wolfgang [Tue, 27 May 2003 21:14:22 +0000 (21:14 +0000)]
[project @ 2003-05-27 21:14:21 by wolfgang]
Fix two bugs in the PowerPC NCG:
1. it generated a 'subfi' (subtract from with immediate) instruction,
which doesn't exist in the PowerPC architecture.
2. didn't correctly handle switch tables (test case cg048.hs).
MERGE TO STABLE
simonmar [Tue, 27 May 2003 14:52:34 +0000 (14:52 +0000)]
[project @ 2003-05-27 14:52:34 by simonmar]
Small updates for 6.x
simonmar [Tue, 27 May 2003 14:15:40 +0000 (14:15 +0000)]
[project @ 2003-05-27 14:15:40 by simonmar]
Use ErrUtils.showPass for printing the '*** Core Linted...' message
instead of rolling our own (also has the effect of sending this to
stderr for consistency).
Tidy up unused imports while I'm here.
simonmar [Tue, 27 May 2003 14:05:11 +0000 (14:05 +0000)]
[project @ 2003-05-27 14:05:11 by simonmar]
Send "Result size =" messages to stderr.
simonmar [Tue, 27 May 2003 14:02:06 +0000 (14:02 +0000)]
[project @ 2003-05-27 14:02:04 by simonmar]
Enable large-file support using AC_SYS_LARGEFILE.
HEADS UP: autoconf 2.5 is now needed to process configure.in. I
figured this was preferable to copying the AC_SYS_LARGEFILE macro from
the autoconf sources and trying to make it work with autoconf 2.13.
simonpj [Tue, 27 May 2003 12:47:55 +0000 (12:47 +0000)]
[project @ 2003-05-27 12:47:55 by simonpj]
Track changes in Generics.hs
simonmar [Tue, 27 May 2003 12:40:19 +0000 (12:40 +0000)]
[project @ 2003-05-27 12:40:19 by simonmar]
If we're going to surround all arguments to external functions with
quotes (see SysTools.lhs rev. 1.88) then we shouldn't add quotes to
the arguments to -D and -U passed to cpp.
simonpj [Tue, 27 May 2003 11:06:41 +0000 (11:06 +0000)]
[project @ 2003-05-27 11:06:41 by simonpj]
Advise against SplitObjs for Win32
stolz [Sun, 25 May 2003 20:54:18 +0000 (20:54 +0000)]
[project @ 2003-05-25 20:54:18 by stolz]
Fix quoted source filenames on *nix
Noticed by: Calle Lejdfors
(please merge)
wolfgang [Fri, 23 May 2003 21:40:04 +0000 (21:40 +0000)]
[project @ 2003-05-23 21:40:04 by wolfgang]
Enable Native Code Generator for PowerPC by default.
MERGE TO STABLE
panne [Fri, 23 May 2003 16:36:58 +0000 (16:36 +0000)]
[project @ 2003-05-23 16:36:58 by panne]
Nuked unused GL/GLU version detection magic, extensions are now
completely handled at runtime, not compile time. This affects only
configure's --enable-hopengl option.
panne [Fri, 23 May 2003 16:24:57 +0000 (16:24 +0000)]
[project @ 2003-05-23 16:24:57 by panne]
Fixed tests for _SC_GETGR_R_SIZE_MAX and _SC_GETPW_R_SIZE_MAX.
Whitespace is sometimes important in autconf scripts. :-P
MERGE TO STABLE
stolz [Fri, 23 May 2003 16:11:35 +0000 (16:11 +0000)]
[project @ 2003-05-23 16:11:35 by stolz]
Two more Solaris fixes:
- ln -s -f doesn't really work (but doesn't matter)
- 'echo -n' neither necessary nor portable
(please merge)
stolz [Fri, 23 May 2003 14:31:46 +0000 (14:31 +0000)]
[project @ 2003-05-23 14:31:45 by stolz]
No (un)setenv until SUSv3 (e.g. Solaris 2.9). (fallback untested)
simonmar [Fri, 23 May 2003 11:03:39 +0000 (11:03 +0000)]
[project @ 2003-05-23 11:03:39 by simonmar]
The HEAD is version 6.1 now.
