;
>were.
>
> ** **
>
> * We give the ClassOp 'op2' a BuiltinRule that extracts the right piece***
> *
>
>iff its argument satisfies exprIsConApp_maybe. This is done in
>
>MkId mkDictSelId
>
> ** **
>
> * We make 'df' C
Hi,
I was investigating how GHC optimises dictionaries. So, I'm using the
following variant of the Eq type class:
class XEq a where
xeq, xne :: a -> a -> Bool
xne a b = not (a `xeq` b) }
instance XEq Bool where
xeq True True = True
xeq False False = True
On 14 October 2012 15:28, Gábor Lehel wrote:
> The total change would be to replace
>
> Prelude.foldl :: (a -> b -> a) -> a -> [b] -> a
> Prelude.scanl :: (a -> b -> a) -> a -> [b] -> [a]
> Data.List.foldl' :: (a -> b -> a) -> a -> [b] -> a
> Data.Foldable.foldl :: (a -> b -> a) ->
Wouldn't that make haskell-src-exts a boot package? I think that would
mean that users could upgrade haskell-src-exts independently, but won't be
able to use the upgraded version and TH in the same program. Or am I
missing something?
On 15 September 2012 09:32, Simon Peyton-Jones wrote:
> Nikl
The following code in from cmm/CmmMachOps.hs (lines 255-282) looks very wrong:
==
-- Inverting conditions
-- Sometimes it's useful to be able to invert the sense of a
-- condition. Not all conditional tests are invertible: in
-- particular, flo
Are these flags documented somewhere else? How would a future
developer find out that these flags exist?
On 12 April 2012 21:54, David Terei wrote:
> Repository : ssh://darcs.haskell.org//srv/darcs/ghc
>
> On branch : master
>
> http://hackage.haskell.org/trac/ghc/changeset/f1a633a0b0d2bfdf527e0
I'm pretty sure Clang doesn't come with 10.5. I don't remember having
to do anything special with 10.6, but I don't know whether 10.6's
version is compatible with the latest version. How can I find out?
As for my personal stance on upgrading: My 10.5 laptop is a 1st-gen
MacBook which are 32 bit
I while ago I saw a commit to Safe Haskell changing modules from
default Unsafe to default Safe.
This seems wrong to me (and Duncan agreed on IRC) -- in security the
usual advice is to use white listing instead of black listing. The
reasoning is that if you forget to white list something safe, it
This might help: http://progit.org/book/ch3-2.html#basic_merge_conflicts
In GHC you have to figure out which subdirectory the conflict happened
in and then fix it in that repository.
If your error is happening in the haddock subdirectory, then you
probably need to delete that directory and run ./
I have a patch for making mappend = (<>) that I need to port to the
new Git tree. The main reason it hasn't gone in yet is because I
don't know how many client libraries it would break. It was actually
agreed on before 7.0.0 (perhaps even 6.12) but somehow never made it
in.
2011/5/6 Simon Peyton
2011/4/29 José Pedro Magalhães :
> commit 1cf00bfef1c35b89c21d1eaa9f6be7354a40f016
> Author: Jose Pedro Magalhaes
> Date: Fri Apr 29 16:02:45 2011 +0200
>
> First go at making Representable0 just a standard derivable class.
>
> Not yet complete; attaching deriving statements to datatypes
2011/4/13 José Pedro Magalhães
> Hi Simon,
>
> 2011/4/12 Simon Peyton-Jones
>
>> I believe I’ve made a new branch
>>
>> ghc-generics
>>
>> in the Git repos for
>>
>> ghc
>>
>> ghc-prim
>>
>> base
>>
>> that reflects the current sta
On 31 January 2011 09:14, Simon Marlow wrote:
> On 30/01/2011 00:42, Daniel Peebles wrote:
>>
>> In a slightly related proposal on the same topic, what do people think of:
>>
>> newArrayWith# :: (Int# -> a) -> Int# -> State# s -> (# State# s,
>> MutableArray# s a #)
>>
>> that would do the obvious
On 13 January 2011 08:54, Roman Leshchinskiy wrote:
> On 12 Jan 2011, at 23:31, "Edward Z. Yang" wrote:
>
>> Excerpts from Roman Leshchinskiy's message of Wed Jan 12 18:20:25 -0500 2011:
>>> How would we get the current functionality of darcs-all pull? Is it even
>>> possible?
