efi_loader: Check for the native OP-TEE result on mm_communicate calls
Currently we only check for the return value of tee_invoke_func().
Although OP-TEE and StMM will correctly set param[1].u.value.a and we'll
eventually return an error, the correct thing to do is check for the
OP_TEE return code as well.
So let's check for that and move tee_shm_free() and tee_close_session()
before exiting with an error to make sure we always clear the registered
memory.
Fixes: f042e47e8fb43 ("efi_loader: Implement EFI variable handling via OP-TEE") Signed-off-by: Ilias Apalodimas <[email protected]>
Use EFI_DEVICE_ERROR for TEE communication problems. Reviewed-by: Heinrich Schuchardt <[email protected]>
efi_loader: Rename and correct values for ARM_SMC_MM_*
Instead of adding the definition for the specific MM SVC used in
StandAloneMM we added the one used in the standard SMC calls.
So change the value from -4 to -5 to match the correct one defined in
EDK2 and rename them to avoid future confusion
Fixes 23a397d2e2fb: ("efi_loader: Add headers for EDK2 StandAloneMM communication") Signed-off-by: Ilias Apalodimas <[email protected]> Reviewed-by: Heinrich Schuchardt <[email protected]>
efi_loader: don't use memmove() in efi_var_mem_del()
efi_var_mem_del() is in __efi_runtime because it would be needed for a
runtime implementation of SetVariable(). memmove() is not in __efi_runtime.
So we should not use it in efi_var_mem_del().
efi_loader: efi_current_var after SetVirtualAddressMap
Variable efi_current_var is a pointer to a physical memory address that
becomes invalid after SetVirtualAddressMap(). Instead of converting it via
ConvertPointer() simply set it to NULL.
Fixes: b02a707152dc ("efi_loader: enable UEFI variables at runtime") Signed-off-by: Heinrich Schuchardt <[email protected]>
Dave Gerlach [Thu, 16 Jul 2020 04:39:59 +0000 (23:39 -0500)]
dm: soc: Introduce soc_ti_k3 driver for TI K3 SoCs
Introduce an soc_ti_k3_driver that allows identification and selection
of SoC specific data based on the JTAG ID register for device
identification, as described for AM65x[0] and J721E[1] devices.
Dave Gerlach [Thu, 16 Jul 2020 04:39:57 +0000 (23:39 -0500)]
dm: soc: Introduce UCLASS_SOC for SOC ID and attribute matching
Introduce UCLASS_SOC to be used for SOC identification and attribute
matching based on the SoC ID info. This allows drivers to be provided
for SoCs to retrieve SoC identifying information and also for matching
device attributes for selecting SoC specific data.
This is useful for other device drivers that may need different
parameters or quirks enabled depending on the specific device variant in
use.
Dave Gerlach [Thu, 16 Jul 2020 04:39:56 +0000 (23:39 -0500)]
doc: Add new doc for soc ID driver model
Add a new documentation file for UCLASS_SOC and its usage to describe
the SoC Device ID framework that allows SoC identification and device
data matching.
Nicolas Boichat [Mon, 13 Jul 2020 02:50:00 +0000 (10:50 +0800)]
patman: Make sure sendemail.suppresscc is (un)set correctly
Setting sendemail.suppresscc to all or cccmd leads to --cc-cmd
parameter being ignored, and emails going either nowhere, or
just to the To: line maintainer.
Michal Simek [Fri, 10 Jul 2020 11:16:48 +0000 (13:16 +0200)]
ARM: rmobile: Switch back to fdtdec_setup_memory/banksize_fdt()
The commit 361377dbdbc9 ("ARM: rmobile: Merge prior-stage firmware DT
fragment into U-Boot DT on Gen3") reverted changes introduced by commit 175f5027345c ("ARM: renesas: Configure DRAM size from ATF DT fragment")
that's why there is no reason to use functions with _fdt() suffix because
parameter is gd->fdt_blob as is already for functions without _fdt()
suffix.
