Andre Przywara [Sat, 3 Mar 2018 23:49:02 +0000 (23:49 +0000)]
sunxi: move CONFIG_SYS_TEXT_BASE out of defconfigs
Commit 278b90ce786f ("configs: Migrate CONFIG_SYS_TEXT_BASE") made
CONFIG_SYS_TEXT_BASE a proper Kconfig variable, with the consequence
of moving the common definition shared by almost every sunxi board
into 123 individual defconfig files. But the U-Boot start address
for Allwinner boards is a platform decision which has been around for
ages, so defining it in each *board* config file seems a bit over the
top.
Define the standard values (160MB into DRAM for most SoCs, with two
SoC exceptions) if ARCH_SUNXI is selected, and delete the lines from
the individual defconfigs.
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Acked-by: Maxime Ripard <maxime.ripard@bootlin.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
A20-SOM204 board has option with onboard 16GB eMMC. The chip is wired
to MMC2 slot.
This patch adds defconfig and dts files for this board. The dts is same
with mainline kernel.
Signed-off-by: Stefan Mavrodiev <stefan@olimex.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
This is new System-On-Module platform with universal dimm socket for
easy insertation. The EVB board is designed to be universal with
future modules.
Base features of A20-SOM204 board includes:
* 1GB DDR3 RAM
* AXP209 PMU
* KSZ9031 Gigabit PHY
* AT24C16 EEPROM
* Status LED
* LCD connector
* GPIO connector
There will be variants with the following options:
* Second LAN8710A Megabit PHY
* 16MB SPI Flash memory
* eMMC card
* ATECC508 crypto device
The EVB board has:
* Debug UART
* MicroSD card connector
* USB-OTG connector
* Two USB host
* RTL8723BS WiFi/BT combo
* IrDA transceiver/receiver
* HDMI connector
* VGA connector
* Megabit ethernet transceiver
* Gigabit ethernet transceiver
* SATA connector
* CAN driver
* CSI camera
* MIC and HP connectors
* PCIe x4 connector
* USB3 connector
* Two UEXT connectors
* Two user LEDs
Some of the features are multiplexed and cannot be used the same time:
CAN and Megabit PHY. Others are not usable with A20 SoC: PCIe and USB3.
This patch adds defconfig and dts files for this board. The dts is same
with mainline kernel, except some nodes are removed to make file
compatible with existing dtsi file.
Signed-off-by: Stefan Mavrodiev <stefan@olimex.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
CONFIG_PHY_ADDR is used for old-style configuration. This makes
impossible changing the PHY address, if multiple boards share a same
config header file (for example include/configs/sunxi-common.h).
Moving this to Kconfig helps overcoming this issue. It's defined
as entry inside PHYLIB section.
After the implemention, moveconfig was run. The issues are:
- edb9315a - CONFIG_PHYLIB is not enabled. Entry is
deleted.
- ds414 - CONFIG_PHYLIB is in incompatible format:
{ 0x1, 0x0 }. This entry is also deleted.
- devkit3250 - The PHY_ADDR is in hex format (0x1F).
Manually CONFIG_PHY_ADDR=31 is added in
the defconfig.
After the changes the suspicious defconfigs passes building.
Signed-off-by: Stefan Mavrodiev <stefan@olimex.com> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com>
[jagan: rebased on master] Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Samuel Holland [Sun, 28 Jan 2018 05:53:20 +0000 (23:53 -0600)]
net: sun8i_emac: Fix PHY initialization
The previous code tried to update the PHY parameters without waiting for
autonegotiation to complete. This caused wrong values to be written to
the EMAC in sun8i_adjust_link(). As a result, any commands that called
eth_start() before autonegotiation completed would find the network
nonfunctional. Fix this by using the correct function to start up the
PHY.
Signed-off-by: Samuel Holland <samuel@sholland.org> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
ARM: sunxi: Build u-boot-sunxi-with-spl.bin on ARM64 as well
In README.sunxi64 we tell the user how to optionally create
u-boot-sunxi-with-spl.bin by manually running cat. Instead, have the
build system create the file automatically just like it does for 32-bit
sunxi boards.
Signed-off-by: Tuomas Tynkkynen <tuomas@tuxera.com> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Acked-by: Maxime Ripard <maxime.ripard@bootlin.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
ARM: sunxi: Move u-boot-sunxi-with-spl.bin rule to Makefile
We're going to need this logic for 64-bit builds as well, so move it
out from under arch/arm/cpu/armv7.
Signed-off-by: Tuomas Tynkkynen <tuomas@tuxera.com> Reviewed-by: Andre Przywara <andre.przywara@arm.com> Acked-by: Maxime Ripard <maxime.ripard@bootlin.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
With CONFIG_DM_MMC the BananaPi does not detect SD cards.
