Kever Yang [Mon, 22 Jul 2019 11:59:12 +0000 (19:59 +0800)]
rockchip: add common spl board file
The common spl board file handles board_init_f() in SPL,
and with board_early_init_f() and arch_cpu_init() callback,
other operateion after board_init_f() should go to board specific
spl_board_init().
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
include: configs: Increase CONFIG_SYS_BOOTM_LEN to 16MB
The default value of CONFIG_SYS_BOOTM_LEN is 0x800000 i.e, 8MB which
causes board reset because of larger uImage size.
This was tested on rk3288 Amarula Vyasa and rk3288 Asus Tinker
boards.
Error log snippet:
Booting using the fdt blob at 0x1f00000
Loading Kernel Image ... Image too large: increase CONFIG_SYS_BOOTM_LEN
Must RESET board to recover
resetting ...
Reviewed-by: Jagan Teki <jagan@amarulasolutions.com> Signed-off-by: Shyam Saini <shyam.saini@amarulasolutions.com> Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Sekhar Nori [Thu, 11 Jul 2019 09:00:24 +0000 (14:30 +0530)]
clk: initialize clk->data when using default xlate
Right now when using clk_of_xlate_default(), clk->data
remains un-initialized because clk_get_bulk() does not
initialize memory on allocation of clock structure.
This can cause problems when data is used to match if
two clocks pointers are exactly the same underlying
clocks, for example.
Fix it by initializing clk->data to 0.
Suggested-by: Lokesh Vutla <lokeshvutla@ti.com> Signed-off-by: Sekhar Nori <nsekhar@ti.com>
Masahiro Yamada [Thu, 27 Jun 2019 07:39:57 +0000 (16:39 +0900)]
fdt: make fdt_get_base_address() return OF_BAD_ADDR when "reg" not found
Currently, fdt_get_base_address() returns 0 if the "reg" property is
missing. Since 0 is a valid value, it is not suitable for the error
handling. Return OF_BAD_ADDR instead.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Mon, 8 Jul 2019 20:25:53 +0000 (14:25 -0600)]
binman: Add a test for nested and aligned sections
Current test coverage is likely sufficient for the logic used to place
sections in the image. However it seems useful to add a test specifically
for nested sections, since these could have some unusual interactions.
Add a new test for this and aligned sections. This test failed before the
refactor to drop the bsection.py file (Section class), but passes now.
Simon Glass [Mon, 8 Jul 2019 20:25:51 +0000 (14:25 -0600)]
binman: Support reading from CBFS entries
CBFS is a bit like a section but with a custom format. Provide the list of
entries and the compression type to binman so that it can extract the data
from the CBFS, just like any other part of the image.
Simon Glass [Mon, 8 Jul 2019 20:25:50 +0000 (14:25 -0600)]
binman: Allow reading an entry from an image
It is useful to be able to extract entry contents from an image to see
what is inside. Add a simple function to read the contents of an entry,
decompressing it by default.
Simon Glass [Mon, 8 Jul 2019 20:25:49 +0000 (14:25 -0600)]
binman: Allow for logging information to be displayed
Binman generally operates silently but in some cases it is useful to see
what Binman is actually doing at each step. Enable some logging output
with different logging levels selectable via the -v flag.
Simon Glass [Mon, 8 Jul 2019 20:25:47 +0000 (14:25 -0600)]
binman: Convert Image to a subclass of Entry
When support for sections (and thus hierarchical images) was added to
binman, the decision was made to create a new Section class which could
be used by both Image and an Entry_section class. The decision between
using inheritance and composition was tricky to make, but in the end it
was decided that Image was different enough from Entry that it made sense
to put the implementation of sections in an entirely separate class. It
also has the advantage that core Image code does have to rely on an entry
class in the etype directory.
This work was mostly completed in commit:
8f1da50ccc "binman: Refactor much of the image code into 'section'
As a result of this, the Section class has its own version of things like
offset and size and these must be kept in sync with the parent
Entry_section class in some cases.
In the last year it has become apparent that the cost of keeping things in
sync is larger than expected, since more and more code wants to access
these properties.
An alternative approach, previously considered and rejected, now seems
better.
Adjust Image to be a subclass of Entry_section. Move the code from Section
(in bsection.py) to Entry_section and delete Section. Update all tests
accordingly.
This requires substantial changes to Image. Overall the changes reduce
code size by about 240 lines. While much of that is just boilerplate from
Section, there are quite a few functions in Entry_section which now do not
need to be overiden from Entry. This suggests the change is beneficial
even without further functionality being added.
A side benefit is that the properties of sections are now consistent with
other entries. This fixes a problem in testListCmd() where some properties
are missing for sections.
