Simon Glass [Mon, 3 Aug 2015 00:07:21 +0000 (18:07 -0600)]
x86: Enable debug UART for Minnowmax
Enable the debug UART and emit a single 'a' early in the init sequence to
show that it is working.
Unfortunately the debug UART implementation needs a stack to work. I cannot
seem to remove this limitation as the absolute 'jmp %eax' instruction goes
off into the weeds.
So this means that the character output cannot be any earlier than
car_init_ret, where memory is available for a stack.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Lukasz Majewski <l.majewski@samsung.com>
Simon Glass [Thu, 30 Jul 2015 19:40:39 +0000 (13:40 -0600)]
dm: core: Add a way to set a device name
Some devices are bound entirely by probing and do not have the benefit of
a device tree to give them a name. This is very common with PCI and USB. In
most cases this is fine, but we should add an official way to set a device
name. This should be called in the device's bind() method.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Tue, 28 Jul 2015 17:53:14 +0000 (11:53 -0600)]
sandbox: Enable devres subsystem
This should be used for sandbox. We can convert at least one driver to use
it, but in the meantime, enable the feature so that the code is
build-tested.
devres: add debug command to dump device resources
This new command can dump all device resources associated to
each device. The fields in every line shows:
- The address of the resource
- The size of the resource
- The name of the release function
- The stage in which the resource has been acquired (BIND/PROBE)
Currently, there is no driver using devres, but if such drivers are
implemented, the output of this command should look like this:
Currently, Devres requires additional 16 byte for each allocation,
which is not so insignificant in some cases.
Add CONFIG_DEVRES to make this framework optional.
If the option is disabled, devres functions fall back to
non-managed variants. For example, devres_alloc() to kzalloc(),
devm_kmalloc() to kmalloc(), etc.
Because devres_head is also surrounded by an ifdef conditional,
there is no memory overhead when CONFIG_DEVRES is disabled.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Suggested-by: Simon Glass <sjg@chromium.org> Acked-by: Simon Glass <sjg@chromium.org>
devres: add devm_kmalloc() and friends (managed memory allocators)
devm_kmalloc() is identical to kmalloc() except that the memory
allocated with it is managed and will be automatically released
when the device is removed/unbound.
Likewise for the other variants.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Simon Glass <sjg@chromium.org>
In U-Boot's driver model, memory is basically allocated and freed
in the core framework. So, low level drivers generally only have
to specify the size of needed memory with .priv_auto_alloc_size,
.platdata_auto_alloc_size, etc. Nevertheless, some drivers still
need to allocate/free memory on their own in case they cannot
statically know the necessary memory size. So, I believe it is
reasonable enough to port Devres into U-boot.
Devres, which originates in Linux, manages device resources for each
device and automatically releases them on driver detach. With devres,
device resources are guaranteed to be freed whether initialization
fails half-way or the device gets detached.
The basic idea is totally the same to that of Linux, but I tweaked
it a bit so that it fits in U-Boot's driver model.
In U-Boot, drivers are activated in two steps: binding and probing.
Binding puts a driver and a device together. It is just data
manipulation on the system memory, so nothing has happened on the
hardware device at this moment. When the device is really used, it
is probed. Probing initializes the real hardware device to make it
really ready for use.
So, the resources acquired during the probing process must be freed
when the device is removed. Likewise, what has been allocated in
binding should be released when the device is unbound. The struct
devres has a member "probe" to remember when the resource was
allocated.
CONFIG_DEBUG_DEVRES is also supported for easier debugging.
If enabled, debug messages are printed each time a resource is
allocated/freed.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com> Acked-by: Simon Glass <sjg@chromium.org>
Simon Glass [Wed, 8 Jul 2015 02:53:44 +0000 (20:53 -0600)]
dm: Support address translation for simple-bus
The 'ranges' property can be used to specify a translation from the system
address to the bus address. Add support for this using the dev_get_addr()
function, which devices should use to find their address.
Simon Glass [Wed, 8 Jul 2015 02:53:41 +0000 (20:53 -0600)]
net: smsc95xx: Prepare for conversion to driver model
At present struct eth_device is passed around all over the place. This does
not exist with driver model. Add explicit arguments instead, so that with
driver model we can pass the correct things.
