Hi!
Please forgive me that i'm a lazy slacker, even too lazy to send a
proper git patch, but anyway here's a small simplifcation diff for the
SIGTSTP handler of GHC's rts (explanations below):
--- rts/posix/Signals.c.origMon Jun 13 19:10:06 2011
+++ rts/posix/Signals.c Tue Dec 27 19:58:52 201
While here, make the code a little bit more readable.
---
sync-all | 22 --
1 files changed, 8 insertions(+), 14 deletions(-)
diff --git a/sync-all b/sync-all
index 7ccc71d..3ccbc86 100755
--- a/sync-all
+++ b/sync-all
@@ -310,20 +310,14 @@ sub scmall {
if (-d "
rom 9594339a9286f550addc592083de07e89d14dd62 Mon Sep 17 00:00:00 2001
From: Matthias Kilian
Date: Mon, 18 Apr 2011 23:59:51 +0200
Subject: [PATCH] do not use sed for version date processing but rather cut and
tr
Based on an idea from Karel Gardas, who had troubles with the original
sed version (
On Sun, Apr 17, 2011 at 11:44:33PM +0200, Karel Gardas wrote:
> This patch fixes issue with version date processing on Solaris while using
> Solaris' provided sed. I don't know sed enough to remove GNUism in version
> date processing so I replaced sed usage in this task by few cut calls.
Out of cu
Hi,
because http://www.haskell.org/haddock/ seems to be out-of date,
and http://trac.haskell.org/haddock doesn't accept new tickets (even
with a newly registered account), i'll dump this here:
- cross-package module references don't work. The broken link to
System.IO in the first paragraoh from
se the list was in the cc:
> Cc: cvs-ghc
> Subject: Re: Sending git patches
> From: "Edward Z. Yang"
> To: Matthias Kilian
> Date: Sat, 16 Apr 2011 16:24:34 -0400
> Message-Id: <1302985368-sup-3888@ezyang>
So someone should probably have a look
---
sync-all |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/sync-all b/sync-all
index 06c183a..7ccc71d 100755
--- a/sync-all
+++ b/sync-all
@@ -389,6 +389,10 @@ sub scmall {
scm ($localpath, $scm, "grep", @args)
unless $scm eq "darcs";
---
sync-all |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/sync-all b/sync-all
index 06c183a..7ccc71d 100755
--- a/sync-all
+++ b/sync-all
@@ -389,6 +389,10 @@ sub scmall {
scm ($localpath, $scm, "grep", @args)
unless $scm eq "darcs";
---
sync-all |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/sync-all b/sync-all
index 06c183a..7ccc71d 100755
--- a/sync-all
+++ b/sync-all
@@ -389,6 +389,10 @@ sub scmall {
scm ($localpath, $scm, "grep", @args)
unless $scm eq "darcs";
Hi,
what's the correct way to send patches in these modern days?
I'm using
git format-patch origin"
and then
git send-email --to=cvs-ghc@haskell.org 0001-whatever
Is this correct? And is there some reason why I don't see this mail
sent back to be over the mailing list?
Sorry
---
sync-all |5 +
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/sync-all b/sync-all
index 06c183a..7ccc71d 100755
--- a/sync-all
+++ b/sync-all
@@ -389,6 +389,10 @@ sub scmall {
scm ($localpath, $scm, "grep", @args)
unless $scm eq "darcs";
On Thu, Apr 14, 2011 at 11:57:51AM -0400, Edward Z. Yang wrote:
> The remote repository has been reset to 9eebc6dec9a2271a5179, removing
> the two merge commits that Simon pushed. If you do not have any commits
> more recent then that in your working copy, you don't have to worry about
> this.
