This change allows one to build the TF-A (Trusted Firmware A) for
32bit architectures Armv7 and Armv8 platforms. TF-A supports these
configuration since release v1.5.
Change dependency on BR2_aarch64 into a dependency on either
BR2_ARM_CPU_ARMV8A/BR2_ARM_CPU_ARMV7A to better define
the target architectures supported by the package.
Set package configuration directive ARM_ARCH_MAJOR according to
BR2_ARM_CPU_ARMV8A/BR2_ARM_CPU_ARMV7A. Set package configuration
directive ARCH based on BR2_aarch64/BR2_arm.
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This upgrades to version 48 plus the last commits done after that.
Note, that after that, gummiboot is officially dead, so there will
be no more upstream commits after this.
For various reasons, there are still users of gummiboot, so let's base our
work on the latest gummiboot "version".
Signed-off-by: Esben Haabendal <esben@haabendal.dk>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds a new variable OPTEE_OS_SDK, define by the optee-os
package, which points to the OPTEE SDK installed in STAGING_DIR. This
will allow other pakage to find this SDK more easily.
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
[Thomas: improve commit log, rename variable to just OPTEE_OS_SDK]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Commit 1a437fd22f bumped gnu-efi to version 3.0.9. This breaks the build
of syslinux with EFI support due to multiple definitions of 'memset' and
'memcpy'. Backport a patch already applied upstream to fix the problem.
Signed-off-by: Carlos Santos <casantos@datacom.com.br>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
OP-TEE OS is maintained by the OP-TEE project. It provides an
open source solution for development and integration of secure
services for Armv7-A and Armv8-A CPU based platforms supporting
the TrustZone technology. This technology enables CPUs to
concurrently host a secure world as the OP-TEE OS and a non-secure
world as a Linux based OS.
The OP-TEE project maintains other packages to leverage OP-TEE on
Linux kernel based OSes. An OP-TEE interface driver is available
in the Linux kernel since 4.12 upon CONFIG_OPTEE.
This change references in Buildroot the today's latest OP-TEE
revision release tagged 3.4.0.
https://www.op-tee.org/https://github.com/OP-TEE/optee_os
Signed-off-by: Etienne Carriere <etienne.carriere@linaro.org>
[Thomas:
- pass CFG_ARM32_core=y when building for ARMv7, otherwise the build
fails
- add a check that verifies that BR2_TARGET_OPTEE_OS_PLATFORM is not
empty
- minor formatting fixes/adjustements.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Fixes the following check-package warning:
boot/uboot/Config.in:185: attributes order: type, default, depends on, select, help (http://nightly.buildroot.org/#_config_files)
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Rockchip's rk3328 encapsulates the BL31 part of ATP but it needs
image in ELF format. This is extension to
2c4809e608
This commit introduces a BR2_TARGET_UBOOT_NEEDS_ATF_BL31_ELF variable
to achieve this.
Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Allow user to choose specific, stable version from official ATF
repository
Signed-off-by: Michał Łyszczek <michal.lyszczek@bofc.pl>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
At kconfig time, dependencies are not built, and therefore host-ccache
is not ready. Due to this, using $(HOSTCC) as the host compiler in
KCONFIG_OPTS does not work: a "make uboot-menuconfig" invocation from
a clean tree with ccache enabled fails.
This commit fixes this by using $(HOSTCC_NOCCACHE). We cannot rely on
the default value of HOSTCC passed by the kconfig-package
infrastructure, because $(UBOOT_MAKE_OPTS) also contains a HOSTCC
definition that would override the one passed by the kconfig-package
infrastructure.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This commit adds a package for 'shim', an EFI bootloader for secure
boot chain loading.
While gnu-efi supports 32bit ARM, this is currently broken in shim.
