imx8m: config: convert to bootm_size

Restrict the memory range available for image processing in the
"bootm" to 256 MiB so the kernel can access it and FDT or initrd are
not overwritten on ARM64.

Signed-off-by: Grygorii Tertychnyi <grygorii.tertychnyi@leica-geosystems.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Marek Vasut <marex@denx.de>
Cc: Andrey Zhizhikin <andrey.zhizhikin@leica-geosystems.com>
Reviewed-by: Tom Rini <trini@konsulko.com>

ARM: imx6q_logic: Fix broken booting by moving fdt_addr_r address

The loading address is too close to the kernel address, so newer kernels
may overlap memory space, so loading the device tree may corrupt zImage.

This patch moves the fdt_addr_r to 0x14000000 which is also consistent
with guidance that the kernel be allocated 32MB.  This places it
in the same place as the ramdisk, so this patch moves the ramdisk address
512KB after the fdt.

Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>

pico-imx6: Remove unneeded CONFIG_DM_MDIO

As explained in the CONFIG_DM_MDIO text inside drivers/net/Kconfig:

"Useful in particular for systems that support
DM_ETH and have a stand-alone MDIO hardware block shared by multiple
Ethernet interfaces."

i.MX6 has a single FEC instance, so there is no need to select
CONFIG_DM_MDIO.

Signed-off-by: Fabio Estevam <festevam@gmail.com>

arm: mx6: Make all i.MX6 SoCs user-selectable

We have a number of platforms that are a combination of a carrier board
and System-on-Module (SoM) that in turn allows for the board to have
different SoCs on it.  In some cases, this is handled via board-specific
Kconfig options.  In other cases we make use of
CONFIG_SYS_EXTRA_OPTIONS.  This latter case however can lead to invalid
configurations as we will not in turn get options that in Kconfig are
selected by or depend on that setting.

To resolve this, make the SoC option a choice in Kconfig and make boards
depend on what they can support.  This change opens us up for further
clean-ups in the cases where a single CONFIG_TARGET_xxx can support
different SoCs and today they do not, or do not cleanly do so.

Reported-by: Matt Porter <mporter@konsulko.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: "NXP i.MX U-Boot Team" <uboot-imx@nxp.com>
Cc: Soeren Moch <smoch@web.de>
Cc: Markus Niebel <Markus.Niebel@tq-group.com>
Cc: Igor Opaniuk <igor.opaniuk@toradex.com>
Cc: Heiko Schocher <hs@denx.de>
Cc: Hannes Schmelzer <hannes.schmelzer@br-automation.com>
Cc: Otavio Salvador <otavio@ossystems.com.br>
Cc: Nikita Kiryanov <nikita@compulab.co.il>
Cc: Andreas Geisreiter <ageisreiter@dh-electronics.de>
Cc: Ludwig Zenz <lzenz@dh-electronics.de>
Cc: Lukasz Majewski <lukma@denx.de>
Cc: Akshay Bhat <akshaybhat@timesys.com>
Cc: Ken Lin <Ken.Lin@advantech.com.tw>
Cc: Ian Ray <ian.ray@ge.com>
Cc: Tim Harvey <tharvey@gateworks.com>
Cc: Jagan Teki <jagan@amarulasolutions.com>
Cc: Raffaele RECALCATI <raffaele.recalcati@bticino.it>
Cc: Simone CIANNI <simone.cianni@bticino.it>
Cc: Adam Ford <aford173@gmail.com>
Cc: Marcin Niestroj <m.niestroj@grinn-global.com>
Cc: "Eric Bénard" <eric@eukrea.com>
Cc: Baruch Siach <baruch@tkos.co.il>
Cc: Jason Liu <jason.hui.liu@nxp.com>
Cc: Ye Li <ye.li@nxp.com>
Cc: Eric Nelson <eric@nelint.com>
Cc: Troy Kisky <troy.kisky@boundarydevices.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Parthiban Nallathambi <parthiban@linumiz.com>
Cc: Marek Vasut <marex@denx.de>
Cc: "Sébastien Szymanski" <sebastien.szymanski@armadeus.com>
Cc: Christian Gmeiner <christian.gmeiner@gmail.com>
Cc: Niel Fourie <lusus@denx.de>
Cc: Martyn Welch <martyn.welch@collabora.com>
Cc: Richard Hu <richard.hu@technexion.com>
Cc: Stefan Roese <sr@denx.de>
Cc: Boris Brezillon <bbrezillon@kernel.org>
Cc: Arkadiusz Karas <arkadiusz.karas@somlabs.com>
Cc: Breno Lima <breno.lima@nxp.com>
Cc: Francesco Montefoschi <francesco.montefoschi@udoo.org>
Cc: Silvio Fricke <open-source@softing.de>
Tested-by: Matt Porter <mporter@konsulko.com> [colibri_imx6]
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Marcin Niestroj <m.niestroj@grinn-global.com>