A
---
libraries/gen_contents_index |2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/libraries/gen_contents_index b/libraries/gen_contents_index
index fcf30e3..72be850 100644
--- a/libraries/gen_contents_index
+++ b/libraries/gen_contents_index
@@ -7,7 +7,7 @@ HADDOCK_ARGS=
On Fri, Dec 31, 2010 at 10:29:33PM +0100, Matthias Kilian wrote:
> 1 patch for repository http://darcs.haskell.org/ghc:
>
> Fri Dec 31 20:23:43 CET 2010 Matthias Kilian
> * Drop a seven years old workaround for happy
Forgot to mention: tested with ghc-7.0 and ghc-head branche
1 patch for repository http://darcs.haskell.org/ghc:
Fri Dec 31 20:23:43 CET 2010 Matthias Kilian
* Drop a seven years old workaround for happy
New patches:
[Drop a seven years old workaround for happy
Matthias Kilian **20101231192343
Ignore-this: a9348c91292c113bd967464fbe859f1f
] hunk
1 patch for repository http://darcs.volkswurst.de/ghc:
Sun Dec 19 19:02:39 CET 2010 Matthias Kilian
* Drop GhcWithLlvmCodeGen configuration bits
The LLVM code generator is always built unconditionally, so both the
configuration variable in mk/config.mk.in as well as the string in
On Thu, Oct 28, 2010 at 09:51:02AM +0100, Simon Marlow wrote:
> >>custom data stored in ghc's package database. For now this will be
> >>just another field `distPackageId'
> >
> >I think "localInfo" would be a better name. Just a free text field
> >people can put whatever they like in.
First: sorr
On Mon, Jun 28, 2010 at 09:52:22AM +0100, Simon Marlow wrote:
> >After some more poor man's debugging (sprinkling some puts(3) into
> >rts/posix/Signals.c), the problem seems to be that ioManagerDie()
> >assumes that writing that `kill byte' (IO_MANAGER_DIE) wakes the
> >service loop immediately, a
[sorry for constantly replying to myself...]
On Sat, Jun 26, 2010 at 10:42:02AM +0200, Matthias Kilian wrote:
> However, the munmaps happen after what looks like GHC cleanup. For
> example this comes before the munmaps:
[...]
> then, after some more stuff (like getrusage, gettimeof
On Fri, Jun 25, 2010 at 09:16:58PM +0200, Matthias Kilian wrote:
> > So perhaps something called munmap() at some point? Have you tried
> > strace? (or equivalent on OpenBSD, truss?).
>
> Yep (it's ktrace on OpenBSD). Took me some time (this bug tries to
> hide reall
On Tue, Jun 22, 2010 at 09:18:20AM +0100, Simon Marlow wrote:
> So perhaps something called munmap() at some point? Have you tried
> strace? (or equivalent on OpenBSD, truss?).
Yep (it's ktrace on OpenBSD). Took me some time (this bug tries to
hide really hard if you *want* it to happen).
All I
Hi,
Short story: I get impossible segfaults with ghc-6.12.3 on OpenBSD
(unless I completely disable threading support when building ghc).
Long story: in libraries/base/GHC/Conc.lhs, service_loop, the call
to c_select sometimes segfaults because of inaccessible (not mapped)
readfds and writefds a
On Sat, May 29, 2010 at 11:41:36AM +0100, Ian Lynagh wrote:
> Hmm, build 48 used those flags when compiling, but still failed.
> Any idea what's wrong?
>
> http://darcs.haskell.org/ghcBuilder/builders/pgj2/48.html
I'm not a FreeBSD user, but it's strange that the build tries to
pick up the static
[re-sending to the list and Gabor, not only to Ian...]
On Wed, May 19, 2010 at 04:46:24PM +0100, Ian Lynagh wrote:
> On Wed, May 19, 2010 at 05:19:31PM +0200, Matthias Kilian wrote:
> > On Wed, May 19, 2010 at 12:34:15PM +0200, Gabor PALI wrote:
> > > I got a validation error o
On Wed, May 19, 2010 at 12:34:15PM +0200, Gabor PALI wrote:
> I got a validation error on FreeBSD/i386 for Linker.c. I attach a
> tiny patch to fix it, hopefully other systems will like it too :)
This is because dlerror(3) is const char*, right?
Do you just get a warning or any kind of error? I'
On Mon, May 17, 2010 at 08:52:14PM +0100, Ian Lynagh wrote:
> Solaris doesn't have whereis. Does this work for you?
>
> if type make > /dev/null 2> /dev/null
> then
> echo yes
> else
> echo no
> fi
>
> if type gmake > /dev/null 2> /dev/null
> then
> echo yes
> else
> echo no
> fi
On Mon, May 10, 2010 at 09:08:21AM -0700, Simon Peyton Jones wrote:
> Mon May 10 06:56:06 PDT 2010 simo...@microsoft.com
> * Fix Trac #4003: update the knot-tied type environment with boot dfuns
Indeed. Thanks!
