PMUFW requires top 1MB of the lower DDR memory reserved for its operation,
this is missing in k26/k24 sm static dts files because of which U-Boot
throws warning messages "efi_free_pool: illegal free" as EFI puts
some code to that space which shouldn't happen.
Signed-off-by: Sharath Kumar Dasari <sharath.kumar.dasari@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@amd.com>
Link: https://lore.kernel.org/r/57452d16df2d48593de206bebf877d2c2cfe7bf1.1685966389.git.michal.simek@amd.com
The H616 SoC has support for several types of DRAM: DDR3, LPDDR3,
DDR4 and LPDDR4.
At the moment, the driver only supports DDR3 memory.
Let's extend the driver to support the LPDDR3 memory. All "magic"
values obtained from the boot0.
Signed-off-by: Mikhail Kalashnikov <iuncuim@gmail.com>
Acked-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Allwinner H616 SoC supports several types of DRAM memory. To further
integrate other types of memory, we need to add this delimitation.
Signed-off-by: Mikhail Kalashnikov <iuncuim@gmail.com>
Reviewed-by: Andre Przywara <andre.przywara@arm.com>
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Currently there is one DRAM parameter struct for the Allwinner H616 DRAM
"driver". It contains many fields that are compile time constants
(set by Kconfig variables), though there are also some fields that are
probed and changed over the runtime of the DRAM initialisation.
Because of this mixture, the compiler cannot properly optimise the code
for size, as it does not consider constant propagation in its full
potential.
Help the compiler out by splitting that structure into two: one that only
contains values known at compile time, and another one where the values
will actually change. The former can then be declared "const", which will
let the compiler fold its values directly into the code using it.
We also add "const" tags for some new "struct dram_config" pointers, to
further increase code optimisation.
To help the compiler optimise the code further, the definition of the
now "const struct dram_para" has to happen at a file-global level, so
move that part out of sunxi_dram_init().
That results in quite some code savings (almost 2KB), and helps to keep
the code small with the LPDDR3 support added later.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
There are quite some functions in the Allwinner H616 DRAM "driver", some
of them actually change the parameters in the structure passed to them,
but many are actually not.
To increase the optimisation potential for the code, mark those functions
that just read members of the passed dram_para struct as "const".
This in itself does not decrease the code size, but lays the groundwork
for future changes doing so.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
The usage of the C keyword "inline" seems to be a common
misunderstanding: it's a *hint* only, and modern compilers will inline
(or not) functions based on their own judgement and provided compiler
options.
So while marking functions as "inline" does not do much, missing the
"static" keyword will force to compiler to spell out a version of the
function for potential external callers, which actually increases the
code size (though hopefully the linker will drop the function).
Change the "inline" attribute for the mbus_configure_port() functions in
some Allwinner DRAM drivers to "static", so that the explicit version
can actually be dropped from the object file, reducing the code size.
"static inline" has a use case in header files, where it avoids a warning
if a .c file including this header does not use the particular function.
In a .c file itself "static inline" is not useful otherwise, so just use
static here as well.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Now that the PHY driver supports the H616 USB PHY, we can enable USB
support for the two H616 boards.
As the OrangePi Zero2 has a USB-C port hard-wired to peripheral mode,
let's enable USB gadget mode for port 0, so people can use fastboot,
ethernet or mass storage functionality.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Now that the Allwinner USB PHY driver supports the H616 quirk, let's
enable support for USB ports on that SoC.
We connect the compatible string to a new struct describing the SoCs USB
PHY properties, and unblock the PHY driver selection in Kconfig.
A later patch will enable USB support in the H616 boards' defconfigs.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
The H616 USB PHY is some kind of special snowflake: Only port2 works out
of the box, but all other ports need some help from this port2 to work
correctly: The CLK_BUS_PHY2 and RST_USB_PHY2 clock and reset need to be
enabled, and the SIDDQ bit in the PMU PHY control register needs to be
cleared. For this register to be accessible, CLK_BUS_ECHI2 needs to be
ungated. Don't ask ....
Follow the respective Linux patch (b45c6d80325b) and add a quirk bit,
triggering the special sequence as outlined above, for PHYs other than
PHY2: ungate this one special clock, and clear the SIDDQ bit. We also
pick the clock and reset from PHY2 and enable them as well.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
So far we were assigning some crude "type" (SoC name, really) to each
Allwinner USB PHY model, then guarding certain quirks based on this.
This does not only look weird, but gets more or more cumbersome to
maintain.
Remove the bogus type names altogether, instead introduce flags for each
quirk, and explicitly check for them.
This improves readability, and simplifies future extensions.
Port of Linux patch 8dd256bae653.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
At the moment we use "select" in each Allwinner SoC's Kconfig section to
include the USB PHY driver in the build. This means it cannot be disabled
via Kconfig, although USB is not really a strictly required core
functionality, and a particular board might not even include USB ports.