Patches to fix this have been submitted upstream but are not included
here for now.
https://github.com/rhboot/shim/pull/162
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
[Thomas: use BR2_PACKAGE_GNU_EFI_ARCH_SUPPORTS, add separate depends
on to exclude ARM32 build.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This will be used in packages that depend on gnu-efi, and we take this
opportunity to propagate this dependency where it was missing in
gummiboot and syslinux. In practice, it was not a problem because
gummiboot and syslinux are only available on i386 and x86-64, which is
a subset of the architectures supported by gnu-efi.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Following d0f4f95e39 ("Makefile: rework
main directory creation logic"), BINARIES_DIR is not implicitly
created by the main Makefile at the beginning of the build, leaving
that up to whatever piece of code needs to install something in
$(BINARIES_DIR).
The afboot-stm32 package didn't pay attention to this, which this
commit fixes.
While at it, we move the afboot-stm32 installation into
<pkg>_INSTALL_IMAGES_CMDS, because using <pkg>_INSTALL_TARGET_CMDS to
install only files to BINARIES_DIR is a bit strange.
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/131217111
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Acked-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Add notes to test grub2 running on ARM using qemu. The arm section
describes how to run it using u-boot and aarch64 shows how to do it
using efi, which is similar to what has to be done for x86_64.
The source for OVMF builds is also changed to
https://www.kraxel.org/repos/jenkins/edk2/ which is the source for
nightly builds (as rpms but which can be extracted in any distribution),
as the sourceforge link provided only very old builds.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
[Thomas:
- formatting fixes
- simplify the AArch64/EFI example by using the aarch64_efi_defconfig]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This commit enables the arm-uboot, arm-efi and aarch64-efi grub2
platforms in Buildroot.
With the uboot platform, the grub2 image gets built as a u-boot image
and is loaded from u-boot through a regular "bootm". The only
requirement from the u-boot side in order to allow this is that u-boot
is built with CONFIG_API enabled. CONFIG_API seems to not be enabled
by default in most in-tree configurations, however, it seems to be
available for quite some time now. So it might be possible to use this
even on older u-boot versions. This is available only for arm
(32-bit).
With the efi platform, grub2 gets built as an EFI executable. This
allows EFI firmware to find and load it similarly as it can be done
for x86_64. Also, since u-boot v2016.05, u-boot is able to load and
boot an EFI executable, so the uboot efi platform can also be used
from u-boot in recent versions. This has been enabled (mostly) by
default for ARM u-boot. efi platform is available for both arm and
aarch64.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
[Thomas: move the BR2_USE_MMU dependency in
BR2_TARGET_GRUB2_ARCH_SUPPORTS]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add an option to install grub2 support tools to the target.
In the context of Buildroot, some useful target tools provided are
grub2-editenv, grub2-reboot, which provide means to manage the grub2,
environment, boot order, and others.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
grub2 requires the host grub2-mkimage tool to build some of its target
images. The current way of building this tool in the grub2 package is
to perform a simultaneous host-tools/target-bootloader build during
the grub2 build step.
This method makes the recipe complex to understand, and proved to be a
complication during the work to enable grub2 support for architectures
other than x86.
This patch tries to do a better separation between the build of grub2
host tools and target boot loader image, as a partial step to enable
grub2 to build for other architectures.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When building with glibc 2.28+, the "major", "minor" and "makedev"
functions are defined in "sys/sysmacros.h". This commit backports
upstream commit 1a74985b2a404639b08882c57f3147229605dfd5 to fix the
build with glibc 2.28.
Link: https://www.syslinux.org/wiki/index.php?title=Building
Signed-off-by: Alexander Sverdlin <alexander.sverdlin@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
We already use $(MKIMAGE_ARCH) in lots of places; use it here too.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
We already use $(MKIMAGE) instead of $(HOST_DIR)/bin/mkimage in xvisor, linux,
and cpio; use it here too.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
We already use $(MKIMAGE) instead of $(HOST_DIR)/bin/mkimage in xvisor, linux,
and cpio; use it here too.
Signed-off-by: Danomi Manchego <danomimanchego123@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
U-boot has provided SPL support as a viable replacement for the
2nd stage bootloader on TI platforms. The X-loader project
hasn't had a commit in the log since 12-05-2011.
