We can handle the sysc interconnect target module in a generic way for
many TI SoCs. Initially let's just enable domain clocks before the
children are probed.
The code is loosely based on the drivers/bus/ti-sysc.c of the Linux
kernel version 5.9-rc7.
For DT binding details see:
- Documentation/devicetree/bindings/bus/ti-sysc.txt
Dario Binacchi [Tue, 29 Dec 2020 23:06:28 +0000 (00:06 +0100)]
dt-bindings: bus: ti-sysc: resync with Linux 5.9-rc7
Add support for PRUSS SYSC type:
The PRUSS module has a SYSCFG which is unique. The SYSCFG has two
additional unique fields called STANDBY_INIT and SUB_MWAIT in addition
to regular IDLE_MODE and STANDBY_MODE fields. Add the bindings for this
new sysc type.
Add support for MCAN on dra76x:
The dra76x MCAN generic interconnect module has a its own format for the
bits in the control registers.
Felix Brack [Fri, 18 Dec 2020 08:03:50 +0000 (09:03 +0100)]
arm:pdu001: Use pseudo partition UUID for LINUX kernel boot paramter root
As more and more LINUX drivers are modified to use asynchronous probing
instead of synchronous probing, relying on device names being equal in
U-Boot and LINUX is not possible anymore. This is also true for block
device names like mmc0, mmc1 ect.
With LINUX kernel commit a1a4891 the probing type for the sdhci-omap
driver has been set to asynchronous mode too (probe_type is now
PROBE_PREFER_ASYNCHRONOUS).
In the case of the PDU001 board this results in the devices mmc0 and
mmc1 being swapped between U-Boot and LINUX. Device mmc0 in U-Boot
becomes mmc1 in LINUX an vice versa. Hence using device name identifiers
with LINUX kernel parameter root does not work anymore.
This patch changes the LINUX kernel boot parameter root to use the
pseudo (since we use MBR not GPT) partition UUID to locate the partition
hosting the root file system.
McSPI IP provides per CS specific speed and mode selection. Therefore it
is possible to apply these settings only after CS is known. But
set_speed and set_mode can be called without bus being claimed, this
would lead driver to set up wrong CS (or previously used CS).
Fix this by apply set_speed and set_mode only if bus is already claimed.
Ovidiu Panait [Sat, 28 Nov 2020 08:11:28 +0000 (10:11 +0200)]
spi: ti_qspi: Fix "spi-max-frequency" error path in ti_qspi_ofdata_to_platdata
struct ti_qspi_priv->max_hz is declared as unsigned int, so the following
error path check will always be false, even when "spi-max-frequency"
property is invalid/missing:
priv->max_hz = fdtdec_get_int(blob, node, "spi-max-frequency", -1);
if (priv->max_hz < 0) {
...
}
Replace the fdtdec call with dev_read_u32_default() and use 0 as the
default value. Error out if max_hz is zero.
Tomas Novotny [Wed, 25 Nov 2020 17:42:16 +0000 (18:42 +0100)]
gpio: tca642x: fix input subcommand for gpio banks > 0
The value of input pin for bank > 0 is always 0 for input subcommand.
The reason is that gpio_bank variable is computed only for invert and
output subcommands (it depends on number of arguments). The default
value of zero causes to shift the mask away for banks > 0.
Please note that info subcommand works as expected, because the input
pin values are accessed differently.
Fixes: 61c1775f16ed ("gpio: tca642x: Add the tca642x gpio expander driver") Cc: Dan Murphy <dmurphy@ti.com> Signed-off-by: Tomas Novotny <tomas@novotny.cz>
configs: am65x_evm: Define the maximum file size for DFU
In include/dfu.h, if CONFIG_SYS_DFU_MAX_FILE_SIZE is not defined then it is
defined as CONFIG_SYS_DFU_DATA_BUF_SIZE. This is 128 KiB for a53 core and
20 KiB for r5 core. If a larger file is transferred using dfu then it
fails.