Simon Glass [Fri, 17 Jul 2015 15:22:07 +0000 (09:22 -0600)]
dm: Make regmap and syscon optional
Not all boards use garbage collection in their link step, so we should avoid
adding options that rely on this for prevention of code bloat. Add separate
Kconfig options for syscon and regmap uclasses.
York Sun [Mon, 3 Aug 2015 19:02:04 +0000 (12:02 -0700)]
lib/fdtdec: Fix fdt_addr_t and fdt_size_t typedef
fdt_addr_t is a physical address. It can be either 64-bit or 32-bit,
depending on the architecture. It should be phys_addr_t instead of
u64 or u32. Similarly, fdt_size_t is changed to phys_size_t.
Signed-off-by: York Sun <yorksun@freescale.com> CC: Simon Glass <sjg@chromium.org>
Simon Glass [Mon, 3 Aug 2015 14:19:37 +0000 (08:19 -0600)]
exynos: Add support for spring
Spring is the first ARM-based HP Chromebook 11. It is similar to snow
and it uses the same Samsung Exynos5250 chip. But has some unusual
features. Mainline support for it has lagged snow (both in kernel and
U-Boot). Now that the exynos5 code is common we can support spring just
by adding a device tree and a few lines of configuration.
Simon Glass [Mon, 3 Aug 2015 14:19:27 +0000 (08:19 -0600)]
exynos: Add common board code for exynos5 boards that use device tree
Some boards use device tree for almost all board-specific configuration.
They therefore do not need their own separate board code, but can all use
the same version. Add a common version of the board code. It uses the
PMIC, regulator and video bridge uclasses. This will support smdk5250,
smdk5420, snow, spring, pit and pi.
Simon Glass [Mon, 3 Aug 2015 14:19:26 +0000 (08:19 -0600)]
exynos: dts: Drop the old TPS65090 I2C node
While the AP can access the main PMIC on snow, it must coordinate with the
EC which also wants access. Drop the old definition, which can in principle
generate collision errors. We will use the new arbitration driver instead.
Simon Glass [Fri, 3 Jul 2015 00:16:16 +0000 (18:16 -0600)]
dm: gpio: Check a GPIO is valid before using it
Since a gpio_desc is allowed to be invalid we should return an error
indicating that the operation cannot be completed. This can happen if the
GPIO is optional - e.g. some devices may have a reset line and some may
not.
Simon Glass [Fri, 3 Jul 2015 00:16:11 +0000 (18:16 -0600)]
exynos: spi: Convert the timeout to debug()
Since the timeout is reported through normal channels, and is sometimes
expected (e.g. if the bus is being probed for a non-existent device),
don't display the message in the driver.
In general, drivers should not write to the console as this limits their
usefulness in error conditions.
Simon Glass [Fri, 3 Jul 2015 00:16:10 +0000 (18:16 -0600)]
dm: video: Add support for the NXP PTN3460 bridge
This chip provides an eDP to LVDS bridge which is useful for SoCs that don't
support LVDS displays (or it would waste scarce pins). There is no setup
required by this chip, other than to adjust power-down and reset pins, and
those are managed by the uclass.
Simon Glass [Fri, 3 Jul 2015 00:16:09 +0000 (18:16 -0600)]
dm: video: Add support for the Parade PS8622/625 bridge
This chip provides an eDP to LVDS bridge which is useful for SoCs that don't
support LVDS displays (or it would waste scarce pins). The setup is included
in the device tree.
Simon Glass [Mon, 3 Aug 2015 14:19:20 +0000 (08:19 -0600)]
video: Work around lack of pinctrl
We haven't quite got pinctrl ready to apply to mainline. We don't want to
GPIO pull-up/down support to the driver model GPIO layer either. So work
around this for now.
Simon Glass [Fri, 3 Jul 2015 00:16:08 +0000 (18:16 -0600)]
dm: video: Add support for video bridges
A video bridge typically converts video from one format to another, e.g.
DisplayPort to LVDS. Add driver model support for these with a simple
interface to control activation and backlight. The uclass supports GPIO
control of power and reset lines.
Simon Glass [Fri, 3 Jul 2015 00:16:06 +0000 (18:16 -0600)]
dm: power: Don't return an error when regulators are not autoset
Not all regulators can be set up automatically. Adjust the code so that
regulators_enable_boot_on() will return success when some are skipped.
