simonpj@microsoft.com [Wed, 20 Jun 2007 16:26:56 +0000 (16:26 +0000)]
Use -X for language extensions
We've often talked about having a separate flag for language extensions,
and now we have one. You can say
-XImplicitParams
-X=ImplicitParams
-Ximplicit-params
as you like. These replace the "-f" flags with similar names (though
the -f prefix will serve as a synonym for -X for a while).
There's an optional "=", and the flag is normalised by removing hyphens
and lower-casing, so all the above variants mean the same thing.
The nomenclature is intended to match the LANGUAGE pramgas, which are
defined by Cabal. So you can also say
{-# LANGUAGE ImplicitParams #-}
But Cabal doesn't have as many language options as GHC does, so the -X
things are a superset of the LANGUAGE things.
The optional "=" applies to all flags that take an argument, so you can,
for example, say
-pgmL=/etc/foo
I hope that's ok. (It's an unforced change; just fitted in.)
I hope we'll add more -X flags, to replace the portmanteau -fglasgow-exts
which does everything!
I have updated the manual, but doubtless missed something.
simonpj@microsoft.com [Wed, 20 Jun 2007 16:13:41 +0000 (16:13 +0000)]
Unused import
andy@galois.com [Wed, 20 Jun 2007 15:02:44 +0000 (15:02 +0000)]
turning back on case liberation when using hpc
Simon Marlow [Wed, 20 Jun 2007 12:57:42 +0000 (12:57 +0000)]
remove debugging code accidentally left in
Ian Lynagh [Wed, 20 Jun 2007 12:07:23 +0000 (12:07 +0000)]
Use .NOTPARALLEL for the libraries Makefile
simonpj@microsoft.com [Wed, 20 Jun 2007 10:28:28 +0000 (10:28 +0000)]
Improve the handling of deriving, in error cases
I'd been too ambitious with error handling for 'deriving', and got it
entirely wrong. This fixes it. See extensive
Note [Exotic derived instance contexts]
in TcSimplify. (Most of the extra lines are comments!)
simonpj@microsoft.com [Wed, 20 Jun 2007 09:36:46 +0000 (09:36 +0000)]
More refactoring in TcSimplify
This re-jig tides up the top-level simplification, and combines in one
well-commented function, approximateImplications, the rather ad-hoc
way of simplifying implication constraints during type inference.
Error messages get a bit better too.
simonpj@microsoft.com [Wed, 20 Jun 2007 07:56:48 +0000 (07:56 +0000)]
Fix egregious sharing bug in LiberateCase
Andy Gill writes: consider the following code
f = g (case v of
V a b -> a : t f)
where g is expensive. Liberate case will turn this into
f = g (case v of
V a b -> a : t (letrec f = g (case v of
V a b -> a : f t)
in f)
)
Yikes! We evaluate g twice. This leads to a O(2^n) explosion
if g calls back to the same code recursively.
This may be the same as Trac #1366.
andy@galois.com [Wed, 20 Jun 2007 06:49:43 +0000 (06:49 +0000)]
Turning off case liberation when using the hpc option, for now
Consider the following code
f = g (case v of
V a b -> a : t f)
where g is expensive. Liberate case will turn this into
f = g (case v of
V a b -> a : t (letrec f = g (case v of
V a b -> a : f t)
in f)
)
Yikes! We evaluate g twice. This leads to a O(2^n) explosion
if g calls back to the same code recursively.
This happen sometimes in HPC, because every tick is a liberate-able case,
but is a general problem to case liberation (I think).
Ian Lynagh [Tue, 19 Jun 2007 20:05:46 +0000 (20:05 +0000)]
Add --core-only flag to push-all
Ian Lynagh [Tue, 19 Jun 2007 19:28:20 +0000 (19:28 +0000)]
Add a push-all script
simonpj@microsoft.com [Tue, 19 Jun 2007 16:53:54 +0000 (16:53 +0000)]
Improve misleading warning (Trac #1422)
simonpj@microsoft.com [Tue, 19 Jun 2007 16:26:13 +0000 (16:26 +0000)]
Fix a bug in the handling of implication constraints (Trac #1430)
Trac #1430 showed up quite a nasty bug in the handling of implication
constraints when we are *inferring* the type of a function.
