simonpj@microsoft.com [Thu, 19 Aug 2010 10:48:04 +0000 (10:48 +0000)]
Be a bit less aggressive in mark-many inside a cast
simonpj@microsoft.com [Wed, 18 Aug 2010 18:57:38 +0000 (18:57 +0000)]
Wibble
simonpj@microsoft.com [Wed, 18 Aug 2010 06:54:36 +0000 (06:54 +0000)]
Pretty printing change
simonpj@microsoft.com [Wed, 11 Aug 2010 14:35:55 +0000 (14:35 +0000)]
Remember to zonk FlatSkols!
simonpj@microsoft.com [Fri, 30 Jul 2010 15:12:17 +0000 (15:12 +0000)]
De-polymorphise
simonpj@microsoft.com [Fri, 30 Jul 2010 12:24:05 +0000 (12:24 +0000)]
Work around missing type signature in Happy
Happy generates
notHappyAtAll = error "Blah"
without a type signature, and currently the new
typechecker doesn't generalise it. This patch
says "no monomorphism restriction" which makes it
generalise again.
Better would be to add a type sig to Happy's template
simonpj@microsoft.com [Thu, 29 Jul 2010 15:26:11 +0000 (15:26 +0000)]
Add two local type signatures
simonpj@microsoft.com [Thu, 29 Jul 2010 09:17:03 +0000 (09:17 +0000)]
Second test from Simon's laptop
simonpj@microsoft.com [Thu, 29 Jul 2010 09:13:44 +0000 (09:13 +0000)]
Test commit from Simon's laptop
simonpj@microsoft.com [Wed, 28 Jul 2010 07:55:25 +0000 (07:55 +0000)]
Add OutsideIn flag
simonpj@microsoft.com [Tue, 27 Jul 2010 14:15:39 +0000 (14:15 +0000)]
Layout only
simonpj@microsoft.com [Tue, 27 Jul 2010 13:16:59 +0000 (13:16 +0000)]
Improvement to SimplUtils.mkLam
simonpj@microsoft.com [Tue, 27 Jul 2010 13:15:38 +0000 (13:15 +0000)]
Give the correct kind to unsafeCoerce#
simonpj@microsoft.com [Tue, 27 Jul 2010 09:45:49 +0000 (09:45 +0000)]
Suppress warnings about recursive INLINE in output of desugarer
simonpj@microsoft.com [Mon, 26 Jul 2010 23:12:53 +0000 (23:12 +0000)]
Rename CorePrep.tryEtaReduce to tryEtaReducePrep
This avoids the name clash with the similar but
not identical CoreUtils.tryEtaReduce
simonpj@microsoft.com [Mon, 19 Jul 2010 21:11:11 +0000 (21:11 +0000)]
Add a trace message
simonpj@microsoft.com [Mon, 19 Jul 2010 17:04:41 +0000 (17:04 +0000)]
Don't use RelaxedPolyRec in the compiler; it's built in now
Matthias Kilian [Sun, 12 Sep 2010 18:15:18 +0000 (18:15 +0000)]
Remove duplicated #defines for FreeBSD
Matthias Kilian [Sun, 12 Sep 2010 15:05:06 +0000 (15:05 +0000)]
Don't fail with absolute silence
Ian Lynagh [Sat, 11 Sep 2010 19:58:09 +0000 (19:58 +0000)]
Add a release note: "-dynload wrapper" removed
Ian Lynagh [Sat, 11 Sep 2010 19:34:34 +0000 (19:34 +0000)]
put back the conversion of warn-lazy-unlifted-bindings into an error until 7.2
I think we'll currently still have too many people with old versions of
alex/happy to want to make this an error now.
Ian Lynagh [Sat, 11 Sep 2010 19:28:37 +0000 (19:28 +0000)]
6.14 -> 7.0
Ian Lynagh [Sat, 11 Sep 2010 16:20:59 +0000 (16:20 +0000)]
Add a couple more release notes
Ian Lynagh [Sat, 11 Sep 2010 16:20:05 +0000 (16:20 +0000)]
Document -dsuppress-module-prefixes
Ian Lynagh [Fri, 10 Sep 2010 19:13:01 +0000 (19:13 +0000)]
Enable -fregs-graph with -O2; fixes #2790
Ian Lynagh [Thu, 9 Sep 2010 17:02:07 +0000 (17:02 +0000)]
Remove unused code
Ian Lynagh [Thu, 9 Sep 2010 15:43:48 +0000 (15:43 +0000)]
Fix warnings
Ian Lynagh [Thu, 9 Sep 2010 15:09:57 +0000 (15:09 +0000)]
Fix warnings
lykahb@gmail.com [Wed, 1 Sep 2010 16:01:53 +0000 (16:01 +0000)]
Remove context completion
Now completion suggests to remove only modules added to context before.