>>
>> Here is the
On 11 January 2011 19:07, Roman Leshchinskiy wrote:
> On 11/01/2011, at 16:14, Tony Finch wrote:
>
>> On Mon, 10 Jan 2011, Roman Leshchinskiy wrote:
>>>
>>> It also seems to make finding buggy patches rather hard.
>>
>> Have a look at `git bisect`.
>
> I'm aware of git bisect. It doesn't do what I
I just want to point out that since the last discussion we collected
some migration advice at
http://hackage.haskell.org/trac/ghc/wiki/GitForDarcsUsers
Some of it may be untested (or wrong), but it should be a good starting point.
On 10 January 2011 22:15, Neil Mitchell wrote:
>> As another non-
I'd be for a move, but haven't contributed much lately. I use Git for
all my personal projects, so I consider Git to be useful. I
personally find sending patches via Git to be harder than with Darcs,
but if we use Github the pull-request-based model should work well.
I used Git on Windows two ye
Would it be worthwhile to try and turn off TNTC? I don't know which
section info tables live in without TNTC, but if it were to move info
tables into a separate section this might help.
On 16 September 2010 04:16, Ben Lippmeier wrote:
>
>>> I've been poking through the source of MacOS ld, and th
Just a few bits, Simon will be able to help you with the rest.
On 27 May 2010 00:08, Marco Túlio Gontijo e Silva wrote:
> Hi Simon.
>
> Excerpts from Simon Marlow's message of Ter Mai 25 06:31:50 -0300 2010:
>> On 24/05/2010 23:02, Marco Túlio Gontijo e Silva wrote:
> (...)
>> >> From includes/rt
Hm, I thought that flag was there in order to make SL-built binaries
work with Leopard. This caused problems with the Haskell Platform,
whose GHC version was built with SL. I don't think it would be a good
idea to break Leopard support just yet.
On 20 May 2010 17:57, Ian Lynagh wrote:
> On Fr
Ok, I think I should clarify.
I believe that the framed view with a long list of modules on the left
and the haddocks on the right is still useful. What I don't mind
getting rid off is the third frame which shows the contents of the
mini_* files. I would have preferred to have something similar
I think it will no longer be needed once Haddock outputs table-less
layout code. Frames caused problems with the back-button, so they
weren't really an improvement. A simple CSS float:right + smaller
font on the div containing the index would be a lot better.
I think it would be best to keep the
The only part of ghc-syb that depends on syb is the GHC.SYB.Utils module. I
moved that into a separate package. ghc-syb-0.2 now only depends on base and
the ghc package. The new package is ghc-syb-utils-0.2.
HTH
On 28 Mar 2010, at 16:10, David Waern wrote:
> 2010/3/22 Thomas Schill
I wonder if the ghc-syb package really needs to depend on syb. Just
data may be enough if we split it up. How would you make use of syb
without depending on the syb package?
On 22 March 2010 11:08, David Waern wrote:
> 2010/3/22 Simon Peyton-Jones :
>> | We've discussed support for generic prog
word in such matters. :-)
>
> Explain to me how parsers are not relevant to a compiler? I have had it up
> to here with this nonsense. They seem to fail to understand that I'm making
> a charitable contribution.
>
> ------
> From: &
You/we are talking about parser performance and parsing libraries in
general, not about GHC development. I agree that this conversation
should continue on haskell-cafe.
