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

When u-boot is compiled for PXA25x processor, pxa/start.S is calling
cpu_init_crit by BL instruction. BL is overwriting lr register so
relocate_code is going into infinite loop. This patch preservs lr
register in r12 before calling cpu_init_crit and after function returns
restores it.

Signed-off-by: Lukasz Dalek <luk0104@gmail.com>
Acked-by: Marek Vasut <marex@denx.de>

Signed-off-by: Peter Meerwald <p.meerwald@bct-electronic.com>

In commit abbab703 we added a prototype for hw_watchdog_init, in a
Freescale i.MX watchdog driver.  We did not however guard it with an
__ASSEMBLY__ check and broke some platforms.  Move this to the end of
the file with other prototypes and guard it.

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

Signed-off-by: Wolfgang Denk <wd@denx.de>
cc: Anatolij Gustschin <agust@denx.de>

Commit fb6a9aab (LCD: display 32bpp decompressed bitmap image)
broke the check that allowed U-Boot to display 8 bpp BMPs on a 16
bpp LCD screen, effectively turning this feature off.

Restore this feature by changing the check back to the same meaning
it originally had.
To avoid future confusion, the check has also been rephrased to make
its meaning clear.

Signed-off-by: Nikita Kiryanov <nikita@compulab.co.il>
Signed-off-by: Igor Grinberg <grinberg@compulab.co.il>

Previously, the call to draw_logo() was happening irrespective
of whether we have selected logo or LCD console.
With this patch we call draw_logo() only when CONFIG_CMD_BMP is selected.

This would even fix the following compilation warning:
exynos_fb.c: In function 'draw_logo':
exynos_fb.c:74:8: warning: variable 'addr' set but not used
[-Wunused-but-set-variable]
exynos_fb.c:73:9: warning: variable 'y' set but not used
[-Wunused-but-set-variable]
exynos_fb.c:73:6: warning: variable 'x' set but not used
[-Wunused-but-set-variable]

Signed-off-by: Ajay Kumar <ajaykumar.rs@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

Remove unused variable disp_info to fix the following compilation warning:
exynos_dp.c: In function 'exynos_init_dp':
exynos_dp.c:860:23: warning: variable 'disp_info' set but not used
[-Wunused-but-set-variable]

Signed-off-by: Ajay Kumar <ajaykumar.rs@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

This patch resolves the following build errors for I2C driver in
VCMA9:

In file included from s3c24x0_i2c.c:40:0:
s3c24x0_i2c.h:38:17: error: field 'id' has incomplete type
s3c24x0_i2c.c:70:39: error: 'CONFIG_MAX_I2C_NUM' undeclared here (not
in a function) s3c24x0_i2c.c:70:31: warning: 'i2c_bus' defined but
not used [-Wunused-variable] s3c24x0_i2c.c:72:12: warning:
'i2c_busses' defined but not used [-Wunused-variable]

Signed-off-by: Rajeshwari Shinde <rajeshwari.s@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

Add DVI output support and make it the default video output.

Currently the CLAA WVGA panel is supported, but this panel has to be purchased
separately, so using the DVI output as the default would allow more people to
try the splash screen feature on a mx51evk.

If someone still wants to use the CLAA WVGA, just set the panel variable as:
set panel claa

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>

Since commit de7d02ae (mx6qsabresd: add usdhc2 and usdhc4 support)
SDHC3 device node is no longer 0, which breaks loading a uImage from SDHC3.

Fix it by adapting the default environment to use CONFIG_SYS_MMC_ENV_DEV as the
correct mmc node for loading the kernel from.

While at it, go back to using SDHC3 as the default mmc, since we have Yocto
images that generate an SD card containing U-boot,kernel and rootfs, so it is
more convenient to keep using SDHC3 as it was originally.

Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>
Tested-by: Fabio Estevam <fabio.estevam@freescale.com>

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>

The CONFIG_SYS_CBSIZE has been change to 512 to avoid runtime errors as:

,---[ Runtime error ]
| Hit any key to stop autoboot:  0
| MX53LOCO U-Boot > pri netboot
| netboot=echo Booting from net ...; run netargs; if test ...
| prefetch abort
| pc : [<20747368>]          lr : [<20747365>]
| sp : af566e20  ip : 00000000     fp : 00000000
| r10: 00000002  r9 : af6dfc28     r8 : af566f58
| r7 : af6dfc10  r6 : 00000001     r5 : 00000002  r4 : 74206669
| r3 : 00000000  r2 : 00000060     r1 : 00000020  r0 : 0000018e
| Flags: nZCv  IRQs off  FIQs off  Mode SVC_32
| Resetting CPU ...
|
| resetting ...
`---

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>

For a generic environment, we shouldn't have a fixed rootfs filesystem
so we drop it from env.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>

This rework the environment to use tabs for environment setting as
done in other boards.

Signed-off-by: Otavio Salvador <otavio@ossystems.com.br>

Use a common watchdog driver for all these cpus.