simonmar [Fri, 23 May 2003 10:00:04 +0000 (10:00 +0000)]
[project @ 2003-05-23 10:00:04 by simonmar]
For STANDALONE_PACKAGE: Let the package configuration file refer to
$(libdir) as ${pkglibdir}. Note we can't use ${libdir} because
ghc-pkg already redefines it to point to GHC's libdir (bug or
feature?).
simonmar [Fri, 23 May 2003 08:49:22 +0000 (08:49 +0000)]
[project @ 2003-05-23 08:49:22 by simonmar]
Make hackery to give $(GhcCanonVersion) the right value even when we
have a single-digit $(GhcMinVersion).
simonmar [Fri, 23 May 2003 08:28:48 +0000 (08:28 +0000)]
[project @ 2003-05-23 08:28:48 by simonmar]
Remove ancient vestiges of StgHugs: the COMPILER #define.
simonmar [Thu, 22 May 2003 15:36:08 +0000 (15:36 +0000)]
[project @ 2003-05-22 15:36:08 by simonmar]
First hack at supporting standalone packages.
What this means is that a non-GHC fptools project which sets
$(PACKAGE) in a Makefile should do something reasonable. It works by
testing whether $(ProjectNameShort) == "ghc", and if it isn't, then:
- it turns off all the machinery that deals with installing the
package for use with the in-place compiler and a (future)
installation of GHC.
- instead, it enables a rule that installs the package using
the installed $(GHC_PKG). The configure script still needs
updating to detect this, so 'make install' won't work yet.
- $(GHC) is used instead of $(GHC_INPLACE) for compiling Haskell
code.
Apart from the detection of ghc-pkg in configure, this should be all
that's needed to get source distributions and 'make install' going.
simonmar [Thu, 22 May 2003 13:22:39 +0000 (13:22 +0000)]
[project @ 2003-05-22 13:22:39 by simonmar]
Comment only
simonmar [Thu, 22 May 2003 10:53:53 +0000 (10:53 +0000)]
[project @ 2003-05-22 10:53:53 by simonmar]
Fix obscure bug in GHCi: when generating code for tag2enum#, we were
wrongly using the source name for the DataCons rather than the worker
name, which lead to spurious link errors.
This fixes galois_raytrace(ghci).
ross [Thu, 22 May 2003 10:27:52 +0000 (10:27 +0000)]
[project @ 2003-05-22 10:27:52 by ross]
Make it work with newer versions of flex, which produce an extern
declaration for yywrap() unless YY_SKIP_YYWRAP is defined.
igloo [Wed, 21 May 2003 23:40:10 +0000 (23:40 +0000)]
[project @ 2003-05-21 23:40:08 by igloo]
Rename and reorder the internals for unique ids etc. Also fixed a couple
of THSyntax names.
igloo [Wed, 21 May 2003 18:07:30 +0000 (18:07 +0000)]
[project @ 2003-05-21 18:07:13 by igloo]
The Great Renaming. I hope I've kept everything in sync - and all the tests
pass. Now datatypes follow the
data Foo = <a kind of Foo>Foo
| <another kind of Foo>Foo
convention and the smart constructors can be uniformly created by
lowercasing the first letter of the constructor.
stolz [Wed, 21 May 2003 16:02:44 +0000 (16:02 +0000)]
[project @ 2003-05-21 16:02:42 by stolz]
Solaris2 needs _POSIX_PTHREAD_SEMANTICS for the getpw*_r() prototypes.
Make libraries/unix/Makefile use a new variable unix_SRC_HSC2HS_OPTS
which we configure in mk/config.mk.
simonpj [Wed, 21 May 2003 15:49:54 +0000 (15:49 +0000)]
[project @ 2003-05-21 15:49:54 by simonpj]
Add a microscopic amount of documentation for functional dependencies; who would like to volunteer to write more?
simonmar [Wed, 21 May 2003 15:20:37 +0000 (15:20 +0000)]
[project @ 2003-05-21 15:20:37 by simonmar]
The 6.0 release notes.
stolz [Wed, 21 May 2003 13:22:46 +0000 (13:22 +0000)]
[project @ 2003-05-21 13:22:46 by stolz]
Call strerror() instead of sys_errlist[errno] (deprecated & smells bad)
(Broke Solaris-build here)
simonmar [Wed, 21 May 2003 13:05:49 +0000 (13:05 +0000)]
[project @ 2003-05-21 13:05:49 by simonmar]
Restore the correct partitioning of command-line arguments into
objects and compilation-manager inputs for --make mode, and restore a
comment explaining the behaviour.
