It's not used until then, but by moving it below including .config,
we can use CONFIG_* in the .xcompile file in the future.
Change-Id: I672f444dd28b5fae1fc339a1e0e78a249c9b7875
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/10422
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <edward.ocallaghan@koparo.com>
There has been a problem with out of tree build directories specified
using relative paths, as in
$ make obj=../build/peppy
while specifying full path to obj works fine. This patch fixes the
problem, making sure that make's path manipulation string substitute
command is applied to both source and build roots.
To test this ran the following script
echo > /tmp/build.log
for build_root in ./ ../ ''; do
build_dirs="${build_root}build/peppy"
if [ -n "${build_root}" ]; then
build_dirs+=" $(realpath ${build_root})/build/peppy"
fi
for build_dir in ${build_dirs}; do
rm -rf $build_dir .config* build* ../build*
make obj=${build_dir} menuconfig # configure for google peppy board
echo "building in ${build_dir}" >> /tmp/build.log
if ! make obj=${build_dir}; then
exit
fi
done
done
and then checked the generated file:
$ cat /tmp/build.log
building in ./build/peppy
building in /home/vbendeb/old_projects/coreboot/source_code/build/peppy
building in ../build/peppy
building in /home/vbendeb/old_projects/coreboot/build/peppy
building in build/peppy
Change-Id: If46b046108e906796fe84716e93bf341b3785f14
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: http://review.coreboot.org/10127
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
Make fails to update .xcompile in case there is an mv alias preventing
silent overwrites of existing files. To avoid ambiguity, invoke mv
from standard location.
While we are at it, drop the code creating a unique .xcompile instance
before copying it into destination: this was done for the case of
running multiple independent make processes in parallel, but there is
a problem with this approach with multiple processes copying their
unique copies into the same destination file.
The tool running multiple make invocations in parallel should
explicitly invoke xcompile first.
Change-Id: Iedc5e288fbcc5dfc18ce39de5c067bb869a13275
Signed-off-by: Vadim Bendebury <vbendeb@chromium.org>
Reviewed-on: http://review.coreboot.org/9994
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <pgeorgi@google.com>
This patch disables implicit make rules (like %.o: %.c) from our build
system, since we don't use them and they can cause unexpected results
when there's a mistake in the Makefiles.
BRANCH=None
BUG=chromium:466469
TEST=emerge-falco coreboot fails with only this patch and succeeds with
this and the $(src-to-obj) fix.
Change-Id: I7478adaddbbeaa2226fd941ffacfce2577ba59ba
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: c2543f25f7db61f93be96784eca3f4f65e3ffaa0
Original-Change-Id: Ia6a91f040ad4ef556cf7912cd82a73173dfec8cb
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/259947
Original-Reviewed-by: Mike Frysinger <vapier@chromium.org>
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Original-Reviewed-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9865
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The build system uses the $(src-to-obj) function in multiple places to
figure out the corresponding output object file for a certan source
file, most importantly when generating the rule to build it. Usually
this is pretty simple, but some odd cases are a little tricky... such as
the auto-generated intermediary C files containing an array definition
generated from an ASL file. These files have already been compiled per
stage, so they contain the stage as a suffix and reside in the build
directory (e.g. build/.../dsdt.ramstage.c).
The previous $(src-to-obj) implementation just blindly appends the stage
again and turns this into build/.../dsdt.ramstage.ramstage.o. This isn't
very useful, so to avoid confusion this patch makes it strip additional
stage suffixes for those intermediary files.
This also fixes a bug with the ASL postprocessor, which didn't take this
double suffix into account: it added build/.../dsdt.ramstage.o to
ramstage-objs which should've been build/.../dsdt.ramstage.ramstage.o.
This only worked by accident because make compiled the file with its
implicit %.o: %.c rule instead.
BRANCH=none
BUG=chromium:466469
TEST=emerge-falco coreboot with the new make 4.1. Also build Falco and
Veyron_Jerry with make -r -R to make sure there are no other accidental
uses of implicit rules in our build system.