CONFIG_SYS_DFU_DATA_BUF_SIZE can not be increased as there is not enough
heap memory to be allocated for the buffer in case of R5 spl.
Fix this by defining CONFIG_SYS_DFU_MAX_FILE_SIZE as the default
CONFIG_SYS_DFU_DATA_BUF_SIZE value.
arm: dts: am654-base-board-uboot: Add aliases for USB subsystems
The sequence number assigned for USB subsystem in a uclass is dependent on
the order of occurrence in the device tree. If the dr_mode of USB3SS0
controller is varied then the sequence number of USB3SS1 controller also
changes. If aliases are added then sequence numbers are assigned using the
alias number. This makes the sequence number of USB3SS1 controller
independent of USB3SS0 controller's dr_mode.
Therefore, add aliases to fix the sequence number assigned to the USB
subsystems.
board: ti: am65x: Set SERDES0 mux to PCIe to use USB 2.0 interface
It has been observed that setting SERDES0 lane mux to USB prevents USB 2.0
operation on USB0. Setting SERDES0 lane mux to non-USB when USB0 is used in
USB 2.0 only mode solves this issue. However, for USB3.0+2.0 operation this
issue is not present.
Implement this workaround by writing 1 to LANE_FUNC_SEL field in
CTRLMMR_SERDES0_CTRL register.
Lukasz Majewski [Sat, 26 Dec 2020 00:09:02 +0000 (01:09 +0100)]
xea: config: Disable CONFIG_SPL_OF_PLATDATA_PARENT on XEA (imx28)
On the XEA board (imx28) one needs in the SPL support for GPIO, MMC and
SPI. Two last ones are necessary for booting the device. The GPIO support
allows deciding which medium will be used. For example the GPIO DTS node
(gpio@0 at imx28.dtsi) has pinctrl parent (pinctrl@80018000) for which we
don't need driver asigned for correct operation.
In the spl/dts/dt-platdata.c the gpio@0 has index 4 and its parent -
pinctrl@80018000 has index 5.
In the bind_drivers_pass() function (at drivers/core/lists.c) call to
device_bind_by_name() for `fsl_imx23_pinctrl` returns -2, which is
expected.
With current setup - when the SPL_OF_PLATDATA_PARENT=y
The gpio@0 node with index 4 is skipped as its parent with 5 is not yet
bound. It cannot be as we don't need and provide the driver for it.
As a result the gpio@0 is never bound and we end up with bricked board in
the SPL stage.
When CONFIG_SPL_OF_PLATDATA_PARENT is NOT set, all entries from
spl/dts/dt-platdata.c are scanned in ascending index order, so gpio@0 is
properly initialized. For `fsl_imx_pinctrl` we simply check 10 times if
the driver for is available (which is not) and exit.
As a result the GPIOs are initialized and can be used in early SPL stage.
This commit fixes XEA regression introduced with e41651fffda7da55f6.
Signed-off-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
Lukasz Majewski [Sat, 26 Dec 2020 00:09:01 +0000 (01:09 +0100)]
xea: config: Use CONFIG_PREBOOT from Kconfig
The usage of the preboot feature is now controlled via a separate Kconfig
option - namely CONFIG_USE_PREBOOT.
It must be enabled for preboot code executing commands now defined in
CONFIG_PREBOOT (also moved to the Kconfig).
After defining both CONFIG_USE_PREBOOT and CONFIG_PREBOOT in
imx28_xea_defconfig the define of CONFIG_PREBOOT shall be removed from
xea.h as it is redundant.
Hugh Cole-Baker [Sun, 22 Nov 2020 13:03:44 +0000 (13:03 +0000)]
rockchip: pinebook-pro: default to SPI bus 1 for SPI-flash
SPI flash on this machine is located on bus 1, default to using bus 1
for SPI flash and stop aliasing it to bus 0.
