The data caches should be invalided once during startup. This should
also be done when we do not have the MMU enabled in barebox because
the Kernel does not invalidate the caches during start.
To make this sure this patch enables the arm_early_mmu_cache_invalidate
function even if MMU support is disabled. Additionally this patch adds
calls to arm_early_mmu_cache_invalidate in start.c and uncompress.c.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Some of the irq can still be on after a reset or power on as the IP are
powered by the backup power. This could lead to an interrupt dead lock
when the kernel boot. So disable them before booting.
Signed-off-by: Jean-Christophe PLAGNIOL-VILLARD <plagnioj@jcrosoft.com>
Cc: Nicolas Ferre <nicolas.ferre@atmel.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
we need to power on the PLL when enabling the USB clock.
Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
this fix :
arch/arm/boards/crystalfontz-cfa10036/cfa10036.c: In function 'cfa10036_devices_init':
arch/arm/boards/crystalfontz-cfa10036/cfa10036.c:123:3: warning: implicit declaration of function 'imx_gpio_mode' [-Wimplicit-function-declaration]cfa10036: fix warning
Signed-off-by: Eric Bénard <eric@eukrea.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
For the MPC85xx family of SOCs Linux expects any boot firmware
information to be passed in the first 64MiB of memory. This adds support
to ensure that the device tree is relocated to a valid location if it is
outside that address range.
For the other SOC family currently present in the ppc architecture, the
default is not to relocate as at Linux startup the virtual address
equals the physical address.
Signed-off-by: Renaud Barbier <renaud.barbier@ge.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds support for populating derived properties of the mpc85xx SOC
which are not typically included in the dtb directly.
This update is based on U-Boot code from common/fdt_support.c
and arch/powerpc/cpu/mpc85xx/fdt.c - version git-2b26201.
Signed-off-by: Renaud Barbier <renaud.barbier@ge.com>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
- compile in thumb2 mode for a smaller binary
- increase TEXT_BASE to get more memory space
- set malloc size to zero for autodetection
- enable some more commands
- disable MCI_STARTUP, it's no longer needed and saves some time
during boot
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Convert to static inline functions and use lower case letters for function
names. Also, include mach/debug_ll.h when an architecture provides support
for debug_ll, not only when it's actually enabled. This allows architecures
to put some UART initialization code into mach/debug_ll.h which is compiled
out when debug_ll is disabled.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Card detect should be rather muxed as gpio then
mmc0_sdcd, as driver does not support this.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Add function for indirect access of the mmd registers, based on linux.
phy_read_mmd_indirect
phy_write_mmd_indirect
Also clean some private mmd functions
Signed-off-by: Jan Weitzel <j.weitzel@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Using here now the initialisiation functions in
am33xx_generic.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Use now the initialisation functions from am33xx_generic.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
There is a lot of duplicate lowlevel code between the
am33xx boards. Move this code to am33xx_generic and
create structs for sdram settings.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Function should use lower case letters. Also give this function
more namespace. While at it include io.h rather than asm/io.h.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
omap_save_bootinfo derefences the argument passed to barebox without
checking it for validity. This breaks 2nd stage booting where r0
is undefined. The best we can do is to check whether the pointer is
somewhere in SRAM and is word aligned. This at least makes sure that
we do not oops. This introduces SoC specific xxx_save_bootinfo variants
since the SRAM addresses/sizes differ between SoCs.
Additionally fix the prototype for omap_save_bootinfo. It uses r0, so
it must be passed this variable.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
This adds a omap_set_bootmmc_devname function which can be used
to specify the MMC boot device. In an xload configuration this
is then used to load the next stage. In a regular build this is
used to mount /boot.
Also a device_detect_by_name is added to allow to disable probing
for sd/mmc cards during startup.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
SPI NOR flash can be detected by driver so no platform data
is necessary.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
-Added mlo spi NOR copy handler
-This handler will convert the MLO to big endian
-Tested with pcm051 board
Signed-off-by: shravan <shravan.k@phytec.in>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The i.MX53 qsb has 1GiB of memory, but it's divided into two non
contigous banks. Remove the wrong memory node since it's overwritten
with the correct values during runtime anyway.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The oscillator frequency varies on different AM33xx boards.
Pass the osc frequency from lowlevel board code
to set the correct one on every board.
Signed-off-by: Teresa Gámez <t.gamez@phytec.de>
Reviewed-by: Jan Luebbe <jlu@pengutronix.de>
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
The esdctl devices are currently not in the devicetrees. this means
they are not registered when booting from the devicetree. Move the
device registration from soc_devices_init to soc_init which is called
even with devicetree support so that we get esdctl devices.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
gpiolib user have nothing to define in their machine
specific gpio.h, so do not include it.
The only thing they could define would be ARCH_NR_GPIOS,
but currently no architecture defines it. Should an architecure
feel the need to do it this would be a good opportunity to
get rid of this limitation.
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Sascha Hauer
Jean-Christophe PLAGNIOL-VILLARD
Eric Bénard
Renaud Barbier
Teresa Gámez
Jan Luebbe
Antony Pavlov
Jan Weitzel
shravan