spi: dw: Add SoC-specific compatible strings

This adds SoC-specific compatible strings to all users of the designware
spi device. This will allow for the correct driver to be selected for each
device. Where it is publicly documented, a compatible string for the
specific device version has also been added. Devices without
publicly-documented device versions include MSCC SoCs, and Arc Socs. All
compatible strings except those for SoCFPGAs and some of the versioned
strings have been taken from Linux.

Since SSI_MAX_XFER_SIZE is determined at runtime, this is not strictly
necessary. However, it is a good cleanup and brings things closer to Linux.

Signed-off-by: Sean Anderson <[email protected]>
Tested-by Eugeniy Paltsev <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Rearrange struct dw_spi_priv

This should reduce the size of the struct, and also groups more similar
fields together.

Signed-off-by: Sean Anderson <[email protected]>
Tested-by Eugeniy Paltsev <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Remove spi_enable_chip

This function does nothing but wrap dw_write.

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Rename registers to match datasheet

A few registers had slightly different names from what is in the datasheet.

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Use generic function to read reg address

Using an fdt-specific function causes problems when compiled with a live
tree.

Signed-off-by: Sean Anderson <[email protected]>
Tested-by Eugeniy Paltsev <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Rename "cs-gpio" to "cs-gpios"

This property is named differently than other SPI drivers with the same
property, as well as the property as used in Linux.

Signed-off-by: Sean Anderson <[email protected]>
Tested-by Eugeniy Paltsev <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Convert calls to debug to dev_*

This allows different log levels to be enabled or disabled depending on the
desired level of verbosity. In particular, it allows for general debug
information to be printed while excluding more verbose logging which may
interfere with timing.

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: dw: Fix driving MOSI low while recieving

The resting state of MOSI is high when nothing is driving it. If we drive
it low while recieving, it looks like we are transmitting 0x00 instead of
transmitting nothing. This can confuse slaves (like SD cards) which allow
new commands to be sent over MOSI while they are returning data over MISO.
The return of MOSI from 0 to 1 at the end of recieving a byte can look like
a start bit and a transmission bit to an SD card. This will cause the card
to become out-of-sync with the SPI device, as it thinks the device has
already started transmitting two bytes of a new command. The mmc-spi driver
will not detect the R1 response from the SD card, since it is sent too
early, and offset by two bits. This patch fixes transfer errors when using
SD cards with dw spi.

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

mtd: spinand: enable erasing of bad mtd blocks

U-Boot is able to erase bad mtd blocks on raw nand devices, but this
is not true for spinand flashes. Lets enable this feature for spinand
flashes as well. This is extemelly useful for flash testing.

Signed-off-by: Mikhail Kshevetskiy <[email protected]>

mtd: spinand: Do not erase the block before writing a bad block marker

Currently when marking a block, we use spinand_erase_op() to erase
the block before writing the marker to the OOB area. Doing so without
waiting for the operation to finish can lead to the marking failing
silently and no bad block marker being written to the flash.

In fact we don't need to do an erase at all before writing the BBM.
The ECC is disabled for raw accesses to the OOB data and we don't
need to work around any issues with chips reporting ECC errors as it
is known to be the case for raw NAND.

Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs")
Cc: [email protected]
Signed-off-by: Frieder Schrempf <[email protected]>
Reviewed-by: Boris Brezillon <[email protected]>
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://lore.kernel.org/linux-mtd/[email protected]

mtd: spinand: Explicitly use MTD_OPS_RAW to write the bad block marker to OOB

When writing the bad block marker to the OOB area the access mode
should be set to MTD_OPS_RAW as it is done for reading the marker.
Currently this only works because req.mode is initialized to
MTD_OPS_PLACE_OOB (0) and spinand_write_to_cache_op() checks for
req.mode != MTD_OPS_AUTO_OOB.

Fix this by explicitly setting req.mode to MTD_OPS_RAW.

Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs")
Signed-off-by: Frieder Schrempf <[email protected]>
Reviewed-by: Boris Brezillon <[email protected]>
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://lore.kernel.org/linux-mtd/[email protected]

mtd: spinand: Stop using spinand->oobbuf for buffering bad block markers

For reading and writing the bad block markers, spinand->oobbuf is
currently used as a buffer for the marker bytes. During the
underlying read and write operations to actually get/set the content
of the OOB area, the content of spinand->oobbuf is reused and changed
by accessing it through spinand->oobbuf and/or spinand->databuf.