Rev. 1.120 made some subtle changes to the semantics, in particular in
cases where a file given on the command line is neither a source nor
an object file.
I believe the behaviour for one-shot mode has not changed. The
behaviour for GHCi mode is now the same as --make mode (previous to
rev. 1.120 it was subtly broken in this respect).
simonmar [Wed, 21 May 2003 12:46:19 +0000 (12:46 +0000)]
[project @ 2003-05-21 12:46:19 by simonmar]
revert rev. 1.16; looks like this was an experiment that got
accidentally committed.
simonmar [Wed, 21 May 2003 12:38:37 +0000 (12:38 +0000)]
[project @ 2003-05-21 12:38:36 by simonmar]
Put the call to saveDynFlags back in one place.
reid [Wed, 21 May 2003 10:51:31 +0000 (10:51 +0000)]
[project @ 2003-05-21 10:50:55 by reid]
Improved greencard-related definitions
igloo [Wed, 21 May 2003 02:58:40 +0000 (02:58 +0000)]
[project @ 2003-05-21 02:58:39 by igloo]
Added support for newtypes to TH and altered a test for them.
igloo [Wed, 21 May 2003 02:48:56 +0000 (02:48 +0000)]
[project @ 2003-05-21 02:48:56 by igloo]
When --make is used we need to save the initial dynamic flags too
igloo [Tue, 20 May 2003 22:39:33 +0000 (22:39 +0000)]
[project @ 2003-05-20 22:39:33 by igloo]
Fix building on 5.02
stolz [Tue, 20 May 2003 11:07:54 +0000 (11:07 +0000)]
[project @ 2003-05-20 11:07:54 by stolz]
Generate
int main (int argc, char *argv [])
instead of
int main (void)
Suggested by: Nicolas Oury
simonmar [Tue, 20 May 2003 10:03:25 +0000 (10:03 +0000)]
[project @ 2003-05-20 10:03:25 by simonmar]
Skeleton 6.0 release notes.
simonmar [Tue, 20 May 2003 10:03:06 +0000 (10:03 +0000)]
[project @ 2003-05-20 10:03:06 by simonmar]
defaultsHook is now ghc_rts_opts.
simonmar [Tue, 20 May 2003 10:02:13 +0000 (10:02 +0000)]
[project @ 2003-05-20 10:02:13 by simonmar]
Update version numbering policy
simonmar [Tue, 20 May 2003 09:56:58 +0000 (09:56 +0000)]
[project @ 2003-05-20 09:54:42 by simonmar]
Version 6.0
sof [Tue, 20 May 2003 03:46:48 +0000 (03:46 +0000)]
[project @ 2003-05-20 03:46:48 by sof]
--with-gcc=<path>: for mingw32 plat, canonicalise path using cygpath
simonpj [Mon, 19 May 2003 15:39:18 +0000 (15:39 +0000)]
[project @ 2003-05-19 15:39:17 by simonpj]
---------------------------------
Improve the dynamic-linking story
---------------------------------
Arrange proper initialisation for the dynamic linker. Whenever the dynamic linker does
anything (Linker.linkExpr), it first initialises itself, arranging to only do so once
of course.
"Initialising itself" includes loading any .o files, libraries, and packages specified
on the command line. The main effect of all this is to fix a Template Haskell problem,
which happened when a TH link needed some C library that it couldn't link. Now it does.
While I was at it, I tidied up main/Main.hs quite a bit. This is a delicate area (handling
the command line arguments), but I don't believe I broke anything! All the libraries
build, and lots of tests run.