The sunxi device trees use the cd-inverted property to indicate that
the card detect is inverted.
This property is documented in Linux kernel devicetree/bindings/mmc/mmc.txt
The property is not marked as deprecated.
A similar patch was posted by Tuomas but is in status "Changes Requested".
https://patchwork.ozlabs.org/patch/850377/
This patch is a stripped down version of his patch.
Suggested-by: Tuomas Tynkkynen <tuomas@tuxera.com> Signed-off-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> Reviewed-by: Jagan Teki <jagan@openedev.com>
Chen-Yu Tsai [Mon, 5 Feb 2018 14:03:25 +0000 (22:03 +0800)]
sunxi: Add reg property for USB OTG node in sun8i-a83t.dtsi
When the OTG node was added, its reg property for its address space was
missing. With commit f4f9896ac310 ("musb: sunxi: Use base address from
device tree"), the OTG controller's address is derived from the device
tree exclusively. The missing property results in U-boot crashing when
MUSB is initialized.
Fixes: b0bea6678981 ("sunxi: Add USB and R_PIO nodes to sun8i-a83t.dtsi") Signed-off-by: Chen-Yu Tsai <wens@csie.org> Acked-by: Maxime Ripard <maxime.ripard@bootlin.com> Reviewed-by: Jagan Teki <jagan@openedev.com> Tested-by: Jagan Teki <jagan@openedev.com>
Andre Przywara [Sat, 27 Jan 2018 00:52:22 +0000 (00:52 +0000)]
sunxi: Pine64: temporarily remove extra Pine64 non-plus DT
With the merge of the new u-boot.itb size check now the build for
pine64_plus_defconfig breaks, as this file gets too large:
=============
u-boot.itb exceeds file size limit:
limit: 516096 bytes
actual: 521852 bytes
excess: 5756 bytes
make: *** [u-boot.itb] Error 1
=============
One easily fixable reason is that we actually have two .dtbs in our FIT
image, one for the regular Pine64+ board, and one for the non-plus version.
The only difference U-Boot cares about is the 100Mbit Ethernet PHY used
on the non-plus version, however Ethernet isn't enabled in the non-plus
DT anyway.
So we could avoid the non-plus special handling, and remove that extra
.dtb from the FIT image, which saves a few KBs and brings us back below the
limit. The Pine64 would boot with a Pine64+ .dtb, and would fail to
enable Ethernet. Given that it didn't work in the first place, this is not
a regression.
Once we switch to a non-MMC environment, we can bring this back, then
with a proper .dtb and hopefully working Ethernet.
Signed-off-by: Andre Przywara <andre.przywara@arm.com> Signed-off-by: Maxime Ripard <maxime.ripard@bootlin.com>
commit 8c4037a09a5c ("imx: hab: Ensure the IVT DCD pointer is Null prior
to calling HAB authenticate function.") makes the DCD field being NULL a
dependency.
This change though will break loading and executing of existing pre-signed
binaries on a u-boot update i.e. if this change is deployed on a board you
will be forced to redo all images on that board to NULL out the DCD.
There is no prior guidance from NXP that the DCD must be NULL similarly
public guidance on usage of the HAB doesn't call out this NULL dependency
(see boundary devices link).
Since later SoCs will reject a non-NULL DCD there's no reason to make a
NULL DCD a requirement, however if there is an actual dependency for later
SoCs the appropriate fix would be to do SoC version checking.
Earlier SoCs are capable (and happy) to authenticate images with non-NULL
DCDs, we should not be forcing this change on downstream users -
particularly if it means those users now must rewrite their build systems
and/or redeploy signed images in the field.
Fixes: 8c4037a09a5c ("imx: hab: Ensure the IVT DCD pointer is Null prior
to calling HAB authenticate function.")
Breno Lima [Thu, 22 Feb 2018 00:42:56 +0000 (00:42 +0000)]
doc: mxc_hab: Update i.MX HAB documentation
The README.mxc_hab is outdated and need improvements, add the following
modifications:
- Reorganize document and remove duplicate content
- Add CST download link
- Update CST package name
- Align command lines with CST v2.3.3
- Update U-Boot binary name
- Remove CSF padding since is not documented in AN4581
commit ed286bc80e9d ("imx: hab: Check if CSF is valid before authenticating
image") makes use of "__packed" as a prefix to the "struct hab_hdr"
declaration.