Change-Id: I4aeaa60add1ef4215cb6c0b222c3886395c7a045
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: d9ea2e082eca1045409ea1f403082c97dedc70d8
Original-Change-Id: I951edbc9f653321a9084543a65009c6e9154d819
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/259950
Original-Reviewed-by: Mike Frysinger <vapier@chromium.org>
Original-Reviewed-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9861
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Dependency tracking in incremental builds is currently broken for the
ramstage, due to the intermediate linking step into one ramstage.o file
per directory. The original xxx.ramstage.o files are removed from
ramstage-objs, so they don't end up in allobjs and won't get translated
into DEPENDENCIES. This patch explicitly adds them to DEPENDENCIES
beforehand to resolve the issue.
BRANCH=None
BUG=None
TEST=Built, ran 'touch src/include/cbmem.h' and built again
incrementally. Confirmed that objects dependent on the modified header
such as timestamp.ramstage.o get rebuilt correctly.
Change-Id: I3ba411e4073b38e038445aadceeccfe6c09670c8
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Original-Commit-Id: 9c57d6a8421a109ee3e87567c9add579f9ae761e
Original-Change-Id: Ife529ad8f5c011456c1e0c380356f1b1bb5047cb
Original-Signed-off-by: Julius Werner <jwerner@chromium.org>
Original-Reviewed-on: https://chromium-review.googlesource.com/233571
Original-Reviewed-by: Aaron Durbin <adurbin@chromium.org>
Reviewed-on: http://review.coreboot.org/9745
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
This allows combining and simplifying linker scripts.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: Ie5c11bd8495a399561cefde2f3e8dd300f4feb98
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9303
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Instead of keeping this separate variable around, add linker scripts
to the $(class)-y source lists and let the build system sort things out.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: I4af687becf2971e009cb077debc902d2f0722cfb
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9289
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
So far we assumed that all files in *-srcs are below src/
which wasn't really true actually and will be less true with
future changes.
Fix up crt0.S handling on x86, which is covered by default rules
due to this change.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: Icae563c2d545b1aea809406e73faf3b417796a1b
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9288
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Provide a mechanism to define rules that apply for a given filetype
(.c, .asl, ...) across all classes (ramstage, ...)
Change-Id: If45c526d294e0374c32efef01f0213c6b78b1e43
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Reviewed-on: http://review.coreboot.org/9286
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tell gcc what to use as target part of the generated rule.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: Ie4814143337abb3cf1e9e8db7e96201a517a17b8
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9285
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
It also creates file names in the build directory and with
the stage sliced in, but keeps the extension for anything
not .c or .S.
Also some handling for non-.c/.S files was adapted to match.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: If8f89a7daffcf51f430b64c3293d2a817ae5120f
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9175
Reviewed-by: Aaron Durbin <adurbin@google.com>
Tested-by: build bot (Jenkins)
It can now be used on any list of files instead of assuming
to work on $($(class)-src).
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: Ib77afedd4c3f847963497beea503f5447a7c6e28
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9161
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Introduce generic-$(type)-ccopts and $(class)-generic-ccopts
to declare compiler flags that apply to all files of a certain
type or of a certain class. Then use them.
This is inspired by the commit listed below, but rewritten to match
upstream, and split in smaller pieces to keep intent clear.
Change-Id: I655688e82a0cc5bad89b6f55dc217b9f66b64604
Signed-off-by: Patrick Georgi <pgeorgi@chromium.org>
Based-On-Change-Id: I50af7dacf616e0f8ff4c43f4acc679089ad7022b
Based-On-Signed-off-by: Julius Werner <jwerner@chromium.org>
Based-On-Reviewed-on: https://chromium-review.googlesource.com/219170
Reviewed-on: http://review.coreboot.org/9114
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@google.com>
coverity isn't too happy with ccache, and given the current setup
it also isn't too useful.
Change-Id: I420fdd7350dff29296d7101569cb183afe1f92d6
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: http://review.coreboot.org/8478
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
So don't treat them as such. Fixes interactions with
class initialization (eg. verstage) on lint.
Change-Id: I8b5f2a56e73ba934590c45494e6a49f93c42096f
Signed-off-by: Patrick Georgi <pgeorgi@google.com>
Reviewed-on: http://review.coreboot.org/8472
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
This is a doxygen target I'm using for cleaning up the coreboot doxygen
output. It runs in about a minute instead of.... a lot longer, and
only generates documentation for the coreboot/src directory, excluding
vendorcode.
