LTO (Link-Time Optimisation) is an very useful feature which can
significantly reduce the size of U-Boot binaries. So far it has been
made available for selected ARM boards and sandbox.
However, incremental builds are much slower when LTO is used. For example,
an incremental build of sandbox takes 2.1 seconds on my machine, but 6.7
seconds with LTO enabled.
Add a NO_LTO parameter to the build, similar to NO_SDL, so it can be
disabled during development if needed, for faster builds.
Add some documentation about LTO while we are here.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rename the sections used to implement linker lists so they begin with
'__u_boot_list' rather than '.u_boot_list'. The double underscore at the
start is still distinct from the single underscore used by the symbol
names.
Having a '.' in the section names conflicts with clang's ASAN
instrumentation which tries to add redzones between the linker list
elements, causing expected accesses to fail. However, clang doesn't try
to add redzones to user sections, which are names with all alphanumeric
and underscore characters.
Signed-off-by: Andrew Scull <ascull@google.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
For ARMv8-A, NEON is standard, so the compiler can use it even when no
special target flags are provided. For example, it can use stores from
NEON registers to zero-initialize large structures. GCC 11 decides to
do this inside the DRAM init code for the Allwinner H6.
However, GCC 11 has a bug where it generates misaligned NEON register
stores even with -mstrict-align. Since the MMU is not enabled this early
in SPL, the misaligned store causes an exception and breaks booting.
Work around this issue by restricting the compiler to using GPRs only,
not vector registers. This prevents any future surprises relating to
NEON use as well.
Signed-off-by: Samuel Holland <samuel@sholland.org>
Acked-by: Andre Przywara <andre.przywara@arm.com>
- Provide a default Kconfig value of the default script
- Largely continue to define this via the board Kconfig file
- For the boards that select a script based on defconfig rather than
TARGET, keep this within the defconfig.
Signed-off-by: Tom Rini <trini@konsulko.com>
When using LTO, we can throw away the --gc-sections flag, but only if
using private libgcc.
When using system's libgcc, --gc-sections is still needed, otherwise
linking will fail due to undefined references to libc's symbols.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
When building with LTO, using -ffunction-sections/-fdata-sections is not
useful anymore.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add support for stack protector for UBOOT, SPL, and TPL
as well as new pytest for stackprotector
Signed-off-by: Joel Peshkin <joel.peshkin@broadcom.com>
Adjust UEFI build flags.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
When binman is in use, most of the targets built by the Makefile are
inputs to binman. We then need a final rule to run binman to produce the
final outputs.
Rename the variable to indicate this, and add a new 'inputs' target.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Some times gcc may generate data that is then used within code that may
be part of an efi runtime section. That data could be jump tables,
constants or strings.
In order to make sure we catch these, we need to ensure that gcc emits
them into a section that we can relocate together with all the other
efi runtime bits. This only works if the -ffunction-sections and
-fdata-sections flags are passed and the efi runtime functions are
in a section that starts with ".text".
Up to now we had all efi runtime bits in sections that did not
interfere with the normal section naming scheme, but this forces
us to do so. Hence we need to move the efi_loader text/data/rodata
sections before the global *(.text*) catch-all section.
With this patch in place, we should hopefully have an easier time
to extend the efi runtime functionality in the future.
Signed-off-by: Alexander Graf <agraf@suse.de>
[agraf: Fix x86_64 breakage]
When U-Boot started using SPDX tags we were among the early adopters and
there weren't a lot of other examples to borrow from. So we picked the
area of the file that usually had a full license text and replaced it
with an appropriate SPDX-License-Identifier: entry. Since then, the
Linux Kernel has adopted SPDX tags and they place it as the very first
line in a file (except where shebangs are used, then it's second line)
and with slightly different comment styles than us.
In part due to community overlap, in part due to better tag visibility
and in part for other minor reasons, switch over to that style.
This commit changes all instances where we have a single declared
license in the tag as both the before and after are identical in tag
contents. There's also a few places where I found we did not have a tag
and have introduced one.
