This is already invoked a few cycles later in monitor mode by
_secure_monitor (_sunxi_cpu_entry calls _do_nonsec_entry which triggers
_secure_monitor via smc #0). Drop it here, it serves no purpose.

CC: Marc Zyngier <marc.zyngier@arm.com>
Signed-off-by: NJan Kiszka <jan.kiszka@siemens.com>
Reviewed-by: NTom Rini <trini@konsulko.com>
Reviewed-by: NThierry Reding <treding@nvidia.com>
Tested-by: NThierry Reding <treding@nvidia.com>
Tested-by: NIan Campbell <ijc@hellion.org.uk>
Tested-by: NHans de Goede <hdegoede@redhat.com>
Signed-off-by: NHans de Goede <hdegoede@redhat.com>

Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Move arch/arm/include/asm/arch-zynq/*
  -> arch/arm/mach-zynq/include/mach/*
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Move arch/arm/cpu/armv7/zynq/* -> arch/arm/mach-zynq/*
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

The comment line in arch/arm/cpu/armv7/zynq/config.mk says that
the option "-mfpu=neon" is necessary for compiling lowlevel_init.S.
We do not have to give it to all the source files.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Enable SDHCI0 for zynqmp.
Add empty gpio.h because of sdhci requirement.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Provide an option to write filesystem independend commands.
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Added the SPI driver support for ZynqMP
The controller is same as zynq SPI controller
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Reviewed-by: NJagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>

Enable the i2c driver for ZynqMP
Also enable the eeprom for read and writes
to eeprom on ZynqMP
ZynqMP uses the same i2c controller as in Zynq
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Add support for EMMC bootmode.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Add support for Veloce - zynqmp emulation platform.
Platform doesn't support SDHCI.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Fix wrong timer calculation in get_timer_masked incase of
overflow.
This fixes the issue of getting wrong time from get_timer()
calls.
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Remove the quirk SDHCI_QUIRK_NO_CD as it is not
required.
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Add support for loading sw for R5 with enabling for zynqmp.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>

Define the mmu table till 2MB granularity
enable dcaches for zynqmp.
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Disable all level shifters before enabling
the PS-to-PL level shifters as it would
be good to disable all level shifters before
enabling the PS-to-PL in order to ensure that
it is in proper state
Signed-off-by: NSiva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

We are about to change the location for ps7_init files, breaking the
current work-flows.  It is good time to drop the legacy ps7_init.c/h
support.

Going forward, please use ps7_init_gpl.c/h all the time.
If you are still using old Xilinx tools that are only able to
generate ps7_init.c/h, rename them into ps7_init_gpl.c/h.
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Suggested-by: NSören Brinkmann <soren.brinkmann@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

The PicoZed is a System-on-Module board which is marketed as part of
the ZedBoard/MicroZed/etc. collection. It includes a Zynq-7000
processor.

This patch adds support that covers all the variants of the PicoZed
including the SKUs with Z7010/Z7020 and Z7015/Z7030 Zynq chips. This
patch set however only covers support for the System-on-Module and does
not cover any extra components that are available on carrier boards
(except those that are fanned out of the module itself).

More information on this board, its variants and available carrier
boards is available at: http://zedboard.org/product/picozedSigned-off-by: NNathan Rossi <nathan.rossi@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

On slow platforms not all baudrate setting is valid.
Check it directly in the driver and setup maximum possible
frequency.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Enable GPIO driver and GPIO commands.
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

Most of the code is taken (and adapted) from Linux kernel driver.

Just add CONFIG_ZYNQ_GPIO to you config to enable it
Signed-off-by: NAndrea Scian <andrea.scian@dave.eu>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

It is possible for CONFIG_XILINX_EMACLITE to be defined without
XILINX_EMACLITE_BASEADDR being defined as the EMAC Lite driver support
OF init. Check that the driver is enabled and the base address is
available before initializing with a static base address.
Signed-off-by: NNathan Rossi <nathan.rossi@xilinx.com>
Signed-off-by: NMichal Simek <michal.simek@xilinx.com>

These functions now rely on uclass_find_first/next_device() and assume that
they will either return failure (-ve error code) or a device. In fact,
coming to the end of a list is not considered failure and they return 0
in that case.