Unfortunately this is a very large commit since it is not feasible to do
the migration piecemeal. Given the substantial tests available and the
100% code coverage of binman, we should be able to do this safely.
Simon Glass [Mon, 8 Jul 2019 20:25:46 +0000 (14:25 -0600)]
binman: Support reading an image into an Image object
It is possible to read an Image, locate its FDT map and then read it into
the binman data structures. This allows full access to the entries that
were written to the image. Add support for this.
Simon Glass [Mon, 8 Jul 2019 20:25:43 +0000 (14:25 -0600)]
binman: Allow listing the entries in an image
It is useful to be able to summarise all the entries in an image, e.g. to
display this to this user. Add a new ListEntries() method to Entry, and
set up a way to call it through the Image class.
Simon Glass [Mon, 8 Jul 2019 20:25:40 +0000 (14:25 -0600)]
binman: Use the cbfs memlen field only for uncompressed length
The purpose of this badly named field is a bit ambiguous. Adjust the code
to use it only to store the uncompressed length of a file, leaving it set
to None if there is no compression used. This makes it easy to see if the
value in this field is relevant / useful.
Also set data_len for compressed fields, since it should be the length of
the compressed data, not the uncompressed data.
Simon Glass [Mon, 8 Jul 2019 20:25:39 +0000 (14:25 -0600)]
binman: Provide the actual data address for cbfs files
At present a file with no explicit CBFS offset is placed in the next
available location but there is no way to find out where it ended up.
Update and rename the get_data() function to provide this information.
Simon Glass [Mon, 8 Jul 2019 20:25:36 +0000 (14:25 -0600)]
binman: Add a control for post-pack entry expansion
We plan to support changing the size of entries after they have been
packed. For now it will always be enabled. But to aid testing of both
cases (in the event that we want to add a command-line flag, for example),
add a setting to control it.
Simon Glass [Mon, 8 Jul 2019 20:25:35 +0000 (14:25 -0600)]
binman: Add a return value to ProcessContentsUpdate()
At present if this function tries to update the contents such that the
size changes, it raises an error. We plan to add the ability to change
the size of entries after packing is completed, since in some cases it is
not possible to determine the size in advance.
An example of this is with a compressed device tree, where the values
of the device tree change in SetCalculatedProperties() or
ProcessEntryContents(). While the device tree itself does not change size,
since placeholders for any new properties have already bee added by
AddMissingProperties(), we cannot predict the size of the device tree
after compression. If a value changes from 0 to 0x1234 (say), then the
compressed device tree may expand.
As a first step towards supporting this, make ProcessContentsUpdate()
return a value indicating whether the content size is OK. For now this is
always True (since otherwise binman raises an error), but later patches
will adjust this.
Simon Glass [Mon, 8 Jul 2019 20:25:34 +0000 (14:25 -0600)]
binman: Call ProcessUpdateContents() consistently
SetContents() should only be called to set the contents of an entry from
within the ObtainContents() call, since it has no guard against increasing
the size of the contents, thus triggering incorrect operation.
Change all such calls to use ProcessUpdateContents() instead.
Simon Glass [Mon, 8 Jul 2019 20:25:33 +0000 (14:25 -0600)]
binman: Fix up ProcessUpdateContents error and comments
This function raises an exception with its arguments around the wrong way
so the message is incorrect. Fix this as well as a few minor comment
problems.
Simon Glass [Mon, 8 Jul 2019 20:25:32 +0000 (14:25 -0600)]
binman: Allow easy importing of entry modules
At present entry modules can only be accessed using Entry.Lookup() or
Entry.Create(). Most of the time this is fine, but sometimes a module
needs to provide constants or helper functions useful to other modules.
It is easier in this case to use 'import'.
Simon Glass [Mon, 8 Jul 2019 20:25:30 +0000 (14:25 -0600)]
binman: Move compression into the Entry base class
Compression is currently available only with blobs. However we want to
report the compression algorithm and uncompressed size for all entries,
so that other entry types can support compression. This will help with
the forthcoming 'list' feature which lists entries in the image.
Move the compression properties into the base class. Also fix up the docs
which had the wrong property name.
Simon Glass [Mon, 8 Jul 2019 20:25:29 +0000 (14:25 -0600)]
binman: Convert to use ArgumentParser
This class is the new way to handle arguments in Python. Convert binman
over to use it. At the same time, introduce commands so that we can
separate out the different parts of binman functionality.
Simon Glass [Mon, 8 Jul 2019 20:25:28 +0000 (14:25 -0600)]
binman: Add an image header
It is useful to be able to quickly locate the FDT map in the image. An
easy way to do this is with a pointer at the start or end of the image.