It was requested that the comments be stripped to make it easier to
read. This was done through the following command (split for gerrit):
head -n 1 Doxyfile.coreboot_simple > Doxyfile.tmp ;
cat Doxyfile.coreboot_simple | sed 's|^\s*#.*$||' | sed '/^$/d' | sed 's|[[:space:]]\+$||' >> Doxyfile.tmp ;
mv Doxyfile.tmp Doxyfile.coreboot_simple
I'm including the command here because any time the file is updated
with the doxygen wizard, it will need to be run again. It might be
desirable to turn this into a script in the documentation directory.
Change-Id: Ic0cbbcd21aa647e80a037481546bdcd2aab8949e
Signed-off-by: Martin Roth <gaumless@gmail.com>
Reviewed-on: http://review.coreboot.org/8068
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Too many false positives, not useful enough for us at the moment
at the cost of not having the whole tree build without warnings
as errors.
Change-Id: I9f9910b7f66ebf3a82d42e7732e413ba27dbbbe7
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/7778
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Tested-by: build bot (Jenkins)
We build with either gcc or clang, no need to keep both around
Change-Id: I9af2cc7636bdc791a68ba8ed6e7c5a81973c5dfd
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/7552
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
The ccopts mechanism is needed for passing ARM assembler flags to GCC.
There are many gotchas in adding ASFLAGS. As things have moved
around, the revert doesn't remove cleanly, so this reverts and cleans
up the ccopts.
This reverts commit 25b56c3af5.
Change-Id: I44c025535258e6afb05a814123c10c24775a88e8
Signed-off-by: Marc Jones <marc.jones@se-eng.com>
Reviewed-on: http://review.coreboot.org/7352
Tested-by: build bot (Jenkins)
Reviewed-by: Julius Werner <jwerner@chromium.org>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Make use of '-print-librt-file-name' over '-print-libgcc-file-name'
to use Compiler-RT runtime glue over libgcc glue.
NOTE: *** Requires at least clang 3.6.x
Change-Id: I7f63284473d6067bf775409970c8dd98f5d5a8d5
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6144
Reviewed-by: Marc Jones <marc.jones@se-eng.com>
Tested-by: build bot (Jenkins)
There are ARM systems which are essentially heterogeneous multicores where
some cores implement a different ARM architecture version than other cores. A
specific example is the tegra124 which boots on an ARMv4 coprocessor while
most code, including most of the firmware, runs on the main ARMv7 core. To
support SOCs like this, the plan is to generalize the ARM architecture so that
all versions are available, and an SOC/CPU can then select what architecture
variant should be used for each component of the firmware; bootblock,
romstage, and ramstage.
Old-Change-Id: I22e048c3bc72bd56371e14200942e436c1e312c2
Signed-off-by: Gabe Black <gabeblack@google.com>
Reviewed-on: https://chromium-review.googlesource.com/171338
Reviewed-by: Gabe Black <gabeblack@chromium.org>
Commit-Queue: Gabe Black <gabeblack@chromium.org>
Tested-by: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 8423a41529da0ff67fb9873be1e2beb30b09ae2d)
Signed-off-by: Isaac Christensen <isaac.christensen@se-eng.com>
ARM: Split out ARMv7 code and make it possible to have other arch versions.
We don't always want to use ARMv7 code when building for ARM, so we should
separate out the ARMv7 code so it can be excluded, and also make it possible
to include code for some other version of the architecture instead, all per
build component for cases where we need more than one architecture version
at a time.
The tegra124 bootblock will ultimately need to be ARMv4, but until we have
some ARMv4 code to switch over to we can leave it set to ARMv7.
Old-Change-Id: Ia982c91057fac9c252397b7c866224f103761cc7
Reviewed-on: https://chromium-review.googlesource.com/171400
Reviewed-by: Gabe Black <gabeblack@chromium.org>
Tested-by: Gabe Black <gabeblack@chromium.org>
Commit-Queue: Gabe Black <gabeblack@chromium.org>
(cherry picked from commit 799514e6060aa97acdcf081b5c48f965be134483)
Squashed two related patches for splitting ARM support into general
ARM support and ARMv7 specific pieces.
Change-Id: Ic6511507953a2223c87c55f90252c4a4e1dd6010
Signed-off-by: Isaac Christensen <isaac.christensen@se-eng.com>
Reviewed-on: http://review.coreboot.org/6782
Tested-by: build bot (Jenkins)
That file will be generated, but not before make managed
to complain about it.
