If regular NAND booting fails to find a valid uImage in the
kernel partition in NAND, try to boot using a zImage and dtb found
in a UBI volume in the rootfs partition.  This is the NAND analog
of mmc zImage booting for device-tree based kernels.

Signed-off-by: Ash Charles <ashcharles@gmail.com>
Signed-off-by: Arun Bharadwaj <arun@gumstix.com>

Overo COMs have NAND flash that requires 4-bit ECC or better except for
the first sector which can use 1-bit ECC.  The boot ROM expects to load
a payload from NAND written using 1-bit hardware-based ECC.  In short,
write SPL to NAND something like this (4 times for redundancy):
 #> nandecc hw
 #> nand write ${loadaddr} 0x0 ${filesize}
 #> nand write ${loadaddr} 0x20000 ${filesize}
 #> nand write ${loadaddr} 0x40000 ${filesize}
 #> nand write ${loadaddr} 0x60000 ${filesize}

Then, switch back to software-based BCH8 for everything else:
 #> nandecc sw bch8

After [1], enlarge the max size of the SPL so the BCH code can fit.

[1] https://www.mail-archive.com/u-boot@lists.denx.de/msg163912.html



Signed-off-by: Ash Charles <ashcharles@gmail.com>

Signed-off-by: Ash Charles <ashcharles@gmail.com>

Signed-off-by: Arun Bharadwaj <arun@gumstix.com>
Signed-off-by: Ash Charles <ashcharles@gmail.com>

The Gumstix Arbor43C expansion board [1] uses a customized device tree
file.  When this expansion board is detected, load this file.

[1] https://store.gumstix.com/index.php/products/635/



Signed-off-by: Ash Charles <ashcharles@gmail.com>

Signed-off-by: Peter Howard <phoward@gme.net.au>

Fix the usage info from:
part part uuid <interface> <dev>:<part>
to
part uuid <interface> <dev>:<part>

Signed-off-by: Maxin B. John <maxin.john@enea.com>
Reviewed-By: Sjoerd Simons <sjoerd.simons@collabora.co.uk>

Add the stm32F4 board's serial ports support.
User can use it easily.
The user only need to edit the number of the usart.
The patch also fix the serial print out.

Last, this version of patch fix the first patch checkpatch.pl error.
Thanks to Kamil Lulko.

Signed-off-by: kunhuahuang <huangkunhua@gmail.com>

This patch fix the serial output.
The source is from Kamil Lulko's "stm32f429-discovery board support"

Thanks, Varcain. I learned a lot.

Signed-off-by: kunhuahuang <huangkunhua@gmail.com>

This adds support for providing a partition number instead of a sector for eMMC.

Signed-off-by: Paul Kocialkowski <contact@paulk.fr>

This refactors spl_mmc_load_image to use a switch/case structure and easier
to understand spl_start_uboot checks. This also introduces some more automatic
fallback on the next mmc boot mode as long as it keeps failing.

Lines that go beyond 80 chars are also reduced by reducing the number of tabs.
Debug and error strings are refctored to match a common style.

Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
[trini: Fix a thinko in the SPL_EXT_SUPPORT + SPL_OS conversion, make part be
 __maybe_unused]
Signed-off-by: Tom Rini <trini@konsulko.com>

Use clrsetbits_le32() to replace clrbits_le32() + setbits_le32().

Signed-off-by: Axel Lin <axel.lin@ingics.com>

usbupdate in real does allways load some script from usb-storage and execute
it, on all B&R targets.

So we do following 2 things:
- rename it to what it really does
- move it from boards to common environment

Signed-off-by: Hannes Petermaier <oe5hpm@oevsv.at>

Since we display in future the splash screen out of linux-os, we don't need
this support anymore within the common section.

But kwb-target is still using BMP_DISPLAY feature, so we move the related
from the common section into the target-specific.

Also the default environment of tseries will be adapted to this.

Signed-off-by: Hannes Petermaier <oe5hpm@oevsv.at>

Since the used AM3352 SoC doesn't have GPU it isn't allways necessary to build
in complete drm-stuff into linux kernel. In very small applications only we use
the simple-framebuffer.

So we have 2 use-cases:
- device operating on drm-driver (let simplefb node disabled)
- device operating on simplefb-driver (activate simplefb node and reserve mem)

The decision is made by means of "simplefb" environment variable.

simplefb = 0
we don't enable the (maybe) existing simplefb node and all the rest around
display is up to the linux-kernel. We just disable the backlight, beceause we
do not want see the flicker during take over of drm-driver.

simplefb = 1
we enable the (maybe) existing simplefb node and reserve framebuffers size
in memory.

Signed-off-by: Hannes Petermaier <oe5hpm@oevsv.at>

the CONFIG_LCD_NOSTDOUT feature never had become mainline in uboot due to the
fact that the problem of "not writing out whole console to lcd" can be solved
with another way.

