-local exit from the block. In fact it should be doing this on a whole-function basis, rather than individual basic blocks. ] [Fix bootstrap with 6.10.1 on Windows Simon Marlow **20081110134318 ghc-pkg doesn't understand the old syntax any more, so 'ghc-pkg -l' fails ] [Perform case-insensitive matching of path components in getBaseDir on Windows (Fixes bug 2743) Neil Mitchell **20081105134315] [Documentation only. Clarify that 'load*' may indeed throw SourceErrors. Thomas Schilling **20081110175614 I don't think errors during dependency analysis should be passed to the logger. ] [Fix documentation (to say the opposite). Thomas Schilling **20081110153819] [Fix line numbers in TAGS files. Thomas Schilling **20081110153621] [Documentation only. Thomas Schilling **20081110153456] [Add 'packageDbModules' function to GHC API. Thomas Schilling **20081110143510 This function returns a list of all modules available through the package DB. MERGE TO 6.10 ] [We now require GHC 6.6, so we always have Applicative Ian Lynagh **20081108144723] [Remove a CPP test that's always true (__GLASGOW_HASKELL__ >= 605) Ian Lynagh **20081108144544] [Remove some dead code now that __GLASGOW_HASKELL__ >= 606 Ian Lynagh **20081108144459] [Remove some flag duplication from a Makefile Ian Lynagh **20081108144412] [ghc_ge_605 is now always YES Ian Lynagh **20081108144328] [Remove the GHC 6.4 unicode compat stuff; we can now just use Data.Char Ian Lynagh **20081108143423] [Fix libffi bindist Clemens Fruhwirth **20081108094725] [Replace couple of fromJust with expectJust Clemens Fruhwirth **20081107160735] [Bugfix for patch "Do not filter the rts from linked libraries..." (#2745) Simon Marlow **20081107092925 The sense of the #ifdef was wrong ] [fix via-C compilation: import ghczmprim_GHCziBool_False_closure Simon Marlow **20081107090432] [disable instance MonadPlus CoreM for GHC <= 6.6 Simon Marlow **20081107085250] [re-instate counting of sparks converted Simon Marlow **20081106160810 lost in patch "Run sparks in batches" ] [fix ASSERT_SPARK_POOL_INVARIANTS(): top>bottom is valid Simon Marlow **20081106155826] [pruneSparkQueue(): fix bug when top>bottom Simon Marlow **20081106155648] [don't yield if the system is shutting down Simon Marlow **20081106155356] [leave out ATTRIBUTE_ALIGNED on Windows, it gives a warning Simon Marlow **20081106132105] [Cope with ThreadRelocated when traversing the blocked_queue Simon Marlow **20081106114045 Fixes "invalid what_next field" in ioref001 on Windows, and perhaps others ] [Remove dead code. Thomas Schilling **20081031162036] [Run sparks in batches, instead of creating a new thread for each one Simon Marlow **20081106113639 Signficantly reduces the overhead for par, which means that we can make use of paralellism at a much finer granularity. ] [allocateInGen(): increase alloc_blocks (#2747) Simon Marlow **20081106113714] [disable MonadPlus instance that doesn't compile with 6.6 Simon Marlow **20081106100411] [don't yield the Capability if blackholes_need_checking Simon Marlow **20081105154928] [deadlock fix: reset the flag *after* checking the blackhole queue Simon Marlow **20081105150542] [retreat the top/bottom fields of the spark pool in pruneSparkPool() Simon Marlow **20081105150359] [fix the :help docs for :set stop (#2737) Simon Marlow **20081104092929] [bugfix: don't ingore the return value from rts_evalIO() Simon Marlow **20081104142147] [Document the new SPARKS statistic, and xref from the parallelism section Simon Marlow **20081024120236] [Move the freeing of Capabilities later in the shutdown sequence Simon Marlow **20081024104301 Fixes a bug whereby the Capability has been freed but other Capabilities are still trying to steal sparks from its pool. ] [Pad Capabilities and Tasks to 64 bytes Simon Marlow **20081023080749 This is just good practice to avoid placing two structures heavily accessed by different CPUs on the same cache line ] [Fix desugaring of record update (fixes Trac #2735) simonpj@microsoft.com**20081103110819] [Refuse to register packages with unversioned dependencies; trac #1837 Ian Lynagh **20081031181746] [We now require GHC 6.6 to build the HEAD (and thus 6.12) Ian Lynagh **20081031171506] [:set prompt now understand Haskell String syntax; trace #2652 Ian Lynagh **20081031145227] [Comments only simonpj@microsoft.com**20081031140236] [Quickfix for warning. Thomas Schilling **20081031113125] [Export typeclasses for accessing compiler results. Thomas Schilling **20081028182310 MERGE TO 6.10. ] [Minor refactoring. Thomas Schilling **20081028182202] [Include record fields in tags. Thomas Schilling **20081028180538] [Fix imports simonpj@microsoft.com**20081031092306] [Improve error reporting for non-rigid GADT matches simonpj@microsoft.com**20081030143947 Following suggestions from users, this patch improves the error message when a GADT match needs a rigid type: tcfail172.hs:19:10: GADT pattern match in non-rigid context for `Nil' - Solution: add a type signature + Probable solution: add a type signature for `is_normal' In the pattern: Nil In the definition of `is_normal': is_normal Nil = True Now GHC tries to tell you what to give a type signature *for*. Thanks to Daniel Gorin and others for the suggestions. ] [Add (a) CoreM monad, (b) new Annotations feature simonpj@microsoft.com**20081030125108 This patch, written by Max Bolingbroke, does two things 1. It adds a new CoreM monad (defined in simplCore/CoreMonad), which is used as the top-level monad for all the Core-to-Core transformations (starting at SimplCore). It supports * I/O (for debug printing) * Unique supply * Statistics gathering * Access to the HscEnv, RuleBase, Annotations, Module The patch therefore refactors the top "skin" of every Core-to-Core pass, but does not change their functionality. 