pico-imx6ul: convert ethernet function to DM_ETH

- Remove pinmux definition from pico-imx6ul.c
- Enable NET_RANDOM_ETHADDR for temporary solution, because micrel_ksz8xxx
driver does not support DM_ETH yet, so cannot read MAC address directly.

Before enable DM_ETH:
  Net:   FEC [PRIME]

After enable DM_ETH:
  Net:
  Warning: using random MAC address - ca:3f:43:8f:67:d4
  eth1: ethernet@20b4000

Here is the test commands:
  => dhcp
  BOOTP broadcast 1
  DHCP client bound to address 10.88.88.94 (139 ms)
  *** ERROR: `serverip' not set
  Cannot autoload with TFTPGET
  => ping 8.8.8.8
  Using ethernet@20b4000 device
  host 8.8.8.8 is alive

Signed-off-by: Wig Cheng <wig.cheng@technexion.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>

configs: pico-imx6ul: convert DM_VIDEO

It's due to the warning messages issue after compiled:

  ===================== WARNING ======================
  This board does not use CONFIG_DM_VIDEO Please update
  the board to use CONFIG_DM_VIDEO before the v2019.07 release.
  Failure to update by the deadline may result in board removal.
  See doc/driver-model/migration.rst for more info.
  ====================================================

Enable DM_VIDEO can fix it on the stage of compilation.

Signed-off-by: Wig Cheng <wig.cheng@technexion.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>

configs: pico-imx6ul: convert DM_USB

Here is the test commands:

  => ums 0 mmc 0
  UMS: LUN 0, dev 0, hwpart 0, sector 0x0, count 0x710000

Signed-off-by: Wig Cheng <wig.cheng@technexion.com>
Reviewed-by: Fabio Estevam <festevam@gmail.com>

tools/imximage: fix DCD Blocks message output order

The correct order is load address, offset, length. The order was
accidentally switched a while ago; make it match the HAB Blocks output and
what CST expects again.

Fixes: e97bdfa5da70 ("tools/imximage: share DCD information via Kconfig")
Signed-off-by: Matthias Schiffer <matthias.schiffer@ew.tq-group.com>

ARM: imx: ddr: Add deskew register programming

Fill is code for programming the DDR_PHY_CMD_DESKEW_CONx registers,
which are optional, but can be used to fill in the byte lane delays.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP i.MX U-Boot Team <uboot-imx@nxp.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>

ARM: imx: Add support for reading out the primary/secondary bmode to MX7

Implement the 'getprisec' subcommand of 'bmode' command for i.MX7 by
reading out the SRC GPR10 bit 30. This bit is either set by the BootROM
if it switched to the secondary copy due to primary copy being corrupted
OR it can be overridden by the user.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP i.MX U-Boot Team <uboot-imx@nxp.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>

ARM: imx: Add support for reading out the primary/secondary bmode

Add new 'getprisec' subcommand to 'bmode' command, which sets the return
value of the 'bmode' command to either 0 if the system booted from primary
copy or to 1 if the system booted from secondary copy. This can be used
e.g. in 'test' command to determine which copy of the system is running.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP i.MX U-Boot Team <uboot-imx@nxp.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>

