Rasmus Villemoes authored 1 year ago and Tom Rini committed 1 year ago

Currently, all in-tree .dts files (apart from some under test/ and
tools/), reside in arch/$ARCH/dts. However, in the linux kernel tree,
dts files for arm64 boards, and probably in the not too distant
future [1], arm boards as well, live in subdirectories of that.

For private forks, using a vendor or project subdirectory is also more
convenient to clearly separate private code from upstream - in the
same way that code under board/ is also split and easy to maintain.

In order to prepare for us to follow suit and do the splitting of the
in-tree .dts files, and to make life a little easier for private forks
that already place dts files not directly in arch/$ARCH/dts, change
the $(srctree)/arch/$(ARCH)/dts path to instead refer to the directory of
the .dts file being compiled. This should be a no-op for all existing
cases.

[1] https://lore.kernel.org/lkml/20220328000915.15041-1-ansuelsmth@gmail.com/



Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>

Rasmus Villemoes authored 1 year ago and Tom Rini committed 1 year ago

[Linux commit 363547d2191c]

Currently DTB Overlays (.dtbo) are build from source files with the same
extension (.dts) as the base DTs (.dtb). This may become confusing and
even lead to wrong results. For example, a composite DTB (created from a
base DTB and a set of overlays) might have the same name as one of the
overlays that create it.

Different files should be generated from differently named sources.
 .dtb  <-> .dts
 .dtbo <-> .dtso

We do not remove the ability to compile DTBO files from .dts files here,
only add a new rule allowing the .dtso file name. The current .dts named
overlays can be renamed with time. After all have been renamed we can
remove the other rule.

[Import notes: Adapt to U-Boot by using the cmd_dtco function instead
of cmd_dtc just like the current .dts -> .dtbo rule.]

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Simon Glass <sjg@chromium.org>

Rasmus Villemoes authored 1 year ago and Tom Rini committed 1 year ago

[Linux commit b8fc5b2157b1]

If dtc is rebuilt, we should rebuild .dtb files with the new dtc.

[Import notes: Back then there was no .dtbo rule in Linux's
Makefile.lib, but the current .dtbo rules in Linux also have the
$(DTC) dependency, so also add it to our .dtbo rule.]

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Reviewed-by: Simon Glass <sjg@chromium.org>

Rasmus Villemoes authored 1 year ago and Tom Rini committed 1 year ago

The idea in 3609e1dc (dts: automatically build necessary .dtb
files) was fine, but the implementation was suboptimal due to some
misunderstandings on my part (and possibly defects in some defconfig
files):

- Sometimes DEFAULT_DEVICE_TREE is not included in OF_LIST or
  SPL_OF_LIST

- SPL_OF_LIST is not always a subset of OF_LIST

- While SPL_OF_LIST governs the list of dtbs relevant to SPL (i.e.,
  may be built into an
  SPL-with-bunch-of-dtbs-to-choose-between-at-runtime), those dtbs are
  not actually _built_ during the SPL build phase, i.e. when $(SPL_)
  would expand to SPL_. fdtgrep runs on the artifacts produced during
  the ordinary U-Boot build.

Tweak the logic so that we simply add the union of all dtbs mentioned
in either DEFAULT_DEVICE_TREE, OF_LIST and SPL_OF_LIST to dtb-y. That
should, for real, ensure that we always build all the dtbs that is
relevant to the current board, and should in turn enable us to
massively simplify arch/*/dts/Makefile.

Signed-off-by: Rasmus Villemoes <rasmus.villemoes@prevas.dk>
Tested-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

Martin Hundebøll authored 1 year ago and Tom Rini committed 1 year ago

Building dtc from git causes the version number to start with a 'v'
(e.g. v1.7.0). printf then fails to parse 'v1' as a decimal value, and
prints '000700' instead of '010700'. Subsequently, the build fails,
because '000700' is less than the required '010400' version.

Signed-off-by: Martin Hundebøll <martin@geanix.com>

Simon Glass authored 2 years ago and Heinrich Schuchardt committed 2 years ago

Add an option to package a kernel into the debugging script used for
EFI.

The name of the kernel must be added to the script. By default it is
assumed that the kernel is built in the /tmp/kernel directory.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Heinrich Schuchardt committed 2 years ago

When running with video enabled, the pre-relocation output of U-Boot is
currently lost. Add a -serial flag to show it on the terminal.