So we remove this unnary define.

Signed-off-by: Hannes Petermaier <oe5hpm@oevsv.at>

We take use of the new LCD_ROTATION feature.
The information about how the display is rotated is taken from B&R specific
(/factory-settings/rotation) information in the devicetree.

The information there is stored as string (cw, ud, ccw, none) since starting
support of this devices and cannot be changed, so we have to convert it into

none = 0
cw   = 1
ud   = 2
ccw  = 3

Signed-off-by: Hannes Petermaier <oe5hpm@oevsv.at>

The two tools that create android boot images, mkbootimg and the fastboot
client, set the kernel address by default to 0x10008000.

U-boot always honors this field, and will try to relocate the kernel to
whatever value is set in the header, which won't be mapped to the actual RAM on
most platforms, resulting in the kernel obviously not booting.

All the targets in U-Boot right now will download the android boot image to
CONFIG_SYS_LOAD_ADDR, which means that it will already have been downloaded to
some location that is suitable for execution.

In order to have the kernel booting even with the default boot image kernel
address, if that address is used, just execute the kernel where it is.

Signed-off-by: Maxime Ripard <maxime.ripard@free-electrons.com>

I checked with Detlev, and sadly removing that maintainers entry is
right thing to do.

Signed-off-by: Pavel Machek <pavel@denx.de>

This patch fixes cross-compiling U-Boot tools with the musl C library:
  * including <sys/types.h> is needed for ulong
  * defining _GNU_SOURCE is needed for loff_t

Tested for target at91sam9261ek_dataflash_cs3.

Signed-off-by: Jörg Krause <joerg.krause@embedded.rocks>
Cc: Tom Rini <trini@konsulko.com>

This patch adds a driver for the PCA9551 LED controller.

Originated-by: Timo Herbrecher <t.herbrecher@gateware.de>
Signed-off-by: Stefan Roese <sr@denx.de>
Reviewed-by: Tom Rini <trini@konsulko.com>
Cc: Fabio Estevam <festevam@gmail.com>
Reviewed-by: Fabio Estevam <fabio.estevam@freescale.com>

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
Reviewed-by: Tom Rini <trini@konsulko.com>

Across several devices network environment variables are duplicated.
Move these variables to a common include file which insures the environment
variables are reused and insures devices across product lines share the same
values.

Signed-off-by: Franklin S Cooper Jr <fcooper@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>

Signed-off-by: Dinh Nguyen <dinguyen@opensource.altera.com>

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Replace CONFIG_SOCFPGA with CONFIG_ARCH_SOCFPGA.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Move headers to mach-socfpga as well.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Our recent trend is to collect SoC files into arch/arm/mach-(SOC).

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Switch to a more standard way of board select; put the SoC select
into arch/arm/Kconfig and move the board select menu under
arch/arm/mach-socfpga/Kconfig.

Also, consolidate SYS_BOARD, SYS_VENDOR, SYS_SOC, SYS_CONFIG_NAME.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Because all the SOCFPGA boards define CONFIG_SPL_FRAMEWORK
(see include/configs/socfpga_common.h), u-boot.img is automatically
added to the target image list by the top Makefile.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

The compiler option "-Iboard/$(VENDOR)/$(BOARD)" just exists here
for iocsr_config.c to be able to include iocsr_config.h.

Use "..." instead of <...> to include a header in the same directory.

Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>

Default name of spi flash like this "0:0", update it to "spi_flash@0:0".

Signed-off-by: Haikun Wang <haikun.wang@freescale.com>
Acked-by: Simon Glass <sjg@chromium.org>

Drop the code that doesn't use driver model for USB.

Signed-off-by: Simon Glass <sjg@chromium.org>

Enable this command for sandbox.

Signed-off-by: Simon Glass <sjg@chromium.org>

Sometimes the time functions are incorrect due to bad time support on a
board. Add a unit test which tries to detect this.

Signed-off-by: Simon Glass <sjg@chromium.org>

For the distro_bootcmds to succeed on the sandbox a bit of setup is
required (e.g. network configured or host image bound), so running them
by default isn't that useful.

Add a -b/--boot command to the sandbox binary, which triggers the
distro_bootcmds to run after the other command-line commands.

Signed-off-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>
Acked-by: Simon Glass <sjg@chromium.org>

Interrupts transfers timing out is normal, so do not log an error for this.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Marek Vasut <marex@denx.de>

Add proper cache flushing / invalidating for non cache coherent cpus, for now
only enable this for new (driver-model) usb code to avoid regressions.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Reviewed-by: Marek Vasut <marex@denx.de>

Fix taken from the Linux kernel ohci driver.

Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Marek Vasut <marex@denx.de>