Hi all,
validate still doesn't pass on OS X with SRC_CC_OPTS += -Werror. Now the
error is:
RtsUtils.c: In function 'time_str':
RtsUtils.c:328:0:
warning: implicit declaration of function 'ctime_r'
The problem is that in /usr/include/time.h, we have
#if !defined(_ANSI_SOURCE) && !define
Judah Jacobson wrote:
IIUC, the NS* stuff is deprecated in OS X 10.4 and dlsym should be used
instead. This is a change from 10.3 and below, where NS* is better.
I'm not that familiar with those functions, but from the following
page it looks like dlsym will work on 10.3 and later:
http://deve
Tue Sep 4 22:22:13 PDT 2007 Roman Leshchinskiy <[EMAIL PROTECTED]>
* Use dlsym on OS X if available
On OS X 10.4 and newer, we have to use dlsym because the old NS* interface has
been deprecated. The patch checks for HAVE_DLFCN_H instead of switching on
the OS version.
There is on
Tue Sep 4 22:39:53 PDT 2007 Roman Leshchinskiy <[EMAIL PROTECTED]>
* Another failing type families test
A ./tests/ghc-regress/indexed-types/should_compile/GADT10.hs
M ./tests/ghc-regress/indexed-types/should_compile/all.T +1
___
Cvs-ghc mail
Tue Sep 4 18:46:27 PDT 2007 Roman Leshchinskiy <[EMAIL PROTECTED]>
* Add dummy stderr file for panic test
A ./tests/ghc-regress/indexed-types/should_fail/Simple14.stderr
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman
Tue Sep 4 05:36:38 PDT 2007 Tom Schrijvers <[EMAIL PROTECTED]>
* tests for type family completion algorithm and higher order types
A ./tests/ghc-regress/indexed-types/should_compile/GivenCheck.hs
A ./tests/ghc-regress/indexed-types/should_compile/GivenCheckDecomp.hs
A ./tests/ghc-r
Tue Sep 4 01:20:41 PDT 2007 Tom Schrijvers <[EMAIL PROTECTED]>
* recursive equational class context
A ./tests/ghc-regress/indexed-types/should_compile/InstEqContext3.hs
M ./tests/ghc-regress/indexed-types/should_compile/all.T +1
___
Cvs-ghc
Tue Sep 4 00:57:39 PDT 2007 Tom Schrijvers <[EMAIL PROTECTED]>
* equational class context
expect_fail is now normal because of bug fix
M ./tests/ghc-regress/indexed-types/should_compile/InstEqContext2.hs -1 +1
M ./tests/ghc-regress/indexed-types/should_compile/all.T -1 +1
__
Tue Sep 4 05:39:45 PDT 2007 Tom Schrijvers <[EMAIL PROTECTED]>
* bug fix in Decomp step of completion algorithm for given equations
M ./compiler/typecheck/TcTyFuns.lhs -4 +5
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mai
Tue Sep 4 01:00:14 PDT 2007 Tom Schrijvers <[EMAIL PROTECTED]>
* fix of wanted equational class context
Previously failed to account for equational
class context for wanted dictionary contraints, e.g. wanted C a
in
class a ~ Int => C a
instance C Int
should giv
On Tue, Sep 04, 2007 at 11:34:28PM +0200, Sven Panne wrote:
>
> Therefore, we *must* have a line
>
>datarootdir := @datarootdir@
Good point, thanks. Now fixed.
Ian
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listi
Tue Sep 4 18:32:39 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Set datarootdir to the value configure gives us (if any) so datadir works
We then set datarootdir to something else later on so that things still
work when configure doesn't set it.
M ./mk/config.mk.in -1 +3
Tue Sep 4 18:32:39 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Set datarootdir to the value configure gives us (if any) so datadir works
We then set datarootdir to something else later on so that things still
work when configure doesn't set it.
M ./mk/config.mk.in -1 +3
Tue Sep 4 18:02:17 PDT 2007 Manuel M T Chakravarty <[EMAIL PROTECTED]>
* FIX: Correct Leave/EnterCriticalSection imports
M ./rts/HeapStackCheck.cmm -2 +2
M ./rts/PrimOps.cmm -2 +2
M ./rts/StgMiscClosures.cmm -2 +2
___
Cvs-ghc mailing lis
Hello again there,
Thank you for all attention & the information, now I'm going to start my own
task, by the way I might ask for more later, so I hope you don't mind with it.
Cheers,
Isma
-
Need a vacation? Get great deals to amazing places on Yahoo!