Ciao,
Kili
___
Cvs-ghc mailing
On Wed, Apr 28, 2010 at 10:50:39PM +0100, Ian Lynagh wrote:
> > WHat exactly does this fix? I've some similar patch around for some
> > time, but it #ifdef's the whole ELF stuff out (for OpenBSD).
> >
> > Can you show me the warnings you get without it?
>
> http://darcs.haskell.org/ghcBuilder/bui
On Wed, Apr 28, 2010 at 09:16:33AM -0700, Ian Lynagh wrote:
> Wed Apr 28 08:07:00 PDT 2010 Ian Lynagh
> * Fix some cpp warnings when building on FreeBSD; patch from Gabor PALI
>
> M ./rts/Linker.c +8
WHat exactly does this fix? I've some similar patch around for some
time, but it #ifdef's
On Mon, Apr 19, 2010 at 01:40:28PM +0100, Simon Marlow wrote:
> The panic is obviously a bug, but it's not clear whether it is a bug in
> the build system (a missing dependency) or in GHC. Can you give us
> instructions to reproduce?
I just created a ticket (#4012) for a different (and very sca
On Mon, Apr 19, 2010 at 11:29:31AM +0200, Matthias Kilian wrote:
> [Warning: much confusion below ;-)]
Even if I'm often confused: did anyone yet manage to build ghc-head
with ghc-6.12?
___
Cvs-ghc mailing list
Cvs-ghc@haskell.
On Mon, Apr 19, 2010 at 05:19:24PM +0200, Matthias Kilian wrote:
> I don't know if it is of any help, but below are the lat (failing)
> command line with -v, the output, and a ghc-pkg list -v (from the
> ghc-6.12 I'm using for bootstrapping):
And if it's of any furt
On Mon, Apr 19, 2010 at 01:40:28PM +0100, Simon Marlow wrote:
> >This patch appears to uncover some strange problem during bootstrapping
> >a ghc-6.13 stage1 compiler with an existing ghc-6.12 compiler.
> >Initially, it lead to the famous panic involved with outdated
> >interface files when compili
[Warning: much confusion below ;-)]
Hi,
On Fri, Apr 09, 2010 at 04:48:05AM -0700, Ian Lynagh wrote:
> Mon Mar 29 18:10:20 PDT 2010 David Waern
> * Add Data and Typeable instances to HsSyn
>
> The instances (and deriving declarations) have been taken from the ghc-syb
> package.
[...]
T
On Sun, Apr 18, 2010 at 07:29:16PM +0200, Matthias Kilian wrote:
> * Increase the stack size for 3677 a little bit
> With 8k of stack, this test fails for WAY=normal, at least on
> OpenBSD/amd64. Increase it to 10k.
Actually it fails for other WAYs, too with 8k stack. And what&#
On Sat, Apr 17, 2010 at 09:44:13PM +0100, Ian Lynagh wrote:
> > Please note that this patch is super-unimportant and should only
> > be pushed if it has been tested at least by the MacOS X and Solaris
> > people (I don't expect any fallout on Windows, Linux, FreeBSD and
> > NetBSD).
>
> The diff o
On Sat, Apr 17, 2010 at 03:30:14PM +0200, Matthias Kilian wrote:
> Fri Apr 16 15:31:06 CEST 2010 Matthias Kilian
> * Run diff(1) with -a when comparing output.
>
> This stops diff(1) on OpenBSD (and maybe other systems) from treating
> some output files as binary.
>
1 patch for repository http://darcs.volkswurst.de/testsuite:
Fri Apr 16 13:46:27 CEST 2010 Matthias Kilian
* Use ${PYTHON} instead of relying on #!/usr/bin/env python
New patches:
[Use ${PYTHON} instead of relying on #!/usr/bin/env python
Matthias Kilian **20100416114627
Ignore-this
On Wed, Apr 14, 2010 at 09:10:00PM +0100, Ian Lynagh wrote:
> > Perhaps just use make like Matthias suggested?
>
> That feels very klunky to me; a shell script /and/ a Makefile for this
> simple task.