This is a flaw in the original design of the SPI NAND core and at the
latest from 13c15e07eedf ("mtd: spinand: Handle the case where
PROGRAM LOAD does not reset the cache") on, it results in not having
the bad block marker written at all, as the spinand->oobbuf is
cleared to 0xff after setting the marker bytes to zero.

To fix it, we now just store the two bytes for the marker on the
stack and let the read/write operations copy it from/to the page
buffer later.

Fixes: 7529df465248 ("mtd: nand: Add core infrastructure to support SPI NANDs")
Cc: [email protected]
Signed-off-by: Frieder Schrempf <[email protected]>
Reviewed-by: Boris Brezillon <[email protected]>
Signed-off-by: Miquel Raynal <[email protected]>
Link: https://lore.kernel.org/linux-mtd/[email protected]

spi: Fix typo in header

Spelling.

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

doc: Fix typo in FIT documentation

u_boot should be u-boot

Signed-off-by: Sean Anderson <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

mtd: spi-nor-ids: Add SECT_4K to mx25l12805d

According to the mx25l12805d datasheet it supports using 4K or 64K sectors.
So lets add the SECT_4K to enable 4K sector usage.

Datasheet: https://www.mxic.com.tw/Lists/Datasheet/Attachments/7321/MX25L12805D,%203V,%20128Mb,%20v1.2.pdf

Signed-off-by: Robert Marko <[email protected]>
Cc: Luka Perkov <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

mtd: spi-nor-ids: Add Winbond W25M512JW flash entry

Add Winbond W25M512JW flash device description.

Signed-off-by: Biju Das <[email protected]>
Reviewed-by: Lad Prabhakar <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

mtd: spi-nor-ids: Add Winbond W25M512JV flash entry

Add Winbond W25M512JV flash device description.

Linux already has the flash entry present. A snippet below:
{ "w25m512jv", INFO(0xef7119, 0, 64 * 1024, 1024...},

Signed-off-by: Lad Prabhakar <[email protected]>
Reviewed-by: Biju Das <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

spi: migrate trace to dev and log macro in spi uclass

Define LOG_CATEGORY and change printf and pr_*
to dev_ (when dev is available) or log_ macro.

This patch adds the support of logging feature with log command
(filtering, display of device name in trace) and allows to
suppress traces via the syslog driver.

Signed-off-by: Patrick Delaunay <[email protected]>
Reviewed-by: Jagan Teki <[email protected]>

dh_imx6: Switch to full DM-aware

Enable DM_SPI/DM_SPI_FLASH with a related config option.

Build fine, but not tested.

Cc: Ludwig Zenz <[email protected]>
Cc: Andreas Geisreiter <[email protected]>
Signed-off-by: Jagan Teki <[email protected]>

cm_fx6: Switch to full DM-aware

Enable DM_SPI/DM_SPI_FLASH with a related config option.

Build fine, but not tested.

Cc: Nikita Kiryanov <[email protected]>
Signed-off-by: Jagan Teki <[email protected]>

cl-som-imx7: Switch to DM_SPI/DM_SPI_FLASH

Enable DM_SPI/DM_SPI_FLASH with associated config
options.

Build fine, but not tested.

Cc: Uri Mashiach <[email protected]>
Signed-off-by: Jagan Teki <[email protected]>

configs: Resync with savedefconfig

Rsync all defconfig files using moveconfig.py

Signed-off-by: Tom Rini <[email protected]>

Merge https://gitlab.denx.de/u-boot/custodians/u-boot-x86

- Update cmdline parameter to be an env var in zimage boot
- Various minor fixes for Intel Edison board

x86: edison: Switch to DM_USB_GADGET

DM is the modern default approach for the drivers in U-Boot.
It also allows to configure code via Device Tree.

Move Intel Edison to use DM_USB_GADGET and drop hard coded values.

Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: edison: Use dwc3-generic driver for Intel Edison

Use generic Synopsys DesignWare 3 driver on Intel Edison.
For now it's just a stub which allows future refactoring.

Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: edison: BINMAN selection is specific to the board

The platforms based on Intel Tangier may have different requirements
how to create bootloader bundle to supply to a device. Currently
the BINMAN approach is for Intel Edison only.

Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: edison: Add CPU to compatible string

Like in the rest of x86 boards append CPU to the board compatible string.

Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: edison: Drop unneeded DM_PCI_COMPAT

None of the driver for Edison is using DM_PCI_COMPAT, hence drop it.

Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: tangier: Find proper memory region for relocation

It appears that U-Boot works by luck on Intel Edison board because the amount
of RAM is less than 1 GB and standard way of calculating the top of it work
for this configuration. However, this won't work if the amount of RAM is
different and split differently in address space. We have to find the suitable
window correctly.

Find proper memory region for relocation by scanning MMAP SFI table in
board_get_usable_ram_top() callback.

According to the address map documentation the Main Memory is guaranteed to lie
in the 0..2 GB range, that's why we limit search by this range.

Fixes: e71de54a4943 ("x86: Add Intel Tangier support")
Signed-off-by: Andy Shevchenko <[email protected]>
Reviewed-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
[bmeng: fixed a typo in the commit message]
Signed-off-by: Bin Meng <[email protected]>

x86: coral: Update the boot script

Make use of the new bootargs substitution mechanism and zboot command
syntax.

Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

x86: zimage: Update cmdline parameter to be an env var

With the updated changes to bootargs substitution[1], the zboot command
needs to be updated to get its command line from an environment variable
instead of a memory address. This is because the command-line string must
be updated to convert %U to ${uuid}, etc.

In any case it is more flexible to use a environment variable and it is
best to do this before the release to avoid a subsequent change.

Update the command accordingly.

[1] http://patchwork.ozlabs.org/project/uboot/list/?series=212481

Signed-off-by: Simon Glass <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

Merge tag 'doc-2021-01-rc4' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi

Documentation fixes for v2020.01-rc4

Move several README files to reStructured text for the HTML documentation.
Describe register for global data on x86.
Allow building HTML documentation with Sphinx 3

distro_bootcmd: set devtype for dhcp boot

Set $devtype for DHCP boot, which can be handy for the boot.scr
for detection of devtype used (for example, when the same boot.scr is
used for both mmc/dhcp boot):

if test ${devtype} = "dhcp"; then
...
fi

Signed-off-by: Igor Opaniuk <[email protected]>
Reviewed-by: Oleksandr Suvorov <[email protected]>

Merge tag 'u-boot-amlogic-20201215' of https://gitlab.denx.de/u-boot/custodians/u-boot-amlogic

- fix pinctrl BIAS setup which caused failure on some eMMC models
- fix dr-mode configuration in dwc3-meson-g12a glue driver

doc: uefi: remove leading $ from bash commands

Use the same formatting for all bash commands.

Signed-off-by: Heinrich Schuchardt <[email protected]>

doc: move README.bootmenu to HTML doc

Convert README.bootmenu to reStructured text and move it to
usage/bootmenu.rst.

Adjust the text concerning configuration settings as these now are managed
via Kconfig.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Pali Rohár <[email protected]>

doc: move README.trace to HTML documentation

Convert README.trace to reStructured text and move it to develop/trace.rst.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Simon Glass <[email protected]>

doc: move README.NetConsole to HTML documentation

Convert README.NetConsole to reStructured text and move it to
doc/usage/netconsole.rst.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Simon Glass <[email protected]>

doc: use code-block in pstore.rst

Use syntax highlighting for a Linux console session.

Signed-off-by: Heinrich Schuchardt <[email protected]>

doc: move pstore.rst to usage/pstore.rst

Let's have a separate chapter dedicated to using U-Boot.

Signed-off-by: Heinrich Schuchardt <[email protected]>

doc: move README.commands to HTML doc

Reformat README.commands as reStructured text and add it to the HTML
documentation as develop/commands.rst.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Simon Glass <[email protected]>

doc: allow building htmldoc with Sphinx 3+

Due to removed function c_funcptr_sig_re building with Sphinx 3 fails.

With the patch building succeeds with a lot of warnings if the '-W' flag is
removed from doc/Makefile. Most of the documentation is correct

This follows the approach taken by the Linux kernel.

Signed-off-by: Heinrich Schuchardt <[email protected]>

doc: global data pointer on x86, x86_64

On x86 the global data pointer is stored in register fs.
On x86_64 no register is used for the global data pointer.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Simon Glass <[email protected]>

doc/build/gcc.rst: required packages for SUSE

Describe the packages required to build U-Boot on SUSE.

