With LTO activated, the buildman tools failed with an error on my
configuration (Ubuntu 20.04, stm32mp15_trusted_defconfig) with the error:
../arm-linux-gnueabi/bin/nm:
scripts/gen_ll_addressable_symbols.sh: file format not recognized
It seems the shell variable initialization NM=$(NM) is not correctly
interpreted when shell is started in the Makefile, but I have not this
issue when I compile the same target without buildman.
I don't found the root reason of the problem but I solve it by
providing $(NM) as script parameter instead using a shell variable.
The command executed is identical:
cmd_keep-syms-lto.c := NM=arm-none-linux-gnueabihf-gcc-nm \
u-boot/scripts/gen_ll_addressable_symbols.sh arch/arm/cpu/built-in.o \
.... net/built-in.o >keep-syms-lto.c
cmd_keep-syms-lto.c := u-boot/scripts/gen_ll_addressable_symbols.sh \
arm-none-linux-gnueabihf-gcc-nm arch/arm/cpu/built-in.o \
... net/built-in.o > keep-syms-lto.c
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Some toolchains are compiled so that they pass a --build-id=something
parameter to the linker implicitly.
This causes U-Boot LTO linking to fail with something like:
ld: section .note.gnu.build-id LMA ... overlaps section .text LMA ...
because U-Boot's link scripts do not currently handle .note.gnu.build-id
section.
Fix this by explicitly disabling build-id.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
Add plumbing for building U-Boot with Link Time Optimizations.
When building with LTO, $(PLATFORM_LIBS) has to be in --whole-archive /
--no-whole-archive group, otherwise some functions declared in assembly
may not be resolved and linking may fail.
Note: clang may throw away linker list symbols it thinks are unused when
compiling with LTO. To force these symbols to be included, we refer to
them via the __ADDRESSABLE macro in a C file generated from compiled
built-in.o files before linking.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
Currently we use incremental linking (ld -r) to link several object
files from one directory into one built-in.o object file containing the
linked code from that directory (and its subdirectories).
Linux has, some time ago, moved to thin archives instead.
Thin archives are archives (.a) that do not really contain the object
files, only references to them.
Using thin archives instead of incremental linking
- saves disk space
- apparently works better with dead code elimination
- makes things easier for LTO
The third point is the important one for us. With incremental linking
there are several options how to do LTO, and that would unnecessarily
complicate things.
We have to use the --whole-archive/--no-whole-archive linking option
instead of --start-group/--end-group, otherwise linking may fail because
of unresolved symbols, or the resulting binary will be unusable.
We also need to use the P flag for ar, otherwise final linking may fail.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Simon Glass <sjg@chromium.org>
Indent the linking commands so that they look cosmetically better.
Signed-off-by: Marek Behún <marek.behun@nic.cz>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
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>
Which files we generate depends on the setting of OF_PLATDATA_INST in the
build. This might change between builds, but the build directory may be
reused.
Leaving old files around is confusing and switching the OF_PLATDATA_INST
setting does not necessarily regenerate the files, e.g. if the devicetree
has not changed.
Remove all the files before regenerating new ones.
Signed-off-by: Simon Glass <sjg@chromium.org>
There are actually two generated files but only one is currently mentioned
in the Makefile as a dependency. Put them both in a Makefile variable and
use that instead, to avoid inconsistency.
Signed-off-by: Simon Glass <sjg@chromium.org>
When OF_PLATDATA_INST is enabled, we need dt-uclass.c and dt-device.c for
the build to work. When OF_PLATDATA_INST is not enabled, we only need
dt-plat.c
Update the Makefile rules to indicate this.
At present all files are generated and compiled regardless of which are
actually needed.
Signed-off-by: Simon Glass <sjg@chromium.org>
Since the dst_dir rule always runs, it causes a rebuild of the of-platdata
files even if not needed.
Create the directory inside the rule instead.
Signed-off-by: Simon Glass <sjg@chromium.org>
Now that dtoc generates some new C files, add these to the build so that
the instantiated devices and uclasses can be used.
Signed-off-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Simon Glass <sjg@chromium.org>
Add Kconfig options to support build-time device instantiation. When
fully implemented, this will allow dtoc to create U-Boot devices (i.e.
struct udevice records) at build time, thus reducing code space in
SPL.