ARM: imx: Add support for switching primary/secondary boot mode to bmode

The i.MX6/i.MX7 is capable of booting a secondary "redundant" system
image in case the primary one is corrupted. The user can force this
boot mode as well by explicitly setting SRC GPR10 bit 30. This can be
potentially useful when upgrading the bootloader itself. Expose this
functionality to the user.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP i.MX U-Boot Team <uboot-imx@nxp.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Reviewed-by: Stefano Babic <sbabic@denx.de>

ARM: imx: Add bmode support for iMX7

Add the basic differentiation between i.MX6 and i.MX7 into the bmode
command, the mechanism really works almost the same on both platforms.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Fabio Estevam <festevam@gmail.com>
Cc: NXP i.MX U-Boot Team <uboot-imx@nxp.com>
Cc: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>

Merge tag 'efi-2020-10-rc3-2' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi

Pull request for UEFI sub-system for efi-2020-10-rc3 (2)

This series includes bug fixes for:

* UEFI secure boot - images with multiple signatures
* UEFI secure boot - support for intermediate certificates
* corrections for UEFI unit tests
* missing loadaddr on MAIX board

Merge branch '2020-08-14-assorted-updates'

- Xen guest and some paravirt driver support.
- Aspeed SoC updates
- Broadcom IPROC PCIe RC driver

cmd: demo: Remove duplicated help message for list subcommand

There is no need to show demo list description twice when help demo is
performed. The patch removes duplicated entry.

Current state:
=> help demo
demo - Driver model (dm) demo operations

Usage:
demo list                     List available demo devices
demo hello <num> [<char>]     Say hello
demo light [<num>]            Set or get the lights
demo status <num>             Get demo device status
demo list                     List available demo devices

Fixes: a02af4aeece4 ("dm: demo: Add a simple GPIO demonstration")
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

doc: xen: Add Xen guest ARM64 board documentation

Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

board: xen: De-initialize before jumping to Linux

Free resources used by Xen board before jumping to Linux kernel.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

xen: pvblock: Print found devices indices

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

xen: pvblock: Implement front-back protocol and do IO

Implement Xen para-virtual frontend to backend communication
and actually read/write disk data.

This is based on mini-os implementation of the para-virtual block
frontend driver.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

xen: pvblock: Read XenStore configuration and initialize

Read essential virtual block device configuration data from XenStore,
initialize front ring and event channel.
Update block device description with actual block size.

Use code for XenStore from mini-os.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

xen: pvblock: Enumerate virtual block devices

Enumerate Xen virtual block devices found in XenStore and
instantiate pvblock devices.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

xen: pvblock: Add initial support for para-virtualized block driver

Add initial infrastructure for Xen para-virtualized block device.
This includes compile-time configuration and the skeleton for
the future driver implementation.
Add new class UCLASS_PVBLOCK which is going to be a parent for
virtual block devices.
Add new interface type IF_TYPE_PVBLOCK.

Implement basic driver setup by reading XenStore configuration.

Signed-off-by: Andrii Anisov <andrii_anisov@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>

xen: Port Xen grant table driver from mini-os

Make required updates to run on u-boot.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

xen: Port Xen bus driver from mini-os

Make required updates to run on u-boot and strip test code.

Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>

lib: sscanf: add sscanf implementation

Port sscanf implementation from mini-os and introduce new
Kconfig option to enable it: CONFIG_SSCANF. Disable by default.

Signed-off-by: Andrii Anisov <andrii_anisov@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>

linux/compat.h: Add wait_event_timeout macro

Add  wait_event_timeout - sleep until a condition gets true or a
timeout elapses.

This is a stripped version of the same from Linux kernel with the
following u-boot specific modifications:
- no wait queues supported
- use u-boot timer to detect timeouts
- check for Ctrl-C pressed during wait

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
[trini: Drop atomic_read from gadget/ether.c as this has existed for a
        while and now causes problems]
Signed-off-by: Tom Rini <trini@konsulko.com>

serial: serial_xen: Add Xen PV serial driver

Add support for Xen para-virtualized serial driver. This
driver fully supports serial console for the virtual machine.