Signed-off-by: Heinrich Schuchardt <[email protected]>

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

- Fix wrong amoswap t1 usage in startup.
- Reset the board after crash.
- Enable distro booting from an attached SCSI disk for QEMU.
- Support the optional header fields in efi header.

pinctrl: meson: fix bit manipulation of pin bias configuration

This fixes the wrong usage of clrsetbits_le32(), badly setting the set argument.

Fixes: c4c726c26b ("pinctrl: meson: add pinconf support")
Reported-by: Anton Arapov <[email protected]>
Reported-by: Otto Meier <[email protected]>
Signed-off-by: Neil Armstrong <[email protected]>

usb: dwc3-meson-g12a: always configure dr-mode

dwc3_meson_g12a_force_mode() sets the dr-mode of the USB PHY. However
it skips setting the mode if it matches the one done during driver probe
(stored in private structure). This fails if the mode has been changed
to other value and then back to initial one. Fix this by configuring the
dr-mode always, regadless of the one set at the driver probe).

This fixes operation of USB gadget based drivers when they are initialized
for the second time.

Signed-off-by: Marek Szyprowski <[email protected]>
Acked-by: Neil Armstrong <[email protected]>
Signed-off-by: Neil Armstrong <[email protected]>

riscv: Complete efi header for RV32/64

This patch depends on Atish's patch.
(https://patchwork.ozlabs.org/project/uboot/patch/20201013192331.3236458[email protected]/)

Add fields to complete Optional Header "Data Directories" specified in the document.
(https://docs.microsoft.com/en-us/windows/win32/debug/pe-format)

Signed-off-by: Leo Yu-Chi Liang <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Reviewed-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Atish Patra <[email protected]>

riscv: Fix efi header size for RV32

This patch depends on Atish's patch.
(https://patchwork.ozlabs.org/project/uboot/patch/20201013192331.3236458[email protected]/)

Modify the size of the Optional Header "Windows-Specific Fields" to fit with the specification.
(https://docs.microsoft.com/en-us/windows/win32/debug/pe-format)

Signed-off-by: Leo Yu-Chi Liang <[email protected]>
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]
Cc: [email protected]

riscv: Fix efi header for RV32

RV32 should use PE32 format instead of PE32+ as the efi header format.
This requires following changes
1. A different header magic value
2. An additional parameter known as BaseOfData. Currently, it is set to
   zero in absence of any usage.

Signed-off-by: Atish Patra <[email protected]>
Reviewed-by: Bin Meng <[email protected]>
Reviewed-by: Rick Chen <[email protected]>

riscv: qemu: enable distro boot from scsi

Booting via distro boot fails for:

    qemu-system-riscv64
    -drive if=none,file=sct-riscv64.img,format=raw,id=mydisk \
    -device ich9-ahci,id=ahci -device ide-hd,drive=mydisk,bus=ahci.0

Enable distro booting from an attached SCSI disk.

Signed-off-by: Heinrich Schuchardt <[email protected]>
Reviewed-by: Bin Meng <[email protected]>

riscv: reset after crash

If an exception occurs on ARM or x86, we call panic() which will try to
reset the board. Do the same on RISC-V.

To avoid -Werror=format-zero-length move a '\n' to the string passed to
panic. We don't need a message here as depending on CONFIG_PANIC_HANG we
will either see

    ### ERROR ### Please RESET the board ###

or

    resetting ...

as next message.

Reviewed-by: Rick Chen <[email protected]>
Signed-off-by: Heinrich Schuchardt <[email protected]>

riscv: fix the wrong swap value register

Not s2 register, t1 register is correct
Fortunately, it works because t1 register has a garbage value

Signed-off-by: Brad Kim <[email protected]>
Reviewed-by: Lukas Auer <[email protected]>
Reviewed-by: Leo Liang <[email protected]>

Merge tag 'u-boot-atmel-fixes-2021.01-b' of https://gitlab.denx.de/u-boot/custodians/u-boot-atmel

Second set of u-boot-atmel fixes for 2021.01 cycle

This set includes very important fixes for: MMC booting on several
boards, drive strength on sam9x60ek mmc lines, compile issues for
timer.c old driver, removal of unwanted access to sam9x60 bit for
oscillator bypass mode, and eeproms read on sama5d2_icp.

ARM: dts: at91: sama5d2_icp: fix i2c eeprom compatible

The correct compatible for this eeproms is microchip,24aa02e48
The previous compatible string was working up to U-boot 2020.04.