See Note [Inference and implication constraints]:
We can't (or at least don't) abstract over implications. But we might
have an implication constraint (perhaps arising from a nested pattern
match) like
C a => D a
when we are now trying to quantify over 'a'. Our best approximation
is to make (D a) part of the inferred context, so we can use that to
discharge the implication. Hence getImplicWanteds.
My solution is not marvellous, but it's better than before. I transferred
function getDefaultableDicts from Inst to TcSimplify (since it's only
called there). Many of the remaining 50 new lines are comments. But
there is undoubtedly more code than before (sigh).
Test is tc228.
simonpj@microsoft.com [Tue, 19 Jun 2007 16:26:03 +0000 (16:26 +0000)]
Comments only
simonpj@microsoft.com [Tue, 19 Jun 2007 08:46:34 +0000 (08:46 +0000)]
Remove erroneous requirement to import Control.Monad.Fix when using mdo
See Trac #1426
andy@galois.com [Tue, 19 Jun 2007 05:56:54 +0000 (05:56 +0000)]
First cut at documentation for HPC option in GHC
Manuel M T Chakravarty [Tue, 19 Jun 2007 01:15:10 +0000 (01:15 +0000)]
Build package ndp if present
Simon Marlow [Mon, 18 Jun 2007 11:18:17 +0000 (11:18 +0000)]
typo
Ian Lynagh [Mon, 18 Jun 2007 10:28:50 +0000 (10:28 +0000)]
More debugger output order consistency
Pepe Iborra [Sun, 17 Jun 2007 19:34:35 +0000 (19:34 +0000)]
Several changes to the code dealing with newtypes in :print
I simplified the code, killed some unreachable blocks, and renamed it so that it corresponds more accurately with what is explained in the technical report
http://www.dsic.upv.es/docs/bib-dig/informes/etd-
04042007-111431/papernew2.pdf
Also, fixed a bug related to newtypes in the pretty printer
Pepe Iborra [Wed, 13 Jun 2007 09:21:02 +0000 (09:21 +0000)]
Remove now non-existant "Breakpoints" entry from package.conf.in
Ian Lynagh [Sun, 17 Jun 2007 21:52:05 +0000 (21:52 +0000)]
Sort names before printing them in the debugger so output order is consistent
Ian Lynagh [Sat, 16 Jun 2007 19:37:45 +0000 (19:37 +0000)]
Use %d rather than %zd on Windows
Ian Lynagh [Fri, 15 Jun 2007 18:45:27 +0000 (18:45 +0000)]
Add missing quotes in generated script
Ian Lynagh [Fri, 15 Jun 2007 18:23:37 +0000 (18:23 +0000)]
Fix size mismatch errors in mkDerivedConstants.c
Simon Marlow [Thu, 14 Jun 2007 09:57:27 +0000 (09:57 +0000)]
workaround for #1421 (Solaris linker being picky about .size)
Simon Marlow [Wed, 13 Jun 2007 14:45:05 +0000 (14:45 +0000)]
I didn't quite fix #1424 completely - hopefully this gets it right
Simon Marlow [Wed, 13 Jun 2007 14:24:31 +0000 (14:24 +0000)]
FIX #1424: x86_64 NCG generated wrong code for foreign call with >8 double args
I guess we have a missing test... I'll add one
Simon Marlow [Wed, 13 Jun 2007 11:15:52 +0000 (11:15 +0000)]
another fix for -hb: we appear to be freeing the hash table and arena twice
Simon Marlow [Wed, 13 Jun 2007 10:29:28 +0000 (10:29 +0000)]
FIX #1418 (partially)
When the con_desc field of an info table was made into a relative
reference, this had the side effect of making the profiling fields
(closure_desc and closure_type) also relative, but only when compiling
via C, and the heap profiler was still treating them as absolute,
leading to crashes when profiling with -hd or -hy.