On 11 March 2010 16:22, John D. Earle wrote:
> How is it off topic? Am I not following the instructions that were given on
> th
No, it's not lazy vs. strict. Parsec's "try" switches the default
LL(1) parser to an LL(N) parser. Happy is always LALR(1), which is
somewhere in the middle. ReadP is always LL(N) but is efficient by
using a breadth-first search instead of depth-first search. The
efficiency of ReadP depends on
Create a blog, put it up there, then post a link to the blog on the
mailing list.
I don't think the GHC development list is best for this kind of
discussion, though. I think haskell-c...@haskell.org is much better
for discussing your ideas. The GHC team already expressed their
opinion: They wil
On 6 March 2010 07:35, Neil Mitchell wrote:
>
>> http://chadaustin.me/2010/03/your-version-control-and-build-systems-dont-scale-introducing-ibb/
>
> This link is crazy. He's abusing big O notation, complaining about
> constant factors (Python starting up) then suggesting he needs to
> reduce the a
You may want to consider something like this, as well.
http://chadaustin.me/2010/03/your-version-control-and-build-systems-dont-scale-introducing-ibb/
On 4 March 2010 21:06, John D. Earle wrote:
> Hello,
>
> I wish to discuss my feature request
> http://hackage.haskell.org/trac/ghc/ticket/3912 en
On 27 Jan 2010, at 18:17, Neil Mitchell wrote:
> Hi
>>> | These have always distressed me. Would it be feasible to refactor so
>>> | that this isn't necessary, e.g.
>>> |
>>> | typecheck :: HsSyn Name () -> HsSyn Name PostTcInfo
>
>> The problem is that these annotations tend to be construct
On 27 Jan 2010, at 15:12, Simon Peyton-Jones wrote:
>
> | I was wondering whether there would be any concerns against redefining the
> | above type synonyms into newtypes? Either each one of the above gets its
> own
> | newtype wrapper or new phase-describing wrappers:
> |
> | newtype PostTc
Hi,
[This question is probably mostly for Simon PJ.]
The Haskell AST used by GHC unfortunately has a few fields that are only
defined after certain phases have been run. For example:
PostTcType, PostTcExpr -- only defined after type checking
SyntaxExpr, Fixity -- only defined after ren
2010/1/13 Duncan Coutts
> I'm not sure I can. Apparently I just don't get what other people think
> is wrong with rpath.
Quoting from: http://en.wikipedia.org/wiki/Rpath_(linking)
"he primary disadvantage of using RPATH is that it overrides the
LD_LIBRARY_PATH settings which makes things like ru
I think a bogus hunk snuck into this patch. See the TcBinds stuff: Here's
a quick link to the diff
http://github.com/ghc-hq/ghc/commit/f3aed3b7323bae808bdb406edd06844d57e7da32
.
2009/9/17
> Thu Sep 17 06:27:06 PDT 2009 Manuel M T Chakravarty >
> * Use CONF_CC_OPTS
> Ignore-this: 3dd1a67c04
Tue Sep 15 08:03:53 PDT 2009 Thomas Schilling
* Put context information for warnings in errMsgExtraInfo.
For type checker warnings, the context information ("In the expression
...") was simply appended to the main message while for proper errors
they live in errMsgExtraI
Mon Aug 24 11:22:52 PDT 2009 Thomas Schilling
* Make FastString thread-safe.
This is needed both for per-session parallelism and for allowing
multiple concurrent sessions in the same process. With the help of
atomicModifyIORef and unsafePerformIO it is also quite fast--an MVar
Sat Aug 22 16:09:37 PDT 2009 Thomas Schilling
* Make consIORef atomic.
Ignore-this: c4570e17a21be724b45c574bd25d0330
M ./compiler/utils/Util.lhs -4 +2
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20090822230937-7c5c6-d2f0fad973180ed02faade0eef714998ac9fe616.gz
Fri Aug 21 16:53:55 PDT 2009 Thomas Schilling
* Make updates to external package state idempotent.
Without this, concurrent updates to the EPS could introduce
overlapping instances (even though they came from the same module).
