U-Boot has up until now built with -fpic for the MIPS architecture,
producing position independent code which uses indirection through a
global offset table, making relocation fairly straightforward as it
simply involves patching up GOT entries.
Using -fpic does however have some downsides. The biggest of these is
that generated code is bloated in various ways. For example, function
calls are indirected through the GOT & the t9 register:
8f998064 lw t9,-32668(gp)
0320f809 jalr t9
Without -fpic the call is simply:
0f803f01 jal be00fc04 <puts>
This is more compact & faster (due to the lack of the load & the
dependency the jump has on its result). It is also easier to read &
debug because the disassembly shows what function is being called,
rather than just an offset from gp which would then have to be looked up
in the ELF to discover the target function.
Another disadvantage of -fpic is that each function begins with a
sequence to calculate the value of the gp register, for example:
3c1c0004 lui gp,0x4
279c3384 addiu gp,gp,13188
0399e021 addu gp,gp,t9
Without using -fpic this sequence no longer appears at the start of each
function, reducing code size considerably.
This patch switches U-Boot from building with -fpic to building with
-fno-pic, in order to gain the benefits described above. The cost of
this is an extra step during the build process to extract relocation
data from the ELF & write it into a new .rel section in a compact
format, plus the added complexity of dealing with multiple types of
relocation rather than the single type that applied to the GOT. The
benefit is smaller, cleaner, more debuggable code. The relocate_code()
function is reimplemented in C to handle the new relocation scheme,
which also makes it easier to read & debug.
Taking maltael_defconfig as an example the size of u-boot.bin built
using the Codescape MIPS 2016.05-06 toolchain (gcc 4.9.2, binutils
2.24.90) shrinks from 254KiB to 224KiB.
Signed-off-by: Paul Burton <paul.burton@imgtec.com>
Cc: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Cc: u-boot@lists.denx.de
Reviewed-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Tested-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
FreeBSD recently switch to it's BSDL dtc. While it support most of the
features of the GPL one it still lacks the incbin directive.
Add the possibility to specify which dtc we want to use for compiling dts
and generating fit image.
Signed-off-by: Emmanuel Vadot <manu@bidouilliste.com>
Acked-by: Simon Glass <sjg@chromium.org>
Build the upstream python libfdt module. At present the legacy module is
still built and is the one that it used. Future work will switch this
over.
Signed-off-by: Simon Glass <sjg@chromium.org>
Now that this module has been accepted upstream we should stop using the
local U-Boot one. In preparation for this, rename it to indicate it is for
legacy use.
Signed-off-by: Simon Glass <sjg@chromium.org>
At present only a subset of source files are build. Add the rest and
refactor this so that a source file list is available also. This will be
used in later commit.
Signed-off-by: Simon Glass <sjg@chromium.org>
Not force to use python from PATH. Issue was noted when building with
Yocto, because python from the distro is always taken instead of
python-native built during Yocto process.
Signed-off-by: Stefano Babic <sbabic@denx.de>
CC: Simon Glass <sjg@chromium.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
This program generates raw SPL images that can be flashed on the NAND with
the ECC and randomizer properly set up.
This has been copied (and tweaked to find the right headers) from the
sunxi-tools (https://github.com/linux-sunxi/sunxi-tools) upstream
repository, commit 1c3a6ca5.
Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>
Acked-by: Boris Brezillon <boris.brezillon@free-electrons.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Jagan Teki <jagan@openedev.com>
When building u-boot tools in cross-build environment CFLAGS environment
variable set up for target is taken into an account when building code
for host. Make it empty on invocation of python.
This fixes the following build errors when cross-compiling for xtensa:
cc1: error: unrecognized command line option "-mlongcalls"
cc1: error: unrecognized command line option "-mauto-litpools"
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Tom Rini <trini@konsulko.com>
Richard reported U-Boot tools issues in OpenEmbedded/Yocto project.
OE needs to be able to change the default compiler. If we pass in
HOSTCC through the make command, it overwrites all HOSTCC instances,
including ones in tools/Makefile and tools/env/Makefile, which breaks
"make cross_tools" and "make env", respectively.
Add "override" directives to avoid overriding HOSTCC instances that
really need to point to the cross-compiler.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reported-by: Richard Purdie <richard.purdie@linuxfoundation.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
This define is not used in tools sources and can be removed
to avoid unnecessary link between tools and defconfig
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The previous commit ("add a new command to show .config contents")
improves the basic infrastructure of "embed a compressed file into
the U-Boot image, and print it by a command". The same pattern for
the "license" command.