This patch fixes up the story to be consistent: these fields really
should be relative (otherwise we couldn't make shared versions of the
profiling libraries), so I've made them relative and fixed up the RTS
to know about this.
Simon Marlow [Wed, 13 Jun 2007 09:52:01 +0000 (09:52 +0000)]
should be using GET_CON_DESC() to get the constructor name
Simon Marlow [Wed, 13 Jun 2007 09:51:44 +0000 (09:51 +0000)]
warning police
Simon Marlow [Wed, 13 Jun 2007 08:51:38 +0000 (08:51 +0000)]
Use $(if...) to get lazy tests instead of if..endif in a few places
This means it should be possible to set GhcUnregisterised=YES in
build.mk and the rest of the settings should follow automatically
(GhcWithNativeCodeGen, SplitObjs, GhcWithSMP).
Ian Lynagh [Tue, 12 Jun 2007 21:07:38 +0000 (21:07 +0000)]
Tweak banner printing
* -{short,long}-ghci-banner are now dynamic options, so you can put
":set -short-ghci-banner" in .ghci
* The -v2 banner information now always tells you what compiler booted GHC,
and what stage the compiler is. Thus we no longer assume that stage > 1
iff GHCI is defined.
cdsmith@twu.net [Fri, 1 Jun 2007 05:36:44 +0000 (05:36 +0000)]
FIX #1378 Add option for a shorter banner on GHCi startup
Add -short-ghci-banner and -long-ghci-banner. The default is long, which is
the current behavior. The short banner prints a one-line introduction with
only the version, web site, and ":? for help" message.
Ian Lynagh [Sat, 9 Jun 2007 15:14:38 +0000 (15:14 +0000)]
Windows bindist tweaking
Ian Lynagh [Sat, 9 Jun 2007 13:43:02 +0000 (13:43 +0000)]
Fix the ghc package in bindists
Ian Lynagh [Sat, 9 Jun 2007 11:22:35 +0000 (11:22 +0000)]
Fix up whitespace
Ian Lynagh [Sat, 9 Jun 2007 11:19:20 +0000 (11:19 +0000)]
Fix bindists on Windows
Ian Lynagh [Tue, 5 Jun 2007 13:39:56 +0000 (13:39 +0000)]
Install the RTS from a bindist correctly
Ian Lynagh [Mon, 4 Jun 2007 13:47:14 +0000 (13:47 +0000)]
Only add the extra Windows libraries if we are on Windows
Ian Lynagh [Mon, 4 Jun 2007 13:46:45 +0000 (13:46 +0000)]
Follow Cabal change; .setup-config is now dist/setup-config
Ian Lynagh [Sat, 2 Jun 2007 19:56:51 +0000 (19:56 +0000)]
Copy more of base's hacks into installPackage; *sigh*
Ian Lynagh [Sat, 2 Jun 2007 19:38:26 +0000 (19:38 +0000)]
Hack libsubdir so that it does the right thing on both Windows and non-Windows
Ian Lynagh [Sat, 2 Jun 2007 12:11:59 +0000 (12:11 +0000)]
Remove now-unused files
Ian Lynagh [Sat, 2 Jun 2007 12:10:17 +0000 (12:10 +0000)]
Fix ghcii.sh creation
Ian Lynagh [Sat, 2 Jun 2007 01:51:01 +0000 (01:51 +0000)]
Explicitly set datadir
Cabal defaults to the value we want on Linux, but uses
"C:\\Program Files\\Common Files" on Windows.
