If CONFIG_OF_CONTROL is defined, search device tree nodes that are
compatible with "panasonic,uniphier-ehci" and take the base address
from their "reg" property.
Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>
Cc: Marek Vasut <marex@denx.de>

This commit implements the ofdata_to_platdata handler for the UniPhier
serial driver and adds serial device nodes to the device tree sources.
Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>

This commit adds basic device tree sources for UniPhier SoCs/boards.
Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>

If CONFIG_OF_CONTROL is enabled, lib/fdtdec.c is compiled.
It includes <asm/gpio.h> and then <asm/gpio.h> includes
<asm/arch/gpio.h>.  Consequently, all the SoCs that enable
CONFIG_OF_CONTROL must have <asm/arch/gpio.h> even if they do not
support GPIO.

In the first place, GPIO has nothing to do with OF_CONTROL.
It is wrong that lib/fdtdec.c includes GPIO functions; it should
be split into two files, FDT-common things and GPIO things.
It is, however, a pretty big work to fix that correctly.

This is a compromised commit to add a dummy <asm/arch/gpio.h>
to support OF_CONTROL for UniPhier platform.  This dummy header
will be removed after FDT-GPIO stuff is fixed correctly.
Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>

Signed-off-by: NMasahiro Yamada <yamada.m@jp.panasonic.com>

Conflicts:
	drivers/mmc/fsl_esdhc.c
Signed-off-by: NTom Rini <trini@ti.com>

Commit 9f12cd0e has broken SPL EXT support.
This patch update error code check to get SPL EXT support working again.

Tested on a Pandaboard (rev. A3).
Reviewed-by: NSuriyan Ramasami <suriyan.r@gmail.com>
Reviewed-by: NSimon Glass <sjg@chromium.org>
Signed-off-by: NGuillaume GARDET <guillaume.gardet@free.fr>
Cc: Tom Rini <trini@ti.com>

Old kernels cannot handle booting in non-secure (hyp) mode, so when
OLD_SUNXI_KERNEL_COMPAT is set, also set ARMV7_BOOT_SEC_DEFAULT.

Note that whether to booting secure or non-secure can always be overriden
using the bootm_boot_mode environment variable.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

It is not used anywhere.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

Allwinner tells us that this bit of code is the rtc ram being used to detect
coming out of "super-standby" mode, and if that is the case, going out of
self-refresh mode.

Since we do not support "super-standby" mode, this can be dropped.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

Hookup the gmac found on the sun6i / A31 SoCs.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

I noticed that the kernel and u-boot settings were different, double checking
has confirmed that the kernel settings are correct.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

If the target power is connected through a gpio, then give the target some
time to power up before continuing with ahci / sata probing, this avoids
link timeouts, without penalizing other boards where there is no target
power gpio.

Why 500 ms ? I started with 200, that was not enough, then I went to 500 which
worked, lowering it to 350 broke things again, so 500 seems the minimum my
vertex2 needs to be ready to get probed.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

For use together with the hdmi console.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ian.campbell@citrix.com>

Add simplefb support, note this depends on the kernel having support for
the clocks property which has recently been added to the simplefb devicetree
binding.
Signed-off-by: NLuc Verhaegen <libv@skynet.be>
[hdegoede@redhat.com: Use pre-populated simplefb node under /chosen as
 disussed on the devicetree list]
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>.

Change lcd_dt_simplefb_configure_node into a wrapper around the new generic
fdt_setup_simplefb_node helper function.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Tested-by: NStephen Warren <swarren@wwwdotorg.org>
Acked-by: NSimon Glass <sjg@chromium.org>

Add a generic helper to fill and enable simplefb nodes.

The first user of this will be the sunxi display code.

lcd_dt_simplefb_configure_node is also a good candidate to be converted
to use this, but that requires someone to run some tests first, as
lcd_dt_simplefb_configure_node does not honor #address-cells and #size-cells,
but simply assumes 1 and 1 for both.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Tested-by: NStephen Warren <swarren@wwwdotorg.org>
Acked-by: NSimon Glass <sjg@chromium.org>

Besided needing the usual sun6i specific ahb1_reset bits poking, it turns out
that sun6i also needs the drc to be taken out of reset and clocked even though
it is in pass-through mode.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NAnatolij Gustschin <agust@denx.de>

This adds a fixed mode hdmi driver for the sunxi platform. The fixed
mode is a relatively safe 1024x768, more complete EDID handling is
currently not provided. Only HDMI is supported today.

This code is enabled when HPD detects an attached monitor.