2. It adds a completely new facility to GHC: Core "annotations". The idea is that you can say {#- ANN foo (Just "Hello") #-} which adds the annotation (Just "Hello") to the top level function foo. These annotations can be looked up in any Core-to-Core pass, and are persisted into interface files. (Hence a Core-to-Core pass can also query the annotations of imported things.) Furthermore, a Core-to-Core pass can add new annotations (eg strictness info) of its own, which can be queried by importing modules. The design of the annotation system is somewhat in flux. It's designed to work with the (upcoming) dynamic plug-ins mechanism, but is meanwhile independently useful. Do not merge to 6.10! ] [Fix Trac #2674: in TH reject empty case expressions and function definitions simonpj@microsoft.com**20081030094528] [Change naming conventions for compiler-generated dictionaries and type functions simonpj@microsoft.com**20081029140858 Up to now, the data constructor dictionary for class C as been called ":DC". But there is no reason for the colon to be at the front; indeed it confuses the (simple-minded) pretty-printer for types. So this patch changes it to be "D:C". This makes Core a lot easier to read. Having a colon in the middle ensures that it can't clash with a user-written data type. Similarly I changed T:C Data type corresponding a class dictionary (was :TC) D:C Data constructor for class dictionary (was :DC) NTCo:T Coercion mapping from a newtype T to its representation type (was :CoT) TFCo:R Coercion mapping from a data family to its respresentation type R (was :CoFR) Rn:T The n'th respresentation data type for a data type T (was :RnT) Do not merge to 6.10. HEADS-UP: you'll need to recompile libraries from scratch. ROMAN: you could do the same for OccName.mkVectTyConOcc etc, if you wanted. ] [Fix tcrun031: yet more tidying up in TcDeriv simonpj@microsoft.com**20081029130155] [Add Outputable instance for CoercionI simonpj@microsoft.com**20081029130114] [Fix Trac #2720: inlining and casts simonpj@microsoft.com**20081028140828 The issue here is what happens when we have (f |> co) x where f is itself marked INLINE. We want callSiteInline to "see" the fact that the function is applied, and hence have some incentive to inline. I've done this by extending CoreUnfold.CallCtxt with ValAppCtxt. I think that should catch this case without messing up any of the others. ] [Clarify documentatoin simonpj@microsoft.com**20081028133009] [Update library version numbers in the release notes Ian Lynagh **20081023144018] [various updates to the release notes Simon Marlow **20081007151647] [Add library release notes Ian Lynagh **20080920155722] [Add release notes for the compiler Ian Lynagh **20080920114857] [Doc fix Ian Lynagh **20081028150534] [Rename some variables in docs Ian Lynagh **20081028150447] [Fix typos Ian Lynagh **20081028144655] [Mostly-fix Trac #2595: updates for existentials simonpj@microsoft.com**20081028115427 Ganesh wanted to update records that involve existentials. That seems reasonable to me, and this patch covers existentials, GADTs, and data type families. The restriction is that The types of the updated fields may mention only the universally-quantified type variables of the data constructor This doesn't allow everything in #2595 (it allows 'g' but not 'f' in the ticket), but it gets a lot closer. Lots of the new lines are comments! ] [Fix Trac #2723: keep track of record field names in the renamer simonpj@microsoft.com**20081028110445 The idea here is that with -XNamedFieldPuns and -XRecordWildCards we don't want to report shadowing errors for let fld = in C { .. } But to suppress such shadowing errors, the renamer needs to know that 'fld' *is* a record selector. Hence the new NameSet in TcRnFypes.RecFieldEnv ] [Remove dead code simonpj@microsoft.com**20081028074639] [Fix Trac #2713: refactor and tidy up renaming of fixity decls simonpj@microsoft.com**20081027222738 In fixing #2713, this patch also eliminates two almost-unused functions from RnEnv (lookupBndr and lookupBndr_maybe). The net lines of code is prety much unchanged, but more of them are comments! ] [Fix Trac #2701: make deriving check better for unlifted args simonpj@microsoft.com**20081025171211 Getting the automatic deriving mechanism to work really smoothly is surprisingly hard. I keep finding myself in TcDeriv! Anyway, this is a nice clean fix to Trac #2701. ] [Use pdflatex rather than latex for building Ian Lynagh **20081024112400 The Windows builder is having problems running ps2pdf, so this works aroudn the problem. ] [Remove an unmatched } in core.tex Ian Lynagh **20081024111750] [Add a usepackage{url} Ian Lynagh **20081024111458] [Update ANNOUNCE Ian Lynagh **20081022164423] [Fix a bug in the new scheduler Simon Marlow **20081023155017 If the current thread blocks, we should yield the Capability immediately, because the thread and hence