The unsupported Commit-xxx option are silently skipped
and removed as 're_remove=Commit-\w*', this patch adds
warning message in this case to detect misspelled issue
for the 2 supported options:
Commit-notes:
Commit-changes:
For example: the final 's' is missing (Commit-note:)
NB: no issue for Series-xxx option as only the supported
options are accepted (see valid_series in series.py)
The function fdt_find_regions look in the exclude list for each
property, even if the name is NULL. It could happen if the fit
image is corrupted. On sandbox, it generates a segfault.
To avoid this issue, if the name of a property is NULL, we report
an error and avoid looking in the exclude list.
Simon Glass [Fri, 10 Jul 2020 00:39:45 +0000 (18:39 -0600)]
binman: Add support for generating a FIT
FIT (Flat Image Tree) is the main image format used by U-Boot. In some
cases scripts are used to create FITs within the U-Boot build system. This
is not ideal for various reasons:
- Each architecture has its own slightly different script
- There are no tests
- Some are written in shell, some in Python
To help address this, add support for FIT generation to binman. This works
by putting the FIT source directly in the binman definition, with the
ability to adjust parameters, etc. The contents of each FIT image come
from sub-entries of the image, as is normal with binman.
Simon Glass [Fri, 10 Jul 2020 00:39:43 +0000 (18:39 -0600)]
mkimage: Allow updating the FIT timestamp
Normally the FIT timestamp is created the first time mkimage is run on a
FIT, when converting the source .its to the binary .fit file. This
corresponds to using the -f flag. But if the original input to mkimage is
a binary file (already compiled) then the timestamp is assumed to have
been set previously.
Add a -t flag to allow setting the timestamp in this case.
Simon Glass [Fri, 10 Jul 2020 00:39:42 +0000 (18:39 -0600)]
binman: Allow zero-length entries to overlap
Some binary blobs unfortunately obtain their position in the image from
other binary blobs, such as Intel's 'descriptor'. In this case we cannot
rely on packing to work. It is not possible to produce a valid image in
any case, due to the missing blobs.
Allow zero-length overlaps so that this does not cause any problems.
Simon Glass [Fri, 10 Jul 2020 00:39:38 +0000 (18:39 -0600)]
binman: Allow external binaries to be missing
Sometimes it is useful to build an image even though external binaries are
not present. This allows the build system to continue to function without
these files, albeit not producing valid images.
U-Boot does with with ATF (ARM Trusted Firmware) today.
Add a new flag to binman to request this behaviour.
Simon Glass [Fri, 10 Jul 2020 00:39:36 +0000 (18:39 -0600)]
binman: Add an etype for external binary blobs
It is useful to be able to distinguish between ordinary blobs such as
u-boot.bin and external blobs that cannot be build by the U-Boot build
system. If the external blobs are not available for some reason, then we
know that a value image cannot be built.
Simon Glass [Fri, 10 Jul 2020 00:39:33 +0000 (18:39 -0600)]
binman: Adjust pylibfdt for incremental build
If the pylibfdt shared-object file is detected, then Python assumes that
the libfdt.py file exists also.
Sometimes when an incremental build aborts, the shared-object file is
built but the libfdt.py is not. The only way out at this point is to use
'make mkproper', or similar.
Fix this by removing the .so file before it is built. This seems to make
Python rebuild everything.
Simon Glass [Fri, 10 Jul 2020 00:39:31 +0000 (18:39 -0600)]
binman: Add support for calling mkimage
As a first step to integrating mkimage into binman, add a new entry type
that feeds data into mkimage for processing and incorporates that output
into the image.
Simon Glass [Fri, 10 Jul 2020 00:39:30 +0000 (18:39 -0600)]
binman: Set a default toolpath
When binman is run from 'make check' it is given a toolpath so that the
latest tools (e.g. mkimage) are used. When run manually with no toolpath,
it relies on the system mkimage. But this may be missing or old.