Ian Lynagh [Fri, 1 Jun 2007 22:43:01 +0000 (22:43 +0000)]
Fix a -- that somehow turned into a ----
Ian Lynagh [Fri, 1 Jun 2007 22:33:55 +0000 (22:33 +0000)]
Fix creation of hsc2hs-inplace.bat
Ian Lynagh [Fri, 1 Jun 2007 21:49:53 +0000 (21:49 +0000)]
ghc-pkg-inplace now has .bat extension on Windows
Ian Lynagh [Fri, 1 Jun 2007 21:05:55 +0000 (21:05 +0000)]
Add omitted files (mk/{install,recurse}.mk)
Ian Lynagh [Fri, 1 Jun 2007 19:06:59 +0000 (19:06 +0000)]
Fix bindist creation
Bindists should now work again, when doing "make install" at least.
"make in-place" is probably still broken.
Ian Lynagh [Thu, 31 May 2007 14:35:05 +0000 (14:35 +0000)]
Rework the build system a bit
Key changes:
* Always build as if BIN_DIST is 1. BIN_DIST is thus removed.
* Libraries are configured with prefix set to $$topdir rather than $(prefix)
Ian Lynagh [Wed, 30 May 2007 14:46:33 +0000 (14:46 +0000)]
Make package.conf files a bit more readable
andy@galois.com [Tue, 12 Jun 2007 07:46:55 +0000 (07:46 +0000)]
Adding new ffi calls into the Hpc rts subsystem
foreign import ccall unsafe hs_hpc_write :: CString -> IO ()
foreign import ccall unsafe hs_hpc_read :: CString -> IO ()
These write a Hpc description of the state of the world to a file,
or read a description into the current Hpc tickbox subsystem.
Simon Marlow [Tue, 12 Jun 2007 08:17:03 +0000 (08:17 +0000)]
don't add -threaded in stage2/3 if we didn't build the threaded RTS
Simon Marlow [Tue, 12 Jun 2007 07:46:57 +0000 (07:46 +0000)]
fix compile error in the !GHCI_TABLES_NEXT_TO_CODE case
Thorkil Naur [Fri, 18 May 2007 16:59:57 +0000 (16:59 +0000)]
Fix PPC Mac OS X memory access problem in SMP.h (#1362)
andy@galois.com [Sat, 9 Jun 2007 00:00:21 +0000 (00:00 +0000)]
Using blacklist of places not to cover, rather than reverse-engineer deriving.
andy@galois.com [Fri, 8 Jun 2007 22:40:16 +0000 (22:40 +0000)]
Not adding ticks to compiler generated derived code.
Manuel M T Chakravarty [Fri, 8 Jun 2007 06:10:46 +0000 (06:10 +0000)]
Fix worker/wrapper ids for newtype instances
- mkDataConIds forced an algebraic worker/wrapper on newtype instances (not
on vanilla newtypes). I am not sure where this came from (from code I
wrote or from the recent restructuring of MkIds), but its not the right
thing to do. In particular, it requires newtype instances to have con_info
labels in the generated code, which makes no sense.
- Worker construction for newtypes is smart enough that it will construct the
right signature for newtype instances (ie, one that mentions the family
type constructor, not the instance tycon).
Manuel M T Chakravarty [Fri, 8 Jun 2007 04:44:00 +0000 (04:44 +0000)]
Fix deriving of indexed data types
- Fix typo introduced when moving deriving-specific code from TcEnv to TcDeriv
- Thanks to Roman for the bug report
Ian Lynagh [Thu, 7 Jun 2007 13:55:46 +0000 (13:55 +0000)]
Add instructions for adding a library to a build tree
simonpj@microsoft.com [Wed, 6 Jun 2007 13:34:37 +0000 (13:34 +0000)]
Don't suggest -fno-monomorphism-restriction if it's already set
This patch implements the suggestion in Trac #1398. It's obviously
stupid to suggest -fno-monomorphism-restriction if the user is already
using it.