possibly the current Task are now owned by someone else. This worked in the old scheduler, but we moved where the yield happens in the new scheduler which broke it. ] [add an assertion Simon Marlow **20081023154551] [fix a warning Simon Marlow **20081023082213] [traverse the spark pools only once during GC rather than twice Simon Marlow **20081022115233] [Refactoring and reorganisation of the scheduler Simon Marlow **20081022092744 Change the way we look for work in the scheduler. Previously, checking to see whether there was anything to do was a non-side-effecting operation, but this has changed now that we do work-stealing. This lead to a refactoring of the inner loop of the scheduler. Also, lots of cleanup in the new work-stealing code, but no functional changes. One new statistic is added to the +RTS -s output: SPARKS: 1430 (2 converted, 1427 pruned) lets you know something about the use of `par` in the program. ] [Work stealing for sparks berthold@mathematik.uni-marburg.de**20080915132846 Spark stealing support for PARALLEL_HASKELL and THREADED_RTS versions of the RTS. Spark pools are per capability, separately allocated and held in the Capability structure. The implementation uses Double-Ended Queues (deque) and cas-protected access. The write end of the queue (position bottom) can only be used with mutual exclusion, i.e. by exactly one caller at a time. Multiple readers can steal()/findSpark() from the read end (position top), and are synchronised without a lock, based on a cas of the top position. One reader wins, the others return NULL for a failure. Work stealing is called when Capabilities find no other work (inside yieldCapability), and tries all capabilities 0..n-1 twice, unless a theft succeeds. Inside schedulePushWork, all considered cap.s (those which were idle and could be grabbed) are woken up. Future versions should wake up capabilities immediately when putting a new spark in the local pool, from newSpark(). Patch has been re-recorded due to conflicting bugfixes in the sparks.c, also fixing a (strange) conflict in the scheduler. ] [include an elapsed time table Simon Marlow **20081023080648] [generate a valid summary for older GHC versions too Simon Marlow **20081023080629] [Fix Trac #2714 (a minor wibble) simonpj@microsoft.com**20081022145138 In boxy_match (which is a pure function used by preSubType) we can encounter TyVars not just TcTyVars; this patch takes account of that. ] [Reject programs with superclass equalities for now Manuel M T Chakravarty **20081021131721 - The current implementation of type families cannot properly deal with superclass equalities. Instead of making a half-hearted attempt at supporting them, which mostly ends in cryptic error message, rejecting right away with an appropriate message. MERGE TO 6.10 ] [Fix doc syntax Ian Lynagh **20081021183317] [Don't put the README file in the Windows installer; fixes trac #2698 Ian Lynagh **20081021162543 The README file talks about getting and building the sources, which doesn't make sense for the installer. ] [Comments and parens only simonpj@microsoft.com**20081021151401] [Do proper cloning in worker/wrapper splitting simonpj@microsoft.com**20081021143156 See Note [Freshen type variables] in WwLib. We need to clone type variables when building a worker/wrapper split, else we simply get bogus code, admittedly in rather obscure situations. I can't quite remember what program showed this up, unfortunately, but there definitely *was* one! (You get a Lint error.) ] [Don't float an expression wrapped in a cast simonpj@microsoft.com**20081021143019 There is no point in floating out an expression wrapped in a coercion; If we do we'll transform lvl = e |> co [_$_] to lvl' = e; lvl = lvl' |> co and then inline lvl. Better just to float out the payload (e). ] [Fix Trac #2668, and refactor TcDeriv simonpj@microsoft.com**20081021142922 TcDeriv deals with both standalone and ordinary 'deriving'; and with both data types and 'newtype deriving'. The result is really rather compilcated and ad hoc. Ryan discovered #2668; this patch fixes that bug, and makes the internal interfces #more uniform. Specifically, the business of knocking off type arguments from the instance type until it matches the kind of the class, is now done by derivTyData, not mkNewTypeEqn, because the latter is shared with standalone derriving, whree the trimmed type application is what the user wrote. ] [Spelling error in comment simonpj@microsoft.com**20081019233511] [White space only simonpj@microsoft.com**20081019233352] [Comments to explain strict overlap checking for type family instances simonpj@microsoft.com**20081019184208] [Allow type families to use GADT syntax (and be GADTs) simonpj@microsoft.com**20080923140535 We've always intended to allow you to use GADT syntax for data families: data instance T [a] where T1 :: a -> T [a] and indeed to allow data instances to *be* GADTs data intsance T [a] where T1 :: Int -> T [Int] T2 :: a -> b -> T [(a,b)] This patch fixes the renamer and type checker to allow this. ] [Improve crash message from applyTys and applyTypeToArgs simonpj@microsoft.com**20080923135419] [Wibble to ungrammatical error message simonpj@microsoft.com**20080920232256] [Comments only: replace ":=:" by "~" (notation for equality predicates) simonpj@microsoft.com**20080920232024] [FIX #2693 