Ross Paterson [Thu, 9 Sep 2010 12:55:31 +0000 (12:55 +0000)]
avoid Foreign.unsafePerformIO
Simon Marlow [Thu, 9 Sep 2010 11:14:50 +0000 (11:14 +0000)]
updates to the release notes
Simon Marlow [Thu, 9 Sep 2010 11:08:05 +0000 (11:08 +0000)]
newAlignedPinnedByteArray#: avoid allocating an extra word sometimes
benl@ouroborus.net [Thu, 9 Sep 2010 06:13:11 +0000 (06:13 +0000)]
Finish breaking up vectoriser utils
benl@ouroborus.net [Thu, 9 Sep 2010 04:24:51 +0000 (04:24 +0000)]
Move VectType module to Vectorise tree
benl@ouroborus.net [Thu, 9 Sep 2010 03:51:47 +0000 (03:51 +0000)]
Sort all the PADict/PData/PRDict/PRepr stuff into their own modules
benl@ouroborus.net [Thu, 9 Sep 2010 02:57:59 +0000 (02:57 +0000)]
Break out Repr and PADict stuff for vectorisation of ADTs to their own modules
benl@ouroborus.net [Thu, 9 Sep 2010 02:33:32 +0000 (02:33 +0000)]
Break out conversion functions to own module
benl@ouroborus.net [Thu, 9 Sep 2010 02:21:17 +0000 (02:21 +0000)]
Comments and formatting only
Ian Lynagh [Wed, 8 Sep 2010 21:32:51 +0000 (21:32 +0000)]
Remove "-dynload wrapper"; fixes trac #4275
Ian Lynagh [Sun, 5 Sep 2010 12:24:42 +0000 (12:24 +0000)]
Don't set visibility on Windows
With gcc 4.5.0-1, using visibility hidden gives:
error: visibility attribute not supported in this configuration; ignored
Ian Lynagh [Sun, 5 Sep 2010 11:12:01 +0000 (11:12 +0000)]
Fix warnings on Windows
Ian Lynagh [Sun, 5 Sep 2010 00:18:07 +0000 (00:18 +0000)]
Fix gcc wrapper for new mingw binaries
Ian Lynagh [Sun, 5 Sep 2010 10:31:29 +0000 (10:31 +0000)]
Don't pass our gcc options to stage0 ghc's gcc; they may not be suitable
Ian Lynagh [Sat, 4 Sep 2010 22:55:59 +0000 (22:55 +0000)]
Update intree-mingw creation
Ian Lynagh [Sat, 4 Sep 2010 21:51:12 +0000 (21:51 +0000)]
Update commands to build in-tree mingw
benl@ouroborus.net [Wed, 8 Sep 2010 07:41:02 +0000 (07:41 +0000)]
Break out hoisting utils into their own module
benl@ouroborus.net [Wed, 8 Sep 2010 07:20:40 +0000 (07:20 +0000)]
Break out closure utils into own module
benl@ouroborus.net [Wed, 8 Sep 2010 06:59:04 +0000 (06:59 +0000)]
Move VectVar module to Vectorise tree
benl@ouroborus.net [Wed, 8 Sep 2010 06:51:28 +0000 (06:51 +0000)]
Break out vectorisation of expressions into own module
benl@ouroborus.net [Wed, 8 Sep 2010 06:31:11 +0000 (06:31 +0000)]
Break out TyCon classifier into own module
benl@ouroborus.net [Wed, 8 Sep 2010 05:20:04 +0000 (05:20 +0000)]
Break out vectorisation of TyConDecls into own module
benl@ouroborus.net [Tue, 7 Sep 2010 11:03:11 +0000 (11:03 +0000)]
Break out type vectorisation into own module
Ian Lynagh [Sat, 4 Sep 2010 14:42:34 +0000 (14:42 +0000)]
Tidy up the ArchHasAdjustorSupport definition
Sergei Trofimovich [Sun, 29 Aug 2010 19:28:59 +0000 (19:28 +0000)]
ppc: switch handling of 'foreign import wrapper' (FIW) to libffi
Joseph Jezak reported darcs-2.4.4 SIGSEGV in interactive mode in ghc-6.12.3.
So I've concluded ppc also has rotten native adjustor. I don't have hardware
to verify the patch (ticket #3516 should help to test it), but I think it will
help (as similar patch helped for ia64 and ppc64).
