* as Ingo has pointed out, the remote user settings doesn't
relate to the build configuration at all, so setting the
remote user via HAIKU_REMOTE_USER in UserBuildConfig or
via shell environment is the way to go
* additionally: drop debug output
* add option --remote-user to configure, which sets HAIKU_REMOTE_USER
* add evaluation of HAIKU_REMOTE_USER variable when ssh-ing
into git.haiku-os.org
This allows FreeBSD with a pure /bin/sh (not a symlink to /bin/bash) to build
the cross-tools to a dedicated directory, outside of the generated folder.
Pass --enable-hybrid-secondary to gcc's configure when building it as
a secondary compiler. Doesn't make a difference for building Haiku
itself, but makes it easier to match the Haiku cross devel package with
the compiler when building bootstrap packages with haikuporter.
gcc 4 only ATM.
The goal is to do hybrid builds in a single jam (instead of calling a
sub-jam to build parts with the secondary tool chain). This changeset
adds support to configure to prepare multiple tool chains.
configure:
* Merge option --build-cross-tools-gcc4 into --build-cross-tools. The
option does now always require a packaging architecture parameter,
i.e. x86_gcc2 for the legacy tool chain.
* Multiple occurrences of the --build-cross-tools and
--cross-tools-prefix options are allowed. The first one specifies the
primary tool chain, the subsequent ones the secondary tool chains.
* All architecture dependent jam variables are now suffixed with the
name of the packaging architecture. The new HAIKU_PACKAGING_ARCHS
contains the packaging architectures for the prepared tool chains. The
first element is for the primary tool chain.
* No longer generate a separate libgccObjects file. Just put the
respective variable into BuildConfig as well.
build_cross_tools[_gcc4]:
* Replace the <haiku output dir> parameter by a <install dir>
parameter. This allows to create different cross-tools directories.
They are simply suffixed by the packaging architecture.
Jamrules:
* For the moment map the variables for the primary tool chain to the
respective suffix-less variables, so that everything still works as
before.
The next step is to actually support the secondary tool chains in the
jam build system. This will require quite a bit more butchering, though.
* Add configure option --bootstrap which allows specifying the
haikuporter and HaikuPorts repository paths.
* Add rules for supporting a second repository type. The
PackageRepository rule is now private and RemotePackageRepository is
used for remote repositories. The new BootstrapPackageRepository rule
is for defining a bootstrap repository (there will probably be only
the HaikuPorts cross repository) whose packages can be built as needed
via haikuporter.
* Rename DownloadPackage to FetchPackage.
* Define repository HaikuPortsCross.
* HaikuCrossDevel package(s): There are now two sets of packages: A
"stage1" set with the same content as before and a final set
additionally containing the libraries libbe, libnetwork, libpackage.
Those are needed for building the libsolv bootstrap package while for
building them we need other bootstrap packages (ICU, libz).
This is basically all that's required to build a bootstrap Haiku
completely from sources, with a few caveats:
* There's no ICU bootstrap recipe yet (so one has to cheat and use the
prebuilt package ATM).
* Probably doesn't work on Haiku yet (tested on Linux only).
* A 32 bit environment must be used (otherwise building the gcc 2
bootstrap package fails).
* Building with multiple jobs doesn't work yet, since haikuporter uses
common directories for building different packages and there's no
explicit serialization yet.
* Haven't tested the resulting image save for booting it. So it probably
needs a bit more work before it can actually build the final
HaikuPorts packages.
The new configure option "--use-xattr-ref" enables an xattr assisted
variant of the generic attribute emulation. Instead of using the inode
ID of a node to identify its attribute directory, we use a reasonably
unique random 128 bit number, which we generate and attach as an
attribute to the node. This way, when a node changes its inode ID
(defragmentation?) or the inode ID of a removed node with a left-over
attribute directory is reused, attributes won't get mixed up.
The old method is still used for symlinks (since on Linux only
priviledged users can write attributes on symlinks), but those usually
only have a rather boring BEOS:TYPE attribute, so mix-ups wouldn't be
that problematic anyway.
* The GCC2 buildtool sources were updated, so change the date
* A new GCC2 optional package will be added later today
+alpha 4 (build will break otherwise)
Turns out that libgcc is needed, for some reason building the kernel
with -O0 does not end up referencing libgcc but -O2 does. A separate
build of it is done with -mno-red-zone, same reason as for libsupc++.
Ended up being easy to rebuild with different CFLAGS: previously I'd
tried doing `CFLAGS="-mno-red-zone" make` in the libgcc dir which
didn't override, the correct way is `make CFLAGS="-mno-red-zone"`
Kernel mode code on x86_64 needs to be built with -mno-red-zone as
interrupts would corrupt the red zone if it were in use. However, the
kernel is linked with libsupc++, which was not compiled with
-mno-red-zone. If an interrupt occurred in libsupc++ code the red zone
would get corrupted. This was causing random panics, particularly under
heavy system load. Therefore, on x86_64 a separate build of libsupc++
with -mno-red-zone is now done for the kernel to use. Note: this commit
will require a rerun of configure and rebuild of cross tools.
The elfedit tool doesn't exist with the old binutils, so configure was
failing when it tried to get the path to it. Only try to search for it
if building GCC 4.
Added the necessary build flags for modules, and added a module (dpc)
to the floppy image for x86_64 builds for testing purposes. The module
gets loaded correctly and its code runs without issue. Only non-trivial
addition is the different method for generating kernel.so, this is
explained in the kernel Jamfile.
* Use gcc and g++ rather than cc and c++, as the latter now point to
clang with recent Xcode versions and compilation of the host tools
fail for various reasons with it.
* Replace the case-sensitive filesystem check with a more basic one,
as diskutil no longer supports the behaviour of getting info for the
volume that any path is on.
* Updated ReadMe with a correct list of prerequisites for OS X.
* GCC 2 builds are still broken due to a strange error that only
occurs with a GCC 2 built on OS X 10.7