Not a *Makefile*, just some make -f - <<'EOM' ... EOM goo. I actually
had something that *almost
[slightly offtopic now]
On Mon, Apr 12, 2010 at 11:43:23AM +0200, Gabor PALI wrote:
> Ironically, I needed to hack my systems under the build slaves towards
> the GNU direction to see other interesting things like built snapshots
> :) I installed bash, and symlinked to /usr/bin/bash;
To my knowl
On Mon, Apr 12, 2010 at 11:47:26AM +1000, Ben Lippmeier wrote:
> The trouble is that the boot scripts use GNU specific features.
> I was the one that changed it to bash because I was tired of the
> build breaking on Solaris. I think the original problem was that
> the Solaris version of 'test' does
On Tue, Mar 30, 2010 at 07:28:56PM +0200, Matthias Kilian wrote:
> > Thanks for the report; I've made a proper test for the feature, so it
> > should work again now.
>
> I'll give it a try this evening or tomorrow.
Works fine here. Thank
On Tue, Mar 30, 2010 at 05:04:29PM +0100, Ian Lynagh wrote:
> > > * Tweak the Makefile code for making .a libs; fixes trac #3642
> > > The main change is that, rather than using "xargs ar" we now put
> > > all the filenames into a file, and do "ar @file". This means that
> > > ar adds all t
On Tue, Mar 30, 2010 at 03:34:11PM +1100, Roman Leshchinskiy wrote:
> boot-pkgs contains this line:
>
> tar -zxf ../../$tarball
>
> This fails at least on Solaris where we want gtar. We probably
> want to look for gnutar, gtar and then for tar, just like in
> configure.
Or just u
On Wed, Mar 24, 2010 at 12:55:23PM -0700, Ian Lynagh wrote:
> Tue Mar 23 15:13:25 PDT 2010 Ian Lynagh
> * Tweak the Makefile code for making .a libs; fixes trac #3642
> The main change is that, rather than using "xargs ar" we now put
> all the filenames into a file, and do "ar @file". This
On Wed, Mar 17, 2010 at 11:49:12PM +1100, Roman Leshchinskiy wrote:
> DPH is already built by stage 2. You were getting errors from the
> stage 2 compiler which didn't support ghci. Here is the error
> message from your original posting:
Oh, then I got confused by my different builds here.
In thi
On Wed, Mar 17, 2010 at 11:25:30AM +1100, Roman Leshchinskiy wrote:
> > I've a patch which removes DPH from ghc-head, and a merge which
> > removes it from ghc-6.12, and I could successfully build
>
> Unfortunately, we can't remove DPH from the boot packages yet.
> However, as I suggested in my pr
On Mon, Mar 15, 2010 at 09:50:41AM +, Simon Marlow wrote:
> >Does anyone here care about portability? Some time ago (about three
> >years) someone told me that portability is an imporant goal for
> >ghc. Given that you now even need ghci to build ghc, this portability
> >claim is obviously a li
On Tue, Mar 16, 2010 at 01:21:12PM +1100, Roman Leshchinskiy wrote:
> > It happens when you try to build ghc-6.13 (from darcs) using a
> > ghc-6.12 (from darcs) with GHCi disabled.
>
> I suppose we should disable DPH if ghci and/or TH are disabled.
> I'm not sure how to do it properly, though. Sim
On Sun, Mar 14, 2010 at 02:54:02PM -0700, Don Stewart wrote:
> kili:
> > ghc-stage2: panic! (the 'impossible' happened)
> > (GHC version 6.13.20100314 for x86_64-unknown-openbsd):
> > Cant do annotations without GHCi
> > {59:20-31}ghc-6.13.20100314:SpecConstr.NoSpecConstr{d rap}
> >
This can't be true:
"inplace/bin/ghc-stage2" -H32m -O-package-name dph-seq-0.4.0
-hide-all-packages -i -ilibraries/dph/dph-seq/../dph-common
-ilibraries/dph/dph-seq/dist-install/build
-ilibraries/dph/dph-seq/dist-install/build/autogen
-Ilibraries/dph/dph-seq/dist-install/build
-Ilibrari
On Fri, Mar 05, 2010 at 09:26:08PM +, Simon Marlow wrote:
> >Switching GHC to a completely new build system written completely
> >in Haskell would be the most stupid idea ever. (You know why)
>
> You're referring to bootstrapping, I presume?
