GitLab

Re-arrange the interface to TcMatches to allow typechecking of case
commands (part of arrow notation):

* replace the export of the internal tcGRHSs with a more specific
  tcGRHSsPat for checking PatMonoBinds.

* generalize match contexts in the same way as stmt contexts, to include
  a typechecker for the bodies of alts.

This should probably be reviewed, but I hope it can make it into STABLE
after a while.

fixes to desugaring of arrow notation:

* fix free variable calculation for if's

* various fixes for case

please merge to STABLE

POSIX threads should be pthread_detach-ed after they are created.
Otherwise they'll stick around waiting for a pthread_join.

(This wasn't really a problem because our worker threads never exit anyway).

The whole family of index<blah>OffForeignPtr# primops has been broken
for some time now.  This commit fixes them.

I'm tempted to get rid of them altogether, since arguably the
withForeignPtr interface subsumes all this.  But if you want to index
off a ForeignPtr in pure code, you have to use unsafePerformIO, which
is hard to optimise away (GHC doesn't do it), so these primops still
have their uses.

Be a bit more honest about how easy it is to port the linker.

Add irix_TARGET_OS support.

usage_text[]: -k doc wibble

Simplified my previous quick-fix-after-some-beers

Switch examples on/off

Make "one of the most hammered bits in the whole compiler" (quotation
from the source code :-) compile without _ccall_

CasmTarget is gone

-funfold-casms-in-hi-file is no more.

Legacy Removal
~~~~~~~~~~~~~~

The following features have been consigned to the bit bucket:

  _ccall_
  _casm_
  ``....''  (lit-lits)
  the CCallable class
  the CReturnable class

factor out code that handles depositing of work items on the
  thread pool's request queue.
- when it looks as if a new worker thread needs to be created, give
  up our quantum first in the hope that this might at the last minute
  turn up more idle worker threads.
- add comment re: trimming pool size.

Merged to STABLE eventually; I may continue tinkering with this code
some more over the next day or two.

Update comment for GhcUnregisterised.

comment wibble
(merge to stable)

* Corrected some outdated comments re variables and type checking
* Added some more details about type checking

sparc-unknown-openbsd exists and is supported.

Also, we run on Solaris 2.6 as well as 2.7.

Update IA64 status (supported except no NCG)

delayzh_fast: comment only

- Sleep()'s resolution is millisecs, not microsecs.
- adopt a more agressive policy for augmenting the thread pool
  to handle incoming requests (see code comments for details.)

  The previous policy ran the risk of starvation in rare (and hard
  to reproduce) cases, as spotted after having chased a bug
  for two days.

Merge to STABLE

- awaitRequests() comments
- code reformatting

Merge to STABLE

comment wibble

Ignore lines containing

  #pragma .*

generated by GCC 3.3's CPP on MacOS X, apparently.

Make sure that exprIsTrivial responds "False" to
	(scc "foo" x)

We do not treat (_scc_ "foo" x) as trivial, because
  a) it really generates code, (and a heap object when it's
     a function arg) to capture the cost centre
  b) see the note [SCC-and-exprIsTrivial] in Simplify.simplLazyBind

The FP_CHECK_CONSTS test for the errno constants requires stdio.h to be
included, at least on Mac OS X.

MERGE TO STABLE.

Mac OS X:

Bypass the dlfcn compatibility layer in lookupSymbol, use Apple's
NSModule functions directly instead. This protects us from differences
between different implementations of the dlcompat library (regarding
treatment of prefixed underscores).
Note that the compatibility layer is still used in addDLL (it's needed by
the POSIX library anyway).

Also, add a little more sanity checking (checkProddableBlock) to the Mach-O
linker itself.

MERGE TO STABLE

Document -fth and -fimplicit-params

--------------------------
	    Allow recursive dictionaries
	   --------------------------

In response to various bleatings, here's a lovely fix that involved simply
inverting two lines of code, to allow recursive dictionaries.  Here's
the comment.  (typecheck/should_run/tc030 tests it)

Note [RECURSIVE DICTIONARIES]
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Consider
    data D r = ZeroD | SuccD (r (D r));

    instance (Eq (r (D r))) => Eq (D r) where
        ZeroD     == ZeroD     = True
        (SuccD a) == (SuccD b) = a == b
        _         == _         = False;

    equalDC :: D [] -> D [] -> Bool;
    equalDC = (==);

We need to prove (Eq (D [])).  Here's how we go:

	d1 : Eq (D [])

by instance decl, holds if
	d2 : Eq [D []]
	where 	d1 = dfEqD d2

by instance decl of Eq, holds if
	d3 : D []
	where	d2 = dfEqList d2
		d1 = dfEqD d2

But now we can "tie the knot" to give

	d3 = d1
	d2 = dfEqList d2
	d1 = dfEqD d2

and it'll even run!  The trick is to put the thing we are trying to prove
(in this case Eq (D []) into the database before trying to prove its
contributing clauses.

Add comments, change a WARN to an ASSERT

Undo bogus claim that
	_scc_ "foo" x  ==  x

This just isn't right in general, for example when x is a
top-level function.

HEAD is now 6.3

When setting TARGETPLATFORM, we should also set foo_TARGET_ARCH,
bar_TARGET_OS, etc.

Add missing 'cd T' to the porting instructions.

The reify keywords should be valid when template haskell is turned on.

hGetStringBuffer: open file in binary mode

Fix race condition re: WorkItems and SubmitWork().

It is unsafe to access a WorkItem after it has been queued, since
a worker thread may have already serviced (and freed) it.
Hence, return request IDs without looking at the WorkItem.

If desired, this one could be merged to STABLE.

awaitRequests(): canonicalize removal of thread from blocked_queue

New flags for individual syntax extensions:

  -fth                enables template haskell
  -fimplicit-params   enables implicit parameters

These extensions are still implied by -fglasgow-exts, but they can now
be switched off individually with -fno-th and -fno-implicit-params
respectively.  Also, -fno-ffi now works as expected.

I cleaned up the interface to the lexer a bit while I was here.

Quick hack to work around the dynamic exception crashes in GHCi.  The
problem is this: the Data.Typeable library (Data.Dymamic in previous
versions of GHC) maintains an internal cache mapping type names to
unique numbers, and this cache is used for fast comparisons on
TypeReps.  In GHCi, there are actually two versions of the
Data.Typeable library loaded: one linked to GHCi itself, and the other
dynamically loaded, so there are two copies of the hash tables.

The problem is that if we have a Dynamic value generated using one set
of hash tables, it will erroneously appear to be of a different type
when the other hash tables are used.

The hack I've instigated is to use the central RTS genSym (which
already exists) to generate the unique Ids, so that the two copies of
the dynamic library will be using distinct Ids, and Dynamics from one
will never be recognisable to the other.