Signed-off-by: Troy Kisky <troy.kisky@boundarydevices.com>
Acked-by: Stefano Babic <sbabic@denx.de>

Add dataflash boot support on at91sam9x5ek board

Signed-off-by: Bo Shen <voice.shen@atmel.com>
Acked-by: Andreas Bießmann <andreas.devel@googlemail.com>

Fix coldfire serial driver bindings no more relocated to ram after last
changes to drivers/serial/serial.c (regression).
Serial initialization in ram has to be called after that gd->reloc_off is
calculated.

Signed-off-by: Angelo Dureghello <sysamfw@gmail.com>
Cc: Jason Jin <jason.jin@freescale.com>
Acked-by: Jens Scharsig (BuS Elektronik) <esw@bus-elektronik.de>
Tested-by: Jens Scharsig (BuS Elektronik) <esw@bus-elektronik.de>

We enable console multiplexing and use both serial and LCD for stdout/stderr.
Initially, u-boot output console is observed via serial port.
If you also have a DP panel connected onto your SMDK5250 board,
you can switch to LCD console by typing "setenv stdout lcd".
You can always switch back to serial using "setenv stdout serial".
You can switch error console(stderr) as well, using similar commands.

Signed-off-by: Ajay Kumar <ajaykumar.rs@samsung.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

u-boot compilation fails for smdk5250 when we deselect CONFIG_LCD
from the main config file.

Following error was observed:
drivers/video/libvideo.o: In function `exynos_lcd_init':
/home/ajay/u-boot-samsung/drivers/video/exynos_fb.c:68: undefined reference to `lcd_set_flush_dcache'

This is because exynos video drivers have dependency on CONFIG_LCD.

Signed-off-by: Ajay Kumar <ajaykumar.rs@samsung.com>
Acked-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

Mobile storage is used the CLK_DIV_FSYS3 value.

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

To get exactly clock value for mmc, support the get_mmc_clk() like
set_mmc_clk().

Signed-off-by: Jaehoon Chung <jh80.chung@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
Signed-off-by: Minkyu Kang <mk7.kang@samsung.com>

IGEP board PC16550D (ns16550) UART doesn't set the
Transmitter Empty (TEMT) Bit in SPL. This makes
U-Boot to hang while waiting for TEMT. Add the
CONFIG_SYS_NS16550_BROKEN_TEMT config option to
avoid this issue.

Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>

Some platforms (e.g. IGEPv2 board) has a broken ns16550 UART that
does not set the TEMT bit when the transmitter is empty in SPL.
This makes U-Boot to hang while waiting for TEMT to be set.

Add a new option to avoid this:

CONFIG_SYS_NS16550_BROKEN_TEMT

16550 UART set the Transmitter Empty (TEMT) Bit when all output
has finished and the transmitter is totally empty. U-Boot waits
for this bit to be set to initialize the serial console. On some
broken platforms this bit is not set in SPL making U-Boot to
hang while waiting for TEMT. Define this option to avoid it.

Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>

On clobber, do a cleanup even for SPL build else, we'd have
MLO.byteswap remaining for certain platforms like am33xx SPL builds

Signed-off-by: Nishanth Menon <nm@ti.com>

When booting an IGEPv2 board from NAND with SPL, U-Boot hangs
trying to read the OMAP General Purpose Memory Controller (GPMC).

The reason is that the GPMC initialization function is called
inside spl_board_init() and this function is only executed when
CONFIG_SPL_BOARD_INIT is defined.

Signed-off-by: Javier Martinez Canillas <javier.martinez@collabora.co.uk>
Acked-by: Enric Balletbo i Serra <eballetbo@iseebcn.com>

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

To make bootm sub-command feature working on archs with CONFIG_NEEDS_MANUAL_RELOC,
the name pointers in the sub-command table need to be relocated.

Signed-off-by: Daniel Schwierzeck <daniel.schwierzeck@gmail.com>

The two modified lines were indented with spaces.
They are now indented with tabs.

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

The timer may be inited in arch_cpu_init() so it is not safe to make a
bootstage mark before this is called. Arrange the code to fix this.

Note: The question was raised as to why we don't keep all archs in sync.
PowerPC doesn't have specific bootstage markers at present (although it
does use boot progress). I hope that the generic board series will solve
this problem in general, but in the meantime this is a real problem, and
only in ARM.

We now get a correct time for board_init_f:

Timer summary in microseconds:
       Mark    Elapsed  Stage
          0          0  reset
    100,000    100,000  spl_start
    848,530    748,530  board_init_f
    907,301     58,771  board_init_r
    910,478      3,177  board_init

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

SPL u-boot may call do_reset() which depends on interrupts.o and reset.o.
So make them also appear in SPL.

Signed-off-by: Tom Wai-Hong Tam <waihong@chromium.org>
Signed-off-by: Simon Glass <sjg@chromium.org>

Some boards want to report more than just memory size. For example, it
might be useful to display the memory type (DDR2, DDR3) or manufacturer.

Add a weak function to support this requirement, accessed through a new
'meminfo' command.

Any example of the DRAM: output is below, just for illustration:

SMDK5250 # meminfo
DRAM:  2 GiB Elpida DDR3 @ 800MHz

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