Yes.
> I did think about mentioning that. Of cour
On Fri, Mar 05, 2010 at 12:17:55PM +, Simon Marlow wrote:
> I suggest that the way to start would be to design and build the
> infrastructure first, and then think about replacing GHC's build system.
> I have to admit, having rewritten GHC's build system less than a year
> ago, I'm not part
Hi,
has the packages/parallel repository been removed from darcs.haskell.org
by intention?
Ciao,
Kili
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
http://www.haskell.org/mailman/listinfo/cvs-ghc
Hi,
I just noticed that ghc-6.10/packages/mtl is missing on the new
darcs.haskell.org server. Since it's still referenced from
ghc-6.12/ghc/packages, could someone please copy it over?
Ciao,
Kili
___
Cvs-ghc mailing list
Cvs-ghc@haskell.org
htt
On Tue, Dec 22, 2009 at 06:26:16AM +, Duncan Coutts wrote:
> > GNUisms, it's not Linux' fault, although I neither like Linux do I
> > know any Linux distribution not using the GNU tools.
>
> I don't think this was a GNUism. I think this change was in response to
> a bug I reported where the pr
On Mon, Dec 21, 2009 at 02:46:41PM +1100, Manuel M T Chakravarty wrote:
> > 'tr' -d '\t' < compiler/main/DynFlags.hs | '/usr/bin/sed' '/^xFlags/,/]/s/^
> > *( *\("[^"]*"\)[^"]*/ [\1] ++/p;d' >>
> > utils/ghc-cabal/dist-dummy-ghc/build/dummy-ghc.hs
> > sed: 1: "s|/Users/chak/Code/ghc- ...": bad f
FYI: I'm now able to build a boot.tar.gz on i386 for i386 and use
that to build a working ghc on a i386 with no ghc preinstalled. I
didn't yet try this chain on amd64, but I guess it works there,
too.
Doing a real port (like building boot.tar.gz on amd64 and using it
on i3986) still doesn't work.
On Fri, Nov 20, 2009 at 02:46:08PM +, Simon Marlow wrote:
> >Do we really have to rely on a specific version of tar? Dependencies
> >like that always come back to bite you.
>
> It sounds like the --force-local option is required iff tar is GNU tar.
> So we should have a configure test for
On Thu, Nov 19, 2009 at 06:13:29PM +, Ian Lynagh wrote:
> > Why not just use a command line that's more portable? I don't know what
> > that `--force-local' means
>
> It tells tar that c:/mytar.tar is a local file, not a file on the host
> "c".
Yuk! Is this `a file on the host c' a general ms
On Thu, Nov 19, 2009 at 12:57:15PM +, Ian Lynagh wrote:
> On Thu, Nov 19, 2009 at 09:58:17PM +1100, Roman Leshchinskiy wrote:
> > cd a/b/c/ && /usr/bin/tar --force-local -jxf
> > ../../../../bindistprep/ghc-6.13.20091119-i386-apple-darwin.tar.bz2
> > tar: Option --force-local is not supported
On Thu, Nov 12, 2009 at 09:43:42AM +, Simon Marlow wrote:
> >Would it be ok to move the -I$(GHC_INCLUDE_DIR) portion to ghc/ghc.mk?
> >
> >For example, add this one to the BootingFromHC section of ghc/ghc.mk:
> >
> > ghc_CC_OPTS += -I$(GHC_INCLUDE_DIR)
>
> I take it back, I think I underst
On Thu, Nov 12, 2009 at 03:24:32PM +, Simon Marlow wrote:
> On 08/11/2009 18:18, Matthias Kilian wrote:
> >Sun Nov 8 12:02:25 CET 2009 Matthias Kilian
> > * When booting from HC files, search includes for headers
>
> I just committed a version of this. I believe th
On Mon, Nov 09, 2009 at 10:53:18AM +, Simon Marlow wrote:
> ifeq "$(BootingFromHc)" "YES"
> -SRC_CC_OPTS += -DNO_REGS -DUSE_MINIINTERPRETER
> -D__GLASGOW_HASKELL__=$(ProjectVersionInt)
> +SRC_CC_OPTS += -I$(GHC_INCLUDE_DIR) -DNO_REGS -DUSE_MINIINTERPRETER
> -D__GLASGOW_HASKELL__=$(ProjectVer
s one (I think you already said it's ok):
Sun Nov 8 20:01:39 CET 2009 Matthias Kilian
* Don't try to rebuild compiler/primop-*.hs-incl when BootingFromHc
And this one (or rather a cleaner replacement of it, since probably only
hschooks.c is affected):
Sun Nov 8 12:02:25 CET 2009 Ma
On Mon, Nov 09, 2009 at 08:29:25PM +0100, Matthias Kilian wrote:
> The rules in hs-suffix-rules-srcdir.mk didn't get pulled in by
> rts/ghc.mk. I'll retry with adding hs-suffix-rules-srcdir to it.