Manuel M T Chakravarty **20081021120107 - As long as the first reduceContext in tcSimplifyRestricted potentially performs improvement, we need to zonk again before the second reduceContext. It would be better to prevent the improvement in the first place, but given the current situation zonking is definitely the right thing to do. MERGE TO 6.10 ] [Restore the terminal attributes even if ghci does not exit normally. Judah Jacobson **20081020164109] [FIX #2688 Manuel M T Chakravarty **20081021044213 - Change in TcSimplify.reduceContext: We do *not* go around for new extra_eqs. Morally, we should, but we can't without risking non-termination (see #2688). By not going around, we miss some legal programs mixing FDs and TFs, but we never claimed to support such programs in the current implementation anyway. MERGE TO 6.10 ] [Move documentation within 80 column boundary. Thomas Schilling **20081014134016] [Improve haddock documentation for 'GHC.topSortModuleGraph'. Thomas Schilling **20081014133833] [Improve haddock documentation for HsExpr module. Thomas Schilling **20081014133740] [Improve Haddock-markup for HsDecls module. Thomas Schilling **20081014133556] [Run Haddock with compacting GC and show RTS statistics. Thomas Schilling **20081020111410] [FIX (partially) #2703: bug in stack overflow handling when inside block Simon Marlow **20081020121103 As a result of a previous ticket (#767) we disabled the generation of StackOverflow exceptions when inside a Control.Exception.block, on the grounds that StackOverflow is like an asynchronous exception. Instead we just increase the stack size. However, the stack size calculation was wrong, and ended up not increasing the size of the stack, with the result that the runtime just kept re-allocating the stack and filling up memory. ] [Re-export Located(..) and related functions Simon Marlow **20081020102422 So that clients don't need to import SrcLoc ] [whitespace fix Simon Marlow **20081020101241] [FIX #2691: Manually reset the terminal to its initial settings; works around a bug in libedit. Judah Jacobson **20081016024838] [Eliminate duplicate flags in the tab completion of ghci's :set command. Judah Jacobson **20081016015721] [Add dph haddock docs to the doc index Ian Lynagh **20081019133208] [Clean the bootstrapping extensible-exceptions package Ian Lynagh **20081017195942] [Fix trac #2687 Ian Lynagh **20081015163412 OtherPunctuation, ConnectorPunctuation and DashPunctuation are now treated as symbols, rather than merely graphic characters. ] [Fix trac #2680; avoid quadratic behaviour from (++) Ian Lynagh **20081015163235] [Fix the build when the bootstrapping compiler has a newer Cabal than us Ian Lynagh **20081015144023 We need to forcibly use the in-tree Cabal, or we get version mismatch errors ] [Fix the name of prologue.txt when making bindists Ian Lynagh **20081014114714] [Comments only simonpj@microsoft.com**20081015084501] [Fix Trac #2497; two separate typos in Lexer.x simonpj@microsoft.com**20081015084344 The patch to switch on lexing of 'forall' inside a RULES pragma wasn't working. This fixes it so that it does. ] [Update manual: tidy up instances, say more about type families in instance decls simonpj@microsoft.com**20081015080509] [Make tags work on Unices, too. Thomas Schilling **20081014211236] [Undefine __PIC__ before defining it to work around "multiple definitions of __PIC__" warnings Clemens Fruhwirth **20081014143206] [Add "dyn" to GhcRTSWays when compiling --enable-shared Clemens Fruhwirth **20081014125123] [Fill out the ghc package's cabal file Ian Lynagh **20081013235817] [Patching libffi so it can be built as DLL Clemens Fruhwirth **20081014103459 libffi-dllize-3.0.6.patch should be pushed upstream ] [Add 'etags' makefile target. Thomas Schilling **20081013170927 This only works with stage2 since `ghctags' is built against stage1 but not against the bootstrapping compiler. Also note that all of GHC must typecheck for this target to succeed. Perhaps we should not overwrite the old TAGS file by default then. ] [Use cabal information to get GHC's flags to `ghctags'. Thomas Schilling **20081013170658 By giving the dist-directory to ghctags we can get all the GHC API flags we need in order to load the required modules. The flag name could perhaps be improved, but apart from that it seems to work well. ] [Version bump for libffi to 3.0.6 Clemens Fruhwirth **20081014081300] [Encode shared/static configuration into stamps to do the right thing when rebuilding Clemens Fruhwirth **20081013221530] [Add a link to the GHC API docs from the library haddock index Ian Lynagh **20081013195943] [Link to the GHC API documentation from the main doc page Ian Lynagh **20081013200927] [Whitespace only in docs/index.html Ian Lynagh **20081013200625] [Tweak gen_contents_index Ian Lynagh **20081013192548 It now works again after it has been installed, as well as while it is in a source tree. After it's been installed it filters out the ghc package, as that currently swamps everything else in the index. ] [Build fixes for DLLized rts Clemens Fruhwirth **20081013201608] [Do not filter the rts from linked libraries in linkDynLib as Windows does not allow unresolved symbols Clemens Fruhwirth **20081013201426] [Add HsFFI.o to INSTALL_LIBS Clemens Fruhwirth **20081013200945] [Rename symbol macros to a consistant