Only genuine errors are reported.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>
Simon Glass [Fri, 3 Jul 2015 00:16:01 +0000 (18:16 -0600)]
dm: power: Add support for S5M8767 regulators
This PMIC is used with SoCs which need a combination of BUCKs and LDOs. The
driver supports changing voltage and enabling/disabling each regulator. It
supports the standard device tree binding and supports driver model.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>
Simon Glass [Fri, 3 Jul 2015 00:16:00 +0000 (18:16 -0600)]
dm: power: Add support for the S5M8767 PMIC
This PMIC is used with SoCs which need a combination of BUCKs and LDOs. The
driver supports probing and basic register access. It supports the standard
device tree binding and supports driver model. A regulator driver can be
provided also.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>
Simon Glass [Fri, 3 Jul 2015 00:15:59 +0000 (18:15 -0600)]
dm: power: Add support for TPS65090 FETs
The TPS65090 has 7 FETs which are modelled as regulators. This allows them
to be controlled by drivers easier, accessed through the 'regulator' command
and used by other drivers.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>
Simon Glass [Fri, 3 Jul 2015 00:15:58 +0000 (18:15 -0600)]
dm: power: Add a new driver for the TPS65090 PMIC
The existing TPS65090 driver does not support driver model. Add a new one
that does. This can be used as a base for a regulator driver also. It uses
the standard device tree binding.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Przemyslaw Marczak <p.marczak@samsung.com>
Simon Glass [Fri, 3 Jul 2015 00:15:53 +0000 (18:15 -0600)]
exynos: serial: Refactor init code for debug UART
The debug UART code needs to perform the same init as the normal UART
driver. In preparation for this, move the init code into two functions, one
for the basic init and one for setting the baud rate. This will make adding
debug UART support easier.
Simon Glass [Mon, 3 Aug 2015 14:19:24 +0000 (08:19 -0600)]
cros_ec: Support the LDO access method used by spring
Add a driver to support the special LDO access used by spring. This is a
custom method in the cros_ec protocol - it does not use an I2C
pass-through.
There are two implementation choices:
1. Write a special LDO driver which can talk across the EC. Duplicate all
the logic from TPS65090 for retrying when the LDO fails to come up.
2. Write a special I2C bus driver which pretends to be a TPS65090 and
transfers reads and writes using the LDO message.
Either is distasteful. The latter method is chosen since it results in less
code duplication and a fairly simple (30-line) implementation of the core
logic.
The crosec 'ldo' subcommand could be removed (since i2c md/mw will work
instead) but is retained as a convenience.
Simon Glass [Mon, 3 Aug 2015 14:19:23 +0000 (08:19 -0600)]
dm: cros_ec: Convert the I2C tunnel code to use driver model
The Chrome OS EC supports tunnelling through to an I2C bus on the EC. This
currently uses a copy of the I2C command code and a special 'crosec'
sub-command.
With driver model we can define an I2C bus which tunnels through to the EC,
and use the normal 'i2c' command to access it. This simplifies the code and
removes some duplication.
Add an I2C driver which tunnels through to the EC. Adjust the EC code to
support binding child devices so that it can be set up. Adjust the existing
I2C xfer function to fit driver model better.
For now the old code remains to allow things to still work. It will be
removed in a later patch once the new flow is fully enabled.
Simon Glass [Fri, 3 Jul 2015 00:15:50 +0000 (18:15 -0600)]
exynos: dts: Support EC tunnel and main TPS65090 regulator
On pit and pi the TPS65090 regulator is connected only to the EC and we
must use a tunnel to get to it. The existing U-Boot support relies on a
special driver. Add a tunnel definition so that the new device-model
TPS65090 driver can be used unmodified.
Simon Glass [Fri, 3 Jul 2015 00:15:47 +0000 (18:15 -0600)]
exynos: i2c: Tidy up the driver model code
The existing driver model implementation uses the old non-driver-model code
to operate, but has become impossibly tangled as a result. The actual
algorithm is quite simple.
Also the normal-speed and high-speed buses are quite different and it
doesn't seem that useful to put them in the same driver.
Finally, there is a bug which breaks communication with the Maxim sound
codec and may cause problems with other device.