Signed-off-by: Simon Glass <sjg@chromium.org>

Create a script which can package a tool for use with PyPi and the 'pip'
tool. This involves quite a few steps so is best automated. Future work
will enable use of this for some of the tools in U-Boot.

Signed-off-by: Simon Glass <sjg@chromium.org>

The patman directory has a number of modules which are used by other tools
in U-Boot. This makes it hard to package the tools using pypi since the
common files must be copied along with the tool that uses them.

To address this, move these files into a new u_boot_pylib library. This
can be packaged separately and listed as a dependency of each tool.

Signed-off-by: Simon Glass <sjg@chromium.org>

This is only for posterity, since once the conversion is done, the script
is of no use.

Signed-off-by: Simon Glass <sjg@chromium.org>

Help ensure that these don't creep into development by adding a check in
checkpatch for them.

Signed-off-by: Simon Glass <sjg@chromium.org>

Add a CONFIG option to deal with this automatically, printing a warning
when U-Boot starts up. This can be useful if the device tree comes from
another project.

We will maintain this through the 2023.07 release, providing 6 months
for people to notice.

Signed-off-by: Simon Glass <sjg@chromium.org>
Version 4:
Acked-by: Michal Simek <michal.simek@amd.com>

Ville Skyttä authored 2 years ago and Tom Rini committed 2 years ago

`egrep` has been deprecated in GNU grep since 2007,
and since 3.8 it emits obsolescence warnings:
https://git.savannah.gnu.org/cgit/grep.git/commit/?id=a9515624709865d480e3142fd959bccd1c9372d1


Acked-by: Dhruva Gole <d-gole@ti.com>

In some cases, we might not have the sic portion of setuputils
available. Make our import and use of this be done in try/except blocks
as this is done to suppress a run-time warning that is otherwise
non-fatal.

Reported-by: Pali Rohár <pali@kernel.org>
Fixes: 14165918 ("pylibfdt: Fix disable version normalization")
Signed-off-by: Tom Rini <trini@konsulko.com>

Philippe Schenker authored 2 years ago and Tom Rini committed 2 years ago

On Arch Linux based systems python setuptools does not contain
"setuptools.extern" hence it is failing with the following
error-message:
"
ModuleNotFoundError: No module named 'setuptools.extern'
"

According to a eschwartz `setuptools.extern` is not a public API and
shall not be assumed to be present in the setuptools package. He
mentions that the setuptools project anyway wants to drop this. [1]

Use the correct solution introduced by python setuptools developers to
disable normalization. [2]

[1] https://bbs.archlinux.org/viewtopic.php?id=259608
[2] https://github.com/pypa/setuptools/pull/2026


Fixes: 440098c4 ("pylibfdt: Fix version normalization warning")
Signed-off-by: Philippe Schenker <philippe.schenker@toradex.com>
Reviewed-by: Marek Vasut <marex@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>

At this point all listed adhoc CONFIG symbols have been migrated to
Kconfig or removed from the tree or renamed to CFG (or similar). We also
now have CI tests that will error on any new introductions, and
checkpatch.pl also looks. We can now remove these hooks and related
scripts.

Signed-off-by: Tom Rini <trini@konsulko.com>

Now that all CONFIG symbols are in Kconfig, checkpatch.pl should check
for and error on any case of define/undef CONFIG_*.

Signed-off-by: Tom Rini <trini@konsulko.com>

At this point in the conversion there should be no need to have logic to
disable some symbol during the SPL build as all symbols should have an
SPL counterpart.

The main real changes done here are that we now must make proper use of
CONFIG_IS_ENABLED(DM_SERIAL) rather than many of the odd tricks we
developed prior to CONFIG_IS_ENABLED() being available.

Signed-off-by: Tom Rini <trini@konsulko.com>

Perform a simple rename of CONFIG_BOARDDIR to CFG_BOARDDIR

Signed-off-by: Tom Rini <trini@konsulko.com>

Marek Vasut authored 2 years ago and Tom Rini committed 2 years ago

Fix the following version normalization warning:
"
/usr/lib/python3/dist-packages/setuptools/dist.py:530: UserWarning: Normalizing '2023.01' to '2023.1'
"

Using suggestion from Richard Jones:
https://github.com/pypa/setuptools/issues/308#issuecomment-405817468



Signed-off-by: Marek Vasut <marex@denx.de>
Reviewed-by: Simon Glass <sjg@chromium.org>

There is a CONFIG_MMC_SUNXI_SLOT definition in our sunxi_common.h config
header, which was used to note the first MMC controller to initialise.
The definition in that header was always set to 0, with no easy way of
overriding this, and certainly none of the existing boards made any use
of that (non-)feature.
Remove that definition and replace it with a constant 0 in the only
user, in board.c. It turns out that this is safe, as this is only used
in the SPL, and the BROM also unconditionally initialises MMC0.
This also removes the last legacy config symbol with SUN*I in it from
the whitelist.

