LoongArch64 OVMF target is introduced to EDK2 in edk2-202408.
Enable architecture support here.
Signed-off-by: Jiaxun Yang <jiaxun.yang@flygoat.com>
Reviewed-by: Romain Naour <romain.naour@smile.fr>
Signed-off-by: Julien Olivain <ju.o@free.fr>
When building a firmware for the MACCHIATObin with edk2 and
arm-trusted-firmware, the build can randomly fail with the
following make error:
make[1]: Circular output/build/edk2-edk2-stable202411/.stamp_configured <- arm-trusted-firmware dependency dropped.
The message appears also when the build is not failing, depending on
the number of parallel jobs and the build order.
The issue can be observed with the following commands:
cat >.config <<EOF
BR2_aarch64=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_TARGET_ARM_TRUSTED_FIRMWARE=y
BR2_TARGET_ARM_TRUSTED_FIRMWARE_PLATFORM="a80x0_mcbin"
BR2_TARGET_ARM_TRUSTED_FIRMWARE_EDK2_AS_BL33=y
BR2_TARGET_BINARIES_MARVELL=y
BR2_TARGET_EDK2=y
BR2_TARGET_EDK2_PLATFORM_SOLIDRUN_ARMADA80X0MCBIN=y
BR2_TARGET_MV_DDR_MARVELL=y
EOF
make olddefconfig
utils/brmake
grep -FC5 'dependency dropped' br.log
The circular dependency happen due to [1] and [2].
In fact, only TF-A depends on EDK II (passed as BL33) for building and
not vice versa. See [3]. The EDK II "SolidRun MacchiatoBin" platform
build does not need any TF-A image, compared to some other platforms
such as "Socionext DeveloperBox" or "QEMU SBSA" which are referencing
TF-A images in a hook added in EDK2_PRE_BUILD_HOOKS.
Drop the false dependency on TF-A to fix the build.
This issue has been present since the EDK2 introduction in commit [4].
[1] https://gitlab.com/buildroot.org/buildroot/-/blob/2025.02/boot/arm-trusted-firmware/arm-trusted-firmware.mk#L121
[2] https://gitlab.com/buildroot.org/buildroot/-/blob/2025.02/boot/edk2/edk2.mk#L118
[3] https://github.com/Semihalf/edk2-platforms/wiki/Build_firmware
[4] https://gitlab.com/buildroot.org/buildroot/-/commit/1074a37e783b023bd0491ba4f4f7acab7f289e1d
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Dick Olsson <hi@senzilla.io>
[Julien: add extra info in commit log]
Signed-off-by: Julien Olivain <ju.o@free.fr>
Release notes:
https://github.com/tianocore/edk2/releases/tag/edk2-stable202405
We can't bump to edk2-stable202408 yet due to a runtime issue while
testing qemu_aarch64_sbsa_defconfig following the switch generic
ArmPsciResetSystemLib [1]
ERROR: sbsa_sip_smc_handler: unhandled SMC (0xc20000ca) (function id: 202)
Qemu 9.0/9.1 is using a slightly older edk2 firmware based on
edk2-stable202405 release [2] that contains a commit [3] fixing a
bootloader crash produced with qemu_aarch64_sbsa_defconfig [4] since
Qemu 9.0.
From [5]:
"The version of the sbsa-ref EDK2 firmware we used to use in this test
had a bug where it might make an unaligned access to the framebuffer,
which causes a guest crash on newer versions of QEMU where we enforce
the architectural requirement that unaligned accesses to Device memory
should take an exception."
This commit [5] is backported to edk2-stable202405.
For the same reason, we have to update edk2-platforms to a specific
version [6]:
"QemuSbsa: enable WriteCombine for the FrameBuffer
QEMU no longer permits misaligned access to device memory, which breaks
QemuVideoDxe on SbsaQemu.
c1d1910be6e04a8b1a73090cf2881fb698947a6e commit in EDK2 fixed it by
enabling WriteCombine for Framebuffer memory. This change enables that
fix."