Ian Lynagh [Sat, 4 Sep 2010 13:33:39 +0000 (13:33 +0000)]
Binary no longer has unusable UNPACK pragmas, so no need to turn of -Werror
Ian Lynagh [Fri, 3 Sep 2010 23:19:21 +0000 (23:19 +0000)]
Don't haddock packages that we aren't going to install
Ian Lynagh [Fri, 3 Sep 2010 22:13:35 +0000 (22:13 +0000)]
Give haddock per-package source entity paths; fixes #3810
Simon Marlow [Fri, 3 Sep 2010 10:51:31 +0000 (10:51 +0000)]
update for containers-0.4
Simon Marlow [Fri, 27 Aug 2010 10:28:18 +0000 (10:28 +0000)]
Fix for nursery resizing: the first block's back pointer should be NULL
I'm not sure if this could lead to a crash or not, but it upsets +RTS -DS
Might be related to #4265
Ian Lynagh [Thu, 2 Sep 2010 22:47:26 +0000 (22:47 +0000)]
Add some -no-user-package-conf flags
Stops user-installed packages breaking the build
Ian Lynagh [Thu, 2 Sep 2010 20:43:42 +0000 (20:43 +0000)]
Fix warnings: Remove unused imports
benl@ouroborus.net [Tue, 31 Aug 2010 10:07:24 +0000 (10:07 +0000)]
Finish breaking up VectBuiltIn and VectMonad, and add comments
benl@ouroborus.net [Mon, 30 Aug 2010 08:37:46 +0000 (08:37 +0000)]
Fix warnings
benl@ouroborus.net [Mon, 30 Aug 2010 07:09:00 +0000 (07:09 +0000)]
Break up vectoriser builtins module
benl@ouroborus.net [Mon, 30 Aug 2010 05:34:15 +0000 (05:34 +0000)]
Move VectCore to Vectorise tree
benl@ouroborus.net [Mon, 30 Aug 2010 05:02:52 +0000 (05:02 +0000)]
Split out vectoriser environments into own module
benl@ouroborus.net [Mon, 30 Aug 2010 04:27:22 +0000 (04:27 +0000)]
Comments and formatting to vectoriser, and split out varish stuff into own module
benl@ouroborus.net [Mon, 30 Aug 2010 04:03:40 +0000 (04:03 +0000)]
Fix warnings
benl@ouroborus.net [Mon, 30 Aug 2010 03:52:43 +0000 (03:52 +0000)]
Fix warning about multiply exported name
benl@ouroborus.net [Mon, 30 Aug 2010 03:29:41 +0000 (03:29 +0000)]
Vectorisation of method types
benl@ouroborus.net [Mon, 30 Aug 2010 03:25:16 +0000 (03:25 +0000)]
Comments and formatting to vectoriser
benl@ouroborus.net [Mon, 30 Aug 2010 03:24:28 +0000 (03:24 +0000)]
Implement -dsuppress-module-prefixes
benl@ouroborus.net [Thu, 27 May 2010 04:56:29 +0000 (04:56 +0000)]
Whitespace only
Ian Lynagh [Sat, 28 Aug 2010 22:58:27 +0000 (22:58 +0000)]
Disambiguate a function name
Ian Lynagh [Sat, 28 Aug 2010 22:57:51 +0000 (22:57 +0000)]
users_guide.xml is now generated
Ian Lynagh [Fri, 27 Aug 2010 20:45:37 +0000 (20:45 +0000)]
Pass more -pgm flags in the ghc wrapper; fixes #3863
Ian Lynagh [Fri, 27 Aug 2010 17:16:16 +0000 (17:16 +0000)]
Add a new-IO manager release note
Ian Lynagh [Fri, 27 Aug 2010 17:15:11 +0000 (17:15 +0000)]
Merge a duplicate release note
Ian Lynagh [Fri, 27 Aug 2010 15:39:14 +0000 (15:39 +0000)]
Typo, spotted by Johan Tibell
Ian Lynagh [Thu, 26 Aug 2010 22:08:11 +0000 (22:08 +0000)]
First pass at 6.14.1 release notes
Ian Lynagh [Tue, 24 Aug 2010 20:13:30 +0000 (20:13 +0000)]
Fix typo
Simon Marlow [Thu, 26 Aug 2010 09:43:19 +0000 (09:43 +0000)]
FIX BUILD: add rts_isProfiled to the symbol table
Simon Marlow [Thu, 26 Aug 2010 08:44:36 +0000 (08:44 +0000)]
Fix the DPH package cleaning/profiled mess even more (the build was broken)
Simon Marlow [Tue, 24 Aug 2010 11:35:37 +0000 (11:35 +0000)]
Remove the debugging memory allocator - valgrind does a better job
I got fed up with the constant bogus output from the debugging memory
allocator in RtsUtils.c. One problem is that we allocate memory in
constructors that then isn't tracked, because the debugging allocator
hasn't been initialised yet.