Signed-off-by: Tom Rini <trini@konsulko.com>
As of clang-5.0, things have changed a bit. First, we cannot
automatically guess -target values as if we do not pass one with CC then
cc-option will fail. Second, to disable movt/movw relocations the
argument has become -mno-movt.
Related to the target part, we cannot use arm-none-eabi as that ends up
being too generic of an ARM target for things like say rpi_3_32b to
work.
Signed-off-by: Tom Rini <trini@konsulko.com>
This area of the image contains symbols whose values are filled in by
binman. If this feature is not used, the table is empty.
Add this to the ARM SPL link script.
Signed-off-by: Simon Glass <sjg@chromium.org>
The dtb blob section must always be present in the resulting image.
Either if OF_EMBEDED is used or if unit tests include dtb blobs.
Signed-off-by: Pantelis Antoniou <pantelis.antoniou@konsulko.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This is needed in order to allow building it for other archs.
Move relocation comment to a better place.
Remove no longer needed dts FIXME.
Signed-off-by: Álvaro Fernández Rojas <noltari@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
In Linux, CONFIG_ARCH_OMAP2PLUS is used for OMAP2 or later SoCs.
Rename CONFIG_ARCH_OMAP2 to CONFIG_ARCH_OMAP2PLUS to follow this
naming.
Move the OMAP2+ board/SoC choice down to mach-omap2/Kconfig to slim
down the arch/arm/Kconfig level.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
There are more and more cases where if we do not use gcc-6.0 or later we
run into problems where our binaries are too large for the targets.
Given the prevalence of gcc-6.0 or later toolchains at this point in
time, we give notice now that starting with v2018.01 we will require
gcc-6 (or later) for ARM.
Signed-off-by: Tom Rini <trini@konsulko.com>
Recent Linux distributions (e.g. Debian 9) include cross-compilers for
AArch64, but only for the aarch64-linux-gnu triplet only. It can thus
be expected that users will attempt to use the system cross-compiler
(instead of an aarch64-elf variant) to compile U-Boot for their ARMv8
target systems.
One key differences between an aarch64-linux-gnu and an aarch64-elf
compiler are the default settings regarding position-independent: with
the aarch64-linux-gnu compiler, the default will create and use the
global offset table.
This change-set adjusts the list of sections copied on ARMv8 to include
the GOT sections. With this added, the list matches the previous setup
for AArch32 closely.
Note that this is not an 'academic' issue, but was in fact encountered
by our QA during testing of the RK3399-Q7 BSP and resulted in an
early failure of the SPL stage during FDT setup.
Signed-off-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Tested-by: Klaus Goger <klaus.goger@theobroma-systems.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Today, we have cases where we wish to build all of U-Boot in Thumb2 mode for
various reasons. We also have cases where we only build SPL in Thumb2 mode due
to size constraints and wish to build the rest of the system in ARM mode. So
in this migration we introduce a new symbol as well, SPL_SYS_THUMB_BUILD to
control if we build everything or just SPL (or in theory, just U-Boot) in
Thumb2 mode.
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
This patch adds secure_text, secure_data and secure_stack sections for ARMv8 to
hold PSCI text and data, and it is based on the legacy implementation of ARMv7.
ARMV8_SECURE_BASE defines the address for PSCI secure sections, ARMV8_PSCI and
ARMV8_PSCI_NR_CPUS are firstly used in this patch, so they are introduce here
in Kconfig too.
Signed-off-by: Hongbo Zhang <hongbo.zhang@nxp.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: York Sun <york.sun@nxp.com>
With the move to arch/arm/mach-omap2 there are now very few uses of
CONFIG_OMAP_COMMON and further they can all be replaced with
CONFIG_ARCH_OMAP2, so do so.
Signed-off-by: Tom Rini <trini@konsulko.com>
Add support for EFI apps on ARM. This includes start-up and relocation
code, plus a link script and some compiler setting changes.
Signed-off-by: Simon Glass <sjg@chromium.org>
[agraf: Remove whitespace change, add kconfig dep]
Signed-off-by: Alexander Graf <agraf@suse.de>
This format can be flashed directly at address 0 of
the NAND FLASH, as it contains all necessary headers.