As a side note, the edk2-non-osi package does not need to be updated, so
it is left untouched.
Finally, this EDK2 version requires a recent Qemu version. The version
included in the reference Docker image (5.2.0) is too old and will not
allow to start this EDK2 version. Therefore, this bump is known to be
incompatible with runtime tests which are using EDK2 on Aarch64
architecture. At the authoring time of this commit, broken tests are:
- tests.boot.test_edk2
- tests.boot.test_grub.TestGrubAArch64EFI
- tests.package.test_fwts
Note that the EDK2/Qemu incompatibility is only affecting Aarch64
tests. Other tests using EDK2 on i386 or x86_64 are not affected.
To prevent this commit introducing broken runtime tests, those have
been disabled in a previous commit.
[1] https://github.com/tianocore/edk2-platforms/commit/2d0668d399708fc242336e310cc7e7bec0aec891
[2] https://gitlab.com/qemu-project/qemu/-/commit/24a7cd6a7c21f82f1ce9bd5ecf6fb54eb7bf4602
[3] https://github.com/tianocore/edk2/commit/c1d1910be6e04a8b1a73090cf2881fb698947a6e
[4] https://gitlab.com/buildroot.org/buildroot/-/jobs/7764483764
[5] https://gitlab.com/qemu-project/qemu/-/commit/6c84daac58f0d75a908626faf89a832fc0532140
[6] https://github.com/tianocore/edk2-platforms/commit/3f08401365d67e10924c774e6c3f64be56bc15b6
Fixes:
https://gitlab.com/buildroot.org/buildroot/-/jobs/7764483764
Signed-off-by: Romain Naour <romain.naour@smile.fr>
Cc: Waldemar Brodkorb <wbx@openadk.org>
[Julien: add comments in commit log]
Signed-off-by: Julien Olivain <ju.o@free.fr>
Replace all git and svn packages archive names in hash files:
$ sed -r -i -e 's/-br2/-git3/; s/-br3/-svn4/' $(
git grep -l -E -- '-br2|-br3' '*.hash'
)
$ sed -r -i -e 's/(\.tar\.gz)$/-go1\1/' $(
git grep -l -E '\$\(eval \$\((host-)?golang-package\)\)' '*.mk' \
|sed -r -e 's/\.mk$/.hash/' \
|sort -u
)
$ sed -r -i -e 's/(\.tar\.gz)$/-cargo1\1/' $(
git grep -l -E '\$\(eval \$\((host-)?cargo-package\)\)' '*.mk' \
|sed -r -e 's/\.mk$/.hash/' \
|sort -u
)
Then a bit of make source (based on: git diff --name-only), a lot of
sweat, and carefully checking the new archives to verify that only
modes have changed...
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
The version suffix for the git-generated archives has changed, so update
the filenames accordingly in the hash files. The content of the archives
has not changed, though.
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Arnout Vandecappelle <arnout@mind.be>
Yann reported in [1] that edk2 build could sometimes fail. The issue
can be reproduced when per-package directories is enabled, or also
when building on a system with GNU Make >= 4.4 using the
"--shuffle=reverse" option (such as Fedora 39). Those are pointing
toward a Makefile dependency issue.
The issue can be reproduced with commands:
cat > .config <<EOF
BR2_riscv=y
BR2_RISCV_64=y
BR2_TOOLCHAIN_EXTERNAL=y
BR2_TARGET_EDK2=y
EOF
make olddefconfig
Then, building either with:
make --shuffle=reverse
Or:
utils/config -e BR2_PER_PACKAGE_DIRECTORIES
make olddefconfig
make -j$(nproc)
It is interesting to mention that when using "make --shuffle=reverse"
to build, the build can be completed if restarted only with "make". It
will not pull any other Buildroot package. This fact hints toward a
Makefile dependency issue internal to the EDK2 build system, rather
than in the Buildroot recipe.
The EDK2 build system is quite unique. See [2]. It generates files,
makefiles and internally uses GNU Make to compile code. This system is
likely not tested as being a sub-Make process in a complex Makefile
such as Buildroot.