(https://gitorious.org/x-loader/x-loader)
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
As reported by Jeff Wittrock in bug #11396, the U-Boot environment
image checksum is invalid for big endian targets, because the test on
the BR2_ENDIAN Config.in option doesn't take into account that it is
double quoted.
The fix was provided by Jeff himself on bugzilla.
Fixes bug #11396.
Reported-by: Jeff Wittrock <jwittrock@faultrecorder.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Passing CFLAGS/CPPFLAGS in the environment confuses the ATF version
provided by Marvell, and we in fact only need to pass CROSS_COMPILE,
which is already in ARM_TRUSTED_FIRMWARE_MAKE_OPTS.
We however keep TARGET_MAKE_ENV so that the PATH with $(HOST_DIR) is
kept.
This change has been tested with all current defconfigs that build
ATF:
- arm_juno
- bananapi_m64
- freescale_imx8mqevk
- friendlyarm_nanopi_a64
- friendlyarm_nanopi_neo2
- nitrogen8m
- olimex_a64_olinuxino
- orangepi_pc2
- orangepi_prime
- orangepi_win
- orangepi_zero_plus2
- pine64
- pine64_sopine
- solidrun_macchiatobin_mainline
- solidrun_macchiatobin_marvell
- zynqmp_zcu106
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Note that the license file has been entirely rewritten (hence the
change in the hash), but it is still GPL-2.0 with FreeRTOS exception.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
This is necessary to be able to bump the ARM Trusted Firmware version
used on Marvell platforms.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add the RISC-V Proxy Kernel (pk) package which provides the Berkeley
Boot Loader for booting RISC-V kernel images.
Signed-off-by: Mark Corbin <mark.corbin@embecosm.com>
[Thomas:
- fix prompt of option to be just "riscv-pk"
- add blank line between help text and upstream URL in Config.in file
- use $(INSTALL) instead of cp to install the generated image]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When using the legacy buildsystem, the kconfig parser may still be used,
as reported by Thomas:
>>> uboot 2018.09 Configuring
...]
HOSTCC scripts/basic/fixdep
HOSTCC scripts/kconfig/conf.o
YACC scripts/kconfig/zconf.tab.c
bin/sh: 1: bison: not found
ake[3]: *** [scripts/kconfig/zconf.tab.c] Error 127
ake[3]: *** Waiting for unfinished jobs....
LEX scripts/kconfig/zconf.lex.c
bin/sh: 1: flex: not found
However, in that case, the kconfig parser is only generated during the
'configure' step, so we can add bison/flex as standard dependencies.
Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
U-Boot fails to build in a GitLab CI context because the kconfig-package
build stage is unable to find bison or flex even though they are
installed in HOST_DIR.
To fix this, set UBOOT_MAKE_ENV so that UBOOT_KCONFIG_MAKE uses the
correct PATH.
Signed-off-by: Thomas Preston <thomas.preston@codethink.co.uk>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Recent U-Boot no longer ship the flex/bison generated kconfig parser, as
of commit e91610da7c8a9fe42f3e5a75f06c3d1a0cb5f815 (kconfig: re-sync
with Linux 4.17-rc4).
So, add the conditional kconfig dependencies, as we just did for the
kernel.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Cc: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Some U-Boot configurations use lzop, which may not be available on the
build machine since it's not a mandatory dependency of Buildroot.
To solve this, a new option BR2_TARGET_UBOOT_NEEDS_LZOP is introduced,
and handled in a similar way to the existing
BR2_TARGET_UBOOT_NEEDS_OPENSSL option.
This fixes the build of the dra7xx_evm U-Boot defconfig as of U-Boot
2018.07, on a build machine without lzop installed.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
U-Boot SPL configures pinmuxes, clocks and other low-level devices. On
the Xilinx ZynqMP SoCs the code to do this resides in a file called
psu_init_gpl.c which is initially generated by the Xilinx development
tools. Add an option to pass these files from the outside (e.g. in the
board files).