M ./compiler/iface/LoadIface.lhs +1
View patch online
Tue Aug 18 14:32:43 PDT 2009 Thomas Schilling
* Remove the lock around NameCache for readBinIface.
Turns out using atomic update instead of a full-blown lock was easier
than I thought. It should also be safe in the case where we
concurrently read the same interface file. Whichever
Mon Aug 17 07:23:52 PDT 2009 Thomas Schilling
* Make the dynamic linker thread-safe.
The current implementation is rather pessimistic. The persistent
linker state is now an MVar and all exported Linker functions are
wrapped in modifyMVar calls. This is serves as a big lock around
Sun Aug 16 17:48:19 PDT 2009 Thomas Schilling
* Make access to NameCache atomic. Sometimes needs a lock.
'readBinIface' updates the name cache in a way that is hard to use
with atomicModifyIORef, so this patch introduces a lock for this case.
All other updates use atomicM
Sun Aug 16 16:19:20 PDT 2009 Thomas Schilling
* Make updates to the Finder caches atomic. Well, almost.
Flushing and uncaching a single module is not completely atomic since
both caches a cleared separately. However, flushing is only done when
changing the working directory which
Sun Aug 16 16:13:16 PDT 2009 Thomas Schilling
* Make updates to the external package state atomic.
Ignore-this: a700ff0c6725264941d6a5458ce5dade
M ./compiler/typecheck/TcRnMonad.lhs -18 +17
M ./compiler/utils/IOEnv.hs -2 +15
View patch online:
http://darcs.haskell.org/ghc/_darcs
Wed Jul 22 16:13:42 PDT 2009 Thomas Schilling
* Minor documentation fixes.
M ./compiler/basicTypes/OccName.lhs -1 +1
M ./compiler/cmm/OptimizationFuel.hs +6
M ./compiler/deSugar/Desugar.lhs +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20090722231342-7c5c6
I think the inconsistencies are mostly because SrcSpan was added
later. Patches to improve the problematic cases would (most likely)
be welcome.
2009/7/20 Isaac Dupree :
> is a confusing mess!
> The most important module is named SrcLoc, which is acceptable.
> It contains three types,
> SrcLoc,
>
.
It's a feasible GSoC project (that's why I suggested it!). Thomas
Schilling looked into it a little when he was here as an intern, he
might be able to tell you more. I recall we discussed the details a
bit, but can't remember exactly what we concluded. You should read
not just
I think I'm not subscribed to the haddock mailing list, so I'm posting
this here.
The attached two patches fix the Template Haskell bugs with Haddock
and a recent stable release. I already pushed the patch to GHC
stable; these two patches enable the code path in Haddock to make use
of th
Mon Feb 23 08:56:30 PST 2009 Thomas Schilling
* Backport: Let 'loadModule' generate proper code depending on the
'hscTarget'.
With this change it should be possible to perform something similar to
'load' by traversing the module graph in dependen
Wed Nov 26 10:34:02 PST 2008 Thomas Schilling
* Force recompilation of BCOs when they were compiled in HscNothing mode.
Previously, loading a set of modules in HscNothing mode and then
switching to HscInterpreted could lead to crashes since modules
compiled with HscNothing were
No, both stable and head are still darcs1 repos.
The switch to Git for development is scheduled after the 6.10.2
release, since the 6.10 stable branch will have to remain in darcs.
The next stable release process will start some time this month.
2009/1/2 Tim Chevalier :
> Hello all,
>
> I haven't
Fri Dec 19 06:42:23 PST 2008 Thomas Schilling
* Add test for #2739 (tests that it works with HscInterpreted).
A ./tests/ghc-regress/ghc-api/2739/
View patch online:
http://darcs.haskell.org/testsuite/_darcs/patches/20081219144223-7c5c6-7ec90675201285fbe030f215ef73771df455bda3.gz
Thu Nov 27 02:59:29 PST 2008 Thomas Schilling
* Add first test for GHC API features.