In order to prevent Buildroot to pass unexpected Make flags to the
EDK2 sub-Make, this commit unset the MAKEFLAGS variable in the EDK2
build environment. This will put the EDK2 build script in a more
common and tested state. See GNU Make documentation about recursive use
of Make, more specifically [3].
Note: as mentioned, the build failure is likely due to an internal
issue of the EDK2 build system. The failure points to a missing
dependency in the EDK2 generator itself. This commit does not fix this
issue, but rather put the EDK2 build system in a normalized
environment, avoiding Buildroot flags being passed to the internal
EDK2 sub-Make invocation. The upstream EDK2 build system most likely
need a fix too.
Fixes:
make[2]: *** No rule to make target '/buildroot/output/build/edk2-edk2-stable202308/Build/RiscVVirtQemu/RELEASE_GCC5/RISCV64/MdeModulePkg/Application/UiApp/UiApp/DEBUG/UiApp.efi', needed by '/buildroot/output/build/edk2-edk2-stable202308/Build/RiscVVirtQemu/RELEASE_GCC5/FV/Ffs/462CAA21-7614-4503-836E-8AB6F4662331UiApp/UiApp.offset'. Stop.
build.py...
: error 7000: Failed to execute command
make tbuild [/buildroot/output/build/edk2-edk2-stable202308/Build/RiscVVirtQemu/RELEASE_GCC5/RISCV64/MdeModulePkg/Application/UiApp/UiApp]
build.py...
: error F002: Failed to build module
/buildroot/output/build/edk2-edk2-stable202308/MdeModulePkg/Application/UiApp/UiApp.inf [RISCV64, GCC5, RELEASE]
[1] https://lists.buildroot.org/pipermail/buildroot/2023-December/681507.html
[2] https://tianocore-docs.github.io/edk2-BuildSpecification/draft/4_edk_ii_build_process_overview/42_build_process_overview.html
[3] https://www.gnu.org/software/make/manual/make.html#Options_002fRecursion
Reported-by: Yann E. MORIN <yann.morin.1998@free.fr>
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
RISC-V 64bit qemu virt machine support has been added in edk2
version "stable202302". See [1].
Since edk2-stable202308, introduced in buildroot in commit 5c9f310
"boot/edk2: bump to version edk2-stable202308", it is now possible
to boot the edk2 UEFI shell in qemu.
This commit adds this early RISC-V support to edk2.
The RISC-V edk2 UEFI shell can be booted in Buildroot with the
following commands:
# Build EDK2 images
cat > .config <<EOF
BR2_riscv=y
BR2_RISCV_64=y
BR2_PACKAGE_HOST_QEMU=y
BR2_PACKAGE_HOST_QEMU_SYSTEM_MODE=y
BR2_TARGET_EDK2=y
EOF
make olddefconfig
make
# edk2 image size should fit the 32MB of qemu pflash memories
truncate -s 32M output/images/RISCV_VIRT_CODE.fd
truncate -s 32M output/images/RISCV_VIRT_VARS.fd
# Start qemu:
output/host/usr/bin/qemu-system-riscv64 \
-M virt,pflash0=pflash0,pflash1=pflash1,acpi=off \
-nographic \
-blockdev node-name=pflash0,driver=file,read-only=on,filename=output/images/RISCV_VIRT_CODE.fd \
-blockdev node-name=pflash1,driver=file,filename=output/images/RISCV_VIRT_VARS.fd
Note: a Qemu version >= 8.0.0 is needed to properly start edk2. A qemu
version on the host system might not be sufficient. This is why the
Buildroot host-qemu is built in this config example.
[1] https://github.com/tianocore/edk2/releases/tag/edk2-stable202302
Signed-off-by: Julien Olivain <ju.o@free.fr>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
When EDK2 OVMF is built with debug, messages are printed to
IO port 0x402. Those messages are not shown in the normal
Qemu emulated serial port. Enabling this option will print
debug messages on the emulated serial port, potentially
mixing messages with UEFI serial console output.