Reviewed-by: Samuel Holland <samuel@sholland.org>
Tested-by: Samuel Holland <samuel@sholland.org>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>

Holger Brunck authored 2 years ago and Tom Rini committed 2 years ago

Instead of having these defines in a header file, move them to
a simple text file.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>

Holger Brunck authored 2 years ago and Tom Rini committed 2 years ago

Use like the other boards a text file for the environment.
As this is the last user of keymile-common.h we can now remove this
file completely.

Signed-off-by: Holger Brunck <holger.brunck@hitachienergy.com>

We borrow from the Linux Kernel 0d362be5b142 ("Makefile: link with -z
noexecstack --no-warn-rwx-segments") here to disable the RWX segment
linking warnings. We do not also bring in -z noexecstack as that
requires auditing and using ".note.GNU-stack" on assembly functions
which do need this feature. Further, we now introduce KBUILD_EFILDFLAGS
so that we can also pass --no-warn-rwx-segments when linking EFI
applications, and those do explicitly pass -z execstack.

Cc: Heinrich Schuchardt <xypron.glpk@gmx.de>
Cc: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>

Perform some deeper investigation on the remaining symbols listed in
this file and remove more.

Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

Rsync all defconfig files using moveconfig.py

Signed-off-by: Tom Rini <trini@konsulko.com>

This symbol is specific to the PowerPC SPL implementation, so rename
this to reflect that it's in SPL and used / tested there, so that we can
then safely migrate it to Kconfig.

Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

This is out-of-date now. Fix it.

Signed-off-by: Simon Glass <sjg@chromium.org>

This converts the following to Kconfig:
    CONFIG_SYS_MIPS_TIMER_REQ

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>

Resolve all uses of CONFIG_SYS_MHZ with the currently defined value.
Remove code which depends on CONFIG_SYS_MHZ but where no board configs
actually use that code.

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This has not been migrated to DM_VIDEO since 2019. Drop it.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This is not used anymore. Drop it.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This relies on the old LCD implementation which is to be removed. Drop it.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This option is not needed now that the LCD implementation is being
removed. Drop it.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This converts the following to Kconfig:
   CONFIG_VIDEO_LOGO_MAX_SIZE

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This option should not have the SYS_ in it. Drop it so it fits in with the
other video options.

Also simplify the alignment code in gunzip_bmp(), since malloc() always
returns a 32-bit-aligned pointer.

Signed-off-by: Simon Glass <sjg@chromium.org>

Simon Glass authored 2 years ago and Anatolij Gustschin committed 2 years ago

This converts the following to Kconfig:
   CONFIG_HIDE_LOGO_VERSION

Signed-off-by: Simon Glass <sjg@chromium.org>

The distutils package is deprecated. The upstream libfdt repo uses
setuptools for building the pylibfdt module, so bring in that code,
suitably modified for U-Boot. Also bring in the README.

The modifications include setting the version correctly, making use of
the environment variables provided by the Makefile and various tweaks
to the directories.

Note that the version omits the minus character at the start of
EXTRAVERSION, since this creates a warning. The build is really just used
within U-Boot itself, so it doesn't matter too much if the version matches
upstream, or exactly matches U-Boot.

Signed-off-by: Simon Glass <sjg@chromium.org>

Michal Suchanek authored 2 years ago and Simon Glass committed 2 years ago

Currently sandbox configuration defautls to 64bit and there is no
automation for building 32bit sandbox on 32bit hosts.

Use _LP64 macro as heuristic for detecting 64bit targets.

Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Reviewed-by: Simon Glass <sjg@chromium.org>

Michal Suchanek authored 2 years ago and Simon Glass committed 2 years ago

Python 3.10 requires defining PY_SSIZE_T_CLEAN. This will be fixed in
swig 4.10 but it is not clear when it will be released. There was a
warning since python 3.8.

Link: https://github.com/swig/swig/pull/2277



Signed-off-by: Michal Suchanek <msuchanek@suse.de>
Reviewed-by: Simon Glass <sjg@chromium.org>