Tegra's GPIO driver currently enables pins as GPIO as soon as they're
requested. This is not safe, since the desired direction and output value
are not yet known. This could cause a glitch on the output pins between
gpio_request() and gpio_direction_*(), depending on what values happen to
be in the GPIO controller's in/out and out-value registers vs. the final
desired configuration.
To solve this, defer enabling pins as GPIOs until some gpio_direction_*()
is invoked, and the desired configuration is explicitly programmed.
In theory this change could cause regressions, if code exists that claims
a GPIO, never explicitly sets a direction, and then gets/sets the GPIO
value based on that assumption. However, I've read through all the Tegra-
related board files and device drivers that touch GPIOs and I do not see
such buggy code anywhere.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Tegra's gpio_config_table() currently uses common GPIO APIs. These used
to work without requesting the GPIO, but since commit 2fccd2d96b "tegra:
Convert tegra GPIO driver to use driver model" no longer do so. This
prevents any of the GPIO initialization table from being applied to HW.
Fix gpio_config_table() to directly program the HW to solve this.
Fixes: 2fccd2d96b ("tegra: Convert tegra GPIO driver to use driver model")
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Warren <twarren@nvidia.com>
In order to avoid any assumptions about any device connected to
P2371-2180's expansion connector, the latest pinmux spreadsheet
configures all muxable pins on that connector to be GPIO inputs, with
on-chip pulls where appropriate.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Add DFU support.
Tested by flashing SPL and u-boot.img into SPI NOR flash with the
following commands:
=> setenv dfu_alt_info ${dfu_alt_info_spl}
=> run dfuspi
On the host PC:
$ sudo dfu-util -D SPL -a spl
On the target:
CTRL+C
=> setenv dfu_alt_info ${dfu_alt_info_img}
=> run dfuspi
On the host PC:
$ sudo dfu-util -D u-boot.img -a u-boot
Signed-off-by: Fabio Estevam <fabio.estevam@freescale.com>
The driver assumed that I2C1 and I2C2 were always enabled,
and if they were not, then an asynchronous abort was (silently)
raised, to be caught much later on in the Linux kernel.
Fix this by making I2C1 and I2C2 optional just like I2C3 and I2C4
are.
To make the change binary-invariant, declare I2C1 and I2C2 in
every include/configs/ file which defines CONFIG_SYS_I2C_MXC.
Also, while updating README about CONFIG_SYS_I2C_MXC_I2C1 and
CONFIG_SYS_I2C_MXC_I2C2, add missing descriptions for I2C4 speed
(CONFIG_SYS_MXC_I2C4_SPEED) and slave (CONFIG_SYS_MXC_I2C4_SLAVE)
config options.
Signed-off-by: Albert ARIBAUD (3ADEV) <albert.aribaud@3adev.fr>
The VF610 DDRMC driver code contains settings which are
board-specific. Move these out to boards so that new boards
can define their own without having to modify the driver.
Signed-off-by: Albert ARIBAUD (3ADEV) <albert.aribaud@3adev.fr>
Set missing boot address in bootm command. This fixes the error:
Wrong Image Format for bootm command
ERROR: can't get kernel image!
Reported-by: Uwe Scheffler <scheffler.u@web.de>
Signed-off-by: Soeren Moch <smoch@web.de>
Tested-by: Uwe Scheffler <scheffler.u@web.de>
We should follow 'read->set/clr bit->write' flow for enable_fec_anatop_clock,
otherwise we may overridden configuration before enable_fec_anatop_clock.
Signed-off-by: Peng Fan <Peng.Fan@freescale.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Cc: Fabio Estevam <fabio.estevam@freescale.com>
If defined CONFIG_ENV_IS_IN_MMC, then u-boot environment is saved in
mmc's raw sectors. Otherwise, u-boot environment is saved as a file:
uboot.env.
Signed-off-by: Josh Wu <josh.wu@atmel.com>
Reviewed-by: Bo Shen <voice.shen@gmail.com>
Currently, kwboot only allows dynamic UART boot mode patching for SoCs
with header version 0 (Orion, Kirkwood). This patch now enables this "-p"
feature also for SoCs with header version 1 (Armada XP / 38x etc). With
this its possible now to use the UART boot mode without on images that
are generated for other boot devices, like SPI. So no need to change
BOOT_FROM to "uart" for UART xmodem booting any more.