With the attached patch, a normal build and a build bootstrapping
from hc files are
On Mon, Nov 09, 2009 at 10:58:24AM +, Simon Marlow wrote:
> [Some rules for booting from hc files
> Matthias Kilian **20091108110827
> Ignore-this: bf6a3f7db25010b2de013f98377cbd92
> ] hunk ./rules/cmm-suffix-rules.mk 51
>
> endif
>
> +## For booting from .
On Mon, Nov 09, 2009 at 10:58:24AM +, Simon Marlow wrote:
>
> On 08/11/2009 18:30, Matthias Kilian wrote:
> endif
>
> +## For booting from .hc files.
> +$1/$2/build/%.$$($3_osuf) : $1/%.hc
> + "$$(MKDIRHIER)" $$(dir $$@)
> + "$$(CC)"
On Mon, Nov 09, 2009 at 01:58:09PM +, Simon Marlow wrote:
> >>Sun Nov 8 18:48:50 CET 2009 Matthias Kilian
> >> * Don't try to redefine ELF types
> >> At least on OpenBSD, ELFCLASS and all the Elf_* and ELF_* stuff is
> >> already defined (
On Mon, Nov 09, 2009 at 10:53:18AM +, Simon Marlow wrote:
> >Sun Nov 8 12:02:25 CET 2009 Matthias Kilian
> > * When booting from HC files, search includes for headers
>
> ifeq "$(BootingFromHc)" "YES"
> -SRC_CC_OPTS += -DNO_REGS -DUSE
On Sun, Nov 08, 2009 at 08:08:12PM +0100, Matthias Kilian wrote:
> Date: Sun, 8 Nov 2009 20:08:12 +0100 (CET)
> Message-Id: <200911081908.na8j8cwb031...@nutty.outback.escape.de>
> To: cvs-ghc@haskell.org
> From: Matthias Kilian
> Subject: darcs patch: Don't try to redef
Sun Nov 8 18:48:50 CET 2009 Matthias Kilian
* Don't try to redefine ELF types
At least on OpenBSD, ELFCLASS and all the Elf_* and ELF_* stuff is
already defined (in sys/exec_elf.h, which is included from elf_abi.h).
Hopefully, using Elf_Addr as a sentinel is enough to not brea
Sun Nov 8 20:01:39 CET 2009 Matthias Kilian
* Don't try to rebuild compiler/primop-*.hs-incl when BootingFromHc
New patches:
[Don't try to rebuild compiler/primop-*.hs-incl when BootingFromHc
Matthias Kilian **20091108190139
Ignore-this: 5e6a6ceea33c361213cef5bacee7eda6
] {
hunk
Sun Nov 8 12:08:27 CET 2009 Matthias Kilian
* Some rules for booting from hc files
New patches:
[Some rules for booting from hc files
Matthias Kilian **20091108110827
Ignore-this: bf6a3f7db25010b2de013f98377cbd92
] hunk ./rules/cmm-suffix-rules.mk 51
endif
+## For booting from .hc
Sun Nov 8 12:02:25 CET 2009 Matthias Kilian
* When booting from HC files, search includes for headers
New patches:
[When booting from HC files, search includes for headers
Matthias Kilian **20091108110225
Ignore-this: 97b69c0fe499f8c1cbbd8cbf92589e11
] hunk ./ghc.mk 1041
endif
ifeq
On Tue, Oct 06, 2009 at 09:38:01PM -0400, Jonah Petri wrote:
> sed: 1: "/^xFlags/,/]/{/^ (/{s/ ...": extra characters at the end of
> p command
> make[1]: *** [utils/ghc-cabal/dist-dummy-ghc/build/dummy-ghc.hs] Error 1
> make[1]: *** Deleting file `utils/ghc-cabal/dist-dummy-ghc/build/dummy-
>
On Tue, Oct 06, 2009 at 10:08:37AM +0100, Simon Marlow wrote:
> Apparently everyone's iconv is different. You'd think that "UTF32LE"
> would be a universal encoding name, but no.