So let's just generate it if missing - it won't hurt the
dependency tracking some lines later which is looking at
time stamps.
Change-Id: I615f38457eb27a8ffb4352b5234e262ee95d84ac
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/6305
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
It's probably safe to say that .xcompile needs an update if
util/xcompile/xcompile changed, so tell make about this
dependency.
Updates are honored immediately due to GNU make's feature of
reinterpreting everything when an included file changes. See "How
Makefiles Are Remade" in the GNU make documentation for details.
Change-Id: Ide2f028eaddcee66028c6403688cc83e1622fa6b
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/6255
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Currently we set HOSTCC=clang a little late meaning some minor bits
(utils/kconfig) are built with GCC. Move the assignment up the Makefile.
Change-Id: Ic72ad808eba0c0bf508bde34fb9bf0390c0b1d4d
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/6140
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Idwer Vollering <vidwer@gmail.com>
We now use the slightly more familiar CFLAGS_* and CPPFLAGS_*
for the same purpose.
Change-Id: Ifd2bd13f67f71fa0a15611a6d11a6a4c7994271b
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5875
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
The ccache support was mostly disabled because it
didn't hook onto most compilers anymore.
Caveat: ccache and scan-build don't work together since
scan-build doesn't like arguments in its compiler command
line (eg. "ccache gcc").
Change-Id: I7c1c6e22cb662f2b08e774ea484ac1c412fdd2db
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5775
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
This drops the scan-build related Kconfig options
since it's now possible to simply run
scan-build [-o outdir] make
and get coreboot built with its report.
There's also no inner make process anymore, and the way
things work should be clearer now.
Also adapt abuild to this new reality.
Change-Id: I03e03334761ec83f718b3235ebf811834cd2e3e3
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5774
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
Allow ccache and scan-build to wrap romcc.
This works a bit different from the other compilers
because we only define it later.
Change-Id: I3adce91d3dde9dd50aa6a2baad5b457744f35575
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5773
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
There are a couple of places where CPPFLAGS are
pasted into CFLAGS, eliminate them.
Change-Id: Ic7f568cf87a7d9c5c52e2942032a867161036bd7
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5765
Tested-by: build bot (Jenkins)
Reviewed-by: Kyösti Mälkki <kyosti.malkki@gmail.com>
The LLVM integrated assembler has some deficiencies in support for
building AGESA. See:
LLVM PR18918 - [RFE]: Missing altmacro support in integrated assembler
Disable llvm-mc for the moment until these have been addressed fully
upstream.
Change-Id: Id4131d1de04d01c0bec284f976f0ba9662b950ab
Signed-off-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-on: http://review.coreboot.org/5711
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Tested-by: build bot (Jenkins)
This might break a bunch of stuff (eg. win32 support),
but otherwise introduces nconfig (ncurses based configuration
frontend), partial configuration headers for improved dependency
tracking (which requires some more build system support) and
various bug fixes.
Change-Id: I5d8a280810c6a26fc3fd056d5d94cb9e591a0ff5
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5487
Tested-by: build bot (Jenkins)
Reviewed-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Make all three coreboot stages (bootblock, romstage and ramstage) aware of the
architecture specific to that stage i.e. we will have CONFIG_ARCH variables for
each of the three stages. This allows us to have an SOC with any combination of
architectures and thus every stage can be made to run on a completely different
architecture independent of others. Thus, bootblock can have an x86 arch whereas
romstage and ramstage can have arm32 and arm64 arch respectively. These stage
specific CONFIG_ARCH_ variables enable us to select the proper set of toolchain
and compiler flags for every stage.
These options can be considered as either arch or modes eg: x86 running in
different modes or ARM having different arch types (v4, v7, v8). We have got rid
of the original CONFIG_ARCH option completely as every stage can have any
architecture of its own. Thus, almost all the components of coreboot are
identified as being part of one of the three stages (bootblock, romstage or
ramstage). The components which cannot be classified as such e.g. smm, rmodules
can have their own compiler toolset which is for now set to *_i386. Hence, all
special classes are treated in a similar way and the compiler toolset is defined
using create_class_compiler defined in Makefile.
In order to meet these requirements, changes have been made to CC, LD, OBJCOPY
and family to add CC_bootblock, CC_romstage, CC_ramstage and similarly others.