Please note that as the driver is initialized late, so no banner
nor memory size is visible.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

xen: Port Xen event channel driver from mini-os

Make required updates to run on u-boot. Strip functionality
not needed by U-boot.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

xen: Port Xen hypervisor related code from mini-os

Port hypervisor related code from Mini-OS. This is referencing the code
of Mini-OS from [1] by Huang Shijie and Volodymyr Babchuk which is for
ARM64.
Update essential arch code to support required bit operations, memory
barriers etc.

Copyright for the bits ported belong to at least the following authors,
please see related files for details:

Copyright (c) 2002-2003, K A Fraser
Copyright (c) 2005, Grzegorz Milos, gm281@cam.ac.uk,Intel Research Cambridge
Copyright (c) 2014, Karim Allah Ahmed <karim.allah.ahmed@gmail.com>

[1] - https://github.com/zyzii/mini-os.git

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
[trini: Drop wmb() from musb-net/linux-compat.h now]
Signed-off-by: Tom Rini <trini@konsulko.com>

board: Introduce xenguest_arm64 board

Introduce a minimal Xen guest board running as a virtual
machine under Xen Project's hypervisor [1], [2].

Part of the code is ported from Xen mini-os and also uses
work initially done by different authors from NXP: please see
relevant files for their copyrights.

[1] https://xenbits.xen.org
[2] https://wiki.xenproject.org/

Signed-off-by: Andrii Anisov <andrii_anisov@epam.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

riscv: load addresses for Sipeed MAIX

Define default load addresses and the device tree name for the Sipeed MAIX.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

xen: Add essential and required interface headers

Add essential and required Xen interface headers only taken from
the stable Linux kernel stable/linux-5.7.y at commit
66dfe4522160 Linux 5.7.5.

These are better suited for U-boot than the original headers
from Xen as they are the stripped versions of the same.

At the same time use public protocols from Xen RELEASE-4.13.1, at
commit 6278553325a9 update Xen version to 4.13.1
as those have more comments in them.

Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>
Acked-by: Peng Fan <peng.fan@nxp.com>

Kconfig: Introduce CONFIG_XEN

Introduce CONFIG_XEN to make U-Boot could be used as bootloader
for a virtual machine.

Without bootloader, we could successfully boot up android on XEN, but
we need need bootloader to support A/B, dm verify and etc.

Signed-off-by: Peng Fan <peng.fan@nxp.com>
Signed-off-by: Oleksandr Andrushchenko <oleksandr_andrushchenko@epam.com>
Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

Add MIT License

Signed-off-by: Anastasiia Lukianenko <anastasiia_lukianenko@epam.com>

configs: evb-ast2500: Convert to OF_SEPARATE

Switch DTB provider form OF_EMBED to OF_SEPARATE
to avoid the compile warning message:

  ==================== WARNING ======================
  CONFIG_OF_EMBED is enabled. This option should only
  be used for debugging purposes. Please use
  CONFIG_OF_SEPARATE for boards in mainline.
  See doc/README.fdt-control for more info.
  ====================================================

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

configs: evb-ast2500: Move BOOTCOMMAND from header to defconfig

Move the BOOTCOMMAND definition from the board inclusion
header to the default configuration file.

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

include/configs: aspeed: Remove hardcoded variables

The hardcoded platform variables such as DRAM base address are not
common to Aspeed SoCs AST24xx/AST25xx/AST26xx. This patch replaces
those hardcoded with macros defined in a newly added header, where
the basic SoC HW information are assigned accordingly.

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

cosmetic: aspeed: ast2500: Rename board file

Rename the ast2500-board.c to board_common.c and
place the renamed file under the ast2500 folder.

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

aspeed: ast2500: Add lowlevel_init assembly

The original lowlevel_init function of AST2500 is written
in C. However, the C runtime environment is not ready until
_main execution.

This patch adds the assembly version of the lowlevel_init
function. Additional initialization to DRAM configuration
and LPC reset source are also added.