This commit reworks the command to improve the following:
[1] Improve log style
Kbuild style log
GZIP cmd/license_data.gz
CHK cmd/license_data_gz.h
UPD cmd/license_data_gz.h
CHK cmd/license_data_size.h
UPD cmd/license_data_size.h
instead of the bare Make log:
cat ./Licenses/gpl-2.0.txt | gzip -9 -c | \
tools/bin2header license_gzip > ./include/license.h
[2] Collect related code into the "cmd" directory
Prior to this commit, the license.h was created by tools/Makefile,
placed under the "include" directory, included from cmd/license.c,
and deleted by the top-level Makefile. It is not a good idea to
scatter related code.
[3] Drop the fixed-malloc size LICENSE_MAX
Just allocate the minimum required size of buffer because we know
the size of the original gpl-2.0.txt.
[4] Fix more issues
Terminate the buffer with zero to prevent puts() from over-running.
Add "static" to do_license.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
This patch adds a little tool that takes a generic MAC address and
generates a CRC byte for it. The output is the full MAC address without
any separators, ready written into an EEPROM.
Signed-off-by: Olliver Schinagl <o.schinagl@ultimaker.com>
Signed-off-by: Olliver Schinagl <oliver@schinagl.nl>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch enables crc8 to be used from within the tools directory using
u-boot/crc.h.
Signed-off-by: Olliver Schinagl <o.schinagl@ultimaker.com>
Reviewed-by: Joe Hershberger <joe.hershberger@ni.com>
Signed-off-by: Olliver Schinagl <oliver@schinagl.nl>
The patch implements secure booting for the mvebu architecture.
This includes:
- The addition of secure headers and all needed signatures and keys in
mkimage
- Commands capable of writing the board's efuses to both write the
needed cryptographic data and enable the secure booting mechanism
- The creation of convenience text files containing the necessary
commands to write the efuses
The KAK and CSK keys are expected to reside in the files kwb_kak.key and
kwb_csk.key (OpenSSL 2048 bit private keys) in the top-level directory.
Signed-off-by: Reinhard Pfau <reinhard.pfau@gdsys.cc>
Signed-off-by: Mario Six <mario.six@gdsys.cc>
Reviewed-by: Stefan Roese <sr@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Stefan Roese <sr@denx.de>
When we have python building tools for the host it will not check HOSTXX
variables but only XX variables, for example LDFLAGS and not
HOSTLDFLAGS.
Cc: Simon Glass <sjg@chromium.org>
Reported-by: Heiko Schocher <hs@denx.de>
Fixes: 1905c8fc71 ("build: Always build the libfdt python module")
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Tested-by: Heiko Schocher <hs@denx.de>
Do not rely on CONFIG_SPL_OF_PLATDATA to build the libfdt python module.
If swig is present, this will be build
Signed-off-by: Emmanuel Vadot <manu@bidouilliste.com>
The Makefile in tools/ tries to find the "swig" utility by calling "which".
If nothing is found in the path, some versions of which will print an error
message:
$ make clean
which: no swig in (/usr/local/bin:/usr/bin:/bin)
This does not apply to all version of "which", though:
$ echo $0
bash
$ type which
which is aliased to `type -path'
$ which foo <== this version is OK
$ /usr/bin/which foo <== this one is chatty
/usr/bin/which: no foo in (/usr/local/bin:/usr/bin:/bin)
$ sh <== make uses /bin/sh
sh-4.3$ which foo <== no alias here
which: no foo in (/usr/local/bin:/usr/bin:/bin)
This error message is rather pointless in our case, since we just have
this very check to care for this. So add stderr redirection to suppress
the message.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Commit fed329aebe ("tools: add mksunxiboot to tools-all target") added
mksunxiboot to the tools-all target, but used the CONFIG_SUNXI symbol
to enable its build. Now commit aec9a0f19f ("sunxi: Rename CONFIG_SUNXI
to CONFIG_ARCH_SUNXI"), merged before that, renamed that symbol, so that
the first patch basically gets ineffective.
Adjust the symbol name in tools/Makefile to make it build again.
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
mksunxiboot is useful outside of u-boot, it is e.g. used by sunxi-tools.