On 9/4/07, ISMA ISWANDY <[EMAIL PROTECTED]> wrote:
> Hi, I'm new guy here,
>
> Is there sombody could help me, I want to port ghc to BeOS plattform, is
> there any guidelines for it.
> I'm Indonesian so pardon my English.
>
Hi, Isma--
There are instructions on the GHC wiki for porting GHC
Hi, I'm new guy here,
Is there sombody could help me, I want to port ghc to BeOS plattform, is
there any guidelines for it.
I'm Indonesian so pardon my English.
Thanks,
Isma
-
Park yourself in front of a world of choices in alternative vehicle
Just a note, because it is too late for me to fix and test this: The logic in
mk/config.mk.in regarding datarootdir is wrong and leads to incorrect
directories. Autoconf 2.60 introduced datarootdir, and defines the following
defaults:
datadir = ${datarootdir}
infodir = ${datarootdir}/inf
Build description = HEAD on i386-unknown-linux
(cam-02-unx.europe.corp.microsoft.com)
Build location= /playpen/ghc/nightly/HEAD-cam-02-unx
Build config file = /home/simonmar/nightly/site/msrc/conf-HEAD-cam-02-unx
Nightly build started on cam-02-unx at Tue Sep 4 19:30:00 BST 2007.
checkin
Tue Sep 4 10:28:55 PDT 2007 [EMAIL PROTECTED]
* Don't hardwire the build path into the Haddock docs
Formerly, the ghc-pkg was called to get the HTML dirs for other packages, but
of course doing this at *build* time is totally wrong. Now we use a relative
path, just like before. This is pr
Tue Sep 4 07:28:53 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* put the @N suffix on stdcall foreign calls in .cmm code
This applies to EnterCriticalSection and LeaveCriticalSection in the RTS
M ./compiler/cmm/CLabel.hs +7
M ./compiler/cmm/CmmParse.y -5 +19
M ./rts/HeapStackCheck.
Ian Lynagh wrote:
On Tue, Sep 04, 2007 at 03:18:45PM +0100, Ben Lippmeier wrote:
I just un-pulled the last 3 patches I submitted because one of them had
conflicts.
Hmm, you shouldn't be able push patches that have unresolved conflicts.
You can push a conflict+resolution in a single bundle,
Tue Sep 4 07:10:28 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Add a -Warn flag
M ./compiler/main/DynFlags.hs -6 +7
M ./docs/users_guide/flags.xml -1 +7
M ./docs/users_guide/using.xml +10
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http:/
Tue Sep 4 07:03:24 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Always turn on -Wall -Werror when compiling the compiler, even for stage 1
M ./compiler/Makefile -2
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo
Tue Sep 4 07:01:15 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Fix CodingStyle#Warnings URLs
M ./compiler/basicTypes/BasicTypes.lhs -1 +1
M ./compiler/basicTypes/DataCon.lhs -1 +1
M ./compiler/basicTypes/Demand.lhs -1 +1
M ./compiler/basicTypes/Id.lhs -1 +1
M ./compiler/basic
Tue Sep 4 03:06:23 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* OPTIONS_GHC overrides the command-line, not the other way around
M ./docs/users_guide/using.xml -4 +3
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/list
Tue Sep 4 03:05:26 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* fix cut-and-pasto
M ./docs/users_guide/runghc.xml -1 +1
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo/cvs-ghc
Tue Sep 4 00:25:42 PDT 2007 Manuel M T Chakravarty <[EMAIL PROTECTED]>
* FIX #1651: unBox types when deferring unification
- This fixes the first part of #1651; ie, the panic in ghci.
M ./compiler/typecheck/TcUnify.lhs -3 +9
___
Cvs-ghc mailin
Mon Sep 3 00:45:28 PDT 2007 Manuel M T Chakravarty <[EMAIL PROTECTED]>
* Better error message for unsolvable equalities
M ./compiler/typecheck/TcTyFuns.lhs -4 +5
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinf
Hi Ben,
On Tue, Sep 04, 2007 at 03:18:45PM +0100, Ben Lippmeier wrote:
>
> I just un-pulled the last 3 patches I submitted because one of them had
> conflicts.
Hmm, you shouldn't be able push patches that have unresolved conflicts.
> Apparently we're keeping patches with conflicts out of the
Tue Sep 4 06:40:26 PDT 2007 Ian Lynagh <[EMAIL PROTECTED]>
* Add a test num014 for #1658: CSE [of Doubles] changes semantics
M ./tests/ghc-regress/numeric/should_run/all.T +1
A ./tests/ghc-regress/numeric/should_run/num014.hs
A ./tests/ghc-regress/numeric/should_run/num014.stdout
I'd like to volunteer my server, uml (user mode linux) virtualized running
ubuntu 6.06.
