Alex Marginean [Fri, 7 Jun 2019 08:24:24 +0000 (11:24 +0300)]
test: dm: Add a test for PCI Enhanced Allocation
This test is built on top of the existing swap_case driver. It adds EA
capability structure support to swap_case and uses that to map BARs.
BAR1 works as it used to, swapping upper/lower case. BARs 2,4 map to a
couple of magic values.
Signed-off-by: Alex Marginean <alexm.osslist@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com> Tested-by: Bin Meng <bmeng.cn@gmail.com>
Alex Marginean [Fri, 7 Jun 2019 08:24:22 +0000 (11:24 +0300)]
pci: fixed dm_pci_map_bar comment
The comment now indicates that the input argument bar is a register offset,
not a BAR index.
It also mentions which BARs are supported for type 0/1 and that the
function can return 0 on error.
Signed-off-by: Alex Marginean <alexm.osslist@gmail.com> Reviewed-by: Bin Meng <bmeng.cn@gmail.com>
Up to now we had hard coded values for the call depth up to which trace
records are created: 200 for early tracing, 15 thereafter. UEFI
applications reach a call depth of 80 or above.
Provide customizing settings for the call trace depth limit and the early
call trace depth limit. Use the old values as defaults.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
An UEFI application may change the value of the register that gd lives in.
But some of our functions like get_ticks() access this register. So we
have to set the gd register to the U-Boot value when entering a trace
point and set it back to the application value when exiting the trace
point.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
Compiling with TRACE but without TRACE_EARLY results in an error
aarch64-linux-gnu-ld.bfd:
common/built-in.o:(.rodata.init_sequence_f+0x10):
undefined reference to `trace_early_init'
trace_early_init() should not be called if CONFIG_TRACE_EARLY is not
defined.
Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Reviewed-by: Simon Glass <sjg@chromium.org>
Patrick Delaunay [Tue, 21 May 2019 17:19:12 +0000 (19:19 +0200)]
fdt: Allow indicating a node is for U-Boot proper only
This add missing parts for previous commit 06f94461a9f4
("fdt: Allow indicating a node is for U-Boot proper only")
At present it is not possible to specify that a node should be used before
relocation (in U-Boot proper) without it also ending up in SPL and TPL
device trees. Add a new "u-boot,dm-pre-proper" boolean property for this.
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Patrick Delaunay [Tue, 21 May 2019 17:19:11 +0000 (19:19 +0200)]
test: check u-boot properties in SPL device tree
Add a test to check the management of the U-boot relocation properties
for device tree SPL generation (fdtgrep result) and platdata:
- 'dm-pre-proper' and 'dm-tpl' not included in SPL
- 'dm-pre-reloc' and 'dm-spl' included in SPL
Signed-off-by: Patrick Delaunay <patrick.delaunay@st.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Thierry Reding [Mon, 20 May 2019 16:05:04 +0000 (18:05 +0200)]
fdtdec: test: Fix memory leak
Free the memory allocated to store the test FDT upon test completion to
avoid leaking the memory. We don't bother cleaning up on test failure
since the code is broken in that case and should be fixed, in which case
the leak would also go away.
Reported-by: Tom Rini <tom.rini@gmail.com> Suggested-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Signed-off-by: Thierry Reding <treding@nvidia.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Vabhav Sharma [Thu, 31 Jan 2019 12:08:10 +0000 (12:08 +0000)]
drivers: serial: lpuart: Enable Little Endian Support
By default LPUART driver with compatible string "fsl,ls1021a-lpuart"
support big-endian mode.On NXP SoC like LS1028A LPUART IP is
little-endian,Added support to Fetch LPUART IP Endianness from lpuart
device-tree node.
Signed-off-by: Vabhav Sharma <vabhav.sharma@nxp.com> Reviewed-by: Simon Glass <sjg@chromium.org>
Simon Glass [Sat, 18 May 2019 17:59:54 +0000 (11:59 -0600)]
spi: Avoid using malloc() in a critical function
In general we should avoid calling malloc() and free() repeatedly in
U-Boot lest we turn it into tianocore. In SPL this can make SPI flash
unusable since free() is often a nop and allocation space is limited.
In any case, these seems no need for malloc() since the number of bytes
is very small, perhaps less than 8.
Signed-off-by: Simon Glass <sjg@chromium.org> Fixes: d13f5b254a (spi: Extend the core to ease integration of SPI
memory controllers)
Simon Glass [Sat, 18 May 2019 17:59:53 +0000 (11:59 -0600)]
blk: Allow control of the block cache in TPL
Generally block devices are not enabled in TPL, but in case they are,
add a Kconfig option for the block cache. This allows the setting (default
off) to be found with CONFIG_IS_ENABLED().