Signed-off-by: Hugh Cole-Baker <sigmaris@gmail.com> Suggested-by: Simon Glass <sjg@chromium.org> Fixes: c4cea2bb ("rockchip: Enable building a SPI ROM image on bob")
Marek Vasut [Sat, 2 Jan 2021 16:44:47 +0000 (17:44 +0100)]
ARM: dts: stm32: Fix GPIO hog flags on DHCOM DRC02
The GPIO hog flags are ignored by gpiolib-of.c now, set the flags to 0.
Since GPIO_ACTIVE_HIGH is defined as 0, this change only increases the
correctness of the DT.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@st.com> Cc: Patrick Delaunay <patrick.delaunay@st.com> Reviewed-by: Patrice Chotard <patrice.chotard@st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Marek Vasut [Sat, 2 Jan 2021 16:44:46 +0000 (17:44 +0100)]
ARM: dts: stm32: Fix GPIO hog flags on DHCOM PicoITX
The GPIO hog flags are ignored by gpiolib-of.c now, set the flags to 0.
Due to a change in gpiolib-of.c, setting flags to GPIO_ACTIVE_LOW and
using output-low DT property leads to the GPIO being set high instead.
Signed-off-by: Marek Vasut <marex@denx.de> Cc: Patrice Chotard <patrice.chotard@st.com> Cc: Patrick Delaunay <patrick.delaunay@st.com> Reviewed-by: Patrice Chotard <patrice.chotard@st.com> Reviewed-by: Patrick Delaunay <patrick.delaunay@foss.st.com>
Tom Rini [Wed, 6 Jan 2021 03:34:43 +0000 (22:34 -0500)]
Merge tag 'dm-pull-5jan21' of git://git.denx.de/u-boot-dm into next
Driver model: make some udevice fields private
Driver model: Rename U_BOOT_DEVICE et al.
dtoc: Tidy up and add more tests
ns16550 code clean-up
x86 and sandbox minor fixes for of-platdata
dtoc prepration for adding build-time instantiation
Simon Glass [Tue, 29 Dec 2020 03:35:06 +0000 (20:35 -0700)]
dtoc: Split source-code scanning to a separate file
Before expanding the scanning features any more, move this into a separate
file. This will make it easier to maintain in the future. In particular,
it reduces the size of dtb_platdata.py and allows us to add tests
specifically for scanning, without going through that file.
The pieces moved are the Driver class, the scanning code and the various
naming functions, since they mostly depend on the scanning results.
So far there is are no separate tests for src_scan. These will be added
as new functionality appears.
Simon Glass [Tue, 29 Dec 2020 03:35:05 +0000 (20:35 -0700)]
dtoc: Drop dm_populate_phandle_data()
This has not been needed since parent information was added and we started
using indicies for references to other drivers instead of pointers. It was
kept around in the expectation that it might be needed later.
However with the latest updates, it doesn't seem likely that we'll need
this in the foreseeable future.
Drop dm_populate_phandle_data() from dtoc and driver model.
Simon Glass [Tue, 29 Dec 2020 03:35:04 +0000 (20:35 -0700)]
dtoc: Output nodes in order
Previously we had to worry about nodes being output before those that they
depended on, thus causing build errors. So the current algorithm is
careful to output nodes in the right order.
We now use a different method for outputting phandles that does not
involve pointers. Also we plan to add a 'declarations' header file to
declare all drivers as 'extern'.
Update the code to drop the dependency checking and output in a simple
loop. This makes the output easier to follow since drivers are in order of
thier indices (0, 1, ...), which is also the order it appears in in the
linker list.
Simon Glass [Tue, 29 Dec 2020 03:35:03 +0000 (20:35 -0700)]
dtoc: Allow specifying the base directory for tests
The base directory of U-Boot, where the source is, it currently calculated
from the directory of the dtb_platdata.py script. If this is installed
elsewhere that will not work. Also it is inconvenient for tests.
Add a parameter to allow specifying this base directory.
To test this, pass a temporary directory with some files in it and check
that they are passed to scan_driver().
Simon Glass [Tue, 29 Dec 2020 03:35:02 +0000 (20:35 -0700)]
dtoc: Add the method for each command to OutputFile
Rather than the if/else construct, update OutputFile with the method to
call to process each command. This is easier to maintain as the number of
commands increases.