See OVMF README:
https://github.com/tianocore/edk2/blob/master/OvmfPkg/README
This option has been useful to debug problems such as [1], in which the
emulator was hanging at startup without any log. Enabling this option
show the debug message on the console, before hanging:
...
PlatformCpuCountBugCheck: Present=0 Possible=1
PlatformCpuCountBugCheck: Broken CPU hotplug register block found. Update QEMU to version 8+, or
PlatformCpuCountBugCheck: to a stable release with commit dab30fbef389 backported. Refer to
PlatformCpuCountBugCheck: <https://bugzilla.tianocore.org/show_bug.cgi?id=4250>.
...
[1] http://lists.busybox.net/pipermail/buildroot/2023-July/670807.html
Signed-off-by: Julien Olivain <ju.o@free.fr>
Tested-by: Romain Naour <romain.naour@gmail.com>
Reviewed-by: Romain Naour <romain.naour@gmail.com>
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
The Marvell Armada Devicetree files have been moved out of edk2-platforms
by commit 4b53da6b12a8 ("Marvell/Armada7k8k: Remove device tree sources
from edk2-platforms") and they are now in edk2-non-osi.
Therefore update the MACCHIATObin recipe to depend on the new edk2-non-osi
package and rework a bit the packages path to support that.
Also, drop the backported patch as it is not necessary anymore.
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Dick Olsson <hi@senzilla.io>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The edk2 project is licensed under the BSD-2-Clause license with a patent
grant, as per commit 304bff7223a8 ("edk2: Change License.txt from 2-Clause
BSD to BSD+Patent").
There is a BSD-2-Clause-Patent SPDX license identifier[1] for this case,
therefore refine the edk2 package to use this more specific identifier.
[1]: https://spdx.org/licenses/BSD-2-Clause-Patent.html
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Dick Olsson <hi@senzilla.io>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
The edk2 package can be configured for platform Arm Sgi575 but this
does not build correctly:
Usage: build.exe [options] [all|fds|genc|genmake|clean|cleanall|cleanlib|modules|libraries|run]
build.exe: error: option -a: invalid choice: '-b' (choose from 'IA32', 'X64', 'EBC', 'ARM', 'AARCH64', 'RISCV64')
make[1]: *** [package/pkg-generic.mk:293: /home/thomas/buildroot/buildroot/output/build/edk2-edk2-stable202102/.stamp_built] Error 2
make: *** [Makefile:84: _all] Error 2
Add the necessary definitions to fix the build.
Signed-off-by: Vincent Stehlé <vincent.stehle@arm.com>
Cc: Dick Olsson <hi@senzilla.io>
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Add BR2_TARGET_EDK2_PLATFORM_OVMF_I386 for x86_64 architecture.
Update the management of EDK2_ARCH to follow the edk2 platform type in
place of BR2 architecture.
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
Signed-off-by: Arnout Vandecappelle (Essensium/Mind) <arnout@mind.be>
Add support the build the firmware for QEMU i386 pc machine
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
[yann.morin.1998@free.fr:
- do not make it available for BR2_x86_64
- introduce BR2_TARGET_EDK2_ARCH_SUPPORTS
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
EDK2 is a modern, feature-rich, cross-platform firmware development
environment for the UEFI and PI specifications.
The initial version of this bootloader package makes it possible to
build firmware for the following seven configurations:
* QEMU x86-64 pc machine
* QEMU aarch64 virt machine, booting directly from flash
* QEMU aarch64 virt machine, booting via the kernel protocol
* QEMU aarch64 sbsa-ref machine
* ARM FVP vexpress machine
* Socionext SynQuacer Developerbox
* SolidRun MacchiatoBin
Support the use of EDK2 UEFI payloads as BL33 in ARM Trusted Firmware.
Signed-off-by: Dick Olsson <hi@senzilla.io>
[yann.morin.1998@free.fr:
- duplicate defaults in Config.in
]
Signed-off-by: Yann E. MORIN <yann.morin.1998@free.fr>
Jiaxun Yang