Signed-off-by: Stefan Roese <sr@denx.de>
Tested-by: Kevin Smith <kevin.smith@elecsyscorp.com>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Cc: Dirk Eibach <eibach@gdsys.de>
As reported by Simon Guinot, commit ade741b3
"arm: mvebu: Call timer_init early before PHY and DDR init" breaks
Kirkwood platforms. As the static variable "init_done" is not
available at that early boot time. This patch moves it to explicitly
to the data section, making it available at that time.
Signed-off-by: Stefan Roese <sr@denx.de>
Reported-by: Simon Guinot <simon.guinot@sequanux.org>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Tested-by: Simon Guinot <simon.guinot@sequanux.org>
Currently booting on A38x is broken. As the current code tries to detect
the SoC family to disable the MMU for the A38x at runtime. But before the
internal registers are switched to the new location (0xf100.0000), this
runtime detection does not work. As all macros / defines are already
assigned to the new location at 0xf100.0000. But the registers are sill
mapped to the default location at 0xd000.0000.
This patch now makes sure, no such runtime detection is used before
the internal registers are configured to the new location. After this,
the remaining cache cleanup is executed.
Signed-off-by: Stefan Roese <sr@denx.de>
Reported-by: Kevin Smith <kevin.smith@elecsyscorp.com>
Cc: Luka Perkov <luka.perkov@sartura.hr>
Simon Glass and Joe Hershberger suggested adding at least one
test case for the CONFIG_DM_ETH plus CONFIG_NETCONSOLE options.
This patch enables NetConsole as a default for the "Banana Pi/Pro"
sunxi boards.
(By the nature of this patch it could probably be extended later
to include all sunxi boards using CONFIG_SUNXI_[EG]MAC.)
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch introduces CONFIG_NETCONSOLE as an option to the
Kconfig system.
Joe Hershberger pointed out that it may not be entirely free of
problems, as many boards predating the driver model define this
symbol directly via include files. In case they're not properly
migrated, their NetConsole might 'vanish' if they start to use
CONFIG_NET or CONFIG_NETDEVICES.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
CONFIG_NETCONSOLE causes common/bootm.c to call eth_unregister()
for network device shutdown. However, with CONFIG_DM_ETH this
function is no longer defined.
This is a workaround to avoid the call in that case, and solely
rely on eth_halt(). In case this is insufficient, a proper way
to unregister / remove network devices needs to be implemented.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This patch uses the eth_is_active() function to work around
issues that prevented compilation with the newer driver model.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
The previous eth_device struct returned by eth_get_dev() allowed
code to directly query the state member field. However, with
CONFIG_DM_ETH this data gets encapsulated (i.e. private), and
eth_get_dev() returns a udevice struct 'abstraction' instead.
This breaks legacy code relying on the former behaviour - e.g.
netconsole.
(see http://lists.denx.de/pipermail/u-boot/2015-June/216528.html)
The patch introduces a method to retrieve the ethernet device
state in a 'clean' and uniform way, supporting both legacy code
and driver model. The new function eth_is_active() accepts a
device struct pointer and tests it for ETH_STATE_ACTIVE.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Currently when phy device is created the link variable is
initialized to 1 which denoted phy link is already up. On a power
reset there is no issue as phy status register link status will
not be set, so phy auto negotiate will be started. But when a cpu
reset is issued (ex: dra72x-evm) phy's link status bit is already
set which leads to assume that link is already setup in
genphy_update_link() initial check which results in ehternet not
working. So do not assume that link is already up and on phy
device create set link to zero. This is verified on dra72x-evm.
Reported-by: Franklin S Cooper Jr <fcooper@ti.com>
Signed-off-by: Mugunthan V N <mugunthanvnm@ti.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
The 7" Q8 tablet enclosure is used for a ton of slightly different cheap
chinese tablets. There are some differences in which accelerometer /
wifi is used, but other then that these are all the same from a u-boot /
kernel pov.