Wolfgang: I also jiggled the darwin_TARGET_OS stuff a little in Linker.lhs, but again
I think the net effect is zero. You might want to check.
simonpj [Mon, 19 May 2003 15:11:15 +0000 (15:11 +0000)]
[project @ 2003-05-19 15:11:15 by simonpj]
Comments only
simonpj [Mon, 19 May 2003 15:10:41 +0000 (15:10 +0000)]
[project @ 2003-05-19 15:10:40 by simonpj]
--------------------------
Minor Template Haskell bug
--------------------------
This bug meant that spliced-in class declarations yielded a 'op not in scope',
where op was the class operation. Thanks to Andre Pang for spotting this.
Some consequential tidying up in parsing too.
simonmar [Mon, 19 May 2003 13:29:15 +0000 (13:29 +0000)]
[project @ 2003-05-19 13:29:15 by simonmar]
- rearrange the file a little
- build a binary distribution from stage 2 by default, unless
overriden from the command line.
simonmar [Mon, 19 May 2003 13:01:23 +0000 (13:01 +0000)]
[project @ 2003-05-19 13:01:23 by simonmar]
Remove obsolete references to DriFT, DtdToHaskell and Xtract.
simonmar [Mon, 19 May 2003 12:11:14 +0000 (12:11 +0000)]
[project @ 2003-05-19 12:11:14 by simonmar]
tryM again: make it work with GHC 5.02.x (I hope). Seems we changed
the representation of UserErrors several times in the last few
releases...
simonmar [Mon, 19 May 2003 11:37:07 +0000 (11:37 +0000)]
[project @ 2003-05-19 11:37:07 by simonmar]
Fix a pair of off-by-one errors that meant that BCOs were normally two
words larger than they needed to be.
simonmar [Mon, 19 May 2003 11:33:06 +0000 (11:33 +0000)]
[project @ 2003-05-19 11:33:06 by simonmar]
Eeek! intsToReverseBitmap was borked for bitmaps with more than
wORD_SIZE_IN_BITS entries.
simonpj [Mon, 19 May 2003 08:31:11 +0000 (08:31 +0000)]
[project @ 2003-05-19 08:31:11 by simonpj]
More Win32 help
reid [Sat, 17 May 2003 14:51:28 +0000 (14:51 +0000)]
[project @ 2003-05-17 14:49:45 by reid]
Add checks for ghc-pkg and green-card 3.00 (used in x11 lib)
ross [Sat, 17 May 2003 00:11:30 +0000 (00:11 +0000)]
[project @ 2003-05-17 00:11:27 by ross]
Rename per-package configuration files from $(PACKAGE).conf.* to
package.conf.*, making them easier to find (since each package is
in a separate directory anyway).
simonmar [Fri, 16 May 2003 14:39:29 +0000 (14:39 +0000)]
[project @ 2003-05-16 14:39:29 by simonmar]
Make retainer profiling work again after the recent SRT changes.
simonmar [Fri, 16 May 2003 14:16:53 +0000 (14:16 +0000)]
[project @ 2003-05-16 14:16:53 by simonmar]
heapCensus should grok IND_OLDGEN objects, because compacting GC
doesn't always eliminate them (perhaps it should).
simonmar [Fri, 16 May 2003 10:14:24 +0000 (10:14 +0000)]
[project @ 2003-05-16 10:14:20 by simonmar]
Now that we have auto packages, it makes sense to keep all the
interfaces for hierarchical libraries in the same directory tree. So
now, instead of putting interfaces for package P in $libdir/imports/P,
we put them all in $libdir/imports.