Add an 'image header' entry, which places a magic number followed by a
pointer to the FDT map. This can be located at the start or end of the
image, or at a chosen location.
As part of this, update GetSiblingImagePos() to detect missing siblings.
Simon Glass [Mon, 8 Jul 2019 20:25:27 +0000 (14:25 -0600)]
binman: Add an FDT map
An FDT map is an entry which holds a full description of the image
entries, in FDT format. It can be discovered using the magic string at
its start. Tools can locate and read this entry to find out what entries
are in the image and where each entry is located.
Simon Glass [Mon, 8 Jul 2019 20:25:23 +0000 (14:25 -0600)]
binman: Simplify the entry test
The current test for the 'entry' module is a bit convoluted since it has
to import the module multiple times. It also relies on ordering, in that
test1EntryNoImportLib() must run before test2EntryImportLib() if they are
running in the same Python process.
This is unreliable since neither the ordering of tests nor the process
that they run in is defined.
Fix this by always reloading the entry in these two tests. Also add a
check that the expected value of have_importlib is obtained.
This corrects a code-coverage problem in the 'entry' module on some
systems.
Simon Glass [Mon, 8 Jul 2019 19:18:56 +0000 (13:18 -0600)]
binman: Add support for fixed-offset files in CBFS
A feature of CBFS is that it allows files to be positioned at particular
offset (as with binman in general). This is useful to support
execute-in-place (XIP) code, since this may not be relocatable.
Simon Glass [Mon, 8 Jul 2019 19:18:55 +0000 (13:18 -0600)]
binman: Pad empty areas of the CBFS with files
When there is lots of open space in a CBFS it is normally padded with
'empty' files so that sequentially scanning the CBFS can skip from one to
the next without a break.
Simon Glass [Mon, 8 Jul 2019 19:18:52 +0000 (13:18 -0600)]
binman: Add a utility library for coreboot CBFS
Coreboot uses a simple flash-based filesystem called Coreboot Filesystem
(CBFS) to organise files used during boot. This allows files to be named
and their position in the flash to be set. It has special features for
dealing with x86 devices which typically memory-map their SPI flash to the
top of 32-bit address space and need a 'boot block' ending there.
Create a library to help create and read CBFS files. This includes a
writer class, a reader class and associated other helpers. Only a subset
of features are currently supported.
Simon Glass [Mon, 8 Jul 2019 19:18:50 +0000 (13:18 -0600)]
binman: Pass the toolpath to tests
Tools like ifwitool may not be available in the PATH, but are available in
the build. These tools may be needed by tests, so allow tests to use the
--toolpath flag.
Patrick Delaunay [Mon, 22 Jul 2019 12:19:20 +0000 (14:19 +0200)]
psci: Fix warnings when compiling with W=1
This patch solves the following warnings:
arch/arm/mach-stm32mp/psci.c:
warning: no previous prototype for ‘psci_set_state’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_arch_cpu_entry’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_features’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_version’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_affinity_info’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_migrate_info_type’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_cpu_on’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_cpu_off’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_system_reset’ [-Wmissing-prototypes]
warning: no previous prototype for ‘psci_system_off’ [-Wmissing-prototypes]
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Suman Anna [Fri, 19 Jul 2019 15:27:56 +0000 (10:27 -0500)]
remoteproc: Fix potential build issues with SPL remoteproc
The rproc uclass driver can either be built with SPL_REMOTEPROC
or REMOTEPROC, but the function prototypes in remoteproc.h are
defined only when CONFIG_REMOTEPROC is defined. This can cause
build issues in SPL if CONFIG_REMOTEPROC is not selected.
Fix this by replacing the existing precompiler macro usage with
CONFIG_IS_ENABLED.
Fixes: ddf56bc7e3ef ("drivers: Introduce a simplified remoteproc framework") Signed-off-by: Suman Anna <s-anna@ti.com> Reviewed-by: Nishanth Menon <nm@ti.com> Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Jonathan Corbet [Sun, 14 Jul 2019 08:35:45 +0000 (10:35 +0200)]
doc: Cope with Sphinx logging deprecations
Recent versions of sphinx will emit messages like:
doc/sphinx/kerneldoc.py:103:
RemovedInSphinx20Warning: app.warning() is now deprecated.
Use sphinx.util.logging instead.
Switch to sphinx.util.logging to make this unsightly message go away.
Alas, that interface was only added in version 1.6, so we have to add a
version check to keep things working with older sphinxes.
Signed-off-by: Jonathan Corbet <corbet@lwn.net>
Rebased for U-Boot Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
'ab_select' command is used to decide which slot should be used for
booting up. A/B metadata resides in 'misc' partition.