A ./tests/ghc-regress/ghc-api/
A ./tests/ghc-regress/ghc-api/Makefile
A ./tests/ghc-regress/ghc-api/apirecomp001/
A ./tests/ghc-regress/ghc-api/apirecomp001/A.hs
A ./tests/ghc-regress/ghc
Thu Dec 11 07:46:57 PST 2008 Thomas Schilling
* Comments only. Haddockify parts of TcRnTypes.
M ./compiler/typecheck/TcRnTypes.lhs -108 +135
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081211154657-7c5c6-8c3122cbc97a38548b28a1efb2a37ba2cda94650.gz
Thu Dec 11 07:31:04 PST 2008 Thomas Schilling
* Comments only. Fix typo.
M ./compiler/ghci/ByteCodeGen.lhs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081211153104-7c5c6-60277c35ad4dedfe6229785c330dd12db0ef68e9.gz
Thu Dec 11 07:30:05 PST 2008 Thomas Schilling
* Include PprTyThings in tags file.
M ./compiler/Makefile -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081211153005-7c5c6-7e6afee922417e57091cae9098cc741f03278621.gz
My validate always has process008 failing (missing printf). The rest is fine.
2008/12/10 Simon Marlow <[EMAIL PROTECTED]>:
> Simon Marlow wrote:
>>
>> Simon Peyton-Jones wrote:
>>>
>>> | Hooray for the testsuite and validate: out of 2000+ tests, there was
>>> | exactly 1 (ONE) that demonstrated t
t;
> endif
>
> That doesn't seem to build stage 2 with profiling. So what are you using?
>
> S
>
> | -Original Message-
> | From: Thomas Schilling [mailto:[EMAIL PROTECTED]
> | Sent: 08 December 2008 19:58
> | To: Simon Peyton-Jones
> | Cc: cvs-ghc@has
rProfiling, but I could only guess how this could look
like or whether this should perhaps never occur and be converted to a
closure or an StgLet somewhere.
2008/12/8 Thomas Schilling <[EMAIL PROTECTED]>:
> That seems to have fixed one problem, but I now get:
>
> < (2 samples),
y adding more
> eta_expand n us (Lam v body) ty
>
>
>
> If that fixes it, commit the patch and I'll add comments.
>
> Note for Simon M: this is relying on the fact that
>scc "f" (\x.e) == \x. scc "f" e
> which is, I hope, true!
>
&g
With HEAD from today, I get the following GHC panic when building
stage2 in profiling way (BuildFlavour = devel2):
<>
/Users/nominolo/code/ghc/ng-api/ghc/stage1-inplace/ghc
-DGHCI_TABLES_NEXT_TO_CODE -DSTAGE=2 -package-name ghc-6.11.20081205
-hide-all-packages -no-user-package-conf -i -idist-stage
Tue Dec 2 07:48:00 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Document 'loadModule'.
M ./compiler/main/GHC.hs -1 +7
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081202154800-7c5c6-ae2c12db76e51aab8740e46abcf
Tue Dec 2 07:23:58 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Add 'needsTemplateHaskell' utility function and document why one might
want to use it.
M ./compiler/main/DynFlags.hs -11 +19
M ./compiler/main/GHC.hs +10
View patch online:
http://darcs.haskel
Tue Dec 2 07:01:58 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Documentation only.
M ./compiler/main/GHC.hs -7 +36
M ./compiler/main/HscTypes.lhs -1 +2
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081202150158-7c5c6-3aed3093943074cadd7f886c93121bb4281df
Tue Dec 2 06:44:51 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Export 'succeeded' and 'failed' helper functions.
M ./compiler/main/GHC.hs -1 +2
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081202144451-7c5c6-5e1b05d269bff225f
rg/haddock) though. So please check that repo before
> pushing patches so we can avoid conflicts. Yes, I should push to the
> GHC repo more often, but I don't currently have a working build.