Rewrite the driver model code for normal-speed operation so that it is
easier to understand, and fix the bug. Add a TODO to split the drivers.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Heiko Schocher <hs@denx.de>
Simon Glass [Mon, 3 Aug 2015 14:19:22 +0000 (08:19 -0600)]
i2c: Add a mux for GPIO-based I2C bus arbitration
While I2C supports multi-master buses this is difficult to get right.
The implementation on the master side in software is quite complex.
Clock-stretching and the arbitrary time that an I2C transaction can take
make it difficult to share the bus fairly in the face of high traffic.
When one or more masters can be reset independently part-way through a
transaction it is hard to know the state of the bus.
This driver provides a scheme based on two 'claim' GPIOs, one driven by the
AP (Application Processor, meaning the main CPU) and one driven by the EC
(Embedded Controller, a small CPU aimed at handling system tasks). With
these they can communicate and reliably share the bus. This scheme has
minimal overhead and involves very little code. It is used on snow to
permit the EC and the AP to share access to the main system PMIC and
battery. The scheme can survive reboots by either side without difficulty.
This scheme has been tested in the field with millions of devices.
Since U-Boot runs on the AP, the terminology used is 'our' claim GPIO,
meaning the AP's, and 'their' claim GPIO, meaning the EC's. This terminology
is used by the device tree bindings in Linux also.
Simon Glass [Mon, 3 Aug 2015 14:19:21 +0000 (08:19 -0600)]
dm: i2c: Add support for multiplexed I2C buses
Add a new I2C_MUX uclass. Devices in this class can multiplex between
several I2C buses, selecting them one at a time for use by the system.
The multiplexing mechanism is left to the driver to decide - it may be
controlled by GPIOs, for example.
The uclass supports only two methods: select() and deselect().
The current mux state is expected to be stored in the mux itself since
it is the only thing that knows how to make things work. The mux can
record the current state and then avoid switching unless it is necessary.
So select() can be skipped if the mux is already in the correct state.
Also deselect() can be made a nop if required.
Simon Glass [Fri, 3 Jul 2015 00:15:42 +0000 (18:15 -0600)]
dm: i2c: Add a function to transfer messages
Sometimes it is useful to be able to transfer a raw I2C message. This
happens when the chip address needs to be set manually, or when the data to
be sent/received is in another buffer.
Add a function to provide access to this.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Heiko Schocher <hs@denx.de>
Simon Glass [Fri, 3 Jul 2015 00:15:38 +0000 (18:15 -0600)]
dm: core: Support finding a device by phandle
It is common for one node to reference another via a phandle. Add support
for obtaining an attached device by this method. As an example, a node may
have a 'power-supply' property which references a regulator, allowing the
driver to turn on its power.
Marcel Ziswiler [Wed, 5 Aug 2015 14:58:17 +0000 (16:58 +0200)]
dm: usb: fix USB Ethernet without CONFIG_DM_ETH regression
The following commit enforces CONFIG_DM_ETH for USB Ethernet which
breaks any board using CONFIG_USB_HOST_ETHER without CONFIG_DM_ETH
which this patch fixes.
Tested on Colibri T20/T30 as well as Apalis T30 with
CONFIG_USB_HOST_ETHER and CONFIG_USB_ETHER_ASIX enabled and a LevelOne
USB-0301 ASIX AX88772 dongle.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Acked-by: Simon Glass <sjg@chromium.org>
Marek Vasut [Wed, 5 Aug 2015 01:19:22 +0000 (03:19 +0200)]
usb: Fix device detection code
The code in question polls an USB port status via USB_REQ_GET_STATUS
to determine whether there is a device on the port or not. The way to
figure that out is to check two bits. Those are wPortChange[0] and
wPortStatus[0].
The wPortChange[0] indicates whether some kind of a connection status
change happened on a port (a device was plugged or unplugged). The
wPortStatus[0] bit indicates the status of the connection (plugged or
unplugged).
The current code tests whether wPortChange[0] == wPortStatus[0] and
if that's the case, considers the loop polling for the presence of a
USB device on port finished.
This works for most USB sticks, since they come up really quickly and
trigger the USB port change detection before the first iteration of the
detection loop happens. Thus, both wPortChange[0] and wPortStatus[0]
are set to 1 and thus equal. The loop is existed in it's first iteration
and the stick is detected correctly.