Signed-off-by: Albert ARIBAUD (3ADEV) <albert.aribaud@3adev.fr>
The secure monitor may need to store global or static values within the
secure section of memory, such as target PC or CPU power status.
Signed-off-by: Chen-Yu Tsai <wens@csie.org>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
- Move most of the flags required into LLVM_RELFLAGS to test at build
time instead of requiring them to be passed in.
- Update doc/README.clang to reflect this
- Switch to rpi_2 as the example as it's closer to working out of the
box than rpi is.
Cc: Jeroen Hofstee <jeroen@myspectrum.nl>
Signed-off-by: Tom Rini <trini@konsulko.com>
After booting has finished, EFI allows firmware to still interact with the OS
using the "runtime services". These callbacks live in a separate address space,
since they are available long after U-Boot has been overwritten by the OS.
This patch adds enough framework for arbitrary code inside of U-Boot to become
a runtime service with the right section attributes set. For now, we don't make
use of it yet though.
We could maybe in the future map U-boot environment variables to EFI variables
here.
Signed-off-by: Alexander Graf <agraf@suse.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Android's tool chain enable the -mandroid at default.
This option will enable the -fpic, which cause uboot compilation
failure:
"
LD u-boot
u-boot contains unexpected relocations: R_ARM_ABS32
R_ARM_RELATIVE
"
In my testcase, arm-linux-androideabi-gcc-4.9 internally
enables '-fpic', so when compiling code, there will be
relocation entries using type R_ARM_GOT_BREL and .got
section. When linking all the built-in.o using ld, there
will be R_ARM_ABS32 relocation entry and .got section
in the final u-boot elf image. This can not be handled
by u-boot, since u-boot only expects R_ARM_RELATIVE
relocation entry.
arm-poky-linux-gnueabi-gcc-4.9 default does not enable '-fpic',
so there is not .got section and R_ARM_GOT_BREL in built-in.o.
And in the final u-boot elf image, all relocation entries are
R_ARM_RELATIVE.
we can pass '-fno-pic' to xxx-gcc to disable pic. whether
the toolchain internally enables or disables pic, '-fno-pic'
can work well.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Cc: Albert Aribaud <albert.u.boot@aribaud.net>
Reviewed-by: Tom Rini <trini@konsulko.com>
For compatibility clang will report some gcc version. However since we
are checking gcc versions in order to then fail to build, we should
limit these tests only to when we are using gcc and not clang.
Signed-off-by: Tom Rini <trini@konsulko.com>
Commit 47ed5dd0 dropped the .got section from U-Boot binaries. This is needed
for some relocations, and causes failures if missing. Add it back.
Signed-off-by: Simon Glass <sjg@chromium.org>
Move the option to Kconfig renaming it to CONFIG_HAVE_GENERIC_BOARD.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Alexey Brodkin <abrodkin@synopsys.com>
Resynchronize memcpy/memset with kernel 3.17 and build them in
Thumb2 mode (unified syntax). Those assembler files can be built
and linked in ARM mode too, however when calling them from Thumb2
built code, the stack got corrupted and the copy did not succeed
(the exact details have not been traced back). However, the Linux
kernel builds those files in Thumb2 mode. Hence U-Boot should
build them in Thumb2 mode too when CONFIG_SYS_THUMB_BUILD is set.
To build the files without warning, some assembler instructions
had to be replaced with their UAL compliant variant (thanks
Jeroen for this input).
To build the file in Thumb2 mode the implicit-it=always option need
to be set to generate Thumb2 compliant IT instructions where needed.
We add this option to the general AFLAGS when building for Thumb2.
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Stefan Agner <stefan@agner.ch>
In anticipation of refactoring the HYP/non-secure code to run
from secure RAM, add a new linker section that will contain that
code.
Nothing is using it just yet.
Signed-off-by: Marc Zyngier <marc.zyngier@arm.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
After Kbuild introduction, the CROSS_COMPILE environment variable has been
set to some default value (prefix arm-linux-).