Make some effort to find the built-from-soruce version by looking in the
current directory and in the builds created by 'make check'.
Simon Glass [Fri, 10 Jul 2020 00:39:29 +0000 (18:39 -0600)]
binman: Specify the toolpath when running test coverage
At present binman's test coverage runs without a toolpath set. This means
that the system tools will be used. That may not be correct if they are
out of date or missing and this can result in a reduction in test coverage
below 100%.
Simon Glass [Fri, 10 Jul 2020 00:39:28 +0000 (18:39 -0600)]
binman: Correct the search patch for pylibfdt
Now that binman uses tools/ as its base directory for importing modules,
the path to the pylibfdt build by U-Boot is incorrect. Fix it with a new
path.
Simon Glass [Fri, 10 Jul 2020 00:39:26 +0000 (18:39 -0600)]
binman: Output errors to stderr
At present binman outputs errors to stdout which means that fails are
effectively silent when printed by buildman, for example. Fix this by
outputing errors to stderr.
Simon Glass [Mon, 6 Jul 2020 03:41:55 +0000 (21:41 -0600)]
patman: Add a 'test' subcommand
At present we use --test to indicate that tests should be run. It is
better to use a subcommand for list, like binman. Change it and adjust
the existing code to fit under a 'send' subcommand, the default.
Give this subcommand the same default arguments as the others.
Simon Glass [Mon, 6 Jul 2020 03:41:54 +0000 (21:41 -0600)]
patman: Allow different commands
At present patman only does one thing so does not have any comments. We
want to add a few more command, so create a sub-parser for the default
command ('send').
Simon Glass [Mon, 6 Jul 2020 03:41:52 +0000 (21:41 -0600)]
patman: Allow skipping patches at the end
The -s option allows skipping patches at the top of the branch. Sometimes
there are commits at the bottom that need to be skipped. At present it is
necessary to count the number of commits and then use -c to tell patman
how many to process.
Add a -e option to easily skip a number of commits at the bottom of the
branch.
Simon Glass [Mon, 6 Jul 2020 03:41:50 +0000 (21:41 -0600)]
patman: Add a test that uses gitpython
It is convenient to use gitpython to create a real git repo for testing
patman's operation. Add a test for this. So far it just checks that patman
produces the right number of patches for a branch.
Simon Glass [Mon, 6 Jul 2020 03:41:49 +0000 (21:41 -0600)]
patman: Move main code out to a control module
To make testing easier, move the code out from main into a separate
'control' module and split it into four parts: setup, preparing patches,
checking patches and emailing patches.
Add comments and fix a few code-style issues while we are here.
Add SPINAND_HAS_CR_FEAT_BIT flag to identify the SPI NAND device with
the Continuous Read mode.
Some of the Micron SPI NAND devices have the "Continuous Read" feature
enabled by default, which does not fit the subsystem needs.
In this mode, the READ CACHE command doesn't require the starting column
address. The device always output the data starting from the first
column of the cache register, and once the end of the cache register
reached, the data output continues through the next page. With the
continuous read mode, it is possible to read out the entire block using
a single READ command, and once the end of the block reached, the output
pins become High-Z state. However, during this mode the read command
doesn't output the OOB area.
Simon Glass [Fri, 17 Jul 2020 14:48:28 +0000 (08:48 -0600)]
x86: mtrr: Restructure so command execution is in one place
At present do_mtrr() does the 'list' subcommand at the top and the rest
below. Update it to do them all in the same place so we can (in a later
patch) add parsing of the CPU number for all subcommands.
Simon Glass [Fri, 17 Jul 2020 14:48:23 +0000 (08:48 -0600)]
x86: Don't enable SMP in SPL
SMP should be set up in U-Boot where possible, not SPL. Disable it in SPL.
For 64-bit U-Boot we should find a way to allow SMP operations in U-Boot,
but this is somewhat more complicated. For now that is disabled too.
