GitLab

We no longer use dummy-ghc; instead we don't configure most packages
until the stage1 compiler is available.
  
We also now use Cabal for building the ghc-bin package.

There are a couple more sanity checks too.

Using Haskell conditionals means the compiler sees all the code, so
there should be less rot of code specific to uncommon arches. Code
for other platforms should still be optimised away, although if we want
to support targetting other arches then we'll need to compile it
for-real anyway.

The LLVM code generator is always built unconditionally, so both the
configuration variable in mk/config.mk.in as well as the string in
compilerInfo can be removed.

If the GHCi .o lib doesn't exist, load the .a instead

We now use the CONF_CC_OPTS_STAGEn C flags in machdepCCOpts, rather than
repeating them there.

This also means the file is generated in a dist directory, not a
source directory.

Simon M and I looked at this, and we think GhcLibProfiled is
(a) not needed (b) confusing.

Ian should review.

Really, if GhcProfiled is on we should also 
check that 'p' is in the GhcLibWays

    /usr/bin/ld -Wl,--relax -r -o dist-stage1/build/HSghc-6.10.4.o \
                                  dist-stage1/build/BasicTypes.o dist-stage1/build/DataCon.o ...
    /usr/bin/ld: unrecognized option '-Wl,--relax'

If we just drop '-Wl,' part it will not help as '-r' and '--relax' are incompatible.

Looks like '-Wl,--relax' was skipped by earlier binutils' ld as unknown option.
Removing ia64 specific path.

We do this through a gnu as feature called subsections,
where you can put data/code into a numbered subsection
and those subsections will be joined together in descending
order by gas at compile time.

This was done as part of an honours thesis at UNSW, the paper describing the
work and results can be found at:

http://www.cse.unsw.edu.au/~pls/thesis/davidt-thesis.pdf

A Homepage for the backend can be found at:

http://hackage.haskell.org/trac/ghc/wiki/Commentary/Compiler/Backends/LLVM

Quick summary of performance is that for the 'nofib' benchmark suite, runtimes
are within 5% slower than the NCG and generally better than the C code
generator.  For some code though, such as the DPH projects benchmark, the LLVM
code generator outperforms the NCG and C code generator by about a 25%
reduction in run times.

We still need the workaround for when compiling HEAD with 6.12.2

I had incorrectly "optimised" checkHiBootIface so that it forgot
to update the "knot-tied" type environment. 

This patch fixes the HEAD

The new Makefile logic was enabling the stage 1 rules when stage=2,
so "make 2" was rebuilding stage 1.

In particular, this fixes a problem where stage3 bits weren't being cleaned

Partly this is cleaner as we only have to preprocess the source files
once, but also it is necessary to avoid Haddock recompiling source
files when Template Haskell is in use, saving some time in validate
and fixing a problem whereby when HADDOCK_DOCS=YES, make always
re-haddocks the DPH packages.  This also needs an additional fix to
GHC.

HsColour support still uses Cabal, and hence preprocesses the source
files again. We could move this into the build system too, but there
is a version dependency that would mean adding extra autoconf stuff.

It wasn't fatal, but better to avoid it anyway

CPP finds the Rts.h, RtsFlags.h etc from the tree, rather than the
bootstrapping compiler, and then fails because it doesn't think
RtsFlags.h should be used any more.

It was munging 6.12.1 into 62

We now just call gcc to get the dependencies directly

Patch submitted by Matthias Kilian <kili@outback.escape.de>

GHC's recompilation checker doesn't take into account #included files,
which is really a bug.  We work around it here by adding dependencies
and using -fforce-recomp in a couple of places.

There are now 104 calls to mkdirhier, down from 1201, when validating.

It confuses the build system.