To activate the A/B boot process, the following config options must be
set:
CONFIG_ANDROID_AB=y
CONFIG_CMD_AB_SELECT=y
For successful A/B boot, the corresponding A/B infrastructure must be
involved on Android side [1] (including mounting system as root), and
disk must be partitioned accordingly.
When A/B boot is enabled, there are some known limitations currently
exist (not related to A/B patches, need to be implemented later):
1. The 'Verified Boot' sequence is not supported
2. dev path to system partition (system_a or system_b) is passed via
'bootargs' as 'root=' argument like 'root=/dev/mmcblk1p12', but
further we'll need to rework it with respect to dm-verity
requirements [2]
In case when A/B partitions are not present in system (and A/B boot is
enabled), boot up process will be terminated and next message will be
shown:
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
Add a short documentation for A/B enablement and 'ab_select' command
usage.
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Signed-off-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
For A/B system update support the Android boot process requires to send
'androidboot.slot_suffix' parameter as a command line argument. This
patch implementes 'ab_select' command which allows us to obtain current
slot by processing the A/B metadata.
The patch was extracted from commit [1] with one modification: the
separator for specifying the name of metadata partition was changed
from ';' to '#', because ';' is used for commands separation.
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
This patch determines the A/B-specific bootloader message structure
that is the basis for implementation of recovery and A/B update
functions. A/B metadata is stored in this structure and used to decide
which slot should we use to boot the device. Also some basic functions
for A/B metadata manipulation are implemented (like slot selection).
The patch was extracted from commits [1], [2] with some coding style
fixes.
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
This patch adds part_get_info_by_dev_and_name_or_num() function which
allows us to get partition info from its number or name. Partition of
interest is specified by string like "device_num:partition_number" or
"device_num#partition_name".
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org>
These have been reported by Simon in [1] and fixed in [2].
However, since [1] has already been pushed to u-boot/master, the
improvements incorporated in [2] are now extracted and resubmitted.
The changes are in the area of coding style and best practices:
* s/field/fieldp/, s/size/sizep/, to convey that the variables return
an output to the caller
* s/err_1/err_read_fail/, s/err_2/err_too_small/, to be more descriptive
* Made sure 'static int do_bcb_load' appears on the same line
* Placed a `/*` on top of multi-line comment
cmd: bcb: Use strcmp() instead of strncmp() for string literals
Quote from https://patchwork.ozlabs.org/patch/1104244/#2210814:
----------8<-----------
strncmp() is chosen for the sake of paranoid/defensive programming.
Indeed, strncmp() is not really needed when comparing a variable
with a string literal. We expect strcmp() to behave safely even if the
string variable is not NUL-terminated.
In the same scenario, Linux v5.2-rc7 uses both strcmp() and strncmp(),
but the frequency of strcmp() is higher:
At this moment only -C option is handled, but specifying -A and -T options
are recommended for compatibility reasons in the future.
Kernel that compressed with LZ4 can be already used without wrapping into
the uImage, but I recommend wrapping it into the uImage in order to avoid
situations when by some mistake legacy LZ4 is used, that is interpreted as
raw Image and causes CPU to enter Exception Handler without providing any
meaningful explanation to the user.
Signed-off-by: Roman Stratiienko <roman.stratiienko@globallogic.com>
Leon Yu [Fri, 21 Jun 2019 04:12:39 +0000 (12:12 +0800)]
menu: don't bother going interactive with just one menu item
If there is only one menu item available, prompting user to enter
choice makes little sense and just causes unnecessary boot delay. This
change makes menu_get_choice return the only one item when there is no
other choices.
Signed-off-by: Leon Yu <leoyu@nvidia.com> Cc: Tom Warren <twarren@nvidia.com> Cc: Stephen Warren <swarren@nvidia.com> Cc: Thierry Reding <treding@nvidia.com>
This sub-command serves for getting the partition index from
partition name. Also it can be used to test the existence of specified
partition.
Use case:
For example, in most CI environments this U-Boot command for automatic
testing of Linux rootfs is used:
=> setenv bootpart 1:f
where 0xf is "userdata" partition. But the number of "userdata"
partition can be changed any time, when partition table is changed.
So it would be nice to get rid of that 0xf magic number and use
partition name instead, like this:
=> part number mmc 1 userdata part_num
=> setenv bootpart 1:${part_num}
Signed-off-by: Ruslan Trofymenko <ruslan.trofymenko@linaro.org> Signed-off-by: Igor Opaniuk <igor.opaniuk@gmail.com> Reviewed-by: Alistair Strachan <astrachan@google.com> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org> Reviewed-by: Simon Glass <sjg@chromium.org> Reviewed-by: Sam Protsenko <semen.protsenko@linaro.org>