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

MAINTAINERS: Add maintainers for Aspeed SoCs

Update maintainers for Aspeed SoC platforms.

Signed-off-by: Chia-Wei, Wang <chiawei_wang@aspeedtech.com>

drivers: pcie: add Broadcom IPROC PCIe RC driver

Add support for IPROC PAXC PCIe RC driver.

Signed-off-by: Srinath Mannam <srinath.mannam@broadcom.com>
Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
Reviewed-by: Simon Glass <sjg@chromium.org>

drivers: pci: add api to get dma regions

Add api to get dma regions.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>

Merge https://gitlab.denx.de/u-boot/custodians/u-boot-riscv

- Fix HiFive Unleashed the broken problem by call fix_fdt() before
  reserve_fdt().  Please refer to
  https://www.mail-archive.com/u-boot@lists.denx.de/msg379444.html for
  master u-boot broken for HiFive Unleashed.
- Add unaligned exception cmd.
- Refine sifive/fu540 spl flow.
- Add  additional crash information for efi.
- Update sipeed/maix doc.
- Two minor refine.

test/py: efi_secboot: modify 'multiple signatures' test case

The test case 5 in test_signed (multiple signatures) must be modified
and aligned with the change introduced in the previous commit
("efi_loader: signature: correct a behavior against multiple signatures").

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>

efi_loader: signature: correct a behavior against multiple signatures

Under the current implementation, all the signatures, if any, in
a signed image must be verified before loading it.

Meanwhile, UEFI specification v2.8b section 32.5.3.3 says,
    Multiple signatures are allowed to exist in the binary’s certificate
    table (as per PE/COFF Section “Attribute Certificate Table”). Only
    one hash or signature is required to be present in db in order to pass
    validation, so long as neither the SHA-256 hash of the binary nor any
    present signature is reflected in dbx.

This patch makes the semantics of signature verification compliant with
the specification mentioned above.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

common/board_f: make sure to call fix_fdt() before reserve_fdt()

There may be a chance that board specific fix_fdt() will change the
size of FDT blob so it's safe to call reserve_fdt() after fix_fdt()
otherwise global data (gd) will overwrite with FDT blob values.

Fixes: a8492e25ac71 ("riscv: Expand the DT size before copy reserved memory node")

Signed-off-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Atish Patra <atish.patra@wdc.com>

riscv: additional crash information

If an exception occurs, the relocated program counter and return address
are required for an analysis.

With this patch you get:

    => exception undefined

    Unhandled exception: Illegal instruction
    EPC: 0000000080595908 RA: 000000008059c0c6 TVAL: 000000008030c01e
    EPC: 0000000080007908 RA: 000000008000e0c6 reloc adjusted

We can use the relocated addresses to find the involved functions in
u.boot.map:

    .text.do_undefined
                0x0000000080007908        0x8 cmd/built-in.o
    .text.cmd_process
                0x000000008000dfcc      0x11a common/built-in.o
                0x000000008000dfcc                cmd_process

If an exception occurs in an UEFI binary additionally the load addresses of
the UEFI binaries are needed. With this patch:

    => setenv efi_selftest exception
    => bootefi selftest

    Unhandled exception: Illegal instruction
    EPC: 000000008042e18a RA: 000000008042e18a TVAL: 000000008030c01e
    EPC: 000000007fea018a RA: 000000007fea018a reloc adjusted

    UEFI image [0x0000000000000000:0xffffffffffffffff] '/\selftest'
    UEFI image [0x000000008042e000:0x000000008042e43f] pc=0x18a '/bug.efi'

The value pc=0x18a matches the position of the illegal instruction in
efi_selftest_miniapp_exception.efi (loaded as /bug.efi);

    asm volatile (".word 0xffffffff\n");

    00000180   93 85 C5 11  1C 64 22 85  82 97 FF FF  FF FF 1C 64

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Sean Anderson <seanga2@gmail.com>
Tested-by: Sean Anderson <seanga2@gmail.com>
Reviewed-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>

cmd: exception: unaligned data access on RISC-V

The command 'exception' can be used to test the handling of exceptions.