Additionally, CC_x86_32 and CC_armv7 handle all the special classes. All the
toolsets are defined using create_class_compiler.
Few additional macros have been introduced to identify the class to be used at
various points, e.g.: CC_$(class) derives the $(class) part from the name of
the stage being compiled.
We have also got rid of COREBOOT_COMPILER, COREBOOT_ASSEMBLER and COREBOOT_LINKER
as they do not make any sense for coreboot as a whole. All these attributes are
associated with each of the stages.
Change-Id: I923f3d4fb097d21071030b104c372cc138c68c7b
Signed-off-by: Furquan Shaikh <furquan@google.com>
Reviewed-on: http://review.coreboot.org/5577
Tested-by: build bot (Jenkins)
Reviewed-by: Aaron Durbin <adurbin@gmail.com>
Under some circumstances the coreboot toolchain test prevented
building crossgcc, which is counter-productive: If a .config file
exists but no suitable .xcompile.
Don't assume anything about the tree when building crossgcc or
crosstools targets.
Change-Id: I4d6e7a88908dc967342daf30df0fcbcc269ae63d
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/5584
Tested-by: build bot (Jenkins)
Reviewed-by: Edward O'Callaghan <eocallaghan@alterapraxis.com>
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
This reverts commit 1287d1cc80.
This commit has the side-effect of making abuild fail, and as such is
reverted until a safe solution can be found.
Change-Id: Ib8cb78468c2922322b490e0b52c0bd24f3de7ef9
Signed-off-by: Alexandru Gagniuc <mr.nuke.me@gmail.com>
Reviewed-on: http://review.coreboot.org/3269
Tested-by: build bot (Jenkins)
Reviewed-by: Vladimir Serbinenko <phcoder@gmail.com>
If xcompile can't find out suitable GCC compiler for i386/armv7, it
will not set $CC_i386/$CC_armv7 variable. Makefile sets $CC variable
from xcompile, and will print strange error messages when executing
$CC program if $CC is empty.
Add checking to avoid this problem. If $CC is empty, also delete
invalid .xcompile file, so Make can recreate this file next time.
Change-Id: Ia8d481d76ca52f3351cb99f05779d06947161c5d
Signed-off-by: Andrew Wu <arw@dmp.com.tw>
Reviewed-on: http://review.coreboot.org/3905
Tested-by: build bot (Jenkins)
Reviewed-by: Jonathan A. Kollasch <jakllsch@kollasch.net>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
SeaBIOS’ Makefile requires cpp (C Preprocessor) to build. Modify
the xcompile script to search for cpp program path, and pass it to
SeaBIOS’ `Makefile.inc`. Also pass the program path for as (GNU assembler).
This is needed, so the crossgcc toolchain to build the SeaBIOS payload
under Mac OSX. OSX ships a cpp program, but it works differently
from GNU CPP, so we need to override it.
Change-Id: If996ffbb76ec4bd16079b54b41f3fac07bfe25be
Signed-off-by: Andrew Wu <arw@dmp.com.tw>
Reviewed-on: http://review.coreboot.org/3896
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Jonathan A. Kollasch <jakllsch@kollasch.net>
It looks like the inclusion of dependency files was broken for all
ramstage objects since the list of those gets processed through the
ramstage-postprocess macro. Fix that by taking the unaltered list
for dependency files.
The output of `make printall` (look for DEPENDENCIES=) shows which
dependency files will be included.
See also:
commit 79f9010e80
Author: Patrick Georgi <patrick@georgi-clan.de>
Date: Sun Nov 25 14:31:08 2012 +0100
build system: Add hook to postprocess classes (object lists)
and:
commit f33e395213
Author: Patrick Georgi <patrick@georgi-clan.de>
Date: Sun Nov 25 17:10:47 2012 +0100
build system: Split linking into multiple steps
Change-Id: If93b1773c5d53240f98382aab11bf7f5a4649ee8
Signed-off-by: Nico Huber <nico.huber@secunet.com>
Reviewed-on: http://review.coreboot.org/3258
Tested-by: build bot (Jenkins)
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
REQUIRES_BLOB assumes that all blob files come from the 3rdparty directory,
builds failed when all files were configured to point to other sources.
This change modifies the blob mechanism so that cbfs-files can be tagged as
"required" with some specification what is missing.
If the configured files can't be found (wrong path, missing file), the build
system returns a list of descriptions, then aborts.