Signed-off-by: Stefan Brüns <stefan.bruens@rwth-aachen.de>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
CONFIG_SUNXI -> CONFIG_ARCH_SUNXI
and removed CONFIG_SUNIX from config_whitelist.txt
Cc: Simon Glass <sjg@chromium.org>
Cc: Ian Campbell <ijc@hellion.org.uk>
Cc: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Jagan Teki <jteki@openedev.com>
Reviewed-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
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>
When swig is not available, we can still build correctly. So make this
optional. Add a comment about how to enable this build.
Signed-off-by: Simon Glass <sjg@chromium.org>
Add a Python version of the libfdt library which contains enough features to
support the dtoc tool. This is only a very bare-bones implementation. It
requires the 'swig' to build.
Signed-off-by: Simon Glass <sjg@chromium.org>
Because a gpimage cannot be detected, a false
GP header is printed instead of checking
for further image types.
Move gpimage as last to be linked, letting check
all other image types and printing a GP header just
in case no image is detected.
Signed-off-by: Stefano Babic <sbabic@denx.de>
Add support for the zynqmpimage to mkimage.
Only basic functionality is supported without encryption and register
initialization with one partition which is filled by U-Boot SPL.
For more detail information look at Xilinx ZynqMP TRM.
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The license command isn't usually built and has a few problems:
- The rules to generate license.h haven't worked in a long time,
re-write these based on the bmp_logo.h rules.
- 'tok' is unused and the license text size has increased
- bin2header.c wasn't grabbing unistd.h to know the prototype for
read().
Cc: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
At present this file is omitted. It is used to build up a binary device
tree. We plan to do this in mkimage, so include this file in the build.
Signed-off-by: Simon Glass <sjg@chromium.org>
Since OpenSSL is deprecated on OS X in favour of Common Crypto API disable the
warning for this host OS.
Another solution would be to add some glue layer for crypto stuff, but I think
this is not worth the effort.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Until now, the SoC selection for the ARCH_MVEBU platforms has been done
in the config header. Using CONFIG_ARMADA_XP in a non-clear way. As
it needed to get selected for AXP and A38x based boards. This patch
now changes this to move the SoC selection to Kconfig. And also
uses CONFIG_ARCH_MVEBU as a common define for both AXP and A38x.
This makes things a bit clearer - especially for new board additions.
Additionally the defines CONFIG_SYS_MVEBU_DDR_AXP and
CONFIG_SYS_MVEBU_DDR_A38X are replaced with the already available
CONFIG_ARMADA_38X and CONFIG_ARMADA_XP.
And CONFIG_DDR3 is removed, as its not referenced anywhere.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Our chips may have different max spl size and spl header, so
we need to add configs for that.
Signed-off-by: Jeffy Chen <jeffy.chen@rock-chips.com>
Acked-by: Simon Glass <sjg@chromium.org>
Dropped CONFIG_ROCKCHIP_MAX_SPL_SIZE from rk3288_common.h,
Added $(if...) to tools/Makefile to fix widespread build breakage
Signed-off-by: Simon Glass <sjg@chromium.org>
Series-changes: 8
- Drop CONFIG_ROCKCHIP_MAX_SPL_SIZE from rk3288_common.h,
- Add $(if...) to tools/Makefile to fix widespread build breakage
As with other platforms vendors love to create their own boot header
formats. Xilinx is no different and for the Zynq platform/SoC there
exists the "boot.bin" which is read by the platforms bootrom. This
format is described to a useful extent within the Xilinx Zynq TRM.
This implementation adds support for the 'zynqimage' to mkimage. The
implementation only considers the most common boot header which is
un-encrypted and packed directly after the boot header itself (no
XIP, etc.). However this implementation does take into consideration the
other fields of the header for image dumping use cases (vector table and
register initialization).
Signed-off-by: Nathan Rossi <nathan@nathanrossi.com>
Cc: Michal Simek <michal.simek@xilinx.com>
Cc: Tom Rini <trini@konsulko.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
The Rockchip boot ROM requires a particular file format. It consists of
64KB of zeroes, a 512-byte header encoded with RC4, and then some executable
code.
Add support to mkimage so that an SPL image (u-boot-spl-dtb.bin) can be
converted to this format.
Signed-off-by: Simon Glass <sjg@chromium.org>
Rockchip SoCs require certain formats for code that they execute, The
simplest format is a 4-byte header at the start of a binary file. Add
support for this so that we can create images that the boot ROM understands.
Signed-off-by: Simon Glass <sjg@chromium.org>
This is commented out in the Makefile for more than 10 years.