>
> David
>
> 2008/11/29 Thomas Schilling <[EMAIL PROTECTED]>:
>> I changed
I changed the behaviour of 'loadModule', which exposes a problem in
Haddock's implementation. I pushed the respective fix to Haddock, so
make sure you do a "darcs-all pull" instead of a simple pull next time
you update GHC. If you don't, Haddock *will overwrite your compiled
files* which can lead
Fri Nov 28 08:44:12 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Let 'loadModule' generate proper code depending on the 'hscTarget'.
With this change it should be possible to perform something similar to
'load' by traversing the module graph in de
Fri Nov 28 08:37:46 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Expose a separate 'hscBackend' phase for 'HsCompiler' and change
parameter to 'InteractiveStatus' to a 'Maybe'.
M ./compiler/main/DriverPipeline.hs -2 +5
M ./compiler/mai
Fri Nov 28 07:07:27 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Whoops, *don't* reset the complete session in 'withLocalCallbacks'.
M ./compiler/main/HscTypes.lhs -1 +2
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches
Fri Nov 28 04:19:47 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Use a record instead of a typeclass for 'HsCompiler'. This is mostly
equivalent to a typeclass implementation that uses a functional
dependency from the target mode to the result type.
M ./compiler/
Fri Nov 28 02:36:28 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Use a per-session data structure for callbacks. Make 'WarnErrLogger'
part of it.
Part of the GHC API essentially represents a compilation framework.
The difference of a *framework* as opposed to a *
Thu Nov 27 05:57:25 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Improve error message for #2739 (but no fix).
This patch changes 'loadModule' to define a fake linkable. The
previous implementation of providing no linkable at all violated a
pre-condition in the
Thu Nov 27 02:25:34 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Collect instead of print warnings in 'warnUnnecessarySourceImports'.
M ./compiler/main/GHC.hs -4 +4
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/2008
Wed Nov 26 10:34:02 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Force recompilation of BCOs when they were compiled in HscNothing mode.
Previously, loading a set of modules in HscNothing mode and then
switching to HscInterpreted could lead to crashes since modules
compile
Wed Nov 26 05:43:44 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Documentation only: Add module description for HscMain.
M ./compiler/main/HscMain.lhs -3 +6
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081126134344-7c5c6-f3a94ea8ee75f6281d574136fcabdd41cd55e
Wed Nov 26 04:28:01 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Include GHCi files in ctags/etags.
M ./compiler/Makefile -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081126122801-7c5c6-1010ed4b16db8aae8471e7d3b020e0729f694
Tue Nov 25 07:32:01 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Major clean-up of HscMain.
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
Sat Nov 22 07:41:51 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Return errors instead of dying in myParseModule.
M ./compiler/main/HscMain.lhs -37 +34
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081122154151-7c5c6-edb4cbe33929b65691d90f752bc05f8c9ccab
Sat Nov 22 06:30:18 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Comments/Haddockification only.
M ./compiler/hsSyn/HsDecls.lhs -14 +25
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081122143018-7c5c6-9f3b8eb8f9a0eea8da804970f4d6c0bb23673
Sat Nov 22 06:26:41 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Report source span instead of just source location for unused names.
M ./compiler/rename/RnEnv.lhs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081122142641
Sat Nov 22 05:06:58 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Change 'handleFlagWarnings' to throw exceptions instead of dying.
It now uses the standard warning log and error reporting mechanism.
M ./compiler/ghci/InteractiveUI.hs -3 +3
M ./compiler/main/Driv
Fri Nov 21 06:53:07 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Document exported functions in main/HeaderInfo.
M ./compiler/main/HeaderInfo.hs -5 +16
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081121145307-7c5c6-621b03a81dc0365619a3a4e00f9711e12fa42
Fri Nov 21 06:41:55 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Remove warning supression klugde in main/HeaderInfo
M ./compiler/main/HeaderInfo.hs -8 +4
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081121144155-7c5c6-e0daf6d967c538e3cd101ca5135213d538e30
Fri Nov 21 08:02:35 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Follow changes in pragma error reporting.