For now this defaults to off, but will be enabled when the rest of
the implementation is in place.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add target to Makefile to generate "u-boot-spl-dtb.hex" for Intel
SOCFPGA SOC64 devices (Stratix 10 and Agilex). "u-boot-spl-dtb.hex"
is hex formatted spl with and offset of CONFIG_SPL_TEXT_BASE. It
combines the spl image and dtb. "u-boot-spl-dtb.hex" is needed to
generate the final configuration bitstream for Intel SOCFPGA SOC64
devices.
Signed-off-by: Dalon Westergreen <dalon.westergreen@intel.com>
Signed-off-by: Siew Chin Lim <elly.siew.chin.lim@intel.com>
Switch the SPL boot image generation from using mksunxiboot to the new
sunxi_egon format of mkimage.
Verified to create identical results for all 152 Allwinner boards.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@siol.net>
Reviewed-by: Samuel Holland <samuel@sholland.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Samuel Holland <samuel@sholland.org>
Use this new name to be consistent with the rest of U-Boot, which talks
about 'plat' for the platform data, which is what this file holds.
Signed-off-by: Simon Glass <sjg@chromium.org>
Update the Makefile to run dtoc only once, generating all required files.
This saves time since there is a lot of processing in each invocation of
dtoc.
We already have a variable for the object files to build, so use that
instead of repeating the same filenames. Add a C version of this also,
for the same reason.
This makes it easier to add new C files (generated by dtoc) to the build
later, as needed.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present the dtoc commmand line is repeated twice in the Makefile. Use a
variable to avoid this, so it is easier to add more conditional arguments.
Signed-off-by: Simon Glass <sjg@chromium.org>
Use the SPL_NAME variable to simplify the rules. Drop the SPL targets
clean-files since the SPL and TPL dts/ directories are removed by
existing rules. Move the SPL rules into a new spl_dtbs to avoid the
complicated $(if) construct.
Also drop unused pieces from the 'targets' variable.
With this, SPL and TPL have separate dtbs which respect the various
u-boot,dm-spl / u-boot,dm-tpl tags.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rather than putting these in the top-level dts/ directory (which is
intended for U-Boot proper), put them in the correct subdirectory for
SPL (either spl/ or tpl/). This is where other SPL targets are kept,
so this is more consistent.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present both SPL and TPL use the same devicetree binary. While there
is logic to run fdtgrep separately on each one, it does not actually
happen.
Add a new TPL rule and use that instead. Make this rule conditional on
there actually being a TPL. Do the same for SPL for consistency.
Note that the SPL and TPL dtbs are build by a Makefule rule used for
U-Boot proper. This is the 'dtbs' target in dts/Makefile. So the check
for CONFIG_TPL_BUILD in cmd_fdtgrep never actually works at present.
We don't support CONFIG_OF_EMBED for TPL at present.
Signed-off-by: Simon Glass <sjg@chromium.org>
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>
This avoids needless rebuilding.
Fixes: 2f57c95100 ("spl: dm: Make it possible for the SPL to pick its own DTB from a FIT")
CC: Jean-Jacques Hiblot <jjhiblot@ti.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
The dts dir must exists when running this rule.
That missing dependency broke e.g. "make -j" for the am65x targets.
Fixes: 2f57c95100 ("spl: dm: Make it possible for the SPL to pick its own DTB from a FIT")
CC: Jean-Jacques Hiblot <jjhiblot@ti.com>
Signed-off-by: Jan Kiszka <jan.kiszka@siemens.com>
Currently generic-asm-offsets.h and asm-offsets.h are generated based
on U-Boot proper config options. The same asm-offsets headers are used
for building U-Boot SPL/TPL, which causes potential offset mismatch if
U-Boot proper has different config options from U-Boot SPL/TPL.
This commit adds:
spl/include/generated/(generic-)asm-offsets.h
tpl/include/generated/(generic-)asm-offsets.h
spl/include/generated/(generic-)asm-offsets.h is generated if
CONFIG_SPL=y, and included when building SPL.
tpl/include/generated/(generic-)asm-offsets.h is generated if
CONFIG_TPL=y, and included when building TPL.
They are created before Kbuild descends into SPL/TPL object directories
and builds $(obj)/dts/dt-platdata.o because $(obj)/dts/dt-platdata.c
includes a bunch of headers.