Currently the exception command only allows to create an illegal
instruction exception on RISC-V.

Provide a sub-command 'exception unaligned' to cause a misaligned load
address exception.

Adjust the online help for 'exception undefined'.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Rick Chen <rick@andestech.com>

doc: riscv: debug UART for MAIX

Provide the required settings for the debug UART.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Sean Anderson <seanga2@gmail.com>

doc: riscv: Update documentation for Sipeed MAIX boards

The MAIXDUINO runs fine with the sipeed_maix_bitm_defconfig but a different
board id parameter should be passed to kflash.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Sean Anderson <seanga2@gmail.com>

riscv: sifive: fu540: redundant initialization

We should not initialize a variable if the value is overwritten before
being read.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Reviewed-by: Rick Chen <rick@andestech.com>

riscv: remove redundant logical constraint.

After

    if (ret) return ret;

we know that ret is zero. Don't check it again.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>
Reviewed-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>

riscv: sifive/fu540: Move SPL related functions to spl.c

It's better to keep all SPL related functions in the same spl.c.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Leo Liang <ycliang@andestech.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

riscv: sifive/fu540: Drop NET_RANDOM_ETHADDR

This option was enabled during the earlier U-Boot porting time. Now
we already have the OTP driver in place and the unique MAC address
is read from the OTP, there is no need to turn on this option.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Leo Liang <ycliang@andestech.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

riscv: sifive/fu540: kconfig: Move FU540 driver related options to the SoC level

All FU540 driver related options should be in the SoC level Kconfig.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

riscv: sifive/fu540: spl: Rename soc_spl_init()

spl_soc_init() seems to be a better name, as all SPL functions
names start from the spl_ prefix.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

riscv: sifive/fu540: spl: Drop our own version of board_init_f()

Use the generic board_init_f() provided by the RISC-V library codes.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

riscv: Call spl_board_init_f() in the generic SPL board_init_f()

The generic SPL version of board_init_f() should give a call to
board specific codes to initialize board in the SPL phase.

Signed-off-by: Bin Meng <bin.meng@windriver.com>
Reviewed-by: Rick Chen <rick@andestech.com>
Reviewed-by: Pragnesh Patel <pragnesh.patel@sifive.com>
Tested-by: Pragnesh Patel <pragnesh.patel@sifive.com>

efi_loader: variable: fix secure state initialization

Under the new file-based variable implementation, the secure state
is always and falsely set to 0 (hence, the secure boot gets disabled)
after the reboot even if PK (and other signature database) has already
been enrolled in the previous boot.

This is because the secure state is set up *before* loading non-volatile
variables' values from saved data.

This patch fixes the order of variable initialization and secure state
initialization.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Fixes: 5f7dcf079de8 ("efi_loader: UEFI variable persistence")

test/py: efi_secboot: add test for intermediate certificates

In this test case, an image may have a signature with additional
intermediate certificates. A chain of trust will be followed and all
the certificates in the middle of chain must be verified before loading.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>

test/py: efi_secboot: small rework for adding a new test

It won't be very useful to customize HELLO_PATH and EFI_SECBOOT_IMAGE_NAME
under the current code base. So just remove them.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>

efi_loader: signature: rework for intermediate certificates support

In this commit, efi_signature_verify(with_sigdb) will be re-implemented
using pcks7_verify_one() in order to support certificates chain, where
the signer's certificate will be signed by an intermediate CA (certificate
authority) and the latter's certificate will also be signed by another CA
and so on.

What we need to do here is to search for certificates in a signature,
build up a chain of certificates and verify one by one. pkcs7_verify_one()
handles most of these steps except the last one.

pkcs7_verify_one() returns, if succeeded, the last certificate to verify,
which can be either a self-signed one or one that should be signed by one
of certificates in "db". Re-worked efi_signature_verify() will take care
of this step.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>

efi_loader: variable: keep temporary buffer during the authentication

This is a bug fix; Setting an authenticated variable may fail due to
a memory corruption in the authentication.