Rename generate_tables to generate_plat since it better describes what is
being generated ('plat' is the U-Boot name for platform data).
With this, each output method needs to have the same signature. Store the
output structures in a member variable instead of using parameters, to
accomplish this.
Simon Glass [Tue, 29 Dec 2020 03:35:00 +0000 (20:35 -0700)]
dtoc: Add a header comment to each generated file
It is currently fairly obvious what the two generated files are for, but
this will change as more are added. It is helpful for readers to describe
the purpose of each file.
Add a header commment field to OutputFile and use it to generate a comment
at the top of each file.
Simon Glass [Tue, 29 Dec 2020 03:34:59 +0000 (20:34 -0700)]
dtoc: Run tests using test_util
Use the standard function for running tests and reported results. This
allows the tests to run in parallel, which is a significant speed-up on
most machines (e.g. 4.5 seconds -> 1.5s on mine).
Simon Glass [Tue, 29 Dec 2020 03:34:58 +0000 (20:34 -0700)]
concurrencytest: Fix Python3 warning
This gives a warning in some situations:
File "tools/dtoc/../concurrencytest/concurrencytest.py", line 95,
in do_fork
stream = os.fdopen(c2pread, 'rb', 1)
File "/usr/lib/python3.8/os.py", line 1023, in fdopen
return io.open(fd, *args, **kwargs)
RuntimeWarning: line buffering (buffering=1) isn't supported in binary
mode, the default buffer size will be used
Simon Glass [Tue, 29 Dec 2020 03:34:57 +0000 (20:34 -0700)]
dm: Rename U_BOOT_DRIVER_ALIAS to DM_DRIVER_ALIAS
We use the U_BOOT_ prefix (i.e. U_BOOT_DRIVER) to declare a driver but
in every other case we just use DM_. Update the alias macros to use the
DM_ prefix.
We could perhaps rename U_BOOT_DRIVER() to DM_DRIVER(), but this macro
is widely used and there is at least some benefit to indicating it us a
U-Boot driver, particularly for code ported from Linux. So for now, let's
keep that name.
Simon Glass [Tue, 29 Dec 2020 03:34:56 +0000 (20:34 -0700)]
dm: Rename DM_GET_DRIVER() to DM_DRIVER_GET()
In the spirit of using the same base name for all of these related macros,
rename this to have the operation at the end. This is not widely used so
the impact is fairly small.
Simon Glass [Tue, 29 Dec 2020 03:34:55 +0000 (20:34 -0700)]
dm: Rename DM_GET_DEVICE() to DM_DRVINFO_GET()
This does not get a device (struct udevice *) but a struct driver_info *
so the name is confusing.
Rename it accordingly. Since we plan to have several various of these
macros, put GET at the end instead of the middle, so it is easier to spot
the related macros.
Simon Glass [Tue, 29 Dec 2020 03:34:54 +0000 (20:34 -0700)]
dm: Rename U_BOOT_DEVICE() to U_BOOT_DRVINFO()
The current macro is a misnomer since it does not declare a device
directly. Instead, it declares driver_info record which U-Boot uses at
runtime to create a device.
The distinction seems somewhat minor most of the time, but is becomes
quite confusing when we actually want to declare a device, with
of-platdata. We are left trying to distinguish between a device which
isn't actually device, and a device that is (perhaps an 'instance'?)
It seems better to rename this macro to describe what it actually is. The
macros is not widely used, since boards should use devicetree to declare
devices.
Rename it to U_BOOT_DRVINFO(), which indicates clearly that this is
declaring a new driver_info record, not a device.
Simon Glass [Tue, 29 Dec 2020 03:34:53 +0000 (20:34 -0700)]
Makefile: Invoke dtoc only once
Update the Makefile to run dtoc only once, generating all required files.
This saves time since there is a lot of processing in each invocation of
dtoc.
