In light of an upcoming ICFP submission I'm trying to refactor some
back-end code. Unfortunately with the new build system I haven't figured
out how to make a new file work. I added compiler/cmm/DFLattice.hs,
and a DFLattice.o gets built all right, but it's not linked with the
other files. I add
> Simon PJ and I had a talk about the build system earlier today, I thought
> I'd float the idea we discussed...
> I propose we do this:
>
> - Extract the code from Cabal that generates Makefiles, and treat it as
> part of the GHC build system. Rather than generating a Makefile
>
> this place seems to allude to "hints" as well as the hacky tuple that I
> think is the one I fixed by turning it into a data type (so it could be put
> into typeclasses reasonably), but maybe isn't up-to-date, and doesn't
> explain it:
> http://hackage.haskell.org/trac/ghc/wiki/Commentary/Co
ying and
implementing procedure calling conventions, which is the use to which
they are put within GHC.
Norman
P.S. I have placed this information in the Commentary at
http://tinyurl.com/4kmpw9
> [EMAIL PROTECTED] wrote:
> > Sat May 3 15:45:14 PDT 2008 Norman Ramsey <[EMAIL P
while taking extra care not to screw up, I dran
darcs whatsnew -ls
and I got a flood of entries from libffi. Perhaps someone
with more knowledge than I can figure out how to expunge them?
And perhaps someone tell me which of the following files are boring?
./compiler/ghci/LibFFI.hs
./compile
> Any chance this was a partial repository?
No chance.
> For what it's worth, I now have darcs 1.0.9 as well as darcs 2.0.0 on my
> machine, so if there's any specific steps I can take to reproduce this, let
> me know.
Pushing a 100+MB tarball up through my ADSL line will knock out
Internet
>
>
> Simon/Norman: what version of darcs was this? I'm hoping this is something
> that darcs2 has already fixed.
>
> More specifically, looking at issue279, I understand that David has fixed a
> get_extra commuting patch bug between darcses 1.0.9 and 2.0.0.
This is in fact 1.0.9 (Debian
Could someone who knows something about ./configure have a look at
http://hackage.haskell.org/trac/ghc/ticket/2257
and suggest how I could work around this bug in order to build
GHC? Right now I can't do a thing because ./configure hangs...
Norman
___
I note that from
http://hackage.haskell.org/trac/ghc/ticket/1558
testing has to run single-threaded. ./validate is frequently on
my critical path, and I would love to be able to use two (soon, four)
processors. The last activity on this ticket was around September
from Simon Marlow. The note
I'm trying to get my copy of the head up to date,
and I'm tripping over a bug in darcs:
: [EMAIL PROTECTED] 9937 ; ./darcs-all pull -av
== running darcs pull -av
Pulling from "http://darcs.haskell.org/ghc";...
This is the GHC darcs repository (HEAD branch)
For more information, visit the GHC dev
Dear Fellow GHC Hackers,
I'll be buying new computers soon, and right now the only thing I do
for which computer performance is a problem is GHC hacking---especially
running 'validate'. I've got a good performance boost on my current
machine by using two SATA drives striped in a RAID-0 configurat
> Norman Ramsey <[EMAIL PROTECTED]> writes:
>
> > The need for both autoconf *and* automake is clearly stated at
> > http://hackage.haskell.org/trac/ghc/wiki/Building/Prerequisites
> > but it might be wise to have a Build-Depends on these packages in
&g
but it might be wise to have a Build-Depends on these packages in
debian/control. (Maybe it's there but I missed it.)
Norman
--- Forwarded Message
From: Ben Pfaff <[EMAIL PROTECTED]>
To: [EMAIL PROTECTED]
Cc: [EMAIL PROTECTED], Norman Ramsey <[EMAIL PROTECTED]>,
I tried using [[GhcFile(...)]] to hard link some source code into the
developers' wiki pages, so that people could easily follow what's going on.
Unfortunately this has come back to bite me because instead of bringing
up something my browser understands, it comes back as a '.hs file'
which Firefox
> Another perspective is that no command that you have ever done (on
> a given repository) while connected will fail while disconnected.
Ah that's a very nice, succinct, pleasant property.
My fears are allayed.
Norman
___
Cvs-ghc mailing list
Cvs-
> For those that don't follow the darcs-devel list, David has just announced
> a pre-release of Darcs 2:
I did see an announcement on the Haskell list, and I was slightly
alarmed that some unspecified operations will no longer work
disconnected. I don't have time to keep up with the darcs mail
> We've been using darcs for nearly 2 years now, I feel it's time for a
> retrospective and to think seriously about whether darcs is still the best
> option to continue with...
Simon,
Your post about darcs was so thoughtful that I have almost nothing to add.
I did learn a bit about distrib
> > perhaps, this is of more interest, when I also mention that I'd like
> > to fix the Sparc native code generator. I will not be able to do it
> > very quickly, because I have to read (and understand) a lot of stuff
> > beforehand, so if anybody else is quicker, no problem. Otherwise, I'd
>
I pushed to the head despite a problem with tc236(normal) failing.
There are some testsuite patches I've been unable to pull owing
to a bug in darcs. The error has to do with a parse error on forall,
which is not where I've been making any messes :-)
Norman
_
> validate dies for me with the appended error on MacOS with the
> latest HEAD.
Sorry about this; I had some bad code lurking behind an untriggered #ifdef.
I've now put something on the developers' wiki about how to avoid
#ifdef DEBUG in many situations. Maybe this will stop somebody else
fro
20 matches
Mail list logo