naming scheme Clemens Fruhwirth **20081013162433] [Fix #2685: two Bool arguments to tidyTypeEnv were the wrong way around Simon Marlow **20081013121339 So -XTemplateHaskell was behaving like -fomit-interface-file-pragmas, and vice versa. ] [Simplify the "is $bindir in $PATH" test Ian Lynagh **20081011191008] [Correct the "is $bindir in $PATH" test Ian Lynagh **20081011191030 We were testing neq instead of eq ] [Fix a typo which was causing ghci to quit on commands errors pepe **20081011114720] [Drop libm from the linker dependencies for libffi Clemens Fruhwirth **20081011074524] [Do not generate haddock documentation when running install-docs in libffi Clemens Fruhwirth **20081010192318] [When waking up thread blocked on TVars, wake oldest first (#2319) Josef Svenningsson **20081010150322 StgTVarWatchQueue contains the threads blocked on a TVar in order youngest first. The list has to be traversed backwards to unpark the threads oldest first. This improves the fairness when using STM in some situations. ] [add readTVarIO :: TVar a -> IO a Simon Marlow **20081010131545] [fix #2636: throw missing module errors as SourceErrors, not ErrMsg Simon Marlow **20081010131535] [atomicModifyIORef: use a local cas() instead of the global lock Simon Marlow **20081008154702 This should improve scaling when using atomicModifyIORef ] [Delay building libffi until package.conf is created and fix bindist Clemens Fruhwirth **20081010073106] [Install a versioned ghc-pkg script; fixes trac #2662 Ian Lynagh **20081009164946] [Fix bindist creation: Only the main RTS was being put in the bindists Ian Lynagh **20081009163451] [pushAtom: add missing case for MachNullAddr (#2589) Simon Marlow **20081009091118] [undo incorrect assertion, and fix comments Simon Marlow **20081009085118] [remove old GRAN/PARALLEL_HASKELL code Simon Marlow **20081009085051] [FIX #2639 Manuel M T Chakravarty **20081009132328 MERGE TO 6.10 ] [Cover PredTy case in Type.tyFamInsts Manuel M T Chakravarty **20081009061435 MERGE TO 6.10 ] [Drop ghcconfig.h/RtsConfig.h from libffi's package.conf.in Clemens Fruhwirth **20081009071342] [Don't use sed's -i flag as Solaris doesn't know it in libffi/Makefile Clemens Fruhwirth **20081008234455] [Don't use /dev/null trick to create empty object files in libffi/Makefile Clemens Fruhwirth **20081008232902] [Turn libffi into a Haskell package Clemens Fruhwirth **20081008170443] [Make 'getModSummary' deterministic. Thomas Schilling **20081008144032] [Add accessors to 'HsModule' and haddockify it. Thomas Schilling **20081007235656] [fix syntax errors in src-dist publish rules Simon Marlow **20081008103432] [add comments and an ASSERT_LOCK_HELD() Simon Marlow **20081008112627] [Fix #2663: we had a hard-wired capabilities[0] Simon Marlow **20081008112609 For some unknown reason in schedulePostRunThread() we were always passing capabilities[0] rather than the current Capability to throwToSingleThreaded(). This caused all kinds of weird failures and crashes in STM code when running on multiple processors. ] [Fix #1955 for heap profiles generated by +RTS -hT Simon Marlow **20081003150745] [add a section id for +RTS -hT Simon Marlow **20081007151007] [update documentation for PostfixOperators Simon Marlow **20081007150957] [fix markup Simon Marlow **20081007150943] [Fix bug in DPH docs Roman Leshchinskiy **20081008101618] [Add short DPH section to users guide Roman Leshchinskiy **20081008064754 MERGE TO 6.10 ] [Users Guide: added type family documentation Manuel M T Chakravarty **20081008061927 MERGE TO 6.10 ] [Track changes to package dph Roman Leshchinskiy **20081008032859 MERGE TO 6.10 ] [Build a profiled GHC API by default if p is in GhcLibWays Ian Lynagh **20081007152318] [Check whether mk/validate.mk defines anything after validating Ian Lynagh **20081007144855] [Remove #define _BSD_SOURCE from Stg.h Ian Lynagh **20081006101959 It's no longer needed, as base no longer #includes it ] [Make ghctags compile again. Thomas Schilling **20081007135705] [Revert AutoLinkPackages change for dynamic libraries. Cabal handles that now. Clemens Fruhwirth **20081007100417] [Change suffix for dyn. linked executables from _real to .dyn Clemens Fruhwirth **20081007100750] [Add accessors to 'Target' fields and haddockify. Thomas Schilling **20081006222940 MERGE TO 6.10 ] [Make 'gblock' and 'gunblock' part of 'ExceptionMonad'. This way the Thomas Schilling **20081006222831 default implementations of 'gbracket' and 'gfinally' just work. MERGE TO 6.10 ] [Add Word8 support to vectoriser Roman Leshchinskiy **20081007004416 MERGE TO 6.10 ] [Fix generating OS X installers: Set COMMAND_MODE=unix2003 Ian Lynagh **20081005222715 If we don't specify COMMAND_MODE=unix2003 then xcodebuild defaults to setting it to legacy, which means that ar builds archives without a table of contents. That makes the build fail later on. ] [We need to set datadir = $(libdir) in bindists Ian Lynagh **20081005143307 We already do in the normal Makefiles. This is because GHC needs package.conf and unlit to be in the same place (and things like ghc-pkg need to agree on where package.conf is, so we just set it globally). ] [prep-bin-dist-mingw complains if it finds a bad version of windres Ian Lynagh **20081004175351] [Fix a build problem with GHC 6.4.2 Ian Lynagh **20081003195700] [No AutoLinkPackages for dynamic library