Interfaces for old non-auto non-hierarchical packages now go in
$libdir/hslibs-imports/P for package P.
simonmar [Fri, 16 May 2003 08:38:59 +0000 (08:38 +0000)]
[project @ 2003-05-16 08:38:59 by simonmar]
Doc the -L ghc-pkg option.
simonmar [Fri, 16 May 2003 08:35:49 +0000 (08:35 +0000)]
[project @ 2003-05-16 08:35:49 by simonmar]
Add -L (--list-packages-local) flag which lists packages in the
topmost configuration file only.
ross [Thu, 15 May 2003 12:28:29 +0000 (12:28 +0000)]
[project @ 2003-05-15 12:28:29 by ross]
update, including a short note on Nhc98.
simonmar [Wed, 14 May 2003 11:55:11 +0000 (11:55 +0000)]
[project @ 2003-05-14 11:55:11 by simonmar]
Should have been committed with recent SRT changes (see rev. 1.39 of StgSyn.lhs).
simonmar [Wed, 14 May 2003 09:14:02 +0000 (09:14 +0000)]
[project @ 2003-05-14 09:13:52 by simonmar]
Change the way SRTs are represented:
Previously, the SRT associated with a function or thunk would be a
sub-list of the enclosing top-level function's SRT. But this approach
can lead to lots of duplication: if a CAF is referenced in several
different thunks, then it may appear several times in the SRT.
Let-no-escapes compound the problem, because the occurrence of a
let-no-escape-bound variable would expand to all the CAFs referred to
by the let-no-escape.
The new way is to describe the SRT associated with a function or thunk
as a (pointer+offset,bitmap) pair, where the pointer+offset points
into some SRT table (the enclosing function's SRT), and the bitmap
indicates which entries in this table are "live" for this closure.
The bitmap is stored in the 16 bits previously used for the length
field, but this rarely overflows. When it does overflow, we store the
bitmap externally in a new "SRT descriptor".
Now the enclosing SRT can be a set, hence eliminating the duplicates.
Also, we now have one SRT per top-level function in a recursive group,
where previously we used to have one SRT for the whole group. This
helps keep the size of SRTs down.
Bottom line: very little difference most of the time. GHC itself got
slightly smaller. One bad case of a module in GHC which had a huge
SRT has gone away.
While I was in the area:
- Several parts of the back-end require bitmaps. Functions for
creating bitmaps are now centralised in the Bitmap module.
- We were trying to be independent of word-size in a couple of
places in the back end, but we've now abandoned that strategy so I
simplified things a bit.
simonmar [Wed, 14 May 2003 09:11:49 +0000 (09:11 +0000)]
[project @ 2003-05-14 09:11:49 by simonmar]
Comments only
simonmar [Wed, 14 May 2003 09:11:06 +0000 (09:11 +0000)]
[project @ 2003-05-14 09:11:06 by simonmar]
Calling mmap() with a size > 3Gig results in EINVAL on Linux. Catch
this case and report it as an out of memory condition instead of a
panic.
simonmar [Wed, 14 May 2003 09:07:28 +0000 (09:07 +0000)]
[project @ 2003-05-14 09:07:28 by simonmar]
Comment only
simonpj [Tue, 13 May 2003 15:37:23 +0000 (15:37 +0000)]
[project @ 2003-05-13 15:37:23 by simonpj]
Add optimistic-eval page; not linked yet
simonpj [Tue, 13 May 2003 08:17:13 +0000 (08:17 +0000)]
[project @ 2003-05-13 08:17:13 by simonpj]
Improve notes on Cygwin installation
simonmar [Mon, 12 May 2003 14:12:52 +0000 (14:12 +0000)]
[project @ 2003-05-12 14:12:52 by simonmar]
Another bugfix to tryM, this time for GHC 5.00--5.04
simonmar [Mon, 12 May 2003 09:22:35 +0000 (09:22 +0000)]
[project @ 2003-05-12 09:22:35 by simonmar]
Updates:
- Loading module Main without a main binding is no longer an error.
- Clarify the section on libraries to note that it means object-code
libraries, not source code.