When we get to adding accelerometer support the plan is to add some kind
of autodetection and mangle the dt accordingly (likely using the new quirks
mechanism).
For now this is a non issue as we do not yet have accelerometer
support, and in the future, some sort of auto-detect is the way to go
as we cannot expect users to exactly know what is inside their tablet.
The dts[i] files this commit adds are identical to the ones submitted
to the upstream kernel.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
In recent allwinner kernel sources the mmc/sdio clk-delay settings have
been slightly tweaked, and for sun9i they are completely different then
what we are using.
This commit brings us in sync with what allwinner does, fixing problems
accessing sdcards on some A33 devices (and likely others).
For pre sun9i hardware this makes the following changes:
-At 400Khz change the sample delay from 7 to 0 (first introduced in A31 sdk)
-At 50 Mhz change the sample delay from 5 to 4 (first introduced in A23 sdk)
-Above 50 MHz change the out delay from 2 to 1 (first introduced in A20 sdk)
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
We know when u-boot is written to its own partition, in this case the
layout always is:
eb 0 spl
eb 1 spl-backup
eb 2 u-boot
eb 3 u-boot-backup
eb: erase-block
So if we cannot load u-boot from its primary offset we know exactly where
to look for it.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
This patch makes use of the previous changes to add a new "fel" boot
target for sunxi boards.
When booting via FEL, it's often desirable to work around the absence
of other (usable) boot devices - or to be able to override them,
deviating from the standard boot sequence. To achieve this, the "fel"
boot target gets the highest priority, but won't actually do anything
unless certain criteria are met.
The "bootcmd_fel" implementation proposed here first tests if an actual
FEL boot takes place (using the "fel_booted" env var), and secondly
checks that "fel_scriptaddr" was set (originating from the 'loader',
i.e. the sunxi-tools fel utility). If both checks pass, then it will
try to execute the boot script (boot.scr) at the given address. In case
of an error (e.g. an invalid image), the source command might return
"false", causing "distro_bootcmd" to proceed with the next boot target.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
This patch extends the misc_init_r() function on sunxi boards
to test for the presence of a suitable "sunxi" SPL header. If
found, and the loader ("fel" utility) provided a non-zero value
for the boot.scr address, then the corresponding environment
variable fel_scriptaddr gets set.
misc_init_r() also sets (or clears) the "fel_booted" variable depending
on the active boot device, using the same logic as spl_boot_device().
The goal is to provide sufficient information (within the U-Boot
environment) to make intelligent decisions on how to continue the boot
process, allowing specific customizations for the "FEL boot" case.
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
This patch follows up on a discussion of ways to improve support
for the sunxi FEL ("USB boot") mechanism, especially with regard
to boot scripts, see:
https://groups.google.com/d/msg/linux-sunxi/wBEGUoLNRro/rHGq6nSYCQAJ
The idea is to convert the (currently unused) "pad" bytes in the
SPL header into an area where data can be passed to U-Boot. To
do this safely, we have to make sure that we're actually using
our "sunxi" flavor of the SPL, and not the Allwinner boot0.
The modified mksunxiboot introduces a special signature to the
SPL header in place of the "pub_head_size" field. This can be
used to reliably distinguish between compatible versions of sunxi
SPL and anything else (older variants or Allwinner's boot0).
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Siarhei Siamashka <siarhei.siamashka@gmail.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The sunxi platform currently doesn't seem to make any use of the
asm/arch-sunxi/spl.h file. This patch moves some declarations from
tools/mksunxiboot.c into it.
This enables us to reuse those definitions when extending the
sunxi board code (boards/sunxi/boards.c).
Signed-off-by: Bernhard Nortmann <bernhard.nortmann@web.de>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
crt0.S will both memset the bss sectioan and call board_init_r for us,
so there is no need to do either ourselves.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
For the upcoming nand support we need a bigger heap, esp. ubi[fs] uses
quite a bit of memory, increase the heap size to 64 MB.
Our video code returns unused reserved framebuffer memory to the kernel
before booting it. Drop the #ifdef-ery and simply always reserve 16 MB.