A temporary buffer will, if needed, be allocated to parse a variable's
authentication data, and some portion of buffer, specifically signer's
certificates, will be referenced by efi_signature_verify().

So the buffer should be kept valid until the authentication process
is finished.

Signed-off-by: AKASHI Takahiro <takahiro.akashi@linaro.org>
Tested-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

cmd/efidebug: missing initialization of load_options

Variable load_options must be initialized to NULL to avoid a segmentation
fault when freeing the memory this variable points to.

Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de>

Merge https://gitlab.denx.de/u-boot/custodians/u-boot-marvell

- Fix dtc warnings for some MVEBU boards

Merge tag 'u-boot-stm32-20200813' of https://gitlab.denx.de/u-boot/custodians/u-boot-stm

  - Use IS_ENABLED to prevent ifdef in board_key_check for STM32MP
  - Add STM32 FMC2 EBI controller driver
  - Fix dwc3-sti-glue which allows STiH410-B2260 to boot again
  - Add fitImage its entry for 587-200 DHCOR SoM
  - Add both PDK2 and DRC02 DT into DHCOM fitImage its
  - Fix DHCOM KS8851 ethernet MAC address
  - Remove stm32mp1 board.c file
  - Use const for struct node_info in board stm32mp1.c file

arm: mvebu: armada-3720-turris-mox.dts: Fix dtc warning

Fix this dtc warning:
Warning (avoid_default_addr_size)

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Marek Behún <marek.behun@nic.cz>

arm: mvebu: armada-388-gp.dts: Fix dtc warnings

Fix these dtc warnings:
Warning (reg_format)
Warning (avoid_default_addr_size)

Signed-off-by: Stefan Roese <sr@denx.de>
Cc: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Chris Packham <judge.packham@gmail.com>

arm: mvebu: armada-xp-maxbcm.dts: Fix dtc warning

Fix this dtc warning:
Warning (avoid_default_addr_size)

Signed-off-by: Stefan Roese <sr@denx.de>

arm: mvebu: armada-xp-theadorable.dts: Fix dtc warning

Fix this dtc warning:
Warning (avoid_default_addr_size)

Signed-off-by: Stefan Roese <sr@denx.de>

board: stm32mp1: use const for struct node_info

Use const for the variable nodes in ft_board_setup,
this patch follow fdt_fixup_mtdparts prototype and no more use stack.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: remove board.c

Remove the file board/st/stm32mp1/board.c which is not more
compiled since commit 156732cc8939 ("board: stm32mp1: move the
function board_debug_uart_init in spl.c")

Fixes: 4fb46816c7e2 ("board: stm32mp1: move the function board_debug_uart_init in spl.c")
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

ARM: dts: stm32: Update eth1addr from EEPROM if eth1 present

The STM32MP1 DHCOM has two ethernet interfaces, the on-SoM DWMAC and KS8851.
Set eth1addr for the KS8851 to a MAC address of the DWMAC incremented by 1.
The MAC of the DWMAC is set from on-SoM EEPROM already, but the MAC address
of KS8851 was left uninitialized, so fix this.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

ARM: stm32: Add fitImage its entry for 587-200 DHCOR SoM

The new 587-200 DHCOR SoM is compatible with the 587-100 prototype,
hence just replicate the entries, as there is no pattern matching.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

ARM: stm32: Add both PDK2 and DRC02 DT into DHCOM fitImage its

Include both PDK2 and DRC02 DTs in the DHCOM fitImage .its and implement
support in SPL to select the correct configuration entry for U-Boot by
using the machine compatible string from SPL DT.