For this to work properly, a patch to U-Boot is needed. However this
patch must be applied by each defconfig using
BR2_TARGET_UBOOT_ZYNQMP_PSU_INIT_DIR. If it were in boot/uboot/ to be
applied unconditionally, it would break the build for configs using a
U-Boot version where the patch is already applied.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Joel Carlson <joelsoncarl@gmail.com>
Reviewed-by: Joel Carlson <JoelsonCarl@gmail.com>
Tested-by: Joel Carlson <JoelsonCarl@gmail.com>
[Thomas: add comment about U-Boot version requirements to use this
option.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
In order to boot on the Xilinx ZynqMP SoCs, U-Boot SPL requires a
recent PMU firmware loaded. Instruct U-Boot to add pmufw.bin to the
boot.bin file together with U-Boot SPL, and the boot ROM will load
both.
For this to work properly, a patch to U-Boot is needed. However this
patch must be applied by each defconfig that wishes to use
BR2_TARGET_UBOOT_ZYNQMP_PMUFW. If it were in boot/uboot/ to be applied
unconditionally, it would break the build for configs using a U-Boot
version higher than 2018.7-rc1 where the patch is already applied.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Joel Carlson <joelsoncarl@gmail.com>
Reviewed-by: Joel Carlson <JoelsonCarl@gmail.com>
Tested-by: Joel Carlson <JoelsonCarl@gmail.com>
[Thomas:
- indicate that this feature requires U-Boot 2018.07, since commit
c7df098a71e05dc81cee818747759e8060b59626 is needed.
- define UBOOT_ZYNQMP_KCONFIG_PMUFW only once, and instead use a
variable named UBOOT_ZYNQMP_PMUFW_PATH to store the path to the PMU
firmware]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Some i.MX boards which use device tree to describe the hardware
for U-Boot generate a u-boot-dtb.imx binary.
Introduce the BR2_TARGET_UBOOT_FORMAT_DTB_IMX option to
handle such case.
Signed-off-by: Fabio Estevam <festevam@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
So the host ncurses includes and library are used instead of a mix of both,
causing corrupted characters. Similar to the linux fix in commit
6d3d09e232 (linux: don't override HOSTCC for kconfig), except that we
pass the linker flags in HOSTLDFLAGS.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Acked-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
U-Boot SPL for the Xilinx ZynqMP SoCs needs ATF in this format to load
it.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The path to the binary images is very long. Since we are about to make
a larger use of it, let's use a variable to make it somewhat shorter.
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
By default, the builsystem for uboot defaults to 'legacy', while the
default version is very well capable of using the 'kconfig' buildsystem
instead.
Having the 'legacy' buildsystem be the default in that case makes it
quite inconvenient for users: they would expect to be able to use e.g.
uboot-menuconfig et al. with the default uboot version.
Switch to using 'kconfig' when we use the latest version. Keep the
'legacy' as default for everything else.
Also, invert the 'legacy' and 'kconfig' entries in the choice: it is
nicer to have the recent and future-proof entry first.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Tested-by: Petr Vorel <petr.vorel@gmail.com>
Reviewed-by: Yegor Yefremov <yegorslists@googlemail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
In commit 2a27294e9a ("grub2: force
-fno-stack-protector in CFLAGS"), a fix was made to the grub2 package
to make it build properly even when SSP support is enabled.
However, commit 20a4583ebf ("security
hardening: add RELFO, FORTIFY options") reworked how SSP options are
passed, and they are now passed in CPPFLAGS instead of CFLAGS, making
the fix introduced by 2a27294e9a no
longer operating.
This commit will force no-stack-protector in CPPFLAGS instead of
CFLAGS.
Fixes bug #10961.
Signed-off-by: Tarek El-Sherbiny <tarek_el-sherbiny@waters.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Instead of DL_DIR, the package should now use $(PKG)_DL_DIR to ease the
transition into a new directory structure for DL_DIR.