Simon Glass [Sat, 18 May 2019 17:59:47 +0000 (11:59 -0600)]
sandbox: Allo sdl-config to be overridden
When cross-compiling, sometimes sdl-config must come from a different path
from the default. Add a way to override it, by adding SDL_CONFIG to the
environment before building U-Boot.
Simon Glass [Sat, 18 May 2019 17:59:45 +0000 (11:59 -0600)]
sandbox: spl: Lower priority of standard loader
We normally want to load U-Boot from SPL, but if a board wants to do
something else, it is currently not possible since the standard loader
has the top priority. Lower it to allow other SPL_LOAD_IMAGE_METHOD()
declarations to override it.
Simon Glass [Sat, 18 May 2019 17:59:44 +0000 (11:59 -0600)]
sandbox: Quit when hang() is called
It doesn't make a lot of sense to hang on sandbox when hang() is called,
since the only way out is Ctrl-C. In fact, Ctrl-C does not work if the
terminal is in raw mode, which it will be if the command-line has not been
reached yet. In that case, Ctrl-Z / kill -9 must be used, which is not
very friendly.
Avoid all of this by quiting when hang() is called.
Simon Glass [Sat, 18 May 2019 16:07:17 +0000 (10:07 -0600)]
Add a simple script to remove boards
This script attempts to create a git commit which removes a single board.
It is quite fallible and everything it does needs checking. But it can
help speed up the process.
Signed-off-by: Simon Glass <sjg@chromium.org> Reviewed-by: Chris Packham <judge.packham@gmail.com>
Simon Glass [Sat, 18 May 2019 04:00:41 +0000 (22:00 -0600)]
dtoc: Add a unit test for BytesToValue()
Add a simple unit test for one of the cases of this function, so that any
fault can be seen directly, rather than appearing through the failure of
another test.
Simon Glass [Sat, 18 May 2019 04:00:38 +0000 (22:00 -0600)]
dtoc: Convert the Fdt.Node class to Python 3
Update this class to work correctly on Python 3 and to pass its unit
tests. The only required change is to deal with a difference in the
behaviour of sorting with a None value.
Simon Glass [Sat, 18 May 2019 04:00:35 +0000 (22:00 -0600)]
dtoc: Updates BytesToValue() for Python 3
The difference between the bytes and str types in Python 3 requires a
number of minor changes to this function. Update it to handle the input
data using the 'bytes' type. Create two useful helper functions which can
be used by other modules too.
Simon Glass [Sat, 18 May 2019 04:00:32 +0000 (22:00 -0600)]
dtoc: Sort platdata output from dtoc
At present the order of struct field emitted by this tool depends on the
internal workings of a Python dictionary. Sort the fields to remove this
uncertainty, so that tests are deterministic.
Simon Glass [Tue, 14 May 2019 21:53:53 +0000 (15:53 -0600)]
patman: Adjust functional tests for Python 3
Change the code so that it works on both Python 2 and Python 3. This works
by using unicode instead of latin1 for the test input, and ensuring that
the output is converted to a string rather than a unicode object on
Python 2.
Simon Glass [Tue, 14 May 2019 21:53:52 +0000 (15:53 -0600)]
patman: Avoid unicode type in settings unit tests
The unicode type does not exist in Python 3 and when displaying strings
they do not have the 'u' prefix. Adjusts the settings unit tests to deal
with this difference, by converting the comparison value to a string, thus
dropping the 'u'.
Simon Glass [Tue, 14 May 2019 21:53:51 +0000 (15:53 -0600)]
patman: Sort series output for repeatabily
We use sets to produce the list of To and Cc lines for a series. This does
not result in stable ordering of the recipients. Sort each list to ensure
that the output is repeatable. This is necessary for tests.
Simon Glass [Tue, 14 May 2019 21:53:50 +0000 (15:53 -0600)]
patman: Move unicode helpers to tools
Create helper functions in the tools module to deal with the differences
between unicode in Python 2 (where we use the 'unicode' type) and Python 3
(where we use the 'str' type).
Simon Glass [Tue, 14 May 2019 21:53:47 +0000 (15:53 -0600)]
binman: Handle repeated bytes for Python 3
The method of multiplying a character by a number works well for creating
a repeated string in Python 2. But in Python 3 we need to use bytes()
instead, to avoid unicode problems, since 'bytes' is no-longer just an
alias of 'str'.