Parse errors in pragmas are now normal error messages and printed as
such. The full span info is still there, but not by default we only
print the start of the regio
Fri Nov 21 06:13:39 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Throw SourceErrors instead of ProgramErrors in main/HeaderInfo.
Parse errors during dependency analysis or options parsing really
shouldn't kill GHC; this is particularly annoying for GHC API clients.
M
Ian, this should also go into the 6.10 branch.
2008/11/13 Thomas Schilling <[EMAIL PROTECTED]>:
> Thu Nov 13 08:26:53 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
> * Fix a bug in the recompilation checking logic.
>
> Previously, using target HscNothing
Thu Nov 13 08:26:53 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Fix a bug in the recompilation checking logic.
Previously, using target HscNothing resulted in unnessesary
recompilation because 'upsweep_mod' treated HscInterface specially.
This patch changes relaxes
Mon Nov 10 07:38:19 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Fix documentation (to say the opposite).
M ./compiler/basicTypes/SrcLoc.lhs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081110153819-7c5c6-414117511a22400f9d733c6cc58a27eb00e2f
Mon Nov 10 09:56:14 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Documentation only. Clarify that 'load*' may indeed throw SourceErrors.
I don't think errors during dependency analysis should be passed to
the logger.
M ./compiler/main/GHC.hs -1 +9
Vi
Mon Nov 10 07:36:21 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Fix line numbers in TAGS files.
M ./utils/ghctags/GhcTags.hs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081110153621-7c5c6-20f268e9c543af0423b2016bb3e31da85fc86
Mon Nov 10 07:34:56 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Documentation only.
M ./compiler/utils/Maybes.lhs -13 +4
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081110153456-7c5c6-e29b87cb71c07b8a8fd4dbf2332d16875fb92
Mon Nov 10 06:35:10 PST 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Add 'packageDbModules' function to GHC API.
This function returns a list of all modules available through the
package DB.
MERGE TO 6.10
M ./compiler/main/GHC.hs +21
View patch online:
http://d
Fri Oct 31 09:20:36 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Remove dead code.
M ./compiler/typecheck/TcSplice.lhs -1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081031162036-7c5c6-0176403f131f4a211fa0e548de5a4d76a2a38
2008/10/31 Thomas Schilling <[EMAIL PROTECTED]>:
> Fri Oct 31 04:31:25 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
> * Quickfix for warning.
>
>M ./compiler/typecheck/TcSplice.lhs -1 +1
>
> View patch online:
> http://darcs.haskell.
Fri Oct 31 04:31:25 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Quickfix for warning.
M ./compiler/typecheck/TcSplice.lhs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081031113125-7c5c6-a7179657c6ef36b5f6afb5710224365826c4a
Tue Oct 28 11:23:10 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Export typeclasses for accessing compiler results.
MERGE TO 6.10.
M ./compiler/main/GHC.hs +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081028182310
Tue Oct 28 11:22:02 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Minor refactoring.
M ./compiler/main/GHC.hs -3 +1
M ./compiler/main/HscTypes.lhs -1 +1
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081028182202-7c5c6-07ca58c6bdde5bea672050be19c7090b305b1
Tue Oct 28 11:05:38 PDT 2008 Thomas Schilling <[EMAIL PROTECTED]>
* Include record fields in tags.
M ./utils/ghctags/GhcTags.hs -10 +5
View patch online:
http://darcs.haskell.org/ghc/_darcs/patches/20081028180538-7c5c6-6163009315c61d7fb86a4b817f819cbfb1315
FWIW, I had some validate failures under windows dependening on which
shell I used. I have to check again which one it was, though.
2008/10/22 Mitchell, Neil <[EMAIL PROTECTED]>:
> Hi
>
>
>> > Could it be that buffering has a different default on Windows?
>>
>> 2228 is now an expected failure on
1 - 100 of 174 matches
Mail list logo