Rework the Kconfig part by removing the "select" lines for each SoC's
section, and instead letting it default to "y" in the PHY driver section
itself. We use "depends on !" to exclude the few SoCs we don't support
(yet). The Allwinner V3s does not enable USB (PHY) support at the moment,
even though it should work: let the PHY default to "n" to keep the
current behaviour.
Also the MUSB USB driver directly calls some functions from the PHY
driver, so let the former depend on the PHY driver.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Tested-by: Sam Edwards <CFSworks@gmail.com>
The Allwinner F1C100s implements a single USB PHY, connected to its MUSB
OTG controller. The USB PHY is of the simpler, older type (like the A10),
the only real difference is that it's indeed only one PHY.
Add a struct describing those F1C100s USB PHY properties, and connect it
to the new compatible string.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
In its of_xlate() function, the Allwinner USB PHY driver compares the
args_count variable against the number of implemented USB PHYs, although
this is the *number of arguments* to the DT phandle property. Per the DT
binding for this PHY device, this number is always one, so this check
will always fail if the particular SoC implements exactly one USB PHY.
So far this affected only the V3s (which has USB support disabled), but
the F1C100s also sports one PHY only.
Fix that check to compare args_count against exactly 1, and the args[0]
content (requested PHY number) against the number of implemented PHYs.
This fixes USB operation on the Allwinner V3s and allows to enable USB
on the Allwinner F1C100s SoC.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Reviewed-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Add support for writing symbols and determining the assumed position of
binaries inside a mkimage image. This is useful as an example for other
entry types which might want to do the same thing.
Signed-off-by: Simon Glass <sjg@chromium.org>
Collections can used to collect the contents of other entries into a
single entry, but they result in a single entry, with the original entries
'left behind' in their old place.
It is useful to be able to specific a set of entries ones and have it used
in multiple images, or parts of an image.
Implement this mechanism.
Signed-off-by: Simon Glass <sjg@chromium.org>
This permits implementation of a simple templating system, where a node
can be reused as a base for others.
For now this adds new subnodes after any existing ones.
Signed-off-by: Simon Glass <sjg@chromium.org>
Fix the check for the __bss_size symbol, since it may be 0. Unfortunately
there was no test coverage for this.
Signed-off-by: Simon Glass <sjg@chromium.org>
This is not needed since the linker script sets it up. Drop the variable
to avoid confusion.
Fix the prototype for main() while we are here.
Signed-off-by: Simon Glass <sjg@chromium.org>
Sometimes multiple boards are built with binman and it is useful to
specify a different FDT list for each. At present this is not possible
without providing multiple values of the of-list entryarg (which is not
supported in the U-Boot build system).
Allow a fit,fdt-list-val string-list property to be used instead.
Signed-off-by: Simon Glass <sjg@chromium.org>
This is needed to handle mkimage with inner section located itself in a
section.
Signed-off-by: Marek Vasut <marex@denx.de>
Use BuildSectionData() instead of ObtainContents(), add tests and a few
other minor fixes:
Signed-off-by: Simon Glass <sjg@chromium.org>
Update the LookupAndWriteSymbols() function to return the number of
symbols written. Also add some logging for when debugging is not
enabled.
Signed-off-by: Simon Glass <sjg@chromium.org>
These boards use SPL in a mkimage entry and apparently access the symbol
containing the image position of U-Boot, but put U-Boot in another
image. This means that binman is unable to fill in the symbol correctly
in the SPL binary.
This doesn't matter at present since mkimage doesn't support symbol
writing. But with the upcoming conversion to a section, it will. So add
a property to disable symbol writing.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some boards don't use symbol writing but do access the symbols in SPL.
Provide an option to work around this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Some section types don't have a simple _entries list. Use the GetEntries()
method in GetEntryContents() and other places to handle this.
This makes the behaviour more consistent.
Signed-off-by: Simon Glass <sjg@chromium.org>
This should be set up in the init function, to avoid a warning about a
property not set up there. Fix it.
Signed-off-by: Simon Glass <sjg@chromium.org>
When building from source, setuptools would complain about not finding
package via its auto-discovery mechanism. Manually specify how to
locate the files, relative to the package's directory.
* tools/u_boot_pylib/pyproject.toml: New tool.setuptools.packages.find
section.
Signed-off-by: Maxim Cournoyer <maxim.cournoyer@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Change the device type from "usb_mass_storage" to "usb" when
booting a script.
Before this change:
=> printenv devtype
devtype=usb_mass_storage
After this change:
=> printenv devtype
devtype=usb
Signed-off-by: John Clark <inindev@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The loads and saves commands crash on the sandbox due to illegal memory
access.