This commit has been generated with the following scripts:
for i in $(find . -iname "*.mk"); do
if ! grep -q "\$(DL_DIR)" ${i}; then
continue
fi
pkg_name="$(basename $(dirname ${i}))"
[ "${pkg_name}" = "package" ] && continue
raw_pkg_name=$(echo ${pkg_name} | tr [a-z] [A-Z] | tr '-' '_')
pkg_dl_dir="${raw_pkg_name}_DL_DIR"
sed -i "s/\$(DL_DIR)/\$($pkg_dl_dir)/" ${i}
done
Signed-off-by: Maxime Hadjinlian <maxime.hadjinlian@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Like the utilities, it is meant to run on the host machine, hence must
be built using the host toolchain.
Signed-off-by: Carlos Santos <casantos@datacom.ind.br>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
... to follow the convention: type, default, depends on, select, help.
Signed-off-by: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Use only one space before backslash.
Indent with tabs.
Signed-off-by: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The change in commit bf73334232 only works
for newer uboot source trees. Add a check that scripts/dtc/libfdt
directory exists before making this change.
[Peter: add comment explaining why]
Signed-off-by: Stefan Becker <chemobejk@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
When libfdt-devel is installed system-wide, the build of U-Boot host
programs currently fails with lots of:
HOSTCC tools/aisimage.o
In file included from tools/../include/../lib/libfdt/libfdt.h:10:0,
from tools/../include/libfdt.h:1,
from tools/fdt_host.h:11,
from tools/imagetool.h:24,
from tools/aisimage.c:8:
/usr/include/libfdt_env.h:70:30: error: conflicting types for ‘fdt64_t’
typedef uint64_t FDT_BITWISE fdt64_t;
^~~~~~~
In file included from <command-line>:0:0:
././include/libfdt_env.h:19:16: note: previous declaration of ‘fdt64_t’ was here
typedef __be64 fdt64_t;
^~~~~~~
In file included from ././include/libfdt_env.h:12:0,
from <command-line>:0:
/usr/include/libfdt_env.h:90:24: error: expected ‘)’ before ‘x’
static inline uint32_t fdt32_to_cpu(fdt32_t x)
^
This commit adds a fixup in the U-Boot code to fix this problem. The
fixup is equivalent to applying upstream commit
http://git.denx.de/?p=u-boot.git;a=commitdiff;h=e0d20dc1521e74b82dbd69be53a048847798a90a,
but we can't use a patch for the uboot package, since people are using
arbitrary versions.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Commit baae5156ce (uboot: use local fdt headers) changed the uboot logic
to pass the host include directories with -idirafter instead of -I, so
include files local to u-boot would be preferred over host includes.
This unfortunately breaks configurations using
BR2_TARGET_UBOOT_NEEDS_OPENSSL on hosts with incompatible openssl headers
installed in the system default include directories as explained here:
http://lists.busybox.net/pipermail/buildroot/2018-March/214651.html
The problem is that -idirafter directories gets added to the very end of the
search order, AFTER the system default directories.
Instead use -isystem which causes the directories to be added after -I but
before the system default directories. With this in place, the include
directories of u-boot will first be scanned, followed by the host includes
and finally the system default include directories.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This reverts commit 3a6573ccee.
It is no longer necessary after solving the problem differently.
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
After commit b8c3e94173 ("package/dtc: qemu system build need libfdt")
changed the dtc install target from 'install-bin' to 'install', uboot
compilation failures occurred because libfdt related headers were now
suddenly taken from output/host/include rather than from the uboot sources
itself.
Commit 3a6573ccee ("uboot: use local libfdt.h") solved this by patching
one specific uboot source file, tools/fdtgrep.c, to replace '<...>'-style
includes by '"..."'-style includes.
However, depending on the uboot version, this may not be enough: there may
be other references to fdt header files. In particular taking into account
that it is not uncommon to have vendor-provided uboot trees which have
custom changes.
The root of the problem is that the uboot.mk file passes the host compiler
as follows:
UBOOT_MAKE_OPTS += \
...