Create a function to handle this detail and call it from the relevant
places in binman.
Simon Glass [Tue, 14 May 2019 21:53:45 +0000 (15:53 -0600)]
binman: Use binary mode when compressing data
The lz4 utility inserts binary data in its output which cannot always be
converted to unicode (nor should it be). Fix this by using the new binary
mode for program output.
Simon Glass [Tue, 14 May 2019 21:53:44 +0000 (15:53 -0600)]
patman: Provide a way to get program output in binary mode
At present cros_subprocess and the tools library use a string to obtain
stdout from a program. This works fine on Python 2. With Python 3 we end
up with unicode errors in some cases. Fix this by providing a binary mode,
which returns the data as bytes() instead of a string.
Simon Glass [Tue, 14 May 2019 21:53:42 +0000 (15:53 -0600)]
tools: binman: Open all binary files in binary mode
At present some files are opened in text mode despite containing binary
data. This works on Python 2 but not always on Python 3, due to unicode
problems. BC&D are not my favourite people. Adjust the affected open()
statements to use binary mode.
Simon Glass [Tue, 14 May 2019 21:53:38 +0000 (15:53 -0600)]
binman: Don't show errors for failed tests
An unfortunate new feature in Python 3.5 causes binman to print errors for
non-existent tests, when the test filter is used. Work around this by
detecting the unwanted tests and removing them from the result.
Simon Glass [Sat, 11 May 2019 18:46:39 +0000 (12:46 -0600)]
patman: Update cros_subprocess to use bytes
At present this function uses lists and strings. This does not work so
well with Python 3, and testing against '' does not work for a bytearray.
Update the code to fix these issues.
Chuanhua Han [Mon, 8 Jul 2019 03:45:25 +0000 (11:45 +0800)]
rtc: pcf2127: Fixed bug with rtc settings and getting error time
The previous pcf2127 RTC chip could not read and set the correct time.
When reading the data of internal registers, the read address was the
value of register plus 1. This is because this chip requires the host
to send a stop signal after setting the register address and before
reading the register data.
This patch sets the register address using dm_i2c_write and reads
the register data using the original dm_i2c_xfer in order to generate
a stop signal after the register address is set, and fixes the bug of
the original read and write time.
Signed-off-by: Biwen Li <biwen.li@nxp.com> Signed-off-by: Chuanhua Han <chuanhua.han@nxp.com> Reviewed-by: Lukasz Majewski <lukma@denx.de> Reviewed-by: Heiko Schocher <hs@denx.de>
Cyrille Pitchen [Tue, 18 Jun 2019 08:52:01 +0000 (08:52 +0000)]
ARM: at91: sama5: add common environment for QSPI
Use the same memory layout as we use for the NAND boot on the other boards.
QSPI flashes are present on the following boards:
sama5d2_xplained RevB: 32 Mbyte flash (mx25l3273fm2i-08g)
sama5d2_xplained RevC: 8 Mbyte flash (sst26vf064b-104i/sn)
sama5d27_som1_ek: 8 Mbyte flash (sst26vf064b-104i/sn)
sama5d2_ptc_ek: 8 Mbyte flash (sst26vf064b-104i/sn)
The 8 Mbyte limit is enough to cope with the memory layout used in the NAND
boot. rootfs exceeds the 8 Mbyte limit and will stay in eMMC in the
sama5d2_xplained case. The final scope is to use a single memory layout for
all boot medias.
Cyrille Pitchen [Tue, 18 Jun 2019 08:51:58 +0000 (08:51 +0000)]
ARM: dts: at91: sama5d2_xplained: fix QSPI0 node
Fix the following:
- use "jedec,spi-nor" binding, we use jedec compatible flashes
- set bus width to 4, we use quad capable flashes
- differentiate bewteen data and clk and cs pins
- drop partions as we don't use them in u-boot.
Melin Tomas [Fri, 28 Jun 2019 12:08:40 +0000 (12:08 +0000)]
xilinx_xiic: Fix transfer initialisation
Prior to starting a new transfer, conditionally wait for bus to not
be busy.
Reinitialise controller as otherwise operation is not stable.
For reference, see linux kernel
commit 9656eeebf3f1 ("i2c: Revert i2c: xiic: Do not reset controller before every transfer")
hs: Fixed DOS line endings
added missing '\n'
Fixed git commit description style
Signed-off-by: Tomas Melin <tomas.melin@vaisala.com>
Jun Chen [Wed, 5 Jun 2019 07:23:16 +0000 (15:23 +0800)]
i2c: designware_i2c: Restore enable state after set speed
Before calling __dw_i2c_set_bus_speed(),
the I2C could already be set as ether enable or disable,
we should restore the original setting instead of enable i2c anyway.