For command line arguments the sandbox uses a virtual address space which
does not equal the addresses of the memory allocated with memmap(). Add the
missing address translations for the loads and saves commands.
Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
With sandbox and sandbox64 configurations:
In file included from .../u-boot/include/test/test.h:156,
from .../u-boot/include/test/lib.h:9,
from .../u-boot/test/lib/test_crc8.c:8:
.../u-boot/arch/sandbox/include/asm/test.h: In function ‘sandbox_sdl_set_bpp’:
.../u-boot/arch/sandbox/include/asm/test.h:323:17: error: ‘ENOSYS’ undeclared (first use in this function)
323 | return -ENOSYS;
| ^~~~~~
Per Tom Rini's suggestion:
move that function prototype over to arch/sandbox/include/asm/sdl.h
and make test/dm/video.c include <asm/sdl.h>
Cc: Simon Glass <sjg@chromium.org>
Suggested-by: Tom Rini <trini@konsulko.com>
Signed-off-by: Sergei Antonov <saproj@gmail.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The behaviour of dev_read_addr_size() is surprising as it does not
handle #address-cells and #size-cells but instead hardcodes the values
based on sizeof(fdt_addr_t).
This is different from dev_read_addr_size_index() and
dev_read_addr_size_name() both of which do read the cell sizes from the
device tree.
Since dev_read_addr_size() is only used by a single driver and this
driver is broken when CONFIG_FDT_64BIT does not match the address size
in the device tree, fix the function to behave like all of the other
similarly named functions. Drop the property name argument as the only
caller passes "reg" and this is the expected property name matching the
other similarly named functions.
Signed-off-by: John Keeping <john@metanate.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org> # chromebook_jerry
Tested-by: Simon Glass <sjg@chromium.org> # chromebook_bob
of_property_match_string calls of_find_property to search for the
string property.
If the device node does not exist, of_find_property returns NULL, and
of_property_match_string returns -EINVAL, which is correct.
However, if the device node exists, but the property is not found,
of_find_property still returns NULL, but it will place -FDT_ERR_NOTFOUND
in the *lenp variable.
of_property_match_string does not use the lenp parameter, thus this error
case is being lost, and treated as if the node is NULL, and returns
-EINVAL, which is incorrect.
The callers of of_property_match_string treat the error differently if
the return value is -EINVAL or -ENOENT, e.g. in dwc3 driver:
ret = generic_phy_get_by_name(dev, "usb3-phy", &phy);
if (!ret) {
ret = generic_phy_init(&phy);
if (ret)
return ret;
} else if (ret != -ENOENT && ret != -ENODATA) {
debug("could not get phy (err %d)\n", ret);
return ret;
} else {
phy.dev = NULL;
}
So the caller drivers will just consider the property missing if -ENOENT
is returned, versus the case of -EINVAL, which means something else.
To fix this situation, changed the code to call the of_find_property
with the right third argument to catch this error code and treat it
accordingly.
Signed-off-by: Eugen Hristev <eugen.hristev@collabora.com>
It is not uncommon for some of the QEMU-based jobs to fail not because
of a code issue but rather because of a timing issue or similar problem
that is out of our control. Make use of the keywords that Azure and
GitLab provide so that we will automatically re-run these when they fail
2 times. If they fail that often it is likely we have found a real issue
to investigate.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
The job to build all imx8 and imx9 platforms is currently close to, or
sometimes exceeding the allowed build time. Exclude some platforms that
are already being built under their vendor-specific job as well to
reduce the time.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Currently the 64bit "rk" job is close to and sometimes goes over the job
time limit. Let us rework this in to one job for "rk" and "rv" (which
are the SoC prefixes) jobs which include or exclude "rockchip" the board
vendor. This gives us two jobs of similar numbers of platforms to build
now instead.
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
As this is the current version of the public cross toolchains we use,
upgrade to this now.
Suggested-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Alexey Brodkin <abrodkin@synopsys.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Back in the day we relied a lot on Synopsys own build of the GNU tools
for ARC processors, but since then we worked hard on getting all our changes
upstream and for a couple of years now we have ARCompact (AKA ARCv1)
and ARCv2 processors supported very well in upstream GCC, Binutils, GDB etc.
And so there's no need to use Synopsys forks any longer, thus we remove
all the references to that form and use upstream components as majority
of other architectures in U-Boot.
Thanks to Tom for pointing to that left-over!
Signed-off-by: Alexey Brodkin <abrodkin@synopsys.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Tom Rini <trini@konsulko.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
When building qemu, all required submodules (of which we need more than
just dtc) are handled automatically. Currently trying to init the
submodule the way we do results in a git failure.
Reported-by: Alexey Brodkin <Alexey.Brodkin@synopsys.com>
Signed-off-by: Tom Rini <trini@konsulko.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