The problem is with some obscure USB sticks, which take some time before
they pop up on the bus after the port was enabled. In this case, both
the wPortChange[0] and wPortStatus[0] are 0. They are equal again, so
the loop again exits in the first iteration, but this is incorrect, as
such USB stick didn't have the opportunity to get detected on the bus.
Rework the code such, that it checks for wPortChange[0] first to test
if any connection change happened at all. If no change occured, keep
polling. If a change did occur, test the wPortStatus[0] to see there is
some device present on the port and only if this is the case, break out
of the polling loop.
This patch also trims down the duration of the polling loop from 10s
per port to 1s per port. This is still annoyingly long, but there is
no better option in case of U-Boot unfortunatelly. This change will
most likely increase the duration of 'usb start' on some platforms,
but this is needed to fix a bug.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Simon Glass <sjg@chromium.org> Cc: Hans de Goede <hdegoede@redhat.com>
Marcel Ziswiler [Wed, 5 Aug 2015 15:16:59 +0000 (17:16 +0200)]
net: asix: fix operation without eeprom
This patch fixes operation of our on-board AX88772B chip without EEPROM
but with a ethaddr coming from the regular U-Boot environment. This is
a forward port of some remaining parts initially implemented by
Antmicro.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com> Acked-by: Marek Vasut <marex@denx.de>
sunxi: usb: USB_MUSB_SUNXI move to musb-new Kconfig
Now that the musb-new driver has a Kconfig, we can move Kconfig options to
enable controllers to it, so that it's easier in e.g. menuconfig.
In addition, this allows declaring support for USB_MUSB_HOST/GADGET in
defconfigs instead of the USB_MUSB_SUNXI controller, that will get selected
automatically when needed.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
usb: musb-new: Kconfig support for USB_MUSB_HOST and USB_MUSB_GADGET
Having MUSB_HOST and MUSB_GADGET in Kconfig allows more flexibility with regard
to what Kconfig options to enable, such as USB_STORAGE or USB_KEYBOARD.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
usb: musb-new: CONFIG_MUSB prefix replacement with CONFIG_USB_MUSB
USB-related options are usually prefixed with CONFIG_USB and platform-specific
adaptation for the MUSB controller already have a CONFIG_USB_MUSB prefix, so
this switches all MUSB-related options to a CONFIG_USB_MUSB prefix, for
consistency.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
usb: Generic USB Kconfig option, that fits both host and gadget and comments
There is no particular reason why the USB Kconfig option should be specific to
host mode. In prevision of adding MUSB host and gadget to Kconfig, this moves
the title and help message of the USB Kconfig option to a more generic format.
Adding comments to the usb Kconfig allows for a better separation and more
readability in generated configs and in menuconfig.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Simon Glass [Tue, 4 Aug 2015 18:34:04 +0000 (12:34 -0600)]
x86: Gracefully disable the vesa driver when running from EFI
We cannot use this driver when running from EFI as we have no direct hardware
access. In fact coreboot uses a different driver which uses tables provided
by coreboot. So far it does not seem possible to use a normal video driver
when booting from EFI.
Signed-off-by: Simon Glass <sjg@chromium.org> Acked-by: Anatolij Gustschin <agust@denx.de> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Tue, 4 Aug 2015 18:34:02 +0000 (12:34 -0600)]
x86: baytrail: Support operation as an EFI payload
Disable a few things which interfere with the EFI init. This allows the
Minnowboard MAX to boot into EFI, load a U-Boot payload then boot to the
U-Boot prompt.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Tue, 4 Aug 2015 18:34:01 +0000 (12:34 -0600)]
x86: Add helper code for running from EFI
When U-Boot is running from EFI some of the x86 init is replaced with
EFI-specific init. For example, since DRAM has already been set up, we only
need to find it, not init it. Add these functions so that boards can easily
allow booting from EFI if required.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Simon Glass [Tue, 4 Aug 2015 18:34:00 +0000 (12:34 -0600)]
x86: Handle running as EFI payload
When U-Boot runs as an EFI payload it needs to avoid setting up the CPU
again. Also U-Boot currently does not handle interrupts for many devices, so
run with interrupts disabled.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
projects / u-boot / log