- Update the section on packages now that most packages are
auto-loaded.
simonpj [Wed, 7 May 2003 16:20:21 +0000 (16:20 +0000)]
[project @ 2003-05-07 16:20:21 by simonpj]
Comment only
simonpj [Wed, 7 May 2003 09:30:09 +0000 (09:30 +0000)]
[project @ 2003-05-07 09:30:09 by simonpj]
Allow deriving(Show) for data types with unboxed fields
simonpj [Wed, 7 May 2003 08:30:08 +0000 (08:30 +0000)]
[project @ 2003-05-07 08:30:08 by simonpj]
Print type of data constructors correctly in GHCi
simonpj [Wed, 7 May 2003 08:29:48 +0000 (08:29 +0000)]
[project @ 2003-05-07 08:29:42 by simonpj]
Comments and formatting
simonpj [Tue, 6 May 2003 14:48:12 +0000 (14:48 +0000)]
[project @ 2003-05-06 14:48:12 by simonpj]
Note about context in data type decls
simonmar [Tue, 6 May 2003 10:36:17 +0000 (10:36 +0000)]
[project @ 2003-05-06 10:36:17 by simonmar]
Update: we now follow Haskell 98 w.r.t. exports from module Main.
simonpj [Tue, 6 May 2003 10:28:33 +0000 (10:28 +0000)]
[project @ 2003-05-06 10:28:32 by simonpj]
-------------------------------------
Implement deriving( Data )
-------------------------------------
Implements deriving( Data ), where the Data class is defined
in Data.Generics; its the "scrap your boilerplate" Term class.
Ralf is still converging on the exact definition of the Data class,
so the details may change.
simonpj [Tue, 6 May 2003 10:26:48 +0000 (10:26 +0000)]
[project @ 2003-05-06 10:26:48 by simonpj]
Import wibbles
simonpj [Tue, 6 May 2003 10:26:29 +0000 (10:26 +0000)]
[project @ 2003-05-06 10:25:32 by simonpj]
-------------------------------------
Main module exports
-------------------------------------
Make it so that
module Main where ....
exports everything defined in Main, as the report says it should.
simonpj [Tue, 6 May 2003 10:23:16 +0000 (10:23 +0000)]
[project @ 2003-05-06 10:23:16 by simonpj]
Formatting only
simonpj [Tue, 6 May 2003 10:22:54 +0000 (10:22 +0000)]
[project @ 2003-05-06 10:22:54 by simonpj]
Comments about eta expansion
igloo [Sun, 4 May 2003 13:21:49 +0000 (13:21 +0000)]
[project @ 2003-05-04 13:21:48 by igloo]
Add support for unboxed Ints, Floats and Doubles to Template Haskell.
ross [Thu, 1 May 2003 10:46:19 +0000 (10:46 +0000)]
[project @ 2003-05-01 10:46:19 by ross]
Make -t option to the script override the template file, just as -c
overrides the compiler, as the the documentation implies. (Formerly -t
added an extra template file, which is not so useful.)
simonmar [Thu, 1 May 2003 10:25:38 +0000 (10:25 +0000)]
[project @ 2003-05-01 10:25:38 by simonmar]
Use the DocBook 4.1 DTD instead of 3.1. Seems to work fine...
simonmar [Thu, 1 May 2003 08:43:54 +0000 (08:43 +0000)]
[project @ 2003-05-01 08:43:54 by simonmar]
Remove special cases for the setting of $(libdir) and $(datadir) on
mingw32, after discussion with Sigbjorn.
wolfgang [Tue, 29 Apr 2003 21:37:31 +0000 (21:37 +0000)]
[project @ 2003-04-29 21:37:31 by wolfgang]
Unbreak Mac OS X build (#ifdef mixup)
simonmar [Tue, 29 Apr 2003 09:32:20 +0000 (09:32 +0000)]
[project @ 2003-04-29 09:32:20 by simonmar]
update aLL_NON_PTRS for the new RET_DYN stack frame layout.
simonmar [Mon, 28 Apr 2003 10:02:15 +0000 (10:02 +0000)]
[project @ 2003-04-28 10:02:15 by simonmar]
Add a compile-time sanity check that RESERVED_STACK_WORDS is right.
simonmar [Mon, 28 Apr 2003 09:57:12 +0000 (09:57 +0000)]
[project @ 2003-04-28 09:57:12 by simonmar]
Comments to note the relationship between the RET_DYN stack frame size
and RESERVED_STACK_WORDS.
simonmar [Mon, 28 Apr 2003 09:55:20 +0000 (09:55 +0000)]
[project @ 2003-04-28 09:55:20 by simonmar]
Following the recent change to the layout of the StgRetDyn frame, we
now need to bump RESERVED_STACK_WORDS because this governs the amount
of room which is guaranteed to be available on the stack in the event
of a stack check failure.