Adjust bootm_size for the above changes.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
Move some #define-s around from one #ifdef block to another to
reduce the number of #ifdef blocks (note this causes no functional
changes even though the conditions are not always exactly the same)
and move generic #include statements to the top.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
Acked-by: Ian Campbell <ijc@hellion.org.uk>
We only prefix the board defconfig name with the SoC if the SoC is part
of the official board-name, such as with Olimex OLinuxIno boards, for the
Wobo i5 this is not the case so remove the A10s prefix from the defconfig
filename.
Also fix the double listing of A10s-OLinuXino-M_defconfig in MAINTAINERS
which the original Wobo i5 addition commit introduced.
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
The logic to calculate the number of E820 table entries is wrong
when walking through the FSP HOB tables. Fix it.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Tested-by: Jian Luo <jian.luo4@boschrexroth.de>
Acked-by: Simon Glass <sjg@chromium.org>
Now that PCI bridges are probed before relocation we need additional memory.
Each PCI bridge takes 240 bytes at present since it uses the same uclass as
the PCI controller. Probably we should split this out so that bridges have
their own uclass.
Expand the memory on link so that it works correctly.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
The commit: d9dbb97be0
"mmc: dw_mmc: Zap endless timeout" removed endless loop waiting for end
of dw mmc transfer.
For some workloads - dfu test @ Odroid XU3 (sending 8MiB file) -
and SD cards (e.g. MicroSD Kingston 4GiB, Adata 4GiB)
the default timeout is to short.
The new value - 4 minutes (240 seconds) - is the same as the one used in
Linux kernel driver. Such fix should be good enough until we come up
with better fix for this issue.
Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
Cc: Marek Vasut <marex@denx.de>
Cc: Pantelis Antoniou <panto@antoniou-consulting.com>
Cc: Tom Rini <trini@konsulko.com>
Tested-by: Przemyslaw Marczak <p.marczak@samsung.com>
Give a full URL for a working nds32 toolchain for U-Boot.
Signed-off-by: Bin Meng <bmeng.cn@gmail.com>
Reviewed-by: Masahiro Yamada <yamada.masahiro@socionext.com>
As 'time(0) | getpid()' will have a lot of duplicated value. It is not a
expected behavior. We expect different value for the seed when when run
it in many times.
So this patch will left shift the getpid() and add to time(0). That
avoid duplicated value.
Test command is like:
% RUN=0; while [ $RUN -lt 10000 ]; do
tools/gen_eth_addr; RUN=$(($RUN+1)); done | sort | uniq | wc -l
10000
This patch is incorporated with suggestions made by Wolfgang Denk and Andreas
Bießmann. Thanks them a lot.
Signed-off-by: Josh Wu <josh.wu@atmel.com>
Acked-by: Andreas Bießmann <andreas.devel@googlemail.com>
Acked-by: Wolfgang Denk <wd@denx.de>
Tested-by: Wolfgang Denk <wd@denx.de>
DRAM region 0x3f000000 - 0x3fffffff is reserved for OP-TEE. Touching
0x3f000000 memory location from unsecure world causes the board
to hang.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Use DM for the pl01x serial driver on hikey. Also allow UART0 or
UART3 to be chosen via Kconfig.
By default we now output to UART3 as the latest version of ATF outputs
to this UART. Also UART3 comes out on the LS connector, as opposed to
UART0 which goes to a unpopulated header.
As part of this change we also enable CONFIG_BOARD_EARLY_INIT_F and
call the pinmux configuration code for the UART. Before we were
relying on ATF having already configured the pin configuration.
NB: Upstream Linux kernel doesn't yet support UART3, so serial console
will still be output on UART0 when booting a upstream kernel.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
Most platforms enable these options from Kconfig rather
than the configs header file.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
This causes exceptions and other strange behaviour
when enabling CONFIG_SYS_MALLOC_F_LEN which is required to
migrate the serial driver over to DM_SERIAL.
As GD_FLG_FULL_MALLOC_INIT flag gets reset, after relocation
we don't end up using the full malloc which ultimately ends up
causing a synchronus abort.
Signed-off-by: Peter Griffin <peter.griffin@linaro.org>
Reviewed-by: Simon Glass <sjg@chromium.org>