HOSTCC="$(HOSTCC) $(HOST_CFLAGS)" \
...
where HOST_CFLAGS contains the string '-I$(HOST_DIR)/include'
The uboot makefiles then use constructs of the form:
$(CC) $(CPPFLAGS) $(CFLAGS) .....
where CPPFLAGS may contain -I references pointing to local directories.
On the expanded compiler command-line, Buildroot's '-I$(HOST_DIR)/include'
is thus present _before_ any -I to local directories, and thus takes
precedence. And that becomes a problem for header files present both
locally as in the Buildroot host directory, which is the case for libfdt.
To fix this problem without having to patch u-boot sources, use '-idirafter'
rather than '-I' to pass the Buildroot host include directory. '-idirafter'
is basically the same thing, but adds the specified directory at the end
of the include precedence chain, rather than at the beginning.
Signed-off-by: Thomas De Schampheleire <thomas.de_schampheleire@nokia.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
ATF >= 1.3 builds a host program called fiptool which uses
OpenSSL, so we need to build host-openssl. We could have made it an
optional dependency like U-Boot does, but since most ATF versions are
going to be >= 1.3 in the near future, we simply make host-openssl a
mandatory dependency.
However, the ATF build system is not very good, and you can't easily
pass flags that will affect the build of host programs. Therefore, we
take the approach of building fiptool separately before triggering the
real build process.
It would obviously be better to fix ATF itself, but as usual with
those bootloader packages, we fetch different versions depending on
the platform/configuration, making it difficult to use patches.
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/44868961
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since 0542bb79e8 (uboot: Support multiple environment source files),
missing user-supplied environment source files is no longer detected.
This is because we cat them all, and feed the concatenation to the stdin
of mkenvimage. So, if one source file is missing, the cat exits in error,
but the compound command exits with the exit code of the last command,
which is that of mkenvimage, which happens to be happy with whatever it
is fed on its stdin, even is empty.
We fix that by creating a temporary file, that we even leave afterward
for the user to inspect.
We also move it out of the _CMDS block and into a macro of its own, so
that it is easier to write and maintain.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Cam Hutchison <camh@xdna.net>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add Marvell Armada 7k/8k SoCs support to arm-trusted-firmware package.
Marvell ATF needs two additional dependencies:
DDR training code and SCP_BL2 image.
Signed-off-by: Sergey Matyukevich <geomatsi@gmail.com>
[Thomas: adjust to previous ATF changes.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Following our feedback, Marvell has added a README.md file into the
branch that contains the firmware. Thereore, this commit bumps the
version to the commit that includes the README.md file (it's the only
change, the firmware files are unmodified), updates
<pkg>_LICENSE_FILES to point to README.md, and adds the hash for this
license files.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some systems, including Marvell Armada 7k/8k SoCs, have a separate
System Control Processor (SCP) for power management, clocks, reset
and system control. ATF Boot Loader stage 2 (BL2) loads optional
SCP_BL2 image into a platform-specific region of secure memory.
This package adds SCP_BL2 firmware for Marvell Armada 7040 and 8040 SoCs.
Signed-off-by: Sergey Matyukevich <geomatsi@gmail.com>
[Thomas:
- Use "binaries-marvell" in the Config.in prompt and in the .mk file
header.
- Change the license information.
- Adjust license information: it is GPL-2.0 with the FreeRTOS
exception, and therefore can be redistributed.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
This package adds Marvell Armada SoC DDR training algorithms.
This code is not built separately, it is needed as dependency
to build ATF firmware for Marvell Armada SoCs.
Signed-off-by: Sergey Matyukevich <geomatsi@gmail.com>
[Thomas:
- Remove MV_DDR_MARVELL_SRC_SYMLINK, ATF will directly pick up the
mv-ddr-marvell code from where it is.
- Use "mv-ddr-marvell" as the Config.in prompt, and in the comment
header of the .mk file.
- Add upstream URL in Config.in help text
- Adjust license information.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some ARM64 platforms (such as Allwinner A64/H5) have a boot process
where U-Boot encapsulates the BL31 part of the ARM trusted
firmware. For such platforms, we need to build ATF before U-Boot, and
pass a BL31 variable pointing to ATF bl31.bin to the U-Boot build
process.