The logic to deal with this was replaced in commit acb9ca2a with just using
uclass_get_device_tail(). Add back the missing logic. This bug was
caught by unit tests but since they were broken for other reasons at the
time, this was not noticed.
Signed-off-by: NSimon Glass <sjg@chromium.org>

This was accidentally added by commit dd0b0122
"serial: ns16550: Add an option to specify the debug UART register shift".
Remove it.
Signed-off-by: NAxel Lin <axel.lin@ingics.com>
Acked-by: NSimon Glass <sjg@chromium.org>

Trivial bug fix for commit 5a87c417 (dm: core: Drop device
removal error path when not supported).
Signed-off-by: NMasahiro Yamada <yamada.masahiro@socionext.com>
Acked-by: NSimon Glass <sjg@chromium.org>

Upto now flash sector_size is assigned from params which isn't
necessarily a sector size from vendor, so based on the SECT_*
flags from flash_params the erase_size will compute and it will
become the sector_size finally.

Bug report (from Bin Meng):
=> sf probe
SF: Detected SST25VF016B with page size 256 Bytes, erase size 4 KiB,
total 2 MiB, mapped at ffe00000

=> sf erase 0 +100
SF: 65536 bytes @ 0x0 Erased: OK
Signed-off-by: NJagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>
Reported-by: NBin Meng <bmeng.cn@gmail.com>
Tested-by: NBin Meng <bmeng.cn@gmail.com>

With SPI flash moving to driver model, commit fbb09918 "dm: Convert
spi_flash_probe() and 'sf probe' to use driver model" ignored the
SST flash-specific write op (byte program & word program), which
actually broke the SST flash from wroking.

This commit makes SST flash work again under driver model, by adding
SST flash-specific handling in the spi_flash_std_write().
Signed-off-by: NBin Meng <bmeng.cn@gmail.com>
Reviewed-by: NJagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>

Add a new member 'flags' in struct spi_flash to store the flash flags
during spi_flash_validate_params().
Signed-off-by: NBin Meng <bmeng.cn@gmail.com>
Reviewed-by: NJagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>

The timeout value is never reset during the transfer. This means that when
transferring more data we eventually trigger the timeout.

This was reported on the mailing list:
"Spansion SPI flash read timeout with AM335x"
Signed-off-by: NDavid Dueck <davidcdueck@googlemail.com>
CC: Tom Rini <trini@konsulko.com>
CC: Stefan Roese <sr@denx.de>
CC: Andy Pont <andy.pont@sdcsystems.com>
Tested-by: NDavid Dueck <davidcdueck@googlemail.com>
Reviewed-by: NJagannadha Sutradharudu Teki <jagannadh.teki@gmail.com>

Instead of hardcoding the 'fdtfile' variable, let's detect the SoC and
board variant on the fly and change the dtb name.

Based on the scheme done on am335x board.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Tested-By: NVagrant Cascadian <vagrant@debian.org>
Reviewed-by: NTom Rini <trini@konsulko.com>

README file suggests to use 'fdtfile' for the dtb file name and
'fdt_addr_r' for the dtb address in RAM, so do as suggested.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: NTom Rini <trini@konsulko.com>

Introduce is_hummingboard() function that reads GPIOs that can distinguish
between Cubox-i and Hummingboard.

Print the board name accordingly.

Based on a patch from Rabeeh Khoury.
Signed-off-by: NRabeeh Khoury <rabeeh@solid-run.com>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: NTom Rini <trini@konsulko.com>

Cubox-i and Hummingboard support several MX6 SoCs: mx6solo, mx6dual-lite,
mx6dual and mx6quad. Add support for the different SoC/memory sizes
combinations.

DDR initialization values were extracted from Solid-run internal U-boot.

Tested on a CuBox-i4Pro, HummingBoard-i2eX and HummingBoard-i1.
Signed-off-by: NRabeeh Khoury <rabeeh@solid-run.com>
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: NTom Rini <trini@konsulko.com>

Cubox-i and Hummingboard support several MX6 SoCs: mx6solo, mx6dual-lite,
mx6dual and mx6quad.

Use IOMUX_PADS() macro in order to prepare for the multi-SoC support.
Also pass 'MX6QDL' in the defconfig to indicate it.
Signed-off-by: NFabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: NTom Rini <trini@konsulko.com>