linking Clemens Fruhwirth **20081003185304] [use ghcError for error in command line Clemens Fruhwirth **20081001125648] [Fix warnings simonpj@microsoft.com**20081003171207] [Always use extensible exceptions in ghc-pkg, rather than using ifdefs Ian Lynagh **20081003161247] [Use a proper exception for IOEnvFailure, not just a UserError Ian Lynagh **20081003160129] [Use an extensible-exceptions package when bootstrapping Ian Lynagh **20081003140216 Ifdefs for whether we had extensible exceptions or not were spreading through GHC's source, and things would only have got worse for the next 2-3 years, so instead we now use an implementation of extensible exceptions built on top of the old exception type. ] [Expunge ThFake, cure Trac #2632 simonpj@microsoft.com**20081003140423 This patch fixes a dirty hack (the fake ThFake module), which in turn was causing Trac #2632. The new scheme is that the top-level binders in a TH [d| ... |] decl splice get Internal names. That breaks a previous invariant that things like TyCons always have External names, but these TyCons are never long-lived; they live only long enough to typecheck the TH quotation; the result is discarded. So it seems cool. Nevertheless -- Template Haskell folk: please test your code. The testsuite is OK but it's conceivable that I've broken something in TH. Let's see. ] [Make a debug check more refined simonpj@microsoft.com**20081003140144] [Add ASSERTs to all calls of nameModule simonpj@microsoft.com**20081003135334 nameModule fails on an InternalName. These ASSERTS tell you which call failed. ] [Let parseModule take a ModSummary like checkAndLoadModule did. Thomas Schilling **20081002230412 To get the ModSummary for a ModuleName getModSummary can be used. It's not called find* or lookup* because it assumes that the module is in the module graph and throws an exception if it cannot be found. Overall, I'm not quite sure about the usefulness of this function since the user has no control about which filetype to grab (hs or hs-boot). ] [Remove some out-of-date entries from .darcs-boring Ian Lynagh **20081002201519] [TFs: Allow repeated variables in left-hand sides of instances Manuel M T Chakravarty **20081002134539 MERGE TO 6.10 ] [Clean up some comments Manuel M T Chakravarty **20081002074642 MERGE TO 6.10 ] [Make the new binder-swap stuff in OccurAnal work right for GlobalIds simonpj@microsoft.com**20081002133002 See Note [Binder swap on GlobalId scrutinees]. I hadn't got this right before, so repeated cases on imported Ids weren't getting optimised. ] [Minor refactoring only simonpj@microsoft.com**20081002132929] [Comments only simonpj@microsoft.com**20081002132833] [Zap dead-ness info appropriately in SpecConstr simonpj@microsoft.com**20081002132657 SpecConstr can make pattern binders come alive, so we must remember to zap their dead-variable annotation. See extendCaseBndrs. (This was triggering a Core Lint failure in DPH.) ] [Suppress invalid Core Lint complaint about lack of constructors simonpj@microsoft.com**20081002132426] [add some more GC roots (fixes conc048, and possibly some others) Simon Marlow **20081001164427] [Document +RTS -hT Simon Marlow **20081001163222 We forgot to document this in GHC 6.8 ] [fix new-qualified-operators link Simon Marlow **20081001163105] [Proper error message for unsupported pattern signatures Manuel M T Chakravarty **20081001144339 - Pattern signatures must be identical to the type expected for the pattern; see Note [Pattern coercions] - We now signal an appropriate error if an equality coercion would be needed (instead of just generating Core that doesn't typecheck) MERGE TO 6.10 ] [Prevent excessive inlining with DPH Roman Leshchinskiy **20081002012055 This adds a new flag -finline-if-enough-args which disables inlining for partially applied functions. It is automatically set by -Odph. This is a temporary hack and should remain undocumented. MERGE TO 6.10 ] [On Windows, check that we have a good version of windres when configuring Ian Lynagh **20081001171133] [Call $(PERL) rather than perl when making the manpage Ian Lynagh **20080930155054] [don't install the installPackage program Ian Lynagh **20080930145714] [Fix #2637: conc032(threaded2) failure Simon Marlow **20081001135549 There was a race condition whereby a thread doing throwTo could be blocked on a thread that had finished, and the GC would detect this as a deadlock rather than raising the pending exception. We can't close the race, but we can make the right thing happen when the GC runs later. ] [Remove outdated link to OGI webpage Simon Marlow **20080930150912] [TFs: Fixed InstContextNorm (and simplification of IPs) Manuel M T Chakravarty **20081001131303 MERGE TO 6.10 ] [TcSimplify.reduceImplication: clean up Manuel M T Chakravarty **20081001091315 - This cleans up some of the mess in reduceImplication and documents the precondition on the form of wanted equalities properly. - I also made the back off test a bit smarter by allowing to back off in the presence of wanted equalities as long as none of them got solved in the attempt. (That should save generating some superfluous bindings.) MERGE TO 6.10 ] [Make sure to zonk the kind of coercion variables Manuel M T Chakravarty **20081001053243 MERGE TO 6.10 ] [Remover PROT_EXEC flag from mmap() Simon Marlow **20080930141842 Needed for #738 fix ] [Fix #2410: carefully generate unique names for CAF CCs Simon Marlow **20080930141812] [fix #2594: we were erroneously applying masks, as the reporter suggested Simon Marlow **20080930115611 My guess is that this is left over from when we represented Int8 and friends as zero-extended rather than sign-extended. It's amazing it hasn't been noticed earlier. ] [Unconditionalize definition of DYNAMIC_* so that libffi.so/.dll is removed even when BuildSharedLibs is reset to NO Clemens Fruhwirth **20080930085449] [Type families: need to instantiate flexible skolems before other flexibles Manuel M T Chakravarty **20080930053559 MERGE TO 6.10 ] [Fix warnings Manuel M T Chakravarty **20080929142227] [Type families: consider subst rules both way Manuel M T Chakravarty **20080929141040 - applySubstFam, applySubstVarVar & applySubstVarFam need to return their second argument -to be put into the todo list- if the rule would be applicable if the equalities would be supplied in the opposite order. MERGE TO 6.10 ] [removed Data.Generics.Basics, added Data.Data 'Jose Pedro Magalhaes '**20081002082808] [Clean up a bit and improve an error message pepe**20080926211429] [Don't capture error calls in tryUser pepe**20080926204836 A previous patch slightly changed the semantics of tryUser. This patch restores the original behaviour (as expected in :print) ] [tweaks to this section of the docs Simon Marlow **20080927141834] [Add -outputdir flag (#2295) Simon Marlow **20080927141822] [oops, forgot to add -XNewQualifiedOperators to the flags table Simon Marlow **20080923140449] [Fix making OS X installers from source tarballs Ian Lynagh **20080927150507 I'm not sure why it works in the HEAD, but when making an installer from the 6.10.1 beta configure hangs when doing the CHECK_HIST_ERRORS test (during rl_initialize, I believe). Giving make /dev/null as stdin fixes it. ] [Make the matching of the filename ghc.exe case insensitive, fixes bug #2603 Neil Mitchell **20080916160311] [Fix #2411: missing case for CATCH_STM_FRAME in raiseAsync() Simon Marlow **20080926232806] [Fix parsing of -ignore-package flag. Bertram Felgenhauer **20080925053820] [Add an example of how to use SCCs to the user guide Ian Lynagh **20080926203832] [Add some description of the +RTS -t/-s/-S output Ian Lynagh **20080926200203] [Remove a redundant options pragma Ian Lynagh **20080926152731] [Split ShowVersion etc off into a different type to DoInteractive etc Ian Lynagh **20080926140539 This fixes trac #1348 (ghci --help gave ghc's help), and also tidies things up a bit. Things would be even tidier if the usage.txt files were put into a .hs file, so that ShowUsage wouldn't need to be able to find the libdir. ] [Pass SRC_HC_OPTS to GHC when building GHC's Main.hs Ian Lynagh **20080926131609] [Improve runghc docs; fixes trac #2477 Ian Lynagh **20080926124425] [Type families: fixes in flattening & finalisation Manuel M T Chakravarty **20080925225324 * Finalisation didn't do the right thing for equalities x ~ y, where x was instantiated, but not zonked and y flexible (need to do y := x) * During flattening we weren't careful enough when turning wanteds intermediates into locals Both bugs showed up in a small example of SPJ: linear :: HasTrie (Basis v) => (Basis v, v) linear = basisValue class HasTrie a where type family Basis u :: * basisValue :: (Basis v,v) basisValue = error "urk" ] [Fix the behaviour of flags like --help and --version; fixes trac #2620 Ian Lynagh **20080925165618 They should override other mode flags, not conflict with them ] [Follow the integer package changes Ian Lynagh **20080925133855] [Type families: fix decomposition problem Manuel M T Chakravarty **20080925084139 * Fixes the problem reported in ] [Don't exit ghci if :info is called on an undefined identifier. Judah Jacobson **20080924212422] [Fix maintainer-clean Ian Lynagh **20080924230553] [Use -f when making the runhaskell symlink Ian Lynagh **20080924124255 Otherwise installation fails if runhaskell already exists. ] [Use -perm -100 rather than -perm /a+x when looking for executable files Ian Lynagh **20080924124137 /a+x doesn't work on some Solaris and OS X machines. Spotted by Christian Maeder. ] [Use $(FIND) rather than find, as the former may be gfind Ian Lynagh **20080924123323] [Look for gfind as well as find Ian Lynagh **20080924123046] [In configure, don't call FPTOOLS_HADDOCK Ian Lynagh **20080924122558 We now use the in-tree haddock, so we don't need to look for it. ] [Use $(TAR) rather than tar Ian Lynagh **20080924121759 Fixes problems on Solaris, where we need to use gtar instead of tar ] [Add $(strip) to a Makefile test Ian Lynagh **20080924120940 Fixes making bindists on solaris. Patch from Christian Maeder. ] [Use test -f rather than test -e, for portability (Solaris) Ian Lynagh **20080924120840] [Remove some dependencies on bootstrapping.conf from libraries/Makefile Ian Lynagh **20080923205755 They were causing some unnecessary work: Running make in a built tree reregisters the GHC package in bootstrapping.conf, and the build system thought that this updated timestamp meant that the configure stamps were out of date. This is particularly bad for the libraries with configure scripts, as those take a while to run. The bootstrapping.conf is built in an earlier phase ("make boot") so one shouldn't rely on the