With my compiler "gcc version 7.2.1 20171011 (Linaro GCC 7.2-2017.11)" we
get:
./arch/arm/include/asm/mach-imx/hab.h:42:25: error: expected ‘=’, ‘,’, ‘;’,
‘asm’ or ‘__attribute__’ before ‘{’ token
struct __packed hab_hdr {
Jagan Teki [Tue, 6 Mar 2018 11:45:59 +0000 (17:15 +0530)]
ARM: dts: imx6dl-icore-rqs: Fix to include correct dtsi
This patch fixes the wrongly included dtsi file which was
breaking mainline support for Engicam i.CoreM6 DualLite/Solo RQS.
Linux commit details for the same change as
"ARM: dts: imx6dl: Include correct dtsi file for Engicam i.CoreM6
DualLite/Solo RQS"
(sha1: c0c6bb2322964bd264b4ddedaa5776f40c709f0c)
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com> Reviewed-by: Fabio Estevam <fabio.estevam@nxp.com>
Stefan Theil [Thu, 8 Mar 2018 08:00:13 +0000 (09:00 +0100)]
tools/mkimage: Use proper output parameter in dtc-system call
The system call used by mkimage to run dtc redirects stdout to a
temporary file. This can cause problems on Windows (with a MinGW
cross-compiled version). Using the "-o" dtc parameter avoids
this problem.
Signed-off-by: Stefan Theil <stefan.theil@mixed-mode.de> Reviewed-by: Tom Rini <trini@konsulko.com>
CONFIG_SYS_CBSIZE determines the maximum length of the kernel command
line, and the default value of 256 is too small for booting some Linux
images in the wild.
Signed-off-by: Tuomas Tynkkynen <tuomas.tynkkynen@iki.fi>
Yasushi SHOJI [Thu, 8 Mar 2018 04:21:10 +0000 (13:21 +0900)]
imx: syscounter: make sure asm is volatile
Without the volatile attribute, compilers are entitled to optimize out
the same asm(). In the case of __udelay() in syscounter.c, it calls
`get_ticks()` twice, one for the starting time and the second in the
loop to check the current time. When compilers inline `get_ticks()`
they see the same `mrrc` instructions and optimize out the second one.
This leads to infinite loop since we don't get updated value from the
system counter.
Fabio Estevam [Fri, 9 Mar 2018 11:25:00 +0000 (08:25 -0300)]
imximage: Remove failure when no IVT offset is found
Sometimes imximage throws the following error:
CFGS board/freescale/vf610twr/imximage.cfg.cfgtmp
CFGS board/freescale/vf610twr/imximage.cfg.cfgtmp
MKIMAGE u-boot-dtb.imx
Error: No BOOT_FROM tag in board/freescale/vf610twr/imximage.cfg.cfgtmp
arch/arm/mach-imx/Makefile:100: recipe for target 'u-boot-dtb.imx' failed
Later on, when running mkimage for the u-boot.imx it will succeed in
finding the IVT offset.
Looks like some race condition happening during parallel build when
processing mkimage for u-boot-dtb.imx and u-boot.imx.
A proper fix still needs to be implemented, but as a workaround let's
remove the error when the IVT offset is not found.
It is useful to have such message, especially during bring-up phase,
but the build error that it causes is severe, so better avoid the
build error for now.
The error checking can be re-implemented later when we have a proper
fix.
Reported-by: Breno Lima <breno.lima@nxp.com> Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com> Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Masahiro Yamada [Sun, 4 Mar 2018 16:20:11 +0000 (01:20 +0900)]
libfdt: move headers to <linux/libfdt.h> and <linux/libfdt_env.h>
Thomas reported U-Boot failed to build host tools if libfdt-devel
package is installed because tools include libfdt headers from
/usr/include/ instead of using internal ones.
This commit moves the header code:
include/libfdt.h -> include/linux/libfdt.h
include/libfdt_env.h -> include/linux/libfdt_env.h
and replaces include directives:
#include <libfdt.h> -> #include <linux/libfdt.h>
#include <libfdt_env.h> -> #include <linux/libfdt_env.h>
Reported-by: Thomas Petazzoni <thomas.petazzoni@bootlin.com> Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
tools: Include U-Boot libfdt headers from their actual path
There are no headers for libfdt in lib/libfdt, as they are instead
located in scripts/dtc/libfdt. Specifying lib/libfdt for headers
inclusion in host tools results in using the system libfdt headers,
which is not what we want. Change this to the proper path.
Signed-off-by: Paul Kocialkowski <contact@paulk.fr>
Mario Six [Thu, 1 Mar 2018 13:45:04 +0000 (14:45 +0100)]
gpio: pca953x_gpio: Support label setting from DT
The PCA953x driver uses "gpio@%x_" as the GPIO bank name, where "%x" is
instantiated with the I2C address of the chip. While this works, it
becomes very confusing if a board has multiple PCAs with the same
address on different I2C busses, and it also becomes an issue when a
GPIO's value is to be set via the 'gpio' command, because this command
only ever sets the value of the first device it encounters, leaving the
other devices inaccessible to the command.
