When performing a cache disable function, code must not access DRAM.
That is because when the cache is disabled, it will be bypassed and all
loads and stores will be serviced by RAM. This prevents accessing any
dirty data in the cache. In turn, this means the stack cannot be
used, since that is in RAM. To guarantee that code doesn't use RAM (and
in particular the stack) __asm_flush_l3_cache() must be manually
implemented in assembly, rather than implemented in C since the compiler
won't know not to touch RAM.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
nvtboot_boot_x0 is a 64-bit variable and hence must be 64-bit aligned.
So far this has happened by accident! Fix the code so this is guaranteed.
This fixes the following build error:
... relocation truncated to fit: R_AARCH64_LDST64_ABS_LO12_NC
against symbol `nvtboot_boot_x0' ...
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Signed-off-by: Tom Warren <twarren@nvidia.com>
Use random ethernet address if the ethernet address found
is invalid, not zero and config for random address
is defined.
Signed-off-by: Siva Durga Prasad Paladugu <sivadur@xilinx.com>
Signed-off-by: Michal Simek <michal.simek@xilinx.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Commit 5a49f17481 ("net: mii: Use spatch to update miiphy_register")
updated the mvgbe implementation of smi_reg_read/smi_reg_write. Prior to
that change mvgbe_phy_read and mvgbe_phy_write where used as wrappers to
satisfy the phylib APIs. Because these functions weren't updated in that
commit build errors where triggered when CONFIG_PHYLIB was enabled.
Fix these build errors by removing mvgbe_phy_read and mvgbe_phy_write
and using smi_reg_read/smi_reg_write directly.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Like [1], reset the FLP burst timing for the KSZ9031 to the 16ms
specified by the IEEE802.3 standard from the chip's default of 8ms.
For more details, see the "Auto-Negotiation Timing" section of the
KSZ9031RNX datasheet.
[1] https://patchwork.kernel.org/patch/6558371/
Signed-off-by: Ash Charles <ash.charles@savoirfairelinux.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Enable the Ethernet device in DT, provide board-specific configuration,
and enable the driver in Kconfig.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Tegra186 includes a Synopsys DWC EQoS (Ethernet) device. Add this to the
Tegra186 SoC DT so that boards can make use of it.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
On Tegra186, the bootloader which runs before U-Boot passes the Ethernet
MAC address to U-Boot using device tree. Extract this value and write it
to the environment, so that the Ethernet uclass picks it up and uses it
for the built-in Ethernet device.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Extend the Tegra186 implementation of board_late_init() to call a per-SoC
"hook" function. This will allow SoC-specific (rather than Tegra-wide)
functionality to be implemented without the core Tegra code needing to be
aware of the details. While board186.c is currently only used for
Tegra186, it should be applicable to any other future SoC, and perhaps its
simple design could be back-ported to older SoCs in the future too.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
This driver supports the Synopsys Designware Ethernet QoS (Quality of
Service) a/k/a eqos IP block, which is a different design than the HW
supported by the existing designware.c driver. The IP supports many
options for bus type, clocking/reset structure, and feature list. This
driver currently supports the specific configuration used in NVIDIA's
Tegra186 chip, but should be extensible to other combinations quite
easily, as explained in the source.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org> # V1
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
The Synopsys DWC EQoS is a configurable Ethernet MAC/DMA IP block which
supports multiple options for bus type, clocking and reset structure, and
feature list.
This patch imports the binding from the Linux kernel, including my V3
patch to extend the binding to cover the Tegra186, which is applied for
next-20160912. So far, my changes have been acked by Lars Persson, the
original author of the binding.
Signed-off-by: Stephen Warren <swarren@nvidia.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Since commit ce412b79e7 ("drivers: net: phy: atheros: add separate
config for AR8031") ethernet does not work on mx6sabresd.
This commit correctly assigns ar8031_config() as the configuration
function for AR8031 in the same way as done in the Linux kernel.
However, on wandboard design we need some additional configuration,
such as enabling the 125 MHz AR8031 output that needs to be done
in the board file.
This also aligns with the same method that the kernel performs
the AR8031 fixup in arch/arm/mach-imx/mach-imx6q.c.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Acked-by: Joe Hershberger <joe.hershberger@ni.com>
Adds -i option that allows specifying a ramdisk file to be added to the
FIT image when we are using the automatic FIT mode (no ITS file).