dependencies anyway. ] [Bump the version number to 6.11 Ian Lynagh **20080923165613] [Generalise type of 'defaultErrorHandler' so it can be used inside a Ghc session. Thomas Schilling **20080921085647] [Make "sh -e boot" work Ian Lynagh **20080921111508] [Use -f rather than -e for portability Ian Lynagh **20080921111436] [Add some special cases for putting dph in bindists Ian Lynagh **20080921000406] [Escape a hash in the Makefile (it was breaking source dist creation) Ian Lynagh **20080920232945] [Disallow package flags in OPTIONS_GHC pragmas (#2499) Simon Marlow **20080923173904] [#2566: emit a warning for 'ghc -c foo.bar' Simon Marlow **20080923144956 $ ghc -c foo.bar Warning: the following files would be used as linker inputs, but linking is not being done: foo.bar ghc: no input files Usage: For basic information, try the `--help' option. ] [Fix to new executable allocation code (fixed print002 etc.) Simon Marlow **20080922210915 The problem here is caused by the fact that info tables include a relative offset to the string naming the constructor. Executable memory now resides at two places in the address space: one for writing and one for executing. In the info tables generated by GHCi, we were calculating the offset relative to the writable instance, rather than the executable instance, which meant that the GHCi debugger couldn't find the names for constructors it found in the heap. ] [clean sm/Evac_thr.c and sm/Scav_thr.c Simon Marlow **20080922152827] [add -XNewQualifiedOperators (Haskell' qualified operator syntax) Simon Marlow **20080922152340] [Fix Trac #2597 (first bug): correct type checking for empty list simonpj@microsoft.com**20080920212010 The GHC front end never generates (ExplicitList []), but TH can. This patch makes the typechecker robust to such programs. ] [Fix Trac #2597 (second bug): complain about an empty DoE block simonpj@microsoft.com**20080920211101 When converting an empty do-block from TH syntax to HsSyn, complain rather than crashing. ] [Update dependencies Ian Lynagh **20080920183534] [Fix building with GHC 6.6 Ian Lynagh **20080920162918] [Remove fno-method-sharing from the list of static flags Ian Lynagh **20080920010635 It is now a dynamic flag ] [Tidy up the treatment of dead binders simonpj@microsoft.com**20080920175238 This patch does a lot of tidying up of the way that dead variables are handled in Core. Just the sort of thing to do on an aeroplane. * The tricky "binder-swap" optimisation is moved from the Simplifier to the Occurrence Analyser. See Note [Binder swap] in OccurAnal. This is really a nice change. It should reduce the number of simplifier iteratoins (slightly perhaps). And it means that we can be much less pessimistic about zapping occurrence info on binders in a case expression. * For example: case x of y { (a,b) -> e } Previously, each time around, even if y,a,b were all dead, the Simplifier would pessimistically zap their OccInfo, so that we can't see they are dead any more. As a result virtually no case expression ended up with dead binders. This wasn't Bad in itself, but it always felt wrong. * I added a check to CoreLint to check that a dead binder really isn't used. That showed up a couple of bugs in CSE. (Only in this sense -- they didn't really matter.) * I've changed the PprCore printer to print "_" for a dead variable. (Use -dppr-debug to see it again.) This reduces clutter quite a bit, and of course it's much more useful with the above change. * Another benefit of the binder-swap change is that I could get rid of the Simplifier hack (working, but hacky) in which the InScopeSet was used to map a variable to a *different* variable. That allowed me to remove VarEnv.modifyInScopeSet, and to simplify lookupInScopeSet so that it doesn't look for a fixpoint. This fixes no bugs, but is a useful cleanup. * Roman pointed out that Id.mkWildId is jolly dangerous, because of its fixed unique. So I've - localied it to MkCore, where it is private (not exported) - renamed it to 'mkWildBinder' to stress that you should only use it at binding sites, unless you really know what you are doing - provided a function MkCore.mkWildCase that emodies the most common use of mkWildId, and use that elsewhere So things are much better * A knock-on change is that I found a common pattern of localising a potentially global Id, and made a function for it: Id.localiseId ] [Gix the ghcii script Ian Lynagh **20080919174651 The ghc executable name doesn't have a version number on Windows, so don't put one in the script. ] [Create runhaskell as well as runghc Ian Lynagh **20080919153010] [On Linux use libffi for allocating executable memory (fixed #738) Simon Marlow **20080919134602] [Move the context_switch flag into the Capability Simon Marlow **20080919102601 Fixes a long-standing bug that could in some cases cause sub-optimal scheduling behaviour. ] [Fix building the extralibs tarball Ian Lynagh **20080919133555 We now need to dig the appropriate lines out of packages, rather than just catting libraries/extra-packages, in order to find out what the extralibs are. ] [Install libffi when installing frmo a bindist Ian Lynagh **20080919130332] [Fix how we put libffi into bindists Ian Lynagh **20080919125528] [TAG 6.10 branch has been forked Ian Lynagh **20080919123437] Patch bundle hash: 33bc121a485cb82b8d51deff964beae6c6552c69