Prepend -I$(obj)/include to $(UBOOTINCLUDE) so (generic-)asm-offsets.h
is searched in {spl,tpl}/include/generated/.
Requested-by: Bin Meng <bmeng.cn@gmail.com>
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Tested-by: Bin Meng <bmeng.cn@gmail.com>
if_changed must have FORCE as a prerequisite.
Add $(obj)/dts/dt-platdata.o to 'targets' so that the corresponding
.cmd file is included.
Signed-off-by: Masahiro Yamada <masahiroy@kernel.org>
Align Kconfig and Kbuild logic to Linux 4.19 release with minimal impact
on files outside of this scope.
Our previous Kconfig sync was done by commit 5972ff077e ("kconfig /
kbuild: re-sync with Linux 4.18").
In this particular re-sync in order to keep clang support working a
number of related changes needed to be pulled in that had been missed
previously. Not all of these changes we easily traceable and so have
been omitted from the list below.
The imported Linux commits are:
[From prior to v4.18]
9f3f1fd29976 kbuild: Add __cc-option macro
d7f14c66c273 kbuild: Enable Large File Support for hostprogs
6d79a7b424a5 kbuild: suppress warnings from 'getconf LFS_*'
24403874316a Shared library support
86a9df597cdd kbuild: fix linker feature test macros when cross compiling with Clang
0294e6f4a000 kbuild: simplify ld-option implementation
[From v4.18 to v4.19]
96f14fe738b6 kbuild: Rename HOSTCFLAGS to KBUILD_HOSTCFLAGS
10844aebf448 kbuild: Rename HOSTCXXFLAGS to KBUILD_HOSTCXXFLAGS
b90a368000ab kbuild: Rename HOSTLDFLAGS to KBUILD_HOSTLDFLAGS
8377bd2b9ee1 kbuild: Rename HOST_LOADLIBES to KBUILD_HOSTLDLIBS
f92d19e0ef9b kbuild: Use HOST*FLAGS options from the command line
4ab3b80159d4 kconfig: check for pkg-config on make {menu,n,g,x}config
693359f7ac90 kconfig: rename SYMBOL_AUTO to SYMBOL_NO_WRITE
f60b992e30ff kbuild: replace $(LDFLAGS) $(ldflags-y) with $(ld_flags)
2fb9279f2c3e kbuild: change ld_flags to contain LDFLAGS_$(@F)
c931d34ea085 arm64: build with baremetal linker target instead of Linux when available
5accd7f3360e kconfig: handle format string before calling conf_message_callback()
a2ff4040151a kconfig: rename file_write_dep and move it to confdata.c
0608182ad542 kconfig: split out useful helpers in confdata.c
adc18acf42a1 kconfig: remove unneeded directory generation from local*config
79123b1389cc kconfig: create directories needed for syncconfig by itself
16952b77d8b5 kconfig: make syncconfig update .config regardless of sym_change_count
d6c6ab93e17f kbuild: remove deprecated host-progs variable
56869d45e364 kconfig: fix the rule of mainmenu_stmt symbol
c151272d1687 kconfig: remove unused sym_get_env_prop() function
1880861226c1 kconfig: remove P_ENV property type
e3fd9b5384f3 scripts/dtc: consolidate include path options in Makefile
4bf6a9af0e91 kconfig: add build-only configurator targets
f1575595d156 kconfig: error out when seeing recursive dependency
5e8c5299d315 kconfig: report recursive dependency involving 'imply'
f498926c47aa kconfig: improve the recursive dependency report
98a4afbfafd2 kconfig: fix "Can't open ..." in parallel build
9a9ddcf47831 kconfig: suppress "configuration written to .config" for syncconfig
87a32e624037 kbuild: pass LDFLAGS to recordmcount.pl
d503ac531a52 kbuild: rename LDFLAGS to KBUILD_LDFLAGS
217c3e019675 disable stringop truncation warnings for now
bc8d2e20a3eb kconfig: remove a spurious self-assignment
fd65465b7016 kconfig: do not require pkg-config on make {menu,n}config
5a4630aadb9a ftrace: Build with CPPFLAGS to get -Qunused-arguments
Note that this adds new cleanup work to do in that we should adapt the
shared library support we have to what is now upstream.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Masahiro Yamada <masahiroy@kernel.org>
Right now enabling SPL_FRAMEWORK will also enable it for the TPL in all
cases, making the TPL bigger. There may be cases where the TPL is really
size constrained due to its underlying ram size.