Change-Id: Icc128e3afcee8acf49bff9409b93af7769db3517
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/2418
Tested-by: build bot (Jenkins)
Reviewed-by: Martin Roth <martin.roth@se-eng.com>
Reviewed-by: Paul Menzel <paulepanter@users.sourceforge.net>
Reviewed-by: Marc Jones <marcj303@gmail.com>
When running "make clean" the build system used to parse the entire
build system. Besides costing time, it prevents cleaning the tree
if a blobs-board is selected but blobs are not enabled.
Instead, clean always removes all of $(obj) and .xcompile, while
distclean additionally removes .config and the like.
Besides cleaning up more completely (eg. dependency files), a side
effect is that this also removes $(obj)/util, if it exists
(default location for build tools).
Change-Id: Ief6362460d4eb7edcb4b0a47ec76cb9a61bf3b86
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/2338
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
build system hygiene, not known if this actually matters.
Change-Id: Ic800a2acecff123fc2055047fab67df107ac43ab
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/2356
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: build bot (Jenkins)
Adding support for the multiple architectures broke the scan-build
option. The new CC setting needed to be wrapped and not run again
when doing the scan-build second pass.
Change-Id: Ieb418f51d44803308040926a4154fb5fdc3ba67f
Signed-off-by: Martin Roth <martin@se-eng.com>
Reviewed-on: http://review.coreboot.org/2031
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
The ARM architecture is really many architectures, and most
of them need their own toolchain. After discussing with Ron
and David, we decided that we're going to call the architecture
of our ARM board armv7.. This patch cleans out the remainders
of ARCH_ARM in the tree and moves on to consistent ARMv7 naming.
As of now, we only support little endian ARMv7 CPUs. We can
fix that for big endian if/when it comes our way.
Change-Id: Id70c7ef615f600e4d09961d811e7ac974fce4811
Signed-off-by: Stefan Reinauer <reinauer@google.com>
Reviewed-on: http://review.coreboot.org/1968
Tested-by: build bot (Jenkins)
Reviewed-by: Ronald G. Minnich <rminnich@gmail.com>
This will be used to minimize the ramstage class, to avoid command
line lengths to exceed the limit on mingw, esp. after we got rid
of ramstage.a
Change-Id: I80582d04476545c275e8d1d08fb52a99f58cebcc
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/1910
Tested-by: build bot (Jenkins)
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
With this change the the xcompile script now creates environment variables
for more than one architecture.
Signed-off-by: David Hendricks <dhendrix@chromium.org>
Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Signed-off-by: Stefan Reinauer <stepan@coresystems.de>
Change-Id: I349a1fd1d865ef16979f1dfd6aeca12b1ee2eed6
Signed-off-by: Ronald G. Minnich <rminnich@gmail.com>
Reviewed-on: http://review.coreboot.org/1915
Reviewed-by: Stefan Reinauer <stefan.reinauer@coreboot.org>
Tested-by: build bot (Jenkins)
make distclean causes error on mingw:
-------
rm: cannot lstat `build/util': Permission denied
make: *** [distclean] Error 1
-------
Guess, When the distclean is made by multi-process, the mkdir
in the Makefile will execute when build is removed. That causes
conflicts.
Change-Id: Ia41ecc5d1db2fa9d3328c81ac1d33fa94779492d
Signed-off-by: Zheng Bao <zheng.bao@amd.com>
Signed-off-by: Zheng Bao <fishbaozi@gmail.com>
Reviewed-on: http://review.coreboot.org/1602
Tested-by: build bot (Jenkins)
Reviewed-by: Patrick Georgi <patrick@georgi-clan.de>
This reverts commit 645f2dd5d9.
Instead of adding a special case to nvramtool to avoid it
picking up Kconfig's regex.h, have the host compiler only
consider util/kconfig for includes (ie. -Iutil/kconfig)
for kconfig related object files.
Change-Id: Ie4f97ce38cb3e911f6e6c1e5b6f86f6998d93f69
Signed-off-by: Patrick Georgi <patrick@georgi-clan.de>
Reviewed-on: http://review.coreboot.org/1509
Tested-by: build bot (Jenkins)
Reviewed-by: Zheng Bao <zheng.bao@amd.com>
Reviewed-by: Anton Kochkov <anton.kochkov@gmail.com>
Patrick Georgi