This patch fix a bug happened in init function:
__dw_i2c_init(){
/* Disable i2c */
...
__dw_i2c_set_bus_speed(i2c_base, NULL, speed);
writel(slaveaddr, &i2c_base->ic_sar);
/* Enable i2c */
}
In this case, enable i2c inside __dw_i2c_set_bus_speed() function
will cause ic_sar write fail.
Hamish Guthrie [Wed, 15 May 2019 13:15:59 +0000 (15:15 +0200)]
ubispl: add support for loading volumes by name
The motivation is to use the UBI atomic volume rename functionality to
allow double copy software updates on UBI. To that end the SPL is
configured to always load the same volume name (e.g. "u-boot"),
whereas a software updater always installs into the secondary volume
"u-boot_r". After successful installation, these two volume names are
switched.
This extension is protected by #ifdefs as it will somewhat slow down
loading of volumes by id. This is because the code needs to disable
the optimization of ignoring all volume ids which are not
to-be-loaded, since these can only be resolved after attaching.
This adds two vtbl related functions from Linux, which are taken from
the same kernel version as the current main U-Boot UBI code (Linux 4.2 64291f7db5bd8).
Signed-off-by: Hamish Guthrie <hamish.guthrie@kistler.com> Signed-off-by: Markus Klotzbuecher <markus.klotzbuecher@kistler.com> Reviewed-by: Heiko Schocher <hs@denx.de> Cc: Kyungmin Park <kmpark@infradead.org>
Migrate the ubispl configuration for the omap3_igep00x0 and
am335x_igep003x boards to KConfig. Both boards were built with
SOURCE_DATE_EPOCH=0 and found to be equal before and after.
Signed-off-by: Markus Klotzbuecher <markus.klotzbuecher@kistler.com> Cc: Heiko Schocher <hs@denx.de> Cc: Kyungmin Park <kmpark@infradead.org> Cc: Javier MartÃnez Canillas <javier@dowhile0.org> Cc: Enric Balletbo i Serra <eballetbo@gmail.com> Acked-by: Enric Balletbo i Serra <eballetbo@gmail.com>
at91, omap2plus: configs: migrate CONFIG_ENV_ to defconfigs
Enable the extended ENV options for AT91 and OMAP2PLUS in order to be
able to use CONFIG_ENV_UBI_* on these architectures.
As this change also makes the configs ENV_SIZE, ENV_SECT_SIZE,
ENV_OFFSET visible to AT91 and OMAP2PLUS, migrate users of these to
KConfig.
This migration was run using an extended moveconfig.py which evaluates
expressions such as "(512 << 10)". See patch ("moveconfig: expand
simple expressions").
All modified boards were built with SOURCE_DATE_EPOCH=0 before and
after the change and successfully confirmed that the identical binary
is generated (the only exception was igep00x0, which does not define
CONFIG_ENV_IS_IN_UBI in the original board header. Once that is
defined, the test passes too).
hs: rebased patch to: 68b90e57bc: "configs: tinker-rk3288 disable CONFIG_SPL_I2C_SUPPORT"
Signed-off-by: Markus Klotzbuecher <markus.klotzbuecher@kistler.com> Cc: Heiko Schocher <hs@denx.de> Cc: Eugen Hristev <eugen.hristev@microchip.com> Cc: Tom Rini <trini@konsulko.com>
Introduce the KConfig option CONFIG_ENV_UBI_VOLUME_REDUND for defining
the name of the UBI volume used to store the redundant environment.
Signed-off-by: Markus Klotzbuecher <markus.klotzbuecher@kistler.com> Reviewed-by: Heiko Schocher <hs@denx.de> Cc: Kyungmin Park <kmpark@infradead.org>
hs: get rid of stm32mp1* build errors
Tom Rini [Mon, 8 Jul 2019 20:02:51 +0000 (16:02 -0400)]
Merge tag 'efi-2019-10-rc1' of https://gitlab.denx.de/u-boot/custodians/u-boot-efi
Pull request for UEFI sub-system for v2019.10-rc1
Fix a possible overflow for GUID partition tables.
For some runtime services we only have implementations valid at boottime.
So we replace them when leaving boottime. Move this from
SetVirtualAddressMap() to ExitBootServices() as SetVirtualAddressMap() is
not called by all operating systems. Adjust the Python tests accordingly.
Bump the supported UEFI specification version to 2.8.
projects / u-boot.git / log