Therefore introduce a new TPL_FRAMEWORK option and make the relevant
conditionals check for both. The default is set to "y if SPL_FRAMEWORK"
to mimic the previous behaviour where the TPL would always get the
SPL framework if it was enabled in SPL.
Signed-off-by: Heiko Stuebner <heiko.stuebner@theobroma-systems.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Add support for USB3 host and gadget driver. This is a direct sync of
Linux kernel Cadence USB stack that from v5.4-rc1 release.
Driver has been modified so that it compiles without errors against
U-Boot code base.
Features not required for U-Boot such as scatter-gather DMA and OTG
interrupt handling has been dropped.
Signed-off-by: Jean-Jacques Hiblot <jjhiblot@ti.com>
[jjhiblot@ti.com: Add PHY support]
Signed-off-by: Vignesh Raghavendra <vigneshr@ti.com>
At present these two sections of code are linked together into a single
2KB chunk in a single file. Some Intel SoCs like to have a FIT (Firmware
Interface Table) in the ROM and the pointer for this needs to go at
0xffffffc0 which is in the middle of these two sections.
Make use of the new 'reset' entry and change the existing 16-bit entry to
include just the 16-bit data.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present we support having 16-bit init be in SPL or U-Boot proper, but
not TPL. Add support for this so that TPL can boot.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
At present only OF_SEPARATE is considered valid for building a device-tree
file in SPL. However sandbox uses OF_HOSTFILE instead. Update the logic to
handle this and make it easier to understand.
Note that the new logic is not quite the same as the old logic. It was
previously assumed that checking for:
$(CONFIG_$(SPL_TPL_)OF_CONTROL)
$(CONFIG_OF_SEPARATE)
$(CONFIG_$(SPL_TPL_)OF_PLATDATA)
producing 'yy' meant that the first two were 'y' and the last was empty.
Strictly speaking it would be possible for any two of the three to be 'y'
and still yield the same result. However, that was not the intention of
the new logic, since OF_PLATDATA always ensures that no device-tree file
is included. So in effect the new logic is the same, with the addition of
OF_HOSTFILE as an option for OP_SEPARATE.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
This patch moves the AT91SAM NAND booting SPL image "boot.bin" which
includes the ECC values from the root directory into the spl directory,
where all SPL related images are located.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Heiko Schocher <hs@denx.de>
Cc: Andreas Bießmann <andreas@biessmann.org>
Cc: Eugen Hristev <eugen.hristev@microchip.com>
Reviewed-by: Heiko Schocher <hs@denx.de>
Tested on the taurus board:
Tested-by: Heiko Schocher <hs@denx.de>
The sfp file is only valid for Gen5 (Cyclone5 & Arria5) and Arria10
devices. The file should only be built for these devices.
Signed-off-by: Dalon Westergreen <dalon.westergreen@intel.com>
Make use of "IMAGE_MAX_SIZE" and "IMAGE_TEXT_BASE" rather than
CONFIG_SPL_MAX_SIZE and CONFIG_SPL_TEXT_BASE. This lets us re-use the
same script for both SPL and TPL. Add logic to scripts/Makefile.spl to
pass in the right value when preprocessing the script.
Cc: Stefano Babic <sbabic@denx.de>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Cc: Jagan Teki <jagan@openedev.com>
Cc: Maxime Ripard <maxime.ripard@bootlin.com>
Cc: Andreas Bießmann <andreas@biessmann.org>
Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Cc: Michal Simek <monstr@monstr.eu>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: York Sun <york.sun@nxp.com>
Cc: Bin Meng <bmeng.cn@gmail.com>
Cc: Heiko Schocher <hs@denx.de>
Cc: Adam Ford <aford173@gmail.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Adam Ford <aford173@gmail.com> #da850evm & omap3_logic_somlv
Reviewed-by: Simon Goldschmidt <simon.k.r.goldschmidt@gmail.com>
Rather than checking for CONFIG_TPL_BUILD and then re-defining
CONFIG_SPL_TEXT_BASE make use of CONFIG_TPL_TEXT_BASE directly.