Quoting the holy (and, in this case, not very helpful) open group
standard:
Settings of fromcode and tocode
On Fri, Oct 02, 2009 at 02:12:07AM +0100, Ian Lynagh wrote:
> > On Wed, Sep 30, 2009 at 11:54:24PM +0200, Matthias Kilian wrote:
> > > Wed Sep 30 23:26:29 CEST 2009 Matthias Kilian
> > > * System.Console.Terminfo isn't available when bootstrapping.
>
> Ap
On Fri, Oct 02, 2009 at 05:12:28PM +0100, Ian Lynagh wrote:
> > I'm worried that every time we update gcc we'll get an immortal copy in
> > darcs/patches, which will significantly increase the size of the
> > snapshots over time...
>
> This patch was 16M, and I have 152M of patches.
Size doesn
On Wed, Sep 30, 2009 at 11:54:24PM +0200, Matthias Kilian wrote:
> Wed Sep 30 23:26:29 CEST 2009 Matthias Kilian
> * System.Console.Terminfo isn't available when bootstrapping.
FYI: This should be appropriate for both head *and* the 6.12 branch.
Ciao,
Kili
--
> Are t
On Tue, Sep 29, 2009 at 11:23:52PM +0200, Matthias Kilian wrote:
> Tue Sep 29 23:21:10 CEST 2009 Matthias Kilian
> * Follow renaming of binary to ghc-binary.
oh, now i can build ghc-6.12 from ghc-6.10, but i can't build
ghc-6.12 from ghc-6.12 (see below).
Could you please st
Tue Sep 29 23:21:10 CEST 2009 Matthias Kilian
* Follow renaming of binary to ghc-binary.
New patches:
[Follow renaming of binary to ghc-binary.
Matthias Kilian **20090929212110
Ignore-this: f28a7deafa283ef1c9c57ed38a90dbbf
] hunk ./utils/ghc-pkg/ghc.mk 64
-ilibraries/filepath
On Wed, Sep 16, 2009 at 09:52:41AM +0100, Simon Marlow wrote:
> Ok, so concrete steps we can take:
>
> - I think we should include install/binary-dist in validate for the
>run up to the release at least, and turn it off afterward.
>
> - People need to use the recommended 2-tree workflow to
And here's a little rant...
On Tue, Sep 15, 2009 at 11:40:50PM +0200, Matthias Kilian wrote:
> Tue Sep 15 22:42:54 CEST 2009 Matthias Kilian
> * Follow the builtin:rts vs. builtin_rts renaming
>
> This unbreaks make install again.
Why are some parts of the build cont
On Thu, Sep 10, 2009 at 05:02:14AM -0700, Simon Marlow wrote:
> Thu Sep 10 03:27:03 PDT 2009 Simon Marlow
> * Change the representation of the package database
> Ignore-this: 7c9b38ded7f753d5bb95743695700dbc
>
>- the package DB is a directory containing one file per package
> inst
[warning: long and broad build output logs -- better use a large
window for reading this mail]
Hi,
When trying to build ghc-6.11 using a previously built ghc-6.11 for
bootstrapping, linking ghc-stage1 fails with the following error:
"/home/kili/src/ghc/2/bin/ghc" -o ghc/stage1/build/tmp/ghc-stag
Sat Aug 29 12:23:58 CEST 2009 Matthias Kilian
* Make this test a little more stable
On slow machines, ghci can take more than one second to launch,
which produces confusing failure output like
+cat: 3171.pid: No such file or directory
+usage: kill [-s signame | -signum | -signame
On Wed, Aug 26, 2009 at 11:16:02PM +0200, Matthias Kilian wrote:
> Wed Aug 26 23:00:44 CEST 2009 Matthias Kilian
> * Use -W, not -Werror, for gcc older than 3.4
This shouldn't be pushed without checking first that it doesn't
break builds that use gcc-3.4 or newer.