The bigger problem is that for a given piece of leaking memory it's
impossible to find out where it was allocated; however Valgrind gives
output like this:
==6967== 8 bytes in 1 blocks are still reachable in loss record 1 of 7
==6967== at 0x4C284A8: malloc (vg_replace_malloc.c:236)
==6967== by 0x4C28522: realloc (vg_replace_malloc.c:525)
==6967== by 0x6745E9: stgReallocBytes (RtsUtils.c:213)
==6967== by 0x68D812: setHeapAlloced (MBlock.c:91)
==6967== by 0x68D8E2: markHeapAlloced (MBlock.c:116)
==6967== by 0x68DB56: getMBlocks (MBlock.c:240)
==6967== by 0x684F55: alloc_mega_group (BlockAlloc.c:305)
==6967== by 0x6850C8: allocGroup (BlockAlloc.c:358)
==6967== by 0x69484F: allocNursery (Storage.c:390)
==6967== by 0x694ABD: allocNurseries (Storage.c:436)
==6967== by 0x6944F2: initStorage (Storage.c:217)
==6967== by 0x673E3C: hs_init (RtsStartup.c:160)
which tells us exactly what the leaking bit of memory is. So I don't
think we need our own debugging allocator.
Simon Marlow [Tue, 24 Aug 2010 11:26:06 +0000 (11:26 +0000)]
free the entries in the thread label table on exit
simonpj@microsoft.com [Wed, 25 Aug 2010 09:16:14 +0000 (09:16 +0000)]
Panic in the right way
simonpj@microsoft.com [Wed, 25 Aug 2010 09:16:02 +0000 (09:16 +0000)]
Fix the DPH/profiled make thing (again)
Ian Lynagh [Tue, 24 Aug 2010 13:14:01 +0000 (13:14 +0000)]
Don't test for gcc flags before we've found gcc
Ian Lynagh [Tue, 24 Aug 2010 13:09:38 +0000 (13:09 +0000)]
Change how the dblatex/lndir problem is worked around
Hack: dblatex normalises the name of the main input file using
os.path.realpath, which means that if we're in a linked build tree,
it find the real source files rather than the symlinks in our link
tree. This is fine for the static sources, but it means it can't
find the generated sources.
We therefore also generate the main input file, so that it really
is in the link tree, and thus dblatex can find everything.
Ian Lynagh [Tue, 24 Aug 2010 10:58:27 +0000 (10:58 +0000)]
Clean the generated userguide sources
simonpj@microsoft.com [Mon, 23 Aug 2010 13:34:39 +0000 (13:34 +0000)]
DPH should not even be built if GhcProfiled
It's not just when cleaning!
simonpj@microsoft.com [Mon, 23 Aug 2010 13:33:53 +0000 (13:33 +0000)]
The templateHaskellOk check should only run in stage2
Because rtsIsProfiled is only available in stage2
Simon Marlow [Mon, 23 Aug 2010 10:46:54 +0000 (10:46 +0000)]
Add a couple of missing tests for EAGER_BLACKHOLE
This was leading to looping and excessive allocation, when the
computation should have just blocked on the black hole.
Reported by Christian Höner zu Siederdissen <choener@tbi.univie.ac.at>
on glasgow-haskell-users.
Marco Túlio Gontijo e Silva [Sun, 18 Jul 2010 22:55:26 +0000 (22:55 +0000)]
Don't check for swept blocks in -DS.
The checkHeap function assumed the allocated part of the block contained only
alive objects and slops. This was not true for blocks that are collected using
mark sweep. The code in this patch skip the test for this kind of blocks.
Ian Lynagh [Sun, 22 Aug 2010 18:35:42 +0000 (18:35 +0000)]
Fix "darcs get"
Ian Lynagh [Sun, 22 Aug 2010 16:34:19 +0000 (16:34 +0000)]
Add "darcs-all upstreampull"
This pulls from the upstream repos, for those packages which have
upstreams
Ian Lynagh [Sun, 22 Aug 2010 15:55:14 +0000 (15:55 +0000)]
Generate the bit in the user guide where we say what -fglasgow-exts does
Stops the docs going out of sync with the code.
Ian Lynagh [Sun, 22 Aug 2010 15:48:13 +0000 (15:48 +0000)]
Factor out the packages file parsing in darcs-all
Ian Lynagh [Sun, 22 Aug 2010 13:45:30 +0000 (13:45 +0000)]
Document --supported-extensions
Ross Paterson [Sun, 22 Aug 2010 09:00:22 +0000 (09:00 +0000)]
fix extraction of command stack of arguments of arrow "forms" (fixes #4236)
The command stack was being extracted (by unscramble) with the outermost
type first, contrary to the comment on the function.