Cc: York Sun <york.sun@nxp.com>
Cc: Po Liu <po.liu@nxp.com>
Cc: Qiang Zhao <qiang.zhao@nxp.com>
Cc: Timur Tabi <timur@tabi.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Bootrom is not capable to work with non align bootloader partition
that's why it is necessary to align it before boot.bin creation.
The patch is creating new spl/u-boot-spl-align.bin which is used only
for boot.bin.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The function part_init() will only be built when we have both
CONFIG_PARTITIONS and CONFIG_HAVE_BLOCK_DEVICE set. Protect the call to
this function with both of these tests now.
Cc: Simon Glass <sjg@chromium.org>
Cc: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: York Sun <york.sun@nxp.com>
Cc: Prabhakar Kushwaha <prabhakar.kushwaha@nxp.com>
Cc: Mingkai Hu <mingkai.hu@nxp.com>
Cc: Stefan Roese <sr@denx.de>
Cc: Marek Behún <marek.behun@nic.cz>
Cc: Vanessa Maegima <vanessa.maegima@nxp.com>
Cc: Eugen Hristev <eugen.hristev@microchip.com>
Cc: Adam Ford <aford173@gmail.com>
Cc: Jagan Teki <jagan@amarulasolutions.com>
Cc: Tom Warren <twarren@nvidia.com>
Cc: Stephen Warren <swarren@nvidia.com>
Cc: Vitaly Andrianov <vitalya@ti.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This patch adds support for MTK boot image generation.
Signed-off-by: Weijie Gao <weijie.gao@mediatek.com>
Signed-off-by: Ryder Lee <ryder.lee@mediatek.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
At present // in a device-tree file or link script causes a warning. But
this is used in the standard license header. Update the compiler flags to
use C99, which permits this.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present the build fails in strange ways if CONFIG_SPL is defined by
CONFIG_SUPPORT_SPL is not. Add a warning for this case as it can be very
confusing to debug.
Signed-off-by: Simon Glass <sjg@chromium.org>
Update Makefile to generate
- tispl.bin: First stage bootloader on ARMv8 core
- u-boot.img: Second stage bootloader on ARMv8 core.
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Lokesh Vutla <lokeshvutla@ti.com>
The value of PMUFW_INIT_FILE is prefixed with "$(srctree)/", thus
forcing it to be a relative path inside the U-Boot source tree. Since
the PMUFW is a binary file generated outside of U-Boot, the PMUFW
binary must be copied inside the U-Boot source tree before the
build.
This generates a few problems:
* if the source tree is shared among different out-of-tree builds,
they will pollute (and potentially corrupt) each other
* the source tree cannot be read-only
* any buildsystem must add a command to copy the PMUFW binary
* putting an externally-generated binary in the source tree is ugly
as hell
Avoid these problems by accepting an absolute path for
PMUFW_INIT_FILE. This would be as simple as removing the "$(srctree)/"
prefix, but in order to keep backward compatibility we rather use the
shell and readlink to get the absolute path even when starting from a
relative path.
Since 'readlink -f' produces an empty string if the file does not
exist, we also add a check to ensure the file configured in
PMUFW_INIT_FILE exists. Otherwise the build would exit successfully,
but produce a boot.bin without PMUFW as if PMUFW_INIT_FILE were empty.
Tested in the 12 possible combinations of:
- PMUFW_INIT_FILE empty, relative, absolute, non-existing
- building in-tree, in subdir, in other directory
Signed-off-by: Luca Ceresoli <luca@lucaceresoli.net>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Emmanuel Vadot <manu@bidouilliste.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
When building compressed (lzop, gzip) multi-dtb fit images, the
compression tool may embed the time or umask in the image.
Work around this by manually setting the time of the source file using
SOURCE_DATE_EPOCH and a hard-coded 0600 umask.
With gzip, this could be accomplished by using -n/--no-name, but lzop
has no current workaround:
https://bugs.debian.org/896520
Signed-off-by: Vagrant Cascadian <vagrant@debian.org>
Generate SoCFPGA boot header version 1 instead of version 0 for Arria10.
Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Dinh Nguyen <dinguyen@kernel.org>
Cc: Chin Liang See <chin.liang.see@intel.com>