We already have a variable for the object files to build, so use that
instead of repeating the same filenames. Add a C version of this also,
for the same reason.
This makes it easier to add new C files (generated by dtoc) to the build
later, as needed.
Simon Glass [Tue, 29 Dec 2020 03:34:51 +0000 (20:34 -0700)]
dtoc: Allow outputing to multiple files
Implement the 'output directory' feature, allowing dtoc to write the
output files separately to the supplied directories. This allows us to
handle the struct and platdata output in one run of dtoc.
Simon Glass [Tue, 29 Dec 2020 03:34:50 +0000 (20:34 -0700)]
dtoc: Allow providing a directory to write files to
At present dtoc writes only a single file on each invocation. U-Boot
writes the two files it needs by separate invocations of dtoc. Since dtoc
now scans all U-Boot driver source, this is fairly slow (about 1 second
per file).
It would be better if dtoc could write all the files at once.
In preparation for this, add a way to specify an output directory for the
files.
Simon Glass [Tue, 29 Dec 2020 03:34:49 +0000 (20:34 -0700)]
dtoc: Test the stdout output
Normally dtoc outputs to a file but it also offers a way to write output
to stdout. At present the test for that does not actually check that the
output is correct. Add this to the test.
This uses a member variable to hold the expected text, so it can be used
in muitiple places.
Simon Glass [Sat, 19 Dec 2020 17:40:17 +0000 (10:40 -0700)]
dm: core: Allow the uclass list to move
At present the uclass list head is in global_data. This is convenient
but with the new of-platdata we need the list head to be declared by
the generated code.
Change this over to be a pointer. Provide a 'static' version in
global_data to retain the current behaviour.
Simon Glass [Sat, 19 Dec 2020 17:40:16 +0000 (10:40 -0700)]
dm: core: Split out scanning code to dm_scan()
Move the code related to scanning for devices to bind, into a new
function. This will make it easier to skip this step with the new
of-platdata improvements.
Simon Glass [Sat, 19 Dec 2020 17:40:14 +0000 (10:40 -0700)]
dm: core: Access device ofnode through functions
At present ofnode is present in the device even if it is never used. With
of-platdata this field is not used, so can be removed. In preparation for
this, change the access to go through inline functions.
Simon Glass [Sat, 19 Dec 2020 17:40:13 +0000 (10:40 -0700)]
dm: core: Use dev_has_ofnode() instead of dev_of_valid()
We have two functions which do the same thing. Standardise on
dev_has_ofnode() since there is no such thing as an 'invalid' ofnode in
normal operation: it is either null or missing.
Also move the functions into one place.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Simon Glass [Sat, 19 Dec 2020 17:40:09 +0000 (10:40 -0700)]
dm: core: Rename sqq to seq_
Now that the sequence-numbering migration is complete, rename this member
back to seq_, adding an underscore to indicate it is internal to driver
model.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Simon Glass [Sat, 19 Dec 2020 17:40:08 +0000 (10:40 -0700)]
dm: core: Split out alloc code into a new function
Add a new function to handle the allocation of private/platform data for
a device. This will make it easier to skip this feature when using the new
of-platdata.
Simon Glass [Sat, 19 Dec 2020 17:40:05 +0000 (10:40 -0700)]
x86: apl: Adjust how the UART gets its platform data
At present this driver calls malloc() to start a new platform data
structure, fills it in and tells driver model to use it.
We want to avoid malloc, particularly with the new version of of-platdata.
Create a new struct which encompasses both the dtd struct and the ns16550
one, to avoid this. Unfortunately we must copy the data into the right
place for the ns16550 driver. Add some comments about this.
Simon Glass [Sat, 19 Dec 2020 17:40:02 +0000 (10:40 -0700)]
x86: apl: Use struct spi_nor instead of struct spi_flash
This construct effectively uses struct spi_nor due to a #define in
spi-nor.h so we may as well use that struct here. This allows dtoc to
parse it correctly.