Signed-off-by: Eugen Hristev <[email protected]>
Tested-by: Codrin Ciubotariu <[email protected]>

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

- Add lx2162 soc, lx2162qds support.
- Bug-fixes related ls102x-usb, ifc, bootcmd, secure-boot header,
- rgmii, vid, fdt, env variable, pci for Layerscape products

configs: lx2162aqds: Enable gpio driver in defconfig

make lx2162aqds platform to enable gpio driver.

Signed-off-by: Hui Song <[email protected]>
Signed-off-by: Ran Wang <[email protected]>
[Rebased]
Signed-off-by: Priyanka Jain <[email protected]>

arm: ls102xa: select USB PHY erratum's only if USB is enabled

The USB support is not by default enabled on all designs, so it does not
make seance to have USB specific erratum's enabled on such a designs.

On our internal Hitachi-Powergrids design not using the USB controller
there is a crash when accessing those specific memory locations selected
by the erratum flags.

Signed-off-by: Aleksandar Gerasimovski <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

drivers: ifc: add define for IFC_CSPRn TE bit

To drive TE pin high is supported IFC configuration that can be used on
some designs.

Signed-off-by: Aleksandar Gerasimovski <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

powerpc: mpc85xx: Allow boards to override CONFIG_USB_MAX_CONTROLLER_COUNT

If the board isn't strapped to enable USB1 then attempting to access it
will result in a hang. Avoid this by allowing boards to define
CONFIG_USB_MAX_CONTROLLER_COUNT.

Signed-off-by: Chris Packham <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

armv8: fsl-layerscape: Fix automatic setting of bootmcd with TF-A

When booting from TF-A there is a logic that attempt to detect if the
default environment is used, if this is the case it then set the
`bootcmd` and `mcinitcmd` depending of the device we booted from.
This detection logic is dubious as it access internals of the env
implementation and it doesn't always work correctly.

First of all it detect any valid environment as not being the
default, so after running `env default -a && saveenv` the board
doesn't boot anymore as `bootcmd` is then empty.
But it also fails in some other ways, for example it always detect a
default environment when redundant env is enabled on MMC, so in that
case `bootcmd` is overwritten on every boot.

Instead of increasing the complexity of the detection just check if
`bootcmd` and `mcinitcmd` are set in the environment and set them if
they are not.

Signed-off-by: Alban Bedel <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

net: pfe_eth: read PFE ESBC header flash with spi_flash_read API

Read PFE ESBC header flash with spi_flash_read API
- logs as follows,
  Net:   SF: Detected s25fs512s with page size 256 Bytes, erase size 256
  KiB, total 64 MiB
  "Synchronous Abort" handler, esr 0x96000210
  elr: 000000008206db44 lr : 0000000082004ea0 (reloc)
  elr: 00000000b7ba6b44 lr : 00000000b7b3dea0
  x0 : 00000000b79407e8 x1 : 0000000040640000
  x2 : 0000000000000050 x3 : 0000000000000000
  x4 : 000000000000000a x5 : 0000000000000050
  x6 : 0000000000000366 x7 : 00000000b7942308
  x8 : 00000000b76407c0 x9 : 0000000000000008
  x10: 0000000000000044 x11: 00000000b7634d1c
  x12: 000000000000004f x13: 0000000000000044
  x14: 00000000b7634d98 x15: 00000000b76407c0
  x16: 0000000000000000 x17: 0000000000000000
  x18: 00000000b7636dd8 x19: 0000000000000000
  x20: 00000000b79407d0 x21: 00000000b79407e8
  x22: 0000000040640000 x23: 00000000b7634e58
  x24: 0000000000000000 x25: 0000000003800000
  x26: 00000000b7bdd000 x27: 0000000000000000
  x28: 0000000000000000 x29: 00000000b7634d10

  Code: d2800003 eb03005f 54000101 d65f03c0 (f8636826)
  Resetting CPU ...

Signed-off-by: Biwen Li <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

lx2160a: Fix address for secure boot headers

Update kernel_size_sd variable with correct value for lx2160a.

Signed-off-by: Manish Tomar <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

ls1043a: Fix address for secure boot headers

Update kernelheader_addr and kernelheader_addr variables with
correct values for ls1043a.