One thing that I suspect might be a little unusual is that I don't have
much disk space - 4 gb, which I think I can expand up to 8. If that's not
enough, then I guess too bad. But if it's doable, I'd like t
On Mon, Sep 03, 2007 at 05:58:49PM +0100, Duncan Coutts wrote:
> http://openafp.org/dist/rts.c99.diff
>
> currently rts/ requires C99, unlike the rest of the tree
> so GCC 2.* barfs on it
> that patch simply moves decls upfront.
> thanks in advance :)
Thanks for the patch! I've applied it to
Hi All,
I just un-pulled the last 3 patches I submitted because one of them had
conflicts.
They had timestamps of
Tue Sep 4 04:30:47
Mon Sep 3 09:34:04
Mon Sep 3 06:46:25
Hopefully no-one would have got these, but if you have you should unpull.
Apparently we're keeping patches
On Tue, Sep 04, 2007 at 08:39:45AM +0100, Simon Peyton-Jones wrote:
> | This also changes the URL referenced for the -w options from
> | WorkingConventions#Warnings to CodingStyle#Warnings for the compiler
>
> Alas not. You put the URL as
> http://hackage.haskell.org/trac/ghc/wiki/Cod
Tue Sep 4 04:30:47 PDT 2007 [EMAIL PROTECTED]
* warning police
M ./compiler/nativeGen/GraphBase.hs -6 +1
M ./compiler/nativeGen/GraphColor.hs -10 +4
M ./compiler/nativeGen/GraphOps.hs -10 +3
M ./compiler/nativeGen/GraphPpr.hs -11 +24
M ./compiler/nativeGen/RegAllocColor.hs
Mon Sep 3 09:34:04 PDT 2007 [EMAIL PROTECTED]
* Do conservative coalescing in register allocator
Avoid coalescing nodes in the register conflict graph if the
new node will not be trivially colorable. Also remove the
front end aggressive coalescing pass.
For typical Haskell code
Mon Sep 3 06:46:25 PDT 2007 [EMAIL PROTECTED]
* warning police
M! ./compiler/nativeGen/GraphBase.hs -2 +2
M! ./compiler/nativeGen/GraphColor.hs -2 +2
M! ./compiler/nativeGen/GraphOps.hs -2 +2
M! ./compiler/nativeGen/GraphPpr.hs -2 +2
M! ./compiler/nativeGen/RegAllocColor.hs
Tue Sep 4 03:06:23 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* OPTIONS_GHC overrides the command-line, not the other way around
M ./docs/users_guide/using.xml -4 +3
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/list
Tue Sep 4 03:05:26 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* fix cut-and-pasto
M ./docs/users_guide/runghc.xml -1 +1
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo/cvs-ghc
Tue Sep 4 00:25:42 PDT 2007 Manuel M T Chakravarty <[EMAIL PROTECTED]>
* FIX #1651: unBox types when deferring unification
- This fixes the first part of #1651; ie, the panic in ghci.
M ./compiler/typecheck/TcUnify.lhs -3 +9
___
Cvs-ghc mailin
Manuel M T Chakravarty wrote:
The new test ffi019 misses the .hs file (ie, the recent patch added
.stdout and the addition of the test to all.T), but the darcs patch did
not include the .hs file.
Fixed now, sorry about that.
Simon
___
Cvs-ghc maili
Tue Sep 4 01:18:14 PDT 2007 Simon Marlow <[EMAIL PROTECTED]>
* add missing file
A ./tests/ghc-regress/ccall/should_run/ffi019.hs
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo/cvs-ghc
| > Any reason why /= 1? Why not always?
|
| To avoid failures caused by erroneous warnings in old GHCs. Although
| given most of the time validate will be using the most recent released
| GHC, maybe turning warnings on for stage 1 too would be worthwhile so
| that if it's going to fall over it do
| This also changes the URL referenced for the -w options from
| WorkingConventions#Warnings to CodingStyle#Warnings for the compiler
Alas not. You put the URL as
http://hackage.haskell.org/trac/ghc/wiki/CodingStyle#Warnings
but actually it's
http://hackage.haskell.org/trac/gh
Build results:
x86-64 Linux head: fail (failed stage1) lost
x86 Windows head fast: pass pass pass pass pass pass
gbesh Intel x86_64 Linux head: pass
kahl G5 Gentoo Linux head: pass
tnaur x86 Linux head: lost fail (failed stage1)
x86-64 Linux head unreg: fail (
46 matches
Mail list logo