This makes adding Depthcharge support to LAVA much more convenient, as
no additional configuration files need to be kept around in the machine
that dispatches jobs to the boards.
Signed-off-by: Tomeu Vizoso <tomeu.vizoso@collabora.com>
Cc: Simon Glass <sjg@chromium.org>
Cc: Matt Hart <matthew.hart@linaro.org>
Cc: Neil Williams <codehelp@debian.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
The way that we have things broken down currently allows for some
combinations of vendor or CPU to not be built. To fix this, create a
new catch-all job that excludes everything we've built elsewhere. For
the sake of simplicity we are allowing for the possibility of some
overlap between the vendor-based jobs and the CPU-based jobs. While
we're in here, make a failed build provide the summary of failure.
Signed-off-by: Tom Rini <trini@konsulko.com>
The README for buildman says that we can use any field in boards.cfg to
decide what to build. However, we were not saving the options field
correctly.
Cc: Simon Glass <sjg@chromium.org>
Signed-off-by: Tom Rini <trini@konsulko.com>
Acked-by: Simon Glass <sjg@chromium.org>
Since commit ce412b79e7 ("drivers: net: phy: atheros: add separate
config for AR8031") ethernet does not work on mx6sabresd.
This commit correctly assigns ar8031_config() as the configuration
function for AR8031 in the same way as done in the Linux kernel.
However, on mx6sabresd design we need some additional configuration,
such as enabling the 125 MHz AR8031 output that needs to be done
in the board file.
This also aligns with the same method that the kernel performs
the AR8031 fixup in arch/arm/mach-imx/mach-imx6q.c.
Signed-off-by: Fabio Estevam <fabio.estevam@nxp.com>
Config file names on MAINTAINERS and README in
board/engicam/icorem6 seems to be wrong, hence fixed the same.
Cc: Stefano Babic <sbabic@denx.de>
Cc: Michael Trimarchi <michael@amarulasolutions.com>
Signed-off-by: Jagan Teki <jagan@amarulasolutions.com>
For 64-bit ARM systems we provide just a timer_read_counter()
implementation and rely on the generic non-uclass get_ticks() function
in lib/time.c to call the former.
However this function is actually not 64-bit safe, as it assumes a
"long" to be 32-bit. Beside the fact that the resulting uint64_t
isn't bigger than "long" on 64-bit architectures and thus combining two
counters makes no sense, we get all kind of weird results when we try
to OR in the high value shifted by _32_ bits.
So let's avoid that function at all and provide a straight forward
get_ticks() implementation for ARMv8, which also is in line with ARMv7.
This fixes occasional immediate time-out expiration issues I see on the
Pine64 board. The root cause of this needs to be investigated, but this
fix looks like the right thing anyway.
Signed-off-by: Andre Przywara <andre.przywara@arm.com>
Check return value when open the plugin file.
Coverity report:
** CID 153926: Error handling issues (NEGATIVE_RETURNS)
/tools/imximage.c: 542 in copy_plugin_code()
ifd = open(plugin_file, O_RDONLY|O_BINARY);
>>> CID 153926: Error handling issues (NEGATIVE_RETURNS)
>>> "ifd" is passed to a parameter that cannot be negative.
Signed-off-by: Peng Fan <peng.fan@nxp.com>
Cc: Stefano Babic <sbabic@denx.de>
Cc: Tom Rini <trini@konsulko.com>
Reported-by: Coverity (CID: 153926)
Reviewed-by: Tom Rini <trini@konsulko.com>
Coverity complains that this can overflow. If we later increase the size
of one of the strings in the table, it could happen.
Adjust the code to protect against this.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Coverity (CID: 150964)
As the upcoming driver model integration takes up some more precious flash
space first make sure to drop expensive LZMA support.
Signed-off-by: Marcel Ziswiler <marcel.ziswiler@toradex.com>
- Add in system aarch64-linux-gnu toolchain
- Now that all VMs will have aarch64 available, don't exclude them from
other jobs but instead exclude them from the catch-all aarch64 build
- Add JOB= to the Freescale/ARM build to be clear about what it does.