I assume it is proof that this tool is unused.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Cc: Pantelis Antoniou <panto@intracom.gr>
To use this offset for other boot device (like SDIO/MMC), lets rename
it to a more generic name. This will be used be the SDIO/MMC SPL boot
support for the A38x.
Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <eibach@gdsys.de>
This tool allows us to extract subsets of a device tree file. It is used by
the SPL vuild, which needs to cut down the device tree size for use in
limited memory.
This tool was originally written for libfdt but it has not been accepted
upstream, so for now, include it in U-Boot. Several utilfdt library
functions been included inline here.
If fdtgrep is eventually accepted in libfdt then we can bring that version
of libfdt in here, and drop fdtgrep (requiring that fdtgrep is provided by
the user).
If it is not accepted then another approach would be to write a special
tool for chopping down device tree files for SPL. While it would use the
same libfdt support, it would be less code than fdtgrep.c because it would
not have general-purpose functions.
Another approach (which was used with v1 of this series) is to sprinkler all
the device tree files with #ifdef. I don't like that idea.
Signed-off-by: Simon Glass <sjg@chromium.org>
Instead of hardcoding -lssl -lcrypto as the flags needed to build
mkimage with FIT signature enabled, use pkg-config when
available. This allows to properly support cases where static linking
is used, which requires linking with -lz, since OpenSSL uses zlib
internally.
We gracefully fallback on the previous behavior of hardcoding -lssl
-lcrypto if pkg-config is not available or fails with an error.
Signed-off-by: Thomas Petazzoni <thomas.petazzoni@free-electrons.com>
Commit a93648d197 introduced linker generated
lists for imagetool which is the base for some host tools (mkimage, dumpimage,
et al.). Unfortunately some host tool chains do not support the used type of
linker scripts. Therefore this commit broke these host-tools for them, namely
FreeBSD and Darwin (OS/X).
This commit tries to fix this. In order to have a clean distinction between host
and embedded code space we need to introduce our own linker generated list
instead of re-using the available linker_lists.h provided functionality. So we
copy the implementation used in linux kernel script/mod/file2alias.c which has
the very same problem (cause it is a host tool). This code also comes with an
abstraction for Mach-O binary format used in Darwin systems.
Signed-off-by: Andreas Bießmann <andreas.devel@googlemail.com>
Cc: Guilherme Maciel Ferreira <guilherme.maciel.ferreira@gmail.com>
We can't use config.h directly as some platforms include headers that
aren't safe to use in normal Linux userland.
Signed-off-by: Tom Rini <trini@ti.com>
Public exponentiation which is required in rsa verify functionality is
tightly integrated with verification code in rsa_verify.c. The patch
splits the file into twp separating the modular exponentiation.
1. rsa-verify.c
- The file parses device tree keys node to fill a keyprop structure.
The keyprop structure can then be converted to implementation specific
format.
(struct rsa_pub_key for sw implementation)
- The parsed device tree node is then passed to a generic rsa_mod_exp
function.
2. rsa-mod-exp.c
Move the software specific functions related to modular exponentiation
from rsa-verify.c to this file.
Signed-off-by: Ruchika Gupta <ruchika.gupta@freescale.com>
CC: Simon Glass <sjg@chromium.org>
Acked-by: Simon Glass <sjg@chromium.org>
The registration was introduced in commit f86ed6a8d5
This commit also removes all registration functions, and the member "next"
from image_type_params struct
Signed-off-by: Guilherme Maciel Ferreira <guilherme.maciel.ferreira@gmail.com>
Some Intel CPUs use an 'FSP' binary blob which provides an inflexible
means of starting up the CPU. One result is that microcode updates can only
be done before RAM is available and therefore parsing of the device tree
is impracticle.
Worse, the addess of the microcode update must be stored in ROM since a
pointer to its start address and size is passed to the 'FSP' blob. It is
not possible to perform any calculations to obtain the address and size.
To work around this, ifdtool is enhanced to work out the address and size of
the first microcode update it finds in the supplied device tree. It then
writes these into the correct place in the ROM. U-Boot can then start up
the FSP correctly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Newer Intel chips require a Management Engine which requires a particular
format for the SPI flash that contains the boot loader. Add a tool that
supports creating and modifying these ROM images.
This tool is from Chrome OS but has been cleaned up to use U-Boot style
and to add comments. A few features have been added also.
Signed-off-by: Simon Glass <sjg@chromium.org>