Current config is such that 8MB is shaved off at the top of the RAM.
This avoids several memory handling issues, most significant is the fact
that on linux on ARM you are not allowed to remap known RAM as IO. A
clued in display driver will be able to recycle this reserved RAM in
future though.

cfbconsole was chosen as it provides the most important functionality: a
working u-boot console, allowing for the debugging of certain issues
without the need for a UART.
Signed-off-by: NLuc Verhaegen <libv@skynet.be>
[hdegoede@redhat.com: Major cleanups and some small bugfixes]
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NAnatolij Gustschin <agust@denx.de>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

This is a preparation patch for adding support for HDMI out.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

The data sheet just calls it DRAM_CLK_REG, and on sun6i we've both a
dram_clk_cfg and dram_clk_gate, and the sun4i reg matches dram_clk_gate on
sun6i, so name it the same on sun4i.
Signed-off-by: NHans de Goede <hdegoede@redhat.com>
Acked-by: NIan Campbell <ijc@hellion.org.uk>

This patch enhances the boot of origen board by adding support to ext2, bootz, initrd, bootenv loading and boot script.
It still keeps the previous mmc load command if boot script fails.
Signed-off-by: NGuillaume GARDET <guillaume.gardet@free.fr>
Signed-off-by: NMinkyu Kang <mk7.kang@samsung.com>

Daniel discovered a better solution to the problem this was solving, so
don't do what this patch was doing anymore.

This reverts commit 666ba844.
Signed-off-by: NTom Rini <trini@ti.com>

Signed-off-by: NTom Rini <trini@ti.com>

EC1 pins in RCW can be selected as RGMII1, GPIO3, CAN1/2, FTM1 or
SAI1/2. There is a bug that EC3 RGMII could not work when selecting EC1
as other functionality except RGMII. The workaround is to select
ge2_clk125 for eTSEC clock muxing in register SCFG_ETSECCMCR.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

SystemID information could be read through I2C1 from EEPROM
on LS1021ATWR board.

As LS1 is a little-endian processor, getting the version ID by
be32_to_cpu() is wrong. Fix it by using e.version directly.
This change will be compatible for both ARM and PowerPC.

As there is an errata that I2C1 could not work in SD boot,
reading EEPROM through I2C1 is disabled too in SD boot.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Add define CONFIG_SYS_WRITE_SWAPPED_DATA.
For LS1021AQDS and LS1021QTWR nor flash write should swap the
bytes when handle unaligned tail bytes.

Because of the ending, if the date bus width is 16-bits and the
number of bytes is odd, we should swap the byte when write the
last one.
Signed-off-by: NYuan Yao <yao.yuan@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

SCFG_SCFGREVCR is SCFG bit reverse register. This register
must be written with 0xFFFFFFFF before writing to any other
SCFG register. Then other SCFG register could be written in
big-endian mode.

Address: 157_0000h base + 200h offset = 157_0200h
Bit   0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15|16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31
W/R                                   SCFGREV
Reset 0 0 0 0 0 0 0 0 0 0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0  0
0-31
SCFGREV SCFG Bit Reverse Control Filed
32'h 0000_0000 - No bit reverse is applied
32'h FFFF_FFFF - Bit reverse is applied; so 31:0 will be stored/read as
0:31

This patch removes the bit reversing for SCFG registers in
u-boot. It will be implemented through PBI commands in RCW
.pbi
write 0x570200, 0xffffffff
.end
So other SCFG register could be written in big-endian mode
in u-boot or kernel directly.
Signed-off-by: NAlison Wang <alison.wang@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Disable the snoop for slave interface 0, 1 and 2
to avoid the interleaving on the CCI400 BUS.
Signed-off-by: NJason Jin <Jason.Jin@freescale.com>
Signed-off-by: NMinghuan Lian <Minghuan.Lian@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Enable USB EHCI Host on LS1021AQDS
Signed-off-by: NNikhil Badola <nikhil.badola@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Define USB configs for LS1021XA such as CONFIG_SYS_FSL_USB1_ADDR,
CONFIG_USB_MAX_CONTROLLER_COUNT
Signed-off-by: NNikhil Badola <nikhil.badola@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

Set TXFIFOTHRESH to adjust ddr pipeline delay for successful large
usb writes
Signed-off-by: NNikhil Badola <nikhil.badola@freescale.com>
Reviewed-by: NYork Sun <yorksun@freescale.com>

128059b9 ("hush: fix segfault on syntax error") attempted to fix a
segfault on syntax errors, but it broke Ctrl-C handling, and the
assumption that it made, that rcode could not be -1, is incorrect.
Revert this change.
Reported-by: NStephen Warren <swarren@wwwdotorg.org>
Reported-by: NPrzemyslaw Marczak <p.marczak@samsung.com>
Signed-off-by: NRabin Vincent <rabin@rab.in>

Conflicts:
	drivers/serial/serial-uclass.c
Signed-off-by: NTom Rini <trini@ti.com>