This accounts for at least one cause of recent crashes in the HEAD.
qrczak [Sun, 27 Apr 2003 10:37:01 +0000 (10:37 +0000)]
[project @ 2003-04-27 10:37:01 by qrczak]
Strings in perl are compared by eq, not ==.
Fixes creating "nul" files on Unix.
qrczak [Sun, 27 Apr 2003 10:30:13 +0000 (10:30 +0000)]
[project @ 2003-04-27 10:30:13 by qrczak]
s/symbol link/symbolic link/
panne [Fri, 25 Apr 2003 20:58:09 +0000 (20:58 +0000)]
[project @ 2003-04-25 20:58:09 by panne]
Unbreak 2nd stage
simonmar [Fri, 25 Apr 2003 09:17:05 +0000 (09:17 +0000)]
[project @ 2003-04-25 09:17:05 by simonmar]
- Add a note about the incorrect handling of the 'module Main where'
header.
- While I'm here: fix out of date stuff, reformat and generally tidy up.
simonmar [Wed, 23 Apr 2003 10:42:36 +0000 (10:42 +0000)]
[project @ 2003-04-23 10:42:36 by simonmar]
Make a leading '~' expand to $HOME in commands that load source files
(:load, :add) in addition to :cd which already had this behaviour.
simonmar [Wed, 23 Apr 2003 08:54:45 +0000 (08:54 +0000)]
[project @ 2003-04-23 08:54:45 by simonmar]
Oops, forgot to update this file after the RET_DYN changes.
igloo [Tue, 22 Apr 2003 20:40:00 +0000 (20:40 +0000)]
[project @ 2003-04-22 20:39:59 by igloo]
Order declarations in reifications in order of source line number.
The bugs still there but it bites less often now...
Also remove the type parameterisation and do some type renaming as
discussed on the template-haskell list.
simonmar [Tue, 22 Apr 2003 16:25:12 +0000 (16:25 +0000)]
[project @ 2003-04-22 16:25:08 by simonmar]
Fix an obscure bug: the most general kind of heap check,
HEAP_CHECK_GEN(), is supposed to save the contents of *every* register
known to the STG machine (used in cases where we either can't figure
out which ones are live, or doing so would be too much hassle). The
problem is that it wasn't saving the L1 register.
A slight complication arose in that saving the L1 register pushed the
size of the frame over the 16 words allowed for the size of the bitmap
stored in the frame, so I changed the layout of the frame a bit.
Describing all the registers using a single bitmap is overkill when
only 8 of them can actually be pointers, so now the bitmap is only 8
bits long and we always skip over a fixed number of non-ptr words to
account for all the non-ptr regs. This is all described in StgMacros.h.
simonpj [Tue, 22 Apr 2003 11:44:17 +0000 (11:44 +0000)]
[project @ 2003-04-22 11:44:17 by simonpj]
Fix the context for derived Typeable instances
simonpj [Tue, 22 Apr 2003 09:30:53 +0000 (09:30 +0000)]
[project @ 2003-04-22 09:30:52 by simonpj]
Stage-2 wibbles to the Expected type changes
sof [Mon, 21 Apr 2003 14:45:28 +0000 (14:45 +0000)]
[project @ 2003-04-21 14:45:28 by sof]
drop curious *_debug_opts_prefix tables; no longer used
simonmar [Fri, 18 Apr 2003 09:45:32 +0000 (09:45 +0000)]
[project @ 2003-04-18 09:45:32 by simonmar]
Add a comment about the entry convention to stg_ap_* and why it's done
this way (thanks to Andy Moran for the prodding).
simonmar [Fri, 18 Apr 2003 09:40:10 +0000 (09:40 +0000)]
[project @ 2003-04-18 09:40:10 by simonmar]
Just noticed that all the stack checks in stg_ap_[1-8]_entry are too
low by one word. Thanks to Andy Moran for (unwittingly) causing me peer
at this code.