Ci
On Wed, Aug 19, 2009 at 07:06:02PM +1000, Ben Lippmeier wrote:
> Perhaps what we need is some better build bots. It would be good to have
> a mechanism to run a build with "the current head + these patches" on
> all architectures easily.
It would also be helpful for those who are running build s
On Wed, Jul 22, 2009 at 10:23:31PM +0200, Krister Walfridsson wrote:
> The testsuite has two hard-coded paths to /usr/bin/python, wich prevents
> it from running on systems that installs python in other places. Most
> python scripts seems to solve this using the /usr/bin/env trick, as in
> the att
On Fri, Jul 10, 2009 at 07:17:24PM +0200, Matthias Kilian wrote:
> Yes, I missed the important bits, sorry for the noise. First, the
> `$(REALGOALS) all' should depend on *.old (which is ugly), second,
> the `touch $@' should be a `touch -r $@ $<', and I'm not sur
On Fri, Jul 10, 2009 at 10:19:20AM +0100, Simon Marlow wrote:
> >%.old: %
> > @set -x&& [ -f $@ ]&& cmp -s $< $@ || cp -p $< $@
> > touch $@
> >
> >And then simply change the `$(REALGOALS) all' rule to
> >
> >$(REALGOALS) all:mk/config.mk mk/project.mk compiler/ghc.cabal
> > @ec
On Thu, Jul 09, 2009 at 08:47:23AM -0700, Simon Marlow wrote:
> Thu Jul 9 05:46:16 PDT 2009 Simon Marlow
> * Use /usr/bin/test if it exists, and fix test syntax.
> Ignore-this: 83a75ba7c3ce2a1d02bddb7bfe414bfe
> Should fix Solaris build failures
>
> M ./Makefile -2 +7
>
> View patch
On Thu, May 07, 2009 at 11:42:18AM +1000, Manuel M T Chakravarty wrote:
> Are you sure, you got that right? I am getting
>
> >limitingfactor chak 16 (.../Code/ghc-test): sh validate
> >Booting .
> >Booting libraries/base
> >Booting libraries/directory
> >Booting libraries/old-time
> >Booting libr
On Mon, Mar 23, 2009 at 03:08:57AM -0700, Simon Marlow wrote:
> This checks if darwin_HOST_OS is defined and, if so, we call
> sysctlbyname() on the "hw.ncpu" property to get the processor count.
>
> M ./rts/posix/OSThreads.c +9
This shoulld work on all *BSDs, too.
__
On Mon, Mar 02, 2009 at 08:45:42PM +0100, Matthias Kilian wrote:
> > kili stable: fail (failed stage1)
> > tnaur PPC OSX stable 2: fail (failed stage1)
>
> This happens when building from ghc-6.6:
[...]
> Patch attached.
Now that poor stupid guy named "ki
n
Patch attached.
Ciao,
Kili
New patches:
[Don't ignore warnings when bootstrapping from ghc-6.6
Matthias Kilian **20090228130006
Ignore-this: d3d970230b88a85b365ce69361dc3316
] {
hunk ./Makefile 9
include $(TOP)/mk/boilerplate.mk
include $(TOP)/mk/cabal.mk
+ifeq "$(ghc_ge
On Wed, Jan 14, 2009 at 01:44:18PM -0800, Ian Lynagh wrote:
> Wed Jan 14 12:59:45 PST 2009 Ian Lynagh
> * Fix "make install": Put "[]" in the install package.conf
>
> M ./Makefile +2
>
> View patch online:
> http://darcs.haskell.org/ghc/_darcs/patches/20090114205945-3fd76-d4e6a9b3a11bad74
On Wed, Jan 28, 2009 at 05:56:12PM +, Ian Lynagh wrote:
> > If we have a dependency on make 3.81, we ought to check that in configure
> > (and document it). Ian - would you mind doing that? Thanks!
>
> Hmm, we can't do it in configure, because we don't know if you're about
> to type "make",
I'm not sure which patch did the magic (iirc, there where two or
three patches to the linker within the last five weeks), but i just
noticed that ghci now works on OpenBSD/amd64, too. Thanks!
Ciao,
Kili
___
Cvs-ghc mailing list
Cvs-ghc@haskell.o
1 - 100 of 130 matches
Mail list logo