Signed-off-by: Manish Tomar <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

layerscape: fdt.c: Check for NULL return value from fdt_getprop()

Check for NULL return value from fdt_getprop() in
fdt_fixup_remove_jr()

Signed-off-by: Priyanka Singh <[email protected]>
[Fixed checkpatch errors/warnings]
Signed-off-by: Priyanka Jain <[email protected]>

board: freescale: powerpc: add support for all RGMII modes

Make sure all RGMII internal delay modes are covered.

Signed-off-by: Madalin Bucur <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

armv8: ls1043/ls1046aqds: add support for all RGMII modes

Make sure all RGMII internal delay modes are covered.

Signed-off-by: Madalin Bucur <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

board: freescale: vid.c: Initialize variable 'i2caddress'

Initialize variable 'i2caddress' in adjust_vdd() to zero

Signed-off-by: Priyanka Singh <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

include/configs: ls1012aqds: add default environment variable

This adds default environment variable for ls1012aqds

Signed-off-by: Biwen Li <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

board/freescale/common: fix a bug that failed to read/write eeprom on ls1021atsn

Fix a bug that failed to read/write eeprom on ls1021atsn

Signed-off-by: Biwen Li <[email protected]>
Tested-by: Vladimir Oltean <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

pci: layerscape: fix a dead loop issue

Fixes: commit 8ec619f8fd84 ("pci: layerscape: Fixup PCIe EP
mode DT nodes for LX2160A rev2")

This added the PCIe EP nodes fixup of LX2160A, but it
didn't update the condition value when there isn't a
property 'apio-wins'.

Signed-off-by: Hou Zhiqiang <[email protected]>
[Fixed checkpatch error]
Signed-off-by: Priyanka Jain <[email protected]>

configs: lx2162aqds: enable eMMC HS400 mode support

Enable eMMC HS400 mode support on LX2162AQDS.

Signed-off-by: Yangbo Lu <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

configs: lx2162aqds: enable CONFIG_BOARD_EARLY_INIT_R

Enable CONFIG_BOARD_EARLY_INIT_R for SDHC adapter card
identification and configuration.

Signed-off-by: Yangbo Lu <[email protected]>
[Rebased]
Signed-off-by: Priyanka Jain <[email protected]>

configs: lx2162a: Enable OPTEE support

Enable support to compile OPTEE driver, access AVB TA
and RPMB API's access via RPC from OPTEE for lx2162

Signed-off-by: Ruchika Gupta <[email protected]>
Signed-off-by: Gaurav Jain <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

armv8: lx2162aqds: Add support for LX2162AQDS platform

This patch add base support for LX2162AQDS board.
LX2162AQDS board supports LX2162A family SoCs.
This patch add basic support of platform.

Signed-off-by: Ioana Ciornei <[email protected]>
Signed-off-by: Zhao Qiang <[email protected]>
Signed-off-by: hui.song <[email protected]>
Signed-off-by: Manish Tomar <[email protected]>
Signed-off-by: Vikas Singh <[email protected]>
Signed-off-by: Meenakshi Aggarwal <[email protected]>
[Rebased]
Signed-off-by: Priyanka Jain <[email protected]>

armv8: lx2162a: Add Soc changes to support LX2162A

LX2162 is LX2160 based SoC, it has same die as of LX2160
with different packaging.

LX2162A support 64-bit 2.9GT/s DDR4 memory, i2c, micro-click module,
microSD card, eMMC support, serial console, qspi nor flash, qsgmii,
sgmii, 25g, 40g, 50g network interface, one usb 3.0 and serdes
interface to support three PCIe gen3 interface.

Signed-off-by: Meenakshi Aggarwal <[email protected]>
[Fixed whitespace errors]
Signed-off-by: Priyanka Jain <[email protected]>

drivers/net/phy: Add CORTINA_NO_FW_UPLOAD to Kconfig

Move CORTINA_NO_FW_UPLOAD to Kconfig file so that it can
be controlled via defconfig files.

Signed-off-by: Meenakshi Aggarwal <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

pci: ls_pcie_g4: Add size check for config resource

resource "config" is required to have minimum 4KB space
to access all config space of PCI Express EP.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

pci: layerscape: Add size check for config resource

resource "config" is required to have minimum 8KB space
as per hardware documentation.

Signed-off-by: Hou Zhiqiang <[email protected]>
Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls1028a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls1043a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls1012a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls1088a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls2080a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: ls1046a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

arm: dts: lx2160a: add label to pcie nodes in dts

Add label to pcie nodes in dts so that these nodes
are easy to refer.