Simon Glass [Fri, 17 Jul 2020 14:48:22 +0000 (08:48 -0600)]
x86: mtrr: Use MP calls to list the MTRRs
Update the mtrr command to use mp_run_on_cpus() to obtain its information.
Since the selected CPU is the boot CPU this does not change the result,
but it sets the stage for supporting other CPUs.
Simon Glass [Fri, 17 Jul 2020 14:48:21 +0000 (08:48 -0600)]
x86: mp: Add iterators for CPUs
It is convenient to iterate through the CPUs performing work on each one
and processing the result. Add a few iterator functions which handle this.
These can be used by any client code. It can call mp_run_on_cpus() on
each CPU that is returned, handling them one at a time.
Simon Glass [Fri, 17 Jul 2020 14:48:20 +0000 (08:48 -0600)]
x86: mp: Park CPUs before running the OS
With the new MP features the CPUs are no-longer parked when the OS is run.
Fix this by calling a special function to park them, just before the OS is
started.
Simon Glass [Fri, 17 Jul 2020 14:48:19 +0000 (08:48 -0600)]
x86: mp: Allow running functions on multiple CPUs
Add a way to run a function on a selection of CPUs. This supports either
a single CPU, all CPUs, just the main CPU or just the 'APs', in Intel
terminology.
It works by writing into a mailbox and then waiting for the CPUs to notice
it, take action and indicate they are done.
When SMP is not yet enabled, this just calls the function on the main CPU.
Simon Glass [Fri, 17 Jul 2020 14:48:17 +0000 (08:48 -0600)]
global_data: Add a generic global_data flag for SMP state
Allow keeping track of whether all CPUs have been enabled yet. This allows
us to know whether other CPUs need to be considered when updating
CPU-specific settings such as MTRRs on x86.
Simon Glass [Fri, 17 Jul 2020 14:48:16 +0000 (08:48 -0600)]
x86: mp: Support APs waiting for instructions
At present the APs (non-boot CPUs) are inited once and then parked ready
for the OS to use them. However in some cases we want to send new requests
through, such as to change MTRRs and keep them consistent across CPUs.
Change the last state of the flight plan to go into a wait loop, accepting
instructions from the main CPU.
Simon Glass [Fri, 17 Jul 2020 14:48:13 +0000 (08:48 -0600)]
x86: mp_init: Set up the CPU numbers at the start
At present each CPU is given a number when it starts itself up. While this
saves a tiny amount of time by doing the device-tree read in parallel, it
is confusing that the numbering happens on the fly.
Move this code into mp_init() and do it at the start.
Simon Glass [Fri, 17 Jul 2020 14:48:08 +0000 (08:48 -0600)]
x86: Move MP code into mp_init
At present the 'flight plan' for CPUs is passed into mp_init. But it is
always the same. Move it into the mp_init file so everything is in one
place. Also drop the SMI function since it does nothing. If we implement
SMIs, more refactoring will be needed anyway.
Stefan Roese [Tue, 30 Jun 2020 10:08:58 +0000 (12:08 +0200)]
mips: octeon: Add minimal Octeon 3 EBB7304 EVK support
This patch adds very basic minimal support for the Marvell Octeon 3
CN73xx based EBB7304 EVK. Please note that the basic Octeon port does
not support DDR3/4 initialization yet. To still use U-Boot on with this
port, the L2 cache (4MiB) is used as RAM. This way, U-Boot can boot
to the prompt on this board.
Stefan Roese [Tue, 30 Jun 2020 10:33:17 +0000 (12:33 +0200)]
mips: octeon: use mips_mach_early_init() to copy to L2 cache
This patch adds the code to copy itself from bootrom location to a
different location (TEXT_BASE) to the Octeon platform. Its used in
this case to copy the complete U-Boot image into L2 cache, which
greatly improves the bootup time - especially in regard to the
very long and complex DDR4 init code.
The Kconfig symbol CONFIG_MIPS_MACH_EARLY_INIT is enabled with this
patch for Octeon.