As to not break boards that rely on this naming scheme, we introduce a
new device tree string property "label" for the driver. If it exists, it
is used to build a bank name of the form "%s@%x_" (where %x is still
instantiated with the I2C address). If it does not exist, the legacy
labeling scheme is used.
Stefan Agner [Thu, 1 Mar 2018 13:06:32 +0000 (14:06 +0100)]
tools/env: allow equal sign as key value separation
Treat the first equal sign as a key/value separation too. This makes
the script files compatible with mkenvimage input file format. It
won't support variables with equal signs anymore, but this seems not
really like a loss.
Signed-off-by: Stefan Agner <stefan.agner@toradex.com>
Jeremy Boone [Mon, 12 Feb 2018 22:56:37 +0000 (17:56 -0500)]
Atmel TPM: Fix potential buffer overruns
Ensure that the Atmel TPM driver performs sufficient
validation of the length returned in the TPM response header.
This patch prevents memory corruption if the header contains a
length value that is larger than the destination buffer.
Signed-off-by: Jeremy Boone <jeremy.boone@nccgroup.trust>
Jeremy Boone [Mon, 12 Feb 2018 22:56:36 +0000 (17:56 -0500)]
Infineon TPM: Fix potential buffer overruns
Ensure that the Infineon I2C and SPI TPM driver performs adequate
validation of the length extracted from the TPM response header.
This patch prevents integer underflow when the length was too small,
which could lead to memory corruption.
Signed-off-by: Jeremy Boone <jeremy.boone@nccgroup.trust>
Tom Rini [Thu, 1 Mar 2018 17:41:21 +0000 (12:41 -0500)]
Merge tag 'xilinx-kconfig-for-v2018.03' of git://www.denx.de/git/u-boot-microblaze
Xilinx Kconfig changes for v2018.03
- Moving some macros to Kconfig
# gpg: Signature made Thu 01 Mar 2018 10:45:58 AM EST using DSA key ID 294A0C21
# gpg: Good signature from "Michal Simek <monstr@monstr.eu>"
# gpg: aka "Michal Simek (Xilinx) <michals@xilinx.com>"
# gpg: aka "Michal Simek (Xilinx) <michal.simek@xilinx.com>"
Eugeniy Paltsev [Tue, 6 Feb 2018 14:12:09 +0000 (17:12 +0300)]
NET: designware: fix clock enable
After commit ba1f966725223 ("net: designware: add clock support")
we got NET broken on axs101 and axs103 platforms.
Some clock don't support gating so their clock drivers don't
implement .enable/.disable callbacks. In such case clk_enable
returns -ENOSYS.
Also some clock drivers implement .enable/.disable callbacks not for all
clock IDs and return -ENOSYS (or -ENOTSUPP) for others.
If we have such clock in 'clocks' list of designware ethernet controller
node we fail to probe designware ethernet.
Fix it.
Signed-off-by: Eugeniy Paltsev <Eugeniy.Paltsev@synopsys.com> Reviewed-by: Patrice Chotard <patrice.chotard@st.com> Acked-by: Joe Hershberger <joe.hershberger@ni.com> Reviewed-by: Philipp Tomsich <philipp.tomsich@theobroma-systems.com>
Mario Six [Mon, 15 Jan 2018 10:08:25 +0000 (11:08 +0100)]
net: phy: marvell 88e151x: Fix handling of bare RGMII interface type
Commit 68e6ecadc551 ("net: phy: marvell 88e151x: Fix handling of RGMII
interface types") fixed the initialization of 88e151x phys, but made it
so that interfaces of type PHY_INTERFACE_MODE_RGMII had both RX and TX
delay bits cleared. The default (like in m88e1111s_config) is to have
both bits set.
Hence, this patch changes the behavior in the PHY_INTERFACE_MODE_RGMII
case so that both bits are set.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Mario Six [Mon, 15 Jan 2018 10:08:22 +0000 (11:08 +0100)]
net: tsec: Fix memory leak in error path
tsec_initialize allocates a private driver structure using malloc.
Should the memory allocation of this private structure fail, the
function execution is aborted with a return 0, but the previously
allocated device structure is never freed, hence leaked.
Free the device structure in the error case.
Signed-off-by: Mario Six <mario.six@gdsys.cc> Acked-by: Joe Hershberger <joe.hershberger@ni.com>
projects / u-boot / log