This shall be removed since it breaks building u-boot for native arm target
(like qemu ARM).
Moreover not all compilers have arm-linux- prefix.
Additionally the u-boot cross compiles with CROSS_COMPILE= set explicitly-
e.g.:
CROSS_COMPILE=/ .... /arm-v7a-linux-gnueabi- make
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
Cc: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
"checkthumb" makes sense only for ARM architecture.
Move it to arch/arm/config.mk.
To make sure gcc supports THUMB mode before beginning build,
run "checkthumb" during "archprepare".
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
The syntax
CROSS_COMIPLE ?= <cross_compiler_prefix>
does not work because config.mk is parsed after
exporting CROSS_COMPILE.
Like Linux Kernel's arch/$(ARCH)/Makefile,
we must write as follows:
ifeq ($(CROSS_COMPILE),)
CROSS_COMPILE := <cross_compiler_prefix>
endif
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
We had switched to Kbuild so now we can specify
PLATFORM_LIBS/PLATFORM_LIBGCC with relative path.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Acked-by: Tom Rini <trini@ti.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
This commit avoids generating ./SPL twice.
- Fist time descending to spl/
- Second time as a prerequisite of u-boot-with-spl.imx,
u-boot-with-nand-spl.imx.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Rename OBJCFLAGS to OBJCOPYFLAGS beforehand to use
"cmd_objcopy" in scripts/Makefile.lib in an upcoming commit.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Before this commit, all arch/arm/cpu/${CPU}/config.mk except ARMv8
had the same option:
$(call cc-option,-mshort-load-bytes,$(call cc-option,-malignment-traps,))
This commit moves it into arch/arm/config.mk.
If the compiler does not support the option,
it is ignored by $(call cc-option,...).
So this commit gives no harm to ARMv8.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
Some targets will build fine but not boot if sections .hash and
.got.plt are not present in the binary. Add them back.
Also, Exynos machines require .machine_param section in SPL.
Add it.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Tested-by: Rajeshwari S Shinde <rajeshwari.s@samsung.com>
Relocation code based on a patch by Scott Wood, which is:
Signed-off-by: Scott Wood <scottwood@freescale.com>
Signed-off-by: David Feng <fenghua@phytium.com.cn>
Current LDS files /DISCARD/ a lot of sections when linking ELF
files, causing diagnostic tools such as readelf or objdump to
produce partial output. Keep all section at link stage, filter
only at objcopy time so that .bin remains minimal.
Signed-off-by: Albert ARIBAUD <albert.u.boot@aribaud.net>
Reviewed-by: Benoît Thébaudeau <benoit.thebaudeau@advansee.com>
To be more EABI compliant and as a preparation for building
with clang, use the platform-specific r9 register for gd
instead of r8.
note: The FIQ is not updated since it is not used in u-boot,
and under discussion for the time being.
The following checkpatch warning is ignored:
WARNING: Use of volatile is usually wrong: see
Documentation/volatile-considered-harmful.txt
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Every ARM cpu config.mk (arch/arm/cpu/{CPUDIR}/config.mk) defines:
PLATFORM_RELFLAGS += -fno-common -ffixed-r8 -msoft-float
So, this patch moves the common compiler options to arch/arm/config.mk.
Signed-off-by: Masahiro Yamada <yamada.m@jp.panasonic.com>
The movt/movw instruction can be used to hardcode an
memory location in the instruction itself. The linker
starts complaining about this if the compiler decides
to do so: "relocation R_ARM_MOVW_ABS_NC against `a local
symbol' can not be used" and it is not support by U-boot
as well. Prevent their use by requiring word relocations.
This allows u-boot to be build at other optimalization
levels then -Os.
Signed-off-by: Jeroen Hofstee <jeroen@myspectrum.nl>
Cc: TigerLiu@viatech.com.cn
Cc: Albert ARIBAUD <albert.u.boot@aribaud.net>
Acked-by: Simon Glass <sjg@chromium.org>