Signed-off-by: Marek Vasut <marex@denx.de>
Cc: Patrice Chotard <patrice.chotard@st.com>
Cc: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

ARM: dts: stm32: add FMC2 EBI support for stm32mp157c

This patch adds FMC2 External Bus Interface support on stm32mp157c.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

configs: stm32mp: add CONFIG_STM32_FMC2_EBI

This patch enables the support of FMC2 EBI.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: update fdt fixup partitions table

This patch adds "st,stm32mp1-fmc2-nfc" compatible string in the
fdt fixup partitions table.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: get resources from parent node

FMC2 EBI support has been added. Common resources (registers base
address and clock) can now be shared between the 2 drivers using
"st,stm32mp1-fmc2-nfc" compatible string. It means that the
common resources should now be found in the parent device when EBI
node is available.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

memory: stm32-fmc2-ebi: add STM32 FMC2 EBI controller driver

The driver adds the support for the STMicroelectronics FMC2 EBI controller
found on STM32MP SOCs.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: use clrsetbits_le32

This patch uses clrsetbits_le32 function instead of multiple instructions.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: use FIELD_PREP/FIELD_GET macros

This patch removes custom macros and uses FIELD_PREP and FIELD_GET macros.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: cosmetic change to use nfc instead of fmc2 where relevant

This patch renames functions and local variables.
This cleanup is done to get all functions starting by stm32_fmc2_nfc
in the FMC2 raw NAND driver when all functions will start by
stm32_fmc2_ebi in the FMC2 EBI driver.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: use FMC2_TIMEOUT_5S for timeouts

FMC2_TIMEOUT_5S will be used each time that we need to wait.
It was seen, during stress tests in an overloaded system,
that we could be close to 1 second, even if we never met this
value. To be safe, FMC2_TIMEOUT_MS is set to 5 seconds.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: remove useless inline comments

Remove inline comments that are useless since function label are
self explanatory.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

mtd: rawnand: stm32_fmc2: fix a buffer overflow

The chip select defined in the device tree could only be 0 or 1.

Signed-off-by: Christophe Kerello <christophe.kerello@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

arm: stm32mp: stm32prog: use IS_ENABLED to prevent ifdef

Use IS_ENABLED to prevent ifdef in stm32prog command.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

arm: stm32mp: bsec: use IS_ENABLED to prevent ifdef

Use IS_ENABLED to prevent ifdef in bsec driver.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in ft_board_setup

Use IS_ENABLED to prevent ifdef in ft_board_setup.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in env functions

Use IS_ENABLED to prevent ifdef in env functions:
- env_get_location
- env_ext4_get_intf
- mmc_get_env_dev

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in board_late_init

Use IS_ENABLED to prevent ifdef in board_late_init.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in dk2_i2c1_fix

Use IS_ENABLED to prevent ifdef in dk2_i2c1_fix.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in set_dfu_alt_inf

Use CONFIG_IS_ENABLED to prevent ifdef in set_dfu_alt_inf.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in sysconf_init

Use IS_ENABLED to prevent ifdef in sysconf_init.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in board_check_usb_power

Use IS_ENABLED to prevent ifdef in board_check_usb_power.

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in g_dnl_board_usb_cable_connected

Use IS_ENABLED to prevent ifdef in g_dnl_board_usb_cable_connected
and in g_dnl_bind_fixup

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

board: stm32mp1: use IS_ENABLED to prevent ifdef in board_key_check

Use IS_ENABLED to prevent ifdef in board_key_check

Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com>
Reviewed-by: Patrice Chotard <patrice.chotard@st.com>

usb: host: dwc3-sti-glue: Fix ofnode_valid() parameter

node varaible is used as iterator into ofnode_for_each_subnode()
loop, when exiting of it, node is no more a valid ofnode.
Use dwc3_node instead as parameter of ofnode_valid()

Fixes: ac28e59a574d ("usb: Migrate to support live DT for some driver")
Signed-off-by: Patrice Chotard <patrice.chotard@st.com>
Cc: Kever Yang <kever.yang@rock-chips.com>

Makefile: fix annoying sunxi hack message

every compilation shows this error

 Hack for sunxi which doesn't have a proper binman definition for
 64-bit boards

not only for sunxi-boards/arm64

fix this by changing to real comments

Fixes: 9f55ee259d0c ("Makefile: sunxi: Don't use binman to build ATF image")
Signed-off-by: Frank Wunderlich <frank-w@public-files.de>