Signed-off-by: Wasim Khan <[email protected]>
[Rebased]
Signed-off-by: Priyanka Jain <[email protected]>

pci: ls_pcie_g4: Print pcie controller number starting from 1

Print pcie controller number starting from 1

Signed-off-by: Wasim Khan <[email protected]>

pci: layerscape: Update print of pcie controller

Print pcie controller number starting from 1

Signed-off-by: Wasim Khan <[email protected]>
[Trimmed subject]
Signed-off-by: Priyanka Jain <[email protected]>

configs: lx2160a: Enable CONFIG_PCIE_LAYERSCAPE_GEN4

LX2160A-Rev1 uses PCIe layerscape Gen4 controller.
Enable CONFIG_PCIE_LAYERSCAPE_GEN4 for lx2160a.

Signed-off-by: Wasim Khan <[email protected]>
Reviewed-by: Priyanka Jain <[email protected]>

Merge tag 'u-boot-stm32-20201209' of https://gitlab.denx.de/u-boot/custodians/u-boot-stm

- Manage CONFIG_ENV_EXT4_DEVICE_AND_PART in stm32mp1 board
- Update ARM STI and ARM STM STM32MP Arch maintainers emails
- Enable internal pull-ups for SDMMC1 on DHCOM SoM

ARM: dts: stm32: Add USB OTG ID pin on DH AV96

Add USB OTG ID pin mux and switch the USB OTG port
from peripheral to OTG mode.

Signed-off-by: Marek Vasut <[email protected]>
Cc: Patrice Chotard <[email protected]>
Cc: Patrick Delaunay <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

ARM: dts: stm32: Enable SDMMC3 on DH DRC02

The DH DRC02 board has an on-board microSD slot,
add DT properties to enable the slot.

Signed-off-by: Marek Vasut <[email protected]>
Cc: Patrice Chotard <[email protected]>
Cc: Patrick Delaunay <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

ARM: dts: stm32: Disable SDMMC1 CKIN feedback clock

The STM32MP1 DHCOM SoM can be built with either bus voltage level shifter
or without one on the SDMMC1 interface. Because the SDMMC1 interface is
limited to 50 MHz and hence SD high-speed anyway, disable the SD feedback
clock to permit operation of the same U-Boot image on both SoM with and
without voltage level shifter.

Signed-off-by: Marek Vasut <[email protected]>
Cc: Patrice Chotard <[email protected]>
Cc: Patrick Delaunay <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

ARM: dts: stm32: Enable internal pull-ups for SDMMC1 on DHCOM SoM

The default state of SD bus and clock line is logical HI. SD card IO is
open-drain and pulls the bus lines LO. Always enable the SD bus pull ups
to guarantee this behavior on DHCOM SoM. Note that on SoMs with SD bus
voltage level shifter, the pull ups are built into the level shifter,
however that has no negative impact.

Signed-off-by: Marek Vasut <[email protected]>
Cc: Patrice Chotard <[email protected]>
Cc: Patrick Delaunay <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

.mailmap: map Patrick Delaunay and my email address

Add our new email address dedicated for upstream activities.

Signed-off-by: Patrice Chotard <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

treewide: Update email address Patrick Delaunay and Patrice Chotard

Update Patrick and my email address with the one dedicated to
upstream activities.

Signed-off-by: Patrice Chotard <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

MAINTAINERS: Update ARM STI and ARM STM STM32MP Arch maintainers emails

Update Patrick and my email address with the one dedicated to
upstream activities.

Signed-off-by: Patrice Chotard <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

add check for ignored CONFIG_ENV_EXT4_DEVICE_AND_PART definition

Check whether user has explicitly defined device and partition where
environment file will be located before using 'auto' i.e. bootable
partition

Voids the need to set such partition as bootable to work with the
'dev:auto' tuple

Signed-off-by: Manuel Reis <[email protected]>
Cc: Patrick Delaunay <[email protected]>
Cc: Patrice Chotard <[email protected]>
Tested-by: Michael Opdenacker <[email protected]>
Reviewed-by: Patrick Delaunay <[email protected]>

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

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

- Espressobin: Simplify DT handling of board variants (Pali)
- Add Luka Perkov to maintainers of Puzzle-M801 (Luka)
- Armada 38x: Enable board specific USB2 high-speed impedance
  threshold configuration (Joshua)