This commit introduces a BR2_TARGET_UBOOT_NEEDS_ATF_BL31 variable to
achieve this.
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
[Thomas:
- Rename option to BR2_TARGET_UBOOT_NEEDS_ATF_BL31
- Drop changes to arm-trusted-firmware.mk, they are taken care of by
previous commits.
- Improve Config.in help text
- Add missing dependency on arm-trusted-firmware when
BR2_TARGET_UBOOT_NEEDS_ATF_BL31 is enabled.
- Use bl31.bin from $(BINARIES_DIR) instead of taking it from ATF's
build dir.]
[Peter: depend on toplevel atf option and select bl31 option. Ensure it
cannot be enabled together with ARM_TRUSTED_FIRMWARE_UBOOT_AS_BL33
as that would cause circular dependencies]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Some platforms (e.g Allwinner ARM64) don't build a FIP image out of
ATF, but only a BL31 image, which is then encapsulated in U-Boot. This
commit adds an ATF option to build such an image.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Tested-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
ATF can be used in different ways:
- ATF encapsulates U-Boot as the BL33, which is what is done on ARM
Juno (currently supported in Buildroot) and Marvell platforms (soon
to be supported)
- U-Boot encapsulates ATF's BL31, which is what is done on Allwinner
ARM64 platforms.
Until now we were assumming the former was always the case, but
obviously it isn't. Therefore, this patch adds an option that allows
to explicitly tell ATF that it encapsulates U-Boot as its BL33.
We adapt the only defconfig that uses ATF so that it enables this
option as appropriate.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Currently, our arm-trusted-firmware unconditionally builds a FIP
(Firmware Image Package). While this is often needed on platforms
where ATF encapsulates U-Boot, it is not the case on some other
platforms where it's U-Boot that encapsulates parts of ATF.
In order to prepare the support for the later platforms, we make
building the FIP image optional, and update the only defconfig we have
that uses ARM Trusted Firmware.
Note: we considered adding a "default y" here to preserve backward
compatibility, but there really isn't any default that is sane:
whether a FIP image needs to be built or not is purely platform
specific.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Reviewed-by: Sergey Matyukevich <geomatsi@gmail.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Since the BR2_TARGET_UBOOT_SPL_NAME option accepts a space-separated
list of binaries, the same option can be reuses for TPL binaries as
well. This commit updates the string and help text to indicate that
the same option can be used for SPL and TPL.
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Use of libfdt.h by u-boot is currently using
the <system-path>.
Jan submitted the following upstream
https://patchwork.ozlabs.org/patch/833760/
Resolves a failure like this one observed on uboot-tools
http://autobuild.buildroot.net/results/347cde4b5c0e6ca76d354396385be4ec1294da73
[Peter: only sed if file exists]
CC: Jan Kundrát <jan.kundrat@cesnet.cz>
CC: Arnout Vandecappelle <arnout@mind.be>
Signed-off-by: Matthew Weber <matthew.weber@rockwellcollins.com>
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
The license file got reformatted as reStructuredText, but the license itself
didn't change.
Drop unneeded md5sum and add license hash.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
U-Boot nowadays also uses kconfig, so we can handle config fragments like we
do for barebox/busybox/linux.
Signed-off-by: Peter Korsgaard <peter@korsgaard.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
grub is no longer maintained: it is stuck at version 0.97 with huge
patches that have no opportunity to be applied upstream, as upstream
has even renamed it grub-legacy.
Besides, it no longer builds correctly with recent binutils versions,
and even the huge patches we could grab from Debian do not help the
slightest.
Since upstream really considers it dead, and there are at least two
alternatives (grub2 and syslinux), just remove grub.
Add a legacy entry.
Remove the test cases as well.