- Add uniphier as a stand-alone job
Signed-off-by: Tom Rini <trini@konsulko.com>
The allocated memory should be freed. Fix it.
Signed-off-by: Simon Glass <sjg@chromium.org>
Reported-by: Coverity (CID: 150963)
Reviewed-by: Tom Rini <trini@konsulko.com>
Prevent cache misalignment message by ensuring that a whole cache line
is flushed.
Signed-off-by: Chris Packham <judge.packham@gmail.com>
Reviewed-by: Stefan Roese <sr@denx.de>
Cover-Letter: Fixes several spelling errors for the words "resetting",
"extended", "occur", and "multiple".
Signed-off-by: Vagrant Cascadian <vagrant@debian.org>
Reviewed-by: Simon Glass <sjg@chromium.org>
The autodetect feature doesn't allow users to specify the device tree.
This fix will make it only autodetect if 'fdtimage' is not defined.
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Not all networks have a DHCP server configured properly, so these
scripts make it easier to boot in that scenario.
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
The LCD isn't supported in U-Boot and the LCD is now configured in
the device tree, so this code is pointless.
V2: Eliminiate erroneous newline.
Signed-off-by: Adam Ford <aford173@gmail.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
Current logic for query of revision, board_name, config returns
NULL. Users of these functions do a direct strncmp to compare.
Unfortunately, as per conventions require two valid strings to compare
against and the current implementation causes a crash when compared
with NULL.
We'd still like to maintain the simplistic usage of these APIs instead
of redundant if (string) res=strncmp(fn(),"cmp",n); flowing all over
the place.
Hence, since the version, name and config is already pre-initialized
with empty string, just dont check for invalid header in the first
place and return the empty string to the caller.
Reported-by: Brad Griffis <bgriffis@ti.com>
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
[trini: Correct was'nt -> wasn't typo]
Signed-off-by: Tom Rini <trini@konsulko.com>
config should have been initialized along with others as defaults.
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
We should have used TI_DEAD_EEPROM_MAGIC in the first place.
Fixes: d3b98a9eb9 ("ti: common: dra7: Add standard access for board description EEPROM")
Signed-off-by: Nishanth Menon <nm@ti.com>
Reviewed-by: Lokesh Vutla <lokeshvutla@ti.com>
Reviewed-by: Tom Rini <trini@konsulko.com>
SPL_MMC_SUPPORT defined in rockchip top level Kconfig instead of
inside rk3288 and default to disable if ROCKCHIP_SPL_BACK_TO_BROM
defined.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Reviewed-by: Simon Glass <sjg@chromium.org>
Unify the partitions of each chip then it will be more easy for us to
write scripts, tools or guides for rockchip chips.
Those extra partitions mostly are used to be compatible with our
internal loaders (such as miniloader which was same as spl, or
android loader then we can support dual boot)
Signed-off-by: Jacob Chen <jacob2.chen@rock-chips.com>
Reviewed-by: Kever Yang <kever.yang@rock-chips.com>
Since we implement the dram capacity auto detect, we don't
need to set the channel number and sdram-channel in dts.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Acked-by: Simon Glass <sjg@chromium.org>
Tested-by: Simon Glass <sjg@chromium.org>
Tested-by: Vagrant Cascadian <vagrant@debian.org>
Add support for rk3288 dram capacity auto detect, support DDR3 and
LPDDR3, DDR2 is not supported.
The program will automatically detect:
- channel number
- rank number
- column address number
- row address number
The dts file do not need to describe those info after apply this patch.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
Tested-by: Simon Glass <sjg@chromium.org>
Tested-by: Vagrant Cascadian <vagrant@debian.org>
Tested-by: Vagrant Cascadian <vagrant@debian.org>
Enable ROCKCHIP_SPL_BACK_TO_BROM and disable CONFIG_SPL_MMC_SUPPORT
to save memory in order to enable add source code for dram capacity
auto-detect.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
rk3399 has two mmc dev controller:
mmc 0: SD card;
mmc 1: EMMC
U-Boot will scan the mmc boot device configured by CONFIG_SYS_MMC_ENV_DEV,
since evb has emmc on board, let's set the EMMC as default.
Signed-off-by: Kever Yang <kever.yang@rock-chips.com>