Simon Glass [Sat, 19 Dec 2020 17:40:01 +0000 (10:40 -0700)]
spi: Tweak a few strange SPI NOR features for of-platdata
The #define of one struct to another has been around for a while. It
confuses dtoc and makes it think that struct spi_flash does not exist.
Make a few changes to improve things while we wait for migration to be
completed:
- Move the 'struct spi_flash' to column 1 so dtoc scans it
- Remove the #define when compiling dt-platdata.c
- Update the strange mtd_get/set_of_node() functions
- Use struct spi_nor in the drivers, so dtoc sees the correct struct
Simon Glass [Sat, 19 Dec 2020 17:40:00 +0000 (10:40 -0700)]
spi: Tidy up get/set of device node
This code is a bit odd in that it only reads and updates the livetree
version of the device ofnode. This means it won't work with flattree.
Update the code to work as it was presumably intended.
Simon Glass [Sat, 19 Dec 2020 17:39:56 +0000 (10:39 -0700)]
sandbox: Update dts files to reduce SPL size
At present there are require a few devices in the devicetree which are
not actually used in SPL. This will cause problems with the new
of-platdata, since it will try to instantiate devices which are not
compiled into U-Boot.
Update the devicetree to remove these devices from SPL.
Patrick Delaunay [Fri, 20 Nov 2020 08:48:33 +0000 (09:48 +0100)]
sandbox: remove ram buffer file when U-Boot is loaded by SPL
Update management of "--rm_memory" sandbox's option and force
this option when U-Boot is loaded by SPL in os_spl_to_uboot()
and remove the ram file after reading in main() as described
in option help message: "Remove memory file after reading".
This patch avoids that the file "/tmp/u-boot.mem.XXXXXX" [created in
os_jump_to_file() when U-Boot is loaded by SPL] is never deleted
because state_uninit() is not called after U-Boot execution
(CtrlC or with running pytest for example).
This issue is reproduced by
> build-sandbox_spl/spl/u-boot-spl
and CtrlC in U-Bot console
> make qcheck
One temp file is created after each SPL and U-Boot execution
(7 tims in qcheck after test_handoff.py, test_ofplatdata.py,
test_spl.py execution).
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass [Wed, 23 Dec 2020 15:11:23 +0000 (08:11 -0700)]
dtoc: Convert _drivers to a dict
At present this member holds a simple list of driver names. Update it to
be a dict of DriverInfo, with the name being the key. This will allow more
information to be added about each driver, in future patches.
Simon Glass [Wed, 23 Dec 2020 15:11:19 +0000 (08:11 -0700)]
dtoc: Fix a few pylint warnings in dtb_platdata
These have crept in again. Update the file to fix all but these ones:
dtb_platdata.py:143:0: R0902: Too many instance attributes (10/7)
(too-many-instance-attributes)
dtb_platdata.py:713:0: R0913: Too many arguments (6/5)
(too-many-arguments)
Simon Glass [Wed, 23 Dec 2020 15:11:18 +0000 (08:11 -0700)]
test: Move some test drivers into their own file
At present several test drivers are part of the test file itself. Some of
these are useful for of-platdata tests. Separate them out so we can use
them for other things also.
A few adjustments are needed so this driver can build for sandbox_spl as
well.
Simon Glass [Wed, 23 Dec 2020 15:11:17 +0000 (08:11 -0700)]
test: Use a simple variable to record removed device
At present the entire test state is effective passed into a test driver
just to record which device was removed. This is unnecessary and makes it
harder to track what is going on.
Use a simple boolean instead.
Also drop the unused 'removed' member while we are here.
At present it is not possible to include spl.h in on these architectures
since the asm/spl.h file is not present. We want to be able to use the
spl_phase() function, so add empty headers to make things build.
Simon Glass [Wed, 23 Dec 2020 02:30:29 +0000 (19:30 -0700)]
dm: core: Use access methods for dev/uclass private data
Use these functions in the core code as much as possible. With this, there
are only two places where each priv/plat pointer is accessed, one for read
and one for write.
projects / u-boot.git / log