(Maybe another suggestion would be good, but this one clearly bogus.)
simonpj@microsoft.com [Wed, 6 Jun 2007 13:25:21 +0000 (13:25 +0000)]
Remove unnecessary free-variables from renamer
The renamer used to be responsible for making sure that all interfaces
with instance decls (other than orphans) were loaded. But TH makes that
impossible, so the typechecker does it, via checkWiredInTyCon.
This patch simply removes redundant additions to the free-variable set
in the renamer, which were there, I believe, solely to ensure that the
instances came in. Removing them should change nothing, but it's a
useful clean up.
simonpj@microsoft.com [Wed, 6 Jun 2007 13:00:56 +0000 (13:00 +0000)]
FIX #1385: make sure tuple instances are loaded
Typechecking of explicit tuples is done "by hand" (rather than using
boxySplitTyConApp) so that rigidity is maintained. In making that change
I'd forgotten to retain the call to checkWiredInTyCon, which is important
to make sure instances are brought into scope.
Test is tc227.
simonpj@microsoft.com [Tue, 5 Jun 2007 16:14:54 +0000 (16:14 +0000)]
Typo in comments
Isaac Dupree [Tue, 5 Jun 2007 23:53:01 +0000 (23:53 +0000)]
remove #if branches for pre-ghc-6.0
I skipped utils/hsc2hs/Main.hs since its ifs also involved
checking for old versions of nhc98 (I don't want to figure that out),
but removed everything else I found relating to building with pre-6.0
Ian Lynagh [Tue, 5 Jun 2007 19:46:43 +0000 (19:46 +0000)]
Typo, spotted by Stefan Holdermans
Ian Lynagh [Tue, 5 Jun 2007 17:47:25 +0000 (17:47 +0000)]
ifBuildable now prints a warning if the package is not buildable
Ian Lynagh [Tue, 5 Jun 2007 14:50:43 +0000 (14:50 +0000)]
Fix normalisation of path to find
The old code didn't work on Mac OS X, as "which" returns successfully
when it can't find the program.
Simon Marlow [Tue, 5 Jun 2007 08:44:38 +0000 (08:44 +0000)]
fix panic in #1379
bjpop@csse.unimelb.edu.au [Tue, 5 Jun 2007 02:09:30 +0000 (02:09 +0000)]
Add missing newline from ghci :help output
Isaac Dupree [Fri, 1 Jun 2007 20:49:48 +0000 (20:49 +0000)]
implement FastMutInt in non-GHC using IORefs (#1405)
ghc still works, also the module was tested in hugs and ghc
Simon Marlow [Fri, 1 Jun 2007 15:19:32 +0000 (15:19 +0000)]
FIX #1110: the linker also needs the workaround
Simon Marlow [Mon, 21 May 2007 14:17:45 +0000 (14:17 +0000)]
FIX BUILD: GHC doesn't use filepath (yet)
Simon Marlow [Fri, 25 May 2007 08:22:35 +0000 (08:22 +0000)]
tiny cleanup
Simon Marlow [Fri, 1 Jun 2007 08:26:48 +0000 (08:26 +0000)]
remove OpenBSD-specific initAdjustor() code that shouldn't be required now
Submitted by: Matthias Kilian <kili@outback.escape.de>
Simon Marlow [Thu, 31 May 2007 13:41:13 +0000 (13:41 +0000)]
fix locations for uploading src dists
Simon Marlow [Thu, 31 May 2007 13:28:42 +0000 (13:28 +0000)]
FIX unregisterised build
Can't check $(GhcUnregisterised) eagerly, because it might not have
been set yet (it's set in build.mk).
Simon Marlow [Thu, 31 May 2007 11:05:06 +0000 (11:05 +0000)]
FIX: loading package ghc in GHCi (added a couple of missing symbols)
simonpj@microsoft.com [Wed, 30 May 2007 14:14:44 +0000 (14:14 +0000)]
Fix bogus check for strictness in newtypes
Simon Marlow [Wed, 30 May 2007 13:31:46 +0000 (13:31 +0000)]
export noSrcLoc, noSrcSpan
Ian Lynagh [Wed, 30 May 2007 10:56:29 +0000 (10:56 +0000)]
Fix copy-and-paste-o
simonpj@microsoft.com [Wed, 30 May 2007 08:12:40 +0000 (08:12 +0000)]
Fix bug in tcSimplifyInfer (Trac #1382)
When I rejigged constraint simplification when inferring types, I missed
a corner case. Somethign that it's distressingly easy to do. Anyway
this fixes it; see the comments in tcSimplifyInfer with the second call
to partition and extendLIEs.