Signed-off-by: "Yann E. MORIN" <yann.morin.1998@free.fr>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Cc: Ricardo Martincoski <ricardo.martincoski@gmail.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
grub2 fails to configure when BR2_SSP_ALL is enabled, with the following
configure error:
checking whether -fno-asynchronous-unwind-tables works... yes
checking whether -fno-unwind-tables works... yes
checking for target linking format... unknown
configure: error: no suitable link format found
This can be worked around by enforcing -fno-stack-protector in the
package CFLAGS in a way that overrides the SSP flag, as is already done
for the valgrind package.
Fixes bug #10261.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Reported-by: Dr I J Ormshaw <ian_ormshaw@waters.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
When we use the cross-compiler to build syslinux with a recent binutils
version, it fails with:
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld -Bsymbolic -pie -E --hash-style=gnu -T
/builds/arnout/buildroot/output/build/syslinux-6.03/core/i386/syslinux.ld -M -o ldlinux.elf ldlinux.o \
--start-group libcom32.a --whole-archive /builds/arnout/buildroot/output/build/syslinux-6.03/bios/com32/lib/libcom32core.a libldlinux.a --end-group -N
--no-omagic \
> ldlinux.map
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld: ldlinux.elf: Not enough room for program headers, try linking with -N
/builds/arnout/buildroot/output/host/bin/i586-buildroot-linux-uclibc-ld: final link failed: Bad value
/builds/arnout/buildroot/output/build/syslinux-6.03/core/Makefile:167: recipe for target 'ldlinux.elf' failed
Backport an upstream patch that reorganises the i386 bios build by
removing some symbols and making others hidden. To simplify the
backport, an additional patch that also touches the link script is
also included - it anyway looks like that patch could be relevant as
well.
Partially fixes: https://gitlab.com/arnout/buildroot/-/jobs/28979377
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Also the last two are regenerated, their context has changed due to
the patches that have been removed earlier.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Add a new config option BR2_TARGET_UBOOT_NEEDS_PYLIBFDT for U-Boot
targets, like sunxi, needing the Python libfdt library.
Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
[Thomas: drop bogus selects on BR2_PACKAGE_HOST_{PYTHON,SWIG}.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
As discussed in the mailing list, grub2 usage notes were growing too big
for a Config.in documentation, and so it was agreed that a readme.txt in
the package directory is a better place to put them.
This commit simply moves the documentation as-is to preserve the
original contents as they were in Config.in which can be worked on in
further commits.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
After many years since the last release and a long time with grub 2.02
in beta, there is finally a release and it brings many bug fixes and
interesting features such as support for ARM.
Patch boot/grub2/0001-remove-gets.patch doesn't seem to be required
anymore as grub-core/gnulib/stdio.in.h has changed significantly since
"053cfcd Import new gnulib." and has another treatment for gets.
Patch
boot/grub2/0002-grub-core-gettext-gettext.c-main_context-secondary_c.patch
was a backport which is present after the bump and therefore is also no
longer necessary.
Since we're adding a Config.in comment, we also introduce a
BR2_TARGET_GRUB2_ARCH_SUPPORTS hidden boolean, in order to avoid
repeating the architecture dependencies.
Signed-off-by: Erico Nunes <nunes.erico@gmail.com>
Cc: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
[Thomas: add BR2_TARGET_GRUB2_ARCH_SUPPORTS, remove bogus dependencies
on ARM and AArch64, since enabling Grub2 on those architectures is
done in another commit.]
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Xtensa core configuration must be added to U-Boot before it can be
built for that xtensa CPU variant. Extract configuration files from the
xtensa overlay as is done for other packages that need to be configured
for a specific xtensa core.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
grub2 builds for the target but installs with DESTDIR=$(HOST_DIR). Since
we set prefix to /usr in TARGET_CONF_OPTS, this results in installing
things in $(HOST_DIR)/usr.
To make sure we don't install in $(HOST_DIR)/usr, override --prefix and
--exec-prefix.
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Since things are no longer installed in $(HOST_DIR)/usr, the callers
should also not refer to it.
This is a mechanical change with
git grep -l '$(HOST_DIR)/usr/share' | xargs sed -i 's%$(HOST_DIR)/usr/share%$(HOST_DIR)/share%g'
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