The test is tcfail181.
simonpj@microsoft.com [Wed, 30 May 2007 07:16:25 +0000 (07:16 +0000)]
Reject newtypes with strictness annotations; fixes read008
This used to be a parse error, but ! annotations are now handled
further downstream in the compiler, and I'd forgotten to check
that newtypes do not have strictness annotations.
The test read008 is technically in the wrong place (it's a typechecker
test now) but that doesn't matter
iavor.diatchki@gmail.com [Fri, 25 May 2007 16:43:08 +0000 (16:43 +0000)]
refactor: combine repeated code in file reloading
iavor.diatchki@gmail.com [Fri, 25 May 2007 16:31:04 +0000 (16:31 +0000)]
Changes the behavior of the "edit" command in GHCi.
Currently, when the user invokes the GHCi editor without a parameter
GHCi chooses the last file that loaded _sucessfully_.
This patch changes this behavior to start the editor with the
last file that _faild_ (this being what I usually want to edit).
Caveat: we use the module graph to determine what happened
and so we do not catch errors that occured during dependecy
analysis (e.g., a mis-typed module name) but we could fix this later.
Ian Lynagh [Tue, 29 May 2007 15:28:48 +0000 (15:28 +0000)]
Update Windows installation docs
Isaac Dupree [Sat, 26 May 2007 23:00:45 +0000 (23:00 +0000)]
#1318: lex negative unboxed literals
I reorganized the lexing of numeric literals a bit so the code didn't
get too ugly, after trying a few ways, and also considering possible plans
to be able to conditionally lex negative _boxed_ literals.
Isaac Dupree [Sat, 26 May 2007 21:50:42 +0000 (21:50 +0000)]
#1318: remove negative-prim-literal old hackish implementation
Isaac Dupree [Sat, 26 May 2007 21:22:04 +0000 (21:22 +0000)]
parseInteger->parseUnsignedInteger to clarify meaning
I decided against adding parseSignedInteger since octal
and hex literals often have junk between the '-' and the
digits, but, compare to Util.readRational which does handle
signed numbers. Also since Integers - mathematically and
in Haskell - can be negative, normally.
Ian Lynagh [Tue, 29 May 2007 10:21:43 +0000 (10:21 +0000)]
Build Windows installer when making a bindist if ISCC is set
It doesn't work yet, but I'm pretty sure that's because the bindist is
broken rather than the installer is broken.
Simon Marlow [Tue, 29 May 2007 07:42:24 +0000 (07:42 +0000)]
don't build the threaded RTS when GhcUnregisterised=YES
Isaac Dupree [Sat, 26 May 2007 18:08:00 +0000 (18:08 +0000)]
only comments, spacing, alpha-renaming
Michael D. Adams [Fri, 25 May 2007 20:51:26 +0000 (20:51 +0000)]
Removed defunct compiler/codeGen/CgUsages.hi-boot-6
Simon Marlow [Fri, 25 May 2007 14:42:16 +0000 (14:42 +0000)]
-package pretty is required for GHC >= 6.7
Simon Marlow [Fri, 25 May 2007 14:42:02 +0000 (14:42 +0000)]
add $(ghc_ge_607)
Simon Marlow [Fri, 25 May 2007 08:09:44 +0000 (08:09 +0000)]
System.Directory.Internals is now in package directory
Ian Lynagh [Thu, 24 May 2007 17:50:31 +0000 (17:50 +0000)]
Say where we're booting to aid debugging